High consistency enzymatic pretreatment of eucalyptus and softwood kraft fibres for regenerated fibre products

Sustainability of regenerated cellulosic fibres could be improved by using paper grade pulp instead of dissolving pulp as a raw material in the fibre spinning process. However, the use of paper grade pulp calls for adjustment of the molar mass distribution (MMD) prior to dissolution to obtain good solubility and spinnability. The objective of this work was to adjust MMD of softwood and eucalyptus kraft pulps by enzymatic treatments at high pulp consistency. The reduction of the MMD of eucalyptus kraft pulp was found to require a nearly 30-fold higher dose of endoglucanase compared to the treatment of softwood pulp. Interestingly, when xylanase was used in combination with endoglucanase to treat eucalyptus kraft pulp, 27% of the xylan was dissolved and the required endoglucanase dose could be decreased from 0.57 to 0.06 mg/g. The endoglucanase dose could be further decreased to 0.028 mg/g when 67% of xylan was removed chemically before the enzymatic treatment. This suggests that xylan hinders endoglucanase action on eucalyptus kraft pulp. For softwood pulp, the addition of xylanase and mannanase had only a minor impact on the treatment efficiency. The different processabilities of softwood and eucalyptus kraft pulps are suggested to originate from the deviating cellulose accessibility which is affected by the fibre structures as well as their hemicellulose composition and localisation. The effect of the treatment consistency was further studied with softwood kraft pulp. Treatment at high consistency clearly enhanced the endoglucanase action whereas the effect of solid content on the hemicellulase action was modest.


Introduction
Broader raw material base for regenerated textile fibres holds potential for further improving the environmental sustainability of man-made cellulosic textile fibres. The use of less processed paper grade pulp, instead of dissolving grade pulp, would decrease chemical consumption and improve resource efficiency as also hemicellulose would be utilised as part of the regenerated fibres. Expanding the raw material base for the lyocell process and even for the ionic liquid based process has been studied with promising results in recent years (Chen et al. 2015a, b;Ma et al. 2018;Kosan et al. 2020;Jadhav et al. 2021). However, the use of paper grade pulp calls for adjustment of the molar mass distribution of cellulose prior to pulp dissolution to obtain spinnable dope. Previously reported ways to adjust molar mass distribution of pulps include enzymatic treatment, acid hydrolysis, electron beam irradiation and steam explosion (Kamide et al. 1992;Graveson et al. 2002;Håkansson et al. 2005b;Köpcke et al. 2008;Palme et al. 2016). Enzymatic treatments are an environmentally benign option for adjusting the molar mass distribution of kraft pulps. The use of enzymes enables processing of pulp at close to neutral pH and at ambient temperature (40-60 °C).
Enzymes are specific catalysts that can be exploited for targeted depolymerisation of both cellulose and hemicellulose fractions in kraft pulps (Suurnäkki et al. 1997;Tenkanen et al. 1997;Köpcke et al. 2008;Rahikainen et al. 2019). In contrast to chemical pretreatments of pulp, enzymes enable selective modification of e.g., the cellulose polymer leaving the hemicelluloses intact (Rahikainen et al. 2019). The endoglucanase enzymes catalyse internal scission of cellulose chains leading to a sharp decrease in pulp viscosity and degree of polymerization of cellulose. It is commonly known that a reduction in cellulose molar mass improves cellulose reactivity (Duan et al. 2016;Wang et al. 2018), but interestingly, endoglucanase treatment has also been shown to improve reactivity more than the sole decrease in molar mass would suggest. As an example, endoglucanase treated pulp has been reported to have higher cellulose reactivity than acid or hypochlorite treated pulp at the same viscosity level (Engström et al. 2006;Duan et al. 2016). Furthermore, enzymatic pretreatment has enabled cellulose dissolution in non-toxic solvent systems such as NaOH/ZnO (Vehviläinen et al. 2008). The improved reactivity is suggested to originate from the morphological changes in the fibre walls caused by the enzymatic treatment (Engström et al. 2006;Duan et al. 2016). Traditionally, enzyme treatments have been carried out at low fibre consistency (Engström et al. 2006;Köpcke et al. 2008;Gehmayr et al. 2011). However, recently it has been shown that at high consistency conditions the endoglucanase action on fibre modification is enhanced (Grönqvist et al. 2015;Wang et al. 2015;Rahikainen et al. 2020;Banvillet et al. 2021). This effect has been attributed to the combined outcome of the enzyme action and mechanical shearing of fibres that occurs at high fibre consistency. As a result, enzyme accessibility to the fibre wall polymers is improved.
