Combined computational and experimental studies of the mechanism and scope of the retro-Nazarov reaction

Article abstract of DOI:10.1021/ja048942h

Density functional theory calculations (B3LYP/6-31+G*) demonstrate that conjugating and electron-donating substituents at carbons three and four of a cyclopentenyl oxyallylic cation should have a rate-accelerating effect on the retro-Nazarov reactions of these species. The retro-Nazarov reaction of these intermediates is predicted to exhibit significant torquoselectivity when carbon three is substituted with a methoxy and a methyl group. Experimental studies show that oxyallylic cations can undergo effective retro-Nazarov reactions when two alkyl and one aryl/vinyl groups are on carbons three and four. An equal number of alkyl substituents or a single aryl substituent is not effective in promoting the reaction. Interestingly, a single alkoxy substituent at carbon three is sufficient for the retro-Nazarov reaction to occur. The methodology developed was used in a total synthesis of the natural product turmerone.

Full text of DOI:10.1021/ja048942h

Published on Web 08/14/2004
Combined Computational and Experimental Studies of the
Mechanism and Scope of the Retro-Nazarov Reaction
Michael Harmata,*,† Peter R. Schreiner,*,‡ Dong Reyoul Lee,† and
Patrick L. Kirchhoefer†
Contribution from the Department of Chemistry, UniVersity of Missouri-Columbia,
Columbia, Missouri 65211, and Institut fu¨r Organische Chemie, Justus-Liebig-UniVersity,
Heinrich-Buff-Ring 58, D-35392 Giessen, Germany
Received February 25, 2004; E-mail: harmatam@missouri.edu; prs@org.chemie.uni-giessen.de
Abstract: Density functional theory calculations (B3LYP/6-31+G*) demonstrate that conjugating and
electron-donating substituents at carbons three and four of a cyclopentenyl oxyallylic cation should have
a rate-accelerating effect on the retro-Nazarov reactions of these species. The retro-Nazarov reaction of
these intermediates is predicted to exhibit significant torquoselectivity when carbon three is substituted
with a methoxy and a methyl group. Experimental studies show that oxyallylic cations can undergo effective
retro-Nazarov reactions when two alkyl and one aryl/vinyl groups are on carbons three and four. An equal
number of alkyl substituents or a single aryl substituent is not effective in promoting the reaction. Interestingly,
a single alkoxy substituent at carbon three is sufficient for the retro-Nazarov reaction to occur. The
methodology developed was used in a total synthesis of the natural product turmerone.
Scheme 1
We recently reported the first examples of the retro-Nazarov
reaction in which 3-aryl-2-bromo-4-tert-butoxy-cyclopentanones
underwent ring opening upon treatment with base.1 This reaction
presumably occurred through the mechanism depicted in Scheme
1. The putative oxyallylic cation intermediate 4 underwent
unimolecular retroelectrocyclization faster than potential bimo-
lecular reactions such as elimination or nucleophilic attack.
However, this intermediate is both kinetically and thermody-
namically biased toward ring opening by virtue of its substitution
pattern. Both the alkoxy and aryl (or vinyl) groups should
weaken the bond to be cleaved and provide a driving force for
opening through conjugative stabilization in the product.
point vibrational energies (ZPVE). All absolute energies and
As part of our effort to explore the scope of the reaction, we
geometries are given in the Supporting Information.
embarked on a computational study of the key mechanistic steps
All of the reactions are computed to be exothermic (Table
1). As expected, substitution adjacent to the C3-C4 bond with
atoms bearing lone pairs of electrons lowers the barrier
significantly (relative to parent 6, ∆H0q ) 14.9 kcal mol-1) with
in the hope of gleaning some information about substituent
effects in the process. This information would provide the basis
for further experimental work. In this article, we report the
results of a density functional theory (DFT) study of the barriers
associated with ring opening of various cyclopentenyl oxyallylic
cations and report studies that define the present experimental
limits of the reaction.
q
the acceleration being greatest for the amino (17, ∆H0 ) 4.2
kcal mol-1) and methoxy (7, ∆H0 ) 4.6 kcal mol-1) groups
‡
(2) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M.
A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann,
R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin,
K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi,
R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.;
Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz,
J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;
Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,
M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision x.x; Gaussian,
Inc.: Pittsburgh, PA, 1998.
Energies for oxyallylic cations, transition states, and ring-
opened dienones were computed with Gaussian 98,2 using
Becke’s three-parameter hybrid functional3 (B3LYP) with a
6-31+G* basis set. A restricted approach was used for all
geometry optimizations, energy evaluations, and frequency
analyses of the reactants. All stationary structures were identified
as minima or transition structures by computing analytic
vibrational frequencies, which were also used to compute zero-
(3) (a) Becke, A. D. Phys. ReV. A 1988, 38, 3098-3100. (b) Becke, A. D. J.
Chem. Phys. 1993, 98, 5648. The Lee-Yang-Parr nonlocal correlation
functional: (c) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
(4) Kallel, E. A.; Houk, K. N. J. Org. Chem. 1989, 54, 6006.
† University of Missouri-Columbia.
‡ Justus-Liebig-University.
(1) Harmata, M.; Lee, D. R. J. Am. Chem. Soc. 2002, 124, 14328-14329.
9
10954
J. AM. CHEM. SOC. 2004, 126, 10954-10957
10.1021/ja048942h CCC: $27.50 © 2004 American Chemical Society

