Toward Shape-Controlled Metal Oxynitride and Nitride Particles for Solar Energy Applications Dileka Abeysinghe and Sara E. Skrabalak* Department of Chemistry, Indiana UniversityBloomington, 800 East Kirkwood Avenue, Bloomington, Indiana 47405, United States ABSTRACT: Photocatalytic water splitting is a promising platform for clean and renewable energy. Photocatalysts that absorb visible light are critically needed to efficiently harness solar energy, and metal oxynitrides and nitrides have the potential to address this need owing to their narrower band gaps compared to oxides. The catalytic activity of these materials could be enhanced through the synthesis of structurally defined particles with reduced recombination sites and the preferential expression of catalytically active sites. This Perspective highlights recent progress and current challenges associated with the design and synthesis of metal (d0/d10) oxynitride and nitride particles with defined structural features for photocatalytic applications. Nitriding oxide templates with defined morphology and topotactic transformations of structurally defined templates are promising routes to metal oxynitrides and nitrides with shape control. Case studies show that such synthetic advances are often coupled with enhanced photocatalytic activity; however, greater synthetic control is still required to understand the structure−function relationships of these promising materials.
D
on photocatalytic water splitting has been carried out and many metal oxide photocatalysts capable of splitting water under UV light have been reported.6 Here, we highlight recent advances and current challenges in the design and synthesis of shapecontrolled oxynitride and nitride particles, which are capable of absorbing visible light. Figure 1 illustrates the main processes in a photocatalytic reaction. When the energy of incident light is greater than that of the band gap of the photocatalytic material, electrons are excited from the valence band to the conduction band, generating electron−hole pairs that can facilitate redox reactions at the surface (e.g., eqs 2 and 3). Thus, the standard potential for water electrolysis (E0cell = E0cathode − E0anode) is −1.23 V at 25 °C at pH 0, and the theoretical minimum band gap for a suitable photocatalyst is 1.23 eV (λ = 1100 nm). Of the materials studied as heterogeneous photocatalysts, many are metal oxides that respond only to UV light, which is a small fraction (ca. 4%) of incoming solar energy.7 Therefore, current interests lie in developing photocatalysts that can function under visible light in order to use sunlight efficiently and effectively. Most metal oxide photocatalysts can be categorized into two classes: d0 type metal (Ti4+, Ta5+, Nb5+, Zr4+ etc.) and d10 type metal (Ga3+, In3+, Ge4+, Sn5+, Cd2+, Zn2+ etc.) semiconductors.
evelopment of clean, renewable energy is essential to mitigate the effects of current and future global energy demands. The advancement of platforms that use sunlight is of urgent need, as this naturally occurring resource delivers enough energy to power the world’s yearly consumption within an hour.1 However, a large gap between the present usage of solar energy and its potential to deliver energy exists. Photocatalytic splitting of water into hydrogen and oxygen is an attractive method to store solar energy in the form of clean, renewable hydrogen fuel.2−4 This process is accompanied by a large positive Gibbs free energy change, as shown in eq 1. H 2O → H 2 +
1 O2 2
ΔG 0 = 237 kJ/mol
(1)
Anode (oxidation): H 2O + 2h+ → 2H+ +
1 O2 2
E 0 = + 1.23 V vs NHE
(2)
Cathode (reduction): 2H+ + 2e− → H 2
E 0 = 0.00 V vs NHE
(3)
The uphill reaction has been coined “artificial photosynthesis”, and in 1970, Honda and Fujishima achieved water splitting using a TiO2 photoanode connected to a Pt counter electrode in a cell composed of an electrolyte and an external bias.5 Ultraviolet (UV) light was required for photoexcitation on account of the large band gap of TiO2. Since this discovery, extensive research © XXXX American Chemical Society
Received: March 29, 2018 Accepted: May 4, 2018
1331
DOI: 10.1021/acsenergylett.8b00518 ACS Energy Lett. 2018, 3, 1331−1344
Perspective
Cite This: ACS Energy Lett. 2018, 3, 1331−1344
ACS Energy Letters
Perspective
redox reaction.6,19,21 The latter helps avoid recombination of electron−hole pairs (Figure 2). For these reasons, the design of
Figure 2. Scheme depicting the effects of defects and particle size on photocatalytic activity.
Figure 1. Scheme highlighting the key processes in a photocatalytic reaction.
The conduction bands of these d0 and d10 compounds typically consist of d and sp orbitals, respectively, whereas their valence bands consist of O 2p orbitals. However, such metal oxides often suffer from a wide band gap as their valence bands are positioned at ∼3 eV (vs NHE at pH 0).6,8 Strategies to circumvent this limitation are to either dope oxide sites with less electronegative anions9,10 or substitute some or all the oxygen ions in metal oxides to create compounds such as metal oxynitrides or nitrides, among others.11−13 These compounds have narrower band gaps compared to their oxide counterparts and often perform better than doped oxides, which can contain detrimental electron−hole traps inherent to the electronic structure of the material.14,15 Oxynitrides contain hybridized valence bands of O 2p and N 2p orbitals with reduced band gaps compared to corresponding metal oxides, and nitrides have even narrower band gaps as their valence bands consist of only N 2p orbitals.16 Nitrogen introduction is facilitated by similar ionic radii, coordination environments, electronegativity, and polarizability when comparing an oxide to an oxynitride or nitride.17 There are only a few oxynitride and nitride examples known to date that can function as photocatalysts under visible light, and our focus is on those prepared as structurally defined particles.18−28 Conventionally, the synthesis of metal oxynitride and nitride particles is accomplished by heating metals, metal oxides, or chlorides under NH3 at high temperatures (900−1100 °C) for many hours (>10 h).11,29 Here, NH3 acts as the nitrogen source. Organic nitrogen sources such as urea and amines also have been used with sol−gel, hydrothermal, and other synthetic strategies to produce oxynitride and nitride particles, but their decomposition often introduces carbon to materials.30,31 Regardless of the nitrogen source, morphology control and nanostructuring of metal oxynitrides and nitrides have been difficult to achieve. The harsh nitridation conditions usually introduce defects within the crystals and yield structurally ill-defined oxynitrides and nitrides on account of crystallization and sintering processes. Thus, these particles can contain trap sites and recombination centers that reduce the efficiency of photocatalytic processes. The photocatalytic efficiency of these particles could be enhanced through the synthesis of crystals with defined morphology, which typically have reduced defect populations and can facilitate charge separation by driving electrons and holes to different facets.32−34 Moreover, shape-controlled crystals selectively express specific facets, which may provide more active sites for a given catalytic process or better stabilize a cocatalyst.19,21,34 Also, smaller particle sizes can enhance photocatalytic activity by increasing the expression of active sites and decreasing the distances that photogenerated electrons and holes have to travel to facilitate the
synthetic strategies to obtain metal oxynitrides and nitride particles with controlled structural features is attractive. By analyzing specific case studies, this Perspective highlights recent progress toward metal oxynitride and nitride particles with defined structural features to realize high photocatalytic activity. We conclude with an analysis of this work and future directions toward which the synthesis of oxynitride and nitride particles can be directed to achieve high-performance materials for a variety of energy applications. Tantalum Nitride and Tantalum Oxynitride. Tantalum nitride (Ta3N5) is a widely studied and promising material for solar water splitting. Although there are several d0 metal nitrides and oxynitrides available for solar applications, Ta3N5 is the only reported nitride of this kind. With an anosovite (Ti3O5) structure type consisting of corner- and edge-shared TaN6 octahedra, this simple nitride has a narrow band gap of 2.1 eV and can convert water to H2 or O2 under visible light (λ < 590 nm) illumination in the presence of a sacrificial electron donor or acceptor, respectively.18 Compared to bulk samples, nanostructured Ta3N5 has shown higher photocatalytic activity.19,35−40 For instance, Domen et al. synthesized mesoporous Ta3N5 by nitridation of SiO2-coated mesoporous Ta2O5 with NH3.35 When loaded with Pt, the mesoporous Ta3N5 was three times more active for photocatalytic H2 evolution from an aqueous methanol solution under visible light (λ > 420 nm) than that of bulk Ta3N5. Recently, Li and co-workers reported a record photocurrent of 12.1 mA cm−2 at 1.23 V vs RHE from a photoanode (AM 1.5 G simulated sunlight at 100 mW cm−2 in 1 M NaOH aqueous solution, pH 13.6) consisting of the Ta3N5 semiconductor, a Ni(OH)x/ferrihydrite hole-storage layer, and a TiOx blocking layer; this value is nearly at its photocurrent limit of 12.9 mA cm−2.40 These examples highlight the potential utility of Ta3N5 as a visible light photocatalyst, and its performance could be enhanced by incorporating strategies toward shape-controlled platforms. Recently, Domen et al. synthesized Ta3N5 with defined morphologies and modified surfaces.41 Specifically, TaCl5 or Ta2O5 was dispersed in a molten flux of NaCl, Zn, or Na2CO3 during ammonolysis; the nitridation temperature was in the range of 1023−1273 K depending on the specific synthesis and produced well-crystallized Ta3N5 particles. For example, a powder consisting of nearly monodispersed particles, a few tens of nanometers in size, were produced by nitriding TaCl5−NaCl at 1073 K for 10 h. Increasing the temperature and nitridation time to 1173 K and 20 h, respectively, formed larger particles, with a rectangular parallelepiped shape. An SEM image of the rectangular parallelepiped particles is shown in Figure 3A. The authors speculated that this difference in morphology with increasing 1332
DOI: 10.1021/acsenergylett.8b00518 ACS Energy Lett. 2018, 3, 1331−1344
ACS Energy Letters
Perspective
Figure 3. SEM images of (A) Ta3N5 prepared from TaCl5/NaCl mixture nitridated at 1173 K for 20 h and (B) Ta3N5 prepared from Ta2O5 in Na2CO3 flux. (C) Proposed scheme for evolution of conventional Ta3N5 (bottom) and Na2CO3/ Ta3N5 (top). (D) Rate of O2 gas evolution over Ta3N5 synthesized with different fluxes without a cocatalyst. TEM images of (E) Ta3N5 octahedra and (F) Ta3N5 nanoplates. (G) OER and (H) HER time-dependent study of AMSS-derived nanoplates and octahedra compared to the reference Ta3N5. Reprinted with permission from refs 41, 36, and 19. Copyright 2011 and 2012 American Chemical Society and 2016 The Royal Society of Chemistry.