Compared to chemicals, enzymes are bulky spherical macromolecules. As enzymes can be several nm in diameter, they have restricted access to the porous fibre wall. In a study by Imai et al., a mix of cellulase enzymes was found to penetrate through the cell wall of a bleached softwood kraft fibre after 4-25 hours of incubation (Imai et al. 2019). Due to the limited accessibility, fibre type (softwood or hardwood) will also have an essential role in the processability of the pulp using enzymes. Fibre morphology and the composition, content and distribution of hemicelluloses in the fibre vary between softwoods and hardwoods and even between wood species (Alén 2011). Hemicellulose in the pulp is suggested to limit the cellulose accessibility since its removal chemically or enzymatically is shown to enhance enzymatic hydrolysis of cellulose in low consistency treatments of softwood and hardwood pulps (Várnai et al. 2011;Penttilä et al. 2013;Malgas et al. 2017;Chen et al. 2018). At low pulp consistency, the combined action of cellulases and hemicellulases on softwood pulps is shown to result in higher sugar release than the sum of the individual cellulase and hemicellulase treatments which suggests that cellulases and hemicellulases act in a synergistic manner promoting the action of each other (Várnai et al. 2011).
Previously, enzymatic treatment at high consistency has been shown to enhance endoglucanase action in the modification of softwood kraft fibres (Rahikainen et al. 2019(Rahikainen et al. , 2020. The effect of high consistency conditions on other types of kraft fibres has been studied less and the effect of high consistency on hemicellulase action has not been investigated. Therefore, this study aims to compare endoglucanase treatment on softwood and eucalyptus kraft fibres at high consistency to assess the effect of fibre type on high consistency enzymatic treatments. In addition, this study demonstrates the effect of high consistency on hemicellulase (endo-xylanase and endo-mannanase) action in fibre modification.

Pulps and enzymes
Dry sheets of bleached softwood kraft pulp with intrinsic viscosity of 860 ml/g were received from Metsä Fibre from a Finnish pulp mill. Dry sheets of bleached eucalyptus kraft pulp with intrinsic viscosity of 930 ml/g were received from Altri group (Portugal). The carbohydrate composition of the pulps was analysed by ion chromatography (Dionex ICS-3000) after a two-step sulphuric acid hydrolysis according to NREL (Sluiter et al. 2008). Compositional data of the pulps is shown in Table 1.
Hydrodynamic diameters of hemicellulase catalytic domains were estimated based on the structural models from protein data bank (mannanase 1QNR and xylanase 6K9O) using an online HullRad calculator according to Fleming and Fleming (2018).
Protein concentrations of the enzyme preparations were determined using a commercial kit (DC protein assay, Bio-Rad Laboratories, Town, Counry) based on the Lowry method (Lowry et al. 1951). Protein precipitation with acetone and re-dissolution in water containing 2% (w/v) Na 2 CO 3 and 0.4% (w/v) NaOH was applied for the commercial preparation prior to the protein assay to remove impurities that may interfere with the assay. For purified enzymes (hemicellulases) the precipitation was not used prior to the protein assay. Information on the protein concentrations was used to dose enzymes as mg of protein per g of dry pulp in the pulp treatments.

Pulp treatments
Treatments for softwood and eucalyptus kraft pulps are presented in Fig. 1. Prior to the enzymatic and chemical pulp treatments, the pulp sheets were disintegrated as follows: The sheets were shredded and soaked in reverse osmosis water (2 kg/20 l) for 2 h followed by transfer to a pilot scale wet-disintegrator equipped with pins (providing gentle disintegration) where the total volume was adjusted to 50 l. The pulp was disintegrated by mixing at 300 rpm for 15 min. Excess water was removed and the pulp was homogenised.

Enzymatic treatments at high consistency
High consistency pulp treatments were carried out using a farinograph mixer (Brabender, Germany) equipped with two z-shaped mixing blades. Prior to the treatments, enzymes diluted in buffer were evenly sprayed to the pulp manually using a spray bottle. The amount of enzymes added varied between 0.02 and 0.9 mg of protein/g of pulp. The treatments were performed at 20 or 24% (w/w) consistency with 50 or 60 g of dry pulp at 50 °C for 2 h using 25-30 rpm mixing. Sodium phosphate buffer (pH 6) was used to obtain an ionic strength of 75-100 mM in the liquid phase. Increasing ionic strength of 50-200 mM was used in an additional experiment to study electrostatic repulsion between endoglucanase and eucalyptus xylan. After the high consistency treatment, the enzyme activity was terminated by diluting the pulp with boiling water approx. to 10% (w/w) consistency followed by incubation at 100 °C for 20 minutes. The sample was vacuum filtrated with a 60 µm filtration cloth and a filtrate sample was collected to determine the dissolved sugars and oligosaccharides (yield loss). After the treatment, the pulp sample was washed with cold reverse osmosis water and homogenised with a kitchen mixer.

