On the Diels-Alder reactions and the Lewis acid induced rearrangements of 6-fumaryl 1,3,8-nonatrienes

Article abstract of DOI:10.1016/S0040-4039(02)00377-5

6-Fumaryl 1,3,8-nonatrienes substituted at the C5 position by a vinyl group were found to undergo competing tandem sigmatropic rearrangement/Diels-Alder cyclisation when heated under standard Diels-Alder cyclisation conditions. This rearrangement became t

Full text of DOI:10.1016/S0040-4039(02)00377-5

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 2753–2756
On the Diels–Alder reactions and the Lewis acid induced
rearrangements of 6-fumaryl 1,3,8-nonatrienes
a,
a
b
Paul A. Clarke, * Rebecca L. Davie and Simon Peace
a
School of Chemistry, University of Nottingham, University Park, Nottingham NG7 2RD, UK
b
Medicinal Chemistry, GlaxoSmithKline R&D, Gunnels Wood Road, Stevenage SG1 2NY, UK
Received 10 January 2002; revised 5 February 2002; accepted 21 February 2002
Abstract—6-Fumaryl 1,3,8-nonatrienes substituted at the C5 position by a vinyl group were found to undergo competing tandem
sigmatropic rearrangement/Diels–Alder cyclisation when heated under standard Diels–Alder cyclisation conditions. This rear-
rangement became the exclusive pathway when the reactions were performed in the presence of a Lewis acid. © 2002 Elsevier
Science Ltd. All rights reserved.
As part of our studies towards the total synthesis of
Alder cyclisation reaction, followed by functional group
manipulations.
1
,2
FR182877 1 and related compounds, we have investi-
gated the potential use of an intramolecular Diels–
Alder cyclisation to install the B-ring with appropriate
functionality to allow the synthesis of the A- and
C-rings via a simultaneous two-directional reductive
annulation strategy. Such a route would provide func-
tionality on the C-ring suitable to expedite its union
with a DEF-ring precursor unit. To this end our initial
target was 2 (Scheme 1) which we reasoned could be
synthesised from 4 by way of an intramolecular Diels–
Compound 4 was synthesised in six steps from known
3,4
Stille coupling precursors (Scheme 2). Protection of
the hydroxyl group of 6 as a TBS ether followed by
reduction of the ester and re-oxidation to the aldehyde
furnished 8, which was converted into 4 by the addition
of vinyl magnesium bromide and subsequent esterifica-
tion with monoethyl fumaryl chloride.
CO2Et
(ii)
CO2Et
(iii-iv)
CHO
Bu3Sn
(i)
(v-vi)
4
HO
H
OH
O
OH
OH
H
A
OH
OTBS
OTBS
H
H
H
CHO
5
6
7
8
H
O
F
O
B
E
H
C
D
H
Scheme 2. Reagents and conditions: (i) EtO
(MeCN) PdCl , DMF, 84%; (ii) TBSCl, imidazole, DMF,
2
00%; (iii) DIBAL-H, CH Cl , −78°C–rt, 95%; (iv) Swern
oxidation, 75%; (v) vinyl magnesiumbromide, THF, rt, 99%;
vi) EtO CCHꢀCHCOCl, py, CH Cl , 0°C, 100%.
CCHꢀCHI,
2
H
O
H
H
2
2
1
2 2
FR182877 (1)
(
O
O
2
2
2
H
CO2Et
OTBS
O
O
If 4 were to undergo an intramolecular Diels–Alder
cyclisation four possible products could be formed (Fig.
1). These products result from both exo- and endo-
cyclisation and cyclisation onto either of the a or b
faces of the diene unit. The stereochemistry required for
our synthesis would necessitate exo-cyclisation from the
CO2Et
OTBS
H
H
3
4
Scheme 1.
5
a-face of the diene unit (i.e. the exo-trans compound
3
). We predicted that this would be the case due to a
minimisation of steric interactions between the exo-
cyclic vinyl group and H3 and H8 in the transition
*
0
Corresponding author. E-mail: paul.clarke@nottingham.ac.uk
040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)00377-5

