sensors

Article
A New Low-Temperature Electrochemical
Hydrocarbon and NOx Sensor
Praveen Kumar Sekhar 1, * ID
, Zachary Moore 1 , Shyam Aravamudhan 2 and Ajit Khosla 3
1 Nanomaterials and Sensors Laboratory, School of Engineering and Computer Science,
Washington State University Vancouver, Vancouver, WA 98686, USA; [email protected]
2 Joint School of Nanoscience and Nanoengineering, North Carolina A & T State University,
Greensboro, NC 27401, USA; [email protected]
3 Faculty of Engineering, Yamagata University, Yonezawa, Yamagata 992-8510, Japan; [email protected]
* Correspondence: [email protected]

Received: 22 October 2017; Accepted: 27 November 2017; Published: 29 November 2017

Abstract: In this article, a new investigation on a low-temperature electrochemical hydrocarbon

and NOx sensor is presented. Based on the mixed-potential-based sensing scheme, the sensor is
constructed using platinum and metal oxide electrodes, along with an Yttria-Stabilized Zirconia
(YSZ)/Strontium Titanate (SrTiO3 ) thin-film electrolyte. Unlike traditional mixed-potential sensors
which operate at higher temperatures (>400 ◦ C), this potentiometric sensor operates at 200 ◦ C with
dominant hydrocarbon (HC) and NOx response in the open-circuit and biased modes, respectively.
The possible low-temperature operation of the sensor is speculated to be primarily due to the
enhanced oxygen ion conductivity of the electrolyte, which may be attributed to the space charge
effect, epitaxial strain, and atomic reconstruction at the interface of the YSZ/STO thin film.
The response and recovery time for the NOx sensor are found to be 7 s and 8 s, respectively. The sensor
exhibited stable response even after 120 days of testing, with an 11.4% decrease in HC response and
a 3.3% decrease in NOx response.

Keywords: electrochemical; YSZ; STO; oxygen ion conductivity; NOx ; hydrocarbon

1. Introduction
Electrochemical gas sensors [1–6] offer a portable, selective, sensitive, rapid, and compact solution
to detect analytes for various applications including clean coal power plants, automotive emissions,
hydrogen safety, air quality, and breathe analysis to identify chronic disorders. Among the various
detection methodologies in electrochemical (EC) gas sensors, the mixed-potential-based scheme [7]
possesses some unique features such as (a) simple device architecture; (b) compactness, to fit into
critical areas; (c) fast response; and (d) direct voltage read-out, circumventing the need for additional
conditioning circuitry.
Mixed-potential electrochemical sensors fabricated using well-established commercial
manufacturing methods present a promising avenue to enable the widespread utilization of
gas sensors. These devices are fundamentally simple and robust, owing to their close relationship to the
ubiquitous automotive Lambda sensor. The mixed-potential sensors develop a non-Nernstian potential
due to differences in the redox kinetics of various gas species at each electrode/electrolyte gas interface
in the presence of oxygen [8–11]. The difference in the steady-state redox reaction rates between the
fast and slow electrodes gives rise to the measured mixed-potential response. A stable sensor design
incorporates dense electrodes and a porous electrolyte, using a high-temperature ceramic co-fire (HTCC)
approach [12]. The use of dense electrodes (a) minimizes deleterious heterogeneous catalysis (leading to
reduced sensitivity); and (b) increases the morphological stability, yielding a robust electrochemical
interface, thereby increasing lifetime durability. The downside of mixed-potential sensors is the high

Sensors 2017, 17, 2759; doi:10.3390/s17122759 www.mdpi.com/journal/sensors

Sensors 2017, 17, 2759 2 of 10

power consumption due to the high temperature of sensor operation. While high-temperature operation
is suitable for certain applications such as emissions control, the majority of the sensor applications
demand ultra-low power consumption. In addition, to realize low-power operation, the sensor should
be operable at low temperature without compromising on its performance. The high-temperature
operation is necessary mainly due to the presence of the well-known solid electrolyte Yttria-Stabilized
Zirconia (YSZ) in the sensor fabrication.
Yttria-Stabilized Zirconia (YSZ) is an oxygen ion conductor, where oxygen vacancies are
introduced by aliovalent substitution of Zr by Y. The substitution of Zr in a 4+ oxidation state with
Y3+ produces an oxygen vacancy for every two Y atoms. As mentioned earlier, the high-temperature
operation of YSZ is necessary for oxygen ion conductivity and to minimize the ohmic losses.
In recent years, approaches making use of confinement effects at the nanoscale to improve
device performance have emerged as an interesting alternative [13]. Decreasing the size of the
material to the nanometer-length scale may lead to a dominant influence of the interface on the
overall ionic conductivity of the system. In addition, heterostructures combining transition metal
oxides can accommodate very large amounts of epitaxial strain without breaking into islands or
structural domains, leading to enhanced ion diffusivity. A well-studied material combination is
Y2 O3 -ZrO2 /SrTiO3 (YSZ/STO) heterostructures where different structures (fluorite vs perovskite) are
combined with a large lattice mismatch of 7%. The interface between highly dissimilar structures
stabilizes a disordered oxygen sublattice with an increased number of oxygen vacancies, which,
in turn, may promote oxygen diffusion even at a low temperature [14]. In the past, this YSZ/STO
heterostructure has been used as a solid electrolyte in low-temperature fuel cell applications [15].
Inspired by the available literature, we have investigated the use of nanoscale heterostructures of
YSZ/STO in the design of a mixed-potential gas sensor. The goal of this article is to study the
heterostructure sensor response at reduced operating temperatures (around 200 ◦ C). In this work,
the low-temperature mixed-potential sensors were designed to detect hydrocarbons and NOx in the
open-circuit and biased modes, respectively. The electrode choice and the analyte selection were based
on previous studies [16].

