january/february 2002 volume 27 number

Transcription

1 january/february 2002 volume 27 number (ISSN )

2 JANUARY/FEBRUARY 2002 VOLUME 27 NUMBER Aim and Scope Operative Dentistry publishes articles that advance the practice of operative dentistry. The scope of the journal includes conservation and restoration of teeth; the scientific foundation of operative dental therapy; dental materials; dental education; and the social, political, and economic aspects of dental practice. Review papers, book reviews, letters, and classified ads for faculty positions are also published. Operative Dentistry (ISSN ) is published bimonthly by Operative Dentistry, Indiana University School of Dentistry, Room S411, 1121 West Michigan Street, Indianapolis, IN Periodicals postage paid at Indianapolis, IN, and additional mailing offices. Postmaster: Send address changes to: Operative Dentistry, Indiana University School of Dentistry, Room S411, 1121 West Michigan Street, Indianapolis, IN Subscriptions: Fax (317) Current pricing for individual, institutional and dental student subscriptions (both USA and all other countries) can be found at our website: or by contacting our subscription manager: Fax 317/ jmatis@indy.rr.com Information on single copies, back issues and reprints is also available. Make remittances payable (in US dollars only) to Operative Dentistry and send to the above address. Credit card payment (Visa, MasterCard, or JCB Japanese equivalent) is also accepted by providing card type, card number, expiration date, and name as it appears on the card. Contributions Contributors should study the instructions for their guidance printed in this journal and should follow them carefully. Permission For permission to reproduce material from Operative Dentistry please apply to Operative Dentistry at the above address. The views expressed in Operative Dentistry do not necessarily represent those of the Academies or of the Editors. Editorial Office Operative Dentistry Indiana University School of Dentistry, Room S West Michigan Street, Indianapolis, IN Telephone: (317) , Fax: (317) URL: Editorial Staff Editor: Michael A Cochran Editorial Assistant/Subscription Manager: Joan Matis Editorial Associate: Karen E Wilczewski Associate Editors: Bruce A Matis, Edward J DeSchepper and Richard B McCoy Managing Editor: Timothy J Carlson Assistant Managing Editors: Joel M Wagoner and Ronald K Harris Editorial Board Kinley K Adams Maxwell H Anderson Steven R Armstrong Tar-Chee Aw Wayne W Barkmeier Douglas M Barnes Mark W Beatty K. Birgitta Brown Lawrence W Blank Murray R Bouschlicher William W Brackett William Browning Fred J Certosimo Daniel CN Chan David G Charlton Gordon J Christensen Kwok-hung Chung N Blaine Cook David Covey Gerald E Denehy Joseph B Dennison E Steven Duke William J Dunn Frederick C Eichmiller Sigfus T Eliasson Omar M El-Mowafy John W Farah Dennis J Fasbinder Mark Fitzgerald Kevin B Frazier Toh Chooi Gait James C Gold Valeria V Gordan William A Gregory Charles B Hermesch Harald O Heymann Van B Haywood Richard J Hoard Ronald C House Poonam Jain Gordon K Jones Barry Katz Robert C Keene Edwina A M Kidd George T Knight Kelly R Kofford Harold R Laswell Mark A Latta Xavier Lepe Walter Loesche Melvin R Lund Dorothy McComb Jonathan C Meiers Georg Meyer Ivar A Mjör Michael P Molvar B Keith Moore Graham J Mount David F Murchison Jennifer Neo Jacques E Nör John W Osborne Michael W Parker Craig J Passon Tilly Peters Frank E Pink T R Pitt Ford Jeffrey A Platt L Virginia Powell James C Ragain John W Reinhardt Philip J Rinaudo André V Ritter Frank T Robertello Henry A St Germain, Jr David C Sarrett Gregory E Smith W Dan Sneed Ivan Stangel James M Strother James B Summitt Edward J Swift, Jr Peter T Triolo, Jr Karen Troendle Richard D Tucker Martin J Tyas Marcos Vargas Douglas R Verhoef Joel M Wagoner Charles W Wakefield Steve W Wallace Nairn H F Wilson Peter Yaman Adrian U J Yap 2002 Operative Dentistry, Inc. Printed in USA

3 1 In Memoriam Wilmer B Eames Wilmer B Eames Wilmer Ballou Eames, truly one of dentistry s greatest men, had a distinguished career spanning more than 50 years. In his lifetime, he had a wide variety of interests including a love of music and photography. An outstanding dental practitioner and academician, Dr Eames was a man of great integrity and character who lived his life with a quick wit and a wonderful sense of humor. Wilmer Eames was born in Kansas City, MO on May 8, His family moved many times in his early years, but finally settled in Grand Junction, Colorado, where they grew cherries. Will attended Boulder High School and became editor of the school paper, where he developed an interest in journalism and photography. Will was taking photos of the office of family dentist Dr John Smith one day when, during the session, Dr Smith convinced Will that he should apply to dental school. Dr Eames entered dental school in 1935 at Kansas City Western Dental College and supplemented his income by singing in a trio, working as a night switchboard operator and doing portrait photography. He met his wife during a portrait shoot and married Elaine in 1939, the same year he graduated from dental school. He first practiced in Grand Junction, but soon was called into the US Army, reaching the rank of Major at Lowery Field in Denver. After WW II, he practiced in Denver with Dr Miles Markley for two years, then went to Glenwood Springs, Colorado for 14 years. While at Glenwood Springs, he was active in his community and served as its school board president. While in private practice, Dr Eames began leading study clubs throughout the West and lecturing worldwide. Dr Eames developed his now famous 1:1 ratio for dental amalgam while practicing in western Colorado. Publication of this technique in JADA in 1959 became one of the most highly influential articles ever printed. Within 18 months, most manufacturers had changed their instructions, and dental schools had revised clinical procedure for amalgam trituration. Mercury hygiene was dramatically improved and clinical amalgam restorations were better standardized. Called the Eames technique, this procedure was readily incorporated into the high copper amalgam revolution that started in the late 60s. Today, no one would think of triturating amalgam using an 8:5 that was used prior to Dr Eames publication. Fittingly, the minimal mercury technique for amalgam is so routinely used today that Dr Eames name has been dropped from it. He had clocks in his office and house set at 12:59 or 1 to 1, and it was reported that his car license plate read 12:59. Because Dr Eames developed severe hand dermatitis, he was forced to quit practicing and took the position of Professor at Northwestern University in Chicago in He went on to become the associate dean at Northwestern, then in 1967, accepted an appointment as director of Applied Materials at Emory University in Atlanta, Georgia. During his 12 years at Emory, Dr Eames received multiple grants from NIDR and industry to support research. His work with stu-

4 2 Operative Dentistry dents enriched his and the students lives. While at Emory, Dr Eames published many of his 113 articles and 59 abstracts. Dr Eames left Emory in 1979 as Professor Emeritus to retire in Aurora, Colorado. Even though retired, he continued to write, review for journals, direct research and lecture worldwide. He was appointed clinical professor at the University of Colorado School of Dentistry and became the editor of the Journal of the Colorado Dental Association. Dr Eames lectured more than 600 times worldwide, with an emphasis on techniques not often taught or frequently overlooked in the scientific world but which were useful and practical for the practicing general dentist. His awards were many, including: the Wilmer Souder Award from the IADR; the Distinguished Service Award from the Board of Regents of the University of Colorado; the William John Gies Award, American College of Dentists; the Honus Maximus Award, Denver Dental Society; the Alumni Achievement Award, University of Missouri-Kansas City; honorary member of Alpha Omega; the Silver Fox Award, Emory University; the Hollenback Prize, Academy of Operative Dentistry; the Schweitzer Research Award, Greater New York Academy of Prosthodontics; the Hinman Distinguished Service Medallion, honorary member of American Dental Society of Europe; Dentist of the Year, Colorado Academy of General Dentistry; Legends of Dentistry, American Academy of Gold Foil Operators; Man of the Year Award, American College of Dentists; the Callahan Memorial Award, Ohio Dental Association, Who s Who in America and the list continues. Dr Eames dry sense of humor was ever-present. One of his infamous publications was on the #00 bur. Since round burs were getting smaller and smaller, he decided to beat everyone to the punch and publish on a bur with no head. He reported with some glee that the article was first sent to a highly respected journal, which accepted it without realizing the joke. Wilmer was then obligated to withdraw the article but was finally persuaded to publish it with co-author, Alfred E Neuman of Mad Magazine fame. In another instance, he developed a contoured matrix band for the Gets company. Since he was living in Atlanta and enjoyed big band music, he called the matrix band Dixieland Bands. Dr Eames service to dentistry has been enormous. He published work on cements, composite resins and high-speed instrumentation besides his work on amalgam. His honest, forthrightness got him into trouble, but his spirit will live on in those who knew him. Dentistry has lost a great friend, and he will be missed. Dr Eames died Sept 6, 2001 after a long illness. He is survived by his wife, Elaine, son, Douglas and daughter, Alice. John W Osborne Clifford S Litvak

5 Operative Dentistry, 2002, 27, 3 Editorial Solicitation Soliloquy Ireally dislike solicitations, particularly since most of the ones I receive are completely unsolicited. The endless piles of junk brochures and letters in our mailboxes, homes, offices and garbage dumps should certainly be viewed by the EPA as a major threat to the ecosystems of the world (perhaps even more dangerous than mercury). The constant stream of telephone calls that are perfectly timed to interrupt our concentration, important work, meals, sleep and those rare moments of leisure time are definite drains on productivity, sanity and the sanctity of the family unit. The most obnoxious are the completely impersonal. The flyers directed to occupant, the individual on the phone who horribly mispronounces your name and the dehumanizing recorded solicitation that does not even allow you to respond. Moreover, many telephone solicitors do not seem to know the meaning of the word solicitation, since I frequently ask, Is this a solicitation? and am told No, this is a courtesy call. (This tells me that they do not understand the meaning of the word courtesy, either.) I was raised not to be rude to people, so I try to be polite when attempting to end these unwanted conversations but occasionally my only recourse is to just hang up. To add insult to injury, I have had the solicitor call me back to complain about my rudeness for hanging up on them! This tells me that I can add rude to the list of words they do not understand. The burning question in my mind is: who are the individuals who respond favorably to this abuse and make it financially viable for companies to continue the process? There must be a number of masochistic folks out there who get some perverse pleasure from these invasions of time and privacy, and are willing to contribute substantial portions of their income to perpetuate the practice. Unfortunately, attempts to have my name removed from mailing and telephone lists have met with very limited success, and an unlisted phone number creates too many problems of its own. I am also loath to spend money for the various electronic screening programs and equipment created to keep solicitors at bay although I have received a number of solicitations to purchase these devices. Perhaps I just need to quit complaining and reconcile myself to the fact that solicitations, like death and taxes, are an unavoidable fact of life in today s world. Now that I have vented my dislike for solicitors in general as well as my distaste for the practice of solicitation, I find myself in an extremely awkward and embarrassing position. As I begin my third year as editor of Operative Dentistry, I really need to solicit your help, yet I cannot even personalize my request by using your name. I can only address my plea to the readership of this journal. There are three things that would be of immense help to this publication that only you can provide: feedback, clinical technique/case report articles and assistance in increasing our subscription base. First, as an editorial team, we have received hardly any feedback from you, our readers, related to the format, quality or content of the journal. It is easy to assume that no feedback is actually positive feedback and the fact that we do not hear from you indicates satisfaction with each issue, but we are not convinced that this is really the case. We have changed the appearance of the journal, increased its size, pushed to include more clinical research and clinical technique papers (more on this in a minute) and added different publication queues to expedite printing of clinical information. We have reduced the publication time of accepted manuscripts to 12 months or less, resulting in a more than 60% increase in submission of papers during the last year. We have continued to maintain our high ranking by the Institute for Scientific Information as a premier research journal. However, none of this means anything unless it provides information that you find relevant and important. In this context, we welcome and encourage letters to the editor either for publication or simply to voice an opinion. Please take the time to send us your suggestions on how we can improve the journal. You can write to the address on the inside front cover, us at editor@jopdent.org, fax us at or leave a phone message at and we will get back to you as quickly as

6 4 Operative Dentistry possible. We are not seeking praise; we want your input on how we can make the journal better for you. Second, we know that a vast percentage of our readers are engaged in the practice of dentistry and spend the majority of their time providing health care to their patient population. Many of you have indicated in the past that you would like to see more clinically relevant articles, along with the laboratory research. Since our readership is comprised of some of the finest, most skilled clinicians in the world, it would seem logical that they (you) supply the journal with outstanding articles on successful clinical techniques and interesting case studies. Unfortunately, we have encountered a great deal of difficulty obtaining this type of submission. Dr Max Anderson made a similar plea when he was editor and even published some examples to encourage others to share their expertise (Operative Dentistry 18(3) ). I know that many of you have the information to share but feel that your writing experience is limited and are uncomfortable with the process. Please, just send us your rough ideas and photographs and our editorial team will help transform them into pertinent papers. Finally, Operative Dentistry exists on the financial base provided by our subscriptions. If you have pride in your journal and feel that it is a worthwhile publication, then help expand our readership by sharing that knowledge and encouraging others to subscribe to Operative Dentistry. Does your Alma Mater s library carry Operative Dentistry? If not, urge them to purchase a subscription or buy them one as a taxdeductible donation. Do you have colleagues who could benefit from reading the journal? Introduce them to our publication and get them to subscribe. Are you a member of both parent academies? Use the double subscription payment in your dues to give a year s gift of Operative Dentistry to a colleague or a dental student in your family. The bottom line is that we want to make Operative Dentistry truly your journal and a publication that you both benefit from and take pride in. To do this we must solicit your assistance. The positive point regarding this particular solicitation is that it already comes with something you want (the journal) and you can read it and hopefully respond at whatever time best fits your personal schedule. Who knows? If this works, I may even change my own mind about solicitations well, maybe not. Michael A Cochran Editor

7 Operative Dentistry, 2002, 27, 5-11 Clinical Research A Clinical Evaluation of a Bleaching Agent Used With and Without Reservoirs BA Matis YS Hamdan MA Cochran GJ Eckert Clinical Relevance There is no clinical difference in tooth whitening after two hours of tray use whether or not reservoirs are present. SUMMARY This in vivo study evaluated the variation of tray fabrication (trays constructed with or without reservoirs) on the degree of color change of teeth and sensitivities associated with using a 15% carbamide peroxide bleaching agent for two hours once daily for 14 days. Patients returned in one, two, three, six and 12 weeks. Color changes were evaluated by subjective shade matching, comparing clinical photographs and through measurements obtained using a color-measuring device. Subjects were asked to keep a daily record of any tooth and gingival sensitivity on the right and left side of their maxillary dental arch for three weeks. Indiana University School of Dentistry, Clinical Research Section, 1121 West Michigan Street, Indianapolis, IN Bruce A Matis DDS, MSD, director, Clinical Research Section Yaser S Hamdan BDS, MSD, graduate student, Operative Dentistry Michael A Cochran DDS, MSD, director, Graduate Operative Dentistry Program George J Eckert MS, biostatistician, School of Medicine Colorimeter data showed that teeth lightened with agent with reservoirs were significantly lighter than teeth lightened with the same agent without reservoirs. However, the amount of lightening was below the threshold of visual differentiation. Shade guide and slide photography data showed no significant differences between teeth lightened with agent with reservoirs compared to teeth lightened with the same agent without reservoirs. In addition, no significant differences in tooth and gingival sensitivity were found between the tray side with reservoirs and those without reservoirs. INTRODUCTION At-home tooth bleaching using peroxide-containing materials has become very popular (Clinical Research Associates, 1997; Dental Advisors, 2000), and currently, more than 45 different products are available. Palmer (1995) showed that among dental practitioners in the US, those providing professional in-office tooth bleaching had decreased from 56% in 1993 to 44% in 1995; however, dentists dispensing at-home whitener had increased from 79% to 95% during the same period. A very relevant question in the science of bleaching is whether reservoirs are necessary. In 1997, Haywood concluded in a pilot study that there was no apparent difference in the bleaching rate with or without reservoirs. Without using reservoirs in the bleaching tray,

8 6 Operative Dentistry the amount of bleaching material used to lighten teeth would be reduced as well as the time for tray fabrication. This in vivo study evaluated the effect of using reservoirs vs not using them on the degree of color change, rebound effects and sensitivities associated with the daytime use of an at-home bleaching agent. A half-arch design with reservoirs on one side of each subject s bleaching tray and no reservoirs on the other was used. METHODS AND MATERIALS The manufacturer (Rembrandt Xtra-Comfort Non- Sensitizing Bleaching Gel Regular Strength, Den-Mat Corp, Santa Maria, CA 93456, USA) supplied the bleaching agent used in this double-blind study. It had a 15% concentration of carbamide peroxide. Subjects who met the inclusion/exclusion criteria (Table 1) were randomly divided into two groups. Two alginate impressions of each subject s maxillary arch were taken. The first model was used to fabricate the bleaching tray. Paint-On Dental Dam (Den-Mat Corp) was applied as a block-out material to the central and lateral incisor and the canine on one side of the arch to create tray reservoirs on these teeth. The block-out was applied so that the labial surface was covered with the exception of 1 mm mesially, distally and cervically. The other half arch had no reservoirs (Figure 1). A study monitor randomly assigned which side of the maxillary arch would have reservoirs. The trays were made by a vacuum-formed process using Sheet Resin (Den-Mat Corp). As recommended by the manufacturer, the excess was trimmed on the labial and lingual surfaces to the gingival junction. The subjects were instructed by the study monitor regarding how to place the correct amount of whitening agent in the tray. The second study model was used to construct a positioning jig with full palatal coverage. The jig was indexed with a dual-prong precision attachment (Coltene/Whaledent Mahwah, NJ 07430, USA) to ensure that the light-measuring device could be precisely repositioned at each evaluation. Extrinsic stains of the teeth were removed with a dental prophylaxis using Nupro prophylaxis fluoride paste (Dentsply, Preventive Care, York, PA 17404, USA). The prophylaxis was performed at least two weeks prior to initiating the active study phase. Preoperative evaluation was done on the maxillary anterior teeth and their surrounding soft tissues. During the preoperative evaluation, a Loe & Silness Gingival Index was conducted to qualify patients for the study. The maxillary teeth were then evaluated by: 1) clinical photographs recorded with Ektachrome Elite 100, 35 mm color slide film (Kodak, Rochester, NY 14650, USA) with a shade tab of B-54 in each slide frame as a constant color; 2) shade matching of right and left anterior teeth with Trubyte Bioform Color Ordered Shade Guide (Dentsply, Trubyte, York, PA 17405, USA) and 3) Colorimeter Table1: Inclusion/Exclusion Criteria to Qualify for Study Inclusion Criteria Have all six maxillary anterior teeth. None of the maxillary anterior teeth can have more than 1/6 of the labial surfaces of the natural tooth covered with a restoration, and the location must not interfere with colorimeter placement. All six anterior teeth must be darker than B54 and lighter than B85 on the Trubyte Bioform Color Ordered Shade Guide. None of the maxillary anterior teeth can be excessively rotated, such as mesiorotation or distorotation, which interferes with colorimeter placement. Be willing to sign a consent form. Be at least 18 years of age. Be able to return for periodic examinations. Be willing to refrain from the use of tobacco products during the study period. Exclusion Criteria A history of any medical condition that may interfere with the study or require special attention, such as tuberculosis, hepatitis, AIDS, AIDS-related disease, other infectious disease, blood disorders, pacemaker, any other condition requiring pre-medication and other conditions left up to the judgment of the principal investigator. Use of any kind of tobacco products during the preceding 30 days. Subjects who have used professionally applied or prescribed tooth whiteners, whether in-office or at-home, in the preceding five years. Gross pathology in the oral cavity (excluding caries). Gingival index score greater than 1.0. Intrinsic discolored teeth due to tetracycline staining. Pregnant or lactating women. readings (Chroma Meter Model SR-321 Minolta, Ramsey, NJ 07446, USA) to measure L*, a* and b* of the six maxillary anterior teeth using the custom-fitted positioning jig (Figure 2). The L*, a* and b* color space system was defined by the Commission International de l Eclairage in 1979 and is referred to as CIELAB (International Commission on Illumination, 1978). The L* represents the value where white is 100 and black is 0. A positive a* value indicates the red direction, a negative a* value the green direction, a positive b* value the yellow direction and a negative b* value the blue direction (Matis & others, 1998). Total color differences between two colors ( E) are calculated using the formula (International Commission on Illumination, 1978): E ab* = [( L*) 2 + ( a*) 2 + ( b*) 1/2 ]. All subjects were instructed to insert the mouth guard containing the bleaching agent for two hours, once a day for 14 days. They were told to brush their teeth at least twice daily for oral hygiene standardization. Subjects were also asked to keep a daily record in five categories (1: no sensitivity; 2: slight sensitivity; 3: moderate sensitivity; 4: considerable sensitivity; 5: severe sensitivity) of any tooth or gingival sensitivity.

9 Matis & Others: A Clinical Evaluation of a Bleaching Agent Used With and Without Reservoirs 7 Figure 1. Maxillary arch with one side blocked out to fabricate reservoirs in trays. Figure 2. Custom-made jig on the maxillary arch with the cone head, which fits on the colorimeter, seated in place. colorimeter was connected to a PC running Spectra QC software (Minolta, Ramsey, NJ 07446, USA) capable of directly recording and analyzing the readings, similarly to what had been accomplished in a previous study (Mokhlis, 1998). Figure 3. Standard B-54 shade tab positioned in place during baseline photograph of maxillary arch. During the active phase of treatment and for seven days after the cessation of bleaching, they were to indicate whether any sensitivity was present on the left or right side. Subjects experiencing more than a moderate degree of sensitivity after using the bleaching agent were asked to notify the study monitor. They were to be given Desensitizing Gel (Den Mat Corp) containing 5% potassium nitrate to use in their tray to reduce sensitivity. Patients were recalled at one, two, three, six and 12 weeks. Identical three-step examinations were conducted at each recall by the same examiners who conducted the pre-operative evaluation. The photographs were compared by two experienced independent evaluators for color changes on the right and left sides of the maxillary arch. The evaluators categorized each side of the maxillary arch into one of four gradients (0: no difference; 1: slight; 2: moderate; 3: significant). While taking photographs, the subjects held a standard shade guide (B-54) beside their anterior teeth (Figure 3). If differences existed between the right or left sides, the evaluators were required to develop a consensus to determine the lighter side. The Statistical Methods Assignment to half-tray type (half with reservoirs and the other half without) was examined for baseline differences in colorimeter measurements and shade guide rank order using analysis of variance (ANOVA). The ANOVA models included fixed effects for tooth type, half-tray and tray-by-tooth interaction, and a random subject effect to correlate multiple measurements from each subject. Change in colorimeter measurements and shade guide rank orders were computed by subtracting the baseline from the follow-up measurements. The L*, a*, b*, E and change in shade guide comparisons were made using ANOVA. The ANOVA included fixed effects for tooth type, half-tray type, examination and interactions between those effects. Baseline values were included as covariates. Random subject effects were included to correlate measurements on multiple teeth and to correlate measurements on the same teeth at multiple examinations. Pairwise comparisons between half-tray types were made using Tukey s multiple comparisons procedure to control the significance level for each comparison at 5%. At each exam, Wilcoxon Signed Rank tests were used to determine whether using the reservoir resulted in significantly lighter shades according to the clinical slide assessments. Gingival and tooth sensitivity comparisons were performed using ANOVA with fixed effects for half-tray type, day, half-tray type by-day interaction and random subject effects to correlate the sensitivities within and between days.

10 8 Operative Dentistry RESULTS Without Reservoir With Reservoir Weeks Figure 4. Change in L* over 12 weeks for teeth in quadrant lightened with agent in tray with reservoirs and the adjacent quadrant without reservoirs Without Reservoir With Reservoir Weeks Figure 5. Change in a* over 23 weeks of teeth in quadrant lightened with agent in tray with reservoirs and the adjacent quadrant without reservoirs Without Reservoir With Reservoir Weeks Figure 6. Change in b* over 12 weeks for teeth in quadrant lightened with agent in tray with reservoirs and the adjacent quadrant without reservoirs. Twenty-seven subjects were enrolled and completed the study, including 12 males and 15 females ranging in age from 23 to 68 years (average age being 48.3). Thirteen patients were assigned custom trays with reservoirs on the left anterior quadrant and 14 patients were assigned custom trays with reservoirs on the right anterior quadrant of their maxillary arch. Chroma Meter Data Neither side had a significantly different baseline L*(p=0.20), a* (p=0.57), b* (p=0.24) or shade guide (p=0.77). L* Quadrants lightened with agent in a tray with reservoirs had significantly higher L* than quadrants lightened with agent without reservoirs overall (p=0.0045) and for week 1 (p=0.0006), week 2 (p=0.0063), week 3 (p=0.0367) and week 6 (p=0.0271), but not for week 12 (p=0.33). L* continued to change over time until week 6, but there was no significant change between week 6 and 12 (p=0.93) (Figure 4). a* Quadrants lightened with agent in a tray with reservoirs and adjacent quadrants lightened without reservoirs had significantly different a* for week 2 (p=0.0424) and marginally different a* for week 1 (p=0.06). No significant difference between quadrants lightened with reservoirs and without reservoirs was found for a* overall (p=0.14) or for week 3 (p=0.20), week 6 (p=0.15) or week 12 (p=0.92) (Figure 5). b* Quadrants lightened with agent in a tray with reservoirs had significantly different b* than an adjacent quadrants lightened without reservoirs overall (p=0.0055) and for week 1 (p=0.0418), week 2 (p=0.0031), week 6 (p=0.0063) and week 12 (p=0.0203). Quadrants lightened in a tray with reservoirs had marginally different b* than adjacent quadrants lightened without reservoirs for week 3 (p=0.07). b* continued to change over time until week 3, but there were no significant changes after week 3 (p>0.08) (Figure 6). E Quadrants lightened with agent in trays with reservoirs had significantly higher E than adjacent quadrants lightened without reservoirs overall (p=0.0028) and for week 1 (p=0.0006), week 2 (p=0.0035), week 3 (p=0.0326) and week 12 (p=0.0188), but not for week 6

11 Matis & Others: A Clinical Evaluation of a Bleaching Agent Used With and Without Reservoirs (p=0.11). E continued to change over time until week 6, but there was no significant change between weeks 6 and 12 (p=0.30) (Figure 7). Without Reservoir With Reservoir Weeks Figure 7. Change in E* over 12 weeks of teeth in quadrants and lightened with agent in tray with reservoirs and the adjacent quadrant without reservoirs Without Reservoir With Reservoir Weeks Figure 8. Change in Shade Guide Tabs over a 12 week period with and without reservoirs Without Reservoir With Reservoir Days Figure 9. Change in soft tissue sensitivity over 14 days of bleaching and 7 days postbleaching of teeth with and without reservoirs Shade Guide Data Figure 8 shows shade guide readings with and without reservoirs. No significant difference between teeth lightened with reservoirs and adjacent teeth lightened without reservoirs was present for shade guide readings overall (p=0.78) or at any specific exam (p>0.25). Shade continued to change over time until week 6; however, there was no significant change between week 6 and 12 (p=0.27) (Figure 7). Slide Data The presence of the reservoir did not have a significant effect on lightness, as determined by the clinical slide assessments at baseline (p=1.00) or any follow-up exam (p>0.62) (Table 2). Due to an overexposure of film during processing, only 11 of the 27 patients had slides that were of diagnostic value at week 2. Sensitivity Data No significant difference was present between a maxillary quadrant lightened with reservoirs and an adjacent quadrant lightened without reservoirs for either gingival sensitivity (p=0.46) (Figure 9) or tooth sensitivity (p=0.90) (Figure 10). DISCUSSION Few studies have evaluated the effects of tray fabrication design (Haywood, Leonard & Nelson, 1993; Javaheri & Janis, 2000; Bosma & others, 2000). Those studies used subjective shade guide matching and clinical photographs without the inclusion of an objective color-measuring device. Colorimeter data in this study showed that quadrants lightened with reservoirs produced significantly higher L*, b* and E than adjacent quadrants lightened without reservoirs. However, the subjective shade matching and slide evaluation showed no significant difference between teeth lightened with reservoirs and those lightened without reservoirs. The difference between subjective and objective readings probably resulted from limitations of the human eye. The colorimeter expresses minute differences in color in numerical form, while subjective perception of color may be affected by color adaptation, background of viewing area or the light source illuminating the color. Colorimeters have sensitivities corresponding to those of the human eye, but, because they always take measure ments using the same light source and illumination method, the measurement conditions are much more standardized.

12 10 Operative Dentistry Table 2: Slide Evaluation of Sides With and Without Reservoirs in Trays Weeks Without Reservoir No Difference With Reservoir Total Slightly Lighter Slightly Lighter Without Reservoir With Reservoir Days Figure 10. Change in tooth sensitivity over 14 days of bleaching and 7 days of postbleaching of teeth with and without reservoirs. A recent crossover study by Yousef (2002) compared degradation of nine different products. Six used no reservoirs and three used reservoirs. The results showed that the percentage of carbamide peroxide recovered after two hours (50%) was significantly higher for trays designed with reservoirs than for trays designed without them (19%). However, the carbamide peroxide percentage in teeth samples in Yousef s study were not affected by whether or not reservoirs were used, within the same concentration (3.5 +/- 1.0 without reservoir vs 3.7 +/- 0.4 with reservoir for 10% CP, 5.2 +/- 1.4 vs 6.9 +/- 0.6 for 15% or 16% CP, 7.5 +/- 1.6 vs 7.5 +/- 0.9 for 20% or 22% CP). Similar results were obtained for the carbamide peroxide percentage in tray samples. This indicates that the availability of gel and not the bulk of material present is important. A study by Panich (1999) compared 15% carbamide peroxide and 5.5% hydrogen peroxide applied for a half-hour, twice daily for 14 days. In that study, all trays were designed with reservoirs. They reported a mean E of 4.42 and 3.44, at 2 weeks and 6 weeks, respectively. In this study, the mean E for two-hour daily exposure of 15% carbamide peroxide at 2 weeks was 4.56 without reservoirs, 5.33 with reservoirs. At 6 weeks, E in this study was 3.21 for trays without reservoirs and 3.43 with reservoirs. The current study used the same shade guide as the aformentioned study, and the same evaluators performed all subjective color evaluations. Panich reported a mean shade of and at 2 weeks and 6 weeks, respectively. In this study, the mean shade at 2 weeks was for both reservoir and non-reservoir sides. At 6 weeks the mean shade guide value was with reservoirs and for the sides without reservoirs. Comparing the values of E and shade from both studies gives an indication that using 15% concentration of carbamide peroxide applied for one half-hour, twice daily for 2 weeks can give results similar to the two-hour application of 15% carbamide peroxide for the same time period. The L*, a*, b* and E for the reservoirs group showed color relapses at a higher rate when compared to the non-reservoirs group during the first four weeks after termination of bleaching. Trays with reservoirs had significantly higher E than trays without reservoirs at weeks 1, 2, 3 and 12, but not for week 6. E continued to change over time until day 46. There was no significant difference between quadrants lightened with reservoirs and those lightened without reservoirs for shade guide readings overall or at any specific exam. This agrees with three other studies (Bosma & others, 2000; Javaheri & Janis, 2000; Haywood & others, 1993) that evaluated the design of bleaching trays using shade guides, but their results do not agree with those obtained with the colorimeter values in the current study, which indicated that trays with reservoirs had significantly higher b*, L* and E overall. The question then is, does objective statistical difference always translate to clinical significance? Ruyter, Nilner & Moller (1987) and Um & Ruyter (1991) suggest that a E of 1 unit is visually perceptible and 3.3 units is clinically acceptable. In this study, change in colorimeter measurements with and without reservoirs for E value was 0.77 at 2 weeks and 0.30 at 12 weeks. Statistical difference was documented, but clinical difference was below visual perception in this study. The shade guide data agreed, however, with colorimeter data in the sense that, for both sides, the majority of lightening occurred during the first week

13 Matis & Others: A Clinical Evaluation of a Bleaching Agent Used With and Without Reservoirs 11 and to a lesser extent during the second week of active bleaching. The shade guide rank data showed that color relapse started following discontinued bleaching, with most of the relapse taking place during the first week postbleaching. During all visits, use of the reservoirs did not have a significant effect on lightness determined by the clinical slide assessments. This agrees with the finding of Bosma & others (2000) in a similar study, and the subjective shade matching in this study. There was no significant difference between quadrants lightened with reservoirs and adjacent quadrants lightened without reservoirs for gingival sensitivity or tooth sensitivity. None of the participating subjects experienced greater than mild gingival or tooth sensitivity, and none was given a desensitizing gel. An exit questionnaire for this whitening study included the question, Did you notice a difference in the color between your upper and lower teeth? Twenty-four subjects answered yes, and three responded no. Another question was, Did you notice a difference in the color of your upper teeth between the right and the left side? If yes, which side is lighter in color? Twenty-four subjects answered no, three subjects replied yes. Two of the three subjects who answered yes pointed to the side with the reservoirs. CONCLUSIONS This three-month, double-blind clinical study was conducted to evaluate the effects of tray design on the degree of color change, rebound effects and sensitivities associated with using a daytime at-home bleaching agent containing 15% carbamide peroxide. The degree of color change and color relapse was evaluated objectively by using a colorimeter and subjectively by using a shade guide and photographs. Patients self-evaluated any tooth and gingival sensitivity they experienced by recording maxillary right or left side sensitivity during the first 21 days of the study. This study concluded: 1. Objective measurements with a colorimeter indicated the bleaching with tray reservoirs produced significantly greater tooth lightening than bleaching without reservoirs. 2. Subjective evaluations using the shade guide, slide photography and subject feedback indicated no significant difference between teeth lightened with or without using reservoirs in the tray. 3. There was no significant difference for tooth and gingival sensitivity in the sides with or without reservoirs. (Received 1 May 2001) Acknowledgements This study was funded by Den-Mat Corporation. References Bosma M, Bowman J, Dorfman W & Soll K (2000) Clinical evaluation of a tray fabrication design and effects on vital tooth bleaching Hilltop Research, Inc Project No , August 7. Clinical Research Associates (1997) Tooth bleaching, state-of-art 97 Clinical Research Associates Newsletter Dental Advisors (2000) Bleaching The Dental Advisor 17(7) 1-5. Haywood VB (1997) Nightgard vital bleaching; current concepts and research Journal of the American Dental Association S-25S. Haywood VB, Leonard RH Jr & Nelson CF (1993) Efficacy of foam liner in 10% carbamide peroxide bleaching technique Quintessence International 24(9) International Commission on Illumination (1978) Recommendations on uniform color spaces-color-difference equations psychometric color terms Supplement 2 to CIE Publication 15 Paris: Bureau Central de la CIE. Javaheri DS & Janis JN (2000) The efficacy of reservoirs in bleaching trays Operative Dentistry 25(3) Matis BA, Cochran MA, Eckert G & Carlson TL (1998) The efficacy and safety of a 10% carbamide peroxide bleaching gel Quintessence International 29(9) Mokhlis, GR (1998) A three month clinical evaluation of 20 percent carbamide peroxide and 7.5 percent hydrogen peroxide whitening agents during daytime use [Thesis] Indianapolis, IN: Indiana University School of Dentistry. Palmer R (1995) Popularity of at-home whiteners continues Dental Products Report December 62B-66B. Panich M (1999) In vivo evaluation of 15-percent carbamide peroxide and 5.5-percent hydrogen peroxide whitening agent during daytime use [Thesis] Indianapolis, IN: Indiana University, School of Dentistry. Ruyter IE, Nilner K & Moller B (1987) Color stability of dental composite resin materials for crown and bridge veneers Dental Materials 3(5) Um CM & Ruyter IE (1991) Staining of resin-based veneering materials with coffee and tea Quintessence International 22(5) Yousef M (2002) Degradation of bleaching gels after two hours in vivo as a function of tray design and carbamide peroxide concentration Operative Dentistry 27(1)

14 Operative Dentistry, 2002, 27, Degradation of Bleaching Gels In Vivo as a Function of Tray Design and Carbamide Peroxide Concentration BA Matis M Yousef MA Cochran GJ Eckert Clinical Relevance Products have been found to degrade at the same rate in trays with or without reservoirs; however, more tray areas without reservoirs had no measurable bleaching agent remaining at the end of two hours. SUMMARY This study determined the degradation of nine bleaching agents with different concentrations after two hours in vivo following the manufacturers recommendations. The nine carbamide peroxide products are 10%, 15% and 20% Opalescence, 10%, 15% and 22% Rembrandt and 10%, 16% and 22% Nite White Excel 2. Each subject wore the tray with the bleaching agent for two hours on three separate occasions. The amount of remaining carbamide peroxide was determined after each use. Evaluation of remaining amount of carbamide peroxide was calculated by the US Pharmacopeia method. Indiana University School of Dentistry, 1121 West Michigan Street, Indianapolis, IN Bruce A Matis, DDS, MSD, Director, Clinical Research Section Mohammed Yousef, DDS, MSD, Graduate Student, Operative Dentistry Michael A Cochran, DDS, MSD, Director, Graduate Operative Dentistry Program George J Eckert, MAS, Biostatistician, School of Medicine The study showed that the total carbamide peroxide percent recovered was significantly higher for Opalescence products (47% to 54%) compared to Nite White (22% to 25%) and Rembrandt bleaching gels (15% to 16%). It concluded that this difference was mostly due to the use of facial reservoirs with Opalescence products, and also that whitening gel in trays with reservoirs and trays without reservoirs degraded at the same rate. INTRODUCTION One of the reasons for seeking cosmetic dental care is discoloration of the anterior teeth. Even persons whose teeth are a normal color often want them whiter. With careful case selection, diagnosis and treatment planning, bleaching can change a patient s smile dramatically (Haywood, 1997). Bleaching is almost always less invasive and less expensive than other treatment options, such as crowns or veneers, but can be used in conjunction with such procedures when bleaching alone does not completely correct the discoloration or when shape modification is also necessary (Small, 1998). Although the mechanism by which bleaching agents work is not fully understood, it involves an oxidation process by which the molecules causing the discoloration are chemically modified. The bleaching process

15 Matis & Others: Degradation of Bleaching Gels In Vivo as a Function of Tray Design 13 therefore depends on the penetration of the bleaching agent into the source of the discoloration. The extent to which bleaching works depends on the type of stain, its etiology, the length of time the stain has been present, the bleaching agent itself, how frequently it is applied, how long it remains on the teeth and the concentration that is applied to the teeth (Leonard, Sharma & Haywood, 1998). It is therefore important to discover how much degradation occurs with different agents. This study determined the degradation of carbamide peroxide (CP) for nine products after two hours in vivo. Patient Selection METHODS AND MATERIALS Ten healthy adults were enrolled in this study. Each subject received a dental screening and oral tissue examination prior to starting the study. Inclusion criteria included: (1) six caries-free, unrestored maxillary anterior teeth; (2) a periodontal disease Gingival Index (Loe & Silness, 1963) <1.0, indicating the absence of periodontal disease; (3) the anatomical crown of the right central incisor being at least 9 mm in length and 8 mm in width; (4) being in a non-smoking state for at least 15 days prior to and during the study and (5) a willingness to sign an informed consent form. The Institutional Review Board of Indiana University-Purdue University Indianapolis approved the protocol and informed consent form. Nine bleaching agents were used in this study (Table 1). Procedures The subjects received a prophylaxis treatment with rubber cup and pumice at least four weeks prior to the study. During the same visit, an alginate impression (Jeltrate Plus, Caulk Division Dentsply International Inc, Milford, DE 19963, USA) of the upper arch of each subject was taken. The impression was then poured with Super-Die fast setting die stone (Whip Mix Corporation, Louisville, KY 40217, USA) to fabricate a study cast. Three bleaching trays were made for each subject according to manufacturers recommendations. Two of the trays had no facial reservoirs (Resin Sheets, Rembrandt Bleaching Gel, Den-Mat Corporation, Santa Maria, CA 93456, USA and EVA Material, Nite White Bleaching Gel, Discus Dental Inc, Los Angeles, CA 90232, USA), and one tray had facial reservoirs (Sof-Tray, Opalescence Tooth Whitening Gels, Ultradent Products Inc, South Jordan, UT 84095, USA). To fabricate a tray with facial reservoirs, a wax layer of 0.5 mm thickness was applied on the facial surface of the six anterior teeth using 26-gauge casting wax sheets (Kerr Corp, Orange, CA 92867, USA). One millimeter of tooth surface on the mesial, distal and incisal, and 2 mm of tooth surface on the cervical were left without wax. The cast was then immersed in water for five minutes, and a new impression of this cast was taken with alginate and poured to make a second study cast. From the second cast, a customized vacuumformed stint was fabricated. In all cases, the margins of the tray were trimmed just shy of the gingival margins. In this crossover design study, three different brands of bleaching gels were tested for degradation of carbamide peroxide after two hours in vivo. Each brand had three different concentrations of carbamide peroxide (CP) and each subject tested every product three times using different bleaching trays for each brand. Subjects were provided comfortable chairs and were instructed to refrain from talking, drinking or eating while using the product. The bleaching agents were analyzed in triplicate before and after the study to determine the carbamide peroxide percentage in the product at both times. A Table 1: Bleaching Agents Used with Manufacturers, Lot Numbers of Products, Labeled CP%, Assayed CP% Before Beginning of Study and Assayed CP% After Completion of Study Bleaching Agent/Manufacturer Lot # Label Assayed Assayed CP% CP% CP% Before After Nite White Excel 2/Discus Dental Inc, Culver City CA FP-9ES 10% Nite White Excel 2/Discus Dental Inc, Culver City CA DS-9EA 16% Nite White Excel/Discus Dental Inc, Culver City CA GC 22% Opalescence/Ultradent Products Inc, South Jordan UT GK 10% Opalescence FP/Ultradent Products Inc, South Jordan UT CS7 15% Opalescence FP/Ultradent Products Inc, South Jordan UT D8K 20% Rembrandt Xtra-Comfort/Den-Mat Corp, Santa Maria, CA % Rembrandt Xtra-Comfort/Den-Mat Corp, Santa Maria, CA % Rembrandt Xtra-Comfort/Den-Mat Corp, Santa Maria, CA %

16 14 Operative Dentistry Table 2: CP Concentration and Amount of CP Delivered and Recovered (Mean Values) Label CP Concentration (%) Amount of CP (mg) Delivered Recovered Delivered Recovered Percent Tray and Rinse Total Teeth Sample Sample Nite White 10% Nite White 16% Nite White 22% Opalescence 10% Opalescence 15% Opalescence 20% Rembrandt 10% Rembrandt 15% Rembrandt 22% Table 3: Repeatability Assessment of Carbamide Peroxide Percentage Recovered %CP Tray %CP Teeth % Recovered Bleaching Agent S b S w ICC S b S w ICC S b S w ICC Nite White 10% Nite White 16% Nite White 22% Opalescence 10% Opalescence 15% Opalescence 20% Rembrandt 10% Rembrandt 15% Rembrandt 22% Sb : between subject standard deviation Sw : within subject standard deviation ICC: intraclass correlation coefficients chemistry laboratory technician who was unaware of the specific products being tested. All the bleaching samples for a specific product were from the same lot. Opalescence and Rembrandt bleaching gels were refrigerated during the study as recommended by their manufacturer s instructions. Nite White gels were not refrigerated, as the manufacturer does not recommended it. The gel was added to a pre-weighed vacuum-formed tray. In the trays without facial reservoirs, the bleaching gel was dispensed in the shape of a line on the facial surface of the tray. For the tray with a reservoir, the gel was dispensed in a zig-zag shape to avoid dispensing excess material into the facial reservoirs. The filled tray was weighed to the nearest milligram on an analytical balance, Mettler AE100 (Mettler Instruments Corp, Hightstown, NJ 08520, USA). The tray was then seated into position in the subject s mouth. After two hours, the tray with the adherent gel was removed from the mouth and weighed. This was called the tray sample. It was placed in a beaker with a solution of 100 ml of water and 20 ml of acetic acid. The gel that remained on the teeth was scraped off using a spatula and placed on a tared polystyrene weighing dish and weighed on the analytical balance. This sample was called the teeth sample. It was placed in a beaker with 100 ml of water and 20 ml of acetic acid. The third sample was called the rinse sample. It consisted of four gentle mouth rinses using deionized water which the subject expectorated into a beaker after the tray was removed and the gel scraped from the teeth. Twenty ml of acetic acid was added per 100 ml of solution. This sample was not weighed. All samples were analyzed for peroxide content by the method stated in US Pharmacopeia (2000). All the samples were stirred with a stir bar over a stir plate ( rpm) to dissolve the acetic acid into

