A study of the ring-opening of lactides and related cyclic esters by Ph2SnX2 and Ph3SnX compounds (X = NMe 2, OR)

Article abstract of DOI:10.1039/b300101f

Ph₃SnX reacts with L-lactide in the order X = NMe ₂～OMe ? OPrⁱ > OBu^t in the initial ring-opening event. The rate of ring-opening of methyl substituted 1,4-dioxane-2,5-diones decreases with methyl substitution and ring-opening of 3,3,6,6-tetramethyl-1,4-dioxane-2,5-dione is not observed. From studies of the reaction between Ph₃SnOPrⁱ and L-lactide the activation parameters ΔH^≠ = 13(1) kcal mol^-1 and ΔS^≠ = -37(3) eu have been determined. Ph ₃SnNMe₂ reacts with cyclic esters and propylene carbonate at low temperatures to give isolable ring-opened products. The compound Ph ₃Sn[OCHMeC(O)OCHMeC(O)X] (where X = NMe₂, OMe) are labile in solution at room temperature, yielding Ph₃Sn[OCHMeC(O)X] and Ph₃Sn[OCHMeC(O)]_nX, where n ≥ 3, by transesterification, along with other minor products due to phenyl/OR group transfer (i.e., Ph₄Sn). Ph₂Sn(NMe₂) ₂ and L-lactide react to give Ph₂Sn[OCHMeC(O)NMe ₂]₂ by ring-opening of L-lactide followed by rapid amidation. A related compound, Ph₂Sn[OCHMeC(O)OPrⁱ] ₂, is also formed in the reaction between Ph₂Sn(OPr ⁱ)₂ and L-lactide but is more labile toward further ring-opening of L-lactide.

Full text of DOI:10.1039/b300101f

View Article Online / Journal Homepage / Table of Contents for this issue
A study of the ring-opening of lactides and related cyclic esters by
Ph2SnX2 and Ph3SnX compounds (X = NMe2 , OR)y
Malcolm H. Chisholm* and Ewan E. Delbridge
Department of Chemistry, The Ohio State University, 100 W 18th Avenue, Columbus, OH,
43210-1185, USA. E-mail: chisholm@chemistry.ohio-state.edu; Fax: +1 614 292 0368;
Tel: +1 614 292 7216
Received (in New Haven, CT, USA) 6th January 2003, Accepted 5th March 2003
First published as an Advance Article on the web 7th July 2003
Ph3SnX reacts with L-lactide in the order X ¼ NMe2 ꢀ OMe ꢁ OPri > OBut in the initial ring-opening
event. The rate of ring-opening of methyl substituted 1,4-dioxane-2,5-diones decreases with methyl substitution
and ring-opening of 3,3,6,6-tetramethyl-1,4-dioxane-2,5-dione is not observed. From studies of the reaction
¼
¼
between Ph3SnOPri and L-lactide the activation parameters DH ¼ 13(1) kcal molꢂ1 and DS ¼ ꢂ37(3)
eu have been determined. Ph3SnNMe2 reacts with cyclic esters and propylene carbonate at low temperatures to
give isolable ring-opened products. The compound Ph3Sn[OCHMeC(O)OCHMeC(O)X] (where
X ¼ NMe2 , OMe) are labile in solution at room temperature, yielding Ph3Sn[OCHMeC(O)X] and
Ph3Sn[OCHMeC(O)]nX, where n ꢃ 3, by transesterification, along with other minor products due to phenyl/
OR group transfer (i.e., Ph4Sn). Ph2Sn(NMe2)2 and L-lactide react to give Ph2Sn[OCHMeC(O)NMe2]2 by
ring-opening of L-lactide followed by rapid amidation. A related compound, Ph2Sn[OCHMeC(O)OPri]2 , is
also formed in the reaction between Ph2Sn(OPri)2 and L-lactide but is more labile toward further ring-opening
of L-lactide.
We reasoned that much might be learned from a study of the
Introduction
reactivity of Ar3SnX and Ph2SnX2 precursors in their reactions
with lactides and related cyclic esters since, unlike Tp or beta-
diiminate (BDI) ligands, reversible metal-ligand dissociation in
solvents such as benzene and methylene chloride could safely
be ruled out. To date, there have been a number of interesting
papers that have dealt with the polymerization of cyclic esters
such as LA by tin(II) catalyst precursors8 but there are
few examples utilizing tin(IV) catalysts.9 These Ar3SnX and
Ph2SnX2 precursors are kinetically slow relative to tin(II) and
other zinc(II) and magnesium(II) catalysts. This allows for the
possibility to study in isolation the three components of
metal-mediated anionic lactide polymerization, namely the
initial ring-opening event, the propagating reaction (also
known as ring-opening polymerization, ROP) and the side
Interest in polylactides, PLAs, arises from their numerous
potential applications, which range from environmentally
friendly packaging materials1 to drug-delivery agents,2 biode-
gradable sutures2,3 and polymer scaffolds for tissue engineer-
ing.4 Ring-opening polymerization (ROP) of lactides (LA; L-,
D-, rac- and meso- forms) to form polylactides by single-site
metal alkoxide precursors has attracted considerable recent
attention5 since the properties of a given PLA are determined
by its molecular weight, molecular weight distribution and
molecular microstructure. Recent successes in the development
of single-site catalysts include the preparation of syndiotactic5a
and heterotactic5c polymers along with control of polydisper-
sity. An excellent review documents recent endeavors in this
field.6 Much attention has been directed to understanding the
stereosequences (the arrangement of the R and S chiral centers)
of PLA with recent work by Hillmyer and coworkers addres-
sing this issue by use of high resolution NMR spectroscopy.7
It is, however, fair to state that little is really understood in
terms of what controls the intimate mechanistic steps of poly-
merization, how stereoselectivity is achieved and what factors
influence the relative rates of ROP versus the often deleterious
side reactions involving transesterification and chain transfer.
For example, it is not clear why sterically demanding and
chiral Tp*M(OR) compounds,5e A, of magnesium or zinc
should yield only poor stereoselectivity when less sterically
demanding achiral diiminato compounds B achieve significant
stereoselectivity ( > 90%) under similar reaction conditions
(Fig. 1).5d,5g
y Electronic supplementary information (ESI) available: preparation
and characterization data for Ar3SnX, Ph2SnNMe2 and Ph2Sn(OPri)2
compounds [X ¼ NMe2 , OMe, OPri, OBut, OPh, OCH(CF3)2 , PPh2
and Ar ¼ Ph, o-MeC6H4]; NOE data and NMR spectra available.
See http://www.rsc.org/suppdata/nj/b3/b300101f/
Fig. 1 Zinc and magnesium catalyst precursors for the polymeriza-
tion of lactide.
DOI: 10.1039/b300101f
New J. Chem., 2003, 27, 1167–1176
1167
This journal is # The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2003

View Article Online
Scheme 1 The three components of lactide polymerization: ring-opening, propagation and side reactions (transesterification and chain transfer).
reactions such as chain transfer and transesterification
(Scheme 1).
