Dynamic equilibria in the products of intramolecular buchner additions of diazoketones to aryl rings bearing methoxy substituents

Article abstract of DOI:10.1021/jo015750l

Rhodium carboxylate catalyzed aromatic addition reactions of a range of diazoketones bearing methoxy-substituted aryl rings have been explored. While the existence of norcaradiene-cycloheptatriene equilibria in related compounds is well established, the aromatic addition products in this study display more complex dynamic equilibria due to conjugation with the methoxy group; the experimental evidence for this is discussed in detail. In the azulenone products 21-26 derived from p-methoxy-substituted diazoketones 14-16, the diastereomers interconvert via a spiro intermediate 39. A related mechanistic process in the azulenones 43-46 derived from the o-methoxy-substituted diazoketones 17, 18 interconverts regioisomers, explaining the conflicting reports for the regioselectivity of the cyclization of diazoketone 1. With the m-methoxy-substituted diazoketone 19, involvement of the methoxy group through a different pathway results in fragmentation of the azulenone to form the tetralone 47. With the azulenones 21-26 exclusive trapping of the norcaradiene associated with the less thermodynamically stable diastereomers in a cycloadduct with N-phenylmaleimide is observed. Due to the presence of the activating methoxy substituent on the aromatic ring, the aromatic addition reactions of the diazoketones studied were not very sensitive to the nature of the rhodium catalyst.

Full text of DOI:10.1021/jo015750l

7166
J . Org. Chem. 2001, 66, 7166-7177
Dyn a m ic Equ ilibr ia in th e P r od u cts of In tr a m olecu la r Bu ch n er
Ad d ition s of Dia zok eton es to Ar yl Rin gs Bea r in g Meth oxy
Su bstitu en ts
Anita R. Maguire,*,† Patrick O’Leary,† Francis Harrington,† Simon E. Lawrence,† and
Alexander J . Blake‡
Department of Chemistry, University College Cork, Cork, Ireland, and School of Chemistry,
The University of Nottingham, University Park, Nottingham NG7 2RD, U.K.
a.r.maguire@ucc.ie
Received May 14, 2001
Rhodium carboxylate catalyzed aromatic addition reactions of a range of diazoketones bearing
methoxy-substituted aryl rings have been explored. While the existence of norcaradiene-
cycloheptatriene equilibria in related compounds is well established, the aromatic addition products
in this study display more complex dynamic equilibria due to conjugation with the methoxy group;
the experimental evidence for this is discussed in detail. In the azulenone products 21-26 derived
from p-methoxy-substituted diazoketones 14-16, the diastereomers interconvert via a spiro
intermediate 39. A related mechanistic process in the azulenones 43-46 derived from the o-methoxy-
substituted diazoketones 17, 18 interconverts regioisomers, explaining the conflicting reports for
the regioselectivity of the cyclization of diazoketone 1. With the m-methoxy-substituted diazoketone
19, involvement of the methoxy group through a different pathway results in fragmentation of the
azulenone to form the tetralone 47. With the azulenones 21-26 exclusive trapping of the
norcaradiene associated with the less thermodynamically stable diastereomers in a cycloadduct
with N-phenylmaleimide is observed. Due to the presence of the activating methoxy substituent
on the aromatic ring, the aromatic addition reactions of the diazoketones studied were not very
sensitive to the nature of the rhodium catalyst.
In tr od u ction
between the resulting norcaradiene (NCD) and cyclohep-
tatriene (CHT) tautomers were of particular interest. The
presence of methoxy substituents has been shown to
strongly affect the relative stability of NCD and CHT
tautomers.3g
The intramolecular Buchner reaction of diazoketones
has attracted considerable attention in recent years.1 As
part of an ongoing investigation of stereocontrol in this
transformation,2 we extended our study to reactions with
substituted benzene derivatives to determine the effect
of substituents on the aromatic ring on both the efficiency
and stereocontrol in the cyclization. The activating meth-
oxy group was among the substituents selected for
investigation.3 The influence of the methoxy substituent
on the efficiency of the cyclization and regio- and stereo-
selectivity and its effect on the position of equilibrium
During the course of this study, rearrangements in the
resulting methoxy-substituted norcaradiene derivatives
were uncovered. In addition to the intrinsic mechanistic
interest in the dynamic equilibria observed, these results
explained the conflicting results that have been reported
for cyclization of 1-diazo-4-(2-methoxyphenyl)butan-2-one
1.4,5 The original description4 of the cyclization of diazo-
ketone 1 reported that the azulenone 2 formed was that
which corresponded to the cyclization toward the meth-
oxy substituent, which on treatment with TFA produced
the tetralone 3. However, in 1993, Cordi et al.5 reported
that the tetralone isolated was the regioisomeric com-
pound 5 and therefore assigned structure 4 to the
azulenone (Scheme 1). A recent report of related work
conducted independently by Manitto et al.6 is consistent
with our observations.
* To whom correspondence should be addressed. Tel: 353 21 490
2125. Fax: 353 21 427 4097.
† University College Cork.
‡ The University of Nottingham.
(1) For a comprehensive review of intramolecular addition of diazo-
ketones to aromatic systems, see: Modern Catalytic Methods for
Organic Synthesis with Diazo Compounds; Doyle, M. P., McKervey,
M. A., Ye, T., Eds.; Wiley: New York, 1998; pp 298-336.
(2) Maguire, A. R.; Buckley, N. R.; O’Leary, P.; Ferguson, G. Chem.
Commun. 1996, 2595. Maguire, A. R.; Buckley, N. R.; O’Leary, P.;
Ferguson, G. J . Chem. Soc., Perkin Trans. 1 1998, 4077.
(3) For examples of intramolecular diazoketone additions to methoxy
substituted aromatic derivatives, see: (a) Ledon, H.; Cannic, G.;
Linstrumelle, G.; J ulia, S. Tetrahedron Lett. 1970, 3971. (b) Ledon,
H.; Linstrumelle, G.; J ulia, S. Tetrahedron 1973, 29, 3609. (c) Pirrung,
M. C.; Morehead, A. T., J r. J . Am. Chem. Soc. 1996, 118, 8162. (d)
Manitto, P.; Monti, D.; Zanzola, S.; Speranza, G. J . Org. Chem. 1997,
62, 6658. (e) Mander, L. N. Synlett 1991, 134. (f) Mander, L. N.; Wells,
A. P. Tetrahedron Lett. 1997, 38, 5709. (g) Morris, J . C.; Mander, L.
N.; Hockless, D. C. R. Synthesis 1998, 455. (h) Zhan, H.; Appels, D.
C.; Hockless, D. C. R.; Mander, L. N. Tetrahedron Lett. 1998, 39, 6577.
(i) Matsumoto, M.; Shiono, T.; Mutoh, H.; Amano, M.; Arimitsu, S. J .
Chem. Soc., Chem. Commun. 1995, 101. (j) Doyle, M. P.; Protopopova,
M. N.; Peterson, C. S.; Vitale, J . P.; McKervey, M. A.; Fernandez
Garcia, C. J . Am. Chem. Soc. 1996, 118, 7865.
Resu lts a n d Discu ssion
A series of diazoketones 13-19 bearing methoxy sub-
stituents on the aryl ring was prepared under standard
(4) Kennedy, M.; McKervey, M. A.; Maguire, A. R.; Tuladhar, S. M.;
Twohig, M. F. J . Chem. Soc., Perkin Trans. 1 1990, 1047.
(5) Cordi, A. A.; Lacoste, J .-M.; Hennig, P. J . Chem. Soc., Perkin
Trans. 1 1993, 3.
(6) Manitto, P.; Monti, D.; Zanzola, S.; Speranza, G. Chem. Commun.
1999, 543.
10.1021/jo015750l CCC: $20.00 © 2001 American Chemical Society
Published on Web 09/21/2001

Intramolecular Buchner Additions of Diazoketones
J . Org. Chem., Vol. 66, No. 21, 2001 7167
Sch em e 1
Sch em e 2
conditions from the analogous carboxylic acids 6-12 as
outlined in Scheme 2 by initial formation of the acid
chlorides and then treatment with an excess of diazo-
ethane. The precursor carboxylic acids 6, 10, and 12 are
commercially available, while 7-9 and 11 were synthe-
sized as summarized in Scheme 2. The derivatives 13-
19 were selected for investigation to determine the
influence of the position of the methoxy substituent and
side-chain substituents on the rhodium acetate catalyzed
cyclization. Following the Cordi report, the terminal
diazoketone 1 was added to the study. Pure samples of
each of the diazoketones were obtained by chromato-
graphic purification and could be stored in a freezer for
several months without decomposition.
were very rapid, with the diazoketone decomposition
complete once the addition was finished. Cyclization of
the diazoketone 13 without any side-chain substituent
produced the azulenone 20 in good yield.
1
Analysis of the H NMR spectra of the crude product
mixtures from reactions of the methyl and isopropyl
substituted diazoketones 14 and 15 showed that both the
cis and trans diastereomers were formed in 3-5:1 ratio
with the trans isomers 21 and 23 predominating in each
case. Following chromatographic purification on silica gel,
the ratio of azulenones observed changed only slightly
with the trans isomers 21 and 23 still predominating.
As the azulenone products exist as rapidly equilibrating
norcaradiene-cycloheptatriene tautomers for which time-
1
averaged signals are observed, the H NMR character-
Rhodium acetate catalyzed decompositions of the
4-methoxy-substituted derivatives 13-16 were explored
initially (Scheme 3 and Table 1). All reactions were
conducted with <1 mol % of Rh2(OAc)4 in refluxing
dichloromethane unless otherwise stated, and in general
istics are very informative as to the nature of the
products, especially the position of NCD-CHT equilibria,7
(7) Hannemann, K. Angew. Chem., Int. Ed. Engl. 1987, 27, 284.

