Optimizing non-pb radiation shielding materials using bilayers

Transcription

1 Optimizing non-pb radiation shielding materials using bilayers J. P. McCaffrey, a E. Mainegra-Hing, and H. Shen Institute for National Measurement Standards, National Research Council of Canada, Building M-35, 1200 Montreal Road, Ottawa K1A 0R6, Canada Received 22 May 2009; revised 8 October 2009; accepted for publication 19 October 2009; published 10 November 2009 Purpose: The objective of this study was to demonstrate that the weight of non-pb radiation shielding materials can be minimized by structuring the material as a bilayer composed of different metal-powder-embedded elastomer layers. Methods: Measurements and Monte Carlo MC calculations were performed to study the attenuation properties of several non-pb metal bilayers over the x-ray energy range kev. Metals for the layers were chosen on the basis of low cost, nontoxicity, and complementary photoelectric absorption characteristics. The EGSnrc user code cavity.cpp was used to calculate the resultant x-ray fluence spectra after attenuation by these metal layers. Air kerma attenuation was measured using commercially manufactured metal/elastomer test layers. These layers were irradiated using the primary standard calibration beams at the Institute for National Measurement Standards in Ottawa, Canada utilizing the six x-ray beam qualities recommended in the German Standard DIN Both the measurements and the calculations were designed to approximate surface irradiation as well as penetrating radiation at 10 mm depth in soft tissue. The MC modeling point and the position of the measurement detector for surface irradiation were both directly against the downstream face of the attenuating material, as recommended in DIN Results: The low-z upstream/high-z downstream ordering of the metal bilayers provided substantially more attenuation than the reverse order. Optimal percentages of each metal in each bilayer were determined for each x-ray radiation beam quality. Conclusions: Depending on the x-ray quality, appropriate choices of two complementary metalembedded elastomer layers can decrease the weight of radiation shielding garments by up to 25% compared to Pb-based elastomer garments while providing equivalent attenuation American Association of Physicists in Medicine. DOI: / Key words: unleaded protective clothing, radiation shielding, x-ray attenuation, shielding weight, bilayer I. INTRODUCTION Radiation shielding materials have historically been constructed from a rubber, polymer, or elastomer polymer with elastic properties matrix material embedded with small lead particles. These materials are incorporated into garments, covers, and curtains for radiation protection purposes. Objections to the weight of Pb-based material and concerns about environmental disposal have encouraged the study and development of shielding materials incorporating non-pb materials either as single metal powders or as mixtures of metal powders Related studies have been undertaken to optimize protection for physicians and operators during cardiac catheterization and radiotherapy procedures using non-pb materials The attenuation properties of these materials are traditionally rated as 0.5 mm Pb-equivalent, 0.25 Pbequivalent, etc. Recent standards for radiation protection materials now include the requirement that such non-pb materials be tested over the energy range kev. 12 Because of variations related to different metals characteristic photoelectric absorption edges, it is difficult to choose a single metal that supplies effective attenuation over this entire energy range. The attenuation provided by commercial non-pb garment materials is in some cases less than what is claimed, 9 at least partly due to inefficient use of the attenuating properties of the mixtures of metals embedded in the matrix material. The photoelectric effect overwhelmingly dominates energy transfer and absorption in the kev energy range for the materials of interest. The characteristic K-absorption edges and emission lines for various metal attenuators result in substantial differences in spectra and consequently attenuation capabilities. As an illustration of spectral variation, Fig. 1 displays Monte Carlo MC -calculated photon fluence spectra for a 100 kvp HVL 6.3 mm Al x-ray beam attenuated by single layers of W, Bi, Sb, and Ba. Pb, the other metal of relevance in this study not shown, has a very similar spectral shape to Bi, but with the K-absorption edge and the emission peaks approximately 2 kev lower in energy. All metal layers in Fig. 1 have a mass thickness of g/cm 2, corresponding to the 0.5 mm Pb-equivalent definition. This mass thickness provides very strong attenuation, allowing only a small fraction of the initial x-ray beam energy to be transmitted. From the spectra in Fig. 1, the ratios of the calculated transmitted air kerma for the four metals to the calculated transmitted air kerma for Pb was 0.94, 0.98, 0.76, and Med. Phys , December /2009/36 12 /5586/9/$ Am. Assoc. Phys. Med. 5586

