Access to this full-text is provided by Wiley.
Content available from International Journal of Food Science & Technology
This content is subject to copyright. Terms and conditions apply.
Review
Effects of postharvest processing on aroma formation in roasted
coffee –a review
Xiaotong Cao,
1
Hanjing Wu,
1
Claudia G. Viejo,
1
Frank R. Dunshea
1,2
& Hafiz A. R. Suleria
1
*
1 Faculty of Veterinary and Agricultural Sciences, School of Agriculture and Food, The University of Melbourne, Parkville 3010,VIC,
Australia
2 Faculty of Biological Sciences, The University of Leeds, Leeds, UK
(Received 1 December 2022; Accepted in revised form 13 December 2022)
Summary Postharvest processing of coffee cherries significantly influences sensory characteristics and commercial
values. Aroma is one of the critical elements in product qualification and differentiation of coffees from
different origins, roasting levels and brewing methods. Except for primary coffee volatile organic com-
pounds (VOCs) (furans and pyrazines), which are generated during postharvest processing (dry, honey,
wet processing and roasting), aldehydes, ketones, phenols, sulphur compounds and others could also con-
tribute to the complex coffee flavour. Desirable flavour requires a balance between pleasant and defective
VOCs. This review comprehensively discussed the mechanisms of conventional and novel postharvest pro-
cessing of coffee beans, their impact on the sensorial profile of green and roasted coffee, and the composi-
tion, generation and analysis techniques of coffee VOCs. This review shows the feasibility of GC–MS and
electronic nose (E-nose) in coffee VOCs and flavour detection, meanwhile building a comprehensive link-
age between postharvest processing and coffee sensory characteristics.
Keywords Aroma, coffee flavour, e-nose, GC–MS, postharvest processing, VOCs.
Introduction
Nowadays, coffee is the second-largest commodity in
the global market after crude oil (Haile & Kang, 2019;
Zakidou et al., 2021). According to the International
Coffee Organisation (ICO), 166.35 million (60 kg) bags
of roasted coffee were consumed in 2020/2021, with a
1.0% increase compared to 2017/2018. It is also the
second most consumed beverage after tea owing to its
energising function and a broad spectrum of aroma
and flavour. The demand for speciality coffee, made
from certain varietals or specific processing methods,
has been rising lately (Zakidou et al., 2021). For
instance, the Geisha varietal is favoured mainly
because of its floral and sweet taste, while wet process-
ing could generate more outstanding coffee sensory
characteristics than dry processing. A total of 43% of
consumers chose speciality coffee in 2022, with an
increase of 20% compared to 2021, and reached the
highest level to date (da Silva Portela et al., 2022).
Generally, the genetic varietal of coffee plants and
the geographic conditions could directly or indirectly
influence the size, shape and chemical compositions
(sugar, fat, protein) of raw coffee beans (Kitzberger
et al., 2016). The elements of geography and environ-
ment generally include geographic altitude, light expo-
sure, projected temperature, water resource and
precipitation regime, cultivation methods and pest and
disease management (Ahmed et al., 2021). Some influ-
ential aromatic makers found in roasted coffee, such
as 2,3-butanedione, 2,3-pentanedione, 2-methylbutanal
and 2,3-dimethyl pyrazine, could be significantly
impacted by geographic variations (de Toledo
et al., 2017; Dryahina et al., 2018). Increased light
exposure could also reduce the phenolic metabolites
and increase lipids content (Muschler, 2001; Delaroza
et al., 2017). Decazy et al.(2003) observed that more
intensive light exposure might reduce coffee sensory
quality due to the lack of aroma. L€
aderach
et al.(2017) also reported an improved body of
brewed coffee with expanded shade on the farm.
Besides, crop quality and productivity can be impacted
by the nutrients available in the soil. The potassium
fertilisation practised on the farm is believed to be an
essential controller to regulate many flavour-related
components, such as sugars, chlorogenic acids and
phenols (Clemente et al., 2015; Vinecky et al., 2017).
Although over 100 species have been identified
worldwide, only two are most commercially valued,
namely Coffea arabica and Coffea robusta (Canephora)
*E-mail: hafiz.suleria@unimelb.edu.au
International Journal of Food Science and Technology 2023, 58, 1007–1027
doi:10.1111/ijfs.16261
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
This is an open access article under the terms of the Creative Commons Attribution License, which permits use,
distribution and reproduction in any medium, provided the original work is properly cited.
1007
(Davis et al., 2006). Nowadays, 75% of the total coffee
production in the global market is from Arabica spices
due to their higher chemical complexity than Robusta,
which is the potential to make roasted coffee more
favourable (Toledo et al., 2016). The relatively higher
concentration of 2-methylisoborneol in Robusta is con-
sidered as a factor of its typical earthy note (Kny-
sak, 2017). Another critical role of postharvest
practice (fermentation and roasting) is regulating the
conversion of these resources into distinguishing VOCs
in the final cup (Toledo et al., 2016; de Melo Pereira
et al., 2019). The choice of washed (wet), dry (natural)
or semi-dry (honey) processing is considered one of
the most quality-differentiative steps in the coffee pro-
duction (Sunarharum et al., 2014; de Melo Pereira
et al., 2019) since it decides the presence of de-pulping
and mucilage, resulting in different chemical composi-
tions, which functions as the aromatic precursors in
the following steps (Gonzalez-Rios et al., 2007a,
2007b). This is discussed in detail in the later section.
Many factors need to be considered when selecting the
optimal method for a specific roasting batch of coffee,
including coffee species, targeted flavour, farm scale
and expected cost (Lee et al., 2013a; Toledo
et al., 2016).
Coffee cherry, the raw and unroasted coffee beans,
usually possess undesirable peasy off-odours because
of alkyl-methoxy pyrazines (such as 3-isobutyl-2-
methoxypyrazine) (Flambeau et al., 2017). Roasting
can suppress these compounds and convert them into
a pleasantly roasty aroma (Sunarharum et al., 2014).
Most of the volatile compounds (VOCs) related to the
coffee aroma, including pyrazines, furans, aldehydes,
ketones, phenolic compounds and sulphur-containing
compounds, are principally produced from Maillard
reaction, caramelisation, Strecker degradation and
pyrolysis during roasting (Akiyama et al., 2003;
Franca et al., 2005; Toledo et al., 2016; Zakidou
et al., 2021). In roasted coffee, furans and pyrazines
are arguably the most quantitative compounds.
Sulphur-containing compounds (despite their relatively
lower contents in roasted coffee) and pyrazines could
be the most influential to the sensory quality (Sunar-
harum et al., 2014). The mechanisms of producing
these compounds are interactive and highly complex,
contributing to various concentrations and unique sen-
sory properties of brewed coffee. The health benefits of
coffee consumption are always associated with its
antioxidant capacity (the ability of retarding oxidative
reactions), which is believed to be achieved mainly by
phenolics, such as flavonoids and tannins (Haile
et al., 2020; Chindapan et al., 2022). Ludwig
et al.(2014) reported that heterocyclic VOCs (furans,
pyrroles and thiophenes) showed the highest antioxi-
dant activities among the detected VOCs in roasted
coffee. Chlorogenic acid, one of the most abundant
esters found in coffee, is reported as a good source of
antioxidants consumed from dietary sources (Buffo &
Cardelli-Freire, 2004). Nevertheless, the presence of
aromatic furans is a concern over the possible negative
impact on human liver function (EFSA Panel on Con-
taminants in the Food Chain (CONTAM) et al., 2017,
Haile et al., 2020). Long-term observational investiga-
tion on the health effect of coffee consumption is still
required. Due to the volatility of those compounds,
analytic detection also requires high sensitivity, kinet-
ics monitoring and trace detection (Chang
et al., 2016).
Coffee provides an intricate blend of various fla-
vours after postharvest processing, roasting and brew-
ing, which develop a range of sensory experiences.
Specific sensory characteristics differentiate the types
of coffee varietals, processes and roasts (Stroe-
baek, 2013). Generally, the human sense of smell can
identify over 10 000 different odorants via retronasal
and orthonasal systems, while the sense of taste can
only distinguish five official tastes (sweet, sour, bitter,
salty and umami) and two recently acknowledged
basic tastes (starch and fat) (Petracco, 2001; Keast &
Costanzo, 2015; Besnard et al., 2016; Low
et al., 2017). Therefore, the coffee flavour plays a
prominent role in the coffee sensory qualification as it
is the mix of aromas, tastes and trigeminal sensations
(Stevenson, 2012; Sunarharum et al., 2014). The
odours or aromas chiefly come from the VOCs pro-
duced during postharvest preparation and roasting,
which will be particularly reviewed in the following
sections. This review aimed to address the effect of
coffee postharvest process and roasting conditions on
the VOCs composition and the aroma formation in
roasted coffee. The application of traditional (GC-MS)
and novel sensor technologies (Electronic nose) to
assess VOC is discussed. Those two methods have
been widely used and validated for chemical and sen-
sory analysis in the food and beverage industry.
Besides, other methods such as Nuclear Magnetic
Resonance (NMR), Proton-transfer Reaction Mass
Spectrometry (PTR-MS), and Gas chromatography-
olfactory (GC-O) have also been used in the industry.
Postharvesting processing
Postharvest processing with the botanical structure
changes in coffee seeds is illustrated in Fig. 1. The coffee
cherry pulp (mesocarp) in the middle is rich in nutrients
such as carbohydrates (pectin, glucose and fructose),
fat, protein and other substances, which could be
the resource of aromatic compounds produced in the
following processing (Janissen & Huynh, 2018; de Melo
Pereira et al., 2019). The endocarp is a polysaccharide
layer mainly composed of cellulose, hemicellulose, lig-
nin and ashes (Esquivel & Jimenez, 2012). Noticeably,
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technolo gy 2023
Postharvest practice and coffee aroma X. Cao et al.1008
Figure 1 Illustration of three post-harvesting processing of coffee cherries.
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technology 2023
Postharvest practice and coffee aroma X. Cao et al. 1009
silver skin is also abundant in polysaccharides and
monosaccharides, proteins, polyphenols and other
minor compounds (Janissen & Huynh, 2018; de Melo
Pereira et al., 2019).
