Atom Transfer Polymerisation of methyl methacrylate: Use of chiral aryl/alkyl pyridylmethanimine ligands; with copper(I) bromide and as structurally characterised chiral copper(I) complexes

Article abstract of DOI:10.1039/a800467f

The use of chiral catalysts in the living radical polymerisation of methyl methacrylate via atom transfer polymerisation (ATP) has been investigated in an effort to control the stereochemistry of the polymer backbone. Two enantiomerically pure chiral cata

Full text of DOI:10.1039/a800467f

J O U R N A L O F
Atom transfer polymerisation of methyl methacrylate: use of chiral
aryl/alkyl pyridylmethanimine ligands; with copper( ) bromide and as
structurally characterised chiral copper( ) complexes
I
I
David M. Haddleton,*† David J. Duncalf, Dax Kukulj, Alex M. Heming, Andrew J. Shooter and
Andrew J. Clark
C H E M I S T R Y
Department of Chemistry, University of Warwick, Coventry, UK CV 4 7AL
The use of chiral catalysts in the living radical polymerisation of methyl methacrylate via atom transfer polymerisation (ATP) has
been investigated in an effort to control the stereochemistry of the polymer backbone. Two enantiomerically pure chiral catalysts
have been prepared and used in the ATP of methyl methacrylate: the structurally characterised complex, bis[N-(1-phenylethyl)-2-
pyridylmethanimine]copper() tetrafluoroborate, [Cu(C H N ) ][BF ], and the reaction product of copper() bromide with N-
14 14 2 2
4
(1-cyclohexylethyl)-2-pyridylmethanimine, [Cu(C H N ) ][Br]. Both catalysts were found to be suitable for the ATP of methyl
14 20 2 2
methacrylate in conjunction with either ethyl 2-bromo-2-methylpropanoate in xylene at 90 °C or 4-methoxybenzenesulfonyl
chloride in diphenyl ether at 90 °C. The system yields polymer of relatively narrow polydispersity; however, the use of these chiral
catalysts did not significantly affect the stereochemistry of the polymer backbone. This may be due to the chiral centre being too
distant from the propagating site to exert any influence over the monomer addition step, or that the reaction proceeds via a
completely free-radical mechanism. This is, however, the first time a discrete, structurally characterised copper complex has been
used as an effective ATP catalyst. The bond lengths of the complexed ligand indicate the oxidation state of the copper to be +1 in
the complex.
Controlled polymerisation of acrylic and methacrylic mon-
omers is of continuing interest,1 providing the potential for a
wide range of novel materials from currently available mon-
omers. This has been the objective of many research programs
which have produced a number of different systems to achieve
this aim, viz. living/pseudo-living polymerisation of metha-
crylates.1 A major challenge is to avoid complications arising
from side reaction with the ester group, or deprotonation of
the polymer backbone in the case of poly(acrylics). These
secondary reactions which destroy the living nature of the
polymerisations have proven problematic as most develop-
ments have centred around anionic or anionic-type polymeris-
ations. For example, group transfer polymerisation (GTP) has
been perhaps the most widely studied system in recent years,
and utilises silyl ketene acetals in conjunction with nucleophilic
catalysts to control polymerisation, resulting in polymers with
very narrow polydispersity indexes (PDI) in tetrahydrofuran
at temperatures up to 80 °C.2,3 GTP, as with most other
systems which rely on anionic type propagation, is highly
sensitive to protic impurities necessitating the use of rigorously
dried and purified monomers and solvents. It is for this reason
that we have been focusing on living free-radical polymeris-
ation, which should be tolerant to many different impurities
and functional monomers/solvents rendering the system more
commercially viable. Although stable free radical mediated
polymerisation using TEMPO and related nitroxide stable
organic radicals is relatively successful for styrene and substi-
tuted styrene,4,5 it is less so for acrylics and methacrylics. Also
it is relatively easy to achieve living polymerisation of styrene
by conventional anionic polymerisation, which is the basis for
commercially available styrene-butadiene block copolymers.
We are particularly interested in achieving controlled poly-
merisation of alkyl methacrylates and methacrylates containing
functional groups in the ester side chain which can be used to
synthesise polymers with a wide range of properties for differing
applications. The use of transition metal mediated polymeris-
ation is very attractive for this type of monomer.
