Abstract
Autohydrolysis and kraft pulping were sequentially applied to Eucalyptus urograndis wood to obtain added-value products from hydrolysis liquor. A biorefinery approach was used to bleach the resulting solid phase containing the cellulose pulp with an optimized O–D–(EP)–D bleaching sequence, where O denotes delignification with the sequence D bleaching with chlorine dioxide and EP alkaline extraction with soda and hydrogen peroxide. The pulp was then beaten to obtain paper sheets. The two-stage process yielded pulp with a small Kappa number relative to conventional delignification (4.9 to 13.7 vs 17.1). Using autohydrolysis temperatures of 160 °C to 170 °C during 0 min to 15 min made the process selective towards the extraction of hemicelluloses (xylan up to 43.9%). Additionally, the Kappa number and brightness obtained after bleaching with an O-D-EP-D sequence were suitable for preparing dissolving grade pulp. The brightness of the pulp obtained was identical with that of the reference pulp (91.0%) and the Kappa number was smaller (0.2 to 0.6). Beating the pulp for 4500 PFI revolutions produced paper sheets with slightly better tensile strength and tear index than those for sheets from the reference pulp (viz., 90.3 N·m·g-1 and 9.9 mN·m²·g-1, respectively, versus 89.8 N·m·g-1 and 9.7 mN·m²·g-1, respectively).
Download PDF
Full Article
Autohydrolysis, Pulping, and Bleaching of Eucalyptus urograndis in a Biorefinery Framework
Javier M. Loaiza Rodriguez,a,* Jorge L. Colodette,b Juan C. García,a and Francisco López a
Autohydrolysis and kraft pulping were sequentially applied to Eucalyptus urograndis wood to obtain added-value products from hydrolysis liquor. A biorefinery approach was used to bleach the resulting solid phase containing the cellulose pulp with an optimized O–D–(EP)–D bleaching sequence, where O denotes delignification with the sequence D bleaching with chlorine dioxide and EP alkaline extraction with soda and hydrogen peroxide. The pulp was then beaten to obtain paper sheets. The two-stage process yielded pulp with a small Kappa number relative to conventional delignification (4.9 to 13.7 vs 17.1). Using autohydrolysis temperatures of 160 °C to 170 °C during 0 min to 15 min made the process selective towards the extraction of hemicelluloses (xylan up to 43.9%). Additionally, the Kappa number and brightness obtained after bleaching with an O-D-EP-D sequence were suitable for preparing dissolving grade pulp. The brightness of the pulp obtained was identical with that of the reference pulp (91.0%) and the Kappa number was smaller (0.2 to 0.6). Beating the pulp for 4500 PFI revolutions produced paper sheets with slightly better tensile strength and tear index than those for sheets from the reference pulp (viz., 90.3 N·m·g-1 and 9.9 mN·m²·g-1, respectively, versus 89.8 N·m·g-1 and 9.7 mN·m²·g-1, respectively).
Keywords: Autohydrolysis; Biorefinery; Dissolving pulp; Hemicelluloses; Kraft pulping
Contact information: a: Research Center in Technology of Products and Chemical Processes, PRO²TECS- Chemical Engineering Department, University of Huelva, Av. 3 de Marzo s/n, 21071 Huelva, Spain; and b: Department of Forest Engineering, Federal University of Viçosa, Avenue Purdue, s/n 36570-900, Viçosa, Brazil; *Corresponding author: javiermauricio.loiza@diq.uhu.es
INTRODUCTION
The related paper properties depend not only on the characteristics of the raw material and the pulping method used, but also on how the pulp is refined. Thus, an adequate hemicellulose content is essential for obtaining strong paper. This is difficult to accomplish because of the complex biorefining processes in which the lignocellulosic materials are subjected to the sequential separation of the hemicellulose, cellulose, and lignin fractions for use as feedstock for biofuel and bioplastic productions (Duarte et al. 2010).
