Physicochemical characteristics of Ethiopian Coffea arabica cv. Heirloom coffee extracts with various roasting conditions

Abstract

This study aimed to investigate the physicochemical characteristics of Ethiopian Coffea arabica cv. Heirloom coffee extracts with various roasting conditions. Green coffee beans were roasted at four different conditions (Light-medium, Medium, Moderately dark, and Very dark) and used to extract espresso and drip coffee. Moisture content in coffee beans was decreased as the roasting degree increased. The contents of crude fat and ash were lower in the Light-medium roasted coffee beans than in green coffee beans but increased as the roasting degree increased. The values of lightness (L^*), redness (a^*), yellowness (b^*), and browning index of coffee extracts were decreased as the roasting degree increased. Total dissolved solids in espresso coffee were increased with increasing roasting degree but decreased in drip coffee. In both the extracts, the contents of reducing sugar, titratable acidity, organic acids, and chlorogenic acid were decreased, but that of caffeine was increased with the roasting degree increased.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10068-020-00865-w.

Introduction

Coffee is one of the widely consumed beverages in the world. According to the International Coffee Organization (ICO), world coffee consumption in 2019–2020 is estimated at 10,160.4 million kg, 8.2% higher than in 2015–2016 (ICO, 2020). The number of plants included in the genus Coffea is approximately 103 or more (Davis et al., 2006). Among them, the two species named Arabica (Coffea arabica) and Robusta (Coffea canephora) are widely cultivated for commercialization (ICO, 2020). Even with the same species, the appearance and flavor of the coffee vary according to cultivation regions. Ethiopia is the largest coffee producer in Africa, with a production of 466.8 million kg during the crop year in 2018 (ICO, 2020). Ethiopia mainly grows and exports Arabica and has a unique feature to grow heirloom varietals that have not yet been classified and cultivated in the world (Boot, 2011). Therefore, it is difficult to generalize the characteristic of Ethiopian coffee. Nevertheless, the fruit flavors, especially berry and citrus, and chocolate aromatics, are known to be typical in the coffee of all Ethiopia regions (Boot, 2011).

Roasting is an essential process that coffee beans produce their distinct colors and flavors by physicochemical changes. During the roasting, coffee beans evaporate their moisture, become porous, lower their density, and change their chemical compositions due to the Maillard reaction, caramelization, lipid oxidation, and decomposition of phenolic compounds (Stefanello et al., 2019). The impact of the roasting degree of coffee extracts also depends on the type of coffee beans and how they are brewed. Therefore, it is necessary to investigate the effect of roasting on various coffee species and coffee extracts. Previous studies on the physicochemical properties of coffee have mainly focused on the espresso and drip methods, mainly because these types of coffee extracts are commonly consumed (Cordoba et al., 2020). However, little has been reported about the chemical composition of coffee extracts prepared by the espresso and drip methods under the same conditions, except for the degree of roasting. Therefore, this study was conducted to investigate the impact of various roasting methods on the selected chemical constituents of Ethiopian Coffea arabica cv. Heirloom coffee extracts prepared by an expresso and drip method.

Materials and methods

Materials

The green coffee beans (Coffea arabica cv. Heirloom, Natural, G-4) used in this study were kindly supplied by Moi Coffee Company (Yongin, Korea). The coffee beans were collected from the plants grown in Ethiopia, harvested in December 2017.

Roasting

Green coffee beans were roasted using a roaster (Aillio Bullet R1 Roaster, Aillio, Taipei, Taiwan) preheated to 170 °C. The Agtron number was measured by a coffee roast degree analyzer (CM-100, Lighttells, Zhubei, Taiwan) to measure roasting degree based on the method of the Specialty Coffee Association of America (SCAA). The green coffee beans were roasted in four steps: Light-medium (Agtron 70.17–60.36), Medium (Agtron 60.07–50.00), Moderately dark (Agtron 49.75–45.18), and Very dark (Agtron 30.07–20.50). The temperatures, roasting time, and Agtron number of the coffee beans are presented in supplementary Table 1. The roasted coffee beans were stored at  − 18 °C until use.

