Amorphous Fe2O3 for photocatalytic hydrogen evolution

Zhaoyong Lin , Chun Du , Bo Yan and Guowei Yang *
State Key Laboratory of Optoelectronic Materials and Technologies, Nanotechnology Research Center, School of Materials Science & Engineering, Sun Yat-sen University, Guangzhou 510275, Guangdong, P. R. China. E-mail:

Received 13th August 2019 , Accepted 2nd September 2019

First published on 3rd September 2019

Fe2O3 has drawn significant attention in photocatalysis due to its natural abundance, thermodynamic stability, environmental compatibility, low toxicity and narrow bandgap. Here, for the first time, we demonstrate that amorphous Fe2O3 nanoparticles can act as efficient and robust photocatalysts for solar H2 evolution without any cocatalysts. We also establish a plausible mechanism involving the amorphization-induced thermodynamic and dynamic behaviors of amorphous Fe2O3 upon photocatalytic hydrogen evolution. Thermodynamically, amorphization provides more surface states and larger carrier density, and thus elevates the conduction band edge to go across the H2 evolution potential level. Dynamically, amorphization-induced crystal field splitting weakening delocalizes the photogenerated carriers, and thus overcomes the excitation-wavelength-dependent small polaron trapping effect. These findings imply that amorphization may be a promising approach to functionalize and tailor other photocatalysts.


H2 is one kind of high-density, clean energy that can be extracted through solar-driven water splitting, which is named solar H2 evolution (SHE).1 The development of SHE relies on the discovery of efficient, stable and low-cost photocatalysts with high utilization of solar irradiation.2 Since O and Fe are respectively the first and fourth most abundant elements in the earth's crust,3 Fe2O3 has attracted great attention since 1976,4 only four years after the first discovery of the Fujishima–Honda effect at a TiO2 electrode.5 Compared with TiO2, Fe2O3 exhibits a narrower bandgap for visible light absorption, further leading to its popularity in photocatalysis.6

Due to its short carrier lifetime and unsuitable conduction band (CB) edge (lower than the H+/H2 potential level), SHE over Fe2O3 has almost been concentrated on photoelectrocatalysis (PEC) relying on an external bias to drive carrier separation and provide additional energy.7–9 However, the obtained photoconversion efficiency is still much lower than the theoretical maximum of 12.9%.10,11 Carneiro et al. ascribed it to the excitation-wavelength-dependent small polaron trapping of photoexcited carriers (ESPT),12 which results in a mismatch between the absorption and action spectra.13,14 The absorption in the visible light region is actually wasted absorption.15 Among them, nanotechnology is a valid way.16,17 On the one hand, photogenerated carriers can transport to the photocatalyst surface to take part in the reactions when the diffuse distance is less than the localization length (2–4 nm).18 On the other hand, nanoscaling can result in an up-shift of the CB edge of Fe2O3.19,20 Note that SHE through the process of powdery photocatalysts in suspension over Fe2O3 has rarely been realized.21,22

Thanks to its abundant coordination defects, amorphous Fe2O3 has made great contributions to the fields of sensors, storage media, solar energy transformation, electrocatalysis and water oxidation.23–26 Although amorphous materials are traditionally considered as photocatalytically inactive, they have emerged as a new class of photocatalysts based on various mechanisms.27–30 Here, for the first time, we demonstrate that amorphous Fe2O3 nanoparticles (AFNs) can act as efficient and robust photocatalysts for solar H2 evolution without any cocatalysts. In view of short-range order and structural flexibility (long-range disorder), amorphization of nanoparticles can be regarded as further nanoscaling, which elevates the CB edge to go across the H2 evolution potential level. AFNs are thermodynamically capable of accomplishing unassisted SHE. Dynamically, amorphization overcomes ESPT of photoexcited carriers, resulting in the non-zero photoconversion efficiency in the visible light region. Thus, these findings enable Fe2O3 to achieve SHE with high utilization of solar irradiation, and may be instructive for improving other photocatalysts.

Results and discussion

AFN preparation and characterization

AFNs were prepared by laser ablation in liquids (LAL), which has been developed for the amorphization of many metal oxides by our group before.30,31 Crystalline hematite micropowders (CHPs) were employed as the raw material, whose transmission electron microscopy (TEM) images and selected-area electron diffraction (SAED) pattern are shown in Fig. S1 (ESI). The X-ray diffraction (XRD) pattern in Fig. S2 (ESI) clearly indicates the hematite crystalline phase of the raw material (JCPDS No. 33-0664). From the TEM image in Fig. 1a, we can find that the LAL product is an aggregation of irregular nanoparticles. The average size of the nanodots is determined to be 3.71 nm according to the Gaussian fitting of the size distribution (Fig. S3, ESI). The high-resolution TEM (HRTEM) image in Fig. 1b shows no distinguishable lattice fringes, indicating the disordered atomic arrangement. It can be confirmed by the SAED pattern in the inset, which only shows a full halo ring without defined spots. Further, only a broad bread-shaped XRD peak centered at 35.5° appears in the XRD pattern, verifying the amorphous nature of the product.
image file: c9cy01621j-f1.tif
Fig. 1 Physical structure characterization of AFNs and CMNs. TEM (a) and HRTEM (b) images of AFNs. TEM (c) and HRTEM (d) images of CMNs. The corresponding SAED patterns are shown in the insets of (b) and (d), respectively. Raman (e) and FTIR spectra (f) of AFNs and CMNs. Red lines are AFNs and blue lines are CMNs.

The crystallinity can be increased by annealing the amorphous nanodots in N2 at 200 °C. In contrast, an XRD pattern consistent with the standard maghemite one (JCPDS No. 39-1346) was obtained. The peaks at 30.5, 35.5, 43.2, 57.3 and 62.7° are assigned to the (220), (311), (400), (511) and (442) facets of the maghemite phase, respectively. The TEM image (Fig. 1c) suggests that the annealed product is still an aggregation of nanodots. The slight increase of the average size from 3.71 to 4.20 nm (Fig. S4, ESI) should be due to the grain coalescence reported by Mendili et al.32 An apparent interplanar spacing of 0.342 nm can be observed in the HRTEM image (Fig. 1d), corresponding to the (211) crystallographic plane of maghemite. Clear diffraction spots in the SAED pattern (the inset) further indicate the higher crystallinity. The annealed sample should be crystalline maghemite nanoparticles (CMNs).