Technologists serving the pulp and paper industry have long assumed that over-extraction of hemicelluloses detracts from some properties of paper sheets obtained from the solid phase remaining after an autohydrolysis treatment (Saukkonen et al. 2012). In fact, some strength-related properties of paper are impaired by treatments such as alkaline extraction, vapour explosion, hot water extraction, or autohydrolysis (Martín-Sampedro et al. 2010; Mendes et al. 2011; Vila et al. 2012). Hydrolysis pretreatments are used to open the wood structure of mechanical pulp to facilitate beating and thus save energy. However, with dissolving pulp, acid hydrolysis substantially affects the cellulose fraction (Kencaly et al. 2007; Jahan 2009). In any case, the resulting loss of handsheet strength can be reduced by avoiding excessive extraction of hemicelluloses, especially xylan (Chirat et al. 2012). Thus, the viscosity of hydrolysed pulp can be preserved by using milder operating conditions than that used in conventional pulping processes (Vila et al. 2012). Obviously, those processes initially extract hemicelluloses via autohydrolysis and diminish the pulp yields by over-extraction of hemicelluloses and by boosting the peeling reactions in the cellulose fraction (Chirat et al. 2012).
Although autohydrolysis is especially interesting for biorefinery schemes, the hemicelluloses can be efficiently extracted with alternative hydrolytic, acidic, or alkaline treatments, which provide liquor containing soluble oligomers or low-molecular weight polymers. These liquid phases can be used to obtain a variety of oligomers with high applicability in the chemical, oil, pharmaceutical, and food industries (Garcia et al. 2013; Adom 2012; Cheng et al. 2008). In a conventional process (i.e., one involving no prior hydrolysis), such polymers are degraded to a great extent and were incorporated into the black liquor, which can thus be used for energy production (Bhaskaran and Von Koeppen 1970). As new technologies affording renewable and sustainable production emerge (Ragauskas et al. 2006; Kamm et al. 2009), biorefinery is increasingly expected to raise the benefits of processing lignocellulosic biomass fractions and hence to expand business opportunities for the silvicultural sector (Hess et al. 2007; Chambost et al. 2008). In fact, cellulose pulp mills can be the starting point for the integral production of energy and a wide range of chemicals from lignocellulosic biomass (Clark 2007; Kamm and Kamm 2007; Sanders et al. 2009). Additionally, extracting hemicelluloses should facilitate the delignification, thereby expediting reactions or reducing the reagents needed to obtain a given target Kappa number (Vila et al. 2011). However, in some cases, extracting hemicelluloses reduces intrinsic viscosity in the resulting pulp through depolymerisation of cellulose.
Accepting the idea that the extraction of hemicelluloses provides additional value (in the use of lignocellulosic material), kraft pulping process can be used for purposes other than conventional paper production. Dissolving pulp is one example of such an alternative purpose. In fact, the dissolving pulp market requires pulp with a high cellulose purity of typically 95% to 98% α-cellulose, only 1% to 10% hemicellulose, and less than 0.05% lignin (Christov et al. 1998). With kraft pulps, these cellulose purity levels can be obtained with a bleaching process carried out to reach a high brightness degree. Because the yields rarely exceed 30% to 35%, the dissolving pulp is more expensive to produce than the conventional cellulose pulp. For this reason, the dissolving pulp is often used to obtain various value-added cellulose derivatives such as cellulose acetate, viscose, cellulose nitrate, carboxymethyl cellulose, or methylcellulose.
The dissolving pulp can be obtained not only from wood but also from other raw materials, such as cotton linters, which are kraft-pulped following sulphite-hydrolysis (Barba et al. 2002).
Returning to the argument of the first paragraphs about that over-extraction of hemicelluloses detracting from some properties of paper sheets, hemicelluloses are hydrophilic and account for most of the water-retention capacity of fibres in beaten cellulose pulp. In fact, the hemicelluloses make fibres more flexible and increase fibre contact areas, thereby improving its intertwining property. The resulting increased fibre cohesion reduces the beating energy required for specific levels of mechanical strength to be obtained (Barzyc et al. 1997; Laine and Hynynen 1997; Warberg and Annergren 1997). However, an increased water-retention capacity of fibres diminishes drainability in the resulting paper and can compromise the operation of papermaking machines, which usually work at their top drainage rate. Whether pulp can be properly refined under these conditions depends on a balance between the mechanical strength and fibre intrinsic performance (Ferreira et al. 1998). Because hemicelluloses facilitate refining, they should only be removed to produce pulp for tissue paper (Santos 2005).
