Food Research International 62 (2014) 315–325

Contents lists available at ScienceDirect

Food Research International

j ourna l homepage: www.e lsev ie r .com/ locate / foodres

Review

Complexity of coffee flavor: A compositional and sensory perspective

Wenny B. Sunarharum a,b, David J. Williams c, Heather E. Smyth a,⁎a Queensland Alliance for Agriculture and Food Innovation (QAAFI), The University of Queensland, PO Box 156 Archerfield BC, Queensland 4108, Australiab Department of Food Science and Technology, Faculty of Agricultural Technology, University of Brawijaya, JL. Veteran Malang 65145, Indonesiac Agri-Science Queensland, Department of Agriculture, Fisheries and Forestry (DAFF), PO Box 156, Archerfield BC, Queensland 4108, Australia

⁎ Corresponding author. Tel.: +61 7 32766035.E-mail address: h.smyth@uq.edu.au (H.E. Smyth).

http://dx.doi.org/10.1016/j.foodres.2014.02.0300963-9969/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 30 November 2013Accepted 23 February 2014Available online 1 March 2014

Keywords:CoffeeFlavorCoffea arabicaAromaSensoryReview

For the consumer, flavor is arguably the most important aspect of a good coffee. Coffee flavor is extremelycomplex and arises fromnumerous chemical, biological andphysical influences of cultivar, coffee cherrymaturity,geographical growing location, production, processing, roasting and cup preparation. Not surprisingly there is alarge volume of research published detailing the volatile and non-volatile compounds in coffee and that are likelyto be playing a role in coffee flavor. Further, there is much published on the sensory properties of coffee. Never-theless, the link between flavor components and the sensory properties expressed in the complex matrix ofcoffee is yet to be fully understood. This paper provides an overview of the chemical components that arethought to be involved in the flavor and sensory quality of Arabica coffee.

© 2014 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3152. Flavor perception of coffee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316

2.1. Aroma and taste sensations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3162.2. Mouthfeel and chemesthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316

3. The biochemical generation of coffee flavor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3164. The compositional drivers of coffee flavor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318

4.1. Non-volatile components and their contribution to coffee flavor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3184.2. Volatile components and their contribution to coffee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319

5. Determination of key volatile aroma compounds in coffee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3206. Relationship between sensory properties and composition of coffee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3207. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322

1. Introduction

Coffee is a very popular brewed beverage that is consumed world-wide. In terms of financial value, it is the most important agriculturalcommodity after petroleum. Coffee consumption is increasing acrossthe globe and in 2011 roasted coffee was estimated to be worth aroundUS $75.4 billion dollars in the global retail market and close to US $175.7billion in total gross value including out of home sales (ICO, 2011).World leaders in coffee production are Brazil, Vietnam, Indonesia and

Colombia while the leading consumers are the USA, Germany, Japan,Italy and France (ICO, 2012).

Good quality coffee flavor has been described as a pleasant sensa-tion, a balanced combination of flavor, body and aroma in the absenceof faults (Mori et al., 2003). Flavor remains themost important consum-er parameter and warrants thorough investigation from a sensory andcompositional perspective (Mori et al., 2003). Not surprisingly, the com-position and sensory properties of coffee has been a target for researchfor over a century.

The flavor and distinctive sensory qualities of coffee varies enor-mously across the globe due to influences of genetic strain, geographicallocation, unique climates, differing agricultural practices and variations

316 W.B. Sunarharum et al. / Food Research International 62 (2014) 315–325

in processing method applied. While there is a volume of work onindividual or groups of chemical components present in coffee, thelikely importance of individual flavor components to certain coffeetypes, as well as the sensory properties and consumer preferences forcoffee, there is limited information available that directly links per-ceived sensory properties of coffee to specific compositional compo-nents. Further, most studies of coffee flavor are limited to identifyingkey flavor components in a single coffee type, style or geographicallocation. Not surprisingly, the ‘key’ flavor compounds identified in onestudy are often a different set of ‘key’ flavor components identified byanother, depending on the specific coffee studied. It is clear that for acomprehensive understanding of coffee flavor, that includes the spec-trum of coffee flavor types, new studies are needed that investigatecoffee flavor from a broad perspective so that we may truly understandthe compositional drivers of coffee flavor.

This review focuses on the processes in coffee preparation thatinfluence flavor and details the volatile and non-volatile chemical com-ponents identified in coffee which are thought to contribute to thediverse flavor and sensory properties of Arabica (Coffea arabica) coffeefrom around the world. While coffee flavor has been previouslyreviewed (Buffo & Cardelli-Freire, 2004; Grosch, 1998), the reviewextends our understanding on coffee flavor by focusing on the sourceof flavor generation and the missing links in our knowledge of thecause and effect relationship between components and sensory proper-ties in coffee globally.

2. Flavor perception of coffee

2.1. Aroma and taste sensations

Flavor is a complex sensation which can be described as a combina-tion of aroma, taste, texture andmouthfeel (Taylor & Roozen, 1996) andchemesthesis or trigeminal sensations (Cliff & Green, 1994). The aroma,or odor, is arguably the most important component of coffee flavor.

Many consumers generally perceive and explain taste as what theysmell which has led to flavor being sometimes defined as the ‘olfactorycomponent of taste’ perceived retronasally (Petracco, 2001). Retronasalperception occurswhen food volatiles flow from themouth through theback of the throat reaching the nasal cavity through the pharynx(Petracco, 2001) whereby volatiles interact with receptors on theolfactory epithelium, generating olfactory nerve stimulus and signaltransmission via the olfactory bulb to the brain which then processesthe sensory information as odor recognition (Mombaerts, 2001a,2001b).Orthonasal perception occurs when volatiles are inhaled through thenose and interact with the olfactory system directly (Petracco, 2001).While it has been speculated that the human sense of smell can distin-guish more than 10,000 different odorants (Lancet, 1986), the humansense of taste (detected from the tongue receptors) can detect fivebasic taste sensations, namely sweet, bitter, sour, salty, and umami(savory) (Rawson & Li, 2004). Consequently, the aroma component offlavor for complex products such as coffee is thought to be exceptionallyimportant and is primarily responsible for flavor diversity (Lawless &Heymann, 2010; Murphy, Cain, & Bartoshuk, 1977).

The sensory properties of coffee have been studied for many yearsand, with increasing consumption worldwide, interest in coffee flavorand aroma has gained momentum from industry and scientists alike.Recent examples of sensory language that has been used to describeflavor properties of coffee include attributes such as astringency, body,bitter flavor, burned aroma, ‘typical’, and burned tastes (Bicho, Leitão,Ramalho, de Alvarenga, & Lidon, 2013), sweet-caramel, earthy, roast/sulfur and smoky characteristics (Czerny, Mayer, & Grosch, 1999;Mayer, Czerny & Grosch, 2000). Further detail on aroma lexicon (term/description) includes coffee, roasted, burnt/acrid, brown, beany, nutty,cocoa, musty/earthy, floral, fruity, green, ashy/sooty, sweet aromatic,sour aromatic, and pungent (Bhumiratana, Adhikari, & Chambers, 2011).

Previous studies have concluded that brewing results in an increaseof sweet-caramel aroma of Arabica coffee while more spicy, harsh,earthy aroma prevailed in Robusta (Blank, Sen, & Grosch, 1991). Furtherstudies of Arabica and Robusta coffee flavor roasted across three differ-ent levels (Bicho et al., 2013) indicate that the characteristic odor,astringency, body, bitter flavor, burned aroma, and residual, typical,and burned tastes, citric acid flavor and aroma accounts for the differ-ence between these two species (Bicho et al., 2013).

The above-mentioned sensory properties measured through asensory evaluation involving humans as the assessors. One of theindustrial standard measurement of sensory quality (cup quality) ofcoffee (R. Teixeira, Teixeira, & Brando, 2005) is a cup evaluation orcoffee ‘cupping’. This method involves trained industry assessors whoevaluate the coffee grounds and fresh brew for aroma and flavor withsubsequent visual evaluation of the green and roasted beans.

2.2. Mouthfeel and chemesthesis

Besides the aroma and taste, texture, mouthfeel and chemesthesisare other components that influences flavor perception and are influ-enced by food structure interaction with the lining of the mouth duringconsumption (Taylor & Roozen, 1996). These sensations typicallyinclude properties such as crunchiness, oiliness, grittiness, viscosity,softness or hardness and also include more complex sensations createdfrom interactions of food componentswith the surface of themouth andtongue due to chemical sensitivity of the skin and mucous membranesto burning, tickling, prickling, and cooling sensations (Cliff & Green,1994). Importantly, texture, mouthfeel and chemesthesis sensationsare not detected via the olfactory system or taste receptor pathways.

Sensory language to describe the texture andmouthfeel of commer-cial coffee has been developed by research from Japan and Korea andincludes terms such as having body (viscosity), astringency, round,smooth, thick, coarse, grainy, rough, oily, and sticky, with overallimpressions of being crisp, pure, non-persistent, clear, sharp, mild,round, soft, delicate, balanced, intense, strong, heavy, hard, light, neutral,monotonous, flat, simple and light, mellow, winey, rich, nippy, piquant,pungent, tangy, acrid, alkaline, easy to swallow and refreshing(Hayakawa et al., 2010; Seo, Lee, Jung, & Hwang, 2009). An Italianstudy on the sensory classification of espresso developed descriptorssuch as thick, lingering, full-mouthed, viscous, resistance to tongue-palate movements, syrupy, consistency, velvety, pasty/doughy, creamy,mouth-coating, smooth, round, clinging/tongue coating, particulate,bulky, rich/heavy (Navarini, Cappuccio, Suggi-Liverani, & Illy, 2004).

3. The biochemical generation of coffee flavor

The generation of coffee flavor begins in the coffee plant whereflavor precursors form as the coffee cherries develop. Flavor complexityfurther develops throughout the varying steps of coffee processing andsubsequent cup preparation techniques (Fig. 1).

Coffee from the species Coffea arabica and Coffea robusta are the twomost commonly grown in commercial production and differ distinctlyin flavor (Bicho et al., 2013). Within the Coffea arabica species, numer-ous varieties can be distinguished which suit the myriad environmentswhere coffee is grown around the world. The environmental factorssuch as geographical origins (Bhumiratana et al., 2011; Costa Freitas &Mosca, 1999), climate, altitude and temperature elevation (Bertrandet al., 2006, 2012), shading (Bosselmann et al., 2009), and nutritionalor fertilizers (Poltronieri, Martinez, & Cecon, 2011) had been suggestedto have an impact on coffee quality.

