Sustainable production of benzene from lignin

Benzene is a widely used commodity chemical, which is currently produced from fossil resources. Lignin, a waste from lignocellulosic biomass industry, is the most abundant renewable source of benzene ring in nature. Efficient production of benzene from lignin, which requires total transformation of Csp2-Csp3/Csp2-O into C-H bonds without side hydrogenation, is of great importance, but has not been realized. Here, we report that high-silica HY zeolite supported RuW alloy catalyst enables in situ refining of lignin, exclusively to benzene via coupling Bronsted acid catalyzed transformation of the Csp2-Csp3 bonds on the local structure of lignin molecule and RuW catalyzed hydrogenolysis of the Csp2-O bonds using the locally abstracted hydrogen from lignin molecule, affording a benzene yield of 18.8% on lignin weight basis in water system. The reaction mechanism is elucidated in detail by combination of control experiments and density functional theory calculations. The high-performance protocol can be readily scaled up to produce 8.5 g of benzene product from 50.0 g lignin without any saturation byproducts. This work opens the way to produce benzene using lignin as the feedstock efficiently.


Synthesis of the model compounds
The syntheses of the model compounds were performed using the methods reported 1, 2 . 1 H NMR and 13 C NMR analyses were performed on a Bruker Avance III 400 HD using DMSO-d6 as the solvent. 1 H chemical shifts were referenced to TMS at 0 ppm, and 13 C chemical shifts were referenced to DMSO-d6 at 39.6 ppm. Multiplicities are described using the following abbreviations: s, singlet; d, doublet; t, triplet; q, quartet; and m, multiplet. The details of the methods and procedures for the syntheses of the model compounds are described as follows: 1-(4-Methoxyphenyl)-1-propanol (1a) Sodium borohydride (3.8 g, 100 mmol) was added into the solution of 4'methoxypropiophenone (16.4 g, 100 mmol) in THF/H2O (100 mL/30 mL) at 0 °C . The reaction mixture was stirred at room temperature for 24 h. After the reaction, the mixture was sat for 10 min, and then the organic layer was separated. Ethyl acetate (50 mL) was added into the stirred organic layer at room temperature. The organic layer was then successively washed with deionized water (50 mL×3) and saturated brine (50 mL), dried by anhydrous MgSO4. After the concentration in vacuum rotavap, 15.1 g of 1a was obtained finally. 1  3, 4-Dimethoxybenzaldehyde (1.7 g, 10 mmol) was added into the solution of Grignard reagent, freshly prepared from bromoethane (1.6 g, 15 mmol) and magnesium turnings (0.4 g, 15 mmol) in anhydrous THF (20 mL) at 0 °C . The reaction mixture was stirred at room temperature for 1 h. After the reaction, the mixture was quenched with cold water (1 mL) and acidified with saturated NH4Cl solution (20 mL). Ethyl acetate (50 mL) was added into the stirred organic layer at room temperature. The organic layer was then successively washed with deionized water (50 mL×3) and saturated brine (50 mL), dried by anhydrous MgSO4. After the concentration in vacuum rotavap, 1.7 g of 2a was obtained finally. 1  4-Hydroxy-3-methoxybenzaldehyde (1.5 g, 10 mmol) was added into the solution of Grignard reagent that was freshly prepared from bromoethane (3.3 g, 30 mmol) and magnesium turnings (0.7 g, 30 mmol) in anhydrous THF (20 mL) at 0 °C . The reaction mixture was stirred at room temperature for 1 h. After the reaction, the mixture was quenched with cold water (1 mL) and acidified with saturated NH4Cl solution (20 mL). Ethyl acetate (50 mL) was added into the stirred organic layer at room temperature. The organic layer was then successively washed with deionized water (50 mL×3) and saturated brine (50 mL), dried by anhydrous MgSO4. After the concentration in vacuum rotavap, 0.9 g of 3a was obtained finally. 1  1-(4-isopropoxy-3-methoxyphenyl)propan-1-ol (4a) 4-Hydroxy-3-methoxybenzaldehyde (3.0 g, 20 mmol) was dissolved in 20 mL DMF at room temperature. 2-Bromopropane (4.9 g, 40 mmol) and anhydrous K2CO3 (5.5 g, 40 mmol) were subsequently added and stirred at 100 °C for 6 h. After the reaction, the mixture was poured into water (200 mL) and extracted by ethyl acetate (50 mL×3). The organic layer was then successively washed with saturated brine (20 mL×6), dried by anhydrous MgSO4 and concentrated in vacuum. Finally, 3.5 g of 4-isopropoxy-3-methoxybenzaldehyde was obtained by silica column chromatography. 