Cocontinuous Cellulose acetate/polyurethane composite nanofiber fabricated through electrospinning.

INTRODUCTION

Electrospinning is a simple and versatile technique in the generation of continuous ultrafine fibers with diameter down to submicrometers or even 10s of nanometers (1). Nanofiber possesses the characteristic features of high length-to-diameter ratio and specific surface areas, enabling it to be applied for protective clothing, filter, catalyst support, reinforced composite, and tissue engineering (2). Recently, there is growing interest in the design and preparation of novel composite nanofiber with improved properties, because it may combine the merits of each component. One type of composite nanofibers reported so far are prepared by multiple-jet electrospinning (3), (4), mainly because a cosolvent is hard to be found for each component. For instance, cellulose acetate (CA)/poly(vinyl alcohol) (PVA) composite nanofiber was generated by side-by-side electrospinning of CA in 2:1 acetone/dimethylacetamide and PVA in water (4). It was found that the mechanical properties of the CA nanofibrous mats are significantly enhanced with the increasing of PVA component in the CA/PVA composites (4). Nevertheless, this type of composite nanofibrous mat is composed of two kinds of nanofibers formed by individual pure polymers. Another type of composite nanofiber is synthesized from electrospinning of polymer blends in a cosolvent. Most composite nanofibers of this type consist of two components. One component is hard to be electrospun into nanofiber alone such as chitosan and Bombyx mori silk, and the other is a good nanofiber-forming polymer such as poly(ethylene oxide) (5-7). Generally, such nanofibers are of the "islands-in-the-sea" structure with the former dispersing into the latter, which is continuous fiber morphology due to phase separation and fast solvent evaporation during fiber formation. Selective removal of the dispersed component often results in nanoporous nanofibers with higher specific surface areas than untreated nanofibers, whereas the removal of the fiber-forming component would cause complete disintegration of the composite nanofiber [8, 9]. With special design of the spinneret, side-by-side (10) or core-sheath (11) bicomponent polymer nanofibers have been reported as well. Obviously, none of these bicomponent composite nanofibers presents a well-blended and cocontinuous fiber structure within a single nanofiber.

Cocontinuous composite nanofiber is defined as each component forms fiber structure in the composite. Comparing with the composite nanofibers from the side-by-side electrospinning, the cocontinuous structure enables the components in the nanofiber to be blended in the nanolevel, and to interact with each other through molecular forces such as hydrogen bonding. Such structure would certainly impart new mechanical and thermal properties for the composite nanofiber. However, little work on the cocontinuous composite nanofiber has been found so far (9). In this work, we aim to prepare cocontinuous composite nanofibers of CA/PU through electrospinning their blends in cosolvent N,N-dimethylacetamide (DMAc)/ acetone. Each component alone is readily to form nanofiber through electrospinning. CA nanofibers show poor mechanical property with stress of 1.3 MPa, but good dimensional stability with strain as small as 0.007% due to its semirigid chain structure (12), (13). Moreover, it can be easily deacetylated to regenerate more biocompatible and reactive cellulose nanofiber (14). PU nanofiber exhibits excellent tensile strength of over 10 MPa but high strain of more than 300% (15). Depending on the structure, PU nanofibers have been tested in the application areas of protective clothing (16) and tissue engineering (17). CA/PU cocontinuous composite nanofibers are possible to exhibit excellent mechanical strength and relative stable dimensionality. To test this assumption, the electrospinning of CA/PU is investigated in terms of composition ratio in the binary mixtures. The structure, morphology, mechanical behavior, and thermal property of the CA/PU composite nanofibers are examined. The structure of residual nanofibers obtained from selective removal of one of the components from the CA/PU composite nanofiber is characterized.

EXPERIMENTAL

Materials

Cellulose acetate ([M.sub.n] =3.0 X [10.sup.4], acetyl content 39.8 wt%) was from Aldrich Chemical Company. Elastollan B64D, a polyester-based polyurethane, was purchased from BASF Company. Both CA and PU can dissolve in DMAc and 2:1 (v/v) DMAc/acetone. To facilitate their capabilities in the fiber formation during electrospinning, the latter was chosen as the cosolvent for CA and PU. All materials were used as received.

