1. Introduction

With the swift growth of the economy, the total demand for energy in the world is skyrocketing year by year. Fossil resources satisfy the majority of the world's energy consumption. However, the use of massive fossil energy has also exacerbated global warming and caused frequent haze. To meet the growing energy needs and alleviate the environmental problems, it is urgent to find a green, abundant, and economic alternative. Biomass is the largest renewable carbon resource, which is considered a promising alternative. It is a highly functionalized feedstock with an organic carbon structure, which is suitable to produce fine chemicals [1,2]. Various processes such as thermochemical and biological transformation have been proposed to convert biomass to high-value-added chemicals.

Nevertheless, for thermochemical transformation, harsh reaction conditions (such as high pressure or high temperature) and enormous energy consumption are inevitable [3-5]. In addition, it is hard to accurately transfer energy to the target reactants or chemical bonds, generating a low energy utilization rate.

Moreover, biomass usually has multiple functional groups, so selectively activating a specific functional group is out of the question, causing the low selectivity of the target product with various by-products [6-8]. As for biological transformation, the obstacle that must be overcome is the huge investment including costly reagents and equipment [9,10]. As an emerging technology, photocatalysis has attracted increasing concerns thanks to its harmlessness, environmental protection, sustainability, mild operational conditions, and cheapness [11,12]. Photocatalysis is a process in which inexhaustible solar energy can be converted into chemical energy and trigger the transformation of material (such as biomass) under the irradiation of light [13,14]. When biomass is the substrate of photocatalytic reaction, a variety of key fine chemicals, such as 2,5-dicarbonyl furan (DFF) [15] and 5-hydroxymethylfurfural (HMF) [16] can be obtained. Further, H₂ can also be got if an appropriate amount of electron donors is added [17].

The development and design of high-efficiency photocatalysts, the key to photocatalytic reaction, are a critical part of the photocatalytic transformation of biomass. There are two types of photocatalysts, homogeneous and heterogeneous photocatalysts [18-20]. The latter has better practical potential as it is more inexpensive and easier to separate [21-23]. The semiconductor material is a kind of the most common heterogeneous biomass photocatalyst due to its high selectivity, stable catalytic performance, long activation time, and low cost, with the most perfect examples being TiO₂, CdS, and g-C₃N₄ [24]. To the best of our knowledge, some excellent reviews have appeared in the literature. Wu et al. [25] provide a critical review of photocatalytic transformations of bio-based platform compounds to high-value chemicals. Alberto V. Puga [26] summarized a specific overview of biomass-to-H₂ routes. Considering the prospect of semiconductor material in the field of photocatalysis research, a comprehensive summarization of semiconductor photocatalytic mechanism and modification methods is needed to clarify the emerging research field of solar energy-driven biomass conversion.

This review mainly introduces the semiconductor materials represented by TiO₂, CdS, and g-C₃N₄ in the photocatalytic transformation of biomass, details the photocatalytic mechanism and the ways to promote the photocatalytic efficiency, such as structural modification, metal co-catalyst loading, and introduction of heterojunction. The existing problems, future development direction, and trend toward biomass photocatalysis prospects are in the last section.

2. Different semiconductors on biomass photocatalytic transformation

The semiconductor has a unique energy band structure, which consists of a low-energy valence band (VB) and a high-energy conduction band (CB). And the band gap between CB and VB is called the forbidden band, i.e., energy gap (E_g). In a typical photocatalytic process (Fig. 1a), the semiconductor absorbs light energy and generates high-energy electron (e^–) and hole (h⁺) pairs, and chemical reactions are activated by detached charge carriers. Importantly, the capacity of the photogenerated carriers (e⁻-h⁺ pairs) to facilitate a redox reaction depends on the potential of VB and CB.

