Amberlyst 15 is a supported Brønsted acid catalyst, and
has been previously used by us to produce the cyclic and
symmetrical ethers in scCO2.3,5 As the temperature was
increased in the reaction of 1 over this catalyst (>80 °C), the
initial product was methyl octyl carbonate (1c). The observation
of this product was expected, as the transesterification of DMC
with ethanol has previously been observed over acidic ion-
exchange resins.24 As the temperature was increased further
(>110 °C), the desired methyl ether, 1-methoxyoctane (1b) was
formed, reaching a maximum yield of around 30% at 140 °C.
At these high temperatures octenes (1e) and dioctyl ether (1f)
were also formed in small amounts respectively due to the
unimolecular and bimolecular dehydration of 1. However,
the thermal instability of Amberlyst 15 precludes increasing the
yield of 1b by further raising the temperature.
Amberlyst 70 and Nafion SAC-13 are also supported
Brønsted acid catalysts but possess greater thermal stability than
Amberlyst 15, thus permitting higher reaction temperatures.25,26
When using Amberlyst 70 and Nafion SAC-13 as catalysts,
similar effects of temperature on product composition were
observed; 1c being formed at lower temperatures, followed by
1b at increased temperatures with both catalysts having a
maximum yield of 54%, achieved at 168 °C with Amberlyst
70 and 191 °C with Nafion. However, the competing dehydra-
tion reaction to form 1e also occurs at these high temperatures
and becomes the dominant process as the temperature is
increased further. Additionally, at these elevated temperatures,
small amounts of 1f and dioctyl carbonate (1d), which is
produced by the transesterification of 1c with 1, were also
detected.
Zeolite H-beta is a protic zeolite related to the NaY and
NaX faujasites used by Selva et al.20 Again at lower temper-
atures the main product of the reaction is 1c followed by 1b at
higher temperature with a maximum yield of 56% at 200 °C.
Additionally, selectivity towards 1b was generally observed to
be poor with this catalyst, due to significant formation of 1f
and 1e at temperatures above 180 °C.
Base-activated alumina and acidic γ-alumina were studied
as a comparison to the basic and acidic aluminas used by Tundo
et al.19 Acidic γ-alumina is a well-known catalyst for the
formation of ethers Via the dehydration of alcohols,27-30 which
we have also demonstrated in a scCO2 environment.31 Whilst
both these aluminas are believed to be somewhat amphoteric
in nature, there is a clear difference in their performance in this
reaction. The basic alumina exhibited a marked selectivity for
the undesired 1c at ∼180 °C. However, high yields of 1b were
achieved at temperatures in excess of 220 °C, with a maximum
yield of 75% obtained at 270 °C, and an operating window of
>70% yield between 260 and 290 °C. By contrast, the acidic
γ-alumina exhibited a much higher selectivity toward the
formation of 1b over a larger temperature range than the basic
alternative. The acidic catalyst produced >80% yield of 1b over
a temperature window that extended from 220 to 280 °C, with
a maximum yield of 86% at 262 °C. In a separate experiment,
this catalyst showed no decline in activity over a period of 20 h
at a constant temperature of 250 °C.
As a result of these initial screening studies, the acidic
γ-alumina was chosen as the catalyst to be used in further MA
comparison studies, because of its wide temperature range, high
catalytic activity and good selectivity.
Comparison of the Performance of DMC, DME and
MeOH as Methylating Agents. The relative performance of
each of these MAs, were compared by mixing solutions of 1
and each MA (6:1 molar ratio of MA to 1) with scCO2 (11:1
molar ratio of CO2 to MA) and passing the resulting mixture
over the acidic γ-alumina. The molar ratio of 1 to MA was
raised from the 1:2 used in the screening studies because it was
known from previous work that greater ratios of MeOH were
needed to achieve etherification at significant yield. The reactor
temperature was gradually increased from 100 to 350 °C, while
the system pressure was maintained at a constant 100 bar. The
variations in product distribution with temperature for all three
reactions are shown in Figure 1. The key results, summarized
in Table 2, show that the formation of 1b proceeded in higher
yield with DMC than with either DME or MeOH, and that
DMC required significantly lower operating temperatures than
MeOH.
Furthermore, from Figure 1a, it can also be seen that the
undesired byproduct, 1c, was the major product when using
DMC at lower temperatures (<150 °C), presumably resulting
from the direct attack of 1 at the carbonyl group of DMC. At
higher temperatures (160-250 °C), the selectivity of the reaction
rapidly switched to favor the formation of the methyl ether,
1b, to a maximum yield of 96%. Figure 1b shows that, when
using DME as the MA, ether formation is favored over a much
narrower temperature range (170-275 °C) and the yield is
consistently lower (75%) than that of the comparable reaction
with DMC.
The methylation of 1 using MeOH, shown in Figure 1c,
required much higher catalyst bed temperatures than with either
DMC or DME. Significant ether formation was not observed
until the temperature was in excess of 280 °C. With all three
MAs, temperatures >300 °C led to dehydration of 1, resulting
in the production of the alkene 1e. This is consistent with
previous reports for the dehydration of 1 over alumina cata-
lysts.32 The dehydration reaction places an upper limit on the
temperature range over which etherification can be carried out.
Overall, our results show that DMC has a significantly wider
operating temperature range than either DME or MeOH due to
the enhanced reactivity of DMC at lower temperatures.
It was observed that during the reaction of 1 with DMC, a
proportion of DMC underwent catalytic thermal decomposition
to form DME and CO2, Scheme 2. This decomposition has been
(24) Zielinska-Nadolska, I.; Warmuzinski, K.; Richter, J. Inz. Chem.
Procesowa 2004, 25, 1861–1866.
(25) Siril, P. R.; Cross, H. E.; Brown, D. R. J. Mol. Catal. A: Chem. 2008,
279, 63–68.
(26) Harmer, M. A.; Sun, Q. Appl. Catal., A 2001, 221, 45–62.
(27) Sohlberg, K.; Pennycook, S. J.; Pantelides, S. T. Chem. Eng. Commun.
2000, 181, 107–135.
(28) Knözinger, H. Angew. Chem., Int. Ed. 1968, 7, 791–805.
(29) Shi, B. C.; Davis, B. H. J. Catal. 1995, 157, 359–367.
(30) Seo, C. W.; Jung, K. D.; Lee, K. Y.; Yoo, K. S. Ind. Eng. Chem. Res.
2008, 47, 6573–6578.
(32) Makgoba, N. P.; Sakuneka, T. M.; Koortzen, J. G.; van Schalkwyk,
C.; Botha, J. M.; Nicolaides, C. P. Appl. Catal., A. 2006, 297, 145–
150.
(31) Amandi, R.; Hyde, J. R.; Ross, S. K.; Lotz, T. J.; Poliakoff, M. Green
Chem. 2005, 7, 288–293.
Vol. 14, No. 2, 2010 / Organic Process Research & Development
•
413