Macromolecules 2006, 39, 3-5 3

Communications to the Editor

Facile and Effective Purification of Polymers Scheme 1. Atom Transfer Radical Polymerization Mechanism
Produced by Atom Transfer Radical Polymerization
via Simple Catalyst Precipitation and
Microfiltration

Santiago Faucher, Paul Okrutny, and Shiping Zhu*

Department of Chemical Engineering, McMaster UniVersity, alyst tethered to commercial Merrifield resins to yield polymer
Hamilton, Ontario, Canada L8S 4L7 containing residual copper concentrations of 3% of the initial
amount of copper loaded.7 This equates to 120 ppm of copper
ReceiVed September 3, 2005 in polymer. Matyjaszewski also uses a dual immobilized/soluble
ReVised Manuscript ReceiVed NoVember 4, 2005 hybrid catalyst system to reduce these concentrations to 15 ppm
of copper in polymer.8 Brittain utilizes JandaJel-supported
Atom transfer radical polymerization (ATRP) is one of the catalysts and reports residual copper concentrations in polymer
many recently discovered mechanisms capable of producing to be 5-7% of the original copper loaded for the polymeriza-
living and well-controlled polymers.1,2 Its tolerance of impurities, tion.9 The equivalent concentrations are 331-463 ppm of copper
mild reaction conditions, and catalytic nature make the process in polymer per our calculations. Dubois achieves residual nickel
desirable for industrial production of polymers. ATRP mediates concentrations of 34 ppm in polymer following precipitation
the growth of the polymer chain using a catalyst, unlike other of the polymerization solution in heptane.10 More recently, Jones
living free radical polymerization methods. This catalyst used catalyst chemically tethered to Cab-O-Sil to produce
activates and deactivates the growing polymer chains, as shown polymer solutions containing less than 1 ppm of copper.3 This
in Scheme 1.1,2 Unfortunately, ATRP catalyst activities are low, equates to copper concentrations in polymer of less than 3 ppm
requiring large amounts be used to drive the polymerization at as calculated from the given data.
acceptable rates. The cost of the catalyst and the need to remove Biphasic systems are also successful at reducing residual
it from the polymer product have, however, limited ATRP’s copper concentrations in polymer.11,12 Haddleton uses a fluorous
use primarily to the laboratory. biphasic catalyst and reports a residual copper concentration in
Several catalyst-polymer separation and recovery strategies polymer of 880 ppm.11 Vincent et al. studied a nonfluorous
have been applied to overcome the low catalytic activities in biphasic catalyst and report a residual copper concentration in
ATRP and reduce catalyst residue in the polymer product. polymer of 200 ppm.12 Further reductions to less than 50 ppm
Soluble/recoverable, biphasic, and solid supported catalysts have are attainable if the product is further treated with activated
been designed to achieve these ends.3-16 They have generally carbon in 1,4-dioxane.
been successful for catalyst recovery and recycling, but further Soluble and recoverable catalysts yield similar results.13-15
reductions of residual catalyst concentrations in polymer are Zhu uses poly(ethylene-b-ethylene glycol) supported catalyst
sought. and reports polymerization solutions containing as little as 0.51%
Physically adsorbed catalysts on silica supports are found to of the original copper loaded. This translates to 33 ppm of
be recyclable and are effective at reducing the residual metal copper in polymer. Refrigeration of the polymerized solution
concentration in polymer.3,4 Jones reports residual copper to 0 °C prior to catalyst product separation is necessary to
concentrations in polymer solution of 37 ppm when using achieve these lower values.13 Brooks uses catalyst supported
catalyst adsorbed to Cab-O-Sil.3 On the basis of the provided onto polyglycidol and reports copper concentration in polymer
conversion data, catalyst-to-monomer ratios, and monomer of 100 ppm following precipitation and filtration of 10 g of
concentration, we calculate an equivalent copper concentration polymeric solution in 1 L of methanol.14 Weberskirch uses a
in polymer of 250 ppm. For a similar system developed by Zhu bipyridine-functionalized block copolymer support and reports
consisting of CuBr/1,1,4,7,10,10-hexamethyltriethylenetetramine residual copper concentrations as low as 100 ppm in polymer
physically adsorbed to silica gel, we measure 110 ppm of copper following precipitation and two washing steps in ice cold
in the polymerization solution, which equates to 360 ppm of methanol.15 Brittain uses a precipiton tethered catalyst and
copper in polymer.4 An improvement to simple physical reports residual copper concentrations in polymer of less than
adsorption of the catalyst is reversible self-assembly on catalyst 1% of the copper used.16 This equates to less than 70 ppm of
sponges as developed by Shen.5 Using this method, residual copper in polymer per our calculations.
copper concentration in polymer is reported to be as low as Recently, we studied the solubility and kinetics of hetero-
0.38% of the total copper used in the first polymerization. This geneous ATRPs using the catalyst complexes of CuIBr and
translates to 39 ppm of copper in polymer. CuIIBr2 with N,N,N′,N′′,N′′-pentamethyldiethylenetriamine
Comparable results are achieved with catalysts chemically (PMDETA (1)), 1,1,4,7,10,10-hexamethyltriethylenetetramine
grafted to the underlying support.3,6-10 Haddleton reports 1000 (HMTETA (2)), and N,N,N′,N′′,N′′′,N′′′′,N′′′′′,N′′′′′-octameth-
ppm of residual ruthenium in polymer using 3-aminopropyl ylpentaethylene-hexamine (OMPEHA (3)).17 The complexes
silica complexed with RuCl2(PPh3)3.6 Matyjaszewski uses cat- of CuIIBr2 with PMDETA and HMTETA at metal salt to ligand
10.1021/ma051920a CCC: $33.50 © 2006 American Chemical Society
Published on Web 12/02/2005
4 Communications to the Editor Macromolecules, Vol. 39, No. 1, 2006

