Azeotropic reflux chromatography: An efficient solution to a difficult separation in the scale-up synthesis of spongistatin 1

Article abstract of DOI:10.1039/b719569a

Azeotropic reflux chromatography (in which the eluent is continuously recycled by means of refluxing) was used to separate a mixture of spiroketal intermediates in the scale-up synthesis of spongistatin 1, leading to an improved separation and an approxim

Full text of DOI:10.1039/b719569a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/obc | Organic & Biomolecular Chemistry
Azeotropic reflux chromatography: an efficient solution to a difficult
separation in the scale-up synthesis of spongistatin 1
Matthew O’Brien,* Alejandro Die´guez-Va´zquez, Day-Shin Hsu, Helmut Kraus, Yukihito Sumino and
Steven V. Ley
Received 19th December 2007, Accepted 25th January 2008
First published as an Advance Article on the web 19th February 2008
DOI: 10.1039/b719569a
Azeotropic reflux chromatography (in which the eluent is
continuously recycled by means of refluxing) was used to
separate a mixture of spiroketal intermediates in the scale-up
synthesis of spongistatin 1, leading to an improved separation
and an approximately 35-fold reduction in the amount of
solvent used.
quantities of spongistatin 1 (1) for further biological and clinical
evaluation.
Synthetic strategy
Our retrosynthetic analysis of this molecule is based upon a late
stage C28–C29 Wittig olefination and C1–C41(OH) macrolac-
tonisation, resulting in the advanced ABCD fragment 2 and EF
fragment 3, thus allowing for a convergent synthesis (Scheme 1).
Further examination of the ABCD fragment 2 reveals a high level
of latent C2-symmetry about C15 (Scheme 2). With this in mind, a
synthetic route was developed in which both the C1–C15 fragment
4 and C16–C28 fragment 5 can each be differentially elaborated
from the diastereomeric precursors 6 and 7 respectively. These, in
turn, can both be derived from the acid catalysed deprotection–
spiroketalisation of ketone 8, which is a 1 : 1 mixture of epimers at
C5 (each epimer leading to the C11 and C19 stereochemistries in 4
and 5 respectively). Thus, a single synthetic sequence can be used
to supply the common intermediate. Compared with our previous
Introduction
Since the seminal publication over one hundred years ago by
Tswett,1 liquid chromatography has grown in popularity and
has been used in different forms to separate a great variety of
compounds. Although other techniques, such as crystallisation
and distillation, prove to be more efficient in many cases, they
often entail some optimization and require the compounds to be
either crystalline or volatile. Liquid chromatography, on the other
hand, frequently provides satisfactory results with a minimum of
groundwork and can, by and large, be applied to a much wider
range of compounds. Indeed, in many academic research labs it
has become the accepted paradigm for compound purification and
separation.2 A major disadvantage, however, is that the method
requires the use of solvent, often in relatively large quantities
that scale with the amount of material to be separated. Typically,
the solvents used are flammable and, as such, their use is highly
undesirable from a safety standpoint. In addition, there may be
a significant environmental impact. Cases where the compounds
to be separated elute at very similar rates generally necessitate
using larger amounts of solvent. During the scale-up synthesis of
spongistatin 1 (1), we encountered a similarly difficult separation.
This led to the development of the chromatographic method that
we describe herein.
Spongistatin 1 (1), with typical GI50 values of 0.025–0.035 nM
when tested against the US National Cancer Institute’s panel of
sixty human cancer cell lines, is the most active of the spongistatin
family of antitumour marine macrolides and has potential for use
in anticancer chemotherapy.3 Unfortunately, these compounds are
only available in minute quantities from nature and thus total
synthesis remains as the only viable source of further material.
