The continuous acid-catalyzed dehydration of alcohols in supercritical fluids: A new approach to the cleaner synthesis of acetals, ketals, and ethers with high selectivity

Article abstract of DOI:10.1021/ja991562p

We report a new continuous method for forming ethers, acetals and ketals using solid acid catalysts, DELOXAN ASP or AMBERLYST 15, and supercritical fluid solvents. In the case of ether formation, we observe a high selectivity for linear alkyl ethers with little rearrangement to give branched ethers. Such rearrangement is common in conventional syntheses. Our approach is effective for a range of n-alcohols up to n-octanol and also for the secondary alcohol 2-propanol. In the reaction of phenol with an alkylating agent, the continuous reaction can be tuned to give preferential O- or C- alkylation with up to 49% O-alkylation with supercritical propene. We also investigate the synthesis of a range of cyclic ethers and show an improved method for the synthesis of THF from 1,4-butandiol under very mild conditions.

Full text of DOI:10.1021/ja991562p

J. Am. Chem. Soc. 1999, 121, 10711-10718
10711
The Continuous Acid-Catalyzed Dehydration of Alcohols in
Supercritical Fluids: A New Approach to the Cleaner Synthesis of
Acetals, Ketals, and Ethers with High Selectivity
William K. Gray,† Fiona R. Smail,† Martin G. Hitzler,† Stephen K. Ross,‡ and
Martyn Poliakoff*,†
Contribution from the School of Chemistry, UniVersity of Nottingham, UniVersity Park,
Nottingham, NG7 2RD, UK, and Thomas Swan and Co. Ltd., Crookhall, Consett,
Co. Durham, DH8 7ND, UK
ReceiVed May 10, 1999
Abstract: We report a new continuous method for forming ethers, acetals and ketals using solid acid catalysts,
DELOXAN ASP or AMBERLYST 15, and supercritical fluid solvents. In the case of ether formation, we
observe a high selectivity for linear alkyl ethers with little rearrangement to give branched ethers. Such
rearrangement is common in conventional syntheses. Our approach is effective for a range of n-alcohols up to
n-octanol and also for the secondary alcohol 2-propanol. In the reaction of phenol with an alkylating agent,
the continuous reaction can be tuned to give preferential O- or C-alkylation with up to 49% O-alkylation with
supercritical propene. We also investigate the synthesis of a range of cyclic ethers and show an improved
method for the synthesis of THF from 1,4-butandiol under very mild conditions.
Introduction
Scheme 1
In recent years there has been an increasing interest in
environmentally more sustainable chemical processes.1,2 This
has led to increased research into the use of supercritical fluids
(SCFs),3-5 especially carbon dioxide, scCO2, as an environ-
mentally benign and economically feasible reaction medium for
heterogeneously catalyzed reactions.3,4,6
SCFs are gases heated above their critical temperature (TC)
and compressed above their critical pressure (PC).7,8 CO2 has a
critical point close to ambient temperature (TC ) 31.3 °C, PC
) 72.9 atm)4 and is nonflammable, nontoxic, and inexpensive.
With extremely low viscosity and good thermal and mass
transport properties, SCFs are attractive solvents for continuous
reactions. Moreover, their properties can be tuned within certain
limits by adjusting pressure and temperature (and hence the
density). In a continuous flow reactor, such parameters can be
adjusted independently, giving very precise control over reaction
conditions. Furthermore, continuous reactors are generally
smaller and safer than batch reactors with equivalent production
capacity.9
10,11
We have previously reported
how SCFs, particularly
scCO2, can be used for the continuous hydrogenation of a wide
range of organic substrates using a fixed-bed reactor packed
with a supported noble metal catalyst (e.g. Pt, Pd, etc.). This
supercritical process combines cleaner processing with very high
conversions and high product selectivities. More recently, with
slightly modified equipment, we have carried out selective
Friedel-Crafts alkylation reactions using solid acid catalysts
and alcohols as the alkylating agents.12 In some of these
alkylation reactions, we observed small amounts of ethers as
byproducts, formed by condensation of two molecules of the
alcohol, Scheme 1. It was this observation which prompted the
study of etherification in scCO2, reported here.
† University of Nottingham. Internet: http://www.nottingham.ac.uk/
supercritical
‡ Thomas Swan and Co. Ltd.
(1) Anastas, P. T.; Williams, T. C. Green Chemistry: Designing
Chemistry for the EnVironment; Oxford University Press: Oxford, 1998.
(2) Anastas, P. T.; Warner, J. C. Green Chemistry: Theory and Practice;
Oxford University Press: Oxford, 1998.
(3) Savage, P. E. Chem. ReV. 1999, 99, 603.
(4) Baiker, A. Chem. ReV. 1999, 99, 453.
(5) Darr, J. A.; Poliakoff, M. Chem. ReV. 1999, 99, 495.
(6) Savage, P. E.; Gopalan, S.; Mizan, T. I.; Martino, C. J.; Brock, E. E.
AIChE J. 1995, 41, 1723.
(7) Eckert, C. A.; Knutson, B. L.; Debendetti, P. G. Nature 1996, 383,
313.
(8) McHugh, M. A.; Krukonis, V. J. Supercritical Fluid Extraction:
Principles and Practice; Butterworth-Heinemann: Boston, 1994.
(9) Tundo, P. Continuous Flow Methods in Organic Synthesis; Ellis
Horwood: Chichester, UK, 1991.
Acid-Catalyzed Reactions in the Supercritical Phase. A
greater number of fundamental catalytic organic reactions are
promoted by acid than all other catalytic reactions added
together.13,14 Solid acid catalysts such as clays, zeolites and even
(10) Hitzler, M. G.; Poliakoff, M. Chem. Commun. 1997, 1667.
(11) Hitzler, M. G.; Smail, F. R.; Ross, S. K.; Poliakoff, M. Org. Process
Res. DeV. 1998, 2, 137.
(12) Hitzler, M. G.; Smail, F. R.; Ross, S. K.; Poliakoff, M. Chem.
Commun. 1998, 359.
(13) Ertl, G.; Kno¨zinger, H.; Weitkamp, J. Handbook of Heterogenious
Catalysis; Wiley-VCH: Weinheim, 1997; Vol. 5.