Biomass photocatalytic transformation in general involves the hydrogen evolution reaction (HER) and biomass oxidation reaction (BOR). These two steps may be realized concurrently in one step by the photogenerated electrons and holes over semiconductor photocatalysts. The HER typically occurs over the photocatalyst with CB electrons to reduce protons (H⁺) and generate H₂. In contrast, the biomass oxidation reaction is complex and strongly depends on the substrate, catalyst surface, and reaction conditions. The holes with high oxidation potential are commonly considered the major active species responsible for biomass oxidation. In this process, the biomass is moved to a different potential energy surface and resulting in the evolution from the reactant to the product state (Fig. 1b). Essentially, the energy of the charge carrier is converted into the vibrational energy of the activated chemical bond [27].

More generally, photocatalytic transformations of biomass are mostly carried out under mild conditions. Through a well-designed semiconductor material, photon-generated carriers can carry out selective cleavage or functionalization of the targeted chemical bond. Thus, photocatalysis has greater potential to accomplish the selective conversion of biomass compared to conventional transformations.

2.1. TiO₂-based material

TiO₂ was the first example used to investigate the photocatalytic reforming of carbohydrates in 1980 [28]. It was found that the conversion of cellulose under light irradiation (500 W Xe lamp) produced H₂ with a quantum yield of 0.3%. Fig. 2 showed the photocatalytic mechanism of TiO₂. It can be summarized that CB is lower than normal hydrogen electrode (NHE) and VB is higher than biomass oxidation. Low CB is favorable for photo-generated electron hydrogenolysis [29], and high VB could efficiently drive the fission of C—C and C—O bonds. Meanwhile, the band gap of TiO₂ (3.0~3.3 eV) is less than 3.4 eV, which can maintain the stable catalytic effect under the irradiation of ultraviolet (UV) (< 387 nm) [30]. Generally, raw TiO₂ exhibits lower efficiency in biomass photocatalysis, because of the rapid electron-hole recombination, which limits its industrial application [31]. Time-resolved spectroscopic showed that the diffusion of electrons from the semiconductor to the active site occurs on a timescale of picosecond to nanoseconds [32]. Thus, 90% of electrons and holes would recombine after excitation, which often results in a lower quantum yield. These drawbacks can not be overcome by only optimizing the material itself. Various other strategies to cope with these problems have also been developed [33]. In the following part of this section, the method of modifying the catalyst structure of TiO₂ to improve the target yield will be discussed from the following three aspects: (1) Loading of metal elements, (2) construction of heterojunction, and (3) selection of catalyst carrier.

2.1.1. Loading of metal elements

Metal species play a special part in photocatalytic reactions. In detail, the metals act as active sites for H₂ production on the surface of TiO₂ [34]. Generally, Pt is one of the most effective and investigated co-catalyst due to its low overpotential of NHE and large work function. [35]. Bellardita et al. [36] examined the conversion of glucose to arabinose with Pt-TiO₂ under photocatalysis. Results showed that the conversion yield of glucose under 700 W medium pressure Hg lamp was 92%, about 7.6 times higher than the raw TiO₂ (12%). Authors found that Pt was uniformly scattered on the surface or matrix of TiO₂ particles in the shape of clusters or mononuclear complexes. About 0.5 electrons per Pt atom transferred from the d-orbital to the s-orbital in the reaction, thus, the 5d-orbital would have an excess of vacant orbitals, which can better capture photo-generated electrons and inhibit electron/hole recombination [37].

To further improve the conversion yield of the substrate, researchers have studied the effect of multi-metal co-loading in TiO₂. Bi/Pt-TiO₂ was used to convert lignin to valuable compounds in the photocatalytic process [38]. Results indicated that Pt provided active sites to alleviate the problem of photo-generated e⁻-h⁺ pairs recombination. In this system, Bi served as a promoter to boost the generation of O₂^•− from trapped photo-electrons. The conversion yield of lignin reached 84.5% after 1 h irradiation, while in the Pt-TiO₂ system, the conversion rate of lignin was only 28.23%. This research provides a novel opportunity for the transformation of biomass using Bi and noble metal-based co-catalysts under solar radiation.