7168 J . Org. Chem., Vol. 66, No. 21, 2001
Maguire et al.
Sch em e 3
Ta ble 1. Rh od iu m Aceta te Ca ta lyzed Cycliza tion of
4-Meth oxyp h en yl-Su bstitu ted Dia zok eton es 13-16
Ta ble 2. Dia ster eoselection in Rh 2(OAc)4 Cycliza tion of
â-Su bstitu ted Dia zok eton esa
dr
azulenone dr (crude)a
(purified)a
T
yieldb
(%)
R
diazoketone (°C) trans cis
trans/cis
trans/cis
H
13
14
∆
∆
0
∆
0
20
80
82
51
77
63
72
Me
Pri
But
21
23
25
22
24
26
3:1
4.6:1
3:1
>98:2
98:2
4:1
4:1
9:1c
diazoketone
R
azulenone
yieldb (%)
trans/cisc
15
16
4:1
>98:2d
95:5d
27
28
29
30
31
32
Me
Et
33
34
35
36
37
38
87
80
74
74
70
72
89:11
96:4
97:3
> 98:2
98:2
∆
Pr
Diastereomeric ratios determined by 1H NMR spectroscopy of
the crude product and the product following chromatography on
a
Pri
Bu
But
b
silica gel. Yields reported for the azulenones following purifica-
> 98:2
tion by chromatography on silica gel. c This sample was stored in
the freezer prior to recording the 1H NMR spectrum; accordingly,
the ratio had shifted from the thermodynamic ratio of 4:1 to 9:1.
a
The data for cyclization of diazoketones 28-32 is described
b
in detail in ref 2. Yield of azulenones isolated following chroma-
tography on silica gel. c Diastereomeric ratios in the crude product
mixtures estimated by 1H NMR spectroscopy; chromatographic
separation of the diastereomers is possible.
d
The sample isolated directly from chromatography consisted of
25 only; however, on holding in CDCl3 for 2-3 days, equilibration
to the thermodynamic ratio 95:5 was observed.
tone 15 was conducted at 0 °C, the reaction was margin-
and furthermore, the signals for the cis and trans isomers
are well distinguished, allowing ready estimation of the
isomeric ratios. The signals (CDCl3, 20 °C) for the C(8)H
in the trans isomers 21 and 23 appear at 3.47 and 3.69
(J ) 7.3 Hz) while those of the cis isomers 22 and 24
appear at 4.51 and 4.68 (J ) 9.5 Hz), respectively.
Therefore, the position of equilibrium in the trans
isomers favors the NCD tautomer to a greater extent
than in the cis isomers.
As we have already reported,2 Rh2(OAc)4-catalyzed
cyclization of related diazoketones 28-32 onto phenyl
rings without any substituents is highly diastereoselec-
tive (see Table 2). Further to our earlier report, we have
investigated reaction of the methyl-substituted diazoke-
tone 27. Significantly, in the context of this work, the
diastereoselection observed when the â-substituent is a
methyl group is notably lower at 89:11 compared to that
observed with other alkyl groups, and indeed, a distinct
trend of increased diastereoselection as the â-substituent
is increased in size from Me to Et, Pr, Bu can be
distinguished. With branched alkyl side chains, Pri and
But, only a single diastereomer can be detected in the
crude product mixtures.
ally slower, the efficiency of cyclization reduced some-
1
what, but critically the H NMR spectrum of the crude
product ratio showed the presence of the trans isomer
23 only. Following chromatographic purification on silica
gel, however, a diastereomeric mixture of 23 and 24 in a
similar ratio to that observed before (4:1) was obtained.
Thus, it seems that the trans isomer 23 is the kinetic
product of the cyclization and that this can then equili-
brate with the cis isomer 24; the thermodynamic ratio
of 23/24 at room temperature appears to be 4:1. In
contrast, with the methyl-substituted diazoketone 14, the
crude product of the cyclization conducted at 0 °C
contained the cis isomer in a ratio comparable to that
found after chromatography. This suggests that either
cyclization of 14 is less stereoselective than that of 15 or
equilibration of the methyl-substituted product 21/22 is
more rapid than that of the isopropyl derivative 23/24.
More likely, both factors are involved. This difference is
presumably due to the decreased steric demands of the
methyl substituent compared to the isopropyl group. As
outlined above (Table 2), the diastereoselection in the
cyclization of â-methyl-substituted derivatives is lower
than those observed with the larger isopropyl substituent
at this position. However, we were never able to fraction-
ate the two isomers 21 and 22 on silica gel or in any other
manner, whereas 23 and 24, the isopropyl-substituted
derivatives, could be separated chromatographically and
then monitored as they returned to the equilibrium
mixture. Therefore, the rate of equilibration is apparently
faster for the methyl-substituted derivatives than for the
isopropyl-substituted derivatives, and in fact, we have
Our initial interpretation of the results of cyclization
of the 4-methoxyphenyl-substituted diazoketones 14 and
15 (Table 1) was that introduction of the activating
methoxy group to the aromatic system decreased the
diastereoselection compared to those observed with un-
substituted phenyl rings (Table 2) resulting in the
observed mixtures of diastereomers. However, when the
rhodium(II) acetate catalyzed cyclization of the diazoke-

Intramolecular Buchner Additions of Diazoketones
J . Org. Chem., Vol. 66, No. 21, 2001 7169
Sch em e 4
lowered the position of equilibrium shifts in favor of the
trans isomer in each case. At 218 K, the ratio of trans/
cis observed in each case was ∼6:1, compared to 4:1 at
298 K. We did not establish if this is the thermodynamic
ratio at this temperature; the rate of equilibration of the
isopropyl derivatives 23/24 in particular is rather slow,
and therefore, the position of equilibrium may not have
been attained while recording these NMR spectra.
Furthermore, samples of the isopropyl-substituted
azulenone 23/24, which were stored in a freezer, on
occasion showed ratios of trans/cis greater than the
equilibrium ratio of 4:1, e.g., 9:1 (see Table 1), again
indicating that the position of equilibrium is temperature
sensitive. With the isopropyl derivative, the rate of
equilibration is relatively slow, with the result that NMR
spectra of samples that are not at the position of
thermodynamic equilibrium can be recorded. With the
methyl-substituted derivative, the rate of interconversion
of 21 and 22 is much more rapid, so alteration of the
diastereomeric ratio following storage in the freezer was
not observed.
The sample of 23 that had been heated in ethyl acetate
(Table 3) showed a ratio of 23/24 of 2.5:1, indicating that
as the temperature is increased the position of equilib-
rium shifts even further toward 24; again, the rate of
equilibration is sufficiently slow that, even though the
NMR spectrum was recorded at room temperature in
CDCl3, the sample had not yet returned to the thermo-
dynamic ratio for this temperature of 4:1.
Ta ble 3. Equ ilibr a tion of th e tr a n s-Azu len on e 23 w ith
th e Cis Dia ster eom er 24
diastereomeric ratioa
conditions
cyclization at 0 °C
23:24
> 98:2
∼7:1
∼4:1
∼4:1
7 days at 20 °C
18 h on silica gel (dry)
18h in EtOAc/hexane mixture (1:9)
with silica gel
1
(d) Resolution of H NMR signals for the bridgehead
methyl groups of the two enantiomers of the azulenones
formed in this work through addition of Eu(hfc)3 as chiral
shift reagent is generally readily achieved, e.g., with
azulenones 33-38 and 23/24. However, in the case of the
azulenone 20 resolution of the signals for the two
enantiomers was not possible. This observation is ex-
plained by the dynamic behavior illustrated in Scheme
4; when R ) H the two enantiomers of 20 are intercon-
verted via the spiro enolate intermediate 39. Thus, the
azulenone 20 is nonresolvable. Furthermore, the rate of
this interconversion is likely to be enhanced by the
addition of Eu(hfc)3. In contrast, with the substituted
derivatives, e.g., 23/24, the dynamic equilibrium converts
only the diastereoisomers, not the enantiomers of each
component, and therefore, on addition of Eu(hfc)3, signals
for the bridgehead methyl groups of the two enantiomers
of 23 were clearly distinguished.
18h in refluxing EtOAc
2.5:1
a
Diastereomeric ratios determined by integration of 1H NMR
spectra.
never isolated a sample of 21 and 22 at other than the
thermodynamic ratio. Interestingly, the thermodynamic
ratio at 20 °C for both systems is the same at 4:1 trans/
cis.
The mechanism envisaged for the interconversion of
21 and 22, and also 23 and 24, is illustrated in Scheme
4. The electron-donating methoxy group triggers opening
of the cyclopropane ring to form a spiro enol or enolate
intermediate 39 that on reclosure can form either the cis
or trans isomer. On silica gel, acidic catalysis of this
process results in rapid formation of the thermodynamic
ratio of cis and trans isomers. To confirm this proposal,
the following experiments were conducted:
(a) Interconversion of a pure sample of the trans isomer
23 (formed by conducting the rhodium acetate catalyzed
cyclization at 0 °C) on storage at room temperature for 7
days, exposure to silica gel, either dry or in solution
(hexane/ethyl acetate 9:1, the chromatography eluant
used), and on heating (refluxing ethyl acetate for 18 h)
was observed as shown in Table 3. A thermodynamic
ratio of 4:1 for 23:24 at room temperature is again
indicated from each of these experiments. The experi-
ment involving heating in ethyl acetate is discussed
further in (c).
(e) Reaction of the azulenones 21/22 and 23/24 with
N-phenylmaleimide in refluxing ethyl acetate resulted
in formation of cycloadducts 40 and 41 as single diaster-
eomers (see Scheme 5); X-ray crystallographic analysis
(Figure S1, Supporting Information) of the cycloadduct
40 formed from the methyl-substituted azulenone mix-
ture 21/22 established that the relative stereochemistry
of the methyl substituents is cis; i.e., the cycloadduct is
formed selectively from the norcaradiene tautomer of the
minor diastereomer 22 present in the equilibrating
mixture. As the yield of the cycloadducts is greater than
that which would be possible if the minor diastereomer
only in the mixtures reacted, this provides further
evidence for the dynamic equilibration of the cis and
trans isomers. By analogy, the cycloadduct 41 from the
isopropyl substituted azulenone 23/24 is assigned the
same relative stereochemistry. Evidently, approach of the
dienophile, through transition state A (Figure 1), is easier
in the less hindered cis isomers 22 and 24 than in the
(b) A sample of the azulenone which contained pre-
dominantly the cis isomer 24 was isolated by chroma-
tography (ratio 23:24 1:11.5). Within 21 h at room
temperature in CDCl3, the ratio had altered to 1:6.2, and
after 190 h a ratio of 4:1, the thermodynamic ratio, was
observed.
1
(c) Variable-temperature H and 13C NMR studies of
21/22 and 23/24 showed that as the temperature is

7170 J . Org. Chem., Vol. 66, No. 21, 2001
Maguire et al.
On the basis of these preliminary observations, it was
not clear whether the dynamic behavior that had been
observed with 21/22 and 23/24 was also occurring in 25/
26. The altered ratio of 25/26 on purification by chroma-
tography is consistent with interconversion of 25 and 26
on silica gel, but in this case resulting in isolation of 25
exclusively or separation of two isomers 25 and 26, which
do not interconvert. The formation of 25 as the exclusive
product following chromatography is very different from
the behavior of 21/22 and 23/24 where the thermody-
namic ratio is ∼4:1 at 20 °C. Furthermore, as the
cycloadducts 40 and 41 were formed from the minor cis
isomers 22 and 24, the fact that 25/26 did not produce a
cycloadduct could indicate that in this case dynamic
interconversion between the isomers is not possible, and
with a maximum of 5% of the cis isomer 26 present (NMR
studies in CDCl3 at 20 °C) then the cycloadduct is not
formed to a detectable extent. Therefore, this system was
explored further, and firm evidence for dynamic inter-
conversion of 25 and 26 was obtained as outlined below.
(a) A sample of the azulenone 25 in which 26 could
not be detected was obtained following chromatographic
purification (i.e., dr 25/26 > 98:2). Monitoring of this
sample in CDCl3 over 5 days showed that the dr of 25/
26 gradually altered to 95:5, indicating that intercon-
version takes place slowly and that the position of
thermodynamic equilibrium in this solvent at 20 °C is
dr 25/26 19:1. Evidently, the sterically demanding tert-
butyl substituent alters the conformation of the product
and, as a result, the relative stability of the isomers, with
the trans isomer 25 more favored than in the derivatives
21/22 and 23/24 with the smaller substituents.
F igu r e 1. Stereoselectivity of cycloaddition.
Sch em e 5
(b) The isolation of the trans azulenone 25 as a single
diastereomer following chromatography (hexane/ethyl
acetate), together with the fact that no cycloaddduct could
be formed in refluxing ethyl acetate, led us to propose
that 25 and 26 undergo interconversion in the same
manner as 21/22 and 23/24 as illustrated in Scheme 4,
but that in ethyl acetate the position of equilibrium of
25/26 favors 25 exclusively with none of 26 present. To
test this proposal, a sample of the azulenone 25/26 with
1
a dr of 98:2 (estimated by H NMR in CDCl3 at 20 °C)
was divided in three portions. The first was stirred with
silica gel in CH2Cl2, the second was stirred with silica
gel in EtOAc and the third portion was stirred in EtOAc,
each for 2h. Each of the samples was filtered, concen-
trated and redissolved in CDCl3 for 1H NMR spectro-
scopy; the first sample had a dr of ∼95:5, while the two
samples that had been in contact with EtOAc consisted
of 25 only. These experiments confirmed that in ethyl
acetate the position of equilibrium is such that only 25
is present. They also explain why only 25 was isolated
following chromatography, as during evaporation of the
eluent the sample of 25 would be in solution in ethyl
acetate as the more volatile hexane was removed. In
dichloromethane or chloroform, the thermodynamic ratio
of 25/26 is 95:5 at 20 °C. Interestingly, the rate of
interconversion of 25 and 26 is sufficiently slow that the
samples that had been in contact with ethyl acetate
consisted of 25 only, even after they had been concen-
trated and redissolved in CDCl3.
trans isomers 21 and 23 in which the alkyl substituent
hinders approach of the dienophile as shown in transition
state B.
Interestingly, when the tert-butyl-substituted diazoke-
tone 16 was investigated, its behavior was quite different
from that of the methyl- or isopropyl-substituted deriva-
tives 14 and 15 (Table 1). The crude product of the
rhodium acetate catalyzed cyclization typically had a
diastereomeric ratio of 25/26 of 98:2, and following
purification on silica gel, the product was isolated with
a dr of >98:2; i.e., 26 could not be observed in the 1H
NMR spectrum. On holding for 2-3 days either neat or
in CDCl3, the ratio altered slightly to 25:26 95:5. When
the azulenone 25/26 was treated with N-phenylmaleim-
ide in refluxing ethyl acetate under identical conditions
used to form the cycloadducts 40 and 41 from 21/22 and
23/24 respectively (Scheme 5), there was no evidence of
formation of a cycloadduct.
(c) The failure to form a cycloadduct from 25/26 on
refluxing with N-phenylmaleimide in ethyl acetate could
now be rationalizedsthe cycloadducts 40 and 41 formed
from the cis isomers of the azulenones 22 and 24 only.
However, in ethyl acetate the cis isomer 26 is not present,