17 Matis & Others: Degradation of Bleaching Gels In Vivo as a Function of Tray Design 15 Figure 1. CP concentration in tray and teeth samples combined after two hours (mean and standard error of the mean). Figure 2. Amount of CP recovered in rinse samples after two hours (mean and standard error of the mean). Figure 3. Total percentage of CP recovered after two hours (mean and standard error of the mean). the solution. After adding acetic acid to all the samples, the beakers were covered using a glass dish, then stirred for approximately five minutes or until the gel was completely dissolved. Thereafter, 2.0 gm of potassium iodide was added to each sample. The sample colors changed from transparent to yellow depending on the concentration of peroxide in the sample. Two drops of ammonium molybdate were added to each sample by a dropper. The solution changed from yellow to darker yellow or orange depending on the concentration of peroxide in the sample. The glass cover was then replaced on the beakers, and the specimens were placed in a dark area for a minimum of 10 minutes. Each sample was removed and titrated with N sodium thiosulfate. As the titrant was added, the sample turned paler. When the color of the sample was very pale yellow, starch was added. The sample turned purple when the starch was added. The sample then was titrated to the end point, which was reached when the sample color became transparent. To determine the percentage of carbamide peroxide which was recovered, the following formula was used: CP% = V x x 4.704/W, where (W) is the total weight of a sample and volume (V) is the amount of thiosulfate used in titration. This formula was applied on tray and teeth samples to obtain the actual concentration of carbamide peroxide. Only the amount of carbamide peroxide was determined in the rinse sample. The result of each sample gave an estimate of the concentration. The percentage of carbamide peroxide recovered was obtained by dividing the total amount of CP recovered by the amount of CP delivered and multiplying by 100. Statistical Methods Three replicates per bleaching agent per subject were accomplished. Estimated within and between subject standard deviations and intra-class correlation coefficients (ICCs) were used to represent the repeatability of

18 16 Operative Dentistry the degradation measurements. These values were computed using separate ANOVA models for each bleaching agent, with a random effect for subject. The tray and teeth samples were combined and the average of the three replicates was computed. Comparisons between concentrations and different brands were made using repeated measures ANOVA. Brands, concentrations and the interaction between brands and concentrations were included in the models as fixed effects. The repeated subject effect was necessary because measurements of each of the bleaching agents made on the same subject may be correlated. Pairwise comparisons between the bleaching agents were made using the Sidak method to control the overall significance level at 5%: adjusted p-value = 1-(1- original p-value) # comparisons. RESULTS Ten subjects participated in this study eight females and two males. The subjects mean age was 50 years; they ranged from 27 to 84 years of age. Table 1 shows the percentage of carbamide peroxide for each bleaching gel before initiation and after study completion. After the study was completed, evaluating the bleaching agents showed a slight decrease of values compared to baseline. CP Concentration in Tray and Teeth Samples CP concentration in the combined tray and teeth samples was not significantly different for 10% Nite White, 10% Opalescence and 10% Rembrandt (p=0.70). CP concentration in the samples for 15% Opalescence was significantly higher than for 16% Nite White (p=0.0005) and marginally higher than for 15% Rembrandt (p=0.08). However, CP concentration in samples was not significantly different for 16% Nite White and 15% Rembrandt (p=0.99). CP concentration in the samples was not significantly different for 22% Nite White, 20% Opalescence and 22% Rembrandt (p=0.40) (Table 2, Figure 1). Amount of CP Recovered in Rinse Samples Opalescence 20% had a significantly lower amount of CP in the rinse sample than Nite White 22% (p=0.0125). Rembrandt 22% did not have significantly different amount of CP in the rinse sample from Nite White 22% (p=0.78) or Opalescence 20% (p=0.71). The amount of CP in the rinse sample was not significantly different for Nite White, Opalescence and Rembrandt for initial CP percent of 10% (p=0.58) or 15%/16% (p=0.79) (Table 2, Figure 2). Total Percentage of CP Recovered The percentage of CP recovered was significantly higher for Opalescence bleaching gels than for Nite White (p=0.0003) and Rembrandt (p=0.0003). The percentage of CP recovered was marginally higher for Nite White than Rembrandt (p=0.0955) (Table 2, Figure 3). DISCUSSION The methodology used in this study simulates the clinical situation for daytime bleaching. The six anterior teeth were selected for gel retrieval. For this study, subjects with larger teeth were selected to facilitate gel collection from the surface of the teeth. Restoration-free surfaces were chosen to allow for maximum contact of bleaching gel with natural tooth surface. Subjects with no periodontal disease were also chosen to avoid reaction of the bleaching gel with a larger number of microorganisms (Gurgan, Bolay & Alacam, 1996). Trays were filled but not over or underfilled with bleaching gel. Overfilling would cause physical loss of the material, while underfilling would produce voids that would reduce the reaction of the bleaching gel with the tooth surface. The Nite White Gel labeled 10% was assayed to have 10.89% carbamide peroxide; the 16% was assayed to have 11.35% carbamide peroxide and Nite White 22% was assayed to have 22.70% carbamide peroxide. Since the 10% and 22% concentrations of Nite White products were within the labeled percentage, the low assay of carbamide peroxide in the product labeled 16% could be due to a quality control problem. Table 3 shows the repeatability assessment of carbamide peroxide percentage recovered. The repeatability assessment was performed to determine the withinsubject consistency of the measurements from three uses of each product. When measurements are inconsistent from multiple uses, measurements should be made multiple times within a subject, then averaged to result in an accurate estimate for each subject. However, when measurements are consistent, multiple measurements from each subject is not necessary to achieve an accurate estimate. Repeatability assessment for tray samples was better than that for teeth samples, however, repeatability was good for all samples. The concentration of carbamide peroxide in teeth and tray samples, and the amount of carbamide peroxide in rinse samples after two hours in vivo were similar among the three brands. For all samples, the higher the concentration of the bleaching gel, the higher the concentration of carbamide peroxide recovered (Figures 1 and 2). This means that using reservoirs did not matter in the rate of degradation and all materials seemed to degrade at the same rate. This shows that teeth should lighten at the same rate regardless of whether or not reservoirs are used as long as there is material present. This would lead the authors to believe that reservoirs are not necessary when bleaching for two hours or less, and other factors besides tooth reactivity with CP, such as heat (temperature of the oral cavity) and saliva, are

19 Matis & Others: Degradation of Bleaching Gels In Vivo as a Function of Tray Design 17 involved in the degradation of bleaching agents. However, the total percentage of carbamide peroxide remaining after two hours was significantly higher for Opalescence products (47%-54%) than for Nite White gels (22%-25%) and Rembrandt gels (15% to 16%) (Figure 3). Using facial reservoirs with Opalescence products would explain this. Trays with facial reservoirs are passive in holding the bleaching gel in the trays and their use limits the amount of bleaching gel that is physically lost during active treatment. The loss in the total recoverable CP is probably due to the dynamic action of the non-reservoir trays trying to reach a resting state. The degradation results for Opalescence products agree with those found by Gaiao (1997). Gaiao tested the degradation of 10% Opalescence bleaching gel after one, two, four, six and 10 hours in vivo. He found that the percentage of carbamide peroxide remaining after two hours was 52%. The use of facial reservoirs is probably the important factor that resulted in a higher percentage of carbamide peroxide for Opalescence products compared to Rembrandt and Nite White bleaching gels. In the past, reference has been made to the more rapid degradation of peroxide in proximity to the tooth surface (Matis & others, 1999). This study documents that a minimal amount of peroxide is used in the bleaching process itself, as the concentrations of CP do not differ in the tray and teeth samples. In retrospect, the authors have determined that in the Gaiao (1997) study, salivary contamination increased the weight of the sample and, with the amount of peroxide remaining the same, it manifested itself as decreased concentration on the facial and tray surfaces. Wattanapayungkul & others (1999) suspected this concentration gradient error in her study. Clinical Research Associates (CRA) (1989) found that the percentage of carbamide peroxide remaining after one hour in vivo was less than half the original volume. In another study, CRA (1997) found that the percentage of carbamide peroxide remaining after two hours ranged from zero to 30% as the maximum. Ten percent Nite White had 30% of the initial carbamide peroxide remaining after two hours, while 10% Opalescence had 20% of the initial carbamide peroxide remaining after two hours. However, in CRA studies, facial reservoirs were not used. The method of calculating CP was also different from the method used in this study. The results of the current study could explain the relationship between the percentage of carbamide peroxide and bleaching efficacy. Each manufacturer had 30 trials in each of their three products. In 90 trials of each brand using Rembrandt, no active agent was recovered from seven trials; when using Nite White, no active agent was recovered in nine trials and no active agent was recovered in one trial using Opalescence. Other studies (Mousa, 1998; Leonard & others, 1998) have shown that the higher the carbamide peroxide percentage in bleaching gels, the faster the bleaching process. In the two-week in vitro laboratory study on extracted teeth reported by Leonard & others (1998), a 16% carbamide peroxide solution (Nite White) successfully obtained a three-unit Vita shade change more quickly than 5% or 10% carbamide peroxide solutions. At the end of the two-week regimen of in vitro whitening, no significant differences existed in tooth shade among the 10% or 16% carbamide peroxide gel treated groups. However, teeth treated with 10% and 16% carbamide peroxide gel were significantly lighter than those treated with the 5% carbamide peroxide gel. Extending treatment to a third week for the teeth treated with 5% carbamide peroxide gel resulted in lightened shades that approached the two-week values for the 10% and 16% concentrations. The study concluded that lower concentrations of carbamide peroxide gel will whiten teeth and can achieve the same results as higher concentrations; the process just takes longer. In a clinical in vivo study, Hamdan (2002) used 15% Rembrandt bleaching gel in trays with and without reservoirs to evaluate the color change of teeth. Shade guide measurements and a colorimeter device were used to evaluate the color change. He reported that teeth bleached by trays with reservoirs had significantly higher E (total color difference) than teeth bleached by trays without reservoirs. E was 3.36 for the reservoir group, while it was 3.05 for the no reservoir group. This means that teeth bleached by trays with reservoirs were significantly lighter than the other group. However, this difference was not significant for shade guide readings or photographic assessment. Clinically, the difference between both groups would be detectable by the human eye if the difference between E for both groups was one point or higher (Seghi, Hewlett & Kim, 1989; Kuehni & Marcus, 1979; Rubino & others, 1994). These results would indicate that for two hours of daytime home bleaching, the presence of reservoirs would not lead to a clinically significant color change compared to trays without reservoirs. This study showed that much more bleaching gel is lost into the oral cavity and probably ingested by subjects when reservoirs are not used. In this study, more than 25% of the CP was ingested when the subjects did not use reservoirs. Wattanapayungkul & others (1999) reported that the collected saliva from subjects using bleaching trays with reservoirs revealed an average of 2.10 mg or 6.6% of the carbamide peroxide placed in the tray after one hour of bleaching. This number, however, would increase if the bleaching trays did not have facial reservoirs. Also, this number could decrease if the subject wore the tray at night when the salivary flow and the frequency of swallowing is decreased.

20 18 Operative Dentistry This study has shown conclusively that much more whitening gel is lost to the oral environment when reservoirs are not used. In the past, when a negligible amount of gel was lost, one could refer to it as degradation of the gel; however, now one must begin referring to the kinetics of active ingredient release (American Dental Association, 1998). The amount of active ingredient recovered needs to be determined whether the breakdown is a chemical breakdown of the agent, a breakdown due to antioxidants on the surface of the teeth or a physical loss of the product. The ADA guidelines for acceptance of a product specifically refer to the necessity of determining the kinetics of the active ingredient release of whitening gels before accepting any product. In the future, one must refer to this physical and chemical loss of active agent as the kinetics of whitening agents, now that whitening agents are being used in trays without reservoirs. CONCLUSIONS This study evaluated the degradation of nine bleaching gels after two hours in vivo. Ten healthy adults participated in the study over a period of four months. Three different brands of bleaching gels were tested (Nite White, Opalescence and Rembrandt). Each brand had three different concentrations of carbamide peroxide. The manufacturers of Nite White and Rembrandt recommended no reservoirs in trays holding their products, and the manufacturer of Opalescence recommended facial reservoirs in trays holding its products. Each subject tested every product three times using different bleaching trays for each brand. Each subject made a total of 27 visits. After two hours of wearing the trays, the bleaching gel was retrieved from the surfaces of teeth, from the bleaching trays and intraorally by four mouth rinses with deioninzed water. Analyses of carbamide peroxide were conducted following the US Pharmacopeia method. The study concluded that: 1. All nine products tested degraded at the same rate. 2. Total recovery of carbamide peroxide was significantly higher when reservoirs were placed in trays. 3. More carbamide peroxide is ingested when used in trays without reservoirs. Acknowledgements This study was funded by Den-Mat Corporation and Ultradent Products, Inc. References American Dental Association (1998) Acceptance program guidelines for home-use tooth whitening products Council on Scientific Affairs Chicago, IL. Clinical Research Associates (1989) Tooth bleaching, home-use products in vivo assays for presence of gel in trays Clinical Research Associates Newsletter 13(12) 2. Clinical Research Associates (1997) Tooth bleaching, state-of-art 97 Clinical Research Associates Newsletter 21(4) 3. Gaiao U (1997) A clinical study of 10-percent carbamide peroxide degradation in bleaching trays [Thesis] Indianapolis IN Indiana University School of Dentistry. Gurgan S, Bolay S & Alacam R (1996) Antibacterial activity of 10 percent carbamide peroxide bleaching agents Journal of Endodontics 22(7) Hamdan Y (2002) A clinical evaluation of a bleaching agent used with and without reservoirs Operative Dentistry 27(1) Haywood VB (1997) Historical development of whiteners: Clinical safety and efficacy Dental Update 24(3) Kuehni RG & Marcus RT (1979) An experiment in visual scaling of small color differences Color Research Applications Leonard RH, Sharma A & Haywood VB (1998) Use of different concentration of carbamide peroxide for bleaching: An in vitro study Quintessence International 29(8) Loe H & Silness J (1963) Periodontal disease in pregnancy (Pt 1) Prevalence and severity Acta Odontologica Scandinavica Matis BA, Gaiao U, Blackman D, Schultz FA & Eckert GJ (1999) In vivo degradation of bleaching gel used in whitening teeth Journal of the American Dental Association 130(2) Mousa HN (1998) A clinical evaluation of bleaching agents at different concentrations [Thesis] Indianapolis IN Indiana University School of Dentistry. Rubino M, Garcia JA, Jimenz del Barco L & Romero J (1994) Color measurement of human teeth and evaluation of a color guide Color Research and Application Seghi RR, Hewlett ER & Kim J (1989) Visual and instrumental colorimetric assessments of small color differences on translucent porcelain Journal of Dental Research 68(12) Small BW (1998) The application and integration of at-home bleaching into private dental practice Compendium of Continuing Education in Dentistry 19(8) , 802, 804. United States Pharmacopeia (2000) The National Formulary United States Pharmacopeial Convention, Inc, Rockville, Maryland. Wattanapayungkul P, Matis BA, Cochran MA & Moore BK (1999) A clinical study of the effect of pellicle on the degradation of 10% carbamide peroxide within the first hour Quintessence International 30(11) (Received 25 May 2001)

21 Operative Dentistry, 2002, 27, Laboratory Research Microleakage of Direct and Indirect Composite Restorations with Three Dentin Bonding Agents AA Alavi N Kianimanesh Clinical Relevance The dentin bonding agents tested significantly reduce microleakage, however, they do not completely prevent microleakage. There is no advantage to the indirect technique in small Class V cavities. SUMMARY This study evaluated the marginal sealing ability of direct and indirect (inlay) resin composite restorations with three dentin bonding systems. Forty-eight freshly extracted bovine incisor teeth were randomly assigned to four groups for bonding with Syntac Single-Component, Excite (Vivadent, Liechtenstein), ScotchBond Multi- Purpose Plus (3M Dental Products, St Paul, MN 55144, USA) and a control with no bonding agents. Class V cavities were cut in both buccal and lingual surfaces. The coronal half of each preparation was in enamel and the gingival half was in cementum or dentin. Half of the specimens in each group were restored with direct and the remainder with indirect technique. Operative Department, School of Dental Medicine, Shiraz University of Medical Sciences, Shiraz, Iran A Alavi, DMD, MScD, associate professor and chair N Kianimanesh, DMD, MScD, assistant professor The teeth were stored in 37 C water for 30 days, then thermocycled. After immersion in 0.5% basic fuchsin, the teeth were cut faciolingually and evaluated for dye penetration using a binocular stereomicroscope. There was no significant difference among the bonding systems for either the direct or indirect technique or between the two techniques used for each system, however, the indirect technique showed significantly (p=0.001) less microleakage than the direct technique in control groups. All groups showed more leakage at the cementum margins except Excite with direct technique, where microleakage at the incisal and gingival margins was almost equal. INTRODUCTION Adhesive dentistry promotes the use of preparations that preserve tooth substance. However, in spite of technologic advances, there are still shortcomings in current resin adhesive techniques. Strong, durable adhesion between cavity walls and restorative materials is necessary to produce well-sealed, long lasting restorations.

22 20 Operative Dentistry Previous studies have demonstrated that etched and bonded enamel produces a more consistent seal compared to dentin (Swift, Perdigão & Heymann, 1995). Adhesion to dentin is more complicated because of the composition and histologic structure of the substrate (Pashley & Carvalho, 1997). Current generations of dentin bonding agents (DBA) are designated as fourth generation (multiple-step) and fifth generation (simplified, two-step or one-bottle). New one-bottle dentin adhesives combine primer and adhesive resin into a single solution; some also contain nanoscale fillers to absorb the stresses at the interface of adhesive bond. Perdigão, Baratieri & Lopez (1999); Swift & others (1999); Latta & others (1997) and Ario (1997) reported relatively equal performance within these systems. However, some contradictory results have been reported by other researchers (Castelnuovo, Tjan & Liu, 1996). The major shortcoming of visible light-cured composites is the polymerization contraction that results in gap formation, particularly at the dentin interface. This shrinkage of resin-based restorations, coupled with and masticatory forces generates stresses within the adhesive layer that must be resisted to retain the restoration and maintain marginal integrity (Retief, 1994). Along with material improvements, more sophisticated and multi-layering insertion techniques were developed to help overcome the destructive forces and to achieve better marginal adaptation and seal (Dietschi & others, 1995). Different resin-composite inlay systems were also developed. In theory, shrinkage stress should be minimized with an indirect technique since polymerization occurs prior to cementing the restoration (Ruyter 1992). Resin inlays are reported to have less microleakage than direct resin composite restorations (Robinson, Moore & Swartz, 1987; Douglas, Fields & Fundingsland, 1989). However, the flow-active free surface of the luting resin composite is relatively small at the narrow inlaytooth marginal gap, yielding a high C-factor (ratio of bonded to unbonded surfaces). Consequently, the luting resin composite is not likely to provide enough compensation for the shrinkage stress induced by polymerization of the luting resin composite (Feilzer, de Gee & Davidson, 1987). This in vitro study compared the ability of two onebottle and one multi-step DBA in preventing or reducing microleakage around direct and indirect Class V resin composite restorations. METHODS AND MATERIALS Forty-eight freshly extracted bovine incisor teeth were selected. The teeth were scaled and cleaned with a prophylaxis brush and stored in tap water. Facial and lingual Class V cavities (2.5 mm in height, 3 mm in mesiodistal direction and 1.5 mm in depth) were prepared with a fissure diamond in an air turbine at the CEJ and finished with a #170 carbide bur. Occlusal margins were placed in enamel and cervical margins in cementum/dentin. The teeth were randomly assigned to four groups (n=12) for bonding with Syntac single-component (Vivadent, Liechtenstein), Excite (Vivadent, Liechtenstein), ScotchBond Multi-Purpose Plus (3M Dental Products) and a control with no bonding agents (Table 1). Syntac SC is a self-priming or one-bottle bonding agent; Excite is another one-bottle system which also contains nanofillers. ScotchBond Multi- Purpose Plus (SBMP) is a multiple-step bonding system. Table 2 lists the composition of the adhesives used in this study. After conditioning the cavities with 37% phosphoric acid, all materials were applied according to the manufacturer s recommendations. Special attention was given to avoid desiccation of dentin. Half the teeth in each group (six teeth, 12 cavities) were restored with direct and the remainder with indirect technique using Tetric-Ceram (Vivadent, Liechtenstein) microhybrid composite. In the direct technique, cavities were filled incrementally, first occlusally, then gingivally. Each increment was cured for 40 seconds with a light-curing unit (Coltoux II, 50/60 Hz, 300 mw/cm 2 Coltene AG/Switzerland). Then the restoration surface was finished and polished sequentially with a finishing diamond, gray Politip-Finisher silicon rubber and green Politip-Polisher silicon rubber (Vivadent, Liechtenstein) in a low-speed handpiece. The cavities in the indirect technique were lubricated with Johnson & Johnson water-soluble lubricating gel, then filled with composite (Tetric-Ceram) and cured for 60 seconds. The inlays were then removed from the tooth and their internal surfaces were polymerized for 40 additional seconds. The inlays were subjected to a thermal treatment (100 C for 10 minutes in boiling water) for secondary polymerization. The inner surfaces of the inlays were treated with APF 1.23% for 10 minutes and silanized prior to cementation. The cavities were etched and the bonding agents applied in each group according to the manufacturer s instructions but were left uncured to ensure complete seating of inlays. The inner surface of the inlays was also covered with the same bonding agent used in the cavities in each group. The catalyst and base pastes of the dual cured resin composite cement Enforce (Dentsply/Caulk, Milford,

23 Alavi & Kianimanesh: Microleakage of Direct and Indirect Composite Restorations 21 Table 1: Materials Used Material Manufacturer ScotchBond Multi-Purpose Plus 3M Dental Products, St Paul, MN (SBMP) 55144, USA Syntac Single-Component Vivadent Ets Schaan, Liechtenstein Excite Vivadent Ets Schaan, Liechtenstein Tetric-Ceram Vivadent Ets Schaan, Liechtenstein Enforce Dentsply/Caulk, Milford, DE 19963, USA Table 2: Components of Adhesive Systems Bonding Conditioner Primer Adhesive Resin Agent (Gel) SBMP 37% H 3 PO 4 HEMA Bis-GMA polyalkenoic acid HEMA copolymer,water Syntac 37% H 3 PO 4 maleic acid HEMA methacrylate modified polyacrylic acid water Excite 37% H 3 PO 4 Phosphonic acid acrylate Hydroxy ethyl methacrylate Bis-GMA Dimethacrylate Highly dispersed silica (0.5%) ethanol was finally light cured for 60 seconds. Finishing and polishing procedures for the indirect technique were similar to those for the direct technique. The specimens were then stored for 30 days in tap water at approximately 37 C before being subjected to 200 thermal cycles in 5 C and 55 C water baths with a dwell time of 30 seconds. The root apices were sealed with sticky wax, and the teeth were covered with two layers of nail varnish except for the area of the restoration and a 1 mm border of tooth surrounding each cavity. All teeth were immersed in 0.5% basic fuchsin dye solution for 24 hours. The teeth were then cut faciolingually with a diamond saw (Leitz 1600). The penetration of dye was evaluated under a binocular stereomicroscope (Ziess). Dye penetration at the restoration/ tooth interface was scored for both occlusal and cervical margins on a scale from 0 to 4: 0 = no microleakage 1 =dye penetration within 1/3 of cavity wall 2 =dye penetration within 2/3 of cavity wall 3 =dye penetration within last 1/3 of cavity wall up to the axial wall 4 = dye penetration spreading along the axial wall Table 3: Frequency of Microleakage Scores at Enamel and Cementum Margin Adhesive Enamel Margins SBMP (D) SBMP (I) Syntac (D) Syntac (I) Excite (D) Excite (I) Control (D) Control (I) Cementum Margins SBMP (D) SBMP (I) Syntac (D) Syntac (I) Excite (D) Excite (I) Control (D) Control (I) DE 19963, USA) were mixed in a 1:1 ratio and applied on all internal surfaces of the cavity. The inlay was then seated completely and the excess cement removed. The resin composite luting material RESULTS Table 3 shows the frequency of microleakage scores at the enamel and cementum margins for both direct and indirect bonding techniques. To determine significant differences between the groups, data were analyzed using the Kruskal-Wallis non-parametric analysis of variance test. The Mann-Whitney U-test was used to compare the direct and indirect techniques for each of the three bonding materials and the control, as well as the incisal and gingival margins. The Kruskal-Wallis one-way analysis of variance revealed significant differences among the four groups for the overall, incisal and gingival score in both the direct and indirect technique. But there was no significant difference among the three experimental groups for the overall, incisal and gingival score in both techniques. The Mann-Whitney U-test showed gingival leakage to be significantly more than incisal leakage in all groups except Excite with direct technique where leakage of the incisal and gingival margin were almost equal (p=0.444) (Table 4).

24 22 Operative Dentistry Table 4: Comparison of Incisal and Gingival Leakage in Direct and Indirect Technique by Mann-Whitney U-test Bonding Agent Groups Two-Tailed P-Value Direct Technique Indirect Technique SBMP *non-significant Syntac Excite * Control Table 5: Comparison of Direct and Indirect Leakage in the Incisal and Gingival Margin by Mann-Whitney U-test Bonding Agent Groups Two-Tailed P-Value Incisal Margin Gingival Margin SBMP *significant Syntac Excite Control * * Table 6: Comparison of Direct and Indirect Technique by Mann-Whitney U-test Bonding Agent Groups Two-Tailed P-Value SBMP *significant Syntac Excite Control * There was no significant difference between the direct and indirect technique in the three experimental groups either in incisal or gingival margin, but in the control group, the indirect technique showed significantly less microleakage both in the incisal (p=0.027) and gingival (p=0.003) margin (Table 5). Table 6 shows the comparison between the direct and indirect technique for each of the three experimental groups and the control group with combined incisal and gingival scores. No dye penetration was seen along the inlay/luting composite interface in any of the samples. DISCUSSION Bovine teeth were selected for this study because they are considered an acceptable substitute for human teeth in laboratory studies (Reeves & others, 1995) and they afford less risk of infectious disease transmission. Due to the influence of mechanical properties of various filling materials on the efficacy of dentin bonding systems, one resin composite and one luting composite was utilized to exclude changes in volume or stress distribution of different composite materials (Hasegawa & others, 1999). Data from this study revealed no significant differences in the degree of microleakage between one multi-step and two onebottle adhesive systems. These results are in accordance with Latta & others (1997) but must be supported by long-term studies. There was a significant difference between cementum and enamel margins both in the direct and indirect technique, which is in agreement with several studies (Castelnuovo & others 1996; Saunders & Saunders, 1996; Chan & Swift, 1994). With the exception of Excite in the direct technique, microleakage in enamel and cementum margins were almost equal (Table 4). Excite with the direct technique provided the best seal in cementum margin in this study (Figure 1). Excite is a fifth-generation adhesive and contains nanoscale fillers. In agreement with this result, Pilo & Ben- Amar (1999) and Castelnuovo & others (1996) found One-Step (fifth-generation adhesive) and Optibond FL the best materials for sealing cementum margins. The good performance of Optibond FL adhesive in these studies was attributed to the addition of filler to the bonding system. In our direct restorations, SBMP provided the best seal at the incisal margin (mean score = 0.08) and the worst seal at the cementum/dentin margin (mean score = 0.46). Pilo & Ben-Amar (1999) also found almost similar results. Griffiths & Watson (1995) studied the resin-dentin interface of SBMP adhesive and reported that the primed dentin surface is extremely fragile during subsequent placement of the adhesive. In the regions where the adhesive was in direct contact with the dentin, an optimal bond could not develop. Several studies compared the sealing performance of direct and indirect resin-composite restorations and found better results with the indirect technique (Robinson & others, 1987; Dietschi & others, 1995). In this study, there was no significant difference between direct and indirect technique using the adhesive systems. These results could be attributed to the high sensitivity of the indirect technique, more procedural steps and high polymerization shrinkage stress in resin composite cement. In other words, there was no advantage in this study for the indirect technique on small or standard sized Class V cavities filled incrementally using dentin

25 Alavi & Kianimanesh: Microleakage of Direct and Indirect Composite Restorations 23 If bonding agents are properly applied, there is no advantage to the indirect technique in small Class V cavities. (Received 5 February 2001) Figure 1. Mean score at direct technique. Figure 2. Mean score at indirect technique. bonding agents. However, in the control group where no bonding agent was used, the indirect technique showed significantly better results (p=0.001). CONCLUSIONS The dentin bonding agents tested did not prevent microleakage at either the enamel or cementum/ dentin margins of Class V cavities regardless of which restorative technique was employed. References Ario PD (1997) Shear bond strength of a new 3M dental adhesive Journal of Dental Research (abstract). Castelnuovo J, Tjan AH & Liu P (1996) Microleakage of multi-step and simplified-step bonding systems American Journal of Dentistry 9(6) Dietschi D, Siebenthal G, Neveu- Rosenstand L & Holz J (1995) Influence of the restorative technique and new adhesives on the dentin marginal seal and adaptation of resin composite Class II restoration: An in vitro evaluation Quintessence International 26(10) Douglas WH, Fields RP & Fundingsland J (1989) A comparison between the microleakage of direct and indirect composite restorative systems Journal of Dentistry Feilzer AJ, de Gee AJ & Davidson CL (1987) Setting stress in composite resin in relation to configuration of the restorations Journal of Dental Research 66(11) Griffiths BM & Watson TF (1995) Resindentin interface of Scotchbond Multi- Purpose dentin adhesive American Journal of Dentistry 8(4) Hasegawa T, Itoh K, Koike T, Yukitani W, Hisamitsu H, Wakumoto S &Fujishima A (1999) Effect of mechanical properties of resin composites on the efficacy of the dentin bonding systems Operative Dentistry 24(6) Latta MA, Wilwerding TM, Barkmeier WW & Los SA (1997) Laboratory evaluation of one-component dental adhesive Journal of Dental Research (abstract). Pashley DH & Carvalho RM (1997) Dentine permeability and dentine adhesion Journal of Dentistry 25(5) Perdigão J, Baratieri L & Lopez M (1999) Laboratory evaluation and clinical application of a new one-bottle adhesive Journal of Esthetic Dentistry 11(1)

26 24 Operative Dentistry Pilo R & Ben-Amar A (1999) Comparison of microleakage for three one-bottle and three multi-step dentin bonding agents Journal of Prosthetic Dentistry 82(2) Reeves GW, Fitchie JG, Hembree JH & Puckett AD (1995) Microleakage of new dentin bonding systems using human and bovine teeth Operative Dentistry 20(6) Retief DH (1994) Do adhesive prevent microleakage? International Dental Journal 44(1) Robinson PB, Moore BK & Swartz, ML(1987) Comparison of microleakage in direct and indirect composite resin restorations in vitro Operative Dentistry 12(3) Ruyter IE (1992) Types of resin-based inlay materials and their properties International Dental Journal 42(3) Saunders WP & Saunders EM (1996) Microleakage of bonding agents with wet and dry bonding techniques American Journal of Dentistry 9(1) Swift EJ, Perdigão J & Heymann HO (1995) Bonding to enamel and dentin: A brief history and state of the art Quintessence International 26(2) Swift EJ, Perdigão J, Heymann HO & Ritter A (1999) Shear bond strength of one-bottle adhesives to moist enamel Journal of Esthetic Dentistry 11(2)

27 Operative Dentistry, 2002, 27, Effects of Cyclic Temperature Changes on Hardness of Composite Restoratives AUJ Yap KEC Wee SH Teoh Clinical Relevance The clinical durability of Z100 and Surefil may be better than Ariston as cyclic temperature changes increase the surface hardness of Z100 and Surefil but decrease the hardness of Ariston. SUMMARY The clinical durability of some composite restorative materials may be significantly affected by cyclic temperature changes. This study investigated the effects of cyclic temperature changes on surface hardness of four commercial composite resins (Silux, Z100, Ariston and Surefil). Eighteen specimens of each material were divided into three treatment groups comprising a control and Department of Restorative Dentistry, Faculty of Dentistry, National University of Singapore, 5 Lower Kent Ridge Road, Singapore , Republic of Singapore Adrian UJ Yap, BDS, MSc, PhD, FAMS, FADM, FRSH, associate professor, Department of Restorative Dentistry, Faculty of Dentistry, assistant director, Center for Biomedical Materials Applications and Technology, Faculty of Engineering Kevin EC Wee, student, Faculty of Engineering SH Teoh, PhD, BEng, CPENG MIE (Aust), MASAIO, MIM, MASTM, associate professor, director, Center for Biomedical Materials Applications and Technology, Faculty of Engineering two different thermal cycling regimes. Control specimens were stored in distilled water at 35ºC for 178 hours. Thermal cycled specimens were stored in distilled water at 35ºC for 173 hours and subjected to five hours (300 cycles) of a thermal cycling regime consisting of the cycle ABAC, where A and B represent the fixed temperatures of 35ºC (28 seconds) and 15ºC (two seconds) and C, depending on the treatment group, either 45ºC or 60ºC (two seconds). All specimens were subsequently subjected to hardness testing (KHN) using a digital microhardness tester (load = 500 gf; dwell time = 15 seconds). Results were analyzed using ANOVA/Scheffe s test (p<0.05). The effect of thermal cycling on hardness was materialdependent. While thermal cycling significantly increased the surface hardness of Z100 and Surefil, it significantly decreased the hardness of Ariston. The hardness of Silux was not significantly affected by cyclic temperature changes. For all treatment groups, Z100 was significantly harder than the other composite resins evaluated and Surefil was significantly harder than Silux and Ariston. For both thermal cycled groups, Silux was significantly harder than Ariston.

28 26 Operative Dentistry INTRODUCTION Routine eating, drinking and breathing can produce changes in intra-oral temperatures (Crabtree & Atkinson, 1955; Boehm, 1972; Palmer, Barco & Billy, 1992). Cyclic temperature changes may be pathogenic in two ways. First, mechanical stresses generated by differences in coefficient of thermal expansion can result in bond failure at the tooth-restorative interface (Eakle, 1986; Crim & García-Godoy, 1987). Second, the changing gap dimensions are associated with gap volume changes that pump pathogenic oral fluids in and out of the gaps (Torstenson & Brannström, 1988). The physical properties of composite resins may also be affected by cyclic temperature changes or thermal cycling. Although thermal cycling had no effect upon the fracture toughness of composite resins (Mair & Vowles, 1989), it increased wear rate and decreased compressive strength (Mair, 1991; Chadwick, 1994). The effects of thermal cycling on surface hardness of composite restoratives is not widely reported in the literature (Hirabayashi & others, 1990). Hardness may be defined as the resistance of a material to indentation or penetration (O Brien, 1997a). It is, however, difficult to formulate a definition that is completely acceptable, as the indentation produced results from the interaction of numerous properties. Among the properties related to the hardness of a material are strength, proportional limit and ductility. Hardness has also been used to predict the wear resistance of a material and its ability to abrade or be abraded by opposing dental structures and materials (Anusavice, 1996). The variations in thermal cycling regimens used in laboratory research is considerable, making comparison of reports difficult. A clinically relevant regimen was derived by Gale & Darvell (1999) after assessing reports describing in vivo temperature changes of teeth and an analysis of 130 in vitro studies of thermal cycling of teeth. The regimen advocated was: 35ºC (28 seconds), 15ºC (two seconds), 35ºC (28 seconds) and 45ºC (two seconds) and has been suggested as a benchmark standard. This study investigated the effects of cyclic temperature changes on the hardness of four commercial composite restoratives based on the aforementioned regimen. The hardness of the different composite resins was also compared. METHODS AND MATERIALS Table 1 lists the composite restorative materials investigated. The composite resins were placed in the rectangular recesses (4 mm long, 3 mm wide and 2 mm deep) of customized acrylic molds and covered with acetate strips (Hawe-Neos Dental, Bioggio, Switzerland). A glass slide was placed over the acetate strip and pressure was applied to extrude excess material. The materials were then light cured according to manufacturers cure time through the glass slide/ acetate strip with a curing light (Spectrum; Dentsply Inc, Milford, DE 19963, USA). The mean intensity of the light source (467 ± 4.03 mw/cm 2 ) was determined with a radiometer (Cure Rite, EFOS Inc, Ontario, Canada). A total of 18 specimens were made for each material. The specimens were randomly divided into three treatment groups of six specimens each. Specimens in Group 1 (Control) were stored in distilled water at 35ºC for 178 hours. Group 2 (Cycled 45ºC) specimens were stored in distilled water at 35ºC for 173 hours and subjected to five hours (300 cycles) of thermal Table 1: Technical Profiles of the Composites Evaluated Material Manufacturer Type Polymer Fillers Filler Filler Cure Time Size Content (µm) (% by volume) Silux Plus 3M Dental Products, Microfill BisGMA Silica 0.04 (mean) 40 (Lot # ) St Paul, MN seconds TEGDMA Z100 3M Dental Products, Minifill BisGMA Zirconia (Lot # ) St Paul, MN seconds TEGDMA Silica (mean) Ariston phc Vivadent Midifill BisGMA Ba-Al- 1.3 (mean) 59 (Lot #A06719) Schaan, Liechtenstein 40 seconds UDMA Fluorosilicate glass TEGDMA Alkaline glass Silica Ytterbium Trifluoride Surefil Dentsply-Caulk Minifill Urethane- Ba-Boron- 0.8 (mean) 65 (Lot # ) Milford, DE seconds modified Fluoro- BisGMA silicate glass Silica BisGMA = Bisphenol-A-dimethacrylate TEDGMA = Triethylene glycol dimethacrylate UMDA = Urethane dimethacrylate Composite classification based upon that reported by Ferracane (1995).

29 Yap, Wee & Teoh: Effects of Cyclic Temperature Changes on Hardness of Composite Restoratives 27 Table 2: Mean KHN for the Different Treatment Groups Material Treatment Mean KHN SD Group Silux Group Group Group Z100 Group Group Group Ariston phc Group Group Group Surefil Group Group Group Group 1 = Control; Group 2 = Cycled 45ºC; Group 3 = Cycled 60ºC Mean KHN Group 1 Group 2 Group 3 Silux Z100 Ariston Surefil Materials Figure 1. Mean surface hardness (KHN) for the different treatment groups. cycling with an upper temperature of 45ºC. The specimens in Group 3 (Cycled 60ºC) were also stored in distilled water at 35ºC for 173 hours and subjected to five hours (300 cycles) of thermal cycling but with an upper temperature of 60ºC. One thermal cycle consisted of the cycle ABAC. The temperatures of the water in containers A and B were fixed at 35ºC and 15ºC, respectively. The temperature of the water in container C was either 45ºC or 60ºC depending on the treatment group. Dwell (immersion) time in container A was 28 seconds, while dwell time in containers B and C was two seconds. The total time of each thermal cycle was therefore one minute. After treatment, the specimens were blotted dry and subjected to hardness testing. The specimens were positioned centrally beneath the indenter of a digital microhardness tester (FM7, Future-Tech Corp, Tokyo, Japan) and a 500 g load was applied through the indenter with a dwell time of 15 seconds. Three readings were taken and averaged to form a single value for each specimen. Table 3: Results of Statistical Analysis Material Significance Materials Silux NS Z100 Group 3 > Groups 1 & 2 Ariston phc Group 1 > Groups 2 & 3 Surefil Groups 2 & 3 > Group 1 Treatment Group 1 Z100 > Silux, Ariston & Group Surefil Surefil > Silux & Ariston Group 2 Group 3 Group 1 = Control; Group 2 = Cycled 45ºC; Group 3 = Cycled 60ºC Z100 > Silux, Ariston & Surefil Surefil > Silux & Ariston Silux > Ariston Z100 > Silux, Ariston & Surefil Surefil > Silux & Ariston Silux > Ariston All statistical analysis were conducted at a significance level of Two-way analysis of variance (ANOVA) was performed on hardness data with composite material and treatment groups as main effects. Post-hoc Scheffe s test was used to test for differences among means (SPSS Inc, Chicago, IL 60611, USA). One-way ANOVA was performed for each material to determine the effects of cyclic temperature changes on material hardness. One-way ANOVA was also used to compare the hardness between composite resins after the different treatments. RESULTS The mean Koop hardness value (KHN) for the different treatment groups are shown in Table 2 and Figure 1. Results of statistical analysis are displayed in Table 3. Two-way ANOVA showed significant interaction between materials and treatment group. The effects of cyclic temperature changes on surface hardness was therefore material-dependent. The hardness of Silux was not significantly affected by cyclic temperature changes. For Z100, specimens thermal cycled at an upper temperature of 60ºC (Group 3) were significantly harder than specimens thermal cycled at an upper temperature of 45ºC and those not thermal cycled (Group 1). For Ariston, specimens in the control group (Group 1) were significantly harder than specimens that were thermal cycled (Groups 2 and 3). For Surefil, specimens that were thermal cycled (Groups 2 and 3) were significantly harder those in the control group (Group 1). Ranking of hardness was the same for all treatment groups and was as follows: Z100 > Surefil > Silux > Ariston. For all treatment groups, Z100 was significantly harder than Surefil, Silux and Ariston, and Surefil was significantly harder than Silux and

30 28 Operative Dentistry Ariston. Although no significant difference in hardness was observed between Silux and Ariston for the control group, Silux was significantly harder than Ariston for the thermal cycled groups (Groups 2 and 3). DISCUSSION Thermal cycling is commonly employed in tracer penetration, shear bond strength and tensile bond strength test of dental materials. Although the number of thermal cycles likely to be experienced clinically has yet to be determined, a provisional estimate of approximately 10,000 cycles per year was suggested. The number of thermal cycles previously used for in vitro tests ranges from 1 to 1,000,000 cycles, with a mean of about 10,000 and median of 500 cycles (Gale & Darvell, 1999). The latter was selected for this study. The thermal cycling regimen advocated by Gale & Darvell (1999) was: 35ºC (28 seconds), 15ºC (two seconds), 35ºC (28 seconds) and 45ºC (two seconds). An upper temperature of 60ºC was included as it was commonly employed in other studies investigating the effect of thermal cycling on physical properties (Mair & Vowles, 1989; Chadwick, 1994; Fujii & others, 1999). Although this upper temperature may be encountered in vivo, it is perceived as relatively hot and may cause discomfort (Plant, Jones & Darvell, 1974). The hardness of composite resins is related to the volume fraction of inorganic fillers and the quality of the resin matrix. An increase in filler volume and polymerization of resin leads to increased hardness (O Brien, 1997b). The surface hardness of composite restorative materials is, however, significantly affected by both water sorption and the contact time with the aqueous media (Hansen, 1983; Yap, Low & Ong, 2000). Water sorption results in swelling of the resin matrix and introduces radial tensile stresses at the filler-resin interfaces, thereby straining the Si-O-Si bonds in the fillers (Söderholm, 1983). The high energy levels arising from the strained Si-O-Si bonds makes the fillers more susceptible to stress corrosion attack, resulting in complete and partial debonding of the fillers at the surface layers and decreased surface hardness. For these reasons, the contact time with water (178 hours) was standardized for all treatment groups. The effects of cyclic temperature changes on surface hardness was found to be material-dependent. Thermal cycling with an upper temperature of 60ºC significantly increased the surface hardness of Z100 and Surefil. All changes can be attributed to the effects of thermal cycling, as the contact time with water was kept constant. Thermal cycling at an upper temperature of 60ºC may increase post-cure maturation of Z100 and Surefil (Shinkai & others, 1994; Harris, Jacobsen & O Doherty, 1999). Results concur with those of Kandil & others (1989), who found a marked increase in the mechanical properties of composite resins with increased temperature. Thermal cycling with an upper temperature of 60ºC, however, reduced the hardness of Ariston. This is most likely caused by differences in microstructure but may be attributed in part to variation in base monomer and activator/initiator concentrations. Ariston phc (ph controlled) uses a low oral ph to increase fluoride/other ionic release, and this episodic release prolongs the therapeutic usefulness of the material and optimizes fluoride release. For more effective fluoride and ionic release, the composite resin structure needs to be more open (Combe & Douglas, 1998). This open structure permits increased water diffusion, leading to possible increase in water sorption and solubility. The aforementioned may be accelerated/increased by thermal cycling resulting in decreased hardness observed. For Silux, any increase in hardness resulting from thermal cycling may be mitigated by water sorption due to the higher resin content (Table 1). The hardness of Surefil and Ariston were also significantly affected by thermal cycling at an upper temperature of 45 C. While thermal cycling at an upper temperature of 40 C increased hardness of Surefil, it decreased the hardness of Ariston. An increase in hardness of Z100 was also observed but the increase in KHN was not significant. Hirabayashi & others (1990) found that the hardness of resin composites decreased with an increasing number of thermal cycles between 0 and 30,000 cycles. The discrepancy in results between this and the present study can be accounted for by differences in the composite materials evaluated. As hardness is related to strength, proportional limit, ductility and wear resistance (Anusavice, 1996; Yap, Teoh & Tan, 2001a), the clinical durability of Z100 and Surefil is expected to be better than Ariston since cyclic temperature changes increase the surface hardness of Z100 and Surefil but decrease the hardness of Ariston. The ranking of hardness was the same for all treatment groups and was as follows: Z100 > Surefil > Silux > Ariston. The volume fraction of inorganic fillers of Z100, Surefil, Ariston and Silux were 66%, 65%, 59% and 40%, respectively. The theoretical hardness ranking based on filler volume should be Z100 > Surefil > Ariston > Silux, which was observed by Yap & others (2001b). The disparity in hardness ranking can be ascribed to the difference in water exposure time between current and the latter study as load, dwell time and instrumentation was identical. Microhardness testing was conducted after 24 hours by Yap & others (2001b) but after 178 hours in this study. The hardness of Ariston may therefore be severely reduced with increased water exposure time and warrants further investigation. The high release of methacrylic acid by Ariston could be a contributing factor (Yap, Lee & Sabapathy, 2000). For all treatment groups, Z100 was significantly harder than Surefil, Silux and Ariston and Surefil was significantly harder than Silux and Ariston.

