AIP Conference Proceedings https://www.scimagojr.com/journalsearch.php?

q=26916&tip=sid&clean=0

Enter Journal Title, ISSN or Publisher Name

1 of 4 4/5/2023, 9:34 PM
AIP Conference Proceedings https://www.scimagojr.com/journalsearch.php?q=26916&tip=sid&clean=0

2 of 4 4/5/2023, 9:34 PM
AIP Conference Proceedings https://www.scimagojr.com/journalsearch.php?q=26916&tip=sid&clean=0

4 of 4 4/5/2023, 9:34 PM
Characteristics of array MOS gas sensors in
detection of adulteration on patchouli oil
with candlenut oil
Cite as: AIP Conference Proceedings 2586, 070016 (2023); https://doi.org/10.1063/5.0106394
Published Online: 24 January 2023

Arief Sudarmaji, Agus Margiwiyatno and Susanto Budi Sulistyo

ARTICLES YOU MAY BE INTERESTED IN

Effect of application of types of fertilizer and number of seeds per hill on production of wheat
(Triticum aestivum L.) Dewata 162 variety
AIP Conference Proceedings 2586, 020010 (2023); https://doi.org/10.1063/5.0106869

AIP Conference Proceedings 2586, 070016 (2023); https://doi.org/10.1063/5.0106394 2586, 070016

© 2023 Author(s).
 Characteristics of Array MOS Gas Sensors in Detection of
Adulteration on Patchouli Oil with Candlenut Oil
Arief Sudarmaji1, a), Agus Margiwiyatno1, b), and Susanto Budi Sulistyo1, c)
1
Department of Agricultural Engineering, Faculty of Agriculture, Jenderal Soedirman University, Jl. Dr. Soeparno,
Karangwangkal, Purwokerto 53132 Indonesia
a)
Corresponding author: [email protected]
b)
[email protected]
c)
[email protected]

Abstract. The high price of essential oils, especially Patchouli Oil, leads to adulteration by mixing with cheaper oils. The
similar color and viscosity make it difficult to be recognized or distinguished by human sensing. In general, a technique
for measuring essential oils is to use capillary Gas Chromatography/Mass Spectrometry (GC/MS). However, GC/MS is
high cost, time-consuming (requires sample preparation), and requires adequate skill. This study aims to determine the
characteristics of gas sensors made from Metal Oxide Semiconductor (MOS) which is used to detect adulteration of
patchouli oil with candlenut oil. The measurement uses 9 MOS to capture gas/aroma from samples in a chamber. The
output of MOS is acquired in a PC. The sample tested is pure Patchouli Oil, pure Candlenut Oil, 1% mixture, 5%
mixture, 10% mixture, 15% mixture, and 20% mixture. The mixture is Patchouli Oil with Candlenut Oil. The individual
response characteristics and the MOS series were analyzed using box plot graphs and Principal Component Analysis
(PCA) respectively. It is found that modulation on MOS led the sensors more sensitive and PCA results showed that the
system is adequate to distinguish mixing of in Patchouli Oil.

