Selective Control of Catalysts for Glycerol and Cellulose Hydrogenolysis to Produce Ethylene Glycol and 1,2-Propylene Glycol: A Review

Abstract

The bioconversion of cellulose and the transformation of glycerol can yield various diols, aligning with environmental sustainability goals by reducing dependence on fossil fuels, lowering raw material costs, and promoting sustainable development. However, in the selective hydrogenolysis of glycerol to ethylene glycol (EG) and 1,2-propylene glycol (1,2-PG), challenges such as low selectivity of catalytic systems, poor stability, limited renewability, and stringent reaction conditions remain. The production of diols from cellulose involves multiple reaction steps, including hydrolysis, isomerization, retro-aldol condensation, hydrogenation, and dehydration. Consequently, the design of highly efficient catalysts with multifunctional active sites tailored to these specific reaction steps remains a significant challenge. This review aims to provide a comprehensive overview of the selective regulation of catalysts for producing EG and 1,2-PG from cellulose and glycerol. It discusses the reaction pathways, process methodologies, catalytic systems, and the performance of catalysts, focusing on active site characteristics. By summarizing the latest research in this field, we aim to offer a detailed understanding of the state-of-the-art in glycerol and cellulose conversion to diols and provide valuable guidance for future research and industrial applications. Through this review, we seek to clarify the current advancements and selective control strategies in diol production from glycerol or cellulose, thereby offering critical insights for future investigations and industrial scale-up.

1. Introduction

Diols are a vital class of organic compounds, defined by the presence of two hydroxyl (−OH) functional groups in their molecular structure. These compounds find extensive applications in the chemical industry and materials science due to their distinct physical and chemical properties. Notable examples of diols include ethylene glycol (EG), 1,2-propylene glycol (1,2-PG), and 1,4-butanediol (1,4-BD). Ethylene glycol is widely used in antifreeze formulations and polyester fiber production, while propylene glycol is extensively utilized in food additives, pharmaceuticals, and cosmetics. In polymer synthesis, diols play a crucial role, particularly in the production of polyurethanes and polyesters, where they function as chain extenders and cross-linking agents, significantly influencing the material’s final properties [1]. Various methods exist for synthesizing diols, primarily through the hydration of epoxides and the dihydroxylation of olefins. In recent years, with the growing emphasis on green chemistry and sustainability, the synthesis of diols from biomass and glycerol has garnered substantial attention, aiming to reduce dependence on petrochemical resources.

According to the U.S. Department of Energy, glycerol is a major by-product of the biodiesel industry, accounting for 10% of the total production. By 2020, the global release of glycerol was projected to reach 10 billion liters. In 2023, the global glycerol market size reached USD 1342.03 million, and it is projected to grow to USD 2076.11 million by 2030 [2]. Current market applications which exist and have large-scale use (pharma and personal care 42%, Alkyd resins 13%, Food and Beverage 14%, Polyether polyols 10%, tobacco 9.5%, others 11.5%). Considering future trends, as various industries push for cleaner and more sustainable fuel alternatives, the demand for such derivatives is expected to rise. Over the next 25 years, the role of glycerol in fuel additive production may expand, thanks to innovations in catalyst design that will enhance glycerol conversion efficiency and selectivity [3]. Disposing of glycerol as waste is both costly and undesirable, given its potential for conversion into value-added chemicals due to its unique structural properties. Valuable chemicals produced from glycerol hydrogenolysis include diols such as 1,2-PG, 1,3-PG, and EG. Similarly, cellulose, as a critical biopolymer, has garnered substantial attention because of its distinctive structure and properties. It is a linear polysaccharide composed of glucose molecules linked by β-1,4-glycosidic bonds, forming the structural basis of cellulose (see Figure 1). This specific structure endows cellulose with remarkable tensile strength and chemical stability, making it an ideal engineering material [4]. Its chemical formula is (C₆H₁₀O₅)_n, where n represents the degree of polymerization, typically exceeding 10,000. Shifting production from fossil-fuel-based processes to biomass-based alternatives, such as those utilizing glycerol or cellulose, significantly reduces the overall carbon footprint of these compounds. Producing EG and 1,2-PG from renewable sources can potentially lower the demand for fossil fuels by offering more sustainable processes. In recent years, significant advancements [5,6] have been made in the synthesis of EG and 1,2-PG, covering various aspects such as reaction mechanisms, process optimization, feedstock selection, pretreatment methods, and catalyst structures. Several key studies on glycerol hydrogenolysis for propylene glycol production have been summarized in the literature, focusing on proposed mechanisms, developed catalysts, and specific process conditions. Two notable reviews have comprehensively discussed glycerol hydrogenolysis over non-precious metal catalysts, both with and without the presence of hydrogen. While these works offer valuable insights, they primarily focus on glycerol hydrogenolysis under conventional conditions or with specific catalyst types.

In contrast, this review adopts a broader perspective by focusing not only on glycerol but also on cellulose as a feedstock for hydrogenolysis. Selective control over the synthesis process is particularly critical for achieving high selectivity towards EG and 1,2-PG while minimizing by-products. Current literature lacks comprehensive reviews addressing selectivity optimization of catalysts for both glycerol and cellulose hydrogenolysis. This work fills that gap by providing an in-depth analysis of catalyst structures, reaction pathways, and process methodologies, offering a more complete understanding of the factors influencing catalytic performance. By integrating the use of renewable feedstocks like glycerol and cellulose, this review also highlights sustainable alternatives to fossil-fuel-based production methods. Ultimately, this paper aims to guide future research and contribute to the development of more efficient, selective, and sustainable catalytic systems.

2. Glycerol to Diols Conversion

2.1. The Reaction Pathway of Glycerol Hydrogenolysis

Glycerol can be utilized via various chemical modification methods, including hydrogenolysis [7], dehydration, acetalization [8], oxidation, etherification, and steam reforming [9]. Among these, glycerol hydrogenolysis stands out as a widely researched method for value enhancement. Glycerol undergoes hydrogenolysis to yield several products depicted in Figure 2 [10]. Hydrogenolysis reactions, whether conducted in the liquid or gas phase, have the potential to transform glycerol into a wide range of valuable chemicals. Depending on the extent of hydrogenation, glycerol hydrogenolysis can yield various products, including diols (e.g., 1,2-propylene glycol, 1,3-propylene glycol, ethylene glycol), monoalcohols (e.g., 1-propanol, 2-propanol, ethanol, methanol), and alkanes (e.g., propane, ethane, methane).

Mechanism of Hydrogenolysis of Glycerol

The hydrogenolysis of glycerol involves multiple reaction route and product formation routes. According to existing studies, its reaction mechanisms can be primarily divided into two types: the dehydration–hydrogenation mechanism and the dehydrogenation–dehydration–hydrogenation mechanism. In the dehydration–hydrogenation mechanism, glycerol is converted to 1,2-PG through reactive intermediates such as acetol and its enol tautomers [11]. Route 1 of Figure 3 involves the conversion of glycerol into 1,2-PG through reactive intermediates, specifically acetol and its enol tautomers. These intermediates are subsequently hydrogenated to form 1,2-PG. Route 2 of Figure 3 describes the conversion of glycerol to 3-hydroxypropanal through dehydration, followed by hydrogenation to produce 1,2-PG. Route 3 of Figure 3 outlines the direct hydrogenation of glycerol via C-C bond cleavage, leading to the formation of EG and methanol [12]. Zhang et al. reported [13] that glycerol hydrogenolysis over the PtFe/Y-IE catalyst follows a dehydration–hydrogenation mechanism, as shown in Figure 4. The catalyst’s Brønsted acid sites play a crucial role in the activation of C–H and C–O bonds in glycerol, significantly enhancing its dehydration. The modified surface acidity facilitates the dehydration of glycerol into acetol, which acts as a key intermediate in the production of 1,2-PG and other glycols. Hydrogen species accumulate on Pt sites due to the cleavage of C–H bonds. Fe³⁺-induced surface acidity further aids the hydrogen transfer steps by weakening the Pt–H bond, which promotes the cleavage of the C=O bond in acetol, ultimately yielding 1,2-PG as the final product.

Although theoretically, both 1,2-PG and 1,3-PG can be produced in the dehydration–hydrogenation process, propanol is invariably generated as a major by-product. This by-product formation is a primary reason for the reduced selectivity towards 1,2-PG. Additionally, glycerol can be directly hydrogenated to ethylene glycol and methanol via C-C bond cleavage. Acetol intermediates can achieve high selectivity through catalytic reactive distillation. Suppes et al. [14] demonstrated this approach, providing direct evidence for the dehydration–hydrogenation route and the formation of 1,2-PG under such conditions.

