ABSTRACT
Abstract Chemical modification of lignin without degradation of its polymeric structure could lower the production cost of higher quality paper, particularly with regard to stabilization toward light-induced yellowing, where the presence of so-called α-carbonyls and α-hydroxyl groups is of key importance. In order to explore which chromophore groups can be modified, hydrogenation of aromatic lignin model compounds has been tested, using as catalysts in-situ, water-soluble ruthenium species containing tris(hydroxymethyl)phosphine, P(CH2OH)3. Such systems are active for hydrogenation of activated alkenes, carbonyl groups and, of more interest, hydrogenolysis of some aromatic alcohols. Catalyst activity depends on the phosphine/Ru ratio, the nature of the substrate (particularly substituents of the aromatic ring), and the solvents used. Thus, hydrogenation of a >C=O functionality can give either >CH(OH) or >CH2 depending on the nature of the arene fragment. Introduction The pulp and paper industry is a multibillion-dollar business. Production of high quality printing paper involves the use of low-yield, lignin-free, chemical pulp, while employment of less expensive, high-yield, lignin-rich, mechanical pulp leads to lower quality paper (1). As a sequel to our efforts to find an improved industrial process for the bleaching and brightness stabilizing of lignin-rich mechanical wood pulps, we are developing methods for catalytic hydrogenation of such pulps using H2O-soluble metal complexes (1-6). Hydrogenation using heterogeneous and homogeneous catalysts has been used previously to convert lignin to commercially valuable chemicals and for investigation of lignin structures, but usually at severe conditions of temperature and pressure, with accompanying significant hydrogenolysis/cleavage of substituent side-chains (7, 8). Studies by our group, using milder conditions, have shown that Ru
complexes containing H2O-soluble, sulfonated phosphines (2), and Ru-arene (2) or n-trioctylamine (4) systems, can catalyze the hydrogenation of alkenyl and carbonyl bonds, and aromatics, in lignin model compounds (LMCs), and in milled wood lignin without degradation of its polymeric structure. We then demonstrated that use of in situ Ru/tris(hydroxymethyl)phosphine (THP) [Ru/P(CH2OH)3] hydrogenation catalysts led to a significant bleaching of mechanical pulps (1). In order to understand which chromophoric groups can be hydrogenated with Ru/THP/H2 catalysts, we have now studied such systems for the hydrogenation of LMCs. In this paper, results on hydrogenation of C=O and C=C bonds, and the hydrogenolysis of alcohols in such model compounds, are presented. More generally, reports on transition metal complexes containing THP or other hydroxymethyl functionalized phosphines have been limited (915), and their application in biphasic or aqueous catalysis has been rare (10-13). Experimental Section Solid P(CH2OH)3 (THP, > 90% purity) and [P(CH2OH)4]Cl (THPC, 80% aq. solution) were purchased from Strem Chemicals and Aldrich, respectively; THP was also prepared in similar purity from THPC by a literature procedure (16). Lignin model substrates 1, 3, 5, 6, 10 and 13 (see eqs. 2-6, below), were used as received from Aldrich. Carbinols 11b and 11c (eq. 5) were prepared from a reported borohydride reduction of 10b and 10c, respectively (17). Solutions of in situ catalyst precursors were prepared via a ligand exchange reaction between RuCl2(PPh3)3 and THP, ideally exemplified by eq. 1 (but see Results and Discussion): a selected amount of THP dissolved in deoxygenated H2O (25 mL) was added dropwise via a cannula to a brown solution of RuCl2(PPh3)3 (18) (785 mg, 0.82 mmol) in deoxygenated CH2Cl2 (25 mL) in a Schlenk tube, and the resulting mixture stirred at room temperature (r.t., ~20 ºC) for 2 h. The yellowbrown aqueous layer was then decanted from the almost colourless (n > 3) or light brown (n ≤ 3) CH2Cl2, which was washed with H2O (3 x 15 mL); the water fractions were combined, and the volume typically made up to 100 mL by additional H2O. This solution (containing ~8.2 mM Ru) was filtered in air, and stored in a dark glass flask; solutions with THP/Ru ratios of 0.5 - 6 were prepared. RuCl2(PPh3)3 + n THP ⎯→ n PPh3 + RuCl2(THP)n(PPh3)3-n ; n = 1-3 (1)
