Preparation, characterization, and application of polymer-supported stannols and distannoxanes

Article abstract of DOI:10.1021/om990630d

Novel organotin copolymers derived from poly-3- and poly-4-(2-di-n-butylchlorostannyl)-ethylstyrene-co-divinylbenzene-co-styrene have been prepared. The copolymers were prepared with various proportions of monomers and converted to organotin oxides and organotin carboxylates. Copolymer-bound organotin oxides, as a mixture of stannol and distannoxane, have been shown to catalyze the lactonization of hydroxycarboxylic acids.

Full text of DOI:10.1021/om990630d

Organometallics 1999, 18, 5577-5583
5577
P r ep a r a tion , Ch a r a cter iza tion , a n d Ap p lica tion of
†
P olym er -Su p p or ted Sta n n ols a n d Dista n n oxa n es
Duncan H. Hunter* and Chantelle McRoberts
Department of Chemistry, University of Western Ontario, London, ON N6A 5B7, Canada
Received August 6, 1999
Novel organotin copolymers derived from poly-3- and poly-4-(2-di-n-butylchlorostannyl)-
ethylstyrene-co-divinylbenzene-co-styrene have been prepared. The copolymers were prepared
with various proportions of monomers and converted to organotin oxides and organotin
carboxylates. Copolymer-bound organotin oxides, as a mixture of stannol and distannoxane,
have been shown to catalyze the lactonization of hydroxycarboxylic acids.
In tr od u ction
In recent years, this laboratory has been developing
an approach to radiopharmaceutical and pharmaceuti-
9
The synthetic versatility of organotin reagents is well
documented, and several exhaustive reviews exist.
cal preparation based on this same polymer-bound
trialkylchlorostannane. Investigations have focused on
a polymer-bound chlorostannane with the highest at-
tainable loading capacity. This present study reports on
the preparation and characterization of trialkylchlo-
rostannane copolymers of varying loading capacity.
These copolymers have been converted to distannoxane,
stannol, and stannyl carboxylate derivatives. Their
application as catalysts for esterfication reactions has
been investigated.
1
Common organotin reagents include trialkyltin halides,
which can be converted to trialkylarylstannanes, silyl-
stannanes, and distannanes as well as stannols and
distannoxanes. These tin derivatives have been applied
to numerous organic transformations. Of particular
interest to this work is the application of organostannols
and distannoxanes to esterfication reactions.2
Despite the versatility of organotin reagents, they are
generally toxic and quantitative removal of organotin
impurities from reaction mixtures is often difficult. To
alleviate these pitfalls, an insoluble cross-linked poly-
styrene-supported trialkylchlorostannane has been pre-
Resu lts a n d Discu ssion
Cop olym er P r ep a r a tion . As illustrated in Scheme
1, three copolymers, 2A-C, have been prepared using
different amounts of 1, styrene, and divinylbenzene
(Table 1). Neither 1 nor 2A-C are commercially avail-
able but have been prepared by modification of a
3
pared and applied to the Geise reaction and Stille
4
coupling reaction. In addition, it has been reduced to
5
form a tethered tin hydride. This immobilized trialkyl-
stannane was effective for free radical cyclizations with
a reactivity similar to tri-n-butylstannane. However,
unlike tri-n-butylstannane, tin impurities were removed
by simple filtration. This polymer-supported tin hydride
9
published procedure as described in the Experimental
Section. Each of these copolymers was a granular white
solid, insoluble but swellable in both organic and
alcoholic solvents, including chloroform and ethanol.
The extent of cross-linking, as determined by the
proportion of divinylbenzene employed, remained con-
stant at 15 mol %. However, the mol % of 1 was
decreased from 85% (2A) to 42% (2B) to 8.5% (2C).
Technical grade divinylbenzene as a mixture of meta
and para isomers was used for all preparations. In
addition, the technical mixture contained 20 mol %
ethylvinylbenzene. Because polymer yields were not
quantitative (50-73 wt %) and the relative reactivity
of each monomer is unknown, the actual proportions of
the monomers in the resultant copolymers are unknown
and ethylvinylbenzene might also have been incorpo-
rated. Consequently, the structure drawn in Scheme 1
for 2A-C is not meant to suggest a block copolymer and
the ethylvinylbenzene component has been omitted.
6
was applied to free radical dehalogenations, deamina-
6
7
tions, dehydroxylations, and ring enlargement reac-
8
tions. This insoluble polymer-supported tin system
overcomes purification and toxicity concerns often as-
sociated with tin reagents.
*
To whom correspondence should be addressed. E-mail: dhunter@
julian.uwo.ca.
†
This work contains parts of the Ph.D. Thesis of C. McRoberts, 1999,
University of Western Ontario.
1) (a) Van Der Kerk, G. J . M. In Organotin Compounds: New
Chemistry and Applications Zuckerman, J . J ., Eds.; Advances in
Chemistry Series; 1976; p 157. (b) Davies, A. G. Organotin Chemistry:
VCH: Weinheim, 1997. (c) Smith, P. J . Chemistry of Tin; Blackie:
London, 1998. (d) Pereyre, M.; Quintard, J . P.; Rahm, A. Tin in Organic
Synthesis; Butterworths: London, 1986.
2) Mascaretti, O. A.; Furlan, R. L. E. Aldrichim. Acta 1997, 30, 55,
and references therein.
3) Bokelman, C.; Neumann, W. P.; Peterseim, M. J . Chem. Soc.,
Perkin Trans. 1 1992, 3165.
(
(
(
1
19
Solid-state MAS
Sn NMR spectroscopy of each
(
4) Kuhn, H.; Neumann, W. P. Synlett 1994, 123.
