Evaluation of key aroma compounds in hand-squeezed grapefruit juice (Citrus paradisi Macfayden) by quantitation and flavor reconstitution experiments

Twenty-five odor-active compounds were quantified in the fresh, hand-squeezed juice of White Marsh seedless grapefruits using stable isotope dilution assays. By calculation of the odor activity values of the odorants (ratio of their concentrations in the juice to their odor thresholds in water) it was shown that the fruity esters ethyl 2-methylpropanoate, ethyl butanoate, and (S)-ethyl 2-methylbutanoate, and the fruity, sweet winelactone, as well as the grassy smelling (Z)-hex-3-enal, and trans-4,5-epoxy-(E)-dec-2-enal with metallic odor, were among the most potent odorants of the fresh grapefruit juice. The typical sulfurous, grapefruit-like odor quality was mainly due to the catty, blackcurrant-like 4-mercapto-4-methylpentan-2-one and the grapefruit-like smelling 1-p-menthene-8-thiol. These findings were confirmed by reconstitution experiments to simulate the aroma of the fresh grapefruit juice.     

Evaluation of aroma differences between hand-squeezed juices from Valencia late and Navel oranges by quantitation of key odorants and flavor reconstitution experiments

Twenty-five odor-active compounds were quantified in hand-squeezed juices of Valencia late and Navel oranges using stable isotope dilution assays. Odor activity values (OAVs, ratio of the concentration to odor thresholds) based on odor thresholds in water were calculated for the entire set of aroma compounds in both varieties. It was shown that due to their high OAVs, the fruity-smelling esters ethyl 2-methylpropanoate, ethyl butanoate, (S)-ethyl 2-methylbutanoate, and 3a,4,5,7a-tetrahydro-3,6-dimethyl-2(3H)-benzofuranone (wine lactone), the grassy smelling (Z)-hex-3-enal, and the citrus-like decanal were the most potent odorants in both juices. The weaker fruity note in the Navel oranges was clearly correlated with significantly lower OAVs of all fruity-smelling esters but a higher OAV of (Z)-3-hexenal compared to Valencia late. Model solutions simulating the odor of both orange varieties confirmed the findings of the quantitation studies.     

Critical comparison of three olfactometric methods for the identification of the most potent odorants in cooked mussels (Mytilus edulis)

Three olfactometric methods (olfactometry global analysis, OSME, and AEDA) were compared to evaluate the main impact odorants of cooked mussels. The results obtained from these methods were very similar and well correlated. On the basis of the three techniques, 42 odor-active compounds were detected and 28 were identified. Among these compounds, 6 odorants seem to contribute actively to the aroma of mussels: 2,3-butanedione (4) (buttery, caramel-like odor), (Z)-4-heptenal (14) (boiled potato-like odor), (E)-2-penten-1-ol (17) (mushroom-like odor), 2-ethylpyrazine (19) (nutty odor), methional (25) (boiled potato-like odor), and (E,E)-2,4-octadienal (32) (cucumber-like odor).     

Similarities in the aroma chemistry of Gewürztraminer variety wines and lychee (Litchi chinesis sonn.) fruit

GC/O analysis of canned lychees indicated that cis-rose oxide, linalool, ethyl isohexanoate, geraniol, furaneol, vanillin, (E)-2-nonenal, beta-damascenone, isovaleric acid, and (E)-furan linalool oxide were the most odor potent compounds detected in the fruit extracts. However, on the basis of calculated odor activity values (OAVs), cis-rose oxide, beta-damascenone, linalool, furaneol, ethyl isobutyrate, (E)-2-nonenal, ethyl isohexanoate, geraniol, and delta-decalactone were determined to be the main contributors of canned lychee aroma. When these results were compared with GC/O results of fresh lychees and Gewürztraminer wine, 12 common odor-active volatile compounds were found in all three products. These included cis-rose oxide, ethyl hexanoate/ethyl isohexanoate, beta-damascenone, linalool, ethyl isobutyrate, geraniol, ethyl 2-methylbutyrate, 2-phenylethanol, furaneol, vanillin, citronellol, and phenethyl acetate. On the basis of OAVs, cis-rose oxide had the highest values among the common odorants in the three products, indicating its importance to the aroma of both lychee fruit and Gewürztraminer wines. Other compounds that had significant OAVs included beta-damascenone, linalool, furaneol, ethyl hexanoate, and geraniol. This indicated that while differences exist in the aroma profile of lychee and Gewürztraminer, the common odorants detected in both fruit and wine, particularly cis-rose oxide, were responsible for the lychee aroma in Gewürztraminer wine. When headspace SPME was used as a rapid analytical tool to detect the levels of selected aroma compounds deemed important to lychee aroma in Gewürztraminer-type wines, cis-rose oxide, linalool, and geraniol were found to be at relatively higher levels in Gewürztraminers. No cis-rose oxide was detected in the control wines (Chardonnay and Riesling), while lower levels were detected in the Gewürztraminer-hybrid wine Traminette. Gewürztraminers produced in the Alsace region showed differences in the levels of the 3 monoterpenes when compared to those from New York State, which could be attributed to differences in viticultural and enological practices between regions.     

