Gas-sensing properties of p-type α-Fe₂O₃ polyhedral particles synthesized via a modified polyol method

Abstract

The gas sensing properties of polyhedral α-Fe₂O₃ particles were investigated. The polyhedral α-Fe₂O₃ particles were synthesized via a modified polyol method with the addition of an extra amount of NaBH₄ in ethylene glycol, and fabricated to a thick gas sensing film after calcination at 500 °C and 900 °C. The polyhedral α-Fe₂O₃ particles exhibited a p-type nature, which is a new result opposite to other results of recent reports of semiconductor gas sensors using α-Fe₂O₃. In particular, we suggest that the difference between structure and morphology of α-Fe₂O₃ particles can lead to p-type and n-type characterization. In addition, we suggested that Na ions from NaBH₄ were incorporated into α-Fe₂O₃ oxide. They are the cause of the generation of holes in α-Fe₂O₃ oxide that led to the p-type nature of α-Fe₂O₃ oxide. The measurement of the sensor response to hydrogen, carbon monoxide, toluene, propane and ethanol revealed that the sensor device synthesized using polyhedral α-Fe₂O₃ particles is sensitive to hydrocarbons and ethanol. Importantly, the sensor device using polyhedral α-Fe₂O₃ particles calcined at 900 °C was strongly sensitive to ethanol due to the porous structure of the α-Fe₂O₃ particles.

---

1. Introduction

So far, oxide semiconductor gas sensors have been practically used for detecting gases such as inflammable gases, dangerous hazardous gases and environmental gases because of their high sensor response and stability as well as simplification in electric circuits.¹ The gas sensing mechanism for reducing gases of oxide semiconductor gas sensors has been considered that target gases diffused into the porous gas sensing films and are oxidized by oxygen species (O⁻ and O²⁻) absorbed on the oxide surface, leading to the change in electric resistance. In the case of n-type semiconductors, the electric resistance of the sensing film decreases when the target gases react with the sensing film, because the majority carrier, e.g. electrons, move from absorbed oxygen species to the semiconductor. In contrast, the electric resistance of the sensor film using p-type semiconductor increases when the target gases react with the sensing film, because the concentration of the majority carrier, e.g. holes in p-type semiconductors, decreases when electrons move from absorbed oxygen species to the semiconductor.²

Various typical materials have been reported for use in semiconductor gas sensors, such as SnO₂ (n-type),³ TiO₂ (n-type),⁴ ZnO (n-type),⁵ In₂O₃ (n-type),⁶ WO₃ (n-type),⁷ α-Fe₂O₃ (n-type),⁸ γ-Fe₂O₃ (n-type),⁹ carbon nanotubes (p-type),¹⁰ Co₃O₄ (p-type),¹¹ V₂O₅ (p-type),¹² and CuO (p-type).¹³ Among them, α-Fe₂O₃ has the potential for practical use due to its high sensitivity, high stability and insensitivity to humidity.^14,15 It has been reported that shape and size of semiconductor oxides strongly affects the gas sensing performance of the semiconductor gas sensor. When the grain size of semiconductor oxide decreases, the volume ratio of the space charge layer to the grain increases, leading to the increase in the sensor response.^16,17 The controlling methods of the shape and morphology can lead to improve the facile diffusion of target gases into the gas sensing films.^18,19 Therefore, morphology and size of α-Fe₂O₃ has been controlled to nanosphere,²⁰ nanorod,^21,22 nanotube,²³ hollow sphere^24–26 and porous structure^27–29 in order to improve the gas sensing properties of α-Fe₂O₃.

In this study, we have synthesized polyhedral α-Fe₂O₃ particles for use in sensor materials. Herein, polyhedral α-Fe₂O₃ particles were successfully synthesized by a modified polyol method with the addition of an extra amount of NaBH₄ in ethylene glycol (EG). The large polyhedral α-Fe₂O₃ particles were fabricated for use in a gas sensing film and its gas sensing properties were investigated in detection of hydrogen (H₂), carbon monoxide (CO), toluene (C₆H₅CH₃), propane (C₃H₈), and ethanol (C₂H₅OH) as target gases.