2 5587 McCaffrey, Mainegra-Hing, and Shen: Non-Pb radiation shielding bilayer materials 5587 FIG kvp X-ray fluence spectra resulting from attenuation by equivalent thicknesses of W, Bi, Sb, and Ba. All spectra display the characteristic K-absorption edges of the individual materials. Enhanced W emission peaks contributed by the W X-ray tube target are present in all spectra at approximately 59 and 67 kev. for W, Bi, Sb, and Ba, respectively. All four of these metals provide superior attenuation to Pb at this 100 kvp x-ray quality. Virtually all the energy in the x-ray spectra below 40 kev in this figure is absorbed by the metal layers, with the exception of the peaks corresponding to the Sb and Ba photoelectric emission line energies. W absorbs strongly above its characteristic 69.5 kev K-absorption edge, and then reemits some of this energy at energies lower than the absorption edge, most strongly in the characteristic W K and W K emission lines at approximately 59 and 67 kev, respectively. The W K emission line peak extends beyond the top of the graph to a value of 7 on this figure s ordinate scale, which was truncated in this figure to show more of the detail of the four spectra. All four spectra and particularly W include an energy enhancement at the W emission lines, reflecting the contribution from the W target in the x-ray tube. The different absorption and emission characteristics of these four metals offer opportunities for tailoring the resulting transmitted spectra by judicious use of complimentary absorption effects. Multilayer attenuators have been used for many years for hardening x-ray beams narrowing the x-ray energy spectra. For example, Thoraeus filters employ an initial layer of tin combined with a second layer of copper to absorb the characteristic fluorescent radiation of the tin and a third layer of aluminum to absorb the characteristic radiation of the copper. In these filters, the layers are positioned in the beam in descending order of atomic number Z, to allow subsequent layers to remove the fluorescence x-rays that originate in the higher-z layer upstream. 13 For low-energy kev and medium-energy kev x-rays, this approach minimizes the energy spread, with the intention of approaching a monoenergetic beam. For the purposes or radiation attenuation, the primary requirement is the minimization of the total energy transmitted through the material. Preliminary MC calculations performed for this study revealed that with a bilayer two adjacent metal attenuating layers, an increasing order of Z results in the greatest attenuation for metals with Z 48. An important qualification that became evident during this study was that calculations are very effective in providing the design of the layered material, but the actual attenuating capabilities of any manufactured radiation shielding material must ultimately be determined by measurements, since the effects of the matrix material and the quality of the manufacturing process will influence the attenuating properties of the material. Therefore, the purpose of this study was not to provide exact results or recipes, but to demonstrate the effectiveness of bilayers in maximizing the radiation attenuation properties and minimizing the weight of non-pb radiation attenuating materials. All measurements and calculations in this study pertain to 0.5 mm Pb-equivalent materials. II. MATERIALS AND METHODS Experimental attenuation measurements are timeconsuming, so extensive preliminary MC calculations were performed to lower the number of actual measurements required to a few hundred. Metal layers were chosen on the basis of low cost, nontoxicity, and photoelectric absorption edges in the kev range. Focusing on simple bilayers further narrowed the choices. MC calculations were utilized to optimize the ordering in the bilayer systems. Since materials for radiation shielding are generally discussed under the topic of radiation protection, the MC calculations and experimental measurements were designed to approximate surface irradiation of tissue as well as penetrating radiation at 10 mm depth in tissue. The comparison of measurements to MC calculations was complicated by the manufacturing processes used to produce radiation shielding materials. These processes result in consistency variations between different manufacturers and even between different manufacturing runs. Some complications and compromises