Humidity-Sensing Performance of 3DOM WO3 with Controllable

Jan 16, 2018 - Humidity-Sensing Performance of 3DOM WO3 with Controllable Structural Modification ... performance is a great challenge in the field of...
2 downloads 0 Views 7MB Size
Research Article www.acsami.org

Cite This: ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Humidity-Sensing Performance of 3DOM WO3 with Controllable Structural Modification Zhihua Wang, Xiaoxiao Fan, Chunju Li, Geling Men, Dongmei Han, and Fubo Gu* State Key Laboratory of Chemical Resource Engineering, Beijing University of Chemical Technology, Beijing 100029, China

ABSTRACT: The development of humidity sensors with excellent sensing performance is a great challenge in the field of material chemistry. Here, we synthesized 3DOM WO3 nanomaterials through a poly(methyl methacrylate) template method, and first, we applied it to humidity measurement. For the goal of better sensing performance, the structural modification of Li/Kcodoping was adopted, and the test results showed that Li/K-codoped 3DOM WO3 possessed highly improved humidity sensing performances, such as high response, low-humidity hysteresis, good long-term stability, great repeatability, decent response, and recovery properties. To deeply understand the great effect of Li/K-codoping on sensing performance, the pure, Li-monodoped, and Li/K-codoped 3DOM WO3-based humidity sensors were compared, and we found that the structure defects and adsorbed oxygen as well as the co-effect of Li/K dopants were key factors for the improved sensing performance. Additionally, a possible humidity sensitive mechanism was proposed to further study the promotion effect of Li/K-codoping on humidity sensing process. KEYWORDS: 3DOM WO3, humidity, modification, Li/K-codoping, defects

1. INTRODUCTION Humidity sensing is of great importance in many industrial and domestic applications, including manufacturing process control, storage, living environmental monitoring, and so on. Recently, the humidity sensors gained from organic polymers,1 nanometer-scale metal−oxide semiconductor (MOS),2 and composite materials3 have been frequently reported. In particular, MOS have been attracting more and more interest in humiditysensing fields because of their processing flexibility. However, most humidity materials have poor response that hinders their further development. The realization of humidity sensors based on a novel material exhibiting enhanced humidity-sensing properties (e.g., high response, excellent stability, small hysteresis, fast response, and recovery speeds) has been a big challenge. Tungsten oxide (WO3), with stable chemical and physicochemical properties, has been attracting more and more attention in scientific research on gas sensing. The 6+ valence oxidation states of W can be reduced to other low valence states, which results in more surface defects.4 WO3 is made up of WO6 octahedral unit crystal structures arising from corner/ edge/face-sharing. This structure suggests that WO3 lattices can © XXXX American Chemical Society

withstand plenty of oxygen deficiencies as well as nonstoichiometric defects.5 According to previous report,6 a slight decrease of the oxygen content in WO3 might have huge influences on surface electronic states and greatly increase its reactivity so that big changes in the electrochemical properties will be made. The humidity-sensing process is generally based on the adsorption of water molecules, and structure defects play an important role in the adsorption process. Thus, WO3, with a variety of nonstoichiometric defects and oxygen vacancies, is a very promising material for humidity sensors. Furthermore, the opportunity to effectively generate and take advantage of structure defects in WO3 is of particular importance. One of the most common structural modifications to increase the amount of defects is element doping. Many literature studies have theoretically and experimentally demonstrated that an alkali metal especially Li+-doped WO3 possessed larger lattice distortion and more structure defects.6,7 There are also reports which demonstrate that the adsorbed water molecules could be Received: November 8, 2017 Accepted: January 4, 2018

A

DOI: 10.1021/acsami.7b17048 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 1. Schematic diagram of the experimental process and device.

Figure 2. SEM, TEM, and HRTEM images (inset of (c,g,k) is the corresponding Fourier transformation) of (a−d) pure 3DOM WO3; (e−h) Li monodoped 3DOM WO3; and (i−l) Li/K-codoped 3DOM WO3.

