Regioselectivity of the Claisen rearrangement in meta-allyloxy aryl ketones: An experimental and computational study, and application in the synthesis of (R)-(-)-pestalotheola D

Article abstract of DOI:10.1002/chem.201002757

A study of the regioselectivity of the Claisen rearrangement of meta-allyloxy aryl ketones showed that the electron-withdrawing carbonyl group has a major influence and strongly directs rearrangement to the more hindered ortho position. However, when the ketone is part of a ring structure, its electronic effect can be negated by conversion into its triisopropylsilyl enol ether, which dramatically reverses the regiochemistry of the Claisen rearrangement. DFT calculations suggest that the effect is electronic although there is also a steric effect of the bulky silyl group. This strategy for influencing the regiochemical outcome of the Claisen rearrangement was then employed in a short synthesis of the furo[2,3-g]chromene, (-)-pestalotheola D, that confirms the absolute stereochemistry of the natural product. A new strategy for influencing the regiochemical outcome of the Claisen rearrangement was developed and employed in a short synthesis of furo[2,3-g]chromene, (-)-pestalotheola D (see scheme), which confirms the absolute stereochemistry of the natural product.

Full text of DOI:10.1002/chem.201002757

DOI: 10.1002/chem.201002757
Regioselectivity of the Claisen Rearrangement in meta-Allyloxy Aryl
Ketones: An Experimental and Computational Study, and Application in the
Synthesis of (R)-(À)-Pestalotheol D
Catherine L. Lucas, Barry Lygo, Alexander J. Blake, William Lewis, and
Christopher J. Moody*[a]
Abstract: A study of the regioselectivi-
ty of the Claisen rearrangement of
meta-allyloxy aryl ketones showed that
the electron-withdrawing carbonyl
negated by conversion into its triiso-
propylsilyl enol ether, which dramati-
cally reverses the regiochemistry of the
Claisen rearrangement. DFT calcula-
tions suggest that the effect is electron-
ic although there is also a steric effect
of the bulky silyl group. This strategy
for influencing the regiochemical out-
come of the Claisen rearrangement
was then employed in a short synthesis
group has
a major influence and
strongly directs rearrangement to the
more hindered ortho position. Howev-
er, when the ketone is part of a ring
structure, its electronic effect can be
Keywords: Claisen rearrangement ·
of the furo
A
heterocycles
· natural product ·
lotheol D, that confirms the absolute
stereochemistry of the natural product.
total synthesis
Introduction
viously used by us in the synthesis of other natural prod-
ucts,[7–15] with the requirement for a regioselective sigma-
tropic rearrangement to C7 rather than C5 in the Claisen
precursor 1 (Scheme 1). We had previously observed good
regioselectivity in the related Claisen rearrangement of di-
methyl allyl ether 3,[12] and we now describe a detailed ex-
perimental and computational study of the Claisen rear-
rangement of 3-allyloxy aryl ketones leading to a successful
In the century since its discovery,[1] the thermal rearrange-
ment of aryl allyl ethers, universally known as the Claisen
rearrangement, has been the subject of numerous studies
and has found wide application in organic synthesis.[2] Nev-
ertheless there are still some features of this well known re-
action that are poorly understood. For example, the alkene
geometry resulting from rearrangement of a,a-disubstituted
allyl ethers seems difficult to predict or control, whilst the
regioselectivity in the rearrangement of meta-substituted
aryl allyl ethers remains unreliably predictable in spite of
experimental and computational studies.[3–6]
In connection with a planned synthesis of the furoACTHUNTGRNEUNG[2,3-
g]chromene pestalotheol D (see below), it was reasoned
that the dihydrofuran ring of the natural product could be
accessed using the Claisen rearrangement methodology pre-
[a] C. L. Lucas, Prof. Dr. B. Lygo, Prof. Dr. A. J. Blake, Dr. W. Lewis,
Prof. Dr. C. J. Moody
School of Chemistry, University of Nottingham
University Park, Nottingham NG7 2RD (UK)
Fax : (+44)114-951-3564
E-mail: c.j.moody@nottingham.ac.uk
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201002757: Full experimental
details, X-ray crystal structures of (Æ)-32 and (R)-31 (including .cif
files), HPLC data, coordinates for all transition-state structures re-
Scheme 1. Regioselective Claisen rearrangements: a) required for a syn-
thesis of pestalotheol D; b) previously observed regioselectivity in a re-
lated rearrangement.
1
ported, and copies of H and 13C NMR spectra).
1972
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 1972 – 1978

