Interactions between volatile and nonvolatile coffee components. 1. Screening of nonvolatile components

This study is the first of two publications that investigate the phenomena of coffee nonvolatiles interacting with coffee volatile compounds. The purpose was to identify which coffee nonvolatile(s) are responsible for the interactions observed between nonvolatile coffee brew constituents and thiols, sulfides, pyrroles, and diketones. The overall interaction of these compounds with coffee brews prepared with green coffee beans roasted at three different roasting levels (light, medium, and dark), purified nonvolatiles, and medium roasted coffee brew fractions (1% solids after 1 or 24 h) was measured using a headspace solid-phase microextraction technique. The dark roasted coffee brew was slightly more reactive toward the selected compounds than the light roasted coffee brew. Selected pure coffee constituents, such as caffeine, trigonelline, arabinogalactans, chlorogenic acid, and caffeic acid, showed few interactions with the coffee volatiles. Upon fractionation of medium roasted coffee brew by solid-phase extraction, dialysis, size exclusion chromatography, or anion exchange chromatography, characterization of each fraction, evaluation of the interactions with the aromas, and correlation between the chemical composition of the fractions and the magnitude of the interactions, the following general conclusions were drawn. (1) Low molecular weight and positively charged melanoidins present significant interactions. (2) Strong correlations were shown between the melanoidin and protein/peptide content, on one hand, and the extent of interactions, on the other hand (R = 0.83-0.98, depending on the volatile compound). (3) Chlorogenic acids and carbohydrates play a secondary role, because only weak correlations with the interactions were found in complex matrixes.     

Chemical interactions between odor-active thiols and melanoidins involved in the aroma staling of coffee beverages.

Comparative aroma dilution analyses of the headspaces of aqueous solutions containing either the total volatiles isolated from a fresh coffee brew, or these volatiles remixed with the melanoidins isolated from coffee brew, revealed a drastic decrease in the concentrations of the odorous thiols 2-furfurylthiol, 3-methyl-2-butenthiol, 3-mercapto-3-methylbutyl formate, 2-methyl-3-furanthiol, and methanethiol when melanoidins were present. Among these thiols, 2-furfurylthiol was affected the most: e.g., its concentration decreased by a factor of 16 upon addition of melanoidins. This was accompanied by a decrease in the overall roasty-sulfury aroma. Quantitations performed by means of stable isotope dilution assays confirmed the rapid loss of all thiols with increasing time while keeping the coffee brew warm in a thermos flask. Using [2H2]-2-furfurylthiol as an example, [2H]-NMR and LC/MS spectroscopy gave strong evidence that thiols are covalently bound to the coffee melanoidins via Maillard-derived pyrazinium compounds formed as oxidation products of 1,4-bis-(5-amino-5-carboxy-1-pentyl)pyrazinium radical cations (CROSSPY). Using synthetic 1,4-diethyl diquaternary pyrazinium ions and 2-furfurylthiol, it was shown that 2-(2-furyl)methylthio-1,4-dihydro-pyrazines, bis[2-(2-furyl)methylthio]-1,4-dihydro-pyrazines, and 2-(2-furyl)methylthio-hydroxy-1,4-dihydro-pyrazines were formed as the primary reaction products. Similar results were obtained for models in which either 1,4-diethyl diquaternary pyrazinium ions were substituted by Nalpha-acetyl-L-lysine/glycolaldehyde, or the 2-furfurylthiol by 2-methyl-3-furanthiol and 3-mercapto-3-methylbutyl formate. On the basis of these results it can be concluded that the CROSSPY-derived pyrazinium intermediates are involved in the rapid covalent binding of odorous thiols to melanoidins, and, consequently, are responsible for the decrease in the sulfury-roasty odor quality observed shortly after preparation of the coffee brew.     