10.1021/ja991562p CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/06/1999

10712 J. Am. Chem. Soc., Vol. 121, No. 46, 1999
Gray et al.
graphite are commonly used in such reactions, but rapid catalyst
deactivation and a lack of selectivity has meant that such
catalysts are far from ideal. In this paper, we describe the use
of solid acid catalysts for the continuous conversion of alcohols
to linear and branched ethers, to cyclic ethers, to aryl ethers
and to acetals and ketals in good yields with high selectivities.
Although the first report of an acid catalyzed reaction in an
SCF was over fifty years ago,15 it is only in the last 10 years
thatthenumberofpapersonthesubjecthasgrownsubstantially.16-22
Eckert16 has reported Friedel-Crafts alkylation of phenol in near
critical water, with the water itself acting as the acid catalyst.
There are a number of reports23,24 of the oligomerization of
supercritical isopentane or isobutane over a solid zeolite catalyst,
with the supercritical reactant also acting as the solvent. This
method is attractive because heterogeneous catalysts provide
an environmentally and economically acceptable alternative to
conventional homogeneous catalysts such as H2SO4 and HF.13
Furthermore, the catalyst lifetime can, in many cases, be
dramatically improved under supercritical conditions, due to
reduced coking.25,26 Long chain ester formation from the acid-
catalyzed reaction of oleic acid with methanol in scCO2 has
also been reported.18,19 Very recent work by Subramaniam,27
on the oligomerization of butane/1-butene, has shown that, using
scCO2 as a solvent (rather than supercritical isobutane/1-butene),
reactions can be performed at lower temperatures, with greater
selectivity for C8 oligomers, although coking is a major problem
at subcritical temperatures with respect to reactants28 ( ∼50 °C).
Fortunately, however, catalyst coking is not a serious problem
in the reactions reported in this paper.
Ether synthesis. There are a variety of methods for synthe-
sizing ethers, all of which can encounter problems on scale-up.
The most widely used method, the Williamson ether synthesis,29
in which an alkyl halide is reacted with an alkoxide or a
phenoxide, is generally high-yielding and clean, with a salt (e.g.,
NaCl) as a byproduct. However, it is far from an ideal synthetic
route because the most common methods of preparing alkyl
halides (e.g., alcohol substitution or the addition of hydrogen
halides or phosphorus polyhalides to alkenes) generate large
amounts of halogenated waste and solid salts. The reaction of
alcohols with inorganic esters (e.g., dimethyl sulfate) also gives
ethers, but the process again involves the disposal of aqueous
effluent containing large amounts of inorganic salts. A recent
communication by Strauss et al.30 described a new method for
the formation of symmetrical ethers from alcohols by use of
the corresponding alkyl bromide as a promoter. Although ethers
were formed in good yield, reaction times were up to 24 h in
some cases. Halogenated waste was still produced, albeit in
much reduced amounts, and this might cause problems on scale-
up.
One of the most cost-effective and atom efficient processes
for ether formation is the dehydration of alcohols with an acid
catalyst.31-34 However, this reaction often leads to a mixture
of products because the intermediate carbocation can rearrange
from the primary to the secondary and thence to the tertiary
position. In addition, the use of homogeneous acid catalysts,
such as H2SO4, requires a final separation step to recover or
neutralize the acid. Therefore, these reactions are usually carried
out in a batch or semi-batch reactor, and the long contact time
between catalyst and starting materials often leads to further
isomerization.
Supercritical Etherification. There are a few reports of
supercritical etherification. In 1995 Antal35 reported acid-
catalyzed formation of t-BuOEt from EtOH and t-BuOH at high
temperature and pressure. The reported yields and selectivities
were good, and interestingly, the ether could only be formed
from the reaction of EtOH with t-BuOH, but not from EtOH
and isobutene. Dehydration of alcohols to ethers and alkenes
in supercritical water (scH2O) has been studied by a number of
authors;17,20,36-41 all report good yields and high selectivities.
Some of the most important work in this area has been done
by Antal et al. They report the formation of Et2O from EtOH
in scH2O,20 as well as a number of other acid-catalyzed
dehydrations.39-41 Finally, work by Richter and Vogel42 reports
the batch dehydration and cyclization of 1,4-butanediol in
scH2O, at >374 °C to give tetrahydrofuran (THF) in quantitative
yield.
In this paper, we describe a continuous and selective general
route to ethers, using scCO2 as the solvent. We begin by
describing the reactor and then outline how it can be used to
generate cyclic ethers from diols. Particular attention is paid to
the cyclization of 1,4-butandiol to give THF. We then investigate
the condensation of a range of primary alcohols to give linear
ethers and show that relatively little isomerization to branched
products occurs. Finally we describe O-alkylation versus C-
alkylation selectivities in the reaction of phenol with i-PrOH
or propene.
Experimental Section
Safety Hazard. CAUTION! The experiments described in this paper
inVolVe the use of relatiVely high pressures and require equipment with
(14) Weissermel, K.; Arpe, H.-J. Industrial Organic Chemistry; Wiley-
VCH: New York, 1997.
(15) Patat, F. Monatsch. Chem. 1945, 77, 352.
(16) Chandler, K.; Deng, F.; Dillow, A. K.; Liotta, C. L.; Eckert, C. A.
Ind. Eng. Chem. Res. 1997, 36, 5175.
(17) Crittendon, R. C.; Parsons, E. J. Organometallics 1994, 13, 2587.
(18) Vieville, C.; Mouloungui, Z.; Gaset, A. 3rd International Symposium
on Supercritical Fluids; Strasbourg, France, 1994; Vol. 3, p 19.
(19) Vieville, C.; Mouloungui, Z.; Gaset, A. Ind. Eng. Chem. Res. 1993,
32, 2065.
(20) Xu, X.; De Almeida, C.; Antal, M. J. Ind. Eng. Chem. Res. 1991,
30, 1478.
(30) Bagell, L.; Cablewski, T.; Strauss, C. R. Chem. Commun. 1999,
283.
(31) Kotkar, D.; Ghosh, P. K. J. Chem. Soc., Chem. Commun. 1986,
650.
(32) Tagliavini, G.; Marton, D.; Furlani, D. Tetrahedron 1989, 45, 1187.
(33) Traynelis, V. J.; Hergenrother, W. L.; Hanson, H. T.; Valicenti, J.
A. J. Org. Chem. 1964, 29, 123.
(34) Costa, A.; Riego, J. M. Synth. Commun. 1987, 17, 1373.
(35) Habenicht, C.; Kam, L. C.; Wilschut, M. J.; Antal, M. J. Ind. Eng.
Chem. Res. 1995, 34, 3784.
