Extraction of espresso coffee by using gradient of temperature. Effect on physicochemical and sensorial characteristics of espresso

Food Chemistry 214 (2017) 622–630

Contents lists available at ScienceDirect

Food Chemistry journal homepage: www.elsevier.com/locate/foodchem

Extraction of espresso coffee by using gradient of temperature. Effect on physicochemical and sensorial characteristics of espresso C. Alejandra Salamanca a, Núria Fiol a,⇑, Carlos González b, Marc Saez c, Isabel Villaescusa a a

Chemical Engineering Department, Escola Politècnica Superior, Universitat de Girona, Mª Aurèlia Capmany, 61, 17071 Girona, Spain Rancilio España S.A., Gran Via Carles III, 83, Barcelona, Spain c Department Economics, Facultat de Ciències Econòmiques I Empresarials, Universitat de Girona, Campus Montilivi, 17071 Girona, Spain b

a r t i c l e

i n f o

Article history: Received 19 November 2015 Received in revised form 13 July 2016 Accepted 20 July 2016 Available online 21 July 2016 Keywords: Espresso brewing Gradient of temperature Physicochemical properties Sensorial evaluation Robusta Arabica natural Washed Arabica

a b s t r a c t Espresso extraction is generally carried out at a fixed temperature within the range 85–95 °C. In this work the extraction of the espressos was made in a new generation coffee machine that enables temperature profiling of the brewing water. The effect of using gradient of temperature to brew espressos on physicochemical and sensorial characteristics of the beverage has been investigated. Three different extraction temperature profiles were tested: updrawn gradient (88–93 °C), downdrawn gradient (93–88 °C) and fixed temperature (90 °C). The coffee species investigated were Robusta, Arabica natural and Washed Arabica. Results proved that the use of gradient temperature for brewing espressos allows increasing or decreasing the extraction of some chemical compounds from coffee grounds. Moreover an appropriate gradient of temperature can highlight or hide some sensorial attributes. In conclusion, the possibility of programming gradient of temperature in the coffee machines recently introduced in the market opens new expectations in the field of espresso brewing. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction The quality of brew coffee depends on multiple factors such as coffee species (Campa, Doulbeau, Dussert, Hamon, & Noirot, 2005; Maeztu et al., 2001) coffee blends (Fujioka & Shibamoto, 2008); bean roasting conditions (Blumberg, Frank, & Hofmann, 2010; Cavaco Bicho, Leitao, Cochicho Ramalho, de Alvarenga, & Cebola Lidon, 2011; Ludwig, Bravo, de Peña, & Cid, 2013; Nunes, Coimbra, Duarte, & Delgadillo, 1997), grinding of the roasted coffee beans (Andueza, de Peña, & Cid, 2003a) and in a great extension the brewing method i.e. drip, espresso (Gloess et al., 2013). One of the most popular presentations of coffee brews in South-Europe is the Italian espresso. Italian espresso coffee has been defined as a beverage prepared on request by extraction of ground roasted coffee beans (6.5 ± 1.5 g), with hot water (90 ± 5 °C) under pressure (9 ± 2 bar) for a defined short time (30 ± 5 s) (Illy & Viani, 1995). In the espresso brewing process, key operation variables of an espresso coffee are: grinding, ground coffee portion, tamping, water quality (Navarini & Rivetti, 2010), and also some parameters controlled by the coffeemaker such as extraction temperature, ⇑ Corresponding author. E-mail addresses: [email protected] (C.A. Salamanca), nuria.fiol@ udg.edu (N. Fiol), [email protected] (C. González), [email protected] (M. Saez). http://dx.doi.org/10.1016/j.foodchem.2016.07.120 0308-8146/Ó 2016 Elsevier Ltd. All rights reserved.

water pressure and percolation time that clearly influence the final quality of the coffee brew. During the last decade, there has been growing interest in studying the influence of some of these parameters on the quality of espresso coffee. The comparison of espressos obtained by different common coffee makers such as capsules or automatic and semi-automatic coffeemakers was studied by Gloess et al. (2013) in terms of chemical analytical methods and sensory analysis. Andueza et al. (2002) studied the influence of water pressure on the quality of arabica espresso coffee. The same authors investigated the influence of extraction temperature in the quality of espressos obtained from different coffee varieties (Andueza et al., 2003b). Results of this study evidenced differences in terms of body, flavor characteristics and overall acceptability of espresso coffees due to extraction temperature, and also differences as a function of the coffee varieties. Thus, the selection of the optimal extraction temperature in terms of overall acceptability should consider aspects related to physico-chemical characteristics, and odour and flavor components. In some cases, optimal extraction temperature is selected when positive and negative aspects are balanced or when positive ones are predominant in the cup. In this sense, the extraction of compounds that increase components related to positive aspects while avoiding the extraction of components related to decrease the coffee quality is a challenge in the coffee industry research. In this context, the initial hypothesis of our research was that by

C.A. Salamanca et al. / Food Chemistry 214 (2017) 622–630

controlling the espresso dispensing temperature the extraction of the different components of coffee and resulting flavor would be satisfactorily modulated in the espresso cup so as to obtain an espresso of good quality. Recently, Rancilio Group (www.ranciliogroup.com), a reputed espresso machines manufacturer has implemented in one model of their coffee machines a revolutionary technology for the temperature profiling of the brewing water for espresso coffee. In this study, the influence of using extraction temperature ramps on physico-chemical properties and sensorial attributes of espressos obtained from three different coffee varieties has been investigated.

2. Materials and methods 2.1. Coffee samples Samples of ground roasted coffee, Robusta natural (pure Coffea canephora) (R), Arabica natural (pure Coffea Arabica) (A) and washed Arabica (WA) were provided by a local company. The differences between natural and washed coffee are related to the methods used to extract the seed from the coffee cherry. Natural coffee is obtained by drying the cherries in the sun and washed coffee by immersing the cherries in water tanks where fermentation takes place (Odello & Odello, 2002). The beans of the selected types of coffee were roasted by a coffee roasting company at a roast point 7 over 10 (dark medium). The beans were ground just before the espresso preparation by means of a coffee grinder (MACAP MXDL, Italy) that allows regulation of ground coffee size. In order to obtain espressos in an appropriated volume (25 ± 2.5 mL) and in a fixed percolation time (25 ± 5 s) (Odello & Odello, 2002) it was necessary to find the appropriate grade of grinding for each coffee type and ramp of extraction temperature. The search of the appropriate grinding point was made by trial and error that is, by grinding the beans using different positions of the grinder blades and measuring the volume and the extraction time of the obtained espressos by following the procedure specified in Section 2.4. As it was reported that particle size distribution of ground coffee beans can affect aroma and flavor of espressos (Andueza et al., 2003a), particle size distribution of all the selected grinding grades was determined For that purpose 100 g of each ground coffee sample was grinded and sieved by using a digital sieve shaker (CISA BA200N, Spain). Particle size fractions were separated by using 8 sieves (1000, 630, 500, 400, 250, 150, 100 and 50 lm). Coffee fraction of each sieve was weighed and expressed as percentage. Particle distribution was determined in triplicate and the average is presented.

