Sulfonate-bonded tin complexes for the production of diphenyl carbonate

Article abstract of DOI:10.1016/j.jorganchem.2004.02.039

Various tin complexes including dibutyltin oxide and dibutyltin diacetate were tested for their activities in the transesterification between dimethyl carbonate (DMC) and phenol to produce diphenyl carbonate (DPC). The activities of tin complexes were significantly enhanced by the co-presence of alkyl or arylsulfonic acid, possibly due to the in situ formation of sulfonate-bonded tin complexes. Highly active triflate-bonded tin species, [Bu₂Sn(OH) (OTf)]₂ and [Bu₂Sn(OAc)(OTf)]₂, were isolated from the reaction of triflic acid with dibutyltin oxide and dibutyltin diacetate, and characterized by single crystal X-ray diffraction study.

Full text of DOI:10.1016/j.jorganchem.2004.02.039

Journal of Organometallic Chemistry 689 (2004) 1816–1820
www.elsevier.com/locate/jorganchem
Sulfonate-bonded tin complexes for the production of
diphenyl carbonate
Hyunjoo Lee, Jin Yong Bae, O-Sung Kwon, Sung Joon Kim, Sang Deuk Lee,
*
Hoon Sik Kim
Reaction Media Research Center, Korea Institute of Science and Technology, 39-1 Hawolgokdong, Seongbukgu, Seoul 136-791, Republic of Korea
Received 9 January 2004; accepted 26 February 2004
Abstract
Various tin complexes including dibutyltin oxide and dibutyltin diacetate were tested for their activities in the transesterification
between dimethyl carbonate (DMC) and phenol to produce diphenyl carbonate (DPC). The activities of tin complexes were sig-
nificantly enhanced by the co-presence of alkyl or arylsulfonic acid, possibly due to the in situ formation of sulfonate-bonded tin
complexes. Highly active triflate-bonded tin species, [Bu2Sn(OH)(OTf)]2 and [Bu2Sn(OAc)(OTf)]2, were isolated from the reaction
of triflic acid with dibutyltin oxide and dibutyltin diacetate, and characterized by single crystal X-ray diffraction study.
Ó 2004 Elsevier B.V. All rights reserved.
Keywords: Diphenyl carbonate; Transesterification; Sulfonato tin complexes; Dimethyl carbonate; Non-phosgene process
1. Introduction
Polycarbonate, one of the most important engineer-
ing thermoplastics, is being commercially produced by
reacting bisphenol-A with phosgene [1,2]. This conven-
ð1Þ
However, due to the thermodynamically unfavorable
equilibrium (K ¼ 3 ꢀ 10ꢁ4 at 453 K) toward DPC, the
reaction suffers from low yield and selectivity even at
elevated temperature [9]. Various tin complexes have
been applied as catalysts for the production of DPC by
the transesterification between DMC and phenol, but
their catalytic activities are not high enough for the in-
dustrial purpose [10,11].
We now report highly efficient catalytic system con-
sisting of an alkyl or arylsulfonic acid and a tin complex
for the transesterification between DMC and phenol as
well as the X-ray structural characterization of
[Bu2Sn(OH)(OTf)]2 and [Bu2Sn(OAc)(OTf)]2.
tional phosgenation process, however, has the serious
environmental problems such as the use of highly toxic
phosgene, the formation of a stoichiometric amount of
NaCl, and the use of a copious amount of methylene
chloride as a solvent. Accordingly, there have been tre-
mendous efforts to produce polycarbonate by using non-
phosgene methods [3–5].
The most practical non-phosgene process seems to be
the melt polymerization of bisphenol-A with diphenyl
carbonate (DPC) which can be prepared from step-wise
transesterification of dimethyl carbonate (DMC) with
phenol with the continuous removal of methanol as
shown in the following equation [6–8]:
2. Results and discussion
The activities of various tin compounds have been
tested for the transesterification between DMC and
phenol. The transesterification reactions were performed
* Corresponding author. Tel.: +82-2-958-5855; fax: +82-2-958-5859.