Espresso extraction

Four types of espresso coffee extracts (Light-medium, E1; Medium, E2; Moderately dark, E3; and Very dark, E4) were prepared using the roasted coffee beans described above. Roasted coffee beans were ground into a size of 45 mesh using a semi-automatic grinder (900 N, Yang-Chia Machine Works Co., Ltd., Taiwan). The ground coffee powder (7 g) was extracted up to 30 mL using a semi-automatic espresso machine (E98 President A2, Faema, Milano, Italy).

Drip extraction

Four types of drip coffee extracts (Light-medium, D1; Medium, D2; Moderately dark, D3; and Very dark, D4) were prepared using the roasted coffee beans described above. The roasted coffee beans were ground into 25 mesh size using a grinder (Fuji coffee mill R-440, Fujikouki Co., Ltd., Osaka, Japan). The coffee powder (15 g) was extracted up to 210 mL with a Clever Dripper (Mr. Clever, EK Int'l Co., Ltd., Taipei, Taiwan).

Weight loss

The weight loss of coffee beans due to roasting was measured according to the Association of Official Analytical Chemists (AOAC, 1995) method and expressed as a percentage of the sample weight before roasting.

$$\mathrm{Weight}\mathrm{loss}\left(\%\right)=\left[\left(\mathrm{Initial}\mathrm{weight}-\mathrm{Final}\mathrm{weight}\right)/\mathrm{Initial}\mathrm{weight}\right]\times 100$$ 1

Proximate composition

The proximate composition of the coffee beans was analyzed by the method of AOAC (1995). The moisture content was determined by oven-drying at 105 °C. The crude protein content was measured by a Micro-Kjeldahl method. The crude fat content was analyzed using the Soxhlet apparatus and petroleum ether as the extraction solvent. The ash content was analyzed by ignition at 550 °C in an electric furnace. Proximate compositions are expressed as a percentage of the sample weight. The content of carbohydrates was calculated by subtracting the moisture content, crude protein content, crude fat content, and ash content from 100.

Color measurement

The colors of coffee beans and coffee extracts were measured using a colorimeter (LC100, Tintometer Limited, Amesbury, England), which relies on the L^*, a^*, and b^* parameters. L^* indicates lightness, and a^* and b^* indicate redness and yellowness, respectively. Browning index (BI) was calculated from the values of L^*, a^*, and b^* using equations previously described (Korley Kortei et al., 2015).

$$\mathrm{BI}=\left(\left(z-0.31\right)/0.17\right)\times 100$$ 2

where, $z=\left(a^{\ast }+1.75L^{\ast }\right)/\left(5.645L^{\ast }+a^{\ast }-3.012b^{\ast }\right)$.

Total dissolved solids

The content of total dissolved solids (TDS) in coffee extracts was calculated by multiplying the value of °Brix by 0.85 according to the method of Gómez (2019). The °Brix value was measured using a refractometer (HI 96,801, HANNA Instruments, Woonsocket, RI, USA) and expressed as an average value after three repeated measurements.

Reducing sugars

The determination of the content of reducing sugar (RS) in coffee extracts was performed based on the method of Al Loman et al. (2017). The espresso coffee was diluted five times, and drip coffee was diluted two times and used as a sample. After adding 0.3 mL of 3,5-dinitrosalicylic acid (DNS, Sigma-Aldrich, St. Louis, MO, USA) reagent to 0.3 mL of the sample, the mixture was reacted for 5 min in a water bath at 100 °C. After adding 0.9 mL of distilled water to the solution and incubating for 15 min at room temperature, the absorbance was measured at 550 nm using a microplate reader (Spectramax M2E, Thermo Fisher Scientific, Waltham, MA, USA). A standard curve was prepared using glucose (Duksan, Seoul, Korea). The RS content was back-calculated to the established standard curve and expressed in mg/mL.

pH and titratable acidity

The pH value of coffee extracts was determined using a pH-meter (ORION STAR A211, Thermo Fisher Scientific, Waltham, MA, USA). It was expressed as an average value after three repeated measurements. The titratable acidity (TA) of the extract was measured based on the method of AOAC (1995). The espresso and drip coffee were each diluted five times and used as samples. The TA was determined by titration with 0.1 N NaOH (Duksan) solution up to pH 8.0 using a 1% phenolphthalein indicator (Showa chemicals Inc., Tokyo, Japan). The volume of the NaOH solution consumed was converted to milliequivalents of citric acid (Sigma-Aldrich) and expressed as a TA.

