Gas Sensing with Nano-Indium Oxides (In2O3) Prepared via

Dec 21, 2011 - A rapid, clean, and continuous hydrothermal route to the synthesis of ca. 14 nm indium oxide (In2O3) nanoparticles using a superheated ...
0 downloads 26 Views 5MB Size
Download PDF
Article pubs.acs.org/Langmuir

Gas Sensing with Nano-Indium Oxides (In2O3) Prepared via Continuous Hydrothermal Flow Synthesis Sofia Elouali,† Leanne G. Bloor,† Russell Binions,*,†,‡ Ivan P. Parkin,† Claire J. Carmalt,† and Jawwad A. Darr*,† †

Department of Chemistry, Christopher Ingold Laboratories, University College London, 20 Gordon Street, London, WC1H 0AJ, United Kingdom ‡ School of Engineering and Materials Science, Queen Mary University of London, Mile End Road, London E1 4NS, United Kingdom S Supporting Information *

ABSTRACT: A rapid, clean, and continuous hydrothermal route to the synthesis of ca. 14 nm indium oxide (In2O3) nanoparticles using a superheated water flow at 400 °C and 24.1 MPa as a crystallizing medium and reagent is described. Powder X-ray diffraction (XRD) of the particles revealed that they were highly crystalline despite their very short time under hydrothermal flow conditions. Gas sensing substrates were prepared from an In2O3 suspension via drop-coating, and their gas sensing properties were tested for response to butane, ethanol, CO, ammonia, and NO2 gases. The sensors showed excellent selectivity toward ethanol, giving a response of 18−20 ppm.

■

INTRODUCTION Bulk indium oxide (In2O3) is a wide bandgap semiconductor (Eg = 3.7 eV), well-known for its useful optoelectronic properties in both doped and undoped forms. Indium oxide films can be prepared via various deposition methods1−3 and have found use in a variety of electronics applications from flat panel displays to solar panels and heat reflecting surfaces. They are also of some interest as chemical sensors as In2O3 has been found to be particularly sensitive to reducing gases such as ethanol and ammonia4 and oxidizing gases such as ozone.5 Many studies have been conducted in order to improve the performance of In2O3 gas sensors by using crystallites with reduced dimensions since Yamazoe demonstrated that a reduction in crystallite size can significantly increase sensor performance.6 Indeed, smaller particle sizes have been found to result in improved gas responses during sensing.7−11 For example, a study carried out by Korotecenkov reported as much as a 103 increase in response to ozone at 300 °C, with a drop concomitant in response time, when the average In2O3 primary particle size was reduced from 80 to 15 nm.10 While several direct and rapid syntheses of In2O3 colloids have been reported in the literature via organometallic synthetic methods, these are air-sensitive and require the use of solvents and other chemical modifiers over a period of 3−7 h.12−14 Most investigations into In2O3 nanoparticle synthesis have involved batch solvothermal or hydrothermal methods which are lengthy, often requiring 12−24 h to complete and usually resulting in the production of In(OH)3 due to the low reaction temperatures.9,15−19 These In(OH)3 nanoparticles then require calcination at temperatures higher than 400 °C in order to convert them to In2O3, often causing extensive particle growth © 2011 American Chemical Society

or agglomeration which may subsequently affect the overall properties of the final gas sensor film. In2O3 formed around this temperature is poorly crystalline,8 while excessively high sintering temperatures result in large crystallite sizes with broad distributions which may further detrimentally affect gas sensor properties. In contrast to batch processes, continuous hydrothermal flow synthesis (CHFS) reactions offer a rapid and controllable method of producing semiconductor nanoparticles including TiO2,20,21 rare earth oxide solid solutions,22 potential SOFC cathodes,23 Ni/YSZ anodes,24 and bioceramics.25,26 In CHFS, a superheated water stream (near or above the critical temperature, Tc = 374 °C, and pressure, Pc = 22 MPa) is brought into contact with a room temperature flow containing metal salt(s), resulting in the rapid precipitation and crystallization of nanoparticulate products which are then cooled in-line within minutes of the initial nucleation. This very short duration under such extreme hydrothermal conditions limits particle growth, while the relatively high temperature of reaction can produce crystalline materials that either are only obtained under batch processes after many hours or are difficult to access directly such as In2O3 or other kinetic products. In this work, a rapid and continuous hydrothermal flow synthesis method was used to prepare In2O3 nanopowders directly in water only, without the use of further chemical modifiers. The resulting slurry was used to prepare gas sensor devices by drop-coating which were then tested for sensitivity Received: September 12, 2011 Revised: December 20, 2011 Published: December 21, 2011 1879

