Advances in Direct Mass Spectrometry Techniques Coupled with Chemometric Modelling for the Rapid Detection of Food Fraud Ken Rosnack Principal Market Development Manager [email protected] ©2019 Waters Corporation 1 Overview Food Fraud / Adulteration Introduction Vanilla Authenticity Belgian Butter (PDO) Olive Oil ©2019 Waters Corporation 2 Introduction to food fraud A food safety incident is typically an unintentional act with unintentional harm – e.g. German E. coli O104:H4 outbreak in 2011 A food defence incident is an intentional act with intentional harm – e.g. Punjab sweet poisoning with chlorfenapyr in 2016 Food fraud is most commonly referred to as the intentional defrauding of food and food ingredients for economic gain Food fraud encompasses the terms: – Food authenticity; ensuring that food offered for sale or sold is of the nature, substance and quality expected by the purchaser – Economically motivated adulteration (EMA); intentional substitution or addition of a substance in a product for the purpose of increasing the apparent value of the product or reducing the cost of its production ©2019 Waters Corporation 3 Food Products at Risk for Food Fraud Meat Milk Olive oil Fish / Seafood Organic foods Cereals, grains and rice Honey and maple syrup Coffee and tea Select herbs and spices Wine Fruit juices ©2019 Waters Corporation 4 Mass Spectrometric Profiling Traditional • Simple, sample preparation • GC-MS • LC-MS, LC-MS/MS • Few thousand components in 5-30 min Direct + • No sample prepared + • Formation of sample originated + gaseous ions Gas-phase ions Sample • Spatially well defined • Few hundred components in 1-5 sec ©2019 Waters Corporation 5 Direct MS for flexible & rapid screening Sample preparation Determination No requirement for No chromatographic step sample extraction Ambient ionisation Results in Data analysis near-real (2s) time Point-of-control analysis ©2019 Waters Corporation 6 Vanilla Authenticity Testing ©2019 Waters Corporation 7 ASAP – how does it work? ASAP Heated N2 desolvation gas Vaporisation step + [H(H2O)n] N* [M+H]+ + + * [M+H] [H(H2O)n] N N +• 4 N* Corona discharge induced ionisation MS inlet Sample loaded onto Licensed from M&M Mass Spec Consulting melting point capillary ©2019 Waters Corporation 8 Vanilla Extract Vanilla Bean Vanilla Orchid ©2019 Waters Corporation 9 ASAP: Sample Loading Procedures ABC < 5 seconds ©2019 Waters Corporation 10 Vanilla Flavor Related Compounds Major components of pure vanilla extracts Synthetic flavor marker ©2019 Waters Corporation 11 Ingredient Authenticity - Food and Beverage Products Vanillin Ethyl vanillin MS AP+ 1a, Imitation vanilla A 167 100 153 Compound m/z Vanillic acid 169 %%%% Caffeine 195 0 1b, Imitation vanilla B Vanillin Ethyl vanillin 167 100 153 Ethyl vanillin Vanillin 153 %%%% Vanillic acid 167 169 4-Hydroxybenzaldehyde 123 0 Vanillin 1c, Pure vanilla extract C 153 Vanillin 100 3a, Cookie A MS AP+ 100 153 Ethyl vanillin %%%% 4-hydroxybenzaldhyde Vanillic acid 167 127 145 169 123 137 139 163 0 %%%% m/z 120 125 130 135 140 145 150 155 160 165 170 175 139 0 Vanillin 3b, Cookie B 100 153 169 167 195 153…. 