highly polarized so that more free ions of H+/H3O+ could be offered to conduction as a result of large charge density of Li+ ions.8−10 Thus, Li+-doped WO3 is an excellent material for humidity sensors. Moreover, the doping of K ions could further increase the conductivity of WO3 owing to their contribution as mobile carriers.11,12 Combining the above two points, such a material of WO3 modified by Li/K-codoping would thus create new opportunities for the development of humidity sensing. Notably, the three-dimensional (3D) pore structure, especially, 3D ordered materials (3DOM) can be applied in humidity sensors13 because they are not affected by mass transfer limitations. Water molecules are allowed to diffuse easily, thus improving humidity-sensing performance. Herein, this paper first reports a Li/K-codoped 3DOM WO3 humidity sensor with interesting features for practical applications, for instance, small hysteresis, short response− recovery time, good stability, and large response. By combining with the characterizations, we proposed that the excellent humidity-sensing performance for Li/K-codoped 3DOM WO3 is attributed to mainly two aspects: (1) large amounts of structure defects and adsorbed oxygen and (2) the co-effect of Li/K dopants, including the polarization effect on H2O of Li+ and the promoting effect on conductivity of K+. Moreover, the

unique well-interconnected pore structure might have great effect on enhancing humidity-sensing capability.

2. EXPERIMENTAL SECTION 2.1. Synthesis. The analytically pure agents were used for material synthesis. Some specific experimental steps are as follows: the orderly arranged poly(methyl methacrylate) (PMMA) microspheres with an average particle size of 330 nm were prepared by the hard template method as described elsewhere.14,15 Then, Li/K-codoped 3DOM WO3 was obtained from the PMMA colloidal crystal template method. A certain amount of 2.5 mol L−1 (NH4)6W7O14·5H2O and dilute solutions of LiNO3 and KNO3 (with Li/W and K/W molar ratios of 0.03) were added to the mixed solution of 3 mL deionized water and 3 mL anhydrous methanol by magnetic stirring. Subsequently, 1.0 g of PMMA microspheres were immersed in the previously prepared solution for 4 h, and the sample was separated by vacuum filtration. Finally, the sample was dried and annealed at 475 °C (ramping rate of 1 °C/min) for 3 h. To observe and study the co-effect of Li/K dopants on the humidity sensor, the pure and Li-monodoped 3DOM WO3 (also with Li/W molar ratio of 0.03) were synthesized through the same method but without addition of corresponding LiNO3 or KNO3. The schematic diagrams of the experimental process and the experimental device are illustrated in Figure 1. 2.2. Apparatus. S-4700 scanning electron microscopy (SEM) and H-800 transmission electron microscopy (TEM) recorded the material B

DOI: 10.1021/acsami.7b17048 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 3. (a) XRD spectra of 3DOM WO3 materials and (b) partial enlarged details of the (002), (020), and (200) peaks. morphology. High-resolution transmission electron microscopy (HRTEM) images were obtained by the J-3010 microscope. The material surface area was provided by the Brunauer−Emmett−Teller (BET) method (Micromeritics ASAP 2020). The crystal phase information of the 3DOM samples was measured by Bruker-D8 ADVANCE powder X-ray diffraction (XRD) with Cu Kα radiation (λ = 0.15406 nm). 2θ angle was scanned at a rate of 10° min−1 from 5° to 90°. The structural properties were characterized by VG Scientific ESCALAB 250X X-ray photoelectron spectroscopy (XPS). The binding energy reference was the C 1s peak at 284.0 eV. An X-ray fluorescence spectrometer (XRF-1800) was utilized to detect quantitatively the element contents of the samples. Diffuse reflectance UV−visible (UV−vis) spectra were collected using a JMNU-3010 UV−vis spectrophotometer. 2.3. Humidity-Sensing Measurements. The properties of humidity-sensing materials were estimated by detecting the impedance variation values of 3DOM WO3 sensors during exposure to different relative humidities (RHs) in the test apparatus. First, the prepared 3DOM materials were mixed with several drops of ethanol to form a homogeneous paste. Next, the paste was sprinkled on ceramic substrates with Ag−Pd interdigital arrayed electrodes. To improve their durability and stability, the prepared sensors were aged at 95% RH for 24 h.16 The humidity-sensing performances were assayed on a CHS-1 intelligent moisture-sensitive detection device (Beijing Elite Tech Co., Ltd, China). The operating voltage was kept at 1 V, and the working frequency changed between 50 Hz and 100 kHz. In our work, the response−recovery properties were tested using a saturated salt solution as a humidity generator. The sensor was switched between the 11% RH (saturated LiCl solution) and 95% RH (saturated KNO3 solution). The frequency characteristics, impedance characteristics, and hysteresis curves were tested by the DHD-II dual-flow humidity generation system, which provided humidity atmospheres for humidity testing by the dynamic air flow. The system automatically adjusted the proportion of moisture and dry air flow according to the preset humidity value to achieve a high-precision humidity atmosphere. The sensing measurements were performed at 25 °C.