2. Experimental section

2.1. Chemical

Poly(vinylpyrrolidone) (PVP) (FW = 55 000), FeCl₃·4H₂O, sodium borohydride (NaBH₄) and ethylene glycol (EG) were purchased from Sigma-Aldrich Co. and used for the synthesis of α-Fe₂O₃ particles. Ethanol, acetone and hexane were purchased from Wako Pure Chemical Industries, Ltd. and used for the dispersal of α-Fe₂O₃ particles synthesized. α-Terpineol purchased from Kishida chemical Co. Ltd. was used for the binder of gas sensing films. All chemicals were used as received without further purification.

2.2. Materials synthesis

The details and steps of the procedures for the controlled synthesis of polyhedral α-Fe₂O₃ particles were previously presented in the controlled synthesis of Pt nanoparticles.³⁰ Briefly, 3 mL of EG, 1.5 mL of 0.0625 M FeCl₃, 3 mL of 0.375 M PVP, and 0.028 g NaBH₄ were used for making Sample 1 in a typical process of the controlled synthesis of the polyhedral α-Fe₂O₃ oxide particles. The known polyol processes were previously presented.^30,42 In general, FeCl₃ was completely reduced with the extra amount of NaBH₄ in EG at 200–230 °C for 30 min. As a result, the dark-brown solutions containing polyhedral α-Fe₂O₃ oxide particles with large sizes, shapes and morphologies were obtained as the final product. They have the particle size of 1–5 μm and with polyhedral shape and morphology. Similar to the synthetic process for Sample 1, we have used the same processes for Samples 2 and 3 for XRD and SEM measurements. Then, Sample 1 was also used for XRD and SEM measurement and analysis. Sample 2 was heated at 500 °C for 1 h for SEM measurement and analysis (Fig. 1(a)). Sample 3 was heated at 900 °C for 1 h for SEM analysis (Fig. 1(b)). The crystal phase of samples was identified by means of an X-ray diffractometer (XRD; RINT2100, Rigaku Co., Ltd. Japan) with Cu Kα radiation (1.54056 Å) as an X-ray source. The morphology of samples was observed by field emission scanning electron microscopy (FE-SEM, JSM-6340F, JEOL, Japan). The samples for the SEM observation were obtained by dropping the suspension of α-Fe₂O₃ particles on copper grids or copper brass.

Fig. 1 SEM images of the samples of the as-prepared Fe-based particles calcined at (a) 500 and (b) 900 °C.

2.3. Sensor fabrication and measurements. Gas sensing films were fabricated on alumina substrates (9 × 13 × 0.38 mm) equipped with a pair of comb-type Au electrodes (electrode gap: 90 μm), as shown in Fig. 2. Experimentally, a sol suspension of large as-synthesized α-Fe₂O₃ particles was evaporated at 80 °C for their drying in an oven for 6 h. Then, the fine powder of the as-synthesized α-Fe₂O₃ was mechanically mixed with α-terpineol to form a paste for gas sensing films. Next, the paste obtained was fabricated to a thick film (ca. 10 μm) through screen-printing on an alumina substrate attached with a pair of comb-type Au electrodes (electrode gap: 90 μm). The gas sensing film on the alumina substrate was sintered at 500 and 900 °C for 3 h in atmospheric air. The sensor devices were then settled in quartz tubes in a tube-type electric furnace. Next, the sensor device was connected with a standard resistor in series. To evaluate the electrical resistance of the devices in air and in air containing target gases (H₂, CO, C₆H₅CH₃, C₃H₈ and C₂H₅OH), the voltage across the standard resistor was measured with an electrometer (Model 2701, Keithley Instruments Inc.) under an applied voltage of dc 4 V.

Fig. 2 Structure of the thick-film type sensor device fabricated by a screen-printing.