Atom transfer polymerisation (ATP) has emerged as an
effective, living, transition metal mediated polymerisation for a
wide range of vinyl monomers; ATP has been derived
from atom transfer chemistry developed for a range of
organic transformations in more conventional organic syn-
thesis,6–8 in particular atom transfer cyclisation.9 This was first
demonstrated, almost simultaneously, by Sawamoto and
Matyjaszewski. Sawamoto described the use of RuCl (PPh )
2
3 3
in conjunction with strong Lewis acids for the living polymeris-
ation of methyl methacrylate.10–12 In related system,
a
Matyjaszewski has been investigating the use of copper()
halides with bipyridine and 4,4∞-dialkyl substituted bipyridine
ligands for the living polymerisation of styrene, methacrylates
and acrylates.13–17 Teyssie has used a well characterised NiII
compound for the living polymerisation of alkyl metha-
crylates.18 In all of these systems, polymerisation is thought to
be mediated by a formal M n/M n+1 redox couple and initiated
t
t
by homolysis, or at least partial homolysis, of an alkyl halide
[Scheme 1(a)]. Percec has extended the range of initiators to
include sulfonyl halides which, due to the ease of homolysis of
the sulfur–halogen bond, give a fast rate of initiation relative
to propagation for virtually all monomers, and are so called
‘universal initiators’ for ATP.19–21
We have been concentrating on a copper based ATP pro-
cess22–24 which uses Schiff bases of type 1 as electron accepting
ligands on the metal.25 These ligands stabilise formal CuI,
relative to CuII, and promote effective ATP. Ligands based on
CuIX
+ CuIIX2
•
Rn
(a)
Rn—X
+
Monomer
Addition
Rn
X
CuIIX
(b)
Rn—X
+
CuIX
Monomer
Addition
†E-mail: msrg@csv.warwick.ac.uk
Scheme 1
J. Mater. Chem., 1998, 8(7), 1525–1532
1525

(R)-,(S)- and ( )-N-(1-Phenylethyl)-2-pyridylmethanimine
[(R)-2,(S)-2 and ( )-2]
N
N
N
A typical preparation was carried out as follows: (R)-1-phenyl-
ethylamine (2.5 g, 0.02 mol) was added dropwise to a stirred
N
*
2
N
N
R
solution of pyridine-2-carbaldehyde (2.2 g, 0.02 mol) in Et O
2
(ca. 20 ml). To the reaction mixture was added MgSO (1–2 g).
4
*
The reaction was stirred at room temperature for 4 h after
which time the solvent was removed in vacuo yielding (R)-N-
(1-phenylethyl)-2-pyridylmethanimine (3.8 g, 90%) as a clear,
R = n-alkyl
3
1
slightly yellow, oil, bp 107 °C at 0.5 mmHg; d (250 MHz,
H
298 K, CDCl ) 8.50 (d, 1H, Py-H), 8.36 (s, 1H, CHNN), 7.96
3
the general structure 1 are extremely easy to synthesize by a
simple condensation of pyridine-2-carbaldehyde with an appro-
priate primary amine, the reaction proceeding efficiently and
quickly at ambient temperature and yielding the products as
yellow oils which are readily purified by distillation. The wide
variety of commercially available primary amines gives a large
family of potential ligands and allows control over catalyst
properties such as solubility in organic media and the CuI/CuII
redox potential.
(d, 1H, Py-H), 7.57 (t, 1H, Py-H), 7.20 (m, 6H, Py-H/Ph-H),
4.51 [q, 1H, Ph(Me)CH], 1.50 (d, 3H, Me); d (250 MHz,
298 K, CDCl ) 159.6 (CHNN), 153.9, 148.4, 135.6, 127.7, 125.9
(Py), 143.5, 126.1, 123.2, 120.7 (Ph), 68.7 (CH ), 23.8
C
3
3
[CH(Me)Ph]; n (neat)/cm−1 1646s, 1587s, 1567s (CNN);
acetone).
max
[a] −1.342[(R)-2], +1.321[(S)-2], 0.000 [( )-2] (c 1,
D
ATP has already been demonstrated to give excellent control
over molecular weight and polydispersity. In addition it would
be desirable to be able to control the stereochemistry of the
polymer backbone. The stereochemistry of the polymer can
have significant effects on materials properties such as the
(R)- and (S)-N-(1-Cyclohexylethyl-2-pyridylmethanimine [(R)-
3 and (S)-3]
A typical preparation was carried out as follows: (R)-1-cyclo-
hexylethylamine (6.5 g, 0.051 mol) was added dropwise to a
stirred solution of pyridine-2-carbaldehyde (5.47 g, 0.051 mol)
in Et O (ca. 30 ml). To the reaction mixture was added MgSO
glass transition temperature (T ), as well as other mechanical
properties. Previously we have suggested that ATP of MMA
g
2
4
(1–2 g). The reaction was stirred at room temperature for 4 h
after which time the solvent was removed in vacuo yielding
(R)-N-(1-cyclohexylethyl-2-pyridylmethanimine (9.4 g, 85%) as
utilising a CuI catalyst with ligands based on 1 may not
proceed via a free-radical mechanism, and we have postulated
a concerted propagation mechanism as being a possibility
where the halogen is not fully abstracted from the propagating
polymer24 [Scheme 1(b)]. Indeed, Teyssie has suggested, in his
studies of the ATP of MMA using a NiII complex, that
propagation may occur via a coordinate mechanism, whereby
the halide from the initiator remains partially bonded to both
the growing polymer chain and the metal centre.18 If the
catalyst is bound to the growing polymer chain, or even in
the vicinity as a ‘radical cage complex’ it may be possible to
influence the stereochemistry of the polymer backbone by the
use of chiral metal catalysts, produced from enantiomerically
pure, chiral Schiff base ligands. It is noted that the stereochem-
istry of all poly(methyl methacrylate) produced via ATP to
date (including CuI, RuII and NiII systems) has been reported
to be consistent with free radical polymerisation, i.e. described
by Bernoullian statistics with the persistence ratio close to
unity.
