Abstract

Nanostructured Sn-doped TiO2 have been prepared by ball milling using SnO2 and TiO2 as raw materials. The as-prepared powders are characterized by XRD, SEMs and EDAX to identify the structural phases, surface morphology, and composition of the materials. The materials are prepared with the addition of tin of different molar ratios (0, 0.05, 0.10, 0.15, 0.20, 0.25, and 1.0) to TiO2 and sintered at 800°C for 3 h. They are subjected to dc resistance measurements as a function of relative humidity (RH) in the range of 30%–97% in a self-designed humidity chamber, and the results revealed that the sensitivity factor increased with an increase in tin molar ratio. Among them, TiO2—20 wt% of SnO2 possessed the highest humidity sensitivity, while the pure TiO2 and SnO2 composite possessed a low sensitivity.

1. Introduction

The development of humidity sensors has received much attention during the last years due to the necessity of controlling and monitoring environments in many different fields like industrial processes and domestic comfort [13]. Since each application field requests distinct operating conditions and generally sensor elements work in narrow ranges of humidity and temperature, the selection of a material should be based on certain conditions, in order to assure the satisfactory operation of the humidity sensor, that include good sensitivity, linearity over the range of application, fast response, low hysteresis, and stability in the exhibition to impurities present in the environment [36].

As a consequence, a wide variety of materials have been researched with the objective to study their sensitivity, usually variations of electrical parameters in relation to the humidity in the atmosphere. These materials are based on polymers, electrolytes, and especially ceramics [1, 3].

At present, ceramic materials possess a certain prominence due to their properties, which exhibit advantages regarding their mechanical resistance, and physical and chemical stability. The humidity-sensitivity properties of a ceramic are mostly influenced through the porous microstructure and the surface reactivity with moisture. Porous ceramic materials based on metals oxides have been largely used as humidity sensors [2]. The principle of humidity measurement with ceramic sensors is the change in electrical capacitance and conductance owing to water vapor chemisorption and physisorption and/or capillary condensation in the pores of the ceramic sensor.

TiO2-based humidity sensors are studied vigorously, as it is likely to show better sensitivity because of its hydrophilic property [7]. It is studied more in thin film form [8]. However, it is reported that thin film sensors show lower sensitivity than those shown by porous ceramic sintered counterparts [79]. TiO2 is known to have a hysteresis in humidity sensitivity curve. It was thought that the use of additives would be useful to get over this drawback. The additives are reported to minimize the hysteresis in TiO2 humidity sensors [10]. Additives are also reported to improve performance parameters of TiO2 humidity sensors [1115]. SnO2 is known as isostructural to a rutile phase of TiO2, and fairly good electrical conductor was considered preferentially as an additive. Although thin films have many advantageous properties for sensor applications, achieving porous thin films by use of energy intensive thin film formation processes has remained an unsolved problem for the scientists.

The aim of this work is to synthesize and investigate humidity sensing properties of Sn-doped nano-TiO2 ceramic material by ball milling.

2. Experimental

To synthesize Sn-doped nano-TiO2, raw materials 99.5% pure TiO2 (Aldrich) and 98% pure SnO2 (Aldrich) are used. We have prepared pure TiO2, SnO2, and Ti1-xSnxO2 composition, where = 5, 10, 15, 20, 25 wt% of SnO2. A planetary ball mill (Fritsch Company) was used with its vial rotation of 300 rpm. The vial and balls are made of steel. The raw materials are milled for 5 hours. The total mass of powder was about 50 gm and the ratio of balls to powder is 1 : 1. Milling was interrupted for 15 minutes to suppress the excessive temperature rise of the vial and balls. Now, pellets of these materials were made using a hydraulic pressing machine under the pressure of 55158 KPascal at room temperature. Size of each pellet was 5 cm diameter and 4 mm thickness. Each pellet was annealed at 800°C for 3 hours inside an electric furnace. Each pellet was kept within an electrode holder and has been exposed to humidity in the self-designed humidity chamber, and variations in resistance with humidity have been recorded. Relative humidity is measured using standard hygrometer. Variations in resistance were noted by using sinometer.