Significant interest in these molecules has led to several total
syntheses being reported.4 We recently disclosed a milligram scale
total synthesis of spongistatin 1 (1)5 and, confident of our synthetic
strategy, we sought to repeat the synthesis on a significantly larger
scale. More specifically, we aim to produce meaningful gram
Department of Chemistry, University of Cambridge, Lensfield Road,
Cambridge, UK CB2 1EW. E-mail: mo263@cam.ac.uk; Fax: +44 (0) 1223
336442; Tel: +44 (0) 1223 336398
Scheme 1 Structure and retrosynthetic analysis of spongistatin 1 (1).
Org. Biomol. Chem., 2008, 6, 1159–1164 | 1159
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Scheme 2 Pseudo-C2-symmetric approach to the ABCD fragment.
syntheses of the AB and CD fragments,6 this approach is far more
efficient from a scale-up point of view, dramatically decreasing the
total number of steps required to construct this key unit. However,
whilst this has obvious benefits, it relies on the ability to separate
the very similar AB and CD spiroketal fragments prior to further
elaboration. Indeed, it was recognised at an early stage that the
success of the entire scale-up synthesis would hinge on this crucial
separation. In the small scale synthesis it was found that treatment
of ketone 8 with dilute aqueous perchloric acid in acetonitrile–
dichloromethane smoothly effected cleavage of the acetonide and
triethylsilyl protecting groups with concomitant cyclisation to
generate three different spiroketals corresponding to the AB bis-
anomeric stabilised unit 6 (ABAA) and the two CD anomeric
isomers, bis-anomeric 11 (CDAA) and mono-anomeric 7 (CDAE) in
a ratio of 1.00 : 0.74 : 0.26 respectively (Scheme 3). The undesired
CDAA isomer 11 can easily be converted to the desired CDAE
spiroketal 7 under calcium perchlorate epimerisation conditions
(Scheme 3). As such, 7 and 11 can be considered as a single entity
in terms of separation. The key goal then is the separation of AB
spiroketal 6 from the CD spiroketals 7 and 11. During the small
scale synthesis it was found that flash column chromatography
could be used to efficiently separate the two CD spiroketals 7
and 11 but attempts to separate the AB spiroketal from the CD
spiroketals using this technique failed. In the end, the separation
was carried out using preparative HPLC (silica gel). However,
the maximum loading capacity of the columns at hand (tens of
milligrams) meant that multiple HPLC runs had to be performed
in order to obtain sufficient quantities of separated material. This
Scheme 3 Reagents and conditions: (a) 10% aq. HClO4, MeCN–CH2Cl2, r.t., 30 min, 86%; (b) 3.5% aq. HClO4, 5 equiv. Ca(ClO4)2·4H2O, MeCN–CH2Cl2,
r.t., 18 h, 87% after three recycles.
1160 | Org. Biomol. Chem., 2008, 6, 1159–1164
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Fig. 1 TLC (SiO2) chromatograms of the mixture (m) of 6, 7 and 11 against reference samples. (a) Et2O × 1 (b) 80 : 20 Et2O–PE × 8 (c) 1 : 1 Et2O–DCM ×
1 (d) 1 : 1 Et2O–pentane × 15 (e) DIPE × 10 (all plates stained with vanillin).
process would quite clearly be impractical for the amounts of
material required for our scale-up synthesis (ca. 40 g each of ABAA
(6) and CDAE (7)). Rather than spend time investigating the use of
expensive, high-capacity HPLC columns, attention turned to the
reinvestigation of traditional silica-gel column chromatography.
In order to determine a suitable solvent for flash chromatog-
raphy, TLC studies were carried out on a variety of binary and
ternary solvent systems. Unfortunately, no combination led to the
resolution of the three spiroketals into individual spots on the
chromatogram. An example chromatogram (neat diethyl ether)
is shown in Fig. 1a. It can be seen that the spots streak and
tend to merge into each other. Multiple-elution TLC (whereby
the processes of elution and drying are repeated several times
prior to development) was then attempted in an effort to afford
resolution. Of the solvent mixtures initially assayed, those based
on diethyl ether–petroleum ether were found to provide the best
resolution. Shown in Fig. 1b is a chromatogram of the crude
spiroketal mixture run 8 times in 80 : 20 diethyl ether–petroleum
ether. Three separate spots, corresponding to each spiroketal, can
easily be seen. Solvents based on the diethyl ether–DCM system
led to a greater separation of 7 and 11 (even in single elution) but
no separation between 6 and 7. Pleasingly, when the diethyl ether–
petroleum ether solvent system was used in preparative TLC, three
resolved bands were obtained (using UV light to visualise) from
which analytically pure (1H NMR) samples could be extracted.
