Triorganotin 4-isopropylbenzoates as model transesterification catalysts for triorganotin carboxylates grafted to cross-linked polystyrene

Article abstract of DOI:10.1016/j.jorganchem.2006.01.002

Starting from 4-isopropylbenzoic acid, three new triorganotin carboxylates bearing methyl, butyl and phenyl substituents at tin, respectively, were prepared and fully characterized by spectroscopic and thermal techniques, with particular regard to the coordination number of tin atom, in solution as well as in the solid state. The triorganotin compounds, tested as transesterification catalysts in the reaction between ethyl acetate and primary, secondary or tertiary alcohol, respectively, displayed, as expected, a strong decrease of activity on passing from the primary to the tertiary alcohol reactant. Different activities by the tin carboxylates were also observed in the reaction between primary alcohol and ethyl acetate. The reaction mechanism, as elucidated by Sn NMR, involves coordination of both ester substrate and alcohol reactant to the triorganotin compound, the reaction conversion appearing related not only to the Lewis acidity of the tin atom, but also to the nature of the reactants. Preliminary catalytic tests were also carried out in the reaction between glyceryl tridodecanoate (as a model of natural triglyceride) and ethanol, mimicking the preparation of biodiesel fuel. Although in this case lower conversions were obtained with respect to the reactions on ethyl acetate, the catalytic activity of organotin derivatives appears considerable.

Full text of DOI:10.1016/j.jorganchem.2006.01.002

0
Journal of Organometallic Chemistry 691 (2006) 1965–1972
www.elsevier.com/locate/jorganchem
Triorganotin 4-isopropylbenzoates as model transesterification
catalysts for triorganotin carboxylates grafted to
cross-linked polystyrene
a,
a
a
a
Luigi Angiolini *, Daniele Caretti , Laura Mazzocchetti , Elisabetta Salatelli ,
b
b
Rudolph Willem , Monique Biesemans
a
Dipartimento di Chimica Industriale e dei Materiali, University of Bologna, Viale Risorgimento 4, I-40136 Bologna, Italy
b
High Resolution NMR Centre (HNMR), Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussel, Belgium
Received 24 October 2005; received in revised form 29 December 2005; accepted 3 January 2006
Available online 14 February 2006
Abstract
Starting from 4-isopropylbenzoic acid, three new triorganotin carboxylates bearing methyl, butyl and phenyl substituents at tin,
respectively, were prepared and fully characterized by spectroscopic and thermal techniques, with particular regard to the coordination
number of tin atom, in solution as well as in the solid state. The triorganotin compounds, tested as transesterification catalysts in the
reaction between ethyl acetate and primary, secondary or tertiary alcohol, respectively, displayed, as expected, a strong decrease of activ-
ity on passing from the primary to the tertiary alcohol reactant. Different activities by the tin carboxylates were also observed in the
reaction between primary alcohol and ethyl acetate. The reaction mechanism, as elucidated by Sn NMR, involves coordination of both
ester substrate and alcohol reactant to the triorganotin compound, the reaction conversion appearing related not only to the Lewis acid-
ity of the tin atom, but also to the nature of the reactants. Preliminary catalytic tests were also carried out in the reaction between glyceryl
tridodecanoate (as a model of natural triglyceride) and ethanol, mimicking the preparation of biodiesel fuel. Although in this case lower
conversions were obtained with respect to the reactions on ethyl acetate, the catalytic activity of organotin derivatives appears
considerable.
ꢀ 2006 Elsevier B.V. All rights reserved.
Keywords: Carboxylate ligands; IR spectroscopy; NMR spectroscopy; Organotin; Transesterification
1. Introduction
Their actual use remains however limited by their toxic-
ity, particularly when they can be released into the environ-
ment, or when they are retained in the final product as
catalyst residues [10,11].
The pollution problems raising from the use of organo-
tins can be conveniently addressed by covalently grafting
the organometallic moiety to a macromolecular support
[12], thus lowering or even preventing tin release into the
environment [13], leading to the so-called ‘‘clean organotin
derivatives’’ [14–19]. Tin compounds grafted to polymeric
carbon chains experience growing interest since alkyltin
moieties in polymers or copolymers mimic well the proper-
ties of their non-polymeric organotin analogues.
Since their first biocidal application in the early 1920s
[1], the production of organotin compounds [2] has
increased for various purposes mainly in the last decades,
covering a wide variety of applications, being used in anti-
fouling paints [3–5], PVC stabilization [6], as homogeneous
catalysts [7], in antitumour therapy [8] and as ion carriers in
electrochemical membranes design [9].
*
Corresponding author. Tel./fax: +39 051 2093687.
E-mail address: luigi.angiolini@unibo.it (L. Angiolini).
0022-328X/$ - see front matter ꢀ 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2006.01.002

1966
L. Angiolini et al. / Journal of Organometallic Chemistry 691 (2006) 1965–1972
In this context, the preparation of non-polymeric model
compounds may conveniently parallel the synthesis of
related polymeric derivatives when is aimed to reproduce
the repeating unit of the polymer, thus allowing to assess
the influence of the macromolecular chain on the behav-
iour of the organometallic moiety.
x
y
z
In particular, when the insolubility of cross-linked poly-
meric derivatives prevents the application of many common
analytical techniques, non-polymeric model compounds,
more amenable to full characterization, offer a key support
to the assessment of features which would be otherwise
undetectable by direct observation.
y
O
OSnR3
R= Me, Bu, Ph
Chart 2. General structure of cross-linked polymers.
reaction for the biodiesel production by transesterification
of renewable materials such as vegetable oils [21–23] is
presented.
