Effect of the pH pre-adjustment on the formation of In2W3O12 and In6WO12 powders: Cluster coordination and optical band gap

Abstract

In this communication, diindium tristungstate (In₂W₃O₁₂) and hexaindium tungstate (In₆WO₁₂) powders were synthesized by the co-precipitation method at different pH values and heat treatment at 1073 K for 2 h. X-ray diffraction patterns and Rietveld refinements were employed to analyze the long-range structural ordering as well as quantify the presence crystallographic phases, respectively. All powders obtained in pH ranging from 2 to 3.5 exhibited a mixture of In₂W₃O₁₂ and gamma-tungsten oxide (γ-WO₃) phases. When prepared at pH close to 4, the powders presented both In₂W₃O₁₂ and In₆WO₁₂ phases. All these crystalline structures were simulated in a three-dimensional visualization program by adopting the Rietveld refinements analysis as input data. The ultraviolet–visible spectra were employed to estimate the optical band gap energy of all powders for the first time.

Resumen

En esta comunicación, los polvos tristungstate de diindio (In₂W₃O₁₂) y tungstato de hexaindio (In₆WO₁₂) se sintetizaron mediante el método de co-precipitación a diferentes valores de pH y tratamiento térmico a 1.073 K durante 2 h. Se emplearon patrones de difracción de rayos X y refinamientos de Rietveld para analizar el ordenamiento estructural de largo alcance y cuantificar las fases cristalográficas de presencia, respectivamente. Todos los polvos obtenidos en un pH de 2 a 3,5 mostraron una mezcla de fases In₂W₃O₁₂ y óxido de tungsteno gamma (γ-WO₃). Cuando se prepararon a un pH cercano a 4, los polvos presentaron las fases In₂W₃O₁₂ e In₆WO₁₂. Todas estas estructuras cristalinas se simularon en un programa de visualización tridimensional mediante la adopción del análisis de refinamiento de Rietveld como datos de entrada. Los espectros ultravioleta-visibles se emplearon para estimar la energía del intervalo de banda óptica de todos los polvos por primera vez.

Introduction

In the last years, trivalent tungstates, as diindium tristungstate [In₂(WO₄)₃ or In₂W₃O₁₂] and hexaindium tungstate (In₆WO₁₂), have been especially synthesized by using top down (TD) approaches [1–6]. The TD approach involves the breaking down of bulk materials up to achieve nanometer-size particles. The most popular TD methods employ mixture of precursor powders, repeated attrition or milling cycles, and heat treatments performed at high temperatures for long reaction times [7–11]. Despite the simplicity and easy handling, the final products are able to present large grains, reduced surface area, nonuniform particle shapes and sizes, broad particle size distribution, and a significant amount of impurities [12].

In order to minimize theses drawbacks, a good strategy to prepare crystalline In₂W₃O₁₂ or In₆WO₁₂ oxides is by means of bottom-up (BU) approaches. In BU approach, small building blocks (atomic and/or molecular species) are self-assembled to produced complex nanostructures, but with an efficient control on the particle sizes and shapes (nano-, meso-, or microscale) [13,14]. In general, trivalent tungstates (powder or thin film) have been synthesized by co-precipitation reaction, sol–gel, salt molten, and hydrothermal processing [15–18].

In terms of crystalline structure, In₂W₃O₁₂ exhibits an orthorhombic structure with space group (Pnca) at room temperature. On the other hand, at approximately 531 K, this tungstate crystallize in a monoclinic structure with space group (P2₁/a) [19,20]. A curious feature of In₂W₃O₁₂ is that its orthorhombic structure has a negative thermal expansion coefficient for temperatures above 523 K, while the monoclinic structure exhibit a positive thermal expansion coefficient for temperatures lower than 523 K [21].

Different electronic properties of both In₂W₃O₁₂ and In₆WO₁₂ oxides have been investigated, such as inorganic scintillators [22], luminescence [23], electrochromic [24], n-type semiconductor with low conductivity [25], trivalent-cation conductor [26], and photocatalysis [27]. However, in our understanding, the literature yet not reported the optical band gap behavior at room temperature of these kinds of trivalent tungstates.

Therefore, in this short communication, we report on the formation of In₂W₃O₁₂ and In₆WO₁₂ powders prepared by the co-precipitation method (at 363 K for 1 h) at different pH pre-adjustment values (from 2 to 4). The obtained amorphous precipitates were crystallized by means of heat treatments at 1073 K for 2 h. The structural ordering as well as crystalline phase identification and quantification were investigated in details by X-ray diffraction (XRD) and Rietveld refinement analysis, respectively. The ultraviolet–visible (UV–Vis) spectroscopy was a powerful tool in order to estimate the optical band gap energies of these tungstates.