In alkaline conditions, the dehydrogenation–dehydration–hydrogenation mechanism is widely adopted for glycerol hydrogenolysis [15]. In Route 1 of Figure 5, the process begins with the reversible dehydrogenation of glycerol on metal sites to form glyceraldehyde. Subsequently, the glyceraldehyde intermediate undergoes dehydration in an alkaline environment to yield 2-hydroxyacrolein, which is then hydrogenated on metal sites to produce both acetol and 1,2-PG. Route 2 of Figure 5 outlines the dehydrogenation of glycerol to glyceraldehyde, which may proceed via a retro-aldol condensation reaction on alkaline sites, leading to the formation of intermediates such as 2-hydroxyethanal and formaldehyde. These intermediates are then hydrogenated on metal sites to produce ethylene glycol and methanol, respectively. The hydrogenolysis of methanol and hydrogenation of CO on metal sites may also lead to the formation of methane. It is important to note that this route could involve side reactions, demonstrating the potential for multiple product formation routes and mechanisms during glycerol hydrogenolysis. By modulating the dehydration rate of glyceraldehyde (to form 1,2-PG) and the dehydrogenation rate (to form ethylene glycol), the selectivity between 1,2-PG and ethylene glycol can be controlled.

Hybrid mechanisms [16] have also been reported, which integrate some reactions from the dehydration–hydrogenation mechanism with parts of the dehydrogenation–dehydration–hydrogenation mechanism, forming a combined process. Wu et al. proposed two possible pathways for Acetol formation: the traditional route, which involves the dehydration of glycerol over support sites, and another pathway involving the dehydrogenation of glycerol to glyceraldehyde over metal sites, followed by the dehydration of glyceraldehyde to pyruvic aldehyde, and finally the hydrogenation of pyruvic aldehyde to yield (Figure 6).

The reaction routes and mechanisms for glycerol hydrogenolysis provide insights into the potential production of a range of valuable chemicals. The reaction mechanisms highlight the pivotal role of both metal and alkaline sites in directing the reaction’s outcome, where a delicate balance between dehydration, dehydrogenation, and hydrogenation steps is crucial. The insights gained from such mechanistic studies can guide the design of catalysts and reaction conditions optimized for the selective production of desired glycerol hydrogenolysis products.

3. Selectivity Modulation in Glycerol Hydrogenolysis

I In the hydrogenolysis of glycerol using heterogeneous catalysts, metals such as Cu, Ni, and Co are commonly employed due to their cost-effectiveness. Copper-based catalysts are particularly noted for their strong ability to cleave C-O bonds, which is crucial for the production of diols like 1,2-PG. This process involves two main steps: the removal of −OH groups, typically facilitated by acidic or basic functional groups present in metal oxides or supports, followed by hydrogen activation and hydrogenation through redox reactions catalyzed by the metal. Nickel-based catalysts are also effective but generally exhibit lower efficiency in C-O bond cleavage compared to copper. Cobalt-based catalysts are less commonly used but can be effective under specific conditions.

Transition metal catalysts (e.g., Cu, Ni) generally exhibit a higher efficiency for glycerol hydrogenolysis, with the efficiency order being Cu > Ni > W > Mo. Noble metal catalysts (e.g., Ru, Pt, Pd) play a significant role in enhancing selectivity towards ethylene glycol (EG). The efficiency order for these catalysts is Ru > Pt > Pd > Ir. These noble metals often influence the reaction pathways to favor the production of EG, especially under high glycerol conversion rates (>75%). Catalysts containing Group VIII metals, such as Ru, Rh, Pt, and Ni, are also known to cleave C-C bonds in glycerol, leading to the formation of EG and other lower alcohols. While supported noble metal catalysts like Pt, Ru, Rh, Pd, Re, and Ag show notable activity in glycerol hydrogenolysis, their high cost remains a major consideration. The selective use of these catalysts is crucial for achieving efficient conversion and high selectivity to diols, underscoring the importance of tailored catalyst design to balance conversion rates and product selectivity. The following

Table 1. Reaction performance of EG and 1,2-PG prepared by different catalytic systems of hydrogenolysis of glycerol.

Catalyst	Reaction Conditions	Conversion/%	S EG/%	S1,2-pG/%	Ref.
Cu-ZnO-Al2O3	473 K, 6.0 MPa H2, 6 h	31.8	9.9	90.1	[17]
CuCr2O4	403 K, 2.0 MPa H2, 4 h			99.6	[18]
Cu/MgO	453 K, 3.0 MPa H2	72	1.3	97.6	[19]
Cu/SiO2	513 K, 8.0 MPa H2, 5 h	52	1.8	96.2	[20]
35%Cu/MgO	483 K, 4.5 MPa H2, 12 h	96.6	5.1	92.6	[21]
Raney-Ni6Ag	483 K, 4 MPa H2, 6 h	69.2	5.9	65.9	[22]
20 wt% Cu-Ni (1:1)/γ-Al2O3	493 K, 0.75 MPa H2	98.5	1.8	89.3	[23]
Ru/C	473 K, 4 MPa H2	20	68	32	[24]
Ru/m-ZrO2	473 K, 6 MPa H2, 3 h	22.9	21	45.7	[25]
Rh/m-ZrO2		22.8	12.4	65.6	[25]
Pt/m-ZrO2		20.4	7.4	85.7	[25]
Ru/CeO2	473 K, 6 MPa H2, 12 h	72.4	5.2	66	[26]
PtSn/SiO2	498 K, 1.6 MPa H2	16		83	[27]
Pt/Al2O3-AC	453 K, 0.6 MPa H2	14	17	36	[28]

summarizes the conversion rates and selectivity of glycerol to ethylene glycol and 1,2-PG under various catalysts:

3.1. Transition Metal Catalyst

3.1.1. Cu-Based Catalyst

Researchers have extensively investigated various transition metal catalysts for the selective hydrogenolysis of glycerol. Among these, copper-based catalysts are particularly noteworthy for their application in both batch and continuous processes. Due to copper’s ability to cleave C-O bonds while exhibiting low activity towards C-C bonds, these catalysts demonstrate high selectivity to 1,2-PG hydrogenolysis reactions. Wang et al. [29] achieved a high selectivity to 1,2-PG (83.6%) and a combined selectivity to 1,2-PG and EG (94.3%) using a Cu-ZnO catalyst prepared by co-precipitation at 473 K. This study revealed the dual functionality of the catalyst, where the acidic ZnO surface promotes glycerol dehydration to intermediates such as acetol and glycidol, while hydrogenation occurs on the Cu surface. This finding underscores the high efficiency of Cu/ZnO catalysts in glycerol conversion, offering significant industrial potential. Interestingly, a higher Cu/ZnO ratio resulted in lower glycerol conversion rates but increased selectivity to EG. Additionally, increasing the reaction temperature (453–513 K) was crucial in enhancing both glycerol conversion rates and selectivity to EG. This effect can be attributed to the differential adsorption/desorption characteristics and conversion rates of dehydrated glycerol intermediates on Cu and ZnO surfaces. At lower temperatures, the adsorption of intermediates is more favorable, leading to higher decomposition rates. Conversely, at higher temperatures, the desorption process is enhanced, promoting C-O and C-C bond cleavage and thereby increasing selectivity for EG [30]. A ZnO/poly(acrylic acid) (PAAH) hybrid nanomaterial was utilized as a support for copper catalysts, prepared via the urea deposition-precipitation method for glycerol hydrogenolysis. The study highlighted that variations in synthesis parameters influenced the morphology, specific surface area, and consequently, the size and dispersion of the copper nanoparticles. These effects significantly impacted the catalytic performance of the material. With a high Cu loading of 35 wt%, the catalysts maintained copper stability without sintering and achieved impressive conversions (over 60–70%) and high selectivity (above 90%) for 1,2-PG under conditions of 200 °C, 30 bar H₂, and 24 h.

