Abstract—Airborne pollution and explosive gases threaten
human health and occupational safety, therefore generating high
demand for a wearable autonomous multi-analyte gas sensor
system for real-time environmental monitoring. This paper
presents a system level solution through synergistic integration
of sensors, electronics, and data analysis algorithms.
Electrochemical sensors featuring ionic liquids were chosen to
provide low-power room-temperature operation, rapid response,
high sensitivity, good selectivity, and a long operating life with
low maintenance. The system
electrochemical instrumentation circuit that combines all signal
condition functions within a single microelectronics chip to
minimize system cost, size and power consumption. Embedded
sensor array signal processing algorithms enable gas
classification and concentration estimation within a real-world
mixture of analytes. System
methodologies are described, and preliminary results are shown
for a first generation SO2 sensor and a thumb-drive sized
utilizes a multi-mode
design and integration
xposure to air pollution consistently ranks among the
leading global causes of illness and mortality , and
explosive gases are an increasing threat to occupational
safety as energy demands rise. Airborne pollutants and
explosive gases vary in both time and space. For example, CO
and CH4 can be released from boilers and stoves in homes, and
dangerous levels of CH4 and SO2 can be found in underground
coal mines . To improve scientific understanding of the
health impacts of personal exposure to these pollutants, a
miniaturized monitoring device is needed that individuals can
wear or carry to constantly examine their surrounding
environment. To be the most effective, the monitoring device
must be cost-effective, allow real-time data collection, and
operate autonomously with no user training or regular
maintenance. Toward this goal, significant energy has been
devoted to the research and development of gas sensors. For
example, an autonomous gas measurement system with
infrared gas sensors and wireless transmission has been
reported , but its handheld size and other restrictions make
it unsuitable for long-term individual use. Also, a single-chip
gas recognition system has been reported , but this device
can not measure gas concentrations and lacks many elements
of a complete system. Finally, a real-time gas sensor array test
system was presented in , but it is neither wearable nor
autonomous. In summary, although there have been many
important advances in gas sensor technologies, development
of a wearable, autonomous multi-analyte gas sensor system
remains an open challenge.
This paper presents a new system designed to meet the
goals for personal exposure assessment and environmental
monitoring by integrating sensor arrays, electronics, and data
analysis algorithms into an autonomous microsystem.
Electrochemical sensors were chosen to provide rapid
response, high sensitivity, and good selectivity. The sensors
utilize ionic liquid interfaces for low-power room-temperature
operation with low maintenance requirements. CMOS
techniques is employed to enable an autonomous
measurement system with minimum cost, size and power
consumption. Sensor array signal processing algorithms
running within the system provide the gas classification and
concentration estimation necessary for the mixture of analytes
found in real-world environments. Integrating these
components into a thumb-drive sized package, the resulting
system supports continuous, real-time measurement of
hazardous gases within a wearable autonomous platform.
II. MICROSYSTEM DESIGN
Simultaneously achieving all of the desired features for a
personal exposure monitoring system introduces many
challenges. To be wearable, the system needs to be
miniaturized, light weight, battery powered and either store
data, transmit it wirelessly, or both. A thumb-drive size system
with power consumption less than 1mW would be suitable for
this goal. To be autonomous the system needs to perform
measurements intelligently and without the need of any
external equipment, software, etc. To measure target gases in
real time, the entire measurement path, from appearance of the
analyte to storage of measurement results, must be completed
within seconds. This section analyzes these goals and
constraints from a system-level perspective and shows how
they can be mapped to component-level decisions.
A. Gas sensor technologies
A wearable autonomous multi-gas sensor microsystem for
real-time measurement requires gas sensors to be small
(chip-scale), low cost, free of maintenance, highly sensitive,
highly specific, fast responding, and inherently low power.
The sensor choice must also consider the required
measurement techniques and instrumentation electronics,
which must adhere to the same requirements as the sensor.
The three types of gas sensors that are most commonly used
are metal oxide gas sensors (e.g. stannic oxide),
Wearable Autonomous Microsystem with Electrochemical Gas
Sensor Array for Real-Time Health and Safety Monitoring
Haitao Li, Xiaoyi Mu, Zhe Wang†, Xiaowen Liu, Min Guo†, Rong Jin, Xiangqun Zeng†, Andrew J. Mason
Electrical and Computer Engineering, Michigan State Univ., East Lansing, MI, USA
†Chemistry Department, Oakland University, Rochester, MI, USA
non-dispersive infrared (NDIR) sensors, and electrochemical
(EC) sensors [6, 7]. A comparison among these sensors is
summarized in Table I. Metal oxide sensors have poor
selectivity because all reducing gases in the atmosphere are
detected on its surface. NDIR sensors show good sensitivity
and good selectivity but necessitate a relatively complicated
optical system making it expensive and bulky and requiring
operation training. Furthermore, they cannot provide real-time
monitoring. Compared with the first two sensor types, EC
sensors have good selectivity, low power consumption, low
cost and wide dynamic range. In addition, they have also
proven to be effective for detecting relevant gases such as CH4,
CO, CO2, NO, NO2, SO2, H2, and O2 [8-10]. Although
traditional EC sensors suffer from limited specificity and
interference, these disadvantages can be overcome with a
multi-mode EC sensor array using novel materials and
structures. Therefore, EC sensors were chosen in our system
for multi-gas measurement.
