PREPARATION OF DIMETHYL CARBONATE FROM METHANOL AND CARBON DIOXIDE IN THE PRESENCE OF Sn(IV) and Ti(IV) ALKOXIDES AND METAL ACETATES

Article abstract of DOI:10.1135/cccc19950687

The synthesis of dimethyl carbonate by the reaction of methanol with carbon dioxide in the presence of metal alkoxides and metal carboxylates was studied.The best results have been achieved with Ti(IV) and Sn(IV) alkoxides which at 130 to 180 deg C and low CO2 pressures yield dimethyl carbonate in 30 to 100 mole percent or 40 to 130 mole percent yields with respect to the metal alkoxides, depending on the carbon dioxide (gaseous and solid one, respectively).The yields can be further increased up to 70-190 mole percent and 90-270 mole percent, respectively, by the use of chemical scavengers of the reaction water.

Full text of DOI:10.1135/cccc19950687

Short Communication
687
PREPARATION OF DIMETHYL CARBONATE FROM METHANOL AND
CARBON DIOXIDE IN THE PRESENCE OF Sn(IV) and Ti(IV) ALKOXIDES
AND METAL ACETATES
Juraj KIZLINK and Ivan PASTUCHA
Department of Organic Technology,
Slovak Technical University, 812 37 Bratislava, The Slovak Republic
Received January 12,1995
Accepted February 3,1995
The synthesis of dimethyl carbonate by the reaction of methanol with carbon dioxide in the presence
of metal alkoxides and metal carboxylates was studied. The best results have been achieved with
Ti(IV) and Sn(IV) alkoxides which at 130 to 180 °C and low CO2 pressures yield dimethyl carbonate
in 30 to 100 mole % or 40 to 130 mole % yields with respect to the metal alkoxides, depending on
the carbon dioxide (gaseous and solid one, respectively). The yields can be further increased up to 70 – 190
mole % and 90 – 270 mole %, respectively, by the use of chemical scavengers of the reaction water.
Dimethyl carbonate is an important methylating agent which replaces former toxic
methylating agents, and also an intermediate of synthesis of higher carbonates and fuel
additive. It is produced industrially by carbonylation of methanol in the presence of
copper or palladium catalysts1.
An alternative route is the reaction of carbon dioxide with methanol in the presence
of organotin compounds, especially dialkyltin dialkoxides2, as well as the use of
Sn(IV), Ti(IV) or Zr(IV) alkoxides3,4. In our previous studies we reported on the activ-
ity of other organotin compounds, using both solid5 and gaseous6 carbon dioxide. We
attempted at improving reaction conditions and product yields compared to the other
authors2–4, which, however, has been achieved only in the presence of chemical
scavengers of the reaction water6. This led us to examine further the activity of metal(IV)
alkoxides and metal carboxylates in this process which is not a typical catalytic reac-
tion5. The effectiveness of metal alkoxides for similar types of reactions was already
generally surveyed7. Of interest is also the use of mercury(II) acetate for synthesis of
dialkyl carbonates, even though the preparation of dimethyl carbonate has not been
specified8. In fact, many metal compounds of variable valence might be suitable agents
in this reaction. Unfortunately, the presence of reaction water reduces markedly their
effectiveness12.
Collect. Czech. Chem. Commun. (Vol. 60) (1995)

