- Email: [email protected]
Functionalization of syndiotactic polystyrene with succinic anhydride in the presence of aluminum chloride
European Polymer Journal 41 (2005) 823–829
EUROPEAN POLYMER JOURNAL www.elsevier.com/locate/europolj
Functionalization of syndiotactic polystyrene with succinic anhydride in the presence of aluminum chloride Juan Li, Hua-Ming Li
*
Department of Chemistry, Institute of Polymer Science, Xiangtan University, Xiangtan 411105, Hunan Province, PR China Received 8 October 2004; accepted 29 October 2004 Available online 13 January 2005
Abstract Syndiotactic polystyrene has been chemically modified with succinic anhydride by use of Friedel–Crafts acylation reaction in the presence of anhydrous aluminum chloride in carbon disulfide. The modified syndiotactic polystyrene containing –COCH2CH2COOH fragments in side phenyl rings, named succinoylated syndiotactic polystyrene (s-sPS), was characterized by FTIR and 1H NMR spectroscopy. The effects of reaction conditions on the degree of succinoylation of s-sPS were investigated. In addition, the effects of incorporation of carboxyl groups into syndiotactic polystyrene on the thermal behavior were studied by differential scanning calorimetry in comparison with pure syndiotactic polystyrene. It was found that the crystallization temperature, melting temperature, and degree of crystallinity of the modified polymer decreased with increasing the degree of succinoylation, while the glass transition temperature increased. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Syndiotactic polystyrene; Acylation; Succinic anhydride; Modification
1. Introduction Since the first synthesis of syndiotactic polystyrene (sPS) by Ishihara [1], this new semicrystalline polymer has been the subject of intense investigation. Along with high melting temperature (270 °C), high crystallinity and rapid crystallization rate, sPS exhibits low dielectric constant, high modulus, and good chemical resistance, which has been made an attractive engineering thermoplastic for many applications in the electronic, packaging, and automotive industries. However, sPS resembles
*
Corresponding author. Tel.: +86 732 8293606; fax: +86 732 8293264. E-mail address: [email protected], huamingli8@ 163.com (H.-M. Li).
atactic polystyrene (aPS) polymer with poor impact strength, inherent brittleness, and low surface energy. Recently, attempts have been made to improve the physical properties and processability of sPS through several procedures. Physical bending with other polymers or substrates (e.g., engineering thermoplastics and elastomers) may extend the commercial utility of sPS [2,3]. Except for a few polymers such as aPS and PPO, blending with other polymers (e.g., polyamides) usually leads to phase separation due to lack of compatibility. Therefore, chemical modified sPS polymer with functional groups was expected to be a very desirable material. In previous articles dealing with the preparation of functionalized sPS, experimental observations were interpreted as two aspects. One involves direct polymerization of styrenic monomer or copolymerization with a second monomer by using metallocene catalyst systems
0014-3057/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.eurpolymj.2004.10.048
824
J. Li, H.-M. Li / European Polymer Journal 41 (2005) 823–829
to produce syndiotactic polystyrene derivatives containing functional groups [4–8]. For example, Chung developed a route to prepare functionalized sPS through the direct copolymerization of styrene with a borane-containing styrenic monomer and extended this sPS to prepare functionalized sPS and sPS graft copolymers [7]. Chung also demonstrated a family of syndiotactic polystyrene derivatives containing primary amino groups via stereospecific polymerization of a styrene derivative containing a masking N,N-bis(trimethylsilyl)amino group followed by acid hydrolysis leading to the complete recovery of primary amino groups in sPS derivatives [8]. On the other hand, functionalization of sPS can also be achieved by introducing other types of functional groups into the polymer or by modifying existed functional groups. From a research and development point of view, the latter one is usually more efficient and less expensive. So far, there are only a few reports on discussing the modification of sPS-based polymer, such reports including sulfonated sPS [9,10], brominated sPS [11], acetylated sPS [12], maleated sPS [13,14], and a series of sPS graft copolymers synthesized using ATRP technique with brominated sPS as initiator [15]. In addition, Liu et al. used a mixture of [Ni(p-methallyl)(Br)]2and AlCl3, to graft branched oligoethene onto the pendant aromatic groups of sPS [16]. In our previous paper, we have reported a method for preparing acetylated syndiotactic polystyrene (AsPS) by Friedel–Crafts reaction [12]. The outstanding virtue of this method lies in that it is well suited for the preparation of high molecular mass of styrenic polymer based ionomers with substituted groups situated randomly along the polymer chain under mild conditions [12, 17,18]. In addition, all the reagents used in the reaction are commercially available. Following our previous work, the present work is design to introduce carboxyl groups onto pendant aromatic groups of sPS through Friedel–Crafts acylation reaction. There is limited information about the preparation of sPS bearing pendant carboxyl groups in the side phenyl rings. Pendant functional groups, such as carboxyls, will be potentially interesting for preparing amphiphilic copolymers and ion-containing sPS. The aim of this work is to modify sPS with succinic anhydride via Friedel–Crafts acylation reaction. Furthermore, differential scanning calorimetry (DSC) was used to investigate the thermal properties of functionalized polymers in view of the crystallization and melting behavior along with the neat sPS.