In green coffee bean processing, twomajor processing techniques areapplied to the harvested coffee fruits to green bean and include dry pro-cessing (natural) and wet processing (washed) (Clarke & Macrae,1985). Semi-dry (semi-washed) processing is an additional methodwhich comprises components of both the dry and wet processingmethods (A. A. Teixeira, Brando, Thomaziello, & Teixiera, 2005). The

Fig. 1. Factors that influence coffee flavor complexity from farm to cup.

317W.B. Sunarharum et al. / Food Research International 62 (2014) 315–325

main difference between thesemethods is the pulping operation aswellas the fermentation and washing process (Clarke & Macrae, 1985) andthese differences in processing can impact coffee flavor. Dry processingproduces a ‘hard’ coffee with amedicinal flavor (Clarke &Macrae, 1985)whilewet processing yields a better quality coffeewith less body, higheracidity and more aroma than the dry processing (Mazzafera & Padilha-Purcino, 2004). The semi-dry (semi-washed or pulped natural) whichis a compromise between the dry and wet method, offers a coffee withintermediate body (Duarte, Pereira, & Farah, 2010).

As well as the above-mentioned green coffee bean processes, thereare certain ‘specialty’ styles of coffee that are produced using moreuncommon processing methods such as a ‘digestive bio-processing’(e.g. fermentation inside the intestine of Luwak or civet mammals)(Marcone, 2004, 2011; Ongo et al., 2012) and monsooning. The latterwas developed in India after unique and desirable flavor traits were dis-covered in coffee beans that had been shipped under humid (monsoon)conditions for an extended period of time (Ahmad, Tharappan, &Bongirwar, 2003).

Further commercial processing of the green coffee beans involvesroasting, grinding, and brewing which are arguably themost importantfactors contributing to flavor of the coffee beverage. Roasting has themost significant influence on coffee flavor and has been the focus ofmuch research (Buffo & Cardelli-Freire, 2004; Eggers & Pietsch, 2001;Esquivel & Jiménez, 2012). Roasting temperatures can vary and aretypically between 180 °C to 240 °C for periods of between 8 to15 min. During roasting endothermic and exothermic processes beginfrom heat transferred to the bean through hot gases or contact withthe metal surface of the coffee roaster which reduces water content ofthe coffee beans and causes puffing and cooling to produce desirablecharacteristics. The impact of roasting on flavor comes from the degra-dation and formation or release of numerous chemical compoundsthrough Maillard reactions, Strecker degradation, break down ofamino acids, degradation of trigonelline, quinic acid, pigments, lipids

and interaction between intermediate products (Buffo & Cardelli-Freire, 2004). Importantly, roasting relates directly to cup quality as itconverts the pea-like, green smell of raw green coffee into the pleasantaromas characteristic of roasted coffee due to a drastic increase ofnumerous aroma compounds (Czerny & Grosch, 2000; Czerny et al.,1999). Typically, while more complex aromas are formed at a mediumroasting level, a light roast produces sweet, cocoa, and nutty aromasand dark roasting is responsible for burnt/acrid, ashy/sooty, sour,pungent, coffee, and roasted characteristics (Bhumiratana et al., 2011).While roasting level is a matter of personal preference, certain roastingconditions may better suit coffees of different variety, style, geographi-cal origin or end use, depending on the aroma characteristics desiredin the resulting beans (Bhumiratana et al., 2011). Medium roasting, forexample,will express the regionalflavors derived from the geographicaloriginmuch better than dark roasting, whichwouldmask these charac-teristics in the coffee beans.

Grinding of roasted bean releases coffee flavor for the purpose ofextraction or infusion in coffee beverage preparation (Akiyama et al.,2003) and therefore higher intensities of aroma tend to be perceivedafter grinding of the roasted beans (Bhumiratana et al., 2011). Thegrind level and particle size influences the extraction and thus thequality of prepared beverage. Too fine a coffee grind could produce alow volume and bitter coffee due to over extraction while too coarse agrind could decrease extraction due to reduction in surface arearesulting in a weak insipid coffee brew (Andueza, de Peña, & Cid, 2003).

Itmust be acknowledged that the global trendof coffee consumptionis toward convenience and health with a growing consumer conscienceand interest for origin, variety, brewing and grinding, flavor, packaging,social ‘content’ and ambience (Ponte, 2002). This has resulted in anincrease in specialty and convenience coffee products such as instant(including decaffeinated), but more recently flavored flavor coffeecapsules or pods. The latter might include the addition of other naturalor artificial flavor (Petracco, 2001). These products are highly processed

318 W.B. Sunarharum et al. / Food Research International 62 (2014) 315–325

and often involve formulation and addition of other ingredients toretain consistency. Consequently, the flavors of these products are out-side the scope of this review.

Brewing of coffee is a crucial step in translating coffee flavors fromthe ground bean into thewatermatrix of the beverage for consumption.There are a number of brewing methods applied for coffee beveragepreparation which can generally be classified under decoction (boiled,Turkish, percolator, vacuum coffees), infusion or steeping (filter,Napoletana) and pressure methods (plunger, moka, espresso) assummarized below (Petracco, 2001):

• Decoction methods: involve contact or continuous reflux of groundcoffee in water for certain time and high temperature, extract moreintensively and faster than other methods but results in some flavorloss as the common method is through boiling with a direct contactto heat or at high temperature.

• Infusion: conducted by soaking or steeping ground coffee (extracoarse–medium grind) under heated or cold water for a period beforefiltering, produces a milder coffee than decocted coffee with enhanceacidity and flavor.

• Pressure methods: involve the percolation of fluid through a porousmediumor a filter by application of high pressure and heat to enhancethe body of the beverage such as in an espresso style.

4. The compositional drivers of coffee flavor

4.1. Non-volatile components and their contribution to coffee flavor

Non-volatile compounds present in roasted coffee beans whichmaybe important to coffee flavor include alkaloids (caffeine, trigonelline),chlorogenic acids, carboxylic acids, carbohydrates and polymericpolysaccharides, lipids, protein, melanoidins and minerals (Buffo &Cardelli-Freire, 2004). The occurrence of these components in commer-cial roasted coffee beans is quite diverse, due to the variability in coffeecultivation and processing (as described previously).

Compared to Robusta, Arabica has been reported to contain a highercaffeine content as green or roasted bean and as instant coffee(Oestreich-Janzen, 2010; Wasserman, 1992). Caffeine, a nitrogenoussecondary metabolite, is thought to influence the perceived strength,body and bitterness of a brewed coffee (Clarke & Macrae, 1985).Alkaloids, which in neat form have a bitter flavor, are extractable inwater and may give a physiological stimulating effect (Higdon & Frei,2006; Lean & Crozier, 2012).

Trigonelline (N-methylpyridinium-3-carboxylate) and its twoderivatives (nicotinic acid and N-methylnicotinamide) are other alka-loids present in coffee (Buffo & Cardelli-Freire, 2004). Unlike caffeine,these components can be found at higher levels in the Arabica varietycompared to other cultivars (Wasserman, 1992). They are thought tocontribute to the overall aroma perception of both roasted coffeebeans and a brewed coffee beverage (Oestreich-Janzen, 2010).

Chlorogenic acids are a family of esters formed between certaintrans-cinnamic acids (phenolic acids that usually are caffeic acid, ferulicacid and p-coumaric acids) and quinic acid (Clifford, 1985, 1999). Thesesecondary metabolites are present in coffee beans and contribute to theastringency (Buffo & Cardelli-Freire, 2004) and bitterness of a coffeebeverage, and have potential as an antioxidant for human health(Oestreich-Janzen, 2010). Chlorogenic acids are one of the most abun-dant polyphenols present in plant and plant-based foods and coffee hasbeen reported to be one of the richest sources of chlorogenic acids inthe human diet compared to other beverages (Clifford, 1999, 2000). Acup of Arabica coffee brew (200 ml) contains 70–200 mg of chlorogenicacid, while in Robusta it may reach 70–350 mg (Clifford, 1999). Due tothermal instability, further processing, particularly roasting of greencoffee beans, has been reported to progressively degrade chlorogenicacids (Clifford, 1972) up to 93% for dark roasting (Farah, Monteiro,Calado, Franca, & Trugo, 2006). Chlorogenic acids and quinic acid may

form chlorogenic lactones during coffee roasting (Farah, de Paulis,Trugo, & Martin, 2005) which contribute to increased bitterness of thecoffee brew (Ginz & Engelhardt, 2001).

Acidity, or the tartness, is an important attribute of coffee quality incombination with sweetness, bitterness and aroma profile. In coffee,acidity is often conversely correlated to sweetness. Arabica coffeebrews are more acidic than Robusta, with pH ranges of 4.85–5.15 and5.25–5.40, respectively (Vitzthum, 1976). The acid content of greencoffee bean is around 11%, mainly comprising citric, malic, chlorogenicand quinic acids, while roasted bean contains around 6% due todecreases in citric, malic and chlorogenic acids (Ginz, Balzer, Bradbury,& Maier, 2000; Stegen & Duijn, 1987; Urgert et al., 1995). Duringroasting of coffee beans, these acids form other compounds such aslactones, from the reaction between chlorogenic and quinic (Bennat,Engelhardt, Kiehne, Wirries, & Maier, 1994), and volatile phenols, suchas guaiacol and 4-vinylguaiacol from chlorogenic acid degradation(Vitzthum, Weissmann, Becker, & Kohler, 1990). These breakdownproducts are volatiles that can influence coffee aroma.

While some acids degrade during coffee bean roasting, othersincrease in concentration including formic, acetic, glycolic and lacticacids. While the first two aliphatic acids increase only up to mediumroast before beginning to degrade, the latter two continue to increaseduring roasting (Ginz et al., 2000; Weers, Balzer, Bradbury, & Vitzthum,1995). Dark roasting is the most efficient way to reduce acid contentand perceived acidity in coffee (Clifford, 1985).

Polysaccharides are a major component of green and roastedcoffee beans, in the form of arabinogalactans, mannans and cellulose(Bradbury, 2001). Polysaccharides play an important role in retainingvolatiles and therefore flavor, and also contribute to the perceivedviscosity of the coffee brew (Buffo & Cardelli-Freire, 2004). Other carbo-hydrate compounds such as glucose and fructose are mainly found inimmature beanswhile higher amounts of sucrose accumulate inmaturebeans and contribute to perceived coffee sweetness (Oestreich-Janzen,2010; Wasserman, 1992).