dx.doi.org/10.1021/la203565h | Langmuir 2012, 28, 1879−1885

Langmuir

Article

cylinders of synthetic air (79% nitrogen, 21% oxygen) containing target analytes: ethanol (0−20 ppm), ammonia (0−10 ppm), butane (0−10 ppm), carbon monoxide (0−200 ppm), or nitrogen dioxide (0−16 ppb). The devices gas response behavior was investigated at operating temperatures between 300 and 600 °C in 50 °C increments against all five target analytes. 300 °C was found to be the lowest temperature the sensors could be operated at. At temperatures lower than 300 °C it was found the baseline resistance was in the tens of MΩ region, and there was a poor signal-to-noise ratio rendering the data unreliable.

to a variety of gases in environmentally relevant concentrations.27 The sensors showed remarkable sensitivity to ethanol, giving a gas response of 18−20 ppm ethanol at an operating temperature of 300 °C.

■

EXPERIMENTAL METHODS

Materials Synthesis. Indium(III) nitrate monohydrate [In(NO3)3·H2O, 99%] was obtained from Alfa Aesar (Lancashire, UK) and initially prepared as a 0.1 M aqueous solution. 10 MΩ deionized (DI) water deionizer was used throughout. The basic design of the CHFS system has been reported elsewhere,25,28 but in summary, it consists of three Gilson HPLC pumps, 316SS Swagelok stainless steel tubing and fittings, and a 2000 W electrically powered water preheater. The entire system was pressurized to 24.1 MPa using a Tescom backpressure regulator (model no. 26-1762-24) fitted at the system outlet. The [In(NO3)3·H2O] solution was pumped into the CHFS system at 10 mL min−1 and diluted in-line with a further 10 mL min−1 flow of DI water at a T-piece premixer. This second pump flow has been used in previous investigations to add auxiliary reagents or modifiers such as base which were not required in this work. Its use was retained in order to maintain typical CHFS conditions in terms of relative pump flow rates. The diluted room temperature feed then met a 20 mL min−1 flow of DI water (preheated to 400 °C) within a stainless steel cocurrent mixer (UK patent pending ref GB335328) made from 1/4 in. and 1/16 in. pipe and a 1/4 in. cross (see Supporting Information Figure S1), resulting in the rapid crystallization and precipitation of nanosized In2O3 particles. The temperature in the jet mixer was calculated to be ca. 307 °C.29 The product flow was cooled in-line via a water jacket cooler and continuously collected at the exit of the backpressure regulator before being concentrated via centrifugation and resuspended in a volume of 40 mL. The resulting concentrated lemonyellow suspension was cleaned of waste solutes via dialysis using visking tubing (SLS Ltd., UK); water changes were carried out at regular intervals until the conductivity of the immersion water was no longer above 30 μS. Supporting Information Figure S2 shows the full CHFS system scheme. Materials Characterization. Particle size and morphology of the In2O3 nanocrystals were investigated using a JEOL transmission electron microscope after a few drops of the product suspension were diluted in methanol and dropped onto carbon-coated copper grids (Holey Carbon Film, 300 mesh Cu, Agar Scientific, Stansted, UK) and allowed to air-dry. Powder X-ray diffraction (XRD) measurements on the as-prepared In2O3 were carried out on a freeze-dried portion of the suspension using a Bruker Gadds D8 diffractometer using Cu Kα radiation (λ = 0.154 18 nm). Data were collected over the 2θ range 15°−80° with a step size of 0.04° and a count time of 2 s. Raman spectra of the sensor films were recorded before and after sensing tests at room temperature with a Renishaw 1000 Raman microscope system equipped with a diode laser operating at 514 nm. Scanning electron microscopy (SEM) of the In2O3 sensor films was carried out on a Phillips XL30 ESEM instrument. Powder X-ray diffraction (XRD) measurements on the sensor film were carried out on a Bruker Gadds D8 diffractometer. Gas Sensor Characterization. For gas response, films were deposited onto commercially produced 3 × 3 mm sensor substrates (City Technology, Portsmouth, UK). 12 μL drops of the In2O3 nanoparticle slurry (0.225 g mL−1) were deposited onto the substrate using a calibrated Gilson pipet, with air-drying between separate depositions. The sensors were fired at 600 °C for 2 h in air using a muffle furnace. The sensors consist of a gold track printed on the top of an alumina tile and a platinum heater track printed on the reverse side of the tile. Gold electrodes were formed by laser trimming to produce an interdigitized section with gap and finger widths of 50 μm. Contacts to the devices were formed by spot-welding 50 μm diameter platinum wire to pads of the track material in the corner of the sensor chip. The sensor heater was kept at constant resistance and hence constant temperature by incorporating it into a Wheatstone bridge. Electrical experiments were formed on a locally constructed test rig.11,30 Various concentrations of test gases were diluted from

