Thin film of indium iron oxide, InFeO3, with a two-dimensional natural superlattice structure was proposed as one of the candidate photoelectrodes for visible- and ultraviolet-light-driven water splitting. A first principle calculation shows that InFeO3 is a charge transfer-type insulator with a bandgap energy (Eg) of 2.5 eV. Flat band potentials are calculated from Mulliken electronegativities, and they show an appropriately placed conduction band to enable H2 evolution. Despite its wide Eg, the photocurrent of InFeO3 in the visible region is comparable to that of α-Fe2O3 (Eg = 2.1 eV), a conventional oxide photoelectrode, possibly owing to its intrinsic structural anisotropy, reducing recombination losses of photogenerated electron–hole pairs. Furthermore, InFeO3 shows enhanced light absorption and photocurrent in the ultraviolet region compared with α-Fe2O3. The onset potential of photocurrent of InFeO3 is significantly lower than that of α-Fe2O3. These properties suggest that InFeO3 is a promising candidate for water splitting by visible and ultraviolet light.

Photoelectrochemical (PEC) water splitting has been studied extensively as a promising technique to produce clean and sustainable hydrogen fuel. A wide variety of semiconducting materials have been investigated as photoelectrodes for water splitting assisted by sunlight.1–4 Among them, iron oxides including α-Fe2O3 (hematite) have attracted much attention because of their high response to visible light, together with the advantages of their natural abundance, electrochemical stability, and low toxicity.5–8 An indium iron oxide, InFeO3 (IFO), with a two-dimensional triangular lattice structure9,10 is also expected to exhibit a high PEC response to visible light, judging from the optical bandgap energies (Eg) of its related oxides such as InFeZnO4 (Eg = 2.86 eV) and InFeZn2O5 (Eg = 2.96 eV).11 However, there is no report on the PEC properties of InFeO3. The lattice of InFeO3 is represented as an alternation stack of In–O and Fe–O triangular lattice planes along the c-axis [see Fig. 1(a)]. It has been suggested that such an intrinsic self-polarized structural anisotropy is expected to promote the separation of photogenerated hole–electron pairs,11,12 possibly resulting in the enhanced photoelectrochemical properties. We have focused on the InFeO3-based photoelectrode, and its PEC properties are investigated in comparison with those of α-Fe2O3. The unique electronic and band structure resulting from such a structural anisotropy makes IFO potential materials for addressing recombination issues associated with semiconductor photoelectrodes.

FIG. 1.

XRD 2θω scans for the (a) IFO and (b) α-Fe2O3 films grown on the TTO/α-Al2O3 (110) substrate. Crystal structures of α-Fe2O3 and IFO are shown next to their XRD patterns.

FIG. 1.

XRD 2θω scans for the (a) IFO and (b) α-Fe2O3 films grown on the TTO/α-Al2O3 (110) substrate. Crystal structures of α-Fe2O3 and IFO are shown next to their XRD patterns.

Close modal

Thin films of IFO were fabricated by pulsed laser deposition (PLD). A Ta-doped SnO2 (TTO; Ta content = 3.0 at. %) film grown on an α-Al2O3 (110) substrate was used as the bottom electrode for PEC measurements.13–15 The IFO and TTO targets were prepared via a standard solid-state reaction. The electrical resistivity of the TTO films was found to be ∼10−1 Ω cm at 300 K. The thickness of the TTO films was 120 nm. IFO films were deposited on the substrate at 700 °C under the oxygen pressure of 0.1 Pa. The as-grown films were annealed at 700 °C in air for 1 h. The α-Fe2O3 films were also prepared as the reference samples, using PLD under the same growth conditions as those of IFO. The thickness of the IFO and α-Fe2O3 layers was 100 nm and 90 nm, respectively. Structural properties of the films were analyzed by x-ray diffraction (XRD). The PEC measurements were carried out using a Solartron SI 1287 electrochemical interface analysis instrument (Solartron Corp.) operated with the CorrWare program in a three electrode configuration with a Pt counterelectrode, a KCl saturated Ag/AgCl reference electrode, and thin films of α-Fe2O3 or IFO as working electrodes in a quartz cell. The electrolyte consisted of N2 degassed 1.0M solution of NaOH in DI water (Milli-Q, Millipore Corp.; 18.2 MΩ) at pH 13.6, which was continuously degassed with N2 (in order to suppress the oxygen reduction reaction at the counterelectrode). Optical absorption measurements on the IFO and α-Fe2O3 films deposited directly on α-Al2O3 (110) substrates were performed using a UV-VIS spectrometer (JASCO, V-670). The first principles calculations of IFO were performed for the analysis of its band structure. We used the Advance/PHASE package program (Advancesoft Co., Ltd.), which employs projector augmented wave pseudopotentials within the generalized gradient approximation (GGA) of the density functional theory (DFT).16–18 Details of calculation are provided in the supplementary material.

