A novel ethanol gas sensor based on TiO2/Ag0.35V2O5 branched nanoheterostructures

Much greater surface-to-volume ratio of hierarchical nanostructures renders them attract considerable interest as prototypical gas sensors. In this work, a novel resistive gas sensor based on TiO2/Ag0.35V2O5 branched nanoheterostructures is fabricated by a facile one-step synthetic process and the ethanol sensing performance of this device is characterized systematically, which shows faster response/recovery behavior, better selectivity, and higher sensitivity of about 9 times as compared to the pure TiO2 nanofibers. The enhanced sensitivity of the TiO2/Ag0.35V2O5 branched nanoheterostructures should be attributed to the extraordinary branched hierarchical structures and TiO2/Ag0.35V2O5 heterojunctions, which can eventually result in an obvious change of resistance upon ethanol exposure. This study not only indicates the gas sensing mechanism for performance enhancement of branched nanoheterostructures, but also proposes a rational approach to design nanostructure based chemical sensors with desirable performance.

suitable not only for TiO 2 , but also for other metal oxide based gas sensors 23 . Therefore, establishing heterostructures in sensor materials has long been regarded as the best strategy.
Recently, Silver vanadium oxides, such as AgVO 3 and Ag 0.35 V 2 O 5 , have attracted increasing attention for their application in batteries because of their unique electronic structure 24 . In particular, it has been reported that the electrical conductivity of Ag 0.35 V 2 O 5 nanowires is 0.5 S/cm, about 6-7 times higher than that of V 2 O 5 nanowires 25 , and the amine sensitivity of Ag 0.35 V 2 O 5 is much higher compared with V 2 O 5 particles 26 . Accordingly, it may be an interesting role to modify TiO 2 with Ag 0.35 V 2 O 5 to get enhanced gas sensitivity. However, to the best of our knowledge, there has been no report so far on the gas sensing performance of TiO 2 /Ag 0.35 V 2 O 5 composite. Furthermore, the emergence of nanostructures, such as one-dimensional (1D) nanomaterials (nanowires, nanorods, nanofibers), have led to improved sensitivity compared with conventional thin film due to their largely increased surface to volume ratio and rich surface chemistry on the nanostructure surfaces 27 .
Accordingly, in this paper, a novel ethanol gas sensor based on TiO 2 /Ag 0.35 V 2 O 5 nanoheterostructures with branched fiber-structures prepared by a facile one-step synthetic process is presented, in which well-matched energy levels are induced by the formation of effective heterojunctions between TiO 2 and Ag 0.35 V 2 O 5 , and at the same time, the branched-nanofiber structures display large Brunauer-Emmett-Teller (BET) surface area and complete electrons depletion for the nanobranches. By this way, the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures sensor exhibits higher selectivity, shorter response and recovery time, and higher sensitivity than pure TiO 2 nanofibers.

Results and Discussion
Structure and morphology. The TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures are composed of two phases: crystalline TiO 2 as the host, Ag 0.35 V 2 O 5 is introduced as the activators (right hand side of Fig. 1a). The process for fabricating the TiO 2 /Ag 0.35 V 2 O 5 heterostructures is based on a one-step electrospinning approach (Fig. 1a). Briefly, continuous PVP/tetrabutyl titanate/silver nitrate/vanadyl acetylacetonate (PVP/TBT/AgNO 3 / VO(acac) 2 ) nanofibers are prepared by means of electrospinning, and then the nanofibers are annealed in air ambient to crystallize the oxides and remove the PVP support (See the methods for details). The microstructures of the samples are investigated by SEM images. As shown in Fig. 1b,c, pure TiO 2 nanofibers with rough surface and uniform morphology can be observed, diameter of the nanofibers is approximately 220 nm and the length is about several micrometers. After introducing Ag 0.35 V 2 O 5 , nanofibers become thoroughly rougher and a great many nanobranches owing to the secondary growth of Ag 0.35 V 2 O 5 distribute uniformly on the surface of them, where diameter of the nanofibers is about 190 nm and that of nanobranches is about 20 nm (as shown in Fig. 1d,e). These novel branched nanostructures can provide more active sites for absorption of gas molecular and reaction of gas molecular with surface-adsorbed oxygen ions, thus would be benefit to the gas sensing response.
