вход по аккаунту


The BelluЦClaisen Rearrangement.

код для вставкиСкачать
J. Gonda
Sigmatropic Rearrangement
The Belluš–Claisen Rearrangement
Jozef Gonda*
allylic compounds · ketenes · macrocycles · ring
expansion · sigmatropic rearrangement
Dedicated to Professor Daniel Belluš
on the occasion of his 65th birthday
Among the reactions available to synthetic chemists for the
construction of new CC bonds, the Claisen rearrangement is one of
the most powerful, elegant, and well-characterized methods. A genuinely new variant, the Belluš–Claisen rearrangement came to light a
quarter of a century ago: The reaction of an allylic ether, thioether, or
amine with a ketene leads through a [3,3] sigmatropic bond reorganization of a zwitterionic intermediate to an E unsaturated ester, thioester, or amide. When applied to cyclic allylic substrates, a ringenlargement by four carbon atoms in one step provides medium-ring
unsaturated E-configured lactones, thiolactones, and lactams. The
scope of the Belluš–Claisen rearrangement and the optimum reaction
conditions will be discussed in this Minireview.
up for a [3,3] sigmatropic rearrangement to form the unsaturated ester 6 of
the Johnson–Claisen type.[3] In a seminal paper demonstrating the synthetic
usefulness of the newly discovered
reaction, Malherbe and Belluš[4] employed readily accessible, highly electrophilic dichloroketene,[5] which reacted with allylic ethers in
a similar way to the rather exotic ketene 1. Thus, an
1. Introduction, Mechanism, and Scope
The synthesis of complex organic molecules from simple
chemical building blocks represents an enticing and profound
intellectual challenge, which lies at the heart of chemistry and
has immense practical value. In this context, the discovery of a
new carbon–carbon bond-forming reaction, especially a
simple and reliable one, is of evident benefit. A quarter of a
century ago, a serendipitous experimental result led Belluš
and Malherbe to the discovery of a conceptually new
intermolecular method for creating very reactive zwitterionic
intermediates that are poised for a [3,3] sigmatropic Claisentype rearrangement.[1]
In an attempt to prepare the 2-chlorocyclobutanone
derivative 4 (Scheme 1), a key compound in a versatile
synthesis of pyrethroid insecticides,[2] Belluš and Malherbe
treated the chloro(trichloroethyl)ketene 1 (generated in situ)
with the allylic ether 2 at ambient temperature. Besides the
expected [2+2] cycloaddition product 4, the g,d-unsaturated
ester 6 was also isolated. Apparently, the nucleophilic oxygen
atom of the allylic ether in 2 can compete successfully with the
double bond in 2 for the electrophilic ketene. Thus, not only is
the expected 1,4-dipolar intermediate 3 formed, but also the
zwitterionic intermediate 5. The zwitterion 5 is beautifully set
[*] Prof. Dr. J. Gonda
Institute of Chemistry, P. J. Safarik University
Moyzesova 11, 041 54 Kosice (Slovak Republic)
Fax: (+ 421) 556-222-124
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Scheme 1. First observation of the Belluš–Claisen rearrangement.
DOI: 10.1002/anie.200301718
Angew. Chem. Int. Ed. 2004, 43, 3516 –3524
Belluš–Claisen Rearrangement
unprecedented two-component Claisen-type rearrangement
had been discovered, for which Belluš and Malherbe used the
name “ketene-Claisen rearrangement”.
It was later shown by other authors that in the case of
allylic amines, the aza analogue of the zwitterion 8 (X = N-R;
Scheme 2) can be formed from the allylic amine 7 (X = N-R)
Scheme 2. Mechanism of the Belluš–Claisen rearrangement.
upon reaction with a ketene 7 a. This ketene may be employed
as an isolated substance,[6] or may be formed in situ either by
Zn-induced dehalogenation of a 2-chloroacyl chloride,[7]
photolysis of a chromium–carbene complex,[8] or dehydrohalogenation of an acyl chloride.[9] Under the latter reaction
conditions the allylic amine 7 (X = N-R) may react with the
acyl chloride (e.g. 7 b) to form an acyl ammonium salt 7 c (X =
N-R), which, of course, is structurally unable to undergo a
[3,3] sigmatropic rearrangement. Only after fast in situ
dehydrohalogenation of 7 c occurs to afford the ketenederived aza analogue 8 can the fast [3,3] bond reorganization
take place.
The base-induced transformation of the adduct 7 c (X =
N-R) into the zwitterion 8 (X = N-R) can be facilitated by the
complexation of the acyl carbonyl group with a variety of
Lewis acids, which may be added in either stoichiometric or
catalytic[10] amounts. When used stoichiometrically, both
achiral[11] and chiral[12] Lewis acids may be employed. Moreover, Lewis acid activation of 7 c allows the participation of
ketenes that are less electrophilic than dichloroketene in the
ketene Claisen rearrangement to give 9 (X = N-R). When X =
O or S, direct formation of 8 is at present confined to
Jozef Gonda, born in 1957 in Presov (Slovakia), received his PhD in 1986 from the
University of P. J. Safarik in Kosice, where
he has been a full professor since 2003.
From 1991 to 1995 he undertook postdoctoral research with Prof. Daniel Belluš at
the University of Fribourg and Ciba-Geigy in
Basel on [3,3] sigmatropic reactions. He
completed his habilitation in Kosice in
1995. His main research interests include
the stereoselective synthesis of natural products, in particular rare amino acids and
Angew. Chem. Int. Ed. 2004, 43, 3516 –3524
electrophilic ketenes: dichloroketene chloro(trichloroethyl)ketene @ chloro(methyl)ketene > chloroketene chloro(cyano)ketene. They represent synthetic equivalents of the
corresponding chloro-free ketenes, as the reductive removal
of chloro groups from the position a to the carbonyl group in
the rearrangement products 9 is easy to achieve preparatively.