Intramolecular Buchner Additions of Diazoketones
J . Org. Chem., Vol. 66, No. 21, 2001 7171
Sch em e 6
Ta ble 4. Rh od iu m Ca r boxyla te Ca ta lyzed Cycliza tion of
2-Meth oxyp h en yl-Su bstitu ted Dia zok eton es
and therefore, cycloaddition is not observed. To confirm
this, the azulenone 25/26 (dr ∼98:2 estimated by 1H NMR
in CDCl3 at 20 °C) was heated in dichloromethane with
N-phenylmaleimide for 31 days with NMR monitoring;
33% of 25/26 was recovered together with 21% of the
cycloadduct 42 (Scheme 5). As the cycloaddition was
inefficient in refluxing dichloromethane, the reaction was
repeated in refluxing chloroform; following 6 days at the
higher reaction temperature, only the cycloadduct could
be detected in the 1H NMR spectrum of the crude product
with none of the azulenone 25/26 remaining. Chromato-
graphic purification resulted in isolation of the analyti-
cally pure cycloadduct 42 in 76% yield. X-ray crystallog-
raphy (Figure S2, Supporting Information) confirmed
that the relative stereochemistry of this cycloadduct was
identical to that of 40 confirming that the cycloaddition
occurred to the minor cis isomer 26 even though only 5%
of this is present at equilibrium.
crude purified
ratioa
ratioa
yieldb
(%)
SM R1 R2 T (°C) catalyst
A
B
A/B
A/B
17
H
Me
0
∆
∆
Rh2(pfb)4 43 44
Rh2(pfb)4 43 44 2.3:1
9:1
2:1
3:1
58
58
18 Pri Me
Rh2(OAc)4 45 46
Rh2(OAc)4
4:1
>96:4c
1
H
H
∆
0
4
2
1:2.3
2:1e
∼95d
a
Ratios determined by integration of 1H NMR spectra (MeO
signals) of both the crude product mixtures and the products
b
following purification by chromatography on silica gel. Yields
following chromatographic purification c The ratio of isomers had
changed after storage in the freezer at -20 °C for 1 year; no
d
noticeable decomposition of the azulenone was observed. This
sample was not purified as the 1H NMR spectrum of the crude
product was very clean; yield estimated by 1H NMR spectroscopy
of the crude product e This is the result reported by Manitto et
al.6 after these experiments. The rhodium acetate catalyzed
cyclization was in this case conducted at 0 °C, and therefore, the
ratio of the kinetic isomer 4 is higher than in our experiment
conducted at reflux in DCM where the amount of the thermody-
namic isomer 2 is increased.
The slow cycloaddition in refluxing dichloromethane
had two possible reasons, first that the rate of cycload-
dition to 26 was slow at this temperature and second that
the rate of interconversion of 25 to 26 was limiting. The
latter possibility was ruled out: following 3 days, a
ing 2- and 3-methoxyphenyl substituents, envisaging
similar dynamic behavior for the products from the
2-substituted derivatives 17 and 18 but not the 3-sub-
stituted compound 19. Cordi et al.5 had described cy-
clization of 1-diazo-4-(2-methoxyphenyl)butan-2-one 1
and reported regiochemistry opposite to that which had
been earlier described as shown in Scheme 1.4 On the
basis of our results, we immediately recognized that the
source of these conflicting reports was most likely due
to a dynamic equilibrium between norcaradienes as
illustrated in Scheme 6 and therefore expanded our study
to include the terminal diazoketone 1 also.
1
sample of the reaction mixture was withdrawn. The H
NMR spectrum indicated that 25 and 26 were present
in ∼94:6 ratio; i.e., thermodynamic equilibrium had been
attained.
Thus, dynamic interconversion of the azulenones 25/
26, mechanistically identical to the behavior of 21/22 and
23/24, was confirmed; however, the bulky tert-butyl group
influences both the position of equilibrium (25 exists
exclusively in ethyl acetate, and is more favored in
chloroform or dichloromethane, dr 25/26 19:1 cf. 4:1 for
21/22 and 23/24) and the rate of interconversion. As the
substituent decreases in size the rate of the interconver-
sion increases, with interconversion of 21/22 occurring
so quickly that they were never observed at anything
other than the thermodynamic ratio, while 23/24 and 25/
26 bearing the larger isopropyl and tert-butyl substitu-
ents interconvert more slowly, and NMR spectra at ratios
other than the thermodynamic equilibrium could be
recorded.
The results of these studies, summarized in Table 4,
support the proposed dynamic equilibria between regioi-
someric structures as illustrated in Scheme 6. Rhodium
acetate catalyzed cyclization of diazoketone 17 proved
rather inefficient. However, rhodium perfluorobutyrate
catalyzed decomposition of the 2-methoxy-substituted
diazoketone 17, without any side-chain substituent,
1
displayed in the H NMR spectra of the crude product a
Having established that the 6-methoxyazulenones 21/
22 and 23/24 could be interconverted via the spiro
intermediate 39, we commenced investigation of Rh2-
(OAc)4-catalyzed cyclization of diazoketones 17-19 bear-
mixture of the two possible regioisomers 43 and 44. When
the cyclization was conducted at 0 °C, 43 and 44 were
recovered in a 9:1 ratio, while at reflux the ratio was 2.3:

7172 J . Org. Chem., Vol. 66, No. 21, 2001
Maguire et al.
Sch em e 7
1, with the principal product in each case corresponding
to cyclization away from the methoxy group. Rhodium
acetate catalyzed cyclization of the analogous diazoketone
bearing an isopropyl substituent on the side chain 18 was
more efficient and again produced a mixture of regioiso-
mers 45 and 46, but this time in a ratio of 4:1, with the
same regioisomer A predominating. However, both com-
pounds displayed altered ratios of regioisomers on expo-
sure to silica gel during chromatography resulting in
thermodynamic ratios of 2:1 (43/44) and 3:1 (45/46).
Furthermore, when the sample of 45 and 46 (3:1) was
stored in the freezer at -20 °C for 1 year, the ratio altered
to >96:4 as shown by 1H NMR spectroscopy with no
significant level of decomposition of the azulenone evident
over this period. Therefore, with these two azulenones
the kinetic isomer is A, i.e., 43 and 45, corresponding to
cyclization away from the methoxy substituent, and this
is also the thermodynamically favored isomer at 20 °C,
with the equilibrium shifting even more toward A on
storage at -20 °C.
methoxy group, but in fact derived by rearrangement of
the kinetic product 4. Thus, the directive effect of the
o-methoxy group on the carbenoid cyclization, away from
the substituent, is the same as that reported with other
substituents such as methyl at the ortho position.
However, the tetralone formed when the azulenone 2,
or a mixture of 4 and 2, is treated with trifluoroacetic
acid is the 8-methoxy-2-tetralone 5 as described by Cordi
et al,5 derived from the kinetic product of the cyclization
4 as mechanistically the norcaradiene tautomer of 2 is
not readily rearomatized to form tetralone 3 (see Scheme
7). The formation of tetralone 5 from the azulenone 2
even when the rearrangement of 4 to 2 is essentially
complete by NMR spectroscopy, provides evidence for the
existence of a dynamic equilibrium between the two
regioisomers 4 and 2. The sample of the tetralone 5 which
we generated during the course of this research by
treatment of the azulenones 2 and 4 with TFA in
dichloromethane had identical spectroscopic character-
istics to those reported by Cordi et al.5
NMR spectroscopic analysis of the azulenone products
is particularly informative; two tautomers, NCD and
CHT, are possible for each isomer and furthermore the
position of this NCD-CHT equilibrium is very sensitive
to the nature and position of the substituents on the
compound;3g,7 for example, 45 exists almost entirely as
the NCD tautomer. The rate of tautomerisation is fast
on the NMR time scale while interconversion via the
spiro intermediates of the diastereomers for the 6-meth-
oxy azulenones 21-24 and the regioisomers for the 4/8-
methoxy azulenones 43-46, 2, 4 is slow. Therefore,
distinct time-averaged signals for the NCD/CHT tau-
tomers of each of the diastereomers/regioisomers are
observed in each case.
In light of the report by Cordi et al.,5 we decided to
look at the cyclization of diazoketone 1 as we believed
that equilibration of the regioisomeric azulenones 4 and
2 via the mechanism outlined in Scheme 6 explains the
conflicting reports outlined in Scheme 1.4,5 When the Rh2-
(OAc)4-catalyzed cyclization of 1 was conducted at reflux
and the 1H NMR spectra recorded as soon as possible
once the reaction was complete, the crude product
contained a mixture of regioisomers 4 and 2 in a 1:2.3
ratio. Treatment of an ether solution of this sample of 4
and 2 with TFA for 10 min resulted in isomerization to
a 3:97 ratio, indicating that conversion to the thermody-
namically more stable isomer 2 was catalyzed by the acid.
These results are in agreement with the result subse-
quently described by Manitto et al.6 with a ratio of 4/2 of
2:1 when the cyclization was conducted at 0 °C. As
described by Manitto et al.6 4 disappears and is gradually
converted to the thermodynamic isomer 2. Therefore,
when the cyclization was conducted at reflux the equili-
bration of the initially formed kinetic isomer 4 to the
thermodynamic isomer 2 proceeded more rapidly. With
this compound, the thermodynamically favored isomer
2 is regioisomer B, opposite to that observed when there
is a bridgehead methyl substituent where 43 and 45, i.e.
regioisomer A, were thermodynamically preferred. These
results confirm McKervey’s original assignment4 of the
product of cyclization of diazoketone 1 as azulenone 2,
the product formed by apparent cyclization toward the
The azulenones synthesized in this work were rela-
tively stable compounds that could be purified on silica
gel and stored in a freezer for extended periods without
any significant level of decomposition.
At this point, on the basis of our results, we had
demonstrated that the origin of the controversy4,5 in the
regioselectivity of cyclization of the methoxy-substituted
diazoketone 1 was due to the dynamic equilibrium via a
spiro intermediate as illustrated in Scheme 6. The
interconversion of the cis and trans isomers of the
6-substituted azulenones supports this mechanism. Fur-
ther evidence supporting this mechanistic interpretation
was secured when the 3-methoxyphenyl-substituted dia-
zoketone 19 was examined. On treatment with rhodium-

