- Email: [email protected]
Integration of a kraft pulping mill into a forest biorefinery: Pre-extraction of hemicellulose by steam explosion versus steam treatment
Bioresource Technology 153 (2014) 236–244
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Integration of a kraft pulping mill into a forest biorefinery: Pre-extraction of hemicellulose by steam explosion versus steam treatment Raquel Martin-Sampedro ⇑, Maria E. Eugenio, Jassir A. Moreno, Esteban Revilla, Juan C. Villar Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria INIA, Carretera de la Coruña, km 7.5, 28040 Madrid, Spain
h i g h l i g h t s Steam explosion and steam pre-treatments were carried out before kraft pulping. The rate of extraction of hemicelluloses was similar with both pre-treatments. Steam explosion enhanced delignification more efficiently. Pre-treatments reduced paper mechanical properties owing to the fiber morphology. Paper optical properties were boosted when pre-treatments were carried out.
a r t i c l e
i n f o
Article history: Received 11 October 2013 Received in revised form 25 November 2013 Accepted 30 November 2013 Available online 8 December 2013 Keywords: Autohydrolysis Biorefineries Kraft pulping Steam explosion Eucalyptus globulus
a b s t r a c t Growing interest in alternative and renewable energy sources has brought increasing attention to the integration of a pulp mill into a forest biorefinery, where other products could be produced in addition to pulp. To achieve this goal, hemicelluloses were extracted, either by steam explosion or by steam treatment, from Eucalyptus globulus wood prior to pulping. The effects of both pre-treatments in the subsequent kraft pulping and paper strength were evaluated. Results showed a similar degree of hemicelluloses extraction with both options (32–67% of pentosans), which increased with the severity of the conditions applied. Although both pre-treatments increased delignification during pulping, steam explosion was significantly better: 12.9 kappa number vs 22.6 for similar steam unexploded pulps and 40.7 for control pulp. Finally, similar reductions in paper strength were found regardless of the type of treatment and conditions assayed, which is attributed to the increase of curled and kinked fibers. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Integration of a lignocellulose biorefinery into a pulp mill is a feasible possibility to add value to the production of pulp, a product considered as a commodity with a moderate added value. The integrated biorefinery-pulp mill would bring forth products of higher added-value, such as furfural, carbon fibers, biopolymers or biofuels, in addition to pulp (Helmerius et al., 2010). One strategy to accomplish this integration is to use a fraction of hemicelluloses which, in a conventional kraft cooking, are dissolved into the cooking liquor and burned in the recovery boiler jointly with the lignin degradation products. While, combustion of the non-cellulosic compounds provides the necessary supply of energy to maintain the pulp mill or even to have a surplus, removal of hemicelluloses prior to cooking does not necessarily
⇑ Corresponding author. Tel.: +34 913476834; fax: +34 913476767. E-mail address: [email protected] (R. Martin-Sampedro). 0960-8524/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.biortech.2013.11.088
affect the final energy balance because their heating value is approximately half that of lignin, and cooking pre-extracted chips requires considerably less energy than cooking normal chips, as the cooking time of the former is significantly shorter (Martín-Sampedro et al., 2011a). The removed hemicelluloses could be hydrolyzed into sugars and then fermented to ethanol or derived to other value-added products. These compounds are considered promising platform chemicals in the synthesis of many other valuable products (Tuck et al., 2012). Hemicelluloses extraction has been done by a variety of methods such as autohydrolysis, steam explosion, acid hydrolysis or alkali extraction. However, steam explosion and autohydrolysis pre-treatments only use water/steam at high temperature causing the formation of acetic acid from the acetylated wood component, which catalyzes hydrolytic reactions in the wood polymers. Thus, these hydrothermal pre-treatments can be considered as a green and competitive technology to remove hemicelluloses in hardwood, since reaction media contain only lignocellulosic feedstock
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244
and water, which prevents corrosion problems and the formation of neutralization sludges (Garrote et al., 1999; Martín-Sampedro et al., 2011a). The effect of these hydrothermal pre-treatments on a subsequent pulping process has been studied using a diversity of raw materials. Autohydrolysis has been applied as a pre-treatment on kraft, soda or ethanol–water pulping using pine (Kautto et al., 2010; Saukkonen et al., 2012), a mixture of maple, poplar and birch (Li et al., 2010), eucalyptus (Chirat et al., 2012; Mendes et al., 2009; Vila et al., 2011, 2012), aspen (Al-Dajani et al., 2009), birch (Helmerius et al., 2010), bagasse (Hamzeh et al., 2013) and Leucaena leucocephala K360 (Feria et al., 2012). Although in less extent than autohydrolysis, steam explosion pre-treatment has also been studied using different raw materials such as pine (San Martin et al., 1995), Norway spruce (Jedvert et al., 2012), a mixture of marple, birch and aspen (Ahvazi et al., 2007), Hesperaloe funifera (Martín-Sampedro et al., 2012b) and eucalyptus (Martín-Sampedro et al., 2011a,b). With both pre-treatments, hemicelluloses removal results in an improvement of the delignification rate, which reduces the cooking time and/or the chemical charge required for an equal target of kappa number, and enhances the subsequent bleaching process. In some cases, it has also been observed a decrease in pulp viscosity, due to the reduction of the cellulose polymerization degree and a general loss of mechanical properties. Nevertheless, the loss of pulp mechanical properties is moderate when the xylan removal is limited (Chirat et al., 2012), and the pulp viscosity could be maintained if preextracted chips are cooked at milder conditions than nonextracted wood (Vila et al., 2012). Finally, the decrease in pulp yield observed could be explained by the hemicellulose removal and an increase of the peeling reactions of a more exposed cellulose (Chirat et al., 2012). As all hydrothermal pre-treatments (stem treatment, hot water treatment, steam explosion, etc.) lead to a similar release of acids from acetylated wood components which catalyze hydrolytic reactions in the wood polymers (autohydrolysis), the main difference between steam treatment (sometimes called autohydrolysis treatment) and steam explosion is the rapid decompression that takes place at the end of steam explosion but not in steam treatment. This decompression forces the fibrous material to ‘‘explode’’ into separated fibers and fiber bundles, generating a solid fraction with a more open structure (Ahvazi et al., 2007; Martín-Sampedro et al., 2011a,c) that may enhance the efficient diffusion of cooking liquor into the fibers. Although both pre-treatments have been studied in different articles, as mentioned above, a comparison between them has not been previously reported. Therefore, the main objective of this study is to compare the effect of steam explosion and steam treatment on the subsequent kraft pulping of Eucalyptus globulus, in order to elucidate the separate effect of the steam treatment (causing autohydrolysis in both pre-treatments) and of the ‘‘explosion’’. Furthermore, the influence of the severity factor on the subsequent pulping process and pulp quality was also studied with both pre-treatments.
237
2.2. Steam explosion and steam treatments Steam explosion and steam treatments were performed in a 26 L stainless steel digester (manufactured by Cadepla S.L.) capable of temperatures to 190 °C and a pressure of 1.37 MPa (14 kgf cm2). The digester was connected to a blowing tank into which chips were discharged at the end of the treatment. It was also equipped with electrovalves for steam admission, and a ball valve of discharge. The steam generator was a Babcock Wanson VAP 250RR boiler, with a maximum steam production rate of 270 kg h1 and a working pressure of 1.37 MPa. According to previous reports (Martín-Sampedro et al., 2011a; Martín-Sampedro et al., 2011c), chips were immersed in water at 25 °C for 16 h in order to improve the efficiency of the subsequent steam pre-treatments. In all of the experiments, 500 g of E. globulus chips were treated with steam at 183 °C (10 kg-f cm2). The variable operational conditions were: number of cycles of treatment (one or two), duration of the first cycle (5 or 10 min), and discharge pressure (6 kg-f cm2 for steam explosion treatments or atmospheric pressure for steam treatments). When a second cycle was carried out, the pre-treated chips obtained in the first cycle were washed with cold water and then subjected to a second cycle of 3 min following the same procedure as in the first cycle. After treatment, the samples were thoroughly washed with water, dried at room temperature and stored in sealed polyethylene bags. The severity factor of each treatment was calculated according to the following equation (Eq. (1)) defined by Overend and Chornet (1987).
T100 S0 ¼ log e 14:75 t
ð1Þ
in which T is the temperature (°C) and t the duration of the treatment (min). Water retention, or hydration, capacity of the treated and untreated chips was defined as the weight of water absorbed by the chips after being immersed in water at 25 °C for 6 h. It was expressed as grams of water per 100 g of oven dry wood, according to the following equation:
WR ¼
Ww Wd 100 Wd
ð2Þ
where Ww was the weight of the wood chips after the water immersion; then, wood chips were dried at 104 °C during 24 h, and weighted again (dry weight, Wd). 2.3. Chemical analysis In order to carry out the chemical analysis, all the samples were dried at room temperature and then milled in a Wiley mill. The samples were sieved using standard sieves to obtain 20 g of wood meal sized between 0.30 and 0.40 mm. Acetone extractives (UNE-EN ISO 14453), hot water extractives (UNE 57-013-82), lignin content (TAPPI T 222 om-88), holocellulose content (Wise et al., 1946) and pentonsans content (TAPPI T 223 cm-84) were then measured in the wood meal. All determinations were duplicated.
2. Experimental
2.4. Kraft cooking process
2.1. Raw material
Kraft cooking was performed over control (non-treated), steam exploded and steam unexploded chips in pressurized 1-liter reactors. Four reactors were placed in a 20 L rotatory pressurized vessel that contained hot water for indirect heating of the reactors. The rotatory vessel had a jacket-type electrical heater controlled by a computer to set the cooking temperature. Cooking conditions were: 150 g of dry chips, 4 L/kg liquor to wood ratio, 16% active
E. globulus chips were kindly provided by La Montañanesa pulp mill (Torraspapel – Lecta Group, Spain). The material was air dried and then homogenized in a single stock (by conditioning inside polyethylene bags) to avoid differences in composition and water content. The chips were stored in polyethylene bags at 25 °C.
238
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244
alkali, 20% sulfidity, 160 °C cooking temperature, 60 min to maximum temperature and 30 min at maximum temperature. After processing each batch, the reactors were removed from the vessel, and the cooked chips were washed, disintegrated and screened to determine their reject content and screened kraft yields (based on the solid content of the material submitted to kraft pulping). Moreover, the overall yield was calculated taking into account both the steam explosion or steam pre-treatments and the kraft process. Black liquor was titrated with HCl to determine the chemical reagent consumption. The kappa number (TAPPI T-236/UNE 57034) and viscosity (UNE 57039-1) were determined in the pulp samples. 2.5. Morphological characterization of kraft fibers In order to elucidate the effect of the steam pre-treatments on fiber morphology, kraft fibers were analyzed with the morphological analyser Morfi V7.9.13.E (Techpap, France). The procedure was based on that described by Jarabo et al. (2012) and adapted for eucalyptus kraft pulps. Previously, the samples were prepared for morphological characterization by disintegrating 1 g of dry pulp into 600 ml of water in an ENJO-model 692 lab disintegrator. The characterization was done in duplicate. The Morfi fiber morphological characterization is based on an image analysis system, consisting of a diode which emits unpolarized light. The image acquisition is maintained until 5000 elements are analyzed. Some parameters are required to discriminate between the elements of interest (fibers, vessels and fines) and the others. These parameters were: fiber length, 100–6000 lm; fiber width, 5–75 lm; maximum fiber kink, 178°; maximum fiber curl, 98%; maximum vessels length/ width ratio, 1.5; and fines length, 1–100 lm. Fiber kink is defined as an abrupt change in the direction of the fiber and it is expressed in degrees. The curl is determined from the ratio between the arc (distance between the two ends of the fiber) and the developed length of the fiber, and it is given by the classical definition shown in Eq. (1), which for a completely straight fiber (arc equals to developed length) gives a 0% curl.
