Recoverypropertiesofhydrogengassensorwith Pd/titanate
and Pt/titanate nanotubes photo-catalyst by UV radiation
from catalytic poisoning of H
2
S
Dae Ung Hong
a,b
, Chi-Hwan Han
b,
*
, Sang Hyun Park
b
, Il-Jin Kim
b
, Jihye Gwak
b
,
Sang-Do Han
b
, Hyun Jae Kim
a
a
Department of Electrical and Electronic Engineering, Yonsei University, 134, Shinchon-dong, Seodaemoon-ku, Seoul 120-749, Republic of Korea
b
Electrical and Electronic Materials Research Center, Korea Institute of Energy Research, 71-2, Jangdong, Yuseong, Daejeon 305-343, Republic of Korea
Received 15 October 2007; received in revised form 17 January 2008; accepted 17 January 2008
Available online 1 February 2008
Abstract
Recovery properties after H
2
S catalytic poisoning of catalytic-type gassensorwith photo-catalysts and UV radiation have been exam-
ined. Each sensing material of the sensor consists of Pd, Pt supported on c-Al
2
O
3
and Pd/titanate, Pt/titanate nanotubes or TiO
2
particles.
Pd/titanate and Pt/titanate nanotubes photo-catalyst were synthesized by hydrothermal synthesis method. All the sensors were deactivated
after 500 ppm H
2
S exposure for 20 h. The sensors with Pd/titanate or Pt/titanate nanotubes showed regenerated voltage response under UV
radiation. However the sensorwith TiO
2
particles showed negligible regenerated voltage response. Regenerated voltage response with Pd/
titanate or Pt/titanate nanotubes may stem from location of Pd or Pt catalyst on the titanate nanotube photo-catalyst.
Ó 2008 Elsevier B.V. All rights reserved.
PACS: 07.07.Df
Keywords: Catalytic poisoning; H
2
S; H
2
sensor; Photo-catalyst; Titanate nanotubes
1. Introduction
The research interest on hydrogen as a clean energy
resource or a fuel gas has been increased remarkably
because it is renewable, abundant and efficient with zero
emissions. It is extensively used to make ammonia, metha-
nol, gasoline, heating oil, and rocket fuel, etc. The amount
of energy produced by hydrogen is three times bigger than
the energy contained in equal weight of gasoline and about
seven times that of coal. Hydrogen can replace natural gas
in warming home and powering hot water heaters [1–5].
Like any other gas type fuel, hydrogen is flammable and
potentially dangerous. Safety is the first priority in using
hydrogen gas as fuel. Sensing hydrogen leakage from stor-
age and transportation equipment is essent ial. Hydrogen
also demands a careful ha ndling, because a 4% (v/v) mix-
ture in air is its lower explosive limit (LEL) [4]. The mon-
itoring of the concentration of this gas close to its
production and consu mption plants is necessary to avoid
accidents due to hydrogen explosions.
One of the simplest forms for H
2
monitoring is the use
of catalytic combustion sensors . Catalytic type gas sensors
have been developed by many research groups [5–9]. Com-
bustible gas mixtures do not burn until they reach an igni-
tion temperature. However, in the presence of certain
chemical media, the gas can ignite and burn at lower tem-
perature. This phenomenon is known as a catalytic com-
bustion. A gas molecule oxidizes on the catalyzed surface
of the sensor at a much lower temperature than its normal
ignition temperature. Every conductive material has its
own coefficient of temperature resistance (C
t
). Platinum
1567-1739/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.cap.2008.01.010
*
Corresponding author. Tel.: +82 42 860 3449; fax: +82 42 860 3307.