2. Experimental 2.1. Materials The sPS used in these studies was synthesized by bulk polymerization of styrene with a Cp*Ti (OCH2C6H5)3/
MAO catalytic system at 80 °C [19]. The resulting polymer was stirred in a 10 wt.% methanol solution of HCl for 5 h to remove the residual metal catalyst, the polymer was then filtered and dried under vacuum at 70 °C for 72 h after which was extracted with methylethyl ketone (MEK) to remove the atactic component. The purified polymer was characterized to have a very high steric purity (>99% in syndio units) and its number average molecular weight and polydispersity were 210,000 and 2.2, respectively. Succinic anhydride (SA) was purified by recrystallization from chloroform before used. Carbon disulfide was dried overnight with anhydrous calcium chloride, filtered and fractionally distilled in the presence of phosphorus pentoxide before used. All the other reagents and solvents were commercially available and of analytical grade. 2.2. Modification In a typical run, 0.50 g (5 mmol) of SA and 2.00 g (15 mmol) of finely powdered anhydrous aluminum chloride (AlCl3) was treated with 50 ml of carbon disulfide in a 150 ml three-neck round-bottom flask equipped with condenser, dropping funnel, gas inlet/outlet, and a magnetic stirrer. After being rapidly stirred for 2 h, 0.52 g (5 mmol based on benzene ring) of sPS (200 mesh) was added to the mixture. The reaction was continued under nitrogen atmosphere until the product turned to a dark red. Then, the product was decomposed with ice water followed by dilute hydrochloric acid, thoroughly washed with water to remove any residual acid, filtered and dried overnight under vacuum at 70 °C. The modified polymer thus obtained was refined with 1,1,2-trichloroethane/methanol mixture (99/1, v/v), then precipitated with methanol, filtered, and subsequently dried under vacuum. 2.3. Characterization Fourier transform infrared (FTIR) spectra were recorded on a Perkin–Elmer Spectrum One spectrometer. Samples films were cast in aluminum pans from a 1.0 wt.% solution in chloroform/methanol mixture (99/ 1, v/v) and dried under vacuum at 70 °C, which is sufficiently high for removal of residual solvent. 1 H NMR spectra were obtained at 25 °C on a Bruker AV 400 NMR spectrometer. Samples for 1H NMR spectroscopy were prepared by dissolving about 10 mg of products in 5 ml of deuterated chloroform. Tetramethylsilane was used as an internal reference. Quantitative analysis corresponding to the amount of pendant carboxyl groups incorporated onto sPS was done by a titration method as follows: 0.2 g of modified polymer was put in 50 ml refluxing chloroform/methanol mixture (99/1, v/v) for 2 h. Then the hot solution was directly titrated without permitting it to cool to
J. Li, H.-M. Li / European Polymer Journal 41 (2005) 823–829
a phenolphthalein end point using sodium hydroxide (0.05 mol L1) in methanol. Results were expressed as the degree of succinoylation, which is defined as the mole percentage of the styrene units succinoylated. Sample without modification was also titrated, yielding the blank value. Thermal analysis was performed using a TA instruments Q10 differential scanning calorimeter equipped with a RCS accessory under nitrogen atmosphere. For all samples, the standard procedure is as follows: the samples (about 5 mg) were heated at 300 °C for 5 min in order to eliminate the influence of thermal history and the effect of heat treatment on the crystalline structure of the materials, then cooled down to 50 °C to record the crystallization temperatures, and then reheated to 300 °C to record the melting temperatures, all at a rate of 20 °C min1. The recorded temperatures were calibrated using Indium as standard.
3. Results and discussion 3.1. Acylation reaction Friedel–Crafts acylation reactions are aromatic substitution reactions in which benzene (or a substituted benzene) undergoes acylation when treated with carboxylic acid derivatives (usually acyl halide or anhydride) and a Lewis acid catalyst, such as AlCl3 [20]. These reactions were widely used to modify polystyrene through the side groups (phenyl rings) of macromolecules [21]. However, crosslinking reaction of acylated macromolecules usually occurs in Friedel–Crafts acylation reactions, which leads to changes in the molecular mass and the solubility of the modified polymers. In order to overcome this problem, Hird and Eisenberg [18] reported a simple method for the preparation of partial p-carboxylation of linear polystyrene without degradation or crosslinking of the polymer. It is well established that, in Friedel–Crafts acylation reactions, when aluminum chloride and acetyl chloride are allowed to react together prior to addition to the substrate, the ratio of catalyst to acyl component remains constant throughout the reaction, and the results are reproducible. In this study, Friedel–Crafts acylation reaction was used to prepare slightly succinoylated syndiotactic polystyrene (s-sPS) in a heterogeneous process. However, conducting the succinoylation reaction in the solution state proved to be difficult, since sPS only dissolves in high-boiling chlorinated solvents, such as 1,2,4-trichlorobenzene and 1,1,2-trichloroethane, at elevated temperatures. It is well known that chlorinated solvents and high temperatures will have opposite influences on the acylation procedure [20]. Thus, in the first stage of the heterogeneous sPS succinoylation experiments, a charge–transfer complex is formed between AlCl3 and
825
succinic anhydride in carbon disulfide. After stirring this complex about 30 °C for 2 h, powder sPS (200 mesh) was added, then a formation of HCl occurred and the polymer was functionalized (Scheme 1). The following parameters, such as the amount of AlCl3, reaction temperature and reaction time, were changed in order to optimize the process. The work-up procedure involves treatment with ice water followed by dilute hydrochloride acid to decompose the complex and dissolve the aluminum salts. The degree of succinoylation corresponding to carboxylic acid value of the polymers was determined by chemical titration, and the data are presented in Table 1. As shown in Table 1, a noticeable increase of the degree of succinoylation can be observed initially with increasing catalyst concentration. The results indicate that a relatively higher equimolar catalyst concentration which depends on the [AlCl3]/[SA] molar ratio is desired to promote succinoylation efficiency. Data for the succinoylation reactions at different temperature show an increase in the succinoylation efficiency with increasing reaction temperature. On the other hand, a high reaction temperature would lead to crosslinking of sPS. As well, time plays an important role on the succinoylation percentage in the Friedel–Crafts reactions. When succinic anhydride and AlCl3 was first added as to form a charge–transfer complex in the reaction medium to eliminate undesirable side reactions, crosslinking was still observed in the presence of high levels of AlCl3, i.e., [AlCl3]/[SA] molar ratio above 3/1, in parallel with high reaction temperature, i.e., above 40 °C. This is attributed to a low reactivity between succinic anhydride and AlCl3 generated on the sPS backbone, which is believed to be responsible for the increased crosslinking. With respect to the sPS acetylation reactions [12], it is worth noting that the low sPS succinoylation efficiency was achieved due to the lower reactivity of succinic anhydride and aluminum chloride complex in comparison with acetyl chloride. 3.2. FTIR analysis To aid in the structural elucidation of the succinic anhydride-functionalization chemistries, sPS with carboxyl moieties along the backbone was analyzed using FTIR spectroscopy, and assignments for the characteristic groups were developed. FTIR spectra of pure sPS and the s-sPS with a degree of succinoylation of 3.7 mol% in the range 2000– 1500 cm1 are given in Fig. 1a and b, respectively. Compared with Fig. 1a, two new bands appeared at 1685 and 1713 cm1 in Fig. 1b, which confirmed the presence of carbonyl groups in the s-sPS. In the s-sPS molecule, two different carbonyl functional groups separated by two carbon atoms do not lie in the same plane and can be assigned to individual keto and acid groups.