The Raman spectrum of CMNs in Fig. 1e matches well with the previously reported maghemite Raman spectra. The emission peaks at 408.5, 513.6, 704.6 and 1344.9 cm−1 are ascribed to the T2g, Eg, A1g and 2M (two-magnon) modes, respectively. The A1g mode is generally considered to be characteristic for Fe2O3 with a maghemite phase.33 The height ratio between the A1g and T2g modes approximates to 2, which is typical for crystalline maghemite.34 For the amorphous nanodots, it can be clearly found that the four peaks are all broadened, weakened and slightly shifted, which is due to the disordered atomic arrangement. The crystallite size-sensitive 2M mode nearly disappears, implying the rather small ordered domains.30 The A1g and T2g peak height ratio is about 1, deviating from the expected value of 2. This further suggests the amorphous nature of the product.35 Considering the matching between the two Raman spectra and the absence of the hematite characteristic mode at around 600 cm−1, the short-range atomic arrangement for Fe and O atoms in the LAL product should be similar to that in crystalline maghemite.7 Spectroscopically, the obtained amorphous Fe2O3 should be related to maghemite. It is known that the short-range order of an amorphous material plays a great role in its electronic structure and photocatalytic behavior.36,37 Actually, the long-range atomic arrangement in amorphous Fe2O3 should be disordered.33 Here, in terms of the short-range order, the obtained amorphous Fe2O3 nanoparticles are named AFNs.38

The Fourier transform infrared (FTIR) spectra in Fig. 1f provide more information. CMNs show two FTIR bands at 458.8 and 593.7 cm−1, which are the Fe–O–Fe stretching vibration modes (T1u) for maghemite.39 For AFNs, only one T1u mode appears in the FTIR spectrum. It is located in the vicinity of the two T1u modes of CMNs, and broadened significantly. This may be on account of the short-range order and long-range disorder of the amorphous materials. The other band ranging from 2500 to 3600 cm−1 should be assigned to the O–H stretching vibrations of the adsorbed H2O molecules. The weakening of the O–H band in CMNs should be related to the evaporation of H2O molecules by annealing. At this point, it can be concluded that AFNs and CMNs have been prepared.

Photocatalytic performance evaluation

Photocatalytic H2 evolution was first evaluated upon simulated solar irradiation (AM 1.5). Sodium sulfite (Na2SO3) with sodium sulfide (Na2S), triethanolamine (TEOA) and methanol (MeOH) were respectively used as the hole scavengers. As expected, no H2 was released over CMNs regardless of which hole scavenger was selected. However, AFNs achieved this goal. As shown in Fig. S5 (ESI), the concentrations of Na2SO3, TEOA and MeOH aqueous solutions for optimizing H2 evolution over AFNs are 0.2 M, 10 vol% and 15 vol%, respectively. Excess scavengers hinder the adsorption of H+ on the photocatalyst surface. Fig. 2a shows that the optimal H2 evolution amounts in 3 h are 27.13, 13.08 and 5.02 μmol for Na2SO3, TEOA and MeOH, respectively. It is known that the oxidation potentials for the three scavengers decrease in sequence. A higher oxidation potential leads to faster hole scavenging and weaker carrier recombination.40 This illustrates that the carrier recombination in AFNs is still present, which may be the reason for the sluggishness of the photocatalytic overall water splitting (OWS).
image file: c9cy01621j-f2.tif
Fig. 2 Comparison of the photocatalytic behaviors of AFNs and CMNs. a–c) Photocatalytic H2 evolution over AFNs (20.0 mg photocatalyst, 100 mL aqueous solution with different hole scavengers). a) Comparison of the optimal H2 evolution amounts using Na2S/Na2SO3 (0.2 M Na2SO3, 0.28 M Na2S), TEOA (10 vol%) and MeOH (15 vol%) as the hole scavengers for 3 h. b) Typical time courses of H2 evolution in Na2S/Na2SO3 (0.2 M Na2SO3, 0.28 M Na2S) aqueous solution upon the irradiation of simulated solar light (AM 1.5), ultraviolet light (UV), visible light (vis) and near-infrared light (NIR), respectively. c) Action (wavelength-dependent QE for H2 evolution) and absorption spectra. d–f) Photocatalytic O2 evolution over AFNs and CMNs (0.03 M AgNO3 aqueous solution). d) Typical time courses of O2 evolution. Action spectra for CMNs (e) and AFNs (f). The corresponding absorption spectra are added for comparison. The error bars represent the bandwidths of the filters.

According to the time course in Fig. 2b, we can find that H2 evolution proceeds continuously upon AM 1.5 irradiation over AFNs using Na2SO3 (0.2 M) as the hole scavenger. No H2 was detected when either the photocatalyst or the irradiation was removed from the system, suggesting that it is a photo-driven process in AFNs. The photocatalytic H2 evolution rate is determined to be 8.99 μmol h−1 by linearly fitting the plot. The photocatalytic test was repeated five times. As shown in Fig. S6a (ESI), the photocatalytic activity of AFNs after five cycles exhibits little deterioration. Fig. S6b (ESI) shows the long-term photocatalytic activity evaluation. H2 can still be produced at the end of 24 h irradiation. Moreover, AM 1.5 irradiation was divided into three regions: ultraviolet (UV, <420 nm), visible (vis, 420–760 nm) and near-infrared (NIR, >760 nm) regions, by respectively adding the corresponding optical filter to trigger the photocatalytic reactions. Irradiated by UV light individually, AFNs exhibit a photocatalytic rate of 6.62 μmol h−1. The rate becomes 3.07 μmol h−1 upon vis irradiation. H2 evolution stops when NIR irradiation is applied.

Further, wavelength-dependent quantum efficiency (QE) was measured to investigate the conversion of incident monochromatic photons to reacted electrons. For H2 evolution (two-electron process), QE is defined by

image file: c9cy01621j-t1.tif(1)

For the measurements, various monochromatic filters centered at 365, 405, 420, 500, 580, 600, 670 and 780 nm were used. The action spectrum for H2 evolution is shown in Fig. 2c, along with the absorption spectrum. Clearly, the QE at 780 nm is zero, which is consistent with previous results. Interestingly, the QE variation trend tracks the absorption spectrum well in the UV and vis regions, and QEs in the vis region are non-zero. It is divergent from the above discussion that the ESPT effect will result in a mismatch between the action and absorption spectra in Fe2O3.12–14

To further monitor the effective utilization of solar irradiation, photocatalytic O2 evolution was evaluated with an AgNO3 electron scavenger. As shown in Fig. 2d, both AFNs and CMNs can realize O2 evolution upon AM 1.5 irradiation with photocatalytic rates of 2.41 and 1.67 μmol h−1, respectively. The O2 evolution rate of AFNs is less than half of the H2 evolution rate because O2 evolution is a four-electron process that generally requires a larger overpotential.1 Similarly, the reactions can be triggered by UV irradiation over both AFNs and CMNs (1.34 and 1.59 μmol h−1). The photocatalytic rates upon NIR irradiation are zero, in accordance with the H2 evolution behavior. It should be noticed that vis irradiation plays different roles in O2 evolution over AFNs and CMNs. The O2 evolution rate of AFNs is 0.97 μmol h−1 whereas no O2 can be released by CMNs. The QE for O2 evolution was also analyzed by

image file: c9cy01621j-t2.tif(2)

According to Fig. 2e, the ESPT effect, that is, the mismatch between the action and absorption spectra, can be obviously found in CMNs. The absorption in the vis region is nearly wasted. In contrast, the action spectrum matches well with the absorption spectrum in AFNs. Based on the action spectra for H2 and O2 evolution, it can be concluded that the ESPT effect is avoided in AFNs whereas it is still present in CMNs.

