LiCl as a Solvent and

Jan 3, 2014 - Dual-Component System Dimethyl Sulfoxide/LiCl as a Solvent and Catalyst for Homogeneous Ring-Opening Grafted Polymerization of ε-Caprol...
0 downloads 9 Views 2MB Size
Article pubs.acs.org/JAFC

Dual-Component System Dimethyl Sulfoxide/LiCl as a Solvent and Catalyst for Homogeneous Ring-Opening Grafted Polymerization of ε‑Caprolactone onto Xylan Xue-Qin Zhang,† Ming-Jie Chen,† Chuan-Fu Liu,*,† and Run-Cang Sun†,‡ †

State Key Laboratory of Pulp and Paper Engineering, South China University of Technology, Guangzhou 510640, People’s Republic of China ‡ Institute of Biomass Chemistry and Technology, Beijing Forestry University, Beijing 100083, People’s Republic of China ABSTRACT: The preparation of xylan-graf t-poly(ε-caprolactone) (xylan-g-PCL) copolymers was investigated by homogeneous ring-opening polymerization (ROP) in a dual-component system containing Lewis base LiCl and strong polar aprotic solvent dimethyl sulfoxide (DMSO). DMSO/LiCl acted as solvent, base, and catalyst for the ROP reaction. The effects of the parameters, including the reaction temperature, molar ratio of ε-caprolactone (ε-CL) to anhydroxylose units (AXU) in xylan, and reaction time, on the degree of substitution (DS) and weight percent of PCL side chain (WPCL) were investigated. The results showed that xylan-g-PCL copolymers with low DS in the range of 0.03−0.39 were obtained under the given conditions. The Fourier transform infrared spectroscopy (FTIR), 1H nuclear magnetic resonance (NMR), 13C NMR, 1H−1H correlation spectroscopy (COSY), and 1H−13C correlation two-dimensional (2D) NMR [heteronuclear single-quantum coherence (HSQC)] characterization provided more evidence of the attachment of side chains onto xylan. Only one ε-CL was confirmed to be attached onto xylan with each side chain. Integration of resonances assigned to the substituted C2 and C3 in the HSQC spectrum also indicated 69.23 and 30.77% of PCL side chains attached to AXU at C3 and C2 positions, respectively. Although the attachment of PCL onto xylan led to the decreased thermal stability of xylan, the loss of unrecovered xylan fractions with low molecular weight because of the high solubility of xylan in DMSO/LiCl resulted in the increased thermal stability of the samples. This kind of xylan derivative has potential application in environmentally friendly and biodegradable materials considering the good biodegradability of xylan and PCL. KEYWORDS: xylan, ε-caprolactone, DMSO/LiCl, ring-opening polymerization



INTRODUCTION Hemicelluloses, the second most abundant class of renewable and biodegradable polysaccharides found in nature after cellulose, account for on average about 20−35% of most plant materials.1,2 Because of the easy availability, good biodegradability, biocompatibility, and renewability, low price, and good mechanical properties, much attention has been paid to the use of hemicelluloses to produce performance biomaterials, biofuels, and platform biochemicals with increased environmental awareness and the development of green chemistry.3,4 Xylan-type hemicelluloses are the main hemicellulosic components of cell walls of hardwoods and grasses. As heterogeneous polymers with low molecular weight, they possess backbone chains of β-(1 → 4)-D-xylopyranose units and are branched by short carbohydrate chains.5 Besides xylose, xylans may also contain arabinose, glucuronic acid or its 4-O-methyl ether, and acetic, ferulic, and p-coumaric acids. Usually, mostly the α-L-arabinofuranosyl residue may attach to some of the C3 positions of the xylan backbone as the principal substituent and also carry minor amounts of 4-O-methlglucuronic or α-1,2-linked glucuronic acid residues, which are mainly linked to the C2 positions of the xylan backbone.6 In addition, there can be O-acetyl substitution on the backbone.1 Recently, much attention has been focused on the use of lignocellulosic biomass and its components.7 In comparison to © 2014 American Chemical Society

cellulose and lignin, the exploiting of hemicelluloses was paid little attention until the last decade. Because of their inherent low molecular weight and heterogeneous structure, chemical modification of hemicelluloses becomes an important way to prepare material with the desired properties.8,9 In comparison to heterogeneous functionalization, homogeneous functionalization is simple and the degree of substitution (DS) of the products can be easily controlled and reproduced.10−12 Because of the excellent biodegradability, biocompatibility, and permeability, considerable efforts have been paid to aliphatic polyesters,13 among which poly(ε-caprolactone) (PCL) is one of the most important polyesters. PCL is a tough, flexible, and crystalline polymer with a low glass transition temperature (about −60 °C) and melting point (60 °C).14−17 Ring-opening polymerization (ROP) of ε-caprolactone (ε-CL) is the most common way to prepare PCL, which can be accomplished in melt or in solution, catalyzed with 4-dimethylaminopyridine (DMAP), stannous octoate [Sn(Oct)2], LiCl, or other catalysts.18−20 However, because of the cytotoxicity and difficulties in the removal of the catalyst from the resulting polymers, the use of Sn(Oct)2 and DMAP has been limited in many cases. LiCl, a very economic and Received: Revised: Accepted: Published: 682