Several methods are available, for example, treatment with
Zn/AcOH at elevated temperature,[4] with Zn/Ag alloy,
MeOH/AcOH (5:1) at room temperature,[7] with Zn, NH4Cl,
MeOH at room temperature or above,[13] with H2, catalytic Pt/
C,[14] or with Bu3SnH, AIBN, toluene at reflux.[15]
Alternative names, such as the “zwitterionic aza-Claisen
reaction”[16] and “acyl-Claisen rearrangement”[17] have also
appeared in the literature for the aza variant of the keteneClaisen rearrangement, even though the prerequisite and at
present unquestioned occurrence of 8 had already first been
proposed 25 years ago in the original paper of Malherbe and
Belluš.[4] Other individual names bearing at least one surname
of the discoverers have been used for the underlying reaction
in the literature.[18] Herein I will use the term Belluš–Claisen
rearrangement (BC rearrangement), in analogy with the
already established terms Eschenmoser–, Johnson–, and
Ireland–Claisen rearrangements.[19]
In general, the BC rearrangement proceeds very fast at
room temperature or below. This is in accordance with the
well-known fact that the presence of a positively charged or
chelated heteroatom in position 3[20, 21] of the rearranging 1,5hexadiene system, or a negatively charged substituent at
C2,[19c, d, 22] provides a significant rate acceleration of the
Claisen [3,3] bond reorganization. For the first time within the
Claisen family of rearrangements, the rearranging intermediate 8 contains both rate-accelerating charges in one molecule
working in synergy![23]
In terms of the design of target molecules that can be
prepared by using the BC rearrangement, at least one
remarkable synthetic transformation deserves particular
attention because of its uniqueness amongst the Claisen
family of [3,3] rearrangements: If the nonparticipating alkyl
substituent on the heteroatom X in 7 (X = O, S, or N-R) is
connected to the adjacent sp3 carbon atom of the allylic group
to form a ring, then the BC rearrangement results in a ring
enlargement by four carbon atoms in a single step, thus
providing medium-sized lactones (Schemes 3–5), thiolactones
(Schemes 7 and 8), or lactams (Schemes 9, 11, and 12). As a
rule, the BC reaction generally affords the E isomer of the
unsaturated rearrangement product stereoselectively. The
stereoselective synthesis of these compounds is much more
tedious by other routes. The yields of the ring-enlarged
products are typically moderate to excellent, up to 96 %.
Herein the research efforts of the first twenty-five years
since the publication of the seminal paper in 1978 are
summarized,[4] with a focus on the synthetic aspects of
chemo-, regio-, stereo-, and enantioselective applications of
the carbon–carbon bond-forming Belluš–Claisen rearrangement.
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Gonda
2. Allylic Ethers
Shortly after the first observation of the Belluš–Claisen
rearrangement with the chloro(trichloroethyl)ketene 1
(Scheme 1), a number of acyclic allylic ethers were also
subjected to reaction with the well-investigated ketene[5]
dichloroketene (11).[4, 14] The ketene 11 was generated in situ
from trichloroacetyl chloride by dechlorination with activated
zinc in diethyl ether, and the expected a,a-dichloro-g,dunsaturated esters, analogous to 6, were formed in moderate
yields (40–65 %). The moderate yields suggest that competitive formation of the corresponding cyclobutanones (analogous to 4) takes place when allylic ethers containing a
nucleophilic trialkyl-substituted double bond are used. The
use of 1,2-dimethoxyethane as a co-solvent capable of
complexing the Lewis acid ZnCl2 generated in situ favors
the [2+2] cycloaddition pathway over the [3,3] sigmatropic
rearrangement pathway.[24]
In spite of the limited chemoselectivity and moderate
yields, Belluš and Malherbe immediately decided to further
elucidate the newly discovered [3,3] sigmatropic rearrangement mode, this time with the cyclic allylic ether cis-2-vinyl-6methyltetrahydro-4H-pyran (10). A successful outcome
would result in a very useful synthetic methodology, that is,
a conceptually new and simple one-step route to mediumsized lactones, whose synthesis is not only a challenge,
because their formation is disfavored by both enthalpy and
entropy, but also of practical importance, as they occur in a
variety of interesting natural products.[25] In fact, the straightforward synthesis of the naturally occurring 10-memberedring macrolides ( )-phoracantholide I (13) and ( )-phoracantholide J (15)[26] was successful on the first attempt, just as
designed![4] (Scheme 3). The synthesis exemplified the unique
product (ñmax = 1810 cm1) was observed in the crude reaction
mixture in 10–15 % yield, but did not survive chromatography
on silica gel.
Along similar lines, several authors investigated the scope
of the BC rearrangement in terms of the possible ring sizes of
the cyclic allylic ethers used as starting materials. Taylor and
co-workers[15] sought an efficient procedure for the construction of unsaturated nine-membered lactones for use in the
total synthesis of diverse bioactive marine metabolites. After
due consideration of the methods known at the time for
cyclizations into medium-sized rings, they decided to utilize
the BC rearrangement (Scheme 4) and reported: a) that this
Scheme 4. Ring expansion to nine-membered lactones. AIBN = azobisisobutyronitrile.
chemistry is extremely useful for the preparation of lactones
such as 17 and 18 from the corresponding 2-vinyltetrahydrofurans 16 and dichloroketene (11; generated in situ by a
CCl3COCl/Zn procedure), b) that the methodology is compatible with a bulky silyl ether substituent, for example, as in
16 c, and c) that the cyclic E alkenes 17 thus formed can
undergo dechlorination in the presence of Bu3SnH and AIBN
with concomitant double-bond isomerization to provide the
thermodynamically preferred Z-unsaturated lactones 18.