INTRODUCTION

Patchouli Oil is one of Indonesian potential export products which has high price in the international market.
Indonesia holds up to 85% world market share of Patchouli Oil [1], and is the three larger of International supplier
of Patchouli Oil [2][3]. Patchouli Oil has a high-priced value in the oil industry. The price of Indonesia’s Patchouli
Oil in August 2020 was 52.30 USD to 62.10 USD per kg for a minimum order quantity of 1000 kgs [4].
Patchouli Oil is mostly resulted by Small and Medium Enterprises (SMEs) from steam distillation of patchouli
plant (Pogostemon cablin Benth) is obtained through a distillation process which is taken from a separator tank. In
general, the separation is done by waiting for the oil to cool in order to avoid the water particles move to Patchouli
Oil. Oil contaminated with water will reduce shelf life that leads to lower price. Many Patchouli Oil made by SMEs
has not met the quality requirement to be exported.
The high economic value and high demand of Patchouli Oil, lead to its adulteration with other cheap oils to get
economical profit. SMEs intentionally added Patchouli Oil with cheaper oils to increase the weight. The practice of
adulteration becomes an issue for consumers, producers, and regulatory bodies. Candlenut Oil is one of cheaper oils
and has same color which can disguise its presence in Patchouli Oil. The lack of knowledge and the availability of
tools to detect the presence of impurities in Patchouli Oil cause local collector/SME to experience a lot of economic
losses when selling their products to exporter or industries [1]. Thus, it is needed an instrumental technique to assure
the authenticity of Patchouli Oil by indicating the presence of sufficient accuracy for the level of SME.
The determination of Patchouli Oil contents accurately and widely applied using the Gas Chromatography-Mass
Spectrometry method. GC and HPLC physical-chemical analysis are the most advanced methods among them.
Some adulterations can be detected simply by GC-MS tests with technology such as GC-IRMS and SNIF-NMR

The Third International Symposium on Food and Agrobiodiversity (ISFA 2021)

AIP Conf. Proc. 2586, 070016-1–070016-7; https://doi.org/10.1063/5.0106394
Published by AIP Publishing. 978-0-7354-4231-3/$30.00

070016-1
[5][6]. However, this method may not be used by SMEs because of the complexity of use and the price of the
equipment is very expensive, cannot be done by SMEs.
The gases and volatiles of essential oils can be captured using an array of Metal Oxide Semiconductor (MOS)
gas sensors. The MOS gas sensor is small and compact and are multi-series to detect a wide variety of compounds.
Several studies that have utilized the MOS sensor gas line, including [7] distinguished essential oils from sweet
oranges, chamomile flowers, and jasmine flowers, and [8] detected and classified the variety of Rosa Damascena oil
into three categories. And one technique that has been studied for a long time and succeeded in increasing the
sensitivity and selectivity of MOS is temperature modulation [9][10][11][12]. This technique was further developed
by [13] known as Temperature Modulation with Specified Detection Point (Temperature Modulation-SDP). The use
of MOS in detection of Patchouli Oil has been also done by [14] that successfully distinguished between Patchouli
Oil and Clove Oil clearly. However, there are no studies yet that test the modulations on MOS gas sensors to detect
the Patchouli Oil adulteration (such as with Candlenut Oil). Since the applied modulation leads unique response to a
particular substance, it is believed that the modulations will change according to the adulterant of Patchouli Oil.
Therefore, this paper presents the characterization of array MOS gas sensors that driven in Temperature Modulation
to indicate the adulteration substance (i.e., Candlenut Oil) in Patchouli Oil.

MATERIAL AND METHODS

The measurement materials include MOS gas sensors, sensor chamber, sample headspace, reference headspace,
bottle vial, syringe/vapor injector, mini air pump, acquisition unit, and computer unit with acquisition interface
software. The solution samples consist of pure Patchouli Oil, pure Candlenut Oil, and mixture of Patchouli Oil and
Candlenut (i.e. 5%, 10%, 15%, and 20%). Thus, we examined 6 categories of solution samples. While the content of
Patchoulol (patchouli alcohol) of Patchouli Oil were tested using the Gas Chromatography-Mass Spectrometry (GC-
MS) method.
The measurement setup and apparatus of measurement are shown in Fig. 1 and Fig. 2 respectively. The nine
MOS gas sensors (Table 1), which aimed to sense several odorous gases and volatiles, were used to capture the
vapor of Patchouli Oil. All sensors were put inside a sensor chamber made of 5 mm acrylic and formed into 693 cm3
square box (11x9x7) cm. MOS gas sensors are operated in dynamic mode with temperature modulation-SDP
technique. We tested 3 modulations: without modulation, 0.25 Hz modulation, and 1 Hz modulation. The outputs are
acquired to a computer through PSoC CY8C28445-24PVXI based interface unit and xBee wireless communication.
The excel file is created to store the output data.
Array MOS gas
sensors
Acquisition
unit
P1
inlet 1 Injection outlet
silica hole
gel
Sample Reference sensor chamber P2
headspace headspace

