Reactivity of carbon dioxide with n-butyl(phenoxy)-,(alkoxy)-, and (oxo)stannanes: insight into dimethyl carbonate synthesis

Article abstract of DOI:10.1021/om000397f

The CO₂ insertion into Sn-O bonds of a series of butyl(phenoxy)-, (alkoxy)-, and (oxo)-stannanes has been investigated. The tributyl derivatives Bu₃SnOR (2a, R = Me; 3a, R = ⁱPr; 4a, R = ^tBu; 5a, R = SnBu₃)¹ give quantitatively Bu₃Sn(OCO₂R), 2b-5b; the analogous tributylphenoxystannane, 1, is less reactive. For the dibutyl series, Bu₂Sn(OR)₂, steric effects of ^tBu groups in OR (8a) suppress carbonation under atmospheric pressure. With R = Me (6a) or R = ⁱPr (7a), only one Sn-OR bond reacts, resulting in Bu₂Sn(OR)(OCO₂R), 6b or 7b. Treating 6a with 2-propanol affords under CO₂ the mixed compound Bu₂Sn(OMe)-(OCO₂ⁱPr), selectively. Facile deinsertion of CO₂ is a common property of all compounds, occurring more readily in the dibutyl series. The stoichiometric transformation of the carbonato ligand in 2b, 5b, or 6b to dimethyl carbonate (DMC) on reaction with MeI requires nucleophilic assistance by F^- to proceed. In the presence of MeOH, 2b and 5b are almost inactive for DMC formation, in contrast with 6b. The best yield is obtained under supercritical CO₂-methanol conditions.

Full text of DOI:10.1021/om000397f

Organometallics 2000, 19, 4563-4567
4563
Rea ctivity of Ca r bon Dioxid e w ith n -Bu tyl(p h en oxy)-,
(a lk oxy)-, a n d (oxo)sta n n a n es: In sigh t in to Dim eth yl
Ca r bon a te Syn th esis
Danielle Ballivet-Tkatchenko,*,† Olivier Douteau, and Stefanie Stutzmann
Institut de Recherches sur la Catalyse, CNRS-UPR 5401 Lie´e par Convention a` l’Universite´
Claude Bernard Lyon 1 et a` l’Ecole Normale Supe´rieure de Lyon, 2 Avenue Albert Einstein,
69626-Villeurbanne Cedex, France
Received May 10, 2000
The CO2 insertion into Sn-O bonds of a series of butyl(phenoxy)-, (alkoxy)-, and (oxo)-
stannanes has been investigated. The tributyl derivatives Bu3SnOR (2a , R ) Me; 3a , R )
iPr; 4a , R ) tBu; 5a , R ) SnBu3)1 give quantitatively Bu3Sn(OCO2R), 2b-5b; the analogous
tributylphenoxystannane, 1, is less reactive. For the dibutyl series, Bu2Sn(OR)2, steric effects
t
of Bu groups in OR (8a ) suppress carbonation under atmospheric pressure. With R ) Me
i
(6a ) or R ) Pr (7a ), only one Sn-OR bond reacts, resulting in Bu2Sn(OR)(OCO2R), 6b or
7b. Treating 6a with 2-propanol affords under CO2 the mixed compound Bu2Sn(OMe)-
i
(OCO2 Pr), selectively. Facile deinsertion of CO2 is a common property of all compounds,
occurring more readily in the dibutyl series. The stoichiometric transformation of the
carbonato ligand in 2b, 5b, or 6b to dimethyl carbonate (DMC) on reaction with MeI requires
nucleophilic assistance by F- to proceed. In the presence of MeOH, 2b and 5b are almost
inactive for DMC formation, in contrast with 6b. The best yield is obtained under supercritical
CO2-methanol conditions.
In tr od u ction
their low activity is a problem. To circumvent this
drawback, an important step is an understanding of the
reaction mechanism. Since the first reports of CO2
insertion into the Sn-O bond of Bu3SnOMe, (Bu3-
Sn)2O,11 and Bu2Sn(OMe)2,12 only bis(trialkyltin)car-
bonates were characterized by X-ray structure deter-
mination,13,14 until very recently, when the structure of
Me2Sn(OMe)(OCO2Me) was published.15
Dimethyl carbonate (DMC) has a number of specialty
chemical applications,2 but its primary use to date is in
the production of polycarbonates.3 It is now considered
as an option for fuel additives.4 This potential market
constitutes a challenge for DMC production via envi-
ronmentally friendly processes to avoid the use of
phosgene and formation of the coproduct HCl (eq 1).