C ð%Þ ¼ 100 1
arc Developed length
ð3Þ
The characterization of the fibers was made in terms of fiber length, fiber width, coarseness, percentage of kinked fibers and angle or curl. Vessels were characterized by their length and width. The percentages of fibers, vessels and fines in the pulp suspension were also determined. 2.6. Refining and handsheets characterization In consistency with results reported elsewhere (Martín-Sampedro et al., 2011a), pulps were refined in a PFI mill (UNE-EN ISO 5264-2/TAPPI T-248) at 2000 revolutions. The refining degree (Schopper-Riegler) (UNE-EN ISO 5267-1) of the pulps was determined before and after refining. Handsheets were then formed from unrefined and refined pulps, in accordance with ISO UNE-EN ISO 5269-2, and characterized in terms of stretch at break, tensile and tear indexes (UNE-EN ISO 5270), brightness (UNE 57062), and yellowness index (TAPPI T-1216). 3. Results and discussion 3.1. Impact of steam explosion and steam pre-treatments on the resulting solid fraction Table 1 shows the chemical composition of the chips before (control) and after the different steam explosion or steam
pre-treatments. It can be observed that there was an increase in the content of extractives (acetone and hot water soluble material) in the wood chips after the different pre-treatments (steam explosion and steam treatment) compared to the control. These findings could be owing to the fact that during the pre-treatments some macromolecules are partially degraded becoming extractible in a subsequent solvent extraction. Similar findings have been reported by other authors (Chandra et al., 2007; Martín-Sampedro et al., 2011a; Rahikainen et al., 2013). Although the pre-treatments result in an increase of both acetone and hot water extractives, a different trend was observed under our operational conditions. Namely, the content of acetone extractives increased with the severity of the pre-treatment (duration and number of cycles) regardless of the pressure of discharge (6 kg-f cm2 and atmospheric pressure for steam explosion and steam treatment, respectively). However, the amount of extractives in hot water was greater in the steam unexploded samples compared to the steam exploded samples, likely owing to the removal of some hydrosoluble material during the rapid expansion of steam inside the chips. On the other hand, a decrease in hot water extractives content was observed when a second cycle was carried out. The reason could be the partial degradation of polysaccharides (mainly hemicellulose), which renders them readily extractible in the subsequent steam treatment (second cycle). Furthermore, the reduction in water extractives after the second cycle was more pronounced when using steam explosion as pre-treatment likely as a result of the first explosion (first cycle), which opens the wood internal structure and facilitates subsequent extraction during the second cycle. Percentages of lignin, holocellulose and pentosans removed from the original wood were calculated for each pre-treatment (Fig. 1), taking into account the chemical composition of the samples (Table 1) and the yield of the pre-treatments. Although steam explosion and steam treatments are known as selective methods for the removal of hemicelluloses from the wood matrix, a significant amount of lignin is colloidally dissolved during these pretreatments (Gutsch et al., 2012; Leschinsky et al., 2009). The amounts of lignin removed with one cycle treatments were similar with steam explosion and steam pre-treatments. However, more lignin was removed during the second cycle of steam explosion (22% and 32% for 2C:5 + 3 and 2C:10 + 3, respectively compared with 18% and 23% for similar steam treatments). These findings bear importance for the evaluation of these pre-treatments with a view to integrate a kraft mill into a biorefinery. On the one hand, a smaller amount of lignin in the solid residual after the pre-treatment makes it more susceptible than raw wood to being delignificated (yielding pulps of lower kappa numbers under the same conditions or yielding similar kappa numbers under milder conditions) (Vila et al., 2012). On the other hand, lignin present in the hydrolyzates (containing mainly hemicelluloses) can cause problems during further processing (e.g. enzymatic hydrolysis or fermentation) to obtain value-added products due to lignin precipitation after cooling steps (Gutsch et al., 2012). Fig. 1 also shows that, as expected, the main components extracted in both pre-treatments were pentosans. A similar trend was observed regarding pentosans and holocellulose removal as no cellulose degradation was expected under these conditions (Martín-Sampedro et al., 2011d; Martín-Sampedro et al., 2012a). The percentage of pentosans removed from the original wood was similar with steam explosion and with steam pre-treatments, and increased from 32% to 67% with longer cycles. Therefore, the solubilization of hemicelluloses seems to depend only on the duration of the pre-treatments (severity factor varying from 3.1 to 3.5 according to Overend and Chornet (1987)), but not on the number of cycles nor on the pressure of discharge. When adjusting by the severity factor, several studies report similar outcomes in hemicelluloses extraction during hydrothermal processes. Thus,
239
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244 Table 1 Chemical composition of E. globulus chips before (control) and after steam explosion and steam treatment. Acetone extractives
Hot water extractives
Klason Lignin
Acid Sol. lignin
Holocellulose
Pentosans
Control Steam explosion 1C:5 1C:10 2C:5 + 3 2C:10 + 3
1.4
1.9
21.8
2.4
78.8
16.7
2.8 6.2 7.2 11.9
4.9 8.6 5.3 2.0
20.7 21.4 19.7 18.5
1.9 1.1 1.5 1.0
68.1 60.9 65.6 63.1
11.4 6.5 9.7 6.6
Steam treatment 1C:5 1C:10 2C:5 + 3 2C:10 + 3
3.9 7.3 7.5 12.2
10.6 12.2 10.9 9.3
19.5 19.1 18.4 17.6
1.6 1.2 1.6 1.2
63.9 56.6 57.7 55.1
10.6 7.4 9.1 5.6
(b)
80%
80%
70%
70%
% Removed from solid
% Removed from solid
(a) 60% 50% 40% 30% 20% 10% 0%
60% 50% 40% 30% 20% 10% 0%
1C:5
1C:10 Lignin
2C:5+3
2C:10+3
1C:5
1C:10
Holocellulose
2C:5+3
2C:10+3
Pentosans
Hamzeh et al. (2013) observed a removal of hemicelluloses of 40% and 66% during autohydrolysis of bagasse with severity factors (S0) of 3.1 and 3.5, respectively; Ruiz et al. (2008) reported a complete solubilization of arabinans and about 50% solubilization of xylan after subjecting sunflower stalks to steam explosion with S0 equal to 3.45; and Li et al. (2005) extracted 32% and 75% of hemicelluloses with steam treatment of aspen wood with S0 of 3.3 and 4.2, respectively. The water retention capacity of the solid fractions obtained after the different pre-treatments is showed in Fig. 2. This parameter provides information about the changes in the internal structure of the material and the macro-structural effects of the different pre-treatments on the chips. Based on the increase in water retention capacity of all treated samples compared to control chips, both pre-treatments rendered the chips more accessible, mainly as a result of the extraction of hemicelluloses. However accessibility was greater in chips subjected to steam explosion because the sudden decompression forces fibrous material to ‘‘explode’’ into separated fibers and fiber bundles (Li et al., 2005; Martín-Sampedro et al., 2011a,c). Furthermore, with more severe pre-treatments, water retention capacity of exploded chips increased from 97% to 212%. In contrast, while the steam pre-treatment resulted in greater water retention capacity compared to control pulp, it did not change significantly with the severity of the treatment (75–81% for different severities of steam pre-treatments against 61% for control chips). This result suggests that the removal of 32% of hemicelluloses (achieved with 1C:5 pre-treatments) is enough to increase the accessibility of the material and that further removal of hemicelluloses (up to 67%) caused only small changes in the macro-structure of the material.
g water adsorbed/100 g wood
Fig. 1. Percentanges of lignin, holocellulose and pentosans extracted from the original wood after steam explosion (a) and steam treatment (b).
250 200 150 100 50 0
1C:5 Exp
1C:10 NoExp
2C:5+3
2C:10+3
Control
Fig. 2. Water retention capacity of steam exploded (Exp), steam unexploded (NoExp) and non-treated (control) chips.
3.2. Impact of steam explosion and steam hydrolysis pre-treatments on pulp production Pre-treated and non pre-treated chips were subjected to kraft pulping and the resulting pulps were characterized by determining their kappa number and viscosity (Fig. 3a and b). The overall pulp yield (including yield losses through the pre-treatments and through the cooking process) and consumption of chemicals during kraft pulping were also evaluated (Fig. 3c and d).
3.2.1. Effect on delignification The main effect of the pre-treatments on pulp production was a significant enhancement in delignification. As Fig. 3a shows, kappa numbers of all pre-treated pulps were much lower than those of
240
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244
(b)
40 35 30 25 20 15 10 5 0
1,200 1,000 800 600 400 200 0 1C:5 Exp
1C:10 NoExp
2C:5+3
2C:10+3
1C:5
Control
(c)
Exp
1C:10 NoExp
2C:5+3
2C:10+3
Control
(d) 100%
60%
% NaOH Consumption
50%
Overall Yield
1,400
Viscosity ml/g)
Kappa Number
(a) 45
40% 30% 20% 10% 0% 1C:5 Exp
1C:10 NoExp
2C:5+3
2C:10+3
Control
80% 60% 40% 20% 0% 1C:5 Exp
1C:10 NoExp
2C:5+3
2C:10+3
Control
Fig. 3. Kappa number (a) and viscosity (b) of pulps, overall yield (c) and NaOH consumption (d) during pulping of non pre-treated (control), steam exploded (Exp) and steam unexploded (NoExp) samples.
