Open Access
Add to BibSonomy
Abstract
Mesoporous WO3 is synthesized by using SBA-15 and KIT-6 silica as templates, and Pt nanoparticles are then incorporated into the mesochannels of mesoporous WO3. TEM and BET results demonstrate the existence of the pore structure. Hydrogen sensing tests show that the repeatability of the sensors based on mesoporous WO3 has been significantly improved by an order of magnitude. Based on XRD, TEM and Raman spectrum results, a model of Pt nanoparticles encapsulated into channels of mesoporous WO3 is proposed to explain the improved repeatability of mesoporous WO3.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
With the depletion of fossil fuels and global warming caused by CO2, hydrogen has drawn widespread research interest for its reproducibility and environmental friendliness. However, hydrogen is a volatile, odorless and flammable gas with the ignition energy of 0.02 mJ, it is easy to explode in air when volume concentration exceeds 4%. Hence, a highly sensitive and selective hydrogen sensor is required to monitor hydrogen concentration in real time. Among the various types of hydrogen sensors, optical fiber hydrogen sensor has its unique advantages such as avoidance of spark generation, immunity to electromagnetic interference, distributed detection and environmental benignity. Various optical fiber hydrogen sensors have been developed such as evanescent sensor [1, 2], surface plasmon resonance sensor [3, 4] and fiber brag grating (FBG) sensor [5, 6]. Among these sensors, FBG hydrogen sensors are most likely to put into use for its distributed detection ability.
At present, many attempts have been made to improve the sensitivity and response time of the hydrogen sensors. Caucheteur et.al [7] first reported a FBG hydrogen sensor with Pt/WO3 coating, which was based on the exothermic reaction of the sensitive layer in hydrogen atmosphere and the threshold of this sensor is 6000 ppm. In our previous work, a FBG hydrogen sensor with detection limit of 200 ppm was fabricated by using a sol-gel method prepared Pt/WO3 as sensing material [6]. However, like most of hydrogen sensors, it exhibits a gradual degradation phenomenon after long time use, which is a severe problem for its application. There are various factors affecting the sensor’s repeatability such as water content [8], phase transition [9] and catalyst deactivation [10]. Among these factors, catalyst deactivation caused by the separation of Pt from WO3 support may result in the sensor’s degradation significantly. Mesoporous material has been recognized as an ideal support for catalyst for its ordered mesopores and high specific surface area. SBA-15 and KIT-6 has attracted much interests due to its ability to absorb and interact with guest species on their outer and inner surfaces, and in the pore space [11]. Besides, their uniform mesochannels and hydrothermal stability make them excellent templates to synthesize mesoporous metal oxides. Numerous research have reported methods incorporating nanoparticles into mesoporous silica (and metal oxides) in order to enhance the stability of the catalyst [12–14]. In this paper, a novel hydrogen sensing material mesoporous Pt/WO3 is proposed to improve the repeatability of FBG hydrogen sensor. Mesoporous WO3 is prepared by using a hard-templating method, where mesoporous silica is used as a template for accommodating adequate precursors with the mesochannels. Two kinds of mesoporous tungsten trioxide are prepared by using two different mesoporous silica as templates: 1) SBA-15 silica, which has a 2D hexagonal structure (space group p6mm) with large pores and thick walls; 2) KIT-6 silica, which possess a complex 3D pore network (space group Ia3d) integrated by a double-gyroidal mesostructure with channels running along the [100] and [111] directions [15]. Pt nanoparticles are then incorporated into the replica by calcination to obtain the gas sensing material. The two samples are named as S-Pt/WO3 (based on SBA-15 template) and K-Pt/WO3 (based on KIT-6 template). In addition, a sample of Pt/WO3, which was synthesized by doping Pt nanoparticles into the commercialized WO3 (Macklin) by using the same method above, was prepared as a control group. Experiment results show that the sensitivity and repeatability of the sensor based on mesoporous Pt/WO3 have been significantly improved, the relative standard deviation has been reduced by an order of magnitude. Besides, a lower detection limit of 100 ppm is achieved by mesoporous WO3 compared to 400 ppm of commercialized WO3.