Intramolecular Buchner Additions of Diazoketones
J . Org. Chem., Vol. 66, No. 21, 2001 7173
Sch em e 8
electronic effect of the catalysts. Rhodium(II) caprolactam
was noticeably less efficient in the cyclization of diazo-
ketones 13-15 typically producing the azulenones in 50-
60% yield. Use of this catalyst with the diazoketones 1,
17, and 18 was not explored. Also, cyclization of diazo-
ketone 17 was notably more efficient with the more
electron-withdrawing ligands in the catalysts Rh2(pfb)4
and Rh2(mand)4 than with Rh2(OAc)4. However, other
than these two observations the efficiency of the diazo-
ketone cyclizations was not very sensitive to the nature
of the catalyst ligand. In particular the outcome of the
cyclization of the tert-butyl substituted diazoketone 16
was very insensitive to the catalyst employed, with good
yields (64-82%) observed in each case even with Rh2-
(cap)4 (79%). Interpretation of the influence of the catalyst
on the selectivity of the cyclizations, diastereoselectivity
with the 4-methoxy and regioselectivity with the 2-meth-
oxy derivatives, was difficult due to the dynamic equi-
libria discussed above which result in product ratios
which are more dependent on the history of the sample
in terms of age, reaction temperature, temperature of
storage etc. than on the kinetic properties of the catalyst.
In conclusion, rhodium(II)-catalyzed cyclizations of
diazoketones to aromatic rings bearing methoxy substit-
uents proceed efficiently; the activating effect of the
methoxy substituent overcomes to a large extent the
electronic effect of the ligands on the catalyst and results
in relatively minor catalyst sensitivity. Interesting dy-
namic equilibria are observed in the cyclization products
via rearrangements of the methoxy substituted norcara-
dienes. These observations explain the conflicting results
reported for the regioselectivity of cyclization of diazoke-
tone 1.4,5,6
Sch em e 9
(II) perfluorobutyrate in refluxing dichloromethane, only
the tetralone 47 was observed in agreement with
McKervey’s report4 that the tetralone 49 was formed
directly from the terminal diazoketone 48 (Scheme 8).
Therefore, as illustrated in Scheme 8, the 5-methoxy
group triggers an alternative fragmentation of the cy-
clopropane ring of the norcaradiene intermediate result-
ing in formation of the tetralone. Interestingly, there was
no evidence in the cyclization of 19 of a regioisomeric
product; McKervey4 also reported the formation of a
single tetralone from 48.
Following the completion of this work, a report of a
related study by Manitto et al.6 appeared which is in
agreement with our results. Involvement of a methoxy
substituent in cleavage of norcaradienes is precedented;
Mander et al.3g have proposed the mechanism illustrated
in Scheme 9 to explain the instability of norcaradiene
50 on exposure to silica gel. Moriarty et al.8 have also
reported related rearrangements in intermediates gener-
ated by reaction of iodonium ylides with methoxy sub-
stituted aromatic rings.
It is well established that variation of the ligand on
rhodium carboxylate and carboxamide catalysts can have
a significant effect on the reactivity of carbenoids formed
from the catalysts; chemo-, regio-, and stereoselectivity
can be altered by variation of the ligand.1,9 While the
aromatic addition reaction would be expected to be
catalyzed more efficiently by catalysts such as Rh2(tfa)4
and Rh2(pfb)4 which form highly electrophilic carbenoids
and less efficiently by Rh2(cap)4 compared to Rh2(OAc)4,
in practice it was observed that the activating effect of
the methoxy group overcame to a large extent the
Exp er im en ta l Section
All reactions were carried out under an inert atmosphere
of nitrogen. Infrared (IR) spectra were obtained as films on
NaCl plates or as KBr disks, and wavelengths (ν) are reported
in cm-1. 1H and 13C nuclear magnetic resonance (NMR) spectra
were recorded in CDCl3 (270 and 67.8 MHz, respectively)
unless indicated otherwise. Chemical shifts (δ) are given in
ppm downfield from tetramethylsilane. Flash chromatography
was conducted using the indicated solvent mixture and Merck
silica gel 60 (0.040-0.063 mm). Thin-layer chromatography
(TLC) was carried out on silica gel coated aluminum sheets
(Merck silica gel F254). Melting points are uncorrected. All
solvents were rountinely dried and distilled prior to use. Ether
refers to diethyl ether. All extractions were usually followed
by water and saturated NaCl aqueous solution washings,
MgSO4 drying, filtration, and evaporation. Diazoethane and
diazomethane solutions in ether were freshly prepared prior
to use and were distilled unless otherwise stated. All com-
mercially available reagents were used as received unless
otherwise stated. Thionyl chloride was distilled first from
quinoline and then from linseed oil.10 Rhodium acetate was
obtained from J ohnson Matthey. Other rhodium catalysts were
prepared according to standard procedures.11 In the cases of
the minor diastereomers of the azulenones spectral charac-
teristics reported are those which were easily distinguished
from those due to the major diastereomer. While acids 6, 10,
and 12 are commercially available, these were prepared for
this work by hydrogenation of the analogous cinnamic acid
derivatives.
3-(4′-Meth oxyp h en yl)bu ta n oic Acid (7). A suspension of
but-2-enoic acid (4.00 g, 4.65 × 10-2 mol) in ether (50 mL) was
(8) Moriarty, R. M.; May, E. J .; Prakash, O. Tetrahedron Lett. 1997,
38, 4333.
(9) Miah, S.; Slawin, A. M. Z.; Moody, C. J .; Sheehan, S. M.; Marino,
J . P., J r.; Semones, M. A.; Padwa, A.; Richards, I. C. Tetrahedron 1996,
52, 2489; Padwa, A.; Austin, D. J . Angew. Chem., Int. Ed. Engl. 1994,
33, 1797.
(10) Vogel, A. Textbook of Practical Organic Chemistry; Longman:
London, 1978.
(11) (a) Doyle, M. P.; Westrum, L. J .; Walthius, W. N. E.; Sie, M.
M.; Boone, W. P.; Bagheri, V.; Pearson, M. M. J . Am. Chem. Soc. 1993,
115, 958. (b) Drago, R. S.; Long, R. F.; Cosmano, R. Inorg. Chem. 1982,
21, 2196.