Thermodynamic aspect

Thermodynamically, Fe2O3 is incapable of achieving unassisted SHE since its CB edge is lower than the H+/H2 potential level. In this case, photocatalytic H2 evolution was realized by AFNs. Therefore, the thermodynamic band structures of AFNs and CMNs should be investigated. Fig. 3a shows their absorption spectra. Clearly, absorption in UV, vis and NIR regions can be observed. According to previous reports, the bands in the UV region should be related to ligand-to-metal charge transfer (LMCT) transitions.41–43 Due to the crystal structure and coordination environment, Fe2O3 is known as a correlated electron system with crystal field splitting of Fe 3d orbitals into t2g and eg states.13 For CMNs, the two peaks centered at 252 and 384 nm are assigned to O 2p → t2g and O 2p → eg transitions, labelled LMCT1 and LMCT2, respectively.14 For the maghemite crystal structure, most of the Fe cations are octahedrally coordinated. The coupling of two adjacent octahedrally coordinated Fe cations will result in a double excitation process (DEP, Fe 3d → Fe3d) and absorption in the vis region.44 In addition, the NIR absorption should be associated with single ligand field intraband transitions (LF, Fe d–d transitions).
image file: c9cy01621j-f3.tif
Fig. 3 Thermodynamic investigation for AFNs and CMNs. a) Absorption spectra of AFNs and CMNs. Fe 2p XPS spectra of CMNs (b) and AFNs (c), and the multiplet fits of the Fe 2p3/2 core levels. d) VB XPS spectra of AFNs and CMNs. e) Mott–Schottky plots of AFNs and CMNs. f) Band structure diagram of AFNs (red) and CMNs (blue). The solid bands, semitransparent bands and dotted lines represent the conduction bands, valence bands and quasi-Fermi levels, respectively. The H2 evolution and O2 evolution potential levels are also shown by purple dashed lines.

Actually, a similar phenomenon can be found in the related literature reporting the absorption spectra of Fe2O3 materials. For example, Song et al. have prepared three Fe2O3 polymorphs. All of them exhibit LMCT, DEP and LF transitions as well.39 Generally, hematite shows obvious LMCT transitions.3,13,14,41 Due to the stronger Fe–Fe interaction, the DEP transition becomes greater for maghemite.33 Therefore, a big DEP peak appears in the absorption spectrum of CMNs in this work. It should be noticed that the absorption spectrum of a material may be influenced by many parameters, such as morphology and size. For example, Zhu et al. have also prepared hematite. In spite of the similar absorption edge at around 600 nm in the absorption spectrum, it shows no apparent boundary between LMCT and DEP transitions because of its hollow morphology.8 Moreover, Zou's group has demonstrated that the size makes a great difference in the absorption of hematite and maghemite.45

On account of the short-range order characteristic of amorphous materials, the absorption spectrum of AFNs is similar to that of CMNs with LMCT, DEP and LF transitions. Nevertheless, long-range disorder, that is, the loss of far-field symmetry, will lead to a difference in the density of state (DOS) and some novel phenomena. The undefined atomic arrangement generates undefined DOS distribution. Therefore, the sharp absorption peaks are broadened in AFNs with a smeared-out behavior. The Tauc plot slopes much more gently in Fig. S7 (ESI). This is the so-called tailing effect for amorphous materials, which narrows the bandgap from 1.82 to 1.69 eV. Moreover, we can find that LMCT transitions are enhanced whereas DEP and LF transitions are weakened in AFNs. This indicates the stronger Fe–O interaction.

The chemical environments of Fe and O atoms are monitored by X-ray photoelectron spectroscopy (XPS) measurements. Both of the high-resolution O 1s spectra in Fig. S8 (ESI) are deconvoluted into two peaks at 529.8 and 531.2 eV. They are respectively ascribed to the lattice oxygen (Fe–O) and chemisorbed oxygen (H2O). Compared with AFNs (33.2%), CMNs exhibit a lower H2O content (17.9%). This should be due to the annealing-induced evaporation. The high-resolution Fe 2p XPS spectra are shown in Fig. 3b and c. The peaks around 710 and 724 eV are assigned to Fe 2p3/2 and Fe 2p1/2 core levels, respectively. The others are the corresponding Fe3+ satellite peaks. In this work, the deconvolution of the Fe 2p3/2 core level peaks is performed based on the multiplet splitting proposed by Gupta and Sen (GS) rather than a simple splitting method.7,43 It has been demonstrated that the GS multiplet splitting can fit well most Fe 2p3/2 XPS spectra of Fe2+ and Fe3+ high-spin compounds by taking the crystal field effects into consideration.42 As shown in the Fe 2p XPS spectra, in addition to the GS multiplets (peaks 1–4), a low-binding energy pre-peak (708.4 eV), a high-binding energy surface peak (714.6 eV) and a shake-up satellite peak (718.2 eV) can be found. Both of the obtained envelopes are well consistent with the measured spectra.

The presence of a surface peak should be due to the large surface–bulk ratio of nanodots. For AFNs, the structural flexibility brings a larger surface–bulk ratio in the amorphous phase, and the surface peak content is increased from 11.2% to 18.3%, by a factor of 1.63.41 According to the size distribution in Fig. S3 and S4 (ESI), the ratio between the geometric specific surface areas (GSSAs) of AFNs and CMNs is calculated to be 1.13. Brunauer–Emmett–Teller (BET) measurements show that the BET surface areas for AFNs and CMNs are 77.4 and 39.2 m2 g−1, respectively, with a ratio of 1.97. It is larger than the GSSA ratio, indicating the more open structure of AFNs. This would be the reason for the larger surface peak content in AFNs.

