Application Note No. 301/2017

Transcription

1 Application Note No. 301/ MCPD and glycidol in infant formula SpeedExtractor E-916, Multivapor P-6: Extraction of 3-MCPD and glycidyl-esters from infant formula Quality in your hands

2 1. Introduction Esters of 2- and 3-MCPD (monochloropropanediol) are formed during the refining process of vegetable oils. The formation of 2- and 3-MCPD from glycerol in presence of chloride ions takes place at high temperatures in the deodorization step followed by esterification with free fatty acids. Therefore, any food that contains vegetable oils can potentially contain 2- or 3-MCPD esters. The content depend on the type of oil used. The widely used palm oil has the highest content of 3- MCPD [1, 2]. Free 3-MCPD is also formed during heating processes, e.g. during grilling of meat and during the hydrolysis of vegetable proteins, used for instance in soy sauce. Figure 1: 3-MCPD and 3-MCPD esters [4]. The only identified source of glycidyl esters in food is refined vegetable oil where they appear to be formed during the heating of diacylglycerol or monoacylglycerol under the high temperature conditions of deodorization [3]. Figure 2: Glycidol and glycidyl esters (R1: fatty acid). 3-MCPD has been classified as potentially carcinogenic to humans, whereas glycidol is carcinogenic and genotoxic [4]. The esters themselves are not toxic, but they are released in human digestion and therefore the total 3-MCPD (free and bound) are taken into consideration to estimate the safety of the food [3]. A tolerable daily intake (TDI) for 3-MCPD of 2 μg/kg body weight was set in 2001 by the Joint FAO/WHO Expert Committee on Food Additives JECFA. This value is also supported by the German Federal Institute for Risk Assessment (BfR). The EFSA (European Food Safety Authority) defined a lower TDI of only 0.8 μg/kg body weight. For glycidol and glycidyl esters there is no tolerable daily intake value available, but the presence should be avoided according to the ALARA principle (As Low As Reasonable Achievable). Among all foods, the infant formula is the most critical one because infant formula is the only food that is consumed by infants that are not breastfed, and the body weight is very low. The German Federal Institute for Risk Assessment (BfR) communicated about the risk for infants in 2007 [4]. The extraction of solid and semisolid food samples is usually done by pressurized solvent extraction. The extraction using methods with acidic or alkaline hydrolysis such as Weibull-Stoldt or Mojonnier should be avoided because significant amounts of 3-MCPD esters are formed during the procedure [5]. The SpeedExtractor E-914 / E-916 has been successfully used for the extraction of free 3-MCPD in grilled meat and meat products [6, 7]. Application Note 301/ June /12

3 There is no official standard method for extraction prior to the analysis of 3-MCPD and glycidol available. The EFSA published a method [8] for the determination of 3-MCPD in various food matrices using pressurized solvent extraction. The BfR Method 22 also includes pressurized solvent extraction. This method has been validated in a ring test with different food matrices [9]. The BfR method 22 had been adapted for infant formula and was published by Wöhrlin et al. [10], the authors are employees of the BfR. The main difference is this method uses another solvent mixture. The present application note follows the method published by Wöhrlin et al. [10]. 2. Equipment SpeedExtractor E-916, equipped with 40 ml extraction cells Multivapor TM P-6 Recirculating Chiller F-308 Analytical balance (accuracy +/- 0.1 mg) Eppendorf pipette Beaker 3. Chemicals and materials Chemicals: Acetone, for HPLC, 99.8 %; Sigma Aldrich ( L) Isohexane, Suprasolv for gas chromatography; Merck ( ) Diatomaceous earth (DE), BUCHI (053201) Cellulose top filter, BUCHI (049572) Cellulose bottom filter, BUCHI (049569) Ethanol, for residue analysis; Sigma Aldrich (02851) tert-butyl methyl ether (MTBE), for HPLC, 99.8 %; Sigma Aldrich (34875) Water, distilled Standards: rac 1.2-Bis-palmitoyl-3-chloropropanediol, Toronto Research Chemicals (B mg) Glycidyl stearate, Toronto Research Chemicals (G mg) Surrogate standards: See DGF method C-VI (10) Standard solutions: Solution 1: 20.1 mg rac 1.2-Bis-palmitoyl-3-chloropropanediol was diluted in 100 ml ethanol (0.201 mg/ml); 1 ml was diluted to 10 ml in MTBE ( mg/ml) % of rac 1.2-Bis-palmitoyl-3-chloropropanediol is equal to 3-MCPD, therefore standard solution 1 contains mg/ml 3-MCPD. Solution 2: 22.6 mg glycidyl stearate was diluted in 100 ml ethanol (0.226 mg/ml); 1 ml was diluted to 10 ml in MTBE ( mg/ml) % of glycidyl stearate is equal to glycidol, therefore standard solution 2 contains mg/ml glycidol. For safe handling please pay attention to all corresponding MSDS. s: Infant formula A, PRE, from birth, expected fat content: % Infant formula B, stage 2, after 6 month; expected fat content: % The samples were purchased at a local supermarket. Due to the fact that there are no certified reference materials available for the determination of 3-MCPD and glycidol in infant formula, the samples were spiked in this study. Application Note 301/ June /12

