�������� ����� ��

Direct Infusion Electrospray Ionization Mass Spectrometry Applied to theDetection of Forgeries: Roasted Coffees Adulterated with their Husks

Francisco J.T. Aquino, Rodinei Augusti, Junia de O. Alves, Maria E.R.Diniz, Sergio A.L. Morais, Blyeny H.P. Alves, Evandro A. Nascimento, AdaoA. Sabino

PII: S0026-265X(14)00119-2DOI: doi: 10.1016/j.microc.2014.06.016Reference: MICROC 1980

To appear in: Microchemical Journal

Received date: 14 November 2013Revised date: 16 April 2014Accepted date: 12 June 2014

Please cite this article as: Francisco J.T. Aquino, Rodinei Augusti, Junia de O. Alves,Maria E.R. Diniz, Sergio A.L. Morais, Blyeny H.P. Alves, Evandro A. Nascimento, AdaoA. Sabino, Direct Infusion Electrospray Ionization Mass Spectrometry Applied to theDetection of Forgeries: Roasted Coffees Adulterated with their Husks, MicrochemicalJournal (2014), doi: 10.1016/j.microc.2014.06.016

This is a PDF file of an unedited manuscript that has been accepted for publication.As a service to our customers we are providing this early version of the manuscript.The manuscript will undergo copyediting, typesetting, and review of the resulting proofbefore it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers thatapply to the journal pertain.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

1

Direct Infusion Electrospray Ionization Mass Spectrometry Applied to the Detection of

Forgeries: Roasted Coffees Adulterated with their Husks

Francisco J. T. Aquinoa,b

, Rodinei Augustia*

, Júnia de O. Alvesc, Maria E. R. Diniz

a,

Sérgio A. L. Moraisb, Blyeny H. P. Alves

b, Evandro A. Nascimento

b, Adão A. Sabino

a

a Department of Chemistry, Federal University of Minas Gerais, Belo Horizonte, MG, Brazil,

31270-901

b Laboratory of Natural Products and Chromatography (LPNC), Institute of Chemistry, Federal

University of Uberlândia, Uberlândia, MG, Brazil, 38400-902

c Department of Chemistry, Federal Center of Technological Education of Minas Gerais, Belo

Horizonte, MG, Brazil, 30421-169.

Corresponding author: Rodinei Augusti, Department of Chemistry, Federal University of Minas

Gerais, Belo Horizonte, MG, Brazil, 31270-901. Phone: 55-31-34095734; fax: 55-31-34095700; e-

mail: augusti.rodinei@gmail.com

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

2

Abstract

Recognition of samples of coffee adulterated with their own husks has been a challenging task. The

quite similar physical aspect of roasted grinded coffee husk when compared to ground coffee

hampers a prompt visual distinction among both types of products. Several methodologies that

make use of distinct analytical techniques have been developed for the authentication of coffee

quality. In the present work, we demonstrate that direct infusion electrospray ionization mass

spectrometry (ESI-MS) can be applied to detect counterfeit samples of roasted coffee adulterated by

the addition of coffee husks (at a level of 10 % w/w) in a quick and reliable way. The ESI-MS

fingerprints (in both the negative and positive modes) revealed the presence of diagnostic markers,

such as carbohydrates (for instance, saccharose), chlorogenic acids, caffeine, and other components

related to the coffee flavor, that characterize each type of sample (coffee and rusk). Furthermore,

the PCA (principal component analysis) methodology, applied to the whole set of the ESI-MS data

(in the negative mode), grouped the samples into three clearly distinct categories: coffees, husks and

blends. The results presented herein describe therefore an innovative and rapid methodology

potentially useful in the diagnosis of such hardly-detectable type of adulteration.

Keywords: Roasted

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

3

1. Introduction

A significant amount of commercial roasted coffee produced in Brazil is regarded as low

quality due to the presence of defects in the coffee beans as well as adulterant material added before

or during the roasting process.[1] The addition of impurities in the roasted coffee (now a common

practice) helps to reduce costs, thus allowing for higher earnings coming from the product

commercialization in the internal market.[1] Roasted fine grinded coffee husk present a similar

physical aspect when compared to ground coffee; because of that it has been difficult to

characterize samples of coffee adulterated with husk by a simple visual inspection. Chemometric

methods have been applied in order to verify adulteration patterns using coffee husk and other

matrixes by means of its carbohydrate concentrations. There is a great correspondence of

carbohydrates from the coffee husk in relation to coffee bean, with higher levels of mannose,

galactose, and arabinose.[2] Moreover, there are few reports in the literature referring to the

chemical composition of roasted coffee and coffee husks.[2, 3]

