Recent Developments in Coffee Flavour Formation Using In-Bean Experiments

Transcription

1 23 rd ASIC Conference, Bali (Indonesia), ctober 3-8, 2010 Coffee Chemistry Recent Developments in Coffee Flavour Formation Using In-Bean Experiments L. Poisson, J. Kerler, T. Davidek, and I. Blank (estle PTC rbe, Switzerland)

2 From the plant to the cup: Roasting is the critical step in flavour formation Genetic, botanic Cultivation, harvesting Post-harvest treatment Coffee technology Roasting Instant coffee PTC rbe Cup preparation Roast & ground coffee

3 Coffee roasting: Some compounds formed from green bean precursors H Furaneol H C 2H + icotinic acid H C H 2 HMF Pyridine and derivatives C H Caramelization Hydrolysis Coffee Flavour Precursors Pyrolysis SUGARS Sucrose, glucose, fructose (arabinogalactans) + AMI ACIDS (Strecker-active AA) TRIGELLIE CHLRGEIC ACIDS Cyclisation Chlorogenic acid lactones Fragmentation Hydrolysis R Diketones 2-Furfurylthiol Aldehydes S H H HC 2 H C H 3 C 2 H H C 2 C H 3 C 2 H C 2 H Maillard, Strecker R 1 R 2 R 3 R 3 S H Pyrazines Formic acetic glycolic lactic R 1 R 2 H 3-Mercapto-3-methylbutyl-formate H H Quinic acid H + Quinic acid lactone C 2 LIPIDS RGAIC ACIDS xidation Carbonyls Caffeic C H + Ferulic acids R E- -Damascenone xidative cleavage CARTEIDS PTC rbe H Vanillin H Guaiacols

4 bjective of the study Get a better understanding of the relative importance of flavour precursors in coffee under realistic conditions Impact coffee odorants Formation: Food (in-bean experiment) vs. Model studies (only precursor) Conclusions PTC rbe

5 ew experimental set-up Biomimetic in-bean study Water extraction at 95 C for 2 h Green coffee beans (GC) Spiking with precursors Exhausted beans (EB) Spiked green beans + Roasting Roasting Biomimetic recombinate (BR) Roasted beans spiked with precursors EB + BR Precursor omission Roasting Incorporation of labelled precursors Roasting mission EB + BR omitted in sugars/aa EB + BR + labeled precursors Spiking PTC rbe Labelling

6 Composition of the biomimetic recombinate (BR) Components Amount (g) Components Amount (g) Phenols Amino acids (cont.) Chlorogenic acid L-Proline Trigonelline L-Tyrosine Caffeine L-Valine rganic acids L-Isoleucine D-(-)-Quinic acid L-Leucine L-(+) -Lactic acid L-Phenylalanine Potassium acetate L-Lysine Malic acid Sugars Potassium citrate D-(-)-Arabinose Amino acids L-(+) -Rhamnose L-Alanine D-(+) -Galactose L-Threonine D-(+) -Sucrose L-Serine D-(+) -Mannose L-Aspartic acid Metals L-Glutamic acid Iron(III) chloride L-Glycine Manganese(II) chloride L-Histidine Copper(II) sulfate L-Arginine Zinc(II) sulfate, heptahydrate PTC rbe

7 1. Formation of 2-furfurylthiol (FFT): Model studies Model System Tressl et al. (1993): FFT is formed from arabinose/cysteine via 3-deoxypentosone (3-DP) and furfural while maintaining the intact carbon chain R H H Arabinose H + HS H 2 Cysteine H 3-DP 2-Furaldehyde Figure: Hypothetical formation of FFT in coffee from arabinogalactans or arabinose (R=H) and cysteine (protein-bound) H 2 S H 2-Furdurylthiol (FFT) SH Grosch (1999): Arabinogalactans suggested as precursor of FFT by isolating the polysaccharide from green coffee and roasting it in the presence of cysteine PTC rbe