2.2. The coffee machine Espressos were prepared by using a XCELSIUS Classe 9 Rancilio coffeemaker (Rancilio Group, Italy). The XCelsius system enables the temperature of the brew water to be set dynamically, with an increase or decrease of up to 5 °C during the 25–30 s it takes for each individual delivery. The coffee machine used in the present study possesses three extractors. The independent nature of the extractor group heads allows programming upward and descent gradients of temperature within the range (85–95 °C) for each extractor independently. The temperatures of the three extractors were programmed according to three specific thermal profiles: increasing temperature profile from 88 to 93 °C (ramp-up), decreasing temperature profile from 93 to 88 °C (ramp-down) and fixed temperature at 90 °C. The relative water pressure was fixed at 9 atm.

623

2.3. Water for coffee extraction For coffee extraction commercial bottled water was used. The water quality analysis shown on the bottle label reports the following composition: bicarbonate (113 mg L 1), calcium (27.7 mg L 1), chloride (5.7 mg L 1), fluoride (0.9 mg L 1), magnesium 1 1 (4.5 mg L ), potassium (4.9 mg L ), dry residue (139 mg L 1), sodium (11.9 mg L 1) and sulphate (11.2 mg L 1). Taking into account calcium and magnesium concentration, the mineral water used in this study can be classified as soft water. 2.4. Espressos extraction Two espressos were prepared in each extractor from 15 ± 0.5 g of coffee powder (Istituto Nazionale Espresso Italiano (www. espressoitaliano.org) compacted by using a dynamometer coffee press (MACAP NS 1106/0791). The volume of the two espressos obtained in each extractor was of 25 ± 2.5 mL and the percolation time was 25 ± 5 s to avoid over-extraction of substances with poor flavours (Illy & Viani, 1995). An experimental design of two factors, type of coffee (R, A and WA) and extraction temperature profile (88–93 °C, 90 °C and 93– 88 °C) was considered. A total of 18 samples of espressos obtained in the same conditions (type of coffee and temperature profile) were analyzed. 2.5. Physicochemical characterisation of espressos 2.5.1. Foam index and persistence of foam Foam volume and coffee volume of the espressos were measured immediately after the extraction by using a graduated borosilicate glass cylinder. Foam index was calculated as the ratio of foam and liquid volumes measured after the extraction. Foam persistence was measured as the time that foam persists while the espresso is cooling down at room temperature. Persistence time was measured when dark round spot appeared under the foam layer (Andueza et al., 2003a). 2.5.2. Total volume, density, pH and viscosity Espressos were cooled to room temperature (20 ± 2 °C) before measuring total coffee brew volume in a 100 mL graduated cylinder; pH in a pH meter (Crison micro pH 2000), density in a 20 mL glass pycnometer and viscosity with an Ostwald viscometer (Proton 100) (Andueza et al., 2003a). Turbidity was measured in a turbidity meter (Hanna LP2000). For turbidity measurements, 0.4 mL of espresso was diluted to 100 mL in a volumetric flask. 2.5.3. Total solids and filtered solids Total solids were determined by drying 10 mL of espresso coffee in an oven at 100 °C for 24 h. Filtered solids were determined by weighting the dried solids retained in a glass microfiber filter 1.2 lm (Filter-LabÒ) after filtering 10 mL of espresso. Drying was performed in an oven at 100 °C until constant weight. 2.5.4. Lipid content Total lipids content was determined by liquid-liquid extraction with hexane (n-Hexane 99%, reagent grade Scharlau) following the methodology proposed by Parenti et al., 2014. 50 mL of espresso was extracted by adding 20 mL of hexane three times in a separating funnel. The organic fraction was washed with 60 mL of distilled water three times. Solid sodium sulphate anhydrous powder (Scharlau) was added to remove water from the hexane extract and then the solid was filtered by a paper filter. The total lipid content was calculated by weight after solvent evaporation.

624

C.A. Salamanca et al. / Food Chemistry 214 (2017) 622–630

2.5.5. Caffeine Caffeine in espresso coffee was determined by HPLC analysis following the methodology proposed by Fox, Wu, Yiran, & Force, 2013. with an analytical HPLC-UV unit (Agilent 1200 Series) equipped with a 20 lL loop injector and UV detector at 270 nm. A Nucleosil 100 C18 column (5 mm particle size, 100  4.6 mm) was used. The mobile phase was methanol:water (40:60) at a constant flow rate (1 mL min 1). Coffee samples were diluted and filtered through a 0.2 lm nylon syringe filter before injection. The analytical standard of caffeine was purchased from Sigma-Aldrich, and methanol, 250 gradient quality super purity solvent from Romil. 2.5.6. 5-Chlorogenic acid 5-Chlorogenic acid (5-O-caffeoylquinic acid (5-CQA)) was analyzed following the method developed by ChromaDex, Application TN-1134 (Aqeel, Truong, Preston, & Lazzaro, 2012), by HPLC-UV analysis in an Eclipse XDB C-18 column (5 lm particle size, 250  4.6 mm) and detection at 325 nm at room temperature. The mobile phase was acetonitrile and water, both eluents containing 1% formic acid. The mobile phase composition was kept constant at 5% acetonitrile for 15 min, followed by a linear change up to 100% acetonitrile The eluent flow rate was 0.8 mL/min. The analytical standard of 5-O-caffeoylquinic acid (5-CQA) was purchased from Sigma-Aldrich, acetonitrile UV-HPLC grade 240 nm/far and formic acid 85% extra pure from Scharlau. Before HPLC analysis, proteins and other high molecular weight compounds were eliminated. 0.2 mL of each Carrez’s solutions, (K4[Fe(CN)6]
3H2O) (Scharlau Carrez’s Reagent I) and (ZnSO4
7H2O) (Scharlau Carrez’s Reagent II) and 1.6 mL of ethanol were added to 8 mL of espresso sample in a centrifuge tub. The mixture was shaken for 5 s, let stand for 10 min and centrifuged for 10 min at 5000 rpm (Hettich Universal 320 centrifuge) (Trugo & Macrae, 1984). The supernatants were collected, filtered through a 0.2 lm nylon syringe filter and analyzed by HPLC. The injection volume was 20 lL. 2.5.7. Total polyphenol content The total polyphenol content in the espresso samples was determined by spectrophotometry in a Sequential Injection Analysis (SIA) equipment using gallic acid as standard, according to the Folin-Ciocalteu assay (Singleton & Rossi, 1965). The calibration curve was obtained by preparing different standard concentrations of gallic acid within the range 0.1–0.6 mg L 1. Briefly, a 100 lL aliquot of extracts, gallic acid standard solutions (0.1–0.6 mg L 1) and a blank (deionized water) were put in different tubes. Then, 4 mL of the Folin-Ciocalteu’s phenol reagent diluted 1:10 were added to each tube, the tubes were shaken and allowed to react for 5 min. After this time, 4 mL of 7.5% Na2CO3 solution was added. After incubation of the mixture in a thermostatic bath for 15 min at 45 °C, the absorbance against a blank was determined spectrophotometrically at 765 nm (Hitachi U-2000 VIS/UV spectrophotometer). Total phenolic content was expressed as milligrams of gallic acid equivalents (GAE) per litre of espresso. 2.6. Sensory evaluation Sensory evaluation was performed by a trained sensory panel of eleven panelists recruited from members of Forum Cultural del