(
5) Gerigk, U.; Gerlach, M.; Neumann, W. P.; Vieler, R.; Weintritt,
copolymer showed only one signal at 148-150 ppm,
which is consistent with a trialkylchlorostannane such
V. Synthesis 1990, 448.
6) Gerlach, M.; J ordens, F.; Kuhn, H.; Neumann, W. P.; Peterseim,
M. J . Org. Chem. 1991, 56, 5971.
(
(
7) Neumann, W. P.; Peterseim, M. Synlett 1992, 801.
(9) (a)Hunter, D. H.; Marinescu, A. M.; Loc’h, C.; Maziere, B. J .
Labelled Cmpd. Radiopharm. 1995, 37, 144. (b) Hunter, D. H.; Zhu, X.
Z. J . Labelled Cmpd. Radiopharm 1999, 42, 653.
(8) Dygutsch, D. P.; Neumann, W. P.; Peterseim, M. Synlett 1994,
3
63.
1
0.1021/om990630d CCC: $18.00 © 1999 American Chemical Society
Publication on Web 11/20/1999

5
578 Organometallics, Vol. 18, No. 26, 1999
Hunter and McRoberts
Sch em e 1
Solution phase organotin reagents including hexa-n-
butyldistannoxane, tri-n-butylchlorostannane, and di-
n-butyltin oxide are known to catalyze esterfication
2
reactions. Tri-n-butylchlorostannane and hexa-n-bu-
tyldistannoxane have also found use in protecting group
1
1
chemistry. The polymer-supported tin oxides (4A-C),
obtained by hydrolysis, provide a potential solid phase
substitute for these solution phase reagents. As well as
overcoming solution phase problems of purification and
toxicity, the solid phase may impart special reactivity
due to the spatially discrete catalytic centers and the
possible effects of anchoring.
Con ver sion to Tin Ca r boxyla tes. Trialkylsilyl
groups are commonly used as protecting groups, and
trialkylstannyl groups have also been investigated.12
With trialkystannyl groups, carboxylic acids can be both
protected and deprotected under mild conditions, al-
though tri-n-butylstannyl carboxylates prove sensitive
to moisture. Attachment of the trialkylstannyl protect-
ing group to an insoluble solid support could provide
the advantage of simplifying isolation and purification
procedures during multistep syntheses, as well as
reducing the rate of tin carboxylate hydrolysis.
Ta ble 1. Qu a n tities of Mon om er s Used in th e
P r ep a r a tion of P oly-3- a n d
P oly-4-(2-(d ibu tylch lor osta n n yl)-
eth ylstyr en e-co-d ivin ylben zen e-co-styr en e
Cop olym er s
divinyl-
benzene,
mmol
loading
capacity,
mmol/g
As a model for solid-phase carboxylic acid protecting
groups, the copolymer-supported stannyl toluate 3 was
prepared using the highest loading capacity trialkyl-
chlorostannane 2A. As illustrated in Scheme 2, reaction
of copolymer 2A with sodium toluate led to a new
polymeric material, 3. The solid-state MAS 1 Sn NMR
spectrum showed a disappearance of the signal at 148
ppm and the appearance of a new tin signal at 104 ppm
1
,
styrene,
mmol
yield,
wt %
copolymer
mmol
2
2
2
A
B
C
225
112
22.5
0
112
202
40
40
40
50
73
55
1.6
1.1
0.24
19
as tributylchlorostannane at 141 ppm1a or 1 at 143 ppm.
The absence of additional peaks indicated that there
was no hydrolysis of the chlorostannane copolymers
during polymerization despite the refluxing biphasic
1
3
consistent with a tin carboxylate. In addition, IR
spectroscopy further confirmed the structure with a
characteristic tin carboxylate CdO stretch at 1645
1
3
aqueous conditions employed. The MAS
C NMR
-
1 14
spectrum of copolymer 2A showed signals consistent
with the anticipated structure. No signals typical of
vinyl carbons were observed, indicating complete reac-
tion of each double bond of the divinylbenzene. Energy-
dispersive X-ray (EDX) analyses were consistent with
approximately equal amounts of tin and chlorine atoms
within each copolymer.
cm
.
The loading capacity of copolymer 3 was deter-
mined from the amount of toluic acid released upon
treatment with trifluoroacetic acid and was found to be
1.57 mmol/g.
The stability of this tethered tin carboxylate was
determined under conditions of aqueous acetic acid,
anhydrous trifluoroacetic acid, and methanol. The
amount of toluic acid cleavage was quantified by HPLC,
and a time course for each reaction was obtained. The
polymer-supported stannyl toluate 3 displayed poor
stability under these conditions, ranging from 50% to
100% cleavage within 15 min of exposure. Under protic
conditions. transesterification was observed. For ex-
ample, trifluoroacetic acid in methylene chloride re-
sulted in quantitative release of toluic acid. An IR
spectrum of the recovered copolymer indicated a tin
Overnight treatment of each copolymer in ethanolic
sodium hydroxide solution led to complete hydrolysis,
1
19
as evidenced by solid-state MAS
Sn NMR spectros-
copy. The signal at 148-150 ppm, attributed to a
trialkylchlorostannane, completely disappeared and was
replaced by overlapping signals at 91 and 101 ppm.
These signals are attributed to trialkyltin oxides and
will be discussed later.
The effective proportion of 1 present in each copoly-
1
0
-1
mer was evaluated by argentometric titration of the
chloride released upon basic hydrolysis. These values,
in mmol Cl/g polymer, represent the loading capacity
for all three copolymers and are included in Table 1.