In the course of our current investigations dealing with
the synthesis and characterization of cross-linked systems
bearing triorganotin carboxylate residues [20], intended as
grafted transesterification catalysts active at the liquid–solid
interface, we have prepared the trimethyltin (2), the tribu-
tyltin (3) and the triphenyltin (4) esters of 4-isopropylben-
zoic acid (1, Chart 1) as structural models for the related
styrenic repeating unit of stannylated cross-linked polymers
possessing the general structure depicted in Chart 2.
The present paper is aimed therefore at investigating
structure and catalytic activity of model compounds 2–4,
in order to gain a preliminar insight into the properties
of the related cross-linked derivatives, which are presently
under investigation. Particular attention is focused not
only onto the tin atom coordination expansion which is
believed to play a key role in the catalytic performances,
but also onto the evaluation of the Lewis acidity of the dif-
ferently substituted organotin derivatives, as an useful indi-
cator of their catalytic potential in transesterifications
involving different hindered alcohols. In addition, a model
2. Results and discussion
2.1. Synthesis and structural characterization
Direct triorganostannylation of a carboxylic acid with
the appropriate bis(triorganotin) oxide is one of the most
commonly used methods [24] to obtain triorganotin carb-
oxylates, along with reaction of the carboxylic acid with
the suitable triorganotin hydroxide [25]. The water pro-
duced in both cases as a by-product can be conveniently
removed from the reaction mixture by azeotropic distilla-
tion with toluene, in order to push the reaction to
completion.
As represented in Chart 1, we have applied this well
established synthetic pathway to synthesize both the tribu-
tyltin (3) and triphenyltin (4) carboxylic esters, starting
from commercial 4-isopropylbenzoic acid (1), which was
allowed to react with bis(tributyltin)oxide and triphenyltin
hydroxide, respectively. While in the latter case separation
of the reaction product from unreacted stannylating
reagent was straightforward, the isolation of 3 from unre-
acted tin hydroxide required the treatment of the crude
reaction mixture with a weak ionic exchange resin (Fluka
Amberlite IRC-86), which is able to entrap the organome-
tallic reagent present in excess [26], thus allowing to obtain
pure 3.
The synthesis of the trimethyltin ester 2 from 4-isopro-
pylbenzoic acid (1) was achieved by direct esterification
with trimethyl tin hydroxide, but using in this case acetone
as the reaction solvent in order to operate at a refluxing
temperature low enough to prevent undesired sublimation
of the reacting trimethyltin hydroxide during the reaction.
The water produced was separated in this case by 5A
molecular sieves added to the reaction mixture and the final
elimination of unreacted reagent was simply achieved by
sublimation under reduced pressure and gentle heating.
Model compounds 2–4 were all characterized by FT-IR
as well as NMR spectroscopy.
Me3SnOH
Acetone
O
O
O
OSnMe3
OSnBu3
OSnPh3
2
(Bu3Sn)2O
Toluene
O
OH
3
1
Ph3SnOH
Toluene
The infrared asymmetric stretching vibration of the
newly formed triorganotin carboxylate group appears in
the range 1630–1635 cmÀ1 for both 3 and 4, while the sym-
4
Chart 1. Synthesis of compounds 2–4.

L. Angiolini et al. / Journal of Organometallic Chemistry 691 (2006) 1965–1972
1967
troscopy, using in particular two-dimensional correlation
techniques.
The 1H NMR resonances were fully assigned on the
basis of COSY and HSQC experiments, since straightfor-
ward assignment of the aromatic proton resonances, based
1
solely on standard H NMR, failed. This was particularly
the case for 4, due to the presence of overlapping signals
of the protons belonging to different aromatic rings.
The 119Sn NMR spectra of 2–4 in dilute CDCl3 solution
(Table 1) show the same trend as the related styryl mono-
mers [21], the actual chemical shift values of model com-
pounds 2–4 being only slightly higher than those of the
monomers, probably due to lower p conjugation. As
expected in solution, all the observed resonances are typical
of tetracoordinated tin atom [31].
Solid-state 117Sn MAS NMR spectra could be obtained
only for 2 and 4 (Table 1), given the oily state of the trib-
utyltin derivative 3. The observed resonances are in line
with the FT-IR data, displaying d(117Sn) = À110 ppm for
4, a value typical of tetracoordinated triphenyltin moiety,
similar to the value reported in solution [32]. By contrast,
the trimethyltin derivative 2 shows a single isotropic chem-
ical shift at 26 ppm, which can be assigned to pentacoordi-
nated tin when the coordination with the carbonyl oxygen
occurs intramolecularly [31].
Fig. 1. FT-IR spectra of 2 (a), 3 (b) and 4 (c).
The model compounds 2–4 were also investigated by
DSC in order to get a deeper insight into their thermal
properties, given the unusual tin tetracoordination in 3 in
the solid state, as revealed by FT-IR.
metric one lies around 1330 cmÀ1 (Fig. 1b and c), these
bands being related to a carbonyl moiety free from any
interaction with the metal atom [27], therefore evidencing
tetracoordinated tin in 3 and 4 in the bulk.
This behaviour, well established for analogue non-poly-
meric compounds bearing the triphenyltin carboxylate
moiety, is uncommon for tributyltin derivatives. These last
compounds are in fact usually pentacoordinated [28,29], at
least when not hindered by the presence of a macromolec-
ular backbone, as demonstrated by X-ray crystallography
[30].
When submitted to heating/cooling cycles in the temper-
ature range 0–160 ꢁC (Table 2), 2 shows an endothermic
peak in the first heating, the enthalpy of which is similar
to the one of the related exothermic process occurring dur-
ing the cooling step. These thermal transitions can be
referred to melting and subsequent crystallization of crys-
talline domains induced by pentacoordinated tin, as previ-
ously reported [26,33]. The same pattern is present in a
second heating cycle, with melting enthalpy close to the
values found in the first one, evidencing the complete
reversibility of the transition.