31 Yap, Wee & Teoh: Effects of Cyclic Temperature Changes on Hardness of Composite Restoratives 29 Results were consistent with volume fraction of inorganic fillers. CONCLUSIONS Under the conditions of this in vitro study: 1. The effect of cyclic temperature changes on hardness was material dependent. 2. Hardness of Z100 was significantly increased by thermal cycling at an upper temperature of 60ºC. 3. Hardness of Surefil was significantly increased by thermal cycling at upper temperatures of 45ºC and 60ºC. 4. Hardness of Ariston was significantly decreased by thermal cycling at upper temperatures of 45ºC and 60ºC. 5. Hardness of Silux was not significantly affected by thermal cycling. 6. For all treatment groups, Z100 was significantly harder than Surefil, Silux and Ariston and Surefil was significantly harder than Silux and Ariston. (Received 18 February 2001) References Anusavice KL (1996) Mechanical properties of dental materials in Phillip Science of Dental Materials 10 th edition Philadelphia WB Saunders Co p 69. Boehm RF (1972) Thermal environment of teeth during open mouth respiration Journal of Dental Research 51(1) Chadwick RG (1994) Thermocycling the effects upon the compressive strength and abrasion resistance of three composite resins Journal of Oral Rehabilitation 21(5) Combe EC & Douglas WH (1998) The future of dental materials Dental Update 25(9) Crabtree MG & Atkinson HF (1955) Preliminary report on the solubility of decalcified dentine in water Australian Journal of Dentistry Crim GA & García-Godoy F (1987) Microleakage: The effect of storage and cycling duration Journal of Prosthetic Dentistry 57(5) Eakle WS (1986) Effect of thermal cycling on fracture strength and microleakage in teeth restored with a bonded composite resin Dental Materials 2(3) Ferracane JL (1995) Current trends in dental composites Critical Review in Oral Biology and Medicine 6(4) Fujii K, Miura K, Omori K, Arikawa H, Kanie T & Inoue K (1999) Effects of thermal cycling on dynamic viscoelastic properties of four commercial resins for crown and bridge Dental Materials Journal 18(4) Gale MS & Darvell BW (1999) Thermal cycling procedures for laboratory testing of dental restoratives Journal of Dentistry 27(2) Hansen EK (1983) After-polymerization of visible light activated resins: Surface hardness vs light source Scandinavian Journal of Dental Research 91(5) Harris JS, Jacobsen PH & O Doherty DM (1999) The effect of curing light intensity and test temperature on the dynamic properties of two polymer composite Journal of Oral Rehabilitation 26(8) Hirabayashi S, Nomoto R, Harashima I & Hirasawa T (1990) The surface degradation of various light-cured composite resins by thermal cycling Shika Zairyo Kikai 9(1) Kandil SH, Kamar AA, Shaaban SA, Taymour NM & Morsi SE (1989) Effect of temperature and aging on the mechanical properties of dental polymeric composite materials Biomaterials 10(8) Mair LH & Vowles R (1989) The effect of thermal cycling on the fracture toughness of seven composite restorative materials Dental Materials 5(1) Mair LH (1991) Effect of surface conditioning on the abrasion rate of dental composites Journal of Dentistry 19(2) O Brien WJ (1997a) Physical properties in Dental Materials and Their Selection Illinois Quintessence Publishing Company p 18. O Brien WJ (1997b) Polymeric Restorative Materials in Dental Materials and their selection Illinois Quintessence Publishing Company p Palmer DS, Barco MT & Billy EJ (1992) Temperature extremes produced orally by hot and cold fluids Journal of Prosthetic Dentistry 67(3) Plant CG, Jones DW & Darvell BW (1974) The heat evolved and temperatures attained during setting of restorative materials British Dental Journal 137(6) Söderholm KJ (1983) Leaking of fillers in dental composites Journal of Dental Research 62(2) Shinkai K, Suzuki S, Leinfelder KF & Katoh Y (1994) How heat treatment and thermal cycling affect wear of composite resin inlays Journal of the American Dental Association 125(11) Torstenson B & Brannström M (1988) Contraction gap under composite resin restorations: Effects of hygroscopic expansion and thermal stress Operative Dentistry 13(1) Yap AUJ, Lee HK & Sabapathy R (2000) Release of methacrylic acid from dental composite Dental Materials 16(3) Yap AUJ, Low JS & Ong LFKL (2000) Effect of food-simulating liquids on surface characteristics of composite and polyacidmodified composite restoratives Operative Dentistry 25(3) Yap AUJ, Teoh SH & Tan KB (2001a) Three-body abrasive wear of composite restoratives Operative Dentistry 26(2) Yap AUJ, Chew CL, Teoh SH & Ong LFKL (2001b) Influence of contact stress on OCA wear of composite restoratives Operative Dentistry 26(2)

32 Operative Dentistry, 2002, 27, Microhardness of Resin Composites Polymerized by Plasma Arc or Conventional Visible Light Curing SH Park I Krejci F Lutz Clinical Relevance A newly developed plasma arc curing unit, Apollo 95E, does not properly cure composite when the layer thickness exceeds 2 mm and requires longer curing time than is recommended by the manufacturer. To take advantage of the newly developed unit, 12 seconds of light curing is recommended for shallow cavities not exceeding 2 mm. SUMMARY This study evaluated the effectiveness of the plasma arc curing (PAC) unit for composite curing. To compare its effectiveness with conventional quartz tungsten halogen (QTH) light curing units, the microhardness of two composites (Z100 and Tetric Ceram) that had been light cured by the PAC or QTH units, were compared according to the depth from the composite surface. In addition, linear polymerization shrinkage was compared using a custom-made linometer between composites which were light cured by PAC or QTH units. Department of Conservative Dentistry, NHIC Ilsan Hospital, 1232, Baeksok-dong, Ilsan-gu, Koyang-shi, Kyonggi-do, , Korea (South) Sung-Ho Park, DDS, PhD, associate professor, Department of Conservative Dentistry, Yonsei University, Seoul, Korea I Krejci, professor dr, Department of Preventive and Restorative Dentistry, University of Geneva, Geneva, Switzerland F Lutz, professor dr, Department of Preventive Dentistry, Periodontology, and Cariology, Zurich University, Zurich, Switzerland Measuring polymerization shrinkage for two resin composites (Z100 and Tetric Ceram) was performed after polymerization with either QTH or PAC units. In the case of curing with the PAC unit, the composite was light cured with Apollo 95E for two (Group 1), three (Group 2), six (Group 3) or 2 x 6 (Group 4) seconds. For light curing with the QTH unit, the composite was light cured for 60 seconds with Optilux 500 (Group 5). The linear polymerization shrinkage of composites was determined in the linometer. Two resin composites were used to measure microhardness. Two-mm thick samples were light cured for three seconds (Group 1), six seconds (Group 2) or 12 (2 x 6) seconds (Group 3) with Apollo 95E or they were conventionally light cured with Optilux 500 for 30 seconds (Group 4) or 60 seconds (Group 5). For 3 mm thick samples, the composites were light cured for six seconds (Group 1), 12 (2 x 6) seconds (Group 2) or 18 (3 x 6) seconds (Group 3) with Apollo 95E or they were conventionally light cured with Optilux 500 for 30 seconds (Group 4) or 60 seconds (Group 5). Twenty samples were assigned to each group. The microhardness of the upper and lower surfaces was measured with a Vickers hardness-measuring instrument under load. The difference in microhardness between the upper and lower surfaces in each group was analyzed by paired t-test. For the

33 Park, Krejci & Lutz: Microhardness of Resin Composites Polymerized by Plasma Arc or Conventional VLC 31 upper or lower surfaces, one-way ANOVA with Tukey was used. For Tetric Ceram, the amount of polymerization shrinkage was lower when cured with the Apollo 95E for two or three seconds than when cured for six and 12 (2 x 6) seconds, or for 60 seconds with Optilux 500 (p<0.05). For Z100, the amount of linear polymerization shrinkage was lower when cured with the Apollo 95E for two, three and six seconds than for 12 (2 x 6) seconds with Apollo 95E or for 60 seconds with the Optilux 500 (p<0.05). The results of the microhardness test indicated that there was no statistically significant difference in microhardness between groups for the upper surface. However, for the lower surface, when the composites were light cured with Apollo 95E for three seconds as recommended by the manufacturer, microhardness of the lower surface was usually lower than that of the upper surface and did not cure sufficiently. Conclusively, when compared with conventional QTH unit, the PAC unit, Apollo 95E did not properly cure the lower composite surface when the layer thickness exceeded 2 mm. In addition, three seconds of curing time, which the manufacturer recommended, was insufficient for optimal curing of composites. INTRODUCTION Adequate polymerization of a resin composite is considered to be a very important factor in obtaining adequate physical (Asmussen, 1982) and biological properties (Caughman & others, 1991). It has been reported that the power density required for effective polymerization of a composite should be more than 280mW/cm 2 (Rueggeberg, Caughman & Curtis, 1994). To effectively polymerize a composite, light curing units have been developed with improved power density. Currently, widely used quartz tungsten halogen (QTH) units have power densities that range between 400 and 900mW/cm 2. When the composites are cured with QTH unit, 60-second cure time of resin increments not greater than 2 mm have been recommended (Rueggeberg & others, 1994). In addition, argon laser curing units, which have a more consistent light output over distance, have been used to cure the composites (Blankenau & others, 1991a,b; Kelsey & others, 1989). Recently, a new type of light curing system, Plasma arc curing (PAC) unit has been introduced. It uses a highfrequency electrical field to generate its plasma energy and matter is thereby transformed into a mixture of ions, electrons and molecules. The significant amount of energy released during this process is used for curing photosensitive composites. Compared with conventional QTH units, the PAC unit emits lights of higher power density (1370 mw/cm 2 ). The wavelength of emitted light of PAC is around 470 nm, whereas in a conventional QTH unit, it is between 400 and 520 nm. The manufacturers assert that PAC units cure composite material in one-to-three seconds and also decrease the amount of polymerization shrinkage. However, reduced polymerization shrinkage in one-to-three seconds may be related to improper cure. In recent publications, some researchers have indicated the possibility of improper cure when composites were light cured with PAC units (Peutzfeldt, Sahafi & Asmussen, 2000; Stritikus & Owens, 2000; Hofmann & others, 2000). Since most dentists are familiar with the QTH visible light curing technique, comparing the effectiveness of the newly developed PAC unit with a conventional QTH unit would provide them with much clinical information. In resin composite, the physical properties are closely related to the degree of conversion, and hardness measurement is an effective way to evaluate the degree of cure (Rueggeberg & Craig, 1988). The polymerization shrinkage of resin composite is easily and accurately measured by a specially designed linometer (de Gee, Feilzer & Davidson, 1993; Park, Krejci & Lutz, 1999). This study evaluated the effectiveness of the PAC unit for composite curing. To compare its effectiveness with conventional QTH light curing units, the microhardness of two composites (Z100 and Tetric Ceram) that had been light cured by the PAC unit (Apollo 95E, Dental/Medical Diagnostic Systems Inc, Clovis, CA 93612, USA) or QTH unit (Optilux 500, Demetron/Kerr, Danbury, CT 06810, USA) were compared according to the depth from the composite surface. In addition, polymerization shrinkage between composites that were light cured by PAC or QTH unit was compared using a custom-made linometer. METHODS AND MATERIALS Measurement of the Polymerization Shrinkage Z100 (A2 shade, Lot # , 3M Dental Products, St Paul, MN 55144, USA) and Tetric Ceram (A2 shade, Lot #913072, Vivadent AG, Schaan, Liechtenstein) were used in this study. The composites, expressed from a syringe, were transferred to a Teflon mold to ensure the same amount of composite for each linometer sample. After a slight amount of composite was added or subtracted in the mold, the composite was transferred to the disk in the custom-made linometer that had been previously coated with a separating glycerin gel (Airblock, De Trey Division, Dentsply Ltd, Weybridge, Surrey, England). The resin composite was then covered with a slide glass and an aluminum shield. The surface of the slide glass facing the composite had also been coated with separating gel. The shield was then covered and fastened under constant pressure to the base metal

34 32 Operative Dentistry Figure 2. Change in the amount of linear polymerization shrinkage versus time in Z100 that was light cured with Apollo 95E or Optilux 500. Figure 1. Schematic drawing of the linometer with a composite sample in place. of the mold. The position of the disk was adjusted to its zero position with the height adjustment screw. The composites were light cured with Optilux 500 or Apollo 95E. The tip of the curing light was positioned as close as possible to the aluminum shield surface and fixed into position with screws (Figure 1). In the case of light curing with Apollo 95E, the composite was light cured for 2 seconds (Group 1), three seconds (Group 2), six seconds (Group 3) or 12 (2 x 6) seconds (Group 4). When the composites were light cured for 12 seconds, two seconds of time elapse was needed after six seconds of light activation due to the inherent properties of the light curing unit. In the case of conventional light curing, the composite was light cured for 60 seconds with the Optilux 500 (Group 5), the power density of which was determined at 900 mw/cm 2 by the installed radiometer. As the composite under the slide glass was cured, it shrank toward the light source and the aluminum disk under the composite moved upward. The amount of displacement of the disk, which resulted from linear shrinkage of the resin composite, was measured with a custom-made infrared micrometer. The analog signal was converted into digital data by an A/D converter and recorded in a computer every second for 60 seconds using the software program Excel 5.0 (Figure 1). Ten measurements were made for each group. For each composite, the amount of linear shrinkage from each group was calculated to obtain a mean and compared by ANOVA using Tukey s test as a post-hoc test at the confidence level of 95%. The linear shrinkage value was converted to volumetric shrinkage according to the method used by de Gee & others (1993) as follows: The percentage of linear polymerization shrinkage (lin%) was calculated by: lin% = L/(L + L) X 100 L: Recorded displacement L: Thickness of the sample after polymerization The thickness of the light cured samples were measured to 0.01 mm. The percentage of volumetric shrinkage (vol%) derived from the linear shrinkage was given by: Vol % = 3lin% (lin%) (lin%) 3 For each composite, the amount of linear polymerization shrinkage in the first 10 seconds from each group was compared with one way ANOVA with Tukey at 95% levels of confidence. Measurement of the Microhardness Cylinders with a diameter of 6 mm and a thickness of 2 mm and 3 mm were made in aluminum plates covered Table 1: Curing Methods for Hardness Measurements Sample Thickness Groups Curing Methods 2 mm 1 3 seconds, Apollo95E 2 6 seconds, Apollo95E 3 12 seconds, Apollo95E 4 30 seconds, VLC 5 60 seconds, VLC 3 mm 1 6 seconds, Apollo95E 2 12 seconds, Apollo95E 3 18 seconds, Apollo95E 4 30 seconds, VLC 5 60 seconds, VLC When the composites were light cured for 12 and 18 seconds with Apollo 95E, two seconds of time elapse was needed after every six seconds of light activation.

35 Park, Krejci & Lutz: Microhardness of Resin Composites Polymerized by Plasma Arc or Conventional VLC 33 Figure 3. Change in the amount of linear polymerization shrinkage versus time in Tetric Ceram that was light cured with Apollo 95E or Optilux 500. with glass slides placed on the underside of each hole of the mold. Titanium-coated instruments (Composite Instrument, Coltene, Switzerland) were used to place the resin composites into the mold. Glass slides were placed on top of the resin composite and pressed flat. The composites were light cured with the conventional visible light curing unit or Apollo 95E. For the 2 mm samples, the composites were light cured for three seconds (Group 1), six seconds (Group 2) or 12 (2 x 6) seconds (Group 3) with Apollo 95E or for the conventionally light cured sample for 30 seconds (Group 4) or 60 seconds (Group 5). For the 3 mm samples, the composites were light cured for six seconds (Group 1), 12 (2 x 6) seconds (Group 2) or 18 (3 x 6) seconds (Group 3) with Apollo 95E, or conventionally light cured for 30 seconds (Group 4) or 60 seconds (Group 5) (Table 1). After the composites were light cured with the visible light curing unit or Apollo 95E, the specimens were removed from the mold and the upper surface, (closer to the light source) and the lower surfaces were marked with a pen. Twenty samples were assigned to each group. The samples were then stored in the dark in 100% humidity at 37 C for seven days. The microhardness of the upper and lower surfaces were measured with a Vickers hardness-measuring instrument (Optidur, Göttfert, Feinwerktechnik GmbH, Germany). The difference in microhardness between the upper and lower surfaces in each group was analyzed by paired t-tests. For the upper or lower surfaces, one-way ANOVA with Tukey test were used to determine significant differences in hardness among the groups. Figure 4. Comparison of change in the amount of linear polymerization shrinkage versus time in Z100 and Tetric Ceram that was light cured with Apollo 95E or Optilux 500. Table 2: Linear Shrinkage, Converted % Linear and % Shrinkage Value Groups Linear Shrinkage %lin %vol Value (µm) (0.82) (0.92) Z (0.84) (0.74) (0.69) (1.01) (0.95) Tetric Ceram (0.87) (0.85) (0.83) The values in the parentheses are SDs (n=10). Groups joined by vertical lines were not significantly different at p=0.05) level. lin%: The percentage of linear polymerization shrinkage. vol%: The percentage of volumetric shrinkage. RESULTS Polymerization Shrinkage The average thickness of the samples was 1.60 ± 0.06 mm. Table 2 lists the converted linear % shrinkage and % volumetric shrinkage. For Z100, the amount of linear polymerization shrinkage was lower when it was light cured with Apollo 95E for two seconds, three seconds and six seconds compared with 12 seconds for Apollo 95E, or 60 seconds with the Optilux 500 (p<0.05) (Table 2, Figure 2). Z100 showed more rapid polymerization in the first 10 seconds when cured with Optilux 500 than when cured with Apollo 95E (Table 3, Figure 2). For Tetric Ceram, the amount of polymerization shrinkage was lower when it was light cured with Apollo 95E for two seconds or three seconds than when cured for six seconds and 12 seconds, or for 60 seconds with Optilux 500 (p<0.05) (Table 2, Figure 3).

36 34 Operative Dentistry Table 3: Amount of Linear Polymerization Shrinkage (µm) at 10 seconds of Light Curing Z100 Tetric Ceram Group (0.16) a 4.50 (0.15) a (0.31) b 5.92 (0.28) b (0.39) c 7.17 (0.25) d (0.29) c 7.25 (0.30) d (0.30) d 6.47 (0.30) c The values are amount of linear polymerization shrinkage with SDs in parentheses (n=10). The letters in the back of the value represent difference in linear shrinkage between groups at p=0.05 level. Table 4: Microhardness of 2 mm Samples Groups Upper Surface Lower Surface (55.5) (43.7) * (48.3) (50.2) * Z (40.2) (56.3) * (60.3) (62.8) Tetric Ceram, which was light cured with Apollo 95E, showed a more rapid polymerization shrinkage rate in the first 10 seconds than when it was cured with Optilux 500 (Table 3, Figure 3). In comparison, Z100 showed a more rapid polymerization shrinkage rate than Tetric Ceram (Figure 4). B. Microhardness (50.6) (69.4) (40.3) (34.5) * Tetric (43.2) (36.3) * Ceram (36.8) (33.7) (44.5) (34.6) (39.7) (38.7) Values are microhardness with SDs in parentheses (n=20). * Significant difference in microhardness between upper and lower surface at the 95% level of confidence. Groups joined by vertical lines were not significantly different at p=0.05 level. Table 5: Microhardness of 3 mm Samples Groups Upper Surface Lower Surface (51.2) (37.7) * (53.4) (32.8) * Z (55.7) (40.2) * (33.3) (40.2) (45.8) (42.8) (40.6) (33.5) * Tetric (36.8) (36.7) * Ceram (43.6) (44.2) * (44.1) (41.3) (36.7) (40.7) Values are microhardness with SDs in parentheses (n=20). * Significant difference in microhardness between upper and lower surface at the 95% level of confidence. Groups joined by vertical lines were not significantly different at p=0.05 level. In the 2 mm Z100 samples, there was no difference in microhardness between the upper surfaces of each group. There was no difference in microhardness between the upper and lower surfaces in Groups 4 and 5, whereas the microhardness of the lower surface was less when compared to the upper surface in Groups 1, 2 and 3. In Groups 1, 2 and 3, the microhardness of the lower surface increased as the exposure time increased for three seconds and six-to-12 seconds (Table 4). In the 3 mm Z100 samples, there was no difference in microhardness between the upper surfaces of each group. In Groups 4 and 5, there was no difference in microhardness between the upper and lower surfaces. On the other hand, the microhardness of lower surfaces was less compared to the upper surfaces in Groups 1, 2 and 3. In these groups, the microhardness of the lower surface increased as the exposure time increased from six seconds, 12 seconds and 18 seconds (Table 5). In the 2 mm Tetric Ceram samples, there was no difference in microhardness between the upper surfaces of each group. On the lower surface, however, the microhardness in Groups 3, 4 and 5 was higher than in Groups 1 and 2. In Groups 3, 4 and 5, there was no difference in microhardness between the upper and lower surfaces, whereas the microhardness of the lower surface was less than the upper surface in Groups 1 and 2 (Table 4). In the 3 mm Tetric Ceram samples, there was no difference in microhardness between the upper surfaces of each group. In Groups 4 and 5, there was no difference in microhardness between upper and lower surfaces. On the other hands, the microhardness of the lower surface was less than the upper surface in Groups 1, 2 and 3 (Table 5). DISCUSSION For the polymerization shrinkage test, the amount of polymerization shrinkage in Tetric Ceram was lower when the composite was cured for two or three seconds with Apollo 95E than when cured for six seconds or 12 seconds with Apollo 95E or for 60 seconds with visible light. For Z100 the amount of linear polymerization shrinkage was lower when light cured with Apollo 95E for two seconds, three seconds and six seconds than when cured for 12 seconds with Apollo 95E or for 60 seconds with Optilux 500. Polymerization shrinkage is closely related with the degree of conversion for a composite. Thus, it can be concluded that the composites showed low polymerization shrinkage in two and three seconds of light curing with Apollo 95E because it did not cure sufficiently. The results of shrinkage measurement in this study would indicate that curing time for 1.6 mm-thick com-

37 Park, Krejci & Lutz: Microhardness of Resin Composites Polymerized by Plasma Arc or Conventional VLC 35 posites with Apollo 95E should be extended to six seconds in Tetric Ceram and to 12 seconds in Z100 even though the manufacturer of Apollo 95E claims that it cures composite material in one-to-three seconds. In this study, polymerization shrinkage was measured for only 60 seconds. One drawback of linometer is that it cannot measure the composite shrinkage value if the measuring time is extended for hours because the separation between disk and composite would occur with time. The absolute shrinkage value would be more accurate if the shrinkage measure could be extended for hours. However, the greatest degree of increase in hardness was observed in the first few minutes (Hansen, 1983) and the polymerization shrinkage rate of light cured composite was greatest during the first seconds (Sakaguchi & others, 1992). Thus, in a comparative study, recording the shrinkage value during the first 60-to-90 seconds mirrored the relative long-term shrinkage values (Park & others, 1999). As this was a comparative study, it was designed to record the shrinkage value for 60 seconds. The use of PAC unit with higher power density may cause more rapid development of polymerization contraction force. However, considering the results of this study, the high power density in Apollo 95E was not always related to rapid polymerization. Tetric Ceram when light cured with Apollo 95E for more than six seconds showed slightly more rapid polymerization shrinkage in the first 10 seconds than when cured with Optilux 500, whereas Z100 showed more rapid polymerization when cured with Optilux 500 (Table 3, Figures 2, 3 and 4). According to the manufacturer, the wavelength of emitted light of Apollo 95E is around 470 nm, whereas it is between nm in a conventional QTH unit. For composites containing both camphorquinone (CQ) and photo initiators absorbing at shorter wavelength, the QTH unit may cure them more efficiently than the PAC unit. As the initiator system of Z100 is CQ only, another factor seems to contribute to the more efficient polymerization of Z100 in QTH unit. One factor that could be considered is a new initiator-activator system in Z100. In Z100, a three-component initiation system (camphorquinone, tertiary amine and Iodonium salt) was introduced for initiation and activation of the composite (US Patent #5,545,676). According to Dr Oxman, JD (3M Dental Products), Iodonium salt plays an important role in increasing curing efficiency. The new initiator-activator system in Z100 may be more efficient for Optilux than for Apollo 95E. Another factor that may affect the results is the power density of QTH unit. The power density of Optilux 500 in this study was 900mW/cm 2. In a study which has been ongoing in this author s laboratory but has not yet been published, Z100 cured with the QTH unit of low power density (400mW/cm 2 ) showed slower polymerization than when cured with the PAC unit. Thus, in this study, more rapid polymerization of Z100 with a QTH unit may be related to the 900mW/cm 2 power density of Optilux 500. It could be hypothesized that the new initiator-activator system in Z100 and the 900mW/cm 2 power density of Optilux 500 have a synergistic effect and result in more efficient polymerization in the QTH unit. A linear relationship between light intensity and polymerization contraction has been demonstrated (Sakaguchi & others, 1992). The contraction rate of light-cured composite is highest during the first seconds of the polymerization reaction (Sakaguchi & others, 1992). This is clinically important because the integrity of tooth-composite interface is rapidly challenged during the early phase of polymerization when the bond between the hard tissue and composite is still maturing. It has been pointed out that composite cured at lower power density has a better marginal adaptation (Uno & Asmussen, 1991a). However, this procedure leads to inferior material properties (Uno & Asmussen, 1991a). Another way to minimize the wallto-wall contraction is to allow the flow of resin composite during setting by means of controlled polymerization. This can be done by pre-polymerization at low power density followed by final cure at high power density (Mehl, Hickel & Kunzelmann, 1997). The reduced rate of polymerization may allow for increased flow of the material, decreasing the polymerization shrinkage stress in a restoration (Uno & Asmussen, 1991b). This may produce a more favorable marginal integrity (Uno & Asmussen, 1991a; Feilzer & others, 1995; Mehl & others, 1997). The use of a PAC unit with higher power density may cause more rapid development of polymerization contraction force. However, there was no difference in marginal gap between the PAC unit and the QTH unit (Peutzfeldt & others, 2000). Peutzfeldt & others indicated the possibility that a low degree of conversion was compensated for by the rapid cure resulted in gaps that were not different from those obtained with a conventional QTH unit. Based on the results of this study, it is likely that composites do not cure enough if the thickness layer exceeds 2 mm or the composites are cured for one-to-three seconds as recommended by the manufacturer. The higher power intensity of Apollo 95E did not increase the microhardness of the upper composite surface even though the manufacturer claims it does. When the composites were cured with Apollo 95E on the upper surface, there was no difference in microhardness among the groups. For the 3 mm samples, the microhardness of the lower surface was less than that of the upper surface even though the light activation time was extended to 18 seconds in Z100 and Tetric Ceram. In contrast, 30 seconds of light curing with Optilux 500 cured the lower surface to the same

38 36 Operative Dentistry degree as that of the upper surface. It was indicated that the duration of exposure will allow the excited CQ molecule to diffuse and react with the amine to help initiate polymerization, and the duration of exposure becomes important, especially when the power density is not the rate-limiting step in polymerization (Rueggeberg, Caughman & Curtis, 1993). Relative short curing time in a PAC unit may be insufficient for diffusion of the CQ molecule to the deeper portion; it may then limit the reaction with amine. The second factor to be considered is energy density. The energy density of Optilux 500 at 30 seconds in this study was 27 J/cm 2 (0.9 W/cm 2 X 30 seconds). According to the manufacturer, the power density of Apollo 95E is about 1370 mw/cm 2. Thus, the energy density of Apollo 95E at 18 seconds was 24.7 J/cm 2. Considering the energy density of Optilux 500 at 30 seconds, the curing time of Apollo 95 E should be extended to cure the 3 mm composite samples. The wavelength of emitted light is another important factor in determining the efficiency of the light source in composite curing. The wavelength of emitted light is nm for conventional light sources and nm for Apollo 95E. It is possible that the emitted light from Optilux 500, which has a broader spectrum than Apollo 95E, may enhance the cure of the composites tested. According to Rueggeberg & Craig (1988), the Knoop hardness value of composite under the cured composite overlay decreased significantly as the overlay thickness increased. In addition, the light transmission through the cured overlay decreased significantly as the overlay thickness increased. In this study, Apollo 95E cured the composite surface very rapidly. The rapidly cured upper surface might act as a precured overlay, and hence block light transmission through it. This hypothesis can be applied to the study by Vargas, Cobb & Smith (1998) in which resin composite polymerization was compared as indicated by microhardness at increasing depth using an argon laser versus a conventional QTH unit. In their study, the laser effectively reduced the polymerization time to a depth of 2 mm. However, in their study they also showed a higher energy density did not correlate with a significantly greater hardness in the deeper portions of samples. That is, the hardness of the composites that were light cured with conventional light curing methods was greater in the deeper portions than that of composites cured with an argon laser even though the energy density was higher for the laser cured composites. This hypothesis needs further research. On the lower surface in the 2 mm Tetric Ceram samples, when the light activation with Apollo 95E was extended to 12 seconds, it reached the same microhardness as the visible light cured composite which was light cured for 30 or 60 seconds, whereas 12 seconds was not enough for 2 mm of Z100. In the polymerization shrinkage test, Z100 showed more rapid polymerization shrinkage than Tetric Ceram. Considering the hypothesis previously mentioned, it might be more difficult for Z100, which showed more rapid polymerization to cure to a greater depth than Tetric Ceram. This hypothesis needs further testing. As mentioned previously, the new initiator-activator system in Z100 may be more efficient for conventional light curing units. In this study, aluminum was used as a mold. As the tooth is more transparent than aluminum, the tooth may pass more light energy in the deeper portion. Further study in the cavity of human teeth would be desirable to confirm the results of this study. Considering the results of this study, the newly developed composite curing unit that uses plasma energy could effectively cure the composite at a reduced light activation time. However, the time required to achieve adequate curing was greater than that claimed by the manufacturer. In addition, the curing efficiency was reduced when the sample thickness exceeded 2 mm. To take advantage of this newly developed unit, 12 seconds of light curing to a depth of 2 mm would seem necessary to adequately cure samples. CONCLUSIONS 1. When the composites were light cured with a PAC unit (Apollo 95E) for two or three seconds, it showed a lower amount of linear polymerization shrinkage than when cured for 12 (2 x 6) seconds or for 60 seconds with a QTH unit (Optilux 500). 2. The PAC unit did not always induce a rapid polymerization shrinkage rate compared to the QTH unit. 3. For 2 mm composite samples, 12 seconds of curing with Apollo 95E was needed to properly cure the lower surface of Tetric Ceram. For Z100, the lower surface was not cured properly even though the curing time was extended to 12 seconds. 4. For 3 mm composite samples, the lower surfaces of Z100 and Tetric Ceram were not cured properly even though the curing time was extended to 18 seconds, whereas 30 seconds of light curing of Optilux 500 properly cured it. Acknowledgements This work was supported by the Brain Korea 21 Project. The authors most gratefully acknowledge the help of Dr Burrow MF of the University of Melbourne for English language assistance. (Received 25 February 2001)

39 Park, Krejci & Lutz: Microhardness of Resin Composites Polymerized by Plasma Arc or Conventional VLC 37 References Asmussen E (1982) Restorative resins: Hardness and strength vs quantity of remaining double bonds Scandinavian Journal Dental Research 90(6) Blankenau RJ, Powell GL, Kelsey WP & Barkmeier WW (1991) Post-polymerization strength values of an argon laser cured resin Lasers in Surgery Medicine 11(5) Blankenau RJ, Kelsey WP, Powell GL, Shearer GO, Barkmeier WW & Cavel WT (1991) Degree of composite resin polymerization with visible light and argon laser American Journal of Dentistry 4(1) Caughman WF, Caughman GB, Shiflett RA, Rueggeberg F & Schuster GS (1991) Correlation of cytotoxicity, filler loading and curing time of dental composites Biomaterials 12(8) de Gee AJ, Feilzer AJ & Davidson CL (1993) The linear polymerization shrinkage of unfilled resins and composites determined with a linometer Dental Materials 9(1) Feilzer AJ, Dooren LH, de Gee AJ & Davidson CL (1995) Influence of light intensity on polymerization shrinkage and integrity of restoration-cavity interface European Journal of Oral Science 103(5) Hansen EK (1983) After-polymerization of visible light activated resins: Surface hardness vs light source Scandinavian Journal of Dental Research 91(5) Hofmann N, Hugo B, Schubert K & Klaiber B (2000) Comparison between a plasma arc light source and conventional halogen curing units regarding flexural strength, modulus, and hardness of photo-activated resin composites Clinical Oral Investigation 4(3) Kelsey WP 3d, Blankenau RJ, Powell GL, Barkmeier WW, Cavel WT & Whisenant BK (1989) Enhancement of physical properties of resin restorative materials by laser polymerization Lasers in Surgery Medicine 9(6) Mehl A, Hickel R & Kunzelmann KH (1997) Physical properties and gap formation of light-cured composites with and without softstart polymerization Journal of Dentistry 25(3-4) Park SH, Krejci I & Lutz F (1999) Consistency in the amount of linear polymerization shrinkage in syringe-type composites Dental Materials 15(6) Peutzfeldt A, Sahafi A & Asmussen E (2000) Characterization of resin composites polymerized with plasma arc curing units Dental Materials 16(5) Rueggeberg FA & Craig RG (1988) Correlation of parameters used to estimate monomer conversion in light-curing composite Journal of Dental Research 67(6) Rueggeberg FA, Caughman WF, Curtis JW Jr & Davis HC (1993) Factors affecting cure at depths within light-activated resin composites American Journal of Dentistry 6(2) Rueggeberg FA, Caughman WF & Curtis JW Jr (1994) Effect of light intensity and exposure duration on cure of resin composite Operative Dentistry 19(1) Sakaguchi RL, Peter MC, Nelson SR, Douglas WH & Poort HW (1992) Effects of polymerization contraction in composite restorations Journal of Dentistry 20(3) Stritikus J & Owens B (2000) An in vitro study of microleakage of occlusal composite restorations polymerized by a conventional curing light and PAC curing light Journal of Clinical Pediatric Dentistry 24(3) Uno S & Asmussen E (1991a) Marginal adaptation of a restorative resin polymerized at reduced rate Scandinavian Journal of Dental Research 99(5) Uno S & Asmussen E (1991b) Effect on bonding of curing through dentin Acta Odontologica Scandinavica 49(5) Vargas MA, Cobb DS & Schmit JL (1998) Polymerization of composite resins: Argon laser vs conventional light Operative Dentistry 23(2)

40 Operative Dentistry, 2002, 27, Effect of Collagen Removal on Microleakage of Resin Composite Restorations V de PA Saboia LAF Pimenta GMB Ambrosano Clinical Relevance The use of sodium hypochlorite to remove collagen may decrease microleakage at the resin/dentin interface when an acetone-based adhesive system is used. SUMMARY This study evaluated the effects of collagen removal on the microleakage of two single-bottle adhesive systems. Forty human third molars were selected and each received two root preparations. The roots were randomly assigned for restoration using Prime & Bond 2.1 (Dentsply Ltda, Petrópolis, RJ 90915, Brazil) or Single Bond (3M Dental Products, St Paul, MN 55144, USA). One root in each tooth was treated with 36% H 3 PO 4 for 15 seconds and the other received an additional treatment with 10% NaOCl for 60 seconds to remove the collagen layer before adhesive was applied. All preparations were restored with Z100 restorative resin (3M Dental Products). The specimens were submitted to 5,000 thermal Dentistry School of Piracicaba UNICAMP Av Limeira, 901 Piracicaba SP Brazil Vicente de Paulo Aragão Saboia, DDS, MS, PhD, associate professor of Restorative Dentistry, School of Dentistry, Federal University of Ceará, UFC, Brazil Luiz André Freire Pimenta, DDS, MS, PhD, associate professor of Restorative Dentistry Glaucia Maria Boni Ambrosano, Agr Eng, MS, PhD, assistant professor of Biostatistics cycles (5-55 C) and stored in 37 C distilled water for one year. The specimens were then coated with a varnish except for 1 mm of tooth structure surrounding the restoration and immersed in 2% buffered methylene blue for four hours. After rinsing, the restorations were sectioned and two independent observers scored the microleakage at the interface between the restorative material and the tooth using an optical microscope at x45 magnification. The scores were submitted to Fisher s Exact Test and the results showed that collagen removal significantly reduced microleakage for Prime & Bond 2.1 and had no effect on microleakage for Single Bond. INTRODUCTION A number of articles over the past few years have critically evaluated the role of the collagen-rich layer produced by acid etching of dentin (Gwinnett, 1994; Wakabayashi & others, 1994; Uno & Finger, 1995; Gwinnett & others, 1996; Saboia, Rodrigues & Pimenta, 2000). Dissolution of collagen by deproteinization prior to adhesive bonding produces excellent bonding efficacy (Uno & Finger, 1995). With acid conditioning and some bonding combinations, dissolution of collagen had no adverse effects on dentinal adhesion (Gwinnett, 1994; Uno & Finger, 1995; Gwinnett & others, 1996; Chersoni & others, 1998). Other works showed that bond strength increased

41 Saboia, Pimenta & Ambrosano: Effect of Collagen Removal on Microleakage of Resin Composite Restorations 39 when an acetone-based adhesive system was applied to collagen-depleted dentin (Ciucchi, Sano & Pashley, 1994; Fujita & others, 1996; Vargas, Cobb & Armstrong, 1997; Inai & others, 1998; Saboia & others, 2000). In contrast to the tendency toward improved shear bond strengths after demineralization and deproteinization, marginal quality was clearly enhanced when a collagen layer was present and a hybrid layer had formed (Uno & Finger, 1995; Vichi, Ferrari & Davidson, 1997, Frankenberger & others, 2000). On the other hand, a leakage pathway was found at the adhesive interface even with formation of a hybrid layer (Sano & others, 1995). It appears that the hybrid layer is a weak link in the coupling of resin to dentin (Wakabayashi & others, 1994) and may be susceptible to hydrolytic attack as was suggested by the increased porosity seen at the top of the hybrid layer over time (Sano & others, 1999). Leakage of the resin composite restoration may be attributed to the contraction gap that forms under the restoration from polymerization shrinkage and expansion and contraction of the substrates with temperature changes because the coefficient of thermal expansion of resin composite is different from that of dental hard tissues (Bullard, Leinfelder & Russell, 1988). The extent of curing shrinkage determines the creation of marginal gaps if the filling material did not sufficiently adhere to tooth structure, or it could cause cohesive failures in the material (van Dijken, 1996). Marginal gaps could result in microleakage, which is considered a major factor influencing the longevity of dental restorations (Anusavice, 1988). Microleakage may lead to staining at the margins of restorations, a hastening of the breakdown at the marginal areas of the restorations, recurrent caries (Going, 1972), hypersensitivity of restored teeth and development of pulpal pathology (Alani & Toh, 1997). This study evaluated the effects of collagen removal on microleakage of resin restorations placed with two single-bottle adhesive systems after 12 months storage time: an acetone-based (Prime & Bond 2.1) and an ethanol/water-based (Single Bond). METHODS AND MATERIALS Forty sound, non-carious human third molars extracted within a six month period and stored in solution of 10% formalin (ph 7.0) and two one-bottle dentin adhesive systems were used in this study (Table 1). After sectioning from crowns at the level of the cementoenamel junction, the roots were cleaned with scalers and flour of pumice, then stored in distilled water for one week prior to the study. The pulp chamber and apex were filled with Z100 composite resin. Each root received two preparations that were 2 mm in diameter and 2 mm Table1: Adhesive Systems Used in This Study Prime & Bond 2.1 Single Bond UDMA, PENTA, R5-62-1resin, BPDM Butylated hydroxytoluene, 4-ethyl dimethylaminobenzoate, cetylamine hydrofluoride, acetone, PI HEMA, Bis-GMA, dimethacrylates, methacrylates pendant polyalkenoic acid copoly mer, ethanol, water, PI UDMA: urethane di-methacrylate; PENTA: phosphoric penta-acrylate ester; PI: photoinitiator; BPDM: bisphenol di-methacrylate; Bis-GMA: bisphenol-glycidyl methacrylate; HEMA: 2-hydroxyetyl methacrylate. deep, using a diamond bur #2294 (KG Sorensen, São Paulo, SP ). The roots were randomly assigned for restoration with Z100 composite resin and Prime & Bond 2.1 or Single Bond. Each root received two restorations with the same adhesive system; however, one followed the manufacturer s instructions, while in the other, the adhesive was applied on collagen-depleted dentin. The preparations were assigned to four groups of 20 each and received the following treatments. Group 1 (PB) Prime & Bond 2.1 (Dentsply). Dentin surfaces were conditioned with 36% phosphoric acid (H 3 PO 4 ) gel (Dentsply) for 15 seconds, rinsed with water for 15 seconds and blot-dried, leaving a moist surface. Prime & Bond 2.1 was applied according to manufacturer s directions and light cured with an Optilux 500 (Demetron Research Corp, Danbury, CT 06810, USA) visible light-activation unit. The light output measured with a curing radiometer was 650 mw/cm 2. Z100 composite resin was inserted in one increment, then light cured for 40 seconds. In the first 10 seconds, the light was kept 1 cm from the restorative material. In the remaining 30 seconds, the light touched the restorative material. Group 2 (PBH) Prime & Bond NaOCl. The same procedures as Group 1 were followed, but a solution of 10% NaOCl was applied for one minute after acid-conditioning, then rinsed for 30 seconds and blotdried before applying the adhesive. Group 3 (SB) Single Bond (3M Dental Products). Dentin surfaces were conditioned with 36% H 3 PO 4 gel for 15 seconds, rinsed with water for 15 seconds and blot-dried, leaving a moist surface. Single Bond adhesive was applied and light cured following manufacturer s instructions. The composite resin application and its light cure were the same as in Group 1. Group 4 (SBH) Single Bond + NaOCl. The same procedures were followed as for Group 3 except that a solution of 10% NaOCl was applied after acid conditioning, then rinsed for 30 seconds and blot-dried before applying the adhesive. After 15 minutes, the excess composite was removed with a diamond bur #3195F (KG Sorensen, São Paulo, SP

42 40 Operative Dentistry ) mounted in a high-speed handpiece with water spray and the specimens were kept in distilled water at 37 C for 24 hours. Final finishing of the restorations was done with a graded series of Sof-Lex disks (3M Dental Products). The specimens were subjected to a thermocycling regimen of 5,000 cycles between 5 (+/-2) C and 55 (+/-2) C waterbaths (Instrumental MCT-2, São Paulo, SP ). The dwell time was one minute, with a five-second transfer time between baths. After thermal cycling, the specimens were immersed in distilled water at 37 C for 12 months. During this period the water was changed once a week. In preparation for the dye penetration test, the specimens were dried superficially and two coats of nail varnish applied to the entire specimen surface, leaving a 1 mm window around the cavity margins. The samples were then immersed in a 2% buffered methylene blue solution for four hours. After removal from the dye, the roots were rinsed in coolant water and sectioned longitudinally by a cut through the center of the restorations. A diamond saw (KG Sorensen, São Paulo, SP ) mounted in a low-speed handpiece was used to make the cuts. This procedure exposed four margins of the adhesive interface, allowing for identification of the areas where there was dye penetration. The degree of marginal microleakage was determined by the penetration of the tracer agent, starting from the margins of the restoration and moving toward the axial wall. Two independent observers analyzed the margins separately by viewing them under an optical microscope (EMZ-TR Meiji Techno Co, China) with a x45 magnification. The following scoring system was used: 0 = no marginal penetration. 1 = penetration up to 1/3 the length of the lateral wall. 2 = penetration from 1/3 up to 2/3 of the lateral wall. 3 = penetration of greater depth than in score 2, but not including the axial wall. 4 = penetration involving the axial wall. Table 2: Distribution of Microleakage Scores of Groups 1 (PB) and 2 (PBH) The highest score (maximum amount of leakage) of the four margins was recorded for the later analysis. If conflict occurred in scoring, consensus between observers was obtained. Scores for each group of restorations were compared using Fisher s Exact non-parametric statistical test (p 0.05). RESULTS The results of Fisher s Exact test showed that dentin treated with NaOCl significantly reduced the scores of microleakage when using the acetone-based adhesive system (Prime & Bond 2.1). When the collagen layer was left intact (Group 1 PB), all the specimens presented some degree of microleakage. Fifty-five percent showed severe microleakage with penetration of the dye to the axial wall. When Prime & Bond 2.1 was applied in collagen-depleted dentin (Group 2 PBH), only 21.05% of the specimens showed microleakage up to the axial wall (score 4) and 21.05% showed no degree of microleakage (score 0). Table 2 shows the percentage scores of microleakage and the sample size of Groups 1 and 2. For the water-ethanol based adhesive system (SB), the NaOCl treatment had no significant influence on microleakage scores. The percentage scores of microleakage and the sample size of Groups 3 and 4 are shown in Table 3. DISCUSSION The study s results showed that removal of collagen significantly reduced the scores of microleakage for the acetone-based adhesive system (PB). This can be explained by two factors: (1) the higher diffusibility of the acetone as well as its higher ability to displace water (Jacobsen & Soderhold, 1995). These factors could improve the contact of the monomer with the irregular intertubular dentin structure exposed by NaOCl treatment, resulting in a homogeneous interface with no voids; (2) there were no collagen fibers directly exposed to the oral environment, and consequently, the degradation of adhesive interface by hydrolysis, which probably would start by exposed collagen (Sano & others, 1999), would not happen. The decrease in microleakage observed when the PB adhesive sys- % n % n % n % n % n G1-PB (n=20) G2-PBH (n=19) p=0.02 Table 3: Distribution of Microleakage Scores of Groups 3 (SB) and 4 (SBH) % n % n % n % n % n G3-SB (n=16) G4-SBH (n=15) p=0.72

