Conversion of Cellulose to γ-Valerolactone over Raney Ni Catalyst Using H₂O as a Hydrogen Source

Abstract

The sustainable valorization of lignocellulosic biomass into high-value platform chemicals presents a crucial pathway for reducing reliance on fossil resources. Gamma (γ)-valerolactone (GVL) has gained recognition as a versatile bio-derived compound with broad applications in renewable energy systems and green chemical synthesis. While conventional GVL production strategies from carbohydrate biomass typically depend on noble metal catalysts paired with high-pressure hydrogen gas, these approaches face substantial technical barriers including catalyst costs, hydrogen storage requirements, and operational safety concerns in large-scale applications. This work develops an innovative catalytic system utilizing earth-abundant iron for in situ hydrogen generation through water splitting, integrated with Raney Ni as the hydrogenation catalyst. The designed two-stage process enables direct conversion of cellulose—first through acid hydrolysis to levulinic acid (LA) followed by catalytic hydrogenation to GVL without intermediate purification. Through systematic parameter optimization, a remarkable 61.9% overall GVL yield from cellulose feedstock was achieved. Furthermore, the methodology’s versatility was demonstrated through wheat straw conversion experiments, yielding 24.6% GVL. This integrated methodology explores a technically feasible pathway for direct cellulose-to-GVL conversion utilizing abundant water as the hydrogen source, effectively overcoming the critical limitations associated with conventional hydrogenation technologies regarding hydrogen infrastructure and process safety.

1. Introduction

The overexploitation of fossil resources and heavy reliance on them have not only triggered energy depletion crises but also precipitated severe environmental degradation, including but not limited to global warming and acid rain deposition. A promising mitigation strategy involves developing utilization of biomass, which is the Earth’s most abundant renewable carbon source. However, biomass possesses relatively low energy density and significant geographical constraints have hindered its large-scale application. Therefore, to truly realize the potential of biomass as a substitute for fossil fuels, it is crucial to develop efficient biomass utilization technologies that can convert this low-energy-density resource into clean, value-added products, such as key platform compounds or biofuels [1,2,3].

Cost-effective catalytic strategies have been systematically developed to accommodate the structural complexity of highly functionalized biomass derivatives, aiming to render the resultant chemicals competitive in both quality and market price compared to petroleum-based analogues [4,5,6]. Over the past decade, γ-valerolactone (GVL) has emerged as a preeminent biomass-derived platform chemical, distinguished by its unique physicochemical profile: low melting point (−31 °C), elevated boiling temperature (207 °C), and favorable open-cup flash point (96 °C) [7]. These thermophysical characteristics, combined with its biodegradable nature and non-toxic properties, facilitate its global logistics in bulk quantities while ensuring storage safety. The versatility of GVL extends beyond its conventional roles as a food-grade additive and green aprotic solvent [8,9]. Recent advancements have demonstrated that GVL can be employed as an octane-enhancing fuel additive with superior oxygen content (32 wt%) [10,11], and a reactive intermediate for synthesizing value-added derivatives including linear alkanes [12], valeric esters [13,14], and dimethyl adipate [15] through cascade catalytic processes. Much attention focusses on direct acid-catalyzed conversion of C6 carbohydrates (e.g., glucose, fructose) and their derivatives (notably levulinic acid, LA) to GVL, where the synergy between Brønsted acid sites and hydrogenation catalysts significantly improves conversion rate and process economics [16,17,18].

Current GVL production methods predominantly utilize two hydrogenation approaches differentiated by hydrogen source: exogenous gaseous hydrogen versus liquid-phase hydrogen donors. The conventional gaseous hydrogen route requires high-pressure H₂ injection into sealed reactor vessels, entailing substantial safety risks associated with hydrogen storage, transportation, and high-pressure operations. This methodology necessitates noble metal catalysts (e.g., [Ru(CO)₃I₃]⁻, CuAg/Al₂O₃, Ru/C) to activate molecular hydrogen, as evidenced by prior studies [19,20,21,22]. However, developing cost-effective catalysts with sustained activity remains a critical challenge for practical applications. In contrast, alternative liquid-phase systems employ organic hydrogen donors such as formic acid or alcohols to circumvent gaseous H₂ requirements [23]. Heeres [24] obtained the highest GVL yield of 52% from D-fructose using formic acid as a hydrogen donor and Bo [25] achieved a 97% yield of GVL from ethyl levulinate using 2-butanol as a hydrogen donor. Although both formic acid and 2-butanol as hydrogen sources can be derived from biomass, their application as hydrogen donors in hydrogenation reactions necessitates precise control of catalytic conditions, which inevitably complicates large-scale implementation.