The coffee supply chain starts with coffee cherry
harvesting by handpicking or mechanical stripping (de
Melo Pereira et al., 2019). Proper maturity guarantees
raw coffee beans not only an ideal content of desirable
VOCs but also a reduced phenolic concentration,
which could lower the unwanted astringency and
improve the organoleptic quality of final products
(Viejo et al., 2020). The slight environment and prac-
tising changes, including the climate, humidity, tem-
perature and manufacturing, would significantly affect
the composition and quality of roasted coffee beans
from different batches. Therefore, the postharvest pro-
cessing of coffee beans is supposed to be conducted
immediately to avoid unwanted germination, indige-
nous fermentation and mould generation (de Melo
Pereira et al., 2019). Simple washing, separating and
sorting needs to be conducted primarily to remove the
impurities (dust, stone, etc.) and defective beans. Sub-
sequently, selected beans will be processed in different
procedures (dry, semi-dry and wet processing) to
achieve the optimal chemical composition (moisture,
sugars, lipids, etc.) according to additional quality
requirements and thus preserve coffee profitability.
Conventional (dry, honey and wet) processing of cof-
fee cherries processing and novel processing (carbonic
maceration) could critically influence the pathway and
degree of de-pulping and de-mucilage, which is dis-
cussed in detail in the following content (de Melo Per-
eira et al., 2019; Junior et al., 2021). Those
modifications can trigger and regulate the chemical
and microbiological fermentation of beans before
roasting, producing aromatic precursors (de Melo Per-
eira et al., 2019). Stronger off-aromas were found in
green coffee with incomplete mucilage removal than
those with entirely de-mucilage, probably because of
the unwanted spontaneous fermentation, which could
induce the generation of unwanted substances during
roasting (Gonzalez-Rios et al., 2007a,2007b). How-
ever, since the microflora population present in coffee
is dense and diverse, the fermentation could be incon-
sistent and unpredictable, which remains to be less elu-
cidated.
Natural/dry processing
Natural processing is commonly used in the regions
with limited access to water, such as Ethiopia and Bra-
zil (de Melo Pereira et al., 2019), where the intact cof-
fee cherries (with mucilage) are dried by sunshine after
washing and de-pulping. It depends more on the regio-
nal climate, including rainfall (humidity), temperature
and atmospheric conditions (Toledo et al., 2016). With
the mucilage intact, the natural process gives the
roasted coffee a rich body and sweet characteristic
since abundant polysaccharides and minerals in muci-
lage, which are the essential precursors of mostly
VOCs. The quality of final coffee products may not be
uniform because of the dependence on weather factors
such as rainfall, sunlight exposure time and tempera-
ture. One of the corrective actions conducted by the
coffee farmers in Vietnam is to exclusively pick ripe
fruits (up to 98%) and process them in a fast and
meticulous way, leading to higher coffee quality and
stability (Le et al., 2020).
Washed/wet processing
Soaking in water is aimed to remove all soft fruit resi-
due of the coffee cherry. The wet environment facili-
tates microorganism growth and fermentation,
especially lactic acid bacteria (LAB), such as Leuconos-
toc mesenteroides,Lactobacillus plantarum and Lb. bre-
vis, yeasts (e.g., Pichia guilliermondii,P. anomala
Kluvyeromyces marxianus and Saccharomyces cere-
visiae) and mould (Vilela et al., 2010; Evangelista
et al., 2014; de Melo Pereira et al., 2019). Water addi-
tion could increase the richness and community of
bacteria and fungi (da Silva et al., 2022). Using specific
microorganisms such as Saccharomyces cerevisiae and
Lactobacillus rhamnosus under controlled conditions is
a vital regulator in coffee quality improvement via
microbial process (da Silva et al., 2022; Krajangsang
et al., 2022). Washed coffee tends to have a cleaner
and fruity aroma than natural ones. The reasons can
be (1) less sugar (fructose and glucose) metabolism
involved due to the wet anoxic environment and the
removal of fruit pulp and skin; (2) higher content of
free amino acids because of the increased protein
hydrolysis (Gonzalez-Rios et al., 2007b; de Melo Per-
eira et al., 2019). Although the wet process might con-
tribute to the VOCs quantity in the final coffee due to
higher microbial diversity, the washed process faces a
higher uncertainty on the VOCs quality because of the
unknown microbial activity and uncontrolled end-
point, which can cause either a positive or negative
impact on the final cup quality (de Melo Pereira
et al., 2019). Excessive defective VOCs, such as sul-
phur and phenolic compounds, aminobutyric acid and
acetic acid, can be generated if overfermentation due
to prolonged exposure to drought stress, which may
induce unfavourable bitterness or sourness in the final
coffee brew (Toledo et al., 2016; de Melo Pereira
et al., 2019). Recently, the standardisation of fermen-
tation by selecting microbial strains has been high-
lighted to solve this problem while the operational
feasibility remains consideration due to the compli-
cated microbial metabolites (Zhang et al., 2019).
Besides, proper waste treatment access to sustain the
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technolo gy 2023
Postharvest practice and coffee aroma X. Cao et al.1010
ecosystem around coffee plants is necessary for farmers
with mostly wet-processed coffee production fields
(Nguyen & Sarker, 2018).
Honey processing
Honey processing combines wet and dry processing,
removing the pulp but keeping mucilage before drying.
Thus, impurities from pulp could be avoided and poten-
tially allows the well-controlled fermentation of the
mucilage around the coffee seeds (de Melo Pereira
et al., 2019). The glucose and fructose contents in the
semi-dried coffee beans were found in between wet (low-
est) and dry (highest) processed coffee beans (de Melo
Pereira et al., 2019). Semi-dry processed coffee bean was
reported with higher concentrations of TBARS (thio-
barbituric acid reactive substances), which indicated
higher lipid oxidation. Additionally, semi-dried coffee
beans would show a higher intensity of ‘rested coffee fla-
vour’, defined as a woody taste and pale flavour notes,
after 15-month-storage (Rend
on et al., 2014). Based on
the degree of mucilage removal, the semi-drying/honey
process can be divided into subclasses: white honey
process (80–100% mucilage removal), yellow honey
process (50–75% mucilage removal), red honey process
(0–50% mucilage removal) and black honey process
(the least amount of mucilage removal). The honey
process includes moderate water usage and retains a
certain level of sweetness and cleanness (Toledo
et al., 2016). As a combined method compromising
between washed and dry processes, both benefits and
limitations of these two methods should be considered
in the honey process.
Novel technique - carbonic maceration
The carbonic maceration (CM) technology was ini-
tially used in the wine-making industry to upgrade the
aroma and body and save processing time. It was
introduced to the coffee industry recently and is
designed to leave the intact coffee beans sealed in an
enclosed environment, with expelled O
2
and filled with
CO
2
. It aims to switch aerobic respiration to anaero-
bic respiration, encouraging microbial fermentation
within the fruits (Gonz
alez-Arenzana et al., 2020;
Junior et al., 2021). Although this method has been
rapidly developed, few scientific studies have been
published, and it remains an ‘experimental processing
method’. Only one study has evaluated the possibility
of the CM application in coffee postharvest processing
(Junior et al., 2021). The authors observed a signifi-
cant impact of the CM method on the sensorial,
chemical and microbiological profiles of their Arabia
coffee sample. However, this area remains further elu-
cidated.
Chemical reactions during coffee roasting
Coffee roasting is a complex thermal process, including
colour browning, moisture reduction (from 10% to
30%) and structural porosity improvement. Numerous
VOCs could be generated via multiple chemical reac-
tions, which are essential to the coffee aroma (Ruosi
et al., 2012). Zakidou et al.(2021) found a total of 138
VOCs in 10 types of roasted coffee formed mostly by
Maillard reactions (30–40%), followed by Strecker
degradation (16–18%), and other responses (30–37%)
(pyrolysis and fragmentation). Around 300 VOCs have
been found in raw coffee beans, while over 850 in
roasted coffee, which determines the final cup profile
(Flament, 2001).
Roasting conditions significantly impact those chem-
ical changes, VOCs yields and sensory quality.
Gonzalez-Rios et al.(2007b) found that light-roasted
(240 °C for 6 min) coffee had the lowest concentration
of total VOCs among the three roasting degrees, com-
pared to the medium and dark roasting, which was
240 °C for 7 and 9 min, respectively. Poor-controlled
roasting can also cause the further degradation of
some desirable VOCs and encourage the formation of
undesirable compounds because of intensive pyrolytic
reactions (Toledo et al., 2016). Ethyl acetate and
methyl acetate with fruity, wine-like, grape-like notes
and ethanol with a sweet note, could be reduced when
the roasting degree increases. Meanwhile, pyridine
(sour and fishy notes) would be raised (Moon &
Shibamoto, 2009).
Maillard reaction and Strecker degradation
Maillard reaction
During roasting, the Maillard reaction (non-enzymatic
browning) could occur when the temperature reaches
140 °C, where carbohydrates and amino acids in coffee
beans could start a cascade of reactions along with the
development of many aroma-related compounds. As
shown in Fig. 2, three stages are involved: (1) the reac-
tions between reducing sugar and amino acids, generat-
ing N-substituted glycosylamine and water; (2) the
Amadori rearrangement (isomerisation) of the glycosy-
lamine produced in the first step, generating ketoamines
(Amadori product); (3) the ketoamines conversion via
several pathways into fission products and reductones
(under alkaline condition), hydroxymethylfurfural (un-
der acidic condition) and melanoidins (brown pigments)
(Moreira et al., 2012; de Oliveira et al., 2016).
Temperature, humidity, pH, the presence of certain
metals, and the contents of reducing sugar could affect
Maillard reactions (Toledo et al., 2016). A wide range
of VOCs, including furans, pyrazines and pyrroles,
would be generated and contribute to the various cof-
fee aromas (roasted, malty, nutty, bitter and burnt
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technology 2023
Postharvest practice and coffee aroma X. Cao et al. 1011
aromas) with a significant impact on its sensory qual-
ity (Moreira et al., 2012). A higher roasting degree
could strengthen Maillard reactions and eventually
contribute to a higher yield of pyridines and pyrroles,
leaving dark-roasted coffee with a chocolate-like and
nutty flavour (Moon & Shibamoto, 2009). N-
methylpyridine concentration could also start to build
up and show an increasing trend from the medium
roasting level and reach the highest level at the dark
roasting point (Hu et al., 2020). Furfural and deriva-
tives (sweet and caramel-like) are found at a relatively
higher level in mild-roasted coffee but with a decrease
in their higher-roasted counterparts (Moreira
et al., 2012). Poor Maillard reactions may also cause
inadequate generation of VOCs, especially those with
nutty, coconut- and chocolate-like features, thus low-
ering the roasted coffee (Vel
asquez et al., 2019).
