EU funded project PURE JUICE, G6RD-CT-2002-00760 Workshop 9/03/2006 WP3: Detection of undeclared added organic acids WP leader: Eurofins
Where do organic acids come from?
Citric acid: always derived from sugar (and water) 1 CHO 2 CHOH 3 CHOH 4 CHOH Glucose 5 CHOH 6 CH2OH PEP carboxylase 1/6 2/5 CH2 ll C-O-P Glycolyse Phosphoénolpyruvate 100% C atoms sugar 1/6 1/6 CH2 C=O Malate déshydrogénase CH2 CHOH Enzyme malique 3/4 1/6 Acide oxalo-acétique Acide malique Pyruvate CH3 C=O SCoA Cycle de Krebs H 2 O Acetyl CoA Citrate synthétase Acide citrique 1/6 CH2 1/6 CHO - CH2 - H 2 O 100% H 1 CH 2 Water 1 CH 2 2/3 H Glucose 1, 6, 6 1/3 H Water D/H plant water > D/H tap water
δ 13 C values of citric acid Lemon vs. artificial sources 60 50 40 No of obs 30 20 10 0-28.4-24.68-20.96-17.24-13.52-9.8-26.54-22.82-19.1-15.38-11.66 Industrial biosynthesis Authentic lemons
L-malic acid: biosyntheses 1 CHO 2 CHOH 3 CHOH 4 CHOH Glucose 5 CHOH 6 CH2OH Glycolyse PEP carboxylase 1/6 2/5 CH2 ll C-O-P Phosphoénolpyruvate Enzyme malique 3/4 1/6 Pyruvate CH3 C=O SCoA Acetyl CoA Krebs cycle (n times): CH(OH) 100% H Water CH 2 Citrate synthétase 1/6 CH2 1/6 C=O Acide oxalo-acétique Acide citrique Malate déshydrogénase Cycle de Krebs CH2 1/6 CHO - CH2 - PEP carboxylase: CH(OH) Water CH 2 Glucose 1, 6, 6 1/6 CH2 CHOH Acide malique
L-malic acid: industrial synthesis O O OH 1 C C 2 H C O H C OH 3 H C H C H 4 O C O C OH Maleic anhydride L-malic acid (from fossil material) (adapted from Schmidt et al., Flüssiges Obst, 3/2000, 131-136)
δ 13 C values of L-malic acid Apple vs. artificial sources 33 30 27 24 21 No of obs 18 15 12 9 6 3 0 <= -28 (-28;-26] (-26;-24] (-24;-22] (-22;-20] (-20;-18] ARTIFICIAL APPLE
What has been done in the Pure Juice project?
Flow chart of the protocol used for the isotopic analysis of organic acids Juice or diluted concentrate Filtrate CENTRIFUGATION FILTRATION ION EXCHANGE CHROMATOGRAPHY Amino-acids Sugars Organic Acids PREPARATIVE HPLC Citric L-Malic 13 C D/H 13 C Positional 13 C
Orange sugars & organic acids 13 C deviations -20 Y = X -21 δ 13 C of organic acids ( ) -22-23 -24-25 -26-27 -28 Malic acid Citric acid -29-30 -30-29 -28-27 -26-25 -24-23 -22-21 -20 δ 13 C of sugars C13 ( )
Orange sugars & organic acids 13 C deviations -20 Y = X -21 δ 13 C of organic acids ( ) -22-23 -24-25 -26-27 -28 Malic acid Citric acid -29-30 -30-29 -28-27 -26-25 -24-23 -22-21 -20 δ 13 C of sugars C13 ( ) These correlations are independant from the geographical origin and allow to reduce the detection limits of acids/sugar addition
1 2 CHOH 3 CH2 4 Ce IV H 2 SO 4 δ 13 C of CO 2 from site 1 L-Malic acid: global and positional δ 13 C -20 Y = X -21 C1 of malic acid C13 ( ) -22-23 -24-25 -26-27 -28-29 -30-31 -32-33 -34 CHINA FRANCE GERMANY ITALY POLAND ROUMANIA SPAIN TURKEY UKRAINE Commercial L-malic acids Commercial, out of project -35-30 -29-28 -27-26 -25-24 -23-22 -21-20 Malic acid C13 ( )
New (D/H) analysis of citric acid : principle CH 2 + CaCl 2 C 12 H 10 Ca 3 O 14 130 C D/H by Py-IRMS HO C Calcium citrate CH 2 (residual water) Citric acid
New (D/H) analysis of citric acid : precision Number of samples Repetitions per sample repeatability standard deviation, Sr (ppm) repeatability limit, r ( ppm ) Py-IRMS measurement 10 3 0.2 0.5 Overall method (from the juice) 5 2 0.2 0.5 New (D/H) analysis of citric acid : accuracy Sample (D/H) citrate (ppm) by IRMS (this project) (D/H) TEC (ppm) by SNIF-NMR (previous eurofins study*) Difference (ppm) A 149.0 146.5 2.5 B 147.8 146.9 0.9 C 147.4 146.3 1.1
(D/H) analysis of citric acid : spiking experiment (in orange) 162 160 158 (D/H) Calcium citrate (ppm) 156 154 152 150 148 146 144 0 10 20 30 40 50 60 70 80 90 100 % added citric acid
(D/H) analysis of citric acid : authentic samples 164 (D/H) Calcium citrate (ppm) 162 160 158 156 154 152 150 ORANGE LEMON GRAPEFRUIT PINEAPPLE RASPBERRY STRAWBERRY BLACKCURRANT COMMERCIAL 148 146 144-30 -25-20 -15-10 -5 δ 13 C citric acid Published in «Detection of exogenous citric acid in fruit juices by stable isotope ratio analysis», Jamin et al., JAFC, 2005
(D/H) analysis of citric acid : lemon market samples 161.0 159.0 LEMON (D/H) Calcium citrate (ppm) 157.0 155.0 153.0 151.0 149.0 Passed Failed COMMERCIAL Lemon juice, market samples 147.0 145.0-30.0-25.0-20.0-15.0-10.0-5.0 δ 13 C citric acid ( /V.PDB)
Peer-testing of the methods
Positional C13: Peer-testing design EA-IRMS (method A) 1 2 CHOH Ce IV H 2 SO 4 δ 13 C of CO 2 from site 1 GC-C-IRMS (method B) 3 CH2 4 Sample Type L-malic acid commercial L-malic acid From apple L-malic acid C mixture of A and B Sample Code Test Material Number Experimental Design A 1 + 6 blind duplicates B 2 + 5 blind duplicates C 3 + 4 blind duplicates