Café (http://www.forumdelcafe.com/cursos-forum.php) and coffee roasting companies. Espressos were prepared immediately before analysis and served in white, round, 50 mL cups preheated at 31 °C (Cuadras, 2009). Each espresso was tasted in duplicate at 22 °C in an air-conditioned room with separated booths in two independent sessions. Deionized distilled water, and unsalted crackers were served between samples tasting to cleanse the palette. The panelists were given a scorecard (evaluation form) and asked to score the ten attributes described in Table 1 on a scale from 0 (non detectable) to 10 (strong optic, smell or taste impression). Reference range of scores for each attribute which define an espresso of proved acceptability and quality by coffee quality experts from International Institute of Coffee Tasters (Odello & Odello, 2002) were included in the scorecard evaluation form. 2.7. Statistical analysis The one-way analysis of variance (ANOVA) was used to determine whether there are any significant differences between the means of physicochemical parameters determined at different extraction profile and each type of coffee. The null hypothesis that all means are equal was rejected when difference between means was <0.05. Following the ANOVA test multiple comparisons were made at those statistically significant variables using the test HSD (honestly-significant-difference) Tukey with a significance level of 10%. Concerning sensorial analysis data the second quartile or median, interquartile range and quartile deviation was used as a robust descriptive statistics. This tool provided characteristic information of each set of data at different extraction profile and each type of coffee that could be related to the range of acceptability established by specialists of the International Institute of Coffee Tasters. It was not possible to apply a principal components analysis because when data were discretized by temperature profile and type of coffee the number of individuals was very small respective to the number of variables. The application of this type of analysis in these conditions would imply a matrix of incomplete range of data that would lead to unreliable results. This is the reason why median was used as a robust measurement of central tendency. Percentiles deciles and quartiles were also used to look for characteristic information related to a data group that helped make the discussion in Section 3.4. All statistical analyses were performed using the R Statistical Computing programme. 3. Results 3.1. Particle size distribution The particle size distribution of the selected grinding grades used to obtain espressos from R, NA and WA at different temperature profile (88–93 °C, 90–90 °C and 93–88 °C) is presented in Fig. 1(a) –(c), respectively. As can be observed, particle size distribution of the different grinds presents a bimodal profile and a similar pattern. All these distribution patterns seem appropriate to obtain a good espresso: coarse particles provide a good structure of the cake and finer particles favor the extraction of coffee

Table 1 Score ranges of quality for sensory attributes defined by International Institute of Coffee Tasters (Odello & Odello, 2002). Crema

Scores range

Aroma

Scores range

Texture

Scores range

Flavor

Scores range

Colour Texture

5–7 7–9

Intensity Pleasant odors Unpleasant odors

7–9 7–9 1–3

Body

7–9

Roasty Acidity Bitterness Astringency

5–7 5–7 3–5 1–3

625

C.A. Salamanca et al. / Food Chemistry 214 (2017) 622–630

Percentage (%)

50 40 30 20 10 0 0

200

400

600

800

1000

Particle size (µm) o

o

88-93 C 88-93oC

(a)

90-90 C 90-90oC

o

93-88 C 93-88oC

Percentage (%)

50 40 30 20 10 0 0

200

400

600

800

1000

grinding is suitable for all the extraction temperature profiles investigated. Conversely, from the profiles plotted in Fig. 1(c) a different grade of grinding is needed to obtain a perfect espresso from WA beans with the required volume and extraction time at the studied extraction temperatures. Larger percentage of fine particles and lower percentage of coarse particles were observed in grinding grade for WA espressos extraction at fixed 90 °C. For the other two temperature profiles the amount of fine particles was similar but marked differences were found between percentages of particles of 400 mm and 630 mm. Andueza et al., 2003a who investigated the influence of grinding grade on the quality of expressos extracted at different fixed temperature concluded that finding the optimal grinding grade is essential to obtain a goodquality espresso coffee. Taking into account Andueza’s findings, the physicochemical characteristics of WA could be affected not only by extraction temperature but also by the different grade of grinding used for the extraction of WA espressos. Cuadras, 2009 reported that an appropriate grinding is a key for the correct extraction of the coffee components. The use of a too fine grind results an over extracted espresso as water pours slowly and as a consequence of this the espresso smells roast, its taste is bitter and astringent and there is no crema or a white spot appears on it (Cuadras, 2009).

Particle size ( µm) o

o

90-90 C 90-90oC

88-93 C 88-93oC

(b)

3.2. Physico-chemical analysis

o

93-88 C 93-88oC

Percentage (%)

50 40 30 20 10 0 0

200

400

600

800

1000

Particle size (µm) o

(c)

88-93 C 88-93oC

o

90-90 C 90-90oC

o

93-88 C 93-88oC

Fig. 1. Particle size distribution of ground beans used for different extraction temperature profile of espressos (a) Robusta, (b) Arabica natural and (c) Washed Arabica.

compounds (Andueza et al., 2003a). As seen in Fig. 1(a) and (b) particle size distribution profiles of R and NA ground coffee beans are almost superimposed. Therefore, it seems that a similar grade of

