Bismuth Vanadate Photocatalyst for Solar Energy Thesis

Introduction

The rapidly growing human population has placed tremendous stress on the planet for energy and raw materials. At present, the world relies greatly on non-renewable fossil fuels which account for approximately 80% of the global energy needs to maintain economic growth. Moreover, these limited fossil fuels are being burnt on an unprecedented scale, releasing carbon dioxide into the environment exceeding 30 billion tons annually. Carbon dioxide, one of the so-called “greenhouse gases,” is believed to be the major contributor to global warming and the extraordinary climate change phenomena. Regardless of climate change, the argument of sustainability for the energy product and generation is irrefutable and perhaps a more unifying argument for change. Future generations will eventually be left with a world void of easily accessible fossil fuels, and this will require strategies to find alternatives.

As the global population and affluence increase, the use of various materials has also increased in volume, diversity, and distance transported. For the past 50 or more years, since the start of industrialization, society has been increasingly reliant on the products of the petrochemical industry to supply man-made fibers for clothing, plastic artifacts, cosmetics, and paints. Hence, the increasing global demand for energy, depletion of fossil fuel reserves, and environmental concerns over rising carbon dioxide (CO₂) emissions warrant urgent actions towards the development of renewable sources of chemicals and fuels as well as the mitigation of environmental problems.

Addressing the power generation using renewable or green energy sources would lead to a dramatic change in the global consumption of fossil fuels. Renewable or green energy sources are resources that have an inexhaustible energy source with low or net-zero CO₂ production intensity. Among various renewable energy sources, solar energy is believed to be the most promising alternative as it is one of the most abundant and sustainable natural energy resources. Every hour, the sun emits about 4.3×10²⁰ joules of energy to the Earth’s surface, which is more than what the entire world consumes annually (4.1×10²⁰ joules) from fossil fuels. Visible light, ranging from 380 nm to 700 nm is one of the strongest output range of the Sun’s total irradiance spectrum that reaches the Earth’s surface. The solar spectrum consists of 50% and 45% of the visible light and infrared respectively, while ultraviolet (UV) light only occupies only 5% of the solar spectrum.

Photosynthetic organisms participate in complex biological processes to accumulate and utilize solar energy while involving only natural elements that can be found in different ecosystems to store this energy in carbon compounds. A fundamental feature of the light-harvesting mechanism among green plants is the integration of two visible light absorbers in Photosystems I and II to create a “Z-scheme.” The two chromophores, chlorophylls P680 and P700, are organized so that electrons flow from P680 to P700 with high fidelity by interfacing with both thermodynamically and kinetically favorable intermediates. This process essentially creates a “molecular photodiode” for the effective light absorption and charge separation, allowing P680 to build up the four holes for oxidizing water, while P700 acquires the two electrons to reduce NADP⁺. As a result, it is possible to state that a unique self-sustaining natural system has been created to address people’s demands.

Visible Light Active Photocatalysts

In recent years, pristine visible light-active semiconductors are being actively sought after for the better utilization of the full solar spectrum. It is important to note that different types of visible light-active semiconductors that are characterized by certain activities under sunlight irradiation represent two specific groups. They are metal oxides and metal chalcogenides, such as PbS,(cite) WS₂,(cite) PbSe,(cite) FeS₂,(cite) CdS,(cite) CdSe,(cite) and MoS₂.(cite) They possess relatively narrow bandgaps which can efficiently absorb a broad range of solar radiation, but these types often demonstrate the limited stability because of photo-corrosive phenomena that affect photoactivity in these cases. (cite)