■

RESULTS AND DISCUSSION Materials Characterization. Highly crystalline In2O3 nanopowder was prepared by continuous hydrothermal flow synthesis from a 0.1 M indium nitrate solution. The powder XRD pattern of the as-prepared In2O3 nanoparticles is shown in Figure 1a and confirms the formation of the cubic (bixbyite)

Figure 1. XRD patterns of (a) as-prepared phase pure nano-In2O3 made in the continuous hydrothermal flow synthesis system using superheated water at 400 °C and 24 MPa and (b) In2O3 gas sensor film after testing. Major indices are marked, as matched to the ICDS pattern [640179] for bixbyite phase In2O3.

phase of In2O3, with a characteristic intense (222) peak at 2θ = 30.6° and other strong peaks at 2θ = 35.5°, 51.0°, and 60.7°. The peaks were significantly broadened, indicating a small crystallite size. The strongly defined peaks indicate a high degree of crystallinity in the product. This high crystallinity and the fact that cubic-phase In2O3 was obtained from a hydrothermal reaction between indium nitrate and water with no other additives can be attributed to the extreme hydrothermal conditions reached in the continuous flow process where the temperature in the jet mixer was calculated29 to be 307 °C on the basis of a heater temperature of 400 °C and flow rates of 20 mL min−1 for the hot water feed and 20 mL min−1 for the diluted In(NO3)3·H2O solution. Powder XRD was also carried out on the gas sensor film after its use in testing as shown in Figure 1b, confirming the retention of nanosized cubic In2O3 on the gas sensor substrate. Using the Scherrer equation (τ = 0.9λ/β cos θ where τ is the domain size, λ the wavelength of the incoming radiation, β the line broadening at full width half-maximum, and θ the Bragg angle of the peak) to calculate the crystallite size for the as-synthesized powder and the used sensor material indicated crystallite sizes of 14.8 ± 2 and 15.5 ± 2 nm, respectively. This indicates some sintering, but within experimental error the sizes are essentially the same. 1880

dx.doi.org/10.1021/la203565h | Langmuir 2012, 28, 1879−1885

Langmuir

Article

Figure 2. TEM image (bar = 50 nm) of (a) as-prepared In2O3 nanocrystals made directly in the continuous hydrothermal flow system using superheated water at 400 °C and 24 MPa and (b) details of In2O3 nanocrystal agglomerates showing well-defined semicubic-shaped particles.

subsequently evaluated. The gas sensor substrates consisted of gold electrodes on an alumina tile, and the whole assembly was subsequently annealed at 600 °C for 10 min. The XRD and Raman patterns (Figures 1 and 4, respectively) were collected post-testing and showed the indium oxide retained its cubic crystalline phase.