Enzymatic treatment at low consistency
The low consistency treatments were performed at 3% (w/w) consistency with 20 g of dry pulp at 50 °C for 2 h using 25-30 rpm mixing and sodium phosphate buffer (pH 6, 100 mM final ionic strength). The pulp was suspended in pre-heated (50 °C) buffer with a kitchen mixer and placed in a 50 °C water bath. A diluted enzyme solution was evenly added to the slurry while mixing. The amount of endoglucanase added was 0.02 mg of protein/g of pulp and hemicellulases 0-0.5 mg of protein/g of pulp. After the treatment, the pulp was vacuum filtered, and a filtrate sample was collected for the analysis of dissolved sugars. The enzyme activity was terminated by adding boiling water to the pulp followed by incubation of the pulp and the filtrate sample at 100 °C for 20 minutes. After the treatment the pulp was washed and disintegrated as described earlier.

Cold alkali extraction
Cold caustic extraction (CCE) was done for eucalyptus kraft pulp as pretreatment prior to enzymatic treatments to remove hemicelluloses. The pulp was treated in a reactor equipped with an anchor type mixer at 30 °C for 30 min at 10% (w/v) pulp consistency using 70 g/l NaOH concentration and 200 rpm mixing. Conditions for the treatment were modified from Gehmayr et al. (2011). After washing the pulp was stored at 4 °C.

Analysis of solubilised sugars in the filtrate
The composition of the solubilised sugars was analysed with high performance anion exchange chromatography with pulsed amperometric detection (HPAEC-PAD) after mild (4%) sulphuric acid hydrolysis at 120 °C for one hour. Analysis was run using a Dionex ICS-5000 for softwood samples and a Dionex ICS-3000 for eucalyptus samples. The gradient system used in the detection was based on a previously published protocol by Tenkanen and Siika-Aho (2000) with a modification on the B gradient for the softwood samples by decreasing the NaOH concentration to 10 mM. The monomeric sugars were calculated to correspond to polymeric weights as anhydrous sugars. In the case of softwood kraft pulp; xylose and arabinose were considered to originate from arabinoglucuronoxylan while mannose, galactose, and glucose corresponding to 1/3.5 of total mannose were considered to originate from galactoglucomannan and the rest of the glucose was considered to originate from cellulose. In the case of eucalyptus kraft pulp; xylose was considered to originate from glucuronoxylan while mannose and the amount of glucose corresponding to 2/3 of total mannose were considered to originate from glucomannan and the rest of the glucose was considered to originate from cellulose.

Fibre characterization
The intrinsic viscosity of the pulps was determined in duplicates with a maximum accepted deviation of 2% according to ISO 5351-1 using a PSL Rheotek instrument (Poulten, Selfe & Lee Ltd, UK). Molar mass distributions of eucalyptus kraft samples were determined with size exclusion chromatography-Multi-Angle Laser Light Scattering (SEC-MALLS) after dissolving the samples in dimethylacetamide (DMAc)/9% LiCl with the solvent exchange method based on Potthast et al. (2015). The SEC system consisted of a 1260 bioinert pump (Agilent Technologies, Waldbronn, Germany), an autosampler (G1367B; Agilent Technologies, Waldbronn, Germany), MALLS detector (Wyatt Dawn DSP with a diode laser, λ 0 = 488 nm; Wyatt Technology, Santa Barbara, US), and a refractive index detector (Shodex RI-71; Showa Denko, Japan). Four serial columns and a pre-column (Styragel HMW 6E, 7.8 × 300 mm, Waters Corporation, Milford, MA, USA) were used as the stationary phase. The SEC operating conditions were as follows: 1.00 ml/min flow rate and 100 µl injection volume. DMAc/LiCl (0.9%, w/v) filtered through a 0.02 µm filter was used as an eluent. Data were evaluated with Astra 4.73, Grams 7 and Orig-inPro 2020 software. Molar mass distributions of softwood pulp samples were determined using SEC similarly as in Rahikainen et al. (2020) after dissolution in DMAc/8% LiCl according to the solvent exchange method with additional ethyl isocyanate derivatisation (Berthold et al. 2001). The molar mass distributions were calculated against 8 pullulan (6,100-708,000 g/mol) standards and analysed with Waters Empower 3 software. Molar masses of pullulan standards were converted to correspond to the molar masses of same size cellulose chains based on the method by Berggren et al. (2003).
Morphology of the fibres was studied with optical microscopy after Congo red staining using an Olympus BX61 microscope with a ColorView 12 camera and Analysis Pro 3.1 software (Soft Imaging System GmbH, Germany). The fibre dimension analysis was done using Kajaani FibreLab analyser (Metso, Finland). Sample preparation and analysis were done as in Kangas et al. (2019).