The Retro-Nazarov Reaction
A R T I C L E S
Figure 1. Transition structure for the ring opening of 19 (cis and trans forms).
Table 1. Relative (Including ZPVEs, to Starting Material in kcal
mol-1) at B3LYP/6-31+G*
Å) in TS19c, leading to an appreciable barrier. Alternatively,
one may look at either conrotation of 19c as having a higher
barrier due to untoward steric and stereoelectronic effects.4 The
reverse reactions are equally different as the two electronegative
heteroatoms get into each other’s way in the cis approach when
viewed from the product side. One other important point is that
conrotation of 19c demands the inward rotation of either the
methoxy or amino group. Our calculations suggest that rotation
to place the methoxy group inside is favored substantially
kinetically and to a slight extent thermodynamically.
Interestingly, but as expected in the context of work by the
Houk group,4 the retro-Nazarov reaction is expected to be
torquoselective. As illustrated in Scheme 2, while the formation
of both 15p and 16p is favored to essentially the same extent
thermodynamically (∆∆RH0 ) 0.8 kcal mol-1), the barriers for
the two processes differ considerably (4.9 vs 14.4 kcal mol-1),
and selective formation of 15p is anticipated. As found for cis-
and trans-19, the decisive difference in the transition structures
leading to torquoselectivity traces back to their varying stages
of development toward the respective product. TS15 occurs
earlier on the PES, as evident from the 0.050 Å shorter breaking
central bond; all structural parameters of this TS are closer to
the starting material than for the competing TS16.
R1 ) R2 ) R3 ) H; R4 ) -(6)
14.9
4.6
-26.4
-36.6
-45.1
-34.3
-25.0
-27.3
-31.7
-27.5
-30.9
-34.0
-34.8
-39.9
-32.3
-42.0b
-42.3c
-50.6
-29.5
-17.1
-24.7
R1 ) OMe, R2 ) R3 ) H; R4 ) -(7)
R1 ) OMe, R2 ) OMe, R3 ) H; R4 ) -(8a)
R1 ) OMe, R2 ) OMe, R3 ) CH3; R4 ) -(9)
R1 ) OMe, R2 ) H, R3 ) H; R4 ) H (10)
R1 ) OMe, R2 ) H, R3 ) H; R4 ) Li (11)
R1 ) vinyl, R2 ) H, R3 ) H; R4 ) -(12)
R1 ) OMe, R2 ) H, R3 ) Cl; R4 ) -(13)
R1 ) Me2, R2 ) H, R3 ) H; R4 ) -(14)
R1 ) Me/OMe, R2 ) H, R3 ) H; R4 ) -(15)
R1 ) OMe/Me, R2 ) H, R3 ) H; R4 ) -(16)
R1 ) NH2, R2 ) R3 ) H; R4 ) -(17)
R1 ) SH, R2 ) R3 ) H; R4 ) -(18)
0.0b
2.1
2.7
3.7
10.5
7.6
13.9
4.9
14.4
4.2
9.6
R1 ) OMe, R2 ) NH2, R3 ) H; R4 ) -(19c)
11.5b
7.6c
0.0d
10.4
9.8
R1 ) OMe, R2 ) NH2, R3 ) H; R4 ) -(19t)
R1 ) -(CH2)3- ) R2, R3 ) H; R4 ) -(20)
R1 ) -(CH2)4- ) R2, R3 ) H; R4 ) -(21)
R1 ) -(CH2)5- ) R2, R3 ) H; R4 ) (22)
11.5
With respect to our latest experimental work in this area, it
is the ring openings of 6, 7, 12, and 14 that are most relevant.
Except for 7, these oxyallylic species are expected to undergo
ring opening at a relatively slow rate and thus be more subject
to side reactions. To examine this, we set out to determine the
minimum substitution requirements adjacent to the C3-C4 bond
