Abstract
The objective of this thesis was to test the potential application of Raman spectroscopy for quality monitoring in the meat processing industry. Raman spectroscopy is a spectroscopic technique, which can provide rapid chemical information at the molecular level. It can, therefore, be used for non-destructive quality monitoring in the meat processing industry as it requires no further sample preparation. Water, which is the main component of meat, is also not a problem for Raman spectroscopic measurement because water is a poor Raman scatter. Two major product quality related issues, cooking of meat and fat quality were selected as focus areas: Endpoint Temperature (EPT) and gross fatty acid (FA) composition. Endpoint temperature of a cooked product is an important parameter as it is related to the microbial safety and palatability of the product. The project aimed at using Raman spectroscopy for authentication of the cooked
then chill stored samples to determine if the EPT of the cooked meat reached to the intended EPT or not. For this purpose, pork longissimus dorsi was heat treated at 50 oC to 70 oC (2 oC interval) for 2 to 10 h (2 h interval) (paper-I) and at 50 oC to 80 oC (2 oC interval) for 2 h in water bath then stored under vacuum packaging for day- 0, -4 and -8 (paper-II). Secondly, the thesis work examined the use of Raman spectroscopy in combination with chemometrics to determine the gross FA parameters and individual FAs in pork backfat. Non-targeted predictions of highly
collinear reference variables were also investigated. Backfat samples for this study were provided from Danish Meat Research Institute (DMRI) as part of a large feeding project.
Raman spectroscopy was able to predict EPT of cooked intact muscle with high cross-validated squared coefficient of determination ( = 0.96) and an acceptable accuracy (Root Mean Error of Cross-Validation (RMSECV = 1.78 oC) (Paper-I and II) independent of the storage time (day- 0, -4 and -8) (Paper-II). Furthermore, the results showed that Raman spectroscopy was useful to classify cooked meat with EPT of below and above 65 oC even after the storage time up to 8 days (Paper-II). This EPT was chosen as a critical point as most pathogenic microorganisms are expected to die at this temperature (at a given time). In contrast, Raman spectroscopy was not able to predict cooking time (2-10 h) of the intact muscle (Paper-I). Scrutinizing the Raman spectra provided detailed information about the structural changes of the meat proteins during cooking (Paper-I and II) and demonstrated that cooking temperature has a much higher impact on the structural changes in meat compared to cooking time (Paper-I).
Concerning the use of Raman spectroscopy to assess the fat quality, good correlations between Raman spectra and the gross FA parameters and individual FAs in pork backfat were obtained (Paper-III). However, the predictions of the individual FAs were primarily indirect and to a high degree based on co-variation of the individual FAs and the global FA parameters such as polyunsaturated fatty acids (PUFA) and iodine value (IV). The results show that Raman spectroscopy can be used for prediction of gross FA parameters whereas great care should be
taken when predicting individual FAs as a changed sample matrix will change the correlation pattern and thus provide inaccurate predictions of new samples. A new method for investigating the non-targeted predictions of such interdependent reference values is suggested. In conclusion, this thesis work demonstrated that Raman spectroscopy has a potential to be used in the meat processing industry for prediction of different quality parameters related to the final product quality. The work further shows that Raman spectroscopy has great potential for
authentication of ‘ready to eat’ cooked meat products in the retail market to be used by the quality control authority. Finally, it should be stressed that the results in this thesis were obtained only from a laboratory setup. The next step will require that the technique is tested in an
then chill stored samples to determine if the EPT of the cooked meat reached to the intended EPT or not. For this purpose, pork longissimus dorsi was heat treated at 50 oC to 70 oC (2 oC interval) for 2 to 10 h (2 h interval) (paper-I) and at 50 oC to 80 oC (2 oC interval) for 2 h in water bath then stored under vacuum packaging for day- 0, -4 and -8 (paper-II). Secondly, the thesis work examined the use of Raman spectroscopy in combination with chemometrics to determine the gross FA parameters and individual FAs in pork backfat. Non-targeted predictions of highly
collinear reference variables were also investigated. Backfat samples for this study were provided from Danish Meat Research Institute (DMRI) as part of a large feeding project.
Raman spectroscopy was able to predict EPT of cooked intact muscle with high cross-validated squared coefficient of determination ( = 0.96) and an acceptable accuracy (Root Mean Error of Cross-Validation (RMSECV = 1.78 oC) (Paper-I and II) independent of the storage time (day- 0, -4 and -8) (Paper-II). Furthermore, the results showed that Raman spectroscopy was useful to classify cooked meat with EPT of below and above 65 oC even after the storage time up to 8 days (Paper-II). This EPT was chosen as a critical point as most pathogenic microorganisms are expected to die at this temperature (at a given time). In contrast, Raman spectroscopy was not able to predict cooking time (2-10 h) of the intact muscle (Paper-I). Scrutinizing the Raman spectra provided detailed information about the structural changes of the meat proteins during cooking (Paper-I and II) and demonstrated that cooking temperature has a much higher impact on the structural changes in meat compared to cooking time (Paper-I).
Concerning the use of Raman spectroscopy to assess the fat quality, good correlations between Raman spectra and the gross FA parameters and individual FAs in pork backfat were obtained (Paper-III). However, the predictions of the individual FAs were primarily indirect and to a high degree based on co-variation of the individual FAs and the global FA parameters such as polyunsaturated fatty acids (PUFA) and iodine value (IV). The results show that Raman spectroscopy can be used for prediction of gross FA parameters whereas great care should be
taken when predicting individual FAs as a changed sample matrix will change the correlation pattern and thus provide inaccurate predictions of new samples. A new method for investigating the non-targeted predictions of such interdependent reference values is suggested. In conclusion, this thesis work demonstrated that Raman spectroscopy has a potential to be used in the meat processing industry for prediction of different quality parameters related to the final product quality. The work further shows that Raman spectroscopy has great potential for
authentication of ‘ready to eat’ cooked meat products in the retail market to be used by the quality control authority. Finally, it should be stressed that the results in this thesis were obtained only from a laboratory setup. The next step will require that the technique is tested in an
| Originalsprog | Engelsk |
|---|
| Forlag | Department of Food Science, Faculty of Science, University of Copenhagen |
|---|---|
| Antal sider | 128 |
| Status | Udgivet - 2015 |