An impressive result was obtained by Dean and Taylor.[27]
They subjected 3,3-diethyl-2-vinyloxetane (19) to the usual
ring expansion conditions with 11 generated from CCl3COCl/
Zn (Scheme 5). The reaction with 19 was far more rapid than
with 2-vinyltetrahydrofurans, probably due to the ring-strain
relief during the reaction. The remarkable unsaturated eightmembered lactone 20, which contains an E double bond, was
Scheme 3. Synthesis of the macrolides ( )-phoracantolide I (13) and
( )-phoracantolide J (15).
applicability of the new reaction to the transformation of an
n-membered, 2-vinyl-substituted cyclic ether into an unsaturated, (n + 4)-membered lactone. Interestingly, both 10-membered E-configurated lactones 12 and 14 assume two rapidly
interconverting ground-state conformations, related to each
other by a 1808 flip of the double bond.[14] A cyclobutane by-
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Scheme 5. Ring-enlargement reaction of the 2-vinyloxetane 19 and the
2-vinyloxirane 22. Ts = toluenesulfonyl.
Angew. Chem. Int. Ed. 2004, 43, 3516 –3524
Belluš–Claisen Rearrangement
isolated in 38 % yield. Although the double bond did not
isomerize at room temperature, isomerization to the thermodynamically stable Z isomer 21 occurred quantitatively in the
presence of a catalytic amount of p-TsOH in aqueous
acetonitrile at reflux.
In striking contrast with unsuccessful attempts at the ring
expansion of six-[1] and four-membered[27] 2-vinyl cyclic ethers
by reaction with dichloroketene (11) generated in situ from
dichloroacetyl chloride by dehydrochlorination with an
organic base, these appear to be the only viable reaction
conditions for the ring expansion of a series of 2-vinyloxiranes.[9] For example, 2-styryloxirane 22 reacted with 11
(from CHCl2COCl/Et3N) to furnish, presumably via the
zwitterion 23, the 3,3-dichloro-4-phenyl-2,3,4,7-tetrahydrooxepin-2-one 24 in a good 73 % yield (Scheme 5). However, 24
was not formed if 11 was generated in situ from CCl3COCl
and zinc, probably because of a fast ZnCl2-catalyzed polymerization of the oxirane 22. Only the unsaturated sevenmembered lactones bearing a Z double bond were isolated.[9]
3. Allylic Thioethers
Given the higher nucleophilicity of sulfur versus oxygen, it
is not surprising that in their first publication Malherbe and
Belluš already reported the successful application of the new
[3,3] bond reorganization reaction to several acyclic allylic
thioethers, and, as a curiosity, also to an allylic selenoether
(Scheme 6).[4, 14]
Scheme 7. The contribution of Vedejs et al. to the synthesis of 10- and 11-membered lactones by the Belluš–Claisen rearrangement.
and Buchanan stated that “there is good reason to believe
that efficient (cyclic) sulfide ring expansion with dichloroketene is a general reaction”.[29]
Rossini et al.[30] found that the readily accessible five- and
six-membered cyclic thioketals of a,b-unsaturated cyclopentenones and cyclohexenones, such as 36, underwent an
efficient ring enlargement to bicyclic thiolactones, such as
37, in yields of 78–88 % (Scheme 8). Interestingly, in this case
11 had to be formed in situ by dehydrohalogenation of
dichloroacetyl chloride with Et3N. No cyclobutanone products were observed.[24, 30]
Scheme 6. Examples of rearrangements of the allylic thioether 25 and
selenide 28.
Scheme 8. An example of a four-carbon-atom ring enlargement of a
cyclic thioketal.
The cyclic version of this reaction was originally reported
to proceed in a disappointing yield of 8 %.[14] However, Vedejs
et al. showed that an optimized ring-expansion procedure
with dichloroketene (11) generated by the slow addition with
a syringe pump of excess CCl3COCl to the substrate and Zn/
Cu couple in ether at reflux gave 10-membered thiolactones
in consistently good yields (75–85 %; Scheme 7). The double
bond in the 10-membered ring is formed exclusively with the
E configuration.[28, 29]
Thiolactones such as 32 and 34 are not only interesting as
final targets but can also be used as intermediates in the total
synthesis of natural products. Thus, Vedejs et al. used 32 in
studies towards the stereoselective synthesis of a key fragment of the natural macrolide erythronolide, and 34 was used
to demonstrate a convenient transformation of the initially
formed macrocyclic w-hydroxyalkyl thiolactones into lactones such as 35 by means of an intramolecular S to O acyl
shift. This elegant ring expansion by one carbon atom was
catalyzed by anhydrous camphorsulfonic acid (CSA). Vedejs
Angew. Chem. Int. Ed. 2004, 43, 3516 –3524
Much effort has been invested in the study of chirality
transfer and chirality induction in standard Claisen [3,3]
rearrangements.[19c,d, 31] In a number of early publications,
Belluš, Ernst, and co-workers studied the effect of chiral
substituents on the BC reaction of allylic thioesters. Thus,
nearly complete (> 98 %) 1,3-chirality transfer occurred from
the allylic C center of a CS bond in 41 to the stereogenic
allylic C center in the new CC bond in the chiral thioester 42,
as determined by 1H NMR spectroscopic shift experiments
with (R)-()-TFAE (2,2,2-trifluoro-1-(9-anthryl)ethanol; Table 1).[32, 33] Conceivably, the zwitterionic intermediate 38
(Figure 1) is involved in the suprafacial [3,3] bond reorganization step, thus leading to a new CC bond oriented
exclusively “cis” with respect to the original CS bond. The
stereochemical integrity of 42 was conserved during the
subsequent dechlorination with Zn in AcOH at 100 8C to give
the thioester 43.
