High calcium content of Eucalyptus dunnii wood affects delignification and polysaccharide degradation in kraft pulping

Introduction

Kraft pulping has in the last 50 years definitely established itself as the dominating technique for pulping of wood (Ragnar et al. 2013). It goes back to studies in the nineteenth century by Dahl (1884), but it was not until after the second world war, when the problems with the bleachability of the pulp was technically solved, the real increase of the production volumes were initiated (Torén and Blanc 1997, Ragnar et al. 2013). Despite being a relatively old process, it is still developing in many ways, including better control over alkalinity and temperature (Lindgren and Lindström 1996, Johansson and Germgård 2008).

The reasons for the success of kraft pulping are superior economic and environmental footprint, due to an efficient chemical recovery system (Adams 1997), pulp with excellent mechanical properties mostly due to the cellulose being relatively undamaged (Suckling et al. 2001), and a broad tolerance for raw materials – both hardwood, softwood and herbal plants can be used (Ragnar et al. 2013); even the sea animal tunicates (sea squirts) have successfully been pulped with this method (Zhao and Li 2014). However, there are also drawbacks with the process; the investment costs for equipment are high (Lundberg et al. 2013); there is a substantial loss of non-cellulose polysaccharides during the pulping (MacLeod 2007); and the contents of minerals in the biomass may affect the process negatively (Sundin and Hartler 2000).

Among many varieties of trees used for kraft pulping, the genera of eucalypt are the globally most used for kraft pulping of hardwood today (Coppen 2002, Hart and Santos 2015). This group of plants originate from Australia and Indonesia (Macphail and Thornhill 2016), and are characterised by fast-growing and general resilience (Penín et al. 2020). The cellulose content is often very high, and so is the content of short slim librioform fibers in the wood, making it suitable for paper production (Hart and Santos 2015). These have made Eucalypt species like Eucalyptus globulus and Eucalyptus dunnii to subjects for plantations in for instance South America, southwest Europe and South Africa, and the wood is in addition to kraft pulping also used for other purposes, such as furniture and charcoal for the steel industry (Colodette et al. 2014).

However, strains of E. dunnii cultivated on plantations in various locations in Uruguay have shown irregular behavior in kraft pulping, and some samples showed very slow delignification rate. Investigations made at mills suggested a connection to high calcium content in the wood (Stora Enso, unpublished). Calcium is well-known for causing problems with scaling especially in the chemical recovery system, but different types of inorganic salts are also known to slow down delignification, and the effects have been suggested to decrease the solubility of lignin (Bogren et al. 2009, LéMon and Teder 1973, Mortimer and Fleming 1983, Saltberg et al. 2009, Sundin and Hartler 2000). There are however differences between different salts and calcium appears to be one of the more harmful cations in this aspect, giving stronger effects than magnesium ions (Lundqvist et al. 2006a). This suggests that the negative effect of calcium goes beyond the hydrophobic interaction created by the salts according to the Hofmeister series (Gurau et al. 2004, Hofmeister 1888) leading to low solubility of lignin degradation products and thereby slower delignification.

In this study, we characterise Eucalyptus dunnii samples with different calcium content chemically and morphologically, as well as for kraft pulping properties in order to better understand what kind of problems in the kraft pulping the calcium causes.

Materials and methods

Results

Wood samples of Eucalyptus dunnii cultivated at various plantations in Uruguay, with four different distinct contents of calcium, were chosen for this investigation (Table 1). The samples were subjected to a wood chemical analysis for lignin content and carbohydrate composition together with a reference sample of Eucalyptus globulus in order to investigate if any other significant differences than the calcium content occurred between the samples. The results (Table 2) indicate that the composition of all samples is relatively normal for Eucalypt samples, with a lignin content of 22–25 %, a cellulose content of 38–43 %, xylan content of 10–13 %, where the O-methyl glucuronic acid content is relatively low (around 5 %), and glucomannan content of 2–3 %, and pectin content of 2–3 %, where polygalacturonic acid dominates over rhamnogalactourans. There is no general tendency that the chemical wood composition varies with higher calcium content, but the sample with the highest calcium content 4668 mg/kg is an exception; here both xylan and pectin content appears to be increased by about 25 % (legend to Table 2). Both these polysaccharides contain uronic acids, and thus have the potential to bind to calcium ions. Can this difference in polysaccharide composition be an adaptation of the plant to the high calcium content? Further studies are needed for investigating this.