For a good understanding of the influence of the nanoheterostructure on the gas sensing performance, we use BET method of adsorption and desorption of nitrogen gas to measure the specific surface area of the TiO 2 / Ag 0.35 V 2 O 5 branched nanoheterostructures and pure TiO 2 nanofibers, as shown in Fig. 1f. The BET surface area of TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures calculated from the nitrogen isotherm is 21.15 m 2 g −1 , of about five times that of pure TiO 2 nanofibers (4.78 m 2 g −1 ). Obviously, the enhanced surface area of the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures is mainly attributed to the growth of nanobranches on the nanofibers surface.
XRD patterns have been employed to identify the phase composition and crystal structure of the samples (Fig. 1g). It can be seen that all the samples exhibit strong diffraction peaks, demonstrating the high crystallinity of the samples. The diffraction peaks of the pure TiO 2 nanofibers match the standard patterns of the rutile and anatase phase TiO 2 (PDF#21-1276, PDF#21-1272). As for the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures, several additional diffraction peaks can be clearly observed compared with the pure TiO 2 nanofibers, which can be indexed to the diffraction pattern of monoclinic Ag 0.35 V 2 O 5 (PDF#28-1027), indicating the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures composed of anatase TiO 2 , rutile TiO 2 , and monoclinic Ag 0.35 V 2 O 5 have been successfully prepared by the one-step electrospinning process. Moreover, the color of the two samples is very different, as can be clearly seen in Fig. S1, the color of TiO 2 nanofibers is white, while the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures turn to brown, indicating Ag 0.35 V 2 O 5 are successfully introduced to TiO 2 host, this can also be confirmed by the enhanced visible light absorption of the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures compared with the pure TiO 2 nanofibers (Fig. S2a). Additionally, the incorporation of Ag 0.35 V 2 O 5 leads to an increase of the phase transition of TiO 2 from anatase to rutile, this effect can also be observed in other TiO 2 based materials, such as TiO 2 /V 2 O 5 28 .
To further study the microscopic morphology and structure information of the as-synthesized TiO 2 / Ag 0.35 V 2 O 5 branched nanoheterostructures, TEM analysis is performed, as shown in Fig. 2. Branched-fiber-like structure of the TiO 2 /Ag 0.35 V 2 O 5 nanoheterostructures is clearly evidenced in Fig. 2a,b, where nanobranches of 10-20 nm in diameter are well dispersed on the surface of the nanofibers. HRTEM images of the backbone and branch defined by white boxes in Fig. 2b are shown in Fig. 2c,d, respectively. It can be seen that a strong alignment of two different crystal lattices resulted from the epitaxial growth of Ag 0.35 V 2 O 5 on TiO 2 is displayed obviously. The measured lattice distance of 3.5 Å corresponds to the (101) lattice distance of anatase TiO 2 , and the lattice fringe of 2.1 Å corresponds to the interplanar spacing of (106) planes of monoclinic Ag 0.35 V 2 O 5 . In addition, in order to further identify the elements distribution of the nanoheterostructures, STEM-EDS elemental mapping analysis is employed, as can be seen clearly from Fig To determine the chemical composition of the nanoheterostructures, XPS measurements are carried out in the region of 0-1050 eV ( Fig. 3 and Fig. S3), in which all binding energies are calibrated to the C 1s peak at 284.6 eV (Fig. S3b). The whole survey for all elements detection of the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures is presented in Fig. 3a, where O, V, Ti, Ag and C are detected. For comparison, the XPS whole survey of pure TiO 2 nanofibers is displayed in Fig. S3a, where only O, Ti, and C are detected. The two well resolved peaks at 458.6 and 464.2 eV observed from the Ti 2p core-level spectrum (Fig. 3b) can be ascribed to the Ti 2p3/2 and Ti 2p1/2 spin-orbital components, respectively, which are characteristic of a + 4 oxidation state of titanium 29 . The V 2p core-level spectrum of the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures is shown in Fig. 3c, the V 2p3/2 and V 2p1/2 peaks located at 517.1 and 524.6 eV is consistent with a + 5 oxidation state of the vanadium 30 . In addition, two small peaks at 515.8 eV and 523.0 eV indicate the appearance of V 4+ during the preparation process 30 . It is calculated that the molar ratio of V 4+ to V 5+ is 0.13. Fig. 3d shows that the silver species in the TiO 2 / Ag 0.35 V 2 O 5 sample include Ag + and metallic Ag 25 . The metallic Ag is not explored in XRD pattern may be because the little quantity and the no organization in a long range order. The atomic ratio of metallic Ag to Ag + in the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures is calculated to be 0.08, and thus the chemical composition of Ag 0. 