Template-free preparation of mesoporous single crystal In₂O₃ achieving superior ethanol gas sensing performance

Abstract

Mesoporous single crystal indium oxide (In₂O₃) has been successfully prepared via a hydrothermal annealing route in the absence of templates and surfactants. An In₂O₃ mesoporous single crystal has larger surface areas and smaller pore diameter, and thus exhibits better ethanol gas sensing properties than mesoporous In₂O₃ with a large pore diameter which was prepared by adjusting the pH value.

---

The mesoporous single crystal (MSC) metal oxides have attracted great attention because of their high porosity, large specific surface areas, good thermal stability and superior charge transport capability.^1–3 These intriguing properties endow the MSC metal oxides with a high potential in various fields from energy conversion to gas sensors.^4–7 To date, many synthetic routes using template, including soft and hard template approaches, have been widely employed to construct MSC structures.^8–12 The hard template methods have demonstrated effectiveness in forming MSC metal oxides. However, thermal or chemical treatments are sometimes required to remove the template, and these post-processing techniques are not only tedious but also often destroy the desired product structure. The soft template method requires complicated control of the assembly processes and normally results in mesoporous (M) materials in amorphous or semi-crystalline phases.^13,14 In general, it remains a major challenge to develop a simple and template-free route for preparing MSC nanostructures of metal oxides.

It is evident that the mesoporous structure can dramatically increase the surface area and surface-to-volume ratio of metal oxides, which results in an improvement in the reaction efficiency and enhanced properties in surface related applications. It is well known that gas sensing behavior is a predominantly surface dependent phenomenon and the surface-to-volume ratio of the material shows a significant influence on the gas sensing performance. Thus, it is necessary to control the surface characteristics of the sensing material with the aim of achieving superior performance. In particular, the generation of high specific surface areas and plentiful mesopores will result in a higher probability for gases to diffuse and interact with the sensing materials, which is likely to increase the sensitivity, response/recovery time and stability.^15,16 Therefore, it is of great importance to fabricate a gas sensor using mesoporous metal oxides as the gas sensing material. In addition to the mesoporous characteristic, the ideal structure is considered to be single crystal, because such gas sensing materials can exhibit high electron conductivities, and have abundant interfacial active sites and excellent thermal stability.

Among the numerous metal oxides available, indium oxide (In₂O₃) is a promising material for application as a sensor for various oxidizing gases and reducing gases.^17,18 However, In₂O₃ usually has several critical limitations such as relatively low sensitivity, long response/recovery time and poor stability. For enhancing sensitivity and response/recovery time, In₂O₃ should possess large surface-to-volume ratios and porosity to facilitate high accessibility and reactivity for the gas molecules. For improving the thermal stability and electron transport, In₂O₃ should exhibit single-crystalline characteristics. Accordingly, fabrication of MSC In₂O₃ is of great significance. Until now, mesoporous In₂O₃ was prepared using template methods.^19–23 Nevertheless, as far as is known, there have been few reports on a MSC In₂O₃ gas sensing material prepared without templates or surfactants.

In this work, to extend the preparation method of MSC materials and to extend the understanding of the MSC structure for their related applications, In₂O₃ was selected as a model compound to study the gas sensing properties. MSC and mesoporous In₂O₃ particles with a large pore diameter have been successfully prepared via a hydrothermal annealing route in the absence of templates or surfactants. The pore and particle size were controlled by adjusting the pH value of the reactive solution. To demonstrate the value of MSC In₂O₃ in practical applications, its ethanol (C₂H₅OH) gas sensing capability was tested. Compared with M In₂O₃, the prepared MSC In₂O₃ exhibited a superior C₂H₅OH gas sensing performance thanks to its mesoporous and single crystal characteristics.

