Study On The Flavor Influence Of Dried Pork Slice In Drying And Baking Based On GC-IMS Technique

Dried pork slice(DPS),as a traditional dried meat products with a long history,has high nutritional value and has been welcomed by the consumers in modern time.With the promotion of industrialization trends in traditional food,there is urgent need for a detection technology which can accurately and quickly detect the volatile organic component(VOCs)of DPS to ensure the quality and flavor consistency in processing.Gas chromatography-ion mobility spectrometry(GC-IMS),as a new detection technology of VOCs,has been applied in food industry.In this study,DPS was taken as the research object and GC-IMS technology was used to study the new methods of for monitoring formation,change and control of VOCS in the production process from the point of flavor chemistry and fingerprint of characteristic flavor.The main work of this research is as follows:(1)In order to analyze the quality of DPS in the market,12 kinds of DPS produced by different manufacturers were purchased,all of which were labeled as the original flavor of Jingjiang distinct.The physical and chemical indexes of different grades of DPS specified in GB/T31406-2015 have been tested and graded.The results show that half of the DPS brands are not in line with the packaging grade.Thus,the DPS industry is in urgent need of standardization.(2)In order to compare the advantage and disadvantage of VOC detection,GC-MS,electronic nose and GC-IMS were used to detect the VOCs of 12 brands of DPS.The results showed that GC-MS was more sensitive for alkanes,ethers,aldehydes and acids in DPS,but GC-IMS was more sensitive for aldehydes,heterocyclic compounds(pyrazines and furans),alcohols,ketones and esters.Although the results of the two methods cannot be completely correspondent,GC-MS and GC-IMS can make up for each other’s shortcomings.Electronic nose can respond to alkanes,esters,aldehydes,alcohol and other substances.It can’t effectively detect specific fingerprint spectrum of VOCs.(3)In order to grade DPS products,the mathematical model of grade classification was established.The response values of 10 sensors of the electronic nose are feature extraction by PCA,and the 3 principal components are selected in combination with KNN to establish the model.The established model has a grade accuracy of 89.8%.The GC-IMS characteristic peak intensity is combined with PCA and KNN to establish the grade model that has a grade accuracy of 90%.It is proved that both GC-IMS and electronic nose can be used to quickly identify the grade of DPS.(4)In order to analyze the reasons for flavor difference of VOCs in actual production,the VOCs fingerprints of 90 samples(75 samples in processing and 15 samples after processing)of DPS in process and finished products were detected by GC-IMS.The 62 effective characteristic peaks of GC-IMS have been selected in the processing products,in which intensity of 11 characteristic peaks are quite different,especially the heterocyclic compounds such as pyrazines which are greatly affected by high temperature.In finished products,44 VOCs were identified from 84 characteristic peaks,which flavor substances were more than those of the product in processing.The results demonstrate that the inconsistency of DPS flavor is caused by the uneven temperature and humidity field in the process of drying and baking.(5)In order to study the changing trend of drying temperature,drying time,baking process and VOCs from DPS,the effects of VOCs in drying temperature,drying time and baking process were analyzed.And relationships between VOCs and Maillard reaction degree,L~*a~*b~* value,moisture and free amino acids were tested.The results show that all indexes of DPS tend to be stable after drying at 80℃ for 4 h,and prolonging time had little change.The main changes of VOCs during drying are ester,alcohols and pyrazines formed at high temperature.The GC-IMS characteristic peak intensity was combined with PLSR to establish the model for quantifying the drying time,drying temperature and moisture in processing.The correlation coefficient R_c of moisture correction set was 0.95.The RMSECV was 5.819,and the R_p and RMSEP of prediction set were 0.947 and 7.274.And baking process will increase the number of VOCs of DPS,and add specific flavor characteristics to DPS.It has been proved that GC-IMS technology can not only detect the VOCs changes,but also can be used to predict the drying temperature,time and water content of DPS.(6)For obtaining the best processing parameters of DPS and establish the fitting model of GC-IMS technology and sensory score,the odor sensory score of DPS was selected as evaluation index.A four-factor and three-level response surface analysis model of drying and baking was established.The process parameters of the best flavor fitted by the model are as follows: drying temperature 83.2℃,drying time 5.5 h,baking temperature 192.8℃,and baking time 1.0 min.Three fitting models of PLSR,SWF and SVM combined with GC-IMS characteristic peak intensity and sensory score were established and tested.Among them,PLSR had the highest fitting accuracy,with R_c and RMSEC of 0.9254 and 0.2302,R_p and RMSEP of 0.8761 and 0.2698.The results show that GC-IMS technology can be effectively applied to the prediction of sensory score of dried pork,and it is feasible to realize the numerical and intelligent evaluation of artificial sensory score in the future.