Experimental details

Synthesis of In₂W₃O₁₂ and In₆W₃O₁₂ powders

The synthesis of In₂W₃O₁₂ and In₆W₃O₁₂ powders is described as follows: 3 × 10⁻³ mol of sodium tungstate (VI) dihydrate (Na₂WO₄·2H₂O; 99.0% purity, Sigma–Aldrich) and 2 × 10⁻³ mol of indium (III) nitrate hydrate [In(NO₃)₃·H₂O; 99.99% purity, Sigma–Aldrich] were separately placed in two plastic tubes (Falcon, capacity of 60 mL) and dissolved with 50 mL of deionized water (DI-H₂O). The first solution, containing, Naaq+ and WO4aq2− ions, was transferred into a beaker and the volume was completed with 50 mL of DI-H₂O. This same procedure was adopted for the second solution containing Inaq3+ and NO3aq− ions. The pHmeter (QUIMIS®, Brazil) registered pH values of 9.6 and 3.1 for these solutions, respectively. The complete dissolution of their respective chemical precursors [Na₂WO₄·2H₂O and In(NO₃)₃·H₂O] is described in the following equations:

This experimental procedure was performed five times. Before mixing the solutions for the formation of amorphous precipitates, we perform a fine pH adjustment in the first solution by means of nitric acid (HNO₃; 70% purity, Meck). The pH values used in this solution were 2; 2.5; 3; 3.5 and 4. Similar the first solution, ammonium hydroxide solution (NH₃·H₂O; 30% purity, Vetec) and HNO₃ were employed to adjust the pH of the second solution. In this case, NH₃·H₂O to acquire pH at 3.5 and 4.0, while the HNO₃ for pH from 2 to 3. Finally, after these pH adjustments, the second solution (100 mL) was mixed with the first solution (100 mL) in a same pH condition. This mixture was maintained under constant stirring at 363 K for 1 h. After this process, the formation of a white suspension with precipitates was verified. This reaction is described by Eq. (3):

These suspensions containing white precipitates were washed fifteen times (water and acetone), centrifuged (8500 rpm for 10 min), and subsequently dried in a conventional furnace for 338 K for 10 h. Finally, the amorphous precipitates were calcined at 1073 K for 2 h, maintaining a heating rate of 274 K/min (Electronic Supplementary Material with XRD patterns and TGA analysis of both In₂W₃O₁₂ and In₆W₃O₁₂ powders obtained at pH 4).

Characterizations of In₂W₃O₁₂ and In₆W₃O₁₂ powders

The structural analysis of all powders was performed by means of XRD patterns by using a LabX XRD-6000 diffractometer (Shimadzu®, Japan) with CuKα radiation (λ = 0.154184 nm). Data were collected over 2θ ranging from 10° to 40° with a scanning scan rate and step size of 2°/min and 0.02°, respectively. The PowderCell software [28] was employed to identify all crystallographic phases. For Rietveld analysis, XRD patterns were acquired by using a D/Max-2500PC diffractometer (Rigaku, Japan) with CuKα radiation. Data were collected over 2θ ranging from 10° to 110° with a scanning scan rate and step size of 0.2°/min and 0.02°, respectively. Ultraviolet–Visible (UV–Vis) spectra were measured on a Cary 5G Spectrophotometer (Varian, USA) operated in diffuse-reflectance mode.

Results and discussion

Structural analysis of In₂W₃O₁₂ –In₆W₃O₁₂ powders

Figs. 1(a, b) illustrate XRD patterns of all powders synthesized by the co-precipitation with different pH values and heat treated at 1073 K for 2 h.

XRD patterns were employed to identify the crystallographic phases and analyze the degree of crystallization of In₂W₃O₁₂ powders. Firstly, two different crystalline phases were identified in XRD patterns for the powders synthesized with pH ranging from 2 to 3.5 (Fig. 1(a)). The majority phase was indexed to In₂W₃O₁₂ with a monoclinic structure and space group (P2₁/a) (“Inorganic Crystal Structure Data” (ICSD) No. 99607) [5], while the secondary phase was ascribed to the polymorph of γ-WO₃ with monoclinic structure and space group (P2₁/n) (ICSD No. 50727) [29]. However, when the pH was adjusted at 4, diffraction peaks related to γ-WO₃ disappeared completely (Fig. 1(b)). Besides the In₂W₃O₁₂, the appearing of a new phase indexed to In₆WO₁₂ with rhombohedral structure and space group (R3¯) was clearly detected in this system (ICSD No. 27197) [7]. In principle, pH values superior to 4 considerably favors the crystallization of In₆WO₁₂. All these XRD patterns confirm that the pH plays a fundamental role for the formation of the desired In₂W₃O₁₂ phase. All our results are consistent with the data previously published in the literature [30].

Fig. 1.

XRD patterns of In₂W₃O₁₂ and In₆WO₁₂ powders obtained with pH ranging from 2 to 3.5 (a) and pH at 4 (b) with posterior heat treatment at 1073 K for 2 h.

Quantitative analysis of In₂W₃O₁₂ and In₆W₃O₁₂ powders by Rietveld method

The indexing of the phases was conducted by using the structural characterization data obtained from XRD patterns (experimental) with the respective crystallographic files (theoretical) [ICSD No. 99607 (In₂W₃O₁₂), 50727 (γ-WO₃), and 27197 (In₆WO₁₂)] by using the ReX.Cell software [31]. In addition, the general structure analysis software (GSAS) software was used [32] to perform the Rietveld refinement analysis and quantify the crystalline phases in the samples.

