Synthesis and infrared shielding properties of molybdenum-containing ammonium tungsten bronzes

Abstract

A series of molybdenum-containing ammonium tungsten bronzes was synthesized by a solvothermal process at 200 °C for 72 h using ethylene glycol, acetic acid, ammonium tungstate and ammonium molybdate as starting materials. All the diffraction peaks of XRD for the samples synthesized with starting Mo/W (mol) = 0–0.2 could be rough indexed as a hexagonal ammonium tungsten bronze with the formula of (NH₄)_0.33WO₃, which was accompanied by a shift of the main peaks after the introduction of molybdenum and a split peak for starting Mo/W (mol) = 0.1. The nitrogen content, based on both the thermogravimetry and X-ray photoelectron spectroscopy, rose with the increase of molybdenum content. Optical properties were evaluated using Fourier transform infrared spectroscopy, which proved that the molybdenum-containing ammonium tungsten bronzes could shield infrared rays better than the sample without molybdenum. The shield efficiency was related to the content of molybdenum. The optimum starting Mo/W (mol) to obtain high infrared shielding properties was 0.1. This sample consisted of uniform rods of ca. 50 nm in diameter emerging from the agglomerates.

---

1. Introduction

Tungsten bronzes are a group of nonstoichiometric compounds with the general formula M_xWO₃. In the most cases, the structure of tungsten bronzes is built up of layers containing corner-sharing WO₆ octahedra. The edges of the WO₆ octahedra form triangular, quadrangular, pentagonal or hexagonal channels, which are occupied randomly by cations.¹

It had been deeply researched due to the interesting physical and chemical properties, such as electronic and magnetic properties, superconductivity, gas sensors, and ionic conductivity.^2,3 In recent years, nanoparticles of hexagonal tungsten bronzes attract more and more attention because of the excellent near-infrared (NIR) shielding properties. Guo et al.^4–8 synthesized various nanoparticles of tungsten bronzes with different cations in the channels, such as Na⁺, K⁺, Rb⁺, Cs⁺, or NH₄⁺, by facile hydrothermal or solvothermal approach, and investigated their near-infrared shielding properties. Comparing with the conventional near-infrared shielding materials, such as nanoparticles of rare-earth hexaborides and tin-doped indium oxide (ITO),^9–11 tungsten bronzes have better performance of NIR shielding and visible light transmitting, and can be prepared with relatively facile routes.

Adachi and Sato^12–14 researched and discussed the origin of the NIR absorption of tungsten bronzes which is caused by a localized surface plasmon resonance. This phenomenon is observable for the one-dimensional nanoparticles because transverse and longitudinal surface plasmon resonances correspond to electron oscillations perpendicular and parallel to the length direction, respectively.^6,8 Moreover, amount of free electrons also affects the NIR shielding properties.¹⁵ In the channels of hexagonal tungsten bronzes, each cation contributes one electron to the tungsten conduction band.¹⁶ Consequently, the amount of cations, i.e., the value of x in M_xWO₃ influences the NIR shielding properties. In addition, the optical properties of nanosized particles also relate to the crystallographic defects.^17,18 On the basis of these theories, it is possible to improve the NIR shielding properties by doping of tungsten bronzes.

From the viewpoint of crystal chemistry, most of the reported researches about infrared shielding properties of tungsten bronzes focused on different cations in the channels. These tungsten bronzes have almost identical crystalline structures and similar chemical compositions, i.e., a series of isomorphism. There were few works about doping onto the tungsten site and the effect on the infrared shielding properties. In the present paper, molybdenum-containing ammonium tungsten bronzes were synthesized by a solvothermal process. At a certain molybdenum concentration, molybdenum-containing ammonium tungsten bronzes formed clear rod-like morphology and possessed homogeneous diameter. This sample had lower infrared transmittance than that of the ammonium tungsten bronze without molybdenum.