4 4. Procedure The determination of 3-MCPD and glycidol in infant formula includes the following steps: Extraction of infant formula using the SpeedExtractor E-916 Evaporation of the extract to dryness for fat determination using the Multivapor TM P-6 Determination of 3-MCPD and glycidol in the extracted fat using DGF method C-VI 18 (10) [5]. 4.1 Extraction of infant formula using the SpeedExtractor E Weigh 2 g of the sample and 4.6 g DE into a beaker 2. Add 0.8 ml water and mix intensively using a spatula 3. Load the samples into the cells using a funnel 4. For spiking, add standard solutions to the samples using an Eppendorf pipette: ml of standard solution 1 (corresponds to mg 3-MCPD) ml of standard solution 2 (corresponds to mg Glycidol) 5. Dry 240 ml collection bottles in the drying oven at 102 C for 30 min 6. Cool down the collection bottles to ambient temperature in a desiccator for 60 min 7. Weigh and record the exact bottle weights, accuracy 0.1 mg 8. Place the collection bottles in the receiving unit 9. Carry out the extraction using the parameters in Table 1 Table 1: Extraction method using the SpeedExtractor E-916. Parameter Value Temperature 100 C Pressure Solvent Cell Vial Cycles 3 Heat up Hold Discharge Flush with solvent Flush with gas Total extraction time 100 bar 80 % Acetone 20 % Isohexane 1 40 ml 240 ml 1 min / 1 min / 1 min 10 min / 10 min / 10 min 3 min / 3 min / 3 min 2 min 3 min 1 h 10 min 4.2 Evaporation of extract to dryness for fat determination using the Multivapor TM P Evaporate the solvent according to Tables 2 and 3 Table 2: Parameters for the Multivapor TM P-6. Parameter Value Temperature 55 C Rotation level 8 Temperature Chiller 5 C 1 The solvent mixture is different to the solvent mixture used by Wöhrlin et al. [10]. According to the authors, there is a typing mistake in the publication and the solvent mixture has to be 80% Acetone / 20 % Isohexane, not vice versa. Application Note 301/ June /12

5 Table 3: Evaporation parameters for the Multivapor TM P-6 Step Pressure start [mbar] Pressure end [mbar] Time [min] 11. Dry the collection bottles in a drying oven for 30 min at 102 C 12. Cool down the bottles in a desiccator for 60 min 13. Weigh the bottles containing the extract and record the weight. 14. Calculate the fat content using equation (1), see section Determination of 3-MCPD and glycidol in the extracted fat The determination was done according to the DGF C-VI 18 (10) standard method [5]. This method comprises the determination of free and bound 3-MCPD after alkaline catalyzed ester cleavage and derivatization with phenylboronic acid (PBA). It also allows the indirect determination of glycidyl esters under the assumption that no other substances are present that react at room temperature with inorganic chloride to generate 3-MCPD. The quantification of the derivatized compounds is done by GC-MS, using deuterated surrogate standards. The method has the following principle: Assay A: Determination of the sum of bound 3-MCPD and bound glycidol. Assay B: Determination of the bound 3-MCPD. The total amount of bound glycidol is then calculated by the difference between these two assays. Proceed as described below. For more detailed information, please refer to the DGF method [5]. 15. Take an aliquot of 100 mg of extracted fat for each assay 16. Weigh this aliquot into a 1.5 ml screw cap vial 17. Add 100 µl of the surrogate standard and 100 µl of MTBE to each assay and shake it until the sample is completely dissolved 18. Add 200 ml of sodium hydroxide to each assay and shake it for a short time. 19. The reaction time for the ester cleavage is between 3.5 to 5.5 min 20. Stop the ester cleavage by adding 600 µl of acidified sodium chloride solution to assay A 600 µl acidified chloride solution to assay B 21. Add to both assays 600 µl of isohexane, shake them and wait for 5 min 22. Shake both assays and separate and discard the organic phase, repeat step 20 and 21 once 23. Extract the aqueous phases of both assays three times with 600 µl of a mixture of diethyl ether (ethyl acetate). Combine the organic phases in a new vial containing sodium sulphate 24. Add the derivatization reagent ( ml, depending on the capability of the chromatographic system) 25. Evaporate both assays to dryness using a gentle nitrogen stream 26. Redissolve the dried extracts with 500 µl of iso-octane The GC-MS conditions are shown in Table 4. Application Note 301/ June /12