Asides from Pt, Ag is another common co-catalyst of TiO₂ in the photocatalytic process. However, the catalytic effect mechanism of Ag is quite different from Pt, which is affected by the dispersion model and electronic properties [39]. The better photocatalytic effect of Ag-TiO₂ is owing to the difference in the Fermi energy of TiO₂ and Ag [40]. Higher Fermi energy in Ag can form a Schottky barrier, which contributes to the electron-hole pair separation. Xu et al. [41] also studied the synergistic effect of interfacial lattice Ag⁺ and Ag⁰ clusters in enhancing the photocatalytic. Results showed that photo-generated electrons can transfer CB (TiO₂) to the Ag⁰ cluster quickly and smoothly, resulting in improving the photocatalytic activity of Ag⁺/TiO₂@Ag⁰.

Besides Pt and Ag, some typical metals, like Au [42], Pd [43], Bi [44] and Ni [45], also can improve photocatalytic activity by building numerous oxidation reaction active sites (Table 1). The choice of loading metal to TiO₂ is vital to photocatalytic efficiency. The metals have a low overpotential for H₂, which can make it easier to generate H₂ [46,47]. When TiO₂ is loaded with these metals, the electrons migrate from TiO₂ to the noble metal. Thus, the metals can serve as active sites for H₂ production on the TiO₂ surface. To sum it up, the lower the overpotential of the metal, the higher the activity it shows. However, metal loading amount is also another key parameter that affects the photocatalytic reaction. Deposited metal can promote photo-generated carrier separation effectively, whereas excessive metal loading reduces photocatalytic efficiency [48,49]. Specifically, superfluous metal particles form metal clusters or blocks on the surface of the semiconductor, impeding light absorption. In addition, in most studies, the reusability and stability of catalysts were not detail discussed due to the well-known stability of TiO₂ [50].

2.1.2. Construction of heterojunction

Constructing heterojunction is an effective way to regulate the band gap of TiO₂ in photocatalytic transformation [51,52]. TiO₂ is a typical n-type semiconductor (electron-rich), which can combine with a p-type semiconductor (hole-rich) to form a typical staggered gap structure (n-p junction). Two semiconductors possess different band gap structures, and they can form the built-in electric field due to the spontaneous diffusion of electrons. The photo-generated electrons under suitable irradiation could be forced to move through the built-in electric field, therefore inhibiting the recombination of e⁻-h⁺.

NiO is a common p-type semiconductor. It can construct an n-p heterojunction with TiO₂ on account of staggered band structures. Zhao et al. [53] prepared TiO₂-NiO through a one-step hydrothermal method. TiO₂ nanoparticles were coated by NiO nanoclusters and results showed that TiO₂-NiO had high reaction activity (Fig. 3). The apparent quantum yield (AQY) of TiO₂-NiO reached 6.46% under 365 nm monochromatic light, which was around 2.9 times greater than that of raw TiO₂. Besides the core-shell structure of the n-p junction, Wang et al. [54] reported a large 2D nano junction photocatalyst by growing 2D MoS₂ nanosheets on the superficial of TiO₂ (Fig. 4). 2D-2D MoS₂-TiO₂ photocatalysts were constructed to convert lignocellulose to H₂. An AQY of 1.45% (H₂) was acquired under 380 nm (Xe) over MoS₂(2%)-TiO₂, which is better than the current TiO₂-based photocatalysts.

The construction of heterojunctions has been demonstrated as one of the most efficient methods to upgrade the spatial separation of carriers. However, in the process of synthesizing heterojunction, some defects at the interface or surface states would be formed, which is harmful to the form of the built-in electric field. Therefore, a simple and steady synthetic method is very necessary to gain impeccable nanocomposite architecture.

2.1.3. Selection of catalyst carrier

Pure nanoscale TiO₂ has the disadvantage of being easily aggregated in solution, which decreases the specific surface area of the catalyst. A better carrier should not only provide a rich reaction site and large surface area but also promote the reaction through its physical and chemical properties [55]. The most common catalyst carrier is acidic zeolite. Wang et al. [56] combined TiO₂, Au, and zeolite HY to form Au-HYT. Au-HYT was used to convert cellulose to glucose with a yield of 48% under light irradiation at 140 ℃. Zeolite HY plays a pre-decomposition role in cellulose, which makes the reaction easier. As the reaction continues, HMF was also produced as a co-product, and the overall yield of glucose and HMF reached 60%.