By contrast, the triphenyltin derivative 4 does not dis-
play any thermal transition after the melting observed in
the first heating step, thus confirming the absence of rele-
vant crystalline domains formation, at least in the time
scale of the DSC measurements.
The unusual non-coordinated arrangement of the tin
atom in 3 can tentatively be attributed to the concomitant
contribution of both the conformational rigidity of the
molecule due to p conjugation, inducing restricted rotation
around the phenyl ring–carboxylate bond, and the steric
hindrance of the butyl groups, apparently preventing the
carbonyl oxygen from interaction with tin. The steric hin-
drance of the Me3Sn moiety in 2 being lower, such a restric-
tion to tin coordination by the carbonyl oxygen atom is no
longer holding, as demonstrated by the carbonyl absorp-
tions in the FT-IR spectrum of 2 (Fig. 1-a), with asymmet-
The tributyltin derivative 3 when submitted to the same
heating/cooling cycles in the range 0–160 ꢁC as 2 and 4,
does not display any thermal transitions at all. However,
ric and symmetric stretching bands at 1595 and 1371 cmÀ1
,
Table 1
respectively, indicative of pentacoordination [30] at tin.
The frequency values found for the triorganotin derivatives
2–4 parallel those of the related triorganotin p-vinyl benzo-
ate monomeric derivatives [21], displaying a strictly similar
coordination behaviour at tin.
119Sn and 117Sn NMR chemical shifts (ppm) of 2–4 in dilute (1%) CDCl3
solution and in the solid state, respectively
Sample
Solution
Solid state
2
3
4
133
109
26
–
With the aim of assessing their structure, products 2–4
1
À115
À110
have been investigated by H, 13C and 119Sn NMR spec-

1968
L. Angiolini et al. / Journal of Organometallic Chemistry 691 (2006) 1965–1972
Table 2
Thermal behaviour of 2–4
The question raises as to whether, and, if so, how, the
observed increase of catalytic activity from compounds 2
to 4 can be correlated to their Lewis acidity.
Sample
First heating step
First cooling step
Second heating
step
In order to address this issue, 119Sn NMR spectra were
recorded for compounds 2–4 in the presence of the actual
reagents ethyl acetate and 1-octanol. For the sake of com-
parison in the assessment of the degree of tin complexation,
acetone-d6, likewise containing a carbonyl but no alkoxy
function as in the ester, was first used as a reference solvent,
so as to enable us to find out whether complexation occurs
in the ester predominantly by its carbonyl or alkoxy func-
tion. As seen in Table 4, the 119Sn resonances of 2–4 in ace-
tone-d6 are shifted to lower frequency, as expected for such
a coordinating solvent. The trimethyltin derivative 2 dis-
plays the highest low frequency shift, and 4 the smallest
one, which suggests a complexation stability toward tin
by acetone decreasing from 2 to 4. The smallest low fre-
quency shift observed for the triphenyltin derivative 4 is
somewhat surprising since the latter compound is generally
assumed to have higher Lewis acidity [9,32,34].
a
b
a
Tm
(ꢁC)
DHm
Tc
(ꢁC)
DHc
Tm
(ꢁC)
DHm
(kJ molÀ1
)
(kJ molÀ1
)
(kJ molÀ1
)
2
147
–
109
32.6
–
23.5
108
À24
–
26.7
16.6
–
143
25
–
26.0
18.2
–
3c
4
a
Melting temperature at maximum endothermic peak height.
Crystallization temperature at maximum exothermic peak height.
Temperature range from À60 to 160 ꢁC.
b
c
when the sample is cooled to À60 ꢁC (starting temperature
of the second cycle), an exothermic peak appears at
À24 ꢁC, the enthalpy of which is similar to that one mea-
sured for the endothermic transition at around 25 ꢁC
observed in the subsequent heating step. This observation
explains the oily state of the product at room temperature,
in addition to confirm its scarce tendency to give crystalline
aggregates.
The 117Sn NMR spectra of compounds 2–4 in the pres-
ence of the transesterification reagents, either ethyl acetate
or 1-octanol or mixtures of both (with a small amount of
CDCl3 as a deuterated lock solvent), shed some light into
these observations (Table 5). In the presence of ethyl ace-
tate alone, the low frequency shifts are only slightly smaller
than in acetone-d6, indicating that complex formation with
an ester gives rise to only a slightly smaller shielding at the
tin nucleus than acetone, in line with the well-known lower
nucleophilicity of the oxygen atom of an ester carbonyl
function with respect to that of a ketone; however the pres-
ence of CDCl3 in ethyl acetate, by contrast with pure ace-
tone-d6, can also contribute to the lower shielding at the
117Sn nucleus. Upon addition of pure 1-octanol a more
pronounced low frequency shift is observed, thus suggest-
ing a stronger complex formation with the alcohol than
with the ester.
2.2. Assessment of the catalytic activity in transesterification
reactions
The catalytic activity of the triorganotin carboxylates 2–
4 was investigated in a transesterification reaction on ethyl
acetate with differently hindered alcohols, namely 1-octa-
nol, cyclohexanol and 3-ethyl-3-pentanol, in order to assess
the influence on the catalytic performance of the accessibil-
ity to tin by the alcohol and ester carbonyl functions [19].
The amount of added catalyst was measured so as to
achieve a 1% tin-to-alcohol molar ratio in all experiments.
No attempt was made to remove the added catalyst from
the reaction mixture, the reaction being performed in
homogeneous phase for 48 h. All the reactions were also
carried out in the absence of catalyst, no transesterification
at all being detected, as required.