The work outlined in this paper investigates the possibility
of using the chiral Schiff bases 2 and 3 with CuI halides as
ATP catalysts. Results for both the R and S enantiomers of 2
and 3, as well as for a racemic mixture in the case of 2, are
reported. In all of the studies of copper() mediated ATP
published to date, the active copper complex has been formed
in situ by addition of excess ligand to a suspension of CuI
halide in the reaction solution. We report the use of well
defined, fully characterised copper() compounds (arising from
2) which were isolated by recrystallisation, characterised and
subsequently used as discrete compounds in the polymeris-
ation reaction.
a clear, slightly yellow, oil, bp 96 °C at 0.5 mmHg; d (250 MHz,
H
298 K, CDCl ) 8.48 (d, 1H, Py-H), 8.17 (s, 1H, CHNN), 7.85
3
(d, 1H, Py-H), 7.54 (t, 1H, Py-H), 7.11 (m, 1H, Py-H), 2.97 [q,
1H, Cy(Me)CH], 1.65 (m, 4H, Cy-H), 1.36 (m, 1H, Cy-H),
1.09 [s, 3H, C(H)(Cy)CH ], 1.07 (m, 4H, Cy-H), 0.81 (m, 4H,
3
Cy-H); d (250 MHz, 298 K, CDCl ) 159.4 (CHNN), 154.7,
149.0, 136.1, 124.1, 121.0 (Py), 71.4 (CH ), 43.4, 29.6, 26.0, 19.5
C
3
3
(Cy), 26.1 [CH(Me)Cy]; n (neat)/cm−1 1647s, 1588s, 1568s
max
(CNN); [a] −2.003 [(R)-3], +2.022 [(S)-3] (c 1, acetone).
D
(R,R)-, (S,S)- and ( )-Bis[N-(1-phenylethyl )-2-
pyridylmethanimine)copper(
I ) tetrafluoroborate
[Cu(C H N ) ][BF ] [(R)-4, (S)-4 and ( )-4]
14 14 2 2
4
[Cu(MeCN) ][BF ] was prepared by the method of Kubas.27
4
4
A
[Cu(MeCN) ][BF ] (2.0 g, 6.47 mmol) in MeOH (35 ml)
typical preparation was carried out as follows: to
4
4
was added (R)-N-(1-phenylethyl-2-pyridylmethanimine, (R)-2
(2.7 g, 12.94 mmol). The reaction immediately became a deep
red–brown colour and was stirred for 4 h. After this time the
solution was filtered and concentrated to ca. 20 ml and then
allowed to cool slowly to −40 °C, whereupon red crystals of
[Cu(C H N ) ][BF ] formed in 70% yield (2.57 g), mp
14 14 2 2
4
96–99 °C; d (250 MHz, 298 K, [2H ]acetone) 8.85 (s, 1H,
H
6
CHNN), 7.96 (m, 1H, Py-H), 7.86 (m, 1H, Py-H), 7.42 (br,
1H, Py-H), 7.12, 6.99 (br, 5H, Ph-H), 4.91 [br, 1H,
Ph(Me)CH], 1.55 (br, 3H, Me); d (250 MHz, 298 K,
C
[2H ]acetone) 161.04 (CHNN), 151.89, 149.98, 139.16, 125.94,
6
120.36 (Py), 142.98, 129.4, 127.1, 115.72 (Ph), 67.98
[C(CH )PhH], 38.47 [CCH (Ph)H];
1612m, 1586s (CNN).
n
(Nujol)/cm−1
3
3
max
Experimental
General
Crystal structure determinations
Methyl methacrylate (Aldrich, 99%) was purified by passing
through a column of activated basic alumina to remove
inhibitor. Copper() bromide (Aldrich, 98%) was purified
according to the method of Keller and Wycoff.26 Xylene
(Fisons, 99.8%), ethyl 2-bromo-2-methylpropanoate (Aldrich,
98%), 1-phenylethylamine (Aldrich, 98%), 1-cyclohexylethyl-
amine (Aldrich, 98%) and 4-methoxybenzenesulfonyl chloride
(Avocado, 98%) were used as received.
A suitable crystal of (R)-4 was quickly glued to a quartz fiber,
coated in dry Nujol and cooled in the cold nitrogen gas stream
of the diffractometer. The structure was solved by direct
methods. Anisotropic thermal parameters were used for all
non-H atoms whilst hydrogen atoms were inserted at calculated
positions and fixed, with isotropic thermal parameters (U=
˚
0.08 A3), riding on the supporting atom. The structure solutions
were carried out using SHELXTL28 version 5.0 software on a
1526
J. Mater. Chem., 1998, 8(7), 1525–1532

Silicon Graphics Indy workstation, refinements were carried
out using SHELXTL29 software, minimising on the weighted
R factor wR2.
+
+
*
*
Full crystallographic details, excluding structure factors,
have been deposited at the Cambridge Crystallographic Data
Centre (CCDC). See Information for Authors, J. Mater. Chem.,
1998, Issue 1. Any request to the CCDC for this material
should quote the full literature citation and the reference
number 1145/96.
N
N
N
N
Cu
Cu
N
N
N
N
*
*
Typical polymerisation procedures
4
5
Ethyl 2-bromo-2-methylpropanoate initiator in xylene solvent.
Compound 4 (0.25 g, 4.65×10−4 mol)] [or CuIBr (0.066 g,
4.65×10−4 mol) and 3 (0.30 g, 1.40×10−3 mol)] was placed
into a nitrogen filled flask and subjected to three vacuum/
nitrogen fill cycles. An aliquot of MMA (5.0 ml 4.65×10−2
mol) together with xylene solvent (5.0 ml) was added to the
flask, which was then placed into an oil bath at 90 °C with
stirring. To this mixture was added 6 (0.09 g, 4.65×10−4 mol)
to initiate polymerisation. Periodically, 1–2 ml samples were
removed for conversion and molecular weight analysis.