3. Results and discussion

3.1. Materials characterizations

Prior to gas sensing measurements, the crystal phase and the morphology of samples were evaluated by means of XRD and FE-SEM, respectively. Fig. 3 shows the two XRD patterns of the samples after calcination at 500 and 900 °C. The pure crystal phases of both samples were assigned to α-Fe₂O₃ (JCPDS 33-0664) without any other impurity phases. Fig. 1(a) and (b) shows the FE-SEM image of α-Fe₂O₃ obtained at 500 °C and 900 °C, respectively. As shown in Fig. 1(a), α-Fe₂O₃ particles after the calcination at 500 °C were dense polyhedral particles with side lengths of 1–3 μm. Considering the crystalline size of α-Fe₂O₃ particles calculated by XRD, it seems that small size α-Fe₂O₃ nanocrystals was found in the dense polyhedral particles. On the other hand, in the case of α-Fe₂O₃ particles after the calcination at 900 °C, although the particle size is almost the same as α-Fe₂O₃ particles after calcination at 500 °C, many grain boundaries and small pores can be seen in the particles. It seems that small crystallites in one polyhedral particle were sintered, resulting in the formation of grain boundaries and small pores. However, the reason why the morphology of polyhedral particles is changed by the calcination temperature will be discussed in another paper.

Fig. 3 XRD patters of the samples calcined at (a) 500 and (b) 900 °C.

3.2. Gas sensing properties

In a semiconductor gas sensor, target gases react with oxygen species absorbed on the surface of the oxide semiconductor, resulting in the change in the electric resistance of the gas sensing film. Therefore, the dependence of the electric resistance on the oxygen partial pressure is an important factor for the operation of a semiconductor gas sensor. Fig. 4 shows the dependence of the electric resistance on the oxygen partial pressure (P_O2) for the sensor device calcined at (a) 500 °C and (b) 900 °C. The electric resistance of both sensors decreased with an increase in the oxygen partial pressure, indicating the p-type nature of the α-Fe₂O₃ gas sensing film.^31–33

Fig. 4 Dependence of the electric resistance on the oxygen partial pressure for the sensor device calcined at (a) 500 and (b) 900 °C. Operating temperature is 300 °C.

In the case of p-type oxide semiconductors, oxygen attracts electrons in the valence band of the oxide, resulting in an increase in the hole concentrations on the large crystal surfaces of oxides. Therefore, the electric resistance of p-type α-Fe₂O₃ obtained in this study decreased with increasing the oxygen partial pressure.³⁴ The main reason for the relatively large difference in the electric resistance between the two samples may be due to the possible effect of the crystalline size of α-Fe₂O₃. In the case of the p-type oxide semiconductor, the grain boundaries have lower electric resistance than the bulk oxide due to the formation of Schottky barriers. Hence, the larger number of grain boundaries tends to result in a lower electric resistance.

As shown in Fig. 3, the crystalline size of α-Fe₂O₃ calcined at 500 °C is smaller than α-Fe₂O₃ calcined at 900 °C. Therefore, it seems that the number of the grain boundaries in the sensor device calcined at 500 °C is larger than that of the device calcined at 900 °C, leading to a large decrease in the electric resistance of the sensor device calcined at 500 °C. Fig. 5 shows the response transients to 500 ppm H₂ at various temperatures for the sensor device calcined at (a) 500 °C and (b) 900 °C. As shown in Fig. 5, the electric resistance increased when switching the flow gas from air to 500 ppm H₂ attributed to the p-type nature of α-Fe₂O₃ obtained in this study. In the previous reports about semiconductor gas sensors with the use of α-Fe₂O₃, α-Fe₂O₃ oxide material works as an n-type oxide semiconductor, meaning that the electric resistance of α-Fe₂O₃ decreased under exposure to the reducing gases.^35–38 However, our gas sensor exhibited the opposite tendency to the above reports. The origin of the p-type nature of α-Fe₂O₃ obtained in our research has not been elucidated yet. However, it seems that low concentration of Na ions from NaBH₄ affected the major carrier of α-Fe₂O₃ obtained in our research. When Na ions from NaBH₄ for the synthesis of samples are doped into α-Fe₂O₃ bulk, Na ions in the α-Fe₂O₃ lattice can generate holes as follows:³⁹

Na₂O → 2Na_Fe′′ + O_O + 2V_O +4h˙

Fig. 5 The response transients to 500 ppm H₂ for the sensor device calcined at (a) 500, and (b) 900 °C.

The holes generated by the above formula can act as a source of p-type nature of α-Fe₂O₃ obtained in our research.