43 Saboia, Pimenta & Ambrosano: Effect of Collagen Removal on Microleakage of Resin Composite Restorations 41 tem was applied on collagen-depleted dentin (G2-PBH) compared to the control group (G1-PB) (Table 2) can reinforce the hypothesis that NaOCl treatment would promote better contact between bonded surfaces. This results in an interface that resists the challenge of the oral environment simulated herein by thermal cycling and water storage for one year. When the collagen layer was left intact (G1-PB), all specimens restored with PB/Z100 showed some degree of microleakage, and in 55% of the specimens, the dye reached the axial wall. These findings could be related with the degradation of exposed collagen, as suggested by Sano & others (1999). Collagen removal by NaOCl treatment increases the surface roughness of dentin and its wettability (Toledano & others, 1999). Inai & others (1998) showed that deproteinization exposes a labyrinth of lateral secondary tubules that were not observed on etched dentin surfaces, which could cause some increase in wettability. After deproteinization, dentin turns into a porous structure with multiple irregularities that allows for good mechanical retention (Vargas & others, 1997). This substrate is also rich in hydroxyapatite crystals, and from a cristalographic viewpoint, it is similar to enamel (Sakae, Mishima & Kozawa, 1998). It may result in a stable interface over time because it is essentially made of minerals (Toledano & others, 1999). The polymerization of resin inside the porous surface ensures a solid anchorage for resin composite restorations and probably ensures for an adequate sealing of the dentin (Toledano & others, 1999). Related to the water-ethanol based adhesive system (Single Bond), removal of collagen had no influence on microleakage. Although 60% of specimens in Group 4 (SBH) showed a maximum microleakage score, under the experimental conditions of this study, there were no statistical differences between groups treated or not treated with NaOCl. In accordance with manufacturer s instruction, the ethanol/water-based adhesive system (Single Bond) was applied twice with an interval of five seconds between each application. As this kind of adhesive system diffuses more slowly than acetonebased systems (Jacobsen & Soderhold, 1995; Kanca III, 1998), the short dwell time may have been insufficient to permit a full diffusion of the monomer into the substrate. In this way, nanometric porosities of intertubular dentin created by NaOCl treatment were not reached by monomer, leaving the adhesive interface with voids. These voids are similar to those left in the hybrid layer after hydrolysis. For the ethanol-water adhesive system (Single Bond) in the presence or absence of collagen, voids would form at the adhesive interface, resulting in marginal defects. Similar results obtained by these techniques can reinforce this hypothesis. By increasing the time between applying the ethanol/water-based adhesive, it is hypothesized that complete penetration into the substrate would occur, resulting in values of adhesion similar to those obtained by the acetone-based adhesive system (Prime & Bond 2.1). The bond strength and microleakage relationship is complex and poorly understood (Fortin & others, 1994). Although there is no statistically significant correlation between these phenomena, they are strongly associated with each other (Retief, Mandras & Russell, 1994; Fortin & others, 1994; Frankenberger & others, 2000). Retief & others (1994) hypothesized that a shear bond strength of approximately MPa may reduce microleakage to nearly zero at the dentin/resin bonding system interface. Saboia & others (2000) showed an increase in shear bond strength for Prime & Bond 2.1 when applied to collagen-depleted dentin. A decrease in microleakage for this material after NaOCl treatment was observed in this work, reinforcing the idea that higher values of bond strength can be associated with lower scores of microleakage. The results of this study agree with those by Blunck, Speyer & Roulet (1997) that found a significant increase in the percentage of excellent margins of restorations after thermocycling specimens treated with NaOCl and restored using an acetone-based adhesive system (Prime & bond 2.0). For Gluma CPS, Optibond FL and Scotchbond MP, no statistically significant differences were found between groups treated and not treated with NaOCl. It was concluded that the hybrid layer is not essential as an elastic band to improve the marginal adaptation of composite resin in combination with dental adhesives in dental cylindrical cavities. Kobaslija (1999) showed an absence of nanoleakage for specimens treated with NaOCl and restored with adhesive systems acetone-based (Syntac) and waterglutaraldheyde-based (Gluma CPS). Probably, a longer dwell-time used for Gluma CPS permitted a full diffusion of the monomer into the substrate, resulting in a better seal of the margins and consequently an absence of leakage. Uno & Finger (1995), using a water-glutharaldheydebased adhesive system (Gluma), showed that the marginal quality was clearly enhanced when a hybrid layer was present. It was assumed that the resin-impregnated layer has a lower Young s modulus of elasticity than the restorative resin and thus acts as an inherent elastic buffering layer that absorbs the resin composite s curing contraction stress. Regarding this viewpoint, the marginal quality of the enamel-resin interface is expected to be worse than that of dentin-resin. There is no formation of the hybrid layer at this interface and Young s modulus of elasticity is higher than the dentin-resin interface, nevertheless, in contrast to what is expected, its marginal quality is better than of resin/dentin inter-

44 42 Operative Dentistry face. Perhaps the most important part is Young s modulus of elasticity of the adhesive layer instead of formation of the hybrid layer. The presence of elastomeric resins on Prime & Bond 2.1 (Table 1) may have aided its adhesive interface to absorb the stress from thermal cycling resulting in a decrease of microleakage. Vichi & others (1997) showed a statistically significant increase in microleakage for a water-based adhesive system (Scotchbond MP Plus) when applied to collagen-depleted dentin (NaOCl treatment). They concluded that a hybrid layer formation improves the sealing ability of this adhesive system. On the other hand, Toledano & others (2000) showed gap formation occurred at the dentin/cementum margins of Class V restorations whether or not a hybrid layer was present. They pointed out that the elastomeric resins included in the formulation of Prime & Bond 2.1 may not be sufficiently thick to compensate for shrinkage stress. CONCLUSIONS According to the results of this study, collagen removal may be important to reduce microleakage when using an acetone-based adhesive system (PB) and has no influence on microleakage for ethanol/water-based adhesive system (SB). However, further investigations would help to confirm the results in order to elucidate the effectiveness of this dentin treatment. (Received 26 February 2001) References Alani AH & Toh CG (1997) Detection of microleakage around dental restorations: A review Operative Dentistry 22(4) Anusavice KJ (1988) Quality evaluation of dental restorations Chicago Quintessence Publishing Co. Blunck U, Speyer F & Roulet J-F (1997) Effect of hypochlorite treatment of conditioned dentin on the marginal adaptation of composite resin restorations Journal of Dental Research 76 Abstract of Papers p 19 Abstract #46. Bullard RH, Leinfelder KF & Russell CM (1988) Effect of coefficient of thermal expansion on microleakage Journal of the American Dental Association 116(7) Chersoni S, Prati C, Montanari G & Mongiorgi R (1998) Effect of collagen layer on self-etching bonding systems adhesion Journal of Dental Research 77(1) Abstract of Papers p 238 Abstract #1062. Ciucchi B, Sano H & Pashley DH (1994) Bonding to sodium hypochlorite treated dentin Journal of Dental Research 73(1) Abstract of Papers p 296 Abstract #1556. Fortin D, Swift EJ Jr, Denehy GE & Reinhardt JW (1994) Bond strength and microleakage of current dentin adhesives Dental Materials 10(4) Frankenberger R, Krämer N, Oberschachtsiek H & Petschelt A (2000) Dentin bond strength and marginal adaptation after NaOCl pre-treatment Operative Dentistry 25(1) Fujita E, Yamashita A, Terada & Suzuki K (1996) Effect of NaOCl pre-conditioning to bovine root canal dentin Journal of Dental Research 75 Abstract of Papers p 391 Abstract #2990. Going RE (1972) Microleakage around dental restorations: A summarizing review Journal of the American Dental Association 84(6) Gwinnett AJ (1994) Altered tissue contribution to interface bond strength with acid conditioned dentin American Journal of Dentistry 7(5) Gwinnett AJ, Tay FR, Pang KM & Wei SHY (1996) Quantitative contribution of the collagen network in dentin hybridization American Journal of Dentistry 9(4) Inai N, Kanemura N, Tagami J, Watanabe LG, Marshall SJ & Marshall GW (1998) Adhesion between collagen depleted dentin and dentin adhesives American Journal of Dentistry 11(3) Jacobsen T & Soderhold KJ (1995) Some effects of water on dentin bonding Dental Materials 11(2) Kanca III, J (1998) Effect of primer dwell time on dentin bond strength General Dentistry 46(6) Kobaslija S (1999) The effect of NaOCl dentin treatment on leakages within the hybrid layer Thesis Bosnia University of Sarajevo. Retief H, Mandras RS & Russell CM (1994) Shear bond strength required to prevent microleakage at the dentin/restoration interface American Journal of Dentistry 7(1) Saboia V de PA, Rodrigues AL & Pimenta LAF (2000) Effect of collagen removal on shear bond strength of two single-bottle adhesive systems Operative Dentistry 25(5) Sakae T, Mishima H & Kozawa Y (1988) Changes in bovine dentin mineral with sodium hypochlorite treatment Journal of Dental Research 67(9) Sano H, Takatsu T, Ciucchi B, Horner JA, Matthews WG & Pashley DH (1995) Nanoleakage: Leakage within the hybrid layer Operative Dentistry 20(1) Sano H, Yoshikawa T, Pereira PRN, Kanemura N, Morigami M, Tagami J & Pashley DH (1999) Long-term durability of dentin bonds made with a self-etching primer, in vivo Journal of Dental Research 78(4) Toledano M, Osorio R, Perdigão J, Rosales JI, Thompson JY & Cabrerizo-Vilchez MA (1999) Effect of acid etching and collagen removal on dentin wettability and roughness Journal of Biomedical Materials Research 47(2) Toledano M, Perdigão J, Osorio R & Osorio E (2000) Effect of dentin deproteinization on microleakage of Class V composite restorations Operative Dentistry 25(5) Uno S & Finger WJ (1995) Function of the hybrid zone as a stress-absorbing layer in resin-dentin bonding Quintessence International 26(10) Van Dijken JW (1996) 3-year clinical evaluation of a compomer, a resin-modified glass ionomer and a resin composite in Class III restorations American Journal of Dentistry 9(5) Vargas MA, Cobb DS & Armstrong SR (1997) Resin-dentin shear bond strength and interfacial ultrastructure with and without hybrid layer Operative Dentistry 22(5)

45 Saboia, Pimenta & Ambrosano: Effect of Collagen Removal on Microleakage of Resin Composite Restorations 43 Vichi A, Ferrari M & Davidson CL (1997) In vivo leakage of an adhesive system with and without NaOCl as pretreatment Journal of Dental Research 76 Abstract of Papers p 398 Abstract #3077. Wakabayashi Y, Kondou Y, Suzuki K, Yatani H & Yamashita A (1994) Effect of dissolution of collagen on adhesion to dentin International Journal of Prosthodontics 7(4)

46 Operative Dentistry, 2002, 27, Effectiveness of Composite Cure with Pulse Activation and Soft-start Polymerization AUJ Yap MS Soh KS Siow Clinical Relevance The use of certain pulse activation and soft-start polymerization regimens may reduce the effectiveness of cure at the bottom surfaces of composite restorations. SUMMARY The study investigated the effectiveness of composite cure with pulse activation and soft-start polymerization. A light-cure unit (BISCO VIP, BISCO Dental Products, Schaumburg, IL 60193, USA) that allowed for independent command over time and intensity was used. The six lightcuring modes examined were: Control (C)-400 mw/cm 2 [40 seconds]; Pulse Delay I (PDI) -100 Department of Restorative Dentistry, Faculty of Dentistry, National University of Singapore, 5 Lower Kent Ridge Road, Singapore , Republic of Singapore Adrian UJ Yap, BDS, MSc, PhD, FAMS, FADM, FRSH, associate professor, Department of Restorative Dentistry, Faculty of Dentistry, assistant director, Centre for Biomedical Materials Applications and Technology, Faculty of Engineering, National University of Singapore MS Soh, BSc (Hons), Department of Chemistry, Faculty of Science, National University of Singapore KS Siow, BSc, MSc, PhD, associate professor, Department of Chemistry, Faculty of Science, National University of Singapore mw/cm 2 [3 seconds] delay [3 minutes] 500 mw/cm 2 [30 seconds]; Pulse Delay II (PDII) mw/cm 2 [20 seconds] delay [3 minutes] 500 mw/cm 2 [30 seconds]; Soft-start (SS) mw/cm 2 [10 seconds] 600 mw/cm 2 [30 seconds]; Pulse Cure I (PCI) 400 mw/cm 2 [10 seconds] delay [10 seconds] 400 mw/cm 2 [10 seconds] delay [10 seconds] 400 mw/cm 2 [20 seconds]; and Pulse Cure II (PCII) mw/cm 2 [20 seconds] delay [20 seconds] 400 mw/cm 2 [20 seconds]. Effectiveness of cure with the different modes was determined by measuring the top and bottom surface hardness of 2 mm thick composite (Z100) specimens using a digital microhardness tester (load=500 gf; dwell time=15 seconds). The effectiveness of cure of the bottom surface of the composite was also established by Fourier Transform Infrared (FTIR) spectroscopy using the KBr technique. Data obtained was analyzed using one-way ANOVA/Scheffe s post-hoc test (p<0.05). No significant difference in top Knoops Hardness Number KHN was observed except for PDI and PDII. At the bottom surfaces, KHN obtained with the control was significantly greater than with PDII, SS and PCII. FTIR results ranked well with the hardness of the bottom surfaces. The absorbance ratio of carbon double bonds to aromatic ring obtained with the control group was

47 Yap, Soh & Siow:Effectiveness of Composite Cure with Pulse Actvation and Soft-start Polymerization 45 significantly greater than with PDII and PCII. Effectiveness of the cure at the bottom surfaces of composites may be reduced by some pulse activation and soft-start polymerization regimens. INTRODUCTION An inherent disadvantage of visible light-activated resin composites is that they shrink during light polymerization (Sakaguchi & others, 1991; Yap & others, 2000). Composite shrinkage can be divided into pre-gel and post-gel phases. During pre-gel polymerization, the composite flows and stresses within the structure are relieved (Davidson & de Gee, 1984). Flow is the amount of shrinkage stress that exceeds the elastic limit (Davidson & de Gee, 1984) and the ability of molecules within the forming polymer to slip into new positions before being restricted by cross-linking. After gelation, flow ceases and cannot compensate for shrinkage stresses. Thus, post-gel polymerization results in significant stresses in the surrounding tooth structure and composite-tooth bond (Feilzer, de Gee & Davidson, 1987). Feilzer, de Gee & Davidson (1993) measured polymerization strain and found that lightactivated composites developed higher setting stresses than chemically activated ones. This was attributed to the capacity for flow to occur in the chemically activated materials. Flow tended not to occur in light-activated composites because of its characteristically more rapid polymerization and the more rapid achievement of cross-linking and elastic limit. Stresses arising from post-gel polymerization shrinkage may produce defects in composite tooth bond, leading to bond failure with associated postoperative sensitivity, microleakage and recurrent caries (Eick & Welch, 1986). If the composite-tooth bond is good, it may also cause deformation of the surrounding tooth structure (Sheth, Fuller & Jensen, 1988), resulting in microcracks in the cervical enamel (Bowen, Nemoto & Rapson, 1983) that predisposes the tooth to fracture. One way of minimizing polymerization shrinkage of light-activated composites is to allow flow Light-curing Mode Regimen through controlled poly- (CC) (40 Control 400 mw/cm 2 seconds) merization during setting. This may be achieved by applying shortpulses of light energy, prepolymerization at low light Table 1: The Different Light-curing Modes Examined intensity followed by final cure at high intensity (softstart polymerization) or a combination of both. Studies have shown that these polymerization modes result in smaller marginal gap, increased marginal integrity and improved material properties (Uno & Asmussen, 1991; Goracci, Casa de Martinis & Mori, 1996; Mehl, Hickel & Kunzelmann, 1997; Kanca & Suh, 1999). However, only a few studies have reported the effectiveness of cure of these new curing modes (Yap, Ng & Siow, 2001; Bouschlicher & Rueggeberg, 2001; Bouschlicher, Rueggeberg & Boyer, 2000; Silikas, Eliades & Watts, 2000). These studies focused on softstart polymerization, not pulse activation or a combination of both curing modes. The effectiveness of composite cure may be assessed directly or indirectly. Direct methods that assess the degree of conversion, such as infrared spectroscopy and laser Raman spectroscopy, have not been accepted for routine use as they are complex, expensive and timeconsuming (Rueggeberg & Craig, 1988). Indirect methods have included visual, scraping and hardness testing. Incremental surface hardness has been shown to be an indicator of the degree of conversion (Asmussen, 1982a) and a good correlation between Knoops hardness and infrared spectroscopy has also been reported (DeWald & Ferracane, 1987). This study investigated the effectiveness of composite cure associated with different pulse activation and soft-start polymerization regimens by hardness testing and infrared spectroscopy. METHODS AND MATERIALS An A2 shade mini-filled resin composite (Z100; 3M Dental Products) with a mean particle size of 0.5 to 0.7 µm and a commercial light-cure unit that allowed for independent command over time and intensity (Variable Intensity Polymerizer (VIP); BISCO Inc, Schaumburg, IL 60193, USA) was selected for this study. VIP has an output wavelength range of 400 to 500 nm and is programmed with preset exposure times of 2 to 5, 10, 20 and Pulse Delay I 100 mw/cm 2 Delay 500 mw/cm 2 (PDI) (3 seconds) (3 minutes) (30 seconds) Pulse Delay II 200 mw/cm 2 Delay 500 mw/cm 2 (PDII) (20 seconds) (3 minutes) (30 seconds) Soft-start 200 mw/cm mw/cm 2 (SS) (10 seconds) (30 seconds) Pulse Cure I 400 mw/cm 2 Delay 400 mw/cm 2 Delay 400 mw/cm 2 (PCI) (10 seconds) (10 seconds) (10 seconds) (10 seconds) (20 seconds) Pulse Cure II 400 mw/cm 2 Delay 400 mw/cm 2 (PCII) (20 seconds) (20 seconds) (20 seconds)

48 46 Operative Dentistry 30 seconds and a continuous mode up to 225 seconds. The light intensities for each light-curing mode were checked with the in-built radiometer prior to use. The programmed intensity settings are 100, 200, 300, 400, 500 and 600 mw/cm 2. The six light-curing modes are detailed in Table 1. The control mode (C) involves light irradiation at 400 mw/cm 2 for 40 seconds. Pulse Delay I (PDI), recommended by BISCO, uses an initial low energy dose (100 mw/cm 2 for three seconds) followed by a waiting time of three minutes and a final cure at high-energy dose (500 mw/cm 2 for 30 seconds). The three-minute waiting period is to be used for finishing and polishing the composite restoration. Pulse Delay II (PDII) is similar to PDI with the exception of a higher initial energy dose (200 mw/cm 2 for 20 seconds). The soft-start (SS) mode uses an initial low-light intensity (200 mw/cm 2 for 10 seconds) immediately followed by final cure at high light intensity (600 mw/cm 2 for 30 seconds). Pulse Cure I (PCI) involves two 400 mw/cm 2 10-second pulses and one 400 mw/cm 2 20-second pulse with 10 second intervals in-between. For Pulse Cure II (PCII), two 400 mw/cm 2 20-second pulses with 20-second intervals in-between was employed. The hardness testing methodology used to assess the effectiveness of cure was based upon that used by Yap (2000). The composite was placed in black delrin molds with square cavities 2 mm deep and 4 mm wide/long and confined between two opposing acetate strips (Hawe-Neos Dental, Bioggio, Switzerland). The bottom of the molds was blacked-out to prevent transmission of ambient light. A glass slide (1 mm thick) was then placed on the molds and excess material was extruded by pressure application. The composite was then irradiated from the top through the glass slide and an acetate strip using the different light-curing modes. Immediately after light polymerization, the acetate strips were removed and the specimens were positioned in their molds centrally beneath the indenter of a digital microhardness tester (FM7, Future-Tech Corp, Tokyo, Japan) to assess the Knoop s Hardness Number (KHN) of the top and bottom surfaces. A 500g load was then applied through the indenter with a dwell time of 15 seconds. The KHN corresponding to each indentation was computed by measuring the dimensions of the indentations using the formula KHN=1.451 x (F/d 2 ), where F is the test load in Newtons and d is the longer diagonal length of an indentation in millimeters. Five specimens were made for each light-curing mode. Three readings were taken for each specimen and averaged to form a single value for that specimen. The mean KHN and hardness ratio of the five specimens were then calculated and tabulated using the formula: Hardness ratio=khn of bottom surface/khn of top surface. Fourier Transform Infrared (FTIR) spectroscopy was used to confirm the effectiveness of cure of the bottom surfaces. A scalpel blade was used to obtain biopsies of the bottom surfaces. The number of scrapings was standardized at 50 for each specimen. Specimen biopsies were then ground down with a mortar and pestle until a fine powder was obtained. The powder was then mixed with IR grade potassium bromide and pressed into a pellet. The degree of cure (DC) was evaluated by performing FTIR analysis (FTS 165, Bio-Rad Laboratories, Hercules, CA 94547, USA) of the pellets in transmission mode. Sixty-four scans over cm -1 range and a resolution of 4 cm -1 were used in the acquisition of the FTIR spectra. The spectra were obtained under a nitrogen gas purge to ensure a clean, stable background. DC was calculated by comparing the absorbance ratio of the C=C (methacrylate carbon double bond) absorbance peak at cm -1 to the aromatic ring peak at cm -1. The interaction between light-curing modes and specimen surfaces was examined using two-way analyses of variance (ANOVA). Hardness and DC data were subjected to one-way ANOVA and Scheffe s post-hoc test. All statistical analysis were carried at a significance level of RESULTS The mean KHN and hardness ratio associated with the different light-curing modes are shown in Table 2. The mean absorbance ratio is reflected in Table 3. Results of statistical analysis are shown in Table 4. Two-way ANOVA revealed significant interaction between light-curing modes and specimen surfaces. Therefore, the effects of the light-curing modes on hardness were surface dependent. At the top surface, KHN after polymerization with PDII was significantly higher than with PDI. No significant difference in top KHN was observed between the other light-curing modes. Ranking of KHN at the bottom surfaces was as follows: C>PDI>PCI>SS>PDII>PCII. The bottom KHN of specimens in the control was significantly higher than specimens polymerized by PDII, SS and PCII. The bottom surfaces of specimens polymerized with PDI were significantly harder than specimens polymerized with PDII and PCII. Bottom KHN of PCI Table 2: Mean KHN and Hardness Ratio for the Different Light-curing Modes Light-curing Modes Top KHN Bottom KHN Hardness Ratio Control (0.70) (0.32) 0.93 (0.01) Pulse Delay I (5.88) (5.28) 0.95 (0.02) Pulse Delay II (0.18) (1.21) 0.76 (0.01) Soft-start (2.44) (2.53) 0.81 (0.03) Pulse Cure I (0.32) (0.59) 0.86 (0.01) Pulse Cure II (0.82) (1.88) 0.75 (0.02) Standard deviations in parentheses.

49 Yap, Soh & Siow:Effectiveness of Composite Cure with Pulse Actvation and Soft-start Polymerization 47 Table 3: Mean Absorbance Ratio with the Different Light-curing Modes Light-curing Modes Mean Absorbance Ratio Control (0.018) Pulse Delay Cure I (0.014) Pulse Delay Cure II (0.004) Soft Start (0.006) Pulse Cure I (0.005) Pulse Cure II (0.009) Standard deviations in parentheses. Table 4: Results of Statistical Analysis Variable KHN Top KHN Bottom Hardness Ratio Absorbance Ratio Significance Pulse Delay II > Pulse Delay I Control>Pulse Delay II, Soft-start, Pulse Cure II Pulse Delay I>Pulse Delay II, Pulse Cure II Pulse Cure I>Pulse Cure II Control, Pulse Delay I>Pulse Delay II, Soft-start, Pulse Cure I, Pulse Cure II Soft-start>Pulse Delay II, Pulse Cure II Pulse Cure I>Pulse Delay II, Soft-start, Pulse Cure II Control>Pulse Delay II, Pulse Cure II Results of one-way ANOVA/Scheffe s post-hoc test at significance level > denotes statistical significance. specimens was also significantly higher than PCII. The hardness ratio associated with C and PDI were significantly greater than all the other curing modes. The hardness ratio of SS and PCI was significantly greater than PDII and PCII. The hardness ratio of PCI was significantly greater than SS. Absorbance ratio ranged from for PCII to for C. Ranking of absorbance ratio was as follows: C>PDI>PCI>SS>PDII>PCII. This was identical to the ranking for KHN at the bottom surfaces. The absorbance ratio observed with C was significantly greater than with PDII and PCII. No significant difference in absorbance ratio was observed between C and the other curing modes. DISCUSSION Effective cure of light-activated composites is vital, not only to ensure optimum physico-mechanical properties (Asmussen, 1982b), but also to ensure that clinical problems do not arise due to cytotoxicity of inadequately polymerized material (Caughman & others, 1991). The effectiveness of cure not only depends on the chemistry of the material and concentration of initiator, but also upon the filler particle type, size and quantity. In addition, polymerization is dependent on the effectiveness of the radiation sources, including spectral distribution, intensity, exposure time and position of the light-cure tip (Harrington, Wilson & Shortall. 1996). The composite material investigated and the distance of the light-cure tip from the composite (1 mm via usage of glass slide) were therefore standardized in this study. Versluis, Sakaguchi & Douglas (1993) examined the contraction stresses of 10 resin composites and found that Z100 exhibited the greatest contraction stress of all materials tested. This makes it a suitable candidate for attempts at stress reduction using pulse activation and softstart polymerization. Two mm thick composite specimens were used as it ensured uniform and maximum polymerization (Yap, 2000). A2 shade was selected to minimize the effects of colorants on light polymerization (Bayne, Heymann & Swift, 1994). As a minimum intensity of 400 mw/cm 2 has been suggested for routine polymerization (Rueggeberg, Caughman & Curtis, 1994; Manga, Charlton & Wakefield, 1995), this light intensity, together with the manufacturer s recommended cure time of 40 seconds, was used as control. The pulse-delay modes utilize a combination of pulse activation and soft-start polymerization. PDII is a modification of PDI. A higher initial energy dose (200 mw/cm 2 for 20 seconds) was employed as usage of the low pulse energy (100 mw/cm 2 for three seconds) recommended by BISCO could result in a soft surface that may be more prone to damage and excessive removal during finishing and polishing procedures (Yap & Seneviratne, 2001). The degree to which light-activated composites polymerize is proportional to the amount of light to which they are exposed (Rueggeberg & others, 1994). For resin composites 2 mm in depth, source intensity and exposure duration are two important factors that influenced composite cure (Rueggeberg & others, 1993). Denehy & others (1993) found that the top surface hardness of composites was less dependent on light intensity than the bottom surface. At the top surface where no overlying composites interfere with light transmission, it has also been been established that even relatively low intensity lights can cure the resin matrix to an extent almost equal to when high-intensity lights are used (Rueggeberg & Jordan, 1993). The general lack of significance between the curing modes in top KHN found in this study corroborates the aforementioned studies. At the top surface, sufficient light energy reaches and activates the photoinitiator, initiating polymerization. Continued exposure sustains activation of photoinitiator molecules near the surface. Hence, the duration of exposure is the more important factor for polymerization at the top surface. The latter explains the significant difference in top KHN between PDII and PDI. Total light exposure times were 50 and 33 seconds for PDII and PDI, respectively. Total light

50 48 Operative Dentistry exposure time for the control and other curing modes was 40 seconds. At the bottom surfaces, a significant difference in KHN was observed between specimens in the control group and specimens polymerized with PDII, SS and PCII. As total light exposure time was identical or greater than the control group, any difference in KHN can be attributed to the light intensity or pulse activation and soft-start polymerization regimen used. For PDII and SS, usage of an intensity of 200 mw/cm 2 for 10 to 20 seconds appeared to minimize the compensatory effects of high light intensities (Yap & Seneviratne, 2001). Polymerization of the resin associated with the initial cure may be sufficient to interfere with light transmission and severely decrease the amount of light reaching the bottom surface. The bottom KHN of PCII was the lowest among the different light-curing modes. Use of a higher light energy density (400 mw/cm 2 for 20 seconds) during the first pulse cure may result in greater resin polymerization leading to increased interference with light transmission. Both light scattering and absorption by the composite are possible mechanisms. Since a higher light intensity was not used for the final cure, the presence of a low bottom KHN is anticipated. The ranking of DC and KHN of the bottom surfaces were identical. This finding agreed with that of DeWald & Ferracane (1987), who found a good correlation between Knoops hardness testing and infrared spectroscopy. FTIR was not conducted on the top surfaces, as there was generally no significant difference in effectiveness of cure at this surface and the intensity of the light source is a more critical factor in determining the effectiveness of cure for the bottom surface (Rueggeberg & others, 1993; Denehy & others, 1993; Sakaguchi & Berge, 1998). Determining effectiveness of cure by FTIR may, however, be more sensitive than hardness testing. Significant differences in DC were only observed between the control group and specimens polymerized with PDII and PCII. Ideally, the degree of polymerization of the composite should be the same throughout its depth and the hardness ratio should be very close or equal to one. As light passes through the composite, the light intensity is greatly reduced due to light scattering, thus decreasing the effectiveness of cure at the bottom surface (Ruyter & Øysæd, 1982). It has been suggested that the top-tobottom hardness gradient should not exceed 10-20% (that is, hardness ratio should be greater than 0.8) for light-activated resin composite to be adequately polymerized (Pilo & Cardash, 1992). The hardness ratio for both PDII and PCII were below 0.8 and hardness ratios associated with these curing modes were significantly lower than all the other modes investigated. Based on the bottom KHN, DC and hardness ratio, clinical usage of Pulse Delay II and Pulse Cure II is not advocated. CONCLUSIONS 1. The effectiveness of cure at the top surface of composites was generally not affected by pulse activation and soft-start polymerization regimens. 2. The effectiveness of cure at the bottom surface of composites was significantly affected by some pulse activation and soft-start polymerization regimens. 3. Polymerization with Pulse Delay II and Pulse Cure II modes resulted in significantly lower hardness ratio, hardness and degree of cure of the bottom surface as compared to the control and should not be used clinically. 4. Use of high energy densities during the initial cure phase of pulse activation and soft-start polymerization regimens may adversely affect the effectiveness of cure at the bottom of composite restorations. (Received 19 March 2001) References Asmussen E (1982a) Factors affecting the quantity of remaining double bonds in restorative resin polymers Scandinavian Journal of Dental Research 90(6) Asmussen E (1982b) Restorative resins: Hardness and strength vs quantity of remaining double bonds Scandinavian Journal of Dental Research 90(6) Bayne SC, Heymann HO & Swift EJ Jr (1994) Update on dental composite restorations Journal of the American Dental Association 125(6) Bouschlicher MR & Rueggeberg FA (2001) Effect of ramped light intensity on polymerization force and conversion in a photoactivated composite Journal of Esthetic Dentistry Bouschlicher MR, Rueggeberg FA & Boyer DB (2000) Effect of stepped light intensity on polymerization force and conversion in a photo-activated composite Journal of Esthetic Dentistry 12(1) Bowen RL, Nemoto K & Rapson JE (1983) Adhesive bonding of various materials to hard tooth tissues: Forces developing in composite materials during hardening Journal of the American Dental Association 106(4) Caughman WF, Caughman GB, Shiflett RA, Rueggeberg F & Schuster GS (1991) Correlation of cytotoxicity, filler loading and curing time of dental composites Biomaterials 12(18) Davidson CL & de Gee AJ (1984) Relaxation of polymerization contraction stresses by flow in dental composites Journal of Dental Research Denehy GE, Pires JAF, Cuitko E & Swift EJ (1993) Effects of curing tip distance on light intensity and composite resin microhardness Quintessence International 24(7)

51 Yap, Soh & Siow:Effectiveness of Composite Cure with Pulse Actvation and Soft-start Polymerization 49 DeWald JP & Ferracane JL (1987) A comparison of four modes of evaluating depth of cure of light-activated composites Journal of Dental Research 66(3) Eick JD & Welch FH (1986) Polymerization shrinkage of posterior composite resins and its possible influence on postoperative sensitivity Quintessence International 17(2) Feilzer AJ, de Gee AJ & Davidson CL (1987) Setting stress in composite resin in relation to configuration of the restoration Journal of Dental Research 66(11) Feilzer AJ, de Gee AJ & Davidson CL (1993) Setting stresses in composites for two different curing modes Dental Materials 9(1) 2-5. Goracci G, Casa de Martinis L & Mori G (1996) Curing light intensity and marginal leakage of composite resin restorations Quintessence International Harrington E, Wilson HJ & Shortall AC (1996) Light-activated restorative materials: A method of determining effective radiation times Journal of Oral Rehabilitation 23(3) Kanca J III & Suh BI (1999) Pulse activation: Reducing resinbased composite contraction at the enamel cavosurface margins American Journal of Dentistry 12(3) Manga RK, Charlton DG & Wakefield CW (1995) In vitro evaluation of a curing radiometer as a predictor of polymerization depth General Dentistry 43(3) Mehl A, Hickel R & Kunzelmann KH (1997) Physical properties and gap formation of light-cured composites with and without soft-start polymerization Journal of Dentistry 25(3-4) Pilo R & Cardash HS (1992) Post-irradiation polymerization of different anterior and posterior visible light-activated resin composites Dental Materials 8(5) Rueggeberg FA, Caughman WF & Curtis JW Jr (1994) Effect of light intensity and exposure duration on cure of resin composite Operative Dentistry 19(1) Rueggeberg FA, Caughman WF, Curtis JW Jr & Davis HC (1993) Factors affecting cure at depths within light-activated resin composites American Journal of Dentistry 6(2) Rueggeberg FA & Craig RG (1988) Correlation of parameters used to estimate monomer conversion in a light-cured composite Journal of Dental Research 67(6) Rueggeberg FA & Jordan DM (1993) Effect of light-tip distance on polymerization of resin composite International Journal of Prosthodontics 6(4) Ruyter IE & Øysæd H (1982) Conversion in different depths of ultraviolet and visible light activated composite materials Acta Odontologica Scandinavica 40(3) Sakaguchi RL & Berge HX (1998) Reduced light energy density decreases post-gel contraction while maintaining degree of conversion in composites Journal of Dentistry 26(8) Sakaguchi RL, Sasik CT, Bunczak MA & Douglas WH (1991) Strain gauge method for measuring polymerization contraction of composite resin Journal of Dentistry 19(5) Sheth JJ, Fuller JL & Jensen ME (1988) Cuspal deformation and fracture resistance of teeth with dentin adhesives and composites Journal of Prosthetic Dentistry 60(5) Silikas N, Eliades G & Watts DC (2000) Light intensity effects on resin-composite degree of conversion and shrinkage strain Dental Materials 16(4) Uno S & Asmussen E (1991) Marginal adaptation of a restorative resin polymerized at reduced rate Scandinavian Journal of Dental Research 99(5) Versluis A, Sakaguchi R & Douglas W (1993) Post-gel shrinkage measurements by means of strain gauges Journal of Dental Research (Abstract #2263). Yap AUJ (2000) Effectiveness of polymerization in composite restoratives claiming bulk placement: Impact of cavity depth and exposure time Operative Dentistry 25(2) Yap AUJ, Ng SC & Siow KS (2001) Soft-start polymerization: Influence on effectiveness of cure and post-gel shrinkage Operative Dentistry 26(3) Yap AUJ & Seneviratne C (2001) Influence of light energy density on effectiveness of composite cure Operative Dentistry 26(5) Yap AUJ, Wang HB, Siow KS & Gan LM (2000) Polymerization

52 Operative Dentistry, 2002, 27, Effect of Prophylactic Polishing Protocols on the Surface Roughness of Esthetic Restorative Materials AL Neme KB Frazier LB Roeder TL Debner Clinical Relevance Prophylactic polishing protocols can be used to restore a smooth surface on esthetic restorative materials following simulated tooth brushing. SUMMARY Many polishing protocols have been evaluated in vitro for their effect on the surface roughness of restorative materials. These results have been useful in establishing protocols for in vivo application. However, limited research has focused on the subsequent care and maintenance of esthetic restorations following their placement. This investigation evaluated the effect of five polishing protocols that could be implemented at recall on the surface roughness of five direct esthetic restorative materials. Specimens (n=25) measuring 8 mm diameter X 3 mm thick were fabricated in an acrylic mold using five light-cured resin-based University of Detroit Mercy School of Dentistry, Department of Restorative Dentistry, 8200 West Outer Drive, PO Box 19900, Detroit, MI 48219, USA Ann-Marie L Neme, DDS, MS, associate professor Kevin B Frazier, DMD, associate professor, Medical College of Georgia Leslie B Roeder, DDS, associate professor, University of Texas Houston Health Science Center Tara L Debner, RDH, graduate student, University of Texas Houston Health Science Center materials (hybrid composite, microfilled composite, packable composite, compomer and resinmodified glass ionomer). After photopolymerization, all specimens were polished with Sof-Lex Disks to produce an initial (baseline) surface finish. All specimens were then polished with one of five prophylactic protocols (Butler medium paste, Butler coarse paste, OneGloss, SuperBuff or OneGloss & SuperBuff). The average surface roughness of each treated specimen was determined from three measurements with a profilometer (Surface 1). Next, all specimens were brushed 60,000 times at 1.5 Hz using a brush-head force of 2 N on a Manly V-8 cross-brushing machine in a 50:50 (w/w) slurry of toothpaste and water. The surface roughness of each specimen was measured after brushing (Surface 2) followed by re-polishing with one of five protocols, then final surface roughness values were determined (Surface 3). The data were analyzed using repeated measures ANOVA. Significant differences (p=0.05) in surface roughness were observed among restorative materials and polishing protocols. The microfilled and hybrid resin composite yielded significantly rougher surfaces than the other three materials following tooth brushing. Prophylactic polishing protocols can be used to restore a smooth surface on resinbased esthetic restorative materials following simulated tooth brushing.

53 Neme & Others: Effect of Polishing on Surface Roughness of Esthetic Restorative Materials 51 INTRODUCTION The importance of a smooth surface to the success of a restoration has been well-documented (Fruits, Miranda & Coury, 1996; Goldstein, 1989; Goldstein & Goldstein, 1988). Smooth, highly polished restorations have been shown to be more esthetic and more easily maintained than restorations with rougher surfaces (Strassler & Baum, 1993; Weitman & Eames, 1975). A number of polishing protocols has been evaluated in vitro for their effect on the surface roughness of restorative materials. These results have been useful in establishing protocols for in vivo application. The continued development of esthetically acceptable adhesive restorative materials has made a variety of tooth-colored materials available for clinical use. Currently, the clinician has resin composite, polyacidmodified composite, resin-modified glass ionomer and traditional glass ionomer restoratives as options for direct restorations. In addition, resin composite materials are available with a variety of filler types that affect both their handling characteristics and physical properties. The ultimate esthetics of these tooth-colored restoratives is strongly influenced by the final surface polish (Yap, Lye & Sau, 1997). Reported surface roughness values following initial finishing and polishing have traditionally been higher for glass ionomer and lower for resin-modified glass ionomer, with resin composite restorative materials yielding the smoothest finish (Hoelscher & others, 1998; Hondrum & Fernandez, 1997). Finishing has been defined as the gross contouring or reduction of a restoration to obtain ideal anatomy. Polishing refers to the reduction of the roughness and scratches created by finishing instruments (Yap & others, 1997). Although there has been extensive research into initial finishing and polishing of tooth-colored restoratives, comparatively little has been done to investigate the effects of hygiene maintenance on esthetic restorations by patients or professional health care providers. Goldstein & Garber (1995) stated that a high percentage of restorations are compromised by the lack of ongoing maintenance on the part of both patient and dentist. Problems associated with abrasion of dentin and enamel by dentifrices and polishing pastes have been investigated (Barbakow, Lutz & Imfeld, 1987). Many clinicians use only one prophylaxis paste, although patients present with a wide range of polishing requirements. Medium- and coarse-grade prophylaxis pastes are popular because of their stain-removing potential although they can produce rough surfaces on tooth structure and restorations. These rough Material Restorative Material surfaces enhance the deposition of plaque, stain and calculus (Lutz & others, 1993). Researchers have demonstrated that current methods of oral hygiene by the patient and the health care provider to remove plaque and calculus may adversely affect the surface characteristics of restorations for implant abutments, creating a rougher surface for plaque and calculus accumulation (Balshi, 1986; Rapley & others, 1990). Momoi & others (1997) comment that toothbrush-dentifrice abrasion can take place on all restoration surfaces; however, it is more commonly seen on restorations in anterior teeth and cervical erosion/abrasion lesions. In their 1997 in vitro investigation, Momoi & others reported that following brushing with a dentifrice, the abraded surface characteristics of glass ionomer and resin-modified glass ionomer restorative materials were apparent under microscopic examination. In addition, it has been shown that abrasives used in prophylaxis significantly roughen the surface of resin-modified glass ionomers (Berry, Berry & Powers, 1994). The surface finish of dental restorations can have a substantial effect on their overall long-term clinical success. The finished surface can affect the esthetic qualities and longevity of the restoration as well as the biocompatibility with the surrounding oral tissues. It has been suggested that all esthetic restorations require ongoing maintenance, including periodic repolishing (Goldstein & others, 1992; Goldstein 1989). This investigation evaluated the effects of five polishing protocols that could be implemented at recall on the surface roughness of five direct esthetic restorative materials. METHODS AND MATERIALS The restorative materials used in this study were all light cured and included three resin composites, a resin-modified glass ionomer (hybrid ionomer) and a polyacid-modified resin composite (compomer). Table 1 lists the restorative materials (F-2000, Fuji II LC, Heliomolar, SureFil and TPH Spectrum), finishing/polishing agents (OneGloss, Butler medium and coarse prophy paste, Sof-Lex Disk system and SuperBuff) and their respective manufacturers. The sample size for Table1: List of Materials and Their Manufacturers Manufacturer F-2000 (Polyacid-modified composite) 3M Dental Products, Minneapolis, MN Fuji II LC (Resin-modified glass ionomer) GC America, Chicago, IL Heliomolar (Microfilled composite) Ivoclar Vivadent, Amherst, NY SureFil (Packable composite) Dentsply/Caulk, Milford, DE TPH Spectrum (Hybrid composite) Dentsply/Caulk, Milford, DE Finishing/Polishing Materials OneGloss Shofu, Menlo Park, CA Prophy Paste (medium, coarse) Butler, Chicago, IL Sof-Lex Disk System 3M Dental Products, St Paul, MN SuperBuff Shofu, Menlo Park, CA 94025

54 52 Operative Dentistry Table 2: Experimental Groups Treatment n Restorative Surface 1 Surface 2 Surface 3 Group Material 1 5 TPH Coarse prophy paste + simulated brushing + Coarse prophy paste 2 5 TPH Medium prophy paste + simulated brushing + Medium prophy paste 3 5 TPH OneGloss + simulated brushing + OneGloss 4 5 TPH SuperBuff + simulated brushing + SuperBuff 5 5 TPH OneGloss+SuperBuff + simulated brushing + OneGloss+SuperBuff 6 5 SureFil Coarse prophy paste + simulated brushing + Coarse prophy paste 7 5 SureFil Medium prophy paste + simulated brushing + Medium prophy paste 8 5 SureFil OneGloss + simulated brushing + OneGloss 9 5 SureFil SuperBuff + simulated brushing + SuperBuff 10 5 SureFil OneGloss+SuperBuff + simulated brushing + OneGloss+SuperBuff 11 5 Heliomolar Coarse prophy paste + simulated brushing + Coarse prophy paste 12 5 Heliomolar Medium prophy paste + simulated brushing + Medium prophy paste 13 5 Heliomolar OneGloss + simulated brushing + OneGloss 14 5 Heliomolar SuperBuff + simulated brushing + SuperBuff 15 5 Heliomolar OneGloss+SuperBuff + simulated brushing + OneGloss+SuperBuff 16 5 F-2000 Coarse prophy paste + simulated brushing + Coarse prophy paste 17 5 F-2000 Medium prophy paste + simulated brushing + Medium prophy paste 18 5 F-2000 OneGloss + simulated brushing + OneGloss 19 5 F-2000 SuperBuff + simulated brushing + SuperBuff 20 5 F-2000 OneGloss+SuperBuff + simulated brushing + OneGloss+SuperBuff 21 5 Fuji II LC Coarse prophy paste + simulated brushing + Coarse prophy paste 22 5 Fuji II LC Medium prophy paste + simulated brushing + Medium prophy paste 23 5 Fuji II LC OneGloss + simulated brushing + OneGloss 24 5 Fuji II LC SuperBuff + simulated brushing + SuperBuff 25 5 Fuji II LC OneGloss+SuperBuff + simulated brushing + OneGloss+SuperBuff this research protocol was calculated using the range and standard deviations from data presented in previous literature (Hoelscher & others 1998; Herrgott, Ziemiecki & Dennison, 1989). Based on those published values, a sample size of three units per cell was calculated (significance level of alpha = 0.05; power of 0.80). To prevent Type II error, the final sample size was increased to five samples per group with five groups for each of the five materials. Twenty-five samples were fabricated for each of the five resin-based restorative materials. Cylindrical cavities were fabricated from a sectioned acrylic rod measuring 8 mm in diameter and 3 mm deep. The cavities were slightly over-filled with material, covered on each side with a Mylar matrix strip (SS White Co, Philadelphia, PA 19177, USA) and placed between two glass slides. Each side of the two-sided sample was irradiated with a halogen light-curing unit (Curing Light 2500, 3M Dental Products, St Paul, MN 55144, USA) for 30 seconds. A Dentek photometric tester (Dentek, Inc, Buffalo, NY 14207, USA) verified light intensity (750 mw/cm 2 ). After the initial two-way light-curing step, samples were irradiated for an additional 60 seconds without the matrices in place. Next, the samples were stored at 100% humidity and 37 C for 24 hours prior to baseline polishing and surface roughness measurement. Following the storage period, each sample was polished with the medium, fine and superfine Sof-Lex Finishing and Polishing Disks (3M Dental Products, St Paul, MN 55144, USA) in sequence to attain a baseline surface. A single operator applied disks per manufacturer s directions with a light pressure in a circular pattern for 20 seconds per disk. Prior to additional polishing procedures, surface roughness was measured on five randomly selected samples from each of the restorative materials. The baseline surface was evaluated using a Talysurf 10 Profilometer (Taylor-Hobson, Leicester England) set for a tracing length of 2 mm and a cutoff value of 0.25 mm. Each sample was rotated 120 to record three readings per surface. The instrument provided a readout of average surface roughness (Ra) in microns. The average roughness value represents the arithmetic mean of the height of all surface

55 Neme & Others: Effect of Polishing on Surface Roughness of Esthetic Restorative Materials 53 irregularities over a predetermined linear segment of each sample. Five samples from each of the five restorative materials were randomized into one of five polishing treatments. Table 2 lists treatment group, restorative material and surface preparation. Every effort was made to standardize the different aspects of the methodology of this investigation. To control variability, one investigator finished all samples according to manufacturers directions. A single, slow-speed handpiece, prophy angle and unit were used for the investigation. Polishing using the various finishing/polishing systems was accomplished according to the manufacturers directions as follows: OneGloss (Shofu Dental Corporation, Menlo Park, CA 94025, USA) disk was used for 20 seconds with a moderate to light pressure followed by a feather-light touch; SuperBuff (Shofu Dental Corporation) was moistened with water and rotated at slow speed on its edge, applying firm-to-light pressure, adjusting to a feather-light buffing over a 20 seconds. The two prophy pastes (John O Butler Company, Chicago, IL 60630, USA), either coarse or medium grit, were applied for 20 seconds using an individual disposable prophy cup per sample with moderate-to-light pressure. The final polishing sequence was a combination of OneGloss and SuperBuff, respectively, for a total of 40 seconds polishing time. Samples were then rinsed in tap water and stored at 100% relative humidity. Each sample was measured by the profilometer as previously described and three average roughness values (Ra) per surface were recorded as Surface (treatment) for each material. Following surface analysis, each sample was further treated with a simulated toothbrushing technique. The toothbrushing device (Manly V-8 Crossbrushing Machine, Sabri Enterprises, Lombard, IL) contains eight brushing stations, each consisting of a toothbrush mounted on a moveable head and a sample holder. The brushing heads are connected to a camshaft driven by an electric motor. A toothbrush with nylon bristles (Oral B soft, straight head #35, Oral B Laboratories, Belmont, CA 94002, USA) was fitted into each head, and the samples were mounted in the corresponding sample holders. Care was taken to ensure that the toothbrush bristles were perpendicular to the surface of each sample and touched the surface evenly. A 50:50 (w/w) slurry of toothpaste (UltraBrite, Colgate- Palmolive, New York, NY 10022, USA) and deionized water was used as the abrasive medium. The volume of slurry needed to maintain a constant supply of abrasive between the brushes and the sample surfaces was provided by 37.5 grams of dentifrice and water (75 grams total). Each surface was brushed 60,000 times at 1.5 Hz using a brush-head force of 2 N. After the simulated brushing technique, samples were rinsed with tap water and stored in 100% humidity until roughness values (Surface 2) were obtained using the profilometer as previously described. After analyzing the brushed surfaces, each sample was re-polished with the original polishing protocol that was used to produce Surface 1. Samples were measured as previously described for final surface roughness (Surface 3). Data were analyzed using repeated-measures ANOVA and Tukey-Kramer technique for multi-wise comparison (α=0.05). RESULTS Table 3 lists the average material surface roughness measurements of the baseline group (Sof-Lex polishing only) and treatment groups. To compare baseline surfaces with treated surfaces, data were analyzed using two-factor ANOVA (α=0.05) and Tukey HSD pairwise comparison. Statistical analysis revealed a significant difference between polishing treatments and restorative materials with a significant interaction (p<0.05). No significant difference in surface roughness was apparent among the three resin composite materials; however, the polyacid-modified composite and resin-modified glass ionomer yielded a statistically rougher surface than the resin com- Table 3: Mean Ra Values (µm) and Standard Deviations for the Various Materials and Polishing Systems Evaluated Polishing Restorative Material Method Hybrid Microfilled Packable Polyacid Resin-modified Composite Composite Composite Modified Glass Ionomer Composite Coarse paste 0.28 (0.07) 0.30 (0.10) b 0.50 (0.17) ac 0.59 (0.15) 0.66 (0.23) Medium paste 0.31 (0.04) ad 0.35 (0.16) c 0.47 (0.23) bd 0.54 (0.23) 0.44 (0.15) OneGloss 0.34 (0.05) be 0.60 (0.24) abcde 0.39 (0.15) 0.47 (0.18) 0.56 (0.21) SuperBuff 0.19 (0.02) def 0.14 (0.03) d 0.17 (0.05) cd 0.42 (0.14) 0.48 (0.15) OneGloss + SuperBuff 0.30 (0.07) cf 0.28 (0.06) e 0.24 (0.04) 0.50 (0.16) 0.50 (0.14) Baseline (Sof-Lex) 0.21 (0.02) abc 0.19 (0.04) a 0.16 (0.06) ab 0.37 (0.04) 0.36 (0.13) Within material groups, same superscript letters indicate statistically different value (p 0.05).