E-mail address: khs@kist.re.kr (H. Sik Kim).
0022-328X/$ - see front matter Ó 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2004.02.039

H. Lee et al. / Journal of Organometallic Chemistry 689 (2004) 1816–1820
1817
Table 1
Activities of various tin complexes for the transesterification of dimethyl carbonate with phenola
Entry
Catalyst
DMC conv. (%)
Yield (%)
DPC
MPC
Anisole
1
2
Bu2SnO
Bu2Sn(OAc)2
21.0
17.0
22.9
58.1
56.2
60.2
57.5
58.0
60.9
58.1
4.0
3.8
16.7
12.0
17.2
39.8
36.5
39.4
38.4
37.6
40.9
38.5
0.3
1.2
3
[Bu2Sn(OPh)]4(l3-O)2, 3
Bu2SnO/CF3SO3H
Bu2SnO/CH3SO3H
4.9
0.8
4
18.3
19.7
20.8
19.1
20.4
20.0
19.6
trace
trace
trace
0.1
5
6
Bu2SnO/CH3C6H4SO3H
Bu2Sn(OAc)/CF3SO3H
[Bu2Sn(OPh)]4(l3-O)2, 3/CF3SO3H
[Bu2Sn(OH)(OTf)]2, 1a
[Bu2Sn(OAc)(OTf)]2, 2
7
8
trace
trace
trace
9
10
a Dimethyl carbonate (40 mmol), phenol (200 mmol), catalyst (0.4 mmol based on Sn atom), sulfonic acid (0.4 mmol), benzene (40 ml), and
molecular sieves (30 g), 180 °C, t ¼ 3 h.
in the presence of molecular sieves (4A) at 180 °C for 3 h
in a 100-ml stainless steel reactor equipped with an
electrical heater.
As can be seen in Table 1, Bu2SnO gave low yields of
DPC and methyl phenyl carbonate (MPC) along with
the formation of a by-product, anisole. However, when
an alkyl or aryl sulfonic acid was used in conjunction
with dibutyltin oxide, the yields of DPC and MPC
greatly increased, while the formation of anisole de-
creased. Similarly, the activity of dibutyltin diacetate
also increased significantly by the presence of a sulfonic
acid. No appreciable changes in catalytic activities of
Fig. 1. Molecular structures of 1 (1a ꢂ 1b). The hydrogen atoms on
dibutyltin oxide and dibutyltin diacetate were observed
with the variation of substituent on the sulfonic acid.
ꢀ
carbon atoms are omitted for clarity. Selected bond lengths (A) and
angles (°): Sn(1)–O(4) 2.090(4), Sn(1)–O(4ꢃ) 2.120(4), Sn(1)–O(3)
2.492(5), Sn(2)–O(8) 2.064(4), Sn(2)–O(8ꢃ) 2.138(4), Sn(2)–O(7)
2.863(3), Sn(2)–O(9) 2.364(5), O(4)–Sn(1)–O(4ꢃ) 80.07(2), Sn(1)–O(4)–
Sn(1ꢃ) 109.59(2), O(8)–Sn(2)–O(8ꢃ) 71.11(2), Sn(2)–O(8)–Sn(2ꢃ)
108.89(2).
To have a better understanding of the role of sulfonic
acid, we have carried out a reaction of dibutyltin oxide
with an equimolar amount of triflic acid in CH2Cl2 for 2
h at an ambient temperature Eq. (2). Following addition
of n-hexane into the reaction mixture produced a white
solid, which turned out to be a highly active interme-
diate species for the transesterification reaction,
ð2Þ
1H, 13C,19F, and 119Sn NMR spectra show that the
white solid is a tin complex containing n-butyl and tri-
flate ligands. Due to the ambiguity of the spectroscopic
data of the complex, we have conducted a single crystal
X-ray diffraction study. Interestingly, as shown in Figs.