E-mail address: hanchi@kier.re.kr (C H. Han).
www.elsevier.com/locate/cap
www.kps.or.kr
Available online at www.sciencedirect.com
Current Applied Physics 9 (2009) 172–178
which has large C
t
in comparison with other metals is a
good candidate for the catalytic combustible sensor
because it can detect flammable gases by measuring resis-
tance change of heater metal. In addition, its C
t
is linear
between 500 °C and 1000 °C, which is the temperature
range at which the sensor needs to operate. The catalytic
surface is generally prepared by sintering noble metal par-
ticles (Pt, Pd) on a high surface area material like c-Al
2
O
3
[4–7]. However, there are still certain limitations associated
with the catalytic sensors to be applied. These sensors show
low sensitivity due to the lack of adsorption sites for hydro-
gen and are affected by a small amount of poisonous gases.
Therefore, removing catalyst poisoning is extremely impor-
tant. The poisoning has been reported in many classes of
chemical products such as molecules containing sulfur,
hexamethyldisiloxane (HMDS), nitrogen, silicon, nitric
oxide, etc [10–15]. Recently, a remarkable recoveryof a
semiconductor type titania nanotubes hydrogen sensor
from sensor poisoning through UV photo-catalytic oxida-
tion of the contaminants was reported [16].
The recoverypropertiesof catalytic type hydrogen sen-
sor with various catalysts from H
2
S poisoning are shown
in this paper. H
2
S was selected as the catalytic poisoning
species because it is one of the worst and most commonly
encountered catalyst deactivating compounds among sul-
fur containing compounds. Many kinds of catalysts were
used, such as Pd/titanate, Pt/titanate nanotubes, TiO
2
,Pt
and Pd. The performance of the sensor after H
2
S poisoning
and recoveryproperties by UV radiation was tested.
2. Experimental
The synthesis of Pd/titanate and Pt/titanate nanotubes
was processed with several steps. All the chemicals were
purchased from Aldrich. Anatase-type titanate powder
(4 g) was dispersed into an aqueous solution of NaOH
(10 M, 80 ml). Then, PdCl
2
(4 g) or PtCl
2
(1 g) was added
into the solution, which charged into a Teflon-lined auto-
clave. The au toclave was heated at 150 °C for 15 h. After
the hydrothermal treatment, the precipitate was washed
with deionized water and separated by filtration. Final
product was obtained through air-drying at 120 °C in oven
[17,18]. Morphology of the samples was observed by a field
emission scanning electron microscopy (FE-SEM) using
Hitachi S-4300 and by field emission transmission electron
microscopy (FE-TEM) using JEOL JEM-2100 F. The ele-
ments ratio of the sensor surface was observed by disper-
sive X-ray spectroscopy (EDS) using Horiba 7200-H.
BET surface area of each sensor material was measured
by the nitrogen sorption method at the liquid nitrogen tem-
perature using Micromeritics ASAP 2010. Before each
measurement, samples were degassed at 200 °C in vacuum
until constant pressure ($3 lm Hg) was obtained.
The sensing materials of the sensors were listed in Table
1 and EDS results of each sensor surface were also listed in
Table 2. The reference material was an inactive c-Al
2
O
3
.
The metal oxide powder material was mixed with an
organic and inorganic vehicle at a concentration of
15 wt.% followed by ball-milling for 24 h, to prepare the
pastes suitable for drop coating.
Fig. 1 shows the structure and size of the present sensor
device. The sensor device was fabricated in the following pro-
cedure. First, a platinum micro-heater was formed on an alu-
mina plate by a screen-printing method with platinum paste
(METECH, Platinum conductor PCC 11417) followed by
heat treatment at 950 °C for 10 min. Second, the sensing ele-
ment was formed by drop coating of a catalytic layer on the
platinum heater, followed by firing at 650 °C for 1 h. Finally,
the sensing and compensating elements were linked to signal
pins of the sensor body by spot welding (WITH Corpora-
tion, WMH-V1) with platinum wire (ø 50 lm).
The compensating element forms one arm of the wheat-
stone bridge, which is shown in Fig. 2a. The sensor element
is connected in series with the bridge. The surface temper-
ature of the sensor at each applied voltage was measured
by a radiation thermo tracer (NEC TH9100MLN).
Measurements were carried out using an environmental
test chamber with simulation gas containing 1% H
2
. The
schematic of measur ement settings is shown in Fig. 2b.
