Iron(III) chloride/R-Cl/tributylphosphine for metal-catalyzed living radical polymerization: A unique system with a higher oxidation state iron complex


  • Kotaro Satoh,

    1. Department of Applied Chemistry, Graduate School of Engineering, Nagoya University, Furo-Cho, Chikusa-Ku, Nagoya 464-8603, Japan
    Search for more papers by this author
  • Hiroshi Aoshima,

    1. Department of Applied Chemistry, Graduate School of Engineering, Nagoya University, Furo-Cho, Chikusa-Ku, Nagoya 464-8603, Japan
    Search for more papers by this author
  • Masami Kamigaito

    Corresponding author
    1. Department of Applied Chemistry, Graduate School of Engineering, Nagoya University, Furo-Cho, Chikusa-Ku, Nagoya 464-8603, Japan
    • Department of Applied Chemistry, Graduate School of Engineering, Nagoya University, Furo-Cho, Chikusa-Ku, Nagoya 464-8603, Japan
    Search for more papers by this author


original image

Iron(III) chloride (FeCl3), a higher oxidation state iron species, was employed for the metal-catalyzed living radical polymerization of styrene in the presence of tributylphosphine and a chloride initiator without any added reducing agents or radical initiators. The polymerization smoothly proceeded in toluene at 100 °C to produce polymers with controlled molecular weights and narrow molecular weight distributions (MWDs) (Mw/Mn = 1.1–1.2). The copolymerization of styrene and methyl methacrylate resulted in the simultaneous consumption of both monomers to give copolymers with controlled molecular weights and narrow MWDs, indicating that the FeCl3-catalyzed polymerization proceeded via a radical mechanism.


Recent advances in the controlled/living radical polymerization have been making significant impacts on not only polymer science, but also other wide areas related to the polymeric materials based on the controlled polymer structures.1–7 One of the most widely employed and efficient living radical polymerizations is the one based on a metal catalyst that generates the growing radical species from the carbon–halogen polymer terminal via the reversible redox reaction of the metal center. The various effective central metals, such as Ru(II),8 Cu(I),9–11 Ni(II),12, 13 Fe(II),14–33 and so forth, give one electron upon the formation of the radical species from the dormant carbon–halogen covalent bond and thus should be basically in a low oxidation state.

However, these metal complexes in a low oxidation state are generally sensitive to air or oxygen, especially in solution, which may render them difficult to practically use. Thus, when using metal catalysts for living radical polymerization, it requires careful handling and procedures including the removal of oxygen and oxidants not only from the reaction systems, but also from the transportation packages. This problem can be resolved by using a higher oxidation state metal species, such as Cu(II), in the presence of reducing agents or radical generators, which in situ change the metals into a lower oxidation state so that they can be active for the carbon–halogen dormant terminal.34–44 Another problem for the metal catalysts is the possible toxicity of the metals, which may remain in the polymer products because of the difficulty in the complete removal. This can be more or less overcome by the development of highly active catalysts, which can efficiently work even at a low concentration, or by the use of reducing agents, which can reactivate the oxidized metal species, because the use of a small amount of the metal catalysts diminishes the contamination of metals in the products.37–39 However, the best solution would be the use of a nontoxic metal, such as iron, as well as an easily removable metal catalyst. Iron catalysts are highly promising in all the chemical and industrial areas from the viewpoint of being environmentally benign and naturally abundant in nature,45 and thus the development of effective iron-based catalytic systems are still required.

Under such circumstances, there are now many iron-based catalysts for the living radical polymerization, which are mostly iron(II) halides with various phosphorous and/or nitrogen-based ligands, including phosphine, amine, bipyridine, diamine, diimine, aminopyridine, iminopyridine, bisoxazoline, pyridylphosphine, triamine, diaminopyridine, diiminopyridine, and so forth.14–33 Alternatively, iron(III) halides were employed in conjunction with similar ligands in the presence of a radical initiator like 2,2′-azobisisobutyronitrile (AIBN), in which Fe(III) was reduced to Fe(II) via the reaction with the initiating radical species (R·) along with the formation of an alkyl halide (R[BOND]X).35–43 However, there are no reports on the direct use of the Fe(III) complex as a catalyst for living radical polymerizations without any intentionally added reducing agents, because a higher oxidation state iron(III) complex is believed to be inactive for the homolytic activation of the dormant C[BOND]X bond.46 This communication reports the unprecedented living radical polymerization using iron(III) chloride47 in the presence of a phosphine ligand (PnBu3) and a chloride initiator without any added reducing agents or radical initiators (Scheme 1).48

Scheme 1.

Living radical polymerization of styrene with R-Cl/FeCln/PnBu3.