If the relative rates of initial ring-opening, kro , and sub-
Characterization data for all compounds reported here are
given in the Electronic supplementary information (ESI). The
Ar3SnOR compounds (R ¼ But and Pri) were stable at 100 ꢄC
1
in anhydrous toluene for several days as evidenced by H and
sequent propagation, kprop , markedly differ such that kro
>
kprop it should be possible to isolate products where just one
LA monomer has been ring-opened. We first observed this
for Ph3SnNMe2 and L-lactide10 and have expanded our find-
ings here. Specifically we have systematically investigated the
effect of three parameters on the ring-opening event. (1) The
effect of X, the initiating group, in Ph3SnX on the ring-
opening of LA. (2) The effect of methyl substitution on the
cyclic monomer as in LA vs. GL (where GL ¼ glycolide).
(3) The effect of altering the Ar groups (the spectator ligands)
on the Sn(IV) initiator. The results from this study are dis-
cussed below whilst a detailed description of the propagation
kinetics11 and side reactions of transesterification and chain
transfer utilizing R2SnX2 and R3SnX catalyst precursors will
be discussed in subsequent publications.
119Sn NMR spectroscopy.
Discrete ring-opened lactide and related cyclic ester derivatives
of Sn(IV)
These new compounds, described below, have been character-
ized by NMR [1H, 13C{1H}, 119Sn] and IR spectroscopy, mass
spectrometry and, in many cases, by elemental analyses. Data
are reported in the Experimental in addition to our previous
preliminary report on this chemistry.10
(a) Ring-opening of L-lactide by Ph3SnX where X ¼ NMe2 ,
OMe, OPri, OCH(CF3)3 , OBut, OPh and PPh2. In order to
investigate the effect of the initiating group on the ring-opening
of L-lactide a variety of Ph3SnX compounds [X ¼ NMe2 ,
OMe, OPri, OCH(CF3)2 , OBut, OPh and PPh2] were synthe-
sized and treated with equiv. of L-lactide (Scheme 2). It is
apparent from Scheme 2 that electronic and steric properties
of the initiating group influence not only the ring-opening of
L-lactide but also the stability of the ring-opened product.
Scheme 2 reveals that Ph3SnNMe2 and Ph3SnOMe rapidly
ring-open L-lactide at room temperature in benzene to form
the discrete and isolable products Ph3Sn[OCHMeC(O)OCH-
MeC(O)X] where X ¼ NMe2 or OMe. The 1H NMR spectrum
of Ph3Sn[OCHMeC(O)OCHMeC(O]NMe2) is shown in Fig. 2,
along with an assignment of the ligand resonances. The latter
is possible from a COSY spectrum.
Results and discussion
Ar3SnX and Ph2SnX2 compounds
The dimethylamido compounds were all synthesized in a simi-
lar manner from the metathetic reaction between Ar3SnCl and
LiNMe2 in benzene or THF to yield the corresponding
Ar3SnNMe2 compounds, where Ar ¼ Ph and o-MeC6H4 ,
which were air-sensitive, white or colorless oils or waxy solids.
Ph2Sn(NMe2)2 was prepared in an analogous fashion from
Ph2SnCl2 and 2 equiv. of LiNMe2 .
The corresponding alkoxides Ar3SnOR and Ph2Sn(OR)2
were prepared by alcoholysis of the appropriate dimethyl-
amides (R ¼ Me, Pri, CH(CF3)2 , But and Ph). Ph3SnPPh2
was likewise synthesized by treatment of Ph3SnNMe2 with
excess HPPh2 . These compounds were similarly air-sensitive,
white or colorless waxy solids or oils and were purified by
vacuum distillation or crystallization.
Ph3Sn[OCHMeC(O)OCHMeC(O)OMe] is less stable than
the analogous NMe2 compound and is more susceptible to
intermolecular transesterification leading to Ph3Sn[OCHMe-
C(O)OMe] and Ph3Sn[OCHMeC(O)]nOMe, where n > 3 (see
below for a discussion of stabilities and side reactions of these
complexes).
1168
New J. Chem., 2003, 27, 1167–1176

View Article Online
Scheme 2 Reactions between Ph3SnX compounds and L-lactide indicating the effect of initiating group. Compounds in boxes are isolable species.
The steric effect of X on ring-opening of L-lactide is best illu-
strated in the comparison between Ph3SnOMe, Ph3SnOPri and
Ph3SnOBut, where the order of reactivity decreases with
increasing steric bulk of X. This is the inverse of the basicity
of the alkoxide anions: ButO > PriO > MeO. Unlike
Ph3SnOMe, both Ph3SnOPri and Ph3SnOBut require heating
in benzene (45–50 ꢄC) to accomplish appreciable ring-opening.
This decrease in reactivity for the more bulky X groups may
reflect the increased steric crowding about the tin(IV) center,
thereby suppressing the ability of L-lactide to coordinate to
the tin(IV) metal center prior to ring-opening (Scheme 1). In
addition, the X group may be too bulky to attack the ester car-
bonyl carbon atom. Furthermore, it appears that Ph3Sn[OCH-
MeC(O)OCHMeC(O)X] with X ¼ OPri is more stable (Fig. 3)
(forming a stable, isolable product) than for X ¼ OBut, which
probably is a result of two effects: (1) the relative rates of ring-
opening, kro , and propagation or ROP,11 kprop , are such that
kro > kprop for X ¼ OPri but kprop > kro for X ¼ OBut, and
(2) the poorer ability of the –C(O)OBut group to chelate either
inter- or intramolecularly to the tin(IV) center (see below for
a discussion of the stability of these compounds).
A more pronounced effect of kprop > kro was seen in the
reactions involving Ph3SnX and L-lactide (1:1 equiv.) where
X ¼ OCH(CF3)2 , OPh and PPh2 . Here heating in benzene-
d6 was required (73 and 94 ꢄC, Scheme 2) before any reaction
was observed and then only a small amount of Ph3SnX
was consumed and no discrete compounds of formula
Ph3Sn[OCHMeC(O)OCHMeC(O)X)] were detected by 1H
NMR spectroscopy. We can conclude that the observed rate
order for the initial rate of ring-opening of L-lactide, which is
i
X ¼ NMe2ꢀ OMe > OPr > OBut > OPh > OCH(CF3)2 > PPh2,
Fig. 2 1H NMR spectrum in benzene-d6 of Ph3Sn[OCHMe-
C(O)OCHMeC(O)ONMe2]. Notice the inequivalent NMe2 methyl
signals.
Fig. 3 1H NMR spectrum in benzene-d6 of Ph3Sn[OCHMe-
C(O)OCHMeC(O)OPri]. Notice the inequivalent OPri methyl signals.
New J. Chem., 2003, 27, 1167–1176
1169

View Article Online
represents a combination of electronic and steric factors. The
rate is enhanced by the nucleophilicity/basicity of the anion
Xꢂ and is suppressed by steric factors. For X ¼ NMe2 ,
OMe and OPri, kro > kprop but for X ¼ OCH(CF3)2 , OPh
and PPh2 , kprop > kro . The influence of steric factors is
further underscored by the observation that (o-MeC6H4)3-
SnNMe2 and L-lactide do not react!.