Traces of fucose and glucuronic acid was also detected. Methanolysis was used for the uronic acid measurements and acid hydrolysis for the other sugars. The molar content was estimated from the molecular weight of the anhydrous sugars. The only clear trend for the E. d samples with varying calcium content, is that the sample with the highest calcium content (E. d 4668 mg/kg), had a higher content of xylose, arabinose, galacturonic acid, and 4-O-methylglucuronic acid, and a lower amount of glucose. The polysaccharides of the eucalypt cell wall consist mainly of cellulose (100 % glucose), xylan (xylose, 4-O-methyl glucuronic acid, small amounts of arabinose and galactose), glucomannan (mannose and glucose, and pectin, that consists of two main structures polygalacturonic acid (mostly galacturonic acid), and rhamnogalacturonan (rhamnose, galacturonic acid and galactose). Thus, it seems plausible, that the difference between E. d 4668 and the other E. dunnii samples is that the earlier have more xylan and polygalacturonic acid and less cellulose. A sample of E. globulus was also analyzed for lignin- and carbohydrate content as a control of a well-known eucalypt wood, and showed similar values as the E. dunni samples and of literature values, underlining that the lignin and polysaccharides content was rather normal of eucalypt wood (Table 2).

No significant difference in the wood density was observed with different calcium content in the wood (Table 3). Scanning electron microscopy (SEM) investigation indicate presence in the lumen of crystals in the E. dunnii wood samples (Figure 1). Many studies have reported the presence of calcium oxalate in wood (Dutton et al. 1993, Dutton and Evans 2011, Margaret B. Nagy et al. 2012, Uloth et al. 2015). The number and shape of calcium oxalate crystals vary considerably (Franceschi and Horner 1980). Thus, it appears likely that the crystals in Figure 1 consist of calcium oxalate.

In summary, the structural and chemical characterisation did not indicate any other significant difference other than calcium content in the form of crystals for the samples investigated, with the exception that the E. d 4668 sample have more xylan and pectin.

Figure 1

Calcium oxalate crystals in the lumen of E. dunnii wood cells with high calcium content determined using SEM.

Lab scale kraft pulping was performed on the respective wood samples using a L:W of 4:1, an effective alkali of 18 % and a sulphidity of 35 % with an impregnation temperature of 110 °C and a cooking temperature of 145 °C. Samples with a calcium content ≤1500 mg/kg were found to defibrillate well for these conditions, with reject contents ≤1.5 %, while the sample with a calcium content of 3366 mg/kg (Table 1) required an elevated cooking temperature of 165 °C to achieve similar defibration. The sample with the highest calcium content (E. d 4668 mg/kg, Table 1) could not be defibrated with acceptable levels of reject, even at elevated cooking temperatures. This corroborated experiences from the industry, where wood with higher calcium content was found to be worse for kraft pulping, although the magnitude of this effect appeared to be even higher in lab scale pulping than it was in the industrial scale pulping. It is not clear why this was the case; one possibility could be that industrial white liquor, unlike white liquor in the lab, contains carbonates, and that these carbonates might inactivate calcium ions by precipitation (Li 2003, Lundqvist et al. 2006b, Saltberg et al. 2009).

