NC State
Joubert, A. J., Chimphango, A. F. A., and Görgens, J. F. (2016). "Effect of integrating xylan extraction from E. grandis into the kraft pulping process on pulp yield and chemical balance," BioRes. 11(1), 2417-2437.


Kraft mills have the potential to pre-extract hemicellulose from wood as another value added product. The impacts of pre-extracting xylan on pulp and handsheet properties, sodium and sulfur balances, and chemical make-ups in the kraft pulping process of Eucalyptus grandis were assessed. Xylan extractions using white liquor, green liquor, and NaOH were done under varying extraction times, temperatures, and alkaline concentrations; residues were subsequently pulped at 170 °C for 45 min. The highest xylan yield (15.15% w/w) was obtained with 2 M NaOH, at 120 °C for 90 min followed by white liquor (13.27% w/w), utilizing 20% AA at 140 °C for 90 min. Green liquor extraction with 2% total titratable alkali (TTA), at 160 °C and an H-factor of 800 produced the lowest yield (7.83% w/w). However, the green liquor extractions were the most practical for integration into the kraft process due to their limited effect on pulp yield and properties of handsheets produced from the pre-extracted woodchips and the sulfur and sodium make-up increase. White liquor extractions would favour high pulp yield with low kappa number and reduced chemical charge and cooking time. These results are important for technical-economic assessment of integrated kraft pulp biorefineries.

Download PDF

Full Article

Effect of Integrating Xylan Extraction from Egrandis into the Kraft Pulping Process on Pulp Yield and Chemical Balance

Andre J. Joubert, Annie F.A. Chimphango,* and Johann F. Görgens

Kraft mills have the potential to pre-extract hemicellulose from wood as another value added product. The impacts of pre-extracting xylan on pulp and handsheet properties, sodium and sulfur balances, and chemical make-ups in the kraft pulping process of Eucalyptus grandis were assessed. Xylan extractions using white liquor, green liquor, and NaOH were done under varying extraction times, temperatures, and alkaline concentrations; residues were subsequently pulped at 170 °C for 45 min. The highest xylan yield (15.15% w/w) was obtained with 2 M NaOH, at 120 °C for 90 min followed by white liquor (13.27% w/w), utilizing 20% AA at 140 °C for 90 min. Green liquor extraction with 2% total titratable alkali (TTA), at 160 °C and an H-factor of 800 produced the lowest yield (7.83% w/w). However, the green liquor extractions were the most practical for integration into the kraft process due to their limited effect on pulp yield and properties of handsheets produced from the pre-extracted woodchips and the sulfur and sodium make-up increase. White liquor extractions would favour high pulp yield with low kappa number and reduced chemical charge and cooking time. These results are important for technical-economic assessment of integrated kraft pulp biorefineries.

Keywords: Hemicellulose; Xylan pre-extraction; Kraft cooking; Green liquor extraction; White liquor extraction; NaOH extraction; Sodium balance; Sulfur balance

Contact information: Department of Process Engineering, Stellenbosch University, Private Bag X1, Stellenbosch 7600, South Africa; *Corresponding author:


Pulp mills are facing increased environmental and economic pressures. Consequently, it is important that mills maximize the economic value derived from woody biomass that they process. Establishing integrated biorefineries (IFBRs) where additional value added products can be produced instead of pulp alone, entails important economic benefits (Ragauskas et al. 2006). Kraft pulp mills are especially well suited for conversion into IFBRs as they already have the required infrastructure, support networks, and specialized manpower (van Heiningen 2006; Mateos-Espejel et al. 2011).

A promising concept towards establishing IFBRs is the extraction of hemicelluloses prior to the pulping process (van Heiningen 2006; Tunc et al. 2010). Hemicelluloses have a wide variety of potential industrial uses, including not only the pulp and paper industry itself, but also the food and pharmaceutical industries. In the standard kraft pulping process, half of the hemicellulose, equivalent to about 15% of the wood weight, and almost all of the lignin present in woodchips raw material, is dissolved into the black liquor (Christopher 2013).

The fairly high heating value of lignin, 26.9 MJ/kg, makes it cost-effective to recycle it back to the reboiler for combustion. The heating value of hemicellulose, on the other hand, is about 13.6 MJ/kg, thus causing it to be underutilized when it is incinerated in the reboiler, as it provides only about a quarter of the energy recovery (Christopher 2013). It is more valuable to extract hemicellulose prior to the pulping process to generate an additional stream of high value products (Um and van Walsum 2010). Any hemicellulose pre-extraction method needs to be compatible with the existing kraft mill processes, while ensuring that the pulp and resulting paper qualities, obtained after hemicellulose pre-extraction, are still within the required industry standard. Importantly, the quantity of xylan extracted also needs to be sufficient to ensure that the pre-extraction process is viable, but without compromising subsequent pulp production.

A number of hemicellulose pre-extraction methods have been investigated by researchers, ranging from hot water extraction, to more “mild” extraction methods using alkaline chemicals. The recovery of the hemicellulose, in particular xylan, can vary considerably during alkali extraction (Al-Dajani and Tschirner 2008; Jun et al. 2012). Relatively higher xylan yields of approximately 25% of the original xylan fraction in the aspen hardwood chips were reported by Al-Dajani and Tschirner (2008) and Jun et al. (2012) when using white liquor and NaOH, respectively.

