We report a combined experimental and theoretical study on the optoelectronic properties of α-SnWO4 for UV-Vis excitation. The experimentally measured values for thin films were systematically compared with high-accuracy density functional theory and density functional perturbation theory using the HSE06 functional. The α-SnWO4 material shows an indirect bandgap of 1.52 eV with high absorption coefficient in the visible-light range (>2 × 105 cm−1). The results show relatively high dielectric constant (>30) and weak diffusion properties (large effective masses) of excited carriers.

The solar water splitting reaction involves many fundamental mechanisms that begin with solar light absorption, e/h+ generation/dissociation, diffusion of the charge carriers, and finally, water redox reactions.1–4 Visible light photoactive oxides, such as hematite,5 BiV O4,6 WO3,7 and a class of metal (oxy-)nitride materials8–13 are promising photocatalytic components for water splitting. Alternatively, Sn2+ incorporated tungsten oxide can produce a new SnWO4 material phase that exhibits an extended visible-light-response due to additional s-electrons from Sn2+.14–21 SnWO4 presents two polymorphs, the α- and β-phase, which have an orthorhombic and a cubic phase structure, respectively.15–17 Although the α-SnWO4 has an attractive narrow bandgap (∼2 eV), the photocatalytic and photoelectrochemical performance has been limited to date.16,19,21

In this study, we present a detailed experimental and theoretical investigation of the photophysical properties of α-SnWO4. The sputtered thin film configuration allows the accurate determination of its absorption coefficient, dielectric constant, and charge carrier effective masses. These experimental characterizations were systematically compared with first-principles calculations on the basis of density functional theory (DFT, including the perturbation approach DFPT) and employing the range-separated hybrid HSE06 exchange-correlation functional. The choice of the HSE06 functional was based on previous reports on widely used semiconductors in photocatalysis and photovoltaics, showing the high prediction accuracy when compared to experimental data.22–26 This systematic study on α-SnWO4 has never been reported in previous published works. The experimental and computational details used in this paper are described in the supplementary material.27 

SnWO4 films were deposited by direct current magnetron sputtering using W and SnO2 targets. They were synthesized with measured thicknesses of 80, 180, and 350 nm, and the samples are denoted as SnWO4 − (80), SnWO4 − (180), and SnWO4 − (350), respectively. The XRD results are presented in Fig. 1(a). Because an as-grown film does not indicate any XRD patterns from the material (see Fig. S1 of supplementary material),27 annealing treatments were performed under N2 to facilitate the formation of the crystalline structure. Although the N2 treatment at 500 °C produces only a small amount of crystal growth, annealing at 600 °C for 1 h significantly improved the crystallinity of the film (Fig. S1).27 Further treatment for 2 h did not improve the crystallinity, and thus, 1 h of treatment was chosen thereafter (Fig. S1).27 We indexed the (002), (111), (121), and (040) peaks according to the JCPDS 29-1354. The crystallization increases with an increased film thickness. In this particular system, the (121) plane is usually the most prominent peak,15,16 but in film growth, a preferential crystal orientation with the (0k0) plane was observed. SnWO4-(350) shows a small presence of a SnW3O9 new phase, as denoted by an asterisk in Fig. 1(a). The scanning electron microscopy images (Fig. S2)27 show that the annealing caused some grain formation and produced rougher surfaces compared with the as-grown sample. The DFT-optimized unit cell that was designed for this material is presented in Fig. 1(b). The crystal lattice of the α-SnWO4 is orthorhombic (space group Pnna) and composed of two-dimensional (2D) sheets of distorted WO6 octahedra that are separated by layers of Sn2+ ions, which are four-fold coordinated by O. Three distinct W–O bond lengths (2 × 1.802, 2 × 1.889, and 2 × 2.141 Å) and three different Sn–O bonds (2 × 2.184, 2 × 2.392, and 2 × 2.826 Å) were obtained. Our calculated lattice parameters (a = 5.592 Å, b = 11.632 Å, c = 4.983 Å, and α = β = γ = 90°) were found to be in excellent agreement with the measured XRD data.

FIG. 1.

