IOP Conference Series: Earth and
Environmental Science
     
PAPER • OPEN ACCESS
The authentication of peaberry and civet ground
roasted robusta coffee using UV-visible
spectroscopy and PLS-DA method with two
different particle sizes
To cite this article: Diding Suhandy et al 2019 IOP Conf. Ser.: Earth Environ. Sci. 258 012043
 
View the article online for updates and enhancements.
You may also like
Enumeration and characterization of
bacteria from civet gastrointestinal tract
M Muzaifa, D Hasni, Y Abubakar et al.
-
Numerical calculation of the turbulent flow
past a surface mounted cube with
assimilation of PIV data
Konstantinos Kellaris, Nikolaos Petros
Pallas and Demetri Bouris
-
Multi-channel EEG-based sleep staging
using brain functional connectivity and
domain adaptation
Wenhao Yuan, Wentao Xiang, Kaiyue Si
et al.
-
This content was downloaded from IP address 103.19.37.202 on 21/09/2024 at 20:38

Content from this work may be used under the terms of theCreative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd (ICoSITeR) 2018
IOP Conf. Series: Earth and Environmental Science258 (2019) 012043
IOP Publishing
doi:10.1088/1755-1315/258/1/012043
1
The authentication of peaberry and civet ground roasted
robusta coffee using UV-visible spectroscopy and PLS-DA
method with two different particle sizes
Diding Suhandy
1,4
, Meinilwita Yulia
2,4
and Kusumiyati
3

1
Department of Agricultural Engineering, The University of Lampung, Jl. Prof. Dr.
Soemantri Brojonegoro No.1, Bandar Lampung, 35145, Indonesia.
2
Department of Agricultural Technology, Lampung State Polytechnic, Jl. Soekarno
Hatta No. 10, Rajabasa Bandar Lampung, 35141, Indonesia.
3
Crop Science Department, Faculty of Agriculture, Padjadjaran University, Bandung,
Indonesia.


4
Spectroscopy Research Group (SRG), Laboratory of Bioprocess and Postharvest
Engineering, Faculty of Agriculture, The University of Lampung, Bandar Lampung,
Lampung, Indonesia.

E-mail: [email protected]
Abstract. The objective of this research was to investigate the influence of particle size of
ground roasted coffee samples in the authentication of peaberry and civet coffee using UV-
visible spectroscopy combined with partial least squares-discriminant analysis (PLS-DA)
method. For this purpose, we provide a total of 80 samples of peaberry and civet coffee with two
different particle size of 1680 micrometer (mesh 12) and 297 micrometer (mesh 50). The all
coffee samples were extracted using hot distilled water (90-98ºC). UV-visible spectral data of
the all coffee samples were acquired using a UV-visible spectrometer in transmittance mode.
The result showed that the different absorbance spectra were observed for the different particle
sizes of peaberry and civet coffee samples. The absorbance of samples having particle size of
297 micrometer (mesh 50) is higher than that of particle size of 1680 micrometer (mesh 12). The
calibration model was developed using PLS-DA method for samples having same particle size
and resulted in acceptable result with high coefficient of determination (R
2
) (R
2
=0.99 both for
mesh 12 and 50, respectively), low root mean square error of calibration (RMSEC), and high
residual prediction deviation (RPD). However, the performance of the prediction results was
very low for samples having different particle size with high bias and high root mean square
error of prediction (RMSEP).
1. Introduction
Peaberry coffee (‘kopi lanang’ in Indonesian language) and civet coffee (‘kopi luwak’ in Indonesian
language) are one of the expensive specialty coffees in the world. A peaberry (also called caracol or
snail in Spanish) is a natural mutation of the coffee bean inside the cherry. Normally two coffee beans
grow in a fruit (dicotyledonous)—flat against each other like halves of a peanut; however, on rare
occasions a single bean is produced (monocotyledon) [1]. Civet coffee is any coffee bean (arabica or
robusta) which has been eaten and passed through the digestive tract of Asian palm civet (Paradoxurus

