recovery properties of hydrogen gas sensor with pdtitanate

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 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 gas sensor with 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 sensor with 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 recovery of a semiconductor type titania nanotubes hydrogen sensor from sensor poisoning through UV photo-catalytic oxida- tion of the contaminants was reported [16]. The recovery properties of 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 recovery properties 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 properties of gas 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 of hydrogen on the titanate nanotube surface which facilitates the oxidation of hydrogen 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 hydrogen gas combustion of Pd/titanate nanotube is considerably better than that of Pt/titanate nanotube. 3.3. The response properties of 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 recovery properties of sensor response after UV radiation Fig. 7 shows the recovery properties of 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 recovery properties 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 recovery properties of hydro- gen sensor with titanate nanotube catalysts by UV radia- tion from catalytic poisoning of H 2 S. The catalytic-t ype hydrogen sensor with 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 gas sensor 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 of hydrogen 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 hydrogen gas 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 hydrogen sensor 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 hydrogen gas 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