123 %%%% 145 177 149 0 m/z ©2019 Waters Corporation 130 135 140 145 150 155 160 165 170 175 12 Ingredient Authenticity - Food and Beverage Products Compound m/z Caffeine 2a, French vanilla coffee MS AP+ Vanillic acid 169 100 195 Caffeine 195 Ethyl vanillin 167 Vanillin Vanillin 153 %%%% 153 Ethyl vanillin 167 175 0 Caffeine 2b, Irish cream coffee 100 195 %%%% Ethyl vanillin 167 175 0 m/z 145 150 155 160 165 170 175 180 185 190 195 200 ©2019 Waters Corporation 13 Coumarin Screening Gives Imitation Vanilla a better taste Coumarin m/z = 147 ©2019 Waters Corporation 14 Product Ion Spectra of Coumarin Coumarin standard (2 mg/Kg) Fragments of 147 AP+ 100 91 3.39e5 % 103 65 77 147 0 Wafer cookie (blank, un-spiked) 100 3.39e5 % X10 magnified 67 79 91 105 119 0 Coumarin spiked wafer cookie (2 mg/Kg) 100 3.39e5 91 % 103 77 65 147 0 m/z 40 60 80 100 120 140 160 ©2019 Waters Corporation 15 REIMS Rapid Evaporative Ionization Mass Spectrometry ©2019 Waters Corporation 16 REIMS workflow Model training & validation Different production Different species Quality & product consistency Real time recognition Biomarker discovery ©2019 Waters Corporation 17 REIMS - How does it work? To the atmospheric inlet of a mass spectrometer REIMS (& LA-REIMS) • No Voltage required, aerosol droplets formed from tissue disintegration. • Small droplets with unequal distribution of anions and cations • Accelerated onto Hot (900oC) coil (abrupt heat) • Rapid desolvation forms gaseous ions in the first stage pumping region ©2019 Waters Corporation 18 Ionisation Method An atmospheric sampling technique – Uses rapid heating to evaporate the sample o Joule heating ‐ o Contact heating o Radiative heating - - - o IR LASER +Gas+ phase ions-- + + - ‐ Sample the aerosol produced ©2019 Waters Corporation 19 Product authenticity Belgian butter – PDO status (1996) ©2019 Waters Corporation 20 Fatty acids Tof MS neg ion Oleic acid 250–300 m/z region Essential C18H34O2 Ω6 fatty acid Palmitic acid C16H32O2 Linoleic acid C18H32O2 Ω9 ©2019 Waters Corporation 21 Saturated fatty acid in cow’s milk Myristic acid (tetradecanoic acid) C14H28O2 HMDB 00806: Myristic acid is a saturated 14-carbon fatty acid occurring in most animal and vegetable fats, particularly butterfat and coconut, palm, and nutmeg oils ©2019 Waters Corporation 22 OPLS-DA Scores Plot ©2019 Waters Corporation 23 S-Plot Selective markers S‐Plot (Belgian butter = ‐1, Industrial butter = 1) 1.0 0.8 Belgian butter 0.6 0.4 0.2 -0.0 p(corr)[1] (Correlation) -0.2 Commercial butter -0.4 -0.6 -0.8 -1.0 -0.58 -0.56 -0.54 -0.52 -0.50 -0.48 -0.46 -0.44 -0.42 -0.40 -0.38 -0.36 -0.34 -0.32 -0.30 -0.28 -0.26 -0.24 -0.22 -0.20 -0.18 -0.16 -0.14 -0.12 -0.10 -0.08 -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12 p[1] (Loadings) EZInfo controlled by Progenesis QI 3 - Belgian butter.usp (M14: OPLS-DA) - 2015-10-12 16:06:29 (UTC+0) ©2019 Waters Corporation 24 S-Plot Selective markers S‐Plot (Margarine (Flora) = ‐1, Spanish olive oil = 1) 1.0 Spanish Olive Oil 0.8 0.6 0.4 0.2 -0.0 p(corr)[1] (Correlation) -0.2 Margarine (Flora) -0.4 -0.6 -0.8 -1.0 -0.30 -0.28 -0.26 -0.24 -0.22 -0.20 -0.18 -0.16 -0.14 -0.12 -0.10 -0.08 -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24 p[1] (Loadings) EZInfo