- Email: [email protected]
CO gas sensing properties of direct-patternable TiO2 thin films containing multi-wall carbon nanotubes
Thin Solid Films 529 (2013) 89–93
Contents lists available at SciVerse ScienceDirect
Thin Solid Films journal homepage: www.elsevier.com/locate/tsf
CO gas sensing properties of direct-patternable TiO2 thin films containing multi-wall carbon nanotubes Hyuncheol Kim a, Min-Hee Hong a, Ho Won Jang b, Seok-Jin Yoon b, Hyung-Ho Park a,⁎ a b
Department of Materials Science and Engineering, Yonsei University, Seoul 120‐749, Republic of Korea Electronic Materials Center, KIST, Seoul 130‐650, Republic of Korea
a r t i c l e
i n f o
Available online 22 July 2012 Keywords: MWCNTs Direct-patterning TiO2 thin films Electron conductivity Surface area Gas sensor
a b s t r a c t The sensing properties of direct-patternable TiO2 thin films prepared by photochemical solution deposition were improved by the incorporation of multi-wall carbon nanotubes (MWCNTs). Upon incorporation of the MWCNTs, the crystallinity of the TiO2 thin films containing MWCNTs did not change. However, the electron conductivity was improved due to the π-bond nature of the MWCNT surface. In addition, an increase in the oxidation ratio on the surface of the TiO2 thin films upon MWCNT incorporation caused an increase in surface area. Because of the high electron conductivity and surface area of the MWCNT-incorporated TiO2 sensor, the sensitivity of the sensor was enhanced from 2.19 to 89.2. Direct-patterning of MWCNT-incorporated TiO2 thin films can be performed without a photoresist or etching process. These results suggest that a micro-patterned sensor can be simply fabricated at a low cost and the sensitivity of TiO2 thin films can be improved by incorporating MWCNTs. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Metal oxide semiconductor based gas sensors have attracted considerable attention owing to their low cost, high sensitivity, and high compatibility with microelectronic processing. Titanium dioxide (TiO2) was a widely used semiconductor material for gas sensor applications [1]. CO gas detection in the anatase TiO2 phase can be attributed to a surface reaction, which takes place at lower temperatures (under 500 °C) [2]. When O2 molecules are adsorbed on the surface of a TiO2 thin film, they trap the electrons at the surface in the form of ions. On the contrary, when a gas sensor is exposed to CO gas, CO is oxidized by O − and releases electrons to the surface. O− is believed to be dominant at operating temperatures of 300–450 °C [3]. Investigation about hybridization using carbon nanotubes (CNTs) has been increasing in recent studies. Multi-wall and single-wall carbon nanotubes (MWCNTs and SWCNTs, respectively) are chemically inert, exhibit excellent electrical and thermal conductivity, and have superior mechanical strength [4]. CNTs have a good potential for detecting gases due to their large surface area resulting from big aspect ratio and outer walls [5]. As a result, MWCNTs have been employed in semiconductor gas sensors. A micro-patterning process is inevitably introduced for the fabrication of gas sensor devices [6]. The conventional etching process is
⁎ Corresponding author at: Department of Materials Science and Engineering, Yonsei University, 134 Sinchon-dong, Seodaemun-gu, Seoul 120‐749, Republic of Korea. Tel.: +82 2 2123 2853; fax: +82 2 312 5375. E-mail address: [email protected] (H.-H. Park). 0040-6090/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.tsf.2012.07.062
accompanied by physical defect generation, property degradation, and pollution due to use of hazardous-material, and so on. However, if we use photosensitive additives, neither a photoresist nor dry etching is required for micro-scaled patterning [7]. Therefore, we can avoid many problems associated with conventional etching process. The effects of MWCNTs on the gas sensing properties of TiO2 thin films were investigated in this study. Through MWCNT incorporation, the gas sensitivity of TiO2 thin film gas sensors was improved without significant differences in response and recovery times. Furthermore, direct-patterning of MWCNT-incorporated TiO2 thin films was achieved by using a photosensitive solution and UV light exposure to avoid damage from dry etching and to simplify the micro-scaled patterning procedure. 2. Experimental details TiO2 thin films containing MWCNTs were prepared by a photochemical solution deposition process. The Ti precursor and solvent used for TiO2 thin films were titanium (IV) isopropoxide (C12H28O4Ti) and 1-propanol (C3H8O), respectively. Titanium (IV) isopropoxide was dissolved in 1-propanol and 2-nitrobenzaldehyde (NBAL), which was introduced into the solution as a photosensitive additive. MWCNTs (CM-95, Hanwha Nanotech) were used as dopant. To obtain MWCNTincorporated TiO2 thin films, 0.04, 0.055, 0.07, 0.105, 0.14, and 0.29 wt.% MWCNTs were incorporated into the TiO2 photosensitive solution. The MWCNTs in the solution were dispersed using a sonicator and the mixed solution was stirred at room temperature for 24 h. The solution of MWCNT-incorporated TiO2 was spin-coated at 2000 rpm for
90
H. Kim et al. / Thin Solid Films 529 (2013) 89–93