35  Gas sensing properties. The resistance of the sensor is measured under the conditions of exposing the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures based sensor to ethanol vapor and dry air alternately. Sensor response to the gas is expressed with the normalized value R a /R g , where R a is the initial value in air and R g is the initial value in ethanol vapor exposure. In addition, the sensor's repeatability and sensor drift are studied by subsequent exposure-cleaning cycles. Due to good work function matching, the role of the contact between the semiconducting TiO 2 /Ag 0.35 V 2 O 5 and the gold electrodes seems to have a negligible effect on the conduction. It is well known that the response of a semiconductor metal oxide gas sensor is highly influenced by its operating temperature. Therefore, to begin with, ethanol vapor is used as the probe gas to perform gas-sensing tests at varying operating temperature to determine the optimum operating temperature. As shown in Fig. 4a, the sensing properties of two sensors to 100 ppm ethanol vapor are measured under different operating temperatures. Evidently, the output signal currents slightly increase with the increases of operating temperatures, indicating the decrease of resistance with temperature increasing. This temperature-dependent behavior of the samples is consistent with the normal semiconducting behavior. In addition, the relationship between the different operating temperatures and the corresponding sensor response is shown in insert figure of Fig. 4a. The sensitivity of the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures increases in relation to the operating temperature and reaches a maximum value of 31.8 at 350 °C. When the operating temperature increases beyond this value, the response value decreases due to the competition between adsorption and desorption of the chemisorbed gases. As for the pure TiO 2 nanofibers, the sensitivity value increases with the operating temperature marginally and reaches 4.4 at 450 °C. By this token, the introducing of Ag 0.35 V 2 O 5 can reduce the operating temperature evidently due to the heterojunction between TiO 2 and Ag 0.35 V 2 O 5 and the optimal operating temperature is determined to be 350 °C. Therefore, all sensing responses tests are further carried out at 350 °C for comparison.
The gas sensing performances of the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures and TiO 2 nanofibers for ethanol vapor are circularly tested and plotted in Fig. 4b. The change in resistance of sensors is measured during a time period of 50 seconds at a temperature of 350 °C in all the cases. It shows that the resistance decreases after the introduction of ethanol gas and reaches a saturation stage. When the supply of ethanol gas is stopped, the resistance starts to increase again and returns to its original value. This typically shows an n-type semiconducting behavior. It can be inferred that the TiO 2 nanofibers undergo a sensitivity value of about 3.5, whereas TiO 2 / Ag 0.35 V 2 O 5 branched nanoheterostructures exhibit a sensitivity value of about 31.8, which is more than 9 times compared with the pure TiO 2 nanofibers. For comparison, the gas sensing response of pure Ag 0.35 V 2 O 5 nanofibers and TiO 2 /V 2 O 5 fiber-like nanoheterostructures are also tested here (Fig. S4), where the response of Ag 0.35 V 2 O 5 nanofibers is about 5.8, while the TiO 2 /V 2 O 5 fiber-like nanoheterostructures exhibit improved gas sensing response of 24.8, indicating the hybridization of two semiconductors is much benefit to improve the gas sensing properties. Moreover, the better sensitive property of TiO 2 /Ag 0.35 V 2 O 5 nanoheterostructures sensor compared with TiO 2 /V 2 O 5 fiber-like nanoheterostructures sensor implies Ag 0.35 V 2 O 5 is an outstanding choice for TiO 2 modification to get enhanced ethanol sensitivity because of its excellent electrical conductivity 25,26 . In addition, the ethanol sensing properties of the TiO 2 /Ag 0.35 V 2 O 5 nanoheterostructures and other n-n type TiO 2 -based nanoheterostructures published in recent literatures are compared and shown in Tab. S1. It can be seen that the TiO 2 / Ag 0.35 V 2 O 5 nanoheterostructures sensor exhibits much higher ethanol gas sensing response compared with other competing nanoheterostructures 20,22,[31][32][33][34] , this highly sensitive ethanol sensing property demonstrates high potential of TiO 2 /Ag 0.35 V 2 O 5 nanoheterostructures for application in ethanol analysis. Furthermore, reproducibility, another important factor, is checked by repeating the response for ten times. It can be seen from Fig. 4b that both two samples exhibit outstanding reproducibility. The value for response and recovery times is also measured. The response time for TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures and pure TiO 2 nanofibers is calculated as 7 and 12 s, respectively, for 100 ppm of ethanol gas from the insert figure in Fig. 4b. Similarly, the recovery time of  Fig. 4c). This confirms the improvement in sensing for TiO 2 / Ag 0.35 V 2 O 5 branched nanoheterostructures. In summary, the sensor fabricated from TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures exhibits higher sensitivity, shorter response time/recovery time, and broader detection range from 20 to 1000 ppm for ethanol sensing, compared with those obtained by pure TiO 2 nanofibers.