826
J. Li, H.-M. Li / European Polymer Journal 41 (2005) 823–829 O
O O
AlCl3
O
AlCl3
(1)
O
O O H2 C
CH
CH2
m
CH
O
n
AlCl3
O
H+ /H2O H2C
CH
m
CH2
CH
n O O
AlCl3
O
H2C
CH
CH2
m
(2)
CH
n
H2C
O
H2C
O OH
Scheme 1.
Table 1 Synthesis of s-sPS by Friedel–Crafts reactiona [AlCl3]/[SA] (molar ratio)
Timeb (h)
Temperature (°C)
DSc (mol%)
1 2 3 4 5 6 7
2 2 2 3 4 3 3
2 2 2 2 2 4 6
20 30 40 30 30 30 30
0.5 1.5 4.1 2.8 4.6 3.7 5.9
a Conditions: sPS, 0.52 g (5 mmol); SA, 0.50 g (5 mmol); CS2, 50 ml. b The reaction time are refereed to duration of reaction between the AlCl3–SA charge–transfer complex and sPS polymer. c DS referred to the degree of succinoylation obtained by titration analysis.
Conjugation with an aromatic group leads a lower frequency, the keto absorbs at 1685 cm1, while the free acid exhibited bands at 1713 cm1 which attributed to
a
Transmittance
Run
2000
b
1900
1800
1700
1600
1500
Wavenumber (cm-1) Fig. 1. FTIR spectra of pure sPS (a) and s-sPS (b) with degree of succinoylation of 3.7 mol% in the range of 2000– 1500 cm1.
absorbance of isolated and hydrogen-bonded carbonyl groups [22].
J. Li, H.-M. Li / European Polymer Journal 41 (2005) 823–829
3.3. NMR analysis Supporting evidence for the structural elucidation was revealed by 1H NMR analysis. Fig. 2 shows the 1 H NMR spectra of starting sPS and the s-sPS with degree of succinoylation of 3.7 mol%. The resonances at about 1.8 and 1.3 ppm are assigned to CH and CH2 units in the sPS backbone, respectively. After succinoylation, two new broad peaks at about 2.8 ppm and 3.2 ppm, due to methylene (CH2) proton of –COCH2CH2COOH moiety, are observed. Furthermore, in the aromatic region, a new peak due to the protons ortho to the succinoyl group appears around 7.6 ppm [23]. A similar chemical shift was observed for the published acetylated sPS [12]. The degree of succinoylation of the resultant polymer can be estimated from the ratio of the integrated area under the peaks resulting from the aliphatic and aromatic protons in the 1H NMR. The signals of respective protons in the modified polymer were assigned as a, b, c and d as shown in Fig. 2. Aa, Ab, Ac and Ad denote integrated area under the signals of respective protons maximum at d = 1.80, 1.29, 3.22 and 2.84 ppm. The degree of succinoylation (DS) of partially succinoylated sPS was calculated from the relative intensities of the respective signals in 1H NMR spectra according to the following equation: DS ¼ f½3 ðAc þ Ad Þ=½4 ðAa þ Ab Þg 100% The value obtained by a NMR quantitative analytical method through the equation (4.0 mol%) was found to be in agreement with the titration analysis (3.7 mol%). 3.4. Thermal analysis
827
of a semicrystalline polymer has a dramatic effect on the thermodynamics and kinetics of crystallization. Relative to the behavior observed with homopolymer, crystallizable copolymers usually exhibit low melting temperature, low degrees of crystallinity, and a significant decrease in the overall rate of crystallization [24]. In attempt to understand the link between succinoyl moieties and crystallization of s-sPS, the thermal behavior of s-sPS was investigated by means of DSC. The sample subjected to the DSC experiments are used the following protocol: equilibrium at 300 °C and kept at this temperature for 5 min, then cooling from 300 to 50 °C, and finally reheating from 50 to 300 °C, both the heating and cooling rate are 20 °C min1. Figs. 3 and 4 exhibit DSC scans of sPS (a) and related s-sPS with different degree of succinoylation ((b) 0.5 mol%, (c) 1.5 mol%, (d) 3.7 mol% and (e) 5.9 mol%). Table 2 lists the thermal data for each of the samples shown in Figs. 3 and 4. From crystallization temperature (Tc) recorded from the cooling scans of the samples (Fig. 3), the crystallization endotherm of pure sPS occurs at the highest temperature and has the sharpest crystallization exotherm, while crystallization temperature (Tc) and enthalpy of cryatallization (DHc) for s-sPS samples from the melt decreased with increasing degree of succinoylation. Furthermore, a more broadened transition temperature range has been observed for all succinoylated samples with increasing degree of succinoylation, which indicates that the nonisothermal crystallization rate decreases with increasing degree of succinoylation. This suggests that the crystallization rate can be retarded by the presence of covalently attached carboxyl groups. Generally, the degree of crystallinity of crystallizable polymer materials can be estimated by measuring the
The random incorporation of small quantities of noncrystallizable comonomer units into the backbone
CH
m
b
a
CH2
CH
b n
c H2C
O
d H2C
O
a
Endo
H2 C
b a c
OH
d c d
B
e
A
50
100
150
200
250
300
Temperature (°C) 10
9
8
7
6
5
4
3
2
1
0 ppm
Fig. 2. 1H NMR spectra pure sPS (A) and s-sPS (B) with degree of succinoylation of 3.7 mol%.
Fig. 3. DSC cooling scans of pure sPS (a) and s-sPS with different degree of succinoylation, (b) 0.5 mol%, (c) 1.5 mol%, (d) 3.7 mol% and (e) 5.9 mol%.