temperature and time may arise from an enhanced growth rate relative to elution rate. Similar results were obtained from ammonolysis of Ta2O5−Na2CO3 at either 1123 K for 80 h or 1173 K for 40 h, with rectangular parallelepiped shaped crystals produced. No differences between the results obtained at nitridation temperatures of 1123 and 1173 K were evident, indicating that annealing and crystal growth were promoted by Na2CO3. In contrast, nitridation of the TaCl5−Zn mixture at 1073 K for 10 h yielded monodispersed fine particles; the powder X-ray diffraction (PXRD) pattern showed broader peaks compared to the sample prepared with the NaCl flux, indicating smaller crystallite sizes. Interestingly, direct ammonolysis of Ta2O5 or TaCl5 resulted in pseudomorphous and sintered Ta3N5 particles, respectively. Thus, this example shows that the synthesis of morphologically defined particles of Ta3N5 is aided with a molten flux. Typically, a flux is used to obtain large, highquality crystals of materials by dissolving precursors at a lower temperature compared to the melting point of the precursor.42,43 Then, through judicious control of the heating conditions and supporting atmosphere, supersaturation can be achieved through precursor reaction, decomposition, or recrystallization to facilitate nucleation of product. In the example provided by Domen et al., the different fluxes provided a variety of conditions for crystallization, while also inhibiting the agglomeration
of the Ta3N5 crystals. Unfortunately, these samples were not evaluated as photoanodes or photocatalysts, but this study demonstrated that morphology control and surface modification of nitrides can be obtained by using fluxes as a novel media for ammonolysis. In a subsequent study, Domen et al. used a related technique to synthesize Ta3N5 and study the photocatalytic activities of the products.36 Ta2O5 was mixed and annealed in air with different alkaline metal (AM) salts (NaCl, Na2CO3, KCl, K2CO3, LiCl, and Li2CO3; AM/Ta = 0.1 by mole) at 773 K for 2 h. The Ta2O5 samples were then isolated from the salts and nitrided with a NH3 flow at 1123 K for 20 h. An SEM image of the Ta3N5 particles obtained with the Na2CO3 flux are shown in Figure 3B, revealing a powder composed of particles with an average size of ∼80 nm. The other AM/Ta2O5 combinations produced similar products, except for LiCl and Li2CO3, which produced powders with larger crystallites. This difference is likely the result of the Li salts having lower melting points (886 K, LiCl and 1005 K, Li2CO3) compared to the other AMs studied as well as the relatively close ionic radii of Li+ (90 pm) and Ta5+ (78 pm); these features can enable interdiffusion of the cations and growth of larger Ta3N5 particles. Regardless of the system, the morphology of the Ta3N5 produced with a flux contrasts with Ta3N5 prepared by direct ammonolysis of Ta2O5, in which porous structures were obtained. PXRD studies provide mechanistic insight into why such different products are obtained when a flux is employed versus not. Specifically, samples were analyzed at different stages of nitridation for the Na2CO3/Ta2O5 system and revealed that the formation of Ta3N5 goes through tantalum oxynitride (TaON) and NaTaO3 as intermediates. Both intermediates, TaON and
Morphology control and surface modification of nitrides can be obtained by using fluxes as a novel media for ammonolysis. 1333
DOI: 10.1021/acsenergylett.8b00518 ACS Energy Lett. 2018, 3, 1331−1344
ACS Energy Letters
Perspective
from Ba(NO3)2 and SrCl2 fluxes, respectively, and nominally ascribed to Ta2O5. Broad PXRD peaks along with diffuse rings in electron diffraction confirmed the poorly crystalline nature of both the nanoplates and octahedra. In the absence of the flux, polycrystalline microspheres were produced, indicating that the flux was essential to obtain different morphologies. Polycrystalline microspheres are consistent with products obtained by traditional ultrasonic spray synthesis, where multiple particles nucleate and agglomerate within each droplet as the solvent evaporates. The various oxide particles were nitrided in an NH3 flow. Remarkably, the octahedral and nanoplate shaped particles were preserved during nitridation as revealed by transmission electron microscopy (TEM) (Figures 3E,F), but both particles were porous and PXRD confirmed Ta3N5. This porosity arises because the replacement of every 3 oxygen atoms with 2 nitrogen atoms ultimately leads to volume shrinkage during nitridation. The oxygen evolution rate (OER: 300 W xenon lamp with a 400 nm long-pass filter, 20 mg of catalyst, 20 mg of La2O3 as pH buffer, 12 mL of 0.05 M AgNO3) and hydrogen evolution rate (HER: 300 W xenon lamp with a 400 nm long-pass filter, 20 mg of catalyst, 12 mL aqueous solution containing 10 vol % methanol) were measured independently for the nanoplates, octahedra, and reference Ta3N5. Without a cocatalyst, the O2 and H2 evolution rates were low. However, the addition of 2 wt % CoOx cocatalyst and 3 wt % Pt onto the Ta3N5 samples to examine OER and HER, respectively, led to improved photocatalytic performance compared to the reference (Figure 3G and H). The OER for the Ta3N5 nanoplates, octahedra, and reference were 7.98 (3.34% for AQE), 3.32 (1.39% for AQE), and 1.23 μmol h−1 (0.53% for AQE), respectively. Similarly, Ta3N5 nanoplates showed higher HER activity 0.53 μmol h−1 (0.22% for AQE) compared to the octahedra 0.34 μmol h−1 (0.14% for AQE) and reference 0.03 μmol h−1 (0.01% for AQE) samples. The band gaps of the Ta3N5 samples were similar (2.14, 2.15, and 2.11 eV for nanoplates, octahedra, and reference, respectively); however, the diffuse reflectance data for the reference material showed absorption at longer wavelengths, around 700 nm. This feature is indicative of reduced Ta5+ defects on the surface, and such defects are detrimental to photocatalytic performance.45−47 We attributed this difference to the longer nitridation time required for the reference (15 h) compared to the nanoscale templates (9 h). The BET surface areas were 23.70, 9.37, and 8.08 m2 g−1 for the nanoplates, octahedra, and reference Ta3N5, respectively. Nanoplates had the highest specific surface area, which could be the reason for their enhanced activity compared to the octahedra. Significantly, this work highlights that advances in oxide synthesis can enable the synthesis of metal nitrides with defined morphology through judicious control of nitridation conditions. Tantalum oxynitride is also a visible light-absorbing d0 metal catalyst with a high quantum yield for O2 evolution in the presence of a sacrificial agent. Its band gap is slightly greater than that of Ta3N5, 2.4−2.8 eV depending on the polymorph versus ∼2.1 eV.48−50 Unfortunately, there are few studies reporting morphology control of TaON, although presumably the methods developed for Ta3N5 can be applied to the oxynitride through control of the ammonolysis conditions. Just as with Ta3N5, hightemperature nitridation of Ta2O5 powders under a flow of NH3 is the most common way to synthesize TaON. For instance, hollow urchinlike β-TaON and γ-TaON nanostructures were synthesized by Zhu and co-workers through nitridation of Ta2O5.20 β-TaON is a yellow-green oxynitride with the same structure as baddeleyite monoclinic ZrO251 and γ-TaON is a light brown
NaTaO3, were formed within 1 h of nitridation. However, only the target phase, Ta3N5, is ultimately produced because of the volatilization of Na species. In contrast, Ta3N5 was produced directly from Ta2O5 without the Na2CO3 flux. These different pathways and their effect on morphology are illustrated in Figure 3C. Notably, the Na2CO3 flux facilitates dissolution of reagents at elevated temperatures to nucleate TaON and NaTaO3 as intermediates in the early stages of the synthesis, which is followed by the formation of highly crystalline Ta3N5. The flux inhibits crystallite aggregation to achieve smaller particle sizes than direct nitridation, which depends on the morphology and crystallinity of the initial oxide. The rates of photocatalytic O2 evolution of the flux-assisted and reference Ta3N5 samples were measured (0.1 g of catalyst, 0.1 g of La2O3 as pH buffer, 100 mL of 50 mM AgNO3, 300 W xenon lamp with 420 < λ < 800 nm; Figure 3D). The flux-assisted samples outperformed the reference Ta3N5 in all cases, with the Na2CO3-modified Ta3N5 showing the highest activity of all. This enhancement may be due to the smaller particle sizes that allow for rapid migration of the charge carriers within the particles to the surface active sites as well as the improved crystallinity that reduced the concentration of recombination centers. Ir-based and CoOx cocatalysts were also added to the Ta3N5 samples, and the samples were then evaluated as photocatalysts for O2 evolution. The highest activity was obtained with 2 wt % of CoOx loading, followed by NH3 treatment at 773 K for 1 h of the product from the Na2CO3/Ta3N5 system. The apparent quantum efficiencies (AQEs), which is the ratio of photoproduct molecules formed per photon of incident light, of conventional Ta3N5 and CoOx loaded Na2CO3/Ta3N5 were calculated. AQE is estimated to be less than the real quantum yield because the number of absorbed photons is usually less than that of incident light; however, caution in comparing AQEs must be taken as the amount of scattered and reflected light may change as a function of particle size and shape. The AQE of conventional Ta3N5 increased from 1.3% to 5.2% at 500−600 nm with CoOx-loaded Na2CO3/Ta3N5, which is one of the highest values reported at longer wavelengths. Thus, the synthesis of Ta3N5 samples with modified surfaces also enhances the performance of cocatalysts, which are an integral component of most photocatalytic systems. H2 evolution also was investigated for the Na2CO3/Ta3N5 and Ta3N5 samples, both loaded with Pt cocatalysts; however, no significant improvement was observed with the addition of Na2CO3 during nitridation. The authors noted that further investigations on the cocatalyst loading and modification methods of Ta3N5 were necessary to fully understand the effect of the AM salt modification on H2 evolution. In addition to modifying the nitridation process, flux methods can be used to achieve novel oxide templates for the synthesis of oxynitride and nitride materials. In fact, our group has spatially and temporally confined molten salt syntheses within aerosol droplets to achieve oxide crystals with far from equilibrium forms.44 This aerosol-assisted molten salt synthesis (AMSS) technique limits crystal growth within micrometer-sized aerosol droplets, with the flux itself inhibiting the agglomeration of individual crystallites. Adopting this technique, Ta3N5 nanocrystals with defined morphology were produced that displayed enhanced photocatalytic activity compared to structurally ill-defined particles.19 First, shape-defined oxide particles were produced by AMSS. An aqueous solution of (NH4)4[Ta2(C3H4O3)4(O2)2O] and either Ba(NO3)2 or SrCl2 was nebulized, and the aerosol droplets were passed through a hot wall reactor at 900 and 950 °C, respectively. Nanoplates and octahedra were achieved 1334
DOI: 10.1021/acsenergylett.8b00518 ACS Energy Lett. 2018, 3, 1331−1344
ACS Energy Letters
Perspective
Figure 4. (A) Schematic illustration of the formation of hollow urchinlike Ta2O5 hierarchical nanostructures and the formation of γ-TaON, β-TaON, and Ta3N5. SEM images of urchinlike (B) Ta2O5, (C) γ/β-TaON, and (D) Ta3N5. (E) H2 production rates of urchin(u)-Ta2O5, a; commercial(c)-Ta2O5, b; γ-TaON, c; γ/β-TaON, d; β-TaON, e; c-TaON, f; u-Ta3N5, g; and c-Ta3N5, h. Reprinted with permission from ref 20. Copyright 2013 The Royal Society of Chemistry.