Results and discussion
High consistency enzymatic treatment of eucalyptus and softwood kraft pulps Enzymatic modification of pulp viscosity with endoglucanase was studied with eucalyptus and softwood kraft pulps at high consistency. The effect of endoglucanase dosage (0.02-0.9 mg/g) on pulp viscosity was assessed (Fig. 2). The targeted viscosity level was 420-450 ml/g, which is shown to be optimal to obtain spinnable solutions in the Ioncell F-process (Ma et al. 2018). Interestingly, to reach the target level of viscosity, i.e. 430 ml/g, softwood kraft pulp required 0.02 mg/g of endoglucanase while eucalyptus kraft pulp required an almost 30-fold higher endoglucanase dosage (0.57 mg/g). Characteristic for both softwood and eucalyptus kraft pulps was a viscosity level-off that was reached with an increasing enzyme dosage. The intrinsic viscosity was found to level-off at approx. 380 ml/g and approx. 450 ml/g for softwood and eucalyptus kraft pulp, respectively (Fig. 2). With chemical treatments like hydrochloric acid hydrolysis intrinsic viscosity has been found to level off with time at 170 ml/g (0.4 M HCl, 80 °C) and at 130 ml/g (4 M HCl, 80 °C) with softwood kraft pulp (Palme et al. 2016) and at 200 ml/g (2 M HCl, 80 °C) with eucalyptus kraft pulp (Håkansson et al. 2005b). The lower viscosities achieved with acid treatments are likely due to the small size of the acid catalyst with better accessibility to cellulose in the fibres. However, the levelling-off viscosity of eucalyptus kraft pulp was higher and required higher doses than softwood kraft pulp with both enzymatic and acid treatments suggesting some fundamental differences in cellulose organisation in the two pulps.
Fibre length distributions of unmodified softwood and eucalyptus kraft pulps were compared to those of pulps treated with endoglucanase at high consistency Fig. 2 Intrinsic viscosity of softwood and eucalyptus kraft pulps as a function of endoglucanase dose in high consistency treatments (20-24% consistency, 2 h, 50 °C, pH 6). Maximum of 2% deviation was accepted for the replicates of intrinsic viscosity results (Fig. 3). The high consistency enzymatic pretreatment was found to cut fibres and alter the length distributions of softwood and eucalyptus kraft fibres to a very different extent. The fibre cutting as a result of the combined action of the enzymes and the shear forces at the high consistency mixing was much more prevalent in the inherently longer softwood fibres compared to the eucalyptus fibres. The treatment of softwood pulp strongly modified the shape of the length distribution; the most prevalent length fraction of the unmodified softwood pulp was the fraction of 2.2 mm fibres (Fig. 3a) whereas the ca. 0.1 mm fraction was the most prevalent after the treatment (Fig. 3b). The treatment of eucalyptus kraft pulp, which resulted in a similar decrease in intrinsic viscosity, was less efficient to fragment the fibres (Fig. 3c, d). At 20% (w/w) consistency, the viscosity of the shorter eucalyptus fibres in the treatment suspension is only half the viscosity of the long softwood fibres (Silveira et al. 2002) which probably affects the efficiency of the mechanical treatment. The lower viscosity and therefore weaker mechanical treatment of eucalyptus fibres might partly explain the need for a higher enzyme dose. Fibre cutting, seen for the softwood fibres did likely lead to improved enzyme accessibility to the polymers in the fibre wall. However, it seems that the combined action of enzymes and the mechanical forces in the studied conditions were not able to cut the eucalyptus fibres. The higher enzyme dose might be needed to compensate for the lower level of fibre structure modification.
