Roasting is the most common method of processing coffee. During roasting, aromatic compounds are generated due to various reactions, which are important for developing color, flavor and aroma. Acrylamide is an undesirable carcinogenic substance that is metabolically activated and formed during the coffee roasting process. Coffea arabica was first found in Ethiopia, and Ethiopia can produce a large volume of coffee. The major coffee-producing areas in Ethiopia are Hararghe, Sidama, Gimbi/Nekemte, Yergachefe and Limu. The primary purpose of this study was to quantify the acrylamide contents of brewed and roasted coffee collected from street coffee sellers and industrial processors found in Addis Ababa, Ethiopia, and optimize the roasting conditions for Sidama coffee. The acrylamide contents were determined by HPLC using a DAD at 210 nm, the antioxidant property were examined using a UV–spectrophotometer, and moisture and nutrient composition of coffee was determined using the method described by the AOAC (Association of Official Analytical Chemists). The roasting temperature and time were optimized based on the acrylamide content, nutritional composition and antioxidant property of the coffee using central composite design. The roasting temperature and time significantly affected (p<0.05) the acrylamide level, nutritional composition and antioxidant property of the coffee. The acrylamide contents of street and industrial processed powdered coffee were 346 ±19 to 701±38μg/kg and 442±14 to 906±7μg/kg, respectively. Brewed coffee from street vendors and industrial processing had acrylamide contents of 25±2 to 49±1μg/L and 63±2 to 89±4μg/L, respectively. The EC50 values for scavenging radicals for the optimized coffee ranged from 171±0 to 111±4 μg/L. The optimal roasting temperature and time were 190°C and 6 minutes, at this temperature and time the acrylamide content decreased, and the antioxidant and nutritional compositions of the coffee improved.
Citation: Endeshaw H, Belay A (2020) Optimization of the roasting conditions to lower acrylamide content and improve the nutrient composition and antioxidant properties of Coffea arabica. PLoS ONE 15(8): e0237265. https://doi.org/10.1371/journal.pone.0237265
Editor: Branislav T. Šiler, Institute for Biological Research "S. Stanković", University of Belgrade, SERBIA
Received: April 23, 2020; Accepted: July 22, 2020; Published: August 25, 2020
Copyright: © 2020 Endeshaw, Belay. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: This work was supported by the Department of Food Science and Applied Nutrition, Addis Ababa Science and Technology University, Ethiopia (www.aastu.edu.et). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Coffee is the most frequently consumed beverage around the world . It has a good fragrance, taste, health benefits and caffeine content, which make it a highly acceptable beverage throughout the world. Coffea arabica is cultivated in Ethiopia, and a large volume of coffee can be produced in this country. According to Chauhan et al. , Ethiopia’s cool climatic conditions and rainfall create favorable conditions for coffee cultivation.
Food processing is an important factor in improving the nutritional qualities, microbiological safety, and sensory qualities while removing some compounds with potential negative health impacts . Roasting is the most well-known processing method and is important for coffee preparation. During roasting, aromatic compounds are generated due to different reactions, such as the Maillard reaction, Strecker degradation, the degradation of sugars and the breakdown of amino acids, which are important for developing color, flavor and aroma . Acrylamide is a thermal processing contaminant that is produced in the Maillard reaction, a series of nonenzymatic reactions between reducing sugars and free amino acids [4,5]. The products of fat breakdown, mainly acrolein and acrylic acid, have also been found to be precursors for acrylamide formation [5,6]. Acrylamide is not found in raw and unprocessed food items. It is formed when food containing certain precursors are thermally processed at temperatures of 120°C or above, such as by frying, baking, roasting, grilling and toasting [7,8].
According to Morales et al. , compounds with health benefits, such as antioxidants, can be generated during roasting. However, heat-induced toxic food contaminants with carcinogenic and genotoxic properties such as heterocyclic amines, acrylamide and furfurals can also be generated during roasting, especially if the processing temperature and time are not tightly controlled . The International Agency for Research on Cancer (IARC) has classified acrylamide as a probable human carcinogen . Roasting coffee at very high temperatures also causes the degradation of antioxidant compounds, such as caffeine, chlorogenic acid (CGA) and Maillard reaction products (melanoidins) .
The European Commission (EC) developed recommendations and references in 2013 to control the acrylamide content of food products. The reference values for roasted and instant coffee were 450 μg/kg and 900 μg/kg, respectively . The newly revised EC recommendations for acrylamide in roasted and instant coffee are 400 and 850 μg/kg, respectively . Dibaba et al.  reported that the level of acrylamide in ‘Keribo’, a traditional Ethiopian fermented beverage, was as high as 3440 μg/kg. However, the content of acrylamide and the optimal time and temperature for coffee roasting were not investigated for Ethiopian coffee. In this study, the acrylamide contents of brewed and roasted coffee samples that were prepared by street coffee brewers and coffee processors was examined, and the roasting temperature and time were optimized for Sidama coffee. Accordingly, the acrylamide content, antioxidant properties and nutritional composition were determined.
Materials and methods
Coffee samples were collected based on Madihah et al. . Street and industrially processed brewed and powdered coffee were obtained from local markets in Addis Ababa sub-cities (Arada, Akaki-Kality, Nifas Silk-Lafto and Kirkose); using randomized lottery sampling methods; and Sidama coffee was used in the optimization of roasting temperature and time. Sidama coffee is one of the majorly produced and highly exported coffee type in Ethiopia, and was collected from Ethiopian Coffee Quality Inspection and Certification Center (ECQICC), Addis Ababa.