Although water and acid methods for hemicellulose extraction may achieve high hemicellulose yields, the severity of such extraction methods reduces the final pulp yield and paper quality obtained by pulping of solid residues from hemicellulose extraction (Mendes et al. 2010; Vena et al. 2013). These are key concerns of pulp and paper manufacturers (Jun et al. 2012).

Alkaline extraction methods alleviate some of these concerns, as they have minimal effect on pulp yield reduction when compared to other extraction methods, such as acid and water extraction methods (Vena et al. 2013). Moreover, acid-based methods for hemicellulose pre-extraction will neutralize pulping chemicals in subsequent process steps, resulting in inefficient use of these chemicals. The water and acidic extraction methods are also more prone to degradation of extracted xylan biopolymers to oligosaccharides and monomers when compared to that of alkaline extraction methods, thus impeding the potential to produce high value biopolymers (Fiserova and Opalena 2012).

The qualities of paper produced from solid residues of woods from xylan pre-extraction using alkaline methods such as NaOH, green liquor as well as white liquor, have been comparable with those obtained by pulping of non-extracted woodchips. In some instances, the qualities have improved properties (Helmerius et al. 2010; Walton et al. 2010; Jun et al. 2012; Vena et al. 2013). In striving for commercial feasibility, increased attention is being directed towards using pre-existing alkaline chemicals within kraft mills for hemicellulose pre-extraction (Johakim and Andrew 2013).

Eucalyptus grandis species is one of the most important sources of fiber in the pulp and paper industry (Magaton et al. 2009) and is widely used by South African mills due to its high pulping yield and low production costs. The aim of this study was to determine the impact of the xylan pre-extraction by green liquor, white liquor, and NaOH on pulp yields and paper properties from kraft process of E. grandis at a South African pulp mill. Furthermore, the study assessed the impact on the sodium and sulfur balances of the kraft mills, and the requirement for make-up chemicals to the kraft pulping cycle.


Green liquor, white liquor, and NaOH were used to extract xylan from E. grandis woodchips prior to kraft pulping. Pre-extracted woodchips from selected extraction conditions were subjected to varying pulping conditions, to compare pulp yields and properties with those obtained from untreated E. grandis chips with conventional pulping. Handsheets were also made from the pulps produced under the highest pulp yield conditions, and these were tested for pulp quality properties. Furthermore, mass balances were performed to gauge the impact that hemicellulose pre-extraction would have using green liquor, white liquor, and NaOH on the sodium and sulfur balances of the mill. The most feasible method for hemicellulose (xylan) pre-extraction in an existing kraft process could thus be identified.

Feedstock Utilized

Experiments were performed using wood chips obtained from an 11-year-old E. grandis plantation grown in Tzaneen in South Africa. The wood was chipped and the resulting wood chips were screened for size fractions ranging between 4 to 8 mm. The screening was done by a six-component chip screen driven by an electric motor. Before use, the wood chips were conditioned in a conditioning room allowing for constant temperature of 23 °C coupled with 55% relative humidity.

Compositional Analysis of E. grandis and Solid Residue from Alkaline Pre-Extraction

The E. grandis chips were sub-sampled as prescribed by British Standards DD CEN/TS 14780 (2005).E. grandis chips were milled to a particle size below 625 μm with the use of a Condux-Werk mill (Netzsch, Germany) as well as in a Retsch ZM 200 mill (Retsch GmbH, Germany). The particles produced were then sieved using a Retsch AS200 vibratory shaker (Retsch GmbH, Germany), with the 425 μm and 625 μm size fractions being utilized for further analysis.

Moisture content was determined using the NREL-TP 510-42621 (Sluiter et al. 2008a) method, with the extractives content being determined by a two-step extraction process, where both water and ethanol/cyclohexane extractives were quantified as described by the NREL-TP 510-42619 method. Klason lignin as well as carbohydrate composition of both the feedstock and the residue from alkaline pre-extraction were determined using NREL-TP 510-42618 (Sluiter et al. 2007), whereas the ash contents were determined via NREL-TP 510-42622 (Sluiter et al. 2008b).

Following xylan pre-extraction, the remaining solid residue was air-dried and conditioned at constant temperature of 23 °C coupled with 55% relative humidity. Moisture content of the residue was determined subsequently using infrared drying. The wood chips were subjected to two consecutive size reduction processes of the Condux-Werk mill and Retsch ZM 200 mill for characterization and determination of ash content. The subsequent wood weight loss of the pre-extracted chips was determined using Eq. 1 on ash free and dry weight bases,

where WIO = the oven dry weight of the initial sample subjected to pre-extraction; WEO = the oven dry weight of the remaining residue after pre-extraction; and AshE = the ash fraction of the remaining residue after pre-extraction.