oxynitride that crystallizes in the monoclinic VO2(B) structure.52 First, uniform, urchinlike nanostructures of Ta2O5, which consist of nanoneedles with lengths of 100−200 nm and an average diameter of 5 nm, were synthesized from Ta powders and HF. H2TaF7 was produced then hydrolyzed to form the Ta2O5 phase. Nitridation proceeded through the following phase transformations: Ta2O5 → γ-TaON → β-TaON → Ta3N5 with the hollow urchinlike morphology of the oxide template preserved throughout (Figure 4A). Specifically, Ta2O5 powder was heated at 850 °C for 5, 10, and 15 h under an NH3 flow of 20 mL min−1 to obtain γ-TaON, γ/β-TaON, and β-TaON, respectively, whereas Ta2O5 powder was heated at 850 °C for 10 h under NH3 flow of 40 mL min−1 to obtain Ta3N5. The SEM images of urchinlike Ta2O5, γ/β-TaON, and Ta3N5 nanostructures are shown in panels B, C, and D of Figure 4, respectively. After nitridation, the urchinlike morphology of the oxide was preserved for the γ-TaON, β-TaON, and Ta3N5 nanostructures, but the nanoneedles were thicker (20−30 nm) because of the high-temperature treatment.
The photocatalytic H2 evolution rates of the hollow structures and commercial Ta2O5 and Ta3N5 were measured (300 W xenon lamp with a cut off filter λ > 420 nm, 0.3 g of catalyst, 200 mL of 10:1 water:methanol). The hollow structures exhibited excellent photocatalytic activities compared to commercial Ta2O5, TaON, and Ta3N5 (Figure 4E). Of the hollow structures, 0.1 wt % Ru loaded γ-TaON performed the best (381.6 μmol h−1 and AQE of 9.5%), followed by 0.1 wt % Ru loaded γ/β-TaON (351.9 μmol h−1 and AQE of 8.7%), 0.1 wt % Ru loaded β-TaON (278.7 μmol h−1 and AQE of 6.9%), and 0.1 wt % Pt loaded Ta3N5 (127.5 μmol h−1 and AQE of 3.1%). The photocatalytic activity of γ-TaON was about 47.5 times higher than that of the conventional TaON. Furthermore, the photocurrent density of hollow urchinlike TaON films was compared with the commercial TaON (300 W xenon lamp with a cut off filter λ > 420 nm, 0.5 M Na2SO4 electrolyte, Co3O4 cocatalyst). Specifically, TaON films were prepared by dispersing each sample in ethanol and polyethylene glycol, which was then dip-coated onto a fluorine-doped tin oxide (FTO) glass electrode followed by the calcination at 400 °C for 1 h 1335
DOI: 10.1021/acsenergylett.8b00518 ACS Energy Lett. 2018, 3, 1331−1344
ACS Energy Letters
Perspective
Figure 5. (A) Crystal structure model for the topotactic transition from Sr2Ta2O7 (along [010] direction) to SrTaO2N (along [100] direction). SEM images of SrTaO2N (B) columnar crystals, (C) nanoplates, (D) polyhedra, and (E) polycrystals. Amount of (F) H2 and (G) O2 evolved in the first hour of irradiation from SrTaO2N nanoplates, polyhedra, and polycrystals. Reprinted with permission from ref 58 and 21. Copyright 2013 The Royal Society of Chemistry and 2017 Wiley-VCH Verlag GmbH & Co. KGaA.
under N2 flow. All the hollow TaON samples provided a higher photocurrent compared to commercial TaON, with the γ-TaON sample showing the highest photocurrent density (∼1.4 mA cm−2), which was about 2.5 times greater than that of the commercial TaON at 0.8 V vs SCE. The high specific surface areas and short migration distances for charge carriers achieved with the hollow nanostructures accounted for the enhanced photocatalytic and photoelctrochemical performances compared to bulk samples. Perovskite-Related Oxynitrides. Apart from Ta3N5 and TaON, the rest of the d0 metal oxynitrides adopt a perovskite-based crystal structure.12 Perovskites belong to the ABO3 structure type and often have many useful properties, including superconductivity,53 colossal magnetoresistance,54 high dielectric property,55 and nonlinear optical properties.56 Introducing N into the perovskite structure yields the ABO2N crystal structure that consists of irregular, corner-shared BO(N)6 octahedra that are connected via A cations. This incorporation can lead to narrower band gap materials compared to the parent oxide. This narrowing is on account of the higher N 2p orbital energy compared to the O 2p orbital of the parent oxide and also makes them promising candidates for visible light-absorbing photocatalysts. Studied compounds include LaBO2N, (B = Ti and Ta),28,57 ATaO2N (A = Ca, Sr, and Ba),23 and ANbO2N (A = Ca, Sr, Ba),24 with
control of structural features through innovations in synthesis reported in some cases and leading to better-performing materials as highlighted herein. Strontium Tantalum Oxynitride (SrTaO2N). For example, there are recent reports of SrTaO2N synthesized as nanoplates, polyhedra, polycrystals, nanowires, and more.21,58,59 Typically, SrTaO2N is synthesized by ammonolysis of Sr2Ta2O7 and has a tetragonally distorted perovskite structure with a space group of I4/mcm. Structurally ill-defined powders are common.60−62 However, analysis of precursor and product crystal structures reveal that topotactic pathways to product are possible. Specifically, Sr2Ta2O7 belongs to the orthorhombic crystal system with the Cmcm space group. The structure consists of layers of cornershared TaO6 octahedra. There are two unique crystallographic Sr centers, with one lying within and one lying between the layers. When converting to SrTaO2N, the layered structure of the parent oxide disappears, and the two unique Sr centers become one, forming corner-shared TaO(N)6 connected via Sr cations. As the crystallinity and crystal faceting of the product are determined by the precursor phase in topotactic chemistry, synthesis of oxides with defined morphology should template the oxynitride phase as evident in comparing Sr2Ta2O7 and SrTaO2N along the [010] and [100] directions, respectively (Figure 5A). 1336
DOI: 10.1021/acsenergylett.8b00518 ACS Energy Lett. 2018, 3, 1331−1344
ACS Energy Letters
Perspective
high surface area and low defect density of the nanoplates were cited reasons for their outstanding activity. Barium Tantalum Oxynitride (BaTaO2N). BaTaO2N is another promising member of the perovskite-related oxynitride family of photocatalysts for water splitting. The oxynitride has the space group of Pm3̅m. Although there are several synthetic approaches to form BaTaO2N,63−66 molten flux methods as described in the Ta3N5 and SrTaO2N sections of this Perspective have been shown to be suitable for shape-controlled samples.67 These methods can produce high-quality crystals and prevent particle agglomeration, which result in better photocatalytic activity. As the previous examples highlighted, nitride and oxynitride formation typically proceeds through two steps. First, an oxide precursor is formed within a flux. Then, the oxide is converted to the nitride or oxynitride by ammonolysis. However, in the case of BaTaO2N, Teshima et al. showed that the oxynitride could be produced in one step, greatly simplifying the synthesis.67 They investigated the effects of the flux (including KCl, KI, KF, MgCl2, CaCl2, SrCl2, BaCl2, K2SO4, K2MoO4, and K2CO3), solute concentration (1−50 mol %), reaction time (0−10 h), and temperature (700 °C−950 °C) on forming BaTaO2N crystals from BaCO3 and Ta2O5 under an NH3 flow. They found that cubic BaTaO2N crystals with an average size of 125 nm could be grown from the KCl flux at 950 °C while flowing NH3 for 10 h, as shown in the TEM image in Figure 6A. Critical to this outcome was the solute concentration, in which only BaTaO2N was produced when the solute concentration was ≥10 mol %. In contrast, when the solute concentration was ≤10 mol %, Ta3N5 was produced as a secondary phase. These findings highlight the importance of controlling supersaturation conditions during synthesis to achieve phase-controlled, high-quality samples. This concept is well-established in traditional colloidal routes to nanomaterials and can be similarly exploited in molten salt syntheses through judicious selection of flux, precursors, and heating profile.44 The growth of BaTaO2N cubes was proposed to proceed through the decomposition of BaCO3 and subsequent dissolution of BaO and Ta2O5 in the KCl flux. Then, BaTa2O6 and Ba5Ta4O15 intermediates crystallized, and their formations were supported by PXRD studies at different times and temperatures of the synthesis. Finally, these intermediates dissolved and nonporous BaTaO2N cubes crystallized directly from the flux once the system had enough active atomic N centers. The proposed reaction mechanism is summarized in Figure 6B. The cubic morphology is attributed to the expression of lower-energy facets and the capping effect of K+ ions from the flux.
