Hydrolysis of Lignin α‑O‑4 Model Compounds Using Strong Acid Catalyst in Water

Abstract

Efficient and sustainable cleavage of α-O-4 ether linkages is a critical initial step in lignin valorization. In this study, we present two highly effective catalytic protocols for the cleavage of C–O bonds in lignin α-O-4 model compounds under mild conditions (100 °C) that do not require organic solvents. Benzyl phenyl ether and its derivatives undergo quantitative conversion at 100 °C in the absence of an organic solvent under two contrasting conditions: aqueous trifluoromethanesulfonic acid (TfOH) yields the hydrolysis products phenol and benzyl alcohol, whereas solid superacid Nafion NR50 yields benzyl-migrated recombination products predominantly under solvent-free conditions. Nafion NR50 was recovered and reused for five consecutive cycles with conversion maintained within experimental error of activity. Non-fluorinated resins undergo the same transformation more slowly, which suggests that a perfluorinated acid is unnecessary. The medium-control principle established here defines the conditions under which condensation can be suppressed during acid-mediated α-O-4 cleavage.

Introduction

The transition toward a sustainable bio-based economy is driven by the depletion of fossil fuels and escalating environmental concerns. , Lignin, a major constituent of lignocellulosic biomass, is the most abundant renewable source of aromatic building blocks. − However, the full valorization of this heterogeneous biopolymer is impeded by its recalcitrant and complex three-dimensional network (Figure ). The primary strategy for lignin depolymerization involves the cleavage of inter-unit ether linkages. Among these, the α-O-4 bond is of particular strategic importance. Its bond dissociation energy is 80–118 kJ/mol, which is lower than that of the β-O-4 bond (150 kJ/mol). , Consequently, cleavage of the α-O-4 bond often constitutes the rate-determining step in initial depolymerization and is critical for releasing lignin from the lignocellulosic matrix. , Given the analytical difficulties inherent in processing real lignin streams, the use of model compounds remains an indispensable approach for elucidating the precise reaction mechanisms and kinetics that are essential for optimizing these degradation pathways.

1.

Typical structural model of lignin derived from woody biomass, alongside the target α-O-4 model.

A diverse array of depolymerization strategies has been developed to convert lignin into value-added low-molecular-weight compounds, primarily categorized into thermal (e.g., pyrolysis, gasification) and chemical (e.g., oxidation, reduction) approaches. − Thermochemical methods, despite extensive investigation, often exhibit low product selectivity and promote the formation of recalcitrant char or tar due to uncontrolled radical repolymerization. Accordingly, considerable attention has shifted toward chemical depolymerization, specifically catalytic hydrolysis, in addition to emerging electrochemical − or photochemical − methods, which offer better control over product distribution. However, a major limitation of these conventional catalytic processes is their reliance on harsh reaction conditions, which typically involve high temperatures (>200 °C), high pressures, and the use of organic solvents to ensure substrate solubility and suppress condensation. The dependence on organic solvents increases the environmental footprint and separation costs and contradicts the principles of green chemistry. , Furthermore, even in aqueous systems, achieving efficient ether bond cleavage while preventing the rapid recombination of reactive intermediates remains a challenge unless extreme supercritical conditions are employed.

In this context, water has garnered increasing attention as an ideal “green” solvent, and organic synthesis in aqueous media has been widely investigated. − Recent comprehensive studies have described the decomposition of lignin model compounds in water, highlighting the potential of water to replace environmentally harmful organic solvents. For instance, the cleavage of the α-O-4 ether linkage in benzyl phenyl ether (BPE) has been achieved using various Lewis or Brønsted acid catalysts (Scheme a and b, Table ). ,,− Despite these advances, two limitations persist across the reported systems. Firstly, the cleavage of the α-O-4 linkage in aqueous media generally requires temperatures of 180–400 °C and, in some cases, supercritical or near-supercritical conditions to overcome the activation barrier. Secondly, more fundamentally, the benzylic cation formed when the C–O bond is broken is quickly captured by electron-rich aromatic rings. This means that depolymerization and condensation occur concurrently. It is this competition, rather than the cleavage step itself, that limits the amount of monomer produced.