(a) The XRD pattern of the α-SnWO4 thin films and (b) the DFT-optimized crystal structure of α-SnWO4. The blue, gray, and red atoms denote Sn, W, and O, respectively. The rigid frameworks highlighted in the structure represent the SnO4 tetrahedral and the WO6 octahedral species. Asterisk denote the SnW3O9 phase impurities.

FIG. 1.

(a) The XRD pattern of the α-SnWO4 thin films and (b) the DFT-optimized crystal structure of α-SnWO4. The blue, gray, and red atoms denote Sn, W, and O, respectively. The rigid frameworks highlighted in the structure represent the SnO4 tetrahedral and the WO6 octahedral species. Asterisk denote the SnW3O9 phase impurities.

Close modal

The computed electronic density of states (DOS) and energy dispersion diagram that were produced using the DFT/HSE06 method are shown in Fig. 2(a). The valence band states that are located within the 0-1.5 eV range below the Fermi level are dominated by a strong mixing of fully filled Sn 5s and O 2p orbitals. The conduction band states are mainly composed of empty W 5d orbitals. Our calculations predict this material to be an indirect-type (Γ-X) semiconductor with a lowest-energy bandgap of 1.52 eV that originates from the Sn 5s2 + O 2p6 → W 5d0 orbital transitions. Note that we found no spin-orbit coupling effect in the electronic structure of this compound.

FIG. 2.

(a) DOS and k-space band structure diagram of the α-SnWO4, as obtained using the DFT/HSE06 method. Color legend: Total DOS is shown in black, DOS projected on the Sn is shown in red, DOS projected on the W is shown in blue, and DOS projected on the O is shown in green. The top of the valence band, EVB, is represented by the horizontal dotted line. The Fermi level is set at 0 eV. (b) The absorption coefficient spectra of the SnWO4 that was measured with different thicknesses and calculated using the DFPT/HSE06 method.

FIG. 2.

(a) DOS and k-space band structure diagram of the α-SnWO4, as obtained using the DFT/HSE06 method. Color legend: Total DOS is shown in black, DOS projected on the Sn is shown in red, DOS projected on the W is shown in blue, and DOS projected on the O is shown in green. The top of the valence band, EVB, is represented by the horizontal dotted line. The Fermi level is set at 0 eV. (b) The absorption coefficient spectra of the SnWO4 that was measured with different thicknesses and calculated using the DFPT/HSE06 method.

Close modal

The UV-Vis transmittance and reflectance spectra of the films are shown in Fig. S3.27 The figure depicts a continuous decrease of the transmittance towards higher wavelengths with an increasing film thickness, as expected from the Beer-Lambert law.28 From both the transmittance and reflectance contributions by accurately accounting for the film thicknesses, we obtained the absorption coefficient of the material using the Lodenquai formalism.29 As shown in Fig. 2(b), the computed spectrum using the DFPT/HSE06 method is compared in Fig. 2(b) to the experimental spectra. It reveals the appearance of high-intensity absorption features (>4.5 × 105 cm−1) in the UV range and a lower-intensity absorption band (<2 × 105 cm−1) in the visible range with a broad edge extending up to 800 nm. Our calculated absorption spectrum is in excellent agreement with the experimental data obtained for SnWO4 − 80 and SnWO4 − 180 samples, while SnWO4 − 350 shows a different behavior in the 650–800 nm. This is probably due to the presence of secondary phase with lower absorption coefficient (see Fig. 1(a)). The weak edge absorption in the visible region suggests that the measured absorption coefficient (see Tauc plots for direct and indirect bandgaps; Fig. S4) is consistent with its calculated indirect bandgap (Fig. 2(a)).