The mean value and standard deviation of physicochemical parameters analyzed in R, A and WA espressos extracted at different temperature profiles are presented in Tables 2–4, respectively. In those tables superscripts (a, b and c) indicate significant differences (p < 0.1) between mean values corresponding to different extraction temperature profiles. Results of foam layer persistence were not included in the tables because in all the cases the value of this parameter was higher than 8 min. Timing was stopped after 8 min when temperature of espresso dropped to 45 °C. At this temperature the espresso loses most of its sensorial attributes. The results presented in Table 2 correspond to the physicochemical parameters values of R espressos. As seen, espresso coffee density seems not be affected by extraction temperature while significant differences are observed between the rest of the parameters analyzed. Illy & Viani, 2005 reported that ‘‘the crema” should represent at least the 10% of the volume of an expresso. In this work, R espressos presented the higher values of foam indexes (around 33%) when extracted at ramp-up and fixed 90 °C temperature profiles. The high values of foam indexes found in this work (>10%) are probably due to the immediate use of ground coffee for making the espressos. Carbon dioxide responsible for foam formation is released during coffee grinding. Illy &

Table 2 Influence of extraction temperature profile on physico-chemical parameters of Robusta espressos. Extraction temperature profile

Foam index (%) Density (g mL 1) Viscosity (mN m 2 s 1) Turbidity (NTU)a Total solids (g L 1) Filtered solids (g L 1) pH Total polyphenolic (GAE g L 1) Caffeine (g L 1) Total lipids (g L 1) Chlorogenic acid (5-CQA) (g L 1)

(88–93) ± 0.35 °C

(90–90) ± 0.35 °C

(93–88) ± 0.35 °C

33.24 ± 2.6a 1.026 ± 0.003a 1.36 ± 0.02a 973.0 ± 160c 63.83 ± 3.2b 2.05 ± 0.4b 5.01 ± 0.82b 12.16 ± 2.0a 2.81 ± 0.5a 0.06 ± 0.01b 0.43 ± 0.02a

33.47 ± 2.0a 1.03 ± 0.01a 1.30 ± 0.01b 1844.6 ± 79a 66.87 ± 2.3a 3.05 ± 0.7a 5.58 ± 0.03b 8.59 ± 1.2b 2.45 ± 0.4b 0.09 ± 0.01a 0.35 ± 0.01c

23.21 ± 1.4b 1.03 ± 0.01a 1.25 ± 0.02c 1110.9 ± 105b 56.27 ± 2.8c 2.25 ± 0.5b 5.71 ± 0.01a 7.37 ± 0.8c 1.19 ± 0.3c 0.03 ± 0.01c 0.31 ± 0.01b

All values are presented as mean±standard deviation (n = 18). Letters next to values indicate statistical significance. Values with the same letter are not significantly different according to HSD test (P-value = 0.1). a Nephelometric Turbidity Units.

626

C.A. Salamanca et al. / Food Chemistry 214 (2017) 622–630 Table 3 Influence of extraction temperature profile on physico-chemical parameters of Natural Arabica espressos. Extraction temperature profile

Foam index (%) Density (g mL 1) Viscosity (mN m 2 s 1) Turbidity (NTU)a Total solids (g L 1) Filtered solids (g L 1) pH Total polyphenolic (GAE g L 1) Caffeine (g L 1) Total lipids (g L 1) Chlorogenic acid (5-CQA) (g L 1)

(88–93) ± 0.35 °C

(90–90) ± 0.35 °C

(93–88) ± 0.35 °C

22.87 ± 2.0ab 1.03 ± 0.01a 1.27 ± 0.02b 943.06 ± 276b 68.92 ± 7.5a 5.97 ± 1.2a 5.28 ± 0.07b 11.79 ± 3.5a 1.74 ± 0.3c 0.08 ± 0.01b 0.51 ± 0.03a

23.78 ± 2.8a 1.026 ± 0.003a 1.29 ± 0.01a 1282.50 ± 246a 59.88 ± 3.1b 3.31 ± 0.6b 5.74 ± 0.04a 7.37 ± 0.9c 2.56 ± 0.5a 0.11 ± 0.02a 0.49 ± 0.01a

21.94 ± 1.5b 1.02 ± 0.02a 1.27 ± 0.02b 1291.08 ± 56a 60.76 ± 3.5b 3.78 ± 0.8b 5.17 ± 0.01c 7.92 ± 1.2b 2.14 ± 0.2b 0.06 ± 0.01c 0.47 ± 0.03b

All values are presented as mean ± standard deviation (n = 18). Letters next to values indicate statistical significance. Values with the same letter are not significantly different according to HSD test (P-value = 0.1). a Nephelometric Turbidity Units.

Table 4 Influence of extraction temperature profile on physico-chemical parameters of Washed Arabica espressos. Extraction temperature profile

Foam index (%) Density (g mL 1) Viscosity (mN m 2 s 1) Turbidity (NTU)a Total solids (g L 1) Filtered solids (g L 1) pH Total polyphenolic (GAE g L 1) Caffeine (g L 1) Total lipids (g L 1) Chlorogenic acid (5-CQA) (g L 1)

(88–93) ± 0.35 °C

(90–90) ± 0.35 °C

(93–88) ± 0.35 °C

19.63 ± 2.1a 1.028 ± 0.002a 1.43 ± 0.06a 1212.22 ± 184c 53.68 ± 2.4b 3.57 ± 0.4b 5.30 ± 0.04a 15.15 ± 2.3a 1.60 ± 0.4b 0.07 ± 0.01b 0.56 ± 0.03a

17.78 ± 2.5b 1.02 ± 0.01a 1.26 ± 0.01c 2492.17 ± 357a 63.76 ± 2.2a 4.81 ± 0.9a 5.34 ± 0.01a 8.04 ± 1.1b 2.32 ± 0.4a 0.07 ± 0.01b 0.50 ± 0.01b

16.82 ± 1.8b 1.022 ± 0.002a 1.289 ± 0.005b 1759.17 ± 136b 54.52 ± 1.8b 2.77 ± 0.4c 5.14 ± 0.02b 6.06 ± 0.9c 0.73 ± 0.4c 0.10 ± 0.02a 0.46 ± 0.01c

All values are presented as mean ± standard deviation (n = 18). Letters next to values indicate statistical significance. Values with the same letter are not significantly different according to HSD test (P-value = 0.1). a Nephelometric Turbidity Units.