The lipid fraction of coffee, also known as the coffee oil, is the largestconstituent of green coffee beans and consists of triglycerides (75%),total free and esterified diterpene alcohols (19%), total free and esteri-fied sterols (5%), and a small quantity of other lipid types such astocopherols (Kaufmann & Gupta, 1964; Kaufmann & Hamsagar, 1962).The diterpenes kahweol and cafestol in coffee are often mentioned ashaving a positive effect on health in relation to cholesterol (Speer &Kölling-Speer, 2001). Roasting of coffee beans results in migration ofcoffee oil to the bean's surface (Savonitti, 2005). While some changesin coffee lipid profile occurs during roasting, sterols and most triglycer-ides remain unchanged (Maier, 2005). These lipid fractions of the beansare extracted into the coffee brew and provide the crema emulsion ofespresso coffee that carries flavor volatiles and fat-soluble vitamins,and contributes to perceived texture and mouthfeel of the coffee brew(Oestreich-Janzen, 2010).

Protein content of Arabica is slightly lower than Robusta, eventhough total amino acid composition is similar (Wasserman, 1992).The amino acids content of green bean has an important contributionto flavor development during roasting through Maillard reactions(Liu & Kitts, 2011). Maillard or caramelization reactions occur due toa reaction between the amine group of amino acids or nitrogen-containing compounds and the carboxyl group of reducing sugars,hydroxy-acids and phenols to yield aminoaldoses and aminoketonesby condensation (Buffo & Cardelli-Freire, 2004). The resulting productis the brownish colormelanoidins andother components suchas severalnitrogen and/or sulfur containing heterocyclic compounds whichare thought to be important flavor compounds in coffee (Shibamoto,1983).

Amongotherminor constituents of coffee are theminerals. Potassiumis the major mineral present in roasted coffee, however, manganese,iron, and copper are also present in smaller amounts and act as impor-tant catalysts of certain biochemical reactions which facilitate the

319W.B. Sunarharum et al. / Food Research International 62 (2014) 315–325

production and release of flavor components in coffee bean during pro-cessing (Oestreich-Janzen, 2010).

Non-volatiles present in coffee beans and brew are important to thesensory quality of coffee and have been related to positive as well asnegative aspects of coffee flavor. Carbohydrates impact on sweetness,caramel notes arise from Maillard reactions between sugars andamino acids, and caffeine and chlorogenic acids contribute to bitterness.Specifically, trigonelline, 3,4-dicaffeoilquinic acid and, to some extent,caffeine, have been associated with a good cup quality in BrazilianArabica coffee (Farah et al., 2006). Elevated amounts of chlorogenicacids mainly 5-caffeoylquinic acid, and to some extent feruloylquinicacid and associated oxidation products, are related with poor cupquality and off-flavors such as harsh medicinal, phenolic or iodine-likeflavors (Spadone, Takeoka, & Liardon, 1990).

4.2. Volatile components and their contribution to coffee

Aroma volatiles produced during coffee bean roasting are arguablythe most important quality-determinant of coffee (Andueza et al.,2003; Baltes & Knoch, 1993; Grosch, Czerny, Mayer, & Moors, 2000;Kumazawa &Masuda, 2003) and as such have been of research interestfor almost a century with intensive profiling over the past 50 years.Aroma volatiles characterize not only the different cultivars, styles andprocessing techniques used, but also the geographical origins of thecoffee (Costa Freitas & Mosca, 1999). To date, there have been morethan 1000 volatiles identified in coffee (Nijssen, Visscher, Maarse,Willemsense, & Boelens, 1996) ranging in concentration from part permillion (ppm) to part per trillion (ppt) levels, however, only a smallnumber of these are important to the flavor and aroma characteristicsof coffee (Buffo & Cardelli-Freire, 2004; Grosch, 2001a). Some authorssuggest that as few as 20–30 individual volatiles may be important tothe aroma of any single type or style of coffee (Blank et al., 1991;Blank, Sen, & Grosch, 1992; Czerny & Grosch, 2000; Czerny et al.,1999; Deibler, Acree, & Lavin, 1998; Grosch, 2000; Mayer, Czerny, &Grosch, 2000; Mayer & Grosch, 2001; Sanz, Maeztu, Zapelena, Bello, &Cid, 2002; Semmelroch & Grosch, 1996; Semmelroch, Laskawy, Blank,& Grosch, 1995).

Coffee volatiles are derived from numerous precursors found in thebean and from chemical reactions occurring particularly duringroasting, but also during processing and storage (Buffo & Cardelli-Freire, 2004). The generation of aroma compounds has been previouslyreviewed (Buffo & Cardelli-Freire, 2004; Grosch, 2001b). The mainchemical reactions that occur during roastingwhich generate importantaroma volatiles include Maillard reactions (non-enzymatic browning),phenolic acid and carotenoid degradation (Holscher & Steinhart, 1992;Reineccius, 1995; Tressl, 1980); Strecker degradation; breakdown ofsulfur amino acids, hydroxy-amino acids, proline and hydroxyproline;degradation of trigonelline, chlorogenic acids and quinic acid, pigments,and lipids; as well as reactions between other intermediate products(Buffo & Cardelli-Freire, 2004; Ribeiro, Augusto, Salva, Thomaziello, &Ferreira, 2009).

Coffee volatile compounds comprise several chemical classesincluding hydrocarbons, alcohols, aldehydes, ketones, carboxyclic acids,esters, pyrazines, pyrroles, pyridines, other bases (e.g. quinoxalines,indoles), sulfur compounds, furans, furanones, phenols, oxazolesamong others. Quantitatively, the top two classes in coffee are furansand pyrazines, while qualitatively, sulfur-containing compoundstogether with pyrazines are considered the most significant to coffeeflavor (Nijssen et al., 1996). These compounds vary significantly in con-centration and sensory potency which makes coffee flavor extremelycomplex, and explains why different coffee types may exhibit suchdiverse, unique and specific flavors (Risticevic, Carasek, & Pawliszyn,2008).

Furans are among the most abundant group of volatiles present incoffee (Grosch, 2001a) and are found in sensorily or aroma active signif-icant concentrations in roasted coffee (Akiyama et al., 2007; Bicchi et al.,

2011; Cheong et al., 2013; Gianturco, Friedel, & Giammarino, 1964;Ribeiro, Augusto, Salva, & Ferreira, 2012). They are formed throughthermal degradation of carbohydrates, ascorbic acid, or unsaturatedfatty acids during roasting (Crews & Castle, 2007; Ribeiro et al., 2009)and range in concentration from 3–115 ppb in coffee brew (Kuballa,Stier, & Strichow, 2005). Volatile furans exhibit malty and sweet roastedaromas (Akiyama et al., 2007; Burdock, 2010; Fors, 1983) with sensorythresholds that are relatively high compared to other groups of coffeevolatiles (Burdock, 2010) although due to their high concentrationsare still considered of importance to coffee flavor. In relation to health,there is concern over possible negative health impacts of furans andtherefore commercial coffee roasting has been optimized to minimizethe presence of furans (Bicchi et al., 2011; EFSA, 2004).

Due to their sensory potency, sulfur-containing compounds such asthiols are among the most important contributors to coffee flavordespite their presence at relatively low concentration. For example,2-furfurylthiol which has a very low sensory threshold (0.01 ppb)(Semmelroch et al., 1995) exhibits a strong roasted aroma (Blanket al., 1992). This compound is considered by many as a key impactaroma compound in coffee and has been reported widely in roastedand brewed coffee (Akiyama et al., 2007; Blank et al., 1992; Czerny &Grosch, 2000; Grosch et al., 2000; Holscher & Steinhart, 1992; Mayeret al., 2000; Michishita et al., 2010; Semmelroch et al., 1995). Otherthiols, such as 2-methyl-3-furanthiol and 3-methyl-2-butene-1-thiol,are also present in coffee and have very low sensory thresholds whileexhibiting meaty characters (Akiyama et al., 2003; Blank et al., 1991,1992; Grosch et al., 2000). Also belonging to the class of sulfur-containing compounds, 3-methylthiophene (Ribeiro et al., 2009) and2,4-dimethyl-5-ethylthiazole (Blank et al., 1992) are present in coffeeat sensorily significant levels and exhibit roasted and meaty flavors(Maga, 1975).

Pyrazines arewell known class of compounds that arise as a productof roasting various foods and horticultural products including coffee.They are an abundant class of compounds present in coffee, with lowsensory threshold concentrations and they are of key importance tothe flavor of coffee. Generally, pyrazines have been described asexhibiting nutty, earthy, roasty, green aromas (Akiyama et al., 2007;Blank et al., 1991; Czerny & Grosch, 2000; Czerny, Wagner, & Grosch,1996; Czerny et al., 2008; Holscher & Steinhart, 1992; Semmelroch& Grosch, 1996; Wagner, Czerny, Bielohradsky, & Grosch, 1999).Ethylpyrazines and ethenylalkylpyrazines have been reported tocontribute to the earthy aroma characteristic of Robusta (Blank et al.,1991). The volatile 3-isobutyl-2-methoxypyrazine, with an exceptionallylow sensory threshold of 0.002 ppb (Belitz, Grosch, & Schieberle, 2009) ispresent at low concentrations in roasted Arabica coffee beans butstill have been reported to have a significant impact on roasted Arabicacoffee (Czerny & Grosch, 2000). Arguably the other two most importantaroma compounds are 2-ethyl-3,5-dimethylpyrazine and 2,3-diethyl-5-methylpyrazine (Akiyama et al., 2003; Blank et al., 1991; Czerny et al.,1999; Grosch et al., 2000; Mayer & Grosch, 2001; Mayer et al., 2000;Semmelroch et al., 1995).

Furanones are generated in coffee mainly via the Maillard reactionand subsequent aldol condensation (Grosch, 2001b). They are a signifi-cant group of volatiles in coffee in terms of abundance and potency.Major flavor contributors are thought to be 4-hydroxy-2,5-dimethyl-3(2H)-furanone and2(5)-ethyl-4-hydroxy-5(2)-methyl-3(2H)-furanone,3-hydroxy-4,5-dimethyl-2(5H)-furanone (sotolon), and 4-ethyl-3-hydroxy-5-methyl-2(5H)-furanone (abhexon) and are thought to beresponsible for the sweet caramel aroma of roasted coffee (Akiyamaet al., 2003, 2007; Blank et al., 1992).