For Fe2O3 materials, the Fe 2p3/2 spectra are sensitive to the structure symmetry and crystal field strength.43 The coordination environments of Fe3+ cations will influence the areas of the multiplet peaks.35 We can find that most of the multiplets are widened in AFNs, which should be due to the long-range disorder and the resulting complicated coordination environments. Moreover, they are also slightly blue shifted. Actually, amorphization leads to more surface states and dangling bonds in AFNs. The oxygen coordination defects reduce the electronegativity of the ligand around the Fe atoms and the shielding effect. Therefore, more energy is required to generate a photoelectron.44 In spite of the slight blue-shift, the spacings of the multiplets for AFNs are in accordance with the reported values for maghemite.42,44 This also implies that the obtained AFNs are spectroscopically assigned to maghemite in terms of the short-range order.

Valence band (VB) XPS measurements were performed (Fig. 3d). Obviously, amorphization transforms the two peaks of CMNs into a broad plateau for AFNs. It is also due to the long-range disorder, demonstrated by the tailing effect from 0 to 2 eV. The VB edge positions of AFNs and CMNs are determined to be 1.41 and 1.93 V (vs. NHE), respectively. Therefore, the CB edges are located at −0.28 and 0.11 V (vs. NHE), respectively. It should be noticed that the cavity at 4.38 eV for CMNs should be due to the above-mentioned crystal field splitting.33 In detail, the peaks at around 3 and 7 eV should be respectively assigned to the eg and t2g states of Fe 3d orbitals for octahedrally coordinated Fe cations. For AFNs, both of them are weakened, and the cavity is filled up. This could be due to the increased tetrahedrally coordinated Fe cations.45 The above Fe 2p XPS analysis shows that abundant surface states and dangling bonds are present in AFNs. These increased tetrahedrally coordinated Fe cations (coordination defects) should be due to the surface reconstruction. It also leads to the stronger Fe–O interaction, as demonstrated by the light absorption analysis.46 It is believed that the coordination environment has a great influence on the crystal field splitting. Long-range disorder endows AFNs with a new coordination environment and makes the energetic distribution of DOS wider. Therefore, crystal field splitting is weakened in AFNs, which is also demonstrated by the fact that the boundary between LMCT1 and LMCT2 in the absorption spectrum becomes much more blurred.

Further, Mott–Schottky analysis was conducted. Both of the Mott–Schottky plots show a positive slope, suggesting their n-type feature. The Fermi level can be roughly estimated by the flat-band potential, which is determined by the lateral axis intercept.47 Therefore, the quasi-Fermi levels for AFNs and CMNs are −0.03 and 0.60 V (vs. NHE), respectively. Carrier density Nc can be calculated using the equation:

Nc = 2e0−1ε−1ε0−1[d(C−2)/dV]−1(3)
where e0, ε0 and ε are the electron charge, vacuum permittivity and dielectric constant, respectively.48 Apparently, the carrier density of AFNs (1.24 × 1020 cm−3) is much larger than that of CMNs (4.10 × 1019 cm−3). This should result from the coordination defects and increased unpaired electrons.

According to the above results, the band structures of AFNs and CMNs are schematically shown in Fig. 3f. It can be obviously found that amorphization affords an up-shift of the band structure for AFNs. A similar phenomenon was reported by Savateev et al. in carbon nitride with poor crystallinity.49 In addition, Yao et al. reported that the band structure of maghemite can be elevated by grafting with reduced graphene oxide.50 It was demonstrated that the introduction of oxygen defects into hematite also results in the up-shift.51 As discussed above, nanoscaling is a valid approach for elevating energy band edge positions of many semiconductors (CoO, α-Fe2O3 and CdSe).7,20 It is worth mentioning that the commonality of these methods is lifting of the Fermi level by providing additional electrons. Actually, nanoscaling can produce surface states and dangling bonds, and the resulting unpaired electrons. Amorphization of the maghemite nanodots can be regarded as further nanoscaling, which endows AFNs with more surface states and larger carrier density. The Fermi level is therefore pushed up. The CB edge is also lifted to go across the H2 evolution potential level, and the VB edge is still below the O2 evolution potential level, making AFNs thermodynamically capable of achieving photocatalytic H2 and O2 evolution.52

Dynamic aspect

Above, we have demonstrated the photocatalytic H2 evolution activity of AFNs, which depends on the oxidation potential of the used hole scavenger. This reflects that carrier recombination is still present in AFNs during the photocatalytic process. Since CMNs are thermodynamically unable to achieve photocatalytic H2 evolution, O2 evolution was performed to evaluate and compare the carrier dynamic behaviors of AFNs and CMNs. As discussed above, AFNs perform better than CMNs in O2 evolution upon AM 1.5 irradiation in spite of their higher VB edge and weaker oxidizability. One may consider the increased specific surface area as the reason. However, it is inapplicable when individual vis or UV irradiation was employed. Upon UV irradiation, the photocatalytic rate of AFNs (1.34 μmol h−1) is slightly less than that of CMNs (1.59 μmol h−1). Upon vis irradiation, AFNs exhibit a rate of 0.97 μmol h−1 whereas a trace amount of O2 was detected for CMNs. The action spectra also imply that the wavelength-dependent carrier dynamic behaviors of AFNs and CMNs are different. Therefore, steady-state photoluminescence (PL) and transient time-resolved PL (TRPL) decay spectra were recorded to monitor the carrier dynamic behaviors (Fig. 4).
image file: c9cy01621j-f4.tif
Fig. 4 Carrier dynamic behaviors for AFNs and CMNs. Steady-state PL spectra of AFNs (red) and CMNs (blue) at room temperature with excitation wavelengths of 380 (a), 500 (b) and 700 nm (c). Room-temperature transient TRPL decay spectra for AFNs (red) and CMNs (blue) at 410 (d), 545 (e) and 780 nm (f) with the excitation of the femtosecond laser (395 nm).

As shown in Fig. 4a, under excitation in the UV region (385 nm), two PL peaks at around 410 and 545 nm can be found for AFNs and CMNs. They can be respectively attributed to the LMCT and DEP transitions.7Fig. 4b shows the PL spectra with an excitation wavelength of 500 nm (vis region).53 Both AFNs and CMNs exhibit fluorescence at around 545 and 780 nm, which can be assigned to DEP and LF transitions, respectively.54 LF transitions can be also observed by NIR light excitation at 700 nm (Fig. 4c).55 In contrast to CMNs, AFNs have smaller PL intensity, which may indicate the weaker carrier recombination. Certainly, one may consider that the fluorescence difference should be related to the light absorption difference between AFNs and CMNs. Therefore, TRPL decay curves were measured with excitation of a femtosecond laser (395 nm). The decay transients at 410, 545 and 780 nm are shown in Fig. 4d–f, and fitted using a biexponential decay model56

I(t) = A1[thin space (1/6-em)]exp(−t/τ1) + A2[thin space (1/6-em)]exp(−t/τ2)(4)
where τ1 and τ2 are the short and long lifetime components, respectively. They should be respectively attributed to the surface-related non-radiative recombination and the recombination of free excitons.57A1 and A2 are the corresponding amplitudes. The average emission lifetime (τ) is calculated according to the equation:
τ = (A1τ12 + A2τ22)/(A1τ1 + A2τ2)(5)

The decay parameters are shown in Tables S1–S3 (ESI).