Carbon nanotubes (CNT) are another common catalyst carrier. Photo-generated electrons can transfer from TiO₂ to CNTs, thus retards the recombination of photo-generated e⁻-h⁺. Li et al. [57] prepared CNT-TiO₂ by sol-gel technique, and the structure of TiO₂-CNTs nanocomposites was shown in Fig. 5. Results revealed that the conversion of CNT-TiO₂ was about two times as high as the conversion of pure TiO_2. Moreover, Takenaka et al. [58] loaded Pt on TiO₂-CNT to boost photocatalytic efficiency. When the loading amount of Pt was 1.5%, the photocatalytic activity of TiO₂-CNT@Pt was about 1.6 times over that of TiO₂-CNT and 2.9 times over that of pure TiO₂. As shown in Fig. 6, part of Pt was deposited in the cavity of CNT and the rest was on the outer surface of CNTs. Pt on the outer surface provided a catalytic active center, in addition, the deposited Pt in the cavity would inhibit the reverse reaction.

Generally, the oxidative potential of the photogenerated holes in TiO₂ is around 3.2 eV, which can be viewed as strong oxidants. Nevertheless, this is also the bottleneck of TiO₂, these photo-induced oxidants can oxidize reactant to the final product in one step, thus the selectivity is typically not high. There is still often a trade-off between product selectivity and conversion.

2.2. CdS-based materials

Although the research regarding TiO₂ photocatalytic biomass has obtained a huge development, the inherent shortcomings of TiO₂ photocatalyst like limited light absorption (< 387 nm) and serious recombination of photo-generated carriers, still urge scientist to find more types of photocatalyst, and CdS is among these. CdS is a typical Ⅱ-Ⅵ semiconductor photocatalyst, and it has become one of the most potential biomass photocatalyst materials given the advantages of special band structure, diverse modifiability, low price, and simple preparation method [59,60]. At 300 K, the band gap of CdS is 2.4 eV, which allows it to absorb solar energy in a wider spectrum and promote the utilization of light energy (wavelengths shorter than 516 nm). Moreover, comparing the edge position of CB in photocatalyst, CdS is more negative than others (such as TiO₂, SrTiO₂, and ZnO), which makes the electrons generated in CdS photocatalytic reaction have a stronger reduction ability.

Research also discovered that the microstructure of CdS had a major effect on photocatalytic activity. Song et al. [61] probed into the key role of the microstructure of CdS in the photocatalytic production of amino acids from biomass-based feedstocks. Results showed that the ultra-thin CdS nanosheets had higher activity for lactic acid to produce alanine compared with other forms of CdS (such as nanospheres and nanorods). Moreover, CdS quantum dots (QDs, < 6.5 nm) acted greater activity than CdS nanoparticles (20~40 nm) due to the more surface sites and the improved redox ability [25].

Similar to pure TiO₂, pure CdS does not have better photocatalytic efficiency. During the last decades, extensive efforts have been made to enhance the photocatalytic activity and stability of CdS semiconductors. Thus, some important and representative strategies, including loading of co-catalyst, functionalized with ligands, and avoiding photo-corrosion are discussed in the following sections.

2.2.1. Loading of co-catalyst

To better realize the oxidation of biomass and the generation of H₂, the co-catalyst can be loaded on CdS [62]. Loading a co-catalyst on the surface of photocatalysts is a favor for capturing photo-generated electrons or holes so that they can be transferred to the catalyst surface quickly. Han et al. [63] reported a photocatalyst of ultrathin CdS nanosheets loaded Ni as a co-catalyst (Ni-CdS), which can selectively oxidize the biomass-derived alcohol under visible light and obtain H₂ in the meantime. In this nano-architecture, the delivery of photogenerated electrons from the CdS to the metal surface, promotes the separation of electron and hole carriers, thereby leading to the performance improvement of photocatalytic H₂ evolution. In this research, the photocatalyst showed excellent catalytic activity and stability. The corresponding products (aldehyde or acid) can be selectively obtained through adjusting the catalyst and reaction conditions. Besides Ni, other co-catalyst, such as NiS [64], MoS₂ [65], and Pt [66] were also used to improve the photocatalyst efficiency (Table 2). Metallic particles deposited on the surface of the semiconductor often serve as collectors for photo-generated electrons and catalyze the reduction of protons to H₂, depending on the work function of the metal and the potential of the redox processes involved.