20 s onto Corning 1737 glass substrates. The spin-coated films were exposed to 365 nm UV light for direct-patterning. Then, the UV exposed films were washed with methanol (CH3OH) to remove the unexposed area of the film. After washing, the films were aged at 50 °C for 12 h in a dry oven and annealed at 450 °C for 1 h in a tubular furnace under an air atmosphere to remove the solvent and organic residues. Finally, the films were annealed in a vacuum furnace at 450 °C for 12 h. The crystallinity was analyzed using an X-ray diffractometer (XRD, D/MAX-2000, Rigaku) with Cu Kα radiation. Scan was performed using theta/2theta (θ/2θ) method in the procedure of 0.05° step with a 4°/min scan speed. The final thickness of the film was measured using ellipsometry. The surface roughness and morphology of the films were analyzed by atomic force microscopy in non-contact mode (AFM, XE-100, Park Systems) with a base line noise of 0.01 nm and scanning electron microscope (SEM, JSM-6390, Jeol). The electrical properties of the film were measured using a Hall effect measurement (HMS-3000, Ecopia). To investigate the surface chemical bonding state of the films, X-ray photoelectron spectroscopy (XPS, ESCALAB 220i-XL, VG Scientific) was used with an Al Kα source. The accelerating voltage and emission current of X-ray source were 15 kV and 20 mA, respectively. Avantage 3.25 was used for peak fitting software supplied by XPS manufacturer. The binding energies were corrected using the reference C 1s peak at 284.5 eV [8]. A Shirley subtraction and Gaussian shape were used for background subtraction and peak fit analysis, respectively. During peak fit analysis of measured spectra, the parameters such as binding energy and FWHM were constrained refer to the published references [9–11]. A gas sensor for sensing CO gas was fabricated using a SiO2/Si substrate with Pt interdigitated electrodes (IDE) in which the gap between each electrode was 5 μm. The thickness of Pt was 200 nm and the IDE patterns were fabricated using photolithography and dry etching. The responses of the fabricated gas sensors to CO gas were measured at 400 °C by monitoring the change in sensor resistances while changing the flow gas from dry air to test gases (100 ppm CO balanced with dry air). To eliminate interfering effects, we used a constant flow rate of 1000 sccm for dry air and the test gases. The film resistance was measured under a DC bias voltage of 3 V using a source measurement unit (Keithley 2635a). 3. Results and discussion Fig. 1 presents the XRD patterns of TiO2 thin films containing various amounts of MWCNTs annealed at 450 °C. All the films were indexed as (101), (004), and (200) diffraction peaks of the anatase TiO2 crystalline phase [12]. When the MWCNTs were incorporated, similar diffraction patterns were observed. In other words, the presence of MWCNTs did not affect the crystallization of the TiO2 thin films significantly. The possibility of direct-patterning of MWCNT-incorporated TiO2 thin films was examined by removing the area unexposed to UV light using a solvent. The optical micrograph in Fig. 2 indicates a relatively dark area (greenish) corresponding to the MWCNT-incorporated TiO2 thin film and a bright area (reddish area) corresponding to the glass substrate. The photolysis reaction stage of the photosensitive additive is explained as follows. The metal ion in a metal-alkoxide is an acid and an aldehyde is base and they are strongly attracted to each other. Hydrolysis and condensation reactions are limited due to this attraction. After exposure to UV, intermolecular proton transfer and hydrogen abstraction occur, then a 2-nitrosobenzoic anion is produced. The metal ion in the metal-alkoxide formed by a condensation reaction is attracted to a nitryl ligand and a cross-linked structure is formed. This cross-linked structure is no longer in the fluid state and it is not removed during the next solvent-rinsing step. As shown at the edge of the centered pattern in Fig. 2, micron-scaled patterning was achieved by lithographic patterning using photochemical solution deposition. The gas response characteristics of gas sensors based on TiO2 thin films are shown in Fig. 3. During exposure to 100 ppm CO at 400 °C,
Fig. 1. XRD patterns of TiO2 thin films containing various amounts of MWCNTs annealed at 450 °C under vacuum.
the sensor resistance decreased, which indicates that the TiO2 thin films were n-type semiconductors. From the difference between the sensor resistance in air (Rair) and the sensor resistance in CO (RCO), we can predict the sensitivity of the sensors (sensitivity = Rair/RCO). We know that the sensitivity of the 0.07 wt.% MWCNT-incorporated TiO2 sensor is higher than the TiO2 sensor because the sensor resistance decreased dynamically compared to the TiO2 sensor when exposed to CO gas. Sensing properties such as sensitivity, 90% response time, and 90% recovery time of MWCNT-incorporated TiO2 sensor were shown in Fig. 4(a). The sensitivity, 90% response time, and 90% recovery time of the TiO2 sensor were 2.19, 5.16 s, and 2.72 s, respectively. The sensitivity of the 0.07 wt.% MWCNT-incorporated TiO2 sensor was increased to 89.2. The general gas sensing model is based on modulation of the depletion layer by oxygen absorption. When a reducing gas such as CO is exposed to the sensors, it reacts with the adsorbed oxygen molecules and releases the trapped electrons to the conduction band, thereby increasing the electron concentration and mobility of the sensors. The reasons for the increasing sensitivity were the low sensor resistance, high surface area, and larger depletion region of MWCNTs-incorporated TiO2 sensors. Because MWCNTs have a large surface to volume ratio,
Fig. 2. Optical image of direct-patterned MWCNTs incorporated into a TiO2 thin film washed with solvent.
