Dissolution of Cellulosic Fibers: Impact of Crystallinity and Fiber

Feb 2, 2018 - With the aim of informing the selection of biomass pretreatment options and to assist in interpreting experimental results from differen...
2 downloads 8 Views 4MB Size
Article Cite This: Biomacromolecules XXXX, XXX, XXX−XXX

pubs.acs.org/Biomac

Dissolution of Cellulosic Fibers: Impact of Crystallinity and Fiber Diameter Mohammad Ghasemi, Paschalis Alexandridis, and Marina Tsianou* Department of Chemical and Biological Engineering, University at Buffalo, The State University of New York (SUNY), Buffalo, New York 14260-4200, United States ABSTRACT: With the aim of informing the selection of biomass pretreatment options and to assist in interpreting experimental results from different biomass/solvent combinations, this study addresses the impact of crystallinity and size on the kinetics of semicrystalline cellulose fiber swelling and dissolution. To this end, a newly developed phenomenological model is utilized that accounts for the role of decrystallization and disentanglement as two rate-determinant steps in the cellulose dissolution process. Although fibers with lower crystallinity swell more and faster, the degree of crystallinity does not affect the dissolution rate. Fibers of smaller diameter swell more and become amorphous faster. When decrystallization is important, the solubility of thinner fibers increases more with a reduction in the crystallinity compared to the diameter. However, when the dissolution is controlled by chain disentanglement, or in the case of dissolution of fibers having larger diameters, milling the fibers to reduce the particle size could increase the solubility.

1. INTRODUCTION Cellulosic biomass (comprising cellulose, hemicelluloses, and lignin) provides a valuable resource for the sustainable production of functional polymers, specialty chemicals, and fuels on the basis of its widespread availability, low cost, and environmentally benign production.1−3 The compact packing of cellulose molecules, their crystalline structure, and extended noncovalent interactions render the cellulosic biomass recalcitrant to processing and constrain its efficient utilization.4−6 The disruption of the interaction network of crystalline cellulose by solvents can increase the accessibility of cellulose chains to reagents to improve the processing of this renewable resource.7−9 However, there are only a few solvent systems capable of directly dissolving cellulose, and they are effective under limited and strict operating conditions.10 Pretreatment of cellulose11,12 to reduce its degree of crystallinity, through methods either physical (e.g., mechanical comminution) or chemical (e.g., immersion in acids or bases), affects the solubility and reactivity of cellulose.13−15 For example, in the lithium chloride plus N,N-dimethylacetamide (LiCl/DMAc) solvent system, the pretreatment (i.e., mercerization) of cotton linter, which reduced the crystallinity from 0.8 to 0.7 (11% reduction), improved the dissolution of samples.16 In the same solvent system, it was shown that the reactivity and solubility of cellulose increased as the crystallinity decreased.17,18 Furthermore, it has been argued that LiCl/ DMAc dissolves sisal pulp better than cotton linters because the degree of crystallinity and the crystallite size of the sisal pulp (0.67 and 3.9 nm, respectively) are smaller than those of cotton linters (0.8 and 5.9 nm).19 Although such experimental evidence signifies the impact of cellulose crystallinity on its © XXXX American Chemical Society

dissolution, some controversial results have appeared in the literature. For instance, highly crystalline cotton linters were found to dissolve in LiCl/DMAc more rapidly than low crystallinity Kraft pulp.20 Furthermore, it is believed that, at least for rather low molecular weights (DP < 200), the crystalline and amorphous regions of cellulose dissolve at approximately the same rate in NaOH aqueous solution.21,22 It has also been argued that crystallinity is not the hampering factor in cellulose dissolution; otherwise, the complete transformation of crystalline into amorphous should be the solution, but this is not the case.23 The reduction of particle size through pretreatment increases the reactivity of cellulose in biorefinery processes such as enzymatic hydrolysis24−26 and anaerobic digestion27,28 due to an increase in the surface area and higher accessibility of cellulose to the reagents (e.g., enzymes). For example, a reduction in the average particle size of cellulose from 82 to 38 μm doubled the hydrolysis rate after 10 h reaction.29 Milling the biomass particles to decrease the size distribution from ROD. When the value of ROS, which decreases over time, reaches the value of ROD, then the fiber diameter

Table 1. Physical Properties for Cotton Fiber Swelling and Dissolution in [bmim]Cl parameter

symbol

initial av fiber diameter (μm)a initial av degree of cellulose crystallinitya cellulose degree of polymerizationa diffusion coefficient of disentangled cellulose chains in solution (m2/s)b diffusion coefficient of solvent in glassy cellulose (m2/s)b steepness of the glassy−rubbery transition of celluloseb diffusivity ratio of solvent in crystalline to amorphous domainsc solvent volume fraction at the surface of the swollen gel-like fiberb dissolution induction time (min)d ratio of boundary layer thickness to the fiber radiusc

value

D0 dc DP Dp

7.14 0.69 514 1.53 × 10−13

Dsa0

3 × 10−15

as

2

Dsc0/ Dsa0 ϕs,i

0.75

tind δ/R

2 0.05

0.1

a Given in the experimental work.36 bCalculated using appropriate equations.31 cOn the basis of the dissolution conditions.31 dConsidered to be the first instance at which the concentration of dissolved cellulose chains was experimentally detected.36

begins to decrease. Finally, after saturating the fiber with solvent, ROS = 0 and the fiber diameter decreases with a constant velocity, which is controlled by the effectiveness of solvent toward cellulose chain disentanglement (captured by the disentanglement rate, rdis).31

3. RESULTS AND DISCUSSION 3.1. Effect of Fiber Crystallinity. 3.1.1. On the Kinetics of Cellulose Swelling and Dissolution. The dissolution of cellulose involves the two interrelated phenomena of decrystallization and disentanglement, the balance of which determines the overall kinetics of dissolution. When the solvent ability to disentangle chains is lower than its capability in crystal disruption, then the dissolution is disentanglement controlled, whereas if the disentanglement rate is higher than the decrystallization rate, the dissolution process becomes decrystallization controlled.31 For example, the dissolution of cellulose fiber (with properties presented in Table 1) with a degree of crystallinity of 0.69 at rdis = 2.5 × 10−10 m/s is controlled by decrystallization when kdec < 9 × 10−4 1/s and by disentanglement when kdec > 9 × 10−4 1/s.44 The impact of the degree of crystallinity, dc, on the kinetics of fiber swelling and dissolution was examined here by varying dc in the range 0−0.8 for fixed rdis = 2.5 × 10−10 m/s and at two conditions (Figure 2): (i) kdec = 8 × 10−4 1/s, where the dissolution process is decrystallization-controlled, and (ii) at kdec = 2.5 × 10−3 1/s, where the dissolution process is disentanglement-controlled. Panels a and b in Figure 2 reveal that increasing the crystallinity results in an increase in the fiber decrystallization time (DcT) at both decrystallization constant values considered here. At kdec = 2.5 × 10−3 1/s, at all degrees of crystallinity values the fiber becomes completely amorphous much earlier than the total dissolution time (DcT < TDT) (refer to Figure 1.b for definition of DcT and TDT). This is because the solvent molecules are sufficiently efficient to destroy the cellulose crystalline network at all dc values (disentanglement control). Therefore, under these conditions, the reduction of fiber crystallinity does not have a considerable impact on the total dissolution time (TDT) (see Figure 2a). At kdec = 8 × 10−4 1/s, C