Three wood qualities, with calcium contents of (705, 870 and 1500 mg/kg, were chosen for a more thorough characterisation of pulping properties, and pulped at different H factors, and characterised for kappa-number, yield and viscosity. The result in Figure 2 indicates clearly that the delignification goes strongly and significantly slower with higher calcium content. This observation is in line with what Saltberg et al. (2009) got on different hardwoods with various calcium contents. However, as seen in the figure, the lines are virtually parallel (Figure 2). Since there is no significant difference in original lignin content between these samples (Table 2), this indicates that the large differences in delignification rate between the samples are in the beginning of the kraft cook, i. e., the negative effects of calcium peaks early in the process and decline thereafter. The content of hexeneuronic acid in the pulps were also examined, but the results indicated no significant difference between the wood samples with different calcium content (Figure 3), indicating that the calcium does not affect the hexeneuronic acid formation. The reject content was also higher for the pulping experiments with wood of higher calcium content than for the one with lower calcium content (Figure 4), which is in line with a lower delignification rate (Figure 2).

Figure 2

Corrected kappa number plotted against H-factor. Corrected kappa number means that the influence from hexeneuronic acid have been removed, and that the kappa number therefore mainly correlates to lignin content.

Figure 3

Hexenuronic acid formation during pulping.

Figure 4

Reject content of pulps made from wood with different calcium contents.

On the other side, the effects on yield of higher calcium content are obvious (Figure 5), with higher yield losses when calcium content is higher. Also, the lines in the figure are virtually parallel, indicating that also the negative effects on yield are present mainly in the initial stage of the pulping. These two effects slow delignification rate (Figure 2) and higher yield losses (Figure 5) give together considerably effects on the selectivity of the kraft pulping by calcium content, as shown in Figure 6.

Figure 5

Yield of kraft pulping of wood with different calcium content as a function of H-factor.

Figure 6

Effects of selectivity of pulping. At a given kappa number, the differences in yield of the kraft pulping is considerable.

What are the reasons for the yield losses? The differences are probably too large for being caused purely by dissolution of calcium crystals (Figure 1, Table 2), and since the delignification is slower when calcium content is higher, the most likely explanation is that the yield losses are due to polysaccharide degradation. Therefore, carbohydrate analysis was performed on the pulps. The results shown in Figure 7 indicated that both cellulose and hemicellulose, i. e., xylan, were degraded (Most of the glucose and xylose in the wood are expected to be part of cellulose and xylan respectively); however, also in Figure 7, the lines appears to be parallel, which indicates that the critical effects were early in the pulping processes.

Figure 7

(a) Cellulose degradation during the pulping. (b) xylan degradation during the pulping.

How shall the carbohydrate losses in the high calcium woods during pulping be explained? There are mainly three types of reactions that are related to carbohydrate losses during kraft pulping:

1. Alkaline hydrolysis, that shortens polysaccharide chains in addition to degradation of polysaccharides to soluble sugars (Janson and Lindberg 1960).

2. Peeling reaction, that take away monosaccharide residues one-by-one from the reducing chain end (Hansson and Hartler 1970).

3. Dissolutions, where oligosaccharides and shorter polysaccharides are solubilised in the pulping liquor (Aurell and Hartler 1965).

Figure 8

Viscosity decrease of different pulps made of wood with different calcium content.

Discussion

The characterization of the wood samples did not indicate any other difference in the wood samples than calcium content, with the exception of the sample with the highest calcium content, that appeared to have increased pectin and xylan levels. Since the only significant difference between the samples that we could detect is the content of calcium, it is our hypothesis that the pulping problems are connected to this element, although we of course could not exclude that there are any other undetected differences. Evaluation of pulping experiments presented in this paper suggests that the difficulties to pulp the high-calcium E. dunnii are dual as follows:

1. Slower delignification than for wood with a lower calcium content.

2. Higher yield and viscosity losses for wood with a higher calcium content, most likely explained by increasing alkaline hydrolysis.

Figure 9

Hypothetical explanations for the negative effect of calcium on lignin degradation.

For the slower delignification caused by calcium, we could see at least four principal ways this can happen:

1. The delignification reactions, i. e., the lignin degrading reactions, caused by the strong nucleophile HS − , are going slower. The Ca(II) ion might form strong complexes with hydrogen sulphide ions and may this lower the activity of this strong nucleophile (Figure 9a)?