56 54 Operative Dentistry Surface Roughness in µm Average Surface Roughness of Resin Composite Materials Hybrid Microfilled 0.3 Packable Baseline C-paste M-paste OG SB OG + SB Figure 1. Average surface roughness of resin composite materials. posite materials at baseline polish. Although additional polishing appears to roughen the baseline surface in both the polyacid-modified composite and resin-modified glass ionomer treatment groups, no significant difference in surface roughness was determined from baseline. Further polishing on baseline surfaces of the packable and hybrid composite with prophy paste resulted in significantly rougher surfaces. Additional polishing with One-Gloss, either alone or in combination with SuperBuff, significantly increased roughness in the hybrid composite material compared to the baseline surface. No significant difference in surface roughness from baseline was determined for the microfilled resin compos- Table 4. Mean Ra Values (µm) and Standard Deviations for the Various Materials and Polishing Systems Evaluate Following Simulated Brushing Polishing Restorative Material Method Hybrid Microfilled Packable Polyacid Resin-Modified Composite Composite Composite Modified Glass Ionomer Composite Coarse paste 0.46 (0.19) 1.45 (1.24) 0.38 (0.13) 0.30 (0.16) 0.44 (0.14) Medium paste 0.48 (0.22) 1.92 (0.89) 0.24 (0.13) 0.25 (0.11) 0.35 (0.13) OneGloss 1.00 (0.89) 1.88 (1.10) 0.18 (0.08) 0.22 (0.05) 0.37 (0.24) SuperBuff 0.51 (0.26) 2.34 (1.60) 0.18 (0.15) 0.22 (0.06) 0.44 (0.22) OneGloss + SuperBuff 0.28 (0.14) 1.67 (1.49) 0.31 (0.10) 0.21 (0.10) 0.39 (0.11) Table 5. Mean Ra Values (µm) and Standard Deviations for the Various Materials and Polishing Systems Evaluated Following Re-polish After Simulated Brushing Polishing Restorative Material Method Hybrid Microfilled Packable Polyacid Resin-Modified Composite Composite Composite Modified Glass Ionomer Composite Coarse paste 0.35 (0.14) 2.20 (1.24) 0.45 (0.11) 0.42 (0.25) 0.49 (0.19) Medium paste 0.47 (0.18) 1.43 (1.12) 0.73 (0.20) 0.39 (0.17) 0.67 (0.17) OneGloss 0.38 (0.15) 1.10 (0.82) 0.42 (0.13) 0.42 (0.18) 0.49 (0.20) SuperBuff 0.30 (0.20) 2.05 (1.19) 0.32 (0.11) 0.26 (0.90) 0.56 (0.23) OneGloss + SuperBuff 0.35 (0.15) 1.45 (1.06) 0.25 (0.07) 0.37 (0.14) 0.60 (0.17)

57 Neme & Others: Effect of Polishing on Surface Roughness of Esthetic Restorative Materials 55 ite material except with the application of OneGloss, which resulted in a significantly rougher surface than the other surface finishes (Figure 1). Table 4 represents the mean Ra values and standard deviations of each treatment group following simulated brushing. The microfilled and hybrid resin composite yielded significantly rougher surfaces following toothbrushing than the other three materials. Figures 2a-2e reveal the average surface roughness of each material following the three treatments (initial polish, toothbrushing and re-polish). Figures 2a (hybrid composite) and 2b (microfilled composite) demonstrate the increased roughness following brushing compared to Figures 2c-2e, where no apparent increase in roughness is demonstrated. Mean Ra values and standard deviations for each treatment group following re-polishing with one of five standardized polishing protocols are listed in Table 5. Repeated measures ANOVA yielded significant differences (p=0.05) in surface roughness among restorative materials and polishing treatment. DISCUSSION The clinicians objective for placement of esthetic restorations is to achieve the smoothest surface that will minimize plaque and stain retention and be easily maintained. Considerable emphasis has been placed on the surface roughness of various restorative materials following initial polishing. Mechanical profilers or profilometers have been employed to measure surface roughness for in vitro investigations. Although the profilometer provides somewhat limited two-dimensional information, an arithmetic average roughness is calculated and used to represent various material/polishing surface combinations that assist clinicians in treatment decisions. In this study, five resin-based restorative materials were evaluated for surface changes following initial polishing with the Sof-Lex disk series. As with similar studies (Hoelscher & others, 1998; Hondrum & Fernandez, 1997), the resin composite materials exhibited a statistically smoother surface than the polyacid-modified composite and resinmodified glass ionomer. These results support the findings that resin composites lead the other resinbased materials in surface smoothness (Note: esthetics per se was not evaluated in this study) following initial polishing. In addition, there was no significant difference in surface roughness among the three resin composite materials following initial polishing with the Sof-Lex disk series. This polishing Roughness µm Hybrid Composite Tx 1 Tx 2 Tx 3 Treatment Figure 2a. Hybrid composite. Roughness µm Microfilled Composite Tx 1 Tx 2 Tx 3 Treatment Figure 2b. Microfilled composite. Roughness µm Packable Composite Tx 1 Tx 2 Tx 3 Treatment Figure 2c. Packable composite. Coarse Medium OneGloss SuperBuff OG+SB Coarse Medium OneGloss SuperBuff OG+SB Coarse Medium OneGloss SuperBuff OG+SB

58 56 Operative Dentistry Roughness µm Polyacid-Modified Composite Tx 1 Tx 2 Tx 3 Treatment Figure 2d. Polyacid modified composite. Roughness µm Resin-Modified Glass Ionomer Tx 1 Tx 2 Tx 3 Treatment Figure 2e. Resin-modified glass ionomer. Coarse Medium OneGloss SuperBuff OG+SB Coarse Medium OneGloss SuperBuff OG+SB technique has been recommended with anterior restorations (Herrgott & others, 1989), although posterior usage may be compromised due to difficult access. Serio & others (1988) reported that surface roughness is increased in composite materials with subsequent use of prophylaxis pastes. In this investigation, both the hybrid and packable resin composite yielded a significantly rougher surface following application of prophylaxis paste compared to their baseline (Sof-Lex polish) surface. Although additional polishing with prophylaxis paste increased the surface roughness of resin-modified, polyacid-modified and microfilled composite, the values were not significantly different from baseline values. However, additional treatment of the microfilled composite surface with OneGloss resulted in significantly more roughness than the other polishing treatments among the microfilled composite group. Hondrum & Fernandez (1997) suggest that one variable affecting final surface polish of restorations may be flexibility of the backing material in which the abrasive is embedded. Of the finishing materials tested, OneGloss demonstrated the least flexibility. Unfortunately, there have been a limited number of studies investigating the subsequent care and maintenance of esthetic restorations following placement. Some agree that toothbrush abrasion can occur on all restoration surfaces (Hefferren, 1976; Momoi & others, 1997; Settembrini, Penugonda & Fischer, 1993). For years, it has been accepted that during toothbrushing, the matrix supporting inorganic filler particles in composite materials may wear away and leave particulate matter or irregularities projecting from the surface (Aker, 1982). In this study, both microfilled and hybrid resin composite yielded significantly rougher surfaces following simulated brushing compared to packable composite, polyacid-modified composite and resin-modified glass ionomer. The lower resin ratio in the packable composite compared to the traditional hybrid and microfilled composites may be responsible for the resistance to roughening. Also, the more favorable modulus of elasticity may play a role in the resistance to toothbrush abrasion in packable composite material. However, in the case of the resin-modified glass ionomer, the results are in contrast to Momoi & others (1997), who observed greater surface roughness in a resin-modified glass ionomer compared to a hybrid composite. The authors comment that the visible porosity following toothbrushing could have been associated with hand mixing of the powder and liquid. This investigation used pre-encapsulated material with no visible signs of porosity associated with the delivery system. All esthetic restorations require ongoing maintenance that may include periodic repolishing (Nash, 1991; Goldstein 1989). The life expectancy of esthetic restorations can be affected by improper home-care (Strassler & Moffitt, 1987) or aggressive prophylaxis treatments (Goldstein & others, 1992). Current methods for home-care maintenance include brushing, flossing, oral irrigation and interdental stimulation. Although effective oral hygiene techniques and dietary awareness may help reduce surface discoloration, some staining will eventually occur. Most often, surface staining of esthetic restorations can be removed with a thorough prophylaxis. If stains cannot be removed during routine recall or if an increase in surface roughness occurs, subsequent re-surfacing may be required to enhance the appearance and increase the longevity of the esthetic restoration. To simulate a routine prophylaxis, both a coarse and medium grit polishing paste were evaluated. A relatively new polishing paste

59 Neme & Others: Effect of Polishing on Surface Roughness of Esthetic Restorative Materials 57 impregnated on a flexible disk (SuperBuff) was also evaluated by itself and following resurfacing with a resin-impregnated polishing disk (OneGloss). Overall, there were no statistically significant differences among the polishing treatments evaluated. Although no significant differences among polishing systems were determined following toothbrushing, some important trends have been noted. Surface retreatment for hybrid composite materials resulted in a similar surface for each polishing system compared with post-brushed surfaces, with the exception of samples in the OneGloss group. The samples re-surfaced with OneGloss following simulated toothbrushing improved, yielding a reduction in overall surface roughness, although this group had a higher surface roughness following simulated brushing (1.00 µm) compared to others in the hybrid composite group (range µm). This protocol may prove to be beneficial if additional re-surfacing is required on restorations that present with increased roughness. Each sample in the microfilled composite group decreased in roughness following re-treatment except for those in the coarse prophy paste group. Microfilled samples re-surfaced with coarse prophy paste following simulated toothbrushing resulted in an increase in overall roughness values. CONCLUSIONS This in vitro investigation evaluated the effect of five finishing/polishing protocols that could be implemented at recall on the surface roughness of five direct esthetic materials. The three resin composite materials (hybrid, microfilled and packable) yielded significantly less roughness following initial polishing than that of the polyacid-modified composite (compomer) and resin-modified glass ionomer (hybrid ionomer). An increase in surface roughness was reported for each material investigated following additional polishing. However, only the resin composite materials yielded significant differences between initial polishing and additional polishing protocols. Both the hybrid and microfilled composite materials demonstrated a significant increase in roughness following simulated toothbrushing. Application of the polishing protocols following simulated toothbrushing on the hybrid and microfilled composite materials decreased their overall surface roughness. As with other in vitro investigations, additional research is required to: (1) evaluate operator effect on surface finish with various protocols/material combinations, (2) simulate clinical situations under different physical treatments, that is, wear and (3) substantiate the clinical relevance of the results of the current study through clinical trials. Acknowledgements The authors express their sincere appreciation to Drs Frank E Pink and John M Powers for their assistance with the statistical analysis of this investigation. (Received 27 March 2001) References Aker J (1982) New composite resins: Comparisons of their resistance to toothbrush abrasion and characteristics of abraded surfaces Journal of the American Dental Association 105(4) Balshi T (1986) Hygiene maintenance procedures for patients treated with the tissue-integrated prosthesis (osseointegration) Quintessence International 17(2) Barbakow F, Lutz F & Imfeld T (1987) Relative dentin abrasion by dentifrices and prophylaxis pastes: Implications for clinicians, manufacturers, and patients Quintessence International Berry LL, Berry EA & Powers JM (1994) Prophylaxis abrasives affect surface roughness of composites and hybrid ionomers Journal of Dental Research Abstract #948. Fruits TJ, Miranda FJ & Coury TL (1996) Effects of equivalent abrasive grit sizes utilizing differing polishing motions on selected restorative materials Quintessence International 27(4) Goldstein RE (1989) Finishing of composites and laminates Dental Clinics of North America 33(2) Goldstein RE & Goldstein CE (1988) Is your case really finished? Journal Clinical Orthodontics 22(11) Goldstein RE & Garber DA (1995) Maintaining esthetic restorations a shared responsibility Journal of Esthetic Dentistry 7(5) 187. Goldstein RE, Garber DA, Schwartz CG & Goldstein CE (1992) Patient maintenance of esthetic restorations Journal of the American Dental Association 123(1) Hefferren JJ (1976) A laboratory method for assessment of dentifrice abrasivity Journal of Dental Research 55(4) Herrgott AM, Ziemiecki TL & Dennison JB (1989) An evaluation of different composite resin systems finished with various abrasives Journal of the American Dental Association 119(6) Hoelscher DC, Neme AM, Pink FE & Hughes PJ (1998) The effect of three finishing systems on four esthetic restorative materials Operative Dentistry 23(1) Hondrum SO & Fernandez R Jr (1997) Contouring, finishing, and polishing Class V restorative materials Operative Dentistry 22(1) Lutz F, Sener B, Imfeld T, Barbakow F & Schupbach P (1993) Self-adjusting abrasiveness: A new technology for prophylaxis pastes Quintessence International 24(1) Momoi Y, Hirosaki K, Kohno A & McCabe JF (1997) In vitro toothbrush-dentifrice abrasion of resin-modified glass ionomers Dental Materials 13(2)

60 58 Operative Dentistry Nash LB (1991) Maximizing aesthetic restorations: The hygienist s role Practical Periodontics Aesthetic Dentistry 3(3) Rapley JW, Swan RH, Hallmon WW & Mills MP (1990) The surface characteristics produced by various oral hygiene instruments and materials on titanium implant abutments The International Journal of Oral and Maxillofacial Implants 5(1) Serio FG, Strassler HE, Litkowski LJ, Moffitt WC & Krupa CM (1988) The effect of polishing pastes on composite resin surfaces, A SEM study Journal of Periodontology 59(12) Settembrini L, Penugonda B & Fischer E (1993) Dentifrice abrasiveness on microfill composite resin and dentin: A comparative study Journal of Clinical Dentistry 4(2) Strassler HE & Baum G (1993) Current concepts in polishing composite resins Practical Periodontics Aesthetic Dentistry 3(1) Strassler HE & Moffitt WC (1987) The surface texture of composite resin after polishing with commercially available toothpaste Compendium 8(10) Weitman RT & Eames WB (1975) Plaque accumulation on composite surfaces after various finishing procedures Oral Health 65(12) Yap AU, Lye KW & Sau CW (1997) Surface characteristics of tooth-colored restoratives polished utilizing different polishing systems Operative Dentistry 22(3)

61 Operative Dentistry, 2002, 27, Effects of Delayed Polishing on Gap Formation of Cervical Restorations M Irie K Suzuki Clinical Relevance Delaying polishing for one day resulted in improved gap formation for cervical restorations of resin-modified glass ionomers. Delaying was not necessary for a compomer. SUMMARY This study evaluated the effect of polishing after one-day storage in water on the gap formation around a Class V restoration completely bordered by enamel (coronal cavity) using one resin-modified glass ionomer, one compomer and one conventional glass ionomer as a control. The study also examined gap formation of these materials in two different cervical restorations a cervical cavity incisally bordered by enamel and cervically bordered by dentin and a root surface cavity completely bordered by dentin. When the specimens of the two types of glassionomer material were polished immediately after the setting procedure, gaps around the coronal restorative cavity were observed. In contrast, only gaps were observed when the specimens were polished after one-day storage. Significant differences in the two glass-ionomer restorative materials were observed between immediate polishing and polishing after one-day storage. The compomer not show this pattern. Department of Biomaterials, Okayama University Graduate School of Medicine and Dentistry, , Shikata-cho, Okayama , Japan Masao Irie, DDS, PhD, assistant professor, Department of Biomaterials Kazuomi Suzuki, PhD, professor and chair, Department of Biomaterials Restorations placed in enamel/dentin and alldentin gave results similar to those in all-enamel. INTRODUCTION Restoration of cervical carious and abrasion/erosion lesions has been evaluated using conventional glassionomer cement (GIC), resin-modified glass-ionomer cement (RMGIC) and polyacid-modified resin composite (compomer) (Ferrari & others, 1998; Gladys & others, 1998; Tyas, 2000; Folwaczny & others, 2000a; Folwaczny & others, 2000b; Di Lenarda & others, 2000; Brackett & others, 2001). Conventional glass ionomer cement has several beneficial properties, including physicochemical bonding to tooth substrate, fluoride release and uptake and tooth color. However, it also demonstrates brittle fracture, erosion and wear in the oral environment (Wilson & McLean, 1988). In an attempt to correct these deficiencies, two types of combination materials, RMGIC and compomer, have been developed. RMGIC is unlike a light-cured resin composite or a GIC. It has a dual setting reaction that consists of an acid-base reaction and a photochemical polymerization process. The final set materials have been described as having a complex structure in which glass particles are sheathed in a matrix consisting of two networks one derived from the glass ionomer, the other from the resin (Wilson, 1990; Mitra, 1991). In these dual-setting systems, the resin reinforcement gives higher mechanical strength (Uno, Finger & Fritz, 1996; Irie & Suzuki, 1999; Irie & Suzuki, 2000; Yap & others, 2001) and higher bond strength to tooth sur-

62 60 Operative Dentistry faces compared to GICs. RMGIC claims to have an improved marginal seal by hygroscopic expansion (Sidhu, Sherriff & Watson, 1997) and improved bonding ability (Fritz, Finger & Uno, 1996a,b; Irie & Suzuki, 1999; Irie & Suzuki, 2000; Irie & Suzuki, 2001) after storage in water. A compomer may be either a single- or two-component material that contains either or both of the essential components of a glass ionomer; however, the components do not react as part of the setting process (McLean, Nicholson & Wilson, 1994). It consists of a resin matrix and a fluoroaluminosilicate glass filler. The monomer in the resin matrix ionizes following water uptake during storage after light-activation. The released hydrogen ions then react with the glass filler to initiate an acid-base reaction. Ionic cross-linking also occurs and fluoride is released (Hammesfahr, 1994). Although the marginal seal may be improved by enhancing the bond ability, it does not undergo dimensional change by expansion and does not improve the marginal seal after storage in water for a day (Irie & Suzuki, 2000). Cavity preparation locations serve as other factors that may influence the seal ability around a restoration. A cervical restoration may have margins completely bordered by enamel, completely bordered by dentin or bordered partially by enamel and partially by dentin (Davidson & Kemp-Scholte, 1989). Although a RMGIC and a GIC showed consistent bonding ability to both enamel and dentin, a compomer showed better bonding ability to dentin, alone (Irie & Suzuki, 2000; Di Lenarda & others, 2000). Currently, no data is available regarding the gap formation behavior around cervical lesions of a RMGIC or a compomer. In this study, the effect of storage in water on the gap formation around cervical restorations bordered entirely by enamel was evaluated using a RMGIC and a compomer. The gap formation around cervical restorations entirely in dentin and with margins in both enamel and dentin was also examined. Table 1: Restorative Materials Investigated METHODS AND MATERIALS The basic properties of one RMGIC, one compomer and one GIC are summarized in Tables 1 and 2. All processing of these materials was carried out according to the manufacturer s instruction. For light activation of the priming and restorative materials, a curing unit (New Light VL-II, GC, Tokyo, Japan, irradiated diameter: 13 mm) was used. Human premolars extracted for orthodontic reasons were used for the experiment. After extraction, each tooth was immediately stored in cold distilled water at 4 C for one-to-two months before testing. For each material and storage period, 10 specimens were made. All procedures except for cavity preparation and mechanical testing were performed in a thermo-hygrostatic room kept at 23±0.5 C and 50±2% relative humidity. The Cavity Region: Coronal Cavity Restoration Cavity preparations were placed on the facial surface of premolars. A cylindrical cavity was prepared with a tungsten carbide bur (200,000-rpm) and a fissure bur (8,000-rpm) under wet conditions to a depth of 1.5 mm with a diameter of 3.5 mm. A cavity preparation was placed parallel to the cemento-enamel (CEJ) with the preparation extended 1.0 mm above the CEJ (Figure 1, completely bordered by enamel: Coronal cavity). Cavosurface walls were finished to a butt joint. One cavity was prepared in each tooth. In total, cavities were prepared for this study in 150 teeth. The prepared cavity surface was treated with the conditioner/primer according to each manufacturer s Material Manufacturer Batch # Type (Powder/Liquid)/Components Fuji II LC GC Corp P: RMGIC (3.0g/1.0g) Tokyo, Japan L: P: fluoroaluminosilicate glass L: copolymer of acrylic and maleic acid, HEMA, water Dyract DeTrey/Dentsply B Compomer Konstanz, Germany fluoroaluminosilicate, polyacrylic acid, UDMA Fuji II GC Corp P: Conventional GIC (2.7g/1.0g) Tokyo, Japan L: P: fluoroaluminosilicate glass L: copolymer of acrylic and maleic acids, polybase carboxylic acid, water Key: GIC, Glass ionomer cement; HEMA, 2-Hydroxtethyl methacrylate; UDMA, urethane dimethacrylate Table 2: Conditioner/Primer Agents Investigated Material Manufacturer Batch # Components and Surface Treatment Dentin Conditioner GC Corp Polyacrylic acid, water Tokyo, Japan Apply with brush 20 seconds rinse 15 seconds gently dry 5 seconds Dyract-PSA DeTrey/Dentsply PENTA, TEGDMA, acetone Konstanz, Germany Apply 30 seconds gently dry 5 seconds light cure 10 seconds Key: PENTA, dipentaerythritolpenta-acrylatemonophosphate; TEGDMA, tri-ethylene-glycol dimethacrylate

63 Irie & Suzuki: Effect of Delayed Polishing on Cervical Restorations 61 Three kinds of cervical restoration and each measured point in the cervical restorations. E: Enamel substrate, D: Dentin substrate instructions. Dentin Conditioner was applied for 20 seconds and rinsed with water. An ample amount of Dyract- PSA was applied and left undisturbed for 30 seconds. Excess solvent was removed by compressed air. The light curing time was 10 seconds. A second layer of Dyract-PSA was applied with a brush. The same conditioning/ priming pretreatment was performed on enamel as for dentin. The cavity was filled with mixed materials using a syringe tip (Centrix C-R Syringe System, Centrix Inc, Shelton, CT 06464, USA) and covered with a plastic strip and hardened. Fuji II LC and Dyract were exposed to a visible light source with irradiation times of 20 and 40 seconds, respectively. Fuji II was stored in an incubator at 37 C at 100% relative humidity for four minutes after mixing. Inspection Procedure #1: Immediately after light curing or setting, each tooth was sectioned in a buccolingual direction through the center of the restoration with a low-speed diamond saw (Isomet, Buehler Ltd, Lake Bluff, IL 60044, USA). Then, the presence or absence of marginal gaps was measured at 14 points (each 0.5 mm apart) along the cavity restoration interface (n=10; total points measured = 140). This was done with a traveling microscope (x1000, Measurescope, MM-11, Nikon, Tokyo, Japan) positioned parallel with the cavity wall and bottom on each half of the sample (Figure 1). The number of gaps in each sample was totaled and expressed as the sum of each sample. Inspection Procedure #2: Immediately after light curing or setting, the surface of the restorations was polished with abrasive points (Silicone Mide, Shofu, Kyoto, Japan) and rinsed with distilled water. Measuring a specimen involved the same procedure as described above. Inspection Procedure #3: After polishing and inspecting as described above (Inspection Procedure #2), the specimen was stored in distilled water at 37 C for one day. Then, the presence of gaps was re-inspected as described above. Inspection Procedure #4: The specimen was stored in distilled water at 37 C for one day after light curing or setting. Next, the surface of the restorations was polished as described above. Then, the presence of gaps was inspected as described above. The Cavity Region: Cervical and Root Surface Restorations: Cylindrical cavities were prepared for both groups with the same instrumentation and techniques used for the coronal preparations. Cavosurface walls were finished to a butt joint. Previous work by the authors (Irie & Suzuki, 2000) had confirmed that delayed polishing enhanced gap formation so the routine established for inspection procedure #4 was utilized on these restorations. Cervical Cavity: A cavity preparation was placed parallel to the cemento-enamel junction with the center at the cemento-enamel junction (Figure 1, incisally bordered by enamel, cervically by cement or dentin, Cervical cavity). Root Surface Cavity: A cavity preparation was placed parallel to the cemento-enamel junction with the preparation extending 1.0 mm below the cementoenamel junction (Figure 1, completely bordered by cementum or dentin, Root Surface Cavity). Inspection Procedure #4: The restorative, polishing and inspection procedures were performed as described for the Inspection Procedure #4. The results were statistically analyzed using Duncan s New Multiple-Range Test (non-parametric) (Conover & Iman, 1981; Irie & Suzuki, 2000). RESULTS Table 3 summarizes data for gap formation of the RMGIC observed in cervical restorations with various cavity regions and inspection procedures. With the two inspection procedures, cutting and immediate inspection after the setting (#1 and #2), 86 and 101 gaps around the restorative cavities, respectively, were

64 62 Operative Dentistry Table 3: Effect of Cavity Region and Inspection Procedure on Gap Formation Around Restoration: RMGIC (Fuji II LC) Number of Specimens Showing Gaps Coronal Axial Cervical Sum Cavity region Inspection Procedure Coronal # (A) Coronal # (A) Coronal # (A) Coronal # (B) Cervical # (B) Root surface # (B) #1: light-activation (20 seconds) cut inspection (not polished) #2: light-activation (20 seconds) polish cut inspection #3: #2 storage in water for one day re-inspection #4: light-activation (20 seconds) storage in water for one day polish cut inspection N=10 (total measuring points, 1 ~ 14 = 140) Values with the same letters were not significantly different by Duncan s New Multiple-Range Test (Conover & Iman, 1981) (p>0.05) Table 4: Effect of Cavity Region and Inspection Procedure on Gap Formation Around Restoration: Compomer (Dyract) Number of Specimens Showing Gaps Coronal Axial Cervical Sum Cavity region Inspection Procedure Coronal # (A) Coronal # (A) Coronal # (A) Coronal # (A) Cervical # (A) Root surface # (A) #1: light-activation (20 seconds) cut inspection (not polished) #2: light-activation (20 seconds) polish cut inspection #3: #2 storage in water for one day re-inspection #4: light-activation (20 seconds) storage in water for one day polish cut inspection N=10 (total measuring points, 1 ~ 14 = 140) Values with the same letters were not significantly different by Duncan s New Multiple-Range Test (Conover & Iman, 1981) (p>0.05) Table 5: Effect of Cavity Region and Inspection Procedure on Gap Formation Around Restoration: GIC (Fuji II) observed. When the same specimens as described above (#2) were stored in distilled water for one day, then reinspected (#3), 74 gaps around the restorative cavities were observed. No significant differences among the three conditions were observed. The severest points, 1 and 14, showed the most gaps in the three conditions. The cervical corner area, 9~11, also showed many gaps. The axial regions showed half the number of gaps in the same conditions. However, when the specimen was cut and inspected after storing in water for one day (#4), 17 gaps around the c o r o n a l restorative cavities were observed. Significant differences were observed among the three conditions (#1, #2 and #3) and the three #4 conditions. No significant differences among the sum of gaps of the three different cavity region restorations (#4) were observed. However, when the cavity region was coronal, the polished point, 14 (enamel substrate), showed half the number of gaps in the restorative cavities. The cervical corner, 11, also showed Number of Specimens Showing Gaps Coronal Axial Cervical Sum Cavity region Inspection Procedure Coronal # (AB) Coronal # (A) Coronal # (B) Coronal # (C) Cervical # (C) Root surface # (C) #1: light-activation (20 seconds) cut inspection (not polished) #2: light-activation (20 seconds) polish cut inspection #3: #2 storage in water for one day re-inspection #4: light-activation (20 seconds) storage in water for one day polish cut inspection N=10 (total measuring points, 1 ~ 14 = 140) Values with the same letters were not significantly different by Duncan s New Multiple-Range Test (Conover & Iman, 1981) (p>0.05)

65 Irie & Suzuki: Effect of Delayed Polishing on Cervical Restorations 63 many gaps. When the cavity region was the root surface, there were 2-3 gaps in all the inspected points in the axial region 5~10. Table 4 summarizes the data for gap formation of the compomer observed in cervical restorations with various cavity regions and inspection procedures. When the four inspection procedures were compared, the #1 procedure was 18, the #2 was 38, the #3 was 40 and the #4 was 35, respectively. No significant differences were observed among the four sum numbers of gaps in the coronal cavity-restoration interfaces. There was no effect for delaying polishing for one day for the compomer. When the restoration surface was not polished (#1), no gaps were observed at the two polished points 1 and 14; however, most gaps appeared after the polishing procedure in the three conditions (#2, #3 and #4), respectively. The authors observed several gaps nearer points 2 and 13. No significant differences were observed among the sum of gaps of three different cavity region restorations (#4 conditions). Namely, there was no significant difference between both cavity regions and inspection procedures. The enamel margins 1 and 14 of the coronal cavity and #4 inspection procedure condition, and 1 of the cervical cavity and #4 inspection procedure condition, showed nine gaps in 10 specimens. However, the dentin margin showed little or no gap in the two conditions. Table 5 summarizes data for gap formation of Fuji II observed in cervical restorations with various cavity regions and inspection procedures. With both procedures cutting and immediate inspection after the setting (#1 and #2) 104 and 117 gaps around the restorative cavities, respectively, were observed. When the same specimens as described above (#2) were stored in distilled water for one day, then re-inspected (#3), 87 gaps around the restorative cavities were observed. The #3 condition was a significant improvement in gap formation compared with that of #2. The severest points, 1 and 14, showed the most gaps in the three conditions. The cervical corner area, 9~11, also showed more gaps. The axial regions showed more gaps in the same conditions. However, when the specimen was cut and inspected after storing in water for one day (#4), only seven gaps were observed around the restorative cavities. Significant differences were observed among the three conditions (#1, #2 and #3) and the three #4 conditions. No significant differences were observed among the sum of gaps of the three different cavity region restorations (#4). The polished points, 1 and 14, showed few gaps in the three restorative cavities. The presence of gaps was almost zero at all inspected points in the axial region 5~11. DISCUSSION This study clearly demonstrated that polishing two filled glass ionomers (RMGIC & GIC) should not be performed immediately after filling and setting in the coronal cavity. For example, it demonstrated that polishing (#2) or not polishing (#1) did not prevent gap formation immediately after setting. Polishing/unpolishing should be delayed to a later time to prevent gap formation at the material-tooth cavity interface. In contrast to the presence of approximately of 140 (total measured points) gaps at the material-tooth cavity interface in the coronal cavity of specimens polished immediately after setting, the gap was near zero when the specimen was polished after storage in water for one day. RMGIC or GIC shrinks during setting reaction. A gap was formed as the adhesion between the tooth cavity and glass ionomer did not resist the stress formed by cement shrinkage (Feilzer, de Gee & Davidson, 1988; Sidhu, 1994). One reason for this dependence of the gap on the storage period may be the hygroscopic expansion of the glass ionomer due to an apparent correlation of the marginal gap in the tooth cavity with the marginal gap in a Teflon mold observed for the condition after one-day storage (Irie & Suzuki, 2000). This effect was reported for the uptake of water by the matrix of RMGICs forming a poly-hema complex (Wilson, 1990). In addition, GIC forms a hydrogel of calcium and aluminum polyacrylates by the uptake of water (Wilson & McLean, 1988). After one day of water storage, the curing contraction stresses of the materials are effectively compensated for or even converted into expansion stress due to water uptake and swelling (Feilzer & others, 1995). Water absorption of RMGICs and GICs reportedly affects cavity adaptation and reduces microleakage (Irie & Nakai, 1987; Fritz & others, 1996b; Irie & Suzuki, 2000). Although hygroscopic expansion may not be enough to compensate for the setting shrinkage, it plays an important role in reducing the shrinkage caused by the cement setting reaction and thus improves the marginal seal (Sidhu & others, 1997; Irie & Suzuki, 2001). The cement is expected to show higher bond and mechanical strengths when fully set rather than during setting reaction. It is suggested that the bond ability to the tooth substrate increases with the development of the glass ionomer/tooth substrate interaction during storage in water, and the cohesive strength of the cement itself improves with the setting process (Irie & Suzuki, 2000). The ph, an index of the degree of hardening reaction of set glass ionomer, is reported to be lower at the initial stage regardless of the type of cement, that is, GICs or RMGICs. The ph value of the set cement gradually increases for 24 hours (Tosaki & Hirota, 1994; Anusavice, 1996). Therefore, it can be presumed that completing the setting reaction of an RMGIC or GIC requires 24 hours. Thus, 24 hours are required until an RMGIC or GIC has adequate mechanical strength, which has a close relationship with the bond strength (Irie & Suzuki, 2000). RMGIC has a dual setting reaction: one is light-initiated cross-linking of

66 64 Operative Dentistry methacrylate groups similar to the setting of light-cured resin composites; the other is an acid-base reaction similar to that of a GIC (Wilson, 1990; Mitra, 1991). When the same specimens as described above (#2) were stored in distilled water for one day and re-inspected (#3), 74 (the RMGIC case) and 87 (the GIC case) gaps were observed around the restorations (Tables 3 and 5). No significant difference was observed between the two conditions (#2 and #3) of RMGIC. One possible reason is that the opened enamel margin could not be sealed by hygroscopic expansion during subsequent storage in water for a day because debris caused by the polishing procedure would be forced into the open gap area between the enamel margin and restorative material (Irie & Nakai, 1987). Another possibility was that gaps in the axial could not be prevented by subsequent storage in water for a day. Since axial gap was caused by setting shrinkage of the restorative material itself. The stress that occurred at the tooth substrate/restorative material interface was caused by sectioning immediately after light curing (Davidson, Leloup & de Gee, 1994), and the tooth substrate-restorative material interface was destroyed. However, a significant difference was observed between the two results (#2 and #3) of GIC. The reason is that the #2 result of GIC was too bad compared with that of RMGIC, although almost the same degree of improvement exists in the number of gaps of RMGIC and GIC, respectively. As a result, it was suggested that only hygroscopic expansion for one day did not prevent gaps in the axial. The cervical corner of the coronal cavity restorations showed more gaps than the coronal corner in RMGIC and GIC. This was expected, as bond strength to coronal dentin is usually higher than bond strength to cervical dentin because cervical dentin is a less favorable bonding substrate (Heymann & Bayne, 1993). The presence of gaps at the material-tooth cavity interface in the cervical and root surface cavities was near zero in 140 (total measured points) when the specimen was polished after storage in water for a day. The reasons for this are mentioned above and the RMGIC and GIC showed constant bonding abilities to both enamel and dentin (Irie & Suzuki, 2000). This study demonstrated that delaying the polishing of the compomer for a day does not prevent gap formation between the material and the cervical cavity. Therefore, it did not prevent bonding to the tooth structure and hygroscopic expansion of the compomer, itself (Irie & Suzuki, 2000). The marginal gaps observed even after the specimen was stored in water for one day indicated that the hygroscopic expansion does not fully compensate for the shrinkage caused by the setting reaction. The degree of hygroscopic expansion of the compomer was smaller than that of RMGICs because the composition of the compomer is similar to the resin composite system (Hammesfahr, 1994; Irie & Suzuki, 2000). The values of bond strength to enamel and dentin measured after one-day storage were not significantly higher than those measured immediately for the compomer (Irie & Suzuki, 2000). Since no micro-mechanical interlocking with conditioned enamel was available when Dyract-PSA was used, a low shear bond strength to enamel was shown even after storage for one day (Fritz, Finger & Uno, 1996a). This study demonstrated that the polishing period of the compomer showed no effect to prevent gaps around the cervical restoration, especially at the enamel margin. The gap width of the enamel margin was found to result from polishing procedures. The manufacturer suggested that tooth etching is unnecessary for most restorative procedures with Dyract. Although the recommended primer is considered adequate to condition tooth-hard tissues, it has been demonstrated that compomer restorations are lacking in their marginal adaptation (Ferrari & other, 1998; Tyas, 2000; Di Lenarda & others, 2000; Brackett & others, 2001). A modified formulation of Dyract is now available (Dyract AP, Dentsply/DeTrey, Konstanz, Germany). However, the manufacturer still considers etching unnecessary even for this material. To improve the enamel/compomer interface, the manufacturer s protocol should be modified, introducing enamel etching as achieving the compomer restorations clinical success. The restorative materials tested should not be polished at the placement appointment except for the compomer. Instead, the prepared cavity should be filled at the placement appointment, then polished at the next appointment (Irie & Suzuki, 1999 & 2000). CONCLUSIONS When specimens of the two types of glass-ionomer material, a RMGIC and a GIC, were polished immediately after the setting procedure, this study showed gaps around the cervical restoration (coronal cavity). In contrast, only gaps around the cervical restorations (coronal, cervical and root surface cavities), respectively, were observed when the specimens were polished after one-day storage. The compomer did not show this pattern. Acknowledgements The authors wish to thank GC and Dentsply/DeTrey for the free supply of materials. (Received 2 April 2001) References Anusavice KJ (1996) Phillips science of dental materials 10 th ed WB Saunders Co, Philadelphia