controlled by Progenesis QI 3 - Oil and butter.usp (M6: OPLS-DA) - 2015-10-13 17:23:22 (UTC+0) ©2019 Waters Corporation 25 Real time model builder software PCA & LDA space PCA Model 4 different Belgian butter samples LDA Model ©2019 Waters Corporation 26 Laser REIMS Olive Oil ©2019 Waters Corporation 27 REIMS Using Lasers or LA-REIMS IR Laser used to vaporise Sample – Similar spectral response as D-REIMS – Currently o CO2 ( = 10.6 m) o OPO ( = 2 - 3 m) • Higher Performance – Advantages o Non conductive samples o Minimal Damage o Non Contact • Reduced Contamination ©2019 Waters Corporation 28 LA-REIMS Workflow Three regional IGP/DOP EVOO samples 8 x 5 uL spots on microscope slides Analysed using LA-REIMS platform (5 s per spot) Data processing using LiveID and analysis with PLS-DA and Random Forest ©2019 Waters Corporation 29 LA-REIMS for IGP / DOP EVOO traceability Sesame oil Olive oil - Valdi Olive oil - Terradi S‐Plot (Olive oil ‐ Terradi = ‐1, Olive oil ‐ Toscano = 1) 1.0 417.0978 1075.7260 1101.7406 838.5306 1102.7457 1103.7507 0.9 1073.7087497.07951100.7305837.52551099.7273 217.0451 337.0946 215.0297 439.0911647.2435 1179.9080 537.1455 218.0485 361.0664 170.07011180.9170 675.2681551.17911179.8278 199.0647 553.1417385.1242811.5071495.2108661.2837 395.3780 0.8 538.1518396.3811515.13321076.7316423.4068336.1126 335.1078 1181.9290 839.5387216.0342521.25561097.7221959.70661155.8319 1104.7611533.1979631.26391213.8915565.4925904.73741185.86651171.85371194.84161170.8491 253.1391 227.0789687.3138328.1949889.67501141.9361341.19721222.8716 1207.85371182.84991198.87341169.84541181.84381224.88821211.87951183.8575 226.1243144.0530500.1971262.0557523.2411353.1959549.1494499.28041142.9441537.4646219.0491877.72061180.83111226.90091199.87781221.86521212.8846327.19031225.89341184.8636 1197.8691 1210.8750 299.14601156.8353323.1637 1223.8835 339.1948 343.1070633.2514237.0724605.2186367.3494519.21951113.9084295.1332225.06011208.86001193.8409 326.1824 1195.8554 155.0824536.4556424.4098805.4300387.1325566.4963247.0435905.7491835.5213258.0699645.28321167.8291875.7089298.15301139.9253 340.1987257.0664 1196.86071209.8720 325.1794 393.1755765.4355255.1042573.2962285.04171143.9070876.7145 891.6768903.7303 311.1646 0.7 591.1897357.08751157.8406267.1052239.0752 312.1681 297.1499 200.0679857.3418719.2780617.2599354.2028565.19111111.8931517.1501626.51641153.8136659.3014961.7345313.1702 591.2388364.1091409.16361115.7201579.2012933.67811138.9145915.6927 269.0719158.0679 677.2625689.26371186.9400314.1695873.6960899.7082185.0311253.09051140.9313 917.6931 354.1124284.13461153.8983271.05771237.8838381.3627269.1157703.2710321.2741701.2272277.0708318.1018 1058.8665559.1273991.5595427.1635458.0618368.3528261.0741509.1792184.0212932.76661137.9114890.6967197.0319 268.1649403.0992169.0896761.25951100.8665589.2268879.3770743.3480338.1880182.06082082.5407308.2612 914.6905211.0467317.0982918.6850 149.0700602.4780960.7970475.1011169.0576225.1140186.0428240.1337273.0701578.3458342.1944525.29071114.9162457.0593384.34581168.9547 902.7266181.0560901.7214 1081.84951156.9315143.0403165.1058186.1123359.1370409.0978685.2525671.23272080.5227287.0706665.53411227.9121289.0877928.73661167.9485355.3139