(b) Kinetics of the ring-opening event. A reaction scheme
for the initial ring-opening event can be represented by
eqn. (1),
k1
*
ðaÞ A þ B)
C
Keq
k
ꢂ1
k2
ðbÞ
C
D
ð1Þ
ƒƒƒ!
where A represents the Ph3SnX compound, B the L-lactide, C a
kinetically labile adduct or reactive intermediate and D, the
product of the ring-opening event. In the reaction scheme
the rate of formation of the ring-opened product, d[D]/dt, is
given by eqn. (2), according to the steady-state approximation.
d½Dꢅ k2:k1½Aꢅ½Bꢅ
Fig. 4 Plot of ln(kobs/T ) vs. 1/T for the reaction between Ph3SnOPri
and L-lactide (1:1).
¼
ð2Þ
dt
k
ꢂ1 þ k2
There are two limiting situations: (i) for
k
ꢂ1 ꢁ k2 ,
kobs ¼ k2ꢆk1/kꢂ1 ꢀ k2ꢆKeq and (ii) for k2 ꢁ k when kobs
¼
ꢂ1
expected for attack on a carbonyl group coordinated to the
metal by an adjacent nucleophilic alkoxide ligand. The nega-
tive and modest magnitude of the entropy of activation is
typical of many bimolecular processes and does not imply a
uniquely high order in the transition state.
k1 . In other cases the reaction scheme does not simplify and
kobs ¼ k2ꢆk1/(kꢂ1 + k2).
If the ring-opened product D closely resembles A then
further reaction with substrate B can be expected and we will
arrive at a kinetic situation where kprop ꢀ kro . However, in
order to study the initial ring-opening event and have this
make a meaningful or at least a chemically interesting compari-
son with the ring-opening polymerization (propagation), kro
must be greater but not very much greater than kprop . In this
study this situation is obtained for Ph3SnOPri and we have
examined the kinetics of the ring-opening event as a function
of temperature (29 to 81 ꢄC) in benzene-d6 . The reactions were
Since 4-coordinate tin(IV) compounds are well known to be
able to increase their coordination numbers in associative reac-
tions we examined the kinetics of the ring-opening of L-lactide
by Ph3SnOPri in the presence of added donors, specifically
propylene carbonate (PC) and pyridine (py) (Table 1). Given
that neither is susceptible to ring-opening but that both mole-
cules could compete in a reversible association to tin(IV) with L-
lactide, according to the scheme outlined in eqn. (1), the rate of
ring-opening, kro , could be supressed. The added ligand, py or
PC, was in five-fold excess of the lactide concentration. At
1
followed by H NMR spectroscopy by simultaneously follow-
ing the decrease in L-lactide, the loss of SnOPri signals and the
growth of the product Ph3Sn[OCHMeC(O)OCHMeC(O)OPri]
(see Experimental). The reaction proceeded as expected based
on the reaction scheme shown in eqn. (1). Most significantly
there was no evidence by NMR spectroscopy for the formation
of a 1:1 adduct, supporting the notion that C is a reactive inter-
mediate, with Keq being small, not detected by NMR spectro-
scopy. This is further corroborated by the fact that 119Sn
NMR spectra of Ph3SnOPri in toluene-d8 at low temperatures
are insensitive to the presence of excess lactide. Kinetic data
are summarized in Table 1.
29 ꢄC we observed no supression in the rate constant, kobs
,
and for py a slight increase in kobs was observed, which may
be due to subtle changes in solvent polarity that, in turn,
enhance the rate of ring-opening (Table 1).
(c) Ring-opening of related cyclic esters by Ph3SnX where
X ¼ NMe2 and OPri. It has been shown that the nature of
the initiating group X and the ancillary ligands Ar can control
the ring-opening event of L-lactide. However, it is not clear
whether related cyclic esters have the same propensity to
ring-open as L-lactide and, if so, whether there will be any
regiospecificity in ring-opening of an unsymmetrical cyclic
ester. To address these issues we have studied the reactions
between Ph3SnX compounds, where X ¼ NMe2 or OPri, and
a variety of cyclic esters that are closely related to L-lactide
(Scheme 3).
From a plot of ln(k/T ) vs. 1/T (Fig. 4) we can obtain an
¼
estimate of the activation parameters: DH ¼ 13(1) kcal molꢂ1
¼
and DS ¼ ꢂ37(3) eu. These are entirely reasonable values for
the metal-mediated anionic ring-opening reaction proposed in
Scheme 1. The enthalpy of activation is quite modest as
Table 1 Kinetics data for the 1:1 reaction between Ph3SnOPri and L-
lactide leading to formation of Ph3Sn[OCHMeC(O)OCHMeC(O)OPri]
at various temperatures and in the presence of donors
Ph3SnNMe2 was shown to react regioselectively with the
asymmetrically substituted cyclic ester 3,3,6-trimethyl-1,4-diox-
ane-2,5-dione to give Ph3Sn[OCMe2C(O)OCHMeC(O)NMe2]
whilst Ph3SnOPri reacts with 3-methyl-1,4-dioxane-2,5-dione
(methylglycolide) to give Ph3Sn[OCHMeC(O)OCH2C(O)OPri]
(95%). In both cases nucleophilic attack occurs at the less-
hindered (and more electrophilic) carbonyl moiety (Scheme
3). The H NMR spectrum of the compound Ph3Sn[OCMe2-
C(O)OCHMeC(O)NMe2] can be found in the ESI.
As is apparent, there is only one regioisomer for
Ph3Sn[OCMe2C(O)OCHMeC(O)NMe2] and the specific iso-
mer is established with confidence based on the presence
or absence of coupling to 119Sn/117Sn in the 1H and
Entry
kobs/molꢂ1 dm3 sꢂ1
)
T/ꢄC
1
2.0(2) ꢇ 10ꢂ5
4.1(2) ꢇ 10ꢂ5
1.3(2) ꢇ 10ꢂ4
6.1(2) ꢇ 10ꢂ4
1.6(2) ꢇ 10ꢂ5
3.0(2) ꢇ 10ꢂ5
29
49
61
81
29
29
2
1
3
4
5a
6b
a
With 5 equiv. of propylene carbonate. b With 5 equiv. of pyridine.
1170
New J. Chem., 2003, 27, 1167–1176

View Article Online
Scheme 3 Reactions between Ph3SnX (X ¼ NMe2 , OPri) and methyl substituted lactide derivatives.
13C{1H} spectra. This compound is notably less kinetically
labile than its L-lactide counterpart and does not participate
in self transesterification as observed for Ph3Sn[OCHMe-
C(O)OCHMeC(O)NMe2]. (See below for a discussion of
the stabilities of these compounds.)