Scheme 2. Ligand-Metal Salt Network Complex

molar ratios of 4 are found to have no detectable solubility in

toluene as measured by inductively coupled plasma atomic
emission spectroscopy (ICP-AES).17 While the reason for this
insolubility remains unclear, we believe it results from the
formation of a network-like complex between the metal center
and the multidentate amine ligand, as shown in Scheme 2. On Figure 1. Poly(methyl methacrylate) produced by microfiltration and
the basis of the results of the solubility study, the structure of drying of a treated (left) and untreated (right) ATRP polymer solution.
this network is expected to be of the form [(CuIIBr2)3 or 4(L)]n. Polymerization conditions: toluene/MMA ) 2 (w/w), [MMA]/[initia-
A similar mechanism has been proposed to explain the solubility tor]/[CuIBr]/[ligand] ) 100:1:1:1 (molar), 90 °C, stirred. Treatment
behavior of copper(II) in the presence of plant polyphenols.18 conditions: to the polymerized solution add CuIIBr2, 3 equiv to the
initial CuIBr loaded for the ATRP, and stir at 90 °C.
We hypothesized that the formation of this insoluble network
could be used to generate a new method for catalyst-polymer atmosphere and passed through 0.1 µm syringe filter. This
purification post-ATRP. We hoped that the addition of excess filtered sample was colorless compared to the green control
CuIIBr2 at the end of the polymerization would precipitate out sample previously collected. Thus, the copper-ligand precipitate
the soluble catalyst and allow the product to be siphoned off appeared to be effectively retained by the filter while the
catalyst free. polymer passed through it freely. The collected filtrates (post-
ATRPs of methyl methacrylate (MMA) (33 wt %) in toluene and pre-CuIIBr2 addition) following drying overnight in a
at 90 °C were completed using the three ligands. The catalyst vacuum oven are shown in Figure 1. The clear polymer was
complexes consisted of CuIBr coordinated with PMDETA, digested and subsequently assayed for copper using ICP-AES.
HMTETA, or OMPEHA in a one-to-one molar ratio. Methyl The copper assay results are presented in Table 1 along with
R-bromophenylacetate (MBP) was used as the initiator. Molar details of the polymer produced in the ATRPs. The copper
ratios were fixed to MMA/MBP/CuIBr/ligand ) 100/1/1/1. assays corroborate the visual observations. Microfiltration on
Following 315 min of polymerization time a small sample of its own only removes, on average, 95.7% of the heterogeneous
the polymerizing solution was hot filtered (90 °C) through a catalyst as part of it, the active catalyst, remains soluble.17,19
0.1 µm PTFE membrane. The solvent and residual monomer, Average concentrations of residual copper in polymer were
in this control sample, were evaporated under vacuum, leaving measured to be 310 ppm post-filtration. Following the addition
a green polymer (see Figure 1). This polymer was then digested of CuIIBr2, however, average residual copper concentrations in
for subsequent copper analysis using ICP-AES. the filtered polymer drop to 30 ppm, representing an average
Following the removal of the control sample, CuIIBr2 (3 equiv removal of 99.6% of the catalyst loaded. In the case of the
to the original CuIBr loaded) was added to the polymerizing HMTETA-based catalyst, the addition of CuIIBr2 results in
solution under a nitrogen atmosphere. The reaction flask was 99.9% of the catalyst being removed. Residual copper concen-
then left to stir at 90 °C for 3 h under a nitrogen atmosphere. trations in polymer following filtration drop from 470 to 10
Over this period the originally green translucent polymeric ppm when the precipitating agent (CuIIBr2) is used.
solution turned dark brown and opaque for all three catalyst The precipitation and microfiltration method reported here
types. Unfortunately, the precipitate formed did not settle freely provides a practical means of purifying polymers produced
even when centrifuged. The formation of the precipitate by ATRP. The residual copper concentrations in polymer it
increased the viscosity of the polymer solution. This high produces are lower than most other methods thus far de-
viscosity explains the lack of settling. veloped.3-16 Little catalyst synthesis and preparation work is
To overcome this challenge, microfiltration of the product necessary to yield polymer with very low residual copper
was attempted. Following the 3 h complexing period, a sample concentrations as compared to chemically grafted catalysts.3,6-16
of the polymer solution was withdrawn under a nitrogen While catalyst adsorption onto silica gel also offers a simple