August 14, 2013 January 3, 2014 January 3, 2014 January 3, 2014 dx.doi.org/10.1021/jf4036047 | J. Agric. Food Chem. 2014, 62, 682−690

Journal of Agricultural and Food Chemistry

Article

almost nontoxic chemical, could effectively catalyze the ROP reaction.19 The novel dissolution systems for lignocelluloses and their components received much attention because of the complexity of the plant cell walls. Dimethyl sulfoxide (DMSO)/LiCl was proposed as a novel dual-component system containing a Lewis base and strong polar aprotic solvent for the dissolution and pretreatment of plant materials.21,22 In the present study, the DMSO/LiCl system was investigated as the reaction medium for the ROP reaction of ε-CL onto xylan. The physicochemical properties of xylan derivatives were characterized with Fourier transform infrared spectroscopy (FTIR), 1H nuclear magnetic resonance (NMR), 13C NMR, 1H−1H correlation spectroscopy (COSY), 1H−13C correlation two-dimensional (2D) NMR [heteronuclear single-quantum coherence (HSQC)], scanning electron microscopy (SEM), and thermal analysis.



Scheme 1. Grafting Copolymerization of PCL onto the Xylan Backbone in DMSO/LiCl

MATERIALS AND METHODS

Materials. Xylan with a xylose content of over 85% was purchased from Yuan-Ye Biological Technology Co., Ltd. (Shanghai, China). LiCl with a purity of 97% was provided by Kermel Chemical Reagent Co., Ltd. (Tianjin, China). DMSO with 99.5% purity and ε-CL with 99% purity were supplied by Aladdin Reagent Co. (Shanghai, China). All chemicals were of analytical reagent grade and directly used without further purification. Preparation of Xylan-g-PCL Copolymers in DMSO/LiCl. A series of xylan-g-PCL copolymers were prepared in dual-component system DMSO/LiCl under the conditions in Table 1. Dry xylan (0.33 g,

Table 1. Detailed Structural Factors of Xylan-g-PCL Copolymers Obtained under Various Conditions in DMSO/LiCl sample

temperature (°C)

time (h)

ε-CL/AXU

DS

WPCL (%)

1 2 3 4 5 6 7 8 9 10 11 12 13

80 90 100 110 120 100 100 100 100 100 100 100 100

24 24 24 24 24 24 24 24 24 3 6 12 48

4:1 4:1 4:1 4:1 4:1 1:1 8:1 12:1 20:1 8:1 8:1 8:1 8:1

0.04 0.20 0.18 0.09 0.04 0.21 0.23 0.39 0.03 0.07 0.15 0.21

3.34 14.73 13.45 7.21 3.34 15.35 16.57 25.20 2.53 5.70 11.47 15.35

Figure 1. FTIR spectra of unmodified xylan and xylan-g-PCL copolymers obtained with an increased reaction temperature (A) from 90 °C (sample 2) to 100 °C (sample 3) and 110 °C (sample 4) and increased molar ratio of ε-CL to AXU (B) from 1:1 (sample 6), 4:1 (sample 3), 8:1 (sample 7), 12:1 (sample 8), and 20:1 (sample 9). (Lorentzen and Wettre; precision = 1 μm) at five different locations, and the mean value was used for the measurements of stress strength, tensile strain, and elasticity modulus on a universal tensile machine (DRK101). Detailed Physicochemical Characterization. FTIR spectra of the unmodified xylan and xylan-g-PCL copolymers were recorded on a Bruker spectrophotometer (TENSOR 27, Switzerland) from finely ground samples (1%) in KBr pellets in the range of 4000−500 cm−1. The 1H NMR, 13C NMR, 1H−1H COSY, and HSQC spectra of unmodified xylan and xylan-g-PCL copolymers were recorded from 40 mg samples in 0.5 mL of DMSO-d6 on a Bruker Avance III 400 M spectrometer (Germany) with a 5 mm multinuclear probe. For the 1H NMR analysis, the detailed collecting and processing parameters were listed as follows: number of scans, 11; receiver gain, 62; acquisition time, 4.0894 s; relaxation delay, 1.0 s; pulse width, 11.0 s; spectrometer frequency, 400.13 MHz; and spectral width, 8012.8 Hz. For the 13C NMR analysis, the detailed collecting and processing parameters were listed as follows: number of scans, 1932; receiver gain, 1440; acquisition time, 1.2583 s; relaxation delay, 1.5 s; pulse width, 12.0 s; spectrometer frequency, 100.61 MHz; and spectral width, 26 041.7 Hz. For the 1 H−1H COSY analysis, the detailed collecting and processing parameters were listed as follows: number of scans, 16; receiver gain, 355; acquisition time, 0.4588 s; relaxation delay, 2.0 s; pulse width, 8.5 s;

