DEVELOPMENT OF METHODS TO QUANTITATIVELY EXTRACT BIOLOGICALLY ACTIVE PRINCIPLES FROM COMPLEX FOODS, FLAVOURINGS AND HERBS AND SPICES, TO ALLOW THEIR SUBSEQUENT ANALYSIS FINAL REPORT Report Number FD 10/01 Authors M Scotter and G Rees Date March 2010 Sponsor Food Standards Agency Aviation House 125 Kingsway London WC2B 6NH Sponsor's Project No. A01067 Fera Project No. R6NL Fera File No. FLN 8882 Principal Workers M Scotter, D Roberts and G Rees Contract Manager M Scotter Distribution: 1. Dr W Dixon (FSA) (2 x printed plus electronic) 2. Fera Information Centre DEFRA Food and Environment Research Agency York YO41 1LZ (UK) Telephone: 01904 462000 Fax: 01904 462111 Page 1 of 79 Executive summary 1. This report details the findings of Food Standards Agency Project A01067: Development of methods to quantitatively extract biologically active principles (BAPs) from complex foods, flavourings and herbs and spices, to allow their subsequent analysis. These flavouring compounds cover a range of chemical types and this makes their extraction and determination in foods and beverages analytically challenging. 2. The analytical strategy adopted therefore was to develop three separate methods for each chemical / physical class of BAP: I. Volatile BAPs by simultaneous distillation-extraction (SDE) of the sample followed by GC-MS quantitation. II. Coumarin and quassine by solvent extraction of the sample, clean-up using solid phase extraction cartridges and then quantitation by HPLC with UV detection. III. Hydrogen cyanide by extraction of the sample using acidic medium, enzymatic hydrolysis of glycosidic cyanogens to cyanohydrins, hydrolysis of these to cyanide, and lastly derivatisation using a modified König reaction to form a coloured complex which is determined spectrophotometrically. 3. The methods were validated in single-laboratory tests of a range of different food and beverages. The samples for this validation were selected according to the types of food and beverage likely to contain the BAPs via the addition of herbs and spices and/or flavouring preparations and the maximum levels prescribed in EU Page 2 of 79 regulations. Eight different matrices (breath fresheners, mint confectionery, yoghurt, canned soup, soft drink, vegetable product, chewing gum and fish product) were validated for five volatile BAPs (menthofuran, estragole, pulegone safrole and methyl eugenol) plus isosafrole. Ten matrices were validated for coumarin and quassine (bakery ware, breakfast cereal, rice pudding, gelatine confectionery, biscuit, sugar confectionery, carbonated soft drink, fruit-flavoured drink, herbal infusion and mixed spice). Three matrices were validated for cyanogenic potential (canned stone fruit, marzipan and alcoholic beverage). 4. The SDE and GC-MS method was subjected to cross validation by a second (independent) laboratory. Samples of breath freshener, chewing gum, fish product, cordial, herbal infusion and mint confectionery were analysed for all five volatile BAPs. The results were mixed. The method is technically challenging and method transfer to the second laboratory was not successful despite a high degree of communication and assistance. Notwithstanding the transferability issue, it is concluded that the use of SDE works well for the extraction and purification / concentration of volatile BAPs from a wide range of food matrices. Sample preparation is the key in providing an adequately dissolved or dispersed matrix that lends itself to SDE analysis. The GC-MS conditions are reasonably straightforward but the GC column needs to display good separation characteristics in order to separate BAPs from other co-extractives. 5. The HPLC method was also subjected to cross validation by the second laboratory. Soft drink, rice pudding, bakery ware, biscuit, breakfast cereal, curry paste and herbal infusion were analysed for coumarin and quassine, and canned Page 3 of 79 stone fruit, marzipan and alcoholic beverage analysed for cyanogenic potential. There was acceptable agreement in the results for coumarin for all samples except for the herbal infusion. The second laboratory experienced problems with chromatography (split peaks, drifting retention times, interferences) and no conclusions can be drawn on analysis for quassine. It is concluded that coumarin and quassine can be determined in a range of sample types using solvent extraction, clean-up using solid phase extraction cartridges and then quantitation by HPLC with UV detection. The chromatographic conditions are reasonably straightforward but the LC column needs to display good separation characteristics and retention time stability in order to separate coumarin and quassine from other co-extractives and use of a DAD is necessary to avoid false positives. 