- Email: [email protected]
Template-free synthesis of high-yield phosphonated tin oxides with high specific surface area
Materials Letters 236 (2019) 85–88
Contents lists available at ScienceDirect
Materials Letters journal homepage: www.elsevier.com/locate/mlblue
Template-free synthesis of high-yield phosphonated tin oxides with high specific surface area Bing Guo, Xiuzhen Lin ⇑, Peng Liu, Yanyan Zeng, Hongbo Fan School of Environment and Civil Engineering, Dongguan University of Technology, Dongguan 523808, Guangdong, China
a r t i c l e
i n f o
Article history: Received 3 January 2018 Received in revised form 12 September 2018 Accepted 14 October 2018 Available online 15 October 2018 Keywords: Phosphonated tin oxide High surface area Porous materials Nanoparticles
a b s t r a c t High-yield porous phosphonated tin oxides (SnEDTMP) were successfully synthesized through a facile hydrothermal route with SnCl45H2O as tin source, ethylene diamine tetra(methylene phosphonic acid) (EDTMP) as the organophosphorus and NaOH as the pH regulator. Without any additive template during synthetic process, the obtained product had a high specific surface area of 377 m2/g, which was considered to stem from the aggregation of nanoparticles resulting in sufficient void spaces. It was found that the order of NaOH added during synthetic process could effectively dominate the aggregation of nanoparticles such to form tight or loose accumulation, which was the key to successful preparation of highspecific-surface-area phosphonated tin oxides. Ó 2018 Elsevier B.V. All rights reserved.
1. Introduction Various tin-contained compounds, because of their Lewis-acidic and redox properties from tin species, have been widely used in photocatalytic degradation of organic dyes [1], photovoltaic devices [2], rechargeable lithium batteries [3,4], gas-sensing materials [5], and so on. Phosphonating the tin-contained compounds to form organic-inorganic hybrid structures would enrich their functionalities and widen their potential application prospects in many aspects, such as heterogeneous catalysis, adsorption, ion exchange, and as supports [6–9]. To improve application performance of phosphonated tin oxides, incorporating porosity to fabricate materials with high specific surface areas would be a feasible strategy. Surfactant templating routes have been developed to serve this purpose. For example, in 2005 Fujiwara et al. [10] fabricated mesoporous hybrid tin phenylphosphonate with a specific surface area of 371 m2/g in the presence of anionic surfactant sodium dodecylsulfate. In 2012, Dutta et al. [11] synthesized a hybrid porous tin (IV) phosphonate with SnCl55H2O as tin source, pentaethylenehexamine-o ctakis-(methyl phosphonic acid) hexadecasodium salt solution as the phosphonate source and cetyl trimethylammonium bromide as the structure directing agent. This material showed a BET surface area of 723 m2/g, and good catalytic activity in one-pot liquid phase oxidation of cyclohexanone to adipic acid under eco-friendly ⇑ Corresponding author. E-mail address: [email protected] (X. Lin). https://doi.org/10.1016/j.matlet.2018.10.086 0167-577X/Ó 2018 Elsevier B.V. All rights reserved.
conditions. In 2016, Wang et al. reported a category of porous tin (IV) phosphonates, showing specific surface areas in the range of 27–168 m2/g, in presence of nonionic templating agent F127 through solvothermal treatment [9]. For templating synthesis strategy, surfactant molecules have to be further removed whether by acid-ethanol extraction or by calcination. This method would make tedious synthetic steps and lead to increased cost, as well as bring environmental pollution in the post-treatment. Surfactant-free synthesis of such kind of materials with high specific surface area is intriguing. However, the related reports are rare [12]. In this research, high-yield porous phosphonated tin oxide (SnEDTMP) was hydrothermally synthesized with SnCl45H2O as tin source, ethylene diamine tetra(methylene phosphonic acid) (EDTMP) as the organophosphorus and NaOH as the pH regulator. Without any surfactant templating agent or organic solvent involved in the hydrothermal media, SnEDTMP with a high specific surface area of 377 m2/g could be obtained by simply controlling the adding order of NaOH in the synthetic process. 2. Materials and methods The typical synthesis of porous phosphonated tin oxides could be described as follows. An EDTMP aqueous solution was prepared by dissolving 0.76 g of EDTMP into 25 mL of water and then NaOH (0.56 g) was added under stirring. Afterward an aqueous solution of SnCl45H2O (1.78 g in 5 mL of H2O) was added dropwise. The mixture was continuously stirred at 45 °C for 2 h and then trans-
86
B. Guo et al. / Materials Letters 236 (2019) 85–88
ferred into an autoclave aging at 120 °C for 24 h. Finally white powder were collected by filtering, washed with water and dried at 80 °C. The as-prepared material was designated as SnEDTMP. For the purpose of comparison, SnO2 sample was prepared following the processing described above without addition of phosphorus. Details about the material characterizations can be found in Supplementary material.