Organic acid contents

The content of organic acids in coffee extracts was determined using high-performance liquid chromatography (HPLC) system (Ultimate 3000, Thermo Fisher Scientific, Waltham, MA, USA) equipped with a UV–vis detector (VWD-3100, Thermo Fisher Scientific) using the method of Jham et al. (2002) with slight modification. Espresso and drip coffee extracts were diluted 10 and 3 times, respectively, filtered through a syringe filter (pore size of 0.45 μm), and used as samples. A Capcell Pak C18 UG 120 column (4.6 × 150 mm, 5 μm; Shiseido Co., Ltd., Tokyo, Japan) set at 30 °C was used to separate the analysts. The 0.1% phosphoric acid (Samchun, Pyeongtaek, Korea) solution was used as eluent. The flow rate of the mobile phase was 0.4 mL/min. The injection volume was 20 μL. The detector was set at 220 nm for organic acids. Acetic acid, citric acid, L-malic acid, and oxalic acid from Junsei Chemical Co., Ltd. (Tokyo, Japan) were used as standard materials. Quantitative analysis was performed using calibration curves, and the contents of organic acids are expressed as mg/mL.

Caffeine and chlorogenic acid contents

Caffeine and chlorogenic acid in coffee extracts were analyzed using the HPLC system (Ultimate 3000, Thermo Fisher Scientific) equipped with a UV–vis detector (VWD-3100, Thermo Fisher Scientific) based on the method of Jung et al. (2017). Espresso and drip coffee extracts were diluted 25 and 10 times, respectively, filtered through a syringe filter (pore size of 0.45 μm), and used as samples. Each sample was separated using a Capcell Pak C18 UG 120 (4.6 × 150 mm, 5 μm; Shiseido) set at 40 °C. The mobile phase consisted of acetonitrile (13% A, Burdick and Jackson, Porter, IN, USA) and 0.4% phosphoric acid (87% B, Samchun). The flow rate of the mobile phase was 0.5 mL/min. The volume of injection was 20 μL. The detector was set at 290 nm. The caffeine and chlorogenic acid from Sigma-Aldrich were used as standard materials. Quantitative analysis was performed using calibration curves, and the contents of caffeine and chlorogenic acid are expressed as mg/mL.

Statistical analysis

All data are expressed as mean ± standard deviation (SD). Statistical analysis was performed using SPSS version 25 software (IBM Corporation, Armonk, NY, USA). Significant differences among the groups were confirmed by one-way analysis of variance (ANOVA), followed by Duncan's multiple tests. P-values less than 0.05 were considered significant.

Results and discussion

Weight loss

The weight loss of the coffee beans was increased as the roasting degree increased. Both Light-medium and Medium roasting reduced the weight of coffee beans by 12% (data not shown). At Moderately dark roasting, the weight of beans was decreased by 16%, and at Very dark roasting, the weight of beans was decreased by 20% (data not shown). These results follow the results by Perrone et al. (2010), which showed increased weight loss and the increasing roasting time. It might be presumably due to the loss of moisture and the generation of carbon dioxide and volatile substances from the organic compounds by heat (Jokanović et al., 2012).

Proximate composition

The changes in the contents of the moisture, crude protein, crude fat, ash, and carbohydrates in coffee beans during roasting are presented in Table 1. As was expected, the moisture content was decreased from 8.64 g/100 g in green coffee beans to 0.89 g/100 g in the Very dark roasted coffee beans as the roasting degree increased. The roasting did not affect the content of crude protein significantly. The crude fat content in Light-medium roasted coffee beans was significantly lower than that of the green coffee beans, however, the crude fat content was increased again as the roasting degree increased. The ash content in Light-medium roasted coffee beans was slightly lower than that of the green coffee beans, however, the content of ash increased as the roasting degree increased. The content of carbohydrates was the lowest in the green coffee beans. Light-medium, Medium, and Moderately dark roasting increased the content of carbohydrates. However, in Very dark roasted coffee beans, the carbohydrate content significantly was decreased compared to the Light-medium, Medium, and Moderately dark roasted coffee beans.

