Article pubs.acs.org/JAFC
Characterization of the Key Odorants in Chinese Zhima Aroma-Type Baijiu by Gas Chromatography−Olfactometry, Quantitative Measurements, Aroma Recombination, and Omission Studies Yang Zheng,†,‡ Baoguo Sun,‡ Mouming Zhao,† Fuping Zheng,‡ Mingquan Huang,*,‡ Jinyuan Sun,‡ Xiaotao Sun,‡ and Hehe Li‡ †
School of Food Science and Engineering, South China University of Technology, Guangzhou, China 510640 Beijing Key Laboratory of Flavor Chemistry, Beijing Technology and Business University, Beijing, China 100048
‡
ABSTRACT: Zhima aroma-type Baijiu with typical sesame aroma is particularly popular in northern China. To our knowledge, it is still uncertain which components are important to make contributions to its unique aroma, although a few pieces of research have reported many volatile compounds in this Baijiu. The aroma-active compounds from the Baijiu were researched in this paper. A total of 56 odorants were identified in Chinese Zhima aroma-type Baijiu by aroma extract dilution analysis (AEDA). Their odor activity values (OAVs) were determined by different quantitative measurements, and then 26 aroma compounds were further confirmed as important odorants due to their OAVs ≥ 1, and these had higher values, such as ethyl hexanoate (OAV 2691), 3-methylbutanal (2403), ethyl pentanoate (1019), and so on. The overall aroma of Zhima aroma-type Baijiu could be simulated by mixing of the 26 key odorants in their measured concentrations. The similarity of the overall aroma profiles between the recombination model and the commercial sample was judged to be 2.7 out of 3.0 points. Omission experiments further corroborated the importance of methional and ethyl hexanoate for the overall aroma of Chinese Zhima aroma-type Baijiu. KEYWORDS: Chinese Zhima aroma-type Baijiu, AEDA, OAV, aroma recombination, omission/addition experiments, methional, ethyl hexanoate
■
chromatography−mass spectrometry (GC-MS).7,8 The authors thought that methionol needed to be further confirmed as the characteristic aroma compound for Zhima aroma-type Baijiu, because it was not detected in all Zhima aroma-type Baijius. Besides, Sun et al. researched 18 Zhima aroma-type Baijiu samples, which were selected from five different companies, and they found there was no methionol in six Baijiu samples.9 Up to now, there are a number of volatile compounds identified in Zhima aroma-type Baijiu, but it is still a puzzling question which components are the most important or unique aroma-active compounds in Zhima aroma-type Baijiu. These aroma-active compounds in alcoholic beverages can be differentiated from the bulk of “non-active” volatiles by gas chromatography−olfactometry (GC-O) with aroma extract dilution analysis (AEDA).10−14 Important odorants can be further identified from alcoholic beverages by odor activity values (OAVs), and aroma reconstituted and omission experiments are being used to confirm the key odor-active constituents of alcoholic beverages, such as wine,15 brandy,16 and beer.17 The aim of this work was to identify important aroma compounds in Zhima aroma-type Baijiu by GC-O analysis and OAVs, and to confirm the key odorants of Zhima aroma-type Baijiu by aroma recombination and omission experiments.