More generally, new applications for organic carbonates
can be anticipated through safer synthetic methods.5 In
this context, the catalytic oxidative carbonylation of
methanol was studied and is now operative on a
commercial scale (eq 2).6 However, substitution of
phosgene by carbon dioxide is more appealing, through
the carbonation of alcohols (eq 3).7 Dibutyldialkoxy-
In this paper we report a detailed study focused on
key steps of reaction 3 to proceed in the presence of
organostannane derivatives by (i) the evaluation of
carbon dioxide insertion reactivity into the Sn-O bond
according to the nature of the R groups in the series
Bu3SnOR (R ) Ph, Me, iPr, tBu, SnBu3) and Bu2Sn(OR)2
i
t
(R ) Me, Pr, Bu)1 and (ii) the conditions of DMC
formation from the corresponding carbonato species.
Our approach employs 1H, 13C, and 119Sn NMR spec-
troscopy, volumetric experiments, and reaction either
with methanol in supercritical CO2 or with methyl
iodide. IR spectroscopy was also used to monitor the
presence of absorption bands of the carbonato moieties.16
stannanes are known to promote reaction 3,8-10 but
(8) Yamazaki, N.; Nakahama, S. Ind. Eng. Chem. Prod. Res. Dev.
1979, 18, 249.
(9) Genz, J .; Heitz, W. Ger. Offen. DE 3 203 190, 1983; Chem. Abstr.
1983, 99, 139320s.
(10) Kizlink, J .; Pastucha, I. Collect. Czech. Chem. Commun. 1994,
59, 2116.
(11) Bloodworth, A. J .; Davies, A. G.; Vasishtha, S. C. J . Chem. Soc.
(C) 1967, 1309.
(12) Davies, A. G.; Harrison, P. G. J . Chem. Soc. (C) 1967, 1313.
(13) Tiekink, E. R. T. J . Organomet. Chem. 1986, 302, C1.
(14) Ku¨mmerlen, J .; Sebald, A.; Reuter, H. J . Organomet. Chem.
1992, 427, 309.
* Corresponding author. Fax: + 33 3 80 39 37 72. E-mail: ballivet@
u-bourgogne.fr.
† Present address: LSEO, Universite´ de Bourgogne, Faculte´ des
Sciences Mirande, BP 47870, 21078 Dijon Cedex, France.
(1) In all cases where Bu is written, it is n-C4H9 groups that are
present.
(2) Mauri, M.; Romano, U.; Rivetti, F. Ing. Chim. Ital. 1985, 21, 6.
(3) Chem. Br. 1994, 970.
(4) Pacheco, M. A.; Marshall, C. L. Energy Fuels 1997, 11, 2.
(5) Shaikh, A.-A. G.; Sivaram, S. Chem. Rev. 1996, 96, 951, and
references therein.
(6) Romano, U.; Tesei, R.; Massi Mauri, M.; Rebora, P. L. Ind. Eng.
Chem. Prod. Res. Dev. 1980, 19, 396.
(7) Aresta, M.; Quaranta, E. CHEMTECH 1997, March, 32.
(15) Choi, J .-C.; Sakakura, T.; Sako, T. J . Am. Chem. Soc. 1999,
121, 3793.
(16) Palmer, D. A.; Van Eldik, R. Chem. Rev. 1983, 83, 651.
10.1021/om000397f CCC: $19.00 © 2000 American Chemical Society
Publication on Web 10/06/2000

4564 Organometallics, Vol. 19, No. 22, 2000
Ballivet-Tkatchenko et al.
Resu lts a n d Discu ssion
Rea ction of CO2 w ith Bu 3Sn OR. To systematically
investigate the reactivity of Sn-OR bonds toward CO2,
we prepared a series of compounds for which the R
group may influence the reactivity through electronic
i
t
and steric effects: R ) Ph, 1; Me, 2a ; Pr, 3a ; Bu, 4a ;
SnBu3, 5a . These tributylstannane derivatives are
colorless liquids at room temperature and monomeric,
as evidenced by 119Sn NMR spectroscopy.17,18 The inser-
tion of CO2 into the Sn-O bond has been known for a
long time for 2a and 5a .11,12 In our study, volumetric
experiments showed no significant reaction with 1 (CO2:
Sn ≈ 0.1) over several hours of exposure under atmos-
pheric pressure of CO2 at room temperature, neither in
toluene or DMF, nor in the presence of pyridine or 2,2′-
bipyridine. However, a progressive CO2 uptake occurred
for the other compounds, either as neat samples or in
solution (heptane, toluene, or THF); the equilibrium was
reached within 30 min for 2a and 3a , but needed a
longer time, 18 h, for 4a . The maximum uptake cor-
responded to a CO2:Sn molar ratio of 1. The difference
of reactivity observed between 1 and 2a (or 5a ) follows
the basicity of the oxygen atom as previously reported,19
whereas for 4a steric hindrance of the tert-butyl sub-
stituent causes a decrease in the rate of CO2 fixation.