Table 1.

Proximate composition of coffee beans with various roasting degrees

Properties (g/100 g)	Green coffee beans	Roasted coffee beans
		Light-medium	Medium	Moderately dark	Very dark
Moisture	8.64 ± 0.41a	2.56 ± 0.11b	2.17 ± 0.16c	1.29 ± 0.04d	0.89 ± 0.04e
Crude protein	11.56 ± 0.35 ns (12.65 ± 0.44)ns,1)	12.90 ± 1.52 (13.24 ± 1.56)	12.49 ± 0.37 (12.77 ± 0.39)	11.81 ± 0.89 (11.96 ± 0.89)	12.92 ± 1.23 (13.04 ± 1.25)
Crude fat	11.69 ± 1.42a (12.80 ± 1.61)a	8.59 ± 1.25b (8.82 ± 1.28)b	9.08 ± 1.80b (9.28 ± 1.85)b	10.46 ± 1.12ab (10.59 ± 1.13)ab	12.99 ± 1. 07a (13.10 ± 1.07)a
Ash	4.13 ± 0.05c (4.52 ± 0.06)b	4.05 ± 0.09c (4.16 ± 0.09)d	4.10 ± 0.05c (4.19 ± 0.05)d	4.26 ± 0.06b (4.31 ± 0.06)c	4.61 ± 0.02a (4.66 ± 0.03)a
Carbohydrates2)	63.99 ± 2.11c (70.03 ± 2.01)bc	71.90 ± 2.75a (73.79 ± 2.84)a	72.16 ± 1.94a (73.76 ± 1.86)a	72.19 ± 0.33a (73.13 ± 0.31)ab	68.59 ± 0.83b (69.20 ± 0.82)c

Means with different letters in the same row are significantly different (p < 0.05)

^a−eDuncan's multiple range test in all samples

^nsNot significant

¹⁾Values in parenthesis indicate the percentage on a dry weight basis

²⁾Carbohydrates: 100 – (Moisture + Crude protein + Crude fat + Ash)

These results indicate that the roasting intensity has a significant effect on the content of moisture and the contents of crude fat, ash, and carbohydrates, as in other studies. It has been reported that in Sidama coffee beans of Ethiopia, the moisture and protein contents are decreased as the roasting temperature increases, while the contents of carbohydrate, fat, and ash are increased (Endeshaw and Belay, 2020). There are two main phases of coffee roasting, which are dehydration and pyrolysis. In the dehydration phase, most of the water is steeply lost, reaching deficient levels. During pyrolysis, there is still a loss of moisture with CO₂ and CO, but it proceeds at a prolonged rate (Geiger et al., 2005). The protein in coffee is closely related to the foaming properties of coffee. The crude protein content hardly changes during roasting, but the individual free amino acids decrease significantly above 180 °C (Casal et al., 2005). Lipids are one of the main components in coffee beans, of which triacylglyceride accounts for 75% of total coffee lipids (Toci et al., 2013). Most of the lipids in coffee beans are trapped within the cellular structure and are hardly affected by roasting. However, when coffee beans are strongly roasted, most of the cells are destroyed, and lipids can quickly move to the surface (Toci et al., 2013). The ash content of coffee beans is 3–4.5%, most of which are potassium, magnesium, and calcium. There is little change in ash level during roasting because roasting minimally affects mineral contents except for sulfur and phosphorus (Clarke, 2012). The polysaccharides in coffee are mainly composed of galactomannan and arabinogalactan contributing to the viscosity and foam stability in coffee extracts (Redgwell et al., 2002). Polysaccharides are relatively heat-stable compared to disaccharides and monosaccharides, but arabinogalactans are prone to decomposition during prolonged roasting (Redgwell et al., 2002). Therefore, it is thought that the contents of protein and ash hardly changed, but their overall proportions increased in part by the weight loss during roasting. It is also believed that the decomposition of polysaccharides due to the high heat and increased roasting time may be related to a sudden decrease in the other components.