The measured reflectance and transmittance were further utilized to obtain dielectric constant. The thin film was also subject to apply for electrode for Mott-Schottky analyses to obtain donor density, which can lead to estimates of effective mass for electrons. Detailed description of derivation can be found in the supplementary material27 (Figs. S5 and S6). The dielectric constant (electronic contribution) was 7.6 for SnWO4 − (80), as shown in Table I. The donor density was determined to be 1.8 × 1020 cm−3 (Fig. S6).27 Its electron effective mass was estimated to be m e * = 0 . 47 m 0 (Table I). The electronic contribution (ε) to the dielectric constants was also calculated using the DFT/HSE06 method, and the values are 6.2, 7.1, and 6.3 along the [100], [010], and [001] directions, respectively (see Table I). Our calculated average value (arithmetic mean) of 6.5 is in the same order of the experimental value. The ionic contribution (εvib) to the dielectric constant was also computed using the DFPT/Perdew-Burke-Emzerhof exchange-correlation functional (PBE) method, and much larger values of 29.1, 106.5, and 31.8 were obtained in the three principal directions with an average value of 55.8 (see Table I). This leads to a static dielectric constant of 62.3. The major contribution of the static dielectric constant that is due to the ionic portion originates from the strong ionic character of the crystal. We have also computed the effective masses of the photogenerated holes ( m h * ) and electrons ( m e * ) at the band edges using the finite difference method based on the band structure for α-SnWO4 in the three principal crystallographic directions. Our results (Table I) show that both holes and electrons have large effective masses along some directions, e.g., m h * = 1 . 76 m 0 and m e * = 1 . 23 m 0 along the [010] direction.

TABLE I.

Calculated electronic (ε) and ionic (εvib) contributions to the dielectric constants in the three principal directions of α-SnWO4, as determined using the DFT/HSE06 method, with the effective masses of the holes ( m h * ) and electrons ( m e * ) . The m0 is the free electron mass. The calculated values are compared to the current experimental data.

ε εvib m e * / m 0 m h * / m 0
Compound Expt. DFT DFT Expt. DFT DFT
α-SnWO4  SnWO4 − (80): 7.6  6.2 [100]  29.1 [100]  SnWO4 − (80): 0.47  0.18 [100]  0.67 [100] 
7.1 [010]  106.5 [010]  1.23 [010]  1.76 [010] 
6.3 [001]  31.8 [001]  0.20 [001]  0.44 [001] 
  6.5 (average)  55.8 (average)    0.35 (average)  0.80 (average) 
ε εvib m e * / m 0 m h * / m 0
Compound Expt. DFT DFT Expt. DFT DFT
α-SnWO4  SnWO4 − (80): 7.6  6.2 [100]  29.1 [100]  SnWO4 − (80): 0.47  0.18 [100]  0.67 [100] 
7.1 [010]  106.5 [010]  1.23 [010]  1.76 [010] 
6.3 [001]  31.8 [001]  0.20 [001]  0.44 [001] 
  6.5 (average)  55.8 (average)    0.35 (average)  0.80 (average) 

Fig. S7(A) presents the photoelectrochemical (PEC) measurements of the α-SnWO4 thin films.27 We used the cathodic scan and chose to start at 0.3 V vs. RHE to avoid any surface Sn2+ redox reactions. All of the films are photoresponsive under solar AM 1.5 G illumination. In terms of the photocurrent density, we have measured 40, 25, and 16 μA cm−2 for SnWO4 − (80), SnWO4 − (180), and SnWO4 − (350), respectively. The thinner film exhibits the highest photoactivity. As illustrated in Fig. S7(B),27 the photocurrent density of SnWO4 − (80) decreased after just one scan. It is known that Sn2+ easily oxides into Sn4+ (Refs. 30 and 31) and acts as a possible trap site of the electron. The PEC performances for the SnWO4 − (350) in Na2SO4 aqueous solution adjusted with H2SO4 to pH 3 and with NaOH to pH 13 are also shown in Fig. S8.27 We can notice a high improvement of the photocurrent with pH 13 compared to pH 3. Unfortunately, the material is not stable at this pH and the material dissolves after several scans. At this stage, the photocurrent is assigned to originate from material oxidation, likely Sn2+ to Sn4+. A protection layer, such as TiO2, is essential to stabilize the PEC performance for future study.32 The measured photocurrent densities were still very low compared to other visible active photoanodes, but for the flat band potential that was determined by the Mott-Schottky measurements (Fig. S6(B)),27 we found an interesting potential of −0.14 V vs. RHE that suggested that the photoanode has potential for the non-biased water splitting reaction.