Much effort has been devoted to the development of analytical methods for the authentication

of coffee quality.[4-7] Analyses

based on the determination of the total xylose content in

commercial soluble coffees adulterated with husks or parchments have been investigated.[7]

Analogously, Lago and coworkers [5] described the application of capillary zone electrophoresis to

detect adulteration in processed coffee by the addition of cereals and coffee husk. In another work,

the separation of authentic and counterfeit coffees (adulterated with roasted barley) was

accomplished by the application of principal component analysis (PCA) on chromatographic data

(obtained via Solid Phase Microextraction and Gas Chromatography Coupled to Mass

Spectrometry: SPME-GC-MS).[6] Electrospray ionization mass spectrometry (ESI-MS) has been

widely used for the characterization and analysis of complex matrices, especially food samples.[8-

11] For instance, ESI-MS has enabled the characterization of non-defective and defective beans of

Arabica and Robusta coffees,[12] and also the differentiation between green and roasted Arabica

coffees, at different stages of ripeness during the post-harvesting process and in cup quality.[13]

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

4

However, there is a lack of information on the characterization of coffee husk in the literature by

ESI-MS.

This work aims at the attainment and comparison of ESI-MS fingerprints of the methanolic

extracts of typical roasted coffees and coffee husks. The principal aim is, however, to verify

whether this simple and cost-effective approach can be applied to identify forgery samples of

roasted coffee containing coffee husk (at 10 % w/w). Finally, a non-supervised methodology (PCA:

principal component analysis) is also employed on the whole set of ESI-MS data to make sure that

the three types of samples (coffees, husks and blends) are split into well-defined groups.

2. Material and methods

2.1. Sample Description

Green coffee beans from the 2007 crop (from 5 distinct varieties: Acaiá, Catuaí, Topázio and

Conilon) and coffee husks were collected at two different regions of the State of Minas Gerais

(Cerrado and South) and were provided by COOCACER (Cooperativa dos Cafeicultores do

Cerrado, Araguari, State of Minas Gerais, Brazil). Coffee husks used herein (named Araguari and

Expresso) were non-dehydrated and contained shell (endocarp). These husks arose from the coffee

beans produced in the State of Minas Gerais after drying on a sun terrace with temperatures ranging

from 35 to 40 ºC. The husks were then collected and stored in plastic raffia bags. They contained

mucilage, parchment (cone) and pulp.

2.2. Sample preparation

All the samples (coffees and husks) were dried in an oven (Quimis-Kett 600, Brazil) at 65 °C

without ventilation for 36 to 48 hours to reach a moisture content of 11-15 % and a constant weight.

These samples were then roasted in a commercial electric bench roaster (Pinhalense, model TC-0,

Brazil) at a temperature of 190±10 °C. The color of the roasting was determined by visual

assessment using the Roast Color Classification System (AGTRON – SCAA, 1995).[14] The light

roasting degree employed herein was achieved in approximately 10 ± 1 min. In sequence, the

samples were ground in a household mill (KrupsTM

, model 203, Mexico) for 60 s and sieved

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

5

through a stainless steel mesh (0.71 mm), packed in polyethylene/ aluminum, sealed and stored at -

18 ± 5 °C until analysis. As a proof-of-principle, five artificially adulterated admixtures were

prepared by mixing roasted coffees with rusks in a proportion of 90: 10 w/w. The samples were

then numbered accordingly to the following classification: coffee (1-9), husk (10-14), and blend

(15-24).

2.3. Methanolic extraction

Coffees, husks and blends (1 g) in methanol (15 mL) were submitted to reflux for 2 h

(condenser at 5 ± 2 °C). The samples were then filtered through a filter paper. The extracts were

kept in a freezer at -18 ± 5 °C until analysis.

2.4. ESI-MS: sample preparation

Prior to the ESI-MS analysis, aliquots of 20.0 μL of each extract were transferred to 1.5-mL

tubes. Then, 600 µL of Milli-Q water (Millipore, Schwalbach, Germany) and 600 µL of methanol

(HPLC grade, Merck, São Paulo, Brazil) were added. These solutions were basified with 10.0 µL of

concentrated ammonium hydroxide (Aldrich, USA) or acidified with 1.0 µL of concentrated formic

acid (Fluka, Germany) to be analyzed in the negative or positive modes, respectively. They were

then submitted to a vigorous stirring for 30 s using a vortex apparatus (Phoenix AP-56, Araraquara,

Brazil). Sample introduction was performed using a syringe pump (500 μL, Hamilton Gastight

#175, Nevada, USA) at a flow rate of 10.0 μL min-1

and pumped through an uncoated fused silica

capillary.