1. Decomposition of the Lignin α-O-4 Model in Water as the Solvent.

1. Optimization of Cleavage Conditions of 1a C–O Bonds .

Entry	Acid(mol %)	H2O(mL)	time(h)	Conv.(%)	Yield (%) 2a:3a:4a:5a
1	HCl (100)	1.0	5	<5	-
2	H2SO4 (100)	1.0	5	<5	-
3	AcOH (100)	1.0	5	6	<5:0:<5:<5
4	CF3CO2H (100)	1.0	5	9	<5:0:0:7
5	CH3SO3H (100)	1.0	5	<5	-
6	PTSA (100)	1.0	5	6	<5:0:<5:<5
7	TfOH (100)	1.0	5	100	84:41:13:<5
8	TfOH (50)	1.0	5	85	26:<5:27:24
9	TfOH (100)	2.0	1	72	27:<5:10:<5
10	TfOH (100)	0.5	1	97	68:3:10:20
11	TfOH (100)	1.0	1	74	53:<5:14:7

^aReaction conditions: 1a (1.0 mmol) at 100 °C under air in a vial tube.

^bDetermined by GC, using naphthalene as an internal standard.

In this study, the hydrolysis of BPE and its derivatives is demonstrated using trifluoromethanesulfonic acid (TfOH) as a Brønsted acid promoter in water, achieving selective cleavage of the α-O-4 linkage (Scheme c). However, the use of soluble fluorinated acids raises environmental concerns regarding their disposal and toxicity, which requires the development of recoverable catalytic systems. , To address this, the reaction is examined using Nafion NR50, a perfluorinated sulfonic acid resin, to directly compare the catalytic performance of the homogeneous and heterogeneous systems. Through this comparative study, supported by Hammett plot analysis, mechanistic insights into the reaction pathways are provided, and the critical role of the aqueous phase in suppressing the recombination of reactive intermediates is elucidated. Finally, the recovered resin remains active for up to five cycles. This addresses the issue of condensation-induced deactivation, which commonly limits the use of solid acids in lignin chemistry. To the best of our knowledge, this is the first report of α-O-4 undergoing complete hydrolysis at 100 °C without the use of an organic solvent. Furthermore, we demonstrated that the cleavage/condensation ratio in α-O-4 scission can be altered by the reaction medium alone.

Results and Discussion

Optimization of Reaction Conditions

To determine the optimal conditions for hydrolytic cleavage of the α-O-4 bond, BPE (1a) was chosen as the model substrate. Initial screening was conducted to evaluate several Brønsted acids in water at 100 °C (Table ).

Strong inorganic acids, such as hydrochloric and sulfuric acids, were unsuitable for this transformation. Substrate 1a remained almost unreacted (entries 1 and 2). Subsequent evaluation of acetic and trifluoroacetic acids revealed a slight degree of C–O bond cleavage (entries 3 and 4). However, the target hydrolysis products, phenol (2a) and benzyl alcohol (3a), were scarcely detected. Instead, recombination processes dominated the reaction, preferentially yielding rearranged products 4a and 5a.

Screening of sulfonic acid (entries 5–7) markedly improved the hydrolytic reactivity. Electron-rich sulfonic acids, such as methanesulfonic acid and p-toluenesulfonic acid, failed to improve the conversion (entries 5 and 6). However, the use of TfOH accelerated the reaction (entry 7). − The highly electron-withdrawing trifluoromethyl group of TfOH promoted complete substrate conversion. The hydrolysis proceeded with high selectivity, yielding phenol (2a) in 84% yield and benzyl alcohol (3a) in 41% yield.