2. Experimental
2.1 Synthesis of mesoporous silica
SBA-15 silica was synthesized according to the procedure described by Zhao et.al [11], with tetraethylortosilicate (TEOS, 98%, Sigma–Aldrich) and non-ionic tri-block copolymer (EO20PO70EO20, Pluronic P123, Aldrich) as silica source and structure directing agent, respectively. 6 g P123 was dissolved in the mixture of 30 g 35% HCl and 195 ml deionized water under vigorous stirring at 35°C for 6 hours. 12.49 g TEOS was added into the solution and stirred for 24 hours at 35°C,followed by a hydrothermal treatment at 100°C for 24 hours. The obtained product was filtered, washed, dried at room temperature in air and calcined at 550°C for 4 hours with a heating rate of 5°C /min. KIT-6 silica was prepared under acidic conditions by using a mixture of Pluronic P123 and butanol. 6 g P123 was dissolved in the mixture of 12 g 35% HCl and 220 g deionized water. The solution was stirred for 6 hours at 35°C and 6 g butanol was added under stirring for another 1 hour. 12.49 g TEOS was added into the solution and stirred for 24 hours at 35°C,followed by a hydrothermal treatment at 100°C for 24 hours. The solid product was filtered, washed, dried at room temperature in air and calcined at 550°C for 4 hours with heating rate of 5°C /min.
2.2 Synthesis of mesoporous Pt/WO3
S-Pt/WO3 and K-Pt/WO3 were synthesized by a hard template method by taking SBA-15 and KIT-6 silica as templates respectively. These silica materials were used as hard templates and phosphotungstic acid (Macklin) was used as a precursor for WO3. Impregnation of silica hosts with phosphotungstic acid was carried out in ethanol in two stages. In the first impregnation step, 0.3 g phosphor-tungstic acid hydrate and 0.3 g as-synthesized SBA-15 silica (same amount for KIT-6 silica) were mixed and dissolved into 15 g of ethanol under 2 hours of magnetic stirring to form a uniform solution. The mixture was dried and then calcined in air at 350°C for 4 hours. After that, the second impregnation was carried out by dispersing the silica/tungsten oxide product into ethanol with 0.3 g phosphor-tungstic acid hydrate, the solution was dried at room temperature and then calcined at 550°C for another 6 hours to obtain WO3 inside the hosting silica. Later, the obtained material was immersed in a 10 wt % HF solution under stirring to remove silica templates. After that, the obtained precipitates were separated by centrifugation, washed several times with distilled water and absolute ethanol, then dried in air at room temperature. After that, acetyl acetone platinum (Maclin Pt(acac)2) was mixed with commercialized tungsten trioxide, SBA-15 and KIT-6 WO3 replica respectively, where the mass ratio of Pt(acac)2 and WO3 was kept at 1159:207. The mixture was fully grinded and then annealed in air for 2 h at 315°C with heating rate of 5°C/min to obtain the gas sensing material, which were named as C-Pt/WO3, S-Pt/WO3 and K-Pt/WO3.
2.3 Characterization
The X-ray diffraction (XRD) measurement was performed to investigate the crystal information using a Bruker D8 Advance X-ray diffractometer with Cu Kα X-ray source operating at 40 kV. The morphology of the sensing material was characterized by using a Zeiss Ultra Plus field emission scanning electron microscope (FE-SEM). Transmission electron microscopic (TEM), high resolution transmission electron microscopic (HRTEM) images and selected area electron diffraction (SAED) pattern were collected using a JEOL JEM-2100F STEM/EDS microscope. BET analysis was performed using Tristar II3020 Surface Area and Porosity Analyzer.
2.4 Fabrication of sensors and sensing performance tests
Figure 1(a) illustrates the configuration of the FBG hydrogen sensing system. The system consists of a computer, a light source, a 3dB optical fiber coupler, an interrogator and a probe. There are two FBGs in the probe, FBG1 was covered with Pt/WO3 to detect H2, while FBG2 was used for temperature compensation. As shown in Fig. 1(b), the sensor devices were fabricated by uniformly covering the as-synthesized Pt/WO3 powder on the FBG fiber. Firstly, the FBG fiber was fixed on an A3 quartz substrate with a groove by using glue, then 0.03 g obtained Pt/WO3 was mixed with 15 μL de-ionized water to form a slurry. After that, the slurry was transformed into the groove and heated by the hot air gun. After the water was evaporated, the powder was covered on the surface of the FBG fiber. During H2 concentration test, the reflected wavelength was collected by the interrogator with resolution of 1 pm and the data can be displayed on computer in real time. The test was conducted in room temperature and the varying concentration of H2 were provided by changing flow rate of H2 and air. H2 concentration at 400, 1000, 1500, 3000, 6000, 9000, 12000 and 15000 ppm were prepared. When H2 was injected into the chamber, Pt/WO3 will react with H2 and the heat released by the reaction will cause a wavelength shift of the FBG covered with Pt/WO3. The relationship between central wavelength shift of the FBG and the temperature change has been reported in our previous work [5], which can be described as the following equation:
where “Δλ” means the wavelength shift of FBG and x is the H2 concentration, “a” is related to the hydrogen sensing material, “b” and “n” are constants of the reaction parameters. Besides, another 3 sensors based on C- Pt/WO3, S-Pt/WO3 and K-Pt/WO3 were fabricated by using the same method, they were placed in the chamber at hydrogen concentration of 10000 ppm (volume ratio of 1%) for 30 minutes and the test were repeated for 100 times in a week.