Positional C13: Peer-testing results -20-21 -22-23 -24-25 -26 mixture natural Delta 13C of CO2 ( ) -27-28 -29-30 -31-32 -33-34 -35-36 -37-38 -39-40 artificial 0 1 2 3 4 5 6 7 Lab1 (method B) Lab2 (method A) Lab3 (method A) Lab4 (method A)
(D/H) analysis of citric acid : peer-test citrate (D/H) corrected using familiarisation sample (median) 170.0 165.0 Labs results 160.0 155.0 Reference sample ex-fruit Lab1 Lab2 Lab3 Lab4 150.0 Mixture 145.0 artificial 140.0 144.0 146.0 148.0 150.0 152.0 154.0 156.0 158.0 160.0 Mean of 4 labs
Blind tests
Blind test: apple SGF % added δ 13 C δ 13 C malic acid δ 13 C C1 from malic acid Difference reference malic acid ( ) ( ) ( ) Conclusion AS1 23-25.93-25.61 0.32 SUSPECT AS2 16-25.87-25.33 0.55 IN AS3 27-25.06-24.17 0.88 IN AS4 none -25.99-25.66 0.33 SUSPECT AS5=AS1 23-25.99-25.90 0.09 OUT The deviations observed are small, due to the nature of the added malic acid («worst case» situation), hence some undetected additions On the other hand there was no false positive The correlation with sugars should also be taken into account for routine
Blind test: citric acid in orange SGF % added δ 13 C citric acid δ 13 C sugar (D/H) calcium citrate Conclusion reference Origin citric ( ) ( ) (ppm) with origin OS7 Mexico 20% -24.7-26.4 157.3 OUT OS8 Cuba 30% -24.3-25.2 157.0 OUT OS9=OS8 Cuba 30% -24.6-25.3 157.9 OUT OS10 Brazil none -24.5-25.4 158.2 IN OS11 Spain 25% -23.9-25.0 155.5 OUT OS1 Mexico 75%, Brazil 25% 25% -24.1-25.2 155.3 OUT No false positive / all additions detected Carbon 13 profile was not able to detect the additions Knowledge of the origin was needed to evidence the adulteration in samples OS7, 8 and 9
Blind test: citric acid in pineapple SGF reference % added citric δ 13 C citric acid ( ) δ 13 C sugar ( ) (D/H) calcium citrate (ppm) Conclusion AN1 30-17.5-13.9 155.5 OUT AN2 1 g/l -15.2-12.8 155.1 OUT AN3 20-15.4-12.8 155.2 OUT AN4=AN1 30-17.4-14.0 155.8 OUT AN5 none -11.6-11.8 153.8 OUT All additions have been detected Carbon 13 profile was sufficient to detect the additions in this case (because a C3 source was used) One «false positive»: further investigations* have shown that this sample is not authentic
*Pineapple juice concentrate AN5: (D/H)I ethanol (norm/vsmow) = 105.1 ppm (AIJN range 107-111.5) Pineapple juice used to make this concentrate δ 18 O water = -6.2 Conclusion: this can not be considered as an authentic sample Nevertheless pineapple results for citrate D/H should be interpreted with care (because water isotope ratios vary a lot in this fruit)
Outcomes: 1) 13 C multi-component profiles Improved detection of added C4 sugars in C3 plants (resp. C3 sugars in pineapple), when considering the correlation between sugars and acids (down to ca 10%) Combined with the site-specific information, the 13 C profile also improves the detection of added citric and malic acid (down to ca 10% in case of a different metabolism)
Outcomes: 2) positional 13 C analysis of malic acid The positional analysis of L-malic acid allows to detect additions of the artificial source previously undetectable by other methods Based on the Peer-testing results, a close agreement can be obtained between laboratories As confirmed by the blind test the detection limit varies between 20 and 30% (depending on the adulterant used), which can be improved using the correlation with sugars
Outcomes: 3) D/H analysis of citric acid A new method for analysing the D/H ratio of nonexchangeable hydrogen in citric acid has been developed and published. It allows to detect all sources of artificial citric acid in fruit juices Based on the preliminary Peer-testing results, a close agreement can be obtained between laboratories provided that a common citric standard is used for the IRMS analysis As confirmed by the blind test the detection limit lies around 20%, and can be improved using the correlation with sugars Knowledge of the origin helps to detect low additions
Overall conclusion for organic acids A large database covering the main origins of apple and orange juices was built, and partly extended to other fruits such as pineapple, blackcurrant, lemon and grapefruit The methods cover all known sources of artificial L- malic acid and citric acid The detection level of added acids were found to be in the range 10-30%, which is still more efficient than any other method For routine analysis the combination with other analyses / parameters will improve the efficiency of these methods