While topotactic transition was not discussed, such a mechanism may account for the synthesis of columnar SrTaO2N by Teshima and co-workers.58 Specifically, columnar Sr2Ta2O7 crystals were grown by a molten salt method and converted to SrTaO2N by ammonolysis. Stoichiometric amounts of SrCO3 and Ta2O5 were heated along with SrCl2 as a flux between 600− 1000 °C; the solute concentration was varied between 1−50 mol % to identify the optimal conditions. Crystals of Sr2Ta2O7 were produced at 20 mol % and 1000 °C. These crystals had a columnar structure in the size range of 0.1−30 μm, with relatively smooth surfaces. The Sr2Ta2O7 crystals were then heated at 950 °C for 15 h under an NH3 flow to obtain SrTaO2N, as confirmed by PXRD. The resultant orange powder consisted of the same columnar crystal morphologies as the parent oxide powder, with crystals 0.1−25 μm in size (Figure 5B). The similarity in morphology between the parent oxide and product support a topotactic process. The band gap of the oxynitride was 2.1 eV (compared to 4.5 eV for the parent oxide) and is consistent with the orange color of the SrTaO2N sample. Although no photocatalytic measurements were carried out with the SrTaO2N sample, this study demonstrates that the size and shape of oxynitrides can be templated by the precursor oxide phase when there are similarities in crystal structure. Motivated by this idea that morphology control can be achieved through topotactic transition, our group synthesized oxide nanoplates using AMSS.21 These nanoplates were then used as templates for topotactic nitridation to the oxynitride phase, with preservation of nanocrystal shape. Specifically, an aqueous solution of TaClx(OCH3)5−x and SrCl2 in a 1:12 molar ratio was nebulized and transported into a furnace (800 °C), where product formation occurred. SrCl2 served as both a reagent and the flux, and Sr2Ta2O7 nanoplates were produced. They were then nitrided to SrTaO2N at 950 °C for 15 h, producing plates templated by the original oxide phase. Slight irregularities in morphology were evident, along with porosity (Figure 5C), and these defects likely arise from the reduction in unit cell dimension. In addition to direct nitridation, the AMSS-derived Sr2Ta2O7 nanoplates were also nitrided in a SrCl2 flux, in a manner similar to that of Teshima and cowokers.58 In this case, SrTaO2N polyhedra were formed instead of nanoplates because of dissolution of the oxide template particles within the flux (Figure 5D). While well-defined shape-controlled nanocrystals were not produced, these particles clearly have different faceting, which is a first step toward such samples. These samples were compared to conventional SrTaO2N polycrystals synthesized from BaCO3 and Ta2O5 (Figure 5E). The surface areas of these powders were 11.7, 5.4, and 6.6 m2 g−1 for the nanoplates, polyhedra, and polycrystals, respectively. HER and OER for the samples (bare and with either 3 wt % Pt or 2 wt % CoOx cocatalysts) were measured (300 W xenon lamp with a 400 nm ≤ λ ≤ 500 nm cutoff filter; HER: 20 mg photocatalyst was dispersed in 12 mL of aqueous solution containing 10 vol % methanol; and OER: 20 mg photocatalyst and La2O3 as a pH buffer were dispersed in 12 mL of 0.05 M AgNO3; Figure 5F,G). The O2 evolution and H2 evolution was roughly three and ten times higher for the SrTaO2N nanoplates than that of the polyhedra and polycrystals, respectively. The AQE measurements for OER and HER were 6.1% and 0.20%, 1.6% and 0.075%, and 0.60% and 0.019% for the SrTaO2N nanoplates, polyhedra, and polycrystals, respectively. The reported AQE measurement for OER of the SrTaO2N nanoplates is among the highest reported for this material. This work highlights that control over the structural features of photocatalysts can greatly enhance the AQE of photocatalysts. In this case, the
Controlling supersaturation conditions is important during synthesis to achieve phase-controlled, high-quality samples. The UV−visible diffuse reflectance spectrum of the cubic BaTaO2N has an absorption onset of about 600 nm. OER and HER of the samples were measured (λ > 420 nm; OER, 100 mg photocatalyst and La2O3 as a pH buffer were dispersed in 100 mL of 10 mM AgNO3; HER, 100 mg photocatalyst was dispersed in 100 mL of aqueous solution containing 10 vol % methanol). Without a cocatalyst, no H2 and O2 evolution was observed. BaTaO2N crystals were loaded with Pt (0.3 wt % for HER) and CoOx (2 wt % Co for OER), followed by heat treatment at 500 °C for 1 h under NH3 and H2 flow, respectively. Photocatalytic activity was greater with cocatalysts as was the case in literature 1337
DOI: 10.1021/acsenergylett.8b00518 ACS Energy Lett. 2018, 3, 1331−1344
ACS Energy Letters
Perspective
Figure 6. (A) TEM image, (B) possible growth mechanism, and reaction time courses of (C) H2 and (D) O2 evolution over cubelike BaTaO2N crystals. Reprinted from ref 67. Copyright 2015 American Chemical Society.
TiO2 in an equal molar flux of NaCl/KCl that was heated to 1423 K at a rate of 10 K min−1 for 5 h followed by cooling to 1073 K at the same rate before cooling to room temperature. The samples prepared by both techniques were nitrided at 1223 K to obtain LaTiO2N. Figure 7A is a SEM image of the well-dispersed bricklike LaTiO2N particles synthesized from the flux method. These particles were porous. The polymerized LaTiO2N were aggregated nanoparticles with no porosity, as can be seen in Figure 7B. Both samples have the anticipated absorption edge of ∼600 nm. However, the LaTiO2N synthesized by the flux-assisted method has a lower absorption background, which suggests a lower defect density compared to the sample prepared by the polymerization complex route. The O2 evolution rates of the samples with and without a cocatalyst (300 W xenon lamp with λ > 420 nm, 0.2 g of catalysts, 0.2 g of La2O3 buffer, and 200 mL of 0.05 M AgNO3 solution) were measured. The LaTiO2N produced by the flux-assisted method had the highest activity, which was promoted by cocatalyst addition, either 2 wt % CoOx (736 μmol h−1) or 2 wt % IrO2 (170 μmol h−1) as shown in Figure 7C. In contrast, the LaTiO2N synthesized from the polymerized complex had O2 evolution rates of 489 and 146 μmol h−1 with 2 wt % CoOx and 2 wt % IrO2, respectively. The flux-assisted LaTiO2N had an AQE of 27.1 ± 2.6% in the initial 15 min, which is 5−6 times greater than that of LaTiO2N prepared from the polymerized complex with an IrO2 cocatalyst. The rate of O2 evolution decreased with increasing time because of the decrease in Ag+ concentration and metallic Ag deposition on the surface of LaTiO2N, which covers the active sites available for photocatalysis. The higher activity of the LaTiO2N synthesized by the flux-assisted method was attributed in part to the high surface area and low defect density of the sample, all of which minimize the density of recombination sites for the photogenerated carriers. As with BaTaO2N, Teshima et al. also applied their flux-assisted direct nitridation approach to LaTiO2N.72 This method yielded BaTaO2N samples with low defect density, high crystallinity, and enhanced photocatalytic activity. Similar outcomes were achieved with LaTiO2N. Numerous conditions were examined, where the solute concentration (0.1−50 mol %), the flux
reports of BaTaO2N modified with 0.3 wt % Pt and 1.5 wt % IrO2 cocatalyst (Figure 6C,D).68,69 The higher activity was attributed to fewer defects in the crystals. This claim was supported by transient absorption spectroscopy (TAS) (20 000−1000 cm−1) of the BaTaO2N samples. Three peaks were observed. These include a relatively sharp peak at ∼16 000 cm−1, a broad absorption at 13 000−6000 cm−1, and one at 3000−1500 cm−1, which were assigned to trapped holes, deeply trapped electrons at defects, and free or shallowly trapped electrons, respectively.70,71 The intensity of the ∼16 000 cm−1 peak is much greater for BaTaO2N prepared by the two-step process compared to BaTaO2N prepared directly, and this feature indicates that the BaTaO2N produced directly has more photogenerated charge carriers available for the desired redox processes. This study showed that the method of nitridation can influence the defect density within nitrides and oxynitrides and in turn the catalytic performance of the material.
The method of nitridation can influence the defect density within nitrides and oxynitrides and in turn the catalytic performance of the material. Lanthanum Titanium Oxynitride (LaTiO2N). LaTiO2N is another promising perovskite-based photocatalyst similar to SrTaO2N and BaTaO2N. LaTiO2N belongs to the Imma space group and has a band gap of 2.1 eV, with a ∼600 nm absorption edge. Like the other oxynitrides, synthesis of LaTiO2N can proceed through a two-step process in which an oxide phase is first produced and then nitrided to the desired material. For example, Domen et al. synthesized La2Ti2O7 by both polymerization complex and flux methods.28 The polymerized La2Ti2O7 was prepared by first mixing titanium tetraisopropoxide, ethylene glycol, and citric acid which was heated at 333 K, followed by the stoichiometric addition of La(NO3)3 in methanol. With additional heating, a transparent gel formed, which was then carbonized and finally calcined at 773 K for 12 h. Flux-assisted La2Ti2O7 was synthesized by mixing stoichiometric amounts of La2O3 and 1338
DOI: 10.1021/acsenergylett.8b00518 ACS Energy Lett. 2018, 3, 1331−1344
ACS Energy Letters
Perspective
process was enhanced with the cocatalyst, CoOx (2 wt % Co). Within the first 2 h, O2 evolution showed a greater rate for bare LaTiO2N prepared directly (82 μmol h−1) compared to the bare LaTiO2N synthesized in a two-step process (37 μmol h−1). Similarly, with the CoOx cocatalyst, LaTiO2N prepared directly (204 μmol h−1) showed a greater rate compared to LaTiO2N from the two-step process (177 μmol h−1), as shown in Figure 7F. In both cases, LaTiO2N prepared directly performed the best and is likely from its greater crystallinity, better morphology, and fewer defects, just as was observed with the BaTaO2N system. Finally, we note that porous LaTiO2N microspheres were prepared by ammonolysis of oxide spheres from ultrasonic spray synthesis;74 given the flux routes reported for LaTiO2N and its precursor oxide, we anticipate that far from equilibrium forms may be possible by AMSS, as reported for SrTaO2N. d10 Metal Oxynitrides. In contrast to d0 metal nitrides or oxynitrides, there are fewer d10 metal nitride photocatalysts reported. Ga3N4 was the first example of a non-oxide powdered photocatalyst reported for water splitting under UV irradiation.75 Even though bare GaN is not active, when loaded with Rh2−xCrxO3 as a cocatalyst or doped with Mg, Be, or Zn, GaN is active under UV radiation.76−78 A particularly interesting system is the solid solution of GaN and ZnO, (Ga1−xZnx)(N1−xOx) (GZNO), which absorbs visible light even though its parent semiconductors, GaN and ZnO, are wide band gap semiconductors (3.4 and 3.2 eV, respectively).79−82 In the presence of a suitable cocatalyst, GZNO is capable of overall water splitting under visible light irradiation (λ > 400 nm). In fact, a quantum yield of 5.9% under visible light irradiation between 420 and 440 nm in the absence of sacrificial agents has been reported with GZNO.83 According to DFT calculations, the band gap decrease upon solid solution formation arises from the repulsion of N 2p and Zn 3d electrons in the valence band, which defines the valence band maximum and can be tuned by the Zn fraction within the material.84 GZNO has a wurtzite crystal structure just like the parent semiconductors and can be achieved by heating (>800 °C) a mixture of Ga2O3 and ZnO under a flow of NH3.79,81 This reaction proceeds through a spinel intermediate, ZnGa2O4. Because of this spinel intermediate and the Zn volatility at high temperatures, the fraction of ZnO (x) is limited by this method. Significant efforts have been directed toward increasing the Zn content to narrow the band gap of this oxynitride for solar water splitting.81,83,85−87 For example, Dukovic and co-workers synthesized GZNO (0.30 ≤ x ≤ 0.87) nanocrystals by heating a mixture of ZnO (∼10 nm) and ZnGa2O4 (∼5 nm) at 650 °C for 10 h under NH3 flow.85 Here, low nitridation temperatures could be used because of the nanoscale precursor, and these milder conditions inhibited the loss of Zn and produced highquality GZNO nanocrystals. The particles are faceted and similar to the ZnO nanocrystal precursors, which were hexagonal-cone shaped. The synthesis method enabled the particle size to be held constant (dimensions, ∼18 nm) over a compositional range of 0.30 ≤ x ≤ 0.87. The band gap of the GZNO samples decreased from 2.7 eV (x = 0.30) to 2.2 eV (x = 0.87) with increasing x, which corresponds to a 260% improvement in solar photon absorption. Unfortunately, no photocatalytic measurements were evaluated in this study. However, the nanocrystalline nature and tunable composition and band gap of GZNO (0.30 ≤ x ≤ 0.87) are beneficial in realizing high photocatalytic activity. In a subsequent study, Dukovic et al. expanded the compositional range of GZNO from 0.06 to 0.98 by heating a mixture of ZnO (∼10 nm) and ZnGa2O4 (∼5 nm) at either 750 °C
Figure 7. SEM images of LaTiO2N prepared by nitriding La2Ti2O7 synthesized via (A) flux method using 0.5 NaCl−0.5 KCl and (B) polymerized complex method. (C) Time course of O2 evolution on LaTiO2N samples with different cocatalysts under visible light irradiation (λ > 420 nm). SEM images of LaTiO2N crystallites grown (D) directly from NH3-assisted KCl flux at 950 °C for 10 h (one-step) and (E) from nitridation of La2Ti2O7 at 950 °C for 10 h (two-step). (F) Time courses of O2 evolution of bare and CoOx-loaded LaTiO2N crystallites. Reprinted from refs 28 and 72. Copyright 2012 and 2015 American Chemical Society.