Ring-opening is not observed when Ph3SnNMe2 is treated
with the tetramethylated cyclic ester derivative, 3,3,6,6-tetra-
methyl-1,4-dioxane-2,5-dione. This result complements the
regiospecific ring-opening observed for the 3,3,6-trimethyl-
1,4-dioxane-2,5-dione cyclic ester by Ph3SnNMe2 , for which
no activation was observed for the more hindered carbonyl
ester.
Hence the degree of methyl substitution on the cyclic ester
has a considerable effect on the relative rates and regiospecifi-
city of ring-opening: the less hindered cyclic esters more
rapidly ring-open in a less regiospecific fashion. In addition,
the reactivity of the 1:1 ring-opened products is also related
to the degree of methyl substitution in that the most kinetically
persistent compounds are those with a high degree of methyl
substitution of the cyclic ester. This ‘‘stability’’ may arise from
the higher degree of steric crowding about the tin(IV) metal
center in the more heavily substituted ring-opened systems,
which supresses further bimolecular reactions.
Treatment of Ph3SnOPri with 1,4-dioxane-2,5-dione (glyco-
lide) (1:1) does not lead to the isolable compound
Ph3Sn[OCH2C(O)OCH2C(O)OPri] but rather to polyglycolide
(Scheme 3), uggesting that the ratesof ring-opening and chain
propagation are similar such that subsequent glycolide mole-
cules are able to insert readily into the already formed
ring-opened product. The relative stability of the Ph3Sn[CMe2-
C(O)OCHMeC(O)NMe2] and Ph3Sn[CHMeC(O)OCHMe-
C(O)X] (X ¼ NMe2 , OPri) complexes (see above) is presum-
ably related to the poorer substrate activation of the cyclic
ester monomers by the ring-opened product relative to initial
activation by Ph3SnOPri or Ph3SnNMe2 . This leads to the
initial rate of ring-opening being greater than the subseq-
uent rate of ROP (kro > kprop). In contrast, the less
hindered glycolide (and to a lesser extent methylglycolide)
molecules can more readily coordinate to the tin center than
Ph3SnNMe2 has also been shown to ring-open the cyclic
carbonate, propylene carbonate, to give a mixture of the two
regioisomers Ph3Sn[OCH2CHMeOC(O)NMe2] and Ph3Sn-
[OCHMeCH2OC(O)NMe2]. The relative ratio of these regio-
isomers was little changed in reactions carried out at ꢂ78 ꢄC.
L-lactide or 3,3,6-trimethyl-1,4-dioxane-2,5-dione, resulting in
kprop ꢃ kro
12
.
New J. Chem., 2003, 27, 1167–1176
1171

View Article Online
Scheme 4 Reaction between Ph2SnX2 compounds (X ¼ NMe2 , OPri) and L-lactide, indicating rapid intramolecular inter-chain esterification/
amidation.
(d) Reaction between Ph2SnX2 and L-lactide where
X ¼ NMe2 or OPri. Ph2Sn(NMe2)2 and L-lactide (1:1) react in
benzene to give Ph2Sn[OCHMeC(O)NMe2]2 by an initial
ring-opening event, following by a rapid intramolecular ami-
dation (Scheme 4). This reaction is conceptually similar to that
seen between Ph3SnNMe2 (2 equiv.) and L-lactide (1 equiv.),
where the initially formed Ph3Sn[OCHMeC(O)OCHMe-
C(O)NMe2] compound is also seen to react quickly with
Ph3SnNMe2 to give Ph3Sn[OCHMeC(O)NMe2] by an inter-
molecular amidation (see below).
The compound Ph2Sn[OCHMeC(O)NMe2]2 has two IR
bands, at 1596 and 1609 cmꢂ1, assignable to the organic amide
moiety. These values are considerably lower than those found
for Ph3Sn[OCRMeC(O)OCHMeC(O)NMe2] (R ¼ H or Me)
compounds whose infrared spectra of the C(O)NMe2 moieties
show characteristic bands at ca. 1600–1650 cmꢂ1, notably
lower than the ester carbonyl groups, which are at ca. 1700–
1750 cmꢂ1. Furthermore, the 119Sn NMR spectrum of
Ph2Sn[OCHMeC(O)NMe2]2 is also significantly different from
that of the other compounds and shows complex variable tem-
perature behavior similar to that recently observed for the ana-
logous Me2Sn(OAr)2 compound.13 It is believed that this,
together with the lower values of n(CO), is evidence for intra-
or intermolecular amide group coordination to Sn (see below).
Regrettably, crystals suitable for a single crystal X-ray deter-
mination were not obtained.
None of these compounds are, however, indefinitely persistent
in solution or at elevated temperatures {with the exception of
Ph2Sn[OCHMeC(O)NMe2]2}. It is clear from the above dis-
cussions that some of these compounds exhibit greater stability
than others. It is believed that this stability arises from the
formation of chelating C(O)X end groups to tin(IV) (Fig. 5).
NOE difference spectra were obtained for Ph3Sn[OCHMe-
C(O)OCHMeC(O)NMe2], Ph3Sn[OCMe2C(O)OCHMeC(O)N-
Me2] and Ph2Sn[OCHMeC(O)NMe2]2 compounds (see ESI).
These indicate a significant Hꢆ ꢆ ꢆ H interaction involving the
ortho-phenyl protons and the methine protons of the ligand.
We propose that this arises from reversible chelation of the
ligand through 5- and 8-membered rings of the type shown
in C, D and E (Fig. 5). Although 5-membered rings are pre-
ferred over 8-membered rings, the latter may be favored by the
terminal amide groups in the Sn[OCHMeC(O)NMe2]2 moiety.
The compounds Ph3Sn[OCHMeC(O)OCHMeC(O)X] (where
X ¼ NMe2 , OPri, OMe) display kinetic lability toward a
ligand redistribution reaction that arises from intermolecular
transesterification (Scheme 5, reaction i). This reaction pro-
ceeds over several hours at room temperature and is acceler-
ated by heat in benzene-d6 or toluene-d8 solution and
can readily be followed by 1H NMR spectroscopy (see ESI
for X ¼ NMe2). From these studies the stability of
1
1
In an analogous fashion Ph2Sn(OPri)2 reacts with L-lactide
(1:1) to form Ph2Sn[OCHMeC(O)OPri]2 (Scheme 4). This com-
pound is much more susceptible to further L-lactide ring-open-
ing than Ph2Sn[OCHMeC(O)NMe2]2 , making it impossible
to isolate the OPri containing compound in its pure form.
Presumably the OCHMeC(O)NMe2 moiety forms a stronger,
more stable chelate to Sn than does the OCHMeC(O)OPri
moiety. (See below for a discussion of stability and NOE data.)
The relative stability of these two compounds manifests
itself in the polymerization studies where, depending upon
which initiating group is used, NMe2 vs. OPri, striking differ-
ences are observed in the molecular weight distribution and
polymer type (chains or cycles).11
(e) Stability and reactivity of the ring-opened products,
Ph3Sn–(cyclic ester)–X and Ph2Sn–[(cyclic ester)–X]2. 1:1 reac-
tions between Ph3SnX or Ph2SnX2 and cyclic esters have lead
to the formation of, in some cases, discrete compounds that
have sufficient stability for isolation and characterization.
