四川省酒业集团有限责任公司

To gain an in-depth understanding of the flavor components of black rice wine from Yang county of Shanxi province,the volatile flavor components of three different black rice wines from Yang County(A,B C)were detected using headspace solid-phase microextraction coupled with gas chromatography-mass spectrometry(HS-SPME-GC-MS).The key volatile aroma components were screened by odor activity value(OAV),and the key differential aroma compounds were screened by orthogonal partial least squares-discriminant analysis(OPLS-DA)based on variable importance in pro-jection(VIP)value.The results showed that a total of 111 volatile flavor compounds were identified in the three kinds of black rice wines from Yang county of Shanxi province,with samples A,B,and C containing 75,75,and 85 volatile aroma compounds,respectively.Among them,12 key volatile aroma compounds(OAV ≥ 1)were identified,including ethyl butyrate,ethyl 2-methylbutyrate,2-phenylethanol,and damascenone,with samples A,B,and C containing 11,9,and 9 compounds,respectively.The contents of ethyl butyrate,isoamyl acetate,and 2-phenylethanol exhibited significant dif-ferences(P＜0.05)among the different black rice wines from Yang county of Shanxi province.By combining VIP values,2-phenylethanol and acetoin(3-hydroxy-2-butanone)were identified as the key differential volatile flavor substances(OAV ≥ 1,VIP＞1)distinguishing the different black rice wines.The study provided a theoretical basis for further research on the aroma components of black rice wine from Yang county of Shanxi province as a geo-graphical indication product.

Activated carbon,owing to its highly developed pore structure,is widely employed in the purification and aging acceleration of Baijiu.However,existing research predominantly focuses on process optimisation for activated carbon models,and there is a lack of systematic analysis of influence of how pore structure parameters on Baijiu quality during adsorption process.In this study,using 11 types of activated carbon for liquor(1#-11#)and different quality of strong-flavor(Nongxiangxing)Baijiu as research subjects,a quality evaluation model was constructed,and the effect of activated carbon pore size on Baijiu quality and the removal patterns of off-flavors were analyzed.The results demonstrated that the actual quality grade of Baijiu was completely consistent with the model(100％concordance)by analytic hierarchy process.After treatment with the 11 activated car-bon,the off-flavors such as bitterness,astringency,and foul odours were markedly reduced,the total acid,total ester and volatile flavor components contents in Baijiu all decreased significantly(P＜0.05),while the solid content showed no significant change(P＞0.05).In addition,the sensory score was significantly improved(P＜0.05).Further analysis revealed that mesoporous activated carbon demonstrated superior efficacy in enhancing Baijiu quality compared to microporous activated carbon.These findings provided scientific theory and technical support for the improvement of the quality level of strong-flavor Baijiu by activated carbon for alcohol.

Using different rounds of sauce-flavor(Jiangxiangxing)Baijiu as research object,the volatile flavor components were analyzed using gas chromatography-mass spectrometry(GC-MS)and gas chromatography with flame ionization detection(GC-FID).Key flavor compounds were screened based on odor activity value(OAV),and partial least squares-discriminant analysis(PLS-DA)model was established to differentiate the base liquors.Differential flavor compounds were screened according to the variable importance in projection(VIP)values,aiming to investigate the flavor charac-teristics and differences among base liquors from different rounds.The results showed that the alcohol content of the base liquor from round 1 to 7 ranged from 53.01％vol to 56.48％vol,the total acid content ranged from 2.53 g/L to 5.02 g/L,and the total ester content ranged from 3.20 g/L to 5.21 g/L.A total of 75 flavor compounds were detected in the 7th round base liquor,including 31 esters,20 alcohols,13 acids,5 aldehydes,2 ketones,2 phenols,1 furan,and 1 pyrazine.The highest number of volatile flavor compounds(62)was found in the 1st-round base liquor.Esters,alcohols,and acids accounted for a relatively large proportion in each round.Through OAV analysis,41 key flavor compounds(OAV ≥ 1)were identified,among which esters such as ethyl acetate,ethyl lactate,and ethyl butyrate contributed significantly to the flavor of the base liquor from each round.PLS-DA could effectively distinguish the base liquors from different rounds,and 17 differential volatile flavor compounds(VIP＞1)were screened based on VIP values.

A rapid and non-destructive method for predicting the alcohol content of original liquor of strong-flavor(Nongxiangxing)Baijiu was devel-oped using near-infrared(NIR)spectroscopy combined with quantitative modeling techniques.Infrared spectral data of 129 original liquor samples of strong-flavor Baijiu were collected,and 6 spectral pretreatment methods(mean centralization,convolution smoothing,first derivative,second deriva-tive,standard normal transformation,and multivariate scattering correction)were compared.The alcohol content prediction models were constructed by using 5 modeling methods:multiple linear regression(MLR),principal component regression(PCR),partial least squares regression(PLSR),sup-port vector machine regression(SVR),and backpropagation artificial neural network(BP-ANN).The results showed that when the standard normal transform was combined with the MLR model,the coefficient of determination R2t of the training set was 0.983 1,the coefficient of determination R2v of the validation set was 0.967 3,and the residual prediction deviation(RPD)was 5.602 1,demonstrating a good predictive ability.In addition,the first-order derivative combined with the SVR model also showed a relatively high prediction accuracy.The coefficient of determination R2t of the training set was 0.971 2,the coefficient of determination R2v of the validation set was 0.980 5,and the residual prediction deviation(RPD)was 7.257 3.In conclusion,the method based on the combination of near-infrared spectroscopy and machine learning could effectively predict the alcohol content of the original liquor of strong-flavor Baijiu,and had good application potential.It could provide a new rapid non-destructive detection means for the alcohol content analysis of the original liquor of Baijiu.