6 Table 4: GC-MS conditions for the quantification of derivatized 3-MCPD and glycidol. Type of GC/MS Mode Thermo TSQ 8800 Evo SRM Injector PTV, splitless 220 C Column Carrier Gas Oven Program MEGA-5 MS, 30m x 0.25mm ID, 0.25µm film Helium 2.0 ml/min 85 C (hold 0.5 min) 6 C/min 150 C (hold 0.0 min) 12 C/min 180 C (hold 0.0 min) 25 C/min 280 C (hold 6.0 min) The determination was carried out by Labor Veritas in Zurich, a service laboratory accredited according to ISO/IEC 17025: Calculation Fat content The fat content of the infant powder samples are calculated as percentage of fat according to equation (1). (m %Fat m - m ) 100 total vial (1) %Fat mtotal mvial m : percentage of fat in the sample : vial weight + extract : vial weight : sample weight Content of 3-MCPD and glycidol Calculation of the spike amount: Solution 1: 20.1 mg rac 1.2-Bis-palmitoyl-3-chloropropanediol was diluted in 100 ml ethanol (0.201 mg/ml); 1 ml was diluted to 10 ml in MTBE ( mg/ml) % of rac 1.2-Bispalmitoyl-3-chloropropanediol is equal to 3-MCPD, therefore standard solution 1 contains mg/ml 3-MCPD ml of standard solution 1 was added. Therefore mspike = mg 3-MCPD. Solution 2: 22.6 mg Glycidyl stearate was diluted in 100 ml ethanol (0.226 mg/ml); 1 ml was diluted to 10 ml in MTBE ( mg/ml) % of glycidyl stearate is equal to glycidol, therefore standard solution 2 contains mg/ml glycidol ml of standard solution 2 was added. Therefore mspike = mg glycidol. Calculation of recovery for spiked samples: m fat, cal msample %Fat (2) 100 Theoretically added amount of spike solution: mspike m spike, cal 1000 (3) mfat,cal manalyte, total = mspike, cal + manalyte (4) Application Note 301/ June /12

7 Recovery is calculated as a relation of measured analyte to total analyte (spike amount + measured analyte): m spike Recovery analyte, 100 (5) m analyte,total mfat, cal : calculated fat amount msample : used sample amount %Fat : measured fat content [%] mspike, cal : relation of spike to fat mspike : amount of added spike [mg] manalyte, total : total amount of analyte: 3-MCPD resp. glycidol manalyte : measured amount of 3-MCPD resp. glycidol in non spiked sample Recovery : relation of measured analyte to calculated analyte [%] manalyte, spike : measured amount of 3-MCPD resp. glycidol in spiked sample 5. Result 5.1 Determination of the fat content The fat content of both infant formulas was determined, the results are presented in Tables 5 and 6. Table 5: Fat determination; sample: Infant formula A; expected fat content: 25.2 %. m mvial mtotal %Fat A A A Average [%] rsd [%] 2.64 Recovery [%] Table 6: Fat determination; sample: Infant formula B; expected fat content: %. m mvial mtotal %Fat B B B Average [%] rsd [%] 4.29 Recovery [%] The determined fat content of sample A shows a good recovery compared to the declared fat content. For sample B the determined fat content is lower as expected. Wöhrlin et al. [10] has shown that determined fat contents in infant formula using pressurized solvent extraction are typically in a recovery range of %. In this study only infant formula of stage PRE and 1 were analyzed. Due to a different composition of the infant formulas of different stages, the recoveries for the fat determination vary. The extraction method used in this application note is not optimized for the fat determination but for the determination of 3-MCPD and glycidol. The determination of the fat content is necessary for the calculation and interpretation of the spiked samples. Therefore, the fat determination of sample B is acceptable. Application Note 301/ June /12