Synthesis and structure determination of covalent conjugates formed from the sulfury-roasty-smelling 2-furfurylthiol and di- or trihydroxybenzenes and their identification in coffee brew

Recent investigations demonstrated that the reaction of odor-active thiols such as 2-furfurylthiol with thermally generated chlorogenic acid degradation products is responsible for the rapid aroma staling of coffee beverages. To get a clear understanding of the molecular mechanisms underlying this aroma staling, the existence of putative phenol/thiol conjugates needs to be verified in coffee. The aim of the present study was therefore to synthesize such conjugates for use as reference substances for LC-MS screening of coffee. To achieve this, catechol, 3-methyl-, 4-methyl-, and 4-ethylcatechol, pyrogallol, hydroxyhydroquinone, 5-O-caffeoylquinic acid, and caffeic acid, respectively, were reacted with 2-furfurylthiol in the presence of iron(III) chloride and air oxygen. After purification, the structures of 25 phenol/thiol conjugates were identified by means of LC-MS/MS and 1D/2D NMR experiments. Using these compounds as reference materials, four conjugates, namely, 3-((2-furylmethyl)sulfanyl)catechol, 3-((2-furylmethyl)sulfanyl)-5-ethylcatechol, 4-((2-furylmethyl)sulfanyl)hydroxyhydroquinone, and 3,4-bis((2-furylmethyl)sulfanyl) hydroxyhydroquinone, were identified for the first time in coffee brew by means of HPLC-MS/MS(MRM). These findings clearly demonstrate catechol, 4-ethylcatechol, and hydroxyhydroquinone as the primary thiol trapping agents involved in the aroma staling of coffee beverages.     

Changes in headspace volatile concentrations of coffee brews caused by the roasting process and the brewing procedure

Headspace-solid-phase microextraction technique (HS-SPME) coupled with gas chromatography-mass spectrometry (GC-MS) and gas chromatography-olfactometry (GC-O) were used to characterize the aroma compounds of coffee brews from commercial conventional and torrefacto roasted coffee prepared by filter coffeemaker and espresso machine. A total of 47 volatile compounds were identified and quantified. Principal component analysis (PCA) was applied to differentiate coffee brew samples by volatile compounds. Conventional and torrefacto roasted coffee brews were separated successfully by principal component 1 (68.5% of variance), and filter and espresso ones were separated by principal component 2 (19.5% of variance). By GC olfactometry, a total of 34 aroma compounds have been perceived at least in half of the coffee extracts and among them 28 were identified, among which octanal was identified for the first time as a contributor to coffee brew aroma.     

Quantitative studies on the formation of phenol/2-furfurylthiol conjugates in coffee beverages toward the understanding of the molecular mechanisms of coffee aroma staling

To gain a more comprehensive knowledge of the contribution of recently identified phenol/thiol conjugates to the storage-induced degradation of odorous thiols, the concentrations of the sulfury-roasty smelling key odorant 2-furfurylthiol and the concentrations of the putative thiol-receptive di- and trihydroxybenzenes pyrogallol (1), hydroxyhydroquinone (2), catechol (3), 4-ethylcatechol (4), 4-methylcatechol (5), and 3-methylcatechol (6), as well as of the phenol/thiol conjugates 3-[(2-furylmethyl)sulfanyl]catechol (7), 3-[(2-furylmethyl)sulfanyl]-5-ethylcatechol (8), 4-[(2-furylmethyl)sulfanyl]hydroxyhydroquinone (9), and 3,4-bis[(2-furylmethyl)sulfanyl]hydroxyhydroquinone (10) were quantitatively determined in fresh and stored coffee beverages by means of stable isotope dilution analyses (SIDA). Although 2 was found to be the quantitatively predominant trihydroxybenzene in freshly prepared coffee brew, this compound exhibited a very high reactivity and decreased rapidly during coffee storage to generate the conjugates 9 and 10. After only 10 min, about 60% of the initial amount of 2-furfurylthiol in a coffee beverage reacted with 2 to give 9 and 10. In contrast, conjugate 7 was found to be exclusively formed during coffee roasting because its initial concentration as well as the amount of its putative precursor, phenol 3, was not affected by storage. It is interesting to note that the concentration of 8 was increased with increasing incubation time, but its putative precursor 4 was not affected, thus indicating another formation pathway most likely via the chlorogenic acid degradation product 4-vinylcatechol. This study demonstrates for the first time that the loss of 2-furfurylthiol during coffee storage is mainly due to the oxidative coupling of the odorant to hydroxyhydroquinone (2), giving rise to the conjugates 9 and 10.     

