- Email: [email protected]
Sensing properties of CdS-doped tin oxide thick film gas sensor
Sensors and Actuators B 144 (2010) 37–42
Contents lists available at ScienceDirect
Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb
Sensing properties of CdS-doped tin oxide thick film gas sensor Lallan Yadava a,∗ , Ritesh Verma a , R. Dwivedi b a b
Thin Film Laboratory, Department of Physics, Deen Dayal Upadhyaya Gorakhpur University, Gorakhpur 273009, India Centre for Research in Microelectronics Engineering, Department of Electronics Engineering, Institute of Technology, Banaras Hindu University, Varanasi 221005, India
a r t i c l e
i n f o
Article history: Received 25 January 2009 Received in revised form 3 October 2009 Accepted 7 October 2009 Available online 6 November 2009 Keywords: Tin oxide CdS-doping Screen-printing Methanol sensor
a b s t r a c t The tin oxide (SnO2 ) thick film gas sensor is fabricated by employing screen-printing technology. This pure SnO2 thick film is doped with 1 or 2 wt% of cadmium sulphide (CdS) by its weight and, thereby, the effect of dopant is presented. X-ray diffraction (XRD) analyses are administrated, which suggest that CdS dopant inhibits the crystallite growth leading to nanometric reduction in grain size. The fabricated gas sensor is responsively studied on exposure to liquid petroleum gas (LPG), methanol, and acetone. It is observed that CdS (2 wt%) doped structure exhibited highest response and is more selective to methanol (70 for 5000 ppm) over LPG and acetone at the operating temperature 200 ◦ C. The CdSdoping improved response- and recovery-time from 90 s and 200 s, for undoped-film, to 40 s and 110 s for methanol (5000 ppm at 200 ◦ C). © 2009 Elsevier B.V. All rights reserved.
1. Introduction Unlike traditional chemical detection of noxious gases, there is inexorable shift towards designing and fabricating electronic sensors (structures). Amongst various kind of sensors, the chemical gas sensors, recently, received surge of interests on account of their better sensitivity and selectivity for inflammable, combustible and pollutant gases/odors [1]. Semiconducting SnO2 films are extensively used for sensing and discrimination of several reducing gases [2]. The noble metals such as Pt, Pd, and Ni are utilized as dopants in SnO2 thick/thin film to enhance its sensitivity, selectivity and to improve response and recovery times [3]. It is investigated that adsorbed oxygen gas molecules on the surface modify the conductivity of SnO2 thick film layer. Moreover, these oxygen species reacts with exposing (reducing) gases resulting in enhancement of conductance. Other oxides such as ZnO, TiO2 , Fe2 O3 , WO3 , ZrO3 , and V2 O5 are also applied to detect gases/odors [4–11]. For a film, structural stability, porosity and large surface-volume ratio are salient factors in determining its suitability in gas-sensor technology. It is established fact that reduction in grain size of a material increases sensitivity [12,13]. Yamazoe et al. [14] studied additives’ effects in semiconductor gas sensors. Dopant like CaO, MgO are seen to inhibit grain growth and, thereby, increasing gas-sensitivity. Other additives such as ZnO, CuO, and MnO affect conductivity of SnO2 considerably leading into reduction of resistivity. Castro et al. [15] reported the effect of nickel oxide doping in
∗ Corresponding author. Tel.: +91 9450253084. E-mail address: [email protected] (L. Yadava). 0925-4005/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.snb.2009.10.013
SnO2 , which resulted maximum sensitivity for ethyl alcohol. Jain et al. [16] studied the effect of Ni and Al doping on a screen-printed SnO2 thick film and showed that Ni/Al doping results the broadening in the XRD peaks, thus lowers the grain size and improved the gas sensitivity for LPG. Tianshu et al. [17] investigated Cd-doped SnO2 -based sensor for the detection of ethanol and hydrogen. He reported that CdO doping suppresses SnO2 crystallite growth effectively. ZnO, Fe2 O3 are used by Arshak and Gaidan [18] and the structure is reported to be appropriate detector for methanol, ethanol and propanol. In the present study, we investigated the role and effect of CdS on SnO2 thick film. The undoped and doped structure is analyzed and its suitability for detection of LPG, methanol and acetone is studied. In Section 2, the preparation of SnO2 paste and fabrication of undoped and doped thick film sensors are described. The measurement results, its discussions and mechanism responsible for sensing the incoming gases are presented in Section 3 and Section 4 concludes the findings.
2. Experimental 2.1. Preparation of tin oxide paste Doped and undoped paste of SnO2 is prepared in the laboratory. The tin oxide (SnO2 ) powder and glass binder (10 wt% of SnO2 ) are weighed using electronic balance. Taking SnO2 powder along with glass binder in the ball mill (Zirconia Ball Mill, Retsch), a mixture is obtained after 3–4 h of processing. Fine grains are, then, mixed with organic binder (diethyl glycol monobutyl) and organic solvent (␣-terpinol) in ball mill for 1–2 h, which results an undoped SnO2
38
L. Yadava et al. / Sensors and Actuators B 144 (2010) 37–42
paste. For doped pastes of SnO2 , weighed SnO2 powder with glass binder (10 wt% of SnO2 ) and CdS powder (1 wt%, 2 wt%), all these mixed in ball mill and same organic binder and solvent are used to get the sensing paste. To optimize the firing temperature of the paste, the prepared paste is screen-printed on alumina substrate and then dried followed by fired at different temperatures in the central zone of thick film furnace. The drying temperature is kept at 150 ◦ C for 15 min and optimized firing temperature at 800 ◦ C for 30 min.