Response surface design through central composite design (CCD) used to optimize the roasting time and temperature with 14 runs. The roasting temperature and time ranged from 180°C to 200°C and 4 to 8 minutes, respectively. A quadratic model was used to find the optimal roasting conditions.
Roasting of the coffee beans
The coffee beans were roasted based on Madihah et al. , with some modifications. Accordingly, a Probat roaster machine (PROBAT, Gimborn Maschinenfabrik GmbH, Germany, Emmerich am Rhein) was used for the roasting process. The machine was first heated to 170°C, and 100 g of green coffee beans put into the roasting cylinder. The coffee beans were roasted for different combinations of roasting time and temperature using CCD (170 to 200°C for 4 to 8 minutes). The roasted beans were tipped out into a cooling tray and allowed to cool for an average of 4 minutes by blowing cold air over the cooling plate. Visual inspection was used to separate the roasted coffee beans based on color using an AGTRON- Specialty Coffee Association of America (SCAA, 1995). The roasted coffee beans were cooled, finely ground in a coffee grinder (Mahlkoning grinder, Hemro Manufacturing, Germany, Hamburg) to a screen size of 3.5 (0.30 mm), packed in a sealed plastic bag and stored at 4°C until analysis.
The roasted coffee samples were extracted with water using SPE cleanup for acrylamide analysis, following the method described by Oroian et al. . Accordingly, 2.5 g of coffee powder and 10 ml of brewed coffee were weighed in a 50 ml centrifuge tube. Ten milliliters of water and 5 ml of n-hexane were added, and the sample was mixed for 5 minutes and then centrifuged (Universal 320, Andreas Hettich GmbH & Co. KG, Tuttlingen, Germany) at 2509 g (4000 rpm) for 5 minutes. The upper hexane layer was then discarded, and 1 ml of Carrez I solution and 1 ml of Carrez II solution were added. Then, the mixture was centrifuged at 2509 g (4000 rpm) for 5 minutes. The clear intermediate layer was again purified by adding 1 ml of Carrez I solution and 1 ml of Carrez II. The sample was mixed vigorously for 5 minutes and then centrifuged at 2509 g (4000 rpm) for 5 minutes. The clear layer was transferred to an SPE column previously conditioned with 1 ml of methanol and 1 ml of water. The extract was passed through the cartridge. The first five drops were discarded, and 4 ml of filtrate was collected and concentrated to dryness using nitrogen gas. The residue was reconstituted in 2 ml of water. It was then filtered through a 0.20 μm filter disk and transferred to autosampler vials for HPLC-DAD determination of the acrylamide.
The acrylamide contents of different roasted coffee samples were determined by HPLC (Agilent 1260 infinity HPLC, Agilent technologies, USA, Minnesota). The separation of the compounds was achieved using an Agilent ZORBAX Eclipse Plus column) (Agilent ZORBAX Eclipse plus column 4.6*100mm, 3.5μm, Agilent technologies, USA, Minnesota.). The data were processed with Agilent Open Lab CDS software (Agilent Open Lab CDS software, Agilent technologies, USA, Minnesota). The flow rate was 0.50 ml/ minutes, and the column oven was fixed at 30°C. The total run time was 15 minutes with 10 minutes for acquisition and 5 minutes for postrun. The injection volume was 20 μl. The mobile phase was 99% HPLC-grade water and 0.1% acetic acid solution. Acrylamide was detected using a diode array detector set at 210 nm . The linearity of the calibration curve was evaluated by plotting peak area as a function of analyte concentration and assessing the goodness of fit of the linear regression . Sensitivity of the method was determined with the Limit of Detection (LOD) 5 μg/l and Limit of Quantification (LOQ) 20.40 μg/kg . Relative standard deviation percent (RSD%) was evaluated by analyzing replicates of acrylamide and was obtained 2.63% .
Nutrient composition determination
The moisture content was determined using AOAC  method 925.09. AOAC  method 941.12 was used to determine the ash content. AOAC  method 4.5.01 was used to determine the total crude fat content. The protein content of roasted coffee beans was determined using AOAC  method 979.09. The crude fiber content of the roasted coffee powder was determined using AOAC  method 920.169. AOAC  method 985.29 was used to determine the percentage of total carbohydrates. The gross energy was determined based on Kim et al. . Accordingly, the fat, carbohydrate and protein contents were calculated by conversion factors.
The extraction of the coffee samples for antioxidant determination was based on Abdeltaif . Accordingly, 5 g of coffee was weighed into a flask, and 50 ml of methanol was added. The sample was shaken for 24 h and then filtered through fat-free filter paper (Whatman filter paper). Again, 50 ml of methanol was added, and the mixture was shaken for 2 h and then filtered. The filtrate was transferred into a round bottom flask and then concentrated to a rotary evaporator (RE-2000A, Germany RE 501 5L Rotovap, N/A, Canada, Oshawa). The sample was reconstituted in methanol and stored in a refrigerator at 4°C until analysis.
The total phenol content of roasted coffee was analyzed using a UV-vis spectrometer (Lambda 950Uv-Vis/NIR spectrophotometer, EUROIMMUN a PerkinElmer company, Germany, Lübeck) according to Singleton and Rossi . Briefly, 1 ml of Folin-Ciocalteu phenol reagent was added to the solution. After 5 minutes, 10 ml of 7% Na2CO3 was added, and the total volume was brought to 25 ml by deionized water. Then, the sample was incubated for 90 minutes, and the absorbance was measured at 765 nm. The TPC is expressed as mg of GAE/g of extract and was determined using the following equation. Model 1 where C: gallic acid equivalent concentration obtained from the calibration curve (mg/ml); V: volume of stock solution of extract (ml); M: dry weight of extract in the stock solution (g); TPC: total phenol content (mg of GAE/g dry extract); and D: dilution factor.