To explore the selectivity of the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures sensor, other volatile organic pollutants (VOPs) including acetone, ammonia, methanol, and toluene are also measured under the same conditions and the result is shown in Fig. 4d. It is clear to see that the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures based sensor possesses a much higher response, not only to ethanol but also to ammonia and methanol, which are 31.8, 2.3, and 2.7, respectively, and are around 2-9 times compared with those of the pure TiO 2 nanofibers sensor. Selectivity is another important aspect of the gas sensing performances. In fact, a sensor with good selectivity can be used to detect a specific target gas when it is exposed to a multicomponent gas environment. From Fig. 4d, it can be concluded that among all the five tested gases, the response of the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures based sensor to ethanol is the highest, and is 17.7, 13.8, 11.8, and 19.9 times higher than those to acetone, ammonia, methanol, and toluene, respectively, indicating its good selectivity in detecting ethanol.
Gas sensing mechanism. Based on the above results, the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures based sensor shows excellent sensing properties. Herein, we propose an analogous model for the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures based sensor (as shown in Fig. 5). First, a heterojunction can be formed at the interface between TiO 2 and Ag 0.35 V 2 O 5 . Since the band gaps of TiO 2 extrapolated from the UV-Vis spectrum using Tauc's plot is close to the reported values in previous literature 20,22 (Fig. S2), we employ the standard literature energy levels of TiO 2 (conduction band of − 3.9 eV, valance band of − 7.1 eV, and Fermi level of − 4.2 eV, vs. vacuum level, respectively) for the energy bands matching analysis here. In addition, Mott-Schottky testing is used to ascertain the conduction band of Ag 0.35 V 2 O 5 here, the result shows that the conduction band of Ag 0.35 V 2 O 5 is − 5.12 eV vs. vacuum level (Fig. S5a). Considering the band gap of 2.1 eV extrapolated from the UV-Vis spectrum ( (Fig. S2), the valance band of Ag 0.35 V 2 O 5 should be − 7.22 eV. Obviously, when the n-type semiconductor TiO 2 and n-type semiconductor Ag 0.35 V 2 O 5 contact with each other, an n-n type heterojunction can be formed. Because the Fermi level of TiO 2 (− 4.2 eV) is higher than that of Ag 0.35 V 2 O 5 (− 5.37 eV, Fig. S6), the electrons in the Ag 0.35 V 2 O 5 will transfer to the TiO 2 and result in a band bending between TiO 2 and Ag 0.35 V 2 O 5 interfaces, thus an energy barrier can be formed at the heterostructure interface (Fig. 5a). Second, oxygen species are adsorbed on the surface of the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures in the air condition, and then are ionized into oxygen ions (O − , O 2− and O 2 − ) by capturing free electrons from the nanoheterostructures, thus leading to the formation of a thick depletion layer at the oxides surface and an increase of energy barrier height at the heterostructure interface (in air in Fig. 5a, step 1, 2, and 3 of Fig. 5b). Third, ethanol is a typical reductive gas, so when the sensor is exposed to ethanol gas, ethanol can react with the adsorbed oxygen species leading to the release of adsorbed electrons, the thinning of depletion layer at the oxides surface, and the decrease of energy barrier height at the heterostructure interface (in ethanol in Fig. 5a, step 4 and 5 of Fig. 5b). The mechanism can be explained by several chemical reactions, which are shown as follows: From the above reactions, it can be seen that the trapped electrons will be released to the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures after the supply of ethanol gas, thereby the carrier concentration and electron mobility on the sensor surface will be increased, then the depletion layer width and the energy barrier height will Scientific RepoRts | 6:33092 | DOI: 10.1038/srep33092 decrease and the resistance decrease accordingly. On the other hand, electrons on the conduction band will be captured by oxygen molecules adsorbed on the surface of the materials to form oxygen ions (O − , O 2− and O 2 − ) after stopping ethanol gas supply, the depletion layer width and the energy barrier height will increase again, thus leading to an increase in resistance.