(Na2CO3, K2CO3, KCl, NaCl, BaCl2, and KF), and heating time (10 min, 1, 3, 5, 7, and 10 h) were varied. La2O3 and TiO2 were precursors, along with NH3 as the nitrogen source. Interestingly, the carbonate fluxes Na2CO3 and K2CO3 produced LaTiO2N crystallites consisting of platelet (ca. 9.3 μm) and stack-layered (ca. 2.7 μm) morphologies, respectively, whereas halide fluxes (chloride and fluoride) yielded rounded crystallites with average sizes ranging from 100 nm to 2 μm. The differences in crystal size and shape are likely the result of differences in solute solubility in a given flux. Yet again, supersaturation can be manipulated in flux methods to achieve crystals with different morphologies. Highquality crystallites of LaTiO2N (average size 120 ± 39 nm) synthesized from the KCl flux with a solute concentration of 5 mol % at 950 °C for 10 h (Figure 7D) were selected for photocatalytic evaluation. For comparison, LaTiO2N was also synthesized by nitriding La2Ti2O7 prepared using the flux method described by Domen (Figure 7E).28 Both samples had the anticipated absorption edge of ∼600 nm but also demonstrated a strong background absorption at longer wavelengths, which is indicative of reduced Ti defects and anion vacancies.73 TAS measurements showed that the LaTiO2N prepared directly had a higher number of reactive holes and lower defect density compared to LaTiO2N prepared by the two-step process. Water oxidation with both LaTiO2N samples were undertaken (300 W xenon lamp with λ > 420 nm, 0.1 g of catalysts, 0.2 g of La2O3 buffer, and 200 mL of 10 mM AgNO3 solution), and the 1339
DOI: 10.1021/acsenergylett.8b00518 ACS Energy Lett. 2018, 3, 1331−1344
ACS Energy Letters
Perspective
(for x values of 0.06 and 0.24) or 650 °C (for x values >0.24) for 10 h under NH3 flow.86 The particle diameters were in the range of 10−20 nm. The authors proposed that the formation of GZNO nanocrystals occurs at the interface of ZnO and ZnGa2O4 by nucleating GZNO topotactically on ZnO. For the compositional range studied, the lowest band gap achieved was 2.25 eV with x = 0.87, and a blue shift was observed above x = 0.87. (The band gaps of x = 0.06 and 0.98 were 2.9 and 2.6 eV, respectively.) Bulk GZNO (x = 0.40) were prepared using a previously reported synthetic method.80 The bulk and nano GZNO (x = 0.40, 0.52, and 0.87) thin films were prepared by mixing each sample with ethylcellulose and α-terpinol, which was then applied onto a FTO glass electrode followed by calcination at 400 °C for 1 h under air. The PEC activities of bulk and nanoscale GZNO films were measured (300 W xenon lamp with a 460 nm longpass filter, 0.5 M Na2SO4 electrolyte at pH 4.5). The bulk GZNO film produced photocurrents of 1−2 μA cm−2 at 0.5 V. The nanoscale GZNO films, on the other hand, showed higher photocurrents by a factor of 2 and 4 for x = 0.40 and x = 0.52, respectively. Even though x = 0.87 had the lowest band gap, the highest PEC activity was observed for x = 0.52, indicating the observed photocurrents were a result of multiple parameters, including charge transport efficiency and band edge potentials. Many studies of GZNO have focused on obtaining lower band gap compositions by changing the Zn fraction; however, less effort has emphasized morphological control of GZNO. Still, some examples exist beyond the work from Dukovic et al. For instance, Wang et al. synthesized GZNO (0.46 < x < 0.81) by nitridation of layered double hydroxides (LDHs) containing Zn3+, Ga3+, and CO32− ions.83 Specifically, the ZnGa-LDH was prepared by dissolving Ga2O3 and ZnO with [Zn]/[Ga] = 1, 2, 3, 4, and 5 in an aqueous HNO3 (6 M) solution followed by pH adjustment to 8 with a mixture of NaOH (2 M) and Na2CO3 (1 M). The as-synthesized LDHs were hexagonal nanoplates with diameters of ∼300 nm and a thickness of 30−50 nm. The nanoplate morphology became a bit irregular with higher [Zn]/[Ga]. Shown in Figure 8A are the high-resolution TEM (HRTEM) and selected area electron diffraction (SAED) images of the sample prepared at [Zn]/[Ga] = 2, which demonstrates well-crystallized, hexagonally shaped particles. These ZnGa-LDH precursors were then converted to GZNO (0.46 < x < 0.81) by heating at 800 °C under an NH3 flow for 30 min. Figure 8B shows the TEM and SAED of the GZNO sample from the ZnGa-LDH precursor where [Zn]/[Ga] = 2. Clearly, the hexagonal shape was preserved after nitridation although the interlayer region of the precursor was lost, leading to a porous structure. The porosity was due to the loss of CO2 and H2O during nitridation. For photocatalytic comparison, a reference GZNO sample ([Zn/Ga] = 2) was synthesized by heating ZnO and Ga2O3 at 850 °C under an NH3 flow for 12 h. The use of the LDH precursor greatly decreased the nitridation time as the layered structure eases NH3 migration into the interlayers and the Zn2+ and Ga3+ ions are homogeneously distributed at an atomic scale. The band gaps of the GZNO samples decreased with increasing x (2.60 eV at x = 0.46 and 2.37 eV at x = 0.81). The photocatalytic activity of the LDH-derived and reference GZNO samples were evaluated by monitoring the reduction of Cr6+ ions under visible light irradiation (300 W xenon lamp with a 400 nm cutoff filter, 0.2 g of photocatalyst, 100 mL of 40 mg L−1 K2Cr2O7). Figure 8C shows the photoreduction of Cr6+ ions over the GZNO solid solutions loaded with 1 wt % Pt. The LDHderived GZNO samples exhibited higher activity compared to the reference, with the highest activity observed from the sample
Figure 8. (A) HRTEM and SAED images of ZnGa-LDH. (B) TEM and SAED images of GZNO. (C) The photoreduction of Cr6+ ions over 1 wt % Pt loaded GZNO solid solutions as a function of the visible irradiation time. (x = 0.46, 0.66, 0.72, 0.78, and 0.81 represent sample A, B, C, D, and E, respectively.) SEM images of (D) ZnGaLDH, (E) GZNO-NH3, and (F) GZNO-O2/NH3. Red and yellow segmented lines represent the preserved platelike shape in GZNONH3 and GZNO-O2/NH3, respectively. The yellow arrows indicate the presence of pores in GZNO-NH3. The solid orange box represents a high-magnification image of the area enclosed in the fragmented orange box, and red arrows indicate the pyramidal features in GZNO-O2/NH3. Reprinted with permission from refs 83 and 87. Copyright 2011 the Royal Society of Chemistry and 2016 American Chemical Society.
with x = 0.66. Significantly, the use of the LDH precursor minimizes Zn loss during nitridation, indicating that the Zn fraction in GZNO can be controlled facilely by adjusting the Zn/Ga ratio in the ZnGa-LDH precursor solution. In contrast, GZNO prepared by traditional solid-state methods can have substantial Zn loss compared to the input ratio due to the reduction of ZnO to metallic Zn during nitridation, which volatilizes at high temperatures. Although not discussed by the authors, the enhanced photocatalytic activity was possibly attributed to the crystallinity, defined nanostructural features, and high surface area of the GZNO nanoplates. We also note that conservation of the platelike morphology during nitridation was likely from topotactic nitridation of ZnGa-LDH to GZNO. Our group has subsequently adapted the ZnGa-LDH topotactic transformation route to Zn-rich GZNO and provided insight into the origins of defects detrimental to photocatalytic activity.87 Specifically, Zn(NO3)2 and Ga(NO3)3 were mixed together in an aqueous solution at a 2:1 mol ratio, and the pH of the solution was adjusted to 8 with NaOH and Na2CO3 to produce LDHs, which could be used as a precursor to GZNO. An SEM image of the GaZn-LDH microcrystals, which adopt a hexagonal platelike morphology, is shown in Figure 8D. The GaZn-LDH precursor was then nitrided under two different 1340
DOI: 10.1021/acsenergylett.8b00518 ACS Energy Lett. 2018, 3, 1331−1344
ACS Energy Letters
Perspective
conditions. The first was a conventional ammonolysis process at 800 °C for 10 h, and the second introduced O2 into the NH3 stream (2 vol % O2). SEM images of the products GZNO-NH3 and GZNO-O2/NH3 are shown in panels E and F of Figure 8, respectively. In both cases, the platelike features of the ZnGaLDH microcrystals were retained. Pores are evident in GZNONH3, which is an indication of Zn loss. Nominal cation concentrations were obtained from ICP-OES, which showed the Zn loss in GZNO-NH3 (x = 0.22) was higher than that of GZNO-O2/NH3 (x = 0.42), indicating the increase in pO2 limits Zn2+ reduction. Fewer pores are introduced into GZNO-O2/NH3; however, pyramidal overgrowths along the basal planes of the platelets appear. Crystallographic analysis of the ZnGa-LDH to GZNO plates support that the product is produced through topotactic transformation. The band gaps of GZNO-NH3 and GZNO-O2/NH3 were 2.71 and 2.30 eV, consistent with the different Zn fractions of the samples. Interestingly, the GZNO-O2/NH3 sample exhibited less Urbach tailing near the absorption edge and less peak broadening in the diffraction patterns, suggesting lower structural disorder associated with GZNO-O2/NH3 versus GZNO-NH3. Even though no photocatalytic measurements were performed, this study highlights a new direction to suppress Zn volatilization by maintaining a low partial pressure of O2, which leads to smaller band gap materials with minor structural defects. Incorporation of this nitridation method with other precursors may yield highperformance metal oxynitride materials. Conclusion and Future Outlook. Photocatalytic water splitting is a promising route toward generating clean and renewable hydrogen fuel. Efficient use of sunlight can be achieved by photocatalysts such as oxynitrides and nitrides owing to their narrow band gaps, which allow for visible light excitation. Moreover, photocatalysts with high crystallinity, high surface area, and smaller particle size can enhance activity by decreasing the probability of electron−hole recombination. This Perspective highlighted recent advances toward three types of shape-controlled metal nitride and oxynitride particles: Ta3N5 and TaON (d0 metal), perovskite-like oxynitrides (d0 metal), and d0 metal oxynitrides. In compiling these case studies, several trends and future research directions emerge. First, many of the examples of metal oxynitride and nitride particles with morphology control were achieved from nitridation of metal oxide or LDH templates which already displayed morphology control. That is, the structure of the precursor phase can be transferred to the final oxynitride or nitride phase during nitridation. For products that contain multiple metals, selecting precursors with the metal stoichiometry can be beneficial and ensure phase control. Also, selecting precursors capable of topotactic transformation can preserve the single-crystallinity in the product phase, which is beneficial to photocatalytic performance. Otherwise, this approach can result in polycrystalline products. Notably, porosity can be introduced because of the differences in metal-to-oxide ratio of the precursor phase and the metal-to-nitride ratio of the product. Such a structural change can be beneficial to photocatalysis by producing high surface area materials but can also be detrimental by introducing defect sites.