7174 J . Org. Chem., Vol. 66, No. 21, 2001
Maguire et al.
added over 10 min to a solution of 4-methoxyphenylmagnesium
bromide [freshly prepared from magnesium (3.35 g, 1.38 × 10-1
mol), 4-bromoanisole (26.10 g, 1.40 × 10-1mol), and iodine
(catalytic amount)] in ether (140 mL) at 0 °C. The reaction
mixture was stirred for 1 h while slowly returning to room
temperature and was then brought to reflux for 6 h. The
reaction mixture was then cooled to 0 °C and poured carefully
onto hydrochloric acid (120 mL, 10%) in ice (150 g). After
extraction with ether, the residue was purified by flash
chromatography with gradient ethyl acetate-hexane as eluant
to give the acid 7 (4.42 g, 49%) as a white solid: mp 67-69 °C
2-Dia zo-5-(4′-m eth oxyp h en yl)p en ta n -3-on e (13). 3-(4′-
Methoxyphenyl)propanoyl chloride [prepared from 3-(4′-meth-
oxyphenyl)propanoic acid 6 (0.46 g, 2.55 × 10-3 mol) and
freshly distilled thionyl chloride (15 mL) and purified by bulb
to bulb distillation (110 °C at 0.3 mmHg): IR (film) 1797, 1613
cm-1] (0.44 g, 2.22 × 10-3 mol), in ether (10 mL), was added
dropwise over 20 min to an ethereal diazoethane solution
[prepared from N-ethyl-N-nitrosourea (7.15 g, 6.11 × 10-2
mol)],15 at -20 °C while stirring under nitrogen. The solution
was allowed to slowly return to room temperature while
stirring for 3h. The ether and residual diazoethane were
evaporated under reduced pressure at room temperature,
using a rotary evaporator fitted with an acetic acid trap.
Purification by chromatography, using ethyl acetate-hexane
(5:95) as eluant, gave the diazoketone 13 (0.42 g, 75%) as a
(lit.12 mp 67-68 °C); IR (KBr) 3600-2400, 1712, 1614 cm-1
;
1H NMR δ 1.29 (d, J ) 7 Hz, 3H), 2.47-2.69 (m, 2H), 3.14-
3.20 (m, 1H), 3.78 (s, 3H), 6.79-6.88 (m, 2H), 7.09-7.20 (m,
2H); 13C NMR δ 22.00, 35.44, 42.83, 55.26, 114.03, 127.64,
137.65, 158.26, 178.11. Anal. Calcd for C11H14O3: C, 68.02; H,
7.27. Found: C, 68.20; H, 7.32.
3-(4′-Meth oxyp h en yl)-4-m eth ylp en ta n oic Acid (8). 4′-
Methoxycinnamic acid (3.00 g, 1.69 × 10-2 mol) was added
portionwise over 30 min to a solution of isopropylmagnesium
bromide [freshly prepared from magnesium (1.37 g, 5.62 × 10-2
mol), 2-propyl bromide (6.92 g, 5.62 × 10-2 mol), and iodine
(catalytic amount)] in ether (100 mL) at 0 °C. The reaction
mixture was stirred for 1 h while slowly returning to room
temperature and was then brought to reflux for 1 h. The
reaction mixture was then cooled to 0 °C and poured carefully
onto hydrochloric acid (50 mL, 10%) in ice (50 g). After
extraction with ether and the usual workup, the residue was
purified by flash chromatography with gradient ethyl acetate-
hexane as eluant to give the acid 8 (2.75 g, 73%) as a white
1
yellow oil: IR (film) 2071, 1634 cm-1; H NMR δ 1.92 (br s,
3H), 2.72 (br t, J ) 7 Hz, 2H), 2.90 (br t, J ) 8 Hz, 2H), 3.77
(s, 3H), 6.77-6.88 (m, 2H), 7.04-7.19 (m, 2H); 13C NMR δ 8.03,
29.70, 39.72, 55.25, 62.20, 113.96, 129.32, 132.84, 158.13,
193.63; HRMS (EI) calcd for C12H14N2O2 218.1055 (M+) found
218.1063; EIMS m/z (rel int) 218 (M+, 3), 190 (20), 175 (18),
134 (21), 121 (base). Anal. Calcd for C12H14N2O2: C, 66.13; H,
6.61; N, 13.04. Found: C, 66.13; H, 6.47; N, 12.83.
2-Diazo-5-(4′-m eth oxyph en yl)h exan -3-on e (14). This was
prepared following the procedure described for 13, using 3-(4′-
methoxyphenyl)butanoyl chloride [prepared from 3-(4′-meth-
oxyphenyl)butanoic acid 7 (3.00 g, 1.55 × 10-2 mol) and freshly
distilled thionyl chloride (20 mL)] and purified by bulb-to-bulb
distillation (100 °C, at 0.5 mmHg): IR (film) 1798, 1613 cm-1
]
(1.09 g, 8.94 × 10-3 mol), in ether (20 mL) and ethereal
diazoethane [prepared from N-ethyl-N-nitrosourea (10.00 g,
8.55 × 10-2 mol)].15 Purification by chromatography, using
ethyl acetate-hexane (3:97) as eluant, gave the diazoketone
14 (1.60 g, 45%) as a yellow oil: IR (film) 2065, 1634 cm-1; 1H
NMR δ 1.28 (d, J ) 7 Hz, 3H), 1.83 (br s, 3H), 2.52-2.79 (m,
2H), 3.22-3.39 (m, 1H), 3.78 (s, 3H), 6.76-6.92 (m, 2H), 7.06-
7.22 (m, 2H); 13C NMR δ 7.98, 21.61, 35.77, 46.38, 55.18, 62.53,
113.86, 127.61, 137.84, 158.10, 193.50; EIMS m/z (rel int) 233
(M+ + 1, 27), 204 (35), 189 (41), 135 (base). Anal. Calcd for
solid: mp 82-86 °C; IR (KBr) 3650-2400, 1705, 1612 cm-1
;
1H NMR δ 0.67 (d, J ) 7 Hz, 3H), 0.85 (d, J ) 7 Hz, 3H), 1.68-
1.85 (m, 1H), 2.44-2.58 (m, 1H), 2.65-2.84 (m, 2H), 3.71 (s,
3H), 6.70-6.81 (m, 2H), 6.96-7.06 (m, 2H); 13C NMR δ 20.08,
20.59, 33.19, 38.32, 47.71, 55.17, 113.43, 129.12, 134.64,
158.17, 178.76. Anal. Calcd for C13H18O3: C, 70.24; H, 8.16.
Found: C, 70.47; H, 8.29.
3-(4′-Meth oxyp h en yl)-4,4-d im eth ylp en ta n oic Acid (9).
This was prepared following the procedure described for 8 from
3-(4′-methoxyphenyl)propenoic acid (2.00 g, 1.12 × 10-2 mol)
and tert-butylmagnesium chloride in ether (2 M, 16.8 mL, 3.36
× 10-2 mol). The reaction was refluxed for 3 h. The crude acid
was isolated in 86% yield (2.27 g), which was pure enough to
use without further purification. Recrystallization of a small
portion of the crude product from ethanol gave an analytically
pure sample of the acid 9 as a white solid: mp 126-127 °C
C
13H16N2O2: C, 67.22; H, 6.94; N, 12.06. Found: C, 67.38; H,
7.20; N, 12.30.
2-Dia zo-5-(4′-m et h oxyp h en yl)-6-m et h ylh ep t a n -3-on e
(15). Oxalyl chloride (0.83 mL, 9.51 × 10-3 mol) was added
dropwise over 5 min to 3-(4′-methoxyphenyl)-4-methylpen-
tanoic acid 8 (1.93 g, 8.69 × 10-3 mol) in ether (15 mL) with
stirring at 0 °C. The solution was allowed to slowly return to
room temperature with stirring for 18 h. The solvent and
residual reagent were removed under reduced pressure to give
the acyl chloride, which was used without further purification.
An ethereal diazoethane solution15 was prepared from N-ethyl-
N-nitrosourea (10.16 g, 8.68 × 10-2 mol) and cooled to -20
°C. The crude acyl chloride in ether (20 mL) was added
dropwise over 20 min to the stirring diazoethane solution. The
solution was then allowed to return to room temperature with
stirring for 4 h. The ether and residual diazoethane were
evaporated under reduced pressure, using a rotary evaporator
fitted with an acetic acid trap. Purification by chromatography,
using ethyl acetate-hexane (2:98) as eluant gave the diazo-
ketone 15 (1.78 g, 79%) as a yellow oil: IR (film) 2072, 1632
(lit.13 mp 131-132 °C); IR (KBr) 3680-3030, 1711, 1610 cm-1
;
1H NMR δ 0.85 (s, 9H), 2.66 (dd, J ) 15, 11 Hz, 1H), 2.77 (dd,
J ) 16, 4 Hz, 1H), 2.87 (dd, J ) 11, 4 Hz, 1H), 3.78 (s, 3H),
6.78 (d, J ) 9 Hz, 2H), 7.05 (d, J ) 9 Hz, 2H); 13C NMR δ
27.83, 33.73, 35.54, 51.02, 55.11, 113.01, 130.09, 133.31,
158.05, 179.01; EIMS m/z (rel int) 236 (M+, 15), 221 (8), 179
(base), 137 (90). Anal. Calcd for C14H20O3: C, 71.16; H, 8.47.
Found: C, 71.41; H, 8.22.
3-(2′-Meth oxyph en yl)-4-m eth ylpen tan oic Acid (11). This
was prepared following the procedure described for 8 from
3-(2′-methoxyphenyl)propenoic acid (1.96 g, 1.10 × 10-2 mol)
and isopropylmagnesium bromide [freshly prepared from
magnesium (0.94 g, 3.84 × 10-2 mol), 2-propyl bromide (4.73
g, 3.84 × 10-2 mol) and iodine (catalytic amount)] in ether (100
mL). Purification by flash chromatography with gradient ethyl
acetate-hexane as eluant gave the acid 11 (1.39 g, 57%) as a
cm-1 1H NMR δ 0.73 (d, J ) 7 Hz, 3H), 0.96 (d, J ) 7 Hz,
;
3H), 1.69-1.92 (m, 4H), 2.68-2.82 (m, 2H), 2.83-2.97 (m, 1H),
3.78 (s, 3H), 6.74-6.87 (m, 2H), 7.00-7.15 (m, 2H); 13C NMR
δ 8.02, 20.58, 20.87, 33.01, 41.76, 48.29, 55.13, 62.77, 113.34,
1
white solid: IR (KBr) 3650-2400, 1711, 1602 cm-1; H NMR
128.98, 135.02, 158.03, 193.89; EIMS m/z (rel int) 288 (M+
+
δ 0.76 (d, J ) 7 Hz, 3H), 0.92 (d, J ) 7 Hz, 3H), 1.87-2.06 (m,
1H), 2.58-2.82 (m, 2H), 3.23-3.35 (m,1H), 3.78 (s, 3H), 6.79-
6.92 (m, 2H), 7.03-7.22 (m, 2H); 13C NMR δ 20.52, 20.78,
31.92, 36.98, 42.26, 55.21, 110.91, 120.32, 127.25, 128.72,
131.43, 157.58, 179.55. Anal. Calcd for C13H18O3: C, 70.24; H,
8.16. Found: C, 70.31; H, 8.08. Evidence was seen in the 1H
NMR for 9% of an isomeric compound, which is presumably
similar to those products observed by Mestres et al.14
N2, 12), 261 (M+ + 1, 10), 232 (M+ - N2, 33), 217 (20), 189
(base). Anal. Calcd for C15H20N2O2: C, 69.21; H, 7.74; N, 10.76.
Found: C, 69.19; H, 8.12; N, 9.45.
2-Dia zo-5-(4′-m et h oxyp h en yl)-6,6-d im et h ylh ep t a n -3-
on e (16). This was prepared following a similar procedure to
(14) (a) Aurell, M. J .; Mestres, R.; Munoz, E. Tetrahedron Lett. 1998,
39, 6351. (b) Aurell, M. J .; Banuls, M. J .; Mestres, R.; Munoz, E.
Tetrahedron 1999, 55, 831.
(15) Arndt, F. Organic Syntheses; Wiley: New York, 1943; Collect.
Vol. II, p 461.
(12) Fort, A. W.; Leary, R. E. J . Am. Chem. Soc. 1960, 82, 2494.
(13) Antoniadou-Vyza, A.; Foscolos, G. B.; Costakis, E. Chem. Chron.
1985, 14, 79; Chem. Abstr. 1986, 105, 60354q.