afforded reasonable levels of separation, a larger scale separation
was attempted. The crude product from reaction of 24 g of ketone
8 was loaded onto 2 kg of silica gel (packed in petroleum ether)
with 30 mL 1 : 1 : 1 DCM–diethyl ether–petroleum ether. The
column was then eluted with 50 L 1 : 1 diethyl ether–petroleum
ether, collecting fractions of 500 mL. A solvent gradient was then
started from 1 : 1 diethyl ether–petroleum ether to diethyl ether
over 150 L, after which time all the ABAA (6) had eluted. The
solvent was then drained and the column flushed with acetone
until all the CDAA (11) had eluted (11 L). After combination
of fractions three samples were obtained: 6.19 g clean 6, 2.68 g
1 : 1 : 1 6 : 7 : 11, 5.35 g 1 : 8 7 : 11. This corresponds to
a 92% yield and an 87% separation of the ABAA spiroketal 6.
Although this level of separation was acceptable, two major issues
were encountered. Firstly, the amount of solvent used for the
column was well over 200 L, which is unacceptable in terms of
cost, safety and environmental concerns. Secondly, the time spent
running the column (i.e. refilling the solvent reservoir, changing
fraction vessels and evaporating solvent) was extremely long (over
40 h). As significantly greater amounts of material would need to
be purified, an alternative, more efficient method was sought for
the separation of these spiroketals.
We envisaged that a system in which the solvent is continually
recycled by means of reflux, thereby combining the properties of
the chromatography column with those of the Soxhlet extractor,
might lead to a dramatic reduction in solvent volume and
operating effort. Whilst this concept had been demonstrated on
a limited number of other examples,7 it had not been used to
separate stereoisomers. Additionally, we were concerned about the
stability of our compounds to prolonged heating after elution and
therefore wished to investigate the use of relatively low-boiling
solvents, including low-boiling azeotropic mixtures (which, to
our knowledge, have not been employed previously). To test the
applicability of the general concept on a small scale, we fashioned
a prototype apparatus using equipment to hand (Fig. 2a), which
operates as follows: the solvent in the collecting round bottomed
flask is heated to reflux with stirring and the vapour travels up the
glass tubing at the side of the column, the condensate collects on
the top of the column through which it travels on its way back to
the round bottomed flask. Of course, the difference in elution times
of the compounds to be separated has to be greater than the length
of time collecting each fraction to avoid remixing in the collecting
vessel. As the solvent cycle involves reflux, this would severely limit
Column chromatography
Having established a potential solvent system, small scale column
chromatography was attempted. After some experimentation, the
use of a solvent gradient was found to give optimal results. For
example, the crude product from reaction of 40 mg of the ketone
8 was loaded onto 7 g of silica gel (packed using neat petroleum
ether) with 1 mL 1 : 1 DCM–petroleum ether and was eluted
with 500 mL 1 : 1 diethyl ether–petroleum ether collecting 10 mL
fractions. The column was then eluted with a solvent gradient
from 1 : 1 diethyl ether–petroleum ether to 70 : 30 diethyl ether–
petroleum ether over 800 mL. At this point all of the ABAA
(6) and CDAE (7) had eluted. The remainder of the CDAA (11)
was then eluted with 300 mL diethyl ether. The fractions were
combined into four samples: 10 mg clean 6, 3 mg 2 : 1 6 : 7,
2 mg clean 7, 8 mg clean 11 corresponding to an 89% yield for
the reaction and 83% separation of 6. As these small scale studies
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 1159–1164 | 1161
©

View Article Online
the solvent was changed and collection continued. The ABAA (6)
material then began to elute. At 6 h, slight traces of CDAE (7)
were observed by TLC and the solvent was changed once more.