Navarini, 2011 recommend that the time expired after grinding should not be longer than 30 min to avoid carbon dioxide loss. It’s remarkable the 30% decrease of foam when using the rampdown temperature profile. Other authors who investigated the effect of different fixed temperatures of extraction on physicochemical characteristics of R espresso found that foam index increased with the increase of temperature within the range (88–98 °C)(Andueza et al., 2003b). As seen in Table 2, when the ramp-down temperature profile is used viscosity, total polyphenolic compounds, caffeine and chlorogenic acids values decrease. No clear trend with the extraction profiles was found for the rest of physicochemical properties. Viscosity is influenced by the amount of liquid droplets in the emulsion. A high viscosity is related to a high foam index, body, flavor and odor (Illy & Viani, 2005). Significant differences were found between viscosities of R espressos extracted at different temperature profiles. The highest viscosity was found for espressos brewed at ramp-up gradient of temperature. These results contrast with the ones reported by Andueza et al., 2003b for R espressos extracted at fixed temperature. These authors found lower viscosity values that increased with temperature in the range 88 °C up to 96 °C. Turbidity is defined by the lack of transparency of a liquid due to the presence of suspended particles. This parameter is directly related to total solids, filtered solids and viscosity (Navarini, Ferrari, Liverani, Liggieri, & Ravera, 2004) which are related to the body of espressos. Robusta espressos extracted at 90 °C presented the highest values of turbidity, total solids and filtered

solids. No clear trend of the effect of extraction ramps of temperature on these three parameters was observed. From the scientific side pH is a measure of hydrogen ion concentration in a solution. Cup acidity depends on the carboxylic acids content in the brew. Acetic, malic and citric acids are the most important carboxylic acids influencing the perceived acidity (Maier, 1987). Phosphoric acid and chlorogenic acids also contribute to cup acidity. Therefore, the measure of pH must be considered as a balance between acidic and alkaline compounds whose content varies depending on coffee species and roasting conditions. As seen in Table 2, the lowest acidity (higher pH) was found at ramp-down temperature profile. These values are significantly different from the ones of espressos brewed at the other two temperature profiles. Polyphenolic compounds are the responsible for color, astringency and bitterness of espresso. These compounds are present in different chemical structures and play an important role in quality of espresso and human health due to their strong antioxidant properties (Charurin, James, & del Castillo, 2002). It is noticeable the high value of polyphenolic compounds content in R espresso extracted at 88–93 °C gradient of temperature as compared to the content in the espressos brewed at the other two temperature profiles. At this ramp-up temperature polyphenolic compounds content is around 30% and 60% higher than the ones observed for fixed 90 °C and ramp-down profile, respectively. The antioxidant capacity of coffee is attributed to the presence of polyphenolic compounds, therefore the use of the ramp-up gradient enhances the antioxidant properties of R espresso. No references were found

C.A. Salamanca et al. / Food Chemistry 214 (2017) 622–630

in literature related to the effect of extraction temperature on polyphenols of espressos. Hecimovic, Belscak-Cvitanaovic, Horzic, & Komes, 2011 reported that antioxidant properties are affected by the roasting conditions. Caffeine is an alkaloid that contributes to the bitter taste of espresso (Fox et al., 2013). As seen in Table 2 the use of ramp-up gradient of temperature significantly increases the amount of caffeine in R espresso. Mora & Rodríguez, 2010 reported that after 5 s of extraction 40% of caffeine is already extracted and this percentage reaches 60% after 5 s more. Therefore the extraction temperature that acts during the first 5–10 s seems crucial in caffeine extraction and our study evidences that the use of ramp-up gradient is suitable if a high amount of caffeine is desired. The amount of lipids in the espresso depends on several parameters, coffee variety, roasting, grinding, water quality, pressure, temperature and extraction time. The lipids play an important role in the retention of aroma, thus a high lipids extraction ensures the increase of the aroma in the final cup of espresso (Illy & Viani, 2005; Speer and Kölling-Speer & Kölling-Speer, 2006). Furthermore the lipids concentration strongly affects the foam phase of the espresso (Parenti et al., 2014). Indeed, when looking at Table 2 it can be observed that the highest and lowest content of lipids were found at fixed 90 °C and at ramp-down temperature profiles that correspond to the highest and lowest foam index values measured at both temperatures, respectively. Chlorogenic acid is a trivial name used to describe a range of phenolic acids found in plant materials, included coffee (Trugo & Macrae, 1984). The most abundant chlorogenic acid is 5-CQA. The results found in this study show a decrease of 5-CQA with the increase of initial temperature of the extraction profiles used, the highest value being the one measured in the espresso brewed at 88–93 °C. Results in Table 2 evidence that the use of ramp-up temperature profile highly enhances the extraction of polyphenolic compounds whose concentration is around 30% higher than the one of espressos obtained at the other two temperature profiles. Moreover, the use of ramp-up temperature profiles results in a slightly increase of caffeine, acidic components and chlorogenic acid content in R espressos. Table 3 shows the results obtained when analyzing the physicochemical parameters in Arabica espressos brewed at different temperature profiles. As observed in the table, no significant differences as a function of extraction temperature gradient were observed between density values. Foam indexes were found to be lower than the ones observed for R espressos. Nevertheless, same correlation between foam index and lipids content as the observed for R espressos was found. Some authors also reported lower foam indexes for Arabica than for Robusta (Andueza et al., 2003b; Maeztu et al., 2001). Nevertheless, Illy & Navarini, 2011, concluded that the role played by coffee species in espresso foam is not established and postulates that the interplay between carbon dioxide content and lipid content is more relevant as far as foamability is concerned. Significant differences can be observed between viscosity of A espressos extracted at 90 °C fixed temperature and espressos obtained when ramp gradients of temperature are used. Masella et al., 2015 who investigated the effect on viscosity of fixed temperatures of extraction within the range 75 and 85 °C found that viscosity was not significantly affected. Turbidity and caffeine of A espressos were found to increase when coffee erogation was performed at fixed 90 °C and rampdown gradient of temperature. Conversely, total solids, filtered solids and 5-CQA were found to decrease their concentration value at higher initial temperatures of extraction. The highest values of these parameters were obtained when ramp-up temperature was used. In this study, the content of 5-CQA in A espressos was higher