Certain phenolic compounds which generated and released duringroasting are thought to be of importance to coffee flavor (Ribeiro et al.,2009). Particuarly guaiacol, 4-ethylguaiacol and 4-vinylguaiacol whichare have a spicy phenolic aroma (Akiyama et al., 2007; Blank et al.,1992; Czerny & Grosch, 2000) and vanillin (Czerny & Grosch, 2000). Inroasted Arabica coffee, phenolic compounds range in concentration

320 W.B. Sunarharum et al. / Food Research International 62 (2014) 315–325

from 3 to 56 ppm (Cheong et al., 2013; Czerny et al., 1999; Semmelrochet al., 1995) depending on the variety and geographical source. Thesephenolic compounds arise from thermal degradation of chlorogenicacids (mainly ferulic, caffeic and quinic acids) and their concentrationin roasted bean is proportional to the amount of chorogenic acidspresent in green bean. Given there are significantly more chlorogenicacids present in green bean of Robusta compared to Arabica, thesevolatiles are thought to play a key role in the flavor differentiationbetween these two varieties of coffee (Bicho et al., 2013; Blank et al.,1991; Sanz et al., 2002; Semmelroch & Grosch, 1996; Semmelroch et al.,1995).

5. Determination of key volatile aroma compounds in coffee

In an attempt to unlock the compositional basis of coffee flavor therehave beennumerous studies involving the extraction and analysis of thevolatile profiles of coffee from green bean to roasted ground bean andbrew, from espresso style to instant coffee (Akiyama et al., 2003,2007; Blank et al., 1992; Borém et al., 2013; Czerny & Grosch, 2000;Czerny et al., 1996, 1999; Grosch et al., 2000; Kumazawa & Masuda,2003; Mayer et al., 2000; Michishita et al., 2010; Ribeiro, Teófilo,Augusto, & Ferreira, 2010; Ribeiro et al., 2009; Semmelroch & Grosch,1995, 1996). The overriding objective of these studies has been to iden-tify those volatiles that are key contributors to the flavor of coffee. Froman extensive review of the coffee literature, a summary of individualcompounds thought to be of greatest important to the flavor of Arabicacoffee is provided in Table 1 together with reported concentrationreported in coffee (ppb), aroma description (as a neat compound), sen-sory threshold concentration (ppb), and literature references wherestudies have indicated the importance of each compound to coffeeflavor.

Studies on coffee flavor have relied on a variety of methods todetermine the relative and likely importance of individual volatiles tothe flavor of coffee. These methods include: direct comparisons ofrelative compound concentrations with sensory detection thresholds(Czerny et al., 2008); application of odor activity values (OAVs) (Acree,Barnard, & Cunningham, 1984; Semmelroch et al., 1995) whereby theratio of concentration of each compound with its sensory threshold iscalculated and likely odor-activity of components are ranked (Grosch,2001a); calculation of odor spectrum values (OSV)where likely activityof odor compounds are ranked independent of concentration using gaschromatography-olfactometry (GC–O) methods (Acree et al., 1984);and application of other GC–O methods such as aroma extract dilutionanalysis (AEDA) (Blank et al., 1991, 1992) or Charm analysis (Grosch,2001a).

Given the extensive differences in the genetics, cultivation, processingand geographical origins of coffee, it is not surprising that coffeeexhibits a broad range of flavor and aroma types. Consequently, whilethere has been extensive investigation on the key aroma volatiles ofcoffee, individual studies often report different sets of key volatilesthat are representative of the particular sample of coffee studies.Further, the different methods used for measuring volatile compoundcomposition in coffee may also result in differences in ranking ofkey volatiles of any particular coffee sample. This is especially apparentin the case of detecting components at trace levels, such as themore recent detection of potent 3-mercapto-3-methylbuthyl acetate(Kumazawa&Masuda, 2003), 1-(3,4-dihydro-2H-pyrrol-2-yl)-ethanoneand 4-(4-hydroxyphenyl)-2-butanone (Akiyama et al., 2007) in Arabicacoffee brew.

In a study of roasted and brewedArabica coffee asmany as 13 potentodorants have been identified as key aroma contributors of roastedcoffee based on AEDA experiments (Blank et al., 1992). 2-Ethyl-3,5-dimethylpyrazine was found to be important for roasted and brewedcoffee while 3-mercapto-3-methylbutyl-formate, 2-furfurylthiol, and(E)-β-damascenone were important to the roasted powdered coffeewhilemethional, sotolon, and 4-hydroxy-2,5-dimethyl-3(2H)-furanone

and vanillin were important contributors to the brew (Blank et al.,1992). Another study on roasted Arabica and Robusta coffee quantified14 and subsequently 22 important compounds and highlightedmethional, trialkylated pyrazines, guaiacol, 4-vinylguaiacol, 4-hydroxy-2,5-dimethyl-3(2H)-furanone as key aroma volatiles (Semmelroch &Grosch, 1996; Semmelroch et al., 1995). A geographical comparisonbetween blends and varieties of medium roasted Arabica coffee fromfour countries identified 28 potent odorants which were thoughtto be key to the aroma of those coffees (Mayer, Czerny, & Grosch,1999). Different geographical origins and roasting resulted in differentconcentrations of important compounds identified (Mayer et al.,1999). The compounds 2,3-butanedione, 2,3-pentanedione, 3-isobutyl-2-methoxypyrazine, 4-hydroxy-2,5-dimethyl-3(2H)-furanone, 4-vinyl-guaiacol, 4-ethylguaiacol, 2-furfurylthiol, 3-mercapto-3-methylbutyl-formate and 3-methyl-2-buten-1-thiol were found to be affected byorigins while propanal, 2(5)-ethyl-4-hydroxy-5(2)-methyl-3(2H)-furanone, guaiacol, 4-ethylguaiacol, 2-furfurylthiol, 3-methyl-2-buten-1-thiol andmethanethiol were affected by roasting (Mayer et al., 1999).

A study using flavor omission and aroma model experiments werecarried out with 27 coffee odorants dissolved in an oil/water mixture(Czerny et al., 1999) using previously quantified compounds fromthe headspace of roasted Arabica coffee powder (Mayer et al., 2000).The findings confirmed the importance of 2-furfurylthiol, somealkylpyrazines, furanones and phenols, methional and 3-mercapto-3-methylbutyl formate as key drivers of Arabica coffee (Mayer et al.,2000).

Most of the research conducted on identifying key odor-contributingvolatiles dates back to the90s. In the past decade, there have been only ahandful of studies investigating importance of coffee volatiles. Thesestudies have focused on topics such as investigating key aroma changesin green coffee (Scheidig, Czerny, & Schieberle, 2007), fingerprintingcoffee flavor (Huang et al., 2007), and discrimination of volatiles indefective coffee (Toci & Farah, 2008). More recent studies of coffeecomposition focus on coffee, health or bioactive compounds (Cheonget al., 2013; de Azeredo, 2011; Yeretzian, Pascual, & Goodman, 2012).

6. Relationship between sensory properties and composition ofcoffee

Compositional data only, is not enough to explain the importance ofkey compounds and importantly, the nature of their contribution, tocoffee flavor. Similarly, sensory information of coffee aroma properties,in the absence of good quality chemical data, cannot be used to explainwhat's causing specific sensory attributes. Good quality and compre-hensive research that matches these properties in coffee to explain thecompositional basis of coffee flavor is still limited.

To fully understand the correlation between sensory (consumerdata) and the sensory descriptive analysis results or physicochemicalmeasurements, researchers may use a multivariate data analysis toolsknown as chemometrics (Resurreccion, 1988; Wold & Sjostrom, 1998).Commonly applied methods are a principal component analysis (PCA)and a partial least squares (PLS) regression which are widely used fora food analysis and are useful in identifying compounds accountingfor specific aroma nuances in complex systems such as coffee.

Despite the wide application of chemometrics, correlating composi-tional data with sensory attributes is a complicated task and can beproblematic if the methodology used to collect the information is notsuitably comprehensivewith a degree of accuracy and precision. Conse-quently, there are few studies to date that correlate physicochemicaland sensory attributes of coffee by means of a multivariate tools inunderstanding coffee flavor. A recent application of PCA has successfullydiscriminated aroma characteristics of Arabica coffee from three differ-ent origins and different roasting level (Bhumiratana et al., 2011) aspreviously explained in Section 3. PCA was also applied to successfullydescribe sensory effects of additives on the quality of stored Colombiancoffee brews (Pérez-Martínez, Sopelana, de Peña, & Cid, 2008). Further,

Table 1A summary of important aroma compounds identified in Arabica coffee.

Key odorants identified in coffee (literature cited) Concentration in Arabica coffee (ppb)* Aroma description Sensory threshold (ppb)***

Aldehyde2-methylbutanal 1–4 20,7002 – 1.326

2-methylpropanal5 – buttery oily27 –

3-methylbutanal1–4,6,7 18,6002 malty11 0.3526

(E)-2-nonenal8,9 1914 – 0.0827

3-methylpropanal10 – – –

acetaldehyde1,2,4 139,0002 – 0.7a,28

methylpropanal1,2,4,7 32,3002 – 0.726

p-anisaldehyde11 – minty11 27b,29

phenylacetaldehyde12,13 – sweet fruity12 –

propanal1,2 17,4002 – 1026

Acid2-methylbutyric acid11,12 25,0004 sweaty14 1028

3-methylbutyric acid3,8 18,060–32,1804,15 sweaty11 700c,30

Esterethyl-2-methylbutyrate14 3.914 fruity14 0.5d,14

ethyl-3-methylbutyrate14 1414 fruity14 0.6d,14

Furanfurfural10,15 5880–19,37015 – 280a,28

2-((methylthio)methyl)furan12 – smoke–roast12 –

2-furanemethanol acetate10 24,520–40,04015 – –

2-methylfuran10 – – –

5-methyl-2-furancarboxyaldehyde10,12 – – 6000a,28

furfurylformiate13 – – –

furfurylmethyl ether13 – – –

furfurylformate10 4060–6,42015 – –

furfuryldisulfide13 – – –

Sulfur-containing compoundsdimethyl trisulfide12 2819 cabbage-like28 0.00127

bis(2-methyl-3-furyl)disulphide11 – meaty11 0.00076b,29

methional3,11,16,17 213–24014,22 boiled potato-like11 0.216

Thiols3-mercapto-3-methylbutylformate3,7,11,12,16,17 13016 green blackcurrant12 0.00355