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Gonda
As a result of the very mild experimental conditions, the 1,2-asymmetric
induction in the BC reaction of the
Product of
Product after
treatment with Zn/AcOH
allylic thioether 50, derived from ()-l(Figure 1)
(yield [%])
(yield [%])
lactic acid, with dichloroketene (11) is
close to 100 % (Table 1). The resulting
diastereomer 51 has the syn configura38
tion. Three subtle factors allow 11 to
differentiate between the two possible
41, 96 % ee[32, 33]
42, 95 % ee (80)
43, 95 % ee (92)
diastereotopic faces of 50 in favor of the
depicted intermediate 40 (Figure 1):
First, the minimized 1,3-allylic strain as
a stereochemical controlling factor;[37]
[32, 33]
44, 85 % ee
45, 83 % ee (76)
46, 66.4 % ee (66)
second, the preferred equatorial position of the chiral substituent at the
terminal allylic sp2 carbon in the interlike 39
mediate 40; and third, the highly preferred anti quasi-nucleophilic attack of
47, 85 % ee[34]
48, 83 % ee (86)
49, 83 % ee (98)
the terminal sp2 carbon atom of the
ketene with respect to the (CO) bond
on the chiral substituent, as has also
been observed in the mechanistically
related [2,3] sigmatropic Wittig rear50, 94 % ee
51, d.r. 97:3 (82)
52, cis/trans 97:3 (82)
rangement[38] and later in aza-BC reactions.[16, 39]
In contrast to the complete diasterlike 40
eoselectivity observed in the BC rearrangement of 50, only d.r. 75:25 and
53, 95 % ee[35]
54, d.r. 90:10 (89)
55, d.r. n.d. (95)
57.5:43.5 were observed in the Ireland–
Claisen rearrangement of structurally
[a] R = isopropyl, n.d. = not determined. Bn = benzyl, TBDMS = tert-butyldimethylsilyl.
related silylketene acetals derived from
l-lactaldehyde[40] and d-glyceraldehyde,
An analogous thermal (140 8C) Johnson–
Claisen rearrangement with triethyl orthoacetate furnished
only a 1:1 mixture of diastereomeric products.[42]
In comparison to results with 50, slightly lower 1,2-syndiastereoinduction was observed in the BC reaction of Ntoluenesulfonyl- (53; Table 1) and analogous N-Boc-protected chiral thioethers with dichloroketene (11).[35] As the
Figure 1. Chairlike reactive zwitterionic intermediates derived from
basicity of the allylic nitrogen atom in 53 is greatly reduced
dichloroketene (11) and the chiral allylic thioethers 41, 44, and 50
by either of these protecting groups, chemoselective [3,3]
(Table 1).
rearrangement of the allylic thioester to afford syn-54 was
observed. An attempt was made to understand the difference
between the diastereoselectivities in the rearrangements of 50
and 53 by calculating all possible (two chairlike and two
The E allylic thioethers 44 and 47 served as models for
boatlike) zwitterionic intermediates at a semiempirical AM1
scrutinizing the stereochemical course of the BC reaction with
SCF-MO level. For 50, calculation revealed that there is an
11 in open-chain systems.[32–34a] Very high enantioselectivity
energy difference of 17.15 kJ mol1 in favor of a “syn”
was also observed for the 1,3-chirality transfer from 44 and 47
(85 % ee) to the unsaturated E thioesters 45 and 48 (83 % ee),
chairlike intermediate 40 relative to the next-best chairlike
as well as excellent chemical yields. A chairlike zwitterion 39
intermediate, which would lead to an anti diastereomer of 51.
is responsible for the high level of chirality transfer. InterestThus, the calculated diastereoselectivity of d.r. > 99.9:0.1 is in
ingly, if both allylic S-alkyl and allylic O-alkyl groups are
perfect agreement with the experimental findings. In contrast,
present in the molecule, as in 47, dichloroketene (11) reacts
the calculated best chairlike zwitterionic intermediate arising
highly chemoselectively with the more nucleophilic S atom,
from 53 and 11, which leads to 54 as the main syn isomer, is
thus leading in this case exclusively to the thioester 48.[34] The
favored by only 1.08 kJ mol1 relative to the energetically
recently published Pd-catalyzed synthesis of chiral allylic
closest (boatlike(!)) intermediate. These results are indicative
thioethers could expand the scope of the thia-BC rearrangeof a more-complex electronic and steric interaction in the
ment by making the starting materials more readily accessulfonium intermediate derived from 53 and 11 than in that
derived from 50. Conceivably, electronic and steric fine-
Table 1: Selected examples of the stereo- and chemoselective BC rearrangement of allylic
thioethers with dichloroketene (11).[a]
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2004, 43, 3516 –3524
Belluš–Claisen Rearrangement
tuning of the two substituents on the allylic nitrogen atom in
compounds such as 53 would offer many synthetic opportunities for the use of BC rearrangements with differing
outcomes in terms of syn/anti selectivity or even of chemoselectivity (reaction of the allylic sulfur atom versus the allylic
nitrogen atom).