Köhnke and Gatenholm (2007) altered cellulose fibres to control their adsorption capacity for hemicellulose and examined their impact on the properties of pulp from leafy wood. They found that the hemicellulose adsorption by fibres increases cellulose stability by preventing microfibril aggregation upon drying and hence minimizing hornification of fibres. According to Oliveira and Sabioni (2002), the hornification causes fibres to harden after drying, which detracts from their suitability for papermaking. This phenomenon arises during the cooking, bleaching, and drying of pulp, which are usually separately completed.
In this work, the authors used Eucalyptus urograndis wood as the raw material for pulping. This is one of the most widely grown plant species in Brazil’s forests, and also one of the most widely used species for wood for kraft pulping. The material was subjected to sequential autohydrolysis and kraft pulping as a means to valorize the hydrolysis liquor. Additionally, the remaining solid phase (cellulose pulp) was processed under biorefinery principles by using an optimized OD–EP–D bleaching sequence using pressurized oxygen, hydrogen peroxide, and alkali extraction, and then beaten to obtain handsheets.
EXPERIMENTAL
Raw Material and Autohydrolysis Processing
The raw material used was E. urograndis chips obtained from a pulp and paper mill in central-eastern Brazil (Suzano, Salvador de Bahia, Brazil). The material was subjected to the treatments described below at the cellulose and paper laboratory of the Forestry Engineering Department of the Federal University of Viçosa, Brazil. The wood chips were classified according to the SCANCN 40:94 (1993) standard, dried to approximately 15% moisture, and then stored until further processing. For chemical analysis (the same analysis than hydrolysed material -see below in this section-), the chips were ground and sieved through an 8-mm mesh according TAPPI T-264-cm-07 (2007). No particle diffusion constraints were detected under these conditions. The ground wood samples were allowed to dry at ambient temperature, homogenized in different batches, and stored in air-tight vessels.
The E. urograndis chips (chip size: long 25 mm to 30 mm, width 15 mm to 20 mm and thickness 2 mm to 3 mm ) used in the autohydrolysis tests were processed in a 10-L jacketed reactor from MK-Systems, Inc. (Danvers, MA, USA) furnished with a liquor recycling system. The operating temperature used was 160 °C to 180 °C, and the treatment time was up to 30 min. The time needed for the reactor to reach the processing temperature was 43 min, and the water to wood mass ratio was 10:1. The recycling system allowed the reactor to be externally cooled down to 25 °C before it was opened.
The hydrolysed material was mechanically beaten and sieved through 40/60-mesh on a Wiley mill (Arthur H. Thomas Co., Philadelphia, PA, USA). The hydrolysed material was allowed to dry under ambient conditions (23 °C and 50 ± 2% relative humidity) prior to storage in air-tight containers. These samples were chemically characterized according to the following standards: TAPPI T264 cm-07 (2007) for moisture content and TAPPI T204-om-07 (2007) with Soxhlet extraction (95% ethanol for 5 h) for extractables. After removing ethanol extracts, the solid samples were analysed for lignin and sugars after acid hydrolysis according to TAPPI T249-em-09 (2009). Klason lignin content was determined according to TAPPI T222 om-11 (2011).
The sugar monomers in the hydrolysis liquor were determined according to Wallis et al. (1996) using a 940 Professional IC Vario ion chromatograph from Metrohm AG (Herisau, Switzerland) that was equipped with a pulsed amperometric detector and a CarboPac PAI column (Thermo Fisher Scientific, Waltham, MA, USA). A sample volume of 25 µL was injected at a flow rate of 1 mL/min in each run. The instrument was calibrated with glucose, xylose, galactose, mannose, and arabinose standards, and fructose was used as an internal standard.