828
J. Li, H.-M. Li / European Polymer Journal 41 (2005) 823–829
substituent group, the chain mobility, which is required for significant crystallization, would be less in s-sPS relative to the acetylated sPS. Thus, based on this chain mobility argument, the succinoylated styrene units may interrupt or retard crystal growth, limiting the size of the crystallites achievable and resulting in the depression of melting point. The glass transition temperature (Tg) data also helps in understanding the effects of succinoyl groups on the movement of the polymer chains. It is clear from Table 2 that the Tg value of the modified polymer increases with increasing the degree of succinoylation. For example, as degree of succinoylation increases, Tg increases from 93 °C for neat sPS to 98 °C for the 3.7 mol% ssPS sample. As mentioned above, the substituent groups result in reducing the mobility of the polymer chains and therefore raising Tg values. Furthermore, compared with neat sPS, due to the interactions between the acid groups, e.g., hydrogen bonding, the mobility of the polymer chains is also reduced, thus raising Tgin the modified samples. This result also coincides with the results presented in Refs. [9,12].
e d
Endo
c b a
50
100
150
200
250
300
Temperature (°C) Fig. 4. DSC heating scans of pure sPS (a) and s-sPS with different degree of succinoylation, (b) 0.5 mol%, (c) 1.5 mol%, (d) 3.7 mol% and (e) 5.9 mol%.
enthalpic changes at melt. The melting enthalpy of 100% crystalline sPS has been reported to be 53 J g1 [25]. Using this value, the degree of crystallinity (Xc) of the samples was calculated. The data in Table 2 exhibits a systematic trend of degree of crystallinity (Xc) depression with increasing degree of succinoylation. For the sample with degree of succinoylation of 5.9 mol%, its Xc value is 23%, much lower than that of neat sPS (56%). The melting point (Tm) of succinoylated polymers in Fig. 4, as expected, exhibits a systematic trend of depression with increasing degree of succinoylation. The Tm of neat sPS is around 270 °C, similar to the value previously obtained [26]. For the 5.9 mol% succinoylated polymer sample, the Tm decreases to about 255 °C. This result is quite different from the acetylated syndiotactic polystyrene. For the 42.6 mol% acetylated syndiotactic polystyrene sample, the Tm is about 265 °C [12]. This phenomenon may be explained by comparing the size of acyl substituents between the acylated styrene units. It is expected that in s-sPS due to the larger size of the
4. Conclusions The succinoylated syndiotactic polystyrene was accomplished by Friedel–Crafts reaction in a heterogeneous process by using carbon disulfide as dispersing agent, succinic anhydride as succinoylating agent and aluminum chloride as catalyst. An optimum reaction should be carried out at 30 °C with a molar ratio of aluminum chloride to succinic anhydride of 3/1. The succinoylated syndiotactic polystyrene was confirmed by FTIR and 1H NMR spectroscopy. Moreover, it is found that thermal behavior of the succinoylated syndiotactic polystyrene exhibits considerable differences in comparison to the neat sPS. The melting temperature (Tm), crystallization temperature (Tc), and degree of crystallinity of the succinoylated polymer samples decreases with increasing the degree of succinoylation, while the glass
Table 2 Summary of DSC results for sPS and s-sPS Run
DSa (mol%)
Tgb (°C)
Tmb (°C)
DHm (J g1)
Xcc (%)
Tcb (°C)
DHc (J g1)
1 2 3 4 5
0 0.5 1.5 3.7 5.9
92.7 97.2 97.1 98.3 95.1
270.7 269.8 268.0 263.8 255.6
29.5 27.5 19.6 16.0 13.4
55.7 51.9 37.0 30.2 25.3
239.5 236.7 225.8 213.0 197.6
30.4 26.9 20.3 16.7 16.0
a
DS referred to the degree of succinoylation. The glass transition temperatures, Tgs were determined as the midpoint of the step change in the heat flow. The melting, Tm and crystallization, Tc temperatures were selected as the peak maximum or minimum in endothermic or exothermic transition, respectively. c The degree of crystallinity in the sample, Xc is determined by the equation: X c ¼ ðDH m =DH 0m Þ 100%, where DHm is the melting enthalpy of the sample and DH 0m is the melting enthalpy of 100% crystalline sPS (53 J g1 [25]). b
J. Li, H.-M. Li / European Polymer Journal 41 (2005) 823–829
transition temperature increases. The functionalized sPS offers possibility for the development of novel sPS-based polymer blends and composites, thus extending the application field of sPS.
Acknowledgments The authors thank the Key Project of Scientific Research Funds of Hunan Provincial Education Department (02A011) for support of this work.
References [1] Ishihara N, Seimiya T, Kuramoto M, Uoi M. Macromolecules 1986;19:2464. [2] Li HM, Shen ZG, Zhu FM, Lin SA. Eur Polym J 2002; 38:1255. [3] Hong BK, Jo WH. Polymer 2000;41:2069. [4] Longo P, Oliva L, Grassi A. Macromol Chem 1990;191: 2387. [5] Aaltonen P, Seppala J, Matilainen L, Leskela M. Macromolecules 1994;27:3136. [6] Xu G, Lin SA. Macromolecules 1997;30:685. [7] Dong JY, Manias E, Chung TC. Macromolecules 2002;35: 3439. [8] Xu G, Chung TC. Macromolecules 2000;33:5803. [9] Su ZH, Li X, Hsu SL. Macromolecules 1994;27:287.
829
[10] Orler EB, Yontz DJ, Moore RB. Macromolecules 1993;26: 5157. [11] Jejiku T, Woolman JG. JP 10195133, 1998. [12] Gao Y, Li HM. Polym Int 2004;53:1436. [13] Li HM, Chen HB, Shen ZG, Lin SA. Polymer 2002;43: 5455. [14] Lim JG, Baik JH, Zhang XQ, Son Y, Choi WM, Park OO. Polym Bull 2002;48:397. [15] Liu SS, Sen A. Macromolecules 2000;33:5106. [16] Liu SS, Sen A. J Polym Sci, Polym Chem 2001;39:446. [17] Kurbanova RA, Mirzaoglu R, Akovali G, Rzaev ZMO, Karatas I, Okudan A. J Appl Polym Sci 1996;59:235. [18] Hird B, Eisenberg A. J Polym Sci, Polym Chem 1993;37: 1377. [19] Zhu FM, Lin SA, Zhou WL, Tu JJ, Chen DQ. Chem J Chin Univ 1998;19:1844. [20] Andrew SJ. Introduction to organic chemistry. New York: Wiley; 1971. [21] Sun G, Chen TY, Worely SD. Polymer 1996;37:3753. [22] Bellamy LJ. The infrared spectra of complex molecules, vol. 2. 2nd ed. London and New York: Chapman and Hall; 1980. [23] Nasrullah JM, Raja S, Vijayakumaran K, Dhamodharan R. J Polym Sci, Polym Chem 2000;38:453. [24] Orler EB, Calhoun BH, Moore RB. Macromolecules 1996; 29:5965. [25] De Candia F, Filho AR, Vittoria V. Colloid Polym Sci 1991;269:650. [26] Li HM, Liu JC, Zhu FM, Lin SA. Polym Int 2001;50: 421.
