T-MIR Spectroscopy Applications in Coffee and Cocoa

Version	Summary	Created by	Modification	Content Size	Created at	Operation
1		Aracely Acosta Caudillo	+ 2094 word(s)	2094	2022-02-23 07:44:16	\|
2	format correct	Nora Tang	+ 474 word(s)	2568	2022-03-01 06:32:14	\| \|
3	format correct	Nora Tang	+ 474 word(s)	2568	2022-03-01 06:36:09	\|

This entry is adapted from the peer-reviewed paper 10.3390/foods11040579

Nowadays, coffee and cocoa have broad applications in the food and pharmaceutical industries due to their organoleptic and nutraceutical properties, which have turned them into products of great commercial demand. Consequently, these products are susceptible to fraud and adulteration, especially those sold at high prices, such as saffron, vanilla, and turmeric. This situation represents a major problem for industries and consumers’ health. Implementing analytical techniques, i.e., Fourier transform mid-infrared (FT-MIR) spectroscopy coupled with multivariate analysis, can ensure the authenticity and quality of these products since these provide unique information on food matrices.

spices coffee cocoa quality control adulteration FT-MIR

1. Coffee

Coffee is one of the main traded crops and the second most consumed beverage worldwide ^[1]^[2]. The coffee plant belongs to the genus Coffea L., which comprises more than 100 species, but only Coffea arabica (Arabica coffee) and Coffea canephora (Robusta coffee) are of commercial and economic importance ^[3]^[4]. After harvesting, coffee cherries undergo processing (dry, semi-wet, or wet) to separate the coffee beans (green or raw) from the fruit and reduce their moisture content (up to 10–12%). Likewise, it facilitates coffee beans transportation without a loss of quality ^[5]^[6]. In the international market, the leading coffee exporting countries are Brazil, Vietnam, Indonesia, Colombia, and Ethiopia ^[4]^[7].

Green coffee beans require a roasting process for consumption. Coffee roasting softens the bean, facilitates the detachment of the husk, and augments coffee’s characteristic color, flavor, and aroma. During this method, some coffee compounds are degraded (proteins, trigonelline, chlorogenic acids, and phenols), and others are formed (melanoidins and volatile compounds) ^[6]^[7]^[8]. Initially, roasting is carried out at 180 °C aiming to reduce the moisture content of the beans to 1.5–5%; then, the temperature is raised to 200–300 °C commencing Maillard and caramelization reactions, which lead to the formation of volatile substances that generate coffee characteristics. Once the desired roasting is obtained, coffee beans are cooled by water or air to stop the biochemical reactions and avoid further degradation. Roasted coffee beans consist of polysaccharides (38–42%), lipids (11–17%), proteins (7.5–10%), aliphatic acids (1.6%), chlorogenic acids (2.5–3.8%), caffeine (1.3–2.4%), trigonelline (0.7–1%), minerals (4.5–4.7%), volatile compounds (0.1%) and melanoidins (23–25%) ^[9]^[10]. Besides providing health benefits, (e.g., vasoconstrictor, neuroprotective and neurostimulator, antioxidant, anti-inflammatory, and anticarcinogenic), caffeine, trigonelline, chlorogenic acids, and volatile compounds content are related to coffee quality ^[8]^[10]^[11]^[12]^[13]. Approximately 60–80% of the world’s coffee production is obtained from Coffea arabica species and the other from Coffea canephora ^[6]^[14]. Both commercial species of coffee differ in composition. C. canephora has a high content of caffeine and chlorogenic acids; therefore, C. arabica quality is considered superior and double the price of C. canephora (Robusta coffee). Moreover, the geographical origin of C. arabica can further raise its cost ^[15]^[16]^[17]. After roasting and grinding, those differences become imperceptible; as a result, C. arabica beans can be adulterated with C. canephora to reduce production costs ^[1]. Other low-grade products might be added to C. arabica, for example, coffee husks and grounds, corn wheat, barley, soybean, and rye ^[2]^[15]^[18]. Coffee quality and authenticity are generally measured through chromatography (HPLC and GC), spectroscopy (UV-vis, NMR, fluorescence, IR, and Raman), inductively coupled plasma (ICP), real-time PCR, atomic absorption spectrometry (AAS), and mass spectrometry (MS) ^[13]^[16]^[17].