2. The accessibility of the nucleophile to lignin may be lowered; this can be achieved by the soluble lignin fragments are precepted by calcium ions, and a “layer” that protects the lignin in the fiber from nucleophilic degradation (Figure 9b).

3. The calcium ions form complexes with carboxylic acids in the cell wall and make it more compact and thereby less accessible for delignifying chemicals (Figure 9c).

4. Reactions that counteract the delignification reactions are going faster. This might be a possibility, since the lignin fragments during kraft pulping conditions undergo formation of novel and stable chemical bonds If calcium ions aggregate lignin degradation products this might enhance reactions forming novel bonds that lower the delignification rate (Figure 9d).

How shall the polysaccharide degrading ability of calcium be explained? An explanation could be that calcium ions catalyze degradation of polysaccharides, and indeed have catalytical activity of calcium salts has been reported (Correia et al. 2014). In Figure 10, a hypothetical mechanism of catalysis of alkaline hydrolysis by calcium (II) ions is shown. It shall be stressed that we cannot exclude that there might be other mechanism and explanation to these results.

Figure 10

Proposed role for calcium ions in alkaline hydrolysis of polysaccharides.

Figure 11

Suggested hypothesis for why the negative effects of calcium ions is concentrated to the early stage of pulping.

What is the explanation for why the effects appear to be concentrated in the initial stages of pulping? One possibility could be that the calcium oxalate crystals in the lumen (Figure 1) quickly are dissolved when they come in contact with the high pH in the pulping liquor, and that the calcium ions then migrate into the cell wall, where they disturb pulping twofold as described above. Later in the pulping the calcium leaves the cell wall, or get “inactive” by the formation of stable crystals (Figure 11). Calcium oxalate solubility increases indeed with temperature, and it appears also as high pH can help in dissolving the crystals (Tomazic and Nancollas 1979). It is also known that calcium ions have certain affinity for cellulose, lignin and hemicellulose (Torre et al. 1992). Therefore, it appears plausible that calcium ions might migrate from the lumen into the cell wall early during pulping. A layer rich in calcium will therefore be produced. The calcium might then disturb delignification, and also catalyse alkaline hydrolysis as discussed above (Figure 9, 10). However, when organic black liquor products are formed, the calcium might form stable complex with components here, thereby being chemically inactivated, but the damage has already happened (Figure 11). It shall be underlined that this is a suggested explanation, and we cannot rule out other mechanisms involved. Saltberg et al. (2009), observed a similar effect, and this effect was stronger when eucalypt was used than with other hardwoods, and eucalypt black liquors are known to contain higher amounts of calcium than other hardwoods (Lidén et al. 1996), possibly due to that certain degradation products from heartwood tannins – gallic acid and ellagic acid, form soluble calcium complexes. One explanation might be that also lignin degradation products and soluble xylan chelates calcium.

Conclusions

1. Samples of Eucalyptus dunnii with different calcium content (705–4668 mg/kg) have similar chemical composition otherwise, with the exception of the highest calcium content (4668 mg/kg), where xylan and pectin content increase.

2. The density of wood ( 0.55 ± 0.3 g / cm 3 ) was unaffected by calcium content.

3. Calcium was present in the form of crystals in the lumen.

4. Samples of E. dunnii, were delignified slower in kraft pulping the higher calcium content they had, and reject content was higher (Figure 2 and 4).

5. Higher calcium content in wood gives higher yield losses in kraft pulping.

6. Both cellulose and hemicellulose degradation is enhanced with higher calcium content.

7. Higher calcium content gives a higher decrease in viscosity during kraft pulping (Figure 8).

8. All these effects are concentrated at the beginning of the pulping.

9. Hexenuronic acid formation seems not to be significantly affected (Figure 3).

10. The calcium might be present in the form of oxalate crystals in the lumen of fibers.

11. The effects in lab scale pulping appear to be stronger than in industrial scale pulping.