0.005 mol of hydroxyl group in xylan) was added to 20 mL of DMSO/ LiCl (6 wt % LiCl) in a 50 mL dried three-neck flask. The mixture was agitated with magnetic stirring at 80 °C for about 2 h under the protection of nitrogen to achieve a homogeneous solution. Then, the required quantities of ε-CL [the molar ratio of ε-CL to anhydroxylose units (AXU) in xylan were 1:1, 4:1, 8:1, 12:1, and 20:1] were added at 80, 90, 100, 110, and 120 °C, respectively. The reaction was carried out under the protection of nitrogen with magnetic stirring for 3, 6, 12, 24, and 48 h, respectively. After the required time, the resulting solution was cooled to room temperature. The solid residues after precipitation into 150 mL of ethanol were filtered out, washed thoroughly with ethanol and dichloromethane to eliminate DMSO, LiCl, ε-CL, and byproducts, and then dried in vacuum for 48 h at 50 °C. Film Preparation and Physical Property Determination. The xylan-g-PCL copolymer sample (0.3 g) was dissolved in 10 mL of deionized water. The solution was poured onto a polystyrene dish (9 × 9 cm) and allowed to dry at ambient temperature to obtain xylan-gPCL film. The film thickness was measured using a micrometer 683

dx.doi.org/10.1021/jf4036047 | J. Agric. Food Chem. 2014, 62, 682−690

Journal of Agricultural and Food Chemistry

Article

Figure 2. 1H NMR spectra of unmodified xylan (A) and xylan-g-PCL copolymer sample 9 (B).



spectrometer frequency, 400.13/400.13 MHz; and spectral width, 2231.1/2231.1 Hz. For the 1H−13C HSQC analysis, the detailed collecting and processing parameters were listed as follows: number of scans, 20; receiver gain, 879; acquisition time, 0.1645 s; relaxation delay, 2.0 s; pulse width, 8.5 s; spectrometer frequency, 400.13/100.61 MHz; and spectral width, 3113.3/22 727.3 Hz. The detailed structural factors of xylan-g-PCL copolymers, including DS and weight percent of PCL side chains (WPCL), were calculated from the integral area of the resonances assigned to the characteristic protons from AXU (H1) and the attached PCL (the methylene protons). The thermal stability of the samples was performed using thermogravimetric analysis (TGA) and derivative thermogravimetry (DTG) using a Q500 theromgravimetric analyzer (TA Instruments, New Castle, DE). The apparatus was continually flushed with nitrogen. The sample between 9 and 11 mg was heated from 30 to 600 °C at a heating rate of 10 °C/min. The surface morphology was examined by SEM on a field emission scanning electron microscope (LEO 1530 VP, LEO, Germany). The samples for SEM were prepared by casting a few solids onto a mica sheet followed by gold plating.

RESULTS AND DISCUSSION

Effects of Reaction Conditions on the DS of Xylan-gPCL Copolymers. Xylan-g-PCL copolymers were synthesized via homogeneous ROP of ε-CL onto xylan in DMSO/LiCl. Scheme 1 illustrates the ROP reaction of ε-CL with the hydroxyl groups of AXU in xylan in DMSO/LiCl. The preparation conditions and the detailed structural factors of xylan-g-PCL copolymers are shown in Table 1. The ROP reaction was easily achieved in DMSO/LiCl at 100 °C, and the xylan derivatives with DS of 0.20 were obtained within 24 h. However, the reaction occurred much more difficultly at higher or lower temperatures and even did not occur at 80 °C. The optimum reaction temperature for the ROP reaction in DMSO/LiCl was 100 °C. Clearly, the increase of the derivatizing reagent dosage and reaction time had favorable effects on the ROP reaction because of the increased molecule collisions of ε-CL with reactive hydroxyl groups. An enhancement of the molar ratio of ε-CL to AXU in xylan from 1:1 to 20:1 resulted in an increase in DS from 684

dx.doi.org/10.1021/jf4036047 | J. Agric. Food Chem. 2014, 62, 682−690

Journal of Agricultural and Food Chemistry

Article

Figure 3. 1H−1H COSY spectrum of xylan-g-PCL copolymer sample 9.