8 5.2 Measurement of the 3-MCPD and glycidol content Non spiked samples The determined 3- MCPD and glycidol content of both infant formulas are shown in Tables 7 and 8. These are the values of the native samples, without addition of the spiking solutions. Table 7: Infant formula A; measured amount of 3-MCPD and glycidol in fat, not spiked sample. m m3-mcpd A A A * Average [%] rsd [%] *outlier, not taken into consideration for average calculation mglycidol Table 8: Infant formula B; measured amount of 3-MCPD and glycidol in fat, not spiked sample. m m3-mcpd B B B Average [%] rsd [%] mglycidol In 2007 the BfR communicated the risk of 3-MCPD and glycidol for infants [4]. Since then, the producers have reduced the contents of 3-MCPD and glycidol in infant formula. Wöhrlin et al. [10] has shown in recent studies that in infant formula the 3-MCPD content is mg/kg extracted fat resp. the glycidol content is mg/kg extracted fat [10]. The 3- MCPD and glycidol content determined in this application note are comparable to these results Spiked samples In Tables 9 and 10 the results for the spiked samples are shown. The addition of spike solution is equal to 1.5 mg/kg -2 mg/kg fat. Therefore, the results of the analysis of the spiked samples show a good reproducibility compared to the non spiked samples. Table 9: Infant formula A; measured amount of 3-MCPD and glycidol in fat, spiked sample. m m3-mcpd, spike A A A Average [%] rsd [%] mglycidol, spike Table 10: Infant formula B; measured amount of 3-MCPD and glycidol in fat, spiked sample. m m3-mcpd, spike B B B Average [%] mglycidol, spike rsd [%] Application Note 301/ June /12

9 5.3 Calculation of the recovery of spiked samples The recoveries of the spiked amount of 3-MCPD and glycidol have been calculated, the results are shown in Tables 11, 12, 13 and 14. Table 11: Infant formula A; measured fat content: %; recovery of 3-MCPD spiked. m mfat, cal mspike, cal manalyte, total m3-mcpd, spike Recovery [%] A A A Average [%] rsd [%] 8.45 Table 12: Infant formula B; measured fat content: %; recovery of 3-MCPD spiked. m mfat, cal mspike, cal manalyte, total m3-mcpd,spike Recovery [%] B B B Average [%] rsd [%] 5.04 Table 13: Infant formula A; measured fat content: %; recovery of glycidol spiked. m mfat, cal mspike, cal manalyte, total mglycidol, spike Recovery [%] A A A Average [%] rsd [%] Table 14: Infant formula B; measured fat content: %; recovery of glycidol spiked. m mfat, cal mspike, cal manalyte, total mglycidol, spike Recovery [%] B B B Average [%] rsd [%] Wöhrlin et al. have shown recoveries in a range of % for the 3-MCPD resp. for the glycidol determination [10]. The results for infant formula A correspond to this range. The recoveries for infant formula B are higher compared to this publication. This is most probably related to the lower recovery of the fat extraction (see results 5.1, Table 6). Assuming that the spiked stearates are extracted more easily than the native fat of the sample, this results in an over-estimation of the 3- MCPD and the glycidol per gram of fat. A solution could be to correct the recovery by this factor if the fat recovery is too low. If the determined amount of 3-MCPD and glycidol is corrected by the fat recovery, the recovery is around % and thus comparable to the results of Wöhrlin et al. [10], see Table 15. For infant formula A, no correction is necessary, since the fat determination was at a 100 % recovery (Table 5). No correction is needed for native samples, since the 3-MCPD is bound to the native fat matrix. Application Note 301/ June /12

10 Table 15: Infant formula B; measured fat content: % (fat recovery 88.7 %); corrected recovery of spiked 3-MCPD and glycidol. m3-mcpd, spike m3-mcpd, spike corrected Recovery [%] mglycidol, spike mglycidol, spike corrected Recovery [%] B B B Average [%] rsd [%] Comparison to standard methods Alternative pressurized solvent extraction method EFSA method The EFSA has published an alternative pressurized solvent extraction method [8]. The parameters are shown in Table 16 and compared to the parameters used in this application note taken from the BfR [10]. Table 16: Comparison of the parameters used in this Application Note (BfR method) and EFSA method for extraction with the SpeedExtractor E-916. Parameter Value Application Note Value EFSA method amount 2 g 5 g preparation mixed with 4.6 g DE, addition of 0.8 ml water, mixed in beaker Temperature 100 C 40 C Pressure 100 bar Ambient Solvent 80 % Acetone 20 % Isohexane 2 Cell 40 ml 40 ml Vial 240 ml 240 ml Cycles 3 2 Heat up Hold Discharge 1 min / 1 min / 1 min 10 min / 10 min / 10 min 3 min / 3 min / 3 min Flush with solvent 2 min 2 min Flush with gas 3 min 3 min mixed with 5 g polyacrylic acid and 15 g quartz sand. Mixed in mortar 100 % tert-butyl methyl ether 1 min / 1 min 5 min / 5 min 3 min / 3 min Total extraction time 1 h 10 min 45 min 7. Conclusion The sample infant formula A can be used directly after birth. It is recommended to feed the child in this age with 100 ml liquid infant formula six times per day. For 100 ml liquid infant formula 13.5 g powder is required resp. 81 g per day. 81 g of powder contain g of fat (fat content measured: 25%). The measurements have shown a 3-MCPD content of 0.52 mg/kg fat resp. glycidol content of mg/kg fat in infant formula A. Which means per day mg 3-MCPD and mg glycidol will be consumed by the child. Assuming a body weight of 3.5 kg, the 2 See foot note 1 on page 4. Application Note 301/ June /12