INTRODUCTION Chinese Baijiu is a traditional distillate from grain fermentation and the most popular alcoholic beverage in China. Different Jiuqus, fermentation, and distillation processes generate a large number of different compounds in different Chinese Baijius. Based on aroma characteristics, Chinese Baijius could be classified into diverse aroma types according to their aroma characteristics, including soy sauce, strong, light, rice, Zhima, Chixiang, complex, herblike, Fengxiang, Laobaigan, and Texiang aroma-type Baijiu.1,2 Chinese Zhima aroma-type Baijiu is famous for its unique aroma combining fruity with sweaty, roasted sesame-like, and floral flavor, especially in northern China. Numerous investigations have been performed to identify volatile compounds in different aroma-types of Chinese Baijius, and today more than 1730 compounds have been reported.3 The volatile composition of Zhima aroma-type Baijiu is also quite complex. Hu and Lu had analyzed Zhima aroma-type Baijiu by GC-FPD and first reported that methionol was the characteristic aroma compound.4 Wu found that nitrogenous heterocyclic compounds were the main aroma components of Zhima aroma-type Baijiu, and the flavor profile of the liquor was just among but different from light, soy sauce, and strong aroma-type Baijiu. 5 Zheng and Zhao reported 179 volatile compounds in Zhima aroma-type Baijiu, including 54 esters, 51 alkanes, 20 acids, 19 alcohols, 17 aromatic compounds, 7 aldehydes, 7 furans, and 4 ketones.6 Dimethyl disulfide, dimethyl tetrasulfide, difurfuryl disulfide, and methionol were detected in Zhima aroma-type Baijius, which were identified by gas chromatography−olfactometry (GC-O) and gas © 2016 American Chemical Society
Received: Revised: Accepted: Published: 5367
March 26, 2016 June 4, 2016 June 5, 2016 June 5, 2016 DOI: 10.1021/acs.jafc.6b01390 J. Agric. Food Chem. 2016, 64, 5367−5374
Article
Journal of Agricultural and Food Chemistry
■
and 1.0 μL of each dilution was analyzed by GC-O. The flavor dilution (FD) factor of each odorant represented the maximum dilution in which the compound could be perceived. Most aroma compounds were identified by comparing their odorants, retention index (RI), and mass spectra with those of pure standards. The AEDA was performed by three experienced assessors. All analyses were repeated in triplicate by each panelist. Before experiments, they underwent a GC-O training by sniffing 40 reference compounds in their concentrations 5 times above their odor thresholds in water or air. Quantitative Analysis of Aroma Compounds. Gas Chromatography with Flame Ionization Detector. Ethyl lactate, 2-phenylethanol, 3-methyl-1-butanol, 3-methylbutanoic acid, ethyl hexanoate, and 2-furfural were quantitated by gas chromatography with a flame ionization detector (GC-FID) according to ref 21. Baijiu samples were spiked with internal standard 1(2-octanol), whose final concentration was 100 mg/L. One microliter of Baijiu sample was directly analyzed by GC with the splitless mode. Nitrogen was used as carrier gas at a constant flow rate of 1.0 mL/min. The separations were performed on a DB-FFAP column (60 m × 0.25 mm i.d. × 0.25 μm film thickness). The oven temperature was set as for the GC-O analysis described above. After 12 mixed standard solutions with different concentrations were detected by GC-FID, the correction factors of six compounds were calculated and are shown in Table 2. The mixed standard solutions were prepared as follows. At first, the mixed stock solution was prepared through dissolving standard compounds in absolute ethanol, and then was diluted to 12 different concentrations. Each standard solution was spiked with IS1 to the final concentration 100 mg/L. The standard curves were carried out by plotting the response ratio of standard compounds and IS1 against their concentration ratio (Table 2). The limits of detection (LOD) were estimated as the analyte concentration of a standard that produced a signal-to-noise ratio of 3. All analyses were repeated in triplicate. Liquid−Liquid Extraction. Fatty acids were quantitated by liquid− liquid extraction (LLE). Twenty-five mL of Baijiu sample with 10 μL of IS2 solution (5.00 mg/L final concentration) was extracted 3 times with 25 mL of dichloromethane, and then washed by 20 mL of saturated sodium chloride solution. The organic phase was dried over anhydrous sodium sulfate, and concentrated to 500 μL by distillation with a vigreux column. The concentrated extract was stored at −20 °C prior to GC-O and GC-MS analysis. The GC-MS conditions were set as GC-O analysis on a DB-FFAP column (60 m × 0.25 mm i.d., 0.25 μm film thickness) described above. The odorants were quantitated with selective ion monitoring (SIM) of MS. The monitored ion of internal