The combined action of endoglucanase and hemicellulases in the high consistency enzymatic treatment was found to affect softwood and eucalyptus kraft pulps differently. The used hemicellulases, xylanase and mannanase, initiate the hydrolysis of intra-chain linkages of xylan (Tenkanen et al. 1992) and mannan or heteromannan (Stålbrand et al. 1993) respectively. For softwood kraft pulp, the addition of 0.1 mg/g xylanase or 0.1 mg/g of xylanase and mannanase in combination with endoglucanase did not show a clear additional decrease in the intrinsic viscosity of the pulp (Fig. 4b) and dissolved less than 20% of xylan and ca. 5% of mannan. Interestingly, Fig. 3 Weighted distributions of fibre lengths of untreated softwood (a) and eucalyptus (c) kraft pulps and endoglucanase treated softwood (b) and eucalyptus (d) kraft pulps (20% con-sistency, 50 °C, 2 h, 0.04 mg/g and 0.9 mg/g of endoglucanase for softwood and eucalyptus pulps respectively). The treated pulp samples had similar intrinsic viscosity levels when xylanase (0.1 mg/g) was added in combination with endoglucanase to treat eucalyptus kraft pulp, 27% of the xylan was dissolved and the intrinsic viscosity could be lowered to the same viscosity (430 ml/g) with a ten-fold smaller endoglucanase dose (Fig. 4a). However, the combined use of xylanase and endoglucanase was not found to fragment fibres more than a sole endoglucanase treatment when the resulting pulps had similar viscosities (data not shown). This suggests that xylan prevents endoglucanase access to cellulose in eucalyptus kraft pulp and the addition of xylanase improved the accessibility of the endoglucanase. Considerable improvement in endoglucanase action was also achieved when two thirds of xylan were removed from eucalyptus kraft pulp by cold caustic extraction (CCE) prior to the enzymatic treatment (Fig. 4a). Due to the CCE pretreatment, cellulose accessibility was increased and a similar viscosity level was achieved with a 20-fold smaller endoglucanase dose than what was required without the pretreatment. This suggests that the lower the xylan content is, the lower the required endoglucanase dose gets.
The combined action of xylanase and endoglucanase was highlighted in microscopic images taken of eucalyptus fibres after treatment with sole xylanase (Fig. 5a), sole endoglucanase (Fig. 5b) and combinations of xylanase and endoglucanase ( Fig. 5c, d). For eucalyptus fibres, enzymatic treatment at high consistency with a high dose of sole xylanase did not result in visible changes in fibre morphology (Fig. 5a), whereas treatment of eucalyptus kraft fibres with sole endoglucanase resulted in clear fibre fragmentation (Fig. 5b). The fibre fragmentation was found to enhance considerably when the combination of endoglucanase and xylanase was used (Fig. 5c, d). The fibre fragmentation is clearly caused by the endoglucanase action which weakens the cell walls and makes it more amenable to fragmentation under shear during the high consistency treatment. Clearly the use of a small dose of xylanase in combination with endoglucanase further enhances endoglucanase action in eucalyptus kraft fibres and thus disassembly of the fibre wall structure. On the contrary, the appearance of softwood kraft fibres treated with sole endoglucanase was similar to those softwood fibres treated with a combination of endoglucanase and hemicellulases (Fig. 5e, f).
The results strongly suggest that xylan acts as a barrier preventing endoglucanase access to the eucalyptus kraft fibre structure, and the xylan barrier needs to be disrupted either by enzymatic or chemical means to improve endoglucanase accessibility to the fibre wall. A similar observation was made by Gehmayr and Sixta who reported a strong Fig. 4 Intrinsic viscosity of eucalyptus (a) and softwood (b) kraft pulps as a function of endoglucanase dose in high consistency treatments (20-24% consistency, 2 h, 50 °C, pH 6). Xylanase and mannanase were used in some treatments together with endoglucanase and a cold caustic extraction (CCE) pretreatment was done before enzymatic treatment for some eucalyptus kraft pulp samples. Maximum of 2% deviation was accepted for the replicates of intrinsic viscosity results enhancement for low consistency endoglucanase treatment of eucalyptus kraft pulp after xylan degrading pretreatments (Gehmayr and Sixta 2012). In general, the hindering effect of hemicellulose is reported earlier for several lignocellulosic materials at low pulp consistency and especially the removal of xylan is found to improve the enzymatic hydrolysis of cellulose (Hu et al. 2011;Várnai et al. 2011;Penttilä et al. 2013;Malgas et al. 2017;Chen et al. 2018). However, xylan is not only limiting the action of bulky enzymes but its removal is shown to enhance acid hydrolysis of bleached birch and eucalyptus kraft pulps (Håkansson et al. 2005a, b;Borrega et al. 2018).
The hindering effect of xylan is suggested to originate from the physical barrier of the xylan layer (Öhgren et al. 2007;Hu et al. 2011). Interestingly, xylan has a classical three-fold conformation in solution but it is shown to adopt a rigid cellulose-like twofold helical conformation when it binds tightly to the surface of cellulose microfibrils by hydrogen bonds (Simmons et al. 2016;Falcoz-Vigne et al. 2017). Falcoz-Vigne et al. showed that extracted xylan can adopt this two-fold conformation when it re-adsorbs on different cellulose samples (2017). In their study, a higher proportion of xylan was found to adopt the two-fold conformation on microfibrillated cellulose (MFC) made of birch kraft pulp than on MFC made of pine sulphite pulp (Falcoz-Vigne et al. 2017), which might suggest that hardwood fibres would contain more of this tightly bonded xylan in two-fold conformation. The mechanical action deriving from the high consistency mixing might be incapable to separate the tightly bonded xylan in two-fold conformation from the surface of cellulose microfibrils. Besides the physical barrier of xylan, electrostatic repulsion between endoglucanase and xylan was suggested to hinder cellulose hydrolysis but an additional pulp treatment series with increasing ionic strength did not support the assumption.
