Atom transfer radical polymerization of styrene catalyzed by iron^II chloride/EDTA

Deng Kuilin, Zhang Yanzhe, Liu Yinghai^*, Shi Zengqian, Wang Yanhuan
(College of Chemistry & Environmental Science, Hebei University, Baoding, 071002, China)

Abstract The atom transfer radical polymerization of styrene with FeCl₂/EDTA as a novel catalytic system in bulk and in solution was successfully performed at 110ºC. Well-defined polymers with relatively low polydispersities(Mw/Mn<1.5) and high initiator efficiency have been obtained. Block copolymer was synthesized to confirm the controlled nature of the polymerization system in the presence of FeCl₂/EDTA. Various parameters such as the effects of the different initiators, the temperature, the ratio of catalyst to initiator and the solvents were investigated to optimize the polymerization conditions. The results show that FeCl₂/EDTA is an particularly effective catalyst for the ATRP of styrene.
Keywords atom transfer radical polymerization; styrene; ethylenediamine tetraacetic acid (EDTA)

1 INTRODUCTION                  
Since its discovery^[1-2] in 1995, the transition-metal-mediated living radical polymerization, sometimes known as atom transfer radical polymerization (ATRP), has been shown to be a powerful tool to prepare polymers with predictable molecular weights and low polydispersities^[3,4]. With this controlled radical polymerization method, a lot of polymer architectures and compositions are accessible, e.g., block polymer^[5], star polymer^[6]. Recent researches have aimed at the direction of new ligands and new metals that affect the activity and selectivity of the ATRP catalysts. For examples, well-controlled polymerization of acrylates can be achieved at ambient temperature when tris[2-(dimethylamino)ethyl]amine (Me₆TREN) was used as the ligand^[7]. Particularly, monomethoxy-capped oligo-(ethylene glycol) methacrylate (OEGMA) can be polymerized to over 95% monomer conversion within 0.50 h at 20ºC using 2,2^'-bipyridine (bpy) as a ligand for the copper catalyst^[8]. However, organic amine and phosphorus compound which have been extensively used in the previous ATRP studies are harmful to human beings and rather expensive. In this paper, a new kind of ligand based on common organic acid has been developed.
    Some organic acids, such as iminodiacetic acid^[9], succinic acid^[10], acetic acid^[11] and isophthalic acid^[12], have been successfully employed as new ligands in the atom transfer radical polymerization of vinyl monomers, such as styrene (St) and methyl methacrylate (MMA). EDTA is a multidentate organic acid which is widely used in complex chemistry and analytical chemistry. This new catalytic system, including EDTA and FeCl₂, is much cheaper than the conventional ligands and transition metals applied previously in the ATRP. Moreover, EDTA and FeCl₂ with very low toxicity are safer in respect of human health than triphenyl phosphine, bipyridine and their derivatives. On the other hand, FeCl₂/EDTA-FeCl₃/EDTA has the similar redox potential(0.1-0.3v) to some reported ATRP catalysts^[13,9,10,11]. It is the ligand that determines the versatility and the catalytic activity of the complexes in the atom radical transfer polymerization^[14]. The equilibrium constant k_act/k_deact, which represents the activity of the catalyst and has close relation with redox potential of catalyst, should be appropriate^[15]. By this catalytic system, the resulting polymer with relatively narrow molecular weight distribution (Mw/Mn<1.5) has been obtained successfully. The measured molecular weights are close to the theoretical value, indicating the controlled "living" radical polymerization.

2 EXPERIMENTAL
2.1 Materials
Styrene and butyl acrylate (BuA) were purified by washing with 5% NaOH solution and distilled water to remove the stabilizers, then vacuum distilled over CaH₂ and stored at –5ºC in a refrigerator. FeCl₂ was washed with acetone and dried under vacuum at 60ºC before use. EDTA was dried at 60ºC for 24 h. The initiators 1-phenylethyl chloride (1-PECl) and 1-phenylethyl bromide were distilled prior to polymerization and all other reagents were used as received.