(21) Boix, C.; Poliakoff, M. Tetrahedron Lett. 1999, 40, 4433.
(22) Boix, C.; Poliakoff, M. 6th International Symposium on Supercritical
Fluids; Nottingham, UK, 1999; p 247.
(36) Narayan, R.; Antal, M. J. J. Am. Chem. Soc. 1990, 112, 1927.
(37) Narayan, R.; Antal, M. J. Supercrit. Fluid Sci. Technol. ACS
Symposium Series 406; American Chemical Society: Washington, DC,
1989: p 226.
(23) Hussain, A. U.S. Patent 5,304,698, 1994.
(24) Fan, L.; Nakamura, I.; Ishida, S.; Fujimoto, K. Ind. Eng. Chem.
Res. 1997, 36, 1458.
(25) Madras, G.; Erkey, C.; Akgerman, A. Ind. Eng. Chem. Res. 1993,
32, 1163.
(38) Xu, X.; De Almeida, C.; Antal, M. J. J. Supercrit. Fluids 1990, 3,
228.
(39) Ramayya, S.; Brittain, A.; De Almeida, C.; Mok, W.; Antal, M. J.
Fuel 1987, 66, 1364.
(26) Baptist-Nguyen, S.; Subramaniam, B. AIChE J. 1992, 38, 1027.
(27) Clark, M. C.; Subramaniam, B. Ind. Eng. Chem. Res. 1998, 37,
1243.
(28) Corma, A.; Martinez, A.; Martinez, C. J. Catal. 1994, 146, 185.
(29) March, J. AdVanced Organic Chemistry: Reactions, Mechanisms
and Structure, 3rd ed.; John Wiley and Sons: New York, 1995.
(40) Xu, X. D.; Antal, M. J.; Anderson, D. G. M. Ind. Eng. Chem. Res.
1997, 36, 23.
(41) Antal, M. J.; Carlsson, M.; Xu, X.; Anderson, D. G. M. Ind. Eng.
Chem. Res. 1998, 37, 3820.
(42) Richter, T.; Vogel, H. International Meeting on High Pressure
Chemical Engineering; Karlsruhe, Germany, 1999; p 115.

Dehydration of Alcohols in Supercritical Fluids
J. Am. Chem. Soc., Vol. 121, No. 46, 1999 10713
Scheme 2
Table 1. Effect of Reactor Conditions on the Conversion of 3 into
a
4 in scCO2
product
4 (%)b 3 (%)
temperature pressure flow rate of 3
(°C)
(atm)
(mL/min)
varying T
varying P
100
125
150
175
200
100
100
100
100
150
150
150
150
150
100
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
1.0
2.0
3.0
63
96
100
100
100
69
79
81
87
100
100
100
98
37
4
0
0
0
31
21
19
13
0
0
0
Figure 1. Schematic view of the apparatus used for etherification.
The parts are labeled as follows (alphabetically): Cyl, cylinder of gas
(CO2, propane or propene); EF, excess flow cut-out valves; E1 and E2,
two expansion valves for pressure reduction and flow control; F, flow
meter; M, mixer; P, HPLC pump (0.3-20.0 mL/min, Gilson model
303); PP, pneumatic pump 20-200 atm (module PM 101, NWA
GmbH); Prod, product(s); R, reactor made from 316 stainless steel
tubing (9 mm i.d.), length 152 mm or 304 mm (10 mL or 20 mL volume
respectively), containing the solid catalyst; S, regulator which determines
the system pressure, Sub, organic substrate; T, trap for any droplets
inadvertently carried through with the gas (CO2, propane or propene).
100
100
100
100
125
150
175
200
60-200
varying
flow rate
of 3
100
100
100
100
2
7
93
the appropriate pressure rating. It is the responsibility of indiVidual
researchers to Verify that their particular apparatus meets the necessary
safety requirements. The components described below work well, but
they are not necessarily the only equipment of this type aVailable nor
the most suitable for the purpose.
a Other reaction conditions: DELOXAN ASP catalyst; 10 mL
reactor; flow rate of CO2 0.65 L/min. b In all tables presented in this
paper the symbol (%) next to a numerical identifier refers to the
percentage yield of that particular product, as found upon analysis of
the material collected.
Figure 1 shows the apparatus used in these experiments, which is
similar to that used for hydrogenations11 but without an H2 supply.
The gaseous solvent (CO2, propane, or propene) is pressurized using
the pneumatic pump, PP, and then reduced to system pressure across
valve S. The solvent enters the mixing area, M, where it is meets the
organic substrate, the mixture then passes through the reactor, R,
containing the catalyst. After the fluid passes through the reactor, the
pressure is dropped stepwise to separate the product(s) from the fluid
using an expansion module (PE 103, NWA GmbH). In this module,
the expansion valve E2 controls the flow rate of the exhaust gases.
Additionally, a flow meter, F, measures the flow rate of the fluid, which
is typically set to 0.65 L/min of gas at 1 atm and 20 °C, corresponding
to 1.06 g of CO2/min. The organic products are recovered free of any
solvent and thus they can immediately be analyzed by NMR (Bruker
DPX 300), GC (Philips PU 4500), or GC-MS (Shimadzu QP-5000)
without further workup.
R is heated by three cartridge-heaters in an aluminum block, which
encases the reactor. The heaters are thermostatically controlled by a
thermocouple reading the temperature of the catalyst bed. Two
additional thermocouples are used to monitor the system, one reading
the temperature of the Al heating block and the other in the fluid stream
at the end of the reactor. For all of the reactions described in this paper,
the temperature quoted is that of the catalyst bed.
in USA being used in the manufacture of THF.14 Convention-
ally, 3 is passed through a hot tube in the presence of H2SO4 or
H3PO4 at 100 atm and 300 °C, with 4 being generated in 90-
100% yield in a continuous process.
For our experiments 3 (mp 16 °C) was first dissolved in a
small amount of MeOH, (25 g 3 in 10 mL) to ensure that 3 did
not solidify in any of the pipework. The cyclization of 3 in
scCO2 is highly successful. Flow rates of up to 1.0 mL/min of
3 can be maintained for up to 15 h over the same catalyst with
no loss of activity. Table 1 shows the effect of temperature and
pressure on the yield of 4. Thus, increasing the reactor
temperature from 100 to 125 °C at 100 atm results in a 33%
increase in the yield of 4, and increasing it to 150 °C gives full
conversion. At 150 °C, varying the pressure inside the reactor
had no effect on conversion, even at only 60 atm (below the
critical pressure of CO2).43 However, at 100 °C, a clear
relationship between reactor pressure and yield can be seen,
with an 18% increase in conversion when the pressure was
raised from 100 to 200 atm. Such an effect is to be expected,
considering that, in fixed bed flow reactors and under these
conditions, an increase in pressure, at constant temperature,
increases the residence time of the substrate over the catalyst.