FT-MIR spectroscopy and multivariate analysis are applied for coffee authentication and quality control (Table 1). Wang et al. ^[19] implemented these techniques to detect and quantify adulteration of Kona coffee with lower-quality coffee. They used ground and brewed coffee in spectra range (1900–800 cm⁻¹), first and second derivative pretreatments, and PCR and PLS analysis to calibrate the models. The best calibration results were obtained with the spectra of brewed coffee (R² = 0.999), applying the second derivative and PLS algorithm. The unfavorable results for ground coffee spectra were explained by low spectral data precision due to particle shape and size.

Table 1. Applications of FT-MIR spectroscopy in coffee quality control.

Spectral Range (cm⁻¹)	Sampling Technique	Algorithm	Purpose of the Analysis	Reference
Arabica coffee variety Kona typica
1900–800	ZnSe ATR	PCR PLS	Detection and quantification of adulteration of coffee grown in Kona, Hawaii, with coffee from another region.	^[19]
Brazilian coffee
3600–2820 1800–784 1900–800	DRIFT	PCA PLS-DA	Discrimination of decaffeinated coffee and classification according to roasting degree.	^[20]
3600–2820 1800–784 1900–800	KBr pellets	RBF (ANN)	Coffee classification by geographic and genotypic origin.	^[21]
Green Arabica coffee
4000–700	ZnSe ATR DRIFT	PCA, LDA HCA	Discrimination of immature coffee (defective) and mature coffee (non-defective).	^[22]
4000–700	KBr pellets ZnSe ATR, DRIFT	PCA HCA	Discrimination of defective and non-defective coffee using three different sampling techniques.	^[23]
3600–600	DRIFT	PCA, LDA	Discrimination of defective and non-defective roasted coffee.	^[24]
1800–800	KBr pellets	SVM	Geographical classification of different coffee genotypes.	^[25]
4000-600 2920–2850 1745	ZnSe ATR	PCA	Discrimination of coffee beans according to their origin (Brazil, Colombia, Ethiopia, Kenya, and Yemen).	^[26]
3000–900	ZnSe ATR	PLS	Prediction of quality scores given by cuppers for coffee beverage samples.	^[27]
Roasted Arabica coffee
3200–700	DRIFT	PCA LDA	Discrimination between roasted coffee, corn, coffee husk, coffee-corn, and coffee-husk blends.	^[28]
3200–700	DRIFT	PCA LDA	Discrimination between roasted coffee, coffee husks, coffee grounds, corn, barley, and coffee-adulterant blends.	^[29]
3200–700	DRIFT	PLS	Prediction of adulteration levels of roasted coffee with different adulterants (pure and blended).	^[30]
4000–700	ZnSe ATR	PLS	Simultaneous quantification of four adulterants (coffee husk, coffee grounds, barley, and corn) in roasted coffee.	^[31]
4000–525	Diamond ATR	PCA PLS	Detection and quantification of adulteration of roasted coffee with corn.	^[32]
3200–700 4000–600	DRIFT ZnSe ATR	PLS-DA HM DF	Discrimination between roasted coffee and adulterated coffee using two sampling techniques and merging data.	^[33]
3200–700 4000–600	ZnSe ATR	PCA PLS-DA	Classification of cup quality of coffee with different roasting degrees.	^[34]
3500–2800 1800–800	Diamond ATR	SIMCA PCR PLS1, PLS2	Identification and quantification of adulterated coffee with coffee husks, corn, barley, soybeans, oats, and rice.	^[35]
Arabica and Robusta coffee
4000–600 2000–1500 3000–2750	ZnSe ATR	PLS	Quantification of Robusta coffee content in blends with Arabica coffee.	^[36]
1800–800	ATR	PCA PLS-DA	Comparison of three spectroscopic techniques (1H-NMR, NIR, and MIR) for the discrimination of coffee by species and origin.	^[37]
Commercial coffee capsules
3000–600	ZnSe ATR	PCA PLS-DA	Discrimination of espresso coffee according to sensory characteristics.	^[38]