Fresh air was introduced then inlet and outlet of the cham-
ber were closed. The device was exposed to a hydrogen gas
sample for around 30 s for gas response test and device was
Table 1
Compositions of the fabricated sensor materials in wt.% ratios and voltage response of each sensor
Sample number Composition of the used materials (wt.%) DV at 310 °C (mV)
c-Al
2
O
3
Pt/titanate nanotube Pd/titanate nanotube TiO
2
Pd Pt
S1 80 10 10 61.5
S2 60 20 10 10 61.7
S3 50 40 10 73.3
S4 70 20 10 73.3
S5 60 30 10 84.5
Table 2
EDS results and BET surface area of each sensor
Sample number EDS results (wt.%) BET surface area (m
2
/g)
OAlTiPdPt
S1 49.2 40.5 5.2 5.1 18.0
S2 54.8 25.0 13.1 3.1 4.0 17.5
S3 49.5 20.0 16.8 5.3 7.5 43.4
S4 45.7 22.4 15.6 7.8 6.2 25.9
S5 48.7 24.3 16.2 5.8 5.0 32.5
D.U. Hong et al. / Current Applied Physics 9 (2009) 172–178 173
recovered by exposing to purified air again. A mass flow
controller (MFC) was employed to fix the gas flow rate.
The gas concentration was controlled by selecting appro-
priate values of the flow rates. For practical poisoning test
for the hydrogen sensor, 500 ppm H
2
S gas was fed for 20 h
before the simulation gas was fed. Then sensitivity of the
sensors for 1% H
2
was measured. UV light was irradiated
by blacklight blue lamps which efficiently emit near ultravi-
olet rays at 315–400 nm. Gas sensitivity (DV) was defined
as the difference between the outpu t voltage in a sample
gas (V
g
) and that in air (V
a
): DV=V
g
À V
a
.
3. Results and discussion
3.1. Characterization of photo-catalyst
Fig. 3 shows typical TEM images demonstrating uni-
form sized titanate nanotubes over which Pt or Pd nano-
particles are randomly distributed. The outer diameter of
nanotubes in TEM images is approximately 100 nm.
Detailed characteristics of Pd/titanate and Pt/titanate
nanotubes have been reported in our previous study [19] .
3.2. Sensor performance with different catalyst
Compositions of the fabricated sensor materials and the
voltage responses of each sensor were listed in Table 1.
Fig. 4 shows the response of sensors to 1% hydrogen gas
before H
2
S gas exposure. The maximum response (DV)of
sensors can be achieved at nearly 310 °C. It is clear from
Fig. 4 that sensing performance of the sensors using Pd/
titanate or Pt/titanate nanotubes catalyst supported on c-
Al
2
O
3
(S3–S5) is better than that of the sensors employing
other catalysts like Pd, Pt and TiO
2
at operating tempera-
ture of 310 °C. This may stem from increased adsorption
sites and surface area of Pd/t itanate and Pt/titanate nano-
tube catalysts. BET surfaces areas of the sensor materials
were listed in Table 2. It was clearly observed that BET
surface area increased as titanate nanotube increased. S3
Fig. 1. Schematic diagram of (a) the sensor structure and (b) the
fabricated sensor. Dimensions are in lm.
Fig. 2. (a) Bridge circuit for the output voltage and (b) schematic view of test system.
174 D.U. Hong et al. / Current Applied Physics 9 (2009) 172–178
Fig. 3. TEM images of (a) Pd/titanate and (b) Pt/titanate nanotubes.
0 100 200 300 400 500
-0.01
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
S1
S2
S3
S4
S5
Δ
V (V)
Heater temperature (ºC)
Operating temperature
(310
o
C)
Fig. 4. The voltage responses of sensors using various catalysts to 1%
hydrogen.
0 100 200 300 400 500
0.030
0.035
0.040
0.045
0.050
0.055
0.060
0.065
0.070
0.075
Δ
V (V)
H
2
S concentration (ppm)
Fig. 5. The voltage response property of S1 sensor for 1% H
2
with
different H
2
S exposure conditions at 100 °C.