We do not expect significant changes in crystal size to occur; several hours at 600 °C while carbonaceous material is being burnt out is too low a temperature for significant sintering to occur. The In2O3 particle suspension was analyzed by transmission electron microscopy (TEM), images of which are shown in Figure 2. The nanoparticles were well-defined, with the majority appearing as rounded cubes 5−20 nm in size. Figure 2a shows the TEM of the as-prepared particles. Figure 2b shows a higher magnification TEM of the relatively loose packing of the In2O3 nanoparticles within a larger agglomerate; the majority of the particle edges are still easily observable. A sample of 1000 imaged particles was measured in order to calculate the mean particle size and obtain a size distribution (Figure 3). All particles imaged were smaller than 35 nm, and a

Figure 4. Raman spectra of (a) the as-prepared CHFS-made nanoIn2O3 gas sensor film and (b) heat-treated In2O3 film after gas sensing, measured at room temperature with a 514 nm laser.

Raman spectroscopy was used to further investigate the local chemical structure of the In2O3 sensor films (Figure 4). Peaks were observed at 308, 367, 497, and 630 cm−1, and these were in general agreement with those values previously reported in the literature for bulk body-centered-cubic In2O3.31,32 The peak at 308 cm−1 corresponds to the δ(InO6) bending vibration of InO6 octahedra. The 367 cm−1 peak can be assigned to the stretching vibrations of In−O−In linkages, while peaks at 497 and 630 cm−1 can be attributed to octahedral stretching vibrations v(InO6). No significant differences in shifts were observed between the as-prepared gas sensor films (Figure 4a) and those that had been heat-treated and used in sensing tests (Figure 4b). Scanning electron microscopy (SEM) of the prepared sensor films showed a film thickness over the sensor electrode grid of 64 ± 5 μm for the two-coat sensor and 80 ± 6 μm for the three-coat sensor (Figure 5A,B). The sensor films appeared to

Figure 3. Particle size distribution plot of a sample population of directly synthesized CHFS In2O3 observed via TEM.

relatively broad distribution was observed compared with the calculated mean particle diameter of 13.7 nm. The standard deviation of the sample was found to be 4.3 nm. These sizes compare well with the sizes calculated from the XRD data. Sensor Synthesis and Characterization. The In2O3 sensors were fabricated by drop-coating two or three 12 μL drops of a water based slurry of nanoparticles (0.225 g mL−1) onto gas sensor substrates using a calibrated Gilson pipet. The sensors were fired at 600 °C for 2 h in a furnace to improve substrate adhesion. Their gas sensing properties were 1881

dx.doi.org/10.1021/la203565h | Langmuir 2012, 28, 1879−1885

Langmuir

Article

Figure 5. SEM micrographs of (A) cross section of two-coat nano-In2O3 gas sensor film, (B) cross section of three-coat nano-In2O3 gas sensor film, and (C) surface details of the gas sensor film.

The gas response was rapid, and a near plateau was reached within minutes of exposure to ethanol, as shown in Figure 6. This almost square response shape observed indicates that the sensors first respond rapidly to ethanol, quickly achieving a near steady state.11 The response then changes more slowly as the analyte gas diffuses through the material and occupys the remaining surface reaction sites. When the ethanol flow was stopped, the response drops to near baseline level and then tails off of at a similar rate to the baseline. The resistive gas response for each concentration is shown in Table 1. The highest response (R0/R) of two- and three-coat sensors were 13.4 and 14.7, respectively, to 20 ppm concentration of ethanol. The response of the thicker three-coat sensor is greater than that of the thinner two-coat sensor. The difference in the two sensors can be explained by the thicker three-coat sensor having a larger quantity of material and most likely a larger total surface area. Therefore, a larger number of surface sites are available for absorption and reaction to occur. In spite of both sensors appearing to be quite dense (Figure 5), they must retain sufficient porosity that the analyte gas can easily diffuse throughout the sensor material reaching all available surface reaction sites, or we would expect to observe a slower gas response. The response to ethanol was also tested at different temperatures. Table 1 summarizes the maximum response of the two- and three-coat sensors at different concentrations of ethanol in dry air as a function of temperature. The response decreases linearly with increasing temperature, showing a maximum response of 13.4 for the two-layer sensor and 14.7 for the three-coat sensor at 300 °C. The In2O3 gas sensors were also tested against the reducing gas ammonia, as indium oxide has previously been reported to show good sensitivity to NH3.4 Figure 7 shows the variation in

