研究目的
To evaluate the feasibility of using near infrared spectroscopy combined with chemometric approaches to assess the shelf-life of cloudy apple juice, particularly focusing on the effects of sonication treatments and spoilage over time.
研究成果
NIR spectroscopy combined with chemometrics is a feasible tool for monitoring the shelf-life of apple juice, identifying spoilage thresholds at 7 days for non-sonicated and 14 days for sonicated samples. The method shows promise for non-destructive quality control but requires further validation with larger datasets.
研究不足
The study is exploratory and based on a limited number of samples; discrimination between sonicated and non-sonicated juices was not fully achieved, indicating a need for a larger sample size and more robust calibration models. The NIR technique may not capture all aspects of spoilage, and the results are specific to the experimental conditions used.
1:Experimental Design and Method Selection:
The study involved two experiments: a storage test comparing sonicated and non-sonicated non-inoculated juices, and a challenge test with inoculated juices subjected to sonication. NIR spectroscopy at 1300–2500 nm was used with chemometric analyses including PCA, PLS-DA, and CDA to monitor shelf-life.
2:Sample Selection and Data Sources:
Juice samples from Golden delicious apples (Melinda protected designation of origin 4021) were used. Samples included non-inoculated non-sonicated (NINS), non-inoculated sonicated (NIS), inoculated non-sonicated (INS), and inoculated sonicated (IS) juices, analyzed at T0, T7, T14, T21, T28, and T60 days under refrigeration.
3:List of Experimental Equipment and Materials:
UP 400 S sonicator (Hielscher USA, Inc.), SpectraStar 2500TW spectrometer (Unity Scientific), glass beakers (Duran Group), sterile polypropylene containers, ice for temperature control, and yeasts (Rhodotorula glutinis and Candida parapsilosis) for inoculation.
4:Experimental Procedures and Operational Workflow:
Sonication was applied for 360 s per 100 mL at 100% amplitude with temperature control. Inoculation was done with a mix of yeasts at 3 log10 CFU/mL. NIR spectra were recorded in transreflectance mode with duplicate scans averaged. Data were processed using Matlab and PLS Toolbox for PCA, PLS-DA, and CDA with various pre-treatments.
5:Data Analysis Methods:
Statistical analyses included PCA to explore spectral features, fitting functions to model PC1 trends, PLS-DA for classification with cross-validation, VIP index for wavelength importance, and CDA for sample segregation based on storage time.
独家科研数据包,助您复现前沿成果,加速创新突破
获取完整内容