# ADVANCED NANOMATERIALS FOR SENSING APPLICATIONS

EDITED BY : Zhongchang Wang, Wen Zeng and Zhenyu Li PUBLISHED IN : Frontiers in Chemistry

### Frontiers Copyright Statement

© Copyright 2007-2019 Frontiers Media SA. All rights reserved. All content included on this site, such as text, graphics, logos, button icons, images, video/audio clips, downloads, data compilations and software, is the property of or is licensed to Frontiers Media SA ("Frontiers") or its licensees and/or subcontractors. The copyright in the text of individual articles is the property of their respective authors, subject to a license granted to Frontiers.

The compilation of articles constituting this e-book, wherever published, as well as the compilation of all other content on this site, is the exclusive property of Frontiers. For the conditions for downloading and copying of e-books from Frontiers' website, please see the Terms for Website Use. If purchasing Frontiers e-books from other websites or sources, the conditions of the website concerned apply.

Images and graphics not forming part of user-contributed materials may not be downloaded or copied without permission.

Individual articles may be downloaded and reproduced in accordance with the principles of the CC-BY licence subject to any copyright or other notices. They may not be re-sold as an e-book.

As author or other contributor you grant a CC-BY licence to others to reproduce your articles, including any graphics and third-party materials supplied by you, in accordance with the Conditions for Website Use and subject to any copyright notices which you include in connection with your articles and materials.

All copyright, and all rights therein, are protected by national and international copyright laws.

The above represents a summary only. For the full conditions see the Conditions for Authors and the Conditions for Website Use. ISSN 1664-8714 ISBN 978-2-88963-123-0 DOI 10.3389/978-2-88963-123-0

### About Frontiers

Frontiers is more than just an open-access publisher of scholarly articles: it is a pioneering approach to the world of academia, radically improving the way scholarly research is managed. The grand vision of Frontiers is a world where all people have an equal opportunity to seek, share and generate knowledge. Frontiers provides immediate and permanent online open access to all its publications, but this alone is not enough to realize our grand goals.

### Frontiers Journal Series

The Frontiers Journal Series is a multi-tier and interdisciplinary set of open-access, online journals, promising a paradigm shift from the current review, selection and dissemination processes in academic publishing. All Frontiers journals are driven by researchers for researchers; therefore, they constitute a service to the scholarly community. At the same time, the Frontiers Journal Series operates on a revolutionary invention, the tiered publishing system, initially addressing specific communities of scholars, and gradually climbing up to broader public understanding, thus serving the interests of the lay society, too.

### Dedication to Quality

Each Frontiers article is a landmark of the highest quality, thanks to genuinely collaborative interactions between authors and review editors, who include some of the world's best academicians. Research must be certified by peers before entering a stream of knowledge that may eventually reach the public - and shape society; therefore, Frontiers only applies the most rigorous and unbiased reviews.

Frontiers revolutionizes research publishing by freely delivering the most outstanding research, evaluated with no bias from both the academic and social point of view. By applying the most advanced information technologies, Frontiers is catapulting scholarly publishing into a new generation.

### What are Frontiers Research Topics?

Frontiers Research Topics are very popular trademarks of the Frontiers Journals Series: they are collections of at least ten articles, all centered on a particular subject. With their unique mix of varied contributions from Original Research to Review Articles, Frontiers Research Topics unify the most influential researchers, the latest key findings and historical advances in a hot research area! Find out more on how to host your own Frontiers Research Topic or contribute to one as an author by contacting the Frontiers Editorial Office: researchtopics@frontiersin.org

# ADVANCED NANOMATERIALS FOR SENSING APPLICATIONS

Topic Editors: Zhongchang Wang, International Iberian Nanotechnology Laboratory (INL), Portugal Wen Zeng, Chongqing University, China Zhenyu Li, Southwest Petroleum University, China

Citation: Wang, Z., Zeng, W., Li, Z., eds. (2019). Advanced Nanomaterials for Sensing Applications. Lausanne: Frontiers Media. doi: 10.3389/978-2-88963-123-0

# Table of Contents


*84 Hydrothermal Synthesis of Hierarchical Ultrathin NiO Nanoflakes for High-Performance CH4 Sensing*

Qu Zhou, Zhaorui Lu, Zhijie Wei, Lingna Xu, Yingang Gui and Weigen Chen

*88 Enhanced H2 S Gas-Sensing Performance of Zn2 SnO4 Lamellar Micro-Spheres* Ting-Ting Xu, Ying-Ming Xu, Xian-Fa Zhang, Zhao-Peng Deng, Li-Hua Huo and Shan Gao

# One-Dimensional Zinc Oxide Decorated Cobalt Oxide Nanospheres for Enhanced Gas-Sensing Properties

Hang Zhou, Keng Xu\*, Yong Yang, Ting Yu, Cailei Yuan\*, Wenyan Wei, Yue Sun and Wenhui Lu

*Jiangxi Key Laboratory of Nanomaterials and Sensors, Jiangxi Normal University, Nanchang, China*

### Edited by:

*Wen Zeng, Chongqing University, China*

### Reviewed by:

*Yan Wang, Henan Polytechnic University, China Ming Ying Xu, Heilongjiang University, China Shunping Zhang, Huazhong University of Science and Technology, China*

### \*Correspondence:

*Keng Xu xukeng@163.com Cailei Yuan clyuan@jxnu.edu.cn*

### Specialty section:

*This article was submitted to Nanoscience, a section of the journal Frontiers in Chemistry*

Received: *10 October 2018* Accepted: *03 December 2018* Published: *17 December 2018*

### Citation:

*Zhou H, Xu K, Yang Y, Yu T, Yuan C, Wei W, Sun Y and Lu W (2018) One-Dimensional Zinc Oxide Decorated Cobalt Oxide Nanospheres for Enhanced Gas-Sensing Properties. Front. Chem. 6:628. doi: 10.3389/fchem.2018.00628* In this study, one-dimensional (1D) zinc oxide was loaded on the surface of cobalt oxide microspheres, which were assembled by single-crystalline porous nanosheets, via a simple heteroepitaxial growth process. This elaborate structure possessed an excellent transducer function from the single-crystalline feature of Co3O<sup>4</sup> nanosheets and the receptor function from the zinc oxide nanorods. The structure of the as-prepared hybrid was confirmed via a Scanning Electron Microscope (SEM), X-ray diffraction (XRD), and a Transmission Electron Microscope (TEM). Gas-sensing tests showed that the gas-sensing properties of the as-designed hybrid were largely improved. The response was about 161 (Ra/Rg) to 100 ppm ethanol, which is 110 and 10 times higher than that of Co3O<sup>4</sup> (Rg/R<sup>a</sup> = 1.47) and ZnO (Ra/R<sup>g</sup> = 15), respectively. And the as-designed ZnO/Co3O<sup>4</sup> hybrid also showed a high selectivity to ethanol. The superior gas-sensing properties were mainly attributed to the as-designed nanostructures that contained a super transducer function and a super receptor function. The design strategy for gas-sensing materials in this work shed a new light on the exploration of high-performance gas sensors.

Keywords: gas sensing, transducer function, receptor function, ethanol detection, single-crystalline

# INTRODUCTION

Metal oxides, as a type of predominant gas sensing material, have provoked considerable attention because of their low cost, excellent electrical properties and controllable preparation (Korotcenkov, 2007; Li et al., 2016; Dong et al., 2018). It is well accepted that the gas-sensing processes of metal oxides, contain not only gas diffusion and gas reaction on its surface, but also signal transformation that transforms the surficial chemical signal into an electrical resistance variation (Rai et al., 2013). In this regard, gas-sensing performances of metal oxides depend largely on the structure factors of sensing materials such as surface area, morphology, as well as signal transmission channels (Zhu et al., 2018). Nowadays, many studies have reported that the gas-sensing properties of metal oxides can be improved by the creation of heterojunctions, which can provide tunable morphologies and compositions (Miller et al., 2014; Cao et al., 2017; Gong et al., 2018). For example, the In2O3/Co3O<sup>4</sup> composites synthesized by Mirzaei et al., exhibited superior gas-sensing performances than pristine In2O<sup>3</sup> did (Mirzaei et al., 2017). However, up to now, most related studies have been based on mechanical mixed composites or inhomogeneous structures, resulting in randomly distributed heterojunctions. In particular, the gas-sensing processes of composites that contain gas diffusion, gas reaction, and signal transformation still need to be further improved.

In this paper, ZnO nanorods anchored on the surface of Co3O<sup>4</sup> nanospheres were synthesized by an epitaxial growth method. ZnO and Co3O<sup>4</sup> have been widely investigated in the field of gas-sensing application, because of their chemical and physical stability and abundance in raw materials (Tan et al., 2017; Zhu and Zeng, 2017). We chose the ZnO nanorods and Co3O<sup>4</sup> microspheres because of their large surface area for gas reaction and because of the effective accesses for gas diffusion (Park et al., 2016). Moreover, the Co3O<sup>4</sup> nanospheres consisted of single-crystalline porous nanosheets which provided effective electrical pathways for charge-carrier transfer and was more beneficial to the signal transformation process (Singh et al., 2011; Zhang et al., 2015). Therefore, these ZnO nanorods acted as "trigger hairs," while Co3O<sup>4</sup> microspheres acted as "channels." And the p-n heterojunctions between ZnO and Co3O<sup>4</sup> also led to an extended depletion region and a high initial resistance, which is more beneficial for the transducer function (Li et al., 2015).

# EXPERIMENT

# Synthesis of ZnO/Co3O<sup>4</sup> Hybrid

All the chemical reagents were purchased from Sinopharm Chemical Reagent and used without further purification. During the synthesis processes (Xu et al., 2017), a mixed solution (80 ml) composed of distilled water and ethylene glycol (1: 79) was added at 0.05 g PVP, as well as 1 g cobalt acetate. The mixed solution was then stirred and poured into a Teflon-lined stainless-steel autoclave (100 mL). The resulting precipitate was collected and washed repeatedly at least seven times and thereafter maintained at 180◦C for 12 h. The final powder was then annealed at 350◦C for 2 h in air atmosphere. To prepare the ZnO/Co3O<sup>4</sup> hybrid, 0.03g as-prepared Co3O<sup>4</sup> and 0.32 g zinc acetate dehydrate was added in methanol through the assistance of ultrasonication. A KOH methanol solution (60 ml) was then also added to the above mixture. After 2 h of stirring, the mixture was centrifuged at 3,000 rpm and washed several times. The powder obtained, was then was added into a mixed solution composed of Zn(NO3)<sup>2</sup> (0.89 g) and HMT (0.42 mol/L) by vigorous stirring. The final as-designed hybrid was obtained after being refluxed at 95 ◦C for 6 h.

## Sensor Fabrication and Measurements

To fabricate a gas sensor, a certain number of as-prepared samples were uniformly dispersed in ethanol, by ultrasonication for 15 min, to form a suspension. The suspension was then coated onto a ceramic flat, which had been coated with Pt interdigitated electrodes and a Pt heat element (about 500µm). The obtained film was sintered at 500◦C for 2 h and aged at 300◦C for 3 days in air atmosphere. Gas-sensing performances were measured in a static testing instrument which was purchased from Wuhan Hua Chuang Rui Ke Co. Ltd. This equipment includes a test chamber (about 30 L in volume) and a personal computer. During the gas-sensing tests, the target gases were introduced into the test chamber by injecting their corresponding liquid, with a calculated amount, which was then vaporized by a heater. The desired concentrations of the testing gases are calculated by the following formula:

$$Q = \frac{V \times \varphi \times M}{22.4 \times d \times \rho} \times 10^{-9} \times \frac{273 + T\_R}{273 + T\_R}$$

where Q (ml) is the liquid volume of the volatile compound, V (ml) is the volume of the testing chamber, φ is the required gas volume fraction, M (g. mol−<sup>1</sup> ) is the molecular weight, d (g. cm−<sup>3</sup> ) is the specific gravity, and ρ is the purity of the volatile testing liquid, T<sup>R</sup> and T<sup>B</sup> ( ◦C) are the temperatures at ambient and the test chamber, respectively. For example, the liquid volume of ethanol is calculated at about 13.05 µL that corresponds to 100 ppm. Two electric fans installed in the chamber were used to make the test gas homogeneous. After a few minutes, the chamber was lifted to introduce ambient air (humidity: 16∼22%). In the meantime, the resistances of the sensors were recorded by a personal computer.

# RESULTS AND DISCUSSION

The crystalline phases of as-prepared samples were characterized by XRD (**Figure 1A**). As can be seen, beside the peaks that belong to Co3O<sup>4</sup> (JCPDS No. 42-1467), all the diffraction peaks left in the pattern of hybrids can be indexed to the reflections of ZnO (JCPDS No. 36–1451).The morphology of Co3O<sup>4</sup> is shown in the **Figures 1B,C**. It is seen that the Co3O<sup>4</sup> is composed of numerous uniform microspheres with a diameter of 2–4µm. These microspheres are composed of many crossed 2D nanosheets with numerous pores, as shown in the enlarged images (**Figure 1C**). It is worth mentioning that the individual nanosheets containing numerous pores are single crystals, since each nanosheet consisted of coherent lattice fringes regardless of the pore in the inset of **Figure 1C**. The single crystal can provide effective electrical pathways for charges, which benefits the gas-sensing performances (Meng et al., 2015). Accordingly, the morphology of ZnO/Co3O<sup>4</sup> was also characterized (**Figures 1D–F**). **Figure 1D** reveals that the ZnO/Co3O<sup>4</sup> hybrids were also composed of numerous microspheres. However, its diameter was much larger than that of pristine Co3O4. The enlarged images (**Figures 1E,F**) reveal a hedgehog-like structure of ZnO/Co3O<sup>4</sup> where ZnO nanorods were formed on the surface of Co3O<sup>4</sup> as trigger hairs. The ZnO/Co3O<sup>4</sup> hybrids were further characterized through EDS (**Figure S1**). Only O, Zn, Co, together with C elements were detected. The average atomic ratio of Zn and Co was about 1:6.5. The distributions of O, Zn and Co are presented in **Figures S1A–D**. It was revealed that the elements O, Zn, and Co were distributed homogeneously. The growth processes of ZnO on the surface of cobalt oxide were displayed under different time conditions as shown in **Figure S2**. The preparation processes of ZnO/Co3O<sup>4</sup> are thus illustrated in **Figures 1G,H**. As seen in step 1, the as-obtained Co3O<sup>4</sup> microspheres are composed of many single-crystalline porous nanosheets. These Co3O<sup>4</sup> microspheres

FIGURE 1 | (A) XRD spectra of Co3O4, ZnO, and ZnO/Co3O4. (B,C) SEM and TEM (inset) of Co3O4. (D–F) SEM and TEM (inset) of ZnO/Co3O4. (G–J) Schematic illustration of the formation processes of ZnO/Co3O4 hybrid.

were then anchored with ZnO crystal seeds by soaking them into a pre-prepared ZnO colloid solution as shown in step 2 (**Figure 1I**). Finally, via a heteroepitaxial growth process, ZnO nanorods were grown on the surface of Co3O<sup>4</sup> microspheres (**Figure 1J**).

The responses of these sensors to 100 ppm ethanol at different operating temperatures are shown in **Figure 2A**. It can be observed that the optimal operating temperature is 300◦C for ZnO/Co3O4. The response was about 161 (Ra/Rg) to 100 ppm ethanol, which is 110 and 10 times higher than that of Co3O<sup>4</sup> (Rg/R<sup>a</sup> = 1.47) and ZnO (Ra/R<sup>g</sup> = 15), respectively. Gas-sensing tests of ZnO/Co3O<sup>4</sup> toward other gases (100 ppm) were carried out as shown in **Figure 2B**. It was found that the responses to ethanol was much higher than to other gases. This could be due to the different volatilities and chemical properties of gases, which prompt the sensors to exhibit different adsorption and catalytic performances toward them. Therefore, the polarity, molecular weight and structure of these gases can exert great effect on the gas-sensing response. The difference between ethanol and methanol in this work may be the molecular weight. Alcohol with larger molecules can be more easily adsorbed and can release more electrons. On the other hand, different constituents and structures with various surface properties are also believed to influence the selectivity of sensing material, due to the diverse chemisorption abilities, which induce different selectivity behaviors between Co3O4, ZnO/Co3O4, and ZnO. The responses to 1∼100 ppm ethanol were tested as shown in **Figure 2C**. Evidently, exposure to ethanol gas led to an increase in resistance for Co3O4, while it led to a decrease for ZnO and ZnO/Co3O4, indicating that Co3O<sup>4</sup> exhibits a p-type semiconducting behavior while ZnO and ZnO/Co3O<sup>4</sup> exhibits an n-type behavior. **Figure 2D** present the responses of Co3O<sup>4</sup> and ZnO/Co3O<sup>4</sup> as a function of the ethanol concentrations (1–100 ppm), from

which the gas responses increase almost linearly with the gas concentration.

The real-time responses of Co3O4, ZnO, and ZnO/Co3O<sup>4</sup> to 100 ppm ethanol with three cycles are displayed in **Figure S3**. The resistance patterns reveal the similar continuous recycles, indicating all of the samples exhibit good repeatability. The response and recovery time are also important parameters in the fields of gas detection. It was revealed that the response time of Co3O4, ZnO, and ZnO/Co3O<sup>4</sup> are 6, 17, and 13 s, respectively. And the recovery times of Co3O4, ZnO, and ZnO/Co3O<sup>4</sup> were 12, 405, and 168 s, respectively. The fast response and recovery of Co3O<sup>4</sup> can be attributed to its single-crystalline porous structure which benefits the gas-diffusion process and charge-transfer process. Compared with pure ZnO, the response and recovery process of ZnO/Co3O<sup>4</sup> were accelerated because of its super single-crystalline porous structure. In addition, the response and recovery speeds of ZnO/Co3O<sup>4</sup> to ethanol with low concentrations was much faster than those in ethanol with high concentrations, because of the slow diffusion speed of ethanol molecules to the active site at a low concentration.

The excellent gas-sensing properties of the ZnO/Co3O<sup>4</sup> to ethanol, are mainly due to the many benefits of the as-designed structure. First, the ZnO nanorods and Co3O<sup>4</sup> microspheres provide numerous channels for gas diffusion and an extremely high surface area for gas reaction. Furthermore, the single-crystalline feature of Co3O<sup>4</sup> nanosheets provides a fast transport of charge carrier, another main factor to improve the signal transformation process of a hybrid. Moreover, the p-n heterojunctions between ZnO and Co3O<sup>4</sup> can enlarge the response signal. Because the conduction band edge of Co3O<sup>4</sup> is more negative than that of ZnO, band bending, and depletion regions are formed at the heterojunctions (Jana et al., 2015). By increasing the initial resistance, due to the as-established depletion region, the modulation of resistance will become more evident (Kim et al., 2017).

# CONCLUSIONS

In this paper, ZnO nanorods anchored on the surface of Co3O<sup>4</sup> nanospheres were synthesized. ZnO nanorods and porous Co3O<sup>4</sup> microspheres provided numerous channels for gas diffusion and a large surface area for gas reaction. The Co3O<sup>4</sup> nanospheres consisted of single-crystalline porous nanosheets to provide effective electrical pathways for charge-carrier transfer. And the p-n heterojunctions between ZnO and Co3O<sup>4</sup> also led to an extended depletion region and a high initial resistance. As expected, this as-designed hybrid exhibited excellent gas-sensing properties with an extremely high response (Ra/R<sup>g</sup> = 161) and a high selectivity to ethanol.

## AUTHOR CONTRIBUTIONS

HZ performed the experiments and analyzed the data with the help from YY, TY, CY, WW, YS, and WL. KX wrote the manuscript with input from all authors. All authors read and approved the manuscript.

## FUNDING

This work was supported by the National Natural Science Foundation of China (Grant No. 51461019, 51661012, 51761017, 51702140, 51602134, 61664005).

### REFERENCES


# SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fchem. 2018.00628/full#supplementary-material


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Zhou, Xu, Yang, Yu, Yuan, Wei, Sun and Lu. This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Electrospun ZnO–SnO<sup>2</sup> Composite Nanofibers and Enhanced Sensing Properties to SF<sup>6</sup> Decomposition Byproduct H2S

Zhaorui Lu<sup>1</sup> , Qu Zhou1,2 \*, Caisheng Wang<sup>2</sup> \*, Zhijie Wei <sup>1</sup> , Lingna Xu<sup>1</sup> and Yingang Gui <sup>1</sup>

<sup>1</sup> College of Engineering and Technology, Southwest University, Chongqing, China, <sup>2</sup> Electrical and Computer Engineering Department, Wayne State University, Detroit, MI, United States

Hydrogen sulfide (H2S) is an important decomposition component of sulfur hexafluoride (SF6), which has been extensively used in gas-insulated switchgear (GIS) power equipment as insulating and arc-quenching medium. In this work, electrospun ZnO-SnO<sup>2</sup> composite nanofibers as a promising sensing material for SF<sup>6</sup> decomposition component H2S were proposed and prepared. The crystal structure and morphology of the electrospun ZnO-SnO<sup>2</sup> samples were investigated by X-ray diffraction (XRD), scanning electron microscopy (SEM) and transmission electron microscopy (TEM), respectively. The composition of the sensitive materials was analyzed by energy dispersive X-ray spectrometers (EDS) and X-ray photoelectron spectroscopy (XPS). Side heated sensors were fabricated with the electrospun ZnO-SnO<sup>2</sup> nanofibers and the gas sensing behaviors to H2S gas were systematically investigated. The proposed ZnO–SnO<sup>2</sup> composite nanofibers sensor showed lower optimal operating temperature, enhanced sensing response, quick response/recovery time and good long-term stability against H2S. The measured optimal operating temperature of the ZnO–SnO<sup>2</sup> nanofibers sensor to 50 ppm H2S gas was about 250◦C with a response of 66.23, which was 6 times larger than pure SnO<sup>2</sup> nanofibers sensor. The detection limit of the fabricated ZnO–SnO<sup>2</sup> nanofibers sensor toward H2S gas can be as low as 0.5 ppm. Finally, a plausible sensing mechanism for the proposed ZnO–SnO<sup>2</sup> composite nanofibers sensor to H2S was also discussed.

### Edited by:

Zhongchang Wang, Laboratório Ibérico Internacional de Nanotecnologia (INL), Portugal

### Reviewed by:

Yong Zhang, Xiangtan University, China Qi Qi, Jilin University, China

### \*Correspondence:

Qu Zhou zhouqu@swu.edu.cn Caisheng Wang cwang@wayne.edu

### Specialty section:

This article was submitted to Nanoscience, a section of the journal Frontiers in Chemistry

Received: 27 August 2018 Accepted: 18 October 2018 Published: 06 November 2018

### Citation:

Lu Z, Zhou Q, Wang C, Wei Z, Xu L and Gui Y (2018) Electrospun ZnO–SnO2 Composite Nanofibers and Enhanced Sensing Properties to SF6 Decomposition Byproduct H2S. Front. Chem. 6:540. doi: 10.3389/fchem.2018.00540 Keywords: ZnO-SnO2 nanofibers, electrospinning, H2S, sensing properties, SF6 decomposition components

## INTRODUCTION

Sulfur hexafluoride (SF6) insulating gas has excellent insulation performance and arc quenching. It is widely applied in gas-insulated switchgear (GIS) of power system as electrical insulator as well as arc-quenching medium (Beroual and Haddad, 2017; Zhang X. et al., 2017). However, partial discharge and disruptive discharge might occur in GIS equipment during the long run, accounting for the SF<sup>6</sup> gas decomposing to various decomposition components, such as H2S, SO2, SOF2, SO2F<sup>2</sup> (Tsai, 2007; Liu et al., 2017). Previous researches have reported that these typical decomposition components are able to accelerate the corrosion rate of the GIS equipment and increase the paralysis possibility of the power system (Zhang X. et al., 2016; Li et al., 2017). Therefore, accurate and effective detection of SF<sup>6</sup> gas decomposition components is significant to estimate and optimize the operation state of GIS power equipment.

Semiconductor metal oxides such as SnO<sup>2</sup> (Qi et al., 2014; Li et al., 2016; Shahabuddin et al., 2017; Zhou et al., 2018a), ZnO (Zhou et al., 2013; Zuo et al., 2013; Zhu et al., 2018), TiO<sup>2</sup> (Zeng et al., 2012; Park et al., 2017; Zhang Y. X. et al., 2018), NiO (Zhang Y. et al., 2016; Zhou et al., 2018b,c) are the most investigated group for gas sensors owing to their outstanding gas response and selectivity. Sensing nanostructure with high surface area and full electron depletion is advantageous to enhance the sensing performances (Hao et al., 2012; Miller et al., 2014). In particular, the 1D nanostructures such as nanofibers (Jiang et al., 2016), nanorods (Zhang et al., 2014; Zou et al., 2016), and nanotubes (Kong et al., 2015) have been extensively applied to improve gas sensing properties (Li T. M. et al., 2015; Long et al., 2018). Besides, many studies indicated that the selectivity and other important sensing parameters of semiconductor metal oxide nanomaterials can be enhanced by compositing semiconductor metal oxides (Zhou et al., 2015; Tomer and Duhan, 2016; Wang et al., 2017). Jae-Hun Kim et al. systematically investigated the sensing applications of xSnO2-(1-x)Co3O<sup>4</sup> composite nanofibers and reported that the 0.5SnO2-0.5Co3O<sup>4</sup> sensor exhibited the most outstanding sensing characteristics (Kim et al., 2017). As one of the most important decomposition components of SF6, H2S has been widely studied in the past few years. A variety of composite metal oxides like Cu2O-SnO<sup>2</sup> (Cui et al., 2013), CeO-SnO<sup>2</sup> (Fang et al., 2000), NiO-ZnO (Qu et al., 2016), and PdO-NiO (Balamurugan et al., 2017) have been reported as promising materials for H2S gas sensing applications. However, the report of ZnO-SnO<sup>2</sup> composite nanofibers for H2S gas sensing has been only investigated in a limited number of reports.

In this present work, we have successfully synthesized ZnO-SnO<sup>2</sup> nanofibers by electrospinning method and systematically investigated their sensing performances to H2S gas. The prepared ZnO-SnO<sup>2</sup> nanofibers exhibited significantly improved sensing properties containing high response, low detection limit, low operating temperature and fast response/recovery times to H2S gas detection, which can be ascribed to the large surface area of nanofiber structure and the formation of n-n heterojunctions at interface between ZnO and SnO2. Finally, a plausible sensing mechanism for the proposed ZnO–SnO<sup>2</sup> composite nanofibers sensor to H2S was also discussed.

## EXPERIMENTAL

### Materials Synthesis

Zinc nitrate hexahydrate (Zn(NO3)2·6H2O), stannic chloride pentahydrate (SnCl4·5H2O), N,N-dimethylformamide (DMF), polyvinylpyrrolidone (PVP, Mw = 1,300,000) and ethanol were of analytical graded and used directly without further purification. All chemicals were purchased from Chongqing Chuandong Chemical Reagent Co., Ltd (Lu et al., 2018).

In the typical synthesis of ZnO-SnO<sup>2</sup> composite nanofibers, 0.7 g of SnCl4·5H2O and 0.6 g Zn(NO3)2·6H2O (the molar ratio was 1:1) were dissolved in 5 ml of mixed solvents of ethanol and DMF (The volume ratio was 1:1) and stirred for 2 h. Then 2 g of PVP was added to the mixture and stirred for 24 h to form a viscous and homogeneous solution at room temperature. During electrospinning, the obtained mixture was delivered to a glass syringe. A voltage of 25 kV was applied between the flat aluminum foil and syringe at an electrode distance of 15 cm as shown in **Figure 1** and the flow rate is 0.7 ml/h. Finally, the electrospun nanofibers were transferred to a tube furnace and the specimens were annealed at 600◦C for 3 h in air for the removal of PVP. For comparison, the pure SnO<sup>2</sup> nanofibers were also synthesized without adding Zn(NO3)2·6H2O. The schematic illustration of producing electrospun ZnO-SnO<sup>2</sup> composite nanofibers was shown in **Figure 1** (Bai et al., 2018).

# Materials Characterization

To investigate the structures of electrospun ZnO-SnO<sup>2</sup> nanofibers, X-ray diffraction (XRD, D/Max-1200X, Rigaku, Japan) analysis was carried out at room temperature using a Rigaku D/Max-1200X diffractometry with Cu-Kα radiation (λ = 1.542 Å) over Bragg angles from 20◦ to 75◦ and the scanning speed of 2 deg/min. The morphologies of the electrospun nanofibers were investigated with a field emission scanning electronic microscopy (FESEM, JSM-6700F, JEOL, Japan) and transmission electron microscopy (TEM, JEM-2100, JEOL, Japan) operated at 120 kV. The energy dispersive X-ray spectrum analysis (EDS, Oxford INCA 250, JEOL, Japan) and X-ray photoelectron spectroscopy analysis (XPS, KRATOS X SAM800, Kratos, Kingdom) were tested to analyze the elemental compositions of the sample (Lu et al., 2018).

### Gas Sensor Measurements

Gas sensors were fabricated with a side heated structure as shown in **Figure 2A** and a theoretic diagram of the test circuit was showed in **Figure 2B**. As shown in **Figure 2A**, there are two gold electrodes connected with platinum wire at both ends of the ceramic tube. Firstly, the as-prepared powder was mixed with appropriate amount of anhydrous ethanol and deionized water to form a homogeneous paste (Xu et al., 2015). Then the obtained paste was coated onto a prefabricated ceramic tube to form the sensing film and dried at room temperature for 2 h. Next a Ni–Cr heating wire was inserted into the ceramic tube to finish the side heated H2S gas sensor. Finally, the stability of the sensing materials was improved by putting the sensor on the aging instrument of the side heated sensor at 120◦C for 10 days. The sensor response was defined as S = Ra/R<sup>g</sup> (Zhang Q. Y. et al., 2017), where R<sup>a</sup> and R<sup>g</sup> were the resistances of the sensing material measured in air and in atmosphere containing target test gas H2S, respectively (Zhu et al., 2017). Gas sensing properties of the obtained sensors were performed with the CGS-8 TP intelligent gas sensing analysis system (Chemical gas sensor-8, Beijing Elite Tech Co., Ltd., China). The response time was defined as the time required by the sensor to reach 90% of the final stable resistance when target gas in. The recovery time is the time required to return to 90% of its original baseline resistance when the sensor was exposed in air again (Nan et al., 2017). The sensing measurement were tested under laboratory condition with room temperature 25◦C and constant humidity (50% relative humidity).

# RESULTS AND DISCUSSION

# Morphology and Structure

**Figure 3** shows the XRD pattern of the electrospun ZnO-SnO<sup>2</sup> composite nanofibers. It can be seen from **Figure 3** that the XRD peaks are in accordance with the hexagonal wurtzite ZnO and tetragonal rutile SnO2, compared with the standard pattern of JCPDS card No. 36-1451 and No. 41-1445, respectively. There is neither apparent peak shift nor any other phase corresponding SnO, ZnSnO<sup>3</sup> and Zn2SnO4, confirming that there are only SnO<sup>2</sup> and ZnO co-exist in the prepared material. The result shows a possibility of developing n–n heterojunction at the interface between ZnO and SnO<sup>2</sup> nanomaterial (Bai et al., 2018). The crystallite sizes of nanoparticels were calculated using Scherrer's equation (D = kλ/βcosθ) (Lu et al., 2018), and the average crystallite sizes of nanofibers were calculated by diffraction peaks (100), (101) for ZnO, and (110), (101), (221) for SnO2. After calculation, the average crystallite size of ZnO is 20.1 nm, while it is 19.8 nm of SnO2.

**Figure 4** presents the FESEM and TEM images of the as-prepared ZnO-SnO<sup>2</sup> nanofibers after annealing. **Figure 4a**

consists of randomly oriented nanofibers with diameters of 80–150 nm and lengths of 0.5–2µm. Besides, the ZnO-SnO<sup>2</sup> nanofibers consist of nanoparticles and surface of nanofibers

is rough, which can be attributed to thermal decomposition of PVP caused by annealing. More structural information of the ZnO-SnO<sup>2</sup> nanofibers is researched by TEM characterization as shown in **Figure 4b**, the single fiber is composed of grain-like nanoparticles with around 20 nm in size.

**Figure 5** demonstrates the EDS spectrum of the electrospun ZnO-SnO<sup>2</sup> nanofibers. It can be seen that the ZnO–SnO<sup>2</sup> composite nanofibers are composed of Zn, Sn, and O elements and the atomic ratio of Zn and Sn is about 19.63:20.76.

To further analyze the compositions and element valences of the ZnO-SnO<sup>2</sup> nanofibers, XPS tests were investigated. **Figure 6A** shows the full range XPS survey spectra of the sample. It confirms the presence of Zn, O, Sn from prepared nanomaterial and C element which is due to the carbon contamination. The binding energies were calibrated using C 1 s hydrocarbon peak at 284.54 eV. **Figures 6B**,**C** demonstrate the high resolution spectra of the Zn 2p and Sn 3d energy state, respectively. The Zn 2p XPS spectrum (**Figure 6B**) presents the doublet peaks located at binding energies of 1045.6 eV and 1022.6eV, which corresponds to Zn 2p1/<sup>2</sup> and Zn 2p3/2, respectively (Li W. Q. et al., 2015). The result indicates that the Zn2<sup>+</sup> is the dominant species in the prepared material and in good agreement with the reported data for ZnO (Zhao et al., 2015). **Figure 6C** shows the binding energy of Sn 3d5/2, Sn 3d3/<sup>2</sup> are 487.6eV and 496.1eV respectively, which are assigned to the highest oxidation state of Sn4<sup>+</sup> for SnO<sup>2</sup> (Hamrouni et al., 2014; Chen et al., 2018). It further confirmed that ZnO and SnO<sup>2</sup> coexist in the samples.

