Comparative study on key aroma components of Hongqu Huangjiu with different sugar contents through aroma analysis technology

Abstract

Hongqu Huangjiu (HQHJ) is a representative traditional Chinese fermented beverage with medicinal and nutritional value. However, composition of aroma-active components in HQHJ with different sugar contents remains unclear. Herein, solid-phase extraction combined with gas chromatography–olfactometry–mass spectrometry revealed 56 aroma-active compounds (20 newly identified) in HQHJ. Difurfuryl sulfide was identified as a odorants in Huangjiu for the first time. After quantification and odour activity value calculation, β-phenethyl alcohol, dimethyl trisulfide, 1,1-diethoxyethane and ethyl isobutyrate were identified as important odorants. Combined with partial least squares regression analysis and addition tests, ethyl 4-oxopentanoate (imparting Chen aroma to HQHJ) and 2-methoxy-4-vinylphenol were validated as key differential odorants in HQHJ with varying sugar contents. Notably, semi-sweet HQHJ promoted contents accumulation of aroma compounds, while semi-dry HQHJ enhanced aroma intensities and diversity of compounds. Conversely, sweet HQHJ, through its brewing process involving adding alcohol, inhibited formation of certain compounds. Therefore, they exhibited distinctive aromatic characteristics.

Highlights

• •Twenty odorants were detected for the first time in Hongqu Huangjiu (HQHJ).
• •Difurfuryl sulfide was firstly identified as an aroma-active compound in Huangjiu.
• •The majority of aroma attributes exhibited highest intensity in semi-dry HQHJ.
• •Differential odorants of HQHJ with different sugar contents were explored.
• •Ethyl 4-oxopentanoate was the key contributor of Chen-aroma in HQHJ.

1. Introduction

Huangjiu is one of the three major ancient liquors in the world, alongside beer and wine (Sun, 2019). It is made from Qu (a starter) as the saccharification and fermentation agent and rice as the main raw material through a process involving saccharification, fermentation, filtration, sterilisation and ageing. The Qu used for Huangjiu includes Maiqu (wheat Qu), Miqu (rice Qu) and Hongqu (red Qu). Hongqu, a unique fermented starter, is produced by fermenting rice with Monascus to achieve a brownish-red or purplish-red colour (Wang et al., 2014). As a medicine and food homology component, Hongqu contains bioactive compounds such as Monascus pigments, γ-aminobutyric acid and monacolin K, which help reduce blood pressure, blood lipid levels and blood glucose levels, suppress tumour growth, prevent corrosion, and inhibit bacterial growth (Shen et al., 2014). Hongqu Huangjiu (HQHJ), which is made from Hongqu as the starter, is mainly produced in Fujian and Zhejiang Provinces in China. During brewing, Hongqu not only enhances the medicinal and nutritional value of HQHJ but also imparts a unique aroma to the beverage (Bai et al., 2023).

Flavor is not only a key factor determining consumer acceptance and preference but also an important indicator of food quality (Wang et al., 2020; Wang et al., 2023). Recently, some studies have investigated the aroma profile of HQHJ. For example, eight aroma compounds exhibiting a positive correlation with ageing duration were identified in 10 Wuyi HQHJ samples through headspace solid-phase microextraction (HS-SPME) coupled with gas chromatography–mass spectrometry (GC–MS) (Su et al., 2024). Ethyl acetate, 2-methyl-1-propanol, and 3-methyl-1-butanol were identified as the characteristic compounds between two HQHJ fermented by Gutian Qu and Wuyi Qu via HS-SPME combined with gas chromatography-olfactometry (GC-O) (Chen et al., 2023). Isobutanol, isopentanol, ethyl acetate and 11 other compounds were identified as major contributors to the characteristic aroma of sweet Chengang HQHJ with three different vintages through HS-SPME–GC–MS (Zheng et al., 2014). Wang (2023) investigated the composition of aroma-active compounds in Gutian and Wuyi HQHJ through HS-SPME and solid-phase extraction (SPE) combined with GC–O and GC–MS. They found that phenylacetaldehyde, ethyl lactate, ethyl isobutyrate, isovaleric acid, vanillin and ethyl acetate were key aroma compounds in both HQHJ types. A majority of the studies on HQHJ aroma have used HS-SPME–GC–MS for semi-quantitative analysis. Despite its advantages of high sensitivity, ease of operation, and being solvent-free, HS-SPME has key limitations including relatively poor reproducibility and low extraction efficiency for strongly polar, high-molecular-weight, and low-volatility compounds (Godage & Gionfriddo, 2019). In addition, existing studies have primarily focused on volatile compounds instead of aroma-active compounds in HQHJ. Currently, studies on key aroma-active compounds in HQHJ are insufficient. Furthermore, the differences in aroma characteristics between HQHJ with varying sugar contents remain unclear, and the aroma-active compounds responsible for these distinctions have yet to be identified.

This study compared the aroma profiles of HQHJ with varying sugar contents using the following methods: (1) isolation of aroma compounds using SPE, (2) identification of aroma-active compounds through gas chromatography–olfactometer–mass spectrometry (GC–O–MS) and odour specific magnitude estimation (Osme), (3) quantification of aroma-active compounds and calculation of their odour activity values (OAVs), (4) analysis of the relationship between sensory data and aroma-active compounds and determination of the impact of different odorants on HQHJ aroma through partial least squares regression (PLSR) analysis, and (5) validation of key odorants responsible for aroma differences among HQHJ with varying sugar contents through addition tests.

2. Materials and methods

2.1. Samples

Chengang-light type (CG-L), Chengang-blue label (CG-B), Chengang-three years of mellowness (CG-T) and Danxi Huangjiu (DX) were obtained from Longyan Chengang Wine Co., Ltd. (Longyan, Fujian, China) and Yiwu Danxi Wine Co., Ltd. (Yiwu, Zhejiang, China). These four types of HQHJs were characterised as follows: CG-L, 12%vol alcohol and dry type; DX, 15%vol alcohol and semi-dry type; CG-B, 12%vol alcohol and semi-sweet type; CG-T, 15%vol alcohol and sweet type. A total of 500 mL of each HQHJ was stored at −20 °C until further analysis.

2.2. Chemicals and reagents

The purity of all chemical standards was greater than 95%. The basic information of the chemical standards are shown in Table S1.

2.3. Sensory evaluation

As described in a previous study (Wang et al., 2020), 20 panellists (10 men and 10 women, aged 21–30 years) with experience in sensory evaluation were recruited from the Key Laboratory of Brewing Molecular Engineering of China Light Industry. The sensory evaluation experiment was reviewed and approved by the Ethics Committee of Beijing Technology and Business University (Ethics Approval No.: Beijing Technology and Business University Ethics Review 2024 No. 195). All panellists signed an informed consent form, and their rights and privacy were protected during the execution of the research. Nine aroma attributes and their corresponding standards were evaluated: alcoholic (ethanol), Zao (fermented grains, called Jiuzao), Qu (Xiaoqu, fermentation starter for Huangjiu), acidic (acetic acid), raw grain (raw grains), sweet (ethyl 3-phenylpropanoate), Chen (mature vinegar), Jiang (soy sauce), and herbal (sotolon) (Wang et al., 2020; Wang et al., 2022; Zhang et al., 2024). Briefly, 10.0 mL of each HQHJ sample was transferred into 15 mL glasses coded with three-digit numbers randomly. The panellists rated the intensity of the eight attributes on a 7-point scale: 0, undetectable; 3, moderate; 7, extremely strong. Each HQHJ sample was evaluated three times.

2.4. Isolation of aroma compounds

Aroma compounds were isolated from all HQHJ samples as described previously (Chen, Wang, & Xu, 2013), with minor modifications. Briefly, 20 mL of each HQHJ sample was mixed with 16.0 μL of 2-methyl-3-heptanone (IS1, 1000 mg/L), and 3.0 g of NaCl. The mixture was separated through extraction on an LiChrolut EN SPE columns (3 mL) with LiChrolut EN packing (250 mg) (Merck, Darmstadt, Germany). Before this mixture was added, the column was activated with 6.0 mL of dichloromethane, 6.0 mL of methanol and 6.0 mL of ultrapure water. The flow rate was maintained at below 2 mL/min when adding the sample. After the sample flowed completely through the column, it was eluted with 6.0 mL of ultrapure water. After the column dried, 6.0 mL of dichloromethane was added for elution. The eluate collected was dried with anhydrous sodium sulfate, concentrated to 0.50 mL via nitrogen blowing (99.999%, 10 mL/min) and stored at −40 °C until further analysis.

