Impact of Drying on Wood Ultrastructure: Similarities in Cell Wall

Jul 9, 2010 - For a more comprehensive list of citations to this article, users are ... effect of wetting cycles on moisture behaviour of thermally mo...
1 downloads 0 Views 2MB Size
Biomacromolecules 2010, 11, 2161–2168

2161

Impact of Drying on Wood Ultrastructure: Similarities in Cell Wall Alteration between Native Wood and Isolated Wood-Based Fibers Miro Suchy, Eero Kontturi,* and Tapani Vuorinen Department of Forest Products Technology, School of Science and Technology, Aalto University, P.O. Box 16300, FIN-00076 Aalto, Finland Received May 20, 2010; Revised Manuscript Received June 24, 2010

Ultrastructural alterations of fresh wood caused by initial drying were compared to changes incurred during drying of never-dried wood pulp fibers of different macromolecular composition. Drying induced inaccessibility of a native wood sample exhibited remarkable similarity to wood pulp samples of different lignin contents. The results suggest that the supramolecular rearrangements in native wood matrix upon dehydration are qualitatively identical to the well-known changes occurring in pulp fibers after drying, although the changes are considerably different in quantity. The alterations were observed and quantified by monitoring the conversion of accessible deuterium exchanged OH groups in fresh wood and wood pulp fibers to inaccessible, reprotonation resistant OD groups during drying. The deuteration/FT-IR measurements correlated well with the water retention measurement of the pulp samples. Irreversible reduction of water retention due to the supramolecular changes implies reduced accessibility of wood polymers in various chemical and mechanical treatments, such as enzymatic conversion of biomass or preparation of cellulosic nano-objects for diverse applications.

1. Introduction Current global developments in the traditional wood, pulp, and paper industry, coupled with efforts to reduce fossil fuel dependency have stimulated renewed interest in cellulose research. Its native abundance and renewable nature are appealing factors considered in attempts to devise greener alternatives for synthetic or nonrenewable materials.1,2 In contemporary cellulose research, two areas in particular are enjoying increased attention: (i) production of biofuels and (ii) cellulosic nanoobjects fabrication.3 Production of bioethanol from biomass, of which cellulose is the major constituent, is facilitated through glucose that is a product of cellulose degradation. Preparation of cellulosic nano-objects, on the other hand, is based on isolation of high aspect ratio particles whose dimensions are originally determined by biosynthesis of cellulosic plants.4,5 Successful utilization of these particles in a variety of fields has generated a vast interest within the materials science community.3,6,7 Native and isolated cellulose exists in the form of fibril aggregates (microfibrils) formed by strong intra- and intermolecular hydrogen bonds, dipole and van der Waals interactions. This supramolecular structural organization plays an important role in the accessibility and degradation of cellulose.8,9 For example, molecularly dissolved cellulose degrades much faster than native cellulose within microfibrils.10-12 In the same vein, crystallinity and crystalline structure have a profound impact on the kinetics of heterogeneous cellulose degradation reactions.13 Except for the cellulose molecules on the surface of crystallites or between them, cellulose is not readily accessible to the degrading reagents. Similarly, the supramolecular characteristic of cellulose is relevant in the preparation of cellulosic nano-objects.14 The cellulosic nano-objects are extracted from cellulosic substrates either by mechanical disintegration or * To whom correspondence should be addressed. Tel.: +358 9 470 24250. Fax: +358 9 470 24259. E-mail: [email protected].

selective oxidation (nanofibrillar cellulose) or by controlled acid hydrolysis (cellulose nanocrystals). Depending on the substrate, nanocellulose preparation requires various pretreatments. Wood, for instance, must be stripped of components other than cellulose and its anisotropic network of fibers must be disintegrated. This can be accomplished by, for example, chemical pulping and subsequent enzymatic treatments. Most of these pretreatment methods heavily influence the supramolecular structure of components in the substrate. Chemical pulping, for one, increases the size of microfibril bundles and porosity in wood fibres.15,16 Upon close examination of previous efforts in microfibrilar cellulose preparation, it is interesting to note that all cellulose microfibrils, which have been isolated by mechanical disintegration, are, at least partially, microfibril bundles, rather than individual microfibrils.4,5,17 Another important preprocessing step, which influences the supramolecular arrangement of cellulose, is drying. Although it does not alter the crystallinity or crystalline structure, it does have an effect on the size of microfibril bundles and, subsequently, on the accessibility of cellulose.18 In addition, drying can profoundly alter the response of cellulosic substrates toward, for example, enzymatic19 or acid-catalyzed degradation. Surprisingly little information on the impact of drying on a molecular level is available. Furthermore, the presence of a possible drying step in the experimental procedures is often neglected when cellulosic substrates are subjected to various treatments. Our earlier work has shown that drying does change the supramolecular structure of fresh wood and that these changes can be monitored by FT-IR spectroscopy coupled with deuterium exchange.20 In this study, we follow the impact of drying on several wood-based, industrially produced pulps and compare their behavior to that of a native wood matrix by applying the same experimental concept. The trends observed for the pulp samples with known consequences to the fiber properties and mechanisms involved were then correlated with the results observed for drying of fresh native wood sample. Wood pulps