Styrene (KISHIDA, 99.5%), methyl methacrylate (MMA; Tokyo Kasei; >98%), methyl acrylate (MA; Tokyo Kasei; >99%), and butyl acrylate (BA; KANTO, >99%) were distilled over calcium hydride under reduced pressure before use. FeCl2 (Aldrich; 99.99%), FeCl3 (Aldrich; >99.99%), and PnBu3 (KANTO; > 98%) were used as received and handled in a glove-box (VAC Nexus) under a moisture- and oxygen-free argon atmosphere (O2 < 1 ppm). Me2C(CO2Me)CH2C(CO2Me) (Me)Cl (1) was prepared according to the literature.49 Toluene was distilled over sodium benzophenone ketyl and bubbled with dry nitrogen over 15 min just before use. All other reagents were purified by usual methods.


Polymerization was carried out under dry nitrogen in baked glass tubes equipped with a three-way stopcock. Typically, a mixture of FeCl3 (176 mg, 1.08 mmol) and PnBu3 (0.54 mL, 2.16 mmol) in toluene (10.3 mL) was stirred for 12 h at 80 °C to give a homogeneous purple solution. After the solution was cooled to room temperature, 0.7 mL of the FeCl3 solution (0.10 mol/L) was added into styrene (3.20 mL, 28.0 mmol) and toluene (2.64 mL) mixture. And then a toluene solution (0.60 mol/L) of 1 (0.46 mL, 0.28 mmol) was added. The solution was evenly charged in seven glass tubes, and the tubes were sealed by flame under a nitrogen atmosphere. The tubes were immersed in thermostatic oil bath at 100 °C. In predetermined intervals, the polymerization was terminated by cooling the reaction mixtures to −78 °C. Monomer conversion was determined from the concentration of residual monomer measured by gas chromatography, with toluene as an internal standard (528 h, 91% conversion). The quenched reaction mixture was diluted with toluene (30 mL), washed with dilute citric acid or hydrochloric acid solution and water to remove complex residues, evaporated to dryness under reduced pressure, and vacuum-dried to give the product polymers (0.38 g; Mn = 11,100, Mw/Mn = 1.19).


1H NMR spectrum was recorded on a Varian Gemini 2000 spectrometer (400 MHz). The number-average molecular weights (Mn) and molecular weight distributions (MWDs: Mw/Mn) of the polymers were measured by size-exclusion chromatography using THF, at a flow rate 1.0 mL/min at 40 °C on two polystyrene gel columns; both Shodex KF-805L, that were connected to a JASCO PU-980 precision pump and a JASCO RI-930 detector. The molecular weight was calibrated against eight standard polystyrene samples (Mn = 526–900,000). The monomer conversions were determined from the concentration of the residual monomer measured by gas chromatography, using toluene as the internal standard.


We first investigated the possibility of the living radical polymerization of styrene with a lower oxidation state iron chloride, FeCl2, in conjunction with a chloride initiator in the presence of the PnBu3 ligand and then examined the effects of a higher oxidation state FeCl3 on the polymerization. Throughout this study, both iron salts were of the highly purified and anhydrous form (>99.99%), which are commercially available. They were used as received and handled in a glove box under a moisture- and oxygen-free argon atmosphere. The divalent iron chloride (FeCl2) coupled with a chloride initiator [Me2C(CO2Me)CH2C(CO2Me)(Me)Cl (1)] induced the polymerization of styrene in the presence of PnBu3 in toluene at 100 °C to give the polymers with controlled molecular weights, which increased with the monomer conversion (Fig. 1). The MWDs were narrow throughout the polymerization (Mw/Mn ∼ 1.2) (Fig. 2). These results suggest that the FeCl2/PnBu3 initiating system induced the living polymerization of styrene via activation of the C[BOND]Cl bond, which is much less common than the activation of a weaker C[BOND]Br bond for the metal-catalyzed living radical polymerization or atom-transfer radical polymerization.3, 4 The use of the C[BOND]Cl bond is preferable in terms of stability because it is less susceptible to side reactions.3

Figure 1.

Time-conversion, Mn, and Mw/Mn curves for the polymerization of styrene with 1/FeCln/PnBu3 in toluene at 100 °C: [styrene]0 = 4.0 M; [1]0 = 40 mM; [FeCl2]0 = 10 mM (•, ▴), [FeCl2]0 = [FeCl3]0 = 5.0 mM ( equation image, equation image), or [FeCl3]0 = 10 mM (○, ▵); [PnBu3]0 = 20 mM. The diagonal bold line indicates the calculated Mn assuming the formation of one living polymer per 1 molecule.

Figure 2.

Molecular weight distribution curves of polystyrene obtained in the same experiments as for Figure 1.