2-furfurylthiol1-5,8,11,12,14,16,17 1080–5,0809,15,16 roasty (coffee-like)11 0.0116

2-methyl-3-furanthiol1,11 60–682,4 meaty, boiled11 0.00727

3-mercapto-3-methylbutylacetate18 7.518 roasty18 –

3-methyl-2-butene-1-thiol3,19 134 amine-like19 0.00035

methanethiol7,16 4,55023 – 0.0227

Thiophene3-methylthiophene10 – – –

Thiazole2,4-dimethyl-5-ethylthiazole11 – earthy, roasty11 –

Furanonedihydro-2-methyl-3(2H)-furanone13,19 7580-30,00015,24,25 – 0.005e,31

2-ethyl-4-hydroxy-5-methyl-3(2H)-furanone2,3 16,8002 sweet caramel3 207

3-Hydroxy-4,5-dimethyl-2(5H)-furanone (sotolone)2,14,11,12,16,17,19 1.1–1,4702,14,16 sweet caramel3 2016

4-Hydroxy-2,5-dimethyl-3(2H)-furanone (furaneol)2,3,8,11,16,17 10,930–112,0002,15,16 sweet caramel12 1016

5-Ethyl-3-hydroxy-4-methyl-2(5H)-furanone (abhexon)2,11,16, 19 104–1602,16 seasoning-like, caramel-like11 7.516

5-ethyl-4-hydroxy-2-methyl-3(2H)-furanone16 17,30016 sweet caramel, – 1.1516

Ketone1-octen-3-one12 – mushroom-like28 0.0036b,29

2,3-hexadione13 – – –

2,3-butanedione3,4,5 48,400–50,8002,7 buttery–oily12 0.3 a,28

2,3-pentanedione3,4,5,6 3540–39,6002,15,7 buttery–oily12 20 a,28

3,4-dimethylcyclopentenol-1-one19 – caramel-like, sweet11 –

4-(4′-hydroxyphenyl)-2-butanone12,20 120 sweet fruity12 (raspberry ketone) 1–1032

1-(2-furanyl)-2-butanone10 – – –

Norisoprenoid(E)-β-damascenone3,7,11,14,16 195–2552,14,16 honey-like, fruity11 0.0007516

Phenolic compoundsguaiacol2,11,16,17,19 2000–17,9702,15,16,25 phenolic, burnt11 2.516

4-ethylguaiacol10-12,14,16,17,19 800–24,8002,14,15,16,25 spicy11 25 a,28

4-vinylguaiacol1-3,11,12,14,16,19 8000–64,8002,15,16,25 spicy11 0.75 a,28

vanillin11,12,14,16,17 2290–4,8002,14,16 vanilla-like14 2516

Pyrazine2,3-dimethylpyrazine15 2580–6,10015 – 800 a,28

2,5-dimethylpyrazine15 4550–11,73015 – 80 a,28

(continued on next page)

321W.B. Sunarharum et al. / Food Research International 62 (2014) 315–325

Table 1 (continued)

Key odorants identified in coffee (literature cited) Concentration in Arabica coffee (ppb)* Aroma description Sensory threshold (ppb)***

Pyrazine2,3-diethyl-5-methylpyrazine1-3,11,12,16,17,19 73–952,16 nutty–roast12 0.09 a,28

2-ethenyl-3,5-dimethylpyrazine1,2,17,21 522 earthy21 0.000012f,21

2-ethenyl-3-ethyl-5-methylpyrazine2,21 182 earthy21 0.000014f,21

2-ethyl-3,5-dimethylpyrazine1-4,10,11,12,16 55-3302,4,16 nutty–roast12 0.0433

2-ethyl-3,6-dimethyl-pyrazine10 2570-5,98015 – 8.633

2-methoxy-3,5-dimethylpyrazine14 1.123 earthy 0.006f,14

2-methoxy-3,2-methylpropylpyrazine3,12 – green earthy12 –

2-methoxy-3-isopropylpyrazine11,19 2.423 earthy roasty11 0.002a,28

3-ethenyl-2-ethyl-5-methylpyrazine21 – – –

3-isobutyl-2-methoxypyrazine2,11,14,17 59-972,14,16 peasy14 0.00227

6,7-dihydro-5-methyl-5H-cyclopentapyrazine3,11 – nutty–roast27 –

ethylpyrazine13 – – 4000a,28

Pyridinepyridine13 21,280–65,52015 – 7716

Pyrrole1-methyl pyrrole13 – negative notes–defective beans6 –

Terpenelinalool11,12 – flowery11 0.17b,14

limonene11 – – 4 a,28

geraniol12,20 – – 1.129

Compounds identified as important contributors, concentration measured and/or aroma description and sensory threshold provided in literatures 1(Grosch et al., 2000); 2(Czerny et al.,1999); 3(Akiyama et al., 2003); 4(Mayer&Grosch, 2001); 5(Holscher& Steinhart, 1992); 6(Ribeiro et al., 2010); 7(Semmelroch &Grosch, 1996); 8(Michishita et al., 2010); 9(Tressl & Silwar,1981); 10(Ribeiro et al., 2009); 11(Blank et al., 1992); 12(Akiyama et al., 2007); 13(Ribeiro et al., 2012); 14(Czerny & Grosch, 2000); 15(Cheong et al., 2013); 16(Semmelroch et al., 1995);17(Mayer, Czerny & Grosch, 2000); 18(Kumazawa & Masuda, 2003); 19(Blank et al., 1991); 20(Akiyama et al., 2008); 21(Czerny et al., 1996); 22(Clarke & Vitzhum, 2001); 23(Grosch,2001a,); 24(Gianturco et al., 1964); 25(Silwar, Kamperschroer, & Tressl, 1987); 26(Milo & Grosch, unpublished); 27(Belitz et al., 2009); 28(Burdock, 2010); 29(Czerny et al., 2008);30(Salo, 1970); 31(Barrett, Halsey, & Peppard, 1983); 32(Larsen & Poll, 1992); 33(Buttery & Ling, 1997).*authors report compound concentrationwithin the range indicated, these concentrations relate to roasted Arabica coffee grounds or beans (weight/weight) (not coffee brew).Where noconcentration is listed, non could be found in the literature.**aroma description sourced only from coffee literature***where two ormore sensorydetection thresholdswere found, the lowest is presented. All sensory thresholdsweredetermined inwater except: amatrix unknown; b thresholdmeasuredby first diluting compounds in ethanol in a defined concentration and then dissolved inwater, for linalool as R-linalool; c in ethanolic solution 9.5%; dthreshold in cellulose; e in ale; fin air.

322 W.B. Sunarharum et al. / Food Research International 62 (2014) 315–325

PLS has been applied to correlate sensory data to volatile chomato-graphical profiles resulting in adequate predictions of acidity, cleanli-ness, overall quality, bitterness, body, and flavor of Brazillian Arabicacoffee (Ribeiro et al., 2012).

The few studies that have correlated physicochemical and olfactoryor sensory panel data typically analyze a limited range of coffee flavor-types (Akiyama et al., 2008; Bhumiratana et al., 2011; Pérez-Martínezet al., 2008; Ribeiro et al., 2009, 2012). Certainly there is scope for futureresearch efforts to model coffee flavor more comprehensively in termsof the range of sensory properties exhibited in Arabica coffees fromaround the world.

7. Conclusion

Complexity of coffee flavor arises from numerous influences fromcultivation to processing and preparation. Variations in these influencescause differences in the formation of flavor and aroma components inthe green and roasted coffee bean and subsequent brew. From the com-positional point of view, the volatiles and non-volatiles have a greatinfluence on flavor perception and consumer acceptance and enjoy-ment of coffee.

Knowledge on the chemical composition of coffee flavor is impor-tant, but reliable measurement and ranking of aroma componentsin coffee in the absence of good quality sensory information cannoteffectively describe the importance, or the nature of contribution, ofindividual or groups of flavor components in coffee. Further, the coffeematrix itself interacts with volatiles and has a large impact on theperceived flavor assessed through a sensory study. Thus, matching orcreating a comprehensive link on all components of coffee flavor andsensory quality will lead to a deeper understanding of coffee flavor.For example elucidating what compounds cause the nutty, cocoa, cara-mel, fruity, or ‘coffee-type’ flavor which can then be subsequently

tracked back to individual processes involved in their formation. Under-standing on flavor will aid the coffee industry to control desirable flavoroutcomes of coffee through processing or other farm managementtechnique.

Acknowledgements

This research was supported under Australian Research Council'sLinkage Projects funding scheme (project number LP130100376).We acknowledge Green Cauldron Coffee and the Australia AwardsEndeavour Postgraduate Scholarship for their ongoing support of thisresearch.

References

Acree, T. E., Barnard, J., & Cunningham, D. G. (1984). A procedure for the sensory analysisof gas-chromatographic effluents. Food Chemistry, 14(4), 273–286, http://dx.doi.org/10.1016/0308-8146(84)90082-7.

Ahmad, R., Tharappan, B., & Bongirwar, D. R. (2003). Impact of gamma irradiation on themonsooning of coffee beans. Journal of Stored Products Research, 39(2), 149–157,http://dx.doi.org/10.1016/s0022-474x(01)00043-1.

Akiyama, M., Murakami, K., Hirano, Y., Ikeda, M., Iwatsuki, K., Wada, A., et al. (2008).Characterization of headspace aroma compounds of freshly brewed arabica coffeesand studies on a characteristic aroma compound of Ethiopian coffee. Journal of FoodScience, 73(5), C335–C346, http://dx.doi.org/10.1111/j.1750-3841.2008.00752.x.

Akiyama, M., Murakami, K., Ikeda, M., Iwatsuki, K., Wada, A., Tokuno, K., et al. (2007).Analysis of the headspace volatiles of freshly brewed arabica coffee using solid-phase microextraction. Journal of Food Science, 72(7), C388–C396, http://dx.doi.org/10.1111/j.1750-3841.2007.00447.x.

Akiyama, M., Murakami, K., Ohtani, N., Iwatsuki, K., Sotoyama, K., Wada, A., et al. (2003).Analysis of volatile compounds released during the grinding of roasted coffee beansusing solid-phase microextraction. Journal of Agricultural and Food Chemistry, 51(7),1961–1969, http://dx.doi.org/10.1021/jf020724p.