protons at H1, H2, H3, H4, H5e, and H5a of AXU, respectively.26 In comparison, spectrum B provided evidence of the attachment of PCL side chains onto xylan. The methylene proton signals of the attached PCL could be observed at 3.36 ppm (−CH2OH, e, end unit), 2.24 ppm (−COCH2−, a), 1.50 ppm (−CH2−, b), 1.38 ppm (−CH2−, d), and 1.28 ppm (−CH2−, c).18 The signal for the hydroxyl group at the end of PCL side chains (−CH2OH, e−OH) after the ROP reaction was present at 4.36 ppm. Unexpectedly, there was no signal for the methylene proton at the e position in the repeating unit of PCL (−CH2−O−, e), indicating probably only one ε-CL attached onto xylan with each side chain. As expected, no carboxyl proton was found in spectrum B, suggesting the absence of homopolymers of ε-CL and the formation of xylan−PCL ester. To confirm the correct assignment of the primary signals, the 1H−1H COSY spectrum of xylan-g-PCL copolymer sample 9 was collected, as illustrated in Figure 3. To avoid the overlapping of the primary signals, the COSY spectrum is shown at higher contour levels to provide the well-resolved correlations. As shown in Figure 3, the moderate cross-correlations in the lower left corner for H1/H2, H2/H3, H3/H4, and H5e/H5a were clearly observed. More importantly, the strong cross-correlations for PCL side chains in the upper right corner for a/b, b/c, c/d, and d/e and that in the lower left corner for e/e−OH provided the primary proton signals in PCL. Beside the cross-correlation at δH/δH of 3.36/1.38 ppm, there were no other cross-correlations of the methylene proton at the d position, indicating the absence of the PCL repeating unit. This result was consistent with 1H NMR analysis, confirming that only one ε-CL was attached onto xylan with each side chain; that is, the degree of polymerization (DP) of the side chain was only 1. On the basis of the typical proton signals from PCL side chains and AXU in xylan, the DS and weight percent of PCL side chains in Table 1 were calculated with the integral area of the resonances of the corresponding protons according to the following equations:

0.04 to 0.39, and an increasing reaction time from 3 to 48 h led to an improvement of DS from 0.03 to 0.21. The similar results of the ROP reaction of ε-CL onto cellulose were reported in ionic liquid 1-N-butyl-3-methylimidazolium chloride ([Bmim]Cl) with DMAP and Sn(Oct)2 as catalysts at 130 °C.18 The higher DS of cellulose-g-PCL (0.09−2.41) was probably due to the higher temperature in [Bmim]Cl and more hydroxyl groups available in cellulose. Because the ROP reaction of ε-CL onto xylan is acid- or base-catalyzed, in the present study, the favorable effect of Lewis base LiCl in strong polar aprotic solvent DMSO resulted in the occurrence of the ROP reaction of ε-CL with the hydroxyl groups in xylan at relatively low temperatures. This dual-component system acted as a solvent, base, and catalyst in the preparation of xylan-g-PCL copolymers. Similarly, the dissolution of the plant whole cell wall was reported in this system, because of the favorable effects on the destruction of hydrogen bonds extensively present in the cell wall.23,24 FTIR. Figure 1 illustrated the FTIR spectra of unmodified xylan and xylan-g-PCL copolymers obtained with an increased reaction temperature (A) and molar ratio of ε-CL to AXU (B). The absorbances at 3443, 2922, 1731, 1629, 1465, 1036, and 892 cm−1 in the spectrum of unmodified xylan are previously reported. 23 In comparison, the spectra of xylan-g-PCL copolymers provided evidence for the ROP reaction by showing changes of two important bands at 2867 cm−1 for CH2 stretching and 1731 cm−1 for CO stretching.24 The intensities of these two bands increased with the increment of the reaction temperature and the molar ratio of ε-CL to AXU, corresponding to the slight changes of DS in Table 1. 1 H NMR, 1H−1H COSY, 13C NMR, and HSQC. Figure 2 illustrates the 1H NMR spectra of unmodified xylan (A) and xylan-g-PCL copolymer sample 9 (B; DS of 0.39). In spectrum A, the proton signals at 4.90−5.50 ppm originated from the hydroxyl groups in AXU.25 The strong signals at 3.59 and 2.50 ppm relate to water and DMSO, respectively. The signals at 4.26, 3.02, 3.25, 3.48, 3.85, and 3.13 ppm originated from the 685

dx.doi.org/10.1021/jf4036047 | J. Agric. Food Chem. 2014, 62, 682−690

Journal of Agricultural and Food Chemistry

Article

Figure 4. 13C NMR spectra of unmodified xylan (A) and xylan-g-PCL copolymer sample 9 (B).