2. Experimental
Based on the mixed-potential sensor design developed earlier [17], ESL ElectroScience developed
a High Temperature Co-fired Ceramic (HTCC) structure (both substrate tape and screen-printed
patterns for the various layers) using well-established manufacturing methods, along with fully
stabilized YSZ substrate tape. First, on the backside of the 1 mm thick, dense YSZ substrate tape,
a Pt-heater with independent leads was printed to form the core of the new ceramic substrate.
The Pt-heater was applied to a layer of insulating alumina tape to prevent ionic transport through the
sensor body. A standard heater ink composition was used, and the required heater resistance was
obtained by modifying the design of the heater pattern. An insulating overcoat was then applied to the
Pt-heater to protect and reduce the amount of ambient heat loss, in order to provide a more uniform
heat distribution across the substrate. Next, the sensing component, consisting of screen-printed
La0.8 Sr0.2 CrO3 (MO) and Pt electrodes, was printed on the opposite side of the heater. Two special
screens were used to print the electrodes. The MO and Pt electrode pads were made to a thickness of
500 nm so that the electrolyte completely covers the electrode and the leads were made to a thickness
of 1 µm.
Lastly, the YSZ/STO thin-film heterostructure was sputter deposited through a shadow mask
to cover and embed the MO and Pt electrodes. The YSZ and STO sputter targets (99.9% pure) were
obtained from Kurt J. Lesker. First, a 50 nm thin film of STO was deposited at a Radio Frequency (RF)
power of 150 W for 15 min, and then YSZ was deposited at a RF power of 200 W for 20 min to achieve
a thickness of 100 nm. Throughout the sputtering process, the substrate was placed on a heated plate at
175 ◦ C. The sputtering process was repeated to achieve 5 layers of STO (50 nm)/YSZ (100 nm). Finally,
the device structure was sintered at 1100 ◦ C in the presence of Argon. A control device with a plain
Sensors 2017, 17, 2759 3 of 10

YSZ thin film was fabricated to compare the performance of the sensors with and without the STO
multilayer. The plain YSZ thin film was obtained by sputtering for a thickness of 750 nm matching the
YSZ/STO multilayer.
The sintered thin films were then inspected using a Transmission Electron Microscope (TEM).
TEM cross-sections of the multilayer films were prepared using focused ion beam (FIB)-based milling
(FEI). A thin layer of Au was evaporated onto the multilayer film before FIB milling to protect the
region of interest and to prevent charging. Coarse FIB milling was done at 30 kV, and the samples
were then thinned to approximately 50 nm thickness at 2 kV. The samples were then examined in
an aberration-corrected TEM operated at 300 kV. In the presence of air (partial oxygen pressure
of 21%), the electrical conductivity of the YSZ/STO thin film was measured using AC impedance
measurement on the Potentiostat (Princeton Applied Research PARSTAT 4000 potentiostat/galvanostat).
The measurements were done in the impedance mode with a frequency sweep from 13 MHz down to
10 mHz and a perturbation of 10 mV at different temperatures. To compare the oxygen ion conductivity
surrogate from the YSZ/STO multilayer and control YSZ thin film, the impedance spectrum was
collected at 21% oxygen with a temperature reading of 200 ◦ C and 25 ppm of propylene. A Keithley
2400 Source Measure Unit (SMU) was used to record the open-circuit sensor response. The SMU was
connected to the fabricated sensor such that the Pt counter electrode was instrument positive and the
metal oxide electrode was instrument negative.
A specially prepared quartz tube was used to test the fabricated thin-film sensor under the test
gas stream using a method very similar to commercial automotive sensor testing. A commercial flat
plate O2 sensor mount, using a standard configuration of four metal spring-loaded clips, was used
to hold the fabricated sensor. A rubber stopper was used to hold the packaging in place and silicone
gasket sealant was used to seal the gaps around the four wires providing electrical contact to the sensor
electrodes and heater. The sensor was heated to 200 ◦ C with a heater voltage of 4.5 V and a current of
0.2 A (0.9 W).
To get an accurate temperature reading of the devices, a P-L-type precision RTD sensor from
Omega was placed in the tube furnace near the device. Air was used as the base gas. The flow
rates of the various gas mixtures were controlled using an automated Trace Vapor Generator (TVG).
The TVG was also used to generate 25 ppm of NO, NO2 , NH3 , C3 H6 , and CO. In this article, the fixed
concentration of the gases was chosen to test the concept of low-temperature operation of the sensors.
The low-temperature operation of sensors will enable low power consumption, which might be critical
for applications such as air quality monitoring. However, the concentration range of analytes for
such an application is in the order of parts per billion (ppb). These mixed-potential sensors will be
engineered in future to sense ppb levels of pollutants or analytes.
The flow rates were fixed at 500 sccm. Two terminal measurements were performed at a fixed
total gas flow rate and a fixed oxygen partial pressure of 21%. The rise time (t90 , rise) and fall time
(t90 , fall) are defined as the time required for the sensor output to reach 90% of the full-scale value
(maximum voltage) and as the time required for the sensor output to diminish to 10% of the full-scale
value, respectively. The mixed-potential sensor response to an analyte was calculated as the voltage
differential of the base gas (air) and the target/test gas.