### Gas-Sensing Properties

The gas sensor responses of pure SnO<sup>2</sup> and ZnO-SnO<sup>2</sup> nanofibers sensors as a function of temperatures in the range of 100–375◦C toward 50 ppm of H2S gas were tested and shown in **Figure 7**. The sensing responses of all the prepared sensors increase with increasing temperature and attain a maximum value at 300 and 250◦C for pure SnO<sup>2</sup> and ZnO-SnO<sup>2</sup> nanofibers sensor, respectively. With further increase in temperature, the sensing responses begin to decrease because desorption of H2S is dominated and the amount of the adsorbed gas onto the surface decreases (Zhang et al., 2018). The response of gas sensor based on ZnO-SnO<sup>2</sup> nanofibers for 50 ppm H2S gas at operating

temperature of 250◦C is 66.23, while it is 10.49 and 300◦C for the pure SnO<sup>2</sup> nanofibers based sensor. The results indicate that ZnO-SnO<sup>2</sup> composite nanofibers can obviously improve the response to H2S at different working temperatures and reduce the optimal operating temperature.

**Figure 8** shows the H2S gas responses of pure SnO<sup>2</sup> and ZnO-SnO<sup>2</sup> nanofibers based sensors to different concentration of H2S in the range of 0.5–100 ppm at their optimal operating temperatures measured above. The measured results show that gas responses of the as-prepared gas sensors increase in a good linear relationship from 0.5 to 100 ppm. The linear relationship of the response and gas concentration satisfies linear equation y = 1.1003x+6.90664 for electrospun ZnO-SnO<sup>2</sup> nanofibers gas sensor. The higher response of ZnO-SnO<sup>2</sup> nanofibers can be explained by the formation of n-n heterojunctions at the interface between ZnO and SnO2. Moreover, the sensor detection limit was defined as the target gas concentration value at which the response is above 3. The response of the ZnO-SnO<sup>2</sup> nanofibers sensor to 0.5 ppm H2S gas can reach up to 3.45, indicating that the detection limit of the sensor for detecting H2S gas is as low as sub-ppm level.

The dynamic response and recovery curve of the ZnO-SnO<sup>2</sup> nanofibers sensor for 1, 5, 30, and 50 ppm H2S gas was performed and shown in **Figure 9**. The obtained sensing response values

are about 7.25, 13.37, 40.31, and 66.23, respectively. The ZnO-SnO<sup>2</sup> nanofibers sensor responds rapidly and could recover to its initial value when it was exposed to air again, implying a satisfying stability and reproducibility of the proposed H2S gas sensor.

**Figures 10A**,**B** illustrate the response-recovery curve of electrospun ZnO-SnO<sup>2</sup> nanofibers sensor and pure SnO<sup>2</sup>

nanofibers sensor to 50 ppm of H2S gas at their optimal operating temperature mentioned above. From the curves, it is observed that the response and recovery time of the ZnO-SnO<sup>2</sup> nanofibers sensor is about 18 and 32 s, respectively, whereas for pure

SnO<sup>2</sup> nanofibers sensor the corresponding values is 24 and 38 s, respectively.

Finally, the long-term stability of the fabricated ZnO-SnO<sup>2</sup> nanofibers sensor was measured to 10, 30, 50, and 100 ppm H2S gas at 250◦C for 30 days as shown in **Figure 11**. The measured results show that the response has little change for 30 days and confirm a good stability of the fabricated electrospun ZnO-SnO<sup>2</sup> nanofibers sensor.

Experimental results of ZnO-SnO<sup>2</sup> nanofibers sensor have been compared with the results reported by the other workers on H2S sensors and presented in **Table 1**. It can be seen that electrospun ZnO-SnO<sup>2</sup> nanofibers sensor toward H2S can reach a relatively higher response at lower temperature and the response time is relatively shorter than other sensors reported previously. The obtained results indicate that the electrospun ZnO-SnO<sup>2</sup> nanofibers sensor is promising for H2S gas sensing.

### Sensing Mechanism

ZnO and SnO<sup>2</sup> belong to typical n-type semiconductors, characterized by their high free carrier concentration (Hong

et al., 2017; Zhou et al., 2018d). The gas sensing mechanism of ZnO-SnO<sup>2</sup> nanofibers is shown in **Figure 12**. Due to the sensing mechanism of ZnO-SnO<sup>2</sup> sensor follows the surface controlled type, the gas sensing properties are ascribed to the change of the surface resistance, which controlled by the adsorption and desorption of oxygen on the surface of sensing materials (Wei et al., 2014). When ZnO-SnO<sup>2</sup> nanofibers sensor is exposed to air (**Figure 12A**), the resistance of gas sensor depends on the amount of chemisorbed oxygen species (O−, O<sup>−</sup> 2 , and O2−). The free oxygen molecules are absorbed on the surface and capture electrons from the conduction band of the ZnO-SnO<sup>2</sup> nanofibers, which causes a depletion layer around the surface and the increasing the resistance (Cheng et al., 2014). When ZnO-SnO<sup>2</sup> nanofibers sensing materials are exposed to H2S (**Figure 12B**), the target gas reacts with the adsorbed oxygen and then releases the captured electrons into the conduction band of ZnO-SnO<sup>2</sup> nanifibers to reduce the depletion layer and decrease the resistance.

It is well-known that the chemisorbed oxygen depends on the specific surface area of sensing materials and the operating temperature. ZnO-SnO<sup>2</sup> nanofibers show a big surface area as shown in **Figure 4**. It means that adsorption capability of ZnO-SnO<sup>2</sup> nanofibers was greatly enhanced (Zhang W. et al., 2017). Moreover, O2<sup>−</sup> and O<sup>−</sup> species are regarded as the most oxygen adsorption species at 250◦C, and the following H2S sensing reaction can be considered (Kolhe et al., 2017).

$$\text{H}\_2\text{S} + \text{3O}^- \rightarrow \text{SO}\_2 + \text{H}\_2\text{O} + \text{3e}^- \tag{1}$$

$$\text{H}\_2\text{S} + 3\text{O}^{2-} \rightarrow \text{SO}\_2 + \text{H}\_2\text{O} + 6\text{e}^-\tag{2}$$

In accordance with the definition about gas response (S = Ra/Rg), the increasing of R<sup>a</sup> and decreasing of R<sup>g</sup> cause that the response of ZnO-SnO<sup>2</sup> nanofibers is significantly enhanced. Additional electron consumption will occur at the boundaries of ZnO and SnO2, which further enhances the gas response. The contact of ZnO and SnO<sup>2</sup> provides condition for electrons transfer from SnO<sup>2</sup> to ZnO, which has a higher work function of 5.2 eV compared to SnO<sup>2</sup> (4.9 eV). It results in the formation TABLE 1 | Summary of the H2S gas sensing performances of different gas sensor materials.


Temp., the optimal working temperature. τres and τrec - response time and recovery time.

of an additional depletion layer in the vicinity region between the ZnO and SnO2, eventually generating a potential barrier for electron flow (Choi et al., 2013). The boundary barrier may decrease when the gas sensor is exposed in H2S. More electrons of oxygen species transfer to the sensing material by reaction of H2S with oxygen species, which results the resistance of the sensing material decreases and the response of the sensor increases. However, pure SnO<sup>2</sup> does not provide reaction interface, leading to lower gas response. So the ZnO–SnO<sup>2</sup> composite nanofibers exhibit better sensing properties than the SnO<sup>2</sup> nanofibers.

# CONCLUSIONS

In summary, ZnO-SnO<sup>2</sup> composite nanofibers were successfully synthesized by electrospinning method and characterized by various techniques. H2S sensing properties of the electrospun nanofibers sensor were also investigated. Compared to the pure SnO<sup>2</sup> nanofiber sensor, the ZnO-SnO<sup>2</sup> composite nanofibers sensor shows excellent gas sensing response for H2S gas, which is attributed to the large specific surface and the heterojunctions between SnO<sup>2</sup> and ZnO. The proposed ZnO-SnO<sup>2</sup> composite nanofibers sensor exhibits good linear relationship between sensing response and gas concentration in the range of 0.5∼100 ppm and its detection limit is as low as sub-ppm level. Moreover, the proposed sensor achieves good repeatability and longterm stability, making it a promising candidate for detecting H2S gas.

### AUTHOR CONTRIBUTIONS

ZL and ZW performed the experiments and analyzed the data with the help from LX and YG. ZL, QZ, and CW wrote and

### REFERENCES


revised the manuscript with input from all authors. All authors read and approved the manuscript.

### ACKNOWLEDGMENTS

This work has been supported in part by the National Natural Science Foundation of China (No. 51507144), the China Postdoctoral Science Foundation Project (Nos. 2015M580771, 2016T90832), the Chongqing Science and Technology Commission (CSTC) (No. cstc2016jcyjA0400) and the project of China Scholarship Council (CSC).

film for H2S detection. J. Alloy. Compd. 748, 6–11. doi: 10.1016/j.jallcom.2018. 03.123


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Lu, Zhou, Wang, Wei, Xu and Gui. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Gas Sensing Performances of ZnO Hierarchical Structures for Detecting Dissolved Gases in Transformer Oil: A Mini Review

### He Zhang<sup>1</sup> \*, Wei-Gen Chen<sup>1</sup> \*, Yan-Qiong Li <sup>2</sup> and Zi-Hao Song<sup>1</sup>

<sup>1</sup> State Key Laboratory of Power Transmission Equipment & System Security and New Technology, School of Electrical Engineering, Chongqing University, Chongqing, China, <sup>2</sup> School of Electronic and Electrical Engineering, Chongqing University of Arts and Sciences, Chongqing, China

Power transformer is one of the critical and expensive apparatus in high voltage power system. Hence, using highly efficient gas sensors to real-time monitor the fault characteristic gases dissolved in transformer oil is in pressing need to ensure the smooth functionalization of the power system. Till date, as a semiconductor metal oxide, zinc oxide (ZnO) is considered as the promising resistive-type gas sensing material. However, the elevated operating temperature, slow response, poor selectivity and stability limit its extensive applications in the field of dissolved gases monitoring. In this respect, rigorous efforts have been made to offset the above-mentioned shortcomings by multiple strategies. In this review, we first introduce the various ZnO hierarchical structures which possess high surface areas and less aggregation, as well as their corresponding gas sensing performances. Then, the primary parameters (sensitivity, selectivity and stability) which affect the performances of ZnO hierarchical structures based gas sensors are discussed in detail. Much more attention is particularly paid to the improvement strategies of enhancing these parameters, mainly including surface modification, additive doping and ultraviolet (UV) light activation. We finally review gas sensing mechanism of ZnO hierarchical structure based gas sensor. Such a detailed study may open up an avenue to fabricate sensor which achieve high sensitivity, good selectivity and long-term stability, making it a promising candidate for transformer oil monitor.

### Edited by:

Zhongchang Wang, Laboratório Ibérico Internacional de Nanotecnologia (INL), Portugal

### Reviewed by:

Ming-Guo Ma, Beijing Forestry University, China Tianming Li, Leibniz-Institut für Festkörper-und Werkstoffforschung (IFW Dresden), Germany

### \*Correspondence:

He Zhang sophy305410@163.com Wei-Gen Chen weigench@cqu.edu.cn

### Specialty section:

This article was submitted to Nanoscience, a section of the journal Frontiers in Chemistry

Received: 10 August 2018 Accepted: 03 October 2018 Published: 22 October 2018

### Citation:

Zhang H, Chen W-G, Li Y-Q and Song Z-H (2018) Gas Sensing Performances of ZnO Hierarchical Structures for Detecting Dissolved Gases in Transformer Oil: A Mini Review. Front. Chem. 6:508. doi: 10.3389/fchem.2018.00508 Keywords: ZnO, gas sensors, hierarchical structures, sensitivity, selectivity, stability, gas sensing mechanism

# INTRODUCTION

Power transformer is one of the most critical and expensive devices in high voltage power system (Christina et al., 2018). Generally, oil is used inside the transformer for its operation and can release different fault characteristic gases, such as hydrogen (H2), carbon oxides (CO, CO2) and hydrocarbons (CH4, C2H2, C2H4, and C2H6). Hence, real-time detection of dissolved gases in transformer oil is very essential in order to avoid unexpected failures (Mariprasath and Kirubakaran, 2018). At present, dissolved gas analysis (DGA) remains to be the simplest and most effective diagnostic method for checking latent faults of oil-immersed power transformers (Siada and Hmood, 2015; Fan et al., 2018). Therefore, using highly efficient gas sensors to realtime monitor these dissolved gases in transformer oil is a feasible way to ensure the stability and reliability of power system (Uddin et al., 2016).

**19**

Different types of gas sensors have already been applied in the online detection of dissolved gases in transformer oil, such as resistance-type (Benounis et al., 2008; Sun et al., 2015), optical-type (Ma et al., 2012) and electrochemical-type (Ding et al., 2014). Among diverse types of gas sensors, resistancebased sensors stand out owing to advantages like the small, cheap, high sensitivity and low power consumption (Bodzenta et al., 2002; Yang et al., 2011; Zhao et al., 2017; Xu et al., 2018). With the increasing demand for better gas sensors of higher sensitivity and selectivity (Sun et al., 2012; Gardon and Guilemany, 2013), countless endeavors have been poured on hunting for more suitable sensing nanomaterials. Semiconductor metal oxides (MOS), such as zinc oxide (ZnO), tin oxide (SnO2), tungsten oxide (WO3), etc., have received wide research for gas sensing applications and so on. Among these, the gas sensing performance of ZnO-based gas sensor was first investigated by Seiyama et al. (1962). As a typical n-type semiconductor material with a direct wide band gap (Eg ≈ 3.37 ev) and large excitation binding energy (∼60 mev), ZnO has got important status in various MOS nanomaterials due to its high carrier mobility of conduction electrons, good chemical and thermal stability (Zeng et al., 2015; Das and Sarkar, 2017; Ganesh et al., 2017).

The gas sensing properties of ZnO greatly depend on its structure and morphology including surface area, size, orientation and crystal density (Cho et al., 2011). Therefore, tailoring the structure and morphology of ZnO is particularly important to optimize the gas sensing performances (Liao et al., 2008). In particular, the elaborate design of unique three-dimensional (3D) hierarchical architectures can fully achieve this, since such hierarchical structures possess high surface area and fast gas diffusion as well as reduce the agglomerated configuration of low dimensional structures.

ZnO-based gas sensors commonly have the shortcomings of slow response, poor selectivity and lack of long-term stability, which limits the wide applications. To acquire an efficient and reliable dissolved gases sensor, high sensitivity, selectivity, longterm stability, low response / recovery time, low fabrication cost are urgently needed (Wang et al., 2012). This review focuses on the factors that affect the performances (sensitivity, selectivity and stability), the methods to improve these sensor parameters and gas sensing mechanism of ZnO-based gas sensors.

### GAS SENSING PERFORMANCES OF ZnO-BASED GAS SENSOR

## Effects of Morphologies About ZnO Hierarchical Structures on Gas Sensing Performances

Three-dimensional (3D) hierarchical structures are generally recognized as the best candidate for gas sensing performances, compared with low-dimensional structures (Mo et al., 2008; Guo, 2016). They are defined as those assembled by zerodimensional (0D), one-dimensional (1D) and two-dimensional (2D) components, which can be further classified into the following sub-sections. (1) Assembly of 0D structures: Li W. Q. et al. (2015)reported the synthesis of pure ZnO hollow nanofibers by electrospinning method. The walls of ZnO nanofibers consist of the aggregation of many individual nanoparticles, as shown in **Figure 1A**. The sensor based on ZnO hollow nanofibers exhibits excellent sensing performance for acetone detection, which can be attributed to the large aperture and small diameters provide higher specific surface area for gas adsorption. Chen H. et al. (2016) synthesized the uniformly monodispersed ZnO nanospheres via a simply hydrothermal route. In particular, all the microparticles on the surface are sphere-shapes and have a rough surface, as shown in **Figure 1B**. This unique porous structure exhibits perfect sensing performance toward ethanol. (2) Assembly of 1D structures: Lin et al. (2015) reported the hierarchical ZnO microstructures by hydrothermal method. The morphology of the sample likes a bunch of flowers which is made of uniform nanorods, as shown in **Figure 1C**. The sensor based on the sample shows a good response. Chen H. et al. (2016) reported the sea-urchin-like ZnO nanostructures by hydrothermal method. The sample is composed of many strips and radiates from the center, as shown in **Figure 1D**. The sensor based on the ZnO sample toward ethanol exhibits high sensitivity and quick response. (3) Assembly of 2D structures: Gu et al. (2011) reported the porous flower-like ZnO nanostructures by economical hydrothermal synthesis combined with subsequent calcination. Calcination of the precursors produced flower-like ZnO nanostructures which composed of interconnected porous ZnO nanosheets with high porosity, as shown in **Figure 1E**. The as-prepared flower-like ZnO nanostructures are highly promising candidate for applications of gas sensors. Han et al. (2016) reported the ZnO hollow spheres with high crystallinity via a simple template process, as shown in **Figure 1F**. The surfaces of these core-shell spheres are rough, suggesting that polystyrene sphere (PSS) core was coated by ZnO nanoparticles. The sensor based on ZnO hollow spheres exhibits good sensing performances.

In this part, the authors make a brief introduction with respect to the hierarchical structures. Hierarchical hollow or porous ZnO structures exhibit excellent properties for gas sensor applications (Guo et al., 2011, 2012, 2013). These unique hollow structures with large specific surface area and highly porous structures can provide excellent channel and "surface accessibility" for the gas transportation, which is very favorable for facilitating the interaction of ZnO surface with the gas molecules (Gu et al., 2011). No matter how complicated the hierarchical structure, it's all derived from low dimensional nanostructures as building blocks. Hence, the investigation about regulating the structure and morphology is a meaningful and challenge work.

A summary about factors affecting gas sensing performances of ZnO-based gas sensors and improvement approaches is shown in **Table 1**. The details are described in sections Factors affecting the sensitivity of ZnO hierarchical structure based Gas sensor, Factors affecting the selectivity of ZnO hierarchical structure based Gas sensor, and Factors affecting the long-term stability of ZnO hierarchical structure based Gas sensor as follows.

FIGURE 1 | ZnO with different 3D hierarchical structures: (A) Nanofibers assembly by 0D structures. Reprinted with permission from Li X. et al. (2015). Copyright (2015) Elsevier Science BV. (B) Nanospheres assembly by 0D structures. Reprinted with permission from Chen H. et al. (2016). Copyright (2016) Elsevier Science SA. (C) Flower-like microstructure assembly by 1D structures. Reprinted with permission from Lin et al. (2015). Copyright (2015) Elsevier Science SA. (D) Sea-urchin-like ZnO nanostructures assembly by 1D structures. Reprinted with permission from Chen H. et al. (2016). Copyright (2016) Elsevier Science SA. (E) Porous flower-like ZnO nanostructures assembly by 2D structures. Reprinted with permission from Gu et al. (2011). Copyright (2011) Elsevier Science SA. (F) Core-shell hollow spheres assembly by 2D structures. Reprinted with permission from Han et al. (2016). Copyright (2016) Elsevier Science SA.

# Factors Affecting the Sensitivity of ZnO Hierarchical Structure Based Gas Sensor

Recently, numerous reports confirmed that ZnO-based nanomaterials are promising candidates for the fabrication of gas sensors (Gu et al., 2013; Wang et al., 2014). Given this, a number of strategies have been proposed for enhancing the gas sensitivity. It can be introduced from the following four aspects.

• Modulation of the dimensional and the exposed crystal facet of their constituting building blocks.

Zhang et al. (2009) synthesized brush-like hierarchical ZnO nanostructures. The FESEM image (**Supplementary Figure 1A**) shows that this structure is composed of 6-fold nanorod arrays TABLE 1 | A summary about factors affecting gas sensing performances of ZnO hierarchical structure based gas sensors and improvement approaches.


grown on the side surface of core nanowires. The central stems provide its six prismatic facets as growth platforms for branching of multipod units. The sensor based on these structures shows high sensitivity and fast response.

• Enhance the porosity of hierarchical structures.

Lei et al. (2017) successfully synthesized hierarchical porous ZnO microspheres assembled from 2D nanosheets. The high specific surface area and hierarchical pore structure are beneficial to increase the adsorption capacity (**Supplementary Figure 1B**). Song et al. (2018)reported hierarchical porous ZnO microflowers which composed of ultrathin nanosheets. From the SEM image (**Supplementary Figure 1C**), we can see that the surface of nanosheet has lots of pores. The porous structure is favorable for gas sensor to promote the inward/outward gas diffusion and improve gas sensitivity.

• Modification by doping with noble metals and loading other n-type or p-type MOS materials.

It's known that noble metals, such as Pt (Rout et al., 2006), Pd (Yang et al., 2010) and Au (Vallejos et al., 2011) are frequently used in gas sensing materials due to doping can sensitize the ZnO electronic and structural properties. Lin et al. (2015) reported that Au nanoparticles were decorated on the surface of hierarchical flower-like ZnO microstructures, as shown in **Supplementary Figure 1D**. After Au nanoparticle-decoration, the specific surface area is much higher than that of the bare ZnO (**Figure 1C**). Au nanoparticles can act as catalysts to accelerate the chemisorption process and greatly improve the sensitivity. So far, heterostructure composites consisting of two metal oxides, such as (n-n type) SnO2/ZnO (Park et al., 2013) and (n-p type) NiO/ZnO, AgO/ZnO (Gandomania et al., 2014) have been successfully prepared and have improved the sensitivity. Liu et al. (2017) reported the NiO nanoparticles which were decorated onto the surfaces of well-dispersed ZnO hollow spheres (**Supplementary Figures 1E,F**). Such hollow structures with rough surfaces endow the NiO/ ZnO composites high surface areas and abundant active sites, which could facilitate the gas diffusion toward the entire materials and an improvement of the sensitivity (Lee, 2009).

• Control of grain size.

Previous research found that sensors which consist of fine particles of MOS tend to exhibit high sensitivity. Thus, one of the most important factors affecting the sensitivity is grain size (D) of the sensor materials in conjunction with the thickness of the space charge layer (L). **Supplementary Figure 1G** illustrates three kinds of schematic models for grain-size effects (Shimizu and Egashira, 1999). When D >> 2L, the conductance is limited by Schottky barrier at grain boundaries (grain boundary control). If D ≥ 2L, the conductance is limited by necks between grains (neck control). When D < 2L, the conductance is controlled by grains themselves (grain control). Among three models, grain control is the most sensitive condition (Mirzaei et al., 2018). The smaller the grain size, the higher the sensitivity of gas sensor. But, excessive decrease in grain size can reduce structural stability.

# Factors Affecting the Selectivity of ZnO Hierarchical Structure Based Gas Sensor

Selectivity is the ability of gas sensor to recognize the target gas in a mixture of other gases. Generally, there are two approaches for enhancing the selectivity of gas sensor. The first one is to synthesize a material which is selective to the specific compound and has very low cross-sensitivity for other compounds. Moreover, the synergistic effect of two component system is greater than the production effect of the two elements. In fact, noble metals and p-type metal oxides have been extensively applied as good catalysts in the two component systems to promote selectivity of sensors (Li T. M. et al., 2015). Another approach to improve the selectivity is to combine with other methods. Recently, some reports have suggested that lowering the operating temperature can be realized by activating the sensing material under UV illumination (Helwig et al., 2009; Lu et al., 2012; Cui et al., 2015). The possible UVactivated selective photo catalysis plays an important role in the enhancement of the selectivity at low temperature (Li X. et al., 2015). It can be explained based on the selective photocatalytic oxidation. The adsorbed oxygen would be re-activated by the photon generated electron-hole pairs, which is conductive to enhancing their reactivity with target gas. After the target gas reacted with the adsorbed oxygen on ZnO surface, the donated electrons would thus decrease resistance of the sensor and finally reduce the operating temperature (Ho et al., 2015). Chen et al. reported that the mesoporous hollow ZnO microspheres were applied to detect volatile organic vapors (VOCs) with the help of UV LED illumination at lower temperatures (Chen Y. et al., 2016). The sensor with UV activation at 80◦C shows a much higher response to ethanol (**Supplementary Figure 2A**). When the sensor was operated at 220◦C, the UV illumination became ineffective. It shows almost same response to ethanol and acetone (**Supplementary Figure 2B**). This is because the difference about catalytic conversion of O2<sup>−</sup> would have negligible toward them at 220◦C. However, the O<sup>−</sup> possibly indicated higher preference to ethanol at 80◦C, resulting in the better selectivity. When metal doped-ZnO was illuminated by UV light, the sensor had an appreciable selectivity at low temperature, which was attributed to the heterostructure was in favor of chemical interactions, adsorption of gases and changes in electronic bind energies in the composite (Chen et al., 2008). Espid investigated the photoresponsive behavior of ZnO/ In2O<sup>3</sup> composite sensors (Espid and Taghipour, 2017). When the semiconductor composites are irradiated with photons emitted from a UV source, the photogenerated electron/hole pairs will enhance the conductance of the sensing layer and improve the selectivity.

## Factors Affecting the Long-Term Stability of ZnO Hierarchical Structure Based Gas Sensor

Stability is a key parameter for the long-term development of gas sensors, which determines its application state in the real market. Generally, the long-term stability refers to the attenuation degree of gas sensing performances (e.g., sensitivity, selectivity, response and recovery time) during a certain period of time. When the sensor is in working state, working conditions including high temperature and toxic gases can reduce the stability. When the sensor is in normal storage state, changes of humidity, fluctuations of temperature in the surrounding atmosphere may also interfere with the stability of sensor. At present, there is not a recognized method to improve stability of ZnO-based gas sensors. Stability can be increased to some extent by calcination/ annealing as the post-processing treatment (Gu et al., 2011) and reducing the working temperature of gas sensing element. Chen et al. tested the long-term stability of ZnO-based sensor working at 80◦C with UV activitation (Chen Y. et al., 2016). The sensor test lasted 1 month (**Supplementary Figure 2C**). In the first 2 days, the response values dropped significantly, which might be related to the "pre-aging" effect. In the next few days, the sensor response became stabilized and showed a good long-term stability. It might be because the microstructure of the materials had little change under low temperature with low-powered UV activation. In addition, doping noble metal or synthesis of mixed oxides can also increase the stability of the sensors (Dey, 2018).

## GAS SENSING MECHANISM OF ZnO HIERARCHICAL STRUCTURE BASED GAS SENSOR

By summarizing the methods to improve the gas sensing performances in section Gas Sensing Performances of ZnO-Based Gas Sensor, we find that metal doping is an excellent method to promote sufficient reaction between sensing material and target gas.

The gas sensing mechanism of noble metals doped-ZnO hierarchical structures based gas sensors is explained as an example. This process mainly involves two effects: chemical effect and electronic effect (Zhu and Zeng, 2017). Firstly, the chemical effect is related to spillover process (Nakate et al., 2016). Oxygen molecules were adsorbed on the surface and grain boundary of ZnO, forming the oxygen ions. The sensitization of noble metals increases the quantity of oxygen species and accelerates the surface reaction, causing an expansion of charge depletion layer, which results in a higher baseline resistance (**Supplementary Figures 3A,B**). When the reducing gas is introduced, the catalysis of noble metals may give rise to the dissociation of target gas molecules. The trapped electrons are released and transmitted to the conduction band, resulting in a remarkable decrease in depletion layer with a lower resistance. Secondly**,** the electric effect is produced by contact resistance of noble metal modified ZnO gas sensors (Hosseini et al., 2015). Electrons from the conduction band of ZnO transfer into noble metals owing to their work functions are different, forming the Schottky barriers at noble metal-ZnO interface, which leading to generate the additional depletion region near ZnO surface (**Supplementary Figure 3C**).

Therefore, the enhanced sensing performance was ascribed to the spillover phenomenon, the formation of Schottky barriers at the interface between noble metals and ZnO, more introduced surface active sites and effective surface areas (Hosseini et al., 2015).

### CONCLUSION

A study on gas sensing performances of ZnO hierarchical structures has been shortly summarized in this review. Firstly, unique 3D hierarchical architectures with high sensing capabilities are discussed by modifying surface morphologies. Small grain size, high effective specific surface area and porosity are favorable to the enhancement of gas sensing performances. Therefore, the preparation of the desired 3D hierarchical structure can lay a solid foundation for the development of gas sensor. Then, factors that affect the sensitivity, selectivity and stability of ZnO hierarchical structures based gas sensors and their improvement strategies are summarized separately. Among these methods, additive doping and UV-light irradiation are

### REFERENCES


more effective methods to improve gas sensing performances. The former can increase charge carrier concentration and decrease activation energy. The latter can promote the catalytic oxidation reaction between target gases and oxygen ions, thus reduce the working temperature and power consumption. Numerous reports indicate that the integration of metal doped-oxide and UV excitation is one of the most effective and workable attempts to achieve high sensor performances. The composite oxides based sensors under UV illumination have better charge separation, which benefit for the gas performances enhancement of the sensors. We hope our work is helpful for further exploration on higher gas sensing performances of MOS sensing materials for detecting dissolved fault gases in transformer oil. Finally, gas sensing mechanism of noble metal sensitized ZnO is illuminated from the point of view of chemical effect and electronic effect. Nevertheless, the authors suggest only a few possible ways to improve the existed oxygen-absorbed model in recent researchers. Much effort should be made to hunt for an integration of different models which was used to explain the gas sensing reaction.

### AUTHOR CONTRIBUTIONS

HZ and W-GC conceived and designed the experiments, HZ and Y-QL performed the experiments, HZ and Z-HS analyzed the data, HZ wrote the manuscript with input from all authors. All authors read and approved the manuscript.

### ACKNOWLEDGMENTS

This work was supported by the Science Fund for Creative Research Groups of the National Natural Science Foundation of China (Grant No. 51321063), the Joint Fund of the National Natural Science Foundation of China and the Smart Grid of State Grid Corporation of China (Grant No. U1766217), and the Scientific Research Fund of Chongqing University of Arts and Sciences (2017RDQ38).

## SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fchem. 2018.00508/full#supplementary-material

spheres and sea-urchin-like nanostructures. J. Alloy Compd. 687, 342–351. doi: 10.1016/j.jallcom.2016.06.153


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Zhang, Chen, Li and Song. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Enhanced Sub-ppm NH<sup>3</sup> Gas Sensing Performance of PANI/TiO<sup>2</sup> Nanocomposites at Room Temperature

### Chonghui Zhu, Xiaoli Cheng, Xin Dong\* and Ying ming Xu\*

*Key Laboratory of Functional Inorganic Material Chemistry, Ministry of Education, School of Chemistry and Materials Science, Heilongjiang University, Harbin, China*

PANI/TiO<sup>2</sup> nanocomposites spheres were synthesized using a simple and efficient one-step hydrothermal process. The morphology and structure of PANI/TiO<sup>2</sup> nanocomposites spheres were investigated by X-ray diffraction (XRD), scanning electron microscopy (SEM), and transmission electron microscopy (TEM) techniques. The PANI/TiO<sup>2</sup> nanocomposite sphere-based sensor exhibits good selectivity, sensitivity (5.4 to 100 ppm), repeatability, long-term stability and low detection limit (0.5 ppm) to ammonia at room temperature (20 ± 5 ◦C). It also shows a good linearity relationship in the range of 0.5–5 and 5–100 ppm. The excellent NH<sup>3</sup> sensing performance is mainly due to the formation of the p-n heterostructure in the nanocomposites.

Edited by:

*Wen Zeng, Chongqing University, China*

### Reviewed by:

*Jianping Yang, Donghua University, China Zhenmeng Peng, University of Akron, United States*

\*Correspondence:

*Xin Dong xdong92@yahoo.com Ying ming Xu xuyingming@hlju.edu.cn*

### Specialty section:

*This article was submitted to Nanoscience, a section of the journal Frontiers in Chemistry*

Received: *04 June 2018* Accepted: *27 September 2018* Published: *18 October 2018*

### Citation:

*Zhu C, Cheng X, Dong X and Xu Ym (2018) Enhanced Sub-ppm NH3 Gas Sensing Performance of PANI/TiO2 Nanocomposites at Room Temperature. Front. Chem. 6:493. doi: 10.3389/fchem.2018.00493* Keywords: gas sensing, polyaniline, TiO2 , NH3 , nanocomposites

# INTRODUCTION

Polyaniline (PANI), as an intrinsically conductive polymer, is the widely used material in anti-corrosion, anti-static electricity, stealth and so on. Since its low density, excellent processability, excellent flexibility, and good electrical conductivity, (Bhadra et al., 2009; Baker et al., 2017) polyaniline is also attracting more attention in the field of gas sensing. However, pure PANI sensors often show lower sensitivity or higher detection limit than semiconductor metal oxide sensors (Sutar et al., 2007; Kebiche et al., 2012). The addition of semiconducting metal oxides not only improves gas-sensing properties, but also avoids high operating temperatures. In recent years, SnO2, ZnO, TiO2, and In2O<sup>3</sup> have been chosen as the addition of polyaniline to prepare ammonia sensors, which can used at room temperature (Guo et al., 2013; Chen et al., 2015; Dai et al., 2016; Zhang et al., 2018). Most of the materials reported so far have been prepared by electrospinning, interfacial synthesis, or mechanical mixing (Talwar et al., 2014; Li et al., 2015; Nie et al., 2016). Pawar et al. synthesized PANI and TiO<sup>2</sup> by chemical oxidative polymerization and sol-gel method, respectively. Then, they mixed PANI and TiO<sup>2</sup> by mechanical mixing method to prepare PANI-TiO<sup>2</sup> nanocomposites. The response of the sensor to 100 ppm NH<sup>3</sup> is 50% (Pawar et al., 2011a). Liu et al. prepared the PANI-TiO2-Au ternary nanocomposite thin film by in-situ selfassembly method. The response of the sensor is 2.23 toward 50 ppm NH3, and the detection limit is 1 ppm (Liu et al., 2017). Their synthetic methods basically need more than one-step reactions. And the recovery time, sensitivity and other properties of the sensors need to be further improved.