DS =

terminal CL I a/2 = AXU IH1

WPCL =

114DS × 100% 132 + 114DS

Figure 4 shows the 13C NMR spectra of the unmodified xylan (A) and xylan-g-PCL copolymer sample 9 (B; DS of 0.39). The five major signals at 102.1, 75.8, 74.5, 73.2, and 63.7 ppm are attributed to C1, C4, C3, C2, and C5, respectively, of AXU in xylan.25 The signal at 172.1 ppm observed in spectrum B is assigned to the carbonyl group (f position in PCL), indicating the successful attachment of PCL onto xylan, corresponding to the results from FTIR analysis. Moreover, the signals at 33.7 ppm (a), 32.3 ppm (d), 25.0 ppm (b), 24.4 ppm (c), and 60.6 ppm (e) correspond to the methylene carbons of the attached PCL side chains,18 suggesting that the ROP reaction of ε-CL onto xylan does occur in DMSO/LiCl. Two-dimensional 1H−13C HSQC is a powerful tool for qualitative and quantitative analyses of chemical structures. It could provide more detailed information of signals overlapped in 1H and 13C NMR spectra. In the present study, unmodified xylan (A) and xylan-g-PCL copolymer sample 9 (B; DS of 0.39) were characterized with HSQC spectroscopy to better understand the xylan-g-PCL structure and further confirm the attachment of PCL on xylan. The HSQC spectra are illustrated in Figure 5 at a relatively low contour level to exhibit the primary

(1)

(2)

where DS is the degree of substitution of PCL, WPCL is the weight percent of PCL side chains, AXU is anhydroxylose unit, 114 g/mol is the molar mass of ε-CL, and 132 g/mol is the molar mass of an AXU. As shown in Table 1, the data calculated from 1H NMR confirmed the attachment of PCL onto xylan. However, the length of PCL side chains was only 1, which was less than that (2.29−3.05) attached onto cellulose in ionic liquid [Bmim]Cl.18 The differences of the ROP reaction of ε-CL onto xylan and cellulose were probably due to the different chemistries of xylan and cellulose. Cellulose is a linear chain polymer with three reactive hydroxyls on each anhydroglucose unit (AGU), while xylan is heterpolysaccharide with two reactive hydroxyls on each AXU of the backbone and branched with different groups. Cellulose comparatively allows for slightly longer side chains than xylan, resulting in the high DP of PCL attached. 686

dx.doi.org/10.1021/jf4036047 | J. Agric. Food Chem. 2014, 62, 682−690

Journal of Agricultural and Food Chemistry

Article

Figure 5. HSQC spectra of unmodified xylan (A) and xylan-g-PCL copolymer sample 9 (B).

More importantly, the changes of the correlations belonging to the polysaccharides indicated the detailed ROP reaction on AXU. The strong signals, especially C2 and C3, indicated that most of xylan was unsubstituted, corresponding to the relatively low DS (0.39). The two important weak correlations at δH/δC of 4.49/73.6 and 4.79/75.0 ppm are due to substituted C2 and C3,27 respectively, indicating that the attachment of PCL occurred at C2 and C3 positions of AXU. Clearly, more PCL side chains were attached to C3 than to C2. The resonances assigned to substituted and unsubstituted C2 and C3 were integrated to quantitatively compare the ROP reactivity of hydroxyl groups at C2 and C3 in DMSO/LiCl, and the results indicated that 69.23 and 30.77% of PCL side chains were attached to C3 and C2 positions of AXU, respectively.

signals, both substituted and unsubstituted. In spectrum A, the six strong correlations at δH/δC of 4.26/102.1, 3.02/73.2, 3.25/74.5, 3.48/75.8, 3.85/63.7, and 3.13/63.7 ppm are related to C1−H1, C2−H2, C3−H3, C4−H4, C5e−H5e, and C5a−H5a, respectively, in xylan. The weak correlations at δH/δC of 3.78/ 80.8, 3.60/78.3, 3.91/86.4, and 3.41/62.2 ppm are attributed to C2−H2, C3−H3, C4−H4, and C5−H5 in α-L-arabinofuranosyl residues.26 Typically, methylene signals in the PCL side chain are shown in the aliphatic region as strong 1H−13C correlations. The strong correlations at δH/δC of 2.24/33.7, 1.50/25.0, 1.28/24.4, 1.38/32.3, and 3.36/60.6 ppm in spectrum B are assigned to the methylene group at a, b, c, d and e positions of PCL, respectively. The presence of these strong correlations indicated that the PCL side chains were successfully introduced into xylan. 687

