Content uploaded by Adriana Farah
Author content
All content in this area was uploaded by Adriana Farah on Oct 22, 2020
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=batc20
Download by: [UNESP] Date: 15 April 2016, At: 11:58
Critical Reviews in Analytical Chemistry
ISSN: 1040-8347 (Print) 1547-6510 (Online) Journal homepage: http://www.tandfonline.com/loi/batc20
Coffee Adulteration: More than Two Decades of
Research
Aline Theodoro Toci, Adriana Farah, Helena Redigolo Pezza & Leonardo
Pezza
To cite this article: Aline Theodoro Toci, Adriana Farah, Helena Redigolo Pezza & Leonardo
Pezza (2016) Coffee Adulteration: More than Two Decades of Research, Critical Reviews in
Analytical Chemistry, 46:2, 83-92, DOI: 10.1080/10408347.2014.966185
To link to this article: http://dx.doi.org/10.1080/10408347.2014.966185
Accepted author version posted online: 29
Jan 2015.
Submit your article to this journal
Article views: 122
View related articles
Citing articles: 1 View citing articles
Coffee Adulteration: More than Two Decades of Research
Aline Theodoro Toci
a
, Adriana Farah
b
, Helena Redigolo Pezza
a
, and Leonardo Pezza
a
a
Instituto de Qu
ımica, UNESP-Universidade Estadual Paulista “J
ulio de Mesquita Filho,”Araraquara, Brazil;
b
N
ucleo de Pesquisa em Caf
e Prof. Luiz Carlos
Trugo, Instituto de Nutri¸c~
ao, UFRJ-Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil
ABSTRACT
Coffee is a ubiquitous food product of considerable economic importance to the countries that produce
and export it. The adulteration of roasted coffee is a strategy used to reduce costs. Conventional methods
employed to identify adulteration in roasted and ground coffee involve optical and electron microscopy,
which require pretreatment of samples and are time-consuming and subjective. Other analytical
techniques have been studied that might be more reliable, reproducible, and widely applicable. The
present review provides an overview of three analytical approaches (physical, chemical, and biological) to
the identification of coffee adulteration. A total of 30 published articles are considered. It is concluded that
despite the existence of a number of excellent studies in this area, there still remains a lack of a suitably
sensitive and widely applicable methodology able to take into account the various different aspects of
adulteration, considering coffee varieties, defective beans, and external agents.
KEYWORDS
Analytical methods; coffee
adulteration; coffee quality
Introduction
Coffee is globally one of the most widely used food products
and is vital to the economies of countries involved in its pro-
duction and export. In the past two decades there has been con-
tinual growth in the global consumption of coffee, driven by
new products and drink formulations, new domestic coffee
machines, and high street coffee shops, together with changes
in the profile of the typical consumer (Associa¸c~
ao Brasileira da
Ind
ustria de Caf
e [ABIC], 2014). Accompanying this trend, a
growing number of articles have been published concerning the
adulteration of coffee, since growth in the market has made it
necessary to implement regulation and control strategies.
The adulteration of roasted coffee is both frequent and
diversified. It can involve the quality of the beans (considering
species, geographical origin, and defective beans), as well as the
addition of other substances (coffee husks and stems, maize,
barley, chicory, wheat middlings, brown sugar, soybean, rye,
triticale, and a¸ca
ı) to coffee blends in order to make them less
expensive. There are nearly 100 coffee species in the world; the
most commercialized are Coffea arabica (arabica coffee) and
Coffea canephora (robusta coffee), with arabica having higher
commercial value due to its aromatic superiority. Among these
species, there are a variety of cultivars that can be grown and
spread throughout the world, such as the Typica and Bourbon
varieties (arabica), but there are also some more linked to their
producing countries, such as Novo Mundo and Catuai (Brazil),
Jimma and Harar (Ethiopia), and Villa Sarchi (Costa Rica).
The quality of the fresh coffee beans, the proportion of defec-
tive beans, and the type of roasting and grinding process all
influence the final product (Toci and Farah, 2008,2014;Toci
et al., 2008). Due to these factors, and considering that after
roasting and grinding, the addition of others materials cannot
be detected visually (ABIC, 2014), investigation of the adultera-
tion of roasted and ground coffee is highly complex.
For the regulation of coffee adulteration, various different
norms have been adopted by national and international com-
mittees and organizations. For example, the International Cof-
fee Council (ICC) published Resolution 399 on 24 May 2001,
encouraging ICO member countries (45 members) to take
measures to remove defective coffees from the marketplace. A
document was issued outlining a framework for action needed
to implement a Coffee Quality-Improvement Programme
(CQP). The Programme sets target standards for exportable
coffee, stating that exporting countries shall strive not to export
coffee that has the following characteristics: for arabica, in
excess of 86 defects per 300 g of sample (New York green coffee
classification/Brazilian method, or equivalent); for robusta, in
excess of 150 defects per 300 g (Vietnamese, Indonesian, or
equivalent). The main concern about defective beans is the pos-
sible presence of ochratoxin A (OTA). To this end, the Food
and Agriculture Organization (FAO) was appointed to oversee
a project sponsored by the ICO, with the aim of trying to elimi-
nate or reduce the formation of molds. In 2004, the Brazilian
Association of Coffee Industries (ABIC), together with the Bra-
zilian Deliberative Council for Coffee Policy (CDPC), created
the Coffee Quality Program (PQC). This program certifies the
purity of roasted coffees, which is regulated by the Ministry of
Agriculture, Livestock, and Supply (MAPA). Normative
Instruction number 16 (24 May 2010) was issued with the
objective of ensuring the quality of roasted ground coffee, and
sets a maximum 1% limit for the combined total content of
impurities, including husks and stems, sediment (stones,
agglomerates, and sand), and other substances (such as maize,
rye, barley, a¸ca
ıseeds, and sugar). The PQC also sets the quality
CONTACT Aline Theodoro Toci alinettoci@iq.unesp.br Rua Prof. Francisco Degni 55, Jardim Quitandinha, 14800-060, Araraquara, SP, Brazil.
© 2016 Taylor and Francis Group, LLC
CRITICAL REVIEWS IN ANALYTICAL CHEMISTRY
2016, VOL. 46, NO. 2, 83–92
http://dx.doi.org/10.1080/10408347.2014.966185
Downloaded by [UNESP] at 11:58 15 April 2016
of coffee, using the classifications traditional, superior, and
gourmet. In addition to considering the sensorial characteristics
of the drink, this program sets a limit of 40% for the quantity of
defective coffee beans. However, this program does not yet
cover the entire national territory, because it is a private initia-
tive that is not supported by a segment of the coffee roasting
industry, notably the market-leading brands that do not require
certificates to improve their visibility.
Global data about coffee adulteration are practically nonex-
istent, mainly because it involves the domestic economic situa-
tion of each country. In Brazil, which is the world’s largest
coffee producer, an inspection conducted by ABIC analyzed
2400 brands, and among these 583 were adulterated with husks,
maize, rye, a¸ca
ıseeds, or brown sugar, representing 25% of the
national brands (Peixoto, 2009). Most of these brands were not
certified by ABIC. However, this problem is even greater in
some Brazilian states, as in the state of Minas Gerais, which
owns 50% of the Brazilian coffee production. In this state, the
adulterations reach 47% of regional brands (Peixoto, 2009).