Fig. 1a shows a typical X-ray diffraction (XRD) pattern of the MSC In₂O₃ product prepared through annealing corresponding indium(III) hydroxide [In(OH)₃] precursors, which were synthesized via a hydrothermal process in a solution with a pH value of 11 and indexed to cubic In(OH)₃ (JCPDS no. 76-1464) (ESI, Fig. S1†). The sharp diffraction peaks of MSC In₂O₃ were indexed well to In₂O₃ (JCPDS no. 76-0152). The crystallite size of MSC and M In₂O₃ specimens was 28.1 and 33.2 nm, respectively, using Scherrer's formula to calculate it. Scanning electron microscopy (SEM) and low-magnification transmission electron microscopy (TEM) images shown in Fig. 1b and c demonstrate that as-calcined In₂O₃ is in the form of nanoparticles, which is consistent with that of In(OH)₃ (ESI, Fig. S1b†). The high resolution (HR) TEM image in Fig. 1d shows that many mesopores are randomly arranged in In₂O₃ particles and the shape of the pores is irregular. As observed from the TEM image of In(OH)₃ (ESI, Fig. S2†), there are no mesopores, indicating that the mesopores are created in the conversion process of In(OH)₃ to In₂O₃. It can be concluded that in situ release of crystalline water (H₂O) leads to the formation of pores. The clear lattice fringes shown in the high resolution TEM image (Fig. 1e) suggest that the mesoporous In₂O₃ are highly crystalline. The different interplanar distances, d = 0.72 and 0.41 nm correspond to the (110) and (220) lattice planes, respectively, and many mesopores (red circles in Fig. 1e) with a uniform diameter of ∼2.5 nm exist. The selected area (electron diffraction (SAED) pattern consisting of distinct spots, shown in Fig. 1f, further indicates the single crystal characteristic of mesoporous In₂O₃. The previous results discussed demonstrate that the as-calcined In₂O₃ possesses an MSC structure.

Fig. 1 Crystal structure and morphology of MSC In₂O₃ particle: (a) XRD pattern, (b) SEM image, (c, d) TEM images, (e) HRTEM image and (f) SAED pattern.

Brunauer, Emmett and Teller (BET) measurement was performed to confirm the mesoporous characteristic (Fig. 2). The surface areas of the two samples were 39.0 m² g⁻¹ and 26.7 m² g⁻¹, respectively. The surface areas of MSC In₂O₃ prepared in the present work are larger than other In₂O₃ mesostructures prepared using surfactants such as Span-40.²⁴ Compared to MSC In₂O₃ (Fig. 1b), M In₂O₃ have a larger particle size (ESI, Fig. S5†). The isotherm of MSC In₂O₃ in Fig. 2a shows is type-IV with a hysteresis loop. The distinct hysteresis loops can be observed at a relative pressure range from 0.4 to 0.9. The Barrett–Joyner–Halender (BJH) pore size distribution plots of MSC In₂O₃ particles (inset of Fig. 2a) exhibit an average pore diameter of 2.4 nm, which is in agreement with diameters obtained from the TEM images. In addition, the pore size distribution for diameters less than 2 nm cannot be displayed because of the test limitation of the analysis instrument. The isotherm in Fig. 3b shows a type-IV with a hysteresis loop. The distinct hysteresis loops can be observed at a relative pressure range from 0.8 to 1. The BJH pore size distribution plots of M In₂O₃ particles show the average pore size to be about 45.7 nm (inset of Fig. 3b). The different pore structures and gas adsorption ability of MSC and M In₂O₃ were further confirmed by the corresponding Fourier-transform infrared (FTIR) spectra (ESI, Fig. S8†). MSC In₂O₃ exhibits a stronger adsorption ability of carbon dioxide (CO₂) and H₂O than M In₂O₃, which indicates that there is a higher porosity and more surface defects. Furthermore, these characteristics are created after the formation of In₂O₃ because the intensity of CO₂ and H₂O adsorption peaks are almost the same in the FTIR spectra of MSC and M In(OH)₃ precursor (ESI, Fig. S6†).