2.5. GC–O–MS analysis

The extracts obtained through SPE were analysed using a 7890B GC system combined with a 5977 A MSD mass spectrometer (Agilent Technologies, CA), and an ODP-III olfactory detection port (Gerstel, Germany). Briefly, 1.0 μL of each extract was injected into the injection port at 250 °C, and helium (99.999%) was used as the carrier gas at a flow rate of 2.0 mL/min (Zhang et al., 2024). The temperature of the olfactory detection port was maintained at 230 °C, and a DB-WAX column (60 m × 0.25 mm × 0.25 μm) was used for separation. The initial temperature of the oven was 40 °C, increased to 50 °C at 10 °C/min and held for 10 min, then raised to 80 °C at 3.0 °C/min and kept for 10 min, and finally raised to 230 °C at 5.0 °C/min and maintained for 5.0 min. The fraction separated by the column flowed to the olfactory detector and port at a ratio of 1:1 v/v. The temperature of the ion source, quadrupole and transmission line were maintained at 230 °C, 150 °C, and 230 °C, respectively. The electron impact (EI) mode was adopted with an electron energy of 70 eV, and the full scan mode was used within the scanning range of m/z 30–400. Aroma compounds were identified by comparing with the National Institute of Standards and Technology (NIST) 2017 mass spectral library (the match of compound was greater than 700), retention index (RI), aroma attributes of the compounds, and standards.

2.6. Odour specific magnitude estimation (Osme)

Osme was performed by three sensory evaluators (two women females and one man) using GC–O–MS analysis (Zhang et al., 2024). It was recorded when more than two evaluators determined the odour area. The evaluators rated the aroma intensity of volatile compounds on a scale of 0–5, with 0 indicating no aroma, 3 representing moderate intensity, and 5 denoting extremely strong intensity. Each evaluator rated the same sample three times.

2.7. Quantitative analysis of odorants

Odorants were quantified using the internal standard curve method and internal standard method. Each internal standard was prepared in a 10% aqueous ethanol solution, which was diluted to 10 concentrations through a 3-fold dilution. An equal amount of the internal standard was added to each solution. The final concentration of 2-methyl-3-heptanone (IS1) was 0.80 mg/L for SPE, and that of 4-octanol (IS2) was 1.02 mg/L for HS-SPME. The concentrations of target compounds in HQHJ were calculated by plotting internal standard curves from the peak area ratios of the compounds and internal standards versus their concentration ratios.

Nine compounds (9, 15, 21, 22, 24, 25, 26, 32, and 42; Table 1) were quantified using SPE and GC–MS (TRACE1310-ISQ LT GC–MS, Thermo Fisher Scientific, USA). Briefly, 20 mL of each HQHJs sample and 16.0 μL of IS1 (1000 mg/L) were mixed and saturated with 3.0 g of NaCl. The mixture was extracted using an SPE column as described in section 2.4. The eluate of dichloromethane was collected, dried with anhydrous sodium sulfate as described, concentrated to 0.5 mL through nitrogen blowing (99.999%, 10 mL/min) and stored at −40 °C until further analysis. A DB-WAX (60 m × 0.25 mm × 0.25 μm) column was used for separation, and helium (99.999%) was used as the carrier gas at a flow rate of 1.5 mL/min. The injection volume was 1.0 μL, and the split-less mode was selected. The temperature of the injection port was maintained at 250 °C. The initial temperature of the oven was 40 °C, which was increased to 80 °C at 10 °C/min, held for 5 min, then up to 230 °C at 5 °C/min and kept for 10 min, followed by rising to 240 °C at 5 °C/min and held for 5 min. The temperature of the ion source and transmission line was maintained at 230 °C. The EI mode was used at 70 eV. Aroma compounds were quantified via the selective ion monitoring (SIM) mode, and the monitored ions of the compounds are shown in Table 2.

Table 1.

Odour description and aroma intensity (AI) of aroma active compounds in Hongqu Huangjiu.

No.	cRI\lRIa	Compoundsb	Aroma	AI factorsc		Average AI factorsd	Identificatione
				CG-L	DX
Esters
1	878/888	Ethyl acetate	fruity, sweet	2	2	2.0	MS, RI, aroma, S
2	939/953	Ethyl propionate	fruity, sweet	1	2	1.5	MS, RI, aroma, S
3	971/961	Ethyl isobutyrate	fruity, sweet	2	2	2.0	MS, RI, aroma, S
4	1026/1036	Ethyl butanoate	fruity	2	2	2.0	MS, RI, aroma, S
5	1426/1422	Ethyl 2-hydroxy-3-methylbutanoatef	pineapple, strawberry, sweet	–	4	4.0	MS, RI, aroma
6	1505/1515	Ethyl 3-hydroxybutanoatef	green, fruity	2	3	2.5	MS, RI, aroma, S
7	1614/1607	Ethyl 4-oxopentanoatef	fruity, green, waxy	–	1	1.0	MS, RI, aroma, S
8	1644/1658	Ethyl benzoate	floral, fruity	–	5	5.0	MS, RI, aroma, S
9	1672/1681	Diethyl butanedioate	wine, fruity, sweet	2	5	3.5	MS, RI, aroma, S
10	1820/1813	Phenethyl acetate	rose, honey, sweet	–	1	1.0	MS, RI, aroma, S
11	1892/1893	Ethyl 3-phenylpropanoate	floral, sweet	–	2	2.0	MS, RI, aroma, S
12	1768/1783	Ethyl phenylacetate	sweet, floral, honey	–	2	2.0	MS, RI, aroma
13	2065/2047	Diethyl malatef	wine, fruity, apple	5	5	5.0	MS, RI, aroma
14	2195/2195	Diethyl 2-hydroxypentanedioatef	cotton candy	3	1	2.0	MS, RI, aroma
Alcohols
15	1108/1092	2-Methyl-1-propanol	wine, bitter, solvent	1	2	1.5	MS, RI, aroma, S
16	1134/1142	1-Butanol	fusel, wine	3	2	2.5	MS, RI, aroma, S
17	1217/1211	3-Methyl-1-butanol	fusel, fruity	–	5	5.0	MS, RI, aroma
18	1877/1870	Benzyl alcohol	sweet, floral	5	2	3.5	MS, RI, aroma
19	1912/1906	β-Phenethyl alcohol	honey, rose	5	5	5.0	MS, RI, aroma, S
Acids
20	1429/1453	Acetic acidf	sour	2	2	2.0	MS, RI, aroma
21	1581/1556	2-Methylpropionic acid	sour	2	2	2.0	MS, RI, aroma, S
22	1637/1628	Butyric acid	sour, sweaty	2	2	2.0	MS, RI, aroma, S
23	1680/1661	3-Methylbutanoic acid	pungent sour	2	2	2.0	MS, RI, aroma, S
24	1720/1713	Pentanoic acid	sweaty, rancid	2	2	2.0	MS, RI, aroma, S
25	1854/1882	Hexanoic acid	sweaty, cheese	2	2	2.0	MS, RI, aroma, S
26	2038/2050	Octanoic acid	sweaty, cheese	–	3	3.0	MS, RI, aroma, S
27	2144/2150	Nonanoic acid	cheese	2	2	2.0	MS, RI, aroma, S
28	2320/2365	Decanoic acidf	unpleasant rancid	2	–	1.0	MS, RI, aroma
29	2530/2569	Benzeneacetic acidf	honey, floral	3	3	3.0	MS, RI, aroma
Aldehydes
30	1508/1520	Benzaldehyde	almond, burnt sugar	1	2	1.5	MS, RI, aroma, S
31	1933/1907	2-Phenyl-2-butenal	sweet	–	1	1.0	MS, RI, aroma, S
32	2588/2566	Vanillin	vanilla	3	3	3.0	MS, RI, aroma, S
Ketones
33	1132/1111	3-Penten-2-onef	musty	1	2	1.5	MS, RI, aroma
34	1317/1298	1-Octen-3-one	mushroom	2	2	2.0	MS, RI, aroma
35	1627/1647	Acetophenone	sweet, floral	2	5	3.5	MS, RI, aroma, S
Acetals
36	894/892	1,1-Diethoxyethanef	fruit, cream	1	2	1.5	MS, RI, aroma, S
37	962/973	1,1-Diethoxy-2-methylpropanef	fruity	2	2	2.0	MS, RI, aroma
38	1020/994	2,4,5-Trimethyl-1,3-dioxolanef	fruity	2	2	2.0	MS, RI, aroma
39	1061/1065	1,1-Diethoxy-3-methylbutanef	fruity	2	–	1.0	MS, RI, aroma, S
Furans
40	1297/1291	Ethyl furfuryl etherf	spicy, solvent	–	2	2.0	MS, RI, aroma
41	1432/1462	Furfural	bread, almond, sweet	2	2	2.0	MS, RI, aroma
42	1512/1499	2-Acetylfuran	nutty, roast, sweet	2	2	2.0	MS, RI, aroma, S
43	1597/1605	2-Acetyl-5-methylfuran	nutty, milky	–	2	2.0	MS, RI, aroma
Phenols
44	2183/2172	4-Ethylphenol	smoky, herbal	1	5	3.0	MS, RI, aroma
45	2203/2180	2-Methoxy-4-vinylphenolf	lilac, curry, spicy	–	5	5.0	MS, RI, aroma, S
Lactone
46	2204/2210	4-Hydroxy-2,3-dimethyl-2H-furan-5-onef	herbal	5	5	5.0	MS, RI, aroma
Nitrogen-containing compounds
47	1330/1319	2,6-Dimethylpyrazine	nutty, cocoa, roasted beef	–	2	2.0	MS, RI, aroma, S
48	1382/1375	2-Ethyl-6-methylpyrazinef	roasted hazelnut	2	3	2.5	MS, RI, aroma, S
49	1940/1952	2-Acetyl-1H-pyrrole	nutty, walnut, bread	5	2	3.5	MS, RI, aroma, S
50	2036/2030	2-Formyl-1H-pyrrole	musty, beef, coffee	–	2	2.0	MS, RI, aroma, S
51	1436/1444	2,6-Diethylpyrazinef	nutty, hazelnut	2	3	2.5	MS, RI, aroma
Sulfur - containing compounds
52	1362/1376	Dimethyl trisulfide	cabbage	2	3	2.5	MS, RI, aroma, S
53	1785/1755	5-Methyl-2-thiophenecarboxaldehyde	sweet almond, roasted, woody	–	3	3.0	MS, RI, aroma
54	2215/2223	Difurfuryl sulfidef	sofu, sauce, salty, broad bean paste	5	–	5.0	MS, RI, aroma
Others
55	2179/2187	(2S,6R,7S,8E)-(+)-2,7-epoxy-4,8-megastigmadienef	sweet, floral	–	2	2.0	MS, RI, aroma
56	2230/−	Unknown	salty	5	–	5.0	–