2.2. Fabrication of sensor The sensors are fabricated in the form of rectangular (8 mm × 10 mm) thick film deposited on the alumina substrate (25 mm × 25 mm) by screen-printing technology. The utilized substrate is, initially, screen-printed with a finger electrode pattern in rear side and a resistor heating pattern in back side. Three kinds of sensing pastes are screen-printed on designed alumina substrate to make the sensors named as S1 (undoped SnO2 ), S2 (1 wt% CdS-SnO2 ) and S3 (2 wt% CdS-SnO2 ). After each screen-printing the film is dried at 150 ◦ C for 15 min followed by firing at 800 ◦ C for 30 min, as a result of which accurate adherence of active layer to the substrate is obtained.
3. Results and discussion 3.1. X-ray diffraction (XRD) The fabricated sensors S1 , S2 and S3 respectively are analyzed by administrating XRD (Fig. 1). The XRD was performed by RIGAKU MINIFLEX-2 using Cu K␣ radiation of wavelength 1.5408 Å. All the main XRD lines belong to SnO2 with casserite structure and are indexed accordingly. The broadening in the XRD peaks shows the doping of CdS significantly reduces the crystallinity of SnO2 . The crystallite size (R) of SnO2 was evaluated by fitting the (1 0 1) diffraction peak width of SnO2 using Scherrer equation, as—R = 0.9 /( cos ), where is the wavelength of X-ray used, is the diffraction angle, ˇ is the peak full line width at half maximum (FWHM). The crystallite size of undoped and doped SnO2 with CdS composition has been listed in Table 1. It may be seen that CdS act as a growth inhibitor for SnO2 crystallization.
Fig. 1. XRD pattern of (a) S1 (undoped SnO2 ), (b) S2 (1 wt% CdS-SnO2 ) and (c) S3 (2 wt% CdS-SnO2 ) sensors.
Table 1 Crystallite size and composition of samples. Sensor no.
S1
S2
S3
Composition CdS-SnO2 (wt%) Crystallite size, R (nm)
0 50.9
1 30.2
2 24.8
3.2. Resistance measurement and gas sensitivity The sensor heater (heating resistor pattern) has been optimized and calibrated which implies that how much electric power needed to sensor for a desired operating temperature. The electric power consumed to get heating the pattern depends on various parameters like, resistance of heating material, structure and area of pattern also. Fig. 2 shows the variation of resistance of the sensors with temperature in air. It is evident that the resistance decreases on increasing device temperature. Moreover, the reduction in resistance is visible on CdS-doped sensing film. We regarded ‘percentage change in resistance’ to a particular gas concentration as: S=
Ra − Rg × 100 Ra
where Ra = resistance of sensor in air, Rg = resistance of sensor in the presence of gas exposed, the response-parameter. The response of the fabricated sensors (S1 , S2 and S3 ) is measured with exposure to LPG, methanol and acetone concentration (0–5000 ppm) at different operating temperatures in air ambient. The response with varying concentration of methanol at operating temperatures 150 ◦ C, 200 ◦ C and 250 ◦ C for sensor S1 is shown in Fig. 3(a). It is evident from the figure that, as the concentration increases, response increases and it is maximum (18) at 200 ◦ C. The sensing behavior of S2 and S3 sensor for methanol vapor at various operating temperature 150 ◦ C, 200 ◦ C and 250 ◦ C are reported in Fig. 3(b) and (c). It is seen that maximum response to methanol (5000 ppm) for sample S2 is 44 (2.5 times of S1 ) and for sensor S3 is 70 (4 times of S1 ) at 200 ◦ C temperature. The operating temperature has a considerable influence on the response of the sensor. As the operating temperature increases from 150 ◦ C to 200 ◦ C, the response also enhances for each sensor, however above 200 ◦ C, the response decreases. The measurements reveal that the fabricated sensors respond maximally at operating temperature 200 ◦ C. The responses of fabricated sensors S1 , S2 and S3 for acetone and LPG at 200 ◦ C temperature are, respectively, presented in Figs. 4 and 5. Fig. 4 shows the response increases with increasing acetone concentration and its maximum values are 44, 51 and 56 (at 5000 ppm) for sensors S1 , S2 and S3 individually. These measurement reflects that doping of CdS (2 wt%) to SnO2 enhanced the response to acetone 1.3 times of S1 . Fig. 5 indi-
Fig. 2. Resistance vs. temperature for sensor S1 , S2 and S3 .
L. Yadava et al. / Sensors and Actuators B 144 (2010) 37–42
39
Fig. 3. The response of the sensor as function of methanol concentration at different operating temperatures for (a) sensor S1 , (b) sensor S2 and (c) sensor S3 .
cates that the maximum response to LPG (5000 ppm) for sensors S1 , S2 and S3 are 13, 45 and 49, respectively. It is observed that sensor S3 showed maximum response to LPG, which is 3.7 times that of S1 . Fig. 6 represents a comparative responses of the sensor S3 for varying concentration of methanol, LPG and acetone at fixed temperature 200 ◦ C. The ratio of maximum responses for sensor S3 to S1 is 4, 1.3 and 3.7 times for methanol, acetone and LPG relatively. The doping of CdS (2 wt%) caused a remarkable improvement in the response and is more selective to methanol over LPG and acetone. The transient response and recovery curve of sensor S1 and S3 for methanol at 200 ◦ C temperature are plotted in Fig. 7. At time 10 min, the 5000-ppm methanol was injected into test chamber
in air ambient and after getting maximum response the chamber was opened for recovery of initial state of the device. Moreover, the response and recovery time is evaluated by the time taken to get 90% of maximum response and to recover the 90% of starting value, respectively. It is seen from Fig. 7 that doping of CdS into SnO2 films shows fast response (40 s) and recovery time (110 s) by reducing them, respectively, 50 s and 90 s from 90 s and 200 s for methanol. The transient response and recovery curve to acetone and LPG for sensor S1 and S3 are shown in Figs. 8 and 9. The measurement shows that response and recovery time decreases by 30 s and 70 s for acetone however, in case of LPG, response time reduces almost 60 s and recovery time is less affected.