As mentioned previously, the PEC performances are greatly influenced by the photophysical properties. Below, we separately discuss the absorption, dielectric constant (associated with charge separation), and effective mass (associated with charge transport). First, for the absorption coefficient, we measured 4 × 104 cm−1 at 600 nm. Thus, it requires 250 nm film thicknesses to sufficiently absorb the light irradiation in this wavelength range, which cannot explain the highest PEC performance of the thinnest material. Second, regarding the static dielectric constant, in previous experimental studies on common semiconductors, it has been demonstrated that a value of 10 or more is effective for exciton dissociation into free charge carriers.22,33 We measured an electronic contribution of 7.6 (the calculated one is 6.5) and calculated a value of 55.8 for the ionic contribution, which gives 62.3 for the static dielectric constant. Because this value is much higher than 10, this may indicate that α-SnWO4 has excellent dielectric properties that should reflect an efficient exciton dissociation.22 Finally, regarding the charge carrier effective masses, it has been shown that values smaller than 0.5 m0 are needed to obtain an effective mobility.22 The measured electron effective mass of 0.47 m0 for SnWO4 − 80 is in good agreement with the average calculated value (geometric mean) of 0.35 m0. The hole effective mass (average value) was predicted to be 0.8 m0. Moreover, the DFT calculations showed important anisotropies in both the hole and electron effective masses along the three principal orientations of the crystal, and there were much larger values than 0.5 m0 in the [010] direction. Consequently, our calculations and experiments show that α-SnWO4 exhibits a high probability of charge separation due to the anisotropic nature of the carrier effective masses, but it possesses low carrier mobility, which must be detrimental for the PEC activity. When the mobility is low, charge carriers can only travel for short distances. The high photocurrents observed with the thinner film may indicate that transport is the limiting factor. In addition to these aspects, the unstable oxidative character of Sn atoms in the α-SnWO4 system leads to low and unstable PEC performance. Many strategies can be adopted to improve the transport properties, such as metal doping,34 nanostructuration to decrease the film thicknesses (improve transport), while preserving a high film absorption35 or via surface cocatalyst functionalization,33 but this is beyond the scope of this paper.

In summary, a combined experimental and theoretical approach was used to determine the photophysical properties of α-SnWO4, including the bandgap, absorption coefficient, dielectric constant, and charge carrier effective masses. All of the measured and calculated properties are in excellent agreement. Despite the fact that α-SnWO4 presents interesting absorption properties and efficient charge carriers extraction, it suffers from weak transport properties such as high effective masses, which may explain the low PEC performance. The methodology presented in this study provides an excellent tool to deeply understand the photoelectrochemical performance of various materials at high accuracy.36 