3. Results and discussion
Fig. 1. XRD patterns of as-synthesized SnEDTMP and SnO2 fabricated.
The XRD pattern of SnEDTMP (Fig. 1) gives four typical peaks at 2h values of 26.38°, 33.76°, 37.61° and 51.4°, corresponding to (1 1 0), (1 0 1), (2 0 0), and (2 1 1) reflections, respectively, which evidently revealed the presence of a tetragonal SnO2 crystalline structure. The as-obtained SnEDTMP contained tetragonal SnO2
Fig. 2. (A) SEM image, low- (B, C) and high-resolution (D) TEM images, (E) the corresponding particle size distribution from (C), and (F) N2 adsorption-desorption isotherms, of SnEDTMP.
B. Guo et al. / Materials Letters 236 (2019) 85–88
0.330 nm, approximate to the d-spacing of 0.338 nm corresponding to (1 1 0) reflection at 2h = 26.38°. The amorphous area in Fig. 2D might be composed of Sn-O and organophosphonate species. In combination with the XRD analysis, clearly SnEDTMP had a porous structure with tetragonal SnO2 nanocrystals well dispersed within the hybrid framework. The N2 adsorption-desorption isotherms for SnEDTMP were of type II (Fig. 2F). A very strong increase of nitrogen-adsorbed volume was observed at high relative pressure (P/P0 > 0.6), which suggested the presence of an appreciable amount of secondary porosity of very large pores (even macropores) [13], consistent with the interparticle voids observed in SEM and TEM images. The isotherms displayed type H3 hysteresis loops that did not level off at relative pressures close to the saturation vapor pressure, indicating these materials comprised of aggregates of particles forming narrow slitlike pores. Its pore size distribution curve is shown in Fig. S1. SnEDTMP has a specific surface area of 377 m2/ g and a pore volume of 0.32 cm3/g. Without addition of EDTMP in the synthetic media, the SnO2 material prepared followed the same hydrothermal procedure only gives a specific surface area of 200 m2/g with a pore volume of 0.11 cm3/g. Apparently, organophosphonate groups played an important role in enlarging the surface area of SnEDTMP. Interestingly, it was found that when NaOH was added at the last step in the synthetic process, the asprepared material showed a specific surface area as low as 131 m2/g with a lower yield of 0.844 g. The aggregation of nanoparticles from the corresponding SEM images showed a very tight stacking between nanoparticles with little space left (Fig. S2). Apparently, the sequence of NaOH added in the synthesis process played a key role in controlling the particles aggregation, and accordingly resulting in the difference in their specific surface areas. Fig. 3 shows the FT-IR spectrum of SnEDTMP. The strong and wide band at 3400 cm 1 and the sharp peak at 1630 cm 1 corresponded to the surface-adsorbed water and hydroxyl groups [14]. The weak bands around 2900–3000 cm 1 were assigned to the CAH stretching modes. The overlapped bands at 1460 and 1430 cm 1 were assigned to the CAH bending in ACH2A groups and the P-C stretching vibrations, respectively [15]. The small bands at 1373 and 1322 cm 1 could be attributed to phosphoryl (P@O) frequency and CAN stretching, respectively. The broad band around 1000–1200 cm 1 could be disintegrated into two peaks: one centered at 1163 cm 1 assigned to the P–CH2N = groups [16], another around 1049 cm 1 attributed to the P-O-Sn groups. The shoulder peak around 988 cm 1 assigned to PAOH groups were not observed, suggesting the P atoms are connected to Sn atoms