3. Characterization

3.1. XRD

A well-crystallized anatase form has been found for undoped TiO2 (Figure 1(a), which corresponds well with the JCPDS data file no. 21-1272. With 0.05%, 0.10%, and 0.15% of Sn, a mixture of anatase and rutile phases has been observed with rutile being the predominant one. Figure 1(c) suggests that the presence of Sn retards the anatase grain growth and accelerates anatase to rutile phase transformation. As Figure 1(d) shows, for 0.20% of Sn, a completely rutile solid solution is formed with no traces of anatase TiO2. All the peaks observed for this sample correspond to JCPDS data for rutile TiO2 (File no. 1276). Also, no peak for either metallic Sn or SnO2 could be observed even with 0.25 at% of Sn (Figure 1(d)). These results indicate that there is complete solubility of Sn in rutile TiO2 in the studied composition range and for about 0.20 at % of Sn and anatase TiO2 is completely transformed into the rutile form with a nominal composition of Sn0.20Ti0.80O2. The ionic radii being comparable ( 83 pm and 74 : 5 pm in octahedral coordination), Sn2+ ions can easily enter the TiO2 lattice to form a stable solid solution. Figure 1(b) show that the XRD spectra of the pure SnO2 show reflection from the (110), (101), (200), (220), and (310) and XRD pattern matched well with the rutile phase of SnO2. (JCPDS file no. 44-0872).

Apart from forming solid solution with TiO2, addition of SnO2 also affects the crystallite size and particle size of the composition. Table 1 summarizes the observations of structural properties of Sn-doped TiO2- and the particle size was evaluated from XRD by employing the Scherrer equation (d = 0.9λ/2β cosθ). An important observation is that the surface area increases and particle size decreases with increasing concentration of SnO2.

3.2. EDAX

Figure 2 shows the EDAX spectrum of 20 wt% TiO2-SnO2 nanopowder. The peak indicates that no elements other than Sn, O, and Ti were detected. The Au C peak originates from the gold sputtered sample from SEM measurement. The atomic percent of tin to titanium was determined as 80 : 20, which is good in agreement with the expected value. From this spectrum, it is found that strong signals from the tin, titanium, and oxygen atoms in the nanoparticles.

3.3. SEM

In Figure 3 the SEM image reveals that they possess the grain size of nanometer order and have nanoporous structure. It means that the structure is likely to facilitate the adsorption process of water molecules because of the capillary pore and large surface area.

4. Device Assembly

4.1. Fabrication Procedure

Device assembly has been shown in Figure 4. A glass box is designed of dimensions, length 42 cm, breadth 21 cm, and height 27.5 cm with a suitable door at one side. The door is used to place the sample inside the chamber and to analyze the humidity sensing property. This arrangement was helpful to increase the humidity inside the chamber. Inside the glass box, there are two stands, one for placing the solution which varies the relative humidity and the other for placing the sample which is in the pellet form.

The Cu-pellet-Cu electrode arrangement is fixed on a nonconducting base inside the glass box such that connecting wires remain outside the chamber. Saturated aqueous solution of potassium sulphate in a small container has been placed inside the glass box, and it increases the humidity inside box. The percentage relative humidity (RH %) inside the box is increased from 33.1% to 97.6%. Relative humidity in % is measured by using hygrometer. Variations in the resistance have been noted by using sinometer of MΩ order. All the measurements are carried out at room temperature. After each change of the humidity, the sensor element was exposed to the new humidity for 2 minutes before measuring the new resistance. Relative humidities ranging from 33.1% to 97.6% were obtained using saturated solutions as the humidity generation source. A list of saturated salts and their constant humidity values is given in Table 2.