The sensor response to 500 ppm H₂ was calculated from the response transients shown in Fig. 5, and summarized in Fig. 6. The sensor response was defined as the ratio of the electric resistance in air containing target gas (R_gas) to in air (R_air). As shown in Fig. 6, the sensor response increased with a decrease in the operating temperature. When the operating temperature is too high, H₂ can be catalytically oxidized to water (H₂O) without electron transfer between absorbed oxygen and α-Fe₂O₃. Therefore, lower operating temperature is preferable for high sensor response. However, as shown in Fig. 5, both the response and recovery time for H₂ sensing decreased with a decrease in the operating temperature of our gas sensor. We suggested that the desorption of H₂O produced by the reaction between H₂ and the absorbed oxygen on α-Fe₂O₃ oxide surfaces became slow at lower temperature. In particular, the slow desorption of H₂O prevented the adsorption of oxygen species on the surface of α-Fe₂O₃, resulting in the slow recovery of sensor response. Consequently, the optimal operating temperature of the sensor was defined as 300 °C. When the calcination temperature of the sensor devices was compared, the sensor response of the sensor device calcined at 500 °C exhibited a higher sensor response as compared with the sensor device calcined at 900 °C. It seems that the larger number of grain boundaries in the sensor device calcined at 500 °C gave the higher sensor response because the change in the electric resistance at the grain boundaries of oxides is one of the most important factors for a significant change in the electric resistance of the gas sensing film. Fig. 7 shows the dependence of the sensor response on the H₂ concentration. As shown in Fig. 7, the sensor response to hydrogen gas (H₂) was increased with an increase in the H₂ concentration without saturation at high concentrations of H₂. In particular, the sensor device calcined at 500 °C becomes more sensitive to a change in the H₂ concentration due to the small crystalline size of α-Fe₂O₃. The sensor devices were exposed to various reducing gases, and compared with their sensor response. Typically, H₂, CO, C₆H₅CH₃, C₃H₈, and C₂H₅OH were chosen for the gas sensing measurements. Fig. 8 shows the response transients to various gases of the sensor device calcined at (a) 500 and (b) 900 °C. The value of the sensor response to various gases was summarized in Fig. 9. It was found that sensor devices using polyhedral α-Fe₂O₃ were sensitive to hydrocarbons and C₂H₅OH as compared with H₂ and CO. However, as shown in Fig. 8, the recovery time for the hydrocarbons and C₂H₅OH was rather slow due to the slow desorption speed of oxidized intermediates of target gases. Such a slow recovery speed of α-Fe₂O₃ will be improved by doping foreign element or loading catalysts.^40,41 As shown in Fig. 9, the sensor device calcined at 500 °C was most sensitive to C₆H₅CH₃. However, in the case of the sensor device calcined at 900 °C, the highest sensor response was obtained by C₂H₅OH detection. This tendency can be explained by the difference in the porosity of the two sensor devices. As shown in Fig. 1(a), particles calcined at 500 °C are dense polyhedral particles with a side length of around 1–3 μm. In contrast, α-Fe₂O₃ particles calcined at 900 °C are also polyhedral particles but contains small pores in the polyhedral particles. It is likely that C₂H₅OH molecules can diffuse into the small pores in each polyhedral α-Fe₂O₃ particle. However, C₆H₅CH₃ molecules were too large to diffuse into the pores facilely. Therefore, the sensor device using polyhedral α-Fe₂O₃ particles calcined at 900 °C was more sensitive to C₂H₅OH than C₆H₅CH₃. As shown in the above, sensor devices obtained by calcination at 900 °C in this study are more sensitive and selective to ethanol as compared with semiconductor gas sensors using α-Fe₂O₃ reported in the other literature.^21,25,36,38 There is a new phenomenon found in large particles with grain and boundary structure in Fig. 1(b), which is particle and structure deformation on the particle surface and inside the particle structure under heat treatment at 900 °C.^43,44 We have suggested that phenomena of both plastic and elastic deformation on surface and inside α-Fe₂O₃ particles under nanoparticle heat treatments at 900 °C were identified for better gas sensing properties of the sensor device to ethanol. Therefore, the gas sensor using α-Fe₂O₃ particles with grain and boundary structures “new micro-nano structures” can be a good candidate for an ethanol-selective gas sensor without any other additives for high selectivity and sensitivity.⁴⁵

Fig. 6 The dependence of the sensor response to 500 ppm H₂ on the operating temperature for sensor devices calcined at (a) 500 and (b) 900 °C.