Pulping and bleaching
After the autohydrolysis process, a kraft delignification was completed using the following conditions: alkali concentration, 25% (on dry wood weight); sulphidity, 31% (on dry wood weight); temperature, 167 °C; and operating time of 45 min. As noted earlier, the reactor warm-up time was 43 min. However, the amount of water needed to impregnate the material was much smaller than that used for autohydrolysis, and the liquid/solid mass ratio of 4:1 was used. As in the autohydrolysis tests, recycling the liquor ensured efficient mixing and uniform delignification of the previously hydrolysed wood chips.
Cellulose pulps (from the solid phase after autohydrolysis) were collected in the same reactor used for the hydrothermal pretreatment. In addition, a pulp without autohydrolysis was performed under the same operating conditions as a reference experiment.
The black liquor from the pulping process was filtered off, and the pulp was beaten in a PFI mill at 2000 rpm for 10 min. The pulp samples thus obtained were analysed for yield, Kappa number, and intrinsic viscosity according to TAPPI T257 cm- 85 (1999), TAPPI T236 cm-85 (1996), and TAPPI T230 om-99 (1999), respectively. Paper properties were assessed according to TAPPI T205 sp-02 (2014) using handsheets previously obtained with an ENJO-F-39.71 sheet former (IDM Test, San Sebastián, Spain).
Table 1. General Conditions Used for Bleaching of Pre-hydrolysis Kraft Pulps
In a kraft process, bleaching can produce large amounts of effluents containing toxic substances. Replacing the typical bleaching sequences based on chlorinated reagents with elemental chlorine-free (ECF) sequences can help to reduce effluent toxicity (García et al. 2010). In this work, autohydrolysed kraft pulp was subjected to an O–D–(EP)–D bleaching sequence where O denotes delignification with sequence, D bleaching with chlorine dioxide, and EP alkaline extraction with soda and hydrogen peroxide. The operating conditions for each step in the sequence are shown in Table 1.
Methods
PFI beating, sheet formation, and physical testing
Pulp samples were beaten in a PFI mill (IDM Test, San Sebastián, Spain) in accordance with TAPPI T248 sp-08 (2008). As per TAPPI T200 sp-01 (2002), the beating intensity was expressed in Schopper–Riegler degrees (ºSR) or as energy consumption (W·h). The samples (an amount of 30 g oven-dry pulp in each run) were beaten at a variable rate (0 to 4500 PFI revolutions) to obtain three levels of pulp refinement spanning the range of 15 ºSR to 50 ºSR (0 W·h to 70 W·h).
Formation of sheets for physico–mechanical and optical tests
The paper sheets used in the tests were obtained on a sheet former according to TAPPI T205 sp-02 (2014). The sheets, 60 g/m2 in aerial weight according to TAPPI T205 sp-02 (2014), were conditioned at 23 ± 1 °C and a relative humidity of 50% ± 2% for testing.
The light scatter coefficient was determined on a Datacolor Elrepho 450X instrument (Datacolor Inc., Lawrenceville, NJ, USA) at two different points in each sheet. The readings obtained at each beating level were averaged for subsequent processing. Tensile strength tests were conducted on Instron equipment (Instron, Barcelona, Spain), using a grip distance of 100 mm, a test speed of 25 mm/min, a loading cell of 1000 N, and a computer-controlled data acquisition and processing system. The tear strength, burst strength, and air pass were determined on Elmendorf and Müllen equipment (IDM Test, San Sebastián, Spain).
Experimental design for the pulping process – multiple regression modelling
The pulping process was modelled and optimized using a polynomial model approach considering linear and quadratic terms of the process variables, and mutual interactions. Thus, the equations established were fitted by multiple regression. A central composition design was used to reduce the number of tests needed while ensuring the absence of significant covariances between the independent variables. These two tools allow the influence of several process variables on pulp and paper properties to be simultaneously assessed (Dong and Fricke 1996). The dependent variables, such as yield, glucan, xylan, mannose, ethanol extractives, intrinsic viscosity, and Kappa number were thus related to the independent variables of the kraft pulping process after autohydrolysis, viz., autohydrolysis temperature and time.