2.5. ESI-MS: analysis

The aqueous-methanolic solutions were injected directly into the mass spectrometer (LCQ-

Fleet, Thermo Scientific, San Jose, CA) equipped with an electrospray ionization (ESI) source,

operating both in the negative and positive modes. The optimized parameters were as follows:

capillary temperature 275 °C, capillary voltage ± 40.0 V, spray voltage ± 3.0 kV, and cone voltage

ranging from ± 15 to ± 30 V. Each analysis required about 1 minute. High purity nitrogen was used

as the nebulizer and drying gas. Mass spectra were acquired by scanning over the 100-1000 m/z

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

6

range. Xcalibur Analyst internal software (v2.0.7, Thermo Fischer Scientific Inc.) was used for

instrumentation control and data acquisition.

2.6. ESI-MS data handling and statistical treatment

The ions in each mass spectrum were summed into integral m/z values and normalized using an

in-house program. To discard noise, only the ions with a relative abundance higher than 5% were

included into the final data matrix. The m/z values were aligned and compiled to generate a final

matrix where each line was a sample and each column a variable (m/z ratios and relative intensities

of the detected ions). Multivariate analysis by PCA was performed by using the Matlab software (v

7.9, R2009b, USA).

3. Results and discussion

3.1. Distinction between samples of roasted coffee and coffee husk

3.1.1. ESI(-)MS fingerprints

Figure 1 displays the ESI(-)-MS fingerprints of the methanolic extracts of typical samples of a

roasted coffee (Coffea arabica) (Figure 1a) and husk (Figure 1b). The most intense anions (and

their probable identification) detected in the ESI(-)-MS of the coffee sample are displayed in Table

1. Most of the main components presented in Table 1 were previously detected in coffee brews as

reported in the literature.[12-21] Compounds of high nominal mass (above 600 Da) may be possibly

attributed to adducts of chlorogenic acids, phospholipids, and other oligomers.[22]

Figure 1

Table 1

A similar set of prominent ions are observed in the ESI(-)-MS of the coffee husk extracts

(Figure 1b). The following anions are diagnostic for the husk samples: 191 (quinic and citric acids),

255 (palmitic acid), 279 (linoleic acid), 335 (caffeoylquinide), 353 (caffeoylquinic acid), 367

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

7

(feruloylquinic acid), 383 (an unidentified compound), 481 (gallocatechin-O-glucuronide), 497

(dicaffeoylshikimic acid), 515 (dicaffeoylquinic acid) and 529 (caffeoylferuloylquinic acid).

The comparison of Figures 1a and 1b clearly shows that both extracts have different

composition. Hence, whereas the ESI(-)-MS of the coffee extracts indicate the major presence of

caffeoylquinic acid (detected in its deprotonated form of m/z 353, as observed in Figure 1a and

displayed in Table 1), the ESI(-)-MS of the husk extracts are characterized to possess the anion of

m/z 191 (deprotonated form of quinic and citric acids) as the major component. Furthermore, it was

found that phenylacetic acid, dehydrated caffeic acid, dehydrated quinic acid and caffeic acid

(detected in their deprotonated forms of m/z 135, 161, 173 and 179, respectively) are detected

exclusively in the ESI(-)-MS of the coffee (Figure 1a) rather than rusk (Figure 1b) extracts. These

results were lately confirmed by HPLC experiments using standard compounds.

3.1.2. ESI(+)-MS fingerprints

Figure 2 displays typical ESI(+)-MS fingerprints of coffee (Figure 2a) and husk (Figure 2b)

extracts. The most characteristic ions (and their possible structural attribution) in the ESI(+)-MS of

the coffee extracts are displayed in Table 2.