This is in contrast with earlier results on ring opening
polymerisation of e-caprolactone, initiated by n-propanol
and catalysed by dibutyltindichloride, where e-caprolac-
tone displays stronger complexation toward tin than n-pro-
panol [35]. No straightforward explanation of this finding
can be presently proposed, except the observation that
both the catalyst and the reactants in the transesterification
here reported are notably different from those investigated
in that report. In all cases, adding an amount of n-octanol
to the triorganotin carboxylates 2–4 in ethyl acetate,
induces an additional low frequency shift, implying that
The alcohol conversion in each catalysed transesterifica-
tion reaction (Table 3) was determined by GC after evalu-
ation of the response factor of each species present.
The most effective catalyst appears to be 4, giving the
highest conversion after 24 h reaction of ethyl acetate with
1-octanol. The conversion is only slightly lower with 3, and
markedly lower with 2.
Table 3
Alcohol conversions obtained in the catalysed transesterification of ethyl
acetate
Table 4
119Sn NMR chemical shifts (ppm) in CDCl3 and acetone-d6 for
compounds 2–4
Catalyst
Alcohol conversion (%)
1-Octanol
Cyclohexanol
3-Ethyl-3-
pentanol
Sample
d(119Sn)
D d(119Sn)
24 h
48 h
24 h
48 h
24 h
48 h
CDCl3
Acetone-d6
Acetone-d6 vs. CDCl3
2
3
4
19
76
95
43
95
100
5
7
10
6
9
17
–
–
–
Trace
Trace
Trace
2
3
4
133
109
À115
90
84
À134
À43
À25
À19

L. Angiolini et al. / Journal of Organometallic Chemistry 691 (2006) 1965–1972
1969
Table 5
117Sn Chemical shifts in solution of compounds 2–4 in the presence of the transesterification reactants, ethyl acetate (EtOAc) and 1-octanol (OctOH)
Solvent
2 [Sn(Me)3]
d(117Sn)
3 [Sn(Bu)3]
d(117Sn)
4 [Sn(Ph)3]
d (117Sn)
Dda
Dda
Dda
CDCl3
133.0
94.9
77.7
65.6
57.9
52.4
49.2
34.8
36.3
40.8
43.9
46.6
46.7
109.3
87.5
81.7
76.9
72.8
69.6
66.9
73.2
66.5
66.1
66.5
67.8
67.4
À114.9
À126.6
À130.4
À133.7
À136.2
À138.7
À139.3
À145.2
À149.3
À147.3
À145.7
À143.5
À143.1
100 ll CDCl3 + 400 ll EtOAc
À38.1
À17.2
À12.1
À7.7
À21.8
À5.8
À4.8
À4.1
À3.2
À20.6c
À36.1
À6.7
À0.4
+0.4
+1.3
À5.8c
À11.7
À3.8
À3.3
À2.5
À2.5
À12.7c
À30.3
À4.1
+2.0
+1.6
+2.2
+2.1c
100 ll CDCl3 + 400 ll EtOAc + 100 ll OctOH
100 ll CDCl3 + 400 ll EtOAc + 200 ll OctOH
100 ll CDCl3 + 400 ll EtOAc + 300 ll OctOH
100 ll CDCl3 + 400 ll EtOAc + 400 ll OctOH
100 ll CDCl3 + 400 ll EtOAc + 400 ll OctOHb
100 ll CDCl3 + 400 ll OctOH
À5.5
À45.7c
À98.2
+1.5
+4.5
+3.1
100 ll CDCl3 + 400 ll OctOH + 100 ll EtOAc
100 ll CDCl3 + 400 ll OctOH + 200 ll EtOAc
100 ll CDCl3 + 400 ll OctOH + 300 ll EtOAc
100 ll CDCl3 + 400 ll OctOH + 400 ll EtOAc
100 ll CDCl3 + 400 ll OctOH + 400 ll EtOAcb
a
+2.7
+11.9c
Difference in chemical shift relative to previous situation.
After two weeks.
Dd induced by total amount (400 ll) of OctOH or EtOAc.
b
c
the tin atom coordinates both reagents, and that ethyl ace-
tate does not bind strongly enough to tin to prevent n-oct-
anol from complexing likewise to tin, both reagents being
brought in this way into a favourable position for the
transesterification by a straightforward proximity effect.
Table 5 indicates indeed not only that 1-octanol displays
a stronger 117Sn low frequency shift than ethyl acetate for
all three triorganotin carboxylates 2–4, but also that addi-
tion of further aliquots of 1-octanol to the ethyl acetate
solution of the triorganotin carboxylate enhances the low
frequency shift. This happens however to an increasingly
limited extent, as evidenced by the increasingly smaller
low frequency shift increments upon increasing the number
of added 1-octanol aliquots. By contrast, the stronger com-
plexation ability of 1-octanol is in turn supported by the
finding that progressive additions of increasing aliquots
of ethyl acetate to the 1-octanol solution does not enhance
the shielding at the 117Sn nucleus to a significant extent,
deshielding tending even to take place in the case of 2,
while for 3 and 4 an approximately break even situation
is obtained. This now suggests not only that 1-octanol is
interacting sufficiently strongly with tin to be able to pre-
vent significant further complexation of the tin atom by
ethyl acetate, but also, somewhat paradoxically, that the
latter, possibly, slightly destabilizes the 1-octanol–tin inter-
action. Overall, these findings strongly suggest that the
Lewis acidity of the tin atom toward both ethyl acetate
and 1-octanol decreases also in that order from 2 to 4. In
turn this means that, while the joined interaction of both
the alcohol and ester with tin is a necessary condition for
the catalytic activity, too strong a thermodynamic stability
in this interaction, as observed for 2, obviously disfavours
the catalysis kinetics, compound 2 being precisely the one
with the significantly lowest catalytic activity. Stated in
other terms, the explanation for the transesterification pro-
ceeding better with the triphenyltin compound, where coor-
dination of both the ester and alcohol is obviously weaker,
is that although complex formation is indeed necessary, the
complex has to be sufficiently labile in order to achieve fast
exchange between incoming reactants and outgoing prod-
ucts. This conclusion was also reached in a study relating
electrochemical chloride anion response to the structure
of different tributyltin and triphenyltin carboxylates [9].