The flavonoid content in the roasted coffee was analyzed using a UV-vis spectrometer (Lambda 950Uv-Vis/NIR spectrophotometer, EUROIMMUN a PerkinElmer company, Germany, Lübeck) . The samples and quercetin standards were at different concentrations. After 5 minutes, methanol and 1 ml of 10% aluminum chloride were added. Then, the absorbance was measured at 415 nm, and the results are expressed as quercetin equivalents (mg quercetin/g dried extract) and were calculated using the following equation: Model 2 where C: quercetin equivalent concentration obtained from the calibration curve (mg/ml); V: volume of stock solution of the extract (ml); M: dry weight of extract in the stock solution (g); FC: flavonoid content expressed as (mg of CE/g dry extract); and D: dilution factor.
Radical scavenging activity.
The radical scavenging activity (DPPH) of the roasted coffee was analyzed using a UV-vis spectrometer (Lambda 950Uv-Vis/NIR spectrophotometer, EUROIMMUN a PerkinElmer company, Germany, Lübeck) by Abdeltaif et al. . The sample solution and ascorbic acid standards were prepared at different concentrations, and methanol was added. Then, 4 ml of DPPH was added, and the mixtures were placed in the dark for 90 minutes. The absorbance was measured at 517 nm, and the inhibition of DPPH free radical was calculated using the following equation: Model 3 where Ac is the absorbance of the control reaction and As is the absorbance of the test compound.
Ferric reducing power.
The FRAP (ferric reducing antioxidant power) of the roasted coffee was analyzed by using a UV-vis spectrometer (Lambda 950Uv/NIR spectrophotometer, EUROIMMUN a PerkinElmer company, Germany, Lübeck) according to Benzie et al. , with some modifications. Fresh FRAP reagent (2.5 ml of TPTZ (10 mM in 40 mM HCl), 25 ml of acetate buffer and 2.5 ml of FeCl36H2O) was prepared and incubated at 37°C for 15 minutes. The test samples and ascorbic acid standards were taken at different concentrations, and methanol was added. Then, 3 ml of the freshly prepared FRAP reagent was added, and the mixtures were incubated for 30 minutes. The absorbance was measured at 593 nm.
EC50 (Effective concentration).
EC50 (Effective concentration), the amount of sample necessary to decrease the absorbance of DPPH by 50%, computed based on the method stated by Alexander et al. . EC50 was calculated by considering the dose-response curves obtained by plotting the percentage of inhibition versus concentration.
All tests were performed in duplicate. Response surface design (RSD) was used to optimize the roasting time and temperature, and the data were analyzed using SPSS version 20. One-way ANOVA was used to determine the effect of roasting on the acrylamide content, the nutritional composition and the antioxidant properties of roasted coffee. Multiple comparison tests using the least significant difference (LSD, p = 0.05) technique were applied to compare the means of each parameter between different coffee samples.
Results and discussion
Acrylamide content of powdered coffee
The acrylamide contents in powdered coffee from street vendors and industrially processed coffee ranged from 346±19 to 906±0 μg/kg (Table 1). A significant difference (p<0.05) was observed among the acrylamide contents of powdered and brewed coffee among street vendors and industrially processed coffee due to the roasting and brewing conditions. This is in line with the report of Pastoriza et al.  that during dark roasting, acrylamide reacts with melanoides to form the color and flavor of coffee, which causes the content of acrylamide to decrease.
Bagdonaite et al.  reported that the coffee species had an impact on the level of acrylamide formed during roasting, and the acrylamide contents in 12 roasted coffees ranged from 299 to 762 μg/kg, which was in agreement with this finding. The findings of this study for street coffee samples (from sub-city A, B, C, D) and industrially processed coffee samples (from sub-city A, B, C, D) (Table 1) was in agreement with the acrylamide content reported by Surma et al.  (17±1 to 776±8 μg/kg). Bertuzzi et al.  reported that the mean acrylamide level of 66 coffee samples collected from Italy was 465 μg/kg. The acrylamide contents of 291 coffee samples from Europe ranged from 79 μg/kg to 1188 μg/kg . The acrylamide content found in this study agreed with the acrylamide contents of Italian and European processed coffees. The maximum limit of acrylamide for roasted coffee in the EU is 400 μg/kg . The mean acrylamide content in the coffee samples in this study (SB, SC, SD, PA, PB, PC and PD) (Table 1) was higher than the maximum limit set by EU ; however, the content in SA was below the limit. Lachenmeier et al.  reported that the acrylamide content in pure roasted coffee reduced over the years. None of the samples exceeded the EU benchmark. Samples, which have considerably above the EU benchmark, could be a mix with cereals . Hence, further study needs to conduct on the purity of the roasted/brewed samples collected from Addis Ababa.
Acrylamide content of brewed coffee
The acrylamide contents in brewed coffee samples from street vendors and industrial processing were 25 to 89μg/L (Table 1). A significant difference (p<0.05) was observed between the acrylamide contents of brewed coffee from street vendors and industrial processed coffee due to the coffee preparation and roasting conditions. This finding was in agreement with the report of Sirot et al.  (37 μg/L), Alves et al.  (1.7 to 75 μg/L) and Mesías & Morales  (7.7 to 40 μg/L).