10.1021/bm100547n  2010 American Chemical Society Published on Web 07/09/2010

2162

Biomacromolecules, Vol. 11, No. 8, 2010

Scheme 1. Schematic of Wood Samplinga

a

Dimensions are not to scale.

consist of isolated fibers with different macromolecular composition (mainly in their lignin content), whereas in native wood, the matrix is anisotropic with fibers aligned longitudinally with respect to the tree trunk. Understanding the changes induced by drying on a molecular level will help to better explain the different responses of cellulosic substrates to different treatments, such as acid or enzymatic hydrolysis, or mechanical disintegration, wherein the substrates have undergone drying beforehand. The importance of this knowledge is further amplified when considering the inevitable partial or full drying of biomass prior to its actual processing. Knowledge on the fundamental aspects of drying wood-based cellulosic substrates should further promote the utilization of biomass for a variety of purposes.

2. Experimental Section Materials. Freshly felled pine (Pinus sylVestris) wood samples from Eastern Finland were supplied in the form of discs, 7-10 cm thick and 20-40 cm in diameter. The specimens for FT-IR analysis were sampled using a Suunto increment borer (Suunto, Finland). The cores (5.15 mm in diameter) were then sliced to 0.5-1 mm thick discs by using a sharp knife. The dimensions of the specimen were selected to agree with the size requirements of the FT-IR photoacoustic detection cell. The sample preparation is depicted in Scheme 1. Freshly produced never dried wood pulp samples were used in this investigation. A softwood groundwood sample was obtained from a paper mill in South Eastern Finland. Chemical (kraft) pulp samples were also obtained from a pulp mill in South Eastern Finland. The kraft pulps were made from spruce and pine wood (approximately 60% spruce and 40% pine). The collected samples were washed and centrifuged to approximately 30% solid content before shipping. The kappa numbers (an approximation of lignin content) of the kraft pulps, as determined by ISO standard 302:2004, were 23.1 for an unbleached sample and 0.7 for a fully bleached sample. Deuterium oxide (99.9 atom % D, Sigma-Aldrich) was used for deuteration and relative humidity control. The salts used for humidity control were NaOH (p.a. Merck) and NaCl (99.5%, J.T. Baker). A saturated solution of these salts placed in a closed environment created approximately 7 and 75% D2O relative humidity, respectively. Although the approximations are based on literature data for water,21 the D2O

Suchy et al. Scheme 2. Schematic of Deuteration and Controlled Drying Experiments

relative humidity values reported for saturated solutions of these salts in deuterium at 20 °C did not differ significantly from the values reported for water.22 Deuteration and Drying at Controlled Relative Humidity. The deuteration was carried out in 10 mL glass vials by immersing wood samples in an excess of D2O for 60 min. The pulp samples were deuterated in plastic bags for a period of 2 × 20 min. Excess D2O was added to the pulp and then the slurry was mixed by kneading. After 20 min, the free D2O was squeezed out of the bag before fresh deuterium oxide was added and the slurry was mixed again. After the treatment, the samples were dried under different conditions and then flushed with an excess of water for an identical period of time as the deuteration process (2 × 20 and 60 min for pulp and wood, respectively). Both deuteration and flushing were carried out at room temperature. All samples were then dried in a convection oven at 40 °C prior to measurement with the FT-IR spectrometer. The drying at controlled D2O relative humidity was carried out in vacuum desiccators. A schematic of the experiments is shown in Scheme 2. The samples were held in perforated aluminum containers and placed on the porcelain plate of desiccators containing D2O saturated solutions at the bottom. The desiccators were then evacuated and placed into an oven (25, 60, and 80 °C) for conditioning (7 days). FT-IR Measurement. The spectra were collected using a Bio-Rad FTS 6000 spectrometer (Cambridge, MA) with a MTEC 300 photoacoustic cell (Ames, IA) at a constant mirror velocity of 5 kHz, 1.2 kHz filter, and 8 cm-1 resolution. A background spectrum with standard carbon black was measured at the beginning of each set of measurements. After collecting the background spectrum, the wood or pulp sample was put into the sample holder that was then placed into the PA detection cell and the cell was purged with helium gas for 5 min. Then the cell was sealed and the actual spectrum of the sample was recorded. A minimum of 400 scans per spectrum were collected and processed using the Win-IR Pro 3.4 software (Digilab, Randolph, MA). Each spectrum was normalized to have the same value at 1200 cm-1.23 The spectra presented throughout this article are averages of at least four measurements. Each treatment and drying scenario was carried out in triplicate. At least two samples were measured (both sides), with an additional sample measured if a noticeable difference between the two measurements was observed. Pulp Drying and WRV Testing. Pulp samples for water retention value (WRV) measurement were dried under similar conditions to those described for the drying of deuterated samples. The presence of deuterium was not necessary in this testing, thus, the relative humidity was achieved using saturated aqueous solutions. The WRV of the pulps was then determined according to the standard ISO 23714:2007 with a Jouan GR 4.22 centrifuge.