2. Experimental

0.5 g ammonium tungstate and a certain amount of ammonium molybdate were dissolved into 20 ml ethylene glycol at 190 °C. After cooling down this ethylene glycol solution to the room temperature, 10 ml acetic acid was added into the solution. Then, the obtained solution was transferred into an autoclave of 50 ml internal volume, followed by solvothermal reaction in an electric oven at 200 °C for 72 h. After the reaction, the powder was centrifuged, washed 4 times with water and ethanol, respectively, and finally dried.

The phase compositions of the samples were determined by X-ray diffraction analysis (XRD, Rigaku D/max 2200 PC) using graphite monochromatized Cu Kα radiation. The morphology and microstructure of the nanoparticles were investigated using scanning electron microscopy (SEM, Hitachi S-4800). The thermal behavior of the samples was investigated by the thermogravimetry and differential scanning calorimetry (TG-DSC, Netzsch STA449F3) from room temperature to 700 °C at a heating rate of 5 °C min⁻¹ in air. The surface composition and binding energy of the samples were determined by X-ray photoelectron spectroscopy (XPS, ESCALAB 250). Concentrations of elements were determined using inductively coupled plasma-atomic emission spectrometer (ICP-AES, SPECTRO ARCOS EOP).

To investigate the optical performance of molybdenum-containing (NH₄)_xWO₃ powder, an infrared measurement was employed by a Fourier transform infrared spectrometer (FT-IR, Tensor 27). The method is to grind a known mass ratio (100 : 1) of potassium bromide and the as-synthesized sample, and then pressed in a mechanical press with a constant pressure to form a pellet to ensure the results under the same condition.

3. Results and discussion

XRD analysis was employed to identify the phase composition and crystallographic structure of the samples. Fig. 1 shows the XRD patterns of the samples prepared with various Mo/W (mol) in the starting materials and the product synthesized by ammonium molybdate without ammonium tungstate under the same reaction conditions. Fig. 1a shows the XRD pattern of (NH₄)_xWO₃ without molybdenum. All the diffraction peaks of this sample could be well indexed as the hexagonal ammonium tungsten bronze with the formula (NH₄)_0.33WO₃ (JCPDS card no. 42-0452), and no characteristic peaks for impurities are observed. Higher diffraction intensity for the sample synthesized with starting Mo/W (mol) of 0.05 can be seen in Fig. 1b. Fig. 1c shows the sample prepared with starting Mo/W (mol) of 0.1. To understand XRD results more clearly, magnification patterns can offer easy viewing minor response for the introduction of molybdenum, as shown in Fig. 2. It is worth noting that the (002) diffraction peak of the product with Mo/W (mol) = 0.1 splits into two peaks as shown in Fig. 2b. It implies an existence of some species for this sample, although they are not able to be identified. Moreover, an obvious shift for both (002) and (200) diffraction peaks of the sample towards larger degree can be found comparing to the sample without molybdenum, as shown in Fig. 2a. It may be attributed to the doping. Fig. 1d and e show the samples synthesized with starting Mo/W molar ratios of 0.15 and 0.2, respectively. All the peaks are consistent with (NH₄)_0.33WO₃. A magnification result for the ratio of 0.20 as shown in Fig. 2c, a slight shift only for (200) is exhibited. In addition, an amorphous phase was obtained according to the pattern of the sample synthesized from ammonium molybdate without ammonium tungstate, as shown in Fig. 1f. It might be a kind of ammonium-containing molybdenum oxides. An in-depth discussion is to follow.

Fig. 1 XRD patterns of the samples with Mo/W (mol) = (a) 0; (b) 0.05; (c) 0.1; (d) 0.15; (e) 0.2 in the starting materials; and (f) sample synthesized from ammonium molybdate without ammonium tungstate.

Fig. 2 XRD patterns of the samples with Mo/W (mol) = (a) 0; (b) 0.1; (c) 0.2 in the starting materials.