At 7 h, no ABAA (6) was observable, only CDAE (7) and CDAA
(11). The column was flushed with acetone (20 mL) and these
Fig. 2 Prototype reflux chromatography apparatuses.
the solvent system used due to the phenomenon of azeotropic
distillation. In order to keep the eluting solvent isochratic we
could either employ a single solvent or use a low-boiling azeotrope
with the required properties. Both options were investigated. To
simplify matters, we abandoned the use of petroleum ether as this
is an ill-defined mixture of hydrocarbons whose composition may
vary from batch to batch.
Diethyl ether was initially retained as the polar component.
Hexane, which is a constituent of petroleum ether and has
comparable chromatographic qualities, could not be used as the
non-polar partner as it fails to form an azeotrope with diethyl
ether. Instead, we considered pentane, which is inexpensive and
forms an azeotrope with diethyl ether of almost 1 : 1 composition
(b.p. 34 ◦C).8 TLC studies supported this choice. Shown in Fig. 1d
is a multiple-elution chromatogram of the mixture of spiroketals
eluted 15 times in 1 : 1 diethyl ether–pentane before development.
As an alternative to an azeotrope we also considered the use
of a less polar dialkyl ether. In terms of boiling point and
cost/availability, two obvious options were methyl tert-butyl ether
(MTBE, b.p. 56 ◦C) and diisopropyl ether (DIPE, b.p. 69 ◦C).
Whilst MTBE afforded insufficient separation in TLC studies,
DIPE seemed to be the solvent of choice. As shown in Fig. 1e,
when eluted 10 times in DIPE, the CDAE (7) spot runs much
closer to that of the CDAA (11) spiroketal and the separation
between the former and the topmost ABAA (6) spot is higher. Using
the diethyl ether–pentane azeotrope, we conducted a small scale
trial of the process using the apparatus shown in Fig. 2a. Crude
spiroketal mixture resulting from reaction of 50 mg of ketone 8
in 2 mL 1 : 1 : 1 diethyl ether–DCM–pentane was loaded onto
2 g silica gel (loading: 25 mg g−1), which was packed with 1 :
1 diethyl ether–pentane. The 100 mL round bottomed flask was
charged with 80 ml 1 : 1 diethyl ether–pentane and heated to
reflux and collection began. The contents of the collecting flask
were monitored by TLC analysis. After 3 h only spots higher than
the ABAA (6) (mainly due to triethylsilyl byproducts) were present;
Fig. 3 (a) 20 cm jacketed coil condenser; (b) screw-thread connector
(GL 25); (c) side-arm adapter (23 mm o.d.); (d) L-shape glass tubing
(15 mm o.d.); (e) RodavissTM B24 ground glass joint; (f) 80 mm jacket
condenser (B34 male cone incorporated into jacket); (g) RodavissTM B34
ground glass joint; (h) side-arm and insulation (5 mm cotton wool within
10 mm polystyrene insulating foam covered with aluminium foil); (i)
column, 15 cm o.d., height 19 cm (excluding taper), porosity 3 sinter;
(j) retaining ring; (k) screw-thread connector (GL 14, lined with PTFE
tape) to cut-down luer-lock 2 mL glass syringe; (l) polythene luer-luer tap;
(m) 150 mm 14 G stainless steel syringe; (n) B14 latex suba-seal (syringe
R
and seal wrapped with PTFE tape and Parafilmꢀ); (o) 1 L 3-neck round
bottomed flask; (p) aluminium heating block; (q) stirrer-hotplate.