Figure 1 shows 2θω XRD scans of the α-Fe2O3 and IFO films fabricated by PLD. The bottom TTO electrode layer was found to grow epitaxially along [101] on the α-Al2O3 substrate as previously reported.15,18 As shown in Fig. 1(a), the peaks of the IFO layer were observed at 14.5°, 29.3°, and 60.8°, which are assigned to (002), (004), and (006) of IFO, respectively. This suggests the anisotropic growth of IFO with a two-dimensional crystal structure along [001]. The lattice constants of the IFO film are calculated to be a = 3.31 Å and c = 12.2 Å, which are close to those reported for bulk IFO.10 In the XRD pattern of the α-Fe2O3 film grown on TTO, the only (110) and (220) peaks of the corundum crystalline structure [see Fig. 1(b)] are observed, indicating that the α-Fe2O3 film is also a single crystal with an orientation along [110]. High orientation of α-Fe2O3 along the [110] direction on the electrode is desirable because the anisotropic electrical conduction along [110] facilitates the collection of photocarriers in α-Fe2O3.12 The lattice constant of α-Fe2O3 is calculated to be a = 5.04 Å and c = 13.80 Å, which agree with the previously reported values.19,20

The calculated DOS of IFO is displayed in Fig. 2(a). The upper edge of the valence band is dominated mainly by O 2p states, whereas the conduction band minimum (CBM) is mainly composed of Fe 3d states. This electronic structure is similar to that of α-Fe2O318 and typical for the charge-transfer (CT) insulators.21,22 The VBM and CBM of IFO are separated by the energy gap (Eg) of 2.5 eV. The band edge states consist of the O 2p or Fe 3d orbitals, and the In 5s state is located at 5.0 eV, which is significantly far from the CBM. Given that the Fe 3d and O 2p states are somewhat hybridized at the VBM and CBM in the DOS of α-Fe2O3,18 they are hardly hybridized at band edges of IFO due to its two-dimensional superlattice structure. This result suggests that the photogenerated holes and electrons are spatially separated in IFO in contrast to α-Fe2O3, possibly resulting in the lowered recombination of hole–electron pairs and enhanced photocurrent in IFO. For the application of the photoelectrode for water splitting, the control of the position of the CBM is crucial from the viewpoint of energy consumption. We attempted to estimate the band edge positions of α-Fe2O3 and IFO based on the Mulliken electronegativity theory.23,24 The flat band potential Vfb is expressed as

(1)

where Ea and Eref are the electron affinity of the individual atom and the energy of free electrons in the hydrogen scale, Eref = 4.5 eV, respectively.25,26 The electronegativities of α-Fe2O3 and IFO can be calculated as the geometric mean of the electronegativities of their constituent atoms and expressed as follows:11,27–29

(2)
(3)

where Ea(α-Fe2O3), Ea(Fe), Ea(O), Ea(IFO), and Ea(In) are the electronegativities in the Mulliken scale of α-Fe2O3, iron, oxygen, IFO, and indium, respectively. For IFO, Vfb = 2.32 eV is obtained from Eqs. (1) and (3) using the values of Ea(Fe) = 4.06 eV, Ea(O) = 7.53 eV, Ea(In) = 3.1 eV, and Eg = 2.5 eV. Similarly, Vfb of α-Fe2O3 is calculated to be 2.48 eV from Eqs. (1) and (2) based on the values of Ea(Fe), Ea(O), and Eg = 2.2 eV. Figure 2(b) shows the band edge potentials of α-Fe2O3 and IFO determined from the results of the above calculations. The CBM of α-Fe2O3 lies at 0.3 eV positive of the hydrogen evolution potential, which is quite close to the previous theoretical and experimental estimates.30,31 In contrast, IFO can straddle the water redox potential with the conduction band at 0.2 V above the reduction potential of hydrogen, which is suitable for application in the PEC water splitting with a lower external bias.

FIG. 2.

(a) Electronic density of states (DOS) of IFO obtained by the DFT calculations. (b) Valence band and conduction band positions of α-Fe2O3 and IFO.

FIG. 2.

(a) Electronic density of states (DOS) of IFO obtained by the DFT calculations. (b) Valence band and conduction band positions of α-Fe2O3 and IFO.