EUROPEAN POLYMER JOURNAL www.elsevier.com/locate/europolj
Functionalization of syndiotactic polystyrene with succinic anhydride in the presence of aluminum chloride Juan Li, Hua-Ming Li
*
Department of Chemistry, Institute of Polymer Science, Xiangtan University, Xiangtan 411105, Hunan Province, PR China Received 8 October 2004; accepted 29 October 2004 Available online 13 January 2005
Abstract Syndiotactic polystyrene has been chemically modified with succinic anhydride by use of Friedel–Crafts acylation reaction in the presence of anhydrous aluminum chloride in carbon disulfide. The modified syndiotactic polystyrene containing –COCH2CH2COOH fragments in side phenyl rings, named succinoylated syndiotactic polystyrene (s-sPS), was characterized by FTIR and 1H NMR spectroscopy. The effects of reaction conditions on the degree of succinoylation of s-sPS were investigated. In addition, the effects of incorporation of carboxyl groups into syndiotactic polystyrene on the thermal behavior were studied by differential scanning calorimetry in comparison with pure syndiotactic polystyrene. It was found that the crystallization temperature, melting temperature, and degree of crystallinity of the modified polymer decreased with increasing the degree of succinoylation, while the glass transition temperature increased. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Syndiotactic polystyrene; Acylation; Succinic anhydride; Modification
1. Introduction Since the first synthesis of syndiotactic polystyrene (sPS) by Ishihara [1], this new semicrystalline polymer has been the subject of intense investigation. Along with high melting temperature (270 °C), high crystallinity and rapid crystallization rate, sPS exhibits low dielectric constant, high modulus, and good chemical resistance, which has been made an attractive engineering thermoplastic for many applications in the electronic, packaging, and automotive industries. However, sPS resembles
*
Corresponding author. Tel.: +86 732 8293606; fax: +86 732 8293264. E-mail address: [email protected], huamingli8@ 163.com (H.-M. Li).
atactic polystyrene (aPS) polymer with poor impact strength, inherent brittleness, and low surface energy. Recently, attempts have been made to improve the physical properties and processability of sPS through several procedures. Physical bending with other polymers or substrates (e.g., engineering thermoplastics and elastomers) may extend the commercial utility of sPS [2,3]. Except for a few polymers such as aPS and PPO, blending with other polymers (e.g., polyamides) usually leads to phase separation due to lack of compatibility. Therefore, chemical modified sPS polymer with functional groups was expected to be a very desirable material. In previous articles dealing with the preparation of functionalized sPS, experimental observations were interpreted as two aspects. One involves direct polymerization of styrenic monomer or copolymerization with a second monomer by using metallocene catalyst systems
0014-3057/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.eurpolymj.2004.10.048
824
J. Li, H.-M. Li / European Polymer Journal 41 (2005) 823–829
to produce syndiotactic polystyrene derivatives containing functional groups [4–8]. For example, Chung developed a route to prepare functionalized sPS through the direct copolymerization of styrene with a borane-containing styrenic monomer and extended this sPS to prepare functionalized sPS and sPS graft copolymers [7]. Chung also demonstrated a family of syndiotactic polystyrene derivatives containing primary amino groups via stereospecific polymerization of a styrene derivative containing a masking N,N-bis(trimethylsilyl)amino group followed by acid hydrolysis leading to the complete recovery of primary amino groups in sPS derivatives [8]. On the other hand, functionalization of sPS can also be achieved by introducing other types of functional groups into the polymer or by modifying existed functional groups. From a research and development point of view, the latter one is usually more efficient and less expensive. So far, there are only a few reports on discussing the modification of sPS-based polymer, such reports including sulfonated sPS [9,10], brominated sPS [11], acetylated sPS [12], maleated sPS [13,14], and a series of sPS graft copolymers synthesized using ATRP technique with brominated sPS as initiator [15]. In addition, Liu et al. used a mixture of [Ni(p-methallyl)(Br)]2and AlCl3, to graft branched oligoethene onto the pendant aromatic groups of sPS [16]. In our previous paper, we have reported a method for preparing acetylated syndiotactic polystyrene (AsPS) by Friedel–Crafts reaction [12]. The outstanding virtue of this method lies in that it is well suited for the preparation of high molecular mass of styrenic polymer based ionomers with substituted groups situated randomly along the polymer chain under mild conditions [12, 17,18]. In addition, all the reagents used in the reaction are commercially available. Following our previous work, the present work is design to introduce carboxyl groups onto pendant aromatic groups of sPS through Friedel–Crafts acylation reaction. There is limited information about the preparation of sPS bearing pendant carboxyl groups in the side phenyl rings. Pendant functional groups, such as carboxyls, will be potentially interesting for preparing amphiphilic copolymers and ion-containing sPS. The aim of this work is to modify sPS with succinic anhydride via Friedel–Crafts acylation reaction. Furthermore, differential scanning calorimetry (DSC) was used to investigate the thermal properties of functionalized polymers in view of the crystallization and melting behavior along with the neat sPS.