Fig. 5 Possible 5- and 8-membered rings of Ar3SnOP compounds.
1172
New J. Chem., 2003, 27, 1167–1176

View Article Online
Scheme 5 Reactivity of Ph3Sn–(cyclic ester)–X and Ph2Sn–[(cyclic ester)–X]2 compounds towards transesterification and chain transfer reactions.
Ph3Sn[OCHMeC(O)OCHMeC(O)X] compounds follows the
order X ¼ NMe2 > OPri > OMe, where the OMe compound
is impossible to isolate free from any transesterification reac-
tions. This observed order of stability probably reflects both
the superior chelating ability of the –C(O)NMe2 moiety (if
an 8-membered chelate forms) and steric factors where the –
C(O)OMe group provides less steric protection through chela-
tion to tin(IV) versus the corresponding –C(O)OPri moiety. If
chelation to the metal center is strong then Sn–OR migration
(which is necessary for transesterification) will be less facile.
The transesterification products, Ph3Sn[OCHMeC(O)X]
(where X ¼ NMe2 , OMe) can be independently prepared from
the direct reaction between Ph3SnX (2 equiv.) and L-lac-
tide (1 equiv.) (Scheme 5, reaction ii). This reaction is similar to
that seen between Ph2Sn(NMe2)2 and L-lactide (1:1 ratio),
where the initially formed Ph2Sn[OCHMeC(O)OCHMe-
C(O)NMe2](NMe2) rapidly undergoes an intramolecular
transamidation reaction to give Ph2Sn[OCHMeC(O)NMe2]2
(see above).
group migration whilst all Ph3Sn[OCHMeC(O)OCHMe-
C(O)X] compounds do (Scheme 5, reactions iv and v). This
has significant ramifications in ROP studies since the kinetic
analysis is not complicated by the formation of kinetically
inactive species such as Ph4Sn.
Concluding remarks
The reactions between Ph3SnX and Ph2SnX2 , where
X ¼ NMe2 or an alkoxide, and L-lactide have allowed us to
examine the influence of the initiator group on the rate of
ring-opening. The order of reactivity X ¼ NMe2 ꢀ OMe >
OPri > OBut > OPh > OCH(CF3)2 > PPh2 clearly impli-
cates the combined influence of electronic and steric factors.
The importance of steric factors is also seen in the complete
lack of reactivity between (o-MeC6H4)3SnNMe2 and L-lactide
and between Ph3SnNMe2 and the tetramethylated cyclic ester,
3,3,6,6-tetramethyl-1,4-dioxane-2,5-dione. The ring-opened
products Ph3Sn[OCHMeC(O)OCHMeC(O)X] were shown to
be kinetically labile to an intermolecular transesterification
reaction and the stability of the –OCHMeC(O)NMe2 ligand
can be understood in terms of chelation via the nucleophilic
carbonyl group, which is stabilized by the NMe2 group to form
a 5-membered ring involving Sn. From the studies of the
Remarkable stability is seen for the Ph3Sn[OCMe2-
C(O)OCHMeC(O)NMe2] compound where no transesterifica-
tion is seen, even upon heating (Scheme 5, reaction iii). This
increased stability must arise from the presence of the tin(IV)
tertiary alkoxide, SnOCMe2R, and the increased shielding of
=
the ketonic C O group. Both factors suppress intermolecular
reactions.
kinetics of the ring-opening event in the 1:1 reaction between
¼
The lability of ligand exchange in these reactions is also
manifest by the fact that some Ph4Sn and Ph2Sn[OCHMe-
C(O)NMe2]2 are also formed (as identified by mass spectrome-
try and 1H, 13C and 119Sn NMR spectroscopy) in the
decomposition of Ph3Sn[OCHMeC(O)OCHMeC(O)NMe2]
(Scheme 5, reaction iv). These side-reactions become more
prevalent upon heating to 60–80 ꢄC.
Ph3SnOPri and L-lactide the activation parameters DH
¼
¼
13(1) kcal molꢂ1 and DS ¼ ꢂ37(3) eu have been determined.
These are entirely consistent with a bimolecular reaction
wherein the tin-alkoxide moiety adds across the ketonic C–O
bond of the lactide. The metal center serves to assist in the acti-
=
vation of the C O moiety by acting as a weak electrophile
while the neighboring alkoxide ligand participates in the
nucleophilic attack on the carbon of the coordinated carbonyl
moiety. It is reasonable to suppose that this is slow for Sn(IV)
A very important observation was the stability of
Ph2Sn[OCHMeC(O)NMe2]2 , which does not experience aryl
New J. Chem., 2003, 27, 1167–1176
1173

View Article Online
tin(IV) (0.39 g, 1.00 mmol), a toluene (10 mL) solution of L-lac-
relative to say Sn(II), Zn(II) and Mg(II) because the Sn–OR
bond is notably less polar.
tide (0.14 g, 1.00 mmol) was slowly added. The colorless reac-
tion solution was stirred for 30 min after which time the
toluene was removed under vacuum to give an opaque semi-
solid (yield 0.42 g, 78%). Anal. calcd for C26H29NO4Sn: C,
58.02; H, 5.43; N, 2.60; found: C, 58.08; H, 5.88; N, 2.85. IR
(liquid)/cmꢂ1: 3081 w, 2987 w, 2934 w, 1755 m, 1713 m,
1667 s, 1617 m, 1480 m, 1429 s, 1403 w, 1376 w, 1331 w,
1303 w, 1259 m, 1235 m, 1192 m, 1148 s, 1075 s, 1024 m,
997 m, 938 w, 870 w, 812 w, 730 vs, 699 vs, 501 m, 450 m.
Experimental
General considerations
Caution: organotin(IV) compounds are highly toxic and require
appropriate handling!
The manipulation of air-sensitive compounds involved
standard Schlenk line and dry box techniques. All solvents
were distilled under nitrogen from alkali metals (sodium or
1H NMR (400 MHz, benzene-d6):
d 0.99 [d, CHMe-
C(O)NMe2 , 3H], 1.53 (d, SnOCHMe, 3H), 2.11 (s, NMe2 ,
3H), 2.47 (s, NMe2 , 3H), 4.75 (q, SnOCHMe, 1H), 4.83 [q,
CHMeC(O)NMe2 , 1H], 7.16 (m, m-H and p-H, 9H), 7.80
(dd, o-H, 6H, JHH 7.9 and 1.5 Hz, 119/117Sn satellites JSnH
117Sn 64, 119Sn 49 Hz). 13C{1H} NMR (126 MHz, benzene-
d6): d 16.30 [s, CHMeC(O)NMe2], 23.25 (s, SnOCHMe),
35.31 (s, NMe2), 35.74 (s, NMe2), 68.55 [s, CHMeC(O)NMe2],
69.24 (s, SnOCHMe, 119/117Sn satellites JSnC117Sn 46, 119Sn 44
Hz), 128.83 (s, m-C), 129.45 (s, p-C), 137.27 (s, o-C, 119/117Sn
satellites JSnC 47 Hz), 142.21 (s, ipso-C), 168.74 [s, CHMe-
C(O)NMe2], 180.74 [s, SnOCHMeC(O)O]. 119Sn NMR (149
MHz, benzene-d6): d ꢂ124 (s). ESI-HRMS (m/z) calcd for
C26H29NO4Sn [MNa+]: 562.1016; found: 562.1056 (7.1 ppm).