inherent in this study include Measurements and modeling could not exactly match the composition of commercially available radiation shielding materials because of the wide variety of component materials and proprietary processes. The total bilayer thickness and order of the commercial metal/elastomer layers that were available to us for measurements varied with each combination. A 1.2 mm 3/64 in., Lucite layer was required in front of our detector for the measurements approximating surface irradiation, and a 10 mm thick Lucite layer was used to simulate the 10 mm tissue depth required to approximate penetrating radiation. Including the thin window of the detector, up to 14 interfaces comprised the measurement geometry. For MC calculations, these and other complications meant that exact modeling of these geometries became impractical. However, meaningful comparisons between measurements and calculations could still be attained by basing compari-

3 5588 McCaffrey, Mainegra-Hing, and Shen: Non-Pb radiation shielding bilayer materials 5588 TABLE I. Metal/elastomer sheets used in this study. Metal powder in elastomer Atomic No. Z Density g/cm 3 Metal /elastomer sheet density g/cm 3 K-absorption edge kev Antimony Sb Barium Ba N/A 37.4 Tungsten W Bismuth Bi sons on attenuation ratios. MC calculations of transmitted spectra were based on modeling the bilayer as two pure metal layers. The transmitted air kerma through these bilayers was calculated, and this value was then divided by the calculated air kerma transmitted through an equivalent thickness of pure Pb. This provided a calculated attenuation ratio to Pb. Similarly, the transmitted air kerma measured for the bilayer combinations was divided by the measured transmitted air kerma value for an equivalent thickness of Pb to provide measured attenuation ratios to Pb, for comparison to the calculated ratios. At least 11 different metal/polymer layers have historically been used for radiation protection garments. 8 For the current study, combinations of two metal/elastomer layers were investigated, with the metal/elastomer layers designated as either low-z Z=48 58 or high-z Z= Based on the practical considerations of low cost and nontoxicity, the preliminary short-list included three low-z elements Sn, Sb, and Ba and two high-z elements W and Bi, with all subsequent combinations compared with pure Pb. Of the low-z materials, Ba was readily available only in the form of BaSO 4. This added substantially to the thickness and weight of these layers, so Ba attenuation was calculated but no measurements were made. Sn and Sb are adjacent on the periodic table and have similar responses due to their similar K-absorption edges and emission line energies. Preliminary calculations showed that Sb had slightly better attenuation properties, so Sb was chosen as the sole representative low-z metal/elastomer layer for measurements. However, both Sb and Ba were included as low-z metal/elastomer layers for the calculations. The high-z elements W and Bi were substantially different in their characteristics, so both W and Bi were chosen as representative high-z metal/elastomer layers for both calculations and measurements. Extensive MC calculations were undertaken for three low-z/high-z bilayer systems: Sb/W, Sb/Bi, and Ba/Bi. Ba/W was not modeled, since it became clear early in the study that the emission lines for W would render the Ba/W bilayer less effective than Ba/Bi. The two bilayer systems measured were Sb/W and Sb/Bi. A single equivalent thickness layer of Pb was also modeled and measured for comparison with each combination. Characteristics of the Sb, W, and Bi metal/elastomer sheets and Ba metal are given in Table I. The density of the elastomer imbedding material for the Sb, W, and Bi layers was known 0.87 g/cm 3,sothe equivalent thickness of pure metal incorporated in each piece of sample material could be calculated from the measurement results for comparison with MC calculations based on pure metal layers. The six x-ray qualities recommended in DIN 6857 Ref. 12 were chosen for this study, and are given in Table II. The HVL and mean energies of the six x-ray qualities were calculated using SPEC78 Ref. 18 and the effective energy by a standard textbook calculation. II.A. MC calculations Simulation of the Comet MXR-320 x-ray tube was accomplished by using the EGSnrc Refs. 14 and 15 user code BEAMnrc Ref. 25 as a simulation source for the EGSnrc user code cavity.cpp. 16,17 Rayleigh scattering, bound Compton scattering, atomic relaxations, and electron impact ionization were turned on to ensure accurate simulation of lowenergy photon and electron transport. Cutoff energies for the transport of electrons ECUT=512 kev and photons