DOI: 10.1021/acs.biomac.7b01745 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 2. Effect of degree of crystallinity, dc, on (a and b) total dissolution time (TDT) and decrystallization time (DcT) of the fiber. (c and d) Variation of the fiber diameter, and (e and f) fraction of dissolved fiber plotted vs the dimensionless dissolution time. The decrystallization constant, kdec, values are 2.5 × 10−3 1/s (a, c, and e) and 8 × 10−4 1/s (b, d, and f).

condition (kdec = 8 × 10−4 1/s), a reduction of dc from 0.8 to 0.6 (by 25%) decreases TDT by 15%. The impact of crystallinity on the time evolutions of the dimensionless diameter (normalized to the initial value) at kdec values of 2.5 × 10−3 and 8 × 10−4 1/s is presented in Figure 2c and d, respectively. Because the fiber dissolves at different times for different crystallinity values, for better comparing the swelling rates, the time on the x-axis of these figures is presented as a dimensionless dissolution time (time/TDT). The results indicate that the degree of crystallinity affects the swelling rate and the maximum swelling of the fiber, but it does

when dc < 0.4 the amount of crystalline cellulose is low enough as to not affect the overall dissolution kinetics. The TDT values at dc < 0.4 are very similar (Figure 2b), and the dissolution is constrained by disentanglement. However, when dc > 0.4 at kdec = 8 × 10−4 1/s, the ability of the solvent to disrupt the crystalline structure is not sufficient to convert enough crystalline to amorphous domains; consequently, the fiber remains crystalline almost until the end of dissolution (DcT ≈ TDT), and the overall kinetics of the dissolution process are controlled by the decrystallization of cellulose. Under this D

DOI: 10.1021/acs.biomac.7b01745 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules not affect the rate of the fiber diameter reduction. Because the solvent diffusion in amorphous domains is easier than in crystalline domains,45 fibers with lower crystallinity swell faster and to a greater extent (Figure 2c). This modeling prediction is borne by experimental results, for example, mercerized cotton fibers, which have a lower degree of crystallinity and crystallite size, swelled more than the native fibers.37,46 When the dissolution is constrained by decrystallization for certain periods of time (e.g., at 0.1 < time/TDT < 0.2 for dc = 0.8 as shown in Figure 2d), the dissolution might be hampered because the crystalline domains need to be first decrystallized and then disentangled; consequently, the fiber at these time periods only swells with solvent but does not dissolve. When the dissolution is controlled by the cellulose chain disentanglement, fibers with different degrees of crystallinity dissolve almost linearly with time with the same rate: the dissolution rate is almost independent of the crystallinity (Figure 2e). In decrystallization-controlled conditions, the dissolution rate is also not affected by the degree of crystallinity, but the dissolution might be interrupted for some time intervals because of the lack of solvent ability in destroying the crystalline network of cellulose (see Figure 2f for dc > 0.4). 3.1.2. On Dissolution Time and Decrystallization Time. The effect of crystallinity on the kinetics of cellulose dissolution over a wide range of decrystallization constant values (kdec from 2.5 × 10 −5 to 2.5 × 10 −1 1/s) and at a constant disentanglement rate (rdis = 2.5 × 10−10 m/s) is presented in Figure 3a. The impact of crystallinity on the kinetics of cellulose dissolution becomes more important when the solvent is less effective in decrystallization. Figure 3b shows the ratio of the total dissolution time over the decrystallization time (TDT/DcT) as a function of the degree of crystallinity at different decrystallization constant values. The dimensionless TDT/DcT value allows the quantification of the impacts of decrystallization and disentanglement on the cellulose dissolution kinetics. TDT/DcT ≫ 1 means that the fiber becomes completely amorphous much earlier than the dissolution time, and this, the dissolution is controlled by the disentanglement. However, in the case when TDT/DcT ≈ 1 the fiber remains almost crystalline until the end, and thus, the dissolution is mainly controlled by decrystallization. The results indicate that, for kdec ≥ 2.5 × 10−3 1/s, the dissolution is controlled by the cellulose disentanglement for all degrees of crystallinity values considered here (dc > 0.1). However, for kdec < 2.5 × 10−3 1/s, the decrystallization becomes important. For example, at kdec = 8 × 10−4 1/s, increasing the cellulose crystallinity above dc = 0.4 results in a change of the controlling factor of dissolution kinetics from disentanglement to decrystallization. For kdec ≤ 2.5 × 10−4 1/s, the dissolution process is decrystallization controlled at all crystallinity values. 3.1.3. On the Change of the Controlling Factor of Dissolution Kinetics. For identifying the kdec values at which a reduction of fiber crystallinity can change the kinetics of cellulose dissolution from decrystallization control to disentanglement control, kdec has been varied in the range 2.5 × 10−4 to 2.5 × 10−3 1/s, and at each kdec value, the corresponding dc value has been determined at which the controlling factor in cellulose dissolution changes. The results (Figure 4) indicate that, at a constant rdis = 2.5 × 10−10 m/s, the kinetics of dissolution change due to the reduction of crystallinity only in a rather narrow range of kdec values between 2 × 10−4 and 1 × 10−3 1/s. If kdec is low and the cellulose crystallinity is high

Figure 3. Effect of degree of crystallinity on the (a) total dissolution time (TDT) and (b) ratio of total dissolution time (TDT) over decrystallization time (DcT) of the fiber at different decrystallization constant (kdec) values.

