Content uploaded by Mohammed Faeik Ruzaij Al-Okby
Author content
All content in this area was uploaded by Mohammed Faeik Ruzaij Al-Okby on Jan 01, 2024
Content may be subject to copyright.
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
MOBILE GAS SENSING FOR LABORATORY
INFRASTRUCTURE
RADUAN SARIF 1, MOHAMMED FAEIK RUZAIJ AL-OKBY 2,3*,
THOMAS RODDELKOPF1 AND KERSTIN THUROW3
1Institute of Automation, University of Rostock, 18119 Rostock, Germany
2Technical Institute of Babylon, Al-Furat Al-Awsat Technical University (ATU), Kufa 54003, Iraq
3Center for Life Science Automation (Celisca), University of Rostock, 18119 Rostock, Germany
*Corresponding author: mohammed.al_okby@atu.edu.iq
(Received: 31 January 2023; Accepted: 30 November 2023; Published on-line: 1 January 2024)
ABSTRACT: Indoor air quality has become a growing concern in modern society due to
prolonged indoor working hours that lead to the frequent exposure to numerous toxic gases
from various sources. These pollutants, including volatile organic compounds (VOCs), pose
severe health risks such as asthma and lung cancer. To address this critical issue, this project
focuses on developing and evaluating an advanced gas detection system that explicitly targets
VOCs by integrating two novel metal oxide semiconductor (MOX)-based gas sensors, ENS
160 and TED110. Different sensor parameters, such as the air quality index (AQI) and volatile
organic compounds (VOCs), were evaluated using 12 volatile organic chemicals. The
findings revealed that the ENS 160 sensor performs excellently, detecting 60 gas samples out
of 72, with an average detection rate of approximately 83%. In contrast, the TED110 sensor
demonstrated considerably lower performance and response in 24 out of 72 gas samples, with
a detection rate of about 33%. The results contribute insights into the gas sensor's
characteristics, providing essential information to enhance indoor air quality monitoring
technology, particularly in laboratory environments.
ABSTRAK: Kualiti udara dalam bangunan semakin mendapat perhatian di kalangan
masyarakat moden kerana waktu bekerja yang panjang di dalam bangunan dan ini
berpotensi terdedah kepada gas toksik dari pelbagai sumber. Pencemaran ini, adalah
termasuk kepada sebatian organik mudah meruap (VOCs), yang menimbulkan resiko
kesihatan yang teruk, seperti asma dan kanser paru-paru. Bagi menangani isu kritikal ini,
projek ini memfokuskan tentang sistem pembangunan dan penilaian secara eksplisit
mensasarkan VOCs dengan mengintegrasikan dua pengesan gas berasaskan semikonduktor
logam oksida (MOX), ENS 160 dan TED110 yang baru. Parameter berbeza pada pengesan,
seperti indeks kualiti udara (AQI) dan sebatian organik mudah meruap (VOCs), dinilai
menggunakan 12 bahan kimia organik mudah meruap. Dapatan menunjukkan pengesan
ENS 160 berjaya mengesan 60 jenis gas daripada 72 jenis, dengan purata kadar identifikasi
sebanyak 83%. Secara perbandingan pengesan TED110 hanya mengesan 24 daripada 72
sampel gas, dengan kadar pengesanan sebanyak 33%. Dapatan ini menyumbang kepada
pemahaman tentang ciri-ciri pengesan gas, penyumbang kepada pengetahuan penting
tentang teknologi pemantauan kualiti udara iaitu secara khususnya dalam persekitaran
makmal.
KEY WORDS: Mobile gas sensing, hazardous gas detection, volatile organic compounds,
environmental gases, gas sensors, toxic gases.
178
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
1. INTRODUCTION
In laboratory infrastructure, employees should be familiar with the chemicals they may
come across, potentially reactive or explosive liquids and gases that can be highly hazardous.
Accidental or uncontrolled chemical reactions are significant causes of severe personal injury
and property damage. Dangerous gases are sufficiently toxic or reactive substances that
vigorously or violently give off heat and energy, and become poisonous in contact with air,
water, or some other common material. They can be classified in various ways, including
acutely toxic, corrosive, flammable, dangerously reactive, and oxidizing compounds [1–7].
Toxic compounds include hydrogen chloride, benzene or toluene, dioxin, and volatile organic
compounds (VOCs, such as hydrocarbons, fluorocarbons, chlorofluorocarbons.), or elements
such as cadmium, mercury, and chromium. Different VOCs in indoor air are produced from
building materials, for instance, wood, cement, stones, asbestos used during construction, and
utility items placed inside the building, such as carpets. These are sources of hazardous/toxic
gases that can be inorganic, organic, biological, or even radioactive.
In the laboratory infrastructure for learning and experimentation purposes, the laboratory
staff must handle a variety of dangerous, poisonous, and reactive chemicals or gases. A
particular quantity of hazardous/toxic gases pollutes the atmosphere and can significantly
influence human health, creating severe illnesses and threatening worker safety. This
expanding number of dangerous gases is sometimes the cause of catastrophic incidents, ruining
assets and the causing the deaths of many people [4], [8–10]. Therefore, toxic gases may be
acidic, explosive, and extremely dangerous, depending on the concentration and surroundings.
A hazardous gas may harm living tissues, affect the central nervous system, cause severe
disease, or, in the worst situations, result in death when consumed, breathed, or absorbed by
the skin or eyes, according to specialists in gas detection. Furthermore, the employee may
regularly be in contact with various hazardous gases in the chemical research laboratory [11],
[12]. For instance, long-time exposure to the following gases, CO2, carbon monoxide, and
nitrogen oxide (NO2), can cause headaches, dizziness, restlessness, tingling, or pins or needles
feeling, difficulty breathing, sweating, tiredness, increased heart rate, elevated blood pressure,
coma, asphyxia, and convulsions [13–15]. The effects of different VOCs on human health,
such as carbonyl and aromatic compounds, like HCHO, CH3CHO, C6H6, C₆H₅CH₃, and C₈H₁₀,
severely impact human health and are causes of cancer. Besides, inhaling these compounds can
lead to lung cancer [13–17], so experimenting with the effect of these compounds on human
health is worthwhile for researchers. Numerous commercial gas detection sensors are available,
such as MOX, electrochemical, catalytic, and optical infrared, detect hazardous gases including
volatile organic compounds (VOCs). These sensors can be portable or fixed devices and
provide alarms when gas concentrations exceed specified thresholds. Several criteria should be
considered to evaluate the performance of gas sensors, such as sensitivity, selectivity, stability,
response time, reversibility, energy consumption, adsorptive capacity, and fabrication cost.
These factors play a crucial role in determining the effectiveness and reliability of gas detection
systems in different applications [18] [19].
Neubert et al. [20] discussed a modular Internet of Things (IoT)-based sensor node for
hazardous gas detection and monitoring. This experiment used two MOX gas sensors which
are BME688 (Bosch Sensortec, Reutlingen, Germany) and SGP30 (Sensirion AG, Stafa,
Switzerland). Moreover, a WROOM WiFi module (Espressif Systems, Shanghai, China)
transfers the collected data to an IoT cloud for data monitoring and storage. The processing
unit for this project was an NXP MKL27Z128VLH4 (NXP Semiconductors N.V., Eindhoven,
Netherlands) ARM microcontroller. Furthermore, the experiments were done with various
VOCs as a standalone unit and hosted by a stationary and mobile robot.
