Quality of coffee cultivated in Campos Gerais , Minas Gerais

Coffee quality can be defined as a set of physical, chemical, sensory and safety attributes that meet consumer taste preferences. Coffea arabica is the most cultivated species, accounting for 70% national and world production of coffee. The present study analyzed 6 coffee samples from 6 different regions in the municipality of Campos Gerais, subjected to the treatments raw and roasted, in order to assess their physical and chemical quality based on the parameters of titratable acidity, pH, moisture, soluble solids content and caffeine content. All the samples met the standard of at least one of the quality parameters studied. The best value of total acidity in roasted coffee was observed in the Grupiara region. The raw coffee of the Galo region presented the best quality as to the pH. The sample of the Prata region showed the highest caffeine content related to the quality of raw and roasted coffee, the best content for total acidity for raw coffee and the best pH value for roasted coffee. Therefore, the coffee of the Prata region is considered the best quality coffee analyzed in this study.


Introduction
Coffee industry has evolved over the last decade to meet market demand.High technology levels have been required by the coffee sector to increase productivity in Brazil, with a production estimated at 44.6 million bags in 2014, reducing costs and restricting the use of agrochemicals.Moreover, never before coffee quality has been so appreciated due to the increasing expansion of consumption of specialty coffees, about 12 million bags per year in Brazil (Silva, Rabelo et al., 2014).
Brazil, as a major producer and exporter of coffee, needs to deepen the knowledge about the fruit and its quality (Baliza et al., 2012).Seeking efficiency and excellence in quality, favoring the export market and investment in the domestic market (Jöet et al., 2010;Barbosa et al., 2012).Coffee quality can be influenced by climatic, genetic or sanitary factors during the diverse production phases (Silva et al., 2005).Losses are observed in postharvest, processing or during drying process, which is one of the key steps that result in quality loss, especially if it is poorly conducted and performed integrally in terraces (Camargo, 2010;Malerbo-Souza, Halak, 2012).
Several chemical compounds found in coffee beans are used, during and after roasting, through chemical processes, resulting in important constituents of the aroma and flavor of the drink (Franca, Mendonça, & Oliveira, 2005).The chemical composition of coffee can change according to the climate, region, altitude, soil, species and others (Silva, Junior et al., 2014).Desirable and undesirable compounds with perceptible flavor and aroma can be formed depending on the chemical composition of coffee bean (Farah, Monteiro, Calado, Franca, & Trugo, 2006).
In this context, aiming to analyze the quality of coffee grown in Campos Gerais, Minas Gerais State, the present study investigated some physical and chemical characteristics of raw and conventionally roasted coffee samples produced and processed in conventional crops of small farmers in the region and marketed by a particular company of buying and selling coffee.

Raw material
This study analyzed samples of Arabica coffees from the harvest of 2011 purchased in a coffee trading company located in Campos Gerais, Minas Gerais State.Beans were received as raw and dried grains, from conventional crops, dried on brick yards (dry processing).All samples were from the municipality of Campos Gerais, at 843 m above sea level, and geographical coordinates 21°14'07" South latitude and 45°45'12" West longitude, with tropical climate of altitude.At the harvest time, the weather was dry with temperatures ranging from 9 to 25°C and rainfall of 9-23 mm.
All the experiment was performed in the Laboratory of Biochemistry and Food Science, Faculty of Science and Technology of Campos Gerais.

Experimental design
This was a completely randomized design (CRD) in a (2 x 6) factorial arrangement of 2 treatments (raw -ground, roasted -ground) and 6 regions of Campos Gerais, corresponding to the region where coffee was grown, namely: Galo, (GA) with humid tropical climate and 840 m altitude; Amoras (AM) with humid tropical climate and 810 m altitude; Sapezinho (SP) with humid tropical climate and 830 m altitude; Dois Paus (DP) with humid tropical climate and 845 m altitude; Grupiara (GR) with humid tropical climate and 832 m altitude and Prata (PR) with humid tropical climate and 855 m altitude, with 4 repetitions for each treatment.

Statistical analysis
The results were tested by analysis of variance (ANOVA), using the statistical software SISVAR (Ferreira, 2011).All statistical tests were performed at the 5% significance level.

Sample preparation
The samples were divided into two equal portions, one subjected to commercial medium roast (55-65 Agtron) and ground in a conventional Willy grinder; while the other was kept raw and subjected to crushing in Willy mill.
After treatment, all samples were stored in amber package hermetically sealed and kept in a dry, wellventilated and dark room.