Further optimization of the reaction parameters underscored the importance of the established TfOH system. Decreasing the TfOH loading resulted in a corresponding decline in substrate conversion (entry 8). The volume of water was also identified as a critical factor affecting conversion and selectivity (entries 9 and 10). Doubling the amount of water decreased the conversion, whereas performing the reaction at higher concentrations led to selectivity issues, likely as a result of the promotion of unwanted side reactions. Attempts to shorten the reaction time resulted in incomplete conversion of 1a (entry 11). Furthermore, dibenzyl ether was not detected under any of the reaction conditions using TfOH.

While TfOH demonstrates high reactivity, the use of soluble fluorinated acids raises environmental and regulatory concerns, particularly regarding their classification as perfluoroalkyl substances (PFAS). − To develop a more sustainable and recoverable catalytic system, heterogeneous solid acid catalysts were evaluated (Table ). Initial attempts using 100 mg of the sulfonic acid–based ion-exchange resins DUOLITE, DIAION, and Nafion NR50 were performed under aqueous conditions identical to those used for TfOH (Table , entries 1–3). However, in contrast to the homogeneous system, no substrate conversion was observed with any solid catalyst.

2. Optimization of Solid Acid Catalysts for Cleavage Reactions in BPE (1a)­ .

Entry	Acid	Y	Time (h)	Conv. (%)	Yield (%) 2a:3a:4a:5a
1	DUOLITE	1.0	5	<5	-
2	DIAION	1.0	5	<5	-
3	Nafion NR50	1.0	5	<5	-
4	Nafion NR50	0.5	5	<5	-
5	Nafion NR50	0.18	5	<5	-
6	Nafion NR50	0.036	5	78	15:<5:61:21
7	Nafion NR50	none	5	100	13:<5:56:27
8	DUOLITE	none	16	100	26:<5:49:22
9	DIAION	none	16	62	33:<5:50:15
10	Nafion NR50	none	1	98	23:<5:52:23

^aReaction conditions: 1a (1.0 mmol) at 100 °C under air in a vial tube.

^bDetermined by GC, used naphthalene as an internal standard.

Given its super acidic nature and structural similarity to TfOH, optimization efforts focused on the perfluorinated sulfonic acid resin Nafion NR50. − In addition, Nafion NR50 (100 mg) corresponds to ca. 0.08 mmol of sulfonic acid sites (ca. 8 mol % relative to 1a) based on an ion-exchange capacity of 0.8 mmol g^–1. It was hypothesized that the bulk aqueous phase might hinder access of the hydrophobic substrate to the solid active sites. Consequently, the effect of water content was evaluated systematically (entries 4–7). Maintaining 10 equiv of water resulted in no observable decomposition; however, reduction of the water content to 2 equiv. promoted the reaction, affording 78% conversion of 1a. The reaction on Nafion NR50 did not produce the expected hydrolysis product, benzyl alcohol (3a). Instead, the reaction preferentially afforded benzyl-migrated recombination products 4a and 5a. This outcome suggests that the reaction predominantly occurs on the catalyst surface, where the local concentration of benzyl cations is high, whereas the availability of water molecules to trap them is low; this condition favors C–C bond-forming recombination.

Under neat conditions, omitting water and performing the reaction resulted in complete substrate conversion (100%) (entry 7). DUOLITE and DIAION also catalyzed the reaction under these conditions; however, longer reaction times were required to achieve comparable results (entries 8 and 9). The transformation therefore does not require a perfluorinated acid, and a fluorine-free implementation is accessible at the cost of reaction time. Using Nafion NR50 under neat conditions for 1 h enabled efficient C–O bond cleavage, affording phenol (2a) in 23% yield, along with the rearranged products (entry 10). More significantly, despite differing widely in acid strength, the three resins produced similar product distributions (entries 8–10). While acid strength governs the rate of the reaction, it does not affect the partition between cleavage and recombination. The contrast with the homogeneous system is more pronounced when compared to the homogeneous system. TfOH, which has an acid strength comparable to Nafion’s, produced a completely different distribution in water (see Table , entry 7).