PCR: principal component regression; PLS: partial least squares regression; PCA: principal component analysis; PLS-DA: partial least squares discriminant analysis; LDA: linear discriminant analysis; HCA: hierarchical cluster analysis; ANN: artificial neural networks; HM: hierarchical models; DF: data fusion; SVM: support vector machine; SIMCA: soft independent modeling of class analogy; PLS1: partial least squares with single y-variables; PLS2: partial least squares with multiple y-variables.

Ribeiro et al. ^[20] performed discrimination on commercial coffee samples according to their caffeine content (caffeinated and decaffeinated) and classification based on roasting degree. They used DRIFT, smoothing, and multiplicative scatter correction (MSC) pretreatments and selected the spectral regions 3600–2820 and 1800–784 cm⁻¹. Later, they effectuated the discrimination model using PCA and the classification model with PLS-DA achieving 100% classification of the external validation samples. Craig et al. ^[22]^[23] developed multivariate analysis models to discriminate defective and non-defective green coffee beans by comparing different sampling techniques (KBr pellets, ZnSe ATR, and DRIFT) and multivariate analyses (PCA, LDA, and HCA). They executed baseline correction and normalization in all models, except in the KBr pellets’ model, where the first derivative was applied. Subsequently, they developed the classification models based on PCA and LDA algorithms using the DRIFT data, whose predictive capacity ranged between 95% and 100% ^[24]. Finally, the authors classified Arabica coffee according to cup quality. They used ZnSe ATR sampling, baseline correction, normalization, MSC, first and second derivative; employed PCA to separate Arabica and Robusta coffee; and built the classification models with PLS-DA, which presented 100% sensitivity and specificity in calibration but 67–100% in validation ^[34]. Reis et al. ^[28]^[29]^[30] conducted diverse studies on roasted coffee, adulterants (coffee husk, coffee grounds, roasted corn, and roasted barley), and coffee-adulterant mixtures using 3200–700 cm⁻¹ spectral region and DRIFT sampling. Through PCA, they separated the pure samples and each of the adulterants and identified the spectral regions with significant variations. Simultaneously, they elaborated LDA classification models for roasted coffee, pure adulterant, and coffee-adulterant mixtures, obtaining 100% prediction and recognition capacity. After, they tested a PLS algorithm to predict the level of adulteration of the samples (1–66%, w/w), achieving high calibration (0.99) and validation (0.98) coefficients and low percentage of error (1.23% for calibration and 2.67% for validation). Researchers simultaneously quantified, with the ZnSe ATR sampling technique, the above-mentioned adulterants in roasted coffee, obtaining correlation coefficients of 0.99 for validation and calibration and a percentage of error of 0.69% for calibration and 2.00% for validation ^[31]. Lastly, they performed a discriminant analysis of the data by comparing DRIFT and ZnSe ATR techniques, employing hierarchical models (HM), and then spectral data fusion (DF). The percentage of misclassified samples decreased to 0% after DF ^[33].

Brondi et al. ^[32] introduced two similar models to detect and quantify adulteration of roasted coffee with corn (0.5–40% w/w). The models were built from data collected with differential scanning calorimetry (DSC) and FT-MIR techniques coupled to PCA and PLS. They were able to discriminate between pure coffee and coffee-corn mixtures with both techniques. The FT-MIR model presented lower cross-validation mean square error (RMSECV = 2.7%) than DSC; therefore, it has a great application potential. Another approach proposed by Link et al. ^[21] classified Arabica coffee samples based on geographic and genotypic origin using KBr pellets sampling. Spectra (1900–800 cm⁻¹ region) were pretreated with normalization, baseline correction, and smoothing. They employed a radial basis function (RBF) network, an artificial neural network (ANN), to build the classification models, getting better geographic (100%) and genotypic (94.44%) classification results compared with soft independent modeling of class analogy (SIMCA) and multilayer perceptron (MLP). Bona et al. ^[25] also performed geographical classification of Arabica coffee through SVM and NIR and MIR spectroscopy. The best results were obtained with the NIR-SVM approach, where all samples were correctly validated.