67 Irie & Suzuki: Effect of Delayed Polishing on Cervical Restorations 65 Brackett WW, Browning WD, Rose JA & Brackett MG (2001) Two-year clinical performance of a polyacid-modified resin composite and a resin-modified glass-ionomer restorative material Operative Dentistry 26(1) Conover WJ & Iman R (1981) Rank transformations as a bridge between parametric and nonparametric statistics The American Statistician 35(3) Davidson CL & Kemp-Scholte CM (1989) Shortcomings of composite resins in Class V restorations Journal of Esthetic Dentistry 1(1) 1-4. Davidson CL, Leloup G & de Gee AJ (1994) Self-repair of damaged glass ionomer cement Journal of Dental Research 73 Abstracts of papers p181 Abstract #634. Di Lenarda RD, Cadenaro M & DeStefano Dorigo E (2000) Cervical compomer restorations: The role of cavity etching in a 48-month clinical evaluation Operative Dentistry 25(5) Feilzer AJ, de Gee AJ & Davidson CL (1988) Curing contraction of composites and glass ionomer cements Journal of Prosthetic Dentistry 59(3) Feilzer AJ, Kakaboura AI, de Gee AJ & Davidson CL (1995) The influence of water sorption on the development of setting shrinkage stress in traditional and resin-modified glass ionomer cements Dental Materials 11(3) Ferrari M, Vichi A, Mannocci F & Davidson CL (1998) Sealing ability of two compomers applied with and without phosphoric acid treatment for Class V restorations in vivo Journal of Prosthetic Dentistry 79(2) Folwaczny M, Loher C, Mehl A, Kunzelmann KH & Hickel R (2000a) Tooth-colored filling materials for the restoration of cervical lesions: A 24-month follow-up study Operative Dentistry 25(4) Folwaczny M, Mehl A, Kunzelmann KH & Hickel R (2000b) Determination of changes on tooth-colored cervical restorations in vivo using a three-dimensional laser scanning device European Journal of Oral Science 108(3) Fritz UB, Finger WJ & Uno S (1996a) Resin-modified glass ionomer cements: Bonding to enamel and dentin Dental Materials 12(3) Fritz UB, Finger WJ & Uno S (1996b) Marginal adaptation of resin-bonded light-cured glass ionomers in dentin cavities American Journal of Dentistry 9(6) Gladys S, Van Meerbeek B, Lambrechts P & Vanherle G (1998) Marginal adaptation and retention of a glass-ionomer, resinmodified glass-ionomers and a polyacid-modified resin composite in cervical Class V lesions Dental Materials 14(4) Hammesfahr PD (1994) Developments in resinionomer systems in Hunt PR ed Glass Ionomers: The Next Generation Proceedings of the 2 nd International Symposium in Glass Ionomers pp Philadelphia, PA. Heymann HO & Bayne SC (1993) Current concepts in dentin bonding: Focusing on dentinal adhesion factors Journal of the American Dental Association 124(5) Irie M & Nakai H (1987) The marginal gap and bonding strength of glass ionomers Dental Materials Journal 6(1) Irie M & Suzuki K (1999) Water storage effect on the marginal seal of resin-modified glass-ionomers restorations Operative Dentistry 24(5) Irie M & Suzuki K (2000) Marginal seal of resin-modified glass ionomers and compomers: Effect of delaying polishing procedure after one-day storage Operative Dentistry 25(5) Irie M & Suzuki K (2001) Current luting cements: Marginal gap formation of composite inlay and their mechanical properties Dental Materials 17(4) McLean JW, Nicholson JW & Wilson AD (1994) Proposed nomenclature for glass-ionomer dental cements and related materials Quintessence International 25(9) Mitra SB (1991) Adhesion to dentin and physical properties of a light-cured glass-ionomer liner/base Journal of Dental Research 70(1) Sidhu SK (1994) Marginal contraction gap formation of lightcured glass ionomers American Journal of Dentistry 7(2) Sidhu SK & Watson TF (1995) Resin-modified glass ionomer materials A status report for the American Journal of Dentistry American Journal of Dentistry 8(1) Sidhu SK, Sherriff M & Watson TF (1997) The effects of maturity and dehydration shrinkage on resin-modified glass ionomer restorations Journal of Dental Research 76(8) Tosaki S & Hirota K (1994) Current and future trends for light cured systems in Hunt PR, ed Glass ionomers: The Next Generation Proceedings of the 2 nd International Symposium on Glass Ionomers pp Philadelphia, PA. Tyas MJ (2000) Three-year clinical evaluation of a polyacid-modified resin composite (Dyract) Operative Dentistry 25(3) Uno S, Finger WJ & Fritz UB (1996) Long-term mechanical characteristics of resin-modified glass ionomer restorative materials Dental Materials 12(1) Wilson AD & McLean JW (1988) Glass-Ionomer Cement Quintessence Publishing Co Chicago Wilson AD (1990) Resin-modified glass-ionomer cements International Journal of Prosthodontics 3(5) Yap AUJ, Mudambi S, Chew CL & Neo JC (2001) Mechanical properties of an improved visible light-cured resin-modified glass ionomer cement Operative Dentistry 26(3)

68 Operative Dentistry, 2002, 27, The Whitening Effect of Bleaching Agents on Tetracycline-Stained Rat Teeth DH Shin JB Summitt Clinical Relevance All tested bleaching agents had enough power to bleach Tetracycline stains if they were allowed direct contact with the stained area. In the case of severely stained teeth, extended bleaching times should be considered or, if veneers are planned, preparation should be accomplished prior to whitening to allow the bleach more intimate contact with the stained area. SUMMARY This study compared the whitening effect of three bleaching agents on the teeth of rats and demonstrated differences in bleaching where dentin was exposed or enamel was thin. Thirty Albino rats were peritoneally injected with tetracycline solution daily for two weeks. Thirty-two disc-shaped specimens were cut from the crowns of incisors removed from sacrificed rats and were irradiated with UV light for 16 hours. Sections were stored in saline. Eight sections served as controls and were not bleached. Three bleaching agents (Opalescence, Rembrandt and Dankook University School of Dentistry, 7-1 Shinbu-dong, Cheonan, Chungnam, , Republic of Korea DH Shin, DDS, MS, PhD, associate professor, Department of Conservative Dentistry JB Summitt, DDS, MS, professor and head, Division of Operative Dentistry, Department of Restorative Dentistry, The University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX Nite White) were applied to eight specimens each, five times a day for two weeks, and images of the sections were recorded at the following times: before bleaching (baseline), day 1, day 3, day 5, day 7, day 9, day 11 and day 14. Mean colors to demonstrate any change ( E) from baseline for each time period were as follows: control 9.78 (baseline), 9.17, 9.36, 9.65, 9.40, 9.99, 10.57, 11.36; Opalescence 10.08, (baseline) 7.63, 6.72, 6.04, 5.10, 4.87, 4.89, 4.27; Rembrandt 9.83 (baseline), 11.27, 9.55, 8.36, 7.75, 6.94, 7.11, 7.04; Nite White (baseline), 9.92, 7.58, 6.80, 5.45, 5.05, 4.73, All bleached teeth were lightened (p<.01). Another 56 tetracycline-stained rat incisors were UV irradiated for three days. Three different penetration depths were tested: penetration through lingual dentin and labial enamel (DN group), penetration through labial enamel only (RE group) and penetration through labial enamel covered with 1.0 mm human enamel (HE group). Specimens were bleached with Opalescence for one hour five times a day for one week or four weeks. A control group of unbleached teeth was also examined. Results ( E) were as follows: control 11.67; 1-week DN 13.55; 1-week RE 12.80;

69 Shin & Summitt: The Whitening Effect of Bleaching Agents on Tetracycline-Stained Rat Teeth 67 1-week HE 12.07; 4-week DN 7.48; 4-week RE 7.50; 4-week HE The color change in the 4- week DN and the 4-week RE groups showed the greatest reduction (p<.01). INTRODUCTION Tooth shade, which is the result of diffuse reflectance from the inner dentin through the outer translucent enamel layer (O Brien, 1985), is believed to be the most important factor in patients perceptions of dental attractiveness (Dunn, Murchison & Broome, 1996). Because tooth discoloration contributes to psychological distress and poor esthetics, various techniques have long been sought to regain natural tooth color. Using veneers or esthetic crowns to cover unesthetic teeth is one way to accomplish this. However, bleaching techniques are preferred because tooth color can be regained without removing sound, healthy tooth structure. Bleaching has become a routine, popular, conservative option with a high rate of success. Unfortunately, some regression does occur and not all tooth discoloration can be corrected. One such case is discoloration due to tetracycline (TC), which is very difficult to bleach. TC discoloration occurs when TC has been ingested at a young age or by the patient s mother during pregnancy. Mello (1967) found that TC chelates with calcium at the hydroxyapatite surface of mineralizing dentin to form tetracycline-orthophosphate that results in tooth discoloration. It was also noted that TC deposited in enamel or dentin is permanently positioned, while that absorbed in bone could be released through normal bone remodeling (Stewart, 1973). This might be one reason why this stain responds less to bleaching. The discoloration is seen predominantly in dentin and cementum, and only slightly in enamel. This is because of dentin s greater absorbance of TC due to a larger surface area of dentin apatite crystals than that of enamel apatite crystals (Urist & Ibsen, 1963). Several methods, including vital bleaching and intentional non-vital bleaching (Aldecoa & Mayordomo, 1992), have been used to lighten the color of TC-stained teeth. However, the extent of lightening was not comparable to that of non-tc-stained teeth with vital bleaching (Haywood & Heymann, 1989). In one study where teeth were devitalized and a non-vital bleaching technique used, long-term results were not favorable (Friedman, 1997). Furthermore, non-vital bleaching has been associated with a risk of external root resorption (Friedman, 1997; Neuvald & Consolaro, 2000). It is better to perform vital bleaching for TC-stained teeth rather than to devitalize the teeth and use a nonvital bleaching technique; however, vital bleaching agents currently in use have limited efficacy in severe TC stain cases. Haywood (1997), however, reported that extending vital bleaching time to six months would allow moderate to severe TC discoloration cases to respond to treatment, and the best indicator of a good prognosis is not the severity of the discoloration, but the location of the discoloration. Leonard & others (1999) also reported that shade stability of bleached TC stain cases may last at least 54 months after treatment. The reasons for the decreased response of TC-stained teeth may be that the bleaching potential of these agents on TC stain may be ineffective, or they may fail to sufficiently penetrate through the enamel into the dentin. This work compared the whitening effects of three bleaching agents and demonstrated differences in bleaching effect when the depth of penetration of bleaching agent was varied. The amount of color change was quantitated by computer-aided image processing instead of subjectively comparing before- and aftertreatment photographs (Haywood, Leonard & Dickinson, 1996; Haywood, Leonard & Nelson, 1993). METHODS AND MATERIALS Thirty Albino rats (wt gm) were peritoneally injected daily for two weeks with 6 ml of 5 mg/ml TC solution [Chlortetracycline (Teracyclincap, ChongKeun Dang, Korea)]. One week after the final injection, the rats were sacrificed by cervical dislocation, and teeth extracted and coded with random two-digit numbers. Thirty-two mandibular incisors were used in the first phase of this study, which compared the effectiveness of three vital bleaching agents. Fifty-six maxillary teeth were used in the second phase to identify the differences in bleaching effect when the depth within tooth structure of application of the bleaching agent was varied. Phase One: Specimen Preparation For the first phase, 32 teeth were sectioned to separate the coronal and root portions. Pulpal remnants were carefully removed with a barbed broach. The teeth were cleaned in an ultrasonic bath and stored in saline solution in a light-proof case to prevent color change from light exposure. For the first phase, the root and the cervical 1 mm of the stained crown were ground away, and a 1.5 mm thick disc was made by horizontal sectioning with a carborundum disc. Discs were polished with 600- grit sandpaper to prevent any effects from irregular texture (Johnston & Kao, 1989). The sections used in Phase One and the teeth used in Phase Two were exposed to ultraviolet (UV) light since it has been reported that Chlortetracycline-induced stains shift to the yellow-green hue (Shin & Cho, 1996), then gradually darken to brown or gray-brown with continued UV light exposure due to the formation of a red quinone product (Walton & others, 1982). To induce maximum discoloration, the 32 sections were irradiated for 16 hours with UV light (H-5000A, Hanshin

70 68 Operative Dentistry Figure 4. Representative images in phase 1 study; day 5. Figure 1. Schematic drawing shows the ways of preparing specimens. Figure 5. Representative images in phase 1 study; day 9. Figure 2. Representative images in phase 1 study; baseline. Figure 6. Representative images in phase 1 study; day 14. Figure 3. Representative images in phase 1 study; day 1. Medical Co, Seoul, Korea) and the 56 intact maxillary Chlortetracycline-stained teeth were irradiated for 72 hours. These irradiation times were determined to be necessary in a preliminary study to assure significant brown discoloration. Bleaching Procedures Phase 1 Effectiveness of Three Bleaching Agents The 32 discs were randomly divided into four groups of eight discs each, a control group and three bleaching groups. The bleaches used were 10% carbamide peroxide bleaching agents of three brands: Opalescence (Ultradent Products, South Jordan, UT 84095, USA), Rembrandt (Den-Mat, Santa Maria, CA 93456, USA) and Nite White (Discus Dental, Culver City, CA 90237, USA). The bleaching agents were applied to samples for one hour, five times a day, for 14 days. Teeth sections in the control group were stored in saline solution. After each one-hour immersion in the bleaching agent, the specimens were rinsed with tap water. After the fifth one-hour bleaching, the specimens were stored in saline solution until the next day. All bleaching was carried out in a light-proof case. Images of the teeth were taken to check the color space (L*, a*, b*) at the following times: baseline, days

71 Shin & Summitt: The Whitening Effect of Bleaching Agents on Tetracycline-Stained Rat Teeth 69 Table 1. Mean (SD) Baseline and E Values Between Stained and Non-stained Areas of All Groups 1, 3, 5, 7, 9, 11 and 14. Control Opalescence Rembrandt Nite White Baseline (4.18) (3.11) (3.70) (1.92) Day (3.93) (2.51) (3.42) (1.84) Day (4.14) (1.74) (2.77) (1.64) Day (4.23) (1.50) (2.16) (0.89) Day (3.74) (1.95) (1.63) (1.08) Day (5.05) (2.17) (2.05) (1.25) Day (5.33) (2.27) (2.10) (1.40) Day (4.82) (2.25) (2.35) (1.02) Figure 7. Representative images in phase 2 study. a) 1 week DN group; b) 1 week RE group; c) 1 week HE group; d) control group; e) 4-week DN group; f) 4-week RE group; g) 4-week HE group. Phase 2 Effect of Overlying Enamel Thickness Three different penetration depths were tested: penetration of bleach effect through lingual dentin and labial enamel (DN); penetration through labial enamel only (RE) and penetration through a 1.0 mm thickness of human enamel attached to labial rat enamel (HE). This also helped to determine whether increased bleaching is possible when enamel is thinned or the outer dentin is exposed, such as after teeth preparation for veneers in severely stained teeth. Because rat teeth have no lingual enamel, teeth in the DN group were exposed without alteration to the bleaching agent. This group was used to simulate human teeth with some exposed dentin. In the RE group, the exposed lingual dentin was covered with two coats of nail polish and wax so that the bleaching agent would penetrate into tooth structure through the labial enamel only. It was noted that rat labial enamel is composed of an outermost amorphous, aprismatic layer and an inner prismatic layer similar to human enamel, but thinner (Shin & Cho, 1996). This group was used to simulate human teeth in which some part of the labial enamel was thinned by tooth preparation for veneering, without dentin exposure. In the HE group, the bleaching agents had to penetrate into the rat teeth through a 1.0 mm thickness of human enamel that was attached to the labial surface using paper tape (Figure 1). Paper tape started from one side of the human enamel layer and surrounded the lingual part of the rat incisor, where the lingual surface was covered with nail polish and wax, and ended at the other side of the human enamel layer. This group was used to simulate intact human teeth with a more normal enamel thickness. The 56 maxillary intact teeth that had been UV irradiated for three days were randomly divided into seven groups. For the control group, eight teeth were sectioned (1.5 mm thick disc) without bleaching and their color spaces examined. The color spaces from the control teeth were used as baseline values. For the experimental groups, Opalescence was chosen as a bleaching agent because of its strong bleaching potential. Three groups of intact teeth, prepared as described above, were bleached for one hour five times a day for one week. In another three groups, this bleaching procedure was continued for four weeks. After sectioning and obtaining color spaces from bleached teeth, the bleached teeth were compared those in the control group. Analysis of the Color Spaces Tooth color was analyzed with the CIELAB system. This L*, a*, b* color is related to human color perception in all three dimensions or directions of color space.

72 70 Operative Dentistry L* represents the degree of gray and corresponds to value or brightness; a* is a hue-chroma parameter in the red-green direction; b* is a hue-chroma parameter in the blue-yellow axis. Images were captured via a stereomicroscope (15x) using a CCD camera (1k-642k, Toshiba, Japan). The computer software, HumanEye (Professional Scientific Instrument Co, Suwon, Korea) assessed the tooth color. The color difference ( E) between the stained area and the nonstained peripheral dentin area was measured rather than the change of color within the stained area itself. Color spaces at eight different randomly selected stained areas and eight different non-stained areas were measured and their mean values were estimated. Size of 5 x 5 pixels was used to analyze the color spaces at each area. Color difference ( E) between these two values was calculated by the equation: E = [( L*) 2 + ( a*) 2 + ( b*) 2 ] 1/2. Statistical Analysis For the first phase of this study, differences between color shifts of all groups were compared using a repeated-measures analysis of variance (MANOVA) and a Newman-Keuls Post Hoc Test. One-way Analysis of Variance (ANOVA) and a Least-significant Difference post hoc test were used to detect any differences among groups in the second phase of this study. Phase 1 RESULTS Table 1 shows the color differences ( E values) between stained areas and nonstained peripheral dentin areas of all groups. All bleached teeth were lightened in the stained dentin as well as regular dentin, and E values were reduced. The teeth in the control group showed relatively stable values. There was no difference between groups at the beginning of this experiment [Mean (SD)]: control [9.78 (4.18)], Opalescence [10.08 (3.11)], Rembrandt [9.83 (3.70)], Nite White [10.44 (1.92)] (Figure 2). Although the rate at which teeth lightened decreased with time, there was no significant difference between groups at day 1 (Figure 3). At day 5, Opalescence and Nite White bleaching agents showed significantly greater bleaching than the control and Rembrandt groups (Figure 4). Of the three bleaching agents, Rembrandt could not effectively bleach TC-stained teeth until day 9 (Figures 5 and 6) (Table 2). Table 3 shows that all three bleaching agents Table 2: Daily Change of p Values Among Groups p value effectively bleached TC-stained teeth, but Opalescence and Nite White were significantly more effective than Rembrandt (p<.01). Phase 2 Bleaching Effect (lower < higher) Baseline.977 Day Day Day Control, Rembrandt < Nite White, Opalescence Day Control, Rembrandt < Nite White, Opalescence Day Control < Rembrandt, Nite White, Opalescence Day Control < Rembrandt, Nite White, Opalescence Day Control < Rembrandt, Nite White, Opalescence Table 3: MANOVA Result Among Groups (Newman-Keuls Test) Control Opalescence Rembrandt Nite White Control Opalescence ** Rembrandt ** ** Nite White ** ** **p<.01 Table 4: E Values of All Groups in Phase 2 Study Groups Bleaching Time E Mean SD Control DN 1 week weeks 7.48* 2.23 RE 1 week weeks 7.50* 1.96 HE 1 week weeks *p<.05 Table 5: Statistical Differences Between One-Week and Four-Week Groups Mean Difference p value DN 1 week vs 4 weeks E RE 1 week vs 4 weeks HE 1 week vs 4 weeks Figure 7 shows representative images. Means and standard deviations of E values between stained areas and non-stained peripheral dentin areas in all seven groups are given in Table 4. After one week, there was no difference between groups. After four weeks of bleaching with Opalescence, the greater color changes were found in the groups of exposed dentin (DN) and thinned enamel (RE). Compared with the groups bleached for 1 week, the 4-week whitening effect was significantly better (p<.01) in all groups except the HE group (Table 5).

73 Shin & Summitt: The Whitening Effect of Bleaching Agents on Tetracycline-Stained Rat Teeth 71 DISCUSSION Conventional, photographic methods detected bleaching effect until instruments such as a colorimeter were developed (Rustogi & Curtis, 1994). That instrument converts all colors within the range of human perception into a common numerical code, allowing for a more quantitative comparison, and it is now being used more frequently. In this study, the image of rat teeth was captured through a stereomicroscope and CCD camera because rat teeth are too small to use the colorimeter. Chlortetracycline induced a yellow-green hue discoloration. The teeth were exposed to ultraviolet light as a means of obtaining more severe TC-discolored specimens of a brown or gray-brown hue from the formation of a red quinone product (Walton & others, 1982). A preliminary study revealed that TC-stained teeth were darkening to brown within a limited time, then, they were bleached. In this study, a 16-hour irradiation time with UV light in the first phase and 72 hours in the second phase were used. Bleached groups showed decreasing E values with time. Teeth were lightened (p<.05) and their chromas (degree of stain) reduced. Lenhard (1996) reported that the observed tooth color change after bleaching mainly resulted from a change in the color parameters L* and b* and that bleaching caused a shift in the blue direction within the color space and lightened the color of the teeth. Although these factors were not evaluated statistically because they were outside the objectives of this study, the same tendency could also be observed in the stained dentin and regular dentin: increased L*, decreased a* (shift toward the green direction) and decreased b* spaces (decrease in intensity of yellow). Compared to the control group, significantly bleached teeth were seen in all three bleaching groups. There was no significant difference between the Opalescence and Nite White groups, which both showed better bleaching than the Rembrandt group. Opalescence and Nite White showed differential bleaching effects starting with day 5, while Rembrandt showed effects starting with day 9 (Table 2). The bleaching effect of Rembrandt was slower and less effective. Among the three bleaching agents, Opalescence and Nite White showed a similar changing pattern of the E value, however, the day of greatest change differed with each group: day 1 was greatest for Opalescence and day 3 for Nite White. This demonstrated that Opalescence was a faster-acting agent. Another study also showed that carbamide peroxide penetration to the pulp varies significantly for various commercial bleaching products. This may result in different levels of tooth sensitivity or bleaching efficacy (Thitinanthapan, Satamanont & Vongsavan, 1999). Dental bleaching is designed to remove stains so that one could ultimately not discern the difference between the color spaces of bleached areas and normal areas. There are varying opinions regarding the amount of color difference needed to distinguish between the two colors. Kuehni & Marcus (1979) reported that color differences greater than 1.0 are perceivable by more than 50% of observers under uniformly controlled conditions. Johnston & Kao (1989) compared a visual rating scale with colorimetry of composite restorations and natural teeth. They found the average CIELAB color difference between tooth/restoration and restoration/ restoration pairs of 3.7 to be a visually acceptable match. After two weeks of bleaching in this study, Opalescence and Nite White achieved a E value of approximately 4.5. This color difference is detectable by either the 1.0 or 3.7 criterion. However, it is difficult to discern between any normal peripheral dentin area and the stained area after bleaching. Further study needs to determine how this 4.5 difference is perceived. From the second phase of this study, the effect of the thickness of enamel on bleaching was observed. Teeth bleached for one week showed no significant color change compared to the control group. This result differed from that of the first phase of this study. The difference might result because there was direct contact with a tooth structure in the first phase but not in the second. Even a little thickness of permeable dentin retarded the rate of penetration of the bleaching agent. Compared with the results of groups bleached for one week, significant color changes were found in the fourweek groups of exposed dentin (DN) and thinned enamel (RE) (p<.01), however, there was no difference in the HE group (covered with 1.0 mm thick human enamel). This means that increased bleaching effect would be expected with an increase in bleaching time, and the thick enamel might act as a powerful barrier to the vital bleaching agent, Opalescence. It could not penetrate sufficiently to bring about bleaching of deeply-stained enamel or dentin. Thus, when teeth are severely stained to the extent that routine bleaching procedures have no effect when the agent is applied to the surface, it may be possible to whiten them by exposing the deeper enamel directly to the bleaching agent through veneer preparation without exposing dentin, if a veneer is planned. Simulating the intact normal thickness of the enamel layer was difficult to achieve in this study. In some instances, a rat tooth was attached to the human enamel layer using paper tape; however, a space between the two parts may have existed. In this case, molecular diffusion could not be expected, so the study tried to completely surround both structures with paper tape to allow agent retention between the parts. Luting agent might be used to unite both structures; however, the authors were concerned that the luting

74 72 Operative Dentistry agent would act as a barrier. More effective ways of simulating this kind of situation should be considered in order to obtain a more reliable result. CONCLUSIONS All bleaching agents demonstrated sufficient power to bleach TC stains; however, it was the ability of these agents to penetrate the tooth structure that determined clinical efficacy. In severely TC-stained cases that are not responsive to routine vital bleaching, extending bleaching time or enamel reduction for veneering without exposing underlying dentin may allow efficient penetration of bleaching agents into stained dentin. Acknowledgements This study was accomplished with research funds provided by the Korea Research Foundation, Support for Faculty Research abroad. (Received 3 April 2001) References Aldecoa EA & Mayordomo FG (1992) Modified internal bleaching of severe tetracycline discoloration: A 6-year clinical evaluation Quintessence International 23(2) Dunn WJ, Murchison DF & Broome JC (1996) Esthetics: Patient s perceptions of dental attractiveness Journal of Prosthodontics 5(3) Friedman S (1997) Internal bleaching: Long-term outcomes and complications Journal of the American Dental Association 128(Supplement) 51S-55S. Haywood VB (1997) Nightguard vital bleaching: Current concepts and research Journal of the American Dental Association 128(Supplement) 19S-25S. Haywood VB & Heymann HO (1989) Nightguard vital bleaching Quintessence International 20(3) Haywood VB, Leonard RH Jr & Dickinson GL (1996) Efficacy of six months of nightguard vital bleaching of tetracyclinestained teeth Journal of Esthetic Dentistry 9(1) 1-7. Haywood VB, Leonard RH Jr & Nelson CF (1993) Efficacy of foam liner in 10% carbamide peroxide bleaching technique Quintessence International 24(9) Johnston WM & Kao EC (1989) Assessment of appearance match by visual observation and clinical colorimetry Journal of Dental Research 68(5) Kuehni FG & Marcus RT (1979) An experiment in visual scaling of small color differences Color Research Applications Lenhard M (1996) Assessing tooth color change after repeated bleaching in vitro with a 10 percent carbamide peroxide gel Journal of the American Dental Association 127(11) Leonard RH, Haywood VB, Eagle JC, Garland GE, Caplan DJ, Matthews KP & Tart ND (1999) Nightguard vital bleaching of tetracycline-stained teeth: 54 months post treatment Journal of Esthetic Dentistry 11(5) Mello HS (1967) The mechanism of tetracycline staining in primary and permanent teeth Journal of Dentistry for Children 34(6) Neuvald L & Consolaro A (2000) Cementoenamel junction: Microscopic analysis and external cervical resorption Journal of Endodontics 26(9) O Brien WJ (1985) Double layer effect and other optical phenomena related to esthetics Dental Clinics of North America 29(4) Rustogi KN & Curtis J (1994) Development of a quantitative measurement to assess the whitening effects of two different oxygenating agents on teeth in vivo Compendium 17(Supplement) 631S-634S. Shin DH & Cho YB (1996) Tooth color and structure changes induced by tetracycline in rat Journal of the Korean Academy of Conservative Dentistry 21(2) Stewart DJ (1973) The re-incorporation in calcified tissues of tetracycline released following its deposition in the bone of rats Archives of Oral Biology 18(6) Thitinanthapan W, Satamanont P & Vongsavan N (1999) In vitro penetration of the pulp chamber by three brands of carbamide peroxide Journal of Esthetic Dentistry 11(5) Urist M & Ibsen K (1963) Chemical reactivity of mineralized tissue with oxytetracycline Archives of Pathology Walton RE, O Dell NL, Myers DL, Lake FT & Shimp RG (1982) External bleaching of tetracycline stained teeth in dogs Journal of Endodontics 8(12)

75 Operative Dentistry, 2002, 27, An In-Vitro Investigation of Variables Influencing Mercury Vapor Release from Dental Amalgam AL Neme BB Maxson JB Linger LJ Abbott Clinical Relevance Mercury vapor release from dental amalgam can be influenced by variables controlled by the clinician. However, there is no rationale for alteration in clinical behavior based on mercury vapor release alone. SUMMARY Controversy regarding patient exposure to mercury from dental amalgam is more than 150 years old. Researchers continue to investigate the amount of mercury vapor released from amalgam both in vivo and in vitro. In this investigation, an in vitro testing method previously described in the literature was used to quantify the effect of operator-controlled variables on mercury release from dental amalgam. The variables tested were alloy morphology (spherical, admixed or atomized irregular particle), operator skill (inexperienced, novice and expert), operator technique (overfill and evenly fill) and cavity design (standard Class I, double volume and double surface area). Preparations fabricated in sections University of Detroit Mercy School of Dentistry, Department of Restorative Dentistry, 8200 West Outer Drive, PO Box 19900, Detroit, MI Ann-Marie L Neme, DDS, MS, associate professor Barbara B Maxson, DDS, MS, associate professor Jackson B Linger, DMD, MS, associate professor Lawrence J Abbott, DDS, MBA, associate professor of acrylic rod were filled with dental amalgam, placed in 25 ml glass bottles and sealed. Mercury vapor concentrations were measured using a Jerome M-411 at specified times. Standardized mean concentrations for each time and total mercury released over time were calculated and analyzed with ANOVA and Tukey HSD. Statistically significant differences (α = 0.05) were identified for all variables tested. Total mercury vapor release was consistently found to be greater for admixed as compared to spherical amalgam. Amalgam restorations prepared by an inexperienced operator demonstrated statistically less mercury vapor than a novice or experienced clinician for both spherical and admixed morphologies. A statistically significant difference in mercury vapor using different condensation and carving techniques was found for the spherical amalgam but not for the admixed material. Restoration design demonstrated significant differences in total mercury vapor dependent on volume and exposed surface area of the amalgam restoration. In this in vitro investigation, mercury vapor release from amalgam was dependent on alloy morphology, operator experience, operator technique and restoration design.

76 74 Operative Dentistry INTRODUCTION Dental amalgam has effectively served the public for more than 150 years. Although amalgam has inferior esthetics to tooth-colored materials, it continues to be the preferred restorative material for direct application (Berry & others, 1998). Leinfelder (1993) attributes the success of amalgam to factors that traditionally have been limitations of resin-based materials; durability, sealing the restoration/tooth interface, ease of manipulation and finishing, and low technique sensitivity. Improvements in physical properties of dental amalgam during the past 30 years have been instrumental in its persistence. Two major developments in the early 1960s were primarily responsible for the improvement in amalgam characteristics. The copper content was increased in the alloy to produce an amalgam free of the gamma 2 phase. The increase in copper provides more substrate for a copper-tin reaction, greatly reducing or eliminating the tin-mercury (gamma 2) phase, which is the weakest and most corrosive product of the amalgamation reaction. Also, a spraying technique used in manufacturing alloys for jet engines was implemented in the manufacture of dental amalgam alloy. The resultant alloy particles were spherical and crack-free with a much cleaner surface than powders produced with previous methods. Amalgam has not lost favor in the eyes of clinicians or patients due to inferior physical properties. Rather, the two characteristics that best explain the decline in using amalgam as a restorative material are appearance and mercury content. As more reliable bonding agents and improved physical properties of resin composites have improved the efficacy of direct esthetic restorations for posterior teeth, amalgam use has declined. Clinicians currently have a variety of tooth-colored direct and indirect materials available for posterior restorations. Also, although literature supports the safety of dental amalgam (ADA Council on Scientific Affairs, 1998; Berry & others, 1998; Osborne & Albino, 1999) exposure to mercury is still a common concern voiced by many patients. During all phases of manipulation (Vimy & Lorscheider, 1985a; Berglund & others, 1988; Berglund, 1990; Olsson & Bergman, 1992; Berglund, 1993), it is well documented that mercury vapor is released from dental amalgam. Researchers have measured mercury emission from dental amalgam in expired air (Reinhardt & others, 1979; Svare & others, 1981; Reinhardt & others, 1983) and intraoral air of patients (Vimy & Lorscheider, 1985a; Mackert, 1987; Berglund & others, 1988). Both atomic absorption flameless spectrophotometry (Svare & others, 1981; Haikel & others, 1990; Berglund, 1990; Berglund, 1993) and detection of elemental mercury vapor by thin gold film sensors (Vimy & Lorscheider, 1985a; Vimy & Lorscheider, 1985b; Berglund & others, 1988; Ahmad & Stannard, 1990) have been used to evaluate intraoral mercury vapor. In addition, researchers have measured and evaluated the amount of mercury vapor release from amalgam in vitro (Ferracane, Hanawa & Okabe, 1992; Engle & others, 1992; Berdouses & others, 1995). Each investigation has furthered our knowledge in this area to support the continued used of amalgam as a safe, effective restorative material. The potential for adverse health effects from exposure to mercury from dental amalgam has raised concern among many members of the public, government and scientific community. No adverse health effects from dental amalgam have been substantiated, other than rare cases of allergies (WHO/FDI, 1995). Yet, efforts have been made to determine the amount of mercury vapor released from amalgam restorations and to reduce mercury exposure. This study investigated operator-controlled variables that may affect mercury vapor release from dental amalgam and suggested possible techniques to reduce exposure. An in vitro method (Neme, McLaren & O Brien, 1999; Neme, Wagner & O Brien, 1999) for measuring mercury vapor release from dental amalgam was used in the four independent, yet related studies, to evaluate these variables: alloy morphology, operator skill, operator technique and cavity design. METHODS AND MATERIALS For each study in this investigation, preparation of amalgam samples was carried out at room temperature on bench top. Amalgam was triturated according to manufacturer s directions in an Automix amalgamator (Kerr Corporation, Romulus, MI 48174, USA). Following condensation and carving, amalgam samples were stored in 25 ml glass bottles sealed with a rubber stopper as described in a previous study by Neme & others (1999). The bottles sat undisturbed at room temperature until mercury vapor measurements were taken at predetermined times. The Jerome 411 Gold Film Mercury Vapor Analyzer (Arizona Instrument Corp, Tempe, AZ 85281, USA) measured the mercury content in each vapor sample. Using a syringe, vapor samples were extracted through the rubber septum from each of the closed bottles (Figure 1). To ensure a representative sampling at each reading, the plunger of the syringe was raised and lowered twice prior to collecting the 1 cc of vapor. Vapor samples were injected into the sealed inlet valve of the Jerome M-411 and sampled for 10 seconds at room temperature at a flow rate of 750 ml/min. Prior to each daily measurement and throughout the experiment when the gold film became saturated, a film heat was performed per the manufacturer s suggestion. This procedure was done to remove residual mercury vapor from the gold film for accuracy in meas-

77 Neme & Others: Mercury Vapor Release from Amalgam 75 Figure 1. Jerome 411 Gold Film Mercury Vapor Analyzer, syringe and closed bottle. uring and to avoid saturating the gold film. Mercury vapor concentrations were standardized against calibrated temperature readings that compared actual and expected mercury vapor for a given temperature. The average mercury vapor release levels and the total amount of vapor released were calculated and the data was analyzed by ANOVA and Tukey HSD (α=0.05). When appropriate, pairwise comparisons were evaluated for each condition tested. Alloy Morphology Three amalgam alloys with similar chemistry and different morphology were analyzed; a high-copper spherical particle (Valiant, Ivoclar North America, Amherst, NY 14228, USA), a combination of high-copper spherical and high-copper lathe cut particle (Valiant PhD, Ivoclar North America) and an atomized irregular particle (Experimental, Special Metals Corp, Ann Arbor, MI 48103, USA). Fifteen cylindrical cavities fabricated from acrylic rod (5.0 mm diameter x 7.0 mm depth) were used as standardized preparations. Amalgam was triturated according to manufacturer s suggestion and condensed by a single operator into the cavity preparations (n=5) using a standardized technique. Samples were placed into 25 ml glass bottles and sealed within four minutes of trituration. Vapor was extracted and analyzed as previously described at 1, 3, 5 and 24 hours, then daily for 5 days. After each vapor measurement, samples were placed in new 25 ml glass bottles and sealed. Operator Skill Three operators an inexperienced, a novice and an expert participated in the study using both a spherical alloy (Valiant Snap, Ivoclar North America) and an admixed alloy (Valiant PhD). Amalgam was triturated according to manufacturer s directions, and condensed into standardized cavities (5.0 mm diameter x 5.0 mm depth) prepared in acrylic rod. The operators received no instructions for condensation technique. Each operator filled five cavities with each amalgam type. Samples were stored in 25 ml glass bottles and sealed within four minutes of trituration. Measurements were taken as previously described at 1, 3, 5 and 24 hours, then daily for 4 days after trituration. After each measurement of vapor, amalgam samples were placed into new 25 ml glass bottles and sealed. Operator Technique A spherical alloy (Tytin, Kerr Corporation) and an admixed alloy (Dispersalloy, Dentsply/Caulk, Milford, DE 19963, USA) were evaluated for mercury vapor release following condensation into standardized acrylic cavities (5.0 mm diameter x 4.0 mm depth). A single operator placed each alloy using two methods of condensation. For Method 1, the cavity was filled to excess during condensation to produce a mercury-rich layer that was then carved away to leave the amalgam flush with the surface of the cavity. For Method 2, only the amount of amalgam needed to fill the cavity even with the preparation surface was condensed with little or no carving required, leaving the mercury-rich layer at the surface. Five samples were prepared for each amalgam/technique combination. Samples were stored and analyzed as previously described at 1, 3, 5 and 24 hours, then daily for 9 days. After each vapor measurement was taken, amalgam samples were placed into new 25 ml glass bottles and sealed. Cavity Design Three standardized cavities were fabricated in the acrylic rod; standard Class I (5.0-mm diameter x 4.5-mm depth), double volume (5.0 mm diameter x 9.0 mm depth) and double surface area (standard preparation with both ends exposed to air). Ten samples of each preparation type were filled with a spherical alloy (Tytin). Amalgam was triturated according to manufacturer s instructions, condensed into each cavity and carved with a standardized technique by a single operator. Samples were placed into 25 ml glass bottles and sealed three minutes after trituration. Vapor was sampled and analyzed as previously described at 0.5, 1, 3, 5 and 24 hours after trituration, then daily until vapor levels were below the detection level of the instrument (+/-0.03 mg/m 3 ). After each vapor reading, the bottles were left open to room air for four minutes, then resealed. RESULTS For each system investigated, mercury vapor release decreased with time. Also, a significant reduction in measurable mercury vapor was reported within the first five hours following trituration for each system. Alloy Morphology Figure 2 presents the average mercury vapor released from the three alloy morphologies in nanograms (ng). A

78 76 Operative Dentistry Never Novice Expert Never Novice Expert Spherical Alloy Alloy Morphology Spherical Admixed Atomized Figure 2. Total mercury vapor release: Alloy Morphology. 5.8 (1.4) 7.0 (1.1) 10.3 (1.6) Operator Skill 10.9 (2.0) 9.0 (0.8) 12.0 (2.5) 17.4 (2.1) 11.6 (0.9) Admixed Alloy Figure 3. Total mercury vapor release: Operator Skill. significant difference was determined among alloys by time, with a material/time interaction. A significant difference in mercury release between initial reading and 24-hour reading was determined for each alloy. The alloy with the greatest initial reduction in vapor release (spherical alloy) resulted in the least amount of overall mercury vapor. There was no significant difference (p 0.05) among initial readings of the three alloys. However, Tukey pairwise comparison yielded a significant difference in the total amount of mercury vapor released between each alloy pair. The spherical alloy 15.4 (1.8) Day 6 Day 5 Day 4 Day 3 Day 2 Day 1 Day 4 Day 3 Day 2 Day 1 yielded the least total mercury vapor (7.0 ng) over six days of collection, followed by the admixed (9.0 ng) and the atomized irregular particle alloy (11.6 ng). Operator Skill The results (Figure 3) are presented in nanograms for mercury vapor released from the condensation of amalgam alloy by three clinicians with varying degrees of experience. A significant difference was observed among operators and materials (p 0.05). The spherical alloy consistently released significantly less mercury vapor than the admixed alloy for each operator and each time interval. The cavities restored by the operator without prior experience or training (inexperienced) released significantly less total mercury vapor than the novice (three months training) or expert (30 years experience) for both morphologies. Comparison of sample volumes among the three operators suggests that the differences in mercury vapor release were unrelated to sample volume. Final alloy mass was calculated for each sample following the experiment. The inexperienced operator produced samples with an average final mass (9.6 mg) statistically greater (p 0.05) than the novice operator (9.2 mg) for the spherical alloy Valiant. Average final mass for the expert operator fell between the inexperienced and novice operators. Cavities filled with the admixed alloy Valiant PhD yielded weights statistically similar for the inexperienced (9.3 mg), novice (9.5 mg) and expert (9.3 mg) operators. Operator Technique Figure 4 presents the average mercury vapor released from a spherical and admixed alloy using different condensation and carving techniques; either overfilled and carved flush or condensed flush without excess to carve back. Initial mercury vapor readings were significantly greater than subsequent readings over time with both techniques. No significant difference was determined (p 0.05) with either technique for the admixed alloy. For the spherical alloy, overfilling and carving back the mercury-rich layer resulted in statistically less mercury vapor (7.6 ng) compared with filling the cavity flush without excess to carve back (9.7 ng).