Andueza, S., de Peña, M. P., & Cid, C. (2003). Chemical and sensorial characteristics ofespresso coffee as affected by grinding and torrefacto roast. Journal of Agriculturaland Food Chemistry, 51(24), 7034–7039, http://dx.doi.org/10.1021/jf034628f.

323W.B. Sunarharum et al. / Food Research International 62 (2014) 315–325

Baltes, W., & Knoch, E. (1993). Model reactions on roast aroma formation. XIII.The formation of some uncommon N-heterocyclic compounds and furans afterroasting of tryptophan with reducing sugars and sugar degradation products. FoodChemistry, 46(4), 343–349, http://dx.doi.org/10.1016/0308-8146(93)90002-w.

Barrett, J., Halsey, S. A., & Peppard, T. L. (1983). Flavour stability of ales brewed usinghydrolysed maize syrup. Sensory and analytical aspects. Journal of the Institute ofBrewing, 89(5), 356–360.

Belitz, H. -D., Grosch, W., & Schieberle, P. (2009). 5. Aroma compounds. Food Chemistry,714. (pp. 340–402). Berlin Heidelberg: Springer.

Bennat, C., Engelhardt, U. H., Kiehne, A., Wirries, F. M., &Maier, H. G. (1994). HPLC analysisof chlorogenic acid lactones in roasted coffee. Zeitschrift für Lebensmittel-Untersuchungund-Forschung, 199(1), 17–21, http://dx.doi.org/10.1007/bf01192945.

Bertrand, B., Boulanger, R., Dussert, S., Ribeyre, F., Berthiot, L., Descroix, F., et al. (2012).Climatic factors directly impact the volatile organic compound fingerprint in greenArabica coffee bean as well as coffee beverage quality. Food Chemistry, 135(4),2575–2583, http://dx.doi.org/10.1016/j.foodchem.2012.06.060.

Bertrand, B., Vaast, P., Alpizar, E., Etienne, H., Davrieux, F., & Charmetant, P. (2006).Comparison of bean biochemical composition and beverage quality of arabica hybridsinvolving Sudanese–Ethiopian origins with traditional varieties at various elevationsin Central America. Tree Physiology, 26(9), 1239–1248, http://dx.doi.org/10.1093/treephys/26.9.1239.

Bhumiratana, N., Adhikari, K., & Chambers, E. (2011). Evolution of sensory aromaattributes from coffee beans to brewed coffee. LWT - Food Science and Technology,44(10), 2185–2192, http://dx.doi.org/10.1016/j.lwt.2011.07.001.

Bicchi, C., Ruosi, M. R., Cagliero, C., Cordero, C., Liberto, E., Rubiolo, P., et al. (2011).Quantitative analysis of volatiles from solid matrices of vegetable origin by high con-centration capacity headspace techniques: Determination of furan in roasted coffee.Journal of Chromatography A, 1218(6), 753–762, http://dx.doi.org/10.1016/j.chroma.2010.12.002.

Bicho, N. C., Leitão, A. E., Ramalho, J. C., de Alvarenga, N. B., & Lidon, F. C. (2013). Impactof roasting time on the sensory profile of Arabica and Robusta coffee. Ecology ofFood and Nutrition, 52(2), 163–177, http://dx.doi.org/10.1080/03670244.2012.706061.

Blank, I., Sen, A., & Grosch, W. (1991). Aroma impact compounds of Arabica and Robustacoffee. Qualitative and quantitative investigations. Paper presented at the 14th ASICColloquium, San Francisco.

Blank, I., Sen, A., & Grosch, W. (1992). Potent odorants of the roasted powder and brew ofArabica coffee. Zeitschrift für Lebensmittel-Untersuchung und -Forschung, 195(3),239–245, http://dx.doi.org/10.1007/bf01202802.

Borém, F. M., Ribeiro, F. C., Figueiredo, L. P., Giomo, G. S., Fortunato, V. A., & Isquierdo, E. P.(2013). Evaluation of the sensory and color quality of coffee beans stored in hermeticpackaging. Journal of Stored Products Research, 52(0), 1–6, http://dx.doi.org/10.1016/j.jspr.2012.08.004.

Bosselmann, A. S., Dons, K., Oberthur, T., Olsen, C. S., Ræbild, A., & Usma, H. (2009). Theinfluence of shade trees on coffee quality in small holder coffee agroforestry systemsin Southern Colombia. Agriculture, Ecosystems & Environment, 129(1–3), 253–260,http://dx.doi.org/10.1016/j.agee.2008.09.004.

Bradbury, A.G.W. (2001). Chapter 1. Chemistry I:Non-volatile compounds. InR. J. Clarke, &O. G. Vitzhum (Eds.),Coffee: Recent developments (pp. 1–17). Oxford: Blackwell Science.

Buffo, R. A., & Cardelli-Freire, C. (2004). Coffee flavour: An overview. Flavour and FragranceJournal, 19(2), 99–104, http://dx.doi.org/10.1002/ffj.1325.

Burdock, G. A. (2010). Fenaroli's handbook of flavor ingredients (6th ed.). Boca Raton: CRCPress/Taylor & Francis.

Buttery, R. G., & Ling, L. C. (1997). 2-Ethyl-3,5-dimethylpyrazine and 2-ethyl-3,6-dimethylpyrazine: Odor thresholds in water solution. LWT - Food Science andTechnology, 30(1), 109–110, http://dx.doi.org/10.1006/fstl.1996.0139.

Cheong, M. W., Tong, K. H., Ong, J. J. M., Liu, S. Q., Curran, P., & Yu, B. (2013). Volatilecomposition and antioxidant capacity of Arabica coffee. Food Research International,51(1), 388–396, http://dx.doi.org/10.1016/j.foodres.2012.12.058.

Clarke, R. J., & Macrae, R. (Eds.). (1985). Coffee. Chemistry, volume 1. London: Elsevier Ap-plied Science Publishers.

Clarke, R. J., & Vitzhum, O. G. (2001). Coffee: Recent developments. Oxford: BlackwellScience.

Cliff, M. A., & Green, B. G. (1994). Sensory irritation and coolness produced by menthol:Evidence for selective desensitization of irritation. Physiology & Behavior, 56(5),1021–1029, http://dx.doi.org/10.1016/0031-9384(94)90338-7.

Clifford, M. N. (1972). The phenolic compounds of green and roasted coffee beans. Glasgow:University of Strathclyde.

Clifford, M. N. (1985). Chlorogenic acids. In R. J. Clarke, & R. Macrae (Eds.), Coffee(pp. 153–202). Netherlands: Springer.

Clifford, M. N. (1999). Chlorogenic acids and other cinnamates — Nature, occurrence anddietary burden. Journal of the Science of Food and Agriculture, 79(3), 362–372, http://dx.doi.org/10.1002/(sici)1097-0010(19990301)79:3b362::aid-jsfa256N3.0.co;2-d.

Clifford, M. N. (2000). Chlorogenic acids and other cinnamates — Nature, occurrence,dietary burden, absorption and metabolism. Journal of the Science of Food andAgriculture, 80(7), 1033–1043, http://dx.doi.org/10.1002/(sici)1097-0010(20000515)80:7b1033::aid-jsfa595N3.0.co;2-t.

Costa Freitas, A. M., & Mosca, A. I. (1999). Coffee geographic origin — An aid to coffeedifferentiation. Food Research International, 32(8), 565–573, http://dx.doi.org/10.1016/s0963-9969(99)00132-5.

Crews, C., & Castle, L. (2007). A review of the occurrence, formation and analysis of furanin heat-processed foods. Trends in Food Science & Technology, 18(7), 365–372, http://dx.doi.org/10.1016/j.tifs.2007.03.006.

Czerny, M., Christlbauer, M., Christlbauer, M., Fischer, A., Granvogl, M., Hammer, M., et al.(2008). Re-investigation on odour thresholds of key food aroma compounds anddevelopment of an aroma language based on odour qualities of defined aqueous

odorant solutions. European Food Research and Technology, 228(2), 265–273, http://dx.doi.org/10.1007/s00217-008-0931-x.

Czerny, M., & Grosch, W. (2000). Potent odorants of raw Arabica coffee. Their changesduring roasting. Journal of Agricultural and Food Chemistry, 48(3), 868–872, http://dx.doi.org/10.1021/jf990609n.

Czerny, M., Mayer, F., & Grosch, W. (1999). Sensory study on the character impactodorants of roasted Arabica coffee. Journal of Agricultural and Food Chemistry, 47(2),695–699, http://dx.doi.org/10.1021/jf980759i.

Czerny,M.,Wagner, R., &Grosch,W. (1996). Detection of odor-active ethenylalkylpyrazinesin roasted coffee. Journal of Agricultural and Food Chemistry, 44(10), 3268–3272, http://dx.doi.org/10.1021/jf9600824.

de Azeredo, A. M. C. (2011). Coffee roasting: Color and aroma-active sulfur compounds.3497021 Ph.D. United States — Florida: University of Florida (Retrieved from http://search.proquest.com/docview/922403813?accountid = 14723 ProQuest Disserta-tions & Theses Full Text database).

Deibler, K. D., Acree, T. E., & Lavin, E. H. (1998). Aroma analysis of coffee brew by gaschromatography–olfactometry. In E. T. Contis, C. T. Ho, C. J. Mussinan, T. H.Parliment, F. Shahidi, & A. M. Spanier (Eds.), Developments in food fcience, Vol. 40.(pp. 69–78). Elsevier.

Duarte, G. S., Pereira, A. A., & Farah, A. (2010). Chlorogenic acids and other relevantcompounds in Brazilian coffees processed by semi-dry and wet post-harvestingmethods. Food Chemistry, 118(3), 851–855, http://dx.doi.org/10.1016/j.foodchem.2009.05.042.

EFSA (2004). Report of the scientific panel on contaminants in the food chain on provi-sional findings of furan in food. European Food Safety Authority, 137, 1–20.

Eggers, R., & Pietsch, A. (2001). Technology I: Roasting. In R. Clarke, & O. Vitzthum (Eds.),Coffee recent developments (pp. 90–107). Oxford: Blackwell Science.

Esquivel, P., & Jiménez, V. M. (2012). Functional properties of coffee and coffee by-products. Food Research International, 46(2), 488–495, http://dx.doi.org/10.1016/j.foodres.2011.05.028.