8 Relative Conc. [%] Formation of 2-Furfurylthiol (FFT): mission and spiking experiments Relative Conc. [%] In-bean experiment 2-furaldehyde 2-furfurylthiol 175% 175% 2-furaldehyde 2-furfurylthiol % 125% % 75% % 25% EB + BR EB + BR - mission AA EB + BR - mission sugars Green Coffee (GC) GC + Sucrose GC + Arabinose GC + Cysteine mission of sugars favored the generation of FFT, whereas furfural content was highly suppressed SH Spiking with sucrose increased furfural amounts but considerably decreased concentrations of FFT Spiking experiment with cysteine resulted in enhanced FFT amounts, thus indicating cysteine as a suitable sulphur source PTC rbe

9 relative Intensity (%) Formation of 2-furfurylthiol (FFT): Labelling experiments relative Intensity (%) In-bean experiment 10 88% furfural 10 88% 2-furfurylthiol M % M % % 1. M+2 98 M+3 99 M M M % 4.2% 5.1% 0.4% 0.9% M M M M M Incorporation of D-[U- 13 C 5 ]-arabinose did not yield fully labelled furfural nor FFT, but partially labelled FFT with 13 C 1, 13 C 2 and 13 C 3 -moieties 2-Furfurylthiol (FFT) is most likely not generated via the furfural pathway! PTC rbe

10 2. Formation of alkyl pyrazines: Model studies Model System Amrani-Hemaimi et al. (1995): Potential precursors are C6 and C5 sugars (e.g. fructose, glucose, arabinose) and degradation compounds form key intermediates (e.g. α-amino carbonyls) through the Strecker reaction Amrani-Hemaimi et al. (1995): In addition, alanine and glycine play a key role as they are incorporated into the side chain of the alkyl pyrazine CH 2 H H H CH 2 H H Fructose + * H 3 C * H 2 Alanine H 2 H 2 H 2 H 2 H H * * * * EDMP DEMP Figure: Precursors and key intermediates of pyrazines PTC rbe

11 Relative Conc. [%] Relative Conc. [%] Formation of alkyl pyrazines: mission and spiking experiments In-bean experiment mission of amino acids or sugars Spiking of green coffee with sugars (equimolar amounts) or alanine 20 2-ethyl-3,5-dimethylpyrazine 2,3-diethyl-5-methylpyrazine 40 2-ethyl-3,5-dimethylpyrazine 2,3-diethyl-5-methylpyrazine ~95 175% % % % 5 EB + BR EB + BR - mission AA EB + BR - mission sugars Green Coffee (GC) GC + Sucrose GC + Arabinose GC + Alanine mission of free amino acids and spiking with alanine confirmed their importance Sugars omission resulted in highly increased pyrazine concentrations Amino acids, in particular alanine, are key precursors in the formation of ethyl pyrazines Spiking of green beans with sugars had a suppressing effect PTC rbe

12 relative Intensity (%) Formation of alkyl pyrazines: Labelling experiments relative Intensity (%) In-bean experiment Spiking with L-[3-13 C]-alanine Incorporation of [U- 13 C 6 fructose]-sucrose (16% of natural content) ,3-diethyl-5-methylpyrazine * * % 2-ethyl-3,5-dimethylpyrazine % M % 32.3% M M M M M % M % M % M M C 3 -Fragments are formed from free sugars as well as from non-water extractable polysaccharides: Labeled alanine is efficiently incorporated into ethyl pyrazines Pyrazine formation implies competition between bound and free sugars for the water extractable -source PTC rbe

13 3. Formation of Furaneol : Model studies Model System Hofmann & Schieberle (2001): Furaneol is mainly formed from the intact C6-glucose skeleton under roasting conditions (dry heating) Hofmann & Schieberle(2001): Under aqueous conditions, furaneol is also formed by recombination of reactive C 3 sugar fragments (ca. 6) CH 2 - H-CH-CH C= (CH-H) 3 CH 2 -H CH CH-H R Amadori compounds C 3 + C 3 Sugar fragments CH CH-H H CH CH-H (CH-H) 2 CH 2 -H Pentose/ Glycine CH CH-H (CH-H) 3 CH 2 -H Hexose (CH-H) 3 CH 3 6-Deoxyhexose (CH-H) 3 CH 2 -P Hexose-6-P PTC rbe