In this study, it was shown, by combining potentiometric
anionic response data with 119Sn NMR titrations and
AM1 calculations, that the observed potentiometric
response is related to the binding constant, but that carriers
with too high binding constants show lower potentiometric
response. In contrast to the present investigation, the tri-
phenyltin carboxylates in the electrochemical response
investigation showed a much higher affinity for the chloride
anion than the corresponding tributyltin compounds,
which was explained by their higher Lewis acidity. Thus
Lewis acidity is not an intrinsic property, but also depends
on the nature of the Lewis base with which the tin atom
interacts, as well as on the chemical conditions, e.g., the
solvent. In conclusion, toward 1-octanol as well as to ethyl
acetate, the Lewis acidity of the tin atom decreases from 2
to 4, a more labile Lewis base–acid interaction obviously
favouring the catalysis kinetics.
Upon increasing the steric hindrance of the alcohol,
switching to cyclohexanol, the achieved final conversion
of the reactant (Table 3) is dramatically reduced, probably
because of the lower reactivity of the alcohol due to the
higher steric demand of the intermediate, the molecular
mobility of which is lowered by the stiffness of the carbox-
ylic residue, as already outlined for Amberlite derivatives
[20]. Although differences in catalytic performances are
small, an increasing trend is again observed from 2 to 4.
In the case of the tertiary alcohol, only traces of the desired
products can be observed, confirming the determinant
influence of steric hindrance.
Compounds 2–4 were also applied as catalysts in a
model reaction mimicking biodiesel production by transe-
sterification of vegetable oils with methanol. Vegetable oils
are a renewable and potentially inexhaustible source of

1970
L. Angiolini et al. / Journal of Organometallic Chemistry 691 (2006) 1965–1972
Table 6
to obtain favourable kinetic features [39]. In the biodiesel
model reaction between glyceryl tridodecanoate and etha-
nol, the catalytic performance of the system is lowered,
confirming the limiting effect of the reactants bulkiness,
however the triphenyltin catalyst displays again the best
performance.
In spite of only moderate catalytic activities, the present
investigation on model triorganotin carboxylates can serve
as a basis, not only to a better understanding of the role of
the Lewis acidity of the tin atom, in competition with steric
demand, but also to the investigation on related triorgano-
tin carboxylates grafted to cross-linked polymeric systems.
It can also be hypothesized that improved performances as
transesterification catalysts can be obtained by properly
varying the substituents at tin as well as by optimally tun-
ing the length of the spacer between the triorganotin car-
boxylate moiety and the aromatic ring of the cross-linked
support, thus favouring a better, but still kinetically labile,
accessibility by the reactants to the metal atom.
Triglyceride conversion obtained in the catalytic transesterification of
glyceryl tridodecanoate with ethanol
Catalyst
Conversion (%)
24 h
48 h
2
3
4
27
20
43
30
25
49
energy with an energetic content similar to diesel fuel [36],
and their transesterification with both methanol and etha-
nol leads readily to a product that can be directly com-
busted in unmodified diesel engines [36–38].
Our model reaction involves glyceryl tridodecanoate, a
simplified triglyceride model, which is converted with an
excess of ethanol into the corresponding fatty acid ethyl
ester, the biodiesel fuel. The catalyst is added in 1% molar
amount with respect to the triglyceride.
The glyceryl tridodecanoate conversion to ethyl dode-
canoate in each catalytic run was assessed (Table 6) by
1H NMR directly on the reaction mixture, after evapora-
tion of unreacted ethanol.
4. Experimental
The catalytic performances of the triorganotin catalysts
in the triglyceride transesterification to produce biodiesel
are lower for 3 and 4 but of similar efficiency for 2, when
compared to the transesterification of ethyl acetate with
1-octanol, which is ascribed to the low accessibility of the
bulky reactants to the triorganotin carboxylate moiety
directly linked to the aromatic ring. However, similarly
to the reaction between ethyl acetate and 1-octanol, the
activity of the trimethyl- and tributyltin derivatives 2 and
3 remains lower than that of the triphenyltin one 4. The
better score of 2 with respect to 3 is attributed to the supe-
rior accessibility to tin atom by the sterically demanding
triglyceride with the trimethyltin carboxylate. Beyond the
reaction time of 24 h it clearly appears that no further sig-
nificant conversion takes place, indicating that the reaction
with the glyceride cannot go to completion.
All chemicals were supplied by Sigma–Aldrich, with the
exception of trimethyltin hydroxide (Strem Chemicals),
and generally used as received. Solvents were purified using
usual purification techniques.
Elemental analyses were performed by REDOX
s.n.c.(Milano).
1H, 13C and 119Sn NMR spectra were recorded at room
temperature on a Varian Gemini 300, a Varian Mercury
400, a Bruker AMX 500 or a Varian Inova 600 spectrom-
eter using CDCl3 solutions, unless otherwise stated, and
tetramethylsilane as internal standard. 117Sn NMR spectra
in the solid state were recorded on a Bruker Avance 250
NMR spectrometer and tetracyclohexyltin as external stan-
dard. Notations: Ci = ipso carbon of the phenyl ring linked
to triorganotin carboxylate; Co (Ho) = ortho; Cm (Hm) =
meta; Cp (Hp) = para carbon (hydrogen) atom with respect
to carboxylate substituent.