Cell Wall Alterations in Native Wood and Wood Pulp

Biomacromolecules, Vol. 11, No. 8, 2010

2163

3. Results and Discussion 3.1. Deuterium Exchange in Wood and Pulp Fibers and Its Reversibility. The main difficulty associated with the investigation of the native wood ultrastructure is the requirement of sample preparation for the majority of applicable analytical techniques. During preparation, water in the sample is often removed or replaced with a suitable organic solvent. This removal or exchange can alter the original structure of the cell wall, which therefore may not be entirely representative of the wood in its native swollen hydrated state. To investigate the alterations in the ultrastructure of fresh wood during initial drying, it was important to select a methodology that would require minimal sample preparation to preserve the native wood structure. The concept of deuterium exchange in combination with FT-IR spectroscopy selected for this evaluation fulfills these requirements. Cellulose or wood samples can be readily deuterated by simple exposure to deuterium oxide vapor or by immersion in liquid deuterium oxide. Deuteration of cellulose has been described previously and utilized to study the accessibility of cellulose, cellulose derivatives, and cellulose in wood.23-27 The exchanged deuterium in the cellulose sample can be easily detected by IR spectroscopy, because the OD band signal is in an area of spectrum without interference from other signals. Photoacoustic detection FT-IR spectroscopy is optimal for this testing because, apart from partial drying of the sample prior to the measurement, it does not require additional sample preparation. Our previous investigation demonstrated the suitability and reliability of this experimental concept in detecting the ultrastructural alterations in wood upon drying.20 In the present investigation, in addition to further exploring these alterations in the fresh wood, samples of never dried wood pulp fibers of different characteristics were evaluated. Mechanical pulp (groundwood) is produced by separating the fibers from fresh wood using elevated temperature and mechanical forces (grinding).28 During the process, the native wood matrix is disintegrated into fibrous pulp and because mainly mechanical force is applied in the separation, the chemical composition of the produced pulp is not significantly altered and, thus, is similar to that of wood. Chemical pulps are produced by cooking the wood in an aqueous solution of chemicals (i.e., Na2S/NaOH in kraft process) at elevated temperatures and high pressures.29,30 During this process, the lignin in the wood is chemically degraded and extracted from the fiber matrix. Approximately 90% of lignin, up to 60% of hemicelluloses and 15% of cellulose is removed in the course of chemical pulping,30 meaning that the unbleached chemical pulp sample contains 2-5% (dry mass) of residual lignin, while the bleached sample represents the fully delignified carbohydrate segment of the original wood cell wall. In addition to comparison with wood and groundwood samples, a direct comparison of both bleached and unbleached pulps was designed to better understand the impact of residual lignin on the structural alteration of chemical pulps during drying. A fundamental requirement for this experimental concept was to be able to achieve complete reprotonation of the deuterated never dried samples. In our previous work we demonstrated the complete deuteration reversibility for a fresh wood sample.20 Similar experiments were carried out for all wood and pulp samples used in this investigation. Detailed analyses and pertinent spectra from the deuteration reversibility study of wood, groundwood, and chemical pulps, both unbleached and bleached, are available in Supporting Information (Figure S1). As previously reported in the literature for cellulose,23 and in agreement with our initial work, the deuteration of fresh wood

Figure 1. Comparison of maximum deuterium exchange measured for fresh wood and never dried pulp samples. Deuteration time: 60 min (wood) and 2 × 20 min (pulps); 25 °C. Below: Direct OD band size comparison (left) and measured OD band area comparison (right).

and never dried pulp samples is fully reversible by exposing the deuterated sample to water (reprotonation).20 This deuteration reversibility thus allows the presence of residual, reprotonation resistant OD groups in the sample to act as an indicator of irreversible alterations. In addition to reversibility, the maximum extent of deuteration for each substrate was measured and compared. This measurement was carried out by deuterating fresh samples, placing them into desiccators containing drying agent (to prevent any possible interaction with water vapors) and drying at 40 °C prior to FTIR measurement. The OD groups measured in the deuterated samples indicate the amount of accessible OH groups originally present in the sample. The complete spectra, OD band and OD band area comparisons are shown in Figure 1. It is clearly evident that, under the selected conditions, the amount of the deuterated accessible OH groups in groundwood and chemical pulps is considerably greater when compared to wood. Interestingly, the measured amount of accessible OH groups in groundwood is almost twice that of the wood sample. The chemical composition of both wood and groundwood is similar; therefore, the increase in OH group accessibility can be mainly attributed to the effects of actual wood matrix disintegration. These include an increase in surface area and an increase in the cell wall volume of the fibers liberated from the wood matrix due to partial disruption of the fiber cell wall and the removal of the middle lamella. The mechanical pulping exposes significant amounts of fresh surface, which is bound to include regions that were previously inaccessible to water in

2164

Biomacromolecules, Vol. 11, No. 8, 2010

Suchy et al.