It should be noted that under steady state conditions the weight
hourly space velocity of material passing through our apparatus
will remain the same irrespective of the system pressure.
However, the residence time over the catalyst will be affected
by pressure because of the high compressibility of the super-
critical fluid.
We have used two types of solid acid catalyst in the experiments
described here. DELOXAN ASP catalyst (Degussa-Huls AG, Frankfurt,
Germany) has in the past been found12 to be a reliable and convenient
catalyst for use in our continuous flow reactor. During the course of
our study, we have also found that AMBERLYST resin-based acid
catalysts (Rohm and Haas Co.) can also be effective. All reactions
described in this paper were run with commercially available substrates
(Aldrich, Fluka, Acros). The apparatus used in these experiments gave
a high degree of reproducibility.
Etherification of 1,5-Pentanediol (5) and 1,6-Hexanediol
(8). 5 and 8 can be dehydrated with very high selectivity for
tetrahydropyran, 6, and the oxepane ring, 9, respectively (see
Table 2). This is a striking result since one would generally
expect the rearranged 2-alkyltetrahydrofurans, 7 and 10, to
Results and Discussion
Cyclic Ether Formation. Synthesis of Tetrahydrofuran (4)
from 1,4-Butanediol (3). Our preliminary investigations into
ether formation have been carried out using the cyclization of
1,4-butanediol, 3, to THF, 4, as a model reaction, because it is
both facile and clean and has no unwanted byproducts (see
Scheme 2). Furthermore, the cyclization of 3 is also a significant
industrial route to 4, with more than 50% of 3 produced annually
(43) 60 atm is above the vapor pressure of CO2 at room temperature.
Thus, under these conditions liquid CO2 will mix with the reactant(s) in
the cool part of the apparatus upstream of the reactor. This CO2/reactant
mixture is likely to have lower viscosity than the pure reactants themselves.

10714 J. Am. Chem. Soc., Vol. 121, No. 46, 1999
Gray et al.
a
a
Table 2. Formation of Cyclic Ethers from R,ω-diols in scCO2
Table 3. Continuous Formation of Acetals and Ketals in scCO2
a Other reaction conditions: DELOXAN ASP catalyst, 10 mL
reactor; 200 atm; flow rates - CO2 0.65 L/min, substrate 0.5 mL/min.
1 All material not accounted for was recovered as unreacted starting
material. 2 Yields based on 16 consumed; ratio of 16:17 or 19 was 1:2,
respectively.
in the NMR spectra of our product samples, which suggests
that, if Me2OH forms at all, it is not formed in sufficient
quantities to deplete MeOH significantly as a potential reactant
in these experiments.
Di(ethylene Glycol) (13) and Ethylene Glycol (16). 13 gave
a surprisingly clean and high-yielding reaction with excellent
selectivity for 1,4-dioxane 14. However, the formation of small
quantities of acetal, 15, as a side product prompted us to
investigate the supercritical formation of acetals and ketals, see
below. By contrast, the reaction of ethylene glycol, 16, gave
only decomposition products, probably resulting from partial
dehydration and rearrangement of 16 to acetaldehyde. The
product mixture would then be further complicated by the
reaction of the acetaldehyde with 16 to give an acetal. Clearly,
the failure to form the first ether linkage (i.e., 16 is not converted
to 13) is the reason 16 does not cyclize cleanly.
Acetals and Ketals. The most widely used method for
generating acetals and ketals is the reaction of alcohols with
aldehydes or ketones under acidic conditions, with removal of
water. This is precisely the route used in our supercritical reactor.
Thus, we have studied the reaction of 16 with acetone, 17, and
benzaldehyde, 19, as simple examples of ketal and acetal
formation respectively (Table 3).
Usually, a major problem associated with acetal and ketal
synthesis is that the reaction is an equilibrium process; H2O
must be removed to drive the equilibrium toward products.50
However, the yields in our reactor are remarkably high, probably
for two reasons. First, the relatively low solubility of H2O in
scCO2, compared to that of the starting materials and organic
products,51,52 means that H2O is effectively removed from the
reactant/product mixture, thus driving the equilibrium toward
products. Second, a number of studies have shown that high
pressure favors the formation of ketals, and to a lesser extent
of acetals, from the reaction of alcohols with carbonyl com-
pounds. Indeed, it has been claimed that this is a general effect
in all bimolecular acid-catalyzed reactions, due to the negative
volume of activation experienced by reactions proceeding via
an ionic intermediate.53,54
a Other reaction conditions: DELOXAN ASP catalyst; 10 mL reactor;
200 atm pressure; flow rates - CO2 0.65 L/min, substrate 0.5 mL/
min. 1 All material not accounted for was recovered as unreacted starting
material. 2 Yields based on 8 consumed.
predominate.29 5 and 8 are both solids at room temperature and
thus were dissolved in CH2Cl2 (25 g in 30 mL) for these
reactions.
By contrast, quite a different reaction is observed when 8 is
dissolved in MeOH rather than CH2Cl2. The reaction of 8 with
excess MeOH gives high selectivity for the linear monometh-
ylether product 11. Possibly the preferential formation of 11,
rather than 12, is due to 8 being more polar than 11. scCO2 is
a relatively nonpolar solvent, comparable to cyclohexane. In
such a medium, polar species, such as the reactants, intermedi-
ates, and catalyst in our reactor, will be forced together with an
increased chance of reaction. Such effects are well documented
and have been used to achieve high ortho selectivity in the
Friedel-Crafts alkylation of phenol.44 Clark and co-workers45
have reported a similar solvent effect in the base-catalyzed
Knoevenagel condensation, where the choice of solvent has a
marked effect on the rate of reaction, with a general order:
cyclohexane > toluene > 1,2-dichloroethane > chlorobenzene.
It therefore seems likely that the more polar diol, 8, is
preferentially adsorbed onto the ionic catalyst surface and thus
reacts readily, whereas the less polar mono-ether, 11, is resistant
to further etherification.46,47 Hydrogen bonding between 8 and
MeOH may also be an important factor in these reactions.