Figure 2

Table 2

Most of the main compounds presented in Table 2 are also well-known in coffee brews.[12-

21] In the ESI(+)-MS of roasted coffees, the ion of m/z 195 is abundant in all spectra and could be

associated with caffeine. Other ions present are protonated gama-aminobutanoic acid (m/z 104),

protonated trigonelline (m/z 138) and its potassium adduct (m/z 176). In the range of m/z 300 to 400,

ions relative to saccharose (sodium adduct: m/z 365; potassium adduct: m/z 381) and chlorogenic

acid (protonated form: m/z 355; potassium adduct: m/z 393) are noticeable. Another group of ions

important for discrimination among both classes of samples (coffee and rusk) is observed between

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

8

m/z 400 and 1000. Hence, the ions of m/z 535, 719, 758 and 783 are observed exclusively in the

ESI(+)-MS of the coffee extracts. The compounds of high (above 700 Da) nominal masses in the

coffee samples may be attributed to the potassium adducts of chlorogenic acids, phospholipids, and

other oligomers.[14]

Conversely, the ESI(+)-MS of the husk extracts (Figure 2b) shows a quite distinct profile:

only two prominent ions of m/z 138 (protonated trigonelline) and 195 (protonated caffeine) can be

noticed. Hence, whereas the caffeine (m/z 195) and trigonelline (m/z 138) ions are predominant in

both mass spectra, sugars, chlorogenic acids and compounds with higher nominal masses (above

500 Da) seems to be absent (or at a very low concentration) in the rusk samples.

3.1.3. Detecting samples of coffee adulterated with husk

Adulterations of coffee in Brazil are usually conducted by adding adulterants at levels ranging

from 20 to 40 %w/w.[1, 23] Thus, an adulteration of 10 % w/w with rusk was chosen to ensure that

the ESI-MS approach could be efficiently applied to the analysis of real samples. Hence, whereas

lower levels of adulteration (< 10 % w/w) have no practical interest, higher contents (> 10 % w/w)

could be easily detected by the present methodology. Methanolic extracts of the blends were then

analyzed by ESI-MS in both the negative and positive modes. A simple visual inspection revealed

that the ESI-MS of these mixtures (not shown) were roughly an average of those of the coffee and

rusk samples. To better investigate the statistical relevance of the ESI(-)-MS fingerprints in

differentiating among the samples of coffee (1-9), husk (10-14) and blends (15-24) the data were

analyzed by the PCA method. Similar results were obtained when the PCA approach was applied to

the ESI(+)-MS data; hence, these latter results will not be included herein.

Figure 3

In the scores plot displayed in Figure 3, the samples were clearly grouped according to their

nature (coffee, husk and blend). Note that, despite of the distinct cultivar, region, and process

benefiting employed during the production of coffees and husks, a homogeneous grouping were

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

9

achieved since each sample was correctly allocated into its expected class. This can be additionally

confirmed as each group of samples is presented surrounded by an ellipse with a 95 % confidence

level. Furthermore, in this PCA model, PC1, PC2, and PC3 accounted for 42.59%, 22.20%, and

13.14% of the total variance, respectively.

Finally, the ions selected as being the most important for the separation among the three

groups and that presented the greatest influence on PC1, PC2 and PC3 can be distinguished in the

loading plots displayed in Figure 4. These ions, ordered from the highest to the lowest influence in

each PC, are the following ones: PC1- m/z 191 (quinic and citric acids), 353 (caffeoylquinic acid),

335 (caffeoylquinide) and 367 (feruloylquinic acid); PC2- m/z 335 (caffeoylquinide), 353

(caffeoylquinic acid), 367 (feruloylquinic acid), 515 (dicaffeoylquinic acid) and 497

(dicaffeoylshikimic acid); PC3- m/z 335 (caffeoylquinide), 191 (quinic and citric acids), 353

(caffeoylquinic acid) and 497 (dicaffeoylshikimic acid) (for PC3).

Figure 4

4. Conclusions

The present study comprises the first example on the application of ESI-MS to identify

counterfeit roasted coffees adulterated by the addition of their own roasted husks, which were also

characterized by the first time by using ESI-MS. The PCA approach applied on the ESI-MS data

clearly split the samples of coffee, husk and blends of coffee/husk (prepared at a level of 10 % w/w

to simulate adulteration) into three well-defined groups. Finally, this methodology is certainly

promising in detecting such kind of adulteration in real samples.

5. Acknowledgments

The authors gratefully acknowledge PROCAD-CAPES UFU/UFMG, FAPEMIG and CNPq

for the financial support. The authors are also grateful to UFU/UFMG and

NIEAMBAV/DEQ/UFMG for providing technical support for the mass spectral analyses and

inestimable assistance in several fields.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

10

6. References

[1] R.D.M.C. Amboni, A. de Francisco, E. Teixeira, Detection of frauds in ground coffee with

scanning electronic microscopy, Cienc. Tecnol. Aliment., 19 (1999) 311-313.