with the grain size about 6.6 nm, calculated based on the characteristic diffraction at 2h = 33.76° according to the Scherrer equation. In addition, the yield of SnEDTMP is high (1.905 g). Without addition of EDTMP in the synthetic process, the obtained SnO2 sample shows the same reflection as SnEDTMP does, but its intensity became much weaker, as well as giving a low yield (0.195 g). The evident difference in the yield for the synthetic cases with EDTMP added or not reveals that the organophosphonate groups were well contained in the SnEDTMP. According to analysis above, it could be concluded that the structure of SnEDTMP presented organophosphonated SnO2 hybrid framework. The SEM image of SnEDTMP in Fig. 2A presents the aggregation of irregular-shaped particles with submicrometer size. It’s noted that their aggregation produced void spaces with pore sizes in the range of 20–240 nm. The TEM image in Fig. 2B demonstrated the agglomeration of small particles, rendering rich inter-spaces of several tens to hundred nanometers between these particles. This result was consistent with that observed from the SEM image (Fig. 2A). In the enlarged micrograph, abundant clear crystal fringes of the nanocrystals with an average diameter of 6.0 nm are observed (Fig. 2C and E), which was close to that (6.6 nm) calculated from XRD pattern. As shown in the high-resolution TEM micrograph (Fig. 2D), the measured interlayer spacing was
Fig. 3. FT-IR spectrum of SnEDTMP.
Fig. 4. Solid-state
13
C and
87
31
P MAS NMR spectra of SnEDTMP.
88
B. Guo et al. / Materials Letters 236 (2019) 85–88
through PAOASn bonding, and the majority of multidentate phosphonic acid were coordinated to tin atoms. Fig. 4 gives the 13C and 31P MAS NMR spectra of SnEDTMP. The sole signal at 52.98 ppm from 13C MAS NMR spectrum was detected, and it could be attributed to the carbon atoms in the nitrilomethylenephosphonate group (Fig. 4a). 31P MAS NMR spectroscopy was conducted to provide useful information on the nature of the phosphonate groups. The overlapped chemical shift of SnEDTMP in Fig. 4b could be split into signals at 11.73 and 2.25 ppm, ascribed to the P nuclei in the form of RPO2(OH) and RPO3, respectively [17]. No signal attributed to RPO(OH)2 species was observed in the 31P MAS NMR spectroscopy. The change of phosphorus environment in the hybrid material indicates the formation of PAOASn bonds, and also suggests that the organic moieties were well integrated into the hybrid framework, instead of pendent on the pore wall. A combination of all above analyses demonstrated the successful synthesis of porous phosphonated tin oxides with high specific surface areas.
4. Conclusions High-yield porous phosphonated tin oxides (SnEDTMP) were successfully synthesized via a facile hydrothermal route without adding any surfactant templating agent or organic solvent. It was found that the adding sequence of NaOH as the pH regulator in the synthetic process could effectively dominate the aggregation of nanoparticles whether to form a tight stacking or loose accumulation. As a result, SnEDTMP with a high specific surface area of 377 m2/g was obtained when NaOH was added in the first step during the synthetic process. Yet the sample had a rather lower specific surface area of 131 m2/g when NaOH was added at last step.