1162 | Org. Biomol. Chem., 2008, 6, 1159–1164
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
flushings were combined with the 6 h–7 h fraction. After removal
of solvent, two fractions were obtained: 16 mg 10 : 1 6 : 7 (3 h–
6 h), 14 mg 1 : 3 7 : 11 (6 h–7 h plus the acetone flush material)
corresponding to a 92% yield. Although the separation was not
complete, we felt this could be improved by collecting fractions at
shorter intervals. The same experiment was also carried out using
DIPE as solvent. A similar level of separation was observed but,
in this case, the material obtained seemed to be impure by NMR
analysis. This degradation may have been due to the higher boiling
well as to keep the total height of the apparatus within reasonable
working limits. The volume between the sinter and the luer outlet
was the minimum that could be achieved within the practical limits
of construction. This diminished any post-column remixing of
eluates. The side arm was lagged with a layer of cotton wool
inside insulating foam covered with aluminium foil. With this
in place, a sufficient rate of reflux of the diethyl ether–pentane
azeotrope could be maintained with the heating block at around
100 ◦C. The flow rate from the column packed with 1 kg of silica
and solvent at the level of the side arm adapter was an acceptable
80 mL min−1.
◦
point of DIPE (69 C). This, added to the fact that DIPE has a
noted tendency to form explosive peroxides,9 led us to abandon
its use and we settled on the diethyl ether–pentane mixture for
all subsequent work. Using the slightly larger apparatus shown
in Fig. 2b, the separation was carried out with crude spiroketal
mixture from reaction of 1.5 g of ketone 8. In this case, 46 g
of silica gel (loading: 33 mg g−1) was used and three fractions
were obtained: 425 mg 6, 55 mg 5 : 3 6 : 7, 412 mg 1 : 3 7 : 11.
This corresponds to a 92% reaction yield and 93% separation of
the ABAA spiroketal (6). Having established that the concept was
applicable, we designed and built10 a larger apparatus that would
allow for the separation of significant quantities of the spiroketal
materials (Fig. 3). In previously reported systems, where very few
fractions were collected, the column is placed vertically above
the solvent reservoir/collector, connected to it by a ground glass
joint.7 We felt that this would make the process of changing solvent
fractions difficult. As we intended to collect several fractions,
we kept the configuration of our prototype system in which the
column is held separately, horizontally removed from the solvent
reservoir/collector but attached to it by connections that are easily
removed during solvent change. A second condenser (Fig. 3f) was
included, between the column and side arm adapter (Fig. 3c),
to provide cooling to the solvent condensate. The column was
designed to hold at least 1 kg of silica gel and was relatively wide
to provide sufficient solvent flow through the porosity 3 sinter as
On this larger scale, a slightly higher loading (42 mg g−1) of
material was used. The crude product from reaction of 42 g of
ketone 8 in 45 mL 1 : 1 : 1 DCM–diethyl ether–pentane was loaded
onto 1 kg of silica gel packed with 1 : 1 diethyl ether–pentane.
After the solvent had been adsorbed, the silica gel was covered
with a 2 inch layer of sand (low in iron) and 1 : 1 diethyl ether–
pentane was added up to the level of the side-arm adapter. The
1 L round bottomed flask was charged with 800 mL 1 : 1 diethyl
ether–pentane and heated to reflux. After solvent had started to
condense above the column, the tap at the bottom was opened and
elution began. The contents of the collecting flask were monitored
periodically by TLC analysis. Fractions were collected on average
every 7 h. Fraction collection involved turning off the heating,
allowing the heating block to cool until reflux had stopped (about 5
minutes), closing the luer-luer tap, disconnecting the screw thread
connector (Fig 3b) and hypodermic needle (Fig 3m), removing the
side arm (Fig 3d) and changing the solvent in the round bottomed
flask before reversing the process. In total, the time needed to
collect each fraction was around 10 minutes. The ABAA (6) began
to elute at 21 h (fraction 4, Fig. 4). Prior to this, the triethylsilyl
byproducts had completely eluted. The CDAE (7) started to co-
elute with the ABAA at 58 h (fraction 11, Fig. 4). After 93 h (fraction
16, Fig. 4), the last traces of ABAA (6) had eluted and the CDAA
Fig. 4 TLC chromatogram of fractions from reflux chromatography against reference a mixture (M) of 6, 7 and 11. Eluted with 10 × 80 : 20
Et2O–petroleum ether (stained with vanillin solution).