With the emission at 410 nm (Table S1, ESI), the free exciton lifetime τ1 of AFNs is prolonged from 84 to 94 ps, suggesting the enhanced free carrier separation. This would also be the reason for the reduced PL intensity. However, the surface recombination process is strengthened with an increased A2 value due to the amorphization-induced tailing effect. The lifetime component τ2 is shortened from 197 to 142 ps. Therefore, the average PL lifetime of AFNs (129 ps) is smaller than that of CMNs (181 ps), which results in the slightly decreased O2 evolution activity upon UV irradiation. It is discussed above that the ESPT effect resulting from the crystal field splitting of Fe 3d orbitals leads to the O2 evolution sluggishness for CMNs in the vis region. In detail, the atomic arrangement in CMNs is defined and the crystal field splitting is similar for each Fe atom.33 Therefore, separation between t2g and eg states is expected. Experimentally, a clear LMCT1 and LMCT2 absorption boundary and VB XPS cavity can be found. At this point, small polarons are generated since the defined t2g and eg states are highly localized.12,45 Both τ1 and τ2 for CMNs are reduced greatly at 545 nm (Table S2, ESI). The average lifetime (26 ps) is much shorter than that at 410 nm.

In contrast, for AFNs, PL is greatly quenched. The average lifetime at 545 nm is 97 ps, comparable to that at 410 nm. The H2 and O2 evolution action spectra confirm the absence of the ESPT effect in AFNs. Actually, the ordered atomic arrangement is disrupted in AFNs. Coordination defects and undefined DOS distribution are produced. The separation between t2g and eg states therefore becomes blurred. The weakening of crystal field splitting and broadening of the DOS distribution thus delocalize the photogenerated carriers. Table S2 (ESI) reveals that both τ1 and τ2 for AFNs at 545 nm are increased, compared with those for CMNs. The prolonged carrier lifetime and decreased PL intensity endow AFNs with photocatalytic activity upon vis irradiation. In addition, Table S3 (ESI) shows that the lifetimes for CMNs and AFNs at 780 nm are much smaller (12 and 20 ps). No gases can be detected upon NIR irradiation. This is consistent with a previous report claiming the photochemical sluggishness of transient species generated by LF transitions in Fe2O3.14

In this work, particulate photocatalysis for unassisted SHE has been accomplished over AFNs. The H2 evolution rate upon AM 1.5 irradiation using Na2SO3 as the hole scavenger is 8.99 μmol h−1. Indeed, AFNs exhibit mediocre activity among a wide variety of photocatalysts. However, to the best of our knowledge, this is the first time to realize SHE through particulate photocatalysis over Fe2O3 materials. In addition, as shown in Table S4 (ESI), AFNs exhibit favorable activity among amorphous photocatalysts. It should be noticed that this system is noble metal-free, which is in favour of practical application. Although H2 evolution and O2 evolution can be respectively realized over AFNs, they fail to achieve OWS. Actually, OWS is more beneficial to the hydrogen economy with the aim of zero carbon footprint.58 Generally, the H2 evolution rate of half reaction with a sacrificial donor is larger than that of the corresponding OWS. The investigations on half reactions can shed light on the development of OWS. Certainly, further efforts should be made to improve the photocatalytic performance of AFNs and achieve the goal of OWS.

As discussed above, nanoscaling is a strategy to endow thermodynamically photocatalytically inactive semiconductors with unassisted SHE activity through elevating their CB edges to go across the H2 evolution potential level. Unfortunately, it has not been achieved in Fe2O3 materials, even in the ultrathin hematene.7 In spite of the previous progress on band edge shifting by material modification, in this case, amorphization can be regarded as enhanced nanoscaling in view of the short-range order. It pushes the CB edge up greatly, making AFNs thermodynamically capable of achieving unassisted SHE. More material modification methods should be developed to improve the performance of more photocatalysts, which is an important theme in materials science. In addition, crystal field splitting of Fe 3d orbitals leads to the ESPT effect in Fe2O3 materials, which makes absorption in the vis region wasted.13 It can be weakened by amorphization to delocalize the photogenerated carriers. Actually, crystal field splitting is present in many other oxides in the family of correlated electron systems.59,60 Amorphization may be applied to develop these correlated materials to behave as efficient photocatalysts with high utilization of solar irradiation.


In summary, SHE from water splitting has been achieved by AFNs in this case. Thermodynamically, amorphization provides more surface states and larger carrier density, elevating the Fermi level. The CB edge is thus lifted to go across the H2 evolution potential level. Dynamically, amorphization-induced crystal field splitting weakening and DOS distribution broadening delocalize the photogenerated carriers, and overcome the ESPT effect. Non-zero photoconversion efficiency in the vis region is thus obtained. Therefore, amorphization enables AFNs to realize unassisted SHE with high utilization of solar irradiation. This work actually provides a modification method for other photocatalysts.


AFN and CMN preparation

AFNs were prepared by LAL. Typically, 10.0 mg crystalline hematite micropowders and 10 mL deionized water in a 15 mL glass bottle were mixed using a magnetic stirrer with continuous stirring. A laser beam produced by the second harmonic of a Q-switched Nd:YAG laser device was focused into the suspension using a lens above the glass. The focus of the lens is 10 cm. The second harmonic is a pulsed laser with a wavelength of 532 nm and a pulse width of 10 ns. The single pulse energy is 680 mJ. The pulses were repeated with a frequency of 10 Hz. The laser ablation process was carried out for 5 h. Finally, the product was collected by drying at 50 °C. CMNs were prepared by annealing the as-prepared AFN powders in N2 at 200 °C for 3 h.

Materials characterization

XRD patterns were recorded on an XRD setup at a voltage of 40 V and a current of 26 mA (D/MAX-2200, Rigaku). A transmission electron microscope (FTI Tecnai G2 F30) was used to obtain the TEM, HRTEM images and SAED patterns. Raman spectra were measured on a Renishaw InVia Plus laser micro-Raman spectrometer. An FTIR spectrometer (Nicolet 6700, Thermo Scientific) was used to obtain the FTIR patterns. UV-vis-NIR absorption with BaSO4 as the reference was performed on a spectrophotometer (Lambda9500, PerkinElmer). The element chemical environment was detected by XPS measurement on an XPS scanning microprobe spectrometer (Escalab 250, Thermo-VF Scientific). PL spectra were measured on a fluorescence spectrophotometer (FLS920, Edinburgh Instrument, UK). The frequency doubled pulses of a Ti:sapphire oscillator (100 fs, 1 kHz) were used to perform the TRPL measurements. The specific surface area of the sample was determined by BET measurement on an automated gas sorption analyzer (Autosorb-IQ2-MP). Mott–Schottky curves were recorded on an electrochemical workstation with a frequency of 5 kHz in the dark. The measurements were conducted in a three-electrode electrochemical cell with a saturated Ag/AgCl reference electrode and a Pt counter electrode. The electrolyte was 0.5 M Na2SO4 aqueous solution.