688
Kizlink, Pastucha:
EXPERIMENTAL
Chemicals
Dimethyl carbonate, 99% purity (Aldrich). Methanol (analytical grade purity, 99.8% and 0.2% H2O,
Lachema Brno) was dried over sodium sulfate and calcium hydride and distilled through a column
filled with sodium, yielding the product containing max. 0.01% H2O. Tin(IV) ethoxide was obtained9
as a white hygroscopic powder, titanium(IV) methoxide was prepared by reesterification of Ti(IV)
butoxide with methanol10,11 as a yellow powder melting at 202 °C. Titanium(IV) ethoxide was pre-
pared similarly10,11 and obtained as a yellow liquid boiling at 130 – 133 °C/400 Pa. Titanium(IV)
butoxide, a light brown liquid with b.p. 155 – 158 °C/400 Pa, was prepared by esterification of TiCl4
as reported10. Dicyclohexylcarbodiimide (DCC) and trimethyl phosphate (TMP) (both chemical
purity grade, min. 99%, Aldrich) were used as chemical scavengers of water. The other chemicals
were of analytical purity. Carbon dioxide was either the dry ice (water content max. 2.0%) or ga-
seous carbon dioxide (pure, min. 98.5%, max. 0.5% H2O, Tatragas Sala, The Slovak Republic).
Procedure
Solid carbon dioxide. A cooled 250 ml-stainless steel autoclave with frictional magnetic stirring
and electrical heating5 was charged with 150 ml (120 g, 3.8 mol) of cold methanol, a metal(IV)
alkoxide (10 mmol) and drying agent (10 g), followed by an initiator (5 mmol) and/or a water
scavenger (50 mmol). Then, a total of 44 g (1.0 mol) of dry ice were portionwise added. The auto-
clave was closed and its contents were heated to 150 °C for 6 to 12 h. The cooled reaction mixture
was filtered and analyzed by GLC. The results of these experiments are presented in Table I.
Gaseous carbon dioxide. A 150 ml-stainless steel autoclave equipped with magnetic stirring and
immersed in an electrically heated silicone bath6 was charged with 40 ml (30 g, 1.0 mol) of methanol, a
metal(IV) alkoxide (10 mmol), a drying agent (8 g) and eventually also with an initiator (5 mmol)
and/or a water scavenger (50 mmol). The reactor was closed, purged and then pressured with gaseous
carbon dioxide and heated to l50 °C for 20 h. The reaction mixture was worked up as described in
the preceding paragraph. The results obtained are summarized in Table II.
Analyses
GLC analyses were carried out on a Chrom-4 gas chromatograph equipped with CI-100 integrator
(Laboratorni pristroje, Prague) under the following conditions: FID detector, 2 m column (4 mm i.d.)
packed with 10% silicone elastomer SE 30 on Chromaton N-AW-DMCS (0.100 – 0.125 mm), column
temperature 60 °C, injection port temperature 200 °C (other suitable packings were found to be SE-31,
SE-301 or Reoplex-400), nitrogen as a carrier gas. Toluene was used as an internal reference. The
relative error of the GLC analysis was ±3 per cent.
RESULTS AND DISCUSSION
The results presented in Tables I and II document that the best metal compounds for the
title reaction are tin(IV) or titanium(IV) alkoxides, especially butoxides. It seems likely
that mechanism of the reaction is the same as in the presence of previously studied
organotin compounds4,5. These afforded dimethyl carbonate in up to 260 mole %
yields4, and with the use of chemical scavengers of reaction water even in up to 330
mole % yields6, most likely as the result of the better stability of the above tin com-
Collect. Czech. Chem. Commun. (Vol. 60) (1995)

Short Communication
689
TABLE I
Activity of metal(IV) alkoxides (10 mmol) in the reaction of solid carbon dioxide (1.0 mol) with
methanol (3.8 mol) in the presence of initiators (NaCl or Hg(OAc)2 5 mmol) and/or chemical
scavengers of reaction water (TMP or DCC, 50 mmol) at 150 °C (total pressure 6 MPa, reaction time 6 h)
Yield, %
Metal (IV)
Addenda
alkoxide
EtOH
AcOMeb
(MeO)2COc
Sn(OEt)4
none
60
85
68
62
67
62
63
60
0
0
0
86
111
82
none (200 °C)
NaBr
0
Hg(OAc)2
TMP
37
0
88
105
125
122
185
33
DCC
0
Hg(OAc)2 + TMP
NaBr + DCC
none
54
0
Ti(OMe)4
0
none (200 °C)
NaBr
0
0
58
0
0
53
Hg(OAc)2
Hg(OAc)2 + TMP
NaBr + DCC
none
0
44
42
0
48
0
75
0
88
Ti(OEt)4
Ti(OBu)4
66
68
65
70
70
67d
85d
64d
66d
63d
65d
68d
84d
70d
80d
0
50
NaBr
0
55
Hg(OAc)2
Hg(OAc)2 + TMP
NaBr + DCC
none
35
38
0
55
103
147
75
0
none (200 °C)
NaBr
0
105
76
0
Hg(OAc)2
TMP
0
78
0
90
DCC
0
108
113
137
175
190
Hg(OAc)2 + TMP
Hg(OAc)2 + TMP (200 °C)
NaBr + DCC
NaBr + DCC (200 °C)
43
58
0
0
a TMP trimethyl phosphate, DCC dicyclohexylcarbodiimide. b Expressed as mole % per mole of the
addend. c Calculated as mole % per mole of metal(IV) alkoxide. d The yield of butanol.
Collect. Czech. Chem. Commun. (Vol. 60) (1995)

690
Kizlink, Pastucha:
TABLE II
Activity of metal(IV) alkoxides (10 mmol) in the reaction of gaseous carbon dioxide (total pressure
2.5 MPa) with methanol (1 mol) in the presence of initiators (NaBr or Hg(OAc)2, 5 mmol) and/or
chemical scavengers of reaction water (TMP or DCC, 50 mmol) at 150 °C (reaction time 20 h)
Yield, %
Metal(IV)
Addenda
alkoxide
EtOH
AcOMeb
(MeO)2COc
Sn(OEt)4
none
80
93
80
80
80
80
83
86
0
0
0
150
163
160
156
166
220
175
233
40
none (180 °C)
NaBr
0
Hg(OAc)2
48
0
TMP
DCC
0
Hg(OAc)2 + TMP
NaBr + DCC
none
60
0
Ti(OMe)4
0
none (200 °C)
NaBr
0
0
68
0
0
48
Hg(OAc)2
0
55
50
0
48
Hg(OAc)2 + TMP
NaBr + DCC
none
0
94
0
111
63
Ti(OEt)4
Ti(OBu)4
82
90
77
70
80d
100d
85d
88d
88
80
95
100
90
100
0
none (180 °C)
Hg(OAc)2 + TMP
NaBr + DCC
none
0
126
123
183
105
140
122
133
144
200
165
188
237
268
53
0
0
none (180 °C)
NaBr
0
0
Hg(OAc)2
55
0
TMP
DCC
0
Hg(OAc)2 + TMP
Hg(OAc)2 + TMP (180 °C)
NaBr + DCC
NaBr + DCC (180 °C)
83
85
0
0
a TMP trimethyl phosphate, DCC dicyclohexylcarbodiimide. b Expressed as mole % per mole of the
addend. c Calculated as mole % per mole of metal(IV) alkoxide. d The yield of butanol.
Collect. Czech. Chem. Commun. (Vol. 60) (1995)