2.2.2. Functionalized with ligands

Wu et al. [67] applied CdS QDs to the photocatalysis of lignin and realized efficient conversion of native lignin and full utilization of lignocellulose under visible light and mild conditions for the first time. CdS QDs highly dispersed in the solvent can contact native lignin well during the reaction process, so they present excellent catalytic performance. The cleavage of the β-O-4 bond in lignin is achieved by an electron-hole coupled photoredox mechanism on account of Cα radical intermediate, where both photo-generated e⁻-h⁺ take part in the reaction (Fig. 7a). For further confirmation of the role of organic ligands stuck to CdS QDs in the photocatalysis of lignin, Wu et al. [68] systematically researched the photocatalytic conversion of native lignin by employing CdS QDs with various surface ligands (Fig. 7b). Photocatalytic behavior of CdS QDs is strongly linked to the organic ligands structure in the transformation from raw lignin to high-value functionalized aromatics.

The hydrophilicity/hydrophobicity of ligands is vital for the joint connection between QDs and lignin, bringing about the better catalytic activity. Ligands also take part in the key step of the photocatalysis-electron transfer process with the mechanism of ligand-mediated electron tunneling.

The photocatalytic activity of CdS can be also enhanced by surface modification. Zhang et al. [69] prepared [SO₄]-CdS catalyst by importing surface sulfate ions on CdS catalyst. The surface sulfate ion [SO₄] can be used as a proton receptor to effectively shorten the distance of proton transmission and promote the transfer of proton-coupled electrons. In the photocatalytic reaction of biomass polyol to synthesis gas, the production rate of CO and H₂ are is 9 times and 3 times than pure CdS, respectively.

We ascribed the high activity of this hybrid photocatalytic system to improvement in charge carrier separation and the efficient electron transfer from the CdS to ligands. On the other hand, the hydrophilic ligands rendered the water-insoluble catalyst well-dispersed in an aqueous solution. The encapsulation of the nanoscale semiconductor particles by ligands prevented the catalyst from aggregating and improved the efficiency of active sites.

2.2.3. Avoiding photo-corrosion

CdS could suffer from serious photo-corrosion and continual recombination of photo-generated e⁻-h⁺ in the photocatalytic process. Photo-corrosion of CdS is attributable to an irreversible oxidation reaction driven by photo-generated holes. Under long-term visible light, CdS will be decomposed (CdS + 2h⁺ = Cd²⁺ + S²⁻) and then Cd²⁺ will be dissolved into liquid part [70], which seriously affects the photocatalytic stability. In addition, due to the narrow band gap (2.4 eV) of CdS, the electrons in VB transited to CB will recombine with the holes in a few picoseconds. To upgrade the photocatalytic performance of CdS, it is essential to optimize the contact environment between CdS and the reactant interface, thus promoting the separation and transfer efficiency of photo-generated e⁻-h⁺.

Among synthetic methods aiming at increasing the stability of CdS, the formation of a surface layer on CdS proved beneficial over successive cycles. This strategy forms core/shell structures with protective layers on the surface of CdS nanoparticles, inhibiting corrosion effectively. Wakerley et al. [71] constructed a CdO_x protective shell on the outside surface of CdS under alkaline conditions (Fig. 8). The obtained photocatalyst (CdS/CdO_x QDs) can be applied in the oxidation of unprocessed lignocellulosic substrates to achieve the light-driven H₂ evolution without photo-corrosion. Meanwhile, the quantum dots in CdS/CdO_x provide a place for biomass oxidation reaction and enhance the separation of photo-generated e⁻-h⁺. The system could operate stably under visible light beyond six days and produce high H₂ production rate (1~9 mmol g⁻¹ h⁻¹).