The conventional methods that are most widely used in lab-
oratories to identify the adulteration of roasted and ground cof-
fee involve the use of optical and electron microscopy.
Complementary physicochemical analyses are also performed,
including moisture content, mineral residues, ether-extractable
substances, and caffeine. The analyses based on microscopy are
frequently slow and subjective, and can produce conflicting
results. Other techniques have therefore been studied in order
to provide analyses that are more reliable and reproducible and
able to identify the many different types of possible adultera-
tion. These alternative techniques include chromatographic
analysis and infrared spectroscopy. More recently, polymerase
chain reaction (PCR) techniques have been used to characterize
the DNA of samples.
The present review considers the physical, chemical, and
biological techniques that have been used to identify coffee
adulteration. A total of 30 articles are described (Table 1),
almost all of which used statistical tools to complement the
analytical technique (these statistical tools are not the focus of
the present work).
Spectrometric methods
Microscopy
Starting in the 1950s, methods based on optical microscopy
were introduced for the detection of adulterated roasted and
ground coffee. These techniques require sample pretreatment,
such as degreasing with an organic solvent, drying, and sifting
(Menezes and Bicudo, 1958). Optical microscopy is based on
the use of objective and ocular lenses to focus on the substance
being studied. The analyses depend on the degree of agreement
obtained between the characteristics of the unknown substance
and those of roasted coffee particles. Resolution can be
improved by illuminating the object with radiation at a wave-
length shorter than that of visible light. The depth of field is
inversely proportional to the magnification, so that perfect
smoothness of the surface under observation is ideally required,
which is not feasible for coffee analysis. In this case, the micro-
scope slides are prepared using chemical reagents, and the
quantification of impurities is based on comparison of the
aqueous extract percentage of the sample with that of pure cof-
fee (Menezes and Bicudo, 1958). The method requires consid-
erable technical ability, and is therefore subjective.
Scanning electron microscopy (SEM) uses a beam of elec-
trons in place of the photons used in conventional optical
microscopy. SEM can provide rapid information concerning
the morphology of a solid sample, as well as identify the chemi-
cal elements present. The principle of SEM is based on the use
of a small-diameter beam of electrons to explore the surface of
the sample, point by point, along successive lines, and transmit
the detector signal (which can consist of electrons or photons)
to a cathodic screen whose scan rate is perfectly synchronized
with that of the incident beam. The method does not require
prior sample preparation, but as in optical microscopy, it
entails a series of comparisons between samples and potential
adulterants (Lopez, 1983), which makes it as subjective as opti-
cal microscopy.
Infrared spectroscopy
Spectrometric techniques, previously used only for the identifi-
cation of analytes, have been greatly improved with the intro-
duction of the direct injection (direct infusion) technique,
which avoids the need for prior separation of analytes and pro-
vides an accurate digital representation of the sample. Infrared
(IR) spectroscopy is one of the most widely used physical tech-
niques for the identification of adulterants. Its advantages
include reduced analysis time, less sample manipulation, and
less chemical waste generation than conventional methods.
Although the IR spectra of coffee products are complex, due to
strong overlapping of peaks originating from many chemical
species, studies have demonstrated the viability of this analyti-
cal technique for the detection of adulteration in roasted coffee.
Briandet et al. (1996) presented an efficient method using Four-
ier transform-infrared (FT-IR) spectroscopy for the detection
of adulteration of freeze-dried instant coffees. Adulteration
with glucose, starch, and chicory was investigated using two
different FT-IR procedures: diffuse reflectance and attenuated
total reflectance. Three different statistical treatments of the
spectra were carried out. First, the spectra were compressed by
principal component analysis (PCA), and a linear discriminant
analysis (LDA) was performed. With this approach, a 98% clas-
sification success rate was achieved. Second, a simultaneous
partial least squares (PLS) regression was carried out for the
content of three added carbohydrates (xylose, glucose, and
fructose) in order to assess the potential of FT-IR spectroscopy
for determining the carbohydrate profile of instant coffee. Last,
the discrimination of pure from adulterated coffee was per-
formed using an artificial neural network (ANN), which
achieved a perfect assignment rate. The performance of the
ANN was validated using an independent data set and 100%
correct classification was once again achieved.
Pizarro et al.(2007) studied the adulteration of C. arabica
blends with the C. robusta variety using near-infrared spectros-
copy (NIRS) combined with multivariate calibration methods.
A total of 108 arabica blends were used, with C. robusta percen-
tages ranging from 0 to 60% (w/w). A method employing PLS
regression and wavelet-based preprocessing was developed,
84 A. T. TOCI ET AL.
Downloaded by [UNESP] at 11:58 15 April 2016
called OWAVEC, which was used to simultaneously perform
two crucial preprocessing steps in multivariate calibration: sig-
nal correction and data compression. Several other preprocess-
ing methods (mean centering, first derivative, and two
orthogonal signal correction methods, OSC and DOSC) were
additionally applied in order to obtain calibration models with
the best possible predictive capacity and to evaluate the perfor-
mance of the OWAVEC method, comparing the qualities of
different regression models. The calibration model developed
after preprocessing derivative spectra using OWAVEC pro-
vided high-quality results (0.79% root mean squared error of
prediction [RMSEP]). The percentage of the C. robusta variety
present was predicted with greater reliability than with models
constructed from the raw spectra or using other orthogonal sig-
nal correction methods. The OWAVEC model also offered
greater simplicity. The work reported was only a feasibility
study, and the authors suggested that further research would be
needed before it could be used for authentication purposes.
More recently, Tavares et al. (2012) used mid-infrared spec-
troscopy to distinguish 13 coffee blends containing different
percentages (0.5 to 30%) of coffee husks. Samples adulterated
with husks were identified by PCA, and quantitative estimation
of adulteration was achieved by PLS regression. The minimum
quantity detectable by this method (0.5%) was lower than the
maximum level (1.0%) of impurities permitted by law.
Reis et al. (2013) evaluated the feasibility of employing dif-
fuse reflectance infrared Fourier transform spectroscopy
(DRIFTS) for discrimination among roasted coffee, roasted
maize, and coffee husks. Arabica coffee beans, coffee husks,
and ground maize kernels were submitted to light, medium,
and dark roasts. Principal component analysis of the DRIFTS
spectra enabled separation of the samples into three groups:
coffee, coffee husks, and maize. The calibration set consisted of
a total of 116 samples: 33 samples of roasted coffee, 27 samples
of roasted coffee husks, 30 samples of roasted maize, and 26
samples of adulterated coffee, with adulteration levels ranging
from 10 to 50% of one or both adulterants. Classification mod-
els based on linear discriminant analysis provided complete
discrimination (100% recognition and prediction) among
roasted coffee, pure adulterants (maize and coffee husks), and
adulterated coffee samples.
Multispectral imaging
Assad et al. (2002) developed a method based on reflectance to
identify adulteration with barley, maize, and stems at percen-
tages ranging from 1 to 50%. The method was based on image
analysis, considering that different organic materials found in
ground coffees present distinct spectral signatures. Multispec-
tral images of coffee samples were generated using a glass mag-
nifier connected to a charge coupled device (CCD) camera. The
camera captured images of the sample in visible spectral bands
(RGB: red, green, and blue), which were then analyzed using a
digital image processing program to calculate the relative pro-
portions of the components of the sample. Capture of the
images was fast, but sample preparation steps were required,
including cleaning, drying, and homogenization. Nevertheless,
this method achieved a minimum accuracy of 95% for quantifi-
cation of adulteration in powdered coffee. In an extension of
the work, the same method was proposed for the identification
of adulteration in arabica coffee blends mixed with coffee husks
and straw, maize, brown sugar, and soybean (Sano et al., 2003).
