Photocatalytic Reduction of CO2 by Halide Perovskite: Recent Advances and Future Perspectives

Photocatalytic CO2 reduction to generate energy-riching fuels through solar energy provides an attractive route to alleviate the global energy crisis and environmental concerns. Searching for various photocatalysts with high catalytic...


INTRODUCTION
Tremendous natural environmental changes have made people around the world increase crisis awareness when we are facing unprecedented nature. Great attention has been plunged into the depletion of natural resources and the impact of the emission of greenhouse gases on the environment due to the burning of fossil fuels. 1-4 Solar-driven carbon dioxide (CO 2 ) conversion is an encouraging strategy to alleviate the negative effect of greenhouse gases, 5,6 which uses artificial photosynthesis and photocatalysis of solar energy for carbon feedbacks and is also a straightforward and flexible pathway of CO 2 valorization. 7,8 Since the landmark breakthrough in 1972, the pioneering work on ultraviolet (UV) driven photocatalytic production of hydrogen by TiO 2 9 has spurred enormous interest and motivated decades of scientific research and progress, which are of particular importance for energy transformation (e.g., reduction of CO 2 and water splitting), chemical transformations, and the decontamination of organic pollutants. 10 During recent years, the rapid increase in CO 2 concentration in the environment has drawn widespread concern. 11 Extensive efforts have been devoted into stabilizing and controlling the concentration of CO 2 in the environment, leading to CO 2 capture and renewable energy production. Particularly, the stateof-the-art photocatalytic reduction of CO 2 has been considered as one of the most promising ways. 12 From the perspective of long-term development, it is essential to explore the costefficient photocatalysts with outstanding properties and performances. An ideal photocatalytic material should possess wide-range and superior light absorption, effective charge separation, excellent stability as well as proper redox ability. 13 Upon illumination, the photogenerated charge carriers are produced in the conduction band (CB) and valence band (VB) in semiconductors, respectively. When the energy structure of these semiconductors straddles the redox potentials, their photogenerated charge carriers would involve the surface Up to now, there have been some review works of halide perovskites as photocatalyst focusing on the stability issues, or modification strategies through structural engineering and interfacial modulations. [53][54][55][56][57] However, a timely and comprehensive review is still in demand concerning the photoreduction mechanism, reductants and products, the structure and photocatalytic performances, and the exploration efforts of halide perovskites and their composites for further improvements. In this review, we start with the basic mechanism of the photoreduction of CO 2 , and the reductants and products. After that, the fundamental photoelectrical properties of halide perovskites are described collectively, including their crystal structure, defect tolerance toxicity, luminescence properties, and photostability issues.
Insights into the molecular calculations of halide based perovskites concerning their unique electronic structure and physio-chemical properties are also included. Then, we summarize the existing exploration efforts of halide perovskites and their composites in tuning the photocatalytic reduction of CO 2 selectivity as well as increasing the active sites. Finally, we present the outlook for upcoming directions and the great potentials of halide perovskites toward photocatalytic CO 2 reduction.