Fig. 2 Nitrogen adsorption/desorption isotherms and pore size distributions (inset) of In₂O₃ samples: (a) MSC In₂O₃ and (b) M In₂O₃.

Fig. 3 Gas sensing performance of In₂O₃ sensors: (a) gas responses to 1000 ppm of C₂H₅OH gas as a function of operating temperature, (b) response and recovery time to 1000 ppm of C₂H₅OH gas, (c, d) transient response curves to 100–1000 ppm of C₂H₅OH gas, (e) selectivity to 1000 ppm of different gases and (f) repeated response of MSC In₂O₃ sensor to 1000 ppm of C₂H₅OH gas.

To evaluate the gas sensing performance of In₂O₃ materials with different pore diameters, C₂H₅OH was chosen as the target gas for detection. For semiconducting metal oxide gas sensors, operating temperature is a critical factor influencing gas sensing performance because the amount of ionized oxygen species (O₂⁻, O⁻, O²⁻) on the surface of the metal oxide changes with temperature, leading to changes in both sensor resistance and response.^25,26 As shown in Fig. 3a, each sensor has an optimal operating temperature, at which the sensor exhibits the highest response to C₂H₅OH gas. The MSC In₂O₃ sensor has not only a lower optimal operating temperature but also a higher gas response than the M In₂O₃ sensor. The optimal operating temperature of 225 °C is then applied in all investigations described further on in this paper. Fig. 3b shows that the response time of the two kinds of sensors is about 3–6 s and the recovery time is 5–6 s after exposure to 1000 ppm C₂H₅OH gas, which indicates that the mesoporous structure is favorable for the adsorption and desorption of gas molecules. To illustrate the excellent properties of the MSC In₂O₃ sensor, the response of the MSC In₂O₃ sensor was also compared with that of the M In₂O₃ sensor to different concentrations of C₂H₅OH gas. Fig. 3c and d shows the dynamic response curves of two kinds of In₂O₃ sensors to different C₂H₅OH concentrations ranging from 100 ppm to 1000 ppm. As can be seen, the sensitivities of the gas sensors increase rapidly with the enhancement of C₂H₅OH concentration. Sensors based on MSC In₂O₃ have a higher sensitivity than the M In₂O₃ sensor. Fig. 3e shows the selectivity of the two sensors based on MSC and M In₂O₃ to different gases, including C₂H₅OH, hydrogen (H₂), CO₂, methane (CH₄) and carbon monoxide (CO). It was observed that both kinds of In₂O₃ sensors prefer to respond to C₂H₅OH out of all the different gases of the same concentration, and the MSC In₂O₃ exhibited better selectivity than the M In₂O₃. The highest response value of the MSC In₂O₃ sensor was 152 to 1000 ppm C₂H₅OH gas, whereas the responses to other gases were less than 10. Selectivity of the In₂O₃ sensor could be ascribed to two reasons. One possible reason is that In₂O₃ has a low optimal operating temperature (225 °C) to C₂H₅OH gas. Some gases, such as CH₄, cannot dissociate and react with ionized oxygen species on In₂O₃ at such a low temperature.²⁷ Another reason might be that In₂O₃ has an enhanced adsorption and catalytic capability towards C₂H₅OH gas.²⁸ Fig. 3f shows a series of repeated responses of the MSC In₂O₃ sensor to 1000 ppm of C₂H₅OH gas, which provides a high response, short response/recovery time and good stabilization. The sensor's response did not decay after eight cycles of switching on/off from dry air to gas and back to dry air. This excellent performance could be attributed to its MSC characteristics.