Fig. 1 a) Configuration of FBG hydrogen sensing system b) schematic illustration of the sensor’s fabrication process
3. Results and discussion
The structure and binding information of SBA-15 and KIT-6 replica WO3 were examined by X-ray diffraction (XRD) and Raman spectroscopy as shown in Fig. 2. Low-angel X-ray diffraction patterns of the two kinds of replica exhibit a strong diffraction peak at about 2θ = 0.5°, which indicate an ordered mesoporous structure of the replica. Further structure information of K-Pt/WO3, S-Pt/WO3 and C-Pt/WO3 were examined by wide-angel X-ray diffraction to explore any structure difference among the samples. The XRD patterns of the three samples before cycling interaction with hydrogen show no obvious differences, the diffraction peaks are well indexed to monoclinic phase of WO3 with lattice parameters of a = 7.3 Å, b = 7.53 Å and c = 7.68 Å (JCPDS No. 01-072-1465). Besides, there are two diffraction peaks at 2θ = 39.805° and 46.380°, corresponding to the (111) and (200) planes of the cubic phase of Pt (JCPDS No.00-004-0802). Raman peaks at 254, 327, 700 and 804 cm−1 can be observed for all three samples. According to the literature [16, 17], two bands at 804 and 700 cm−1 with high intensity can be attributed to the W-O-W stretching mode and the bands at 327 and 254 cm−1 can be ascribed to the W-O-W bending mode. Thus, it can be concluded that different template silica will not cause any structure or binding change to WO3.
Fig. 2 a) Low-angel XRD patterns of K-Pt/WO3 and S-Pt/WO3. b) XRD patterns of K-Pt/WO3, S-Pt/WO3 and C-Pt/WO3 before cycling. c) Raman spectrum of K-Pt/WO3, S-Pt/WO3 and C-Pt/WO3 before cycling.
SEM and TEM images of WO3 derived from different silica are presented in Fig. 3. WO3 KIT-6 replica (Fig. 3(b) exhibits a flower-like morphology with smaller particle size than that of SBA-15 replica, which are likely to aggregate together. According to the literature [15], agglomeration of the SBA-15 templated WO3 particles could be due to the solvent used in the synthesized process. TEM analysis was performed to observe the characteristics of the samples’ morphology. For SBA-15 WO3 replica (Fig. 3(c)), the pores are not very uniform and ordered. However, an ordered hexagonal lattice of pores with diameters of about 5 nm can be observed. Some of the WO3 nanoparticles are randomly oriented in the mesostructured framework, which may block the pores and make it hard to be observed in TEM images. Besides, mesochannels can be clearly seen in the images, which provide an evidence of the existed mesoporous structure. For KIT-6 WO3 replica (Fig. 3(d)), the pores are not very uniform and ordered and the porous framework is not very observable than that of SBA-15 WO3 replica, this phenomenon can be ascribed to the KIT-6 templated WO3 mainly occupy uncoupled subframeworks which originated from the double-gyroidal mesostructure with channels of KIT-6 silica. The uncoupled frameworks may form a sandwich-like structure and cover the pores, which make it difficult to be observed in TEM images.
Fig. 3 SEM images of a) S-Pt/WO3 and b) K-Pt/WO3 before cycling, TEM images of c) S-Pt/WO3 and d) K-Pt/WO3 before cycling
Brunauer–Emmett–Teller (BET) analysis was performed to explore pore structure and surface area of the samples and the results are depicted in Fig. 3. Figure 4(a) displays the nitrogen adsorption–desorption isotherms at 77 K for S-Pt/WO3 and K-Pt/WO3. Both of the samples exhibited a typical IV N2 sorption isotherm with a H1-type hysteresis loop, which is believed to be related to the capillary condensation associated with large pore channels. Pore size distributions were estimated by application of the BJH model to the adsorption branch of these isotherms, being illustrated in Fig. 4(b). The average pore diameter of SBA-15 replica is 7.2 nm and the BET specific surface area is 24.7 m2/g, while the pore diameter and the BET specific surface area of KIT-6 replica are 6.8 nm and 53.4 m2/g respectively. Combining with SEM and TEM results, mesoporous WO3 templated by KIT-6 exhibits a larger surface area with a smaller pore size than SBA-15 replica, the agglomeration phenomenon occurred in SBA-15 replica resulted in a smaller surface area, which may lead to a lower sensitivity of the sensor due to the decreased active sites.
Fig. 4 a) Nitrogen adsorption–desorption isotherms at 77 K for K-Pt/WO3 and S-Pt/WO3. b) Pore size distributions of K-Pt/WO3 and S-Pt/WO3
Hydrogen sensing tests were performed in a fiber optic hydrogen sensing system showed in Fig. 1(a), the sensing mechanism and configuration are described in experimental section in detail. In order to explore the detection limit of the three sensors, hydrogen at concentration of 100, 200, 400, 1000 and 1500 ppm were injected into the chamber accordingly, the response of the three sensors are shown in Fig. 5(a). At concentration of 100 ppm, K-Pt/WO3 and S-Pt/WO3 exhibited an obvious wavelength shift of 3 and 2 pm, while C-Pt/WO3 had no response until 400 ppm, which means the sensors based on mesoporous tungsten trioxide are more sensitive and have a lower detection limit. This can be ascribed to the larger surface area induced by the mesoporous structure, the pore structure provides more active sites for the reaction between sensing material with hydrogen and thus improving the detection limit. For all three sensors, when hydrogen concentration reach the detection limits of their own, wavelength shift increases accordingly with the ascending hydrogen concentration. Shifts in FBG wavelength at concentration of 1500 ppm for K-Pt/WO3, S-Pt/WO3, C-Pt/WO3 are 35, 30 and 17 pm respectively, sensitivity of the sensor based on S-Pt/WO3 is 0.0233 pm/ppm, which is twice as much as that of C-Pt/WO3. Figure 5(b) shows the response of the most sensitive sensor (K-Pt/WO3) at hydrogen concentration varying from 100 ppm to 4% (40000 ppm). It should be noted that the sensor was kept at each concentration for about 20 minutes, no obvious change was observed during these periods, which provide an evidence of good stability of the sensor. The testing result illustrates that the sensor has a high sensitivity with a wide range of measurement, which is suitable for application.
Fig. 5 a) Response of K-Pt/WO3, S-Pt/WO3 and C-Pt/WO3 at low hydrogen concentration ranging from100 ppm to 1500 ppm. b) Response of K-Pt/WO3 at hydrogen concentration ranging from 100 ppm to 4%.
In order to investigate the repeatability of the sensor, each of the sensor have been tested for 50 times. The first 40 tests were competed in a month, after that the sensors were held in the chamber for over two month and then the last ten tests were finished in half a month. In a typical performance test, 7 different concentration of hydrogen at 400, 1500, 3000, 6000, 9000, 12000 and 15000 ppm were injected into the chamber by changing the flow rate of hydrogen and air. However, the whole flow rate containing hydrogen and air was kept at 1000 sccm to avoid the influence of the flow rate. Wavelength shift at different concentration were recorded and the relationship between wavelength shift and hydrogen concentration is obtained by using a non-linear fit. Testing results are shown in Fig. 6. As shown in Fig. 6(a), (b) and (c), there is a clear degradation phenomenon during 50 tests in all three samples. However, C-Pt/WO3 suffered a most decline in wavelength shift while K-Pt/WO3 falls the least. The wavelength shift of C- Pt/WO3 based sensor at hydrogen concentration of 15000 ppm decreased half from 378 pm to 182 pm during the 50 tests, while the wavelength shift of K-Pt/WO3 (S-Pt/WO3) decreased from 530 pm (480 pm) to 500 pm (400 pm) respectively. It should be noted that in the last ten tests, K-Pt/WO3 exhibit almost the same response to hydrogen compared with the first few tests although three month have passed, which means that the mesoporous structure can efficiently improve the sensor’s repeatability.
Fig. 6 a) Sensing performance of K-Pt/WO3, b) S-Pt/WO3 and c) C-Pt/WO3 in 50 tests. d) Wavelength shift and the standard deviation of K-Pt/WO3, S-Pt/WO3 and C-Pt/WO3 in 50 tests.
Figure 6(d) illustrates the average wavelength shifts and the corresponding standard deviations during 50 tests at various hydrogen concentration. To some extent, the dispersion degree of the 50 fitting curves of the three sensors can indicate their repeatability. It can be concluded that the relative standard deviation (RSD) of C-Pt/WO3 is about 27% at hydrogen concentration of 15000 ppm, while the RSD of K-Pt/WO3 and S-Pt/WO3 are less than 2.5% and 9%. Meanwhile, the average FBG wavelength shifts at hydrogen concentration of 15000 ppm based on K-Pt/WO3, S-Pt/WO3 and C-Pt/WO3 reached 512 pm, 444 pm and 265 pm respectively, which means the average hydrogen sensitivity are 0.034, 0.029 and 0.017 pm/ppm. As the resolution of the interrogator is 1 pm, the equivalent H2 concentration resolution of the sensors are calculated to be 29.4 ppm and 34.5 ppm. The sensing parameters of three sensors are listed in Table 1. It’s clear that the sensors based on mesoporous material have higher sensitivity and resolution, besides the sensors have a lower detection limit with better repeatability. For K-Pt/WO3, the relative standard deviation in 50 tests is reduced from 27% to 2.5% by an order of magnitude. In addition, each of the performance parameters of K-Pt/WO3 is better than those of S-Pt/WO3.
Table 1. Sensing parameters of K-Pt/WO3, S-Pt/WO3 and C-Pt/WO3
In order to further explore the sensing performance of the three sensors, another 3 sensors based on C- Pt/WO3, S-Pt/WO3 and K-Pt/WO3 were fabricated by using the same method, they were placed in the gas chamber at hydrogen concentration of 10000 ppm (volume ratio of 1%) for 30 minutes and the test were repeated for 100 times in a week. Testing results are shown in Fig. 7, the initial response of K-Pt/WO3 and S-Pt/WO3 are 335 pm and 300 pm, compared to 229 pm of C-Pt/WO3. After 100 tests, the response decreased to 307 pm, 238 pm and 114 pm for K-Pt/WO3, S-Pt/WO3 and C-Pt/WO3. As shown in Fig. 8, all of three samples exhibit a gradual degradation phenomenon, no obvious sharp declines in performance are found during the test. We can conclude that the degradation is greatly alleviated by introduction of pore structure into WO3 and KIT-6 template is more effective than SBA-15 to improve the repeatability of hydrogen sensor. However, the modified WO3 still suffer loss in performance during hydrogen tests, this can be ascribed to the water content produced in the reaction between Pt/WO3 with hydrogen.
Fig. 7 Cycling performance at hydrogen concentration of 10000 ppm (1%) of K-Pt/WO3, S-Pt/WO3 and C-Pt/WO3.
Fig. 8 a) XRD patterns and b) Raman spectrum of K-Pt/WO3, S-Pt/WO3 and C-Pt/WO3 before and after cycling. TEM images of c) K-Pt/WO3 and d) C-Pt/WO3 after cycling.
After hydrogen sensing performance was completed, XRD, Raman and TEM analysis were performed to investigate the structure change during the tests. XRD patterns of the samples before and after cycling are shown in Fig. 8(a). In Fig. 8(a), the offsets of the 6 XRD patterns from bottom to top are 0, 1600, 3200, 4800, 6400 and 8000 respectively, while the offsets of the 6 Raman spectrum from bottom to top in Fig. 8(b) are 0, 1000, 4000, 5000, 9000 and 10000 respectively. No obvious change in diffraction peaks are found, which means no phase change occurred during the test. Raman spectrum of the samples before and after cycling are demonstrated in Fig. 8(b). Similarly, Raman shifts have no obvious change during the tests, which prove that the binding structure remained the same. TEM images of C-Pt/WO3 after cycling exhibit an obvious agglomeration of Pt particles (the darker area), while Pt nanoparticles distribute uniformly in the WO3 matrix of K-Pt/WO3. Thus, it can be concluded that the degradation phenomenon is mainly resulted by the agglomeration of Pt particles. Numerous research have found that catalyst aggregation will reduce the active sites and thus decreasing the catalytic activity significantly. As our hydrogen sensors work by the heat released by the reaction between Pt/WO3 with hydrogen, any catalyst deactivation will result in worse repeatability. However, excellent repeatability of K-Pt/WO3 and S-Pt/WO3 can be ascribed to the encapsulation effect of mesoporous WO3. As shown in Fig. 9, Pt particles can be incorporated into the channels of the WO3 porous framework during the calcination process and will be anchored in the channels by binding force with the thick walls. Although the reaction with hydrogen will reduce the binding force between them, the resistance caused by the channel makes it difficult for Pt nanoparticles to deviate from the channels to the surface of the mesoporous matrix. However, resistance caused by channels varies from different pore structure. Compared to SBA-15 silica with a 2D hexagonal structure, KIT-6 silica possesses a double-gyroidal channels, which makes it more difficult for Pt particles to move away from the channels to surface. This property makes the sensor based on KIT-6 templated WO3 (K-Pt/WO3) more stable than S-Pt/WO3, which can be proved by the test results. On the other hand, for C-Pt/WO3, Pt nanoparticles are distributed on the surface of WO3. Binding force between Pt and WO3 will be reduced gradually during the tests, which will result in catalyst agglomeration and decline in sensing performance. Besides, the mesoporous structure provide a larger surface area and more active sites, thus K-Pt/WO3 and S-Pt/WO3 based sensors have a lower detection limit and a higher sensitivity than -Pt/WO3.
Fig. 9 a) Structure model of K-Pt/WO3, b) S-Pt/WO3 and a schematic representation describing Pt nanoparticles encapsulated in the mesochannels of c) K-Pt/WO3 and d) S-Pt/WO3.
4. Conclusion
In summary, two mesoporous WO3 have been successfully synthesized by using SBA-15 and KIT-6 silica as templates. Compared to commercialized WO3, hydrogen sensors based on mesoporous have a higher sensitivity with improved detection limit of 100 ppm. In addition, the sensor’s repeatability has been significantly improved. The relative standard deviation is reduced from ± 27% to ± 9% and ± 2.5% for S-Pt/WO3 and K-Pt/WO3. Besides, mesoporous WO3 based on different template exhibit different performance in sensitivity and repeatability. Based on XRD, Raman spectrum SEM and TEM analysis, K-Pt/WO3 exhibits a larger specific surface area and porous structure, which contributed to the higher sensitivity than that of S-Pt/WO3; XRD and Raman spectrum reveal that no obvious phase change or binding information occurred for all three samples after cycling, variation in repeatability can be ascribed to the agglomeration of Pt nanoparticles observed in TEM images. Mesochannels will hold Pt nanoparticles tightly to alleviate the dropping process and thus enhancing the sensor’s repeatability. Compared to S-Pt/WO3 with a 2D hexagonal structure, K-Pt/WO3 possesses a 3d pore network with double-gyroidal channels, which makes movement of Pt nanoparticles more difficult and prevent them separating with WO3, resulting a better repeatability than S-Pt/WO3.
Funding
National Natural Science Foundation of China, NSFC (Project Number: 61575151).
References and links
1. S. Okazaki, H. Nakagawa, S. Asakura, Y. Tomiuchi, N. Tsuji, H. Murayama, and M. Washiya, “Sensing characteristics of an optical fiber sensor for hydrogen leak,” Sens. Actuators B Chem. 93(1-3), 142–147 (2003). [CrossRef]
2. H. Nakagawa, N. Yamamoto, S. Okazaki, T. Chinzei, and S. Asakura, “A room-temperature operated hydrogen leak sensor,” Sens. Actuators B Chem. 93(1-3), 468–474 (2003). [CrossRef]
3. R. Jiang, F. Qin, Q. Ruan, J. Wang, and C. Jin, “Ultrasensitive Plasmonic Response of Bimetallic Au/Pd Nanostructures to Hydrogen,” Adv. Funct. Mater. 24(46), 7328–7337 (2014). [CrossRef]
4. M. Hosoki, M. Nishiyama, H. Igawa, A. Seki, Y. Choi, and K. Watanabe, “A surface plasmon resonance hydrogen sensor using Au/Ta2O5/Pd multi-layers on hetero-core optical fiber structures,” Sens. Actuators B Chem. 185, 53–58 (2013). [CrossRef]
5. X. Zhong, M. Yang, C. Huang, G. Wang, J. Dai, and W. Bai, “Water photolysis effect on the long-term stability of a fiber optic hydrogen sensor with Pt/WO3,” Sci. Rep. 6(1), 39160 (2016). [CrossRef] [PubMed]
6. J. Dai, M. Yang, Z. Yang, Z. Li, Y. Wang, G. Wang, Y. Zhang, and Z. Zhuang, “Performance of fiber Bragg grating hydrogen sensor coated with Pt-loaded WO3 coating,” Sens. Actuators B Chem. 190, 657–663 (2014). [CrossRef]
7. C. Caucheteur, M. Debliquy, D. Lahem, and P. Megret, “Hybrid fiber gratings coated with a catalytic sensitive layer for hydrogen sensing in air,” Opt. Express 16(21), 16854–16859 (2008). [CrossRef] [PubMed]
8. J. C. Shi, G. M. Wu, G. Gao, T. Liang, J. Shen, B. Zhou, X. Y. Ni, and Z. H. Zhang, “Effects of water on the gasochromc properties of WO3 thin films,” Journal of Functional Materials & Devices 15, 459–465 (2009).
9. G. Gao, J. Wu, G. Wu, Z. Zhang, W. Feng, J. Shen, and B. Zhou, “Phase transition effect on durability of WO3 hydrogen sensing films: An insight by experiment and first-principle method,” Sens. Actuators B Chem. 171–172, 1288–1291 (2012). [CrossRef]
10. J. Y. Luo, L. Gong, H. D. Tan, S. Z. Deng, N. S. Xu, Q. G. Zeng, and Y. Wang, “Study of the catalyst poisoning and reactivation of Pt nanoparticles on the surface of WO3 nanowire in gasochromic coloration,” Sens. Actuators B Chem. 171–172, 1117–1124 (2012). [CrossRef]
11. D. Y. Zhao, J. L. Feng, Q. S. Huo, N. Melosh, G. H. Fredrickson, B. F. Chmelka, and G. D. Stucky, Triblock Copolymer Syntheses of Mesoporous Silica with Periodic 50 to 300 Angstrom Pore,” Science 279, 548 (1998).
12. Y. Shi, C. Zhao, H. Wei, J. Guo, S. Liang, A. Wang, T. Zhang, J. Liu, and T. Ma, “Single-atom catalysis in mesoporous photovoltaics: the principle of utility maximization,” Adv. Mater. 26(48), 8147–8153 (2014). [CrossRef] [PubMed]
13. G. Laugel, J. Arichi, M. Molière, A. Kiennemann, F. Garin, and B. Louis, “Metal oxides nanoparticles on SBA-15: Efficient catalyst for methane combustion,” Catal. Today 138(1-2), 38–42 (2008). [CrossRef]
14. J. van der Meer, I. Bardez, F. Bart, P.-A. Albouy, G. Wallez, and A. Davidson, “Dispersion of Co3O4 nanoparticles within SBA-15 using alkane solvents,” Microporous Mesoporous Mater. 118(1-3), 183–188 (2009). [CrossRef]
15. E. Rossinyol, A. Prim, E. Pellicer, J. Arbiol, F. Hernández-Ramírez, F. Peiró, A. Cornet, J. R. Morante, L. A. Solovyov, B. Tian, T. Bo, and D. Zhao, “Synthesis and Characterization of Chromium-Doped Mesoporous Tungsten Oxide for Gas Sensing Applications,” Adv. Funct. Mater. 17(11), 1801–1806 (2007). [CrossRef]
16. W. Feng, G. Wu, and G. Gao, “Ordered mesoporous WO3film with outstanding gasochromic properties,” J. Mater. Chem. A Mater. Energy Sustain. 2(3), 585–590 (2014). [CrossRef]
17. S.-H. Lee, H. M. Cheong, P. Liu, D. Smith, C. E. Tracy, A. Mascanrenhas, J. R. Pitts, and S. K. Deb, “Gasochromic mechanism in a-WO3 thin films based on Raman spectroscopic studies,” J. Appl. Phys. 88(5), 3076–3078 (2000). [CrossRef]