Electrospinning

Precalculated amount of CA and PU was dissolved in 2:1 DMAc/acetone mixture solvent to make CA/PU composition ratio of 100/0, 80/20, 60/40, 40/60, 20/80, and 0/100. The total polymer weight concentrations were 20% by weight. It was observed that the miscibility of these two polymers in solutions is poor, because the blend solutions are not completely transparent. The CA/PU spinning solution was added to a 10-ml syringe attached to a stainless needle with an inner diameter of 0.84 mm. An electrode was clamped on the needle and connected to a power supply. Grounded counter electrode was connected to collector aluminum foil. All solutions were electrospun at ambient temperature, flowing rate at 10 [micro] 1/min, 10 kV, and a 20-cm distance between the collector and needle tip. The products were dried under vacuum at 50[degrees]C for 5 h to remove any residual solvents.

Selective Removal of One Component from

CA/PU Nanofiber

To remove PU component in the composite nanofiber, the CA/PU nanofibrous mat was first hydrolyzed in 0.05 M NaOH/ethanol solution to turn CA to cellulose at room temperature for 24 h. Because cellulose is hard to dissolve in many solvents, the hydrolyzed fibrous mats were subsequently immersed in DMAc for 24 h to dissolve PU component while cellulose nanofiber mat was left behind. Such obtained cellulose nanofibrous mat was further washed with DMAc three times, acetone two times to completely remove PU component. The selective removal of CA component was achieved by immersing CA/PU nanofibrous mat in acetone for 12 h. After it was washed with acetone five times, PU nanofiber was remained. The residual cellulose and PU nanofibrous mats were vacuum dried at 50[degrees]C for 10 h. The mass of dried nanofibrous mat before and after selective removal of CA or PU component was measured to calculate the weight loss. Scheme 1 illustrated the preparation and treatment procedure of CA/PU nanofiber.

Characterization

The fiber morphology was examined by scanning electron microscope (SEM, JEOL JSM-6380LV) and transmission electron microscope (TEM, JEOL JEM-2010). All samples were sputter-coated with gold before SEM observation. For TEM measurement, the fibers were dispersed ultrasonically in ethanol. The tensile strength of the fibrous mats was measured on Twin Column testing machine (LLOYD LR5K) with cross-head speed of 10 mm/min at room temperature. Eight replicates were tested for each sample, and their average value was given. ATR was recorded on Thermo-Nicolet 5700 spectrometer. The thermal properties of the nanofibers were measured on a differential scanning calorimeter (DSC TA Q200) under nitrogen atmosphere. Specimen was first heated to 120 [degrees]C at a heating rate of 20 [degrees]C/min to eliminate the thermal history. The second scan was recorded at a heating rate of 10[degrees]C/min from -70[degrees]C to 270[degrees] C.