Farah, A., de Paulis, T., Trugo, L. C., & Martin, P. R. (2005). Effect of roasting on the forma-tion of chlorogenic acid lactones in coffee. Journal of Agricultural and Food Chemistry,53(5), 1505–1513, http://dx.doi.org/10.1021/jf048701t.

Farah, A., Monteiro, M. C., Calado, V., Franca, A. S., & Trugo, L. C. (2006). Correlationbetween cup quality and chemical attributes of Brazilian coffee. Food Chemistry,98(2), 373–380, http://dx.doi.org/10.1016/j.foodchem.2005.07.032.

Fors, S. (1983). Sensory properties of volatile Maillard reaction products and related com-pounds. The Maillard reaction in foods and nutrition, Vol. 215. (pp. 185–286). AmericanChemical Society, 185–286.

Gianturco, M. A., Friedel, P., & Giammarino, A. S. (1964). The volatile constituents ofcoffee — III : The structures of two heterocyclic compounds and the synthesis oftetrahydrofuranones. Tetrahedron, 20(7), 1763–1772, http://dx.doi.org/10.1016/S0040-4020(01)99177-X.

Ginz, M., Balzer, H. H., Bradbury, A. G. W., & Maier, H. G. (2000). Formation of aliphaticacids by carbohydrate degradation during roasting of coffee. European Food Researchand Technology, 211(6), 404–410, http://dx.doi.org/10.1007/s002170000215.

Ginz, M., & Engelhardt, U. (2001). Identification of new diketopiperazines in roastedcoffee. European Food Research and Technology, 213(1), 8–11, http://dx.doi.org/10.1007/s002170100322.

Grosch, W. (1998). Flavour of coffee. A review. Nahrung, 42(6), 344–350.Grosch, W. (2000). Specificity of the human nose in perceiving food odorants. Paper

presented at the the Ninth Weurman Flavour Research Symposium. Garching: DeutscheForschungsanstalt für Lebensmittelchemie.

Grosch, W. (2001). Chemistry III: Volatile compounds. In R. J. Clarke, & O. G. Vitzthum(Eds.), Coffee: Recent developments (pp. 68–89). Oxford: Blackwell Science.

Grosch, W. (2001). Evaluation of the key odorants of foods by dilution experiments,aroma models and omission. Chemical Senses, 26(5), 533–545, http://dx.doi.org/10.1093/chemse/26.5.533.

Grosch,W., Czerny, M., Mayer, F., & Moors, A. (2000). Sensory studies on the key odorantsof roasted coffee. Caffeinated beverages, Vol. 754. (pp. 202–209). American ChemicalSociety.

Hayakawa, F., Kazami, Y., Wakayama, H., Oboshi, R., Tanaka, H., Maeda, G. O. U., et al.(2010). Sensory lexicon of brewed coffee for Japanese consumers, untrained coffeeprofessionals and trained coffee tasters. Journal of Sensory Studies, 25(6), 917–939,http://dx.doi.org/10.1111/j.1745-459X.2010.00313.x.

Higdon, J. V., & Frei, B. (2006). Coffee and health: A review of recent human research.Critical Reviews in Food Science and Nutrition, 46(2), 101–123, http://dx.doi.org/10.1080/10408390500400009.

Holscher, W., & Steinhart, H. (1992). Investigation of roasted coffee freshness with an im-proved headspace technique. Zeitschrift für Lebensmittel-Untersuchung und-Forschung,195(1), 33–38, http://dx.doi.org/10.1007/bf01197836.

Huang, L. F., Wu,M. J., Zhong, K. J., Sun, X. J., Liang, Y. Z., Dai, Y. H., et al. (2007). Fingerprintdeveloping of coffee flavor by gas chromatography–mass spectrometry and com-bined chemometrics methods. Analytica Chimica Acta, 588(2), 216–223, http://dx.doi.org/10.1016/j.aca.2007.02.013.

ICO (2011). Coffee value chain in selected importing countries. London: International CoffeeOrganization (ICO).

ICO (2012). Trends in coffee consumption in selected importing countries. London: Interna-tional Coffee Organization (ICO).

Kaufmann, H. P., & Gupta, A. K. S. (1964). Über die lipoide der kaffeebohne V: die triterpeneund kohlenwasserstoffe. Fette Seifen Anstrichm, 66, 461–466 (in Germany).

Kaufmann, H. P., & Hamsagar, R. S. (1962). Zur kenntnis der lipoide der kaffeebohne I:Über fettsäure-ester des cafestols. Fette Seifen Anstrichm, 64, 206–213 (inGermany).

Kuballa, T., Stier, S., & Strichow, N. (2005). Furan concentrations in coffee and coffeebeverages. Deutsche Lebensmittel-Rundschau, 101(6), 229–235.

324 W.B. Sunarharum et al. / Food Research International 62 (2014) 315–325

Kumazawa, K., & Masuda, H. (2003). Identification of odor-active 3-mercapto-3-methylbutyl acetate in volatile fraction of roasted coffee brew isolated by steamdistillation under reduced pressure. Journal of Agricultural and Food Chemistry,51(10), 3079–3082, http://dx.doi.org/10.1021/jf021190v.

Lancet, D. (1986). Vertebrate olfactory reception. [Review]. Annual Review of Neuroscience,9, 329–355, http://dx.doi.org/10.1146/annurev.ne.09.030186.001553.

Larsen, M., & Poll, L. (1992). Odour thresholds of some important aroma compounds instrawberries. Zeitschrift für Lebensmittel-Untersuchung und -Forschung, 195(2),120–123, http://dx.doi.org/10.1007/bf01201770.

Lawless, H. T., & Heymann, H. (2010). Physiological and psychological foundations ofsensory function Sensory Evaluation of Food: Principles and Practices. Food ScienceText Series. (pp. 19–56)New York: Springer New York.

Lean, M. E., & Crozier, A. (2012). Coffee, caffeine and health: What's in your cup?[Editorial]. Maturitas, 72(3), 171–172, http://dx.doi.org/10.1016/j. maturitas .2012.04.005.

Liu, Y., & Kitts, D. D. (2011). Confirmation that the Maillard reaction is the principlecontributor to the antioxidant capacity of coffee brews. Food Research International,44(8), 2418–2424, http://dx.doi.org/10.1016/j.foodres.2010.12.037.

Maga, J. (1975). The role of sulfur compounds in food flavor. Part I: thiazoles. CRC CriticalReviews in Food Technology, 6(241), 153–176.

Maier, H. G. (2005). Chapter 4. Roasting, 4.3 Changes produced by roasting. In A. Illy, & R.Viani (Eds.), Espresso coffee: The science of quality (pp. 191–197) (2nd ed.). London,UK: Elsevier Academic Press.

Marcone, M. F. (2004). Composition and properties of Indonesian palm civet coffee (KopiLuwak) and Ethiopian civet coffee. Food Research International, 37(9), 901–912,http://dx.doi.org/10.1016/j.foodres.2004.05.008.

Marcone, M. F. (2011). 4.56 — Origins and compositional analysis of novel foods:Kopi Luwak coffee and bird's nest soup. In M. -Y. Murray (Ed.), ComprehensiveBiotechnology (pp. 683–702) (2nd ed.). Burlington: Academic Press.

Mayer, F., Czerny, M., & Grosch, W. (1999). Influence of provenance and roast degree onthe composition of potent odorants in Arabica coffees. European Food Research andTechnology, 209, 242–250.

Mayer, F., Czerny, M., & Grosch, W. (2000). Sensory study of the character impact aromacompounds of a coffee beverage. European Food Research and Technology, 211(4),272–276, http://dx.doi.org/10.1007/s002170000169.

Mayer, F., &Grosch,W. (2001). Aroma simulation on thebasis of theodourant compositionof roasted coffee headspace. Flavour and Fragrance Journal, 16(3), 180–190, http://dx.doi.org/10.1002/ffj.975.

Mazzafera, P., & Padilha-Purcino, R. (2004). Post harvest processing methods and physio-logical alterations in the coffee fruit. Paper presented at the 20th International ScientificColloquium on Coffee, Bangalore, India.

Michishita, T., Akiyama, M., Hirano, Y., Ikeda, M., Sagara, Y., & Araki, T. (2010). Gaschromatography/olfactometry and electronic nose analyses of retronasal aroma ofespresso and correlation with sensory evaluation by an Artificial Neural Network.Journal of Food Science, 75(9), S477–S489, http://dx.doi.org/10.1111/j.1750-3841.2010.01828.x.

Milo, C., & Grosch, W. (1996). Studies on Character Impact Odorants of Coffee Brews. In P.Semmelroch, & W. Grosch (Eds.), Journal of Agricultural and Food Chemistry, 44. (pp.537–543) (unpublished).

Mombaerts, P. (2001). How smell develops. Natural Neuroscience, 4, 1192–1198.Mombaerts, P. (2001). The human repertoire of odorant receptor genes and pseudogenes.

Annual Review of Genomics and Human Genetics, 2, 493–510.Mori, E. E. M., Bragagnolo, N., Morgano, M. A., Anjos, V. D. A., Yotsuyanagi, K., Faria,

E. V., et al. (2003). Brazil coffee growing regions and quality of natural, pulpednatural and washed coffees. Food and Food Ingredients Journal of Japan, 208(1),416–423.

Murphy, C., Cain, W. S., & Bartoshuk, L. M. (1977). Mutual action of taste and olfaction.Sensory Processes, 1(3), 204–211.

Navarini, L., Cappuccio, R., Suggi-Liverani, F., & Illy, A. (2004). Espresso coffee beverage:Classification of texture terms. Journal of Texture Studies, 35(5), 525–541, http://dx.doi.org/10.1111/j.1745-4603.2004.35504.x.

Nijssen, L. M., Visscher, C. A., Maarse, H., Willemsense, L. C., & Boelens, M. H. (1996).Volatile compounds in food. Zeist, The Netherlands: TNO Nutrition and Food ResearchInstitute.

Oestreich-Janzen, S. (2010). 3.25— Chemistry of coffee. In M. Lew, & L. Hung-Wen (Eds.),Comprehensive natural products II, Vol. 3. (pp. 1085–1117). Oxford: Elsevier.