179
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
Two metal oxide (MOX) gas sensors were tested with Al-Okby et al., namely SPG30 and
SPG40 (Sensirion AG, Stäfa, Switzerland), to measure the indoor air quality parameters IAQ
index and the total volatile organic compound TVOC [7]. The WeMos D1 Mini is a WiFi-
based Microcontroller for IoT applications that has been used in the following project with the
chip ESP8266 (Espressif Systems, Shanghai, China). The sensors have been tested on various
VOC compounds in two different test conditions. Several sensor placements, including a
moving robot, were utilized to assess the effectiveness of the two sensors based on the recorded
characteristics (IAQ-index and TVOC) [19–21].
Demonstrating a gas sensor system for domestic air quality monitoring, K. Gupta et al.
[24] applied Tin Dioxide (SnO2) based MOX gas sensors MQ-135, MQ-6, and MQ-4 ( Winsen
Electronics Technology, Zhengzhou, China.). They detected ammonia (NH3), nitrous gases
(NOx), nicotine, benzene, carbon dioxide (CO2), butanes, LPG, propane, and LNG, and natural
gases (methane, CH4) with an Arduino UNO microcontroller and ESP8266 (Espressif Systems,
Shanghai, China) for the WiFi communication interface. Data was shown on the LCD screen
and stored on the server. This system can be used as a wireless sensor network for
environmental monitoring. Similar technology is utilized for multiple purposes in [23–26].
W. Wojnowski et al. employed electrochemical sensors in E-noses [29]. The sensors
included DGS-CO 968-034, DGS-Ethanol 968-035, DGS-H2S 968-036, DGS-NO2 968-037,
DGS-SO2 968-038, DGS-RESPIRR 968-041, 2E 50, 3E 100 SE (SPEC Sensors LLC, Newark,
CA, USA). They were used for the measurement of carbon monoxide (C.O.), ethanol, hydrogen
sulfide (H2S), nitrogen dioxide (NO2), sulfur dioxide (SO2), VOC, and ammonia (NH3). The
communication interface used a USB driver with an LTC2433 ADC (Linear Technology,
Milpitas, California, United States) to transfer the data to a PC-class computer.
A. Somov et al. presented a wireless sensor-actuator system for methane detection using a
catalytic sensor (NTC-IGD, Stockport, Russia) [30]. Using the Micro Controller Unit (MCU)
ADuC836 (One Technology Way, Norwood, USA), the node was connected to the WSN via
the ETRX3 module; the communication interface was UART. The calibration notices were
recorded in EEPROM M95640, connected to MCU using SPI. Apart from the calibration
information, the memory chip stored information on the occurring events, e.g., emergencies.
S. Esfahani et al. developed an electrical nose for VOC, carbon dioxide (CO2), and
methane (CH4) using optical and infrared sensors such as LFP3144C-337, LFP-3850C-337,
LFP-8850-337, and 90 V LFP- 8850-337 (InfraTec GmbH, Dresden, Germany) [31]. This work
used a Teensy 3.6 (PJRC, Portland, USA) microcontroller with a UART communication
interface. A laptop with a USB serial port was connected for storing and displaying data. This
portable optical e-nose can be applied to a robot for environment monitoring.
A fumigant gas trace detection system (FGTDS) based on a photoionization detector (PID)
was designed for the inspection and quarantine port to monitor the gas leakage within the
dosing room of the fumigant warehouse [32]. It used PID-A1 (Alphasense, Braintree, U.K.)
with MCU STC12LE5A60S2 (Shenzhen LCSC Electronics Technology, Shenzhen, China)
and ADS8325 A/D (Texas Instruments, Texas, United States) converter, which was applied for
MCU controls.
This research focused on developing and evaluating an advanced volatile organic
compounds (VOCs) detection system using novel metal oxide semiconductor (MOX)-based
gas sensors. Besides, the study investigated the sensor performance, such as sensitivity,
selectivity, and detectable gas limit, focusing on distinct sensor parameters, including the air
quality index (AQI) and total volatile organic compounds (VOCs). Overall contribution was to
develop an advanced gas detection system specifically for VOCs using two novel MOX-based
180
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
gas sensors, ENS160 (Sciosense B.V., Eindhoven, Netherlands) and TED110 (Wise Control
Inc, Seoul, Korea). In addition, it evaluated distinct sensor parameters, such as AQI and
TVOCs, in a typical chemical hood environment with 12 distinct VOCs. It also analyzed the
sensors' different characteristics, such as sensitivity, accuracy, response, and recovery time.
The comparison of the system's stability, accuracy, and effectiveness with existing design from
the University of Rostock, including sensors BME688 (Bosch Sensortec, Reutlingen,
Germany), SGP 40, and SGP 30 (Sensirion AG, Stäfa, Switzerland), offers valuable
benchmarks and validates the system's performance. In a nutshell, the findings from this
research enhance the understanding of gas detection technology and provide essential insights
for improving indoor air quality monitoring systems and safety, particularly in laboratory
environments [7], [20].
2. MATERIALS AND METHODS
2.1. Sensor Selection
One of the significant challenges of a gas detection system is determining the appropriate
gas sensor type. Different gas sensor technologies have limitations; none can be used for all
gas types or applications. The primary goal of this project is the detection of indoor TVOC
with a specific focus on a quick response for the safety laboratory employees. The initially
chosen MOX gas sensor ENS160 (Sciosense B.V., Eindhoven, Netherlands; see Fig. 1) can
measure three separate parameters, AQI (100 to 500), TVOC (0-65,000 PPB-Parts per billion),
and eCO2 (0-65,000 PPM-Parts per Million). The TED110 (WISE Control Inc, Seoul, Republic
of Korea), shown in Fig. 2, was chosen as a second gas sensor for the detection of a wide range
of gases in concentrations between 1 and 1,000 ppm, including VOCs, carbon monoxide,
ethanol, methane, nitrogen dioxide, toluene, and hydrogen sulfide. Detailed descriptions of
both sensor specifications are provided in Table 1:
Fig. 1: ENS 160 Gas Sensor, Sciosense B.V., Eindhoven, Netherlands [33]
Fig. 2: TED110 sensor, Wise control inc., Seoul, Republic of Korea [34].
181
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
Table 1: Selected Sensor Specifications
Sensor Specifications
ENS 160
TED 110
Structure
Metal Oxide (MOX)
MOX type Micro-Electro-
Mechanical System (MEMS)
Measure Gases
AQI (100-500), TVOC (0 –
65,000 PPB), (400 – 65,000
PPM)
TVOC (VOCs, CO, EtOH,
CH4, NO2, Toluene, H2S et)
(1-1000 PPM)
Humidity and temperature
Yes
Yes
Response time
1s
10s
Warm-up
< 3 min
< 50 seconds
Communication Interface
I2C and SPI
I2C
Positive supply
1.8V(VDD) & 3.6V(VDDIO)
3.3 V
Lifetime
10 years
5 years
Package dimension
3.0 × 3.0 × 0.9 mm3
3 × 3 × 1 mm3
Cost
$6.06
$12.50
Manufacturers
Sciosense B.V., Eindhoven,
Netherlands
Wise control inc., Seoul,
Republic of Korea.