79 Neme & Others: Mercury Vapor Release from Amalgam 77 Cavity Design Figure 5 presents the average daily and total amount of mercury vapor released in nanograms over a 14 day period in three different cavity designs with a spherical alloy. ANOVA and pairwise comparisons demonstrated a statistically significant difference among all three restoration designs (p 0.05). The double volume cavity design released a significantly different amount of total mercury vapor (6.4 ng) than the standard design (4.4 ng). Mercury vapor from the double surface design (9.2 ng) was statistically different from the standard and double volume designs. As with operator skill, mean weight ratios [single volume (1): double volume (1.5): double surface (1.2)] do not account for differences in mercury vapor release. DISCUSSION The alloys evaluated in this investigation represent varying morphologies and chemistries. Alloys were chosen for each of the individual studies based on their unique characteristics. To test for 10 the effect of morphology on mercury 9 vapor release, alloys with similar chemistry, yet different alloy particle shape, 8 were evaluated (spherical, admixed 7 and an atomized irregular particle). For 6 the other individual studies, operator 5 skill, operator technique and cavity 4 design, a mixture of conventional alloys 3 was tested to evaluate possible trends in mercury vapor release that were not 2 brand-specific. 1 A consistent finding among all variables tested in this study was a typical 0 pattern of mercury vapor release from setting dental amalgam. This pattern is characterized by higher levels of mercury release early in the setting reaction, steadily decreasing over time, until the release rate is undetectable. This suggests that mercury is released throughout the initial amalgamation reaction and continues to be released until the material mass reaches a steady state (Okabe, 1987). However the results also demonstrate that mercury vapor release from amalgam is affected by many variables and challenges some basic concepts regarding amalgam technique. The finding that alloy morphology consistently demonstrated significant differences when evaluated (1.8) Overfilled Even Overfilled Even Spherical Alloy Operator Technique Admixed Alloy Figure 4. Total mercury vapor release: Operator Technique. 4.4 (2.0) 9.7 (2.0) Cavity Design 6.4 (1.4) Standard Double Volume Double Surface Figure 5. Total mercury vapor release: Cavity Design. 8.8 (1.9) 7.9 (1.6) 9.2 (1.2) Day 5-10 Day 4 Day 3 Day 2 Day 1 Day 8-14 Day 7 Day 4 Day 3 Day 2 Day 1 alone and with other variables suggests it is important in predicting mercury vapor release. There are several possibilities for this finding. First, the differences in mercury vapor release over time between the spherical and admixed alloy may, in part, be determined by the dynamics of the amalgamation reaction related to the particle morphology. The spherical alloy has an increased surface area for immediate reaction with the liquid mercury. The more irregular surfaces of the admixed particle may not be as accessible to the liquid mercury initially, and therefore, free mercury remains

80 78 Operative Dentistry in the mass for a longer period of time during the reaction. Second, differences in chemical composition between materials may affect mercury vapor release, especially with regard to tin content (Mahler, Adey & Fleming, 1994; Ferracane & others, 1995). However, the chemical composition of materials used in the individual studies was very similar (that is, Valiant/Valiant PhD) with the exception of operator technique (Tytin/Dispersalloy). Also, a lower mercury-to-alloy ratio is typical for spherical alloys to provide an acceptable consistency compared to admixed. It seems logical that if one starts with a greater percent of mercury, more will be left for release. The effect of operator experience and operator technique somewhat challenges traditional beliefs related to amalgam. One might expect the experienced operator to produce the restoration with the least measurable mercury vapor given the traditional training in dental school. By applying deliberate condensation forces to squeeze the liquid mercury out of the material to produce optimal physical properties, decreased overall mercury in the set mass might be expected. However, the opposite trend was found. The untrained operator produced a restoration that released the least total mercury vapor. One might also hypothesize that the untrained operator may not condense as carefully as the trained or expert operator, and thus produce an amalgam sample with more voids internally. If so, the final amalgam sample weight should be less, which would therefore explain the reduction in mercury vapor seen. However, the only significant difference in weight was between the inexperienced operator (9.6 mg) compared to the novice operator (9.2 mg) when handling the spherical alloy Valiant. This finding alone does not support the theory that the inexperienced operator would produce a less dense cavity resulting in lower mercury vapor release. The assumption that individuals in this portion of the study represented a typical inexperienced, novice and expert operator needs to be demonstrated by repeating the study with more of each type of operator. The technique of overfilling and carving back to remove the mercury-rich layer had the predicted result of reducing mercury vapor for the spherical alloy, but this result was not demonstrated for the admixed material. Although the experienced operator using careful condensation and carving technique may produce a restoration with improved physical properties due to proper adaptation to cavity walls and floors and reduction of internal voids, such operator may not necessarily bring about a reduction in mercury vapor release. This is another example of the significance of alloy morphology on the characteristics of dental amalgam. Research has shown that spherical alloys have demonstrated a higher degree of technique sensitivity regarding handling characteristics (Mahler & Nelson, 1984; Brown & Miller, 1993), sealing potential (Brown & Miller, 1993) and post-operative sensitivity (Mahler & Nelson, 1984; Mahler & Nelson, 1994). Comparing vapor release by preparation design suggests that the larger the mass of amalgam presented in the final restoration, the more mercury vapor will be released. However, the most important factor for predicting mercury vapor release from amalgam restorations does not appear to be volume, but the square area of the restoration exposed to air. In the case of mercury vapor exposure, it seems most reasonable to talk about the surfaces of amalgam, rather than the number of spills used. Mercury in its many forms is widely distributed throughout the environment and trace levels are present in air, water and food. Mercury vapor (the monatomic gas, Hg0, released by metallic, liquid mercury) is probably the most important form of mercury that determines human exposure from dental amalgam fillings. The World Health Organization (1991) established that the absorption of elemental mercury via the respiratory organs is approximately 80%. It is important to clarify whether an amount of elemental mercury large enough to create adverse biological effects can be released from amalgam restorations. Estimates of inhaled elemental mercury from ambient air (unadjusted for absorption) range from 40 to 120 nanograms per day (Clarkson & others, 1988; US EPA, 1984). The amount of mercury vapor released from amalgam restorations tested in this study ranged from less than 1.0 nanograms per day to 11.4 nanograms per day. The worse case scenario of 11.4 ng/day, with inhalation approximately 80%, maintains the contact to that below the normal exposure from air. The results of this investigation demonstrate some consistent patterns for mercury vapor release in vitro. Although clinical evaluation is necessary to evaluate the relation between the amount of mercury vapor released in vitro to that released in the oral cavity, the results of this study indicate that mercury vapor release can be affected by operator technique and restoration design. Furthermore, as stated earlier, there is no evidence to support that mercury exposure due to amalgam usage has any adverse health effects (WHO, 1995; ADA Council on Scientific Affairs, 1998). The traditional methods of amalgam manipulation used in this study have affects beyond mercury vapor release. Many reports in the literature show benefits in terms of reduced porosity and improved mechanical properties resulting from following traditional methods for manipulation of various alloys. The results of this investigation demonstrate one more area in which the clinician s behavior may modify the outcome of the restorative material. Although traditional methods for amalgam manipulation may not coincide with pre-

81 Neme & Others: Mercury Vapor Release from Amalgam 79 dictable reduction in mercury vapor release, the literature does not support change in these methods based upon release of mercury vapor alone. CONCLUSIONS The variables tested in this in vitro investigation demonstrated interesting findings. Regardless of alloy type or manipulation, mercury vapor release decreased with time. Spherical alloys consistently released less mercury vapor than admixed alloys except when the spherical alloy had its remaining mercury-rich layer present. Variables including alloy morphology, operator skill and technique, and cavity design influence total mercury vapor release in vitro. As with other physical and chemical properties of materials, overall operator-handling techniques can influence mercury vapor release from dental amalgam. (Received 3 April 2001) References ADA Council on Scientific Affairs (1998) Dental amalgam; update on safety concerns Journal of the American Dental Association 129(4) Ahmad R & Stannard JG (1990) Mercury release from amalgam: A study in vitro and in vivo Operative Dentistry 15(6) Berdouses E, Vaidyanathan TK, Dastane A, Weisel C, Houpt M & Shey Z (1995) Mercury release from dental amalgams: An in vitro study under controlled chewing and brushing in an artificial mouth Journal of Dental Research 74(5) Berglund A (1990) Estimation by a 24-hour study of the daily dose of intra-oral mercury vapor inhaled after release from dental amalgam Journal of Dental Research 69(10) Berglund A (1993) An in vitro and in vivo study of the release of mercury vapor from different types of amalgam alloys Journal of Dental Research 72(5) Berglund A, Pohl L, Olsson S & Bergman M (1988) Determination of the rate of release of intra-oral mercury vapor from amalgam Journal of Dental Research 67(9) Berry TG, Summitt JB, Chung AK & Osborne JW (1998) Amalgam at the new millennium Journal of the American Dental Association 129(11) Brown IH & Miller DR (1993) Alloy particle shape and sensitivity of high-copper amalgams to manipulative variables Journal of the American Dental Association 6(5) Clarkson TW, Hursh JB, Sager PR & Syversen TLM (1988) Mercury in Clarkson TW, Friberg L, Nordberg GF, Sager PR, eds Biological Monitoring of Toxic Metals New York Plenum Press Engle JH, Ferracane JL, Wichmann J & Okabe T (1992) Quantitation of total mercury vapor released during dental procedures Dental Materials 8(3) Ferracane JL, Hanawa T & Okabe T (1992) Effectiveness of oxide films in reducing mercury release from amalgams Journal of Dental Research 71(5) Ferracane JL, Adey JD, Nakajima H & Okabe T (1995) Mercury vaporization from amalgams with varied alloy compositions Journal of Dental Research 74(7) Haikel Y, Gasser P, Salek P & Voegel JC (1990) Exposure to mercury vapor during setting, removing, and polishing amalgam restorations Journal of Biomedical Material Research 24(11) Leinfelder K (1993) Current developments in dentin bonding systems: Major progress found in today s products Journal of the American Dental Association 124(5) Mackert JR (1987) Factors affecting estimation of dental amalgam mercury exposure from measurements of mercury vapor levels in intraoral air and expired air Journal of Dental Research 66(12) Mahler DB & Nelson LW (1984) Factors affecting the marginal leakage of amalgam Journal of the American Dental Association 108(1) Mahler DB, Adey JD & Fleming MA (1994) Hg emission from dental amalgam as related to the amount of tin in the Ag-Hg (ϒ 1 ) phase Journal of Dental Research 73(10) Mahler DB & Nelson LW (1994) Sensitivity answers sought in amalgam alloy microleakage study Journal of the American Dental Association 125(3) Neme AL, McLaren JD & O Brien WJ (1999) Investigation of two mercury vapor collection techniques Dental Materials 15(6) Neme AL, Wagner WC & O Brien WJ (1999) Effects of palladium addition on emission of mercury vapor from dental amalgam Dental Materials 15(6) Okabe T (1987) Mercury in the structure of dental amalgam Dental Materials 3(1) 1-8. Olsson S & Bergman M (1992) Daily dose calculations from measurements of intra-oral mercury vapor Journal of Dental Research 71(2) Osborne JW & Albino JE (1999) Psychological and medical effects of mercury intake from dental amalgam; a status report American Journal of Dentistry 12(3) Reinhardt JW, Boyer DB, Ga DD, Cox R, Frank CW & Svare CW (1979) Mercury vapor expired after restorative treatment: Preliminary study Journal of Dental Research 58(10) Reinhardt JW, Boyer DB, Svare CW, Frank CW, Cox RD & Gay DD (1983) Exhaled mercury following removal and insertion of amalgam restorations Journal of Prosthetic Dentistry 49(5) Svare CW, Peterson LC, Reinhardt JW, Boyer DB, Frank CW, Gay DD & Cox RD (1981) The effect of dental amalgams on mercury levels in expired air Journal of Dental Research 60(9) US Environmental Protection Agency (1984) Mercury health effects update: Health issue assessment Washington, DC

82 80 Operative Dentistry Office of PHS publication #EPS-600/ F. Vimy MJ & Lorscheider FL (1985a) Intra-oral air mercury released from dental amalgam Journal of Dental Research 64(8) Vimy MJ & Lorscheider FL (1985b) Serial measurements of intra-oral air mercury: Estimation of daily dose from dental amalgam Journal of Dental Research 64(8) World Health Organization/Federation Dentaire International consensus statement on dental amalgam (1995) Journal Dental Association Scientific Affairs World Health Organization (1991) Environmental Health Criteria 118 Inorganic mercury Geneva Switzerland World

83 Operative Dentistry, 2002, 27, Post-gel Shrinkage with Pulse Activation and Soft-start Polymerization AUJ Yap MS Soh KS Siow Clinical Relevance When compared to a continuous, one-level output method, pulse-activation and soft-start polymerization regimens did not significantly reduce post-gel shrinkage. SUMMARY This study investigated the influence of pulse activation and soft-start polymerization regimens on the post-gel shrinkage of a visible lightactivated composite resin (Z100). A light-cure unit (BISCO VIP) that allowed for independent command over time and intensity was used. The six light-curing modes that were examined include: Control (C) 400 mw/cm 2 [40 seconds]; Department of Restorative Dentistry, Faculty of Dentistry, National University of Singapore, 5 Lower Kent Ridge Road, Singapore , Republic of Singapore Adrian UJ Yap, BDS, MSc, PhD, FAMS, FADM, FRSH, associate professor, Department of Restorative Dentistry, Faculty of Dentistry, assistant director, Centre for Biomedical Materials Applications and Technology, Faculty of Engineering, National University of Singapore MS Soh, BSc (Hons), Department of Chemistry, Faculty of Science, National University of Singapore KS Siow, BSc, MSc, PhD, associate professor, Department of Chemistry, Faculty of Science, National University of Singapore Pulse Delay I (PDI) 100 mw/cm 2 [3 seconds], delay [3 minutes], 500 mw/cm 2 [30 seconds]; Pulse Delay II (PDII) 200 mw/cm 2 [20 seconds], delay [3 minutes], 500 mw/cm 2 [30 seconds]; Soft-start (SS) 200 mw/cm 2 [10 seconds], 600 mw/cm 2 [30 seconds]; Pulse Cure I (PCI) two 400 mw/cm 2 [10 seconds] and one 400 mw/cm 2 [20 seconds] pulses with 10 seconds interval between; and Pulse Cure II (PCII) two 400 mw/cm 2 [20 seconds] pulses with 20 seconds interval between. A strain-monitoring device measured the linear polymerization shrinkage associated with the various cure modes during and post light polymerization up to 60 minutes. Five specimens were made for each cure mode. Data was analyzed using one-way ANOVA and Scheffe s posthoc test at significance level Post-gel shrinkage associated with PDI was significantly lower than with PDII, SS and PCI immediately post light-polymerization. At one-minute post light polymerization, PDI had significantly lower shrinkage compared to PDII and SS. Significant differences in shrinkage were observed between PDI and SS only at 10, 30 and 60 minutes. At all time intervals, no significance in post-gel shrinkage was observed between the control and allpulse activation/soft-start polymerization regimens.

84 82 Operative Dentistry Scholte & Davidson, 1990). The most recent method INTRODUCTION Light-activated composite restorative materials have revolutionized clinical dentistry. They have several advantages including control of contour during restoration placement, improved color stability and increased polymerization compared to chemically activated composite resins (Burgess & others, 1999). A critical limitation of composite resins is polymerization shrinkage (Sakaguchi & others, 1991; Yap & others, 2000). The shrinkage of composites, which is about 1-5 volume% (Davidson & Feilzer, 1997), can be divided into two phases: the pre-gel and post-gel phases. During pre-gel polymerization, the composite flows and stresses within the structure are relieved (Davidson & de Gee, 1984). After gelation, flow ceases and cannot compensate for shrinkage stresses. Thus, post-gel polymerization results in significant stresses in the surrounding tooth structure and composite-tooth bond (Feilzer, de Gee & Davidson, 1987). Setting stresses of light-activated composites are higher than chemically activated composites (Feilzer, de Gee & Davidson, 1993). This has been attributed to the capacity of flow that occurs and the slower development of modulus (Braem & others, 1987) in chemically activated materials. Flow is thought to be METHODS AND MATERIALS the ability of molecules to slip into new orientations during the polymerization process. Flow tended not to occur in light-activated composites because of their more rapid polymerization, achievement of cross-linking and elastic limit. Stresses arising from post-gel polymerization shrinkage may contribute to postoperative pain, microleakage and recurrent caries (Eick & Welch, 1986). Composite-dentin gap formation was found to be greater with the use of light-activated materials than with chemically-activated ones (Itoh & Wakumoto, 1985). Fracture in the enamel-resin interface at the cavosurface margin (Fusayama, 1992) often accompanies the use of light-activated composites. The effects of post-gel polymerization shrinkage can be minimized in several ways, including applying liners/ bases, incremental placement of composites and allowing composites to contract freely to the adhesive surface Figure 1. Set up for determination of curing depth. (Davidson, 1986; Kemp- Table 1: The Different Light-curing Modes Examined for minimizing polymerization shrinkage of lightactivated composites is to allow flow during setting by means of controlled polymerization. This can be done by pre-polymerization at low light intensity followed by final cure at high intensity (soft-start polymerization), applying short pulses of light energy (pulse activation) or a combination of both. While some studies have shown that these polymerization modes may result in smaller marginal gap and increased marginal integrity (Goracci, Casa de Martinis & Mori, 1996; Mehl, Hickel & Kunzelmann, 1997; Kanca & Suh, 1999), others found no significant improvement in marginal adaptation (Friedl & others, 2000). The polymerization shrinkage associated with these new curing modes, however, has not been widely reported and the few studies that were conducted were mostly on soft-start polymerization (Koran & Kurschner, 1998; Sakaguchi & Berge, 1998; Dennison & others, 2000; Silikas, Eliades & Watts, 2000; Yap, Ng & Siow, 2001). This study investigated the influence of different pulse activation and soft-start polymerization regimens on the post-gel shrinkage of a visible light-activated composite resin. Shrinkage profiles during the light-polymerization process were also examined. Light-curing Mode Regimen Control 400 mw/cm 2 (CC) (40 seconds) A mini-filled composite resin (Z100, 3M Dental Products, St Paul, MN 55144, USA) of A2 shade and a Pulse Delay I 100 mw/cm 2 Delay 500 mw/cm 2 (PDI) (3 seconds) (3 minutes) (30 seconds) Pulse Delay II 200 mw/cm 2 Delay 500 mw/cm 2 (PDII) (20 seconds) (3 minutes) (30 seconds) Soft-start 200 mw/cm mw/cm 2 (SS) (10 seconds) (30 seconds) Pulse Cure I 400 mw/cm 2 Delay 400 mw/cm 2 Delay 400 mw/cm 2 (PCI) (10 seconds) (10 seconds) (10 seconds) (10 seconds) (20 seconds) Pulse Cure II 400 mw/cm 2 Delay 400 mw/cm 2 (PCII) (20 seconds) (20 seconds) (20 seconds)

85 Yap, Soh & Siow: Post-gel Shrinkage with Pulse Activation and Soft-start Polymerization 83 Table 2: Mean Linear Percent Shrinkage for the Different Light-curing Modes Time 0 Minutes 1 Minute 10 Minutes 30 Minutes 60 Minutes Control (0.03) (0.05) (0.05) (0.04) (0.05) PDI (0.04) (0.04) (0.04) (0.04) (0.04) PDII (0.04) (0.03) (0.03) (0.03) (0.04) SS (0.06) (0.06) (0.06) (0.06) (0.06) PCI (0.03) (0.04) (0.05) (0.05) (0.05) PCII (0.01) (0.01) (0.02) (0.02) (0.02) Standard deviation in parentheses. With the strain gauge in place, the composite resin was placed in the cavity of the Teflon frame. Care was taken to ensure the frame was completely filled and excess composite material was extruded using pressure applied through a second glass slide, then removed. The surface tack of the composite adequately ensured adhesion between the strain gauge and the composite materials. The leads from the strain gauge were connected to a strain-monitoring device commercial light-cure unit that allowed for independent command over time and intensity (Variable Intensity Polymerizer [VIP]; BISCO Inc, Schaumburg, IL 60193, USA) were selected for this study. VIP has an output wavelength range of nm. It is programmed with pre-set exposure times of 2-5, 10, 20, 30 seconds and a continuous mode up to 225 seconds, and intensity settings of 100, 200, 300, 400, 500 and 600 mw/cm 2. The diameter of the exit window of the light cure tip was 11 mm. Light intensities for each light-curing mode were checked with the in-built radiometer prior to use. Table 1 details the six light-curing modes investigated. The control mode (C) involves light irradiation at 400 mw/cm 2 for 40 seconds. Pulse Delay I (PDI), the regimen recommended by BISCO, uses an initial low energy dose (100 mw/cm 2 for three seconds) followed by a waiting time of three minutes and a final cure at highenergy dose (500 mw/cm 2 for 30 seconds). The threeminute waiting period is to be used for finishing and polishing the composite restoration. Pulse Delay II (PDII) is similar to PDI with the exception of a higher initial energy dose (200 mw/cm 2 for 20 seconds). The soft-start (SS) mode uses an initial low light intensity (200 mw/cm 2 for 10 seconds) that is immediately followed by final cure at high light intensity (600 mw/cm 2 for 30 seconds). Pulse cure I (PCI) involves using two 400 mw/cm 2 10-second pulses and one 400 mw/cm second pulse with 10 second intervals between. For pulse cure II (PCII), two 400 mw/cm 2 20-second pulses with 20 second intervals was employed. The test configuration for measuring polymerization shrinkage involved a stiff, white Teflon mold (inner length 10.0 mm, width 5.0 mm and height 2.0 mm) which circumscribed the composite sample (Figure 1). A glass slide served as the base of the set-up. A foil electrical resistance strain gauge (Foil Strain Gauge, RS Components Ltd, Singapore) was attached onto the flat glass surface. The gauge was 2 mm in length and had an electrical resistance 120 Ω and gauge factor Table 3: Results of Statistical Analysis Time Significance 0 minute PDI < PDII, SS, PCI 1 minute PDI < PDII, SS 10 minutes PDI < SS 30 minutes PDI < SS 60 minutes PDI < SS Results of one-way ANOVA/Scheffe s post-hoc test at significance level 0.05.>denotes statistical significance Time (seconds) Figure 2. Mean linear percent shrinkage of the specimens in the control group during light polymerization Time (seconds) Figure 3. Mean linear percent shrinkage during light polymerization with Pulse Delay I.

86 84 Operative Dentistry Time (seconds) Figure 4. Mean linear percent shrinkage during light polymerization with Pulse Delay II Time (seconds) Figure 5. Mean linear percent shrinkage during light polymerization with soft-start polymerization Time (seconds) Time (seconds) Figure 6. Mean linear percent shrinkage during light polymerization with Pulse Cure I Time (minutes) Figure 8. Mean linear percent shrinkage post-light polymerization with the various light-curing modes. (Strain Gauge Recorder, Cole Parmer Instruments, IL 60061, USA) initially balanced at zero. The composite specimens were light polymerized using the different light-curing modes. Dimensional change during and post-light polymerization was monitored at room temperature (25 ± 1 C). During the light-polymerization process, shrinkage measurements were taken continuously every 10 seconds with the Control PDI PDII SS PCI PCII Figure 7. Mean linear percent shrinkage during light polymerization with Pulse Cure II. exception of PDI, where the first measurement was taken after three seconds. For PDI and PDII, readings were taken at 60-second intervals during the threeminute delay period. Post-light polymerization measurements were taken at 0 (immediately after removal of light source), 1, 10, 30 and 60 minutes. The linear percent shrinkage was derived from the following equation: Linear percent shrinkage ( L/L x 100) = ( R/R)/K x 100), where L is the change in length, L is the original length, R is the change of resistance, R is the original resistance and K is the gauge factor (that is, 2). Five composite specimens were used for each curing parameter. Data at 0, 1, 10, 30 and 60 minutes postlight polymerization was subjected to one-way ANOVA and Scheffe s post-hoc test at 0.05 significance level. RESULTS Figures 2-7 show the mean linear percent shrinkage during the light-polymerization process. Table 2 and Figure 8 reflect the mean linear percent shrinkage post-light polymerization. Table 3 shows the results of statistical analysis. For the control group (Figure 2), a steep increase in shrinkage was observed during the first 20 seconds of

87 Yap, Soh & Siow: Post-gel Shrinkage with Pulse Activation and Soft-start Polymerization 85 light polymerization. The rate of shrinkage decreased during the remaining 20 seconds of light curing. For PDI (Figure 3), no shrinkage was observed during the initial low energy cure and three-minute waiting time. Subsequent light polymerization at higher energy resulted in a steep increase in shrinkage. The shrinkage profile of PDII (Figure 4) during light polymerization differed markedly from PDI. Shrinkage associated with the initial low-energy cure resulted in progressive shrinkage even during the waiting period. The rate of shrinkage with the high-energy cure was lower than with PDI. With SS (Figure 5), an initial low rate of shrinkage was followed by a high rate of shrinkage. For both PCI and PCII (Figures 6-7), shrinkage profiles were similar. The composite continued to shrink during the 10- and 20-second time intervals between pulses. Linear shrinkage immediately after light polymerization ranged from 0.18% to 0.30%. The composite continued to shrink regardless of curing modes, and linear shrinkage ranged from %, 60 minutes after removing the light source. Immediately after light polymerization (0 minutes), using PDI resulted in significantly lower shrinkage than with PDII, SS and PCI. At one minute post-light polymerization, the shrinkage associated with PDI was significantly less than with PDII and SS. Results at 10, 30 and 60 minutes were identical. Shrinkage observed with PDI was significantly lower than with SS. At all time intervals, no significant difference in shrinkage was observed between the control and any of the light-curing modes. DISCUSSION The experimental set-up for measuring post-gel polymerization shrinkage was based upon that conducted by Yap & others (2000). Versluis, Sakaguchi & Douglas (1993) measured the post-gel shrinkage of 10 composite resins by means of strain gauges and found that Z100 exhibited the greatest contraction stress of all materials tested. Z100 was therefore selected for this study. Factors influencing the transmission of light include the thickness of the restorative material, the presence and size of filler particles and the distance of the light tip to the restoration surface (Tate, Porter & Dosch, 1999). As all of these factors were standardized in this study, any reduction in polymerization shrinkage may be attributed to the light-curing regimen. Two-mm thick composite specimens were used to ensure uniform, maximum polymerization (Yap, 2000). A minimum light intensity of 400 mw/cm 2 has been suggested for routine polymerization (Rueggeberg, Caughman & Curtis, 1994; Manga, Charlton & Wakefield, 1995). This light intensity, together with the manufacturer s recommended cure time of 40 seconds, was used as control. Volumetric curing contraction determinations (Attin & others, 1995; Rueggeberg & Tamareselvy, 1995) are basically free shrinkage measurements and can offer the total (pre- and post-gel) curing contractions. Dimensional changes in linear curing contraction determinations employed in this and other studies (de Gee, Feilzer & Davidson, 1993; Uno & Shimokobe, 1994) are, however, more or less hindered and should be regarded as a post-gel curing phenomenon. As the displacement transducer (strain gauge) employed requires activation by force, it can only monitor the post-gel part of the curing shrinkage when the composite is sufficiently strong to exert forces (Davidson & Feilzer, 1997). The rate of shrinkage during light polymerization in the control group was not uniform. Shrinkage was very rapid during the initial 20 seconds and decreased during the latter 20 seconds of the 40-second cycle. Softstart polymerization, pulse activation or a combination of these curing modes could reduce the high initial shrinkage by allowing composite flow to occur during this period. Pulse-delay modes are a combination of pulse activation and soft-start polymerization, and PDI is the pulse-delay regimen recommended by BISCO. No linear shrinkage was detected during the initial low energy cure and the three-minute waiting period for PDI. Two possible hypotheses may account for this. Either the light energy density (intensity x time) was insufficient to initiate polymerization or only pre-gel polymerization took place. As stated earlier, the latter will not be detected by strain gauges. Using an initial low-light energy density must, however, be compensated for by a final cure at high light energy density (Yap & Seneviratne, 2001) to ensure optimal physico-mechanical properties and to prevent clinical problems caused by the cytotoxicity of inadequately polymerized materials (Caughman & others, 1991). Although post-gel shrinkage immediately after light polymerization was lower than the control, the difference was statistically insignificant. Shrinkage of PDI was, however, significantly lower than PDII, SS and PCI immediately after light polymerization. The low initial energy density used with PDI could result in a soft surface that may be prone to damage and excessive removal during finishing/polishing procedures during the three-minute waiting period. PDII was therefore introduced. The higher initial light energy density used (200 mw/cm 2 for 20 seconds) ensures adequate surface hardness for finishing/polishing procedures (Yap & Seneviratne, 2001) but resulted in composite post-cure after removal of the light source. The latter accounts for the continued shrinkage during the three-minute waiting period with PDII. The significantly higher shrinkage immediately after light polymerization with PDII compared to PDI may be attributed to the greater total light exposure time employed. The soft-start regimen used was based upon a commercial curing unit with a soft or ramp cure (Epilar

88 86 Operative Dentistry Highlight, Espé Dental AG, Seefeld, Germany). It used an initial 10-second cure with an intensity of 200 mw/cm 2 and a final cure at 750 mw/cm 2 for 30 seconds to complete a 40-second curing cycle. With VIP only having a maximum intensity of 600 mw/cm 2, an initial 10-second cure of 200 mw/cm 2 and a final 30-second cure at 600 mw/cm 2 were selected to characterize softstart polymerization. The shrinkage profile correlated well to the light energy density employed. Shrinkage immediately after light polymerization was significantly higher than PDI and could be attributed to the higher light intensity used during the final cure. A 30-second 500 mw/cm 2 or 20-second 600 mw/cm 2 cure may actually be sufficient to ensure effective cure (Yap & Seneviratne, 2001) and will possibly reduce the shrinkage observed. As the total light-exposure time and light intensity are the same for both pulse cure techniques and the control, any decrease in shrinkage present may be ascribed to second delays between pulses where flow and stress relaxation can occur. No significant difference in shrinkage was observed between the two pulse cure techniques and the control immediately after light polymerization. Shrinkage of PCI immediately after light polymerization was significantly higher than PDI. The short 10-second pulses and delay could actually increase contraction stresses as the material is subjected to shrinkage stresses just when flow is occurring. A longer waiting period, as with PDI and PCII, may therefore be more beneficial as composite flow is allowed to compensate for shrinkage stresses arising from the initial cure prior to the final cure. Regardless of the curing mode, the composite continued to shrink after removing the light source. This can be attributed to the post-curing of composite resins. Polymerization is approximately 75% complete at 10 minutes after light exposure and curing continues for a period of at least 24 hours (O Brien, 1997). Thermal contraction due to loss of radiant heat may also contribute to the initial post-light polymerization shrinkage. At all time intervals, no significant difference in post-gel shrinkage was observed between the control group and all the other curing modes (Table 3). Findings agree with other shrinkage studies conducted on soft-start polymerization (Koran & Kurschner, 1998; Silikas & others, 2000; Yap & others, 2001). Dennison & others (2000) and Sakaguchi & Berge (1998), however, they found that soft-start polymerization significantly reduced polymerization shrinkage. The apparent discrepancy may be explained by the fact that using initial lower light intensities was not compensated for by a final cure at higher intensities in the latter studies. Both studies used a 20-second final cure at 100% of the original intensity. Commercially available soft-start polymerization light-cure units usually use a final cure of 500 mw/cm 2 or greater. The beneficial effect of the initial low intensity cure may therefore be annulled by the high intensity final cure. Independent literature measuring the shrinkage of composites with the pulse-delay and pulse activation techniques is not currently available. Kanca & Suh (1999) evaluated the effect of pulse-delay curing of composites on stress reduction at the enamel surface. The rate of polymerization was examined by hardness testing, while the effects of polymerization stresses were examined by using dye penetration. They found that the rate of surface hardness development with pulsedelay curing was slower than the control. This may reflect a slower rate of polymerization and thus the development of modulus. The accompanying low incidence of dye penetration suggests that with pulse-delay curing, flow may occur and provide stress relief within the composite. The use of a low initial light energy density and relatively long delay period allowed for prolongation of the pre-gel state. The aforementioned observations were consistent with this study. No post-gel shrinkage was observed during the initial cure and three-minute delay for PDI. Among the different pulse activation and soft-start polymerization regimens, only PDI had lower shrinkage than the control. The shrinkage associated with PDI was significantly less than SS at all post-light polymerization time intervals. The conclusions drawn from this study will probably also apply to a wider range of composites with comparable formulations. Photosensitizer concentrations in the different composites may, however, have a profound effect on the responsiveness to particular light intensities. More research involving the use of other materials, multiple permutations and combinations of light energy density and delay periods are warranted. CONCLUSIONS Under the conditions of this in vitro study: 1. Post-gel shrinkage immediately after light-polymerization ranged from 0.18% to 0.30% for PDI and PDII, respectively. 2. Composite post-curing was observed for the control and all pulse activation/soft-start polymerization regimens. 3. Post-gel shrinkage at 60 minutes post light-polymerization ranged from 0.32 to 0.44% for PDI and SS, respectively. 4. At all time intervals, no significant difference in postgel polymerization shrinkage was observed between the control and different pulse activation/soft-start polymerization regimens. 5. At all time intervals, post-gel shrinkage with PDI was significantly lower than SS.

89 Yap, Soh & Siow: Post-gel Shrinkage with Pulse Activation and Soft-start Polymerization 87 (Received 9 April 2001) References Attin T, Buchalla W, Kielbassa AM & Helwig E (1995) Curing shrinkage and volumetric changes of resin-modified glass ionomer restorative materials Dental Materials 11(6) Braem M, Lambrechts P, Vanherele G & Davidson CL (1987) Stiffness increase during the setting of dental composite resins Journal of Dental Research 66(12) Burgess JO, DeGoes M, Walker R & Ripps AH (1999) An evaluation of four light-curing units comparing soft and hard curing Practical Periodontics and Aesthetics Dentistry 11(1) Caughman WF, Caughman GB, Shiflett RA, Rueggeberg F & Schuster GS (1991) Correlation of cytotoxicity, filler loading and curing time of dental composites Biomaterials 12(8) Davidson CL (1986) Resisting the curing contraction with adhesive composites Journal of Prosthetic Dentistry 55(4) Davidson CL & de Gee AJ (1984) Relaxation of polymerization contraction stresses by flow in dental composites Journal of Dental Research 63(2) Davidson CL & Feilzer AJ (1997) Polymerization shrinkage and polymerization shrinkage stress in polymer-based restoratives Journal of Dentistry 25(6) de Gee AJ, Feilzer AJ & Davidson CL (1993) True linear polymerization shrinkage of unfilled resins and composites determined with a linometer Dental Materials 9(1) Dennison JB, Yaman P, Seir R & Hamilton JC (2000) Effect of variable light intensity on composite shrinkage Journal of Prosthetic Dentistry 84(5) Eick JD & Welch FH (1986) Polymerization shrinkage of posterior composite resins and its possible influence on postoperative sensitivity Quintessence International 17(2) Feilzer AJ, de Gee AJ & Davidson CL (1987) Setting stress in composite resin in relation to configuration of the restoration Journal of Dental Research 66(11) Feilzer AJ, de Gee AJ & Davidson CL (1993) Setting stresses in composites for two different cuing modes Dental Materials 9(1) 2-5. Friedl KH, Schmaltz G, Hiller KA & Markl A (2000) Marginal adaptation of Class V restorations with and without softstart polymerization Operative Dentistry 25(1) Fusayama T (1992) Indications for self-cured and light-cured adhesive composite resins Journal of Prosthetic Dentistry 67(1) Goracci G, Casa de Martinis L & Mori G (1996) Curing light intensity and marginal leakage of composite resin restorations Quintessence International Itoh K & Wakumoto S (1985) Wall-to-wall polymerization contraction of five composites Journal of the Showa Dental Society 5(2) Kanca J & Suh BI (1999) Pulse activation: Reducing resin-based composite contraction at the enamel cavosurface margins American Journal of Dentistry 12(3) Kemp-Scholte CM & Davidson CL (1990) Complete marginal seal of Class V resin composite restorations effected by increased flexibility Journal of Dental Research 69(6) Koran P & Kurschner R (1998) Effect of sequential versus continuous irradiation of a light-cured resin composite on shrinkage, viscosity, adhesion, and degree of polymerization American Journal of Dentistry 11(6) Manga RK, Charlton DG & Wakefield CW (1995) In vitro evaluation of a curing radiometer as a predictor of polymerization depth General Dentistry 43(3) Mehl A, Hickel R & Kunzelmann KH (1997) Physical properties and gap formation of light-cured composites with and without soft-start polymerization Journal of Dentistry 25(3-4) O Brien WJ ed (1997) Dental Materials and Their Selection 2 nd edition Illinois Quintessence Publishing Co pp Rueggeberg FA, Caughman WF & Curtis JW (1994) Effect of light intensity and exposure duration on cure of resin composite Operative Dentistry 19(1) Rueggeberg FA & Tamareselvy K (1995) Resin cure determination by polymerization shrinkage Dental Materials 11(4) Sakaguchi RL & Berge HX (1998) Reduced light energy density decreases post-gel contraction while maintaining degree of conversion in composites Journal of Dentistry 26(8) Sakaguchi RL, Sasik CT, Bunczak MA & Douglas WH (1991) Strain gauge method for measuring polymerization shrinkage of composite resin Journal of Dentistry 19(5) Silikas N, Eliades G & Watts DC (2000) Light intensity effects on resin-composite degree of conversion and shrinkage strain Dental Materials 16(4) Tate WH, Porter KH & Dosch RO (1999) Successful photocuring: Don t restore without it Operative Dentistry 24(2) Uno S & Shimokobe H (1994) Contraction stress and marginal adaptation of composite restorations in dentinal cavity Dental Materials Journal 13(1) Versluis A, Sakaguchi R & Douglas W (1993) Post-gel shrinkage measurements by measurements by means of strain gauges Journal of Dental Research (Abstract #2263). Yap AUJ (2000) Effectiveness of polymerization in composite restoratives claiming bulk placement: Impact of cavity depth and exposure time Operative Dentistry 25(2) Yap AUJ, Ng SC & Siow KS (2001) Soft-start polymerization: Influence on effectiveness of cure and post-gel shrinkage Operative Dentistry 26(3) Yap AUJ & Seneviratne C (2001) Influence of light energy density on effectiveness of composite cure Operative Dentistry 26(5) Yap AUJ, Wang HB, Siow KS & Gan LM (2000) Polymerization shrinkage of visible-light-cured composites Operative Dentistry 25(2)

90 Operative Dentistry, 2002, 27, Adhesion of Single Application Bonding Systems to Bovine Enamel and Dentin M Miyazaki K Iwasaki H Onose Clinical Relevance Bond strengths of single application bonding systems are comparable to those of a compomer restorative system. SUMMARY Single application bonding systems have recently been developed in an effort to simplify and shorten bonding procedures. This study compared their bonding ability to enamel and dentin. Four commercial single application systems, Reactmer Bond (Shofu Inc), One-Up Bond F (Tokuyama Co), AQ Bond (Sun Medical Co, Japan) and Prompt L- Pop (ESPE, Germany) were used. F2000 compomer (3M Dental Products, St Paul, MN 55144, USA) was used as a control material. Bovine mandibular incisors were mounted in self-curing resin and the facial surfaces were ground to expose either enamel or dentin. Restoratives were bonded after adhesive application to tooth surface according to the manufacturer s instructions. Fifteen samples per test group were stored in 37 C water for 24 hours, then shear tested at a Department of Operative Dentistry, Nihon University School of Dentistry, , Kanda- Surugadai, Chiyoda-Ku, Tokyo , Japan Masashi Miyazaki, DDS, PhD, instructor Keisuke Iwasaki, DDS, PhD, instructor Hideo Onose, DDS, PhD, professor and chair crosshead speed of 1.0 mm/minute. Statistical analysis was accomplished with a one-way ANOVA followed by the Duncan test (p<0.05) were done. The enamel bond strengths of the newly developed one step bonding systems were not significantly different from the compomer except for Prompt L-Pop, which showed the highest value. The dentin bond strengths of single application bonding systems did not differ from the compomer. The results of this study suggest that the adhesive properties of the newly developed single application bonding systems were comparable to a compomer restorative. INTRODUCTION For many years the dental profession has strived to achieve good adhesion of resin composite to tooth substrate since reliable bonding should produce less microleakage and restoration stability (Van Meerbeek & others, 1998; Perdigão & Lopes, 1999). The introduction of the acid-etch technique (Buonocore, 1955) produced an optimal micromechanical retention between the resin composite and the etched enamel (Gwinnett & Matsui, 1967; Buonocore, Matsui & Gwinnett, 1968). However, early attempts to create a chemical bond with dentin relying on the smear layer provided low bond strength values (Tyas & Chandler, 1993a; Van Meerbeek & others, 1993a; Wilson & Wilson, 1995).

91 Miyazaki, Iwasaki & Onose: Bonding Performance of Single Application Bonding Systems 89 Since the advent of dentin conditioning with acid and the introduction of bi-functional dentin primers (Munksgaard & Asmussen, 1984), three-step bonding systems have increased the acceptance and reliability of resin bonding to dentin (Tyas, 1992; Triolo & Swift, 1992; Jordan, Suzuki & Davidson, 1993; Senda & others, 1995). These systems require dentin conditioning and priming steps prior to applying the bonding agent to ensure maximum adhesive strength (Erickson, 1992; Van Meerbeek & others, 1998). Because bonding procedures require multiple-step clinical approaches, clinical success with these adhesive systems may depend on technique-sensitive and material-related factors (Sano & others, 1998; Finger & Balkenhol, 1999; Miyazaki, Onose & Moore, 2000). Recently, single application bonding systems that combine the function of self-etching primer and bonding agent have been developed (Kitasako & others, 2000). Theoretically, the acidic one-application adhesive dissolves the smear layer, incorporating it into the mixture and demineralizing the superficial dentin. It then hardens after light irradiation. Since these are recently developed bonding systems, more details about bonding to tooth structure needs to be evaluated. This study evaluated the bond strengths to enamel and dentin of new, single application bonding systems and compared them to the shear bond strength of a compomer restorative system. The hypothesis tested was that single application bonding systems would show the same bonding ability as commercially available compomer restorative systems used in clinics. METHODS AND MATERIALS Four commercial single application systems with the combination of resin composites shown in Table 1, Reactmer Bond/ Reactmer (RB, Shofu Inc, Kyoto, Japan), One-Up Bond F/Palfique Estelite (OF, Tokuyama Co, Tokyo, Japan), AQ Bond/Metafil (AQ, Sun Medical Co, Shiga, Japan), and Prompt L-Pop/Filtek Z250 (PL, ESPE, Germany) were used. Clicker/F2000 compomer (F2, 3M Dental Products) was used as a control material. All adhesive systems were applied according to the manufacturers instructions. An Optilux 500 (Demetron/ Kerr, Danbury, CT 06810, USA) curing unit was connected to a variable voltage transformer in order to adjust the light intensity to 600mW/cm 2. Mandibular incisors extracted from two-to-three year old cattle and stored frozen (-20 C) were used as a substitute for human teeth. After removing the roots with an Isomet low-speed saw (Buehler Ltd, Lake Bluff, IL 60044, USA), the pulps were removed and the pulp chamber of each tooth was filled with cotton to avoid penetration of the embedding media. The labial surfaces of the bovine incisors were ground on wet 240-grit SiC paper to make a flat surface in enamel or dentin. Each tooth was then mounted in cold-curing acrylic resin to expose the flattened area and placed in tap water to minimize the temperature rise from the exothermic polymerization reaction of the acrylic resin. Final finish was accomplished by grinding on wet 600- grit SiC paper to expose an area of enamel or dentin approximately 4.0 mm in diameter sufficient for bond strength testing. After ultrasonic cleaning in distilled water for one minute to remove any debris, these surfaces were washed and dried with a three-way syringe. The teeth, mounted with enamel or dentin exposed, were randomly assigned to five restorative material groups with a sample size of 15 per group for enamel and dentin, respectively. Table 1. Bonding Systems Used in This Study Code System Adhesive Lot # Restorative Lot # (Manufacturer) (Main components) RB Reactmer Reactmer Bond A: Reactmer (Shofu Inc) (4-AET, 4-AETA, UDMA, B: HEMA, PRG filler, F aluminosilicate glass, acetone, water, initiator) OF One-Up Bond F One-Up Bond F A:001 Palfique Estelite 219 (Tokuyama Corp) (MAC-10, HEMA, MMA, B: 501 F aluminosilicate glass, water, initiator) AQ AQ Bond AQ Bond VL-3 Metafil TF3 (Sun Medical Co) (4-META, UDMA, Initiator Water, Acetone) PL Prompt L-Pop Prompt L-Pop Z (ESPE) (di-hema-phosphate, (3M Dental Products) Complex fluoride, Water) F2 F2000 Compomer Primer/Adhesive in Clicker ALAR F2000 OBN (3M Dental Products) (Vitremer copolymer, HEMA Bis-GMA, maleic acid, water, initiator)

92 90 Operative Dentistry Tooth surfaces were treated and adhesives applied according to each manufacturer s recommendations. Adhesive tape was used to define the area of the tooth for bonding and a Teflon mold, 2.0 mm high and 4.0 mm in diameter, was used to form and hold the materials to the tooth surface. The restorative was condensed into the mold and light activated for 30 seconds. The Teflon mold and adhesive tape were removed from the specimens 10 minutes after light irradiation. The specimens were then stored in 37 C water for 24 hours and tested in shear mode using knife-edge shear testing apparatus in an Instron testing machine (Instron Corp, Canton, MA, USA) at a crosshead speed of 1.0 mm/minute. Shear bond strength values in MPa were calculated from the peak load at failure divided by the specimen surface area. After testing, the specimens were examined in an optical microscope at a magnification of x10 to determine location of the bond failure. The test area on the tooth was divided into eight segments, and the percentage of remaining substrate was estimated. The types of failures were determined based on the predominant percentage of substrate-free material as adhesive failure, cohesive failure in resin composite and cohesive failure in enamel or dentin. The results were analyzed by calculating the mean shear bond strength (MPa) and standard deviation for each group. The data for each group were tested for homogeneity of variance using Bartlett s test, then subjected to an ANOVA followed by the Duncan multiple range test at p<0.05. The statistical analysis was carried out with the Sigma Stat software system (SPSS Inc, Chicago, IL, USA). The treated tooth surface and restorative/tooth interface were observed by scanning electron microscopy (SEM). For the treated tooth surface observation, the enamel and dentin surfaces were treated with adhesive/primer/etchant according to each manufacturer s instructions, then rinsed with acetone and water. For the ultrastructure observation of the restorative-tooth interface, bonded specimens stored in 37 C distilled water for 24 hours were embedded in epoxy resin, then longitudinally sectioned with a diamond saw. The sectioned surfaces were polished with abrasive discs and diamond pastes down to a 0.1 µm particle size. All SEM specimens were dehydrated in ascending concentrations of tert-butanol (50% for 20 minutes, 75% for 20 minutes, 95% for 20 minutes and 100% for two hours). They were then transferred to a critical-point dryer for 30 minutes. The polished surfaces were subjected to argon-ion beam etching for 30 seconds with the ion beam (Elionox Ltd, Tokyo, Japan) with accelerating voltage of 1.0 kv and ion current density of 0.4 ma/cm 2 directed perpendicular to the polished surface (Inokoshi & others, 1993). The surfaces were coated with a thin film of Au in a vacuum evaporator. The specimens were observed in a scanning electron microscope (JSM-5400, JEOL Ltd, Tokyo, Japan). RESULTS Table 2 shows the mean shear bond strengths to bovine enamel and fracture modes after the test. The enamel Table 2. Mean Shear Bond Strength to Bovine Enamel and Failure Mode After the Test Code n Mean SD CV Duncan s Failure Mode (MPa) (%) Group (A/CR/CE)* RB a 15/0/0 OF a 15/0/0 AQ a 15/0/0 PL b 15/0/0 F a 15/0/0 SD: standard deviation CV: coefficient of variance Values with the same letter are not significantly different at p<0.05. *: A: adhesive failure, CR: cohesive failure in resin, CE: cohesive failure in enamel Table 3. Mean Shear Bond Strength (MPa) to Bovine Dentin and Failure Mode After the Test Code n Mean SD CV Duncan s Failure Mode (MPa) (%) Group (A/CR/CD)* RB a 10/5/0 OF a 8/7/0 AQ a 12/3/0 PL a 13/2/0 F a 15/0/0 SD: standard deviation C.V: coefficient of variance Values with the same letter are not significantly different at p<0.05. *: A: adhesive failure, CR: cohesive failure in resin, CD: cohesive failure in dentin

93 Miyazaki, Iwasaki & Onose: Bonding Performance of Single Application Bonding Systems 91 bond strength ranged from 12.1 ± 1.3 to 21.7 ± 4.7 MPa, and PL showed the highest bond strength. For the enamel bond strength tests, the mode of failure used for all the systems was adhesive failure. Table 3 shows the mean shear bond strengths to bovine dentin and fracture modes. The dentin bond strength ranged from 10.7 ± 3.3 to 13.9 ± 1.8 MPa and OF and RB showed the highest bond strength. In the dentin bond strength tests, the predominant failure was cohesive failure in adhesive for RB and OF, and adhesive failure for AQ, PL and F2. The SEM observations of the treated enamel and dentin surfaces are shown in Figures 1 and 2, respectively. Based on post-application morphology, PL appears to have produced the most aggressive etching pattern. In the case of treated dentin, the smear layer was generally removed but smear plugs remained in some tubule orifices for F2. Figure 3 shows the representative SEM micrographs of the dentin-resin interface after argon-ion beam etching. A thin layer with low resistance to argon ion etching RB Figure 1. Shear bond strength to bovine enamel of the different restorative materials. RB OF OF AQ AQ Figure 2. Shear bond strength to bovine dentin of the different restorative materials. PL PL F2 F2 was identified as the hybrid layer for RB, OF and F2, and the width of this layer was about 0.5~1.0 µm. Though the thickness of the hybrid layer was different among the adhesive systems, excellent adaptation between dentin and restoratives was observed for all systems employed. DISCUSSION In this study, bovine teeth were used as reported in previous studies (Nakamichi, Iwaku & Fusayama, 1983; Fowler & others, 1992) because it is difficult to obtain large numbers of intact, extracted human teeth for conducting bond strength tests. It has been reported that adhesion to the superficial layer of dentin showed no significant differences between human and bovine dentin, and the dentin bond strength decreased with the depth of dentin due to the lower density of dentinal tubules (Schilke & others, 1999). Because the differences in tubule diameters and the number of lateral branches may have some effect on dentin bond strength (Ferrari & Davidson, 1996), bovine superficial dentin was used as a substitute for human dentin. The bond strength values measured depend on the bonding systems used, the site of the tooth, the type of the tooth structure (Suzuki & Finger, 1988, Ishioka & Caputo, 1989; Eick & others, 1991) and the different clinical situations (Sano & others, 1998; Finger & Balkenhol, 1999; Miyazaki & others, 2000). Studies conducted under standard laboratory conditions might lead to inappropriate conclusions with respect to clinical situations (Söderholm, 1991). The adhesive in the one-application bonding systems is a hydrophilic solution that is extremely effective in wetting the tooth surface (Kitasako & others, 2000). The etching effect of these systems is related to the acidic monomers or organic acid solutions that may interact with the mineral component of tooth substrate and enhance monomer penetration (Ikemura, Kouro & Endo, 1996). These one-application adhesives may form a continuum between the tooth surface and adhesive by simultaneous demineralization and resin penetration followed by polymerization. Penetration of acidic monomers into tooth surface creates resin tags for enamel (Gwinnett & Matsui, 1967; Buonocore & others, 1968) and a hybrid layer for dentin (Nakabayashi, Kojima & Masuhara, 1982; Van Meerbeek & others, 1992; Inokoshi & others, 1993) with the application of adhesives. The SEM observations of tooth surfaces treated with one-application adhesives showed some etching of enamel prisms after rinsing off the uncured adhesive with an acetone solution. The adhesive etches and penetrates the etched enamel and hardens after evaporation of the solvent and exposure to light. This process creates a mechanical retention between the enamel and adhesive.