Unfortunately, we were unable to establish whether Me2O
was formed in our reactor since it is too volatile (bp -24.8 °C)
to be trapped in the recovery system. Thus, formation of Me2O
may be occurring in this reaction, and in the others discussed
below, where MeOH is used as a reactant under acidic
conditions.48,49 However, MeOH is detected in large quantities
Reaction of Primary Alcohols. Table 4 shows that a range
of n-alkyl ethers can be selectively formed in good yield from
(44) Sartori, G.; Maggi, R.; Bigi, F.; Arienti, A.; Porta, C.; Predieri, G.
Tetrahedron 1994, 50, 10587.
(45) Clark, J. H.; MacQuarrie, D. J. Chem. Commun. 1998, 853.
(46) Nakagawa, K.; Okamura, M.; Ikenaga, N.; Suzuki, T.; Kobayashi,
T. Chem. Commun. 1998, 1025.
(47) Nishiguchi, T.; Kawamine, K. J. Chem. Soc., Chem. Commun. 1990,
1766.
(48) Olah, G. A.; Iyer, P. S.; Prakash, G. K. S. Synthesis 1986, 7, 513.
(49) Kaspi, J.; Olah, G. A. J. Org. Chem. 1978, 43, 3142.
(50) Meskens, F. A. J. Synthesis 1981, 501.
(51) Francis, A. W. J. Phys. Chem. 1954, 58, 9.
(52) Mesiano, A. J.; Beckman, E. J.; Russell, A. J. Chem. ReV. 1999,
99, 623.
(53) Matsumoto, K.; Sera, A.; Uchida, T. Synthesis 1985, 1.
(54) Dauben, W. G.; Gerdes, J. M.; Look, G. C. J. Org. Chem. 1986,
51, 4964.

Dehydration of Alcohols in Supercritical Fluids
J. Am. Chem. Soc., Vol. 121, No. 46, 1999 10715
a
Table 4. Continuous Formation of Acyclic Ethers in scCO2
the corresponding n-alcohols over the DELOXAN ASP catalyst
in scCO2. For all alcohols, the yields in a 20 mL reactor were
significantly higher than those in a 10 mL vessel, with little
effect on selectivity.
productsb,c
alkene/
rearranged
alcohol (%)
T
(°C)
reactor
size (mL)
(n-Pr)2O
22 (%)
n-PrOi-Pr
23 (%)
substrate
Poor yields were obtained for n-octanol, 30, which is
dehydrated readily under these conditions to give octene,
probably as a result of the reduced solubility of long chain
alcohols in scCO2.51 The reaction between n-BuOH, 24, and
MeOH is interesting because, even when a 2.5:1 excess of
MeOH is used, n-Bu2O, 25, and n-BuOMe, 33, are produced
in comparable amounts, rather than the expected large excess
of 33. This suggests either that MeOH must be forming its
corresponding cation less readily than 24 or that 33 may possibly
be formed via trans-etherification of Me2O with 24.
200
150
10
20
29
60
4
8
5
n-PrOH
21
10
alkene/
rearranged
alcohol (%)
(n-Bu)2O
25 (%)
n-BuOs-Bu
26 (%)
substrate
n-BuOH
24
200
10
33
1
5
alkene/
(n-C6H13)2O n-C6H13Os-C6H13 rearranged
substrate
28 (%)
29 (%)
alcohol (%)
In all of the reactions of n-alcohols, we observed some alkene
production, resulting from the competing dehydration process,
with significant amounts only occurring when high temperatures
were used (typically above 200 °C in the 10 mL reactor).
Usually the predominant olefin product is the corresponding
2-alkene, although some 1-alkene is observed. Production of
these olefins causes the yield of, and the selectivity for, the
n-ether to fall because Markovnikov protonation of the olefin
will result in formation of a secondary carbocation, leading to
branched ethers. For this reason, it is important to maintain a
balanced temperature. Pressure effects in these reactions are
similar to those already observed for the formation of cyclic
ethers, acetals, and ketals and is likely to be due to the same
reasons (i.e., longer residence time and the negative volume of
activation for bimolecular reactions).
Origin of the High Selectivity for n-Ethers: Solvent
Effects. To establish whether scCO2 itself affects the selectivity,
a number of experiments were carried out in supercritical
propane (scPropane; TC ) 96.6 °C, PC ) 41.9 atm).4 The results,
Table 5, suggest that scCO2 is a better solvent than scPropane
for these reactions, although it is not yet clear why. Possibly,
the ether products are more soluble in scCO2 than in scPropane
and, thus are more efficiently separated from H2O, thereby
avoiding a back reaction.51 It should be noted that in a flow
reactor it is difficult to make precise comparisons between two
supercritical fluids because some of the reaction parameters will
usually be different. For example, at 200 °C and 200 atm the
density of CO2 (0.2629 g/mL) is lower than that of propane
(0.3145 g/mL) at the same temperature and pressure.55 To match
the fluid densities inside the reactor, the pressure during the
scPropane experiment could be lowered, but this might decrease
the residence time of the reactants inside the reactor, making a
quantitative comparison between CO2 and propane more dif-
ficult.
200 10
150 20
34
54
1
8
11
16
n-C6H13OH
27
alkene/
(n-C8H15)2O n-C8H15Os-C8H15 rearranged
31 (%)
substrate
32 (%)
alcohol (%)
150 10
200 10
14
0
0
0
10
95
n-C8H15OH
30
(n-Bu)2O n-BuOs-Bu n-Bu-O-Me
substrate
25 (%)
26 (%)
33 (%)
n-BuOH + MeOH 200 10
24
1
27
24d
1:2.5
alkene/
(n-Bu)2O n-BuOn-Pr rearranged
substrate
25 (%)
34 (%)
alcohol (%)
n-BuOH + n-PrOH 200 10
20
28
9
24e+ 21
1:1.2
alkene/
rearranged
alcohol (%)
(n-Pr)2O
22 (%)
n-PrOi-Pr
23 (%)
substrate
n-PrOH +
i-PrOH
21f+ 1
1:1.2
150
200
20
12
41
7
PhCH2OMe
36 (%)
p-PhCH2-C6H4-CH2OH
37 (%)
substrate
PhCH2O +
MeOH
35g
10
84
7
1:2.5
(Ph(CH2)2)2O Ph(CH2)2-O-CH(CH3)Ph
substrate
39 (%)
40 (%)
Ph(CH2)2OH 200 10
58
2
38
alkene/
rearranged alcohol
(%)
(i-Bu)2O
42 (%)
i-BuOt-Bu
43 (%)
substrate
The supercritical solvent appears to play a definite role in
determining the yield, but not the selectivity. Thus when
n-PrOH, 21, was passed over the catalyst in the flow reactor at
200 °C and 200 atm without any supercritical fluid, it gave
predominantly the n-Pr2O, although in a yield (7% 22:4% 23)
much lower than that with scCO2. Similarly, when n-hexanol,
27, was refluxed ( ∼156 °C) over DELOXAN ASP in the
absence of any solvent (but with the possibility of H2O venting
off), n-hexyl ether, 28, (12.5% after 2 h) was formed much more
slowly than in our flow reactor, but with only 1% of the
branched ether, 29. This selectivity is similar to that obtained
in the flow reactor.