2.2. Microcontroller Selection
Microcontrollers are used to analyze and process the measured data, in decision-making,
and in sending the proper action signals to the output ports. The Kinetics KL27
Microcontroller, illustrated in Fig. 3, was chosen for this experiment, which uses an
MKL27Z128VLH4 processor (NXP Semiconductors, Eindhoven, Netherlands). The project
selected this Microcontroller [18], which was tested before with two gas sensors (BME 688
and SGP 30). It is optimized for cost-sensitive and battery-powered applications requiring low-
power USB connectivity. The specification of the MCU is shown in Table 2.
Table 2: Specification of the Microcontroller
2.3. Experimental Setup
This project is the extension of an existing developed system called CELISCA at the
University of Rostock, Germany [20], which consists of two sensors in the sensing layers, such
as BME688 (Robert Bosch GmbH, Stuttgart, Germany) and SGP40 (Sensirion AG, Stäfa,
Switzerland), as shown in Figure 5(a). The sensor layer was tested with the processing layer
Specifications
Values
Core Type
Arm Cortex-M0+
Operating Frequency (MHz)
48
Number of bits
32bit
Temperature range (°C)
-40° to 105 °C
Flash (kB)
128
SRAM (kB)
32
Serial Communication
2 × I²C,2 × SPI,3 × UART
Supply Voltage (V)
1.71 V to 3.6 V
Power supply and data Transfer
USB-C
182
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
NXP MKL27Z128LH4 MCU (NXP Semiconductors N.V., Eindhoven, Netherlands) board.
Since gas sensors are heat sensitive, these sensor layers were designed in two fingers to keep a
particular air gap. The heat generated by the sensors may interfere with the degradation of their
function. The main objective of the current project was the extension of a novel sensor layer
and the design of a relevant system. For the extension, two novel gas sensors, ENS160
(Sciosense B.V., Eindhoven, Netherlands), TED110 (Wise Control Inc, Seoul, Korea), and the
processing layer MKL27Z128LH4 MCU board (NXP Semiconductors, Eindhoven,
Netherlands) were added. Developing the design idea from the previous project, the first
selected sensor, ENS160, was placed in the middle of the previous two sensors, shown in Fig.
5 (b). Finally, by adding the TED110, the overall sensor board then has four fingers - TED110,
BME688, ENS160, and SGP40 as seen in Fig. 5(c). All selected sensors and necessary
electronics were designed on a printed circuit board (PCB) by Autodesk Eagle software
(Autodesk, San Rafael, California, USA). After manufacturing the PCB, sensors and relevant
electronics were mounted on the board.
Fig. 3: MKL27Z128VLH4 MCU board
This project consists of two main layers, as shown in Fig. 4. The first layer is the sensing
layer, which detects various gas parameters. The second layer is the processing layer (MCU)
with the power supply, which receives the measured data from the sensing layer and processes
it. The MCU acts as a master, communicating with the sensors (working as slaves) using the
I2C communication protocol. The MCU sends a register address to initiate the sensor's I2C
clock, baud rate, and data length; after receiving the initiate acknowledgment command from
the sensor, the MCU requests the sensor data. Then the MCU writes register addresses on the
sensors for individual parameters, reads them as gas data, and converts all the data into
appropriate units. The MCU's programming was performed using MCUxpresso IDE (NXP
Semiconductors N.V., Eindhoven, Netherlands) and debugged using J-Link EDU (SEGGER
Microcontroller GmbH, Monheim am Rhein, Germany). The detection results were displayed
in Tera Term (open-source software under the BSD License) serial monitor connected via
USB-C and saved in a CSV file.
Fig. 4: Overall project structure
USB Power
Supply
MKL27Z128L
H4
(MCU)
ENS160, TED
110 (gas sensor)
Processing layer
Sensing layer
(I2C)
Bus
183
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
Fig. 5: (a) Sensor board with BME688 and SGP40; (b) BME688 and SGP40 and ENS
160; (c) TED110, BME688, ENS160 and SGP40
2.4. Experimental Procedure
To evaluate the sensor characteristics, 12 different low volatile organic compounds were
selected, which are commonly used in the laboratory [7], [12], [35]. Among the 12 VOCs,
benzene, toluene, and formic acid are exceptionally toxic and carcinogenic, posing significant
risks of long-term health effects, including leukemia and cancer, even at low exposure levels
[36]. Additionally, acetone, acetonitrile, dichloromethane, diethyl ether, ethanol, heptane,
hexane, and iso-propanol are harmful to human health as they can cause respiratory irritation,
dizziness, and in severe cases, damage to organs such as the liver and nervous system when
exposed to elevated concentrations [2], [37]. So, detecting all these gases is crucial for ensuring
safety in the laboratory and assisting in developing better gas detection technologies. Table 3
displays the selected 12 gases with their molecular formulas and the boiling point [7]. The
entire experiment was done in a classical chemical hood (Waldner Holding GmbH and Co.
KG, Wangen im Allgäu, Germany); Fig. 6 illustrates a hood design for a laboratory for
chemical and analytical purposes. Eppendorf pipettes were used to inject the testing samples
within a Petri dish (Eppendorf AG, Hamburg, Germany). In addition, the sensor node was
mounted on a movable stand with a manually adjustable height. The experiment was performed
for 5 minutes, during which data was taken for 300 s. The amount of the gas samples were 5μL,
10μL, and 50μL; all experiments were done from 40 cm and 100 cm sensor node distance from
the testing vapors. For the ENS160, the target parameter is the air quality index AQI (100 to
500), total volatile organic compounds concentration TVOC (0 – 65,000 ppb), and CO2 (400 –
65,000 ppm). The target parameter for the TED 110 gas sensor is the Gas Density (0 – 1,000
ppm). Every second of data was stored in a CSV file for further graphical visualization. Table
4 displays the various levels of gases for ENS160.
Table 3: Selected 12 gases with the molecular formula and the boiling point [7]
(a)
(b)
(c)
184
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
Table 4: ENS160 concern for different gas concentrations [33]
To avoid any influence from air ventilation, the ventilation system of the hood was shut
off during system testing. The chemical hood is shown in Fig. 6.
Fig. 6: Experiment in a traditional chemical hood
3. RESULTS AND DISCUSSION
The goal of the tests in all scenarios and positions was to determine the smallest quantity
of VOC detected by the two utilized sensors from the testing distance (e.g., the length between
the VOC leakage source and the sensors). The sensor responses were analyzed by presenting
all the experimented data in a graph. The result can be defined for ENS 160 from Table 4 as
low response (all parameters are in excellent level), moderate response (all parameters are good
& moderate level), and excellent response (all parameters are poor & unhealthy level).
Additionally, TED110 has low performance in this experiment, so in this research, just its
response was tested.
The ENS 160 gas sensor demonstrates high effectiveness and performance for acetone
(C3H6O) detection, showing excellent response to sample amounts >5μL and all desired
parameters, including AQI, TVOC, and CO2 . Its reliable performance at lower sample
concentrations showcases its sensitivity and ability to detect acetone accurately. On the other
hand, the TED110 sensor's low response to acetone, except at >50μL and 100 cm distance,
indicates a low gas detectable limit, assuming a reason of cross-sensitivity or environmental
interference. The principle of the MOX gas sensor is that its surface is adsorbed by oxygen,
changing the sensor's response quickly [38], so that the ENS160 gas sensor strongly reacts to
oxygen containing compounds, but TED 110 needs intensive investigation to improve
performance and expanding detection limit. Both sensor responses for acetone are presented in
Fig. 7.