dx.doi.org/10.1021/jf4036047 | J. Agric. Food Chem. 2014, 62, 682−690

Journal of Agricultural and Food Chemistry

Article

low-molecular-weight xylan was not recovered after the ROP reaction under the selected conditions, which was probably due to the high solubility of the dual-component system containing a strong polar aprotic solvent and Lewis base. The primary peak in DTG curves showed the maximum degradation rate at 270 °C for unmodified xylan, 260 °C for sample 7, and 255 °C for sample 9. These data suggested that the attachment of PCL onto xylan led to the decrease of thermal stability of xylan, especially with the increased DS of xylan-g-PCL copolymers, which implied that the increased thermal stability of xylan after modification was probably due to the loss of unrecovered low-molecular-weight fractions in xylan. The TGA/DTG curves of unmodified xylan and xylan-g-PCL copolymers indicated that the ROP reaction actually led to the reduction in thermal stability of xylan. SEM and Physical Properties. The surface morphologies of unmodified xylan and the dried xylan-g-PCL copolymer samples 6 (DS of 0.04), 7 (DS of 0.17), 8 (DS of 0.23), and 9 (DS of 0.39) were examined by SEM, as shown in Figure 7. Unmodified xylan exhibited a rough and irregular surface with different sizes of agglomerates and blocks. The dissolution, modification, and regeneration of xylan destroyed the blocks, and the morphology of xylan-g-PCL copolymer samples changed from an irregular surface with blocks to relatively less rough surface. With the increased DS, the surface of the xylan derivatives gradually became regular and homogeneous, resulting in a more compact structure with decreased agglomerates, which was probably due to the attached PCL side chains on the xylan backbone. In addition, hemicelluloses from plant biomass are known to be non-film-forming polymers.30 The rough and irregular surface with different sizes of agglomerates and blocks of unmodified xylan was probably responsible for the cracks and shrinkage during the failed film forming. Unexpectedly, the easily filmforming capability of xylan-g-PCL copolymers was found in comparison to that of unmodified xylan. The film from xylan-gPCL copolymer sample 9 exhibited a flat and smooth surface. In addition, the film with a thickness ranging from 20 to 30 μm exhibited a relatively low tensile strength, and the stress strength, tensile strain, and elasticity modulus were 17.8 MPa, 5.33%, and 532.8 MPa, respectively. In conclusion, the xylan-g-PCL copolymers were successfully prepared by the homogeneous ROP reaction in DMSO/LiCl under the given mild conditions. The dual-component system, containing a Lewis base LiCl and strong polar aprotic solvent DMSO, acted as a solvent, base, and catalyst for the ROP reaction. The results showed xylan-g-PCL copolymers with a low DS in the range of 0.03−0.39 were obtained under the given conditions. The FTIR and NMR characterization provided more evidence of the occurrence of the ROP reaction and the attachment of PCL side chains onto xylan. The results also indicated that only one ε-CL was confirmed to be attached onto xylan with each side chain. A total of 69.23 and 30.77% of PCL side chains were attached to C3 and C2 positions of AXU, respectively. Thermal analysis indicated that the attachment of PCL onto xylan led to the decreased thermal stability of xylan; however, the loss of unrecovered xylan fractions with a low molecular weight resulted in the increased thermal stability of the samples. The smooth surface of derivatived xylan was obtained with increased DS. Considering the good biodegradability of xylan and PCL, this kind of xylan derivative has potential application in environmentally friendly and biodegradable materials, such as edible food packaging film.

In addition, the HSQC spectrum indicated that the proton signal at C1 was not overlapped with other signals after the ROP reaction, whereas that at C4 was overlapped with methylene at the e position from PCL, C3 and C5 from AXU, and C3 from arabinofuranose. Similarly, the proton signal at the a position had less overlapped signals than other positions. Therefore, H1 and methylene at the a position were selected as the typical proton signals in the DS calculation in the present study. Thermal Analysis. The thermal properties of unmodified xylan and xylan-g-PCL copolymers were investigated using TGA and DTG, and the TGA/DTG curves are illustrated in Figure 6.

Figure 6. TGA/DTG curves of xylan and xylan-g-PCL copolymer samples 7 and 9.