H. Kim et al. / Thin Solid Films 529 (2013) 89–93
Fig. 3. Transient response of the TiO2 film sensor and that containing 0.7 wt.% MWCNTs measured at 400 °C with 100 ppm CO gas.
the surface area of the film was increased upon MWCNT incorporation [13]. The larger specific surface area allows more gas molecules to be absorbed on the surface of the sensing film. By incorporation of MWCNTs into TiO2 film, the depletion region was expanded than pristine TiO2 film.
(a)
(b)
Fig. 4. (a) Sensitivity (■), 90% response time ( ), and 90% recovery time ( ) of MWCNT-incorporated TiO2 sensors measured at 400 °C in dry air and 100 ppm CO gas. (b) Sensor resistance of a MWCNT-incorporated TiO2 sensor in dry air or CO gas and the resistivity, and the mobility of the MWCNT-incorporated TiO2 thin films annealed at 450 °C under vacuum atmosphere.
91
The reason of expanded depletion region was the electrons move easily through π-bonding on surface of MWCNTs [14]. From these reasons, oxidation and reduction occurring on the specific surface were increased, making the movement of the surface electrons quite rapid. The sensor resistance of TiO2 sensors was decreased with MWCNT incorporation, as shown in Fig 4(b). Because π-bonding on the surface of the MWCNTs enhances the electron mobility on the TiO2 thin films, the sensor resistance was decreased with increasing MWCNT content relative to the pristine TiO2 sensor [15]. The sensor resistance was increased after incorporating 0.07 wt.% MWCNTs due to a decrease in mobility by the percolation effect [16,17]. Because the sensor resistance changed due to the increase in electron concentration resulting from oxidation and reduction reaction facilitated by the reaction gas, the response and recovery times were increased with an increase in the difference between Rair and RCO. Increased response and recovery times would be expected for sensor with highly fluctuating resistance as observed in our experiment; however, as shown in Fig. 5(a), the response and recovery times were similar and did not change considerably with increased sensitivity. The reasons for this phenomenon were high electrical conductivity and increased surface area by MWCNT incorporation. The increase in specific surface area of the TiO2 thin film upon incorporation of MWCNTs was confirmed by the increase in surface morphology and roughness of the film observed on SEM and AFM analyses. As shown in Fig. 5, the surface roughness of the 0.29 wt.% MWCNT-incorporated TiO2 thin film was found to be about 3.2 times higher than that of the pristine TiO2 thin film. The morphology of the films was also shown in the inset SEM image of Fig. 5. From SEM and AFM results, it could be concluded that the bumps were formed by MWCNT incorporation and they increased the surface area of the films. The TiO2 and 0.07 wt.% MWCNT-incorporated TiO2 thin films were analyzed by XPS to determine the electrical bonding state of the MWCNTs within the TiO2 matrix and the results are given in Fig. 6. Both films were annealed at 400 °C for 12 h under an air atmosphere and the C 1s, O 1s, and Ti 2p core levels were determined. The atomic percentage was calculated from core level XPS spectra using atomic sensitivity factors. The atomic percentage of elements and the atomic ratio by Ti in pristine TiO2 and 0.07 wt.% MWCNT-incorporated TiO2 thin films were shown in Table 1. Ion-etching of the film surface was not executed to confirm the surface bonding state under sensing condition. The expected carbon content in the 0.07 wt.% MWCNT-incorporated TiO2 thin films was 0.46 at.%; calculated by atomic weight. However, the difference in C/Ti between TiO2 and MWCNTs-TiO2 was 0.02. The reason for the discordance between the calculated and measured values is the increased surface area as shown in the AFM observation results (Fig. 5). C 1s peaks of TiO2 and MWCNT-incorporated TiO2 thin films were fitted as shown in Fig. 6(a) [9]. The intensity of the C_O peak in the MWCNT-incorporated TiO2 thin film was higher than in the TiO2 thin film. A similar result was observed in the O 1s spectra. The O 1s peaks of TiO2 and MWCNT-incorporated TiO2 thin films were fitted as shown in Fig. 6(b), and the fitted binding energies were located at 529.6, 531.1, and 532.1 eV, respectively. The peak binding energy around 529.6 eV is attributed to oxygen in the O\Ti bonds. The other peaks, positioned at 531.1 and 532.1 eV, are ascribed to C\OH and C_O, respectively, on the surface [10]. The intensities of the C\OH and C_O peaks in the spectrum of the MWCNT-incorporated TiO2 were higher than in the spectrum for pristine TiO2 thin film. The reason for the increase was increased oxidation on the larger surface area of the film by MWCNT incorporation [18]. These results agreed well with the AFM observation results. Changes in the core level spectra of Ti 2p would indicate a chemical bond between the MWCNT and TiO2 matrix. However, as shown in Fig. 6(c), the Ti 2p core level spectra of the 0.07 wt.% MWCNT-incorporated TiO2 film were identical to those of the TiO2 film. Hence, we can conclude that no chemical bond formed between the TiO2 and MWCNTs under our experimental conditions [11]. Therefore, we can infer that electron transfer facilitated by the MWCNTs in the TiO2 matrix or vice-versa was maintained by π-bonding on the
92
H. Kim et al. / Thin Solid Films 529 (2013) 89–93
(a)
100 nm
(b)
100 nm
Fig. 5. AFM images of a pristine TiO2 thin film and a 0.29 wt.% MWCNT-incorporated TiO2 thin film. SEM image of the films was inserted in the AFM images.