Eucalyptus kraft pulp is reported to have xylan enriched on the fibre surfaces (Bacarin et al. 2017) while softwood fibres are thought to contain hemicellulose more uniformly distributed through the fibre walls (Suurnäkki et al. 1996;Alén 2011). Hemicellulose location in the fibres seems to at least partly explain why hemicellulose acts as a barrier for eucalyptus fibres and why the addition of xylanase and mannanase in combination with endoglucanase had Fig. 5 Optical microscopy images of eucalyptus a-d and softwood e-f kraft pulps after high consistency enzymatic treatment (20-24% consistency, 2 h, 50 °C, pH 6). The enzyme doses are shown in the images, the abbreviations endo, xyl and man correspond to endoglucanase, xylanase and mannanase respectively. The scale bar of the images is 200 µm. Intrinsic viscosities of the samples were 950 ml/g a, 420 ml/g b, 330 ml/g c, 300 ml/g d, 450 ml/g e and 430 ml/g f much less effect on pulp viscosity for softwood kraft pulp (Fig. 4). The location of xylan on the eucalyptus fibre surface is emphasized when the mechanical action is not able to open the fibre structure and overcome the xylan barrier.
Enzymatic processing of eucalyptus kraft pulp is clearly restricted by hemicellulose but other structural features in the fibre wall may also contribute to the different processability of eucalyptus and softwood fibres. This is supported by a study by Ibarra et al. (2010) which showed a smaller decrease in viscosity for low consistency endoglucanase treatment of eucalyptus dissolving pulp than for softwood dissolving pulp even when the hemicellulose contents of the pulps were under 2.4% (Ibarra et al. 2010). The difference was thought to originate from the higher cellulose crystallinity of the eucalyptus fibres as well as from the different localizations of hemicellulose on the fibres. Additionally, lignin is well known to hinder enzymatic hydrolysis by adsorping enzymes (Berlin et al. 2006;Guo et al. 2014). However, the bleached kraft pulps studied in this work had low and similar lignin contents. Therefore, lignin was assumed not to cause differences in the processabilities of the pulps.

Molar mass distributions of eucalyptus kraft pulps after high consistency enzymatic treatment
Molar mass distributions (MMD) of differently treated eucalyptus kraft pulps with similar intrinsic viscosities were compared. The aim was to study how the enzymatic and the CCE pretreatments affected the molar mass distribution of the samples. Molar mass distributions of untreated and modified eucalyptus kraft pulps revealed how the different treatments modified cellulose and xylan (Fig. 6). The unmodified eucalyptus kraft pulp showed a bimodal distribution where the larger peak in the high molar mass area corresponded to cellulose and the smaller peak in the low molar mass area mainly corresponded to xylan. The treatments caused the expected shift of the MMD to smaller molar mass values. Chain cleavage of long chains caused the desired loss in chain lengths. Interestingly, even if the endoglucanase doses in the different treatments for eucalyptus kraft pulp varied between ten and 20-fold, the losses in the high molar mass area were similar (Fig. 6). The main differences were visible on the hemicellulose shoulders. The sole endoglucanase treatment resulted in a higher shoulder in the low molar mass fraction as the endoglucanase treatment left xylan nearly intact while it fragmented cellulose into shorter chains that increased signals in this region. This is supported by analysis of dissolved sugars showing only minimal degradation of xylan (data not shown). On the other hand, the combination of endoglucanase and xylanase decreased the fraction of the hemicellulose significantly. Simultaneously, the CCE pretreatment hydrolysed and solubilised xylan so efficiently that the hemicellulose shoulder disappeared completely.
The comparison between the distributions shows how enzymes can be utilised to modify the molar mass distribution of pulp in a specific and tailored way. However, enzymes are not ideal for strong modifications such as the removal of hemicellulose because high enzyme doses would be required to compensate for lower efficiency.