To clarify the effects of the introduction of Mo on the characteristics of (NH₄)_xWO₃, the morphologies of the samples were observed by SEM, as shown in Fig. 3. Fig. 3a shows the sphere-like morphology of the product synthesized from ammonium molybdate without ammonium tungstate. These sphere-like particles should be the amorphous ammonium-containing molybdenum oxides combined with Fig. 1f. Nonuniform rod-like particles can be observed for the samples synthesized with starting Mo/W (mol) of 0 and 0.05, as shown in Fig. 3b and c. Fig. 3d displays uniform rods of ca. 50 nm in diameter emerging from the agglomerates for the sample synthesized with starting Mo/W (mol) of 0.1. Although there might be some unknown species combined with the results from XRD (see Fig. 2b), there is no obviously difference in the morphology for this sample. When the starting Mo/W (mol) is over 0.15, more and more sphere-like particles, which are similar to Fig. 3a, can be found (see Fig. 3e and f). Obviously, optimum amount of Mo, i.e., Mo/W (mol) is 0.1, is beneficial to an evolution of uniform rod-like ammonium tungsten bronze. Related the morphologies (see Fig. 3a, e and f) with the results from XRD (see Fig. 1d–f and 2c), the amorphous ammonium-containing molybdenum oxides were slipped through for the XRD analysis, however, caught up in the electron microscopy observation. With the increase of Mo content in the starting materials, the excessive ammonium molybdate is adverse to fabricate doped sample and makes the internal stress in the lattice partly relax. Thus, the shift for the sample with Mo/W (mol) = 0.2 is not strong enough than that of 0.1, as shown in Fig. 2b and c. Accompanying with a synchronous variation, it enables amorphous ammonium-containing molybdenum oxides separate out from the solid solution of Mo-containing ammonium tungsten bronzes, as shown in Fig. 3e and f. Consequently, the above results suggest that the starting Mo/W (mol) = 0.1 might be close to a saturation point for doping concentration, which leads to greater shift of diffraction peaks and maintains no sphere-like amorphous ammonium-containing molybdenum oxides.

Fig. 3 SEM images of the samples (a) without tungsten; and with Mo/W (mol) = (b) 0; (c) 0.05; (d) 0.1; (e) 0.15; and (f) 0.2 in the starting materials.

The thermal behaviors of the samples synthesized with starting Mo/W (mol) = 0–0.2 were further investigated. As shown in Fig. 4a, the weight loss of (NH₄)_xWO₃ without molybdenum from room temperature to 100 °C could be related to the loss of surface adsorbed water. After that, the weight loss up to about 270 °C could be attributed to the loss of structural water. The weight loss in the temperature range of 270–530 °C can be explained as the release of NH₃ and H₂O because of the decomposition of (NH₄)_xWO₃. Previous researches had confirmed that the powder remained after 550 °C was WO₃.⁸ Based on the weight of NH₃ decomposed from (NH₄)_xWO₃ and the weight of WO₃, the value of x (i.e., molar ratio of N/W) in (NH₄)_xWO₃ without molybdenum could be calculated in terms of its thermogravimetric curve to be 0.26 which is lower than the theoretical maximum value of 0.33. Similar result of x was reported.⁸ The trends of thermogravimetric curves for the products synthesized with starting Mo/W (mol) of 0.05 and 0.1 are similar to Fig. 4a. In the case of the sample with Mo/W (mol) = 0.05, the value was calculated to be 0.32, which tracks more closely with the theoretical value of 0.33. Furthermore, this value is 0.46, over the theoretical value of 0.33 for the sample with Mo/W (mol) = 0.1. However, there is an obvious difference after 600 °C for the thermogravimetric curves of samples with Mo/W (mol) over 0.15, as shown in Fig. 4b. It may be caused by the sublimation of molybdenum oxide based on the images of SEM and sublimation point of 600 °C for molybdenum oxide. The values of x for the samples synthesized with Mo/W (mol) = 0.15 and 0.2 are 0.54 and 0.56, respectively. Clearly, the content of nitrogen rises with the increase of the Mo concentration in terms of the thermogravimetric curves. Combined with the split and shift of the diffraction peaks in XRD (see Fig. 2b), the unreasonable value of 0.46 for Mo/W (mol) = 0.1 might be attributable to the molybdenum introduction. On the other hand, the values of 0.54 and 0.56 might be related to the amorphous ammonium-containing molybdenum oxides which are inclined to form molybdenum oxide during thermal decomposition.