2.
K.
Takanabe
and
K.
Domen
,
Green
1
,
313
(
2011
).
3.
D.
Nocera
,
Acc. Chem. Res.
45
(
5
),
767
(
2012
).
4.
S.
Cho
,
J.-W.
Jang
,
K.-H.
Lee
, and
J. S.
Lee
,
APL Mater.
2
,
010703
(
2014
).
5.
K.
Sivula
,
F.
Le Formal
, and
M.
Grätzel
,
ChemSusChem
4
,
432
(
2011
).
6.
T. W.
Kim
and
K.-S.
Choi
,
Science
343
,
990
(
2014
).
7.
R.
Solarska
,
R.
Jurczakowski
, and
J.
Augustynski
,
Nanoscale
4
,
1553
(
2012
).
8.
M.
Higashi
,
K.
Domen
, and
R.
Abe
,
J. Am. Chem. Soc.
134
,
6968
(
2012
).
9.
K.
Maeda
and
K.
Domen
,
J. Phys. Chem. C
111
,
7851
(
2007
).
10.
Y.
Li
,
L.
Zhang
,
A.
Torres-Pardo
,
J. M.
Gonzalez-Calbet
,
Y.
Ma
,
P.
Oleynikov
,
O.
Terasaki
,
S.
Asahina
,
M.
Shima
,
D.
Cha
,
L.
Zhao
,
K.
Takanabe
,
J.
Kubota
, and
K.
Domen
,
Nat. Commun.
4
,
2566
(
2013
).
11.
G.
Liu
,
J.
Shi
,
F.
Zhang
,
Z.
Chen
,
J.
Han
,
C.
Ding
,
S.
Chen
,
Z.
Wang
,
H.
Han
, and
C.
Li
,
Angew. Chem., Int. Ed.
53
,
7295
(
2014
).
12.
A.
Ziani
,
E.
Nurlaela
,
D. S.
Dhawale
,
D.
Alves-Silva
,
E.
Alarousu
,
O. F.
Mahammed
, and
K.
Takanabe
,
Phys. Chem. Chem. Phys.
17
,
2670
(
2015
).
13.
J. M.
Morbec
,
I.
Narkeviciute
,
T. F.
Jaramillo
, and
G.
Galli
,
Phys. Rev. B
90
,
155204
(
2014
).
14.
C.
Janáky
,
K.
Rajeshwar
,
N. R.
de Tacconi
,
W.
Chanmanee
, and
M. N.
Huda
,
Catal. Today
199
,
53
(
2013
).
15.
W.
Jeitschko
and
A. W.
Sleight
,
Acta Cryst.
B28
,
3174
(
1972
).
16.
I. S.
Cho
,
C. H.
Kwak
,
D. W.
Kim
,
S.
Lee
, and
K. S.
Hong
,
J. Phys. Chem. C
113
,
10647
(
2009
).
17.
A.
Kuzmin
,
A.
Anspoks
,
A.
Kalinko
,
J.
Timoshenko
, and
R.
Kalendarev
,
Sol. Energy Mater. Sol. Cells
(published online
2014
).
18.
J.
Ungelenk
and
C.
Feldmann
,
Chem. Commun.
48
,
7838
(
2012
).
19.
G.
Zhu
,
W.
Que
,
J.
Zhang
, and
P.
Zhong
,
Mater. Sci. Eng. B
176
,
1448
(
2011
).
20.
J. L.
Solis
,
V.
Lantto
,
L.
Haggstromb
,
B.
Kalska
,
J.
Frantti
, and
S.
Saukko
,
Sens. Actuators, B
68
,
286
(
2000
).
21.
K. J.
Pyper
,
T. C.
Evans
, and
B. M.
Bartlett
,
Chin. Chem. Lett.
26
,
474
(
2015
).
22.
T.
Le Bahers
,
M.
Rérat
, and
P.
Sautet
,
J. Phys. Chem. C
118
,
5997
(
2014
).
23.
M.
Harb
,
D.
Masih
,
S.
Ould-Chikh
,
P.
Sautet
,
J.-M.
Basset
, and
K.
Takanabe
,
J. Phys. Chem. C
117
,
17477
(
2013
).
24.
M.
Harb
,
D.
Masih
, and
K.
Takanabe
,
Phys. Chem. Chem. Phys.
16
,
18198
(
2014
).
25.
M.
Harb
,
P.
Sautet
,
E.
Nurlaela
,
P.
Raybaud
,
L.
Cavallo
,
K.
Domen
,
J.-M.
Basset
, and
K.
Takanabe
,
Phys. Chem. Chem. Phys.
16
,
20548
(
2014
).
26.
M.
Harb
,
J. Phys. Chem. C
119
,
4565
(
2015
).
27.
See supplementary material at http://dx.doi.org/10.1063/1.4930005 for detailed experimental methods and data.
28.
B.
Commoner
and
D.
Lipkin
,
Science
110
,
41
(
1949
).
29.
30.
Y.
Hosogi
,
Y.
Shimodaira
,
H.
Kato
,
H.
Kobayashi
, and
A.
Kudo
,
Chem. Mater.
20
,
1299
(
2008
).
31.
D.
Noureldine
,
D. H.
Anjum
, and
K.
Takanabe
,
Phys. Chem. Chem. Phys.
16
,
10762
(
2014
).
32.
S.
Hu
,
M. R.
Shaner
,
J. A.
Beardslee
,
M.
Lichterman
,
B. S.
Brunschwig
, and
N. S.
Lewis
,
Science
344
,
1005
(
2014
).
33.
Y.
Inoue
,
Energy Environ. Sci.
2
,
364
(
2009
).
34.
G.
Ma
,
T.
Hisatomi
, and
K.
Domen
,
Molecules to Materials
(
Springer International Publishing
,
2015
), Vol.
1
, ISBN: 978-3-319-13800-8.
35.
K. X.
Wang
,
Z.
Yu
,
V.
Liu
,
M. L.
Brongersma
,
T. F.
Jaramillo
, and
S.
Fan
,
ACS Photonics
1
,
235
(
2014
).
36.
K.
Takanabe
,
Top. Curr. Chem.
(published online
2015
).

Supplementary Material