Strecker degradation
Strecker degradation is considered a subset of Maillard
reactions because of the close linkage. Some of the
carbonyl derivatives generated from the Maillard reac-
tion may participate in Strecker degradation with free
amino acids, the building blocks of proteins in coffee
beans. Two steps are involved: (1) the oxidisation
between amino acids (reductant) and carbonyl deriva-
tives (oxidant), and (2) the break-down of the resulting
molecules into aldehydes, ammonia and carbon
dioxide in the presence of water (Toledo et al., 2016).
The aldehyde group (acetaldehyde and hexanal) is a
crucial part of coffee aroma, which gives coffee a fruity
fragrance and indicates coffee freshness (Buffo &
Cardelli-Freire, 2004). Besides, the CO
2
created in
Strecker degradation can increase the internal pressure,
expand beans, make them crack fields and further
assist the volatilisation of coffee aroma (Wang &
Lim, 2014; Toledo et al., 2016).
Caramelisation
Similarly, caramelisation is also a non-enzymatic
browning, while it occurs in the absence of amino
acids at a relatively higher temperature. As shown in
Fig. 3, it focuses on the thermal decomposition of
more complex carbohydrates (such as sucrose) in cof-
fee beans into smaller sugar molecules (fructose and
glucose) with higher water-solubility at a temperature
around 170 °C (Knopp et al., 2006). The loss of water
molecules from individual sugars is also known as a
condensation step. Subsequently, maltol with a cara-
mel taste, acetic acid with sourness and furans with
nutty notes could be produced, resulting in a compli-
cated coffee flavour (Knopp et al., 2006). Therefore,
this step is one of the primary sources of sweet, cara-
mel and almond-like aroma, which could improve the
perceived sweetness of final brewed coffee.
Figure 2 Three stages of Maillard reaction.
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technolo gy 2023
Postharvest practice and coffee aroma X. Cao et al.1012
Pyrolysis
Pyrolysis reactions happen once the beans are heated
to 190 °C. Sucrose pyrolysis can lead to caramelisa-
tion, which darkens the bean’s colour and produce
volatile carbocyclic compounds at 190–200 °C (Mon-
tavon et al., 2003). Roasting can cause 3-
methylbutanoyl disaccharides (3MDs) pyrolysis in cof-
fee. The products mainly consist of a 3-
methylbutanoic acid, which may enhance the richness
of coffee flavour (Iwasa et al., 2015; Iwasa
et al., 2021). Protein also undergoes pyrolysis and
releases aromatic compounds such as alanine and
asparagine (Dong et al., 2015). Importantly, coffee
lipids/oils pyrolysis can provide strong coffee flavour
and antioxidant ability as a functional food and bever-
age (Vila et al., 2005; Ferrari et al., 2010). The lipid
fraction in coffee beans, located in the endosperm, is
primarily composed of triacylglycerol, sterols, coffea-
diol, arabitol and tocopherols (de Melo Pereira
et al., 2019). Although those compounds are found in
small thresholds, they can influence central carbon and
nitrogen metabolism. Thermal roasting can lead to the
self-oxidation of triacylglycerol and lipids, producing
alcohols and ketones. Besides, the condensation
between fatty acids and alcohol molecules can generate
many esters, which have a pivotal role in fruity and
flower-like notes in the coffee sensory (de Melo Pereira
et al., 2019). Roasting does not trigger dramatic
changes in coffee lipid fraction, which may be related
to the presence of lipid-soluble products from the
Maillard reactions (Amarowicz, 2009), while the roast-
ing conditions can influence the lipid pyrolysis pro-
gress. The total polyunsaturated fatty acids (PUFA)
content was found to be the highest (45.5%) under
roasting at 210 °C with high moisture of heating air
among all conditions applied (Budryn et al., 2012).
Other reactions
Based on sugars, dehydration, decarboxylation, frac-
tionation, isomerisation and polymerisation of carbo-
hydrates, are also believed to contribute to coffee
flavour development (Wu et al., 2021). During roast-
ing, aroma-related reactions are not only based on
sugars and lipids in the coffee matrix but also on pro-
tein molecules, such as loss of protein nitrogen, denat-
uration of protein and degradation of specific amino
acids (proline and hydroxy amino acids), which may
lead to undesirable sulphur compounds with cabbage-
or onion-flavour in the final product. Trigonelline, for
instance, is an essential flavour precursor and its
degradation products cover a range of pyridine and
pyrroles (Montavon et al., 2003). Other nitrogen-
containing compounds such as alkaloids and volatile
acids (acetic, butanoic and propanoic acids) are also
detected in high amounts in green coffee beans, which
will be decomposed during roasting and produce influ-
ential sensory metabolites (pyridines and pyrroles)
(Sunarharum et al., 2014).
Major volatile compounds in roasted coffee
Furans
The most volatile compound contributing to the coffee
flavour is furan derivatives (25–41%) with sweet, bread-
like and caramel aromas, especially 5-methylfurfural,
Figure 3 The caramelisation of sucrose during coffee roasting.
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technology 2023
Postharvest practice and coffee aroma X. Cao et al. 1013
furfural and 2-furanmethanol (Flament, 2001; Moon &
Shibamoto, 2009). It could be generated by Maillard
reactions, caramelisation and thermal-oxidative degra-
dation of the polyunsaturated fatty acids (Moreira
et al., 2012; Zakidou et al., 2021). Therefore, the washed
coffee generally has a lower content of furans with lower
sweetness and chocolate flavour after roasting since the
removal of mucilage causes less sugar to be available as
substrates for these reactions (Gonzalez-Rios
et al., 2007a; Zakidou et al., 2021). 5-Methylfurfural is
one of the most important contributors to the desirable
caramel aroma of roasted coffee. Wet processing has
been confirmed to improve the volatilisation of more 5-
methylfurfural in roasted coffee beans due to the muci-
lage removal (Gonz
alez-Arenzana et al., 2020). Besides,
furan and derivatives are the most influenced by burn-
ing conditions, especially 5-hydroxymethylfurfural sig-
nificantly, which decreased from 26.8% (230 °C
12 min) to 0% (250 °C 21 min) and furfural from
19.9% to 2% (Moon & Shibamoto, 2009). This decrease
may result from VOCs’ more significant decomposition
and polymerisation as roasting degrees rise. However,
other furans identified in coffee and formed during
roasting, such as 3-Methylfuran are considered poten-
tially carcinogenic and toxic and are still under investi-
gation by the International Agency for Research on
Cancer (IARC) (Becalski et al., 2016; Rahn & Yeret-
zian, 2019; Gonzalez Viejo et al., 2021b).
Pyrazines
Pyrazines are the second most abundant VOCs (25–
39%) in roasted coffee generated through complex
sugar-amino-acid interactions, known for their hazel-
nut aroma in the coffee brew (Zakidou et al., 2021).
2-Ethyl-3,5-dimethylpyrazine, 2,6-dimethylpyrazine, 2-
ethyl-6-methyl pyrazine and 2-ethyl-3,5-dimethylpyrazine
have been confirmed as the potent odorants in the final
products (Toci & Farah, 2014). Under the same roasting
conditions, a drier postharvest process gives higher pyra-
zine formation after coffee roasting along with a more
robust hazelnut flavour (Gonzalez-Rios et al., 2007a;
Hameed et al., 2018).
Aldehydes
Aldehydes, formed via Strecker degradation, are highly
aromatic compounds that contribute to the spectrum
of coffee aromas, including chocolate-like, floral,
honey-like, fruity, roasting and earthy (Gigl
et al., 2021). For example, phenylacetaldehyde has a
floral and honey-like aroma; methylbutanal is usually
perceived as malty or chocolate-like, while propanal
and hexanal obtain fruity and green odour (Fla-
ment, 2001; Barrios-Rodr
ıguez et al., 2021). The
postharvest process influences the generation and
extraction of aldehydes. Compared to the wet pro-
cess, the dry process gives coffee a relatively higher
aldehyde content, with a more pungent fruity/wine-
like taste and heavier body in the coffee brew, where
the different aldehydes may be generated from muci-
lage remnants (Hameed et al., 2018). However, a
drop in 3-methylbutanal and 2-phenylethanal (malty
aroma) could be observed after proteolytic and
lipolytic yeast fermentation (Lee et al., 2017a,2017b).
Furthermore, aldehydes tend to be influenced more
readily by other VOCs. A strong positive correlation
(r=0.752, P<0.001) between aldehyde and ketone
groups but negative correlations with pyrazines and
phenols were observed (Caporaso et al., 2018). Simi-
larly, Zakidou et al.(2021) also found that aldehydes
and ketones corporately contributed to 1–6% of the
VOCs in coffee, while none of the other single com-
pounds was more than 1%. Thus, aldehydes can be
generated via the oxidation of amino acids and
polyphenols when polyphenol oxidase is present
(Caporaso et al., 2018).
Ketones
Ketones are initially found in raw coffee beans as the
products of oxidative degradation of fatty acids. During
roasting, the main pathway of ketone formation includes
Maillard reactions, caramelisation, alcohol oxidation and
auto-oxidation of unsaturated fatty acids (Flament, 2001).
In this class, the potential coffee-odorant compounds con-
tributing to creamy, buttery and caramel flavour notes
include acetoin, 2,3-pentanedione, b-Damascenone, 1-
hydroxy-2-propanone, 1-hydroxy-2-butanone and 2,3-
butanedione (Ribeiro et al., 2018; de Melo Pereira
et al., 2019). Besides, the variability of single-origin coffee
VOCs and reported ketones as the most uniform VOCs
with a coefficient of variation (CV) <20% (Caporaso
et al., 2014). Roasting speed also affects the performance
of ketone formation. Slow and medium roasting can give
rise to a higher level of ketones in roasted coffee (Hameed
et al., 2018). However, over-fermentation caused by
improper postharvest processing (such as uncontrolled
microbial growth or over-dense drying area >20 kg m
2
)
could result in a foul smell and undesirable taste in the
final cup. Among these defective/degraded compounds,
ketones are one of the most significant off-flavour makers
(Toci & Farah, 2014).