PCUT=1 kev in BEAMnrc were set to 1 kev in both cases. The ability to simulate several geometries in one run with the user code cavity.cpp was used to provide calculated MC spectra of the six DIN 6857 x-ray qualities after attenuation by the many combinations of metals studied and efficiently compute air kerma ratios of metal/pb using a correlated sampling technique. ECUT and PCUT were set to 521 and 1 kev, respectively. Inclusion of electron transport at this stage was necessary to account for the production of characteristic x-rays in the shielding materials. Since no photons below 10 kev were found at the scoring plane of a full simulation with an ECUT of 512 kev 1 kev electron transport cutoff, a value of 512 kev was deemed accurate enough for these calculations. All MC calculations and measurements were intended to approximate surface and penetrating radia- TABLE II. Draft German Standard DIN x-radiation qualities. X-ray tube voltage kv Filter mm Cu HVL mm Al Mean energy kv Effective energy kv

4 5589 McCaffrey, Mainegra-Hing, and Shen: Non-Pb radiation shielding bilayer materials 5589 metal-embedded elastomer layers. Since the density of the elastomer embedding material was known, the equivalent thickness of pure metal incorporated in each slightly different metal/elastomer layer in the bilayer could be determined, i.e., layer A 1 of metal A was not identical to A 2, which was not identical to A 3, etc. The layers were thin enough to allow five layers to comprise one filter, i.e., the measured layers each comprised approximately 20% of the total bilayer weight. This allowed several measurements of varying ratios of the A and B metal layer components of the bilayer, i.e., A 1 A 2 A 3 A 4 A 5, A 1 A 2 A 3 A 4 B 1, A 1 A 2 A 3 B 1 B 2, A 1 A 2 B 1 B 2 B 3, A 1 B 1 B 2 B 3 B 4, and B 1 B 2 B 3 B 4 B 5. The measured transmitted air kerma provided by each variation in bilayer was divided by the measured transmitted air kerma provided by an equivalent thickness of pure Pb, to provide a ratio of the attenuation of the bilayer normalized to the equivalent attenuation by pure Pb. In this way the measured attenuation ratios could be compared to MC calculations. The MC calculations were made for 10% weight increments for each bilayer. FIG. 2. Experimental setup for attenuation measurements and MC modeling, adapted from DIN a Setup to approximate surface radiation in soft tissue and b penetrating radiation at 10 mm depth in soft tissue. tion. The experimental measurements incorporated a thin 1.2 mm Lucite layer between the metal/elastomer layers and the detector for the surface irradiation approximation, and a 10 mm thickness of Lucite was used for the penetrating radiation approximation. These Lucite layers were also modeled. II.B. Measurements All measurements were performed in the low-energy and medium-energy x-ray primary standards laboratory of the National Research Council in Ottawa, Canada. The x-ray tube utilized was a 320 kv Comet model MXR-320/26 with a 5.5 mm focal spot size, tungsten target, and 20 target angle. Reference qualities were verified with the mediumenergy exposure standard primary standard free-air chamber. The experimental setup for attenuation measurements of test layers, adapted from DIN 6857, is shown in Fig. 2. Figure 2 a is the geometry employed to approximate surface irradiation and Fig. 2 b is the geometry approximating penetrating radiation. Two attenuating materials A and B were mounted against a 1 cm diameter Pb aperture. A 1.2 mm sheet of Lucite was utilized behind layer B to provide support and to flatten the layers. A reference-quality Radcal RC6M plane parallel ion chamber was used as the detector, mounted flush against the 1.2 mm Lucite. A second set of air kerma measurements were made with a 10 mm thick piece of Lucite between the downstream material and the ion chamber Fig. 2 b to approximate penetrating radiation measurements. All measurements were compared on the basis of the pure metal content of commercially manufactured Sb, W, and Bi III. RESULTS III.A. MC calculations Preliminary calculations investigated the most effective order of metal bilayers for maximum attenuation. Consequently, two orientations were calculated; high-z upstream with low-z downstream high-z/low-z, and low-z upstream with high-z downstream low-z/high-z. As an illustration of the effect of the order of the layers in the bilayer, Fig. 3 a displays three transmitted spectra for the 60 kv X-ray quality passing through Sb-upstream/W-downstream low-z/high- Z, W-upstream/Sb-downstream high-z/low-z and Pb layers. The total thickness of the bilayers as well as the single Pb layer was an identical g/cm 2, with each individual layer in the bilayer equal to g/cm 