Table 1. Residual Copper Concentrations Pre- and Post-Treatment with CuIIBr2 in Filtered Poly(methyl methacrylate) Synthesized by Atom
Transfer Radical Polymerization Using Various Catalysts
copper in poly(methyl methacrylate) (ppm)
filteredc monomer conversion (%)d polymer characteristicsd,f
catalyst useda theorb pre-CuIIBr2d post-CuIIBr2e pre-CuIIBr2d post-CuIIBr2e MN (g/mol) MW/MN
CuIBr/PMDETA 7650 240 30 83 83 15 400 1.29
CuIBr/HMTETA 6900 470 10 92 92 14 600 1.25
CuIBr/OMPEHA 6830 210 50 93 95 15 800 1.34
a PMDETA)N,N,N′,N′′,N′′-pentamethyldiethylenetriamine;HMTETA)1,1,4,7,10,10-hexamethyltriethylenetetramine;OMPEHA)N,N,N′,N′′,N′′′,N′′′′,N′′′′′,N′′′′′-

octamethylpentaethylene-hexamine. b Theoretical calculated as follows: (copper(I) added for catalysis/(monomer weight × conversion reached)) × 1 ×
106. c Measured by ICP-AES, filtered using a 0.1 µm PTFE filter. d Following 315 min of polymerization time. e 180 min after the addition of CuIIBr2 at 315
min. f MN ) number-average molecular weight, and MW ) weight-average molecular weight, measured by gel permeation chromatography.
Macromolecules, Vol. 39, No. 1, 2006 Communications to the Editor 5