1 and 2, the X-ray diffraction analysis reveals that the
crystal unit cell contains mutually interacting two an-
hydrous and hydrated molecules, [Bu2Sn(OH)(OTf)]2,
1a and [Bu2Sn(OH)(OTf)(H2O)]2, 1b (Table 2). Each tin
atom in 1a and 1b is six-coordinated and adopts a dis-
torted octahedral geometry [12,13]. The distance be-
tween Sn and the triflato oxygen for the hydrated
Fig. 2. Packing diagram of the complexes 1 (1a ꢂ 1b) viewed along the c
axis of the unit cell. The butyl and trifluoromethyl groups have been
omitted for clarity. Intermolecular interactions between triflato oxygen
in complex 1a and tin atom in complex 1b are represented by dashed
line.

1818
H. Lee et al. / Journal of Organometallic Chemistry 689 (2004) 1816–1820
Table 2
Crystallographic data for compounds 1 and 2
1 (1a ꢂ 1b)
2
Molecular formula
(empirical formula)
Formula weight
Crystal system
Space group
C36H80F12O18S4Sn4
(C18H40F6O9S2Sn2)
816.06
C22H26F6O12S2Sn2
(C11H23F3O6SSn)
459.54
Triclinic
ꢁ
Triclinic
ꢁ
P1
P1
Z
4
4
Cell constants
ꢀ
a (A)
9.811(2)
11.995(2
14.628(3)
87.78(3)
72.63(3)
81.97(3)
1626.8(6)
1.736
8.963 (2)
9.120(2)
12.128(2)
105.69(3)
96.06(3)
100.10(3)
927.3(3)
1.538
ꢀ
b (A)
ꢀ
c (A)
a (°)
b (°)
c (°)
3
ꢀ
Fig. 3. Molecular structures of 2. The hydrogen atoms on carbon at-
ꢀ
Volume (A )
l (mmꢁ1
Dcal (g/cm3)
)
oms are omitted for clarity. Selected bond lengths (A) and angles (°):
Sn(1)–O(4) 2.402 (4), Sn(1)–O(5) 2.256(3), Sn(1)–O(4ꢃ) 2.612(4), Sn(1)–
O(1) 2.472(4), Sn(1)–O(6) 2.210(3), Sn(1)–O(4)–Sn(1ꢃ) 115.51(7), O(4)–
Sn(1)–O(1) 143.30(1), O(4)–Sn(1)–O(5) 55.13(1), C(4)–Sn(1)–C(8)
163.4(2).
1.666
51.8
6372
1.644
51.8
3625
2hmax (°)
No. of reflections
used in refinement
ðI > 2rðIÞÞ
No. of parameters
Residualsa: R1; wR2
Goodness-of-fit
338
0.05; 0.13
1.03
202
0.04; 0.11
0.75
The higher activity of 1a compared with dibutyltin
oxide can be largely ascribed to the presence of elec-
tronegative triflate ligands coordinated to Sn atoms. As
a result of a triflate bonding, the Sn atom becomes
highly electrophilic or Lewis acidic, thereby facilitating
the nucleophilic attack of DMC on Sn atom. On the
other hand, in the absence of triflic acid, dibutyltin oxide
easily reacts with phenol to produce a less active
and sterically crowded tetrameric tin complex,
[Bu2Sn(OPh)]4(l3-O)2, 3 [16],
P
2
1=2
GOF ¼ ½ h k lðwðjFo2j ꢁ jFc2ÞÞ =ðndata ꢁ nvariÞꢄ
:
P
P
P
P
2
2 1=2
a R1 ¼ ðkFojꢁjFckÞ= jFoj; wR2 ¼ ½ ½wðFo2ꢁFc2Þ ꢄ= ½wðFo2Þ ꢄꢄ
;
2
w ¼ 1=½r2ðFo2Þ þ ð0:0932PÞ þ 0:0000Pꢄ for 1 and 1=½r2ðFo2Þ þ ð0:1156
2
PÞ þ 2:1798Pꢄ for 2; where P ¼ ðFo2 þ 2Fo2Þ=3:
ꢀ
complex 1b is longer by 0.372 A in comparison with that
in the non-hydrated complex 1a, implying a significant
electron donation from H2O to the electrophilic Sn
atom. The anhydrous and hydrated tin complexes are
likely to form a polymeric structure through interaction
of triflate oxygen of hydrated complex 1b to the tin atom
of anhydrous complex 1a [14,15]. The interaction be-
tween two molecules can be clearly seen in the packing
diagram shown in Fig. 2, where triflato oxygen O(5) is
ꢀ
closely located to Sn(1) with the distance of 2.630(4) A.