B. EC gas sensor array
EC gas sensors require the use of an electrolyte, which is an
ionically conducting medium that transports charge within
electrochemical cells, contacts all electrodes effectively,
solubilizes the reactants and products for efficient mass
transport, and is chemically and physically stable under all
conditions of sensor operation. Traditional electrolytes are
classified as liquid electrolytes (aqueous and non-aqueous)
and solid electrolytes [11, 12]. Liquid electrolytes have high
electrical conductivity but suffer from solvent exhaustion and
require periodic electrolyte maintenance. The solid
electrolytes overcome this problem but suffer from low
Room-temperature ionic liquids (ILs) are novel gas sensing
at room temperature.
materials that combine the benefits of both solid and liquid
electrolytes. Furthermore, the use of ILs as electrolytes can
eliminate the need for a membrane to simplify sensor design.
Electrochemical oxidation of NH3, NO2 and SO2 and the
electrochemical reduction of O2 have been reported using
ionic liquids (ILs), and some of these IL-based gas sensors
have shown wide detection limits, high sensitivity and
excellent reproducibility [13, 14].
EC sensors can be operated in different modes. The
amperometric mode measures the current generated by
reaction of an analyte at an electrode at a fixed or variable
potential. The electrochemical impedance spectroscopy (EIS)
mode measures impedance changes in the double-layer
capacitance and the charge-transfer conductance. The
sensitivity of an amperometric gas sensor is proportional to its
electrochemically active surface and drift over the time. EIS
mode can monitor for changes in the active electrode area
independent of gas concentrations. Therefore, EIS permits
self-monitoring of the sensor’s stability and automated
calibration for drift mechanisms. EIS also provides an
orthogonal detection mode to amperometric mode.
A gas sensor’s selectivity is important to a multi-gas
measurement system. This characteristic of EC sensors is
often based on the value of the working electrode potential,
which depends on the electrolyte and working electrode
material. Therefore, by optimizing the combination of sensor
operating potentials, electrolytic medium (e.g. ILs) and
working electrode material, the best transducer interfaces for
the IL-based EC sensor array can be achieved for high
selectivity of mixed gases. Table II summarizes how the
characteristics of the sensor array affect the desired system
COMPARISON OF GAS SENSOR TECHNOLOGIES
Sensor Type Advantages Disadvantages
Metal oxide gas
gases adsorb or
combust with the
e.g. Pt or Pd
coated on a wire.
parameter stability; low
cost; wide range of
Poor selectivity; major
ambient temperature and
humidity; large power
oxygen; degrades as it is
absorption of IR
Immune to catalyst
Detects limited number
of gases due to fairly
generally used; more
activity of various
Good selectivity; low
low cost; wide
(0-100% LEL and
High maintenance if
liquid electrolytes are
used; suffering from
LEL = lower explosive limit, UEL=upper explosive limit, % is a volume
DESIRED SENSOR PERFORMANCE AND SENSOR ARRAY CHARACTERISTICS
Relevant Sensor Characteristics
IL is an exciting material for EC gas sensors and has
been show to be very sensitive to the gas analytes at
ppb or ppm level detection limits.
Two dimensional IL amperometric and EIS sensor
array with sensor array pattern recognition increases
accuracy of detection.
Both amperometric and EIS sensors permit rapid
(near real time) measurement.
Long life time
ILs are chemically stable in normal atmosphere.
Their thermostability allows the detection interface
to be regenerated by heat.
Absence of moving parts and semi-solid state sensor
design. Complementary transducer types and can
cross verify each other.
Amperometric and EIS are low cost and low power
transducers readily miniaturized using low-cost
batch-fabrication microsystem technologies.
The transducers require no power and can be
measured using low
electronics compatible with portable and long-life
C. Instrumentation circuits and sensor data processing
A sensor array instrumentation circuit is needed to capture
and digitize the sensor signal for storage, transmission or
further signal processing. The system-level goals place the
following requirements on theinstrumentation circuit: low
cost, small size, low power consumption, rapid response.