Therefore, a probable reason for the enhanced sensing properties of the TiO 2 /Ag 0.35 V 2 O 5 is related to the extraordinary branched-nanofiber structures with branch diameter of about 20 nm and fiber diameter of about 160 nm according to the SEM results. On the one hand, the large BET surface area of the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures can be ascribed as one of most important factor for enhanced sensing performance. With the introducing of Ag 0.35 V 2 O 5 , the pure TiO 2 nanofibers are transformed into branched-nanofibers, and the BET surface area of the nanoheterostructures is increased to 21.15 m 2 g −1 , while for TiO 2 nanofibers it is only 4.78 m 2 g −1 (Fig. 1f). This can provide more active sites for absorption of ethanol and reaction of ethanol with surface-adsorbed oxygen ions, thus the resistance decrease becomes more noticeable, and the gas sensing response is enhanced accordingly. On the other hand, electron exchange between the surface states and materials occurs within the surface layer, and the width of it is the order of the Debye length L D , which can be expressed by the following equation: where k is the Boltzmann constant, T is the absolute temperature, ε is the static dielectric constant, ε 0 is the permittivity of vacuum, q is the electrical charge of the carrier, and n c is the carrier concentration. For the TiO 2 / Ag 0.35 V 2 O 5 branched nanoheterostructures fabricated in this study, n c of Ag 0.35 V 2 O 5 extrapolated from the Mott-Schottky plot is about 9.6 × 10 18 cm −3 (Fig. S5a), ε of Ag 0.35 V 2 O 5 is measured to be 360 (see the methods for details). Accordingly, L D is estimated to approximately 10 nm for Ag 0.  channel of TiO 2 /Ag 0.35 V 2 O 5 nanoheterostructures is much influenced by the energy barrier. The resistance of the heterojunctions can be expressed by the following equation: (6) where B is a constant, k is the Boltzmann constant, T is the absolute temperature and qΦ is effective energy barrier at the heterojunction. For air condition, the effective energy barrier (qΦ ) increases because the free electrons are captured by oxygen species to ionize into oxygen ions (O − , O 2− and O 2 − ) (as shown in the first figure in Fig. 5a). After exposure in ethanol gas, ethanol can react with the adsorbed oxygen species and lead to the release of adsorbed electrons, thus leading to the decrease of the energy barrier (the second figure in Fig. 5a). It is obvious that R a /R g is in direct proportion to the value of exp(Δ qΦ ), so the remarkable changes of energy barrier of the heterojunctions can induce great change in the conductivity and improvement of the gas-sensing performance 32 , which can be entitled as synergistic effect. Additionally, the heterojunctions can also be used for additional active sites, leading to an improvement in the sensing performances 37,31 . What is more, the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures act as a more efficient catalyst than pure TiO 2 nanofibers 38 , which can promote the sensing reaction between the reductive VOPs and adsorbed oxygen species 34 .
From all the above, the high performance of the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures gas sensor for ethanol can be ascribed to the following two factors. First, the enhancement in gas sensing is believed to be related to the novel branched-nanofiber structure, which display larger BET surface area and completely electrons depletion for nanobranches compared with the pure TiO 2 nanofibers. Secondly, the synergistic effect, additional active sites, and efficient catalytic capability induced by the effective heterojunctions between TiO 2 and Ag 0.35 V 2 O 5 also contribute to the gas sensing enhancement.