RESULTS AND DISCUSSION

Morphologies and Structures of CA/PU Composite Nanofibers

The SEM images of CA/PU composite nanofibers are presented in Fig. 1. CA alone shows beads-on-string morphology (Fig. la), with bead size and fiber diameter in the range of 0.9-2.5 [micro]m and 0.14-0.3 [micro]m, respectively. This morphology is consistent with that found in a previous work (13). The beads formation is caused by insufficient CA chain entanglement in the solution with low polymer concentration (18). Although total polymer concentration was kept at 20 wt%, the addition of PU could facilitate the bead-free nanofiber formation (Fig. lb-e). In experiment, it was noticed that the addition of PU greatly increased the viscosity of the CA/PU blend-spinning solution, implying that the chain entanglement was enhanced in the solution. Such enhancement would favor the formation of less-bead or even no-bead nanofiber. Similar phenomenon was observed in the manufacture of PVC/PU blend nanofiber. Beaded PVC nanofiber became bead-free PVC/PU nanofiber with the addition of PU in the spinning solution while the total polymer concentration was kept constant at 13 wt% (19). CA/PU80/20 nanofiber was uniform in morphology and size of 360 nm (Fig. 1b). With further increasing of PU content to 80%, fiber size increased but with no significant variation from 700 nm for CA/PU 60/40, 40/60, and 20/80 (Fig. Ig). It is well known that solvent system is one of the key factors affecting the eiectrospinning process and the morphology of nanofiber. 2:1 DMAc/acetone is a poor solvent system for the nanofiber formation of CA solutions with concentration less than 25% (13). The current findings suggest that mixing PU with CA at actual CA concentration down to 12 wt% (Fig- Ic) could greatly facilitate the electro-spinning of CA in this solvent. Bead-free PU nanofiber with mean diameter of 900 nm was easily fabricated from 2:1 DMAc/acetone mixture solvent (Fig. If). Our work indicates that DMAC/acetone is also a feasible solvent for the preparation of PU nanofiber besides the widely used solvent N, N-dimethylformamide (DMF) or DMF/tetrahy-drofuran (THF) (15), (19). Meanwhile, it is interesting to find that PU fibers adhered together at some bonding sites (pointed by the arrows in Fig. If). Similar observations have been reported in the preparation of PVC/PU nanofibers (19). These bonding sites are particularly useful for the enhancement of mechanical properties for PU nanofi-brous membrane, which will be discussed in the following section. Because of high [T.sub.b] at 164[degrees]C, DMAc evaporates slowly in the short course of fiber traveling to the collector, resulting in some DMAc remain in the fiber. Given that the remaining DMAc is not removed in time, PU fibers are readily to be redissolved and fused together at some locations. In our research, freshly generated thick PU nanofibrous mat could transform into film whose morphology is similar to cast-film if it was put in a sealed container, while it could keep nanofibrous mat morphology if it was in an open environment.

Morphologies of CA/PU Composite Nanofthers After Selective Removal of One Component

To obtain the information on the internal morphology of composite fiber, that is, how each component distributes within the composite nanofiber, selective removal of one component followed by microscopical analysis is an established very effective technique (9). Up to now, two very common distribution morphologies, that is. "islands-in-the-sea" dispersion (8) and core-sheath (20) have been reported. We first analyzed the fiber structure of cellulose nanofiber after the selective removal of PU component. For CA/PU60/40 and 40/60 nanofibers hydrolyzed in 0.05 M NaOH/EtOH, the residual cellulose/PU fiber morphology was kept unchanged though average fiber sizes decreased dramatically from originally 650 nm and 710 nm to 370 nm and 460 nm, respectively (Fig. 2a and b). With the further removal of PU component by washing with DMAc, the morphologies of the resultant cellulose nanofibers were still preserved very well, but the mean diameter of nanofibers was further reduced to 280 and 320 nm, respectively (Fig. 2c and d). It should be pointed out that cellulose nanofibers with uniform diameter of less than 300 nm and defects (beads) free morphology are first ever synthesized in this work. Cellulose nanofibers from the deacetylation of CA nanofibers (12) or from direct electrospinning cellulose/DMAc-LiCI (21) or cellulose/ NMMO (22) often exhibit broad fiber size distribution and beads defect when average fiber diameter is down to 500 nm.

The mass losses of CA/PU nanofibrous mats after treatment were listed in Table 1. The actual mass losses after hydrolysis (data in column 4) were higher than those of theoretical acetyl content (data in column 2) in the composite nanofibers. However, actual PU loss after treatment of hydrolyzed mat with DMAc was less than that of theoretical PU content. Considering that the actual total loss closely matches theoretical loss, we believe that part of the polyester-based PU is hydrolyzed along with the deacetylation of CA in NaOH/EtOH solution, and the deacetylation of CA and PU removal are processed completely. In another parallel hydrolysis experiment in NaOH/[H.sub.2]O (values in the last two rows in Table 1), we found that the actual mass loss is a little bit more than that theoretical acetyl content after hydrolysis, with no significant hydrolytic degradation of PU. PU is almost removed in the subsequent treatment with DMAc. These results suggest that aq. NaOH has little effect on the hydrolysis of PU, which is believed to be caused by the poor penetration of aq. NaOH solution into the hydropho bic PU nanofiber. In a similar research, it is reported that alcoholic NaOH is more efficient in the hydrolysis of hydrophobic CA than aq. NaOH (12). FTIR results (not shown) confirm that acetyl groups and PU were completely removed, and cellulose nanofibers were obtained after such treatments.