control pulps, likely due to the modified structure of the pre-treated samples, which is more accessible to cooking reagents, and the breakdown of covalent bonds in lignin-carbohydrates (Kautto et al., 2010; Li et al., 2007). Moreover, the kappa number’s lower values could be explained by the absence of hexenuronic acids (HexA), as the pre-treatment can hydrolyze glucuronic acid groups from the xylan backbone and, therefore, prevent HexA formation (Liu and Amidon, 2007; Mendes et al., 2009). As the kappa numbers indicate, compared to control kraft pulp obtained under the same pulping conditions, the steam explosion pre-treatment caused a 30–68% increase in delignification, depending on the severity of the pre-treatment (S0 between 3.1 and 3.5). Previous reports also describe an increase in delignification with different raw materials subjected to steam explosion before kraft pulping. San Martín et al. (1995) found a 44% increase in delignification when Pinus radiata was steam exploded at 220 °C for 30 s (S0 = 3.2). Jedvert et al. (2012) observed a smaller reduction of the kappa number (4–19%, depending on the duration of the pulping process) with Picea abies, but these researchers used a milder steam explosion pre-treatment (160 °C, 10 min, S0 = 2.8). In a previous work, we used non-wood material Hesperaloe funifera, and found delignification increased 18–28%, depending on the severity of the pre-treatment (S0 between 2.9 and 3.5) (Martín-Sampedro et al., 2012b). However, this delignification increase was lower than that observed with E. globulus under the same steam explosion conditions (Martín-Sampedro et al., 2011a), probably because H. funifera’s readily accessible tissue structure is not rendered more accessible by steam explosion in a way that means any significant improvement in reagent diffusion from baseline and comparable to what steam explosion does to wood. Nevertheless, regardless of which raw material was used, steam exploded material cooked faster and more efficiently, compared to the same untreated material, so that the pulping time could be shortened or the pulping temperature lowered obtaining pulps of similar characteristics
(kappa number and viscosity) while consuming less energy (Ahvazi et al., 2007). Similarly, in a previous work we observed that pulping time could be reduced from 50 to 20 min using a steam explosion pre-treatment similar to 2C:10 + 3 (Martín-Sampedro et al., 2011a). Steam pre-treatments (without explosion) resulted in kappa numbers which were 33–45% lower compared to control pulps obtained under the same pulping conditions. These findings are consistent with results reported in the literature after autohydrolysis of E. globulus. Chirat et al. (2012) reported a 12–60% reduction in kappa numbers depending on the severity of the autohydrolysis pre-treatment (160 °C, 20–50 min, S0 = 3.1–3.5); and Mendes et al. (2009) found a 40% reduction after autohydrolysis at 150 °C for 180 min (S0 = 3.7) when control and pre-treated chips were pulped under the same conditions. Vila et al. (2012) also reported increases in delignification ranging between 23% and 35% even when pre-treated E. globulus chips (pre-treated at 165 °C for 40 min, S0 = 3.5) were pulped for 40 min and control pulps for 50 min. Similarly, using baggase as raw material, Hamzeh et al. (2013) found reductions of up to 26% after an autohydrolysis pre-treatment at 170 °C for 10 min (S0 = 3.1). These delignification enhancements support the use of shorter pulping times when a hydrothermal pre-treatment is applied. Al-Dajani et al. (2009) reported a 70% reduction in pulping time with pre-treated aspen chips (150 °C, 4.5 h, S0 = 3.9); and Kautto et al. (2010) and Saukkonen et al. (2012) reported 30–40% reductions with pre-treated Scot pine (150 °C, 93 min, S0 = 3.4). Comparing both pre-treatments (Fig. 3a), under conditions of low severity (1C:5 pre-treatments), there is a clear similarity between the kappa number of the steam exploded pulp (28.4) and the kappa number of the steam unexploded pulp (27.3). However, under more severe conditions of pre-treatment, kappa numbers for both kinds of pulps differ considerably, with a more pronounced reduction in steam exploded pulps (12.9 vs 22.6 for
241
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244
2C:10 + 3 treatments). These findings can be explained by considering the accessibility of the pulped chips, which determines the efficiency of diffusion of the reagent inside the chips. As mentioned above, the water retention capacity showed that steam exploded samples had a more open structure than steam unexploded chips. Furthermore, while the accessibility of steam treated chips did not significantly changed with the severity of the pre-treatment, steam exploded chips showed a direct increase in water retention capacity from 97%, with 1C:5, to 212%, with 2C:10 + 3. Therefore, based on accessibility and the kappa numbers, steam explosion could be a more efficient pre-treatment for kraft pulping than steam pretreatment, without ‘‘explosion’’. 3.2.2. Effect on pulp viscosity As far as pulp viscosity is concerned (Fig. 3b), no significant differences were observed between pulps obtained from steam explosion or steam pre-treatments. Higher viscosity values (around 1200 ml/g) were found in all pre-treated pulps compared to the control pulp (880 ml/g). Previously, Mendes et al. (2009), Al-Dajani et al. (2009) and Saukkonen et al. (2012), who studied autohydrolysis before kraft pulping of E. globulus, Populus tremuloides and Pinus sylvestris L., respectively, found a slight increase in pulp viscosity compared to the control pulp with similar kappa numbers to the pre-treated pulps. Mendes et al. (2009) attributed the increase in viscosity to the greater dissolution of low molecular weight carbohydrates. However, Martín-Sampedro et al. (2012b) and Sánchez et al. (2011) observed a decrease in viscosity when steam explosion or autohydrolysis, respectively, was applied before subsequent pulping of Hesperaloe funifera. The reason of these findings could be that this non-wood raw material has a more open structure than wood which would enhance the degradation of cellulose under severe hydrothermal pre-treatments. Similarly, Vila et al. (2011, 2012) reported that harsh autohydrolysis conditions could cause a viscosity drop in E. globulus kraft pulps. Therefore, mild hydrothermal pre-treatments, such as those carried out in this study, are recommended (Jedvert et al., 2012). Nevertheless, mild pulping conditions (shorter cooking time or lower cooking temperature) would also contribute to avoid overcooking of pretreated samples, which would imply a reduction in pulp viscosity (Martín-Sampedro et al., 2011a; Vila et al., 2012). 3.2.3. Effect on pulp yield Fig. 3c shows that the overall pulp yields (considering both the pre-treatment and the pulping yield) obtained with pre-treated chips were 7–13% lower than those observed with control chips. These findings were expected and can be attributed to the solubilization of hemicelluloses during the pre-treatments. Furthermore, remaining hemicellulose in the chips after the pre-treatments
might be more susceptible to alkaline peeling and solubilization during pulping, thus causing further yield losses (Chirat et al., 2012; Saukkonen et al., 2012). For these reasons, the screening kraft yield resulted to be lower with pre-treated pulps (45–50%) compared to control pulps (56%). Similar findings have been reported elsewhere with the use of different raw materials and different cooking processes after autohydrolysis (Al-Dajani et al., 2009; Mendes et al., 2009; Vila et al., 2011) and steam explosion pre-treatments (Ahvazi et al., 2007; Martín-Sampedro et al., 2011a). However, Hamzeh et al. (2013) and Martín-Sampedro et al. (2012b) found greater pulping yields with pre-treated nonwood materials (autohydrolyzed baggase and steam exploded Hesperaloe funifera, respectively), compared to un-treated controls. One plausible explanation for this could be the smaller amount of extractives found in pre-treated samples compared to untreated samples (Hamzeh et al., 2013). The higher the pre-treatment severity, the lower the overall yield; with a 43–44% yield (with and without explosion, respectively) when the most severe pre-treatment was used, compared to the control (56%). Vila et al. (2011) also showed a similar maximum reduction in yield after kraft pulping of autohydrolyzed E. globulus chips (40.2% yield compare to 55.1% with their control). Finally, the use of steam exploded versus steam unexploded chips did not result in clear differences in yields after kraft pulping.
3.2.4. Effect on chemical consumption As Fig. 3d shows, all pre-treatments increased the consumption of NaOH during the kraft pulping compared to the control pulp. Similar rates of increase in the consumption of chemicals have been reported by several authors using either autohydrolysis or steam explosion as pre-treatments (Martín-Sampedro et al., 2011a; Mendes et al., 2009; Saukkonen et al., 2012). These findings can be put down to the fact that the amounts of extractives present in the pre-treated pulps were greater. These extractives could consume pulping reagent because, generally, extraneous substances present in the raw materials require soda chemicals for their removal (Hamzeh et al., 2013). Nevertheless, it should be taken into account that this higher chemical consumption produced pulps with lower kappa numbers (Fig. 2a) suggesting a higher efficiency of the chemical in delignification.
3.3. Effect of steam pre-treatments on the morphological properties of fibers The results of the morphological analysis of kraft pulps are summarized in Table 2, which lists the characteristics of fibers, vessels
Table 2 Morphological analysis of E. globulus kraft pulps. Values in brackets indicate standard deviations. Steam explosion
Steam treatment
Control
1C:5
2C:10 + 3
1C: 5
2C:10 + 3
Fibers Length (lm) Width (lm) Coarseness Kinked fibers (%) Average Curl (%)
943.5 (9.2) 19.0 (0.1) 93.5 (0.1) 18.5 (0.4) 5.5 (0.2)
917.0 (4.2) 16.2 (0.2) 74.0 (1.1) 26.1 (0.6) 7.5 (0.2)
875.5 (14.8) 15.0 (0.0) 64.3 (1.8) 33.5 (1.0) 9.4 (0.0)
907.5 (14.8) 15.9 (0.0) 76.5 (2.0) 32.3 (0.3) 8.8 (0.0)
906.0 (8.5) 15.3 (0.1) 65.0 (1.3) 32.4 (0.7) 9.1 (0.0)
Vessels Length (lm) Width (lm)
728.5 (51.6) 136.3 (2.1)
776.0 (41.0) 149.0 (2.4)
722.0 (12.3) 157.2 (6.2)
783.0 (66.5) 161.5 (3.1)
682.5 (33.2) 162.8 (1.6)
56.2 (2.8) 5.2 (0.9) 38.6 (1.8)
52.2 (2.4) 2.7 (1.2) 45.1 (1.3)
54.4 (0.2) 3.8 (0.5) 41.8 (0.6)
50.8 (0.2) 3.5 (0.4) 45.8 (0.5)
Composition (in number of elements) Fibres (%) 53.1 (4.0) Vessels (%) 8.8 (1.3) Fines (%) 38.1 (5.3)
242
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244
(b)
8.0
30.0
6.0
25.0
5.0 4.0 3.0 2.0 1.0 0.0 0 Exp
0
0
NoExp
10 5 0
0.0 1C:10 NoExp - R
2C:5+3
2C:10+3
Control - R
80 70 60 50 40 30 20 10 0
0 Exp
0
0
NoExp
0
1C:5
Control
Exp - R
(f) Tear index (mNm2/g)
Tear index (mNm2/g)
5.0 1C:5
(d)
15
4.0
10.0
Exp - R
20
5.0
15.0
Control
25
(e)
20.0
0
Tensile index (Nm/g)
Tensile index (Nm/g)
(c)
35.0
7.0 Stretch at break (%)
Stretch at break (%)
(a)
3.0 2.0 1.0 0.0
1C:10 NoExp - R
2C:5+3
2C:10+3
Control - R
12 10 8 6 4 2 0
0 Exp
0 NoExp
0
0 Control
1C:5 Exp - R
1C:10 NoExp - R
2C:5+3
2C:10+3
Control - R
Fig. 4. Stretch at break (a and b), tensile index (c and d) and tear index (e and f), for unrefined (Exp, NoExp and Control) and refined pulps (Exp-R, NoExp-R and Control-R).
and fines for some of the steam exploded and steam treated samples. The effect of both treatments on the fiber characteristic is a slight reduction in fiber length and a more noticeable reduction of fiber width and coarseness (the weight of a given length of fibers). All these changes are more striking when the pre-treatments increase in severity, although no significant differences were observed between the fibers obtained from steam explosion or steam pre-treatments. This finding is the result of a better accessibility of the pre-treated chips by cooking chemicals which leads to a more intensive removal of hemicelluloses and lignin and, consequently to narrower fibers. The width decrease and the rigidity increase, caused by the loss of hemicelluloses, clearly affects the internal strength of the fibers and therefore there is an increase in the percentage of kinked fibers increases from the control pulp (18.5%) to the pre-treated kraft pulps (26.1–33.5%). Also, the curl of the fibers increases due to the fiber weakness, from 5.5% (control) to 9.1– 9.4% in the fibers obtained with the most severe pre-treatments. The vessels, the other characteristic element in eucalyptus pulps, have shown minor changes and maintain similar length both in control samples and in pre-treated samples of pulp. Unlike
fibers, vessel width experienced a significant increment with respect to control pulps of up to a 20% (2C:10 + 3 steam unexploded sample). The cause of this radial swelling could be a more effective penetration of chemicals and water in the pre-treated chips. Simultaneously to swelling, the percentage of vessels in the pulp suspension decreases as the severity of the treatment rises. This finding is especially significant for 2C:10 + 3 sample, obtained with the most severe steam explosion conditions, which reduced the percentage of vessels to 2.7%, only 31% of the fine content of the control pulp. Parallelly, the percentage of fines increases in the same way, ranging from 7.0% to 7.7% for the more severe conditions. This suggests that the vessels are de-structured as a result of the action of pulping chemicals, which is more aggressive after the pre-treatment. What is measured as fine material are fragments of these vessels, and that would explain why the percentage of fines becomes larger. Unlike what it does to the percentage of vessels, the introduction of a pretreatment does not change the percentage of fibers significantly. This difference can be put down to the fact that fibers are more massive than vessels and, although reduced in thickness, they would endure the same treatment for longer before their integrity is affected.