Composition Analysis of Liquor Fraction Obtained after Alkaline Pre-Extraction

The liquor fractions from NaOH and white liquor extractions (pH > 13), were neutralized using 37% HCl to pH 7. Subsequently, the samples were frozen at -4 °C until they were prepared for HPLC analyses. Liquid fractions obtained from green liquor extractions, with pH between pH 5 and pH 7, depending on the initial alkali charge, were frozen at -4 °C without neutralization, before high performance liquid performance (HPLC) analysis. The sugars (glucose, xylose, galactose, and arabinose) and acetate composition were analyzed using HPLC according to the method described by the NREL-TP 510-42623 (Sluiter et al. 2008c). The HPLC system used consisted of a Thermo Separations Spectra P 2000, which was equipped with both an Aminex HPX-87H column, operating at 65 °C, and an IG cation H guard column, with a Shodex RI-101 refractive index detector. The mobile phase was 5 mM H2SO4 at a flow rate of 0.6 mL min-1, with the IG cation H guard column.Furthermore, the liquor samples were subjected to FT-IR analysis using a Thermo Nicolet Nexus 870 Fourier transform infrared (FT-IR) system via the attenuated total reflection (ATR) Golden Gate measurement device. The data analysis was done with the use of Omnic 7 software which was subsequently exported to Microsoft Excel.

Alkaline Pre-Extraction of Xylan from E. grandis

White liquor xylan extractions were performed at both mild as well as more severe extraction conditions. The mild extraction experiments were performed via a face centered central composite design with temperatures from 50 °C to 90 °C coupled with active alkali (AA) charges ranging from 16% to 20%. Extractions were performed by mixing 50 g (over dry weight, O.D.W.) E. grandis chips with white liquor (4:1 liquor to wood ratio) in 500 mL Schott bottles and subsequently placing them in a shaking hot water bath. Xylan content in the liquor fraction was considered a response of the central composite design with data interpreted using Stat Soft® Statistica 9.0. Table 1 shows both the extraction conditions as well as the corresponding results.

E. grandis chips were also subjected to green liquor, white liquor, and NaOH at liquor to wood ratio of 4:1 at extractions exceeding 100 °C, through the use of a pressurized oscillating digester. Extractions were carried out using batches of 50 g (O.D.W.) of E. grandis. During the extraction cycles the digester oscillated through 45° in order to ensure sufficient liquid contact with the wood chips. The temperature and pressure of the digester were monitored with a thermocouple and pressure gauge fitted to the lid of the digester. A programmable logic computer (PLC) was used to control the entire extraction cycle. Pressure in the digester was liberated through the use of a blow valve, with operating pressure during pulping being around 8 bar and that of the extractions varying between 2 and 7 bar.

The white liquor extractions were carried out at a temperature range from 100 °C to 140 °C using AA charges of 16% and 20% for a period of 90 min at each given temperature interval. The NaOH extractions were performed at temperatures of 100 °C, 110 °C, 120 °C, and 130 °C and at molar concentrations of 1 M and 2 M for 90 min.

The green liquor extractions were carried out at 160 °C with H-factors between 200 and 800. The total titratable alkali (TTA) charges that were investigated included 2% and 6% for extractions performed using green liquor. For the green liquor extractions that were performed, a charge of 0.05% of anthraquinone (based on oven dry weight of wood chips) was added (Walton et al. 2010). The extraction conditions that were used are summarized in Table 1.

Micro-Pulping of Pre-Extracted Residues

As a benchmark, non xylanextracted E. grandis was subjected to pulping by simulating batch cooking conditions. Pulping conditions included a liquor to wood ratio of 4.5:1, 170 °C pulping temperature, 18% AA coupled with 25% sulfidity. Cooking conditions were based on suggested pulping conditions of South African E. grandis (Myburgh 1967). Pulping time at 170 °C was set to 45 min to achieve a kappa number of 20.

After pre-extraction, the wood chips residues were removed from the micro-bombs and the resulting liquor was drained from the chips through a 100 mesh screen. The solid residue (unwashed) from the extractions was placed in micro-bombs for further pulping.

Table 1. Process Conditions for Extraction of Xylan at Temperatures Exceeding 100 °C using Green Liquor, White Liquor, and NaOH

Testing of Physical Properties of Handsheets

Pulp fibers were separated from the pulping black liquor by washing it through a 10 mesh screen with running water until all the pulp had passed through with the uncooked fibers (rejects) retained on the screen. Thereafter, the pulp was screened with a packer slotted laboratory screen. Excess water from the pulp was removed through spin drying, to a moisture content of approximately 70%. The pulp yield was determined as the percentage of the original oven dry mass of the E. grandis chips using Eq. 2 (Vena et al. 2013).


The kappa number of the pulps was determined using the method that is outlined by ISO 302:2004. The unbeaten pulps, which were comparable in pulp yields to that of non-extracted chips, were subsequently formed into handsheets according to TAPPI T 205 om-88 method.

The physical properties of handsheets formed from the pulps that were produced were also analyzed. These properties included the burst, tear, as well as the tensile indices of the handsheets produced. Methods used for the determination of the physical properties of the handsheets are presented in Table 2.

Table 2. Analytical Standard Methods Used to Determine Physical Properties of Handsheets Formed.