These loading capacities translate to 68, 46, and 10 wt
carboxylate with an IR stretch at 1645 cm and solid-
1
19
state MAS Sn NMR signal at 151 ppm, attributed to
stannyl trifluoroacetate.
From these limited cleavage studies, it can be antici-
pated that most carboxylates will react readily and
quantitatively with the copolymeric chlorostannanes to
produce tin carboxylates. The cleavage studies indicate
that tethering the protecting group to a solid support
%
of monomer 1 in copolymers 2A, 2B, and 2C,
respectively. In comparison, the proportions of monomer
1
7
1
used in the polymer preparations correspond to 94,
3, and 25 wt %, respectively. Thus it seems monomer
is being incorporated into these copolymers in decid-
edly lower proportions than is present in the polymer-
ization reaction mixture.
(11) Frankel, M.; Gertner, D.; Wagner, D.; Zilka, A. J . Org. Chem.
1
965, 30, 1596.
12) Greene, T. Protective Groups in Organic Synthesis; J ohn Wiley
and Sons: New York, 1981.
13) Klein, J .; Borsdorf, R. J . Prakt. Chem. 1993, 335, 465.
(
(
(
10) Day, R. A., J r.; Underwood, A. L. Quantitative Analysis, 2nd
(14) Sandhu, G. K.; Kaur, G.; Holecek, J .; Lycka, A. J . Organomet.
Chem. 1988, 345, 51.
ed.; Prentice Hall: Englewood Cliffs, NJ , 1967.

Polymer-Supported Stannols and Distannoxanes
Organometallics, Vol. 18, No. 26, 1999 5579
Sch em e 2
Sch em e 3
Ta ble 2. Dista n n oxa n e/Sta n n ol Ra tios of
Cop olym er s 4A-C a n d Th eir Loa d in g Ca p a cities
prepared and structurally analogous polymer-supported
trialkyldistannoxanes were reported to give a signal
at 83 ppm.
16
b
loading
ratio of
a
capacity, distannoxane distannoxane, stannol,
In addition to presenting the observed ratio of the
peaks at 91 and 101 ppm, Table 2 also includes the
observed loading capacity of the chlorostannane copoly-
mers 2A-C, used to prepare the tin oxide derivatives,
copolymer
mmol/g
to stannol
mmol/g
mmol/g
4
4
4
A
B
C
1.6
1.1
0.24
1.5
1.0
0.25
0.6
0.37
0.04
0.4
0.37
0.16
4
A-C. On the basis of these loading capacities and the
a
Calculated from the loading capacity of the appropriate
ratio of NMR signal intensities, loadings of the distan-
noxanes and stannols were calculated and are listed.
The copolymer with the highest loading capacity, 4A,
also gave the highest ratio of distannoxane relative to
stannol, 1.5, and also the largest amount of distannox-
ane, 0.6 mmol/g. As the total loading capacity decreased,
so did the distannoxane/stannol ratio.
b
chlorostannane polymer, 2A-C, prior to hydrolysis. Calculated
from the ratio of the peaks at 92 and 101 ppm in the MAS 1 Sn
NMR spectra.
19
does not improve chemical stability of the tin carboxy-
late bond under acidic conditions. As a result of this
chemical instability, no tin carboxylates were prepared
at reduced loading capacities from copolymers 2B and
In solution, an equilibrium between hexa-n-butyl-
distannoxane and tri-n-butylstannol occurs, and the
observed distannoxane/stannol ratios may reflect a
similar equilibrium within the copolymer backbone. The
trend of decreasing distannoxane/stannol ratio with
decreasing loading capacity is in accordance with a
lower “effective concentration” of monomer 1 in the
copolymer as a result of dilution with styrene. Attempts
to drive additional water from 4A at reflux in benzene
and convert the remaining stannol to distannoxane were
unsuccessful. The ratio of the overlapping peaks at 91
and 101 remained unchanged. This indicates more than
a simple equilibrium is involved in determining the
distannoxane/stannol ratio. Also, proximity of tin sites
may play a critical role. Two tin sites in close proximity
are needed to allow for distannoxane formation, while
tin sites in isolation result in stannols.
Ap p lica tion of Tin Oxid e Cop olym er s to ω-Hy-
dr oxycar boxylic Acid Lacton ization s. Although there
are many existing methods for lactonization17 of ω-hy-
droxycarboxylic acids, there are limitations to their
general usefulness. Formation of medium and large
ring-size lactones from the corresponding ω-hydroxy-
carboxylic acids are often plagued by a competition
between intramolecular and intermolecular reaction.
Intermolecular reaction leads to oligomeric products,
resulting in lower lactonization yields. In addition, these
existing methods utilize either acidic or basic conditions,
2
C. While this chlorostannane copolymer system is not
a practical protecting group for carboxylic acids, perhaps
it has applicability to amino acids, as has been re-
ported11 for the tri-n-butylstannyl group.