i-BuOH
200
10
8
5
5
41
CH2dCHCH2-
(CH2dCHCH2)2O CH(CH3)CHO
substrate
45 (%)
46 (%)
60 10
120 10
1
9
0
CH2dCHCH2OH
44
1h
a Other reaction conditions: DELOXAN ASP catalyst; 200 atm; flow
rates - CO2 0.65 L/min, substrate 0.5 mL/min. b All material not
accounted for was recovered as unreacted starting material. c Yields
of volatile alkenes (propene, butene, etc.) were calculated by mass
balance. d Yields based on 24 consumed; 5% alkene/rearranged alcohol
products also formed. e Yields based on 24 consumed, yield of 22
(formed as a result of self-condensation of 21) was 13%, based on 21
consumed, 4% mixed branched ether products also formed. f Yields
based on 21 consumed; yield of 2 (formed as a result of self-
condensation of 1) was 6%, based on 1 consumed. g Yields based on
35 consumed. h 2% propionaldehyde 47 also formed.
Effect of Catalyst. Comparison of DELOXAN ASP with
AMBERLYST 15. DELOXAN ASP was used as the catalyst
(55) NIST Standard Reference Database 12, NIST Thermophysical
Properties of Pure Fluid Version 3.0.

10716 J. Am. Chem. Soc., Vol. 121, No. 46, 1999
Gray et al.
Table 5. Comparison of scCO2 and scPropane as Solvents for
Scheme 3
Etherificationa
productsb,c
alkene/
rearranged
alcohol
(%)
(n-Bu)2O n-BuOs-Bu
substrate T (°C)
solvent
25 (%)
26 (%)
n-BuOH
24
200
200
scCO2
scPropane
33
24
1
2
5
5
Table 7. Effect of Flow Rate of n-PrOH, 21, on Linear/Branched
Selectivity of the Etherificationa
alkene/
productsb,c
n-C6H13Os- rearranged
(n-C6H13)2O
28 (%)
C6H13
29 (%)
alcohol
(%)
flow rate
of 21
(mL/min)
(n-Pr)2O
22
(%)
n-PrOi-Pr
23
i-PrOH
1
(%)
substrate
solvent
propene
(%)
(%)
n-C6H13OH 150 scCO2
27
150 scPropane
14
3
0
0
10
0
0.2
0.5
1.0
2.0
3.0
5.0
38
28
15
15
12
9
5.0
4.0
2.3
2.0
2.0
1.5
2.0
1.0
0.5
0.3
0.2
0.1
5.2
5.1
4.8
4.9
4.8
4.7
a Other reaction conditions: DELOXAN ASP catalyst; 200 atm; flow
rates - CO2 0.65 L/min, substrate 0.5 mL/min. b All material not
accounted for was recovered as unreacted starting material. c Yields
of butene were calculated by mass balance.
Table 6. Comparison of DELOXAN ASP and AMBERLYST 15
a Other reaction conditions: DELOXAN ASP catalyst; 200 °C; 200
atm; 10 mL reactor, flow rates of CO2 0.65 L/min. b All material not
accounted for was recovered as unreacted starting material. c Yields
of propene were calculated by mass balance.
as Catalysts for Etherificationa
productsb,c
alkene/
rearranged
alcohol
(%)
molecule of MeOH weakened, and therefore polarized, the C-O
bond and thus promoted nucleophilic attack from a second
MeOH molecule (see Scheme 3). If such a mechanism were in
operation in our reactor it would lead to predominantly n-ether
products.
(n-Pr)2O n-PrOi-Pr
substrate T (°C)
catalyst
22 (%)
23 (%)
200 DELOXAN
29
4
5
n-PrOH
21
ASP
175 AMBERLYST
15
42
3
1
Effect of Alcohol Flow Rate on Linear/Branched Selectivi-
ties. Makarova et al,57 observed that the concentration of alcohol
around the acidic sites of the catalyst was crucial in determining
the selectivity for n-ethers in the dehydration of n-BuOH, 24,
over zeolites. They found that reducing the concentration of 24
in their reactor dramatically increased the ratio of butene and
mixed ethers to n-ethers in the product mixture. By contrast,
Table 7 shows that there is little or no difference in selectivity
for linear ether formation when the flow rate, and hence the
concentration, of 21 is varied in our reactor (with a constant
flow rate of scCO2). The increasing yield of i-PrOH, 1, with
decreasing flow rate suggests that the concentration of 21 may
be a limiting factor at the very lowest flow rates; the cation is
still rearranging, but there is insufficient 21 to react with the
cation so that it is re-attacked by H2O, giving 1.
alkene/
rearranged
(n-Bu)2O n-BuOs-Bu alcohol
substrate T (°C)
catalyst
25 (%)
26 (%)
(%)
200 DELOXAN
33
1
5
n-BuOH
ASP
175 AMBERLYST
15
24
47
6
2
a Other reaction conditions: 200 atm, 10 mL reactor, flow rates -
CO2 0.65 L/min, substrate 0.5 mL/min. b All material not accounted
for was recovered as unreacted starting material. c Yields of propene
and butene were calculated by mass balance.
in all of the reactions listed in Tables 1-5 and has been found
to be an excellent catalyst, mainly due to its thermal stability
and high acidity. Table 6 summarizes the results of experiments
carried out to compare the ion-exchange resin AMBERLYST
15 with DELOXAN.
It can be seen from Table 6 that there are only small
differences between the product yields obtained with the two
catalysts. AMBERLYST 15 is an excellent catalyst for these
reactions, although it is not as thermally stable as DELOXAN
ASP. Whereas DELOXAN ASP performs well at temperatures
up to 350 °C, AMBERLYST 15 rapidly deactivates above 225
°C. Furthermore, large quantities of alkenes were produced
above 175 °C with AMBERLYST 15, at least 25 °C lower than
that for DELOXAN ASP.