Taking these mechanistic fundamentals into account, an efficient and stable photocatalyst system based on CdS semiconductors should achieve high efficiency in visible-light harvesting, and facilitate the photogenerated charge carrier separation. Despite this progress, some challenges still need to be addressed. Firstly, the current CdS photocatalyst systems are still far from being used for industrial production with both economic and environmental benefits. There are environmental and toxicity concerns about the potentially harmful release of cadmium. Additionally, the photo corrosion of CdS and the poor photostability are still key problems obstructing large-scale applications.

2.3. g-C₃N₄ based materials

Nitrogen-rich organic precursors, such as thiourea, urea, cyanamide, dicyandiamide, as well as melamine, can be used to fabricate g-C₃N₄ by thermal polymerization method [72]. g-C₃N₄ is a distinctive carbon material, which is one of the most steady allotropes of carbon nitrides. g-C₃N₄ is a better photocatalyst due to its mild bandgap energy of 2.7 eV (460 nm) [73] as well as proper CB (−1.1 eV) and VB (+1.6 eV) [74]. In the recent decade, among various semiconductors, g-C₃N₄ is one of the most utilized semiconductors in various photocatalytic processes, due to its numerous advantages like non-toxicity, bio-compatibility, low cost, and high photochemical and thermal stability.

Rao et al. [75] peeled monolayer g-C₃N₄ from bulk g-C₃N₄ by nitrogen-protected ball milling method (Fig. 9). The obtained monolayer g-C₃N₄ retains the intrinsic structure with an atomic layer. This structure can degrade the exciton binding energy, cut down the charge transport distance, and stimulate the separation efficiency of carriers. As a result, effectively ameliorative its performance of photocatalytic degradation of lignocellulose. Results showed that monolayer had better photocatalyst performance in cellulose (872 µmol H₂·g_cat.⁻¹), hemicellulose (719 µmol H₂·g_cat.⁻¹), and lignin (249 µmol H₂·g_cat.⁻¹) conversion under visible light illumination (300 W) after 12 h.

So far, g-C₃N₄ has harvested extensive attention due to its unique two-dimensional structure [76]. It has also emerged as one of the satisfactory stable organic polymers [77,78]. Many modifications have been proposed to ameliorate the photocatalytic capability of g-C₃N₄ by employing various strategies, such as surface defects (N-defected g-C₃N_x) [79], ultrafine structures [80] and QDs/g-C₃N₄ [81].

2.3.1. Loading of metal elements

Similar to TiO₂ and CdS, the catalyst efficiency of g-C₃N₄ can be tuned by modifying the band structure, for instance, by loading metals in biomass photocatalytic transformation [82]. Yang et al. [83] used bimetal (Ni-Au) to modify g-C₃N₄ (Ni-Au/g-C₃N₄) by a step-by-step photo-deposition method, thus promoting the co-production of furfural and H₂ from furfural-alcohol aqueous solution. The product yields with Ni-Au/g-C₃N₄ catalyst were significantly improved and almost 3.2 and 2.9 times higher than bare g-C₃N₄ and Au/CN photocatalyst, respectively. Results showed that g-C₃N₄ was the main light-harvesting substance for generating e⁻-h⁺ pairs. Photo-generate carriers could be separated efficiently at Ag/g-C₃N₄ interface and Ni could be worked as both electron capture and active sites for H₂ evolution.

Nonetheless, g-C₃N₄ has some inherent disadvantages such as inefficient visible-light absorption (below 460 nm), less active sites for interfacial reactions, and inefficient specific surface area (about 10 m²/g). The design of perfect-activity g-C₃N₄ is highly dependent on the morphology, plentiful reactive sites, and especially extended light-harvesting ability. However, the superior metal-free properties of g-C₃N₄ have not received enough importance, with only a few research focusing on the photocatalytic transformation of selected biomass derivatives.