Con ver sion to Tin Oxid es. As previously men-
tioned, stannols, distannoxanes, and stannoxides have
found application as catalysts for esterification, dees-
2
terification, and lactonization reactions. Since hydroly-
sis of the chlorostannane copolymers led to mixtures of
polymer-bound stannols and distannoxanes 4, as indi-
cated in Scheme 3, the three copolymers were prepared
and characterized for their potential as catalysts. Thus
each of the copolymers, 2A-C, was treated with alco-
holic sodium hydroxide and characterized spectroscopi-
cally. The solid-state MAS 1 Sn NMR spectra of each
of these organotin oxides, 4A-C, showed two overlap-
ping signals, one at 91 and a second at 101 ppm. The
ratio of these peaks differed according to the loading
capacity of the original chlorostannane copolymers,
Table 2. On the basis of solution and solid phase
precedent, the peak at 91 ppm was assigned to distan-
noxane and the peak at 101 ppm to stannol. Our own
solution phase 1 Sn NMR spectra (CDCl3) for hexa-n-
butyldistannnoxane gave a peak at 93.7 ppm and for
tri-n-butylstannol at 105.8 ppm. Hexa-n-butyldistannnox-
ane is reported to give a peak at 92.0 ppm in CDCl3
and at 84.8-84.3 ppm in benzene-d6.15 Previously
19
19
(16) Dumartin, G.; Kharboutli, J .; Delmond, B.; Pereyre, M. Orga-
nometallics 1996, 15, 19.
(17) Ishihara, K.; Kubota, M.; Kurihara, H.; Yamamoto, H. J . Org.
Chem. 1996, 61, 4560, and references therein.
(
15) Lockhart, T. P., Manders, W. F., Brinckman, F. E. J . Orga-
nomet. Chem. 1985, 286, 153.

5
580 Organometallics, Vol. 18, No. 26, 1999
Hunter and McRoberts
Ta ble 4. La cton iza tion of
Ta ble 3. Effect of 3A Stoich iom etr y on
Hexa d econ olid e a n d Oligom er Yield s
16-Hyd r oxyh exa d eca n oic Acid , 5, Usin g 4A-C a s
1
0% Ca ta lysts
lactone yield,b
oligomer yield,c
a
molar ratio of 4A/5
mol %
wt %
polymer
loading,a
mmol/g
lactone
yield,b
mol %
oligomer
yield,
wt %
5 remaining,b
c
no catalyst
0
0
55
29
18
7
0
6
27
27
25
1
copolymer
mol %
0
0
0
1
2
.05
.10
.50
.0
4A
4B
4C
1.6
1.1
0.24
e1
e1
e1
55
35
21
27
41
20
.0
a
Calculated from the loading capacity of the appropriate
a
b
Calculated assuming 1.6 mmol/g of tin sites on copolymer 4A.
chlorostannane polymer, 2A-C, prior to hydrolysis. Determined
b
c
by HPLC. c Determined by gravimetry.
Determined by HPLC. Determined by gravimetry.
limiting the compatibility with other functional groups
present. Tin mediation has been used under nearly
neutral reaction conditions in an attempt to solve this
latter problem. Hexa-n-butyldistannoxane, dibutyltin-
methanol. The lactones were soluble in the methanol
fraction. The methanol-insoluble material was com-
prised of oligomeric lactones, the nature of which will
be described below. The yield of lactone was determined
by HPLC analysis against known standards, and the
yield of oligomeric lactones was determined by gravim-
etry.
1
8
oxide,
tri-n-butylchlorostannane, and tri-n-butyl-
methoxystannane19 have been shown to catalyze ω-hy-
droxycarboxylic acid lactonizations, although mixtures
of lactones and oligomers resulted. Longer chain ω-hy-
droxycarboxylic acids resulted in moderate yields of
lactone accompanied by substantial amounts of oligo-
meric byproducts. Attempts to prepare medium ring-
size lactones (8-12) resulted in oligomeric byproducts
almost exclusively.
As the data in Table 3 indicate, copolymer 4A acts as
a catalyst for lactonization. There is an optimum ratio
for both lactone and oligomer yields, which parallel each
other. A maximum lactone yield, as well as maximum
conversion, was observed with a 4A/5 ratio of 0.1. The
lactone yield drops sharply at smaller ratios of catalytic
sites/5 and more slowly for higher ratios. The observed
optimum lactone yield of 55% is comparable to other
Copolymers 4A-C were applied to lactonization reac-
tions of ω-hydroxycarboxylic acids to determine the
effect of loading capacity on their reactivity as lacton-
ization catalysts. Also, there is a possibility that lower
loading capacities or altered distannoxane/stannol ratios
might optimize lactonization over oligomerization yields.
Since organotin oxides activate ω-hydroxycarboxylic
acids toward esterification, spatially discrete tin oxide
sites could improve the competition between lactoniza-
tion and oligomerization without resorting to the in-
convenience of infinite dilution techniques. In addition,
the tin-containing solid phase catalyst could be easily
removed following lactonization.
1
8,19
reported methods using tin catalysis.
Copolymers 4B and 4C were employed under the
same conditions described above for 4A. In Table 4 are
the observed yields of lactone and oligomers. Since
copolymers 4B and 4C had reduced loading capacities,
the amount of polymer was increased to keep the ratio
of catalytic sites/5 constant at 0.1. Although this ratio
was kept constant, the yield of lactone decreased with
decreasing loading capacity. The oligomer yield in-
creased and declined again with decreasing loading
capacity. For synthetic purposes, the maximum yield
of the 17-membered lactone product and the highest
overall conversion was observed with the highest load-
ing capacity polymer, 4A. Presumably, the lactone/
oligomer ratio could be improved by lowering the
concentration of 5 below 0.03 M, but this was not
pursued.
The four ω-hydroxycarboxylic acids chosen for this
study were 16-hydroxyhexadecanoic acid 5, 12-hydroxy-
dodecanoic acid, 10-hydroxydecanoic acid, and 7-hy-
droxyheptanoic acid. These starting materials represent
two large and two medium-sized lactones.