Close modal

Figure 3 shows the optical absorption spectra of α-Fe2O3 and IFO grown on the α-Al2O3 (110) substrates. As for α-Fe2O3, the absorption peak at approximately 3.1 eV corresponds to the O 2p → Fe 3d CT optical transition,32,33 whereas the peak at approximately 2.4 eV is associated with the dd crystal field splitting transition.34,35 The spectrum of the IFO film shows a similar profile to that of α-Fe2O3. Judging from its bandgap energy and electronic band structure obtained by the DFT calculation, the absorption peak observed at around 3.2 eV can be attributed to the O 2p to Fe 3d CT optical transition. On the other hand, the small peak appearing at 2.4 eV in the spectrum of IFO can be attributed to the dd transition. The indirect bandgaps Eg of the as-deposited α-Fe2O3 and IFO films are estimated to be 2.2 eV and 2.5 eV, respectively, using the Tauc relationship αhν ∝ (Eg)2, where α, h, and ν denote the optical absorption coefficient, Planck’s constant, and frequency of light.36 These values agree with those expected by the above DFT calculation. It should be noted that the optical absorption of IFO significantly exceeds that of α-Fe2O3 in the energy region of E > 3.0 eV, which is the enhanced photocurrent of IFO in the ultraviolet region as will be mentioned later.

FIG. 3.

Optical absorption spectra of α-Fe2O3 and IFO films.

FIG. 3.

Optical absorption spectra of α-Fe2O3 and IFO films.

Close modal

Figures 4(a) and 4(b) show the current–potential curves of the films in 1M NaOH solution with and without light irradiation. α-Fe2O3 and IFO show similar values of photocurrent in the visible region. In the ultraviolet region, the photocurrent of IFO is significantly larger than that of α-Fe2O3 in the ultraviolet region. The incident photon to current efficiency (IPCE) values of the films are shown in Fig. 4(c). The IPCE is defined as IPCE (%) = 100 × (hc/e) × I (mA/cm2)/[P(mW/cm2) × λ(nm)], where h, c, e, I, P, and λ denote Planck’s constant, velocity of light in a vacuum, elementary charge, photocurrent density, power per unit area of the incident light, and wavelength of light, respectively.37 When λ < 400 nm, the IPCE of IFO is significantly larger than that of pure hematite, possibly owing to the enhanced light absorption of IFO in this wavelength region. On the other hand, the IPCE of IFO is close to that of α-Fe2O3 for λ > 400 nm despite larger Eg and smaller light absorption of IFO. The values of the applied bias photon to current efficiency (ABPE) are also shown in Figs. 4(d) and 4(e). The ABPE is defined as ABPE (%) = 100 × I (mA/cm2) × [1.23 − V(V vs RHE)]/[P(mW/cm2)], where V is the applied voltage.38 The ABPE values of IFO are significantly larger than those of α-Fe2O3 in the ultraviolet and visible regions. Although the dynamics of photocarriers in IFO remains to be experimentally analyzed, the enhanced PEC properties of IFO are possibly attributed to its electronic structure caused by structural anisotropy, as discussed above. The photocurrent onset potential is 0.65 V vs RHE for the IFO electrode that is shifted approximately −250 mV as compared to the α-Fe2O3 electrode, as shown in the insets of Figs. 4(a) and 4(b). Figures 5(a) and 5(b) show Mott–Schottky plots measured at a frequency of 100 Hz, 500 Hz, and 1000 Hz. The positive slopes of the plots indicate that α-Fe2O3 and IFO are n-type semiconductors. Our results indicate that the donor densities are 6.59 × 1017 cm−3 and 2.52 × 1017 cm−3 for α-Fe2O3 and InFeO3, respectively. Thus, the donor concentration of the films is not significantly affected by In substitution. The averaged flat band potential obtained from Mott–Schottky analysis was −450 mV and −200 mV for IFO and α-Fe2O3, respectively. This shift in the flat band potential of 250 mV to a more negative value compared to that of α-Fe2O3 is consistent with the observation of the photocurrent onset potential shift. This shift can be explained by an elevated conduction band minimum that is suggested by the results of the optical measurement and DFT calculation.

FIG. 4.

Photocurrent–potential (I–V) curves of the (a) α-Fe2O3 and (b) IFO photoelectrodes illuminated with visible (λ = 400–700 nm, 100 mW/cm2) and ultraviolet (λ = 300–400 nm, 34 mW/cm2) light. (c) IPCE spectra of the α-Fe2O3 and IFO photoelectrodes measured at 1.5 V vs RHE. [(d) and (e)] ABPE values of the films as a function of the applied potential bias under visible and ultraviolet light irradiation.

FIG. 4.