standard 2 (2-methylhexanoic acid) was m/z 74. The standard curves of fatty acids were carried out as GC-FID analysis described above. The limits of detection (LODs) were estimated with the method as described above. The recovery of target compounds was estimated by the equation (C1 − C0)/C2 × 100%, where C0 is the concentration before addition, C1 is the detected concentration after addition, and C2 is the added concentration. Headspace Solid-Phase Microextraction−Gas Chromatography− Mass Spectrometry. Each Baijiu sample was diluted with Milli-Q water (Millipore, Bedford, MA) to 10% ethanol by volume. Eight mL of diluted solution with 10 μL of propyl hexanoate (IS3, 305.2 mg/L in ethanol) was put into a 20 mL screw-capped vial, and then saturated with NaCl. An automatic headspace sampling system (MultiPurpose Sample MPS 2 with a SPME adapter, from Gerstel Inc., Mülheim, Ruhr, Germany) with a carboxen/polydimethylsiloxane (CAR/PDMS; 75 μm; Supelco, Inc., Bellefonte, PA, USA) was used for the extraction of volatile components. Subsequently, the SPME fiber was inserted into the headspace and adsorbed for 40 min at 45 °C. After extraction, the loaded SPME fiber was immediately removed from the sample vial and inserted into the injection port of GC-MS for further analysis. The GC−MS conditions were set as GC−O analysis on a DB-FFAP column (60 m × 0.25 mm i.d., 0.25 μm film thickness) described above. The monitored ion of IS3 was m/z 99. Every standard stock solution was prepared through dissolving a standard compound in the model solution, and then diluted to 12 different concentrations. The model solution was prepared by diluting absolute ethanol to 42% ethanol by volume with Milli-Q water, and then the pH was adjusted
MATERIALS AND METHODS
Chemicals. All chemicals were of analytical reagent grade, with at least 97% purity. Diethyl acetal, 3-methylbutanal, ethyl propanoate, ethyl acrylate, ethyl butanoate, ethyl 2-methylbutanoate, ethyl 3-methylbutanoate, 1,1-diethoxypentane, ethyl pentanoate, ethyl 4-methylpentanoate, 3-methyl-1-butanol, ethyl hexanoate, difurfuryl ether, ethyl lactate, dimethyl trisulfide, 2-ethyl-6-methyl pyrazine, trimethyl pyrazine, ethyl 2-hydroxy-butanoate, ethyl 2-hydroxy-3methylbutanoate, ethyl octanoate, 2-ethyl-3,5-dimethyl pyrazine, 2-furaldehyde diethyl acetal, acetic acid, 2-furfural, methional, 2-acetyl furan, propanoic acid, 5-methyl furfural, 2-acetyl-5-methylfuran, ethyl furoate, butanoic acid, ethyl decanoate, ethyl benzoate, furfuryl alcohol, 3-methylbutanoic acid, diethylbutanedioate, methionol, pentanoic acid, ethyl phenylacetate, 2-phenylethyl acetate, hexanoic acid, 3-(furyl)-2propenal, benzyl alcohol, ethyl 3-phenylpropanoate, 2-phenylethanol, heptanoic acid, phenol, 4-ethyl-2-methoxyphenol, octanoic acid, 4-methylphenol, ethyl hexadecanoate, benzoic acid, phenylacetic acid, benzenepropanoic acid, and guaiacol were purchased from SigmaAldrich (Beijing, China). Dichloromethane, anhydrous sodium sulfate, and hydrochloric acid were purchased from Sinopharm Chemical Reagent Co., Ltd. (Beijing, China). The purity of the standards, before being used for quantitation, recombination, and omission experiments, was checked out according to ref 18. Samples. Both base distillate (BD) and commercial Baijiu (CB) with sesame-like flavor were used for GC-O analysis from Jingzhi Liquor Co., Ltd Shandong, China. All liquors (500 mL for each bottle) were stored at 4 °C until analysis. Mentioning of a brand name does not imply any research contact with the Baijiu manufacturer nor is it for advertising purposes. Isolation of the Volatiles. According to a previous work,19 a total of 25 mL of Baijiu sample was diluted to 10% ethanol by volume with Milli-Q water (Millipore, Bedford, MA). Before being used, the water was boiled for 5 min, and then cooled to room temperature in a 1 L flask. The diluted Baijiu sample was saturated with NaCl, and extracted 3 times with freshly distilled dichloromethane (50 mL each time). Prior to GC-O analysis, based on ref 20, the combined extract was separated into acidic and neutral/basic fractions (AF and NBF) as follows. The dichloromethane extract (about 150 mL) was extracted 3 times with sodium carbonate solution (50 mL each time; 0.50 mol/L; pH 10.0) and then washed with 20 mL of saturated sodium chloride solution. The organic phase, containing the neutral/ basic volatiles, was dried with anhydrous sodium sulfate. The combined aqueous phase was acidified to pH 2.0 with hydrochloric acid (1.0 mol/L), and extracted three times with dichloromethane (75 mL each time). The extract was then dried with anhydrous sodium sulfate. Both the acidic fraction (AF) and neutral/basic fractions (NBF) were concentrated to 500 μL by distillation with a vigreux