TABLE 1. Mass changes of CA/PU nanofibrous membranes after hydrolyzed
in NaOH/EtOH or NaOH/[H.sub.2]O and washed in DMAc.

                                                  Mass loss after
          [C.sub.acetyl]  Theoretical [C.sub.pu]      NaOH/EtOH
CA/PU         (wt%)               (wt%)           hydrolysis (wt%)

80/20          32                  20                  39.4
60/40          24                  40                  40.9
40/60          16                  60                  35.1
20/80           8                  80                  36.0
60/40 (a)      24                  40                  25.4
40/60 (a)      16                  60                  18.4

            PU loss after     Theoretical total   Actual total
CA/PU     DMAc washing (wt%)   mass loss (wt%)   mass loss (wt%)

80/20             8.3                 52             47.7
60/40            23.3                 64             64.2
40/60            41.1                 76             76.2
20/80            50.0                 88             86.0
60/40(a)         35.8                 64             61.2
40/60(a)         55.3                 76             73.7

(a) The values in the last two rows were from hydrolysis in the
NaOH/[H.sub.2]O.

[FIGURE OMITTED1]

[FIGURE OMITTED1]

Continuous PU nanofibers were obtained after selective removal of CA in the composite nanofiber (Fig. 2e and f). Because of the loss of 80% CA for the acetone-treated CA/PU80/20 nanofibrous mat, the resultants are still in fibrous morphology but they tend to twinning together to form "crosslinked" fibrous film (Fig. 2e). For treated CA/ PU 40/60, the diameter of the residual PU fiber decreased from 710 nm to 660 nm after 40% CA was removed (Fig. 2f). This insignificant fiber size reduction should result from the swelling of PU in acetone. Upon removal of acetone while drying, swelled PU nanofiber tends to collapse and makes its size look bigger than round morphology. This phenomenon is also found for some hydrogel nanofibers when they are dried from swelled state (23), (24). It is also found that the residual PU fibers are distorted because of its flexible molecular chains and the loss of support from semirigid CA. The mass losses of CA/PU composite nanofibrous mats after acetone treatment are nearly same as the theoretical content of CA in the original bicomponent fibers (Table 2), suggesting the complete removal of CA.

TABLE 2. Mass changes of CA/PU nanofibrous membranes after washed with
acetone.

CA/PU  Theoretical CA content (wt%)  Actual mass loss (wt%)

80/20              80                      81.4
60/40              60                      59.3
40/60              40                      39.0

The smooth surface of CA/PU40/60 composite nanofiber was clearly demonstrated in TEM image (Fig. 3a). After the removal of PU, the surface of residual cellulose nanofiber becomes slightly coarse with shallow and elongated indents with width of several tens of nanometers were observed (Fig. 3b). Similarly, when CA was removed from the CA/PU bicomponent fibers, the remained PU fiber surface displayed shallow and elongated indents (Fig. 3c). The presence of indents, along with fiber size reduction after selective removal of one of the components suggested that CA and PU were elongated together along fiber axis in the electrospinning process, and each component formed continuous fiber structure within the binary composite nanofiber. As mentioned earlier, the miscibility of CA and PU is poor in solution so that microphase separation took place. Hence, each microphase separated component in the blend solution is elongated to form fiber structure simultaneously under electrical force, resulting in cocontinuous composite nanofiber. In the study of poly(vinylpyrrolidone) (PVP)/poly-L-lactide (PLLA) 1:1 composite nanofiber, the selective removal of component PVP not only greatly reduced the fiber size but also generated porous fiber surface topology. The authors concluded that this is a predominant cocontinuous composite nanofiber but some of PVP aggregated to form domains with dimensions in the order of 100-200 nm (9). Because the selective removal of any one of the two components in the CA/PU composite nanofiber did not give rise to noticeable pore structure on the residual fiber surface (Fig. 3b and c), we believe that the phase separation during fiber formation did not generate component domains with dimensions over 10s of nanometers in the CA/PU composite nanofiber, instead most part of each component was elongated by the electrical force, and was stabilized due to quick solvent evaporation to form nanofiber structure in the composite. For CA/PU blend cast-film in a wide range of PU content of 25-60 wt%, PU form domains dispersing in the continuous CA phase because of phase separation. During the drying of cast film, the slow solvent evaporation allows PU molecules reorganize and aggregate to form microphase in the CA continuous phase (25). Hence, in the electrospinning technique, the persistent electric drawing force and fast solvent evaporation rate make it possible that CA and PU form cocontinuous nanofiber structure in the composite fiber.