2, i.e., a 50%-50% bilayer by weight. The ratios of the layers transmitted air kerma to the Pb transmitted air kerma were 0.73, 1.31, and 1.00 for Sb/W, W/Sb, and Pb, respectively, with the high value of the W/Sb ratio due to significant emission of radiation from the downstream Sb layer at the characteristic Sb K and Sb K emission energies. The Sb/W low-z/ high-z order provided the most attenuation least transmission, and a shielding material manufactured from this bilayer would allow a substantial reduction in weight, while providing the same protection as a Pb-based material at this x-ray quality. Comparing the above ratios of Sb/W low-z/ high-z to W/Sb high-z/low-z, 0.73/1.31=0.56, i.e., the low-z/high-z combination transmits only 56% of the energy that the opposite order transmits. This example of a bilayer order ratio transmitted energy of low-z/high-z divided by transmitted energy of high-z/low-z is included as the second data point for the Sb W: Penetrating curve, as shown in Fig. 3 b. Figure 3 b displays the calculated bilayer order ratios of all six DIN 6857 x-ray qualities for the two bilayer combinations Sb W and Sb Bi the two bilayers that could be confirmed by measurement in this study. The abscissa val-

5 5590 McCaffrey, Mainegra-Hing, and Shen: Non-Pb radiation shielding bilayer materials 5590 FIG. 4. MC calculations of the penetrating radiation approximation for three bilayers, calculated at the six x-ray qualities. Lower values on the ordinate represent better attenuation lower transmitted air kerma. FIG. 3. The effects of the order of bilayer materials on attenuation. a Attenuated fluence spectra resulting from the two possible orientations for a Sb W bilayer at the 60 kvp x-radiation quality: Sb-upstream/Wdownstream and W-upstream/Sb-downstream, plus the attenuated spectra resulting from an equivalent thickness of Pb. b Bilayer order ratios of the radiation attenuation of Sb W and Sb Bi bilayers for surface and penetrating radiation. Each calculated point represents the transmitted air kerma ratio for bilayers with the low-z material upstream low-z/high-z divided by the same set of bilayers with the high-z material upstream high-z/low-z. ues are defined by the effective energies of the six x-ray qualities as given in Table II. In all of the penetrating cases 10 mm depth in soft tissue equivalent, the low-z/high-z configuration resulted in lower transmitted energy than the reverse. Similarly for the surface cases except for the bilayer containing W at the 150 kvp quality the low-z/high-z order resulted in lower transmitted energy than the reverse. For the bilayer containing W at the 150 kvp quality, the anomalous behavior is a result of the influence of the 69.5 kev K-absorption edge of W. All bilayers for this figure were calculated with 50%:50% equivalent weight for the two metal components. The attenuating capabilities of the bilayers can be further optimized by varying the percentages of low-z to high-z materials. The calculation of these results is shown in Fig. 4 for the 10 mm depth penetrating radiation approximation for the Sb/W, Sb/Bi and Ba/Bi bilayers. The lowest ordinate values for each curve represent the best attenuation lowest transmitted air kerma. This figure illustrates the difficulty in trying to choose a single bilayer combination that provides the best attenuation at all six of the x-ray qualities. The calculated Ba/Bi bilayer response is superior to the other two bilayers at all percent combinations for the four higher energy x-ray qualities 150, 120, 100, and 80 kv, but superior at 60 kv only if the Ba percentage is 65% and at 50 kv if the Ba percentage is 40%. From this data, maximum attenuation occurs with the Ba/Bi bilayer at 150 kv for Ba 30% ; at 120 kv, Ba 70% ; at 100 and 80 kv, Ba 80% ; at 60 kv, Ba 50% ; and at 50 kv, Ba 30%. This illustrates the difficulty in choosing one bilayer composition to provide the most effective attenuation across the entire kv energy range. In attempting to choose one specific bilayer combination to provide the best overall attenuation for all six x-ray qualities, the most effective calculated combination would be Ba 60% /Bi 40%. The total air kerma transmitted through this bilayer measured for each of the six x-ray qualities divided by the total air kerma transmitted through Pb measured for each of the six x-ray qualities would be %. However, this calculation supposes that an elemental form of Ba were available for embedding in an elastomer sheet. For this study we were not able to obtain such layers; only elastomer layers containing BaSO 4 were available, which added unacceptably to the overall material weight. For the other two available bilayers Sb/W and Sb/Bi, the best