The Rietveld refinement method is a least squares refinement procedure wherein the experimental step-scanned values are adapted to the calculated ones. The profiles are considered to be known, and a model for the crystal structure is available [33]. This structural refinement method presents several advantages over conventional quantitative analysis methods. As the method uses a whole pattern fitting algorithm, all lines for each phase are explicitly considered, and even severely overlapped lines are usually not a problem. Thus, it is not necessary to decompose patterns into separate Bragg peaks, as is often the case in traditional methods. The use of all reflections in a pattern rather than only the strongest ones minimizes both the uncertainty in the derived weight fractions and the effects of preferred orientation, primary extinction, and nonlinear detection systems [34].

The experimental lattice parameters, unit cell volumes, phase percentage, length of the chemical bonds, and bond angles between (OInO) and (OWO) in In₂W₃O₁₂, γ-WO₃, and In₆WO₁₂ powders were obtained and calculated by using the Rietveld refinement method [35,36] with the GSAS software [32]. The structural refinement results of In₂W₃O₁₂ powders obtained at different pH values and heat treated at 1073 K for 2 h are showed in Figs. 2(a–e). In addition, below the respective Rietveld refinements are illustrated in the pie charts indicating the weight percentage for each of the phases found in XRD patterns. The Rietveld refinement results are displayed in Tables 1a-1e (Electronic Supplementary Material with validation of CIF files and structure factors in PLATON).

Fig. 2.

Rietveld refinement plot of In₂W₃O₁₂–In₆WO₁₂ powders synthesized at (a) pH = 2; (b) pH = 2.5; (c) pH = 3; (d) pH = 3.5; and (e) pH = 4 with posterior heat treatment at 1073 K for 2 h. The weight percentage of all crystallographic phases for each sample is shown as pie chart.

Table 1a.

Rietveld refinement results for In₂W₃O₁₂ and In₆WO₁₂ powders synthesized by the co-precipitation method at (a) pH = 2, (b) pH = 2.5, (c) pH = 3, (d) pH = 3.5 and (e) pH = 4, and heat treated at 1073 K for 2 h.

Atoms	Wyckoff	Site	x	y	z
(a)
In1	4e	1	0.38100	0.96400	0.31700
In2	4e	1	0.37200	0.45700	0.05100
In3	4e	1	0.37200	0.47300	0.18700
In4	4e	1	0.10700	0.98400	0.41900
W1	4e	1	−0.00800	0.24900	0.48560
W2	4e	1	0.35620	0.11100	0.13040
W3	4e	1	0.14010	0.11300	0.25580
W4	4e	1	0.14900	0.61500	0.38260
W5	4e	1	0.35160	0.63000	0.21690
W6	4e	1	−0.00200	0.74900	0.01840
O1	4e	1	0.58170	0.38720	−0.00010
O2	4e	1	0.98820	0.40750	0.17150
O3	4e	1	0.82800	0.19370	0.09490
O4	4e	1	0.77440	0.49740	0.05440
O5	4e	1	0.52250	0.42530	0.14430
O6	4e	1	0.73210	0.51940	0.27280
O7	4e	1	0.42040	0.10990	0.41390
O8	4e	1	0.18430	0.28540	0.24640
O9	4e	1	0.54820	0.30880	0.44370
O10	4e	1	0.38770	0.30880	0.97530
O11	4e	1	0.06440	0.38180	0.06740
O12	4e	1	0.41150	0.36270	0.50100
O13	4e	1	0.41150	0.38290	0.23250
O14	4e	1	0.24840	0.03960	0.51280
O15	4e	1	0.11640	0.10650	0.33240
O16	4e	1	0.53270	0.94320	0.36260
O17	4e	1	0.74220	0.95970	0.19260
O18	4e	1	0.66540	0.93780	0.30070
O19	4e	1	0.96190	0.93390	0.31160
O20	4e	1	0.09450	0.31110	0.59390
O19	4e	1	0.15680	0.80620	0.39130
O22	4e	1	0.04510	0.65770	0.12100
O23	4e	1	0.36290	0.60270	0.12280
O24	4e	1	0.34380	0.81710	0.22570

W1	4e	1	0.25300	0.02600	0.28300
W2	4e	1	0.24600	0.03300	0.78110
O1	4e	1	0.00000	0.03000	0.21800
O2	4e	1	0.00200	0.46400	0.22100
O3	4e	1	0.28200	0.26400	0.27700
O4	4e	1	0.21400	0.25700	0.74200
O5	4e	1	0.27700	0.02800	0.00020
O6	4e	1	0.28800	0.49800	0.00000

[Phase 1: In₂W₃O₁₂; P2₁/a (14) – Monoclinic (a = 16.3624(8) Å, b = 9.6366(3) Å, c = 19.0316(8) Å; V = 2448.7(2) ų; β = 125.3149(3)°, Z = 8 and ρ = 5.27932 g/cm³)] and [Phase 2: γ-WO₃; P2₁/n (14) − Monoclinic (a = 7.3072(2) Å, b = 7.5376(2) Å, c = 7.6961(2) Å, V = 423.86(1) ų; β = 90.735(4)°, Z = 8 and ρ = 7.26599 g/cm³]; R_p = 5.88%; R_wp = 7.79%; R_Bragg = 4.12%; χ² = 2.89 and S = 1.70. Weights of the phases in the In₂W₃O₁₂ Monoclinic = 51.11% and γ-WO₃ Monoclinic = 48.89%.

Table 1b.

aaaa.