Water emerges as a superior sustainable hydrogen donor compared to conventional organic sources, providing three distinct merits: (1) global abundance and environmental benignity; (2) facile hydrogen liberation via spontaneous water splitting by Fe; (3) enhanced hydrogenation efficiency from in situ generated atomic hydrogen exhibiting greater reactivity than molecular H₂ [26,27,28]. These inherent properties position water as a potential hydrogen carrier for green GVL synthesis.

We reported the hydrogenation of LA to GVL via zinc-induced water splitting, achieving a 56% GVL yield without adding external catalysts. The results demonstrate that in situ hydrogen generation through metal-enabled water splitting can effectively facilitate the LA-to-GVL conversion [29]. This work extends the strategy toward practical biorefinery integration. Whereas our previous efforts focused on purified LA feedstocks, the direct conversion of native cellulose, which is the predominant constituent of lignocellulosic biomass, to GVL without intermediate isolation represents a critical advancement for industrial feasibility [30]. While existing cellulose-to-GVL systems [31,32] invariably depend on pressurized H₂ and noble metal catalysts, our approach implements two key technological innovations (Figure 1): (1) replacement of zinc with economically favorable iron powder for enhancing water-splitting sustainability; (2) deployment of commercial Raney Ni as a robust heterogeneous catalyst for in situ hydrogenation of cellulose-derived LA. Notably, the iron is converted into iron oxides during water splitting for hydrogen production and can be reduced back to metallic iron via biomass-derived reductants such as glycerol, enabling an iron—iron oxide chemical looping [26,33]. This integrated one-pot system successfully couples acid-catalyzed cellulose depolymerization with metal-induced hydrogenation, effectively eliminating energy-intensive separation steps in conventional biomass processing.

2. Results and Discussion

2.1. Conversion of Cellulose to GVL by One-Step Process

A preliminary feasibility assessment was conducted to evaluate the single-step conversion of microcrystalline cellulose to GVL via an integrated acid hydrolysis–hydrogenation process under mild hydrothermal conditions (200 °C, 1 h). Intriguingly, HPLC analysis (Figure S1a) revealed substantial accumulation of 5-hydroxymethylfurfural (HMF) and organic acids (predominantly LA and formic acid) as primary liquid-phase products. Complementary GC-MS characterization (Figure S1b) confirmed the absence of GVL in the product spectrum, indicating incomplete cascade conversion. This observation aligns with established reaction pathways [34], wherein cellulose-derived glucose undergoes Brønsted acid-catalyzed dehydration to HMF, followed by HMF fragmentation into LA and formic acid.

Critically, the complete lack of GVL formation suggests thermodynamic or kinetic limitations in the hydrogenation step. The reason for no GVL formation in one-step conversion is probably due to preferential HCl-Fe redox interactions (i.e., 2Fe + 6HCl = 2FeCl₃ + 3H₂ ↑) over cellulose depolymerization. This competitive metal–acid reaction likely depletes proton availability for sustained cellulose hydrolysis and compromises hydrogen generation efficiency, thereby arresting the cascade process at the LA intermediate stage.

2.2. Conversion of Cellulose to GVL by Two-Step Process

2.2.1. Hydrolysis of Cellulose to LA

A sequential two-stage strategy was employed to achieve GVL production from cellulose through (i) acid-catalyzed hydrolysis to LA, followed by (ii) hydrogenation using Raney Ni catalyst with H₂ generated via water splitting by Fe. Systematic optimization of reaction parameters for cellulose hydrolysis was first conducted to maximize LA yield. Post-hydrolysis characterization of liquid products by HPLC (Figure S2) and GC-MS (Figure S3) confirmed the formation of LA along with byproducts including HMF, formic acid, and volatile fatty acids.

Time-dependent studies at 200 °C (Figure 2a) revealed that LA yield increased from 30.8% to 45.1% as reaction duration was extended from 60 to 90 min. However, prolonging exposure to 120 min resulted in decreased LA yield (42.5%), indicating LA decomposition under extended hydrothermal conditions [32,34]. Elevated temperature experiments at 250 °C demonstrated comparable yields (44.4% at 60 min; 48% at 90 min), though temporal extension similarly proved detrimental to LA accumulation. Considering both process efficiency and energy consumption, optimal hydrolysis conditions were established at 200 °C for 90 min.