As discussed previously, the mesopores on the sample are created in the conversion process of In(OH)₃ to In₂O₃. The formation of pores is attributed to the loss of crystalline water, as described previously in the literature.^29,30 According to the thermogravimetry/differential scanning calorimetry (TG/DSC) curves of MSC and M In(OH)₃ precursors (ESI, Fig. S4†), two dehydration reactions could take place at about 140 °C and 260 °C, respectively. The reaction process can be described as follows:

In(OH)₃ → InOOH + H₂O (1)

2InOOH → In₂O₃ + H₂O (2)

In these procedures, the grain boundaries divide and move to opposite directions, H₂O molecules are released from the precursor, and molecule sized pores are formed. When the In(OH) precursor is a small particle, a small amount of crystalline water is released and the molecule sized pores continually grow into a mesoporous structure. When the In(OH)₃ precursor is a large particle, a large amount of crystalline water is released and the molecule sized pores continually grow into a large mesopore structure. Contrasted with M In₂O₃, the MSC In₂O₃ has a relatively high specific surface area because the smaller mesoporous structure provides plenty of inner surfaces. The dehydration process leads to the decrease of the In₂O₃ particle size. Compared with the size of MSC and M In(OH)₃ precursor in size distribution histograms (ESI, Fig. S3†), the average size of the MSC and M In₂O₃ particles all shrink (ESI, Fig. S7†). Finally, in situ dehydration and shrinkage result in the formation of MSC In₂O₃.

The formation of mesoporous structure could produce more oxygen vacancies and lead to the enhanced gas sensing property.³¹ Recent studies also reveal that the surface structures might be the essential factor for determining the efficiency of gas sensing properties.^32,33 In order to obtain the surface chemical status of two samples with different pore diameter, X-ray photoelectron spectroscopy (XPS) analyses were carried out to investigate the oxygen vacancy defects of MSC and M In₂O₃ samples (Fig. 4). The In 3d core level spectra of the two In₂O₃ samples showed two peaks located at 443.9 and 451.4 eV in Fig. 4b, which could be attributed to the characteristic spin–orbit split 3d_5/2 and 3d_3/2, respectively. Fig. 4a displays the O1s core level spectra of the two In₂O₃ products, and two peaks can be clearly identified from the spectra.

Fig. 4 The XPS spectra of MSC and M In₂O₃ particles: (a) the O1s core level spectra; (b) In 3d core level spectra.

One peak at 529.3 eV is deemed to be the binding energy of O atoms in the In–O–In bond which corresponded to the oxygen-sufficient state, in which In atoms are in full complement with their neighboring O atoms (crystal lattice oxygen). Whereas the other peak located at 531.1 eV can be attributed to oxygen defects, which corresponded to the oxygen-deficient state.³⁴ However, the ratios of the oxygen-deficient area to the oxygen-sufficient area for MSC and M In₂O₃ are 0.734 and 0.708, respectively. It was also confirmed that there are more oxygen vacancies for MSC In₂O₃ than M In₂O₃. The oxygen vacancies induced by the mesoporous structure make MSC In₂O₃ more sensitive and selective than M In₂O₃.

In principle, the gas sensing of semiconducting metal oxides involves an adsorption–oxidation–desorption process, which happens on the surface of the semiconductors. Therefore, the gas sensing ability of In₂O₃ is theoretically very sensitive to its surface structure. As highlighted in the preceding sections, the mesoporosity coupled with single crystallinity provides for the ease of gas diffusion and more active sites for the formation of reactive oxygen species. As a result, the enhanced gas sensing performances should be ascribed to higher surface areas and oxygen vacancy concentration of MSC In₂O₃ than M In₂O₃. When In₂O₃ sensors are exposed to air (Fig. 5a), more oxygen molecules would be adsorbed on the MSC In₂O₃ surface than on the M In₂O₃ surface and generate more ionized oxygen species (O₂⁻, O⁻ and O²⁻) by trapping more electrons from the conduction band of In₂O₃ (ref. 35), leading to the high resistance state (ESI, Fig. S9†). Once In₂O₃ sensors are exposed to C₂H₅OH gas (Fig. 5b), the ionized oxygen species on the In₂O₃ surface will directly react with C₂H₅OH molecules to produce CO₂ and H₂O. Those electrons trapped by chemisorbed oxygen are released back to the conduction band, resulting in the low resistance state. As shown in Fig. S9 (ESI),† MSC In₂O₃ exhibits a larger resistance change than M In₂O₃, and thus leads to a better gas sensing performance.

