Selection and pre-testing of enzymes
The goal of enzyme pre-selection was to set up an objective variety of 10 to 20 preparations for functional application pre-testing (DP reduction test). This aimed at high opportunity of finding suitable preparations within a reasonable number of application pre-tests. On the basis of their enzyme activities, the preparations of the collection were classified in 9 activity profiles. The profiles were very distinct from each other and contained 5–10 preparations with similar ratios of activities. The targeted diversity of preparations was sufficient and adequate.
A subset of 19 preparations was thus chosen in different profiles and from various microbial and industrial origins. When tested towards chromatographic paper, the studied enzyme preparations reduced the DP from 0% to 27.4%. Seven preparations produced a DP-reduction of more than 21%. Three of them were active at low to moderate dosage (0.3%—1%). From the application pre-tests results, availability and diversity, the enzyme samples P4, P10 and P11 were chosen for application tests of enzymatic pulp modification. The activities of these preparations as well as their performance in application pre-tests are presented in Table 3.
Protein content was highest for P4, middle for P10 and lowest for P11. CMCase and ß-glucanase varied in the same way. Xylanase was highest in P10. Pectinase was low for all selected samples. Remarkably, no FPU was detectable in P11 even though this preparation displayed the highest DP reduction.
The preparations P4, P10 and P11 were compared to the pre-selected samples using their ranks of protein specific activities within the complete collection. Ranks ranged in the same interval for all activities and were thus more convenient than activities for comparison. Specific activity ranks were displayed in 5-ray spider charts, where each ray represented the activity rank for FPU, CMCase, β-glucanase, xylanase and pectinase, respectively (Fig. 1). With each preparation represented by a pentagon, the shape of the pentagon was used to classify the preparations in different categories (Fig. 1b-f).
In Fig. 1a, the selected preparations P4, P10 and P11 were displayed. In many cases, similar enzymatic activities combinations corresponded to similar DP reduction performance. Three preparations similar to P4 (Fig. 1b) performed at middle level in the range [17–18%]. In the category of P11 (Fig. 1c) characterized by undetectable FPU and pectinase, three preparations performed in the range [23–25%]. In this group, P13 yielded only 10% DP-reduction. This preparation was produced by a bacterium for brewery applications. P14 displayed a similar profile to P10 (Fig. 1d) but produced a DP-reduction of only 11%. In Fig. 1e three preparations focusing on Xylanase/CMCase were displayed. In this group, performance correlated to beta-glucanase (P17: 22%; P2: 17%; P20: 0%). Further preparations tested performed poorly or not at all (Fig. 1f, P5: 15%; P8: 12%; P16: 0%, further not shown).
From these observations of biochemical and functional tests, trends could be derived. DP reduction was performed by preparations displaying diverse enzymatic activity combinations. The presence of FPU was not necessary for DP-reduction. Xylanase activity was not sufficient. Presence of beta-glucanase activity was a good indicator for DP-reduction performance, but not sufficient.
The DP reduction from solid cellulose may be performed by catalytic proteins which are partially detected by standard biochemical measurement using beta-glucan soluble substrate. This assessment is suitable for the detection of endo-glucanase activities as a sum. The example of P11 which performs best with poor specific activities suggests that this measurement is not specific for DP reduction activity. The active proteins performing DP reduction may belong to further categories not assessed in this study. Indeed, even after decades of investigations, cellulase research is still in progress (Kostylev and Wilson 2012). In recent years, new classes of cellulases enzymes were discovered. For example, Vermaas et al. (2015) reported the discovery of a new class of enzymes that utilize an oxidative mechanism to cleave glycosidic linkages called lytic polysaccharide monooxygenases (LPMO). As far as known by the authors, such activities are not available at industrial scale. A sensitive, specific and miniaturized test for DP reduction would help in developing a molecular approach of this process, necessary for the directed development of cellulase complexes with better performance.
Investigations for enzymatic treatment of NH-P pulp sample and following dissolution and spinning tests
Enzymatic modification of the pulp sample NH-P
Because of its very wide molecular weight distribution (MWD) containing very high molecular cellulose parts, a direct usage of the paper pulp sample NH-P for Lyocell applications is not possible. The solubility as well as the accessibility of the pulp sample for the process solvent NMMO is too low to achieve acceptable solution states for dry–wet spinning to cellulose fibres. Enzymatic treatments were carried out using 0.3% enzyme sample P4, 1% P10 as well as 1% P11. This enzymatic pulp modification resulted in a suitable reduction of the Cuoxam-DP of the pulp sample and especially in a significant decrease of the polydispersity of the modified pulp samples. The results from SEC characterisation and the measured Cuoxam-DP of the modified samples in comparison to the starting NH-P pulp sample as well as to the Lyocell pulp sample LYO are listed in Table 4. Figure 2 contains the graph with the MWD of the samples.