Subsequent investigation of acid concentration effects (0.1–1.0 mmol/L HCl) identified 0.5 mmol/L as optimal, achieving 54.0% LA yield. The observed inverse correlation between HCl concentration (>0.5 mmol/L) and LA production likely stems from competitive humin formation pathways [32]. This acid-mediated side reaction provides a mechanistic explanation for the counterproductive effects of excessive proton concentrations.

2.2.2. Hydrogenation of LA to GVL Without Additional Catalyst

Preliminary evaluation of cellulose hydrolysate hydrogenation in an exogenous catalyst-free environment revealed the system’s inherent catalytic capacity. Chromatographic analysis (Figure S4) demonstrated distinct GVL formation alongside residual LA and formic acid, confirming the feasibility of in situ H₂ utilization from water splitting without exogenous catalysts. Solid-phase product analysis via XRD (Figure S5) identified metallic iron and magnetite as dominant species, suggesting partial iron oxidation during hydrothermal processing.

A dose-dependent relationship between metallic iron loading (5–25 mmol) and GVL yield was established (Figure 3). Maximum GVL production (13.6%) occurred at 25 mmol Fe (200 °C, 4 h). More Fe facilitating the GVL production could be attributed to the enhancement of hydrogen production by Fe reduction of water. The generated hydrogen participates in LA hydrogenation possibly through the proton-coupled electron transfer processes.

2.2.3. Hydrogenation of LA to GVL with Catalyst

Systematic catalyst screening was performed to enhance GVL productivity through optimized hydrogenation processes. Commonly used metal catalysts (Cu, Ni, Co, Mo, Sn, Cr) powder were comparatively evaluated using H₂ formed in situ via water splitting. Post-reaction XRD analysis (Figure S6) confirmed the structural integrity of supplemental metals, with all additives retaining metallic phases except iron species that transformed into Fe₃O₄ through oxidation.

Catalytic performance assessment (Figure 4) identified the Ni catalyst as exhibiting exceptional LA-to-GVL conversion efficiency, achieving a peak GVL yield of 16.1%—representing a marked 41.2% enhancement relative to non-catalyzed conditions (11.4% yield). This pronounced Ni effect aligns with its known hydrogen spillover capacity and optimal d-band electronic structure for carbonyl group activation. Thus, porous Ni which possessed higher surface area was selected as the hydrogenation catalyst in the following study.

To obtain the optimal conditions of hydrogenation of LA with Raney Ni, the effect of Fe amount on the yield of GVL was initially studied. As shown in Figure 5a, the increase of Fe dosage from 5 to 25 mmol led to the remarkable growth of the GVL yield from 38.9% to 55.7% with 3 mmol Raney Ni at 200 °C. This represents a 4.3-fold enhancement in GVL yield compared to the system without added catalyst (Figure 3, 13.1% yield). The GVL yield rose slowly with levels of Fe above 15 mmol; GVL increased by 1.4% with 25 mmol Fe over the level at 20 mmol Fe, suggesting hydrogen saturation at active sites. Thus, 20 mmol Fe was selected for the following investigation.

Subsequent Ni loading studies mirrored this threshold behavior: GVL yield surged from 30.1% (1 mmol Ni) to 55.7% (3 mmol Ni), followed by marginal enhancement to 58.0% at 4 mmol Ni. Economic evaluation justified 3 mmol Ni as optimal, balancing catalytic performance and resource utilization. For the effect of reaction temperature and time, as shown in Figure 5b, the change of GVL yield with the increase of reaction time showed rapid growth followed by slow increase, with little change discernible between 200 °C and 220 °C. The GVL yield at 220 °C was slightly higher than that at 200 °C and a 61.9% yield was reached for 5 h. Interestingly, GVL yields were universally slightly higher than LA yields, indicating the kinetic competition between precursor generation and consumption.

Finally, the reusability of the Raney Ni catalyst was investigated through consecutive recycling experiments (Figure 6). Cyclic performance analysis revealed an initial activity reduction (4.8% yield) over two cycles, followed by stabilized catalytic behavior with less than 2% performance fluctuation in subsequent cycles. This stabilization phenomenon may arise from surface reconstruction eliminating metastable active sites while preserving the catalyst’s core nickel crystallites.

2.3. Proposed Pathway of GVL Synthesis from Cellulose

Based on the experimental results and the previous study [29] on GVL synthesis, a possible pathway of production of GVL from cellulose using hydrogen from water splitting by Fe was proposed, as illustrated in Figure 7. The cascade reaction initiates with acid-catalyzed cellulose depolymerization generating glucose intermediates, which undergo successive dehydration to HMF and rehydration to LA under hydrothermal conditions. Parallel iron oxidation drives proton-coupled electron transfer through the water-splitting reaction: The in situ formed hydrogen was adsorbed and activated on the surface of Raney Ni. Meanwhile, the keto carbonyl group of LA was adsorbed on the Ni surface. Then, the activated hydrogen attacked the carbon of the keto carbonyl of LA, leading to the reduction of the keto carbonyl group. Finally, the formed O^- in the keto carbonyl attacked the carbon in the carboxyl, resulting in hydroxyl and GVL formation with H₂O.