627

than in R espressos. Marin & Puerta, 2008, also found higher content of this acid in A than in R espressos. Significant differences on pH of espressos as a consequence of extraction temperature were noticed. The highest acidity was recorded for espressos brewed at ramp-down gradient of temperature. The extraction of polyphenolic compounds of A was favored at 88–93 °C as happened in the case of R espressos. This time the increase of extraction yield was 37% and 33% higher than the extraction at 90 °C and 93–88 °C, respectively. The effect of brewing temperature profiles on physicochemical properties of WA can be observed in Table 4. The comparison between parameters values as a function of the extraction temperature highlights that at a fixed 90 °C turbidity, total solids, filtered solids and caffeine values are higher than those obtained for espressos when upward and downward gradients of temperature were used. Polyphenolic compounds and chlorogenic acid were higher extracted at 88–93 °C. The lowest concentration of these two parameters was found when the ramp-down temperature of extraction was used. Same trend was observed in R and A espressos (Tables 2 and 3). Acidity of WA espressos brewed at ramp-down and 90 °C was found to be lower than at ramp-up temperature. Concerning total lipids an opposite trend to the observed for R and A espressos with extraction temperature profiles was found. The total lipids highest extraction yield was obtained when espressos were brewed at 93–88 °C temperature profile. Unlike the results obtained for foam index of R and A espressos the highest WA foam index was observed in the espressos extracted at 88– 93 °C. It is remarkable the low values of foam index found in WA espressos analysis as compared to the ones for R and A presented in Tables 2 and 3. It has been reported that foam depends on grinding among other factors as extraction temperature, extraction grade (Illy & Navarini, 2011). As seen in Fig. 1 the grinding grade and consequently the particle distribution of fine and coarse particles of WA grounds is different for each temperature and differs from the ones of R and A. This fact could explain the differences observed between foam index of WA and the other espressos. It has been reported that grinding destroys the roasted coffee cell structure and this results in a remarkable release of carbon dioxide, with obvious consequences on crema (Illy & Navarini, 2011). Andueza et al., 2003a who studied the effect of grinding on physicochemical properties of R and A espressos found the lowest values of the studied properties when coarse particles were the major portion of ground coffee and espressos were extracted at the highest tested fixed temperature (92 °C). In this study, foam values did not show any trend with the increase of neither fine nor coarse percentage shown in Fig. 1. The highest coarse particles percentage of WA grounds was the one corresponding to the grinding grade used for WA espressos extraction at 93–88 °C profile temperature at which the lowest values of pH, total polyphenols, caffeine and 5-CQA attained their lowest value. Therefore it seems that the effect of temperature is stronger than the one of grinding as the trend of these parameters with extraction temperature was the same as the observed for R and A espressos. As previously discussed results presented in Tables 2–4 show few trends of increasing or decreasing values of studied parameters with the three temperature profiles used in this study. Coffee extraction is a dynamic process that lasts around 25 s during which various compounds are sequentially being extracted influenced by temperature and other variables. The resulting cup of espresso has a complex physical and chemical nature. In spite of this, some aspects related to the use of 90 °C fixed temperature and ramps of temperature that are common to the espressos obtained from the three varieties of coffee investigated can be highlighted: (1) Any effect of extraction temperature profiles on density (2) The lowest foam indexes were found when

628

C.A. Salamanca et al. / Food Chemistry 214 (2017) 622–630

using ramp-down extraction profile (3) The highest values of turbidity and total solids were obtained at 90 °C fixed temperature (4) The use of ramp up temperature profile produced a slight increase of chlorogenic acids (5-CQA) and an increase higher than 30% in polyphenolic compounds content. These two last aspects are very important to take into account as turbidity and total solids are related to body of the espresso and polyphenolic compounds and chlorogenic acids are responsible for the color, astringency and bitterness in espressos. Therefore, the selection of extraction temperature profile might enable to exalt or depress these sensorial attributes in the brewed coffee. 3.3. Sensory evaluation Median of sensory attributes scores of R, A and WA espressos extracted at ramp-up, fixed 90 °C and ramp-down temperature profiles are shown in Fig. 2(a–c). In the same figure the range of acceptability established by specialists of the International Institute of Coffee Tasters for each attribute was also plotted. The plot of the scores given by the panelists relative to R espressos (Fig. 2a) shows that scores for crema color, roast flavor and unpleasant odors, were within the reference range independently of the extraction temperature profile used; some attributes scores

ROBUSTA

(a)

Cream color

Astringency

Bitterness

10 9 8 7 6 5 4 3 2 1 0

as pleasant odors and acidity were lower that the lowest value of the range for all temperature profiles, aroma intensity and body when using ramp-down profile and fixed 90 °C and crema texture when the espressos were brewed at ramp-up and fixed 90 °C temperature; and astringency and bitterness scores were higher than the upper bound of the range of quality for all the tested extraction temperature profiles. According to sensory evaluation the tested R espressos can be qualified as espressos with low acidity and aroma and high bitterness and astringency. It is noticeable that the use of ramp-up temperature profile was found to increase the aroma intensity, pleasant odors and body of the espressos up to values within the range of quality. The plot in Fig. 2b evidences that scores of aroma intensity, unpleasant odors, roasty flavor and acidity of A espressos extracted at the three temperature profiles were within the reference range. Other attributes were also qualified within the reference range when brewed at upward gradient as crema color, body, bitterness and astringency. Bitterness and crema texture were also found to be in that range when A espressos were extracted at fixed 90 °C temperature. Lower scores than the lowest bound of the reference range were obtained by crema texture at ramp-up profile and pleasant odors at the three tested temperature profiles. Like it happened in the case of R espresso astringency and bitterness got

ARABICA

Cream texture

Acidity

Bitterness

Pleasent odors

Acidity

Unpleasant odors

Rosty

10 9 8 7 6 5 4 3 2 1 0

Astringency

Aroma intensity

88-93 ºC

Cream texture

Aroma intensity

Pleasent odors

Unpleasant odors

Rosty Body

Body Range Quality

(b)

Cream color

90-90 ºC

93-88 ºC

Range Quality

88-93 ºC

WASHED ARABICA

90-90 ºC

93-88 ºC

(c) Cream color

Astringency

Bitterness

10 9 8 7 6 5 4 3 2 1 0

Cream texture

Aroma intensity

Pleasent odors

Acidity

Unpleasant odors

Rosty Body Range Quality

88-93 ºC

90-90 ºC

93-88 ºC

Fig. 2. Spider plots showing consensus mean scores of espressos sensory evaluation and espresso quality range (a) Robusta (b) Arabica natural (c) Washed Arabica.