The reductive treatment of the rearrangement products
45, 51, or 54 with Zn in acetic acid at 100 8C initiates three
simultaneous reactions: a,a-dechlorination, deprotection of
O or N, and a ring closure to afford the novel chiral gbutyrolactones 46 or 52, or the lactam 55 (Table 1). These
compounds possess interesting substitution patterns and
retain the stereochemical integrity of their precursors.
4. Allylic Amines
Curiously enough, the aziridine 56 was the first substrate
used in a Belluš–Claisen rearrangement of a tertiary allylic
amine and dichloroketene (11).[9] In analogy with 24, the
double bond in the resulting 2,3,4,7-tetrahydro-1H-azepinone
57 has the Z configuration. (Scheme 9). So far, this remains
the only reported BC reaction of a 2-vinylaziridine.
Scheme 10. Preparative access to the azabicyclo[4.3.0] skeleton of
monoterpene alkaloids.
A number of papers report on that highly characteristic
transformation of the BC rearrangement: the ring expansion
of 2-vinyl cyclic amines by four carbon atoms in one step to
prepare nine and ten-membered lactams, which can be useful
precursors to natural products or their subunits. However,
only Edstrom[7, 44] has utilized 11, generated in situ from
trichloroacetyl chloride and a Zn/Cu couple, to access the
unsaturated lactams 65 and 66 (Scheme 11). Evidence for a
Scheme 9. The first reported example of an aza-BC reaction.
The second reported case of an aza-BC rearrangement is
also rather unconventional, as it represents the only known
reaction that utilizes a stable isolated ketene. Roberts et al.[6]
described an interesting skeletal rearrangement of the Nmethylazanorbornene 58 a with isolated diphenylketene
(59 a) to form the bicyclic piperidone 61 in moderate yield
(Scheme 10). For the N-benzyl derivative of 58, a dramatic
increase in the reaction rate was reported when the reaction
was performed in an ultrasonic bath.
Pombo-Villar and co-workers ingeniously applied the
principle of the transformation of 58 described by Roberts
et al. to an enantioselective total synthesis of the three
monoterpene alkaloids ()-2-skytanthine, (+)-epidihydrotecomanine, and ()-N-demethyl-d-skytanthine, from a common enantiomerically pure tricycle 62, which in turn was
prepared in six steps from the product 61 b of the aza-BC
rearrangement of 58 b with dichloroketene (11;
Scheme 10).[13, 43] The cis ring junction in 61 is the result of a
concerted [3,3] rearrangement of a transient intermediate 60
via a boat conformation. Because of structural constraints,
this system is unable to adopt a chair conformation. As a
precursor for 58 b, (S)-1-phenylethylamine was used as the
original source of chirality for the whole synthesis. For the BC
reaction with 58 b, 11 had to be generated in situ from
dichloroacetyl chloride by using diisopropylethylamine in
CH2Cl2 at 0 8C.
Angew. Chem. Int. Ed. 2004, 43, 3516 –3524
Scheme 11. A route to functionalized indolizidine and quinolizidine ring
systems based on the aza-BC rearrangement.
zwitterionic intermediate of the type 8 (from 63) was obtained
by 1H and 13C NMR spectroscopy. Subsequent exposure of
the lactams 65 and 66, or their a-dechlorinated derivatives, to
certain electrophiles triggered a simultaneous transannular
cyclization–debenzylation. For example, the indolizidine 67
and quinolizidone 68 were formed upon treatment with
iodine. Their skeletons are widespread among alkaloid
natural products.
Analogous sequences that start with a BC rearrangement,
which is then followed by a stereoselective transannular
reaction, have been thoroughly investigated by Nubbemeyer
and co-workers.[11, 16, 45] As a result of their pioneering
expansion of the scope of the aza-BC reaction by elaboration
of an alternative approach to reactive 1,5-hexadiene intermediates of the type 8 via their N-acyl ammonium precursors
of the type 7 c, they were not confined to the use of highly
electrophilic ketenes. Additional substitution patterns, intro-
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Gonda
duced in the form of substituents on the acyl group thus
became accessible in the products of the [3,3] rearrangement.
The experimental variant of the BC rearrangement of
Nubbemeyer and co-workers will be exemplified by means of
the reaction of the optically active 2-pyrrolidine acrylic ester
69 with various acyl chlorides (Scheme 12).[11, 16] They found
1) The rearrangement is not restricted to activated ketenes.
2) Special two-phase conditions are recommended: A slurry
of the allylic amine and solid K2CO3 in anhydrous CHCl3
or CH2Cl2 is treated with the acyl chloride and the Lewis
acid trimethylaluminum (> 20 mol %).
3) The major competing reaction is a von Braun process
involving nucleophilic attack of a chloride ion on the
intermediate acyl ammonium salts, for example, on 70 to
give 75.[11]
4) The use of acyl fluorides instead of acyl chlorides as in situ
sources of ketenes seems to completely suppress the
von Braun process, as no formation of N-acyl ammonium
fluorides of the type 70 is observed.[45a,c]
5) Complete 1,3-chirality transfer takes place; the orientation of the CN bond dictates that of the new CC bond.
Deprotonation of the acyl ammonium salt (e.g. 70)
generates predominantly the Z-enolate structure corresponding to 71 because of minimized 1,3-diaxial and other
steric interactions. As a result, 3,4-trans-difunctionalized
2-azonines 73 are the major diastereomers produced by
the reaction.
6) Because of restricted flipping of the E double bond with
respect to the ring, in nine- and ten-membered lactams,
such as 66, 73, and 74, planar-chirality phenomena
frequently occur.[45d] One can take advantage of these
phenomena in introducing new stereogenic centers in
total syntheses of bicyclic natural products, such as the
Scheme 12. The diastereoselective synthesis of nine-membered
lactams via the ketene precursors 70 by Nubbemeyer and co-workers.