To focus on the two PC applications, researchers mainly study metal oxides because of their chemical and photo-stability, the ease of fabrication, and a comparably low cost. Still, in most cases, binary metal oxides accumulate significant bandgap energies related to positive valence bands (VB), usually including O 2p orbitals that are situated at +3.0 V in comparison to a normal hydrogen electrode (NHE) restricting their ability to absorb light. Bandgap energies of 3.2–3.3 eV for ZnO, 3.0–3.2 eV for TiO₂, and 2.7 eV for WO₃ became observed in this case. To use solar energy efficiently, the design of metal oxides with smaller bandgaps is required. Moreover, to provide a sufficient photo-response, metal cations of metal oxides semiconductors include empty d orbitals, which is known as d⁰ electronic configuration, and filled d orbitals, which is known as d¹⁰ electron configuration, therefore, decreases in these bands are thermodynamically undesirable. To eliminate the band gaps of metal oxides, it is necessary to use a new VB formed by orbitals of the elements that are not related to O 2p. Thus, the focus is on ternary metal oxides that have VB related to atomic orbitals of more than one element. Orbitals of Ag 4d in Ag⁺, Bi 6s in Bi³⁺, Sn 5s in Sn²⁺, and Pb 6s in Pb²⁺ influence VB formation by O 2p orbitals in metal oxides.⁵ Therefore, it is possible to note that Bi, Ag, Pb, and Sn (BiVO₄,6 Bi₂WO₆,7 AgNbO₃,8 PbCrO₄,9 and SnNb₂O₆ 10) in ternary metal oxides have somewhat narrower band gaps with the promising ability to absorb visible light.

Introduction to Bismuth Vanadate (BiVO₄) Photocatalyst

In recent years, BiVO₄-based materials have emerged as an alternative to binary oxides due to their optimal electronic and optical properties that make them a suitable candidate for efficient visible-light-driven photocatalysis. In 1963, BiVO₄ was first synthesized by Roth and co-workers with just heating vanadium(V) oxide and the bismuth oxide mixture at high temperatures. Due to their non-toxicity in nature, BiVO₄-based compounds became actively studied because of the potential to be used as substitutes for the cadmium-, lead-, and chromate-based pigments. These components are generally utilized in the coating and plastics industry (for example, BASF). During the past four decades, researchers demonstrate the interest in studying BiVO₄ concerning examining its technological properties, including acousto-optical, ferroelasticity, and ionic conductivity, as they significantly depend on the crystal structures of BiVO₄.

Crystal Structures and Phase Transitions

There are three main crystal structures commonly obtained with synthetic BiVO₄. They are a scheelite structure with a tetragonal phase, a scheelite structure with a monoclinic phase, and a zircon structure with a tetragonal phase. Vanadium (V) ions in these structures are coordinated with four oxygen (O) atoms in a specific tetrahedral site. Bismuth (Bi) ions, in their turn, are coordinated with eight O atoms. There is a difference between scheelite and zircon structures, and in a scheelite structure, each Bi ion is connected to eight vanadates (VO₄) tetrahedral units. In a zircon structure, there are only six VO₄ related to one Bi ion.¹⁵ For BiVO₄ with a specific scheelite structure, a monoclinic scheelite (m-s) BiVO₄ is distorted with two different V–O bond lengths (1.77 Å and 1.69 Å), while the V–O bonds in tetragonal scheelite (t-s) BiVO₄ are all of an equal length (1.72 Å).¹⁶

The (m-s) BiVO₄ structure is stable at a high temperature, which can be obtained from the irreversible phase transformation from a tetragonal zircon (t-z) structure at 400 to 500°C.¹⁷ Irreversible phase transition from (t-z) to (m-s) BiVO₄ structures was also achieved through mechanical crushing.¹² The (m-s) BiVO₄ can be obtained via the aging of a crystallization process of (t-s) BiVO4 at room temperature.^17-18 Thus, the formation of (t-s) BiVO₄ under these conditions is kinetically favorable, but (m-s) BiVO₄ can be viewed as a more thermodynamically stable form. Also, it is important to note that the heat treatment can catalyze phase transitions because the transformation of scheelite BiVO₄ with monoclinic and tetragonal phases is observed at 255°C.