PHOTOCATALYTIC MECHANISM FOR CO 2 REDUCTION
In the process of photocatalysis, the abundantly available solar energy is transformed into electrical/chemical/thermal energies through semiconductor materials. 13,[58][59][60][61] In general, the photocatalysis process for CO 2 reduction mainly consists of three critical synergistic steps, as illustrated in to CO 2 molecule. In general, larger surface area of the photocatalyst can lead to higher rate of CO 2 adsorption thanks to more active sites. [62][63][64] The photocatalytic activity is generally associated with two main conditions: (i) The bandgap energy (E g ) of the photocatalyst should be lower than the energy of the incident light (hv; the absorption energy). (ii) The reduction potential of the reacting species should be positioned between the CB and VB values of the photocatalyst material. The former condition specifies a narrower bandgap that can help the effective utilization of incident photon. Conversely, the later condition reveals that a higher value of CB potential and a lower value of VB potential are thermodynamically advantageous toward the reduction as well as oxidation reactions of the reacting species, respectively. Therefore, it is inevitable to seek the balance point owing to the existence of the noticeable inconsistency as discussed in the above conditions. However, this is quite a challenging task to obtain both broad light absorption and robust redox capability simultaneously for pure component semiconductor materials. Moreover, the photogenerated electrons in CB can easily recombine with holes, or be trapped in the defect state, or easily come back to the VB for a pure-component material, which would decrease the efficiency of solar energy utilization. 65 Hence, designing a suitable heterogeneous photocatalyst is an effective approach to overcome the above-mentioned issues. Generally, there are four vital rules played by cocatalysts: (i) Enhancing the separation and transfer of charge, (ii) Boosting the performance, and the selectivity photoreduction of CO 2 reduction. (iii) Improving the durability of photocatalysts in the peculiar environment (e.g. high humidity and high temperature). (iv) Overwhelming the adverse reactions (e.g., H 2 reduction). Photocatalytic reduction of CO 2 with semiconductor-based cocatalyst is affected by many aspects, for example, photocatalyst loading, particle size, structure, composition, dispersion, crystal facets, alloy phase, morphology, and valence states. In a photocatalyst system, the maximum photocatalytic performance is accomplished at an optimum loading of   Photocatalytic CO 2 reduction reaction is a high energy-driven process, because of the carbon-based compound that has a thermodynamically stable and high bond energy of C=O (750 kJ/mol). 72,73 Therefore, the photocatalytic reduction reaction of CO 2 also needs prior activation. The process of mass transfer of reactants, CO 2 adsorption and creation of active sites for CO 2 reduction can be improved by increasing surface area of a photocatalyst. A large number of nanostructured materials for photocatalyst were developed to achieve larger surface areas, including zero dimensional (0D) nanoparticles, 74 one dimensional (1D) nanowires/nanotubes, 75 two dimensional (2D) nanosheets, 76,77 and three dimensional (3D) hierarchical nanostructures. 78 Furthermore, various porous and hollow materials were extensively studied for CO 2 conversion because of their increased surface area and porosity.
The adsorption and activation of CO 2 can also be facilitated by tuning surface defects of photocatalysts, which can be quite helpful in improving the reactivity and photochemical properties of the photocatalytic processes. 79,80 Strongly negative reduction potential is required for the activation of CO 2 , which occurs on the surface of photocatalysts by chemisorption method to convert a linear molecule to bend carbonate anion radical (CO 2 • -) via mono-or bi-dentate coordination. Hence, this reactive carbonate anion is the critical

ABX 3 structure
Halide perovskites employed by photocatalytic CO 2 reduction facilitate band alignment with CB position and hinder the photogenerated charge carrier from undergoing recombination. For achieving the higher productivity and selectivity of halide perovskite nanocrystals (NCs) for photoreduction of CO 2 , fundamental understandings of the relationships between the structure (chemical composition as well as morphology) and CO 2 selectivity need to be determined. 89 In this section, the impact stemming of halide perovskites from the unit cell will be reviewed. The unit cell is the simplest and smallest volume of a material. The properties of materials derive from special characteristics of the unit cell, e.g., the bond angle, bond length, and symmetry. Moreover, the diffusion lengths and lifetimes of charge carriers in optoelectronic materials are more affected by crystal boundaries and crystal defects.
One of the main chemical formulas of halide perovskites can be symbolized as ABX 3 , where A and B represent two different kinds of cations and X represents halide anion (i.e., Cl, Br, and I). Cation B restricts at the median of the octahedron body composed of six halide anions ( Figure 4a). [90][91][92] Conventionally, A and B are cations with 12-and 6-fold coordination with X anions at the corner in ABX 3 halide perovskites (Figure 4a), respectively. [93][94][95] Recently, various high-performing perovskite materials were developed with A being formamidinium (FA), methylammonium (MA), or Cs; B being Pb, Bi, or Sn; and X being Cl, Br or I. [93][94][95][96][97][98][99][100][101][102][103][104][105][106] When A site is employed with higher organic cations, low-dimensional halide perovskite structures may form, where the inorganic network connectivity has degenerated to 2D sheets, 1D chains, or 0D clusters (Figure 4b). 107 Conversely, the crystal symmetry will reduce and the cubic structure is distorted. In bulk lead halide perovskites, three polymorphs are usually noticed: cubic, tetragonal, and orthorhombic phases in the order of decreasing symmetry ( Figure 4a). 92 At higher temperatures, the more stable phase is the cubic phase, and the temperature of phase transitions is clearly defined. For NCs, surface effects may regulate the relative stabilities of several polymorphs, which gains limited attention. All as-synthesized lead halide perovskite NCs crystallize into three-dimensional (3D) phases as follows: MAPbI 3 NC is tetragonal; FAPbBr 3 , FAPbI 3 , and MAPbBr 3 NCs are pseudocubic; and CsPbI 3 and CsPbBr 3 NCs are orthorhombic at room temperature. 92 One of the most significant features of halide perovskites is their higher tolerance for defects. The nature of halide perovskites is defect-tolerant generally owing to their electronic band structure, where VB maximum is mostly composed of anti-bonding orbital, [108][109][110] and the CB minimum becomes stabilized via the strongest spin-orbit coupling. Suchlike defect tolerance behavior involves the conservation of a pristine E g upon the formation of typical defects, due to their defect energy levels residing entirely in either VB or CB, rather than within the E g itself. The large hollow between octahedral (A-site) is preoccupied through one or a combination of three big cations (CH(NH 2 ) 2+ , CH 3 NH 3 + , or Cs + ), yielding generally the structure of ABX 3 .
The tolerance factor calculated by where R A , R B , and R X are the ionic radii of the corresponding ions, should be near to 1 to keep a higher symmetry cubical structure of the perovskite. Figure 4c presents the tolerance factors for the best widespread Pb or Sn halide perovskites. Due to the larger numbers of Pb or Sn atoms occupying the B sites of halide perovskites, the A site must be larger enough to placate the tolerance factor. 111