In this work, the PANI/TiO<sup>2</sup> nanocomposite was synthesized by one step hydrothermal method. The PANI/TiO<sup>2</sup> nanocomposite exhibits good selectivity, sensitivity (5.4 to 100 ppm), repeatability, long-term stability, and low detection limit (0.5 ppm) to ammonia at room temperature. Meanwhile, the sensor exhibits a good liner relationship in the range of 0.5–5 and 5-100 ppm.

**26**

# EXPERIMENTAL

All reagents were of analytical grade, offered by Aladdin Reagent Company, and used without further purification.

# Preparation of PANI/TiO<sup>2</sup> Nanocomposites and Pure PANI

1.147 g of ammonium persulfate (APS) was added to 10 mL of HCl (1 mol·L −1 ) to obtain solution A. In 10 mL of anhydrous ethanol, tetrabutyl titanate (TBT), and 0.47 mL of aniline were added to obtain solution B. Solution C is the mixture of 10 mL absolute ethanol and 0.47 mL aniline. The mixture solution of A and B were poured into a 50 mL Teflon-lined stainless steel autoclave and kept at 100◦C for 2 h. The mixtures of solution A and C were treated as the same. The PANI/TiO<sup>2</sup> nanocomposite and pure PANI was obtained after washing several times by deionized water and ethanol. With other conditions remaining constant, the amount of TBT added in solution B is 0.05, 0.1, 0.15, or 0.2 mol·L −1 , respectively, to make products PT1, PT2, PT3, PT4. The synthetic routes of the samples are shown in **Scheme S1**.

# Characterization of PANI/TiO<sup>2</sup> Nanocomposites

The phase structure of the products were analyzed by X-ray powder diffractometer (XRD, Rigaku, D/MAX-3B) with Cu Kα1 radiation (λ = 1.54059 Å). Raman spectra of products were tested by a LabRAM HR800 laser confocal microscopic Raman spectrometer. The morphology and structure of products were investigated by scanning electron microscope (SEM, FEI/Philips, XL-30) and transmission electron microscope (TEM, Jeol, Jem-2100). Thermogravimetric analysis (TGA) was executed by a Perkin-Elmer instruments corporation thermogravimetric analyzer. The specific surface area and pore size distribution were detected by nitrogen adsorption-desorption measurement at 77 K using the TriStar II 3020 system.

# Fabrication and Gas Sensing Measurement of the PANI/TiO<sup>2</sup> Nanocomposites Sensors

The products were made into a paste with ethanol and coated on the surface of Al2O<sup>3</sup> tube whose length was 4 mm and internal diameter was 0.8 mm. The Al2O<sup>3</sup> tube brought with two gold electrodes which spaced 1 mm and the distance between two gold electrodes was 1 mm. Two platinum wires were connected with each gold electrode. After a Ni/Cr wire passed through the Al2O<sup>3</sup> tube, it was welded to the base together with the platinum wires.

The responses of sensors were recorded by JF02E type gas sensor measurement system (Kunming Guiyan Jinfeng Technology Co. Ltd., China). The static testing method was used to measure the property of sensors. After a certain amount of NH<sup>3</sup> was injected into a vacuum 10 L glass bottle, the pressure was returned to the atmospheric pressure. The gas sensor was placed in the glass bottle for testing. The sensor's response to NH<sup>3</sup> was defined as S = Rg/Ra. Rg and Ra were the resistance values of the sensor in NH<sup>3</sup> and air, respectively. The response time is 100s, the recovery time is defined as the time that the resistance changed 90%. The sensor response to humidity measured by the saturated solution of LiCl (11.3 RH%), CH3COOK (23.4 RH%), MgCl<sup>2</sup> (32.8 RH%), K2CO3(43 RH%), Mg(NO3)<sup>2</sup> (54.3 RH%), CuCl<sup>2</sup> (67 RH%), NaCl (75.3 RH%), KCl (85 RH%), and KNO<sup>3</sup> (93.5 RH%; Yang et al., 2017).

# RESULTS AND DISCUSSION

The XRD peaks of the nanocomposites samples PT1, PT2, PT3, and PT4, shown in **Figure 1a** at 25◦ , is consistent with the XRD diffraction peak of pure PANI. As is shown in **Figure S1**, the diffraction peak of the pure PANI appears at about 25◦ . Other peaks at 25.3◦ , 37.8◦ , and 48.0◦ are corresponding to (101), (004), and (200) crystal planes of anatase crystal structure TiO<sup>2</sup> (JCPDS Card NO.21-1272). This reminds that TiO<sup>2</sup> is present in the nanocomposites. Further, the Raman spectrum of the nanocomposites was analyzed, as shown in **Figure S2**. The peaks at 164, 406, 635 cm−<sup>1</sup> are the characteristic peaks of TiO2, and 1,595 cm−<sup>1</sup> is attributed to the C-C stretching vibration of benzenoid ring in PANI, which further proves the presence of TiO<sup>2</sup> in the nanocomposites (Chen and Mao, 2007). Through the thermogravimetric analysis (**Figure S3**) to determine the TiO<sup>2</sup> content of PT1, PT2, PT3 and PT4 is 3.04, 19.48, 21.09, and 38.75%, respectively. The response of the PT1, PT2, PT3, PT4 nanocomposites and pure PANI to 20–100 ppm NH<sup>3</sup> at room temperature is shown in **Figure S4**. As can be seen from the figure, the PT3 nanocomposites sensor has better response to NH<sup>3</sup> than the other four sensors. **Figure S5** exhibits the pore structure and specific surface area of the PT1, PT2, PT3, and PT4 nanocomposites, which were detected by N<sup>2</sup> adsorptiondesorption measurement and Barrett-Joyner-Halenda (BJH) pore size distribution analysis. The BET specific surface area of the PT1, PT2, PT3, and PT4 nanocomposites are 26.76, 46.13, 47.73, and 55.80 m<sup>2</sup> ·g −1 , respectively. The result shows that the specific surface area increase as the TiO<sup>2</sup> content rises. The gassensing performance of the sensor is generally related to the micro-structure of sensing material, specific surface area and so on. Compared with other four sensors, the morphology of the PT3 nanocomposites is uniform (**Figure S6**), many TiO<sup>2</sup> nanoparticles are uniformly dispersed on the surface of the PANI nanospheres, which makes the nanocomposites have many p-n heterostructure and a large specific surface area. These will increase the contact area of PT3 nanocomposites, make the NH<sup>3</sup> molecules diffuse easily and provide more active sites for the efficient adsorption of NH3, which is more conducive to the reaction of the nanocomposite with NH3. Therefore, a more detailed characterization of the PT3 nanocomposites was performed.

The morphology and structure of the PT3 nanocomposites were tested by SEM and TEM. In **Figure 1b**, it is found that the PT3 nanocomposites were composed by the stacking of nanoparticles whose diameter is about 80 nm. **Figure 1c** indicates the same result that the PT3 nanocomposites is composed of nanoparticles, stacked on top of each other. In addition, the

HRTEM image of the PT3 nanocomposites (**Figure 1d**) shows characteristic lattice fringes of TiO<sup>2</sup> with a pitch of 0.35 nm, corresponding to the (101) crystal plane of TiO2. This is consistent with the XRD result that there is TiO2.

The gas sensing properties of the PT3 nanocomposites sensor was further studied. The response-recovery curves for 0.5–100 ppm NH<sup>3</sup> at room temperature are shown in **Figure 2A**. To 100 ppm NH3, the sensitivity of PT3 nanocomposite sensor is 5.4, the response time is 100 s, and the recovery time is 232 s. Moreover, the sensitivity of the PT3 nanocomposite sensor presents a clear linear relationship with the NH<sup>3</sup> concentration between 0.5 and 5 ppm (R <sup>2</sup> = 0.9992) and 5 and 100 ppm (R <sup>2</sup> = 0.9945) (**Figure 2B**). The PT3 nanocomposite sensor performed four consecutive tests on 50 ppm of NH<sup>3</sup> at room temperature (**Figure 2C**). The corresponding sensitivities are 2.59, 2.6, 2.56, and 2.56, respectively. And the relative deviation is 1.75%. This shows that the PT3 nanocomposite sensor has a satisfactory reproducibility. The interference of other seven gases to the sensor was further researched (**Figure 2D**). The responds at room temperature to 100 ppm ethanol (C2H5OH), acetone (CH3COCH3), triethylamine (TEA), ethyne (C2H2), NO, H2, styrene (C8H8), and NH<sup>3</sup> are 1.020, 1.034, 1.027, 1.006, 1.008, 1.007, 1.010, and 5.423, respectively. The result proves that the PT3 nanocomposite sensor is more sensitive to NH<sup>3</sup> at room temperature. The influence of humidity was studied in the humidity range of 11.3–93.5% at room temperature (**Figure S7a**). The maximum response of the sensor is 1.4. It means that the effect of humidity to the sensor is much small. In order to deeply investigate the long-term stability, the sensor responds to 50 ppm NH<sup>3</sup> every 5 days at room temperature. During the 2 months monitoring time (**Figure S7b**), the response of the sensor was reduced by <2% after 60 days. Our investigation leads us to conclude that the sensor has a good long-term stability.

Compared the gas sensing property between the PANI-TiO<sup>2</sup> nanocomposites sensor with the reported sensor in **Table 1**. The results show that the PT3 nanocomposites sensor exhibits excellent sensitivity and good response to NH<sup>3</sup> gas at room temperature. The detection limit of PANI-TiO<sup>2</sup> nanocomposites sensor is lower and sensitivity is better than the other reported sensors. The improvement of the NH<sup>3</sup> sensing property of PT3 sensor is might attributed to the following reasons. First, One-step synthesis is more conducive to the dispersion of titanium dioxide nanoparticles on the surface of PANI. Secondly, the p-n heterostructure formed between TiO<sup>2</sup> and PANI can provide a synergistic effect, which can effectively improve the ability to adsorb NH3. Therefore, the PT3 nanocomposites has potential application value in detecting NH<sup>3</sup> at room temperature.

The possible sensing mechanism of the PT3 nanocomposites sensor for detecting NH<sup>3</sup> is as shown in **Figure S8**. TiO<sup>2</sup> is an n-type semiconductor with a 3.2 eV forbidden band width. And PANI is a p-type semiconductor with a 2.8 eV forbidden band width. At the contact interface, the TiO<sup>2</sup> and PANI interact to

FIGURE 2 | The response-recovery curves (A) and the linear relationship (B) of the PT3 nancomposites sensor to different concentrations of NH3; the reproducibility of the PT3 nancomposites sensor to 50 ppm NH3 (C); the selectivity of the PT3 nanocomposites sensor to 100 ppm of eight gases (D).


form a p-n heterostructure. The p-n heterostructure will make a positively charged depletion layer. So, the activation energy and enthalpy of physisorption for NH<sup>3</sup> will reduce as a result to cause the enhancement of gas sensitivity (Costello et al., 1996; Tai et al., 2010). In addition, the LUMO level of PANI and the conduction band of TiO<sup>2</sup> are well helps charge transfer. It can also effectively improve the gas sensing performance (Li et al., 2006; Tai et al., 2008). When the sensor is exposed to NH3, H<sup>+</sup>

on -NH- site of PANI combines with NH3, cause the electron hole concentration in the PANI to be low. The resistance of the sensor will increase. When the sensor is exposed to air after NH3, PANI gets H<sup>+</sup> from NH<sup>+</sup> 4 , the electron hole concentration of PANI recovers. The resistance decreases to the initial value (**Figure S8b**).

## CONCLUSION

In summary, the PANI/TiO<sup>2</sup> nanocomposites, which is consisted of uniform nanoparticles, are synthesized by one-step hydrothermal synthesis. At room temperature, the sensor based on PANI/TiO<sup>2</sup> nanocomposites has a good linear relationship (0.9945 to 5–100 ppm), high response to NH<sup>3</sup> (5.4 to 100 ppm), and the detection limit is 0.5 ppm. The response and recovery time to 100 ppm NH<sup>3</sup> is 100 and 232 s, respectively. With small humidity effects, the sensor exhibits excellent selectivity, good reproducibility and long-term stability to NH3. Moreover, the excellent gas sensing property of PANI/TiO<sup>2</sup> nanocomposites can be rewarded to p-n heterostructure. The more charge transfer on the surfaces where PANI is in contact with TiO2, the beter gas-sensing performance of PANI/TiO2. These results illustrate that the PANI/TiO<sup>2</sup> nanocomposites sensor has great potential application for detecting NH<sup>3</sup> at room temperature.

### REFERENCES


## AUTHOR CONTRIBUTIONS

CZ performed the experiments, analyzed the data with the help from YX and XC. XD wrote the manuscript with input from all authors. YX conceived the study. All authors read and approved the manuscript.

### ACKNOWLEDGMENTS

This work is financial supported by International Science & Technology Cooperation Program of China (2016YFE0115100), National Natural Science Foundation of China (21771060, 21547012, and 21305033), the Project of Natural Science Foundation of Heilongjiang Province (No. B2015007), and the Project of Nitro Graduate Innovation Research of Heilongjiang University (YJSCX2018-062HLJU). We thank the Key Laboratory of Functional Inorganic Material Chemistry, Ministry of Education, Heilongjiang University for supporting this study.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fchem. 2018.00493/full#supplementary-material


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Zhu, Cheng, Dong and Xu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# 3D Flower-Like NiO Hierarchical Structures Assembled With Size-Controllable 1D Blocking Units: Gas Sensing Performances Towards Acetylene

He Zhang<sup>1</sup> , Wei-Gen Chen<sup>1</sup> \*, Yan-Qiong Li <sup>2</sup> , Ling-Feng Jin<sup>1</sup> , Fang Cui <sup>1</sup> and Zi-Hao Song<sup>1</sup>

*<sup>1</sup> State Key Laboratory of Power Transmission Equipment & System Security and New Technology, School of Electrical Engineering, Chongqing University, Chongqing, China, <sup>2</sup> School of Electronic and Electrical Engineering, Chongqing University of Arts and Sciences, Chongqing, China*

### Edited by:

*Zhenyu Li, Deakin University, Australia*

### Reviewed by:

*Ming-Guo Ma, Beijing Forestry University, China Hua Wang, Beihang University, China Wei Wang, Harbin Institute of Technology, China*

> \*Correspondence: *Wei-Gen Chen weigench@cqu.edu.cn*

### Specialty section:

*This article was submitted to Nanoscience, a section of the journal Frontiers in Chemistry*

Received: *12 May 2018* Accepted: *18 September 2018* Published: *11 October 2018*

### Citation:

*Zhang H, Chen W-G, Li Y-Q, Jin L-F, Cui F and Song Z-H (2018) 3D Flower-Like NiO Hierarchical Structures Assembled With Size-Controllable 1D Blocking Units: Gas Sensing Performances Towards Acetylene. Front. Chem. 6:472. doi: 10.3389/fchem.2018.00472* Acetylene gas (C2H2) is one of the main arc discharge characteristic gases dissolved in power transformer oil. It is of great potential to monitor the fault gas on-line by applying gas sensor technology. In this paper, gas sensors based on nanorods and nanoneedles assembled hierarchical NiO structures have been prepared. Herein, we focus on investigate the relationship between the sizes of the assembling blocking units and gas sensing properties. It can be found that the addition of CTAB/EG plays a vital role in controlling the sizes of blocking unit and assembly manner of 3D hierarchical structures. A comparison study reveals that an enhanced gas sensing performance toward C2H<sup>2</sup> for the sensor based on nanoneedle-assembled NiO flowers occurs over that of nanorod-assembled NiO. This enhancement could be ascribed to the larger specific area of needle-flower, which provides more adsorption and desorption sites for chemical reaction as well as effective diffusion channels for C2H2. Besides, a method of calculating the specific surface area without BET testing was presented to verify the results of gas sensing measurement. The possible growth mechanism and gas sensing mechanism were discussed. Such a synthesis way may open up an avenue to tailor the morphologies and control the sizes of blocking units of some other metal oxides and enhance their gas sensing performances.

Keywords: NiO, hierarchical structures, blocking units, sensor, gas sensing performances

# INTRODUCTION

As we all know, the stable and reliable operation of power transformers is particularly important for the safety and stability of power system. When the oil-immersed power transformers work for a long time, the insulating oil, and paper will gradually deteriorate and produce various trace characteristic gases, which actually dissolve in transformer insulation oil (Singh and Bandyopadhyay, 2010). When the power transformer has the spark of oil or arc discharge fault, it will generate the fault characteristic gas with acetylene gas (C2H2) as the main component. It is of great potential to monitor the fault gas on-line by applying gas sensor technology. The monitoring Zhang et al. NiO Based Gas Sensors

of C2H<sup>2</sup> gas content can predict the internal latent failure of the transformer (Chen et al., 2013; Jin et al., 2017). So, C2H<sup>2</sup> gas is selected as the target gas in this paper.

A common method for detecting C2H<sup>2</sup> gas dissolved in transformer oil is metal oxide semiconductor (MOS) based gas sensor (Zhu and Zeng, 2017). Among various MOS sensing materials, nickel oxide (NiO) as a wide band gap (Eg = 3.6– 4.0 ev) p-type semiconductor has taken a dominated position due to its outstanding physical and chemical properties. Recently, NiO has been extensively applied in multifarious application fields such as electrode materials (Zhang et al., 2004), solar cells (Nakasa et al., 2005), catalysts (Kaminski et al., 2018), and gas sensors (Cao et al., 2015a; Yu et al., 2017).

It's believed that one-dimensional (1D) nanostructures with large surface to volume have great potential to improve the sensing properties. However, there are some shortcomings for 1D structure, i.e., inevitable serious stacking configuration and thermal/ chemical instability. Given this, the sensing materials can be designed into three–dimensional (3D) hierarchical structure assembled by 1D blocking units, which not only prevents the 1D blocking units from serious stacking but also inherits the merits of 1D nanomaterials (Duo et al., 2016; Zhu et al., 2018). The ability to control the assembly configuration, the morphology and size of building units in hierarchical architectures is of utmost importance for the realization of multifunctional nanodevices (Kim and Yong, 2011). Recently, assembly of 1D blocking units into hierarchical structures has been a hot topic in the research. But there are few explorations about the synthesis of hierarchical structures with size-controllable blocking units and studying the influence of the size of the assembling units on gas sensing performances.

In this paper, nanorods, and nanoneedles assembled 3D flower-like NiO hierarchical structures were successfully synthesized via hydrothermal synthesis. Herein, we focus on investigate the relationship between the sizes of the assembling blocking units and gas sensing properties. It can be found that the addition of CTAB/EG plays a vital role in controlling the sizes of blocking unit and assembly manner of 3D hierarchical structures. A comparison study reveals that an enhanced gas sensing performance toward C2H<sup>2</sup> for the sensor based on nanoneedle-assembled NiO flowers occurs over that of nanorodassembled NiO. In order to shed light on this phenomenon, a method of calculating the specific surface area without BET testing was presented to verify the results of gas sensing measurement. Based on our experimental results, the possible formation mechanism of two kinds of NiO nanoflowers is primarily discussed. It's expected that this study can promote the development of gas sensing materials via lower dimensional assembly.

## EXPERIMENTAL

## Synthesis of the Nanorods-Assembled Hierarchical NiO Nanoflowers

In a typical experiment of nanorods assembled NiO nanoflowers, 0.4 g of Ni(NO3)2·6H2O was added to 40 ml of distilled water under vigorous stirring for 10 min. 0.18 g of cetyltrimethyl ammonium bromide (CTAB) was introduced into the above solution. Then, under continuous magnetic stirring, ammonia (NH3·H2O, 25%) was dripped into the mixed solution to obtain the pH = 9. After thorough mixing, the resulting mixture was transferred to a 50 ml autoclave and maintained overnight at 180◦C. After cooling naturally, the precipitates were washed sequentially and dried in air at 60◦C. Finally, the powder was calcined at 500◦C for 2 h. The sample was labeled as rod-flower.

# Synthesis of the Nanoneedles-Assembled Hierarchical NiO Nanoflowers

Typically, 0.4 g of NiCl2·6H2O and 0.08 g of Na2C2O<sup>4</sup> were poured into 15 ml of distilled water. Then 25 ml ethylene glycol (EG) was added into the solution with sequentially stirring. The mixed solution was loaded into a 50 ml autoclave and heated to 160◦C for 12 h. The subsequent process including centrifugation, washing, drying and calcining are the same as the above. The sample after annealing was designated as needle-flower.

## Characterization

Crystal structure of as-prepared samples was examined through X-ray diffraction (XRD, D/Max-1200X, Rigaku). The surface morphologies and nanostructures of the samples were inspected by scanning electronic microscopy (SEM, JEM-6700F) and transmission electron microscopy (TEM, JEM-1200EX).

The detailed process about the fabrication of planar gas sensor and gas-sensing test is as follows Jin et al. (2017). Firstly, the appropriate amount of as-prepared NiO powders was fully ground and mixed with diethanolamine and ethanol to form a slurry suspension. The pastes were evenly coated onto the electrodes of sensor's substrate. Then, the sensor was placed in aging platform and maintained at 120◦C for 100 h to improve the stability of the sensor. Gas sensing properties toward C2H<sup>2</sup> were measured using a CGS-1TP (Chemical Gas Sensor-1 Temperature Pressure) intelligent gas sensing analysis system. The sensor was placed on the heating table of the gas chamber and two probes were adjusted to ensure good electrical signals of the sensing materials. Then, the working temperature was set and air was delivered into the chamber at a constant flow rate. When the resistance of the sensor was stable in the air, it's denoted as Ra. Then, a certain amount of target gas was injected into the chamber through the injection hole. The change of resistance curve in the software was observed until the resistance value was stable again, denoted as Rg. The target gas flow was vented and the sensor was exposed to air again. The concentration of target gas (C2H2) was controlled by the mass flow controllers (MFC) with the following equation (Equation 1):

$$\begin{aligned} & \text{Gas concentration (ppm)}\\ & = \frac{\text{Flow rate (target gas)} \times \text{Gas cylinder (target gas)}}{\text{Flow rate (target gas)} + \text{ Flow rate (air)}} \end{aligned} \tag{1}$$

The response (S) of the sensor was defined as the ratio of Rg to Ra. And the response (recovery) time was regarded as the time required reach 90% of the total resistance change.

### Zhang et al. NiO Based Gas Sensors

### RESULTS AND DISCUSSION

### Morphology and Structure

**Figure 1A** shows the XRD patterns of the obtained samples. The identified peaks in two curves can be well matched with the cubic crystalline structure of NiO (JCPDS Card no. 04-0835) without observable impurity peaks, demonstrating that high purity of NiO. Morphologies and structural features of the samples are characterized by SEM and TEM, as shown in **Figures 1B–E**. From **Figure 1B**, the NiO hierarchical nanoflowers are assembled from a bunch of well-defined nanorods. The roots of these nanorods come together while the tips are detached. **Figure 1C** illustrates that each individual nanorod from the flower shared the same geometric center. The average diameter of these nanorods is ∼900 nm and the length is ∼6µm. Additionally, some rods are scattered around flower-like structures. As observed in the inset of **Figure 1C**, the size of nanorods was similar to what we had observed in SEM images. In **Figure 1D**, the nanoneedles are assembled into homogeneously distributed flower-like structures (AlHadeethi et al., 2017). The magnified SEM image in **Figure 1E** displays the nanoneedles are thicker at roots with sharper emanative ends. Each needle is ∼2.5 um in length and ∼80 nm in diameter at the middle, which is in consistence with the observation in TEM image (the inset of **Figure 1E**).

### Formation Mechanism

Based on the above observations, we proposed a possible formation mechanism for the morphologies evolution of the nanorod-assembled NiO nanoflowers, as shown in **Figure 2A**. Firstly, ammonia aqueous acts as an alkaline reagent to release OH<sup>−</sup> ions. CTAB is a surfactant with a hydrophobic part (Li Y. Q. et al., 2015; Liu et al., 2017). When the Ni(OH)<sup>2</sup> comes across CTAB, Ni(OH)<sup>2</sup> will be preferably absorbed on the CTA<sup>+</sup> heads. Subsequently, the grown Ni(OH)<sup>2</sup> nanoparticles are connected with each other by orientation attachment to form many nanorods. It's proposed that CTAB seemingly acts as an adhesive to gather the nanorods together (Li T. M. et al., 2015; Miao et al., 2017). Finally, the nanorods self-assemble into the ultimate flower-like architectures driven by the minimum surface energy theory.

For the formation of the nanoneedle-assembled hierarchical NiO nanoflowers (**Figure 2B**), firstly, Ni2<sup>+</sup> and C2O 2− 4 can be illustrated as a NiC2O4·2H2O polymer type ribbon owing to the complexation of Ni<sup>+</sup> and C2O 2− 4 . EG is a surfactant with symmetrical structures and functional group-OH, which serves as a ligand to Ni and blocked the crystal surface paralleled to [0,1,1] direction (Cao et al., 2015b). And then the above microstructures are connected with each other along [0,1,1] direction fabricate the needle-like structures. With the reaction time goes by, NiC2O4·2H2O nanoneedles aggregate with each other to assemble into hierarchical needle-flower. Finally, NiO nanoflowers are obtained by thermal calcination.

### Gas Sensing Properties

To further study the connection between the size of the assembling units and gas sensing performances, we conduct gas sensing experiments. Firstly, we investigate gas response curves with temperature changing toward 200 ppm C2H2. In **Figure 2C**, there is a volcano-shaped trend for the changes in gas response of both kinds of nanoflowers. Apparently, the gas response of the needle-flower NiO is higher than that of rod-flower. The responses at peaks are 15.76 and 25.71 at 300◦C, respectively (Long et al., 2018; Zhang et al., 2018). Herein, we determine the

optimum working temperature to be 300◦C for the later testing. **Figure 2D** demonstrates the response and recovery characteristic of hierarchical NiO nanoflowers toward 200 ppm C2H<sup>2</sup> at 300◦C. Both sensors exhibit excellent gas sensing performances. It's clearly seen that the response of the needle-flower NiO is higher than that of rod-flower. Besides, the needle-flower NiO exhibits a shorter response and recovery time (23 and 26 s) compared with that of the rod-flower NiO (34 and 32 s). In addition, a comparison about the sensing performances of NiO sensors in this work and other literature reports is summarized in **Table 1**. It is not difficult to find that the needle-flower NiO based gas sensor in this paper has excellent gas sensing characteristics and has great potential to be a promising candidate for gas-sensitive materials (Lin et al., 2015; Lu et al., 2016; Majhi et al., 2018; San et al., 2018).

Whether the working temperature vs. response or the response and recovery characteristic, sensor based on needleflower NiO absolutely prevails over that of rod-flower. This may be attributed to the needle-flower's high surface area. In order to verify this hypothesis, we use a simple simplified model to calculate the surface area of the hierarchical NiO structures (**Figure 2E**) from associated literature (Lee, 2009; Zhang et al., 2012). We made a slight change according to our data based on the theory. Whether constituent blocking units are nanorods or naononeedles, this proposal has reasonable guiding significance to conduct qualitative analysis. In this modified model, the specific surface area (Equation 2) is

$$\mathcal{S} \cong \frac{(\pi \mathbf{r}^2 + 2\pi \mathbf{r} \mathbf{h})\mathbf{n}}{\mathbf{n}\pi \mathbf{r}^2 \rho} \sim \frac{1}{\rho} (\mathbf{l} + \frac{2\mathbf{h}}{\mathbf{r}}) \tag{2}$$



Where S stands for the specific surface area, r is the equivalent radius of 1D unit, h is the length of 1D unit which can be also expressed as the radius of hierarchical structures, n is the number and ρ is the density of 1D unit. To a specific material, ρ can be considered as a constant. So S is proportional to h/r. Through the measurement and calculation, the h/r value (66.7) of the nanoneedles is ∼5 times that of the nanorods (13.4). So the S of the needleflower is larger. It can explain why the needle-flower NiO shows higher gas response and rapid response/recovery behavior. The larger specific area will provide many adsorption and desorption sites for oxygen, leading to the increasement in the conductivity.

### Gas Sensing Mechanism

The sensing mechanism of NiO-based gas sensors involves three serial reactions: adsorption-oxidation-desorption (Zhu et al., 2017). In the case of p-type semiconductor, its carrier is the hole with positive charge. Specifically, when the sensor is in the air, oxygen molecules react with NiO surface (Equations 3, 4). Due to the above reaction, electrons on the NiO surface combine with O<sup>2</sup> to form oxygen negative ions (O<sup>−</sup> 2 , O−, and O2−). This process cause the decrease of electrons and the increase of holes to form a hole accumulation layer, resulting in the resistance of the sensor decreases correspondingly. When NiO surface comes into contact with C2H<sup>2</sup> gas, oxygen ions will oxidize gas molecules into CO<sup>2</sup> and H2O, and releases electrons to recombine with holes (Equations 5, 6), leading to the decrease of carriers in hole accumulation layer and an increase in the resistance (Balamurugan et al., 2014; San et al., 2015).

$$\text{O}\_2 \text{ (gas)} \rightarrow \text{ O}\_2 \text{ (ads)}\tag{3}$$

$$\text{O}\_2 \text{ (ads)} + \text{ne}^- \rightarrow \text{ O}^{\text{n-}} \text{ (ads)}\tag{4}$$

$$\text{C}\_2\text{H}\_2 \text{ (gas)} \rightarrow \text{ C}\_2\text{H}\_2 \text{ (ads)}\tag{5}$$

C2H<sup>2</sup> (ads) + O <sup>n</sup>−(ads)<sup>−</sup> <sup>→</sup> CO<sup>2</sup> <sup>+</sup> <sup>H</sup>2<sup>O</sup> <sup>+</sup> ne<sup>−</sup> (6)

### CONCLUSION

In summary, nanorods and nanoneedles assembled NiO hierarchical structures have been successfully synthesized via a hydrothermal method and annealing process. Based on the comparative studies, we draw a conclusion that the size (length and diameter) of blocking units has a great influence on gas sensing properties of hierarchical structures. The integral morphologies and sizes of blocking units can be controlled by tuning the additives. Here, CTAB/EG was introduced as a structure-directing agent to regulate the aggregation and assembly. Compared with rod-flower NiO, the needle-flower NiO based sensor exhibits an enhanced gas sensing performance. This enhancement could be ascribed to the larger specific area of needle-flower, which provides more adsorption and desorption sites for chemical reaction as well as abundant effective diffusion channels for C2H2. The results hold a novel point in constructing highly efficient gas sensors. The detection capability of gas sensors determines the effectiveness of transformer on-line monitoring. Therefore, optimize the morphology and structure of gas sensitive materials is very meaningful work. Gas sensors with the advantages of miniaturization structure, high sensitivity, and fast response speed have very high practical value in power system security.

## AUTHOR CONTRIBUTIONS

HZ and W-GC conceived and designed the experiments. HZ, Y-QL, and L-FJ performed the experiments. FC and Z-HS analyzed the data. HZ wrote the manuscript with input from all authors. All authors read and approved the manuscript.

# FUNDING

This work was supported by the National 111 Project of the Ministry of Education of China (No. B08036), the Plan of Innovation Team Construction from Chongqing Universities (No. CXTDX201601001), and the Scientific Research Fund of Chongqing University of Arts and Sciences (2017RDQ38).

### ACKNOWLEDGMENTS

We thank the State Key Laboratory of Power Transmission Equipment and System Security and New Technology, Chongqing University for supporting this study.

## REFERENCES


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Zhang, Chen, Li, Jin, Cui and Song. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Synthesis and Gas-Sensing Property of Highly Self-assembled Tungsten Oxide Nanosheets

Liangbin Hu, Pengfei Hu, Yong Chen\*, Zehui Lin and Changjun Qiu\*

School of Mechanical Engineering, University of South China, Hengyang, China

We report the synthesis of tungsten oxide (WO3) nanosheets using a simple yet efficient hydrothermal technique free of surfactantand template. The WO<sup>3</sup> nano-sheets are self-assembled as well to form ordered one-dimensional chain nanostructure. A comprehensive microscopic characterization reveals that the nano-sheets have triangular and circular two different shape edges, dislocation and stacking faults are also observed, which should have implications for our understanding of catalytic activity of ceria. We also propose a growth mechanism for the nano-sheets. As a result of this unique morphology, this WO<sup>3</sup> nano-sheets are found to show excellent gas-sensing properties which can use as promising sensor materials detecting ethanol with low concentration.