94 92 Operative Dentistry Figure 3. Treated enamel surface with the adhesives followed by acetone and water rinsing. Different etching patterns were observed with different adhesive systems. Figure 4. Treated dentin surface with the adhesives followed by acetone and water rinsing. Removal of smear layer and the opening of dentinal tubules were observed. With the use of acidic solutions, the surface enamel is permanently lost from the tooth surface during the etching procedure. Chemical treatment of enamel by acidic solution provides a modified surface that offers an attractive approach for getting stable bonding between resin composite and enamel substrate (Buonocore, 1955; Gwinnett & Matsui, 1967; Buonocore & others, 1968). From the time that phosphoric acid was introduced as an etching solution for enamel bonding, other etching agents have been investigated for use in clinical dentistry (Bates & others, 1982; Barkmeier, Shaffer & Gwinnett, 1986; Shaffer, Figure 5. Dentin-resin interface after argon-ion etching. Perfect adaptation to the dentin with a thin hybrid layer (resin impregnated dentin) was observed. Barkmeier & Kelsey, 1987; Blosser, 1990). One recently developed acidic solution is a self-etching primer that contains acidic functional monomers, such as 4-AET (Ikemura & others, 1996), phenyl-p and MDP (Barkmeier, Los & Triolo, 1995; Gordan & others, 1997). The self-etching primer, which is believed to be a precursor of the single application system, enables etching and priming of the tooth surface simultaneously. The depth of the etching pattern and the amount of surface enamel removed during etching depends on the type of acid, acid concentration, duration of acid application and composition of the surface enamel. For single application bonding systems, applying the adhesive allows mineralized tissue to be etched and roughened. From the morphological observations by SEM in this study, applying the adhesive did not create a deep enamel pattern like applying phosphoric acid did. However, the specific etching pattern may not be a critical factor in determining enamel bond strength. It was reported that enamel bond strengths after 37% phosphoric acid etching for 15 or 60 seconds were similar to those obtained with a 5% solution for 15 seconds, although the etching patterns observed with SEM were quite different (Barkmeier, Shatter & Gwinnett, 1986). Creating the etching pattern required for stable enamel bond seems to differ among the bonding systems used and other factors, such as age, site and amount of mineral removal from the tooth. This may contribute to variations in enamel bond. The dentin bonding mechanism of resin composite is believed to consist of three steps dentin conditioning, priming and adhesive application (Van Meerbeek & others, 1998). It is generally accepted that the smear layer that forms on ground dentin should be removed or altered with acidic conditioners to achieve good adhesion between the demineralized dentin substrate and an applied bonding system (Prati & others, 1990; Pashley, 1991; Nakabayashi & Saimi, 1996). The SEM observation indicates that single application bonding systems have the ability to dissolve the smear layer. After removing the smear layer, the conditioned dentin

95 Miyazaki, Iwasaki & Onose: Bonding Performance of Single Application Bonding Systems 93 surface should be wetted by hydrophobic resin monomers. Improved wetting of the tooth surface may encourage the penetration of resin monomers into the dentin substrate in order to enhance adaptation and create a hybrid layer. The role of a dentin primer is to improve wettability of the dentin surface by the adhesive and enhance monomer penetration into the hydrophilic dentin substrate. Single application bonding systems rely on their adhesives for wetting in order to create good adaptability to the restored tooth. In order to create a hybrid layer, resin monomers have to penetrate the demineralized dentin surface (Nakabayashi & others, 1982). The hybrid layer thickness of one-application bonding systems was very thin after argon-ion etching of the resin-dentin interface. Thickness of the hybrid layer depends on the bonding systems used, and thickness variations were observed among specimens using the same bonding system and different dentin substrates (Uno & Finger, 1996; Vargas, Cobb & Denehy, 1997; Prati & others, 1998). Though different hybrid layer thicknesses were observed, no correlation with dentin bond strength was found. The hybrid layer and the presence of resin tags may not be the only mechanisms influencing dentin strengths. The hybrid and adhesive layers may act as strainabsorbing layers in bonding systems (Kemp-Schölte & Davidson, 1990). Dentin-resin interface is comprised of several layers of materials with differing mechanical properties. The elastic modulus of the successive layers across a resin-dentin bonding area was determined by a nano-indentation technique, and a gradient of elastic modulus was observed from the rather stiff dentin over a more flexible hybrid layer and bonding agent layer to the stiffer resin composite (Van Meerbeek & others, 1993b). An elastic bonding layer might relieve stresses induced by polymerization shrinkage of resin composites, thereby improving the marginal integrity and retention of the restoration (Perdigão & others, 1996). Adhesives of RB and OF create a rather thick layer between the restoratives and tooth substrate. Though their bond strengths are lower than two-step bonding systems, a thick layer of the adhesives of the single application bonding systems with sub-micron filler addition might assist in close adaptation of the restoration without gap formation. CONCLUSIONS Current developments in adhesive systems have focused on simplifying the application methods by decreasing the time and steps required for placement. From the results of this study, the bonding ability of single application adhesive systems tested in this study was compatible with the compomer restorative system, which also utilizes the single application system. While these simplified adhesive systems are convenient, long-term clinical trials are needed to fully understand the performance of the newly developed bonding systems. Acknowledgements This work was supported, in part, by grant-in-aid Scientific Research (A) (1) and (B) (2) from the Ministry of Education of Japan. (Received 16 April 2001) References Barkmeier WW, Shaffer SE & Gwinnett AJ (1986) Effects of 15 vs 60 second enamel acid conditioning on adhesion and morphology Operative Dentistry 11(3) Barkmeier WW, Los SA & Triolo PT Jr (1995) Bond strengths and SEM evaluation of Clearfil Liner Bond 2 American Journal of Dentistry 8(6) Bates D, Retief DH, Jamison HC & Denys FR (1982) Effects of acid etch parameters on enamel topography and composite resin-enamel bond strength Pediatric Dentistry 4(2) Blosser RL (1990) Time dependence of 2.5% nitric acid solution as an etchant on human dentin and enamel Dental Materials 6(2) Buonocore MG (1955) A simple method of increasing the adhesion of acrylic filling materials to enamel surfaces Journal of Dental Research Buonocore MG, Matsui A & Gwinnett AJ (1968) Penetration of resin dental materials into enamel surfaces with reference to bonding Archives of Oral Biology 13(1) Eick JD, Cobb CM, Chappell RP, Spencer P & Robinson SJ (1991) The dentinal surface: Its influence on dentinal adhesion Part I Quintessence International 22(12) Erickson RL (1992) Surface interactions of dentin adhesive materials Operative Dentistry 5(Supplement) Ferrari M & Davidson CL (1996) In vivo resin-dentin interdiffusion and tag formation with lateral branches of two adhesive systems Journal of Prosthetic Dentistry 76(3) Finger J & Balkenhol M (1999) Practitioner variability effects on dentin bonding with an acetone-based one-bottle adhesive Journal of Adhesive Dentistry 1(4) Fowler CS, Swartz ML, Moore BK & Rhodes BF (1992) Influence of selected variables on adhesion testing Dental Materials 8(4) Gordan VV, Vargas MA, Cobb DS & Denehy GE (1997) Evaluation of adhesive systems using acidic primers American Journal of Dentistry 10(5) Gwinnett AJ & Matsui A (1967) A study of enamel adhesives. The physical relationship between enamel and adhesive Archives of Oral Biology 12(12) Ikemura K, Kouro Y & Endo T (1996) Effect of 4-acryloxyethyltrimellitic acid in a self-etching primer on bonding to ground dentine Dental Materials Journal 15(2)

96 94 Operative Dentistry Inokoshi S, Hosoda H, Harnirattisai C & Shimada Y (1993) Interfacial structure between dentin and seven dentin bonding systems revealed using argon ion beam etching Operative Dentistry 18(1) Ishioka S & Caputo AA (1989) Interaction between the dentinal smear layer and composite bond strengths Journal of Prosthetic Dentistry 61(2) Jordan RE, Suzuki M & Davidson DF (1993) Clinical evaluation of a universal dentin bonding resin: Preserving dentition through new materials Journal of the American Dental Association 124(11) Kemp-Schölte CM & Davidson CL (1990) Complete marginal seal of Class V resin composite restorations effected by increased flexibility Journal of Dental Research 69(6) Kitasako T, Nakajima M, Pereira PN, Okuda M, Sonoda H, Otsuki M & Tagami J (2000) Monkey pulpal response and microtensile bond strength beneath a one-application resin bonding system in vivo Journal of Dentistry 28(3) Miyazaki M, Onose H & Moore BK (2000) Effect of operator variability on dentin bond strength of two-step bonding systems American Journal of Dentistry 13(2) Munksgaard EC & Asmussen E (1984) Bond strength between dentin and restorative resin mediated by mixture of HEMA and glutaraldehyde Journal of Dental Research 63(8) Nakabayashi N, Kojima K & Masuhara E (1982) The promotion of adhesion by the infiltration of monomers into tooth substrates Journal of Biomedical Materials and Research 16(3) Nakabayashi N & Saimi Y (1996) Bonding to intact dentin Journal of Dental Research 75(9) Nakamichi I, Iwaku M & Fusayama T (1983) Bovine teeth as possible substitutes in the adhesion test Journal of Dental Research 62(10) Pashley DH (1991) Dentin bonding: Overview of the substrate with respect to adhesive material Journal of Esthetic Dentistry 3(2) Perdigão J, Lambrechts P, Van Meerbeek B, Braem M, Yildiz E, Yucel T & Vanherle G (1996) The interaction of adhesive systems with human dentin American Journal of Dentistry 9(4) Perdigão J & Lopes M (1999) Dentin bonding Questions for the new millennium Journal of Adhesive Dentistry 1(3) Prati C, Biagini G, Rizzoli C, Nucci C, Zucchini C & Montanari G (1990) Shear bond strength and SEM evaluation of dentinal bonding systems Journal of Dentistry 3(6) Prati C, Chersoni S, Mongiorgi R & Pashley DH (1998) Resininfiltrated dentin layer formation of new bonding systems Operative Dentistry 23(4) Sano H, Kanemura N, Burrow MF, Inai N, Yamada T & Tagami J (1998) Effect of operator variability on dentin adhesion: Students vs dentists Dental Materials Journal 17(1) Schilke R, Bauss O, Lisson JA, Schuckar M & Geurtsen W (1999) Bovine dentin as a substitute for human dentin in shear bond strength measurements American Journal of Dentistry 12(2) Senda A, Kamiya K, Gomi A & Kawaguchi T (1995) In vitro and clinical evaluations of a dentin bonding system with a dentin primer Operative Dentistry 20(2) Shaffer SE, Barkmeier WW & Kelsey WP III (1987) Effects of reduced acid-conditioning time on enamel microleakage General Dentistry 35(4) Söderholm KJ (1991) Correlation of in vivo and in vitro performance of adhesive restorative materials: A report of the ASC MD156 task group on test methods for the adhesion of restorative materials Dental Materials 7(2) Suzuki T & Finger WJ (1988) Dentin adhesives: Site of dentin vs bonding of composite resins Dental Materials 4(6) Tyas MJ (1992) One-year clinical performance of PMDM-based dentine bonding agents Australia Dental Journal 37(6) Tyas MJ & Chandler JE (1993) One-year clinical evaluation of three dentin bonding agents Australia Dental Journal 38(4) Triolo PT Jr & Swift EJ Jr (1992) Shear bond strengths of ten dentin adhesive systems Dental Materials 8(6) Uno S & Finger WJ (1996) Effects of acidic conditioners on dentine demineralization and dimension of hybrid layers Journal of Dentistry 24(3) Vargas MA, Cobb DS & Denehy GE (1997) Interfacial micromorphology and shear bond strength of single-bottle primer/adhesives Dental Materials 13(5) Van Meerbeek B, Inokoshi S, Braem M, Lambrechts P & Vanherle G (1992) Morphological aspects of the resin-dentin interdiffusion zone with different dentin adhesive systems Journal of Dental Research 71(8) Van Meerbeek B, Braem M, Lambrechts P & Lambrechts G (1993a) Two-year clinical evaluation of two dentine-adhesive systems in cervical lesions Journal of Dentistry 21(4) Van Meerbeek B, Willems G, Celis JP, Roos JR, Braem M, Lambrechts P & Vanherle G (1993b) Assessment by nanoindentation of the hardness and elasticity of the resin-dentin bonding area Journal of Dental Research 72(10) Van Meerbeek B, Perdigão J, Lambrechts P & Vanherle G (1998) The clinical performance of adhesives Journal of Dentistry 26(1) Wilson HHF & Wilson MA (1995) The outcome of a clinical trial of a dentin bonding system: Justice or injustice? American Journal of Dentistry 8(2)

97 Operative Dentistry, 2002, 27, Clinical Technique/Case Report A Technique of Occlusal Adjustment for Food Impaction in the Presence of Tight Proximal Contacts DH Newell V John S-J Kim Clinical Relevance While food impaction is generally associated with open contacts, little information is available regarding the management of food impaction associated with tight contacts. This article presents a technique of occlusal adjustment for food impaction associated with tight tooth contacts. The outcomes suggest this is a very successful, clinically relevant technique of patient management in the routine practice of restorative dentistry. SUMMARY Food impaction occasionally occurs in interproximal sites even though the contacts are tight. In this study of 14 patients with food impaction involving tight contacts, the lack of adequate food escape grooves was common to all 14 sites. Uneven marginal ridges and prominent opposing cusps were less common and, together, made up slightly more than half of the contact sites. An occlusal adjustment technique to create adequate food escape grooves, as well as reduce Department of Periodontics and Allied Dental Programs, Indiana University School of Dentistry, 1121 West Michigan Street, Indianapolis, Indiana Donald H Newell, DDS, MS, associate professor Vanchit John, DDS, MSD, assistant professor Seok-Jin Kim, DDS, MSD, clinical associate professor prominent opposing cusps and correct uneven marginal ridges, completely eliminated food impaction in all but one site. This site achieved an 80% reduction in food impaction. Based on the results of this retrospective study, the lack of adequate food escape groves in teeth adjacent to a contact point manifesting food impaction appears to be the primary factor. Creating food escape grooves adjacent to the marginal ridges eliminates, or nearly eliminates, food impaction in tight contact sites. INTRODUCTION Hirschfeld (1930) defined food impaction as the forceful wedging of food through occlusal pressure into interproximal spaces. He classified it into five etiologic categories: Class I (occlusal wear); Class II (loss of support proximally); Class III (extrusion beyond the occlusal plane); Class IV (congenital tooth abnormalities) and Class V (improper restorative design). Food impaction

98 96 Operative Dentistry may result from uneven marginal ridges, plunger cusps, excessive anterior overbite, open contacts and defective restorations (Carranza, 1996). It may contribute to plaque-induced inflammation and root caries in the presence of poor oral hygiene (Hallmon & others, 1996). A number of early publications stressed the importance of marginal ridge relationships (Mosteller, 1950; Goldman, Schluger & Fox, 1956a; Kraus, Jordan & Abrams, 1969; Prichard, 1972a; Johnson, 1973 and Burch, 1975). Prichard (1972a) reported that uneven, adjacent marginal ridges create a step that encourages food impaction. Goldman & others, (1956a) stated that uneven marginal ridges are a potential source of food impaction with their undesirable sequellae. Mosteller (1950) believed that uneven ridges were a significant predisposing factor in periodontal disease. These opinions were mostly based on clinical observations rather than evidence from controlled clinical studies. Later reports of a more scientific nature have shown contradictory results. Kepic & O Leary (1978) examined marginal ridge discrepancies (MRD) in the posterior teeth of 100 patients. They correlated probing depths, plaque index, calculus, gingival status and attachment loss with MRD and found the correlations to be low. They concluded that effective hygiene could maintain interdental papillary health in the presence of MRD. Hancock & others (1980) examined the effect of interdental contacts on the periodontal status. They assessed food impaction, gingival inflammation, plaque, caries, calculus, restorations and overhangs. No significant relationship was noted between tight, loose or open contacts and gingival index or pocket depth. Only 4% of all interdental contacts manifested food impaction, and probing depths were greatest at open contacts with food impaction. Despite open contacts correlating with food impaction and food impaction correlating with increased probing depths, there was no direct correlation between open/loose contacts and probing depths. However, the authors concluded that the presence of food impaction contributed to periodontal pathosis. Koral, Howell & Jeffcoat (1981) reviewed 90 patient records that demonstrated evidence of open contacts on bite-wing radiographs with no evidence of open contacts on the contralateral side. Based on record entries and radiographic bone loss, the severity of periodontal disease was classified using the code established by the American Dental Association (Grant, Stern & Listgarten, 1988): Type I (gingivitis), Type II (early periodontitis), Type III (moderate periodontitis) and Type IV (advanced periodontitis). When compared by site, only the Type II group (early periodontitis) showed bone loss around the open contact site that exceeded bone loss around the control. The authors concluded that with the exception of a possible association in incipient disease, the results demonstrated that open contacts were not associated with a greater degree of bone loss than closed contacts. This may occur because food impaction loses its progressive detrimental effect after attachment loss progresses further apically from the occlusal surface. In a study of 104 patients, Jernberg, Bakdash & Keenan (1983) compared periodontal status adjacent to unilateral open contacts and contralateral closed contacts. Food impaction and occlusal interference were significantly more prevalent at open contacts. In addition, 60.6% of patients had greater clinical attachment loss (CAL) and 49% had deeper probing depths (PD) at open contact sites compared to 17.3% CAL and 22.1% deeper PD at closed contact sites. The literature indicates that interdental papillae associated with weak or open contacts may remain healthy if there is no associated food impaction and effective hygiene is maintained (Kepic & O Leary, 1978; Hallmon & others, 1996). However, if food impaction occurs, a significant likelihood of inflammation of the interdental gingival papilla and increased probing depth appears likely (Hancock, & others, 1980; Jernberg & others, 1983) (Figure 1A and B). While many dentists do not associate tight contacts with food impaction, there is evidence that this can occur. Hancock & others (1980) found that 10 out of 26 anterior sites and five out of 16 posterior sites associated with food impaction had tight contacts, although the mean pocket depth was less than for loose or open contacts. They also noted that for each type of contact where food impaction was present, mean pocket depth was greater than the overall mean pocket depth for that type of contact, which included those with no food impaction. Since tight contacts do not always ensure food impaction will not occur, it behooves therapists to recognize other potential causes that can be corrected to alleviate the problem. Goldman & others (1956a) stated, an important factor in keeping food from packing between the teeth is the occlusal anatomy. They were among the early authors who stressed the importance of good marginal ridge relationships along with proper occlusal and interproximal embrasures (Goldman & others, 1956b). In the presence of improper contacts, they recommended dressing down of the external surfaces of the teeth in these areas to prevent food impaction by providing adequate excursive exits for food deflection (Goldman & others, 1956a, Goldman & others, 1956b). Prichard (1972a) was also one of the few who included food escape grooves as one of the changes that may be made by selective grinding. In Prichard s textbook, Figure 19-4 demonstrated the restoration of good proximal contact relationships after occlusal adjustment in a 38-year-old woman (Prichard, 1972b).

99 Newell, John & Kim: A Technique of Occlusal Adjustment for Food Impaction 97 Figure 1A. Buccal view of a loose contact between teeth #3-4 with a prominent opposing mesial-buccal cusp of tooth #30. The patient complained of persistent food impaction. The literature has been inconclusive regarding the relationship between food impaction and periodontal irritation and attachment loss. Nevertheless, food impaction can be a significant psychological irritant to patients. If related to weak or open contacts, it is commonly felt that treatment for food impaction should be restoring the contacts of the involved teeth with a new restoration. On the other hand, if food impaction occurs with tight contacts, other factors may be involved that can be corrected with occlusal adjustment. Figure 2A. Diagram of an occlusal surface of teeth #3-4 with overly prominent buccal and lingual cusp ridges as they merge with the proximal marginal ridges (Arrows). These prominent ridges form a bowl with no food spillways to help deflect food away from the contact. PURPOSE In one of the authors practice (DHN), 14 patients over 12 years have presented with chief complaints of food impaction in interproximal sites with tight contacts. These situations frequently involved restorations in contact with each other or a restoration in contact with a nonrestored tooth. Occasionally, tight contacts between nonrestored teeth manifested food impaction, although this was not as common. In the tight contacts with food impaction, it was noted that the cuspal ridges were high as they merged with the marginal ridges, resulting in a bowl-like contour that would not allow buccal and lingual food deflection (Figures 2A and 3A, B and C). This would allow food particles to be deflected over the contact area as well as toward the main occlusal surface. This could then allow food to be wedged between the contacts interdentally by the opposing cusp. Figure 1B. Palatal view of the loose contact between teeth #3-4. Note the inflamed interdental gingival papilla. Occlusal adjustment that used diamond and polishing stones was incorporated to create buccal and lingual food escape grooves where the cuspal ridges joined the marginal ridges. This would allow food to escape bucally and lingually from the occlusal surfaces rather than being forced over marginal ridges wedged between the contacts by an opposing cusp. Prominent opposing cusps could also be reduced, if present. Based on Figure 2B. Diagram of a 5 mm long, 3.5 mm wide, coarse, friction grip, football diamond stone used to create food escape grooves (a) and a 7 mm long, 2.5 mm wide, friction grip, fine grit, flame-shaped, white Arkansas stone to smooth the altered surfaces (b). treatment documentation of the 14 patients treated with this technique, this paper assessed the effectiveness of creating food escape grooves to eliminate food impaction in sites with tight contacts. Figure 2C. Diagram shows food escape grooves prepared in teeth # 3-4 at adjacent to the marginal ridges. Figure 2D. Buccal view diagram of the contact between teeth #3-4 showing the shallow food escape grooves created by occlusal adjustment. TECHNIQUE The occlusal anatomy and marginal ridge relationship of each patient with a chief complaint of food impaction associated with a tight contact were evaluated for the presence of 1) adequate food escape grooves, 2) uneven marginal ridges and 3) a prominent cusp opposing the contact area. Table 1 shows the presence of these occlusal struc-

100 98 Operative Dentistry Table 1: Occlusal Presentation Patient Age Inadequate Food Uneven Marginal Prominent Opposing Restoration/ Restoration/ Natural Tooth/ Escape Grooves Ridges Cusp Restoration Natural Tooth Natural Tooth = the presence of each category. tures. When no food escape grooves were detected, grooves were prepared in the cuspal ridges just adjacent to the marginal ridges (Figures 2B, C and D, and 3 D and E) with a 5 mm long, 3.5 mm wide coarse, friction grip, football diamond stone (Figure 4). If the marginal ridges were uneven, the more prominent ridge was reduced. If a prominent opposing cusp was detected, it was reduced at the contact point while maintaining light occlusal contacts. All areas altered by the diamond stone were then smoothed with a 7 mm long, 2.5 mm wide friction grip, fine grit, flame-shaped, white Arkansas stone (Figure 4). Care was taken to remove a minimal amount of tooth structure or a restoration to accomplish these goals (Figures 2C and D and 3D and E). The patients were reevaluated one to two weeks later and asked to rate the improvement on a scale of 0 to 10, with 0 representing no improvement and 10 representing complete elimination of food impaction. Also, during the first re-evaluation visit, additional occlusal correction was done if Table 2: Treatment Results Patient Age Ranking of Food Impaction Elimination by Appointment* First Appointment Second Appointment = no food impaction reduction. 10 = complete food impaction reduction. there was a noticeable but incomplete reduction of food impaction. The patient returned for a second re-evaluation and was again asked to rate the result. RESULTS Occlusal Presentation (Table 1) The 14 patients ranged in age from years, with a mean age of 33 years. Eight of the 14 tight contact sites had adjacent restorations (57.1%); four had a restoration in contact with a natural tooth (28.6%) and two had adjacent natural teeth in contact (14.3%). All teeth involved in the 14 contact sites had inadequate food escape grooves. Three of the 14 sites had uneven marginal ridges (21.4%). Two of these three sites had adjacent natural teeth in contact (66.7%) and one had a natural tooth in contact with a restoration (33.3%). Five of the 14 food impaction sites occluded with a prominent opposing cusp (35.7%). Of these five prominent cusps, four opposed two restorations in contact (80.0%) and one opposed a restoration in contact with a natural tooth (20.0%). Three of the 14 tight contact sites presented with uneven marginal ridges (21.4%). Two of the uneven marginal ridge sites were associated with adjacent natural teeth (66.7%) and one had a restoration contacting a natural tooth (33.3%). Treatment Results: (Table 2) In 13 of the 14 tight contact sites (92.9%), patients rated elimination of food impaction at 10 (complete elimination) after one or two occlusal adjustment appointments. Of these 13 sites, nine achieved total elimination

101 Newell, John & Kim: A Technique of Occlusal Adjustment for Food Impaction 99 Figure 3A. Occlusal view of teeth #13, 14 and 15 with food impaction between #13-14 despite tight contacts. Note the prominent cusp ridges (Arrows) that merge with the mesial marginal ridge of #14 forming a bowl that blocks food escape. There are slightly similar contours between the tight contact of tooth #2 with #3 but there was no food impaction. Figure 3C. Palatal view of the tight contact area between tooth # Note the prominent palatal cusp ridges (arrows) that also prevent palatal food escape during mastication. Figure 3E. Palatal view of teeth #13, 14 and 15 after occlusal adjustment and polishing to create palatal food escape grooves on the distal of tooth #13 and the mesial of #14. A similar preventive groove was prepared on the distal of tooth #14. of food impaction after one appointment (64.3%), while four sites required a second occlusal adjustment appointment to achieve complete elimination of food Figure 3B. Buccal view of the tight contact of tooth #13 with #14. Note the prominent buccal cusp ridges (arrows) that prevent buccal food escape during mastication and allow food to be passed over the contact area and impacted. Figure 3D. Buccal view of teeth #13, 14 and 15 after occlusal adjustment and polishing to create buccal food escape grooves on the distal of #13 (small arrow) and the mesial of #14 (large arrow). This completely eliminated food impaction. A similar food escape groove was prepared on the distal cusp ridge (angled arrow) as a preventive measure since there was no reported food impaction. Figure 4. A 5-mm long, 3.5-mm wide, coarse, friction grip, football diamond stone (bottom) used to create food escape grooves and a 7 mm long, 2.5-mm wide, friction grip, fine grit, flame-shaped, white Arkansas stone (top) used to smooth the altered surfaces. impaction (28.6%). One site did not achieve complete elimination of food impaction after two occlusal adjustment appointments (7.1%) but achieved a reduction of eight, two points shy of complete reduction. DISCUSSION While food impaction is most common in sites with loose or open contacts, it occasionally occurs with tight contacts (Hancock & others, 1980). Uneven marginal ridges and prominent opposing cusps have also been implicated as a predisposing factor (Goldman & others, 1956a; Prichard, 1972a). However, in this retrospective study of food impaction associated with tight contacts, uneven marginal ridges were present in only three of the 14 impaction sites (21.4%). Two occurred with two natural teeth in contact and one with a natural tooth in contact with a restoration. Five prominent opposing cusps were associated with the 14 impaction sites (35.7%) but seven sites had neither uneven marginal ridges or prominent opposing cusps (50.0%). Inadequate food escape grooves were the most common occlusal discrepancy occurring in the teeth adjacent to 100% of the tight contact sites with food impaction. Of the five sites that failed to achieve complete elimination of food impaction after one occlusal adjustment, three involved uneven marginal ridges. The other two sites were involved with a prominent opposing cusp. One failed to achieve complete elimination of food impaction even after two occlusal adjustment appointments. This site involved two restorations in contact and a prominent opposing cusp in addition to inadequate food escape grooves. The prominent cusps and inadequate enlargement of the food escape grooves may have accounted for the incomplete

102 100 Operative Dentistry elimination of food impaction. Nevertheless, the patient estimated that food impaction was reduced by approximately 80% after the second adjustment, and he appeared satisfied. No further occlusal adjustment was done in an attempt to complete elimination of food impaction because the patient moved from the area. The only occlusal discrepancy common to all 14 impaction sites was the lack of adequate food escape grooves. It made no difference whether there were restored teeth in contact, a restoration in contact with a natural tooth or two natural teeth in contact food escape grooves were minimal or non-existent. However, slightly more than half involved two restored teeth (57.1%) and 28.5% involved a restoration and a natural tooth. Since 85.6% of the contact sites involved one or two restorations, this would suggest that some restorations are not fabricated with adequate food escape grooves. Creating or enlarging food escape grooves by occlusal adjustment successfully eliminated food impaction in all but one of the 14 tight contact sites. That one site achieved an 80% reduction. Since this was not a designed study and the main objective of therapy was to eliminate food impaction, it is difficult to know to what degree the elimination of uneven marginal ridges and the reduction of prominent opposing cusps contributed to the success of this technique. However, this occlusal adjustment technique was successful in eight of the 14 impaction sites with inadequate food escape grooves but no uneven marginal ridges or prominent opposing cusps. Therefore, creating adequate food escape grooves without correcting existing uneven marginal ridges and reducing prominent opposing cusps might be successful in eliminating food impaction. Therapists attempting this technique to alleviate food impaction in tight contact sites may be concerned about perforating a restoration or exposing dentin with resulting hypersensitivity. This was not a problem in this study because restorations with over-contoured cuspal ridges blocking food escape are thick enough that conservative creation of food escape grooves will not perforate the restoration. Also, in natural teeth, these grooves are usually created in enamel and no dentin is exposed. CONCLUSIONS Within the limitations of this retrospective study, the following conclusions can be reached: 1. Inadequate food escape grooves have a positive association with food impaction in tight contact sites. 2. Creating food escape grooves by occlusal adjustment is highly successful in eliminating or greatly reducing food impaction in tight contact sites. Acknowledgements The authors thank Mark Dirlam, Thomas Meador and Amanda Marlett of the Indiana University School of Dentistry Dental Illustrations Division for their assistance with the illustrations. (Received 1 November 2001) References Burch JG (1975) Periodontal considerations in operative dentistry Journal of Prosthetic Dentistry 34(2) Carranza FA Jr (1996) Glickman s Clinical Periodontology 8 th ed Philadelphia WB Saunders pp Goldman HM, Schluger S & Fox L (1956a) Periodontal Therapy St Louis CV Mosby Co pp Goldman HM, Schluger S & Fox L (1956b) Periodontal Therapy St Louis CV Mosby Co p 390. Grant DA, Stern IB & Listgarten MA (1988) Periodontics 5 th ed St Louis CV Mosby Co p 562. Hallmon WW, Carranza FA Jr, Drisko CL, Rapley JW & Robinson P editors (1996) Periodontal Literature Reviews Chicago American Academy of Periodontology pp Hancock EB, Mayo CV, Schwab RR & Wirthlin MR (1980) Influence of interdental contacts on periodontal status Journal of Periodontology 51(8) Hirschfeld I (1930) Food impaction Journal of the American Dental Association Jernberg GR, Bakdash MB & Keenan KM (1983) Relationship between proximal tooth open contacts and periodontal disease Journal of Periodontology 54(11) Johnson RH (1973) Periodontal therapy Ontario Dentistry 50(11) Kepic TJ & O Leary TJ (1978) Role of marginal ridge relationships as an etiologic factor in periodontal disease Journal of Periodontology 49(11) Koral SM, Howell TH & Jeffcoat MK (1981) Alveolar bone loss due to open interproximal contacts in periodontal disease Journal of Periodontology 52(8) Kraus BS, Jordan RE & Abrams LA (1969) Dental Anatomy and Occlusion Baltimore The Williams and Wilkins Company p 251. Mosteller JH (1950) The etiology of periodontal disease: A review of current literature Journal of Periodontology Prichard JF (1972a) Advanced Periodontal Disease/Surgical and Prosthetic Management ed 2 Philadelphia WB Saunders p 827. Prichard JF (1972b) Advanced Periodontal Disease/Surgical and Prosthetic Management ed 2 Philadelphia WB Saunders p 821.

103 101 Awards American Academy of Gold Foil Operators Clinician of the Year Award Dr Wendell A Foltz The stress of an active daily practice in dentistry often pushes out the precision required of Direct Gold as one of the services offered to our patients. All too frequently, many fine operators give up foil because of its requirements for exactness and attention to detail. The Clinician of the Year award was established to honor our younger members Wendell A Foltz who have maintained an active interest in direct gold as well as being excellent operators in our profession. This year s recipient is the apotheosis of this intention. Dr Wendell Foltz more than fills all the requirements for this honor. Wendell has a native ability for excellence in operative dentistry that most of us struggle to obtain over many years. He has the unique ability to solve dental problems outside of routine textbook design. He is a superior operator with a common sense approach to the complexities of today s dentistry. Wendell s interest in dentistry started at an early age by stealing anesthetic carpules from his dentist s garbage to use in his junior microscope kit. I will not go into his many exploits during his formative years, but after he started dental school in 1973, he used up all of his gold allotment making jewelry. I guess that is one way of getting the attention of the most beautiful girl on campus. In dental school, he received the ICD Outstanding Achievement Award and was one of the top students in his class. Along with his very busy practice, he has devoted much time to the DDS program, Northwest medical teams and numerous dental health screening programs. Wendell is involved in several study clubs, as well as other forms of continuing education. Terry Tanaka has always said that we should read two dental articles each day, and I am sure Wendell does that and more. Besides being a fine dentist and constant student, this man has a tremendous number of outside interests. He is a farmer, pilot, motorcycle rider, fisherman, hunter and white water rafter. If you want a story, get Wendell to tell you about his bike trip to Mt Hood and back. When there is a bear and rattlesnake barbecue, Wendell is there. Wendell and his wife, Gloria, have been married for 32 years and have three children Andrew, who is in the Air Force in Germany, Mindy, a veterinarian and recent first-time mother and Sarah, a homemaker and mother of two boys. Congratulations to you Wendell as the 2001 American Academy of Gold Foil Operators Clinician of the Year. Chester Gibson

104 102 Operative Dentistry American Academy of Gold Foil Operators Distinguished Member Award Dr Melvin R Lund It is a privilege to have an opportunity to publicly honor someone who has had an indelible impact on so many professional careers, including my own. It is a double pleasure to be asked by the recipient, himself, to deliver this honor. Thus, I find myself in the enviable position of presenting the American Academy of Gold Foil Operators Distinguished Member Award to Dr Melvin R Lund, affectionately known as Pug to his friends. Melvin R Lund Over the 28 years of our association, I have learned many things about this man. He was born in Wisconsin and spent his early years there before moving to the State of Washington to finish high school and complete his pre-dental training. I want to note that he was honored in 2001 as Alumnus of the Year by his high school, Auburn Academy. Dr Lund received his DMD from the University of Oregon in 1946, married his wife, Marg and served in Korea with the Army Dental Corps for two years before leaving the service and setting up private practice in Camas, Wash. In 1953, Pug decided on an academic career and went to the University of Michigan to bolster his credentials by obtaining a Master s of Science Degree in Restorative Dentistry. He then joined the founding faculty at Loma Linda University School of Dentistry and spent the next 16 years working with Lloyd Baum and others to create an outstanding educational institution. Our paths first crossed in 1973 when I entered his Graduate Operative Dentistry Program at Indiana University School of Dentistry. He had assumed leadership of the Department of Operative Dentistry in 1971 after chairing the Restorative Department at Loma Linda for 13 years. My first impressions were of a kind and gentle man who exuded an air of wisdom, competence and confidence with no trace of ego. He seemed to be completely comfortable and at peace with himself and quickly made his new residents feel welcome. He and Marg immediately treated us as members of their family and made sure that we understood that we had a new home away from home. As a mentor, Pug had a way of making us want, more than anything else, to please him with our work and accomplishments. He has a knack of bringing out the best in his students by correcting without criticism, demonstrating without intimidating, demanding without discouraging and constantly promoting the concept of excellence. Dr Lund s contributions to his profession, church and community are exemplary and extensive. Whether it is the National Association of Seventh Day Adventist Dentists, Lions International or one of 13 dental academies, he has never just been a member, but always actively served on various committees, boards and councils or held prominent office. Pug is one of the founding members of the American Board of Operative Dentistry, where he continues to serve as an examiner and is an active reviewer for our journal, Operative Dentistry. He has authored or contributed to 12 textbooks, published more than 45 papers and lectured extensively both nationally and internationally. He has also been extremely active in promoting his first love, direct gold. Dr Lund s dedication to his profession has been recognized with awards from both Loma Linda and Indiana universities, from his graduate students and with the 1991 Award of Excellence from the Academy of Operative Dentistry. Somehow, he still finds time to be a loving father to three children Mark, Chris and Kelly, and grandfather to nine grandchildren (we re not sure if they will make it into double figures, but it is a possibility). Pug is also an active snow skier, tennis player and carrier of all the antiques that Marg finds to furnish their home. I can t begin to estimate how many lives have been touched and improved by Pug s humanitarian approach to all things. He is loved and respected by his students and colleagues and, with Marg by his side, has gathered, trained and nurtured an international family of exceptional dental professionals who are carrying the torch of excellence all over the world. Although he has officially retired and holds the title of Professor Emeritus, Pug continues to contribute to our educational program on a weekly basis. He teaches one day a week in the pre-doctoral technique course, another day a week in the Graduate Operative Dentistry clinic and also lectures in several of the graduate classes. He spends hours in the library and on the Internet so that he can stay current for the students. If I were forced to single out one attribute that makes Pug truly special, it would be that he has the unique talent of simply inspiring by example both in his personal relationships and his professional activities. I have been privileged to benefit from his teaching and mentoring and am extremely proud to be his friend. I can t think of anyone who is more deserving of recognition by this Academy than Dr Melvin R Lund, and it is an honor to present him with this Distinguished Member Award. Michael A Cochran

105 103 Departments Book Reviews Bleaching Techniques in Restorative Dentistry by Linda Greenwall Editors: Linda Greenwall, BDS, MGDSRCS, MRDRCS, MSc, FGDP, George A Freedman, DDS, FFACD, Valeria V Gordan, DDS, MS, Martin Kelleher, BDS, MSc, FDS, Gerald McLaughlin, DDS and Ilan Rotstein, DDS Published: The Livery House, 7-9 Pratt Street, London NW1 0AE Dr Linda Greenwall is a private practitioner in London, UK and Cape Town, South Africa, who has studied the techniques of bleaching for more than 10 years. She completed her MSc in 1992 and wrote a thesis entitled The effects of carbamide peroxide on tooth color and structures: An in vitro investigation. She has accumulated a wealth of clinical cases that she shares in this publication. Because bleaching has only recently been accepted in the UK, Dr Greenwall acknowledges that many of the cases are from others who have assisted in the preparation of the manuscript. This book is designed to help dentists to successfully incorporate the wide variety of bleaching treatments into their practices. Dr Greenwall does this effectively through the written word and illustrations. It appears to have been written mainly for clinicians and students; however, since it is very user friendly, it can also be used to educate patients on the limitations and advantages of the different methods of bleaching in restorative dentistry. Dr Greenwall states up front that dentists wishing to provide bleaching services for their patients should be acquainted with their country s legal standing before undertaking any bleaching treatments. Concepts are meticulously referenced. While it is difficult to document a process that is dynamic in nature, such as bleaching, the author pinpoints concepts that have not been previously stressed, such as the importance of preparing a model for the fabrication of a tray. She identifies well-constructed trays as being essential to successful bleaching treatment and patient compliance. Dr Greenwall covers the advantages and disadvantages of reservoirs, then gives a 10-step guide to fabricating a wellfitting tray. This book features more than 500 color illustrations; the color photographs are excellent in their skin tones. The author has gone to great lengths to choose her photographs. In a few cases the author has included some excellent clinical cases at the end of the chapters, introducing new concepts that are not covered in the chapter, such as the bleaching of endodontically treated teeth at the end of Chapter 4. Intracoronal bleaching of non-vital teeth is covered in Chapter 8. Such double coverage is not unusual in first editions of textbooks. The book covers in-office bleaching and microabrasion in the same thorough way that it covers at-home bleaching. The chapter on Safety Issues is covered on a clinical level and gives answers to questions patients ask, however, it does not always address both sides of the controversies in bleaching. This book includes an excellent chapter on marketing of bleaching. This topic is seldom discussed in lectures, and most dentists are unsure how to tap in to an existing market without causing ruffled feathers within their patient population. This is a book that all dentists should have on their shelves to teach their patients and staff about the benefits and concerns related to bleaching. For those of us interested in bleaching, this is a book you will not be able to put down until it has been read from cover to cover. Thanks to Dr Greenwall, this up-to-date guide helps the busy restorative dentist who needs to keep abreast of the latest types of bleaching products available, learn the latest techniques related to how to use them safely and effectively and offers additional information on how to use them in combination with restorative dentistry when treating patients. Not only is Bleaching Techniques in Restorative Dentistry a good reference for dental practitioners, but this reviewer believes it should be available to undergraduate and graduate students in our dental school libraries. Announcements Tucker Institute Course Bruce A Matis, DDS, MSD, professor director of clinical research section Restorative Dentistry Indiana University School of Dentistry 1121 West Michigan Street Indianapolis, IN A five-day clinical course in conservative gold castings, mentored by Dr Richard V Tucker, will be conducted on June 10-14, 2002 at the University of Washington Dental School For course information, please contact Dr Dennis Miya at 206/ or via at diichi@aol.com.

106 104 Operative Dentistry Classifieds: Faculty Positions The University of Texas Houston The University of Texas Health Science Center at Houston, Dental Branch seeks applicants for a fulltime, tenure/clinical educator track position in the Department of Restorative Dentistry and Biomaterials at the Assistant or Associate Professor level. Applicants must have a DDS/DMD degree with prior teaching and/or private practice experience. Advanced training in Operative Dentistry, Prosthodontics or GPR/AEGD certification is preferred. Responsibilities include clinical and preclinical teaching to undergraduate and graduate dental students, research and service. The position is available September 1, Academic rank and salary are commensurate with qualifications and experience. The University of Texas Health Science Center at Houston is an EEO/AA employer. Women and minorities are encouraged to apply. Send a letter of application, a curriculum vitae and a list of three references to: Dr William Tate, University of Texas Dental Branch, Department of Restorative Dentistry and Biomaterials, 6516 M D Anderson Blvd, Suite 493, Houston, TX The University of British Columbia The Faculty of Dentistry invites applications for a fulltime position in the Department of Oral Health Sciences. Applicants should preferably have postgraduate training in restorative dentistry and research experience at the PhD level or the equivalent. The successful applicant will be expected to participate fully in teaching of restorative dentistry in the undergraduate and graduate dental curriculum, in research and faculty service. Preference will be given to individuals who can participate in the active research areas in the facility. In accordance with Canadian immigration requirements, priority will be given to Canadian citizens and permanent residents of Canada. UBC hires on the basis of merit and is committed to employment equity. We encourage all qualified persons to apply. Please respond by November 15, 2001 with curriculum vitae and contact addresses of three referees to: Dr Virginia M Diewert Head, Department of Oral Health Sciences Faculty of Dentistry, UBC 2199 Westbrook Mall Vancouver, BC V6T 1Z3 See our website at: Indiana University Restorative Dentistry Full-time, tenure track faculty position in Restorative Dentistry available September 1, Activities will involve pre-clinical and clinical instruction in Operative Dentistry, as well as some didactic instruction. Research activity will be encouraged. Candidates must be graduates of an ADA-accredited DDS/DMD program. Licensure or eligibility for licensure in Indiana is required. Academic rank and salary commensurate with qualifications and experience. Opportunity for one day of private practice per week is available. Send curriculum vita, three letters of recommendation and a personal statement to Dr E Steven Duke, Chair, Department of Restorative Dentistry, 1121 West Michigan Street, Room S316, Indianapolis, IN Indiana University is an Affirmative Action/Equal Opportunity Employer. Letters Correction Operative Dentistry received a letter from Dr Rosemary Shinkai offering a correction to her article In Vitro Evaluation of Secondary Caries Development in Enamel and Root Dentin Around Luted Metallic Restoration, Operative Dentistry (1) The correction states: There was a mistake in the printed equation of volume % mineral (page 54, mineral loss parameters, 4th line). The equation should read: Volume % mineral = 4.3.KHN-+11.3