[FIGURE OMITTED3]

Mechanical Properties of CA/PU Nanofibrous Mats

[FIGURE OMITTED4]

The typical stress-strain relationship of CA/PU nanofibrous mats is shown in Fig. 4. Unlike smooth stress-strain curve for bulk materials such as cast film, characteristic seesaw curves are presented for nanofibrous mat due to the easily slipping and quick reorientation of short nanofiber (with length up to several centimeters (13)) in the loosely packed isotropic nanofibrous mat during stretching (15). The curves of those mats with high PU content such as CA/PU20/80 and 0/100 are in better shape than the others, mainly because the bonding sites among PU nanofibers keep them more integral than the loosely packed CA/PU60/40 mat (19). The mechanical properties of CA/PU nanofibrous mats were summarized in Table 3. CA nanofibrous mat presents very low [[sigma].sub.b] and [[member of].sub.b] of 0.98 M and 8%, respectively. With increasing of PU content from 20 to 60% in the composite, the [[sigma].sub.b] and E of corresponding composite nanofibrous mat increased gradually but kept in a narrow range of 2.8-3.2 MPa and 78-105 MPa, respectively. As for CA/PU20/80, which has similar fiber diameter to CA/PU 40/60, its [[sigma].sub.b] and E increased substantially to respective 7.46 MPa and 258 MPa, which is about 2.5 times that of CA/PU 40/60. Comparing with the mechanical properties of pure PU nanofibrous mat, the presence of 20% CA in the CA/PU20/80 significantly increased its E to more than 10 folds of CA/PU0/100, while without much loss of tensile strength. This result suggests the semirigid CA molecule chains enhance the rigidity of the CA/PU nanocomposite materials. For CA/PVA nanofibrous mats composed of two individual pure component nanofibers from side-by-side electrospinning, it has been reported that the presence of CA in the CA/PVA nanofibrous mats did no good to the mechanical properties of composite mat (4). Therefore, a cocontinuous fiber structure within a single nanofiber with two components blended in the nanolevel can largely improve its mechanical properties. The [sigma] of hydrolyzed CA/PU nanofibrous mat, that is, cellulose/PU60/40 and 20/80 got improved greatly comparing with that of corresponding untreated mat (Fig. 4 and Table 3). The deacetylation of CA in aq. NaOH regenerates cellulose, in which cellulose molecule chains tend to reorient along fiber axis to form monoclinic unit cell with crystallinity in the range of 20-60% (26). The strong inter-and intra-hydrogen bonding and existence of crystalline phase within cellulose make it possess superior mechanical properties to that of CA with same degree of polymerization. Hence its composite nanofibrous mat (cellulose/PU) shows better tensile strength than that of the CA-reinforced PU mat (CA/PU).

[FIGURE 5 OMITTED]

TABLE 3. Mechanical properties of CA/PU nanofibrous membranes.