Figure 2. Comparison of OD band segment of spectra measured for deuterated wood and pulp samples after drying under controlled D2O relative humidity (7%) at different temperatures (25, 60, and 80 °C). Samples were flushed with water after drying.

Figure 3. Comparison of OD band areas of deuterated wood and pulp samples measured after drying under controlled D2O relative humidity (7%) at different temperatures. Samples were flushed with water after drying.

solid fresh wood. Furthermore, the lignin rich (relatively hydrophobic) middle lamella that holds the wood fibers together is at least partially removed.31 Thus, although the chemical structure is similar, the wood matrix itself is different from mechanical pulp due to alterations and partial removal of the middle lamella. The cell wall swelling is due to the release of compressive forces previously present in the compact native wood structure. In addition to the wood matrix disintegration, the increase in accessibility for the groundwood sample can be a result of the elevated temperatures that are applied in the groundwood production.28 The greater amount of accessible OH groups in chemical pulps correlates with the difference in chemical composition and structural alterations resulting from the chemical pulping process. Lignin removal increases the carbohydrate content from 40-45 in wood to 65-75% in unbleached pulp,30 and the carbohydrates (mostly cellulose) are the main contributors of accessible OH groups in the fiber. The extraction of lignin from the cell wall results in the formation of pores throughout the cell wall,15,16 which increases the swelling capacity of the fibers.32 During bleaching of chemical pulps, in addition to complete removal of the residual lignin, the carbohydrates (hemicelluloses in particular) are also partially degraded and removed. Additional lignin removal and ensuing further increase in cell wall porosity, as well as slightly higher relative cellulose content, thus, are responsible for the greater amount of accessible OH groups in the bleached pulp sample. 3.2. Controlled Drying of Wood and Pulps. In our previous investigation, we demonstrated that irreversible alterations in ultrastructure of fresh wood take place during initial drying. These changes were indicated by conversion of the formerly accessible OD groups (readily reprotonated by exposure to water) to inaccessible, reprotonation-resistant OD groups.20 The drying temperature and D2O relative humidity, the latter particularly at higher temperature (80 °C), were shown to have an impact on the extent of the alterations. In this investigation, the behavior of wood during drying was compared with wood pulp samples of different characteristics. To further examine the impact of drying temperature on the extent of alterations, an additional drying temperature level was evaluated. A comparison of the OD band region of spectra measured from samples dried at different temperatures and 7% D2O relative humidity is shown in Figure 2. Entire spectra are included in Supporting Information (Figure S2).

The comparison in Figure 2 clearly indicates the impact of drying temperature on the extent of ultrastructural alterations, particularly in the wood and groundwood samples. The overall extent of the alteration is visibly greater in the chemical pulps dried at all temperature levels. The differences in the extent of alterations between substrates and the impact of temperature is more evident when the OD band areas are compared directly, as shown in Figure 3. In addition to the clearly obvious impact of the drying temperature, comparison of the OD band areas amplifies the difference in the amount of inaccessible OD groups retained in dried wood and pulp samples. A similar trend to that in Figure 1 for the total amount of accessible OH groups in never dried samples is evident; the amount of residual inaccessible OD groups after drying is significantly lower in the wood and groundwood samples compared to the chemical pulp samples (Figure 3). Comparison of wood and groundwood samples showed that the OD retention and, thus, extent of the alteration in ultrastructure upon drying is similar for both when dried at low temperature. With increasing temperature, however, the difference becomes greater and is more pronounced, particularly in the samples dried at 80 °C. The unbleached and bleached chemical pulps follow a similar trend, with the pulp samples dried at 25 °C having a comparable amount of retained OD groups, while the difference observed for samples dried at higher temperature is markedly greater (Figure 3). When the retained OD groups after drying and flushing are expressed relative to the maximum OD exchanged (accessible OH in original sample), however, the trend changes. A comparison of relative deuterium retention (ratio between band areas of dried and flushed sample vs maximum deuterium exchange) is shown in Figure 4. The mechanical disintegration of wood increased the accessibility of OH groups in never dried groundwood (Figure 1) and subsequent retention of inaccessible OD groups after drying (Figure 3). However, the relative extent of the alterations (inaccessible OD conversion rate) is smaller compared to that of the wood sample, at both temperature levels. Of all the exchanged accessible OH groups within the wood sample, 27% became converted to inaccessible during drying at 25 °C, while only 17% became inaccessible in the groundwood sample dried at this temperature (Figure 4). When dried at 80 °C, both wood and groundwood showed a similar relative amount of retained

Cell Wall Alterations in Native Wood and Wood Pulp

Figure 4. Comparison of relative reduction of accessible OD band areas of deuterated wood and pulp samples dried under controlled conditions. Values calculated as the ratio between OD band area of dried and flushed sample vs maximum OD exchanged in never dried sample (from Figure 1).