comparison with those of reference standards; aroma, identification based on odour description.

^a“cRI/lRI”, represents the calculated and literature retention indices, respectively.

^b“Compound”, represents aroma compounds detected in Hongqu Huangjius.

^c“AI factors”, represents aroma intensity factors; −, not detected. CG-L, DX represent Chengang-light type Hongqu Huangjiu; DX, Danxi Huangjiu, respectively.

^d“Average AI factors”, represents the average valve of the aroma intensity factors of 2 Hongqu Huangjius.

^eMS, identification based on mass spectrometric data in NIST library (Version 2017); RI, identification based on RI reported in literatures; S, identification by.

^fThe aroma-active compound was firstly identified in Hongqu Huangjiu.

Table 2.

Concentrations of odour-active compounds in four kinds of Hongqu Huangjius with different sugar contents.

No.	Compounds	Quantitate ions	Standard curves	R2	Concentrations (μg/L) a				Average Concentrations (μg/L)b
					CG-L	DX	CG-B	CG-T
Esters
1	Ethyl acetatec	43	y = 56.061× - 0.362	0.9987	34,875.95 ± 1201.12	42,570.99 ± 2156.12	46,269.37 ± 6479.42	45,653.01 ± 2329.80	42,342.33
2	Ethyl propionatec	57	y = 38.382× + 0.0429	0.9999	–	54.07 ± 10.40	46.85 ± 3.95	–	50.46
3	Ethyl isobutyratec	43	y = 17.217× + 0.1388	0.9995	1396.47 ± 139.33	1985.41 ± 129.13	3318.07 ± 80.02	3447.44 ± 230.60	2536.85
4	Ethyl butanoatec	71	y = 8.7172× + 0.1055	0.9994	616.81 ± 13.74	706.81 ± 11.78	1459.07 ± 99.85	1025.31 ± 152.52	952.00
6	Ethyl 3-hydroxybutanoatec	43	y = 391.22× + 0.0843	0.9996	3.46 ± 0.17	11.80 ± 0.99	4.41 ± 0.17	4.82 ± 0.37	6.12
7	Ethyl 4-oxopentanoatec	43	y = 107.2× + 0.0693	0.9987	2048.84 ± 71.88	1457.23 ± 275.49	6919.84 ± 255.67	4854.86 ± 526.14	3820.19
8	Ethyl benzoatec	105	y = 0.5157× + 0.0038	0.9991	0.89 ± 0.09	1.89 ± 0.04	0.52 ± 0.01	1.49 ± 0.06	1.20
9	Diethyl butanedioated	101	y = 1.6278× - 0.3604	0.9958	194.94 ± 8.69	655.24 ± 13.32	339.02 ± 7.83	21.62 ± 1.42	302.70
10	Phenethyl acetatec	104	y = 2.6231× - 0.0868	0.9982	208.73 ± 5.92	402.27 ± 5.11	269.19 ± 6.85	258.79 ± 8.76	284.75
11	Ethyl 3-phenylpropanoatec	104	y = 6.3948× + 0.0464	0.9999	–	228.74 ± 40.81	188.24 ± 35.08	131.15 ± 24.00	182.71
12	Ethyl phenylacetatec, e	91	–	–	83.49 ± 6.62	133.70 ± 10.28	381.01 ± 45.66	193.86 ± 13.44	198.02
13	Diethyl malated, e	71	–	–	6.12 ± 1.22	33.29 ± 1.76	11.76 ± 0.34	23.34 ± 2.64	18.63
	Total				39,435.71	48,241.44	59,207.35	55,615.67	50,625.04
Alcohols
15	2-Methyl-1-propanold	43	y = 7.0474× - 0.1028	0.9991	1012.83 ± 66.70	2866.60 ± 53.68	1594.08 ± 5.35	2403.35 ± 351.65	1969.22
16	1-Butanolc	31	y = 278.88× - 1.8093	0.997	6225.74 ± 214.09	6048.83 ± 588.97	12,149.88 ± 2073.63	8575.91 ± 136.22	8250.09
17	3-Methyl-1-butanolc, e	55	–	–	1076.06 ± 11.47	1504.47 ± 36.66	1878.43 ± 280.12	1657.42 ± 5.95	1529.09
18	Benzyl alcoholc, e	79	–	–	55.66 ± 3.48	44.63 ± 3.89	136.42 ± 16.69	128.47 ± 6.67	91.29
19	β-Phenethyl alcoholc	91	y = 14.167× - 0.197	0.9964	15,991.92 ± 2367.75	30,141.66 ± 2036.56	32,075.00 ± 2387.57	16,447.90 ± 299.58	23,664.12
	Total				24,362.21	40,606.18	47,833.81	29,213.06	35,503.81
Acids
20	Acetic acidc, e	43	–	–	379.11 ± 10.02	570.98 ± 44.79	501.14 ± 40.01	477.16 ± 8.53	482.10
21	2-Methylpropionic acidd	43	y = 4.9665× - 0.0241	0.9999	53.10 ± 2.07	207.68 ± 2.43	105.27 ± 0.05	141.86 ± 2.31	126.98
22	Butyric acidd	60	y = 4.9264× - 0.069	0.9998	18.93 ± 1.41	71.53 ± 1.39	25.93 ± 0.19	32.37 ± 0.87	37.19
23	3-Methylbutanoic acidd, e	60	–	–	16.40 ± 1.41	75.78 ± 0.28	24.63 ± 0.71	49.85 ± 1.65	41.66
24	Pentanoic acidd	60	y = 4.6055× - 0.1616	0.9987	13.95 ± 0.27	27.58 ± 4.07	62.58 ± 5.94	16.84 ± 0.99	30.24
25	Hexanoic acidd	60	y = 3.5656× - 0.2022	0.9975	115.68 ± 9.70	108.97 ± 4.34	151.84 ± 15.31	110.43 ± 5.54	121.73
26	Octanoic acidd	60	y = 3.1641× - 0.7862	0.9922	23.80 ± 3.03	73.14 ± 1.72	7.11 ± 0.77	14.92 ± 0.76	29.74
27	Nonanoic acidc	60	y = 20.909× + 2.0561	0.9958	2100.82 ± 52.14	2083.20 ± 38.33	2306.37 ± 110.47	2097.34 ± 41.24	2146.93
28	Decanoic acidc, e	60	–	–	15.06 ± 1.81	3.92 ± 0.39	10.60 ± 0.55	3.98 ± 0.22	8.39
29	Benzeneacetic acidd, e	91	–	–	9.45 ± 1.52	58.37 ± 2.89	12.96 ± 0.53	25.88 ± 2.02	26.67
	Total				2746.29	3281.15	3208.44	2970.62	3051.62
Aldehydes
30	Benzaldehydec	77	y = 2.1674× + 0.003	0.9999	1490.49 ± 80.24	2319.09 ± 105.59	2220.35 ± 407.18	1873.49 ± 20.27	1975.85
31	2-Phenyl-2-butenalc	117	y = 2.5803× + 0.0804	0.9995	192.17 ± 22.91	515.52 ± 97.40	628.93 ± 97.20	353.19 ± 6.61	422.45
32	Vanillind	151	y = 1.9707× - 0.1034	0.9995	5.16 ± 0.81	143.41 ± 3.09	15.15 ± 0.45	15.14 ± 0.57	44.71
	Total				1687.81	2978.03	2864.43	2241.81	2443.02
Ketones