Phenolic compounds
In the category of phenols/phenolic compounds, guaia-
col, 4-vinylguaiacol and 4-ethylguaiacol are the pri-
mary aromatic contributors in roasted coffee, bringing
a well-documented spicy and smoky note (Piccone
et al., 2012; de Melo Pereira et al., 2019; Zakidou
et al., 2021). Although roasting determinately regulates
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technolo gy 2023
Postharvest practice and coffee aroma X. Cao et al.1014
phenol productions because phenols are mainly gener-
ated from the thermal degradation of chlorogenic acids
and the decarboxylation of carboxylic acids, other
postharvest processing also influences the coffee phe-
nols formation (Zakidou et al., 2021). Phenols are
found in higher concentrations in higher/darker
roasting-intensity coffee (Moon & Shibamoto, 2009;
Gonzalez Viejo et al., 2021b). The contents of volatile
phenols, including 4-methoxyphenol, 4-ethylguaiacol
and 4-vinylguaiacol, were the topmost in washed coffee
but decreased in the semi-wet process and dry process
(Toledo et al., 2016). Furthermore, all aroma-related
phenols were detected in roasted coffee, while not all
exist in green beans (Lee et al., 2017a,2017b; de Melo
Pereira et al., 2019). Notably, the flavour contributions
of phenols may differ in different concentrations, as
guaiacol tends to have a burnt flavour at a relatively
higher concentration while sweet at a lower concentra-
tion (Gonzalez-Rios et al., 2007b;Gonz
alez-Arenzana
et al., 2020). Besides, phenol contents were observed
with a higher intra-batch variation (above 40% CV),
even representing over 100% CV in some coffee sam-
ples (Caporaso et al., 2018).
Sulphur compounds
Sulphur compounds are mainly produced during ther-
mal roasting and lead to an undesirable vegetable,
putrid and garlic smell even at a low threshold, which
shows an unignorable effect on the human olfactory
perception (Dulsat-Serra et al., 2016). For example,
methanethiol is described with cabbage and rotten egg
smell and is considered an indicator of low-quality cof-
fee, especially 3-methyl-2-buten-1-thiol (amine-like
contributor) (Caporaso et al., 2018; De Melo Pereira
et al., 2020,2019; Zakidou et al., 2021). The widely
accepted formation of sulphur compounds is the inter-
actions among sulphur-containing amino acids, sugars
and other minor chemicals. Maillard reactions largely
account for thiols generations while pentoses- or
hexoses-cysteine reactions for 2-furfurylthiol and 2-
methyl-3-furanthiol (Dulsat-Serra et al., 2016). Thus,
the roasting process is more influential on coffee thiol
production than other postharvest processes.
Volatile acids and fatty acids
Acids produced via microbial fermentation during cof-
fee processing may induce a pungent and sour flavour;
however, it could enhance the complexity of coffee fla-
vour and bring a rounder cup taste with a small
amount (Bressani et al., 2020). For example, acetic
acid (a simple monocarboxylic acid) could bring a
taste of vinegar on its own but can contribute to a
wine-like and sweet flavour paired with other com-
pounds while a saturated short-chain fatty acid,
isovaleric acid (3-methylbutanoic acid), with cheese-
like rancidity was also identified as a potential odorant
in brewed coffee (Chin et al., 2011). Besides, quantita-
tive analysis confirmed isovaleric acid as an indicator
of excellent coffee quality, over 80 scored by Speciality
Coffee Association (SCA) (Iwasa et al., 2021). Fatty
acid (FAs) fractions of triacylglycerols have a subtle
but desirable flavour note in coffee. It has also been
proven beneficial to human health due to its precursor
role of eicosanoid production, a bioregulator of many
cellular metabolisms (Dong et al., 2015). Linoleic acid
(C18:2) is an essential polyunsaturated fatty acid that
requires humans from external sources and is found in
large amounts in roasted coffee Fields (Couto
et al., 2009). In addition, palmitic acid (C16:0) and
archaic acid (C20:0) are the primary saturated fatty
acids, while oleic acid (C18:1) is the monounsaturated
fatty acid present in the roasted coffee (Wagemaker
et al., 2011; Dong et al., 2015) (Table 1).
Other VOCs
Other VOCs, including esters, pyrroles, pyridines and
alcohol, are also found in roasted coffee and affect cof-
fee sensory quality. Pyrrole derivatives are a subsidiary
class of VOCs found in the coffee brew (5–12%), par-
ticularly 1-methyl-1H-pyrrole-2-carboxaldehyde, 2-
acetylpyrrole and 2-formylpyrrole, contributing
roasted, nutty and cocoa notes (Zakidou et al., 2021).
They could be generated during the thermal degrada-
tion of Amodori intermediates, caramelisation, pyroly-
sis and trigonelline degradation (Flament, 2001;
Zakidou et al., 2021). Ester group represents the third
abundant in roasted coffee (5–8%) by two major com-
pounds, 2-furan methanol acetate (3.7–5.8%) and pro-
panoate (around 1.3%), which develop the fruity
aroma (Zakidou et al., 2021). Gonzalez Viejo
et al.(2021b) found 2-furan methanol acetate as the
most abundant VOC in NespressoÒcoffee pods with
different roasting intensities and reported related aro-
mas such as fruity, banana, floral and ethereal. Like-
wise, fermentation favours coffee esters formation
leading to a higher concentration of ethyl acetate,
methyl acetate and ethyl isovalerate. Ethanol and 2-
phenyl ethanol from the alcohol class has a sweet and
floral aroma. Pyridines could be generated through the
thermal degradation of Amodori intermediates and
pyrolysis of amino acids and trigonelline, contributing
3–6% presence in roasted coffee. However, different
postharvest practices could also randomly alter the
pyridine formation (Flament, 2001).
Coffee sensory
Coffee provides a diverse sensory experience, including
aroma, taste, mouthfeel and aftertaste. Among these,
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technology 2023
Postharvest practice and coffee aroma X. Cao et al. 1015
Table 1 Common VOCs in roasted coffee with molecular formula, chemical structure and description notes; all molecular formula
and chemical structure are cited from PubChem
Compound
classification Compound name
Molecular
formula
Chemical
structure Aroma notes References
Pyrazines Pyrazine C
4
H
4
N
2
Nutty,
roasted,
pungent
De Melo Pereira
et al.(2019),
Zakidou
et al.(2021)
2-Ethyl-6-methylpyrazine C
7
H
10
N
2
Cocoa,
roasted,
hazelnut-like
Akiyama
et al.(2003),
Zakidou
et al.(2021)
2-Ethyl-5-methylpyrazine C
7
H
10
N
2
Nutty, coffee-
like
Caporaso
et al.(2014),
Zakidou
et al.(2021)
2,6-Dimethylpyrazine C
6
H
8
N
2
Nutty, cocoa,
coffee-like
Zakidou
et al.(2021)
2-Ethyl-3,5-dimethylpyrazine C
8
H
12
N
2
Nutty, burnt,
almond-like,
roasted
de Melo Pereira
et al.(2019), de
Morais et al.,
2007, Zakidou
et al.(2021)
2,6-Diethylpyrazine C
6
H
8
N
2
Nutty, toasted Gonzalez-Rios
et al.(2007b),
Caporaso
et al.(2018)
2,6-Dimethylpyrazine C
8
H
12
N
2
Nutty,
roasted,
cocoa-like,
coffee-like
Zakidou
et al.(2021)
Furans 2,5-Dimethylfuran C
6
H
8
O Meaty,
roasted,
ethereal;
coffee-like
Flament (2001),
Gonzalez-Rios
et al.(2007b),
Alstrup
et al.(2020),
Zakidou
et al.(2021)
5-Methylfurfural C
6
H
6
O
2
Caramel,
maple,
spicy, sweet
de Melo Pereira
et al.(2019),
Zakidou
et al.(2021)
furfural C
5
H
4
O
2
or
C
4
H
3
OCHO
Sweet,
woody,
almond-like
Gonzalez-Rios
et al.(2007b),
Zakidou
et al.(2021)
2-Furanmethanol C
5
H
6
O
2
Caramel,
warm-oily,
burnt,
smoky
Flament (2001),
Zakidou
et al.(2021)
Furfuryl acetate C
7
H
8
O
3
Sweet, fruity,
banana
(Evangelista et al.
(2015),
Yanagimoto
et al.(2002)
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technolo gy 2023
Postharvest practice and coffee aroma X. Cao et al.1016
Table 1 (Continued)
Compound
classification Compound name
Molecular
formula
Chemical
structure Aroma notes References
Furfuryl alcohol C
5
H
6
O
2
Sweet,
caramel,
brown sugar
Evangelista et al.
(2015), Poyraz
et al.(2016)
5-Hydroxymethylfurfural C
6
H
6
O
3
Buttery,
caramel
Zakidou
et al.(2021)
Aldehydes Acetaldehyde C
2
H
3
O Fruity, fresh,
green
Caporaso
et al.(2018), de
Melo Pereira
et al.(2019),
Zakidou
et al.(2021)
2-Methylpropanal C
4
H
8
Oor
(CH
3
)
2
CHCHO
Fruity, malty, Caporaso
et al.(2014)
Butanal C
4
H
8
Oor
CH
3
CH
2
CH
2
CHO
Fruity,
chocolate
Caporaso
et al.(2018),
L
opez-Galilea
et al.(2006)
2-Methylbutanal C
5
H
10
O Fruity, malty,
roasted
cocoa,
musty
Caporaso
et al.(2014)
3-Methylbutanal C
5
H
10
O Fruity, malty,
peach, cocoa
Flament (2001),
Caporaso
et al.(2014), de
Melo Pereira
et al.(2019)
Hexanal C
6
H
12
O Fruity, leafy,
fatty, grassy
Flament (2001),
Caporaso
et al.(2014),
Zakidou
et al.(2021)
Phenols Guaiacol C
7
H
8
O
2
Smoky, spicy,
burnt,
Schenker et al.