A 2n central composite factor design was used to minimize the number of tests needed while avoiding correlation in the covariance matrix. Modelling required previously normalizing the ranges spanned by the independent variables according to Eq. 1 and statistically identifying the significant influences in the coefficients. Thus, no term with a coefficient exceeding p = 0.05 in significance as per Student’s t-test, or with a 95% confidence interval, was included in the models,
(1)
where X is the absolute value of the independent variable concerned, X̅ is the mean value, and Xmax and Xmin are the maximum and minimum values, respectively.
Three levels each of the autohydrolysis process (temperature 160 °C, 170 °C, and 180 °C) and operation time (0 min, 15 min, and 30 min) were used. Operational conditions in kraft pulping were fixed: alkali concentration, 25%; sulphidity, 31%; temperature, 167 °C; and operating time, 45 min. A liquid/solid ratio of 4/1 were used in all experiments.
The minimum number of tests needed, N, was calculated to be 2n + 2n + c, where n is the number of independent variables and c that of replicates of the central point in the experimental design (Eq. 2). Therefore, N was 22 + 2×2 + 2 = 10. The experimental results were fitted to the following second-order polynomial,
(2)
where X are the independent variables and Y are the dependent variables. The coefficients, ao, bi, ci, and dij are constant unknown characteristics, estimated from the experimental data.
The results were assessed with STATISTICA 10.0 software (StatSoft, Inc., Tulsa, OK, USA). In addition, global model adjustment statistics, such as R2 and Snedecor-F have been used. Levels of R2 higher than 0.85 or Snedecor-F higher than 5 were found as suitable.
RESULTS AND DISCUSSION
Characteristics of the Raw Material
The chemical characterization of Eucalyptus urograndis was determined. The ethanol extractives content obtained was 3.7%. The chemical compositional analysis revealed that the major content was cellulose, which accounted for 47.0% of the mass, followed by Klason lignin (24.4% after quantitative acid hydrolysis), and hemicelluloses (In the present case, to estimate the efficiency of autohydrolysis, hemicellulose was calculated as the combination of xylose 11.0%, arabinose 0.25%, galactose 0.85%, mannose 1.1%, and acetyl content 3.1%) with 16.3%. According to the Wise et al. (1964) method, the holocellulose content was 71.9%. Therefore, the hemicelluloses content could also be calculated as hollocellulose – glucan (or α-cellulose) = 25.0%
This composition was similar to other eucalyptus varieties used in biorefinery (Table 2): Eucalyptus globulus (Garrote and Parajó 2002; López et al. 2008; Loaiza et al. 2016), E. grandis, and Eucalyptus urograndis (Da Silva Morais et al. 2016).
Autohydrolyzed Solids and Kraft Pulping
The main purpose of an autohydrolysis treatment preceding kraft pulping is to obtain a liquor containing abundantly extracted xylo-oligosaccharides and monomeric sugars, such as xylose and acetic acid. Some authors have suggested that the resulting solid phase can be used to obtain added-value products (Testova et al. 2011). For example, if it contains more than 15% to 20% pentosans, then it can be useful to produce furfural (Yahyazadeh 2011).
Table 2. Results of Chemical Characterization of Eucalyptus globulus, E. grandis, and E. urograndis from the Literature
Table 3 shows the results of the pulping process, and the normalized values of the dependent variables in accordance with the proposed experimental design (autohydrolysis with kraft delignification). It also shows the results of a reference experiment (kraft pulping without autohydrolysis). Each result is the average of 5 replicate determinations and deviations from the mean were all less than 5%.
Table 3. Normalized Values of the Process Variables and Characterization of Pulp Obtained by Autohydrolysis + Kraft Pulping in Accordance with the Experimental Design
The polynomial models relating the dependent variables to the independent variables (Table 4) were obtained by fitting the non-linear regression models of Eq. 2 to the data of Table 3. Fitting was invariably very good (R2 > 0.96; Snedecor’s F > 100). The residuals between the experimental values of the dependent variables and their estimates from Eq. 4 were also very good, in fact, all differences (residuals) were less than 5% of the range spanned by each dependent variable.