-5 0 5 10 15 20 25 30 35 40 45 50 55
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
Δ
V (V)
UV radiation time (h)
S1
S2
S3
S4
S5
Fig. 7. Response changes of the sensors by poisoning of 500 ppm H
2
S and
reactivation treatment with UV radiation.
0 5 10 15 20
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
Δ
V (V)
Time (h)
S1
S2
S3
S4
S5
Fig. 6. The voltage response propertiesofgas sensors for 1% H
2
after H
2
S
500 ppm poisoning at 100 °C.
D.U. Hong et al. / Current Applied Physics 9 (2009) 172–178 175
sensor including highest weight ratio of titanate nanotube
showed the largest BET surface area of 43.4 m
2
/g. Other
cause for the enhanced response may due to the better
adsorption ofhydrogen on the titanate nanotube surface
which facilitates the oxidation ofhydrogen reaction by
the Pd and Pt catalysts on the titanate nanotube surface
[19].
Among S3, S4, and S5 sensors, S5 sensor showed highest
sensor response to the 1% hydrogen above 200 °C as shown
in Fig. 4, although it had smaller surface area than S3. The
response of S5 sensor is 11.2 mV higher than those of S3
and S4 sensors at operating temperature of 310 °C. From
these results, it may be thought that catalytic property
for the hydrogengas combustion of Pd/titanate nanotube
is considerably better than that of Pt/titanate nanotube.
3.3. The response propertiesof sensors after H
2
S poisoning
Fig. 5 shows the voltage response property of S1 sensor
for 1% H
2
before and after different H
2
S exposure condi-
tions at 100 °C. It was observed that the voltage response
of S1 sensor for 1% H
2
decreased with increasing amount
of H
2
S injected. After 500 ppm H
2
S injection, the voltage
response difference of S1 sensor for 1% H
2
became com-
pletely saturated. The response of S1 sensor for 1% H
2
decreased around 27 mV after 500 ppm H
2
S exposure. It
was considered that H
2
S was adsorbed on catalyst surface
and it worked as catalytic poisoning species.
The relationship between the voltage response of the
sensors and H
2
S exposure time are shown in Fig. 6.
500 ppm H
2
S was introduced into the chamber every 5 h
Fig. 8. SEM images of fabricated sensing layer for (a) S1, (b) S2, (c) S3, (d) S4 and (e) S5.
176 D.U. Hong et al. / Current Applied Physics 9 (2009) 172–178
and then the device was recovered by exposing to purified
air again. The voltage responses of the sensors for 1% H
2
were measured after 500 ppm H
2
S exposure. It is clear from
Fig. 6 that the voltage response of various sensors for 1%
H
2
decreased with increasing exposure time of 500 ppm
H
2
S. The responses of S1, S2, S3, S4 and S5 sensors for
1% H
2
decreased by around 25, 22, 21, 18 and 23 mV,
respectively, after 20 h H
2
S poisoning.
3.4. The recoverypropertiesofsensor response after UV
radiation
Fig. 7 shows the recoverypropertiesof sensors with UV
radiation for 50 h. When UV light was illuminated on the
catalytic sensor, a remarkable difference in the recovery
properties of the sensors was observed. The S1 and S2 sen-
sors showed negligible regenerated voltage responses. How -
ever, the voltage responses of S3, S4 and S5 sensors for 1%
H
2
increased by around 20, 22 and 17 mV after UV radia-
tion. A common point of S3, S4, S5 sensors is using Pd or
Pt/titanate nanotube catalyst, as listed in Table 1.