Effect of roasting temperature and time on the acrylamide content of Sidama Coffee
The acrylamide content (mean±sd) of optimally roasted Sidama coffee is given in Table 2. The acrylamide contents of optimally roasted Sidama coffee were 202 and 2549μg/kg for very light and medium dark roasted coffee, respectively. There was a significant difference (p<0.05) between the roasted coffee samples due to the roasting conditions. This was similar to EFSA, Contam  reported that the roasting temperature and time were the major factors affecting the formation of acrylamide in coffee. The report of Claeys et al.  (170 to 2522 μg/kg) was in agreement with this finding. However, the report of Wenzl & Anklam  on European coffee (79 μg/kg to 1188 μg/kg) reported a higher acrylamide content for Sidama coffee. This could be due to the coffee species, roasting conditions and duration of storage. According to Bagdonaite et al. , the coffee roasting conditions affect the formation of acrylamide. The r2 value and adjusted coefficient of acrylamide for optimally roasted coffee were 0.55 and 0.24, respectively. According to Zhang et al. , for good fit, the r2 should be at least 0.80. The r2 for these response variables was less than 0.80. This indicated that the estimated quadratic model was not good. This was related to the unstable formation of acrylamide. According to Madihah et al. , the content of acrylamide formed was not stable. It forms at high temperatures and is also destroyed at high temperatures (oC).
The chromatogram of acrylamide under different roasting conditions are presented in Fig 1. This study revealed that acrylamide content increases with increasing roasting temperature until it reaches a maximum after which it rapidly decreases. In this study, medium dark roasted coffees contained more acrylamide than light and dark roasted coffees. This was in line with Lantz et al. , who found that the acrylamide content increased during the middle phase of roasting and then rapidly decreased.
a) light roasted coffee; b) medium roasted coffee; c) medium dark roasted coffee; d) dark roasted coffee; e) standard.
The acrylamide contents in very light, light, light-medium, medium-light, and medium roasted coffees were below the limit set by the European Commission (400 μg/kg) . However, the contents in medium-dark, dark-medium, dark, and very dark roasted coffees were higher than 400 μg/kg. The presence of acrylamide, in association with the color type of the roasted coffee (light to dark), appears in contrast to the report of Lachenmeier et al. , though this study equally reported that acrylamide content declined, with an increase at a higher extreme temperature. For the reason, that degree of roasting certainly yield different categories of color (very light to very dark), and the variation could be the capacity of coffee beans to possess different color levels during roasting. At a similar temperature and time of roasting, coffee color varies from region to region, and roasting does not yield consistent results. This is mostly governed by size, weight, and moisture content of coffee beans , and coffee bean variety, reducing sugars, sucrose level, available free amino acids, and pH . As a result, color alone could not be a reliable assessment for roasting level , which demands further study. The optimal roasting time and temperature for Sidama coffee were 190°C for 6 minutes, as this resulted in a low acrylamide (336 μg/kg) content (Fig 2A). In relative term, this was in agreement with the report of Chung  (182 oC and 7 minutes).
Nutrient composition of Sidama coffee
The moisture contents of Sidama coffee at different roasting temperatures and times ranged from 0.45 to 3.59% (Table 3). The moisture content in very light roasted coffee (3.59±0) was higher than that in medium (2.23±0) and dark roasted coffee (0.91±0). Accordingly, there was a significant difference (p<0.05) in moisture contents between treatments due to the roasting conditions. The moisture contents of medium, medium-dark (1.43±0.) and dark-medium (1.21±0) roasted coffee were in agreement with the reports of Franca et al.  on Brazilian roasted coffee (1.5%). On the other hand, very dark roasted Sidama coffee (0.45%) had a lower water content than that of Brazilian dark roasted coffee . The moisture contents of very light, light and light-medium roasted coffee agreed with the reports of Tawfik & El Bader .
The fat contents of roasted Sidama coffee ranged from 11.04 to 17.28% (Table 3). The fat content in dark roasted coffee (16.29±0%) was higher than that in medium (14.47±0%) and light roasted coffees (12.93±0%). Accordingly, the fat contents in the samples varied significantly (p<0.05) due to the roasting conditions. This was in agreement with the findings of Liu et al.  that the fat content of coffee beans increases with roasting temperature due to the degradation of carbohydrates and the evaporation of volatile chemicals. In addition, the oil also released to the outer surface of the bean during roasting. Tawfik & El Bader  reported that the mean fat contents of roasted Harar and Berry coffee were 11.83 and 12.30%, respectively. The fat content of roasted Sidama coffee was higher than that of Harar and Berry coffee beans. According to Franca et al. , healthy mature coffee beans have a higher oil content than defective beans.
Accordingly, Sidama coffee can be classified as a good quality coffee. The optimal temperature and time for roasting to achieve better fat formation were 190°C at 6 minutes, which resulted in 14.42% fat (Fig 2B).
The protein contents of the roasted Sidama coffee samples ranged from 13.12 to 14.18% (Table 3). The protein content of very dark roasted Sidama coffee was significantly different (p<0.05) from that of light-medium, light and very light roasted coffee. There were no significant differences (p>0.05) among the protein contents of very light, light, light-medium, medium-light, medium, medium-dark, dark-medium and dark roasted coffee. The protein contents reported by Belitz et al.  ranged from 9–12%, which are lower than the protein contents of roasted Sidama coffee. Farah et al.  reported that the protein contents of green and roasted coffee ranged from 9–16% and 11–16.5%, respectively. The protein content in roasted Sidama coffee was similar to the result reported by Farah . The optimal temperature and time for achieving the highest protein content were 190°C and 6 minutes, respectively, which afforded a protein content of 13.69% (Fig 2D).
The ash contents of roasted Sidama coffee samples range from 3.32 to 4.38% (Table 3). The ash content of very light roasted coffee was significantly different (p<0.05) from those of dark and very dark roasted coffees. According to Bicho , the total mineral content in coffee beans is approximately 4%. The ash content of roasted Sidama coffee is consistent with that report. However, the ash content of roasted Sidama coffee was lower than the ash content reported by Oliveira et al.  (4.8% to 5.8%).
