Development of NIR-HSI and chemometrics process analytical technology for drying of beef jerky
Introduction
The process analytical technology (PAT) framework is defined as a mechanism to design, analyse and control manufacturing processes through the measurement of critical process parameters (CPPs) which affect critical quality attributes (CQAs) of raw and in-process materials and processes (Cullen, O'Donnell, & Fagan, 2014; Food and Drug Administration, 2004). PAT tools provide real-time information, facilitating manufacture of products with consistent quality and reduced waste, while lowering processing costs. PAT is widely employed in the pharmaceutical and chemical industries (Cullen et al., 2014). However, there is a need to validate PAT tools for food industry applications to facilitate increased adoption of PAT in food manufacture.
Drying is a common and longstanding method of food preservation. Prediction of drying kinetics under different conditions may be employed to optimize the drying process which influences the quality of the dried food (Andueza, Agabriel, Constant, Lucas, & Martin, 2013; Luckose, Pandey, & Harilal, 2017). Modelling of the drying process is important for equipment design, process optimization and product quality improvement (Zhao, Downey, & O'Donnell, 2014).
Jerky is a dried salted meat product with low water activity and a high protein to moisture ratio. Salt addition contributes to the preservation, characteristic taste and texture of jerky products. Various thin layer kinetic models have been investigated for drying of meat and meat products (Luckose et al., 2017). Moisture content is the main factor influencing quality, safety and shelf life of the meat-based jerky products. Traditional methods for moisture content determination are time consuming, and not suitable for online monitoring of the drying process. Therefore, the development of a non-destructive rapid technology to monitor the drying process would facilitate the adoption of a process analytical technology approach in jerky manufacture. (Achata, Esquerre, O'Donnell, & Gowen, 2015; Andersen, Frydenvang, Henckel, & Rinnan, 2016).
Ultrasound (US) technology has the potential to enhance drying processes by reducing the drying time and energy consumption (Corrêa, Rasia, Mulet, & Cárcel, 2017; Nian et al., 2017; Zhao, Downey, & O'Donnell, 2015). US has been reported to improve drying of fruits and vegetables (Corrêa et al., 2017; Kowalski & Rybicki, 2017; Kroehnke et al., 2018; Magalhães et al., 2017; Rojas & Augusto, 2018; Tao et al., 2018).
However limited studies have been reported on the drying of meat products in combination with US treatment (Ojha et al., 2018; Ojha, Kerry, & Tiwari, 2017). US can intensify mass transport phenomena in solid – liquid systems (Spellman, McEvoy, O'Cuinn, & FitzGerald, 2003). Application of US has been studied for brining of pork meat at different NaCl concentrations (Kuehler & Stine, 1974), and to improve the diffusion of NaCl in pork meat using different brining treatments (Siró et al., 2009).
Near infrared hyperspectral imaging (NIR-HSI) is a rapid non-destructive technology that does not require sample preparation, and is a suitable process analytical technology tool for on/at-line control and monitoring of food processes (Huang, Yu, Xu, & Ying, 2008; Karunathilaka, Yakes, He, Chung, & Mossoba, 2018; Lieske & Konrad, 1996). NIR spectroscopy is well suited to moisture content determination because of strong water absorption bands in the NIR spectrum. Hyperspectral imaging (HSI) provides both spatial and spectral information of samples by combining imaging and spectroscopic tools. Chemometric methods are employed for the extraction of chemical information from hyperspectral images or hypercubes and to develop prediction maps for quality attributes. Hyperspectral imaging has been investigated for the control of drying of banana slices (Slattery & Fitzgerald, 1998), mango slices (Kumar & Barth, 2010) and potato slices (Cadet, Pin, Rouch, Robert, & Baret, 1995). However the use of HSI to monitor the drying process of beef jerky has not been reported.