To elucidate the different recoveryproperties between
S1, S2 sensors and S3, S4, S5 sensors, the sensor surface
was examined by SEM and shown in Fig. 8. The titanate
nanotubes of S3, S4, S5 sensors were found to be entangled
to other particles and forming net structure on the detect-
ing surface of the sensors. When TiO
2
or titanate nanotube
photo catalyst is irradiated with UV light, electrons and
holes are generated in it. The photogenerated holes in the
valence band can oxidize water to produce highly reactive
hydroxyl radical (
Å
OH), and the photogenerated electrons
in the conduction band can reduce oxygen to form highly
reactive superoxide (O
À
2
Å
) ions, which then assist in oxidiz-
ing adsorbed and gaseous H
2
S into sulfate via SO
2
or SO
2À
3
[15,16].
Maxted has reported that the poisoning effect of sulfate
was less than that of sulfide by comparing compounds [20].
Sulfide compounds are coordinated directly with Pd using
two anti-bonding lone pairs. The activity of Pd catalyst
was drastically reduced by sulfide. In contrast, the sulfur
atom of sulfate is surrounded by oxygen atoms. The struc-
ture of sulfate satisfies with the octet rule and the sulfur
atom of sulfate does not bind directly with Pd. The interac-
tion between Pd and S atom of sulfate is smaller than that
of sulfide, and thus the poisoning effect of sulfate is smaller.
From our experimental results Pd or Pt dispersed titanate
nanotubes catalysts were recovered by UV radiation from
H
2
S catalytic poisoning, however Pd/Pt catalysts mixed with
TiO
2
nano particles (S2) were not recovered by UV radia-
tion. The regenerated voltage response with titanate nano-
tubes may stem from location of Pd or Pt catalyst on the
titanate nanotube photo-catalyst. The life time of hydroxyl
radical (
Å
OH) and superoxide (O
À
2
Å
) ions which is formed
on the photo-catalyst surface is very short and only can oxi-
dize H
2
S to sulfate of adjacent Pd and Pt. For S2 sensor, poi-
soned Pd or Pt can not be recovered because the (Pd, Pt)
catalysts existed apart from the photo-catalyst.
Fig. 9 shows a relationship between change of output
voltage and the hydrogen concentration after UV radiation
at the operating temperature 310 °C. It was observed that
the voltage difference was proportional to the hydrogen
in the concentration range of 0.5–4% (v/v), and the catalyst
after UV radiation was still efficiently active. However the
catalyst without UV radiation became deactivated at the
same reaction time.
4. Conclusion
We have clearly shown the recoverypropertiesof hydro-
gen sensorwith titanate nanotube catalysts by UV radia-
tion from catalytic poisoning of H
2
S. The catalytic-t ype
hydrogen sensorwith Pd or Pt/titanate nanotubes having
large response and recovery property from H
2
S poisoning
by UV radiation can be a good candidate for future hydro-
gen sensor. The regenerated voltage response with titanate
nanotubes may stem from location of Pd or Pt catalyst on
the titanate nano tube photo-catalyst, and can be explained
by oxidizing adsorbed and gaseous H
2
S into sulfate, which
is less poisonous than H
2
S, after UV radiation. The voltage
difference of sensors was proportional to the hydrogen in
the concentration range of 0.5–4% (v/v).
Acknowledgement
This research was performed for the Hydrogen Energy
R&D Center, one of the 21st Century Frontier R&D Pro-
gram, funded by the Ministry of Science and Technology of
Korea.
References
[1] W. Shin, K. Imai, N. Izu, N. Murayama, Thermoelectric thick-film
hydrogen gassensor operating at room temperature, Jpn. J. Appl.
Phys. 40 (2001) 1232.
[2] V.A. Goltsov, T.N. Veziroglu, A step on the road to hydrogen
civilization, Int. J. Hydrogen Energy 27 (2004) 719–723.
01234
0.00
0.05
0.10
0.15
0.20
0.25
0.30
Δ
V (V)
H
2
concentration (%)
S3
S4
S5
Fig. 9. The relationship between the voltage response and hydrogen
concentration after UV radiation at 310 °C.
D.U. Hong et al. / Current Applied Physics 9 (2009) 172–178 177
[3] S. Shukla, S. Seal, L. Ludwig, C. Parish, Nanocrystalline indium
oxide-doped tin oxide thin film as low temperature hydrogen sensor,
Sens. Actuators, B 97 (2004) 256.