4-Isopropoxy-3-methoxybenzaldehyde (3.5 g, 18 mmol) was added into the solution of Grignard reagent, freshly prepared from bromoethane (3.2 g, 30 mmol) and magnesium turnings (0.8 g, 30 mmol) in anhydrous THF (50 mL) at 0 °C . The reaction mixture was stirred at room temperature for 1 h. After the reaction, the mixture was quenched with cold water (1 mL) and acidified with saturated NH4Cl solution (50 mL). Ethyl acetate (100 mL) was added into the stirred organic layer at room temperature. The organic layer was then successively washed with deionized water (50 mL×3) and saturated brine (50 mL), dried by anhydrous MgSO4. After the concentration in vacuum rotavap, 3.9 g of 4a was obtained finally. 1  1-(3, 4, 5-Trimethoxyphenyl) propan-1-ol (5a) 3, 4, 5-Trimethoxybenzaldehyde (2.0 g, 10 mmol) was added into the solution of Grignard reagent that was freshly prepared from bromoethane (1.6 g, 15 mmol) and magnesium turnings (0.4 g, 15 mmol) in anhydrous THF (20 mL) at 0 °C . The reaction mixture was stirred at room temperature for 1 h. After the reaction, the mixture was quenched with cold water (1 mL) and acidified with saturated NH4Cl solution (20 mL). Ethyl acetate (50 mL) was added into the stirred organic layer at room temperature. The organic layer was then successively washed with deionized water (50 mL×3) and saturated brine (50 mL), dried by anhydrous MgSO4. After the concentration in vacuum rotavap, 2.0 g of 5a was obtained finally. 1  4'-Hydroxy-3', 5'-dimethoxybenzaldehyde (1.8 g, 10 mmol) was added into the solution of Grignard reagent that was freshly prepared from bromoethane (3.3 g, 30 mmol) and magnesium turnings (0.7 g, 30 mmol) in anhydrous THF (20 mL) at 0 °C . The reaction mixture was stirred at room temperature for 1 h. After the reaction, the mixture was quenched with cold water (1 mL) and acidified with saturated NH4Cl solution (20 mL). Ethyl acetate (50 mL) was added into the stirred organic layer at room temperature. The organic layer was then successively washed with deionized water (50 mL×3) and saturated brine (50 mL), dried by anhydrous MgSO4. After the concentration in vacuum rotavap, 1.3 g of 6a was obtained finally. 1  1-(4-isopropoxy-3, 5-dimethoxyphenyl)propan-1-ol (7a) 4-Hydroxy-3, 5-dimethoxybenzaldehyde (3.6 g, 20 mmol) was dissolved in 20 mL DMF at room temperature. 2-Bromopropane (4.9 g, 40 mmol) and anhydrous K2CO3 (5.5 g, 40 mmol) were subsequently added and stirred at 100 °C for 6 h. After the reaction, the mixture was poured into water (200 mL) and extracted by ethyl acetate (50 mL×3). The organic layer was then successively washed with saturated brine (20 mL×6), dried by anhydrous MgSO4 and concentrated in vacuum. Finally, 4.0 g of 4-isopropoxy-3, 5-dimethoxybenzaldehyde was obtained by silica column chromatography. 4-Isopropoxy-3, 5-dimethoxybenzaldehyde (4.0 g, 18 mmol) was added into the solution of Grignard reagent, freshly prepared from bromoethane (3.2 g, 30 mmol) and magnesium turnings (0.8 g, 30 mmol) in anhydrous THF (50 mL) at 0 °C . The reaction mixture was stirred at room temperature for 1 h. After the reaction, the mixture was quenched with cold water (1 mL) and acidified with saturated NH4Cl solution (50 mL). Ethyl acetate (100 mL) was added into the stirred organic layer at room temperature. The organic layer was then successively washed with deionized water (50 mL×3) and saturated brine (50 mL), dried by anhydrous MgSO4. After the concentration in vacuum rotavap, 3.9 g of 7a was obtained finally. 1 Bromine (33.6g, 210 mmol) was added dropwise into the solution of 3', 4'dimethoxyacetophenone (36.4 g, 200 mmol) and AlCl3 (667 mg, 5 mmol) in diethyl ether (200 mL) at 0 °C~5 °C . The reaction mixture was then stirred at room temperature for 1h. After the reaction, the mixture was poured into ice water (1000 mL) and extracted by ethyl acetate (200 mL×2). The organic layer was then successively washed with deionized water (200 mL×3 ) and saturated brine (100 mL), dried by anhydrous MgSO4 and concentrated in vacuum. Finally, 33.7 g of 2-bromo-1-(3, 4-dimethoxyphenyl)ethanone was obtained by recrystallization process.