(below the line in Figure 4), then the dissolution process is controlled by decrystallization, whereas at high kdec and low crystallinity (above the line), disentanglement controls the kinetics of dissolution. To extend the above result regarding the impact of dc on the change of dissolution kinetics and cover a wide range of solution conditions, at each dc value we varied kdec (from 2.5 × 10−6 to 2.5 × 10−2 1/s) and rdis (from 2.5 × 10−12 to 2.5 × 10−9 m/s) and determined the corresponding rdis and kdec pairs at which the dissolution kinetics change. The results are plotted in Figure 5 in terms of the logarithm of rdis vs kdec. At each degree of crystallinity (dc) value (0.2, 0.4, 0.6, 0.8), the conditions where the dissolution kinetics change are represented by a given line. In the region above the line, the dissolution is controlled by disentanglement, whereas below the line, cellulose decrystallization is more important. Each line is representative of conditions where the ability of solvent toward chain disentanglement is approximately equal to the solvent capability to decrystallize cellulose: equality of disentanglement and decrystallization rates. Figure 5 reveals that, in all degree of crystallinity values, the equality lines have the same slope (equal to one), indicating that an increase in solvent quality (increasing kdec and rdis) has a similar impact on the dissolution kinetics of cellulose regardless of the initial degree of E

DOI: 10.1021/acs.biomac.7b01745 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

for a limited range of conditions (the area between the upper and lower lines in Figure 5). Accordingly, when a particular solvent (represented by specific kdec and rdis values) falls into the upper region (disentanglement control) and far from the upper line, a possible fiber treatment to reduce its degree of crystallinity would not have much impact on the overall kinetics of cellulose dissolution: the total dissolution time and the time evolution of fiber solubility remain constant (see Figure 2a and e). These results signify that, in experimental studies that have concluded that the crystallinity does not have much impact on dissolution,22,23,47,48 the processes were likely controlled by the disentanglement of cellulose chains, and such solvent systems would fall into the region above the lines in Figure 5. Examples of such behavior are the dissolution of cellulose in [bmim]Cl,36 in 1-ethyl-3-methylimidazolium acetate plus water,47 and in aqueous 8 wt % NaOH.22 However, the dissolution of cellulose in LiCl/DMAc was experimentally shown to be affected by crystallinity.16−19 This indicates that this solvent falls into the region where the dissolution is mainly controlled by decrystallization (below the lines in Figure 5). 3.2. Effect of Fiber Diameter. 3.2.1. On the Kinetics of Cellulose Swelling and Dissolution. The effect of fiber diameter, D, on the kinetics of cellulose dissolution was investigated by varying the fiber diameter by dividing the original (reference) fiber diameter (D0) by a factor of 2, 3, or 4, or by multiplying D0 by a factor of 2, 3, ... 8, while keeping all the other parameters constant (as presented in Table 1 and at d c = 0.69). Because fibers having different diameters decrystallize and dissolve at different times, the variations of the fraction of dissolved fiber and the fiber diameter are presented here in terms of a dimensionless dissolution time (time/TDT), and the evolution of the fraction of converted crystalline domains is plotted vs the dimensionless decrystallization time (time/DcT). The results are presented in Figure 6. The fraction of dissolved fiber increases approximately linearly with time for all D values considered (Figure 6a). The fiber diameter has a small impact on the dissolution rate: during the fiber swelling period, the dissolution rates of larger diameter fibers are slightly greater than those of smaller diameter fibers, whereas following swelling the thinner fibers have higher dissolution rates. Therefore, cellulose fibers with different diameters and degrees of crystallinity have approximately the same diameter reduction rate in a particular solvent (represented by a unique pair of rdis and kdec) (see Figures 6b, 2c, and 2d). Whereas the dissolution rate is mostly independent of the fiber diameter, the swelling rate is affected by the diameter (Figure 6b). The maximum swelling ratio (D/D0) of a fiber is controlled by the balance between solvent diffusion rate and fiber dissolution rate.43 Fibers having small diameters become saturated with solvent faster and swell more because the mass transfer resistance against the solvent is much smaller than that in the larger diameter fibers (Figure 6b). Because of the faster saturation of thinner fibers, more solvent molecules become accessible to the cellulose chains in such fibers. Therefore, a reduction of fiber diameter increases the decrystallization rate of cellulose due to the enhancement of solvent accessibility (Figure 6c). The reduction of biomass particle size would also increase the accessibility of reagents or enzymes and microorganisms (because of the increase of the overall surface area and decrease of mass transfer resistance against solvent diffusion) and would impact the reactivity of cellulose in enzymatic hydrolysis25,49 and anaerobic diges-

Figure 4. Solvent effectiveness in decrystallization as captured by kdec values at which a reduction of the fiber degree of crystallinity (dc) at a fixed disentanglement rate of rdis = 2.5 × 10−10 m/s, results in changing the cellulose dissolution kinetics from decrystallization control to disentanglement control. At conditions above the line, the dissolution process is mostly controlled by disentanglement of cellulose chains, whereas for conditions below the line, cellulose decrystallization controls the dissolution.

Figure 5. Effect of fiber initial degree of crystallinity on conditions (expressed in terms of decrystallization constant, kdec, and disentanglement rate, rdis, values) where the kinetics of cellulose dissolution are controlled by the decrystallization or the disentanglement step. In the region above the line, the dissolution process is mostly controlled by disentanglement of cellulose chains, whereas for conditions corresponding below the line, cellulose decrystallization is the more dominant step.

crystallinity. However, the slope of the equality line changes with fiber diameter (refer to the discussion on Figure 9). The results presented in Figure 5 indicate that crystallinity can change the controlling kinetics of dissolution from decrystallization-controlled to disentanglement-controlled only F

DOI: 10.1021/acs.biomac.7b01745 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

of red-oak sawdust (which has 39.7% cellulose) increased by 55% following particle size decrease from 590 to 33 μm.24 The yield of glucose increased 60% when the average particle size of microcrystalline cotton cellulose was reduced from 25.5 to 0.78 μm.26 3.2.2. On Dissolution Time and Decrystallization Time. The total dissolution time and decrystallization time decrease with a reduction in the fiber diameter (Figure 7). Therefore,

Figure 7. Effect of diameter (D) on the fiber total dissolution time (TDT) and decrystallization time (DcT). D0 denotes the reference fiber diameter (7.14 μm).

