ABSTRACT
Abstract Catalyst deactivation during consecutive lactose and xylose hydrogenation batches over Mo promoted sponge nickel (Activated Metals) and Ru(5%)/C (Johnson Matthey) catalysts were studied. Deactivation over sponge nickel occurred faster than on Ru/C in both cases. Product selectivities were high (between 97 and 100%) over both catalysts. However, related to the amount of active metal on the catalyst, ruthenium had a substantially higher catalytic activity compared to nickel. Introduction The importance of catalyst stability is often underestimated not only in academia but also in many sectors of industry, notably in the fine chemicals industry, where high selectivities are the main objective (1). Catalyst deactivation is inevitable, but it can be retarded and some of its consequences avoided (2). Deactivation itself is a complex phenomenon. For instance, active sites might be poisoned by feed impurities, reactants, intermediates and products (3). Other causes of catalyst deactivation are particle sintering, metal and support leaching, attrition and deposition of inactive materials on the catalyst surface (4). Catalyst poisons are usually substances, whose interaction with the active surface sites is very strong and irreversible, whereas inhibitors generally weakly and reversibly adsorb on the catalyst surface. Selective poisons are sometimes used intentionally to adjust the selectivity of a particular reaction (2). Catalyst deactivation often plays a central role in manufacturing of various alimentary products. Sugar alcohols, such as xylitol, sorbitol and lactitol, are industrially most commonly prepared by catalytic hydrogenation of corresponding sugar aldehydes over sponge nickel and ruthenium on carbon catalysts (5-10). However, catalyst deactivation may be severe under non-optimized process conditions.
Experimental Section Aqueous lactose (40 wt-% in water) and xylose (50 wt-%) solutions were hydrogenated batchwise in a three-phase laboratory reactor (Parr Co.). Reactions with lactose were carried out at 120 û C and 5.0 MPa H2. Xylose hydrogenations were performed at 110 û C and 5.0 MPa. The stirring rate was 1800 rpm in all of the experiments to operate at the kinetically controlled regime. For lactose hydrogenations were used 5 wt-% (dry weight) sponge nickel and 2 wt-% (dry weight) Ru/C catalyst of lactose amount. In case of xylose, 2.5 wt-% (dry weight) sponge nickel and 1.5 wt-% (dry weight) Ru/C catalyst of xylose amount were used. Prior to the first hydrogenation batch, the Ru/C catalyst was reduced in the reactor under hydrogen flow at 200 û C for 2 h (1.0 MPa H2, heating and cooling rate 5 û C/min). The reactor contents were analysed off-line with an HPLC, equipped with a Biorad Aminex HPX-87C carbohydrate column. Results and discussion Xylose hydrogenation gave xylitol as a main product (selectivity typically over 99 %) and arabinitol, xylulose and xylonic acid as by-products. In lactose hydrogenation, the main product was lactitol (selectivity typically between 97 and 99 %) and lactulitol, galactitol, sorbitol and lactobionic acid were obtained as by-products. Studies about xylose hydrogenation to xylitol suggested that the main reasons for the sponge nickel deactivation were the decay of accessible active sites through the accumulation of organic species in the catalyst pores and by poisoning of the nickel surface. Deactivation during consecutive xylose hydrogenation batches over Ru/C catalyst was insignificant (Fig. 1A). Catalyst deactivation during consecutive lactose hydrogenation batches occurred faster than during the xylitol manufacture (Fig. 1B).