A novel method for quantification of 2-methyl-3-furanthiol and 2-furanmethanethiol in wines made from Vitis vinifera grape varieties

A rapid, easy method has been developed for isolating and quantifying 2-methyl-3-furanthiol (2M3F) in wines. Until now, it was not possible to quantify this highly odoriferous compound, with a smell reminiscent of cooked meat, in wine. The original aspect of this method is the specific release of volatile thiols using a cysteamine solution applied in reverse flow to sample percolation on the basis of a p-hydroxymercuribenzoate (pHMB)-volatile thiol conjugate formed by the direct addition of pHMB to 50 mL of wine. Purification of volatile thiols in wines is much faster and easier than our previous method. This method may also be used to assay 2-furanmethanethiol in wine. This thiol's strong aroma of roasted coffee has been shown to contribute to the "roast coffee" aroma of certain wines. Assaying 2M3F by this method showed that it was present in the wines analyzed (red and white Bordeaux, Loire Valley Sauvignon blanc, white Burgundy, and Champagne) at concentrations up to 100 ng/L, i.e., significantly above the olfactory perception threshold for this compound in model dilute alcohol solution.     

Changes in volatile compounds and overall aroma profile during storage of coffee brews at 4 and 25 degrees C

In this work, the chemical changes occurring in the volatile fraction of Arabica coffee brews during storage at 4 and 25 degrees C for 30 days have been characterized for the first time by means of HS-GC-MS. A total of 47 compounds were identified and quantified: 2 sulfur compounds, 7 aldehydes, 3 esters, 15 furans, 5 ketones, 1 alcohol, 2 thiophenes, 4 pyrroles, 1 pyridine, 5 pyrazines, 1 alkene, and 1 acid. No new volatile compounds were detected at the end of the storage time. The changes observed are, in general, slower and less pronounced at refrigeration temperature. Storage also affects the sensory characteristics of the stored coffee brews, which lose part of their aroma intensity and freshness, acquiring some nondesirable notes such as rancid aroma, mainly during storage at 25 degrees C. Furthermore, seven aroma indices have been proposed as indicators of coffee brew staling, which show a good correlation with some sensory attributes, not only for aroma but also overall sensory quality. Consequently, they could be considered useful to monitor both the "age" and the sensory quality of stored coffee brews.     

Screening of raw coffee for thiol binding site precursors using "in bean" model roasting experiments

The purpose of the following study was to investigate the influence of coffee roasting on the thiol-binding activity of coffee beverages, and to investigate the potential of various green bean compounds as precursors of thiol-binding sites by using promising "in bean" model roast experiments. Headspace gas chromatographic analysis on coffee brews incubated in the presence of the roasty-sulfury smelling 2-furfurylthiol for 20 min at 30 degrees C in septum-closed vessels revealed that the amounts of "free" thiol decreased drastically with increasing the roasting degree of the beans used for preparation of the brews. A half-maximal binding capacity (BC(50)) of 183 mg of 2-furfurylthiol per liter of standard coffee beverage was determined for a roasted coffee (CTN value of 67), thus demonstrating that enormous amounts of the odor-active thiol are "bound" by the coffee. Furthermore, biomimetic "in bean" precursor experiments have been performed in order to elucidate the precursor for the thiol-binding sites in the raw coffee bean. These experiments opened the possibility of studying coffee model reactions under quasi-natural roasting conditions and undoubtedly identified chlorogenic acids as well as thermal degradation products caffeic acid and quinic acid as important precursors for low-molecular-weight thiol-binding sites. In particular, when roasted in the presence of transition metal ions, chlorogenic acids and even more caffeic acid showed thiol-binding activity which was comparable to the activity measured for the authentic coffee brew.     

A powerful aromatic volatile thiol, 2-furanmethanethiol, exhibiting roast coffee aroma in wines made from several Vitis vinifera grape varieties

The chemical compound 2-furanmethanethiol (2FM), with a strong roast coffee aroma, has been identified in sweet white wines made from the Petit manseng grape variety, and in certain red Bordeaux wines (made from the Merlot, Cabernet franc, and Cabernet sauvignon grape varieties). This was done by extracting specific volatile thiols using p-hydroxymercuribenzoate. The 2FM has also been found in toasted oak used in barrel-making. All the Petit manseng sweet white wines and some of the red Bordeaux wines analyzed contained between a few ng/L and several dozen ng/L of 2FM. Taking into account its very low perception threshold (0.4 ng/L in a model hydro alcoholic environment), 2FM could therefore contribute to the roast coffee aroma of certain wines.     

