Determination of quality parameters of tomato paste using guided microwave spectroscopy Lu Zhang a , Michelle A. Schultz b , Rick Cash b , Diane M. Barrett a , Michael J. McCarthy a, c, * a Department of Food Science and Technology, University of California, Davis, One Shields Avenue, Davis, CA 95616, USA b Thermo Fisher Scientic, Minneapolis, MN 55433, USA c Department of Biological and Agricultural Engineering, University of California, Davis, One Shields Avenue, Davis, CA 95616, USA article info Article history: Received 19 September 2013 Received in revised form 3 December 2013 Accepted 5 December 2013 Keywords: Tomato paste Soluble solids content Titratable acidity Bostwick consistency Guided microwave spectroscopy Partial Least Square abstract Guided microwave spectrometry was evaluated for in-line determination of quality attributes, soluble solids content, titratable acidity and Bostwick consistency of tomato pastes in continuous processing. The spectrometer was tested for two processing operations, dilution and evaporation. A total of 34 tomato paste samples were prepared by blending 9 different tomato pastes together in various proportions with nal Bostwick consistency ranging from 2 to 11 cm. During the processing operation, spectra of 5 samples were obtained when different soluble solids contents were achieved. At the same time, standard reference methods were used to measure the parameters of interest on the same samples taken from the processing ow loop. The spectrum was recorded from 136 to 2690 MHz in 2 MHz steps. Partial Least Square (PLS) was employed to produce prediction models for each quality attribute from the collected microwave spectra. Cross-validation was performed to validate the calibration models. These models demonstrated good predictability of the soluble solids content, titratable acidity and Bostwick consis- tency in both dilution and evaporation conditions, as shown by their high correlation coefcient (r) and low root mean square error of cross-validation (RMSECV) values. Interval PLS (iPLS) was applied to optimize the models by selecting the most relevant and non-noisy regions in the spectra. The iPLS models showed improved accuracy for prediction of all three of the parameters, especially for Bostwick consistency (r > 0.92 and RMSECV < 0.7 cm). These observations suggest that guided microwave spectroscopy gave accurate estimation of soluble solids content, titratible acidity and Bostwick consis- tency of tomato paste. Different processing conditions required varying prediction models for the changes in the dielectric properties of tomato pastes. Ó 2013 Elsevier Ltd. All rights reserved. 1. Introduction The majority of the processing tomatoes grown in California are thermally processed into concentrated tomato paste. The concen- trated pastes are usually stored and used as an intermediate product with water and other ingredients to be reconstituted into nal products, such as ketchups and sauces. Since tomato paste is the main ingredient in the nal products, maintaining the quality of the paste is crucial for the tomato processing industry. Factors like the cultivar of the tomato and the processing conditions introduce great variation in the quality of the paste (Anthon, Diaz, & Barrett, 2008). These variations pose difculties in achieving a consistent paste quality during production. As a quality control routine, pastes are sampled from each production batch and a range of quality factors are tested; these include color, soluble solids content, pH, acidity and consistency. Because the classication of tomato prod- ucts, e.g., paste or puree, is done according to its soluble solids content, it is a key characteristic of tomato products. During evaporation, the real-time value of the soluble solids helps to determine the end point of the processing. In addition, the avor of tomato paste is closely associated with the soluble solids content and titratable acidity. Among all factors, consistency is a major quality component of tomato paste, sauce and ketchup, determining the rheological properties of the intermediate and nal products. Consistency de- scribes the ow property of non-Newtonian uids with suspended particles and dissolved long chain molecules (Barrett, Garcia, & Wayne, 1998). The consistency of the paste is typically evaluated using a Bostwick consistometer. This empirical method is widely used in the tomato industry, and USDA has established quality * Corresponding author. Department of Food Science and Technology, University of California, Davis, One Shields Avenue, Davis, CA 95616, USA. Tel.: þ1 530 752 8921; fax: þ1 530 752 4759. E-mail address: mjmccarthy@ucdavis.edu (M.J. McCarthy). Contents lists available at ScienceDirect Food Control journal homepage: www.elsevier.com/locate/foodcont 0956-7135/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.foodcont.2013.12.008 Food Control 40 (2014) 214e223