185
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
Acetonitrile (C2H3N) has an excellent response with ENS160's required parameters with a
volume >5μL, illustrated in Figure 8, showing a good sensitivity and detection level. The AQI,
TVOC, and CO2 exhibit an excellent response for >10μL. TED110 gas sensor (gas density) has
no response for any amounts of acetonitrile 40 cm and 100 cm. ENS 160 has good performance
without oxygen-containing compounds, but TED110 failed in this context; further
improvement is necessary.
Detecting low benzene concentrations (C6H6) is essential because inhaling it for a long
time can cause cancer. ENS160 has an excellent response with the benzene samples
concentration of >10μL for all required parameters, such as AQI, TVOC, and CO2, presented
in Fig. 9. Although it has no oxygen compounds, ENS 160 is responding, but the TED110 gas
sensor has no response for benzene. ENS160's ability to detect benzene at concentrations
>10μL demonstrate that it has a low detection limit (has no oxygen compounds); further
calibration is necessary to enhance its gas detectable limit and broaden its application range for
benzene detection.
Dichloromethane (CH2Cl2) increases the risk for several specific cancers, including brain,
liver, and biliary tract cancer. Unfortunately, as seen in Fig. 10, ENS 160 has a low response
to this vapor, and the TED110 gas sensor has no response for all the required parameters. The
low response of both sensors creates concern about the sensor's traceability, testing methods,
sensitivity, and selectivity, probably following the MOX sensor principle. In short, to improve
accuracy and eliminate external factor contamination, a more extensive experimental procedure
must be developed or choosing gas sensors with appropriate sensitivity and selectivity.
ENS160 has an excellent response to diethyl ether (C4H10O) from the sample amount of
>5 μL for all the necessary parameters, indicating its significant effectiveness and detection
capability, illustrated in Fig. 11. The AQI, TVOC, and CO2 showed excellent responses for the
sensor node distances, such as 40 cm and 100 cm. ENS 160 has an excellent response; although
no oxygen compound, the TED110 gas sensor (Gas Density) declines to respond to diethyl
ether.
Ethanol (C2H6O) has an excellent response of >5μL for all the desired parameters in
ENS160. Similarly, the TED110 gas sensor (gas density) has a reaction for ethanol in all gas
sample concentrations. Fig. 12 exhibits both sensor responses for ethanol. Overall, based on
the MOX gas sensor principle, both sensors possess appropriate gas detectable limits and
sensitivity, which makes them suitable for ethanol detection.
Formic Acid (CH2O2) has a low response in the ENS 160 gas sensor for all the expected
parameters (AQI, TVOC, and CO2). On the other hand, the TED110 gas sensor (gas density)
has a response to formic acid, showing the highest reaction for sample size at >5μL at 40 cm.
Similarly, for the 100 cm distance, the TED110 gas sensor (Gas Density) has a response from
>10μL gas samples. Both sensors' response to formic acid are shown in Fig. 13. Based on
oxygen compounds, ENS160 should have had a reaction but failed; TED110 has a response
but low detection level, so both sensors require more calibration and intensive investigation to
improve sensitivity, effectiveness, and performance to the formic acid.
As shown in Fig. 14, heptane (C7H16) has a low response for all the desired parameters at
the ENS160 gas sensor. Similarly, the TED110 gas sensor (Gas Density) has no response. As
no oxygen compounds, both sensors declined to respond. As a result, testing methods for the
current sensors or the adoption of a heptane-specific sensor, need to improve.
186
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
a-1 a-2
b-1 b-2
c-1 c-2
d-1 d-2
Fig. 7: Acetone (C3H6O) response for ENS160 (a, b, c) and TED110 (d) at 40 cm (1) and
100 cm (2)
0
100
200
300
400
500
600
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
100
200
300
400
500
600
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
2000
4000
6000
8000
10000
12000
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
5000
10000
15000
20000
25000
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
1000
2000
3000
4000
5000
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
2000
4000
6000
8000
10000
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
0.2
0.4
0.6
0.8
1
1.2
0 100 200 300
Gas density (PPM)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
0 100 200 300 400
Gas density (PPM)
Time(sec)
5µL 10µL 50µL
(
c)
(
d)
187
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
a-1 a-2
b-1 b-2
c-1 c-2
Fig. 8: Acetonitrile (C2H3N) response for ENS160 (a, b, c) at 40 cm (1) and 100 cm (2).
0
100
200
300
400
500
600
0 100 200 300 400
AQI
Time(sec)
5µL 10µL 50µL
0
50
100
150
200
250
300
350
400
450
0 100 200 300 400
AQI
Time(sec)
5µL 10µL 50µL
0
500
1000
1500
2000
2500
050 100 150 200 250 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
1000
2000
3000
4000
5000
050 100 150 200 250 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
1400
1600
1800
050 100 150 200 250 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
1400
050 100 150 200 250 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
188
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
a-1 a-2
b-1 b-2
c-1 c-2
Fig. 9: Benzene (C6H6) response for ENS160 at 40 cm (1) and 100 cm (2)
0
50
100
150
200
250
300
350
400
450
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
50
100
150
200
250
300
350
400
450
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
1400
1600
1800
2000
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
1400
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
189
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
a-1 a-2
b-1 b-2
c-1 c-2
Fig. 10: Dichloromethane (CH2Cl2) response for ENS160 (a, b, c) at 40 cm (1) and
100 cm (2)
0
50
100
150
200
250
300
350
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
50
100
150
200
250
300
350
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
50
100
150
200
250
300
350
400
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
50
100
150
200
250
300
350
400
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
100
200
300
400
500
600
700
800
900
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
100
200
300
400
500
600
700
800
900
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
190
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
a-1 a-2
b-1 b-2
c-1 c-2
Fig. 11: Diethyl Ether (C4H10O) response for ENS160 (a, b, c) at 40 cm (a) and 100 cm (b),
0
100
200
300
400
500
600
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
100
200
300
400
500
600
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
5000
10000
15000
20000
25000
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
5000
10000
15000
20000
25000
30000
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
2000
4000
6000
8000
10000
12000
14000
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
191
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
a-1 a-2
b-1 b-2
c-1 c-2
d-1 d-2
Fig. 12: Ethanol (C2H6O) response for ENS160 (a, b, c) and TED110 (d) at 40 cm (1) and
100 cm (2)
0
100
200
300
400
500
600
0 100 200 300 400
AQI
Time(sec)
5µL 10µL 50µL
0
100
200
300
400
500
600
0 100 200 300 400
AQI
Time(sec)
5µL 10µL 50µL
0
2000
4000
6000
8000
10000
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
2000
4000
6000
8000
10000
12000
14000
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
1000
2000
3000
4000
5000
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
500
1000
1500
2000
2500
3000
3500
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
0 100 200 300 400
Gas density (PPM)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
0 100 200 300 400
Gas density (PPM)
Time(sec)
5µL 10µL 50µL
192
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
a-1 a-2
b-1 b-2
c-1 c-2
d-1 d-2
Fig. 13: Formic Acid (CH2O2) response for ENS160 (a, b, c) and TED110 (d) at 40 cm (1)
and 100 cm (2)
0
50
100
150
200
250
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
50
100
150
200
250
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
20
40
60
80
100
120
140
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
50
100
150
200
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
100
200
300
400
500
600
700
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
100
200
300
400
500
600
700
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
0 100 200 300
Gas density (PPM)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
0 100 200 300
Gas density (PPM)
Time(sec)
5µL 10µL 50µL
193
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
a-1 a-2
b-1 b-2
c-1 c-2
Fig. 14: Heptane (C7H16) response for ENS160 (a, b, c) at 40 cm (1) and 100 cm (2)
ENS160 hexane (C6H14) has an excellent response from >50μL for 40 cm for all the
required parameters. The AQI, TVOC, and CO2 responded well with >50μL at 40 cm with a
low response at 100 cm. In contrast to the MOX sensor principle, ENS160 can detect hexane
but shows low sensitivity, effectiveness, and detection level. Fig. 15 depicts the hexane
responses for ENS160. The TED110 gas sensor (gas density) hexane has not responded in any
amount.