40
L. Yadava et al. / Sensors and Actuators B 144 (2010) 37–42
Fig. 4. Response vs. acetone concentration for sensor S1 , S2 and S3 at 200 ◦ C operating temperature.
3.3. Gas sensing properties The sensing mechanism of metal oxide semiconductor has been explained on the basis of adsorbed oxygen, O2, on the semiconductor surface. The chemisorbed oxygen molecules remove electrons (e− ) from the conduction band and, thereby, increasing resistance of SnO2 thick film. This process may be, schematically, shown as follows: 1 O 2 2
+ e− + S → O− − S
(1)
Fig. 5. Response vs. LPG concentration for sensor S1 , S2 and S3 at 200 ◦ C operating temperature.
Fig. 6. Variation of response of the sensor S3 (2 wt% CdS-SnO2 ) as function of methanol, LPG and acetone concentration at fixed 200 ◦ C temperature.
where S is an adsorption site on sensing surface. However, during chemisorptions, O2− and O2 − also appear along with O− depending upon the temperature of metal oxides [2]. Among these anions, O2 − is treated as ‘electrophilic’ agent while O2− associated with the lattice on the surface, as a ‘nucleophilic’ agent. Following to Kohl [2,3], O2− species are highly unstable and do not play any role in determining the response. Hence among the adsorbed species, O− plays dominant role in influencing the response of thick film. This dominant adsorbed oxygen species (O− ) (Eq. (1)) creates a space-charge region near the film surface. The resistance of SnO2 thick film, loaded with species O− , at
Fig. 7. Transient response and recovery time of sensor S1 and S3 at 200 ◦ C for methanol (5000 ppm).
L. Yadava et al. / Sensors and Actuators B 144 (2010) 37–42
41
The change in resistance is attributed to the change in barrier height, at grain boundaries, due to modulating surface charge densities which depends on temperature as well as on exposure of gases. Moreover, the XRD measurement of the samples (S1 –S3 ) shows broadening in peaks and clear change in relative intensity (200 peaks) on CdS-doping. The crystallite size of SnO2 is evaluated using Scherrer equation and it is found that the same decreases from 50.9 nm to 24.8 nm on doping. Since the crystallite size falls within the 100 nm thus, we can say that average grain size lies in nanometric range. Thus, CdS-dopant acts as a crystal growth inhibitor and improves the response of sensor. It is, earlier, reported that dopant like Ni and CdO in tin oxide acts as a crystal growth inhibitor which improves sensor response on exposing various gases/vapors [16,17]. Obviously, from the XRD measurement, we may conclude that the CdS-doping lowers the average grain size, therefore, responsible for the increase in sensor response. Future work is to be done to get a definite understanding of reaction mechanism on CdS-SnO2 surface exposed with hydrocarbon vapors. 4. Conclusion
Fig. 8. Transient response and recovery time of sensor S1 and S3 at 200 ◦ C for acetone (5000 ppm).
higher temperature gets decreased on exposure of reducing gases like LPG, methanol and acetone and enhancing the response of the sensor. It is reported that methanol can be decomposed at lower temperature (150–200 ◦ C) using different catalytic metal oxides [19,20]. The highest response of the sensor may be accounted for liberation of more free electrons, when species O− , possibly reacts with methanol as follows: 2CH3 OH + 2[O− ] → CO + CO2 + 3H2 + H2 O + 2e−
(2)
In the present investigation, it is observed that the response of the CdS-doped sample increases with the exposure of LPG, methanol and acetone. The electrical resistance measurement of sensors (S1 , S2 and S3 ) in air with varying temperature indicates that the resistance decreases with increase in temperature (Fig. 2).
The XRD analysis, vividly, indicated that CdS-SnO2 structure lowers the crystallinity of SnO2 . The lowering in crystallinity, which shows fine grain size and, thereby, enhanced the response of sensor. All the sensors showed better result at a relatively low temperature 200 ◦ C. Moreover, this optimized temperature (200 ◦ C) peculiarly depends on preparing conditions and variant compositions of paste. The sensor S3 (2 wt% CdS-SnO2 ) exhibited maximum response and is more selective to methanol over LPG and acetone. The response and recovery time also reduces considerably with CdS-doping. The present study demonstrates that CdS (2 wt%) doped SnO2 sensor is a most suitable detector for methanol. Acknowledgement We are thankful to HOD, Department of Electronics Engineering, Institute of Technology, Banaras Hindu University, Varanasi, India for providing experimental facility during fabrication of sensors. We are also thankful to Mr. Shiv Kumar Singh and Superconductivity and Cryogenics Division, National Physical Laboratory, New Delhi, India for providing XRD measurements. References
Fig. 9. Transient response and recovery time of sensor S1 and S3 at 200 ◦ C for LPG (5000 ppm).