Optical Properties and Electronic Structures

Among the three polymorphs of BiVO₄, (m-s) BiVO₄, in particular, became actively researched after the publication of the paper by Kudo and co-workers on the visible light-triggered water oxidation over BiVO₄ of three polymorphs in the presence of Ag⁺ ions as electron scavengers.¹⁹ They observed higher O₂ evolution activity (water oxidation) from aqueous AgNO₃ solution for (m-s) BiVO₄ mainly due to the wider bandgap of (t-z) BiVO4.²⁰ In the scheelite BiVO₄, it is possible to find out a hybridized band structure with Bi 6s and O 2p orbitals. Therefore, concerning a hybridized Bi 6s–O 2p linked to VB and the O 2p VB, the electron transition to the V 3d CB can result in forming less bandgap energy of scheelite BiVO₄ (2.4 eV) while comparing it to the energy associated with zircon BiVO₄ (2.9 eV) with V–O transition.²⁰

Computational methods using density functional theory (DFT) calculations to study the BiVO₄ crystal structure were reported in 2009, suggesting that the hybridized VB may exist.²¹ However, only in 2011, Payne and co-workers conducted experiments with the help of X-ray photoemission spectroscopy (XPS), X-ray absorption spectroscopy (XAS), and X-ray emission spectroscopy (XES)²² to find evidence to support the hybridization of Bi 6s and O 2p orbitals at the valence band maximum of monoclinic BiVO₄ that were previously explained by Kudo and co-workers.

Contrary to the report by Walsh and co-workers,²¹ BiVO4 is viewed as an indirect bandgap semiconductor which was supported by the recent study by Cooper and co-workers, where they provided a rigorous analysis of the band structure of the BiVO₄. The researchers referred to DFT calculations and concluded concerning experimental evidence found via the combination of X-ray spectroscopies (XAS, XES, and resonant inelastic X-ray scattering (RIXS)) and XPS.²³ Furthermore, the bandgap energy was identified to be 2.5 eV (absorption edge: ca. 500 nm), and the Fermi level at the surface was 2.0 eV above the valence band maximum. These measures showed that the BiVO₄ is an n-type semiconductor. The CB edge was 0.1 V_RHE, and the VB edge was 2.6 V_RHE, as a result of conducted measurements and Mott-Schottky analysis.²⁴ It is possible to state that DFT calculations completed by Walsh and co-workers²¹ and Zhao and co-workers²⁵ demonstrate that oxide materials in most cases are larger than effective masses of both holes and electrons in (m-s) BiVO₄. The drift velocity of holes and electrons is proportional to the effective mass in reverse.²⁶ Thus, smaller effective masses are associated with the easier charge separation and migration of electrons to reach the surface reaction sites. Therefore, monoclinic scheelite BiVO₄ is viewed as an effective photocatalyst.

Limitations of BiVO₄

Based on the near to optimal bandgap of 2.5 eV, and VB edge located at ca. 2.6 eV, as well as the relatively small effective carrier masses mentioned in the previous section, the actual solar-to-hydrogen (STH) conversion efficiency obtained by BiVO₄ is far below its theoretical value of 9.3% with a maximum photocurrent of 7.6 mA.cm⁻² under standard AM 1.5 solar spectrum.²⁷ The limitations are as follows:

1. Poor electron transport.²⁸ One should note that the poor electron transport properties of BiVO₄ can be explained concerning the disconnected VO₄ tetrahedral units observed in the material’s crystal structure. This suggests that the photoexcited electrons in the V 3d conduction band can be bounded between the VO₄ tetrahedra.²⁷
2. Slow water oxidation kinetics (it only applies to water splitting).
3. Low carrier mobility that can significantly add to the poor charge separation of BiVO₄ also leads to approximately 60%–80% of the electron-hole pairs’ recombination before they reach the interfaces.^{2, 23, 29-30}
4. Weak surface absorption property drastically limits the efficiency of the reactions at the interfaces.³¹

The main reason for the poor charge separation of BiVO₄ is its intrinsic low carrier mobility that is only about 0.044 cm² V^-1s^-1 (measured under 1 sun illumination conditions).³² From this perspective, the comparably long carrier lifetime (40 ns) and the charge diffusion length (70 nm) seem to balance this poor carrier mobility. It is important to note that charge carriers move by diffusion if an external field is absent. Diffusion, in this case, depends on the carrier lifetime and specific mobility.