Sodium and Sulfur Mass Balance for the Kraft Process

To investigate the effect the pre-extraction conditions would have on the make-up of chemicals required in the kraft mill, sodium and sulfur mass balances were calculated for the modified kraft processes. The sodium and sulfur mass balances were done using extraction conditions selected for pulping of wood chips to form handsheets. The estimated kraft chemical recovery cycle in the conventional process was assumed to be 97% (Tran and Vakkilainannen 2012). The amount of sodium and sulfur (kg) used for each specific hemicellulose extraction condition, per ton of air dried pulp produced, was calculated with Eqs. 3 to 6.


NaNa2S, NaNaOH, and NaNa2CO3 represent the mass (kg) of sodium, if one ton of air dried pulp is to be produced, from Na2S, NaOH, and Na2COrespectively. SNa2S represents the mass (kg) of sulfur used during hemicellulose extraction, if a ton of air dried pulp is to be produced, from Na2S. In order to calculate the total amount of sodium used during white liquor extraction, the sodium used from Na2S and NaOH was added together, while, as green liquor consists of all three Na2S, NaOH, and Na2CO3, the Na from all these components were added together. With NaOH extraction, only the Na from the NaOH was used for the calculation. The only component containing sulfur was Na2S, which is found only in white and green liquor. The values 23, 32, 78.1, 40, and 106 are the molecular weights (g/mole) of sodium, sulfur, Na2S, NaOH, and Na2CO3, respectively. The value 0.085 is the oven dry mass (kg) of the wood chips used during extraction, while Y represents the pulp yield as a fraction. PMis the moisture fraction for air dried pulp which was assumed to be 10%.


Compositional Analysis of the E. grandis Feedstock

The chemical composition of the E. grandis feedstock for combined hemicellulose extraction and pulping is shown in Table 3. The cellulose fraction is expressed in terms of the glucose content, while the hemicellulose content is expressed as monomeric xylose and arabinose contents.

Table 3. Chemical Composition of E. grandis Utilized

The contents of xylan and cellulose, which were polymers of particular interest in this study, were typical for E. grandis. Previous reports show xylan contents between 11.5% and 21%, and cellulose contents between 43% and 53% (Magaton et al. 2009; Vena et al. 2013).

The lignin content for the species was typical of E. grandis with reported values between 21% and 30% (Cotteril and Macrae et al. 1997; Vena et al. 2013). Low lignin content is especially desirable for pulping purposes due to easier delignification. Moreover, lignin can also hinder the release of sugars during hemicellulose pre-extraction (Studer et al. 2011). The extractives content was higher than those reported in previous reports of 2.9% to 4.2% for E. grandis (Emmel et al. 2003; Magaton et al. 2009).

Yield of Xylan from Alkaline Pre-Extraction of E. grandis with Green Liquor, White Liquor, and NaOH

The xylan yield was influenced more by level of alkalinity than temperature. Higher xylan yields were obtained with alkaline extraction from E. grandis at temperatures above 100 °C. The highest yield of 15.5% was obtained with NaOH (2M) extraction at 120 °C for 90 min, followed by white liquor, 13.27% obtained at 20% AA, 140 °C for the same duration (Fig. 1). The green liquor extraction, which had lowest alkalinity (Table 1), yielded the lowest xylan (7.83%), despite using higher extraction temperatures of 160 °C than the other two methods. This indicated that the higher alkali charges used during the white liquor and NaOH extractions had a more significant impact on xylan yield than the higher extraction temperature utilized by the green liquor extraction.

The specific E. grandis used in this study showed recalcitrance in releasing xylan at lower temperatures. At temperatures ranging from 50 to 90 °C, the maximum recovery of xylose was 5.63% (5.41% predicted by the model) from white liquor extractions (Table 4), which was lower than that ofnearly 25% from aspen under similar conditions (Jun et al. 2012).

Table 4. Central Composite Design Output of White Liquor Xylan Extraction from E. Grandis at Temperatures Lower than 100 °C.

* Significant linear effect (p = 0.007, F = 13.96; MS = 2.54; R2 = 0.78)

Under mild alkali conditions of 1 M NaOH at 90 °C, Vena et al. (2013) extracted 12.4% of the xylan present in E. grandis feedstock that had lower lignin content (21.1%). The low yield may be attributed to the high extractives content (4.3%) and lignin in the E. grandis (Table 3).

Some of the extractives are hydrophobic and can aggregate on the surface of the fiber, blocking both the penetration of chemicals and the diffusion of lignin from the fiber wall (Dai et al. 2004), thus affecting the xylan yield. Feedstock with low lignin is particularly amenable to xylan release, during alkali extraction (Studer et al. 2011). The E. grandis in the present study had a higher lignin (Table 3) than one Vena et al. (2013) used. The decrease with increasing lignin content is possibly due to the presence of lignin-carbohydrate complexes (Jeffries 1990).

White liquor pre-extraction performed at temperatures below 100 °C had a significant effect (p<0.5) on xylose recovery. However, alkaline charge in the range of 16% and 20% for the same temperature range (< 100 °C), had no significant effect (p>0.5) on the xylose recovery. Therefore, the effect alkali charge on xylan extraction is temperature dependent (Liu et al. 2011). The R-squared value of 0.78 of the model itself implies reasonable predictability of xylose recovery.