3. Conclusion and outlook

Semiconductor photocatalysts are always a research hotspot due to their unique energy band structure and excellent photocatalytic performance. Applications of semiconductor materials in photocatalytic reforming of biomass and its derivatives have realized the rational utilization of biomass resources, which is of great emphasis to alleviate the crisis of fossil energy and realize the sustainable development of the environment. Increased requirements for establishing a clean and green energy system, combined with escalating environmental standards and government initiatives, have all contributed to an in-depth study in photocatalysis. Research and development on using semiconductors as efficient catalysts can play a vital role in advancing the practical application of photocatalytic technology. At present, this technology has successfully realized the conversion of a variety of biomass and its derivatives to H₂ and high-value chemicals. In this paper, the influences of TiO₂, CdS, and g-C₃N₄ on valuable products from photocatalytic transformation biomass were discussed thoroughly with representative instances.

Section 2.1 (TiO₂-based materials) focuses on discussing the effects of the conversion of biomass by photocatalytic transformation. Despite many methods that have been realized to enhance its photocatalytic activity, like metal particle loading and constructing heterojunction, industrialization is yet to be realized. For CdS-based materials part (Section 2.2), the influences of photo-corrosion, co-catalysts, and surface ligands on biomass conversion under photocatalytic transformation were summarized. For M-CdS, in many works, it gave desirable photocatalytic transformation activity, but its environmental toxicity remains yet to be worked out. A brief introduction to g-C₃N₄ based material was shown in Section 2.3. Due to its structural and semiconductor optical properties, g-C₃N₄ has shown great potential as a catalyst in energy conversion.

Based on the existent research progress on biomass conversion under photocatalytic reaction, we propose the following views and opinions:

(1) The modification of photocatalyst. As we have seen, the efficiency of photocatalytic transformation of selected biomass derivatives is still low for most semiconductor photocatalysts. The commonly used evaluation parameters are the conversion of biomass, the rate of H₂ generation as well as quantum yield. However, part of the incident sunlight energy can be stored in the final H₂ fuel and value added chemicals. In this regard, the energy efficiency of the overall sum of processes, starting from raw biomass and concluding with H₂, is a parameter of paramount importance. Up to now, there is still a great deficiency between the present studies and industrial expectations in biomass conversion. An ideal photocatalyst is needed to perform the reaction mildly with high efficiency. Photocatalyst development is a major bottleneck for the semiconductor photocatalysis of biomass to obtain valuable products.

(2) The exploration of mechanism. In current studies, the photocatalytic process of semiconductor materials has been clearly defined, but the fracture and formation paths of chemical bonds in the biomass photocatalytic reaction are vague. Most studies only reveal the relationship between material structures and experimental results, and a few articles cover the catalytic mechanism but still do not explain it thoroughly. The development of isotope labeling may help us in further understanding. Further clarification of the catalytic mechanism will greatly heighten the application performance of semiconductors and expand the broader application areas.

(3) The selection of substrate. Most studies chose to use pure substances to explore photocatalytic performance. Biomass is very complex and recalcitrant with various functional groups, which have abundant C—O and C—C bonds. Its composition and proportion vary from source to source, and the mechanism of interaction between components is complex. Expanding the range of substrates and making efficient use of biomass are the directions of further research.

(4) Cost-benefit analysis. There is no doubt that the photocatalytic transformation of selected biomass derivatives helps to take advantage of solar energy. However, before industrialization, the cost-benefit analysis should ensure it conforms to economic feasibility. The reusability and stability of catalysts are crucial factors to reduce the cost of industrialization. In addition, related analysis shows that further engineering of the photoreactor is crucial to improve light penetration and mass transfer for optimal photon absorption. Thus, from an economic standpoint, apart from photocatalyst design, photoreactor engineering should be given equal consideration in future investigations.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The authors are grateful for the financial supports provided by the National Natural Science Foundation of China (Nos. 22108221 and 52203145) and the Shaanxi Natural Science Foundation (No. 2021JQ-028). The authors wish to thank Yayun Ma (Core Facilities and Experiment Center of Xi'an Jiaotong University) for technological guidance.

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2023.108723.