243
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244
After the morphological analysis of the pulp suspension, it can be concluded that the reduction of fiber length due to the pretreatments (only a 7% for the worst value) only has a slight impact on the mechanical properties of the paper, while fiber width reduction (up to 21%) and the associated loss of coarseness (up to 31%) means a significant loss of mechanical properties as the pre-treatment renders the fibers brittle and weaker. 3.4. Impact of steam explosion and steam pre-treatments on refining and paper properties Control and pre-treated kraft pulps were refined in a PFI mill at 2000 revolutions increasing the degree of refining (Schopper-Riegler) from 12 (control pulp) and 14 ± 1 (pre-treated pulps) to 21 in all cases. This result indicates that pre-treated pulps required more energy to obtain similar increase in degree of refining compared to un-treated pulps. This is in consistent with previous findings (Helmerius et al., 2010; Kautto et al., 2010; Martín-Sampedro et al., 2011a), since shorter refining times and less energy absorption in refining is a feature typical of kraft pulps having high hemicelluloses content. Hemicelluloses contribute significantly to swelling of fibers, which increases interfiber contact during refining promoting faster external fibrillation and hence better refining response (Kautto et al., 2010). Fig. 4 shows the stretch at break and tensile and tear indexes of unrefined and refined pulps. As expected, refined pulps presented higher paper strength than non-refining pulps. In general, it can be observed that handsheets formed from pre-treated pulps presented lower mechanical properties in comparison with controls. Both the lower content in hemicelluloses and the greater amount of fiber deformations (kinking and curling) in the pre-treated pulps contribute to the loss of properties associated to strength (Saukkonen et al., 2012). This increase in curled and kinked fibers in pretreated pulps has been shown in Table 2 and is a consequence of the weakness of the fibers (less width and coarseness) which
(b)
50 45 40 35 30 25 20 15 10 5 0
Brightness (%ISO)
Brightness (%ISO)
(a)
1C:5 Exp
(c)
would result in a paper with poor mechanical strength. While a slight decrease of paper strength was observed when the duration of the pre-treatment increased, no clear relation with neither the number of cycles, nor the type of hydrothermal pre-treatment (steam explosion or steam treatment), was found. The lower values of stretch at break and tensile index observed for pre-treated pulps compared to the control pulp, are consistent with results reported previously pre-treating with autohydrolysis (Chirat et al., 2012; Hamzeh et al., 2013; Helmerius et al., 2010; Kautto et al., 2010; Saukkonen et al., 2012; Vila et al., 2012) and steam explosion (Martín-Sampedro et al., 2011a; San Martin et al., 1995). These properties are related to the fiber-to-fiber bonding in the fiber net-work, which decreases with the loss of hemicelluloses and with the increase in fiber deformations caused by the pre-treatments. On the other hand, the tear index is related to fiber degradation and some authors have found it to increase or not significantly change towing to hydrothermal pre-treatments (Hamzeh et al., 2013; Helmerius et al., 2010; Kautto et al., 2010). However, other authors found a decrease in tear index (Chirat et al., 2012; Saukkonen et al., 2012; Vila et al., 2012), which would indicate a degradation of the fibers, and it is consistent with the results obtained here. Nevertheless, viscosity values reported by the latter authors did not indicate a degradation of the cellulose chains during the pre-treatment, as it was discussed above. Therefore, the lower tear index observed could be caused by other factors such as fiber morphology. Moreover, the reduction in coarseness is a reflection of fiber damage, and it is consistent with the observed loss of tear resistance in handsheets, a parameter strongly dependent on individual fiber strength. Contrarily to mechanical properties, optical properties of papersheets improved when a hydrothermal pre-treatment was carried out. Thus, higher brightness and lower yellowness index were observed in pre-treated pulps compared to the control pulp (Fig. 5). This improvement in optical properties is mainly caused by the
1C:10 NoExp
2C:5+3
1C:5
2C:10+3
Control
Exp - R
(d)
40 30
Yellowness Index
Yellowness Index
35 25 20 15 10 5 0 1C:5 Exp
1C:10 NoExp
45 40 35 30 25 20 15 10 5 0
2C:5+3
2C:10+3
Control
1C:10 NoExp - R
2C:5+3
2C:10+3
Control - R
45 40 35 30 25 20 15 10 5 0 1C:5 Exp - R
1C:10 NoExp - R
2C:5+3
2C:10+3
Control - R
Fig. 5. Brightness (a and b) and yellowness index (c and d) for unrefined (Exp, NoExp and control) and refined pulps (Exp-R, NoExp-R and Control-R).
244
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244
lower lignin content (kappa number) and hexenuronic acid content of the pre-treated samples. Therefore, exploded samples, which experienced a clear decrease in kappa number when the severity of the pre-treatment increased (Fig. 2a), provided handsheets with higher brightness and lower yellowness index with more severe pre-treatments. Several authors have also reported better optical properties when pre-treating pulps with autohydrolysis (Hamzeh et al., 2013; Kautto et al., 2010; Vila et al., 2011) and steam explosion (Jedvert et al., 2012; Martín-Sampedro et al., 2011a,b; San Martin et al., 1995) compared to the control pulp. Furthermore, Chirat et al. (2012), Hamzeh et al. (2013), Martín-Sampedro et al. (2011b) and Vila et al. (2011) observed that hydrothermally pretreated pulps showed better bleaching ability than non pre-treated pulps, which means better optical properties consuming fewer chemicals during bleaching. Finally, refining caused a decrease in brightness and an increase in yellowness index. Eugenio et al. (2011), who analyzed unrefined and refined E. globulus pulps by thioacidolysis-SEC, attributed this darkening of the pulps to the formation of new chromophores during refining. 4. Conclusion Steam explosion and steam pre-treatments are appropriate not only to extract hemicelluloses but also to enhance delignification. Nevertheless, steam explosion seems to be more efficient to this end, as it achieved the biggest reduction of the kappa number (68% vs 45% with the steam pre-treatment). Therefore, it is reasonable to assume that these pre-treatments would reduce the pulping time and consequently would increase the pulp production, although consuming more wood. The paper obtained could be used in applications in which good optical properties, rather than good mechanical properties, are required, such as manufacturing tissue paper or dissolving pulps. Acknowledgements The authors wish to thank the Spanish MINECO for funding this study via Project CTQ 2011-28503-C02-01 and Program PTA 20114857-I, and the Community of Madrid for funding via Project PROLIPAPEL–CM (S2009/AMB-1480). Special thanks to Concepción Monte (Complutense University), for her valuable collaboration in the morphological characterization of fibers. References Ahvazi, B., Radiotis, T., Bouchard, J., Goel, K., 2007. Chemical pulping of steamexploded mixed hardwood chips. J. Wood Chem. Technol. 27 (2), 49–63. Al-Dajani, W.W., Tschirner, U.W., Jensen, T., 2009. Pre-extraction of hemicelluloses and subsequent kraft pulping Part II: acid- and autohydrolysis. Tappi J. 8 (9), 30–37. Chandra, R., Bura, R., Mabee, W., Berlin, A., Pan, X., 2007. Substrate pretreatment: the key to effective enzymatic hydrolysis of lignocellulosics? Adv. Biochem. Eng./Biotechnol. 108, 67–93. Chirat, C., Lachenal, D., Sanglard, M., 2012. Extraction of xylans from hardwood chips prior to kraft cooking. Process Biochem. 47 (3), 381–385. Eugenio, M.E., Du, X., Li, J. 2011. Towards improvements of kraft pulp bleaching by additional treatments 16th ISWFPC.Tiajin, China. Feria, M., García, J., Díaz, M., Fernández, M., López, F., 2012. Biorefinery process for production of paper and oligomers from Leucaena leucocephala K360 with or without prior autohydrolysis. Bioresour. Technol. 126, 64–70. Garrote, G., Dominguez, H., Parajó, J.C., 1999. Mild autohydrolysis: an environmentally friendly technology for xylooligosaccharide production from wood. J. Chem. Technol. Biotechnol. 74 (11), 1101–1109. Gutsch, J.S., Nousiainen, T., Sixta, H., 2012. Comparative evaluation of autohydrolysis and acid-catalyzed hydrolysis of Eucalyptus globulus wood. Bioresour. Technol. 109, 77–85.
Hamzeh, Y., Ashori, A., Khorasani, Z., Abdulkhani, A., Abyaz, A., 2013. Pre-extraction of hemicelluloses from bagasse fibers: effects of dry-strength additives on paper properties. Ind. Crop Prod. 43, 365–371. Helmerius, J., von Walter, J.V., Rova, U., Berglund, K.A., Hodge, D.B., 2010. Impact of hemicellulose pre-extraction for bioconversion on birch kraft pulp properties. Bioresour. Technol. 101 (15), 5996–6005. Jarabo, R., Monte, M.C., Blanco, A., Negro, C., Tijero, J., 2012. Characterisation of agricultural residues used as a source of fibres for fibre-cement production. Ind. Crop Prod. 36 (1), 14–21. Jedvert, K., Wang, Y., Saltberg, A., Henriksson, G., Lindstrom, M.E., Theliander, H., 2012. Mild steam explosion: a way to activate wood for enzymatic treatment, chemical pulping and biorefinery processes. Nord. Pulp Pap. Res. J. 27 (5), 828– 835. Kautto, J., Saukkonen, E., Henricson, K., 2010. Digestibility and paper-making properties of prehydrolyzed softwood chips. Bioresources 5 (4), 2502–2519. Leschinsky, M., Sixta, H., Patt, R., 2009. Detailed mass balances of the autohydrolysis of Eucalyptus globulus at 170 °C. Bioresources 4 (2), 687–703. Li, H., Saeed, A., Jahan, M.S., Ni, Y., Van Heiningen, A., 2010. Hemicellulose removal from hardwood chips in the pre-hydrolysis step of the kraft-based dissolving pulp production process. J. Wood Chem. Technol. 30 (1), 48–60. Li, J., Henriksson, G., Gellerstedt, G., 2005. Carbohydrate reactions during hightemperature steam treatment of aspen wood. Appl. Biochem. Biotechnol. 125 (3), 175–188. Li, J., Henriksson, G., Gellerstedt, G., 2007. Lignin depolymerization/ repolymerization and its critical role for delignification of aspen wood by steam explosion. Bioresour. Technol. 98 (16), 3061–3068. Liu, S., Amidon, T.E., 2007. Essential components of a wood based biorefinery. O Papel. 68 (3), 84–105. Martín-Sampedro, R., Eugenio, M.E., Revilla, E., Martín, J.A., Villar, J.C., 2011a. Integration of kraft pulping on a biorefinery by the addition of a steam explosion pretreatment. Bioresources 6 (1), 513–528. Martín-Sampedro, R., Eugenio, M., Villar, J., 2011b. Biobleaching of Eucalyptus globulus kraft pulps: comparison between pulps obtained from exploded and non-exploded chips. Bioresour. Technol. 102, 4530–4535. Martín-Sampedro, R., Martin, J.A., Eugenio, M.E., Revilla, E., Villar, J.C., 2011c. Steam explosion treatment of Eucalyptus globulus wood: influence of operational conditions on chemical and structural modifications. Bioresources 6 (4), 4922– 4935. Martín-Sampedro, R., Capanema, E.A., Hoeger, I., Villar, J.C., Rojas, O.J., 2011d. Lignin changes after steam explosion and laccase-mediator treatment of eucalyptus wood chips. J. Agric. Food Chem. 59 (16), 8761–8769. Martín-Sampedro, R., Eugenio, M.E., García, J.C., Lopez, F., Villar, J.C., Diaz, M.J., 2012a. Steam explosion and enzymatic pre-treatments as an approach to improve the enzymatic hydrolysis of Eucalyptus globulus. Biomass Bioenerg. 42, 97–106. Martín-Sampedro, R., Eugenio, M.E., Villar, J.C., 2012b. Effect of steam explosion and enzymatic pre-treatments on pulping and bleaching of Hesperaloe funifera. Bioresour. Technol. 111, 460–467. Mendes, C.V.T., Carvalho, M., Baptista, C., Rocha, J.M.S., Soares, B.I.G., Sousa, G.D.A., 2009. Valorisation of hardwood hemicelluloses in the kraft pulping process by using an integrated biorefinery concept. Food Bioprod. Process. 87 (C3), 197– 207. Overend, R., Chornet, E., 1987. Fractionation of lignocellulosics by steam-aqueous pretreatments. Philos. Trans. R. Soc. London, A 321 (1561), 523–536. Rahikainen, J.L., Martín-Sampedro, R., Heikkinen, H., Rovio, S., Marjamaa, K., Tamminen, T., Rojas, O.J., Kruus, K., 2013. Inhibitory effect of lignin during cellulose bioconversion: the effect of lignin chemistry on non-productive enzyme adsorption. Bioresour. Technol. 133, 270–278. Ruiz, E., Cara, C., Manzanares, P., Ballesteros, M., Castro, E., 2008. Evaluation of steam explosion pre-treatment for enzymatic hydrolysis of sunflower stalks. Enzyme Microb. Technol. 42 (2), 160–166. San Martin, R., Perez, C., Briones, R., 1995. Simultaneous production of ethanol and kraft pulp from pine (Pinus radiata) using steam explosion. Bioresour. Technol. 53 (3), 217–223. Sánchez, R., Rodríguez, A., García, J., Rosal, A., Jiménez, L., 2011. Exploitation of hemicellulose, cellulose and lignin from Hesperaloe funifera. Bioresour. Technol. 102, 1308–1315. Saukkonen, E., Kautto, J., Rauvanto, I., Backfolk, K., 2012. Characteristics of prehydrolysis-kraft pulp fibers from Scots pine. Holzforschung 66, 801–808. Tuck, C.O., Pérez, E., Horváth, I.T., Sheldon, R.A., Poliakoff, M., 2012. Valorization of biomass: deriving more value from waste. Science. 337 (6095), 695–699. Vila, C., Romero, J., Francisco, J.L., Garrote, G., Parajo, J.C., 2011. Extracting value from Eucalyptus wood before kraft pulping: effects of hemicelluloses solubilization on pulp properties. Bioresour. Technol. 102 (8), 5251–5254. Vila, C., Romero, J., Francisco, J.L., Santos, V., Parajó, J.C., 2012. On the recovery of hemicellulose before kraft pulping. Bioresources 7 (3), 4179–4189. Wise, L.E., Murphy, M., d’Addieco, A., 1946. Chlorite holocellulose, its fractionnation and bearing on summative wood analysis and on studies on the hemicelluloses. Paper Trade J. 122 (2), 35–43.