Correia et al. ^[36] quantified Robusta coffee content in adulterated Arabica coffee samples using ZnSe ATR sampling. They analyzed the samples with FT-MIR and developed a multivariate PLS model that presented low detection (LOD = 1.29%) and quantification (LOQ = 4.3%) limits and a high coefficient of determination (R² = 0.9635) in the cross-validation. They further evaluated the samples by negative-ion mode electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry (ESI(-)FT-ICR MS) and modeled the data by univariate analysis, obtaining slightly better results. However, FT-MIR is a simpler technique than ESI(-)FT-ICR MS.

Medina et al. ^[37] compared three spectroscopic techniques (1H-NMR, NIR, and MIR) for coffee discrimination according to species (Arabica and Robusta) and origin (Colombia and other countries). FT-MIR spectra were collected in the 1800–800 cm⁻¹ range and pretreated with normalization and second derivative; the models were built using PCA and PLS-DA. All techniques successfully discriminated the samples by species, although discrimination by origin FT-MIR and 1H-nuclear magnetic resonance (1H-NMR) showed better results than NIR. The authors emphasized that FT-MIR is a faster and cheaper technique in comparison with 1H-NMR.

Obeidat et al. ^[26] performed another model for origin discrimination in green coffee samples from different countries (Brazil, Colombia, Ethiopia, Kenya, and Yemen) applying ZnSe ATR. They used the 4000–600 cm⁻¹ spectral region pretreated with normalization. PCA algorithm allowed to identify the bands with greater spectral variation (2920–2850 and 1745 cm⁻¹), enabling successful discrimination of samples based on origin.

Belchior et al. ^[38], with the assistance of a panel of tasters, built a model of espresso coffee discrimination according to its sensory characteristics. Unlike other researchers, they used commercial capsules of different coffee brands and diverse roasting degrees. Tasters evaluated the sensory quality of espresso coffee on a 5-point scale. The spectra of the beverages were collected in the 3000–600 cm⁻¹ region, measured with ZnSe ATR. The data were pretreated with autoscaling and analyzed by PCA to group the samples according to aroma and flavor. Then, they developed discrimination models for each sensory attribute with PLS-DA, which showed high sensitivity and specificity for calibration and validation and low classification errors in cross-validation. The same authors built a model for predicting quality scores in beverages prepared from specialty samples of green Arabica coffee (quality score of 81–91 points established by cuppers). FT-MIR spectra were obtained under the same conditions as their previous work. PLS algorithm results were satisfactory for both calibration (R² = 0.99 and RMSEC = 0.23) and validation (R² = 0.97 and RMSEP = 0.23) ^[27]. Lastly, Flores-Valdez et al. ^[35] identified and quantified adulterated samples of Arabica coffee with corn, barley, soybeans, oats, rice, and coffee husks at levels of 1–30%. MIR spectra were obtained with diamond ATR and processed with atmospheric filter pretreatments, normalization, smoothing, and baseline correction; using the spectral regions of 3500–2800 and 1800−800 cm⁻¹, they optimized a discrimination model applying a SIMCA algorithm and developed PCR, PLS1, and PLS2 algorithms for quantification models. The discrimination model presented an accuracy of 100% and differentiated all adulterants, while the best quantification model was obtained with the PLS1 algorithm with outstanding calibration (R² = 0.99 and SEC = 0.39–0.82) and validation results (R² = 0.99 and SEP = 0.45–0.94).