[1] J. Watson, The tin oxide gas sensor and its applications, Sens. Actuators B 5 (1984) 29–42. [2] D. Kohl, Surface processes in the detection of reducing gases with SnO2 -based devices, Sens. Actuators B (1989) 71–116. [3] D. Kohl, The role of noble metals in the chemistry of solid-state gas sensors, Sens. Actuators B (1990) 158–165. [4] G. Sberveglieri, E. Comini, G. Faglia, M.Z. Atashbar, W. Wlodarsk, Titanium dioxide thin films prepared for alcohol microsensor applications, Sens. Actuators B 66 (2000) 139–141. [5] A. Teeramongkonrasmee, M. Sriyudthsak, Methanol and ammonia sensing characteristics of sol–gel derived thin film gas sensor, Sens. Actuators B 66 (2000) 256–259. [6] S.M. Chou, L.G. Teoh, W.H. Lai, Y.H. Su, M.H. Hon, ZnO:Al thin film gas sensor for detection of ethanol vapor, Sensors 6 (2006) 1420–1427. [7] P. Ivanov, J. Hubalek, K. Malysz, J. Parsek, X. Vilonova, E. Llobat, X. Correig, A route toward more selective and less humidity sensitive screen printed SnO2 and WO3 gas sensitive layers, Sens. Actuators B 100 (2004) 221–227. [8] Y. Mo, Y. Okawa, T. Nakai, M. Tajima, K. Natukawa, Preparation of SnO2 films with high sensitivity and selectivity to C2 H5 OH by oxygen radical assisted electron beam evaporation for micro-machined gas sensors, Thin Solid Films 416 (2001) 248–253. [9] Z. Tianshua, P. Hinga, Z. Jiancheng, K. Lingbing, Ethanol-sensing characteristics of cadmium ferrite prepared by co-precipitation, Mater. Chem. Phys. 61 (1999) 192–196.
42
L. Yadava et al. / Sensors and Actuators B 144 (2010) 37–42
[10] W. Xing-Hui, W. Yu-De, L. Yan-Feng, Z. Zhen-Lai, Electrical and gas-sensing properties of perovskite-type CdSnO3 semiconductor material, Mater. Chem. Phys. 77 (2002) 588–593. [11] W. Yude, S. Xiaodan, L. Yenfeng, Z. Zhenlai, W. Xinghui, Perovskite-type NiSnO3 used as ethanol sensitive material, Solid State Electron. 44 (2000) 2009– 2014. [12] G. Sberveglieri, L.E. Depero, A novel method for preparation of nanosized TiO2 thin films, J. Adv. Mater. 8 (4) (1996) 334–337. [13] N. Yamazoe, N. Miura, Some basic aspects of semiconductor gas sensor, in: S. Yamauchi (Ed.), Chemical Sensor Technology, vol. 4, Elsevier, Amsterdam, 1992. [14] N. Yamazoe, Y. Kurokuwa, T. Seiyama, Effect of additives on semiconductor gas sensors, Sens. Actuators B 4 (1983) 283. [15] R.H.R. Castro, P. Hidalgo, R. Muccillo, D. Gouvea, Microstructure and structure of NiO–SnO2 and Fe2 O3 systems, Appl. Surf. Sci. 214 (2003) 172. [16] K. Jain, R.P. Pant, S.T. Lakshmikumar, Effect of Ni doping on thick film SnO2 gas sensor, Sens. Actuators B 113 (2006) 823–829. [17] Z. Tianshu, P. Hing, Y. Li, Z. Jiancheng, Selective detection of ethanol vapor and hydrogen using Cd-doped SnO2 -based sensors, Sens. Actuators B 60 (1999) 208–215. [18] K. Arshak, I. Gaidan, Gas sensing properties of ZnFe2 O4 /ZnO screen-printed thick films, Sens. Actuators B 111–112 (2005) 58–62. [19] W.J.S.Y. Matsumura, Low temperature methanol decomposition to carbon monoxide and hydrogen catalyzed over cationic palladium species in Pd/CeO2 , Phys. Chem. Chem. Phys. 2 (2000) 1519–1522. [20] B. Miremadi, T. Yadav, J.Z. Zhang, J.L. Falconer, Low temperature hydrogen production using electrically activated catalysts, Chem. Commun. (2000) 1875–1876.
Biographies Lallan Yadava received his B.Sc. (Honors) in Physics 1982 and M.Sc. with specialization electronics in 1985 from Banaras Hindu University, Varanasi. He obtained his Ph.D. degree in Physics from Department of Electronics Engineering, Institute of Technology, Banaras Hindu University, Varanasi (India) in the thrust area of MOS Gas Sensors. He has been continuously engaged in research in the area of Thin/Thick film gas sensors. Presently he is working as a Reader in the Department of Physics, Deen Dayal Upadhyaya Gorakhpur University, Gorakhpur (India). He has more than 22 years of research experience and published more than 25 papers in various International/National journals and in proceeding of symposia. He is currently working in the area of gas sensors based on Silicon and Thick film technology. Ritesh Verma received his B.Sc. (Physics) in 2001 and M.Sc. (2003) in Solid State Physics from Deen Dayal Upadhyaya Gorakhpur University, Gorakhpur (India). He is currently pursuing Ph.D. degree in the thrust area of MOS/Thick film gas sensor from the Department of Physics, Deen Dayal Upadhyaya Gorakhpur University, Gorakhpur (India). He is currently working on the fabrication of tin oxide-based thick film sensors to detect gases/organic vapors. R. Dwivedi was born in 1952. He obtained Ph.D. degree in Electronics Engineering in 1978 from Banaras Hindu University, Varanasi (India). Presently, he is working as an Associate Professor in the Department of Electronics Engineering, Institute of Technology, Banaras Hindu University, and Varanasi (India). He has over 75 research papers published in various International/National journals and in proceeding of symposia. He is currently working in the area of sensors based on Silicon and Thick film technologies, MOS devices, photo voltaic and LSI/VLSI.