The crystal phase information of 3DOM materials were characterized using XRD, and all the diffraction peaks can correspond to those of monoclinic WO3 (JCPDS no. 43-1035). Figure 3b shows the peaks of XRD signals of (002), (020), and (200) crystallographic planes of Li/K-codoped sample changing to lower angle degrees, which indicates that Li+ and K+ ions have been incorporated into WO3 lattices and thus induced local distortions. To comprehensively demonstrate the degree of lattice distortions, the lattice strains (ε) of WO3 materials are calculated according to the eq 119 ε = β /4 tan θ

(1)

As shown in Table 1, it is obvious that the Li/K-codoped sample presents the maximum lattice strain, indicating the Table 1. Crystal Structure Information of the WO3 Materials lattice parameters samples

a

b

c

lattice strain ε (%)

pure Li monodoping Li/K-codoping

7.306 7.295 7.332

7.532 7.537 7.566

7.688 7.688 7.714

0.22 0.16 0.30

largest lattice distortions, which may introduce abundant structure defects. In our previous study,6 we have demonstrated that the major reason that leads to structure defects is a typical connection between Li and O. Thus, we think the number of defects in the Li/K-codoped sample should be at least as much as that of the Li-monodoped sample. Besides, the larger lattice parameters of the codoped sample mean the larger lattice expansion due to K doping,20 which may result in the Li/Kcodoped sample possessing more structure defects. To further indicate defect contents and build associations between the defect and humidity-sensing performance, XPS characterization is conducted, as shown in Figure 4. Figure 4a suggests the full-range XPS spectrum. Figure 4b indicates that the W 4f peaks of monodoped and codoped materials shift to lower binding energies, reflecting the electronic interaction between WO3 and alkali metal dopants. The peaks of binding energies at 37.4 and 35.4 eV correspond to W6+ ions, whereas other peaks at 36.4 and 34.4 eV are attributed to W5+ ions in the WO3 lattice.21 The emission peak of Li 1s was not detected and may result from the small Scofield photoionization crosssection or the low doping ratio of Li. Similar results have also been reported.22−24 As for the oxidation state of the Li dopant, both theoretical and experimental studies25−27 have demonstrated that Li is doped in the interstices of monoclinic WO3 based on the way of cationic doping with the +1 oxidation state. The observation of K 2p core level peaks is shown in Figure 4c.

3. RESULTS AND DISCUSSIONS 3.1. Characterizations of 3DOM WO3. Figure 2 indicates the SEM, TEM, and HRTEM images of 3DOM samples, which all exhibit a highly ordered macroporous structure. HRTEM images with a Fourier transformation pattern demonstrate that 3DOM samples possess high crystallinity.17,18 It is worth noting that there is no significant difference existing in the morphology of the pure, Li-monodoped, and Li/K-codoped 3DOM WO3. Besides, the interplanar spacing of WO3 (020) facets and the BET surface area of three 3DOM samples are almost the same as well (12.9, 12.6, and 13.1 m2 g−1, respectively). Thus, the effect of morphology and surface area on humidity-sensing properties could be excluded, and our main focus is to be on the structure defects and adsorbed oxygen, as well as the coeffect of Li/K dopants. C

DOI: 10.1021/acsami.7b17048 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 4. XPS spectra of 3DOM samples: (a) full-range survey spectrum; (b) W 4f spectrum; (c) K 2p spectrum; and (d) O 1s spectrum [(a−c) pure 3DOM WO3, Li-monodoped, and Li/K-codoped 3DOM WO3 in (b,d)].

reports,30,31 the narrowed bandgap results from the large concentration of structure defects, especially oxygen vacancies. Therefore, we could compare the concentration of structure defects by comparing the size of the bandgap width of 3DOM samples. The optical bandgap Eopt is given by using the following eq 232

From its corresponding binding energies of 293.5 and 296 eV, it could be confirmed that K should also exist in the +1 oxidation state, which is supported by previous literature.28 From the results of XRD (Figure 3), no peaks of Li and K were observed, and the XRD signal peaks of (002), (020), and (200) planes of the doped materials shifted to lower degrees, which showed that Li and K were actually incorporated into WO3 lattices. The K content is 3.1% as measured by XRF, which agrees well with experimental amount added. In Figure 4d, the Olatt peak (530.2 eV), Odef peak (531.5 eV), and Oads peak (532.5 eV) belong to the lattice oxygen, the oxygen-deficient species (i.e., oxygen vacancies), and the adsorbed oxygen species (including O2− and O−), respectively.29 Table 2 lists the