2. Cocoa

Cocoa beans are used to produce chocolate; these are native to South America but domesticated in Central America. Afterward, the Spanish sped it to Europe and then distributed it to other countries. Nowadays, it is mainly grown in the hot and humid regions of Africa ^[39], Central and South America, and Asia ^[40]. There are three main varieties of cocoa: Criollo, Trinitario, and Forastero; the Criollo variety is not produced as much as other varieties despite the suitable quality of its cocoa; this is mainly grown in America. On the other hand, Forastero is produced mainly in Africa; even if its cocoa is not as suitable as Criollo, it has better yields. Lately, Trinitario variety is a cross between Criollo and Forastero, and it yields cocoa of reasonably suitable quality. Although the quality of cocoa depends on the variety of origin, another determining factor is the suitable harvest and proper drying and fermentation. During fermentation, the conditions are provided for the characteristic flavor and aroma of chocolate to develop in the grain through the microorganisms that intervene in the process. A correct process could produce suitable cocoa quality ^[41].

Due to the increase in the demand for cocoa, it is important to have a technique that allows determining the suitable quality of the product or a correct classification of the kind of variety. Table 2 summarizes the works on FT-MIR spectroscopy application to cocoa and chocolate.

Table 2. Applications of FT-MIR spectroscopy on quality control of cocoa.

Spectral Range (cm⁻¹)	Sampling Technique	Algorithm	Purpose of the Analysis	Reference
Chocolate
3600–2800 1800–500	ATR cell	PCA PLS	Determination of cocoa solids content in chocolates.	^[42]
1800–700	Diamond ATR	PLS	Quantification and prediction of antioxidant capacity and catechin concentration in chocolate.	^[43]
Chocolate and fermented cocoa beans
4400–600	ZnSe ATR	PLS	Prediction of antioxidant capacity and total phenolic content.	^[44]
Cocoa bean shells
4000–500	Ge ATR	PCA PLS-DA	Identification of systematic patterns related to the geographical origin of the samples.	^[45]

PCA: principal component analysis; PLS: partial least squares regression; SIMCA: soft independent modeling of class analogy; PLS-DA: partial least squares discriminant analysis.

Batista et al. ^[44] used cocoa beans spontaneously fermented and inoculated with Saccharomyces cerevisiae to quantify the antioxidant capacity and the total phenolic compounds of the beans as well as the chocolate produced from them. Results indicated variations in phenolic composition between spontaneously fermented and inoculated samples. The PLS model for total phenolics and antioxidant capacity prediction showed a correlation coefficient >0.94.

Santos et al. ^[42] performed models for predicting cocoa solids in chocolate, which showed excellent prediction and generalization capability for commercial samples by applying PLS to the MIR data set and reported that the cocoa solids content in 14% of tested chocolates differed in more and 10% of the content presented on the label. On the other hand, Hu et al. ^[43] built five PLSR models and cross-validated them to quantify catechin, antioxidant capacity, and total phenolics in chocolate, achieving suitable prediction capability for DPPH (R²_p = 0.89), ORAC (R²_p = 0.90), Folin–Ciocalteu (R²_p = 0.88) and (+)-catechin (R²_p = 0.86) but low accuracy in prediction of (−)-epicatechin (R²_p = 0.72). Finally, Mandrile et al. ^[45] used NIR, FT-MIR, and inductively coupled plasma-optical emission spectroscopy (ICP-OES) in conjunction with PCA to authenticate the geographical origin of cocoa shells through their molecular and elemental composition. The best classification results were obtained using the three spectroscopic techniques and PLS-DA to merge the PCA data obtained with each technique and were for Central African samples with an accuracy of 0.84.