3. Results and Discussion

Figure 1 shows the front, rear and cross-sectional image of the fabricated sensor. The embedded
MO (black) and Pt (gray) electrodes, along with the YSZ/STO electrolyte (whitish strip),
are clearly visible.
Figure 2 shows TEM micrographs of the YSZ/STO heterostructure. Figure 2a displays
a representative high-resolution TEM image of the YSZ thin film, STO thin film, and the interface
region showing the good crystalline quality of the sample aligned in the [001] and [110] direction.
The YSZ/STO interface was seen to be largely flat and continuous with some slight distortion in the
few atomic layers. The YSZ film showed good structural coherence with the STO crystal. From the
 Sensors 2017, 17, 2759 4 of 10

TEM observations, an in-plane interfacial strain of 7.1% in the interface was calculated using the lattice
constants of 3.65 Å and 3.91 Å for YSZ and STO layers respectively.
Figure 2b shows the low-magnification TEM image of the YSZ substrate, five layers of STO and
the YSZ thin film after sintering. The interface between the substrate and the thin film and the interface
Sensors 2017, 17, 2759 4 of 10
Sensors
between2017,YSZ
17, 2759 4 of 10
and STO thin films appear to be well-defined without any inter-diffusion after sintering.

Figure
Figure1.1. Mixed-potential sensing
Mixed-potentialsensing
1.Mixed-potential device
sensingdevice and
deviceand its
andits cross-section
itscross-section image.
cross-sectionimage. (Left)
image.(Left) Front
(Left)Front and
Frontand rear
rearofof
andrear the
ofthe sensor
thesensor
sensor
Figure
showing
showing the
theelectrodes
electrodes(Mo,
(Mo,Pt,Pt,
and YSZ/STO
and YSZ/STOelectrolyte) and the
electrolyte) andPtthe
heater;
Pt (Right)
heater; Cross-section
(Right) image
Cross-section
showing the electrodes (Mo, Pt, and YSZ/STO electrolyte) and the Pt heater; (Right) Cross-section image
showing the electrolyte
image showing embedding
the electrolyte the electrodes.
embedding the electrodes.
showing the electrolyte embedding the electrodes.

Figure 2.
2.TEMTEMmicrograph
micrograph of the device.
of the (A) High-resolution
device. TEM image
(A) High-resolution YSZ/STO
TEM image thin film,
YSZ/STO andfilm,
thin the
Figure 2. TEM micrograph of the device. (A) High-resolution TEM image YSZ/STO thin film, and the
interface indicating
and the interface good crystallinity
indicating of theofsample;
good crystallinity andand
the sample; (B)(B)
Low-magnification
Low-magnificationTEM TEMimage
image of
interface indicating good crystallinity of the sample; and (B) Low-magnification TEM image of
cross-section
cross-section showing
showing the YSZ substrate, five layers of STO thin
thin film
film (thickness
(thickness around
around 5050nm)/YSZ
nm)/YSZ
cross-section showing the YSZ substrate, five layers of STO thin film (thickness around 50 nm)/YSZ
thin film (thickness around 100 nm).
thin film (thickness around 100 nm).