Table 1. List of papers with different analytical approaches (physical, chemical, and biological) to the identification of coffee adulteration.
Method No. Technique used Authors Year Adulterants investigated
Physical 1Infrared Briandet et al. 1996 Glucose, starch, chicory
2 Pizarro et al. 2007 Robusta
3 Tavares et al. 2012 Husks
4 Reis et al. 2013 Husks, maize
5Multispectral imaging Assad et al. 2002 Barley, maize, stems
6 Sano et al. 2003 Coffee husks and straw, maize, brown sugar, soybean
7 Gon¸calves et al. 2007 Husks, straw
8Mass spectroscopy Amorim et al. 2009 Defective seeds
9 Garrett et al. 2012 Robusta
10 Photoacoustic Cesar et al. 1984 Coffee husks, maize, barley
11 Thermal lens Fontes et al. 2001 Maize
12 Photothermal Fontes et al. 2006 Maize
13 NMR Ciampa et al. 2010 Robusta
Chemical 14 HPLC-UV Blanc et al. 1989 Coffee husks, maltodextrin, caramelized sugar
15 Davis et al. 1990 Coffee husks
16 HPLC-UV-vis Pauli et al. 2011 Coffee husks, starch
17 HPAEC Stober et al. 2001 Legumes
18 HPAEC-PAD Garcia et al. 2009 Coffee husks, maize
19 HPLC-HPAEC-PAD Domingues et al. 2014 Triticale, a¸ca
ı
20 HPLC with fluorescence Jham et al. 2007 Maize
21 DH-GC-MS Ruiz et al. 1995 Maize, barley
22 GC Valdenebro et al. 1999 Robusta
23 Jham et al. 2008 Maize
24 SPME-GC-MS Toci and Farah 2008 Defective coffee seeds
25 Toci and Farah 2014 Defective coffee seeds
26 Oliveira et al. 2009 Barley
Biological 27 PCR Martellossi et al. 2005 Robusta
28 Ferreira et al. 2012 Maize, rice, barley
29 Spaniolas et al. 2006 Robusta
30 Spaniolas et al. 2008 Robusta
CRITICAL REVIEWS IN ANALYTICAL CHEMISTRY 85
Downloaded by [UNESP] at 11:58 15 April 2016
The following concentrations of contaminants were prepared:
10, 20, 30, 40, 50, 100, 200, 300, 400, and 500 g/kg. The percent-
age areas of the contaminants in each image were calculated by
the maximum likelihood supervised classification technique.
Best-fit equations relating weight percentages (g/kg) to the per-
centage areas were obtained for each coffee contaminant. To
test the method, 247 coffee samples with different amounts and
types of adulterants were analyzed in the laboratory. It was con-
cluded that the method could provide fast, nondestructive, and
precise analysis of a wide range of adulterants present at differ-
ent concentrations in ground coffee powders. The high correla-
tions obtained for samples containing two different
contaminants also indicated that the method could be success-
fully employed when more than one adulterant was present in
the sample. This method was subsequently validated for the
selective identification of husks and straws (Gon¸calves et al.,
2007), with detection and quantification limits of 0.03 and
0.05%, respectively. The method was considered to be valid,
although for regulatory purposes it was suggested that precision
in the concentration range 0.2–2.2% would need to be
improved.
Mass spectroscopy
Direct infusion electrospray ionization mass spectrometry, in
both negative (ESI(–)-MS) and positive (ESI(C)-MS) ion
modes, has been used to distinguish pure green and roasted
arabica coffees according to bean characteristics (green, ripe,
and overripe [defective seeds]) and post-harvesting processes
(dry, wet, and semi-wet), as well as coffees with different cup
qualities (Amorim et al., 2009). Statistical analysis using PCA
showed that in the ESI(–)-MS mode, ions from chlorogenic
acids and short-chain organic acids derived from sugars were
most important for the discrimination of defective beans. In
the ESI(C)-MS mode, discrimination mainly employed low m/
zions, such as protonated pyridine and alkylpyridines resulting
from trigonelline degradation. Preliminary results showed that
both ESI(C)-MS and ESI(–)-MS modes were able to differenti-
ate the cup qualities of roasted arabica coffees, and the ions
used to perform discrimination were the same as those
employed in ripeness and post-harvesting evaluations.
In subsequent work by the same group, Garrett et al. (2012)
evaluated the use of direct-infusion ESIMS data, combined
with the PLS multivariate calibration technique, to detect and
quantify the adulteration of arabica coffee by robusta coffee.
Five robusta/arabica blends, with robusta percentages of 0, 25,
50, 75, and 100%, were used for the construction of a calibra-
tion curve, and samples with robusta concentrations of 20, 40,
60, and 80% were evaluated. A total of 16 PLS models were
built using ESI(§) quadrupole time-of-flight (QTOF) and ESI
(§) Fourier transform ion cyclotron resonance (FT-ICR) MS
data for hot aqueous extracts of certified coffee samples. ICR is
a phenomenon related to the movement of ions in a magnetic
field. It is used for accelerating ions in a cyclotron and for mea-
suring the masses of ionized analytes in mass spectrometry.
The model using the 30 most abundant ions detected by ESI
(C) FT-ICR MS produced the most accurate coffee blend per-
centage prediction and was later successfully employed to pre-
dict the blend composition of commercial robusta and arabica
coffees. In addition, ESI(§) FT-ICR MS analysis enabled the
identification of 22 compounds in the arabica coffee and 20
compounds in the robusta coffee, most of which were
phenolics.
Other spectroscopic methods
Other promising physical methods that have been proposed
over the years include photoacoustic, photothermal, thermal
lens, and nuclear magnetic resonance (NMR) techniques. Some
of these methodologies have not yet been fully studied or vali-
dated. Nevertheless, they deserve attention due to the good
results that have been achieved, as described below.
The photoacoustic (PA) effect has been recognized in the
past few years as an important tool for studying the optical
absorption properties of crystalline, powdered, and amorphous
solids. Cesar et al. (1984) used this method to evaluate the adul-
teration of coffee by coffee husks, maize, and barley. The suc-
cess of this spectroscopic technique is essentially due to the fact
that only the absorbed light is converted into pressure fluctua-
tions in the gas cell. The primary source of the acoustic signal
in the cell arises from the periodic heat flow from the solid to
the surrounding gas as the solid is cyclically heated by the
absorption of pulsed light. Only a relatively thin layer of gas
adjacent to the surface of the solid is assumed to respond ther-
mally to the periodic heat flow from the solid to the surround-
ing gas. This boundary layer of gas then acts as a piston,
generating the acoustic signal detected by the microphone.
