Unravelling the enigma of ligninOX: can the oxidation of lignin be controlled?

As societal challenges go, the development of efficient biorefineries as a means of reducing our dependence on petroleum refineries is high on the list.

S2 e one possibility is that the b-b unit has been converted to the previously proposed pyran-4one structure (see references [1][2][3]. Additional transformations are possible and are the subject of on-going work in our groups. f Whilst only 40% of the spent DDQ was recovered as DDQ-H 2 (which can be converted back to DDQ using MnO 2 (an excess in Et 2 O, stirred at r.t. for 2 hours and collected by filtration and concentration of filtrate to give DDQ in a ca. 54% yield based on two repeats), the residual 60% (~54% recovered as a crude mixture, 43% isolated as DDQ through a CHCl 3 recrystallization of the crude mixture) is recoverable from the diethyl ether precipitation step required to precipitate lignin a-OX through concentration of the filtrate. All materials were used as received unless otherwise stated. Diisopropylamine was distilled prior to use and stored over KOH. n-BuLi was titrated using diphenylacetic acid. Dry solvents were acquired from an MBRAUN (MB-SPS-800) dry solvent purification machine. 1 H NMR and 13 C NMR was performed on a Bruker Ascend 400 MHz, Bruker Avance 500 MHz, Bruker Avance III 500 with nitrogen cooled broadband probe or Bruker AVANCE III HD 700 5mm 1 H, 13 C, 15 N triple resonance inverse nitrogen cooled probe (TCI Prodigy) spectrometer with solvent peak used as internal standard. Multiplicities reported as following: s = singlet, d = doublet, t = triplet, q = quartet and m = multiplet and J values are reported in Hz. Column chromatography was conducted using Davisil ® silica (40-63 μm, 230-400 mesh) on a Biotage Isolera One Spektra system with ACI. Thin layer chromatography was performed using pre-coated glass plates (Silica gel 60A from Fluorochem) and visualised under UV light (254 nm) or through staining with KMnO 4, 4-dinitrophenylhydrazine (aldehydes) and FeCl 3 (phenols). IR spectra were obtained on a Shimadzu IRAffinity-1 Fourier Transform IR spectrophotometer as thin films. IR analysis was carried out using IResolution v1.50 with only characteristic peaks reported. Mass spectrometry data was acquired through the University of St Andrews School of Chemistry mass spectrometry service or EPSRC Swansea Mass Spectrometry Service. Reactions at -78 o C were achieved using acetone: dry ice. Any reactions requiring anhydrous conditions were run using oven-dried (140 o C) or flame dried glassware under N 2 atmosphere. Lignin Dioxasolv Extraction: as previously described. 4,5

DDQ Oxidation of Lignin
To a stirring solution of lignin (1 wt. eq.) in 1,4-dioxane (100 mg/ 2.33 mL) is added DDQ (varying wt. eq.). The solution is heated to 80 °C for 2 hours, cooled, filtered through a pad of celite and washed with 1,4-dioxane (1 volume). The filtrate is added dropwise to Et 2 O (10 S4 volumes) and the resulting precipitate is filtered and washed with excess Et 2 O. Lignin a-OX is dried to constant weight in a vacuum oven at 40 °C for 24 hours prior to analysis.

2D HSQC NMR Acquisition
Oven-dried lignin samples (70 mg) are dissolved in 0.6 mL of d 6 -DMSO in a 1.5 mL eppendorf tube and subjected to sonication for 10 minutes at 30 °C. Samples are centrifuged at 6000 RPM for 5 minutes. Supernatant is filtered through a 0.45 μM syringe filter into an over dried NMR tube. 2D HSQC NMR spectra were acquired on a Bruker Avance III 500 with nitrogen cooled broadband probe or Bruker AVANCE III HD 700 5mm 1 H, 13 C, 15 N triple resonance inverse nitrogen cooled probe (TCI Prodigy) spectrometer. The central DMSO solvent peak was used as internal reference (δC 39.5, δH 2.49 ppm). The 1H, 13C-HSQC experiment was acquired using standard Bruker pulse sequence 'hsqcetgpsp.2' (phase-sensitive gradientedited-2D HSQC using adiabatic pulses for inversion and refocusing). Composite pulse sequence 'garp4' was used for broadband decoupling during acquisition. 2048 data points was acquired over 12 ppm spectral width (acquisition time 170 ms) in F2 dimension using 24 scans with 1 s interscan delay and the d4 delay was set to 1.8 ms (1/4J, J = 140 Hz). A spectral width of 86 ppm (47-133 ppm) and 128 increments were acquired in F1 dimension (acquisition time 5.9 ms). The spectrum was processed using squared cosinebell in both dimensions and LPfc linear prediction (32 coefficients) in F1. Volume integration of cross peaks in the HSQC spectra was carried out using MestReNova 11.0 for Mac software and figures were prepared using Adobe Illustrator CS6 for Mac for spectral annotation.

S5
Scheme S1: Westwood group methodology for one-pot lignin depolymerisation. 5 Lignin is initially oxidised selectively using a catalytic DDQ system followed by a stoichiometric zinc mediated reductive cleavage of the C-O-aryl ether bond. The reductive cleavage of the C-O-aryl bond leads to the generation of a lignin-bound keto-alcohol, a monomeric keto-alcohol and a free phenol end-group on a lignin chain.  From the temperature screen, the highest amount of the b-O-4 linkage in situations (iv) or (v) was achieved at 80 °C and 100 °C, both yielding comparable results. 80°C was therefore carried through to all the following reactions.

A'
A" S11 should allow for more monomers to be produced by a C-O cleavage-mediated depolymerisation strategy (e.g. Scheme S1). The choice of non-alcoholic solvents also prevented any chance of incorporation of solvent molecules into the lignin structure.

S12
Throughout the course of this study, pre-defined regions were used for all integration on MestReNova 11.0 for Mac.

MestReNova Integral Regions
Minor structures in hardwood lignin α-9: (discussed here)  This discussion is based on the data presented in Figure 3 in the manuscript and reproduced in Figure Figure S10). To examine this in more detail, a competitive DDQ oxidation of model compounds S1 and S2 was conducted (Scheme S2   Table S4 and S5.  ), a value is obtained (in triplicate) for the total aromatic integral obtained from 2D HSQC NMR analysis using the integral regions described below Figure S5. The second grouping  ), a value is obtained for the total aromatic integral obtained from 2D HSQC NMR analysis using the integral regions described below Figure S5. The second grouping    NB. Whilst a 1: 2.5 mixture of S1 to S2 does not represent the exact S:G ratio found in the hardwood lignin we were using, enough of S1 was needed to ensure signal was acquired by NMR.    From the competitive DDQ oxidation reaction with models S1 and S2 (Scheme S2), it would be predicted that at any given time point, the conversion of S1, the G-G b-O-4 model, would be higher due to better stability of the intermediate carbocation (S1a vs. S2a, Figure S15

Pseudo-2D DOSY NMR Analysis of Compound 2b
Initial analysis of 2b led to two possible structures being proposed, a 4-membered ring and an 8-membered ring ( Figure S17h). In an attempt to distinguish between the two structures, molecular weight prediction using DOSY NMR (a non-destructive technique) was conducted 16h; (g) I 2 , acetone, r.t. 16h. R 2 = CO 2 Et.