Optimized reaction conditions were determined for 5
using the highest loading capacity tin oxide copolymer,
The reaction conditions optimized for 5 were applied
to 12-hydroxydodecanoic acid, 10-hydroxydecanoic acid,
and 7-hydroxyheptanoic acid for the preparation of 13-,
4
A, as catalyst. As summarized in Table 3, the molar
ratio of 4A/5 was systematically changed from 0 to 2.0
in significant intervals. As detailed in the Experimental
Section, lactonization reactions were run at reflux in
mesitylene for 19 h with 0.03 M concentration of 5.
Lower boiling solvents resulted in reduced yields. A
concentration of 0.03 M, near the anticipated effective
molarity20 for this lactonization, was chosen. At this
concentration, rates of lactonization and oligomerization
should be equal. Thus a minor effect of catalyst on
product ratios should be fairly apparent.
1
1-, and eight-membered lactones. The 10-hydroxyde-
canoic acid and 7-hydroxyheptanoic acid produced no
lactone products, but modest yields of oligomers did
result, 49 and 17 wt %, respectively. With 12-hydroxy-
dodecanoic acid, a lactone yield of 10 mol % was
observed with a 78 wt % yield of oligomers being
produced. For each of these hydroxycarboxylic acids, the
concentation of 0.03 M is above the anticipated effective
molarity for these ring sizes, particularly for eight- and
The residue obtained after solvent removal was
separated into two components by trituration with
1
1-membered lactones. Unless the copolymers exerted
a special effect on the lactone/oligomer ratio, little
lactone product was anticipated and little was observed.
Com p etition Stu d ies. In an attempt to determine
the preferred mode of reaction of ω-hydroxycarboxylic
acids with the copolymer 4A, three experiments were
performed. In one instance, equimolar amounts of
phenylacetic acid, benzyl alcohol, and hexa-n-butyldi-
(
18) (a) Steliou, K.; Poupart, M. A. J . Am. Chem. Soc. 1983, 105,
7
130. (b) Steliou, K.; Szcygielska-Nowosielska, A.; Favre, A.; Poupart,
M. A.; Hanessian, S. J . Am. Chem. Soc. 1980, 102, 7578. (c) Otera, J .;
Yano, T.; Himeno, Y.; Nozaki, H. Tetrahedron Lett. 1986, 27, 4501.
(
19) White, J . D.; Green, N. J .; Fleming, F. F. Tetrahedron Lett.
993, 34, 3515.
20) Illuminati, G.; Mandolini, L. Acc. Chem. Res. 1981, 14, 95.
1
(

Polymer-Supported Stannols and Distannoxanes
Organometallics, Vol. 18, No. 26, 1999 5581
stannoxane were refluxed in benzene. A 119Sn NMR
spectrum in benzene-d6 showed peaks for unreacted
hexa-n-butyldistannoxane (89.2 ppm) and for tri-n-
butylstannol (102 ppm) and a peak at 104.5 ppm
attributed to tri-n-butylstannyl phenylacetate. A stan-
nyl alkoxide would be anticipated near 80 ppm and
was not observed. Since tin alkoxides are hydrolytically
unstable, care was taken to limit moisture and air
exposure of the reaction products. This solution phase
result suggests that tin carboxylates form in preference
to tin alkoxides under these conditions.
carbonyl absorptions at 1730 cm for a lactone or ester
-1
-1
but no peak at 1684 cm for a carboxylic acid carbonyl,
as seen in the starting hydroxycarboxylic acids. The
typical OH absorptions were also absent. Both the
1
3
C
1
and H NMR spectra had shifts and splitting patterns
for the -CH2-O- more consistent with lactones than
with hydroxycarboxylic acids.
In the mass spectra, ions resulting from polylactone
fragmentation could not be distinguished with certainty
from secondary ions generated in the spectrophotom-
eter. Consequently, the approximate polylactone mo-
lecular weights were obtained using camphor freezing
point depression. In this way, a Mn of 586 ( 16 g/mol,
or an average of 2.3 lactone units per ring, was obtained
for the oligomeric mixture from 5 and of 911 ( 2 g/mol,
or 4.6 lactone units per ring, for 12-hydroxydodecanoic
acid.
2
1
This competition study was extended to the solid
phase using copolymer 4A, phenylacetic acid, and benzyl
alcohol. The IR spectrum of the recovered polymer
2
2
-1
showed a carbonyl stretch at 1649 cm , consistent with
a tin carboxylate. Finally, equimolar amounts of 5 and
copolymer 4A were refluxed for 1 h in benzene. The IR
spectrum of the recovered solid showed a carbonyl
-
1
stretch at 1637 cm , consistent with a tin carboxylate,
Con clu sion s
-
1
and a broad peak centered at 3320 cm attributed to
an alcohol OH. These findings suggest that a ω-hy-
droxycarboxylic acid reacts selectively through the
carboxyl group rather than the hydroxyl group and that
reaction is fairly complete. How this bound ω-hydroxy-
carboxylate proceeds to lactone and oligomeric lactones
is a matter of conjecture.
Three novel copolymers, 2A-C, of monomer 1, sty-
rene, and divinylbenzene, differing in their loading
capacity, have been prepared and characterized. Hy-
drolytic conversion to tin oxides lead to polymers as
mixtures of stannols and distannoxanes, 4A-C. The
ratio of stannol/distannoxane was dependent on loading
capacity of the copolymers and thus, presumably, on the
Role of th e Ca ta lyst. The results in Tables 3 and 4
suggest a complicated catalytic role for the polymer-
supported tin oxides. The data in Table 3 show that at
high ratios of 4A/5 esterification, be it lactonization or
oligomerization, slows or stops. At this high ratio all of
the hydroxyacid is presumably bound to the polymer.