ð3Þ
Triflate-bonded tin complex 2 was similarly prepared
by the reaction of the dibutyltin diacetate with triflic
acid in CH2Cl2 and characterized by 1H, 13C, 119Sn
NMR, and X-ray crystallography (Table 2). Complex 2
was crystallized as a hydrated form. The molecular
structure of 2 depicted in Fig. 3 reveals that Sn(1), O(4),
Sn(1ꢃ), and O(4ꢃ) constitute a square planar and one of
the oxygen atoms of the acetate ligand interacts in-
termolecularly with tin atom of the different molecule.
This is why one of the Sn–O bond lengths of acetate
substituent is slightly longer than the other (Sn(1)–
The catalytic activity of 3 also increased by the co-
presence of triflic acid (entry 8), possibly due to the
formation of an active intermediate species with triflate
ligands. The formation of triflate-bonded complex with
1
the liberation of phenol is evident from the H NMR
spectrum (not shown here). The increase in electrophi-
licity due to the triflate bonding is supported by the
comparison of 119Sn NMR spectral data for 1a and 3,
which show a singlet at )156.3 ppm for 1a and two
singlets at )178.5 and )177.8 ppm for 3. The two
singlets can be ascribed to the presence of the four- and
five-coordinated tins, respectively. The strong Lewis
acidic character of Sn atoms in complex 1a is obvious
from the coordination of water molecules to complex 1a
to give complex 1b. Based on the experimental and
spectroscopic results as well as the structural analysis, a
plausible mechanism of the transesterification with 1a is
ꢀ
ꢀ
O(4) ¼ 2.402(4) A and Sn(1)–O(5) ¼ 2.256(3) A).
As expected, complex 1a exhibited a similar activity
to the corresponding catalytic system consisting of
Bu2SnO and triflic acid. This result strongly suggests
that the transesterification reaction in the presence of
the catalytic system composed of Bu2SnO and triflic acid
is, indeed, catalyzed by in situ formed 1a.

H. Lee et al. / Journal of Organometallic Chemistry 689 (2004) 1816–1820
1819
to remove unreacted dibutyltin compounds. The addi-
tion of n-hexane into the resulting solution produced
triflate-bonded tin complex as a white solid.
3.2.1. Complex 1a
1
Yield: 95.4%; H NMR (300 MHz, CDCl3) d ¼ 0:85
(t, 6H; CH3), 1.34 (m, 4H; CH2), 1.71 (m, 8H; CH2CH2),
5.28 (s, OH); 13C NMR (CDCl3, 75 MHz) d ¼ 13:6
2
(CH3), 26.4 (CH2), 26.7 (CH2, J(13C–117=119Sn) ¼ 115/
1
120 Hz), 29.1 (CH2Sn, J(13C–117=119Sn) ¼ 322/336 Hz),
119.4 (CF3, 1J(13C–19F) ¼ 317.8); 19F NMR (external
reference CF3CO2H) d ¼ ꢁ1:63 (s, CF3); 119Sn NMR
(external reference SnMe4); d ¼ ꢁ156:26(s).