Intramolecular Buchner Additions of Diazoketones
J . Org. Chem., Vol. 66, No. 21, 2001 7175
that described for the preparation of 13, using 3-(4′-methoxy-
phenyl)-4,4-dimethylpentanoyl chloride [prepared from crude
3-(4′-methoxyphenyl)-4,4-dimethylpentanoic acid 9 (25.86 g,
1.09 × 10-1 mol) and thionyl chloride (80 mL)] and purified
by fractional distillation (122 °C, at 0.035 mmHg) to give the
acid chloride [14.15 g, 47% from 3-(4′-methoxyphenyl)prope-
noic acid: IR (film) 1803, 1611 cm-1] (7.10 g, 2.79 × 10-2 mol)
in ether (100 mL) and ethereal diazoethane [prepared from
N-ethyl-N-nitrosourea (26.11 g, 2.38 × 10-1 mol) and not
distilled but dried twice over KOH].15 Purification by chroma-
tography, using ethyl acetate-hexane (1:4) as eluant, gave the
diazoketone 16 (3.99 g, 52% from acid chloride) as a yellow
solid: mp 78-79 °C; IR (film) 2071, 1631 cm-1; 1H NMR (300
MHz) δ 0.88 (s, 9H), 1.76 (br s, 3H), 2.73 (br d, J ) 13 Hz,
1H), 2.85-3.07 (m, 2H), 3.78 (s, 3H), 6.79 (d, J ) 9 Hz, 2H),7.06
(d, J ) 9 Hz, 2H); 13C NMR (75.47 MHz) δ 8.06, 27.97, 33.87,
38.44, 51.53. 55.06, 113.06, 129.92, 133.70, 157.94, 194.07;
EIMS m/z (rel int) 274 (M+, 1), 246 (22), 189 (43). Anal. Calcd
for C16H22N2O2: C, 70.04; H, 8.02; N, 10.20. Found: C, 70.56;
H, 8.31; N, 10.03.
and ethereal diazoethane [prepared from N-ethyl-N-nitrosourea
(5.54 g, 5.04 × 10-2 mol)].15 Purification by chromatography,
using ethyl acetate-hexane (1:4) as eluant, gave the diazoke-
tone 27 (0.755 g, 62% from acid) as a yellow oil: IR (film) 2069,
1633 cm-1; 1H NMR (300 MHz) δ 1.31 (d, J ) 7 Hz, 3H), 1.88
(br s, 3H), 2.57-2.80 (m, 2H), 3.28-3.43 (m, 1H), 7.15-7.32
(m, 5H); 13C NMR δ 8.13, 21.37, 36.48, 46.20, 62.35, 126.34,
126.87, 128.12, 145.87, 193.21; EIMS m/z (rel int) 174 (M+
-
N2, 22), 159 (28), 132 (base), 105 (98). Anal. Calcd for
12H14N2O: C, 71.29; H, 6.93; N, 13.86. Found: C, 71.33; H,
C
7.18; N, 13.56.
1-Dia zo-4-(2′-m eth oxyp h en yl)bu ta n -2-on e (1).4 This was
prepared in a manner similar to that described for 13, using
3-(2′-methoxyphenyl)propanoyl chloride [prepared from 3-(2′-
methoxyphenyl)propanoic acid 10 (0.84 g, 4.67 × 10-3 mol) and
freshly distilled thionyl chloride (10 mL) and purified by bulb-
to-bulb distillation (100 °C, at 0.5 mmHg): IR (film) 1798 cm-1
]
(0.86 g, 4.33 × 10-3 mol) in ether (20 mL) and ethereal
diazomethane [prepared from Diazald (8.00 g, 3.73 × 10-2
mol)].10 Purification by chromatography, using ethyl acetate-
hexane (2:98) as eluant, gave the diazoketone 1 (0.74 g, 78%)
2-Dia zo-5-(2′-m eth oxyp h en yl)p en ta n -3-on e (17). This
was prepared following the procedure described for 15 using
3-(2′-methoxyphenyl)propanoic acid 10 (3.31 g, 1.84 × 10-2
mol) in ether (50 mL), oxalyl chloride (1.87 mL, 2.14 × 10-2
mol), and ethereal diazoethane [prepared from from N-ethyl-
N-nitrosourea (18.34 g, 1.57 × 10-1 mol)].15 Purification by
chromatography, using ethyl acetate-hexane (5:95) as eluant,
gave the diazoketone 17 (2.37 g, 59%) as a low melting yellow
1
as a yellow oil: IR (film) 2105, 1644 cm-1; H NMR δ 2.54-
2.72 (br m, 2H), 2.93 (t, J ) 8 Hz, 2H), 5.20 (br s, 1H), 6.77-
6.93 (m, 2H), 7.10-7.28 (m, 2H); 13C NMR δ 26.46, 40.98,
54.62, 55.49, 110.60, 120.82, 127.89, 129.18, 130.28, 157.73,
194.93.
3,8a -Dih yd r o-6-m et h oxy-8a -m et h yla zu len -1(2H )-on e
(20). 2-Diazo-5-(4′-methoxyphenyl)pentan-3-one 13 (100 mg,
4.59 × 10-4 mol) in CH2Cl2 (100 mL) was added dropwise over
1 h to a refluxing solution of rhodium(II) acetate (0.5 mg) in
CH2Cl2 (100 mL). The reaction was monitored by TLC and was
complete once the diazoketone had been added. Evaporation
of the solvent at reduced pressure gave the crude product as
1
solid: mp 38-39 °C dec; IR (KBr) 2070, 1632 cm-1; H NMR
δ 1.92 (br s, 3H), 2.74 (br t, J ) 8 Hz, 2H), 2.94 (br t, J ) 7
Hz, 2H), 3.80 (s, 3H), 6.79-6.93 (m, 2H), 7.08-7.29 (m, 2H);
13C NMR δ 8.04, 26.50, 37.91, 55.09, 62.15, 110.19, 120.48,
127.59, 128.84, 130.14, 157.44, 194.42. HRMS (EI) calculated
for C12H14N2O2 218.1055 (M+) found 218.1158; EIMS m/z (rel
int) 246 (M+ + N2, 4), 218 (M+, 8), 190 (1), 175 (47), 162 (47),
121 (40). Anal. Calcd for C12H14N2O2: C, 66.04; H, 6.47; N,
12.83. Found: C, 65.91; H, 6.62; N, 12.79.
1
a yellow oil. A H NMR spectrum was recorded to determine
the efficiency of the cyclization (88%) (percent of azulenone
by comparison to aromatic byproducts). Purification by flash
chromatography, with gradient ethyl acetate-hexane as elu-
ant, gave the azulenone 20 (70 mg, 80%) as a colorless oil: IR
2-Dia zo-5-(2′-m et h oxyp h en yl)-6-m et h ylh ep t a n -3-on e
(18). This was prepared following the procedure described for
15 using 3-(2′-methoxyphenyl)-4-methylpentanoic acid 11 (2.00
g, 9.01 × 10-3 mol) in ether (15 mL), oxalyl chloride (1.00 mL,
1.15 × 10-2 mol), and ethereal diazoethane [prepared from
from N-ethyl-N-nitrosourea (10.53 g, 9.00 × 10-2 mol)].15
Purification by chromatography, using ethyl acetate-hexane
(5:95) as eluant gave the diazoketone 18 (1.24 g, 53%) as a
(film) 1745, 1710, 1645, 1577 cm-1 1H NMR δ 0.81 (s, 3H),
;
2.16-2.60 (m, 4H), 3.61 (s, 3H), 4.00 (br d, J ) 7.8 Hz, 1H),
5.49 (dd, J ) 8, 2 Hz, 1H), 5.79 (dd, J ) 7, 2 Hz, 1H), 6.07 (d,
J ) 8 Hz, 1H); 13C NMR δ 11.28, 27.11, 34.28, 54.58, ∼85 (br),
108.68, 111.56, 123.87, 129.29, 156.58, 219.07; HRMS (EI)
calcd for C12H14O2 190.0994 (M+), found 190.0998; EIMS m/z
(rel int) 190 (M+, base), 175 (52), 162 (68), 147 (95), 133 (55).
Anal. Calcd for C12H14O2: C, 75.77; H, 7.40. Found: C, 75.35;
H, 7.52. No resolution of signals for enantiomers was seen in
1
yellow oil: IR (film) 2071, 1634 cm-1; H NMR 0.72 (d, J ) 7
Hz, 3H), 0.97 (d, J ) 7 Hz, 3H), 1.76 (br s, 3H), 1.91-2.13 (m,
1H), 2.69-2.85 (m, 1H), 2.86-3.05 (m, 1H), 3.18-3.36 (br m,
1H), 3.80 (s, 3H), 6.77-6.94 (m, 2H), 6.99-7.24 (m, 2H); 13C
NMR δ 8.08, 20.38, 20.76, 31.82, 40.62, 44.27, 55.20, 62.59,
110.63, 120.22, 127.13, 129.28, 131.24, 157.46, 194.29; EIMS
m/z (rel int) 288 (M+ + N2, <1), 260 (M+, <1), 232 (M+ - N2,
20), 217 (13), 189 (100). Anal. Calcd for C15H20N2O2: C, 69.21;
H, 7.74; N, 10.76. Found: C, 68.93; H, 7.39; N, 10.31.
1
a H NMR chiral shift study using (+)-Eu(hfc)3. It was difficult
to obtain 20 completely free of aromatic impurities.
tr a n s-(3R*,8a R*)-3,8a -Dih yd r o-6-m et h oxy-3,8a -d im e-
th yla zu len -1(2H)-on e (21) a n d cis-(3R*,8a S*)-3,8a -d ih y-
d r o-6-m eth oxy-3,8a -d im eth yla zu len -1(2H)-on e (22). This
was prepared following the procedure described for 20 from
2-diazo-5-(4′-methoxyphenyl)pentan-3-one 14 (100 mg, 4.31 ×
10-4 mol) and rhodium(II) acetate (0.5 mg). An 1H NMR
spectrum was recorded to determine the efficiency of the
cyclization (91%) and the diastereomeric ratio of the azule-
nones formed: diastereomeric ratio, trans-21/cis-22, 76:24 [by
2-Dia zo-5-(3′-m eth oxyp h en yl)p en ta n -3-on e (19). This
was prepared following the procedure described for 15 using
3-(3′-methoxyphenyl)propanoic acid 12 (3.03 g, 1.68 × 10-2
mol) in ether (15 mL), oxalyl chloride (2.57 mL, 2.95 × 10-2
mol), and ethereal diazoethane [prepared from N-ethyl-N-
nitrosourea (18.44 g, 1.68 × 10-1 mol)].15 Purification by
chromatography, using ethyl acetate-hexane (5:95) as eluant,
gave the diazoketone 19 (2.78 g, 76%) as a yellow oil: IR (film)
1
comparison of the H NMR integration for C(8)H cis and C(7)-
H trans]. Purification by flash chromatography, with gradient
ethyl acetate-hexane as eluant, gave a mixture of the dia-
stereomers of the azulenones 21, 22 (72 mg, 82%) as a colorless
oil. The diastereomeric ratio was estimated as trans-21/cis-
1
2070, 1634 cm-1; H NMR δ 1.93 (br s, 3H), 2.75 (br t, J ) 7
Hz, 2H), 2.94 (br t, J ) 8 Hz, 2H), 3.79 (s, 3H), 6.68-6.84 (m,
3H), 7.14-7.28 (m, 1H); 13C NMR δ 8.16, 30.86, 39.44, 55.19,
62.20, 111.70, 114.25, 120.76, 129.41, 142.51, 159.87, 193.43;
EIMS m/z (rel int) 220 (M+ + 2, 5), 204 (1), 180 (<1), 161 (3),
91 (30). Anal. Calcd for C12H14N2O2: C, 66.04; H, 6.47; N, 12.83.
Found: C, 65.80; H, 6.79; N, 11.29.
22, 80:20: IR (film) 1745, 1712, 1646 cm-1
.
Spectral details for trans-21: 1H NMR δ 0.75 (s, 3H), 1.08
(d, J ) 6 Hz, 3H), 1.84 (dd, J ) 18, 9 Hz, 1H), 2.46 (dd, J )
18, 9 Hz, 1H), 2.67-2.87 (m, 1H), 3.47 (d, J ) 7.3 Hz, 1H),
3.61 (s, 3H), 5.29 (dd, J ) 7, 2 Hz, 1H), 5.92 (dd, J ) 9, 2 Hz,
1H), 6.04 (d, J ) 9 Hz, 1H); 13C NMR δ 9.08, 18.84, 33.54,
42.07, 54.58, 66.95, 102.98, 116.42, 124.93, 156.43, 217.76.
Spectral details for cis-22: 1H NMR δ 0.86 (s, 3H), 1.28 (d, J
) 6 Hz, 3H), 2.08-2.18 (m, 1H), 2.45-2.59 (m, 1H), 3.62 (s,
3H), 4.51 (d, J ) 9.5 Hz, 1H), 5.66 (dd, J ) 10, 2 Hz, 1H), 5.78
(dd, J ) 8, 2 Hz, 1H), 6.21 (dd, J ) 8,1 Hz, 1H); 13C NMR δ
13.94, 21.63, 33.25, 43.68, 54.58, 108.79, 113.80, 120.40,
2-Dia zo-5-p h en ylh ep ta n -3-on e (27). This was prepared
following the procedure described for 13, using 3-phenylbu-
tanoyl chloride [prepared from 3-phenylbutanoic acid (1.00 g,
6.09 × 10-3 mol) and thionyl chloride (4.42 mL) and purified
by bulb-to-bulb distillation (75 °C, at 0.1 mmHg): IR (film)
1800 cm-1] (0.930 g, 5.10 × 10-3 mol 84%) in ether (20 mL)