### Edited by:

Zhongchang Wang, Laboratório Ibérico Internacional de Nanotecnologia (INL), Portugal

### Reviewed by:

Jun Wang, Shenzhen University, China Xin Wang, Shenzhen University, China

### \*Correspondence:

Yong Chen chenyongjsnt@163.com Changjun Qiu qiuchangjun@hotmail.com

### Specialty section:

This article was submitted to Nanoscience, a section of the journal Frontiers in Chemistry

Received: 02 June 2018 Accepted: 12 September 2018 Published: 05 October 2018

### Citation:

Hu L, Hu P, Chen Y, Lin Z and Qiu C (2018) Synthesis and Gas-Sensing Property of Highly Self-assembled Tungsten Oxide Nanosheets. Front. Chem. 6:452. doi: 10.3389/fchem.2018.00452 Keywords: crystal structure, defects, nanosheets, highly self-assembled, gas-sensitivity

# INTRODUCTION

Tungsten trioxide WO<sup>3</sup> nanomaterials are extensively applied in electrochromic device, gas sensor and photocatalysts filed (Hai et al., 2016; Zhang et al., 2018). Great effort has been devoted to the control the specific size and shape of WO<sup>3</sup> nanoparticles which can significantly impact properties of materials. Up to now, the WO<sup>3</sup> nanoparticles with avariety of morphologies (such as nanowires, nano-rods, nano-plate, nano-spheres and so on) have been syhthesized successfully via highintensity ultrasound, rapid microwave, hydrothermal synthesis and other methods (Chai et al., 2016, Hu et al., 2017a, Xu et al., 2017; Chen et al., 2018; Parthibavarman et al., 2018; Zhan et al., 2018). Among all of those applied methods, hydrothermal synthesis technique has been concerned due to the merits of simple operation, low energy consumption, the possibility for large-scale industrialization and so on.

Here,we report the WO<sup>3</sup> nano-sheets with highly self-assembled architecture synthesized via the hydrothermal synthesis process. Defects are observed in the nano-sheets, which may play a key role in affecting properties of WO<sup>3</sup> nano-sheets. In addtion,we also test the response and selectivity of the sensor fabricated from the WO<sup>3</sup> nano-sheets.

## MATERIALS AND METHODS

All the reagents were analytical grade and without further purification. We adopted a facile hydrothermal method to synthesize the nanostructures. First of all, 32 ml of deionized water and 8 ml glycerol (C3H8O3) were mixed into a mixture, then 1.6 nmol of sodium tungstate dihydrate (Na2WO4·2H2O) and 3 nmol oleic acid (C18H34O2) were dispersed into the mixture, and stirred for 15 min with a magnetic stirrer. Secondly, the pH of the mixture was adjusted to 1.25 by HCl. After stirring for 15 min, the solution was transferred into the Teflonlined stainless steel autoclave and treated at 120◦C during 12 h under autogenously pressure.

Finally, the obtained particles were washed by deionized and alcohol to remove the unexpected ions by high-speed centrifugation and then dried at 60◦C for 10 h in air.

The characterization of speciemen as our previous work (Chen et al., 2013, 2014; Hu et al., 2017b). The process of measuring the gas sensitivity of the prepared nanomaterials is described in the literature (Guo and Wang, 2016). Response of the sensors was defined as the ratio of Ra (resistances in air) to Rg (resistances in target gases).

### RESULTS

**Figure 1a** shows a typical XRD spectrum of the products, the diffraction peaks match well with those of a standard WO3·H2O with orthorhombic structure (JCPDS No. 84-0886). The WO3·H2O nano-particles exhibit rectangle shape with an average size of ∼400 nm and thickness of 30nm (**Figure 1b**). In addtion, one-dimensional chain nanostructure is self-assembled by the quadrilateral faces on the both sides of the nano-sheets. The face of nano-sheet are flat (**Figures 1b,c**). The corresponding SADP identifies that the structure of WO<sup>3</sup> nano-sheets is orthorhombic (**Figures 1d**). **Figure 1e** shows a high-resolution TEM (HRTEM) image taken around the corner of a nanosheet. We determined the lattice spacing of the perpendicular lattice fringes to be ∼0.280 and ∼0.267 nm which are belong to the WO<sup>3</sup> (010) and (120) planes respectively. **Figure 1g** shows a HRTEM image taken from the corner area of the nanosheet **Figure 1f** from which lattice spacing is determined to be ∼0.280 nm, in line with the distance between {010} planes of WO3. In addtion, the edges of the nano-sheets are in an arc shape (insert of **Figure 1f**). Two different direction edge dislocations: positive edge dislocation (**Figure 1h**) and negative edge dislocation (**Figure 1i**) are deteced by the HR-TEM, which should be critical for the properties of WO3. Apart from the defects on the surface, we also observed the stacking faults in nano-sheet (**Figure 1j**), which should impact the properties of WO<sup>3</sup> nanoparticles as well. The stacking fault was observed obviously by the one-dimensionally filtered HR-TEM images of the WO<sup>3</sup> nano-sheet (**Figure 1k**). Those dislocations and stacking faults may affect the catalytic activity or other properties of nano material (Wang et al., 2011, 2014, 2016; Sun et al., 2015).

**Figure 2** shows the gases (NH3, CH3OH, C6H6) response of the sensor based on WO<sup>3</sup> nanosheets. All of the gases were tested at an operating temperature of 300◦C with a concentration of 30 ppm. In **Figure 2**, the results indicate that the sensor exhibited little responses to NH3, C6H6, to indicated that it was insensitive to NH3, C6H6. For ethanol, the highest response of the sensor was 25.6, while the responses to NH<sup>3</sup> and C6H<sup>6</sup> were no >1.

### DISCUSSION

In light of the aforementioned microstructural characterization, we propose a likely growth mechanism for the nanosheet. First, the Na2WO4·2H2O is ionized to WO<sup>−</sup> 4 . Then, the WO<sup>−</sup> 4 ion react with H<sup>+</sup> which ionized by HCl, forming the H2WO<sup>4</sup> suspension. The H2WO<sup>4</sup> suspension decomposes, at the high temperature and pressure during hydrothermal process, resulting in to the nucleation of WO3. The oleic acid acts as a soft template and controls the growth rate of different crystal plane owing to its selective absorption and desorption behavior. Then most of WO<sup>3</sup> nano-sheets with {010} exposure planes are selfassembled, forming one-dimensional chain nanostructures, due to the addition of oleic acid. The formation process can be described as follows:

$$\text{Na}\_2\text{WO}\_4\cdot2\text{H}\_2\text{O} \rightarrow 2\text{Na}^+ + \text{WO}\_4^- + 2\text{H}\_2\text{O} \tag{1}$$

$$\text{HCl} \rightarrow \text{H}^+ + \text{Cl}^- \tag{2}$$

$$\mathrm{^{2}H^{+}} + \mathrm{WO\_{4}^{-}} \rightarrow \mathrm{H\_{2}WO\_{4}} \tag{3}$$

$$\text{H}\_2\text{WO}\_4 \rightarrow \text{WO}\_3 + \text{H}\_2\tag{4}$$

We imply that the WO<sup>3</sup> nano-sheets can act as an efficient gas-sensing material for selective detection of ethanol. Such the sensing performance due to the fact that the diffusion of ethanol and its oxidation with O<sup>−</sup> or O2<sup>−</sup> are very rapidly in nanoplates structures (Xiao et al., 2017). When sensor prepared by the WO<sup>3</sup> nanosheets is exposed in air, the resistance of the WO<sup>3</sup> nanosheets is increased by oxygen molecules which adsorbed on the surfaces of the WO<sup>3</sup> nanosheets, trapping electrons in the conduction band and forming oxygen species (O−, O2−). As ethanol is introduced to the sensor, the oxygen species (O−, O2−) react with ethanol molecules on the surface of the WO<sup>3</sup> nanosheets, which will release trapped electrons to conduction

### REFERENCES

Ahmad, M. Z., Sadek, A. Z., Ou, J. Z., Yaacob, M. H., Latham, K., and Wlodarski, W. (2013). Facile synthesis of nanostructured WO<sup>3</sup> thin films and their characterization for ethanol sensing. Mater. Chem. Phys. 141, 912–919. doi: 10.1016/j.matchemphys.2013. 06.022

band and resistance of the WO<sup>3</sup> nanosheets is decreased (Ahmad et al., 2013).

### CONCLUSIONS

We have adopted the hydrothermal technique to synthesize highly self-assembled WO<sup>3</sup> nano-sheets using the tungsten resource Na2WO4·2H2O and the soft template oleic acid. We demonstrated that the WO<sup>3</sup> nano-sheets are mainly exposed with {010} planes and crystal defects such as edge dislocations and stacking faults exist in single crystalline nano WO<sup>3</sup> by microscopic investigations, which may be important for the catalytic activity of WO3. We indicate that the WO<sup>3</sup> nano-sheets could be used as promising sensor material for detecting CH3OH with low concentration.

### AUTHOR CONTRIBUTIONS

LH synthesized and characterized the microstructure of highly self-assembled tungsten oxide nanosheets and wrote the manuscript. PH and ZL tested the response and selectivity of the sensor fabricated from the WO<sup>3</sup> nano-sheets. YC and CQ directed the experiment. All authors read and approved the manuscript.

### ACKNOWLEDGMENTS

YC thanks the financial support by the Construct Program of the Key Discipline of Hunan province (Education Department of Hunan Province Bulletin grant no.[2014]85). ZL thanks the financial support by the Project of Research Study and Creativity Experiment Plan for College Students of the Hunan province (Education Department of Hunan Province Bulletin grant no.[2016]283).

Chai, Y., Ha, F. Y., Yam, F. K., and Hassan, Z. (2016). Fabrication of tungsten oxide nanostructure by sol-gel method. Proc. Chem. 19, 113–118. doi: 10.1016/j.proche.2016.03.123

Chen, G. C., Chua, X. F., Qiao, H. B., Ye, M. F., Chen, J., Gao, C., et al. (2018). Thickness controllable single-crystal WO<sup>3</sup> nanosheets: highly selective sensor for triethylamine detection at room temperature. Mater. Lett. 226, 59–62. doi: 10.1016/j.matlet.2018.05.022


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Hu, Hu, Chen, Lin and Qiu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Functionalization of α-In2Se<sup>3</sup> Monolayer via Adsorption of Small Molecule for Gas Sensing

Zhi Xie<sup>1</sup> \*, Fugui Yang<sup>2</sup> , Xuee Xu<sup>1</sup> , Rui Lin<sup>1</sup> and Limin Chen<sup>1</sup>

<sup>1</sup> College of Mechanical and Electronic Engineering, Fujian Agriculture and Forestry University, Fuzhou, China, <sup>2</sup> School of Electronic Information Science, Fujian Jiangxia University, Fuzhou, China

Based on first-principles calculations, the adsorption of NO and NO<sup>2</sup> gas molecules on the α-In2Se<sup>3</sup> monolayer have been studied. The adsorption configuration, adsorption energy, electronic structure and charge transfer properties are investigated. It is found that the charge transfer processes of NO and NO<sup>2</sup> adsorbed on the surface of α-In2Se<sup>3</sup> monolayer exhibit electron donor and acceptor characteristics, respectively. After the adsorption of the molecules, the α-In2Se<sup>3</sup> monolayers have new states near the Fermi level induced by NO and NO2, which can trigger some new effects on the conducting and optical properties of the materials, with potential benefits to gas selectivity. The present work provides new valuable results and theoretical foundation for potential applications of the In2Se3-based gas sensor.

### Edited by:

Zhongchang Wang, Laboratório Ibérico Internacional de Nanotecnologia (INL), Portugal

### Reviewed by:

Shuping Huang, Fuzhou University, China Youzhao Lan, Zhejiang Normal University, China

> \*Correspondence: Zhi Xie xzfjau@126.com

### Specialty section:

This article was submitted to Nanoscience, a section of the journal Frontiers in Chemistry

Received: 17 July 2018 Accepted: 30 August 2018 Published: 26 September 2018

### Citation:

Xie Z, Yang F, Xu X, Lin R and Chen L (2018) Functionalization of α-In2Se3 Monolayer via Adsorption of Small Molecule for Gas Sensing. Front. Chem. 6:430. doi: 10.3389/fchem.2018.00430

Keywords: 2D materials, first-principles calculation, In2Se3 , charge transfer, gas sensor

# INTRODUCTION

In recent years, Layered two-dimensional (2D) materials have received tremendous research attention due to their unique physical and chemical properties (Miró et al., 2014; Bhimanapati et al., 2015; Xie et al., 2018). Because of their ultrahigh flexibility, strength and thickness-dependent electronic properties (Wang et al., 2012; Novoselov et al., 2016), the nanodevices based on 2D materials and tuning the properties of their heterostructures via defects engineering (Cervenka et al., 2009; Wang et al., 2011; Park et al., 2014; Sun et al., 2015) hold great promise for potential applications in nanoscale electronics, optoelectronics and spintronics (Wang et al., 2008, 2016; Geim and Grigorieva, 2013; Lan, 2018). Additionally, the high surface/volume ratio, weak electronic screening and ultrathin thickness of 2D materials induce that their structural stability and electronic properties are very sensitive to environmental molecules, and the relevant effects make them efficient for gas molecules sensing, catalysis, and energy storage technologies (Lightcap and Kamat, 2013; Yang et al., 2016; Zhang et al., 2018). Graphene has exhibited good performance in the field of gas sensor (Kemp et al., 2013). Previous reports have also shown that MoS2 based nanosensors possess excellent sensing ability with high response value, and their molecule adsorption properties can be modulated by applying light, strain, and external electric field (Late et al., 2013; Ma et al., 2016). Recently, InSe monolayer has been found having tunable electronic properties via the molecule adsorption and promising for gas sensing application (Ma et al., 2017). All these studies clearly reveal that external factors can modulate the properties of 2D materials effectively and extend their application fields.

**42**

Indium selenide (In2Se3) is an interesting III-VI group layered chalcogenide compound with multiple phases and excellent properties (Shi et al., 2013), and have attracted extensive research interest for the applications in phase change memory (Yu et al., 2007), lithium batteries (Feng et al., 2016), optoelectronic and photovoltaic devices (Zhai et al., 2010; Jacobs-Gedrim et al., 2014). Among all the phases, 2D materials based on α-In2Se<sup>3</sup> exhibit obvious thickness-dependent shift of band gap and promising prospects for tunable wavelength photodetection (Quereda et al., 2016). It is also reported that the strain sensor fabricated from 2D α-In2Se<sup>3</sup> films possesses good stability, excellent sensitivity, and high spatial resolution in strain distribution, showing attractive properties for e-skin applications in wearable electronics (Feng et al., 2016). However, to best of our knowledge, the investigations on the adsorption of small gas molecules on atomically thin 2D In2Se<sup>3</sup> materials and the related modification of their properties are still lacking so far. It is well-known that NO<sup>2</sup> and NO are common air pollutants and harmful to human health. The detection and control of them are very important for the environmental protection. Hence, in this work, we have made first-principles studies on the α-In2Se<sup>3</sup> monolayers adsorbed by NO and NO2, respectively. The adsorption configuration, structural stability, electronic structure and charge transfer properties have been investigated and discussed in detail.

### COMPUTATIONAL METHODS

All calculations are carried out using the Vienna ab initio simulation package (VASP)(Kresse and Furthmüller, 1996), with the core electrons described by the projected augmented wave (PAW) method. For the exchange-correlation term, the generalized gradient approximation (GGA) with Perdew-Burke-Ernzerhof (PBE) scheme is employed. The cutoff energy for plane-wave basis is set as 450 eV. For simulating the adsorption of the molecules, a 4 × 4 × 1 supercell of the α-In2Se<sup>3</sup> monolayer is built with one NO or NO<sup>2</sup> molecule adsorbed on its surface, and a vacuum space of more than 15 Å is set up to prevent the interactions between the repeated monolayers. The Monkhorst-Pack of 2 × 2 × 1 (4 × 4 × 2) k-point grid is adopted for the Brillion zone sampling in geometry optimization (total energy calculation). The convergence criterion of energy is taken as 10−<sup>5</sup> eV. Structure relaxation is performed until the force on each atom is smaller than 0.02 eV/Å. To estimate the adsorption stability of gas molecules on the surface of α-In2Se<sup>3</sup> monolayer, the adsorption energy (Ead) is calculated by the formula: Ead = E<sup>M</sup> + E<sup>G</sup> – EM+G, where EM, EG, and EM+<sup>G</sup> denote the total energy of the α-In2Se<sup>3</sup> monolayer, the free gas molecule, and the α-In2Se<sup>3</sup> monolayer adsorbed by gas molecules, respectively. According to this definition, a positive value of Ead represents the adsorption is energetically favorable.

### RESULTS AND DISCUSSION

Firstly, the geometry optimizations of free gas molecules were performed. The obtained bond lengths of NO and NO<sup>2</sup> are 1.17 and 1.21 Å, respectively, and the O-N-O bond angle of NO<sup>2</sup> is 133.39◦ . The band gap of the clean α-In2Se<sup>3</sup> monolayer has been calculated to be 0.77 eV (see **Figure 3A**). All these results are in line with the data of previous reports (Debbichi et al., 2015; Ma et al., 2017). In order to find the most stable adsorption configuration, four typical adsorbing sites on the Se atom plane of one side have been considered including the top of Se atom, the center of a Se-In bridge and two centers of the hexagonal void (see **Figure 1A**). Because of the different coordination structures of the Se atom plane on the other side, four similar adsorbing sites were also investigated on the other side (see **Figure 1B**).

For the adsorption of NO molecule, besides the eight adsorbing sites mentioned above, we also considered two different orientations of the molecule with the N-O bond perpendicular or parallel to the surface of α-In2Se<sup>3</sup> monolayer. Therefore, 16 configurations have been examined. **Figure 2A** presents the top and side views of the most stable configuration obtained, where the O atom of NO molecule points away from the α-In2Se<sup>3</sup> surface and the N atom of NO molecule points toward the surface with the smallest distance between the adsorbed NO and the surface atom is 2.65 Å. The N-O bond is a little shortened to 1.16 Å compared with that (1.17 Å) of free NO molecule. The adsorption energy was calculated

to be 208 meV, which is comparable to those of NO adsorptions on the monolayers of InSe, GaSe, and MoS<sup>2</sup> (Yue et al., 2013; Ma et al., 2017; Zhou et al., 2017). This low adsorption energy indicates the NO adsorption capability of α-In2Se<sup>3</sup> monolayer is not very strong, which is applicable for the gas detection since the adsorption-desorption of NO molecule on α-In2Se<sup>3</sup> monolayer can be easily achieved.

For further investigating the interactions and electron transfers between the adsorbed NO molecule and the α-In2Se<sup>3</sup> monolayer, the charge density difference (CDD) is calculated from the formula: 1ρ = ρM+<sup>G</sup> − ρ<sup>M</sup> − ρG, where ρM+G, ρ<sup>M</sup> and ρ<sup>G</sup> represent the total charge densities of the α-In2Se<sup>3</sup> monolayer adsorbed by gas molecules, the α-In2Se<sup>3</sup> monolayer, and the isolated gas molecule, respectively. The ρ<sup>M</sup> and ρ<sup>G</sup> are obtained with each atom at the same position as the adsorption configuration. In the NO<sup>2</sup> adsorption case, the similar calculation has also been performed. As shown in **Figure 2B**, it can be seen that the adsorption make the redistribution of charges around the NO molecule. In the space between the adsorbed NO and the α-In2Se<sup>3</sup> surface, the depletion of electrons is dominant. Based on the Bader charge analysis, the charge transfer has been quantitatively calculated. It is demonstrated that the NO

molecule provides 0.054 e electrons to the α-In2Se<sup>3</sup> surface and acts as an electron donor. This behavior is different from the situation of the NO adsorption on InSe monolayer, where the NO molecule acts as an electron acceptor with the amount (0.018 e) of transferred charges (Ma et al., 2017) smaller than that (0.054 e) between NO and α-In2Se<sup>3</sup> monolayer. The band structure of the most stable configuration is depicted in **Figure 3A**, it is shown that after the NO adsorption the Fermi level (E<sup>f</sup> ) of the system moves upwards to the bottom of the conduction bands compared with that of the clean α-In2Se<sup>3</sup> monolayer (see **Figure 3A**), demonstrating an n-type conducting property of the materials with NO adsorption, which is similar to the situation of property modification in the NO-adsorbed MoS<sup>2</sup> monolayer (Shokri and Salami, 2016). This property changes can be useful to the detection of NO molecule. In addition, some new states are found to be located at the E<sup>f</sup> . To better understand the adsorption effect of NO molecule on the α-In2Se<sup>3</sup> monolayer, the local density of states (DOS) of the adsorbed NO and its nearest Se atom are illustrated in **Figure 3B**. It is clearly shown the new states at the E<sup>f</sup> are from the adsorbed NO, and there is little hybridization between the states of NO molecule and the states of the surface Se atom near it, which is similar to the NO adsorption behavior on InSe monolayer, further confirming that the interaction between the adsorbed NO molecule and the α-In2Se<sup>3</sup> monolayer is not strong.

In the NO<sup>2</sup> adsorption case, two orientations of NO<sup>2</sup> molecule have been considered. One is the two O atoms point toward the α-In2Se<sup>3</sup> surface, and the other is that they point away from the α-In2Se<sup>3</sup> surface. The obtained most stable adsorption configuration is displayed in **Figure 2C**, in which the NO<sup>2</sup> is adsorbed on the Se atom plane of the other side different from that of the NO adsorption case and the two O atoms of NO<sup>2</sup> molecule point toward the α-In2Se<sup>3</sup> surface. The smallest distance between the NO<sup>2</sup> molecule and the α-In2Se<sup>3</sup> surface is 3.57 Å, and the N-O bond is a little elongated to 1.22 Å with the O-N-O angle reduced to 131.89◦ . The adsorption energy was calculated to be 59 meV, suggesting the adsorption of NO<sup>2</sup> on α-In2Se<sup>3</sup> monolayer is weak, which is comparable to that of the NO<sup>2</sup> adsorption on graphene (Leenaerts et al., 2008).

The CDD of the most stable configuration for NO<sup>2</sup> adsorption is displayed in **Figure 2D**. It is shown that the charge redistribution of NO<sup>2</sup> molecule is apparent. The electrons accumulate in the vicinity of the adsorbed NO<sup>2</sup> with a little electron depletion in its core region. The depletion of electrons mainly occurs for the Se atoms around the NO<sup>2</sup> molecule. From the Bader analysis, it is indicated that the NO<sup>2</sup> molecule is an electron acceptor and obtains 0.081 e electrons from the α-In2Se<sup>3</sup> monolayer, which is similar to the situation of NO<sup>2</sup> adsorption on InSe monolayer, and their amounts of transferred charges are comparable (Ma et al., 2017). The band structure of the discussed configuration is shown in **Figure 3A**. It can be seen that there is a new impurity band lying just above the top of the valence bands and the E<sup>f</sup> , which can modify the optical and conducting properties of the materials, benefiting the detection of NO<sup>2</sup> molecule. As shown in **Figure 3B**, the local DOS distributions demonstrate that the impurity states just above the E<sup>f</sup> are from the adsorbed NO<sup>2</sup> molecule, and there are some overlaps of states between the NO<sup>2</sup> molecule and the Se atoms near it.

### CONCLUSION

To explore the gas sensing applications of 2D materials based on In2Se3, the effects of the adsorbed NO and NO<sup>2</sup> molecules on α-In2Se<sup>3</sup> monolayer have been studied using first-principles calculations. When the NO and NO<sup>2</sup> are adsorbed on the surface of the α-In2Se<sup>3</sup> monolayer, the calculated adsorption

### REFERENCES


energies of positive value indicate their adsorption processes are exothermic and energetically favorable. Their low adsorption energies demonstrate the α-In2Se<sup>3</sup> monolayer is applicable for the gas molecules detection. In the most stable configurations, the gas molecules are adsorbed on different Se atom planes for NO and NO2, respectively, and the smallest distance (3.57 Å) between the adsorbed NO<sup>2</sup> and the α-In2Se<sup>3</sup> monolayer is larger than that (2.65 Å) of NO adsorption case. NO provides 0.054 e electrons to the α-In2Se<sup>3</sup> monolayer as the donor gas molecule, while NO<sup>2</sup> acts as the acceptor gas molecule and gains 0.081 e electrons from the α-In2Se<sup>3</sup> monolayer. Both of the adsorbed molecules induce new electronic states near the Fermi level compared with the electronic structure of clean α-In2Se<sup>3</sup> monolayer. These changes of electronic properties can modify the conducting and optical properties of the materials and benefit gas sensing. The theoretical findings of this work suggest the 2D α-In2Se<sup>3</sup> materials hold great promise for the application of gas sensor.

### AUTHOR CONTRIBUTIONS

ZX and FY performed the calculations and analyzed the data with the help of XX, RL, and LC. ZX and FY wrote the manuscript with input from all authors. All authors read and approved the manuscript.

### FUNDING

This study is financially supported by the Natural Science Foundation of Fujian Province (2018J01587) and the Scientific Research Development Funds (107/KF2015096) of Fujian Agriculture and Forestry University.


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Xie, Yang, Xu, Lin and Chen. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Optical Temperature Sensing With Infrared Excited Upconversion Nanoparticles

### Kory Green<sup>1</sup> , Kai Huang<sup>2</sup> , Hai Pan<sup>1</sup> , Gang Han<sup>2</sup> and Shuang Fang Lim<sup>1</sup> \*

*<sup>1</sup> Department of Physics, North Carolina State University, Raleigh, NC, United States, <sup>2</sup> Department of Biochemistry and Molecular Pharmacology, University of Massachusetts Medical School, Worcester, MA, United States*

Upconversion Nanoparticles (UCNPs) enable direct measurement of the local temperature with high temporal and thermal resolution and sensitivity. Current studies focusing on small animals and cellular systems, based on continuous wave (CW) infrared excitation sources, typically lead to localized thermal heating. However, the effects of upconversion bioimaging at the molecular scale, where higher infrared intensities under a tightly focused excitation beam, coupled with pulsed excitation to provide higher peak powers, is not well understood. We report on the feasibility of 800 and 980 nm excited UCNPs in thermal sensing under pulsed excitation. The UCNPs report temperature ratiometrically with sensitivities in the 1 × 10−<sup>4</sup> K −1 range under both excitation wavelengths. Our optical measurements show a ln(I525/I545) vs. 1/T dependence for both 800 nm and 980 nm excitations. Despite widespread evidence promoting the benefits of 800 nm over 980 nm CW excitation in avoiding thermal heating in biological imaging, in contrary, we find that given the pulsed laser intensities appropriate for single particle imaging, at both 800 and 980 nm, that there is no significant local heating in air and in water. Finally, in order to confirm the applicability of infrared imaging at excitation intensities compatible with single nanoparticle tracking, DNA tightropes were exposed to pulsed infrared excitations at 800 and 980 nm. Our results show no appreciable change in the viability of DNA over time when exposed to either wavelengths. Our studies provide evidence for the feasibility of exploring protein-DNA interactions at the single molecule scale, using UCNPs as a reporter.

Keywords: upconversion thermal sensing, pulsed excitation, 800 nm, 980 nm, local thermal heating, DNA denaturation

# INTRODUCTION

Many biological processes occurring within intracellular structures may result in changes in the pH, temperature and electrical potential, to name a few. Exploring thermal changes at the cellular level provides insight into biochemical reactions taking place in a cell. Fluorescent thermometry relies on changes of relative fluorescent intensities, lifetimes and wavelengths to local temperature (Engeser et al., 1999; Sakakibara and Adrian, 1999; Ross et al., 2001; Wang et al., 2002; Löw et al., 2008; Binnemans, 2009; Vetrone et al., 2010a). Conventional fluorescence microscopy uses short-wavelength (UV-blue) excitation, and detection of a longer-wavelength, Stokes-shifted fluorescence (Stokes, 1852; Lichtman and Conchello, 2005). However, this use of short-wavelength excitation leads to autofluorescence, photobleaching, and photodamage to biological specimens

Edited by: *Wen Zeng, Chongqing University, China*

### Reviewed by:

*Ruoxue Yan, University of California, Riverside, United States Daqin Chen, Hangzhou Dianzi University, China*

> \*Correspondence: *Shuang Fang Lim sflim@ncsu.edu*

### Specialty section:

*This article was submitted to Nanoscience, a section of the journal Frontiers in Chemistry*

Received: *04 June 2018* Accepted: *23 August 2018* Published: *24 September 2018*

### Citation:

*Green K, Huang K, Pan H, Han G and Lim SF (2018) Optical Temperature Sensing With Infrared Excited Upconversion Nanoparticles. Front. Chem. 6:416. doi: 10.3389/fchem.2018.00416*

**48**

(Giloh and Sedat, 1982). Molecular dyes bleach under intense illumination (Shaner et al., 2008). Semiconductor nanoparticles (i.e., Qdots) (Tessmer et al., 2013) are stable, but display blinking and toxic behavior (Yao et al., 2005). In contrast, our method is based on rare-earth ion doped, upconversion nanoparticles (UCNPs) (Lim et al., 2006, 2009, 2010; Austin and Lim, 2008; Ungun et al., 2009). UCNPs absorb at 800 and 980 nm in the near infrared (NIR), exhibit no bleaching, are non-toxic, and are not affected by blinking (Chen et al., 2006; Schubert et al., 2006). Due to this lack of bleaching, UCNPs are well suited for long term monitoring of biological events at the high laser intensity levels employed in single cell imaging, as opposed to a dye that may bleach over time. The demonstration of UCNPs as nanothermometers in water has been shown by Vetrone et al. (2010a) and others (Chen et al., 2016; Zhu et al., 2016). The electrons in the 4f shell of rare earths are shielded from the surroundings by the filled 5s and 5p shells, and therefore the influence of the surrounding matrix on the optical transitions within the 4f shell is small, whether in crystals or in solution. Therefore, UCNPs show reduced sensitivity to physiological changes such as salt concentration (Gota et al., 2009) and pH while monitoring cellular temperatures (Vetrone et al., 2010a,b). UCNPs have also been used to measure the temperature of the interior nanoenvironment of magnetically heated iron oxide nanoparticles (Dong and Zink, 2014) and have been shown to enable direct measurement of the local temperature with high temporal (millisecond) and thermal resolution (0.3–2.0 K) (Debasu et al., 2013) and (10−<sup>5</sup> K −1 ) sensitivity (Xu et al., 2012) with simple equipment requirements. The emission of the dopant ions is sensitive to temperature in some configurations due to closely spaced energy levels being thermally coupled (Bai et al., 2007; Lü et al., 2011; Xu et al., 2012; Debasu et al., 2013). Moreover, these thermally coupled energy levels are not sensitive to other environmental factors such as scattering or tissue autofluorescence. Thermally coupled emissions, such as for the Er3<sup>+</sup> rare earth ion, can be in the visible, such as in the intensity ratio (RHS) of the <sup>2</sup>H11/<sup>2</sup> to <sup>4</sup> I15/<sup>2</sup> (525 nm) over <sup>4</sup> S3/<sup>2</sup> to <sup>4</sup> I15/<sup>2</sup> (545 nm) transitions (Bai et al., 2007; Lü et al., 2011; **Figure 1**), or in the red to near infrared, with the Tm3<sup>+</sup> rare earth ion, such as in the intensity ratio (RHS) of the <sup>3</sup>F2,3 to <sup>3</sup>H<sup>6</sup> (700 nm) over the <sup>3</sup>H<sup>4</sup> to <sup>3</sup>H<sup>6</sup> (800 nm) transitions (Xu et al., 2012). In the aforementioned emissions, the energy separation between the nearest excited states Er: <sup>2</sup>H11/<sup>2</sup> and Er: <sup>4</sup> S3/2, is only several hundred wavenumbers. Thus, the population distribution of Er: <sup>2</sup>H11/<sup>2</sup> and Er: <sup>4</sup> S3/<sup>2</sup> is influenced by both thermal distribution and nonradiative relaxation. As a consequence, the population of the Er: <sup>2</sup>H11/<sup>2</sup> level varies as a function of the Boltzmann's distribution between the two states (Lei et al., 2005). Similarly, the small energy separation between the thermally coupled Tm3<sup>+</sup> levels of about 1,850 cm−<sup>1</sup> , gives rise to the same phenomenon (Xu et al., 2012). Measurements of the Boltzmann distribution between the two closely spaced states with varying temperatures show that the natural log of this ratio is inversely proportional to the temperature in the range relevant to most biological systems (Vetrone et al., 2010a).