C.A. Salamanca et al. / Food Chemistry 214 (2017) 622–630

scores higher than the upper value of the reference range but in this case it occurs only when the espressos were extracted at ramp-down temperature profile. In general A espressos were perceived as espressos with fine aroma and balanced acidity and roasty flavor. The use of upward gradient produced in general espressos of good color of crema and well balanced aroma intensity, body and flavor. Conversely, espressos extracted by using downward gradient presented high bitterness and astringency, low body and low level of pleasant odors. The plot of WA sensory evaluation (Fig. 2c) reveals that in general, crema, aroma, texture and flavor attributes were scored within the reference range. The use of upward gradient resulted in well balanced espressos however downward gradient and fixed 90 °C resulted in espressos which were perceived to have high bitterness and astringency and pleasant odors scores below the reference range values. From the obtained results the better qualified espressos were those brewed at ramp-up temperature profile. Therefore, it seems that starting brewing at 88 °C, a quite low temperature compared to the standard 90 °C, and sequentially rising the temperature up to 93 °C results in an espresso of superior sensory quality. 3.4. Relationship between analytical results and the sensory evaluation Physicochemical properties of the espressos presented in Tables 2–4 can be related to a greater or to a lesser degree with the organoleptic attributes perceived by the panel of experts. With regard to taste characteristics, (Farah, 2012). chlorogenic acids, polyphenols and caffeine content are the major responsible for the perception of these attributes in espressos. Another physical property related to acidity is the pH that is in turn related to the amount of chlorogenic acids (Bähre & Maier, 1996) and other organic acids that have not been determined in this study. Boeneke, McGregor, & Aryana, 2006 reported that pH tended to decrease when chlorogenic acids were present in a high concentration. Therefore, any slight increase or decrease of the concentration of these compounds could modify the typical bitterness-acidity balance recognized in espressos that could be only perceived through the sensorial attribute acidity (Parenti et al., 2014). In this study, R espressos showed the highest content of polyphenols, caffeine and chlorogenic acids when espressos were extracted upward gradient and the lowest one when falling temperature profile was used. Conversely, the pH followed the opposite trend being the highest pH the one of espressos extracted at falling temperature profile. In the tasting, all espressos presented low acidity and high bitterness and astringency at all the extracting temperature profiles. Arabica Natural espressos also presented the highest content of polyphenolic compounds and chlorogenic acid at upward gradient but at this temperature profile the content of caffeine was the lower than the obtained at the other two temperature profiles. These espressos brewed at 88–93 °C showed a certain balance between astringency and bitterness was perceived in cup. In the espressos brewed at downward gradient a greater perceived astringency and bitterness was noticed while acidity was scored in the range of quality. Presumably this fact is due to the low content of polyphenols, caffeine and chlorogenic acids. The highest and the middle values of the pH were observed for the fixed 90 °C and ramp-up profile where the content of polyphenols, caffeine and chlorogenic acids was lower. Nevertheless, it seems that the content of these compounds did not affect the acidity since the scores for this attribute were within the target values. In the case of WA espressos, the highest values for polyphenols and chlorogenic acids was observed when using rising and fixed 90 °C temperature profiles which was reflected in cup as higher bitterness for these two profiles than for the one of falling temperature while acidity and bitterness were kept within the target values.

629

The pH values were high, medium and low for the flat, rising and falling temperature profiles at which the acidity was perceived to be within the limits of the reference range. It should be noted that although the highest value of pH and caffeine was observed for espressos brewed at 90 °C the content of both polyphenols and clorogenic acids was intermediate. Results obtained in this study cannot completely describe the aromatic profile of the espressos investigated but some discussion about the content of lipids and aroma attributes can be carried out. Lipids content has been directly related to the aroma of espresso. The lipid fraction protects from the aromatic compounds degradation during the Maillard reaction (De Oliveira, Cruz, Eberlin, & Cabral, 2005). Concerning aroma attributes, aroma intensity values of R espressos were lower that the target values except the ones of espressos brewed at rising temperature profile. No direct relation between the content of lipids was found as the highest lipids content was found at fixed 90 °C profile at which aroma intensity was out of the range of quality. Aroma intensity perceived in Arabica natural espressos extracted at the three temperature profiles was found to be within the range of quality. Conversely the attribute pleasant odors were qualified to be below the reference range of values. As lipids content found in these espressos was in the same range of the other two tested espressos the low level of pleasant odors could be due to degradation of some aromatic compounds responsible of aromatic notes in the aroma of the espressos. The aroma intensity of Washed Arabica was qualified within the reference range. Note that only the use of upward gradient produced espressos with the appropriated level of pleasant odors in spite of the similar content of lipids in espressos obtained at rising and fixed temperature profiles and the highest content observed at falling temperature profile. Body of an espresso is an attribute related to physicochemical properties as turbidity, total solids, filtered solids and viscosity (Illy & Viani, 2005; Masella et al., 2015). In general, espressos obtained from R grounds stand out as having a good body. In spite of this, only the body of R espressos brewed at rising temperature was considered to be within the range of quality. The body of espressos brewed at the other two temperature profiles was qualified to be lower than the target values in spite of the fact that turbidity, total solids and filtered solids at 90 °C were higher than at rising temperature profile. In the case of A espressos the highest total solids and filtered solids content was found at rising temperature profile at which body was evaluated as being within the range of quality. However, at this temperature viscosity and turbidity values were lower than the ones found at the other two temperature profiles. In WA espressos turbidity, total solids and filtered solids were high at fixed 90 °C and intermediate at rising temperature, conversely, viscosity was high at the rising temperature and low at fixed temperature. The discussion carried out in this section reveals that finding a direct relationship between sensory attributes evaluation and the physicochemical properties values presented in Tables 2–4 is not always possible due to the large amount and complex reactions taking place during erogation that constitutes the cup. On the other hand, the sensory attributes perceived in cup are associated to more than one of the studied variables or to other variables that have not been considered in this study.

4. Conclusions The effect of using ramps of temperature to brew espressos from coffee of three different varieties on espressos physicochemical properties and quality has been investigated. The results showed: (i) Significant differences between most of the physicochemical properties as a function of the extraction temperature

630

C.A. Salamanca et al. / Food Chemistry 214 (2017) 622–630

profile in all the espressos investigated (ii) The use of ramp of temperatures affects in a different way the physicochemical properties of espressos extracted from different coffee species (iii) Sensorial evaluation of the espressos also varies depending on the extraction temperature profile leading to espressos of different quality (iv) The use of 88–93 °C gradient of temperature contributes to increasing the quality of all espressos investigated. In conclusion, the possibility of programming gradient of temperature in the coffee machines recently introduced in the market opens new expectations in the field of espresso brewing. The use of gradient temperature for brewing espressos allows increasing or decreasing the extraction of some compounds from coffee grounds according to consumers demand. For instance, caffeine and polyphenols are very appreciated by the consumers for their stimulating and antioxidant effects, respectively. Therefore the physicochemical analysis of espressos is an adequate tool to help making decisions on the temperature profiles to get the preferred content of compounds. Of course, the physicochemical analyses don’t assess quality of the espressos as quality is associated at the consumer’s visual, olfactory and gustative sensations of some attributes of a complex beverage such as an espresso. Nevertheless, an appropriate gradient of temperature can be used to highlight or hide some of the sensorial attributes so as to enhance the quality of an espresso. Acknowledgements The authors wish to express their gratitude to Rancilio group for providing the coffee machine and for their financial support to this study. Thanks are also extended to Cafés Saula for providing the coffee samples, Maria Peiroló for her help in the experimental work and the panel of experts for their assessment. References Andueza, S., de Peña, M. P., & Cid, C. (2003a). Chemical and sensorial characteristics of espresso coffee as affected by grinding and torrefacto roast. Journal of Agricultural and Food Chemistry, 51, 7034–7039. Andueza, S., Maeztu, L., Dean, B., de Peña, M. P., Bello, J., & Cid, C. (2002). Influence of water pressure on the final quality of Arabica espresso coffee. Application of multivariate analysis. Journal of Agricultural and Food Chemistry, 50(7426–7), 431. Andueza, S., Maeztu, L., Pascual, L., Ibáñez, C., de Peña, M. P., & Cid, C. (2003b). Influence of extraction temperature on the final quality of espresso coffee. Journal of the Science of Food and Agriculture, 83, 240–248. Aqeel, Z., Truong, D., Preston, J., & Lazzaro, S. (2012). Chlorogenic Acids from Green Coffee by HPLC, Phenomenex TN-1134 http://www.phenomenex.com/Info/ WebDocumentServe/tn1134.pdf (accessed 25.05.16). Bähre, F., & Maier, G. H. (1996). Electrophoretic clean-up or forganic acids from coffee for the GC/MS analysis. Fresenius Journal of Analytical Chemistry, 355, 190–193. Blumberg, S., Frank, O., & Hofmann, T. (2010). Quantitative studies on the influence of the bean roasting parameters and hot water percolation on the concentrations of bitter compounds in coffee brew. Journal of Agricultural and Food Chemistry, 58, 3720–3728. Boeneke, D., McGregor, J., & Aryana, K. (2006). The effect of sweetners on the acceptability of dairy-based espresso beverages. International Journal of Dairy Technology, 59(1), 12–17. Campa, C., Doulbeau, S., Dussert, S., Hamon, S., & Noirot, M. (2005). Qualitative relationship between caffeine and chlorogenic acid contents among wild coffea species. Food Chemistry, 93, 135–139.