Still, this insight indicates that advances in the synthesis of complex oxide materials can advance the synthesis and study of oxynitrides and nitrides. Second, flux-assisted crystal growth is a promising avenue to achieve structurally defined oxynitride and nitride photocatalysts. In many cases, the flux methods were used to prepare shapecontrolled complex oxides, where conventional colloidal routes are unavailable and solid-state heating methods prohibit shape control; these shape-controlled complex oxides are then used as templates to the desired oxynitride or nitride phase. Exciting developments include the coupling of molten salt chemistry with aerosol methods and direct ammonolysis of fluxes. The former limits crystal growth to the dimensions of the aerosol droplets, with nanoscale templates achieved. The latter provides a onestep route to oxynitride and nitride particles in a fluid phase, with direct nucleation of the product. This advancement provides a means to manipulate the shape of the product crystals through judicious control of supersaturation and the capping effects of the ions that comprise the flux. Still, understanding the size- and shape-effects of metal oxynitride and nitride materials is in its infancy. This gap in knowledge is largely a result of insufficient control over the synthesis of these materials as often only thermodynamically favored shapes can be achieved or when materials with different structural features are achieved, other features of the catalyst change (e.g., composition or crystallite size). These outcomes inhibit rigorous structure−function−reactivity studies. Continued advances in the synthesis of oxynitrides and nitride particles could fill this gap, including lower-temperature and colloidal routes that employ reactive nitrogen sources. Also, with shapecontrolled oxynitride and nitride materials achieved, selfassembly methods20,88 could be adopted to control the architecture of these materials to maximize surface area and light absorption. In addition, advanced characterization of these materials can help to better understand their electronic structures, whether it be the application of TAS to understand defect states as a function of nitridation conditions,71 X-ray and neutron scattering methods to elucidate anion clustering effects in oxynitride solid solutions,87,89−91 or something new. In fact, Bartlett and Brancho noted that complete control over the nitrogen content that is incorporated during ammonolysis of oxides is limited and reliable quantification techniques to determine the nitrogen content of products are needed.92 Hence, new synthetic approaches with precise control of the anion stoichiometry coupled with advanced and/or improved characterization techniques need further exploration. Overall, there have been tremendous advances in the synthesis and use of shape-controlled oxynitride and nitride materials for solar energy applications, and this trajectory is anticipated to accelerate as greater understanding of their formation and fundamental properties increase.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Sara E. Skrabalak: 0000-0002-1873-100X
Advances in the synthesis of complex oxide materials can advance the synthesis and study of oxynitrides and nitrides.
Notes
The authors declare no competing financial interest. Biographies Sara E. Skrabalak is the James H. Rudy Professor in the Department of Chemistry at Indiana UniversityBloomington. Her research group is 1341
DOI: 10.1021/acsenergylett.8b00518 ACS Energy Lett. 2018, 3, 1331−1344
ACS Energy Letters
Perspective
(19) Fu, J.; Skrabalak, S. E. Aerosol Synthesis of Shape-Controlled Template particles: A Route to Ta3N5 Nanoplates and Octahedra as Photocatalysts. J. Mater. Chem. A 2016, 4, 8451−8457. (20) Wang, Z.; Hou, J.; Yang, C.; Jiao, S.; Huang, K.; Zhu, H. Hierarchical Metastable γ-TaON Hollow Structures for Efficient Visible-Light Water Splitting. Energy Environ. Sci. 2013, 6, 2134−2144. (21) Fu, J.; Skrabalak, S. E. Enhanced Photoactivity from SingleCrystalline SrTaO2N Nanoplates Synthesized by Topotactic Nitridation. Angew. Chem., Int. Ed. 2017, 56, 14169−14173. (22) Hitoki, G.; Takata, T.; Kondo, J. N.; Hara, M.; Kobayashi, H.; Domen, K. An Oxynitride, TaON, as an Efficient Water Oxidation Photocatalyst Under Visible Light Irradiation (λ ≤ 500 nm). Chem. Commun. 2002, 16, 1698−1699. (23) Yamasita, D.; Takata, T.; Hara, M.; Kondo, J. N.; Domen, K. Recent Progress of Visible-Light-Driven Heterogeneous Photocatalysts for Overall Water Splitting. Solid State Ionics 2004, 172, 591−595. (24) Siritanaratkul, B.; Maeda, K.; Hisatomi, T.; Domen, K. Synthesis and Photocatalytic Activity of Perovskite Niobium Oxynitrides with Wide Visible-Light Absorption Bands. ChemSusChem 2011, 4, 74−78. (25) Liu, M.; You, W.; Lei, Z.; Zhou, G.; Yang, J.; Wu, G.; Ma, G.; Luan, G.; Takata, T.; Hara, M.; Domen, K.; Li, C. Water Reduction and Oxidation on Pt-Ru/Y2Ta2O5N2 Catalyst Under Visible Light Irradiation. Chem. Commun. 2004, 0, 2192−2193. (26) Liu, M.; You, W.; Lei, Z.; Takata, T.; Domen, K.; Li, C. Photocatalytic Water Splitting to Hydrogen Over a Visible Light-Driven LaTaON2 Catalyst. Chin. J. Catal. 2006, 27, 556−558. (27) Kasahara, A.; Nukumizu, K.; Hitoki, G.; Takata, T.; Kondo, J. N.; Hara, M.; Kobayashi, H.; Domen, K. Photoreactions on LaTiO2N Under Visible Light Irradiation. J. Phys. Chem. A 2002, 106, 6750−6753. (28) Zhang, F.; Yamakata, A.; Maeda, K.; Moriya, Y.; Takata, T.; Kubota, J.; Teshima, K.; Oishi, S.; Domen, K. Cobalt-Modified Porous Single-Crystalline LaTiO2N for Highly Efficient Water Oxidation Under Visible Light. J. Am. Chem. Soc. 2012, 134, 8348−8351. (29) Zhang, P.; Zhang, J.; Gong, J. Tantalum-Based Semiconductors for Solar Water Splitting. Chem. Soc. Rev. 2014, 43, 4395−4422. (30) Baxter, D. V.; Chisholm, M. H.; Gama, G. J.; DiStasi, V. F.; Hector, A. L.; Parkin, I. P. Molecular Routes to Metal Carbides, Nitrides, and Oxides. 2. Studies of the Ammonolysis of Metal Dialkylamides and Hexamethyldisilylamides. Chem. Mater. 1996, 8, 1222−1228. (31) Brown, G. M.; Maya, L. Ammonolysis Products of the Dialkylamides of Titanium, Zirconium, and Niobium as Precursors to Metal Nitrides. J. Am. Ceram. Soc. 1988, 71, 78−82. (32) Choi, K.-S. Shape Effect and Shape Control of Polycrystalline Semiconductor Electrodes for use in Photoelectrochemical Cells. J. Phys. Chem. Lett. 2010, 1, 2244−2250. (33) Choi, K.-S. Shape Control of Inorganic Materials via Electrodeposition. Dalton Trans. 2008, 0, 5432−5438. (34) Steinmiller, E. M. P.; Choi, K. S. Photochemical Deposition of Cobalt-Based Oxygen Evolving Catalyst on a Semiconductor Photoanode for Solar Oxygen Production. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 20633−20636. (35) Hisatomi, T.; Otani, M.; Nakajima, K.; Teramura, K.; Kako, Y.; Lu, D.; Takata, T.; Kondo, J. N.; Domen, K. Preparation of Crystallized Mesoporous Ta3N5 Assisted by Chemical Vapor Deposition of Tetramethyl Orthosilicate. Chem. Mater. 2010, 22, 3854−3861. (36) Ma, S. S.; Hisatomi, T.; Maeda, K.; Moriya, Y.; Domen, K. Enhanced Water Oxidation on Ta3N5 Photocatalysts by Modification with Alkaline Metal Salts. J. Am. Chem. Soc. 2012, 134, 19993−19996. (37) Liu, G.; Shi, J.; Zhang, F.; Chen, Z.; Han, J.; Ding, C.; Chen, S.; Wang, Z.; Han, H.; Li, C. A Tantalum Nitride Photoanode Modified with a Hole-Storage Layer for Highly Stable Solar Water Splitting. Angew. Chem., Int. Ed. 2014, 53, 7295−7299. (38) Cong, Y.; Park, H. S.; Wang, S.; Dang, H. X.; Fan, F.-R. F.; Mullins, C. B.; Bard, A. J. Synthesis of Ta3N5 Nanotube Arrays Modified with Electrocatalysts for Photoelectrochemical Water Oxidation. J. Phys. Chem. C 2012, 116, 14541−14550. (39) Wang, L.; Nguyen, N. T.; Zhou, X.; Hwang, I.; Killian, M. S.; Schmuki, P. Enhanced Charge Transport in Tantalum Nitride
recognized for pioneering syntheses to nanoscale materials with size, shape, and compositional control for diverse applications. Dileka Abeysinghe is a postdoctoral research associate in the research group of Prof. Sara E. Skrabalak where her interests are focused on shape-controlled oxynitride particles for solar energy applications. She completed her Ph.D. at the University of South Carolina with Prof. Hans-Conrad zur Loye.