E = CO2Et.
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Hegedus and co-workers[8] extended the aza-BC rearrangement to include photogenerated ketenes originating
from the photolysis of chromium–carbene complexes, such as
[(CO)5Cr=C(OR)Me] in THF. These ketenes can be used to
install two geminal substituents a to the carbonyl carbon atom
in products of the [3,3] rearrangement, for example, amethyl,a-n-butoxy substitution in skeletons corresponding to
61 (R1 = Bn; 51 % yield), 65 (66 % yield), and 77 (Table 2,
68 % yield). Whereas the strained compounds 58 (R1 = Bn)
underwent an efficient BC reaction with n-butoxymethylketene even in the absence of a Lewis acid, the less-strained 63
required the mild Lewis acid ZnCl2, and the acyclic allylic
amine 76 required the presence a strong Lewis acid, such as
Me2AlCl. More electron-rich carbene complexes, such as
CO5Cr=C(NR2)H, were much less efficient, giving the
corresponding rearranged products in only low yields
(< 22 %). The results of Hegedus and co-workers are
mechanistically significant for the aza-BC rearrangement, as
they reveal that even with electron-rich ketenes an “acyl
ammonium intermediate” of the type 7 c need not be formed,
provided that the ketene part of the reacting zwitterionic 1,5hexadiene system of the type 8 is chelated by an appropriate
Lewis acid.
Nubbemeyer applied the two-phase experimental conditions described under b) above for the cyclic compound 69 to
the BC rearrangement of optically active open-chain N-allyl
pyrrolidones.[39, 46] As in the case of structurally similar chiral
thioethers (e.g. 50), high 1,2-asymmetric induction was
observed in the construction of a new CC bond adjacent
to the stereogenic C center bonded to O. The yields of the
isolated g,d-unsaturated amides thus formed were 60–90 %.
A powerful feature of the aza variant of the BC rearrangement involving N-acyl ammonium intermediates of the
type 7 c, in which both the ketene and allylic E or Z double
bonds can bear electronically diverse substituents, is the
capacity to build diverse functional and stereochemical
arrays. This capacity was demonstrated by Nubbemeyer and
co-workers (with Me3Al as the activating Lewis acid)[11, 16, 45]
and Yu et al. (without a Lewis acid).[47] Recently MacMillan
and co-workers[10] reported the successful catalysis of a
diastereoselective [3,3] rearrangement of the type 8!9 with
a variety of Lewis acids, including Yb(OTf)3, AlCl3,
Ti(OiPr)2Cl2, and TiCl4·2 thf. Although only the BC reactions
with (E)-76 and (Z)-78 can be compared directly so far,
Table 2 suggests the preparative superiority of the reaction
conditions B of MacMillan and co-workers in terms of
chemical yield and diastereoselectivity. Significant structural
variation in the substituent on the double bond of the allyl
group (H, aryl, or Cl instead of Me) as well as in the acyl
chloride (NPht, SPh, or OBn instead of Me) is also possible
without loss of yield or diastereoselectivity. If 3,3-disubstituted allylic morpholines are used, the chairlike transition
state controls the p-facial discrimination, so that one can
introduce elusive quarternary carbon stereocenters stereoselectively in d.r. > 97.5:2.5.
High-yielding experimental conditions enabled Dong and
MacMillan[17] to design and perform a series of tandem azaBC rearrangements that could be used to build up functional
and stereochemical acyclic systems of considerable complex-
Angew. Chem. Int. Ed. 2004, 43, 3516 –3524
Belluš–Claisen Rearrangement
(Scheme 14). A survey of ligand/metal-salt combinations
revealed that the (R,R)-arbox framework 87 exhibited
superior levels of enantioselectivity. For example, its use as
a chiral Lewis acid in the BC reaction of achiral 84 with
Table 2: Comparison of different experimental conditions for the
generation of methylketene for the aza-BC rearrangement.
(yield [%])
77 (41)
79 (38)
77 (92)
79 (74)
> 99:1
[a] Conditions: A: K2CO3, toluene, 0 8C; B: TiCl4·2 thf (5 mol %), iPr2EtN,
CH2Cl2, 23 8C.
Scheme 14. The contribution of Yoon and MacMillan to the development of an enantioselective aza-BC rearrangement.
ity. These systems are not readily accessible by other synthetic
methods. Thus, two consecutive [3,3] rearrangements of 80
with phthaloylglycyl chloride as a surrogate of N-phthaloylimidoketene 81 allowed the rapid construction of 83 with two
masked a-amino acid moieties, with > 95 % diastereoselectivity in favor of the 2,3-syn-3,6-anti isomer and in a superb
overall yield of 98 % (Scheme 13). The intermediate 82 (2,3-
benzyloxyacetyl chloride (85) as a precursor to benzyloxyketene led to 86 in 95 % yield, with syn/anti diastereoselectivity
of 98:2 and, most significantly, with 91 % ee. It seems that a
strongly chelating heteroatom, such as the alkyl-substituted
oxygen atom in the ketene moiety, promotes the enantiofacial
discrimination of the [3,3] rearrangement process, because
poorly chelating ketenes, such as acetoxyketene and tertbutyldimethylsilyloxyketene, exhibited only modest levels of
stereoselectivity (< 38 % ee). Significant variation in the
substituents on the double bond of the allylic amine was
tolerated without loss of enantioselectivity.