The fiber contents in roasted Sidama coffee ranged from 13.13 to 33.1% (Table 3). The fiber content of dark roasted coffee (15.09±0%) was lower than that of medium (23.95±0%) and light roasted coffee (30.12±2%), and a significant variation (p<0.05) was found between very light roasted coffee and light, light-medium, medium-light, medium, medium-dark, dark-medium, dark and very dark roasted coffees. The fiber contents in coffee reported by Ismail et al.  were 15.45 to 21.51%, which is in agreement with the content found for roasted Sidama coffee. This value was also similar to that reported by Martinez-Saez et al.  (19% to 23%). The optimal temperature and time for achieving a high fiber content were 190°C and 6 minutes, respectively, which afforded a fiber content of 22.60%.
The carbohydrate contents of the roasted Sidama coffee samples were 34.77 to 52.64% (Table 3). The carbohydrate content in dark roasted coffee (48.97±0%) was higher than that of medium (41.85±0%) and light roasted coffee (36.28±1%). Accordingly, light-medium and very dark roasted coffee had significantly different contents (p<0.05) from those of very light, light, medium-light, medium, medium-dark, dark-medium and dark roasted coffee due to the roasting conditions. Nogaim et al.  reported the carbohydrate contents of 70 coffee samples from Yemen (7.92 ± 0 to 35.64 ± 2), and their results were in agreement with the carbohydrate contents of Sidama coffee. However, the carbohydrate contents of the roasted Sidama coffees were lower than those reported by Franca et al.  (61.49 to 62.41%). The optimal roasting temperature and time to achieve the highest carbohydrate content was 190°C for 6 minutes, which afforded a carbohydrate content of 42.92%.
The total energy values of the roasted Sidama coffee samples were 295.12 to 409.63 kcal (Table 3). The roasting temperature and time were found to influence the total energy of the coffee. The total energy of the dark roasted coffee (405.58±0 kcal) was higher than that of the medium (353.13±5 kcal) and light roasted coffees (318.14±13 kcal). Accordingly, very light roasted coffee significantly differed (p<0.05) from the light, light-medium, medium-light, medium, medium-dark, dark-medium, dark and very dark roasted coffees in total energy content. The optimal temperature and time for achieving a high total energy content were 190°C and 6 minutes, respectively, which afforded a total energy content of 344.6 kcal.
Antioxidant properties of Sidama coffee
Total phenolic compounds.
The effects of roasting temperature and time on the antioxidant properties of roasted coffee are presented in Table 4. The total phenol contents in roasted Sidama coffee ranged from 26 to 63 mg GAE/g (Table 4). Light-medium roasted coffee (63±5 mg GAE/g) contained more phenolic compounds than light (45±4 mg GAE/g) and dark roasted coffees (33±1 mg GAE/g). Accordingly, there were significant differences (p<0.05) among the phenol contents in the roasted Sidama coffees due to roasting time and temperature. The content of phenolic compounds in coffee depend on coffee variety, origin, degree of roasting and method of beverage preparation . In this study, the content of phenolic compounds decreased with increasing temperature. Tamilmani  reported that roasting temperature causes chlorogenic acid degradation, which consequently reduces the polyphenol content in the coffee. Hečimović et al.  reported the phenol contents of cherry and minas coffee varieties (21mg GAE/g to 42mg GAE/g), which were in agreement with this finding. The phenolic contents of Sidama coffee were also in agreement with Perez-Martinez et al.  (37 to 55 mg GAE/g).
Total flavonoid content.
The flavonoid contents of the roasted Sidama coffee samples ranged from 26 to 73 mg QE/g (Table 4). The flavonoid content in light-medium roasted coffee (73±7 mg QE/g) was higher than those of very light (49±0 mg QE/g) and dark roasted coffee (29±5 mg QE/g). Accordingly, there were significant differences (p<0.05) among the flavonoid contents of roasted Sidama coffee due to roasting time and temperature. Very dark, dark, dark-medium and medium dark roasted coffee had significantly different (p<0.05) contents from those of very light, light, light-medium, medium-light, and medium roasted coffee. Additionally, the flavonoid content of light medium-roasted coffee was significantly different (p<0.05) from those of medium and very light roasted coffee. This finding was similar to that of Cho et al. , who reported that the flavonoid content was high in light roasted coffee and decreased with increasing temperature. The flavonoid contents of Sidama coffee were also in agreement with the results reported by Kim et al.  (22 ±0 to 67.19 ± 0 mg QE/g).
Radical scavenging activity.
The percent inhibition for roasted Sidama coffee ranged from 64 to 90 μg/ml (Table 4). A significant difference (p<0.05) was observed among the inhibitory activities of roasted Sidama coffees due to the roasting conditions. The radical scavenging activities of very dark, dark and light-medium roasted coffees were significantly different (p<0.05) from those of the dark-medium, medium-dark, very light, light, medium-light and medium coffees. Odžaković et al.  stated that the antioxidant activity of coffee decreased during roasting, and the higher antioxidant activity was obtained in medium-roasted coffee compared with dark-roasted coffee, which was in agreement with this study.
A graphical representation of % inhibition for Sidama coffees roasted under different conditions is presented in Fig 3. The percent inhibition and roasting temperature are inversely related. The percent inhibition decreased with temperature. Additionally, Fig 2A shows the direct relation between % inhibition and concentration of sample.
Ferric reducing power.