7176 J . Org. Chem., Vol. 66, No. 21, 2001
Maguire et al.
127.77, 156.43; HRMS (EI) calcd for C13H16O2 204.1150 (M+)
found 204.1152; EIMS m/z (rel int) 204 (M+, 82), 189 (84), 176
(20), 162 (80), 134 (base).
for cis-33: 1H NMR δ 0.82 (s, 3H), 1.31 (d, J ) 7 Hz, 3H), 4.69
(d, J ) 7 Hz, 1H); EIMS m/z (rel int) 174 (M+, 30), 159 (22),
131 (75), 44 (base). Anal. Calcd for C12H14O: C, 82.76; H, 8.05.
Found: C, 82.40; H, 8.36.
tr a n s-(3R*,8a S*)-3,8a -Dih yd r o-3-isop r op yl-6-m eth oxy-
8a -m eth yla zu len -1(2H)-on e (23) a n d cis-(3R*,8a R*)-3,8a -
d ih yd r o-3-isop r op yl-6-m eth oxy-8a -m eth yla zu len -1(2H)-
on e (24). This was prepared following the procedure described
for 20, from 2-diazo-5-(4′-methoxyphenyl)-6-methylheptan-3-
one 15 (100 mg, 3.85 × 10-4 mol) and rhodium(II) acetate (0.5
mg). The efficiency of the cyclization (>90%) and the diaster-
eomeric ratio of the azulenones formed were estimated: dia-
stereomeric ratio, trans 23: cis 24, 76:24. Purification by flash
chromatography, with gradient ethyl acetate-hexane as elu-
ant, gave a mixture of the diastereomers of the azulenones 23,
24 (69 mg, 77%) as a colorless oil. The diastereomeric ratio
was estimated as trans-23/cis-24, 90:10: IR (film) 1746, 1711,
Diels-Ald er Ad d u ct (40). A solution of trans-(3R*,8aR*)-
3,8a-dihydro-6-methoxy-3,8a-dimethylazulen-1(2H)-one 21, cis-
(3R*,8aS*)-3,8a-dihydro-6-methoxy-3,8a-dimethylazulen-1(2H)-
one 22 (21/22 4:1) (121 mg, 5.93 × 10-4 mol), and N-phenylmal-
eimide (103 mg, 5.93 × 10-4 mol) in ethyl acetate (5 mL) was
refluxed for 16 h. The solvent was removed under reduced
pressure. Excess dienophile and aromatic impurities were
removed by trituration with ether. Recrystallization from
ethanol-dichloromethane gave the adduct 40 (101 mg, 45%)
as a colorless solid: mp 189-191 °C; IR (KBr) 1710, 1649 cm-1
;
1H NMR δ 1.18 (d, J ) 7 Hz, 3H), 1.23 (s, 3H), 1.35 (d, J ) 4
Hz, 1H), 1.77-1.93 (m, 1H), 2.33-2.49 (m, 2H), 3.22 (dd, J )
8, 4 Hz, 1H), 3.37 (dd, J ) 8, 4 Hz, 1H), 3.42-3.53 (4H, m),
3.77 (dd, J ) 8, 3 Hz, 1H), 4.68 (dd, J ) 7, 3 Hz, 1H), 7.09-
7.23 (m, 2H), 7.33-7.54 (m, 3H); 13C NMR δ 8.91, 20.11, 30.11,
31.41, 33.76, 38.62, 39.15, 42.14, 44.63, 45.22, 46.92, 55.03,
92.47, 126.26, 128.64, 129.16, 131.84, 156.92, 175.66, 177.00,
213.56; HRMS (EI) calcd for C23H23NO4 377.1627 (M+) found
377.1626; EIMS m/z (rel int) 377 (M+, base), 362 (7), 255 (18),
230 (15), 204 (20), 189 (30). Anal. Calcd for C23H23NO4: C,
73.19; H, 6.14; N 3.71. Found: C, 73.55; H, 5.88; N, 3.61.
Crystal for X-ray crystallography grown from ethanol-
dichloromethane.
1645 cm-1
.
Spectral details for trans-23: 1H NMR δ 0.76 (s, 3H), 0.80
(d, J ) 7, 3H), 0.90 (d, J ) 7 Hz, 3H), 1.58-1.73 (m, 1H), 1.99
(dd, J ) 17, 7 Hz, 1H), 2.30-2.63 (m, 2H), 3.61 (s, 3H), 3.69
(br d, J ) 7.3 Hz, 1H), 5.34 (br dd, J ) 8, 2 Hz, 1H), 5.85 (dd,
J ) 9, 2 Hz, 1H), 6.06 (d, J ) 9 Hz, 1H); 13C NMR δ 9.69,
19.13, 20.93, 32.37, 37.89, 45.41, 54.58, 70-74 (br), 104.05,
114.02, 126.18, 156.58, 217.73. Spectral details for cis-24: 1H
NMR δ 0.84 (s, 3H), 0.88 (d, J ) 7 Hz, 3H), 0.98 (d, J ) 7 Hz,
3H), 3.62 (s, 3H), 4.68 (br d, J ) 9.5 Hz, 1H), 5.72-5.79 (m,
2H), 6.15 (dd, J ) 8, 2 Hz, 1H); 13C NMR δ 13.67, 16.80, 21.13,
29.77, 37.38, 44.86, 55.03, 106.89, 113.38, 120.40, 129.07,
156.58, 219.50; HRMS (EI) calcd for C15H20O2 232.1463 (M+)
found 232.1462; EIMS m/z (rel int) 232 (M+, 13), 217 (17), 189
(16), 175 (17), 161 (76), 134 (base). Anal. Calcd for C15H20O2:
C, 77.55; H, 8.68. Found: C, 77.32; H, 8.83.
Diels-Ald er Ad d u ct (41). This was prepared following the
procedure outlined for the preparation of 40 from trans-(3R*,-
8aS*)-3,8a-dihydro-3-isopropyl-6-methoxy-8a-methylazulen-
1(2H)-one 23, cis-(3R*,8aR*)-3,8a-dihydro-3-isopropyl-6-meth-
oxy-8a-methylazulen-1(2H)-one 24 (23/24 80:20) (69 mg, 2.97
× 10-4 mol) and N-phenylmaleimide (52 mg, 2.99 × 10-4 mol)
in ethyl acetate (5 mL). Recrystallization from ethanol-
dichloromethane gave the adduct 41 (38 mg, 32%) as a
tr a n s-(3R*,8a S*)-3,8a -Dih yd r o-3-ter t-bu tyl-6-m eth oxy-
8a -m eth yla zu len -1(2H)-on e (25) a n d cis-(3R*,8a R*)-3,8a -
Dih yd r o-3-ter t-bu tyl-6-m eth oxy-8a -m eth yla zu len -1(2H)-
on e (26). This was prepared following the procedure described
for 20, from 2-diazo-5-(4′-methoxyphenyl)-6,6-dimethylheptan-
3-one 16 (102 mg, 3.71 × 10-4 mol) and rhodium(II) acetate
(0.5 mg). The efficiency of the cyclization (90%) and the
diastereomeric ratio of the azulenones formed were esti-
mated: diastereomeric ratio, trans-25/cis-26, 98:2. Purification
by flash chromatography, with gradient ethyl acetate-hexane
as eluant, gave a mixture of the diastereomers of the azule-
nones 25, 26 (66 mg, 72%) as a colorless oil. The diastereomeric
ratio was estimated as trans-25/cis-26, >98:2: IR (film) 1746,
colorless solid: mp 229-230 °C; IR (KBr) 1713, 1651 cm-1
;
1H NMR δ 0.68 (d, J ) 7 Hz, 3H), 0.94 (d, J ) 7 Hz, 3H), 1.18
(d, J ) 4 Hz, 1H), 1.23 (s, 3H), 2.04-2.22 (m, 2H), 2.25-2.42
(m, 2H), 3.21 (dd, J ) 8, 4 Hz, 1H), 3.36 (dd, J ) 8, 4 Hz, 1H),
3.42-3.57 (4H, m), 3.75 (dd, J ) 7, 3 Hz, 1H), 4.68 (dd, J ) 7,
4 Hz, 1H), 7.10-7.22 (m, 2H), 7.33-7.54 (m, 3H); 13C NMR δ
8.72, 14.71, 20.56, 28.21, 30.12, 33.45, 34.20, 38.53, 39.87,
40.70, 44.11, 44.51, 45.02, 54.87, 92.26, 126.20, 128.47, 129.02,
131.92, 157.07, 175.42, 176.88, 213.77; HRMS (EI) calcd for
C
25H27NO4 405.1940 (M+), found 405.1941; EIMS m/z (rel int)
1713, 1645 cm-1
.
405 (M+, 1), 362 (1), 327 (1), 173 (base). Anal. Calcd for C25H27
-
Spectral details for trans-25: 1H NMR (300 MHz) δ 0.74 (s,
3H), 0.86 (s, 9H), 2.18 (dd, J ) 18, 8 Hz, 1H), 2.43 (dd, J ) 18,
9 Hz, 1H), 2.61 (dd, J ) 9, 8 Hz, 1H), 3.62 (s, 3H), 3.75-3.87
(br d, 1H), 5.34 (d, J ) 8 Hz, 1H), 5.85 (dd, J ) 9, 2 Hz, 1H),
6.10 (d, J ) 9 Hz, 1H); 13C NMR (75.47 MHz) δ 9.22, 27.81,
33.83, 36.84, 49.54, 54.53, 102.59, 114.35, 122.15, 156.70,
217.75. Selected spectral details for cis-26: 1H δ 0.98 (s, 9H),
3.76 (s, 3H), 5.26 (d, J ) 11 Hz, 1H), 5.72-5.76 (m, 1H), 5.99-
6.04 (m, 1H), 6.31 (d, J ) 7 Hz, 1H); EIMS m/z (rel int) 246
(M+, 50), 231 (32), 189 (55), 105 (base). Anal. Calcd for
NO4: C, 74.06; H, 6.70; N, 3.60. Found: C, 73.97; H, 6.98; N,
3.60.
Diels-Ald er Ad d u ct (42). A solution of trans-(3R*,8aS*)-
3,8a-dihydro-3-tert-butyl-6-methoxy-8a-methylazulen-1(2H)-
one 25, cis-(3R*,8aR*)-3,8a-dihydro-3-tert-butyl-6-methoxy-8a-
methyl azulen-1(2H)-one 26, 25/26 >98:2 (86 mg, 3.49 × 10-4
mol), and N-phenylmaleimide (121 mg, 6.97 × 10-4 mol) in
chloroform (1 mL) was refluxed for 6 days. The solvent was
removed under reduced pressure. Purification by flash chro-
matography, with gradient ethyl acetate-hexane as eluant
gave the adduct 42 (111 mg, 76%) as a colorless solid: mp
98-100 °C dec; IR (KBr) 1775, 1714, 1655 cm-1; 1H NMR (300
MHz) δ 1.07 (s, 9H), 1.21-1.26 (m, 4H), 2.18-2.26 (m, 3H),
3.24 (dd, J ) 8, 3 Hz, 1H), 3.43-3.51 (m, 5H), 4.18 (dd, J ) 7,
3 Hz, 1H), 4.65 (dd, J ) 8, 3 Hz, 1H), 7.15 (d, J ) 7 Hz, 2H),
7.33-7.49 (m, 3H); 13C NMR (75.47 MHz) δ 8.91, 28.81, 32.07,
34.28, 35.62, 37.41, 38.02, 40.09, 44.30, 45.05, 46.37, 55.01,
92.53, 126.17, 128.60, 129.13, 131.74, 156.86, 175.77, 176.94,
214.20; EIMS m/z (rel int) 419 (M+, 25), 362 (20), 256 (18),
215 (20), 57 (base). Anal. Calcd for C26H29NO4: C, 74.44; H,
6.97; N, 3.34. Found: C, 74.33; H, 6.70; N, 3.90.
C
16H22O2: C, 78.01; H, 9.00. Found: C, 77.72; H, 9.08.
tr a n s-(3R*,8a R*)-3,8a -Dih yd r o-3,8a -d im et h yla zu len -
1(2H)-on e a n d cis-(3R*,8a S*)-3,8a -d ih yd r o-3,8a -d im eth yl-
a zu len -1(2H)-on e (33). This was prepared following the
procedure described for 20 from 2-diazo-5-phenylhexan-3-one
27 (100 mg, 4.95 × 10-4 mol) and rhodium(II) acetate (0.5 mg).
A 1H NMR spectrum was recorded to determine the efficiency
of the cyclization (∼86%) and the diastereomeric ratio of the
azulenones 33 formed: diastereomeric ratio, trans/cis 8:1.
Purification by flash chromatography, with gradient ethyl
acetate-hexane as eluant, gave a mixture of the diastereomers
of the azulenones 33 (75 mg, 87%) as a colorless oil. The
diastereomeric ratio was estimated as trans/cis, 7:1: IR (film)
Crystal for X-ray crystallography grown from ethanol-
dichloromethane.
1748, 1715 cm-1
.
3,8a-Dih ydr o-4-m eth oxy-8a-m eth ylazu len -1(2H)-on e (43)
a n d 3,8a -Dih yd r o-8-m eth oxy-8a -m eth yla zu len -1(2H)-on e
(44). (a ) This was prepared following the procedure described
for 20, from 2-diazo-5-(2′-methoxyphenyl)pentan-3-one 17 (100
mg, 4.59 × 10-4 mol) and rhodium(II) perfluorobutyrate (0.5
mg). The efficiency of the cyclization (80%) and the ratio of
Spectral details for trans-33: 1H NMR δ 0.72 (s, 3H), 1.10
(d, J ) 7 Hz, 3H), 1.90 (dd, J ) 18, 9 Hz, 1H), 2.52 (dd, J )
18, 9 Hz, 1H), 2.78-2.97 (m, 1H), 3.75 (d, J ) 7 Hz, 1H), 6.05-
6.19 (m, 2H), 6.25-6.42 (m, 2H); 13C NMR δ10.62, 19.53, 34.03,
42.74, 122.94, 125.20, 126.68, 127.80, 218.21. Spectral details