FIGURE 1. Measurement diagram for capturing vapor of Patchouli Oil sample

TABLE 1. MOS gas sensor for capturing gases and volatiles of Patchouli Oil
No Type Main Gas/Vapor Target Range
1. TGS-2602 Odorous gases (Ammonia, Ethanol) 1-100 ppm
2. TGS-2620 Solvent (organic) vapors 50 – 5,000 ppm
3. TGS-2600 Air Contaminants (H2, CO) 1-30 ppm
4. MQ-5 Natural gas, Coal gas 200-10,000 ppm
5. MQ-135 Air Quality Control 10-200 ppm
6. MQ-138 Wide volatile compound 200-10,000 ppm
7. FIS-12A Methane 300-7,000 ppm
8. FIS-30SB Alcohol 1-100 ppm
9. FIS-AQ1 Volatile organic compound 10-10,000 ppm

070016-2
 (b)

(c)
(a)

FIGURE 2. (a) Measurement apparatus, (b) sample of Patchouli Oil in a 15 ml vial headspace, (c) 10 ml syringe for transporting
gas and vapor of Patchouli Oil to sensor chamber

The samples used to be detected were as follow: (1) Pure (100%) Patchouli Oil, (2) Pure (100%) Candlenut oil,
(3) Mixture of 5% of Candlenut oil in Patchouli Oil, (4) Mixture of 10% of Candlenut oil in Patchouli Oil, (5)
Mixture of 15% of Candlenut oil in Patchouli Oil, and (6) Mixture of 20% of Candlenut oil in Patchouli Oil. We
tested three Modulations to drive the MOS gas sensors: (1) Without Modulation, (2) 0.25 Hz Modulation, and (3) 1
Hz Modulation. Each measurement of sample was five times repeated. Thus, we got 90 measurements.

Vapor Measurement Steps

The solution sample is prepared as follow: An aliquot (1 mL) of sample solution was put inside the static
headspace vial to generate its vapor. A 15 mL glass bottle with rubber cap was used as a static headspace vial. The
vapor of Patchouli Oil was manually delivered to the sensor chamber using a 10 mL syringe when measuring step.
One cycle measurement consists of two steps: Patchouli Oil vapor measurement (Rg), and purging phase as shown
in Fig. 3.
The pumps (P1 and p2) were switched on when warming up and purging phase and stopped when measuring the
Rg. The output was defined as Rg where Rg is resistance when measuring vapor of Patchouli Oil. Dry air (filtered by
silica gel) was constantly pumped to the sensor chamber when warming up and purging process. Acquisition
software was developed under Visual Studio 2019 to set the temperature modulation and to show and store the MOS
outputs into an MS Excel file.

Warming up Measuring Rg Purging

(30 minutes) (1 minute) (10 minute)

Next measurement

FIGURE 3. Measurement steps of Patchouli Oil.

Characterization and Analysis

We applied three ways to analysis the characteristic of MOS gas sensors in discriminating the type of samples
for indicator the adulteration. They are: (1) The individual response characteristics using box plot method, (2) The
significance of each MOS gas sensors to outputs in sensing samples using Loading plot of Principle Component
Analysis (PCA), and (3) Classification Performance using score plot of Principle Component Analysis (PCA). PCA
was performed by means of Minitab 16 Statistical Software.