having a smaller fiber diameter leads to more rapid cellulose dissolution. Similar results have been obtained in the dissolution of amorphous polymers.50−52 According to the results shown in Figure 7, when the fiber diameter is reduced by half, TDT decreases by 53% and DcT decreases by 46%. Cellulose fibers with a relatively large diameter become decrystallized much earlier than their dissolution time (DcT < TDT), and consequently, their overall dissolution is controlled by the disentanglement of cellulose chains. As the fiber diameter decreases, the difference between TDT and DcT is also reduced, and finally, below a certain value, the fiber remains in the crystalline form almost until the end of the dissolution (DcT ≈ TDT) and the process becomes constrained by the decrystallization of cellulose. At the conditions where Figure 7 is obtained (kdec = 2.5 × 10−3 1/s and rdis = 2.5 × 10−10 1/s), the controlling kinetics of dissolution change for fibers with diameters smaller or greater than D0/3: for D ≤ D0/ 3, the process is constrained by decrystallization, whereas for D > D0/3, disentanglement controls the dissolution. 3.2.3. On the Change of the Controlling Factor of Dissolution Kinetics. As discussed in section 3.2.2, a reduction of fiber diameter may change the controlling kinetics of cellulose dissolution and can be potentially helpful to improve the extent and speed of cellulose dissolution. It is important to know up to which level and at what conditions the reduction of particle size would be more effective toward increasing the performance of dissolution. To clarify this issue, we have identified the ability of solvent to disentangle chains (rdis) at which the kinetics of cellulose dissolution change from disentanglement-controlled to decrystallization-controlled. This we did at different decrystallization constant values (kdec

Figure 6. Effect of fiber diameter on (a) fraction of dissolved fiber, (b) swelling of the fiber, and (c) decrystallization of the fiber plotted vs the dimensionless dissolution or decrystallization time. TDT and DcT are the total dissolution time and decrystallization time of the fiber, respectively.

tion.27,28 For example, it has been reported that the hydrolysis rate doubled when the average particle size of α-cellulose (produced from hardwood pulp) was reduced from 82 to 38 μm.29 Further, the production rate of glucose in the hydrolysis G

DOI: 10.1021/acs.biomac.7b01745 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules = 2.5 × 10−5, 2.5 × 10−4, and 2.5 × 10−3 1/s) for different fiber diameters. The results are shown in Figure 8. For the conditions above each line (greater disentanglement rate), the dissolution is

fiber DcT and TDT at each diameter for different dissolution conditions (kdec and rdis values). The results are summarized in Figure 9. Decreasing the fiber diameter moves the line

Figure 8. Solvent effectiveness in disentanglement as captured by rdis values at which a reduction of the fiber diameter, at fixed decrystallization constant values (kdec), results in changing the kinetics of cellulose dissolution from disentanglement to decrystallization control. In the region above each line, the dissolution process is mostly controlled by decrystallization of cellulose, whereas for conditions corresponding to below the line, cellulose disentanglement controls the dissolution. D0 denotes the reference fiber diameter (7.14 μm).

Figure 9. Determining factors of cellulose dissolution kinetics expressed in terms of the decrystallization constant (kdec) and the disentanglement rate (rdis) at different fiber diameters. In the region above the line, the dissolution process is mostly controlled by the disentanglement of cellulose chains, whereas for conditions corresponding to below the line, cellulose decrystallization is the more dominant step.

controlled by the decrystallization step, whereas for the conditions below each line (lower rdis), cellulose chain disentanglement controls the overall dissolution kinetics. Decreasing the fiber diameter (moving from the right to the left side of the plot) changes the kinetics of dissolution from disentanglement-controlled to decrystallization-controlled. However, the impact of diameter reduction in changing the dissolution kinetics for different solvent qualities (captured by kdec and rdis) is not the same. The results presented in Figure 8 signify that, for larger diameter fibers, a small change in the disentanglement rate varies the controlling factor of dissolution (a rdis vs diameter plot has a small slope); however, the dissolution kinetics of smaller diameter fibers are less sensitive to the disentanglement rate (rdis increases steeply with diameter). This indicates that the importance of the solvent ability to disentangle chains is higher in large diameter fibers than that in the case of small diameter fibers. Therefore, when the dissolution is controlled by chain disentanglement, the reduction of fiber diameter increases the solubility of cellulose until the kinetics of cellulose dissolution change (reaching the line in Figure 8); however, further decreasing the fiber diameter has a weaker effect on decreasing the TDT of cellulose (see Figure 7 for D < D0/3). To assess the impact of fiber diameter reduction on the determining factors of cellulose dissolution kinetics over a broad range of dissolution conditions, at each diameter we varied kdec and rdis values and determined the corresponding kdec and rdis pairs at which the kinetics of dissolution change from disentanglement-controlled to decrystallization-controlled. These data were determined by calculation of the

(indicating equality between decrystallization and disentanglement rates) toward the upper part of the plot. This means that the dissolution of smaller diameter fibers is constrained for a wider range of solution conditions because of the lack of solvent ability in decrystallization, and thus, decrystallization is more important in the case of thinner fibers. The controlling line of the larger diameter fibers becomes steeper by increasing the quality of the solvent (increasing kdec and rdis). This signifies that the sensitivity of dissolution kinetics on the disentanglement rate in larger diameter fibers is higher, and supports the result of Figure 8 that disentanglement is more important in larger diameters. Because the equality lines for different fiber diameters (in Figure 9) approach each other at high kdec and rdis values, the reduction of fiber diameter in a high quality solvent has a weaker effect on increasing cellulose solubility compared to that in a poor solvent. 3.3. Comparing the Impact of Reduction of Crystallinity or Fiber Diameter: Insights for Biomass Processing. The results presented in sections 3.1 and 3.2 provide insights that are useful in the processing of cellulosic biomass. Accordingly, when the dissolution is mainly constrained because of the lack of solvent ability in chain disentanglement, physical pretreatments such as particle size reduction can increase the solubility of cellulose, whereas pretreatments to decrease the crystallinity would not affect the solubility. Note that here the solubility (%) is defined as the mass percentage of dissolved fiber (i.e., mass of dissolved fiber/initial fiber mass × 100). H