0
100
200
300
400
500
600
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
50
100
150
200
250
300
350
400
450
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
500
1000
1500
2000
2500
3000
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
100
200
300
400
500
600
700
800
900
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
1400
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
194
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
a-1 a-2
b-1 b-2
c-1 c-2
Fig. 15: Hexane (C6H14) response for ENS160 (a, b, c) at 40 cm (1) & 100 cm (2).
For all the required parameters, isopropanol (C3H8O) shows an excellent response with
the ENS 160 gas sensor from the sample amounts at >5μL. The TED110 gas sensor (gas
density) at a 40 cm distance has an excellent response with isopropanol from >10μL. Based on
the MOX gas sensor concepts, the ENS160 and TED110 gas sensors effectively detect
isopropanol under varying sample amounts and distances. Further investigations specifically
for TED110 are essential to optimize sensor performance and understand the extent of their
accuracy and reliability in real-world environments. Fig. 16 displays the isopropanol sensor
responses for both sensors.
0
100
200
300
400
500
600
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
50
100
150
200
250
300
350
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
1000
2000
3000
4000
5000
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
100
200
300
400
500
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
1400
1600
1800
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
100
200
300
400
500
600
700
800
900
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
195
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
a-1 a-2
b-1 b-2
c-1 c-2
d-1 d-2
Fig. 16: Isopropanol (C3H8O) for ENS160 (a, b, c) and TED110 (Gas density) (d) at 40 cm
(1) and 100 cm (2)
0
100
200
300
400
500
600
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
100
200
300
400
500
600
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
2000
4000
6000
8000
10000
12000
14000
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
2000
4000
6000
8000
10000
12000
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
1000
2000
3000
4000
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
1000
2000
3000
4000
5000
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
0 100 200 300
Gas density (PPM)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
0 100 200 300
Gas density (PPM)
Time(sec)
5µL 10µL 50µL
196
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
a-1 a-2
b-1 b-2
c-1 c-2
d-1 d-2
Fig. 17: Methanol (CH3OH) response for ENS160 (a, b, c) and TED110 (d) at 40 cm (1) &
100 cm (2)
0
100
200
300
400
500
600
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
100
200
300
400
500
600
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
1000
2000
3000
4000
5000
6000
7000
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
2000
4000
6000
8000
10000
12000
14000
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
1000
2000
3000
4000
5000
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
500
1000
1500
2000
2500
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
0 100 200 300
Gas density (PPM)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
0 100 200 300 400
Gas density (PPM)
Time(sec)
5µL 10µL 50µL
197
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
a-1 a-2
b-1 b-2
c-1 c-2
Fig. 18: Toluene (C7H8) response for ENS160 (a, b, c) at 40 cm (1) & 100 cm (2)
0
50
100
150
200
250
300
350
400
450
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
100
200
300
400
500
600
0 100 200 300
AQI
Time(sec)
5µL 10µL 50µL
0
500
1000
1500
2000
2500
3000
3500
4000
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
1400
1600
0 100 200 300
TVOC(PPB)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
0
200
400
600
800
1000
1200
1400
1600
0 100 200 300
CO2 (PPM)
Time(sec)
5µL 10µL 50µL
198
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
(a) (b)
(c) (d)
Fig. 19: ENS160 maximum Response time (s) at 40 cm (a) and 100 cm (b), TED110
maximum Gas Density Response Time(s) at 40 cm (c) and 100 cm (d).
050 100 150
Acetone
Acetronitrile
Benzene
Dichloromethane
Diethyl Ether
Ethanol
Formic Acid
Heptane
Hexane
Isopropanol
Methanol
Toluene
ENS160 Response Time (s) at 40cm
50µL 10µL 5µL
020 40 60 80 100 120
Acetone
Acetronitrile
Benzene
Dichloromethane
Diethyl Ether
Ethanol
Formic Acid
Heptane
Hexane
Isopropanol
Methanol
Toluene
ENS160 Response Time (s) at 100cm
50µL 10µL 5µL
050 100 150
Acetone
Acetronitrile
Benzene
Dichloromethane
Diethyl Ether
Ethanol
Formic Acid
Heptane
Hexane
Isopropanol
Methanol
Toluene
TED110 Response Time (s) at 40cm
50 µL 10 µL 5 µL
050 100 150
Acetone
Acetronitrile
Benzene
Dichloromethane
Diethyl Ether
Ethanol
Formic Acid
Heptane
Hexane
Isopropanol
Methanol
Toluene
TED110 Response Time (s) at 100cm
50 µL 10 µL 5 µL
199
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
(a) (b)
(c) (d)
Fig. 20: ENS160 maximum Recovery time (s) at 40 cm (a) and 100 cm (b); TED110
maximum Gas Density Recovery Time(s) at 40 cm (c) and 100 cm (d).
The ENS160 shows an excellent response for methanol (CH3OH) with gas samples
amounting to>5μL. All sample compounds, such as AQI, TVOC, and CO2, show the highest
response in the desired parameters. The TED110 gas sensor (gas density) for ethanol (C2H6O)
at 40 cm and 100 cm for amounts at >5μL has a response. Both sensors' excellent response to
methanol based on the MOX sensor concept and significant performance among all sample
050 100 150 200 250 300
Acetone
Acetronitrile
Benzene
Dichloromethane
Diethyl Ether
Ethanol
Formic Acid
Heptane
Hexane
Isopropanol
Methanol
Toluene
ENS160 Recovery Time (s) at 40cm
50µL 10µL 5µL
0 100 200 300 400
Acetone
Acetronitrile
Benzene
Dichloromethane
Diethyl Ether
Ethanol
Formic Acid
Heptane
Hexane
Isopropanol
Methanol
Toluene
ENS160 Recovery Time (s) at 100cm
50µL 10µL 5µL
050 100 150 200 250 300
Acetone
Acetronitrile
Benzene
Dichloromethane
Diethyl Ether
Ethanol
Formic Acid
Heptane
Hexane
Isopropanol
Methanol
Toluene
TED110 Recovery Time (s) at 40cm
50 µL 10 µL 5 µL
050 100 150 200 250 300
Acetone
Acetronitrile
Benzene
Dichloromethane
Diethyl Ether
Ethanol
Formic Acid
Heptane
Hexane
Isopropanol
Methanol
Toluene
TED110 Recovery Time (s) at 100cm
50 µL 10 µL 5 µL
200
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
compounds make them valuable for gas detection applications. Fig. 17 depicts both sensor
responses for methanol.