Atoms	Wyckoff	Site	x	y	z
In1	4e	1	0.38100	0.96400	0.31700
In2	4e	1	0.37200	0.45700	0.05100
In3	4e	1	0.12000	0.47300	0.18700
In4	4e	1	0.10700	0.98400	0.41900
W1	4e	1	−0.00800	0.24900	0.48560
W2	4e	1	0.35620	0.11100	0.13040
W3	4e	1	0.14010	0.11300	0.25580
W4	4e	1	0.14900	0.61500	0.38260
W5	4e	1	0.35160	0.63000	0.21690
W6	4e	1	−0.00200	0.74900	0.01840
O1	4e	1	0.58170	0.38720	−0.00010
O2	4e	1	0.98820	0.40750	0.17150
O3	4e	1	0.82800	0.19370	0.09490
O4	4e	1	0.77440	0.49740	0.05440
O5	4e	1	0.52250	0.42530	0.14430
O6	4e	1	0.73210	0.51940	0.27280
O7	4e	1	0.42040	0.10990	0.41390
O8	4e	1	0.18430	0.28540	0.24640
O9	4e	1	0.54820	0.35360	0.44370
O10	4e	1	0.38770	0.30880	0.97530
O11	4e	1	0.06440	0.38180	0.06740
O12	4e	1	0.41150	0.36270	0.50100
O13	4e	1	0.85750	0.38290	0.23250
O14	4e	1	0.24840	0.03960	0.51280
O15	4e	1	0.11640	0.10650	0.33240
O16	4e	1	0.53270	0.94320	0.36260
O17	4e	1	0.74220	0.95970	0.19260
O18	4e	1	0.66540	0.93780	0.30070
O19	4e	1	0.96190	0.93390	0.31160
O20	4e	1	0.09450	0.31110	0.59390
O19	4e	1	0.15680	0.80620	0.39130
O22	4e	1	0.04510	0.65770	0.12100
O23	4e	1	0.36290	0.60270	0.12280
O24	4e	1	0.34380	0.81710	0.22570

W1	4e	1	0.25300	0.02600	0.28300
W2	4e	1	0.24600	0.03300	0.78110
O1	4e	1	0.00000	0.03000	0.21800
O2	4e	1	0.00200	0.46400	0.22100
O3	4e	1	0.28200	0.26400	0.27700
O4	4e	1	0.21400	0.25700	0.74200
O5	4e	1	.27700	0.02800	0.00020
O6	4e	1	0.28800	0.49800	0.00000

[Phase 1: In₂W₃O₁₂; P2₁/a (14) – Monoclinic (a = 16.3664(7) Å, b = 9.6372(3) Å, c = 19.0353(7) Å; V = 2449.7(1) ų; β = 125.3216(25)°, Z = 8 and ρ = 5.27708 g/cm³)] and [Phase 2: γ-WO₃; P2₁/n (14) – Monoclinic (a = 7.3106(3) Å, b = 7.5343(2) Å, c = 7.6991(2) Å, V = 424.05(2) ų; β = 90.613(7)°, Z = 8 and ρ = 7.26282 g/cm³]; R_p = 5.57%; R_wp = 7.46%; R_Bragg = 4.29%; χ² = 2.22 and S = 1.49. Weights of the phases in the In₂W₃O₁₂ Monoclinic = 57.96% and γ-WO₃ Monoclinic = 42.04%.

Table 1c.

aaaa.

Atoms	Wyckoff	Site	x	y	z
In1	4e	1	0.38100	0.96400	0.31700
In2	4e	1	0.37200	0.45700	0.05100
In3	4e	1	0.12000	0.47300	0.18700
In4	4e	1	0.10700	0.98400	0.41900
W1	4e	1	−0.00800	0.24900	0.48560
W2	4e	1	0.35620	0.11100	0.13040
W3	4e	1	0.14010	0.11300	0.25580
W4	4e	1	0.14900	0.61500	0.38260
W5	4e	1	0.35160	0.63000	0.21690
W6	4e	1	−0.00200	0.74900	0.01840
O1	4e	1	0.58170	0.38720	−0.00010
O2	4e	1	0.98820	0.40750	0.17150
O3	4e	1	0.82800	0.19370	0.09490
O4	4e	1	0.77440	0.49740	0.05440
O5	4e	1	0.52250	0.42530	0.14430
O6	4e	1	0.73210	0.51940	0.27280
O7	4e	1	0.42040	0.10990	0.41390
O8	4e	1	0.18430	0.28540	0.24640
O9	4e	1	0.54820	0.35360	0.44370
O10	4e	1	0.38770	0.30880	0.97530
O11	4e	1	0.06440	0.38180	0.06740
O12	4e	1	0.41150	0.36270	0.50100
O13	4e	1	0.85750	0.38290	0.23250
O14	4e	1	0.24840	0.03960	0.51280
O15	4e	1	0.11640	0.10650	0.33240
O16	4e	1	0.53270	0.94320	0.36260
O17	4e	1	0.74220	0.95970	0.19260
O18	4e	1	0.66540	0.93780	0.30070
O19	4e	1	0.96190	0.93390	0.31160
O20	4e	1	0.09450	0.31110	0.59390
O19	4e	1	0.15680	0.80620	0.39130
O22	4e	1	0.04510	0.65770	0.12100
O23	4e	1	0.36290	0.60270	0.12280
O24	4e	1	0.34380	0.81710	0.22570