The neat CO2 adducts, denoted 2b, 3b, and 4b, exhib-
ited a strong ν(CdO) IR band between 1610 and 1590
cm-1, while for 5b the absorption was centered at 1525
cm-1. The 13C{1H} NMR spectra showed a new reso-
nance in the carbonato region at δ 158.4, 157.7, 157.3,
and 162.7 for 2b, 3b, 4b, and 5b, respectively. The
variable-temperature 13C{1H} NMR spectrocopy are
needed to obtain a better insight into the exchange
phenomena, by the determination of the energy barrier
as a function of the nature of R.
Rea ction of CO2 w ith Bu 2Sn (OR)2. Substitution of
one butyl group by alkoxy results in dimeric structures
i
with five-coordinate tin atoms for R ) Me, 6a , and Pr,
7a , whereas for tBu, 8a , steric hindrance favors the
monomeric form.16 In our work, the carbonation reaction
was found to depend on the nature of R; the adducts
were less stable than those of Bu3SnOR, which cor-
relates with the basicity of the oxygen atoms reported
for the series Bu3SnOPh < Bu2Sn(OMe)2 < Bu3-
SnOMe.18 The uptake of CO2 by 6a and 7a corresponded
to a maximum CO2:Sn ) 0.90 at room temperature and
atmospheric pressure of CO2, in solvents such as
toluene, heptane, THF, and acetonitrile. When the
adsorption proceeded at 0 °C, the ratio CO2:Sn increased
to only 1.0 and never approached 2 as reported earlier.12
This observation is in full agreement with very recently
published results on Me2Sn(OMe)2.15 Under the same
conditions, 8a did not fix CO2. However, exposure of
neat 8a to 30 bar of CO2 at room temperature for 16 h
provided a CO2:Sn ratio of 0.4, determined by back-
titration under acidic hydrolysis. Clearly, steric effects
are involved with the tert-butoxy groups, as was found
in the Bu3SnOR series (vide supra).
3
corresponding 13C NMR resonance revealed J C,H cou-
plings for 2b (q, 3.7 Hz) and 3b (d, 2.7 Hz) and none for
5b, which agrees with the presence of Bu3Sn-OC(O)-
OR species. The signals OCH3 (δ 53.5) and OCH(CH3)2
(δ 69.7) appeared as a quartet and a doublet of multi-
As neat liquids, 6a and 7a were transformed into
white solids, 6b and 7b, under atmospheric CO2 at room
temperature, the reaction being reversible under vacuum.
The IR spectra showed ν(CdO) IR bands centered
between 1640 and 1620 cm-1. Multinuclear NMR re-
vealed the disappearance of the starting compound with
the following features: for 6b, two methoxy resonances
1
plet, respectively. The H resonance of OCH(CH3)2 was
a septet centered at δ 4.57. The 119Sn{1H} NMR
spectrum recorded for 3b exhibited one broad signal at
δ -28 (W1/2 ) 850 Hz), shifting upfield to -45 (W1/2
)
680 Hz) upon cooling to -30 °C. These values are close
to those already reported for 2b, indicating the presence
of five-coordinate tin atoms due to the bidentate char-
acter of the carbonate monoester ligand.20 This ligand
is probably better described as a bridging one, by
comparison with the carbonate analogue 5b . For the
latter compound, two 119Sn{1H} NMR signals of equal
intensity belonging to four- and five-coordinate tin
atoms were found (eq 4).16 At room temperature, line
broadening was observed for the resonances of the R
carbon atom of the butyl groups, Sn-CH2-C3H7, and
of dissolved CO2, arising from an exchange process on
the NMR time scale involving reversible insertion of
CO2 into the Sn-OR bond. Under vacuum, 2a was
quantitatively recovered at room temperature, while 3a ,
4a , and 5a were recovered at higher temperatures: 36,
40, and 65 °C, respectively. Further experiments by
1
were present in the H spectrum (δ 3.66 and 3.53; 1:1
ratio) and in the 13C{1H} spectrum (δ 54.09 and 52.71).