The conductivityofofthethe
The conductivity
conductivity multilayer thin film was found to be thermally activated with an
The of themultilayer
multilayer thinthin
filmfilm
was was
found to be thermally
found activatedactivated
to be thermally with an Arrhenius
with an
Arrhenius
behavior. The behavior.
activationThe activation energy for the YSZ/STO films was calculated to be 0.76 eV eV
as
Arrhenius behavior. Theenergy for the
activation YSZ/STO
energy films
for the was calculated
YSZ/STO to becalculated
films was 0.76 eV as opposed
to be 0.76to 0.95
eV as
opposed
for plaintoto
YSZ 0.95 eV for plain YSZ film. The for
activation energy for oxygen ion conductivity was reduced
opposed 0.95film.
eV forThe activation
plain YSZ film.energy oxygenenergy
The activation ion conductivity
for oxygen was reduced for the
ion conductivity wasYSZ/STO
reduced
for
thin the YSZ/STO
film, similar thin
to the film, similar
observation to the
reported observation
earlier reported
[18]. Figure 3earlier
shows [18].
the Figure
impedance 3 shows the
spectrum
for the YSZ/STO thin film, similar to the observation reported earlier [18]. Figure 3 shows the
impedance spectrum
comparisonspectrum
of devices comparison of
with the YSZ/STO devices with
multilayer the YSZ/STO multilayer thin-film electrolyte and
impedance comparison of devices with thethin-film
YSZ/STOelectrolyte
multilayerand the plain
thin-film thin-filmand
electrolyte YSZ
the plain
electrolyte.thin-film YSZ
In an impedance electrolyte.
spectrum In an impedance spectrum obtained from an electrochemical
the plain thin-film YSZ electrolyte. In obtained
an impedance from an electrochemical
spectrum obtaineddevice,
from an theelectrochemical
contributions of
device, the contributions of electrode/electrolyte
the bulk electrolyte and the electrode/electrolyte interface are identified
device, the contributions of the bulk electrolyte and the electrode/electrolyte interface aredispersions.
the bulk electrolyte and the interface are identified by the frequency identified
by the frequency
In general, dispersions.
the low-frequency In
region general, the low-frequency region represents the electrode
by the frequency dispersions. In represents
general, the the electrode polarization/charge-transfer
low-frequency region represents the phenomenon
electrode
polarization/charge-transfer
while the high-frequency component phenomenon can while the high-frequency
be approximated component
to ionic conduction canofbethe
approximated
electrolyte.
polarization/charge-transfer phenomenon while the high-frequency component can be approximated
to ionic conduction of the electrolyte. The semi-circles at the high-frequency region indicate a lower
to ionic conduction of the electrolyte. The semi-circles at the high-frequency region indicate a lower
resistance to oxygen conduction for the YSZ/STO-film-based device as compared with that of the
resistance to oxygen conduction for the YSZ/STO-film-based device as compared with that of the
plain YSZ thin film.
plain YSZ thin film.
Sensors 2017, 17, 2759 5 of 10

The semi-circles at the high-frequency region indicate a lower resistance to oxygen conduction for the
Sensors 2017, 17, 2759 5 of 10
YSZ/STO-film-based device as compared with that of the plain YSZ thin film.

Figure 3.
Figure Comparison of
3. Comparison of impedance
impedance spectra obtained from two devices made of YSZ/STO multilayer
YSZ/STO multilayer
◦
thin-film electrolyte and plain YSZ thin film at 200 °C,
C, 21% O22,, and
and 25
25 ppm
ppm of
of C
C33H
H66. .
thin-film

The
The interface
interfacein inthe
theplanar
planaroxide
oxideheterostructure
heterostructure film
film(YSZ/STO)
(YSZ/STO) seems to act
seems as aas
to act fast oxygen
a fast ion
oxygen
conduction pathway. Oxygen vacancies and oxygen conduction arises
ion conduction pathway. Oxygen vacancies and oxygen conduction arises from the interface strain from the interface strain and
structural
and structuralincompatibility
incompatibility between
between YSZYSZ andand
STO.STO.The TheMO/YSZ-STO/gas
MO/YSZ-STO/gas acts
actsasasthe
the three-phase
three-phase
interface,
interface, otherwise known as the triple-phase boundary. Humidity was not added for thisexperiment.
otherwise known as the triple-phase boundary. Humidity was not added for this experiment.
Figure
Figure 44 shows
shows the open-circuit mixed-potential
the open-circuit mixed-potential response
response of of the
the sensors
sensors in in response
response to various
to various
analytes at 200 °C
◦ at an oxygen partial of 21%. The concentrations of all the
analytes at 200 C at an oxygen partial of 21%. The concentrations of all the analytes were maintained analytes were maintained
at
at 25
25 ppm.
ppm. TheThe sensor
sensor response
response of of the
the device
device fabricated
fabricated withwith just
just the
the YSZ
YSZ thinthin film
film was
was compared
compared withwith
the
the response of the device with the multi-layer YSZ/STO film at 200 C. The device with just the YSZ
response of the device with the multi-layer YSZ/STO film at 200 °C.
◦ The device with just the YSZ
thin
thin film
film did
did not
not show
show any discernible response
any discernible response to to the
the analytes
analytes with
with aa near-zero
near-zero baseline
baseline response.
response.
However,
However, the the sensor
sensor device
devicewith
withthetheYSZ/STO
YSZ/STOthin thinfilm
filmshowed
showed a dominant
a dominant propylene
propylene response.
response.
This sensor device responded to NO, NO 2, and NH3. A slightly stronger signal to ammonia is
This sensor device responded to NO, NO2 , and NH3 . A slightly stronger signal to ammonia is due
due to direct
to direct oxidation
oxidation of ammonia
of ammonia on YSZ/STO/Pt,
on YSZ/STO/Pt, where where
threethree
ammoniumammonium ions areionsformed
are formed for
for every
every four ammonia
four ammonia molecules,
molecules, in an overall
in an overall three-electron
three-electron electrochemical
electrochemical process. process. As expected,
As expected, the
the sensor
sensor
showed selective response to hydrocarbons. The sensor response and recovery time in response in
showed selective response to hydrocarbons. The sensor response and recovery time to
response
C3 H6 wastofoundC3H6 to wasbe found
65 andto 70be 65 and 70 s, respectively.
s, respectively. The supplemental The supplemental
Figure S1 shows Figure
the 1open-circuit
shows the
open-circuit
sensor response sensor responsetocomparison
comparison 25 ppm of Cto3 H25 ppm of C3H6 and a mixture of 25 ppm of C3H6, NO,
6 and a mixture of 25 ppm of C3 H6 , NO, NO2 , and NH3 .
NO 2, and NH3. A 9.5% decrease in sensor response to C3H6 was observed when exposed to a complex
A 9.5% decrease in sensor response to C3 H6 was observed when exposed to a complex mixture in
mixture in comparison to isolated exposure of C3H6. The supplemental Figure 2 shows the sensitivity
of the device as a function of different concentrations of C3H6.
Sensors 2017, 17, 2759 6 of 10

comparison to isolated exposure of C3 H6 . The supplemental Figure S2 shows the sensitivity of the
device as a 17,
Sensors 2017, function
2759 of different concentrations of C3 H6 . 6 of 10