The starting pulp sample NH-P is characterised by bimodal MWD. The determined polydispersity was very high (Mw/Mn: 12.1). The enzymatic treatment resulted in significant decrease of the polydispersities of the modified pulp samples (Mw/Mn: 4.4–4.7). The MWD of all three enzymatic treated samples are very similar and equivalent to the comparative LYO pulp sample. The weight average molecular weights Mw and z-average molecular weights Mz of the pulp samples were more than halved by the enzymatic procedure. By contrast, the number average molecular weight Mn of the enzymatic treated samples was similar or slightly increased compared to the initial pulp sample. The attained Cuoxam-DP values of the modified pulp samples were between 590 and 619, in an appropriate area for the preparation of Lyocell fibres.
Comparative dissolution and spinning tests using starting and enzymatic modified NH-P pulp samples
Dissolution tests in NMMO using the starting pulp sample NH-P did not permit the preparation of spinning dope with acceptable dope qualities. Using reduced cellulose concentration of 9% (w/w), the unmodified pulp sample resulted in a very bad solution state, containing a lot of undissolved fibre fragments (Fig. 3, left). Contrary to this, the enzymatic modified pulp samples could be used for dope preparation in laboratory scale with cellulose concentrations between 12 and 12.5% (w/w). The microscopic images of the prepared dopes showed strongly improved solution states without any undissolved fibre residuals (Fig. 3, right). Because of the high particle content of the 9% dope from the starting pulp with a large number of particles with high aspect ratio, a particle characterisation with laser diffraction was not possible. In contrast, very small particle contents (< 20 ppm) with low particle sizes (up to 51 µm in maximum) were detected by laser diffraction at the prepared dopes from enzymatic modified pulp samples.
Laboratory spinning tests using dope sample NH-P 1 from the unmodified pulp showed very unstable spinning behaviour. Alongside many capillary breaks, the spinning pressure increased from 32 to 38 bar already during the short time of the laboratory spinning test of about 30 min. This pressure increase should be caused by accumulation of particles on the safety filter device because of the worse dope quality containing undissolved fibre residuals or rather gel contents. By contrast, the spinning dopes prepared using the enzymatic modified pulp samples could be spun in a very stable manner at this laboratory scale (capillary diameter: 100 µm, 30 capillaries). The used spinning speed was 30 m/min. Further applied spinning conditions and achieved fibre properties are listed in Table 5.
The rheological properties of the prepared spinning dopes, applying the enzymatic modified NH-P pulp samples, are in a well-suitable area for Lyocell process technology.
The textile physical properties of the prepared fibres were only slightly below the level of conventional Lyocell fibres. It illustrates, that the prepared fibres from paper grade pulps showed slightly increased WRV compared with Lyocell fibres, prepared using typical Lyocell pulps (Michels and Kosan 2006).
The DP degradation between the pulp samples used for dope preparation and the fibres spun thereof, is about 8% at the modified samples and also at the unmodified sample NH-P 1. However, the evaluation of the MWD of the samples (Fig. 4) indicates that significant amounts of the high molecular chains have not been transferred into the fibres, obviously through filtration of undissolved pulp amounts. On the other hand, some low molecular pulp amounts were separated during the spinning process using the starting pulp, which could be enriched in the spinning bath. The enzymatically modified pulp samples showed already minor low molecular pulp fractions, and the MWD of the pulp and fibre samples showed significantly lower differences than in the case of unmodified pulp. For better overview, Fig. 4
contains exemplary only the results from the P11 modified pulp sample.
First upscaling tests for the enzymatic treatment (1% P11) of the pulp sample NH-P and following dope preparation were carried out for staple fibre and filament preparation. Both the enzymatic treatment and the dope preparation were carried out firstly with discontinuous batch trials. First calculations from the mass balance of the enzymatic treatment step showed very low pulp losses between 1 and 2%. The preparation of the spinning dopes was possible using dissolution conditions comparable to typical Lyocell pulps. Semi-technical spinning tests (6 × 80 capillaries, 90 µm outlet diameter) for investigation of the spinning behaviour and stability as well as the preparation of staple fibres and multifilament yarn were carried out using 12% (w/w) cellulose dopes. Solvent recycling of the spinning bathes was carried out without significant changes to spinning dopes from usual Lyocell pulps. Further investigations, especially after longer continuous trials, are necessary for further investigations concerning the influence of pulp impurities during continuous processing. The prepared fibre samples showed good textile-physical properties. The results of these tests were discussed in detail in former publication (Kosan and Meister 2018).