˚
sodium/potassium alloy) and stored over 4 A molecular sieves.
L-Lactide was purchased from Aldrich and sublimed prior to
use. Propylene carbonate was purchased from Aldrich and
was degassed and stored over activated sieves. o-Bromoto-
luene, chlorotriphenyltin(IV), tetrachlorotin(IV), magnesium
turnings, benzene-d6 , toluene-d8 , anhydrous tert-butanol and
iso-propanol were purchased from Acros Scientific whilst
diphenyltin(IV) dichloride was purchased from Alfa Aesar.
Benzene-d6 and toluene-d8 were dried over sodium and
vacuum transferred to a Schlenk flask containing activated
molecular sieves. All aryl bromides were stored over calcium
hydride and tetrachlorotin(IV) was stored over phosphorus
pentoxide whilst all other reagents were used as received.
The (o-MeC6H4)3SnCl complex was synthesized by literature
procedures14 where the Grignard reagent was generated in
either THF or diethyl ether from magnesium and the appropri-
ate arylbromide. This Grignard reagent was then treated in situ
with tetrachlorotin(IV) (4:1), dissolved in either benzene or hex-
anes, to afford Ar4Sn. The Ar3SnCl complex is formed from
reaction between Ar4Sn and tetrachlorotin(IV) (3:1) via the
Kocheshkov redistribution reaction.15 Lithium dimethylamide
was synthesized by treating n-butyllithium in hexanes with
excess dimethylamine at ꢂ78 ꢄC. After allowing the reaction
mixture to warm to room temperature, the solvent and excess
amine were removed to quantitatively yield white lithium
dimethylamide. Synthesis of lactide-like compounds such as
3-methyl-1,4-dioxane-2,5-dione (methylglycolide) and 3,3,
6-trimethyl-1,4-dioxane-2,5-dione were based on reported
methods.16
Ph2Sn[OCHMeC(O)NMe2]2. To a cooled toluene solution
(0 ꢄC 10 mL) of bis(dimethylamido)diphenyltin(IV) (72 mg,
0.20 mmol), a toluene (15 mL) solution of L-lactide (29 mg,
0.20 mmol) was slowly added. The colorless reaction solution
was stirred for 30 min after which time a white precipitate
formed, which was filtered off to give the title complex (yield
78 mg, 77%). Anal. calcd for C22H30N2O4Sn: C, 52.31; H,
5.99; found: C, 51.54; H, 5.92. IR (Nujol)/cmꢂ1: 1609 vs,
1596 vs, 1404 w, 1330 w, 1256 w, 1180 w, 1129 s, 1074 m,
1047 w, 906 w, 781 w, 730 m, 704 m, 656 w, 637 w, 551 w,
456 m, 443 m.1H NMR (400 MHz, benzene-d6): d 1.43 (d,
SnOCHMe, 6H), 1.87 (s, NMe2 , 6H), 2.26 (s, NMe2 , 6H),
4.87 (q, SnOCHMe, 2H), 7.21 (t, p-H, 2H), 7.35 (t, m-H,
4H), 8.37 (dd, o-H, 4H, JHH 7.6 and 1.2 Hz, 119/117Sn satellites:
JSnH 117Sn 75, 119Sn 59 Hz). 13C{1H} NMR (126 MHz, ben-
zene-d6): d 22.85 (s, SnOCHMe), 35.42 (s, NMe2), 35.93 (s,
NMe2), 66.13 [s, SnOCHMeC(O)NMe2 , 119/117Sn satellites:
JSnC 36 Hz], 127.83 (s, m-C), 128.36 (s, p-C), 136.88 (s, o-C,
119/117Sn satellites: JSnC 47 Hz), 152.83 (s, ipso-C), 181.39 (s,
CO). 119Sn NMR (149 MHz, benzene-d6 , 300 K): d ꢂ346 (br
s). 119Sn NMR (187 MHz, THF-d8 , 300 K): d ꢂ350 (br s);
(233 K): d ꢂ360 (s), ꢂ362 (br); (213 K): d ꢂ360 (s, 3Sn),
ꢂ362 (s, 1Sn). ESI-HRMS (m/z) calcd for C22H30N2O4Sn
[MNa+]: 529.1125; found: 529.1130 (0.9 ppm).
The synthesis of Ar3SnNMe2 (Ar ¼ Ph, o-MeC6H4,) and
Ph2Sn(NMe2)2 complexes was based on the reported synth-
esis17 of Ph3SnNMe2 and Ph2Sn(NMe2)2 where Ar3SnCl or
Ph2SnCl2 was treated with 1 or 2 equiv. of LiNMe2 , respec-
tively. The major difference in the synthesis was that LiNMe2
was not generated in situ. The full preparation and character-
ization of Ph3SnX and Ph2Sn(NMe2)2 compounds where
X ¼ NMe2 , OMe, OPri, OPh, OBut, OCH(CF3)2 , PPh2 is
given in the ESI.
Ph3Sn[OCMe2C(O)OCHMeC(O)NMe2]. To a cooled tolu-
ene solution (0 ꢄC 10 mL) of dimethylamidotriphenyltin(IV)
(0.39 g, 1.00 mmol), a toluene (10 mL) solution of 3,3,6-
trimethyl-1,4-dioxane-2,5-dione (0.13 g, 1.00 mmol) was added
slowly. The colorless solution was stirred for 30 min after
which time the toluene was reduced and hexanes were added
to the reaction, which was then cooled to ꢂ20 ꢄC for 12 h,
causing precipitation of the complex. Subsequent filtration
afforded the title complex (yield 0.39 g, 71%). Anal. calcd for
C27H31NO4Sn: C, 58.72; H, 5.63; N, 2.54; found: C, 58.40;
H, 5.60; N, 2.42. IR (Nujol)/cmꢂ1: 3062 m, 3042 m, 1695 m,
1507 w, 1481 m, 1429 s, 1354 m, 1311 m, 1260 w, 1168 s,
1116 w, 1077 s, 1025 m, 1007 w, 997 w, 916 vw, 888 vw, 845
w, 789 w, 762 vw, 731 vs, 702 s, 661 m, 619 vw, 570 w, 531
w, 448 m. 1H NMR (500 MHz, benzene-d6): d 1.06 (d, CHMe-
CONMe2 , 3H), 1.56 (s, SnOCMe2, 3H), 1.65 (s, SnOCMe2,
3H), 2.12 (s, NMe2 , 3H), 2.47 (s, NMe2 , 3H), 4.87 [q, CHMe-
C(O)NMe2 , 1H], 7.17 (m, p-H, 3H), 7.23 (m, m-H, 6H),
7.87 (dd, o-H, 6H, JHH 8.2 and 1.3 Hz, 119/117Sn satellites: JSnH
117Sn 64, 119Sn 48 Hz). 13C{1H} NMR (125 MHz, benzene-d6):
Measurements and analyses
1H, 13C{1H} and 119Sn spectra were obtained from either Bru-
ker DPX-400 or DRX-500 NMR spectrometers using either
benzene-d6 or toluene-d8 solvents. Spectra were referenced
internally to the residual protio impurities for 1H and 13C
(C6D6 d 7.15 1H, 128.0 13C; C6D5CD3 d 2.09) or externally
to Me4Sn (d ¼ 0) for 119Sn. Infrared data were obtained from
a Perkin Elmer Spectrum GX spectrophotometer with samples
sandwiched between potassium bromide or sodium chloride
plates as Nujol1 mulls (solids) or as neat liquids/semi-solids.