N
O
S
Br
CH3O
Cl
O
N
O
OEt
C5H11
6
7
8
tems23). Together with the high M , this suggests inefficient
initiation and/or side reactions such as termination in the
polymerisation.
n
4-Methoxybenzenesulfonyl chloride initiator in Ph O solvent.
We have previously observed that the substituents on the
Schiff base ligand play an important role in determining the
position of the equilibrium between dormant and active poly-
mer chains, which is central to the reaction mechanism30
(Scheme 1). The 1-phenylethyl substituent used in this case is
more sterically demanding than an n-alkyl group and this may
be the reason for the poor molecular weight control observed
when the reaction is performed in xylene. The steric require-
ments of the ligands are clearly an important factor in the
relative stabilities of CuI and CuII, as are electronic factors.
When the initiator is changed to 4-methoxybenzenesulfonyl
chloride 7 with diphenyl ether as solvent, linear first order rate
plots with an apparent induction period of approximately
30 min are obtained (Fig. 3). The rate of reaction is again
2
The procedure detailed above was followed, except that Ph O
2
(5.0 ml) was substituted for xylene and 7 (0.095 g, 4.65×10−4
mol) was used as the initiator.
Polymer analysis. Polymer conversion was determined by
gravimetry for polymerisations performed in xylene and by 1H
NMR spectroscopy in the case of reactions performed in Ph O.
2
At the end of the reaction the polymer was precipitated into
light petroleum (bp 40–60 °C) and dried in a vacuum oven at
70 °C for 8 h, for stereochemical analysis by 13C NMR spec-
troscopy. Molecular weight distributions were measured using
size exclusion chromatography (SEC) on a system equipped
with a guard column, one 30 cm mixed E column (Polymer
Laboratories) and a differential refractive index detector, using
tetrahydrofuran at 1 ml min−1 as the eluent. Poly(methyl
methacrylate) standards in the range (6×104 to 200 g mol−1)
were used to calibrate the SEC analysis. Quantitative 13C
NMR spectra were obtained on a Bruker AM400, using a
Bruker inverse gated acquisition routine, for the analysis of
polymer microstructure.
independent of the catalyst stereochemistry. M increases with
conversion and shows much better agreement with that
n
expected for 100% initiator efficiency (Fig. 4) compared with
the previous case. Thus, the initiator efficiency is higher with
7 than 6. Also, the PDIs are much narrower in all cases (ca.
1.2) and PDI narrows as conversion increases (Table 2). Both
of the results are consistent with Percec’s20 suggestion that the
rate of initiation of MMA using 7 is fast relative to the rate of
propagation which results in narrower PDIs.
The rate of reaction using 7 in diphenyl ether is slightly
lower (ca. 30%) than that using 6 in xylene (Table 3). This
difference in rate is most likely due to the effect of the solvent
on the position of the equilibrium shown in Scheme 1, resulting
in a different concentration of active species and hence a
change in rate. Diphenyl ether may act as a weakly coordinat-
ing solvent which will naturally affect the nature and/or rate
of formation of the active copper species. There may also be a
small difference due to the different initiators used. A lower
concentration of active species would result in a lower rate of
termination due to normal radical–radical reactions and this,
coupled with the faster rate of initiation20 from 7, results in
better molecular weight control.
The polymerisation of MMA in xylene, mediated by cop-
per() catalysts made in situ by mixing CuIBr and (R)-3 or (S)-
3 [resulting in the catalyst (R)-S or (S)-5] and using 6 as the
initiator again gave reproducible rates of polymerisation
(Fig. 5). The rate of polymerisation is independent of the
catalyst stereochemistry and the first order kinetic plot is linear
Results and Discussion
Polymerisations
The polymerisation of MMA with the three forms of 4 [(i.e.
(R)-4, (S)-4 and ( )-4] and with 6 as initiator in xylene at
90 °C gave reproducible number average molecular masses,
M , with PDI between 1.4 and 1.6 (Table 1). In each case the
n
first order rate plot (Fig. 1) shows a linear slope, indicating
the number of active species remains constant throughout the
reaction and that the rate of polymerisation is independent of
the stereochemistry of the catalyst. Each experiment exhibited
an apparent induction period of approximately 35 min. These
induction periods have been observed previously in similar
systems and are ascribed to the time required to establish the
equilibrium shown in Scheme 1. It should be noted that the
apparent induction time may actually be S-shaped due to a
slowly increasing rate of reaction during the time taken to
establish the equilibrium. Although M increases with conver-
sion with all three catalysts, as expected for a living polymeris-
ation, the actual values are consistently higher than expected
for 100% initiator efficiency (Fig. 2). In addition, the PDI is
relatively broad (1.4–1.6 as compared to 1.2 for similar sys-
n
with an apparent 30 min induction period (Fig. 5). M increases
linearly with conversion (Fig. 6), but is slightly higher than
n
the theoretical M , similar to the behaviour seen with 4. The