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 1159–1164 | 1163
©

View Article Online
(11) was the major component of the eluate. After 124 h (fraction
19, Fig. 4), the CDAA (11) appeared to have completely eluted by
TLC analysis. At this point the reflux was stopped and the column
was allowed to drain under gravity and then flushed with 3 L of
acetone. Interestingly, the acetone washes (fractions 19–22, Fig. 4)
contained significant quantities of the CDAA spiroketal (11). After
combination of fractions and evaporation ofsolvent, three samples
were obtained: 11.40 g 6, 3.74 g 26 : 58 : 16 6 : 7 : 11, 10.12 g 11 (a,
b and c in Fig. 4). This corresponds to a 93% reaction yield and a
92% separation of ABAA (6). The total amount of solvent used was
17 L. Using the measured flow rate of 80 mL min−1, the amount of
solvent required to run the same column in the traditional manner
is estimated to be in excess of 590 L. Thus, the requisite volume
of solvent has been significantly reduced, by a factor of around
35. Not only does this signify a considerable reduction in cost,
but it also represents a substantial lessening of the environmental
impact of the process. Furthermore, as our solvent system is an
azeotropic mixture, it has the same composition after distillation
and can be collected from the rotary evaporator after product
isolation to be reused in subsequent separations. This idea is
currently being employed as we continue our scale-up synthesis of
spongistatin 1 (1).
(DSH) and the Cambridge European Trust for a partially funded
studentship (HK).
Notes and references
1 M. S. Tswett, Ber. Dtsch. Bot. Ges., 1906, 24, 316, 384.
2 The following paper greatly contributed to this popularity: W. C. Still,
M. Kahn and A. Mitra, J. Org. Chem., 1978, 43, 2923.
3 G. R. Pettit, Z. A. Cichacz, F. Gao, C. L. Herald, M. R. Boyd, J. M.
Schmidt and J. N. A. Hooper, J. Org. Chem., 1993, 58, 1302; M.
Kobayashi, S. Aoki, H. Sakai, K. Kawazoe, N. Kihara, T. Sasaki and I.
Kitagawa, Tetrahedron Lett., 1993, 34, 2795; N. Fusetani, K. Shinoda
and S. Matsunaga, J. Am. Chem. Soc., 1993, 115, 3977; R. Bai, Z. A.
Cichacz, C. L. Herald, G. R. Pettit and E. Hamel, Mol. Pharmacol.,
1993, 44, 757; R. Bai, G. F. Taylor, Z. A. Chicacz, C. L. Herald, J. A.
Keplar, G. R. Pettit and E. Hamel, Biochemistry, 1995, 34, 9714.
4 (a) D. A. Evans, B. W. Trotter, B. Coˆte´, P. J. Coleman, L. C. Dias and
A. N. Tyler, Angew. Chem., 1997, 109, 2957, (Angew. Chem., Int. Ed.
Engl., 1997, 36, 2744); D. A. Evans, B. W. Trotter, P. J. Coleman, B.
Coˆte´, L. C. Dias, H. A. Rajapaske and A. N. Tyler, Tetrahedron, 1999,
55, 8671; (b) M. M. Hayward, R. M. Roth, K. J. Duffy, P. I. Dalko,
K. L. Stevens, J. Guo and Y. Kishi, Angew. Chem., 1998, 110, 202,
(Angew. Chem., Int. Ed., 1998, 37, 190); (c) A. B. Smith III, Q. Lin,
V. A. Doughty, L. Zhuang, M. D. McBriar, J. K. Kerns, C. S. Brook,
N. Murase and K. Nakayama, Angew. Chem., 2001, 113, 202, (Angew.
Chem., Int. Ed., 2001, 40, 196); (d) I. Paterson, D. Y.-K. Chen, M. J.