In addition, there was a carbonato signal at δ 158.9,
and the R carbon atom of the butyl groups appeared as
a broad resonance at δ 24.03. Similarly for 7b, the
OCH(CH3)2 appeared as two septets (δ 4.62, 3J H,H ) 6.2
3
Hz and 4.12, J H,H ) 6.1 Hz) in a 1:1 ratio, and two
OCH(CH3)2 signals were found at δ 69.9 and 67.4 in the
13C{1H} spectrum. The resonance of the carbonato group
was at δ 158.1, and that for the R carbon atom of the
butyl groups at 21.8 (broad). The identification of 6b
and 7b as Bu2Sn(OR)(OCO2R) species is strengthened
by a very recent single-crystal X-ray diffraction study
of Me2Sn(OMe)(OCO2Me).15 The compound is dimeric
through OMe bridges with a monodentate methyl car-
bonato ligand. Comparison of the NMR data with those
of 6b shows very close OCH3 chemical shifts, the higher
field signal being assigned to the methoxy ligands.
To gain further insight into the reactivity of 6a toward
substitution of the methoxy ligands, exchange with
2-propanol followed by CO2 absorption was carried out.
When 6a was dissolved in 2-propanol at room temper-
ature for 3 h, then treated under vacuum, and finally
(17) Smith, P. J .; White, R. F. M.; Smith, L. J . Organomet. Chem.
1972, 40, 341.
(18) Nadvornik, M.; Holecek, J .; Handlir, K.; Lycka, A. J . Orga-
nomet. Chem. 1984, 275, 43.
(19) Marchand, A.; Mendelsohn, J .; Lebedeff, M.; Valade, J . J .
Organomet. Chem. 1969, 17, 379.
(20) Blunden, S. J .; Hill, R.; Ruddick, J . N. R. J . Organomet. Chem.
1984, 267, C5.

Insight into Dimethyl Carbonate Synthesis
Organometallics, Vol. 19, No. 22, 2000 4565
F igu r e 2. Rate of production of DMC from 2b (1 equiv)
in the presence of MeI (2 equiv) and CsF (1.4 equiv) at 19
°C in DMF, NMP, DME, and acetone.
N,N-dimethylformamide (DMF) failed to afford DMC.
This lack of reactivity was circumvented by the addition
of CsF (1.4 equiv) to the reaction medium. DMF turned
out to be the best solvent for rate enhancement, in
agreement with previous reports on the etherification
of alcohols21 and esterification of carboxylic acids22
mediated by tin compounds. DMC was quantitatively
formed within 6 h, whereas a longer time was needed
in DME, NMP, or acetone in the following order: DMF
< NMP, DME , acetone. The initial rate of formation
is quite similar in DMF, NMP, or DME and much slower
in acetone (Figure 2). Changing the fluoride source to
KF led to an induction period linked to solubility
problems: addition of 18-crown-6 gave a kinetic curve
very similar to that of CsF. When the molar ratio CsF:
2b was <1, the yield of DMC was equal to the amount
of CsF introduced. The selective formation of the organic
carbonate versus dialkyl ether was confirmed with the
system Bu3SnOC(O)OEt/EtI/CsF/DMF, because diethyl
ether could be more easily quantified than dimethyl
ether, if any. Diethyl carbonate was formed in stoichio-
metric amount, whereas no diethyl ether was detected.
On the basis of these results, the behavior of 5b and
6b was studied in DMF at 19 °C with a Sn:MeI:CsF
molar ratio ) 1:2:1.4. The molar ratio DMC:Sn reached
a maximum value after 23 h of 0.48 or 0.94, respectively.
Interestingly, addition of 2.6 equiv of CsF to 6b in-
creased the value DMC:Sn to 1.7 after 22 h, which
indicates that both methoxy groups of the starting
compound 6a can react.
The driving force for the alkylation of organotin
carbonates to form DMC may originate from the inter-
action of fluoride anion with the tin atom by increasing
the nucleophilicity of the masked alkyl carbonate. To
corroborate this hypothesis, the 119Sn{1H} NMR spec-
trum of 2b/DMF revealed one resonance at δ -19
characteristic of five-coordinate tin. After 5 h of reaction
at room temperature with 2 equiv of CsF, the signal
shifted to δ -48, and upon a further 2 h of reaction to
-62 (an authentic sample of Bu3SnF/DMF exhibited a
single resonance at δ -79). In the presence of MeI, after
23 h of reaction (stoichiometric amount of DMC formed),
evaporation of the organics and extraction of the residue
with toluene allowed us to identify Bu3SnF as the only
tin product, by comparison of the 13C NMR and IR
spectra with those of an authentic sample (eq 5).
F igu r e 1. 13C{1H} NMR spectrum of Bu2Sn(OMe)(OiPr)
under CO2, and H NMR spectrum of the alkoxy region.