Luminescence properties
Halide perovskite NCs have excellent luminescence without advance electronic surface passivation. Protesescu et al. presented nearly ideal photoluminescence efficiency from colloidal CsPbBr 3 NCs. 32 The composition of lead halide perovskite NCs can be appropriately adjusted through cation or anion exchange, as shown in Figure 5a. 38 The photoluminescence spectra of lead halide perovskite NCs spans the whole visible light range and their peak positions are tunable through modifying the composition (Figure 5b), size, and shape. 92 Higher photoluminescence quantum yield with tunable emission, low cost, and simple synthesis of halide perovskite NCs makes them appealing in practical applications.
However, the toxicity of lead, sensitivity to atmosphere condition (humidity, oxygen,

Other structure
The lead halide perovskite-based NCs with 3D APbX 3 crystal structure and composition have become the focus of much research interest up to now. However, the inherent toxicity and overall reactivity of these halide perovskites have also motivated the scientist to investigate it in various research directions. Firstly, the structural instability and higher ionicity of LHP NCs could usually be taken as a positive aspect, as APbX 3 lattice could be easily rationalized into other phases. This has motivated the researcher to investigate NCs more extensively by exploring their composition and structural characteristics that are defined as "perovskite-related structures", for example Cs 4 PbX 6 and CsPb 2 X 5 (often called as zero-dimensional (0D) and two dimensional (2D) structures, respectively). Whereas, the  (Figure 6a). The PbX 6 4octahedra in A 4 PbX 6 structures are dissociated in all extents and the halide ions are no longer shared between them (Figure 6b). 114,115 Layered perovskites have recently been under a lot of scrutinization process. Similar to that of layered double hydroxides containing alternating Cs + and [Pb 2 X 5 ] − polyhedron layers, the CsPb 2 X 5 has emerged as a 2D version of lead halide perovskite materials with a tetragonal phase (Figure 6c). 116,117 Another type that is containing an alternating layer of corner-sharing [PbX 6 ] 4octahedra and bulky cations is known as the 2D perovskites A 2 PbX 4 phase (Figure 6d). 118 In APbX 3 NC systems, the lead toxicity and its bioaccumulation in the ecosystem are known as a key drawback, which in turn motivate the researcher to find out alternative materials with similar optoelectronic characteristics, such as Cs 2 SnI 6 NCs. [119][120][121] To date, there has been very little success. Cs 2 SnI 6 crystallizes in the face-centered cubic structure.
Four [SnI 6 ] 2octahedra at the corners and face centers, and eight Cs + cations at the tetragonal interstitials, make up the unit cell (Figure 6e). A Cs 2 SnI 6 structure is a perovskite derivative that is made by removing half of the Sn atoms at regular intervals from each center of the [SnI 6 ] octahedron. 120 Because of this, the structure is also known as a "vacancy ordered double perovskite." Two primary techniques are now being explored in the search for leadfree metal halide compounds: a "simple" substitution of Pb 2+ cations with other less toxic divalent metal ions from the same group IV, such as Sn or Ge; and a "complex" substitution of Pb 2+ cations with other less toxic divalent metal ions from the same group IV, 122 such as Sn or Ge or a substitution of one monovalent M + and one trivalent M 3+ cation for every two divalent Pb 2+ ions (i.e., 2Pb 2+ → B + + B 3+ ), resulting in quaternary A 2 B + B 3+ X 6 systems also known as "double perovskites" in Figure 6f. 123 Other transition or post-transition metals, like Fe 3+ and Bi 3+ , were used to examine the diversity of halide compounds associated to LHPs. [124][125][126][127] Cs 3 M 2 X 9 (M = Fe 3+ , Bi 3+ ) crystallizes in the hexagonal space group P6 3 /mmc. This is made up of isolated clusters, each of which is made up of two face-sharing octahedra and has the M 2 Br 9 3formula, with Cs + acting as a bridging ion between the clusters. Antimony-based halide compounds, on the other hand, form in a layered shape with each Sb 2 Br 9 3cluster sharing corners with three octahedra (Figure 6g)