■
ACKNOWLEDGMENTS We acknowledge financial support from NSF Grant DMR1608711.
■
REFERENCES
(1) Lewis, N. S.; Nocera, D. G. Powering the Planet: Chemical Challenges in Solar Energy Utilization. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 15729−15735. (2) Chen, X.; Shen, S.; Guo, L.; Mao, S. S. Semiconductor-Based Photocatalytic Hydrogen Generation. Chem. Rev. 2010, 110, 6503− 6570. (3) Walter, M. G.; Warren, E. L.; McKone, J. R.; Boettcher, S. W.; Mi, Q.; Santori, E. A.; Lewis, N. S. Solar Water Splitting Cells. Chem. Rev. 2010, 110, 6446−6473. (4) Osterloh, F. E. Inorganic Materials as Catalysts for Photochemical Splitting of Water. Chem. Mater. 2008, 20, 35−54. (5) Fujishima, A.; Honda, K. Electrochemical Photolysis of Water at a Semiconductor Electrode. Nature 1972, 238, 37−38. (6) Kudo, A.; Miseki, Y. Heterogeneous Photocatalyst Materials for Water Splitting. Chem. Soc. Rev. 2009, 38, 253−278. (7) Zou, Z.; Ye, J.; Sayama, K.; Arakawa, H. Direct Splitting of Water Under Visible Light Irradiation with an Oxide Semiconductor Photocatalyst. Nature 2001, 414, 625−627. (8) Xu, Y.; Schoonen, M. A. A. The Absolute Energy Positions of Conduction and Valence Bands of Selected Semiconducting Minerals. Am. Mineral. 2000, 85, 543−556. (9) Asahi, R.; Morikawa, T.; Ohwaki, T.; Aoki, K.; Taga, Y. VisibleLight Photocatalysis in Nitrogen-Doped Titanium Oxides. Science 2001, 293, 269−271. (10) Yang, X.; Wolcott, A.; Wang, G.; Sobo, A.; Fitzmorris, R. C.; Qian, F.; Zhang, J. Z.; Li, Y. Nitrogen-Doped ZnO Nanowire Arrays for Photoelectrochemical Water Splitting. Nano Lett. 2009, 9, 2331−2336. (11) Moriya, Y.; Takata, T.; Domen, K. Recent Progress in the Development of (Oxy)nitride Photocatalysts for Water Splitting Under Visible-Light Irradiation. Coord. Chem. Rev. 2013, 257, 1957−1969. (12) Fuertes, A. Metal Oxynitrides as Emerging Materials with Photocatalytic and Electronic Properties. Mater. Horiz. 2015, 2, 453− 461. (13) Masubuchi, Y.; Sun, S. K.; Kikkawa, S. Processing of Dielectric Oxynitride Perovskites for Powders, Ceramics, Compacts and Thin Films. Dalton Trans. 2015, 44, 10570−10581. (14) Umezawa, N.; Ye, J. Role of Complex Defects in Photocatalytic Activities of Nitrogen-Doped Anatase TiO2. Phys. Chem. Chem. Phys. 2012, 14, 5924−5934. (15) Zhang, J.; Jin, Z.; Feng, C.; Yu, L.; Zhang, J.; Zhang, Z. ESR Study on the Visible Photocatalytic Mechanism of Nitrogen-Doped Novel TiO2: Synergistic Effect of Two Kinds of Oxygen Vacancies. J. Solid State Chem. 2011, 184, 3066−3073. (16) Chun, W.-J.; Ishikawa, A.; Fujisawa, H.; Takata, T.; Kondo, J. N.; Hara, M.; Kawai, M.; Matsumoto, Y.; Domen, K. Conduction and Valence Band Positions of Ta2O5, TaON, and Ta3N5 by UPS and Electrochemical Methods. J. Phys. Chem. B 2003, 107, 1798−1803. (17) Fuertes, A. Synthesis and Properties of Functional Oxynitrides − from Photocatalysts to CMR Materials. Dalton Trans. 2010, 39, 5942− 5948. (18) Hitoki, G.; Ishikawa, A.; Takata, T.; Kondo, J. N.; Hara, M.; Domen, K. Ta3N5 as a Novel Visible Light-Driven Photocatalyst (λ < 600 nm). Chem. Lett. 2002, 31, 736−737. 1342
DOI: 10.1021/acsenergylett.8b00518 ACS Energy Lett. 2018, 3, 1331−1344
ACS Energy Letters
Perspective
Nanotube Photoanodes for Solar Water Splitting. ChemSusChem 2015, 8, 2615−2620. (40) Liu, G.; Ye, S.; Yan, P.; Xiong, F.; Fu, P.; Wang, Z.; Chen, Z.; Shi, J.; Li, C. Enabling an Integrated Tantalum Nitride Photoanode to Approach the Theoretical Photocurrent Limit for Solar Water Splitting. Energy Environ. Sci. 2016, 9, 1327−1334. (41) Takata, T.; Lu, D.; Domen, K. Synthesis of Structurally Defined Ta3N5 Particles by Flux-Assisted Nitridation. Cryst. Growth Des. 2011, 11, 33−38. (42) Bugaris, D. E.; zur Loye, H. C. Materials Discovery by Flux Crystal Growth: Quaternary and Higher Order Oxides. Angew. Chem., Int. Ed. 2012, 51, 3780−3811. (43) Abeysinghe, D.; Smith, M. D.; Yeon, J.; Morrison, G.; zur Loye, H. C. New Lanthanide Mixed-Valent Vanadium(III/IV) Oxosilicates, Ln4V(5‑x)Zn(x)Si4O22 (Ln = La, Ce, Pr, and Nd), Crystallizing in a Quasi Two-Dimensional Rutile-Based Structure. Inorg. Chem. 2016, 55, 1821−1830. (44) Chen, D. P.; Fu, J.; Skrabalak, S. E. Towards Shape Control of Metal Oxide Nanocrystals in Confined Molten Media. ChemNanoMat 2015, 1, 18−26. (45) Dabirian, A.; van de Krol, R. Resonant Optical Absorption and Defect Control in Ta3N5 Photoanodes. Appl. Phys. Lett. 2013, 102, 033905. (46) Chen, S.; Shen, S.; Liu, G.; Qi, Y.; Zhang, F.; Li, C. Interface Engineering of a CoO(x)/Ta3N5 Photocatalyst for Unprecedented Water Oxidation Performance Under Visible-Light-Irradiation. Angew. Chem., Int. Ed. 2015, 54, 3047−3051. (47) Dabirian, A.; van de Krol, R. High-Temperature Ammonolysis of Thin Film Ta2O5 Photoanodes: Evolution of Structural, Optical, and Photoelectrochemical Properties. Chem. Mater. 2015, 27, 708−715. (48) Abe, R.; Higashi, M.; Domen, K. Facile Fabrication of an Efficient Oxynitride TaON Photoanode for Overall Water Splitting into H2 and O2 Under Visible Light Irradiation. J. Am. Chem. Soc. 2010, 132, 11828− 11829. (49) Abe, R.; Takata, T.; Sugihara, H.; Domen, K. Photocatalytic Overall Water Splitting Under Visible Light by TaON and WO3 with an IO3−/I− Shuttle Redox Mediator. Chem. Commun. 2005, 0, 3829−3831. (50) Hara, M.; Nunoshige, J.; Takata, T.; Kondo, J. N.; Domen, K. Unusual Enhancement of H2 Evolution by Ru on TaON Photocatalyst Under Visible Light Irradiation. Chem. Commun. 2003, 0, 3000−3001. (51) Armytage, D.; Fender, B. E. F. Anion Ordering in TaON: A Powder Neutron-Diffraction Investigation. Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem. 1974, 30, 809−812. (52) Schilling, H.; Stork, A.; Irran, E.; Wolff, H.; Bredow, T.; Dronskowski, R.; Lerch, M. γ-TaON: A Metastable Polymorph of Tantalum Oxynitride. Angew. Chem., Int. Ed. 2007, 46, 2931−2934. (53) Maeno, Y.; Hashimoto, H.; Yoshida, K.; Nishizaki, S.; Fujita, T.; Bednorz, J. G.; Lichtenberg, F. Superconductivity in a Layered Perovskite without Copper. Nature 1994, 372, 532−534. (54) Uehara, M.; Mori, S.; Chen, C. H.; Cheong, S. W. Percolative Phase Separation Underlies Colossal Magnetoresistance in MixedValent Manganites. Nature 1999, 399, 560−563. (55) Homes, C. C.; Vogt, T.; Shapiro, S. M.; Wakimoto, S.; Ramirez, A. P. Optical Response of High-Dielectric-Constant Perovskite-Related Oxide. Science 2001, 293, 673−676. (56) Stoumpos, C. C.; Frazer, L.; Clark, D. J.; Kim, Y. S.; Rhim, S. H.; Freeman, A. J.; Ketterson, J. B.; Jang, J. I.; Kanatzidis, M. G. Hybrid Germanium Iodide Perovskite Semiconductors: Active Lone Pairs, Structural Distortions, Direct and Indirect Energy Gaps, and Strong Nonlinear Optical Properties. J. Am. Chem. Soc. 2015, 137, 6804−6819. (57) Zhang, L.; Song, Y.; Feng, J.; Fang, T.; Zhong, Y.; Li, Z.; Zou, Z. Photoelectrochemical Water Oxidation of LaTaO2N Under VisibleLight Irradiation. Int. J. Hydrogen Energy 2014, 39, 7697−7704. (58) Mizuno, Y.; Wagata, H.; Yubuta, K.; Zettsu, N.; Oishi, S.; Teshima, K. Flux Growth of Sr2Ta2O7 Crystals and Subsequent Nitridation to Form SrTaO2N Crystals. CrystEngComm 2013, 15, 8133−8138. (59) Davi, M.; Schrader, F.; Scholz, T.; Ma, Z.; Rokicinska, A.; Dronskowski, R.; Kustrowski, P.; Slabon, A. SrTaO2N Nanowire