Toluene (C7H8) exhibits an excellent response with a sample amount of 10μL for all
necessary parameters for ENS160, as illustrated in Fig. 18. With a quantity of >10μL, the AQI
shows an excellent response at 40 cm sensor node distances. The reaction rate of TVOC and
CO2 increases the high volumes of gases. Above all, ENS160 has a low detection level for
toluene due to no oxygen compounds, which need further investigation to improve the
accuracy. The TED110 gas sensor (Gas density) does not react to amounts of toluene at 40 cm
and 100 cm.
The response time (s) is the time it takes for the sensor output signal to reach 90% of its
highest measured value from its initial settled condition; some strong gas with more sensitivity
has quick response time, e.g., ethanol, methanol. Both sensors require a long response time (s)
for isopropanol to reach its maximum point compared to other samples' response times (s);
ethanol shows a rapid response (it requires less time to get its maximum response point). This
indicates ENS160 has high sensitivity and efficiency in detecting most vapors; has a quick
response for most of the strong gases. Improving response times can enhance gas sensors'
effectiveness and accuracy to provide real-time and accurate measurements in environments.
Fig. 19 presents both sensor's response times (s) for the different distances and specific amounts
of the samples.
Correspondingly, the recovery time (s) is when the sensor response signal returns to its
initial condition from its maximum measured value. Compared to other experimental gas
samples' recovery time (s), the ENS160 gas sensing of ethanol needs longer to recover to its
initial state. In contrast, methanol causes the longest recovery for the TED110 sensor. Gas
sensors can exhibit varying response and recovery times based on the sensitivity to the gases.
Some gases have rapid responses and longer recovery times based on their characteristics. By
improving recovery times, gas sensors can become more responsive and reliable, making them
better suited for safety monitoring in laboratory infrastructure. Fig. 20 displays both sensor's
recovery times (s) for the different distances and specific amounts of the samples.
The data overview for all tested materials that was acquired from the two sensors and their
response graphical presentation is listed below-
❑ ENS 160 successfully detects the AQI, TVOCs, and CO2 60 test out of 72 Tests.
❑ ENS 160 AQI, TVOCs, and CO2 detection rate is around 83%.
❑ TED110 successfully detects the Gas Density 24 test out of 72 Tests.
❑ TED110 overall detection rate is 33%.
The concentration of the exposed analytes directly relates to the change in sensor resistance.
On the surface of MOX, oxygen is adsorbed at high temperatures. The charge carrier
concentration changes due to the adsorbed oxygen capturing electrons from the conduction
band, which impacts the resistance of the MOX sensing Sensor layer. So, the ENS160 gas
sensor strongly reacts to oxygen containing compounds such as acetone (C3H6O), diethyl ether
(C4H10O), isopropanol (C3H8O), methanol (CH3OH), and ethanol (C2H6O): Only formic acid
(CH2O2) has no response. Its response is excellent in specific amounts and sensor node
distances; it maintains a good sensitivity, selectivity, and detection limit. The ENS 160 reaction
201
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
for the other compounds depends on the specific amounts of gas samples and sensor node
distances. The reaction changes as the sample concentration or sensor node distances vary; a
higher sample rate and a lower sensor node distance achieve a good response. The TED110
responds with the strongest gases, such as isopropanol (C3H8O), methanol (CH3OH), ethanol
(C2H6O), and formic acid (CH2O2), where all the gases contain oxygen. According to the
datasheet [34], the TED110 can detect methene and toluene. However, these gases don't contain
oxygen. However, in this experiment, the sensor failed to detect these gases. The reason for
this lack of detection was likely sensor contamination during the electronics assembly or
incorrect sensor purchase, e.g., some aging effect.
Table 5: Comparison of BME88, SGP 40, SGP 30, ENS 160, TED 110
Samples
BME88
SGP 40
SGP 30
ENS 160
TED 110
Acetone
—
Good response
≥ 10μL
Good response
≥ 10μL
Strong
response ≥
5μL
No response
Diethyl ether
—
Good response
≥ 50μL
Weak response
< 100μL
Strong
response ≥
5μL
No response
Isopropanol
—
Good response
≥ 10μL
Weak response
< 100μL
Strong
response ≥
5μL
Strong
response ≥
5μL
Methanol
—
Good response
≥ 2μL
Good response
≥ 100μL
Strong
response≥ 5μL
Strong
response ≥
5μL
Toluene
—
Good response
≥ 100μL
Good response
≥ 100μL
Strong
response ≥
50μL
No response
Ethanol
Strong
response≥ 10μL
Good response
≥ 2μL
Good response
≥ 5μL
Strong
response ≥
5μL
Strong
response ≥
5μL
Hexane
Weak response
Weak response
< 100μL
Weak response
< 100μL
Strong
response ≥
10μL
No response
Acetonitrile
Weak response
Weak response
< 100μL
Good response
≥ 10μL
Good response
≥ 5μL
No response
Benzene
-
Good response
≥ 100μL
Good response
≥ 100μL
Good response
> 5μL
No response
Dichloromethane
Weak response
Good response
≥ 100μL
Good response
≥ 100μL
Weak
response <
50μL
No response
Formic Acid
Strong
response≥ 10μL
Good response
≥ 2μL
Good response
≥ 2μL
Weak
response <
50μL
Strong
response ≥
5μL
Heptane
-
Good response
≥ 5μL
Weak response
< 100μL
Weak
response <
50μL
No response
In [20], Neubert et al. used the BME 688 and SPG30 gas sensors in their project. Those
sensors were tested with various TVOCs (ethanol, formic acid, acetonitrile, dichloromethane,
and hexane). The studies employed two distinct heights, 25 and 40 cm, and four different
202
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
quantities of each component. Dichloromethane, acetonitrile, and hexane were chosen in
quantities of 1, 5, 10, and 20mL, respectively, while ethanol and formic acid were selected in
quantities of 10, 100, 500, and 1000μL. Both sensors produced excellent results with ethanol
and formic acid but had insufficient responses to acetonitrile, dichloromethane, and hexane.
Furthermore, in [7], Al-Okby et al. tested SPG40 and SPG30 with 12 samples of VOC.
This experiment tested VOC in two locations, one directly 1 m below the sensor. The sensor
node was moved one meter horizontally from the bottom for the second position. The volume
was raised according to the sensor's reaction. The quantities used were 2μL, 5μL, 10μL, 50μL,
and 100μL. The test volume was not increased once the lowest detectable volume for a
particular position and distance was determined to avoid sensory overload. The evolution of
the sensor based on TVOC (ppm) and AQI is the SPG30 that has an inadequate response for
diethyl ether, isopropanol, hexane, and heptane. All other VOCs have a good reaction with the
SPG30 gas sensor. Besides, the SPG 40 has a weak hexane and acetonitrile response among
the 12 VOC samples. This project used the ENS 160 sensor, showing a weak response signal
in dichloromethane and formic acid. All other gas samples with this sensor have good
responses. Finally, the TED110 had an excellent reaction for isopropanol, ethanol, methanol,
and formic acid. Almost all other eight gas samples did not respond. The gas sensor comparison
table from the previous project and the current experiment are shown in Table 5.