However, primary excitation of the UCNPs occurs in the near infrared, where the absorption coefficient of the water abundant in biological tissue varies, resulting in some reservations regarding the use of these nanoparticles as a temperature sensor. Those reservations arise from the fact that the absorption coefficient of water at 980 nm is about 20 times larger than that at 800 nm (Weber, 1971; Wang et al., 2013). Specifically, at 980 nm CW excitation, thermal heating of the biological environment, may hamper the measurement process, as seen in small animals, and in cellular systems (Vetrone et al., 2010a; Wang et al., 2013). The more widely used sensitizer, the Yb3<sup>+</sup> rare earth ion absorbs primarily at 980 nm, corresponding to the <sup>2</sup>F7/<sup>2</sup> to <sup>2</sup>F5/<sup>2</sup> transition, whereas the Nd3<sup>+</sup> rare earth ion sensitizer absorbs at 800 nm, corresponding to the <sup>4</sup> I9/<sup>2</sup> to <sup>4</sup>F5/<sup>2</sup> transition. The Nd3<sup>+</sup> ion has an absorption cross section an order of magnitude greater at 800 nm than the Yb3<sup>+</sup> sensitizer(Kushida et al., 1968; Wang et al., 2013; Chen et al., 2015). Most studies focus on small animals and cellular systems, where the infrared excitation source is a continuous wave (CW) laser operating at low intensities. This has led to the prevailing belief that the benefits of bioimaging with CW infrared excitation at 980 nm, with low scattering background, is offset by the thermal cost of cellular heating. At this scale, 800 nm CW excitation has been shown to alleviate some of the issues with heating(Wang et al., 2013). However, given the emergent application of infrared imaging at the single molecule level, efforts to examine the thermal effects of pulsed excitation at 800 and 980 nm have never been made. Infrared upconversion bioimaging at the molecular scale, occurs at a much higher intensity under a tightly focused excitation beam, and is normally coupled with pulsed excitation to provide higher peak powers, for sharper discrimination along the Z axis. Therefore, in this work, particular emphasis is paid to the potential thermal effects of upconversion bioimaging at the molecular scale, under pulsed infrared excitation. An example of single upconversion nanoparticle imaging is shown in **Supplementary Video 1**. Furthermore, we address the proposed potential benefits of pulsed excitation infrared imaging at the single molecule scale, first through spectroscopic studies, and then in single molecule viability studies using DNA tightrope, under pulsed laser excitation at high magnification and intensities. A long-term single molecule imaging technology that is resistant to photobleaching and is excited at longer wavelengths can be a powerful tool to study physiological processes that take time to unfold, such as disease progression.

To that end, comparisons are made between excitations at 980 nm (β-NaYF4:20%Yb, 2%Er UCNPs) and 800 nm (β-NaYF4:40%Yb, 2%Er@NaYF4:20%Yb@NaNdF4:10%Yb coreshell-shell UCNPs), where the thermal response of the UCNP is recorded as a thermally responsive intensity ratio variation in the spectra, with no fluorescence intensity quenching, and with simple equipment requirements. In contrast to other researchers, where both I<sup>525</sup> and I<sup>545</sup> emission lines are derived from a spectral scan over the appropriate wavelength range, we collect both lines simultaneously, in order to avoid heat dissipation effects. Our optical measurements show a ln(I525/I545) vs. 1/T dependence for both 800 nm and 980 nm excitations, in the Yb3+/Er3<sup>+</sup> codoped and triply doped Yb3+/Er3+@Yb3+/Nd3<sup>+</sup> samples. We also observe a strong influence of the laser intensity on the relative spectroscopic ratio of I525/I545. Additionally, we

find that given the pulsed laser intensities appropriate for single particle imaging, at both 800 and 980 nm, that there are no significant differences in the local heating effects. This result is in contrast with that obtained when comparing excitation at both wavelengths under CW excitation. We further demonstrate our observations by comparing the differences in a DNA tightrope denaturation experiment at 800 and 980 nm pulsed irradiation respectively, and find no significant change in denaturation at excitation fluences that have previously been shown to support upconversion imaging utilizing 976 nm excitation at the single particle level at diameters of 10 nm with an intensity of 3e4 W/cm<sup>2</sup> and diameter of 25 nm at an intensity of 4e4 W/cm<sup>2</sup> as demonstrated by Gargas et al. (2014) and Green et al. (2016).

# EXPERIMENTAL

# Synthesis of β-NaYF4: 20% Yb3+, 2% Er3<sup>+</sup>

The β-NaYF4:20%Yb, 2%Er UCNPs were prepared by combining 2.1 mmol of sodium trifluoroacetate, 0.78 mmol of yttrium trifluoroacetate, 0.2 mmol of ytterbium trifluoroacetate, and 0.02 mmol of erbium trifluoroacetate in 6 mL of oleic acid and 6.1 mL of octadecene. The solution was degassed at 120◦C for 2 h with argon purging. Temperature was then increased to 330◦C under argon and allowed to maintain this temperature for 25 min. The particles were then cooled, precipitated, washed in excess ethanol with centrifuging, and dried under vacuum.

# Synthesis of β-NaYF4:40%Yb, 2%Er@NaYF4:20%Yb@NaNdF4:10%Yb Core-Shell-Shell UCNPs

The β-NaYF4:40%Yb, 2%Er Core UCNPs were prepared by a two-step thermolysis method. In the first step, CF3COONa (0.50 mmol), Y(CF3COO)<sup>3</sup> (0.29 mmol), Yb(CF3COO)<sup>3</sup> (0.20 mmol) and Er(CF3COO)<sup>3</sup> (0.01 mmol) precursors were mixed with oleic acid (5 mmol), oleyamine (5 mmol), and 1-octadecene (10 mmol) in a two-neck round bottom flask. The mixture was heated to 110◦C to form a transparent solution followed by 10 min of degassing. Then the mixture was heated to 300◦C at a rate of 15◦C/min under dry argon flow, and maintained at 300◦C for 30 min to form the α-NaYF4:40%Yb, 2%Er intermediate UCNPs. After the mixture cooled to room temperature, the α-NaYF4:40%Yb, 2%Er intermediate UCNPs were collected by centrifugal washing with excessive ethanol (7,500 g, 30 min). In the second step, the α-NaYF4:40%Yb, 2%Er intermediate UCNPs were redispersed into oleic acid (10 mmol) and 1 octadecene (10 mmol) together with CF3COONa (0.5 mmol) in a new two-neck round bottom flask. After degassing at 110◦C for 10 min, this flask was heated to 325◦C at a rate of 15◦C/min under dry argon flow, and maintained at 325◦C for 30 min to complete the phase transfer from α to β. After the mixture cooled to room temperature, the β-NaYF4:40%Yb, 2%Er UCNPs were collected by precipitated with equal volume of ethanol and centrifugation afterwards (7,500 g, 30 min). The β-NaYF4:40%Yb, 2%Er UCNPs were stored in hexane (10 mL).

Next, the as-synthesized β-NaYF4:40%Yb, 2%Er core UCNPs served as cores for the epitaxial growth of core-shell UCNPs. A hexane stock solution of β-NaYF4:40%Yb, 2% Er core UCNPs was transferred into a two-neck round bottom flask, and the hexane was sequentially evaporated by heating. CF3COONa (0.50 mmol), Y(CF3COO)<sup>3</sup> (0.40 mmol) and Yb(CF3COO)<sup>3</sup> (0.10 mmol) were introduced as UCNP shell precursors with oleic acid (10 mmol) and 1-octadecene (10 mmol). After 10 min of degassing at 110◦C, the flask was heated to 325◦C at a rate of 15◦C/min under dry argon flow and maintained at 325◦C for 30 min to complete the shell crystal growth. After the mixture cooled to room temperature, the β-NaYF4:40%Yb, 2%Er@NaYF4:20%Yb coreshell UCNPs were collected by precipitated with equal volume of ethanol and centrifugation afterwards (7,500 g, 30 min). The β-NaYF4:40%Yb, 2%Er@NaYF4:20%Yb core-shell UCNPs were stored in hexane (10 mL).

Afterwards, the as-synthesized β-NaYF4:40%Yb, 2%Er@NaYF4:20%Yb core-shell UCNPs served as cores for the epitaxial growth of shell crystal. A hexane stock solution of β-NaYF4:40%Yb, 2%Er@NaYF4:20%Yb core-shell UCNPs was transferred into a two-neck round bottom flask, and the hexane was sequentially evaporated by heating. CF3COONa (0.50 mmol), Nd(CF3COO)<sup>3</sup> (0.45 mmol) and Yb(CF3COO)<sup>3</sup> (0.05 mmol) were introduced as UCNP shell precursors with oleic acid (10 mmol) and 1-octadecene (10 mmol). After 10 min of degassing at 110◦C, the flask was heated to 325◦C at a rate of 15◦C/min under dry argon flow and maintained at 325◦C for 30 min to complete the shell crystal growth. After the mixture cooled to room temperature, the β-NaYF4:40%Yb, 2%Er@NaYF4:20%Yb@NaNdF4:10%Yb core-shell-shell UCNPs were collected by precipitated with equal volume of ethanol and centrifugation afterwards (7,500 g, 30 min). The β-NaYF4:40%Yb, 2%Er@NaYF4:20%Yb@NaNdF4:10%Yb core-shell-shell UCNPs were stored in hexane (10 mL).

### Characterization and Sample Preparation

UCNP were characterized with Transmission Electron Microscopy (TEM) to determine average diameter. The β-NaYF4:20%Yb, 2%Er (Wirth et al., 2017; **Figure 1b**) and β-NaYF4:40%Yb, 2%Er@NaYF4:20%Yb@NaNdF4:10%Yb coreshell-shell (**Figure 1c**) UCNPs were imaged using a JEOL 2000FX TEM at the Analytical Instrumentation Facility on North Carolina State University's campus. Samples used for spectroscopic experimentation were created by diluting the concentration of UCNPs to 10 ng/mL in ethanol and pipetting 100 uL of solution onto the sample surface. The solution was shaken on the sample surface for 20 min before washing with ethanol and drying under nitrogen.

# Time-Resolved Spectroscopy Measurements

Objects resolved on an optical microscope (40x, 0.9 N.A. air objective) imaged by a Andor NEO sCMOS camera with excitation from a 1,000 hz, tunable Nd: YAG laser with a 4.5 ns pulse width were selected by size, ensuring that imaged objects were less than the size of the diffraction limited spot of the wavelength of collection. Spectral distributions were separated using a half-meter monochromator with a custom exit with two slits. The gap between the slits correspond to a 10 nm wavelength difference centered at the monochromator's single slit location. By setting the monochromator to 535 nm, the 525 nm peak and 545 nm peak were effectively separated at the exit slits and further separated by a prism and coupled into separate SPCM-AQRH avalanche photodiodes. Controlled heating was performed using a peltier heater fixed to the back of the sample slide by epoxy. A thermistor attached to a PTC2.5K-CH temperature controller was embedded in the epoxy to ensure a constant set temperature during operation, and a K-type thermocouple was used to monitor the temperature at the sample surface.

## DNA Tightrope Measurements

The DNA substrates (CpG-free-rich, 7,163 bp) were ligated using the Quick Ligation Kit (New England BioLabs) at room temperature for overnight and then purified by phenolchloroform extraction to remove the ligase (Pan et al., 2017). Coverslips were PEGylated prior to use Flow cells were assembled by using double-sided tape to attach PEGylated coverslips to microscope slides with drilled input holes. After assembling the flow cell, poly-L-lysine (Wako Chemicals) treated silica beads (D∼5 um) were immobilized on the coverslips at a proper density and incubated for 5 min to ensure attachment. Afterwards, 100 uL of 5 ng/uL ligated DNA in a 7.4 pH buffer containing 50 mM HEPES, 100 mM NaCl, and 1 mM MgCl<sup>2</sup> was injected into the flow cell and pushed back and forth for 10 min using a syringe pump at a flow rate of 300 uL/min (Pan et al., 2017). After the DNA was stretched between the beads, the DNA was stained with a YOYO-1 dye. Following that, the DNA tightropes were centered in the laser focal area and imaged for 10 s with a Xenon lamp at 70 mW/cm<sup>2</sup> . The DNA is then exposed to 976 nm at 7 × 10<sup>4</sup> W/cm<sup>2</sup> or 806 nm at 5 × 10<sup>4</sup> W/cm<sup>2</sup> , for 2 min, followed by Xenon lamp imaging again for 10 s. A control DNA tightrope experiment without near infrared exposure was also performed alongside the near infrared exposed DNA tightropes.

# RESULT AND DISCUSSION

To establish the temperature sensing abilities of each type of UCNP under each excitation condition, controlled heating was used to bring the samples to set temperatures as described in the methods section. Spectroscopic measurements were performed at each temperature after a period of equilibration of 10 min. **Figure 2** shows a plot of ln(I525/I545) vs. 1/T measured for the both the β-NaYF4:20%Yb, 2%Er UCNPs and β-NaYF4:40%Yb, 2%Er@NaYF4:20%Yb@NaNdF4:10%Yb core-shell-shell UCNPs at 976 and 806 nm respectively. Since the population of the Er: <sup>2</sup>H11/<sup>2</sup> and Er: <sup>4</sup> S3/<sup>2</sup> levels fluctuates as a function of the Boltzmann's distribution (Lei et al., 2005),

$$\mathcal{R} = \frac{\mathcal{I}\_{525}}{\mathcal{I}\_{545}} = \Lambda \mathbf{e}^{\frac{-\Delta \mathcal{E}}{\mathbf{k} \mathcal{T}}}$$

where by taking the slope of the ln(I525/I545) vs. 1/T plot, a 1E of 887.170 cm−<sup>1</sup> (806 nm, core-shell-shell with Nd) and 966.176 cm−<sup>1</sup> (976 nm core only) is obtained. The calculated difference between 545 and 525 nm peaks is 700 cm−<sup>1</sup> . The sensitivity S is defined as,

$$\mathcal{S} = \frac{\mathcal{d}(\mathcal{R})}{\mathcal{d}\mathcal{T}} = \Lambda \frac{\Delta \mathcal{E}}{\mathcal{k}\mathcal{T}^2} \mathbf{e}^{\frac{-\Delta \mathcal{E}}{\mathcal{k}\mathcal{T}}}$$

where the higher the temperature, the greater the sensitivity. Given the calculated 1E, a plot of the sensitivity against temperature is shown in **Figure 2B**. Our sensitivity values are comparable to other researchers (León-Luis et al., 2012; **Table 1**). The higher sensitivity expands the applicability to environmental and electronics sensing where typical critical operating temperatures are higher.

Local temperature rise with continuous laser irradiation at high pump power intensity and duration has been observed

TABLE 1 | The natural log of the ratio between emissive states was plotted against inverse temperature as seen in Figure 2.

temperature. The sensitivity over the range of interest is plotted in (B) against temperature.


*This table is the result of fitting the resulting curve. The slope is directly equivalent to Delta E over k.*

(Wang et al., 2013), which can potentially impact in vivo applications. To address these concerns, we investigated the influence of the irradiation pump intensity and duration. To investigate the feasibility of experiments utilizing high laser fluence combined with pulsed excitation, we investigate the change in the spectroscopic ratio, with laser intensity and time duration in air and water.

**Figure 3** shows the dependence of the spectroscopic ratio, R, on the pump power intensity. There is a marked difference in the optical response of the particles with respect to near infrared excitation wavelength. In the β-NaYF4:20%Yb, 2%Er UCNPs (**Figure 3A**), R is observed to decrease with increasing pump power intensity at 976 nm excitation. This decrease in ratio apparently contradicts an expected rise in R in the presence of potential local heating. This is expected as a higher pump power intensity enables transition from 2 photon to 3 photon upconversion, where increasing pump power intensity leads to population of the <sup>4</sup>G11/2, after which phonon relaxation to the <sup>2</sup>H9/<sup>2</sup> level occurs. Subsequently radiative relaxation from the <sup>2</sup>H9/<sup>2</sup> to <sup>4</sup> I15/<sup>2</sup> level results in blue emission, as shown in **Figure 4B**. Thereby, due to preferential population of the <sup>4</sup>G11/<sup>2</sup> level, we expect to observe a lower photon population of the <sup>2</sup>H11/<sup>2</sup> (525 nm) level, at high pump power intensities. As the pump power is increased, the increase in phonon coupling to the lattice, and subsequent non-radiative energy transfer from the <sup>2</sup>H9/<sup>2</sup> to <sup>2</sup>H11/<sup>2</sup> and the <sup>4</sup> S3/<sup>2</sup> level to the <sup>4</sup>F9/<sup>2</sup> level occurs. Since the energy gap for the <sup>4</sup> S3/<sup>2</sup> to <sup>4</sup>F9/<sup>2</sup> transition of 3117 cm−<sup>1</sup> coincides with the typical value of 3,000–3,600 cm−<sup>1</sup> for OH vibrations(Kim et al., 2017), a higher pump power results in greater non-radiative relaxation via this pathway, as seen by the decrease in the rise and decay time of the Er3+: <sup>4</sup>F9/<sup>2</sup> to <sup>4</sup> I15/<sup>2</sup> transition (**Figures 2**–**4** and **Table 2**). In comparison, the energy gap for the Er3+: <sup>2</sup>H9/<sup>2</sup> to <sup>2</sup>H11/<sup>2</sup> transition of around 6000 cm−<sup>1</sup> , is much larger than that of the OH absorption energy. The resulting effect of a higher pump power is to promote greater blue and red emissions at the expense of green emission. A similar, but less dramatic, decrease in R is also seen at 806 nm excitation, where the absorption of two photons populates the <sup>2</sup>H9/<sup>2</sup> and subsequent radiative relaxation to the <sup>4</sup> I15/<sup>2</sup> level results in blue emission. We note that the absorption cross-section at 806 nm is comparatively low for this sample. However, increased pump power intensity at 806 nm did further increase blue emission, while also increasing rates of non-radiative transfer, leading to population of the <sup>4</sup>F9/<sup>2</sup> state. Therefore, in the β-NaYF4:20%Yb, 2%Er UCNPs, if used as a temperature sensor at either 806 nm or 976 nm excitation, blue emission is triggered at pump power intensities, which affects the <sup>2</sup>H11/<sup>2</sup> (525 nm) level population significantly and the ratio decrease should be accounted for.

In **Figure 3B**, the β-NaYF4:40%Yb, 2%Er@NaYF4:20%Yb@NaNdF4:10%Yb core-shell-shell UCNPs, at 976 nm excitation, shows a drop in R, but to a smaller degree, with increasing pump power intensity. This is attributed similarly to the β-NaYF4:20%Yb, 2%Er UCNPs as shown earlier. The smaller drop may be due to the protective nature of the core-shell-shell configuration, which limits the effect of lattice vibration effects. In contrast, at 806 nm excitation, an increase in R is seen with increasing pump power. The immediate assumption is that of an increase in local temperature, but results shown later will demonstrate that this is not the case. The increase in R appears to be an optical effect only. Under 806 nm excitation, the Nd3+: <sup>4</sup>F5/<sup>2</sup> and Nd3+: <sup>2</sup>H9/<sup>2</sup> levels are populated due to transitions from the Nd3+: 4 I9/<sup>2</sup> ground state. Subsequent non-radiative relaxation from these two states occur,

FIGURE 3 | Plot of measured ratio between <sup>2</sup>H11/<sup>2</sup> and <sup>4</sup>S3/<sup>2</sup> at increasing intensity for both 806 nm and 976 nm excitation for (A) <sup>β</sup>-NaYF4:20%Yb, 2% Er (A) and (B) β-NaYF4:40%Yb, 2%Er@NaYF4:20%Yb@NaNdF4:10%Yb.

leading to population of the Nd3+: <sup>4</sup>F3/<sup>2</sup> state. The resonant energy levels of Nd3+: <sup>4</sup>F3/<sup>2</sup> and Yb3+: <sup>2</sup>F5/<sup>2</sup> ensures a very high efficiency energy transfer (ET) between Nd3<sup>+</sup> ions and Yb3<sup>+</sup> ions(Wang et al., 2013; Tian et al., 2014; Zhong et al., 2014; Chen et al., 2015). Thereafter, two consecutive energy transfers from the Yb3+: <sup>2</sup>F5/<sup>2</sup> state to the neighboring Er3<sup>+</sup> ions, result in population of the <sup>4</sup>F7/<sup>2</sup> state. Relaxation from the <sup>4</sup>F7/<sup>2</sup> state to the lower <sup>2</sup>H11/2, 4 S3/<sup>2</sup> and <sup>4</sup>F9/<sup>2</sup> states, followed by radiation relaxation from all three of states to the <sup>4</sup> I15/<sup>2</sup> ground state, give rise to emission at 525 nm, 545 nm, and 655 nm, respectively. Since the highly absorbing Nd3<sup>+</sup> ion promotes more efficient energy transfer to the Yb3<sup>+</sup> ion, a higher pump power intensity also leads to a more efficient population of the <sup>4</sup>F7/<sup>2</sup> level, where further non-radiative relaxation to the <sup>2</sup>H11/<sup>2</sup> state occurs, giving rise to preferential emission at 520 nm. Therefore, the Nd3<sup>+</sup> dopant, due to the larger absorption cross-section, and more efficient energy transfer to the Yb3<sup>+</sup> ion, clearly favors upconversion over linear decay. As a result, an increase in laser intensity, leads to greater upconversion.

To further understand the source of the decrease in R with increasing excitation intensity for the β-NaYF4:20%Yb, 2%Er UCNPs, time-resolved decay spectra were measured. **Figures 4A,B** displays the decay spectra for <sup>4</sup>F3/2(blue), <sup>2</sup>H11/2(green), <sup>4</sup> S3/2(green), and <sup>4</sup>F9/2(red) emissions at two different excitation intensities. It can be seen that at a higher excitation intensity (5.6 × 10<sup>4</sup> W/cm<sup>2</sup> ) in **Figure 4B**, that the <sup>4</sup>F9/<sup>2</sup> and <sup>4</sup> S3/<sup>2</sup> states excitation pathways are more strongly coupled as evidenced by the narrowing of the gap between their rise times (See **Table 2**). This has been attributed to the increase in phonon coupling to the lattice OH vibrations at higher laser intensity. At lower laser intensity, the <sup>4</sup>F9/<sup>2</sup> rise time is significantly longer in comparison to that of the <sup>4</sup> S3/<sup>2</sup> state. **Supporting Figure S1** shows normalized time-resolved decay at 545 nm of the β-NaYF4:20%Yb, 2%Er UCNPs and β-NaYF4:40%Yb, 2%Er@NaYF4:20%Yb@NaNdF4:10%Yb core-shell-shell UCNPs. **Supporting Figures S1A,B** show the 545 nm decay of the Yb/Er co-doped sample from excitation at 976 nm and 806 nm respectively. In **Supporting Figure S1A**, the slow rise times and long decay times are indicative of the ETU upconversion process. As the pump power intensity is increased, the rise times and mean lifetimes are both reduced. The corresponding emission at 806 nm excitation is dim, due to the low absorption cross-section at this excitation wavelength, hence the decay curves are noisy, and convey less information. **Supporting Figures S1C,D** shows the decay curves at 545 nm for the core-shell-shell UCNPs for excitation at 976 nm and 806 nm respectively. Both figures clearly show faster rise times and decay,

TABLE 2 | Lifetime and rise times fitted from the time-resolved decay for <sup>β</sup>-NaYF4:20% Yb3+, 2% Er3<sup>+</sup> from Figure 4.


*The lifetime was fitted as a single exponential for the emission's decay and the rise time was defined as the time between the laser excitation and the maximum point of the decay curve.*

which we attribute, for the 976 nm excitation, to the presence of core-shell-shell nanostructure, which reduces phonon coupling to the host matrix. At 800 nm excitation, the highly absorbing Nd3<sup>+</sup> ion further increases the decay rate for the Er3+: 4 S3/<sup>2</sup> to 4 I15/<sup>2</sup> transition.

Since it is unclear as to the impact of pump power intensity on local heating due to the increased probability of higher energy level transitions, we investigated the effect of time duration of excitation on the local temperature rise. **Figures 5A,B** show the dependence of the local temperature with time duration of irradiation in air (**Figure 5A**) and water (**Figure 5B**), respectively. Neither 976 nm, nor 806 nm excitation introduces a local temperature rise over time at laser intensities in the range of 10<sup>4</sup> W/cm<sup>2</sup> . Our results are significantly different from other researchers primarily because our experiments are conducted with pulsed excitation, unlike the effects seen using continuous irradiation as observed by other researchers (Wang et al., 2013). Therefore, the laser mode of operation should be carefully considered regarding the use of upconversion nanocrystals in thermal sensing. While continuous wave lasers are affordable and highly adaptable to many laboratories, the probability of local heating is high. In contrast, pulsed wave excitation under a tightly focused beam, which is typical in single-molecule imaging does not cause appreciable local temperature rise in the sample.

As a demonstration of the effects of 976 vs. 806 nm irradiation at the single-molecule level, we compared DNA tightropes exposed to both wavelengths at intensities shown previously to support single particle imaging (Gargas et al., 2014; Green et al., 2016). In order to clearly identify effects of near infrared exposure, we first stretch DNA on silica beads by flowing the DNA across the beads in an aqueous environment. After the DNA are stretched out and stained with YOYO-1 dye, the laser spot is localized on the DNA tightrope. The tightropes are then exposed to 806 nm or 976 nm excitation separately

FIGURE 5 | Plot of measured temperature converted from spectroscopic ratio against time at a fixed power for particles without Nd (β-NaYF4: 20% Yb3+, 2% Er3+) and with Nd (β-NaYF4: 40%Yb3+, 2% Er3+@NaYF4: 20% Yb3+@NaNdF4: 10% Yb3+) in air (A) and in water (B). The laser was blocked for 10 min before the beginning of the measurement and unblocked at around 20 s to collect any changes in temperature after the sample was exposed. The powers used were 8.0 × 10<sup>4</sup> W/cm<sup>2</sup> and 5.3 × 10<sup>4</sup> W/cm<sup>2</sup> for 976 and 806 nm respectively. Both power intensities were maintained at comparable orders of magnitude as much as possible. The pulse width is ∼4.5 ns and pulse frequency is at 1,000 hz. Each point is at 16 s intervals, so each point is the sum of 16 k pulses.

at power intensities compatible with single UCNP imaging as demonstrated in previous studies (Gargas et al., 2014; Green et al., 2016). **Figure 6** shows that exposure to 800 nm (5 × 10<sup>4</sup> W/cm2) and 976 nm (7 × 10<sup>4</sup> W/cm<sup>2</sup> ) excitation for 2 min durations only cause dsDNA breaks 10% of the time. For a control DNA tightrope experiment where there was no exposure to near infrared excitation, the rate of occurrence of dsDNA breaks were also around 10%. Therefore, any damage observed is consistent with 10 s of UV or blue light nicking of the DNA, from the Xenon Arc Lamp source used to detect YOYO-1 emission. Clearly, from our measurements, the lack of DNA damage under infrared exposure leads us to infer that thermal heating is not significant at the current pump power intensities used in our measurements especially when considered alongside the lack of heating in water, as shown in **Figure 5**.

# CONCLUSION

UCNPs with and without Nd3<sup>+</sup> sensitization may be good candidates for application as nanothermometers in single biological molecule experimental situations. It was demonstrated that each type of UCNP was able to report temperature ratiometrically with sensitivities in the 1 × 10−<sup>4</sup> K −1 range. In addition, it was determined that excitation intensity is a parameter of nanothermometry that must be strongly controlled in the single particle level to avoid ratio suppression due to alteration of the optical pathways involved in ratio reporting. Finally, the excitation intensities previously used to image single nanoparticles were shown to be applicable to DNA tightrope experiments without appreciable change in the viability of DNA over time windows appropriate for the study of protein-DNA interactions.

# AUTHOR CONTRIBUTIONS

The study aims were proposed by SL. KH synthesized the UCNPs. KG designed and performed the spectroscopic experiments. HP prepared the biologicals for the DNA tightrope experiments. Data interpretation and analysis were performed by KG, KH, GH, and SL.

# FUNDING

This work was financially supported by the National Institutes of Health 1R21ES027641-01 grant and the National Science Foundation NSF1706651 grant.

### REFERENCES


# SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fchem. 2018.00416/full#supplementary-material

Supplementary Video 1 | Intensity varies as nanoparticles tumble in the laser spot due to polarization sensitivity. Seen here, single nanoplasmonic UCNPs tumble in 50% sucrose solution, recorded at a frame rate of 10 Hz.

Supporting Figure S1 | Time-resolved decay curves at 545 nm for β-NaYF4:20%Yb, 2%Er at (A) 976 nm, and (B) 807 nm excitation respectively, and β-NaYF4:40%Yb, 2%Er@NaYF4:20%Yb@NaNdF4:10%Yb at (C) 976 nm, and (D) 806 nm excitation respectively. In frame (D) the sample was partially removed from the collection area, resulting in lower signal, noisy collection.


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Green, Huang, Pan, Han and Lim. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Acetone Sensing Properties and Mechanism of Rh-Loaded WO<sup>3</sup> Nanosheets

Zhilei Qiu<sup>1</sup> , Zhongqiu Hua<sup>1</sup> \*, Yan Li <sup>1</sup> , Mengjun Wang<sup>1</sup> \*, Dan Huang<sup>1</sup> , Chen Tian<sup>1</sup> , Chensheng Zhang<sup>1</sup> , Xuemin Tian<sup>1</sup> and Erping Li <sup>2</sup>

*<sup>1</sup> Tianjin Key Laboratory of Electronic Materials and Devices, School of Electronics and Information Engineering, Hebei University of Technology, Tianjin, China, <sup>2</sup> Key Laboratory of Micro-Nano Electronics and Smart System of Zhejiang Province, Department of Information Science & Electronic Engineering, Zhejiang University, Hangzhou, China*

WO<sup>3</sup> nanosheets was prepared by an acidification method and the Rh catalyst was dispersed on the surface of the nanosheets with a wet impregnation method. The morphology of pristine WO<sup>3</sup> and Rh modified WO<sup>3</sup> nanosheets and their responses to acetone gas were studied. According to oxygen adsorption combined with TPR results, the sensing and sensitization mechanism of acetone were discussed. It was found that no visible changes in nanostructures or morphologies were observed in WO<sup>3</sup> nanosheets with Rh, however, the sensor resistance and sensor response were greatly promoted. The basic sensitization mechanism could be caused by the electronic interaction between oxidized Rh and WO<sup>3</sup> surface.

### Edited by:

*Wen Zeng, Chongqing University, China*

### Reviewed by:

*Dachi Yang, Nankai University, China Guotao Duan, Institute of Solid State Physics, Hefei Institutes of Physical Science (CAS), China*

### \*Correspondence:

*Zhongqiu Hua zhongqiuhua@hebut.edu.cn Mengjun Wang wangmengjun@hebut.edu.cn*

### Specialty section:

*This article was submitted to Nanoscience, a section of the journal Frontiers in Chemistry*

Received: *12 July 2018* Accepted: *09 August 2018* Published: *11 September 2018*

### Citation:

*Qiu Z, Hua Z, Li Y, Wang M, Huang D, Tian C, Zhang C, Tian X and Li E (2018) Acetone Sensing Properties and Mechanism of Rh-Loaded WO3 Nanosheets Front. Chem. 6:385. doi: 10.3389/fchem.2018.00385*

### Keywords: WO3 , Rh, acetone, surface modification, gas sensors

# INTRODUCTION

Acetone gas is closely related to people with diabetes. Medical research has shown that there is a significant difference of acetone concentration in the breath for diabetics and healthy people, the former being higher than 1.8 ppm and the latter being below 0.8 ppm (Owen et al., 1982; Natale et al., 2014). Therefore, through the quantitative detection of the acetone concentration in human exhaled gases, it could achieve the purpose of diagnosis and monitoring to the disease condition. Metal oxide semiconductors (MOS) have been widely reported for gas sensors with the significant advantages, such as low cost, simple process and small size (Hübner et al., 2010; Choi et al., 2014). Tungsten trioxide (WO3) as an n-type semiconductor has become a research hotspot in the detection of VOC gases in recent years (Kanda and Maekawa, 2005; Kadir et al., 2015; Li et al., 2017). The adsorption and reaction of VOC gas on WO<sup>3</sup> surface could change the semiconductor resistance, so the gas response can be improved by adding highly efficient catalytic elements. The introduction of ruthenium (Ru) and silicon (Si) improve the sensitivity of WO<sup>3</sup> to acetone (Righettoni et al., 2010; Li et al., 2018). Further, Rh is known as a highly efficient catalyst to the catalytic reaction of acetone gas (Houtman and Barteau, 1991). It has been reported that Rh loaded SnO<sup>2</sup> and In2O<sup>3</sup> significantly improve the response of acetone (Kim et al., 2011; Kou et al., 2018). Therefore, this highly efficient catalyst could be also loaded onto WO<sup>3</sup> surface to increase the response to acetone. In this study, the Rh element was uniformly loaded onto the surface of WO<sup>3</sup> nanosheets based on an impregnation approach. This method has been frequently used in our previous work (Li et al., 2018). The experimental results show that the Rh nanoparticles can significantly enhance the response of WO<sup>3</sup> nanosheets to acetone without changing the surface morphology of WO<sup>3</sup> nanosheets. The basic sensitization mechanism of Rh was also analyzed based on experimental results.

# EXPERIMENTAL

WO<sup>3</sup> nanosheets was obtained by dropping Na2WO<sup>4</sup> solution into H2SO<sup>4</sup> solution (Kida et al., 2009). Aqueous solution of RhCl<sup>3</sup> was impregnated with WO<sup>3</sup> (Rh-WO3) powders and formed a suspension slurry, which was washed by distilled water and dried. Subsequently, the powders were annealed at 500◦C in air. Sensor devices were made by the screen-printing technique. The crystal structures were measured by X-ray diffractometer (XRD; D8FOCUS, Germany). The morphology of sample was analyzed using scanning electron microscopy (FE-SEM; Nova Nano SEM 450, FEI, U. S). Nanosheets were also characterized by a transmission electron microscopy (TEM; Tecnai-F20, FEI, U.S). Energy spectrum analysis of materials uses X-ray photoelectron spectroscopy (XPS, Thermo escalab 250Xi, U. S). The catalyst activity was characterized by H<sup>2</sup> temperature programmed reduction (H2-TPR; TP-5076, China). The experimental procedure of the TPR was descripted in **Figure S1**. Gas sensing tests were carried out by a conventional gas flow apparatus (see **Figure S2**). The gas sample was kept at a constant flow rate of 100 ml/min by mass flow controllers (MFC). The humidity of gas was <20 ppm and temperature of the chamber was about 50◦C. The sensor response (S) was defined as S = Ra/R<sup>g</sup> , where R<sup>a</sup> and R<sup>g</sup> are the sensor resistance in air and in the presence of target gases.