Cavaco Bicho, N. C., Leitao, A. E., Cochicho Ramalho, J., de Alvarenga, N. B., & Cebola Lidon, F. (2011). Identification of nutricional descriptors of roasting intensity in beverages of Arabica and Robusta coffee beans. International Journal of Food Sciences and Nutrition, 62(8), 865–871. Charurin, P., James, J., & del Castillo, M. (2002). Antioxidant activity of coffee model systems. Journal of Agriculture and Food Chemistry, 50, 3751–3756. Cuadras, S. (2009). La cata del café espresso. Revista Fórum del café 26 http:// www.forumdelcafe.com/pdf/F-36_Caf%C3%A9_espresso.pdf (accessed 25.05.16). De Oliveira, A. L., Cruz, P. M., Eberlin, M. N., & Cabral, F. A. (2005). Brazilian roasted coffee oil obtained by mechanical expelling: compositional analysis by GC-MS. Food Science and Technology (Campinas), 25(4), 677–682. Farah, A. (2012). Coffee constituents. In Yi-Fang Chu (Ed.), Coffee: Emerging health effects and disease prevention (pp. 21–58). Illinois: John Wiley & Sons, Inc. by Blackwell Publishing Ltd. Fox, G., Wu, A., Yiran, L., & Force, L. (2013). Variation in caffeine concentration in single coffee beans. Journal of Agriculture and Food Chemistry, 61, 10772–10778. Fujioka, K., & Shibamoto, T. (2008). Chlorogenic acid and caffeine contents in various commercial brewed coffees. Food Chemistry, 106, 217–221. Gloess, A. N., Schönbächler, B., Klopprogge, B., D’Ambrosio, L., Chatelain, K., Bongartz, A., ... Yeretzian, C. (2013). Comparison of nine common coffee extraction methods: instrumental and sensory analysis. European Food Research and Technology, 236, 607–627. Hecimovic, I., Belscak-Cvitanaovic, A., Horzic, D., & Komes, D. (2011). Comparative study of polyphenols and caffeine in different coffee varieties affected by the degree of roasting. Food Chemistry, 129, 991–1000. Illy, E., & Navarini, L. (2011). Neglected food bubbles: the espresso coffee foam. Food Biophysics, 6, 335–348. Illy, A., & Viani, R. (1995). Espresso coffee. The chemistry of quality. London UK: Academic Press Limited. Illy, A., & Viani, R. (Eds.). (2005). Espresso coffee: The science of quality (2nd ed.). London: Elsevier Academic Press. Ludwig, I. A., Bravo, J., de Peña, M. P., & Cid, C. (2013). Effect of sugar addition (torrefacto) during roasting process on antioxidant capacity and phenolics of coffee. LWT-Food Science and Technology, 51, 553–559. Maeztu, L., Andueza, S., Ibáñez, C., de Peña, M. P., Bello, J., & Cid, C. (2001). Multivariate methods for characterization and classification of espresso coffees from different botanical varieties and types of roast by foam, taste and mouthfeel. Journal of Agricultural and Food Chemistry, 49, 4743–4747. Maier, H. G. (1987). The acids of coffee. In Proceedings of the 12th ASIC Colloquium (Montreux). Paris, France: ASIC. Marin, C., & Puerta, G. (2008). Contenido de ácidos clorogénicos en granos de Coffea Arabica L. y Coffea Canephora según el desarrollo del fruto. Cenicafé, 59, 7–28. Masella, P., Guerrini, L., Spinelli, S., Calamai, L., Spugnoli, P., Illy, F., & Parenti, A. (2015). A new espresso brewing method. Journal of Food Engineering, 146, 204–206. Mora, I., & Rodríguez, A. (2010). Comportamiento de la cafeína en un espresso. Revista del Fórum Cultural del Café. number 42. Navarini, L., Ferrari, M., Liverani, F. S., Liggieri, L., & Ravera, F. (2004). Dynamic tensiometric characterization of espresso coffee beverage. Food Hydrocolloi ds, 18(3), 387–393. Navarini, L., & Rivetti, D. (2010). Water quality for espresso coffee. Food Chemistry, 122, 424–428. Nunes, F. M., Coimbra, M. A., Duarte, A. C., & Delgadillo, I. (1997). Foamability, foam stability and chemical composition of espresso coffee as affected by the degree of roast. Journal of Agricultural and Food Chemistry, 45, 3238–3243. Odello, L., & Odello, C. (2002). Espresso Italiano tasting.L’assaggio (8 ed.). Brecia, Italy: International Institute of Coffee Tasters. Parenti, A., Guerrini, I., Massella, P., Spinelli, S., Calamai, L., & Spugnoli, P. (2014). Comparison of espresso coffee brewing techniques. Journal of Food Engineering, 121, 112–117. Singleton, V. L., & Rossi, J. A. (1965). Colorimetry of Total Phenolics with Phosphomolybdic-Phosphotungstic Acid Reagents. American Journal of Enology and Viticulture, 144–158. Speer, K., & Kölling-Speer (2006). I. The lipid fraction of the coffee bean. Brazilian Journal of Plant Physiology, 18, 201–216. Trugo, L. C., & Macrae, R. (1984). A study of the effect of roasting on the chlorogenic acid composition of coffee using HPLC. Food Chemistry, 15, 219–227.