(αhν)2 = A(hν − Eopt)

where α, hν, and A represent absorption coefficient, photon energy, and a constant, respectively. The diffuse reflectance spectra of 3DOM samples show in Figure 5a. Figure 5b indicates a plot of (αhν)2 ∼ hν. The Eopt value of the 3DOM material is the intercept on the X axis obtained by extrapolation. The corresponding Eopt values are marked in Figure 5b. It is evident that Li/K-codoped 3DOM WO3 possesses the narrowest bandgap, indirectly reflecting that it contains the most structure defects. Another reason why the bandgap is narrow might be due to the fact that the carriers with largest density occupy the lowest conduction band of the codoped sample. Thus, the novel Li/K-codoped 3DOM WO3 with more adsorbed oxygen as well as structure defects in addition to K+ as mobile carriers should show a highly improved humiditysensing performance. This reasonable assumption will be confirmed by the humidity-sensing tests. 3.2. Humidity-Sensing Properties of Li/K-Codoped 3DOM WO3. Figure 6a is the humidity-sensing behavior of 3DOM samples through monitoring the impedance variation to different humidities at 100 Hz. According to the results, it is evident that the Li/K-codoped sample exhibits the highest response (5 orders of magnitude of impedance variation). The calibration curves of three samples were found to be linear from 11 to 95% RH, and the slope of the straight line for the Li/Kcodoped sample is the largest, indicating that the sensitivity of this sensor is the best. As assumed, the order of impedance change of the sample is the same as that of the structural defect

Table 2. Relative Content of W and O Species in 3DOM Materials surface element composition samples Pure Li monodoping Li/K-codoping

5+

6+

W /W 0.58 1.07 1.47

Odef/Olatt

Oads/Olatt

0.21 0.39 0.44

0.13 0.27 0.30

(2)

relative contents of W and O species obtained from the fitting peaks. From Table 2, the Li/K-codoped sample possesses the most W5+ states (i.e., the most nonstoichiometric defects), adsorbed oxygen, and oxygen vacancy, which are all helpful to the humidity-sensing performance. Thus, we could draw a conclusion that the codoping of Li+ and K+ to the WO3 lattice may not only increase the density of structure defects but also accelerate the formation of adsorbed oxygen, which is helpful to water molecules adsorbing and enhancing its humidity-sensitive properties. The UV−vis characterization was conducted to further compare the amount of structure defects. According to previous D

DOI: 10.1021/acsami.7b17048 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 5. (a) UV−vis DRS spectra and (b) the plots of (αhν)2 ∼ hν of the 3DOM materials.

Figure 6. (a) Impedance of 3DOM samples under each RH; (b) impedance of the codoped sample at different frequencies under each RH; (c) realtime impedance of the codoped sample under different RHs; and (d) the hysteresis plot of the codoped sample. The error bars represent the standard deviation of three determinations.

As shown in Figure 7, this sensor also reveals good long-term stability in 30 days. To evaluate cross-sensitivity to other gaseous species that might be present in the ambient air, the gas sensing performances were performed on the CGS-4TPs intelligent gas sensitive detection device (Beijing Elite Tech Co., Ltd, China). The response of a sensor to a gas is defined as Ra/Rg

concentration. With more structure defects, the humidity sensor therefore adsorbs more water molecules, leading to a better humidity-sensing performance. For the codoped sample, the co-effect of Li/K dopants on the local charge density and the electrostatic field could further enhance the humiditysensing properties because of the accelerated dissociation of water molecules.9,10 Furthermore, a thicker electron depletion layer due to more adsorbed oxygen is another advantage of the Li/K-codoped sensor. More detailed discussions about the mechanism will be proposed later in this paper. Figure 6b proves that impedance has a good linear relationship with RH at 100 Hz. Therefore, 100 Hz was selected for the following tests. Then, we tested the response− recovery curves for six cycles from 11 to 95% RH. As seen in Figure 6c, the peak values remained nearly unchanged. The response and recovery times are 15 and 10 s, respectively. The Li/K-codoped sensor possesses good repeatability and decent response−recovery properties. Humidity hysteresis is utilized to evaluate humidity sensor reliability. The hysteresis diagram of the Li/K-codoped sensor is shown in Figure 6d. The maximum hysteresis was 3% RH from 11 to 95% RH, indicating that the Li/K-codoped 3DOM WO3 sensor possesses good reliability.

Figure 7. Stability of the Li/K-codoped sample. E

DOI: 10.1021/acsami.7b17048 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 8. (a) Influence of RH on the resistance of the Li/K-codoped 3DOM WO3 sensor. (b) Response of the sensor to 100 ppm different reducing gases (ethanol, acetone, ammonia, and CO) and 500 ppb oxidizing gas (NO2) at 25 °C for different RHs. The error bars represent the SD of four determinations.