3.2.2. Complex 2
Yield: 95.1%; 1H NMR (300 MHz, CDCl3) d ¼ 0:99 (t,
6H; CH3), 1.45 (m, 4H; CH2), 1.80 (m, 4H; CH2), 2.02 (m,
4H; CH2Sn), 2.27 (m, 3H; CH3CO2); 13C NMR (CDCl3,
75 MHz): d ¼ 13:5 (CH3), 21.2 (CH3 CO2), 26.4 (CH2,
3
2J(13C–117=119Sn) ¼ 111/116 Hz, J(13C–117=119Sn) ¼ 36.9
Scheme 1. Plausible mechanism for the transesterification of DMC
with phenol in the presence of 1a.
Hz), 30.2 (CH2Sn, 1J(13C–117=119Sn) ¼ 384/401 Hz), 119.4
(CF3, 1J(13C–19SnF) ¼ 318), 185.2 (CH3CO2); 119Sn
NMR (111 MHz, CDCl3): d ¼ ꢁ165:7 (s).
proposed in Scheme 1. For clarity, only a monomeric
portion of 1a is depicted.
3.3. Transesterification reaction
Coordination of DMC to complex 1a is likely to
occur first to give species I. An attack of phenol on the
carbonyl carbon of the coordinated DMC followed by
the elimination of methanol would give III. The MPC-
coordinated species III either loses MPC to generate
complex 1a or reacts with an additional phenol to give
species IV. The subsequent elimination of methanol and
DPC from species IV would regenerate 1a.
All the transesterification reactions were conducted in
a 100-ml stainless-steel high-pressure reactor equipped
with an electrical heater and a 60-ml stainless steel col-
umn. The reactor was charged with dimethyl carbonate
(40 mmol), phenol (200 mmol), and benzene (40 ml) as a
solvent, an appropriate catalyst or a catalytic system,
and t-butyl benzene (2 g) as an internal standard. Mo-
lecular sieves (30 g) were placed in the 60-ml stainless-
steel column mounted on the lid of the reactor. The re-
actor was evacuated to remove air from the molecular
sieves and then heated to 180 °C at the rate of 10 °C/min.
After the reaction, the reactor was cooled to room tem-
perature and the resulting solution was analyzed by gas
chromatography (HP-6890) and gas chromatography–
mass spectroscopy (GS–MS, HP-6890N GC-5973MSD).
Effort to tune the catalytic activity of complex 1a by
modifying the ligand set is in progress.
3. Experimental
3.1. General
Dimethyl carbonate and phenol were purchased from
Aldrich Chemical Co. and distilled just before use. All
other chemicals were obtained from Aldrich Chemical
Co. and used as received. Catalysts were prepared under
Ar atmosphere. 1H, 13C, 19F, and 119Sn NMR mea-
surements were carried out using a Varian UNITYplus-
300. CF3CO2H and SnMe4 were used as external
references for 19F and 119Sn, respectively.
3.4. X-ray crystallography
Suitable crystals for 1 and 2 were obtained by slow
diffusion of hexane into a methylene chloride solution of
the complexes at ambient temperature. The crystals used
in data collections were glued onto the end of thin glass
fiber. X-ray intensity data were measured at 293 K on an
Enraf CAD-4 automated diffractometer with graphite-
3.2. Preparation of sulfonate-bonded tin complexes
ꢀ
monochromated Mo Ka radiation (k ¼ 0:7107 A). The
Triflic acid (10 mmol) was reacted with a dibutyltin
compound (10 mmol) in 50 ml CH2Cl2 in a 100-ml
round-bottomed flask at room temperature for 3 h.
After the reaction, the reaction mixture was filtered off
unit cells were determined by using search, center, index,
and least-squares routines. The intensity data were
corrected for Lorentz and polarization effects and for
anisotropic decay.