Fig. 1. Mass of solubilized sugars recovered in extraction liquor per 100 grams of E. grandisfeedstock. Temperature on the bottom x-axis is for white liquor and NaOH extractions, while H-factor on the top x-axis is for the green liquor extractions that were performed.

The lower quantities of xylan recovered from the green liquor liquid phase, compared to white liquor and NaOH, reflect the lower xylan solubilisation and release from the residue (Fig. 1), which corresponded to the maximum wood weight losses of 7.67% for green liquor extractions compared to 20.67% and 16.44% for white liquor and NaOH, respectively (Fig. 2).

The xylan was extracted by green liquor with relatively higher purity than when utilizing white liquor and NaOH, as reflected by the presence of lower content of lignin derivatives than those extracted by the latter two methods (Fig. 3). The FTIR spectra for xylan from white liquor and NaOH extractions displayed peaks at 1193/1192 cm-1, associated with Caryl-O (Collier et al. 1992) as well as at 1052 cm-1, which depicts the deformation of C-O coupled with the deformation of aromatic C-H associated with lignin (Collier et al. 1992; Akerholm et al. 2005).

In contrast, peaks at the aforementioned bands were absent in the spectra for green liquor extracted xylan (Fig. 3). Notably, the green liquor extraction conditions were less severe (lower alkalinity) than those of white liquor and NaOH, which may have limited the co-release of lignin and xylan. Depending on the application, a purer xylan may offset cost of purification and lower yield, thus making green liquor pre-extraction more practical for integration in the kraft process.

Fig. 2. Wood weight losses obtained using three different extraction methods. (a) White liquor extraction for 16% and 20% Active Alkali. (b) Green liquor extraction using 2% and 6% TTA. (c) NaOH extraction for 1M and 2M.


Fig. 3. FT-IR spectra of xylan extract from different extraction methods: (Top) White Liquor (Middle) NaOH, and (Bottom) Green liquor

Yield of Pulp and Properties of Handsheets Made from the Solid Residues from Green Liquor, White Liquor, and NaOH of Xylan Extraction

The final selection of the most feasible extraction method, for integrating with kraft pulp production would have to be based on the subsequent properties such as tensile, tear, burst indices of the pulp, and paper products as well as the make-up chemicals required. Green liquor had a pulp yield of 43.19%, which corresponded to a kappa number of 18.70 (Table 5). The pulp yield of the non-extracted chips was 41.67% with a corresponding kappa number of 20.3, indicating that green liquor pre-extraction benefited the subsequent pulping processes in yield and kappa number. The average pulp yield of 41.67% from the non-extracted chips (Table 5) was on the lower side when compared to industrial process yields of above 50% (Vena et al. 2013). Significant variations in wood properties between E. grandis trees, even with trees that grow within the same location, are known to occur (Malan 1988), making the deviation of pulp yield of this particular feedstock not particularly exceptional. The higher extractives content than normal (Table 3) for this E. grandis species might have lowered the pulp yield by preventing chemical penetration during pulping. Therefore, it is desirable that the extractives content should be as low as possible for efficient pulping (Macleod 2007).

The unbeaten handsheets from pulps obtained from green liquor extracted chips, had tensile, tear, and burst indices of 18.2 N.m-1.g-1, 3.2 mN.m2.g-1, and 0.924 kPa.m2.g-1, respectively. These values were almost identical to that of the reference pulp (Fig. 5). The inherent high alkalinity of the white liquor extractions allowed for a greater reduction in chemical charge during the subsequent pulping of the pre-extracted residue when compared to that of green liquor extraction. The resulting pulp yield of 41.60% was comparable with that of the reference pulp of 41.67% (Table 5) but with a lower kappa number of 12.86 compared to kappa number of 20.3 of pulp produced from non-extracted E. grandis(Table 5). Maintaining the pulp yield with a lower kappa number indicates that the white liquor extractions reduced both the cooking time and chemical charge and had selective dissolution of lignin, thus increasing pulping efficiency. There was no statistical significance observed in terms of burst, tear, and tensile indices of handsheets produced from white liquor pre-extracted residues when compared to those of the non-extracted E. grandis (Fig. 4). However, beating could potentially result in more significant deviations being exhibited in the properties of handsheets produced from pulps of non-extracted E. grandis chips.