1
Sch em e 1
submitted to CO2 at room temperature for 1 h, only one
compound was present. Interestingly, the 1H and 13C{1H}
NMR spectra showed that it is neither 6b nor 7b, but a
mixed species resulting from the substitution of one
OMe by OiPr. Besides the 1H multiplets of the alkyl
moieties, only one septet for OCH(CH3)2, δ 4.72 (J (H,H)
) 6.2 Hz), and one singlet for OCH3, δ 3.52, were
present in a 1:3 ratio. In the 13C resonance region of
the alkyl moieties, one signal was observed for (i) each
carbon atom of the butyl groups (CR, br, 24.16; Câ,
26.97; Cγ, 26.69, 3J (13C,119Sn) ) 112, 3J (13C,117Sn) ) 108
Hz; Cδ, 13.48) and (ii) OCH(CH3)2, δ 21.97. The reso-
nances at δ 70.04, 52.77, and 158.2 were attributed to
OCH(CH3)2, OCH3, and OC(O)O, respectively (Figure
1). By comparison with 6b and 7b data, only the low-
1
field H and 13C resonances of OCHMe2 were present,
with the concomitant high-field resonances of OCH3. On
the basis of previous assignments for Me2Sn(OMe)-
(OCO2Me),15 the existence of only the higher field
methoxy signal leads us to propose that CO2 has been
selectively inserted into the Sn-OiPr bond, according
to the following steps (Scheme 1). Selective bridging of
the methoxy versus the isopropoxy ligand can originate
from steric effects, coupled with the higher stability of
the isopropoxy carbonate fragment as found in the
tributyl series (vide supra).
Rea ctivity of 2b, 5b, a n d 6b. The formation of DMC
resulting from electrophilic attack of MeI on the oxygen
atom linked to tin, Sn-OC(O)OMe, was studied first.
Exposure of 2b (1 equiv) to MeI (2 equiv) at 19 °C either
in methanol, dichloromethane, 1,2-dimethoxyethane
(DME), acetone, 1-methyl-2-pyrrolidinone (NMP), or
(21) Ando, T.; Yamawaki, J .; Kawate, T.; Sumi, S.; Hanafuda, T.
Bull. Chem. Soc. J pn. 1982, 55, 2504.
(22) Sato, T.; Otera, J .; Nozaki, H. J . Org. Chem. 1992, 57, 2166.

4566 Organometallics, Vol. 19, No. 22, 2000
Ballivet-Tkatchenko et al.
i
Sn(OR)2 for R ) Me and Pr, forming 1:1 adducts even
in the case of dialkoxides. With the latter compounds,
reversibility occurs more readily. The selective forma-
i
tion of Bu2Sn(OMe)(OCO2 Pr) shows that the isopropoxy
is prompt to (i) insert CO2 and (ii) stabilize the alkyl
carbonate ligand, a situation also encountered with the
monoalkoxide compounds 2a and 3a . The transforma-
tion of the carbonato ligand to dimethyl carbonate is
feasible with MeI, through the assistance of fluoride
anion to increase the nucleophilicity of the alkyl carbon-
ate ligand. The difference in reactivity between mono-
and dimethoxy species is better seen by running the
reaction in methanol: DMC is formed only with 6b. The
best yield is obtained under a supercritical CO2-
methanol mixture. Although an intramolecular pathway
produces DMC, methanol enhances the yield. It is
noteworthy that dibutyl(oxo)stannane, (Bu2SnO)n, has
a comparable reactivity.
F igu r e 3. DMC formation from 6b (4 mmol) in the
presence of MeOH under CO2 pressure (MeOH:CO2 ) 0.8)
at 145 °C as a function of (a) pressure and (b) reaction time.
Exp er im en ta l Section
Accordingly, we propose that the two methoxy groups
of 6a are carboxylated stepwise in the presence of
fluoride as depicted in eq 6. No tin compounds could be
extracted with toluene, which is not surprising if Bu2-
SnF2 is formed.
Gen er a l Com m en ts. All reactions were carried out under
dry argon using Schlenk tube techniques. Chemicals were
purchased from Aldrich and Acros Chimica. Bu3SnX (X ) F,
Cl, I), Bu2SnCl2, (Bu3Sn)2O, and (Bu2SnO)n were used as
received. The anhydrous fluorides were treated under vacuum
at 140 °C for 20 h, just before use. CO2 N45 was purchased
from Air Liquide. The solvents were purified by standard
methods. 1H NMR spectra were recorded on a Bruker AC 100
(100.130 MHz), Bruker AM 250 (250.133 MHz), or Bruker
AMX 400 (400.132 MHz) spectrometer. 13C NMR spectra were
run on a Bruker AC 100 (25.178 MHz), Bruker AC 200 (50.323
MHz), Bruker AM 250 (62.896 MHz), or Bruker AMX 400
(100.623 MHz) spectrometer. Chemical shifts (δ, ppm) were
determined relative to the solvent (1H, CHCl3 δ 7.24; 13C,
CDCl3 δ 77.00) and converted to the δ scale downfield from
Me4Si. Additional DEPT experiments allowed identification of
methyl and methylene resonances, when needed. 119Sn{1H}
NMR spectra were recorded on a Bruker AMX 400 (149.207
MHz) spectrometer. Chemical shifts (δ, ppm) are reported
downfield from Me4Sn. Infrared spectra were obtained with a
Perkin-Elmer 597 spectrometer, the sample being placed
between KBr windows either as neat or dispersed in Nujol.