2. Experimental 2.1. Materials The sPS used in these studies was synthesized by bulk polymerization of styrene with a Cp*Ti (OCH2C6H5)3/
MAO catalytic system at 80 °C [19]. The resulting polymer was stirred in a 10 wt.% methanol solution of HCl for 5 h to remove the residual metal catalyst, the polymer was then filtered and dried under vacuum at 70 °C for 72 h after which was extracted with methylethyl ketone (MEK) to remove the atactic component. The purified polymer was characterized to have a very high steric purity (>99% in syndio units) and its number average molecular weight and polydispersity were 210,000 and 2.2, respectively. Succinic anhydride (SA) was purified by recrystallization from chloroform before used. Carbon disulfide was dried overnight with anhydrous calcium chloride, filtered and fractionally distilled in the presence of phosphorus pentoxide before used. All the other reagents and solvents were commercially available and of analytical grade. 2.2. Modification In a typical run, 0.50 g (5 mmol) of SA and 2.00 g (15 mmol) of finely powdered anhydrous aluminum chloride (AlCl3) was treated with 50 ml of carbon disulfide in a 150 ml three-neck round-bottom flask equipped with condenser, dropping funnel, gas inlet/outlet, and a magnetic stirrer. After being rapidly stirred for 2 h, 0.52 g (5 mmol based on benzene ring) of sPS (200 mesh) was added to the mixture. The reaction was continued under nitrogen atmosphere until the product turned to a dark red. Then, the product was decomposed with ice water followed by dilute hydrochloric acid, thoroughly washed with water to remove any residual acid, filtered and dried overnight under vacuum at 70 °C. The modified polymer thus obtained was refined with 1,1,2-trichloroethane/methanol mixture (99/1, v/v), then precipitated with methanol, filtered, and subsequently dried under vacuum. 2.3. Characterization Fourier transform infrared (FTIR) spectra were recorded on a Perkin–Elmer Spectrum One spectrometer. Samples films were cast in aluminum pans from a 1.0 wt.% solution in chloroform/methanol mixture (99/ 1, v/v) and dried under vacuum at 70 °C, which is sufficiently high for removal of residual solvent. 1 H NMR spectra were obtained at 25 °C on a Bruker AV 400 NMR spectrometer. Samples for 1H NMR spectroscopy were prepared by dissolving about 10 mg of products in 5 ml of deuterated chloroform. Tetramethylsilane was used as an internal reference. Quantitative analysis corresponding to the amount of pendant carboxyl groups incorporated onto sPS was done by a titration method as follows: 0.2 g of modified polymer was put in 50 ml refluxing chloroform/methanol mixture (99/1, v/v) for 2 h. Then the hot solution was directly titrated without permitting it to cool to
J. Li, H.-M. Li / European Polymer Journal 41 (2005) 823–829
a phenolphthalein end point using sodium hydroxide (0.05 mol L1) in methanol. Results were expressed as the degree of succinoylation, which is defined as the mole percentage of the styrene units succinoylated. Sample without modification was also titrated, yielding the blank value. Thermal analysis was performed using a TA instruments Q10 differential scanning calorimeter equipped with a RCS accessory under nitrogen atmosphere. For all samples, the standard procedure is as follows: the samples (about 5 mg) were heated at 300 °C for 5 min in order to eliminate the influence of thermal history and the effect of heat treatment on the crystalline structure of the materials, then cooled down to 50 °C to record the crystallization temperatures, and then reheated to 300 °C to record the melting temperatures, all at a rate of 20 °C min1. The recorded temperatures were calibrated using Indium as standard.
3. Results and discussion 3.1. Acylation reaction Friedel–Crafts acylation reactions are aromatic substitution reactions in which benzene (or a substituted benzene) undergoes acylation when treated with carboxylic acid derivatives (usually acyl halide or anhydride) and a Lewis acid catalyst, such as AlCl3 [20]. These reactions were widely used to modify polystyrene through the side groups (phenyl rings) of macromolecules [21]. However, crosslinking reaction of acylated macromolecules usually occurs in Friedel–Crafts acylation reactions, which leads to changes in the molecular mass and the solubility of the modified polymers. In order to overcome this problem, Hird and Eisenberg [18] reported a simple method for the preparation of partial p-carboxylation of linear polystyrene without degradation or crosslinking of the polymer. It is well established that, in Friedel–Crafts acylation reactions, when aluminum chloride and acetyl chloride are allowed to react together prior to addition to the substrate, the ratio of catalyst to acyl component remains constant throughout the reaction, and the results are reproducible. In this study, Friedel–Crafts acylation reaction was used to prepare slightly succinoylated syndiotactic polystyrene (s-sPS) in a heterogeneous process. However, conducting the succinoylation reaction in the solution state proved to be difficult, since sPS only dissolves in high-boiling chlorinated solvents, such as 1,2,4-trichlorobenzene and 1,1,2-trichloroethane, at elevated temperatures. It is well known that chlorinated solvents and high temperatures will have opposite influences on the acylation procedure [20]. Thus, in the first stage of the heterogeneous sPS succinoylation experiments, a charge–transfer complex is formed between AlCl3 and
825
succinic anhydride in carbon disulfide. After stirring this complex about 30 °C for 2 h, powder sPS (200 mesh) was added, then a formation of HCl occurred and the polymer was functionalized (Scheme 1). The following parameters, such as the amount of AlCl3, reaction temperature and reaction time, were changed in order to optimize the process. The work-up procedure involves treatment with ice water followed by dilute hydrochloride acid to decompose the complex and dissolve the aluminum salts. The degree of succinoylation corresponding to carboxylic acid value of the polymers was determined by chemical titration, and the data are presented in Table 1. As shown in Table 1, a noticeable increase of the degree of succinoylation can be observed initially with increasing catalyst concentration. The results indicate that a relatively higher equimolar catalyst concentration which depends on the [AlCl3]/[SA] molar ratio is desired to promote succinoylation efficiency. Data for the succinoylation reactions at different temperature show an increase in the succinoylation efficiency with increasing reaction temperature. On the other hand, a high reaction temperature would lead to crosslinking of sPS. As well, time plays an important role on the succinoylation percentage in the Friedel–Crafts reactions. When succinic anhydride and AlCl3 was first added as to form a charge–transfer complex in the reaction medium to eliminate undesirable side reactions, crosslinking was still observed in the presence of high levels of AlCl3, i.e., [AlCl3]/[SA] molar ratio above 3/1, in parallel with high reaction temperature, i.e., above 40 °C. This is attributed to a low reactivity between succinic anhydride and AlCl3 generated on the sPS backbone, which is believed to be responsible for the increased crosslinking. With respect to the sPS acetylation reactions [12], it is worth noting that the low sPS succinoylation efficiency was achieved due to the lower reactivity of succinic anhydride and aluminum chloride complex in comparison with acetyl chloride. 3.2. FTIR analysis To aid in the structural elucidation of the succinic anhydride-functionalization chemistries, sPS with carboxyl moieties along the backbone was analyzed using FTIR spectroscopy, and assignments for the characteristic groups were developed. FTIR spectra of pure sPS and the s-sPS with a degree of succinoylation of 3.7 mol% in the range 2000– 1500 cm1 are given in Fig. 1a and b, respectively. Compared with Fig. 1a, two new bands appeared at 1685 and 1713 cm1 in Fig. 1b, which confirmed the presence of carbonyl groups in the s-sPS. In the s-sPS molecule, two different carbonyl functional groups separated by two carbon atoms do not lie in the same plane and can be assigned to individual keto and acid groups.