FULL PAPER
synthesis of pestalotheol D, confirming both its structure
and absolute stereochemistry.
ing in DMF, gave a mixture of regioisomeric products 9 and
10, easily separable and distinguishable by 1H NMR spec-
troscopy, with the 5-prenyl isomer 10 predominating for
both the alcohol and silyl ether substrates (Scheme 3). Al-
though this strategy does influence the outcome of the
Claisen rearrangement and produce some of the desired re-
gioisomer 9, the selectivity remains poor, and, significantly,
it involves an unattractive and avoidable change of oxida-
tion state in the substrate.
Results and Discussion
Regioselectivity in the Claisen rearrangement of 3-allyloxy
aryl ketones: Starting from the known chromanone 4, a nat-
ural product isolated from Calea cuneifolia,[16] and readily
prepared from 2’,5’-dihydroxyacetophenone using a litera-
ture procedure,[17] the dimethylallyl moiety was then incor-
porated via a copper-mediated alkylation[18,19] to give the in-
termediate alkyne 5, followed by hydrogenation over Lin-
dlarꢁs catalyst to give the intermediate dimethyl allyl ether 1
in 78% over the two steps. Claisen rearrangement of 1
under microwave irradiation occurred with complete regio-
selectivity, but unfortunately to the undesired 5-position
rather than the 7-position of the chromanone ring, giving
compound 6 in good yield (Scheme 2).
Scheme 3.
We therefore adopted an alternative strategy that in-
volved temporary masking of the chromanone carbonyl as
its triisopropylsilyl (TIPS) enol ether 11, readily prepared
from chromanone 1 in excellent yield. This silyl enol ether
11 was then subjected to microwave irradiation in DMF to
give a mixture of 7- and 5-prenyl chromenes 12 and 13 in a
1:1 ratio, albeit in poor 26% yield. Pleasingly, however, both
the yield and selectivity were vastly improved by switching
to toluene as solvent, and this gave an excellent yield (98%)
of separable regioisomers in 5.5:1 ratio in favor of the de-
sired 7-prenyl isomer 12 (Scheme 4).
Notwithstanding the aforementioned general effect of
meta electron-withdrawing versus electron-releasing groups,
this tactic of temporary removal of the electron-withdrawing
properties of a carbonyl substituent on an aromatic ring in
order to influence the regiochemical outcome of the Claisen
rearrangement has apparently not been employed previous-
ly, and therefore we briefly explored its generality, both ex-
perimentally and by computation. Thus, 7-hydroxy-1-tetra-
lone was converted into the propargyl ether 14 and hence
the allyl ether 15. Heating ether 15 in toluene under micro-
wave irradiation resulted in a highly regioselective Claisen
rearrangement in favor of the 8-prenyl isomer 17 in excel-
lent yield (92%; ratio 16/17 1:7.5). On the other hand, the
Scheme 2.
Clearly the rearrangement of the chromanone derivative
1 (Scheme 2) is very different to that of chromene 3 previ-
ously studied in our laboratory (Scheme 1),[12] and presuma-
bly reflects the electronic effect of the chromanone carbonyl
group. Although the regiochemistry of the Claisen rear-
rangement of meta-substituted phenyl allyl ethers is not easy
to predict, there is limited evidence that, generally, electron-
withdrawing meta-substituents favored rearrangement to the
more hindered ortho-position, with electron-releasing meta-
substituents favoring rearrangement to the para-position.[3,4]
In order to negate the apparently detrimental effect of the
chromanone carbonyl group, two strategies were considered.
The first involved reduction of the ketone in chromanone 1
to the corresponding chromanol 7 with sodium borohydride,
subsequently protected as its bulky tert-butyldiphenylsilyl
(TBDPS) ether 8. The subsequent Claisen rearrangements
of these two substrates, carried out under microwave heat-
Chem. Eur. J. 2011, 17, 1972 – 1978
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1973