Photocatalytic tests

Photocatalytic tests were performed on an evaluation system designed by CEAuLight (CEL-SPH2N, Beijing, China). 20.0 mg photocatalyst powders were dispersed in a top irradiation Pyrex glass vessel containing 100 mL aqueous solution with scavengers. The hole scavengers for the photocatalytic H2 evolution were respectively Na2S/Na2SO3 with a molar ratio of 1.4, TEOA and MeOH. AgNO3 (0.03 M) was used as the electron scavenger for photocatalytic O2 evolution. The vessel was connected to a closed gas-circulation system with 30 min degassing pretreatment. A Xe lamp with an AM 1.5 filter was used as the light source to simulate solar irradiation. A fiber optic spectrometer (AULTT-P, CEAuLight) was employed for light source spectrum calibration. The light power density was determined to be 100 mW cm−2 using a thermopile optical power meter. To extract UV, vis and NIR irradiation individually, the corresponding band-pass filter was further added. For QE analysis, monochromatic filters were used (365, 405, 420, 500, 580, 600, 670 and 780 nm). The reactions lasted for 5 h. An online gas chromatograph setup (Techcomp-GC7920) was connected to the system to detect the evolved gases with Ar carrier gas (TCD detector). After the photocatalytic H2 evolution test upon AM 1.5 irradiation, the powders were collected by centrifugation and washed with deionized water three times. After air drying, the powders were reused four times.

Conflicts of interest

There are no conflicts to declare.


The National Basic Research Program of China (2014CB931700) and the State Key Laboratory of Optoelectronic Materials and Technologies supported this work.