2.2 Polymerization
To dry and clean glass tubes, the required amounts of EDTA, FeCl₂ and initiator were added respectively. Three freeze-pump-thaw cycles were performed in order to remove the oxygen before polymerization. The tubes were placed in an oil bath at room temperature and stirred for 5h to make ligand complex sufficiently with transition metal. Then the tubes were heated in a thermostated oil bath at 110ºC. At time intervals, the polymerization was stopped by cooling the tubes. The reaction mixture was dissolved in THF and then precipitated in cold methanol for three cycles. The polymer was filtered off and dried for 24 h at 80ºC under vacuum.

2.3 Characterization
The conversion of monomer was determined gravimetrically. With THF as internal standard, molecular weights and molecular weight distributions were measured using Shodex High Performance KF-803 GPC columns. Polystyrene standards were used to calibrate the columns. The IR spectra of PSt and PSt-block-PBuA were recorded on an FTS-40 IR spectrometer by potassium bromide pellet method.

Fig. 1 Dependence of molecular weight and polydispersity index on monomer conversion for the heterogeneous ATRP of St at 110ºC.
[1-PECl]:[FeCl₂]:[EDTA]:[St]=1:2:4:200, [St]=8.7mol/L, [1-PECl]=0.044 mol/L.

3 RESULTS AND DISCUSSION
3.1 ATRP of St Catalyzed by FeCl₂/EDTA               
The polymerization of St with [1-PECl]:[FeCl₂]:[EDTA]:[St]=1:2:4:200 was carried out in bulk at 110ºC. During the reaction, the color of the reaction mixture gradually changed from light yellow to orange as the polymerization proceeded, which indicates the formation of Fe³⁺ complex. The conversion reaches 61% within 5h and no further polymerizations continued to occur with time prolonged. This may be attributed to the fact that the polymer and the catalyst diffuse difficultly in the very viscous system. As shown in Figure 1, the measured molecular weight linearly increases with increasing in the monomer conversion and is slightly lower than the theoretical value after higher monomer conversion due to the chain transfer^[16]. The molecular weight distribution is obviously low compared with the traditional radical polymerization (M_w/M_n=3~10). Figure 2 demonstrates the kinetic plot of the bulk polymerization of styrene.

Fig. 2 Semi-logarithmic kinetic plot for ATRP of St under heterogeneous conditions at 110ºC, [1-PECl]:[FeCl₂]:[EDTA]:[St]=1:2:4:200, [St]=8.7mol/L, [1-PECl]=0.044 mol/L.

The plot of ln([M₀]/[M]) versus time maintains linear before the monomer conversion reaches 60.5%, indicating the bulk polymerization of St was first order with respect to the monomer concentration. The apparent polymerization rate constant (k_p^app) (4.68×10^-5 s^-1) can be obtained from the slope of the ln([M₀]/[M]) vs time dependence, assuming that ln([M₀]/[M])= k_p^appt and k_p^app=k_p[Pn·]^[13]. K_p was evaluated by Robert G. Gilbert ^[17] at 110 ºC and thus [Pn·] can be estimated as 2.98×10^-8 mol/L.

3.2 Effect of different initiators              
Different initiators were selected for the polymerization of styrene in bulk and the results are listed on Table 1. It shows that the lowest molecular weight distribution was obtained using monochloroacetic acid as initiator. Whereas the further work indicates that the "living" species were "dead" in this reaction system, that is, the conversion of monomer increases without the M_n increasing. This may be explained as follows: ClCH₂COOH not only act as initiator in the reaction, it can also react with FeCl₂ to form complex, which leads to the uncontrolled polymerization. The molecular weight distribution is comparably wide in the polymerization initiated by 1-phenylethyl bromide and 1,2-dichloroethane owing to the chain transfer and other side reaction. Therefore, 1-phenylethyl chloride (1-PECl) was selected as the initiator.