Identification of aroma-active compounds in Jiashi muskmelon juice by GC-O-MS and OAV calculation

To identify aromatic compounds in Jiashi melon juice, gas chromatography-mass spectrometry-olfactometry (GC-MS-O) analysis was used. Odor activity values (OAVs) were also calculated on the basis of the qualitative and quantitative analysis of volatile compounds. Results showed that 42 volatiles were identified, among which 4 compounds, namely, diethyl carbonate, isophorone, 2-butoxyethyl acetate, and menthol, were identified or tentatively identified for the first time as volatiles in melon fruit. Twelve compounds, namely, (2E,6Z)-nona-2,6-dienal, (3Z,6Z)-nona-3,6-dien-1-ol, ethyl butanoate, ethyl 2-methylbutyrate, ethyl 2-methylpropanoate, (Z)-non-6-enal, (E)-2-nonenal, heptanal, methyl 2-methylbutyrate, nonanal, hexanal, and 2-methylpropyl acetate, were identified as the potent odorants of Jiashi melon juice by both OAV and detection frequency analysis (DFA). In addition, seven odorants were detected by all of the panelists and showed higher OAVs, indicating that DFA and OAV resulted in relatively similar "Jiashi" melon aroma patterns.     

Identification of potent odorants in Japanese green tea (Sen-cha)

Application of aroma extract dilution analysis using the volatile fraction of a Japanese green tea (Sen-cha) sample resulted in the detection of 36 odor-active peaks with flavor dilution (FD) factors between 10 and 5000. Thirty-six potent odorants were identified from 36 odor-active peaks by gas chromatography/mass spectrometry (GC/MS) and/or the multidimensional GC/MS (MDGC/MS) system. Among these components, 4-methoxy-2-methyl-2-butanethiol (meaty), (Z)-1, 5-octadien-3-one (metallic), 4-mercapto-4-methyl-2-pentanone (meaty), (E,E)-2,4-decadienal (fatty), beta-damascone (honey-like), beta-damascenone (honey-like), (Z)-methyl jasmonate (floral), and indole (animal-like) showed the highest FD factors. Therefore, these odorants were the most important components of the Japanese green tea odor. In addition, 4-methoxy-2-methyl-2-butanethiol, 4-mercapto-4-methyl-2-pentanone, methional, 2-ethyl-3, 5-dimethylpyrazine, (Z)-4-decenal, beta-damascone, maltol, 5-octanolide, 2-methoxy-4-vinylphenol, and 2-aminoacetophenone were newly identified compounds in the green tea.     

Comparison of the most odor-active compounds in fresh and dried hop cones (Humulus lupulus L. variety spalter select) based on GC-olfactometry and odor dilution techniques

Application of aroma extract dilution analysis on the volatiles obtained from dried cones of Spalter Select hops grown in the German hop-growing area of Hallertau revealed 23 odorants in the flavor dilution (FD) factor range of 16-4096, 20 of which could be identified. On the basis of high FD factors, trans-4, 5-epoxy-(E)-2-decenal, linalool, and myrcene were identified as the most potent odorants, followed by ethyl 2-methylpropanoate, methyl 2-methylbutanoate, (Z)-1,5-octadien-3-one, nonanal, (E,Z)-1,3, 5-undecatriene, 1,3(E),5(Z),9-undecatetraene, propyl 2-methylbutanoate, 4-ethenyl-2-methoxyphenol, and 1-octen-3-one. Ten of the high-impact hop aroma compounds had previously not been identified as hop constituents and, in particular, 1,3(E),5(Z), 9-undecatetraene has not yet been reported as a food odorant. In an extract obtained from fresh hops, in addition to the odorants found in dry hops, (Z)-3-hexenal was characterized as a further key odorant rendering an additional green aroma note to the fresh material.     