Mass spectra were obtained with a Mircomass QTOF. Elemen-
tal analyses were performed by Atlantic Microlab, Inc., with
samples sealed under nitrogen in glass ampoules.
Syntheses
Ph3Sn[OCHMeC(O)OCHMeC(O)NMe2]. To
toluene solution (0 ꢄC 10 mL) of dimethylamidotriphenyl-
a
cooled
1174
New J. Chem., 2003, 27, 1167–1176

View Article Online
d 16.38 [s, CHMeC(O)NMe2], 29.50 (s, SnOCMe2), 29.94 (s,
SnOCMe2), 35.38 (s, NMe2), 36.77 (s, NMe2), 67.02 [s, CHMe-
C(O)NMe2], 73.31 (s, SnOCMe2 , 119/117Sn satellites: JSnC29
Hz), 128.72 (s, m-C), 129.36 (s, p-C), 137.32 (s, o-C, 119/117Sn
satellites: JSnC44 Hz), 143.19 (s, ipso-C), 168.66 [s, CHMe-
C(O)NMe2], 183.01 [s, SnOCMe2C(O)]. 119Sn NMR (187
MHz, benzene-d6): d ꢂ141 (s). ESI-LRMS (m/z): 554 [MH+].
CHMe), 4.25 (dd, overlapping Ha and Hb , SnOCHMeCH2),
4.40 (sextet, SnOCHMeCH2), 5.28 (sextet, SnOCH2CHMe)
7.15 (m, m- and p-H), 7.67 (m, o-H).
Kinetics analysis
Standard solutions of Ph3SnOPri and L-lactide were made in
benzene-d6 and stored in the dry box. Appropriate aliquots
of both reagents were transferred to a J. Young1 NMR tube.
The total volume was made up to 800 mL with benzene-d6 to
ensure a constant initial L-lactide concentration (0.039 M).
The reaction temperatures were regulated via a thermosta-
tically controlled oil bath.
An overall second-order process, first-order in both
Ph3SnOPri [A] and L-lactide [B], was assumed. The rate of dis-
appearance of both [A] and [B] {based on the formation
of Ph3SnOCHMeC(O)OCHMeC(O)OPri [C]} was determined
by plotting ln([B]/[A]) vs. time (s). The rate constants for these
reactions were determined from the gradient of the graph via
m ¼ k([Bo] ꢂ [Ao]).
Ph3Sn[OCHMeC(O)OCHMeC(O)OPri]. iso-Proproxytri-
phenyltin(IV) (0.13 g, 0.32 mmol) was added to L-lactide (45
mg, 0.31 mmol) in benzene. The colorless solution was stirred
for 1 day at 60 ꢄC after which time the benzene was removed to
afford a white semi-solid compound (yield 0.10 g, 58%). Anal.
calcd for C27H30O5Sn: C, 58.62; H, 5.47; found: C, 57.90; H,
5.22. IR (Nujol)/cmꢂ1: 1740 vs, 1457 s, 1261 m, 1071 vs, 800
s, 730 s 698 s. 1H NMR (500 MHz, benzene-d6): d 0.78
[d, CHMeC(O)OCHMe2 , 3H], 0.92 (d, OCHMe2, 3H), 1.03
(d, OCHMe2, 3H), 1.58 (d, SnOCHMe, 3H), 4.79 (q, SnOCHMe,
1H), 4.84 [overlapping signals: q, CHMeC(O)OCHMe2 ; sept.,
OCH(CH3)21H], 7.15 (m, p- and m-H, 6H), 7.87 (dd, o-H, 6H
JHH 8.4 and 1.4 Hz, 119/117Sn satellites: JSnH 117Sn 62, 119Sn
46 Hz). 13C{1H} NMR (125 MHz, benzene-d6): d 16.47
[s, CHMeC(O)OCHMe2], 21.32 (s, SnOCHMe2), 21.43
(s, SnOCHMe2), 28.11 (s, SnOCHMe), 68.94 [s, CHMeC(O)-
OCHMe2)], 69.15 (s, SnOCHMe), 70.13 (s, SnOCHMe2)
128.76 (s, m-C), 129.23 (s, p-C), 137.31 (s, o-C), 142.01 (s,
ipso-C), 169.39 [s, CHMeC(O)OCHMe2], 180.58 [s, SnOCH-
MeC(O)]. 119Sn NMR (187 MHz, benzene-d6): d ꢂ123 (s).
Acknowledgements
The authors wish to acknowledge the financial support from
the Department of Energy, Office of Basic Sciences, Chemistry
Division, and the reviewers for their careful reading of the
manuscript and their helpful comments.
Ph3Sn[OCHMeC(O)OCHMeC(O)OMe]. Methoxyoxytri-
phenyltin(IV) (93 mg, 0.24 mmol) dissolved in benzene (40
mL) was slowly added to L-lactide (35 mg, 0.24 mmol) in ben-
zene (40 mL). The colorless solution was stirred for < 5 min at
room temperature after which time the benzene was removed
to afford a clear liquid, which was extracted with hexane to
remove any unreacted L-lactide. Removal of the hexane gave
82 mg (64%) of the title compound. This compound readily
undergoes self-transesterification, hence the product has only
been characterized by NMR spectroscopy. 1H NMR (400
MHz, benzene-d6): d 1.00 [d, CHMeC(O)OMe, 3H], 1.55 (d,
OCHMe, 3H), 3.15 (s, OMe, 3H), 4.77 and 4.81 [overlapping
q, SnOCHMeC(O)OCHMeC(O)OMe, 2H], 7.16 (m, p- and
m-H, 6H), 7.85 (d, o-H, 6H, 119/117Sn satellites: JSnH 117Sn
63, 119Sn 49 Hz). 13C{1H} NMR (100 MHz, benzene-d6): d
16.46 [s, CHMeC(O)OMe], 23.20 (s, SnOCHMe2), 51.68 (s,
OMe), 69.37 [s, CHMeC(O)OMe], 71.86 (s, SnOCHMe),
128.90 (s, m-C), 129.37 (s, p-C), 137.26 (s, o-C, 119/117Sn satel-
lites: JSnH 117/119Sn 45 Hz), 141.89 (s, ipso-C), 170.19
[s, CHMeC(O)OMe], 180.43 [s, SnOCHMeC(O)]. 119Sn NMR
(149 MHz, benzene-d6): d ꢂ121 (s).