Contents lists available at ScienceDirect
Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb
Sensing properties of CdS-doped tin oxide thick film gas sensor Lallan Yadava a,∗ , Ritesh Verma a , R. Dwivedi b a b
Thin Film Laboratory, Department of Physics, Deen Dayal Upadhyaya Gorakhpur University, Gorakhpur 273009, India Centre for Research in Microelectronics Engineering, Department of Electronics Engineering, Institute of Technology, Banaras Hindu University, Varanasi 221005, India
a r t i c l e
i n f o
Article history: Received 25 January 2009 Received in revised form 3 October 2009 Accepted 7 October 2009 Available online 6 November 2009 Keywords: Tin oxide CdS-doping Screen-printing Methanol sensor
a b s t r a c t The tin oxide (SnO2 ) thick film gas sensor is fabricated by employing screen-printing technology. This pure SnO2 thick film is doped with 1 or 2 wt% of cadmium sulphide (CdS) by its weight and, thereby, the effect of dopant is presented. X-ray diffraction (XRD) analyses are administrated, which suggest that CdS dopant inhibits the crystallite growth leading to nanometric reduction in grain size. The fabricated gas sensor is responsively studied on exposure to liquid petroleum gas (LPG), methanol, and acetone. It is observed that CdS (2 wt%) doped structure exhibited highest response and is more selective to methanol (70 for 5000 ppm) over LPG and acetone at the operating temperature 200 ◦ C. The CdSdoping improved response- and recovery-time from 90 s and 200 s, for undoped-film, to 40 s and 110 s for methanol (5000 ppm at 200 ◦ C). © 2009 Elsevier B.V. All rights reserved.
1. Introduction Unlike traditional chemical detection of noxious gases, there is inexorable shift towards designing and fabricating electronic sensors (structures). Amongst various kind of sensors, the chemical gas sensors, recently, received surge of interests on account of their better sensitivity and selectivity for inflammable, combustible and pollutant gases/odors [1]. Semiconducting SnO2 films are extensively used for sensing and discrimination of several reducing gases [2]. The noble metals such as Pt, Pd, and Ni are utilized as dopants in SnO2 thick/thin film to enhance its sensitivity, selectivity and to improve response and recovery times [3]. It is investigated that adsorbed oxygen gas molecules on the surface modify the conductivity of SnO2 thick film layer. Moreover, these oxygen species reacts with exposing (reducing) gases resulting in enhancement of conductance. Other oxides such as ZnO, TiO2 , Fe2 O3 , WO3 , ZrO3 , and V2 O5 are also applied to detect gases/odors [4–11]. For a film, structural stability, porosity and large surface-volume ratio are salient factors in determining its suitability in gas-sensor technology. It is established fact that reduction in grain size of a material increases sensitivity [12,13]. Yamazoe et al. [14] studied additives’ effects in semiconductor gas sensors. Dopant like CaO, MgO are seen to inhibit grain growth and, thereby, increasing gas-sensitivity. Other additives such as ZnO, CuO, and MnO affect conductivity of SnO2 considerably leading into reduction of resistivity. Castro et al. [15] reported the effect of nickel oxide doping in
∗ Corresponding author. Tel.: +91 9450253084. E-mail address: [email protected] (L. Yadava). 0925-4005/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.snb.2009.10.013
SnO2 , which resulted maximum sensitivity for ethyl alcohol. Jain et al. [16] studied the effect of Ni and Al doping on a screen-printed SnO2 thick film and showed that Ni/Al doping results the broadening in the XRD peaks, thus lowers the grain size and improved the gas sensitivity for LPG. Tianshu et al. [17] investigated Cd-doped SnO2 -based sensor for the detection of ethanol and hydrogen. He reported that CdO doping suppresses SnO2 crystallite growth effectively. ZnO, Fe2 O3 are used by Arshak and Gaidan [18] and the structure is reported to be appropriate detector for methanol, ethanol and propanol. In the present study, we investigated the role and effect of CdS on SnO2 thick film. The undoped and doped structure is analyzed and its suitability for detection of LPG, methanol and acetone is studied. In Section 2, the preparation of SnO2 paste and fabrication of undoped and doped thick film sensors are described. The measurement results, its discussions and mechanism responsible for sensing the incoming gases are presented in Section 3 and Section 4 concludes the findings.
2. Experimental 2.1. Preparation of tin oxide paste Doped and undoped paste of SnO2 is prepared in the laboratory. The tin oxide (SnO2 ) powder and glass binder (10 wt% of SnO2 ) are weighed using electronic balance. Taking SnO2 powder along with glass binder in the ball mill (Zirconia Ball Mill, Retsch), a mixture is obtained after 3–4 h of processing. Fine grains are, then, mixed with organic binder (diethyl glycol monobutyl) and organic solvent (␣-terpinol) in ball mill for 1–2 h, which results an undoped SnO2
38
L. Yadava et al. / Sensors and Actuators B 144 (2010) 37–42
paste. For doped pastes of SnO2 , weighed SnO2 powder with glass binder (10 wt% of SnO2 ) and CdS powder (1 wt%, 2 wt%), all these mixed in ball mill and same organic binder and solvent are used to get the sensing paste. To optimize the firing temperature of the paste, the prepared paste is screen-printed on alumina substrate and then dried followed by fired at different temperatures in the central zone of thick film furnace. The drying temperature is kept at 150 ◦ C for 15 min and optimized firing temperature at 800 ◦ C for 30 min.