Fig. 7 (a) The response transients to H₂ at 300 °C and (b) dependence of the sensor response on the H₂ concentration at 300 °C.

Fig. 8 The response transients to various gases of sensor devices calcined at (a) 500 °C and (b) 900 °C.

Fig. 9 Sensor response to various gases of the sensor devices calcined at 500 °C and 900 °C.

4. Conclusion

In this study, polyhedral α-Fe₂O₃ particles was successfully synthesized via a modified polyol method with the addition of an extra amount of NaBH₄ in ethylene glycol, and its gas sensing characteristics were investigated. The measurement of the dependence of the electric resistance on the O₂ concentration revealed that polyhedral α-Fe₂O₃ particles synthesized in this study exhibited a p-type nature; the electric resistance of the sensor device increased with an increase in the O₂ concentration. This tendency is opposite to other reports of gas sensors using α-Fe₂O₃. The measurement of the sensor response to various reducing gases revealed that α-Fe₂O₃ particles prepared in this study are sensitive to hydrocarbons and alcohols as compared with H₂ and CO gases. In particular, a sensor device calcined at 900 °C has sensitively responded to C₂H₅OH due to its porous structure. This result indicates that our sensor device obtained by calcination at 900 °C can be a good candidate for an ethanol-selective gas sensor.

Acknowledgements

In this research, we are very grateful to the precious supports from Structural Ceramics Engineering Center, Shanghai Institute of Ceramics, Chinese Academy of Science, Dingxi Road 1295, Shanghai 200050, China. This study was also supported in part by a fund from the National Natural Science Foundation of China (NSFC, contract nos. 51071167 and 51102266).