(ICoSITeR) 2018
IOP Conf. Series: Earth and Environmental Science258 (2019) 012043
IOP Publishing
doi:10.1088/1755-1315/258/1/012043
2
hermaphroditus), which uses its keen senses to select only the best and ripest berries [2-3].The
production of both peaberry and civet coffees are very limited. For this reason, peaberry and civet
coffees have been a target of food adulteration.
Nowadays, the keyword of food authentication is becoming popular. This authentication system may
give benefit to both producers and customers. For this reason, the area of research for food authentication
is attracting many researchers. The development of new analytical method for food authentication
system is one of the research focus in this area. No exception for coffee authentication system. Several
analytical methods have been proposed for coffee authentication such as NIR and mid spectroscopy [4-
9], fluorescence spectroscopy [10-11], image processing/computer vision [12], nuclear magnetic
resonance (NMR) spectroscopy [13], and electronic nose [14]. However, most of these methods are
expensive and not easy to follow.
Previously, Suhandy and coworkers has developed an authentication and quality evaluation of
ground roasted coffee using UV-visible spectroscopy and chemometrics [1-2, 15-17]. This method is
simpler and cheaper. However, all those reported studies used coffee samples with same particle size
(same mesh) for developing calibration and validation model. Those researches involved procedure of
sieving of ground roasted coffee that is laborious and time consuming. Therefore, for practical
application, it is needed to remove the procedure of sieving and establish calibration model of coffee
authentication which can be universally used for any particle sizes. However, there is lack of research
regarding to relationship between particle sizes of ground roasted coffee and quality of spectral data in
UV-visible region. In this research, an investigation on the influence of particle size of ground roasted
coffee samples in the authentication of peaberry and civet coffee was conducted using UV-visible
spectroscopy combined with partial least squares-discriminant analysis (PLS-DA) method.
2. Materials and Methods
2.1. Civet and peaberry coffee samples
In this research, 80 samples of peaberry and civet coffee with two different particle size of 1680
micrometer (mesh 12) and 297 micrometer (mesh 50) were prepared (1.0 gram weight for each samples).
The samples were collected from coffee farmer in West Lampung, Lampung province of Indonesia. The
extraction of coffee samples was performed by using hot distilled water. The samples were divided into
two groups: calibration and validation set (30 samples for peaberry and civet, respectively) and
prediction set (10 samples for peaberry and civet, respectively).
2.2. Acquisition of UV-visible spectral data
The spectral data of aqueous civet and peaberry coffee samples were acquired in the range of 190-1100
nm by using a UV-Vis spectrometer (Genesys™ 10S UV-Vis, Thermo Scientific, USA). This
spectrometer was equipped with a quartz cell with optical path of 10 mm. The spectral acquisition was
done at spectral resolution of 1 nm at a room temperature (about 27-28ºC). The original spectra were
used for chemometric analysis.
2.3. Chemometric analysis
To study the influence of particle sizes of ground roasted coffee for authentication of peaberry and civet
coffee, two methods were attempted. First is using principal component analysis or PCA (unsupervised
chemometric analysis) method. The PCA was performed using all samples (80 samples) using original
spectral data in the range of 250-350 nm. The second method is using PLS-DA or partial least squares-
discriminant analysis. In this PLS-DA method, a PLS regression was employed with a dummy variable
as a reference value (variable Y). In this research, the response variable Y is composed of 0´s and 1´s,
where the value 1 is assigned to peaberry coffee samples and the value 0 to civet coffee samples. To
evaluate the performance of PLS-DA model, the coefficient of determination (R
2
), root mean square
error of calibration (RMSEC) and root mean square error of cross-validation (RMSECV) are used in
this research. It is expected to have ideal models with lower RMSEC and RMSECV as well as higher

(ICoSITeR) 2018
IOP Conf. Series: Earth and Environmental Science258 (2019) 012043
IOP Publishing
doi:10.1088/1755-1315/258/1/012043
3
R
2
. In the prediction, two statistic parameters were used to assess the model: SEP (standard error of
prediction) and bias. Those two parameters should be as low as possible. PCA and PLS-DA were
performed using the multivariate software of the Unscrambler 9.7 (CAMO software AS, Oslo, Norway).
3. Results and Discussion
3.1. UV-visible spectral data of civet and peaberry coffee with different particle sizes
UV-visible spectral data of peaberry coffee with different particle sizes was depicted in Figure 1. Each
data (peaberry mesh 12 and 50) was averaging of 10 spectral data. From Figure 1, we can notice that
there is a different of peaberry spectral data between mesh 12 and 50 especially in the wavelength of
250-350 nm. The intensity of absorbance of peaberry with mesh 50 is higher than that of mesh 12. The
similar phenomenon was observed for civet coffee as seen in Figure 2. This result showed that the quality
of spectral data in UV-visible spectroscopy of coffee samples is highly influenced by particle sizes of
ground roasted coffee samples.


Figure 1. Average original spectra of peaberry coffee samples with different particle sizes (mesh
12 and 50) in the range of 190-1100 nm.


Figure 2. Average original spectra of civet coffee samples with different particle sizes (mesh
12 and 50) in the range of 190-1100 nm.
3.2. Unsupervised method using PCA analysis
Figure 3 showed the result of PCA analysis of all samples (80 samples) using original spectra in the
range of 250-350 nm. Using two PCs (PC1 and PC2), a clear separation between civet and peaberry
coffee samples was observed. The total of PC1 and PC2 can explain 100% of the total variance of
original data. From here, we cannot see clearly the influence of different particle sizes on separation of

(ICoSITeR) 2018
IOP Conf. Series: Earth and Environmental Science258 (2019) 012043
IOP Publishing
doi:10.1088/1755-1315/258/1/012043
4
civet and peaberry. More quantitative analysis should be conducted using PLS-DA in order to see the
influence of particle sizes on discrimination between civet and peaberry ground roasted coffee.