Chemometric tools used for hyperspectral imaging data analysis include principal component analysis (PCA), which is frequently employed as an exploratory tool and for dimensionality reduction (Burger & Gowen, 2011). Calibration techniques such as partial least squares regression (PLS-R) are routinely employed in hyperspectral imaging analysis for prediction of unknown concentrations and generation of prediction maps to estimate spatial distribution of components in a sample (A. Gowen, Burger, Esquerre, Downey, & O'Donnell, 2014). Band selection methods have been demonstrated to improve the performance of hyperspectral imaging models and to reduce the processing times required by selecting the most informative spectral bands (Achata, Inguglia, Esquerre, Tiwari, & O'Donnell, 2019) and have been studied for the early detection of bruise damage in mushrooms (Esquerre, Gowen, Downey, & O'Donnell, 2011), viability and vigour in muskmelon seeds (Karmali, Karmali, Teixeira, & Curto, 2004) fat and moisture content in ground beef (Schindler, Le Thanh, Lendl, & Kellner, 1998), internal damage in cucumbers and whole pickles (Resa & Buckin, 2011), evaluation of mixing quality of food powders (Achata, Esquerre, Gowen, & O'Donnell, 2018) and the assessment of the brining process for pork meat (Achata et al., 2019). Spectral pre-treatments can be used to correct for the effects of natural variability in the shape and size of samples, light scattering and differences in the effective path length in NIR spectra, which can cause difficulties in the application of HSI for quality assessment (Esquerre, Gowen, Burger, Downey, & O'Donnell, 2012). Prediction models with RPD values >3.5 are considered very good for process control applications of materials which have complex physical characteristics (Williams, 2014) such as meat samples.
The aim of this study was to investigate the development of NIR-HSI and chemometrics as a process analytical technology for drying of beef jerky.
Section snippets
Sample preparation
Six fresh eye of round beef cuts (M. Semitendinosus) were purchased from local supermarkets. Beef cuts' batch codes were checked to ensure that each eye of round came from a different batch code. Fat and external connective tissue was removed from each eye of round cut and stored at −18 °C. Prior to analysis samples were maintained at 4 °C for 16 h, sliced into 3 mm thick slices and trimmed to obtain uniform samples of ca. 75 mm × 65 mm × 3 mm. Two eye of round cuts were used per replicate.
Sample treatments
Four
Moisture content and drying kinetics
The mean moisture content of the three replicates used for the kinetic models is listed in Appendix A. The moisture content (kg water/kg dry matter) during drying for all experimental treatments is presented in Fig. 1a. A gradual reduction in moisture content is observed during the first 90 min of drying, when free or loosely bound water is evaporated. Thereafter the drying curves level off as more strongly bound water is removed. All samples exhibited similar drying profiles. US-WI samples had
Conclusions
The potential of NIR-HSI as a process analytical technology tool for the prediction of moisture content during the drying process of beef jerky with and without US treatment and whether brined or not was demonstrated in this study. Ultrasound treatment accelerated the drying of water immersed beef jerky but was not shown to enhance the drying of brined samples due to salt-protein-water interaction.
The majority of the moisture content prediction models developed in this work had RPD values >4,
Declaration of Competing Interest
The authors, Eva Maria Achata, Carlos Esquerre, K. Shikha Ojha, Brijesh Tiwari & Colm P. O'Donnell declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The authors acknowledge funding for this project from FIRM (13/FM/508) as administered by the Irish Department of Agriculture, Food & the Marine.
References (53)
- et al.
Feasibility of near infrared and Raman hyperspectral imaging combined with multivariate analysis to assess binary mixtures of food powders
Powder Technology
(2018) - et al.
Evaluation of Vis-NIR hyperspectral imaging as a process analytical tool to classify brined pork samples and predict brining salt concentration
Journal of Food Engineering
(2019) - et al.
The potential of laser-induced breakdown spectroscopy for industrial at-line monitoring of calcium content in comminuted poultry meat
Food Control
(2016) - et al.
Using visible or near infrared spectroscopy (NIRS) on cheese to authenticate cow feeding regimes
Food Chemistry
(2013) - et al.
Ultrasonic vacuum drying technique as a novel process for shortening the drying period for beef and chicken meats
Innovative Food Science & Emerging Technologies
(2014) - et al.
Data handling in hyperspectral image analysis
Chemometrics and Intelligent Laboratory Systems
(2011) - et al.
Enzyme kinetics by mid-infrared spectroscopy: β-fructosidase study by a one-step assay
Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology
(1995) - et al.
Performance of some variable selection methods when multicollinearity is present
Chemometrics and Intelligent Laboratory Systems
(2005) - et al.
Influence of ultrasound application on both the osmotic pretreatment and subsequent convective drying of pineapple (Ananas comosus)
Innovative Food Science & Emerging Technologies
(2017) - et al.
Suppressing sample morphology effects in near infrared spectral imaging using chemometric data pre-treatments
Chemometrics and Intelligent Laboratory Systems
(2012)