Although it is clear that DELOXAN ASP itself is not having
any specific effect on the linear/branched ratio, it is still possible
that the acid strength of the two catalysts is the controlling
influence on the selectivities. Theoretical studies by Sandre´ et
al on the condensation of MeOH to Me2O in a zeolite pore
concluded that the transition state involved direct condensation
of two molecules of MeOH, with the zeolite acting only as an
acidic medium.56 Interaction between the zeolite surface and a
All of our reactions show that, under these conditions, there
is a clear preference for formation of n-terminal products over
their branched analogues. Products resulting from the rearrange-
ment of the intermediate cation to a more stable secondary
position are only seen in very low yield. The literature suggests
that rearrangement of an intermediate cation can be avoided by
using a strong acid catalyst. Work by Olah58 on cyclic ether
formation suggests that an exceptionally strong acid will favor
formation of n-terminal products because strong acids can
stabilize the protonated alcohol intermediate, allowing approach
of the nucleophile prior to H2O elimination. In this context we
suggest that the high selectivity for linear ethers reported by
Strauss et al in their process30 may be due to the presence of
HBr, a strong acid catalyst. Possibly, the use of relatively
nonpolar solvents (e.g., scCO2) also increases ionic interactions
between polar molecules and hence the effective strength of
(56) Sandre´, E.; Payne, M. C.; Gale, J. D. Chem. Commun. 1998, 2445.
(57) Makarova, M. A.; Paukshtis, E. A.; Thomas, J. M.; Williams, C.;
Zamaraev, K. I. J. Catal. 1994, 149, 36.
(58) Olah, G. A.; Funo, A. P.; Malhotra, R. Synthesis 1981, 474.

Dehydration of Alcohols in Supercritical Fluids
J. Am. Chem. Soc., Vol. 121, No. 46, 1999 10717
Scheme 4
Scheme 6
Aryl Ethers. We have previously reported that the Friedel-
Crafts alkylation of mesitylene and anisole can be carried out
with 1 in our flow reactor.12 When phenol, 49, is used as the
substrate, competing etherification and alkylation reaction should
occur, as shown in Scheme 6. Conventionally, selective etheri-
fication of phenol (O-alkylation) can be difficult to achieve.
Although many authors have studied such reactions,48,49,61-72
there is still a degree of confusion over exactly which factors
favor O-alkylation versus C-alkylation. Although it is agreed
that increasing reaction temperatures favor C-alkylation because
of the thermodynamics of the process, the role of the catalyst
is much less clearly defined. It has been claimed that the
presence of strongly acidic sites on metal oxides, phosphates,
and sulfates favors O-alkylation and, conversely, that weakly
acidic sites favor C-alkylation. Zeolites also seem to have a
similar effect, although Marczewski64 has observed that strong
acid sites on zeolite USHY favor both O- and C-alkylation to
give cresols, while weak acid sites give only O-alkylation. More
recently61 a high potassium content (3%) in sulfates has been
claimed to give aryl ethers in excellent yields and selectivities
(>90%).
We have studied the alkylation of 49 with 1 in scCO2, and
with scPropene, at a variety of temperatures and pressures. When
1 was used as the alkylating agent, a 3:1 molar excess of 1:49
was used. When scPropene acting both as solvent and reagent
was used, a 230:1 excess of propene was present. The results
(Table 8 and Figure 2) indicate that reaction conditions can be
manipulated to give a good degree of selectivity for etherifi-
cation over Friedel-Crafts alkylation.
When 1 was used as the alkylating agent, the yield of i-PrOPh
53 reached a maximum of 32% at 160 °C and 160 atm. At this
point the combined total of the Friedel-Crafts products 50, 51,
and 52 was 13%. At 200 °C the yield of 53 is very low, and
the dialkylated product predominates. In the Friedel-Crafts
alkylation process, there is a high degree of selectivity for ortho-
Scheme 5
the acid/substrate interaction.45,59,60 Finally, although we observe
no carbonate species resulting from reaction of CO2 with the
alcohol in the product mixture, we cannot rule out the formation
of such species as transient intermediates in the reactions carried
out in scCO2.
2-Phenyl-1-ethanol (38) and Isobutanol (41). Table 4 also
shows that there is a high selectivity (12:1) for the n-terminal
product, 39, in the etherification of 38. This selectivity is striking
because the rearranged benzylic cation, 38b, is stabilized by
the adjacent aromatic ring and thus is much more stable than
the secondary cation, 38a, and hence, rapid rearrangement would
be expected under normal reaction conditions, see Scheme 4.
Tertiary cations are, in turn, more stable than benzylic cations,
and this may explain why i-BuOH (2-methylpropanol), 41, gave
a much lower (8:5) n-ether selectivity, with rapid formation of
the rearranged tert-butyl cation, 41b, and consequent formation
of ether 43, see Table 4 and Scheme 4.
Allyl Alcohol (44). Table 4 shows that the etherification of
44 was an unsuccessful reaction, even in scCO2. In this reaction,
propanal, 47, is formed by a simple rearrangement of the double
bond. The other aldehyde product, 46, arises from rearrangement
of one of the double bonds in the ether, 45, followed by a
Claisen-type sigmatropic rearrangement (see Scheme 5).
Reaction of Secondary Alcohols. i-PrOH, 1, is the only
secondary alcohol that can be dehydrated to give an ether in
our system (29% yield at 200 °C). Even so, propene was the
major reaction product (56%). At temperatures above 200 °C,
propene was formed in >80% yield, with a corresponding
decrease in the yield of ether 2. Surprisingly, the yield of 2
also decreased when a 20 mL reactor was used, instead of 10
mL, under identical conditions. Reaction of longer chain
secondary alcohols, such as s-pentanol, 48, gave exclusively
alkenes with no evidence for any ether formation, even at
reduced temperatures. Clearly, secondary alcohols, apart from
1, cannot be etherified effectively in our system due to the ease
with which they dehydrate.
(61) Samolada, M. C.; Grigoriadou, E.; Kiparissides, Z.; Vasalos, I. A.
J. Catal. 1995, 152, 52.
(62) Sartori, G.; Bigi, F.; Maggi, R.; Arienti, A. J. Chem. Soc., Perkin
Trans. 1 1997, 257.
(63) Marczewski, M.; Perot, G.; Perot, M. Heterogeneous Catalysis and
Fine Chemicals; Elsevier: Amsterdam, 1988.