1820
H. Lee et al. / Journal of Organometallic Chemistry 689 (2004) 1816–1820
[2] S.P. Kim, J.-S. Lee, S.-H. Kim, B.-H. Lee, S.H. Kim, J. Ind. Eng.
Chem. 5 (1999) 268.
Both compounds 1 and 2 crystallized in the triclinic
ꢁ
system. The space group P1 was assumed and confirmed
by the successful solution and refinement of the struc-
ture. All structures were solved by the application of
direct methods using the SHELXL 86 [17] and were re-
fined by full-matrix least-squares on F 2 by using
SHELXL 93 [18]. For each structure, all non-hydrogen
atoms were refined with anisotropic displacement pa-
rameters. All hydrogen atoms were located and refined
using isotropic thermal parameters.
[3] Chem. Brit. 30 (1994) 970.
[4] D. Greitag, G. Fenglar, L. Morbitzer, Angew. Chem., Int. Ed.
Engl. 30 (1991) 1598.
[5] S.M. Gross, W.G. Bunyard, K. Erford, G.W. Roberts, D.J.
Kiserow, J.M. DeSimone, J. Polym. Sci., Part A: Polym. Chem. 40
(2002) 171.
[6] P.R. de Bruin, J.S. Law, V.A. Vriens, US Patent 6,294,684, 2001.
[7] M. Mizukami, Y. Arai, H. harada, T. Ohshida, H. Ohgi, US
Patent 5,980,445, 1999.
[8] M. Fuming, Li. Guangxing, N. Jin, X. Huibi, J. Mol. Catal. A:
Chem. 184 (2002) 465.
[9] P. Tundo, F. Trotta, G. Molaglio, F. Ligorati, Ind. Eng., Chem.
Res. 27 (1988) 1565.
4. Supplementary material
[10] A.G. Shaikh, S. Sivaram, Ind. Eng., Chem. Res. 31 (1992) 1167.
[11] Y. Kiso, M. Yuuichi, US Patent 5,034,557, 1991.
[12] K. Sakamoto, Y. Hamada, H. Akashi, A. Orita, J. Otera,
Organometallics 18 (1999) 3555.
Crystallographic data for the structural analysis have
been deposited with the Cambridge Crystallographic
Data Center, CCDC No. 203808 for compound 1 and
CCDC No. 226718 for 2. Copies of this data may be ob-
tained free of charge from The Director, CCDC, 12 Union
Road, Cambridge BD2 1EZ, UK (fax: +44-1223-336033;
e-mail: deposit@ccdc.cam.ac.uk or http://ccdc.cam.
ac.uk).
[13] K. Jurkschat, N. Pieper, S. Seemeyer, M. Schurmann, M.
Biesemans, I. Verbruggen, R. Willem, Organometalllics 20
(2001) 868.
[14] J. Beckmann, D. Kakternieks, A. Duthie, F.S. Kuan, Organmet-
allics 22 (2003) 4399.
[15] A. Orita, J. Xiang, K. Sakamoto, J. Otera, J. Organomet. Chem.
624 (2001) 287.
[16] W.J. Considine, J.J. Venturea, A.J. Gibbons, A. Ross, Can. J.
Chem. 41 (1963) 1239.
[17] G.M. Sheldrik, in: G.M. Sheldrik, C. Kruger, R. Goddard (Eds.),
Crystallographic Computing, vol. 3, Oxford University Press,
London, 1985, pp. 175–189.
References
[18] G.M. Sheldrick, in: H.D. Flack, L. Parkanyi, K. Simon (Eds.),
Crystallographic Computing, vol. 6, Oxford University Press,
London, 1993, pp. 111–122.
[1] D.C. Clagett, S.J. Shafer, in: G. Allen, J.C. Bevington (Eds.),
Comprehensive Polymer Science, vol. 5, Pergamon press, Oxford,
1989.