References

Barbin, D.F.; Felicio, A.L.; Sun, D.W.; Nixdorf, S.L.; Hirooka, E.Y. Application of Infrared Spectral Techniques on Quality and Compositional Attributes of Coffee: An Overview. Food Res. Int. 2014, 61, 23–32.
Sezer, B.; Apaydin, H.; Bilge, G.; Boyaci, I.H. Coffee Arabica Adulteration: Detection of Wheat, Corn and Chickpea. Food Chem. 2018, 264, 142–148.
Diby, L.; Kahia, J.; Kouamé, C.; Aynekulu, E. Tea, Coffee, and Cocoa. In Encyclopedia of Applied Plant Sciences; Elsevier: Amsterdam, The Netherlands, 2017; pp. 420–425.
Hoseini, M.; Cocco, S.; Casucci, C.; Cardelli, V.; Corti, G. Coffee By-Products Derived Resources. A Review. Biomass Bioenergy 2021, 148, 106009.
Franca, A.S.; Oliveira, L.S. Coffee. In Integrated Processing Technologies for Food and Agricultural by-Products; Academic Press: Cambridge, MA, USA, 2019; pp. 413–438.
Gibson, M.; Newsham, P. Tea and Coffee. In Food Science and the Culinary Arts; Academic Press: Cambridge, MA, USA, 2018; pp. 353–372.
de Melo Pereira, G.V.; de Carvalho Neto, D.P.; Magalhães Júnior, A.I.; do Prado, F.G.; Pagnoncelli, M.G.B.; Karp, S.G.; Soccol, C.R. Chemical Composition and Health Properties of Coffee and Coffee By-Products. In Advances in Food and Nutrition Research; Elsevier: Amsterdam, The Netherlands, 2020; Volume 91, pp. 65–96.
Munyendo, L.M.; Njoroge, D.M.; Owaga, E.E.; Mugendi, B. Coffee Phytochemicals and Post-Harvest Handling—A Complex and Delicate Balance. J. Food Compos. Anal. 2021, 102, 103995.
Wang, X.; Lim, L.-T. Physicochemical Characteristics of Roasted Coffee. In Coffee in Health and Disease Prevention; Academic Press: Cambridge, MA, USA, 2015; pp. 247–254.
Komes, D.; Bušić, A. Antioxidants in Coffee. In Processing and Impact on Antioxidants in Beverages; Academic Press: Cambridge, MA, USA, 2014; pp. 25–32.
Carneiro, S.M.; Oliveira, M.B.P.P.; Alves, R.C. Neuroprotective Properties of Coffee: An Update. Trends Food Sci. Technol. 2021, 113, 167–179.
dos Santos, É.M.; de Macedo, L.M.; Tundisi, L.L.; Ataide, J.A.; Camargo, G.A.; Alves, R.C.; Oliveira, M.B.P.P.; Mazzola, P.G. Coffee by-products in topical formulations: A review. Trends Food Sci. Technol. 2021, 111, 280–291.
Yisak, H.; Redi-Abshiro, M.; Chandravanshi, B.S. Selective Determination of Caffeine and Trigonelline in Aqueous Extract of Green Coffee Beans by FT-MIR-ATR Spectroscopy. Vib. Spectrosc. 2018, 97, 33–38.
Winkler, A. Coffee, Cocoa and Derived Products (e.g., Chocolate). In Food Safety Management; Academic Press: Cambridge, MA, USA, 2014; pp. 251–282.
Combes, M.-C.; Joët, T.; Lashermes, P. Development of a Rapid and Efficient DNA-Based Method to Detect and Quantify Adulterations in Coffee (Arabica versus Robusta). Food Control 2018, 88, 198–206.
Diaz-de-Cerio, E.; Guerra-Hernandez, E.; Garcia-Estepa, R.; Garcia-Villanova, B.; Verardo, V. Analytical Approaches in Coffee Quality Control. In Caffeinated and Cocoa Based Beverages; Woodhead Publishing: Sawston, UK, 2019; pp. 285–336.
Fioresi, D.B.; Pereira, L.L.; Catarina da Silva Oliveira, E.; Moreira, T.R.; Ramos, A.C. Mid Infrared Spectroscopy for Comparative Analysis of Fermented Arabica and Robusta Coffee. Food Control 2021, 121, 107625.
de Morais, T.C.B.; Rodrigues, D.R.; de Carvalho Polari Souto, U.T.; Lemos, S.G. A Simple Voltammetric Electronic Tongue for the Analysis of Coffee Adulterations. Food Chem. 2019, 273, 31–38.