Table 1 Effect of different initiators on polymerization of styrene
T=110ºC t=1h [1-PECl]:[FeCl₂]:[EDTA]:[St]=1:2:4:200

3.3 Effect of temperature                 
The effect of temperature on the polymerization of styrene was studied from 80ºC to 120ºC and shown on Table 2. With increasing in temperature, k_p and k_act/k_deact increase and the polymerization rate increases correspondingly. Due to higher E_p than E_t, K_p/K_t augmented at higher temperature, which results in good control over polymerization. On the other hand, some side reactions also dominate at high temperature and induce blight on the polymers prepared^[18]. When catalyst system and resulting polymers were all taken into account, 110ºC was finally selected for the ATRP of St.

Table 2 Effect of temperature on the polymerization of styrene in bulk
t=2h [1-PECl]:[FeCl₂]:[EDTA]:[St]=1:2:4:200

3.4 Effect of ratio of catalyst to intiator               
The following reaction systems have been studied: [FeCl₂]:[EDTA]:[St] =2:4:200 [1-PECl]:[FeCl₂]:[St]=1:2:20 [1-PECl]:[EDTA]:[St]=1:4:200 [1-PECl]:[St]=1:100. As a result, the polymers with high molecular weight and large polydispersity were obtained, which was in accordance with the conventional radical polymerization. As Table 3 has shown, when the molar ratio of catalyst to initiator ranges from 1:1 to 1:3 with 1-PECl and St fixed, the polymerization rate increases and molecular weight distribution becomes narrow. This may be ascribed to the following fact that large amounts of catalyst make the contact probability with initiator raise in the heterogeneous reaction system, which leads to producing more radicals and accelerating the reaction. On the other hand, Fe²⁺ and Fe³⁺ may deactivate the radicals, thus the equilibrium shifts towards deactivation direction. As a result, the radical concentration declines and the molecular weight distribution of resulting polymer becomes more uniform. With more catalyst added, no evident effect of ratios on M_w/M_n was found owing to the stable radical concentration during the polymerizaion. Compared with the theoretical value, slight excess of catalyst was reasonable in the heterogeneous polymerization of styrene. Furthermore, based on the following reaction system [1-PECl]:[FeCl₂]:[EDTA]:[St]=1:2:4:200, no obvious polymerization was observed, when the same amount of FeCl₃ or phenol to FeCl₂ was added into the reaction system. Thus, the function of characteristic inhibitor of radical polymerization (FeCl₃ and phenol) further verifies the radical nature of the polymerization rather than other ones.

Table 3 Effect of different ratios of catalyst to initiator on polymerization
t=2h [1-PECl]₀:[St]=1:200 [FeCl₂]:[EDTA]=1:2

3.5 Effect of solvents              
A small amount of water was introduced into the reaction system to modify the solubility of the catalyst. The polymerization rate gets fast, because the addition of water into the reaction mixture maybe increases the solubility of FeCl₂ and EDTA and thus more complex has been formed. However, the molecular weight distribution becomes wide under this condition. This may be attributed to the fact that FeCl₂/H₂O can act as the catalyst partly due to the interaction of H₂O with FeCl₂ and its activity is different from that of FeCl₂/EDTA. As to the catalysis of FeCl₂/H₂O, we have reported in the previous paper^[19].
    Two solvents including N,N-dimethylformamide(DMF) and xylene were adopted in the polymerization (solvent:St=1:1 volume ratio) with other conditions fixed. No polymerization occurs with DMF as solvent owing to the dominance of ligand exchange by DMF^[20], indicating that DMF acts as not only a solvent but also a ligand. Polymers with fine control were obtained using xylene as the solvent which has not any complexing role and the living nature still maintained in ATRP of St. The polymerization rate declines without polydispersity changing, within 10h the monomer conversion reaching 82%.