route to purifying the polymer, the residual catalyst concentra- (2) Kamiga, M.; Ando, T.; Sawamoto, M. Chem. ReV. 2001, 101, 3689-
tions in polymer are higher to those reported here by an order 3746.
(3) Nguyen, J. V.; Jones, C. W. J. Catal. 2005, 232, 276-294.
of magnitude.3,4 The precipitation process uses reagents and (4) For silica-supported ATRP polymerizations as described in ref 19
procedures that can be easily procured and applied in common residual copper concentrations in the polymer solution, following
laboratories or industrially. Furthermore, the ligands used to catalyst settling for 90 h at room temperature and under a nitrogen
form the ATRP catalyst (PMDETA and HMTETA) and to atmosphere prior to solution removal by cannula, are measured to
be 110 ppm by ICP-AES as described here. This, based on a
which the method applies are inexpensive, widely used, provide conversion of 91% over 315 min, equates to 360 ppm of copper in
good control of polymer architecture, and form highly active polymer.
catalysts.20,21 Finally, the method described provides an op- (5) Ding, S.; Yang, J.; Radosz, M.; Shen, Y. J. Polym. Sci., Part A:
Polym. Chem. 2004, 42, 22-30.
portunity to recover the catalyst in the filter cake and recycle it (6) Haddleton, D. M.; Duncalf, D. J.; Kukulj, D.; Radigue, A. P.
following its reduction back to its active CuI formsan option Macromolecules 1999, 32, 4769-4775.
presently being investigated along with optimization of the (7) Kickelbick, G.; Paik, H.-J.; Matyjaszewski, K. Macromolecules 1999,
process and polishing using precipitation with nonsolvents. 32, 2941-2947.
(8) Hong, S. C.; Paik, H.-J.; Matyjaszewksi, K. Macromolecules 2001,
In conclusion, we have discovered a simple and highly 34, 5099-5102.
effective method of purifying polymers produced by ATRP (9) Honigfort, M. E.; Brittain, J. Macromolecules 2003, 36, 3111-3114.
using catalyst precipitation and microfiltration. The method (10) Duquesne, E.; Degée, Ph.; Habimana, J.; Dubois, Ph. Chem. Commun.
relies on the precipitation of the CuIBr/ligand catalyst complex 2004, 640-641.
using the addition of CuIIBr2. The precipitate thus formed is (11) Haddleton, D. M.; Jackson, S. G.; Bon, S. A. F. J. Am. Chem. Soc.
2000, 122, 1542-1543.
effectively retained by a 0.1 µm PTFE filter, resulting in up to (12) Barré, G.; Taton, D.; Lastécouères, D.; Vincent, J.-M. J. Am. Chem.
99.9% of the catalyst being removed from the polymer. The Soc. 2004, 126, 7764-7765.
resulting clear polymer filtrate contains little residual copper, (13) Shen, Y.; Zhu, S. Macromolecules 2001, 34, 8603-8609.
down to 10 ppm. Optimization of this purification process is (14) Kumar, K. R.; Kizhakkedathu, J. N.; Brooks, D. E. Macromol. Chem.
Phys. 2004, 205, 567-573.
ongoing; its simplicity makes it very appealing to laboratory (15) Kotre, T.; Nuyken, O.; Weberskirch, R. Macromol. Chem. Phys. 2004,
and industrial application. 205, 1187-1195.
(16) Honigfort, M. E.; Brittain, W. J.; Bosanac, T.; Wilcox, C. S.
Acknowledgment. We thank the Natural Science and Macromolecules 2002, 35, 4849-4851.
(17) Faucher, S.; Okrutny, P.; Zhu, S. Catalyst Solubility and Kinetic Study
Engineering Research Council of Canada (NSERC) for sup- of Heterogeneous Atom Transfer Radical Polymerization. Submitted
porting this research and the Canada Foundation of Innovation, to Macromolecules.
who support our research facilities. (18) McDonald, M.; Mila, I.; Scalbert, A. J. Agric. Food Chem. 1996,
44, 599-606.
Supporting Information Available: Ligand structures, polym- (19) Faucher, S.; Zhu, S. Macromol. Rapid Commun. 2004, 25, 991-
994.
erization, microfiltration, and experimental procedures. This material (20) Xia, J.; Matyjaszewski, K. Macromolecules 1997, 30, 7697-7700.
is available free of charge via the Internet at http://pubs.acs.org. (21) Faucher, S.; Zhu, S. Ind. Eng. Chem. Res. 2005, 44, 677-685.
(22) Icke, N. R.; Wisegraver, B. B.; Alles, A. G.; Snyder, H. R. Org.
References and Notes Synth. 1941, Collect. 3, 723.
(1) Matyjaszewski, K.; Xia, J. Chem. ReV. 2001, 101, 2921-2990. MA051920A