[4] J.G. Firth, A. Jones, T.A. Jones, Principle of the detection of
flammable atmospheres by catalytic devices, Combust. Flame 21
(1973) 303–311.
[5] S.J. Gentry, T.A. Jones, The role of catalysts in solid-state gas
sensors, Sens. Actuators 10 (1986) 141–163.
[6] J.B. Miller, C. Hort, T.B. Scheffler, Catalytic Sensor, US Patent No.
6911180132, 2005.
[7] M.G. Jones, T.G. Nevell, The detection ofhydrogen using catalytic
flammable gas sensors, Sens. Actuators 16 (1989) 215–224.
[8] D.W. Dabill, S.J. Gentry, P.T. Walsh, A fast response catalytic sensor
for flammable gases, Sens. Actuators 11 (1987) 135–143.
[9] V.R. Katti, A.K. Dehnath, S.C. Gadkari, S.K. Gupta, V.C. Sahni,
Passivated thick film catalytic type H
2
sensor operating at low
temperature, Sens. Actuators, B 84 (2002) 219–225.
[10] L. Colin, A. Cassuto, J.J. Ehrhardt, M.F. Ruiz-Lopez, Adsorption
and decomposition of hexamethyldisiloxane on platinum: an XPS,
UPS and TDS study, Appl. Surf. Sci. 99 (1996) 245–254.
[11] J.J. Ehrhardt, L. Colin, D. Jamois, Poisoning of platinum surfaces by
hexamethyldisiloxane (HMDS): application to catalytic methane
sensors, Sens. Actuators B 40 (1997) 117–124.
[12] M. Matsumiya, W. Shin, F. Qiu, N. Izu, I. Matsubara, N.
Murayama, Poisoning of platinum thin film catalyst by hexame-
thyldisiloxane (HMDS) for thermoelectric hydrogengas sensor, Sens.
Actuators B 96 (2003) 516–522.
[13] M.C. Canela, R.M. Alberici, W.F. Jardim, Gas-phase destruction of
H
2
S using TiO
2
/UV–Vis, J. Photochem. Photobiol. A 112 (1998) 73.
[14] J.P. Chen, R.T. Yang, Mechanism of poisoning of the V
2
O
5
/TiO
2
catalyst for the reduction of NO by NH
3
, J. Catal. 125 (1990) 411–420.
[15] T. Kako, H. Irie, K. Hashimoto, Prevention against catalytic
poisoning by H
2
S utilizing TiO
2
photocatalyst, J. Photochem.
Photobiol. A 171 (2005) 131–135.
[16] G.K. Mor, M.A. Carvalho, O.K. Varghege, M.V. Pishko, C.A.
Grimes, A room temperature TiO
2
-nanotube hydrogensensor able to
self-clean photo activity from environmental contamination, J.
Mater. Res. 19 (2004).
[17] R. Ma, Y. Bando, T. Sasaki, Nanotubes of lepidocrocite titanates,
Chem. Phys. Lett. 380 (2003) 577–582.
[18] T. Kasuga, M. Hiramatsu, Titania nanotubes prepared by chemical
processing, Adv. Mater. 11 (1999) 1307–1311.
[19] C H. Han, D.U. Hong, I J. Kim, J. Gwak, S D. Han, K.C. Singh,
Synthesis of Pd or Pt/titanate nanotube and its application to catalytic
type hydrogengas sensor, Sens. Actuators B 128 (2007) 320–325.
[20] E.B. Maxted, Adv. Catal. 3 (1951) 129.
178 D.U. Hong et al. / Current Applied Physics 9 (2009) 172–178
. H
2
S poisoning.
3.4. The recovery properties of sensor response after UV
radiation
Fig. 7 shows the recovery properties of sensors with UV
radiation for 50. Recovery properties of hydrogen gas sensor with Pd/titanate
and Pt/titanate nanotubes photo-catalyst by UV radiation
from catalytic poisoning of H
2
S
Dae