[2] L.M.Z. Garcia, E.D. Pauli, V. Cristiano, C.A.P. da Camara, I.S. Scarminio, S.L. Nixdorf,

Chemometric evaluation of adulteration profile in coffee due to corn and husk by determining

carbohydrates using HPAEC-PAD, J. Chromatogr. Sci., 47 (2009) 825-832.

[3] D. Brand, A. Pandey, S. Roussos, C.R. Soccol, Biological detoxification of coffee husk by

filamentous fungi using a solid state fermentation system, Enzyme Microb. Tech., 27 (2000) 127-

133.

[4] R. Briandet, E.K. Kemsley, R.H. Wilson, Approaches to adulteration detection in instant coffees

using infrared spectroscopy and chemometrics, J. Sci. Food Agr., 71 (1996) 359-366.

[5] T. Nogueira, C.L. do Lago, Detection of adulterations in processed coffee with cereals and

coffee husks using capillary zone electrophoresis, J. Sep. Sci., 32 (2009) 3507-3511.

[6] R.C.S. Oliveira, L.S. Oliveira, A.S. Franca, R. Augusti, Evaluation of the potential of SPME-

GC-MS and chemometrics to detect adulteration of ground roasted coffee with roasted barley, J.

Food Compos. Anal., 22 (2009) 257-261.

[7] J. Prodolliet, M. Bruelhart, M.B. Blanc, V. Leloup, G. Cherix, C.M. Donnelly, R. Viani,

Adulteration of soluble coffee with coffee husks and parchments, J. Aoac Int., 78 (1995) 761-767.

[8] J.D. Alves, W.B. Neto, H. Mitsutake, P.S.P. Alves, R. Augusti, Extra virgin (EV) and ordinary

(ON) olive oils: distinction and detection of adulteration (EV with ON) as determined by direct

infusion electrospray ionization mass spectrometry and chemometric approaches, Rapid Commun.

Mass Spectrom., 24 (2010) 1875-1880.

[9] H.J. Cooper, A.G. Marshall, Electrospray ionization Fourier transform mass spectrometric

analysis of wine, J. Agr. Food Chem., 49 (2001) 5710-5718.

[10] P.P. de Souza, H.G.L. Siebald, D.V. Augusti, W.B. Neto, V.M. Amorim, R.R. Catharino, M.N.

Eberlin, R. Augusti, Electrospray ionization mass spectrometry fingerprinting of Brazilian artisan

cachaca aged in different wood casks, J. Agr. Food Chem., 55 (2007) 2094-2102.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

11

[11] L.S. Santos, R.R. Catharino, C.L. Aguiar, S.M. Tsai, M.N. Eberlin, Chemotaxonomic markers

of organic, natural, and genetically modified soybeans detected by direct infusion electrospray

ionization mass spectrometry, J. Radioanal. Nucl. Chem., 269 (2006) 505-509.

[12] J.C.F. Mendonca, A.S. Franca, L.S. Oliveira, M. Nunes, Chemical characterisation of non-

defective and defective green arabica and robusta coffees by electrospray ionization-mass

spectrometry (ESI-MS), Food Chem., 111 (2008) 490-497.

[13] A.C.L. Amorim, A.M.C. Hovell, A.C. Pinto, M.N. Eberlin, N.P. Arruda, E.J. Pereira, H.R.

Bizzo, R.R. Catharino, Z.B. Morais, C.M. Rezende, Green and roasted Arabica coffees

differentiated by ripeness, process and cup quality via electrospray ionization mass spectrometry

fingerprinting, J. Braz. Chem. Soc., 20 (2009) 313-321.

[14] S.A.L. de Morais, F.J.T. de Aquino, P.M. do Nascimento, E.A. do Nascimento, R. Chang,

Bioactive compounds and antioxidant activity of Conilon coffee submitted to different degrees of

roasting, Quim. Nova, 32 (2009) 327-331.

[15] R.M. Alonso-Salces, C. Guillou, L.A. Berrueta, Liquid chromatography coupled with

ultraviolet absorbance detection, electrospray ionization, collision-induced dissociation and tandem

mass spectrometry on a triple quadrupole for the on-line characterization of polyphenols and

methylxanthines in green coffee beans, Rapid Commun. Mass Spectrom., 23 (2009) 363-383.

[16] Y.C. Chen, A.R. Hu, Simultaneous determination of trace benzodiazepines from drinks by

using direct electrospray probe/mass spectrometry (DEP/MS), Forensic Sci. Int., 103 (1999) 79-88.