rate of polymerisation is slightly slower than with 4 (Table 3),
n
J. Mater. Chem., 1998, 8(7), 1525–1532
1527

Table 1 Experimental results for the polymerisation of MMA using ATP with (R)-4, (S)-4 and ( )-4 and ethyl 2-bromo-2-methylpropanoate
initiator 6 in xylene at 90 °C
(R)-4
(S)-4
( )-4
t/min
conversion
M
PDI
conversion
M
PDI
conversion
M
PDI
n
n
n
60
120
180
240
300
360
0.051
0.152
0.242
0.331
0.408
0.505
3200
4030
4310
4510
5420
6430
1.40
1.41
1.46
1.52
1.42
1.48
0.060
0.179
0.249
0.333
0.412
0.503
4030
4780
5010
5420
6200
6570
1.42
1.47
1.65
1.54
1.46
1.44
0.056
0.180
0.250
0.370
2290
3250
4340
5320
1.38
1.38
1.63
1.51
Fig. 1 First order rate plot for the polymerisation of MMA by ATP
using (&) (R)-4, (#) (S)-4 and (+) ( )-4 catalysts at 90 °C in xylene
using 6 as an initiator
Fig. 3 First order rate plot for the polymerisation of MMA by ATP
using (&) (R)-4, (#) (S)-4 and (+) ( )-4 catalysts at 90 °C in diphenyl
ether using 7 as initiator
Fig. 4 Dependence of molecular weight on conversion for the
polymerisation of MMA by ATP using (&) (R)-4, (#) (S)-4 and (+)
Fig. 2 Dependence of molecular weight on conversion for the
polymerisation of MMA by ATP using (&) (R)-4, (#) (S)-4 and (+)
(
)-4 catalysts at 90 °C in diphenyl ether using 7 as an initiator
(
)-4 catalysts at 90 °C in xylene using 6 as initiator
Polymer microstructure
however the PDI is narrower (Table 4). When 7 is used as the
initiator in diphenyl ether solvent a broadening of PDI is
observed at longer reaction times which may to be due to
termination reactions. The rate of reaction is significantly
lower than in the other systems and may suggest that the
equilibrium shown in Scheme 1 is not positioned for effective
ATP. The first order rate plot shows that the rate of reaction
is unaffected by the stereochemistry of the catalyst (Fig. 7).
The stereochemistry of polymers from all polymerisations was
determined by analysis of 13C NMR spectra (triad region,
44–46 ppm, corresponding to the backbone quaternary
carbon), according to the assignment of Peat and Reynolds.31,32
Table 8 shows the triad fraction, diad fraction and persistence
ratio r [eqn. (1)] of PMMA prepared by ATP with (R)-4, (S)-
4, ( )-4, (R)-5, (S)-5 and a non-chiral ATP ligand, and for
conventional free-radical polymerisation. The stereoregularity
is similar in each case, indicating that the use of these chiral
Schiff base ligands has no significant influence on the stereo-
chemistry of monomer addition. The lack of stereochemical
control can be explained in several ways: (i) the chiral centre
on the ligand may be too far away from the monomer addition
Again M increases with conversion as would be expected for
a controlled polymerisation (Fig. 8). A difference in the poly-
n
merisations mediated by 4 and 5 is that the counter ion to the
copper is BF − and Br−, respectively. This may influence the
4
polymerisation reaction since the Br− may coordinate to the
copper complex, whereas the BF − is non coordinating.
4
1528
J. Mater. Chem., 1998, 8(7), 1525–1532

Table 2 Experimental results for the polymerisation of MMA using ATP with (R)-4, (S)-4 and ( )-4 and 4-methoxybenzenesulfonyl chloride 7 as
initiator in Ph O at 90 °C
2
(R)-4
(S)-4
( )-4
t/min
conversion
M
PDI
conversion
M
PDI
conversion
M
PDI
n
n
n
60
120
180
240
300
0.06
0.11
0.19
0.25
0.28
1170
1590
1820
2130
2320
1.36
1.30
1.27
1.26
1.21
0.04
0.10
0.18
0.26
0.35
860
1220
1690
2570
2970
1.36
1.25
1.20
1.21
1.22
0.04
0.10
0.20
0.26
0.31
640
990
1930
2070
2360
1.43
1.37
1.27
1.16
1.17
Table 3 Comparison of the slopes from first order rate plots of different
systems
Table 4 Experimental results for the polymerisation of MMA using
ATP with (R)-5 and (S)-5 and ethyl 2-bromo-2-methylpropanoate
initiator 6 in xylene at 90 °C
k [RΩ]/s−1
p
(R)-5
(S)-5
catalyst
6
7
t/min
conversion
M
PDI
conversion
M
PDI
n
n
(R)-4, (S)-4, ( )-4
(R)-5, (S)-5
3.46×10−5
2.37×10−5
2.35×10−5
0.52×10−5
60
1140
1440
0.081
0.784
0.895
2070
9480
11800
1.15
1.42
1.31
0.052
0.753
0.857
1300
8780
12500
1.13
1.44
1.37
Fig. 5 First order rate plot for the polymerisation of MMA by ATP
using (&) (R)-5 and (#) (S)-5 catalysts at 90 °C in xylene using 6 as
an initiator
Fig. 7 First order rate plot for the polymerisation of MMA by ATP
using (&) (R)-5 and (#) (S)-5 catalysts at 90 °C in diphenyl ether
using 7 as initiator
Fig. 6 Dependence of molecular weight on conversion for the
polymerisation of MMA by ATP using (&) (R)-5 and (#) (S)-5
catalysts at 90 °C in xylene using 6 as initiator
Fig. 8 Dependence of molecular weight on conversion for the
polymerisation of MMA by ATP using (&) (R)-5 and (#) (S)-5
catalysts at 90 °C in diphenyl ether using 7 as an initiator
site to play a role; (ii) the catalyst may not be rigid enough to
exert significant stereocontrol on the growing polymer chain
since the chirality is in a very flexible ligand (also the complexed
ligand is in dynamic equilibrium with the ligand in solution30);
(iii) the ATP reaction may be purely free-radical and so the
catalyst would play no part in the monomer addition step. We
anticipate that a more strongly complexed and rigid chiral
Schiff base ligand may be able to provide some degree of
stereocontrol in ATP and this is currently under study in our
laboratory.