2
754
P. A. Clarke et al. / Tetrahedron Letters 43 (2002) 2753–2756
state. Indeed, MM2 calculations indicated that a-attack
By this reasoning 14 must be formed by sequential [3,3]
sigmatropic transpositions of the fumarate group down
the chain in the opposite direction to form the other
conjugated triene. Interestingly, when 14 was re-submit-
ted to the Diels–Alder conditions no further rearrange-
ments or cyclisations took place. Rearrangements of
this type are not unprecedented. Eberle reported a
similar observation while studying the Diels–Alder
cyclisations of monoethyl-fumarates of 1-phenyl-2,4-
−
1
would be preferred by approx. 1.32 kcal mol . While
we were engaged in these studies a report by Sherburn
6
et al. appeared detailing both computational and syn-
7
thetic studies on a similar system which indicated that
our predictions were correct.
O
H
CO2Et
O
10
hexadien-1-ol and related compounds.
H
OTBS
H
O
O
H
H
exo-trans
3
CO2Et
CO2Et
OTBS
O
O
O
H
H
O
H
OTBS
H
H
H
CO2Et
O
O
O
4
H
3, 39%
9, 7%
8
9
exo
O
5
OTBS
H
CO2Et
CO2Et
O
2
CO Et
1
O
OTBS
TBSO
exo-cis
9
OTBS
O
3
2
endo
O
H
H
4
CO2Et
12
14, 11%
H
H
+ unisolated diastereomer
OTBS
H
13 combined yield 6%
endo-trans 10
Figure 2.
O
H
CO2Et
OTBS
O
O
O
O
O
O
t
CO2 Bu
OTBS
H
endo-cis 11
OTBS
1
5
16
Figure 1.
O
O
H
H
t
t
While the crude 6-fumaryl 1,3,8-nonatriene 4 was a
black oil, its H NMR showed it to be a single com-
CO2 Bu
CO2 Bu
O
1
H
H
pound and its TLC showed only one spot (R = 0.79;
f
OTBS
OTBS
H
H
1:2 EtOAc–petroleum ether). Decolourisation of 4
could only be achieved through filtration via a plug of
silica gel. This, however, caused decomposition of 4,
generating three additional compounds. Due to the
decomposition it was decided to effect the Diels–Alder
cyclisation on the crude material. To this end 4 was
heated in toluene in the presence of butylated hydroxy-
toluene (BHT) to 110°C, under the conditions reported
17
18
O
O
H
O
2
CO Et
CO2Et
OTBS
O
H
OTBS
H
7
by Sherburn, to generate five new products. The five
19
20
products were identified as 3 39%, 9 7%, 12 and 13 6%
8
O
O
and 14 11% (Fig. 2). Gratifyingly, the major Diels–
H
CO2Et
OTBS
Alder adduct was the desired 3 with the other exo-iso-
mer 9 appearing in small quantities. As far as we could
O
O
CO2Et
OTBS
H
1
detect (by 400 MHz H NMR) neither of the endo-iso-
H
mers were formed. Surprisingly, significant quantities of
rearranged products 12, 13 and 14 were isolated. Com-
pounds 12 and 13 were isolated as a mixture, insepara-
ble by flash column chromatography. The major
component 12 was finally isolated after extensive HPLC
21
22
O
O
H
H
CO2Et
OTBS
CO2Et
OTBS
O
O
9
separations. Unfortunately, useful quantities of 13
could not be isolated in this manner. We postulate that
H
H
1
2 is formed by the thermal [3,3] sigmatropic transposi-
23
24
tion of the fumarate group to the terminal double
bond, thus allowing the formation of a conjugated
triene, which rapidly undergoes Diels–Alder cyclisation.
Figure 3.