The comparison results with other sensors (BME688, SGP 40, and SGP 30) show that the
ENS 160 has an excellent response, making it suitable for laboratory gas detection. On the
other hand, the TED110 shows inadequate response compared to all other gas sensors, possibly
due to contamination or incorrect sensor selection. Without additional testing and
improvement, this sensor may not be appropriate for the project's extension. Since this
experiment was conducted in a real laboratory, measures were taken to avoid destructive
factors such as traditional chemical hood air drafts, air conditioning, opening doors, and human
presence, which can impact measurements due to cosmetics and human body exudation. Above
all, for further accuracy improvement and avoiding external factors, different calibration
methods can be utilized, for instance, reference measurements, and dynamic calibration, to
adapt to changing conditions and minimize the impact of external factors like air drafts and
human presence. Additionally, implementing signal processing techniques, such as noise
filtering and pattern recognition algorithms, can improve the system's performance by
accurately isolating gas-specific signals from background noise. Regular maintenance and
sensor calibration is crucial to maintaining optimal performance and sensitivity. Lastly,
incorporating machine learning algorithms and sensor fusion techniques can enhance the
system's effectiveness and gas detectable limit by intelligently analyzing and combining data
from multiple sensors to improve accuracy and reliability.
4. CONCLUSION
This project aims to implement a mobile gas sensing system to detect hazardous and toxic
gases/chemical vapors. We investigated the performance of two novel gas sensors, ENS160
and TED110, using multiple parameters (AQI, TVOC, CO2, and Gas Density). In the future,
the plan is to extend this project into gas detection with alarming systems; for that reason, it
also helps to find efficient parameters among all parameters (AQI, TVOC, CO2, and Gas
Density). Overall, from the data visualization and the data analysis, both sensors have shown
that they are generally suitable for detecting VOC leakages in laboratories. The ENS 160 sensor
203
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
has a low sensing rate for dichloromethane and formic acid. All other experimented samples
responded very well; it has a detection rate of 60 out of 72 samples; in the three parameters,
such as AQI, TVOCs, and CO2, it has an 83% response rate. In contrast, the TED110 has an
excellent reaction for isopropanol, ethanol, methanol, and formic acid (the other eight
experimental samples have no response), responded to 24 out of 72 tests, and the detection rate
was 33%. Therefore, both sensors must do more tests with higher sample amounts. This system
can be adapted to a flexible IoT platform; the required modules, such as IoT wireless
communication modules and portable power supplies, can be connected and used
independently by connecting to a computer. The drawback of this developed system is that the
sensor node is comparably bigger than the divided processing units, resulting in the system
consuming more energy. In terms of measured parameters and functional qualities, this mobile
gas sensing system can be adapted to various application scenarios, for example, moving
objects such as robots and trolleys. The TED110 gas sensor requires more investigation to
improve its accuracy, sensor data calibration, and more accurate data conversion. Furthermore,
machine learning applications can distinguish different VOCs to precisely identify the natural
hazard and sensor calibration, as well as real-time data analysis and visualization. This mobile
sensing can be used in laboratory robots or moveable equipment so that in the future, the indoor
localization sensor can be implemented to record the position of detection of the robots or
movable objects, for example, roller carriages with laboratory equipment, which need to be
monitored for gases and location.
ACKNOWLEDGEMENT
This research was supported by the Synergy Project ADAM (Autonomous Discovery of
Advanced Materials) funded by the European Research Council (grant number 856405).
REFERENCES
[1] O. US EPA. (2016). Technical Air Pollution Resources. Retrieved from
https://www.epa.gov/technical-air-pollution-resources (Accessed: October 23, 2021).
[2] Pandey, P., & Yadav, R. (2018). A Review on Volatile Organic Compounds (VOCs) as
Environmental Pollutants: Fate and Distribution. Int. J. Plant Environ., 4. doi:
10.18811/ijpen.v4i02.2.
[3] Ghorani-Azam, A., Riahi-Zanjani, B., & Balali-Mood, M. (2016). Effects of air pollution on
human health and practical measures for prevention in Iran. J. Res. Med. Sci. Off. J. Isfahan
Univ. Med. Sci., 21, 65. doi: 10.4103/1735-1995.189646.
[4] Hou, L., Li, Y., Qian, X., Shu, C.-M., Yuan, M., & Duanmu, W. (2021). Large-scale
experimental investigation of the effects of gas explosions in underdrains. J. Saf. Sci. Resil.,
2(2), 90–99. doi: 10.1016/j.jnlssr.2021.03.001.
[5] Al-Okby, M. F. R., Neubert, S., Roddelkopf, T., & Thurow, K. (2021). Mobile Detection and
Alarming Systems for Hazardous Gases and Volatile Chemicals in Laboratories and Industrial
Locations. Sensors, 21(23). doi: 10.3390/s21238128.
[6] Al-Okby, M. F. R., Roddelkopf, T., Fleischer, H., & Thurow, K. (2022). Evaluating a Novel Gas
Sensor for Ambient Monitoring in Automated Life Science Laboratories. Sensors, 22(21). doi:
10.3390/s22218161.
[7] Al-Okby, M. F. R., Neubert, S., Roddelkopf, T., Fleischer, H., & Thurow, K. (2022). Evaluating
of IAQ-Index and TVOC Parameter-Based Sensors for Hazardous Gases Detection and
Alarming Systems. Sensors, 22(4). doi: 10.3390/s22041473.
204
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
[8] Ten health issues WHO will tackle this year. https://www.who.int/news-room/spotlight/ten-
threats-to-global-health-in-2019 (Accessed: October 23, 2021).
[9] Hunter, G. W., et al. (2020). Editors’ Choice - Critical Review - A Critical Review of Solid
State Gas Sensors. J. Electrochem. Soc., 167(3). doi: 10.1149/1945-7111/ab729c.
[10] Liu, X., Cheng, S., Liu, H., Hu, S., Zhang, D., & Ning, H. (2012). A Survey on Gas Sensing
Technology. Sensors, 12(7), 9635–9665. doi: 10.3390/s120709635.
[11] Xu, L., et al. (2022, July). Hybrid Gas Sensor Array to Identify and Quantify Low-Concentration
VOCs Mixtures Commonly Found in Chemical Industrial Parks. IEEE Sens. J., 22(13), 13434–
13441. doi: 10.1109/JSEN.2022.3176049.
[12] Examples of Common Laboratory Chemicals and their Hazard Class.
https://orf.od.nih.gov/EnvironmentalProtection/WasteDisposal/Pages/Examples+of+Common+
Laboratory+ChemicalsandtheirHazardClass.aspx (Accessed: July 09, 2022).
[13] Wyatt, P. (2021). Long-Term Effects of Severe Carbon Monoxide Poisoning. Wyatt Law Firm,
PLLC. https://www.wyattlawfirm.com/long-term-effects-of-severe-carbon-monoxide-
poisoning/ (Accessed: July 06, 2022).
[14] Chrisodgen. (2019). CO2 affects human health at lower levels than previously thought.
AirQualityNews. https://airqualitynews.com/2019/07/10/co2-affects-human-health-at-lower-
levels-than-previously-thought/ (Accessed: July 06, 2022).
[15] Shiwen Huang and Haomin Li and Mingrui Wang and Yaoyao Qian and Kyle Steenland and
William Michael Caudle and Yang Liu and Jeremy Sarnat and Stefania Papatheodorou and
Liuhua Shi (2021). Long-term exposure to nitrogen dioxide and mortality: A systematic review
and meta-analysis. Sci. Total Environ., 776, 145968. doi: 10.1016/j.scitotenv.2021.145968.