The E. grandis chips that were subjected to NaOH pre-extraction had a pulp yield of 40% (Table 5) with a corresponding kappa number of 18.85. The resulting yield was lower than the yields of 43.19% and 41.67% (Fig. 4) for pulps resulting from green liquor and white liquor pre-extraction, respectively. Furthermore, disparities were observed in terms of both the tear and burst indices of pulps resulting from NaOH pre-extracted chips when compared to that of the non-extracted chips (Fig. 4). An increase of 28% was observed for the tear index of pulps derived from NaOH extracted chips; however, a reduced burst index of 9.6% was also observed (Fig. 4). This indicates that the quantity of xylan removed from the NaOH pre-extracted chips was significant enough to start impacting the pulp fibers and subsequently the paper properties. The increase in tear strength becomes apparent when there is significant reduction in flexible inter-fibre bonding that is enhanced by presence of hemicelluloses (Wan et al. 2010). The effects of decreased flexibility of inter-fibre bonding on the tear strength properties of the paper have been discussed before (Helmerius et al. 2010) and are attributable to the contribution of hemicelluloses to the electrostatic character of pulp fibres. As hemicellulose content of pulp is reduced there is reduction in accessible fibril surfaces for water molecules, which results in dried pulps being less conducive to swelling when soaked in water as a result of fiber separation. The effect becomes more challenging for fibre bonding because it provides less contact area with water (Wan et al. 2010; Reyes et al. 2015).

Effect of Pre-Extraction of E. grandis Utilizing Green Liquor, White Liquor, and NaOH on Kraft Chemicals Balance

Pre-extraction of E. grandis with the green liquor reduced the chemicals requirement for subsequent pulping of the solid residues, without compromising the pulp yield or quality. A reduced chemical charge of 35% was obtained with improvements in both the yield and kappa numbers of pulps, which indicates that extraction of hemicellulose preserved the cellulose fiber quality, which in turn is an indicator of how the integration process economics would be affected (Walton et al. 2010). The reduction in chemicals obtained is much higher than the 20% reduction in pulping chemicals for a kappa number of 20 (Johakim and Andrew 2013). A pulp yield of 41.67% obtained from pulping of pre-extracted residues from xylan extraction by white liquor at 120 °C and 20% AA for 90 min, resulted in a 50% reduction in chemical charge and 15 min reduction in cooking time, relative to the standard cooking conditions of non-extracted chips (Table 5).

Table 5. Pulp Yields and Corresponding Kappa Numbers of Pre-Extracted Chips

a No additional chemicals were added to chips after extraction.

b Only half the amount of Na2S required for 85 g (O.D.W.) of non-extracted chips was added. No NaOH added.

Same amount of Na2S as is required for 85 g (O.D.W.) of non-extracted chips was added. No NaOH added.

d Same amount of Na2S and 25% of NaOH as is required for 85 g (O.D.W.) of non-extracted chips added.

e Extraction and pulping combinations selected for handsheet formation.

*Represents percentage reduction of Na2S and NaOH relative to Na2S and NaOH required for 85 g (O.D.W.) of non-extracted chips.

**Wood chips after pulping were not in pulp form as they weren’t sufficiently pulped.

Fig. 4. Properties of handsheets made from xylan pre-extracted residues (a) Tear, (b) Tensile, and (c) Burst indices

Such an effect was reported earlier by Johakim and Andrew (2013). The E. grandis chips that were subjected to NaOH pre-extraction produced the best pulp properties with a 25% reduction in the quantity of NaOH requirement relative to that of the standard pulping conditions when the residues were pulped with the same amount of Na2S as that of the non-extracted chips for 45 min (Table 5).

The practicality of hemicellulose extraction integrated in the kraft pulp process lies both in the extraction technology and the chemical recovery technologies. About half of all the technical equipment in the kraft process is physically used for pulping, with the remaining used for chemical recovery from the spent liquor. About 97% of the alkaline chemicals are recovered (Blechschmidt et al. 2013) to reduce operating costs. In hemicellulose extraction streams the recovery was 90%. The chemical balance assessments to evaluate the losses of sodium and sulfur during hemicellulose pre-extraction against the amount of Na2SO4 and NaOH needed to compensate for the lost sodium and sulfur per ton of pulp produced are depicted in vector diagrams for white liquor (Fig. 5b), and for NaOH and green liquor extractions (Figs 6,a and b, respectively) . When compared with a kraft process without hemicellulose extraction (Fig. 5 a), the sodium and sulfur exiting with the hemicellulose stream in white liquor extraction represented 4.46 and 5.08 times increase in make-up of sodium and sulfur (Fig.5 b). In NaOH extraction the increase in chemical make-up, thus sodium make up was found to be increased 4.96 times (Fig. 6 a). On the other hand, the loss of sodium and sulfur from the green liquor extraction for the same chemical recovery led to 1.73 and 1.29 times increases in the make-up for sodium and sulfur, respectively (Fig. 6 b), which is attributed to the low alkali concentration used during the initial extraction process. The results imply that sodium and sulfur balances in NaOH and white liquor xylan extractions can be closer to those of standard pulping if there is close to 100% recovery of chemicals used in the-extraction-of-the-hemicellulose. This information is important for techno-economic assessment of establishing integrated forest biorefineries.