Photoanode Modified with a Ferrihydrite Hole-Storage Layer for Photoelectrochemical Water Oxidation. ACS Appl. Nano Mater. 2018, 1, 869−876. (60) Atuchin, V. V.; Grivel, J.-C.; Zhang, Z. Core Level Photoemission Spectroscopy and Chemical Bonding in Sr2Ta2O7. Chem. Phys. 2009, 360, 74−78. (61) Kudo, A.; Kato, H.; Nakagawa, S. Water Splitting into H2 and O2 on New Sr2M2O7 (M = Nb and Ta) Photocatalysts with Layered Perovskite Structures: Factors Affecting the Photocatalytic Activity. J. Phys. Chem. B 2000, 104, 571−575. (62) Petrykin, V.; Kakihana, M.; Yoshioka, K.; Sasaki, S.; Ueda, Y.; Tomita, K.; Nakamura, Y.; Shiro, M.; Kudo, A. Synthesis and Structure of New Water-Soluble and Stable Tantalum Compound: Ammonium Tetralactatodiperoxo-μ-oxo-ditantalate(V). Inorg. Chem. 2006, 45, 9251−9256. (63) Kim, Y.-I.; Woodward, P. M.; Baba-Kishi, K. Z.; Tai, C. W. Characterization of the Structural, Optical, and Dielectric Properties of Oxynitride Perovskites AMO2N (A = Ba, Sr, Ca; M = Ta, Nb). Chem. Mater. 2004, 16, 1267−1276. (64) Kim, Y.-I.; Si, W.; Woodward, P. M.; Sutter, E.; Park, S.; Vogt, T. Epitaxial Thin-Film Deposition and Dielectric Properties of the Perovskite Oxynitride BaTaO2N. Chem. Mater. 2007, 19, 618−623. (65) Wang, C.; Hisatomi, T.; Minegishi, T.; Wang, Q.; Zhong, M.; Katayama, M.; Kubota, J.; Domen, K. Synthesis of Nanostructured BaTaO2N Thin Films as Photoanodes for Solar Water Splitting. J. Phys. Chem. C 2016, 120, 15758−15764. (66) Ueda, K.; Minegishi, T.; Clune, J.; Nakabayashi, M.; Hisatomi, T.; Nishiyama, H.; Katayama, M.; Shibata, N.; Kubota, J.; Yamada, T.; Domen, K. Photoelectrochemical Oxidation of Water using BaTaO2N Photoanodes Prepared by Particle Transfer Method. J. Am. Chem. Soc. 2015, 137, 2227−2230. (67) Hojamberdiev, M.; Yubuta, K.; Vequizo, J. J. M.; Yamakata, A.; Oishi, S.; Domen, K.; Teshima, K. NH3-Assisted Flux Growth of Cubelike BaTaO2N Submicron Crystals in a Completely Ionized Nonaqueous High-Temperature Solution and their Water Splitting Activity. Cryst. Growth Des. 2015, 15, 4663−4671. (68) Maeda, K.; Lu, D.; Domen, K. Oxidation of Water Under VisibleLight Irradiation Over Modified BaTaO2N Photocatalysts Promoted by Tungsten Species. Angew. Chem., Int. Ed. 2013, 52, 6488−6491. (69) Higashi, M.; Abe, R.; Teramura, K.; Takata, T.; Ohtani, B.; Domen, K. Two-Step Water Splitting into H2 and O2 Under Visible Light by ATaO2N (A = Ca, Sr, Ba) and WO3 with IO3−/I− Shuttle Redox Mediator. Chem. Phys. Lett. 2008, 452, 120−123. (70) Yamakata, A.; Vequizo, J. J. M.; Kawaguchi, M. Behavior and Energy State of Photogenerated Charge Carriers in Single-Crystalline and Polycrystalline Powder SrTiO3 Studied by Time-Resolved Absorption Spectroscopy in the Visible to Mid-Infrared Region. J. Phys. Chem. C 2015, 119, 1880−1885. (71) Yamakata, A.; Kawaguchi, M.; Nishimura, N.; Minegishi, T.; Kubota, J.; Domen, K. Behavior and Energy States of Photogenerated Charge Carriers on Pt- or CoOx-Loaded LaTiO2N Photocatalysts: Time-Resolved Visible to Mid-Infrared Absorption Study. J. Phys. Chem. C 2014, 118, 23897−23906. (72) Kawashima, K.; Hojamberdiev, M.; Wagata, H.; Yubuta, K.; Vequizo, J. J. M.; Yamakata, A.; Oishi, S.; Domen, K.; Teshima, K. NH3Assisted Flux-Mediated Direct Growth of LaTiO2N Crystallites for Visible-Light-Induced Water Splitting. J. Phys. Chem. C 2015, 119, 15896−15904. (73) Maegli, A. E.; Otal, E. H.; Hisatomi, T.; Yoon, S.; Leroy, C. M.; Schäuble, N.; Lu, Y.; Grätzel, M.; Weidenkaff, A. Perovskite-type LaTiO2N Oxynitrides for Solar Water Splitting: Influence of the Synthesis Conditions. Energy Procedia 2012, 22, 61−66. (74) Rugen, E. E.; Koczkur, K. M.; Skrabalak, S. E. Facile Synthesis of Porous La-Ti-O and LaTiO2N. Dalton Trans. 2017, 46, 10727−10733. (75) Sato, J.; Saito, N.; Yamada, Y.; Maeda, K.; Takata, T.; Kondo, J. N.; Hara, M.; Kobayashi, H.; Domen, K.; Inoue, Y. RuO2-loaded β-Ge3N4 as a Non-Oxide Photocatalyst for Overall Water Splitting. J. Am. Chem. Soc. 2005, 127, 4150−4151. 1343
DOI: 10.1021/acsenergylett.8b00518 ACS Energy Lett. 2018, 3, 1331−1344
ACS Energy Letters
Perspective
(76) Maeda, K.; Teramura, K.; Saito, N.; Inoue, Y.; Domen, K. Photocatalytic Overall Water Splitting on Gallium Nitride Powder. Bull. Chem. Soc. Jpn. 2007, 80, 1004−1010. (77) Arai, N.; Saito, N.; Nishiyama, H.; Domen, K.; Kobayashi, H.; Sato, K.; Inoue, Y. Effects of Divalent Metal ion (Mg2+, Zn2+ and Be2+) Doping on Photocatalytic Activity of Ruthenium Oxide-loaded Gallium Nitride for Water Splitting. Catal. Today 2007, 129, 407−413. (78) Arai, N.; Saito, N.; Nishiyama, H.; Inoue, Y.; Domen, K.; Sato, K. Overall Water Splitting by RuO2-Dispersed Divalent-Ion-Doped GaN Photocatalysts with d10 Electronic Configuration. Chem. Lett. 2006, 35, 796−797. (79) Maeda, K.; Domen, K. Solid Solution of GaN and ZnO as a Stable Photocatalyst for Overall Water Splitting Under Visible Light. Chem. Mater. 2010, 22, 612−623. (80) Maeda, K.; Teramura, K.; Lu, D.; Takata, T.; Saito, N.; Inoue, Y.; Domen, K. Photocatalyst Releasing Hydrogen from Water. Nature 2006, 440, 295. (81) Maeda, K.; Takata, T.; Hara, M.; Saito, N.; Inoue, Y.; Kobayashi, H.; Domen, K. GaN:ZnO Solid Solution as a Photocatalyst for VisibleLight-Driven Overall Water Splitting. J. Am. Chem. Soc. 2005, 127, 8286−8287. (82) Ohno, T.; Bai, L.; Hisatomi, T.; Maeda, K.; Domen, K. Photocatalytic Water Splitting using Modified GaN:ZnO Solid Solution Under Visible Light: Long-Time Operation and Regeneration of Activity. J. Am. Chem. Soc. 2012, 134, 8254−8259. (83) Wang, J.; Huang, B.; Wang, Z.; Wang, P.; Cheng, H.; Zheng, Z.; Qin, X.; Zhang, X.; Dai, Y.; Whangbo, M.-H. Facile Synthesis of Zn-rich (GaN)1−x(ZnO)x Solid Solutions Using Layered Double Hydroxides as Precursors. J. Mater. Chem. 2011, 21, 4562−4567. (84) Phivilay, S. P.; Roberts, C. A.; Puretzky, A. A.; Domen, K.; Wachs, I. E. Fundamental Bulk/Surface Structure−Photoactivity Relationships of Supported (Rh2‑yCryO3)/GaN Photocatalysts. J. Phys. Chem. Lett. 2013, 4, 3719−3724. (85) Lee, K.; Tienes, B. M.; Wilker, M. B.; Schnitzenbaumer, K. J.; Dukovic, G. (Ga1‑xZnx)(N1‑xOx) Nanocrystals: Visible Absorbers with Tunable Composition and Absorption Spectra. Nano Lett. 2012, 12, 3268−3272. (86) Lee, K.; Lu, Y.-G.; Chuang, C.-H.; Ciston, J.; Dukovic, G. Synthesis and Characterization of (Ga1‑xZnx)(N1‑xOx) Nanocrystals with a Wide Range of Compositions. J. Mater. Chem. A 2016, 4, 2927− 2935. (87) Chen, D. P.; Skrabalak, S. E. Synthesis of (Ga1‑xZnx)(N1‑xOx) with Enhanced Visible-Light Absorption and Reduced Defects by Suppressing Zn Volatilization. Inorg. Chem. 2016, 55, 3822−3828. (88) Zhou, J. K.; Lv, L.; Yu, J.; Li, H. L.; Guo, P.-Z.; Sun, H.; Zhao, X. S. Synthesis of Self-Organized Polycrystalline F-doped TiO2 Hollow Microspheres and their Photocatalytic Activity under Visible Light. J. Phys. Chem. C 2008, 112, 5316−5321. (89) Clark, L.; Oró-Solé, J.; Knight, K. S.; Fuertes, A.; Attfield, J. P. Thermally Robust Anion-Chain Order in Oxynitride Perovskites. Chem. Mater. 2013, 25, 5004−5011. (90) Chen, D. P.; Neuefeind, J. C.; Koczkur, K. M.; Bish, D. L.; Skrabalak, S. E. Role of Short-Range Chemical Ordering in (GaN)1‑x(ZnO)x for Photodriven Oxygen Evolution. Chem. Mater. 2017, 29, 6525−6535. (91) James, A.; Esfahani, M.; Woerner, W.; Sinclair, A.; Ehm, L.; Oganov, A.; Parise, J. Theoretical and Experimental Investigations into Novel Oxynitride Discovery in the GaN-TiO2 System at High Pressure. Crystals 2018, 8, 15. (92) Brancho, J. J.; Bartlett, B. M. Challenges in Co-Alloyed Titanium Oxynitrides, a Promising Class of Photochemically Active Materials. Chem. Mater. 2015, 27, 7207−7217.
1344
DOI: 10.1021/acsenergylett.8b00518 ACS Energy Lett. 2018, 3, 1331−1344