W1	4e	1	0.25300	0.02600	0.28300
W2	4e	1	0.24600	0.03300	0.78110
O1	4e	1	0.00000	0.03000	0.21800
O2	4e	1	0.00200	0.46400	0.22100
O3	4e	1	0.28200	0.26400	0.27700
O4	4e	1	0.21400	0.25700	0.74200
O5	4e	1	0.27700	0.02800	0.00020
O6	4e	1	0.28800	0.49800	0.00000

[Phase 1: In₂W₃O₁₂; P2₁/a (14) – Monoclinic (a = 16.3752(2) Å, b = 9.6415(6) Å, c = 19.0521(2) Å; V = 2454.65(2) ų; β = 125.309(6)°, Z = 8 and ρ = 5.26644 g/cm³)] and [Phase 2: γ-WO₃; P2₁/n (14) – Monoclinic (a = 7.3126(6) Å, b = 7.5398(7) Å, c = 7.7023(7) Å, V = 424.65(6) ų; β = 90.615(15)°, Z = 8 and ρ = 7.25259 g/cm³]; R_p = 7.60%; R_wp = 10.16%; R_Bragg = 3.20%; χ² = 1.99 and S = 1.41. Weights of the phases in the In₂W₃O₁₂ Monoclinic = 65.74% and γ-WO₃ Monoclinic = 34.26%.

Table 1d.

aaaa.

Atoms	Wyckoff	Site	x	y	z
In1	4e	1	0.38100	0.96400	0.31700
In2	4e	1	0.37200	0.45700	0.05100
In3	4e	1	0.12000	0.47300	0.18700
In4	4e	1	0.10700	0.98400	0.41900
W1	4e	1	−0.00800	0.24900	0.48560
W2	4e	1	0.35620	0.11100	0.13040
W3	4e	1	0.14010	0.11300	0.25580
W4	4e	1	0.14900	0.61500	0.38260
W5	4e	1	0.35160	0.63000	0.21690
W6	4e	1	−0.00200	0.74900	0.01840
O1	4e	1	0.58170	0.38720	−0.00010
O2	4e	1	0.98820	0.40750	0.17150
O3	4e	1	0.82800	0.19370	0.09490
O4	4e	1	0.77440	0.49740	0.05440
O5	4e	1	0.52250	0.42530	0.14430
O6	4e	1	0.73210	0.51940	0.27280
O7	4e	1	0.42040	0.10990	0.41390
O8	4e	1	0.18430	0.28540	0.24640
O9	4e	1	0.54820	0.35360	0.44370
O10	4e	1	0.38770	0.30880	0.97530
O11	4e	1	0.06440	0.38180	0.06740
O12	4e	1	0.41150	0.36270	0.50100
O13	4e	1	0.41150	0.38290	0.23250
O14	4e	1	0.24840	0.03960	0.51280
O15	4e	1	0.11640	0.10650	0.33240
O16	4e	1	0.53270	0.94320	0.36260
O17	4e	1	0.74220	0.95970	0.19260
O18	4e	1	0.66540	0.93780	0.30070
O19	4e	1	0.96190	0.93390	0.31160
O20	4e	1	0.09450	0.31110	0.59390
O19	4e	1	0.15680	0.80620	0.39130
O22	4e	1	0.04510	0.65770	0.12100
O23	4e	1	0.36290	0.60270	0.12280
O24	4e	1	0.34380	0.81710	0.22570

W1	4e	1	0.25300	0.02600	0.28300
W2	4e	1	0.24600	0.03300	0.78110
O1	4e	1	0.00000	0.03000	0.21800
O2	4e	1	0.00200	0.46400	0.22100
O3	4e	1	0.28200	0.26400	0.27700
O4	4e	1	0.21400	0.25700	0.74200
O5	4e	1	0.27700	0.02800	0.00020
O6	4e	1	0.28800	0.49800	0.00000

[Phase 1: In₂W₃O₁₂; P2₁/a (14) – Monoclinic (a = 16.3679(6) Å, b = 9.6402(3) Å, c = 19.0454(7) Å; V = 2451.63(9) ų; β = 125.3327(3)°, Z = 8 and ρ = 5.27292 g/cm³)] and [Phase 2: γ-WO₃; P2₁/n (14) – Monoclinic (a = 7.30724(2), b = 7.53764(2) Å, c = 7.69613(2) Å, V = 423.863(2) ų; β = 90.735(4)°, Z = 8 and ρ = 7.26628 g/cm³]; R_p = 5.46%; R_wp = 7.30%; R_Bragg = 3.30%; χ² = 1.59 and S = 1.26. Weights of the phases in the In₂W₃O₁₂ Monoclinic = 71.95% and γ-WO₃ Monoclinic = 28.05%.

Table 1e.

aaaa.