OD groups, 59 and 54%, respectively. Significantly greater surface area of the groundwood fibers compared to that of solid wood sample and increase in the cell wall volume of the fibers liberated from the compact wood matrix are mainly responsible for the increased initial accessibility. The impact of the greater surface area of the groundwood fibers in drying experiments can be to some extent minimized because a similar area is then exposed to water during the flushing stage after drying. The lower relative amount of inaccessible OD in dried groundwood compared to dried wood indicates that the increased accessibility of the groundwood is not irreversibly altered by the drying and can for the most part be restored by rewetting. This finding correlates well with the previous research showing that the high yield pulp swelling characteristics are not affected to a great extent by recycling (drying).33 It is also possible that the elevated temperatures (up to 100 °C) that are common in mechanical pulping production28 might have an effect on the property of the fiber cell wall and be somewhat responsible for this difference in the relative extent and reversibility of the alterations. The chemical pulps showed much greater retention of the OD groups compared to wood and groundwood. Almost half (47%) of the OD groups were retained in the unbleached sample dried at 25 °C, while a retention of 42% was measured for the bleached sample (Figure 4). At 80 °C, the retention of the OD groups was significantly higher, with the unbleached pulp showing greater relative OD retention than the bleached sample (73 vs 68%). The increase in the extent of irreversible alteration of the chemical pulp was expected due to greater cell wall porosity and little or no lignin content. Both factors were shown to be instrumental in the reduction of swelling capacity of recycled fibers. The removal of water leads to irreversible closure of the pores and hydrogen bonds are formed between adjacent microfibrils.33,34 These formed bonds are resistant to hydration and do not break during rewetting, thus, the original swelling of the fiber cannot be restored. This phenomenon has been termed hornification and it is further discussed later in the text. It was interesting to note, however, that, despite the greater porosity and negligible quantity of lignin in the bleached pulp sample, the relative amount of inaccessible OD groups formed during drying was slightly greater in the unbleached sample.

Biomacromolecules, Vol. 11, No. 8, 2010

2165

Figure 5. Comparison of the OD band region of FT-IR spectra measured for wood and groundwood samples dried under two D2O RH levels (7 and 75%) at different temperatures (25, 60, and 80 °C).

3.3. Impact of Relative Humidity. Our previous investigation showed that the amount of OD groups retained in the wood structure was affected by the D2O relative humidity, particularly during drying at elevated temperatures. In contrast, no impact of relative humidity was observed in the samples dried at lower (25 °C) temperature. To further investigate the impact of relative humidity on ultrastructure alterations in relation to drying temperature, an additional drying experiment at 60 °C was included in the experimental design. The same experiments were carried out for wood and groundwood samples. A comparison of the OD band region of wood and groundwood spectra is shown in Figure 5. Entire spectra are included in Supporting Information (Figure S3). Similarly, as observed in our previous study,20 the D2O relative humidity did not have an impact on OD retention in wood samples dried at 25 °C (Figure 5). The same trend was observed for the groundwood sample. Neither was the retention of OD groups in wood and groundwood affected by the D2O relative humidity during drying at 60 °C. The D2O relative humidity however, was shown to have an impact on the OD retention in samples dried at 80 °C. Although the OD retention in the groundwood sample was also affected, the extent is markedly greater in the wood sample. Since no chemical changes are expected at this temperature, it appears that additional factors that are triggered at temperatures close to 80 °C play a role in the alteration of wood ultrastructure. Once the wood matrix is disintegrated, however, as it is in the case of groundwood, these factors appear to play a lesser role. The reasons for this difference are not fully understood and need to be further investigated. 3.4. Water Retention of Pulps. The water retention value (WRV) measurement is a standard empirical test that quantifies the total amount of water retained by a pulp sample after centrifugation under defined and controlled conditions.35 During the centrifugation, the free water present between the fibers and inside the cell lumen is removed and only water within the pore structure of the fiber cell wall is retained. The ratio of water to dry fiber after centrifugation is often used as a measure of fiber swelling capacity. Although deficiencies of the test have been broached in the past36 and its explicit physical meaning is unclear, the WRV is still accepted as a reliable and useful description of fiber swelling, particularly in cases where it is sufficient to assess the relative changes in fiber swelling.

2166

Biomacromolecules, Vol. 11, No. 8, 2010

Suchy et al. Table 1. Comparison of Moisture Content (MC) and Water Retention Values (WRV) of Pulp Samples Dried at Different Temperatures at 7% Relative Humidity groundwood

Figure 6. Comparison of water retention values of pulp samples dried under controlled H2O relative humidity (7%) at different temperatures (25, 60, and 80 °C). ND indicates never dried pulp samples.