826
J. Li, H.-M. Li / European Polymer Journal 41 (2005) 823–829 O
O O
AlCl3
O
AlCl3
(1)
O
O O H2 C
CH
CH2
m
CH
O
n
AlCl3
O
H+ /H2O H2C
CH
m
CH2
CH
n O O
AlCl3
O
H2C
CH
CH2
m
(2)
CH
n
H2C
O
H2C
O OH
Scheme 1.
Table 1 Synthesis of s-sPS by Friedel–Crafts reactiona [AlCl3]/[SA] (molar ratio)
Timeb (h)
Temperature (°C)
DSc (mol%)
1 2 3 4 5 6 7
2 2 2 3 4 3 3
2 2 2 2 2 4 6
20 30 40 30 30 30 30
0.5 1.5 4.1 2.8 4.6 3.7 5.9
a Conditions: sPS, 0.52 g (5 mmol); SA, 0.50 g (5 mmol); CS2, 50 ml. b The reaction time are refereed to duration of reaction between the AlCl3–SA charge–transfer complex and sPS polymer. c DS referred to the degree of succinoylation obtained by titration analysis.
Conjugation with an aromatic group leads a lower frequency, the keto absorbs at 1685 cm1, while the free acid exhibited bands at 1713 cm1 which attributed to
a
Transmittance
Run
2000
b
1900
1800
1700
1600
1500
Wavenumber (cm-1) Fig. 1. FTIR spectra of pure sPS (a) and s-sPS (b) with degree of succinoylation of 3.7 mol% in the range of 2000– 1500 cm1.
absorbance of isolated and hydrogen-bonded carbonyl groups [22].
J. Li, H.-M. Li / European Polymer Journal 41 (2005) 823–829
3.3. NMR analysis Supporting evidence for the structural elucidation was revealed by 1H NMR analysis. Fig. 2 shows the 1 H NMR spectra of starting sPS and the s-sPS with degree of succinoylation of 3.7 mol%. The resonances at about 1.8 and 1.3 ppm are assigned to CH and CH2 units in the sPS backbone, respectively. After succinoylation, two new broad peaks at about 2.8 ppm and 3.2 ppm, due to methylene (CH2) proton of –COCH2CH2COOH moiety, are observed. Furthermore, in the aromatic region, a new peak due to the protons ortho to the succinoyl group appears around 7.6 ppm [23]. A similar chemical shift was observed for the published acetylated sPS [12]. The degree of succinoylation of the resultant polymer can be estimated from the ratio of the integrated area under the peaks resulting from the aliphatic and aromatic protons in the 1H NMR. The signals of respective protons in the modified polymer were assigned as a, b, c and d as shown in Fig. 2. Aa, Ab, Ac and Ad denote integrated area under the signals of respective protons maximum at d = 1.80, 1.29, 3.22 and 2.84 ppm. The degree of succinoylation (DS) of partially succinoylated sPS was calculated from the relative intensities of the respective signals in 1H NMR spectra according to the following equation: DS ¼ f½3 ðAc þ Ad Þ=½4 ðAa þ Ab Þg 100% The value obtained by a NMR quantitative analytical method through the equation (4.0 mol%) was found to be in agreement with the titration analysis (3.7 mol%). 3.4. Thermal analysis
827
of a semicrystalline polymer has a dramatic effect on the thermodynamics and kinetics of crystallization. Relative to the behavior observed with homopolymer, crystallizable copolymers usually exhibit low melting temperature, low degrees of crystallinity, and a significant decrease in the overall rate of crystallization [24]. In attempt to understand the link between succinoyl moieties and crystallization of s-sPS, the thermal behavior of s-sPS was investigated by means of DSC. The sample subjected to the DSC experiments are used the following protocol: equilibrium at 300 °C and kept at this temperature for 5 min, then cooling from 300 to 50 °C, and finally reheating from 50 to 300 °C, both the heating and cooling rate are 20 °C min1. Figs. 3 and 4 exhibit DSC scans of sPS (a) and related s-sPS with different degree of succinoylation ((b) 0.5 mol%, (c) 1.5 mol%, (d) 3.7 mol% and (e) 5.9 mol%). Table 2 lists the thermal data for each of the samples shown in Figs. 3 and 4. From crystallization temperature (Tc) recorded from the cooling scans of the samples (Fig. 3), the crystallization endotherm of pure sPS occurs at the highest temperature and has the sharpest crystallization exotherm, while crystallization temperature (Tc) and enthalpy of cryatallization (DHc) for s-sPS samples from the melt decreased with increasing degree of succinoylation. Furthermore, a more broadened transition temperature range has been observed for all succinoylated samples with increasing degree of succinoylation, which indicates that the nonisothermal crystallization rate decreases with increasing degree of succinoylation. This suggests that the crystallization rate can be retarded by the presence of covalently attached carboxyl groups. Generally, the degree of crystallinity of crystallizable polymer materials can be estimated by measuring the
The random incorporation of small quantities of noncrystallizable comonomer units into the backbone
CH
m
b
a
CH2
CH
b n
c H2C
O
d H2C
O
a
Endo
H2 C
b a c
OH
d c d
B
e
A
50
100
150
200
250
300
Temperature (°C) 10
9
8
7
6
5
4
3
2
1
0 ppm
Fig. 2. 1H NMR spectra pure sPS (A) and s-sPS (B) with degree of succinoylation of 3.7 mol%.
Fig. 3. DSC cooling scans of pure sPS (a) and s-sPS with different degree of succinoylation, (b) 0.5 mol%, (c) 1.5 mol%, (d) 3.7 mol% and (e) 5.9 mol%.