C. J. Moody et al.
readily prepared TIPS enol
ether 18 underwent regioselec-
tive Claisen rearrangement in
the opposite sense, with the 6-
prenyl isomer 19 predominating
(100% yield; ratio 19/20 4.6:1)
(Scheme 4). Hence in the relat-
ed chromanone and tetralone
series of compounds,
a very
similar pattern of reactivity is
observed, and on switching the
p-acceptor properties of a car-
bonyl substituent to the p-do-
nating enol ether, the regiose-
lectivity of the Claisen rear-
rangement is inverted.
Since it was not clear wheth-
er the observed change in re-
gioselectivity was entirely elec-
tronic (p-acceptor vs p-donor)
in nature, or whether steric1 ef-
fects are also involved, addi-
tional examples were sought.
Consequently, the rearrange-
ment of the simplest possible
meta-acyl aryl dimethyl allyl
ether was investigated. 3-Hy-
droxyacetophenone was con-
verted into the allyl ether 22 by
way of the propargyl derivative
21. Heating in toluene gave a
mixture (80%) of Claisen rear-
rangement products 23 and 24
with, as expected, the 2-prenyl
isomer predominating (ratio 23/
24 1:2.4). Conversion of ketone
22 into the corresponding TIPS
enol ether 25 followed by heat-
ing in toluene gave a very simi-
lar mixture of regioisomeric
Claisen rearrangement products
26 and 27 in the ratio 1:2.2
(Scheme 4). Hence in this case,
the conversion of the methyl
ketone into its silyl enol ether
has little effect on the regiose-
Scheme 4.
lectivity of the Claisen rearrangement. This suggests that
steric factors do play a role, since the conformational con-
straints provided by the chromanone or tetralone rings
ensure that the steric effect of the bulky silyl group in enol
ethers 11 and 18 is more pronounced than in the enol ether
25 where it can rotate out of the way.
Computational Study
In an attempt to distinguish steric and electronic factors for
the observed changes in regioselectivity, we performed a
series of DFT transition-state calculations in which we com-
pared the effects of an electron-withdrawing ketone sub-
stituent with the corresponding enol and silyl enol ethers
(Table 1). This study was performed using B3LYP/6-31G*,
as previous studies on the Claisen rearrangement of allyl
phenyl ethers have indicated that this method gives consis-
tent results,[20,21] and because computation using a larger
1
Although the difference in regioselectivity in the Claisen rearrange-
ments of the alcohol 7 and its TBDPS-ether 8 suggests that steric fac-
tors may play a role, the substituents are at an sp3 center and therefore
geometrically distinct from the examples in question (1 vs 11, 15 vs 18)
that involve substituents on an sp2 center.
1974
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 1972 – 1978

Claisen Rearrangement in meta-Allyloxy Aryl Ketones
FULL PAPER
Table 1. Relative transition-state energies leading to ortho- and para-Claisen rearrangement products.
basis set (B3LYP/6-311++
G**) on selected examples
(Table 1, entries 1–3, 6) gave
near identical results. The pre-
dicted relative energies of the
chair-transition states for the
two rearrangement pathways
for a series of related 3-allyloxy
aryl ketones and their enol de-
rivatives, including substrates 1,
15, 18, 22 and 25 that had been
studied experimentally as de-
scribed above, are shown in
Table 1.
The results of these calcula-
tions are broadly consistent
with the experimental results
and clearly support the conclu-
sion that both electronic and
steric effects contribute to the
overall regioselectivities. Specif-
ically, the calculations suggest
that:
Entry
Substrate
Relative energy [kcalmolÀ1
]
ortho-TS
para-TS
1
2
0.0
0.0
0.8
3.2
3.1
i) The keto group has
a
strong directing effect fa-
voring the ortho-substituted
product (entry 1 vs entry 3).
ii) The gem-dimethyl has little
influence on the regiochem-
istry of the rearrangement
(entry 2 vs entry 3, entry 6
vs entry 7, and entry 9 vs
entry 10).
iii) Switching from ketone to
enol significantly reduces
the preference for the
ortho-substituted product
(entry 2 vs entry 6, entry 3
vs entry 7, and entry 4 vs
entry 8).
3
4
5
0.0
0.0
0.0
3.9
2.5
6
7
0.0
0.0
0.3
0.4
iv) A combination of the elec-
tronic effect and the steric
bulk of the TIPS-group
would appear to be primar-
ily responsible for the
switch in selectivity to the
para-substituted
product
8
0.0
0.8
(entry 7 vs entry 10 vs
entry 11).
À
v) Rotation about the Ar C
9
0.0
0.0
0.6
0.4
bond allows the acyclic
enol ether to minimize this
latter effect (entry 3 vs
entry 5 and entry 11 vs
entry 12).
10
Chem. Eur. J. 2011, 17, 1972 – 1978
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1975