The ferric reducing activities of roasted Sidama coffee ranged from 5 to 7 mg/g (Table 4). The ferric reducing activity of light roasted coffee (7±1 mg/g) was higher than those of medium (6±1 mg/g) and dark roasted coffees (5±1 mg/g). Likewise, the ferric reducing activities vary significantly (p<0.05) due to the roasting conditions. The ferric reducing powers of the very dark and dark roasted coffee were significantly different (p<0.05) from that of light roasted coffee.
The ferric reducing activity of Sidama coffee decreased as the roasting temperature increased. This finding was similar to Cho et al. ; accordingly, the roasting conditions of the coffee had a significant effect on the FRAP value of the coffee. The FRAP value decreased with increasing roasting temperature due to the thermal degradation of phenolic compounds. The FRAP values of the roasted Sidama coffee samples were similar to those reported by Gebeyehu and Bikila  (6 ± 0 to 9 ± 0 mg/g).
The effect of roasting on the ferric reducing activity of roasted Sidama coffee is presented in Fig 4. Accordingly, ferric reducing power had an inverse relationship with the roasting temperature. Light roasted coffee had a higher ferric reducing power than dark roasted coffee. Fig 2B also shows that the ferric reducing power was directly related to the concentration of the sample. The ferric reducing power increases with increasing concentration.
Effective concentration (EC50).
The EC50 (effective concentration) values of the roasted Sidama coffee samples were 71 to 111 μg/ml (Table 4). The EC50 value of light-medium roasted coffee was higher than that of medium, light and dark roasted coffee. The EC50 values of the roasted Sidama coffees vary significantly (p<0.05) due to the roasting conditions. The values of light-medium, medium, medium-dark and very dark roasted Sidama coffees were significantly different (p<0.05) from those of dark, dark-medium, very light, light, and medium-light roasted coffee. This was in line with the report of Delgado-Andrade et al.  that the EC50 value decreased due to the formation of melanoidins. The content of melanoidins increased with increasing temperature, which was responsible for the reduction in the scavenging activity of the coffee.
The acrylamide contents of coffee samples collected from street vendors and industrial processors were higher than the EU standards. The acrylamide contents in the coffee samples from street vendors were lower than those of industrially processed coffee due to the roasting conditions. Based on our optimization of the roasting conditions of Sidama coffee, the highest acrylamide content was achieved in medium-dark roasted coffee (190°C for 7 minutes), and the lowest was in very light roasted coffee (175.86°C for 5.5 minutes). The roasting time and temperature had an effect on the antioxidant content in the coffee. An increase in roasting decreases antioxidant properties. Accordingly, light-medium roasted coffee had a relatively high antioxidant activity. Roasting also had a significant effect on the composition of the coffee. The moisture, protein, and fiber contents of the coffee decreased as the roasting temperature increased. However, the carbohydrate, fat, ash, and total energy contents increased with roasting temperature. The optimal roasting temperature for Sidama coffee was 190°C for 6 minutes, at which point the acrylamide content decreased and the antioxidant content of coffee improved.
We have a special thanks to Addis Ababa Science and Technology University (AASTU), Department of Food Science and Applied Nutrition for laboratory facility. The Ethiopian Coffee Quality Inspection and Certification Center (ECQICC) is also keenly acknowledged. The authors would like to thank Belew Yemane, Wondu Degaga, and Abas Muhdin for their encouragement and costs covered during the review process.
- 1. Madihah KK, Zaibunnisa AH, Norashikin S, Rozita O, Misnawi J. Optimization of roasting conditions for high-quality Arabica coffee. International Food Research Journal. 2013 Oct 1;20(4).
- 2. Chauhan R, Hooda MS, Tanga AA. Coffee: the backbone of Ethiopian economy. International Journal of Economic Plants. 2015;1(2):82–6.
- 3. Gökmen V. A perspective on the evaluation of safety risks in thermal processing of foods with an example for acrylamide formation in biscuits. Quality Assurance and Safety of Crops & Foods. 2014 Sep 1;6(3):319–25.
- 4. Zyzak DV, Sanders RA, Stojanovic M, Tallmadge DH, Eberhart BL, Ewald DK, et al. Acrylamide formation mechanism in heated foods. Journal of agricultural and food chemistry. 2003 Jul 30;51(16):4782–7. pmid:14705913
- 5. Żyżelewicz D, Oracz J, Krysiak W, Budryn G, Nebesny E. Effects of various roasting conditions on acrylamide, acrolein, and polycyclic aromatic hydrocarbons content in cocoa bean and the derived chocolates. Drying technology. 2017 Feb 17;35(3):363–74.
- 6. Weißhaar R. Acrylamide in heated potato products–analytics and formation routes. European Journal of Lipid Science and Technology. 2004 Nov;106(11):786–92.
- 7. Tareke E, Rydberg P, Karlsson P, Eriksson S, Törnqvist M. Analysis of acrylamide, a carcinogen formed in heated foodstuffs. Journal of agricultural and food chemistry. 2002 Aug 14;50(17):4998–5006. pmid:12166997
- 8. Mesías M, Morales FJ. Acrylamide in coffee: Estimation of exposure from vending machines. Journal of Food Composition and Analysis. 2016 May 1;48:8–12.
- 9. Morales FJ, Jiménez-Pérez S. Peroxyl radical scavenging activity of melanoidins in aqueous systems. European Food Research and Technology. 2004 May 1;218(6):515–20.
- 10. Durling LJ, Busk L, Hellman BE. Evaluation of the DNA damaging effect of the heat-induced food toxicant 5-hydroxymethylfurfural (HMF) in various cell lines with different activities of sulfotransferases. Food and Chemical Toxicology. 2009 Apr 1;47(4):880–4. pmid:19271322
- 11. International Agency for Research on Cancer. Some industrial chemicals. IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans. 1994;60.