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Integration of a kraft pulping mill into a forest biorefinery: Pre-extraction of hemicellulose by steam explosion versus steam treatment Raquel Martin-Sampedro ⇑, Maria E. Eugenio, Jassir A. Moreno, Esteban Revilla, Juan C. Villar Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria INIA, Carretera de la Coruña, km 7.5, 28040 Madrid, Spain
h i g h l i g h t s Steam explosion and steam pre-treatments were carried out before kraft pulping. The rate of extraction of hemicelluloses was similar with both pre-treatments. Steam explosion enhanced delignification more efficiently. Pre-treatments reduced paper mechanical properties owing to the fiber morphology. Paper optical properties were boosted when pre-treatments were carried out.
a r t i c l e
i n f o
Article history: Received 11 October 2013 Received in revised form 25 November 2013 Accepted 30 November 2013 Available online 8 December 2013 Keywords: Autohydrolysis Biorefineries Kraft pulping Steam explosion Eucalyptus globulus
a b s t r a c t Growing interest in alternative and renewable energy sources has brought increasing attention to the integration of a pulp mill into a forest biorefinery, where other products could be produced in addition to pulp. To achieve this goal, hemicelluloses were extracted, either by steam explosion or by steam treatment, from Eucalyptus globulus wood prior to pulping. The effects of both pre-treatments in the subsequent kraft pulping and paper strength were evaluated. Results showed a similar degree of hemicelluloses extraction with both options (32–67% of pentosans), which increased with the severity of the conditions applied. Although both pre-treatments increased delignification during pulping, steam explosion was significantly better: 12.9 kappa number vs 22.6 for similar steam unexploded pulps and 40.7 for control pulp. Finally, similar reductions in paper strength were found regardless of the type of treatment and conditions assayed, which is attributed to the increase of curled and kinked fibers. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Integration of a lignocellulose biorefinery into a pulp mill is a feasible possibility to add value to the production of pulp, a product considered as a commodity with a moderate added value. The integrated biorefinery-pulp mill would bring forth products of higher added-value, such as furfural, carbon fibers, biopolymers or biofuels, in addition to pulp (Helmerius et al., 2010). One strategy to accomplish this integration is to use a fraction of hemicelluloses which, in a conventional kraft cooking, are dissolved into the cooking liquor and burned in the recovery boiler jointly with the lignin degradation products. While, combustion of the non-cellulosic compounds provides the necessary supply of energy to maintain the pulp mill or even to have a surplus, removal of hemicelluloses prior to cooking does not necessarily
⇑ Corresponding author. Tel.: +34 913476834; fax: +34 913476767. E-mail address: [email protected] (R. Martin-Sampedro). 0960-8524/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.biortech.2013.11.088
affect the final energy balance because their heating value is approximately half that of lignin, and cooking pre-extracted chips requires considerably less energy than cooking normal chips, as the cooking time of the former is significantly shorter (Martín-Sampedro et al., 2011a). The removed hemicelluloses could be hydrolyzed into sugars and then fermented to ethanol or derived to other value-added products. These compounds are considered promising platform chemicals in the synthesis of many other valuable products (Tuck et al., 2012). Hemicelluloses extraction has been done by a variety of methods such as autohydrolysis, steam explosion, acid hydrolysis or alkali extraction. However, steam explosion and autohydrolysis pre-treatments only use water/steam at high temperature causing the formation of acetic acid from the acetylated wood component, which catalyzes hydrolytic reactions in the wood polymers. Thus, these hydrothermal pre-treatments can be considered as a green and competitive technology to remove hemicelluloses in hardwood, since reaction media contain only lignocellulosic feedstock
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244
and water, which prevents corrosion problems and the formation of neutralization sludges (Garrote et al., 1999; Martín-Sampedro et al., 2011a). The effect of these hydrothermal pre-treatments on a subsequent pulping process has been studied using a diversity of raw materials. Autohydrolysis has been applied as a pre-treatment on kraft, soda or ethanol–water pulping using pine (Kautto et al., 2010; Saukkonen et al., 2012), a mixture of maple, poplar and birch (Li et al., 2010), eucalyptus (Chirat et al., 2012; Mendes et al., 2009; Vila et al., 2011, 2012), aspen (Al-Dajani et al., 2009), birch (Helmerius et al., 2010), bagasse (Hamzeh et al., 2013) and Leucaena leucocephala K360 (Feria et al., 2012). Although in less extent than autohydrolysis, steam explosion pre-treatment has also been studied using different raw materials such as pine (San Martin et al., 1995), Norway spruce (Jedvert et al., 2012), a mixture of marple, birch and aspen (Ahvazi et al., 2007), Hesperaloe funifera (Martín-Sampedro et al., 2012b) and eucalyptus (Martín-Sampedro et al., 2011a,b). With both pre-treatments, hemicelluloses removal results in an improvement of the delignification rate, which reduces the cooking time and/or the chemical charge required for an equal target of kappa number, and enhances the subsequent bleaching process. In some cases, it has also been observed a decrease in pulp viscosity, due to the reduction of the cellulose polymerization degree and a general loss of mechanical properties. Nevertheless, the loss of pulp mechanical properties is moderate when the xylan removal is limited (Chirat et al., 2012), and the pulp viscosity could be maintained if preextracted chips are cooked at milder conditions than nonextracted wood (Vila et al., 2012). Finally, the decrease in pulp yield observed could be explained by the hemicellulose removal and an increase of the peeling reactions of a more exposed cellulose (Chirat et al., 2012). As all hydrothermal pre-treatments (stem treatment, hot water treatment, steam explosion, etc.) lead to a similar release of acids from acetylated wood components which catalyze hydrolytic reactions in the wood polymers (autohydrolysis), the main difference between steam treatment (sometimes called autohydrolysis treatment) and steam explosion is the rapid decompression that takes place at the end of steam explosion but not in steam treatment. This decompression forces the fibrous material to ‘‘explode’’ into separated fibers and fiber bundles, generating a solid fraction with a more open structure (Ahvazi et al., 2007; Martín-Sampedro et al., 2011a,c) that may enhance the efficient diffusion of cooking liquor into the fibers. Although both pre-treatments have been studied in different articles, as mentioned above, a comparison between them has not been previously reported. Therefore, the main objective of this study is to compare the effect of steam explosion and steam treatment on the subsequent kraft pulping of Eucalyptus globulus, in order to elucidate the separate effect of the steam treatment (causing autohydrolysis in both pre-treatments) and of the ‘‘explosion’’. Furthermore, the influence of the severity factor on the subsequent pulping process and pulp quality was also studied with both pre-treatments.
237
2.2. Steam explosion and steam treatments Steam explosion and steam treatments were performed in a 26 L stainless steel digester (manufactured by Cadepla S.L.) capable of temperatures to 190 °C and a pressure of 1.37 MPa (14 kgf cm2). The digester was connected to a blowing tank into which chips were discharged at the end of the treatment. It was also equipped with electrovalves for steam admission, and a ball valve of discharge. The steam generator was a Babcock Wanson VAP 250RR boiler, with a maximum steam production rate of 270 kg h1 and a working pressure of 1.37 MPa. According to previous reports (Martín-Sampedro et al., 2011a; Martín-Sampedro et al., 2011c), chips were immersed in water at 25 °C for 16 h in order to improve the efficiency of the subsequent steam pre-treatments. In all of the experiments, 500 g of E. globulus chips were treated with steam at 183 °C (10 kg-f cm2). The variable operational conditions were: number of cycles of treatment (one or two), duration of the first cycle (5 or 10 min), and discharge pressure (6 kg-f cm2 for steam explosion treatments or atmospheric pressure for steam treatments). When a second cycle was carried out, the pre-treated chips obtained in the first cycle were washed with cold water and then subjected to a second cycle of 3 min following the same procedure as in the first cycle. After treatment, the samples were thoroughly washed with water, dried at room temperature and stored in sealed polyethylene bags. The severity factor of each treatment was calculated according to the following equation (Eq. (1)) defined by Overend and Chornet (1987).
T100 S0 ¼ log e 14:75 t
ð1Þ
in which T is the temperature (°C) and t the duration of the treatment (min). Water retention, or hydration, capacity of the treated and untreated chips was defined as the weight of water absorbed by the chips after being immersed in water at 25 °C for 6 h. It was expressed as grams of water per 100 g of oven dry wood, according to the following equation:
WR ¼
Ww Wd 100 Wd
ð2Þ
where Ww was the weight of the wood chips after the water immersion; then, wood chips were dried at 104 °C during 24 h, and weighted again (dry weight, Wd). 2.3. Chemical analysis In order to carry out the chemical analysis, all the samples were dried at room temperature and then milled in a Wiley mill. The samples were sieved using standard sieves to obtain 20 g of wood meal sized between 0.30 and 0.40 mm. Acetone extractives (UNE-EN ISO 14453), hot water extractives (UNE 57-013-82), lignin content (TAPPI T 222 om-88), holocellulose content (Wise et al., 1946) and pentonsans content (TAPPI T 223 cm-84) were then measured in the wood meal. All determinations were duplicated.
2. Experimental
2.4. Kraft cooking process
2.1. Raw material
Kraft cooking was performed over control (non-treated), steam exploded and steam unexploded chips in pressurized 1-liter reactors. Four reactors were placed in a 20 L rotatory pressurized vessel that contained hot water for indirect heating of the reactors. The rotatory vessel had a jacket-type electrical heater controlled by a computer to set the cooking temperature. Cooking conditions were: 150 g of dry chips, 4 L/kg liquor to wood ratio, 16% active
E. globulus chips were kindly provided by La Montañanesa pulp mill (Torraspapel – Lecta Group, Spain). The material was air dried and then homogenized in a single stock (by conditioning inside polyethylene bags) to avoid differences in composition and water content. The chips were stored in polyethylene bags at 25 °C.
238
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244
alkali, 20% sulfidity, 160 °C cooking temperature, 60 min to maximum temperature and 30 min at maximum temperature. After processing each batch, the reactors were removed from the vessel, and the cooked chips were washed, disintegrated and screened to determine their reject content and screened kraft yields (based on the solid content of the material submitted to kraft pulping). Moreover, the overall yield was calculated taking into account both the steam explosion or steam pre-treatments and the kraft process. Black liquor was titrated with HCl to determine the chemical reagent consumption. The kappa number (TAPPI T-236/UNE 57034) and viscosity (UNE 57039-1) were determined in the pulp samples. 2.5. Morphological characterization of kraft fibers In order to elucidate the effect of the steam pre-treatments on fiber morphology, kraft fibers were analyzed with the morphological analyser Morfi V7.9.13.E (Techpap, France). The procedure was based on that described by Jarabo et al. (2012) and adapted for eucalyptus kraft pulps. Previously, the samples were prepared for morphological characterization by disintegrating 1 g of dry pulp into 600 ml of water in an ENJO-model 692 lab disintegrator. The characterization was done in duplicate. The Morfi fiber morphological characterization is based on an image analysis system, consisting of a diode which emits unpolarized light. The image acquisition is maintained until 5000 elements are analyzed. Some parameters are required to discriminate between the elements of interest (fibers, vessels and fines) and the others. These parameters were: fiber length, 100–6000 lm; fiber width, 5–75 lm; maximum fiber kink, 178°; maximum fiber curl, 98%; maximum vessels length/ width ratio, 1.5; and fines length, 1–100 lm. Fiber kink is defined as an abrupt change in the direction of the fiber and it is expressed in degrees. The curl is determined from the ratio between the arc (distance between the two ends of the fiber) and the developed length of the fiber, and it is given by the classical definition shown in Eq. (1), which for a completely straight fiber (arc equals to developed length) gives a 0% curl.