5. Conclusions and Outlook
Scheme 13. An example of a tandem aza-Belluš–Claisen rearrangement. Pht = phthaloyl.
syn), the product of the first rearrangement involving the
E allylic morpholine moiety, need not be isolated. Considerable variations in the substituents on the double bond in 80
and in the structure of the acyl chloride are tolerated.
Recently, Yoon and MacMillan[12] used the demonstrated
utility of auxiliary-metal chelation as an organizational
control element in asymmetric synthesis[48] to develop the
first examples of the enantioselective BC rearrangement
Angew. Chem. Int. Ed. 2004, 43, 3516 –3524
The Belluš–Claisen rearrangement offers abundant synthetic potential. Even with its current scope it considerably
broadens the range of possible stereoselective approaches to
E g,d-unsaturated esters, thioesters, and amides. At present
this experimentally simple and reliable reaction is one of the
most efficient and predictable methods for the synthesis of
nine- and ten-membered lactones, thiolactones, and lactams
containing an E double bond. The reactions of allylic ethers
and thioethers with electrophilic ketenes, such as dichloroketene, furnish products of the BC rearrangement in good
yields, whereas the reactions of allylic amines with neutral or
electron-rich ketenes can be efficiently catalyzed by Lewis
acids. When chiral Lewis acids were employed in equimolar
amounts, remarkable enantioselectivity was reported. However, even twenty-five years after its discovery, the Belluš
variant of the Claisen rearrangement is still in its infancy in
comparison with the established Johnson, Eschenmoser, and
Ireland variants. Further imaginative exploration of metalchelation effects might render allylic ethers and thioethers
amenable to reaction with a broader variety of ketenes. The
interaction of aromatic or heteroaromatic allylic systems with
ketenes has not yet been investigated, and neither has the
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Gonda
reaction of allenic and propargylic O, S, and N systems by
single or tandem protocols. It is to be expected that the
Belluš–Claisen rearrangement will become very useful in
organic synthesis in the near future.
Received: October 9, 2003 [M1718]
Published Online: May 24, 2004
[1] D. Belluš, personal communication.
[2] a) P. Martin, H. Greuter, D. Belluš, J. Am. Chem. Soc. 1979, 101,
5853; b) P. Martin, H. Greuter, D. Belluš, Helv. Chim. Acta 1981,
64, 64.
[3] W. S. Johnson, L. Werthemann, W. R. Bartlett, T. J. Brocksom,
T.-T. Li, D. J. Faulkner, M. R. Petersen, J. Am. Chem. Soc. 1970,
92, 741.
[4] R. Malherbe, D. Belluš, Helv. Chim. Acta 1978, 61, 3096 – 3099.
[5] T. T. Tidwell, Ketenes, Wiley, New York, 1995, pp. 336 – 3348.
[6] a) S. M. Roberts, C. Smith, R. J. Thomas, J. Chem. Soc. Perkin
Trans. 1 1990, 1493 – 1495; b) R. Maurya, C. A. Pittol, R. J. Pryce,
S. M. Roberts, R. J. Thomas, J. O. Williams, J. Chem. Soc. Perkin
Trans. 1 1992, 1617 – 1621.
[7] E. D. Edstrom, Tetrahedron Lett. 1991, 32, 5709 – 5712.
[8] C. J. Deur, M. W. Miller, L. S. Hegedus, J. Org. Chem. 1996, 61,
2871 – 2876.
[9] M. Ishida, H. Muramaru, S. Kato, Synthesis 1989, 562 – 564.
[10] T. P. Yoon, V. M. Dong, D. W. C. MacMillan, J. Am. Chem. Soc.
1999, 121, 9726 – 9727.
[11] M. Diederich, U. Nubbemeyer, Chem. Eur. J. 1996, 2, 894 – 900.
[12] T. P. Yoon, D. W. C. MacMillan, J. Am. Chem. Soc. 2001, 123,
2911 – 2912.
[13] M. M. Cid, E. Pombo-Villar, Helv. Chim. Acta 1993, 76, 1591 –
[14] R. Malherbe, G. Rist, D. Belluš, J. Org. Chem. 1983, 48, 860 – 869.
[15] M. R. Kling, G. A. McNaughton-Smith, R. J. K. Taylor, Chem.
Commun. 1993, 1593 – 1595.
[16] M. Diederich, U. Nubbemeyer, Angew. Chem. 1995, 107, 1095 –
1098; Angew. Chem. Int. Ed. Engl. 1995, 34, 1026 – 1028.
[17] V. M. Dong, D. W. C. MacMillan, J. Am. Chem. Soc. 2001, 123,
2448 – 2449.
[18] Malherbe–Belluš variant of the Claisen rearrangement,[8,15]
Belluš–Claisen rearrangement,[11] Belluš–Claisen reaction,[10]
Belluš ketene-Claisen reaction,[17] Belluš reaction.[12]
[19] For reviews, see: a) F. E. Ziegler, Chem. Rev. 1988, 88, 1423 –
1452; b) S. Blechert, Synthesis 1989, 71 – 82; c) P. Wipf in
Comprehensive Organic Synthesis, Vol. 5 (Eds.: B. M. Trost, I.
Fleming), Pergamon, Oxford, 1991, pp. 827 – 873; d) H. Frauenrath, Methods Org. Chem. (Houben-Weyl), Vol.E 21d 1995,
pp. 3300–3756; e) Y. Chai, S. Hong, H. A. Lindsay, C. McFarland, M. C. McIntosh, Tetrahedron 2002, 58, 2905 – 2926.
[20] S. Jolidon, H. J. Hansen, Helv. Chim. Acta 1977, 60, 978.