Quantitative gas chromatography-olfactometry carried out at different dilutions of an extract. Key differences in the odor profiles of four high-quality Spanish aged red wines

Four Spanish aged red wines made in different wine-making areas have been extracted, and the extracts and their 1:5, 1:50, and 1:500 dilutions have been analyzed by a gas chromatography-olfactometry (GC-O) approach in which three judges evaluated odor intensity on a four-point scale. Sixty-nine different odor regions were detected in the GC-O profiles of wines, 63 of which could be identified. GC-O data have been processed to calculate averaged flavor dilution factors (FD). Different ANOVA strategies have been further applied on FD and on intensity data to check for significant differences among wines and to assess the effects of dilution and the judge. Data show that FD and the average intensity of the odorants are strongly correlated (r(2) = 0.892). However, the measurement of intensity represents a quantitative advantage in terms of detecting differences. For some odorants, dilution exerts a critical role in the detection of differences. Significant differences among wines have been found in 30 of the 69 odorants detected in the experiment. Most of these differences are introduced by grape compounds such as methyl benzoate and terpenols, by compounds released by the wood, such as furfural, (Z)-whiskey lactone, Furaneol, 4-propylguaiacol, eugenol, 4-ethylphenol, 2,6-dimethoxyphenol, isoeugenol, and ethyl vanillate, by compounds formed by lactic acid bacteria, such as 2,3-butanedione and acetoine, or by compounds formed during the oxidative storage of wines, such as methional, sotolon, o-aminoacetophenone, and phenylacetic acid. The most important differences from a quantitative point of view are due to 2-methyl-3-mercaptofuran, 4-propylguaiacol, 2,6-dimethoxyphenol, and isoeugenol.     

Impact odorants contributing to the fungus type aroma from grape berries contaminated by powdery mildew (Uncinula necator); incidence of enzymatic activities of the yeast Saccharomyces cerevisiae

Powdery mildew due to the fungus Uncinula necator is an important disease for the vineyard. The development of the fungus at the surface of the berries leads to the occurrence of a very characteristic and sometimes intense mushroom-type odor cited as an important default for grapes quality. Gas chromatography/olfactometry, gas chromatography, and multidimensional gas chromatogaphy/mass spectrometry techniques were used to investigate the most important odorants of grapes diseased by powdery mildew. Among 22 odorants detected, strongly odorant compounds were identified or tentatively identified in purified extracts obtained from grapes diseased by powdery mildew. Aroma extraction dilution analysis (AEDA) analysis revealed that 1-octen-3-one (mushroom odor), (Z)-1,5-octadien-3-one (geranium-leaf odor), and an unidentified odorous zone (fishy-mushroom like odor) were the most potent volatiles of the diseased grapes. In the presence of nonproliferating Saccharomyces cerevisiae yeast cells, and consequently during alcoholic fermentation, the enzymatic reduction of 1-octen-3-one and (Z)-1,5-octadien-3-one to much less odorant compounds, namely 3-octanone and (Z)-5-octen-3-one, was shown. Those results explain to some extent the disappearance of the fungal aroma specific to powdery mildew grapes during alcoholic fermentation.     

Characterization of the most odor-active compounds of Iberian ham headspace

Gas chromatography-olfactometry (GC-O) based on detection frequency (DF) was used to characterize the most odor-active compounds from the headspace of Iberian ham. Twenty-eight odorants were identified by GC-O on two capillary columns, including aldehydes (11), sulfur-containing compounds (7), ketones (5), nitrogen-containing compounds (2), esters (2), and an alcohol. Among them, the highest odor potencies (DF values) were found for 2-methyl-3-furanthiol, 2-heptanone, 3-methylbutanal, methanethiol, hexanal, hydrogen sulfide, 1-penten-3-one, 2-methylpropanal, ethyl 2-methylbutyrate, and (E)-2-hexenal. Nine of the 28 most odor-active compounds were identified for the first time as aroma components of dry-cured ham, including hydrogen sulfide, 1-penten-3-one, (Z)-3-hexenal, 1-octen-3-one, and the meaty-smelling compounds 2-methyl-3-furanthiol, 2-furfurylthiol, 3-mercapto-2-pentanone, 2-acetyl-1-pyrroline, and 2-propionyl-1-pyrroline.     