2.2. Fabrication of sensor The sensors are fabricated in the form of rectangular (8 mm × 10 mm) thick film deposited on the alumina substrate (25 mm × 25 mm) by screen-printing technology. The utilized substrate is, initially, screen-printed with a finger electrode pattern in rear side and a resistor heating pattern in back side. Three kinds of sensing pastes are screen-printed on designed alumina substrate to make the sensors named as S1 (undoped SnO2 ), S2 (1 wt% CdS-SnO2 ) and S3 (2 wt% CdS-SnO2 ). After each screen-printing the film is dried at 150 ◦ C for 15 min followed by firing at 800 ◦ C for 30 min, as a result of which accurate adherence of active layer to the substrate is obtained.
3. Results and discussion 3.1. X-ray diffraction (XRD) The fabricated sensors S1 , S2 and S3 respectively are analyzed by administrating XRD (Fig. 1). The XRD was performed by RIGAKU MINIFLEX-2 using Cu K␣ radiation of wavelength 1.5408 Å. All the main XRD lines belong to SnO2 with casserite structure and are indexed accordingly. The broadening in the XRD peaks shows the doping of CdS significantly reduces the crystallinity of SnO2 . The crystallite size (R) of SnO2 was evaluated by fitting the (1 0 1) diffraction peak width of SnO2 using Scherrer equation, as—R = 0.9 /( cos ), where is the wavelength of X-ray used, is the diffraction angle, ˇ is the peak full line width at half maximum (FWHM). The crystallite size of undoped and doped SnO2 with CdS composition has been listed in Table 1. It may be seen that CdS act as a growth inhibitor for SnO2 crystallization.
Fig. 1. XRD pattern of (a) S1 (undoped SnO2 ), (b) S2 (1 wt% CdS-SnO2 ) and (c) S3 (2 wt% CdS-SnO2 ) sensors.
Table 1 Crystallite size and composition of samples. Sensor no.
S1
S2
S3
Composition CdS-SnO2 (wt%) Crystallite size, R (nm)
0 50.9
1 30.2
2 24.8
3.2. Resistance measurement and gas sensitivity The sensor heater (heating resistor pattern) has been optimized and calibrated which implies that how much electric power needed to sensor for a desired operating temperature. The electric power consumed to get heating the pattern depends on various parameters like, resistance of heating material, structure and area of pattern also. Fig. 2 shows the variation of resistance of the sensors with temperature in air. It is evident that the resistance decreases on increasing device temperature. Moreover, the reduction in resistance is visible on CdS-doped sensing film. We regarded ‘percentage change in resistance’ to a particular gas concentration as: S=
Ra − Rg × 100 Ra
where Ra = resistance of sensor in air, Rg = resistance of sensor in the presence of gas exposed, the response-parameter. The response of the fabricated sensors (S1 , S2 and S3 ) is measured with exposure to LPG, methanol and acetone concentration (0–5000 ppm) at different operating temperatures in air ambient. The response with varying concentration of methanol at operating temperatures 150 ◦ C, 200 ◦ C and 250 ◦ C for sensor S1 is shown in Fig. 3(a). It is evident from the figure that, as the concentration increases, response increases and it is maximum (18) at 200 ◦ C. The sensing behavior of S2 and S3 sensor for methanol vapor at various operating temperature 150 ◦ C, 200 ◦ C and 250 ◦ C are reported in Fig. 3(b) and (c). It is seen that maximum response to methanol (5000 ppm) for sample S2 is 44 (2.5 times of S1 ) and for sensor S3 is 70 (4 times of S1 ) at 200 ◦ C temperature. The operating temperature has a considerable influence on the response of the sensor. As the operating temperature increases from 150 ◦ C to 200 ◦ C, the response also enhances for each sensor, however above 200 ◦ C, the response decreases. The measurements reveal that the fabricated sensors respond maximally at operating temperature 200 ◦ C. The responses of fabricated sensors S1 , S2 and S3 for acetone and LPG at 200 ◦ C temperature are, respectively, presented in Figs. 4 and 5. Fig. 4 shows the response increases with increasing acetone concentration and its maximum values are 44, 51 and 56 (at 5000 ppm) for sensors S1 , S2 and S3 individually. These measurement reflects that doping of CdS (2 wt%) to SnO2 enhanced the response to acetone 1.3 times of S1 . Fig. 5 indi-
Fig. 2. Resistance vs. temperature for sensor S1 , S2 and S3 .
L. Yadava et al. / Sensors and Actuators B 144 (2010) 37–42
39
Fig. 3. The response of the sensor as function of methanol concentration at different operating temperatures for (a) sensor S1 , (b) sensor S2 and (c) sensor S3 .
cates that the maximum response to LPG (5000 ppm) for sensors S1 , S2 and S3 are 13, 45 and 49, respectively. It is observed that sensor S3 showed maximum response to LPG, which is 3.7 times that of S1 . Fig. 6 represents a comparative responses of the sensor S3 for varying concentration of methanol, LPG and acetone at fixed temperature 200 ◦ C. The ratio of maximum responses for sensor S3 to S1 is 4, 1.3 and 3.7 times for methanol, acetone and LPG relatively. The doping of CdS (2 wt%) caused a remarkable improvement in the response and is more selective to methanol over LPG and acetone. The transient response and recovery curve of sensor S1 and S3 for methanol at 200 ◦ C temperature are plotted in Fig. 7. At time 10 min, the 5000-ppm methanol was injected into test chamber
in air ambient and after getting maximum response the chamber was opened for recovery of initial state of the device. Moreover, the response and recovery time is evaluated by the time taken to get 90% of maximum response and to recover the 90% of starting value, respectively. It is seen from Fig. 7 that doping of CdS into SnO2 films shows fast response (40 s) and recovery time (110 s) by reducing them, respectively, 50 s and 90 s from 90 s and 200 s for methanol. The transient response and recovery curve to acetone and LPG for sensor S1 and S3 are shown in Figs. 8 and 9. The measurement shows that response and recovery time decreases by 30 s and 70 s for acetone however, in case of LPG, response time reduces almost 60 s and recovery time is less affected.