Moisture determination
The moisture content was determined by the gravimetric method of the Association of Official Analytical Chemistry (AOAC, 2005), which consists of water loss by dehydration at temperatures of 100 to 105ºC.

Preparation of aqueous extracts
For each sample, it was weighed 2 g coffee powder, which was mixed with 25 mL distilled water and allowed to rest for 48 hours under refrigeration, according to the Association of Official Analytical Chemistry (AOAC, 2005).
The resulting solutions were filtered through filter paper and insoluble parts soaked in additional 25 mL distilled water during the same time period.The new mixtures were filtered and the liquid obtained during both filtrations were combined resulting in the studied extracts.In total, 12 extracts were prepared, 6 from raw coffee powder and 6 from roasted coffee powder, for evaluation of pH, total solids and titratable acidity.

Chemical analysis pH determination
pH was determined by potentiometry with a glass electrode, using a digital pHmeter QUIMIS, following the technique of AOAC (2005).

Total Titratable Acidity (ATT)
Total titratable acidity was determined by titrating the filtrate with a 0.1 M NaOH standard solution, according to AOAC (2005).The results were expressed as g of total acids per 100 g coffee powder.

Determination of caffeine
Caffeine was determined by acid extraction, that is, there is a selective carbonization of organic matter in the sample with sulfuric acid releasing caffeine that was later extracted using chloroform according to the method of AOAC (2005).Results were obtained by spectrometer at 320 nm and expressed in mg caffeine per 100 g solution.

Total Titratable Acidity (ATT)
The values of total titratable acidity of coffee beans can vary according to fermentation levels occuring in the grains, with the different maturity stages and can represent an auxiliary analysis for assessing the quality of the coffee beverage.The total titratable acidity (ATT) of coffee beans has an inverse relationship with the coffee beverage quality, because better quality coffees have lower total titratable acidity (Martinez, Poltronieri, Farah, & Perrone, 2013).
The values of titratable acidity (Table 1) were significantly different between raw and roasted treatments only for Grupiara (13.0 0.1 M NaOH 100 g -1 for raw coffee and 10.7 0.1 M NaOH 100 g -1 for roasted coffee) and Prata (21.3 0.1 M NaOH 100 g -1 for raw coffee and 8.0 0.1 M NaOH 100 g -1 for roasted coffee).

Moisture
Values of moisture content in Table 2 showed significant interaction between raw and roasted treatments.In raw coffee, the lowest average moisture was observed in the sample Dois Paus (9.3%) followed by Amoras (9.5%) and the samples Galo (9.9%), Prata (9.9%), Sapezinho (10.0%) and Grupiara (10.0%), which showed no significant differences to each other.
Moisture content of raw coffee registered in this study is within the recommended range by the Normative Instruction 8 of June 11 st , 2003, the Ministry of Agriculture, which states that the moisture content of processed coffee (raw bean) may not exceed the maximum of 12.5% (Brasil, 2003).
During the drying of coffee, water content is reduced from 60% to close to 11% (Mendonça, Franca, & Oliveira, 2007).The reduction of water in the fruit provides safe storage conditions for long periods in order to minimize the respiration rate and the risk of oxidation, fermentation and development of microorganisms (Isquierdo, Borém, Oliveira, Siqueira, & Alves, 2012).
The values found in this experiment are within the range allowed by ANVISA (National Health Surveillance Agency), which states that roasted coffee must contain a moisture content of at most 5.0% (Brasil, 2005).
pH Changes in pH caused by roasting can be of great importance in the acceptance of the product by the consumer.The pH is one of the quality attributes that helps professionals to classify a coffee as palatable without excess bitterness or acidity among other coffees (Franca et al., 2005).
The pH values found in the experiment are listed in Table 3.There were significant differences between samples and treatments, except for the sample of Amoras.In raw coffee, the lowest pH values were verified in the samples of Prata (5.59) and Sapezinho (5.4), followed by Grupiara (5.75), and subsequently Dois Paus (5.84) and Amoras (5.87) and Galo (6.0).

Table 1 .
Mean total titratable acidity (0.1 M NaOH 100 g -1 ) in different types of raw and roasted coffees.Means followed by different lowercase letters in the same row and uppercase letters in the same column are significantly different by Tukey's test at 5% significance. *

Table 2 .
Moisture (%) in different types of raw and roasted coffees.Means followed by different lowercase letters in the same row and uppercase letters in the same column are significantly different by Tukey's test at 5% significance. *

Table 3 .
Mean pH in different types of raw and roasted coffees.Means followed by different lowercase letters in the same row and uppercase letters in the same column are significantly different by Tukey's test at 5% significance. *