6. Problems were experienced with transfer of the SOP for the cyanogenic potential method and the second laboratory was not able to produce any results within the time and resource available to them. The HCN method is based on a published procedure and it has been improved further and performed well in the single-laboratory validation. Therefore it is recommended that the SOP should be expanded to give further guidance and improve robustness. This report has been prepared by Fera after exercise of all reasonable care and skill, but is provided without liability in its application and use. Page 4 of 79 CONTENTS Section Page GLOSSARY 9 INTRODUCTION 10 Review of published methodology 11 Volatile BAPs 11 Quassine 13 Coumarin 13 Hydrocyanic acid 14 Project objectives 15 PART 1: Development of a method for volatile BAPs 18 Calibration 21 Single laboratory validation 27 Cross validation 28 PART 2: Development of a method for quassine and coumarin 33 Calibration 34 Single laboratory validation 35 Cross validation 38 PART 3: Development of a method for HCN 40 Calibration 41 Single laboratory validation 41 Cross validation 43 Conclusions and recommendations 45 REFERENCES 49 Page 5 of 79 CONTENTS – TABLES Table Title Page 1 Numerical limits for BAPs in foods and beverages 53 2 SDE recovery of BAPs from 50mL of water containing 54 1 mg/L and 0.1 mg/L 3 Single laboratory validation data for SDE method (% 55 recovery) 4 SDE cross validation results for volatile BAPs by SDE 57 5 Coumarin and quassine single laboratory validation 58 data by direct analysis 6 Coumarin and quassine single-laboratory validation 59 data following clean up using SPE 7 Repeat analysis of coumarin-containing foods and 60 quassine-spiked food 8 Coumarin and quassine analysis cross-validation 61 9 HCN single laboratory validation data 62 Page 6 of 79 CONTENTS - FIGURES Figure Title Page 1 Chemical structures of BAPs studied and associated 63 chemicals 2 SDE apparatus 64 3 SDE-GCMS calibration graph 65 4 GC-MS TIC BAPs standard mix 66 5 GC-MS SIM BAPs standard mix 67 6 GC-MS SIM chromatograms of mint flavouring SDE extract 68 7 SIM chromatograms of flavoured crisps SDE extract 69 8 Coumarin and quassine calibration graphs for solvent 70 standards 9 Coumarin calibration for spiked caramel biscuit 71 10 HPLC-PDA of cinnamon powder extract spiked with 72 quassine 11 HPLC of camomile tea extract 73 12 Hydrolysis of amygdalin to hydrogen cyanide 74 13 Comparison of KCN and amygdalin calibration 75 14 HCN colour development time plot 76 Page 7 of 79 ANNEX I. Standard Operating Procedure for the simultaneous distillation-extraction GC-MS method for the determination of biologically active principles in foods and beverages ANNEX II. Standard Operating Procedure for the HPLC method for the determination of coumarin and quassine in foods and beverages ANNEX III. Standard Operating Procedure for the spectrophotometric method for the determination of the cyanogenic potential of foods and beverages Page 8 of 79 GLOSSARY AMG Amygdalin AOAC The Association of Official Analytical Chemists BAPs Biologically active principals COSHH Control of substances hazardous to health DCM Dichloromethane EFSA European Food Safety Authority EU European Union FID Flame ionization detection FSA Food Standards Agency GC Gas chromatography HCN Hydrogen cyanide HCN eq Hydrogen cyanide equivalents HPLC High performance liquid chromatography HS Headspace IL Independent laboratory LC-MS Liquid chromatography-mass spectrometry LN Likens and Nickerson LOD Limit of detection LOQ Limit of quantitation 3-MA 3-methyl acetophenone MS Mass spectrometry MU Methyl umbelliferone PDA Photodiode array RSD Relative standard deviation (%) SCAN Full scan mode SDE Simultaneous distillation extraction SFE Supercritical fluid extraction SIM Selective ion monitoring SOP Standard operating procedure SPME Solid phase micro extraction TDI Tolerable daily intake TIC Total ion chromatogram UV Ultra violet VIS Visible Page 9 of 79 INTRODUCTION Biologically active principles (BAPs) are undesirable chemical substances that are present naturally in some flavourings and food. In the UK, the Flavourings in Food Regulations (SI 1992 No. 1971) as amended (1) , control the use of flavourings by laying down concentration limits for certain foodstuffs and beverages. These regulations enact Directives 88/388/EEC (2) and 91/71/EEC (3) on flavourings for use in foodstuffs. A new EU Regulation (4) has been published recently that focuses on the foods and beverages through which BAPs contribute most to the diet. The Regulation sets numerical limits on specific BAPs in food and beverages (Table 1). This new EU Regulation will apply from 20 th January 2011. Reliable analytical methods are required to enforce the legislation and to carry out surveillance