It has been demonstrated that the properties of recycled (dried) chemical pulp fibers differ from those of fresh never dried fibers. This phenomenon, termed hornification, has been investigated extensively.33,37 In general, hornification refers to reduction of swelling capacity of the wood pulp fibers due to irreversible changes induced by drying. Hornification has been described as a complex process and several mechanisms and factors affecting its extent have been proposed. These include irreversible pore closure, microfibril aggregation, hydrogen bond formation and other mechanisms.37,38 It is mostly a characteristic of low-yield (chemical) pulps.33 Initial studies used the WRV reduction as a clear indicator of hornification. This method is still used in quantifying the extent of hornification in pulps. A comparison of the WRV measured for the pulp samples dried at different temperature is shown in Figure 6. Similarly, as observed in the pulp deuteration studies (Figure 3), the direct comparison clearly demonstrates the impact of drying and drying temperature on WRV reduction (Figure 6). In addition, the difference between the pulp samples of different characteristics is evident. The extent of hornification is most evident in the bleached pulp sample, which agrees with previous investigations.33 The bleached pulp contains the largest segment of cellulose and practically no lignin, compared to the unbleached pulp, which contains a small amount of residual lignin and a slightly greater relative amount of hemicellulose. Both lignin and hemicelluloses were shown to inhibit hornification.39 In groundwood, the lignin is present in significantly greater amounts, yet a reduction of WRV was evident. Although the majority of previous research accounts have described limited impact of drying on the swelling characteristics of high yield pulps,33 the reduction of mechanical pulps swelling ability, particularly after the first drying cycle, has also been previously described.40 All pulp samples dried at 25 °C already exhibited a significant reduction in WRV. Drying at higher temperatures resulted in even greater reduction. The reduction trend is quite steady with temperature for the groundwood sample. The groundwood pulp usually contains a large amount of fines, a fraction of fibers of colloidal dimensions. The fines can retain a considerable amount of water and, thus, could have an impact on overall accuracy of measured swelling. The groundwood sample used in this investigation was thoroughly washed and centrifuged. In addition, the sample was further washed during the deuteration/

unbleached kraft pulp

bleached kraft pulp

sample

MC (%)

WRV (%)

MC (%)

WRV (%)

MC (%)

WRV (%)

control dried at 25 °C dried at 60 °C dried at 80 °C

65.4 5.5 0.2 0.2

146 124 105 96

67.0 5.1 1.0 1.6

146 117 92 89

66.1 4.8 1.3 0.6

143 110 82 81

drying procedure. The raw data from WRV measurements can be used to give an indication of the amount of fines, which was found to be minimal: < 2% of the sample weight (see Supporting Information). Furthermore, it has been shown that the swelling of mechanical pulp fines is not affected by drying.41 Therefore, the small number of fines should not contribute to the dryinginduced alterations within the sample. The chemical pulps, both bleached and unbleached, showed a significant drop in WRV between samples dried at 25 and 60 °C and only a marginal additional decrease between samples dried at 60 and 80 °C (Figure 6). This is in agreement with the previous work of Laivins and Scallan who indicated that hornification is a direct result of the removal of water from the fiber, not the heat treatment associated with drying.33 This is further confirmed by the moisture content of the samples after drying (Table 1). The water remaining in the pulp samples after drying at both temperatures is similar; no substantial further change in moisture content was observed by increasing the drying temperature above 60 °C under the selected conditions. The extent of WRV reduction in the 60 to 80 °C temperature range increased relative to the amount of the lignin in the fibers, with the largest measured for the groundwood and only a marginal decrease observed for the lignin free bleached chemical pulp sample. It appears that drying at 80 °C possibly induces additional mechanisms of alteration, which are not directly related to water removal. These alterations may in some way be related to the presence of hemicellulose/lignin matrix in the groundwood sample and its residues in the unbleached chemical pulp sample, because no significant change took place in the lignin-free sample. Both WRV reduction and OD retention in the cell wall were used to indicate the changes in wood pulp fiber. The correlation between the pulp sample WRV reduction and retained OD band area in the respective IR spectra is shown in Figure 7. The correlation with the WRV of the pulps appears to be linear within the interval of measured values (Figure 7). Although the amount of retained OD groups in groundwood is lower, the incremental OD band area increase with WRV reduction is similar to that of chemical pulps. Because the increase in the OD band area correlates well with the reduction in WRV, which is a widely applied technical definition for an increased extent of hornification in the pulp, it can be safely assumed that hornification can be reliably indicated by detecting the retained inaccessible OD groups within the pulp sample. The wood samples followed the same trend, although to a lesser extent. Because hornification is often described by irreversible aggregation of cellulose microfibrils,18,42 the presence of lignin and hemicellulose in the matrix of native wood should theoretically prevent hornification related phenomena during drying of wood. However, the formation of inaccessible OD groups during drying clearly demonstrate that some kind of irreversible alteration, which, to a certain extent also takes place or proceeds through a similar mechanism in the cell wall of wood pulp fiber,