070016-3
 RESULT AND DISCUSSION

Individual response of MOS gas sensors on various Patchouli Oil adulterations

One measurement (Rg) was acquired as many 30 data/sample, resulted from each 2 second in 1 minute
measurement. We used box plot analysis to show the ability of MOS to sense and distinguish the five kinds of
patchouli oil adulteration.
We found that the MOS sensors can sense vapor of patchouli oil with different adulterations. However, almost
the responses of MOS gas sensors in each modulation driven were highly overlap, as exemplified in Figure 4.
Therefore, there is no single MOS gas sensor used (i.e., general gases and volatiles sensor) that able to distinguish
among the type of samples, especially discriminating the non-adulterated and adulterated Patchouli Oil.
The compound complexity of Patchouli Oil vapor might cause the overlapping response. [15] reported that it is
found 32 detected peaks and their odor descriptions oI 3DWFKRXOL 2LO XVLQJ *&ဨ06 DQDO\VLV %HVLGHV WKH 026
sensors tended to drift [16] and to have poor selectivity (cross-sensitivity) to other gases [17,18]. MOS sensor could
be selective to a certain gas but its cross-sensitivity to other gases was not negligible [19]. Table 2 shows detected
peaks using GC-MS when measuring the 2 samples of Patchouli Oil.

25000 20000

17500
20000
Resistance (ohm)

Resistance (ohm) 15000

15000
12500

10000 10000

5000 7500

5000
0
Type % % % % aO % % % % aO % % % % aO
10 15 20 i x 5 e P 10 15 20 ix 5 e P 10 15 20 ix 5 e P Type % % % % aO % % % % aO % % % % aO
ix ix ix M ur ix ix ix M ur ix ix i x M ur 10 15 20 i x 5 e P 10 15 20 ix 5 e P 10 15 20 ix 5 e P
M M M P M M M P M M M P ix ix ix M ur ix ix ix M ur ix ix i x M ur
M M M P M M M P M M M P
0 2 0
60 60 62 Q
5 35 38
S2 S2 S2 M Q1 Q1
TG TG TG M M

(a) (b)

14000

12000

10000
Resistance (ohm)

8000

6000

4000

2000

0
Type % % % % aO % % % % aO % % % % aO
10 15 20 i x 5 e P 10 15 20 ix 5 e P 10 15 20 ix 5 e P
ix ix ix M ur ix ix ix M ur ix ix i x M ur
M M M P M M M P M M M P
Q1 B3
0 2A
SA SS S1
FI FI FI

(c)

FIGURE 4. Box plot of (a) TGS series, (b) MQ series, and (c) FIS series on 0.25 Hz Modulation.
TABLE 2. Peaks detected by means of GC-MS method of Patchouli Oil
Sample Patchoulol level GC-MS Peak detected
Patchouli Oil 1 26.94% 80
Patchouli Oil 2 29.52% 100

070016-4
 Identification of Patchouli Oil adulterations

We utilized the Principal Component Analysis (PCA) to evaluate the selectivity performance in discriminating
among the 5 adulteration treatments. Figure 8 shows the PCA plot of discrimination of five categories Patchouli Oil
performed by 9 MOS gas sensors used to each Modulation. It can be seen that Modulation on 0.25 Hz leads the
sensors to give clear selective response to discriminate between pure and mixture of Patchouli Oil.
PCA result also reveals that the first two principal components were holding 76% of the data. PCA is commonly
used as a feature extraction part to test selectivity performance and as a linear classification technique. It is usually
utilized in correlation with cluster analysis and visualization of the differences among the treatments. The large
dimension of interrelated variables is reduced into few important principal components. The first two or three
uncorrelated components hold the most significant variation present in all variables and widely used in the various
applications [20].
2000 Ty pe -6000 Ty pe
Mix 10% Mix 10%
Mix 15% Mix 15%
1000 Mix 20% -7000 Mix 20%
Mix 5% Mix 5%
Pure PaO Pure PaO
0 -8000
Second Component

Second Component
-1000 -9000

-2000 -10000

-3000 -11000

-4000 -12000

10000 15000 20000 25000 30000 35000 40000 6000 7000 8000 9000 10000 11000 12000 13000 14000 15000
First Component First Component

(a) (b)
-15000 Ty pe
Mix 10%
Mix 15%
-17500 Mix 20%
Mix 5%
Pure PaO
-20000
Second Component