In TGA curves, unmodified xylan began to decompose at 180 °C, while xylan-g-PCL copolymer samples 7 (DS of 0.17) and 9 (DS of 0.39) began to decompose at 220 and 235 °C, respectively. At 50% weight loss, the decomposition temperature occurred at 270 °C for unmodified xylan, 275 °C for sample 7, and 278 °C for sample 9. These data implied that the thermal stability of xylan increased after the ROP reaction in DMSO/LiCl. It is interesting that the pyrolysis residues at 600 °C for unmodified xylan and samples 7 and 9 were 28, 24, and 16%, respectively, indicating the decreased contents of inorganic salts in the samples after modification in DMSO/LiCl under the selected conditions, which was really different from that of cellulose regenerated from ionic liquids.28 DTG means the degradation rate and can be used to compare the thermal stability between the samples.29 In DTG curves, the small peak at about 185 °C in unmodified xylan disappeared after modification, indicating that the 688

dx.doi.org/10.1021/jf4036047 | J. Agric. Food Chem. 2014, 62, 682−690

Journal of Agricultural and Food Chemistry

Article

Figure 7. Scanning electron microscope images of unmodified xylan (A), xylan-g-PCL copolymers (sample 6, DS of 0.04, B; sample 7, DS of 0.21, C; and sample 8, DS of 0.23, D), and the xylan-g-PCL film (sample 9, DS of 0.39, E).





AUTHOR INFORMATION

Corresponding Author

REFERENCES

(1) Saha, B. Hemicellulose bioconversion. J. Ind. Microbiol. Biotechnol. 2003, 30, 279−291. (2) Ebringerová, A.; Heinze, T. Xylan and xylan derivatives Biopolymers with valuable properties. 1. Naturally occurring xylans structures, isolation procedures and properties. Macromol. Rapid Commun. 2000, 21, 542−556. (3) Gabrielii, I.; Gatenholm, P. Preparation and properties of hydrogels based on hemicellulose. J. Appl. Polym. Sci. 1998, 69, 1661−1667. (4) Heinze, T.; Koschella, A.; Ebringerová, A. Chemical functionalization of xylan: A short review. In Hemicelluloses: Science and Technology; Gatenholm, P., Tenkanen, M., Eds.; American Chemical Society (ACS):

*Telephone: 86-20-87111735. Fax: 86-20-87111861. E-mail: chfl[email protected]. Funding

This work was financially supported by the National Natural Science Foundation of China (31170550), the Program for New Century Excellent Talents in University (NCET-11-0154), and the Fundamental Research Funds for the Central Universities. Notes