Figure 3. The result of PCA analysis of all samples using original spectra in the range of 250-350
nm.
3.3. PLS-DA model using mesh 12
PLS-DA model was developed using calibration and validation samples with mesh 12. These samples
consist of 15 samples of peaberry (mesh 12) and 15 samples of civet (mesh 12). The result was showed
in Figure 4. The coefficient of determination was very high (R
2
= 0.99) for both calibration and
validation. The RMSEC = 0.053 and RMSECV = 0.060. It can be seen that all peaberry samples are
very close to 1 and civet coffee samples are very close to 0.


Figure 4. PLS-DA model developed using mesh 12 of original spectra in the range of 250-350 nm.
3.4. Prediction samples of mesh 12 and 50 using PLS-DA model mesh 12
The performance of PLS-DA model mesh 12 was evaluated in two ways. First, the developed PLS-DA
model mesh 12 was used to predict samples having mesh 12. Second, it was used to predict the samples
having mesh 50. The result was showed in Figure 5. In term of SEP and bias, the prediction of samples
having mesh 12 resulted in low SEP and bias (SEP = 0.085681 and bias = 0.030608). However, the
result of the prediction of samples having mesh 50 became worst with increasing both the value of SEP
and bias (SEP = 0.258729 and bias = 0.089416). This result has confirmed that the performance of PLS-
DA model for civet and peaberry discrimination is highly influenced by particle sizes of the ground
roasted coffee samples. For this, in the future for practical application, it is highly needed to develop
PLS-DA model that can compensate the influence of particle sizes of coffee samples.

(ICoSITeR) 2018
IOP Conf. Series: Earth and Environmental Science258 (2019) 012043
IOP Publishing
doi:10.1088/1755-1315/258/1/012043
5

Figure 5. The result of prediction of mesh 12 and 50 using PLS-DA model mesh 12.
4. Conclusion
In this research, we showed the influence of particle sizes of ground roasted coffee samples in the
authentication of peaberry and civet coffee. The influence was not so clear based on the result of PCA.
However, using PLS-DA method, the influence was seen clearly. It is confirmed that the performance
of PLS-DA model for civet and peaberry discrimination is highly influenced by particle sizes of the
ground roasted coffee samples. For this, in the future for practical application, it is highly needed to
develop PLS-DA model that can compensate the influence of particle sizes of coffee samples. By doing
this, a procedure of sieving may be removed. Finally, faster authentication process of coffee samples
using UV-visible spectroscopy could be established.
Acknowledgement
The authors would like to acknowledge Ministry of Research, Technology and Higher Education,
Republic of Indonesia (Kemenristekdikti) for financial support under Research Grant PENELITIAN
BERBASIS KOMPETENSI (PBK) (Grant Number: 384/UN26.21/PN/2018).
References
Suhandy D and Yulia M 2017 Int. J Food Prop. 20: S331–9
Yulia M and Suhandy D 2017 J. Phys.: Conf. Ser. 835 012010
Suhand D and Yulia M 2017 Int. J. Food Sci. 2017:1–7
Downey G, Boussion J and Beauchene D 1994 J. Near Infrared Spec. 2: 85–92
Suhandy D, Yulia M, Ogawa Y and Kondo N 2018 IOP Conf. Ser.: Earth Environ. Sci. 147
012011
Monteiro P I, Santos J S, Brizola V R A, Deolindo C T P, Koot A, Boerrigter-Eenling R, van
Ruth S, Georgouli K, Koidis A and Granato D 2018 Food Control 91: 276–83
Dupuy N, Huvenne J P, Duponche L and Legrand P 1995 Appl. Spectrosc. 49: 580–5
Obeidat S M, Hammoudeh A Y and Alomary A A 2018 J. Appl. Spectrosc. 84: 1051–5
Reis N, Franca A S and Oliveira L S 2013 LWT - Food Sci. Technol. 50: 715–22
Botelho B G, Oliveira L S and Franca A S 2017 Food Control 77: 25–31
Suhandy D and Yulia M 2018 IOP Conf. Ser.: Mater. Sci. Eng. 334 012059
de Oliveira E M, Leme D S, Barbosa B H G, Rodarte M P and Pereira R G F A 2016 Food Eng.
171: 22–27
Consonni R, Polla D and Cagliani L R 2018 Food Control 94: 284–8
Dong W, Zhao J, Hu R, Dong Y and Tan L 2017 Food Chem. 229:743–51
Yulia M, Asnaning A R and Suhandy D 2018 IOP Conf. Ser.: Earth Environ. Sci. 147 012010
Yulia M and Suhandy D 2018 MATEC Web of Conf. 197 09003
Suhandy D and Yulia M 2018 MATEC Web of Conf. 197 09002