be made up of nano-In2O3 agglomerates (ranging 400−1000 nm in size), which were distinctly visible in cross sections taken of the films. SEM images taken of the top face of the films (Figure 5C) show a degree of surface cracking on the micrometer scale that is attributed to shrinkages from the heat treatment employed after coating. In all cases, the sensors appeared to have low porosity with agglomerates being relatively densely packed. Gas response was measured as the ratio between R0 (the point immediately prior to exposure to ethanol) and R (the resistance when exposed to ethanol). The resistance was measured between the two electrodes of an interdigitized gold electrode structure on an alumina tile. In2O3 gas sensing substrates showed an n-type response to ethanol at different concentrations over time at 300 °C (Figure 6).

Figure 6. Gas response (R0/R) of In2O3 sensors upon exposures to differing concentrations of ethanol in flowing air over time at 300 °C. Arrows indicate when gas flow of ethanol was turned on and off.

Table 1. Maximum Gas Response (R0/R) of Two- and Three-Coat In2O3 Gas Sensors at Varying Temperatures on Exposure to Different Concentrations of Ethanol 2-coat sensor response R0/R

3-coat sensor response R0/R

ethanol concn/ppm

ethanol concn/ppm

temp/°C

4

8

16

20

4

8

16

20

300 400 450 500

5.32 3.14 1.98 1.42

7.69 4.21 2.53 1.66

11.59 5.71 3.31 2.00

13.36 6.35 3.65 2.16

5.78 4.12 3.01 1.86

8.39 5.57 4.09 2.33

12.74 7.59 5.53 2.99

18.12 8.42 6.13 3.27

1882

dx.doi.org/10.1021/la203565h | Langmuir 2012, 28, 1879−1885

Langmuir

Article

Figure 7. Gas response (R0/R) of In2O3 sensors upon exposures to differing concentrations of ammonia in flowing air over time at 450 °C. Arrows indicate when gas flow of ammonia was turned on and off.

Figure 8. Maximum gas response (R0/R) of three-coat In2O3 sensors to different gases in flowing air at the optimum operating temperature for that gas.

the sensor response as a function of the temperature upon exposure to increasing ammonia concentrations. The gas response was rapid, and a shark-fin-like profile was reached within minutes of exposure to ammonia, as shown in Figure 7. The shark-fin response shape observed indicates that both sensors do not saturate during the experiment. When the ammonia flow was stopped, the response drops to near baseline level but takes longer to reach the baseline again compared to the response with ethanol. This response behavior indicates a large number of surface sites suitable for reaction with ammonia and, as such, that the resistance of the sensor has a significant proportion deriving from surface reactions. The resistive gas response for each concentration at a variety of operating temperatures is summarized in Table 2. The highest response of two- and three-coat sensors was 1.53 and 1.76, respectively, to 10 ppm concentration of ammonia at an operating temperature of 450 °C. The optimum operating temperature for ammonia detection (in this work) was found to be 450 °C. Above this temperature, the sensor response decreased until at 600 °C it was barely perceptible. Once again, the three-coat sensor gave larger responses than the two-coat sensor, and this is again attributed to a larger number of surface sites available for reaction. The observed responses compare favorably to those observed by Makija et al.,4 who observed similar magnitude responses but at much higher concentrations of ammoniabetween 100 and 500 ppm. In order to test cross-sensitivity, the In2O3 gas sensors were also tested against a variety of other gases in environmentally relevant concentrations27 (NO2, CO, and butane) at a variety of different sensor operating temperatures. Figure 8 shows the maximum response to the different gases at the optiumum