Table 3. Comparison of the Present Work to Others in Previous Literature Studies material

order of impedance change

response time (s)

recovery time (s)

hysteresis (%)

refs

LiCl−C3N4 crosslinked polyelectrolyte ZrO2 LiCl−PEBAX nanofiber nanoporous polymers Li/K-codoped 3DOM WO3

3 4 4 4 3 5

0.9 19 14 30 3 15

1.4 10 36 80 75 10

3.5 4 4 negligible 4 3

33 34 35 36 37 present work

Figure 9. Humidity-sensing mechanism of 3DOM samples.

(reductive gas) or Rg/Ra (oxidative gas), in which Rg is the resistance of the sensor in the test gas and Ra is the one in air. As seen in Figure 8a, the humidity has an obvious influence on the resistance of the sensor based on Li/K-codoped 3DOM WO3, which is the principle of a humidity sensor design. Figure 8b proves that the response of the sensor to 100 ppm different reducing gases (ethanol, acetone, ammonia, and CO) are almost close to 1; that is, no obvious changes are found between Ra and Rg. Additionally, for 500 ppb oxidizing gas (NO2), when the humidity is equal to or higher than 45% RH, the results are similar to those above. The response of the sensor to NO2 at 25% RH is about 4.5, and the corresponding resistance changes of the sensor between the resistance in air and the one in NO2 is still in an order of magnitude range. Moreover, the concentrations of other gaseous species that might be present in the ambient air are far lower than the concentrations of the test gases. Therefore, the influence of other gaseous species on humidity detection can be neglected.

Table 3 summarizes the comparison of the humidity performance of our sensor with other humidity sensors available in the literature.33−37 According to Table 3, the Li/ K-codoped sample exhibits not only much higher response but also relatively short response−recovery time. Moreover, the reliability of Li/K-codoped 3DOM WO3 is satisfactory because 3% RH hysteresis is lower than that of other sensors. All these results show that Li/K-codoped 3DOM WO3 should be an ideal candidate for the humidity sensor. To intuitively exhibit the sensing mechanism, a depletion layer model based on the influence of adsorbed oxygen on the humidity sensor was suggested. As described in the literature,38 oxygen molecules will be adsorbed on the sensing material surface and extract electrons from the conduction band of WO3 to generate chemisorbed oxygen (O2−) when the temperature is below 100 °C, as shown in eqs 3 and 4

O2(gas) → O2(ads) F

(3) DOI: 10.1021/acsami.7b17048 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces O2(ads) + e− → O2(ads)−

of Grotthuss,39,40 H3O+ + H2O → H2O + H3O+, charge transport occurs among adjacent water molecules by releasing protons of H3O+. At the same time, K+ ions also could act as the mobile carriers to enhance the conductivity of the Li/Kcodoped sample, resulting in a sharp decline in impedance over 5 orders of magnitude. Finally, we can draw a conclusion that the enhanced humidity-sensing properties are mainly attributed to the structure defects and adsorbed oxygen as well as the coeffect of Li/K dopants.

(4)

Then, an electron depletion layer is generated on the material surface. At this time, Li/K-codoped 3DOM WO3 possesses a much increased impedance because the doping of K+ to the WO3 lattice could promote the defect contents of the adsorbed oxygen; that is, a dramatically thicker depletion layer is formed on the codoped sample at lower RH (11−54%), as shown in Figure 9. With RH increasing, the impedance of the codoped sample decreases more obviously; that is, the depletion layer becomes thinner than that of other samples. The reason is as follows. OH− from water molecules are combined with Li+ to form the chemisorbed layer, which contributes to form physisorbed multilayers through hydrogen bonds.35 The physisorbed water can be ionized to form a great quantity of hydronium ions (H3O+) as charge carriers by a Grotthuss chain, H3O+ + H2O → H2O + H3O+.39,40 After water layers of physical absorption gradually exhibit a liquidlike behavior, hydrogen ions can transfer between adjacent water molecules that contribute to conductivity.35 Therefore, Li/Kcodoped 3DOM WO3 will cause a more significant decrease in the depletion layer, which corresponds to a highly improved response of humidity sensors. For the sake of the humidity-sensitive mechanism, complex impedance plots were monitored to explain the conduction process in the Li/K-codoped sensor (Figure 10). To facilitate