The modification of catalyst support composition can further enhance the activity, selectivity, and stability of catalysts for diol production. Liu et al. [17] reported the introduction of Al₂O₃ into Cu-ZnO catalysts as a method to improve catalyst performance. They prepared three types of Cu-ZnO-Al₂O₃ catalysts using homogeneous co-precipitation (CZA-HP), deposition-precipitation (CZA-DP), and conventional co-precipitation (CZA-CP) methods. A Cu-ZnO catalyst prepared by homogeneous co-precipitation (CZ-HP) served as a reference. In recycling experiments at 473 K and 6.0 MPa H₂, CZ-HP and CZA-HP catalysts showed activity loss after four runs, with CZ-HP losing 45% activity in six cycles (36 h total), while CZA-HP only lost 10%. ICP analysis indicated negligible Cu leaching, suggesting that Cu particle agglomeration is the primary cause of deactivation. Despite these observations, the catalysts’ potential for industrial applications must be carefully evaluated. CZ-HP’s significant deactivation presents a challenge for large-scale production, as frequent catalyst replacement or regeneration would be required, increasing operational costs. Conversely, CZA-HP’s lower deactivation rate makes it a more viable candidate for industrial processes, as its stability could translate into more cost-effective and sustainable operations. To make CZ-HP suitable for industrial use, improvements in catalyst formulation to prevent agglomeration and enhance stability are essential.

The acidity and metallic sites of the catalysts play a crucial role in glycerol hydrogenolysis and side reactions. Silica- and silicon-supported nanomaterials are particularly effective for the synthesis of acidic catalysts due to the high value of the silanol group on the surface, which enhances the surface acidity of the catalyst. In Kumar’s study [31], bimetallic Cu-Mg silica-supported catalysts with varying copper loadings were prepared using a chemisorption–hydrolysis method for the hydrogenolysis of glycerol to 1,2-PG. The Cu catalyst, promoted by a base, exhibited the highest activity, achieving a glycerol conversion rate of 89.5% and 1,2-PG selectivity of 92.1%. It was found that the hydrogenolysis of glycerol was highly dependent on the acidity and reduction properties of the Cu-Mg-supported silica catalyst. The high catalytic performance was attributed to the dispersion of copper, as well as the presence of different Cu species and acidic sites.

Copper metal has been investigated on various supports such as Cr₂O₃ [18], MgO [19], and SiO₂ [20] for liquid-phase glycerol hydrogenolysis. Ghalwadkar [32] et al. indicate that Cu catalysts supported on SiO₂ achieved the highest selectivity to EG at 15% during liquid-phase glycerol hydrogenolysis in a batch reactor, attributed to the promotion effect of 3% phosphorus ions in the support material. Bifunctional metal catalysts composed of acidic and/or basic sites are an effective approach to enhance catalytic activity and selectivity to 1,2-PG. Pudi et al. [21] designed a 35% Cu/MgO catalyst, achieving a maximum glycerol conversion rate of 96.6% and a selectivity of 92.6% for 1,2-PG at 210 °C and 4.5 MPa hydrogen pressure. The strong interaction between copper nanoparticles and the MgO support, along with high metal dispersion and the presence of strong acidic and basic sites, were identified as the primary reasons for the high activity and selectivity towards 1,2-PG. The Cu particles, predominantly spherical and uniformly distributed on the MgO support, had an average particle size of 17.5 nm (Figure 7a,b). Post-reuse, the catalyst maintained a similar particle size (Figure 7c,d). The XRD pattern of pure MgO (Figure 7e) exhibited diffraction peaks at 2θ = 36.9°, 42.9°, and 62.2°, corresponding to the typical cubic MgO phase. The introduction of CuO resulted in additional peaks at 2θ = 35.5°, 38.8°, 48.6°, 58.3°, 66.1°, and 68.0°, indicating the presence of the CuO phase. These CuO peaks intensified with increasing copper loading, suggesting that at low copper loadings, CuO exists as small particles, while at higher loadings, CuO tends to agglomerate. Figure 7f demonstrates that the peak intensity corresponding to CuO increases with the copper metal loading in the catalyst.

In the study of 35%Cu/MgO catalysts for glycerol hydrogenolysis, catalyst deactivation and reusability were evaluated through recycle experiments conducted under consistent reaction conditions. Over the course of three recycling cycles, the glycerol conversion rate showed a slight decrease from 96.6% to 89%, with a total reduction of approximately 7.5% after the third cycle. Despite this decrease, the selectivity for 1,2-PG remained high at around 93%, with ethylene glycol and degradation products accounting for approximately 4% and 3%, respectively. The yield of 1,2-PG decreased marginally from 89.4% (fresh catalyst) to 83.3% (third reuse). Characterization of the used catalysts via BET, H₂-chemisorption, XRD, ICP-MS, and TEM revealed that the BET surface area decreased from 26 to 19 m²/g, indicative of increased copper particle size due to agglomeration from repetitive reduction cycles. The minimal deactivation observed is attributed to the intrinsic properties of the Cu metal surface and its acidity, which play a crucial role in maintaining high selectivity for 1,2-PG. For industrial applications, the relatively low deactivation rate and stable selectivity make these catalysts promising. However, addressing the increase in copper particle size and optimizing catalyst stability and recovery will be important for scaling up these catalysts in large-scale processes.

Based on the obtained results, the following conclusions can be drawn: Cu-based catalysts exhibit strong capability in cleaving C-O bonds while showing weaker activity towards C-C bond cleavage. This results in high selectivity to 1,2-PG and relatively low selectivity to EG. The interaction between metal oxides and supports enhances the catalyst’s acid–base properties and stability, making certain Cu catalysts containing multi-metal oxides more effective and attractive. For reactions utilizing Cu-based catalysts, high reaction temperatures, appropriate low pressures, increased metal loading, and the selection of suitable supports and promoters ultimately enhance glycerol conversion rates and selectivity to ethylene glycol.

3.1.2. Ni-Based Catalyst

Similar to Cu catalysts, Ni catalysts primarily target the -OH group on the terminal carbon of glycerol, promoting its hydrogenation to predominantly produce 1,2-PG. Ni, characterized by high electron density, can participate in both C-O and C-C bond cleavage reactions. The formation of products in the presence of inert supports mainly depends on the structural properties of the Ni metal phase. Raney Ni (also called spongy Ni or skeletal Ni) is a fine-grained solid powder that is constructed mainly by nickel. This catalyst is normally derived from a Ni–Al alloy through superbase etching. Due to its high activity and unique property of storing hydrogen within the pores of the generated catalyst during the alkaline leaching process, it has been frequently used for reactions that involve hydrogenation and various hydrofining processes [33]. RANEY^® Ni, under optimal conditions (200 °C, 80 bar), demonstrates a notable glycerol conversion rate of 70% and a selectivity to EG of up to 40% [34].

In the glycerol conversion process, bimetallic effects play a crucial role in the selective cleavage of C-O and C-C bonds. Bimetallic systems can enhance catalyst selectivity by creating specific active sites. The coordination environment and surface energy of the metals in the alloys can modify the adsorption and dissociation characteristics of reaction intermediates, thus influencing reaction pathways and product selectivity. Chen et al. [35] found that the addition of Ag to Ni-Raney led to the formation of new Ni-Ag active sites, which suppressed C-C bond cleavage reactions, thereby increasing the selectivity to 1,2-PG to 73%. The formation of the Ni-Ag alloy improved the hydrogenation capability of the Ni catalyst by reducing the adsorption strength of the substrate and increasing hydrogen adsorption strength on Ni sites, while inhibiting its dehydrogenation capability [36]. Among various nickel catalysts supported on different zeolites, silica, and alumina, Ni supported on NaX zeolite demonstrated the highest performance due to its acidity. Specifically, Ni catalysts on NaMOR and NaZSM-5 achieved glycerol conversions of 14% and 47.8%, respectively, while supports like NaA and SiO₂ provided moderate conversions of 65.3% and 57%. Despite Ni/γ-Al₂O₃ exhibiting the highest glycerol conversion rate of 97%, its selectivity for 1,2-PG was 44.2%, which is lower compared to the 1,2-PG selectivity of 72% observed with Ni/NaX at a glycerol conversion of 94.5%.

The synergistic effect between bimetallic components and supports is influenced by catalyst preparation methods and pretreatments, which in turn affect catalyst performance. Pandhare et al. [37] reported a 20 wt% Cu-Ni (1:1)/γ-Al₂O₃ catalyst achieving a glycerol conversion rate of 98.5% and a 1,2-PG yield of 88% at 220 °C and 0.75 MPa, attributed to the synergistic interactions between Cu-Ni metals and the γ-Al₂O₃ support. The choice of support significantly affects the performance of Ni-based catalysts. For example, Seretis and Tsiakaras [38] reported a novel catalyst with 65% Ni doped into a silica (SiO₂)-modified alumina (Al₂O₃) support. Their study demonstrated excellent glycerol conversion rates exceeding 90% in a batch reactor, and under optimized conditions, the catalyst significantly improved selectivity to EG to 40% [39]. Ni/Al₂O₃–CuCr₂O₄ composite catalyst presents a relatively good performance at low H₂ pressure (14–17 bar) due to the interaction between metal and support as well as the modest hydrogenolytic activity towards C–C bonds.