6 5591 McCaffrey, Mainegra-Hing, and Shen: Non-Pb radiation shielding bilayer materials 5591 TABLE III. Uncertainties in MC calculations and measurements for penetrating radiation. X-ray tube voltage kvp MC: Average statistical uncertainty % Measurement: Average statistical uncertainty % Transmitted air kerma 0.5 mm Pb % Sb/Bi Sb/W Ba/Bi Sb/Bi Sb/W Ba/Bi Pb combinations would be 85% transmitted air kerma by Sb 40% /W 60% for all six x-ray qualities relative to Pb, and 93% transmitted air kerma by Sb 40% /Bi 60% relative to Pb. III.B. Measurements An equivalent thickness of 0.5 mm Pb provides very strong attenuation of radiation beams in these energy ranges. No measurements were possible for the 50 kvp x-ray quality, since 0.5 mm Pb-equivalent attenuation decreases the detector reading by a factor of approximately By contrast, at 150 kvp, 0.5 mm Pb decreases the detector reading by a factor of approximately 9. The attenuating effect of 0.5 mm Pb on each x-ray quality is listed in Table III in the column labeled Transmitted air kerma 0.5 mm Pb %. Table III also lists the statistical uncertainties at 10 mm depth penetrating radiation for the MC calculations and the measurements. The listed values are the averages for the various combinations of that bilayer material. The uncertainty varied with each x-ray quality and each bilayer combination. For measurements, the major source of uncertainty was the density and thickness uniformity of the sheets of bilayer material. The standard deviation for each sheet thickness was approximately 7% for the Sb/W bilayers and 6% for the Sb/Bi bilayers, and typically five bilayer sheets were included in each measurement. Figure 5 compares measurements to MC calculations at 10 mm depth penetrating radiation for the Fig. 5 a Sb/Bi FIG. 5. Transmitted air kerma relative to 0.5 mm Pb for three bilayer materials at six x-ray qualities. The curves represent MC calculations and the symbols represent measurements, all ratioed to 0.5 mm Pb. Data points below the dark horizontal line at 1 on the ordinate scale indicate that equivalent shielding but lower weight can be provided by a bilayer.

7 5592 McCaffrey, Mainegra-Hing, and Shen: Non-Pb radiation shielding bilayer materials 5592 FIG. 6. Approximate weights of a Sb 70% /W 30% bilayer material that would provide equivalent attenuation to 0.5 mm Pb at each of the six x-ray qualities. and Fig. 5 b Sb/W bilayer combinations. Only MC results are included in Fig. 5 c the Ba/Bi combination, to show the possibilities if an elemental source of Ba rather than BaSO 4 were available. All measurements and MC calculations were referenced to transmitted air kerma through 0.5 mm of Pb. Experimental measurements could not be made at the 50 kvp x-ray quality because of the almost complete attenuation provided by the 0.5 mm Pb-equivalent shielding discussed further below. All points on the graphs with an ordinate value less than unity below the dark horizontal lines in Figs. 5 a 5 c represent a bilayer shielding material with less weight than a Pb shielding material, but with an equivalent level of protection. As can be seen, almost all bilayer combinations provide equivalent protection to Pb but with less weight. Bilayers are particularly useful in the kvp range, but the optimum percent of the low-z component given by the abscissa varies for each x-ray quality. Little or no advantage is demonstrated for bilayers at the 150 kvp quality, as close to 100% of the high-z component is required to match or modestly improve upon Pb. While bilayers can be optimized for particular energy ranges, there is no single combination that is best for all of the x-ray qualities measured and modeled here. In Fig. 5, experimental measurements could not be made at the 50 kvp x-ray quality because of the very high attenuation very low transmitted air kerma, see Table III. The signal was not discernible above the leakage rate 99% with a high-quality electrometer. Leakage rates for the other kvp x-ray qualities were 10% 20% at 60 kvp, less than 3% for 80 kvp, and negligible for the 100, 120, and 150 kvp x-ray qualities. Considering the uncertainties for the measurements, the agreement between the measurements and MC calculations is good for both the Sb/Bi and Sb/W bilayers at the 80 kvp and higher energy qualities, and adequate for the 60 kvp quality where the very low signal level introduces excessive uncertainty. Figures 5 a and 5 b demonstrate that when the Sb component of the bilayer comprises approximately 60% of either of these measured bilayer thicknesses, a substantial weight reduction compared to Pb can be achieved for equivalent protection, for all x-ray qualities except the 150 kv quality. For example, a radiation shielding garment made