Figure 4.
Figure 4. Open-circuit mixed-potential sensor response to various analytes (25 ppm) ppm) at 200 ◦°C,
at 200 C, 500 sccm,
and oxygen partial
and oxygen partial pressure pressure of 21%. The sensor made of YSZ/STO thin-film electrolyteshowed
21%. The sensor made of YSZ/STO thin-film electrolyte showeda
adominant
dominantCC 3H3 6Hresponse, whereas
6 response, thethe
whereas sensor made
sensor madeof just the the
of just YSZYSZthinthin
filmfilm
showed no response.
showed The
no response.
letterletter
The ‘u’ stands
‘u’ standsfor for
thethe
symbol ‘µ’.‘µ’.
symbol The baseline
The response
baseline responseis is0 0V.V. Humidity
Humidity was
was not added in in
this experiment.
this experiment.

AA stable
stable sensor
sensor response,
response, eveneven at at such
such aa lowlow temperature
temperature (200 (200 ◦°C), indicates aa strong
C), indicates strong oxygen
oxygen ionion
conductivitywhich
conductivity which may may be attributed
be attributed to space tocharge
spaceeffects,
charge effects,strain,
epitaxial epitaxial strain,reconstruction
and atomic and atomic
reconstruction at the interface of the YSZ/STO thin film [19]. The discontinuity
at the interface of the YSZ/STO thin film [19]. The discontinuity of free energy at the interface of free energy at the
interface
may resultmay result in
in charge chargeprocesses
transfer transfer processes
that break that breakneutrality
charge charge neutrality
and mayand may profoundly
profoundly change
change the carrier density. In ionic compounds, the accumulation
the carrier density. In ionic compounds, the accumulation of defects may create an electric of defects may create an electric
field
fieldcan
that thatbecanscreened
be screenedoverover a space-charge
a space-charge layer
layer byby depletionororaccumulation
depletion accumulationof of mobile
mobile charges.
charges.
The ionic
The ionic interactions
interactions areare found
found to to be
be essential
essential to to reproduce
reproduce the the effective
effective activation
activation energy,
energy, and
and the
the
enhanced oxygen ion mobility may result from a non-random distribution
enhanced oxygen ion mobility may result from a non-random distribution of dopant Y ions at the of dopant Y ions at the
interfacial planes, structural disorder, or a decrease in ionic interactions
interfacial planes, structural disorder, or a decrease in ionic interactions when the layer thicknesswhen the layer thickness is
down
is down to the nanometer
to the nanometer scale.scale.
We believe that the
We believe combination
that of epitaxial
the combination strain andstrain
of epitaxial oxygen andsub-lattice
oxygen
incompatibility between the YSZ and STO structures may be
sub-lattice incompatibility between the YSZ and STO structures may be the key in yielding the key in yielding high oxygen
high
ion conductivity.
oxygen ion conductivity.
Figure 55 shows
Figure shows thethe response
response of of the
the sensor
sensor under
under bias.
bias. The
The bias
bias current
current was
was chosen
chosen by by trial
trial and
and
error in order to achieve selective NO x response and equal magnitude responses for NO and NO2.
error in order to achieve selective NOx response and equal magnitude responses for NO and NO2 .
Thesensor
The sensordevice
device with
with justjust the YSZ
the YSZ thinshowed
thin film film showed a different
a different baseline baseline
comparedcompared with the
with the YSZ/STO
YSZ/STO thin-film sensor.◦ At 200 °C, there was no significant response to
thin-film sensor. At 200 C, there was no significant response to any analytes from the YSZ thin-film any analytes from the YSZ
thin-film sensor whereas the sensor with the YSZ/STO electrolyte
sensor whereas the sensor with the YSZ/STO electrolyte was NOx selective. The supplemental was NO x selective. The

supplemental
Figure S3 shows Figure 3 shows
the device the device
sensitivity as asensitivity
function of asNOa function of NO2 concentration. The sensor
2 concentration. The sensor response and
response and recovery time towards NO x were found to be 7 and 8 s, respectively.
recovery time towards NOx were found to be 7 and 8 s, respectively.
Sensors
Sensors 2017,
2017, 17,
17, 2759
2759 77 of
of 10
10

Figure 5. Current-biasedmixed-potential
5. Current-biased mixed-potentialsensor
sensorresponse
responsetotovarious
variousanalytes
analytes(25
(25 ppm)
ppm) atat 200 ◦ C,
200 °C,
Figure
500 sccm,
500 sccm, and
and oxygen
oxygen partial
partial pressure
pressure of
of 21%.
21%. The
The sensor
sensor made
made with
with aa YSZ/STO
YSZ/STOthin-film
thin-film electrolyte
electrolyte
showed selective NO response, whereas the sensor made of just the YSZ thin film
showed selective NOx response, whereas the sensor made of just the YSZ thin film showed
x showed no response.
no
The letter ‘u’ stands for the symbol ‘µ’. Humidity was not added in this experiment.
response. The letter ‘u’ stands for the symbol ‘µ’. Humidity was not added in this experiment.