CA/PU          [sigma] (MPa)      [[epsilon].sub.b]      E (MPa)
                                    ([degrees]C)

100/0         0.98 [+ or -] 0.09    8 [+ or -] 1      39 [+ or -] 7
80/20         2.80 [+ or -] 0.35   15 [+ or -] 1      78 [+ or -] 11
60/40         2.98 [+ or -] 0.32   23 [+ or -] 5      85 [+ or -] 22
40/60         3.20 [+ or -] 0.84   27 [+ or -] 3     105 [+ or -] 15
20/80         7.46 [+ or -] 0.64  128 [+ or -] 17    258 [+ or -] 9
0/100         8.98 [+ or -] 0.82  193 [+ or -] 26     24 [+ or -] 8
60/40-H (a)   8.52 [+ or -] 0.46   19 [+ or -] 1     412 [+ or -] 45
20/80-H (a)  12.63 [+ or -] 1.50  215 [+ or -] 28    273 [+ or -] 17

(a) Hydrolyzed CA/PU nanofibrous membrane in NaOH/[H.sub.2]O.

Thermal Properties of CA/PU Composite Nanofibers

DSC thermograms of CA/PU composite nanofibers are presented in Fig. 5. The [T.sub.g] of CA and PU appeared at 193.2[degrees]C (Fig. 5a) and -20.8[degrees]C (Fig. 5f), respectively. [T.sub.g] of CA was no longer noticeable in the thermograms of CA/PU composite fiber, because the glass transition history of CA overlapped with the melting of PU component (Fig. 5b and e). In general, [T.sub.g] of PU increased with the presence of CA in the composite nanofiber (Table 4), suggesting that the CA phase may partially miscible with PU at the interface. Interaction between CA and PU compo- nent through hydrogen bonding is readily formed among hydroxyl, carbonyl, and amide groups of CA and PU, which helps improve their miscibility at the interface, which is similar to that found in the CA/PU cast film (25). [T.sub.m] of CA at ~220[degrees]C did not vary very much for CA/PU composite nanofibers with various compositions, also the [DELTA] [H.sub.m-CA] was kept in a narrow range of 8.0-10.0 J/g. As for [T.sub.m] and [DELTA] [H.sub.m-PU] of composite nanofiber, it was found that they changed slightly when compared with that of pure PU nanofiber (Table 4). These findings suggest that the crystal pattern and crystallinity of CA and PU components do not alter very much in the composite nanofiber when compared with single component CA and PU nanofiber. Hence, in the CA/PU composite nanofibers, both components neither undergo strong interaction nor disperse into each other at the molecular level. Instead, each component forms separated microphase within the composite fiber. This is in well agreement with their nanomorphology as suggested by the SEM and TEM analysis.

TABLE 4. Thermal behaviors of CA/PU composite nanofibers.

        [T.sub.g]    [T.sub.m-pu]  [T.sub.m-CA]
CA/PU  ([degrees]C)  ([degrees]C)  ([degrees]C)

100/0     193.2             -         217.0
80/20     -15.5             -         220.2
60/40     -19.4         209.5         217.8
40/60     -14.9         207.4         220.6
20/80     -20.4         207.2         224.0
0/100     -20.8         206.7             -

       [DELTA][H.sub.m-pu]  [DELTA][H.sub.m-CA]
CA/PU     (J [g.sup.-1])    (J [g.sup.-1])

100/0           9.5
80/20          20.7                  7.8
60/40          24.4                 10.3
40/60          22.5                  8.4
20/80          19.3                  8.1
0/100          20.4                    -

CONCLUSION

Cocontinuous CA/PU nanofibers in a broad range of composition ratios were fabricated through electrospinning of microphase-separated CA/PU polymer blend solution. CA/PU composite nanofibers overcome the weak points of each component such as low tensile strength of CA, high strain and low Young's modulus of PU, while combine strong points to make its overall mechanical properties superior to each component. The selective removal of any one of the two components generated diameter-reduced residual cellulose and PU nanofibers. The residual nanofibers showed some elongated indents with width in the order of several 10s of nanometers. No pore structure was observed on residual fiber surface, indicating that the CA/PU nanofiber is a well-dispersed cocontinuous nanocomposite material.

REFERENCES

(1.) D. H. Reneker and I. Chun, Nanotechnology, 7, 216 (1996).

(2.) D. Li and Y. N. Xia, Adv. Mater., 16, 1151 (2004).

(3.) P. Gupta and G. L. Wilkes, Polymer, 44, 6353 (2003).

(4.) B. Ding, E. Kimura, T. Sato, S. Fujita, and S. Shiratori, Polymer, 45, 1895 (2004).