C. J. Moody et al.
chemistry of the other chiral
centers in the compounds by
NMR spectroscopy. Although
the absolute stereochemistry of
pestalotheol D (31) was not re-
ported,[22] we assumed that it
would also have the (R)-config-
uration at C2, and in continua-
tion of our interest in other nat-
urally occurring chromenes,[14,23]
we set out to confirm the struc-
ture and stereochemistry of
pestalotheol D by synthesis
using the regioselective Claisen
rearrangement strategy devel-
Table 1. (Continued)
Entry
Substrate
Relative energy [kcalmolÀ1
]
ortho-TS
para-TS
11
12
0.5
0.0
0.0
2.0
Comparison with experimental data suggests that, for the
most part, the calculations predict the observed trends.
However, it has previously been suggested that B3LYP/6-
31G* may overestimate breaking and forming bond-lengths
in the transition state of the aromatic Claisen rearrangement
by ca. 0.05 ꢂ.[21] This may result in an over-estimation of the
amount of the ortho-product, as increased steric compres-
sion in a tighter transition state structure would be expected
to impact more on the pathway leading to the ortho-prod-
ucts.
oped in the above studies.
With the 7-prenylchromene derivative 12 now available,
silyl deprotection using tetra-n-butylammonium fluoride in
the presence of water restored the ketone 2, poised for the
introduction of the dihydrofuran ring (Scheme 6). Initially
this was carried out to give racemic pestalotheol D by epoxi-
dation (64%) using mCPBA, followed by treatment of the
epoxide 32 under basic conditions to effect ring closure. The
structure of (Æ)-epoxide 32 was confirmed by X-ray crystal-
lography (Figure 1). Alternatively, direct treatment of the
Synthesis of the furoACTHNUTRGNEUNG[2,3-g]chromene natural product pesta-
lotheol D: The pestalotheols were isolated in 2008 from Pes-
talotiopsis theae fungal spores growing on an unidentified
tree on a mountain in Hainan Province in China,[22] and as-
signed the unusual highly oxygenated structures 28–31
(Scheme 5), three of which contain the furoACTHNUGRTENUNG[2,3-g]chromene
nucleus. The structure of pestalotheol A 28 was confirmed
by X-ray crystallographic analysis, and the absolute configu-
rations of the secondary alcohols in both pestalotheols A
and C were determined as (R) by the use of the modified
Mosher method, thereby allowing assignment of the stereo-
Figure 1. X-ray crystal structure of (Æ)-7-((3’,3’-dimethyloxiran-2’-yl)-
methyl)-6-hydroxy-2,2-dimethylchroman-4-one (32).
crude product from the epoxidation reaction with base obvi-
ates the need for isolation of epoxide 32, and gives racemic
pestalotheol D (Æ)-31 in 88% yield over the two steps
1
(Scheme 6). The compound had identical H and 13C NMR
spectra to those reported in the literature.[22]
It remained to perform an asymmetric functionalization
of the prenyl double bond, and we elected to use the relia-
ble Sharpless asymmetric dihydroxylation methodology.[24]
Initial studies into the dihydroxylation of phenol 2 were un-
successful, and we had to resort to its protection as the
benzyl ether 33. Thereafter, dihydroxylation in the presence
of AD-mix-b proceeded smoothly, though owing to the in-
solubility of benzyl ether 33, some tert-butyl methyl ether
(MTBE) was added as a substitute for some of the tert-buta-
nol component of the usual solvent mixture.[25] This deliv-
Scheme 5. Structures of pestalotheols A–D; although the stereochemistry
of pestalotheol D (31) was not reported,[22] we propose it to have the
(R)-configuration at C2 as shown.
1976
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 1972 – 1978