Since the magnitude of the periodic pressure fluctuations in the
cell is proportional to the amount of heat emanating from the
solid absorber, there is a close correspondence between the
strength of the acoustic signal and the amount of light absorbed
by the solid. The feasibility of this method for detection of cof-
fee adulteration was demonstrated. The technique avoided any
need for sample preparation and required only that the sample
exhibit a fairly homogeneous particle size distribution. A linear
relationship was obtained between the PA signal and the mass
concentrations of all the adulterants, and excellent mathemati-
cal models were obtained for maize and husks. However, for
barley, the mathematical model was not as satisfactory, perhaps
due to the choice of a response wavelength very close to that of
the pure coffee. Although it was possible to establish an ade-
quate methodology for the detection of different adulterants in
powdered coffee, there remain difficulties related to sample
uniformity and compaction. In order to obtain compact and
uniform coffee samples, it is necessary to control not only the
grain size and compaction pressure, but also the moisture con-
tent of the sample.
Considering the problems involved in sample uniformity,
the same group published two other studies. The first investi-
gated the viability of thermal lens spectrometry (TLS) to detect
the presence of an adulterant (maize) in brewed coffee (Fontes
et al., 2001). Briefly, this technique involves the illumination of
a sample by a modulated light beam and subsequent measure-
ment of the temperature fluctuation induced in the sample as a
result of non-radiative de-excitation processes within the mate-
rial. Since the photothermal signal is derived only from the
absorbed light, the effects of scattered light play no significant
role in these spectroscopic techniques, which should therefore
86 A. T. TOCI ET AL.
Downloaded by [UNESP] at 11:58 15 April 2016
be suitable for studies involving powdered samples. Thermal
lens experiments were performed using brewed coffee made
with a pure commercial coffee and a commercial coffee mixed
with 4 wt.% of roasted maize meal. The concentrations of the
samples investigated (pure and adulterated) ranged from 0.04
to 20 wt.%. The results indicated that the dn/dT (the analytical
expression for absolute determination of the thermo-optical
properties of the sample) and pH values exhibited similar
behavior as a function of the coffee brew concentration and
were equally sensitive for detecting the presence of adulterants.
Even though these preliminary experiments were limited to
well-controlled samples with a single adulterant concentration,
the results indicated that the combination of the two detection
techniques (thermal lens spectrometry and pH measurements)
could be useful for the routine detection of coffee adulterants,
once a wider range of adulterants and concentrations has been
tested.
The second work reported an alternative photothermal
approach for analysis of the adulteration of coffee by maize
(Fontes et al., 2006), which was actually an extension of the
work initiated in 2001. The approach differed from the previ-
ous photothermal study in two fundamental ways. First, the
proposed method was non-spectroscopic, in that no wave-
length scanning was performed. Instead, the parameter moni-
tored was the temperature coefficient of the refractive index of
the coffee brews at the fixed pumping beam wavelength. Sec-
ond, the new method was based on the use of coffee brews,
instead of powdered samples. This not only eliminated the
need for careful control of sample compaction, but at the same
time also ensured that the samples used were always homoge-
neous. The thermal lens analyses were accompanied by meas-
urements of the pH of the different coffee brews. From a
quantitative perspective, the results suggested that the monitor-
ing of dn/dT as a function of the coffee brew concentration
could discriminate between pure and adulterated brews for
brew concentrations above 0.6%. Even though the results pre-
sented in this article were limited to well-controlled samples,
the thermal lens and pH data indicated that the combination of
these two techniques could be successfully applied to detect
adulterants in brewed coffee.
Finally, Ciampa et al. (2010) demonstrated the potential of
high-resolution magic angle spinning nuclear magnetic reso-
nance (HR-MAS NMR) spectroscopy for discrimination
between arabica and robusta coffees. The innovative approach
to high-resolution magic angle spinning involved the use of a
novel NMR probe head that enabled highly resolved spectra to
be obtained for gel-like or suspended samples. Variations of the
concentrations of relevant species (caffeine, trigonelline, sugar,
amino acids, acrylamide, pyrazines, melanoidins, and chloro-
genic acids) were monitored as a function of roasting tempera-
ture (from green to completely roasted beans). The results
showed that monitoring of analytes such as caffeine, chloro-
genic acids, and sugars might be able to be used to predict the
amount of robusta coffee in the blends. Nonetheless, further
studies are still needed to confirm this possibility. The HR-
MAS NMR technique was demonstrated to be a powerful tool
for fast chemical composition measurements, opening up per-
spectives for novel applications of this approach in coffee qual-
ity control.
Instrumental separation methods (chromatographic
methods)
The most widely used chemical techniques reported for the
analysis of coffee adulteration are liquid chromatography
(seven relevant studies) and gas chromatography (five studies).
Carbohydrates are the analytes most frequently used to distin-
guish between pure and adulterated coffee.
High-performance liquid chromatography (HPLC)
One of the first studies to employ liquid chromatography for
analysis of carbohydrates in order to identify adulteration was
undertaken by Blanc et al. (1989). HPLC-UV was used to char-
acterize the carbohydrate profiles of 122 soluble coffee samples,
pure and adulterated with husks, maltodextrin, and caramel-
ized sugar. The results showed that pure soluble coffee con-
tained maximum levels of ~0.3% total xylose and sucrose, no
maltose, and about 2% total glucose. Higher levels of total
xylose could be explained by the co-extraction of coffee husks
or parchment, while the levels of free fructose and glucose dis-
tinguished whether unroasted or roasted husks/parchments
had been added. Elevated levels of maltose and total glucose
indicated the addition of maltodextrins, and high levels of
sucrose and total glucose reflected the addition of caramelized
sugar. It was concluded that analyses of both free and total
sugar contents were required in order to obtain information on
the nature of the adulterants. In further work by the same
group (Davis et al., 1990), mannitol, a polyhydric sugar alcohol,
was identified in coffee products for the first time using HPLC.
Mannitol is frequently found in exudates of plants such as flow-
ering ash, olive, and plane trees, and is present in marine algae
at concentrations in excess of 20%. It was identified in the car-
bohydrate fraction of dried coffee husks at concentrations of
1.61–2.03%. Its presence in some commercial soluble coffees at
levels above 0.30% was indicative of adulteration by coffee
husks. A total of 145 samples from different countries were
evaluated, and it was concluded that the majority of soluble cof-
fee powders sold globally were manufactured from good quality
coffees, although more than half the samples contained manni-
tol at concentrations exceeding 0.3%.
In 2001, Stober et al. published a preliminary study con-
cerning the detection of legume tissues in soluble coffees by
oligosaccharide profiling combined with protein analyses,
using high-performance anion exchange chromatography
(HPAEC; Stober et al., 2001). This type of chromatography
separates carbohydrates according to specific interactions
between the hydroxyl groups of the glycan and the stationary
phase of the column, at high pH. The glycans are separated
chromatographically as anionic species, and their interaction
with the column is based on their size, composition, and link-
age. Twenty soluble authentic coffees and blends from differ-
ent manufacturers were analyzed. A peak with the retention
time of stachyose (a tetrasaccharide typically found in legumi-
nous plants) was identified in the chromatograms for the
commercial coffee blends, corresponding to concentrations
ranging from about 0.1 to 0.2 g/100 g. No peaks with the
same retention time were observed in the HPAEC chromato-
grams for the authentic soluble coffees. In order to identify
CRITICAL REVIEWS IN ANALYTICAL CHEMISTRY 87
Downloaded by [UNESP] at 11:58 15 April 2016
the stachyose, purification of the oligosaccharide fraction was
performed by classical reversed-phase (RP)-HPLC, followed
by methylation and analysis by GC-MS. It was found that
about 50% of this fraction consisted of stachyose. The authors
concluded that the unusual HPAEC oligosaccharide profile,
together with high protein content and the confirmed pres-
ence of stachyose, was indicative of the use of an adulterant
derived from a legume. This could lead to a new tool for the
screening of soluble coffee authenticity.
