SHORT PAPER
507
A Practical Large-Scale Synthesis of 3-Carbomethoxy-3-sulfolene
a
a
b
b
John Leonard, * Andrew B. Hague, Martin F. Jones, Richard A. Ward
a
Chemistry Department, University of Salford, Salford M5 4WT, UK
Chemical Development, Glaxo−Wellcome Research and Development, Stevenage, Hertfordshire, SG1 2NY, UK
b
E-mail: j.leonard@salford.ac.uk
Received 26 August 1999; revised 11 October 1999
was added too quickly compound 9b was formed in sig-
nificant amounts. The formation of this by-product was
avoided by slow addition of 1b to a solution of paraform-
aldehyde (2 equivalents) over a period of 3 hours. The re-
Abstract: A simple, efficient one-flask procedure for the large-
scale preparation of 3-carbomethoxy-3-sulfolene from trimeth-
ylphosphonoacetate is described. The procedure incorporates a
cheap and simple sulfide to sulfone oxidation using Oxone®. No
chromatography or intermediate purification steps are involved and action was found to be complete after 5 hours, but we
the product is isolated by crystallisation.
normally found it convenient to leave this step of the pro-
Key words: 3-carbomethoxy-3-sulfolene, 2,5-dihydroxy-1,4- cess overnight.
dithiane, Oxone
To convert methyl ether 2a into vinyl phosphonate 3a the
methanol solvent was replaced by toluene and the mixture
heated at reflux with a catalytic quantity of p-toluene sul-
fonic acid. We found that after 4 hours at reflux, the mix-
tures formed from either 1a or 1b had been converted
cleanly into either 3a or 3b respectively, and GC/ GC-
mass spectrometry indicated that the material formed
would be clean enough to use directly in the next step of
the procedure. We thought that it would be useful to avoid
the change of solvent from methanol to toluene during this
first step, so we carried out a range of experiments using
toluene, toluene/methanol and toluene/acetic acid mix-
tures. These trial reactions were also studied at different
temperatures. However, in all cases, even using 50%
methanol/toluene, the rate of formation of 2a was reduced
and significant quantities of by-products were formed.
The most prominent of these was 9a, formed by reaction
of 2a or 3a with 1a. Thus, after much experimentation it
was apparent that the slight practical inconvenience of
changing solvents from methanol to toluene during the
preparation of 3a was worth accepting in order to maxi-
mise the efficiency of its formation.
3
-Carbomethoxy-3-sulfolene (8) is a very useful template
for the construction of functionalised masked dienes. In
our synthetic studies towards various alkaloids we have
exploited the highly selective prenylation of its dianion, to
provide flexible 2,3-substituted diene precursors for use in
inverse demand intramolecular Diels−Alder reactions.1
Although 3-carbomethoxy-3-sulfolene is commercially
available it is prohibitively expensive. A neat synthetic
route was published by McIntosh and Sieler some time
ago and initially we used this to prepare small quantities
-6
7
of the compound. We found that each step of the route
was effective, but overall it was very labour intensive and
expensive to carry out the preparative sequence on a large
scale. We identified the main problems as being: the
mixed ester starting material 1a was expensive; isolation
and distillation of intermediates 3a and 7 were very time
consuming and there were significant material losses dur-
ing these operations; the final oxidation procedure used
m-CPBA which is expensive for large-scale work and re-
quires a cumbersome workup.
In the next step of the McIntosh and Sieler sequence
We required a procedure that would enable us to prepare
>
2
,5-dihydroxy-1,4-dithiane was cracked, at reflux in
100g batches of material routinely without the commit-
dichloromethane, then a solution of distilled 3a in dichlo-
romethane was added to the a-mercaptoacetaldehyde pro-
duced. Instead, we simply allowed the toluene solution
from the formation of 3b (or 3a) to cool to about 50 °C,
then added 2,5-dihydroxy-1,4-dithiane directly and al-
lowed it to react for 3 hours. After this time the vinyl phos-
phonate had completely disappeared (by GC) and a TLC
spot corresponding to dihydrothiophene 7 was the only
significant product detected. We particularly wanted to
avoid isolation of this intermediate, because of the stench
associated with the work-up procedure. An initial attempt
to oxidise the thiophene in situ did not provide a very high
yield of the sulfolene. However, simply washing the tolu-
ene solution with dilute HCl prior to the oxidation step
caused a dramatic improvement. We were able to detect
some of the intermediate 6a in the reaction mixture prior
ment of time consuming and technically demanding ex-
perimental work. Our aim at the outset was to convert the
literature synthesis into a one-flask procedure, avoiding
isolation of any intermediates. Since 3-carbomethoxy-3-
sulfolene is highly crystalline, we hoped that we would
also be able to avoid any purification procedures, apart
from crystallisation of the final product.
1
Our study began by carefully monitoring ( H NMR and
GC-mass spectrometry) the progress of each step of the
synthetic process reported by McIntosh and Sieler. Reac-
tion of 1a with excess paraformaldehyde in methanol at
reflux progressed slowly, but the phosphonate was even-
tually converted completely (~20 h) to a mixture contain-
ing mainly 2a together with a smaller amount of vinyl
phosphonate 3a. We found that the cheaper reagent, trim-
ethyl phosphonoacetate 1b, reacted faster than 1a, but if it
Synthesis 1999, No. 4, 507–509 ISSN 0039-7881 © Thieme Stuttgart · New York