2.4. Conversion of Wheat Straw to GVL by Two-Step Process

Straw, a cellulose-rich biomass, has attracted sustained attention for its resource valorization. To validate the applicability of the proposed one-pot GVL production method, wheat straw conversion experiments were conducted.

Initially, the optimization of LA production via straw acid hydrolysis was investigated. As shown in Figure 8, the yields of LA and FA at 250 °C under varying acid concentrations (0.1, 0.5, and 1 mmol/L) exhibited an initial increase followed by a decline with prolonged reaction time. This trend was consistent with the cellulose hydrolysis pattern for LA production (Figure 2). The maximum LA yield of 23.9% was achieved at 90 min with 1 mmol/L HCl. Furthermore, the effect of reaction temperature on LA yield was investigated under 1 mmol/L acid concentration (Figure 8d). Results indicated that both LA and FA yields initially increased and subsequently decreased with rising temperature, likely due to thermal decomposition of LA and FA at excessive temperatures. Consequently, the optimized conditions for LA production were determined as 240 °C, 90 min, and 1 mmol/L acid concentration, achieving an LA yield of 24.1%. HPLC analysis (Figure S7) revealed that the liquid-phase products from straw acid hydrolysis contained oxalic acid, acetic acid, propionic acid, and acrylic acid, in addition to LA and FA. This observation is consistent with previous findings in biomass conversion studies [23,34].

Subsequently, the effects of Fe and Ni loading on GVL yield during LA hydrogenation were investigated. As illustrated in Figure 9, a low Fe dosage (<10 mmol) was unfavorable for LA-to-GVL conversion, which aligned with the GVL production trend observed in cellulose-derived systems. The GVL yield gradually increased with higher Fe loading, reaching 22.5% at 25 mmol Fe. According to the proposed cellulose conversion pathway (Figure 7), increased catalyst dosage could provide more reactive sites to enhance hydrogenation efficiency. For hydrogenation of LA to GVL, the influence of Raney Ni loading was further examined (Figure 9b). The addition of Raney Ni significantly promoted GVL generation during the reduction step. Specifically, the GVL yield exhibited a progressive increase with elevated Raney Ni loading. However, when the catalyst dosage exceeded 0.25 g, the GVL yield showed negligible improvement, suggesting that the catalytic sites on the catalyst surface were sufficiently available to meet the reaction demands at this threshold.

Finally, the effects of reaction temperature and time on two-step GVL production from straw were investigated. As demonstrated in Figure 10, the GVL yield exhibited an initial increase followed by a decline with prolonged reaction time across all tested temperatures. Notably, the maximum GVL yields were achieved at 5 h (250 °C), 6 h (240 °C), and 7 h (220 °C), respectively, indicating that higher temperatures favor accelerated GVL formation. Under the optimized conditions (0.25 g Raney Ni, 25 mmol Fe, 240 °C, 6 h), the GVL yield reached 24.6% from straw, demonstrating the feasibility of the proposed two-step strategy.

3. Experimental Section

3.1. Materials

Formic acid (FA, ≥96%, Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China), levulinic acid (99.8%, Sigma-Aldrich (Shanghai) Trading Co., Ltd., Shanghai, China), HCl (36–38%, Sino-pharm Chemical Reagent Co., Ltd.), α-cellulose (powder, Sigma-Aldrich) and γ-valerolactone (98%, Sino-pharm Chemical Reagent Co., Ltd.) were used in this study without further purification. The particle size of Raney Ni ordered from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China) was 100-mesh, with a specific surface area of 56 m²/g. Other metal powders (Fe, Ni, Cu, Cr, Mo, Co, Sn) were 200-mesh and purchased from Sino-pharm Chemical Reagent Co., Ltd. (Shanghai, China). Wheat straw was sourced from Zhengzhou, Henan Province, China. After drying, the straw material was ball-milled and sieved through a 100-mesh stainless steel sieve.