2-(4-Methoxyphenyl)propan-2-ol
Methylmagnesium chloride solution (3.0 M in THF, 10 mL) was added into the solution of 4'-methoxyacetophenone (1.4 g, 10 mmol) in THF (20 mL) at 0 °C . The reaction mixture was stirred at room temperature for 1 h. After the reaction, the mixture was quenched with cold water (1 mL) and acidified with saturated NH4Cl solution (20 mL). Ethyl acetate (50 mL) was added into the stirred organic layer at room temperature. The organic layer was then successively washed

Computational methodology
The Vienna Ab Initio Simulation Package (VASP) code based on frozen-coreall-electron projector augmented wave method 3 and plane-wave basis was used to perform the spin-polarized DFT computations. The DFT functional PBE-D3 (Perdew, Burke and Ernzerhof functional with the latest dispersion correction) was used. It corresponds to the PBE functional and the correction for dispersion effects proposed by Grimme et al. (DFT-D3) 4 . A converged kinetic energy cutoff of 400 eV was chosen. The structure optimization was carried out by minimizing forces with the conjugate-gradient algorithm until the force on each ion is below 0.03 eV/A and the convergence criteria for electronic self-consistent interactions are 10 −5 . The zero vibrational energy correction was not taken into account. The minimum energy pathway for elementary reaction steps was performed using climbing-image NEB (Nudged Elastic Band) calculation 5 . A low-symmetry rhombohedral unit cell was used to model faujasite zeolite (a = b = c = 17.21 Å, α = β = γ = 60°). The general formula of a faujasite zeolite with a pure silicon component is Si48O96. The faujasite zeolite framework contains only one crystallographically tetrahedral T site. In view of our experimental results, HY zeolite with a Si/Al ratio of 30 showed the best catalytic performance, indicating that there are approximately two Al atoms in a unit cell. Therefore, two Al atoms were included in our HY unit cell model. Moreover, it is reasonable to assume that these two Al atoms in the unit cell are spatially adjacent, i.e., one is the active site, and the other, due to space constraints, can adsorb or desorb water molecules to promote the reaction. The reaction (∆E) and activation energies (Ea) were calculated using the following two formulas:

∆E = E(FS) − E(IS) and Ea = E(TS) − E(IS), respectively. Here, E(TS), E(IS)
, and E(FS) are the calculated energies of the transition, initial and final states of each elementary step, respectively. The contribution of entropy to free energy in the adsorption/desorption processes was considered in our computational work. Besora et al. 6 reported that the free-energy corrections from the direct application of popular computational packages with its usual simplified models based on IGRRHO (harmonic oscillator, rigid rotor, and particle in-a-box) approaches can give a good estimate of the free-energy changes in the liquid-phase reaction. Greeley and co-workers' work 7 indicated that the most important contributions arise from the translational entropy in the adsorption/desorption process.