C ð%Þ ¼ 100 1
arc Developed length
ð3Þ
The characterization of the fibers was made in terms of fiber length, fiber width, coarseness, percentage of kinked fibers and angle or curl. Vessels were characterized by their length and width. The percentages of fibers, vessels and fines in the pulp suspension were also determined. 2.6. Refining and handsheets characterization In consistency with results reported elsewhere (Martín-Sampedro et al., 2011a), pulps were refined in a PFI mill (UNE-EN ISO 5264-2/TAPPI T-248) at 2000 revolutions. The refining degree (Schopper-Riegler) (UNE-EN ISO 5267-1) of the pulps was determined before and after refining. Handsheets were then formed from unrefined and refined pulps, in accordance with ISO UNE-EN ISO 5269-2, and characterized in terms of stretch at break, tensile and tear indexes (UNE-EN ISO 5270), brightness (UNE 57062), and yellowness index (TAPPI T-1216). 3. Results and discussion 3.1. Impact of steam explosion and steam pre-treatments on the resulting solid fraction Table 1 shows the chemical composition of the chips before (control) and after the different steam explosion or steam
pre-treatments. It can be observed that there was an increase in the content of extractives (acetone and hot water soluble material) in the wood chips after the different pre-treatments (steam explosion and steam treatment) compared to the control. These findings could be owing to the fact that during the pre-treatments some macromolecules are partially degraded becoming extractible in a subsequent solvent extraction. Similar findings have been reported by other authors (Chandra et al., 2007; Martín-Sampedro et al., 2011a; Rahikainen et al., 2013). Although the pre-treatments result in an increase of both acetone and hot water extractives, a different trend was observed under our operational conditions. Namely, the content of acetone extractives increased with the severity of the pre-treatment (duration and number of cycles) regardless of the pressure of discharge (6 kg-f cm2 and atmospheric pressure for steam explosion and steam treatment, respectively). However, the amount of extractives in hot water was greater in the steam unexploded samples compared to the steam exploded samples, likely owing to the removal of some hydrosoluble material during the rapid expansion of steam inside the chips. On the other hand, a decrease in hot water extractives content was observed when a second cycle was carried out. The reason could be the partial degradation of polysaccharides (mainly hemicellulose), which renders them readily extractible in the subsequent steam treatment (second cycle). Furthermore, the reduction in water extractives after the second cycle was more pronounced when using steam explosion as pre-treatment likely as a result of the first explosion (first cycle), which opens the wood internal structure and facilitates subsequent extraction during the second cycle. Percentages of lignin, holocellulose and pentosans removed from the original wood were calculated for each pre-treatment (Fig. 1), taking into account the chemical composition of the samples (Table 1) and the yield of the pre-treatments. Although steam explosion and steam treatments are known as selective methods for the removal of hemicelluloses from the wood matrix, a significant amount of lignin is colloidally dissolved during these pretreatments (Gutsch et al., 2012; Leschinsky et al., 2009). The amounts of lignin removed with one cycle treatments were similar with steam explosion and steam pre-treatments. However, more lignin was removed during the second cycle of steam explosion (22% and 32% for 2C:5 + 3 and 2C:10 + 3, respectively compared with 18% and 23% for similar steam treatments). These findings bear importance for the evaluation of these pre-treatments with a view to integrate a kraft mill into a biorefinery. On the one hand, a smaller amount of lignin in the solid residual after the pre-treatment makes it more susceptible than raw wood to being delignificated (yielding pulps of lower kappa numbers under the same conditions or yielding similar kappa numbers under milder conditions) (Vila et al., 2012). On the other hand, lignin present in the hydrolyzates (containing mainly hemicelluloses) can cause problems during further processing (e.g. enzymatic hydrolysis or fermentation) to obtain value-added products due to lignin precipitation after cooling steps (Gutsch et al., 2012). Fig. 1 also shows that, as expected, the main components extracted in both pre-treatments were pentosans. A similar trend was observed regarding pentosans and holocellulose removal as no cellulose degradation was expected under these conditions (Martín-Sampedro et al., 2011d; Martín-Sampedro et al., 2012a). The percentage of pentosans removed from the original wood was similar with steam explosion and with steam pre-treatments, and increased from 32% to 67% with longer cycles. Therefore, the solubilization of hemicelluloses seems to depend only on the duration of the pre-treatments (severity factor varying from 3.1 to 3.5 according to Overend and Chornet (1987)), but not on the number of cycles nor on the pressure of discharge. When adjusting by the severity factor, several studies report similar outcomes in hemicelluloses extraction during hydrothermal processes. Thus,
239
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244 Table 1 Chemical composition of E. globulus chips before (control) and after steam explosion and steam treatment. Acetone extractives
Hot water extractives
Klason Lignin
Acid Sol. lignin
Holocellulose
Pentosans
Control Steam explosion 1C:5 1C:10 2C:5 + 3 2C:10 + 3
1.4
1.9
21.8
2.4
78.8
16.7
2.8 6.2 7.2 11.9
4.9 8.6 5.3 2.0
20.7 21.4 19.7 18.5
1.9 1.1 1.5 1.0
68.1 60.9 65.6 63.1
11.4 6.5 9.7 6.6
Steam treatment 1C:5 1C:10 2C:5 + 3 2C:10 + 3
3.9 7.3 7.5 12.2
10.6 12.2 10.9 9.3
19.5 19.1 18.4 17.6
1.6 1.2 1.6 1.2
63.9 56.6 57.7 55.1
10.6 7.4 9.1 5.6
(b)
80%
80%
70%
70%
% Removed from solid
% Removed from solid
(a) 60% 50% 40% 30% 20% 10% 0%
60% 50% 40% 30% 20% 10% 0%
1C:5
1C:10 Lignin
2C:5+3
2C:10+3
1C:5
1C:10
Holocellulose
2C:5+3
2C:10+3
Pentosans
Hamzeh et al. (2013) observed a removal of hemicelluloses of 40% and 66% during autohydrolysis of bagasse with severity factors (S0) of 3.1 and 3.5, respectively; Ruiz et al. (2008) reported a complete solubilization of arabinans and about 50% solubilization of xylan after subjecting sunflower stalks to steam explosion with S0 equal to 3.45; and Li et al. (2005) extracted 32% and 75% of hemicelluloses with steam treatment of aspen wood with S0 of 3.3 and 4.2, respectively. The water retention capacity of the solid fractions obtained after the different pre-treatments is showed in Fig. 2. This parameter provides information about the changes in the internal structure of the material and the macro-structural effects of the different pre-treatments on the chips. Based on the increase in water retention capacity of all treated samples compared to control chips, both pre-treatments rendered the chips more accessible, mainly as a result of the extraction of hemicelluloses. However accessibility was greater in chips subjected to steam explosion because the sudden decompression forces fibrous material to ‘‘explode’’ into separated fibers and fiber bundles (Li et al., 2005; Martín-Sampedro et al., 2011a,c). Furthermore, with more severe pre-treatments, water retention capacity of exploded chips increased from 97% to 212%. In contrast, while the steam pre-treatment resulted in greater water retention capacity compared to control pulp, it did not change significantly with the severity of the treatment (75–81% for different severities of steam pre-treatments against 61% for control chips). This result suggests that the removal of 32% of hemicelluloses (achieved with 1C:5 pre-treatments) is enough to increase the accessibility of the material and that further removal of hemicelluloses (up to 67%) caused only small changes in the macro-structure of the material.
g water adsorbed/100 g wood
Fig. 1. Percentanges of lignin, holocellulose and pentosans extracted from the original wood after steam explosion (a) and steam treatment (b).
250 200 150 100 50 0
1C:5 Exp
1C:10 NoExp
2C:5+3
2C:10+3
Control
Fig. 2. Water retention capacity of steam exploded (Exp), steam unexploded (NoExp) and non-treated (control) chips.
3.2. Impact of steam explosion and steam hydrolysis pre-treatments on pulp production Pre-treated and non pre-treated chips were subjected to kraft pulping and the resulting pulps were characterized by determining their kappa number and viscosity (Fig. 3a and b). The overall pulp yield (including yield losses through the pre-treatments and through the cooking process) and consumption of chemicals during kraft pulping were also evaluated (Fig. 3c and d).
3.2.1. Effect on delignification The main effect of the pre-treatments on pulp production was a significant enhancement in delignification. As Fig. 3a shows, kappa numbers of all pre-treated pulps were much lower than those of
240
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244
(b)
40 35 30 25 20 15 10 5 0
1,200 1,000 800 600 400 200 0 1C:5 Exp
1C:10 NoExp
2C:5+3
2C:10+3
1C:5
Control
(c)
Exp
1C:10 NoExp
2C:5+3
2C:10+3
Control
(d) 100%
60%
% NaOH Consumption
50%
Overall Yield
1,400
Viscosity ml/g)
Kappa Number
(a) 45
40% 30% 20% 10% 0% 1C:5 Exp
1C:10 NoExp
2C:5+3
2C:10+3
Control
80% 60% 40% 20% 0% 1C:5 Exp
1C:10 NoExp
2C:5+3
2C:10+3
Control
Fig. 3. Kappa number (a) and viscosity (b) of pulps, overall yield (c) and NaOH consumption (d) during pulping of non pre-treated (control), steam exploded (Exp) and steam unexploded (NoExp) samples.
control pulps, likely due to the modified structure of the pre-treated samples, which is more accessible to cooking reagents, and the breakdown of covalent bonds in lignin-carbohydrates (Kautto et al., 2010; Li et al., 2007). Moreover, the kappa number’s lower values could be explained by the absence of hexenuronic acids (HexA), as the pre-treatment can hydrolyze glucuronic acid groups from the xylan backbone and, therefore, prevent HexA formation (Liu and Amidon, 2007; Mendes et al., 2009). As the kappa numbers indicate, compared to control kraft pulp obtained under the same pulping conditions, the steam explosion pre-treatment caused a 30–68% increase in delignification, depending on the severity of the pre-treatment (S0 between 3.1 and 3.5). Previous reports also describe an increase in delignification with different raw materials subjected to steam explosion before kraft pulping. San Martín et al. (1995) found a 44% increase in delignification when Pinus radiata was steam exploded at 220 °C for 30 s (S0 = 3.2). Jedvert et al. (2012) observed a smaller reduction of the kappa number (4–19%, depending on the duration of the pulping process) with Picea abies, but these researchers used a milder steam explosion pre-treatment (160 °C, 10 min, S0 = 2.8). In a previous work, we used non-wood material Hesperaloe funifera, and found delignification increased 18–28%, depending on the severity of the pre-treatment (S0 between 2.9 and 3.5) (Martín-Sampedro et al., 2012b). However, this delignification increase was lower than that observed with E. globulus under the same steam explosion conditions (Martín-Sampedro et al., 2011a), probably because H. funifera’s readily accessible tissue structure is not rendered more accessible by steam explosion in a way that means any significant improvement in reagent diffusion from baseline and comparable to what steam explosion does to wood. Nevertheless, regardless of which raw material was used, steam exploded material cooked faster and more efficiently, compared to the same untreated material, so that the pulping time could be shortened or the pulping temperature lowered obtaining pulps of similar characteristics
(kappa number and viscosity) while consuming less energy (Ahvazi et al., 2007). Similarly, in a previous work we observed that pulping time could be reduced from 50 to 20 min using a steam explosion pre-treatment similar to 2C:10 + 3 (Martín-Sampedro et al., 2011a). Steam pre-treatments (without explosion) resulted in kappa numbers which were 33–45% lower compared to control pulps obtained under the same pulping conditions. These findings are consistent with results reported in the literature after autohydrolysis of E. globulus. Chirat et al. (2012) reported a 12–60% reduction in kappa numbers depending on the severity of the autohydrolysis pre-treatment (160 °C, 20–50 min, S0 = 3.1–3.5); and Mendes et al. (2009) found a 40% reduction after autohydrolysis at 150 °C for 180 min (S0 = 3.7) when control and pre-treated chips were pulped under the same conditions. Vila et al. (2012) also reported increases in delignification ranging between 23% and 35% even when pre-treated E. globulus chips (pre-treated at 165 °C for 40 min, S0 = 3.5) were pulped for 40 min and control pulps for 50 min. Similarly, using baggase as raw material, Hamzeh et al. (2013) found reductions of up to 26% after an autohydrolysis pre-treatment at 170 °C for 10 min (S0 = 3.1). These delignification enhancements support the use of shorter pulping times when a hydrothermal pre-treatment is applied. Al-Dajani et al. (2009) reported a 70% reduction in pulping time with pre-treated aspen chips (150 °C, 4.5 h, S0 = 3.9); and Kautto et al. (2010) and Saukkonen et al. (2012) reported 30–40% reductions with pre-treated Scot pine (150 °C, 93 min, S0 = 3.4). Comparing both pre-treatments (Fig. 3a), under conditions of low severity (1C:5 pre-treatments), there is a clear similarity between the kappa number of the steam exploded pulp (28.4) and the kappa number of the steam unexploded pulp (27.3). However, under more severe conditions of pre-treatment, kappa numbers for both kinds of pulps differ considerably, with a more pronounced reduction in steam exploded pulps (12.9 vs 22.6 for
241
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244