Acknowledgements This work was supported by the National Natural Science Foundation of China (No. 21503041), Technology Planning Project of Guangdong Province (No. 2015B090927007) and Guangdong Provincial Key Plat-form and Major Scientific Research Projects for Colleges and Universities (No. 2015KCXTD029). Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.matlet.2018.10.086. References [1] J.P. Wilcoxon, J. Phys. Chem. B 104 (2000) 7334. [2] S. Gubbala, V. Chakrapani, V. Kumar, M.K. Sunkara, Adv. Funct. Mater. 18 (2008) 2411. [3] E.J. Kim, D. Son, T.G. Kim, et al., Angew. Chem. Int. Ed. 43 (2004) 5987–5990. [4] G. Derrien, J. Hassoun, S. Panero, B. Scrosati, Adv. Mater. 19 (2007) 2336–2340. [5] X.G. Han, M.S. Jin, S.F. Xie, et al., Angew. Chem. Int. Ed. 48 (2009) 1–5. [6] P. Bhanja, A. Bhaumik, ChemCatChem 8 (2018) 1607–1616. [7] S. Kirumakki, S. Samarajeewa, R. Harwell, et al., Chem. Commun. (2008) 5556– 5558. [8] S. Borah, B. Bhattacharyya, J. Deka, A. Borah, et al., Dalton Trans. 46 (2017) 8664–8672. [9] K. Lv, J. Han, C.T. Yang, et al., Chem. Eng. J. 302 (2016) 368–376. [10] N. Kishor Mal, M. Fujiwara, M. Matsukata, Chem. Commun. (2005) 5199–5201. [11] A. Dutta, M. Pramanik, A.K. Patra, M. Nandi, H. Uyama, A. Bhaumik, Chem. Commun. 48 (2012) 6738–6740. [12] A. Subiah, D. Pyle, A. Rowland, et al., J. Am. Chem. Soc. 127 (2005) 10826– 10827. [13] K.S.W. Sing, D.H. Everett, R.A.W. Haul, et al., Pure Appl. Chem. 57 (1985) 603– 619. [14] T.Z. Ren, Z.Y. Yuan, B.L. Su, Chem. Phys. Lett. 374 (2003) 170–175. [15] E. Jaimez, G.B. Hix, R.C.T. Slade, Solid State Ionics 97 (1997) 195. [16] D.E. Lόpez, J.G. Goodwin Jr., D.A. Bruce, J. Catal. 245 (2007) 381–391. [17] X.Z. Lin, Z.Y. Yuan, Eur. J. Inorg. Chem. (2012) 2661–2664.
Contents lists available at ScienceDirect
Materials Letters journal homepage: www.elsevier.com/locate/mlblue
Template-free synthesis of high-yield phosphonated tin oxides with high specific surface area Bing Guo, Xiuzhen Lin ⇑, Peng Liu, Yanyan Zeng, Hongbo Fan School of Environment and Civil Engineering, Dongguan University of Technology, Dongguan 523808, Guangdong, China
a r t i c l e
i n f o
Article history: Received 3 January 2018 Received in revised form 12 September 2018 Accepted 14 October 2018 Available online 15 October 2018 Keywords: Phosphonated tin oxide High surface area Porous materials Nanoparticles
a b s t r a c t High-yield porous phosphonated tin oxides (SnEDTMP) were successfully synthesized through a facile hydrothermal route with SnCl45H2O as tin source, ethylene diamine tetra(methylene phosphonic acid) (EDTMP) as the organophosphorus and NaOH as the pH regulator. Without any additive template during synthetic process, the obtained product had a high specific surface area of 377 m2/g, which was considered to stem from the aggregation of nanoparticles resulting in sufficient void spaces. It was found that the order of NaOH added during synthetic process could effectively dominate the aggregation of nanoparticles such to form tight or loose accumulation, which was the key to successful preparation of highspecific-surface-area phosphonated tin oxides. Ó 2018 Elsevier B.V. All rights reserved.