(2002)
4-Ethylguaiacol C
9
H
12
O
2
Smoky, spicy,
roasted
Gonzalez-Rios
et al.(2007b), de
Melo Pereira
et al.(2019)
4-Vinylguaiacol C
9
H
10
O
2
Smoky, spicy,
clove-like
Flament (2001),
Caporaso
et al.(2014),
Zakidou
et al.(2021)
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technology 2023
Postharvest practice and coffee aroma X. Cao et al. 1017
Table 1 (Continued)
Compound
classification Compound name
Molecular
formula
Chemical
structure Aroma notes References
3-Methylphenol C
7
H
8
O Woody,
leather-like
Lee et al.(2017a,
2017b), de Melo
Pereira
et al.(2019)
Vanillin C
8
H
8
O
3
Vanilla, sweet Schenker et al.
(2002), Zakidou
et al.(2021)
Ketones 2,3-Pentanedione C
5
H
8
O
2
Buttery, oily-
like, caramel
Caporaso
et al.(2014),
L
opez-Galilea
et al.(2006),
Zakidou
et al.(2021)
b-Damascenone C
13
H1
8
O Floral, fruity,
honey-like
Caporaso
et al.(2014),
Schenker et al.
(2002)
Acetoin C
4
H
8
O
2
Buttery,
creamy,
dairy-like
de Melo Pereira
et al.(2019),
Zakidou
et al.(2021)
1-Hydroxy-2-propanone C
3
H
6
O
2
Sweet,
caramel
Zakidou
et al.(2021)
1-Hydroxy-2-butanone C
4
H
8
O
2
Sweet,
caramel
Gonzalez-Rios
et al.(2007b), de
Melo Pereira
et al.(2019),
Zakidou
et al.(2021)
2,3-Butanedione C
4
H
6
O
2
Buttery,
creamy,
caramel
de Melo Pereira
et al.(2019),
Evangelista
et al.(2015)
Sulphur
compounds
Methanethiol CH
4
S Cabbage-like,
garlic, rotten
egg
Blank et al.(1991)
3-Methyl-2-buten-1-thiol C
5
H
10
S Amine-like Blank et al.(1991)
Thiophene C
4
H
4
S Sulphur-like,
garlic
Zakidou
et al.(2021)
Fatty acids Palmitic acid (SFA) C
16
H
32
O
2
Cheese-like Wagemaker
et al.(2011)
Arachic acid (SFA) C
20
H
40
O
2
N/A Dong et al.(2015)
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technolo gy 2023
Postharvest practice and coffee aroma X. Cao et al.1018
the smell and taste of coffee contribute to the overrid-
ing understanding of consumption. As shown in
Fig. 4, multiple attributes of coffee sensory descrip-
tions are covered and illustrated in detail in the coffee
flavour wheel (Bolger et al., 2017).
Contributed by VOCs, the coffee odour is the first
descriptor perceived when obtaining the sample beside
the visual attributes. Odours are perceived via orthona-
sal, meaning that the volatile compounds travel
through the nose to reach the olfactory bulb when
sniffing. More complex attributes are the aromas,
which are perceived once the product is introduced
into the mouth, where the volatile compounds move
through the larynx towards the olfactory system (Lan-
dis et al., 2005; Small et al., 2005). Concerning
humans, the sense of smell is greatly more sensitive
than taste. Generally, the smell of ground coffee before
pouring water is called fragrance, while the aroma is
always referred to as once water is added, regardless
of the distinction between the odour and aroma used
in sensory and health sciences as described above
(Lingle & Menon, 2017). Thus, the term ‘aroma’ is used
in this review because coffee is consumed in liquid.
VOCs with different functional groups contribute to the
other coffee aromas, influenced by many factors, from
the variety of the beans to postharvest treatment. In
this case, most light will be shed on the effect of
postharvest practices on the VOCs production and thus
sensory quality (Zakidou et al., 2021). Most VOCs
come from Maillard reactions, Strecker degradation
and other heat-related reactions, and thus the VOCs
quantity and quality are primarily associated with the
roasting process (Zakidou et al., 2021). Common sen-
sory attributes related to coffee flavour include acidity,
sweetness and bitterness, all of which are closely associ-
ated with the performance of the postharvest process
(Alstrup et al., 2020). The dry process typically gives
coffee a sweeter taste due to higher fructose and glucose
contents which are significantly reduced in washed cof-
fee (de Melo Pereira et al., 2019).
With the help of a humid environment, those sugars
as the core reactant substances can promote more
robust metabolism and heat-related reactions, produc-
ing enjoyable coffee VOCs such as furans and pyra-
zines. Zakidou et al.(2021) investigated ten types of
coffee beans and summarised that furan derivatives (5-
methyl furfural and 2-furan methanol) were charac-
terised as sweet-like aromas in coffee while pyrazines
(2-ethyl-3,5-dimethyl pyrazine) brought nutty and
chocolaty flavour. The common VOCs found in
roasted coffee with their descriptive notes in previous
research are shown in Table 1. The washed process
Table 1 (Continued)
Compound
classification Compound name
Molecular
formula
Chemical
structure Aroma notes References
Oleic acid (MUFA) C
18
H
34
O
2
N/A Dong et al.(2015)
Linoleic acid (PUFA) C
18
H
32
O
2
Vinegar-like Dong et al.(2017)
Isovaleric acid (SFA) C
5
H
10
O
2
Cheese-like,
dairy, rancid
Dong
et al.(2017), de
Melo Pereira
et al.(2019)
Other organic
acids
Acetic acid C
2
H
4
O
2
Pungent,
sour,
vinegar-like,
fermented
Dong
et al.(2015), de
Melo Pereira
et al.(2019),
Zakidou
et al.(2021)
MUFA, monosaturated fatty acid; PUFA, polyunsaturated fatty acid; SFA, saturated fatty acid.
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technology 2023
Postharvest practice and coffee aroma X. Cao et al. 1019
also facilitates the hydrolysis of proteins. Thus, it
increases the concentrations of free amino acids in
roasted coffee, which is believed to contribute to the
fruity taste (de Melo Pereira et al., 2019). Well-
controlled alcoholic or lactic fermentation may also
develop further under wet conditions due to the diver-
sity and community of more microorganisms (Jo€
et
et al., 2010; de Melo Pereira et al., 2019). Like dairy
products, where extra-cream milk has a heavier mouth-
feel than full-cream and low-fat dairy, coffees can have
a similar taste and aroma but a different mouthfeel -
trigeminal sensations and flavour (combination of aro-
mas, tastes and mouthfeel). Mouthfeel attributes can
be described as thin, heavy/whole, juicy, buttery,
creamy or astringent mouthfeel (Hayakawa
et al., 2010). Two indicators are considered about cof-
fee aftertaste: the lasting time of the taste and whether
it is pleasant. Generally, coffee with a prolonged and
pleasant aftertaste is the most rewarded. Coffee with
an eliminated presence of off-flavour scored higher in
clean cup quality, which indicates a proper postharvest
treatment such as an efficient sorting step and well-
controlled fermentation (Hameed et al., 2018).
From the perspective of sensory analysis, Quantita-
tive Descriptive Analysis (QDAÒ) by the trained panel
is commonly used in the production and development
of the food and beverage industry. The principle of
the QDAÒmethod is to develop a comprehensive and
quantitative product description based on the ability
of a trained party to quantify specific attributes
(Hunaefi et al., 2020). It is a widespread tool to quan-
tify descriptors related to the appearance, aroma, tex-
ture, mouthfeel, taste and flavour characteristics of
brewed coffee (Liseth et al., 2019). However, the cost
of obtaining professional panellists is relatively high,
and the results of this method could be subjective
(Kemp et al., 2011; Fuentes et al., 2018; Fuentes
et al., 2021b).
Advanced technology in the coffee volatile
compounds determination
VOCs analysis –GC-MS
Gas chromatography coupled with mass spectrometry
(GC-MS) is a useful tool for VOCs determination in
Figure 4 The coffee tastes and aromas wheel (Bolger et al., 2017).
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technolo gy 2023
Postharvest practice and coffee aroma X. Cao et al.1020
food matrix such as yoghurt, cheese, roasted coffee
and wine (Delgado et al., 2010; Piccone et al., 2012;
Sunarharum et al., 2014; Caporaso et al., 2018). Along
with the GC-MS test, solid-phase microextraction
(SPME) is commonly used for the extraction and pre-
concentration (using proper absorbing materials to
collect the targeted compounds continuously) of the
volatile fraction, which is simple, fast and requires
minimal treatment and amount of sample, either solid
or liquid (Blake et al., 2009; Delgado et al., 2010). The
validation of SPME/GC-MS to determine and track
the VOCs in roasted coffee is well-documented (De
Toledo et al., 2017; De Melo Pereira et al., 2019;
Angeloni et al., 2020; Lolli et al., 2020). Zakidou
et al.(2021) found over 130 compounds in roasted cof-
fee samples through SPME/GC-MS, with an extensive
range of chemical categories, including pyrazine, ester
furan and pyrrole, among others. SPME/GC-MS was
also utilised to compare the odour compounds in
Ethiopian coffee and detect 80 VOCs with 14 chemical
classes (Akiyama et al., 2005). Also, 81 compounds (of
10 chemical classes) were successfully measured in
commercial coffee via the GC-MS analysis (Lee
et al., 2013b). Over 70 potential odorant compounds
were identified by GC-MS in green and roasted
ground coffees, with 12 chemical classes (Lee
et al., 2017). Besides, numerous studies have also
authorised the capability of GC-MS on adulteration in
the food industry (Cheong et al., 2013; Peris &
Escuder-Gilabert, 2016; Dong et al., 2017). Besides, as
a variant of gas chromatography, Proton-transfer
Reaction Mass Spectrometry (PTR-MS) is almost
exclusively for gaseous compounds detection, making
it suitable for coffee VOCs detection, with the advan-
tages of being fast (<1 min) and high detection sensi-
tivity. However, the PTR-MS technique relies on mass
spectrometry to discriminate compounds, which may
lead to misdetection or identify all the species present
when different molecules have the same mass weight
present in a complex VOCs mixture (Blake
et al., 2009). Hence, applying GC techniques in detect-
ing VOCs shows promise in the coffee area.
E-nose
With the chemical detection of VOCs, sensory analysis
of coffee aromas is also conducted in academic studies.
Traditional methods such as a descriptive sensory
panel or quantitative descriptive analysis (QDAÒ) are
widely used (Dzung et al., 2003). Nonetheless, these
techniques employ human perceptions, which may be
less reliable when panel members are subjected to
biases such as stimulus errors (the assessors make
judgements with additional information) and data
variance based on panellists (Kemp et al., 2011).