Modeling the effect of water activity and storage temperature on chemical stability of coffee brews

This work was addressed to study the chemical stability of coffee brew derivatives as a function of water activity (aw) and storage temperature. To this purpose, coffee brew was freeze-dried, equilibrated at increasing aw values, and stored for up to 10 months at different temperatures from -30 to 60 degrees C. The chemical stability of the samples was assessed by measuring H3O+ formation during storage. Independently of storage temperature, the rate of H3O+ formation was considerably low only when aw was reduced below 0.5 (94% w/w). Beyond this critical boundary, the rate increased, reaching a maximum value at ca. 0.8 aw (78% w/w). Further hydration up to the aw of the freshly prepared beverage significantly increased chemical stability. It was suggested that mechanisms other than lactones' hydrolysis, probably related to nonenzymatic browning pathways, could contribute to the observed increase in acidity during coffee staling. The temperature dependence of H3O+ formation was well-described by the Arrhenius equation in the entire aw range considered. However, aw affected the apparent activation energy and frequency factor. These effects were described by simple equations that were used to set up a modified Arrhenius equation. This model was validated by comparing experimental values, not used to generate the model, with those estimated by the model itself. The model allowed efficient prediction of the chemical stability of coffee derivatives on the basis of only the aw value and storage temperature.     

Incorporation of chlorogenic acids in coffee brew melanoidins

The incorporation of chlorogenic acids (CGAs) and their subunits quinic and caffeic acids (QA and CA) in coffee brew melanoidins was studied. Fractions with different molecular weights, ionic charges, and ethanol solubilities were isolated from coffee brew. Fractions were saponified, and the released QA and CA were quantified. For all melanoidin fractions, it was found that more QA than CA was released. QA levels correlated with melanoidin levels, indicating that QA is incorporated in melanoidins. The QA level was correlated with increasing ionic charge of the melanoidin populations, suggesting that QA may contribute to the negative charge and consequently is, most likely, not linked via its carboxyl group. The QA level correlated with the phenolic acid group level, as determined by Folin-Ciocalteu, indicating that QA was incorporated to a similar extent as the polyphenolic moiety from CGA. The QA and CA released from brew fractions by enzymes confirmed the incorporation of intact CGAs. Intact CGAs are proposed to be incorporated in melanoidins upon roasting via CA through mainly nonester linkages. This complex can be written as Mel=CA-QA, in which Mel represents the melanoidin backbone, =CA represents CA nonester-linked to the melanoidin backbone, and -QA represents QA ester-linked to CA. Additionally, a total of 12% of QA was identified in coffee brew, whereas only 6% was reported in the literature so far. The relevance of the additional QA on coffee brew stability is discussed.     

Headspace GC and sensory analysis characterization of the influence of different milk additives on the flavor release of coffee beverages

Previous investigations of coffee flavor have been confined to the analysis of the aroma substances. These investigations showed that about 30 volatile compounds were substantially responsible for the coffee flavor. The aim of this study was to investigate the influence of different milk additives and one coffee whitener on the release of flavor impact compounds from coffee beverages. For the investigation of these effects an external static headspace technique was developed. With this technique the most potent odorants of the coffee beverage were determined. Analyses were performed by gas chromatography/olfactometry, flame ionization detection, and mass spectrometric detection. In addition, sensory studies of the odor profiles were performed. Milk and vegetable products as additives for coffee beverages affected the release of aroma substances in the brew through their lipid, protein, and carbohydrate components. All beverages with an additive showed reduced, but typical, odor profiles for each additive.     

Solid-phase microextraction method development for headspace analysis of volatile flavor compounds

Solid-phase microextraction (SPME) fibers were evaluated for their ability to adsorb volatile flavor compounds under various conditions with coffee and aqueous flavored solutions. Experiments comparing different fibers showed that poly(dimethylsiloxane)/divinylbenzene had the highest overall sensitivity. Carboxen/poly(dimethylsiloxane) was the most sensitive to small molecules and acids. As the concentrations of compounds increased, the quantitative linear range was exceeded as shown by competition effects with 2-isobutyl-3-methoxypyrazine at concentrations above 1 ppm. A method based on a short-time sampling of the headspace (1 min) was shown to better represent the equilibrium headspace concentration. Analysis of coffee brew with a 1-min headspace adsorption time was verified to be within the linear range for most compounds and thus appropriate for relative headspace quantification. Absolute quantification of volatiles, using isotope dilution assays (IDA), is not subject to biases caused by excess compound concentrations or complex matrices. The degradation of coffee aroma volatiles during storage was followed by relative headspace measurements and absolute quantifications. Both methods gave similar values for 3-methylbutanal, 4-ethylguaiacol, and 2,3-pentanedione. Acetic acid, however, gave higher values during storage upon relative headspace measurements due to concurrent pH decreases that were not seen with IDA.     