Fig. 4 TG curves recorded in air at a heating rate of 5 °C min⁻¹ for the samples (a) without molybdenum and (b) with Mo/W (mol) = 0.15 in the starting materials.

The chemical composition and valence state for Mo-containing (NH₄)_xWO₃ were examined by XPS. Take the sample synthesized with starting Mo/W (mol) = 0.1 for example, the fully scanned spectrum clearly shows that elements of N, W, Mo and C exist in the sample (Fig. 5a). The presence of carbon element in the final product may be related to the residual chemically or physically adsorbed organics originating from the solvent. The XPS peak of N1s may be related to the group of NH₄⁺ in the ammonium tungsten bronze. For tungsten, a complex energy distribution of W4f photoelectrons was obtained as shown in Fig. 5b. The W4f core-level spectrum could be well fitted into two spin-orbit doublets, corresponding to two different oxidation states of W atoms. And, the Mo3d photoelectrons were also detected, as shown in Fig. 5c. The molar ratio of N/(W + Mo) on the surface is 0.43, which was determined by the atomic percentages of N, W and Mo. Additionally, the molar ratio of N/W of (NH₄)_xWO₃ without molybdenum is 0.24, which was determined likewise by XPS (elided). The above molar ratios, 0.24 and 0.43, are close to those calculated results of thermogravimetry. In a sense, this excessive nitrogen of 0.43 for the sample with Mo/W (mol) = 0.1 further confirms the existence of some unknown N-containing species.

Fig. 5 XPS spectra of (a) full range; (b) W4f core-level; and (c) Mo3d core-level for Mo-containing (NH₄)_xWO₃ synthesized with starting Mo/W (mol) = 0.1.

To determine the actual Mo concentration, both the sample and its mother liquid synthesized with starting Mo/W (mol) of 0.1 were examined using ICP-AES. In this way, the actual molar ratio of Mo/W is 0.159 for the sample. This value is greater than 0.1 for the starting materials. Correspondingly, Mo/W (mol) in the mother liquid is 0.016. Therefore, it could be calculated that the recovery ratios of W and Mo are about 59% and 94%, respectively. Thus, a higher recovery ratio of Mo leads to an increase in actual Mo/W molar ratio, which reaches 0.159 for the sample with starting Mo/W (mol) = 0.1.

To investigate the infrared shielding properties of Mo-containing (NH₄)_xWO₃, an infrared measurement was employed in a wide wavelength range. Fig. 6 shows the transmittances of infrared for samples synthesized with Mo/W (mol) in the starting materials between 0 and 0.2. Lower transmittance implies better infrared shielding properties. Fig. 6 exhibits that the sample without molybdenum has the highest transmittance in the spectrum of wave number of 4000–1000 cm⁻¹. This result indicates that the infrared shielding properties of Mo-containing (NH₄)_xWO₃ are superior to the product without molybdenum in this work. Transmittances of samples prepared with starting Mo/W (mol) of 0.05, 0.15 and 0.2 (Fig. 6b, d and e) are closely adjacent to each other and lower than that of (NH₄)_xWO₃ without molybdenum in sequence. Specially, the sample synthesized with starting Mo/W (mol) of 0.1 has the lowest transmittance, i.e., the best infrared shielding properties, as shown in Fig. 6c. The lower transmittance of Mo-containing (NH₄)_xWO₃ may be attributable to the introduction of crystallographic defects deriving from molybdenum, which might be associated with higher content of nitrogen. For the sample with Mo/W (mol) of 0.1, the molybdenum content might be close to the saturation point for the doping concentration, which contributes some more crystallographic defects. Thus more defects together with higher content of nitrogen achieve the best infrared shielding properties. If the Mo/W (mol) is over 0.15, excessive nitrogen tends to precipitate out the second phase of amorphous ammonium-containing molybdenum oxides. It had an effect on the infrared shielding properties.