4. CONCLUSIONS In summary, with the aim of achieving improvements in sensing performance, a novel humidity sensor based on Li/K-codoped 3DOM WO3 was fabricated, which showed outstanding humidity-sensitive performance, including high response, good repeatability, fast response, recovery speed, and small hysteresis. The sensing mechanism was studied by the model of the electron depletion layer and the complex impedance plots so that the structure defects, adsorbed oxygen, and the coeffect of Li/K dopants were considered as the main factors for the improved sensing performance. In the future, more detailed investigations will be carried out to gain deeper understanding on the humidity-sensing mechanism.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: 15810168819. ORCID

Zhihua Wang: 0000-0002-2719-5587 Fubo Gu: 0000-0003-2027-5178 Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (no. 21575011). REFERENCES

(1) Kulkarni, M. V.; Apte, S. K.; Naik, S. D.; Ambekar, J. D.; Kale, B. B. Ink-jet Printed Conducting Polyaniline Based Flexible Humidity Sensor. Sens. Actuators, B 2013, 178, 140−143. (2) Kim, H.-J.; Lee, J.-H. Highly Sensitive and Selective Gas Sensors Using P-type Oxide Semiconductors: Overview. Sens. Actuators, B 2014, 192, 607−627. (3) Rocha, K. O.; Zanetti, S. M. Structural and Properties of Nanocrystalline WO3/TiO2-based Humidity Sensors Elements Prepared by High Energy Activation. Sens. Actuators, B 2011, 157, 654− 661. (4) Barbagiovanni, E. G.; Reitano, R.; Franzò, G.; Strano, V.; Terrasi, A.; Mirabella, S. Radiative Mechanism and Surface Modification of Four Visible Deep Level Defect States in ZnO Nanorods. Nanoscale 2016, 8, 995−1006. (5) Bai, S.; Zhang, K.; Wang, L.; Sun, J.; Luo, R.; Li, D.; Chen, A. Synthesis Mechanism and Gas-sensing Application of Nanosheetassembled Tungsten Oxide Microspheres. J. Mater. Chem. A 2014, 2, 7927−7934. (6) Wang, Z.; Fan, X.; Han, D.; Gu, F. Structural and Electronic Engineering of 3DOM WO3 by Alkali Metal Doping for Improved NO2 Sensing Performance. Nanoscale 2016, 8, 10622−10631. (7) Tosoni, S.; Di Valentin, C.; Pacchioni, G. Effect of Alkali Metals Interstitial Doping on Structural and Electronic Properties of WO3. J. Phys. Chem. C 2014, 118, 3000−3006. (8) Sundaram, R. Comparative Study on Micromorphology and Humidity Sensitive Properties of Thick Film and Disc Humidity

Figure 10. Complex impedance diagram of the Li/K-codoped sample.

comparison of several complex impedance curves, the real part (Re Z) and the imaginary part (Im Z) are enlarged on the same plane. At low humidity (11% RH), the complex impedance curve is an arc with a large radius of curvature, which indicates a “non-Debye” behavior.41 Under this conditions, only a small amount of water molecules are adsorbed on the material surface in the form of chemisorption. Because the coverage of water is not continuous, the conduction process may be originated from the internal electronic conduction of Li/K-codoped 3DOM WO3, leading to a high impedance.42 With humidity rising from 33 to 54% RH, a straight line appears after the semicircle, which represents the Warburg impedance, and is a result from the diffusion process of ions at the electrode/sensing film interface.43 Under this situation, more water molecules are adsorbed onto the Li/K-codoped sample and the water molecules will be polarized because of the high charge density of Li+ ions, which are beneficial to the ion conduction process. At high humidity levels (75% RH or higher), the chemisorbed continuous water layer is formed. The subsequent water molecules are physically adsorbed on the hydroxyl layer by hydrogen bonding. On the basis of the ion-transfer mechanism G