35	Acetophenonec	105	y = 1.9854× + 0.0163	0.9997	116.97 ± 2.62	–	457.98 ± 85.85	231.47 ± 44.79	268.81
	Total				116.97	–	457.98	231.47	268.81
Acetals
36	1,1-Diethoxyethanec	45	y = 15.019× - 0.3525	0.9916	11,997.34 ± 336.11	18,040.77 ± 293.67	17,888.11 ± 1464.22	14,822.51 ± 1072.87	15,687.18
37	1,1-Diethoxy-2-methylpropanec, e	103	–	–	49.45 ± 1.74	222.67 ± 22.81	223.89 ± 8.46	128.64 ± 20.67	156.16
38	2,4,5-Trimethyl-1,3-dioxolanec, e	43	–	–	899.73 ± 55.60	1045.80 ± 82.93	745.69 ± 65.80	786.50 ± 21.90	869.43
39	1,1-Diethoxy-3-methylbutanec	103	y = 0.6913× + 0.0433	0.9995	242.41 ± 4.25	774.05 ± 69.18	1009.36 ± 93.42	709.84 ± 24.62	683.91
	Total				13,188.93	20,083.29	19,867.05	16,447.49	17,396.69
Furans
41	Furfuralc, e	96	–	–	447.01 ± 27.00	534.14 ± 39.01	781.34 ± 7.32	782.91 ± 4.63	636.35
42	2-Acetylfurand	95	y = 2.931× - 0.0988	0.9998	0.19 ± 0.03	4.07 ± 0.07	4.26 ± 0.08	6.86 ± 0.85	3.84
43	2-Acetyl-5-methylfurand, e	109	–	–	0.27 ± 0.02	1.87 ± 0.08	1.10 ± 0.21	1.83 ± 0.32	1.27
	Total				447.48	540.08	786.70	791.59	641.46
Phenols
44	4-Ethylphenolc, e	107	–	–	–	7.53 ± 1.09	9.88 ± 1.60	7.28 ± 0.83	8.23
45	2-Methoxy-4-vinylphenolc	135	y = 1460.3× - 0.0527	0.9949	–	1743.76 ± 116.87	–	–	1743.76
	Total				–	1751.29	9.88	7.28	589.48
Nitrogen-containing compounds
47	2,6-Dimethylpyrazinec	108	y = 26.627× + 0.0447	0.9999	202.86 ± 13.54	288.59 ± 6.55	607.53 ± 6.04	–	366.33
48	2-Ethyl-6-methylpyrazinec	121	y = 56.061× - 0.362	0.9994	98.77 ± 13.71	86.94 ± 1.52	–	–	92.86
49	2-Acetyl-1H-pyrrolec	94	y = 23.017× + 0.0962	0.994	269.98 ± 52.26	338.64 ± 24.65	432.34 ± 81.37	220.45 ± 6.11	315.35
50	2-Formyl-1H-pyrrolec	95	y = 82.196× + 0.2384	0.9916	328.77 ± 8.11	1133.95 ± 11.37	283.07 ± 44.67	274.16 ± 35.76	504.99
51	2,6-Diethylpyrazined, e	135	–	–	–	0.30 ± 0.01	–	–	0.30
	Total				900.38	1848.42	1322.95	494.61	1141.59
Sulfur-containing compounds
52	Dimethyl trisulfidec	126	y = 1.3292× + 0.0207	0.9995	–	46.72 ± 1.74	126.26 ± 6.12	110.60 ± 8.41	94.53
	Total				–	46.72	126.26	110.60	94.53
	In total				82,885.77 ± 4814.57	119,376.59 ± 6353.22	135,684.84 ± 14,322.75	108,124.20 ± 5423.18	111,517.85

^aConcentrations of these compounds in Hongqu Huangjius; values (means ± SD, n = 3). (CG-L, Chengang-light type; DX, Danxi Huangjiu; CG-B, Chengang-blue label; CG-T, Chengang-three years of mellowness).

^bThe average concentrations of the compounds of four Hongqu Huangjiu samples.

^cQuantified by headspace solid-phase microextraction combined with gas chromatography–mass spectrometry (HS-SPME-GC–MS).

^dQuantified by solid-phase extraction combined with gas chromatography–mass spectrometry (SPE-GC–MS).

^eSemi-quantitative via internal standard method.

A total of 23 compounds (1–4, 6–8, 10–11, 16, 19, 27, 30, 31, 35, 36, 39, 45, 47–50, and 52) were quantified using HS-SPME–GC–MS (Wang et al., 2020). Briefly, 6.0 mL of each HQHJ sample saturated with 2.0 g of NaCl and 11.0 μL of IS2 (500 mg/L) was transferred to a 20 mL headspace vial and heated in a thermostatic water bath at 45 °C for 30 min. Thereafter, a divinylbenzene/carboxen/polydimethylsiloxane fibre (DVB/CAR/PDMS, grey, 50/30 μm; Supelco, USA) was inserted into the vial, and the mixture was extracted for 30 min. The fibre was desorbed in the GC injection port at 250 °C for 5 min. Other instrument conditions were consistent with the GC–MS conditions described above.

Through internal standard methods, five compounds (13, 23, 29, 43 and 51) were quantified by SPE combined with GC–MS and nine compounds (12, 17, 18, 20, 28, 37–38, 41, and 44) were quantified using HS-SPME combined with GC–MS. The compounds were isolated and identified as described above. Each sample of HQHJ was injected three times.

2.8. Odour activity values (OAVs) calculation

OAV was calculated as the ratio of the concentration of an aroma-active compounds to its odour threshold. The odour thresholds of most compounds were obtained from previous studies, and their measurement matrix was a 14% ethanol solution (Van Gemert, 2011; Wang et al., 2020; Wang et al., 2022). The odour thresholds of three compounds, namely, 2-ethyl-6-methylpyrazine, 1,1-diethoxy-3-methylbutane and 1,1-diethoxy-2-methylpropane, were determined using the three-alternative forced-choice (3-AFC) method (Wang et al., 2022). Briefly, the compounds were dissolved in a 14% ethanol solution, and a series of solutions with different concentrations was prepared. Each solution was differentiated from the two equal basic solutions (14% ethanol solutions) by 15 professional sensory evaluators from Key Laboratory of Brewing Molecular Engineering of China Light Industry.