Photocurrent–potential (I–V) curves of the (a) α-Fe2O3 and (b) IFO photoelectrodes illuminated with visible (λ = 400–700 nm, 100 mW/cm2) and ultraviolet (λ = 300–400 nm, 34 mW/cm2) light. (c) IPCE spectra of the α-Fe2O3 and IFO photoelectrodes measured at 1.5 V vs RHE. [(d) and (e)] ABPE values of the films as a function of the applied potential bias under visible and ultraviolet light irradiation.

Close modal
FIG. 5.

Mott–Schottky plots of (a) α-Fe2O3 and (b) IFO photoelectrodes.

FIG. 5.

Mott–Schottky plots of (a) α-Fe2O3 and (b) IFO photoelectrodes.

Close modal

In summary, a highly oriented thin film of IFO with a two-dimensional natural superlattice structure was fabricated using pulsed laser deposition and its electronic and photoelectrochemical properties were investigated. The DFT calculations revealed that the VBM of IFO mainly consists of the O 2p band, whereas the CBM is composed of the Fe 3d band. Flat band potentials are calculated from Mulliken electronegativities, and they show an appropriately placed conduction band to enable H2 evolution. The IFO film shows higher photocurrent in the visible and ultraviolet regions compared to the α-Fe2O3 film besides its wider bandgap of 2.5 eV, which is possibly its higher electron mobility in the Fe–O two-dimensional planes. The onset potential of photocurrent is significantly lower than that of α-Fe2O3. These properties suggest that IFO is a promising candidate for solar water splitting without an external bias voltage. Further enhancement of PEC properties will be provided by chemical doping or substitution. The carrier density in Fe–O planes of IFO can be controlled with dopant elements, which may strongly affect its PEC properties. Specifically in IFO, charge carriers can be effectively doped without causing impurity scattering because the conducting layer and doping layer can be separated in the natural superlattice structure. Moreover, chemical substitution is effective for modulating the bandgap energy and band edge positions as indicated in many previous studies. We believe such control of electrical conduction and bandgap engineering in the layered oxides will pave the way for developing novel photoelectrodes for solar water splitting.

See the supplementary material for the details of preparation of photoelectrochemical cells and electronic structure calculations.

This work was supported by the TEPCO Memorial Foundation, JSPS KAKENHI, Grant Nos. JP15H03563 and JP16K14226, and the JSPS Core-to-Core Program, A. Advanced Research Networks.