Elemental analysis was performed at the Laboratoire de
Synthe`se et Electrosynthe`se Organome´talliques, Universite´ de
Bourgogne, Dijon.
A second approach for DMC formation was to study
reaction 3 under catalytic conditions, i.e., replacing MeI
by MeOH. In contrast with the identical behavior of 2b,
5b, and 6b toward MeI, we found that species 2b and
5b treated under 75 bar (148 °C, 20 h) with an excess
of MeOH/CO2 mixture afforded a DMC:Sn molar ratio
that did not exceed 0.1; addition of CsF had no effect.
Interestingly, 6b was found much more effective. This
observation was surprising, since the CO2 adduct is less
stable (vide supra). At 145 °C and for 20 h of reaction,
the DMC:Sn ratio increased with CO2 pressure to a
value of 0.8, which remained constant between 200 and
250 bar (Figure 3a). Addition of CsF (1 equiv) has no
effect on the yield. Under these supercritical conditions,
the DMC:Sn ratio was found to increase to ∼1 with
reaction time during the first 50 h (Figure 3b). To check
if methanol was a reactant, it was replaced by toluene.
DMC did form, but the DMC:Sn ratio remained constant
at a value of 0.4 with an increase in CO2 pressure from
80 to 190 bar. These results show that an intramolecular
pathway is operative with Bu2Sn(OMe)(OCO2Me). How-
ever, methanol plays a role in the reaction mechanism.
We also observed that dibutyl(oxo)stannane, (Bu2SnO)n,
led to DMC under supercritical CO2 in the presence of
methanol (210 bar, 140 °C, 20 h). The DMC:Sn value
was equal to 0.6 versus 0.8 for 6b, under the same
experimental conditions.
Bu 3Sn OP h (1). The synthesis was adapted from ref 23.
Hexabutyldistannoxane, (Bu3Sn)2O (16.4 g, 27.5 mmol), was
mixed with phenol (5.2 g, 55.2 mmol) in 100 mL of heptane.
After heating under reflux for 2 h, water, heptane, and the
excess phenol were removed under vacuum, leaving a colorless
oil. Anal. Calcd for C18H32OSn: C, 56.43; H, 8.42. Found: C,
56.64; H, 8.40. 1H NMR (100 MHz, CDCl3): δ 7.3-6.6 (m, 5
H), 1.8-0.7 (m, 27 H). 13C{1H} NMR (25 MHz, CDCl3): δ butyl
1
CR 15.8 (1J (13C,119Sn) ) 353, J (13C,117Sn) ) 338 Hz), Câ 27.6
(2J (13C119,117Sn) ) 20 Hz), Cγ 26.9 (3J (13C,119Sn) ) 63;
1J (13C,117Sn) ) 60 Hz), Cδ 13.5, phenyl 162.0, 129.1, 119.5,
118.2.
Bu 3Sn OR (2a -4a ) a n d Bu 2Sn (OR)2 (6a -8a ). The syn-
thesis was adapted from ref 24. Butylalkoxystannanes were
prepared from the corresponding butylchlorostannanes and
sodium alkoxide.
Typical procedure for 2a : tributyltin chloride (18 g, 55.3
mmol) in dry toluene (30 mL) was added dropwise to sodium
methoxide (sodium, 1.5 g) in methanol (25 mL) at 0 °C. After
the addition, the reaction mixture was heated at reflux for 6
Con clu sion
(23) Sasin, G. S.; Sasin, R. J . Org. Chem. 1955, 20, 770.
(24) Mehrotra, R. C.; Gupta, V. D. J . Organomet. Chem. 1965, 4,
145.
We have shown that the CO2 insertion into Sn-O
bonds is facile in the alkoxide series Bu3SnOR and Bu2-

Insight into Dimethyl Carbonate Synthesis
Organometallics, Vol. 19, No. 22, 2000 4567
h. The sodium chloride formed was centrifuged, leaving a
supernatant solution from which toluene and methanol were
removed under vacuum. Distillation of the crude oil with a
Bu¨chi GKR51 operating under reduced pressure (100 °C, 5
10-2 mbar) gave tributylmethoxystannane (13.3 g, yield 75%).