2.9. Addition tests

Spiking experiments were performed to investigate the contribution of 2-methoxy-4-vinylphenol, vanillin, phenethyl acetate and ethyl 4-oxopentanoate to the aroma profiles of HQHJ with varying sugar contents (Wang et al., 2020). Briefly, 108.0 μL of ethyl 4-oxopentanoate (1019.50 mg/L dissolved in 12% ethanol/water solution) was added to DX (20.00 mL) until the concentration was equal to that in CG-B (6919.84 μg/L). In addition, 92.0 μL of 2-methoxy-4-vinylphenol (381.00 mg/L dissolved in 12% ethanol/water solution), 77.0 μL of phenethyl acetate (34.90 mg/L dissolved in 12% ethanol/water solution) and 48.0 μL of vanillin (53.60 mg/L dissolved in 12% ethanol/water solution) were separately added to CG-B (20.00 mL) until the concentrations were equal to those in DX (1743.76, 402.27, and 143.41 μg/L, respectively). The aroma intensity of the spiked samples was scored by the same panellists, as mentioned in Section 2.3.

2.10. Statistical analysis

Microsoft Office Excel 2016 was used to process quantitative data. PLSR analysis was performed using XLstat 2016 (Pu et al., 2019). Significance was analysed using the SPSS (version 27.0) software. Figures were drawn using the Origin 2022 software.

3. Results and discussion

3.1. Sensory evaluation

Sensory evaluation results are shown in Fig. 1. Nine aroma attributes, namely, alcoholic, Zao-aroma, Qu-aroma, acidic, raw grain, sweet, Chen-aroma, Jiang-aroma, and herbal, were used to describe the aromatic characteristics of the four HQHJ samples. The average intensity of alcoholic aroma was the highest (4.8), whereas that of the herbal aroma was the lowest (3.1). As shown in Fig. 1, alcoholic (4.3) and Zao-aroma (4.3) were the dominant aroma attributes in CG-L, whereas Zao-aroma (5.1) and sweet (5.0) aroma were the dominant aroma attributes in DX. In addition, alcoholic (4.8, 5.0) and Chen-aroma (5.5, 4.9) were the dominant aroma attributes in both CG-B and CG-T. However, the rice, fruity and floral aromas prominent in Fangxian Huangjiu (FXHJ) brewed with Xiaoqu were not outstanding in HQHJ (Liu, Zhou, et al., 2022). Moreover, the caramel-like, honey and smoky aromas, which are characteristic aromas in Shaoxing Huangjiu (SXHJ) produced using wheat Qu, were weakly perceived in the four HQHJ samples (Chen et al., 2018). Regarding varying sugar content, the scores of alcoholic aroma showed a positive correlation with sugar concentrations, which differed from previous reports and may be influenced by factors such as the type of Qu and brewing techniques (Yu et al., 2019). However, the scores of sweet aroma were unrelated to sugar content, indicating that the perception of sweet aroma was affected by multiple factors such as interactions between compounds (Liu et al., 2025). Furthermore, Chen aroma also obtained higher scores in the samples with higher sugar including CG-B and CG-T, which might be related to the concentration of the compound contributing to the aroma.

Fig. 1.

Sensory evaluation results of the four Hongqu Huangjiu. (CG-L, Chengang-light type; DX, Danxi Huangjiu; CG-B, Chengang-blue label; CG-T, Chengang-three years of mellowness). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3.2. Identification of aroma-active compounds

Of the four HQHJ samples, CG-L (dry type) and DX (semi-dry type), which had high and low scores for aroma attributes in the sensory evaluation, respectively, were selected to represent Chengang Huangjiu and Danxi Huangjiu, respectively. The aroma compounds in the two brands of HQHJ were extracted by SPE, and their aroma-active compounds were identified via GC–O–MS. Overall, 40 and 52 aroma-active compounds were identified in CG-L and DX, respectively (Table 1). A total of 56 compounds were found, including 14 esters, 5 alcohols, 10 acids, 3 aldehydes, 3 ketones, 4 acetals, 4 furans, 2 phenols, 1 lactone, 5 nitrogen-containing compounds, 3 sulfur-containing compounds, 1 other compound and 1 unknown compound. Of these compounds, 20 odorants were identified for the first time in HQHJ.

Esters, which are the most crucial category of odorants in Huangjiu, impart floral, fruity and sweet aromas (Wang et al., 2020). These compounds are produced through microbial metabolism or esterification of alcohols and acids (Yin et al., 2023). Six ester compounds, including ethyl 2-hydroxy-3-methylbutanoate, ethyl 4-oxopentanoate, ethyl benzoate, ethyl phenylacetate, phenethyl acetate and ethyl 3-phenylpropanoate, were identified in DX. The concentrations of these compounds were higher in DX than in GC-L. Diethyl malate (intensity: 5.0) had the highest aroma intensity in both HQHJ samples, displaying wine and fruity aromas. In CG-L, diethyl 2-hydroxypentanedioate with a cotton candy-like aroma had the second highest intensity score (3.0). Besides, ethyl benzoate (5.0) and diethyl butanedioate (5.0) exhibited floral and fruity, wine, fruity and sweet aroma, respectively, in DX.

Alcohols, including 2-methyl-1-propanol, 1-butanol, 3-methyl-1-butanol, benzyl alcohol and β-phenethyl alcohol are common odorants in Huangjiu (Yan et al., 2022). These compounds are produced through amino acid biosynthesis or pyruvate catabolism (Hazelwood et al., 2008), and their contents depend on the raw material, saccharification and fermentation agent, temperature, and post-fermentation process (Peng et al., 2022). β-phenethyl alcohol (5.0; honey and rose aromas) showed the highest aroma intensity in CG-L and DX. This compounds is likely produced from phenylalanine under anaerobic conditions (Huang et al., 2022). The aroma intensity of benzyl alcohol (5.0; sweet and floral aromas) and 3-methyl-1-butanol (5.0; fusel and fruity aromas) was the highest in CG-L and DX, respectively. However, the aroma of 3-methyl-1-butanol from leucine (Huang et al., 2022) was not perceived in CG-L.

Acids not only act as flavor-imparting compounds in Huangjiu but also play an essential role in inhibiting mixed bacteria. They are mainly produced through amino acid metabolism (Wang, 2010). Herein, nine acids were identified, and most of them had an aroma intensity of higher than 2 in both CG-L and DX. Decanoic acid and octanoic acid were perceived in only CG-L and DX, respectively. Short-chain fatty acids such as acetic acid, 2-methylpropionic acid and butyric acid mostly exhibited an acidic aroma. As the chain length increased, hexanoic acid, octanoic acid and nonanoic acid exhibited a cheese-like aroma, whereas decanoic acid presented an unpleasant rancid aroma. All of these acids have been identified as aroma-active compounds in other types of Huangjiu (Wang et al., 2020; Zhang et al., 2024), and significantly contribute to the acidic aroma in HQHJ.

Three aromatic aldehydes were identified, including benzaldehyde, 2-phenyl-2-butenal and vanillin. These compounds facilitate the release of aromas and impart almond and sweet aromas to Huangjiu (Chen, Xu, & Qian, 2013; Yu et al., 2020). Vanillin had the highest aroma intensity in CG-L and DX (3). It is likely produced through the conversion of its precursor, 4-vinylguaiacol, and contributes a sweet aroma to Huangjiu (Vanbeneden et al., 2008). Benzaldehyde has been identified as one of the three essential aldehydes in Huangjiu (Yu et al., 2020). However, it showed a relatively weak aroma intensity in both CG-L (1) and DX (2). Furthermore, three ketones (3-penten-2-one, 1-octen-3-one and acetophenone) were identified, which imparted musty, mushroom, sweet and floral aromas. Four and three acetals were identified in CG-L and DX, respectively, imparting a fruity aroma. Furthermore, furans such as ethyl furfuryl ether (average intensity: 2.0), furfural (2.0), 2-acetylfuran (2.0) and 2-acetyl-5-methylfuran (2.0) had relatively weak aroma intensities in HQHJ. Furfural, which was detected in both CG-L and DX, is primarily generated through the Maillard reaction, caramelisation and carbohydrate degradation (Pereira et al., 2011), presenting bread, almond, nutty and sweet aromas. Phenols are considered one of the functional components of Huangjiu, exhibiting antioxidant, anti-ageing, immune-enhancing and antibacterial properties (Sun, 2019). 4-Ethylphenol and 2-methoxy-4-vinylphenol had the highest aroma intensities in DX (5.0), imparting smoky, herbal, lilac, curry and spicy aromas. Among lactones, 4-hydroxy-2,3-dimethyl-2H-furan-5-one (sotolon) was the only detectable odorant, providing an herbal aroma and exhibiting the highest aroma intensity (5.0) in CG-L and DX. Sotolon is an important contributor to the empty cup aroma of soy sauce aroma type baijiu (Qin et al., 2024), and its olfactory mechanism has been reported (Wang et al., 2024). The odour threshold of sotolone is low (5 μg/L in a 14% ethanol solution) (Van Gemert, 2011). Sotolon has been detected in various types of Huangjiu and is a significant aroma-active compound in Huangjiu (Chen, Wang, & Xu, 2013; Wang et al., 2022; Zhang et al., 2024).