Ongo, E., Falasconi, M., Sberveglieri, G., Antonelli, A., Montevecchi, G., Sberveglieri, V., et al.(2012). Chemometric discrimination of Philippine civet coffee using electronic noseand gas chromatography mass spectrometry. Procedia Engineering, 47, 977–980,http://dx.doi.org/10.1016/j.proeng.2012.09.310.

Pérez-Martínez, M. n, Sopelana, P., de Peña, M. P., & Cid, C. n (2008). Application ofmultivariate analysis to the effects of additives on chemical and sensory qualityof stored coffee brew. Journal of Agricultural and Food Chemistry, 56(24), 11845–11853,http://dx.doi.org/10.1021/jf802146v.

Petracco, M. (2001). Beverage preparation: brewing trends for the newmillenium. In R. J.Clarke, & O. G. Vitzthum (Eds.), Coffee: Recent Developments (pp. 140–164). Oxford:Blackwell Science.

Poltronieri, Y., Martinez, H. E., & Cecon, P. R. (2011). Effect of zinc and its form of supplyon production and quality of coffee beans. Journal of the Science of Food andAgriculture, 91(13), 2431–2436, http://dx.doi.org/10.1002/jsfa.4483.

Ponte, S. (2002). The ‘Latte revolution’? Regulation, markets and consumption in theglobal coffee chain. World Development, 30(7), 1099–1122, http://dx.doi.org/10.1016/S0305-750X(02)00032-3.

Rawson, N. E., & Li, X. (2004). 3. The cellular basis of flavour perception: taste and aroma.In A. J. Taylor, & D. D. Roberts (Eds.), Flavor Perception (pp. 75). Oxford, UK: BlackwellPublishing Ltd.

Reineccius, G. (1995). The Maillard reaction and coffee flavor. Paper presented at the the16th ASIC Colloquium Kyoto.

Resurreccion, A. V. A. (1988). Application of multivariate methods in food quality evalua-tion. Food Technology, 42(11), 128–136.

Ribeiro, J. S., Augusto, F., Salva, T. J. G., & Ferreira, M. M. C. (2012). Prediction models forArabica coffee beverage quality based on aroma analyses and chemometrics.Talanta, 101(0), 253–260, http://dx.doi.org/10.1016/j.talanta.2012.09.022.

Ribeiro, J. S., Augusto, F., Salva, T. J. G., Thomaziello, R. A., & Ferreira, M. M. C. (2009).Prediction of sensory properties of Brazilian Arabica roasted coffees by headspacesolid phase microextraction–gas chromatography and partial least squares. AnalyticaChimica Acta, 634(2), 172–179, http://dx.doi.org/10.1016/j.aca.2008.12.028.

Ribeiro, J. S., Teófilo, R. F., Augusto, F., & Ferreira, M. M. C. (2010). Simultaneous optimiza-tion of the microextraction of coffee volatiles using response surface methodologyand principal component analysis. Chemometrics and Intelligent Laboratory Systems,102(1), 45–52, http://dx.doi.org/10.1016/j.chemolab.2010.03.005.

Risticevic, S., Carasek, E., & Pawliszyn, J. (2008). Headspace solid-phase microextraction–gas chromatographic–time-of-flight mass spectrometric methodology for geographi-cal origin verification of coffee. Analytica Chimica Acta, 617(1–2), 72–84, http://dx.doi.org/10.1016/j.aca.2008.04.009.

Salo, P. (1970). Determining the odor threshold for some compounds in alcoholicbeverages. Journal of Food Science, 35, 95–99.

Sanz, C., Maeztu, L., Zapelena, M. J., Bello, J., & Cid, C. (2002). Profiles of volatilecompounds and sensory analysis of three blends of coffee: influence of differentproportions of Arabica and Robusta and influence of roasting coffee with sugar.Journal of the Science of Food and Agriculture, 82(8), 840–847, http://dx.doi.org/10.1002/jsfa.1110.

Savonitti, O. (2005). Chapter 6. Storage and packaging, 6.1.3 Other physico-chemicalchanges. In A. Illy, & R. Viani (Eds.), Espresso Coffee: The science of quality (pp.241–245). London, UK: Elsevier Academic Press.

Scheidig, C., Czerny, M., & Schieberle, P. (2007). Changes in key odorants of raw coffeebeans during storage under defined conditions. Journal of Agricultural and FoodChemistry, 55(14), 5768–5775, http://dx.doi.org/10.1021/jf070488o.

Semmelroch, P., & Grosch, W. (1995). Analysis of roasted coffee powders and brews bygas chromatography–olfactometry of headspace samples. LWT — Food Science andTechnology, 28(3), 310–313, http://dx.doi.org/10.1016/s0023-6438(95)94411-7.

Semmelroch, P., & Grosch, W. (1996). Studies on character impact odorants of coffeebrews. Journal of Agricultural and Food Chemistry, 44(2), 537–543, http://dx.doi.org/10.1021/jf9505988.

Semmelroch, P., Laskawy, G., Blank, I., & Grosch, W. (1995). Determination of potentodourants in roasted coffee by stable isotope dilution assays. Flavour and FragranceJournal, 10(1), 1–7, http://dx.doi.org/10.1002/ffj.2730100102.

Seo, H. -S., Lee, M., Jung, Y. -J., & Hwang, I. (2009). A novel method of descriptive analysison hot brewed coffee: time scanning descriptive analysis. European Food Research andTechnology, 228(6), 931–938, http://dx.doi.org/10.1007/s00217-009-1006-3.

Shibamoto, T. (1983). Heterocyclic compounds in browning and browning/nitrite modelsystems: Occurrence, formation mechanisms, flavor characteristics and mutagenicactivity. In G. Charalambous, & G. Inglett (Eds.), Instrumental Analysis of Foods, Vol. I.(pp. 229–278). New York: Academic Press.

Silwar, R., Kamperschroer, H., & Tressl, R. (1987). Gas chromatographic–mass spectrometricinvestigation of coffee aroma. Quantitative determination of steam-volatile aromaconstituents. Chemie Mikrobiologie Technologie der Lebensmittel, 10(5–6), 176–187.

Spadone, J. C., Takeoka, G., & Liardon, R. (1990). Analytical investigation of Rio off-flavor ingreen coffee. Journal of Agricultural and Food Chemistry, 38(1), 226–233, http://dx.doi.org/10.1021/jf00091a050.

Speer, K., & Kölling-Speer, I. (2001). Lipids. In R. Clarke, & O. G. Vitzthum (Eds.), Coffee —Recent Developments (pp. 33–49). Oxford: Blackwell Science.

Stegen, G. v. d, & Duijn, J. v (1987). Analysis of normal organic acids in coffee. Paper pre-sented at the 12th International Scientific Colloquium on Coffee, Montreux 29.6.–3.7.1987. Paris, Paris: Imprimerie Jouve.

Taylor, A. J., & Roozen, J. P. (1996). Volatile flavor release from foods during eating. CriticalReviews in Food Science and Nutrition, 36(8), 765–784, http://dx.doi.org/10.1080/10408399609527749.

Teixeira, A. A., Brando, C. H. J., Thomaziello, R. A., & Teixiera, R. (2005). Chapter 3.The raw bean, 3.3. Processing of the harvest. In A. Illy, & R. Viani (Eds.), Espressocoffee: The science of quality (pp. 91–95) (2nd ed.). London, UK: Elsevier Aca-demic Press.

Teixeira, R., Teixeira, A. A., & Brando, C. (2005). Chapter 3. The raw bean, 3.8 Classification:Physical and sensorial analysis. In A. Illy, & R. Viani (Eds.), Espresso coffee: The scienceof quality (pp. 134–141) (2nd ed.). London, UK: Elsevier Academic Press.

Toci, A. T., & Farah, A. (2008). Volatile compounds as potential defective coffee beans'markers. Food Chemistry, 108(3), 1133–1141, http://dx.doi.org/10.1016/j.foodchem.2007.11.064.

Tressl, R. (1980). Formation of aroma compounds by the Maillard reaction. Bildung vonAromastoffen durch Maillard reaktion (In the 9th International Scientific Colloquium onCoffee [see FSTA (1982) 14 11H1751].) (pp. 55–76).

Tressl, R., & Silwar, R. (1981). Investigation of sulfur-containing components in roastedcoffee. Journal of Agricultural and Food Chemistry, 29(5), 1078–1082, http://dx.doi.org/10.1021/jf00107a045.

Urgert, R., van der Weg, G., Kosmeijer-Schuil, T. G., van de Bovenkamp, P., Hovenier, R., &Katan, M. B. (1995). Levels of the cholesterol-elevating diterpenes cafestol andkahweol in various coffee brews. Journal of Agricultural and Food Chemistry, 43(8),2167–2172, http://dx.doi.org/10.1021/jf00056a039.

Vitzthum, O. G. (1976). Chemie und bearbeitung des kaffees. In O. Eichler (Ed.), Kaffee undCoffein (pp. 3–64) (2nd ed.). Berlin: Springer (in Germany).

Vitzthum, O. G., Weissmann, C., Becker, R., & Kohler, H. S. (1990). Identification of anaroma key compound in Robusta coffees. Cafe Cacao Tea, 34, 27–36.

325W.B. Sunarharum et al. / Food Research International 62 (2014) 315–325

Wagner, R., Czerny, M., Bielohradsky, J., & Grosch,W. (1999). Structure–odour–activity re-lationships of alkylpyrazines. Zeitschrift für Lebensmitteluntersuchung und -ForschungA, 208(5–6), 308–316, http://dx.doi.org/10.1007/s002170050422.

Wasserman, G. (1992). Coffee. In Kirk-Othmer (Ed.), (4th ed.)Kirk-Othmer EncyclopediaOf Chemical Technology, Vol. 6, John Wiley & Sons, Inc.

Weers, M., Balzer, H., Bradbury, A., & Vitzthum, O. G. (1995). Analysis of acids in coffee bycapillary electrophoresis. Paper presented at the 16th International Scientific Colloquiumon Coffee, Kyoto 9.4.–14.4.1995. Paris, Kyoto: Imprimerie Jouve.

Wold, S., & Sjostrom, M. (1998). Chemometrics, present and future success. Chemometricsand Intelligent Laboratory Systems, 44(1–2), 3–14.

Yeretzian, C., Pascual, E. C., & Goodman, B.A. (2012). Effect of roasting conditions andgrinding on free radical contents of coffee beans stored in air. Food Chemistry,131(3), 811–816, http://dx.doi.org/10.1016/j.foodchem.2011.09.048.