Contents lists available at ScienceDirect Food Control journal homepage: www.elsevier.com/locate/foodcont ATR-MIR spectroscopy and multivariate analysis in alcoholic fermentation monitoring and lactic acid bacteria spoilage detection Julieta Cavaglia a , Daniel Schorn-García a , Barbara Giussani b , Joan Ferré c , Olga Busto a , Laura Aceña a , Montserrat Mestres a , Ricard Boqué c,∗ a Instrumental Sensometry (iSens), Department of Analytical Chemistry and Organic Chemistry, Campus Sescelades, Edifici N4, Universitat Rovira i Virgili, C/Marcel·lí Domingo s/n, Tarragona, 43007, Spain b Dipartimento di Scienza e Alta Tecnologia, Università degli Studi dell'Insubria, Via Valleggio, 9, Como, 22100, Italy c Chemometrics, Qualimetrics and Nanosensors Group, Department of Analytical Chemistry and Organic Chemistry, Campus Sescelades, Edifici N4, Universitat Rovira i Virgili, C/Marcel·lí Domingo s/n, Tarragona, 43007, Spain ARTICLE INFO Keywords: Process monitoring Alcoholic fermentation Wine Malolactic fermentation contamination ATR-MIR Process analytical technologies ABSTRACT Wine production processes still rely on post-production evaluation and off-site laboratory analyses to ensure the quality of the final product. Here we propose an at-line methodology that combines a portable ATR-MIR spec- trometer and multivariate analysis to control the alcoholic fermentation process and to detect wine fermentation problems. In total, 36 microvinifications were conducted, 14 in normal fermentation conditions (NFC) and 22 intentionally contaminated fermentations (ICF) with different lactic acid bacteria (LAB) concentrations. ATR- MIR measurements were collected during alcoholic and malolactic fermentations and relative density, pH, and L- malic acid were analyzed by traditional methods. Partial Least Squares Regression could suitably predict density and pH in fermenting samples (root mean squared errors of prediction of 0.0014 g mL −1 and 0.06 respectively). With regard to ICF, LAB contamination was detected by multivariate discriminant analysis when the difference in L-malic acid concentration between NFC and ICF was in the order of 0.7–0.8 g L −1 , before the end of mal- olactic fermentation. This methodology shows great potential as a fast and simple at-line analysis tool for de- tecting fermentation problems at an early stage. 1. Introduction The production of wine is based on alcoholic fermentation, which consists in the biochemical transformation of sugar into ethanol by yeasts. There are many factors that have an influence over the com- plexity and quality of the final product such as the grape quality and variety, yeast strain or cellar practices used (Suárez-Lepe & Morata, 2012). However, even with the best raw materials and starting under the optimal conditions, problems during alcoholic fermentation can occur, in which yeast or other microorganisms synthetize undesirable compounds that negatively affect the quality of the wine. Stuck and sluggish fermentations along with contamination-related processes are the most common problems that can appear during alcoholic fermen- tation (Hernández, León, & Urtubia, 2016). Nutrient deficiencies, sudden temperature changes or the imposition of undesired and non- inoculated yeast are the main causes of stuck and sluggish fermenta- tions. Spoilage processes are due to the growth of unwanted micro- organisms in the must, such as acetic acid or lactic acid bacteria (LAB), which are part of the normal microbiota found on the surface of leaves and grapes but can also be found in the environment of wineries (Portillo, Franquès, Araque, Reguant, & Bordons, 2016). Although the “piqûre acétique” is the most widely known spoilage, the “piqûre lac- tique” can also pose very important problems for some wines. LAB are responsible for the biochemical transformation of L-malic acid into L-lactic acid releasing carbon dioxide. This process, called malolactic fermentation, is promoted in red wines to decrease their acidity since, from an organoleptic point of view, a lower acidity is more compatible with the high tannicity of these wines (Cappello, Zapparoli, Logrieco, & Bartowsky, 2017). However, in white wines, this second fermentation is usually undesired because it increases pH and reduces their typical freshness, leading to wines with worse organo- leptic quality (Cozzolino, Mccarthy, & Bartowsky, 2012). In the winemaking industry, a control of the alcoholic fermentation process is required in order to avoid problems that result in low quality wines and consequently, in economic losses. In the cellar, the process is mostly controlled by determining temperature, density and pH, which https://doi.org/10.1016/j.foodcont.2019.106947 Received 2 August 2019; Received in revised form 16 September 2019; Accepted 8 October 2019 ∗ Corresponding author. E-mail address: ricard.boque@urv.cat (R. Boqué). Food Control 109 (2020) 106947 Available online 09 October 2019 0956-7135/ © 2019 Elsevier Ltd. All rights reserved. T