On the other hand since these polymers do act as
catalysts, some amount of bound hydroxyacid must be
involved. These competing effects of polymer-supported
tin oxides first increasing and then decreasing the
extent of esterification lead to an optimum catalytic
ratio. It is not yet clear why high ratios of 4A/5 lead to
decreased esterification yields.
“
effective” concentration of tin sites. These tin oxides
proved to be viable as lactonization catalysts. While no
beneficial effect on lactone/oligomer ratios was observed,
the solid phase has simplified purification procedures.
Copolymers 2A and 4A were converted to tin car-
boxylates. The stannyl toluate, 3, proved to be acid
sensitive. Competition studies between carboxylic acids
and alcohols gave preferential formation of tin carboxy-
lates over tin alkoxides.
Exp er im en ta l Section
Tri-n-butylchlorostannane and hexa-n-butyldistannoxane
were purchased from Gelest. Styrene (Aldrich, tech.) and
divinylbenzene (Aldrich, tech., 80%) were chromatographed
through a silica gel column to remove stabilizers prior to use.
Solution phase NMR spectra were recorded on a Varian
Gemini XL-300 spectrometer and MAS solid-state NMR
spectra on a Bruker 200 spectrophotometer after swelling
samples in chloroform prior to analysis. All spectra were
A mechanism in which lactonization occurs through
an anchored 16-hydroxyhexadecanoate which bites back
on itself is not supported by the results of either Table
3
or 4. Either high ratios of 4A/5 or lower loading
capacities should lead to more spatially discrete sites.
Spatially discrete sites within the polymer matrix would
presumably favor lactone formation over oligomeriza-
tion. The results of Table 4 show the opposite trend and
are suggestive of a scheme in which both ends of 5 are
activated by binding to tin sites. Thus, higher loading
capacity improves the opportunity for both activated
ends to meet and result in lactone formation. Such a
double activation scenario has been previously proposed
1
13
referenced against tetramethylsilane ( H, C) or tetrameth-
ylstannane ( Sn). IR spectra were obtained from neat pow-
119
ders on a Bruker ISS55 FT-IR using a diffuse reflectance
infrared FT spectroscopy (DRIFTS) accessory manufactured
by Spectratech. EDX analyses were performed on an EDAX-
CDU leap detector instrument.
HPLC analyses to quantify lactonization yields employed a
Hewlett-Packard 1047A refractive index detector, SSI 300 LC
pump, Varian 905320 pulse damper, Waters 746 data module,
and Rheodyne injector. Eluent composition, prepared with
spectrophotometric grade solvents, varied depending on the
analysis. All HPLC samples were injected manually onto a
Waters Bondapak C18 10 µm reverse phase column. Oligo-
meric product yields (wt %) were quantified by gravimetry.
Lactone yields (mol %) were determined by comparison to
for solution reactions involving di-n-butyltin oxide
catalysis.18a,b
Oligom er ic Byp r od u cts. The methanol-insoluble
oligomeric byproducts isolated by trituration were char-
acterized by mass spectrometry and 13C, H NMR and
FTIR spectroscopy. All of these anaylyses pointed to
polylactones rather than polyesters. Both EI- and CI-
MS showed peaks for monomeric through pentameric
lactones. The relative proportions of these peaks changed
with the lactone ring size targeted. IR spectra gave
1
2
3
calibration curves for each of the authentic lactones prepared
(
22) Atkins, P. W. Physical Chemistry, 4th ed.; Freeman and
Company: New York, 1990.
23) Meyer, W. L.; Patterson, W. T.; Reed, S. A.; Leister, M. C.;
(
(
21) Smith, P. J .; White, R. F. M.; Smith, L. J . Organomet. Chem.
Schneider, H. J .; Schmidt, G.; Evans, F. E.; Levine, R. A. J . Org. Chem.
1992, 57, 291.
1
972, 40, 341.

5
582 Organometallics, Vol. 18, No. 26, 1999
Hunter and McRoberts
from commercially available cycloketone precursors. Synthesis
of the authentic lactones was confirmed by high-resolution
mass spectrometry and comparison to NMR spectral data
available from the literature.
4B: MAS solid-state 119Sn NMR spectrum: 100.8 (Sn-OH),
-1
92.2 (Sn-O-Sn), 1:1 relative peak intensity. IR (cm ): 3026,
3110 (CHar), 2927, 2850 (CHaliph), 1594, 1490 (Cvinyl), 789, 754,
705.
4C: MAS solid-state 119Sn NMR spectrum: 100.6 (Sn-OH),
P r ep a r a tion of 3- a n d 4-(2-Di-n -bu tylch lor osta n n yl)-
eth yl)styr en e (Mon om er 1). This polymer was prepared
-1
92.0 (Sn-O-Sn), 4:1 relative peak intensity. IR (cm ): 3060,
5
3026 (CH ), 2927, 2850 (CHaliph), 1594, 1488 (Cvinyl), 1453, 754,
following a method similar to Neumann. Di-n-butylstannane
ar
was prepared by lithium aluminum hydride reduction of
dibutyldichorostannane. Equimolar di-n-butylstannane and di-
n-butyldichlorostannane were combined to produce di-n-bu-
tylchlorostannane in situ, and this was reacted with equimolar
divinylbenzene and AIBN, yielding monomer 1. At reaction
completion, an aliquot of the crude reaction mixture was
removed for spectroscopic analysis. No attempt was made to
705.
