engleski

Preliminary study on determination of fish freshness by Raman spectroscopy

Loss of freshness and gradual occurrence of spoilage in fish during storage can be determined by different methods which are often time / labor consuming or require expensive equipment and skilled personnel. Recent scientific publications emphasize the need for fast and easy-to use methodologies for quality control of fish. In this study, possible application of Raman spectroscopy, as rapid and noninvasive method for determination of fish freshness, was evaluated. Five seabass fillets were stored on ice and 10 Raman spectra per sample were recorded each day during 9 days of storage. Raman spectra were recorded using Horiba Jobin Yvon T64000 triple Raman spectrometer equipped with 514.5 nm excitation wavelength laser in spectral range from 200 – 3800 cm-1. Obtained spectra were analyzed by principal component analysis and several models based on different spectral sub-regions were obtained. Best results were obtained by models based on 900 – 1320 cm-1, 820 – 1660 cm-1 regions and model based on selected dominant Raman bands present in the spectra of samples. In addition, artificial neural network model was constructed and used for prediction of storage time with correlation coefficient of prediction R2 > 0, 90. Obtained results show that PCA models are able to classify fish samples according to storage time with 96, 67 % certainty based on Raman spectra in region from 820 – 1660 cm-1 which includes major part of “fingerprint region”, containing vibrational signatures of molecules present in the sample. Although results are promising, further work is needed to identify key indicators of freshness loss in Raman spectra of fish samples which would enable construction of reliable chemometric models for prediction of freshness / duration of storage of fish.

Raman spectroscopy ; fish ; quality ; freshness

nije evidentirano