Substrate Scope Using Nafion

After establishing the optimal solvent-free conditions using Nafion NR50, the substrate scope to investigated to evaluate the dependence of the reaction on various functional groups (Scheme ). Compounds 2 and 3 are phenol and benzyl alcohol derivatives, respectively. Compounds 4 and 5 are benzyl-substituted derivatives bearing one and two substituents, respectively, for which the regiochemistry has not yet been determined. Compound 6, not shown in the scheme, is a phenol derivative bearing three benzyl substituents. These compounds were identified using GC-MS (see Supporting Information). The influence of substituents on the phenoxy ring was first examined. The reaction of substrate 1b, which bears an electron-donating methyl group, proceeded to completion. Similarly, substrates bearing halogen substituents (1c–1e) underwent C–O bond cleavage. Consistent with the optimization results, none of these reactions yielded benzyl alcohol; instead, the corresponding phenols and benzyl-migrated recombination products were obtained instead. Furthermore, substrates 1f–i, which incorporate methoxy groups at various positions, a representative structural motif in native lignin, also reacted to full conversion. Substituents ranging from methyl to methoxy and halogens all gave complete conversion on the phenoxy ring (1b–1i). This indicates that the phenoxy ring, which is on the leaving-group side of the scissile bond, does not influence the rate or product distribution to a measurable extent. Notably, the introduction of these substituents on the phenoxy ring did not alter the product distribution appreciably among free phenol, mono-benzylated, and di-benzylated derivatives.

2. Substrate Scope in the C–O Bond Cleavage Reaction of Different Substituted BPE .

^a Reaction conditions: 1 (1 mmol), Nafion NR50 (100 mg) for 1 h.

In contrast to the phenoxy ring, introducing substituents on the benzyl moiety markedly affected the reaction efficiency. Substrates 1j–m demonstrated that derivatives bearing methyl or fluoro groups achieved 100% conversion. Conversely, substrates bearing stronger electron-withdrawing halogens, such as chloro and bromo substituents, exhibited lower conversions (43% and 36%, respectively). These results indicate that the electronic character of the benzyl ring strongly governs overall reactivity, likely through its influence on the ease of benzyl cation formation. Substrates bearing strong electron-donating methoxy groups on the benzyl ring (1n–q) therefore exhibited high reactivity. For the p-methoxy-substituted substrate 1n, the reaction at 100 °C produced an intractable mixture as a result of overreaction; accordingly, the temperature was lowered to 80 °C to achieve controlled cleavage. Despite the differences in conversion, no consistent trend was observed in the ratio of cleavage to recombination products across these benzyl-substituted derivatives.

The regiochemistry of mono- and di-benzylated products 4 and 5 has not been determined. Since the ortho/para partition in the benzylation step is sterically controlled, a more detailed analysis of the effects of steric interactions on the recombination pathway must await this determination.

To quantitatively assess the electronic influence observed in the substrate scope study and to probe the nature of the transition state, a Hammett plot analysis was conducted. This analysis was performed to evaluate the relative reaction rates of the benzyl-substituted derivatives (see Figure , S1 and Table S2). The plot exhibited linearity, with a large negative reaction constant of ρ = −4.1 (R² = 0.94). However, the correlation against the electrophilic substituent constants, σ⁺, was appreciably poorer (R² = 0.72). , On the other hand, using the Yukawa–Tsuno equation (log­(k _X/k _H) = ρ­[σ + r(σ⁺ – σ)]) yields r = −0.22, indicating that the resonance demand is no enhanced resonance demand (r < 0).

2.

Yukawa-Tsuno plot (r = −0.217).