Wang, J.; Jun, S.; Bittenbender, H.C.; Gautz, L.; Li, Q.X. Fourier Transform Infrared Spectroscopy for Kona Coffee Authentication. J. Food Sci. 2009, 74, 385–391.
Ribeiro, J.S.; Salva, T.J.; Ferreira, M.M.C. Chemometric Studies for Quality Control of Processed Brazilian Coffees Using Drifts. J. Food Qual. 2010, 33, 212–227.
Link, J.V.; Lemes, A.L.G.; Marquetti, I.; dos Santos Scholz, M.B.; Bona, E. Geographical and Genotypic Classification of Arabica Coffee Using Fourier Transform Infrared Spectroscopy and Radial-Basis Function Networks. Chemom. Intell. Lab. Syst. 2014, 135, 150–156.
Craig, A.P.; Franca, A.S.; Oliveira, L.S. Discrimination between Immature and Mature Green Coffees by Attenuated Total Reflectance and Diffuse Reflectance Fourier Transform Infrared Spectroscopy. J. Food Sci. 2011, 76, C1162–C1168.
Craig, A.P.; Franca, A.S.; Oliveira, L.S. Discrimination between Defective and Non-Defective Roasted Coffees by Diffuse Reflectance Infrared Fourier Transform Spectroscopy. LWT Food Sci. Technol. 2012, 47, 505–511.
Craig, A.P.; Franca, A.S.; Oliveira, L.S. Evaluation of the Potential of FTIR and Chemometrics for Separation between Defective and Non-Defective Coffees. Food Chem. 2012, 132, 1368–1374.
Bona, E.; Marquetti, I.; Link, J.V.; Makimori, G.Y.F.; da Costa Arca, V.; Guimarães Lemes, A.L.; Ferreira, J.M.G.; dos Santos Scholz, M.B.; Valderrama, P.; Poppi, R.J. Supp.ort Vector Machines in Tandem with Infrared Spectroscopy for Geographical Classification of Green Arabica Coffee. LWT Food Sci. Technol. 2017, 76, 330–336.
Obeidat, S.M.; Hammoudeh, A.Y.; Alomary, A.A. Application of FTIR Spectroscopy for Assessment of Green Coffee Beans According to Their Origin. J. Appl. Spectrosc. 2018, 84, 1051–1055.
Belchior, V.; Botelho, B.G.; Casal, S.; Oliveira, L.S.; Franca, A.S. FTIR and Chemometrics as Effective Tools in Predicting the Quality of Specialty Coffees. Food Anal. Methods 2020, 13, 275–283.
Reis, N.; Franca, A.S.; Oliveira, L.S. Discrimination between Roasted Coffee, Roasted Corn and Coffee Husks by Diffuse Reflectance Infrared Fourier Transform Spectroscopy. LWT Food Sci. Technol. 2013, 50, 715–722.
Reis, N.; Franca, A.S.; Oliveira, L.S. Performance of Diffuse Reflectance Infrared Fourier Transform Spectroscopy and Chemometrics for Detection of Multiple Adulterants in Roasted and Ground Coffee. LWT Food Sci. Technol. 2013, 53, 395–401.
Reis, N.; Franca, A.S.; Oliveira, L.S. Quantitative Evaluation of Multiple Adulterants in Roasted Coffee by Diffuse Reflectance Infrared Fourier Transform Spectroscopy (DRIFTS) and Chemometrics. Talanta 2013, 115, 563–568.
Reis, N.; Franca, A.S.; Oliveira, L.S. Concomitant Use of Fourier Transform Infrared Attenuated Total Reflectance Spectroscopy and Chemometrics for Quantification of Multiple Adulterants in Roasted and Ground Coffee. J. Spectrosc. 2016, 2016, 4974173.
Brondi, A.M.; Torres, C.; Garcia, J.S.; Trevisan, M.G. Differential Scanning Calorimetry and Infrared Spectroscopy Combined with Chemometric Analysis to the Determination of Coffee Adulteration by Corn. J. Braz. Chem. Soc. 2017, 28, 1308–1314.
Reis, N.; Botelho, B.G.; Franca, A.S.; Oliveira, L.S. Simultaneous Detection of Multiple Adulterants in Ground Roasted Coffee by ATR-FTIR Spectroscopy and Data Fusion. Food Anal. Methods 2017, 10, 2700–2709.