Five nitrogen-containing compounds were identified in HQHJ, mainly generated through the Maillard reaction (Wang et al., 2020). 2-Acetyl-1-pyrrole, which imparts nutty, walnut and bread-like aromas, showed a notably strong aroma intensity in CG-L (5.0). Futhermore, three sulfur-containing compounds were identified in GC-L and DX. These compounds are mostly produced through microbial metabolism and sulfur-containing amino acid degradation (Aurelie et al., 2011). Notably, difurfuryl sulfide was identified for the first time in Huangjiu, with a strong aroma intensity only in CG-L. It is generated through the oxidation of furfuryl mercaptan (Zhang et al., 2014) and imparts a savoury aroma, such as sofu, sauce, salty and broad bean paste–like aroma.

3.3. Quantitative analysis

Aroma-active compounds were quantified using the internal standard curve method and internal standard method. In the four HQHJ samples, esters (average, 50,625.04 μg/L) had the highest overall concentration, followed by alcohols (35,503.81 μg/L) (Fig. 2, Table 2). Similar trends have been observed in FXHJ produced using Xiaoqu (Liu, Wang, et al., 2022). CG-B (semi-sweet, 135,684.84 μg/L) had the highest total content of odorants, followed by DX (semi-dry, 119,376.59 μg/L), GC-T (sweet, 108,124.20 μg/L) and CG-L (dry, 82,885.77 μg/L). This difference could be attributed to the fact that the dry and semi-dry types of Huangjiu are brewed with water, and the volume of water added to the dry type is greater than that added to the semi-dry type (Xie, 2016). Sweet Huangjiu is fermented with Baijiu or edible alcohol instead of water. The high alcohol content in this Huangjiu type likely inhibits the growth and metabolic activity of certain microorganisms, influencing the overall aroma (Qian et al., 2022; Xie, 2016). Semi-sweet Huangjiu is fermented with aged Huangjiu instead of water, which imparts various types of aromas to the final product (Xie, 2016; Yu et al., 2023).

Fig. 2.

Heatmap of quantitative results of the four Hongqu Huangjiu.

Quantitative analysis indicated that ethyl acetate (42,342.33 μg/L) had the highest average content in the four HQHJ samples, followed by β-phenethyl alcohol (23,664.12 μg/L), 1,1-diethoxyethane (15,687.18 μg/L), 1-butanol (8250.09 μg/L) and ethyl 4-oxopentanoate (3820.19 μg/L). Ethyl acetate was the most abundant compound in each HQHJ. Compared with young SXHJ made from wheat Qu, HQHJ has lower concentrations of ethyl acetate and β-phenylethyl alcohol. However, the concentration of 1,1-diethoxyethane is similar between them (Chen, Wang, et al., 2019). With regard to Huangjiu with different sugar contents, the concentration of ethyl acetate is significantly higher in CG-L (34,875.95 μg/L) and CG-T (45,653.01 μg/L) than in Hubei Huangjiu (dry) and Shanxi Huangjiu (sweet) (Liu et al., 2022a; Wang et al., 2020) but is considerably lower in DX (42,570.99 μg/L) than in Zhejiang Huangjiu (semi-dry) (Chen, Xu, & Qian, 2013). Moreover, the concentration of ethyl acetate in CG- in CG-B (46,269.37 μg/L) is similar to that in Jiangsu Huangjiu (semi-sweet) (Liu, Zhou, et al., 2022). The content of β-phenethyl alcohol in CG-L (15,991.92 μg/L), CG-B (32,075.00 μg/L), and CG-T (16,447.90 μg/L) were 3–11 folds higher than that in Hubei Huangjiu (dry and sweet) and Shandong Huangjiu (semi-sweet) (Fan et al., 2021; Liu, Zhou, et al., 2022). However, the content of β-phenethyl alcohol in DX (30,141.66 μg/L) is only half of that in Jiangsu Huangjiu (semi-dry) (Liu et al., 2022b). The internal standard curve method was used to quantify 1,1-diethoxyethane in HQHJ, which was found to significantly contribute to the overall aroma profile of HQHJ with different sugar contents for the first time. The content of 1,1-diethoxyethane in HQHJ was considerably higher than that in Jiangxi Huangjiu (Zhang et al., 2024) and substantially lower than that in Zhejiang Huangjiu (Chen, Liu, et al., 2019).

Furthermore, the contents of 1-butanol (8250.09 μg/L), ethyl 4-oxopentanoate (3820.19 μg/L), ethyl isobutyrate (2536.85 μg/L), nonanoic acid (2146.93 μg/L), benzaldehyde (1975.85 μg/L), 2-methyl-1-propanol (1969.22 μg/L), 2-methoxy-4-vinylphenol (1743.76 μg/L) and 3-methyl-1-butanol (1529.09 μg/L) were relatively high in HQHJ. Compared with SXHJ with the same sugar content, HQHJ contains higher concentrations of 1-butanol, ethyl isobutyrate and 2-methyl-1-propanol and lower concentrations of benzaldehyde and 3-methyl-1-butanol. However, ethyl 4-oxopentanoate, nonanoic acid and 2-methoxy-4-vinylphenol are not detected in SXHJ (Yu et al., 2019). The concentrations of furfural (447.01–782.91 μg/L) and diethyl butanedioate (21.62–655.24 μg/L) are lower in HQHJ than in SXHJ, with furfural showing a significantly higher concentration in sweet and semi-sweet HQHJ and SXHJ than in dry and semi-dry HQHJ and SXHJ. Diethyl butanedioate has the highest concentration in semi-dry HQHJ and SXHJ (Yu et al., 2019). The varying concentrations of these compounds impart distinctive aromas to different types of HQHJ.

In these four HQHJ samples, as the sugar content increased, the concentrations of ethyl isobutyrate, furfural, and 2-acetylfuran also rose, which may be related to the high-sugar environment providing a more favourable condition for the metabolism of yeasts and the Maillard reaction (Deng et al., 2023; Kanzler et al., 2017)). Conversely, the concentration of 2-ethyl-6-methylpyrazine decreased with increasing sugar content, which may be associated with the metabolism of microorganisms such as bacteria and the generation of precursor substances (Yang et al., 2020). The other compounds had the highest concentrations in DX or CG-B, suggesting a non-linear relationship between their content and sugar content. This revealed that sugar not only serves as a precursor for the production of aroma compounds, but also participates in the dynamic regulation of the entire fermentation ecosystem. Under the sugar content in DX or CG-B, it may be most conducive to mutualistic symbiosis among microorganisms, thereby achieving optimal production of aroma compounds (Chen, Liu, et al., 2019).

3.4. Calculation of OAVs

To assess the contribution of the identified odorants to the aroma profile of Huangjiu, OAVs were calculated (Table 3). Overall, 14, 19, 16 and 15 aroma-active compounds with OAVs of ≥1 were identified in CG-L, DX, CG-B and CG-T, respectively. β-phenethyl alcohol (average: 11,832,060) showed the highest OAV in the four HQHJ samples, followed by dimethyl trisulfide (473), 1,1-diethoxyethane (314) and ethyl isobutyrate (169). These compounds are important contributors to the overall aroma of HQHJ.

Table 3.

Odour thresholds and odour activity values of key odorants in Hongqu Huangjiu.