Overall, Ni-based catalysts exhibit lower activity compared to Cu-based catalysts, primarily due to their greater ability to cleave C-C bonds, which results in lower selectivity towards 1,2-PG. However, by optimizing the alloy composition and support properties of Ni catalysts, catalytic performance can be effectively improved, leading to enhanced glycerol conversion rates and increased product selectivity.

3.2. Noble-Metal Catalysts

3.2.1. Ru-Based Catalyst

Ruthenium-based (Ru) catalysts are widely used in the hydrogenolysis of glycerol. Their performance is primarily influenced by Ru particle size, dispersion, and the acidity or basicity of the environment. Under mildly acidic or basic conditions, highly dispersed smaller Ru particles favor the formation of 1,2-PG. However, strongly acidic or basic sites are detrimental to 1,2-PG production as they tend to promote the cleavage of glycerol’s C-C bonds, increasing EG formation. Compared to Cu-based catalysts, Ru catalysts exhibit lower activity for hydrogenating C-O bonds. Maris et al. compared Ru/C and Pt/C catalysts, concluding that Ru promotes C-C bond cleavage reactions, thereby favoring EG formation [24]. The addition of basic promoters (e.g., NaOH and CaO) significantly enhances the glycerol hydrogenolysis rate for Pt/C catalysts but has a minor effect on Ru/C. It is suggested that C-C bond cleavage primarily occurs through metal-catalyzed reactions on Ru, whereas, in the case of Pt, it occurs through base-catalyzed reactions [24].

Wang et al. [25] demonstrated that Ru/ZrO₂ catalysts are the most active compared to Rh, Pt, and Pd catalysts. For catalysts with particle sizes around 2 nm, the activity order is Ru > Rh > Pt > Pd. The tendency for C-O bond cleavage follows the order Ru < Rh < Pt < Pd, whereas the tendency for C-C bond cleavage is the opposite. Ru is more prone to catalyzing extensive C-C bond cleavage, leading to EG formation, while Pd shows the highest selectivity to 1,2-PG formation through C-O bond cleavage. Larger Ru particles exhibit higher glycerol hydrogenolysis activity and higher selectivity towards EG and methane [25]. Figure 8a,b show that the catalysts have similar average metal particle sizes (~2 nm) with narrow distributions. The Ru particle size affects glycerol hydrogenolysis selectivity due to the need for C–C bond cleavage during glyceraldehyde decarbonylation. As seen in Figure 8c, increasing the Ru particle size from 1.8 nm to 4.5 nm reduces 1,2-PG selectivity from 55.8% to 30.6%, while EG selectivity rises from 19.1% to 27.3%. Larger Ru particles favor deeper hydrogenolysis. Figure 8d illustrates that 1,2-PG selectivity increases from 47.2% to 56.5% between 453–483 K under Ru/ZrO₂ catalysis but decreases to 51.1% at 513 K. In the temperature range of 453–483 K, the apparent activation energies calculated for Ru/ZrO₂ and Pt/m-ZrO₂ are 72 and 84 kJ/mol, respectively, indicating superior dehydrogenation capability of Ru over Pt.

The increase in acidity enhances the conversion rate of glycerol over Ru-based catalysts; however, it also decreases the selectivity to specific products such as 1,2-PG and EG while increasing the selectivity to CH₄. Excessive acidity leads to excessive C-C bond cleavage, resulting in methane production rather than the desired diols [22]. Ru-based catalysts exhibit significant catalytic activity in the hydrogenolysis of glycerol to 1,2-PG and EG, with conversion rates influenced by metal particle size and support acidity. Good metal dispersion and moderate support acidity favor higher conversion rates. Therefore, when using Ru-based catalysts, it is essential to optimize metal particle size and support acidity to achieve high conversion rates, minimize by-product formation such as methane, and enhance selectivity to the target products.

3.2.2. Pt-Based Catalyst

Platinum-based (Pt) catalysts are less studied in the hydrogenolysis of glycerol. Under neutral pH conditions, Ru is more active than Pt in converting glycerol to ethylene glycol. However, Ru tends to produce more ethylene glycol and also catalyzes methane formation, whereas Pt, despite its lower activity compared to Ru, selectively catalyzes the formation of 1,2-PG. Maris et al. [24] reported that Pt catalysts have higher selectivity to 1,2-PG production compared to Ru catalysts due to their lower C-C bond cleavage capability. The addition of basic promoters (e.g., NaOH and CaO) significantly enhances the glycerol hydrogenolysis rate for Pt/C catalysts. It is suggested that C-C bond cleavage on Pt primarily occurs through base-catalyzed reactions [24].

Yuan et al. [27] investigated Pt catalysts prepared on various supports and found that the activity of supported Pt catalysts for glycerol hydrogenolysis to 1,2-PG depends on the basicity of the support and the particle size of Pt. The higher the basicity of the support, the higher the conversion rate and selectivity to 1,2-PG, with the activity order: HTL > MgO > Al₂O₃ > Hβ > H-ZSM5. Some researchers have found that acidic supports lead to high levels of activity. Checa et al. [28] reported the activity order of Pt catalysts supported on CeO₂, A_l2O₃, La₂O₃, and ZnO as La₂O₃ > CeO₂ > ZnO > Al₂O₃, attributing the increased catalytic activity to the increase in surface acidity of the supports. In this study, the stability of Pt-based catalysts was evaluated under reaction conditions. The Pt/CeO₂-400 and Pt/La₂O₃-400 catalysts showed a significant reduction in surface area, approximately 25%, after 15 h of reaction. This suggests that the porous structures of the CeO₂ and La₂O₃ supports may degrade over time or become blocked by adsorbed species. On the other hand, the Pt/ZnO-400 catalyst did not experience any loss in surface area, indicating that the ZnO support remains stable under these conditions. Further analysis using ICP-MS of the liquid phase revealed that the amount of Pt leached into the liquid was less than 0.01%, indicating that the Pt metal particles are quite stable. However, the extent of support lixiviation varied: ZnO had a lixiviation percentage of 7.8%, while La₂O₃ and CeO₂ had higher percentages at 28.4% and 29.5%, respectively. This data highlights that while the Pt metal remains stable, the stability of the support materials varies significantly, with ZnO showing superior resistance to lixiviation comparead to CeO₂ and La₂O₃. To ensure the practical application of these catalysts in large-scale processes, it is essential to consider the support material’s stability, as it can significantly impact the overall performance and longevity of the catalyst. Studies on Pt catalysts supported on both basic and acidic supports show that both types can be used to prepare selective Pt catalysts. According to the dehydrogenation–dehydration–hydrogenation mechanism, basic supports favor the dehydration of glyceraldehyde to 2-hydroxypropenal, while acidic supports favor the dehydration of glycerol to acetol. Overall, Pt catalysts are intrinsically less active than Ru catalysts but exhibit higher selectivity to 1,2-PG. The choice of support plays a crucial role in the catalytic process, as its acidity or basicity can facilitate metal dispersion and enhance catalytic activity.

In summary, the catalytic hydrogenolysis of glycerol to 1,2-PG follows different reaction pathways depending on the fundamental properties of the catalytic system. Cu-based catalysts exhibit high selectivity to 1,2-PG due to their ability to cleave C-O bonds, while showing lower selectivity to EG. In contrast, Ni-based catalysts, although less active than Cu, significantly increase selectivity to EG to 40% when 65% Ni is doped into a SiO₂-modified Al₂O₃ support. The performance of Ru-based catalysts is influenced by particle size, dispersion, acidity, and basicity. Under mildly acidic or basic conditions, highly dispersed small Ru particles favor the formation of 1,2-PG, whereas strong acidic or basic conditions tend to cleave C-C bonds in glycerol, producing EG. Ru/C catalysts exhibit high selectivity to EG, up to 68%. For particles around 2 nm, the activity order is Ru > Rh > Pt > Pd [25]. The tendency for C-O bond cleavage follows Ru < Rh < Pt < Pd, while the tendency for C-C bond cleavage is the opposite. Among noble metal catalysts, Ru shows a high tendency for C-C bond cleavage, favoring EG formation, whereas Pd favors C-O bond cleavage, demonstrating high selectivity towards 1,2-PG. While Pt catalysts are less active than Ru, they exhibit higher selectivity for 1,2-PG. To selectively produce EG and 1,2-PG, it is crucial to control the catalytic hydrogenolysis of glycerol to target the cleavage of C-C and C-O bonds. Precise control over the properties of bifunctional catalysts, including the hydrogenation/dehydrogenation capabilities of the metal species and the acidic/basic functionality of the support, is essential. In addition to selecting appropriate metals and acidic/basic solid oxides as key components, efficient preparation methods and precise modulation of catalysts using modifiers or additives are also critical.