Other work published in 2007 used HPLC with fluorescence
detection for the quantification of g-tocopherol, used as a
marker for the adulteration of Brazilian coffee (Coffea arabica
L.) with maize (Jham et al., 2007). This study characterized the
profiles of a-, b-, g-, and d-tocopherol in six coffee varieties
and six maize samples. The results showed a higher concentra-
tion of g-tocopherol in maize samples (91.3 mg/kg) than in
pure coffee samples (61.7 mg/kg). Evaluation was also made of
four coffee/maize blends prepared in the laboratory, together
with six commercial blends. The results indicated that
g-tocopherol could be used as a marker for the adulteration of
Brazilian coffee with maize. Nevertheless, further detailed
examination would be needed of factors that could affect toco-
pherols, such as coffee variety, sample origin, storage, process-
ing, and the presence of other adulterants.
Garcia et al. (2009) reported the carbohydrate profiles of
roasted and ground arabica coffees and their adulteration by
coffee husks and maize, using high-performance anion
exchange chromatography with pulsed amperometric detection
(HPAEC-PAD). In this technique, non-derivatized analytes are
detected by applying various potentials to the working elec-
trode over a specific time period. The hydroxyl groups in car-
bohydrates are oxidized on the electrode surface and the
resulting current is measured. Sensitivity at picomole to femto-
mole levels can be achieved with PAD, making it one of the
most sensitive detection techniques. The results showed higher
levels of galactose and mannose in pure coffee, with concentra-
tions of 8.25 and 9.65% (w/w), respectively. However, in
pure coffee husks, the main carbohydrates were mannitol
(0.64%), arabinose (4.24%), and xylose (3.40%). The highest
concentration of glucose was detected in the maize sample
(52.53%, w/w). Chemometric methods were applied in order to
identify patterns for adulteration by coffee husk and maize,
with different amounts of these contaminants added to the cof-
fee in accordance with a simplex-centroid statistical design.
The results were used to produce linear models describing the
effects on carbohydrate levels caused by the addition of adulter-
ants to the coffee. Factor analysis, supported by hierarchical
cluster analysis, enabled the discrimination of four distinct
groups that differed according to the compositions of the mix-
tures of raw materials used.
In subsequent work, Pauli et al. (2011) validated a chro-
matographic method for the determination of total carbohy-
drates in soluble coffee, using a HPLC-UV-vis system with
post-column derivatization, in order to detect adulterant addi-
tions. The validated method was shown to be accurate and
robust. Adulteration with coffee husks was associated with
increases in the levels of xylose, glucose, and starchy products
in the samples, together with decreases of galactose and man-
nose, which are characteristic carbohydrates present at high
concentrations in soluble coffees produced from arabica and
robusta coffee beans.
Recently, the same group evaluated the performance of two
different chromatographic systems, HPLC-HPAEC-PAD and
HPLC-UV-vis with post-column derivatization, used for carbo-
hydrate determination (ISO method 11292) in order to detect
and quantify adulterants in coffee (Domingues et al., 2014).
Total carbohydrate analyses performed using both methodolo-
gies were effective in determining the concentrations of mono-
saccharides evaluated in roasted and ground coffee adulterated
with triticale and a¸ca
ı, considering the original constituents of
the matrices. Although the determination of carbohydrates
using the HPLC-UV-vis system with post-column derivatiza-
tion resulted in numerically different concentrations, with
poorer chromatographic resolution, sensitivity, and predictive
model fitting than the HPLC-HPAEC-PAD system, the former
technique is faster, easier to use, and could be performed in
most laboratories possessing a UV-vis detector. This system
therefore appears to offer the potential for use in routine qual-
ity control screening of adulterants in coffee. For the purposes
of quantification and forecasting by mathematical modeling,
the HPLC-HPAEC-PAD technique was shown to be superior,
but is more expensive and requires specialist knowledge of elec-
trochemical techniques. The simplex-centroid experimental
design for the three components of the mixtures (arabica coffee,
triticale, and a¸ca
ı) was used to obtain correlations between the
two chromatographic systems. Principal component analysis of
the carbohydrates revealed similar trends for each of the matri-
ces. Galactose was a characteristic component of the arabica
coffee matrix. Glucose and xylose were the predominant carbo-
hydrates in triticale, while at higher concentrations mannose
characterized the a¸ca
ımatrix.
Gas chromatography
The gas chromatography (GC) technique has also been used to
detect adulteration in roasted coffee, albeit to a less extent than
liquid chromatography. One of the first studies using this tech-
nique was undertaken by Ruiz et al. (1995). The profiles of the
volatile components of Colombian coffee at two grades of
roasting, together with those of roasted maize and roasted bar-
ley, were determined by dynamic headspace sampling followed
by GC separation and mass spectrometric detection. The vola-
tile components present at highest concentrations in the maize
samples were furfural and 5-methylfurfural, while in barley the
most prevalent compounds were 2-methylbutanal, 3-methylbu-
tanal, and furan. Sensorial analysis was also performed, which
enabled detection of adulteration with 20% cereal when the cof-
fee was roasted for 8 min. It was difficult to perceive the cereal
aroma when the roasting was performed for 10 min, especially
in the case of adulteration with barley. The authors suggested
that these compounds could be used as indicators of the adul-
teration of coffee with cereals in quality control procedures
designed to ensure the quality and purity of Colombian coffee.
Valdenebro et al. (1999) proposed a method for determina-
tion of the percentage of arabica coffee in blends of roasted
arabica/robusta coffee, based on their sterol contents. Thirteen
blends were prepared, using concentrations of arabica coffee
ranging from 40 to 100% (m/m). Commercial roasted coffee
88 A. T. TOCI ET AL.
Downloaded by [UNESP] at 11:58 15 April 2016
samples were also analyzed. The sterol content was determined
by extracting the coffee oil, saponifying the lipids, and separat-
ing the sterols present in the unsaponifiable fraction using
thin-layer chromatography. The sterols were then converted to
trimethylsilyl derivatives, which were analyzed by gas chroma-
tography. Twelve sterols were determined in roasted coffee
samples consisting of mixtures of the arabica and robusta clas-
ses. Using the sterol profiles in the blends as chemical descrip-
tors, application of principal component regression (PCR) then
provided a method for determination of the composition of
mixtures of roasted commercial coffees. The results showed
that when the descriptor D
5
avenasterol was used alone, it was
possible to predict the percentage of arabica in coffee mixtures
with greater reliability than with the use of the first PC (which
included inputs from all the variables and was therefore more
complex).
Using the same approach, Jham et al. (2008) studied the use
of fatty acid profiles as potential markers for the adulteration of
Brazilian roasted coffee with maize. Six C. arabica varieties
were used (Catuai, Catuca
ı, Bourbom, Mundo Novo, Rub
ı, and
Top
azio), together with three blends containing 5, 10, and 20%
of maize grains. The fatty acid methyl ester (FAME) composi-
tions of the different varieties were determined for the first
time. The method proved to be very fast, with complete charac-
terization (>99%) of the sample being possible in less than
6 min. Although the linoleic/stearic acid ratios were signifi-
cantly different for coffee and maize, it was suggested that in
order for a compound to serve as a marker of adulteration in
coffee, a very large difference (at least a factor of ~30) should
exist between its concentrations in pure and adulterated coffee
samples. Hence, this probe could not be used as a marker to
detect maize adulteration in commercial coffees.