1. Introduction Various tin-contained compounds, because of their Lewis-acidic and redox properties from tin species, have been widely used in photocatalytic degradation of organic dyes [1], photovoltaic devices [2], rechargeable lithium batteries [3,4], gas-sensing materials [5], and so on. Phosphonating the tin-contained compounds to form organic-inorganic hybrid structures would enrich their functionalities and widen their potential application prospects in many aspects, such as heterogeneous catalysis, adsorption, ion exchange, and as supports [6–9]. To improve application performance of phosphonated tin oxides, incorporating porosity to fabricate materials with high specific surface areas would be a feasible strategy. Surfactant templating routes have been developed to serve this purpose. For example, in 2005 Fujiwara et al. [10] fabricated mesoporous hybrid tin phenylphosphonate with a specific surface area of 371 m2/g in the presence of anionic surfactant sodium dodecylsulfate. In 2012, Dutta et al. [11] synthesized a hybrid porous tin (IV) phosphonate with SnCl55H2O as tin source, pentaethylenehexamine-o ctakis-(methyl phosphonic acid) hexadecasodium salt solution as the phosphonate source and cetyl trimethylammonium bromide as the structure directing agent. This material showed a BET surface area of 723 m2/g, and good catalytic activity in one-pot liquid phase oxidation of cyclohexanone to adipic acid under eco-friendly ⇑ Corresponding author. E-mail address: [email protected] (X. Lin). https://doi.org/10.1016/j.matlet.2018.10.086 0167-577X/Ó 2018 Elsevier B.V. All rights reserved.
conditions. In 2016, Wang et al. reported a category of porous tin (IV) phosphonates, showing specific surface areas in the range of 27–168 m2/g, in presence of nonionic templating agent F127 through solvothermal treatment [9]. For templating synthesis strategy, surfactant molecules have to be further removed whether by acid-ethanol extraction or by calcination. This method would make tedious synthetic steps and lead to increased cost, as well as bring environmental pollution in the post-treatment. Surfactant-free synthesis of such kind of materials with high specific surface area is intriguing. However, the related reports are rare [12]. In this research, high-yield porous phosphonated tin oxide (SnEDTMP) was hydrothermally synthesized with SnCl45H2O as tin source, ethylene diamine tetra(methylene phosphonic acid) (EDTMP) as the organophosphorus and NaOH as the pH regulator. Without any surfactant templating agent or organic solvent involved in the hydrothermal media, SnEDTMP with a high specific surface area of 377 m2/g could be obtained by simply controlling the adding order of NaOH in the synthetic process. 2. Materials and methods The typical synthesis of porous phosphonated tin oxides could be described as follows. An EDTMP aqueous solution was prepared by dissolving 0.76 g of EDTMP into 25 mL of water and then NaOH (0.56 g) was added under stirring. Afterward an aqueous solution of SnCl45H2O (1.78 g in 5 mL of H2O) was added dropwise. The mixture was continuously stirred at 45 °C for 2 h and then trans-
86
B. Guo et al. / Materials Letters 236 (2019) 85–88
ferred into an autoclave aging at 120 °C for 24 h. Finally white powder were collected by filtering, washed with water and dried at 80 °C. The as-prepared material was designated as SnEDTMP. For the purpose of comparison, SnO2 sample was prepared following the processing described above without addition of phosphorus. Details about the material characterizations can be found in Supplementary material.
3. Results and discussion
Fig. 1. XRD patterns of as-synthesized SnEDTMP and SnO2 fabricated.
The XRD pattern of SnEDTMP (Fig. 1) gives four typical peaks at 2h values of 26.38°, 33.76°, 37.61° and 51.4°, corresponding to (1 1 0), (1 0 1), (2 0 0), and (2 1 1) reflections, respectively, which evidently revealed the presence of a tetragonal SnO2 crystalline structure. The as-obtained SnEDTMP contained tetragonal SnO2
Fig. 2. (A) SEM image, low- (B, C) and high-resolution (D) TEM images, (E) the corresponding particle size distribution from (C), and (F) N2 adsorption-desorption isotherms, of SnEDTMP.