Again, apart from its low carrier mobility, BiVO₄ seems to be superior in terms of its carrier lifetime and diffusion length compared to the dynamic properties of other typical visible light-absorbing semiconductors, which include WO₃, Fe₂O₃, and Cu₂O. The carrier lifetime of BiVO₄ is 1 to 3 orders of magnitude longer than the other metal oxides, and its diffusion length is just the second to WO₃.³²

In recent years, numerous attempts based on different strategies have been developed to improve the PC performances of BiVO₄. These strategies include doping, loading co-catalysts, surface modification, and other alternative strategies include crystal facet engineering, annealing treatment, and nanoscale. However, before moving on to the strategies to enhance the performance of BiVO4, it is necessary to understand the basic principle of photocatalysis, as well as the requirement for photocatalysis to occur.

Basic Principle of Photocatalysis

The mechanism of photocatalysis should be described in detail. When visible photons are absorbed by the light-harvesting agent, charge separation occurs while generating holes and electrons at the valence band and conduction band, respectively. Subsequently, electrons at the conduction band are transported to the reduction co-catalyst to drive reduction reactions while the oxidation co-catalyst performs oxidation by drawing electrons back into the holes of the valence bands. This process repeats, driving the redox reactions.

When a semiconductor catalyst is illumined with specific photons (their energy is greater than the semiconductor’s bandgap (E_g)), an electron (e^–) can be promoted from the VB to the CB, leaving a hole (h+) in the VB. The excited state conduction band electrons and valence band holes are kinetically, as well as thermodynamically, favored to recombine and dissipate heat or emitted light. However, when the electrons and holes can migrate to the surface of the semiconductor without recombination, they can be involved in electrochemical processes with species adsorbed at the semiconductor surface. Photogenerated electrons act as reducing agents, and holes act as oxidizing agents. The redox ability of the electron/hole pairs can be used for PC water/air remediation and PC hydrogen production.

Strategies to Enhance BiVO₄ PC Performance

Surface Modification

Surface modification is required to improve water oxidation kinetics with the use of various rare-earth transition metal catalysts, such as IrO₂, RuO₂, and RhO₂^34-36. About the cost, co-catalysts containing earth-abundant elements are favorable. Recently, the potential of Co–Pi,^{28, 37} Co₃O₄,³⁸ FeOOH,³⁹, and NiOOH³⁹ as the earth-abundant water oxidation co-catalysts for BiVO₄ has been demonstrated. Among these co-catalysts, Co–Pi is most commonly studied to improve the water oxidation kinetics of BiVO₄ in PEC and PS systems. The performance enhancement (i.e., a negative shift of the onset potential) resulted primarily from a hole transfer process from BiVO₄ to CoPi to suppress surface recombination.

Doping

One approach that has been widely proven to effectively improve the electron transport in BiVO₄ involves doping of the material of interest. Substitutional doping involves replacing one of the host original atoms, i.e., effective doping by atom incorporation. In its turn, interstitial doping is based on attracting a foreign atom to interstitial space, as well as to the lattice of a specific host matrix, when the existing atom remains non-replaced. Still, in comparison to interstitial doping, substitutional doping has more strengths in terms of its electrical and optical properties. Features similar to qualities of substituted atoms or atoms with different valence states can also be possessed by impurities, leading to increases in the number of electrons or holes to stimulate the charge carrier mobility and affect the minority carrier diffusion length. To contribute to the increased photo-activity, transition metal doping stimulates enhanced trapping of electrons that results in the electron-hole recombination observed during irradiation, when the conductivity rises because of doping BiVO₄ with Mo and/or W since Mo⁶⁺_/W⁶⁺ replaces the V⁵⁺ in VO₄ tetrahedron positions.^{31, 34} The mid-gap deep-level states only partially contribute to the visible-light PC activity due to low charge carrier mobility associated with these states that perform as charge recombination centers even though they can broaden the spectral range of light absorption related to bandgap semiconductors.⁴⁰