Odor potency of aroma compounds in Riesling and Vidal blanc table wines and icewines by gas chromatography-olfactometry-mass spectrometry

This study aimed to elucidate the odor potency of aroma compounds in Riesling and Vidal blanc (syn. Vidal) table wines and icewines from the Niagara Peninsula using stir bar sorptive extraction-gas chromatography-olfactometry-mass spectrometry. Dilution analysis determined the most odor-potent compounds in Vidal and Riesling icewines (n = 2) and table wines (n = 2) from a commercial producer. The top 15 odor-potent compounds in each wine were identified and quantified, resulting in 23 and 24 compounds for Riesling and Vidal, respectively. The most odor-potent compounds were β-damascenone, decanal, 1-hexanol, 1-octen-3-ol, 4-vinylguaiacol, ethyl hexanoate, and ethyl 3-methylbutyrate. In general, icewines had higher concentrations of most aroma compounds compared to table wines. Through computation of odor activity values, the compounds with the highest odor activity for the icewines were β-damascenone, 1-octen-3-ol, ethyl octanoate, cis-rose oxide, and ethyl hexanoate. In table wines the highest odor activity values were found for ethyl octanoate, β-damascenone, ethyl hexanoate, cis-rose oxide, ethyl 3-methylbutyrate, and 4-vinylguaiacol. These findings provide a foundation to determine impact odorants in icewines and the effects of viticultural and enological practices on wine aroma volatile composition.     

Aroma components of cooked tail meat of American lobster (Homarus americanus)

Key aroma components of cooked tail meat of American lobster (Homarus americanus) were studied by gas chromatography-olfactometry (GCO) techniques. Components of low and intermediate volatility were evaluated by aroma extract dilution analysis of solvent extracts prepared by direct solvent extraction-high vacuum distillation and vacuum steam distillation-solvent extraction, whereas headspace volatile components were assessed by GCO of decreasing headspace (static and dynamic modes) samples. Forty-seven odorants were detected by all techniques. 3-Methylbutanal (chocolate, malty), 2,3-butanedione (buttery), 3-(methylthio)propanal (cooked potato), 1-octen-3-one (mushroom), 2-acetyl-1-pyrroline (popcorn), and (E,Z)-2,6-nonadienal (cucumber), were identified as predominant odorants by all four isolation methods. The highly volatile compounds methanethiol (rotten, sulfurous) and dimethyl sulfide (canned corn) were detected by headspace methods only. These eight odorants along with three unknown compounds with crabby, amine, fishy odors were found to predominate in the overall aroma of cooked lobster tail meat.     

Identification of characteristic aroma components of Thai fried chili paste

Three forms of Thai fried chili pastes (CP) were prepared, consisting of an unheated CP (UH-CP), a CP heated at 100 degrees C for 25 min (H25-CP, typical product), and a CP excessively heated for 50 min (H50-CP). The potent odorants in the CPs were investigated by two gas chromatography-olfactometry methods: dynamic headspace dilution analysis (DHDA) and aroma extract dilution analysis (AEDA). DHDA revealed that the predominant odorants in heated CPs were mainly sulfur-containing compounds, followed by lipid-derived compounds, Strecker aldehydes, and Maillard reaction products. Dimethyl sulfide, allyl mercaptan, 2- (or 3-) methylbutanal, ally methyl sulfide, 2,3-butanedione, 3,3'-thiobis(1-propene), and methyl propyl disulfide were among the most potent headspace odorants detected by DHDA. By AEDA, 2-vinyl-4 H-1,3-dithiin and diallyl trisulfide had the highest FD factors in H25-CP. On the basis of their high FD factors by both GCO methods, the predominant odorants in H25-CP were 3-vinyl-4 H-1,2-dithiin, allyl methyl disulfide, and allyl methyl trisulfide. Furthermore, dimethyl trisulfide and diallyl disulfide had the highest odor activity values in H25-CP, suggesting that these were also potent odorants in CP. In addition, methional, 3-methylbutanoic acid, 4-hydroxy-2,5-dimethyl-3-(2 H)-furanone, and 3-hydroxy-4,5-dimethyl-2( 5H)-furanone (sotolon) were indicated as potent thermally derived odorants of H25-CP.     