14 Relative Conc. [%] Formation of furaneol: mission experiments relative Intensity (%) In-bean experiment 4-hydroxy-2,5-dimethyl-3[2H]-furanone 175% % % 4-hydroxy-2,5-dimethyl-3[2H]-furanone 75% % EB + BR EB + BR - mission AA EB + BR - mission sugars 2 M 128 1% M M M M M % M Free sugars (sucrose) play a major role in the formation of Furaneol mission of free sugars significantly decreased concentrations (-55%). mission of amino acids promoting generation (+62%): less trapping of reactive C 3 PTC rbe Spiking with [ 13 C 6 -Fru]-sucrose led to exclusive generation of fully labeled and fully unlabelled molecules In agreement with previous model studies: Furaneol is formed from intact C6-glucose Fructose is a much more efficient precursor as compared to glucose 16% labeled Furaneol formed, with only 1/6 of the total sucrose amount added as labeled isotopomer

15 4. Formation of diketones: Model studies Model System Yaylayan et al. (1999): Diketones are formed from mono-, di- & oligosaccharides by Maillard-type reactions in the presence of amino acids like glycine and alanine Schieberle et al. (2003): CAMLA experiments indicated that the recombination of C3 and C1 intermediates is the predominant pathway in 2,3-butanedione formation CH 2 H H H H H + R H 2 H H H e.g. Glucose Glycine (R=H) Alanine (R=CH3) 2,3-Butanedione 2,3-Pentanedione Figure: Precursors and key intermediates of coffee diketones PTC rbe

16 Relative Conc. [%] Formation of diketones: mission and spiking experiments Relative Conc. [%] In-bean experiment mission of amino acids or sugars Spiking of green coffee with sugars (equimolar amounts) or alanine 2,3-butanedione 2,3-pentanedione 2,3-butanedione 125% 175% 2,3-pentanedione % 75% % 25% 5 25% EB + BR EB + BR - mission AA EB + BR - mission sugars Green Coffee (GC) GC + Alanine GC + Sucrose mission of sugars affect the formation of both diketones, in particular of 2,3- pentanedione mission of free amino acids and spiking with alanine in green coffee do not influence the formation of diketones PTC rbe Addition of sucrose increases particularly 2,3-pentanedione (+6), to a lesser extent 2,3-butanedione (+1)

17 relative Intensity (%) Formation of diketones: Labelling experiments % 2,3-pentanedione nly 17% of labelled sucrose compared to original amount Incorporation of [U- 13 C 6 -fructose]-sucrose led also to fully labelled 2,3-pentanedione (ca. 1) 2,3-Butanedione is only formed from the glucose moiety 2 M % 3.5% 0.5% 9.3% M M M M M Free sugars play a key role in diketones formation Amino acids are not the limiting factor in diketone formation ( ) 2,3-butanedione (6) 2,3-pentanedione (7) ion m/z BREB 1 + [ 13 C6]- RGB+ [ 13 C]- m/z BREB b + [ 13 C6]- RGB+ [ 13 C]- SUC ALA SUC ALA M M M M M M M+5 0 Diketones show different formation pathways PTC rbe In-bean experiment

18 Conclusions Biomimetic in-bean experiments: A powerful tool for the elucidation and verification of proposed formation pathways of aroma compounds in complex food systems such as coffee Combination of omission, spiking, and labelling experiments under real food matrix conditions are very useful in providing more precise insights into Maillard-type formation mechanisms Great diversity of precursors and other co-reaction agents present in the green bean lead to competing and even different pathways in flavor formation (e.g. 2-furfurylthiol, diketones) Model versus Food: Results obtained in model systems need to be validated under food processing conditions PTC rbe

19 For more info see Poisson et al. (2009) J. Agric. Food Chem., 57, PTC rbe