Arabinogalactan proteins are incorporated in negatively charged coffee brew melanoidins

The charge properties of melanoidins in high molecular weight (HMw) coffee brew fractions, isolated by diafiltration and membrane dialysis, were studied. Ion exchange chromatography experiments with the HMw fractions showed that coffee brew melanoidins were negatively charged whereas these molecules did not expose any positive charge at the pH of coffee brew. Fractions with different ionic charges were isolated and subsequently characterized by means of the specific extinction coefficient (K(mix 405nm)), sugar composition, phenolic group content, nitrogen content, and the arabinogalactan protein (AGP) specific Yariv gel-diffusion assay. The isolated fractions were different in composition and AGP was found to be present in one of the HMw fractions. The AGP accounted for 6% of the coffee brew dry matter and had a moderate negative charge, probably caused by the presence of uronic acids. As the fraction that precipitated with Yariv was brown (K(mix 405nm) = 1.2), compared to a white color in the green bean, it was concluded that these AGPs had undergone Maillard reaction resulting in an AGP-melanoidin complex. The presence of mannose (presumably from galactomannan) indicates the incorporation of galactomannans in the AGP-melanoidin complex. As the uronic acid content in the more negatively charged melanoidin-rich, AGP-poor HMw fractions decreased, it was hypothesized that acidic groups are formed or incorporated during melanoidin formation.     

Identification of odor-active 3-mercapto-3-methylbutyl acetate in volatile fraction of roasted coffee brew isolated by steam distillation under reduced pressure

In a roasted Arabica coffee brew, the potent roasty odor quality compound was identified as 3-mercapto-3-methylbutyl acetate by comparison of its Kovats gas chromatography retention index, mass spectrum, and odor quality to those of the synthetic authentic compound. 3-Mercapto-3-methylbutyl acetate has been identified for the first time in the coffee, and according to the results of the aroma extract dilution analysis, the contribution of this compound to the flavor of the roasted coffee brew varied depending on the degree of the coffee bean roasting. The concentration of this compound in the coffee brews as with 3-mercapto-3-methylbutyl formate increased with an increase in the degree of roasting. However, the slope of the amount of both esters was different, and 3-mercapto-3-methylbutyl acetate hardly increased with a low degree of roasting at more than a 21 luminosity (L)-value, but it rapidly increased when the roasting degree of the coffee beans reached the L-value of 18. These results suggested that the contribution of 3-mercapto-3-methylbutyl acetate to the overall flavor is peculiar to the flavor of the highly roasted coffee.     

Low molecular weight melanoidins in coffee brew

Analysis of low molecular weight (LMw) coffee brew melanoidins is challenging due to the presence of many non-melanoidin components that complicate analysis. This study focused on the isolation of LMw coffee brew melanoidins by separation of melanoidins from non-melanoidin components that are present in LMw coffee brew material. LMw coffee fractions differing in polarity were obtained by reversed-phase solid phase extraction and their melanoidin, sugar, nitrogen, caffeine, trigonelline, 5-caffeoylquinic acid, quinic acid, caffeic acid, and phenolic groups contents were determined. The sugar composition, the charge properties, and the absorbance at various wavelengths were investigated as well. The majority of the LMw melanoidins were found to have an apolar character, whereas most non-melanoidins have a polar character. The three isolated melanoidin-rich fractions represented 56% of the LMw coffee melanoidins and were free from non-melanoidin components. Spectroscopic analysis revealed that the melanoidins isolated showed similar features as high molecular weight coffee melanoidins. All three melanoidin fractions contained approximately 3% nitrogen, indicating the presence of incorporated amino acids or proteins. Surprisingly, glucose was the main sugar present in these melanoidins, and it was reasoned that sucrose is the most likely source for this glucose within the melanoidin structure. It was also found that LMw melanoidins exposed a negative charge, and this negative charge was inversely proportional to the apolar character of the melanoidins. Phenolic group levels as high as 47% were found, which could be explained by the incorporation of chlorogenic acids in these melanoidins.     