Color and brown index

Color is one of the most critical appearance attributes of coffee, as it influences consumer acceptability. Besides consumer acceptability, the degree of color change is used for roasting process control. The effect of roasting on CIELAB color parameters of coffee beans are shown in Table 2. In coffee beans, the parameter L^*, which shows lightness, was the highest in green coffee beans and decreased progressively with the roasting degrees increased. The parameter a^*, which indicates redness, was the lowest in green coffee beans and steeply increased in Light-medium roasted coffee beans. However, the parameter was decreased significantly as the roasting degree increased. The parameter b^*, showing the degree of yellowness, was the highest in green coffee beans and tended to be decreased as roasting degree increased, to a similar extent to the parameter a^*. Browning index, representing the purity of brown color, was increased in the order of green coffee beans, Light-medium, and Medium as the roasting degree increased, but gradually decreased in the Moderately dark and Very dark roasted coffee beans. Table 3 represents the effect of roasting on L^*, a^*, and b^* parameters of coffee extracts prepared by espresso and drip methods. As roasting degree increased, the L^*, a^*, and b^* values and BI were decreased in both espresso and drip coffee.

Table 2.

Color of coffee beans with various roasting degrees

Properties1)	Green coffee beans	Roasted coffee beans
		Light-medium	Medium	Moderately dark	Very dark
L*	47.70 ± 0.10a,1)	37.49 ± 4.31b	31.83 ± 1.99c	29.99 ± 1.10c	23.84 ± 1.76d
a*	2.85 ± 0.05c	7.12 ± 1.08a	6.27 ± 1.08ab	5.85 ± 0.41b	3.64 ± 0.56c
b*	10.70 ± 0.00a	7.32 ± 1.24b	7.40 ± 1.81b	4.13 ± 1.03c	2.38 ± 0.72c
BI	29.33 ± 0.15b	35.16 ± 1.57ab	40.44 ± 8.16a	28.59 ± 3.30bc	21.09 ± 3.45c

Means with different letters in the same row are significantly different (p < 0.05)

^a−dDuncan's multiple range test in all samples

¹⁾L^*, lightness; a^*, redness; b^*, yellowness; BI, browning index

Table 3.

Color of coffee extracts with various roasting degrees

Properties1)	Espresso				Drip
	E12)	E2	E3	E4	D1	D2	D3	D4
L*	11.60 ± 0.17A	9.93 ± 0.80B	7.13 ± 0.31C	6.17 ± 0.67C	18.77 ± 0.42a	16.73 ± 0.21b	13.60 ± 0.44c	10.13 ± 1.31d
a*	7.17 ± 0.21A	5.60 ± 0.36B	3.03 ± 0.25C	1.90 ± 0.26D	20.40 ± 0.20a	19.77 ± 0.50a	17.03 ± 0.21b	13.80 ± 1.35c
b*	7.87 ± 0.23A	5.07 ± 0.35B	2.30 ± 0.30C	0.53 ± 0.32D	20.30 ± 0.56a	17.03 ± 0.78b	11.57 ± 0.42c	5.57 ± 0.59d
BI	147.82 ± 7.41A	110.17 ± 17.94B	68.75 ± 7.98C	31.27 ± 11.08D	298.79 ± 21.60a	276.93 ± 23.22a	225.71 ± 18.16b	163.01 ± 36.21c

Means with different letters in the same row are significantly different (p < 0.05)

^A−GDuncan's multiple range test in espresso

^a−dDuncan's multiple range test in drip

¹⁾L^*, lightness; a^*, redness; b^*, yellowness; BI, browning index

²⁾E1, Light-medium espresso; E2, Medium espresso; E3, Moderately dark espresso; E4, Very dark espresso; D1, Light-medium drip; D2, Medium drip; D3, Moderately dark drip; D4, Very dark drip