MWCNT surface [19]. Lastly, as shown in Fig. 4(b), the decrease in sensor resistance after incorporation of the MWCNTs into TiO2 film is a result of enhanced charge mobility due to π-bonding of the MWCNTs. 4. Conclusions The presence of MWCNTs does not significantly affect the crystallinity of the films and similar diffraction pattern was observed with an increase in MWCNT content. Furthermore, direct-patterning of TiO2 thin films containing MWCNTs with a line width of 50 μm was performed. By AFM measurements, we ascertained that the roughness increase was due to MWCNT incorporation. The increase in the C/Ti ratio and O/Ti ratio in XPS results indicates that the surface area was increased as a result of the MWCNTs. Moreover, we can infer that the electron transfer facilitated by the MWCNTs in the TiO2 matrix or vice-versa was maintained by π-bonding on the surface of the MWCNTs. Due to a high electrical conductivity and surface area induced by the MWCNT incorporation, the sensitivity of MWCNT-incorporated TiO2 sensors was enhanced relative to a pristine TiO2 sensor.
Fig. 6. XPS spectra of (a) C 1s, (b) O 1s, and (c) Ti 2p core level of TiO2 and 0.07 wt.% MWCNT-incorporated TiO2 thin films after annealing at 400 °C under an air atmosphere.
Table 1 The atomic percentage of elements and atomic ratio by Ti in the pristine TiO2 and 0.07 wt.% MWCNTs incorporated TiO2 thin films measured by using XPS.
Pristine TiO2 MWCNTs-TiO2
C (%) (C/Ti)
O (%) (C/Ti)
Ti (%) (Ti/Ti)
13.0 (0.54) 12.1 (0.56)
63.0 (2.62) 66.4 (3.08)
24.0 (1) 21.5 (1)
H. Kim et al. / Thin Solid Films 529 (2013) 89–93
Acknowledgment This study was supported by a grant from the Fundamental R&D Program (grant no. K0004114) for Core Technology of Materials funded by the Ministry of Knowledge Economy, Republic of Korea. This work (grant no. 00045405) was supported by Business for Cooperative R&D between Industry, Academy, and Research Institute funded by the Korea Small and Medium Business Administration in 2011. References [1] A. Wisitsoraat, A. Tuantranont, E. Comini, G. Sberveglieri, W. Wlodarski, Thin Solid Films 517 (2009) 2775. [2] O. Byl, J.T. Yates Jr., J. Phys. Chem. B 110 (2006) 22966. [3] C. Wang, L. Yin, L. Zhang, D. Xiang, R. Gao, Sensors 10 (2010) 2088. [4] H. Zengin, W. Zhou, J. Jin, R. Czerw, D.W. Smith Jr., L. Echegoyen, D.L. Carroll, S.H. Foulger, J. Ballato, Adv. Mater. 14 (2002) 1480. [5] O.K. Varghese, P.D. Kichambre, D. Gong, K.G. Ong, E.C. Dickey, C.A. Grimes, Sens. Actuators B 81 (2001) 32. [6] A.M. Taurino, S. Capone, A. Boschetti, T. Toccoli, R. Verucchi, A. Pallaoro, P. Siciliano, S. Iannotta, Sens. Actuators B 100 (2004) 177.
93
[7] H. Kim, S.J. Wang, H.H. Park, H.J. Chang, H. Jeon, H.H. Park, Thin Solid Films 516 (2007) 198. [8] J. Chastain, in: Handbook of X-ray Photoelectron Spectroscopy, Perkin-Elmer Corporation Physical Electronics Division, Minnesota, 1992, p. 40. [9] D. Yang, A. Velamakanni, G. Bozoklu, S. Park, M. Stoller, R.D. Piner, S. Stankovich, I. Jung, D.A. Field, C.A. Ventrice Jr., R.S. Ruoff, Carbon 47 (2009) 145. [10] A.V. Fetisov, G.A. Kozhina, V.B. Fetisov, E.A. Pastukhov, J. Appl. Spectrosc. 78 (2011) 240. [11] J.B. Hudson, in: Surface Science: An Introduction, Butterworth-Heinemann, Stoneham, 1992, p. 38. [12] B.S. Shirke, P.V. Korake, P.P. Hankare, S.R. Bamane, K.M. Garadkar, J. Mater. Sci. Mater. Electron. 22 (2011) 821. [13] J. Zie, V.K. Varadan, Mater. Chem. Phys. 91 (2005) 274. [14] Y. Luo, J. Liu, X. Xia, X. Li, T. Fang, S. Li, Q. Ren, J. Li, Z. Jia, Mater. Lett. 61 (2007) 2467. [15] H.J. Kim, R. Koizhaiganova, T. Vasudevan, C. Sanjeeviraja, M.S. Lee, Polym. Adv. Technol. 20 (2009) 736. [16] C. Li, E.T. Thostenson, T. Chou, Appl. Phys. Lett. 91 (2007) 223114. [17] H. Li, J. Xub, Y. Zhua, X. Chen, Q. Xiang, Talanta 82 (2010) 458. [18] J. Zhu, J. Yang, Z.F. Bian, J. Ren, Y.M. Liu, Y. Cao, H.X. Li, H.Y. He, K.N. Fan, Appl. Catal. B Environ. 76 (2007) 82. [19] Y. Tang, J. Gou, Mater. Lett. 64 (2010) 2513.