4. Reaction Path of Cellulose Preparation of Dihydric Alcohol

Cellulose has a robust crystalline structure formed by strong hydrogen bonds between equatorial hydroxyl groups and axial stacking between cellulose molecules, making it insoluble in water and most organic solvents, as well as resistant to mild chemical corrosion [40,41,42]. When isolating technical cellulose, standard methods include chemical pulping (e.g., Kraft or sulfite processes), mechanical pulping, and more sustainable approaches such as organosolv or enzymatic methods [43]. During the isolation process, various contaminants such as lignin, hemicellulose, extractives, or residual chemicals from the pulping process can remain. These impurities can affect the physical and chemical behavior of cellulose during subsequent reactions or processing [44]. For example, residual lignin may interfere with catalytic activity or inhibit specific chemical reactions, while hemicellulose can affect thermal stability and solubility. Higher molecular weights or a broad polydispersity index (PI) can limit the solubility of cellulose, making dissolution more challenging. Longer chains with higher molecular weights often exhibit lower reactivity, whereas a higher PI indicates greater variability in chain length, leading to inconsistent reaction behavior [45]. Polymer properties influence mechanical handling, such as the ease of forming films or fibers from the regenerated cellulose, and these characteristics may directly impact the experimental results.

The hydrolytic hydrogenation of cellulose is a multi-step process that combines cellulose hydrolysis with the hydrogenation of glucose or fructose. This reaction involves both the breakdown of cellulose and subsequent hydrogenation of the sugar intermediates, primarily yielding low-carbon polyols such as 1,2-PG, EG, and glycerol [46]. As shown in Figure 9, the process begins with the hydrolysis of cellulose at the acidic sites of the catalyst, producing glucose. The glucose then undergoes retro-aldol condensation (RAC), which cleaves the C-C bonds to form glycolaldehyde (GA). GA is subsequently hydrogenated at the metal active sites of the catalyst to produce EG. Simultaneously, glucose can isomerize into fructose, which also undergoes RAC and hydrogenation to yield 1,2-PG [47]. The RAC reaction, which involves C-C bond cleavage, primarily occurs under basic conditions or in the presence of transition metal catalysts [48]. The activation energy for RAC (approximately 160 kJ/mol) is higher than that for sugar isomerization (about 140 kJ/mol) and sugar hydrogenation (35–65 kJ/mol) [49,50].

Studies have shown that sugars like glucose, fructose, and xylose can degrade into small C2-C4 molecules through RAC or undergo hydrogenation to form inert C6-C4 polyols, which are more challenging to convert into EG and 1,2-PG [51,52]. Hydrogenation thus plays a dual role in this process: it converts intermediates such as GA and acetol into EG and 1,2-PG, but can also transform sugars into polyols that resist further degradation, reducing the yield of the target diols. Transition metals like tungsten (W), tin (Sn), and others provide essential hydrogenation sites to facilitate cellulose and sugar conversion [52,53]. However, the efficiency and selectivity of the reaction depend heavily on the composition of the catalytic system. A balance must be maintained between the rates of hydrogenation and RAC to prevent the formation of inert polyols [54]. Additionally, selective dehydrogenation by transition metal catalysts can convert hexitols into sugar isomers, which can undergo RAC to produce C2-C4 small molecules [55].

The one-pot catalytic conversion of cellulose, aimed at the preparation of EG and 1, 2-PG. This catalytic conversion involves the hydrolysis of cellulose, retro-aldol condensation (RAC) of glucose, and subsequent hydrogenation of intermediates [56]. The selectivity and efficiency of the entire reaction largely depend on the interplay and rate matching between the individual steps. Each reaction occurs at different active sites, and effective coupling promotes the conversion of cellulose into high-value EG and 1,2-PG [57]. The process of producing EG and 1,2-PG from the hydrolysis and hydrogenolysis of cellulose involves optimizing various catalysts and reaction steps. Through the rational design and regulation of the catalyst active sites, efficient conversion of cellulose to these important chemicals can be achieved.

5. Selective Regulation of Cellulose Preparation Diols

The one-pot process of converting cellulose to EG and 1,2-PG involves three main consecutive reaction steps. First, cellulose undergoes hydrolysis to produce glucose. Next, glucose is converted to glycolaldehyde (GA) through a retro-aldol condensation (RAC) reaction [58,59,60]. Finally, GA is hydrogenated to form EG. The formation of 1,2-PG involves the isomerization of glucose to fructose before undergoing RAC and hydrogenation reactions. Therefore, an ideal catalyst must possess multiple active sites, including acidic sites to facilitate cellulose hydrolysis and catalytic sites to promote RAC and GA hydrogenation reactions [59]. The main challenge in the direct conversion of cellulose to high yields of EG and 1,2-PG lies in the selective cleavage of C-C bonds in C6 sugar molecules to form C2 or C3 intermediates. Hence, the RAC reaction is considered the pivotal step in converting cellulose to small-molecule diols, and various efficient catalysts have been designed specifically for this step.

In the RAC reaction, transition metal catalysts exhibit significant advantages. These transition metals, primarily from Groups 3 to 6 and Group 14 of the periodic table, possess multiple valence states ranging from +2 to +6 [61]. The formation of metal–sugar complexes and carbon rearrangement processes determine the catalytic performance of different valence states of transition metals. Generally, the affinity increases with monovalent, divalent, and trivalent metals, thereby exhibiting varying effects in catalyzing the RAC reaction [49]. Compared to basic catalysts, transition metal catalysts offer distinct advantages in the cellulose conversion process. They enhance the Lewis acidity of the reaction system, which improves the hydrolysis efficiency of cellulose. This enhancement facilitates the conversion of cellulose into monosaccharides, thus increasing both the overall reaction rate and efficiency [61,62]. Transition metal catalysts can maintain stability in acidic reaction media, allowing the introduction of additional acidic substances (such as sulfuric acid) to further improve cellulose conversion rates without significantly reducing selectivity to EG [63,64]. This is unachievable with basic catalysts, which tend to deactivate in acidic environments. Lastly, transition metal catalysts exhibit high selectivity to the RAC reaction, avoiding the promotion of sugar isomerization reactions and thus enabling high selectivity to EG while reducing the formation of by-products like 1,2-PG.

The conversion of glucose to ethylene glycol involves a multi-step process with two primary types of chemical reactions: retro-aldol condensation (RAC) and hydrogenation. Some studies indicate that the RAC reaction, which breaks C-C bonds in sugars to produce low-carbon products, primarily occurs at the acidic sites provided by Group VIB element catalysts. In contrast, the hydrogenation process, which converts unsaturated aldehydes to alcohols, occurs at the hydrogenation sites provided by Group VIII active metals [61]. Therefore, bifunctional catalysts are commonly used for glucose conversion to EG, comprising Group VIII active metals, including noble metals (such as Ru [54], Pd) and non-noble metals (such as Ni [51]), and Group VIB elements (such as W, Mo, Cr [57]), with W being the most widely used element. Besides the pathway leading to EG, glucose can also isomerize to fructose, which theses RAC and hydrogenation reactions to form 1,2-PG. The RAC reaction involving C-C bond cleavage mainly occurs under basic conditions or the catalytic action of transition metal components.

Basic catalysts play a crucial role in the retro-aldol condensation (RAC) reaction of sugars by cleaving the C-C bonds within sugar molecules. Theoretically, all alkali metals and alkaline earth elements can provide basic active sites, thereby facilitating the RAC reaction. Common basic catalysts include hydroxides, metal oxides, strong base–weak acid salts, and basic ammonium salts, ammonium hydroxide, and nitrogen-containing basic organics [65]. In the process of converting cellulose or sugars into 1,2-PG and EG, the combination of basic catalysts with metal catalysts is widely employed [66,67,68]. These basic catalysts provide active sites that promote the RAC reaction and other related reactions. The type of alkali metal cation significantly affects the performance of the RAC reaction. Different cations can regulate reaction conditions and selectivity, thereby optimizing product yields [52,57,68]. For instance, in the conversion of glucose to 1,2-PG, Ca(OH)₂ performs better than other basic catalysts such as Ba(OH)₂, NaOH, and KOH under similar pH conditions. Additionally, perovskite basic catalysts containing Ba (e.g., Pt/BaZrO₃) outperform those containing Ca and Sr [67]. This indicates that selecting appropriate alkali metal cations is crucial for optimizing RAC reaction performance and significantly enhancing the efficiency of glucose conversion to 1,2-PG.