column. The concentrated extracts were stored at −20 °C prior to GC-O and GC-MS analysis. Gas Chromatography-Olfactometric and −Mass Spectrometric Analysis. GC-O and GC-MS analyses were performed on an Agilent 7890 gas chromatograph, equipped with an Agilent 5975 massselective detector and an olfactometer (ODP3, Gerstel, Germany). Samples were analyzed on two different fused silica capillaries: DB-FFAP and HP-5MS (both 60 m × 0.25 mm i.d., 0.25 μm film thickness, J&W Scientific; USA). Helium was used as carrier gas at the fixed flow rate 1.5 mL/min. Each concentrated fraction (1.0 μL) was analyzed by GC-O and GC-MS. The injector temperature was 250 °C. The oven temperature was held at 40 °C first, then raised to 50 °C at 10 °C/min and held for 10 min, then ramped to 80 °C at 3 °C/min, held for 10 min, and finally increased at 5 °C/min to 250 °C and held for 2 min. The temperature of the olfactory port was kept at 250 °C. The electron ionization mode (EI) was used with 70 eV ionization energy. The ion source temperature was 230 °C, and the mass range was from m/z 35 to 350. Aroma Extract Dilution Analysis. Flavor dilution (FD) factors were determined for the NBF and AF by GC-O and AEDA analysis on a capillary column DB-FFAP, respectively. Extracts were diluted stepwise with dichloromethane in a series of 1:2, 1:4, 1:8, ..., 1:4096 dilutions, 5368
DOI: 10.1021/acs.jafc.6b01390 J. Agric. Food Chem. 2016, 64, 5367−5374
Article
Journal of Agricultural and Food Chemistry to 3.8 with hydrochloric acid. These standard solutions were added with the same amount of IS3 as the diluted solutions of Baijiu samples above, and then disposed and analyzed as Baijiu samples too. The standard curves were carried out as for the GC-FID analysis described above. The LOD and recovery of target compounds were detected by the method as for the LLE described above. Sensory Panel. As in the previous study,22 assessors for the aroma sensory tests consisted of 10 healthy, nonsmoking judges, 4 males and 6 females, aged 23−29. All panelists belong to the Beijing Key Laboratory of Flavor Chemistry, Beijing Technology and Business University. The assessors were selected according to their sensitive nose and previous experience in sensory evaluation. Panels were trained by describing and recognizing the odor qualities of 56 standard odorants as shown in the section Chemicals. Descriptive Profile Tests. Sensory analyses were performed in a sensory panel room at (20 ± 1) °C in three different sessions. The odor evaluation of the Baijiu was performed in the following way, as also described in ref 23. The assessors were subjected to a ranking test with a series of eight characteristic aroma solutions (25 mL in Teflon vessels), including ethyl hexanoate (fruity), acetic acid (acidic), butanoic acid and hexanoic acid (fermentation vat-like odor), ethanol (alcoholic), 2-phenylethanol (floral), 2-ethyl-3,5-dimethyl pyrazine (roasted aroma), 3-methylbutanal (malty), and steamed sorghum (grain aroma). The overall aroma profiles of Zhima aroma-type Commercial Baijiu and the aroma model mixtures were evaluated by 10 panelists. The panelists were asked to evaluate the intensities of eight odor qualities represented by the above chemicals using a seven point linear scale. Sensory analyses were performed in a sensory panel room at (21 ± 1) °C in three different sessions. The assessors were asked to rate the odor intensities as 0 (not perceivable), 1 (weak), 2 (significant), and 3 (strong) using a seven point scale of 0, 0.5, 1.0, 1.5, ..., 3.0. The obtained results were averaged for each odor note and plotted in a spider web diagram. The values, judged by the single assessor, differed by no more than 20%. Aroma Recombination of Zhima Aroma-type Commerical Baijiu. An aroma recombination, consisting of 26 odorants with OAVs ≥ 1, was prepared in 42% hydroalcoholic solution (by volume), and the pH was adjusted to 3.8 with hydrochloric acid (1.0 mol/L). The aroma profile of the model mixture was determined in the other room in the same way as for the Zhima aroma-type Commercial Baijiu described above. The similarity between them was estimated by a seven point scale from 0 to 3. Sensory data were analyzed by oneway analysis of variance (ANOVA) by use of SPSS17.0 (SPSS Inc., Chicago, IL). Omission/Addition Experiments. Mixture models were prepared by omitting one or a group of selected components from 26 odorants, and then the sensory panel would evaluate the differences between the omission models and the complete recombinate model in a triangle test as described in ref 24. The significance R of the detected difference was calculated according to ref 25. The sensory panel for the omission experiments was the same as that for the descriptive profile tests.