r=2(m)(r)/(mr)
(1)
J. Mater. Chem., 1998, 8(7), 1525–1532
1529

Table 5 Experimental results for the polymerisation of MMA using
ATP with (R)-5 and (S)-5 and 4-methoxybenzenesulfonyl chloride 7 as
Table 7 Summarised crystallographic data for (R)-4
Crystal parameters
initiator in Ph O at 90 °C
2
Formula
M
C H Cu B F N
56 56 2 2 8
8
(R)-5
(S)-5
1141.79
Tetragonal
P4(3)
16.9382(4)
16.9382(4)
19.5734(5)
90
Crystal system
Space group
t/min
conversion
M
PDI
conversion
M
PDI
n
n
˚
a/A
˚
60
120
180
780
0.005
0.011
0.052
—
484
875
943
—
1.29
1.24
1.23
—
0.004
0.01
—
0.221
0.313
698
1040
—
2420
3890
1.16
1.23
—
1.41
1.66
b/A
˚
c/A
a(°)
b(°)
90
90
1320
0.334
3170
1.53
c(°)
reflections
8192
5615.7(2)
4
0.4×0.4×0.3
1.350
0.828
˚
U/A3
Z
dimensions /mm
D /g cm−3
T /K
c
m(Mo-Ka/mm−1
180
Data collectiona
Data collected h
−17 to 21
−22 to 22
−26 to 20
34133
12021
6965
k
l
total reflections
independent reflections
observed reflections [F ≥4s(F )]
h range (°)
F(000)
Refinement
Rb
o
o
1.20 to 28.58
2352
0.0680
0.1999
0.914
1.062, −0.340
1.000, 0.7479
0.03(2)
wR2c
S
˚
D/e A−3 (max, min)d
Flack parameter
T (max, min)e
Fig. 9 Crystal structure of the ATP catalyst (R)-4 used in polymeris-
ation studies. The BF − counter ion is omitted for clarity.
weighting scheme a,bf
0.087, 12.428
4
aData collected on a Siemens 3 circle diffractometer equipped with a
SMART CCD area detector; graphite-monochromated Mo-Ka radi-
Crystal structure discussion
˚
ation (l=0.71073 A). bR=S|F −F |SF [for F ≥4s(F )]. cwR2=
o
c
o
o
o
[S[w(F 2−F 2)2]/S[w(F 2)2]]1/2 for all data. dPeaks of unassigned
Diazobutadiene (DAB) ligands have been shown to stabilise
low formal oxidation states of the transition metals.33 The
pyridylmethanimine ligands 2 and 3 used in this work have
similar electronic characteristics to DAB type ligands, with
bipyridine based ligands also thought to be similar. In general,
it has been assumed in ATP that the added catalyst contains
copper with a formal oxidation number of +1 since it is
produced from the complexation of CuI and neutral ligands;
however, this may be an oversimplification. Recent studies
of complexes of ytterbium and tert-butyldiazabutadiene
(ButDAB) have shown that when zero valent Yb0 is ligated by
three ButDAB ligands, although the overall complex is neutral,
the formal oxidation state of the Yb metal centre is +3, with
o
c
o
residual electron density. eBy SADABS. fw−1=s2(F 2)+aP+bP,
o
where P=[max(F 2,0)+2F 2]/3, where max(F 2,0) indicates that the
o
c
o
larger of F 2 or 0 is taken, a and b are values set by the program.
o
each ButDAB ligand having
a formal charge of −1,
YbIII(ButDAB−) .34 This difference in the behaviour of DAB
3
ligands, on the one hand stabilising low oxidation states and
on the other causing one electron oxidation of the metal by
the ligand, prompted us to investigate the actual oxidation
state of copper complexes used in this work by examination
of the structure of the complex using X-ray crystallography.
The crystal structure of (R)-4 was determined and it was
Table 6 Comparison of fractions of diads and persistence ratio of poly(MMA) prepared under different conditions
stereochemistryb
initiatora
catalyst
mm
mr
rr
m
r
r
6
6
6
7
7
7
6
6
(R)-4
(S)-4
( )-4
(R)-4
(S)-4
( )-4
(R)-5
(S)-5
(R)-5
(S)-5
8c
0.0315
0.0349
0.0357
0.0256
0.0322
0.0337
0.0440
0.0354
0.0269
0.0310
2.77
0.362
0.367
0.374
0.378
0.372
0.367
0.385
0.373
0.381
0.407
34.8
0.607
0.598
0.590
0.597
0.596
0.600
0.572
0.592
0.593
0.562
62.4
0.212
0.219
0.223
0.214
0.218
0.217
0.236
0.222
0.217
0.235
0.202
0.231
0.788
0.781
0.777
0.786
0.782
0.783
0.764
0.778
0.783
0.766
0.798
0.769
0.925
0.930
0.926
0.892
0.917
0.927
0.939
0.926
0.893
0.882
0.925
0.973
7
7
6
AIBNd
—
4.82
36.5
58.7
a6, reaction in xylene; 7, reaction in diphenyl ether; AIBN=2,2∞-azoisobutyronitrile, reaction in toluene. bmm=isotactic triads, mr=atactic triads,
rr=syndiotactic triads, m=meso diads, r=racemic diads. cATP with nonchiral catalyst, CuIBr and N-n-pentyl-2-pyridylmethanimine ligand 8.