No.	Compounds	Thresholds (μg/L)	OAVse				Averagef
			CG-L	DX	CG-B	CG-T
19	β-Phenethyl alcohol	0.002a	7,995,960	15,070,828	16,037,499	8,223,952	11,832,060
52	Dimethyl trisulfide	0.2c	–	234	631	553	473
36	1,1-Diethoxyethane	50a	240	361	358	296	314
3	Ethyl isobutyrate	15c	93	132	221	230	169
10	Phenethyl acetate	5b	42	80	54	52	57
48	2-Ethyl-6-methylpyrazine	1.68d	59	52	–	–	55
4	Ethyl butanoate	20a	31	35	73	51	48
45	2-Methoxy-4-vinylphenol	40b	–	44	–	–	44
11	Ethyl 3-phenylpropanoate	5a	–	46	38	26	37
7	Ethyl 4-oxopentanoate	350c	6	4	20	14	11
39	1,1-Diethoxy-3-methylbutane	78.28d	3	10	13	9	9
37	1,1-Diethoxy-2-methylpropane	21.54d	2	10	10	6	7
23	3-Methylbutanoic acid	8b	2	9	3	6	5
30	Benzaldehyde	515b	3	5	4	4	4
1	Ethyl acetate	15000a	2	3	3	3	3
12	Ethyl phenylacetate	73a	1	2	5	3	3
27	Nonanoic acid	1100a	2	2	2	2	2
32	Vanillin	29.9b	<1	5	<1	<1	1
38	2,4,5-Trimethyl-1,3-dioxolane	900c	<1	1	<1	<1	<1
47	2,6-Dimethylpyrazine	400c	<1	<1	2	–	<1
22	Butyric acid	173c	<1	<1	<1	<1	<1
25	Hexanoic acid	806.5a	<1	<1	<1	<1	<1
18	Benzyl alcohol	900b	<1	<1	<1	<1	<1
35	Acetophenone	3000b	<1	–	<1	<1	<1
26	Octanoic acid	500c	<1	<1	<1	<1	<1
44	4-Ethylphenol	140a	–	<1	<1	<1	<1
21	2-Methylpropionic acid	2300c	<1	<1	<1	<1	<1
17	3-Methyl-1-butanol	30,000c	<1	<1	<1	<1	<1
15	2-Methyl-1-propanol	40,000a	<1	<1	<1	<1	<1
41	Furfural	14100c	<1	<1	<1	<1	<1
31	2-Phenyl-2-butenal	20000c	<1	<1	<1	<1	<1
9	Diethyl butanedioate	14,417.5a	<1	<1	<1	<1	<1
20	Acetic acid	24,000a	<1	<1	<1	<1	<1
29	Benzeneacetic acid	2500b	<1	<1	<1	<1	<1
2	Ethyl propionate	5000c	–	<1	<1	–	<1
16	1-Butanol	820000c	<1	<1	<1	<1	<1
50	2-Formyl-1H-pyrrole	65000c	<1	<1	<1	<1	<1
49	2-Acetyl-1H-pyrrole	58600c	<1	<1	<1	<1	<1
24	Pentanoic acid	11000c	<1	<1	<1	<1	<1
8	Ethyl benzoate	575c	<1	<1	<1	<1	<1
13	Diethyl malate	10000c	<1	<1	<1	<1	<1
28	Decanoic acid	15000c	<1	<1	<1	<1	<1
42	2-Acetylfuran	10000c	<1	<1	<1	<1	<1
6	Ethyl 3-hydroxybutanoate	67000c	<1	<1	<1	<1	<1
43	2-Acetyl-5-methylfuran	40870c	<1	<1	<1	<1	<1
51	2,6-Diethylpyrazine	–	–	–	–	–	–

^aOdour thresholds from reference (Wang et al., 2022).

^bOdour thresholds from reference (Wang et al., 2020).

^cOdour thresholds from reference (Van Gemert, 2011).

^dOdour thresholds determined in 14% ethanol/water artificial matrix solution in Key Laboratory of Brewing Molecular Engineering of China Light Industry in Beijing Technology & Business University.

^eOAVs represents odour activity values; CG-L, DX, CG-B, and CG-T respectively represent Chengang-light type, Danxi Huangjiu, Chengang-blue label, Chengang-three years of mellowness.

^fAverage odour activity value.

As ethanol content is relatively high in Huangjiu, ethyl esters are commonly considered important aroma compounds in Huangjiu with low odour thresholds (Yin et al., 2023). In the four HQHJ samples, seven ethyl esters showed an OAV of higher than 1, with most of them having higher OAVs than other Huangjiu types. For example, the OAV of ethyl isobutyrate (169) in HQHJ is higher than that in aged Huangjiu produced by wheat Qu. Ethyl isobutyrate is considered an essential contributor to fruity aroma (Chen et al., 2019b). Phenylethyl acetate (57), ethyl 3-phenylpropanoate (37) and ethyl acetate (3) are crucial odorants in Shanxi Huangjiu fermented with Daqu and Jiangxi Huangjiu fermented with Xiaoqu (Wang et al., 2020; Zhang et al., 2024). The compounds had higher OAVs in HQHJ. In addition, ethyl butanoate (48) is an important aroma compound in semi-dry and semi-sweet types of Jiangsu Huangjiu (Liu et al., 2022b). However, its OAV also in Jiangsu Huangjiu is lower than that in HQHJ. Notably, ethyl 4-oxopentanoate (11) was identified as a crucial odorant in HQHJ for the first time. Ethyl phenylacetate (3) was also a characteristic aroma compound in SXHJ with different sugar contents (Yu et al., 2019), contributing to sweet and floral aromas (Wang et al., 2020). It had the highest OAV in CG-B (5) and the lowest OAV in CG-L (1).

Although alcohols had high concentrations, β-Phenethyl alcohol (average OAV: 11,832,060) displayed an OAV of >1, mainly because of the high odour thresholds of alcohols. β-phenethyl alcohol had the highest OAV in CG-B (16,037,499) and the lowest OAV in CG-L (7,995,960), which differed from the maximum OAV in dry SXHJ and the minimum OAV in sweet SXHJ (Yu et al., 2019). With regard to acids, only 3-methylbutanoic acid (5) and nonanoic acid (2) had OAVs of >1. The average OAV of nonanoic acid was two in the four HQHJ samples, whereas it was <1 in the sweet type of Jiangxi Huangjiu and the semi-sweet type of coarse cereal Huangjiu (Wang et al., 2022; Zhang et al., 2024).

Dimethyl trisulfide was not detected in CG-L. However, it had relatively high OAVs in the other three HQHJ samples (average OAV: 473), which were significantly higher than those in Zhejiang Huangjiu and Jiangsu Huangjiu (Chen et al., 2018). The odorant 2-methoxy-4-vinylphenol (44) was detected only in DX, whereas vanillin (5) and 2,4,5-trimethyl-1,3-dioxolane (1) showed an OAV of greater than 1 only in DX. 2-Methoxy-4-vinylphenol had a markedly higher OAV in DX than in Shanxi Huangjiu (Wang et al., 2020), whereas vanillin had a lower OAV in DX. In addition, 1,1-diethoxyethane (314) and benzaldehyde (4) had higher OAVs in HQHJ than in Jiangxi Huangjiu (Zhang et al., 2024). For the three compounds with self-measurement thresholds, 2-ethyl-6-methylpyrazine had an OAV of greater than 1 in CG-L and DX (59 and 52, respectively), and the average OAVs of two acetals, namely, 1,1-diethoxy-3-methylbutane and 1,1-diethoxy-2-methylpropane, were 9 and 7, respectively. Both compounds had a high OAV in CG-B and DX and the low OAV in CG-L.

3.5. Correlation analysis

To clarify the differences between HQHJ samples with varying sugar content, PLSR was employed to explore potential association patterns between sensory attributes and odorants in four HQHJ samples (Pu et al., 2019). Before analysis, the X and Y variables were centred and standardized to have a unit variance. The optimal number of latent components was determined to be two through an iterative process of leave-one-out (LOO) cross-validation. In PLSR, components are extracted to maximize the covariance between X and Y (Wang et al., 2022). Therefore, the first component (Dim1) represents the strongest direction of correlation, followed by the second (Dim2). The correlation loading plots of the sensory data and aroma-active compounds in the four HQHJ samples are presented in Fig. 3A, and these plots are based on dimension 1 and dimension 2. The cumulative R²Y (0.782) and R²X (0.856) values of the two main components corresponding to the correlations between the explanatory (X, aroma-active compounds) and dependent (Y, sensory data) variables were close to one, which indicated that the model possessed a high explanatory ability for the data. Q² was 0.111 and may be related to the small number of samples, indicating that the model exhibited certain limitations in its predictive capability. Therefore, the establishment of the PLS-R model in this study was intended to explore potential trends between the sensory attributions and aromatic active compounds in the four HQHJs, rather than to create a robust predictive model.