## RESULTS AND DISCUSSION

The morphology of WO<sup>3</sup> nanosheets was characterized by SEM and TEM. **Figures 1A,B** show SEM images of pristine and 1wt.% Rh-WO<sup>3</sup> nanosheets. One can see that the sample powders consisted of a large amount of nanoparticles with a lateral size from dozens to several hundred nanometers. According to SEM images, there were no visible changes observed in pristine WO<sup>3</sup> and Rh modified one. For the results of specific surface area, pristine WO<sup>3</sup> was ∼12 m<sup>2</sup> /g and 1wt. %Rh-WO<sup>3</sup> is about 13 m<sup>2</sup> /g, which indicts no significant change. **Figure 2** shows TEM images of the pure WO<sup>3</sup> and 1wt.%Rh-WO<sup>3</sup> nanosheets. It was obvious that the sample powder is actually composed of highly irregular plate-like nanosheets. The insert image of **Figure 2A** presents a selected area diffraction (SAD) pattern of pristine WO<sup>3</sup> nanosheets, suggesting that the nanosheets have a good crystal quality. In addition, some white particles with dozens of nm in size were observed in WO<sup>3</sup> surface, as shown in **Figure 2B**. With the help of SAD in **Figure 2C**, these particles were identified as Rh2O<sup>3</sup> with lattice spacing of 0.26 nm, corresponding to the (110) plane (JCPDS: 25- 0707). It was thought that these large particles of Rh2O<sup>3</sup> could be due to the aggregation of Rh during washing and drying process, which were not effectively removed during the washing process. While the lattice spacing of 0.38 nm in the HRTEM image was belong to monoclinic WO<sup>3</sup> (JCPDS: 43- 1035), which was in a good agreement of XRD results (in **Figure S3**).

The chemical state of Rh on WO<sup>3</sup> surface was also analyzed by XPS. **Figure 3A** presents the XPS spectra of W, detection results of binding energy for W4f7/<sup>2</sup> and W4f5/<sup>2</sup> being 35.7eV and 37.9eV, respectively, which is in good agreement with W6<sup>+</sup> (Dupin et al., 2000). **Figure 3B** shows the XPS spectra of Rh3d obtained from 1wt%Rh-WO3. Among them, the Rh2d5/<sup>2</sup> peaked at 309.45eV is a typical oxide centered on Rh3+. In addition, the Rh2d3/<sup>2</sup> located at 314.5eV is also an oxide centered on Rh3<sup>+</sup> (Kim et al., 2011). Thus, it could be concluded that Rh was present as an oxidized state of Rh2O<sup>3</sup> on WO<sup>3</sup> surface. Additional with XPS results, the oxidized state of Rh could be also evidenced by H2-TPR. **Figure 3C** shows the H2-TPR results of pristine and 1wt%Rh-WO<sup>3</sup> nanosheets. As expected, there was one weak peak around 370◦C observed in pristine WO<sup>3</sup> nanosheets indicating a weak consumption of H2, which may be due to the weak reduction behavior of WO<sup>3</sup> surface at a high temperature (Li et al., 2018). In contrast, large consumptions of H<sup>2</sup> were observed in 1wt%Rh-WO3, suggesting a strong reduction behavior. There were two overlapped peaks of H<sup>2</sup> consumption at a low temperature around 110◦C and the intensities of peaks were relatively high. It was believed that the consumption of H<sup>2</sup> observed at low temperatures could be due to the reduction of Rh2O<sup>3</sup> and peaks located different temperature

may be ascribed to different dispersion states of Rh species. There was a broad but weak peak of H<sup>2</sup> consumption at around 450◦C, which could be attributable to the weak reduction of WO<sup>3</sup> surface, i.e., surface lattice oxygen (OL) reacting with H<sup>2</sup> at a high temperature. The reduction behavior of Rh-WO<sup>3</sup> was much stronger than pristine WO<sup>3</sup> indicating that the reactivity of lattice oxygens is slightly promoted by Rh2O<sup>3</sup> on the surface. We can see significant differences, comparing this reduction behavior with our previous study of Pt-WO<sup>3</sup> nanosheets (in **Figure S4**). At low temperature, Pt-WO<sup>3</sup> produces a negative peak of H<sup>2</sup> desorption. Based on the results of TPR and the resistance behavior under PO2, it is concluded that the main sensitization of Pt-WO<sup>3</sup> may be caused by redox of Pt. nanoparticles (Li et al., 2017). This phenomenon of Rh may cause different sensitization mechanisms.

The sensing properties of pristineWO<sup>3</sup> and Rh-WO<sup>3</sup> nanosheets were characterized with acetone ranging from 0.5 to 10 ppm. **Figure 4A** shows the time dependence of sensor

resistance. It was worth noting that the introduction of Rh greatly increased the sensor resistance of WO3. For 2wt.% Rh-WO3, the sensor resistance was almost three orders of pristine one. This indicated a strong electronic interaction between Rh2O<sup>3</sup> and WO<sup>3</sup> surface, forming the well-known P-N junction or fermi-level control sensitization mechanism. Due to the electronic junction of Rh2O<sup>3</sup> with WO3, the sensors response and responding speed were significantly promoted. **Figure 4B** shows the calibration line of sensors resistance with concentration of acetone at an operation temperature of 250◦C. It was found that sensor based on 1wt.%Rh-WO<sup>3</sup> also responded to 0.5 ppm acetone. One can see that sensor response was increased by 3 times compared with the neat WO3. However, an excess of Rh did not effectively to promote the sensor response. This observation was in conflict with the great enhancement in sensor resistance. In order to explain the reduction in sensor response for 2wt.%Rh-WO3, there were two factors should be considered. Firstly, an excessive amount of Rh could lead to agglomeration of Rh2O<sup>3</sup> and poor dispersion on the surface of WO<sup>3</sup> nanoparticles. Consequently, some electronic interaction of Rh2O<sup>3</sup> with WO<sup>3</sup> leading to the high resistance were not effective to the sensitization. Secondly, with increasing the amount of Rh the surface activity of WO<sup>3</sup> could be enhanced and then leaded to a catalytic reaction of acetone, which inhibit the diffusion of acetone molecule into inside of sensor films. As a result, the sensor response was reduced by a high loading amount of Rh. This reduction in sensor response could be also observed when increasing operation temperatures. This was evidenced by the strong dependence of sensor response on the operation temperatures for Rh-WO<sup>3</sup> as shown in **Figure S5a**. It was thought that increasing the operating temperature leaded to an enhancement in catalytic activity, which reduces the gas diffusion and sensor response to acetone. When operating at a temperature larger than 250◦C, one can note the sensor response greatly decreased with temperatures. It was also found that sensors resistance also obviously decreased with temperature, as shown in **Figure S5b**. For pristine WO3, the sensor response did not change significantly with temperature and exhibited a lower response at different temperatures. This poor response is associated with a weaker oxygen adsorption on WO<sup>3</sup> surface (Zeng et al., 2017). At the same time, the stability of the sensor was also evaluated as shown in **Figures S5c,d**. It can be seen that Rh-WO<sup>3</sup> nanoparticles can work for a long time at 350◦C and has favorable response recovery performance.

It is well-known that oxygen adsorption in the form of O<sup>−</sup> 2 , O−, or O2<sup>−</sup> on the surface serves as the receptor function and determines the sensing ability and mechanism of MOS gas sensors (Hua et al., 2018a). In order to explore the sensitization effect of Rh-WO<sup>3</sup> nanosheets, we analyzed the oxygen adsorption behavior. **Figure 5A** shows a linear plot of sensor resistance (Rg) with the partial pressure of oxygen (PO2) at a double logarithmscale for pristine and 1wt.% Rh-WO<sup>3</sup> sensors. It was observed that a linear relationship indicating a power-law response within all PO2 ranging from 0.06 to 0.99 atm (1 atm = 100% in volume) and the linear fitting coefficients were 0.42 and 0.62 for pristine and Rh-WO3, respectively. This indicated that the main type of oxygen adsorption was in the form of O<sup>−</sup> for both sensors (Hua et al., 2018a,b) at working temperature of 300◦C through:

$$\text{2O}\_2 + 2\text{e}^- \rightleftharpoons 2\text{O}^- \tag{1}$$

In case of 2 wt.% Rh-WO3, the linear plot of lnR<sup>g</sup> with lnPO2 was also valid. Remarkably, the slope, i.e., fitting coefficient was just 0.29, considerably <0.5. However, it was unlikely that a large amount of Rh on the surface could tailor the form of oxygen ionisorption on the surface. The most probably explanation was that with increasing PO2 the oxidized state of Rh, which has been limited to be exposed to atmosphere due to the aggregation of particles, was enhanced and then the electronic interaction between Rh2O<sup>3</sup> and WO<sup>3</sup> surface was promoted. Consequently, new depletion regions formed, leading to an increase in sensor resistance with PO2 and a reduction in the fitting coefficient. This has also been observed in our previous Pt-WO<sup>3</sup> sensor (Li et al., 2017).

According to our recent study, it was found that the powerlaw response of oxygen in the presence of reducing gas such as H2, CO, and acetone can be used to clarify the basic sensing mechanism of gas sensors (Hua et al., 2018c). **Figure 5B** shows the power-law response of oxygen in the presence of acetone (2 ppm) for pure and Rh-WO<sup>3</sup> sensors. A very good linearity was observed for all sensors indicating that the basic sensing mechanism of acetone could be explained by the oxidation of acetone with oxygen adsorbates by:

$$\rm{CH\_3COCH\_3} + 8O^- \rightarrow \rm{3CO\_2} + 3H\_2O + 8e^- \tag{2}$$

For simplicity, it was assumed that acetone catalytic reaction was a complete reaction only producing CO<sup>2</sup> and H2O. However, in fact the oxidation of acetone was rather complex. In addition, it was also found that linear coefficients of the power-law response were all around 0.5, which was consistent with **Figure 5A** and Equation (1). Importantly, for 2wt.%Rh-WO3, the fitting coefficient significantly raised up compared with that in the absence of acetone. This clearly supported our explanation for the degradation of sensitization effect with large loading amount of Rh and the reduction in the exponent of the power-law response to oxygen. In this respect, we believe that the basic sensitization mechanism of Rh on WO<sup>3</sup> could be ascribed to the electronic interaction between Rh2O<sup>3</sup> and WO<sup>3</sup> (p-n junction), which was very similar with the fermi-level control model, popular for Pd-SnO<sup>2</sup> sensors (Tang et al., 2015). The key factor to achieve a good sensitization effect highly relies on an elegant dispersion of Rh2O<sup>3</sup> on WO<sup>3</sup> surface, which can enhance the electronic interaction with WO<sup>3</sup> surface as schematically drawn in **Figure S6**. This finding was similar with the case of Pt and Ru loaded WO<sup>3</sup> nanosheets, however, it was significantly different with Pd and Fe loaded WO3. For the later one, the chemical sensitization effect of Pd and Fe plays a vital role through the reaction of surface lattice oxygens with reducing gases.

### CONCLUSION

In summary, Rh as a noble catalyst was dispersed onto the surface of WO<sup>3</sup> nanosheets through a wet impregnation method. Experimental results indicated that Rh was in a form of oxidized state Rh2O<sup>3</sup> on WO<sup>3</sup> surface and an excessive amount of Rh can lead to an aggregation of Rh2O<sup>3</sup> and poor sensitization effect as well. An electronic interaction between Rh2O<sup>3</sup> and WO<sup>3</sup> surface was evidenced by an extremely high argument in sensor resistance and it was thought that such an electronic was responsible for the observed sensitization effect of Rh loading. To achieve a good sensitization effect, an elegant dispersion of Rh2O<sup>3</sup> is required, which highly relies on an effective dispersion method and a proper loading amount. Additionally, a power-law response to oxygen was observed for both pristine and Rh-WO<sup>3</sup> in the presence of acetone, which indicts that oxygen adsorption on the surface of WO<sup>3</sup> serves as a basic receptor function.

### AUTHOR CONTRIBUTIONS

YL performed the experiments and analyzed the data with help from DH, CT, CZ, and XT. ZH, MW and EL conceived and guided the study. ZQ wrote the manuscript based on experimental data.

### FUNDING

This study was supported by the National Natural Science Foundation of China (Grant NO. 61501167) and Natural Science Foundation of Tianjin (Grant NO. 15JCYBJC52100).

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fchem. 2018.00385/full#supplementary-material

# REFERENCES


In2O<sup>3</sup> hollow spheres. J. Mater. Chem. 21, 18560–18567. doi: 10.1039/C1JM 14252F


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Qiu, Hua, Li, Wang, Huang, Tian, Zhang, Tian and Li. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Application of Graphene Hybrid Materials in Fault Characteristic Gas Detection of Oil-Immersed Equipment

### Lingfeng Jin1,2,3, Weigen Chen1,2 \* and Ying Zhang<sup>3</sup> \*

<sup>1</sup> State Key Laboratory of Power Transmission Equipment & System Security and New Technology, Chongqing University, Chongqing, China, <sup>2</sup> School of Electrical Engineering, Chongqing University, Chongqing, China, <sup>3</sup> School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA, United States

Graphene and its hybrid materials, due to their unique structures and properties, have attracted enormous attention for both fundamental and applied research in the gas sensing field. This review highlights the recent advances in the application of graphene-based gas sensors in fault characteristic gas detection of oil-immersed equipment, which can effectively achieve condition monitoring of the oil-immersed power equipment. In this review, the synthetic methods of graphene hybrid materials with noble metals, metal oxides and their combination are presented. Then, the basic sensing mechanisms of graphene hybrid materials and gas sensing properties of graphene hybrid materials sensors to hydrogen (H2), carbon monoxide (CO), carbon dioxide (CO2), methane (CH4), acetylene (C2H2), ethylene (C2H4), and ethane (C2H6), which are the fault characteristic gas in oil-immersed power equipment, are summarized. Finally, the future challenges and prospects of graphene hybrid materials gas sensors in this field are discussed.

### Edited by:

Zhongchang Wang, Laboratório Ibérico Internacional de Nanotecnologia (INL), Portugal

### Reviewed by:

Ming-Guo Ma, Beijing Forestry University, China Chunli Zhang, Zhejiang University, China

### \*Correspondence:

Weigen Chen weigench@cqu.edu.cn Ying Zhang yzhang@gatech.edu

### Specialty section:

This article was submitted to Nanoscience, a section of the journal Frontiers in Chemistry

Received: 22 June 2018 Accepted: 20 August 2018 Published: 11 September 2018

### Citation:

Jin L, Chen W and Zhang Y (2018) Application of Graphene Hybrid Materials in Fault Characteristic Gas Detection of Oil-Immersed Equipment. Front. Chem. 6:399. doi: 10.3389/fchem.2018.00399 Keywords: graphene, gas sensor, oil-immersed equipment, sensing mechanism, fault characteristic gas

# INTRODUCTION

Graphene is an allotrope of carbon consisting of a single layer of carbon atoms arranged in a hexagonal lattice, which was rediscovered, isolated, and characterized by Geim and Novoselov in 2004 (Novoselov et al., 2004; Geim and Novoselov, 2007). Owing to the unique structure, graphene exhibits excellent physical and chemical properties, and has opened a new and very promising scientific area with a lot of focus on material science and potential applications (Aïssa et al., 2015; Higgins et al., 2016; Long et al., 2018). Among these outstanding properties of graphene, the high electron mobility of up to 200,000 cm<sup>2</sup> /Vs and superior specific surface area of 2,630 m<sup>2</sup> /g make graphene an extremely sensitive material for gas detection (Akturk and Goldsman, 2008; Chen et al., 2008; Bonaccorso et al., 2015). However, pristine graphene, which lacks dangling bonds, is unfavorable for the adsorption of gas molecules on its surface. Therefore, the modification of graphene and its derivatives via physicochemical methods, such as covalent, non-covalent and doping functionalization, has been investigated. Usually, graphene hybrid materials, which are composed of graphene and typical sensing materials (e.g., noble metals, metal oxides, or their ternary hybrids), have significantly improved performance due to the synergistic interaction between graphene and the typical sensing material (Meng et al., 2015). There have been numerous works published on the basic research and the sensing applications of graphene and graphene hybrid materials.

**64**

One such application is condition monitoring of oil-immersed equipment. Insulating oil with high specific heat and dielectric strength is widely used in the high voltage power equipment. However, the insulating oil can be cracked into small related gas molecules due to the long-term electrothermal effect, which reduces the insulation strength and leads to the malfunction of the power equipment. According to previous studies, effective detection of seven typical fault characteristic gases, including H2, CO, CO2, CH4, C2H2, C2H4, and C2H6, can reflect the operation state of oil immersed power equipment (Bakar et al., 2014). Recently, a lot of research has been carried out on using graphene hybrid materials to detect these gases for rapid and accurate fault detection of oil-immersed equipment (Acharyya and Bhattacharyya, 2016; Zhou et al., 2016; Nasresfahani et al., 2017; Zhang et al., 2017b).

Although there are many reviews on the applications of gas sensing using graphene materials (Wang T. et al., 2016; Singhal et al., 2017; Dai et al., 2018), few has focused on detecting the typical gases reflecting the characteristics of faults in oilimmersed equipment. The objective of this review is to provide researchers a systematic understanding of the development of graphene hybrid materials in this application field. The main synthesis method and properties of graphene hybrid materials are introduced in section Synthesis, Properties, and Experimental Testing. Then, the gas sensing mechanism of graphene hybrid materials is discussed in section Sensing Mechanism. In section Application of Gas Sensor, the gas sensing properties of graphene hybrid materials to typical fault characteristic gases of oilimmersed equipment are summarized and compared. Finally, we analyze the challenges of gas sensing application in this field and present the suggestions of the future development trend.

# SYNTHESIS, PROPERTIES, AND EXPERIMENTAL TESTING

### Synthesis and Properties

Since the discovery of graphene in 2004, a lot of research has been conducted on how to synthesize graphene with high quality and large scale. To this date, the preparation of graphene can be mainly divided into top-down routes and bottom-up routes. Top-down routes use graphite as raw materials, and graphene is mainly obtained through separating the carbon atom layer by mechanical exfoliation, chemical oxidation-reduction reaction, and electrochemical methods. In contrast, bottom-up routes use small carbon-based molecules as raw materials, and graphene is obtained by silicon carbide pyrolysis, chemical vapor deposition (CVD), and solvothermal methods. Paton et al. (2014) developed a simple model that shows exfoliation occurs once the local shear rate exceeds 10<sup>4</sup> /s and demonstrated a scalable method for producing relatively large quantities of defectfree graphene. Rathnayake et al. (Yola et al., 2015) presented the methods for preparing graphene oxide (GO) and reduced graphene oxides (rGO) using chemical oxidation and reduction processes, respectively, which are based on needle platy variety of natural vein graphite that has high purity and crystallinity and low cost. Hofmann et al. (2015) investigated the process of electrochemical exfoliation and the impact of its parameters on the produced graphene, and achieved the synthesis of graphene with controllable electronic and mechanical characteristics. Son et al. (2015) reported direct graphene growth over silicon nanoparticles without silicon carbide formation, and the volumetric energy densities are 972 and 700 Wh/L at the first and 200th cycles, respectively. Banszerus et al. (2015) showed that the quality of CVD-grown graphene depends critically on the used transfer process and reported an advanced transfer technique that allows both reusing the copper substrate of the CVD growth and making devices with mobilities as high as 350,000 cm<sup>2</sup> /Vs. Quan et al. (2014) successfully synthesized sulfur-doped and nitrogendoped graphene by using a solvothermal method, and these heteroatom-doped graphene materials exhibited high surface areas and high contents of heteroatoms. With the synthetic routes mentioned above, controllable preparation of graphene can be achieved.

Graphene as an ideal base material can be combined with noble metals, metal oxides, or their ternary hybrids to form graphene hybrid materials with outstanding gas sensing properties. Metal nanoparticles combine the excellent properties of metal and characteristics of nanomaterials, and show great potential in catalysis, sensing and electronics fields. Up to now, nanoparticles of many different metals, such as Au, Ag, Pd, Pt, Cu, Ni, Co, have been successfully combined with graphene. Phan et al. (in press) synthesized Pt nanoparticleloaded 3D graphene for H<sup>2</sup> sensing using a polymer-assisted hydrothermal method and the H<sup>2</sup> sensor had a response value of 16% and response/recovery times of 9/10 s with a 1% H<sup>2</sup> concentration at 200 ◦C. Semiconductor metal oxide such as SnO<sup>2</sup> (Wang et al., 2015), ZnO (Zhu and Zeng, 2017), TiO<sup>2</sup> (Zhang et al., 2018), etc. are used widely in gas detection. The hybrids of graphene and metal oxides are mainly prepared by insitu synthesis, liquid phase method and hydrothermal method. Ye et al. (2015) developed porous graphene embedded with various types of metal oxide nanoparticles through direct laser scribing on the metal-complex-containing polyimide film. Zhang B. et al. (2017) prepared rGO/α-Fe2O<sup>3</sup> hybrids with different rGO contents, which were composed of the round-edged cubic α-Fe2O<sup>3</sup> particles adhering uniformly on both sides of the crumpled and rippled rGO sheets. The local p-n heterojunctions between n-type α-Fe2O<sup>3</sup> and p-type rGO caused an extension of electron depletion layer and potential barriers, which in turn led to significant resistance variation. Zhou et al. (2017) synthesized rGO and rGO/ZnO thin films and the performances of sensing organic vapor molecules are enhanced. In order to combe the advantages of noble metal and metal oxide, the ternary graphene hybrid materials have been developed. Uddin et al. (2015a) synthesized an Ag/ZnO/rGO hybrid via photochemical method and the 5 wt% hybrid enhanced the C2H<sup>2</sup> sensing performance. Wei et al. (2017) synthesized the composite of Ag/SnO2/rGO via a hydrothermal reaction process with a high surface area of 191.583 m<sup>2</sup> /g and showed enhanced sensing properties to ethanol. Esfandiar et al. (2014) synthesized Pd/WO3/ rGO as hybrid sensing material via the facile hydrothermal method and an improvement in H<sup>2</sup> sensing at low temperatures was observed.

### Sensor Fabrication and Testing

Graphene hybrid materials can be used to fabricate gas sensors. Based on the operating mode, the device structure of gas sensor can be classified into directly heated and indirectly heated types. For a directly heated sensor, the sensing materials are in direct contact with the heater, which may make the sensor lose its stability and anti-interference ability. Therefore, the indirectly heated sensors are most widely used in scientific research and commerce, as shown in **Figure 1B**. The planar gas sensor is usually composed of three layers: sensing materials, detection electrodes and substrate (Nguyen et al., 2014). The synthesized sensing materials are covered by the detection electrodes. The interdigital electrodes are used to measure the resistance of the sensing materials. The substrate, made of silicon or alumina, has a good compatibility with integrated circuits and can support the sensing materials.

The responses of the sensor to different gases are carried out by using the gas sensing experimental platform, which composed of gas sources, MFCs (mass flow controllers), testing chamber and computer-controlled data acquisition and analysis system, as presented in **Figure 1A**. The experimental procedure is as follows: firstly, the sensor is placed in the center of the heating stage with two adjustable probes. Then, the heater starts to work and the background gas is delivered into the sealed chamber with a constant speed. When the resistance of the sensor no longer changes, the test chamber is filled with a different concentration of the gas being tested. At last, the background gas will be delivered again after the response of the sensor becomes stable. The response of sensor can be defined as a function of the change of resistance value of sensor (Barsan and Weimar, 2003). All the experiments need to be carried out under the same ambient temperature and relative humidity.

### SENSING MECHANISM

Schedin et al. reported that the graphene gas sensor can effectively detect the adsorption or dissociation behavior of a single gas molecule on the surface of graphene, due to the fact that the change of carrier concentration of graphene leads to the change in electrical conductivity (Schedin et al., 2007). Meanwhile, due to the existence of oxygen functional groups, GO and rGO usually show better sensing properties than graphene. When GO or rGO are exposed to the air, the oxygen functional groups of them mainly combine H2O molecules with hydrogen bonds and the adsorbed H2O molecules are transferred to H3O+, further promoting the formation of ion channels on the surface of the sample. When they are in contact with adsorbed molecules, the hydrogen bonds will be destroyed, which inhibits the ionization reaction between oxygen functional groups and H2O molecules and leads to step-like changes in resistance (Ozcan et al., 2018). Prezioso et al. (2013) prepared a p-type gas sensor through GO drop-cast on standard interdigitated Pt electrodes and its sensing properties to NO<sup>2</sup> was analyzed. They also presented the sensing mechanism of the gas sensor: when NO<sup>2</sup> molecules adsorbed on the oxygen functional groups, the electrons of the adsorption sites are transferred to NO<sup>2</sup> molecules, which leads to the decrease of electron concentration in the surface of the sensing materials, giving a reason for the p-type behavior.

In general, the sensing mechanisms of graphene hybrid are analyzed from the following three aspects. First, the introduction of nanoparticles can effectively prevent the aggregation of graphene sheets, thereby the graphene hybrid material is more favorable to form a 3D porous nanostructure with higher specific surface area, and more adsorption sites, vacancies, defects, and sp<sup>2</sup> -bonded carbon, which are beneficial to the adsorption of gas molecules (Russo et al., 2012; Zhang et al., 2015a). Secondly, the formation of p-n heterojunctions between graphene and metal/metal oxide enhances the gas sensing properties. Once the target gas molecules are in contact with these interfaces, the depletion layers at the heterojunctions will be modulated, the electron state will be changed, and the phenomenon of charge transfer is more active, which lead to a larger relative change of resistance of graphene hybrid material (Tran et al., 2014). Thirdly, when hybrid materials contain metal oxide (e.g., SnO2, ZnO, CuO, Co3O4), the sensing behavior can be explained by the surface-adsorbed oxygen (Bai et al., 2015). For n-type metal oxide material, the oxygen molecules O2(gas) will capture electrons from the surface of metal oxide to form chemisorbed oxygen species (O<sup>−</sup> 2 , O−, or O2−), which leads to a high resistance of the sensor, as observed in the experiments (Jin et al., 2016). As shown in **Figure 1C**, when the sensor is exposed to reducing gas such as methane, the target gas molecules will react with chemisorbed oxygen species and obtain electrons from them, which reduces the concentration of electron on the surface of the sensing materials. Obviously, the gas sensing reaction is a ticklish issue associated with the intricate nanostructure and complicated sensing mechanism (Wang et al., 2011; Sun et al., 2015; Wang Z. et al., 2016). The research on the characterization methods and simulation analysis from atomic level provide a new perspective and starting point for the study on the sensing mechanism of nanomaterials.

## APPLICATION OF GAS SENSOR

Graphene hybrid materials exhibit excellent sensing properties to H2, CO, CO2, and small hydrocarbon gas (e.g., CH4, C2H2, C2H4, C2H6). In this section, we summarize and discuss the related works based on the recently published papers (**Table 1**).

### Hydrogen

The amount of H<sup>2</sup> in the insulating oil will increase significantly before the electrical fault or thermal failure occurs. Therefore, online monitoring of the H<sup>2</sup> content can ensure the security and stability of operation of oil immersed equipment. Noble metal is known for its high superior selectivity for the adsorption of H<sup>2</sup> due to its catalytic activity for H<sup>2</sup> molecules. The gas sensor based on Pd (Alfano et al., 2017) or Pt (Harley-Trochimczyk et al., 2015) nanoparticles loaded graphene demonstrated high sensitivity to H<sup>2</sup> with short response and recovery time. Sharma and Kim (2018) fabricated a MEMS H<sup>2</sup> sensor based on Pd-Ag/graphene, which showed a detection limit of 500 µL/L due to the phase transition of Pd-Ag. In addition, as metal oxide exhibits excellent gas sensing properties for H2, the graphene-metal oxide

hybrid materials are usually used to detect H<sup>2</sup> with low limit of detection (LOD) at room temperature. Wang et al. synthesized MoO<sup>3</sup> nanoribbon/graphene hybrid for H<sup>2</sup> sensing with ultralow LOD of 0.5 µL/L (Yang et al., 2017). Zhang et al. (2017c) reported a high-performance H<sup>2</sup> gas sensor based on CuO/rGO/CuO sandwiched nanostructure. Chen and co-workers successfully fabricated a device based on gasochromic-Pd/WO3/graphene/Si tandem structure with fast response and recovery time for H<sup>2</sup> (Chen et al., 2018).

### Carbon Monoxide and Carbon Dioxide

CO and CO<sup>2</sup> are mainly associated with the presence of overheating fault, which are decomposed from the cellulose of insulating oil paper. Consequently, development of highperformance CO and CO<sup>2</sup> gas sensors is an effective way to monitor the insulation performance of oil-immersed equipment. The gas sensors based on rGO (Hafiz et al., 2014; Panda et al., 2016; Wu et al., 2016) can lower the operating temperature to room temperature. The graphene metal oxide hybrid materials such as NiO/graphene (Khaleed et al., 2017), CuO/rGO (Zhang et al., 2017a) ZnO (Ha et al., 2018), can detect CO with low LOD (as low as 1 µL/L), and quick response and recovery time (9 s/10 s). Balamurugan et al. (2016) demonstrated a selective CO sensor based on rGO decorated mesoporous hierarchical GaInO3, which exhibited a high response rate of 48% to 20 µL/L CO and appreciably fast response (∼14 s) and recovery (∼15 s) at 90◦C. Shojaee and co-workers synthesized Pd-loaded SnO2/rGO hybrid materials by a hydrothermal method for CO sensing application (Shojaee et al., 2018). Additionally, much research on CO<sup>2</sup> sensor based on graphene hybrid materials have been carried out by Nemade's group. They have fabricated sensors with excellent stability, low LOD and short response and recovery time within 30 s based on Sb2O<sup>3</sup> quantum dots (QDs)/graphene (Nemade and Waghuley, 2014b), Al2O<sup>3</sup> QDs/graphene (Nemade and Waghuley, 2014a), and Y2O<sup>3</sup> QDs/graphene (Nemade and Waghuley, 2013).

## Hydrocarbon Gas

The hydrocarbon gas (CH4, C2H2, C2H4, and C2H6) may be generated when electrical or thermal failures occur, such as oil or high-temperature overheating, partial discharge, spark discharge, or arc discharge. These four hydrocarbon gases have a small difference in molecular structure and chemical composition, and the sensing materials usually have the similar response to them. According to the relationship between principle gases and associated fault types (Bakar et al., 2014), when the oil-immersed equipment is suffering an electric or heating fault under any condition, the produced hydrocarbon fault characteristics gas


TABLE 1 | Summary of recent researches on graphene hybrid materials sensor for sensing of fault characteristic gases in oil-immersed equipment.

G, graphene; rGO, reduced graphene oxide; PANI, polyaniline; RT, room temperature; 1I, which is calculated as the current change of gas sensitive response; 1G=|Gair - Ggas|, where Ggas is the conductance exposure to target gas concentration and Gair is the conductance exposure to air or nitrogen; 1R=|Rair - Rgas|, where Rgas is the resistance exposure to target gas concentration and Rair is the resistance exposure to air or nitrogen.

is mainly CH<sup>4</sup> or C2H2, and the amount of each of them is larger than high molecular weight gases (C2H<sup>4</sup> and C2H6) (Sun et al., 2017). Moreover, the detection of CH<sup>4</sup> or C2H<sup>2</sup> is becoming more and more important in many fields as the mining industry, environment monitoring and petrochemical industry. Therefore, the research on the detection of hydrocarbon gas in oil-immersed equipment is mainly focused on CH<sup>4</sup> and C2H2. Wu et al. reported a CH<sup>4</sup> sensor based on polyaniline (PANI)/graphene. Due to the existence of the π-π ∗ conjugation system within the PANI/graphene, this sensor showed a high sensitivity of 10–3,200 µL/L at room temperature (Wu et al., 2013). Zhang et al. (2016) prepared NiO/rGO hybrid materials and demonstrated a high selectivity toward CH<sup>4</sup> against H2, CO, and CO2. The response to 100 µL/L CH<sup>4</sup> was 2.2% at 260◦C. Graphene hybrid with ZnO (Zhang et al., 2015b) and SnO<sup>2</sup> (Navazani et al., 2018) also exhibited unique sensing properties to CH<sup>4</sup> around 190 and 150◦C. Nasresfahani et al. (2017) synthesized Pd-doped SnO2/rGO via hydrothermal route, and the sensing performance to 800–16,000 µL/L CH<sup>4</sup> were carried out at room temperature. Gas sensing characteristics of C2H<sup>2</sup> sensor based on SnO2/rGO were carried out for 0.5–500 µL/L C2H<sup>2</sup> at 180◦C. The p-n heterojunctions between SnO<sup>2</sup> and rGO, oxygen functional groups and high specific surface area of SnO2/rGO hybrid materials contribute to the high performance of sensor (Jin et al., 2016). Furthermore, Ag/ZnO/rGO (Uddin et al., 2015b) and Ag/SnO2/rGO (Jiang et al., 2017) hybrid materials were synthesized, and the sensors were developed for low-temperature C2H<sup>2</sup> sensing. However, with the decrease of operating temperature, the sensitivity of the sensor was reduced with longer response-recovery time.