Moreover, these methods tend to be time-consuming,
expensive and significantly vary between and within
individual fields (Gonzalez Viejo et al., 2021b; Fuentes
et al., 2021a). A new technique, electronic nose
(e-nose), can remarkably save global data and recog-
nise numerous and diverse odorant substances, includ-
ing aldehydes, pyrazines and ketones, for the
chemometric analysis (Pearce et al., 2006; Severini
et al., 2015; Dong et al., 2017).
The Digital Agriculture Food and Wine Group from
The University of Melbourne (DAFW; UoM) devel-
oped a portable and low-cost e-nose composed of an
array of nine gas sensors, coupled with machine learn-
ing modelling as a rapid method to detect the aromas
in food products such as coffee, beer and wine and
agricultural applications such as pest detection in
wheat (Viejo et al., 2020; Gonzalez Viejo et al., 2021a,
2021b; Summerson et al., 2021a,2021b; Fuentes
et al., 2021c). Fig. 5illustrates its sample-handling
part, featuring a 92-mm-diameter and light-weight sys-
tem armed with a temperature and humidity sensor
Figure 5 Structures of (a) the portable e-nose system with nine gas sensors (b) stalked multiple printed circuit board (PCB) design to reduce
the footprint (Viejo et al., 2020).
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technology 2023
Postharvest practice and coffee aroma X. Cao et al. 1021
(AM2320, Guangzhou Aosong Electronics Co., Ltd.,
Guangzhou, China) aimed to ensure all samples are
analysed under similar conditions (ambient and rela-
tive humidity). It consists of nine gas sensors, and
their specifications are shown in Table 2(Henan Han-
wei Electronics Co., Ltd, Zhengzhou, China) (Fuentes
et al., 2020; Gonzalez Viejo & Fuentes, 2020; Viejo
et al., 2020). A Microcontroller with an onboard ADC
was employed to read the voltages at the interval of
500 ms. In their study, significant differences among
aromas in different beers were successfully obtained,
which indicated the discriminability of VOCs of e-
nose. Furthermore, two highly accurate artificial neu-
ral network (ANN) models were developed to predict
(i) 17 volatile aromatic compounds with the GC–MS
targets (accuracy: R=0.97) and (ii) the intensity of 10
sensory descriptors with sensory targets using QDAÒ
(accuracy: R=0.93). The same research group used
the e-nose to assess coffee pods with different roast
intensities along with two ANN models to (i) classify
samples according to the roast intensity level (accu-
racy: 98%) and (ii) predict 45 volatile aromatic com-
pounds with GC-MS targets (accuracy: R=0.99)
(Gonzalez Viejo et al., 2021b).
Recently, the e-nose has been widely used in many
fields (coffee, wine, fish and saffron) to assess the qual-
ity of aromatic products due to its rapid, user-friendly
and non-invasive nature (Kiani et al., 2016; Marek
et al., 2020; Viejo et al., 2020). Marek et al., 2020 uti-
lised e-nose, assisting with GC-MS, to analyse and
compare the volatile substances in roasted coffee sam-
ples from different countries, proving the suitability
for discrimination of coffee aromas. Similarly, Flam-
beau et al.(2017) successfully grouped Rwandan coffee
(Bourbon varietal) into sub-regional classes (such as
northern and southern areas in Rwanda) based on
their aroma profiles via e-nose. Furthermore, com-
bined with the electronic tongue (e-tongue) technique,
the capability of the e-nose system on coffee physio-
chemical determination, including pH value, titrate
acidity (TA), total solids (TS), total soluble solids
(TSS) and TSS/TSA ratio, has been verified in their
study of Chinese Robusta coffee differentiation (Dong
et al., 2017). E-nose is also a valuable tool for deter-
mining the optimal time for coffee to be packaged,
making a difference in the commercial coffee industry
(Falasconi et al., 2003).
Other techniques
Some other techniques in coffee VOCs qualification
and quantification in the past few years can be used.
For instance, Gas chromatography–olfactory (GC-O)
combined with Solid Phase Microextraction (SPME)
has attempted the discrimination of coffee beans with
the different postharvest processes by integrating the
separate VOCs using gas chromatography with an
odour detector called olfactometer (human assessor)
(Sunarharum et al., 2014). Nuclear Magnetic Reso-
nance (NMR) is another valuable tool to reveal the
substantial chemical changes during the coffee posthar-
vest process by detecting the main chemical ingredients
in roasted coffee, such as caffeine and caffeoylquinic
acids (CQAs) and other water-solvent compounds
(Alstrup et al., 2020). Near-Infrared Spectroscopy
(NIRS) can monitor the physical changes during coffee
processing (moisture, lost weight and density) and per-
form with high accuracy using iterative predictor
weighting (IPW) and partial least square regression
(PLS) (Esteban-D
ıez et al., 2004). To broaden the
study scope, more research on those non-volatile com-
pounds that act as precursors in green coffee is
required.
Conclusion
The composition of coffee volatile compounds (VOCs)
is vital for coffee aroma production. Furans, pyrazines,
aldehydes, ketones, phenolic compounds, sulphur com-
pounds, esters, pyrroles, alcohols and pyridines as the
influential VOCs found in roasted coffee. Among
those, furans and pyrazines are the sweet and hazelnut
aroma contributor in roasted coffee, respectively, fol-
lowed by aldehydes (fruity) and ketones (caramel and
buttery), esters (fruity), alcohols (floral and wine-like)
and pyridines (chocolate-like) with desirable aroma
notes. The degree of off-odours largely depends on the
threshold, such as phenolic compounds (roasty and
spice-like), pyrroles (cabbage-like) and sulphur com-
pounds (vegetable-like). Microbial fermentation in wet
processing could significantly influence the VOC pre-
cursor formation and aromatic contribution. Roasting-
Table 2 Specifications of the nine gas sensors and a
temperature-humidity sensor for beer characterisation (Henan
Hanwei Electronics Co., Ltd), modified from Viejo et al.(2020)
Sensor Measurement Sensitivity
MQ3 Ethanol 0.05–10 mg L
1
MQ4 Methane 200–10 000 ppm
MQ7 Carbon monoxide 20–2000 ppm
MQ8 Hydrogen 100–10 000 ppm
MQ135 Ammonia alcohol
benzene
10–300 ppm; 10–300 ppm;
10–1000 ppm
MQ136 Hydrogen sulphide 1–100 ppm
MQ137 Ammonia 5–200 ppm
MQ138 Benzene alcohol
ammonia
10–1000 pp; 10–1000 ppm;
10–3000 ppm
MG811 Carbon dioxide 350–10 000 ppm
AM2320 Humidity
temperature
0–99%; 40–80 °C
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technolo gy 2023
Postharvest practice and coffee aroma X. Cao et al.1022
triggered Maillard reactions, Strecker degradation and
caramelisation in coffee beans are the primary reac-
tions related to roasted coffee aroma generation. GC-
MS and e-nose sensory analysis are commonly used in
coffee quality assurance, with high detectability, accu-
racy and user-friendly nature. However, novel, low-
cost e-noses coupled with machine learning modelling
are a portable, reliable and rapid option to assess cof-
fee, especially for small producers in developing coun-
tries that cannot afford GC-MS. The variation and
interactions among different processing conditions
(method, roasting and storage) and numerous unma-
tured techniques applied in coffee processing still limit
further comparison. Thus, future studies need to pri-
oritise those potential barriers to better understand the
speciality coffee industry.
Funding information
Dr Hafiz Suleria is the recipient of an “Australian
Research Council - Discovery Early Career Award”
(ARC-DECRA - DE220100055) funded by the
Australian Government. This research was funded by
the University of Melbourne under the “McKenzie
Fellowship Scheme” (grant no. UoM-18/21) and “Col-
laborative Research Development Grant (grant no.
UoM-21/23)” funded by the Faculty of Veterinary and
Agricultural Sciences, the University of Melbourne,
Australia.
Conflict of Interest
The authors declare that they have no known compet-
ing financial interests or personal relationships that
could have appeared to influence the work reported in
this paper.
Author contributions
Xiaotong Cao: Conceptualization (equal); visualization
(equal); writing –original draft (lead). Hanjing Wu:
Conceptualization (equal); software (equal); visualiza-
tion (lead); writing –review and editing (lead). Claudia
Gonzalez Viejo: Investigation (equal); supervision
(equal); writing –review and editing (supporting).
Frank R. Dunshea: Investigation (equal); supervision
(equal); writing –review and editing (supporting).
Hafiz A.R. Suleria: Conceptualization (supporting);
data curation (equal); methodology (supporting);
supervision (lead); visualization (supporting); writing –
review and editing (supporting).
Peer review
The peer review history for this article is available at
https://publons.com/publon/10.1111/ijfs.16261.
Acknowledgment
Open access publishing facilitated by The University
of Melbourne, as part of the Wiley - The University of
Melbourne agreement via the Council of Australian
University Librarians.
References
Ahmed, S., Brinkley, S., Smith, E. et al. (2021). Climate change and
Coffee quality: Systematic review on the effects of environmental
and management variation on secondary metabolites and sensory
attributes of Coffea arabica and Coffea canephora.Frontiers in
Plant Science,12, 708013.
Akiyama, M., Murakami, K., Ikeda, M. et al. (2005). Characteriza-
tion of flavor compounds released during grinding of roasted
robusta coffee beans. Food Science and Technology Research,11,
298–307.
Akiyama, M., Murakami, K., Ohtani, N. et al. (2003). Analysis of
volatile compounds released during the grinding of roasted coffee
beans using solid-phase microextraction. Journal of Agricultural
and Food Chemistry,51, 1961–1969.
Alstrup, J., Petersen, M.A., Larsen, F.H. & M€
unchow, M. (2020).
The effect of roast development time modulations on the sensory
profile and chemical composition of the coffee brew as measured
by NMR and DHS-GC–MS. Beverages,6, 70.
Amarowicz, R. (2009). Antioxidant activity of Maillard reaction
products. European Journal of Lipid Science and Technology,111,
109–111.
Angeloni, S., Scortichini, S., Fiorini, D. et al. (2020). Characteriza-
tion of odor-active compounds, polyphenols, and fatty acids in cof-
fee silverskin. Molecules,25, 2993.