11 daily intake would be 3.01 µg/kg of 3-MCPD which is by factor 3.8 higher than the tolerable daily intake (TDI) of 0.8 µg/kg body weight (TDI published by EFSA) [3]. Glycidol should be avoided but the daily intake is 0.61 µg/kg body weight. Infant formula B is recommended for children older than 6 months. At this age children drink about 600 ml of infant formula per day as an average value. For the preparation of 600 ml of infant milk it is necessary to use 78 g powder of infant formula B. 78 g powder contain g of fat (fat content measured: %). The measurements have shown a 3-MCPD content of 0.36 mg/kg fat resp. a glycidol content of 0.32 mg/kg fat in infant formula B. Which means per day mg 3-MCPD and mg glycidol will be consumed by the child. Assuming a body weight of 7.5 kg, the daily intake would be 0.71 µg/kg of 3-MCPD which is below the TDI of 0.8 µg/kg body weight. For glycidol the daily intake is 0.63 µg/kg. The presented procedure for the extraction of infant formula using pressurized solvent extraction with the SpeedExtractor E-916 is a fast and reliable method for the determination of 3-MCPD and glycidol. 8. Acknowledgement We sincerely thank Mr. Pascal Stingl from Labor Veritas for the analysis of the samples and the support in developing this Application Note. 9. References [1] Zelinkova Z, Svejkovska B, Velisek J, Dolezal M: Fatty acid esters of 3-chloropropane-1,2-diol in edible oils. Food Addit Contam 23, (2006). [2] Wöhrlin, F.; Andres, S MCPD and glycidyl-fatty acid esters in food. Presentation slides (in German) Fortbildung für den Öffentlichen Gesundheitsdienst ( [3] EFSA Panel on Contaminants in the Food Chain (CONTAM). Risks for human health related to the presence of 3- and 2-monochloropropanediol (MCPD), and their fatty acid esters, and glycidyl fatty acid esters in food, EFSA Journal 2016;14(5):4426. [4] BfR Opinion No. 047/2007, 11 December 2007; Infant formula and follow-up formula may contain harmful 3-MCPD fatty acid esters. [5] DGF Standard Method C-VI 18 (10) Fatty-acid-bound 3-chloropropane-1,2-diol (3-MCPD) and 2,3-epoxi-propane-1-ol (glycidol) [6] Schallschmidt, K. et al., Occurrence of free 3-MCPD in grilled meat and meat products. Mitteilungsblatt Fleischforschung Kulmbach 2011; 50 (193), [7] Schallschmidt, K. et al Determination of 3-MCPD in grilled meat using pressurized liquid extraction and gas chromatography high resolution mass spectrometry J. Verbr. Lebensm. (2012) 7: 203. [8] Wenzl, T. et al. Development and validation of analytical methods for the analysis of 3-MCPD (both in free and ester form) and glycidyl esters in various food matrices and performance of an ad-hoc survey on specific food groups in support to a scientific opinion on comprehensive risk assessment on the presence of 3-MCPD and glycidyl esters in food. Joint Research Centre, Institute for Reference Materials and Measurements; External Scientific Report. EFSA supporting publication 2015: EN-779. Application Note 301/ June /12

12 [9] Fry, H. et al. Collaborative Study for the Determination of 3-MCPD- and 2-MCPD Fatty Acid Esters in Fat Containing Foods, Federal Institute for Risk Assessment (BfR), [10] Wöhrlin, F. et al. Occurrence of fatty esters of 3-MCPD, 2-MCPD and glycidol in infant formula., Food Additives & Contaminants: Part A, 2015, Vol. 32, No. 11, Operation Manual for the SpeedExtractor E-916 Operation Manual for the Multivapor TM P-6 Application Note 301/ June /12