### CONCLUSIONS

Graphene hybrids are distinctive and promising sensing materials for detection of various gases, due to their ultrahigh electron mobility and large specific surface area. Compared to other types of sensor, the graphene-based sensors exhibit excellent properties and provide a new idea for high sensitivity detection of the oil-immersed power equipment at low temperature.

Although great progress has been made in this field, there are still many challenges that need to be addressed. First, lower LOD and faster response-recovery time are needed. Graphenebased sensors provide a possible approach to detect gases at room temperature, but the reaction activities between gas molecules and sensing materials may be reduced, especially during the desorption, making it difficult to meet the engineering requirements. Ultraviolet light assisted excitation technology and novel multi-dimensional hierarchical nanostructure of graphene hybrid materials are useful ways to solve this problem. Secondly, avoiding the cross-sensitivity in the detection of mixed gas. Gas sensors usually exhibit similar responses toward different gases of different concentrations. The functionalization of graphene, operating temperature modulation and design of filter layer are workable methods to overcome this challenge. Moreover, gas sensor arrays, composed of group of sensors and signal processing algorithms, can realize the qualitative identification and quantitative detection of mixed gas. Thirdly, developing more stable materials and sensors. In practical application, the condition of detection environment, such

### REFERENCES


as vibration, temperature, humidity, and other contaminants, may affect the sensitivity of sensors. The improvement of sensing materials, coating method, and fabrication techniques can avoid or reduce the impact of these factors. Finally, the development in this field is inseparable from the research of new materials as well as the corresponding sensing mechanism.

### AUTHOR CONTRIBUTIONS

WC and LJ summarized and analyzed the related literature. LJ wrote the manuscript under the guidance of YZ and WC. All authors read and approved the manuscript.

### ACKNOWLEDGMENTS

This study was supported by the Science Fund for Creative Research Groups of the National Natural Science Foundation of China (Grant No. 51321063) and the Joint Fund of the National Natural Science Foundation of China and the Smart Grid of State Grid Corporation of China (Grant No. U1766217). LJ thanks the China Scholarship Council (CSC) project (201706050039) for financial support.


Ag-decorated tin dioxide/graphene nanocomposite film. Nanomaterials 7, 278. doi: 10.3390/nano7090278


gas sensing application. J. Mater. Sci. Mater. Electron. 26, 5937–5945. doi: 10.1007/s10854-015-3165-2


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Jin, Chen and Zhang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Recent Advances of SnO2-Based Sensors for Detecting Fault Characteristic Gases Extracted From Power Transformer Oil

Qingyan Zhang<sup>1</sup> , Qu Zhou1,2 \*, Zhaorui Lu<sup>1</sup> , Zhijie Wei <sup>1</sup> , Lingna Xu<sup>1</sup> and Yingang Gui <sup>1</sup>

<sup>1</sup> College of Engineering and Technology, Southwest University, Chongqing, China, <sup>2</sup> Electrical and Computer Engineering Department, Wayne State University, Detroit, MI, United States

Tin oxide SnO2-based gas sensors have been widely used for detecting typical fault characteristic gases extracted from power transformer oil, namely, H2, CO, CO2, CH4, C2H2, C2H4, and C2H6, due to the remarkable advantages of high sensitivity, fast response, long-term stability, and so on. Herein, we present an overview of the recent significant improvement in fabrication and application of high performance SnO2 based sensors for detecting these fault characteristic gases. Promising materials for the sensitive and selective detection of each kind of fault characteristic gas have been identified. Meanwhile, the corresponding sensing mechanisms of SnO2-based gas sensors of these fault characteristic gases are comprehensively discussed. In the final section of this review, the major challenges and promising developments in this domain are also given.

### Edited by:

Zhongchang Wang, Laboratório Ibérico Internacional de Nanotecnologia (INL), Portugal

### Reviewed by:

Yiyi Zhang, Guangxi University, China Song Xiao, Wuhan University, China

> \*Correspondence: Qu Zhou zhouqu@swu.edu.cn

### Specialty section:

This article was submitted to Nanoscience, a section of the journal Frontiers in Chemistry

Received: 22 June 2018 Accepted: 30 July 2018 Published: 29 August 2018

### Citation:

Zhang Q, Zhou Q, Lu Z, Wei Z, Xu L and Gui Y (2018) Recent Advances of SnO2-Based Sensors for Detecting Fault Characteristic Gases Extracted From Power Transformer Oil. Front. Chem. 6:364. doi: 10.3389/fchem.2018.00364

Keywords: tin oxide, gas sensors, fault characteristic gases, power transformer oil, sensing properties, sensing mechanism

## INTRODUCTION

Power transformers are one of the most important apparatuses in power systems, and their reliability is extremely vital to ensure the stable system operation (Zheng et al., 2018). Thermal and electrical faults in oil-filled power transformers may produce typical fault characteristic gases including hydrogen H<sup>2</sup> (David et al., 2018), carbon monoxide CO (Joseph, 1980; Uddin et al., 2016; Zhou et al., 2018b), carbon dioxide CO<sup>2</sup> (2015; Dan et al., 2016; Iwata et al., 2017; Zhang et al., 2017), methane CH<sup>4</sup> (Sedghi et al., 2010), acetylene C2H<sup>2</sup> (Qi et al., 2008), ethylene C2H4, and ethane C2H6. These typical fault characteristic gases could be dissolved in transformer oil or accumulate as free gases if produced rapidly in large quantities. Therefore, detection and analysis of the species, quantities and generation rates of these fault gases presented in the fluid allow for the identification of power transformer fault types such as corona, sparking, overheating, and arcing.

During the past few decades, dissolved gas analysis has been developed to detect the latent faults of oil-immersed power transformers (Morais and Rolim, 2006; Zhan et al., 2015; Uddin et al., 2016; Gong et al., 2018). Gas sensing detection technology is the core of dissolved gas analysis. Different types of sensing technologies have been reported in previous studies for detecting typical fault characteristic gases extracted from transformer oil, such as metal oxide semiconductor gas sensors (Zhu and Zeng, 2017a; Chen et al., 2018), catalytic combustion sensors (Liu et al., 2011), fuel cell sensors (Modjtahedi et al., 2016; Tonezzer et al., 2017), and optical sensors (Trieu-Vuong et al., 2016; Paliwal et al., 2017). Given the remarkable advantages of simple fabrication process, low maintenance cost, fast response and recovery, long service life, and so on, metal oxide semiconductor materials like SnO<sup>2</sup> (Choi et al., 2014; Zeng et al., 2014), ZnO (Zhou et al., 2017; Zhu and Zeng, 2017b; Zhu et al., 2017), TiO<sup>2</sup> (Zeng and Liu, 2010; Zhang et al., 2018b) and In2O<sup>3</sup> (Cao et al., 2015) have received scientific and technological importance for many years and are widely used to detect these gases. Among them, as the most widely used gas-sensitive material, SnO<sup>2</sup> is currently the main sensing materials used in experimental research and commercial application for detecting typical fault characteristic gases extracted from power transformer oil (Zhang et al., 2010; Zhou et al., 2018a).

Herein, the first section of this review will briefly outline the preparation of the currently used SnO<sup>2</sup> sensing materials, the microstructure morphology, and doping modification of SnO<sup>2</sup> sensing materials for detecting typical fault characteristic gases extracted from power transformer oil. The second section addresses the main gas sensing mechanisms of SnO<sup>2</sup> gas sensors for these gases. The third section presents the conclusions, focusing on future challenges and potentialities associated with SnO2-based gas sensors for detecting these typical fault characteristic gases.

# SENSING PERFORMANCES OF SNO2-BASED SENSORS TO FAULT CHARACTERISTIC GASES EXTRACTED FROM POWER TRANSFORMER OIL

For detecting and analyzing typical fault characteristic gases extracted from power transformer oil with SnO2-based gas sensors, the most important concerns are the sensor sensitivity, selectivity, and repeatability (Fan et al., 2017). Combined with the research concerns mentioned above, this section briefly summarizes recent progress in the application of SnO<sup>2</sup> sensing materials to detect typical fault characteristic gases dissolved in transformer oil. The gas sensing performance of some modified SnO<sup>2</sup> gas sensors are listed in **Table 1**.

Toper, temperature at which the best sensor performance, usually in terms of the highest response toward target gas, could be obtained.

S, gas sensing response of a SnO2-based sensor to target gas, which is defined as S =Ra/Rg for reducing gas and S = Rg/Ra for oxidizing gas (where Ra and Rg are the resistance of the sensor in air and in the test gas, respectively).

It is known that the metal oxide semiconductor SnO<sup>2</sup> is a promising material for sensing the typical fault characteristic gases extracted from power transformer oil. Nevertheless, the sensing performance of the proposed sensors considerably depends on the preparation method and surface structure of the sensing materials and the dopants. The structure of modern SnO<sup>2</sup> gas sensors could be classified into two main types including thick/thin films and nanoparticles. The preparation methods of the sensing materials mainly include the hydrothermal method, the sol-gel method, the electrospinning technique, chemical vapor deposition and so on. Different preparation methods could affect the morphology of the sensing materials and further change their gas sensing properties (Long et al., 2018; Zhang et al., 2018a; Zhou et al., 2018c). Noble metals or metal oxide doped on SnO2 based sensors can play an important role in accelerating the sensing process and improving the gas sensor performances.

In power transformers, H<sup>2</sup> is generated from the thermal decomposition of oil at high temperatures, which is a serious problem for transformer oil (Uddin et al., 2016). Various highefficacy and workable SnO2-based sensors have been recently introduced for detecting H2. It can be seen from the data in **Table 1** that various metal particles such as Co (Liu et al., 2010), Au (Yin and Tao, 2017), and Pd (Nguyen et al., 2017) have been added to SnO2-based sensors to enhance H<sup>2</sup> gas sensing performance. Among them, the Au-loaded SnO<sup>2</sup> sensor can detect H<sup>2</sup> down to 1 ppm, which is a good property for detecting low concentrations of H<sup>2</sup> dissolved in transformer oil (Yin and Tao, 2017). K. Inyawilert reported that a SnO<sup>2</sup> sensing film with an optimal Rh-doping level of 0.2 wt% exhibited an ultra-high response of 22,170 and a short response time of 6 s toward 30,000 ppm H<sup>2</sup> at an optimum operating temperature of 300◦C. In addition, the proposed Rh-doped SnO<sup>2</sup> sensor displayed good H<sup>2</sup> selectivity against NO2, SO2, C2H4, C3H6O, CH4, H2S, and CO (Inyawilert et al., 2017). Except for nanoparticles embedded in the metal oxide matrix, the morphology of the SnO<sup>2</sup> materials could be applied to improve the H<sup>2</sup> sensing properties, and low dimensional nanostructures have attracted increasing attention. Nguyen Kien et al. introduced the on-chip growth of SnO<sup>2</sup> nanowire-based gas sensors to detect low concentrations of H<sup>2</sup> gas, which responded with 2.6 for 20 ppm H<sup>2</sup> gas at 400◦C (Nguyen et al., 2017). Taken together, these findings suggest that the Au-loaded SnO<sup>2</sup> gas sensor is the most promising candidate for detecting low concentration H<sup>2</sup> extracted from power transformer oil.

The insulating paper, pressboard and wood blocks in power transformers contain a large number of anhydroglucose rings as well as weak C-O molecular bonds and glycosidic bonds that are thermally less stable than the hydrocarbon bonds in oil. These bonds could decompose to CO and CO<sup>2</sup> at lower temperatures when potential faults occur. As shown in **Table 1**, Pt-doped SnO<sup>2</sup> thick film (Chen et al., 2018), ZnO-SnO<sup>2</sup> nanoparticles (Chen et al., 2015), and Au-doped SnO<sup>2</sup> yolk-shell nanospheres (Wang et al., 2013) could be potential candidates for CO sensors. Among them, Wang et al. compared the gas sensing properties of Au@SnO<sup>2</sup> yolk-shell nanospheres with that of hollow SnO<sup>2</sup> nanospheres. The sensor fabricated with the Au@SnO<sup>2</sup> yolkshell nanospheres showed lower operating temperature (210◦C), lower detection limit (5 ppm), faster response (0.3 s), and better selectivity to CO gas (Wang et al., 2013), which could be attributed to the catalytic effect of Au and enhanced electron depletion at the surface of the Au@SnO<sup>2</sup> yolk–shell nanospheres. Moreover, SnO<sup>2</sup> loaded with Pd nanoparticles may be another kind of promising material for CO gas sensing (Zhou et al., 2018d). Yuasa et al. prepared PdO-loaded SnO<sup>2</sup> nanoparticles by the reverse micelle method and reported that the 0.1 mol% PdOloaded SnO<sup>2</sup> sensor exhibited a high gas response value of 320


TABLE 1 | Comparison of the representative SnO2 based sensors for fault characteristic gases extracted from power transformer oil.

to 200 ppm CO gas (Yuasa et al., 2009). Chen et al. synthesized Pd-doped SnO<sup>2</sup> nanoparticles using a co-precipitation method, and the 1.5 wt% PdO decorated SnO<sup>2</sup> presented the largest gassensitive response of 6.59 at 260◦C in 400 ppm CO atmosphere (Chen et al., 2018). Another study reported that Pd-modified nanocrystalline SnO<sup>2</sup> displayed a fairly high and reversible CO response (2–100 ppm) at room temperature (Marikutsa et al., 2010). Yin et al. prepared Pd-loaded and Fe-doped SnO<sup>2</sup> by the sol-gel method. The composite with 10 mol% Fe and 0.2 mol% Pd had the highest sensitivity and selectivity to CO in the range of 200–3,000 ppm at 350◦C, and the response value to 2,000 ppm CO was increased 13 times compared with pure SnO<sup>2</sup> (Yin and Guo, 2014). Therefore, Pd is a highly effective catalyst for improving the sensing performance of SnO2-based sensors to CO gas (Xiong et al., 2017).

Due to their high chemical stability, conventional binary metal oxides have very low sensitivity to chemically inert gases such as CO<sup>2</sup> (Korotcenkov and Cho, 2017). However, it was reported that La-doped SnO<sup>2</sup> nanocomposites can be used for CO<sup>2</sup> sensing (Kim et al., 2000). And the 8% LaOCl-SnO<sup>2</sup> nanofibers exhibited an optimal response of 3.7 toward 1,000 ppm CO<sup>2</sup> at 300◦C with response/recovery times of 24/92 s (Xiong et al., 2017). Karthik et al. reported the sensing properties of tin oxide (SnO2) and zinc oxide (ZnO) thin films deposited onto macroporous silicon (PS) substrates to CO<sup>2</sup> and found that the obtained SnO2/PS films showed the highest sensing response of 19 to 15 ppm CO<sup>2</sup> gas (Karthik et al., 2018).

For hydrocarbons, namely, CH4, C2H2, C2H4, and C2H6, when high energy discharge occurs, such as arcing, in transformer oil, the content of C2H<sup>2</sup> is relatively high. At low temperature thermal faults (T < 300◦C), the contents of CH<sup>4</sup> and C2H<sup>6</sup> tend to be high. C2H<sup>4</sup> is the main content of hydrocarbon while at high temperature thermal faults (T > 300◦C) (Fan et al., 2017). For detecting these hydrocarbons, an online monitoring system based on a SnO2-based gas chromatographic detector for assessing the running condition of a power transformer was developed (Fan et al., 2017). Qi et al. fabricated a C2H<sup>2</sup> sensor based on 6 wt% Sm2O3-doped SnO2, whose gas response to 1,000 ppm C2H<sup>2</sup> could reach 63.8, 16.8 times larger than that of pure SnO<sup>2</sup> (Qi et al., 2008). Moreover, Jin et al. reported that reduced graphene oxide (rGO) loaded SnO<sup>2</sup> hybrid nanocomposite showed high sensor response (12.4 toward 50 ppm), fast response-recovery time (54 and 23 s), low detection limit (1.3 ppm), good linearity, excellent selectivity and long-term stability to C2H<sup>2</sup> (Jin et al., 2017). These results indicated that rGO would be an effective addition to enhance the sensing properties of SnO2-based sensors to C2H<sup>2</sup> and make a contribution to developing a ppm-level gas sensor for on-line monitoring of C2H<sup>2</sup> gas extracted from transformer oil.

Sensors for CH<sup>4</sup> detection have also been widely studied and are partly summarized in **Table 1**. The 20 mol% Pt-SnO<sup>2</sup> nanofibers exhibited excellent CH<sup>4</sup> sensing properties over a temperature range of 100–350◦C, and an obvious response of 1.11 to 1 ppm CH<sup>4</sup> was measured at 350◦C (Lu et al., 2018). Koeck et al. found a device prepared by ultra-long single crystalline SnO2-nanowires, which was able to detect a few ppm of CO and CH<sup>4</sup> at the operating temperature of 200– 250◦C (Koeck et al., 2009). Chen et al. successfully synthesized Co-doped SnO<sup>2</sup> nanofibers via an electrospinning method and reported that the Co-doped SnO<sup>2</sup> nanofiber sensor exhibited a high response of 30.28 toward 50 ppm CH<sup>4</sup> at 300◦C (Chen et al., 2013b). For C2H<sup>4</sup> and C2H<sup>6</sup> detection, only a few sensors including SnO<sup>2</sup> thin films and Pd-doped SnO<sup>2</sup> nanoparticles are found to be effective. Jadsadapattarakul et al. reported that the sensing response, response and recovery time of SnO<sup>2</sup> thin film sensors for selective detecting of C2H<sup>4</sup> gas could be improved by coating a layer of [010] highly preferred-orientation silicalite-1 polycrystals (Jadsadapattarakul et al., 2010). Ahn et al. obtained SnO<sup>2</sup> thin films by R.F. magnetron sputtering to fabricate high performance C2H<sup>4</sup> gas sensors (Ahn et al., 2010). Chen et al. discovered that the 5 wt% Pd-doped SnO<sup>2</sup> sensor could detect C2H<sup>6</sup> at 400◦C, and the sensor exhibited the largest gas-sensitive response of 5.89 toward 100 ppm C2H<sup>6</sup> (Chen et al., 2013a).

## SENSING MECHANISMS OF SNO2-BASED SENSORS TO FAULT CHARACTERISTIC GASES EXTRACTED FROM POWER TRANSFORMER OIL

It is well agreed that the sensing mechanism of SnO<sup>2</sup> gas sensors is the change in conductivity of the metal oxide layer caused by the interaction with the surrounding atmosphere (Ducéré et al., 2012; Zeng et al., 2013; Korotcenkov and Cho, 2017). When exposed to air, oxygen molecules would be adsorbed on the surface of the SnO<sup>2</sup> nanostructures and capture electrons from the conduction band of SnO<sup>2</sup> to generate chemisorbed oxygen species (O−, O2−, and O<sup>−</sup> 2 , depending on temperatures) (Shahabuddin et al., 2017). At low temperatures (below 150◦C), oxygen molecules exist in the form of molecular ions O<sup>−</sup> 2 , which would change to atomic ions O<sup>−</sup> (150–400◦C) and O2<sup>−</sup> (more than 400◦C) as temperatures rising (Punginsang et al., 2017). The chemical adsorption process can be explained by the following reactions:

$$O\_{2(\text{gas})} \Leftrightarrow O\_{2(\text{abs})} \tag{1}$$

$$\rm O\_{2(ads)} + e^- \Leftrightarrow O\_{2}^{\cdot -} \tag{2}$$

$$2O\_2^{-} \text{ (abs)} + e^- \Leftrightarrow 2O^- \text{ (abs)}\tag{3}$$

$$O^{-}\text{ (abs)} + e^{-} \Leftrightarrow O^{2-}\text{ (abs)}\tag{4}$$

As the electrons transfer from the conduction band of SnO<sup>2</sup> to the chemisorbed oxygen, the electron concentration and electrical conductivity of the SnO<sup>2</sup> film decrease. When the SnO<sup>2</sup> film is exposed to typical fault characteristic gases, the reducing gas would react with the chemisorbed oxygen species, thereby releasing electrons back to the conduction band with increasing electrical conductivity. The sensing mechanisms of the SnO<sup>2</sup> sensor sensing these fault gases can be explained from the following reaction paths, where O<sup>−</sup> is taken as an example (Samerjai et al., 2012).

$$\text{H}\_2 + \text{O}^-\text{ (ads)} \rightarrow \text{H}\_2\text{O} + \text{e}^- \tag{5}$$

$$\text{CO} + \text{O}^-\text{(ads)} \rightarrow \text{CO}\_2 + \text{e}^-\tag{6}$$

$$\text{C}\_2\text{H}\_2 + \text{O}^-\text{(ads)} \rightarrow 2\text{C} + \text{H}\_2\text{O} + \text{e}^- + \text{heat} \tag{7}$$

$$2\text{ CH}\_4 + 2\text{O}^-\text{(ads)} \rightarrow 2\text{H}\_2\text{O} + \text{CO}\_2 + 2\text{e}^-\tag{8}$$

$$\text{C}\_2\text{H}\_4 + 2\text{O}^-\text{(ads)} \rightarrow 2\text{H}\_2\text{O} + 2\text{C} + 2\text{e}^-\tag{9}$$

$$\text{C}\_2\text{H}\_6 + 3\text{O}^-\text{(ads)} \rightarrow 3\text{H}\_2\text{O} + 2\text{C} + 3\text{e}^-\tag{10}$$

Dopants added to SnO2-based gas sensors can accelerate the reaction process mentioned above and improve the sensing performance of gas sensors. The doped catalysts can enhance the sensing performance of a sensor in two ways, namely, chemical sensitization and electronic sensitization (Lin et al., 2017). The high surface area structures such as porous structures and hollow structures could accelerate gas diffusion on the material surface and then shorten the response time.

On the other hand, to understand the sensing mechanism of SnO2-based materials at the atomic and quantum levels, based on the framework of Density Function Theory (DFT), Li et al. established a SnO<sup>2</sup> surface model, gas molecular models and adsorption models. A first principles calculation using the Cambridge Sequential Total Energy Package (CASTEP) program was performed, and the total energies, electronic structures and adsorption properties were investigated in detail. Theoretical calculations provided a qualitative explanation of the sensing properties of the fabricated SnO2-based sensor to various fault characteristic gases (CH4, C2H6, C2H4, and C2H2) extracted from power transformers (Li et al., 2017). Zeng et al. performed a first principles calculation to investigate how H<sup>2</sup> gas interacts with the SnO<sup>2</sup> (110) surface and the effect of metallic ions on the gas response of SnO2. Based on the theoretical calculation, it was reported that the Pd-doped SnO<sup>2</sup> (110) surface could adsorb more H<sup>2</sup> gas and receive larger electrons from adsorbed H<sup>2</sup> molecules (Zeng et al., 2011). Other research involving a possible CO sensing mechanism for Pd-doped SnO<sup>2</sup> sensors was investigated with first principles calculations (Chen et al., 2018), and the theoretical results demonstrated that CO molecules can grab O from the pre-adsorbed oxygen on the Pd<sup>4</sup> cluster or the PdO cluster on the SnO<sup>2</sup> (110) surface. These processes may play an important role in CO sensing for Pd-doped SnO<sup>2</sup> (Chen et al., 2018). In particular, theoretical calculations indicated that CO<sup>2</sup> molecules cannot be adsorbed onto the stoichiometric SnO<sup>2</sup> (110) surface or SnO<sup>2</sup> (110) surface pre-adsorbed by O<sup>−</sup> 2 and O<sup>−</sup> in dry air. However in wet air, CO<sup>2</sup> could react with O of pre-adsorbed OH−, bringing about the formation of carbonates containing (CO3) <sup>2</sup><sup>−</sup> and the dissociation/movement of surface OH<sup>−</sup> groups, accompanying the release of electrons from CO<sup>2</sup> to the SnO<sup>2</sup> surface (Wang et al., 2016).

### CONCLUSION AND PERSPECTIVE

In this mini-review, SnO2-based gas sensors for detecting typical fault characteristic gases extracted from power transformer oil have been briefly summarized. The analysis shows that the detection performances of SnO2-based gas sensors can be obviously enhanced with dopant addition or increasing the surface area of the sensing materials. Despite achieving good results, there is still much room for further development. First of all, due to the cross-sensitivity between these fault characteristic gases, selectivity is a large challenge for developing high performance SnO2-based sensors to detect these gases independently. In the future, to promote engineering

## REFERENCES


applications, a combination of SnO2-based gas sensors and gas chromatography should be further studied for multi-component detection of these gases. For further development of the gassensing performances, the high surface area structures such as hollow and hierarchical nanostructures could be prepared to obtain more active sites for gas diffusion, and different types of dopant elements should be examined. Furthermore, the gas sensing mechanisms are still imperfect and controversial, which cannot effectively guide development of novel SnO2-based sensors and limit the prevailing application of these sensors to fault characteristic gases extracted from power transformer oil. Further research should focus on determining a satisfying mechanism model of SnO2-based sensors to provide a guide for future work.

### AUTHOR CONTRIBUTIONS

All authors listed have made substantial, direct and intellectual contributions to the work and approved it for publication.

### ACKNOWLEDGMENTS

This work has been supported in part by the National Natural Science Foundation of China (Nos. 51507144, 5127785), the China Postdoctoral Science Foundation Project (Nos. 2015M580771, 2016T90832), and the Chongqing Science and Technology Commission (CSTC) (No. cstc2016jcyjA0400).

by interface of n-n heterostructure of CeO2-SnO<sup>2</sup> nanoparticles. Sens. Actuat. B 254, 984–995. doi: 10.1016/j.snb.2017.07.093


of dissolved gas contents in power transformers. Electr. Power Syst. Res. 155, 196–205. doi: 10.1016/j.epsr.2017.10.010


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Zhang, Zhou, Lu, Wei, Xu and Gui. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# SnO2/Graphene Nanoplatelet Nanocomposites: Solid-State Method Synthesis With High Ethanol Gas-Sensing Performance

Run Zhang1,2, Jian-Bo Jia1,2 \*, Jian-Liang Cao1,2 and Yan Wang1,3 \*

*<sup>1</sup> The Collaboration Innovation Center of Coal Safety Production of Henan Province, Henan Polytechnic University, Jiaozuo, China, <sup>2</sup> Department of Chemical Engineering and Technology, College of Chemistry and Chemical Engineering, Henan Polytechnic University, Jiaozuo, China, <sup>3</sup> School of Safety Science and Engineering, State Key Laboratory Cultivation Base for Gas Geology and Gas Control, Henan Polytechnic University, Jiaozuo, China*

Keywords: Graphene nanoplatelet, SnO2 nanoparticles, nanocomposites, ethanol, sensitivity

## INTRODUCTION

Edited by: *Wen Zeng, Chongqing University, China*

### Reviewed by:

*Akihiro Kushima, University of Central Florida, United States Arshad Saleem Bhatti, COMSATS Institute of Information Technology, Pakistan*

### \*Correspondence:

*Jian-Bo Jia jiajianbo@hpu.edu.cn Yan Wang yanwang@hpu.edu.cn*

### Specialty section:

*This article was submitted to Nanoscience, a section of the journal Frontiers in Chemistry*

Received: *29 May 2018* Accepted: *19 July 2018* Published: *09 August 2018*

### Citation:

*Zhang R, Jia J-B, Cao J-L and Wang Y (2018) SnO2/Graphene Nanoplatelet Nanocomposites: Solid-State Method Synthesis With High Ethanol Gas-Sensing Performance. Front. Chem. 6:337. doi: 10.3389/fchem.2018.00337* Graphene nanoplatelet (GNP) is a well-known 2D carbon nanomaterial composed of network structure of sp<sup>2</sup> -hybridized carbon atoms. So far, two main strategies (exfoliation and chemical oxidation-reduction) have been used for the preparation of GNP from graphite. Unfortunately, a lot of defects are caused by the strong oxidizing reagents on the GNP produced by chemical oxidationreduction method, and such GNP lost the remarkable electrical and mechanical properties (Coleman, 2013). Exfoliation method could be adopted to product high-quality GNP from graphite, but the yield of GNP is quite low (∼1 wt.%) (Hernandez et al., 2008). Hence, the requisite large scale production of high-quality GNP remains a challenging task.

Even so, GNP is expected as a promising material for gas sensing due to its unrivaled physiochemical and electronic properties such as excellent flexibility, large specific surface area and high conductivity. Nevertheless, owing to the poor gas-sensing selectivity, GNP exhibited responses to several kinds of gas (Yoon et al., 2012; Nemade and Waghuley, 2013). But when GNP was incorporated with other sensor materials, like metal oxide semiconductors (MOSs), it could remarkably improve the sensing performance of the sensor materials (Eom et al., 2017; Thu et al., 2018). The MOSs phase facilitates the adsorption/desorption process of tested gas, thereby activating the reactions occurring on the carbon surface, which in turn increases the response speed and response/recovery time. In addition, n-p junctions can be formed by GNP with n-type metal oxides, and the resulting novel nanostructures perform much better gas sensing performance than single materials (Neri et al., 2013).

As an n-type metal oxide semiconductor, tin dioxide (SnO2) has a wide band gap of Eg = 3.6 eV and excellent optical and electrical properties. SnO<sup>2</sup> have been one of the most extensive studied materials due to its wide applications including in transparent conductive electrodes and transistors (Liu et al., 2018; Satoh et al., 2018), lithium-ion batteries (Zhao et al., 2016; Shi et al., 2017), dye-sensitized solar cells (Hagfeldt et al., 2010), photocatalysis (Aslam et al., 2018; Praus et al., 2018) and gas sensors (Narjinary et al., 2017; Long et al., 2018; Xu et al., 2018). For gas sensing application, SnO<sup>2</sup> and SnO<sup>2</sup> based composites also show admirable gas sensing properties like lowcost, low detection limit, fast response and recovery, high response and good stability (Yan et al., 2015; Cao et al., 2017; Kim et al., 2017).

Herein, we put forward a simple and potentially scalable method to obtaining massive high-quality GNP from exfoliation of flake graphite in K2FeO4/H2SO4, and use the solid-state method to synthesize SnO<sup>2</sup> decorated graphene nanoplatelet nanocomposites

**79**

(SnO2/GNP) with different mass ratio of SnO<sup>2</sup> and GNP. The as-prepared SnO2/GNP nanocomposites possess the twodimensional (2D) structure, and the 2D GNP accelerating the preferential growth and preventing the agglomeration of the SnO<sup>2</sup> nanoparticles. The gas sensing tests indicated that the sensors based on GNP/SnO<sup>2</sup> nanocomposites possess high sensitivity and excellent characteristic of response and recovery toward ethanol vapor. The sensor response was found to be dependent on the mass ratio of GNP in the composites and it reaches the maximum response when the mass percentage of GNP in the composites is 5%.

## EXPERIMENTAL

### Preparation of the GNP/SnO<sup>2</sup> Nanocomposites

All the reagents were of analytical grade (AR) and used as received without further purification. The preparation of GNP was listed in the **Supplementary Material**. A typical synthesis process of GNP/SnO<sup>2</sup> nanocomposites with 5 wt.% GNP content can be described as follows: 7 g of SnCl4·5H2O, 0.15 g of GNP and 6 ml of PEG-400 were mixed adequately and ground together in an agate mortar. Subsequently, 3.2 g of NaOH was slowly added to the mixtures and ground together for about 30 min. The reaction started readily during the addition process of NaOH, accompanied by release of heat. As the reaction proceeded, the mixture became mushy. Then samples were collected, washed several times with distilled water and absolute ethanol, and dried at 60◦C overnight in a drying oven. Finally, the product was ground to powder, marked as GNP/SnO2-5. 2.5 and 7.5 wt.% GNP of GNP/SnO<sup>2</sup> nanocomposites were prepared using the aforementioned method, and marked as GNP/SnO2- 2.5 and GNP/SnO2-7.5, respectively. For comparison, the same method was used to synthesize SnO<sup>2</sup> nanoparticles without GNP.

The characterization, sensor fabrication and measurement (**Figure S1**) were listed in the Supplementary Material.

## RESULTS AND DISCUSSION

The XRD diffraction pattern of GNP is almost identical to that of pristine graphite (**Figure 1A**), revealing that no structural change occurred during the exfoliating process. The intensity of (002) peak centered at 26.5◦ of GNP decreases obviously compared with that of pristine graphite due to the ultrathin thickness of GNP (Zhang et al., 2016). The XRD diffraction patterns of pure SnO<sup>2</sup> nanoparticles and GNP/SnO<sup>2</sup> nanocomposites are shown in **Figure 1B**. We can see that four distinct diffraction peaks of SnO<sup>2</sup> centered at 2θ of 26.6◦ , 33.9◦ , 51.7◦ , and 65.9◦ , which are corresponding to the reflection from the (110), (101), (211), and (301) planes of the tetragonal rutile SnO<sup>2</sup> (JCPDS Card No. 41- 1445), respectively. This confirmed that the synthesis method that SnO<sup>2</sup> was successfully prepared by solid-state reaction is feasible and complete. However, as seen from **Figure 1B**, there are no diffraction peak around 26.6◦ of SnO<sup>2</sup> observed in the curves, because the diffraction peaks of 26.5◦ of GNP is so high that the peak around 26.6◦ of SnO<sup>2</sup> is covered.

FIGURE 1 | (A) XRD patterns of pristine graphite and GNP. (B) XRD patterns of SnO2 and the SnO2/GNP nanocomposites with different GNP contents. (C,D) TEM and HRTEM images of the SnO2/GNP-5 nanocomposite. (E) TG-DSC profiles of the SnO2/GNP-5 nanocomposites. (F) The pore size distribution curves of the SnO2 and SnO2/GNP-5 nanocomposite.