of Sb/Bi bilayers for use during cardiac catheterization procedures where the kvp would typically not exceed 80 kv would weigh approximately 25% less than a Pb-based apron for equivalent protection. The same 80 kv limit would apply for dentists offices. For CT scans, a bilayer of Sb/W optimized for 120 kvp would be more appropriate. A major finding of this study is that radiation shielding garments can be more effective and weigh less than Pb if they are tailored to match the relevant x-ray energy spectrum for each application, rather than attempting to cover a broad range of energies which may not be relevant to that particular application. As a practical example at the 100 kvp quality, a Sb 70% / W 30% material can provide the same attenuation of a Pbbased material but with only 75% of the weight, depending on the x-ray quality being measured. Figure 6 displays approximate weights of a Sb 70% /W 30% bilayer material that would provide equivalent attenuation to 0.5 mm of Pb at each of the six x-ray qualities. Data points below the dark horizontal line represent situations where the bilayer material weighs less than 0.5 mm Pb for equivalent attenuation. At the 50 kvp quality E eff =11.9 kev, the effect is inconsequential, but at the 100 kvp quality E eff =47.3 kev, the bilayer would weigh only about 75% the weight of an equivalent area of 0.5 mm Pb to provide the same attenuation, for both surface and penetrating radiation. At 150 kvp E eff =72.1 kev, the bilayer would necessarily weigh more than the same area of 0.5 mm Pb to provide the same attenuation for penetrating radiation, but would weigh only 88% of the same area of 0.5 mm Pb for equivalent surface radiation attenuation. Figure 7 compares MC calculations and normalized measurements for both surface irradiation MC represented by dashed lines, measurements by X s, and penetrating radiation MC represented by solid lines, measurements by O s. No measurements were available for 50 kv as discussed above. The measurement normalization was based on values where close agreement between MC and measurements was demonstrated in Fig. 5. The transmitted air kerma is consistently 20% 30% higher for surface radiation than for penetrating radiation for both MC calculations and measurements. This is partly attributable to the effects of fluorescence radiation, which would be prevalent close to the inner surface of a shielding material garment. This has relevance for radiation protection since one would expect these materials to be in close proximity to the surface of the skin. IV. CONCLUSION For radiation shielding materials, an appropriate choice of two complementary metal/elastomer layers bilayers can, in almost all cases, provide equivalent attenuation to pure Pb

8 5593 McCaffrey, Mainegra-Hing, and Shen: Non-Pb radiation shielding bilayer materials 5593 but with less weight, weight reductions of 25% being readily achievable. Based on practical considerations, this study investigated three low-z/high-z bilayer systems Sb/W, Sb/Bi and Ba/Bi and utilized the six x-ray qualities in the energy range kev from the German Standard DIN Measurements and MC calculations were compared, approximating both surface irradiation to soft tissue and penetrating radiation at 10 mm depth in soft tissue. The low-z upstream/high-z downstream ordering of the metal bilayers provides up to five times more attenuation than the reverse order at the 50 kvp x-ray quality, but this difference diminishes gradually and eventually disappears at the 150 kvp quality. Further optimization of the attenuating capabilities of the bilayers is accomplished by varying the percentages of low-z to high-z materials. The Ba/Bi bilayer system would provide the best attenuation if an elemental form of Ba rather than BaSO 4 was available. The more readily available Sb/Bi and Sb/W bilayers offer substantial reductions in weight over Pb-based materials for equivalent attenuation, but the reductions vary with x-ray quality. Bilayer radiation shielding materials can reduce the weight to approximately 75% of Pb-based materials while providing the same protection if they are tailored to match an application at a particular x-ray energy. If the attempt is made to cover the full range of energies from 50 to 150 kvp, there is no optimal combination that provides equally effective attenuation in all cases. The attenuation response at the 150 kvp quality in particular eliminates any advantage of bilayers. The transmitted air kerma through the bilayers was consistently 20% 30% higher for surface radiation than for penetrating radiation for both MC calculations and measurements. This is partly due to the effects of the fluorescence radiation from the metal components of the bilayers