In conclusion, we have demonstrated a high ethanol sensitivity and selectivity for TiO 2 /Ag 0. 35  nanoheterostructures were prepared by an electrospinning process followed by an annealing treatment 39 . First, 0.50 g tetrabutyltitanate (TBT) and 0.20 g polyvinylpyrrolidone (PVP) were dissolved in a mixture of 1.50 ml ethanol and 1.20 ml acetic acid, and stirred for 20 min to give PVP/TBT composite. Then, 0.60 g PVP, 0.20 g VO(acac) 2 , and 0.035 g Ag(NO) 3 were added into 3.70 g dimethylacetamide (DMAC), after stirring for 20 min, the resulting solution was mixed with the PVP/TBT composite prepared in the first step and stirred for 1 h to prepare PVP/TBT/Ag(NO) 3 /VO(acac) 2 composite. Next, the PVP/TBT and PVP/TBT/Ag(NO) 3 /VO(acac) 2 composites were electrospun and then annealed at 450 °C in ambient air for 1 h to remove the PVP support, crystallize TiO 2 and Ag 0.35 V 2 O 5 , and finally resulted in TiO 2 nanofibers and TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures (Ti/V molar ratio is 1), respectively. In a typical electrospinning process, the spinneret had an inner diameter of 0.4 mm. A distance of 15 cm and DC voltage of 15 kV were maintained between the tip of the spinneret and the collector. Additionally, Ag 0.35 V 2 O 5 nanofibers and TiO 2 /V 2 O 5 fiber-like nanoheterostructures for comparison were prepared by the same electrospinning process using PVP/Ag(NO) 3  Characterization and gas sensing measurements. The morphologies of the samples were characterized by field emission scanning electron microscopy (FESEM, Ultra 55) and transmission electron microscopy (TEM, Libra 200FE). X-ray diffraction (XRD, CuKα , λ = 1.5406 Å, X'Pert PRO) and high-resolution TEM (HRTEM) were employed to characterize the crystal structure and elemental analysis of the samples. Nitrogen adsorption-desorption isotherms (ASAP 2020 nitrogen adsorption apparatus) was employed to measure the Brunauer-Emmett-Teller (BET) specific surface areas of the samples. The chemical composition was determined by X-ray photoelectron spectroscopy (XPS), and the measurements were performed in a VG Scientific ESCALAB 210 spectrometer equipped with Mg anode and a source power of 300 W. All binding energies were calibrated to the C 1s peak at 284.6 eV. The UV-Vis absorption spectra were recorded using a UV-3150 spectrophotometer to evaluate the absorption properties. Mott-Schottky testing was performed at an electrochemistry workstation (RST5200) to obtain the semiconductor type, carrier concentration, and conduction band energy of the samples. The measurements were performed in a three-electrode cell with 0.2 M Na 2 SO 4 (PH = 6.5) at a frequency of 1 kHz and scan rate of 10 mV/s, where Pt wire was used as the counter electrode and Ag/AgCl electrode was used as the reference electrode. The potential was measured against an Ag/AgCl reference electrode and converted to NHE potentials using E (NHE) = E (Ag/AgCl) + (0.059 × pH) + 0.197 V. The Fermi energy level of the Ag 0.35 V 2 O 5 sample was measured by the Kelvin probe force microscopy (KPFM) using the SII E-Sweep SPM system in air condition at room temperature across the Au/Ag 0.35 V 2 O 5 border, which was formed at the surface of Ag 0.35 V 2 O 5 by depositing a stripe of Au film. Furthermore, the static dielectric constant was tested using an Agilent 4294A Precision LCE Meter (Agilent Technologies Inc.) at the frequency of 10 MHz.
Scientific RepoRts | 6:33092 | DOI: 10.1038/srep33092 The preparation of the gas sensor was similar to that depicted in previous literature 40 . The sensor device was prepared by dispersing the TiO 2 /Ag 0.35 V 2 O 5 branched nanoheterostructures into ethanol to form a paste and coated onto the outside surface of an alumina tube which was printed a pair of Au electrodes previously. Then, the sensor devices were dried at 150 °C for 3 h in ambient air to form sensor film. Finally, a Ni-Cr alloy wire was inserted into the alumina tube and employed as a heater, the operating temperatures were controlled by adjusting the heating power of the alloy. The gas-sensing properties were measured under a steady-state condition by using a high precision sensor testing system (WS-30A). The device was examined at 50% relative humidity in the temperature range of 250-450 °C at various concentrations of ethanol (20-1000 ppm). The sensor response was defined as S = R a /R g , where R a is the resistance in air and R g is the resistance in the probe gas. The response time was defined as the time needed for the variation in electrical resistance to reach 90% of the equilibrium value after injecting ethanol, and the recovery time was defined as the time needed for the sensor to return to 90% above the original resistance in air after removing the ethanol.