All compounds were characterized by elemental analysis
and NMR spectroscopy; the 13C data for 5a are also included.
2a . Anal. Calcd for C13H30OSn: C, 48.63; H, 9.42. Found:
C, 48.60; H, 9.30. 1H (100 MHz, CDCl3): δ 3.5 (s, 3H), 1.8-0.7
(27H). 13C{1H} (25 MHz, CDCl3): δ butyl CR 14.1 (1J (13C,119Sn)
83.0 Hz), Cδ 13.73, isopropoxy Câ′ 69.68, Cγ′ 22.06, carbonato
157.72. 119Sn{1H} (149 MHz, CDCl3): δ -28.
5b. 13C{1H} (25 MHz, CDCl3): δ butyl CR 17.4 (1J (13C,119Sn)
1
) 412, J (13C,117Sn) ) 403 Hz), Câ 28.1 2J (13C,117,119Sn) ) 23
Hz), Cγ 27.3 3J (13C,117,119Sn) ) 76 Hz), Cδ 13.7, carbonato
162.7.
1
6b. H (250 MHz, CDCl3): δ 3.66 (s, 3H), 3.53 (s, 3H), 1.8-
0.8 (18H). 13C{1H} (63 MHz, CDCl3): δ butyl CR 23.93 (br),
Câ 26.94, Cγ 26.72 3J (13C,117,119Sn) ) 107 Hz), Cδ 13.51,
methoxy 54.02, 52.69, carbonato 158.87.
1
) 360, J (13C,117Sn) ) 344 Hz), Câ 28.3 (2J (13C,119,117Sn) ) 21
7b. 1H (100 MHz, CDCl3): δ 4.62 (sept, J (H,H) ) 6.2 Hz,
1H), 4.12 (sept, J (H,H) ) 6.1 Hz, 1H), 1.8-0.6 (30H). 13C{1H}
(25 MHz, CDCl3): δ butyl CR 25.9 (br), Câ 26.8, Cγ 26.5, Cδ
13.3, isopropoxy Câ′ 69.9, 67.4, Cγ′ 21.8, carbonato 158.2.
Ga som etr y. A Schlenk tube containing 1 mmol of the tin
compound in 1 mL of solvent was connected to a pressure
transducer and to a CO2 reservoir of known pressure. The
volumes of each part of the apparatus were determined, and
the amount of CO2 gas absorbed was calibrated by the
reference experiment in the absence of the tin compound. The
calculated CO2:Sn molar ratio was at (0.05.
Rea ction of 2b, 5b, a n d 6b w ith MeI. Compound 2a , 5a ,
or 6a was submitted to an atmospheric pressure of CO2 at 19
°C for 2 h. Then, a 0.3 M solution was prepared with the
appropriate solvent under CO2, followed by the addition of
toluene (internal standard, 1 equiv), methyl iodide (2 equiv),
and CsF (1.4 equiv). The identification of dimethyl carbonate
was performed by GC-MS (Fisons MD 800, EI 70 eV, J &W
Scientific DB-1 60 m capillary column), and the quantitative
analysis was done by GC (Shimadzu 14 A, FID detector, J &W
Scientific DB-WAX 15 m megabore column).
Rea ction u n d er CO2 P r essu r e. In a 100 mL stainless
steel batch reactor was introduced a methanolic solution or
suspension (10-30 mL) of the tin compound (∼4 mmol). The
reactor was pressurized with CO2 to a MeOH:CO2 molar ratio
of 0.8 and heated to the desired temperature controlled by an
internal thermocouple. At the end of the reaction, the reactor
was cooled to 0 °C and depressurized, and the condensed phase
was transferred to a Schlenk tube for analysis by GC (Fisons
8000, FID detector, J &W Scientific DB-WAX 15 m megabore
column) and identification by GC-MS (Fisons MD 800, EI 70
eV, J &W Scientific DB-1 60 m capillary column). Evaporation
of the volatiles under vacuum at room temperature allowed
the characterization of the residue by NMR.
Hz), Cγ 27.3 (3J (13C,119,117Sn) ) 57 Hz), Cδ 13.7, methoxy 54.3.
3a . Anal. Calcd for C15H34OSn: C, 51.60; H, 9.82. Found:
C, 50.93; H, 9.50. 1H (100 MHz, CDCl3): δ 3.9 (sept, J (H,H) )
6 Hz, 1H), 1.8-0.7 (33H). 13C{1H} (25 MHz, CDCl3): δ butyl
CR 14.7 (1J (13C,119Sn) ) 363, 1J (13C,117Sn) ) 347 Hz), Câ 27.5,
Cγ 27.1, Cδ 13.5, isopropoxy Câ′ 66.5, Cγ′ 28.0.