- 12. Budryn G, Nebesny E, Podsędek A, Żyżelewicz D, Materska M, Jankowski S, et al. Effect of different extraction methods on the recovery of chlorogenic acids, caffeine and Maillard reaction products in coffee beans. European Food Research and Technology. 2009 Apr 1;228(6):913–22.
- 13. European Commission. Commission recommendation of 8 November 2013 on investigations into the levels of acrylamide in food. Official Journal. 2013;50.
- 14. Commission European. Establishing Mitigation Measures and Benchmark Levels for the Reduction of the Presence of Acrylamide in Food. Ref. Ares.2017.2895100-09/06/2017 C.F.R.
- 15. Dibaba K, Tilahun L, Satheesh N, Geremu M. Acrylamide occurrence in Keribo: Ethiopian traditional fermented beverage. Food Control. 2018 Apr 1;86:77–82.
- 16. Oroian M, Amariei S, Gutt G. Acrylamide in Romanian food using HPLC-UV and a health risk assessment. Food Additives & Contaminants: Part B. 2015 Apr 3;8(2):136–41.
- 17. Shamla L, Nisha P. Acrylamide in deep-fried snacks of India. Food Additives & Contaminants: Part B. 2014 Jul 3;7(3):220–5.
- 18. Food and Drug Administration. Guidance for Industry: Validation of Analytical Procedures: Methodology: Final Guidance. Found at http://www.fda.gov/cvm/Guidance/guida64.htm. 1999.
- 19. AOAC. Official methods of analysis. (Vol. ii 17th) of AOAC International Assn of Official Analytical Chemists. Washington, DC, USA, Official Methods 925.09, 923.03, 979.09, 962.09, 4.5.01. 2000.
- 20. Kim E, Choi J, Kim H. Mabolizable energy differences between values calculated using energy conversion factors and actual values determined by metabolic study of Korean starch foods. Journal of food science. 2014 Apr;79(4):H713–8. pmid:24621178
- 21. Abdeltaif SA, SirElkhatim KA, Hassan AB. Estimation of phenolic and flavonoid compounds and antioxidant activity of spent coffee and black tea (processing) waste for potential recovery and reuse in Sudan. Recycling. 2018 Jun;3(2):27.
- 22. Singleton VL, Rossi JA. Colorimetry of total phenolics with phosphomolybdic-phosphotungstic acid reagents. American journal of Enology and Viticulture. 1965 Jan 1;16(3):144–58.
- 23. Kim DO, Jeong SW, Lee CY. Antioxidant capacity of phenolic phytochemicals from various cultivars of plums. Food chemistry. 2003 Jun 1;81(3):321–6.
- 24. Benzie IF, Strain JJ. The ferric reducing ability of plasma (FRAP) as a measure of “antioxidant power”: the FRAP assay. Analytical biochemistry. 1996 Jul 15;239(1):70–6. pmid:8660627
- 25. Alexander B, Browse DJ, Reading SJ, Benjamin IS. A simple and accurate mathematical method for calculation of the EC50. Journal of pharmacological and toxicological methods. 1999 Apr 1;41(2–3):55–8. pmid:10598675
- 26. Pastoriza S, Rufián-Henares JÁ, Morales FJ. Reactivity of acrylamide with coffee melanoidins in model systems. LWT-Food science and technology. 2012 Mar 1;45(2):198–203.
- 27. Bagdonaite K, Derler K, Murkovic M. Determination of acrylamide during roasting of coffee. Journal of agricultural and food chemistry. 2008 Aug 13;56(15):6081–6. pmid:18624446
- 28. Surma M, Sadowska-Rociek A, Cieślik E, Sznajder-Katarzyńska K. Optimization of QuEChERS sample preparation method for acrylamide level determination in coffee and coffee substitutes. Microchemical Journal. 2017 Mar 1;131:98–102.
- 29. Bertuzzi T, Rastelli S, Mulazzi A, Pietri A. Survey on acrylamide in roasted coffee and barley and in potato crisps sold in Italy by a LC–MS/MS method. Food Additives & Contaminants: Part B. 2017 Oct 2;10(4):292–9.
- 30. Wenzl T, Anklam E. European Union database of acrylamide levels in food: update and critical review of data collection. Food additives and contaminants. 2007 Jul 1;24(sup1):5–12.
- 31. Lachenmeier DW, Schwarz S, Teipel J, Hegmanns M, Kuballa T, Walch SG, et al. Potential antagonistic effects of acrylamide mitigation during coffee roasting on furfuryl alcohol, furan and 5-hydroxymethylfurfural. Toxics. 2019 Mar;7(1):1.
- 32. Sirot V, Hommet F, Tard A, Leblanc JC. Dietary acrylamide exposure of the French population: results of the second French Total Diet Study. Food and chemical toxicology. 2012 Mar 1;50(3–4):889–94. pmid:22240413
- 33. Alves RC, Soares C, Casal S, Fernandes JO, Oliveira MB. Acrylamide in espresso coffee: Influence of species, roast degree and brew length. Food Chemistry. 2010 Apr 1;119(3):929–34.
- 34. EFSA Panel on Contaminants in the Food Chain (CONTAM). Scientific Opinion on the risks to human and animal health related to the presence of beauvericin and enniatins in food and feed. EFSA Journal. 2014 Aug;12(8):3802.
- 35. Claeys W, De Meulenaer B, Huyghebaert A, Scippo ML, Hoet P, Matthys C. Reassessment of the acrylamide risk: Belgium as a case-study. Food Control. 2016 Jan 1;59:628–35.