Heterojunction Photocatalysts

The basic principles of semiconductor photocatalysis explain the recombination between holes and electrons as unfavorable for a semiconductor photocatalyst. To increase PC efficiency, electron-hole pairs are expected to be separated. Moreover, to constrain the recombination, charges should be quickly transferred to the surface/interface. The PC performance can be improved in this case concerning forming a semiconductor heterojunction as a result of coupling with any secondary substance. Thus, four typical categories of heterojunction photocatalysts can be determined for this paper:

1. the semiconductor–metal (S–M) heterojunction;
2. the semiconductor– semiconductor (S–S) heterojunction;
3. the semiconductor–carbon group (S–C) heterojunction;
4. the multicomponent heterojunction.

The SC-1 conduction band can be described as being more negative than the band of SC-2 in Type I and Type II. However, the valence band for SC-1 is more positive in comparison to the band of SC-2 in Type I, and it is less positive in Type II. The conduction band and the valence band of SC-2 can be described as lower in comparison to SC-1 in Type III. Thus, with the focus on a semi-conductor heterojunction, electrons relocate to the less negative conduction band, and holes relocate to the less positive valence band. In Type I, electrons and holes then amass in the semi-conductor with a smaller SC-2 bandgap, they do not recombine, and there is no improvement in the photocatalytic performance. In Type II, electrons and holes are separated into two semiconductors, and the overall separation of charge carriers is observed. In Type III, electrons and holes are located in one semiconductor, and they cannot be relocated to another, and there are no heterojunction forms between them.

Referring to Type II, it is important to note that an internal electric field is created with the focus on the direction from SC-1 to SC-2 when the SC-1 Fermi level is higher in comparison to SC-2. In this case, photogenerated electrons relocate to the interior, and holes relocate to the interface, but the flow of charge carriers for SC-2 can be described as opposite to SC-1. As a result, electrons related to SC-2 and holes related to SC-1 remain in the interface, being affected by recombination.⁴¹ On the contrary, SC-1-related electrons and SC-2-related holes are used for catalytic reactions, and they do not drop their oxidation or reduction abilities in this context. As a result, photoexcitation and charge recombination in this form of heterojunction can demonstrate a Z-shape route (the Z-scheme system) in terms of the energy exchange.^42-48

The direction of the internal electric field changes, and it is from SC-2 to SC-1 if the SC-2 Fermi level is higher in comparison to the level of SC-1 in Type II. In this case, SC-1 is described as a p-type semi-conductor, and SC-2 can be described concerning an n-type. As a result, SC-1-related electrons move to the interface, but they can also migrate to the SC-2 conduction band when the holes of SC-2 can relocate to the SC-1 valence band. PC performance becomes improved because of this particular mechanism. From this perspective, certain advantages of heterojunction PC systems that can be identified easily are the following ones:

1. the inhibition of charge recombinations in bandgap SCs;
2. the improvement of the charge separation;
3. the extension of the lifetime for energy excited electrons;
4. the utilization of much sun radiation.

It is found that a specific artificial Z-scheme photocatalyst, which consists of such semiconductor photocatalysts as the oxidation and reduction photocatalysts, often demonstrates the greater photocatalytic performance. A low VB position and a strong oxidation ability are typical of oxidation photocatalysts. In their turn, reduction photocatalysts represent a high CB position, and their reduction ability can also be described as strong. Many advantages are associated with the performance of a Z-scheme photocatalyst that can have a high photocatalytic activity: (1) the preservation of high reduction and oxidation abilities; (2) the separation of specific oxidative and reductive active sites in space; (3) the high separation of photogenerated charge carriers; (4) a broad range of photocatalysts for different reactions, demonstrating that two photocatalysts plus narrower bandgaps can show the higher redox ability in reactions; and (5) the extended light-gathering spectrum.