**Figure S2a** shows the representative FESEM image of pristine graphite. In **Figure S2b** (FESEM image of GNP), twodimensional (2D) structure of the thin layers can be seen clearly. As shown in **Figure S2c**, the FESEM image of the pure SnO<sup>2</sup> exhibits particles with the size of 100–200 nm. The FESEM and TEM images of the GNP/SnO2-5 nanocomposite are presented in **Figure S2d** and **Figure 1C**, respectively, and which show that numerous particles are dispersed on the surface of 2D sheets of GNP. Meanwhile, as can be seen from **Figure 1D**, two phases of GNP and SnO<sup>2</sup> are clearly observed and closely in contact to form an intimate interface. And, the lattice fringes with interplanar spacings of 0.26 nm and 0.34 nm can be corresponding to the (101) and (110) planes of SnO<sup>2</sup> nanoparticles. It can be concluded that the GNP/SnO<sup>2</sup> composites were synthesized successfully using the solid-state method.

TG-DSC analysis revealed the weight change situation of GNP/SnO2-5 nanocomposites from room temperature to 800◦C with the heating rate of 10◦ ·min−<sup>1</sup> . As is shown in **Figure 1E**, there are two stages of weight loss in the TG curve according to the peaks of DSC curve. The first stage in temperature before 300◦C is due to desorption of moisture and solvent. The second stage of weight loss is due to the combustion of GNP in air. This result proves that the GNP/SnO2-5 nanocomposite was not decomposed at the operating temperature of 280◦C in the procedure of measuring gas-sensing properties.

**Figure 1F** displays the pore diameter distribution of the SnO<sup>2</sup> and GNP/SnO2-5 samples. It can be clearly seen that the pore diameters of pure SnO<sup>2</sup> and GNP/SnO2-5 are relatively small, which both the majority concentrate on about 2 nm and 4 nm. The specific surface areas of GNP/SnO2-5 sample is 167.01 m2 ·g −1 , which is higher than SnO<sup>2</sup> (119.67 m<sup>2</sup> ·g −1 ). Increasing specific surface area could be in favor of enhancing gas-sensing properties.

**Figure 2A** shows the response values of pure SnO<sup>2</sup> nanoparticles-based sensor and GNP/SnO2-based sensors to 500 ppm of ethanol at different temperatures. From the curves of GNP/SnO2-2.5, GNP/SnO2-5, and GNP/SnO2-7.5, it can be clearly observed that the response values increased with the increase of the temperature. However, the response values decrease when the temperature is above 280◦C. As a result, the best operating temperature of GNP/SnO2-based sensors is 280◦C. Similarly, the best operating temperature of pure SnO<sup>2</sup> sensors is 300◦C. We can get a conclusion that the best operating temperature is lowered 20◦C because of the joining of GNP. Compared between the different curves, it reaches the maximum response when the mass percentage of GNP in the composites is 5%. The response value of GNP/SnO2-7.5 sample is lower than that of the GNP/SnO2-5 sample. It is because that activation center still focuses on the SnO<sup>2</sup> nanoparticle, and the high content of GNP may lead to the decrease of SnO<sup>2</sup> nanoparticle on the unit specific surface area. Some SnO2-based materials of ethanol sensing from the literature are summarized in **Table S1**. It can be observed that the GNP/SnO<sup>2</sup> composite exhibits superior performances compared with other SnO2-based materials.

**Figure S3** displays the response values of sensors based on pure SnO<sup>2</sup> and GNP/SnO2-5 to different concentrations of ethanol at 280◦C. As shown in the curves, the response values of the two sensors increased with the increasing of ethanol concentrations in the range of 50–2,000 ppm. We can find its regularity through a large number of relevant experiments to establish the relationship between response value and concentration of ethanol. From comparison of two curves, a gradual enhancement in response amplitude was observed for both sensors, and the response amplitudes of GNP/SnO2-5 based sensor are always higher than that of pure SnO2, demonstrating its better sensitivity to ethanol.

It is well known that selectivity is another key criterion for measuring the quality of gas sensors. **Figure 2B** shows the selectivity test results of the pure SnO<sup>2</sup> and GNP/SnO2-5 sensors to five different gases of 500 ppm, including methanol, ethanol, methylbenzene, glycerine and methanal. It can be observed that the GNP/SnO2-5-based sensor has good selectivity to ethanol compared to that of pure SnO<sup>2</sup> sensor at 280◦C. The higher response to ethanol may be because ethanol is more likely to lose electrons in the process of a redox reaction with the absorbed oxygen, and the hydroxyl group (–OH) is much easier to oxidize at the optimum operating temperature.

The response–recovery time curve of GNP/SnO2-5-based sensor to 500 ppm of ethanol is shown in **Figure 2C**. Response and recovery time are defined as change in the resistances from Ra to [Ra−90% × (Ra – Rg)] for gas-in and [Ra + 90% × (Ra – Rg)] to gas-out, respectively (Zhang S.S. et al., 2018; Zhang Y.J. et al., 2018). It can be clearly observed that the response increased and decreased quickly when the GNP/SnO2-5-based sensor was exposed to and separated from ethanol, respectively. The response time and the recovery time of GNP/SnO2-5-based sensor are 26 and 64 s, respectively, which are much shorter than of the pure SnO2-based sensor that are 81 and 171 s. The relatively rapid response and recovery time could be due to the unique structure, which is the SnO<sup>2</sup> nanoparticles are decorated on the 2D sheet of GNP. This indicates that the large specific surface area is favorable to the adsorption of ethanol, which verifies the above conjecture. **Figure S4** depicts the response values of GNP/SnO2- 5-based sensor to 500 ppm of ethanol for every 3 days in 30 days at 280◦C, which fall slightly but are maintained around 295. Therefore, the conclusion could be obtained that the GNP/SnO2-5-based gas sensor to ethanol has a satisfactory stability, which confirms that the sensor might have a practical application.

# CONCLUSION

In conclusion, we reported an easy method to successfully prepare massive high-quality GNP from exfoliation of flake graphite, and GNP/SnO<sup>2</sup> nanocomposites were successfully synthesized by a facile solid-state method. The 2D GNP has no structural change during the exfoliating process from flake graphite, and the SnO<sup>2</sup> nanoparticles were highly distributed

## REFERENCES


on the surface of GNP. The GNP/SnO<sup>2</sup> based sensor showed excellent gas sensing performance toward ethanol, and the ameliorative gas-sensing properties may be due to the accrescent specific surface area and the interaction between 2D GNP and SnO<sup>2</sup> nanoparticles. Due to the procedure is convenient and environment-friendly, and good gas sensing property of the SnO2/GNP nanocomposite, it could be a promising candidate for ethanol detection.

## AUTHOR CONTRIBUTIONS

RZ performed the experiments and analyzed the data with the help from J-LC. J-BJ and YW conceived the study. All authors discussed the results and wrote the manuscript.

# FUNDING

This work was supported by the National Natural Science Foundation of China (U1704146, U1704255), Program for Science & Technology Innovation Talents in Universities of Henan Province (19HASTIT042), the Research Foundation for Youth Scholars of Higher Education of Henan Province (2016GGJS-040, 2017GGJS-053), the Fundamental Research Funds for the Universities of Henan Province (NSFRF1606, NSFRF170201) and Program for Innovative Research Team of Henan Polytechnic University (T2018-2).

# SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fchem. 2018.00337/full#supplementary-material


graphene oxide for high-performance ethanol gas sensor. Sens. Actuators B 255, 3275–3283. doi: 10.1016/j.snb.2017.09.154


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Zhang, Jia, Cao and Wang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Hydrothermal Synthesis of Hierarchical Ultrathin NiO Nanoflakes for High-Performance CH<sup>4</sup> Sensing

Qu Zhou1,2 \*, Zhaorui Lu<sup>1</sup> , Zhijie Wei <sup>1</sup> , Lingna Xu<sup>1</sup> , Yingang Gui <sup>1</sup> and Weigen Chen<sup>2</sup>

<sup>1</sup> College of Engineering and Technology, Southwest University, Chongqing, China, <sup>2</sup> State Key Laboratory of Power Transmission Equipment & System Security and New Technology, Chongqing University, Chongqing, China

Keywords: hydrothermal synthesis, ultrathin NiO nanoflakes, methane, gas sensor, sensing performances

# INTRODUCTION

Edited by:

Zhenyu Li, Deakin University, Australia

### Reviewed by:

Fuping Zeng, Wuhan University, China Peng Li, King Abdullah University of Science and Technology, Saudi Arabia

> \*Correspondence: Qu Zhou zhouqu@swu.edu.cn

### Specialty section:

This article was submitted to Nanoscience, a section of the journal Frontiers in Chemistry

Received: 19 April 2018 Accepted: 14 May 2018 Published: 30 May 2018

### Citation:

Zhou Q, Lu Z, Wei Z, Xu L, Gui Y and Chen W (2018) Hydrothermal Synthesis of Hierarchical Ultrathin NiO Nanoflakes for High-Performance CH4 Sensing. Front. Chem. 6:194. doi: 10.3389/fchem.2018.00194 Methane (CH4), as a colorless and odorless gas, is the main component of natural gas and widely used in various industries and human daily life (Schoonbaert et al., 2015). However, the leakage of natural gas, oil and gas storage, transportation and distribution systems increase atmospheric CH<sup>4</sup> concentration levels and lead to serious climate changes, which must be addressed (Zheng et al., 2017). Additionally, CH<sup>4</sup> will be easy to explode in a range of concentration (5–15%). Therefore, it is necessary to develop rapid and accurate gas sensors for CH<sup>4</sup> detection.

Nickel oxide NiO (Sun et al., 2014) is a significant p-type semiconductor, and has been widely used as catalyst (Yu et al., 2015), lithium-ion battery (Gu et al., 2016; Long et al., 2018), gas sensor (Wang et al., 2016; Zhou et al., 2018b), magnetic material (Cui et al., 2011), and so on. In recent years, many researchers have reported that NiO can be applied to fabricate high performance gas sensors for detecting some special gases such as hydrogen (Sta et al., 2016), NO<sup>2</sup> (Hoa and El-Safty, 2011), ethanol (Miao et al., 2017), etc. Zhang et al. studied a methane gas sensor based on nickel oxide (NiO)/reduced graphene oxide (rGO) nanocomposite film, which exhibited a response of 15% toward 1000 ppm CH<sup>4</sup> gas at 260◦C (Zhang et al., 2016), and the sensing response of the pure NiO film sensor only was 2.5% under the same condition. Moreover, few reports about the synthesis of hierarchical NiO nanostructures and its application for CH<sup>4</sup> detection was reported recently.

Thus, in this study we reported the successful synthesis of hierarchical ultrathin NiO nanoflakes and systematically researched their gas sensing properties to CH4. Interestingly, the proposed sensor exhibited high sensitivity, low optimal operating temperature, good linear relationship and excellent selectivity to CH4.

### EXPERIMENTAL DESIGN, MATERIALS, AND METHODS

### Sample Synthesis

All raw chemicals used for the synthesis of hierarchical ultrathin NiO nanoflakes were analytical graded and used as received without further purifications. In a typical hydrothermal procedure, 3 mmol of nickel nitrate hexahydrate Ni(NO3)2•6H2O and 0.200 g of polyvinylpyrrolidone (PVP) were dissolved in 100 mL DI deionized water under continuous stirring. After 15 min of rigorous stirring, a few drops of NH4OH solution were added to the resultant solution to maintain the pH = 11. The mixture was stirred vigorously for 30 min and then transferred into a Teflon-lined stainless steel autoclave, sealed and heated to 180◦C for 6 h. After cooling to room temperature naturally, the product was collected by centrifugation and washed with DI water and ethanol several times, respectively, and dried at 60◦C overnight. Finally, the dried power was calcined at 400◦C for 2 h.

### Sample Characterization

The phase structure of the prepared products were characterized by X-ray diffraction (XRD) using a Rigaku D/Max-1200X diffractometry with Cu-Kα radiation operated at 30 KV and 100 mA. The morphologies, microstructures and elemental compositions of the synthesized samples were investigated with a Nova 400 Nano field emission scanning electronic microscopy (FE-SEM), equipped with an energy dispersive X-ray spectroscopy (EDS). Gas sensors were fabricated with the side heated structure (Zhou et al., 2018c) and gas sensing properties of the obtained sensors were performed with the CGS-8 (Chemical gas sensor-8) intelligent gas sensing analysis system (Beijing Elite Tech Co., Ltd., Beijing, China).

# RESULTS

# Materials Characterization

**Figure 1A** shows the XRD pattern of the synthesized NiO sample. As shown, all the primary diffraction peaks observed at 37.30◦ , 43.25◦ , 62.85◦ , 75.45◦ , and 79.20◦ could be well assigned to (111), (200), (220), (311), and (222) planes of the cubic form of NiO (JCPDS Card No. 47-1049). **Figure 1B** depicts the EDS spectrum of the prepared NiO sample. As demonstrated, only nickel (Ni) and oxygen (O) peaks are observed with O/Ni molar ratio of nearly 1:1. No other diffraction peaks from impurities and dispersive peaks related with any element were observed, indicating a high purity of the as-prepared hierarchical ultrathin NiO nanoflakes sample.

**Figures 1C,D** demonstrate the FESEM image of the synthesized hierarchical NiO nanostructures, which are constructed by many ultrathin nanoflakes with smooth surface. The diameter of the NiO nanoflakes is in the scope of 300 to 400 nm with thickness ranging from 10 to 15 nm.

### Sensing Performances

The gas response of the fabricated side-heated sensor is defined as Rg/Ra, where Ra and Rg are the resistance values of the sensor in air and in the tested gas, respectively (Zeng et al., 2012; Zhou et al., 2018b). **Figure 2A** shows the relationship between the operating temperature and the gas response of the sensor to 30 ppm of CH<sup>4</sup> with working temperature ranging from 100 to 350◦C. As can be seen, with the increase of the temperature, the sensing response increases at first and attains its maximum value, and then decreases rapidly with further increasing temperature. The optimum operating temperature of the sensor to CH<sup>4</sup> is measured to be about 225◦C, lower than some already reported results (Zhang et al., 2016), and the corresponding response is 46.53.

**Figure 2B** illustrates the gas response of the sensor to various concentration of CH<sup>4</sup> ranging from 0.2 to 50 ppm at 225◦C. It is apparent that the sensing response increases rapidly with increasing gas concentration and a good linear relationship

between the sensing response and gas concentration can be obtained, implying an effective candidate for low concentration CH<sup>4</sup> detection.

**Figure 2C** demonstrates the dynamic response and recovery curve of the as-prepared sensor to 20, 30 ppm CH<sup>4</sup> at 225◦C. As illustrated, the sensor response increases dramatically when CH<sup>4</sup> gas was injected into the test chamber, and rapidly turns back to its initial state when subjected to air for sensor recovering. The time taken by the sensor to reach 90% of the total resistance change was defined as the response (recovery) time in the case of gas adsorption (desorption). According to this definition (Li et al., 2016; Zhang et al., 2018; Zhou et al., 2018a), the response and recovery time of the NiO sensor toward 30 ppm CH4 are calculated to be about 15 s and 20 s, respectively.

**Figure 2D** depicts the sensing response histogram of the NiO sensor to 30 ppm of various gases, including CH3OH, CH2O, NH3, H2S, and NO2. It can be seen that the presented NiO sensor shows extremely high response to CH<sup>4</sup> than other potential interfering gases. The longterm stability of the sensor was also measured and shown in **Figure 2E**, inserted in **Figure 2D**, where the sensor response changes slightly and keeps at a nearly constant value during the long experimental cycles, implying an excellent longtime stability and repeatability of the sensor for CH<sup>4</sup> detection.

It is known to all that NiO is a typical p-type semiconducting material, and its sensing properties are predominantly controlled by the surface resistance (Wang et al., 2016). When the fabricated NiO nanoflakes sensor is exposed to air, oxygen molecules would capture free electrons to form chemisorbed oxygen in the form of O<sup>−</sup> 2ads, O<sup>−</sup> ads and O2<sup>−</sup> ads absorbed on the sensor surface, increasing the number of electron holes of NiO surface and its conductivity. In CH<sup>4</sup> gas ambient, oxidation-reduction reactions would take place between the pre-adsorbed oxygen ions and CH<sup>4</sup> molecules, and then electrons are released back to NiO electron holes, resulting in a decreasing conductivity of the NiO nanoflakes sensor.

# CONCLUSIONS

In summary, hierarchical ultrathin NiO nanoflakes were successfully synthesized via hydrothermal process and characterized by XRD, FESEM and EDS, for the purpose of fabricating highly sensitive CH<sup>4</sup> gas sensors. The diameter of the prepared NiO nanoflakes material is in the scope of 300 to 400 nm with thickness ranging from 10 to 15 nm. Side-heated gas sensor was fabricated with the synthesized ultrathin NiO nanoflakes and methane CH<sup>4</sup> sensing performances were systematically evaluated. The synthesized hierarchical ultrathin NiO nanoflakes sensor exhibited high sensitivity, low optimal operating temperature, rapid response and recovery time, excellent selectivity and stability to CH<sup>4</sup> gas. Moreover, good linear relationship between the sensing response and gas concentration from 0.2 to 50 ppm was also obtained. All results indicate that the synthesized hierarchical ultrathin NiO nanoflakes may be a potential sensing material for fabricating high performance gas sensor for low concentration CH<sup>4</sup> detection.

## AUTHOR CONTRIBUTIONS

QZ and ZL performed the experiments and analyzed the data with the help from ZW and YG. QZ and ZL wrote the manuscript

### REFERENCES


with input from all authors. All authors read and approved the manuscript.

# ACKNOWLEDGMENTS

This work has been supported in part by the National Natural Science Foundation of China (Nos. 51507144, 5127785), China Postdoctoral Science Foundation funded project (Nos. 2015M580771, 2016T90832) and the Chongqing Science and Technology Commission (CSTC) (No. cstc2016jcyjA0400).


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Zhou, Lu, Wei, Xu, Gui and Chen. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Enhanced H2S Gas-Sensing Performance of Zn2SnO<sup>4</sup> Lamellar Micro-Spheres

Ting-Ting Xu, Ying-Ming Xu, Xian-Fa Zhang, Zhao-Peng Deng\*, Li-Hua Huo and Shan Gao\*

*Key Laboratory of Functional Inorganic Material Chemistry, Ministry of Education, School of Chemistry and Materials Science, Heilongjiang University, Harbin, China*

Keywords: Zn2SnO4 , H2S gas, ppb-level, lamellar micro-spheres, sensitivity

# INTRODUCTION

With the rapid development of industry, the discharge of sulfide gas has been increasing in recent decades, which results in severe air pollution. H2S, as the typical representation of sulfide, is a harmful and toxic acidic gas and widely used in various industries. Even at low concentrations, it can cause hypoxia and seriously threatens the safety of human. When the concentration reaches 1 mg/L (659 ppm) or higher, inflammation or death will occur. According to the U. S. Scientific Advisory Board on Toxic Air Pollutants, the acceptable concentration of H2S in the environment is less than 83 ppb (North Carolina Department of Environment and Natural Resources, 2003<sup>1</sup> ). Therefore, it is necessary to fabricate gas sensors which can detect ppb-level H2S in time to reduce the environmental pollution and harm to human.

### Edited by:

*Wen Zeng, Chongqing University, China*

### Reviewed by:

*Jun Zhang, Qingdao University, China Liwu Zhang, Fudan University, China*

### \*Correspondence:

*Zhao-Peng Deng dengzhaopeng@hlju.edu.cn Shan Gao shangao67@yahoo.com*

### Specialty section:

*This article was submitted to Nanoscience, a section of the journal Frontiers in Chemistry*

Received: *12 March 2018* Accepted: *23 April 2018* Published: *14 May 2018*

### Citation:

*Xu T-T, Xu Y-M, Zhang X-F, Deng Z-P, Huo L-H and Gao S (2018) Enhanced H2S Gas-Sensing Performance of Zn2SnO4 Lamellar Micro-Spheres. Front. Chem. 6:165. doi: 10.3389/fchem.2018.00165*

Zinc stannate (Zn2SnO4) is a typical n-type ternary semiconductor, and has been employed as important multifunctional material in the fields of photocatalytic activity (Das et al., 2017), solar cells (Li et al., 2015), lithium ion batteries (Lim et al., 2016), and so forth. Especially, owing to the excellent gas-sensing performance of ZnO and SnO<sup>2</sup> (Sukunta et al., 2017; Zhu and Zeng, 2017; Zhu et al., 2018), the application of Zn2SnO<sup>4</sup> in the field of gas sensors has attracted extensive attention (An et al., 2015; Zhao et al., 2016; Yang et al., 2017). Up to now, only one case has concerned on the detection of H2S with Zn2SnO<sup>4</sup> hollow octahedron (Ma et al., 2012), which showed the response to H2S with the detection limit being 1 ppm at 260◦C. Meanwhile, such reported Zn2SnO<sup>4</sup> sensor presents poor selectivity. Apparently, the sensor cannot satisfy the need of practical application

for the detection of ppb-level H2S, especially in a complex environment involving other interfering gases due to its poor selectivity and higher working temperature. In view of the fact that gas-sensing performance of materials is highly dependent on their micro-structure and surface state (Yu et al., 2017; Zhang et al., 2018), therefore, it would be a meaningful work to prepare Zn2SnO<sup>4</sup> with novel morphology to further decrease the working temperature and improve the selectivity and stability, thus performing the detection of ppb-level H2S.

In this work, Zn2SnO<sup>4</sup> lamellar micro-spheres have been synthesized by a facile ethylenediamine-assisted hydrothermal method followed by calcining at 600◦C. The diameter of micro-spheres is ∼1µm and they are composed of nanosheets with thickness of ∼85 nm. The sensor fabricated from the micro-spheres shows good response and selectivity to H2S at 170◦C, and the lowest detection limit is down to 50 ppb. Moreover, it shows good linear relationship in the range of ppb (50–1000 ppb) and ppm (3–50 ppm) level. Meanwhile, the gas-sensing mechanism is also investigated.

<sup>1</sup>http://daq.state.nc.us/toxics/studies/H2S/ (accessed 2003).

# EXPERIMENTAL

# Preparation of Zn2SnO<sup>4</sup> Lamellar Micro-Spheres

All the reagent were analytical grade (AR) and used as received without supplementary purification. Zn(CH3COO)2·2H2O (1.2 mmol) and SnCl4·5H2O (0.6 mmol) were dissolved in the mixture of 20 mL deionized water and 10 mL ethylenediamine. After stirring for 30 min, 7.2 mmol NaOH was added to the above system followed by further stirring for 1 h. Then, the above turbid solution was transferred into a 50 mL Teflon-lined stainless steel autoclave and kept at 200◦C for 24 h. After cooling to room temperature, the precipitate was centrifuged and washed with deionized water and ethanol for several times. White products were obtained after drying at 60◦C for 12 h, which was then calcined at 600◦C for 2 h in air atmosphere to obtain the Zn2SnO<sup>4</sup> lamellar micro-spheres.

The characterization, sensor fabrication and measurement were listed in the Supporting Information.

# RESULTS AND DISCUSSION

The FT-IR spectra of the precursor and calcined product were illustrated in Figure S1a. After calcination, the corresponding vibrations of acetate disappear. The spectrum of calcined product exhibits one peak at 3417 cm−<sup>1</sup> , corresponding to the stretching vibration of water molecules. The Sn-O and Zn-O vibrations at 572 and 431 cm−<sup>1</sup> in this spectrum are slightly enhanced, which indicates that the calcined product has excellent crystallinity. The TG curve of the precursor (Figure S1b) shows that the marked weight loss of 6.27% between 30 and 600◦C can be attributed to the loss of water molecules and small amount of acetates. After 600◦C, no obvious weight loss occurs with the increase of temperature. Therefore, the precursor is calcined at 600◦C for 2 h. The XRD patterns of the precursor and calcined product are also conducted to confirm their identity and phase purity. As illustrated in **Figure 1A**, all the diffraction peaks in both precursor and calcined product are consistent with standard cubic Zn2SnO<sup>4</sup> (JCPDS Card No. 24-1470). In comparison with the precursor, the calcined product presents stronger diffraction peaks and better consistency. It shows that the calcined product has better crystallinity. The surface components of the calcined product characterized by XPS spectrum indicate that the final product involves Zn, Sn and O elements (Figure S2).

As shown in **Figure 1B**, the micro-spheres of the precursor with diameter of ∼1.5µm are composed of smooth nanosheets with thickness of ∼130 nm. After being calcined at 600◦C in air, the calcined products maintain the morphology of precursor, and the diameter of the micro-spheres is slightly decreased from ∼1.5 to ∼1µm (**Figure 1C** and Figure S3). The thickness of the nanosheets is also sharply reduced to ∼85 nm. The results shown in Figure S4 reveal that the three elements are distributed very homogeneously in the micro-spheres. **Figure 1D** shows TEM image for the Zn2SnO<sup>4</sup> lamellar micro-sphere which presents a clear contrast between the dark and the pale parts, revealing that the micro-spheres are further assembled by nanosheets. The high resolution TEM (HRTEM) image in **Figure 1E** exhibits the clear parallel fringes with d-spacing of 0.261 nm for nanosheet, which corresponds to the (311) lattice planes of cubic Zn2SnO4. Meanwhile, some micro-pores appears on the surface of the nanosheets, which implies better gas performance of the lamellar micro-spheres.

**Figure 2A** is the response of Zn2SnO<sup>4</sup> lamellar micro-spheres sensor toward 100 ppm of H2S at different working temperatures. As can be seen, the response of sensor to H2S decreases with the increase of working temperature. This is mainly because the adsorbed H2S gas molecules escaped from the surface of materials easily before the reaction taking place at high operating temperature so as to a poor response as well. In other word, such commonly observed phenomenon in resistive sensors is mainly ascribed to the fact that the desorption plays dominant role at higher temperature which leads to the decrease in sensitivity. It shows that the best working temperature for the Zn2SnO<sup>4</sup> sensor is 170◦C, which is 90◦C lower than that of Zn2SnO<sup>4</sup> hollow octahedron (Ma et al., 2012). Thus, the following tests were carried out at 170◦C. **Figure 2B** is the response of the sensor toward 100 ppm of CH4, NH3, CH3COCH3, HCHO, C2H5OH and H2S at 170◦C. The responses to the above gases are 1.04, 1.49, 1.08, 1.2, 10.95, 65.13, respectively, which show that the sensor presents the best response to H2S than other gases. The selectivity coefficients (KAB) of H2S to other gases are 62.63, 43.71, 60.31, 54.27, 5.94, respectively, indicating excellent selectivity of the sensor to H2S. Such highly H2S selectivity can be attributed to the reactivities of the reducing test gases. The bond energy of 381 kJ/mol for H-SH in H2S (Liu et al., 2009) is smaller than other inorganic gases and most of organic gases, so that H-SH bond can be easily broken to participate in the reaction with gas sensor during chemical adsorption. Given all that, the sensor has good response and selectivity to H2S at 170◦C.

The relationship between the responses of the sensor and ppb-level concentrations of H2S and the response-recovery characteristics of the sensor to different concentrations of H2S at 170◦C are shown in **Figure 2C**. The gas response of the sensor increases in a good linear relationship from 50 to 1000 ppb (R <sup>2</sup> = 0.9986), indicating that the detection limit of the sensor is 50 ppb (response of 1.29) to H2S. In contrast to the reported Zn2SnO<sup>4</sup> octahedron sensor (Ma et al., 2012), the present sensor displays lower working temperature and ppb-level detection limit. Even compared with other preferable ternary metal oxides, such as La2NiO<sup>4</sup> particles (500◦C, 20 ppb, Hao et al., 2018), K2W4O<sup>13</sup> nano-wires (300◦C, 0.3 ppm, Supothina et al., 2014), Fe2(MoO4)<sup>3</sup> nano-particles and micro-spheres (300◦C, 1 ppm, Liang et al., 2016), the present material still shows better sensing performance toward H2S. To the best of our knowledge, the detection limit of Zn2SnO<sup>4</sup> lamellar micro-spheres sensor is only slightly higher than that of flower-like Bi2MoO<sup>6</sup> (Cui et al., 2017) which has the lowest detection limit of 0.1 ppb to H2S. The response of the sensor in the range from 3 to 50 ppm also increases in a good linear relationship (R <sup>2</sup> = 0.9927) (Figure S5). The responses of the sensor to 10 ppm H2S during 5 consecutive tests at 170◦C is shown in Figure S6a. It shows that the sensor keeps its initial response amplitude after 5 cycles. The sensor measurement also maintains initial response to 10 ppm H2S with the standard error of 4.9% after 60 days (Figure S6b). These

results indicate that this sensor has a satisfactory reproducibility and stability. Meanwhile, the influence of relative humidity on the sensor was considered at 170◦C (Figure S7). It can be seen from Figure S7 that the sensor response has a small change from 1.2 to 1.4 in different humidity environment, suggesting that the influence of humidity on the sensor can be neglected at such temperature. Therefore, the present Zn2SnO<sup>4</sup> lamellar micro-spheres sensor could be utilized as promising material for detecting ppb-level H2S.

It is considered that the gas sensing property of semiconductor oxides is related to the surface adsorption oxygen, therefore, the chemical state changes of each element in the Zn2SnO<sup>4</sup> sensor were analyzed before and after the sensor contacting with H2S at 170◦C. The results indicate that there is no change in the spectra of the Zn 2p and Sn 3d (Figure S8). As shown in Figures S9a,b, the O 1s spectra of Zn2SnO<sup>4</sup> could be deconvoluted into three peaks at 531.9/531.8, 530.8/530.7, 529.4/529.3, corresponding to hydroxyl oxygen, surface adsorbed oxygen and lattice oxygen, respectively. The percentage of surface adsorbed oxygen drops from 30.04 to 27.01% after the sensor contacting with H2S. The obvious decreasing of the adsorbed oxygen indicates that H2S participates in redox reaction with the surface adsorbed oxygen. When the Zn2SnO<sup>4</sup> sensor is exposed to H2S at 170◦C, the S 2p XPS spectrum displays three peaks at 168.9, 163.2 and 161.5 eV that correspond to SO<sup>2</sup> and sulfide (S 2p3/<sup>2</sup> and S 2p1/2) (Figure S9c), respectively. The appearance of the three peaks indicates that H2S is oxidized to SO2. Combined with previous reports (Yu et al., 2015), the redox reaction engenders between H2S and the surface adsorbed oxygen, and generates the product of SO2.

product.

Based on the aforementioned results, the gas sensing mechanism is speculated as follows: the resistance changes of the gas sensor are observed before and after contacting with H2S. When the Zn2SnO<sup>4</sup> sensor is exposed in air, oxygen molecules which is adsorbed on the surface of the sensor (Equation 1) captures electrons in conduction band of Zn2SnO<sup>4</sup> to form of O<sup>−</sup> at 170◦C (Equation 2), and engender electron depletion layer. Then, after the sensor contacting with the H2S, the H2S is oxidized by the O<sup>−</sup> on the surface of sensor and releases electrons back into the conduction band (Equations 3, 4), resulting in an increase in surface electrons and conductivity, and a decrease in resistance.

$$\text{O}\_2\text{(gas)} \rightarrow \text{O}\_2\text{(ads)}\qquad \text{(1)}$$

$$\text{O}\_2\text{(ads)} + 2\text{e}^- \rightarrow 2\text{O}^-\text{(420 - 670 K)}\tag{2}$$

$$\text{H}\_2\text{S(gas)} \rightarrow \text{H}\_2\text{S(ads)}\qquad \text{(3)}$$

<sup>H</sup>2S(ads) <sup>+</sup> 3O−(ads) <sup>→</sup> SO<sup>2</sup> <sup>+</sup> <sup>H</sup>2<sup>O</sup> <sup>+</sup> 3e−(443 K) (4)

### CONCLUSION

In conclusion, a facile ethylenediamine-assisted hydrothermal method followed by calcining at 600◦C led to the formation of Zn2SnO<sup>4</sup> lamellar micro-spheres which comprise of nanosheets with thickness of ∼85 nm. Such sensor exhibits excellent selectivity, sensitivity, humidity resistance and stability to H2S at working temperature 170◦C. The responses of the sensor, increased with the increasing concentrations in the range of 50– 1000 ppb and 3–50 ppm exhibit good relationships with the

### REFERENCES


detection limit of 50 ppb. These results indicate that the Zn2SnO<sup>4</sup> lamellar micro-spheres could be utilized as promising sensor material for detecting ppb-level H2S.

### AUTHOR CONTRIBUTIONS

T-TX performed the experiments and analyzed the data with the help from Y-MX and X-FZ. Z-PD wrote the manuscript with input from all authors. L-HH and SG conceived the study. All authors read and approved the manuscript.

### ACKNOWLEDGMENTS

This work is financial supported by Interational Science & Technology Cooperation Program of China (2016YFE0115100), the Project of Natural Science Foundation of Heilongjiang Province (No. B2015007), the Scientific and Technological Innovation Talents of Harbin (2016RAQXJ005), and the Young Innovation Talents of college in Heilongjiang Province (UNPYSCT-2016074). We thank the Key Laboratory of Functional Inorganic Material Chemistry, Ministry of Education, Heilongjiang University for supporting this study.

### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fchem. 2018.00165/full#supplementary-material


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2018 Xu, Xu, Zhang, Deng, Huo and Gao. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.