It has been reported that the L^* and b^* values of coffee beans are decreased as the roasting degree increased, and a^* value is higher than that of green coffee beans at the light roasting but significantly decreased as the roasting degree increased (Bicho et al., 2012). In a study by Bauer et al., the L^*, a^*, and b^* values of Robusta coffee were decreased as the roasting degree increased (Bauer et al., 2018). It has also been reported that the brownness of the coffee extract is increased with increasing roasting temperature and time, then decreases again, at 191.34 °C and 6.72 min (Chung et al., 2013). The increase in the parameter a^* was attributed to the formation of brown pigments. The brown pigments increase during roasting as the caramelization reaction and the Maillard reaction progress (Friedman, 1996). When coffee beans are roasted, the Malliard reaction produces melanoidins responsible for the browning of coffee beans. The color changes in coffee beans can be attributed to browning reactions and the destruction of pigments existing in coffee beans. Lozano and Ibarz (1997) have reported that thermal processing is one of the causes of color degradation in dehydrated products. Glyceraldehyde and glycolaldehyde, produced by the decomposition of chlorogenic acid, also increases brownness (Rhi and Shin, 1993).

Total dissolved solids

The content of TDS in coffee extracts is shown in Table 4. TDS refers to the ratio of soluble parts of the coffee that are dissolved by water and determines the intensity of coffee flavors. The particle size of a coffee grind is one of the critical factors in coffee extraction in that it affects both the fluid flow through the grind and the extractable mass of the grind. In espresso, coffee beans are ground relatively finely because the brewing time is short (Derossi et al., 2018). Thus, coffee beans were ground into particles of size 45 and 25 mesh in this study and used for espresso and drip coffee extraction, respectively. The TDS in espresso coffee was increased as the roasting degree increased, which varied from 4.25% in E1 to 4.96% in E4. On the other hand, the TDS in drip coffee was decreased as the roasting degree increased, which varied from 1.33% in D1 to 1.08% in D4. It has been reported that coffee extracts are ideal when their TDS is between 1.15% and 1.35% (SCAA, 2016). In a previous study, the total solids of the Brazil espresso coffee are increased with increasing roasting degree (Nunes et al., 1997). In contrast, Kim and Kim (2017) have reported that the total soluble solids content of drip coffee decreases as the roasting time increases. The TDS of coffee varies depending on the roasting degree, particle size, water temperature, and extraction time (Cordoba et al., 2020). When coffee beans are roasted, the melanoidins and caffeine that are easy to dissolve in the water increase, lipids are not technically soluble in water but can be released by water from the coffee cells and remain emulsion. In our results, the TDS in espresso coffee was increased as the roasting degree increased, whereas the drip coffee did not. It is assumed that the ingredients increased by the roasting were not sufficiently eluted in drip coffee partially due to the drip filter, low water pressure, and the large particle size of the coffee powder.

Table 4.

Total dissolved solids, reducing sugar, pH, and, titratable acidity of coffee extracts with various roasting degree

Properties1)	Espresso
	E12)	E2	E3	E4
TDS (%)	4.25 ± 0.09D	4.48 ± 0.10C	4.76 ± 0.00B	4.96 ± 0.05A
RS (mg/mL)	3.64 ± 0.02A	3.08 ± 0.19B	3.18 ± 0.21B	3.26 ± 0.20B
pH	4.79 ± 0.02C	4.84 ± 0.02C	5.05 ± 0.01B	5.67 ± 0.11A
TA (%)	1.51 ± 0.02A	1.41 ± 0.02B	1.44 ± 0.00B	1.10 ± 0.03C

Means with different letters in the same row are significantly different (p < 0.05)

^A−GDuncan's multiple range test in espresso

^a−dDuncan's multiple range test in drip

¹⁾TDS, total dissolved solids; RS, reducing sugar; TA, titratable acidity

²⁾E1, Light-medium espresso; E2, Medium espresso; E3, Moderately dark espresso; E4, Very dark espresso; D1, Light-medium drip; D2, Medium drip; D3, Moderately dark drip; D4, Very dark drip