Fig. 5 Sodium and Sulfur balance in kraft mill: (a) Without hemicellulose extraction (b) during hemicelluloses extraction with white liquor

Fig. 6 Sodium and sulfur balance in kraft mill during hemicelluloses extraction with (a) NaOH and (b) green liquor


  1. Green liquor was found to be the most feasible pre-extraction method to be integrated into the kraft pulp cycle compared to the white liquor and NaOH extraction methods.
  2. The green liquor extraction demonstrated a better sodium and sulfur balance as reflected in the reduced increase in the make-up chemicals of the kraft process.
  3. Green liquor was selective for xylan release and preserved the pulp fibers, thus giving xylan with greater purity and increased pulp yield relative to the pulps obtained from xylan non-extracted wood chips.
  4. Green liquor xylan pre-extraction maintained the burst, tear, and tensile properties of handsheets derived from unbeaten pulp.
  5. About 100% recovery of chemicals from the hemicellulose extract is necessary in white liquor and NaOH xylan pre-extraction to minimize the impact of hemicellulose pre-extraction on the kraft cycle chemical balance.
  6. The white liquor extractions allowed for a greater reduction in chemical charge during the subsequent pulping of the pre-extracted residue and can be used to obtain high pulp yield and low kappa number with reduced cooking time.
  7. Chemical balance and effect of xylan pre-extraction were demonstrated to be important parameters for techno-economic assessment of integrated forest biorefineries based on kraft mills.


The authors are grateful for the financial support of The Paper Manufacturers Association of South Africa (PAMSA), The Technology and Human Resources for Industry Programme (THRIP) of the National Research Foundation (NRF), and the support by the technical staff at the Departments of Wood Science and Forestry and Process Engineering, Stellenbosch University. The study would not have been possible without the technical and material support from The Mondi Group, Richards Bay.


Akerholm, M., and Salmen, L. (2005). “The oriented structure of lignin and its viscoelastic properties studied by static and dynamic FT-IR spectroscopy,” Holzforschung 57(5), 459-465. DOI: 10.1515/HF.2003.069

Al-Dajani, W. W., and Tschirner, U. W. (2008). “Pre-extraction of hemicelluloses and subsequent kraft pulping, Part I: Alkaline extraction,” Tappi Journal 7(6), 3-8.

Blechschmidt, J., Heinemann, S., Putz, H.-J., and Duffy, G. G. (2013). “Fibrous materials for paper and board manufacture,” in: Handbook of Paper and Board, Holik, H. (ed.), Wiley-VCH, Weinheim, Germany.

Christopher, L. P. (2013). “Integrated forest biorefineries: Current state and development potential,” in: Integrated Forest Biorefineries: Challenges and Opportunities, Christopher, L. P. (ed.), The Royal Society of Chemistry, Cambridge.

Collier, W. E., Schultz, T. P., and Kalasinsky V. F. (1992). “Infrared study of lignin: Reexamination of aryl-alkyl ether C-O stretching peak assignment,” Holzforschung 46(6), 523-528. DOI: 10.1515/hfsg.1992.46.6.523

Dai, Q., Jameel, H., Chang, H., and Kadla, J. F. (2004), “Bleachability of kraft pulps from earlywood and latewood of fast-growing loblolly pine,” Journal of Wood Chemistry and Technology 24 (4), 357-370. DOI: 10.1081/WCT-200046256

Emmel, A., Mathias, A. L., Wypych, F., and Ramos, L. P. (2003). “Fractionation of Eucalyptus grandischips by dilute acid-catalysed steam explosion,” Bioresource Technology 86, 105-115

Fiserova, M., and Opalena, E. (2012). “Hemicelluloses extraction from beech wood with water and alkaline solutions,” Wood Research 57(4), 505-514.

Helmerius, J., von Walter, J. V., Rova, U., Berglund, K. A., and Hodge, D. B. (2010). “Impact of hemicelluloses pre-extraction for bioconversion on birch kraft pulp properties,” Bioresource Technology 101(15), 5996-6005. DOI: 10.1016/j.biortech.2010.03.029

Jeffries, W. (1990). “Biodegradation of lignin carbohydrate complexes,” Biodegradation 1(2-3), 163-176. DOI:10.1007/BF00058834

Jun, A., Tscirner, U. W., and Tauer, Z. (2012). “Hemicellulose extraction from aspen chips prior to kraft pulping utilizing kraft white liquor,” Biomass and Bioenergy 37, 229-236. DOI: 10.1016/j.biombioe.2011.12.008

Liu, W., Yuan, Z., Mao, C., Hou, Q., and Li, K. (2011). “Removal of hemicelluloses by NaOH pre-extraction from aspen chips prior to mechanical pulping,” BioResources 6(3), 3469-3480. DOI: 10.15376/biores.6.3.3469-3480

Magaton, A. D. S., Colodette, J. L., Gomes Gouvea, A. D. F., Gomide, J. L., dos Santos Muguet, M. C., and Pedrazzi, C. (2009). “Eucalyptus wood quality and its impact on kraft pulp production and use,” Tappi Journal 8(8), 32-39.

Macleod, M. (2007). “The top ten factors in kraft pulp yield,” Paper and Timber 89(4), 1- 7.

Malan, F. S. (1988). “Genetic variation in some growth and wood properties among 18 full-sib families of South African grown Eucalyptus grandis: (Hill ex Maiden): A preliminary investigation,” South African Forestry Journal 146(1), 38-43.

Mateos-Espejel, E. Moshkelani, M., Keshtkar, and Jaris, P. (2011). “Sustainability of the green integrated forest biorefinery: A question of energy,” Journal of Science and Technology for Forest Products and Processes 1(1), 55-61.