Contents lists available at SciVerse ScienceDirect
Thin Solid Films journal homepage: www.elsevier.com/locate/tsf
CO gas sensing properties of direct-patternable TiO2 thin films containing multi-wall carbon nanotubes Hyuncheol Kim a, Min-Hee Hong a, Ho Won Jang b, Seok-Jin Yoon b, Hyung-Ho Park a,⁎ a b
Department of Materials Science and Engineering, Yonsei University, Seoul 120‐749, Republic of Korea Electronic Materials Center, KIST, Seoul 130‐650, Republic of Korea
a r t i c l e
i n f o
Available online 22 July 2012 Keywords: MWCNTs Direct-patterning TiO2 thin films Electron conductivity Surface area Gas sensor
a b s t r a c t The sensing properties of direct-patternable TiO2 thin films prepared by photochemical solution deposition were improved by the incorporation of multi-wall carbon nanotubes (MWCNTs). Upon incorporation of the MWCNTs, the crystallinity of the TiO2 thin films containing MWCNTs did not change. However, the electron conductivity was improved due to the π-bond nature of the MWCNT surface. In addition, an increase in the oxidation ratio on the surface of the TiO2 thin films upon MWCNT incorporation caused an increase in surface area. Because of the high electron conductivity and surface area of the MWCNT-incorporated TiO2 sensor, the sensitivity of the sensor was enhanced from 2.19 to 89.2. Direct-patterning of MWCNT-incorporated TiO2 thin films can be performed without a photoresist or etching process. These results suggest that a micro-patterned sensor can be simply fabricated at a low cost and the sensitivity of TiO2 thin films can be improved by incorporating MWCNTs. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Metal oxide semiconductor based gas sensors have attracted considerable attention owing to their low cost, high sensitivity, and high compatibility with microelectronic processing. Titanium dioxide (TiO2) was a widely used semiconductor material for gas sensor applications [1]. CO gas detection in the anatase TiO2 phase can be attributed to a surface reaction, which takes place at lower temperatures (under 500 °C) [2]. When O2 molecules are adsorbed on the surface of a TiO2 thin film, they trap the electrons at the surface in the form of ions. On the contrary, when a gas sensor is exposed to CO gas, CO is oxidized by O − and releases electrons to the surface. O− is believed to be dominant at operating temperatures of 300–450 °C [3]. Investigation about hybridization using carbon nanotubes (CNTs) has been increasing in recent studies. Multi-wall and single-wall carbon nanotubes (MWCNTs and SWCNTs, respectively) are chemically inert, exhibit excellent electrical and thermal conductivity, and have superior mechanical strength [4]. CNTs have a good potential for detecting gases due to their large surface area resulting from big aspect ratio and outer walls [5]. As a result, MWCNTs have been employed in semiconductor gas sensors. A micro-patterning process is inevitably introduced for the fabrication of gas sensor devices [6]. The conventional etching process is
⁎ Corresponding author at: Department of Materials Science and Engineering, Yonsei University, 134 Sinchon-dong, Seodaemun-gu, Seoul 120‐749, Republic of Korea. Tel.: +82 2 2123 2853; fax: +82 2 312 5375. E-mail address: [email protected] (H.-H. Park). 0040-6090/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.tsf.2012.07.062
accompanied by physical defect generation, property degradation, and pollution due to use of hazardous-material, and so on. However, if we use photosensitive additives, neither a photoresist nor dry etching is required for micro-scaled patterning [7]. Therefore, we can avoid many problems associated with conventional etching process. The effects of MWCNTs on the gas sensing properties of TiO2 thin films were investigated in this study. Through MWCNT incorporation, the gas sensitivity of TiO2 thin film gas sensors was improved without significant differences in response and recovery times. Furthermore, direct-patterning of MWCNT-incorporated TiO2 thin films was achieved by using a photosensitive solution and UV light exposure to avoid damage from dry etching and to simplify the micro-scaled patterning procedure. 2. Experimental details TiO2 thin films containing MWCNTs were prepared by a photochemical solution deposition process. The Ti precursor and solvent used for TiO2 thin films were titanium (IV) isopropoxide (C12H28O4Ti) and 1-propanol (C3H8O), respectively. Titanium (IV) isopropoxide was dissolved in 1-propanol and 2-nitrobenzaldehyde (NBAL), which was introduced into the solution as a photosensitive additive. MWCNTs (CM-95, Hanwha Nanotech) were used as dopant. To obtain MWCNTincorporated TiO2 thin films, 0.04, 0.055, 0.07, 0.105, 0.14, and 0.29 wt.% MWCNTs were incorporated into the TiO2 photosensitive solution. The MWCNTs in the solution were dispersed using a sonicator and the mixed solution was stirred at room temperature for 24 h. The solution of MWCNT-incorporated TiO2 was spin-coated at 2000 rpm for
90
H. Kim et al. / Thin Solid Films 529 (2013) 89–93