Reducing sugars

The contents of RS in coffee extracts are presented in Table 4. RS is a kind of sugar with a free aldehyde group or a free ketone group acting as a reducing agent. In espresso coffee, E1 had the most RS than E2, E3, and E4. The amounts of RS in E2, E3, and E4 were not statistically different from each other. RS in drip coffee was the highest in D1 and decreased as the roasting degree increased. It has been reported that the RS content is decreased during roasting due to the Maillard reaction, thermal decomposition, hydrolysis, and oxidation (Bertuzzi et al., 2020). In the Maillard reaction, melanoidins are produced between the amino group of the protein and the carbonyl group of the RS. Also, a wide range of flavor compounds is produced from the reaction of the free aldehyde group and free ketone group in RS, which contributes to the complexity of aromas of the brewed coffee.

pH and titratable acidity

The sourness of coffee is a significant factor in determining the taste of coffee, along with bitterness, savory taste, and sweetness, and is related to pH and TA. The pH is an indicator of the acidity by measuring the concentration of H₃O⁺. TA refers to the concentration of total acids contained in foods and better predicts the effect of acids on flavor than pH by reflecting the effect of acids that are not degraded. Table 4 shows the pH and TA of the coffee extract according to the degree of roasting. In both espresso and drip coffee, the pH values were the lowest in Light-medium roasted coffee (E1 and D1) and increased as the roasting degree increased. On the contrary, in both espresso and drip coffee, TA values were the lowest in Very dark roasted coffee (E4 and D4). In the study by Wang and Lim (2012), it has been reported that the TA of the coffee extract is increased sharply from fresh green coffee beans to medium roast due to the maximum formation of total aliphatic acids. It then gradually decreases to dark roast to decompose organic acids (citric, malic, lactic, pyruvic, and acetic acids).

Organic acid contents

Organic acid content is one of the critical factors in coffee quality affecting the aroma and taste of the coffee (Borem et al., 2016). Table 5 shows the organic acid contents of the coffee extracts with various degrees of roasting. As the roasting degree increased, the concentrations of citric acid, L-malic acid, and oxalic acid were decreased in both espresso and drip coffee. In contrast, the acetic acid levels were not significantly different. Consequently, the total organic acid levels in both espresso and drip coffee decreased with the roasting degree increased. These results are in agreement with the results reported by the Coffee Research Institute (CRI) that citric acid and malic acid were found the most in light-roasted coffee (CRI, 2020). Kim and Kim (2017) have reported that coffee using beans roasted for a short time had the highest organic acid content and that coffee using beans roasted for a long time had the lowest organic acid content. By type, citric acid, malic acid, and formic acid were decreased as the roasting time increased, whereas succinic acid and lactic acid were increased as the roasting time increased (Kim and Kim, 2017).

Table 5.

Organic acids content in coffee extracts prepared from coffee beans with various roasting degrees

Properties (mg/mL)	Espresso				Drip
	E12)	E2	E3	E4	D1	D2	D3	D4
Acetic acid	0.64 ± 0.14 ns	0.63 ± 0.10	0.55 ± 0.05	0.48 ± 0.01	0.19 ± 0.03 ns	0.19 ± 0.02	0.19 ± 0.01	0.16 ± 0.04
Citric acid	0.99 ± 0.19A	1.03 ± 0.15A	0.68 ± 0.18B	0.30 ± 0.01C	0.32 ± 0.04a	0.27 ± 0.04a	0.19 ± 0.03b	0.08 ± 0.01c
L-malic acid	0.56 ± 0.05B	0.64 ± 0.06A	0.53 ± 0.03B	0.18 ± 0.01C	0.21 ± 0.07a	0.18 ± 0.01a	0.14 ± 0.01ab	0.08 ± 0.03b
Oxalic acid	5.65 ± 0.25A	5.68 ± 0.13A	4.73 ± 0.07B	2.86 ± 0.29C	1.74 ± 0.09a	1.70 ± 0.01a	1.29 ± 0.06b	0.74 ± 0.08c
Total1)	7.84 ± 0.52A	7.99 ± 0.23A	6.50 ± 0.31B	3.82 ± 0.30C	2.46 ± 0.10a	2.34 ± 0.04a	1.80 ± 0.08b	1.06 ± 0.12c