Effect of pulp consistency in the treatment of softwood pulp Hemicellulase action on softwood was further studied both at high (20%) and low (3%) consistency because Fig. 6 Molar mass distributions of untreated eucalyptus kraft pulp and pulps after enzymatic treatment at high consistency (20% consistency, 2 h, 50 °C, pH 6) and cold caustic extraction (CCE) pretreatment the addition of hemicellulases did not clearly enhance the endoglucanase treatment of softwood kraft pulp at high consistency (Fig. 4). The barrier effect of hemicellulose in softwood kraft fibres was thought to be diminished at high consistency treatment due to the mechanical action and the fragmentation of softwood fibres which opened the fibre wall structure and increased the amount of accessible cellulose. The enzymatic treatments of softwood kraft pulp at high and low consistencies were carried out with a constant dose of endoglucanase and varying doses of hemicellulases. The aim was to clarify how high consistency affects the action of hemicellulases and furthermore how the hemicellulose hydrolysis improves endoglucanase accessibility to cellulose both at high and low consistency. Figure 7 shows the sugar dissolution (yield loss) during the enzymatic treatments with the different enzyme combinations. In all treatments, the degradation of cellulose in softwood kraft pulp by endoglucanase was found to be enhanced when the treatment was done at high pulp consistency (20%). The amount of dissolved sugars originating from cellulose was found to be ca. 2.7 times higher in the high consistency treatments compared to the low consistency treatments (Fig. 7). Additionally, cellulose degradation by endoglucanase was found to be slightly intensified with the addition of mannanase and xylanase. For example, the yield loss of cellulose was ca. 1.1 and 1.3 times higher with the addition of high doses (0.5 mg/g) of both hemicellulases at high and low consistency treatments, respectively. This suggests that cellulose degradation was intensified due to the partial removal of hemicellulose (Várnai et al. 2011;Gehmayr and Sixta 2012;Penttilä et al. 2013). The intensification was slightly weaker at high consistency probably because the mechanical action increases the amount of accessible cellulose by opening the fibre structure and diminishes the effect of hindering hemicellulose on the fibre surface. However, hemicellulose in softwood kraft pulp does not act as a barrier preventing endoglucanase accessibility to cellulose in the same extent as in eucalyptus kraft pulp.
Interestingly, treatment consistency was found to have a less pronounced impact on the hemicellulase action in the combined treatment with endoglucanase and hemicellulase. The high fibre consistency was found to improve hemicellulase action only with a high dose (0.5 mg/g) of mannanase or xylanase, whereas it had no effect on hemicellulose hydrolysis with a low hemicellulase dosage (0.1 mg/g) (Fig. 7). This suggests that the amount of accessible mannan and xylan was already high enough for the low hemicellulase dose and the degradation could not be enhanced by further increasing substrate accessibility.
In a combined endoglucanase and mannanase treatment, the liberation of galactoglucomannanderived sugars with high mannanase dosages (0.5 mg/g) was observed to be ca. 1.3 times higher at high consistency (Fig. 7). The overall mannan degradation was, however, found to be moderate since only 11% of the total mannan was dissolved at most. The poor liberation of mannan could derive from mannan location in the fibres and the pH condition that was not optimal for the used mannanase. Mannan content at the surface of softwood fibres has been reported to be lower than in the whole fibres (Suurnäkki et al. 1996) and its proportion is thought to increase towards the lumen (Alén 2011). Most of the mannan deeper in the fibre structure seems to remain inaccessible for enzymatic hydrolysis despite the mechanical treatment that opens the fibre structure. Previously, enzyme size Fig. 7 Sugar composition of yield loss (solubilised sugars) deriving from softwood kraft pulp treatments performed at high (20%) and low (3%) consistencies with varying enzyme doses (2 h, 50 °C, pH 6). The left bar of each section corresponds to high consistency treatment while the right bar corresponds to low consistency treatment. The dose of endoglucanase was kept constant (0.02 mg/g) while the amount of mannanase and xylanase varied from low (0.1 mg/g) to high (0.5 mg/g). The abbreviations endo, man and, xyl correspond to endoglucanase, mannanase and, xylanase respectively, and characters l and h correspond to low and high enzyme doses has been suggested to influence enzyme accessibility (Rahikainen et al. 2019). The catalytic domain of the used mannanase has a hydrodynamic diameter of 5.32 nm and it could be too large to penetrate deeper into the fibre where most of the mannan locates.
Xylan hydrolysis was found to be improved more than mannan hydrolysis when the combined endoglucanase and xylanase pulp treatments were carried out at high consistency. The effect was most apparent when high dosages (0.5 mg/g) of xylanase were used resulting in ca. two-fold higher xylan yield losses (Fig. 7). At best, 36% of the total arabinoglucuronoxylan content was dissolved. Xylan in softwood is reported to locate rather uniformly in the cell wall (Suurnäkki et al. 1996;Alén 2011). Additionally, the top layer of fibres is reported to contain deacetylated xylan that is reprecipitated on the fibres at the end of the cook (Alén 2000). However, for pine kraft pulp the xylan content on the top layer is shown to be only slightly higher (Suurnäkki et al. 1996) and the deacetylated xylan is shown to be less prone to hydrolysis of the used xylanase (Tenkanen et al. 1992). Therefore, the amount of xylan prone to enzymatic hydrolysis on the surface of the fibres was limited and probably why the mechanical action opening the fibre structure and increasing the amount of accessible xylan was found to enhance the treatment. Additionally, the used xylanase is smaller than the used mannanase as the hydrodynamic diameter of its catalytic domain is 4.16 nm. The smaller size of the xylanase could promote its accessibility to xylan in the mechanically disrupted cell wall and explain the higher yield loss of xylan over mannan together with the different localisations of the hemicelluloses in the fibre structures.