828
J. Li, H.-M. Li / European Polymer Journal 41 (2005) 823–829
substituent group, the chain mobility, which is required for significant crystallization, would be less in s-sPS relative to the acetylated sPS. Thus, based on this chain mobility argument, the succinoylated styrene units may interrupt or retard crystal growth, limiting the size of the crystallites achievable and resulting in the depression of melting point. The glass transition temperature (Tg) data also helps in understanding the effects of succinoyl groups on the movement of the polymer chains. It is clear from Table 2 that the Tg value of the modified polymer increases with increasing the degree of succinoylation. For example, as degree of succinoylation increases, Tg increases from 93 °C for neat sPS to 98 °C for the 3.7 mol% ssPS sample. As mentioned above, the substituent groups result in reducing the mobility of the polymer chains and therefore raising Tg values. Furthermore, compared with neat sPS, due to the interactions between the acid groups, e.g., hydrogen bonding, the mobility of the polymer chains is also reduced, thus raising Tgin the modified samples. This result also coincides with the results presented in Refs. [9,12].
e d
Endo
c b a
50
100
150
200
250
300
Temperature (°C) Fig. 4. DSC heating scans of pure sPS (a) and s-sPS with different degree of succinoylation, (b) 0.5 mol%, (c) 1.5 mol%, (d) 3.7 mol% and (e) 5.9 mol%.
enthalpic changes at melt. The melting enthalpy of 100% crystalline sPS has been reported to be 53 J g1 [25]. Using this value, the degree of crystallinity (Xc) of the samples was calculated. The data in Table 2 exhibits a systematic trend of degree of crystallinity (Xc) depression with increasing degree of succinoylation. For the sample with degree of succinoylation of 5.9 mol%, its Xc value is 23%, much lower than that of neat sPS (56%). The melting point (Tm) of succinoylated polymers in Fig. 4, as expected, exhibits a systematic trend of depression with increasing degree of succinoylation. The Tm of neat sPS is around 270 °C, similar to the value previously obtained [26]. For the 5.9 mol% succinoylated polymer sample, the Tm decreases to about 255 °C. This result is quite different from the acetylated syndiotactic polystyrene. For the 42.6 mol% acetylated syndiotactic polystyrene sample, the Tm is about 265 °C [12]. This phenomenon may be explained by comparing the size of acyl substituents between the acylated styrene units. It is expected that in s-sPS due to the larger size of the
4. Conclusions The succinoylated syndiotactic polystyrene was accomplished by Friedel–Crafts reaction in a heterogeneous process by using carbon disulfide as dispersing agent, succinic anhydride as succinoylating agent and aluminum chloride as catalyst. An optimum reaction should be carried out at 30 °C with a molar ratio of aluminum chloride to succinic anhydride of 3/1. The succinoylated syndiotactic polystyrene was confirmed by FTIR and 1H NMR spectroscopy. Moreover, it is found that thermal behavior of the succinoylated syndiotactic polystyrene exhibits considerable differences in comparison to the neat sPS. The melting temperature (Tm), crystallization temperature (Tc), and degree of crystallinity of the succinoylated polymer samples decreases with increasing the degree of succinoylation, while the glass
Table 2 Summary of DSC results for sPS and s-sPS Run
DSa (mol%)
Tgb (°C)
Tmb (°C)
DHm (J g1)
Xcc (%)
Tcb (°C)
DHc (J g1)
1 2 3 4 5
0 0.5 1.5 3.7 5.9
92.7 97.2 97.1 98.3 95.1
270.7 269.8 268.0 263.8 255.6
29.5 27.5 19.6 16.0 13.4
55.7 51.9 37.0 30.2 25.3
239.5 236.7 225.8 213.0 197.6
30.4 26.9 20.3 16.7 16.0
a
DS referred to the degree of succinoylation. The glass transition temperatures, Tgs were determined as the midpoint of the step change in the heat flow. The melting, Tm and crystallization, Tc temperatures were selected as the peak maximum or minimum in endothermic or exothermic transition, respectively. c The degree of crystallinity in the sample, Xc is determined by the equation: X c ¼ ðDH m =DH 0m Þ 100%, where DHm is the melting enthalpy of the sample and DH 0m is the melting enthalpy of 100% crystalline sPS (53 J g1 [25]). b
J. Li, H.-M. Li / European Polymer Journal 41 (2005) 823–829
transition temperature increases. The functionalized sPS offers possibility for the development of novel sPS-based polymer blends and composites, thus extending the application field of sPS.
Acknowledgments The authors thank the Key Project of Scientific Research Funds of Hunan Provincial Education Department (02A011) for support of this work.
References [1] Ishihara N, Seimiya T, Kuramoto M, Uoi M. Macromolecules 1986;19:2464. [2] Li HM, Shen ZG, Zhu FM, Lin SA. Eur Polym J 2002; 38:1255. [3] Hong BK, Jo WH. Polymer 2000;41:2069. [4] Longo P, Oliva L, Grassi A. Macromol Chem 1990;191: 2387. [5] Aaltonen P, Seppala J, Matilainen L, Leskela M. Macromolecules 1994;27:3136. [6] Xu G, Lin SA. Macromolecules 1997;30:685. [7] Dong JY, Manias E, Chung TC. Macromolecules 2002;35: 3439. [8] Xu G, Chung TC. Macromolecules 2000;33:5803. [9] Su ZH, Li X, Hsu SL. Macromolecules 1994;27:287.
829
[10] Orler EB, Yontz DJ, Moore RB. Macromolecules 1993;26: 5157. [11] Jejiku T, Woolman JG. JP 10195133, 1998. [12] Gao Y, Li HM. Polym Int 2004;53:1436. [13] Li HM, Chen HB, Shen ZG, Lin SA. Polymer 2002;43: 5455. [14] Lim JG, Baik JH, Zhang XQ, Son Y, Choi WM, Park OO. Polym Bull 2002;48:397. [15] Liu SS, Sen A. Macromolecules 2000;33:5106. [16] Liu SS, Sen A. J Polym Sci, Polym Chem 2001;39:446. [17] Kurbanova RA, Mirzaoglu R, Akovali G, Rzaev ZMO, Karatas I, Okudan A. J Appl Polym Sci 1996;59:235. [18] Hird B, Eisenberg A. J Polym Sci, Polym Chem 1993;37: 1377. [19] Zhu FM, Lin SA, Zhou WL, Tu JJ, Chen DQ. Chem J Chin Univ 1998;19:1844. [20] Andrew SJ. Introduction to organic chemistry. New York: Wiley; 1971. [21] Sun G, Chen TY, Worely SD. Polymer 1996;37:3753. [22] Bellamy LJ. The infrared spectra of complex molecules, vol. 2. 2nd ed. London and New York: Chapman and Hall; 1980. [23] Nasrullah JM, Raja S, Vijayakumaran K, Dhamodharan R. J Polym Sci, Polym Chem 2000;38:453. [24] Orler EB, Calhoun BH, Moore RB. Macromolecules 1996; 29:5965. [25] De Candia F, Filho AR, Vittoria V. Colloid Polym Sci 1991;269:650. [26] Li HM, Liu JC, Zhu FM, Lin SA. Polym Int 2001;50: 421.