Table 4. Polynomial Models for the Dependent Variables of the Autohydrolysis + Pulping Process
Dependent variables: Yield (Y, %), Kappa number (K), intrinsic viscosity (V, cm3.g-1), glucan (G, %), ethanol extractives (E, %), and xylan (X, %); Independent variables: Temperature (XT) and time (Xt); Normalized values for the independent variables were used in equations from Table 3
The kraft pulping yields obtained for E. urograndis wood with autohydrolysis were lower than those obtained with the reference pulp. This was because the raw material was subjected to two processes to remove the hemicellulose components. In fact, the Kappa numbers were low for the autohydrolysed raw material than that of the reference pulp.
This observation was consistent with the widespread belief in the paper industry that the pulp should contain a certain amount of hemicellulose, whose content influences the quality of the resulting paper. In fact, more lignin, and a high proportion of hemicellulose were lost in the pulping liquor during conventional industrial processes (Cruz et al. 2012). The most labile fraction, the hemicellulose-lignin complex (containing approximately 15 to 25% of all lignin and 40% of all hemicellulose) was rapidly removed early in the process (Cenunez 2018; Villar 2018). The reference kraft pulp contained approximately 6.02% xylan. This amount was equivalent to 26.6% of all xylan present in the raw material, which was therefore 73.4% extracted into the autohydrolysis liquor. Additionally, based on the total content of hemicelluloses, approximately 16.3% (or 25% as the difference from the glucan content) was extracted 82.1% to 88.3% into the liquor. One of the hypotheses (Hypothesis 1) for this work was that removing hemicelluloses early by autohydrolysis would not have any adverse effect on the hemicellulose content of the final pulp to detract from the strength-related properties in the resulting paper sheets. Another hypothesis (Hypothesis 2) was based on the previous one, and was that the autohydrolysis treatment would provide a valorizable liquor containing a substantial amount of hemicelluloses.
A total of 10 preliminary tests under the temperature and time conditions of Table 3 provided solid phase yields from 75.5% to 95.7% and hemicelluloses contents as the combination of araban, galactan, xylan, and mannose contents from 3% to 13.1%. Therefore, hemicelluloses were extracted to the extent of 86.1% to 90.9% with a hemicellulose content of 16.3% in the raw material and 21.7% to 49.0% (25% based on difference between holocellulose and glucan). The hemicellulose extraction in the autohydrolysis liquors in tests 4, 6, and 8 were those whose samples were closest to the reference pulp in xylan content (Table 3) and hence was chosen for subsequent bleaching and beating. This sample had 21.7% to 32.5% initial hemicellulose content (16.3% is the hemicellulose content in the raw material) or 49.0% to 56.0% (with one of (25%, based on difference between holocellulose and glucan in raw material)
If Hypothesis 1 were to be fulfilled, a given hemicellulose or xylan content of the pulp would exist above which the strength-related properties of the resulting paper would not be substantially better. Based on Eq. 6 in Table 4, the xylan content at experimental points 4, 6, and 8 —non-isothermal autohydrolysis operation at 160 °C for 0 min, non-isothermal conditions at 170 °C, and 160 °C for 15 min, respectively— corresponded to the kraft pulp samples with the highest xylan contents (82.6%, 44.7%, and 52.3%, respectively, of that in the reference pulp). In other words, based on the combined yield of autohydrolysis and delignification, 79.0%, 89.1%, and 87.9%, respectively, of xylan initially present in the raw material (11%) was dissolved versus 73.4% in the reference pulping process. The proportion of dissolved hemicelluloses ranged from 85.8% to 90.7% in test 4, 92.7% to 95.2% in test 6, and 91.8% to 94.7% in test 8 with an initial content of 16.3% (or 25% if expressed as the difference between holocellulose and glucan). By contrast, the hemicelluloses in the reference raw material were dissolved to the extent of 82.1% to 88.3%. The previous experimental points (4, 6, and 8) were used for the subsequent bleaching and refinability tests. The increased hemicellulose contents of the pulp samples obtained under the corresponding operating conditions coincided with the highest viscosity values in the studied operating ranges.