Atoms	Wyckoff	Site	x	y	z
In1	4e	1	0.38100	0.96400	0.31700
In2	4e	1	0.37200	0.45700	0.05100
In3	4e	1	0.12000	0.47300	0.18700
In4	4e	1	0.10700	0.98400	0.41900
W1	4e	1	−0.00800	0.24900	0.48560
W2	4e	1	0.35620	0.11100	0.13040
W3	4e	1	0.14010	0.11300	0.25580
W4	4e	1	0.14900	0.61500	0.38260
W5	4e	1	0.35160	0.63000	0.21690
W6	4e	1	−0.00200	0.74900	0.01840
O1	4e	1	0.58170	0.38720	−0.00010
O2	4e	1	0.98820	0.40750	0.17150
O3	4e	1	0.82800	0.19370	0.09490
O4	4e	1	0.77440	0.49740	0.05440
O5	4e	1	0.52250	0.42530	0.14430
O6	4e	1	0.73210	0.51940	0.27280
O7	4e	1	0.42040	0.10990	0.41390
O8	4e	1	0.18430	0.28540	0.24640
O9	4e	1	0.54820	0.35360	0.44370
O10	4e	1	0.38770	0.30880	0.97530
O11	4e	1	0.06440	0.38180	0.06740
O12	4e	1	0.41150	0.36270	0.50100
O13	4e	1	0.41150	0.38290	0.23250
O14	4e	1	0.24840	0.03960	0.51280
O15	4e	1	0.11640	0.10650	0.33240
O16	4e	1	0.53270	0.94320	0.36260
O17	4e	1	0.74220	0.95970	0.19260
O18	4e	1	0.66540	0.93780	0.30070
O19	4e	1	0.96190	0.93390	0.31160
O20	4e	1	0.09450	0.31110	0.59390
O19	4e	1	0.15680	0.80620	0.39130
O22	4e	1	0.04510	0.65770	0.12100
O23	4e	1	0.36290	0.60270	0.12280
O24	4e	1	0.34380	0.81710	0.22570

In1	3a	1	0.13970	0.32140	0.60440
W1	18f	1	0	0	0
O1	18f	1	−0.04260	0.09740	0.29690
O2	18f	1	0.40570	0.58380	0.82190

[Phase 1: In₂W₃O₁₂; P2₁/a (14) – Monoclinic (a = 16.3757(11) Å, b = 9.6365(4) Å, c = 19.0385(1) Å; V = 2450.56(2) ų; β = 125.3465(3)°, Z = 8 and ρ = 5.27522 g/cm³)] and [Phase 2: In₆WO₁₂; R3¯ (148) – Trigonal (a = b = 6.2344(1) Å, c = 6.2344 Å, V = 232.19(1) ų; Z = 1 and ρ = 7.61446 g/cm³]; R_p = 8.22%; R_wp = 10.93%; R_Bragg = 4.70%; χ² = 3.80 and S = 1.95. Weights of the phases in the In₂W₃O₁₂ Monoclinic = 55.15% and In₆WO₁₂ Trigonal = 44.85%.

The Rietveld refinement results exhibited a good agreement between observed XRD patterns and theoretical results, as illustrated in Fig. 2(a–e). As can be seen, the difference between experimentally observed XRD patterns and the theoretically calculated data was close to zero in terms of the scale of intensity, as indicated by the line (Y_Observed − Y_Calculated). In a first observation, for pH ranging from 2 to 3.5 in the aqueous solution, the powders heat treated at 1073 K for 2 h revealed a reduction of γ-WO₃ as well as an increase of In₂W₃O₁₂ with the increase of pH value (Figs. 2(a–e)). In contrast, the powders obtained with pH adjusted at 4 showed a mixture of In₆WO₁₂ and In₆WO₁₂ phases. This information implies in the difficulty of forming the desired pure In₂W₃O₁₂ phase.

Based on our experimental data, we propose the following reactions induced by the heat treatment at 1073 K for 2 h:

A possible reaction performed pH at 4 is related to the formation of solid In₂O₃ and WO₃ stoichiometric phases (both in amorphous state) when precipitated at 363 K for 1 h (Eq. (4)). More details on the Rietveld refinement analysis are presented in Tables 1a-1e.

In Tables 1a-1e, the statistical parameters (R_p, R_wp, R_Bragg, χy, and S) exhibited small deviations, which indicate the good quality of the structural refinements. The structural refinement data confirmed that all In₂W₃O₁₂ powders were crystallized in a monoclinic structure with space group (P2₁/a), point-group symmetry (C2h5), and eight molecular formula units per unit cell (Z = 8). The secondary or deleterious phase was identified as γ-WO₃, which has a monoclinic structure with space group (P2₁/a), point-group symmetry (C2h5), and eight molecular formula units per unit cell (Z = 8). Therefore, these powders are isostructural with a small difference in their space group due to an inversion in the center of symmetry, which leads to changes in (x, y, z) atomic coordinates [37].

The Rietveld refinement data confirmed that the In₂W₃O₁₂ powders synthesized at pH = 4 exhibited a deleterious phase related to In₆WO₁₂, which presents a rhombohedral structure with space group (R3¯), point-group symmetry (S₆), and one molecular formula unit per unit cell (Z = 1). Moreover, small variations in the atomic positions of In, W, and O atoms were observed because the atoms do not occupy fixed positions in this specific type of structure. These results indicate major variations in the position of O atoms because of distortion caused in the lattice, as demonstrated by the XRD technique. These distortions are able to influence in bond lengths and angles of InO or WO bonds, causing the formation of distorted octahedral [InO₆] clusters and tetrahedral [WO₄] clusters in In₂W₃O₁₂ structure.