The ratio of MMA5CuIBr5N-n-pentyl-2-pyridylmethanimine: ethyl 2-bromo-2-methylpropanoate was 100515251, reaction temperature=90 °C.
dConventional free-radical polymerisation: 25% w/w MMA in xylene, [AIBN]=0.027 mol l−1, 90 °C, precipitated from MeOH, M =14600,
n
PDI=1.74.
1530
J. Mater. Chem., 1998, 8(7), 1525–1532

Table 8 Comparison of average bond length and angle data for [CuI(L) ]+ complexes
2
˚
bond length/A
angle (°)
counter
ion
ligand
CMN
CMN
CMC
CuMN NMCuMN dihedral reference
acyclic
cyclic
N-(1-phenylethyl)-2-pyridylmethanimine (R)-4a,b BF −
1.282
1.273
1.270(4)
1.263(4)
1.470(4)
1.268(7)
1.341
1.370
1.349(4)
1.352(4)
2.035
1.457
1.456
1.467(5)
2.040
2.030
2.035
81.8
82.3
81.89(10)
90.3
94.9
81.9
this work
37
4
N-tert-butyl-2-pyridylmethanimine
BF −
4
81.84(10)
1.453(8)
1.440(15) 2.021(11) 81.5(4)
1.488
1.428
1.450
1.39(3)
tert-butyldiazabutadiene
2,2∞-bipyridine
4,4∞,6,6∞-tetramethyl-2,2∞-bipyridineb
1,10-phenanthrolineb
Br−
1.290(7)c
2.025(4)
82.2(2)
89.5
75.2
68
76.8
85.7
37
35
36
38
39
34
ClO −
4
1.325(14) 1.385(15)
1.347
Cl−
1.360
1.352
1.361
1.51(6)
2.040
2.039
2.063
81.35
82.2
83.4
CuBr − 1.366
—
2
2,9-dimethyl-1,10-phenanthrolineb
NO −
1.360
1.61(6)
3
YbIII(ButDAB−) c,d
3
aThere are two molecules in the asymmetric unit. bLengths are averaged (no estimated standard deviations for averaged data). cSecond acyclic
bond. dThe Yb complex is included as an example of a DAB ligand with a formal −1 charge.
found that the asymmetric unit contains two similar molecules,
one of which is shown in Fig. 9. The important characteristics
are summarised in Table 7. The coordinating nitrogen atoms
surround the metal centre in a distorted tetrahedral arrange-
ment with the intra-ligand dihedral angles, between the mean
planes defined by the central copper atom and each pair of
bidentate nitrogen atoms, measured at 90.3 and 94.9 °. The
We would like to thank the EPSRC (D. J. D. and D. K.
GR/K90364, GR/L10314, A. M. H. and A. J. S. studentships).
References
1
2
T. P. Davis, D. M. Haddleton and S. N. Richards, J. Macromol.
Sci.-Rev. Chem. Phys., 1994, C34, 243.
O. W. Webster, W. R. Hertler, D. Y. Sogah, W. B. Farnham and
T. V. Rajan Babu, J. Am. Chem. Soc., 1983, 105, 5706.
O. W. Webster, Science, 1991, 887.
˚
CuMN distances range from 2.009(5)–2.050(6) A, with an
˚
average distance of 2.040 A. The CNN and CMC bond lengths
for (R)-4 are consistent with those found in similar neutral free
ligands. These bond lengths indicate that the ligands are
neutral and the tetrahedral structure of the complex indicates
that the formal charge of the copper centre is +1. Additional
material to that shown in Table 7, comprising the atomic
coordinates, thermal parameters and all bond lengths and
angles, is available from the Cambridge Crystallographic
Data Centre.
3
4
M. K. Georges, R. P. N. Veregin and G. K. Hamer, T rends Polym.
Sci., 1994, 2, 66.
5
M. K. Georges, R. P. N. Veregin and P. M. Kazmaier,
G. K. Homer, Macromolecules, 1993, 26, 2987.
D. Bellus, Pure Appl. Chem., 1985, 1827.
J. Iqbal, B. Bhatia and N. K. Nayyar, Chem. Rev., 1994, 94, 519.
B. Giese, Radicals in Organic Synthesis: Formation of Carbon-
Carbon Bonds, Pergamon, Oxford, 1988.
6
7
8
Table 8 compares bond length and angle data for compound
(R)-4 and for a selection of related complexes. It can be seen
that, for (R)-4, the CuMN bond distances and the NMCuMN
bond angles are comparable with similar compounds.
Interestingly, the CuMN1N2/CuMN3N4 dihedral angles deter-
mined for (R)-4 are considerably larger than those seen for the
related bipyridine,35,36 N-tert-butyl-2-pyridylmethanimine37
and phenanthroline38,39 complexes, whilst comparing well with
that determined for the similar tert-butyldiazabutadiene
(ButDAB) complex.37 Presumably, steric factors are important
due to the large 1-phenylethylamine group and packing forces,
such as ring stacking, may also play a part. The CMN and
CMC bond distances are very similar to those in the free
9
K. L. Salazar, M. A. Khan and K. M. Nicholas, J. Am. Chem. Soc.,
1997, 119, 9053.