Supplementary Note 1 Traditional Route for benzene production from fossil resources
Currently, industrial production of benzene is highly dependent on petroleum and coal resources 8, 9 . Worldwide, around 30% of commercial benzene is produced via catalytic reforming process, in which aromatic molecules are produced from the dehydrogenation of cycloparaffins, dehydroisomerization of alkyl cyclopentanes, and cyclization/dehydrogenation of paraffins. The feed to the catalytic reformer may be a straight-run, hydrocracked, or thermally cracked naphtha fraction in the range of C6 to 200 °C. Generally, a narrow naphtha cut of 71-104 °C is fed to the reformer for the production of desired benzene product. Supported platinum-rhenium on a high surface area alumina support is the most frequently reforming catalyst. In the reforming process, the operating conditions and type of feedstock determine, to a large extent, the amount of produced benzene. In addition, the benzene product is usually recovered from reformate by solvent extraction techniques.
Benzene can be produced from the hydrodemethylation of toluene under catalytic (Hydeal and DETOL processes) or thermal (HDA and THD processes) conditions, which contribute 25-30% of the global benzene supply. During the catalytic hydrodemethylation, toluene is mixed with hydrogen and passed through a vessel loaded with catalysts, such as supported chromium or molybdenum oxides, platinum oxides, on silica or alumina. The operating temperatures range from 500 to 595 °C and pressures are usually 4-6 MPa. The reaction is highly exothermic and the temperature is controlled by the injection of quench hydrogen at several places along with the reaction.
Benzene can also be obtained from pyrolysis gasoline which is yielded as a liquid by-product during the steam cracking process of heavy naphthas or light hydrocarbons, such as propane or butane, to ethylene. Approximately, 30-35% of global benzene production is derived from pyrolysis gasoline. The typical pyrolysis gasoline contains up to 65% aromatics, and about 50% of which is benzene. The remainder of the product is composed of mono and diolefins, which can be removed by a mild hydrogenation step. Alternatively, pure benzene can also be recovered from the pyrolysis gasoline by solvent extraction and subsequent distillation.
Disproportionation of toluene is another route for industrial benzene production, in which two molecules of toluene can be converted into one molecule of benzene and one molecule of mixed-xylene isomers. In industry, economic feasibility of this process is highly dependent on the relative prices of benzene, toluene and xylene, as well as the excess of toluene and strong demand for benzene. In recent years, the prices of xylene and benzene have generally been higher than that of toluene, which presently makes the disproportionation process an attractive alternative to hydrodealkylation.
The industrial supply of benzene can be achieved by high-temperature carbonization of coal in coke ovens. Carbonization processes produce coke or char as the main product, and tar, light oil, gas and aqueous liquor as the by-products. The light oils contain benzene and its homologs in significant concentration. Besides, the light oils produced in the partial gasification processes can supply benzene and its homologs in considerable concentrations. Pyrolysis process of coal which falls under partial liquefaction, produces more oil than carbonization processes, and the oil can subsequently be converted into benzene and its homologs. Complete liquefaction processes are further divided into hydrogenation and solvent extraction processes, and the benzene product can also be obtained from these two categories.

Entry
Catalytic system a t (h)  : the inner potential correction. f R factor: goodness of fit. Ѕ0 2 were set as 0.90 and 0.86, respectively for Ru-W/Ru-Ru and W-Ru/W-W, which were obtained from the experimental EXAFS fit of Ru foil and W foil references by fixing CN as the known crystallographic value and was fixed to all the samples. Error bounds (accuracies) that characterize the structural parameters obtained by EXAFS spectroscopy were estimated as N ± 15%; R ± 0.02 Å; σ 2 ± 5%; ΔE0± 20%. Cα, the C atom at the aliphatic α-C position. b Cβ, the C atom at the aliphatic β-C position. c Cγ, the C atom at the aliphatic γ-C position. d CMe, the C atom in the methoxy group. Table 7. Solid-state 2D 13 C{ 1 H} HETCOR NMR chemical shifts of the intermediates in the reaction of 1-(4methoxyphenyl)-1-propanol (1a).

Samples
C α c C β d C β1 e C β2 f C β3 g Cβ1, the C atom at the aliphatic β1-C position of the intermediates in the reaction of 1a. f Cβ2, the C atom at the aliphatic β2-C position of the intermediates in the reaction of 1a. g Cβ3, the C atom at the aliphatic β3-C position of the intermediates in the reaction of 1a. h Cγ, the C atom at the aliphatic γ-C position. i CMe, the C atom in the methoxy group. Eq. (2). b Methoxy group/benzene ring ratio, the ratio of the quantity of methoxy group substituted on the benzene ring to the quantity of benzene ring in lignin. The ratio is calculated based on the quantitative analysis of the C atom in the methoxy group (Supplementary Figures 15, 17 to 24), using Eq. (3) as follows: Quantity of C atoms in the methoxy groups Methoxy group/benzene ring ratio Quantity of C atoms in the benzene ring = s 6  Eq. (3). c Yields of benzene product provided are the averages of three experiments conducted in parallel. Reaction conditions: lignin (0.50 g), RuW/HY30 (0.50 g, 3.5 wt% Ru, 20 wt% W), H2O (6.5 ml), 240 º C, 12 h, 0.1 MPa N2, 800 rpm.