References
1
2
3
Ecochem is a polylactide based packaging material developed by
DuPont ConAgra.
Leupron Depot is a product of Takeda Chemical Industries,
Ltd., Japan for drug delivery purposes.
(a) E. E. Schmitt and R. A. Polistina, U.S. Pat., 1969, 3463158; (b)
E. J. Frazza and E. E. Schmitt, J. Biomed. Mater. Res. Symp.,
1971, 1, 43; (c) Dexon and Vicry are products of Davis & Geek
Corporation (Wayne, NJ) and Ethicon, Inc. (Somerville, NJ),
respectively; (d ) J. P. Pennings, H. Dijkstra and A. J. Pennings,
Polymer, 1993, 34, 942.
4
5
(a) J. A. Hubbell and R. Langer, Chem. Eng. News, March 13,
1995, p. 42; (b) R. Langer and J. P. Vacanti, Science, 1993,
260, 920.
(a) T. M. Ovitt and G. W. Coates, J. Am. Chem. Soc., 1999, 121,
4072; (b) C. P. Radano, G. L. Baker and M. R. Smith, J. Am.
Chem. Soc., 2000, 122, 1552; (c) M. Cheng, A. B. Attygalle,
E. B. Lobkovsky and G. W. Coates, J. Am. Chem. Soc., 1999,
121, 11 584; (d ) M. H. Chisholm, J. Gallucci and K. Phomphrai,
Inorg. Chem., 2002, 41, 2785; (e) M. H. Chisholm, N. W. Eilerts,
J. C. Huffman, S. S. Iyer, M. Pacold and K. Phomphrai, J. Am.
Chem. Soc., 2000, 122, 11 845; A. P. Dove, V. C. Gibson, E. L.
Marshall, A. J. P. White and D. J. Williams, Chem. Commun.,
2001, 283; (g) B. M. Chamberlain, M. Cheng, D. R. Moore,
T. M. Ovitt, E. B. Lobkovsky and G. W. Coates, J. Am. Chem.
Soc., 2001, 123, 3229.
Reaction between Ph3SnNMe2 and propylene carbonate. To a
cooled toluene solution (0 ꢄC, 10 mL) of dimethylamidotriphe-
nyltin(IV) (0.39 g, 1.00 mmol) a toluene–hexane (50:10 mL)
solution of propylene carbonate (0.10 g, 1.00 mmol) was added
slowly. The colorless solution was stirred for 30 min at 0 ꢄC
after which time the toluene was removed. An opaque semi-
solid, 0.23 g, of a mixture of Ph3SnOCH2CHMeOC(O)NMe2
and Ph3SnOCHMeCH2OC(O)NMe2 (40:60) was identified by
1H NMR spectroscopy (see below). Modification of the reac-
tion conditions (RT and ꢂ50 ꢄC) did not cause a change in
product ratio. IR (liquid)/cmꢂ1: 3066 m, 3049 m, 3019 w,
2966 m, 2931 m, 2870 m, 1810 br m, 1702 br vs, 1580 w,
1569 wv, 1495 m, 1482 m, 1455 m, 1430 s, 1397 s, 1375 m,
1304 w, 1275 m, 1262 m, 1193 vs, 1153 m, 1120 m, 1077 m,
1023 w, 997 m, 939 w, 860 w, 799 w, 770 m, 730 vs, 699 vs,
6
7
B. J. O’Keefe, M. A. Hillmyer and W. B. Tolman, J. Chem. Soc.,
Dalton Trans., 2001, 2215.
M. T. Zell, B. E. Padden, A. J. Paterick, K. A. M. Thakur, R. T.
Kean, M. A. Hillmyer and E. J. Munson, Macromolecules, 2002,
35, 7700.
8
(a) Recent examples include: K. B. Aubreecht, M. A. Hillmyer
and W. B. Tolman, Macromolecules, 2002, 35, 650; (b) M. Moller,
F. Nederberg, L. S. Lim, R. Kange, C. J. Hawker, J. L. Hedrick,
Y. Gu, R. Shah and N. L Abbott, J. Polym. Sci., Part A: Polym.
Chem.., 2001, 39, 3529; (c) M. Ryner, A.-C. Albertsson and H. R.
Kricheldorf, Macromolecules, 2001, 34, 7281; (d ) J. Libiszowski,
A. Kowalski, A. Duda and S. Penczek, Macromol. Chem. Phys.,
2002, 203, 1694.
9
(a) A. Kowalski, J. Libiszowski, A. Duda and S. Penczek, Macro-
molecules, 2000, 33, 1964; (b) J. E. Kemnitzer, S. P. McCarthy and
R. A. Gross, Macromolecules, 1993, 26, 6144.
1
661 w, 618 w, 539 w, 451 s, 370 vw, 386 w, 374 s. H NMR
(500 MHz, benzene-d6): d 1.21 (d, SnOCHMeCH2) and 1.28
(d, SnOCH2CHMe), 2.37 and 2.56 [overlapping br s,
OC(O)NMe2), 4.05 (dd, overlapping Ha and Hb , SnOCH2-
10 M. H. Chisholm and E. E. Delbridge, Chem. Commun., 2001,
1308.
New J. Chem., 2003, 27, 1167–1176
1175

View Article Online
1986, 122, 45; (c) A. Stern and E. I. Becker, J. Org. Chem.,
1964, 29, 3221; (d ) I. Wharf and M. G. Simard, J. Organomet.
Chem., 1997, 532, 1.
11 M. H. Chisholm and E. E. Delbridge, New J. Chem., 2003, 27,
10.1039/b300351e (following paper).
12 ROP studies of L-lactide, methylglycolide and glycolide suggest
that ROP is more rapid for glycolide and methylglycolide vs.
L-lactide such that kprop ꢃ kro
15 G. Eng and C. R. Dillard, Inorg. Chim. Acta, 1978, 28, 99.
16 (a) Z.-R. Shen, J.-H. Zhu and Z. Ma, Makromol. Chem. Rapid
Commun., 1993, 14, 457; (b) T. A. Augurt, M. N. Rosensaft and
V. A. Perciaccante, U.S. Pat., 1977, 4033938.
11
.
13 H. Yasuda, J.-C. Choi, S.-C. Lee and T. Sakakura, J. Organomet.
Chem., 2002, 659, 133.
14 (a) I. Wharf, Inorg. Chim. Acta, 1989, 159, 41; (b) B. King, H.
Eckert, D. Z. Denney and R. H. Herber, Inorg. Chim. Acta,
17 K. Jones and M. F. Lappert, J. Chem. Soc. A, 1965, 1944.
1176
New J. Chem., 2003, 27, 1167–1176