Claisen Rearrangement in meta-Allyloxy Aryl Ketones
FULL PAPER
b, and confirmed by X-ray crystallography of the final prod-
uct (Figure 2). With the desired chirality installed, the secon-
dary hydroxyl was converted into the corresponding mesy-
late (used without purification), and hence the (S)-epoxide
35 (Scheme 7). Finally the chiral epoxide 35 was hydrogeno-
lyzed under basic conditions to effect the required 5-exo-tet
ring closure of the resulting phenolic epoxide in situ to pro-
vide (R)-pestalotheol D in excellent yield. Comparison with
the racemic sample by HPLC on a chiral stationary phase
suggested that the material had an enantiomeric excess of
87%, and its optical rotation [a]2D3 À7.16 (c = 0.2, MeOH)
is close to that reported.[22] Although the natural product
was described as a gum, we were able to recrystallize it, and
hence confirm the absolute configuration of our synthetic
material as (R) by X-ray crystallography (Flack parameter=
0.05; Figure 2).
Scheme 6. Synthesis of racemic pestalotheol D.
ered the (R)-diol 34 in excellent yield and 93% ee (by
HPLC comparison with the racemate), the stereochemistry
being assigned on the ample precedent when using AD-mix-
Figure 2. X-ray crystal structure of (R)-pestalotheol D (31).
The successful synthesis described herein not only con-
firms the absolute stereochemistry of the natural product,
but also features a new strategy for influencing the regio-
chemical outcome of the Claisen rearrangement of meta-al-
lyloxy aryl ketones.
Experimental Section
Full experimental details are given in the Supporting Information.
Acknowledgements
We thank the EPSRC for support of this work via the DTA Scheme.
[1] L. Claisen, Ber. Dtsch. Chem. Ges. 1912, 45, 3157.
[2] A. M. Martꢃn Castro, Chem. Rev. 2004, 104, 2939.
[3] W. N. White, C. D. Slater, J. Org. Chem. 1961, 26, 3631.
[4] J. M. Bruce, Y. Roshan-Ali, J. Chem. Soc. Perkin Trans. 1 1981,
2677.
Scheme 7. Synthesis of (R)-pestalotheol D.
Chem. Eur. J. 2011, 17, 1972 – 1978
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1977

C. J. Moody et al.
[5] S. Yamaguchi, M. Ishibashi, K. Akasaka, H. Yokoyama, M. Miyaza-
wa, Y. Hirai, Tetrahedron Lett. 2001, 42, 1091.
[6] F. C. Gozzo, S. A. Fernandes, D. C. Rodrigues, M. N. Eberlin, A. J.
Marsaioli, J. Org. Chem. 2003, 68, 5493.
[18] J. D. Godfrey, R. H. Mueller, T. C. Sedergran, N. Soundararajan,
V. J. Colandrea, Tetrahedron Lett. 1994, 35, 6405.
[19] R. J. Detz, H. Hiemstra, J. H. van Maarseveen, Eur. J. Org. Chem.
2009, 6263.
[7] C. J. Moody, J. Chem. Soc. Perkin Trans. 1 1984, 1333.
[8] T. Martin, C. J. Moody, J. Chem. Soc. Perkin Trans. 1 1988, 241.
[9] C. J. Davis, T. E. Hurst, A. M. Jacob, C. J. Moody, J. Org. Chem.
2005, 70, 4414.
[20] O. Wiest, K. A. Black, K. N. Houk, J. Am. Chem. Soc. 1994, 116,
10336.
[21] M. P. Meyer, A. J. DelMonte, D. A. Singleton, J. Am. Chem. Soc.
1999, 121, 10865.
[10] A. M. Jacob, C. J. Moody, Tetrahedron Lett. 2005, 46, 8823.
[11] C. S. P. McErlean, C. J. Moody, J. Org. Chem. 2007, 72, 10298.
[12] C. L. Coombes, C. J. Moody, J. Org. Chem. 2008, 73, 6758–6762.
[13] L. Guillonneau, D. Taddei, C. J. Moody, Org. Lett. 2008, 10, 4505.
[14] M. Bruder, S. J. Smith, A. J. Blake, C. J. Moody, Org. Biomol. Chem.
2009, 7, 2127.
[22] E. Li, R. Tian, S. Liu, X. Chen, L. Guo, Y. Che, J. Nat. Prod. 2008,
71, 664.
[23] M. Bruder, C. J. Moody, Synlett 2008, 575.
[24] H. C. Kolb, M. S. VanNieuwenhze, K. B. Sharpless, Chem. Rev. 1994,
94, 2483.
[25] H. Becker, K. B. Sharpless in Asymmetric Oxidation Reactions (Ed.:
T. Katsuki), Oxford University Press, Oxford, 2001, pp. 81.
[15] M. Bruder, P. L. Haseler, M. Muscarella, W. Lewis, C. J. Moody, J.
Org. Chem. 2010, 75, 353.
[16] T. O. Lourenco, G. Akisue, N. F. Roque, Phytochemistry 1981, 20,
773.
[17] U. Gerlach, J. Brendel, H.-J. Lang, E. F. Paulus, K. Weidmann, A.
Brueggemann, A. E. Busch, H. Suessbrich, M. Bleich, R. Greger, J.
Med. Chem. 2001, 44, 3831.
Received: October 27, 2010
Published online: January 10, 2011
1978
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 1972 – 1978