Next,
Next, thethe effect
effect of of humidity
humidity on on C C33HH6,6 ,CC 3HH
3 8, 8and
, andNHNH 3 analytes
3 analytes were
were tested
testedat at275275°C◦ Cusing
using a
bubbler.
a bubbler.Figure
Figure 6 shows
6 shows thetheeffect of humidity
effect of humidity on the hydrocarbons
on the hydrocarbons andandammonia.
ammonia. WithWith
8% RH, the
8% RH,
sensor hadhad
the sensor a 9.7%
a 9.7%decrease
decrease in in
thethesensor
sensor response
responsetotohydrocarbons
hydrocarbonsand andammonia.
ammonia.The The dominant
dominant
open-circuit
open-circuit sensor
sensor response
response to C33H H66isisrelatively
relativelyunaffected
unaffectedby bythe
theaddition
additionof of humidity.
humidity.
Typically,
Typically, the sensitivity of a mixed-potential
mixed-potential sensor sensor decreases
decreases with
with anan increase
increase in in operating
operating
temperature. To identify the lower bound and upper bound of the
temperature. To identify the lower bound and upper bound of the sensor operating temperature, sensor operating temperature, the
sensitivity to hydrocarbons
the sensitivity to hydrocarbons was monitored
was monitored (figure(figure
not shown) as a function
not shown) of temperature.
as a function Above
of temperature.
400
Above the◦ C,
°C,400 sensitivity to hydrocarbons
the sensitivity to hydrocarbons drops. At 550
drops. °C, ◦the
At 550 sensitivity
C, the sensitivity drops
dropsclose
closetotozero,
zero,
confirming the mixed-potential sensing behavior. On the other hand,
confirming the mixed-potential sensing behavior. On the other hand, below 175 C, the sensitivity below 175 ◦
°C, the sensitivity
drops
drops to 60% of the the value
value at 200◦°C.
at 200 C. AAdetailed
detailedGas GasChromatography–Mass
Chromatography–MassSpectrometry
Spectrometry is is needed
needed to
to completely
completely explain
explain thethe effect
effect of operating
of operating temperature
temperature on the
on the sensor
sensor performance;
performance; this this
will will
formform
a parta
part of future
of future studies. studies.
Next,
Next, the
the long-term
long-term stability (120 days) of the YSZ/STO YSZ/STOthin-film
thin-film sensor
sensor was
was evaluated
evaluated at at 200 ◦C
200 °C
(Figure
(Figure 7).
7). The
The sensor
sensor was was tested
tested every
every 10th10th day day forfor HC
HC and
and NOxx response
response by by maintaining
maintaining all all the
the
experimental
experimental conditions
conditions as as constant.
constant. The Thesensor
sensorwas waskeptkept in
in ambient
ambient airair (oxygen
(oxygen partial
partial pressure
pressure of of
21%)
21%) and
and exposed
exposed to to analytes
analytes (25 (25 ppm)
ppm) on on alternate
alternate days days (idle
(idle days)
days) to
to identify
identify any
any aging
aging or or sensor
sensor
degradation
degradation over over time.
time.
After
After 120
120 days,
days, thethe sensor
sensor showed
showed slightslight decrease
decrease in HC and NOxx sensor sensor response.
response. An An 11.4%
11.4%
decrease
decrease inin HCHC response
response and a 3.3% decrease in NOxx response response werewere observed.
The
The applications
applications for for low-temperature
low-temperature gas gas sensing
sensing withwith minimal
minimal power
power requirements
requirements has has seen
seen
an
an escalated
escalated interest,
interest, particularly
particularlywith withthe theadvent
adventofofthe theInternet
Internetof of
Things
Things (IoT). TheThe
(IoT). features of
features
electrochemical sensors make them an ideal choice for IoT devices.
of electrochemical sensors make them an ideal choice for IoT devices. The promising results The promising results reported in
this articleinindicate
reported this articlethatindicate
solid electrolytes such as YSZ/STO
that solid electrolytes such asthin-film
YSZ/STO heterostructures can enable
thin-film heterostructures
low-temperature and low-power gas sensors. Another application space for these sensors is
air-quality monitoring.
Sensors 2017, 17, 2759 8 of 10

can enable low-temperature and low-power gas sensors. Another application space for these sensors is
air-quality monitoring.
Sensors 2017, 17, 2759
Sensors 2017, 17, 2759
8 of 10
8 of 10

Figure
6. 6. Effect of humidity (8%
Effect (8%RH) on on
the sensor response to Hydrocarbons and Ammonia. The
Figure
Figure 6. Effectofofhumidity
humidity (8% RH)
RH) on thethe sensor
sensor response
response to Hydrocarbons
to Hydrocarbons and Ammonia.
and Ammonia. The
standard
The standard deviation after
deviationafter three
after trials
three of
trialsmeasurements
of measurementsis 0.002 V.
is 0.002 V.
standard deviation three trials of measurements is 0.002 V.