Charge development at the transition state is extensive, comparable in magnitude to systems in which formation of a benzylic cation is rate-determining step (for reference, solvolysis of cumyl chlorides, from which the σ⁺ scale was defined, gives ρ⁺ = −4.54, and that of diphenylcarbinyl chlorides ρ⁺ = −5.1, whereas primary benzyl chlorides give values near −1.9). , Yet the absence of enhanced resonance demand indicates that this charge is not fully delocalized into the aromatic ring at the transition state, as it would be in a free, solvent-separated benzylic cation. We therefore describe the process not as classical S_N1 solvolysis but as rate-determining ionization to a tight ion pair, in which the cationic center remains closely associated with the sulfonate counterion of the resin.

Therefore, the behavior of the heterogeneous system is better described in terms of the reaction environment than acidity. Nafion NR50 is a phase-separated ionomer in which sulfonic acid domains are clustered and embedded in a hydrophobic, perfluorinated matrix. Under solvent-free conditions, the matrix concentrates the nonpolar substrate near the acid sites, accounting for the high observed rate. However, the same hydrophobicity excludes bulk water from the site at which the cation is generated. Therefore, the available nucleophile is not water, but rather the electron-rich ring of a neighboring phenol, which gives the benzyl-migrated products. In the aqueous homogeneous system, the ionization step is followed by hydration because water is present in large excess at the site of cation formation. Thus, the Nafion microenvironment is best understood as hydrophobic rather than as a stronger acid.

Catalyst Recyclability and a Comparison of the Present Study with Previous Research

The stability and recyclability of Nafion NR50 were investigated in the C–O bond cleavage of BPE (1a). After each reaction cycle, the solid catalyst was recovered and regenerated to restore its acidic active sites by washing with methanol and water. The recycled Nafion NR50 catalyst exhibited retained its catalytic activity across five consecutive cycles (Figure ). No decrease in substrate conversion was observed. The yields of products 2–6a remained consistent across runs, suggesting that the catalyst retained its activity.

3.

Recycling test of Nafion in the C–O bond cleavage of 1a.

Visual inspection of the catalyst revealed a noticeable color change over successive cycle (Figure ). The pristine Nafion resin was colorless; however, after the first catalytic cycle, it acquired a reddish tint that progressively darkened to a brown solid after the fifth cycles. Despite this apparent discoloration, which is attributable to the deposition of trace conjugated oligomeric byproducts on the surface, Fourier-transform infrared (FT-IR), solid-state ¹⁹F NMR, SEM-EDX analysis of the recovered catalyst showed no substantial structural differences from the fresh resin (Figure S29). In the case of UV-vis absorption, a clear difference was observed before and after use, leading us to suspect that this was due to the attachment of an unidentified compound originating from the substrate (Figure S10).

4.

Photographs of Nafion under white LED before and after the reaction.

These results indicate that Nafion NR50 possesses high chemical stability under the reaction conditions employed, and its ability to be reused repeatedly without loss of catalytic activity highlights its potential as a sustainable and solid acid catalyst for ether bond cleavage.

Table compares the two systems described here with previously reported catalytic systems for the cleavage of α-O-4 model compounds. The comparison is made in terms of the following: catalyst type, reaction conditions, product selectivity, recyclability, and the principal limitation of each system. Two features distinguish the present work. First, the operating temperature of 100 °C is 80–300 °C lower than that of comparable aqueous systems. Lewis acid catalysis of the same substrate in water requires 300–400 °C. Even at 400 °C, condensation products account for 24–40% of the product distribution, rising to 68–76% at 300 °C. Reductive systems achieve higher monomer yields, but they require elevated hydrogen pressure, an organic co-solvent, and a catalyst prepared by high-temperature pyrolysis. Second, neither of the present systems uses an organic solvent at any stage, unlike the γ-valerolactone-, ethanol-, or tetrahydrofuran-based processes reported previously.