Another important role of basic catalysts is to facilitate the isomerization of aldoses to ketoses, known as the Lobry de Bruyn–van Ekenstein transformation [69]. This reaction proceeds via a proton transfer mechanism through an enediol intermediate, catalyzed by hydroxide groups. Consequently, when the catalyst possesses basic active sites, the main product is 1,2-PG rather than EG. Liu et al. [70] reported that introducing an additional basic catalyst (C_act) into a WO₃ + Ru/C composite catalyst decreased the yield of EG from 51.5% to 24.4%, while increasing the yield of 1,2-PG from 6.7% to 31.9%. This ability to promote sugar isomerization and catalyze RAC can be used to regulate the selectivity of the main product towards 1,2-PG instead of EG.

In 2008, Zhang et al. [51] first reported the direct conversion of lignocellulose to ethylene glycol. They used a tungsten carbide catalyst, achieving an EG yield of 27%. The EG yield further increased to 61% with the addition of a small amount of nickel. By improving the catalyst preparation methods or using novel catalyst supports, the EG yield could be further enhanced to 72–76%. Additionally, novel catalysts such as tungsten phosphide [52], bimetallic tungsten-based catalysts [53], and tungsten oxide-based catalysts [54] have been widely used for the hydrolytic hydrogenation of cellulose to produce EG and 1,2-PG. Moreover, catalytic systems such as Ru/C-NaOH [55], CuO/ZnO/Al₂O₃ [56], and Cu/CrO_x-Ca(OH)₂ [57] have led to variations in reaction products, predominantly yielding 1,2-PG instead of EG. The incorporation of basic co-catalysts in these systems facilitates the isomerization of glucose to fructose. The following table summarizes the yields of diols achieved with various catalysts. These studies illustrate that by choosing appropriate catalysts and optimizing reaction conditions, the efficiency and selectivity of cellulose hydrolysis and hydrogenolysis to produce EG and 1,2-PG can be significantly enhanced.

Table 2. The reaction performance of EG and 1,2-PG was prepared by different catalytic systems of cellulose.

Catalyst	Reaction Conditions	Conversion/%	S EG/%	S1,2-pG/%	Ref.
W2C/AC	cellulose, 518 K, 6 MPa H2, 0.5 h	98	27.9	5.6	[51]
W2C/AC		100	61	7.6	[51]
10Ni-(30WCx/AC)		100	73.0	8.5	[71]
Ni-WP/AC		100	46.0	64	[52]
NiS-W15//SBA-15		100	76.1	3.2	[72]
Ru/AC-H2WO4		100	58.5	3.5	[46]
Raney Ni-H2WO4		100	65.4	3.3	[54]
Ru/AC-WO3		23.4	51.5	6.7	[73]
Ru/C-WO3+ Cact		22.8	24.4	31.9	[73]
NiW/MIL-125	cellulose, 518 K, 6 MPa H2, 2 h	100	68.7	6.5	[74]
Ru/NC@void@MC-SO3H	cellulose, 503 K, 6 MPa H2	97		38	[75]
Co/CeOx	cellulose, 518 K, 6 MPa H2, 6 h	97.3	55.2	33.9	[76]
10% Ni-0.5%Ir/La2O3	cellulose, 518 K, 5 MPa H2	100	63.7		[77]
Fe3O4 @SiO2/10%Ru-20% WOx	cellulose, 518 K, 5 MPa H2, 2 h	96.8	19.71	31.4	[78]

displays the reaction performance for EG and 1,2-PG across different catalytic systems used for cellulose conversion.

5.1. Non-Noble Metal Catalyst

5.1.1. Tungsten-Based Catalyst

Tungsten-based catalysts are extensively studied for their high effectiveness in C-C bond cleavage, particularly in the selective disruption of C-C bonds at the β-position relative to the carbonyl group in glucose intermediates, following the retro-aldol condensation (RAC) mechanism. This process facilitates the formation of EG precursors. In the presence of metal catalysts, glycolaldehyde (GA) is selectively hydrogenated in situ to produce EG. Tungsten compounds are highly effective for achieving high EG yields due to their strong activity in promoting the selective cleavage of C-C bonds in cellulose. Tungsten provides acidic sites that enhance the RAC transformation, while nickel (Ni) is commonly employed for hydrogenation reactions. The synergistic interaction between Ni and W significantly boosts EG yield, with Ni facilitating the formation of tungsten carbide (W₂C).

In 2008, Zhang et al. [51] demonstrated that the combination of nickel and tungsten (Ni-W) catalysts exhibited exceptionally high selectivity towards EG. Compared to monometallic catalysts, these bimetallic catalysts markedly accelerated cellulose conversion and substantially increased selectivity to EG, particularly on Ni-W/SBA-15 catalysts. Fine-tuning the bimetallic content improved EG yield. It was observed that higher nickel content and elevated carbonization–hydrogenation reaction (CHR) temperatures led to significant W₂C sintering, which subsequently reduced the catalyst’s activity in cellulose degradation. The optimal ratio of W:Ni was found to be 3:1, achieving an EG yield of 76.1%. Balancing the bimetallic content facilitated the equilibrium between hydrogenation and RAC reaction rates, thereby enhancing EG production. As shown in Figure 10a, in the absence of nickel, well-crystallized and phase-pure W₂C formed only at 1073 K. Beyond this temperature, some W₂C further carburized to tungsten carbide (WC). Figure 10b illustrates that with the addition of nickel, phase-pure W₂C formed at 973 K. Increasing the CHR temperature to 1073 K intensified the W₂C diffraction peaks, indicating the formation of larger W₂C particles. The presence of nickel lowered the W₂C formation temperature by approximately 100 K, likely due to nickel-facilitated hydrogen dissociation, which reduced the tungsten precursor.

For tungsten-based catalysts, H_xWO₃ is the actual active species involved in the cleavage of C-C bonds in glucose. However, due to the removal of the catalyst support, it is difficult to restore H_xWO₃, which hinders the practical application of the catalyst [79]. Li et al. [74] obtained Ni-W/M catalysts by calcining Ni-W/MIL-125(Ti) precursors. This catalyst achieved a cellulose conversion rate of 100% and an EG yield of 68.7%, with a total polyol yield of 83.9%. The formation of Ti-O-W bonds in the catalyst can connect the tungsten components to the catalyst support, reducing the loss of active tungsten species during the reaction [74].

As shown in Figure 11a, the binding energies at 458.4 and 464.1 eV are characteristic peaks of Ti 2p_3/2 and Ti 2p_1/2 in pure TiO₂, respectively. In Figure 11b, the broadening of the O1s peak and increased binding energy suggest the formation of Ti-O-W bonds. The W 4f spectrum shows peaks at 35.8 eV and 38.0 eV, indicating W⁶⁺, while new peaks at 39.4 eV and 41.6 eV are attributed to Ti-O-W bonds, indicating tungsten–titanate surface species that enhance catalyst stability. Figure 11f reveals that the Ni 2p spectrum shows peaks at 851.9 eV and 869.1 eV, corresponding to Ni0. Overall, the formation of Ti-O-W bonds improves the stability and durability of the catalyst. Under the optimal experimental conditions (20 wt % W and Ni loading, calcination temperature of 773 K, reaction temperature of 518 K, reaction time of 2 h, and initial H₂ pressure of 4 MPa), seven recycling tests were performed. The results indicated that the cellulose conversion rate and EG yield remained stable after each reaction, demonstrating that the catalyst maintained good reactivity with negligible metal leaching and no significant deactivation. The minimal metal leaching observed confirms that the metal components of the catalyst were securely bound and did not migrate into the reaction medium, which is critical for sustaining catalytic efficiency and preventing product contamination. Additionally, the lack of significant deactivation implies that the catalyst’s active sites were not adversely impacted by the reaction conditions or repeated use. These findings underscore the robustness of the catalyst under the tested conditions, making it a promising candidate for large-scale industrial applications where consistent performance and minimal maintenance are essential. Further investigation into the long-term stability and the potential for catalyst regeneration could provide additional insights into its practical applicability in continuous or high-volume processes.