Barrios-Rodr
ıguez, Y.F., Reyes, C.A.R., Campos, J.S.T., Gir
on-
Hern
andez, J. & Rodr
ıguez-Gamir, J. (2021). Infrared spectroscopy
coupled with chemometrics in coffee postharvest processes as com-
plement to the sensory analysis. LWT,145, 111304.
Becalski, A., Halldorson, T., Hayward, S. & Roscoe, V. (2016).
Furan, 2-methylfuran and 3-methylfuran in coffee on the Cana-
dian market. Journal of Food Composition and Analysis,47, 113–
119.
Besnard, P., Passilly-Degrace, P. & Khan, N.A. (2016). Taste of fat:
A sixth taste modality? Physiological Reviews,96, 151–176.
Blake, R.S., Monks, P.S. & Ellis, A.M. (2009). Proton-transfer reac-
tion mass spectrometry. Chemical Reviews,109, 861–896.
Blank, I., Sen, A. & Grosch, W. (1991). Aroma impact compounds
of Arabica and Robusta coffee. Qualitative and quantitative inves-
tigations. In: 14th International Scientific Colloquium on Coffee. Pp.
117–129. San Francisco, CA: ASIC.
Bressani, A.P.P., Martinez, S.J., Sarmento, A.B.I., Bor
em, F.M. &
Schwan, R.F. (2020). Organic acids produced during fermentation
and sensory perception in specialty coffee using yeast starter cul-
ture. Food Research International,128, 108773.
Budryn, G., Nebesny, E.,
_
Zy_
zelewicz, D., Oracz, J., Mi
skiewicz, K.
& Rosicka-Kaczmarek, J. (2012). Influence of roasting conditions
on fatty acids and oxidative changes of Robusta coffee oil. Euro-
pean Journal of Lipid Science and Technology,114, 1052–1061.
Buffo, R.A. & Cardelli-Freire, C. (2004). Coffee flavour: An over-
view. Flavour and Fragrance Journal,19,99–104.
Caporaso, N., Genovese, A., Canela, M.D., Civitella, A. & Sacchi,
R. (2014). Neapolitan coffee brew chemical analysis in comparison
to espresso, moka and American brews. Food Research Interna-
tional,61, 152–160.
Caporaso, N., Whitworth, M.B., Cui, C. & Fisk, I.D. (2018). Vari-
ability of single bean coffee volatile compounds of arabica and
robusta roasted coffees analysed by SPME-GC-MS. Food Research
International,108, 628–640.
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technology 2023
Postharvest practice and coffee aroma X. Cao et al. 1023
Bolger, L., Bria, B., Civille, G.V. et al. (2017). World Coffee
Research Sensory Lexicon. College Station, TX: World Coffee
Research and Kansas State University.
Chang, Y., Tang, N., Qu, H. et al. (2016). Detection of volatile
organic compounds by self-assembled monolayer coated sensor
array with concentration-independent fingerprints. Scientific
Reports,6,1–12.
Cheong, M.W., Tong, K.H., Ong, J.J.M., Liu, S.Q., Curran, P. &
Yu, B. (2013). Volatile composition and antioxidant capacity of
arabica coffee. Food Research International,51, 388–396.
Chin, S.T., Eyres, G.T. & Marriott, P.J. (2011). Identification of
potent odourants in wine and brewed coffee using gas
chromatography-olfactometry and comprehensive two-dimensional
gas chromatography. Journal of Chromatography A,1218, 7487–
7498.
Chindapan, N., Chaninkun, N. & Devahastin, S. (2022). Compara-
tive evaluation of phenolics and antioxidant activities of hot air
and superheated steam roasted coffee beans (Coffea canephora).
International Journal of Food Science & Technology,57, 342–350.
Clemente, J.M., Martinez, H.E.P., Alves, L.C., Finger, F.L. &
Cecon, P.R. (2015). Effects of nitrogen and potassium on the
chemical composition of coffee beans and on beverage quality.
Acta Scientiarum. Agronomy,37, 297–305.
Couto, R.M., Fernandes, J., Da Silva, M.G. & Simoes, P.C. (2009).
Supercritical fluid extraction of lipids from spent coffee grounds.
The Journal of Supercritical Fluids,51, 159–166.
Da Silva, M.C.S., Da Luz, J.M.R., Veloso, T.G.R. et al. (2022). Pro-
cessing techniques and microbial fermentation on microbial profile
and chemical and sensory quality of the coffee beverage. European
Food Research and Technology,248, 1499–1512.
Da Silva Portela, C., De Almeida, I.F., Dos Reis, T.A.D., Hick-
mann, B.R.B. & De Toledo Benassi, M. (2022). Effects of brewing
conditions and coffee species on the physicochemical characteris-
tics, preference and dynamics of sensory attributes perception in
cold brews. Food Research International,151, 110860.
Davis, A.P., Govaerts, R., Bridson, D.M. & Stoffelen, P. (2006). An
annotated taxonomic conspectus of the genus Coffea (Rubiaceae).
Botanical Journal of the Linnean Society,152, 465–512.
De Melo Pereira, G.V., De Carvalho Neto, D.P., J
unior, A.I.M.
et al. (2020). Chemical composition and health properties of coffee
and coffee by-products. Advances in Food and Nutrition Research,
91,65–96.
De Melo Pereira, G.V., De Carvalho Neto, D.P., J
unior, A.I.M.
et al. (2019). Exploring the impacts of postharvest processing on
the aroma formation of coffee beans–a review. Food Chemistry,
272, 441–452.
De Oliveira, F.C., Coimbra, J.S.D.R., De Oliveira, E.B., Zu~
niga,
A.D.G. & Rojas, E.E.G. (2016). Food protein-polysaccharide con-
jugates obtained via the Maillard reaction: A review. Critical
Reviews in Food Science and Nutrition,56, 1108–1125.
De Toledo, P.R., De Melo, M.M., Pezza, H.R., Toci, A.T., Pezza,
L. & Silva, C.M. (2017). Discriminant analysis for unveiling the
origin of roasted coffee samples: A tool for quality control of cof-
fee related products. Food Control,73, 164–174.
Decazy, F., Avelino, J., Guyot, B., Perriot, J.-J., Pineda, C. & Cilas,
C. (2003). Quality of different Honduran coffees in relation to sev-
eral environments. Journal of Food Science,68, 2356–2361.
Delaroza, F., Rakocevic, M., Malta, G.B., Sanchez, P.M., Bruns,
R.E. & Scarminio, I.S. (2017). Factorial design effects of plant den-
sity, pattern and light availability on the caffeine, chlorogenic
acids, lipids, reducing sugars and ash contents of Coffea arabica L.
beans and leaves. Analytical Methods,9, 3612–3618.
Delgado, F.J., Gonz
alez-Crespo, J., Cava, R., Garc
ıa-Parra, J. &
Ram
ırez, R. (2010). Characterisation by SPME–GC–MS of the
volatile profile of a Spanish soft cheese PDO Torta del Casar dur-
ing ripening. Food Chemistry,118, 182–189.
Dong, W., Tan, L., Zhao, J., Hu, R. & Lu, M. (2015). Characteriza-
tion of fatty acid, amino acid and volatile compound compositions
and bioactive components of seven coffee (Coffea robusta) cultivars
grown in Hainan Province, China. Molecules,20, 16687–16708.
Dong, W., Zhao, J., Hu, R., Dong, Y. & Tan, L. (2017). Differentia-
tion of Chinese robusta coffees according to species, using a com-
bined electronic nose and tongue, with the aid of chemometrics.
Food Chemistry,229, 743–751.
Dryahina, K., Smith, D. &
Span
el, P. (2018). Quantification of vola-
tile compounds released by roasted coffee by selected ion flow tube
mass spectrometry. Rapid Communications in Mass Spectrometry,
32, 739–750.
Dulsat-Serra, N., Quintanilla-Casas, B. & Vichi, S. (2016). Volatile
thiols in coffee: A review on their formation, degradation, assess-
ment and influence on coffee sensory quality. Food Research Inter-
national,89, 982–988.
Dzung, N.H., Dzuan, L. & Tu, H.D. (2003). The role of sensory
evaluation in food quality control, food research and development:
A case of coffee study. Proceedings of the 8th Asean food Confer-
ence, Hanoi, Vietnam, 2003. Citeseer, 8-11.
EFSA Panel on Contaminants in the Food Chain (CONTAM),
Knutsen, H.K., Alexander, J. et al. (2017). Risks for public health
related to the presence of furan and methylfurans in food. EFSA
Journal,15, e05005.
Esquivel, P. & Jimenez, V.M. (2012). Functional properties of coffee
and coffee by-products. Food Research International,46, 488–495.
Esteban-D
ıez, I., Gonz
alez-S
aiz, J. & Pizarro, C. (2004). Prediction
of sensory properties of espresso from roasted coffee samples by
near-infrared spectroscopy. Analytica Chimica Acta,525, 171–182.
Evangelista, S.R., Silva, C.F., Da Cruz Miguel, M.G.P. et al. (2014).
Improvement of coffee beverage quality by using selected yeasts
strains during the fermentation in dry process. Food Research
International,61, 183–195.
Evangelista, S.R., Miguel, M.G.C.P., Silva, C.F., Pinheiro, A.C.M.
& Schwan, R.F. (2015). Microbiological diversity associated with
the spontaneous wet method of coffee fermentation. International
Journal of Food Microbiology,210, 102–112.
Falasconi, M., Pardo, P., Sberveglieri, G. et al. (2003). The novel
EOS/sup 835/electronic nose sniffs out the Italian espresso coffee
quality. SENSORS, 2003 IEEE, 2003. IEEE, 26-29.
Ferrari, M., Ravera, F., De Angelis, E., Liverani, F.S. & Navarini,
L. (2010). Interfacial properties of coffee oils. Colloids and Surfaces
A: Physicochemical and Engineering Aspects,365,79–82.
Flambeau, K.J., Lee, W.-J. & Yoon, J. (2017). Discrimination and
geographical origin prediction of washed specialty Bourbon coffee
from different coffee growing areas in Rwanda by using electronic
nose and electronic tongue. Food Science and Biotechnology,26,
1245–1254.