Fig. 3.

(A) Correlation loading plots (based on Dim1 and Dim2) between the aroma compounds (X) and sensory properties (Y) among four Hongqu Huangjius (HQHJs). CG-L, CG-B, CG-T and DX were represented Chengang-light type Huangjiu, Chengang-blue label Huangjiu, Chengang-three years of mellowness Huangjiu, and Danxi Huangjiu, respectively. The compounds corresponding to the numbers were as follows: 1: Ethyl acetate; 2: Ethyl propionate; 3: Ethyl isobutyrate; 4: Ethyl butanoate; 6: Ethyl 3-hydroxybutanoate; 7: Ethyl 4-oxopentanoate; 8: Ethyl benzoate; 9: Diethyl butanedioate; 10: Phenethyl acetate; 11: Ethyl 3-phenylpropanoate; 12: Ethyl phenylacetate; 13: Diethyl malate; 15: 2-Methyl-1-propanol; 16: 1-Butanol; 17: 3-Methyl-1-butanol; 18: Benzyl alcohol; 19: Phenethyl alcohol; 20: Acetic acid; 21: 2-Methylpropionic acid; 22: Butyric acid; 23: 3-Methylbutanoic acid; 24: Pentanoic acid; 25: Hexanoic acid; 26: Octanoic acid; 27: Nonanoic acid; 28: Decanoic acid; 29: Benzeneacetic acid; 30: Benzaldehyde; 31: 2-Phenyl-2-butenal; 32: Vanillin; 35: Acetophenone; 36: 1,1-Diethoxyethane; 37: 1,1-Diethoxy-2-methylpropane; 38: 2,4,5-Trimethyl-1,3-dioxolane; 39: 1,1-Diethoxy-3-methylbutane; 41: Furfural; 42: 2-Acetylfuran; 43: 2-Acetyl-5-methylfuran; 44: 4-Ethylphenol; 45: 2-Methoxy-4-vinylphenol; 47: 2,6-Dimethylpyrazine; 48: 2-Ethyl-6-methylpyrazine; 49: 2-Acetyl-1H-pyrrole; 50: 2-Formyl-1H-pyrrole; 51: 2,6-Diethylpyrazine; 52: Dimethyl trisulfide. (B) Average variable importance for the projection (VIP) values for 19 aroma compounds. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

In dimension 1, CG-B, CG-T and CG-L were located in the positive semi-axis, whereas DX was located in the negative semi-axis. In dimension 2, CG-B, CG-T and DX were located in the positive semi-axis, while CG-L was located in the negative semi-axis. Results indicated that CG-B and CG-T were the most similar. In addition, alcoholic, grain and Chen aromas were positively correlated with dimension 1 and were also associated with CG-B and CG-T. However, DX showed stronger correlations with acidic, sweet, Jiang, herbal, Qu and Zao aromas. A majority of the aroma-active compounds were associated with dimension 2 and showed positive correlations with the aromas of CG-T, CG-B and DX. Of the 46 odorants analysed, 19 compounds with average variable importance in projection (VIP) scores of more than 1.0 were statistically significant (p ≤ 0.05; Fig. 3B) (Chong and Jun, 2005). Furthermore, four of them with OAVs of >1 (2-methoxy-4-vinylphenol, vanillin, phenethyl acetate and ethyl 4-oxopentanoate) were preliminarily considered potential important compounds contributing to aroma differences among HQHJ with different sugar contents. These results provided tentative insights into the relationships among the variables in this study.

3.6. Addition tests

To verify the contribution of 2-methoxy-4-vinylphenol, vanillin, phenethyl acetate and ethyl 4-oxopentanoate (OAV > 1 and VIP > 1) to the aroma profiles of HQHJ, ethyl 4-oxopentanoate was added to DX to obtain a concentration equal to that in CG-B, whereas 2-methoxy-4-vinylphenol, phenethyl acetate and vanillin were separately added to CG-B to obtain concentrations equal to those in DX (based on their concentrations in different Huangjiu types). Subsequently, the samples were subjected to sensory evaluation. As shown in Fig. 4, the Chen-aroma of DX significantly increased after adding ethyl 4-oxopentanoate (Fig. 4A). The Chen-aroma is an important sensory attribute for evaluating the quality or grade of Huangjiu. Therefore, ethyl 4-oxopentanoate could be considered a key odorant in HQHJ. The addition of 2-methoxy-4-vinylphenol enhanced the intensity of sweet, Qu-aroma, herbal, Jiang-aroma, grain aromas in CB-B (Fig. 4B), indicating that 2-methoxy-4-vinylphenol is also a key odorant in HQHJ, which is consistent with the findings of previous studies (Chen, Wang et al., 2019). However, no significant changes in aroma were observed after the addition of phenethyl acetate or vanillin to HQHJ with different sugar contents (Fig. 4C-D).

Fig. 4.

Comparative analyses of aroma profiles among (A) DX and DX spiked with ethyl 4-oxopentanoate, (B) CG-B and CG-B spiked with 2-methoxy-4-vinylphenol, (C) CG-B and CG-B spiked with phenethyl acetate, (D) CG-B and CG-B spiked with vanillin.

4. Conclusion

The composition of aroma-active compounds was compared among four types of HQHJs with different sugar contents. Overall, 56 odorants were identified in HQHJ through GC–O–MS analysis. Among them, 20 odorants were newly identified in HQHJ. The odorant difurfuryl sulfide, which imparts sofu, sauce, salt and broad bean paste-like aromas to Huangjiu for the first time. β-phenethyl alcohol (11,832,060), dimethyl trisulfide (473), 1,1-diethoxyethane (314) and ethyl isobutyrate (169) with high OAVs were identified as important odorants in HQHJ. In addition, 2-methoxy-4-vinylphenol, vanillin, phenethyl acetate and ethyl 4-oxopentanoate significantly contributed to the aroma of HQHJ. Based on PLSR analysis and addition tests, ethyl 4-oxopentanoate and 2-methoxy-4-vinylphenol were confirmed as key odorants contributing to differences in the aroma profiles of different HQHJ samples. In particular, ethyl 4-oxopentanoate was the key contributor of Chen-aroma in HQHJ. In these four samples, ethyl isobutyrate, furfural, and 2-acetylfuran increased with rising sugar content, while 2-ethyl-6-methylpyrazine decreased as sugar content rose. The other compounds exhibited their highest concentrations in DX and CG-B, indicating that these compound contents and sugar contents were not linearly correlated. Sugar contents indirectly regulated aroma compounds through the brewing process. Semi-sweet HQHJ brewed with aged Huangjiu displayed the highest total aroma compound content, whereas sweet HQHJ exhibited suppression due to the addition of a higher concentration of alcohol during the fermentation process. Among semi-dry HQHJ, most aroma attributes has the highest intensity and number of compounds were the greatest. In the future, interaction studies should be conducted to investigate synergistic or masking effects between different aroma compounds, and molecular docking techniques could be utilized to elucidate the mechanisms of taste perception.

CRediT authorship contribution statement

Siman Zheng: Writing – original draft, Visualization, Software, Methodology, Investigation, Data curation, Conceptualization. Yangyang Huo: Writing – original draft, Visualization, Software, Methodology, Investigation, Data curation, Conceptualization. Juan Wang: Writing – review & editing, Writing – original draft, Methodology, Investigation, Data curation. Tianyi Cao: Validation, Supervision, Methodology. Mingquan Huang: Writing – review & editing, Supervision, Methodology, Investigation, Funding acquisition. Hongqin Liu: Validation, Supervision, Investigation. Jihong Wu: Writing – review & editing, Visualization, Software. Jinglin Zhang: Validation, Methodology. Qing Ren: Validation, Supervision. Dongrui Zhao: Validation, Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that may affect the work described herein.

Appendix A. Supplementary data

Supplementary material

Data availability

Data will be made available on request.