Choice of an electrochemical sensor array adds requirements
to provide programmable bias potentials (for electrochemical
selectivity) and stimulus signals generators supporting
amperometric and EIS methods. Our group has prior
experience developing CMOS circuits for electrochemical
amperometry and EIS [15, 16]. Such circuits show that a
highly miniaturized EC sensor array with embedded
instrumentation circuitry is feasible.
To maximize the selectivity of an EC gas sensor array, an
array signal processing algorithm can be employed. Firstly, a
feature extraction method, such as Principal Component
Analysis (PCA) or Discriminate Function Analysis (DCA) can
be used to extract the features from multiple measurements
obtained by the sensor array. Given the extracted features, a
classification algorithm can then be applied to classify the
presence of certain chemicals or gases. Most of the existing
classification algorithms cannot predict concentrations of
multiple gases. A new data processing algorithm capable of
running within the wearable gas sensor system must be
developed to overcome this challenge while maintaining the
power budget of the overall system.
III. MCIROSYSTEM IMPLEMENTATION
To achieve all of the design goals outlined in section II, we
are developing a compact, multi-analyte, monitoring unit
called the intelligent electrochemical gas analysis system, or
iEGAS. As shown in Fig. 1, the iEGAS system combines a
sensor board and a microcontroller board. The sensor board
includes an EC sensor array and its instrumentation circuit. A
multi-mode EC instrumentation chip (MEIC) provides
stimulus signals for EC sensor array and measures sensor
response. Temperature and humidity sensors are included to
compensate for secondary sensitivities of the gas sensor array.
The microcontroller board provides all hardware necessary to
control the system, analyze data and generate alert signals to
warn users of adverse conditions. The iEGAS system can
communicate with a computer through a USB port or with a
mobile device through a plug-in wireless communication
module. The package enclosure houses an air inlet, a tiny fan,
and particle filters to control airflow across the gas sensor
array. The iEGAS system is powered by a rechargeable
battery which can be recharged through the USB port.
The proposed MEIC CMOS instrumentation chip includes
two main blocks: signal generators and a multi-mode readout
circuit. Fig. 2 shows the functional block diagram of the
MEIC and its connection to other iEGAS components. To
generate EC stimulus signals, an AC sinusoid generator and
DC waveform generator are provided for impedance and
amperometry measurements, respectively. The on-chip
potentiostat array provides eight-channel variable sensor
operating potentials. The multi-mode readout circuit amplifies
and records sensor responses and performs analog to digital
conversion with very low power and compact size. The MEIC
also includes a digital control and communication block and
on-chip memory in order to receive commands from and
deliver digital data to the iEGAS microcontroller.
IV. DESIGN RESULTS
An IL-based EC sensor was built to verify selected the
sensor technolgy. POREX® porous PTFE was used as the
permeable membrane for its excellent inertness to electrolyte.
Gold electrodes were deposited on the PTFE in an
interdigitated manner to maximize the reaction current.
High-purity room-temperature ionic liquid [C4mpy][NTf2]
was used as the electrolyte. The sensor structure and electrode
geometry are shown in Fig. 3. EIS measurements were
performed for SO2 as an example pollutant. The impedance
response curve in Fig. 4 shows that Δ|Z| increases with SO2
concentration. In future work, the sensor structure in Fig. 3
Fig. 1 Illustration of iEGAS gas monitoring device, A roughly
thumb-drive sized autonomous sensing system with measurement,
analysis, alert and communication capabilities.
Fig. 2 Block diagram of MEIC instrumentation chip for sensor arrays.
Fig. 3. (a) EC sensor structure; (b) microfabricated electrode geometry.
The electrodes and pads size is 0.75” × 0.5”.
will be expanded to an array suitable for a wide range of Download full-text
A prototype iEGAS system has been constructed and is
shown in Fig. 5. The microcontroller board houses an ultra
low power MSP430 (Texas Instruments), memory for data
collection, and a battery interface. Humidity and temperature
sensors were implemented within the sensor board. The fan
and filter were mounted to the system package. This
first-generation system is fully operational, able to read sensor
data and upload it in real time to a PC for display and storage.
The IL gas sensor array and MEIC discussed above will be
incorporated into the next generation system.