The enhanced endoglucanase action at high consistency was evident also from the pulp intrinsic viscosity data (Fig. 8) and weight-average molar masses (M w ) determined with size exclusion chromatography (SEC). Pulp viscosities were found to be ca. 200 ml/g lower when the treatments were carried out at high consistency. The high consistency treatments decreased the M w from 461 kDa to an average value of 250 kDa while the low consistency treatments resulted in an average value of 316 kDa. These data show that the high consistency conditions enhanced endoglucanase action in the fibre walls leading to intensified scission of cellulosic chains and a decrease in pulp viscosity. This supports the earlier finding that high consistency treatment enhances the action of endoglucanase in the fibre matrix (Grönqvist et al. 2015;Wang et al. 2015;Rahikainen et al. 2019Rahikainen et al. , 2020.

Conclusions
Aim of this work was to compare enzymatic processability of softwood and eucalyptus kraft pulps at high pulp consistency. Additionally, the combined action of endoglucanase and hemicellulases at the high consistency treatments was studied. Interestingly, with the sole endoglucanase, the treatment of eucalyptus kraft pulp was found to require a nearly 30-fold higher endoglucanase dose than the treatment of softwood kraft pulp. However, when xylan in eucalyptus pulp was partly removed either by chemical pretreatment or the addition of xylanase in the enzymatic treatment, the required endoglucanase dose could be drastically decreased. The data strongly suggests that xylan acts as a barrier against the endoglucanase action on eucalyptus kraft fibres. The mechanical treatment deriving from the high consistency mixing was not enough to overcome the xylan barrier in Fig. 8 Intrinsic viscosities of softwood kraft pulp samples treated at high (20%) and low (3%) consistencies with varying enzyme doses (2 h, 50 °C, pH 6). The left bar of each section corresponds to high consistency treatment while the right bar corresponds to low consistency treatment. The dose of endoglucanase was kept constant (0.02 mg/g) while the amount of mannanase and xylanase varied from low (0.1 mg/g) to high (0.5 mg/g). The abbreviations endo, man and, xyl correspond to endoglucanase, mannanase and, xylanase respectively, and characters l and h correspond to low and high enzyme doses. Maximum of 2% deviation was accepted for the replicate measurements. Intrinsic viscosity of the unmodified softwood kraft pulp was 860 ml/g the eucalyptus fibres. For softwood pulp, hemicellulose did not act as a similar barrier since the addition of xylanase and mannanase was found to have only a minor impact on the treatment efficiency. The difference in the processabilities of softwood and eucalyptus kraft pulps are suggested to originate mostly from the deviating hemicellulose composition and localisation but also from different fibre structures and dimensions that can affect the efficiency of the mechanical treatment. The effect of treatment consistency was further studied with softwood kraft pulp. Treatment at high consistency clearly enhanced the endoglucanase action whereas the effect of solid content on the hemicellulase action was modest.
Acknowledgments This research has received funding from the Bio-based Industries Joint Undertaking (JU) under the European Union's Horizon 2020 research and innovation programme under grant agreement No 837527. The JU receives support from the European Union's Horizon 2020 research and innovation programme and the Bio-based Industries Consortium. The analytical expertise of Atte Mikkelson and the technical assistance of Mariitta Svanberg and Nina Vihersola are gratefully acknowledged. Miriam Kellock is acknowledged for helping with English grammar and Anna Borisova for estimating the hydrodynamic diameters of enzymes.
Authors' contribution ES and JR wrote the main manuscript text. SG guided the writing process. AP performed analytics related to Fig. 6, draw the Fig. 6 and contributed in the text, especially in the part concerning the results related to this figure. ES draw all other figures. All authors reviewed the manuscript.
Funding Open Access funding provided by Technical Research Centre of Finland (VTT). This research has received funding from the Bio-based Industries Joint Undertaking (JU) under the European Union's Horizon 2020 research and innovation programme under grant agreement No 837527. The JU receives support from the European Union's Horizon 2020 research and innovation programme and the Bio-based Industries Consortium.

Conflict of interest
The authors have no relevant financial or non-financial interests to disclose.

Consent for publication All authors have given their consent for publishing
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.