Mendes, C. V. T., Rocha, J. M. S., Sousa, G. D. A., and Carvalho, G. V. S. (2011). “Extraction of hemicelluloses prior to kraft cooking: A step for an integrated biorefinery in the pulp mill,” O Papel72(9), 79-83.

Myburgh, H. H. (1967). “Kraft pulping of eucalyptus in South Africa,” Appita 21(2), 49-53.

Ragauskas, A. J., Nagy, M., Kim, D. H., Eckert, C. A., Hallett, J. P., and Liotta, C. L. (2006). “From wood to fuels: Integrating biofuels and pulp production,” Industrial Biotechnology 2(1), 55-65. DOI:10.1089/ind.2006.2.55.

Reyes, P., Pereira, M., and Mendonça, R. T. (2015). “Effect of partial pre-extraction of hemicelluloses on the properties of Pinus radiata chemimechanical pulps,” BioResources 10(4), 7442-7454. DOI: 10.15376/biores.10.4.7442-7454.

Sluiter, A., Hames, B., Hyman, D., Payne, C., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., and Wolfe, J. (2008a). “Determination of total solids in biomass and total dissolved solids in liquid process samples,” National Renewable Energy Laboratory, Golden, Colorado. NREL Technical Report No. NREL/TP-510-42621.

Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., and Templeton, D. (2008b). “Determination of ash in biomass,” National Renewable Energy Laboratory Analytical Procedure (LAP), NREL Technical Report No NREL/TP-510-42622

Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., and Crocker, D. (2007). “Determination of structural carbohydrates and lignin in biomass,” National Renewable Energy Laboratory (NREL) Laboratory Analytical Procedures (LAP) for Standard Biomass Analysis.

Sluiter, A., Ruiz, R., Scarlata, C., Sluiter, J., and Templeton, D. (2008c). “Determination of extractives in biomass,” Technical Report NREL/TP-510-42619. National Renewable Energy Laboratory, Golden, Colorado.

Studer, M. H., DeMartini, J. D., Davis, M. F., Sykes, R. W., Davison, B., Keller, M., Tuskan, G. E., and Wyman, C. E. (2011). “Lignin content in natural Populus variants affects sugar release,” Proceedings of the National Academy of Sciences of the United States of America 108(15), 6300-6305.

TAPPI Test Methods, T 403 om-08. (2008). “Bursting strength of paper,” in: TAPPI Standard Methods for Pulp and PaperTechnical Association of Pulp and Paper Industry, TAPPI Press, Atlanta.

TAPPI Test Methods, T410 om-08 (2008), “Grammage of Paper and Paperboard (Weight per Unit Area),” TAPPI Press, Atlanta.

TAPPI Test Methods, T414 om-98. (1998). “Internal tearing resistance of paper (Elmendorf-type method),” TAPPI Press, Atlanta.

TAPPI Test Methods T494 om-96. (1996). “Tensile properties of paper and paperboard (using constant rate of elongation apparatus),” TAPPI Press, Atlanta.

TAPPI Test Methods, T220 om-88. (1992). “Physical testing of pulp handsheets,” TAPPI Press, Atlanta.

Tran, H., and Vakkilainannen, E. K. (2012). “The kraft chemical recovery process,” [Online Tappi Manuscript] [Accessed 27 March 2012]

Tunc, M. S., Lawoko, M., and van Heiningen, A. R. (2010). “Understanding the limitations of removal of hemicelluloses during autohydrolysis of a mixture of southern hardwoods,” BioResources 5(1), 356-371. DOI: 10.15376/biores.5.1.356-371

Um, B. H., and van Walsum, G. P. (2010). “Mass balance on green liquor pre-pulping extraction of northeast mixed hardwood,” Bioresource Technology 101(15), 5978-5987.

van Heiningen, A. (2006). “Converting a kraft pulp mill into an integrated forest biorefinery,” Pulp and Paper Canada 107(6), 38-43.

Vena, P. F., García-Aparicio, M. P., Brienzo, M., Görgens, J. F., and Rypstra, T. (2013). “Effect of alkaline hemicellulose extraction on kraft pulp fibers from Eucalyptus grandis,” Journal of Wood Chemistry and Technology 33(3), 157-173. DOI: 10.1080/02773813.2013.773040

Walton, S. L., Hutto, D., Genco, J. M., Walsum, G. P. V., and van Heiningen, A. R. P. (2010). “Pre-extraction of hemicelluloses from hardwood chips using an alkaline wood pulping solution followed by kraft pulping of the extracted wood chips,” Industrial and Engineering Chemistry Research 12638-12645. DOI: 10.1021/ie100848p

Wan, J., Wang, Y., and Xiao, Q. (2010). “Effects of hemicellulose removal on cellulose fiber structure and recycling characteristics of eucalyptus pulp,” Bioresource Technology 101(12), 4577-4583. DOI:10.1016/j.biortech.2010.01.026

Article submitted: May 5, 2015; Peer review completed: October 13, 2015; Revised version received and accepted; January 5, 2016; Published: January 26, 2016.

DOI: 10.15376/biores.11.1.2417-2437