3. Conclusions
Infrared spectra were recorded on a Perkin–Elmer 1750
FT-IR spectrometer on KBr pellets.
New catalytic systems containing various triorganotin
carboxylate moieties were prepared by direct triorganost-
annylation of 4-isopropylbenzoic acid. The trimethyltin
derivative exhibits tetracoordinated and pentacoordinated
tin in solution and in the solid state, respectively. By con-
trast, both the tributyltin and triphenyltin esters display
only tetracoordinated tin, regardless of their aggregation
state.
Melting points and DHm values of stannylated com-
pounds have been determined by differential scanning cal-
orimetry (DSC) using a TA Instruments DSC 2920
modulated calorimeter operating under nitrogen at a heat-
ing/cooling rate of 10 ꢁC/min; m.p. values are given as the
temperature at the maximum of the endothermic peak.
4.1. Trimethyltin(4-isopropylbenzoate) (2)
The products displayed catalytic activity in the trans-
esterification of ethyl acetate with decreasing efficiency on
going from primary to secondary and tertiary alcohol.
Interestingly, a significant reverse dependence on Lewis
acidity is displayed in the transesterification with 1-octanol,
the triphenyltin derivative being the most active, which is
explained by the need of weak Lewis acid–base interaction
between tin atom and both the transesterification reactants
In a 250 mL flask 4-isopropylbenzoic acid (1.00 g,
6.09 mmol) and trimethyltin hydroxide (1.10 g, 6.09 mmol)
are dissolved in dry acetone (150 mL). Molecular sieves 5A
are added to remove the water formed as by-product. The
reaction mixture is heated at reflux, the reaction progress
being monitored by FT-IR spectroscopy. The cloudy reac-
tion mixture is finally filtered over a Celite path, in order to

L. Angiolini et al. / Journal of Organometallic Chemistry 691 (2006) 1965–1972
1971
remove traces of molecular sieves, and the solvent evapo-
rated under reduced pressure to leave an oily residue,
which is kept under high vacuum, leading to a white solid.
Elemental analysis (%): calc. for C28H26O2Sn (513.20) C
65.53, H 5.11, O 6.24; found: C 65.39, H 5.09, O 6.26.
1H NMR (300 MHz): d = 1.2 (d, 6H; CH3CH), 2.9
(hept, 1H; CH3CH), 7.3 (dd, 2H; Hm), 7.4–7.5 (m, 9H;
Hm(SnPh) and Hp(SnPh)), 7.8 (bs, 6H; Ho(SnPh)), 8.1 (dd,
2H; Ho) ppm.
Yield 1.39 g (70%); m.p. 147 ꢁC; DHm = 32.6 kJ molÀ1
.
Elemental analysis (%): calc. for C13H20O2Sn (326.99) C
47.75, H 6.16, O 9.79; found: C 47.61, H 6.18, O 9.82.
2
1H NMR (300 MHz): d = 0.6 (s, J(1H,119/117Sn) = 58/
13C NMR (400 MHz): d = 24.0 (CH3CH), 34.5
(CH3CH), 126.5 (Cm), 129.1 (3J(13C–119/117Sn) = 63 Hz;
56 Hz, 9H; SnCH3), 1.3 (d, 6H; CH3CH), 2.9 (hept, 1H;
CH3CH), 7.3 (dd, 2H; Hm), 8.0 (dd, 2H; Ho) ppm.
C
m-SnPh), 130.3 (Cp-SnPh), 131.0 (Co), 131.2 (Ci), 137.1
13C NMR (400 MHz): d = À2.1 (1J(13C–119/117Sn) =
381/399 Hz; SnCH3), 24.0 (CH3CH), 34.4 (CH3CH), 126.4
(Cm), 129.3 (Ci), 130.5 (Co), 153.2 (Cp), 171.8 (COO) ppm.
FT-IR: 2963–2868 (mCH aliph.), 1595 (ma COOSn), 1555,
1508 (mC–C ring), 1371 (ms COOSn), 1179, 1099; 1018 (dCH
(2J(13C–119/117Sn) = 47 Hz; Co-SnPh), 138.8 (Ci-SnPh), 154.4
(Cp), 173.1 (COO) ppm.
FT-IR: 3065–3046 (mCH arom), 2960–2867 (mCH aliph.),
1629 (ma COOSn), 1608, 1566, 1508 (mC–C ring), 1331 (ms COOSn),
1180, 1126, 1077, 1052, 1016 (dCH arom.), 860 (cCH 1,4-
subst. ring), 731 (cCH monosubst. ring), 696 (cC–C mono-
arom.), 860 (cCH 1,4-subst. ring) cmÀ1
.
subst. ring) cmÀ1
.
4.2. Tri-n-butyltin(4-isopropylbenzoate) (3)
4.4. Assessment of the catalytic activity in ethyl acetate
transesterification reactions [19]
In a 100 mL flask equipped with a Dean-Stark appara-
tus, 4-isopropylbenzoic acid (1.00 g, 6.09 mmol) and
bis(tributyltin) oxide (BTBTO) (1.55 mL, 3.05 mmol) are
dissolved in toluene (20 mL). The reaction mixture is
heated until no more water evolution is observed, the reac-
tion progress being monitored by FT-IR spectroscopy. The
oily crude product is separated from unreacted BTBTO by
treatment in heterogeneous phase with dry Amberlite IRC-
86 under stirring for 24 h. The resin is filtered off and
washed with chloroform (3 · 20 mL), the combined chloro-
form extracts are evaporated under reduced pressure to
give a yellowish oil. Yield 2.45 g (76%).