Gas chromatography-olfactometry and chemical quantitative study of the aroma of six premium quality spanish aged red wines

The aroma of six premium quality Spanish red wines has been studied by quantitative gas chromatography-olfactometry (GC-O) and techniques of quantitative chemical analysis. The GC-O study revealed the presence of 85 aromatic notes in which 78 odorants were identified, two of which-1-nonen-3-one (temptatively) and 2-acetylpyrazine-are reported in wine for the first time. Forty out of the 82 quantified odorants may be present at concentrations above their odor threshold. The components with the greatest capacity to introduce differences between these wines are ethyl phenols produced by Brettanomyces yeasts (4-ethylphenol, 4-ethyl-2-methoxyphenol, and 4-propyl-2-methoxyphenol), 2,5-dimethyl-4-hydroxy-3(2H)-furanone (furaneol), (Z)-3-hexenol, thiols derived from cysteinic precursors (4-methyl-4-mercaptopentan-2-one, 3-mercaptohexyl acetate, and 3-mercaptohexanol), some components yielded by the wood [(E)-isoeugenol, 4-allyl-2-methoxyphenol, vanillin, 2-methoxyphenol (guaiacol), and (Z)-whiskylactone], and compounds related to the metabolism (2-phenylethanol, ethyl esters of isoacids, 3-methylbutyl acetate) or oxidative degradation of amino acids [phenylacetaldehyde and 4,5-dimethyl-3-hydroxy-2(5H)-furanone (sotolon)]. The correlation between the olfactometric intensities and the quantitative data is, in general, satisfactory if olfactometric differences between the samples are high. However, GC-O fails in detecting quantitative differences in those cases in which the olfactive intensity is very high or if odors elute in areas in which the odor chromatogram is too complex.     

Quantification and sensory studies of character impact odorants of different soybean lecithins.

Fifty-four potent odorants in standardized, hydrolyzed, and deoiled and hydrolyzed soybean lecithins were quantified by high-resolution gas chromatography/mass spectrometry (HRGC/MS). The characterization of their aroma impact was performed by calculation of nasal (n) and retronasal (r) odor activity values (OAVs). For this, the nasal and retronasal recognition thresholds of 18 odor-active compounds were determined in vegetable oil. The following compounds showed the highest nOAVs: 2,3-diethyl-5-methylpyrazine, methylpropanal, acetic acid, pentanoic acid, 2-ethyl-3,5-dimethylpyrazine, pentylpyridine, (Z)-1,5-octadien-3-one, 2-methylbutanal, and beta-damascenone. In addition to the compounds above, 1-octen-3-one, 1-nonen-3-one, and 3-methyl-2,4-nonandione showed potent rOAVs. The results of quantification and OAV calculation were confirmed by a model mixture of 25 impact odorants, which yielded a highly similar sensory profile to that of the original soybean lecithin. The sensory importance of pyrazines and free acids increased through enzymatic hydrolysis and decreased by the process of deoiling. The impact of unsaturated ketones on the lecithin aroma was not changed by either process.     

Characterization of the most odor-active compounds in an American Bourbon whisky by application of the aroma extract dilution analysis

Application of the aroma extract dilution analysis (AEDA) on the volatile fraction carefully isolated from an American Bourbon whisky revealed 45 odor-active areas in the flavor dilution (FD) factor range of 32-4096 among which (E)-beta-damascenone and delta-nonalactone showed the highest FD factors of 4096 and 2048, respectively. With FD factors of 1024, (3S,4S)-cis-whiskylactone, gamma-decalactone, 4-allyl-2-methoxyphenol (eugenol), and 4-hydroxy-3-methoxy-benzaldehyde (vanillin) additionally contributed to the overall vanilla-like, fruity, and smoky aroma note of the spirit. Application of GC-Olfactometry on the headspace above the whisky revealed 23 aroma-active odorants among which 3-methylbutanal, ethanol, and 2-methylbutanal were identified as additional important aroma compounds. Compared to published data on volatile constituents in whisky, besides ranking the whisky odorants on the basis of their odor potency, 13 aroma compounds were newly identified in this study: ethyl (S)-2-methylbutanoate, (E)-2-heptenal, (E,E)-2,4-nonadienal, (E)-2-decenal, (E,E)-2,4-decadienal, 2-isopropyl-3-methoxypyrazine, ethyl phenylacetate, 4-methyl acetophenone, alpha-damascone, 2-phenylethyl propanoate, 3-hydroxy-4,5-dimethyl-2(5H)-furanone, trans-ethyl cinnamate, and (Z)-6-dodeceno-gamma-lactone.     