Fig. 5 Gas sensing mechanism of MSC (lower particle) and M In₂O₃ (upper particle): (a) high resistance state in air, (b) low resistance state in C₂H₅OH gas.

Conclusions

In summary, an MSC In₂O₃ structure prepared via an easy hydrothermal approach combined with annealing process is reported. MSC In₂O₃ with a small particle size exhibited an enhanced gas sensing performance compared with mesoporous In₂O₃ with a large pore diameter and particle size. The significant role of the MSC structure is demonstrated by it providing higher effective surface areas and oxygen vacancy concentration for gas sensing performance than the large mesopore structure. These findings create new opportunities for designing functional materials with specific structures for different application areas from energy conversion to gas sensors.

Acknowledgements

This work was financially supported by projects of the Natural Science Foundation of China (21271055, 21471040, and 61473095), the Fundamental Research Funds for the Central Universities (HIT. IBRSEM. A. 201410), the Open Project of State Key Laboratory of Urban Water Resources and Environment, Harbin Institute of Technology (QAK201304) and the Natural Science Foundation of Heilongjiang Province (ZD2015014).

Notes and references

1. D. Li, H. Zhou and I. Honma, Nat. Mater., 2004, 3, 65–72.
2. J. N. Kondo and K. Domen, Chem. Mater., 2008, 20, 835–847.
3. Y. Wan, H. Yang and D. Zhao, Acc. Chem. Res., 2006, 39, 423–432.
4. J. Lee, M. Christopher Orilall, S. C. Warren, M. Kamperman, F. J. DiSalvo and U. Wiesner, Nat. Mater., 2008, 7, 222–228.
5. B. Mandlmeier, J. M. Szeifert, D. Fattakhova-Rohlfing, H. Amenitsch and T. Bein, J. Am. Chem. Soc., 2011, 133, 17274–17282.
6. Y. Ren, Z. Ma and P. G. Bruce, Chem. Soc. Rev., 2012, 41, 4909–4927.
7. T. Wagner, S. Haffer, C. Weinberger, D. Klaus and M. Tiemann, Chem. Soc. Rev., 2013, 42, 4036–4053.
8. F. Jiao, A. Harrison, J.-C. Jumas, A. V. Chadwick, W. Kockelmann and P. G. Bruce, J. Am. Chem. Soc., 2006, 128, 5468–5474.
9. A. H. Lu and F. Schüth, Adv. Mater., 2006, 18, 1793–1805.
10. X. Lai, X. Li, W. Geng, J. Tu, J. Li and S. Qiu, Angew. Chem., Int. Ed., 2007, 46, 738–741.
11. K. E. Shopsowitz, A. Stahl, W. Y. Hamad and M. J. MacLachlan, Angew. Chem., Int. Ed., 2012, 51, 6886–6890.
12. X. Lai, D. Wang, N. Han, J. Du, J. Li, C. Xing, Y. Chen and X. Li, Chem. Mater., 2010, 22, 3033–3042.
13. P. Yang, D. Zhao, D. I. Margolese, B. F. Chmelka and G. D. Stucky, Nature, 1998, 396, 152–155.
14. E. L. Crepaldi, G. J. d. A. A. Soler-Illia, D. Grosso, F. Cagnol, F. Ribot and C. Sanchez, J. Am. Chem. Soc., 2003, 125, 9770–9786.