Liu et al. [78] reported the actual mechanism of WO₃ interaction: a tridentate complex is formed through coordination of carbonyl and α-hydroxyl and β-hydroxyl groups with two adjacent tungsten atoms (i.e., W-O-W sites), subsequently triggering the cleavage of the C-C bond at the β position to the carbonyl group, forming C2,3 fragments. Crystalline WO₃ and H₂WO₄, in combination with a hydrogenation catalyst under a hydrogen atmosphere, can effectively and selectively cleave the C-C bonds of sugar intermediates involved in the conversion of cellulose to ethylene glycol and 1,2-PG. The actual state of W-based catalysts during the reaction process remains undiscovered.

Liu et al. [70] demonstrated that both WO₃ and H₂WO₄ tend to be reduced to hydrogen tungsten bronze (H_xWO₃) species during cellulose reactions, comprising H_0.23WO₃ and H_0.33WO₃, which are then re-oxidized to WO₃ upon exposure to ambient air after the reaction. As shown in Figure 12a,b, fresh WO₃ samples show distinct peaks at 23.1° (002), 23.6° (020), and 24.3° (200). After 10 min of cellulose reaction, these peaks fade, and H₀.₂₃WO₃ peaks appear, reaching maximum intensity at 30 min. H₀.₃₃WO₃ peaks then emerge, and both are re-oxidized to WO₃ upon air exposure. In Figure 12c, the UV-Vis spectrum of WO₃ mixed with Ru/C shows strong absorption below 400 nm. Reflectance above 500 nm decreases over 24 h, indicating the formation of H_xWO₃.

5.1.2. Sn-Based Catalysts

Transition metal Sn is a versatile active component in the catalytic conversion of sugars, facilitating various reactions, including glucose isomerization to fructose, sugar retro-aldol condensation (RAC) reactions to glycolaldehyde (GA), and glucose epimerization to mannose. Sn enhances glucose isomerization, leading to lower selectivity for ethylene glycol (EG) and higher selectivity for 1,2-propanediol (1,2-PG). The catalytic effectiveness of Sn in glucose retro-aldol condensation is affected by its valence state, alloy state, and chemical environment. Sun et al. [80] combined Ni/AC with Sn powder, adjusting the Sn valence state to alter the product distribution (EG and 1,2-PG ratios) during cellulose hydrolytic hydrogenation. SnO exhibits activity in RAC and also promotes glucose isomerization to fructose, whereas alloyed Sn in NiSn alloy catalyzes only retro-aldol condensation. Therefore, different Sn species can regulate the EG and 1,2-PG product ratio.

On Ni-Sn/AC, two metal sites were identified: alloyed Sn for glucose retro-aldol condensation to glycolaldehyde and non-alloyed Ni for hydrogenating glycolaldehyde to EG. The Sn loading controls the balance between retro-aldol condensation and subsequent hydrogenation reactions. Ni/AC and Sn powder combination showed excellent activity for EG (57.6%), with a total polyol yield of 86.6%, while Ni/AC and SnO combination favored 1,2-PG (32.2%) with an EG yield of 22.9%. Sn catalysts demonstrate multifunctional capabilities in sugar conversion. Using Sn(II) or alloyed Sn catalysts modulates the product distribution between EG and 1,2-PG. The XRD of Ni–Sn (90)/AC (Figure 13a) shows narrow Ni peaks and weak NiSn peaks, indicating large Ni and small NiSn particles coexist on AC. Figure 13b,c reveal no significant NiSn peaks on ZnO, TiO₂, SiO₂, MgO, or Al₂O₃. Ni and NiSn crystallite sizes are 7.2 nm and 17.2 nm. Ni crystallite size increases sixfold without Sn, but only doubles with Sn, suggesting Sn addition delays Ni sintering. Ni particles are well-dispersed with an average size of 9.7 ± 1.8 nm, compared to 20.6 ± 6.1 nm in metal particles (Figure 13d).

5.2. Noble-Metal Catalysts

Hydrogenation reactions can be performed using catalysts containing noble metals such as Pd, Pt, Rh, and Ru, which have been demonstrated to effectively facilitate cellulose hydrolytic hydrogenation. These noble metal-containing catalysts are commonly used in conjunction with other catalysts that catalyze cellulose hydrolysis and glucose isomerization, forming multifunctional catalysts that efficiently convert cellulose to ethylene glycol (EG) and 1,2-propanediol (1,2-PG). Carbon-based catalysts have received significant attention due to their large surface area, numerous surface acidic groups, and ease of surface modification. Compared to solid acid catalysts, carbon-based catalysts feature carboxyl (−COOH), phenol (−OH), and −SO₃H groups, which form hydrogen bonds with −OH groups, enabling strong adsorption of cellulose onto the catalyst surface.

However, multifunctional catalysts with metal nanoparticles encased by acidic groups may suffer from metal poisoning, leading to reduced product selectivity. Our research group [75] synthesized a yolk–shell structured catalyst, Ru/NC@void@MC-SO₃H, where Ru clusters (1.4 nm) and SO₃H groups are precisely located on the inner shell and outer shell, respectively. This unique core@void@shell structured yolk–shell catalyst functions as a bifunctional one-pot cascade reaction nanoreactor. Unlike supported catalysts, the internal core provides extensive exposed active sites for close interaction with reactants; the cavity facilitates rapid diffusion of reactants/products; and the permeable shell offers protection under harsh reaction conditions, preventing agglomeration, leaching, and damage.

The spatial separation between metal and acid sites not only reduces site poisoning and optimizes acidity but also enhances the metal–acid synergy in hydrolysis, isomerization/hydrogenation, and hydrogenation cascade reactions, providing designated diffusion pathways for reactants [75]. Ru clusters and SO₃H groups are precisely located on the inner core and shell, respectively. The successful synthesis of site-separated Ru/NC@void@MC-SO₃H is attributed to discrete steps for core and shell formation and the protective sacrificial silica layer covering the core [75]. This catalyst achieved a 38% yield of 1,2-PG, further elucidating the yolk–shell catalyst’s mechanism and reaction pathways in cellulose hydrogenolysis, confirming that Lewis acid/base sites in Ru-Nx accelerate glucose isomerization to fructose, providing designated pathways for target product formation. Figure 14A illustrates the synthesis process of Ru/NC@void@MC-SO₃H. Figure 14B shows the possible reaction pathways for cellulose hydrogenolysis to 1,2-PG. Figure 14C presents the TEM and STEM images and elemental maps of the catalyst, indicating the precise location of Ru clusters and SO₃H groups on the inner core and shell, respectively. The successful synthesis of site-separated Ru/NC@void@MC-SO₃H is due to discrete steps for forming the core and shell layers and the protective effect of the sacrificial silica layer covering the core. Our research group initially screened the site-isolated Ru/NC@void@MC-SO₃H catalyst under reaction conditions of 0.20 g cellulose, 0.06 g catalyst, 20 mL H₂O, 6 MPa H₂, 230 °C, and 5 h. After four consecutive runs, there was no significant decrease in both cellulose conversion and 1,2-PG yield. Compared to the first reaction, the yield of 1,2-PG reduced to 28.0%. TEM images of the spent Ru/NC@void@MC-SO₃H catalyst still displayed an intact yolk–shell structure without any metal aggregation, indicating the exceptional stability of the catalyst under harsh reaction conditions. The stability observed in our recycling tests is promising for industrial applications, as consistent performance is crucial for efficient and cost-effective large-scale operations. However, scaling up these processes involves addressing several practical challenges. For instance, ensuring uniform catalyst performance across large reactors and handling the catalyst’s longevity and regeneration are critical. Additionally, the economic feasibility of catalyst materials and the integration of these processes into existing industrial systems must be considered. Moreover, the high pressure and temperature conditions used in the experiments may pose safety and equipment durability concerns at an industrial scale. Therefore, further studies and pilot-scale tests are needed to validate the catalyst’s performance under real-world conditions, optimize reactor design, and develop strategies for maintaining catalyst activity and longevity in a continuous production environment.