Cell Wall Alterations in Native Wood and Wood Pulp

Biomacromolecules, Vol. 11, No. 8, 2010

2167

fibers, it seems reasonable to hypothesize that irreversible aggregation of microfibrils during drying is also responsible for reduced accessibility in native wood matrix. A better understanding of mechanisms involved in ultrastructure alterations during initial drying of wood and cellulosic biomass and their subsequent impact on accessibility is vital in, for example, improving the efficiency of biomass conversion process. Acknowledgment. Prof. Mark Hughes and Petri Huhta (M.Sc.) are thanked for help with wood sampling. Emilia Kauppi is acknowledged for assistance in conducting deuteration experiments, FT-IR measurements, and data transfer. This work was part of projects supported by the Multidisciplinary Institute of Digitalization and Energy (MIDE, http://mide.tkk.fi). Figure 7. Correlation between water retention value reductions and OD band area for pulp samples dried under controlled relative humidity at different temperatures.

occurs in the wood ultrastructure. Reportedly, cellulose microfibrils form lamellar structures in the cell wall already during biosynthesis. It is to be expected that the microfibrils within the lamellae are so close to each other that they can aggregate upon drying. This hypothesis correlates with the findings of Elazzouzi-Hafraoui et al.17 who observed aggregated cellulose crystallites even after severe acid hydrolysis and suggested that already the biological origin promotes the lateral adhesion of cellulose microfibrils. Aggregation could also be the underlying reason on why mechanically separated nanofibrillar cellulose has larger dimensions than the native microfibril even in the case when wood, not pulp, is used as a starting material.4 In our previous paper, we hypothesized that a possible irreversible hardening of the lignin-hemicellulose matrix may also be responsible for accessibility changes during drying. However, the fact that the water retention values (WRV) of dried and never-dried pulps appear to correlate with the deuterium retention supports the hypothesis of microfibril aggregation in wood upon drying, because the drying-induced changes are ascribed to microfbril aggregation in the case of chemical pulps. Fundamentally, the nature and quantity of alterations induced by drying is an important discovery because irreversible changes in swelling characteristics of wood are impossible to track down with conventional measures in swelling. Although the results of this investigation do not present unequivocal hard evidence, they provide a strong indication of the mechanism of changes occurring in the cell wall of wood and wood fibers during initial drying.

4. Conclusions The ultrastructural alterations observed as reduction in accessibility of fresh wood and never dried pulp fibers of different lignin content after initial drying were compared. The drying induced inaccessibility was detected for all evaluated substrates, and although the extent of the reduction was significantly greater in wood pulp fibers and low lignin content pulps, in particular, similarities between the trends were evident. Supramolecular changes monitored with deuterium exchange coupled with FT-IR corresponded quantitatively to the water retention value reduction measured with a standard technical testing of pulps. Irreversible microfibril aggregation upon drying is widely believed to be the underlying reason behind the reduced accessibility of chemical pulps.18,42 In the light of the similar FT-IR trends for both native wood and isolated pulp

Supporting Information Available. Comparison of entire spectra from the deuteration reversibility testing of all wood and pulp samples (S1), comparison of entire spectra of wood and pulps dried at different temperatures (S2), comparison of entire spectra of wood and groundwood dried at different temperatures in two levels of D2O relative humidity environment (S3), and a change of sample weight measured during WRV evaluation of control pulp samples. This material is available free of charge via the Internet at http://pubs.acs.org.