Elemental analysis (%): calc. for C22H38O2Sn (453.23) C
58.30, H 8.45, O 7.06; found: C 58.42, H 8.48, O 7.04.
1H NMR (300 MHz): d = 0.9 (t, 9H; CH3CH2), 1.2 (d,
6H; CH3CH), 1.4–1.2 (m, 12H; CH2Sn and CH3CH2),
1.6 (m, 6H; CH2CH2CH2), 2.9 (hept, 1H; CH3CH), 7.3
(dd, 2H; Hm), 8.0 (dd, 2H; Ho) ppm.
In a 100 mL round-bottomed flask, ethyl acetate
(15 mL) used simultaneously as a transesterification sub-
strate and solvent was added to the appropriate alcohol
(1-octanol, cyclohexanol, 3-ethyl-3-pentanol) so as to
obtain a molar ratio ester/alcohol 7/1. The reaction mix-
ture was heated to reflux, in the presence of 1 mol% of cat-
alyst with respect to the alcohol, the alcohol conversion
being determined after 24 and 48 h by gas chromatography
using an Agilent 6890 instrument equipped with a capillary
methylsilicone column and an Agilent 5973 M mass
detector.
4.5. Assessment of the catalytic activity in glyceryl
tridodecanoate transesterification reactions
In a 100 mL round-bottomed flask, glyceryl tridodecan-
oate (1.0 g, 1.56 mmol) was added to an excess of absolute
ethanol (50 mL). The reaction mixture was heated to reflux
in the presence of 1 mol% of catalyst with respect to the
ester, the reaction progress being determined after 24 and
48 h by integration of two selected resonances at 0.85 and
13C NMR (400 MHz): d = 13.9 (CH3CH2), 16.8
(1J(13C–119/117Sn) = 359/342 Hz; CH2Sn), 24.0 (CH3CH),
27.2 (3J(13C–119/117Sn) = 63 Hz; CH3CH2), 28.1 (2J(13C–
119/117Sn) = 20 Hz; CH2CH2CH2), 34.4 (CH3CH), 126.4
(Cm), 130.0 (Ci), 130.5 (Co), 153.5 (Cp), 171.8 (COO) ppm.
FT-IR: 2958–2871 (mCH aliph.), 1636 (ma COOSn), 1611,
1571, 1509, 1416 (mC–C ring), 1332 (ms COOSn), 1178, 1127,
1
1.15 ppm in the H NMR (600 MHz) spectra in DMSO-
d6 solution of the reaction mixture, after evaporation under
reduced pressure of any unreacted ethanol.
1018 (dCH arom.), 858 (cCH 1,4-subst. ring) cmÀ1
.
4.3. Triphenyltin(4-isopropylbenzoate) (4)
Acknowledgements
In a 100 mL flask equipped with a Dean-Stark appara-
tus, 4-(isopropyl)-benzoic acid (1.46 g, 8.89 mmol) and tri-
phenyltin hydroxide (3.59 g, 9.78 mmol) are dissolved in
toluene (50 mL). The reaction mixture is heated until no
more water evolution is observed, the reaction progress
being monitored by FT-IR spectroscopy. The solvent is
evaporated under reduced pressure and the solid residue
purified by crystallization from ethanol/water. Yield
This work was financially supported by the University
of Bologna (Progetti Pluriennali di Ricerca and Ricerca
Finalizzata Orientata) and Polo Scientifico-Didattico di Ri-
mini (Catalizzatori polimerici a basso impatto ambientale
per reazioni di transesterificazione).
M.B. and R.W. are indebted to the Fund for Scientific
Research–Flanders (Belgium) (FWO) (Grant G.0016.02)
and to the Research Council of the VUB (Grants
GOA31, OZR362, OZR875) for financial support.
3.65 g (80%); m.p. 109 ꢁC; DHm = 23.5 kJ molÀ1
.

1972
L. Angiolini et al. / Journal of Organometallic Chemistry 691 (2006) 1965–1972
[20] L. Angiolini, D. Caretti, L. Mazzocchetti, E. Salatelli, M. Biesemans,
References
R. Willem, Appl. Organomet. Chem. 19 (2005) 841–847.
[21] L. Mazzocchetti, Hybrid organic inorganic polymeric materials
bearing the triorganotin carboxylate moieties, Ph.D. Thesis, Univer-
sity of Bologna, 2005.
[22] U. Schuchardt, R. Sercheli, R.M. Vargas, J. Braz. Chem. Soc. 9
(1998) 199–210.
[23] H.E. Hoydonckx, D.E. De Vos, S.A. Chavan, P.A. Jacobs, Top.
Catal. 27 (2004) 83–96.
[24] R.K. Ingham, S.D. Rosenberg, H. Gilman, Chem. Rev. 60 (1960)
459–539.
[25] K.C. Molloy, K. Quill, S.J. Blunden, R. Hill, Polyhedron 5 (1986)
959–965.
[26] L. Angiolini, D. Caretti, L. Mazzocchetti, E. Salatelli, M. Biesemans,
R. Willem, J. Polym. Sci., Part A: Polym. Chem. 43 (2005) 3091–
3104.
[1] J.A. Thompson, M.G. Sheffer, R.C. Pierce, Y.K. Chau, J.J. Cooney,
W.P. Cullen, R.J. Maguire, Organotin Compounds in the Aquatic
Environment: Scientific Criteria for Assessing their Effects on
Environmental Quality, National Research Council Canada, Ottawa,
1985, p. 284.
[2] A.G. Davies, Organotin Chemistry, VCH, Weinheim, 2004.