DOI: 10.1021/acs.biomac.7b01745 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules For conditions when the dissolution is controlled by decrystallization, it is important to assess what is better to do to dissolve more cellulose: (i) reduce the fiber diameter or (ii) decrease the degree of crystallinity. To clarify this important point, we determined the solubility of cellulose fibers with different diameters (D/D0 ranges from 1 to 8) and degrees of crystallinity (dc ranges from 0.1 to 0.8) after 1 h immersion in hypothetical solvents with fixed disentanglement rate (rdis = 2.5 × 10−10 m/s) and different decrystallization constant values (kdec = 10−5, 10−4, and 10−3 1/s). The results are shown in Figure 10. When the solvent is not very active in decrystallizing cellulose (kdec = 10−5 and 10−4 1/s), a reduction of the fiber diameter or the degree of crystallinity below a certain level can result in increasing the solubility. Therefore, to increase the cellulose solubility, one can perform pretreatment or milling to reduce the degree of crystallinity or particle size, respectively, below a certain value. For example, at kdec = 10−5 1/s and initial degree of crystallinity of 0.6, milling the fibers until D/D0 = 2.5 has almost no impact on solubility, and D/D0 should be lower than 2.5 to increase the cellulose solubility. For fibers with D/ D0 = 4, for the cellulose solubility to increase, the degree of crystallinity should be decreased below 0.4 (Figure 10a). A relevant experimental result involves the dissolution of milled beech and spruce woods in LiCl/DMSO (6 wt % LiCl) where 1 h ball-milling of coarse particles resulted only in a partial dissolution; however, following 2 h milling, the solvent completely dissolved the samples.53 Similarly, the DMSO/Nmethylimidazole (v/v = 2/1) and DMSO/ethylenediamine (v/ v = 1/1) solvent systems completely dissolved the samples after only 5 and 10 h of milling, respectively.53 Furthermore, it was experimentally demonstrated that Norway spruce sawdust could be fully dissolved in the ionic liquid 1-allyl-3methylimidazolium chloride ([amim]Cl) after 48 h ball-milling under mild and acid-free conditions.54 In contrast, 20 min milling of softwood sulfite pulp, which produced smaller particles with no significant crystallinity, did not facilitate the dissolution of this cellulose in LiCl/DMAc (8 wt % LiCl).55 On the basis of the analysis presented here, this case was likely at conditions where reducing crystallinity and size does not affect the solubility (e.g., at kdec = 10−5 or 10−4 1/s for D/D0 > 7 in Figure 10a and b). The point at which the solubility becomes affected by a reduction of the fiber diameter or crystallinity is a function of the initial fiber diameter, initial degree of crystallinity, and solvent quality (which is captured in the model by kdec and rids). Increasing the solvent quality to decrystallize cellulose (which means that disentanglement becomes more important) decreases the range of the parameter space where the solubility is independent of fiber diameter. In the case of low kdec values (10−5 and 10−4 1/s) where decrystallization is important, the solubility of smaller diameter fibers increases significantly with a reduction in the degree of crystallinity. However, deceasing the crystallinity has a weaker effect on increasing the solubility of larger diameter fibers (see Figure 10a, at kdec = 10−5 1/s; when dc decreases from 0.8 to 0.1, the solubility of fiber with D/D0 = 1 increases from 7 to 69%, whereas it only increases from 5.8 to 6.6% for D/D0 = 8). Therefore, for increasing the solubility of highly crystalline, small diameter fibers, it is much better to do pretreatment compared to milling (for example, at kdec = 10−5 1/s, the solubility of fibers with initial D/D0 = 2 and dc = 0.8 increases from 6.4 to 7% and to 30% following a reduction by half of the diameter and crystallinity, respectively). An experimental

Figure 10. Cellulose solubility (%) plotted as a function of the degree of crystallinity (dc) and fiber diameter (D/D0) after 1 h in solvent systems with (a) kdec = 10−5 1/s, (b) kdec = 10−4 1/s, and (c) kdec = 10−3 1/s all at a fixed rdis = 2.5 × 10−10 m/s. The arrows in (a) indicate the reduction of the degree of crystallinity and fiber diameter. D0 is the reference fiber diameter (7.14 μm).

investigation of the dissolution of beech wood (40−80 mesh size) in LiCl/DMSO (8 wt % LiCl) with an ethylenediamine (EDA) pretreatment also revealed that pretreatment other than milling was a prerequisite for the dissolution of such highly crystalline cellulose.56 When the fibers have larger diameter and a lower degree of crystallinity, to dissolve more cellulose, it would be more efficient to mill the particles to reduce their size. For example, at kdec = 10−5 1/s and initial D/D0 = 6 and dc = 0.4, a reduction of the fiber diameter by half triples the cellulose solubility from 6 to 19%; however, pretreatment to decrease the I

DOI: 10.1021/acs.biomac.7b01745 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

from 6 to 30%; however, the solubility from 6% reaches only 7% following a reduction of the fiber diameter by half. • However, when the dissolution is controlled by chain disentanglement, or in the case of dissolution of larger fiber diameters, milling the fibers to reduce the particle size could increase the solubility more than reducing the degree of crystallinity. For example, at kdec = 10−3 1/s and rdis = 5 × 10−10 m/s. the solubility after 1 h of fiber with initial D/D0 = 4 and dc = 0.8 increases from 10 to 21% or to 14% following the reduction by half of the fiber diameter or crystallinity, respectively. This study provides novel results on conditions (captured through the solvent effectiveness toward decrystallization, kdec, and chain disentanglement, rdis) where a reduction of crystallinity (Figure 5) and fiber diameter (Figure 9) can (or cannot) improve the extent and speed of cellulose dissolution. These findings provide guidelines for an efficient pretreatment strategy (i.e., treatment to reduce crystallinity or milling to decrease particle size) to facilitate an enhanced conversion of cellulosic biomass. In the literature there are (i) systems that benefit from chemical pretreatment, such as pretreatment using ethanol and hydrochloric acid that increased solubility of a mixture of spruce and pine pulps up to 4% in a 7% NaOH 12% urea aqueous solvent,57 and EDA pretreatment that resulted in a full dissolution of beech wood without milling in LiCl/DMAc (8 wt % LiCl),56 (ii) systems in which particle size reduction is more favorable, e.g., where milling coarse biomass particles (from beech and spruce wood) increased the cellulose solubility in LiCl/DMSO (6 wt % LiCl), DMSO/N-methylimidazole (v/ v = 2/1), DMSO/EDA (v/v = 1/1),53 and [amim]Cl,54 and (iii) systems where both chemical pretreatment and milling are helpful, such as 5 h mechanical and 2 h chemical (enzyme) pretreatment of softwood sulphite pulp (which has 89.9% cellulose) prior to dissolution into aqueous NaOH/ZnO solution.58 The findings are useful in the preparation and treatment of cellulose fibers to produce filters and membranes,59,60 where changing fiber diameter via pretreatment affects the membrane permeability. The obtained insights are also relevant in utilizing cellulose fibers in the fabrication of functional materials and composites,1 where physical treatment to control porosity, crystallinity, and diameter of cellulose fibers affects the characteristics of the final products.61 The results are further applicable in the use of biodegradable/bioactive semicrystalline polymers (e.g., polyesters) to fabricate scaffolds for tissue engineering62 and water-soluble polymer matrixes for drug release (e.g., poly(ethylene oxide)63 and poly(vinyl alcohol)64), where the initial polymer crystallinity controls the mechanical properties, polymer degradation rate, and drug release rate.

