**1. Introduction**

Brazil is the world's largest coffee producer. The recent export data report that Brazil has shipped around 22.872 million bags (60 kg each) from July 2021 to January 2022. Specialty coffees accounted for 17.4% of total Brazilian exports, with an average price of USD 292.44 per bag, representing 23.4% of the total obtained with the shipments in January 2022 [1]. Specialty coffees, defined as high-quality products, are quite relevant for the coffee industry given the higher prices attained in comparison to commodity coffees. While a regular bag of regular green coffee costs approximately USD 200, specialty coffees can go up to USD 1000 per bag.

The quality of a cup of coffee begins in the field. Several factors including coffee species and variety, harvesting, post-harvesting conditions, blend elaboration, and roasting parameters, have a significant influence on the flavor and aroma of the drink. The delicate taste and aroma obtained from a cup of specialty coffee results from a complex combination of physical transformations and chemical reactions that start on the seed and end on the beverage preparation [2,3].

**Citation:** Belchior, V.; Botelho, B.G.;Franca, A.S. Comparison of Spectroscopy-Based Methods and Chemometrics to Confirm Classification of Specialty Coffees. *Foods* **2022**, *11*, 1655. https:// doi.org/10.3390/foods11111655

Academic Editor: Daniel Cozzolino

Received: 5 May 2022 Accepted: 2 June 2022 Published: 4 June 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

The most common way to evaluate the quality of a green coffee is by cup tasting [3,4]. Several industries, including perfume, coffee and tea, wine, beer, and tobacco, often employ trained personnel for sensory evaluation. In the specific case of coffee, such people are called "Q-graders" and trained to define the sensory profile of different samples. Then, according to the SCA (Specialty Coffee Association) protocol to evaluate coffee, they classify samples by giving different scores [5].

The SCA protocols are based on objective assessment methods, including the presence or absence of sweetness and defects, thus minimizing subjectivity compared to other methodologies. In addition, Q-graders are considered excellent and accurate in giving the scores related to quality, although some errors and inconsistencies regarding the description of a coffee are reported [6].

Furthermore, sensory analysis can lead to a few problems. Bias that comes from the preference and previous knowledge of a specific sample, as well as the influence of some external factors [6] can affect the analysis. Additionally, the Q-grader´s health during the cupping as well as modification on his (her) personal evaluation abilities over time can also affect the results. Such issues can be minimized by using alternative evaluation tools in order to make the coffee trading market more reliable [3,7]. Sensory analysis can also be viewed as a sensitive and time-consuming technique, given the need for well-trained personnel. Considering the economic relevance of specialty coffees in the world trade market, finding alternative tools to confirm coffee quality is of utmost importance.

Many studies have shown the potential of spectroscopic methods in food analysis, with near (NIR) and mid (FTIR) infrared among the most used methods [8]. The employment of such techniques for coffee analysis has been widely reported [9]. Applications include discrimination between coffee species and varieties [10], adulteration of roasted and ground coffee [11–13], and identification of low quality (defective) coffee beans [14–16]. Given that such low quality coffees have a significant effect on the sensory profile of the beverage, spectroscopic methods can also be used to detect differences in sensory parameters. In recent studies, our research group employed chemometrics to develop models for the classification and discrimination between espresso coffee beverages based on generic parameters (intensity and a few sensory aspects) informed by the manufacturers, and also based on sensory analysis performed by a trained panel [7,17]. It was also possible to develop models that classified coffees by cup quality parameters based on classification criteria that are specific to Brazil [4]. Some recent results from another group also showed the feasibility of mid-infrared and chemometrics to discriminate specialty coffees with different roasting profiles [18]. Our latest study showed that FTIR can be successfully used to discriminate specialty coffees classified by Q-graders [3], with models capable of predicting classification scores with high accuracy (validation coefficients above 0.97). Published studies confirm that both FTIR and NIR are promising techniques for coffee quality evaluation. However, in the case of NIR, only qualitative discrimination was performed with respect to coffee quality parameters given by Q-graders, without any attempt to provide an actual score-based classification. Furthermore, a comparison of spectroscopy-based techniques to evaluate specialty coffees has not ye<sup>t</sup> been reported. Since both FTIR and NIR have been shown as reliable techniques for coffee quality definition, a comparison of these methods can indicate which method is more reliable. Although several studies have been described and tested with both techniques, there is still a need for further investigation, in order to improve the quality of predictive models to be applied for food quality evaluation [19].

Therefore, in this study, the potential of NIR was evaluated for establishing sensory characteristics of specialty coffees in terms of quantitative scores. Partial Least Squares (PLS) Regression was employed to build models in order to predict and establish a SCAbased sensory profile. NIR-based models were compared to FTIR ones that were developed in a previous study [3]. To the best of our knowledge, this is the first study in the literature that addresses such comparison for specialty coffee quality evaluation. Furthermore, this is the first work showing that NIR can provide quantitative quality scores.