Reaction medium
Solvent plays a critical role in photocatalytic reactions. Halide perovskites are unstable in a polar solvent, therefore extensive researches have been accomplished to find a suitable medium for photocatalytic reduction of CO 2 . Low-polar ethyl acetate was utilized as the solvent due to its higher solubility of CO 2 , which guarantees durable stability for CsPbBr 3 NCs. 42 Thereafter, various groups reported the boost in the selectivity of CO 2 conversion (99%) with the suppression production of H 2 , with the addition of a small amount of water (˂ 50 µl) in ethyl acetate medium. 45,100,101,113 However, employing acetonitrile/water (0.3 vol. %) mixture showed a photocatalytic reduction of CO 2 with a high conversion rate (149 µmol g -1 h -1 ) of CO 2 to CO, compared to the ethyl acetate/water (0.3 vol. %) mixture upon light irradiation by using CsPbBr 3 NCs combined with porous g-C 3 N 4 . Moreover, enhanced selectivity, as well as productivity of the photocatalytic reduction of CO 2 was obtained with a high content of diluted water (1.2 vol. %). 128 Owing to several polarities, dielectric constant, and CO 2 solubility, the selection of solvent can exert a higher effect on the reaction rate and selectivity. The pairing of a robust co-catalyst is the key feature for enhancing the performance of photocatalysis. Meanwhile, it can not only separate the photogenerated electrons by the creation of Schottky junction within co-catalyst and photocatalyst to hinder the charge recombination, but also lower the kinetic bottlenecks in the activation of CO 2 . 129,130 In the present case, the best nonaqueous solvent was ethyl acetate, which resulted in the reduction reaction of CO 2 with an electron yield rate of 2.74 μmol g -1 h -1 along with 95.2% of the selectivity. Furthermore, a photodeposition of Pt co-catalyst boost the electron yield rate to 5.66 μmol g -1 h -1 . 131 Numerous additional solvents such as toluene, benzene, etc., were also explored for photocatalytic reduction of CO 2 as photocatalysis medium in halide perovskite NCs. 101 Besides, owing to showed that FAPbBr 3 QD, as an alternative to halide perovskites, was a better reduction and capturing agent for CO 2 (Figure 7 a-c). 51 135 For catalysis application, active facets for recanting species or products as well as the reduction of electron-hole recombination were facilitated to the efficient adsorption/desorption process. The highest production rate of CH 4 and CO was 7.6 and 16.4 μmol g −1 , respectively, for the CsPbBr 3 cubic shaped NCs, which matched well with the previously reported literature. 43,99,135 Conversely, the maximum activity was found to be higher for the hexapod nanostructures. The optimum yields of CH 4