10 M. Sawamoto and M. Kamigaito, T rends Polym. Sci., 1996, 4, 371.
11 M. Kato, M. Kamigaito, M. Sawamoto and T. Higashimura,
Macromolecules, 1995, 28, 1721.
12 T. Ando, M. Kamigaito and M. Sawamoto, T etrahedron, 1997,
53, 15445.
13 K. Matyjaszewski and J.-S. Wang, Macromolecules, 1995, 28, 7901.
14 K. Matyjaszewski, T. E. Patten and J. Xia, J. Am. Chem. Soc., 1997,
119, 674.
15 J. S. Wang and K. Matyjaszewski, J. Am. Chem. Soc., 1995, 117,
5614.
16 J. S. Wang and K. Matyjaszewski, Macromolecules, 1995, 28, 7901.
17 J. S. Wang, T. Grimaud and K. Matyjaszewski, Macromolecules,
1997, 30, 6507.
18 C. Granel, P. Teyssie, P. Dubois and R. Jerome, Macromolecules,
1996, 29, 8576.
19 V. Percec and B. Barboiu, Macromolecules, 1995, 28, 7970.
20 V. Percec, B. Barboiu, A. Neumann, J. C. Ronda and M. Y. Zhao,
Macromolecules, 1996, 29, 3665.
21 V. Percec, H.-J. Kim and B. Barboiu, Macromolecules, 1997, 30,
6702.
˚
ligand (1.283 and 1.496 A for ButDAB40) and normal
˚
˚
C(sp)2MN(sp)2 (1.27 A) and C(sp)2MC(sp)2 (1.48 A) bond
lengths. In contrast, the imine CMN and CMC bond lengths
of ButDAB− in the ytterbium complex are significantly longer
and shorter, respectively, since it is a −1 ligand.
22 D. M. Haddleton, C. B. Jasieczek, M. J. Hannon and A. J. Shooter,
Macromolecules, 1997, 30, 2190.
Conclusions
23 D. M. Haddleton, C. Waterson, P. J. Derrick, C. Jasieczek and
A. J. Shooter, Chem. Commun., 1997, 683.
24 D. M. Haddleton, A. J. Clark, M. C. Crossman, D. J. Duncalf,
A. M. Heming, S. R. Morsley and A. J. Shooter, Chem. Commun.,
1997, 1173.
25 H. Dieck and L. Stamp, Z. Naturforsch., T eil. B., 1990, 45, 1369.
26 R. N. Keller and H. D. Wycoff, Inorg. Synth., 1947, 2, 1.
27 G. J. Kubas, Inorg. Synth., 1989, 28, 68.
28 G. M. Sheldrick, SHEL XL 5.0, Siemens Analytical Instruments,
Madison, WI, 1994.
29 G. M. Sheldrick, SHELXL 96, University of Gottengen, 1996.
30 D. M. Haddleton and A. J. Shooter, Unpublished results.
31 I. R. Peat and W. P. Reynolds, T etrahedron L ett., 1972, 14, 1359.
32 J. S. Wang, R. Jerome, R. Warin and P. Teyssie, Macromolecules,
1993, 26, 5984.
This paper has shown that atom transfer polymerisation of
methyl methacrylate can be effectively performed using well
defined copper() complexes. It was found that the enantiomer-
ically pure chiral ligands used had no effect on the stereochemi-
stry of the resulting polymers; however, it may be possible via
the use of more strongly complexed and rigid chiral Schiff
base ligands. The use of a ligand with a methyl group b to the
imine nitrogen significantly broadens PDI and reduces control
over M when the reaction is performed in a nonpolar solvent,
such as xylene, compared to Schiff base ligands with n-alkyl
n
substituents. The control over M and PDI, however, can be
improved if the reaction is performed in a more polar/weakly
coordinating solvent such as diphenyl ether.
n
J. Mater. Chem., 1998, 8(7), 1525–1532
1531

33 G. van Koten and K. Vrieze, Adv. Organomet. Chem., 1982, 21, 151.
34 M. N. Bochkarev, A. A. Trifonov, F. G. N. Cloke, C. I. Dalby,
P. T. Matsunaga, R. A. Anderson, H. Schumann, J. Loebel and
H. Hemling, J. Organomet. Chem., 1995, 486, 177.
D. Kukulj and A. J. Shooter, J. Chem. Soc., Dalton T rans., 1998,
381.
38 P. C. Healy, L. M. Engelhardt, V. A. Patrick and A. White,
J. Chem. Soc., Dalton T rans., 1985, 2541.
35 M. Munakata, S. Kitagawa, A. Asahara and H. Masuda, Bull.
Chem. Soc. Jpn., 1987, 60, 1927.
36 J. F. Dobson, B. E. Green, P. C. Healy, C. H. L. Kennard,
C. Pakawatchai and A. H. White, Aust. J. Chem., 1984, 37, 649.
37 D. M. Haddleton, A. J. Clark, D. J. Duncalf, A. M. Heming,
39 M. Munakata, M. Maekawa, S. K. Kitagawa, S. Matsuyama and
H. Masuda, Inorg. Chem., 1989, 28, 4300.
40 I. Hargittai and R. Seip, Acta Chem. Scand., Ser. A, 1976, 30, 540.
Paper 8/00467F; Received 16th January, 1998
1532
J. Mater. Chem., 1998, 8(7), 1525–1532