40
L. Yadava et al. / Sensors and Actuators B 144 (2010) 37–42
Fig. 4. Response vs. acetone concentration for sensor S1 , S2 and S3 at 200 ◦ C operating temperature.
3.3. Gas sensing properties The sensing mechanism of metal oxide semiconductor has been explained on the basis of adsorbed oxygen, O2, on the semiconductor surface. The chemisorbed oxygen molecules remove electrons (e− ) from the conduction band and, thereby, increasing resistance of SnO2 thick film. This process may be, schematically, shown as follows: 1 O 2 2
+ e− + S → O− − S
(1)
Fig. 5. Response vs. LPG concentration for sensor S1 , S2 and S3 at 200 ◦ C operating temperature.
Fig. 6. Variation of response of the sensor S3 (2 wt% CdS-SnO2 ) as function of methanol, LPG and acetone concentration at fixed 200 ◦ C temperature.
where S is an adsorption site on sensing surface. However, during chemisorptions, O2− and O2 − also appear along with O− depending upon the temperature of metal oxides [2]. Among these anions, O2 − is treated as ‘electrophilic’ agent while O2− associated with the lattice on the surface, as a ‘nucleophilic’ agent. Following to Kohl [2,3], O2− species are highly unstable and do not play any role in determining the response. Hence among the adsorbed species, O− plays dominant role in influencing the response of thick film. This dominant adsorbed oxygen species (O− ) (Eq. (1)) creates a space-charge region near the film surface. The resistance of SnO2 thick film, loaded with species O− , at
Fig. 7. Transient response and recovery time of sensor S1 and S3 at 200 ◦ C for methanol (5000 ppm).
L. Yadava et al. / Sensors and Actuators B 144 (2010) 37–42
41
The change in resistance is attributed to the change in barrier height, at grain boundaries, due to modulating surface charge densities which depends on temperature as well as on exposure of gases. Moreover, the XRD measurement of the samples (S1 –S3 ) shows broadening in peaks and clear change in relative intensity (200 peaks) on CdS-doping. The crystallite size of SnO2 is evaluated using Scherrer equation and it is found that the same decreases from 50.9 nm to 24.8 nm on doping. Since the crystallite size falls within the 100 nm thus, we can say that average grain size lies in nanometric range. Thus, CdS-dopant acts as a crystal growth inhibitor and improves the response of sensor. It is, earlier, reported that dopant like Ni and CdO in tin oxide acts as a crystal growth inhibitor which improves sensor response on exposing various gases/vapors [16,17]. Obviously, from the XRD measurement, we may conclude that the CdS-doping lowers the average grain size, therefore, responsible for the increase in sensor response. Future work is to be done to get a definite understanding of reaction mechanism on CdS-SnO2 surface exposed with hydrocarbon vapors. 4. Conclusion
Fig. 8. Transient response and recovery time of sensor S1 and S3 at 200 ◦ C for acetone (5000 ppm).
higher temperature gets decreased on exposure of reducing gases like LPG, methanol and acetone and enhancing the response of the sensor. It is reported that methanol can be decomposed at lower temperature (150–200 ◦ C) using different catalytic metal oxides [19,20]. The highest response of the sensor may be accounted for liberation of more free electrons, when species O− , possibly reacts with methanol as follows: 2CH3 OH + 2[O− ] → CO + CO2 + 3H2 + H2 O + 2e−
(2)
In the present investigation, it is observed that the response of the CdS-doped sample increases with the exposure of LPG, methanol and acetone. The electrical resistance measurement of sensors (S1 , S2 and S3 ) in air with varying temperature indicates that the resistance decreases with increase in temperature (Fig. 2).
The XRD analysis, vividly, indicated that CdS-SnO2 structure lowers the crystallinity of SnO2 . The lowering in crystallinity, which shows fine grain size and, thereby, enhanced the response of sensor. All the sensors showed better result at a relatively low temperature 200 ◦ C. Moreover, this optimized temperature (200 ◦ C) peculiarly depends on preparing conditions and variant compositions of paste. The sensor S3 (2 wt% CdS-SnO2 ) exhibited maximum response and is more selective to methanol over LPG and acetone. The response and recovery time also reduces considerably with CdS-doping. The present study demonstrates that CdS (2 wt%) doped SnO2 sensor is a most suitable detector for methanol. Acknowledgement We are thankful to HOD, Department of Electronics Engineering, Institute of Technology, Banaras Hindu University, Varanasi, India for providing experimental facility during fabrication of sensors. We are also thankful to Mr. Shiv Kumar Singh and Superconductivity and Cryogenics Division, National Physical Laboratory, New Delhi, India for providing XRD measurements. References
Fig. 9. Transient response and recovery time of sensor S1 and S3 at 200 ◦ C for LPG (5000 ppm).