Other structures
Lead-based halide perovskites are developing as the best encouraging type of materials for new-generation solar conversion energy. [137][138][139] The metal-based halide perovskite (MHP) NCs is estimated to solve the problem affected via the usage of organic photosensitizers, and to form highly effective photocatalytic systems. Despite the considerable success, the key problems of stability and toxicity are yet to be fixed. The photostability of metal-based halide perovskites is usually low in the presence of molecular oxygen (O 2 ). 140,141 The soft nature of MHPs crystal lattice renders their surfaces prone to structural modification and degradation.
Previous research has indicated that a partial or complete replacement of A site organic cations with Cs + can result in many new MHPs compositions (all-inorganic chemical structure), which significantly promotes the photo-and humidity-stability. 142 Some studies have shown that the ionization energy can be used to predict the stability of perovskite. The  acid was hazardous to synthesize pure Cs 3 Sb 2 Br 9 NCs without the CsBr contamination as shown in Figure 8c. 48 Cs 3 Sb 2 Br 9 NCs showed much higher efficiency of photocatalytic reduction reaction of CO 2 , compared to Pb-based CsPbBr 3 NCs. For photocatalytic reduction of CO 2 , a bound state is observed for COOH* on the surface of Cs 3 Sb 2 Br 9 NCs in which a Br ion is partially replaced to allow Sb-C bond formation as shown in Figure 8d. 48    When developing the stronger reduction potential of CsPbBr 3 to produce a Z-scheme heterojunction photocatalyst, it is best to search for a suitable oxidative semiconductor.
Moreover, the oxidative semiconductor has the following properties: (i) strong interactions with CsPbBr 3 ; (ii) visible-light response; (iii) high catalytic activity for the oxidation reaction; and (iv) compatible band structure with CsPbBr 3 . Bismuth tungsten oxide (Bi 2 WO 6 ) is a unique oxide with visible-light response, 154 155,156 For oxidative reactions, Bi 2 WO 6 also has an oxygen-deficient surface that detects on abundant active sites, and it is preferred as an ideal photocatalyst for O 2 evolution. 157,158 Notably, the band structures among CsPbBr 3 and Bi 2 WO 6 were well-matched, and in principle, they can also be incorporated into a Z-scheme heterojunction. Wang et al. developed a 0D/2D heterojunction of CsPbBr 3 QDs/Bi 2 WO 6 nanosheet (CPB/BWO) photocatalyst for photocatalytic reduction of CO 2 . 159 The larger surface area of 2D Bi 2 WO 6 effectively decorated 0D CsPbBr 3 on its surface. The close contact between Bi 2 WO 6 and CsPbBr 3 permitted an excellent interface for the charge transfer and separation. 160,161 The advanced CPB/BWO heterojunction photocatalyst exhibited higher activities for photocatalytic reduction of CO 2 as compared to the Bi 2 WO 6 nanosheets and pristine CsPbBr 3 QDs (Figure 10 a-c), respectively. 159 The separated holes and electrons in CsPbBr 3 QDs and Bi 2 WO 6 nanosheets can be proficiently used for photocatalytic reduction of CO 2 . The yield of CPB/BWO is 503 µmolg -1 , which is 9.5 folds that of CsPbBr 3 .
Besides, two-dimensional (2D) 2D/2D heterojunction is considered as the most promising approach to manipulate the photoreduction efficiency of semiconductor catalysts. This is because that the large surface area of 2D ultrathin nanosheets, rich active sites, special electronic structure, and short charge transport distance, are useful for the catalytic reaction. [164][165][166] Recently, Jiang et al. designed a lead halide perovskite-based 2D/2D direct Z scheme heterojunction by assembling ultrathin Bi 2 WO 6 nanosheets on the surface of CsPbBr 3 nanosheets through an electrostatic self-assembly process (Figure 10 d-f). 167 Bi 2 WO 6 has been identified as one of the best visible-light oxidation photocatalysts, which have a higher oxidation ability, low manufacturing cost as well as controlled synthesis. 168,169 The CsPbBr 3 /Bi 2 WO 6 heterostructure provided higher photocatalytic performance for CO 2 reduction activity. The yield of the 2D/2D CsPbBr 3 /Bi 2 WO 6 hybrid was 153.0 µmol g -1 (10.9 μmol g -1 for H 2 , 56.4 μmol g -1 for CO, and 86.0 μmol g -1 for CH 4 ). 167