References

1. K. Ihokura and J. Watson, The Stanic Oxide Gas Sensor – Principles and Applications, CRC Press, Boca Raton, FL, 1994.
2. P. T. Moseley, Sens. Actuators, B, 1992, 6, 149.
3. N. Yamazoe, Y. Kurokawa and T. Seiyama, Sens. Actuators, 1983, 4, 283.
4. G. S. Devi, T. Hyodo, Y. Shimizu and M. Egashira, Sens. Actuators, B, 2002, 87, 122.
5. L. Liao, H. B. Lu, J. C. Li, H. He, D. F. Wang, D. J. Fu and C. Liu, J. Phys. Chem. C, 2007, 111, 1900.
6. S. N. Malchenko, Y. N. Lychkovsky and M. V. Baykov, Sens. Actuators, B, 1993, 13, 159.
7. J. L. Solisa, S. Saukko, L. Kish, C. G. Granqvist and V. Lantto, Thin Solid Films, 2001, 391, 255.
8. V. V. Malyshev, A. V. Eryshkin, E. A. Koltypin, A. E. Varfolomeev and A. A. Vasiliev, Sens. Actuators, B, 1994, 19, 434.
9. C. V. Gopal Reddy, K. Kalyana Seela and S. V. Manorama, Int. J. Inorg. Mater., 2000, 2, 301.
10. L. Valentini, C. Cantalini, I. Armentano, J. M. Kenny, L. Lozzi and S. Santucci, Diamond Relat. Mater., 2004, 13, 1301.
11. W. Y. Li, L. N. Xu and J. Chen, Adv. Funct. Mater., 2005, 15, 851.
12. M. Yu, X. Liu, Y. Wang, Y. Zheng, J. Zhang, M. Li, W. Lan and Q. Su, Appl. Surf. Sci., 2012, 258, 9554.
13. C. Yang, X. Su, J. Wang, X. Cao, S. Wang and L. Zhang, Sens. Actuators, B, 2013, 185, 159.
14. J. Peng and C. C. Chai, Sens. Actuators, B, 1993, 14, 591.
15. K. Song, Q. Wang, Q. Liu, H. Zhang and Y. Cheng, Sensors, 2011, 11, 485.
16. C. Xu, J. Tamaki, N. Miura and N. Yamazoe, Sens. Actuators, B, 1991, 3, 147.
17. A. Gurlo, M. Ivanovskaya, N. Bârsan, M. Schweizer-Berberich, U. Weimar, W. Göpel and A. Diéguez, Sens. Actuators, B, 1997, 44, 327.
18. M. Tiemann, Chem.–Eur. J., 2007, 13, 8376.
19. M. M. Arafat, B. Dinan, S. A. Akbar and A. S. M. A. Haseeb, Sensors, 2012, 12, 7207.
20. Y. Cao, H. Luo and D. Jia, Sens. Actuators, B, 2013, 176, 618.
21. D. Patil, V. Patil and P. Patil, Sens. Actuators, B, 2011, 152, 299.
22. P. Sun, L. You, D. Wang, Y. Sun, J. Ma and G. Lu, Sens. Actuators, B, 2011, 156, 368.
23. H. Shan, C. Liu, L. Liu, S. Li, L. Wang, X. Zhang, X. Bo and X. Chi, Sens. Actuators, B, 2013, 184, 243.
24. J. Huang, M. Yang, C. Gu, M. Zhai, Y. Sun and J. Liu, Mater. Res. Bull., 2011, 46, 1211.
25. G. Fan, Y. Wang, M. Hu, Z. Luo, K. Zhang and G. Li, Sensors, 2012, 12, 4594.
26. G. Shi, K. Zhang, S. Zhou and Q. Zhang, Mater. Lett., 2013, 107, 228.
27. B. Sun, J. Horvat, H. S. Kim, W. S. Kim, J. Ahn and G. Wang, J. Phys. Chem. C, 2010, 114, 18753.
28. Q. Hao, S. Liu, X. Yin, Z. Du, M. Zhang, L. Li, Y. Wang and Q. Li, CrystEngComm, 2011, 13, 806.
29. D. Su, H. S. Kim, W. S. Kim and G. Wang, Microporous Mesoporous Mater., 2012, 149, 36.
30. N. V. Long, M. Ohtaki, M. Uchida, R. Jalem, H. Hirata, N. D. Chien and M. Nogami, J. Colloid Interface Sci., 2011, 359, 339.
31. S. P. Mitoff, J. Chem. Phys., 1961, 35, 882.
32. Y. Li, W. Wlodarski, K. Galatsis, S. H. Moslih, J. Cole, S. Russo and N. Rockelmann, Sens. Actuators, B, 2002, 83, 160.
33. F. Hossein-Babaei, M. Keshmiri, M. Kakavand and T. Troczynski, Sens. Actuators, B, 2005, 110, 28.
34. P. T. Moseley, Sens. Actuators, B, 1991, 3, 167.
35. C. Cantalini, M. Faccio, G. Ferri and M. Pelino, Sens. Actuators, B, 1994, 19, 437.
36. N. D. Cuong, T. T. Hoa, D. Q. Khieu, T. D. Lam, N. D. Hoa and N. V. Hieu, J. Alloys Compd., 2012, 523, 120.
37. Z. F. Dou, C. Y. Cao, Q. Wang, J. Qu, Y. Yu and W. G. Song, ACS Appl. Mater. Interfaces, 2012, 4, 5698.
38. Y. Kang, L. Wang, Y. Wang, H. Zhang, Y. Wang, D. Hong, Y. Qv and S. Wang, Sens. Actuators, B, 2013, 177, 570.
39. C. V. Santilli, J. P. Bonnet, P. Dordor, M. Onillon and P. Hagenmuller, Ceram. Int., 1990, 16, 25.
40. Z. Jing, Mater. Lett., 2006, 60, 3315.
41. S. C. Lee, B. W. Hwang, S. J. Lee, H. Y. Choi, S. Y. Kim, S. Y. Jung, D. Ragupathy, D. D. Lee and J. C. Kim, Sens. Actuators, B, 2011, 160, 1328.
42. D. L. Huber, Small, 2005, 1(5), 482.
43. F. J. Humphreys and M. Hatherly, Recrystallization and related annealing phenomena, Elsevier Ltd, 2nd edn, 2004.
44. G. E. Totten, Steel Heat Treatment Handbook CRC Press, Taylor & Fracis Group, LLC, 2nd edn, 2007.
45. N. V. Long, Y. Yang, M. Yuasa, T. M. Cao, Y. Cao, T. Nann and M. Nogami, RSC Adv., 2014, 4, 6383–6390.

---