[21] For reviews, see: a) R. P. Lutz, Chem. Rev. 1984, 84, 205; b) B.
Ganem, Angew. Chem. 1996, 108, 1014 – 1023; Angew. Chem.
Int. Ed. Engl. 1996, 35, 936 – 945; c) M. Hiersemann, L. Abraham, Eur. J. Org. Chem. 2002, 1461 – 1471.
[22] U. Kazmeier, H. Mues, A. Krebs, Chem. Eur. J. 2002, 8, 1850.
[23] For remotely related charged aza-Claisen rearrangements, see
also: a) S. Mageswaran, W. D. Ollis, R. Somanathan, I. O.
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Sutherland, J. Chem. Soc. Perkin Trans. 1 1982, 893; b) E.
Vedejs, M. Gingras, J. Am. Chem. Soc. 1994, 116, 579; c) T. H.
Lambert, D. W. C. MacMillan, J. Am. Chem. Soc. 2002, 124,
13 646; d) A. V. Novikov, A. R. Kennedy, J. D. Rainier, J. Org.
Chem. 2003, 68, 993.
B. D. Johnston, E. Czyzewska, A. C. Oehlschlager, J. Org. Chem.
1987, 52, 3693 – 3697.
a) G. Rousseau, Tetrahedron 1995, 51, 2777 – 2849; b) P. von Zezschwitz, K. Voigt, M. Nottemeyer, A. de Meijere, Synthesis
2000, 1327 – 1340.
B. P. Moore, W. V. Brown, Aust. J. Chem. 1976, 29, 1365.
J. C. Dean, MSc thesis, University of East Anglia, Norwich,
1995; J. K. Taylor, personal communication.
E. Vedejs, J. M. Dolphin, H. Mastalerz, J. Am. Chem. Soc. 1983,
105, 127 – 130.
E. Vedejs, R. A. Buchanan, J. Org. Chem. 1984, 49, 1840 – 1841.
G. Rossini, G. G. Spineti, E. Foresti, G. Pradella, J. Org. Chem.
1981, 46, 2228 – 2230.
a) D. Enders, M. Knopp, R. Schiffers, Tetrahedron: Asymmetry
1996, 7, 1847 – 1882; b) H. Ho, T. Taguchi, Chem. Soc. Rev. 1999,
28, 43 – 50; c) B. Werschkun, J. Thiem, Top. Curr. Chem. 2001,
215, 293 – 325; d) U. Nubbemeyer, Synthesis 2003, 961 – 1008.
R. Oehrlein, R. Jeschke, B. Ernst, D. Belluš, Tetrahedron Lett.
1989, 30, 3517 – 3520.
B. Ernst, J. Gonda, R. Jeschke, U. Nubbemeyer, R. Oehrlein, D.
Belluš, Helv. Chim. Acta 1997, 80, 876 – 891.
a) U. Nubbemeyer, R. Oehrlein, J. Gonda, B. Ernst, D. Belluš,
Angew. Chem. 1991, 103, 1533 – 1534; Angew. Chem. Int. Ed.
Engl. 1991, 30, 1465 – 1467; see also: b) V. K. Aggarwal, A.
Lattanzi, D. Fuentes, Chem. Commun. 2002, 2534 – 2535.
J. Gonda, M. Martinkova, B. Ernst, D. Belluš, Tetrahedron 2001,
57, 5607 – 5613.
H.-J. Gais, A. BOhme, J. Org. Chem. 2002, 67, 1153.
R. W. Hoffmann, Chem. Rev. 1989, 89, 1841.
H. Priebke, R. BrPckner, Chem. Ber. 1990, 123, 153.
U. Nubbemeyer, J. Org. Chem. 1996, 61, 3677 – 3686.
S. Hatekayama, K. Saijo, S. Takano, Tetrahedron Lett. 1985, 26,
J. K. Cha, S. C. Lewis, Tetrahedron Lett. 1984, 25, 5263.
S. Takano, A. Kurotaki, M. Takahashi, K. Ogasawara, J. Chem.
Soc. Perkin Trans. 1 1987, 91.
M. M. Cid, U. Eggnauer, H. P. Weber, E. Pombo-Villar, Tetrahedron Lett. 1991, 32, 7233 – 7236.
E. D. Edstrom, J. Am. Chem. Soc. 1991, 113, 6690 – 6692.
a) A. Sudau, U. Nubbemeyer, Angew. Chem. 1998, 110, 1178 –
1181; Angew. Chem. Int. Ed. 1998, 37, 1140 – 1143; b) S. Laabs,
A. Scherrmann, A. Sudau, M. Diederich, C. Kierig, U. Nubbemeyer, Synlett 1999, 25 – 28; c) A. Sudau, W. MPnch, U.
Nubbemeyer, J. W. Bats, J. Org. Chem. 2000, 65, 1710 – 1720;
d) U. Nubbemeyer, Eur. J. Org. Chem. 2001, 1801 – 1816; e) A.
Sudau, W. MPnch, J. W. Bats, U. Nubbemeyer, Eur. J. Org.
Chem. 2002, 3304 – 3314.
U. Nubbemeyer, J. Org. Chem. 1995, 60, 3773 – 3780.
C.-M. Yu, H.-S. Choi, J. Lee, W.-H. Jung, H.-J. Kim, J. Chem. Soc.
Perkin Trans. 2 1996, 115 – 116.
J. S. Johnson, D. A. Evans, Acc. Chem. Res. 2000, 33, 325.
Angew. Chem. Int. Ed. 2004, 43, 3516 –3524
Без категории
Размер файла
263 Кб
rearrangements, belluцclaisen
Пожаловаться на содержимое документа