1.
F.
Yilmaz
,
M. T.
Balta
, and
R.
Selbaş
,
Renewable Sustainable Energy Rev.
56
,
171
178
(
2016
).
2.
A. J.
Bard
and
M. A.
Fox
,
Acc. Chem. Res.
28
,
141
(
1995
).
3.
K.
Ye
,
H.
Li
,
D.
Huang
,
S.
Xiao
,
W.
Qiu
,
M.
Li
,
Y.
Hu
,
W.
Mai
,
H.
Ji
, and
S.
Yang
,
Nat. Commun.
10
,
3687
(
2019
).
4.
L.
Pan
,
Y.
Liu
,
L.
Yao
,
D.
Ren
,
K.
Sivula
,
M.
Grätzel
, and
A.
Hagfeldt
,
Nat. Commun.
11
,
318
(
2020
).
5.
W.
Ji
,
K.
Yao
,
Y.-F.
Lim
,
Y. C.
Liang
, and
A.
Suwardi
,
Appl. Phys. Lett.
103
,
062901
(
2013
).
6.
M.
Gong
and
H.
Dai
,
Nano Res.
8
,
23
(
2015
).
7.
K.
Sivula
,
F.
Le Formal
, and
M.
Grätzel
,
ChemSusChem
4
,
432
(
2011
).
8.
C. A.
Mesa
,
L.
Francàs
,
K. R.
Yang
,
P.
Garrido-Barros
,
E.
Pastor
,
Y.
Ma
,
A.
Kafizas
,
T. E.
Rosser
,
M. T.
Mayer
,
E.
Reisner
,
M.
Grätzel
,
V. S.
Batista
, and
J. R.
Durrant
,
Nat. Chem.
12
,
82
(
2020
).
9.
M.
Seki
,
T.
Konya
,
K.
Inaba
, and
H.
Tabata
,
Appl. Phys. Express
3
,
105801
(
2010
).
10.
I.
Nodari
,
A.
Alebouyeh
,
J. F.
Brice
,
R.
Gérardin
, and
O.
Evrard
,
Mater. Res. Bull.
23
,
1039
(
1988
).
11.
S. B.
Narendranath
,
A. K.
Yadav
,
D.
Bhattacharyya
,
S. N.
Jha
, and
R. N.
Devi
,
ACS Appl. Mater. Interfaces
6
,
12321
(
2014
).
12.
Z.-K.
Tang
,
W.-J.
Yin
,
L.
Zhang
,
B.
Wen
,
D.-Y.
Zhang
,
L.-M.
Liu
, and
W.-M.
Lau
,
Sci. Rep.
6
,
32764
(
2016
).
13.
J.-H.
Kwon
,
Y.-H.
Choi
,
D. H.
Kim
,
M.
Yang
,
J.
Jang
,
T. W.
Kim
,
S.-H.
Hong
, and
M.
Kim
,
Thin Solid Films
517
,
550
(
2008
).
14.
H.
Toyosaki
,
M.
Kawasaki
, and
Y.
Tokura
,
Appl. Phys. Lett.
93
,
132109
(
2008
).
15.
M.
Seki
,
H.
Yamahara
, and
H.
Tabata
,
Appl. Phys. Express
5
,
115801
(
2012
).
16.
W.
Kohn
and
L. J.
Sham
,
Phys. Rev.
136
,
B864
(
1964
).
17.
G.
Kresse
and
D.
Joubert
,
Phys. Rev. B
59
,
1758
(
1999
).
18.
A.
Sathe
,
M.
Seki
,
H.
Zhou
,
J. X.
Chen
, and
H.
Tabata
,
Appl. Phys. Express
12
,
091003
(
2019
).
19.
A.
Hamie
,
Y.
Dumont
,
E.
Popova
,
J.
Scola
,
A.
Fouchet
,
B.
Berini
, and
N.
Keller
,
J. Appl. Phys.
108
,
093710
(
2010
).
20.
A.
Serrano
,
J.
Rubio-Zuazo
,
J.
López-Sánchez
,
I.
Arnay
,
E.
Salas-Colera
, and
G. R.
Castro
,
J. Phys. Chem. C
122
,
16042
(
2018
).
21.
A.
Fujimori
,
M.
Saeki
,
N.
Kimizuka
,
M.
Taniguchi
, and
S.
Suga
,
Phys. Rev. B
34
,
7318
(
1986
).
22.
M.
Seki
,
M.
Takahashi
,
M.
Adachi
,
H.
Yamahara
, and
H.
Tabata
,
Appl. Phys. Lett.
105
,
112105
(
2014
).
23.
R. S.
Mulliken
,
J. Chem. Phys.
2
,
782
(
1934
).
24.
R. S.
Mulliken
,
J. Chem. Phys.
3
,
573
(
1935
).
25.
S. R.
Morison
,
Electrochemistry at Semiconductor and Oxidized Metal Electrodes
(
Plenum Press
,
New York
,
1980
), pp.
49
78
.
26.
M. A.
Butler
and
D. S.
Ginley
,
J. Electrochem. Soc.
125
,
228
(
1978
).
27.
A. J.
Nozik
,
Annu. Rev. Phys. Chem.
29
,
189
(
1978
).
28.
Y. F.
Kim
,
S. J.
Atherton
,
E. S.
Brigham
, and
T. E.
Mallouk
,
J. Phys. Chem.
97
,
11802
(
1993
).
29.
R. T.
Sanderson
,
Chemical Periodicity
(
Reinhold
,
New York
,
1960
).
30.
J.
Li
and
N.
Wu
,
Catal. Sci. Technol.
5
,
1360
(
2015
).
31.
M. J.
Katz
,
S. C.
Riha
,
N. C.
Jeong
,
A. B. F.
Martinson
,
O. K.
Farha
, and
J. T.
Hupp
,
Coord. Chem. Rev.
256
,
2521
(
2012
).
32.
R. F. G.
Gardner
,
F.
Sweett
, and
D. W.
Tanner
,
J. Phys. Chem. Solids
24
,
1183
(
1963
).
33.
A.
Duret
and
M. M.
Grätzel
,
J. Phys. Chem. B
109
,
17184
(
2005
).
34.
L. A.
Marusak
,
R.
Messier
, and
W. B.
White
,
J. Phys. Chem. Solids
41
,
981
(
1980
).
35.
D. M.
Sherman
and
T. D.
Waite
,
Am. Miner.
70
,
1262
(
1985
).
36.
J.
Tauc
,
R.
Grigorovici
, and
A.
Vancu
,
Phys. Status Solidi
15
,
627
(
1966
).
37.
A.
Kay
,
I.
Cesar
, and
M.
Grätzel
,
J. Am. Chem. Soc.
128
,
15714
(
2006
).
38.
Z.
Luo
,
C.
Li
,
S.
Liu
,
T.
Wang
, and
J.
Gong
,
Chem. Sci.
8
,
91
(
2017
).

Supplementary Material