- 36. Zhang Y, Kahl DH, Bizimungu B, Lu ZX. Effects of blanching treatments on acrylamide, asparagine, reducing sugars and colour in potato chips. Journal of food science and technology. 2018 Oct 1;55(10):4028–41. pmid:30228401
- 37. Lantz I, Ternité R, Wilkens J, Hoenicke K, Guenther H, van der Stegen GH. Studies on acrylamide levels in roasting, storage and brewing of coffee. Molecular nutrition & food research. 2006 Nov;50(11):1039–46.
- 38. Winjaya F, Rivai M, Purwanto D. Identification of cracking sound during coffee roasting using neural network. In2017 International Seminar on Intelligent Technology and Its Applications (ISITIA) 2017 Aug 28 (pp. 271–274). IEEE.
- 39. Hamzalıoğlu A, Gökmen V. 5-Hydroxymethylfurfural accumulation plays a critical role on acrylamide formation in coffee during roasting as confirmed by multiresponse kinetic modelling. Food Chemistry. 2020 Jul 15;318:126467. pmid:32145542
- 40. Franca AS, Oliveira LS, Oliveira RC, Agresti PC, Augusti R. A preliminary evaluation of the effect of processing temperature on coffee roasting degree assessment. Journal of Food Engineering. 2009 Jun 1;92(3):345–52.
- 41. Chung HS, Kim DH, Youn KS, Lee JB, Moon KD. Optimization of roasting conditions according to antioxidant activity and sensory quality of coffee brews. Food Science and Biotechnology. 2013 Feb 1;22(1):23–9.
- 42. Franca AS, Oliveira LS, Mendonça JC, Silva XA. Physical and chemical attributes of defective crude and roasted coffee beans. Food Chemistry. 2005 Mar 1;90(1–2):89–94.
- 43. Tawfik MS, El Bader NA. Chemical characterization of Harar and Berry coffee beans with special reference to roasting effect. J. Food Technol. 2005;3(4):601–4.
- 44. Liu Y, Kitts DD. Confirmation that the Maillard reaction is the principle contributor to the antioxidant capacity of coffee brews. Food Research International. 2011 Oct 1;44(8):2418–24.
- 45. Belitz HD, Grosch W, Schieberl P. Food chemistry (4th ed.). Heidelberg: Springer (Chapter 21). 2009.
- 46. Farah A, Monteiro MC, Calado V, Franca AS, Trugo LC. Correlation between cup quality and chemical attributes of Brazilian coffee. Food Chemistry. 2006 Jan 1;98(2):373–80.
- 47. Bicho NC, Leitão AE, Ramalho JC, Lidon FC. Use of colour parameters for roasted coffee assessment. Food Science and Technology. 2012 Sep;32(3):436–42.
- 48. Oliveira LS, Franca AS, Mendonça JC, Barros-Junior MC. Proximate composition and fatty acids profile of green and roasted defective coffee beans. LWT-Food Science and Technology. 2006 Apr 1;39(3):235–9.
- 49. Ismail I, Anuar MS, Shamsudin R. Effect on the physico-chemical properties of liberica green coffee beans under ambient storage. International Food Research Journal. 2013;20(1):255.
- 50. Martinez-Saez N, Ullate M, Martin-Cabrejas MA, Martorell P, Genovés S, Ramon D, et al. A novel antioxidant beverage for body weight control based on coffee silverskin. Food chemistry. 2014 May 1;150:227–34. pmid:24360444
- 51. Nogaim QA, Al-Duais M, Al-Warafi A, Al-Erianee H, Al-Sayadi M. The chemical composition of Yemeni green coffee. Journal of Food Chemistry and Nutrition. 2013;1(2):42–8.
- 52. Farah A, Donangelo CM. Phenolic compounds in coffee. Brazilian journal of plant physiology. 2006 Mar;18(1):23–36.
- 53. Tamilmani P. Optimization and Evaluation of phenolic compounds and their antioxidant activity from coffee beans. International Journal of Advanced Research. International Journal of Advanced Research. 2015 April 3(4):296–306.
- 54. Hečimović I, Belščak-Cvitanović A, Horžić D, Komes D. Comparative study of polyphenols and caffeine in different coffee varieties affected by the degree of roasting. Food chemistry. 2011 Dec 1;129(3):991–1000. pmid:25212328
- 55. Perez-Martinez M, Caemmerer B, De Peña MP, Cid C, Kroh LW. Influence of brewing method and acidity regulators on the antioxidant capacity of coffee brews. Journal of agricultural and food chemistry. 2010 Mar 10;58(5):2958–65. pmid:20141144
- 56. Cho AR, Park KW, Kim KM, Kim SY, Han J. Influence of Roasting Conditions on the Antioxidant Characteristics of C olombian Coffee (C offea arabica L.) Beans. Journal of food biochemistry. 2014 Jun;38(3):271–80.
- 57. Odžaković B, Džinić N, Kukrić Z, Grujić S. Effect of roasting degree on the antioxidant activity of different Arabica coffee quality classes. Acta Scientiarum Polonorum Technologia Alimentaria. 2016 Dec 30;15(4):409–17. pmid:28071018
- 58. Gebeyehu BT, Bikila SL. Determination of caffeine content and antioxidant activity of coffee. Am J Appl Chem. 2015 Mar 31;3(2):69–76.
- 59. Delgado-Andrade C, Rufián-Henares JA, Morales FJ. Assessing the antioxidant activity of melanoidins from coffee brews by different antioxidant methods. Journal of Agricultural and Food Chemistry. 2005 Oct 5;53(20):7832–6. pmid:16190638