2C:10 + 3 treatments). These findings can be explained by considering the accessibility of the pulped chips, which determines the efficiency of diffusion of the reagent inside the chips. As mentioned above, the water retention capacity showed that steam exploded samples had a more open structure than steam unexploded chips. Furthermore, while the accessibility of steam treated chips did not significantly changed with the severity of the pre-treatment, steam exploded chips showed a direct increase in water retention capacity from 97%, with 1C:5, to 212%, with 2C:10 + 3. Therefore, based on accessibility and the kappa numbers, steam explosion could be a more efficient pre-treatment for kraft pulping than steam pretreatment, without ‘‘explosion’’. 3.2.2. Effect on pulp viscosity As far as pulp viscosity is concerned (Fig. 3b), no significant differences were observed between pulps obtained from steam explosion or steam pre-treatments. Higher viscosity values (around 1200 ml/g) were found in all pre-treated pulps compared to the control pulp (880 ml/g). Previously, Mendes et al. (2009), Al-Dajani et al. (2009) and Saukkonen et al. (2012), who studied autohydrolysis before kraft pulping of E. globulus, Populus tremuloides and Pinus sylvestris L., respectively, found a slight increase in pulp viscosity compared to the control pulp with similar kappa numbers to the pre-treated pulps. Mendes et al. (2009) attributed the increase in viscosity to the greater dissolution of low molecular weight carbohydrates. However, Martín-Sampedro et al. (2012b) and Sánchez et al. (2011) observed a decrease in viscosity when steam explosion or autohydrolysis, respectively, was applied before subsequent pulping of Hesperaloe funifera. The reason of these findings could be that this non-wood raw material has a more open structure than wood which would enhance the degradation of cellulose under severe hydrothermal pre-treatments. Similarly, Vila et al. (2011, 2012) reported that harsh autohydrolysis conditions could cause a viscosity drop in E. globulus kraft pulps. Therefore, mild hydrothermal pre-treatments, such as those carried out in this study, are recommended (Jedvert et al., 2012). Nevertheless, mild pulping conditions (shorter cooking time or lower cooking temperature) would also contribute to avoid overcooking of pretreated samples, which would imply a reduction in pulp viscosity (Martín-Sampedro et al., 2011a; Vila et al., 2012). 3.2.3. Effect on pulp yield Fig. 3c shows that the overall pulp yields (considering both the pre-treatment and the pulping yield) obtained with pre-treated chips were 7–13% lower than those observed with control chips. These findings were expected and can be attributed to the solubilization of hemicelluloses during the pre-treatments. Furthermore, remaining hemicellulose in the chips after the pre-treatments
might be more susceptible to alkaline peeling and solubilization during pulping, thus causing further yield losses (Chirat et al., 2012; Saukkonen et al., 2012). For these reasons, the screening kraft yield resulted to be lower with pre-treated pulps (45–50%) compared to control pulps (56%). Similar findings have been reported elsewhere with the use of different raw materials and different cooking processes after autohydrolysis (Al-Dajani et al., 2009; Mendes et al., 2009; Vila et al., 2011) and steam explosion pre-treatments (Ahvazi et al., 2007; Martín-Sampedro et al., 2011a). However, Hamzeh et al. (2013) and Martín-Sampedro et al. (2012b) found greater pulping yields with pre-treated nonwood materials (autohydrolyzed baggase and steam exploded Hesperaloe funifera, respectively), compared to un-treated controls. One plausible explanation for this could be the smaller amount of extractives found in pre-treated samples compared to untreated samples (Hamzeh et al., 2013). The higher the pre-treatment severity, the lower the overall yield; with a 43–44% yield (with and without explosion, respectively) when the most severe pre-treatment was used, compared to the control (56%). Vila et al. (2011) also showed a similar maximum reduction in yield after kraft pulping of autohydrolyzed E. globulus chips (40.2% yield compare to 55.1% with their control). Finally, the use of steam exploded versus steam unexploded chips did not result in clear differences in yields after kraft pulping.
3.2.4. Effect on chemical consumption As Fig. 3d shows, all pre-treatments increased the consumption of NaOH during the kraft pulping compared to the control pulp. Similar rates of increase in the consumption of chemicals have been reported by several authors using either autohydrolysis or steam explosion as pre-treatments (Martín-Sampedro et al., 2011a; Mendes et al., 2009; Saukkonen et al., 2012). These findings can be put down to the fact that the amounts of extractives present in the pre-treated pulps were greater. These extractives could consume pulping reagent because, generally, extraneous substances present in the raw materials require soda chemicals for their removal (Hamzeh et al., 2013). Nevertheless, it should be taken into account that this higher chemical consumption produced pulps with lower kappa numbers (Fig. 2a) suggesting a higher efficiency of the chemical in delignification.
3.3. Effect of steam pre-treatments on the morphological properties of fibers The results of the morphological analysis of kraft pulps are summarized in Table 2, which lists the characteristics of fibers, vessels
Table 2 Morphological analysis of E. globulus kraft pulps. Values in brackets indicate standard deviations. Steam explosion
Steam treatment
Control
1C:5
2C:10 + 3
1C: 5
2C:10 + 3
Fibers Length (lm) Width (lm) Coarseness Kinked fibers (%) Average Curl (%)
943.5 (9.2) 19.0 (0.1) 93.5 (0.1) 18.5 (0.4) 5.5 (0.2)
917.0 (4.2) 16.2 (0.2) 74.0 (1.1) 26.1 (0.6) 7.5 (0.2)
875.5 (14.8) 15.0 (0.0) 64.3 (1.8) 33.5 (1.0) 9.4 (0.0)
907.5 (14.8) 15.9 (0.0) 76.5 (2.0) 32.3 (0.3) 8.8 (0.0)
906.0 (8.5) 15.3 (0.1) 65.0 (1.3) 32.4 (0.7) 9.1 (0.0)
Vessels Length (lm) Width (lm)
728.5 (51.6) 136.3 (2.1)
776.0 (41.0) 149.0 (2.4)
722.0 (12.3) 157.2 (6.2)
783.0 (66.5) 161.5 (3.1)
682.5 (33.2) 162.8 (1.6)
56.2 (2.8) 5.2 (0.9) 38.6 (1.8)
52.2 (2.4) 2.7 (1.2) 45.1 (1.3)
54.4 (0.2) 3.8 (0.5) 41.8 (0.6)
50.8 (0.2) 3.5 (0.4) 45.8 (0.5)
Composition (in number of elements) Fibres (%) 53.1 (4.0) Vessels (%) 8.8 (1.3) Fines (%) 38.1 (5.3)
242
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244
(b)
8.0
30.0
6.0
25.0
5.0 4.0 3.0 2.0 1.0 0.0 0 Exp
0
0
NoExp
10 5 0
0.0 1C:10 NoExp - R
2C:5+3
2C:10+3
Control - R
80 70 60 50 40 30 20 10 0
0 Exp
0
0
NoExp
0
1C:5
Control
Exp - R
(f) Tear index (mNm2/g)
Tear index (mNm2/g)
5.0 1C:5
(d)
15
4.0
10.0
Exp - R
20
5.0
15.0
Control
25
(e)
20.0
0
Tensile index (Nm/g)
Tensile index (Nm/g)
(c)
35.0
7.0 Stretch at break (%)
Stretch at break (%)
(a)
3.0 2.0 1.0 0.0
1C:10 NoExp - R
2C:5+3
2C:10+3
Control - R
12 10 8 6 4 2 0
0 Exp
0 NoExp
0
0 Control
1C:5 Exp - R
1C:10 NoExp - R
2C:5+3
2C:10+3
Control - R
Fig. 4. Stretch at break (a and b), tensile index (c and d) and tear index (e and f), for unrefined (Exp, NoExp and Control) and refined pulps (Exp-R, NoExp-R and Control-R).
and fines for some of the steam exploded and steam treated samples. The effect of both treatments on the fiber characteristic is a slight reduction in fiber length and a more noticeable reduction of fiber width and coarseness (the weight of a given length of fibers). All these changes are more striking when the pre-treatments increase in severity, although no significant differences were observed between the fibers obtained from steam explosion or steam pre-treatments. This finding is the result of a better accessibility of the pre-treated chips by cooking chemicals which leads to a more intensive removal of hemicelluloses and lignin and, consequently to narrower fibers. The width decrease and the rigidity increase, caused by the loss of hemicelluloses, clearly affects the internal strength of the fibers and therefore there is an increase in the percentage of kinked fibers increases from the control pulp (18.5%) to the pre-treated kraft pulps (26.1–33.5%). Also, the curl of the fibers increases due to the fiber weakness, from 5.5% (control) to 9.1– 9.4% in the fibers obtained with the most severe pre-treatments. The vessels, the other characteristic element in eucalyptus pulps, have shown minor changes and maintain similar length both in control samples and in pre-treated samples of pulp. Unlike
fibers, vessel width experienced a significant increment with respect to control pulps of up to a 20% (2C:10 + 3 steam unexploded sample). The cause of this radial swelling could be a more effective penetration of chemicals and water in the pre-treated chips. Simultaneously to swelling, the percentage of vessels in the pulp suspension decreases as the severity of the treatment rises. This finding is especially significant for 2C:10 + 3 sample, obtained with the most severe steam explosion conditions, which reduced the percentage of vessels to 2.7%, only 31% of the fine content of the control pulp. Parallelly, the percentage of fines increases in the same way, ranging from 7.0% to 7.7% for the more severe conditions. This suggests that the vessels are de-structured as a result of the action of pulping chemicals, which is more aggressive after the pre-treatment. What is measured as fine material are fragments of these vessels, and that would explain why the percentage of fines becomes larger. Unlike what it does to the percentage of vessels, the introduction of a pretreatment does not change the percentage of fibers significantly. This difference can be put down to the fact that fibers are more massive than vessels and, although reduced in thickness, they would endure the same treatment for longer before their integrity is affected.
243
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244
After the morphological analysis of the pulp suspension, it can be concluded that the reduction of fiber length due to the pretreatments (only a 7% for the worst value) only has a slight impact on the mechanical properties of the paper, while fiber width reduction (up to 21%) and the associated loss of coarseness (up to 31%) means a significant loss of mechanical properties as the pre-treatment renders the fibers brittle and weaker. 3.4. Impact of steam explosion and steam pre-treatments on refining and paper properties Control and pre-treated kraft pulps were refined in a PFI mill at 2000 revolutions increasing the degree of refining (Schopper-Riegler) from 12 (control pulp) and 14 ± 1 (pre-treated pulps) to 21 in all cases. This result indicates that pre-treated pulps required more energy to obtain similar increase in degree of refining compared to un-treated pulps. This is in consistent with previous findings (Helmerius et al., 2010; Kautto et al., 2010; Martín-Sampedro et al., 2011a), since shorter refining times and less energy absorption in refining is a feature typical of kraft pulps having high hemicelluloses content. Hemicelluloses contribute significantly to swelling of fibers, which increases interfiber contact during refining promoting faster external fibrillation and hence better refining response (Kautto et al., 2010). Fig. 4 shows the stretch at break and tensile and tear indexes of unrefined and refined pulps. As expected, refined pulps presented higher paper strength than non-refining pulps. In general, it can be observed that handsheets formed from pre-treated pulps presented lower mechanical properties in comparison with controls. Both the lower content in hemicelluloses and the greater amount of fiber deformations (kinking and curling) in the pre-treated pulps contribute to the loss of properties associated to strength (Saukkonen et al., 2012). This increase in curled and kinked fibers in pretreated pulps has been shown in Table 2 and is a consequence of the weakness of the fibers (less width and coarseness) which
(b)
50 45 40 35 30 25 20 15 10 5 0
Brightness (%ISO)
Brightness (%ISO)
(a)
1C:5 Exp
(c)
would result in a paper with poor mechanical strength. While a slight decrease of paper strength was observed when the duration of the pre-treatment increased, no clear relation with neither the number of cycles, nor the type of hydrothermal pre-treatment (steam explosion or steam treatment), was found. The lower values of stretch at break and tensile index observed for pre-treated pulps compared to the control pulp, are consistent with results reported previously pre-treating with autohydrolysis (Chirat et al., 2012; Hamzeh et al., 2013; Helmerius et al., 2010; Kautto et al., 2010; Saukkonen et al., 2012; Vila et al., 2012) and steam explosion (Martín-Sampedro et al., 2011a; San Martin et al., 1995). These properties are related to the fiber-to-fiber bonding in the fiber net-work, which decreases with the loss of hemicelluloses and with the increase in fiber deformations caused by the pre-treatments. On the other hand, the tear index is related to fiber degradation and some authors have found it to increase or not significantly change towing to hydrothermal pre-treatments (Hamzeh et al., 2013; Helmerius et al., 2010; Kautto et al., 2010). However, other authors found a decrease in tear index (Chirat et al., 2012; Saukkonen et al., 2012; Vila et al., 2012), which would indicate a degradation of the fibers, and it is consistent with the results obtained here. Nevertheless, viscosity values reported by the latter authors did not indicate a degradation of the cellulose chains during the pre-treatment, as it was discussed above. Therefore, the lower tear index observed could be caused by other factors such as fiber morphology. Moreover, the reduction in coarseness is a reflection of fiber damage, and it is consistent with the observed loss of tear resistance in handsheets, a parameter strongly dependent on individual fiber strength. Contrarily to mechanical properties, optical properties of papersheets improved when a hydrothermal pre-treatment was carried out. Thus, higher brightness and lower yellowness index were observed in pre-treated pulps compared to the control pulp (Fig. 5). This improvement in optical properties is mainly caused by the
1C:10 NoExp
2C:5+3
1C:5
2C:10+3
Control
Exp - R
(d)
40 30
Yellowness Index
Yellowness Index
35 25 20 15 10 5 0 1C:5 Exp
1C:10 NoExp
45 40 35 30 25 20 15 10 5 0
2C:5+3
2C:10+3
Control
1C:10 NoExp - R
2C:5+3
2C:10+3
Control - R
45 40 35 30 25 20 15 10 5 0 1C:5 Exp - R
1C:10 NoExp - R
2C:5+3
2C:10+3
Control - R
Fig. 5. Brightness (a and b) and yellowness index (c and d) for unrefined (Exp, NoExp and control) and refined pulps (Exp-R, NoExp-R and Control-R).