Potent odorants of raw Arabica coffee. Their changes during roasting

Aroma extract dilution analysis of raw Arabica coffee revealed 3-isobutyl-2-methoxypyrazine (I), 2-methoxy-3,5-dimethylpyrazine (II), ethyl 2-methylbutyrate (III), ethyl 3-methylbutyrate (IV), and 3-isopropyl-2-methoxypyrazine (V) as potent odorants. The highest odor activity value was found for I followed by II, IV, and V. It was concluded that I was responsible for the characteristic, peasy odor note of raw coffee. Twelve odorants occurring in raw coffee and (E)-beta-damascenone were also quantified after roasting. The concentration of I did not change, whereas methional, 3-hydroxy-4, 5-dimethyl-2(5H)-furanone, vanillin, (E)-beta-damascenone, and 4-vinyl- and 4-ethylguaiacol increased strongly during the roasting process.     

Evaluation of the key aroma compounds in beef and pork vegetable gravies a la chef by stable isotope dilution assays and aroma recombination experiments

Although the aroma compounds of meat processed as such have been studied previously, data on complete homemade dishes containing beef and pork meat were scarcely studied. Recently, 38 odor-active compounds were characterized in beef and pork vegetable gravies using GC-olfactometry. In the present investigation, the most odor-active compounds were quantitated in a freshly prepared stewed beef vegetable gravy (BVG) as well as a stewed pork vegetable gravy (PVG) by means of stable isotope dilution assays. Calculation of odor activity values (OAVs; ratio of concentration to odor threshold) revealed 3-mercapto-2-methylpentan-1-ol, (E,E)-2,4-decadienal, (E,Z)-2,6-nonadienal, (E)-2-decenal, (E)-2-undecanal, and 3-hydroxy-4,5-dimethyl-2(5H)-furanone as the most potent odorants in both gravies. However, significantly different OAVs were found for 12-methyltridecanal, which was much higher in the BVG, whereas (E,Z)-2,4-decadienal showed a clearly higher OAV in the PVG. Aroma recombination experiments performed on the basis of the actual concentrations of the odorants in both gravies revealed a good similarity of the aromas of both model mixtures containing all odorants with OAVs > 1 with those of the original gravies.     

Characterization of the key aroma compounds in soy sauce using approaches of molecular sensory science

Application of aroma extract dilution analysis (AEDA) to the volatiles isolated from a commercial Japanese soy sauce revealed 30 odor-active compounds in the flavor dilution (FD) factor range of 8-4096, among which 2-phenylethanol showed the highest FD factor of 4096, followed by 3-(methylsulfanyl)propanal (methional), the tautomers 4-hydroxy-5-ethyl-2-methyl- and 4-hydroxy-2-ethyl-5-methyl-3(2H)-furanone (4-HEMF), 4-hydroxy-2,5-dimethyl-3(2H)-furanone (4-HDF), and 3-hydroxy-4,5-dimethyl-2(5H)-furanone (sotolone), all showing FD factors of 1024. Thirteen odorants were quantified by stable isotope dilution assays, and their odor activity values (OAVs) were calculated as ratio of their concentrations and odor thresholds in water. Among them, 3-methylbutanal (malty), sotolone (seasoning-like), 4-HEMF (caramel-like), 2-methylbutanal (malty), methional (cooked potato), ethanol (alcoholic), and ethyl 2-methylpropanoate (fruity) showed the highest OAVs (>200). An aqueous model aroma mixture containing 13 odorants, which had been identified with the highest OAVs, in concentrations that occur in the soy sauce showed a good similarity with the overall aroma of the soy sauce itself. Heat treatment of the soy sauce resulted in a clear change of the overall aroma. Quantitation of selected odorants revealed a significant decrease in sotolone and, in particular, increases in 2-acetyl-1-pyrroline, 4-HDMF, and 4-HEMF induced by heating.