Toci and Farah (2008,2014) conducted two studies using
solid-phase microextraction (SPME)-GC-MS to investigate
thevolatilecompoundprofiles of defective roasted coffee
beans in order to be able to detect substantial defects in
coffee blends. Application of the SPME-GC-MS technique
is straightforward and no previous sample preparation is
required, which makes it potentially promising for routine
chromatographic analysis. In general, defective beans
showed higher numbers and concentrations of volatile com-
pounds than control beans. These compounds included pyr-
azines, pyrroles, and phenols. Several potential volatile
compound markers for defective beans were proposed.
Butyrolactone and hexanoic acid were generally observed
only in raw and roasted defective beans, respectively; 3-
ethyl-2-methyl-1,3-hexadiene was a marker for raw black
beans; b-linalool and 2-butyl-3,5-dimethylpyrazine were
indicators for roasted defective beans in general; and 2-pen-
tylfuran was a marker for roasted black beans. An addi-
tional 16 compounds were suggested as indicators of poor
quality. In addition to demonstrating the ability of the tech-
nique to identify defective coffee beans, it was suggested
that the compounds used as indicators of quality could also
be used in quantitative investigations to determine the per-
centages of defective beans in commercial coffee blends.
Oliveira et al. (2009) evaluated the potential of the SPME-
GC-MS technique to detect the adulteration of ground roasted
coffee with roasted barley, using pure coffee and coffee mixed
with 1, 5, and 50% of barley. Discrimination between unadul-
terated and adulterated samples was achieved using PCA. Pyri-
dine was the substance that showed the greatest influence in
the case of roasted coffee, and the relative peak intensity
increased with roasting time. An unexpected finding was that
adulterated samples could be more easily discriminated at the
highest level of roasting, which enabled detection of adultera-
tion with as little as 1% (w/w) roasted barley in dark roasted
coffee samples. The authors also suggested that the use of aro-
matic compounds to identify adulterants was poorly explored
and inconclusive, and that further investigations were required.
Capillary electrophoresis
One of the first articles reporting the use of capillary electro-
phoresis to identify adulterants in coffee was published by
Nogueira and Lago (2009). The proposed method was based on
the controlled acid hydrolysis of the polysaccharides xylan and
starch present in some plant-based adulterants, followed by
analysis of the corresponding monosaccharides (xylose and
glucose, respectively). Acid hydrolysis by HCl increases the
ionic strength of the sample, which impairs the electrophoretic
separation. A neutralization step based on anion exchange resin
was therefore necessary. The best separations were obtained
using NaOH (80 mmol/L), cetrimonium bromide (CTAB)
(0.5 mmol/L), and methanol (30%, v/v). The high pH of this
electrolyte resulted in the separation of the monosaccharides in
the form of anionic species. The two adulterants evaluated were
maize and coffee husks. The LOQ (limit of quantification) for
both monosaccharides was 0.2 g in 100 g of dry matter, which
conformed to acceptable limits.
The capillary electrophoresis technique has been shown to
be as efficient as liquid chromatography for determination of
carbohydrate profiles of adulterants such as coffee husks and
maize grains. Nevertheless, the need for a hydrolysis step makes
the technique slower than HPLC. There have been few publica-
tions describing the use of this technique, which therefore mer-
its further investigation.
Biological methods
Contemporary techniques such as polymerase chain reaction
(PCR) offer the potential for the unequivocal identification of
DNA, and the use of DNA molecular markers has been
employed to identify different species or varieties. Molecular
markers such as microsatellites have been used to good effect in
the characterization of Coffea species, enabling discrimination
between robusta and arabica and detection of the presence of
adulterants.
Martellossi et al. (2005) demonstrated that PCR-grade
DNA can be obtained from roasted beans and even from
instant coffee. This would permit the analysis of commer-
cial samples, provided that suitable markers for species/
variety identification could be found. A total of 25 samples
of green coffee beans and 12 roasted coffees (3 arabica, 3
robusta, and 6 instant coffees) were used. It was demon-
strated that sufficient DNA survives during roasting and
freeze-drying to enable successful extraction and subse-
quent amplification, provided that a suitable protocol is
CRITICAL REVIEWS IN ANALYTICAL CHEMISTRY 89
Downloaded by [UNESP] at 11:58 15 April 2016
applied. Various methodological considerations were
described for the successful amplification of the DNA, and
the effect of the presence of contaminants that could
inhibit PCR was noted. Notwithstanding, once suitable
markers are found, use of these extraction methods should
enable assessment of the authenticity of coffee varieties,
determination of the percentage of robusta in blends, and
detection of adulterants.
The same analytical approach was adopted by Spaniolas
et al. (2006), who used the PCR technique to determine the
genomic profiles of arabica and robusta coffee. Samples
were also analyzed with a lab-on-a-chip capillary electro-
phoresis system. The choice of target was based on a phylo-
genetic study in which a number of single-nucleotide
polymorphisms (SNPs) were found in several Coffea species,
indicating the existence of different chlorotypes in the arab-
ica and robusta coffees. One of these SNPs resides in a PsuI
restriction site, resulting in the site being present in robusta
but absent in arabica. The purpose of these studies was to
evaluate whether this SNP could be exploited for the quali-
tative or quantitative detection of robusta contamination of
arabica using a PCR-restriction fragment length polymor-
phism (RFLP) approach. In this study, a total of 11 roasted
coffee samples (8 arabica and 3 robusta) with different ori-
gins were used. A 5% limit of detection was achieved for
coffee powder mixtures analyzed using this technique. The
chloroplast target used was the trnL(UAA)-trnF(GAA)
intergenic spacer region, which was found to be discrimina-
tory for all of the arabica and robusta varieties used in this
study. However, although the beans originated from a vari-
ety of geographical regions, further studies will be required
to confirm that the plastid copy number remains relatively
constant across a wider range of varieties, and that there
are no significant influences of factors related to the envi-
ronment or type of cultivation.
More recently, Ferreira et al. (2012) developed a DNA-
based method to detect and quantify the adulteration of
roasted coffee with barley, maize, and rice. The amplifica-
tion capacity was determined by random amplified poly-
morphic DNA (RAPD)-PCR and the products were
analyzed by horizontal electrophoresis in agarose gel.
Marker genes for coffee, barley, maize, and rice were
obtained from the National Centre of Biotechnology Infor-
mation (NCBI, USA). In order to confirm the specificity of
the chosen genes, they were statistically analyzed with
respect to their similarity with those of other species using
the Basic Local Alignment Search Tool (BLAST). The
regions of markers that showed similarity to other species
were discarded and specific regions were selected as DNA
templates to design the primers. The marker genes for bar-
ley (cytochrome C), maize (zein), and rice (hypothetical
protein Chromosome 8) showed no similarity to organisms
of the C. arabica and C. canephora species. Each primer
pair provided specificamplification only with the corre-
sponding target (adulterant) DNA sequences. This promis-
ing method showed the ability to detect the presence of the
aforementioned adulterants, indicating that it could be used
to ensure coffee quality in accordance with international
market specifications.