regarded as the characteristic aroma of Zhima aroma-type Baijiu.4 Ethyl acrylate (256; plastic) was first detected in CB and BD, which exhibited an intense plastic odor. Among the acidic volatiles, the important odor-active compound was identified as hexanoic acid (sour, pungent) followed by butanoic acid (sweaty). Compounds with somewhat lower FD factors were identified as trimethyl pyrazine (nutty), 2-acetyl furan (nutty), and ethyl furoate (caramel) (Table 1). Most of these compounds have been earlier identified as aroma compounds in other aroma-types of Chinese Baijiu.26−28 Concentration of Aroma Compounds in Zhima Aroma-type Baijiu. To gain a deeper insight into the aroma of Zhima aroma-type Baijiu, a total of 56 odorants detected by AEDA were quantitated in BD and CB (Table 2). As exhibited in Table 2, ethyl lactate had the highest concentration (819 mg/L-BD, 509 mg/L-CB), followed by 3-methyl-1-butanol (244 mg/L, 474 mg/L) and ethyl hexanoate (175 mg/L, 148 mg/L). Besides, these odorants had higher concentrations, such as acetic acid (99 mg/L, 79 mg/L), 2-furfural (96 mg/L, 81 mg/L), and 2-phenylethanol (47 mg/L, 16 mg/L). Almost all compounds had higher concentrations in BD than in CB, but these acid compounds, including 3-methylbutanoic acid, hexanoic acid, pentanoic acid, and octanoic acid, had lower concentrations, which could explain why BD had a more pungent aroma. The employed quantitative methods were able to detect all of the volatiles identified by AEDA. The obtained calibration curves were found to have good linearity with correlation coefficient (R2) ≥ 0.99; RSDs in triplicate of samples were ≤10%, which indicated the good precision of the quantitative methods. Odor Activity Values. Although dilution to odor threshold techniques, such as the aroma extract dilution analysis and the aroma dilution analysis, are useful methods for screening of important odorants in foods, these methods involved neither the influence of the food matrix on odorant binding nor the interactions of odorants when matching the overall odor impression of the food.29 Besides, as Grosch reported,11 the influence of ethanol concentration on the volatility of an odorant should not be ignored. Take these cases into consideration, the odor activity values (ratio of concentration to its odor threshold) were calculated to evaluate the contribution of odorants to the overall aroma of CB for each aroma compound listed in Table 3. There were 26 odorants with OAVs ≥ 1, which should be important to the characteristic aroma of CB. The highest OAV in CB was ethyl hexanoate (OAV = 2691), followed by 3-methylbutanal (OAV = 2403), ethyl pentanoate (OAV = 1019), and ethyl octanoate (OAV = 782). These compounds, including dimethyl trisulfide (cabbage, OAV = 388), methional (roasted potato, OAV = 17), 2-ethyl-3,5-dimethyl pyrazine and (roasted potato, OAV = 8), also might contribute to the overall aroma due to their OAVs (≥1). However, the OAVs of trimethyl pyrazine, 2-acethyl furan, 5-methyl furfural, and methionol were not more than 1.0, and so they could not contribute to the overall aroma of CB. The compounds had high OAVs (≥1) and almost have high AEDAs, while 3-methylbutanal had a high OAV (2403) but a low AEDA (8), and propanoic acid had a high AEDA (128) but a low OAV (