crystallinity by half would only increase the solubility from 6 to 9% (Figure 10b). At high kdec values (10−3 1/s), the reduction of crystallinity has a low impact on cellulose solubility; however, the impact of diameter reduction becomes stronger (Figure 10c). The solubility of fiber with initial dc = 0.8 increases from 6 to 55% as a result of the reduction of fiber diameter from D/D0 = 8 to 1 at kdec = 10−3 1/s (compare to the increase from 6 to 7% at kdec = 10−4 1/s). Therefore, by increasing kdec, the impact of pretreatment (reducing the degree of crystallinity) gradually diminishes, and the effect of size reduction (milling) becomes more important.

4. CONCLUSIONS Pretreatment of cellulosic biomass, which involves a reduction of crystallinity and/or particle size, improves its susceptibility to dissolution or hydrolysis and thus renders this renewable resource more amenable for the production of valuable chemicals and fuels.11,12 We compare here the effect of crystallinity and fiber diameter on the kinetics of crystalline cellulose swelling and dissolution to provide insights for efficient processing of cellulosic biomass. To this end, we have employed a newly developed phenomenological model43 that accounts for the several steps (e.g., solvent-induced decrystallization and chain disentanglement) involved in the dissolution process of crystalline cellulose. The results obtained in this work demonstrate the following: • The impact that the reduction of crystallinity and fiber diameter have on the kinetics of cellulose dissolution depends on the following factors: the initial degree of crystallinity, the initial diameter of the fiber; the solvent quality (captured in the model by the decrystallization constant, kdec, and disentanglement rate, rdis, parameters), and the rate-determining step of dissolution (decrystallization or disentanglement). • When the dissolution is controlled by decrystallization, a reduction of crystallinity increases the dissolution rate and decreases the total dissolution time. For example, a 50% reduction in the degree of crystallinity (from 0.8 to 0.4) would result in a 42% decrease in the total dissolution time at kdec = 2.5 × 10−4 1/s, rdis = 2.5 × 10−10 m/s, and D0 = 7.14 μm. • When the cellulose dissolution is controlled by chain disentanglement, a reduction of the cellulose degree of crystallinity has a small impact on the overall dissolution kinetics. For example, a 50% reduction in the degree of crystallinity would result in a 5% decrease in the total dissolution time at kdec = 2.5 × 10−3 1/s, rdis = 2.5 × 10−10 m/s, and D0 = 7.14 μm. • A reduction in fiber diameter can increase the solubility, but the magnitude of this impact depends on the solvent quality: the impact of size reduction is higher in lower quality solvents (relatively low kdec and rdis values). For example, a 20% reduction in the fiber diameter (initial dc = 0.8 and D0 = 7.14 μm) would result in a 32% increase in cellulose solubility after 1 h at kdec = 10−4 1/s and rdis = 5 × 10−12 m/s (low solvent quality) but in only a 4% solubility increase at kdec = 10−2 1/s and rdis = 5 × 10−10 m/s (high solvent quality). • When the dissolution is controlled by decrystallization, the solubility of smaller diameter fibers increases more with a reduction in the degree of crystallinity compared with a reduction in the fiber diameter, signifying that pretreatment of fibers would be beneficial. For example, at kdec = 10−5 1/s and rdis = 5 × 10−10 m/s, a reduction of crystallinity by half increases the solubility of fiber (initial D/D0 = 2 and dc = 0.8) after 1 h



AUTHOR INFORMATION

Corresponding Author

*E-mail: mtsianou@buffalo.edu. ORCID

Paschalis Alexandridis: 0000-0001-6989-8232 Marina Tsianou: 0000-0003-3340-627X Notes

The authors declare no competing financial interest. J

DOI: 10.1021/acs.biomac.7b01745 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules



(20) Schult, T.; Hjerde, T.; Optun, O. I.; Kleppe, P. J.; Moe, S. Characterization of cellulose by SEC-MALLS. Cellulose 2002, 9, 149− 158. (21) Isogai, A.; Atalla, R. H. Dissolution of cellulose in aqueous NaOH solutions. Cellulose 1998, 5, 309−319. (22) Le Moigne, N.; Navard, P. Dissolution mechanisms of wood cellulose fibres in NaOH-water. Cellulose 2010, 17, 31−45. (23) Glasser, W. G.; Atalla, R. H.; Blackwell, J.; Brown, M. R., Jr.; Burchard, W.; French, A. D.; Klemm, D. O.; Nishiyama, Y. About the structure of cellulose: Debating the Lindman hypothesis. Cellulose 2012, 19, 589−598. (24) Dasari, R.; Berson, E. R. The effect of particle size on hydrolysis reaction rates and rheological properties in cellulosic slurries. Appl. Biochem. Biotechnol. 2007, 137, 289−299. (25) Chundawat, S. P. S.; Venkatesh, B.; Dale, B. E. Effect of particle size based separation of milled corn stover on AFEX pretreatment and enzymatic digestibility. Biotechnol. Bioeng. 2007, 96, 219−231. (26) Yeh, A.-I.; Huang, Y.-C.; Chen, S. H. Effect of particle size on the rate of enzymatic hydrolysis of cellulose. Carbohydr. Polym. 2010, 79, 192−199. (27) Mshandete, A.; Björnsson, L.; Kivaisi, A. K.; Rubindamayugi, M. S. T.; Mattiasson, B. Effect of particle size on biogas yield from sisal fibre waste. Renewable Energy 2006, 31, 2385−2392. (28) Angelidaki, I.; Ahring, B. K. Methods for increasing the biogas potential from the recalcitrant organic matter contained in manure. Water Sci. Technol. 2000, 41, 189−194. (29) Gan, Q.; Allen, S. J.; Taylor, G. Kinetic dynamics in heterogeneous enzymatic hydrolysis of cellulose: An overview, an experimental study and mathematical modelling. Process Biochem. 2003, 38, 1003−1018. (30) Düsterhölt, E.-M.; Engels, F. M.; Voragen, A. G. J. Parameters affecting the enzymic hydrolysis of oil-seed meals, lignocellulosic byproducts of the food industry. Bioresour. Technol. 1993, 44, 39−46. (31) Ghasemi, M.; Alexandridis, P.; Tsianou, M. Cellulose dissolution: Insights on the contributions of solvent-induced decrystallization and chain disentanglement. Cellulose 2017, 24, 571−590. (32) Wertz, J.-L.; Bédué, O.; Mercier, J. P. Cellulose science and technology; EPFL Press: Lausanne, Switzerland, 2010; p 350. (33) Klemm, D.; Philipp, B.; Heinze, T.; Heinze, U.; Wagenknecht, W. Comprehensive cellulose chemistry: Fundamentals and analytical methods; Wiley-VCH: Weinheim, Germany, 1998; Vol. 1, p 260. (34) Kang, X.; Kugaa, S.; Wang, L.; Wu, M.; Huang, Y. Dissociation of intra/inter-molecular hydrogen bonds of cellulose molecules in the dissolution processe: A mini review. Journal of Bioresources and Bioproducts 2016, 1, 58−63. (35) Wei, W.; Meng, F.; Cui, Y.; Jiang, M.; Zhou, Z. Room temperature dissolution of cellulose in tetra-butylammonium hydroxide aqueous solvent through adjustment of solvent amphiphilicity. Cellulose 2017, 24, 49−59. (36) Jiang, G.; Huang, W.; Wang, B.; Zhang, Y.; Wang, H. The changes of crystalline structure of cellulose during dissolution in 1butyl-3-methylimidazolium chloride. Cellulose 2012, 19, 679−685. (37) Fidale, L. C.; Ruiz, N.; Heinze, T.; El Seoud, O. A. Cellulose swelling by aprotic and protic solvents: What are the similarities and differences? Macromol. Chem. Phys. 2008, 209, 1240−1254. (38) Medronho, B.; Lindman, B. Brief overview on cellulose dissolution/regeneration interactions and mechanisms. Adv. Colloid Interface Sci. 2015, 222, 502−508. (39) Gubitosi, M.; Duarte, H.; Gentile, L.; Olsson, U.; Medronho, B. On cellulose dissolution and aggregation in aqueous tetrabutylammonium hydroxide. Biomacromolecules 2016, 17, 2873−2881. (40) Lin, L.; Yamaguchi, H.; Suzuki, A. Dissolution of cellulose in the mixed solvent of [bmim] Cl−DMAc and its application. RSC Adv. 2013, 3, 14379−14384. (41) Chaudemanche, C.; Navard, P. Swelling and dissolution mechanisms of regenerated lyocell cellulose fibers. Cellulose 2011, 18, 1−15.

ACKNOWLEDGMENTS We thank the U.S. National Science Foundation (Grant CBET1159981) for supporting research on cellulose dissolution in our laboratory.



REFERENCES

(1) Moon, R. J.; Martini, A.; Nairn, J.; Simonsen, J.; Youngblood, J. Cellulose nanomaterials review: Structure, properties and nanocomposites. Chem. Soc. Rev. 2011, 40, 3941−3994. (2) Langan, P.; Gnanakaran, S.; Rector, K. D.; Pawley, N.; Fox, D. T.; Cho, D. W.; Hammel, K. E. Exploring new strategies for cellulosic biofuels production. Energy Environ. Sci. 2011, 4, 3820−3833. (3) Kobayashi, H.; Fukuoka, A. Synthesis and utilisation of sugar compounds derived from lignocellulosic biomass. Green Chem. 2013, 15, 1740−1763. (4) Beckham, G. T.; Matthews, J. F.; Peters, B.; Bomble, Y. J.; Himmel, M. E.; Crowley, M. F. Molecular-level origins of biomass recalcitrance: Decrystallization free energies for four common cellulose polymorphs. J. Phys. Chem. B 2011, 115, 4118−4127. (5) Selig, M. J.; Tucker, M. P.; Sykes, R. W.; Reichel, K. L.; Brunecky, R.; Himmel, M. E.; Davis, M. F.; Decker, S. R. Lignocellulose recalcitrance screening by integrated high-throughput hydrothermal pretreatment and enzymatic saccharification. Ind. Biotechnol. 2010, 6, 104−111. (6) Pedersen, M.; Meyer, A. S. Lignocellulose pretreatment severity − relating pH to biomatrix opening. New Biotechnol. 2010, 27, 739− 750. (7) Kuo, C.-H.; Lee, C.-K. Enhancement of enzymatic saccharification of cellulose by cellulose dissolution pretreatments. Carbohydr. Polym. 2009, 77, 41−46. (8) Tadesse, H.; Luque, R. Advances on biomass pretreatment using ionic liquids: An overview. Energy Environ. Sci. 2011, 4, 3913−3929. (9) Mai, N. L.; Ha, S. H.; Koo, Y.-M. Efficient pretreatment of lignocellulose in ionic liquids/co-solvent for enzymatic hydrolysis enhancement into fermentable sugars. Process Biochem. 2014, 49, 1144−1151. (10) Liebert, T. Cellulose solvents - remarkable history, bright future. ACS Symp. Ser. 2010, 1033, 3−54. (11) Silveira, M. H. L.; Morais, A. R. C.; da Costa Lopes, A. M.; Olekszyszen, D. N.; Bogel-Łukasik, R.; Andreaus, J.; Pereira Ramos, L. Current pretreatment technologies for the development of cellulosic ethanol and biorefineries. ChemSusChem 2015, 8, 3366−3390. (12) Mosier, N.; Wyman, C.; Dale, B.; Elander, R.; Lee, Y. Y.; Holtzapple, M.; Ladisch, M. Features of promising technologies for pretreatment of lignocellulosic biomass. Bioresour. Technol. 2005, 96, 673−686. (13) Geng, W.; Jin, Y.; Jameel, H.; Park, S. Strategies to achieve highsolids enzymatic hydrolysis of dilute-acid pretreated corn stover. Bioresour. Technol. 2015, 187, 43−48. (14) Yang, B.; Wyman, C. E. Pretreatment: The key to unlocking low-cost cellulosic ethanol. Biofuels, Bioprod. Biorefin. 2008, 2, 26−40. (15) Gan, S.; Zakaria, S.; Chia, C. H.; Padzil, F. N. M.; Ng, P. Effect of hydrothermal pretreatment on solubility and formation of kenaf cellulose membrane and hydrogel. Carbohydr. Polym. 2015, 115, 62− 68. (16) Ramos, L. A.; Assaf, J. M.; El Seoud, O. A.; Frollini, E. Influence of the supramolecular structure and physicochemical properties of cellulose on its dissolution in a lithium chloride/N,N-dimethylacetamide solvent system. Biomacromolecules 2005, 6, 2638−2647. (17) Suzuki, K.; Kurata, S.; Ikeda, I. Homogeneous acetalization of cellulose in lithium chloride and dimethylacetamide. Polym. Int. 1992, 29, 1−6. (18) Buschle-Diller, G.; Zeronian, S. H. Enhancing the reactivity and strength of cotton fibers. J. Appl. Polym. Sci. 1992, 45, 967−979. (19) Ramos, L. A.; Morgado, D. L.; Gessner, F.; Frollini, E.; El Seoud, O. A. A physical organic chemistry approach to dissolution of cellulose: Effects of cellulose mercerization on its properties and on the kinetics of its decrystallization. ARKIVOC 2011, 7, 416−425. K