Coster, J. L. Acena, J. Bach and D. J. Wallace, Org. Biomol. Chem.,
2005, 3, 2431; I. Paterson, D. Y.-K. Chen, M. J. Coster, J. L. Acena, J.
Bach, K. R. Gibson, L. E. Keown, R. M. Oballa, T. Trieselmann, D. J.
Wallace, A. P. Hodgson and R. D. Norcross, Angew. Chem., 2001, 113,
4179, (Angew. Chem., Int. Ed., 2001, 40, 4055); (e) M. T. Crimmins, J. D.
Katz, D. G. Washburn, S. P. Allwein and L. F. McAtee, J. Am. Chem.
Soc., 2002, 124, 5661; (f) C. H. Heathcock, M. McLaughlin, J. Medina,
J. L. Hubbs, G. A. Wallace, R. Scott, M. M. Claffey, C. J. Hayes and
G. R. Ott, J. Am. Chem. Soc., 2003, 125, 12844 and references therein.
5 M. Ball, M. J. Gaunt, D. F. Hook, A. S. Jessiman, S. Kawahara, P.
Orsini, A. Scolaro, A. C. Talbot, H. R. Tanner, S. Yamanoi and S. V.
Ley, Angew. Chem., Int. Ed., 2005, 44, 5433.
Conclusion
Our route to spongistatin 1 (1) exploits the high level of C2
pseudosymmetry present in the ABCD fragment 2, and relies
on the ability to separate the ABAA fragment 6 from the CD
fragments 7 and 11. We have developed a novel system which
is cost-effective and uses a very simple apparatus to continu-
ously recycle the azeotropic solvent mixture. The method allows
for a dramatic reduction in the amount of solvent required,
uses no expensive equipment and is easy to construct and
operate.
6 M. J. Gaunt, D. F. Hook, H. R. Tanner and S. V. Ley, Org. Lett., 2003,
5, 4815; M. J. Gaunt, A. S. Jessiman, P. Orsini, H. R. Tanner, D. F.
Hook and S. V. Ley, Org. Lett., 2003, 5, 4819.
7 For examples see: R. Meier and J. Fletschinger, Angew. Chem., 1956,
68, 373; M. J. S. Dewar, R. B. K. Dewar and Z. L. F. Gaibel, Org. Synth.
Coll. Vol. 5, 1973, 727, (Org. Synth., 1966, 46, 65); J. M. Kauffman and
C. O. Byorkman, J. Chem. Educ., 1976, 53, 33; K. Chatterjee, D. H.
Parker, P. Wurz, K. R. Lykke, D. M. Gruen and L. M. Stock, J. Org.
Chem., 1992, 57, 3253; K. C. Khemani, M. Prato and F. Wudl, J. Org.
Chem., 1992, 57, 3254.
We suggest that this little-used technique could have a much
broader application in organic synthesis programmes in the
future.
Acknowledgements
8 Azeotropic Data III, ed. L. H. Horsley, American Chemical Society,
Washington, 1973 and references therein (from NMR analysis of
distillates in our apparatus, the azeotrope composition seemed to lie
between 50 : 50 and 53 : 47 diethyl ether–pentane).
The authors thank Melvyn Orris and Keith Parmenter for helpful
technical discussions. We also thank the Ministerio de Educacio´n
y Ciencia (Spain) for a postdoctoral fellowship (ADV), the
Taiwan Merit Scholarship Program for a postdoctoral fellowship
9 CRC handbook of laboratory safety, ed. A. K. Furr, CRC Press,
Cleveland, 5th edn, 2002, p. 255 and references therein.
10 Glassware was made by Soham Scientific, Munceyꢀs Mill, 37 Mildenhall
Road, Fordham, Ely, Cambridgeshire, CB7 5NP, UK.
1164 | Org. Biomol. Chem., 2008, 6, 1159–1164
This journal is
The Royal Society of Chemistry 2008
©