-22500

-25000

-27500

-30000

-32500
-30000 -25000 -20000 -15000 -10000 -5000
First Component

(c)
FIGURE 5. Score plots of each MOS gas sensor on (a) without modulation, (b) 1 Hz modulation, and (c) 0.25 Hz modulation
Then, we used the loading plot of Principal Component Analysis (PCA) to know the effect or significance of
each MOS gas sensor to the resistance in sensing the 5 samples of adulterations. As shown in Figure 6, the length of
loading plots of all MOS gas sensors were not similar. Therefore, we chose only 6 MOS gas sensors that give high
significance to evaluate their performance for indicating the presence of adulterant substance in patchouli oil. The 6
MOS gas sensors are TGS-2600, TGS-2602, TGS-2620, MQ-5, MQ-138, and FIS-AQ1.

070016-5
 TGS2600
0.2

0.1
FISSB30
0.0

Second Component
FIS12A
-0.1
MQ135
TGS2602
-0.2
MQ5
-0.3

-0.4
TGS2620
MQ138
-0.5 FISA Q1

-0.6
-0.75 -0.50 -0.25 0.00
First Component

FIGURE 6. The loading plot on MOS gas sensor on 0.25 Hz Modulation

The performance of the 6 chosen MOS gas sensor to distinguish 6 samples of adulteration is shown in Fig. 7. It
can be seen that using the 6 Selected MOS gas sensors, the selectivity performance of MOS to distinguish between
pure Patchouli Oil and Mixed with candle Oil is similar to all MOS. It means the eliminated MOS gas sensors really
less significance, yet the system still have same performance. However, it also reveals that there are many miss-
clustering among groups of Patchouli Oil with adulteration to discriminate the level of composition/percentage. It
seems that there is similar substance between Patchouli Oil and Mixed with candle Oil since they are made from
vegetable oil.
2 Ty pe
30000 Ty pe
Pure CaO Mix 10%
Pure PaO Mix 15%
Mix 20%
1 27500
Mix 5%
Pure PaO
Second Component

Second Component

25000
0

22500
-1
20000

-2
17500

-3 15000

-5 -4 -3 -2 -1 0 1 2 3 4 -30000 -25000 -20000 -15000 -10000 -5000

First Component First Component

(a) (b)
FIGURE 7. The performance of selected 6 MOS gas sensor to discriminate: (a) between pure Patchouli Oil and Candlenut Oil,
and (b) between Pure Patchouli Oil and adulterated Patchouli Oil.

CONCLUSION

The research presents a test of MOS gas sensors (TGS-2600, TGS-2602, TGS-2620, MQ-5, MQ-135, MQ-138,
FIS-AQ1, FIS-SB30, and FIS-12A) for measuring vapor of Patchouli Oil in mixture Candlenut Oil by 5%, 10%, and
15%. It is found that the 6 most significant MOS gas sensors to discriminate between pure Patchouli Oil and mixture
Patchouli Oil are TGS-2600, TGS-2602, TGS-2620, MQ-5, MQ-138, and FIS-AQ1) By PCA classification, the
system which applied modulation of 0.25 Hz on MOS gas sensors is proper to distinguish clearly between pure
Patchouli Oil and adulterated Patchouli Oil (mixed with Candlenut oil). However, it has not been able to determine
the classification of the mixture level of the adulterated Patchouli Oil.

070016-6
 ACKNOWLEDGMENTS

Authors thank to Universitas Jenderal Soedirman and DIKTI for providing fund for this research through scheme
of Riset Unggulan 2021.