DOI: 10.1021/acs.biomac.7b01745 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules (42) Zhang, H.; Wu, J.; Zhang, J.; He, J. S. 1-allyl-3methylimidazolium chloride room temperature ionic liquid: A new and powerful nonderivatizing solvent for cellulose. Macromolecules 2005, 38, 8272−8277. (43) Ghasemi, M.; Singapati, A. Y.; Tsianou, M.; Alexandridis, P. Dissolution of semicrystalline polymer fibers: Numerical modeling and parametric analysis. AIChE J. 2017, 63, 1368−1383. (44) Ghasemi, M.; Tsianou, M.; Alexandridis, P. Assessment of solvents for cellulose dissolution. Bioresour. Technol. 2017, 228, 330− 338. (45) Hancock, B. C.; Zografi, G. Characteristics and significance of the amorphous state in pharmaceutical systems. J. Pharm. Sci. 1997, 86, 1−12. (46) El Seoud, O. A.; Fidale, L. C.; Ruiz, N.; D’Almeida, M. L. O.; Frollini, E. Cellulose swelling by protic solvents: Which properties of the biopolymer and the solvent matter? Cellulose 2008, 15, 371−392. (47) Parviainen, H.; Parviainen, A.; Virtanen, T.; Kilpeläinen, I.; Ahvenainen, P.; Serimaa, R.; Grönqvist, S.; Maloney, T.; Maunu, S. L. Dissolution enthalpies of cellulose in ionic liquids. Carbohydr. Polym. 2014, 113, 67−76. (48) Isogai, A.; Atalla, R. H. Amorphous celluloses stable in aqueous media: Regeneration from SO2−amine solvent systems. J. Polym. Sci., Part A: Polym. Chem. 1991, 29, 113−119. (49) Khullar, E.; Dien, B. S.; Rausch, K. D.; Tumbleson, M. E.; Singh, V. Effect of particle size on enzymatic hydrolysis of pretreated miscanthus. Ind. Crops Prod. 2013, 44, 11−17. (50) Devotta, I.; Ambeskar, V. D.; Mandhare, A. B.; Mashelkar, R. A. The life time of a dissolving polymeric particle. Chem. Eng. Sci. 1994, 49, 645−654. (51) Davey, M. J.; Landman, K. A.; McGuinness, M. J.; Jin, H. N. Mathematical modeling of rice cooking and dissolution in beer production. AIChE J. 2002, 48, 1811−1826. (52) Siepmann, J.; Kranz, H.; Peppas, N. A.; Bodmeier, R. Calculation of the required size and shape of hydroxypropyl methylcellulose matrices to achieve desired drug release profiles. Int. J. Pharm. 2000, 201, 151−164. (53) Wang, Z.; Yokoyama, T.; Chang, H.-m.; Matsumoto, Y. Dissolution of beech and spruce milled woods in LiCl/DMSO. J. Agric. Food Chem. 2009, 57, 6167−6170. (54) Leskinen, T.; King, A. W. T.; Kilpelainen, I.; Argyropoulos, D. S. Fractionation of lignocellulosic materials with ionic liquids. 1. Effect of mechanical treatment. Ind. Eng. Chem. Res. 2011, 50, 12349−12357. (55) Ishii, D.; Tatsumi, D.; Matsumoto, T. Effect of solvent exchange on the supramolecular structure, the molecular mobility and the dissolution behavior of cellulose in LiCl/DMAc. Carbohydr. Res. 2008, 343, 919−928. (56) Wang, Z.; Zhang, L.; Fan, Y.; Yang, Y.; Matsumoto, Y. Solubilization and fractionation of japanese beech wood with LiCl and DMSO. Journal of Bioresources and Bioproducts 2016, 1, 92−99. (57) Trygg, J.; Fardim, P. Enhancement of cellulose dissolution in water-based solvent via ethanol-hydrochloric acid pretreatment. Cellulose 2011, 18, 987−994. (58) Grönqvist, S.; Kamppuri, T.; Maloney, T.; Vehviläinen, M.; Liitiä, T.; Suurnäkki, A. Enhanced pre-treatment of cellulose pulp prior to dissolution into NaOH/ZnO. Cellulose 2015, 22, 3981−3990. (59) Varanasi, S.; Batchelor, W. J. Rapid preparation of cellulose nanofibre sheet. Cellulose 2013, 20, 211−215. (60) Lam, B.; Wei, M.; Zhu, L.; Luo, S.; Guo, R.; Morisato, A.; Alexandridis, P.; Lin, H. Cellulose triacetate doped with ionic liquids for membrane gas separation. Polymer 2016, 89, 1−11. (61) Park, S.; Venditti, R. A.; Abrecht, D. G.; Jameel, H.; Pawlak, J. J.; Lee, J. M. Surface and pore structure modification of cellulose fibers through cellulase treatment. J. Appl. Polym. Sci. 2007, 103, 3833−3839. (62) Rezwan, K.; Chen, Q. Z.; Blaker, J. J.; Boccaccini, A. R. Biodegradable and bioactive porous polymer/inorganic composite scaffolds for bone tissue engineering. Biomaterials 2006, 27, 3413− 3431. (63) Coutts-Lendon, C.; Koenig, J. L. Investigation of the aqueous dissolution of semicrystalline poly (ethylene oxide) using infrared

chemical imaging: The effects of molecular weight and crystallinity. Appl. Spectrosc. 2005, 59, 976−985. (64) Mora-Huertas, C.; Fessi, H.; Elaissari, A. Polymer-based nanocapsules for drug delivery. Int. J. Pharm. 2010, 385, 113−142.

L

DOI: 10.1021/acs.biomac.7b01745 Biomacromolecules XXXX, XXX, XXX−XXX