Caprolactam from renewable resources: catalytic conversion of 5-hydroxymethylfurfural into caprolactone.

Lignocellulosic biomass is a very promising feedstock for the production of biobased chemicals. The C6 sugars (for example d-glucose, d-fructose, and d-mannose) in lignocellulosic biomass are interesting precursors for a broad range of chemicals with high application potential. Apart from fermentation to bioethanol and reforming to CO/H2, [10] the direct conversion of these sugars to useful platform chemicals is highly attractive. Examples of such chemicals are levulinic acid and 5-hydroxymethylfurfural (HMF). HMF can be prepared in high yield from d-fructose, although research is underway to convert d-glucose or even cellulose directly into HMF. It can be converted into a range of derivatives with potential applications as a biofuel (furanics) and as building blocks for the polymer and solvent industry. Herein, we present our work on the conversion of HMF into caprolactam, the monomer for nylon-6, a widely used synthetic polymer with an annual production of about 4 million tons. The proposed reaction for the conversion of HMF into caprolactone, via 1,6-hexanediol (1,6-HD), is shown in Scheme 1. The conversion of caprolactone into caprolactam by the reaction with ammonia is well-established and has already been used on a production scale. A major breakthrough, needed in this research is the conversion of HMF to 1,6-hexanediol. For the feasibility of a bulk chemical process, it is absolutely essential that all conversions proceed with a selectivity in excess of 90%, and preferably even higher. High conversion is desirable, but not a prerequisite, and indeed many bulk processes, and in particular oxidations, are run at very low conversions to maintain a high selectivity. Four different routes were explored involving catalytic hydrogenation and hydrodeoxygenation reactions with various homogeneous and heterogeneous catalysts: 1) The direct hydrogenation of HMF to 1,6-HD; 2) a two-step sequence via 2,5-THF-dimethanol (THFDM); 3) a three-step synthesis via THFDM and 1,2,6-hexanetriol (1,2,6-HT); and 4) a four-step synthesis via THFDM, 1,2,6-HT, and tetrahydro-2H-pyran-2ylmethanol (2-THPM). The last step in the sequence, namely the catalytic conversion of diols into lactones, is a known reaction, but the conversion of 1,6-HD into caprolactone (5) rarely proceeds with high selectivity. Probably the best method in terms of yield and selectivity is the oxidation with 30% H2O2 using heteropolyacids as catalyst which was reported twice and gave caprolactone in 70% and 98% yields, respectively. However, the use of H2O2 may be too expensive for a bulk caprolactam process. Herein, we report a version based on an Oppenauer oxidation that has never been used before on this substrate. The one-step hydrogenation reaction of HMF to 1,6-HD was performed under severe conditions (270 8C, 150 bar), with hydrogen as the reductant and a mixture of copper chromite and Pd/C (1:0.6) as the catalyst following a synthetic procedure reported by Utne and co-workers. After 16 h reaction time, the HMF conversion was 100% and a mixture of products was obtained. The main product was THFDM; the desired product 1,6-HD was present in less than 4% yield. Use of just CuCr or Pd/C led to worse results. Rather worrying was that also some C5 products, such as 1,5pentanediol, were found. A possible pathway towards C5 compounds is by decarbonylation of the aldehyde group. For this reason, it was deemed wiser to first hydrogenate HMF to THFDM under milder conditions and then hydrogenate this compound in a second step to 1,6-HD. The catalytic hydrogenation of HMF to THFDMhas been reported using supported metal catalysts. A catalyst screening study was performed using a variety of catalysts and Raney-Ni (10 wt% catalyst intake, 100 8C, 90 bar hydrogen, 14 h) gave essentially quantitative yields of THFDM (cis/ Scheme 1. Synthetic routes for the conversion of HMF into caprolactam.