The choice of catalyst support is crucial as it plays a key role in the dispersion, accessibility, and stability of active sites. Ji et al. [71] reported that when three-dimensional (3D) mesoporous carbon (MC) was used as a support to disperse tungsten carbide, the yield of EG reached 75% or 73%, regardless of the promotion by transition metals. Zhang et al. [81] investigated the performance of Ru-W₁₈O₄₉/graphene, achieving an EG yield of 62.5% with almost complete cellulose conversion after 60 min at 245 °C. As illustrated in Figure 15a, Ru/graphene composite materials show uniformly distributed Ru nanoparticles on the surface, averaging about 3 nm in size. Figure 15b shows that WO₃ forms micron-sized flakes agglomerated on graphene nanosheets. In Figure 15c, Ru-W₁₈O₄₉/graphene composites display ultrathin graphene nanosheets, decorated with Ru nanoparticles and W₁₈O₄₉ nanowires. Figure 15d confirms the crystalline structure of W₁₈O₄₉ nanowires with a lattice fringe spacing of 0.37 nm, corresponding to the (010) plane. This Ru-W₁₈O₄₉/graphene combination enhances active site stability and interaction with reactants, optimizing catalytic performance for cellulose conversion.

5.3. Lanthanide Catalysts

Recent studies have demonstrated that rare-earth metal catalysts, such as those containing La(III) and Ce, can effectively promote the hydrolytic hydrogenation of cellulose, enabling its conversion to ethylene glycol (EG) and 1,2-propanediol (1,2-PG). Traditional metal–acid catalytic systems typically utilize acidic sites for hydrolysis, retro-aldol condensation (RAC), and isomerization, while hydrogenation occurs at the metal sites. However, these processes often encounter undesirable side reactions, where the acidic environment fosters unwanted dehydration and resinification of humins. To mitigate the adverse effects of acidic conditions, Fu’s research group [76] developed a novel weakly basic Co/CeO_x catalyst for cellulose conversion. This catalyst achieved yields of 55.2% for EG and 33.9% for 1,2-PG after 6 h at 245 °C. The high performance is primarily attributed to the Co^±O^x−Ce³⁺ base–acid electron pairs, which ensure a balance among the different reaction steps, including hydrolysis, RAC, isomerization, and hydrogenation. Previous catalysts containing acidic sites often promoted the formation of humin. In contrast, the newly developed basic catalyst inhibits humin formation and moderately slows cellulose hydrolysis, thereby reducing glucose accumulation and subsequent humin formation. Compared to acidic sites, basic sites facilitate the retro-aldol condensation reaction and prevent resinification, thus improving the yield. Figure 16a,b shows the non-in-situ XPS observation of oxidized Co species on the surface. The Co 2p spectra at 780 and 782 eV identify two peaks corresponding to dissociated Coⁿ⁺ and the interaction between Co and CeOⁿ⁺_x (with n = 2 and 3). The strong metal-support interaction (SMSI) between Co and CeO_x is confirmed by the binding energy changes of Co and Ce, indicating that reduced Ce³⁺ species provide electrons to adjacent Co species, resulting in electron-rich Co and electron-deficient Ce. This Co^n±O^x−Ce³⁺ base–acid electron pair constitutes the essential acidic and basic sites for catalytic cellulose conversion.

Sun et al. [77] reported a catalyst (10% Ni-0.5% Ir/La₂O₃) that achieved cellulose selectivity of 73.2%, with a combined yield of 63.7% for EG and 1,2-PG (1,2-PG). This catalyst (10% Ni-0.5% Ir/La₂O₃) exhibited high catalytic activity for cellulose degradation, with a turnover number (TON) of 339. This method combines RAC catalysts with heterogeneous hydrogenation catalysts to form a binary catalyst, enhancing the substrate-catalyst interaction while maintaining catalyst stability and recyclability.

Density functional theory (DFT) analysis indicated two reaction pathways for cellulose on lanthanum-based catalysts: (1) Similar to W-based and Sn-based catalyst mechanisms, glucose undergoes retro-aldol condensation (RAC) catalyzed by La(III) to cleave into C2 molecules, which are subsequently hydrogenated to form ethylene glycol (EG) and 1,2-propanediol (1,2-PG). (2) Alternatively, glucose can be directly hydrogenated into hexitols, which then undergo hydrogenolysis at the basic sites of La(III) oxide and the hydrogenation active sites of metallic nickel to produce EG and 1,2-PG. Figure 17a provides a detailed mechanistic explanation. Based on these results, the reaction can be divided into four stages: (i) complex formation between La(III)-OH and glucose followed by hydrogen transfer; (ii) C2-C3 bond cleavage via sequential glucose epimerization and 2,3-hydride shift reactions; (iii) loss of glyceraldehyde (GA); (iv) erythrose cleavage into two GA molecules. Figure 17b shows mass spectrometry (MS) and MS/MS results, indicating that the La(III)-glucose complex and glucose degradation fragments (i.e., GA and Erythritol) are formed by the cleavage of the C2-C3 bond in glucose. These findings confirm that glucose activation by La(III) species leads to selective catalytic degradation.

6. Conclusions and Outlook

The selection of suitable catalysts is crucial for the conversion of glycerol and cellulose into low-carbon alcohols. Basic, non-precious, and precious metal catalysts are essential for facilitating the retro-aldol condensation (RAC) of C6 sugars, thereby enhancing the conversion efficiency of cellulose into ethylene glycol (EG) and 1,2-propanediol (1,2-PG). Basic catalysts or Lewis acids that promote glucose isomerization favor the production of 1,2-PG, while catalysts lacking these active sites primarily produce EG. Catalysts with active sites for both hydrogenation and isomerization, such as Ru, W, and Sn, optimize sugar conversion under mild conditions. Transition metals like Cu and Ni, along with precious metals like Ru, play crucial roles in determining product selectivity. Enhancing catalyst durability and developing renewable hydrogen sources are key challenges. Future research should focus on improving catalyst performance and sustainability for large-scale applications.

In the catalytic hydrogenolysis of glycerol, controlling catalyst properties and reaction conditions is essential for precisely regulating selectivity towards EG and 1,2-PG. Transition metal catalysts such as Cu, Ni, and W are primarily responsible for glycerol conversion, whereas precious metal catalysts like Ru, Pt, and Pd play a critical role in selectivity towards EG. Cu-based catalysts, due to their strong C-O bond cleavage capability, exhibit high selectivity towards 1,2-PG but lower selectivity towards EG. In contrast, Ni-based catalysts, when appropriately modified, can significantly enhance selectivity towards EG. Ru-based catalysts demonstrate good 1,2-PG production capabilities under weak acid or weak base conditions, while under strong acid or strong base conditions, they tend to promote C-C bond cleavage in glycerol, yielding EG. Moreover, catalyst properties such as particle size, dispersion, and acid–base characteristics also influence product selectivity. For instance, highly dispersed small Ru particles favor the production of 1,2-PG. Overall, the active sites in the catalytic system play a decisive role in the cleavage of C-C and C-O bonds in glycerol molecules, directly affecting the primary product selectivity between EG and 1,2-PG. By precisely designing and controlling catalyst properties and reaction conditions, it is possible to regulate product selectivity in glycerol hydrogenolysis.

While existing catalysts demonstrate promising activity, many challenges remain regarding their efficiency and durability, particularly for large-scale industrial applications. Catalysts effective in small-scale reactions may suffer from rapid deactivation or insufficient recyclability in industrial processes. Therefore, the development of more robust and durable catalysts with higher efficiency under real industrial conditions is essential. Investigating their long-term stability and reusability will be critical for the practical implementation of these processes on a larger scale.

A deeper understanding of the reaction pathways and mechanisms is necessary to improve selectivity control. Different catalysts can promote various pathways, leading to different intermediate and final products. Thus, further studies combining experimental approaches with computational techniques such as Density Functional Theory (DFT) are needed to provide a comprehensive understanding of how catalysts influence the selectivity and efficiency of hydrogenolysis processes. The integration of renewable hydrogen sources into glycerol and cellulose hydrogenolysis remains a significant challenge. Many current studies rely on conventional hydrogen sources, which may limit the sustainability of these processes. Future research should focus on incorporating hydrogen derived from renewable sources, such as water electrolysis or biomass gasification, to enhance the overall environmental friendliness and sustainability of the production process. There is great potential for the exploration of novel catalytic materials with multifunctional properties. These catalysts not only contribute to hydrogenolysis reactions but also demonstrate synergistic effects in other steps of the conversion processes. Additionally, improving catalyst reusability remains a crucial factor for reducing costs and minimizing environmental impact, especially in large-scale industrial applications. Future research should also investigate the environmental and economic impacts of biomass-derived chemicals in industrial settings, ensuring that these green processes are scalable and commercially viable.