To examine stability of the sensors, the humidity-resistivity characteristics at room temperature, between 33% to 97.6% RH, were measured after 5 days, and no important changes were observed. In fact, the sensor resistance slightly increased in room temperature after aging.

Humidity response was defined as the ratio of the resistance change occurred in a sample under exposure of the humidity to the original resistance where is the resistance in dry air and is the resistance in relative humidity. Response time is calculated by the time required for the sensor to attain the maximum change in resistance. The recovery time is also predicted by the time taken to get back their original resistance.

4.2. Humidity Sensing Properties Studies

Figures 5(a)5(c) show the variation of resistance as a function of the relative humidity for the pure TiO2, pure SnO2, and Ti0.80Sn0.20O2 at room temperature. As the RH increases, the resistance decreases by three orders in RH range 30, to 97%. The change is linear for the all composites except for pure SnO2, suggesting that mechanism of resistance variation is similar over the region of 30%–97%. However, in case of SnO2, the resistance () versus RH plot is clearly nonlinear for the RH > 60%. The nonlinearity of SnO2 sample might be due to the inhomogenous absorption of water molecule.

5. Results and Discussions

For the pure TiO2, pure SnO2 and various composition of SnO2 (5 wt%, 10 wt%, 15 wt%, 20 wt%, and 25 wt%), there will be a large decrease in resistance with increased relative humidity. It is due to the adsorption of water vapours on the surface with large surface area and capillary pores. The water vapors are adsorbed on the grain surface and in the pores. The nanoscaled grains lead to much more grain boundaries and nanopores, leading to more active sites available for condensed water to react. The high surface area provides more sites for water adsorption and produces more charge carriers for electrical conduction.

The conduction behavior of TiO2-SnO2 in the presence of water vapor is mainly due to ions. In this sensing material, Sn4+ cations are assumed to be highly active site for adsorption because of its highest charge density. These ions readily combines with hydroxyl group of adsorbed water and weaken the bond between hydroxyl ions and protons, and subsequently, water molecules gets dissociated. This provides free protons for electrical conduction. The hydroxyl ions thus remain chemically adsorbed at the cationic sites. Therefore, the first layer of adsorbed water molecule forms a chemisorbed layer and this process is irreversible at room temperature. The subsequent layers are physically adsorbed on the first layer to form hydroxyl multilayers by hydrogen bonding. These free mobile protons obtained from the first chemisorbed layer react with physisorbed water layer and form hydronium ions.

When physically adsorbed layer is abundant, the protons from one hydronium ion can hop to another hydronium ion, this proton transfer mechanism provides the mean for electrical conduction. The physisorption of water layer is reversible at room temperature, and it exists in equilibrium with ambient humidity. As the ambient humidity varies, the amount of water vapour physisorbed on the surface or condensed on the capillary pores varies, subsequently changing the concentration of the mobile protons, which in turn changes the conductivity of the ceramic.

6. Conclusions

Sn-doped nano-TiO2 was investigated for its humidity sensing characteristics. The sensing material annealed at 800°C gave nanocrystals of average size 20–50 nm which had good sensing properties for water vapour. Although the doping of Sn4+ increases the surface area of powders, responsible for humidity sensitivity, the segregation of Sn4+ at grain boundaries would be unfavorable factor as hydrophilicity of TiO2 would decrease. This is reflected in the study of various compositions in our study. gives better results with respect to sensitivity and minimum hysteresis than although these compositions are superior to the composition in terms of lower particle size and hence higher surface area.

This sensing element shows the maximum average sensitivity 7.5. Nearly all the sensing elements have shown similar trend that is at minimum RH% the resistance is maximum, and at maximum RH% the resistance is minimum. This study has significance in viewing of variations in properties of TiO2 after addition of tin oxide. The humidity sensor reported in this paper is easy to fabricate, cheap, friendly for users, and sensitive to the entire range of RH. This is applicable to commercial production.