It is important to note that Z-scheme photocatalysts are usually divided into three groups concerning the type of introduction of a charge carrier mediator and its type. These groups are known as traditional Z-scheme photocatalysts, all-solid-state Z-scheme photocatalysts, and direct Z-scheme photocatalysts. In traditional Z-scheme photocatalysts, a reversible redox ion pair (Fe3+/Fe2+ and IO3/I) is used to transport charge carriers. In all-solid-state Z-scheme photocatalysts, electron conductors (Au, Ag nanoparticles (NPs)) are utilized to enhance transporting charge carriers. Furthermore, indirect Z-scheme photocatalysts, two semiconductors contact directly, and charge carrier transfer mediators are not involved in the process. These groups of Z-scheme photocatalysts should be viewed as dissimilar in terms of their working mechanisms, unique properties, and associated synthetic processes.

One should note that a direct Z-scheme photocatalyst is most remarkable in this context because two semiconductors contact closely without any transfer mediator contrary to the traditional variant of the Z-scheme system. As a result, it is important to state that the absence of any redox mediators causes the restraining of backward reactions, as well as the significant shielding effect. It is also necessary to pay attention to the fact that direct Z-scheme photocatalysts are unsusceptible to corrosion in this case. Thus, photogenerated electrons in an Ag-based photocatalyst are accumulated while combining with photogenerated holes, and this aspect reduces any possibilities for Ag-based compounds to be photocorroded into metallic Ag when an Ag-based photocatalyst performs like an oxidation component.

It is also important to state that the charge transfer mode in a specific photocatalytic reaction needs to be studied when two semiconductors with a unique staggered band structure contact directly. In this case, a difference in working functions between two semiconductor photocatalysts can be viewed and discussed as a pre-requisite for stimulating specific charge redistribution. This process also provokes the formation of an internal electric field to influence the photogenerated charge carrier separation and the strengthening of a transfer process.

Photocatalytic systems can be represented in two types where, according to the first type, PC I has the higher CB and VB positions and the smaller work function with the focus on the higher Fermi level than PC II. Another type of the photocatalytic system is when the typical p–n junction is observed. One should note that when PC I and PC II interact directly, free electrons related to PC I move to PC II to stabilize and balance their Fermi levels. In this context, at the interface, the PC I side is usually positively charged in contrast to the negatively charged PC II side. Furthermore, both photocatalysts began to generate specific electron-hole pairs while following the excitation by light, and the Z-scheme charge transfer mode can be viewed as more effective in this context. As a result, the interactions between the photogenerated electrons related to the CB of PC II and photogenerated holes that are observed in the VB of PC I become possible because of the presence of an internal electric field.

An extra potential barrier, an internal electric field, and Coulomb repulsion usually delay the movement of photogenerated electrons from PC I CB to PC II CB. In this case, the electrons related to the CB of PC I and holes related to the VB of PC II become separated and involved in different reductive and oxidative photocatalytic reactions. The recombination between the photogenerated electrons and holes can be restrained in this context by the impact of an induced electric field.

While focusing on p–n junction as the second type of the system, it is important to note that free electrons related to the n-type semiconductor usually move to the p-type semiconductor under the condition of contacting between the p-type semiconductor and the n-type semiconductor. As a result, a specific built-in electric field is created about the contacts between semiconductors. Thus, it is possible to state that a p–n junction charge carrier transfer mode can be realized only in the context of the existed built-in electric field while comparing it with a unique direct Z-scheme charge carrier transfer mode. From this perspective, the photogenerated charge carrier transfer mode that is typically observed in p–n junction cannot be viewed as appropriate for the direct Z-scheme mode.