P. A. Clarke et al. / Tetrahedron Letters 43 (2002) 2753–2756
2755
Having identified the compounds of the thermal Diels–
Alder reaction we were now in a position to recognise
the major two products of the silica gel promoted
decomposition of 4 as 3 and 14, showing that the
rearrangement is very facile. Interestingly, when the
It seems that the fumaryl group, more than the other
esters studied is especially prone to these rearrange-
ments. Although the relative rates of rearrangement of
different esters appears to be dependent upon the
nature of the conjugated system generated: the greater
the conjugation the more facile the rearrangement
process.
Diels–Alder cyclisation was attempted using Et AlCl
2
catalysis, 12 and 13 were the only detectable products
1
by 400 MHz H NMR analysis and were isolated in a
R
R
combined 41% yield. To date we have been unable to
obtain any quantity of uncontaminated 13, so its struc-
ture has not been determined.
O
O
R
O
O
O
O
H
H
Me
Me
H
Intrigued by this result we sought to determine the
structural requirements for the [3,3] sigmatropic rear-
rangements. Contrary to the observations of Eberle,
4
OTBS
H
OTBS
H
OTBS
10
we did not detect any rearrangement products when
acetate 15 (Fig. 3) was heated in toluene in the presence
of BHT to 110°C. Similarly, when 15 was treated with
R
O
O
O
H
H
O
OEt
Et AlCl no rearrangement occurred. When the mono
O
2
11
tert-butyl fumarate 16 was used instead of 4 only the
appropriate exo-trans 17 17% and exo-cis 18 4% prod-
ucts of Diels–Alder cyclisation were detected and iso-
lated, albeit in low yield and overall conversion.
Similarly, when the diene bearing an ethyl group at the
C5 position 19 instead of a vinyl group was heated in
toluene in the presence of BHT to 110°C, the only
products were of Diels–Alder cyclisation, namely the
exo-trans 20 45% and exo-cis 21 7%. Rearrangement
also failed to occur when the unsubstituted diene 22
was heated under our standard conditions. In this case
the exo-product 23 was formed as the major isomer
with only trace amounts of the endo-product 24 being
detected.
Me OTBS
OTBS
1
4
Figure 4.
In summary, we have shown that 6-fumaryl 1,3,8-nona-
trienes substituted at the C5 position by an unsaturated
unit undergo rearrangements which compete with the
‘
desired’ Diels–Alder cyclisation. In some cases this
process is so facile that it completely dominates the
reactivity of these systems. This observation will be of
interest to those seeking to use these 6-fumaryl 1,3,8-
nonatrienes in synthesis.
Acknowledgements
It would seem from these studies (and the earlier work
of Eberle) that the Diels–Alder cyclisations of 6-
fumaryl 1,3,8-nonatrienes bearing pendant unsaturation
are far from straightforward. Diels–Alder cyclisations
compete with a very facile [3,3] sigmatropic rearrange-
ment. As compounds 19 and 22 do not rearrange, the
driving force for the rearrangement is undoubtedly the
formation of a conjugated triene system. The resultant
triene may or may not then undergo Diels–Alder cycli-
sation. In the case of 12, the precursor triene has no
difficulty in adopting the required S-cis conformation
of the diene component, and so Diels–Alder cyclisation
can occur. However, in the case of 14, if one assumes
sequential [3,3] sigmatropic rearrangements proceeding
via chair-like transition states, then this would generate
triene 14 with the double bond geometries shown in
Fig. 4. In this case the adoption of the S-cis conforma-
tion needed for Diels–Alder cyclisation would be disfa-
voured due to the steric interaction of the H and Me
groups indicated, and hence the rate of any subsequent
cyclisation reaction slowed. The tert-butyl compound
We thank the EPSRC and GlaxoSmithKline for a
CASE studentship (R.L.D.). We also thank Clare
Paterson and Eric Hortense (GSK) for HPLC purifica-
tion of 12, Phil Sidebottom (GSK) for 2D NMR spec-
troscopic analysis of compound 12 and Tom Gallagher
(Nottingham) for the preparation of quantities of nona-
triene 4.
References
1
. Yoshimura, S.; Sato, B.; Kinoshita, T.; Takase, S.; Ter-
ano, H. J. Antibiot. 2000, 53, 615.
. Two other groups have reported their efforts towards a
synthesis of FR182877: (a) Vanderwal, C. D.; Vosburg,
D. A.; Weiler, S.; Sorensen, E. J. Org. Lett. 1999, 1, 645;
2
(
b) Armstrong, A.; Goldberg, F. W.; Sandham, D. A.
Tetrahedron Lett. 2001, 42, 4585.
3
. Vinyl stannane see: Betzer, J.-F.; Delaloge, F.; Muller, B.;
Pancrazi, A.; Prunet, J. J. Org. Chem. 1997, 62, 7768.
. Vinyl iodide see: (a) Zoller, T.; Uguen, D. Tetrahedron
Lett. 1998, 39, 6719; (b) Takeuchi, R.; Tanabe, K.;
Tanaka, S. J. Org. Chem. 2000, 65, 1558.
5. The nomenclature exo-trans refers to the product of exo
Diels–Alder cyclisation where the protons at C4 and C5
are trans to each other. Likewise the exo-cis product
16 may not rearrange due to the bulky nature of the
4
tert-butyl group and due to the fact that some other
decompositional pathway competes with rearrangement
and Diels–Alder cyclisation. This is noticeable by the
greatly reduced yield of recognisable products from the
reaction of 16.