Final considerations
This review presents an overview of recent advances in the
detection of adulterants in coffee. The greatest numbers of pub-
lished articles concern spectrometric and instrumental separa-
tions techniques equally, with 13 articles. Spectrometric
methods, such as infrared and multispectral imaging techni-
ques, have practical advantages in that they are fast and do not
require prior sample preparation. In particular, the infrared
method can be used to study a variety of external adulterants,
including glucose, starch, chicory, and maize, as well adultera-
tion involving different coffee varieties. Mass spectroscopy has
also been shown to be a promising technique for the detection
of defective beans and adulteration with different coffee species.
Instrumental separation methods used in this area are based
on chromatographic separation employing liquid chromatogra-
phy, gas chromatography, and capillary electrophoresis.
Despite being time-consuming and more expensive than spec-
trometric methods, these techniques have delivered highly
promising results for the detection of a wide range of adultera-
tions involving different coffee species, coffee husks, maltodex-
trin, caramelized sugar, legumes, maize, starch, triticale, a¸ca
ı,
and barley. The advantages of chromatographic methods based
on chemical screening performed on the samples permits, in
many cases, the identification of potential adulterants’marker
compounds. The carbohydrates found most abundantly in
adulterants have been the class more cited as potential markers.
Among them can be mentioned xylose and mannitol (husks),
stachyose (leguminous), mannose (a¸ca
ıseeds), sucrose, glucose,
and frutose (brown sugar). Other classes of compounds also
showed relevance as markers, like tocopherols (D
5
avenasterol
and g-tocopherol, for arabica and robusta distinction and
maize, respectively), fatty acids (lenoleic/stearic acid profile for
maize), and volatile compounds (furfural and 5-methylfurfural,
for maize; 2-methylbutanal, 3-methylbutanal, and furan, for
barley; butyrolactone, hexanoic acid, 3-ethyl-2-methyl-1,3-hex-
adiene, b-linalool, and 2-butyl-3,5-dimethylpyrazine and 2-
pentylfuran, for defective seeds).
Biological methods based on the PCR technique have shown
considerable advances, despite their relatively low level of
uptake compared to the physical and chemical methods. It has
been shown that complete DNA still remains after the coffee
roasting process (in both roasted and instant coffee), which
indicates that the technique is viable, despite difficulties associ-
ated with amplification of the DNA signal. However, this
method still requires full validation of the DNA data bank and
remains expensive for use in routine analyses.
In general terms, it can be concluded that despite the exis-
tence of valuable studies in this area, it still remains necessary
to develop a widely applicable and sensitive methodology that
can address the various aspects of coffee adulteration. This
includes discrimination between species, detection of defective
beans, and identification of the presence of external agents.
This task will not be straightforward because of the complexity
of the issues involved. Nonetheless, the prospects remain prom-
ising, due to both scientific advances and the interest of regula-
tory agencies. The latter include the Institute for Scientific
Information on Coffee (ISIC), Coop
eration Internationale en
Recherche Agronomique pour le D
eveloppement (CIRAD), the
90 A. T. TOCI ET AL.
Downloaded by [UNESP] at 11:58 15 April 2016
International Coffee Organization (ICO), the Food and Agri-
culture Organization (FAO), and the Food Safety and Inspec-
tion Service of the United States Department of Agriculture
(FSI/USDA).
Funding
The authors would like to thank the Brazilian National Research Council
(CNPq) for financial support.
References
Amorim, A. C. L.; Hovell, A. M. C.; Pinto, A. C.; Eberlin, M. N.; Arruda, N.
P.; Pereira, E. J.; Bizzo, H. R.; Catharino, R. R.; Morais Filho, Z. B.;
Rezende, C. M. Green and Roasted Arabica Coffees Differentiated by
Ripeness, Process and Cup Quality via Electrospray Ionization Mass
Spectrometry Fingerprinting. J. Braz. Chem. Soc. 2009,20(2), 313–321.
Assad, E. D.; Sano, E. E.; Cunha, S. A. R.; Correa, T. B. S.; Rodrigues, H. R.
Identifying Adulteration in Coffee Powders by Spectral Signatures and
Digital Analyses. Pesqui. Agropecu. Bras. 2002,37(2), 211–216.
Associa¸c~
ao Brasileira da Ind
ustria de Caf
e (ABIC). Not
ıcias. http://www.
abic.com.br/exportacao.html (accessed 4/15/14).
Blanc, M. B.; Davis, G. E.; Parchet, J. M.; Viani, R. Chromatographic Profile
of Carbohydrates in Commercial Soluble Coffees. J. Agric. Food Chem.
1989,37, 926–930.
Briandet, R.; Kemsley, E. K.; Wilson, R. H. Approaches to Adulteration
Detection in Instant Coffees Using Infrared Spectroscopy and Chemo-
metrics. J. Sci. Food Agric. 1996,71, 359–366.
Cesar, C. L.; Vargas, H.; Lima, C. A. S.; Mendes Filho, J.; Miranda, L. C. M.
On the Use of Photoacoustic Spectroscopy for Investigating Adulter-
ated or Altered Powdered Coffee Samples. J. Agric. Food Chem. 1984,
32, 1355–1358.
Ciampa, A.; Renzi, G.; Taglienti, A.; Sequi, P.; Valentini, M. Studies on
Coffee Roasting Process by Means of Nuclear Magnetic Resonance
Spectroscopy. J. Food Qual. 2010,33, 199–211.
Davis, G. E.; Garwood, V. W.; Barfuss, D. L.; Husaini, S. A.; Blanc, M. B.;
Viani, R. Chromatographic Profile of Carbohydrates in Commercial
Coffees. 2. Identification of Mannitol. J. Agric. Food Chem. 1990,38(6),
1347–1350.
Domingues, D. S.; Pauli, E. D.; Abreu, J. E. M.; Massura, F. W.; Cristiano,
V.; Santos, M. J.; Nixdorf, S. L. Detection of Roasted and Ground Cof-
fee Adulteration by HPLC by Amperometric and by Post-Column
Derivatization UV–Vis Detection. Food Chem. 2014,146, 353–362.
Ferreira, T.; Farah, A.; Oliveira, T. C.; Lima, I. S.; Vit
orio, F.; Oliveira, E. M.
M. Detection of Barley and Maize as Adulterants in Commercial Cof-
fees Using Real-Time PCR. In Proceedings of 24th International Confer-
ence on Coffee Science,2012, San Ros
e, Costa Rica. http://www.asic-cafe.
org/en/proceedings/5971/toc
Fontes, A. S.; Bento, A. C.; Baesso, M. L.; Miranda, L. C. M. Thermal Lens
and pH Measurements in Pure and Adulterated Brewed Coffee. Ins-
trum. Sci. Technol. 2006,34, 163–181.
Fontes, A. S.; Bento, A. C.; Miranda, L. C. M.; Baesso, M. L. Thermal Lens
Evaluation of the Presence of Adulterants in Brewed Coffee. Anal. Sci.
2001,17, 526–529.
Garcia, L. M. Z.; Pauli, E. D.; Cristiano, V.; Camara, C. A. P.; Scarminio, I.
S.; Nixdorf, S. L. Profile in Coffee Due to Maize and Husk by Determin-
ing Carbohydrates Using HPAEC–PAD. J. Chromatogr. Sci. 2009,47,
825–832.