Figure 7. Long-term testing of the sensor (25 ppm, 200 °C, 500 sccm). A decrease in HC and NO
Figure 7. Long-term
7. Long-term testingofofthe
testing thesensor
sensor (25
(25 ppm,
ppm, 200 °C,
◦ C,500
500sccm). A decrease in HCin and NO
Figure
mixed-potential sensitivity was observed after 120 days200
of testing. sccm).
The A decrease
error bars HC
indicate three and NO
trials
mixed-potential sensitivity was observed after 120 days of testing. The error bars indicate three trials
mixed-potential sensitivity
of analyte exposure. was observed
Humidity afterin120
was not added thisdays of testing. The error bars indicate three trials
experiment.
of analyte exposure. Humidity was not added in this experiment.
of analyte exposure. Humidity was not added in this experiment.
Urban air quality represents a major public health burden and is a long-standing concern to
Urban air quality represents a major public health burden and is a long-standing concern to
citizens. Air pollution is associated with a range of diseases, symptoms, and conditions that impair
Urban air
citizens. Airquality
pollutionrepresents
is associated a major
with a public
range ofhealth burden
diseases, and is
symptoms, a long-standing
and concern to
conditions that impair
health and quality of life. In particular, there is critical need to monitor NO and NO2 levels in ambient
health and quality of life. In particular, there is critical need to monitor NO and NO
citizens. Air pollution is associated with a range of diseases, symptoms, and conditions that impair 2 levels in ambient
Sensors 2017, 17, 2759 9 of 10

health and quality of life. In particular, there is critical need to monitor NO and NO2 levels in
ambient air. Short-term exposures (less than 3 h) to low levels of NO2 can lead to changes in
airway responsiveness and lung function in individuals with preexisting respiratory illnesses. Further,
long-term exposures to NO2 may lead to increased susceptibility to respiratory infection and may
cause irreversible alterations in lung structure. Atmospheric transformation of NOx can lead to the
formation of ozone and nitrogen-bearing particles. Nitrogen oxides contribute to a wide range of
effects on public welfare and the environment, including global warming and stratospheric ozone
depletion. Deposition of nitrogen can also lead to fertilization, eutrophication, or acidification of
terrestrial, wetland, and aquatic (e.g., fresh water bodies, estuaries, and coastal water) systems.
There is a current trend worldwide to increase the collection of air quality data beyond
reference monitoring stations. Low-cost sensor platforms play a key role in air quality monitoring.
Sensor nodes can be deployed as dense networks or mounted on vehicles, facilitating the elaboration of
high-resolution air quality maps. Low-cost sensors with minimal power consumption, robust sensitivity,
and selectivity against interferences with an adequate detection limit are the need of the hour.
For example, the concentrations of NOx in ambient air can vary from 1 ppb to 100 ppb. The sensor
reported in the article can measure reliably up to 5 ppm with a reasonable power consumption.
To improve the selectivity and detection limit, methods such as incorporation of nanomaterials in the
electrodes and a new signal conditioning technique [20], referred to as the pulsed discharge technique,
can be adopted.

4. Conclusions
In this article, a mixed-potential-based electrochemical gas sensor that can be operated at low
temperatures (200 ◦ C) was reported. The sensor was constructed on an YSZ substrate with metal oxide
and Pt electrodes, along with an YSZ/STO multilayer solid electrolyte. The sensor exhibited a dominant
hydrocarbon response at open circuit and a selective NOx response under bias. The device constructed
with just the YSZ thin film and electrodes showed no response at 200 ◦ C, indicating an enhanced oxygen
ion conductivity from the YSZ/STO heterostructure. After 120 days of testing, only an 11.4% decrease
in HC response and a 3.3% decrease in NOx response were observed. In addition, the fabricated
sensor device only consumed less than a 1 W of power; for comparison, the plain YSZ thin-film device
consumes about 5W of power under the standard operating temperature of 500 ◦ C.

Supplementary Materials: The Supplementary Materials are available online at http://www.mdpi.com/1424-

8220/17/12/2759/s1.
Acknowledgments: P.K.S. and the work was supported by the National Science Foundation under grant
No. 1503634.
Author Contributions: Z.M. and P.K.S. conceived and designed the experiments; Z.M. performed the experiments;
P.K.S. analyzed the data; S.A. and A.K. contributed reagents/materials/analysis tools; P.K.S. wrote the paper.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Pang, X.; Shaw, M.D.; Lewis, A.C.; Carpenter, L.J.; Batchellier, T. Electrochemical ozone sensors:
A miniaturized alternative for ozone measurements in laboratory experiments and air-quality monitoring.
Sens. Actuators B Chem. 2017, 240, 829–837. [CrossRef]
2. Wan, H.; Yin, H.; Mason, A.J. Rapid measurement of room temperature ionic liquid electrochemical gas
sensor using transient double potential amperometry. Sens. Actuators B Chem. 2017, 242, 658–666. [CrossRef]
[PubMed]
3. Murray, E.P.; Cui, L. Advances in Electrochemical Nitric Oxide Exhaust Gas Sensors. In Progresses in Chemical
Sensor; Wang, W., Ed.; InTech: Rijeka, Croatia, 24 August 2016; pp. 145–158.
4. Miura, N.; Sato, T.; Anggraini, S.A.; Ikeda, H.; Zhuiykov, S. A review of mixed-potential type zirconia-based
gas sensors. Ionics 2014, 20, 901–925. [CrossRef]
Sensors 2017, 17, 2759 10 of 10