4a . Anal. Calcd for C16H36OSn: C, 52.92; H, 9.99. Found:
1
C, 52.89; H, 9.94. H (100 MHz, CDCl3): δ 1.7-0.75. 13C{1H}
(25 MHz, CDCl3): δ butyl CR 16.1 (1J (13C,119Sn) ) 365,
1J (13C,117Sn) ) 349 Hz), Câ 28.1 (2J (13C,119,117Sn) ) 17 Hz), Cγ
27.1, Cδ 13.6, tert-butoxy Câ′ 70.5, Cγ′ 33.6 (3J (13C,119,117Sn)
) 13 Hz).
5a . Anal. Calcd for C24H54OSn2: C, 48.36; H, 9.13. Found:
C, 48.78; H, 9.07. 13C{1H} (25 MHz, CDCl3): δ butyl CR 16.3
(1J (13C,119Sn) ) 366, 1J (13C,117Sn) ) 350 Hz), Câ 28.1 (2J -
(13C,119,117Sn) ) 19 Hz), Cγ 27.2 (3J (13C,119,117Sn) ) 63 Hz), Cδ
13.5.
6a . Anal. Calcd for C10H24O2Sn: C, 40.72; H, 8.20. Found:
1
C, 40.61; H, 8.41. H (100 MHz, CDCl3): δ 3.52 (s, 6H), 1.8-
0.6 (18H). 13C{1H} (25 MHz, CDCl3): δ butyl CR 19.2
(1J (13C,119Sn) ) 636, 1J (13C,117Sn) ) 608 Hz), Câ 27.3 (2J -
(13C,119,117Sn) ) 30 Hz), Cγ 26.9 (3J (13C,119,117Sn) ) 96 Hz), Cδ
13.5, methoxy 51.9.
7a . Anal. Calcd for C14H32O2Sn: C, 47.89; H, 9.19. Found:
1
C, 48.27; H, 9.32. H (100 MHz, CDCl3): δ 4.11 (sept, J (H,H)
) 6.0 Hz, 2H), 1.8-0.7 (30H). 13C{1H} (25 MHz, CDCl3): δ
butyl CR 18.4 (1J (13C,119Sn) ) 498, 1J (13C,117Sn) ) 478 Hz), Câ
27.1, Cγ 26.9, Cδ 13.4, isopropoxy Câ′ 66.6, Cγ′ 27.5.
8a . Anal. Calcd for C16H36O2Sn: C, 50.69; H, 9.57. Found:
1
C, 50.37; H, 9.46. H (100 MHz, CDCl3): δ 1.7-0.75. 13C{1H}
(25 MHz, CDCl3): δ butyl CR 21.4 (1J (13C,119Sn) ) 489,
1J (13C,117Sn) ) 467 Hz), Câ 27.2, Cγ 26.7 (3J (13C,119,117Sn) )
85 Hz), Cδ 13.4, tert-butoxy Câ′ 71.5, Cγ′ 33.6 (3J (13C,119,117Sn)
) 15 Hz).
Ca r bon a tion Rea ction . The tin compounds were dissolved
in CDCl3, then CO2 was admitted into the Schlenk tube, at
room temperature. After the solution was stirred under CO2
for at least 2 h, it was transferred into the NMR tube for
analysis.
Ack n ow led gm en t. We are grateful for financial
support of this work from the Centre National de la
Recherche Scientifique under the program “Catalyse et
catalyseurs pour l’industrie et l’environnement”. The
authors wish to thank S. Mangematin for technical
assistance.
2b. 1H (100 MHz, CDCl3): δ 3.4 (s, 3H), 1.8-0.7 (27H).
13C{1H} (25 MHz, CDCl3): δ butyl CR 18.1 (br), Câ 28.1
(2J (13C,119,117Sn) ) 26 Hz), Cγ 27.1 (3J (13C,119,117Sn) ) 79 Hz),
Cδ 13.6, methoxy 53.5, carbonato 158.4.
Su p p or tin g In for m a tion Ava ila ble: 1H and 13C{1H}
NMR spectra for 2b, 3b, 5b, 6b, and 7b. This material is
available free of charge via the Internet at http://pubs.acs.org.
3b. 1H (400 MHz, CDCl3): δ 4.57 (sept J (H,H) ) 4.8 Hz,
1H), 1.8-0.7 (33H). 13C{1H} (100 MHz, CDCl3): δ butyl CR
1
1
18.92 (br, J (13C,119Sn) ) 483.4, J (13C,117Sn) ) 462.3 Hz), Câ
28.18 2J (13C,117,119Sn) ) 26.6 Hz), Cγ 27.32 3J (13C,117,119Sn) )
OM000397F