Determination of polycyclic aromatic hydrocarbons in coffee brew using solid-phase extraction

The presence of polycyclic aromatic hydrocarbons (PAHs) in coffee has been reported and is suspected to be due to the degradation of coffee compounds during the roasting step. Due to the high toxicity of these compounds, among which benzo[a]pyrene is known to be the most carcinogenic, their presence in the coffee, especially the coffee brew that is directly ingested by the consumer, is of prime importance. However, due to the low solubility of these compounds, their concentrations are expected to be rather low. As a consequence, reliable and sensitive analytical methods are required. The aim of this study was to develop a reliable and fast analytical procedure to determine these organic micropollutants in coffee brew samples. PAHs were retained on a 0.5 g polystyrene-divinylbenzene cartridge before being eluted by a mixture of methanol/tetrahydrofuran (10:90 v/v), concentrated, and directly analyzed by reversed-phase high-performance liquid chromatography coupled to a fluorescence detector. Application to the determination of PAHs in several coffee brew samples is also given, with mean estimated concentrations in the range of 0-100 ng L(-1) for suspected benzo[b]fluoranthene and benzo[a]pyrene, whereas no fluoranthene could be detected. Tentative identification was made on the basis of UV spectra. However, identification of the suspected traces of PAHs could not be achieved due to matrix effects, so that the presence of coeluting compounds may not be excluded.     

Oxidation of porcine Myosin by hypervalent myoglobin: the role of thiol groups

Oxidation of the myofibrillar muscle protein myosin from pork by hypervalent myoglobin species (MbFe(III)/H 2O2 radical generating system) was investigated in aqueous solution in the pH range of 5.0-7.8 by electron spin resonance (ESR) spectroscopy using N- tert-butyl-alpha-phenylnitrone (PBN) as spin trap and indirectly by determination of the rate of reduction of hypervalent myoglobin species by UV spectroscopy. Cross-linking of myosin was examined by SDS-PAGE. The target for oxidative modification of myosin was studied by thiol blocking by N-acetylmaleimide (NEM) and by determining oxidative modification of myosin thiols. The reaction between myosin and hypervalent myoglobin was fast and showed little dependence on pH. The myosin radicals formed were observed to be short-lived. Myosin thiols are suggested to be the main target for oxidative modification, as NEM-treated myosin did not form radicals in the presence of hypervalent myoglobin. A significant decrease in thiol content was already demonstrated 25 s after initiation of oxidation of myosin. The majority of myosin heavy chain (MHC) was demonstrated to be cross-linked through intermolecular disulfide bonding 1 h after initiation of oxidation. This demonstrates that thiols are important for radical formation and cross-linking of myosin during oxidation with hypervalent myoglobin at the pH of meat products.     

Influence of malt roasting on the oxidative stability of sweet wort

Influence of malt roasting on the oxidative stability of sweet wort was evaluated based on radical intensity, volatile profile, content of transition metals (Fe and Cu) and thiols. Malt roasting had a large influence on the oxidative stability of sweet wort. Light sweet worts were more stable with low radical intensity, low Fe content, and ability to retain volatile compounds when heated. At mild roasting, the Fe content in the wort increased but remained close to constant with further roasting. Dark sweet worts were less stable with high radical intensities, high Fe content, and a decreased ability to retain volatiles. Results suggested that the Maillard reaction compounds formed during the roasting of malt are prooxidants in sweet wort. A thiol-removing capacity was observed in sweet wort, and it was gradually inhibited by malt roasting. It is possibly caused by thiol oxidizing enzymes present in the fresh malt.     

Olfactory perception of cysteine-S-conjugates from fruits and vegetables

Volatile sulfur compounds have a low odor threshold, and their presence at microgram per kilogram levels in fruits and vegetables influences odor quality. Sensory analysis demonstrates that naturally occurring, odorless cysteine- S-conjugates such as S-( R/ S)-3-(1-hexanol)- l-cysteine in wine, S-(1-propyl)- l-cysteine in onion, and S-(( R/ S)-2-heptyl)- l-cysteine in bell pepper are transformed into volatile thiols in the mouth by microflora. The time delay in smelling these volatile thiols was 20-30 s, and persistent perception of their odor occurred for 3 min. The cysteine- S-conjugates are transformed in free thiol by anaerobes. The mouth acts as a reactor, adding another dimension to odor perception, and saliva modulates flavors by trapping free thiols.