(64) Marczewski, M.; Bodido, J. P.; Perot, G.; Perot, M. J. Mol. Catal.
1989, 50, 211.
(65) Campelo, J. M.; Garcia, A.; Marinas, J. M.; Moreno, M. S. Bull.
Soc. Chim. Paris 1988, 2, 283.
(66) Balsama, S.; Beltrame, P.; Beltrame, P. L.; Carmiti, P.; Forni, L.;
Zuretti, G. Appl. Catal. 1984, 13, 161.
(67) Pierantozzi, R.; Nordquist, A. F. Appl. Catal. 1986, 21, 263.
(68) Yamanaka, T. Bull. Chem. Soc. Jpn. 1976, 49, 2669.
(69) Bezouhanova, C.; Al-Zihari, M. A. Appl. Catal. 1992, 83, 45.
(70) Santacesaria, E.; Grosso, D.; Gelosa, D.; Carra, S. Appl. Catal. 1990,
64, 83.
(59) Angeletti, E.; Canepa, C.; Martinetti, G.; Venturello, P. Tetrahedron
Lett. 1988, 2261.
(60) Angeletti, E.; Canepa, C.; Martinetti, G.; Venturello, P. J. Chem.
Soc., Perkin Trans. 1 1989, 29, 105.
(71) Olah, G. A.; Kaspi, J.; Bukala, J. J. Org. Chem. 1977, 42, 4187.
(72) Parlman, R. A. U.S. Patent 4,299,996, 1981.

10718 J. Am. Chem. Soc., Vol. 121, No. 46, 1999
Gray et al.
Table 8. Effect of Reaction Conditions on the Alkylation/
the kinetic product, and it is likely that, at such high propene
concentrations relative to 49, 53 rapidly forms as soon as the
polar, and therefore reactive, phenol enters the reactor. As the
less polar ether product 53 passes down through the reactor,
the large excess of propene molecules may coordinate to, and
hence block, active sites on the catalyst surface, thus preventing
possible rearrangement of 53 to 50.
2. It can be assumed that the polarity of the two reaction
systems PhOH/i-PrOH/scCO2 and PhOH/THF/scPropene will
differ significantly. It is likely that the solubility of 53 in the
scPropene/THF mixture will be greater since THF itself is an
ether, and thus its affinity for the catalyst will also be reduced,
resulting in less rearrangement.
Etherification of Phenol in scCO2 and scPropenea
productsb,c
temperature
pressure
(atm)
49
(%)
50
(%)
51
(%)
52
(%)
53
(%)
(°C)
1401
1601
1701
1801
2001
1602
2002
140
160
160
160
160
160
200
52
36
38
34
8
21
10
9
9
13
14
13
5
1
3
4
5
3
1
4
0
1
3
5
37
0
11
32
28
25
0
56
59
7
0
a Other reaction conditions: 49 dissolved in THF (10 g in 25 mL);
DELOXAN ASP catalyst; 10 mL reactor; flow rates - CO2 0.65 L/min,
substrate 0.5 mL/min; 49:1 ) 3:1. b CO2 used as supercritical solvent.
c Propene used as supercritical solvent and reagent.
Conclusions
We have reported the first examples of continuous etherifi-
cation in scCO2. Our results show that CO2 can provide
significant advantages in these heterogeneous, acid-catalyzed
reactions over a wide range of temperatures and pressures. In
addition to the well documented benefits of reduced coking,25-28
we believe that the use of a nonpolar medium, such as scCO2,
may promote reaction by increasing the interaction of polar
species, such as the alcohols, reaction intermediates, and the
catalyst surface. Furthermore, using scCO2 as the solvent for
ether formation will encourage the phase separation of H2O,
which appears to reduce back reaction via rehydration of the
product.
Our results underline the advantages of continuous reactors
in supercritical fluids. As in hydrogenation10,11 and Friedel-
Crafts alkylation,12 relatively small reactors can give excellent
throughput of materials. At the same time, the relatively small
volume of the reactor reduces the problems of working at high
pressures, compared to batch reactors of similar capacity.
Moreover, scale-up of flow reactors is generally simpler than
for batch processes, and it should be possible to scale-up many
of the reactions reported here to industrially usable levels.
Currently, work is in progress in our laboratory to extend the
use of such fixed-bed reactors to other continuous reactions
under supercritical conditions (e.g., dehydrogenation and hy-
droformylation).
Figure 2. Plot showing the variation in isolated product yields with
temperature in the reaction of 49 with 1. The points are labeled as
follows: (9) Phenol, 49, (2) 2-i-propylphenol, 50, (0) 4-i-propylphenol,
51, (3) 2,4-di-i-propylphenol, 52, (b) phenyl-i-propyl ether, 53.
DELOXAN ASP catalyst, 10 mL reactor, flow rates - CO2 0.65 L/min.
alkylation, 50, over para-alkylation, 51, (14% ortho vs 5% para
at 180 °C, 160 atm), possibly because 50 is produced by either
rearrangement of 53 or by ring alkylation via a complex between
the alkylating agent and the hydroxyl group of phenol.44
We observed a greater degree of selectivity for etherification
over alkylation when the experiment was repeated using
scPropene (scPropene; TC ) 91.6 °C, PC ) 45.4 atm)4 as both
alkylating agent and supercritical solvent and with THF as a
cosolvent for phenol. O/C-alkylation selectivities were as high
as 9:1 at 160 °C, 160 atm (compared to 2.5:1 when 1 was the
alkylating agent under otherwise identical conditions). The
highest yield of 53 (59%) was achieved at 200 °C and 200 atm
in scPropene. We tentatively offer two rationalizations for this
phenomenon:
Acknowledgment. We thank Thomas Swan and Co. Ltd.
for fully funding the work at Nottingham and Mr. J. C. Toler
for his assistance. We are grateful to Degussa-Huls AG for
donating the catalysts. We thank Ms. N. J. Meehan, Dr. P. A.
Hamley, Dr. C. Rehren, Dr. T. Tacke, Dr. S. Wieland, Mr. M.
Guyler, Dr. K.-H. Pickel, and Mr. K. Stanley for their help.
M.P. thanks the EPSRC and the Royal Academy of Engineering
for Fellowship support, and F.R.S. thanks the EPSRC for a
CASE studentship.
1. When the reaction was carried out in scPropene, the ratio
of propene/phenol was calculated to be 230:1. The ether 53 is
JA991562P