[16] Cheng, Shumin and Chang-Chien, Guo-Ping and Huang, Qianli and Zhang, Yu-Bo and Yan,
Ping and Zhang, Jiale and Wang, Yujing and Zhang, Daqing and Teng, Guopeng (2019). Global
Research Trends in Health Effects of Volatile Organic Compounds during the Last 16 Years: A
Bibliometric Analysis. Aerosol Air Qual. Res., 19(8), 1834–1843. doi:
10.4209/aaqr.2019.06.0327.
[17] Clifton, J. (2017, November 03). What are the Uses of Acetone? | The Chemistry Blog. ReAgent
Chemicals. https://www.chemicals.co.uk/blog/what-are-the-uses-of-acetone (Accessed: July 06,
2022).
[18] Liu, X., Cheng, S., Liu, H., Hu, S., Zhang, D., & Ning, H. (2012). A Survey on Gas Sensing
Technology. Sensors, 12(7). doi: 10.3390/s120709635.
[19] Dhall, S., Mehta, B. R., Tyagi, A. K., & Sood, K. (2021). A review on environmental gas sensors:
Materials and technologies. Sens. Int., 2, 100116. doi: 10.1016/j.sintl.2021.100116.
[20] Neubert, S., Roddelkopf, T., Al-Okby, M. F. R., Junginger, S., & Thurow, K. (2021). Flexible
IoT Gas Sensor Node for Automated Life Science Environments Using Stationary and Mobile
Robots. Sensors, 21(21). doi: 10.3390/s21217347.
[21] Al-Okby, M. F. R., Neubert, S., Roddelkopf, T., & Thurow, K. (2021). Integration and Testing
of Novel MOX Gas Sensors for IoT-based Indoor Air Quality Monitoring. In 2021 IEEE 21st
International Symposium on Computational Intelligence and Informatics (CINTI), pp. 000173–
000180. doi: 10.1109/CINTI53070.2021.9668462.
[22] Al-Okby, M. F. R., Roddelkopf, T., Fleischer, H., & Thurow, K. (2022). Robot-based
Environmental Monitoring in Automated Life Science Laboratories. In 2022 IEEE 10th Jubilee
International Conference on Computational Cybernetics and Cyber-Medical Systems (ICCC),
pp. 000395–000400. doi: 10.1109/ICCC202255925.2022.9922865.
205
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
[23] Al-Okby, M. F. R., & Al-Barrak, S. S. (2020). New Approach for Fall Detection System Using
Embedded Technology. In 2020 IEEE 24th International Conference on Intelligent Engineering
Systems (INES), pp. 209–214. doi: 10.1109/INES49302.2020.9147170.
[24] Gupta, K., Krishna, G. G., & Anjali, T. (2020). An IoT based System for Domestic Air Quality
Monitoring and Cooking Gas Leak Detection for a Safer Home. In 2020 International
Conference on Communication and Signal Processing (ICCSP), pp. 0705–0709. doi:
10.1109/ICCSP48568.2020.9182051.
[25] Das, T., Sut, D. J., Gupta, V., Gohain, L., Kakoty, P., & Kakoty, N. M. (2020). A Mobile Robot
for Hazardous Gas Sensing. In 2020 International Conference on Computational Performance
Evaluation (ComPE), pp. 062–066. doi: 10.1109/ComPE49325.2020.9200082.
[26] Joshna, V., Kashyap, M., Ananya, V., & Manitha, P. V. (2019). Fully Autonomous Robot to
Detect and Degasify Hazardous Gas after Flood Disaster. In 2019 2nd International Conference
on Power and Embedded Drive Control (ICPEDC), pp. 134–139. doi:
10.1109/ICPEDC47771.2019.9036703.
[27] Amin, A. B., Patel, H. P., Vaghela, S. P., & Patel, R. R. (2019). IOT Based Vehicle Anti-
Collision And Pollution Control System. In 2019 3rd International conference on Electronics,
Communication and Aerospace Technology (ICECA), pp. 108–110. doi:
10.1109/ICECA.2019.8822059.
[28] Waworundeng, J. M. S., Kalalo, M. A. T., & Lokollo, D. P. Y. (2020). A Prototype of Indoor
Hazard Detection System using Sensors and IoT. In 2020 2nd International Conference on
Cybernetics and Intelligent System (ICORIS), pp. 1–6. doi:
10.1109/ICORIS50180.2020.9320809.
[29] Wojnowski, W., Majchrzak, T., Dymerski, T., Gębicki, J., & Namieśnik, J. (2017). Portable
Electronic Nose Based on Electrochemical Sensors for Food Quality Assessment. Sensors,
17(12), 2715. doi: 10.3390/s17122715.
[30] Somov, A., Baranov, A., & Spirjakin, D. (2014). A wireless sensor–actuator system for
hazardous gases detection and control. Sens. Actuators Phys., 210, 157–164. doi:
10.1016/j.sna.2014.02.025.
[31] Esfahani, S., Tiele, A., Agbroko, S. O., & Covington, J. A. (2020, December). Development of
a Tuneable NDIR Optical Electronic Nose. Sensors, 20(23), 6875. doi: 10.3390/s20236875.
[32] Gong, S., et al. (2014, October). Design and Experiment of Fumigant Gas Trace Detection
System Based on Photoionization. Adv. Mater. Res., 1037, 129–133. doi:
10.4028/www.scientific.net/AMR.1037.129.
[33] SC-001224-DS-7-ENS160-Datasheet.pdf. https://www.sciosense.com/wp-
content/uploads/documents/SC-001224-DS-7-ENS160-Datasheet.pdf (Accessed: October 15,
2022).
[34] TED1X-MEMS-Gas-Sensor-data-1909.pdf. https://www.eoc-inc.com/wp-
content/uploads/TED1X-MEMS-Gas-Sensor-data-1909.pdf (Accessed: July 17, 2023).
[35] EPAs List of Volatile Organic Compounds in order of toxicity. (2020, May 18). Define
Instruments. https://defineinstruments.com/blog/epas-list-of-volatile-organic-compounds-in-
order-of-toxicity/ (Accessed: July 15, 2023).
[36] 6 Hidden Toxins Around Your Home That Can Cause Cancer. (2018).
https://learn.kaiterra.com/en/air-academy/6-hidden-toxins-that-can-cause-cancer (Accessed:
July 15, 2023).
[37] O. US EPA. (2014). Volatile Organic Compounds’ Impact on Indoor Air Quality.
https://www.epa.gov/indoor-air-quality-iaq/volatile-organic-compounds-impact-indoor-air-
quality (Accessed: July 15, 2023).
206
IIUM Engineering Journal, Vol. 25, No. 1, 2024 Sarif et al.
https://doi.org/10.31436/iiumej.v25i1.2727
[38] Spinelle, L., Gerboles, M., Kok, G., Persijn, S., & Sauerwald, T. (2017). Review of Portable and
Low-Cost Sensors for the Ambient Air Monitoring of Benzene and Other Volatile Organic
Compounds. Sensors, 17(7). doi: 10.3390/s17071520.
Table 6: Table of nomenclature
Symbol/Abbreviation
Description
MOX
metal oxide semiconductor
VOC
volatile organic compounds
TVOC
total volatile organic compounds
AQI
air quality index
PPM
Parts per Million
PPB
Parts per billion
MEMS
Micro-Electro-Mechanical System
MCU
Microcontroller unit
PCB
Printed circuit board.
I2C
Inter-Integrated Circuit
IDE
Integrated Development Environment
207