and Pb layer. ACKNOWLEDGMENTS The authors wish to express their appreciation to Martin Lilley at Xenoprene Co. for generously providing the sample materials. The authors also wish to thank Carl Ross for helpful suggestions. FIG. 7. Transmitted air kerma for penetrating radiation and surface radiation at the six x-ray qualities. The solid lines and the circles represent MC calculations and normalized measurements of penetrating radiation. The dashed lines and the crosses represent MC calculations and normalized measurements for surface radiation. a Electronic mail: john.mccaffrey@nrc-cnrc.gc.ca 1 E. W. Webster, Experiments with medium Z materials for shielding against low-energy x-rays, Radiology 86, M. J. Yaffe, G. E. Mawdsley, L. Martin, R. Servant, and R. George, Composite materials for x-ray protection, Health Phys. 60, E. W. Webster, Addendum to Composite materials for x-ray protection, Health Phys. 61, P. H. Murphy, Y. Wu, and S. A. Glaze, Attenuation properties of lead composite aprons, Radiology 186, P. J. Kicken and A. J. J. Bos, Effectiveness of lead aprons in vascular radiology: Results of clinical measurements, Radiology 197, E. G. Christodoulou, M. M. Soodsitt, S. C. Larson, K. L. Darner, J. Satti, and H.-P. Chan, Evaluation of the transmitted exposure through lead equivalent aprons used in a radiology department, including contribution from backscatter, Med. Phys. 30, Y. Takano, K. Okazaki, K. Ono, and M. Kai, Experimental and theoretical studies on radiation protective effect of a lighter non-lead protective apron, Jpn. J. Radiol. Technol. 61, J. P. McCaffrey, H. Shen, B. Downton, and E. Mainegra-Hing, Radiation attenuation by lead and nonlead materials used in radiation shielding garments, Med. Phys. 34, H. Schlattl, M. Zankl, H. Eder, and C. Hoeschen, Shielding properties of lead-free protective clothing and their impact on radiation doses, Med. Phys. 34, H. Warren-Forward, P. Cardew, B. Smith, L. Clack, and K. McWhirter, A comparison of dose savings of lead and lightweight aprons for shielding of 99 m-technetium radiation, Radiat. Prot. Dosim , K. Yue, W. Luo, X. Dong, C. Wang, G. Wu, M. Jiang, and Y. Zha, A new lead-free radiation shielding material for radiotherapy, Radiat. Prot. Dosim. 133, DIN : : Radiation protection accessories for medical use of X-radiation Part 1: Determination of shielding properties of unleaded or lead reduced protective clothing, DIN Deutsches Institut für Normung e. V F. H. Attix, Introduction to Radiological Physics and Radiation Dosimetry Wiley, Toronto, I. Kawrakow, Accurate condensed history Monte Carlo simulation of electron transport. I. EGSnrc, the new EGS4 version, Med. Phys. 27, I. Kawrakow and D. W. O. Rogers, The EGSnrc Code System: Monte Carlo simulation of electron and photon transport, Technical Report PIRS-701, 4th printing National Research Council of Canada, Ottawa, I. Kawrakow, EGSnrc C class library, Technical Report PIRS 898 National Research Council of Canada, Ottawa, E. Mainegra-Hing and I. Kawrakow, Efficient x-ray tube simulations, Med. Phys. 33 8, Institute of Physics and Engineering in Medicine, Catalogue of diagnostic x-ray spectra and other data, Report No. 78 IPEM Publications, York, B. J. McParland, J. Nosil, and B. Burry, A survey of the radiation exposures received by the staff at two cardiac catheterization laboratories, Br. J. Radiol. 63, L. Renaud, A 5-y follow-up of the radiation exposure to in-room personnel during cardiac catheterization, Health Phys. 62 1, M. Zuguchi, K. Chida, M. Raura, Y. Inaba, A. Ebata, and S. Yamada, Usefulness of non-lead aprons for radiation protection of physicians per-

9 5594 McCaffrey, Mainegra-Hing, and Shen: Non-Pb radiation shielding bilayer materials 5594 forming interventional procedures, Radiat. Prot. Dosim , S. Kim, T. T. Yoshizumi, D. P. Frush, C. Anderson-Evans, and G. Tonehava, Dosimetric characterization of bismuth shields in CT: Measurements and Monte Carlo simulations, Radiat. Prot. Dosim. 133, K. P. Kim, D. L. Miller, S. Balter, R. A. Kleinerman, M. S. Linet, D. Kwon, and S. L. Simon, Occupational radiation doses to operators performing cardiac catheterization procedures, Health Phys. 94 3, K. P. Kim and D. L. Miller, Minimizing radiation exposure to physicians performing fluoroscopically guided cardiac catheterization procedures: A review, Radiat. Prot. Dosimetry 133 4, D. W. O. Rogers, B. Walters, and I. Kawrakow, BEAMnrc users manual, Technical Report PIRS 509 rev a I National Research Council of Canada, Ottawa, 2005.

10 Copyright of Medical Physics is the property of American Association of Physicists in Medicine and its content may not be copied or ed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or articles for individual use.