20 s onto Corning 1737 glass substrates. The spin-coated films were exposed to 365 nm UV light for direct-patterning. Then, the UV exposed films were washed with methanol (CH3OH) to remove the unexposed area of the film. After washing, the films were aged at 50 °C for 12 h in a dry oven and annealed at 450 °C for 1 h in a tubular furnace under an air atmosphere to remove the solvent and organic residues. Finally, the films were annealed in a vacuum furnace at 450 °C for 12 h. The crystallinity was analyzed using an X-ray diffractometer (XRD, D/MAX-2000, Rigaku) with Cu Kα radiation. Scan was performed using theta/2theta (θ/2θ) method in the procedure of 0.05° step with a 4°/min scan speed. The final thickness of the film was measured using ellipsometry. The surface roughness and morphology of the films were analyzed by atomic force microscopy in non-contact mode (AFM, XE-100, Park Systems) with a base line noise of 0.01 nm and scanning electron microscope (SEM, JSM-6390, Jeol). The electrical properties of the film were measured using a Hall effect measurement (HMS-3000, Ecopia). To investigate the surface chemical bonding state of the films, X-ray photoelectron spectroscopy (XPS, ESCALAB 220i-XL, VG Scientific) was used with an Al Kα source. The accelerating voltage and emission current of X-ray source were 15 kV and 20 mA, respectively. Avantage 3.25 was used for peak fitting software supplied by XPS manufacturer. The binding energies were corrected using the reference C 1s peak at 284.5 eV [8]. A Shirley subtraction and Gaussian shape were used for background subtraction and peak fit analysis, respectively. During peak fit analysis of measured spectra, the parameters such as binding energy and FWHM were constrained refer to the published references [9–11]. A gas sensor for sensing CO gas was fabricated using a SiO2/Si substrate with Pt interdigitated electrodes (IDE) in which the gap between each electrode was 5 μm. The thickness of Pt was 200 nm and the IDE patterns were fabricated using photolithography and dry etching. The responses of the fabricated gas sensors to CO gas were measured at 400 °C by monitoring the change in sensor resistances while changing the flow gas from dry air to test gases (100 ppm CO balanced with dry air). To eliminate interfering effects, we used a constant flow rate of 1000 sccm for dry air and the test gases. The film resistance was measured under a DC bias voltage of 3 V using a source measurement unit (Keithley 2635a). 3. Results and discussion Fig. 1 presents the XRD patterns of TiO2 thin films containing various amounts of MWCNTs annealed at 450 °C. All the films were indexed as (101), (004), and (200) diffraction peaks of the anatase TiO2 crystalline phase [12]. When the MWCNTs were incorporated, similar diffraction patterns were observed. In other words, the presence of MWCNTs did not affect the crystallization of the TiO2 thin films significantly. The possibility of direct-patterning of MWCNT-incorporated TiO2 thin films was examined by removing the area unexposed to UV light using a solvent. The optical micrograph in Fig. 2 indicates a relatively dark area (greenish) corresponding to the MWCNT-incorporated TiO2 thin film and a bright area (reddish area) corresponding to the glass substrate. The photolysis reaction stage of the photosensitive additive is explained as follows. The metal ion in a metal-alkoxide is an acid and an aldehyde is base and they are strongly attracted to each other. Hydrolysis and condensation reactions are limited due to this attraction. After exposure to UV, intermolecular proton transfer and hydrogen abstraction occur, then a 2-nitrosobenzoic anion is produced. The metal ion in the metal-alkoxide formed by a condensation reaction is attracted to a nitryl ligand and a cross-linked structure is formed. This cross-linked structure is no longer in the fluid state and it is not removed during the next solvent-rinsing step. As shown at the edge of the centered pattern in Fig. 2, micron-scaled patterning was achieved by lithographic patterning using photochemical solution deposition. The gas response characteristics of gas sensors based on TiO2 thin films are shown in Fig. 3. During exposure to 100 ppm CO at 400 °C,
Fig. 1. XRD patterns of TiO2 thin films containing various amounts of MWCNTs annealed at 450 °C under vacuum.
the sensor resistance decreased, which indicates that the TiO2 thin films were n-type semiconductors. From the difference between the sensor resistance in air (Rair) and the sensor resistance in CO (RCO), we can predict the sensitivity of the sensors (sensitivity = Rair/RCO). We know that the sensitivity of the 0.07 wt.% MWCNT-incorporated TiO2 sensor is higher than the TiO2 sensor because the sensor resistance decreased dynamically compared to the TiO2 sensor when exposed to CO gas. Sensing properties such as sensitivity, 90% response time, and 90% recovery time of MWCNT-incorporated TiO2 sensor were shown in Fig. 4(a). The sensitivity, 90% response time, and 90% recovery time of the TiO2 sensor were 2.19, 5.16 s, and 2.72 s, respectively. The sensitivity of the 0.07 wt.% MWCNT-incorporated TiO2 sensor was increased to 89.2. The general gas sensing model is based on modulation of the depletion layer by oxygen absorption. When a reducing gas such as CO is exposed to the sensors, it reacts with the adsorbed oxygen molecules and releases the trapped electrons to the conduction band, thereby increasing the electron concentration and mobility of the sensors. The reasons for the increasing sensitivity were the low sensor resistance, high surface area, and larger depletion region of MWCNTs-incorporated TiO2 sensors. Because MWCNTs have a large surface to volume ratio,
Fig. 2. Optical image of direct-patterned MWCNTs incorporated into a TiO2 thin film washed with solvent.