P r ep a r a tion of P oly-3- a n d P oly-4-(2-(d i-n -bu tyl(sta n -
n yl-4-tolu a te)eth yl)styr en e-co-d ivin ylben zen e (3). Poly-
mer 2A (760 mg, 1.29 mmol) and sodium toluate (203 mg, 1.28
mmol) were refluxed in benzene (15 mL) for 16 h. The polymer
was then isolated by filtration with a sintered glass filter,
rinsed with fresh benzene, and vacuum-dried. Solid-state MAS
1
19
-1
purify this monomer, which was used immediately in subse-
Sn NMR spectrum: 104.0 (Sn-O-CdO). IR (cm ): 3019
1
quent polymerizations. H NMR spectrum (benzene-d
6
): 7.5-
(CHar), 2963, 2927, 2857 (CHaliph), 1639 (CdO), 1601, 1540
(Cvinyl), 1350, 750, 700.
7
5
.1 (m, 4H, CHar), 6.75 (m, 1H, CHvinyl), 5.8 (m, 1H, CHvinyl),
.25 (m, 1H, CHvinyl), 3.0 (m, 2H, CH ), 1.9-0.9 (m, 20H, Sn-
and n-Bu). Ethylvinylbenzene was also visible as ap-
Typ ica l La cton iza tion of ω-Hyd r oxyca r boxylic Acid s
Usin g Ca ta lytic Or ga n otin Oxid e Cop olym er s 4A-C. To
a flask fitted with a Barrett water trap and a condenser was
added 16-hydroxyhexadecanoic acid (345 mg, 1.27 mmol),
organostannol/distannoxane copolymer 4A (1.7 mmol/g, 87.7
mg, 0.13 mmol), and mesitylene (45 mL). After 19 h reflux with
stirring under an argon atmosphere, the reaction solution was
filtered. The solid was washed with fresh mesitylene and
chloroform, and the filtrates were concentrated in a vacuum
to produce a solid residue.
2
CH
proximately a 22% impurity. C NMR spectrum (benzene-
2
1
3
d
1
1
6
): 136.7, 136.5 (CH, Cvinyl m+p), 128.9, 128.1, 128.0, 127.3,
26.5, 125.6, 124.3 (CH and C, CAR m+p+ethylvinylbenzene),
14.1, 113.9 (CH, CHvinyl), 113.1 (CH, CHvinyl,), 31.5 (CH, CH
2
-
-
Ph), 27.7 (CH, nBu), 26.8 (CH, nBu), 19.8, 19.6 (CH, CH
2
1
19
Sn), 17.7 (CH, nBu), 13.6 (CH, nBu).
benzene-d ): +143 (Sn-Cl).
P r ep a r a tion of P oly-3- a n d P oly-4-(2-(d i-n -bu tylch lo-
r osta n n yl)eth yl)styr en e-co-d ivin ylben zen e Cop olym er s
2A, 2B, a n d 2C). The copolymers were prepared in a resin
Sn NMR spectrum
(
6
The resultant residue was triturated with methanol, fol-
lowed by centrifugation to give solids, which were character-
ized as detailed below. The solution, containing primarily
hexadecanolide, was diluted to a known volume for HPLC
(
kettle from the amounts of monomer 1, styrene, and divinyl-
benzene shown in Table 1. Each polymerization was carried
out at reflux with stirring (∼1000 rpm) in the presence of
octanol (70 mL), aqueous methylcellulose solution (0.25%, 250
mL), and azobisisobutyronitrile (1.25 g). After 7.5 h, the
reaction mixture was cooled and the resultant solid was
isolated by decantation. The solid was washed with water until
the supernatant ran clear, and similarly with methanol and
acetone. Drying at 70 °C at about 1 Torr yielded white granular
solids with yields in the range 50-73%. The spectroscopic
characterization of each copolymer follows.
2
analysis. HPLC conditions: 85% methanol, 15% H O (1 mL
-
5
H
4
3
PO
4
/1 L), pH 3.0, flow rate 1 mL/min, 32 × 10 RIU/FS,
0 °C. Calibration curves were constructed for both hexade-
canolide (retention time ) 16.5 min) and 16-hydroxyhexade-
canoic acid (retention time ) 7.6 min). The hexadecanolide
identity was confirmed by co-injection with authentic hexa-
decanolide.
Ch a r a cter iza tion of Solid Byp r od u cts fr om La cton -
iza tion . The solid remaining after trituration of the 16-
hydroxyhexadecanoic acid lactonization was dried at about 1
A: MAS solid-state 13C NMR spectrum (swollen in
CHCl ): 144, 128 (C+CH, CAR), 42 (very broad, CH-Ph, CH
Ph, CH backbone), 28.6 (CH, nBu), 27.7 (CH, nBu), 18.6 (CH,
nBu), 14.6 (CH, nBu). MAS solid-state Sn NMR spectrum
2
1
Torr at room temperature and analyzed by H NMR spectros-
3
2
-
copy (CDCl
3
): 4.1 (m, 2H, CH
2 2
), 2.3 (m, 2H, CH ), 1.6 (m, 4H,
2
-1
1
19
CH ), 1.4-1.2 (m, 22H, CH
2
2
). IR (cm ): 2942, 2857 (aliphatic
-
1
CH), 1728 (ester CdO). LR-MS gave peaks at 255, 490, 744,
and 998, which coincide with monomer through tetrameric
polylactones. Camphor freezing point depression was used to
determine a number average molecular weight for the mixture
of polylactones. Hexadecanolide was used to prepare a stan-
dard calibration curve. The solid isolated after trituration gave
a number average molecular weight of the polylactone products
at 586 ( 16 g/mol.
(
(
swollen in CHCl
3
): +148 (Sn-Cl). IR (cm ): 3060, 3022
CHar), 2960, 2925, 2856 (CHaliph), 1602, 1500 (Cvinyl), 1451, 707.