Craig, A.P.; Botelho, B.G.; Oliveira, L.S.; Franca, A.S. Mid Infrared Spectroscopy and Chemometrics as Tools for the Classification of Roasted Coffees by Cup Quality. Food Chem. 2018, 245, 1052–1061.
Flores-Valdez, M.; Meza-Márquez, O.G.; Osorio-Revilla, G.; Gallardo-Velázquez, T. Identification and Quantification of Adulterants in Coffee (Coffea arabica L.) Using FT-MIR Spectroscopy Coupled with Chemometrics. Foods 2020, 9, 851.
Correia, R.M.; Loureiro, L.B.; Rodrigues, R.R.T.; Costa, H.B.; Oliveira, B.G.; Filgueiras, P.R.; Thompson, C.J.; Lacerda, V.; Romão, W. Chemical Profiles of Robusta and Arabica Coffee by ESI(-)FT-ICR MS and ATR-FTIR: A Quantitative Approach. Anal. Methods 2016, 8, 7678–7688.
Medina, J.; Caro Rodríguez, D.; Arana, V.A.; Bernal, A.; Esseiva, P.; Wist, J. Comparison of Attenuated Total Reflectance Mid-Infrared, Near Infrared, and 1H-Nuclear Magnetic Resonance Spectroscopies for the Determination of Coffee’s Geographical Origin. Int. J. Anal. Chem. 2017, 2017, 7210463.
Belchior, V.; Botelho, B.G.; Oliveira, L.S.; Franca, A.S. Attenuated Total Reflectance Fourier Transform Spectroscopy (ATR-FTIR) and Chemometrics for Discrimination of Espresso Coffees with Different Sensory Characteristics. Food Chem. 2019, 273, 178–185.
Ruf, F.; Schroth, G.; Doffangui, K. Climate Change, Cocoa Migrations and Deforestation in West Africa: What Does the Past Tell Us about the Future? Sustain. Sci. 2015, 10, 101–111.
Franzen, M.; Borgerhoff Mulder, M. Ecological, Economic and Social Perspectives on Cocoa Production Worldwide. Biodivers. Conserv. 2007, 16, 3835–3849.
Nair, K.P. Cocoa (Theobroma cacao L.). In Tree Crops; Springer International Publishing: Cham, Switzerland, 2021; pp. 153–213.
Santos, I.A.; Conceição, D.G.; Viana, M.B.; de Jesus Silva, G.; Santos, L.S.; Ferrão, S.P. NIR and MIR Spectroscopy for Quick Detection of the Adulteration of Cocoa Content in Chocolates. Food Chem. 2021, 349, 129095.
Hu, Y.; Pan, Z.J.; Liao, W.; Li, J.; Gruget, P.; Kitts, D.D.; Lu, X. Determination of Antioxidant Capacity and Phenolic Content of Chocolate by Attenuated Total Reflectance-Fourier Transformed-Infrared Spectroscopy. Food Chem. 2016, 202, 254–261.
Batista, N.N.; de Andrade, D.P.; Ramos, C.L.; Dias, D.R.; Schwan, R.F. Antioxidant Capacity of Cocoa Beans and Chocolate Assessed by FTIR. Food Res. Int. 2016, 90, 313–319.
Mandrile, L.; Barbosa-Pereira, L.; Sorensen, K.M.; Giovannozzi, A.M.; Zeppa, G.; Engelsen, S.B.; Rossi, A.M. Authentication of Cocoa Bean Shells by Near- and Mid-Infrared Spectroscopy and Inductively Coupled Plasma-Optical Emission Spectroscopy. Food Chem. 2019, 292, 47–57.

© Text is available under the terms and conditions of the Creative Commons Attribution (CC BY) license; additional terms may apply. By using this site, you agree to the Terms and Conditions and Privacy Policy.

Upload a video for this entry

Information

Subjects: Spectroscopy

Contributor MDPI registered users' name will be linked to their SciProfiles pages. To register with us, please refer to https://encyclopedia.pub/register :

Aracely Caudillo

View Times: 397

Update Date: 01 Mar 2022

Table of Contents

Video Upload Options

Confirm

1. Coffee

2. Cocoa

References