operating temperature for that gas. (This was found to be 450 °C for all gases other than ethanol where it was 300 °C.) The maximum response (observed from these experiments) of all the gases is quite similar to the exception of ethanol, which has a response of more than 4 times that of the other gases. Consequently, there is selectivity toward ethanol over the other gases at these concentrations. The most widely accepted theory of chemiresistive gas sensing states that “when a metal oxide semiconductor gas sensor is exposed to air, oxygen species are adsorbed on the surface of the sensor and go on to be ionized by electrons from the materials conduction band to form species such as O2−”.33 As such, in air, the measured resistance of In2O3 and other ntype semiconductors will increase due to the lower concentration of free electrons in the materials conduction band.27 On exposure to a reducing gas such as ethanol, surface reaction between the oxygen species and analyte gas can occur, leading to the release of electrons trapped in the ionized oxygen species back into the materials conduction band, thereby lowering the measured resistance. The ability of the sensor material to absorb and ionize oxygen species is fundamental to the sensor performance. The good performance observed here is likely to be the result of two factors. First, the crystallite size is small; a mean size of 13.7 nm was observed (Figure 3). This causes two important effects: first, the surfaces of the crystallites become significantly more reactive and likely to absorb oxygen and form ionized oxygen species.34−36 Second, the surface-to-volume ratio of the materials is also significantly increased. This means that the amount of oxygen that can be absorbed and ionized is maximized.

Table 2. Maximum Gas Response (R0/R) of Two- and Three-Coat In2O3 Gas Sensors Increasing Temperatures on Exposure to Different Concentrations of Ammonia in Flowing Air 2-coat sensor response R0/R

3-coat sensor response R0/R

ammonia concn/ppm

ammonia concn/ppm

temp/°C

2

4

8

10

2

4

8

10

400 450 500 600

1.06 1.11 1.05 1.01

1.18 1.23 1.08 1.01

1.31 1.37 1.14 1.03

1.46 1.53 1.20 1.04

1.05 1.20 1.09 1.01

1.28 1.36 1.15 1.02

1.45 1.57 1.26 1.05

1.61 1.76 1.35 1.07

1883

dx.doi.org/10.1021/la203565h | Langmuir 2012, 28, 1879−1885

Langmuir

Article

support (grant EP/H005803/1). J.A.D. also thanks EPSRC for funding [Lab to pilot plant project (EP/E040551/1)]. Mr. Kevin Reeves is thanked for assistance with scanning electron microscopy. Dr. Steve Firth is thanked for assistance with Raman and transmission electron microscopy.

The sensors prepared in this work were selective toward ethanol (Figure 8) and gave response magnitudes to 20 ppm in great excess of previously reported In2O3 sensors based on other processing methodologies.4,37,38 Our sensors show a response of a 14.69−20 ppm pulse of ethanol comparing favorably with a response of 6.5−50 ppm ethanol by In2O3 nanorods prepared solvathermally,37 1.60−400 ppm by polymer encapsulated In2O3 nanoparticles,4 or 2.0−100 ppm by In2O3 nanowires.38 Our results are comparable to those recently reported for porous nanowires films39 that give a response of 15 toward 20 ppm ethanol. As has been reported previously and demonstrated here, a reduction of the sensor material crystallite size can dramatically increase sensor sensitivity and performance. The continuous hydrothermal synthesis technique has the advantages of being readily scalable and capable of producing highly crystalline nanopowders and is ideally suited for producing materials for metal oxide semiconductor gas sensors. Indeed, the authors have now developed a pilot plant CHFS capable of synthesizing multikg per day quantities of nanomaterials (under EPRSC grant EP/E040551/1).

■

■

CONCLUSION In summary, we have demonstrated for the first time that cubicphase nano-indium oxide can directly be prepared using a continuous hydrothermal flow synthesis reactor. The formation of indium oxide rather than the hydroxide is as a result of the higher temperature of the reaction (307 °C) compared to analogous batch processes, which provides a strongly dehydrating environment. The In2O3 nanopowders were utilized in sensor devices and their gas sensing properties examined. The almost square response shape observed for ethanol indicates that the sensors rapidly respond to ethanol. When the ethanol flow was stopped, the response drops to near baseline level, suggesting a lack of surface sites suitable for reaction, i.e., that the resistance of the sensor is dominated by the bulk contribution. This seems somewhat surprising given that the sensor is made up of nanoparticles for which significant grain boundaries are present. Analysis by XRD and Raman indicated that the nanoparticulate nature of the powders is retained after sensing. It was found that the devices exhibited selectivity toward ethanol and were significantly more responsive to this gas than previously reported In2O3 sensor devices giving a response of 18 toward 20 ppm of ethanol. Continuous hydrothermal flow synthesis presents an exciting new method for the synthesis of nanopowders for use in gas sensing devices.