[17] M.N. Clifford, S. Knight, The cinnamoyl-amino acid conjugates of green robusta coffee beans,

Food Chem., 87 (2004) 457-463.

[18] O. Frank, S. Blumberg, C. Kunert, G. Zehentbauer, T. Hofmann, Structure determination and

sensory analysis of bitter-tasting 4-vinylcatechol oligomers and their identification in roasted coffee

by means of LC-MS/MS, J. Agr. Food Chem., 55 (2007) 1945-1954.

[19] A. Farah, C.M. Donangelo, Phenolic compounds in coffee, Braz. J. Plant Physiol., 18 (2006)

23-36.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

12

[20] R. Garrett, B.G. Vaz, A.M.C. Hovell, M.N. Eberlin, C.M. Rezende, Arabica and Robusta

coffees: identification of major polar compounds and quantification of blends by direct-infusion

electrospray ionization-mass spectrometry, J. Agr. Food Chem., 60 (2012) 4253-4258.

[21] R. Jaiswal, M.A. Patras, P.J. Eravuchira, N. Kuhnert, Profile and characterization of the

chlorogenic acids in green Robusta coffee beans by LC-MSn: identification of seven new classes of

compounds, J. Agr. Food Chem., 58 (2010) 8722-8737.

[22] L.M. de Souza, M. Muller-Santos, M. Iacomini, P.A.J. Gorin, G.L. Sassaki, Positive and

negative tandem mass spectrometric fingerprints of lipids from the halophilic Archaea Haloarcula

marismortui, J. Lipid Res., 50 (2009) 1363-1373.

[23] F.C. Lopez, Quantitative determination of the main substances used to defraud ground roasted

coffee, Rev. Inst. Adolfo Lutz, 43 (1983) 3-8.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

13

Caption to Figures

Figure 1. ESI(-)-MS fingerprint of methanolic extracts of typical samples of: (a) roasted coffee

(Coffea arabica); (b) roasted coffee husk (Coffea arabica).

Figure 2. ESI(+)-MS fingerprint of methanolic extracts of typical samples of: (a) roasted coffee

(Coffea arabica); (b) roasted coffee husk (Coffea arabica).

Figure 3. Scores plot for the first three PCs obtained from the ESI(-)-MS data of the methanolic

extracts of coffee (1-9), coffee husk (10-14), and adulterated (15-24, coffee/ coffee husk

admixtures) samples.

Figure 4. Loadings plot for PC1 (a), PC2 (b), and PC3 (c) arising from the statistical treatment of

the ESI(-)-MS data derived from the 24 samples of coffee (1-9), coffee husk (10-14), and

adulterated (15-24, coffee/coffee husk admixtures).

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

14

Figure 1

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

15

Figure 2

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

16

Figure 3

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

17

Figure 4

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

18

Table 1. The most characteristic ions observed in the ESI(-)-MS of the roasted coffee (Coffea

arabica) extracts (an example is shown in Figure 1a). The ions refer to the deprotonated forms of

the proposed compounds.[12-22]

Ion m/z Possible Compound

135 phenylacetic acid

161 dehydrated caffeic acid

173 dehydrated quinic acid

179 caffeic acid

191 quinic and citric acids

335 caffeoylquinide

353 caffeoylquinic acid

367 feruloylquinic acid

481 gallocatechin-O-glucuronide

497 dicaffeoylshikimic acid

515 dicaffeoylquinic acid

529 caffeoylferuloylquinic acid

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

19

Table 2. The most characteristic ions observed in the ESI(+)-MS of the roasted coffee (Coffea

arabica) extracts (an example is shown in Figure 2a). The ions refer to the protonated forms (or the

sodium/ potassium adducts) of the proposed compounds.[12-22]

Ion m/z Possible Compound

104 gama-aminobutanoic acid

138 trigonelline

176 trigonelline (potassium adduct)

195 caffeine

355 chlorogenic acid

365 saccharose (sodium adduct)

381 saccharose (potassium adduct)

393 chlorogenic acid (potassium adduct)

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

20

Highlights

● Direct infusion ESI-MS shown to be a viable alternative to recognize samples of coffee

adulterated with coffee husks.

● Direct infusion ESI-MS was also able to differentiate among samples of roasted coffee and husk.

● Diagnostic markers were detected for each type of sample.

● The PCA plot grouped the samples into three clearly distinct categories: coffees, husks and

blends.

● This is an innovative and rapid methodology potentially useful in the diagnosis of such a hardly-

detectable type of adulteration.