  1. J. Liu, Y. Liu, N. Liu, Y. Han, X. Zhang, H. Huang, Y. Lifshitz, S.-T. Lee, J. Zhong and Z. Kang, Metal-free efficient photocatalyst for stable visible water splitting via a two-electron pathway, Science, 2015, 347, 970 CrossRef CAS PubMed .
  2. S. Chen, T. Takata and K. Domen, Particulate photocatalysts for overall water splitting, Nat. Rev. Mater., 2017, 2, 17050 CrossRef CAS .
  3. J. Engel and H. L. Tuller, The electrical conductivity of thin film donor doped hematite: from insulator to semiconductor by defect modulation, Phys. Chem. Chem. Phys., 2014, 16, 11374 RSC .
  4. K. L. Hardee and A. J. Bard, Semiconductor electrodes V. The application of chemically vapor deposited iron oxide films to photosensitized electrolysis, J. Electrochem. Soc., 1976, 123, 1024 CrossRef CAS .
  5. A. Fujishima and K. Honda, Electrochemical photolysis of water at a semiconductor electrode, Nature, 1972, 23837 Search PubMed .
  6. Z. Jiang, W. Wan, H. Li, S. Yuan, H. Zhao and P. K. Wong, A hierarchical Z-scheme α-Fe2O3/g-C3N4 hybrid for enhanced photocatalytic CO2 reduction, Adv. Mater., 2018, 30, 1706108 CrossRef .
  7. A. P. Balan, S. Radhakrishnan, C. F. Woellner, S. K. Sinha, L. Deng, C. Reyes, B. M. Rao, M. Paulose, R. Neupane, A. Apte, V. Kochat, R. Vajtai, A. Rharutyunyan, C. Chu, G. Costin, D. S. Galvao, A. A. Marti and P. A. Aken, Exfoliation of a non-van der Waals material from iron ore hematite, Nat. Nanotechnol., 2018, 13, 602 CrossRef PubMed .
  8. J. Zhu, Z. Yin, D. Yang, T. Sun, H. Yu, H. E. Hoster, H. H. Hng, H. Zhang and Q. Yan, Hierarchical hollow spheres composed of ultrathin Fe2O3 nanosheets for lithium storage and photocatalytic water oxidation, Energy Environ. Sci., 2013, 6, 987 RSC .
  9. A. Wei, X. Kie, Z. Wen, H. Zheng, H. Lan, H. Shao, X. Sun, J. Zhong and S.-T. Lee, Triboelectric nanogenerator driven self-powered photoelectrochemical water splitting based on hematite photoanodes, ACS Nano, 2018, 12, 8625 CrossRef CAS PubMed .
  10. Z. Zhou, P. Huo, L. Guo and O. V. Prezhdo, Understanding hematite doping with group IV elements: A DFT+U study, J. Phys. Chem. C, 2015, 119, 26303 CrossRef CAS .
  11. J. A. Glasscock, P. R. F. Barnes, I. C. Plumb and N. Savvides, Enhancement of photoelectrochemical hydrogen production from hematite thin films by the introduction of Ti and Si, J. Phys. Chem. C, 2007, 111, 16477 CrossRef CAS .
  12. L. M. Carneiro, S. K. Cushing, C. Liu, Y. Su, P. Yang, A. P. Alivisatos and S. R. Leone, Excitation-wavelength-dependent small polaron trapping of photoexcited carriers in α-Fe2O3, Nat. Mater., 2017, 16, 819 CrossRef CAS PubMed .
  13. D. A. Grave, N. Yatom, D. S. Ellis, M. C. Toroker and A. Rothschild, The rust challenge: on the correlations between electronic structure, excited state dynamics, and photoelectrochemical performance of hematite photoanodes for solar water splitting, Adv. Mater., 2018, 1706577 CrossRef PubMed .
  14. D. Hayes, R. G. Hadt, J. D. Emery, A. A. Cordones, A. B. F. Martinson, M. L. Shelby, K. A. Fransted, P. D. Dahlberg, J. Hong, X. Zhang, Q. Kong, R. W. Schoenlein and L. X. Chen, Electronic and nuclear contribution to time-resolved optical and X-ray absorption spectra of hematite and insights into photoelectrochemical performance, Energy Environ. Sci., 2016, 9, 3754 RSC .
  15. M. Z. Rahman, J. Moffatt and N. Spooner, Topological carbon nitride: localized photon absorption and delocalized charge carrier separation at intertwined photocatalyst interfaces, Mater. Horiz., 2018, 5, 553 RSC .
  16. T. K. Townsend, E. M. Sabio, N. D. Browning and F. E. Osterloh, Photocatalytic water oxidation with suspendedα-Fe2O3particles-effects of nanoscaling, Energy Environ. Sci., 2011, 4, 4270 RSC .
  17. Q. Du and Q. Gao, Template-free synthesis of mesoporous α-Fe2O3nanoflowerswith short charge-carrier diffuse distance for superior photocatalysis, Mater. Technol., 2017, 32, 724 CrossRef CAS .
  18. J. H. Kennedy, Comments regarding electronic conductivity in solid electrolytes, J. Electrochem. Soc., 1977, 124, 865 CrossRef CAS .
  19. H. L. Wang and J. A. Turner, Characterization of hematite thin films for photoelectrochemical water splitting in a dual photoelectrode device, J. Electrochem. Soc., 2010, 157, F173 CrossRef CAS .
  20. L. Liao, Q. Zhang, Z. Su, Z. Zhao, Y. Wang, Y. Li, X. Lu, D. Wei, G. Feng, Q. Yu, X. Cai, J. Zhao, Z. Ren, H. Fang, F. Robles-Hernandez, S. Baldelli and J. Bao, Efficient solar water-splitting using a nanocrystalline CoO photocatalyst, Nat. Nanotechnol., 2014, 9, 69 CrossRef CAS PubMed .
  21. Q. Wang, T. Hisatomi, Q. Jia, H. Tokudome, M. Zhong, C. Wang, Z. Pan, T. Takata, M. Nakabayashi, N. Shibata, Y. Li, I. D. Sharp, A. Kudo, T. Yamada and K. Domen, Scalable water splitting on particulate photocatalyst sheets with a solar-to-hydrogen energy conversion efficiency exceeding 1%, Nat. Mater., 2016, 15, 611 CrossRef CAS PubMed .
  22. M. Mishra and D.-M. Chun, α-Fe2O3as a photocatalytic material: A review, Appl. Catal., A, 2015, 498, 126 CrossRef CAS .
  23. M. Iacob, M. Cazacu, C. Turta, F. Doroftei, M. Botko, E. Čižmár, A. Zeleňáková and A. Feher, Amorphous iron-chromium oxide nanoparticles with long-term stability, Mater. Res. Bull., 2015, 65, 163 CrossRef CAS .
  24. L. Machala, R. Zboril and A. Gedanken, Amorphous iron(III) oxide-A review, J. Phys. Chem. B, 2007, 111, 4003–4018 CrossRef CAS PubMed .
  25. R. D. L. Smith, M. S. Prévot, R. D. Fagan, Z. Zhang, P. A. Sedach, M. K. J. Siu, S. Trudel and C. P. Berlinguette, Photochemical route for accessing amorphous metal oxide materials for water oxidation catalysis, Science, 2013, 340, 60 CrossRef CAS PubMed .
  26. S. Kakuta and T. Abe, Photocatalysis for water oxidation by Fe2O3 nanoparticles embedded in clay compound: correlation between its polymorphs and their photocatalytic activities, J. Mater. Sci., 2009, 44, 2890 CrossRef CAS .
  27. H. Tüysüz and C. K. Chan, Preparation of amorphous and nanocrystalline sodium tantalum oxide photocatlysts with porous matrix structure for overall water splitting, Nano Energy, 2013, 2, 116 CrossRef .
  28. Y. Kang, Y. Yang, L. Yin, X. Kang and G. Liu, An amorphous carbon nitride photocatalyst with greatly extended visible-light-responsive range for photocatalytic hydrogen generation, Adv. Mater., 2015, 27, 4572 CrossRef CAS PubMed .
  29. M. Z. Rahaman, P. C. Tapping, T. W. Kee, R. Smernik, N. Spooner, J. Moffatt, Y. Tang, K. Davey and S.-Z. Qiao, A benchmark quantum yield for water photoreduction on amorphous carbon nitride, Adv. Funct. Mater., 2017, 27, 1702384 CrossRef .
  30. Z. Lin, C. Du, B. Yan, C. Wang and G. Yang, Two-dimensional amorphous NiO as a plasmonic photocatalyst for solar H2 evolution, Nat. Commun., 2018, 9, 4036 CrossRef PubMed .
  31. L. Li, J. Xiao, P. Liu and G. Yang, Super adsorption capability from amorphousization of metal oxide nanoparticles for dye removal, Sci. Rep., 2015, 5, 9028 CrossRef CAS PubMed .