■

ASSOCIATED CONTENT

S Supporting Information *

Figures S1 and S2. This material is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

(1) Knapp, C. E.; Hyett, G.; Parkin, I. P.; Carmalt, C. J. Chem. Mater. 2011, 23, 1719−1726. (2) Basharat, S.; Carmalt, C. J.; Barnett, S. A.; Tocher, D. A.; Dawson, D. H. Inorg. Chem. 2007, 46, 9473−9480. (3) Bloor, L. G.; Carmalt, C. J.; Pugh, D. Coord. Chem. Rev. 2011, 255, 1293−1318. (4) Makhija, K. K.; Ray, A.; Patel, R. M.; Rrivedi, U. B.; Kapse, H. N. Bull. Mater. Sci. 2005, 28 (28), 9−17. (5) Takada, T.; Suzuki, K.; Nakane, M. Sens. Actuators, B 1993, 13, 404−407. (6) Yamazoe, N. Sens. Actuators, B 1991, 5, 7−19. (7) Gurlo, A.; Ivanovskaya, M.; Bârsan, N.; Schewizer-Berberich, M.; Weimar, U.; Göpel, W.; Diéguez, A. Sens. Actuators, B 1997, 44, 327− 333. (8) Singh, V. N.; Mehta, B. R. Jpn. J. Appl. Phys. 2003, 42, 4226− 4232. (9) Tang, Q.; Zhou, W.; Zhang, W.; Ou, S.; Jian, K.; Yu, W.; Qian, Y. Cryst. Growth Des. 2005, 5, 147−150. (10) Korotcenkov, G.; Cerneavschi, A.; Brinzari, V.; Vasiliev, A.; Ivanov, M.; Cornet, A.; Morante, J.; Cabot, A.; Arbiol, J. Sens. Actuators, B 2004, 99, 297−303. (11) Naisbitt, S. C.; Pratt, K. F. E.; Williams, D. E.; Parkin, I. P. Sens. Actuators, B 2006, 114, 969−977. (12) Seo, W. S.; Jo, H. H.; Lee, K.; Park, J. T. Adv. Mater. 2003, 15, 795−797. (13) Liu, Q.; Lu, W.; Ma, A.; Tang, J. J. Am. Chem. Soc. 2005, 127, 5276−5277. (14) Yao, K.; Caruntu, D.; Zeng, Z.; Chen, J.; O′Connor, C. J.; Zhou, W. J. Phys. Chem. C 2009, 113, 14812−14817. (15) Xu, J. Q.; Wang, X. H.; Shen, J. N. Sens. Actuators, B 2006, 115 (2), 642−646. (16) Li, B. X.; Xie, Y.; Jing, M.; Rong, G. X.; Tang, Y. C.; Zhang, G. Z. I. Langmuir 2006, 22 (22), 9380−9385. (17) Zhu, H.; Wang, X.; Yang, F.; Yang, X. Cryst. Growth Des. 2008, 8, 950−956. (18) Zhu, H. L.; Yao, K. H.; Wo, Y. H.; Wang, N. Y.; Wang, L. N. Semicond. Sci. Technol. 2004, 19, 1020−1023. (19) Du, J.; Yang, M.; Cha, S. N.; Rhen, D.; Kang, M.; Kang, D. J. Cryst. Growth Des. 2008, 8, 2312−2317. (20) Zhang, Z.; Goodall, J. B. M.; Morgan, D. J.; Brown, S.; Clark, R. J. H.; Knowles, J. C.; Mordan, N. J.; Evans, J. R.; Carley, A. F.; Bowker, M.; Darr, J. A. J. Eur. Ceram. Soc. 2009, 29, 2343−2353. (21) Zhang, Z.; Brown, S.; Goodall, J. B. M.; Weng, X.; Thompson, K.; Gong, K.; Kellici, S.; Clark, R. J. H.; Evans, J. R. G.; Darr, J. A. J. Alloys Compd. 2009, 476, 451−456. (22) Weng, X. L.; Perston, B.; Wang, X. Z.; Abrahams, I.; Lin, T.; Yang, S. F.; Evans, J. R. G.; Mordan, N. J.; Evans, J. R. G.; Carley, A. F.; Bowker, M.; Knowles, J. C.; Rehman, I.; Darr, J. A. Appl. Catal., B 2009, 90, 405−415. (23) Weng, X. L.; Boldrin, P.; Abrahams, I.; Skinner, S. J.; Darr, J. A. Chem. Mater. 2007, 19 (18), 4382−4384. (24) Weng, X. L.; Brett, D.; Yufit, V.; Shearing, P.; Brandon, N.; Reece, M.; Yan, H. X.; Tighe, C.; Darr, J. A. Solid State Ionics 2010, 181, 827−834. (25) Chaudhry, A. A.; Haque, S.; Kellici, S.; Boldrin, P.; Rehman, I.; Fazal, A. K.; Darr, J. A. Chem. Commun. 2006, 21, 2286−2288. (26) Chaudhry, A. A.; Goodall, J.; Vickers, M.; Cockcroft, J. K.; Rehman, I.; Knowles, J. C.; Darr, J. A. J. Mater. Chem. 2008, 18 (48), 5900−5908. (27) Fine, G. F.; Cavanagh, L. M.; Afonja, A.; Binions, R. Sensors 2010, 6, 5468−5502.