3. Conversion of Lignin α-O-4 Model in Different Conditions.

Catalyst	Solvent	T (°C)	Time (h)	Main products	Recyclable	Key limitation/advantage	Ref.
H2SO4	EtOH or MeOH/H2O	160–200	4	phenol, ether, benzyl-migrated	no	organic solvent, homogeneous, autoclave	,
H2SO4	γ- valerolactone/H2O	180	5	phenol, benzyl alcohol, benzyl-migrated	no	organic solvent, homogeneous
AlCl3/MCl2 (Fe, Cu, Co, Ni)	H2O	300–400	3	phenol, toluene, benzene, condensates	no	high T/P, condensation homogeneous, non-recoverable
Fe–N/C	THF/H2O	240	12	phenol, toluene	(not reported)	requires 2 MPa H2, organic co-solvent, 800 °C catalyst synthesis
NOBF4	DCM/HFIP	rt	1	benzyl-migrated (up to 84% yield)	no	PFAS solvent
TfOH (This work)	H2O	100	5	phenol, benzyl alcohol	no	PFAS concern, stoichiometric loading	–
Nafion NR50 (This work)	none	100	1	phenol, benzyl-migrated	5 cycles	condensation-selectivity, reusable	–

Conclusions

Two catalytic protocols were developed to cleave the C–O bond of lignin α-O-4 model compounds at 100 °C without an organic solvent. A comparison of the two protocols revealed that the branching between cleavage and condensation is controlled by the reaction medium. In an aqueous medium, the homogeneous superacid TfOH produced hydrolysis products with high selectivity. Under solvent-free conditions, however, the solid superacid Nafion NR50 produced benzyl-migrated recombination products almost exclusively. The finding that three resins of widely differing acid strength gave closely similar product distributions while acid strength determined only the rate confirms that selectivity is governed by the medium rather than by acidity.

A balanced assessment is warranted with respect to green chemistry. Nafion NR50 was recovered by filtration and reused for five cycles or structural alteration. However, both of the employed acids are perfluoroalkyl substances. Immobilization changes the disposition of the fluorinated acid rather than eliminating it. The solid is retained and recovered, so dispersive release into the aqueous waste stream is largely avoided. However, the material remains a fluoropolymer, the manufacture of which involves fluorochemical processing, and the management of which at the end of its life remains unresolved. Therefore, it is relevant that the non-fluorinated resins DUOLITE and DIAION undergo the same transformation under solvent-free conditions. The cost is reaction time rather than chemistry, and a fluorine-free implementation of this chemistry is possible.

The protocol selectively targets the α-O-4 linkage and does not currently apply to β-O-4 motifs or isolated lignin. When a β-O-4 model substrate was subjected to the optimized conditions, the conversion was only 14%, and the resulting product mixture was too complex to quantify the components reliably. A commercially available lignin subjected to the same conditions showed no significant conversion. Three requirements for extension follow directly from the established mechanism: 1) A medium that dissolves or swells the substrate while preserving control over water activity at the acid site so that cleavage/condensation branching remains tunable. 2) Prior activation of the Cα position for β-O-4 linkages (e.g., oxidation to the ketone) so that scission is no longer contingent on benzylic cation stability. 3) A means of intercepting the benzylic cation other than bulk water (e.g., an added nucleophile or capping agent) to allow the hydrolysis-selective regime to be reached without a large aqueous phase. These findings provide a basis for guiding the controlled depolymerization of more complex lignin motifs, including β-O-4 linkages, and ultimately, the sustainable valorization of real lignocellulosic biomass into valuable aromatic chemicals.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.6c05568.

• Experimental procedures, materials, recycling experiments, Hammett analysis, FT-IR, solid-state ¹⁹F NMR, EDX, and UV–vis absorption data of Nafion NR50, and GC and GC–MS data (PDF)

S.M.: Conceptualization, Funding acquisition, Methodology, Project administration, Supervision, Writing – original draft, Writing – review & editing. K.Y.: Investigation, Data curation, Methodology, Formal analysis. Y.K.: Writing – review & editing. T.I.: Writing – review & editing. K.F.: Conceptualization, Funding acquisition, Methodology, Project administration, Supervision, Writing – review & editing.

The authors declare no competing financial interest.