  32. Y. E. Mendili, J. Bardeau, N. Randrianantoandro, F. Grasset and J. Greneche, Insights into the mechanism related to the phase transition from γ-Fe2O3to α-Fe2O3nanoparticles induced by thermal treatment and laser irradiation, J. Phys. Chem. C, 2012, 116, 23785 CrossRef .
  33. C. Lohaus, C. Steinert, J. Brötz, A. Klein and W. Jaegermann, Systematic investigation of the electronic structure of hematite thin films, Adv. Mater. Interfaces, 2017, 4, 1700542 CrossRef .
  34. A. M. Jubb and H. C. Allen, Vibrational spectroscopic characterization of hematite, maghemite, and magnetite thin films produced by vapor deposition, ACS Appl. Mater. Interfaces, 2010, 2, 2804 CrossRef CAS .
  35. S. Chandrasekaran, S. H. Hur, E. J. Kim, B. Rajagopalan, K. F. Babu, V. Senthilkumar, J. S. Chung, W. M. Choi and Y. S. Kim, Highly-ordered maghemite/reduced graphene oxide nanocomposites for high-performance photoelectrochemical water splitting, RSC Adv., 2015, 5, 29159 RSC .
  36. P. A. Fedders, Theoretical study on the nature of band-tail states in amorphous Si, Phys. Rev. B: Condens. Matter Mater. Phys., 1998, 58, 15624 CrossRef CAS .
  37. B. Prasai, B. Cai, M. K. K. Underwood, J. P. Lewis and D. A. Drabold, Properties of amorphous and crystalline titanium dioxide from first principles, J. Mater. Sci., 2012, 47, 7515 CrossRef CAS .
  38. C. Pascal, J. L. Pascal, F. Favier, M. L. E. Moubtassim and C. Payen, Electrochemical synthesis of the control of γ-Fe2O3nanoparticle size. Morphology, microstructure, and magnetic behavior, Chem. Mater., 1999, 11, 141 CrossRef CAS .
  39. T. Song, P. Zhang, J. Zeng, T. Wang, A. Ali and H. Zeng, Boosting the photocatalytic H2 evolution activity of Fe2O3polymorphys (α-, γ- and β-Fe2O3) by fullerene [C60]-modification under visible light irradiation, RSC Adv., 2017, 7, 29184 RSC .
  40. M. J. Berr, P. Wagner, S. Fischbach, A. Vaneski, J. Schneider, A. S. Susha, A. L. Rogach, F. Jäckel and J. Feldmann, Hole scavenger redox potentials determine quantum efficiency and stability of Pt-decorated CdS nanorods for photocatalytic hydrogen generation, Appl. Phys. Lett., 2012, 100, 223903 CrossRef .
  41. I. V. Chernyshova, S. Ponnurangam and P. Somasundaran, On the origin of an unusual dependence of (bio)chemical reactivity of ferric hydroxides on nanoparticle size, Phys. Chem. Chem. Phys., 2010, 12, 14045 RSC .
  42. A. P. Grosvenor, B. A. Kobe, M. C. Biesinger and N. S. Mclntyre, Investigation of multiplet of Fe 2p XPS spectra and bonding in iron compounds, Surf. Interface Anal., 2004, 36, 1564 CrossRef CAS .
  43. R. P. Gupta and S. K. Sen, Calculation of multiplet structure of core p-vacancy levels, Phys. Rev. B: Solid State, 1974, 10, 71 CrossRef CAS .
  44. T. Droubay and S. A. Chambers, Surface-sensitive Fe 2p photoemission spectra of α-Fe2O3(0001): The influence of symmetry and crystal-field strength, Phys. Rev. B: Condens. Matter Mater. Phys., 2001, 64, 205414 CrossRef .
  45. Y. P. He, Y. M. Miao, C. R. Li, S. Q. Wang, L. Cao, S. S. Xie, G. Z. Yang and B. S. Zou, Size and structure effect on optical transitions of iron oxide nanocrystals, Phys. Rev. B: Condens. Matter Mater. Phys., 2005, 71, 125411 CrossRef .
  46. D. M. Sherman, The electronic structure of Fe3+ coordination sites in iron oxides; Applications to spectra, bonding, and magnetism, Phys. Chem. Miner., 1985, 12, 161 CrossRef CAS .
  47. Z. Lin, J. Xiao, L. Li, P. Liu, C. Wang and G. Yang, Nanodiamond-embedded p-type copper(I) oxide nanocrystals for broad-spectrum photocatalytic hydrogen evolution, Adv. Energy Mater., 2016, 6, 1501865 CrossRef .
  48. A. Mettenbörger, T. Singh, A. P. Singh, T. T. Järvi, M. Moseler, M. Valldor and S. Mathur, Plasma-chemical reduction of iron oxide photoanodes for efficient solar hydrogen production, Int. J. Hydrogen Energy, 2014, 39, 4828 CrossRef .
  49. A. Savateev, S. Pronkin, J. D. Epping, M. G. Willinger, C. Wolff, D. Neher, M. Antonietti and D. Dontsova, Potassium ploy(heptazine imides) from aminotetrazoles: shifting band gaps of carbon nitride-like materials by 0.7 V for more efficient solar hydrogen and oxygen evolution, ChemCatChem, 2017, 9, 167 CrossRef CAS .
  50. J. Yao, J. Chen, K. Shen and Y. Li, Phase-controllable synthesis of MOF-templated maghemite-carbonaceous composites for efficient photocatalytic hydrogen production, J. Mater. Chem. A, 2018, 6, 3571 RSC .
  51. A. Boudjemaa, S. Boumaza, M. Trari, R. Bouarab and A. Bouguelia, Physical and photo-electrochemical characterizations of α-Fe2O3. Application for hydrogen production, Int. J. Hydrogen Energy, 2009, 34, 4268 CrossRef CAS .
  52. T. Wang, X. Meng, G. Liu, K. Chang, P. Li, Q. Kang, L. Liu, M. Li, S. Ouyang and J. Ye, In situ synthesis of ordered mesoporous Co-doped TiO2 and its enhanced photocatalytic activity and selectivity for the reduction of CO2, J. Mater. Chem. A, 2015, 3, 9491 RSC .
  53. T. Song, J. Huo, T. Liao, J. Zeng, J. Qin and H. Zeng, Fullerene [C60] modified Cr2-xFexO3 nanocomposites for enhanced photocatalytic activity under visible light irradiation, Chem. Eng. J., 2016, 287, 359–366 CrossRef CAS .
  54. M. Aslam, T. Qamar, A. U. Rehman, M. T. Soomro, S. Ali, I. M. I. Ismail and A. Hameed, The evaluation of the photocatalytic activity of magnetic and non-magnetic polymorphs of Fe2O3 in natural sunlight exposure: A comparison of photocatalytic activity, Appl. Surf. Sci., 2018, 451, 128–140 CrossRef CAS .
  55. L. Stagi, J. A. D. Toro, A. Ardu, C. Cannas, A. Casu, S. S. Lee and P. C. Ricci, Surface effects under visible irradiation and heat treatment on the phase stability of γ-Fe2O3 nanoparticles and γ-Fe2O3-SiO2 core-shell nanostructures, J. Phys. Chem. C, 2014, 118, 2857 CrossRef CAS .
  56. H. Zhang, J. Huang, Y. Fang, M. Zhang, K. Liu and B. A. Dong, A nonmetal plasmonic Z-scheme photocatalyst with UV- to NIR-driven photocatalytic protons reduction, Adv. Mater., 2017, 29, 1606688 CrossRef PubMed .
  57. X. Liu, Q. Zhang, G. Xing, Q. Xiong and T. C. Sum, Size-dependent exciton recombination dynamics in single CdS nanowires beyond the quantum confinement regime, J. Phys. Chem. C, 2013, 117, 10716 CrossRef CAS .
  58. X. Chen, S. Shen, L. Guo and S. S. Mao, Semiconductor-based photocatalytic hydrogen generation, Chem. Rev., 2010, 110, 6503 CrossRef CAS PubMed .
  59. J. Shi, Y. Zhou and S. Ramanathan, Colossal resistance switching and band gap modulation in a perovskite nickelate by electron doping, Nat. Commun., 2014, 5, 4860 CrossRef CAS PubMed .
  60. Z. Zhang, D. Schwanz, B. Narayanan, M. Kotiuga, J. A. Dura, M. Cherukara, H. Zhou, J. W. Freeland, J. Li, R. Sutarto, F. He, C. Wu, J. Zhu, Y. Sun, K. Ramadoss, S. S. Nonnenmann, N. Yu, R. Comin, K. M. Rabe, S. K. R. S. Sankaranarayanan and S. Ramanathan, Perovskite nickelates as electric-field sensors in salt water, Nature, 2018, 553, 68 CrossRef CAS .


Electronic supplementary information (ESI) available. See DOI: 10.1039/c9cy01621j

This journal is © The Royal Society of Chemistry 2019