H. Kim et al. / Thin Solid Films 529 (2013) 89–93
Fig. 3. Transient response of the TiO2 film sensor and that containing 0.7 wt.% MWCNTs measured at 400 °C with 100 ppm CO gas.
the surface area of the film was increased upon MWCNT incorporation [13]. The larger specific surface area allows more gas molecules to be absorbed on the surface of the sensing film. By incorporation of MWCNTs into TiO2 film, the depletion region was expanded than pristine TiO2 film.
(a)
(b)
Fig. 4. (a) Sensitivity (■), 90% response time ( ), and 90% recovery time ( ) of MWCNT-incorporated TiO2 sensors measured at 400 °C in dry air and 100 ppm CO gas. (b) Sensor resistance of a MWCNT-incorporated TiO2 sensor in dry air or CO gas and the resistivity, and the mobility of the MWCNT-incorporated TiO2 thin films annealed at 450 °C under vacuum atmosphere.
91
The reason of expanded depletion region was the electrons move easily through π-bonding on surface of MWCNTs [14]. From these reasons, oxidation and reduction occurring on the specific surface were increased, making the movement of the surface electrons quite rapid. The sensor resistance of TiO2 sensors was decreased with MWCNT incorporation, as shown in Fig 4(b). Because π-bonding on the surface of the MWCNTs enhances the electron mobility on the TiO2 thin films, the sensor resistance was decreased with increasing MWCNT content relative to the pristine TiO2 sensor [15]. The sensor resistance was increased after incorporating 0.07 wt.% MWCNTs due to a decrease in mobility by the percolation effect [16,17]. Because the sensor resistance changed due to the increase in electron concentration resulting from oxidation and reduction reaction facilitated by the reaction gas, the response and recovery times were increased with an increase in the difference between Rair and RCO. Increased response and recovery times would be expected for sensor with highly fluctuating resistance as observed in our experiment; however, as shown in Fig. 5(a), the response and recovery times were similar and did not change considerably with increased sensitivity. The reasons for this phenomenon were high electrical conductivity and increased surface area by MWCNT incorporation. The increase in specific surface area of the TiO2 thin film upon incorporation of MWCNTs was confirmed by the increase in surface morphology and roughness of the film observed on SEM and AFM analyses. As shown in Fig. 5, the surface roughness of the 0.29 wt.% MWCNT-incorporated TiO2 thin film was found to be about 3.2 times higher than that of the pristine TiO2 thin film. The morphology of the films was also shown in the inset SEM image of Fig. 5. From SEM and AFM results, it could be concluded that the bumps were formed by MWCNT incorporation and they increased the surface area of the films. The TiO2 and 0.07 wt.% MWCNT-incorporated TiO2 thin films were analyzed by XPS to determine the electrical bonding state of the MWCNTs within the TiO2 matrix and the results are given in Fig. 6. Both films were annealed at 400 °C for 12 h under an air atmosphere and the C 1s, O 1s, and Ti 2p core levels were determined. The atomic percentage was calculated from core level XPS spectra using atomic sensitivity factors. The atomic percentage of elements and the atomic ratio by Ti in pristine TiO2 and 0.07 wt.% MWCNT-incorporated TiO2 thin films were shown in Table 1. Ion-etching of the film surface was not executed to confirm the surface bonding state under sensing condition. The expected carbon content in the 0.07 wt.% MWCNT-incorporated TiO2 thin films was 0.46 at.%; calculated by atomic weight. However, the difference in C/Ti between TiO2 and MWCNTs-TiO2 was 0.02. The reason for the discordance between the calculated and measured values is the increased surface area as shown in the AFM observation results (Fig. 5). C 1s peaks of TiO2 and MWCNT-incorporated TiO2 thin films were fitted as shown in Fig. 6(a) [9]. The intensity of the C_O peak in the MWCNT-incorporated TiO2 thin film was higher than in the TiO2 thin film. A similar result was observed in the O 1s spectra. The O 1s peaks of TiO2 and MWCNT-incorporated TiO2 thin films were fitted as shown in Fig. 6(b), and the fitted binding energies were located at 529.6, 531.1, and 532.1 eV, respectively. The peak binding energy around 529.6 eV is attributed to oxygen in the O\Ti bonds. The other peaks, positioned at 531.1 and 532.1 eV, are ascribed to C\OH and C_O, respectively, on the surface [10]. The intensities of the C\OH and C_O peaks in the spectrum of the MWCNT-incorporated TiO2 were higher than in the spectrum for pristine TiO2 thin film. The reason for the increase was increased oxidation on the larger surface area of the film by MWCNT incorporation [18]. These results agreed well with the AFM observation results. Changes in the core level spectra of Ti 2p would indicate a chemical bond between the MWCNT and TiO2 matrix. However, as shown in Fig. 6(c), the Ti 2p core level spectra of the 0.07 wt.% MWCNT-incorporated TiO2 film were identical to those of the TiO2 film. Hence, we can conclude that no chemical bond formed between the TiO2 and MWCNTs under our experimental conditions [11]. Therefore, we can infer that electron transfer facilitated by the MWCNTs in the TiO2 matrix or vice-versa was maintained by π-bonding on the
92
H. Kim et al. / Thin Solid Films 529 (2013) 89–93
(a)
100 nm
(b)
100 nm
Fig. 5. AFM images of a pristine TiO2 thin film and a 0.29 wt.% MWCNT-incorporated TiO2 thin film. SEM image of the films was inserted in the AFM images.