Flament, I. (2001). Coffee Flavor Chemistry. Hoboken, NJ: John
Wiley & Sons.
Franca, A.S., Mendonc
ßa, J.C. & Oliveira, S.D. (2005). Composition
of green and roasted coffees of different cup qualities. LWT-Food
Science and Technology,38, 709–715.
Fuentes, S., Gonzalez Viejo, C., Torrico, D. & Dunshea, F. (2018).
Development of a biosensory computer application to assess physi-
ological and emotional responses from sensory panelists. Sensors,
18, 2958.
Fuentes, S., Gonzalez Viejo, C., Torrico, D.D. & Dunshea, F.R.
(2021a). Digital integration and automated assessment of eye-
tracking and emotional response data using the BioSensory App to
maximize packaging label analysis. Sensors,21, 7641.
Fuentes, S., Summerson, V., Gonzalez Viejo, C. et al. (2020). Assess-
ment of smoke contamination in grapevine berries and taint in
wines due to bushfires using a low-cost E-nose and an artificial
intelligence approach. Sensors,20, 5108.
Fuentes, S., Tongson, E. & Gonzalez Viejo, C. (2021b). Novel digital
technologies implemented in sensory science and consumer percep-
tion. Current Opinion in Food Science,41,99–106.
Fuentes, S., Tongson, E., Unnithan, R.R. & Viejo, C.G. (2021c).
Early detection of aphid infestation and insect-plant interaction
Ó2022 The Authors. International Journal of Food Science & Technology published by John Wiley & Sons Ltd
on behalf of Institute of Food, Science and Technology (IFSTTF).
International Journal of Food Science and Technolo gy 2023
Postharvest practice and coffee aroma X. Cao et al.1024
assessment in wheat using a low-cost electronic nose (E-nose),
near-infrared spectroscopy and machine learning modeling. Sen-
sors,21, 5948.
Gigl, M., Frank, O., Barz, J., Gabler, A., Hegmanns, C. & Hof-
mann, T. (2021). Identification and quantitation of reaction prod-
ucts from quinic acid, quinic acid lactone, and chlorogenic acid
with Strecker aldehydes in roasted coffee. Journal of Agricultural
and Food Chemistry,69, 1027–1038.
Gonzalez Viejo, C. & Fuentes, S. (2020). Low-cost methods to assess
beer quality using artificial intelligence involving robotics, an elec-
tronic nose, and machine learning. Fermentation,6, 104.
Gonzalez Viejo, C., Fuentes, S. & Hernandez-Brenes, C. (2021a).
Smart detection of faults in beers using near-infrared spectroscopy,
a Low-cost electronic nose and artificial intelligence. Fermentation,
7, 117.
Gonzalez Viejo, C., Tongson, E. & Fuentes, S. (2021b). Integrating
a low-cost electronic nose and machine learning modelling to assess
coffee aroma profile and intensity. Sensors,21, 2016.
Gonz
alez-Arenzana, L., Santamar
ıa, R., Escribano-Viana, R. et al.
(2020). Influence of the carbonic maceration winemaking method
on the physicochemical, colour, aromatic and microbiological fea-
tures of tempranillo red wines. Food Chemistry,319, 126569.
Gonzalez-Rios, O., Suarez-Quiroz, M.L., Boulanger, R. et al.
(2007a). Impact of “ecological” postharvest processing on coffee
aroma: II. Roasted coffee. Journal of Food Composition and Analy-
sis,20, 297–307.
Gonzalez-Rios, O., Suarez-Quiroz, M.L., Boulanger, R. et al.
(2007b). Impact of “ecological” postharvest processing on the vola-
tile fraction of coffee beans: I. Green coffee. Journal of Food Com-
position and Analysis,20, 289–296.
Haile, M., Bae, H.M. & Kang, W.H. (2020). Comparison of the
antioxidant activities and volatile compounds of coffee beans
obtained using digestive bio-processing (elephant dung coffee) and
commonly known processing methods. Antioxidants,9, 408.
Haile, M. & Kang, W.H. (2019). The harvest and postharvest man-
agement practices’ impact on coffee quality. In: Castanheira, D.T.
(Ed.) Coffee-Production and Research. Pp. 1–18. London: BoD –
Books on Demand.
Hameed, A., Hussain, S.A., Ijaz, M.U., Ullah, S., Pasha, I. & Sule-
ria, H.A.R. (2018). Farm to consumer: Factors affecting the
organoleptic characteristics of coffee. II: Postharvest processing
factors. Comprehensive Reviews in Food Science and Food Safety,
17, 1184–1237.
Hayakawa, F., Kazami, Y., Wakayama, H. et al. (2010). Sensory
lexicon of brewed coffee for Japanese consumers, untrained coffee
professionals and trained coffee tasters. Journal of Sensory Studies,
25, 917–939.
Hu, G., Peng, X., Gao, Y. et al. (2020). Effect of roasting degree of
coffee beans on sensory evaluation: Research from the perspective
of major chemical ingredients. Food Chemistry,331, 127329.
Hunaefi, D., Khairunnisa, W., Fauzan, Z. & Adawiyah, D. (2020).
Sensory profile of commercial coffee products using QDA (Quanti-
tative Descriptive Analysis), flash profile, and CATA (check-all-
that_apply) methods. In: Indonesia. Pp. 20–30. Bogor. SEAFAST
International Seminar-Facing Future Challenges: Sustainable Food
Safety, Quality and Nutrition.
Iwasa, K., Seta, H., Matsuo, Y. & Nakahara, K. (2021). Evaluation
of 3-methylbutanoic acid methyl ester as a factor influencing flavor
cleanness in arabica specialty coffee. Applied Sciences,11, 5413.
Iwasa, K., Setoyama, D., Shimizu, H. et al. (2015). Identification of
3-methylbutanoyl glycosides in green Coffea arabica beans as cau-
sative determinants for the quality of coffee flavors. Journal of
Agricultural and Food Chemistry,63, 3742–3751.
Janissen, B. & Huynh, T. (2018). Chemical composition and value-
adding applications of coffee industry by-products: A review.
Resources, Conservation and Recycling,128, 110–117.
Jo€
et, T., Laffargue, A., Descroix, F., Doulbeau, S., Bertrand, B. &
Dussert, S. (2010). Influence of environmental factors, wet
processing and their interactions on the biochemical composition
of green arabica coffee beans. Food Chemistry,118, 693–701.
Junior, D.B., Guarc
ßoni, R.C., Da Silva, M.D.C.S. et al. (2021).
Microbial fermentation affects sensorial, chemical, and microbial
profile of coffee under carbonic maceration. Food Chemistry,342,
128296.
Keast, R.S. & Costanzo, A. (2015). Is fat the sixth taste primary?
Evidence and implications. Flavour,4,1–7.
Kemp, S.E., Hollowood, T. & Hort, J. (2011). Sensory Evaluation: A
Practical Handbook. Hoboken, NJ: John Wiley & Sons.
Kiani, S., Minaei, S. & Ghasemi-Varnamkhasti, M. (2016). A porta-
ble electronic nose as an expert system for aroma-based classifica-
tion of saffron. Chemometrics and Intelligent Laboratory Systems,
156, 148–156.
Kitzberger, C.S.G., Scholz, M. & Da Silva, J. (2016). Free choice
profiling sensory analysis to discriminate coffees. AIMS Agric
Food,1, 455–469.
Knopp, S., Bytof, G. & Selmar, D. (2006). Influence of processing
on the content of sugars in green arabica coffee beans. European
Food Research and Technology,223, 195–201.
Knysak, D. (2017). Volatile compounds profiles in unroasted Coffea
arabica and Coffea canephora beans from different countries. Food
Science and Technology,37, 444–448.
Krajangsang, S., Mahingsapun, R., Tantayotai, P., Panyachanakul,
T., Samosorn, S., Dolsophon, K., Jiamjariyatam, R., Lorliam, W.
& Srisuk, N. 2022. Enhancement of arabica coffee quality with
selected potential microbial seed culture under controlled fermenta-
tion in wet process. Available at SSRN 4008103.
L€
aderach, P., Ramirez-Villegas, J., Navarro-Racines, C., Zelaya, C.,
Martinez-Valle, A. & Jarvis, A. (2017). Climate change adaptation
of coffee production in space and time. Climatic Change,141,47–62.
Landis, B.N., Frasnelli, J., Reden, J., Lacroix, J.S. & Hummel, T.
(2005). Differences between orthonasal and retronasal olfactory
functions in patients with loss of the sense of smell. Archives of
Otolaryngology–Head & Neck Surgery,131, 977–981.
Le, Q.V., Jovanovic, G., Le, D.-T. & Cowal, S. (2020). Understand-
ing the perceptions of sustainable coffee production: A case study
of the K’Ho ethnic minority in a small village in L^
am
Dồng pro-
vince of Vietnam. Sustainability,12, 1010.
Lee, C., Lee, Y., Lee, J.-G. & Buglass, A.J. (2013a). Development of
a simultaneous multiple solid-phase microextraction-single shot-gas
chromatography/mass spectrometry method and application to
aroma profile analysis of commercial coffee. Journal of Chromatog-
raphy. A,1295,24–41.
Lee, L.W., Tay, G.Y., Cheong, M.W., Curran, P., Yu, B. & Liu,
S.Q. (2017a). Modulation of the volatile and non-volatile profiles
of coffee fermented with Yarrowia lipolytica: I. Green coffee. LWT,
77, 225–232.
Lee, L.W., Tay, G.Y., Cheong, M.W., Curran, P., Yu, B. & Liu,
S.Q. (2017b). Modulation of the volatile and non-volatile profiles
of coffee fermented with Yarrowia lipolytica: II. LWT,80,32–42.
Lee, M.J., Kim, S.E., Kim, J.H., Lee, S.W. & Yeum, D.M. (2013b).
A study of coffee bean characteristics and coffee flavors in relation
to roasting. Journal of the Korean Society of Food Science and
Nutrition,42, 255–261.
Lee, S.J., Kim, M.K. & Lee, K.-G. (2017). Effect of reversed coffee
grinding and roasting process on physicochemical properties
including volatile compound profiles. Innovative Food Science &
Emerging Technologies,44,97–102.
Lingle, T.R. & Menon, S.N. (2017). Cupping and grading—discover-
ing character and quality. In: The Craft and Science of Coffee (edi-
ted