References and Notes (1) Ragauskas, A. J.; Williams, C. K.; Davison, B. H.; Britovsek, G.; Cairney, J.; Eckert, C. A.; Frederick, W. J., Jr.; Hallett, J. P.; Leak, D. J.; Liotta, C. L.; Mielenz, J. R.; Murphy, R.; Templer, R.; Tschaplinski, T. Science 2006, 311, 484–489. (2) Stocker, M. Angew. Chem., Int. Ed. 2008, 47, 9200–9211. (3) Eichhorn, S. J.; Dufresne, A.; Aranguren, M.; Marcovich, N. E.; Capadona, J. R.; Rowan, S. J.; Weder, C.; Thielemans, W.; Toman, M.; Renneckar, S.; Gindl, W.; Veigel, S.; Keckes, J.; Yano, H.; Abe, K.; Nogi, M.; Nakagaito, A. N.; Mangalam, A.; Simonsen, J.; Benight, A. S.; Bismarck, A.; Berglund, L. A.; Peijs, T. J. Mater. Sci. 2010, 45, 1–33. (4) Abe, K.; Iwamoto, S.; Yano, H. Biomacromolecules 2007, 8, 3276– 3278. (5) Pa¨a¨kko¨, M.; Ankerfors, M.; Kosonen, H.; Nyka¨nen, A.; Ahola, S.; ¨ sterberg, M.; Ruokolainen, J.; Laine, J.; Larsson, P. T.; Ikkala, O.; O Lindstro¨m, T. Biomacromolecules 2007, 8, 1934–1941. (6) Azizi Samir, M. A. S.; Alloin, F.; Dufresne, A. Biomacromolecules 2005, 6, 612–626. (7) Kim, J.; Yun, S.; Ounaies, Z. Macromolecules 2006, 39, 4202–4206. (8) Mansfield, S. D.; Mooney, C.; Saddler, J. N. Biotechnol. Prog. 1999, 15, 804–816. (9) Hick, S. M.; Griebel, C.; Restrepo, D. T.; Truitt, J. H.; Buker, E. J.; Bylda, C.; Blair, R. G. Green Chem. 2010, 12, 468–474. (10) Rinaldi, R.; Palkovits, R.; Schu¨th, F. Angew. Chem., Int. Ed. 2008, 47, 8047–8050. (11) Binder, J. B.; Raines, R. T. J. Am. Chem. Soc. 2009, 131, 1979–1985. (12) Binder, J. B.; Raines, R. T. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 4516–4521. (13) Chang, V. S.; Holtzapple, M. T. Appl. Biochem. Biotechnol. 2000, 84-86, 5–37. (14) Iwamoto, S.; Abe, K.; Yano, H. Biomacromolecules 2008, 9, 1022– 1026. (15) Fahle´n, J.; Salme´n, L. Biomacromolecules 2005, 6, 433–438. (16) Fahle´n, J.; Salme´n, L. Holzforschung 2005, 59, 589–597. (17) Elazzouzi-Hafraoui, S.; Nishiyama, Y.; Putaux, J.-L.; Heux, L.; Dubreuil, F.; Rochas, C. Biomacromolecules 2008, 9, 57–65. (18) Hult, E.-L.; Larsson, P. T.; Iversen, T. Polymer 2001, 42, 3309–3314. (19) Nazhad, M. M.; Ramos, L. P.; Paszner, L.; Saddler, J. N. Enzyme Microb. Technol. 1995, 17, 68–74. (20) Suchy, M.; Virtanen, J.; Kontturi, E.; Vuorinen, T. Biomacromolecules 2010, 11, 515–520. (21) O’Brien, F. E. M. J. Sci. Inst. 1948, 25, 73–76. (22) Kou, Y.; Schmidt, S. J. Food Chem. 1999, 66, 253–255. (23) Hofstetter, K.; Hinterstoisser, B.; Salme´n, L. Cellulose 2006, 13, 131– 145. (24) Mann, J.; Marrinan, H. J. Trans. Faraday Soc. 1956, 52, 481–487. (25) Jeffries, R. Polymer 1963, 4, 375–389.

2168

Biomacromolecules, Vol. 11, No. 8, 2010

(26) Tsuchikawa, S.; Siesler, H. W. Appl. Spectrosc. 2003, 57, 667–674. (27) Tsuchikawa, S.; Siesler, H. W. Appl. Spectrosc. 2003, 57, 675–681. (28) Blechschmidt, J.; Heinemann, S. In Handbook of Pulp; Sixta, H., Ed.; Wiley-VCH: Weinheim, Germany, 2006; pp 1079-1111. (29) Sjo¨stro¨m, E. Wood Chemistry. Fundamentals and Applications, 2nd ed.; Academic Press: San Diego, CA, 1993. (30) Ale´n, R. Forest Products Chemistry. In Papermaking Science and Technology; Stenius, P., Ed.; Fapet Oy: Helsinki, Finland, 2000; Vol. 3, pp 62-64. (31) Li, K.; Tan, X.; Yan, D. Surf. Interface Anal. 2006, 38, 1328–1335. (32) Scallan, A. M.; Tigerstro¨m, A. C. J. Pulp Pap. Sci. 1992, 18, 188–193. (33) Laivins, G. V.; Scallan, A. M. Products of Papermaking. Transactions of the 10th Fundamental Research Symposium, Oxford, U.K., September, 1993, Pira International: United Kingdom, 1993; pp 1235-1260.

Suchy et al. (34) Kato, K. L.; Cameron, R. E. Cellulose 1999, 6, 23–40. (35) Jayme, G. Tappi J. 1958, 41, 180A–183A. (36) Maloney, T. C.; Laine, J. E.; Paulapuro, H. Tappi J. 1999, 82, 125– 127. (37) Nazhad, M. M.; Paszner, L. Tappi J. 1994, 77, 171–179. (38) Fernandes Diniz, J. M. B.; Gil, M. H.; Castro, J. A. A. M. Wood Sci. Technol. 2004, 37, 489–494. (39) Oksanen, T.; Buchert, J.; Viikari, L. Holzforschung 1997, 51, 355– 360. (40) Law, K. N.; Valade, J. L.; Quan, J. Tappi J. 1996, 79, 167–174. (41) Luukko, K.; Maloney, T. C. Cellulose 1999, 6, 123–135. (42) Newman, R. H. Cellulose 2004, 11, 45–52.

BM100547N