EDX atom %: Sn 2.32, Cl 2.10.
1
19
2
B: MAS solid-state
3
Sn NMR spectrum (swollen in
-1
CDCl
): +149.8 ppm. IR (cm ): 3060, 3026 (CHar), 2927, 2857
(
7
CHaliph), 1940-1800 (aromatic overtones), 1601, 1495 (Cvinyl),
96, 705, 535. EDX atom %: Sn 0.94, Cl 0.84.
1
19
2
C: MAS solid-state
3
Sn NMR spectrum (swollen in
In a similar manner the solid byproducts from the lacton-
-1
CDCl ): +149.3 ppm. IR (cm ): 3060, 3026 (CHar), 2927, 2857
ization of 12-hydroxydodecanoic acid were dried in vacuo and
(
7
CHaliph), 1940-1800 (aromatic overtones), 1601, 1495 (Cvinyl),
1
analyzed by H NMR spectroscopy (CDCl
3
): 4.15 (m, 2H, CH
2
),
).
96, 705, 535. EDX atom %: Sn 0.37, Cl 0.32.
2
.35 (m, 2H, CH
2
), 1.65 (m, 4H, CH
2
), 1.5-1.25 (m, 14H, CH
2
P r ep a r a tion of P oly-3- a n d P oly-4-(2-(d i-n -bu tylsta n -
LR-MS indicated peaks at 199, 397, 595, 793, and 989,
consistent with monomeric through pentameric lactones.
Camphor freezing point depression using cyclododecanone as
a standard gave an average molecular weight of 911 ( 2 g/mol.
Rea ction of Ben zyl Alcoh ol a n d P h en yla cetic Acid
w ith Hexa -n -bu tyld ista n n oxa n e. To a flask fitted with a
Barrett water trap and a condenser was added benzyl alcohol
(80.4 mg, 0.74 mmol), phenylacetic acid (103.1 mg, 0.76 mmol),
benzene (25 mL), and hexa-n-butyldistannoxane (451.0 mg,
n ol/tetr a -n -bu tyld ista n n oxa n e)eth yl)styr en e-co-d ivin yl-
ben zen e-co-styr en e Cop olym er s (4A, 4B, a n d 4C). Each
of copolymers 2A-C (1.0 g) and sodium hydroxide solution
(50% in ethanol) was shaken for 16 h and then filtered through
a coarse sintered glass filter. The filter cake was washed
liberally with additional water and ethanol. The solid was
dried at about 1 Torr overnight at 70 °C. All hydrolysis filtrates
1
0
were collected and set aside for Mohr analysis, and the
results are included in Table 2. The characterization of each
copolymer follows.
0.76 mmol). After 16 h reflux with stirring under an argon
atmosphere, a reaction aliquot was removed and a 119Sn NMR
3
A: MAS solid-state 119Sn NMR spectrum: 103 (Sn-OH),
6
spectrum (benzene-d ) was obtained: 89.2 (Sn-O-Sn), 102
(Sn-OH), 104.5 (Sn-O-C(O)R)).
Rea ction of Ben zyl Alcoh ol a n d P h en yla cetic Acid
w ith Cop olym er 4A. To a flask fitted with a Barrett water
-
1
9
3
1
1 (Sn-O-Sn), 2:3 relative peak intensity. IR (cm ): 3026,
010 (CHar), 2927, 2850 (CHaliph), 1601, 1492 (Cvinyl), 1440,
360, 755, 703.

Polymer-Supported Stannols and Distannoxanes
Organometallics, Vol. 18, No. 26, 1999 5583
trap and a condenser was added benzyl alcohol (23 µL, 0.22
mmol), phenylacetic acid (30.4 mg, 0.22 mmol), benzene (15
mL), and copolymer 4A (1.67 mmol/g, 298 mg, 0.50 mmol).
After 2 h reflux with stirring under an argon atmosphere, the
polymer was removed by filtration under argon and rinsed
with fresh benzene. DRIFTS analysis was performed attempt-
ing to keep air exposure at a minimum. IR (cm-1): 3088, 3020
minimum for both samples. IR (cm-1) 1 h aliquot: 3320
(alcohol), 3020 (CHar), 2929, 2856 (CHaliph), 1637 (Sn-O-C(O)-
R), 1604, 1557 (Cvinyl), 1062, 794, 707. The 4 h aliquot gave
essentially the same IR spectrum.
Ack n ow led gm en t. The authors wish to give par-
ticular thanks to Dr. Gian Gobbi, Dow Chemical Inc.,
Sarnia, Canada, for his generous and unfailing as-
sistance in running the MAS solid-state NMR spectra.
Mary J ane Walzak and Ross Davidson of Surface
Science Western are to be thanked for their help with
DRIFTS and EDX spectroscopies. Also we wish to thank
NSERC, Canada, and Merck-Frosst, Canada, for their
support of our research through an IOR grant.
(
1
CHar), 2958, 2917, 2862 (CHaliph), 1649 (Sn-O-C(O)R), 1620,
580 (Cvinyl), 1066, 792, 703.
Rea ction of Cop olym er 4A w ith 5. To a flask fitted with
a Barrett water trap and a condenser was added 5 (59 mg,
.22 mmol), polymer 4A (1.8 mmol/g, 123 mg, 0.22 mmol), and
0
benzene (15 mL). After 1 h at reflux, an aliquot was removed
and filtered under an argon atmosphere, rinsing with fresh
benzene. An additional aliquot was removed at 4 h. DRIFTS
analysis was performed attempting to keep air exposure at a
OM990630D