Crystalline structure, symmetry, and coordination clusters of In₂W₃O₁₂, γ-WO₃, and In₆WO₁₂ lattices

Fig. 3(a–c) illustrates schematic representations of crystalline structures, symmetry, geometry, and coordination of In₂W₃O₁₂, γ-WO₃, and In₆WO₁₂ powders, respectively.

Fig. 3.

Schematic representation of (a) monoclinic In₂W₃O₁₂ unit cell, (b) monoclinic γ-WO₃ unit cell, and (c) rhombohedral In₆WO₁₂ unit cell, respectively.

All these structures were modeled using the visualization system for electronic and structural analysis (VESTA) software (version 3.4.3) [38,39] with the lattice parameters and atomic positions obtained from Rietveld refinement results (Tables 1a-1e) as input data. In In₂W₃O₁₂ structure (Fig. 3(a)), In atoms are coordinated to six O atoms, forming distorted octahedral [InO₆] clusters with symmetry group (O_h). These octahedron-type polyhedrons have 6-vertices, 8-faces, and 12-edges [40]. In addition, W atoms are coordinated to four O atoms, forming distorted tetrahedral [WO₄] clusters with symmetry group (T_d). These tetrahedron-type polyhedrons are composed of 4-vertices, 4-faces, and 6-edges [41]. For γ-WO₃ structure (Fig. 3(b)), all W atoms are bonded to six O atoms, resulting in distorted octahedral [WO₆] clusters. The existence of distortion in WOW bonds of [WO₆]–[WO₆] clusters can be easily noted in this structure.

Finally, In₆WO₁₂ structure has its In atoms coordinated to six O atoms, resulting in distorted octahedral [InO₆] clusters. Moreover, W atoms are coordinated to six O atoms, forming octahedral [WO₆] clusters (Fig. 3(c)). The Jahn–Teller (JT) effects can be actuating on octahedral [InO₆] clusters (In₂W₃O₁₂), octahedral [WO₆] clusters (γ-WO₃), and both octahedral [InO₆]/[WO₆] clusters (In₆WO₁₂) [42]. This JT effect is caused by the symmetry breaking in octahedral [InO₆]/[WO₆] clusters, in which the two axial bonds can be shorter or longer than the equatorial bonds. Consequently, this phenomenon reduces the symmetry and energy in the lattice [43]. In principle, In₂W₃O₁₂ powders are able to present variations in both (O–In–O)/(O–W–O) bond angles and lengths, causing distortions in octahedral [InO₆] and tetrahedral [WO₄] clusters with distinct degrees of order-disorder in the lattice.

UV–Vis spectra and optical band gap energy analysis of In₂W₃O₁₂ and In₆W₃O₁₂ powders

The optical band gap energy (E_gap) was calculated by using the Kubelka–Munk equation [44], which is based on the transformation of diffuse reflectance measurements to estimate E_gap values with good accuracy [45]. The Kubelka–Munk equation for any wavelength is described by the following Eq. (5):

where F(R_∞) is the Kubelka–Munk function or the absolute reflectance of the sample. The magnesium oxide was adopted as standard sample in reflectance measurements; R_∞ = R_sample/R_MgO (R_∞ is the reflectance), k is the molar absorption coefficient, and s is the scattering coefficient. The optical band gap and absorption coefficient of semiconductor oxides [46] can be calculated using Eq. (6):

where α is the linear absorption coefficient of the material, hν is the photon energy, C₁ is a proportionality constant, E_gap is the optical band gap, and n is a constant associated to different kinds of electronic transitions (n = 0.5 for direct allowed, n = 2 for indirect allowed, n = 1.5 for direct forbidden, and n = 3 for indirect forbidden). According to the literature [47–49], trivalent tungstates, such as Al₂W₃O₁₂, Gd₂W₃O₁₂, Y₂W₃O₁₂ and In₂W₃O₁₂ powders, have an optical absorption spectrum governed by direct or indirect electronic transitions. In direct electronic transitions, after the electronic absorption process, electrons located in the minimum energy states in the conduction band (CB) return to maximum energy states of the valence band (VB) at the same points in the Brillouin zone [50]. On the other hand, for indirect electronic transitions, the electronic process occurs at distinct points in the Brillouin zone [51]. Based on this information and literature [52], E_gap values of In₂W₃O₁₂ powders were calculated with n = 0.5 in Eq. (6). Finally, by using the absolute reflectance function in Eq. (5) with k = 2α, the modified Kubelka–Munk equation is described as in Eq. (7):

Therefore, by determining the value of F(R_∞) from Eq. (7) and plotting a graph of [F(R_∞)hν]² against hν, E_gap values were calculated for all powders by extrapolating the linear portion of UV–Vis curves.

Figs. 4(a–e) illustrate the UV–Vis spectra of all powders obtained at different pH values and heat-treated at 1073 K for 2 h. The evolution of E_gap as a function of different pH values is illustrated in Fig. 4(f). Insets show the digital photos of each powder.

Fig. 4.

UV–Vis spectra of In₂W₃O₁₂–In₆WO₁₂ powders synthesized at (a) pH = 2, (b) pH = 2.5, (c) pH = 3, (d) pH = 3.5, and (e) pH = 4, (f) evolution of E_gap values as a function of different pH values. Insets show the digital photos of each powder.