2
756
P. A. Clarke et al. / Tetrahedron Letters 43 (2002) 2753–2756
refers to the product of exo cyclisation where the protons
(100 ml) at 0°C and stirred for 3 h. At this time the
mixture was diluted with EtOAc, water and the aqueous
layer acidified with 1 M HCl until pH 2. The aqueous
layer was then extracted with EtOAc until no further
colour was taken into the organic phase. The combined
organics were then washed with brine, dried (MgSO4)
and evaporated to yield a residue which was filtered via a
plug of Celite to give the mono ester as an E/Z mixture
at C4 and C5 have the cis relationship. This terminology
has also been applied to the products of endo Diels–Alder
cyclisation. We believe that the use of this nomenclature
makes the structures of the products instantly recognis-
able: a feature that cannot often be said of the Seebach
like and unlike terminology.
. (a) Lilly, M. J.; Paddon-Row, M. N.; Sherburn, M. S.;
Turner, C. I. Chem. Commun. 2000, 2213; (b) Paddon-
Row, M. N.; Sherburn, M. S. Chem. Commun. 2000,
6
(
5.04 g, 57%).
A portion of this mixture (100 mg, 0.581 mmol) was
added to a solution of potassium tert-butoxide (66 mg,
2
215.
7
. Turner, C. I.; Williamson, R. M.; Paddon-Row, M. N.;
Sherburn, M. S. J. Org. Chem. 2001, 66, 3963.
0.587 mmol) in THF (15 ml) at 0°C and stirred for 3 h.
At this time the mixture was diluted with EtOAc, water
and the aqueous layer acidified with 1 M HCl until pH 2.
The aqueous layer was then extracted with EtOAc until
no further colour was taken into the organic phase. The
combined organics were then washed with brine, dried
8
. All yields refer to the isolated yields of single compounds
1
13
(
by H NMR, 400 MHz and C NMR, 100 MHz
spectroscopy) which have been purified by silica gel chro-
matography.
9
. Separated by HPLC on a chiralpak AD column; 5%
−
1
(MgSO
ester of fumaric acid as a white solid (86 mg, 86%). lH
400 MHz, CDCl ) 6.88 (1H, d, J=15.8 Hz), 6.67 (1H, d,
) and evaporated to yield the mono tert-butyl
4
EtOH/heptane, flow rate 1.0 ml min . Retention time for
1
2 of 9.06 min.
(
1
1
0. Eberle, M. K.; Weber, H.-P. J. Org. Chem. 1988, 53, 321.
1. Mono tert-butyl fumarate was not commercially avail-
able, so it was prepared as follows. Powdered maleic
anhydride (5.00 g, 50.99 mmol) was added to a solution
of potassium tert-butoxide (5.78 g, 51.50 mmol) in THF
3
J=15.8 Hz) and 1.52 (9H, s).
Attempts were made to combine these steps into a single-
pot procedure, but the best results were always obtained
using the two-step sequence detailed above.