REFERENCES
1. W. Haryono, Export News Indonesia: patchouli oil. Jakarta, Indonesia: 2015.
2. E. Schmidt. "Production of Essential Oils", In: Bas‚er KHC, Buchbauer G, editors. Handb. Essent. Oils Sci.
Technol. (Appl., Boca Raton, FL, CRC Press, Taylor & Francis Group, 2016), pp. 1116.
3. A. Hapsari. Indonesia Pensuplai Utama Tiga Komoditi Minyak Atsiri. Suara Merdeka News 2018.
4. Jandico. Aroma Atsiri: Price List August 2020. 2020.
5. E. Schmidt and J. "Wanner, Adulteration of Essential Oils", In: Bas‚er KHC, Buchbauer G, editors. Handb.
Essent. Oils Sci. Technol. (Appl., Boca Raton, FL: CRC Press Taylor & Francis Group, 2016), pp. 1116.
6. F. Swastawati., A. Ambariyanto., B. Cahyono., I. Wijayanti., D. Chilmawati., H. Hadiyanto, and A. N. Al-
Baarri, Physicochemical changes and sensory quality of liquid smoked milkfish nuggets, African Journal of
Food, Agriculture, Nutrition and Development 21, 18261–18278 (2021).
7. S. Kim, H. Kim, I. C. Lee, K. Cheong, and M. Lim, The quantitative analysis of aroma gas with gas
sensors, Kor J. Aesthet Cosmetol 11, 1163–1179 (2013).
8. A. M. Gorji-chakespari., F. Nikbakht., M. Sefidkon., Ghasemi-varnamkhasti, and E. L. Valero,
Classification of essential oil composition in Rosa damascena Mill. genotypes using an electronic nose, J
Dermatol Sci 4, 27–34 (2016).
9. S. Ortega., A. Marco., T. Perera., Šundic., A. Pardo, and J. Samitier, an intelligent detector based on
temperature modulation of a gas sensor with a digital signal processor, Sensors Actuators, B Chem 78, 32–
9 (2001).
10. Y. Sun, X. Huang, F. Meng, J. Liu, Study of influencing factors of dynamic measurements based on SnO2
gas sensor, Sensors 4, 95–104 (2004).
11. X. Huang, F. Meng, Z. Pi, W. Xu, and J. Liu, Gas sensing behavior of a single tin dioxide sensor under
dynamic temperature modulation, Sensors Actuators, B Chem 99, 444–50 (2004).
12. J. Liu, X. Huang, and F. Meng, The Dynamic Measurements of SnO2 Gas Sensors and Their Applications.
(Nova Science Publishers, New York, NY, USA, 2007), pp. 177–214.
13. A. Sudarmaji and Kitagawa, Sensors and transducers temperature modulation with specified detection point
on metal oxide semiconductor gas sensors for e-nose application, Sensors and Transducers 186, 93–103
(2015).
14. A. Sudarmaji, A. Margiwiyatno, R. Ediati, and A. Mustofa, "Vapor measurement system of essential oil
based on mos gas sensors driven with advanced temperature modulation technique", in IOP Conf Ser Earth
Environ Sci, edited by. (IOP Publishing, Bogor, 2018), pp. 1–10.
15. T. A. V. Beek and D. Joulain, The essential oil of patchouli, Pogostemon cablin: A review. Flavour Fragr J
33, 6–51 (2017).
16. A. Hierlemann and R. Gutierrez-Osuna, Higher-order chemical sensing, Chem Rev 108, 563–613 (2008).
17. A. Bermak, S. B. Belhouari, M. Shi, and D. Martinez. Pattern Recognition Techniques for Odor
Discrimination in Gas Sensor Array, Encycl Sensors 10, 1–17 (2005).
18. S. Di Carlo and M. Falasconi, "Drift Correction Methods for Gas Chemical Sensors in Artificial Olfaction
Systems: Techniques and Challenges" in Advances Chemical Sensor, (InTech, London, 2012) pp. 305–26.
19. A. Wilson and M. Baietto, Applications and advances in electronic-nose technologies, Sensors. 9, 99–148
(2009).
20. Z. Haddi, M. Bougrini, K. Tahri, Y. Braham, M. Souiri, N. El Bari., A. Maaref., A. Othmane., N. Jaffrezic-
Renault, and B. Bouchikhi. A hybrid system based on an electronic nose coupled with an electronic tongue
for the characterization of moroccan waters, Sensors & Transducers 27, 190–197 (2014).

070016-7