AUTHOR INFORMATION

Corresponding Author

*Tel 020 7679 4345, e-mail [email protected] (J.A.D.); Tel 020 7882 5305, e-mail [email protected] (R.B.).

■

ACKNOWLEDGMENTS The EPSRC is thanked for funding the synthesis work under Grant Reference EP/F056168/1. R.B. thanks the Royal Society for a Dorothy Hodgkin fellowship and the EPSRC for financial 1884

dx.doi.org/10.1021/la203565h | Langmuir 2012, 28, 1879−1885

Langmuir

Article

(28) Boldrin, P.; Hebb, A. K.; Chaudhry, A. A.; Otley, L.; Thiebaut, B.; Bishop, P.; Darr, J. A. Ind. Eng. Chem. Res. 2007, 46 (14), 4830− 4838. (29) Wagner, W.; Pruss, A. The IAPWS formulation 1995 for the thermodynamic properties of ordinary water substance for general and scientific use. J. Phys. Chem. Ref. Data 2002, 31, 387−535. (30) Henshaw, G. S.; Dawson, D. H.; Williams, D. E. J. Mater. Chem. 1995, 5, 1791. (31) Schwartz, G. P.; Sunder, W. A.; Griffiths, J. E. J. Electrochem. Soc. 1982, 129 (6), 1361−1367. (32) Wang, C.; Dai, Y.; Pezoldt, J.; Lu, B.; Kups, T.; Cimalla, V.; Ambacher, O. Cryst. Growth Des. 2008, 8 (4), 1257−1260. (33) Barsan, N.; Weimar, U. J. Electroceram. 2001, 7, 143−167. (34) Xu, C.; Taamaki, J.; Mirua, N.; Yamazoe, N. Sens. Actuators, B 1991, 3, 147−155. (35) Gurlo, A.; Barsan, N.; Ivanovska, M.; Weimar, U.; Göpel, W. Sens. Actuators, B 1998, 47, 92−99. (36) Franke, M. E.; Koplin, T. J.; Simon, U. Small 2006, 2, 36−50. (37) Xu, J.; Chen, Y.; Shen, J. Mater. Lett. 2008, 62, 1363−1365. (38) Xiangfeng, C.; Caihong, W.; Dongli, J.; Chengmu, Z. Chem. Phys. Lett. 2004, 399, 461−464. (39) Liu, J.; Guo, Z.; Zhu, K.; Wang, W.; Zhang, C.; Chen, X. J. Mater. Chem. 2011, 21, 11412−11417.

1885

dx.doi.org/10.1021/la203565h | Langmuir 2012, 28, 1879−1885