Garrett, R.; Vaz, A. M.; Hovell, C.; Eberlin, M. N.; Rezende, C. M. Arabica
and Robusta Coffees: Identification of Major Polar Compounds and
Quantification of Blends by Direct-Infusion Electrospray Ioniza-
tion¡Mass Spectrometry. J. Agric. Food Chem. 2012,60, 4253¡4258.
Gon¸calves, E. B.; Correa, T. B. S.; Sano, E. E.; Rodrigues, H. R.; Sano, E. E.;
Cunha, S. A. R.; Souza, M. L. M.; Lopes, J. R. H. Valida¸c~
ao intralabora-
torial de m
etodo por imagem para detec¸c~
ao e quantifica¸c~
ao de impure-
zas em caf
e torrado e mo
ıdo: Cascas e paus como exemplo. Braz. J.
Food Technol. 2007,10, 144–150.
Jham, G. N.; Berhow, M. A.; Manthey, L. K.; Palmquist, D. A.; Vaughn, S.
F. The Use of Fatty Acid Profile as a Potential Marker for Brazilian Cof-
fee (Coffea arabica L.) for Corn Adulteration. J. Braz. Chem. Soc. 2008,
19, 1462–1467.
Jham, G. N.; Winkler, J. K.; Berhow, M. A.; Vaughn, S. F. g-Tocopherol as
a Marker of Brazilian Coffee (Coffea arabica L.) Adulteration by Maize.
J. Agric. Food Chem. 2007,55, 5995–5999.
Lopez, F. C. Determina¸c~
ao quantitativa das principais subst^
ancias utiliza-
das para fraudar o caf
e torrado e mo
ıdo. Rev. Inst. Adolfo Lutz 1983,43
(1–2), 3–8.
MAPA Normative Instruction #16. http://www.agricultura.gov.br/comuni
cacao/noticias/2010/12/novo-padrao-para-cafe-vigora-em-fevereiro
(accessed July 20 2015).
Martellossi, C.; Taylor, E. J.; Lee, D.; Graziosi, G.; Donini, P. DNA Extrac-
tion and Analysis from Processed Coffee Beans. J. Agric. Food Chem.
2005,53, 8432–8436.
Menezes, Jr., J. B. F.; Bicudo, B. A. A. Sobre um M
etodo Microsc
opico para
Contagem de Cascas no Caf
eemP
o. Rev. Inst. Adolfo Lutz 1958,11, 13.
Nogueira, T.; Lago, C. L. Detection of Adulterations in Processed Coffee
with Cereals and Coffee Husks Using Capillary Zone Electrophoresis. J.
Sep. Sci. 2009,32, 3507–3511.
Oliveira, R. C. S.; Oliveira, L. S.; Franca, A. S.; Augusti. R. Evaluation of the
Potential of SPME-GC-MS and Chemometrics to Detect Adulteration
of Ground Roasted Coffee with Roasted Barley. J. Food Compos. Anal.
2009,22, 257–261.
Pauli, E. D.; Cristiano, V.; Nixdorf, S. L. M
etodo para determina¸c~
ao de car-
boidratos empregado na triagem de adultera¸c~
oes em caf
e. Qu
ım. Nova
2011,34, 689–694.
Peixoto, P. Fraudes atingem 25% das marcas de caf
e, diz associac~
ao. Folha
de S~
ao Paulo, April 28, 2009. http://www1.folha.uol.com.br/mercado/
2009/04/557291-fraudes-atingem-25-das-marcas-de-cafe-diz-associa
cao.shtml (accessed 2/9/14).
Pizarro, C.; Esteban-D
ıez, I.; Gonz
alez-S
aiz, J. M. Mixture Resolution
According to the Percentage of Robusta Variety in Order to Detect
Adulteration in Roasted Coffee by Near Infrared Spectroscopy. Anal.
Chim. Acta 2007,585, 266–276.
Reis, N.; Franca, A. S.; Oliveira, L. S. Discrimination between Roasted Cof-
fee, Roasted Maize and Coffee Husks by Diffuse Reflectance Infrared
Fourier Transform Spectroscopy. LWT - Food Sci. Technol. 2013,50,
715–722.
Ruiz, A. D.; Suarez, M.; Moreno, E.; Bautista, E. Volatiles of Roasted Maize
(Zea mays), Barley (Hordeum vulgare) and Coffee (Coffea arabica).
Effect of Addition of These Cereals on the Aroma of Coffee. Rev.
Colombiana Quim. 1995, 24, 13–23.
Sano, E. E.; Assad, E. D.; Cunha, S. A. R.; Correa, T. B. S.; Rodrigues, H. R.
Quantifying Adulteration in Roast Coffee Powders by Digital Image
Processing. J. Food Qual. 2003,26, 123–134.
Spaniolas, S.; May, S. T.; Bennett, M. J.; Tucker, G. A. Authentication of
Coffee by Means of PCR-RFLP Analysis and Lab-on-a-Chip Capillary
Electrophoresis. J. Agric. Food Chem. 2006,54, 7466–7470.
Spaniolas, S.; Tsachaki, M.; Bennett, M. J.; Tucker, G. A. Evaluation of
DNA Extraction Methods from Green and Roasted Coffee Beans. Food
Control 2008,19, 257–262.
Stober, P.; Benet, S.; Hischenhuber, C.; Fischer, M.; Fay, L. B. Sta-
chyose: A Marker of the Presence of Legume Adulterants in Solu-
ble Coffee? In Proceedings of 19th International Conference on
Coffee Science, 2001, Trieste, Italy. http://www.asic-cafe.org/en/pro
ceedings/3588/toc
Tavares, K. M.; Pereira, R. G. F. A.; Nunes, C. A.; Pinheiro, A. C. M.;
Rodarte, M. P.; Guerreiro, M. C. Mid-infrared Spectroscopy and Sen-
sory Analysis Applied to Detection of Adulteration in Roasted Coffee
by Addition of Coffee Husks. Qu
ım. Nova 2012,35, 1164–1168.
Toci, A. T.; Farah, A. Volatile Compounds as Potential Defective Coffee
Beans’Markers. Food Chem. 2008,108, 1133–1141.
Toci, A.T.; Silva, C.M.; Fernandes, F.; Farah, A. Effect of the Fluid Flow
Speed Changes on the Chemical Composition of Coffee Samples
Roasted in an Industrial Fluidized Roaster. In Proceedings of 22nd
International Conference on Coffee Science—Coffee Chemistry & Qual-
ity, Campinas, Brazil, Sep 14–19, 2008; http://www.asic-cafe.org/en/pro
ceedings/3591/toc (last accessed Sept 14, 2015).
CRITICAL REVIEWS IN ANALYTICAL CHEMISTRY 91
Downloaded by [UNESP] at 11:58 15 April 2016
Toci, A. T.; Farah, A. Volatile Fingerprint of Brazilian Defective Coffee
Seeds: Corroboration of Potential Marker Compounds and Identifica-
tion of New Low Quality Indicators. Food Chem. 2014,153, 298–314.
Valdenebro, M. S.; Leon-Camacho, M.; Pablos, F.; Gonzalez, A. G.; Mart
ın, M.
J. Determination of the Arabica/Robusta Composition of Roasted Coffee
According to Their Sterolic Content. Analyst 1999,124,999–1002.
92 A. T. TOCI ET AL.
Downloaded by [UNESP] at 11:58 15 April 2016