3.6 Block copolymerization of PSt and BuA                  
In order to verify the living nature of the polymer prepared by ATRP, the copolymerization of BuA and PSt (M_{n, GPC}=7960, M_w/M_n=1.4) was performed using FeCl₂/EDTA as catalyst in bulk. The copolymer has been sufficiently extracted to remove the homopolymer of BuA by acetone for 10 hours before determination. In the Fourier transform infrared spectrum (IR) (Fig 3) of the resulted block copolymer (M_n,GPC=10480, M_w/M_n=1.4), the characteristic peaks at 1731cm^-1(u_c=0) and 1145cm^-1(u_c-0) represent the existence of ester carbonyl of PBuA, and 3050, 3020, 1600, 1500, 760 and 700cm^-1 correspond to the characteristic peak of PSt. It can be deduced that the copolymer was made up of St and BuA units. On the other hand, using a halogen-terminated PSt as the macroinitiator (Mn _GPC=8930 Mw/Mn=1.4) for the ATRP of BuA with FeCl₂/EDTA as the catalyst, a PSt-block-PBuA was obtained (Mn_GPC=I3250 Mw/Mn=1.5). The GPC curve of the block copolymer in Fig.4 is shifted toward higher molecular weight than that of the macroinitiator, which indicates that PSt-block-PBuA has been successfully prepared in the above-mentioned condition. These results demonstrated the living nature of the reaction system.

Fig. 3 IR spectra of PSt (a) and PSt-block-PBuA block copolymer(b)

Fig.4 GPC traces of PSt macroinitiator and PSt-block-PBuA using FeCl₂/EDTA as the catalyst; (a) PSt macroinitiator : Mn=8930,Mw/Mn=1.4; (b) PSt-block-PBuA: Mn=13250, Mw/Mn=1.5.

4 CONCLUSION            
FeCl₂/EDTA was selected as the ATRP catalyst of St in terms of toxicity, price and redox potential. Well -defined PSt with low polydispersity index (M_w/M_n <1.5) was successfully synthesized. Several factors were investigated such as initiator, temperature, ratio of catalyst to initiator and solvent. Block copolymerization was carried out to further confirm the living nature of polymerization using this new catalyst system. IR was used to characterize the resulting copolymer. The results showed that FeCl₂/EDTA complex can be used as an effective catalyst for the ATRP of styrene.

REFERENCES              
[1] Kato M, Kamagaito M, Sawamoto M et al. Macromolecules, 1995, 28: 1721.
[2] Wang J S, Matyjaszewski K. J. Am. Chem. Soc., 1995, 117: 5614.
[3] Hiroko Uegaki, Yuzo Kotani, Masami Kamigaito et al. Macromolecules, 1998, 31: 6756.
[4] Robert M J, Perry S C, Christina Ng et al. Macromolecules, 2000, 33: 7404.
[5] Wang J S, Luo N, Ying S K et al. J. Eur. Polym., 2000, 36 (1): 149.
[6] Alexander K, Douglas R R. J. Polym. Prepr., 1999, 40 (2): 413.
[7] Xia J, Gaynor S G, Matyjaszewski K. Macromolecules, 1998, 31: 5958.
[8] Wang X S, Lascelles S F, Jackson R A et al. Chem. Commun., 1999: 1817.
[9] Shen M Z, De Y Y. Journal of Polymer Science: Part A: Polymer Chemistry, 2000, 38: 4308.
[10] Shen M Z, De Y Y, Guo S Z et al. Macromolecular Chemistry and Physics, 2000, 21 (18): 2666.
[11] Shen M Z, De Y Y, Guo S Z. Polymer Bulletin, 2001, 45: 457.
[12] Shen M Z, De Y Y. Macromol. Rapid Commun., 2000, 21: 1209.
[13] Wang J S. Matyjaszewski K. Macromolecules, 1995, 28 (24): 7901.
[14] Sawamoto M, Kamigaito M. Poly. Mater. Sci. Eng., 2001, 84: 56.
[15] Gregory Chambard, Bert Klumperman, Anton German. Macromolecules, 2000, 33: 4417.
[16] Qiu J, Matyjaszewski K. Macromolecules, 1997, 30 (20): 5643.
[17] Robert G G. Pure & Appl. Chem., 1996, 68 (7): 1491.
[18] Shi Y, Fu Z F, Li B Y et al. J. Beijing University Chemical Technology, 1999, 26 (2): 29.
[19] Deng K L, Zhang Y Z, Liu Z H et al. J. Hebei University(Hebeidaxue xuebao), 2002, 22 (2): 38.
[20] Zhu S M, Yan D Y. Macromolecules, 2000, 33 (22): 8233.

[ Back ] [ Home ] [ Up ] [ Next ]