An important point to be considered is that there is a slight increase of E_gap values with the increase of pH condition for the powders heat treated at 1073 K for 2 h. This behavior is caused by the reduction of intermediary energy levels between the VB and CB. The origin these intermediary energy levels is due to defects arising from modifications in bond lengths and/or distortions in bond angles between O-In_O ([InO₆] clusters) or O-W-O bonds ([WO₆] clusters), respectively [53]. As identified in Rietveld refinement results (Tables 1a-1d), the increase of pH from 2 to 3.5 promoted a reduction of secondary phase (γ-WO₃) in In₂W₃O₁₂ powders. Therefore, there is a decreasing and reorganization of intermediary energy levels in the band gap arising from γ-WO₃, causing an increase of E_gap for In₂W₃O₁₂ powders Fig. 4(a–d). The literature has reported [54,55]E_gap values for γ-WO₃ phase in the range from 2.6 to 2.8 eV [16], while for In₆WO₁₂ phase an E_gap values of approximately 3.3 eV [27] as illustrated in Fig. 4(e, f). This large difference between the E_gap values of the In₂W₃O₁₂–WO₃ phases and In₂W₃O₁₂–In₆WO₁₂ phases illustrated in Fig. 4(d, e) suggest that the In₆WO₁₂ phase with the rhombohedral structure has a high energy value between the valence band and conduction band with respect to In₂W₃O₁₂ phase with the monoclinic structure, as presented in Table 2.

Table 2.

Comparative results between E_gap values of In₂W₃O₁₂–In₆WO₁₂ powders obtained in this study with those reported in the literature.

Method	Material	Shape	Temperature (K)	Time (h)	Egap (eV)	Ref.
Sol–gel	In6WO12	Thin film	873	1	4.1	[16]
Sol–gel	In6WO12	Thin film	973	1	3.8	[16]
Solid state reaction	In6WO12	Ceramics	1473	12	3.3	[27]
PWPPTEC	In6WO12	Crystal	–	–	2.0	[27]
–	In2W3O12	–	–	–	–	d
CP	In2W3O12–γ-WO3	Powder	1073	2	2.88
CP	In2W3O12–γ-WO3	Powder	1073	2	2.92
CP	In2W3O12–γ-WO3	Powder	1073	2	2.95
CP	In2W3O12–γ-WO3	Powder	1073	2	2.90
CP	In2W3O12–In6WO12	Powder	1073	2	3.73

SG = Sol–Gel; SSR = Solid state reaction; PWPPTEC = Plane-Wave Pseudo-Potential Total Energy Calculation; CP = Co-Precipitation; E_gap = optical band gap energy; Ref. = reference; d = Not yet reported in the literature and

= this work.

The results presented in this table indicate a highest E_gap value (3.73 eV) can be attributed to the synergic effect of intermediary energy levels arising from both In₂W₃O₁₂ and In₆WO₁₂ phases (Fig. 4(e, f)). The E_gap values obtained in our study are in good agreement with those previously published in the literature [16,27]. Therefore, we propose that this behavior is promoted owing to the simultaneous presence of these In₂W₃O₁₂ and In₆WO₁₂ complex phases and a change in the color of the powder from the light-green present in this phase mixture (In₂W₃O₁₂ and γ-WO₃) to white (inset Fig. 4(f)), which demonstrates a high reflectance and promotes a significant increase in the E_gap values.

Conclusions

In summary, In₂W₃O₁₂–In₆WO₁₂ powders were synthesized by the co-precipitation method at 363 K for 1 h with different pH values (from 2 to 4) and heat treated at 1073 K for 2 h. The pH pre-adjustment was not able to produce phase-pure In₂W₃O₁₂ powders. According to XRD patterns and Rietveld refinement result, In₂W₃O₁₂ powders formed with pH ranging from 2 to 3.5 exhibited the existence of secondary phase related to γ-WO₃ powders. On the other hand, the synthesis performed with pH at 4 resulted in crystalline powders composed of a mixture of In₂W₃O₁₂ and In₆WO₁₂ phases. In our systematic study, the results suggest that employing the methodology with pH pre-adjustment to obtain the monophasic In₂W₃O₁₂ powder is possibly found between (pH 3.5 and 4.0). The modifications in bond lengths and/or distortions in bond angles of octahedral [InO₆] clusters (In₂W₃O₁₂), octahedral [WO₆] clusters (γ-WO₃), and both octahedral [InO₆]/[WO₆] clusters (In₆WO₁₂) was identified by Rietveld refinement. A slight rise of E_gap values were identified with the increase of pH condition for the powders heat treated at 1073 K for 2 h. This phenomenon was related to decreasing and reorganization of intermediary energy levels in the band gap, as a consequence of the reduction of γ-WO₃ in In₂W₃O₁₂ powders (pH ranging from 2 to 3.5). The combination of intermediary energy levels arising from In₂W₃O₁₂ and In₆WO₁₂ phases resulted in the highest E_gap (at around 3.73 eV).

Acknowledgements

The authors acknowledge the financial support of the Brazilian research financing institutions: CNPq (479644/2012-8 and 304531/2013-8), FAPESP (2012/14004-5 and 2013/07296-2), and CAPES-PNPD (1268069).