Catalytic hydrogenations were performed generally under 500 psig of H2 at 25 or 80 °C in H2O or 1:1 H2O/EtOH media, typically with a total volume of 5 mL; the substrates were essentially insoluble in water, but dissolved readily in the mixed medium. H2O/NaOH and H2O/buffer imply, respectively, use of a 1:1 mixture of the Ru catalyst solution with either 5.0 mM aq. NaOH or a pH 10 buffer solution (made from 44 mM NaHCO3 and 17.6 mM NaOH (19)). Catalyst concentrations were either 4.1 or 2.05 mM, while the substrate (S) concentration was always 100 mM. The reaction mixture was placed into a glass sleeve equipped with a magnetic stir bar in air. The sleeve was placed in a steel
autoclave that was then charged at r.t. with H2 (pressures throughout are given in psig). The reaction mixture was stirred at the required temperature for selected times, after which the H2 was released and the products extracted with 4:1 ethyl acetate/acetone (3 x 10 mL); analysis was done by GC (HP-17; 100 °C for 2 min, 10 °C/min to 220 °C, and 3 min at 220 °C) and/or 1H NMR. Hydrogenated products were identified either by GC-MS or GC co-injection of the commercially available or isolated and fully characterized material. NMR spectra at r.t. were recorded on a Bruker AV300 spectrometer, with residual protons of deuterated solvents (1H, relative to external SiMe4) and external P(OMe)3 being used as references; 31P data are reported relative to 85% aq. H3PO4. UV-Vis spectra were recorded on an HP 8452A Diode Array spectrophotometer. Results and Discussion Nature of the Ru/THP species. Our attempts to characterize the precursor catalyst species have been unsuccessful, as the systems are complicated. Eq. 1 shows a hypothetical, idealized situation with increasing substitution of PPh3 by THP. In-situ 31P{1H} NMR data on the extracted aqueous (D2O) layer from the synthesis at 3:1 THP/Ru with [Ru] ~ 10-2 M do show complete consumption of the added THP (δP –23) and generation of overlapping singlets and multiplets in the δP 18.3-16.3 region; free THP is first seen at a THP:Ru ratio of ~4, implying that 3-4 THP ligands are involved in the chemistry. Of note, another group (14) has isolated the complex Ru(THP)2[PH(CH2OH)2]2Cl2 (A) from the exchange reaction shown in eq. 1, in a 2-phase, H2O/CH2Cl2 system. A was structurally characterized; 2 of the THP groups have eliminated formaldehyde to give the product, which shows the expected 2 triplet, A2B231P{1H} pattern at δP 13.5 and 9.7 (14). Unfortunately, experimental details were not presented (14), but we note that our signal for the monooxide of THP (δP = 50), formed when solutions with THP:Ru > 4 are handled in air, is 5 ppm less than that given in ref. 14, so species A may well be present in our systems. Mixtures of species are certainly present depending on the THP:Ru ratio, but we were unable to form solutions containing just a single species even at high THP:Ru ratios. Further, solutions of the species are somewhat photo-sensitive, and are not stable over 24 h in daylight in the presence or absence of air. Nevertheless, Beer’s Law was obeyed in the UV-Vis for freshly made solutions up to ~ 10-2 M [Ru] in air: λmax 466 nm (ε = 140 M-1 cm-1) at THP/Ru = 3, and λmax 448 (ε = 95) at THP/Ru = 10. Our attempts to isolate characterizable complexes were unsuccessful. Complex A has also been isolated from an EtOH solution of RuCl3 and THP at r.t. (14), while refluxing of the same solution has yielded the structurally characterized, dimeric complex [Ru(µ-PCH2OH){µ-P,O-P(CH2O)(CH2OH)2}Cl2]2,which contains bridged phosphido groups and bridged monoalkoxides derived from THP (20). The chemistry of Ru/THP solutions is clearly complex.