DOI: 10.1021/acsami.7b17048 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces Sensors Based on Semiconducting SnWO4−SnO2 Composites. Sens. Actuators, B 2007, 124, 429−436. (9) Neri, G.; Bonavita, A.; Galvagno, S.; Donato, N.; Caddemi, A. Electrical Characterization of Fe2O3 Humidity Sensors Doped with Li+, Zn2+ and Au3+ ions. Sens. Actuators, B 2005, 111−112, 71−77. (10) Neri, G.; Bonavita, A.; Galvagno, S.; Pace, C.; Patanè, S.; Arena, A. Humidity Sensing Properties of Li−iron Oxide Based Thin Films. Sens. Actuators, B 2001, 73, 89−94. (11) Wang, Z.; Chen, C.; Zhang, T.; Guo, H.; Zou, B.; Wang, R.; Wu, F. Humidity Sensitive Properties of K+-doped Nanocrystalline LaCo0.3Fe0.7O3. Sens. Actuators, B 2007, 126, 678−683. (12) Anbia, M.; Fard, S. E. M. Improving Humidity Sensing Properties of Nanoporous TiO2−10 mol% SnO2 Thin Film by Codoping with La3+ and K+. Sens. Actuators, B 2011, 160, 215−221. (13) Wang, Z.; Tian, Z.; Han, D.; Gu, F. Highly Sensitive and Selective Ethanol Sensor Fabricated with In-doped 3DOM ZnO. ACS Appl. Mater. Interfaces 2016, 8, 5466−5474. (14) Sadakane, M.; Takahashi, C.; Kato, N.; Ogihara, H.; Nodasaka, Y.; Doi, Y.; Hinatsu, Y.; Ueda, W. Three-dimensionally Ordered Macroporous (3DOM) Materials of Spinel-type Mixed Iron Oxides. Synthesis, Structural Characterization and Formation Mechanism of Inverse Opals with a Skeleton Structure. Bull. Chem. Soc. Jpn. 2007, 80, 677−685. (15) Chen, X.; Ye, J.; Ouyang, S. Enhanced Incident Photon-toElectron Conversion Efficiency of Tungsten Trioxide Photoanodes Based on 3D-photonic Crystal Design. ACS Nano 2011, 5, 4310− 4318. (16) Zhang, Y.; Yuan, M.; Jiang, B.; Li, P.; Zheng, X. Effect of Mesoporous Structure on Bi3.25La0.75Ti3O12 Powder for Humidity Sensing Properties. Sens. Actuators, B 2016, 229, 453−460. (17) Waterhouse, G. I. N.; Metson, J. B.; Idriss, H.; Sun-Waterhouse, D. Physical and Optical Properties of Inverse Opal CeO2 Photonic Crystals. Chem. Mater. 2008, 20, 1183−1190. (18) Xie, S.; Deng, J.; Zang, S.; Yang, H.; Guo, G.; Arandiyan, H.; Dai, H. Au-Pd/3DOM Co3O4: Highly Active and Stable Nanocatalysts for Toluene Oxidation. J. Catal. 2015, 322, 38−48. (19) Prajapati, C. S.; Sahay, P. P. Influence of In Doping on the Structural, Optical and Acetone Sensing Properties of ZnO Nanoparticulate Thin Films. Mater. Sci. Semicond. Process. 2013, 16, 200− 210. (20) Mu, W.; Xie, X.; Zhang, R. Characterization of Li-doped WO3 Nanowires and Their Enhanced Electrocatalytic Oxidation of Ascorbic acid. RSC Adv. 2014, 4, 59740−59746. (21) Huo, N.; Yang, S.; Wei, Z.; Li, J. Synthesis of WO 3 Nanostructures and Their Ultraviolet Photoresponse Properties. J. Mater. Chem. C 2013, 1, 3999−4007. (22) Zhao, J.; Xie, C.; Yang, L.; Zhang, S.; Zhang, G.; Cai, Z. Enhanced Gas Sensing Performance of Li-doped ZnO Nanoparticle Film by the Synergistic Effect of Oxygen Interstitials and Oxygen Vacancies. Appl. Surf. Sci. 2015, 330, 126−133. (23) Ravichandran, C.; Srinivasan, G.; Lennon, C.; Sivanathan, S.; Kumar, J. Investigations on the Structural and Optical Properties of Li, N and (Li, N) co-doped ZnO Thin Films Prepared by Sol−gel Technique. Mater. Sci. Semicond. Process. 2010, 13, 46−50. (24) Wang, B.; Tang, L.; Qi, J.; Du, H.; Zhang, Z. Synthesis and Characteristics of Li-doped ZnO Powders for p-type ZnO. J. Alloys Compd. 2010, 503, 436−438. (25) Yang, C.; Chen, J.-F.; Zeng, X.; Cheng, D.; Cao, D. Design of the Alkali-Metal-doped WO3 as a Near-Infrared Shielding Material for Smart Window. Ind. Eng. Chem. Res. 2014, 53, 17981−17988. (26) Lee, J.; Cha, S.; Kim, J.; Nam, H.; Lee, S.; Ko, W.; Wang, K. L.; Park, J.; Hong, J. p-Type Conduction Characteristics of LithiumDoped ZnO Nanowires. Adv. Mater. 2011, 23, 4183−4187. (27) Garduño-Wilches, I. A.; Rodríguez-Fernández, L.; Alonso, J. C. Lithium Content and Low-Temperature Conductivity of NiO:Li Films Deposited by Spray Pyrolysis. J. Electron. Mater. 2015, 44, 4283. (28) Ghosh, S.; Khan, G. G.; Ghosh, A.; Varma, S.; Mandal, K. Zinc Vacancy-induced High-TC Ferromagnetism and Photoluminescence