B. Guo et al. / Materials Letters 236 (2019) 85–88
0.330 nm, approximate to the d-spacing of 0.338 nm corresponding to (1 1 0) reflection at 2h = 26.38°. The amorphous area in Fig. 2D might be composed of Sn-O and organophosphonate species. In combination with the XRD analysis, clearly SnEDTMP had a porous structure with tetragonal SnO2 nanocrystals well dispersed within the hybrid framework. The N2 adsorption-desorption isotherms for SnEDTMP were of type II (Fig. 2F). A very strong increase of nitrogen-adsorbed volume was observed at high relative pressure (P/P0 > 0.6), which suggested the presence of an appreciable amount of secondary porosity of very large pores (even macropores) [13], consistent with the interparticle voids observed in SEM and TEM images. The isotherms displayed type H3 hysteresis loops that did not level off at relative pressures close to the saturation vapor pressure, indicating these materials comprised of aggregates of particles forming narrow slitlike pores. Its pore size distribution curve is shown in Fig. S1. SnEDTMP has a specific surface area of 377 m2/ g and a pore volume of 0.32 cm3/g. Without addition of EDTMP in the synthetic media, the SnO2 material prepared followed the same hydrothermal procedure only gives a specific surface area of 200 m2/g with a pore volume of 0.11 cm3/g. Apparently, organophosphonate groups played an important role in enlarging the surface area of SnEDTMP. Interestingly, it was found that when NaOH was added at the last step in the synthetic process, the asprepared material showed a specific surface area as low as 131 m2/g with a lower yield of 0.844 g. The aggregation of nanoparticles from the corresponding SEM images showed a very tight stacking between nanoparticles with little space left (Fig. S2). Apparently, the sequence of NaOH added in the synthesis process played a key role in controlling the particles aggregation, and accordingly resulting in the difference in their specific surface areas. Fig. 3 shows the FT-IR spectrum of SnEDTMP. The strong and wide band at 3400 cm 1 and the sharp peak at 1630 cm 1 corresponded to the surface-adsorbed water and hydroxyl groups [14]. The weak bands around 2900–3000 cm 1 were assigned to the CAH stretching modes. The overlapped bands at 1460 and 1430 cm 1 were assigned to the CAH bending in ACH2A groups and the P-C stretching vibrations, respectively [15]. The small bands at 1373 and 1322 cm 1 could be attributed to phosphoryl (P@O) frequency and CAN stretching, respectively. The broad band around 1000–1200 cm 1 could be disintegrated into two peaks: one centered at 1163 cm 1 assigned to the P–CH2N = groups [16], another around 1049 cm 1 attributed to the P-O-Sn groups. The shoulder peak around 988 cm 1 assigned to PAOH groups were not observed, suggesting the P atoms are connected to Sn atoms
with the grain size about 6.6 nm, calculated based on the characteristic diffraction at 2h = 33.76° according to the Scherrer equation. In addition, the yield of SnEDTMP is high (1.905 g). Without addition of EDTMP in the synthetic process, the obtained SnO2 sample shows the same reflection as SnEDTMP does, but its intensity became much weaker, as well as giving a low yield (0.195 g). The evident difference in the yield for the synthetic cases with EDTMP added or not reveals that the organophosphonate groups were well contained in the SnEDTMP. According to analysis above, it could be concluded that the structure of SnEDTMP presented organophosphonated SnO2 hybrid framework. The SEM image of SnEDTMP in Fig. 2A presents the aggregation of irregular-shaped particles with submicrometer size. It’s noted that their aggregation produced void spaces with pore sizes in the range of 20–240 nm. The TEM image in Fig. 2B demonstrated the agglomeration of small particles, rendering rich inter-spaces of several tens to hundred nanometers between these particles. This result was consistent with that observed from the SEM image (Fig. 2A). In the enlarged micrograph, abundant clear crystal fringes of the nanocrystals with an average diameter of 6.0 nm are observed (Fig. 2C and E), which was close to that (6.6 nm) calculated from XRD pattern. As shown in the high-resolution TEM micrograph (Fig. 2D), the measured interlayer spacing was
Fig. 3. FT-IR spectrum of SnEDTMP.
Fig. 4. Solid-state
13
C and
87
31
P MAS NMR spectra of SnEDTMP.
88
B. Guo et al. / Materials Letters 236 (2019) 85–88
through PAOASn bonding, and the majority of multidentate phosphonic acid were coordinated to tin atoms. Fig. 4 gives the 13C and 31P MAS NMR spectra of SnEDTMP. The sole signal at 52.98 ppm from 13C MAS NMR spectrum was detected, and it could be attributed to the carbon atoms in the nitrilomethylenephosphonate group (Fig. 4a). 31P MAS NMR spectroscopy was conducted to provide useful information on the nature of the phosphonate groups. The overlapped chemical shift of SnEDTMP in Fig. 4b could be split into signals at 11.73 and 2.25 ppm, ascribed to the P nuclei in the form of RPO2(OH) and RPO3, respectively [17]. No signal attributed to RPO(OH)2 species was observed in the 31P MAS NMR spectroscopy. The change of phosphorus environment in the hybrid material indicates the formation of PAOASn bonds, and also suggests that the organic moieties were well integrated into the hybrid framework, instead of pendent on the pore wall. A combination of all above analyses demonstrated the successful synthesis of porous phosphonated tin oxides with high specific surface areas.
4. Conclusions High-yield porous phosphonated tin oxides (SnEDTMP) were successfully synthesized via a facile hydrothermal route without adding any surfactant templating agent or organic solvent. It was found that the adding sequence of NaOH as the pH regulator in the synthetic process could effectively dominate the aggregation of nanoparticles whether to form a tight stacking or loose accumulation. As a result, SnEDTMP with a high specific surface area of 377 m2/g was obtained when NaOH was added in the first step during the synthetic process. Yet the sample had a rather lower specific surface area of 131 m2/g when NaOH was added at last step.
Acknowledgements This work was supported by the National Natural Science Foundation of China (No. 21503041), Technology Planning Project of Guangdong Province (No. 2015B090927007) and Guangdong Provincial Key Plat-form and Major Scientific Research Projects for Colleges and Universities (No. 2015KCXTD029). Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.matlet.2018.10.086. References [1] J.P. Wilcoxon, J. Phys. Chem. B 104 (2000) 7334. [2] S. Gubbala, V. Chakrapani, V. Kumar, M.K. Sunkara, Adv. Funct. Mater. 18 (2008) 2411. [3] E.J. Kim, D. Son, T.G. Kim, et al., Angew. Chem. Int. Ed. 43 (2004) 5987–5990. [4] G. Derrien, J. Hassoun, S. Panero, B. Scrosati, Adv. Mater. 19 (2007) 2336–2340. [5] X.G. Han, M.S. Jin, S.F. Xie, et al., Angew. Chem. Int. Ed. 48 (2009) 1–5. [6] P. Bhanja, A. Bhaumik, ChemCatChem 8 (2018) 1607–1616. [7] S. Kirumakki, S. Samarajeewa, R. Harwell, et al., Chem. Commun. (2008) 5556– 5558. [8] S. Borah, B. Bhattacharyya, J. Deka, A. Borah, et al., Dalton Trans. 46 (2017) 8664–8672. [9] K. Lv, J. Han, C.T. Yang, et al., Chem. Eng. J. 302 (2016) 368–376. [10] N. Kishor Mal, M. Fujiwara, M. Matsukata, Chem. Commun. (2005) 5199–5201. [11] A. Dutta, M. Pramanik, A.K. Patra, M. Nandi, H. Uyama, A. Bhaumik, Chem. Commun. 48 (2012) 6738–6740. [12] A. Subiah, D. Pyle, A. Rowland, et al., J. Am. Chem. Soc. 127 (2005) 10826– 10827. [13] K.S.W. Sing, D.H. Everett, R.A.W. Haul, et al., Pure Appl. Chem. 57 (1985) 603– 619. [14] T.Z. Ren, Z.Y. Yuan, B.L. Su, Chem. Phys. Lett. 374 (2003) 170–175. [15] E. Jaimez, G.B. Hix, R.C.T. Slade, Solid State Ionics 97 (1997) 195. [16] D.E. Lόpez, J.G. Goodwin Jr., D.A. Bruce, J. Catal. 245 (2007) 381–391. [17] X.Z. Lin, Z.Y. Yuan, Eur. J. Inorg. Chem. (2012) 2661–2664.