The development of a wearable, low-cost, low-power,
autonomous, real-time, multi-analyte gas sensor array system,
from concept to prototype has been reported. Comparison of
gas sensor techniques led to the choice of an electrochemical
sensor featuring ILs to provide small size, high sensitivity,
good selectivity, rapid response, low power consumption and
low cost. A prototype ILs-based EC sensor was built to
demonstrate this chosen design concept. To provide
low-power, small-size instrumentation for the electrochemical
sensors, a CMOS circuit was described that includes multiple
EC stimulus techniques, potentiostat biasing, current readout,
and A/D conversion. Utilization of signal processing
techniques within the miniaturized wearable system was also
discussed for optimizing selectivity of the gas sensor array in
the presence of a real-world mixed gas environment. An
operational first generation prototype integrating the
system-level concepts was described and shown. All the
components necessary for monitoring hazardous gases were
fit into a thumb-drive package with a USB port. Future
generations of the system are expected to provide a valuable
new tool for the assessment of personal exposure to
environmental and occupational hazards.
This work was supported in part by the National Institute
for Occupational Safety and Health (NIOSH) under Grant
 R. D. Brook, “Cardiovascular Effects of Air Pollution,” Clin Sci, vol.
115, pp. 175-87, 2008.
 J. W. Gardner, P. K. Guha, F. Udrea and etal., “CMOS Interfacing for
Integrated Gas Sensors: A Review,” IEEE Sensors J., vol. 10, no. 12,
pp. 1833-1848, Dec. 2010.
 S. Beirne, B. M. Kiernan, C. Fay and et al., “Autonomous Greenhouse
Gas Measurement System for Analysis of Gas Migration on Landfill
Sites,” Proc. of World Academy of Science, Eng. and Tech., 2008, pp.
 K. T. Ng, F. Boussaid, and A. Bermark, “A CMOS Single-Chip Gas
Recognition Circuit for Metal Oxide Gas Sensor Arrays,” IEEE Trans.
Circuits and Systems-I: Regular papers, vol. 58, no. 7, pp. 1569-1580,
 M. Li; X. Chen, J. Liu, “A Practical Testing System for Gas Sensor
Array,” Int. Conf. on Info. and Auto., Hefei, China, pp. 1812-1816,
 F. Rock, N. Barsan and U. Weimar, “Electronic Nose: Current Status
and Future Trends,” Chem. Rev., vol. 108, pp. 705-725, 2008.
 D. James, S. M. Scott, Z. Ali, and etal., “Chemical Sensors for
Electronic Nose Systems,” Microchimica Acta, vol. 149, pp. 1-17,
 R. Knake, P. Jacquinot, and P. C. Hauser, “Amperometric Detection of
Gaseous Formaldehydein the ppb Range,” Electroanalysis, vol. 13, pp.
 G. Schiavon, G. Zotti, G. Bontempelli, G. Farnia and etal.,
“Amperometric Monitoring of Ozone in Gaseous Media by Gold
Electrodes Supported on Ion Exchange Membranes (Solid Polymer
Electrolytes),” Anal. Chem., vol. 62, pp. 293-298, 1990.
 A. W. E. Hodgson, P. Jacquinot, and P. C. Hauser, “Electrochemical
Sensor for the Detection of SO2 in the Low-ppb Range,” Anal. Chem.,
vol. 71, pp. 2831-2837, 1999.
 S. S. Bhoga and K. Singh, “Electrochemical Solid State Gas Sensors:
an Overview,” Ionics, vol. 13, pp. 417-427, 2007.
 P. Jasinski, “Solid-state Electrochemical Gas Sensors,” Materials
Sci.-Poland, vol. 24, pp. 269-278, 2006.
 M. C. Buzzeo. “Kinetic Analysis of the Reaction between
Electrogenerated Superoxide and Carbon Dioxide in the Room
Temperature Ionic Liquids
bis(trifluoromethylsulfonyl)imide,” J. Phys. Chem. B, vol. 108, pp.
 R. Wang, T. Okajima, F. Kitamura, and etal., “A Novel Amperometric
O2 Gas Sensor Based on Supported Room-Temperature Ionic Liquid
Porous Polyethylene Membrane-Coated Electrodes,” Electroanalysis,
vol. 16, pp. 66-72, 2004.
 J. Zhang, Y. Huang, N. Trombly and et al., “Electrochemical Array
Microsystem with Integrated Potentiostat,” IEEE Sensors, vol. 1, no. 1,
pp. 381–385, 2005.
 C. Yang, S. R. Jadhav, R. M. Worden and et al., “Compact Low-Power
Impedance-to-Digital Converter for Sensor Array Microsystems,”
IEEE J. Solid State Circuits, vol. 44, pp. 2844-2855, Oct. 2009.
SO2 concentration (ppm)
Fig 4. EIS test results plot for different SO2 concentrations. The sensor
operation potential was set at -0.5V and the stimulation frequency was set
Fig 5. iEGAS prototype system(a) Outer view; (b) Inner view. Humidity
and temperature sensors were placed on the sensor board. An MSP430
microcontroller was placed on the MCU board to turn on/off the fan can
communicate to remote systems.