15. X. Sun, H. Hao, H. Ji, X. Li, S. Cai and C. Zheng, ACS Appl. Mater. Interfaces, 2014, 6, 401–409.
16. L. Gao, Z. Cheng, Q. Xiang, Y. Zhang and J. Xu, Sens. Actuators, B, 2015, 208, 436–443.
17. G. Korotcenkov, I. Boris, V. Brinzari, V. Golovanov, Y. Lychkovsky, G. Karkotsky, A. Cornet, E. Rossinyol, J. Rodrigue and A. Cirera, Sens. Actuators, B, 2004, 103, 13–22.
18. K. Soulantica, L. Erades, M. Sauvan, F. Senocq, A. Maisonnat and B. Chaudret, Adv. Funct. Mater., 2003, 13, 553–557.
19. T. Wagner, T. Sauerwald, C. D. Kohl, T. Waitz, C. Weidmann and M. Tiemann, Thin Solid Films, 2009, 517, 6170–6175.
20. X. Liu, R. Wang, T. Zhang, Y. He, J. Tu and X. Li, Sens. Actuators, B, 2010, 150, 442–448.
21. T. Waitz, T. Wagner, T. Sauerwald, C.-D. Kohl and M. Tiemann, Adv. Funct. Mater., 2009, 19, 653–661.
22. S. Haffer, T. Waitz and M. Tiemann, J. Phys. Chem. C, 2010, 114, 2075–2081.
23. H. Yang, Q. Shi, B. Tian, Q. Lu, F. Gao, S. Xie, J. Fan, C. Yu, B. Tu and D. Zhao, J. Am. Chem. Soc., 2003, 125, 4724–4725.
24. P. Li, H. Fan and Y. Cai, Colloids Surf., A, 2014, 453, 109–116.
25. K. Wetchakun, T. Samerjai, N. Tamaekong, C. Liewhiran, C. Siriwong, V. Kruefu, A. Wisitsoraat, A. Tuantranont and S. Phanichphant, Sens. Actuators, B, 2011, 160, 580–591.
26. C. Wang, L. Yin, L. Zhang, D. Xiang and R. Gao, Sensors, 2010, 10, 2088.
27. S. S. Sharma, K. Nomura and Y. Ujihira, J. Appl. Phys., 1992, 71, 2000–2005.
28. S. Park, S. Kim, G.-J. Sun and C. Lee, ACS Appl. Mater. Interfaces, 2015, 7, 8138–8146.
29. C. Zhao, J. Fu, Z. Zhang and E. Xie, RSC Adv., 2013, 3, 4018–4023.
30. N. D. Hoa and S. A. El-Safty, Chem.–Eur. J., 2011, 17, 12896–12901.
31. D. Wang, J. Sun, X. Cao, Y. Zhu, Q. Wang, G. Wang, Y. Han, G. Lu, G. Pang and S. Feng, J. Mater. Chem. A, 2013, 1, 8653–8657.
32. M. Batzill and U. Diebold, Phys. Chem. Chem. Phys., 2007, 9, 2307–2318.
33. C. Luan, K. Wang, Q. Yu, G. Lian, L. Zhang, Q. Wang and D. Cui, Sens. Actuators, B, 2013, 176, 475–481.
34. F. Lei, Y. Sun, K. Liu, S. Gao, L. Liang, B. Pan and Y. Xie, J. Am. Chem. Soc., 2014, 136, 6826–6829.
35. Y. Li, J. Xu, J. Chao, D. Chen, S. Ouyang, J. Ye and G. Shen, J. Mater. Chem., 2011, 21, 12852–12857.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details and characterization techniques; gas sensor fabrication and measurements; XRD, SEM, TEM and TG/DSC curves of MSC In(OH)₃ precursors (Fig. S1, S2 and S4); size distribution histograms of MSC/M In(OH)₃ precursors and MSC/M In₂O₃ products (Fig. S3 and S7); SEM images of M In(OH)₃ precursor and M In₂O₃ product (Fig. S5); FTIR spectra of MSC/M In(OH)₃ precursor and MSC/M In₂O₃ products (Fig. S6 and S8). See DOI: 10.1039/c5ra24197a

---