244
R. Martin-Sampedro et al. / Bioresource Technology 153 (2014) 236–244
lower lignin content (kappa number) and hexenuronic acid content of the pre-treated samples. Therefore, exploded samples, which experienced a clear decrease in kappa number when the severity of the pre-treatment increased (Fig. 2a), provided handsheets with higher brightness and lower yellowness index with more severe pre-treatments. Several authors have also reported better optical properties when pre-treating pulps with autohydrolysis (Hamzeh et al., 2013; Kautto et al., 2010; Vila et al., 2011) and steam explosion (Jedvert et al., 2012; Martín-Sampedro et al., 2011a,b; San Martin et al., 1995) compared to the control pulp. Furthermore, Chirat et al. (2012), Hamzeh et al. (2013), Martín-Sampedro et al. (2011b) and Vila et al. (2011) observed that hydrothermally pretreated pulps showed better bleaching ability than non pre-treated pulps, which means better optical properties consuming fewer chemicals during bleaching. Finally, refining caused a decrease in brightness and an increase in yellowness index. Eugenio et al. (2011), who analyzed unrefined and refined E. globulus pulps by thioacidolysis-SEC, attributed this darkening of the pulps to the formation of new chromophores during refining. 4. Conclusion Steam explosion and steam pre-treatments are appropriate not only to extract hemicelluloses but also to enhance delignification. Nevertheless, steam explosion seems to be more efficient to this end, as it achieved the biggest reduction of the kappa number (68% vs 45% with the steam pre-treatment). Therefore, it is reasonable to assume that these pre-treatments would reduce the pulping time and consequently would increase the pulp production, although consuming more wood. The paper obtained could be used in applications in which good optical properties, rather than good mechanical properties, are required, such as manufacturing tissue paper or dissolving pulps. Acknowledgements The authors wish to thank the Spanish MINECO for funding this study via Project CTQ 2011-28503-C02-01 and Program PTA 20114857-I, and the Community of Madrid for funding via Project PROLIPAPEL–CM (S2009/AMB-1480). Special thanks to Concepción Monte (Complutense University), for her valuable collaboration in the morphological characterization of fibers. References Ahvazi, B., Radiotis, T., Bouchard, J., Goel, K., 2007. Chemical pulping of steamexploded mixed hardwood chips. J. Wood Chem. Technol. 27 (2), 49–63. Al-Dajani, W.W., Tschirner, U.W., Jensen, T., 2009. Pre-extraction of hemicelluloses and subsequent kraft pulping Part II: acid- and autohydrolysis. Tappi J. 8 (9), 30–37. Chandra, R., Bura, R., Mabee, W., Berlin, A., Pan, X., 2007. Substrate pretreatment: the key to effective enzymatic hydrolysis of lignocellulosics? Adv. Biochem. Eng./Biotechnol. 108, 67–93. Chirat, C., Lachenal, D., Sanglard, M., 2012. Extraction of xylans from hardwood chips prior to kraft cooking. Process Biochem. 47 (3), 381–385. Eugenio, M.E., Du, X., Li, J. 2011. Towards improvements of kraft pulp bleaching by additional treatments 16th ISWFPC.Tiajin, China. Feria, M., García, J., Díaz, M., Fernández, M., López, F., 2012. Biorefinery process for production of paper and oligomers from Leucaena leucocephala K360 with or without prior autohydrolysis. Bioresour. Technol. 126, 64–70. Garrote, G., Dominguez, H., Parajó, J.C., 1999. Mild autohydrolysis: an environmentally friendly technology for xylooligosaccharide production from wood. J. Chem. Technol. Biotechnol. 74 (11), 1101–1109. Gutsch, J.S., Nousiainen, T., Sixta, H., 2012. Comparative evaluation of autohydrolysis and acid-catalyzed hydrolysis of Eucalyptus globulus wood. Bioresour. Technol. 109, 77–85.
Hamzeh, Y., Ashori, A., Khorasani, Z., Abdulkhani, A., Abyaz, A., 2013. Pre-extraction of hemicelluloses from bagasse fibers: effects of dry-strength additives on paper properties. Ind. Crop Prod. 43, 365–371. Helmerius, J., von Walter, J.V., Rova, U., Berglund, K.A., Hodge, D.B., 2010. Impact of hemicellulose pre-extraction for bioconversion on birch kraft pulp properties. Bioresour. Technol. 101 (15), 5996–6005. Jarabo, R., Monte, M.C., Blanco, A., Negro, C., Tijero, J., 2012. Characterisation of agricultural residues used as a source of fibres for fibre-cement production. Ind. Crop Prod. 36 (1), 14–21. Jedvert, K., Wang, Y., Saltberg, A., Henriksson, G., Lindstrom, M.E., Theliander, H., 2012. Mild steam explosion: a way to activate wood for enzymatic treatment, chemical pulping and biorefinery processes. Nord. Pulp Pap. Res. J. 27 (5), 828– 835. Kautto, J., Saukkonen, E., Henricson, K., 2010. Digestibility and paper-making properties of prehydrolyzed softwood chips. Bioresources 5 (4), 2502–2519. Leschinsky, M., Sixta, H., Patt, R., 2009. Detailed mass balances of the autohydrolysis of Eucalyptus globulus at 170 °C. Bioresources 4 (2), 687–703. Li, H., Saeed, A., Jahan, M.S., Ni, Y., Van Heiningen, A., 2010. Hemicellulose removal from hardwood chips in the pre-hydrolysis step of the kraft-based dissolving pulp production process. J. Wood Chem. Technol. 30 (1), 48–60. Li, J., Henriksson, G., Gellerstedt, G., 2005. Carbohydrate reactions during hightemperature steam treatment of aspen wood. Appl. Biochem. Biotechnol. 125 (3), 175–188. Li, J., Henriksson, G., Gellerstedt, G., 2007. Lignin depolymerization/ repolymerization and its critical role for delignification of aspen wood by steam explosion. Bioresour. Technol. 98 (16), 3061–3068. Liu, S., Amidon, T.E., 2007. Essential components of a wood based biorefinery. O Papel. 68 (3), 84–105. Martín-Sampedro, R., Eugenio, M.E., Revilla, E., Martín, J.A., Villar, J.C., 2011a. Integration of kraft pulping on a biorefinery by the addition of a steam explosion pretreatment. Bioresources 6 (1), 513–528. Martín-Sampedro, R., Eugenio, M., Villar, J., 2011b. Biobleaching of Eucalyptus globulus kraft pulps: comparison between pulps obtained from exploded and non-exploded chips. Bioresour. Technol. 102, 4530–4535. Martín-Sampedro, R., Martin, J.A., Eugenio, M.E., Revilla, E., Villar, J.C., 2011c. Steam explosion treatment of Eucalyptus globulus wood: influence of operational conditions on chemical and structural modifications. Bioresources 6 (4), 4922– 4935. Martín-Sampedro, R., Capanema, E.A., Hoeger, I., Villar, J.C., Rojas, O.J., 2011d. Lignin changes after steam explosion and laccase-mediator treatment of eucalyptus wood chips. J. Agric. Food Chem. 59 (16), 8761–8769. Martín-Sampedro, R., Eugenio, M.E., García, J.C., Lopez, F., Villar, J.C., Diaz, M.J., 2012a. Steam explosion and enzymatic pre-treatments as an approach to improve the enzymatic hydrolysis of Eucalyptus globulus. Biomass Bioenerg. 42, 97–106. Martín-Sampedro, R., Eugenio, M.E., Villar, J.C., 2012b. Effect of steam explosion and enzymatic pre-treatments on pulping and bleaching of Hesperaloe funifera. Bioresour. Technol. 111, 460–467. Mendes, C.V.T., Carvalho, M., Baptista, C., Rocha, J.M.S., Soares, B.I.G., Sousa, G.D.A., 2009. Valorisation of hardwood hemicelluloses in the kraft pulping process by using an integrated biorefinery concept. Food Bioprod. Process. 87 (C3), 197– 207. Overend, R., Chornet, E., 1987. Fractionation of lignocellulosics by steam-aqueous pretreatments. Philos. Trans. R. Soc. London, A 321 (1561), 523–536. Rahikainen, J.L., Martín-Sampedro, R., Heikkinen, H., Rovio, S., Marjamaa, K., Tamminen, T., Rojas, O.J., Kruus, K., 2013. Inhibitory effect of lignin during cellulose bioconversion: the effect of lignin chemistry on non-productive enzyme adsorption. Bioresour. Technol. 133, 270–278. Ruiz, E., Cara, C., Manzanares, P., Ballesteros, M., Castro, E., 2008. Evaluation of steam explosion pre-treatment for enzymatic hydrolysis of sunflower stalks. Enzyme Microb. Technol. 42 (2), 160–166. San Martin, R., Perez, C., Briones, R., 1995. Simultaneous production of ethanol and kraft pulp from pine (Pinus radiata) using steam explosion. Bioresour. Technol. 53 (3), 217–223. Sánchez, R., Rodríguez, A., García, J., Rosal, A., Jiménez, L., 2011. Exploitation of hemicellulose, cellulose and lignin from Hesperaloe funifera. Bioresour. Technol. 102, 1308–1315. Saukkonen, E., Kautto, J., Rauvanto, I., Backfolk, K., 2012. Characteristics of prehydrolysis-kraft pulp fibers from Scots pine. Holzforschung 66, 801–808. Tuck, C.O., Pérez, E., Horváth, I.T., Sheldon, R.A., Poliakoff, M., 2012. Valorization of biomass: deriving more value from waste. Science. 337 (6095), 695–699. Vila, C., Romero, J., Francisco, J.L., Garrote, G., Parajo, J.C., 2011. Extracting value from Eucalyptus wood before kraft pulping: effects of hemicelluloses solubilization on pulp properties. Bioresour. Technol. 102 (8), 5251–5254. Vila, C., Romero, J., Francisco, J.L., Santos, V., Parajó, J.C., 2012. On the recovery of hemicellulose before kraft pulping. Bioresources 7 (3), 4179–4189. Wise, L.E., Murphy, M., d’Addieco, A., 1946. Chlorite holocellulose, its fractionnation and bearing on summative wood analysis and on studies on the hemicelluloses. Paper Trade J. 122 (2), 35–43.