Short Communication
691
pounds. The previously used initiators5,6 such iodine, methyl iodide, mercury(II) iodide
and others turned out to be less suitable in the present case because of their promoting
effect on metal tetraalkoxide decomposition. The better agents are heavy metal (espe-
cially Hg) acetates or alkali metal halides such as NaBr, NaI etc. which cause a partial
metal tetraalkoxide decomposition mainly at higher temperatures12.
One can conclude that the reproducibility of the results of earlier studies2–4 is only
partial or nearly none (the yields of the carbonate did not exceed 20 mole % in the case
of metal acetates (Mn, Co, Sn, and Zn) as such, and 30 mole % with the use of the
mercury salt). The heavy metal acetates thus do not promote dimethyl carbonate pro-
duction but they undergo mostly decomposition, this being accompanied by the esteri-
fication of acetic acid to methyl acetate. In the absence of water scavengers, the yields
of dimethyl carbonate did not exceed 160 mole % with respect to the metal(IV) alkoxide,
because of the already mentioned negative effect of the reaction water. The higher
temperature up to 200 °C facilitates the reaction5 but this trend is not in all cases so
straightforward. Thus for example, at temperatures above 150 °C, the efficiency of
tin(IV) ethoxide does not further increases. It is supposed that the higher temperature
shows favourable effect on both the cleavage of the reaction intermediate to give di-
methyl carbonate and on the subsequent conversion of metal(IV) alkoxide to the metal
metoxide5. It is of interest that titanium(IV) methoxide is the less efficient agent com-
pared to the higher alkoxides, although the latter are transformed into the former by
reesterification with methanol during the studied carbonylation reaction. The use of a
drying agent turned out to be advantageous, the best one being MgSO4, the less suitable
agents then Na2SO4 and molecular sieves (Nalsit, Calsit and Potasit). The determination
of the water content in the reaction mixture after the reaction by Fisher method showed
that the inorganic drying agents ensure in average max. 0.02%, the chemical scavengers
such as trimethyl phosphate (TMP) max. 0.013% and dicyclohexylcarbodiimide (DCC)
so low as max. 0.003%. These data provide evidence for the decisive negative effect of
the reaction water on the yields of the reaction. The yields of dimethyl carbonate ob-
tained (70 – 270 mole %) are comparable to those reported in previous works2–4 (260
mole %), and in some respect even somewhat better as they were achieved after shorter
(ca by 4 h) reaction times. Nevertheless, because of the lower stability of metal(IV)
alkoxides, the dialkyltin(II) compounds studied earlier6 are the more convenient agents
for this purpose since they afford dimethyl carbonate in more than 300 mole per cent
yields.
REFERENCES
1. Romano U., Tesei R., Masuri M. M., Rebora F.: Ind. Eng. Chem., Prod. Res. Dev. 19, 396
(1980).
2. Sakai S., Fujinami T., Furusawa S.: Nippon Kagaku Kaishi 10, 1789 (1975); Chem. Abstr. 84,
5090 (1976).
Collect. Czech. Chem. Commun. (Vol. 60) (1995)

692
Kizlink, Pastucha:
3. Yamazaki N., Nakahama S., Endo K. (Mitsubishi Co.): Japan. Kokai 79 3,012; Chem. Abstr. 90,
168087 (1979).
4. Yamazaki N., Nakahama S.: Ind. Eng. Chem., Prod. Res. Dev. 18, 249 (1979).
5. Kizlink J.: Collect. Czech. Chem. Commun. 58, 1399 (1993).
6. Kizlink J., Pastucha I.: Collect. Czech. Chem. Commun. 59, 2116 (1994).
7. Bradley D. C.: Prog. Inorg. Chem. 2, 303 (1960).
8. Ota H. (Mitsai Toatsu): Japan 70 24,966; Chem. Abstr. 73, 98378 (1970).
9. Bradley D. C., Caldwell E. V., Wardlaw W.: J. Chem. Soc. 1957, 4775.
10. Bischoff F., Adkins H.: J. Am. Chem. Soc. 46, 256 (1924).
11. Bradley D. C., Hancock D. C., Wardlaw W.: J. Chem. Soc. 1952, 2773.
12. Sheldon R. A., Kochi J. K.: Metal-Catalyzed Oxidations of Organic Compounds. Academic
Press, New York 1981.
Collect. Czech. Chem. Commun. (Vol. 60) (1995)