Sulfide with halide perovskite
A high-quality interface is strongly desired in heterojunctions for effectively separating electron-hole pairs both temporarily and spatially. 170 Co-sharing of atoms by two different materials in a heterojunction system could increase the charge separation and boost the carrier lifetime, facilitating the catalytic reaction. Recently, Wang et al. reported that 0D Cs 2 SnI 6 QDs anchored on flower-like 2D SnS 2 nanosheet by co-sharing of Sn atoms can improve the photochemical CO 2 reduction rate. 46 All-inorganic Cs 2 SnI 6 is favored for its outstanding conductivity and strong chemical stability. For example, Cs 2 SnI 6 thinner films own high hole mobility of 3.82 × 10 2 cm 2 V -1 s -1 , 171  reduction activities of SnS 2 , Cs 2 SnI 6 (0.5)/SnS 2 , Cs 2 SnI 6 (1.0)/SnS 2 , and Cs 2 SnI 6 (2.0)/SnS 2 . Reprinted with permission from ref. 46 , Copyright 2019 American Chemical Society.

Carbide with halide perovskite
Recently, carbon-based photocatalysts have aroused great concerns and became more prevalent in photocatalytic reduction of CO 2 owing to their exceptional physicochemical and photo-/electrochemical properties. Various carbon materials (e.g. GO, g-C 3 N 4 ) used as support for many photocatalysts is greatly promising as carbon can regulate photocatalytic reduction of CO 2 performance in the visible light region and extract photogenerated electrons from the surface of semiconductors because of its admirable electrical conductivity. [178][179][180] Figure 12 a-c shows the presentation of the CsPbBr 3 QDs/GO composite photocatalyst and its corresponding TEM image. 42 It was noted that, in ethyl acetate, the pure CsPbBr 3 QDs catalyzed the photocatalytic reduction of CO 2 with an electron consumption rate of 23.75 μmol g -1 h -1 over 99.2% selectivity. Through the combination of highly conducting materials, such as GO, which has charge extraction abilities, the charge recombination was suppressed. 181 Figure 12 d-f. 182 The abovementioned strategy helped to achieve the maximum rate of CO formation in acetonitrile/water mixture compared with bare CsPbBr 3 QDs. 183 Based on the above-mentioned composite structure, Guo et al. functionalized CsPbBr 3 NCs with graphitic carbon nitride, which had a titanium-oxide species (TiO-CN) through N- Br and O-Br bonding, and developed an effective catalyst system for photocatalytic reduction of CO 2 utilizing water as the source of the electron. 98 The introduction of TiO-CN could improve the number of catalytic active sites, along with a rapid interfacial charge separation between CsPbBr 3 and TiO-CN owing to their promising energy-offsets as well as chemical bonding performances (Figure 12 g-i). They also reported that the CsPbBr 3 @TiO-CN composite increased the separation of photogenerated charge and improved the number of catalytic active sites that led to the formation of CO after 10 hours of irradiation, which was 3 times higher than pure CsPbBr 3 . 98

Metal-Organic Framework with halide perovskite
With unique structural architecture and admirable chemical and physical properties, metal-organic framework (MOFs) has recently attracted remarkable consideration regarding encapsulation of halide perovskites owing to their distinctive characteristics, such as tunable structure, higher surface areas, and flexibility. 184  NCs@ZIF-67 composite with the highest electron consumption rate of 29.6 μmol g -1 h -1 , which was a 2.65 times increase compared to bare CsPbBr 3 NCs as shown in Figure 13 a-b.
It was started that CsPbBr 3 NCs covered by ZIFs shell can enhance the moister stability and the CH 4 formation instead of CO as the main product. 102  NCs/UiO-66(NH 2 ) composite toward the reduction of CO 2 in water/ethyl acetate with volume ratio (1/300) solution. 113 The highest conversion of CO 2 was accomplished for 15%-CsPbBr 3 QDs/UiO-66(NH 2 ) heterostructure. 113 The formation of CO was up to 98.5 μmol g -1 , which was higher as compared to the individual CsPbBr 3 QDs and UiO-66(NH 2 ) photocatalysts, as shown in Figure 13 c-d. The enhanced catalytic activity toward the reduction of CO 2 for