The authors declare no competing financial interest. 689

dx.doi.org/10.1021/jf4036047 | J. Agric. Food Chem. 2014, 62, 682−690

Journal of Agricultural and Food Chemistry

Article

Washington, D.C., 2003; ACS Symposium Series, Vol. 864, Chapter 20, pp 312−325. (5) Ebringerová, A.; Hromádková, Z.; Heinze, T. Hemicellulose. In Polysaccharides I; Heinze, T., Ed.; Springer: Berlin, Germany, 2005; Vol. 186, pp 1−67. (6) Puls, J.; Schröder, N.; Stein, A.; Janzon, R.; Saake, B. Xylans from oat spelts and birch kraft pulp. Macromol. Symp. 2005, 232, 85−92. (7) El Seoud, O. A.; Koschella, A.; Fidale, L. C.; Dorn, S.; Heinze, T. Applications of ionic liquids in carbohydrate chemistry: A window of opportunities. Biomacromolecules 2007, 8, 2629−2647. (8) Belmokaddem, F.-Z.; Pinel, C.; Huber, P.; Petit-Conil, M.; Da Silva Perez, D. Green synthesis of xylan hemicellulose esters. Carbohydr. Res. 2011, 346, 2896−2904. (9) Sun, R.; Fang, J. M.; Tomkinson, J. Stearoylation of hemicelluloses from wheat straw. Polym. Degrad. Stab. 2000, 67, 345−353. (10) Liu, C.-F.; Zhang, A.-P.; Li, W.-Y.; Yue, F.-X.; Sun, R.-C. Homogeneous modification of cellulose in ionic liquid with succinic anhydride using N-bromosuccinimide as a catalyst. J. Agric. Food Chem. 2009, 57, 1814−1820. (11) Marson, G. A.; El Seoud, O. A. A novel, efficient procedure for acylation of cellulose under homogeneous solution conditions. J. Appl. Polym. Sci. 1999, 74, 1355−1360. (12) Nawaz, H.; Casarano, R.; El Seoud, O. A. First report on the kinetics of the uncatalyzed esterification of cellulose under homogeneous reaction conditions: A rationale for the effect of carboxylic acid anhydride chain length on the degree of biopolymer substitution. Cellulose 2012, 19, 199−207. (13) Habibi, Y.; Goffin, A. L.; Schiltz, N.; Duquesne, E.; Dubois, P.; Dufresne, A. Bionanocomposites based on poly(ε-caprolactone)-grafted cellulose nanocrystals by ring-opening polymerization. J. Mater. Chem. 2008, 18, 5002−5010. (14) Lonnberg, H.; Fogelstrom, L.; Berglund, M.; Malmstrom, E.; Hult, A. Surface grafting of microfibrillated cellulose with poly(εcaprolactone)Synthesis and characterization. Eur. Polym. J. 2008, 44, 2991−2997. (15) Lonnberg, H.; Fogelstrom, L.; Zhou, Q.; Hult, A.; Berglund, L.; Malmstrom, E. Investigation of the graft length impact on the interfacial toughness in a cellulose/poly(ε-caprolactone) bilayer laminate. Compos. Sci. Technol. 2011, 71, 9−12. (16) Lonnberg, H.; Zhou, Q.; Brumer, H.; Teeri, T. T.; Malmstrom, E.; Hult, A. Grafting of cellulose fibers with poly(ε-caprolactone) and poly(L-lactic acid) via ring-opening polymerization. Biomacromolecules 2006, 7, 2178−2185. (17) Hou, Q.; Grijpma, D. W.; Feijen, J. Preparation of porous poly(εcaprolactone) structures. Macromol. Rapid Commun. 2002, 23, 247− 252. (18) Guo, Y. Z.; Wang, X. H.; Shen, Z. G.; Shu, X. C.; Sun, R. C. Preparation of cellulose-graft-poly(ε-caprolactone) nanomicelles by homogeneous ROP in ionic liquid. Carbohydr. Polym. 2013, 92, 77−83. (19) Xie, W.; Chen, D.; Fan, X.; Li, J.; Wang, P. G.; Cheng, H. N.; Nickol, R. G. Lithium chloride as catalyst for the ring-opening polymerization of lactide in the presence of hydroxyl-containing compounds. J. Polym. Sci., Polym. Chem. 1999, 37, 3486−3491. (20) Wang, Z. D.; Zheng, L. C.; Li, C. C.; Zhang, D.; Xiao, Y. N.; Guan, G. H.; Zhu, W. X. A novel and simple procedure to synthesize chitosangraft-polycaprolactone in an ionic liquid. Carbohydr. Polym. 2013, 94, 505−510. (21) 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. (22) Wang, Z.; Yokoyama, T.; Matsumoto, Y. Dissolution of ethylenediamine pretreated pulp with high lignin content in LiCl/ DMSO without milling. J. Wood Chem. Technol. 2010, 30, 219−229. (23) Ayoub, A.; Venditti, R. A.; Pawlak, J. J.; Sadeghifar, H.; Salam, A. Development of an acetylation reaction of switchgrass hemicellulose in ionic liquid without catalyst. Ind. Crops Prod. 2013, 44, 306−314. (24) Labet, M.; Thielemans, W. Improving the reproducibility of chemical reactions on the surface of cellulose nanocrystals: ROP of εcaprolactone as a case study. Cellulose 2011, 18, 607−617.

(25) Cao, X. F.; Sun, S. N.; Peng, X. W.; Zhong, L. X.; Sun, R. C. Synthesis and characterization of cyanoethyl hemicelluloses and their hydrated products. Cellulose 2013, 20, 291−301. (26) Wen, J. L.; Xiao, L. P.; Sun, Y. C.; Sun, S. N.; Xu, F.; Sun, R. C.; Zhang, X. L. Comparative study of alkali-soluble hemicelluloses isolated from bamboo (Bambusa rigida). Carbohydr. Res. 2011, 346, 111−120. (27) Evtuguin, D. V.; Tomas, J. L.; Silva, A. M. S.; Neto, C. P. Characterization of an acetylated heteroxylan from Eucalyptus globulus Labill. Carbohyd. Res. 2003, 338, 597−604. (28) Lan, W.; Liu, C. F.; Yue, F. X.; Sun, R. C.; Kennedy, J. F. Ultrasound-assisted dissolution of cellulose in ionic liquid. Carbohydr. Polym. 2011, 86, 672−677. (29) Nadji, H.; Diouf, P. N.; Benaboura, A.; Bedard, Y.; Riedl, B.; Stevanovic, T. Comparative study of lignins isolated from Alfa grass (Stipa tenacissima L.). Bioresour. Technol. 2009, 100, 3585−3592. (30) Goksu, E. I.; Karamanlioglu, M.; Bakir, U.; Yilmaz, L.; Yilmazer, U. Production and characterization of films from cotton stalk xylan. J. Agric. Food Chem. 2007, 55, 10685−10691.

690

dx.doi.org/10.1021/jf4036047 | J. Agric. Food Chem. 2014, 62, 682−690