MWCNT surface [19]. Lastly, as shown in Fig. 4(b), the decrease in sensor resistance after incorporation of the MWCNTs into TiO2 film is a result of enhanced charge mobility due to π-bonding of the MWCNTs. 4. Conclusions The presence of MWCNTs does not significantly affect the crystallinity of the films and similar diffraction pattern was observed with an increase in MWCNT content. Furthermore, direct-patterning of TiO2 thin films containing MWCNTs with a line width of 50 μm was performed. By AFM measurements, we ascertained that the roughness increase was due to MWCNT incorporation. The increase in the C/Ti ratio and O/Ti ratio in XPS results indicates that the surface area was increased as a result of the MWCNTs. Moreover, we can infer that the electron transfer facilitated by the MWCNTs in the TiO2 matrix or vice-versa was maintained by π-bonding on the surface of the MWCNTs. Due to a high electrical conductivity and surface area induced by the MWCNT incorporation, the sensitivity of MWCNT-incorporated TiO2 sensors was enhanced relative to a pristine TiO2 sensor.
Fig. 6. XPS spectra of (a) C 1s, (b) O 1s, and (c) Ti 2p core level of TiO2 and 0.07 wt.% MWCNT-incorporated TiO2 thin films after annealing at 400 °C under an air atmosphere.
Table 1 The atomic percentage of elements and atomic ratio by Ti in the pristine TiO2 and 0.07 wt.% MWCNTs incorporated TiO2 thin films measured by using XPS.
Pristine TiO2 MWCNTs-TiO2
C (%) (C/Ti)
O (%) (C/Ti)
Ti (%) (Ti/Ti)
13.0 (0.54) 12.1 (0.56)
63.0 (2.62) 66.4 (3.08)
24.0 (1) 21.5 (1)
H. Kim et al. / Thin Solid Films 529 (2013) 89–93
Acknowledgment This study was supported by a grant from the Fundamental R&D Program (grant no. K0004114) for Core Technology of Materials funded by the Ministry of Knowledge Economy, Republic of Korea. This work (grant no. 00045405) was supported by Business for Cooperative R&D between Industry, Academy, and Research Institute funded by the Korea Small and Medium Business Administration in 2011. References [1] A. Wisitsoraat, A. Tuantranont, E. Comini, G. Sberveglieri, W. Wlodarski, Thin Solid Films 517 (2009) 2775. [2] O. Byl, J.T. Yates Jr., J. Phys. Chem. B 110 (2006) 22966. [3] C. Wang, L. Yin, L. Zhang, D. Xiang, R. Gao, Sensors 10 (2010) 2088. [4] H. Zengin, W. Zhou, J. Jin, R. Czerw, D.W. Smith Jr., L. Echegoyen, D.L. Carroll, S.H. Foulger, J. Ballato, Adv. Mater. 14 (2002) 1480. [5] O.K. Varghese, P.D. Kichambre, D. Gong, K.G. Ong, E.C. Dickey, C.A. Grimes, Sens. Actuators B 81 (2001) 32. [6] A.M. Taurino, S. Capone, A. Boschetti, T. Toccoli, R. Verucchi, A. Pallaoro, P. Siciliano, S. Iannotta, Sens. Actuators B 100 (2004) 177.
93
[7] H. Kim, S.J. Wang, H.H. Park, H.J. Chang, H. Jeon, H.H. Park, Thin Solid Films 516 (2007) 198. [8] J. Chastain, in: Handbook of X-ray Photoelectron Spectroscopy, Perkin-Elmer Corporation Physical Electronics Division, Minnesota, 1992, p. 40. [9] D. Yang, A. Velamakanni, G. Bozoklu, S. Park, M. Stoller, R.D. Piner, S. Stankovich, I. Jung, D.A. Field, C.A. Ventrice Jr., R.S. Ruoff, Carbon 47 (2009) 145. [10] A.V. Fetisov, G.A. Kozhina, V.B. Fetisov, E.A. Pastukhov, J. Appl. Spectrosc. 78 (2011) 240. [11] J.B. Hudson, in: Surface Science: An Introduction, Butterworth-Heinemann, Stoneham, 1992, p. 38. [12] B.S. Shirke, P.V. Korake, P.P. Hankare, S.R. Bamane, K.M. Garadkar, J. Mater. Sci. Mater. Electron. 22 (2011) 821. [13] J. Zie, V.K. Varadan, Mater. Chem. Phys. 91 (2005) 274. [14] Y. Luo, J. Liu, X. Xia, X. Li, T. Fang, S. Li, Q. Ren, J. Li, Z. Jia, Mater. Lett. 61 (2007) 2467. [15] H.J. Kim, R. Koizhaiganova, T. Vasudevan, C. Sanjeeviraja, M.S. Lee, Polym. Adv. Technol. 20 (2009) 736. [16] C. Li, E.T. Thostenson, T. Chou, Appl. Phys. Lett. 91 (2007) 223114. [17] H. Li, J. Xub, Y. Zhua, X. Chen, Q. Xiang, Talanta 82 (2010) 458. [18] J. Zhu, J. Yang, Z.F. Bian, J. Ren, Y.M. Liu, Y. Cao, H.X. Li, H.Y. He, K.N. Fan, Appl. Catal. B Environ. 76 (2007) 82. [19] Y. Tang, J. Gou, Mater. Lett. 64 (2010) 2513.