Means with different letters in the same row are significantly different (p < 0.05)

^A−DDuncan's multiple range test in espresso

^a−dDuncan's multiple range test in drip

^nsNot significant

¹⁾Total: acetic acid + citric acid + L-malic acid + oxalic acid

²⁾E1, Light-medium espresso; E2, Medium espresso; E3, Moderately dark espresso; E4, Very dark espresso; D1, Light-medium drip; D2, Medium drip; D3, Moderately dark drip; D4, Very dark drip

Caffeine and chlorogenic acid contents

Caffeine and chlorogenic acid are the main bioactive components of coffee. Caffeine is a kind of alkaloid that exhibits vasodilation, central nervous system stimulation, and antioxidant activities (Stefanello et al., 2019). Chlorogenic acid is a water-soluble ester compound formed between quinic acid and caffeic acid. It has various physiological activities such as antioxidative action, hypoglycemia, liver protection, and antiviral effect (Stefanello et al., 2019). The caffeine and chlorogenic acid contents of the coffee extract according to the degree of roasting are shown in Fig. 1. In espresso coffee, the content of caffeine was the lowest in E1 and increased as the roasting degree increased. On the contrary, chlorogenic acid's content was the highest in E1 and decreased as the roasting degree increased to a similar extent of the caffeine but in opposite directions. In drip coffee, similar to the espresso, the caffeine content was the lowest in D1 and increased as the roasting degree increased. In contrast, the chlorogenic acid content was the highest in D1 and decreased as the roasting degree increased.

Fig. 1.

Caffeine and chlorogenic acid contents in coffee extracts prepared from coffee beans with various roasting degrees. Coffee extract prepared by an espresso (A) and drip (B) method. Means with different letters are significantly different at p < 0.05 by Duncan's multiple range test. E1, Light-medium espresso; E2, Medium espresso; E3, Moderately dark espresso; E4, Very dark espresso; D1, Light-medium drip; D2, Medium drip; D3, Moderately dark drip; D4, Very dark drip

In the study of Kim and Kim (2017), the caffeine content of the drip coffee was 122.71–129.87 mg/100 g, which increased with increased roasting time. It has been reported that the chlorogenic acid content in the coffee beans are decreased as roasting degree increased, and 5-CQA, which occupies more than 50% of the total chlorogenic acid, are decreased 99.3–99.7% compared to green coffee beans at the French roasting (Moon et al., 2009). Caffeine has little loss during thermal treatment, especially to 250 °C because it has a high sublimation point at 178 °C, and the internal pressure of coffee beans increases by heat so that the sublimation point of caffeine increases (Chindapan et al., 2019). Chlorogenic acid is mostly hydrolyzed and generates quinic acid and various other substances, among which phenols affect the aroma, and some decomposition substances have a sour and astringent taste (Moon et al., 2009). Certain levels of caffeine and chlorogenic acid in roasted coffee can be necessary because of their physiological activities. Therefore, it is desirable to prepare coffee in such a way that the coffee is sensually good and has caffeine and chlorogenic acid sufficient to exert health benefits.

In conclusion, these results indicate that roasting degree (Light-medium, Medium, Moderately dark, and Very dark) influences the color, TDS, RS, pH, TA, organic acids, and the contents of bioactive compounds, such as caffeine and chlorogenic acid, in coffee extracts. The Light-medium roasted coffee showed higher values of the browning index, RS, TA, organic acids (citric acid, L-malic acid, and oxalic acid), and chlorogenic acids than those of the Medium-, Dark- and Very dark-roasted coffee. Very dark-roasted coffee had the highest values of pH and caffeine. It is interesting to note that the TDS in expresso coffee increased while decreasing in drip coffee after roasting, suggesting that roasting degree influenced both positively and negatively depending on the brewing method. Therefore, it is concluded that the degree of roasting is an essential factor determining the characteristics of the coffee extracts.

Supplementary Information

Below is the link to the electronic supplementary material.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.