Intramolecular Buchner Additions of Diazoketones
J . Org. Chem., Vol. 66, No. 21, 2001 7177
the azulenones formed were estimated: 43/44 70:30 (by
comparison of the H NMR integration for OCH3 43 and 44).
(br s, 1H), 3.66 (s, 3H), 5.50 (d, J ) 6, 1H), 5.90 (dd, J ) 8, 6,
1H); 13C NMR δ 27.87, 33.76, 55.01, 104.34, 110.12, 120.05,
126.38, 216.14.
1
Purification by flash chromatography, with gradient ethyl
acetate-hexane as eluant, gave the azulenones 43, 44 (51 mg,
58%) as a colorless oil. The ratio of the two azulenones was
An ether (15 mL) solution of the crude reaction mixture (15
mg) was treated with 1 drop of trifluoroacetic acid for 10 min.
The solution was then washed with sodium hydrogen carbon-
ate (10%, 10 mL) and worked up in the normal manner. A 1H
NMR spectrum was recorded which indicated that the product
of the reaction was the azulenones 2, 4 in the ratio 2/4 97:3.
estimated as 43/44 63:37: IR (film) 1753 (w), 1714 cm-1
.
(b) This reaction was conducted as above except the reaction
temperature was 0 °C. The efficiency of the cyclization (83%)
and the ratio of the azulenones formed were estimated: 43:44
90:10 and the crude sample was subjected to further spectro-
scopic examination.
The spectral details of the two isomers were assigned as
follows. The assignment for 43 was obtained from the NMR
spectra of the crude product, of the reaction carried out at 0
°C and stored in the freezer subsequently, in which this
product was dominant (90%). The assignment for 44 was
obtained from the NMR spectra of the purified products
(reaction at reflux), ratio 43/44 63:37 by subtraction of the
previously identified signals for 43.
3,4-Dih yd r o-5-m eth oxyn a p h th a len e-2(1H)-on e (5).4-6
A
dichloromethane (15 mL) solution of the crude reaction
mixture of 2 and 4 (15 mg) was treated with 1 drop of
trifluoroacetic acid for 10 min, washed with aqueous sodium
hydrogen carbonate, and worked up in the normal manner to
give 5 (15 mg, 100%) as a clear oil: 1H NMR δ 2.51 (t, J ) 7,
2H), 3.08 (t, J ) 7, 2H), 3.56 (s, 1H), 3.85 (s, 3H), 6.75 (dd, J
) 14, 7, 2H), 7.18 (dd, J ) 7, 7, 1H).
3,4-Dih yd r o-1-m et h yl-6-m et h oxyn a p h t h a len e-2(1H )-
on e (47). 2-Diazo-5-(3′-methoxyphenyl)pentan-3-one 19 (100
mg, 4.59 × 10-4 mol) in CH2Cl2 (100 mL) was added dropwise
over 1 h to a refluxing solution of rhodium(II) perfluorobu-
tyrate (0.5 mg) in CH2Cl2 (150 mL). Evaporation of the solvent
at reduced pressure gave the product as a yellow oil. A 1H
NMR spectrum of the crude product was recorded to determine
the efficiency of the reaction and indicated the absence of the
azulenone. Purification by flash chromatography, with gradient
ethyl acetate-hexane as eluant, gave the tetralone 47 (57 mg,
Spectral details for 43: 1H NMR δ 0.71 (s, 3H), 1.83-2.33
(m, 3H), 2.57 (d, J ) 5.1 Hz, 1H), 2.62-2.79 (m, 1H), 3.66 (s,
3H), 5.27 (d, J ) 6.8 Hz, 1H), 5.58 (dd, J ) 9.2, 5.1, 1H), 6.17
(dd, J ) 8.6, 6.5 Hz, 1H); 13C NMR δ 5.05, 22.14, 24.84, 31.51,
40.38, 45.76, 55.25, 94.27, 114.88, 126.21, 156.17, 217.27.
Selected spectral details for 44: 1H NMR δ 0.76 (s, 3H), 2.45-
2.56 (m, 4H), 3.56 (s, 3H), 5.48 (d, J ) 5 Hz, 1H), 6.25-6.35
(m, 3H); 13C NMR δ 12.92, 27.04, 37.89, 56.61, 96.93, 120.97,
122.47, 124.73, 157.49; HRMS (EI) calcd for C12H14O2 190.0994
(M+) found 190.0995; EIMS m/z (rel int) 190 (M+, 34), 148 (52),
121 (62), 91 (base). Anal. Calcd for C12H14O2: C, 75.77; H, 7.40.
Found C, 75.41; H, 7.62.
1
65%) as a colorless oil: IR (film) 1714, 1609 cm-1; H NMR δ
1.45 (d, J ) 7, 3H), 2.42-2.70 (m, 2H), 2.93-3.16 (m, 2H),
3.47 (q, J ) 7, 1H), 3.81 (s, 3H), 6.72-6.85 (m, 2H), 7.06-7.17
(m, 1H); 13C NMR δ 14.65, 28.38, 37.58, 46.74, 55.36, 112.38,
113.33, 128.27, 130.20, 138.01, 158.53, 212.27.
tr a n s-(3R*,8a S*)-3,8a -Dih yd r o-3-isop r op yl-4-m eth oxy-
8a -m eth yla zu len -1(2H)-on e (45) a n d cis-(3R*,8a R*)-3,8a -
Dih yd r o-isop r op yl-8-m et h oxy-8a -m et h yla zu len -1(2H )-
on e (46). This was prepared following the procedure described
for 20, from 2-diazo-5-(2′-methoxyphenyl)-6-methyl-heptan-3-
one 18 (100 mg, 3.85 × 10-4 mol) and rhodium(II) acetate (0.5
mg). The efficiency of the cyclization (90%) and the ratio of
the azulenones formed were estimated: 45/46 80:20. Purifica-
tion by flash chromatography, with gradient ethyl acetate-
hexane as eluant, gave the azulenones 45, 46 (52 mg, 58%) as
a colorless oil. The ratio of the two azulenones was estimated
as 45/46, 74:26: IR (film) 1714 cm-1. The sample was stored
Crystal data for 40: C23H23NO4, 377.42, colorless hexagonal
plate 0.42 × 0.27 × 0.16 mm, monoclinic, P21/c, a ) 10.0820-
(16) Å, b ) 18.328(3) Å, c ) 10.455(3) Å, â ) 95.164(16)°, U )
1924.1(6) Å3, Z ) 4, T ) 293 K, θmax ) 24.92°, Nonius MACH-3
diffractometer, Mo KR radiation (λ ) 0.710 73 Å), R (on F, 1563
reflections with I > 2(σ)I) ) 0.068, wR2 (on F2, all 3368 unique
reflections) ) 0.174, gof (F2) ) 1.13, 256 parameters.
Crystal data for 42: C26H29NO4, 419.50, colorless column
0.30 × 0.12 × 0.10 mm, monoclinic, P21/c, a ) 8.8062(7) Å, b
) 22.772(2) Å, c ) 10.9663(9) Å, â ) 95.080(2)°, U ) 2190.5-
(3) Å3, Z ) 4, T ) 150 K, θmax ) 28.65°, Bruker SMART1000
diffractometer, Mo KR radiation (λ ) 0.710 73 Å), R (on F, 3093
reflections with I > 2(σ)I) ) 0.041, wR2 (on F2, all 5139 unique
reflections) ) 0.093, gof (F2) ) 0.874, 285 parameters.
The X-ray data for compounds 40 and 42 have been
deposited with the Cambridge Crystallographic Data Centre.
The data can be obtained, upon request, from the Director,
Cambridge Crystallographic Data Centre. The deposition
numbers are CCDC 162992 and 162993.
1
in the freezer (∼-20 °C) for 1 year, and the H and 13C NMR
were again recorded. The ratio of the two azulenones was
estimated as 45:46, >96:4. The spectral characteristics for 45
have been obtained from these spectra and those of 46 from
the original ca. 3:1 mixture. Spectral details for 45: 1H NMR
δ 0.69 (s, 3H), 0.85 (d, J ) 7, 3H), 0.86 (d, J ) 7, 3H), 1.44-
1.61 (m, 1H), 1.81 (dd, J ) 18, 9 Hz, 1H), 2.33 (dd, J ) 18, 10
Hz, 1H), 2.61 (d, J ) 5.1 Hz, 1H), 2.70-2.88 (m, 2H), 3.65 (s,
3H), 5.22 (d, J ) 6.8 Hz, 1H), 5.57 (dd, J ) 9.3, 5.1 Hz, 1H),
6.15 (dd, J ) 9.1, 6.7 Hz, 1H); 13C NMR δ 4.92, 20.64, 21.09,
25.51, 32.33, 37.23, 37.60, 40.71, 47.52, 57.80, 93.85, 114.46,
126.29, 157.07, 216.96. Spectral details for 46: 1H NMR δ 0.76
(s, 3H), 0.90 (d, J ) 7, 3H), 0.98 (d, J ) 7, 3H), 3.56 (s, 3H),
5.43-5.52 (m, 1H), 6.25-6.35 (m, 3H); 13C NMR δ 13.00, 18.20,
33.31, 41.41, 47.20, 56.70, 97.10, 122.51, 122.53, 124.60; HRMS
(EI) calcd for C15H20O2 232.1463 (M+), found 232.1465; EIMS
m/z (rel int) 232 (M+, 73), 215 (27), 189 (18), 162 (69), 119 (74),
84 (base). Anal. Calcd for C15H20O2: C, 77.55; H, 8.68. Found:
C, 77.62; H, 8.65.
Ack n ow led gm en t. The financial support of the
following is gratefully acknowledged: Pfizer Pharma-
ceuticals (studentship to P.O.L.), National University
of Ireland (Postdoctoral Fellowship to P.O.L.), Enter-
prise Ireland, University College Cork, especially UCC
President’s Research Fund, J ohnson Matthey for a loan
of rhodium(II) acetate and Cork University Foundation
for the purchase of an X-ray diffractometer. A.J .B.
thanks the EPSRC(UK) for funding the purchase of a
CCD area detector diffractometer. Support from Enter-
prise Ireland International Collaboration Program to
facilitate collaboration between Cork and Nottingham
is gratefully acknowledged.
8-Meth oxy-3,8a-dih ydr oazu len -1(2H)-on e (2)an d 4-Meth -
oxy-3,8a -d ih yd r oa zu len -1(2H)-on e (4).4-6 This was pre-
pared following the procedure described for 20, from 1-diazo-
4-(2′-methoxyphenyl)-butan-2-one 1 (100 mg, 4.90 × 10-4 mol)
and rhodium(II) acetate (0.5 mg). The ratio of the azulenones
formed was estimated: 2/4 2.3:1. The crude product, a clear
oil, was sufficiently pure to use without chromatography.
Spectral details for the major regioisomer 2: 1H NMR δ 2.50-
2.71 (m, 3H), 2.81-3.01 (br t, J ) 9, 2H), 3.59 (s, 3H), 5.41 (d,
J ) 6, 1H), 6.09-6.22 (m, 1H), 6.26-6.42 (m, 2H); 13C NMR δ
27.00, 38.92, 53.73, 56.37, 96.97, 120.95, 124.31, 126.52,
132.67, 145.76, 213.82. Spectral details for the minor regio-
Su p p or tin g In for m a tion Ava ila ble: Detailed spectral
assignments for each of the compounds. ORTEP diagrams
(Figures S1-2) for 40 and 42, and NMR spectra for 15, 16,
18-20, 21/22, 42, and 47. This material is available free of
charge via the Internet at http://pubs.acs.org.
1
isomer 4: H NMR δ 1.33 (br s, 1H), 2.04-2.37 (m, 2H), 3.40
J O015750L