In regards to other pulp properties shown in Table 3 or modelled in Table 4, the autohydrolysis treatment was especially influential on the Kappa number, which ranged from 4.92 under the strongest operating conditions to 13.7 under the mildest, and also on the pulp yield, which was 32.6% and 46.6%, respectively.
To better envision the influence of the operational variables on the chemical properties, and to directly compare with the result obtained from reference pulp, the response surface plots (Figs. 1 to 4) were constructed.
Fig. 1. Variation of solid yield after kraft pulping as a function of temperature (XT) and operation time (Xt) in the pre-autohydrolysis process
As shown in Fig. 1, the solid yield changed corresponding to the Kappa number. Thus, it was markedly influenced by temperature and time in the higher ranges of the operating conditions (viz., 180 °C and 30 min); also, their interaction term in Eq. 1 in Table 4 was statistically significant, but minimally relative to the other coefficients for the model. Using temperatures in the lower range, and times in the central and high ranges, was predicted to provide yields of 40.4%, and hence 16.7% lower than those for the reference pulp.
As shown in Fig. 2, the Kappa number decreased with increasing severity of the operating conditions, which, based on the previous conclusion, should lead to a significant loss of pulp viscosity. However, the interaction term for the two variables in Eq. 2 in Table 4 was not statistically significant. Additionally, the predicted Kappa numbers for temperatures in the lower range (160 °C) and times in the higher range (15 min to 30 min) were 9.5 and 10.5, respectively, which were considerably greater than the calculated values for the higher ranges of both variables (4.7). However the Kappa numbers were substantially (44.5% to 38.7%) smaller than those for the reference pulp with an identical level of viscosity. In fact, the predicted value was 1022 cm3.g-1 and not significantly different from that of the reference pulp (1017 cm3.g-1).
Fig. 2. Variation of Kappa number after kraft pulping as a function of temperature (XT) and operation time (Xt) in the pre-autohydrolysis process
Figure 3 shows the response surface plot for intrinsic viscosity. As shown, an increased autohydrolysis temperature and time had an adverse effect on the pulp viscosity. However, the effect was a result of the synergistic combination of the two independent variables. Thus, the results were approximately 50% smaller under the strongest autohydrolysis conditions (viz., 180 °C and 30 min). The response surface effect was consistent with Eq. 3 in Table 4, where the quadratic terms for the independent variables were not significant but that for their interaction was significant. Additionally, with temperatures close to the lower range, the operating time had virtually no effect on the viscosity. In fact, using temperatures in the lower range resulted in no decrease in viscosity relative to the reference pulp. As a rule, temperatures and time slightly exceeding the central ranges for the two variables (170 °C and 15 min) reduced viscosity by no more than 15% to 20% relative to the highest level calculated with the model (1048 cm3.g-1), which would result in the treated pulp being comparable to or even slightly better than the reference pulp in terms of viscosity.
Fig. 3. Variation of intrinsic viscosity after kraft pulping as a function of temperature (XT) and operation time (Xt) in the pre-autohydrolysis process
Because the pulp was to be bleached after autohydrolysis, the operating conditions used should simultaneously ensure maximal viscosity and yield. Based on Figs. 1 and 3, the pulp yield and viscosity could be increased to some extent by using temperatures in the lower range (160 °C) and times in the central range (15 min), albeit with a slightly greater Kappa number. These conditions should provide a pulp yield of 42.2%, a viscosity of 1034 cm3.g-1, and a Kappa number of 10.7, which were 13.1% lower, only 1.6% higher, and 37.3% smaller —and thus better— than the respective values for the reference pulp. The yield loss relative to the reference pulp under these conditions (6.33%) was largely (45.3%) ascribed to the difference in xylan content between the reference pulp (Table 3) and that calculated from Eq. 6 in Table 4.