Research On Non-destructive And In-situ Detection Of Food Freshness Based On Fiber-optic Evanescent Wave Absorption Spectroscopy

Food quality can be directly affected by its freshness.For some fruits with strong aroma or chilled seafood with strong fishy smell,consumers are not easy to identify the freshness of these foods through their senses.Once consumed by consumers,these unfresh foods will greatly increase the risk of foodborne disease outbreaks.Therefore,non-destructive,in-situ and rapid detection of food freshness is of great significance and an urgent scientific problem to ensure food safety,reduce economic losses and improve human living standards.However,traditional methods of food freshness identification often rely on large laboratory analytical instruments,require sampling,time-consuming and laborious,do not provide immediate data on food freshness.Fiber-optic evanescent wave(FOEW)sensor has been widely used for in-situ detection of liquids,solids and gases in recent years due to its nondestructive,sensitive and miniaturized properties.Therefore,in this paper,Jufeng grapes and Penaeus vannamei were used as the research object,and uses FOEW sensor combined with Fourier transform infrared(FTIR)absorption spectroscopy detection system as the technical means,to carry out research on non-destructive,in-situ and rapid mid-infrared FOEW absorption spectroscopy detection method of food freshness.The main research contents are as follows:(1)A feasibility study was carried out for the online and in situ detection of ethanol released from volatiles without gas collection based on mid-infrared FOEW absorption spectroscopy.The FTIR spectrometer,infrared fiber coupler and spiral FOEW sensor were used to build a complete FOEW sensor system.Using the constructed mid-infrared FOEW absorption spectrometry detection system,the FOEW absorption spectra of volatile compounds released from ethanol solution were acquired in situ,and the characteristic absorption peaks of ethanol were successfully detected,which verified the feasibility and dynamic response characteristics of the system to detect ethanol.The FOEW absorption spectra of volatile compounds from grapes at different storage times were obtained non-destructively and in situ,and specific markers such as ethanol were detected using a FOEW sensor system built for grapes.In addition,the study used principal component analysis(PCA)to downscale the spectral data and partial least squares discriminant analysis(PLS-DA)qualitative prediction model algorithm to model the spectral data of fresh and spoiled grapes,which successfully classified grapes of different freshness classes with accuracy of both calibration and validation sets under the condition of modeling using a total of 40 collected spectra 100.00 %.The experimental results show that midinfrared FOEW absorption spectroscopy can be used for rapid and effective detection of food quality by food volatiles,and that the method is not dependent on light range,does not require pumping into the sample,and has the potential to identify fruit freshness grades nondestructively,in situ,and remotely.(2)Shrimp and other crustacean seafood due to the shell wrapping,resulting in volatiles and not easy to volatilize out.Meanwhile,the deterioration process of shrimp freshness will destroy the relatively stable exoskeletal structure of the shell,which leads to the precipitation of chitin,protein and calcite onto the surface of the shell.In this section,a non-destructive,in-situ and rapid detection method of shrimp freshness was proposed using mid-infrared FOEW absorption spectroscopy of the precipitated material on the shell surface.The mid-infrared FOEW spectra were obtained for different storage times of shrimp samples at indoor temperature(22°C)and in a refrigerated environment(4°C),and the characteristic peaks of chitin,protein and calcite,which are related to the freshness of shrimp,were successfully detected.In addition,PCA and orthogonal partial least squares discriminant analysis(OPLS-DA)were used to downscale the data and successfully classify the shrimp samples into “fresh”,“acceptable” and“spoiled” with reference to the TVB-N values obtained simultaneously.Moreover,combined with the PLS-DA algorithm,a fast qualitative prediction model for the freshness of shrimp samples was established,with the recognition rates of shrimp freshness in the final calibration set and verification set were 87.27 % and 90.28 %,respectively,under the condition that a total of 576 spectra were collected for modeling.The above analysis shows that this method can realize non-destructive and rapid assessment of shrimp freshness only by contacting the micro FOEW sensor with the shell of the shrimp,which has great potential in non-destructive,in situ and rapid recognition surface of shrimp freshness.In summary,the method proposed in this paper can realize non-destructive,in situ and rapid detection of food freshness without sample processing,enrich the technical system of food freshness identification,and provide methods and technical support for rapid identification of food quality and ensuring food safety.