(5.) H.J. Jin, S. V. Fridrikh, G.C. Rutledge, and D. L. Kaplan, Biomacromolecules, 3, 1233 (2002).

(6.) K. E. Park, S. Y. Jung S.J. Lee, B.M. Min, and W.H. Park, Int. J. Biol. Macromol., 38, 165 (2006).

(7.) B. Duan, C.H. Dong, X.Y. Yuan, and K.D. Yao, J. Biomater, Sci. Polym., Ed., 15, 797 (2004).

(8.) L.F. Zhang and Y.L. Hsieh, Nanotechnology, 17, 4416 (2006).

(9.) M. Bognitzki, T. Frese, M. Steinhart, A. Greiner, J.H. Wendorff, A. Schaper, and M. Hellwig, Polym. Eng. Sci., 41, 982 (2001).

(10.) T. Lin, H.X. Wang, and X.G. Wang, Adv. Mater., 17, 2699 (2005).

(11.) M. Wang, J.H. Yu, D. Kaplan, and G. Rutledge, Macromolcules, 39, 1102 (2006).

(12.) H.Q. Liu and Y.L. Hsieh, J. Polym. Sci., Part B: Polym. Phys., 40, 2119 (2002).

(13.) H.Q. Liu and C.Y. Tang, Polym. J., 39, 65 (2007).

(14.) W.K.. Son, J.H. Youk, and W.H. Park, Biomacromolecules, 5, 197 (2004).

(15.) A, Pedicini and R.J. Farris, Polymer, 44, 6857 (2003).

(16.) P. Gibson, H. Schreuder-Gibson, and D. Rivin, Colloid Surf. A: Physicochem. Eng. Asp., 187, 469 (2001).

(17.) J.J. Stankus, J.J. Guan, and W.R. Wagner, J. Biomed. Mater. Res. A, 70, 603 (2004).

(18.) M.G. McKee, G.L. Wilkes, R.H. Colby, and T.E. Long, Macromolecules, 37, 1760 (2004).

(19). K.H. Lee, H.Y. Kim, Y.J. Ryu, K.W. Kim, and S.W. Choi, J. Polym. Sci., Part B: Polym. Phys., 41, 1256 (2003).

(20.) M. Wei, J. Lee, B.W. Kang, and J. Mead, Macromol. Rapid Commun., 26, 1127 (2005).

(21.) C.W. Kim, M.W. Frey, M. Marquez, and Y.L. Joo, J. Polym. Sci., Part B: Polym. Phys., 43, 1673 (2005).

(22.) P. Kulpinski, J. Appl. Polym. Sci., 98, 1855 (2005).

(23.) C.Y. Tang, S. H. Ye, and H. Q. Liu, Polymer, 48, 4482 (2007).

(24.) H.Q. Liu, M. Zheng, and R.H. Wu, Macromol. Chem. Phys., 208, 874 (2007).

(25.) Q. Zhou, L.N. Zhang, M. Zhang, B. Wang, and S.J. Wang, Polymer, 44, 1733 (2003).

(26.) R.D. Gilbert, Cellulosic Polymers, Blends, and Composites,, Hanser, New York (1994).

Correspondence to: Haiqing Liu: e-mail: [e-mail] Contract grant sponsor: Initiative Fund for the Returned Overseas Chinese Scholar administered by the State Education Ministry and the Key Project of Natural Science Foundation of Fuji an Province: contract grant number: E0620001.

DOI 10.1002/pen.2l090

Published online in Wiley InterScienee (www.interscience.vviley.com). [C] 2008 Society of Plastics Engineers

Chunyi Tang,(1) Pingping Chen,(1) Haiqing Liu (1), (2), (3)

(1) College of Chemistry and Materials Science, Fujian Normal University, Fuzhou 35007, China

(2) Fujian Key Laboratory of Polymer Materials, Fuzhou 350007, China

(3) Laboratory of Cellulose and Lignocellulosic Chemistry, Guangzhou Institute of Chemistry, Chinese Academy of Sciences, Guangzhou 510650, China