5. Menil, F.; Coillard, V.; Lucat, C. Critical review of nitrogen monoxide sensors for exhaust gases of lean burn
engines. Sens. Actuators B Chem. 2000, 67, 1–23. [CrossRef]
6. Akbar, S.; Dutta, P.; Lee, C. High-Temperature Ceramic Gas Sensors: A Review. Int. J. Appl. Ceram. Technol.
2006, 3, 302–311. [CrossRef]
7. Garzon, F.H.; Mukundan, R.; Brosha, E.L. Solid-state mixed potential gas sensors: theory, experiments and
challenges. Solid State Ion. 2000, 136, 633–638. [CrossRef]
8. Szabo, N.F.; Dutta, P.K. Correlation of sensing behavior of mixed potential sensors with chemical and
electrochemical properties of electrodes. Solid State Ion. 2004, 171, 183–190. [CrossRef]
9. Fergus, J.W. A review of electrode and electrolyte materials for high temperature electrochemical CO2 and
SO2 gas sensors. Sens. Actuators B Chem. 2008, 134, 1034–1041. [CrossRef]
10. Korotcenkov, G.; Han, S.D.; Stetter, J.R. Review of Electrochemical Hydrogen Sensors. Chem. Rev. 2009, 109,
1402–1433. [CrossRef] [PubMed]
11. Liu, Y.; Parisi, J.; Sun, X.; Lei, Y. Solid-state gas sensors for high temperature applications—A review. J. Mater.
Chem. A 2014, 2, 9919–9943. [CrossRef]
12. Sekhar, P.K.; Brosha, E.L.; Mukundan, R.; Li, W.; Nelson, M.A.; Palanisamy, P.; Garzon, F.H. Application of
commercial automotive sensor manufacturing methods for NOx /NH3 mixed potential sensors for on-board
emissions control. Sens. Actuators B Chem. 2010, 144, 112–119. [CrossRef]
13. Frechero, M.A.; Rocci, M.; Sánchez-Santolino, G.; Kumar, A.; Salafranca, J.; Schmidt, R.; Díaz-Guillén, M.R.;
Durán, O.J.; Rivera-Caldara, A.; Mishra, R.; et al. Paving the way to nanoionics: Atomic origin of barriers for
ionic transport through interfaces. Sci. Rep. 2015, 5, 17229. [CrossRef] [PubMed]
14. Rivera-Calzada, A.; Diaz-Guillen, M.R.; Dura, O.J.; Sanchez-Santolino, G.; Pennycook, T.J.; Schmidt, R.;
Bruno, F.Y.; Garcia-Barriocanal, J.; Sefrioui, Z.; Nemes, N.M.; et al. Tailoring interface structure in highly
strained YSZ/STO heterostructures. Adv. Mater. 2011, 23, 5268–5274. [CrossRef] [PubMed]
15. Sillassen, M.; Eklund, P.; Pryds, N.; Johnson, E.; Helmersson, U.; Bøttiger, J. Low-temperature superionic
conductivity in strained yttria-stabilized zirconia. Adv. Funct. Mater. 2010, 20, 2071–2076. [CrossRef]
16. Mukundan, R.; Teranishi, K.; Brosha, E.L.; Garzon, F.H. Nitrogen Oxide Sensors Based on Yttria-Stabilized
Zirconia Electrolyte and Oxide Electrodes. Electrochem. Solid-State Lett. 2007, 10, J26–J29. [CrossRef]
17. Kreller, C.R.; Sekhar, P.K.; Li, W.; Palaisamy, P.; Brosha, E.L.; Mukundan, R.; Garzon, F.H. Application of
Commercial Manufacturing Methods to Mixed-Potential NOx Sensors. ECS Trans. 2013, 50, 307–314.
[CrossRef]
18. Cavallaro, A.; Burriel, M.; Roqueta, J.; Apostolidis, A.; Bernardi, A.; Tarancón, A.; Srinivasan, R.; Cook, S.N.;
Fraser, H.L.; Kilner, J.A.; et al. Electronic nature of the enhanced conductivity in YSZ-STO multilayers
deposited by PLD. Solid State Ion. 2010, 181, 592–601. [CrossRef]
19. Pennycook, T.J.; Beck, M.J.; Varga, K.; Varela, M.; Pennycook, S.J.; Pantelides, S.T. Origin of colossal ionic
conductivity in oxide multilayers: Interface induced sublattice disorder. Phys. Rev. Lett. 2010, 104, 115901.
[CrossRef] [PubMed]
20. Sekhar, P.K.; Brosha, E.; Mukundan, R.; Farber, B.; Shuk, P.; Garzon, F.H. Investigation of a new conditioning
method for improved performance of non-Nernstian sensors. ECS Trans. 2010, 28, 1–8.

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).