in Group-1 Alkali-metal Substituted p-type ZnO Thin Films. CrystEngComm 2013, 15, 7748−7755. (29) Yang, C.; Zhu, Q.; Lei, T.; Li, H.; Xie, C. The Coupled Effect of Oxygen Vacancies and Pt on the Photoelectric Response of Tungsten Trioxide Films. J. Mater. Chem. C 2014, 2, 9467−9477. (30) Lim, K.-H.; Kim, K.; Kim, S. UV-visible Spectroscopic Analysis of Electrical Properties in Alkali Metal-doped Amorphous Zinc Tin Oxide Thin-film Transistors. Adv. Mater. 2013, 25, 2994−3000. (31) Sian, T. S.; Reddy, G. B. Optical, Structural and Photoelectron Spectroscopic Studies on Amorphous and Crystalline Molybdenum Oxide Thin Films. Sol. Energy Mater. Sol. Cells 2004, 82, 375−386. (32) Xue, S. W.; Zu, X. T.; Shao, L. X. Effects of Annealing on Optical Properties of Zn-implanted ZnO Thin Films. J. Alloys Compd. 2008, 458, 569−573. (33) Zhang, Z.; Huang, J.; Yuan, Q.; Dong, B. Intercalated Graphitic Carbon Nitride: a Fascinating Two-dimensional Nanomaterial for an Ultra-sensitive Humidity Nanosensor. Nanoscale 2014, 6, 9250−9256. (34) Jiang, K.; Fei, T.; Zhang, T. A Novel Crosslinked Polyelectrolyte Synthesized via a One-step Hydrothermal Process as a Humidity Sensor. RSC Adv. 2014, 4, 43189−43194. (35) Lu, Y.; Wang, Z.; Yuan, S. Microwave-Hydrothermal Synthesis and Humidity Sensing Behavior of ZrO2 Nanorods. RSC Adv. 2013, 3, 11707−11714. (36) Liang, S.; He, X.; Wang, F. Highly Sensitive Humidity Sensors Based on LiCl−Pebax 2533 Composite Nanofibers via Electrospinning. Sens. Actuators, B 2015, 208, 363−368. (37) Jiang, K.; Zhao, H.; Fei, T. A Guest/host Composite of Fe(NO3)3/Nanoporous Polytriphenylamine Assembly for Humidity Sensor. Sens. Actuators, B 2016, 222, 440−446. (38) Srinatha, N.; No, Y. S.; Kamble, V. B.; Chakravarty, S.; Suriyamurthy, N.; Angadi, B.; Umarji, A. M.; Choi, W. K. Effect of RF Power on the Structural, Optical and Gas Sensing Properties of RFsputtered Al Doped ZnO Thin Films. RSC Adv. 2016, 6, 9779−9788. (39) Ernsberger, F. M. The Nonconformist Ion. J. Am. Ceram. Soc. 1983, 66, 747−750. (40) Wang, S.; Chen, Z.; Umar, A.; Wang, Y.; Tian, T.; Shang, Y.; Fan, Y.; Qi, Q.; Xu, D. Supramolecularly Modified Graphene for Ultrafast Responsive and Highly Stable Humidity Sensor. J. Phys. Chem. C 2015, 119, 28640−28647. (41) Zhao, Y.; Ding, Y.; Chen, X.; Yang, W. Highly Sensitive and Quickly Responsive Core-shell Hierarchical Porous Li4Ti5O12-TiO2 Humidity Sensors. Sens. Actuators, B 2014, 203, 122−129. (42) Wang, L.; Duan, X.; Xie, W.; Li, Q.; Wang, T. Highly Chemoresistive Humidity Sensing Using Poly(ionic liquid)s. Chem. Commun. 2016, 52, 8417−8419. (43) Qi, Q.; Zhang, T.; Wang, S.; Zheng, X. Humidity Sensing Properties of KCl-doped ZnO Nanofibers with Super-rapid Response and Recovery. Sens. Actuators, B 2009, 137, 649−655.

H

DOI: 10.1021/acsami.7b17048 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX