Single-site β-diiminate zinc catalysts for the ring-opening polymerization of β-butyrolactone and β-valerolactone to poly(3-hydroxyalkanoates)

Article abstract of DOI:10.1021/ja020978r

Polymerization of β-butyrolactone (BBL) and β-valerolactone (BVL) using the zinc alkoxide initiator (BDI-1)ZnO/Pr [(BDI-1) = 2-((2,6-diisopropylphenyl)amido)-4-((2,6-diisopropylphenyl)imino)-2-pentene] proceeds very rapidly under mild conditions to produce poly(3-hydroxybutyrate) (PHB) and poly(3-hydroxyvalerate) (PHV), respectively. For the monomer-to-initiator ratio 2001, PHB number-average molecular weights (M_n) are proportional to conversion throughout the reaction and polydispersity indices (PDIs) are narrow, consistent with a living polymerization. Higher monomer-to-initiator ratios can be used to achieve high molecular weight PHB (Mn > 100 000). End-group analysis verifies that the polymerization of BBL follows a coordination-insertion mechanism, where complexes of the form (BDI-1)ZnOCH(Me)CH2-1)ZnOCH(Me)CH₂-CO₂R are the active species. Variable temperature NMR experiments show that (BDI-1)ZnO/Pr is monomeric in benzene-d₆ solution. In contrast, (BDI-2)ZnO/Pr [(BDI-2) = 2-((2,6-diethylphenyl)amido)-4-((2,6-diethylphenyl)imino)-2-pentene] is a poor initiator at room temperature because it prefers to form a bis-μ-isopropoxide dimer in solution. According to kinetic studies, propagation by (BDI-1)ZnO/Pr is first order in both monomer as well as (BDI-1)ZnO/Pr concentration. These results lead us to propose a monometallic active species. Several results suggest that elimination side reactions are slowly catalyzed by zinc alkoxides, leading to degradation of the polymer.

Full text of DOI:10.1021/ja020978r

Published on Web 11/28/2002
Single-Site â-Diiminate Zinc Catalysts for the Ring-Opening
Polymerization of â-Butyrolactone and â-Valerolactone to
Poly(3-hydroxyalkanoates)
Lee R. Rieth, David R. Moore, Emil B. Lobkovsky, and Geoffrey W. Coates*
Contribution from the Department of Chemistry and Chemical Biology, Baker Laboratory,
Cornell UniVersity, Ithaca, New York 14853-1301
Received July 17, 2002
Abstract: Polymerization of â-butyrolactone (BBL) and â-valerolactone (BVL) using the zinc alkoxide initiator
(BDI-1)ZnOiPr [(BDI-1) ) 2-((2,6-diisopropylphenyl)amido)-4-((2,6-diisopropylphenyl)imino)-2-pentene]
proceeds very rapidly under mild conditions to produce poly(3-hydroxybutyrate) (PHB) and poly(3-
hydroxyvalerate) (PHV), respectively. For the monomer-to-initiator ratio 200:1, PHB number-average
molecular weights (Mn) are proportional to conversion throughout the reaction and polydispersity indices
(PDIs) are narrow, consistent with a living polymerization. Higher monomer-to-initiator ratios can be used
to achieve high molecular weight PHB (Mn > 100 000). End-group analysis verifies that the polymerization
of BBL follows a coordination-insertion mechanism, where complexes of the form (BDI-1)ZnOCH(Me)CH2-
CO2R are the active species. Variable temperature NMR experiments show that (BDI-1)ZnOiPr is monomeric
in benzene-d6 solution. In contrast, (BDI-2)ZnOiPr [(BDI-2) ) 2-((2,6-diethylphenyl)amido)-4-((2,6-dieth-
ylphenyl)imino)-2-pentene] is a poor initiator at room temperature because it prefers to form a bis-µ-
isopropoxide dimer in solution. According to kinetic studies, propagation by (BDI-1)ZnOiPr is first order in
both monomer as well as (BDI-1)ZnOiPr concentration. These results lead us to propose a monometallic
active species. Several results suggest that elimination side reactions are slowly catalyzed by zinc alkoxides,
leading to degradation of the polymer.
Introduction
(3-hydroxyalkanoates) (PHAs) render it impractical in many
applications.
Poly(3-hydroxybutyrate) (PHB) is an aliphatic polyester
produced by bacteria and other living organisms.1,2 This
biodegradable3-5 and biocompatible6 natural polymer is isotactic
with all stereocenters in the (R) configuration. Because isotactic
PHB is highly crystalline (Tm ≈ 180 °C) and has low
thermostability, melt processing is difficult, thereby limiting its
industrial importance. Recently developed fermentation pro-
cesses allow other â-hydroxy acids to be incorporated into the
polymer, thereby altering the thermomechanical properties of
the polymer and improving its processability.7 In fact, an
engineering plastic made from copolymer poly(3-hydroxybu-
tyrate-co-3-hydroxyvalerate), which possesses properties similar
to those of polypropylene, has been commercially developed.1,2,6
However, high production costs of naturally synthesized poly-
Ring-opening polymerization (ROP) of â-substituted â-pro-
piolactones offers another approach for producing PHAs. Most
efforts to date have focused on the ROP of â-butyrolactone
(BBL) to make PHB. Unlike bacteria-mediated polymerization,
which gives only isotactic PHB, controlled ROP of BBL allows
access to a variety of PHB microstructures. Racemic BBL has
been polymerized to make atactic PHB,8-16 as well as PHB
enriched in isotactic17-20 and syndiotactic14,21-25 diads. Optically
(8) Billingham, N. C.; Proctor, M. G.; Smith, J. D. J. Organomet. Chem. 1988,
341, 83-93.
(9) Leborgne, A.; Pluta, C.; Spassky, N. Macromol. Rapid Commun. 1994,
15, 955-960.
(10) Mo¨ller, M.; Kånge, R.; Hedrick, J. L. J. Polym. Sci., Part A 2000, 38,
2067-2074.
(11) Schechtman, L. A.; Kemper, J. J. PCT Int. Appl., WO 0077072, 2000;
Chem. Abstr. 2001, 134, 57105.
(12) Schechtman, L. A.; Kemper, J. J. PCT Int. Appl., WO 9520616, 1995;
Chem. Abstr. 1995, 124, 9719.
* Address correspondence to this author. E-mail: gc39@cornell.edu.
(1) Mu¨ller, H.-M.; Seebach, D. Angew. Chem., Int. Ed. Engl. 1993, 32, 477-
502.
(2) Sudesh, K.; Abe, H.; Doi, Y. Prog. Polym. Sci. 2000, 25, 1503-1555.
(3) Doi, Y.; Kumagai, Y.; Tanahashi, N.; Mukai, K. In Biodegradable Polymers
and Plastics; Vert, M., Feijen, J., Albertsson, A., Scott, G., Chiellini, E.,
Eds.; The Royal Society of Chemistry: Cambridge, U.K., 1992; pp 139-
148.
(13) Schechtman, L. A.; Kemper, J. J. Polym. Prepr. (Am. Chem. Soc., DiV.
Polym. Chem.) 1999, 217, 508-509.
(14) Jedlinski, Z.; Kowalczuk, M.; Kurcok, P.; Adamus, G.; Matuszowicz, A.;
Sikorska, W.; Gross, R. A.; Xu, J.; Lenz, R. W. Macromolecules 1996,
29, 3773-3777.
(15) Connor, E. F.; Nyce, G. W.; Myers, M.; Mo¨ck, A.; Hedrick, J. L. J. Am.
Chem. Soc. 2002, 124, 914-915.
(16) Kurcok, P.; Smiga, M.; Jedlinski, Z. J. Polym. Sci., Part A 2002, 40, 2184-
2189.
(4) Cox, M. K. In Biodegradable Polymers and Plastics; Vert, M., Feijen, J.,
Albertsson, A., Scott, G., Chiellini, E., Eds.; The Royal Society of
Chemistry: Cambridge, U.K., 1992; pp 95-99.
(17) Bloembergen, S.; Holden, D. A.; Bluhm, T. L.; Hamer, G. K.; Marchessault,
R. H. Macromolecules 1989, 22, 1656-1663.
(5) Okada, M. Prog. Polym. Sci. 2002, 27, 87-133.
(18) Leborgne, A.; Spassky, N. Polymer 1989, 30, 2312-2319.
(19) Lenz, R. W.; Yang, J.; Wu, B.; Harlan, C. J.; Barron, A. R. Can. J.
Microbiol. 1995, 41, 274-281.
(6) Holmes, P. A. Phys. Technol. 1985, 16, 32-36.
(7) Holmes, P. A. In DeVelopments in Crystalline Polymers-2; Bassett, D. C.,
Ed.; Elsevier Applied Science: London, 1988; pp 1-66.
(20) Wu, B.; Lenz, R. W. Macromolecules 1998, 31, 3473-3477.
9
10.1021/ja020978r CCC: $22.00 © 2002 American Chemical Society
J. AM. CHEM. SOC. 2002, 124, 15239-15248
15239

A R T I C L E S
Rieth et al.
Scheme 1. Synthesis of (BDI)Zn Alkoxide Complexes
pure BBL can also be polymerized to make highly isotactic
PHB.11-14,16,26-28 With the exception of recently reported
distannoxane21,26 and alkylzinc alkoxide11-13 catalysts, most
systems previously studied for the ROP of BBL are extremely
slow and/or are not capable of producing high molecular weight
PHB (Mn > 100 000) in a controlled manner.
We recently reported that zinc complexes of bulky â-diimi-
nate (BDI) ligands are very active catalysts for the living,
stereoselective ROP of lactide to produce poly(lactic acid)29-31
and living copolymerization of carbon dioxide and epoxides to
make polycarbonates.32-35 Herein, we report that bulky â-di-
iminate zinc alkoxide complexes are also excellent catalysts for
the ROP of BBL and â-valerolactone (BVL). PHAs with narrow
polydispersities can be produced at unprecedented rates under
mild conditions.
Results and Discussion
Polymerization of rac-â-Butyrolactone. Our investigation
of the ROP of â-butyrolactone focused on zinc alkoxides (BDI-
1)ZnOiPr (1a) and (BDI-2)ZnOiPr (2a), which are easily
prepared in two steps from free BDI ligand (Scheme 1).29,30
Previous work on the ROP of lactide showed that OiPr groups
on catalyst 1a directly initiate polymerization at rates competi-
tive with propagation.29,30 The polymerization of racemic BBL
using 1a and 2a produces atactic PHB (eq 1); 13C NMR shows
equal intensity of racemic and meso diad signals of the carbonyl
peak.36 Results of selected polymerizations are summarized in
Table 1. Selected Results of Polymerizations of â-Substituted
â-Propiolactones with (BDI)Zn Alkoxidesa
temp time conversion M (× 10-3
(°C) (min)
)
M /M
(GPC)d
n
w
n
entry monomer cat.
M/Ib
(%)c
(GPC)d
1
2
3
4
5
6
rac-BBL 1a
rac-BBL 1a
rac-BBL 1a
rac-BBL 1a
200 23
220 50
220 75
400 23
70
20
5
150
480
720
70
91
97
94
91
94
90
65
86
95
92
94
84
88
67g
26.1
29.0
24.8
47.0
99.5
144
17.1
23.0
200
48.8
33.7
24.7
12.4
9.0
1.06
1.10
1.14
1.07
1.13
1.20
1.05
1.08
1.70
1.30
1.35
1.47
1.15
1.04
rac-BBL 1a 1000 23
rac-BBL 1a 2000 23
7e rac-BBL 1a
200 23
200 23
200 23 2480
200 50
200 75
200 23
150 23
150 23
8f rac-BBL 1a
9
10 rac-BBL 2a
11 rac-BBL 2a
12 rac-BBL 2b
13 rac-BVL 1a
14 (R)-BBL 1a
70
rac-BBL 2a
60
5
70
120
30
Table 1. When the [BBL]/[Zn] ratio equals 200 (entry 1),
polymerization with 1a surpasses 90% conversion in just over
a All polymerizations run in benzene-d6 except entries 7 (CH2Cl2) and 8
(THF); [lactone] ) 2.45 M. b M/I ) [lactone]/[Zn]. c As determined by the
integration of 1H NMR methine resonances of BBL (or BVL) and PHB (or
PHV); polymerizations were quenched with acetic acid-d4. d Determined
by gel permeation chromatography calibrated with polystyrene standards
in THF. e Polymerization performed in CH2Cl2. f Polymerization performed
in THF. g Polymer precipitates because of insolubility in benzene-d6.
(21) Hori, Y.; Hagiwara, T. Int. J. Biol. Macromol. 1999, 25, 237-245.
(22) Kemnitzer, J. E.; McCarthy, S. P.; Gross, R. A. Macromolecules 1993, 26,
6143-6150.
(23) Kemnitzer, J. E.; McCarthy, S. P.; Gross, R. A. Macromolecules 1993, 26,
1221-1229.
(24) Kricheldorf, H. R.; Lee, S.-R.; Scharnagl, N. Macromolecules 1994, 27,
3139-3146.
(25) Kricheldorf, H. R.; Eggerstedt, S. Macromolecules 1997, 30, 5693-5697.
(26) Hori, Y.; Suzuki, M.; Yamaguchi, A.; Nishishita, T. Macromolecules 1993,
26, 5533-5534.
1 h at room temperature. The number-average molecular weight
(Mn) of the polymer (measured by GPC relative to polystyrene
standards in THF) grows linearly with conversion and polydis-
persities (PDIs) are very narrow (Figure 1), suggesting that the
polymerization is living. At higher temperatures, polymerizations
with 1a proceed more quickly, as expected. At 50 °C, polym-
erization at [BBL]/[Zn] ) 220 reaches 97% conversion in 20
min (entry 2), while, at 75 °C, the same polymerization reaches
94% conversion in 5 min (entry 3). PDIs remain narrow within
all tested temperature ranges.
(27) Tanahashi, N.; Doi, Y. Macromolecules 1991, 24, 5732-5733.
(28) Zhang, Y.; Gross, R. A.; Lenz, R. W. Macromolecules 1990, 23, 3206-
3212.
(29) Chamberlain, B. M.; Cheng, M.; Moore, D. R.; Ovitt, T. M.; Lobkovsky,
E. B.; Coates, G. W. J. Am. Chem. Soc. 2001, 123, 3229-3238.
(30) Cheng, M.; Attygalle, A. B.; Lobkovsky, E. B.; Coates, G. W. J. Am. Chem.
Soc. 1999, 121, 11583-11584.
(31) For other recent work on the polymerization of lactide with â-diiminate
metal alkoxides, see: (a) Chisholm, M. H.; Gallucci, J.; Phomphrai, K.
Inorg. Chem. 2002, 41, 2785-2794. (b) Chisholm, M. H.; Huffman, J. C.;
Phomphrai, K. J. Chem. Soc., Dalton Trans. 2001, 222-224. (c) Dove, A.
P.; Gibson, V. C.; Marshall, E. L.; White, A. J. P.; Williams, D. J. Chem.
Commun. 2001, 283-284.
(32) Cheng, M.; Moore, D. R.; Reczek, J. J.; Chamberlain, B. M.; Lobkovsky,
E. B.; Coates, G. W. J. Am. Chem. Soc. 2001, 123, 8738-8749.
(33) Cheng, M.; Lobkovsky, E. B.; Coates, G. W. J. Am. Chem. Soc. 1998,
120, 11018-11019.
(34) Allen, S. D.; Moore, D. R.; Lobkovsky, E. B.; Coates, G. W. J. Am. Chem.
Soc., in press.
(35) Moore, D. R.; Cheng, M.; Lobkovsky, E. B.; Coates, G. W. Angew. Chem.,
Int. Ed. Engl. 2002, 41, 2599-2602.
(36) Gross, R. A.; Zhang, Y.; Konrad, G.; Lenz, R. W. Macromolecules 1988,
21, 2657-2668.
Molecular weights of the polymer can be increased by altering
[BBL]/[Zn] ratios. When [BBL]/[Zn] ) 400 (entry 4), polym-
erization with 1a yields BBL with Mn ) 47 000 at 91%
conversion, while, for [BBL]/[Zn] ) 1000 (entry 5), Mn )
99 500 at 94% conversion. High molecular weight PHB (Mn >
100 000) is achieved by polymerizing BBL with 1a at
9
15240 J. AM. CHEM. SOC. VOL. 124, NO. 51, 2002

Ring-Opening Polymerization of â-Butyrolactone
A R T I C L E S
Figure 1. Plot of PHB molecular weight (Mn, GPC vs polystyrene
standards) and polydispersity index (PDI) as a function of conversion using
rac-BBL monomer and (BDI-1)ZnOiPr (1a) catalyst ([BBL]0 ) 2.45 M,
[BBL]/[Zn] ) 200).
Figure 2. X-ray crystal structure of 2b. Selected bond distances (Å) and
angles (deg): Zn(1)-N(1), 2.022(3); Zn(1)-N(2), 2.019(3); Zn(1)-O(1),
1.971(3); Zn(1)-O(2), 2.052(2); Zn(1)-O(3), 2.459(3); C(1)-O(1), 1.398-
(4); C(3)-O(3), 1.206(5); O(1)-Zn(1)-N(2), 119.40(1); N(1)-Zn(1)-N(2),
95.59(1); O(1)-Zn(1)-O(2), 76.84(1); N(1)-Zn(1)-O(2), 105.58(1);
N(2)-Zn(1)-O(3), 107.73(1); O(1)-Zn(1)-O(3), 72.79(1); Zn(1)-O(1)-
Zn(2), 101.29(1); C(1)-O(1)-Zn(1), 124.1(2).
[BBL]/[Zn] ) 2000 at room temperature (entry 6). As [BBL]/
[Zn] is increased, however, PDIs broaden and molecular weights
are not linearly proportional with [BBL]/[Zn] ratios. For
instance, comparing entries 1 and 6 shows that the [BBL]/[Zn]
ratio has increased by a factor of 10, but Mn has only increased
by a factor of 5.5. Possible explanations for this observation
are discussed later (see Side Reactions).37
faster at higher temperatures than at room temperature but still
slower than propagation.
Solvent effects were examined by comparing rates of po-
lymerization by 1a in benzene-d6, methylene chloride (CH2-
Cl2), and tetrahydrofuran (THF). At [BBL]/[Zn] ) 200,
polymerization in benzene-d6 (entry 1) reached 91% conversion
in 70 min, while polymerization in CH2Cl2 (entry 7) and THF
(entry 8) achieved only 65% and 86% conversion, respectively,
in the same time period. As a result, all other polymerizations
were conducted in benzene-d6.
In an attempt to facilitate initiation, we sought to synthesize
a (BDI-2)Zn alkoxide complex with an initiator that more aptly
mimics the putative propagating species. Efforts to prepare
(BDI-2)ZnOCH(Me)CH2CO2Me from a reaction of (BDI-2)-
ZnN(SiMe3)229 and methyl 3-hydroxybutyrate failed to generate
clean, X-ray quality crystals of the desired product. However,
a reaction of (BDI-2)ZnN(SiMe3)2 with 1 equiv of racemic
methyl lactate yielded (BDI-2)ZnOCH(Me)CO2Me (2b) in 48%
yield after crystallization. Single-crystal X-ray diffraction
revealed that 2b adopts a dimeric structure in the solid state
(Figure 2), in contrast to monomeric (BDI-1)ZnOCH(Me)CO2-
Me.29 The five-coordinate zinc center of 2b exhibits a highly
distorted trigonal bipyramidal geometry. Dative bonds are
observed between the zinc centers and the carbonyl oxygen of
the methyl lactate initiating group. The bond length of Zn(1)-
O(3) is 2.46 Å, while the bond angle between O(2)-Zn(1)-
O(3) is 139.97°, thus illustrating the distorted geometry around
the zinc. The Zn‚‚Zn separation is 3.10 Å, and the six-membered
chelate is slightly distorted from planar to a boat-shaped
conformation, with a Zn deviation of 0.68 Å from the plane
defined by N(1)-N(2)-C(7). The Zn-O bond lengths within
the four-membered ring of core atoms are 1.97 and 2.05 Å,
and the bond angles of N(1)-Zn(1)-N(2) and O(1)-Zn(1)-
O(2) are 95.59° and 76.84°, respectively.
The polymerization of BBL with catalyst 2a at room
temperature produces polymer slowly, requiring over 40 h to
reach 95% conversion when [BBL]/[Zn] ) 200 (entry 9). In
addition, the polymerization with 2a displays poor control,
giving PDI ) 1.7 and Mn ) 200 000, approximately an order
of magnitude higher than the expected Mn. We attribute these
results to poor initiation by 2a at room temperature.
Increasing temperature in polymerizations with 2a has a
dramatic effect. Whereas the room temperature polymerization
with 2a described in the previous paragraph took over 40 h to
reach 95% conversion (entry 9), the same polymerization
([BBL]/[Zn] ) 200) at 50 °C took only 1 h to reach 92%
conversion (entry 10), and at 75 °C, the polymerization reaches
94% conversion in 5 min (entry 11). The molecular weights
were much lower (Mn ) 48 800 and 33 700, respectively),
although still somewhat higher than expected, and the molecular
weight distributions were narrower (PDI ) 1.30 and 1.35,
respectively). These results suggest that initiation is significantly
The polymerization of BBL using catalyst 2b reached 84%
conversion after 70 min at room temperature (entry 12). The
polydispersity index (PDI ) 1.47) is surprisingly broad, but
the molecular weight (Mn ) 24 700) suggests all Zn complexes
participate in the polymerization. Further details about initiation
and the solution behaviors of 2a and 2b are provided later.
Polymerization of rac-â-Valerolactone. â-Valerolactone
(BVL), prepared by the carbonylation of 1-butene oxide,38 is
(37) In general, GPC number-average molecular weights do not correspond well
with calculated theoretical molecular weights. In fact, for low [BBL]/[Zn]
ratios (e.g., entry 1), GPC molecular weights are somewhat higher than
calculated molecular weights, even though initiaton appears to be very fast
and efficient for 1a (see Initiation). We tentatively attribute this mismatch
to poor correlation between the polystyrene calibration of the GPC and
the actual molecular weights of the PHB chains. However, as the [BBL]/
[Zn] ratio increases, GPC molecular weights do not grow proportionately.
This could also be partly attributable to the GPC calibration, but there is
evidence that side reactions play a role in this unexpected behavior.
(38) Getzler, D. Y. L.; Mahadevan, V.; Lobkovsky, E. B.; Coates, G. W. J.
Am. Chem. Soc. 2002, 124, 1174-1175.
9
J. AM. CHEM. SOC. VOL. 124, NO. 51, 2002 15241

A R T I C L E S
Rieth et al.
also polymerized by 1a to make poly(3-hydroxyvalerate) (PHV)
(eq 2). Rates are somewhat slower than those for BBL, requiring
2 h to polymerize 150 equiv of BVL to 88% conversion at 23
°C (entry 13). Nonetheless, this result suggests that 1a can be
used to produce a variety of copolymers of BBL and BVL.39
Since it is reasonable to assume that BBL and BVL are
polymerized by the same mechanism, the remainder of the paper
will focus on the polymerization of BBL.40
Mechanism. It is well established that the ring-opening of
â-butyrolactone by metal alkoxides can proceed by two mech-
anisms.28,41 In a coordination-insertion mechanism, the BBL ring
is cleaved at the acyl-oxygen bond. The acid-quenched polymer
chains are terminated by ester and hydroxy groups, and the
stereochemistry of the chiral center is undisturbed (Scheme 2,
path a). In an anionic mechanism, the BBL ring is opened at
the alkyl-oxygen bond, leading to carboxylate and ether end-
groups, and the methine carbon undergoes inversion (or epimer-
ization) of configuration (Scheme 2, path b).
Figure 3. Carbonyl region of the 13C NMR spectra of PHB prepared by
the polymerization of (i) rac-BBL and (ii) (R)-BBL.
resulting polymer was shown to be highly isotactic by 13C NMR
(Figure 3b) and displays a Tm ) 155 °C (versus Tm ) 174 °C
for commercially supplied natural-origin poly-(R)-(3-hydroxy-
butyrate) with Mn ) 540 000). Polarimetry at 365 nm revealed
that the polymerization proceeds with retention of the (R)-
stereocenter,42 consistent with a coordination-insertion mech-
anism.
Scheme 2. Possible Routes for the Ring-Opening Polymerization
of â-Butyrolactone (BBL) Using Metal Alkoxide Catalysts: (a)
Coordination-Insertion Mechanism and (b) Anionic Mechanism
Solution Behaviors of (BDI-1)ZnOiPr and (BDI-2)ZnOiPr.
X-ray crystallography of 1a and 2a had previously revealed that
both complexes formed bis-µ-isopropoxide dimers in the solid
state.29,30 To establish the nature of the resting-state species in
the polymerization of BBL, an understanding of the solution
behaviors of 1a and 2a is required. Both 1a and 2a exhibit only
one set of shifts in benzene-d6 under polymerization conditions
(T ) 23 °C, and [Zn] e 12.3 mM); we sought to elucidate
whether the primary species in solution was a monomer or a
dimer.
Previous studies had established that (BDI-2)ZnOMe32 (2c)
exists primarily as a dimer in solution; mixing 1 equiv of (BDI-
2)ZnOMe with 1 equiv of (BDI-3)ZnOMe43,44 (3c), both of
which display only one set of shifts by themselves, leads to the
formation of a third, mixed ligand species after several days in
benzene-d6 solution.44 We extended this methodology to show
that 2a also resides primarily in dimeric form in solution. The
1H NMR spectrum of an approximately equimolar solution of
The polymerization of lactide by 1a was previously shown
to occur by a coordination-insertion route.29 To confirm that
the ROP of â-butyrolactone with 1a also occurs by a coordina-
tion-insertion route, polymerizations of rac-BBL were conducted
with low [BBL]/[Zn] ratios (e.g., [BBL]/[Zn] ≈ 25) (“oligo-
merization”) in order to identify endgroups by 1H NMR.
Oligomerizations run for 30 min at room temperature and
quenched with acetic acid yielded oligomers with isopropyl ester
and hydroxy endgroups, implicating a coordination-insertion
mechanism.
In addition, enantiomerically enriched (R)-BBL (ee ) 95%),
prepared by the carbonylation of (R)-propylene oxide,38 was
polymerized with 1a (Table 1, entry 14). As expected, the
(39) For some recent examples of the copolymerization of â-butyrolactone and
â-valerolactone, see: (a) Bloembergen, S.; Holden, D. A.; Bluhm, T. L.;
Hamer, G. K.; Marchessault, R. H. Macromolecules 1987, 20, 3086-3089.
(b) Bloembergen, S.; Holden, D. A.; Bluhm, T. L.; Hamer, G. K.;
Marchessault, R. H. Macromolecules 1989, 22, 1663-1669. (c) Kobayashi,
T.; Yamaguchi, A.; Hagiwara, T.; Hori, Y. Polymer 1995, 36, 4707-4710.
(d) Noda, I. PCT Int. Appl., WO 0037544, 2000; Chem. Abstr. 2000, 133,
75001. (e) Noda, I. U.S. Patent 6,077,931, 2000; Chem. Abstr. 2000, 133,
59245.
(40) A preliminary experiment shows that 1a also quantitatively polymerizes
2600 equivs of ꢀ-caprolactone in less than 1 min at room temperature to
produce poly(ꢀ-caprolactone) with Mn ) 320 000 and PDI ) 1.59. We are
in the process of optimizing this reaction.
(42) The specific rotation at 25 °C in CHCl3 for entry 14 was [R]25 ) +8.0
(c ) 0.0307 g/mL), compared to [R]25 ) +8.3 (c ) 0.025236g5/mL) for
365
natural origin (i.e., 100% isotactic) PHB. Both values are lower than
literature values ([R]25 ≈ +11 for natural origin PHB; see: Jedlinski,
365
Z.; Kowalczuk, M.; Kurcok, P.; Adamus, G.; Matuszowicz, A.; Sikorska,
W.; Gross, R. A.; Xu, J.; Lenz, R. W. Macromolecules 1996, 29, 3773-
3777). While magnitudes do not precisely correspond to literature values,
the fact that entry 14 rotates light in a positive direction indicates that
stereochemistry is conserved, while 13C NMR indicates that the degree of
isotacticity is very high (>95%).
(41) Lavallee, C.; Grenier, D.; Prud′homme, R. E.; Leborgne, A.; Spassky, N.
In AdVances in Polymer Synthesis; Culbertson, B. M., McGrath, J. E., Eds.;
Plenum Press: New York, 1985; pp 441-460 and references therein.
(43) Coates, G. W.; Cheng, M. PCT Int. Appl., WO 0008088, 2000; Chem Abstr.
2000, 132, 152332.
(44) Cheng, M. Ph.D. Thesis, Cornell University, 2000.
9
15242 J. AM. CHEM. SOC. VOL. 124, NO. 51, 2002

Ring-Opening Polymerization of â-Butyrolactone
A R T I C L E S
Figure 5. 1H NMR spectra of the â-diiminate methine proton of 1a ([Zn]
) 31.7 mM in toluene-d8) at (i) 20 °C, (ii) 10 °C, and (iii) 0 °C.
Figure 4. 1H NMR spectra of the â-diiminate methine protons in an
equimolar mixture of 2a and 2c after (i) 90 min at room temperature and
(ii) 17 h at 70 °C. Both spectra show the presence of a third species, the
mixed alkoxide dimer [2a-2c]. The bottom spectrum represents the
equilibrium mixture of species.
2a and 2c in benzene-d6 ([Zn]tot ) 0.056 M) showed the
presence of a third species 90 min after mixing at room
temperature. After the solution spent an additional 17 h at 70
°C, an equilibrium mixture of the three species had been reached,
with roughly the expected 1:2:1 ratio of species observed (Figure
4). The same ratio was observed after another 24 h at 70 °C,
allowing us to conclude that equilibrium had been reached.
In contrast, an equimolar mixture of 1a and 2a in benzene-
d6 did not give rise to a third species, even after 6 days at 70
°C. This led us to tentatively conclude that 1a is primarily a
monomer in solution. Low temperature 1H NMR studies
provided compelling further evidence that 1a resides in a
monomeric form in solution. We cooled a concentrated solution
of 1a ([Zn] ) 31.7 mM) in toluene-d8 from 20 °C to 0 °C,
recording 1H NMR spectra at 5 °C intervals. At 20 °C, a second
species is barely detectable. As temperature is decreased, the
second species becomes more prevalent (Figure 5). We attribute
the second species to the dimeric form of 1a. Plotting ln(Keq)
Figure 6. van’t Hoff plot for the monomer-dimer equilibrium of 1a in
toluene-d8 ([Zn] ) 31.7 mM).
versus 1/T (van’t Hoff plot) for each temperature reveals that
∆H° ) -18.8 ( 0.5 kcal/mol and ∆S° ) -64 ( 2 eu for the
monomer-dimer equilibrium (Figure 6).45 Under normal po-
(45) The variable temperature NMR data described in the text were combined
with an additional set of spectra recorded at 5 °C intervals from 25 °C to
5 °C with the same sample.
9
J. AM. CHEM. SOC. VOL. 124, NO. 51, 2002 15243

A R T I C L E S
Rieth et al.
Figure 8. van’t Hoff plot for the monomer-dimer equilibrium of 2b in
toluene-d8 ([Zn] ) 58.0 mM).
Figure 7. 1H NMR spectra of the methyl lactate methine proton of 2b
([Zn] ) 58.0 mM in toluene-d8) at (i) 20 °C, (ii) -10 °C, and (iii) -40 °C.
lymerization conditions, 1a resides almost exclusively in mon-
omeric form.
Similar variable temperature experiments with 2a failed to
reveal a second species despite looking at a wide range of
temperatures and concentrations. Nonetheless, the mixed ligand
experiments described above strongly suggest 2a prefers a
dimeric form under polymerization conditions.
Variable temperature 1H NMR experiments with 2b in
toluene-d8 revealed that it, like 1a, resides primarily as a
monomer in solution under polymerization conditions (12.3 mM,
23 °C) despite forming a bis-µ-alkoxide dimer in the solid state
(Figure 2). A 58.0-mM solution of 2b in toluene-d8 was cooled
Figure 9. 1H NMR spectra of the reaction of 2a with 1 equiv of BBL
after (i) 30 min and (ii) 6 h. After 30 min, only BBL (m) and unreacted 2a
are observed. After 6 h, a vast majority of unreacted 2a remains, but BBL
has been replaced with PHB (p).
1
from 20 °C to -40 °C in 10 °C intervals. The H NMR peaks
Scheme 3. Reaction of (BDI-1)ZnOiPr (1a) with 1 Equiv of BBL
i
corresponding to the methine protons of the methyl lactate
initiator are shown in Figure 7. Only one species is visible at
20 °C; we believe this species is a monomeric complex in which
the methyl lactate carbonyl oxygen is coordinated to the zinc
center. As the solution is cooled, however, two new peaks
become visible. We attribute these two peaks to the meso and
racemic forms of the bis-µ-alkoxide dimer, with the larger peak
corresponding to the rac-dimer. This assignment was cor-
roborated by resynthesizing (BDI-2)ZnOCH(Me)CO2Me with
Yields Monoinsertion Product (BDI-1)ZnOCH(Me)CH2CO2 Pr (1b)
1
(S)-methyl lactate; low temperature H NMR revealed that, in
this case, only one additional peak, corresponding to (S,S)-
[(BDI-2)ZnOCH(Me)CO2Me]2, grew in at 4.4 ppm as temper-
ature was reduced.
Initiation. To compare initiation rates of 1a and 2a, we ran
NMR-scale reactions of the Zn catalysts with 1 equiv of BBL
1
By adding the intensities of the two dimer peaks for 2b, we
calculated Keq for the monomer-dimer equilibrium at each
temperature. In this case, plotting ln(Keq) versus 1/T for the
temperature range 0 °C to -40 °C gave ∆H° ) -9.6 ( 0.1
kcal/mol and ∆S° ) -32.2 ( 0.3 eu for the monomer-dimer
equilibrium (Figure 8).
at room temperature. For 1a, the H NMR spectrum taken 15
min after the reagents were combined revealed that all the BBL
had been consumed, and the monoinsertion product 1b (Scheme
3) had formed exclusively; that is, peaks corresponding to the
polymer were not observed. This suggests that, for 1a, the rate
of initiation is greater than the rate of propagation. In contrast,
9
15244 J. AM. CHEM. SOC. VOL. 124, NO. 51, 2002

Ring-Opening Polymerization of â-Butyrolactone
A R T I C L E S
Figure 11. Plot of ln kapp versus ln[Zn] for the polymerization of BBL
with (BDI-1)ZnOiPr (1a) as initiator ([BBL]0 ) 2.45 M).
1.02 ( 0.02.46 Thus, the reaction is first order in zinc and
monomer concentrations, and the overall rate equation is
-d[BBL]/dt ) kp[BBL][Zn]
Since we established that 1a remains monomeric under polym-
erization conditions and can reasonably assume that the zinc
complex of the growing polymer chain is also monomeric, an
order in [Zn] ) 1 suggests that the ROP of BBL proceeds
through a monometallic transition state in the rate-determining
step. We thus propose the polymerization mechanism shown
in Scheme 4: the BBL monomer coordinates to the Zn center
of the monometallic complex; the alkoxide chain end attacks
the activated carbonyl carbon, cleaving the acyl-oxygen bond
and producing a new alkoxide chain end.
Complex 2a is a poor initiator for the ROP of BBL at room
temperature because it prefers to adopt the inactive dimeric form,
which inhibits coordination of the monomer and, therefore,
prevents polymerization. This limitation is minimized by using
higher temperature, which pushes the monomer-dimer equi-
librium toward the active monomer, or by using a “preinitiated”
Zn complex like 2b. Once an insertion has occurred, propagation
can proceed readily.
Side Reactions. As Table 2 indicates, PHB slowly degrades
under polymerization conditions after monomer has been
consumed. When BBL is polymerized by 1a to 99% conversion
at 23 °C, PDIs are still very narrow (entry 1). However, if the
reaction is allowed to continue for an additional 20 h at 23 °C
(entry 2) or is heated to 50 °C for 2 h (entry 3), the PDIs of the
polymer broaden and molecular weights decrease. In addition,
we have already noted that molecular weights of PHB synthe-
sized at different [BBL]/[Zn] ratios do not correlate well with
calculated theoretical molecular weights; as [BBL]/[Zn] is
increased, Mn(GPC) does not grow proportionally. Previous
investigations of lactone polymerization by metal alkoxides have
reported similar behavior.47-50 The most common explanation
for these observations has been transesterification and elimina-
Figure 10. First-order kinetic plots for the polymerization of BBL with
(BDI-1)ZnOiPr (1a) as initiator ([BBL]0 ) 2.45 M: (A) [BBL]/[Zn] )
100; (B) [BBL]/[Zn] ) 150; (C) [BBL]/[Zn] ) 200; (D) [BBL]/[Zn] )
250; (E) [BBL]/[Zn] ) 1000; (F) [BBL]/[Zn] ) 2000).
1
the H NMR spectrum taken 30 min after 2a and 1 equiv of
BBL were combined revealed primarily unreacted 2a and BBL.
After 6 h, the BBL had been consumed. But, a vast majority of
2a remained and polymer had formed (Figure 9). This experi-
ment suggests that initiation by 2a is extremely slow at room
temperature, and only a small percentage of 2a participates in
the polymerization.
Kinetics. A kinetic study was conducted in order to establish
reaction order in monomer and zinc for the polymerization of
1
BBL with 1a. Conversion of rac-BBL was monitored by H
NMR for various catalyst concentrations at room temperature
([BBL]/[Zn] ) 200-2000; [BBL] ) 2.45 M in benzene-d6). In
each case, polymerization proceeds with first-order dependence
on monomer concentration (Figure 10). Thus, the rate of
polymerization can be written as
-d[BBL]/dt ) kapp[BBL]
where kapp ) kp[Zn]x. Plotting ln kapp versus ln[Zn] allows us
to determine x, the order in Zn concentration, from the slope of
the fitted line (Figure 11). On the basis of this analysis, x )
(46) Interestingly, we had previously found an order in [Zn] of 1.56 for the
polymerization of rac-lactide by 1a; see ref 29. We are currently seeking
to understand the reason we obtained a higher order in catalyst for a
seemingly similar polymerization.
9
J. AM. CHEM. SOC. VOL. 124, NO. 51, 2002 15245

A R T I C L E S
Rieth et al.
Scheme 4. Proposed Mechanism for the ROP of BBL by
(BDI)ZnOR′ Catalysts
Figure 12. 1H NMR spectrum of the product of the oligomerization of
BBL ([BBL]/[Zn] ) 25) by 1a for 20 h at room temperature (after acidic
quench).
crotonate (and carboxy) endgroups.49 Tin-based catalysts have
also been shown to catalyze elimination.48,51,52
1H NMR studies of oligomerization reactions revealed that
1a also promotes the elimination reaction. As mentioned above,
oligomerizations run at room temperature for 30 min yield
oligomers with only isopropyl ester and hydroxy endgroups.
However, if the oligomerization is run at elevated temperature
(e.g., 75 °C) or for extended periods (e.g., 20 h), peaks consistent
1
Table 2. Polydispersity Broadening and Molecular Weight
Reduction after Monomer Consumption in Polymerizations of
rac-BBL with (BDI-1)ZnOiPr (1a)a
with trans-crotonate groups are also observed in the H NMR
(Figure 12).
To further investigate the effect of 1a on the elimination
reaction, we stirred 62 mg of commercially supplied natural-
origin poly-(R)-(3-hydroxybutyrate) ((R)-PHB) with 24.2 mg
of catalyst 1a (PHB repeat unit/Zn ) 16) in CHCl3 for 2 h at
50 °C. The proton NMR spectrum of the resulting product
clearly showed the presence of crotonate peaks, and GPC
revealed that the average molecular weight of the polymer had
been reduced from 540 000 to less than 5000.53 In contrast, (R)-
PHB stirred in CHCl3 at 50 °C without catalyst for 20 h
displayed no crotonate groups by NMR. This result, coupled
with the fact that this polymer remained insoluble in THF, led
us to conclude that elimination is minimal in the absence of
1a.
c
temp
(°C)
time
(min)
conversion
(%)
M (× 10-3
)
M /M
n
w
n
entry
cat.
M/Ib
(GPC)d
(GPC)d
1
2
3
1a
1a
1a
200 23
200 23
200 23/50 90/130
90
1200
99
100
100
27.6
25.3
24.6
1.09
1.20
1.21
a Polymerizations run in benzene-d6; [BBL] ) 2.45 M. b M/I ) [BBL]/
[Zn]. c As determined by the integration of H NMR methine resonances
1
of BBL and PHB; polymerizations were quenched with acetic acid-d4.
d Determined by gel permeation chromatography calibrated with polystyrene
standards in THF.
tion reactions promoted by the metal complex. Various zinc,13
aluminum,49 and tin10,50 alkoxide catalysts for the ROP of
â-butyrolactone have been reported to catalyze transesterifica-
tion. In addition, Je´roˆme and co-workers noted the presence of
crotonate groups in PHB prepared by the ROP of BBL with
Al(OiPr)3; they proposed that the metal catalyst promotes
elimination, which leads to chain cleavage and formation of
The ability of 1a to promote elimination reactions was further
demonstrated by treating 1a with 1 equiv of model compound
(51) Melchiors, M.; Keul, H.; Ho¨cker, H. Macromol. Rapid Commun. 1994,
15, 497-506.
(52) Melchiors, M.; Keul, H.; Ho¨cker, H. Macromolecules 1996, 29, 6442-
6451.
(53) The commercially prepared PHB was not soluble in THF, the eluent for
our GPC instrument. Molecular weight data were supplied by Fluka, Mn
) 540 000, but no polydispersity data were available. Stirring the
commercial polymer with 1a in CHCl3, as described, reduced the molecular
weight enough such that the resulting polymer was soluble in THF (Mn )
4160, PDI ) 1.60). Integrating the crotonate methyl groups in the 1H NMR
spectrum versus the PHB methyl gives Mn ) 4300.
(47) Dubois, P.; Jacobs, C.; Je´roˆme, R.; Teyssie´, P. Macromolecules 1991, 24,
2266-2270.
(48) Kricheldorf, H. R.; Scharnagl, N.; Jedlinski, Z. Polymer 1996, 37, 1405-
1411.
(49) Kurcok, P.; Dubois, P.; Je´roˆme, R. Polym. Int. 1996, 41, 479-485.
(50) Kricheldorf, H. R.; Berl, M.; Scharnagl, N. Macromolecules 1988, 21, 286-
293.
9
15246 J. AM. CHEM. SOC. VOL. 124, NO. 51, 2002

Ring-Opening Polymerization of â-Butyrolactone
A R T I C L E S
Scheme 5. Reaction of 1a with 4-Acetoxy-2-butanone
The results presented in this paper bring us a step closer to
an alternative route for making poly(3-hydroxyalkanoates).
Despite a host of potential applications, PHAs produced by
bacterial fermentation and other natural processes remain too
expensive for widespread use. The carbonylation of epoxides
to make optically pure â-substituted â-propiolactones38 followed
by the polymerization of the â-substituted â-propiolactones
using (BDI)Zn alkoxides represents an efficient route to PHAs
from abundant feedstocks.
Experimental Section
General Considerations. All reactions with air- and/or water-
sensitive compounds were carried out under dry nitrogen using a Braun
Labmaster drybox or standard Schlenk line techniques. NMR spectra
were recorded on a Bruker AF300 (1H, 300 MHz; 13C, 75 MHz), Varian
VXR-400 (1H, 400 MHz; 13C, 100 MHz), or Varian Unity 500 (1H,
500 MHz; 13C, 125 MHz) spectrometer and referenced versus residual
solvent shifts. Gel permeation chromatography (GPC) analyses were
carried out using a Waters instrument (M510 pump, U6K injector)
equipped with Waters UV486 and Waters 2410 differential refractive
index detectors and three 5-mm PL gel columns (Polymer Laboratories;
100, 500, and 1000 Å) in series. The GPC columns were eluted with
tetrahydrofuran at 40 °C at 1 mL/min and were calibrated using a
polynomial fit to 10 monodisperse polystyrene standards. Optical
rotation experiments were conducted in chloroform at 25 °C using a
Perkin-Elmer 241 polarimeter. DSC analyses were conducted on a Seiko
DSC 220C instrument using EXSTAR 6000 processing software. The
measurements were made in aluminum crimped pans under nitrogen
with a heating rate of 10 °C/min.
Materials. Benzene-d6, toluene-d8, and tetrahydrofuran were distilled
from sodium benzophenone ketyl under nitrogen and degassed by three
freeze-pump-thaw cycles. Benzene-d6 and toluene-d8 were stored over
activated 4 Å molecular sieves. Methylene chloride was collected from
activated alumina-packed solvent columns and degassed by three freeze-
pump-thaw cycles. Chloroform was distilled from phosphorus pentoxide
under nitrogen and degassed by three freeze-pump-thaw cycles. Racemic
â-butyrolactone was purchased from Aldrich, while (R)-â-butyrolactone
and â-valerolactone were prepared according to published procedures.38
â-Butyrolactone was dried with CaH2 for 3 days, vacuum transferred
onto activated 4 Å molecular sieves, and then fractionally distilled under
reduced pressure; the entire process was then repeated. Note: â-bu-
tyrolactone is a suspected carcinogen; thus, it should be handled with
appropriate safeguards. â-Valerolactone was dried with CaH2 for 3 days,
vacuum transferred onto activated 4 Å molecular sieves, and then
fractionally distilled under reduced pressure. Complexes 1a, 2a, 2c,
and (BDI-2)ZnN(SiMe3) were prepared according to published proce-
dures.29,30,32 Poly-(R)-(3-hydroxybutyrate) was purchased from Fluka
and dried under vacuum for 1 h prior to use.
4-acetoxy-2-butanone in benzene-d6. After 30 min, the reaction
solution was split into volatile and nonvolatile fractions by
1
vacuum transfer. H NMR analysis of the nonvolatile fraction
revealed complete transformation of the zinc species 1a to (BDI-
1)ZnOAc;33 a small amount of protonated ligand was also
observed. In the volatile fraction, the 4-acetoxy-2-butanone was
largely consumed, yielding the expected elimination products
2-propanol and methyl vinyl ketone, plus acetic acid and a small
amount of Michael addition product 4-isopropoxy-2-butanone
(Scheme 5). This experiment provides indirect evidence that
1a induces the elimination reaction on the PHB backbone.
The role of transesterification in polymerizations of BBL with
1a is more difficult to assess. Interestingly, when commercially
prepared poly-(R)-(3-hydroxybutyrate) ((R)-PHB) was stirred
for 2 h with catalyst 1a in CHCl3 at 50 °C (see above), GPC
peaks consistent with low molecular weight PHB cyclics were
not observed in the decomposition product, suggesting that
intramolecular transesterification (backbiting) is negligible
compared with elimination under these conditions. In fact, we
have found little evidence that transesterification is prominent
in this system, on the basis of the narrow PDIs and lack of low
MW cyclics in reaction products, even at 75 °C. Furthermore,
despite the evidence that elimination is facilitated by 1a, we
cannot rule out the possibility that the nonliving behavior we
observed is due to the presence of impurities in the monomer,
since the nonliving behavior is more pronounced at high [BBL]/
[Zn] ratios and BBL is notoriously difficult to prepare in
>99.9% purity.
Conclusions
Zinc alkoxide complex 1a is an easily synthesized and
extremely efficient catalyst for the ring-opening polymerization
of strained lactones. We had previously reported that â-diiminate
zinc alkoxides are extremely active catalysts for the living,
stereoselective ROP of lactide.29,30 In this paper, we demonstrate
that 1a polymerizes BBL and BVL with unprecedented rates
under mild conditions to make PHAs in a controlled manner.
Prior to this report, the active form of the â-diiminate zinc
species for ROP of lactones was largely unresolved.29 The
experimental results described in this paper strongly suggest
that a monometallic zinc complex mediates the insertion of the
â-butyrolactone monomer at the alkoxide chain end. Ring
opening occurs with acyl-oxygen bond cleavage and, therefore,
retention of configuration at the chiral methine carbon. We
believe this model can be extended to describe the behavior of
1a in other lactone polymerizations; we are currently investigat-
ing this hypothesis.
rac-(BDI-2)ZnOCH(Me)CO2Me (2b). To a solution of (BDI-2)-
29
ZnN(SiMe3)2 (1.45 g, 2.47 mmol) in toluene (5 mL) was added
racemic methyl lactate (0.24 mL, 2.5 mmol). After the clear yellow
solution was stirred at room temperature for 6 h, it was dried in vacuo.
The resulting white powder was dissolved in hexanes (20 mL) and
filtered through a pad of Celite. The filtrate was concentrated and cooled
to -30 °C; whereupon, rac-(BDI-2)ZnOCH(CH3)CO2Me (2b) crystal-
1
lized as colorless blocks (0.636 g, 48% yield). H NMR (C6D6, 500
MHz): δ 7.07 (6H, m, ArH), 4.88 (1H, s, â-CH), 4.24 (1H, quartet, J
) 7 Hz, OCH(Me)CO2CHMe2), 3.00 (3H, s, CO2Me), 2.90 (2H, m, J
) 7 Hz, ArCH2Me), 2.85 (2H, m, J ) 7 Hz, ArCH2Me), 2.64 (4H, m,
J ) 7 Hz, ArCH2Me), 1.64 (6H, s, R-Me), 1.25 (12H, m, ArCH2Me),
0.73 (3H, d, J ) 7 Hz, OCH(Me)CO2CHMe2) ppm. 13C{1H} NMR
(C6D6, 100 MHz): δ 192.56, 168.75, 146.72, 138.02, 137.92, 126.31,
125.58, 94.27, 70.91, 53.14, 24.73, 24.00, 23.65, 14.70 ppm. X-ray
analysis of the crystals revealed that the complex exists as a µ-alkoxide-
bridged dimer in the solid state (see Supporting Information).
9
J. AM. CHEM. SOC. VOL. 124, NO. 51, 2002 15247

A R T I C L E S
Rieth et al.
i
(R)-(BDI-1)ZnOCH(CH3)CH2CO2 Pr (1b). To a solution of (BDI-
1)ZnOiPr (1a) (0.124 g, 0.229 mmol) in toluene (3 mL) was added a
1.227 M solution of (R)-â-butyrolactone in C6D6 (0.19 mL, 0.23 mmol).
After the clear colorless solution was stirred at room temperature for
15 min, it was dried in vacuo to yield a white powder. Despite repeated
attempts, efforts to grow X-ray-quality crystals of the complex were
unsuccessful. 1H NMR (C6D6, 500 MHz): δ 7.14 (6H, m, ArH), 4.92
(1H, s, â-CH), 4.82 (1H, m, OCH(Me)CH2CO2CHMe2), 4.14 (1H, m,
OCH(Me)CH2CO2CHMe2), 3.55 (4H, m, ArCHMe2), 1.80 (1H, dd, J
) 14 Hz and J ) 3 Hz, OCH(Me)CH2CO2CHMe2), 1.77 (6H, s, R-Me),
1.62 (1H, dd, J ) 14 Hz and J ) 7 Hz, OCH(Me)CH2CO2CHMe2),
1.42 (12H, d, J ) 7 Hz, ArCHMe2), 1.22 (12H, d, J ) 7 Hz, ArCHMe2),
1.00 (3H, d, J ) 6 Hz, OCH(Me)CH2CO2CHMe2), 0.81 (3H, d, J ) 6
Hz, OCH(Me)CH2CO2CHMe2), 0.80 (3H, d, J ) 6 Hz, OCH(Me)CH2-
CO2CHMe2) ppm. 13C{1H} NMR (C6D6, 100 MHz): δ 169.49, 145.05,
142.93, 125.61, 123.84, 123.81, 93.93, 69.45, 67.48, 47.09, 28.26, 28.16,
24.78, 24.72, 24.67, 24.62, 23.84, 21.64, 21.55 ppm (one resonance
missing, possibly because of resolution or chemical shift equivalence).
General Polymerization Procedure. In the glovebox, the catalyst
was dissolved in 1.0 mL of solvent in a small vial. â-Butyrolactone
(0.25 mL) was added to the vial, and the reaction solution was
transferred to a Teflon-sealed NMR tube. The reaction solution was
quickly removed from the glovebox and inserted into a temperature-
controlled NMR. Conversion was monitored by comparing the relative
magnitude of peaks corresponding to the methine hydrogen for BBL
and PHB. Polymerizations were quenched by adding five drops of acetic
acid-d4. The resulting solution was filtered over a Celite/glass wool
plug 1 h after quenching. The polymer was isolated by removing solvent
in vacuo.
polymerizations conducted at 50 °C or 75 °C corresponds to the time
the reaction solution is introduced to the heat source (i.e., NMR or oil
bath), which is generally 2-3 min after the addition of BBL to the
catalyst solution.
To further purify the polymer for endgroup analysis, the polymer
was redissolved in chloroform; several drops of 1% HCl (in MeOH)
solution were added, and the solution was again filtered over Celite.
The polymer was precipitated by the addition of hexanes; the mother
liquor was removed by decanting, and the polymer was dried in vacuo.
Polymer samples were precipitated twice from chloroform with hexanes
(to remove residual ligand) prior to polarimetry experiments.
Acknowledgment. We thank Yutan Getzler and Dr. Viswanath
Mahadevan for preparation of (R)-â-butyrolactone and â-vale-
rolactone, respectively, and Dr. Scott Hanton and Dr. James
Willis for helpful discussions. This work was generously
supported by the Cornell Nanobiotechnology Center (NBTC),
an STC Program of the National Science Foundation under
Agreement No. ESC-9876771. This work made use of the
Cornell Center for Materials Research Shared Experimental
Facilities, supported through the National Science Foundation
Materials Research Science and Engineering Centers program
(DMR-0079992). G.W.C. gratefully acknowledges an NSF
CAREER Award (CHE-9875261), an Alfred P. Sloan Research
Fellowship, an Arnold and Mabel Beckman Foundation Young
Investigator Award, a Camille Dreyfus Teacher-Scholar Award,
and a David and Lucile Packard Foundation Fellowship in
Science and Engineering.
When nondeuterated solvents were used, the polymerization solution
was sealed in a small vial with a Teflon cap. Temperature was controlled
using an oil bath (except at room temperature). Conversion was
measured after the polymerization was quenched with five drops of
acetic acid-d4 by 1H NMR of an aliquot of the quenched reaction
solution in benzene-d6.
Reaction times for polymerizations conducted at room temperature
reflect the time between the addition of BBL to the catalyst solution
and the quenching of the reaction; in contrast, the start time for
Supporting Information Available: X-ray crystal structure
data for 2b, H NMR spectra for the products of the reaction
between 1a and 4-acetoxy-2-butanone, and a mechanistic
scheme for transesterification and elimination reactions mediated
by metal alkoxides. This material is available free of charge
via the Internet at http://pubs.acs.org.
1
JA020978R
9
15248 J. AM. CHEM. SOC. VOL. 124, NO. 51, 2002