The effects of FeCl3 on the living polymerization of styrene with R-Cl/FeCl2/PnBu3 were then investigated. Styrene was thus polymerized by varying the [FeCl2]0/[FeCl3]0 ratio, while the concentrations of the total iron and the phosphine ligand were fixed ([FeCl2]0 + [FeCl3]0 = 10 mM; [PnBu3]0 = 20 mM), in conjunction with 1 in toluene at 100 °C. Contrary to our expectation and general belief that a higher oxidation state metal species should inhibit or retard the polymerization as previously reported with the Cu(I)/Cu(II) counterparts,4 the polymerization smoothly proceeded even with the use of an equimolar amount of FeCl3 to FeCl2, in which the rate was about half that without FeCl3 [Fig. 1(A)]. Furthermore, only the use of FeCl3 without FeCl2 induced the polymerization at almost the same rate. The number-average molecular weights (Mn) of the polymers obtained with both FeCl2/FeCl3 and FeCl3 alone increased in direct proportion to the monomer conversion and agreed well with the calculated values on the assumption that one initiator molecule generates one living polymer chain [Fig. 1(B)]. The MWDs were also narrow for both systems. Especially, among the three systems, FeCl3 afforded the narrowest MWDs throughout the polymerization (Mw/Mn ∼ 1.1). There have been no reports on the styrene living radical polymerization using Fe(III) species even for the system containing radical initiators. There was an attempt with FeBr3 and AIBN in the presence of ammonium salts, in which the control failed presumably because of the occurrence of the cationic process by the Lewis acidic FeBr3.20 Furthermore, the polymers obtained after the treatment of the reaction solution using acidic water became apparently colorless, which suggests that these simple iron-based catalysts are easily removable, in addition to the inherently less hazardous nature of the iron atom. Thus, the iron(III) complexes with phosphine ligands can induce the living radical polymerization of styrene under the appropriate conditions to give the polymers with controlled molecular weights.

The terminal structure of the polystyrene obtained with 1/FeCl3/PnBu3 in toluene at 100 °C was examined by 1H NMR spectroscopy. Figure 3 shows the 1H NMR spectrum of the product polymers after removing the catalyst with dilute hydrochloric acid solution. The polymer gave the characteristic signals of polystyrene; that is, phenyl groups (e) and main-chain aliphatic protons (c and d). In addition to these large absorptions, small signals due to the end groups appeared. They are the CH3- (α; 0.6–1.1 ppm) and CH3O-group (β; 2.8–3.7 ppm) at the α-end derived from the MMA dimer as an initiator, in which the complicated pattern of the signals is due to the stereoisomer in the initiator, and the [BOND]CH[BOND]Cl group (d'; 4.3 ppm) at the ω-end attributed to the chlorine atom at the growing terminal. The functionalities of the α-end (β: Fn = 0.96) and of the ω-end (d': Fn = 0.94) were almost unity, indicating that one polymer was generated from one initiator in a controlled manner. The reaction most probably did not involve the cationic pathway based on the C[BOND]Cl bond activation by the Lewis acidic FeCl3 because there were no peaks of the olefin or indane ring in the spectrum, which might have appeared because of the low stability of the styryl cation at such a high temperature.

Figure 3.

1H NMR spectrum (CDCl3, 55 °C) of polystyrene (Mn = 10,100, Mw/Mn = 1.08) obtained with 1/FeCl3/PnBu3 (40/10/20 mM, [styrene]0 = 4.0 M) in toluene at 100 °C.

The iron(III) system was then applied to the copolymerization of styrene and other monomers, such as MMA, MA, and BA. For the copolymerization of styrene and MMA, both monomers were polymerized at almost the same rate, similar to the conventional radical copolymerization of the two monomers, and were almost quantitatively consumed in 40 h [Fig. 4(A)]. The copolymerization was much faster than styrene homopolymerization. Styrene was also copolymerized with MA and BA by the 1/FeCl3/PnBu3 system. Independent of the comonomer structures, the obtained polymers had narrow MWDs (Mw/Mn ∼ 1.3) [Fig. 4(B)], and the Mn values of the copolymers increased in direct proportion to the monomer conversions and were close to the calculated ones. These results also indicated that the R-Cl/FeCl3/PnBu3 initiating system induced the living radical polymerization and copolymerization despite the use of the higher oxidation state iron catalyst. The working mechanism of FeCl3 has not been clarified for the iron-catalyzed living radical polymerization. Although the Fe(III) species may work as an activator for the C[BOND]Cl bond cleavage, it can be changed into the Fe(II) species in the presence of phosphine at higher temperatures.50, 51 According to the literature, FeCl3(PR3) (R = Ph, tBu, Cy) decomposed into possibly Fe(II) species at ambient temperature, as suggested by some identified decomposed byproducts.50 Determining the catalytic mechanism and the applicability of the FeCl3-based system are now under way.

Figure 4.

Time-conversion curve for the random copolymerization of styrene (•) and MMA (○) (A) and the molecular weight distribution curves for styrene–methyl methacrylate (MMA), styrene–methyl acrylate (MA), and styrene–butyl acrylate (BA) random copolymerization (B) with 1/FeCl3/PnBu3 in toluene at 100 °C: [styrene]0 = [comonomer]0 = 2.0 M; [1]0 = 40 mM; [FeCl3]0 = 10 mM; [PnBu3]0 = 20 mM.

In conclusion, this communication reveals that the R-Cl/FeCl3/PnBu3 system is effective for the living radical polymerizations and is promising in terms of industrial applications because of the use of a higher oxidation stable and environmentally benign iron complex with a simple ligand.