Noble Metal with halide perovskite
Recently the plasmon-exciton exchange dynamic method has been broadly suggested in metal-semiconductor systems. With the help from the famous localized surface plasmon resonance (LSPR) impact, noble metal Au nanoparticles (NP) was demonstrated by introducing light extinction into visible and even near-infrared zone. 191,192 CsPbBr 3 -Au nanocomposite was successfully produced through the in-situ reduction of AuCl 3 by the surface-bound oleylamine ligand, while the obtained Au NP size is too small to form the LSPR effect. 193,194 Xiao and co-workers examined the energy conversion in the Ag-CsPbBr 3 system as well as dynamics of plasmon-induced hot electron, which showed a significant hot-electron transfer efficiency near 50%. 195 The studies about plasmon-exciton interaction for the hybrid method of plasmonic Au NPs over halide perovskite NCs is still at a nascent development.
CsPbBr 3 -Au nanocomposite was used as photocatalyst for artificial photoreduction of CO 2 by two different types of light sources (λ > 420 nm and λ > 580 nm). Based on the Kelvin probe force microscopy (KPFM) results and the spectroscopic characterizations, twodimensional interaction processes for CsPbBr 3 -Au nanocomposite were shown in Figure   14. 194 Even though there is a significant difference between the Fermi level (E F ) of CsPbBr 3 and the work function (W F ) of Au, energy level alignment can be obtained with their close contact (Figure 14 a-b). Consequently, photogenerated electrons would mostly accumulate in CB of CsPbBr 3 , which raises its E F negatively, and thermodynamically supports the electron to transfer into the Au. 196,197 The electrons accumulated in Au consequently contribute to the photoreduction reaction of CO 2 , as demonstrated in Figure 14c. Only the Au nanoparticle- CO 2 reduction more efficiently, the LSPR-induced hot electrons encompassing higher energy potential in plasmonic Au nanoparticles could be further transferred rapidly to CsPbBr 3 .
Hence, such kind of plasmon-exciton exchange process makes the CsPbBr 3 -Au nanocomposite to work well over the longer wavelength in turn as shown in Figure 14d. played an important role for an electron collector to rapidly separate the electron-hole pairs in CsPbBr 3 NC and controlled the unwanted radiative charge recombination, but also modified the kinetics of catalytic reduction of CO 2 reaction. Therefore, the addition of Pd-NS increased the consumption rate of photoelectron from 9.85 μmol g -1 h -1 to 33.7 μmol g -1 h -1 under visible light (>420 nm). 100 Especially, the incorporation of 2D metallic semiconductor not only formed a Schottky junction at the interface to increase the electrons transfer from semiconductors to metal, but also presented photocatalytic reaction sites to promote the transfer of electrons and increased the resulting chemical reactions 164,198 , as presented in Figure 15 a-c.

Carbon Derivative with halide perovskite
Recently, 2D ultrathin carbon-based materials, such as carbon nanotubes (CNTs), graphene, and graphitic carbon nitride Photo-exciton will be shifted from a high energy level to a lower one, therefore sustaining the charge at a stable state. In this way, electrons are moved to the surface of the photocatalyst, which offers a comparatively negative position based on the band arrangement.
Pan and co-workers demonstrated that the incorporation of MXenen to CsPbBr 3 (CsPbBr 3 NCs/MXene-20) increased the yields of CO and CH 4 linearly. The maximum CH 4 and CO yield rate was 7.25 and 26.32 μmol g -1 h -1 , respectively (Figure 16 a-b), 47  energy barriers of reducing CO 2 to CO) for CO 2 reduction reaction, 215,216 which is thermodynamically more feasible for the reduction reaction.
Liu and co-worker 217

CONCLUSION AND FUTURE OUTLOOK
In this review, we summarize the recent progress achieved in exploring perovskite nanomaterials in terms of photocatalysts for CO 2 reduction. Improvement methods in this field are discussed systematically including utilizing newly emerging halide perovskite nanomaterials along with their modifications in terms of structural engineering, interfacial modulation through the formation of the heterostructure, metal ion doping, surface modification, encapsulation with several types of co-catalyst, or using the conducting substrates. Table 2 displays the performance of recent perovskite-based photocatalysts.
Significant advances have been made to the conversion of CO 2 into useful energy-bearing The utilization of co-catalysts in the photocatalytic process has been proved to be an environmentally benign method with high performance. 56  For artificial photosynthesis, the up to date solar-to-fuel efficiency value is ~1.1% 49 , which is ~10 times that of natural photosynthesis 235 , but is still very low for industrial applications. researches are required to investigate the productivity and selectivity of these C2+ products via photocatalytic reduction of CO 2 by halide perovskites to valuable products such as alkenes, aromatic hydrocarbons and alcohols, rather than less valuable products. These