3.2. Experimental Procedure

All experiments were conducted in a Teflon-lined stainless-steel batch reactor with an inner volume of 28 mL. When conducting reactions in the oven, the vessel can rotate around its central axis at 1 r/s to enhance reactant mixing. The integrated process comprised three consecutive stages: (1) acid-catalyzed cellulose depolymerization to monosaccharides and subsequent LA formation; (2) in situ hydrogen generation via Fe–water redox reaction; (3) catalytic hydrogenation of LA to GVL. According to our previous study on the conversion of glucose to GVL, HCl is a suitable inorganic acid for acidic hydrolysis of carbohydrate biomass. A typical experimental procedure is as follows. First, 1 mmol cellulose and HCl (0.1~1 mmol/L) were put into the reactor and heated in the oven to a set temperature (200~250 °C) with water (~35% filling rate) for hydrolysis. Second, the reactor was cooled to room temperature naturally, after which Fe (5~25 mmol) and Ni (1~4 mmol) powder were added to the reactor for the subsequent reaction. Then, the reactor was heated for 1~5 h for hydrogenation. Finally, the reactor was taken out of the oven and cooled naturally again. The liquid products were collected and filtered with a 0.22 μm filter membrane. The collected solid samples were washed once with deionized water and then twice with absolute ethanol to remove impurities, and dried at 40 °C for 6 h in the vacuum oven. The catalyst reuse experiments were conducted by enclosing the catalyst with an SUS316 mesh for separation of Fe powder. The reaction time was defined as the time that the reactor was placed in the oven. The yield of GVL was defined as the ratio of moles of GVL in the product to moles of glucose (unit of cellulose) in the feedstock input (Equation (1)). Reported values represent mean yields from duplicate experiments with relative standard deviations <5%.

$$Yield of GVL={\displaystyle \frac{\mathrm{moles} \mathrm{o}\mathrm{f} \mathrm{G}\mathrm{V}\mathrm{L} \mathrm{i}\mathrm{n} \mathrm{t}\mathrm{h}\mathrm{e} \mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}}{\mathrm{m}\mathrm{o}\mathrm{l}\mathrm{e}\mathrm{s} \mathrm{o}\mathrm{f} \mathrm{g}\mathrm{l}\mathrm{u}\mathrm{c}\mathrm{o}\mathrm{s}\mathrm{e} \mathrm{i}\mathrm{n} \mathrm{t}\mathrm{h}\mathrm{e} \mathrm{f}\mathrm{e}\mathrm{e}\mathrm{d}\mathrm{s}\mathrm{t}\mathrm{o}\mathrm{c}\mathrm{k} \mathrm{i}\mathrm{n}\mathrm{p}\mathrm{u}\mathrm{t}}}\times 100\%$$

(1)

3.3. Analysis Methods

Liquid samples were analyzed by high performance liquid chromatography (HPLC, Agilent 1200 LC, Agilent Technologies (China) Inc., Beijing, China) equipped with a UV detector and two series Shodex KC-811 columns. The mobile phase was 2 mmol/L HClO₄ and the flow rate was 1 mL/min. The gas chromatography-mass spectroscopy (GC-MS) analysis was conducted on an Agilent 7890A GC system equipped with a 5975C inert MSD. The samples were separated by a HP-INNOWax capillary column with dimensions of 30 m × 250 μm × 0.25 μm and carried by helium gas. Details on the test conditions of HPLC and GC-MS were provided elsewhere [23]. The solid samples were analyzed by X-ray diffraction (XRD, Shimadzu X-ray Diffractometer 6100, Shimadzu Enterprise Management (China) Co., Ltd, Beijing, China) using Cu Kα radiation to determine the solid phase compositions, employing a scanning rate of 2°/min and 2θ angle in the range of 10° to 80°. The accelerating voltage was set at 40 KV with 30 mA flux. Diffraction patterns were compared with reference data in the ICDD PDF-2 database.

4. Conclusions

This work develops an effective catalytic system enabling direct cellulose-to-GVL conversion in a single reactor through acidic hydrolysis and in situ hydrogenation without intermediate separation. The integrated process features iron-driven hydrogen evolution via water splitting, while a Raney Ni catalyst facilitates selective LA hydrogenation to GVL. Under optimized conditions (220 °C, 5 h), a remarkable 61.9% GVL yield was achieved. Furthermore, the methodology’s versatility was demonstrated through wheat straw conversion experiments, yielding 24.6% GVL (240 °C, 6 h). Critically, this methodology explores an efficient biorefinery method that circumvents hazardous gaseous H₂ operation and utilizes Earth-abundant water as hydrogen donor via two-stage cascade reactions in one pot. The demonstrated system holds promise for scalable valorization of lignocellulosic waste, providing a technically robust and economically competitive route toward sustainable production of drop-in biofuels and biobased platform chemicals.