[3] D. Seyferth, T.C. Masterman, Appl. Organomet. Chem. 8 (1994)
335–349.
[4] A.A. Mahmoud, A.F. Shaaban, M.M. Azab, N.N. Messia, Eur.
Polym. J. 28 (1992) 555–559.
[5] J.R. Dharia, C.P. Pathak, G.N. Babu, S.K. Gupta, J. Polym. Sci. 26
(1988) 595–599.
[6] R. Ga¨chter, H. Muller, in: R. Ga¨chter, H. Muller (Eds.), Plastics
¨
¨
Additives Handbook, Hanser Publisher, Munich, 1990, pp. 281–287.
[7] J. Otera, Chem. Rev. 93 (1993) 1449–1470.
[8] M. Gielen, Chim. Nouvelle 19 (2001) 3333–3346.
[9] J.K. Tsagatakis, N.A. Chaniotakis, K. Jurkschat, S. Damoun, P.
Geerlings, A. Bouhdid, M. Gielen, I. Verbruggen, M. Biesemans,
J.C. Martins, R. Willem, Helv. Chim. Acta 82 (1999) 531–542.
[10] E. Merian, Metals and Their Compounds in the Environment, VCH,
Weinheim, 1991.
[11] M.A. Champ, P.F. Seligman, Organotins, Environmental Fate and
Effects, Chapman and Hall, London, 1996.
[12] L. Angiolini, M. Biesemans, D. Caretti, E. Salatelli, R. Willem,
Polymer 41 (2000) 3913–3924.
´ˇ
ˇ
[27] J. Holecˇek, K. Handlır, A. Lycka, T.K. Chattopadhyay, B. Majee,
A.K. Kumar, Collect. Czech. Chem. Commun. 51 (1986) 1100–1111.
[28] L. Angiolini, E. Salatelli, D. Caretti, M. Biesemans, H. Dalil, R.
Willem, N.A. Chaniotakis, E. Gouliaditi, K. Perdikaki, Macromol.
Chem. Phys. 203 (2002) 219–229.
[29] R. Willem, A. Bouhdid, B. Mahieu, L. Ghys, M. Biesemans, E.R.T.
Tiekink, M. Gielen, J. Organomet. Chem. 531 (1997) 151–158.
[30] L. Angiolini, D. Caretti, L. Mazzocchetti, E. Salatelli, C. Femoni, J.
Organomet. Chem. 689 (2004) 3301–3307.
[31] B. Wrackmeyer, Ann. Rep. NMR Spectrosc. 16 (1985) 73–186.
´ˇ
`
´
ˇ
[32] J. Holecˇek, K. Handlır, M. Nadvornık, A. Lycka, J. Organomet.
Chem. 258 (1983) 147–153.
[13] M. Hoch, Appl. Geochem. 16 (2001) 719–743.
[33] L. Angiolini, D. Caretti, L. Mazzocchetti, E. Salatelli, M. Biesemans,
R. Willem, J. Polym. Sci., Part A: Polym. Chem. 42 (2004) 5372–5383.
[34] J.T.B.H. Jastrzebski, G. Van Koten, Advances in Organomet. Chem.
35 (1993) 241–294.
[14] B. Delmond, G. Dumartin, in: M. Gielen, R. Willem, B. Wrackmeyer
(Eds.), Physical Organometallic Chemistry. Solid State Organome-
tallic Chemistry. Methods and Applications, John Wiley & Sons,
Chichester, 1999, p. 445.
[15] M. Biesemans, F.A.G. Mercier, M. Van Poeck, J.C. Martins, G.
Dumartin, R. Willem, Eur. J. Inorg. Chem. 14 (2004) 2908–2913.
[16] J.C. Martins, F.A.G. Mercier, A. Vandervelden, M. Biesemans, J.-
M. Wieruszeski, E. Humpfer, R. Willem, G. Lippens, Chem. Eur. J.
8 (2002) 3431–3441.
´
[35] G. Deshayes, F.A.G. Mercier, P. Degee, I. Verbruggen, M. Biese-
mans, R. Willem, P. Dubois, Chem. Eur. J. 9 (2003) 4346–4352.
[36] F.A. Zaher, Grasas Aceites 41 (1990) 82–91.
[37] M. Mittelbach, P. Tritthard, J. Am. Oil Chem. Soc. 65 (1988) 1185–
1187.
[38] F. Staat, G. Vermeersch, Rev. Fr. Corps Gras 40 (1993) 167–174.
[39] The possibility of formation of an alkoxide intermediate by reaction
of tin carboxylate with the alcohol reactant, as suggested by one
reviewer, can be excluded for these derivatives by the observed lack of
tin leakage from triorganotin carboxylates grafted to cross-linked
resins of similar structure as in Chart 2, in the experiments reported in
Refs. [20] and [21]. Alcoholysis would have in fact produced
detachment of the triorganotin moiety from the matrix and a
variation of the properties of the catalyst.
[17] G. Deshayes, K. Poelmans, I. Verbruggen, C. Camacho-Camacho,
´
P. Degee, V. Pinoie, J.C. Martins, M. Piotto, M. Biesemans, R.
Willem, P. Dubois, Chem. Eur. J. 11 (2005) 4552–4561.
[18] C. Camacho-Camacho, M. Biesemans, M. Van Poeck, F.A.G.
Mercier, R. Willem, K. Darriet-Jambert, B. Jousseaume, T. Tou-
pance, U. Schneider, U. Gerigk, Chem. Eur. J. 11 (2005) 2455–2461.
[19] F.A.G. Mercier, M. Biesemans, R. Altmann, R. Willem, R. Pintelon,
R. Schoukens, B. Delmond, G. Dumartin, Organometallics 20 (2001)
958–962.