CHARACTERISATION OF PULP CARBOHYDRATES BY ENZYMATIC HYDROLYSIS AND DETERMINATION OF PULPING YIELD WITH CARBOHYDRATE PROFILES David Vaaler 1 , Kristin Syverud 2 and Størker T. Moe 1 1 Norwegian University of Science and Technology (NTNU), Departement of Chemical Engineering, Sem Sælands vei 4, NO-7491 Trondheim, Norway 2 Norwegian Pulp and Paper Research Institute (PFI), Høgskoleringen 6B, NO-7491 Trondheim, Norway ABSTRACT A method for determination of carbohydrate composition of glucan, xylan and mannan in pulps has been developed. The carbohydrates are first enzymatic hydrolysed using a mixture of three commercial enzymes. Then the hydrolysate is divided into three equal parts and hydrolysed further with different concentrations of acid (triflouroacetic acid). Neutral monosaccharides are quantified in three analytical runs on a high liquid performance chromatograph (HPLC) system. The total yield of carbohydrates is high (88 - 98% on o.d. lignin-free pulp) and the repeatability is about 0.2 – 0.5% on o.d. pulp lignin-free pulp (95% confidence interval) for each anhydrosaccharides. For Norway Spruce, the content of mannan correlates well with lignin-free pulp yield. When anthraquinone (AQ) is used, the increase of yield is due to both cellulose and glucomannan. On the other hand, when PS or H 2 S is applied, yield increases mainly glucomannan. As a result, the gradient of slope is higher for pulps with AQ than pulps without AQ. INTRODUCTION Accurate determination of carbohydrate profiles in wood and pulp samples with traditional acid hydrolysis can be a difficult and labour-intensive process (1). The polysaccharides in the pulp have to be hydrolysed to monosaccharides that can be detected and quantified by a suitable method. Acidic hydrolysis of fibers with strong mineral acids can also cause non-controllable hydrolysis yield losses (2). Several types of detectors have to make derivatives of monosaccharides. The yield of derivatisation can be less than 100% and incomplete. Recently, several improved methods for carbohydrate analysis of chemical pulps have been published. These methods involve the use of enzymes for depolymerization of the pulp (3,4,5,6). An ideal analysis of carbohydrates would give high and accurate hydrolysis yield i.e. no degradation of monosaccharides, no need for derivatisation and utilise simple and inexpensive equipment. The scope of this work is to try to satisfy most of these requirements utilising commercial enzymes in combination with acid hydrolysis and simple HPLC analysis. The pulping yield is usually hard to determine due to difficulties associated with practical measurements in an industrial pulping mill. Estimates of pulping yield are therefore usually obtained by 3 to 6 months of manufacturing of pulp. Consequently, it is desirable to estimate the yield using the chemical composition of the pulp. Marcoccia et al. (7) have published an interesting method for estimation the pulping yield for northern and southern hardwood. The parameters in the model were viscosity and the cellulose content of the pulp. This method has been further applied on Norway Spruce and simplified using the new carbohydrate analysis. EXPERIMENTAL PULP The pulping procedure is described by Vaaler and Moe (8). Four different pulps were studied in the characterisation method. These include laboratory high-yield unbleached polysulphide/anthraquinone (PS/AQ), laboratory unbleached kraft pulp (Kraft), industrial bleached softwood (Södra 90) and hardwood (Södra EUC) pulps. Properties of the pulps are given in Table 1. Table 1 Pulps used in the characterisation study Pulp Pulp type Addition (% on o.d. wood) Raw materials Kappa PS/AQ Laboratory 0.15 % AQ and 2.0% PS Norway Spruce 29.2 Kraft Laboratory - Norway Spruce 31.7 Södra 90 Industrial unknown quantity of AQ 50% Norway Spruce and 50% Scots Pine ECF (<< 1) Södra EUC Industrial unknown quantity of AQ Eucalyptus spp. ECF (<< 1)