[1] J. Watson, The tin oxide gas sensor and its applications, Sens. Actuators B 5 (1984) 29–42. [2] D. Kohl, Surface processes in the detection of reducing gases with SnO2 -based devices, Sens. Actuators B (1989) 71–116. [3] D. Kohl, The role of noble metals in the chemistry of solid-state gas sensors, Sens. Actuators B (1990) 158–165. [4] G. Sberveglieri, E. Comini, G. Faglia, M.Z. Atashbar, W. Wlodarsk, Titanium dioxide thin films prepared for alcohol microsensor applications, Sens. Actuators B 66 (2000) 139–141. [5] A. Teeramongkonrasmee, M. Sriyudthsak, Methanol and ammonia sensing characteristics of sol–gel derived thin film gas sensor, Sens. Actuators B 66 (2000) 256–259. [6] S.M. Chou, L.G. Teoh, W.H. Lai, Y.H. Su, M.H. Hon, ZnO:Al thin film gas sensor for detection of ethanol vapor, Sensors 6 (2006) 1420–1427. [7] P. Ivanov, J. Hubalek, K. Malysz, J. Parsek, X. Vilonova, E. Llobat, X. Correig, A route toward more selective and less humidity sensitive screen printed SnO2 and WO3 gas sensitive layers, Sens. Actuators B 100 (2004) 221–227. [8] Y. Mo, Y. Okawa, T. Nakai, M. Tajima, K. Natukawa, Preparation of SnO2 films with high sensitivity and selectivity to C2 H5 OH by oxygen radical assisted electron beam evaporation for micro-machined gas sensors, Thin Solid Films 416 (2001) 248–253. [9] Z. Tianshua, P. Hinga, Z. Jiancheng, K. Lingbing, Ethanol-sensing characteristics of cadmium ferrite prepared by co-precipitation, Mater. Chem. Phys. 61 (1999) 192–196.
42
L. Yadava et al. / Sensors and Actuators B 144 (2010) 37–42
[10] W. Xing-Hui, W. Yu-De, L. Yan-Feng, Z. Zhen-Lai, Electrical and gas-sensing properties of perovskite-type CdSnO3 semiconductor material, Mater. Chem. Phys. 77 (2002) 588–593. [11] W. Yude, S. Xiaodan, L. Yenfeng, Z. Zhenlai, W. Xinghui, Perovskite-type NiSnO3 used as ethanol sensitive material, Solid State Electron. 44 (2000) 2009– 2014. [12] G. Sberveglieri, L.E. Depero, A novel method for preparation of nanosized TiO2 thin films, J. Adv. Mater. 8 (4) (1996) 334–337. [13] N. Yamazoe, N. Miura, Some basic aspects of semiconductor gas sensor, in: S. Yamauchi (Ed.), Chemical Sensor Technology, vol. 4, Elsevier, Amsterdam, 1992. [14] N. Yamazoe, Y. Kurokuwa, T. Seiyama, Effect of additives on semiconductor gas sensors, Sens. Actuators B 4 (1983) 283. [15] R.H.R. Castro, P. Hidalgo, R. Muccillo, D. Gouvea, Microstructure and structure of NiO–SnO2 and Fe2 O3 systems, Appl. Surf. Sci. 214 (2003) 172. [16] K. Jain, R.P. Pant, S.T. Lakshmikumar, Effect of Ni doping on thick film SnO2 gas sensor, Sens. Actuators B 113 (2006) 823–829. [17] Z. Tianshu, P. Hing, Y. Li, Z. Jiancheng, Selective detection of ethanol vapor and hydrogen using Cd-doped SnO2 -based sensors, Sens. Actuators B 60 (1999) 208–215. [18] K. Arshak, I. Gaidan, Gas sensing properties of ZnFe2 O4 /ZnO screen-printed thick films, Sens. Actuators B 111–112 (2005) 58–62. [19] W.J.S.Y. Matsumura, Low temperature methanol decomposition to carbon monoxide and hydrogen catalyzed over cationic palladium species in Pd/CeO2 , Phys. Chem. Chem. Phys. 2 (2000) 1519–1522. [20] B. Miremadi, T. Yadav, J.Z. Zhang, J.L. Falconer, Low temperature hydrogen production using electrically activated catalysts, Chem. Commun. (2000) 1875–1876.
Biographies Lallan Yadava received his B.Sc. (Honors) in Physics 1982 and M.Sc. with specialization electronics in 1985 from Banaras Hindu University, Varanasi. He obtained his Ph.D. degree in Physics from Department of Electronics Engineering, Institute of Technology, Banaras Hindu University, Varanasi (India) in the thrust area of MOS Gas Sensors. He has been continuously engaged in research in the area of Thin/Thick film gas sensors. Presently he is working as a Reader in the Department of Physics, Deen Dayal Upadhyaya Gorakhpur University, Gorakhpur (India). He has more than 22 years of research experience and published more than 25 papers in various International/National journals and in proceeding of symposia. He is currently working in the area of gas sensors based on Silicon and Thick film technology. Ritesh Verma received his B.Sc. (Physics) in 2001 and M.Sc. (2003) in Solid State Physics from Deen Dayal Upadhyaya Gorakhpur University, Gorakhpur (India). He is currently pursuing Ph.D. degree in the thrust area of MOS/Thick film gas sensor from the Department of Physics, Deen Dayal Upadhyaya Gorakhpur University, Gorakhpur (India). He is currently working on the fabrication of tin oxide-based thick film sensors to detect gases/organic vapors. R. Dwivedi was born in 1952. He obtained Ph.D. degree in Electronics Engineering in 1978 from Banaras Hindu University, Varanasi (India). Presently, he is working as an Associate Professor in the Department of Electronics Engineering, Institute of Technology, Banaras Hindu University, and Varanasi (India). He has over 75 research papers published in various International/National journals and in proceeding of symposia. He is currently working in the area of sensors based on Silicon and Thick film technologies, MOS devices, photo voltaic and LSI/VLSI.