Fig. 6 Infrared transmittances of the samples with Mo/W (mol) = (a) 0; (b) 0.05; (c) 0.1; (d) 0.15; and (e) 0.2 in the starting materials.

4. Conclusions

Mo-containing (NH₄)_xWO₃ was successfully synthesized by the solvothermal process. The results of XRD confirmed that Mo was doped into (NH₄)_xWO₃. When the actual Mo/W (mol) reached 0.159, i.e., the starting Mo/W (mol) of 0.1, the product was close to the saturated Mo doping concentration according to the results of XRD and SEM. And the value of x in (NH₄)_xWO₃ for this sample of Mo/W (mol) = 0.1 achieved 0.46 in terms of its thermogravimetric curve or 0.43 on the basis of XPS, was higher than the theoretical maximum of 0.33. Its morphology was assembled from more uniform rod-like particles in diameter of ca. 50 nm. Moreover, this sample exhibited the lowest infrared transmittance in a wide wavelength range comparing to all the samples with or without molybdenum.

Acknowledgements

The authors appreciate the financial support from the National Science Foundation of China (NSFC51372006); the Scientific Research Starting Foundation for Returned Overseas Chinese Scholars, Ministry of Education; the Start-Up Fund for High-End Returned Overseas Talents (Renshetinghan 2010, no. 411), Ministry of Human Resources and Social Security, China and the Lab-Installation Foundation of Beihang University for New Teachers.

Notes and references

1. I. M. Szilágyi, J. Madarász, G. Pokol, F. Hange, G. Szalontai, K. Varga-Josepovits and A. L. Tóth, J. Therm. Anal. Calorim., 2009, 97, 11.
2. S. Raj, T. Sato, S. Souma and T. Takahashi, Mod. Phys. Lett. B, 2009, 23, 2819.
3. N. N. Garif'yanov, V. Yu. Maramzin, G. G. Khaliullin and I. A. Garifullin, J. Exp. Theor. Phys., 1995, 80, 301.
4. C. S. Guo, S. Yin and T. Sato, Nanosci. Nanotechnol. Lett., 2011, 3, 413.
5. C. S. Guo, S. Yin, L. Huang and T. Sato, ACS Appl. Mater. Interfaces, 2011, 3, 2794.
6. C. S. Guo, S. Yin, Q. Dong and T. Sato, CrystEngComm, 2012, 14, 7727.
7. C. S. Guo, S. Yin, M. Yan and T. Sato, J. Mater. Chem., 2011, 21, 5099.
8. C. S. Guo, S. Yin, Q. Dong and T. Sato, Nanoscale, 2012, 4, 3394.
9. K. Adachi and M. Miratsu, J. Mater. Res., 2010, 25, 510.
10. H. Takeda, H. Kuno and K. Adachi, J. Am. Ceram. Soc., 2008, 91, 2897.
11. K. L. Purvis, G. Lu, J. Schwartz and S. L. Bernasek, J. Am. Chem. Soc., 2000, 122, 1808.
12. K. Adachia, J. Mater. Res., 2012, 27, 965.
13. Y. Sato, M. Terauchi and K. Adachi, J. Appl. Phycol., 2012, 112, 074308.
14. K. Adachi, Y. Ota, H. Tanaka, M. Okada, N. Oshimura and A. Tofuku, J. Appl. Phys., 2013, 114, 194304.
15. H. Takeda and K. Adachi, J. Am. Ceram. Soc., 2007, 90, 4059.
16. M. R. Skokan, W. G. Moulton and R. C. Morris, Phys. Rev. B: Condens. Matter Mater. Phys., 1979, 20, 3670.
17. C. S. Guo, S. Yin, M. Yan, M. Kobayashi, M. Kakihana and T. Sato, Inorg. Chem., 2012, 51, 4763.
18. C. S. Guo, S. Yin, Y. F. Huang, Q. Dong and T. Sato, Langmuir, 2011, 27, 12172.

---