for the oxyallylic cation intermediate in the ring-opening
process.
We began by removing the alkoxy group and testing a system
that was rather heavily substituted, but possessed only two alkyl
and one aryl substituent adjacent to the breaking bond. The
synthesis of the substrate and subsequent manipulations are
shown in Scheme 3. The reaction of enone 235 with p-tolyl
cuprate afforded bromoketone 24 in 92% yield. This was heated
in refluxing trifluoroethanol (TFE) for 1.5 h in the presence of
excess triethylamine and afforded the dienone 25 in 77% yield.
Subsequent treatment with an organocopper species derived
from methylmagnesium chloride and cuprous iodide gave the
natural product turmerone 26 in 73% yield.6
a
b
c
d
trans. NH2 group inside. OMe group inside. The transition
structure is associated with a negligible barrier that disappears when ZPVE
corrections are applied, i.e., the model reaction is practically barrierless.
q
and smallest for the thiol group (18, ∆H0 ) 9.6 kcal mol-1).
This trend correlates with the exothermicity of the reaction for
these substituents (∆RH0 ) -26.4, -39.9, -36.6, and -32.3
kcal mol-1, respectively). When two methoxy groups straddle
the breaking bond, ring opening should proceed highly exo-
thermically (∆RH0 ) -45.1 kcal mol-1) without a barrier (8).
Geminal dialkyl substitution at C4 provides for some accelera-
tion, but not much, and ring opening of a species such as 14 is
q
not expected to be significantly faster than 6 (∆∆H0 ) 1.0
kcal mol-1). Vinyl substitution (12) is expected to have a more
substantial impact on ring opening, but not so important as an
alkoxy group.
The cis and trans substitution pattern in, for example, 19 also
makes quite a difference (Figure 1) owing to the much earlier
nature of the TS for the trans pathway. In accord with
Hammond’s postulate, TS19t is much less developed (early
transition structure) compared to TS19c, as evident from the
considerable difference in bond lengths for the two transition
structures. Note in particular the key central bond distance of
the breaking bond that is significantly more lengthened (by 0.219
The general substitution pattern present in 24 was effective
in promoting retro-Nazarov reaction in other cases. These results
(5) Paquette, L. A.; Tae, J.; Arrington, M. P.; Sadoun, A. H. J. Am. Chem.
Soc. 2000, 122, 2742.
(6) Motoyoshiya, J.; Miyajima, M.; Hirakawa, K.; Kakurai, T. J. Org. Chem.
1985, 50, 1326.
9
J. AM. CHEM. SOC. VOL. 126, NO. 35, 2004 10955

A R T I C L E S
Harmata et al.
Scheme 2
Scheme 3
are summarized in Table 2. It is anticipated that geminal dialkyl
substitution flanking the breaking bond coupled with an aryl or
alkenyl substituent on the adjacent carbon will be generally
useful for the retro-Nazarov reaction. However, three alkyl
groups are not sufficient to allow efficient reaction to occur
(Table 2, entry 6).
Removing a single alkyl group also allowed the reaction to
proceed. Treatment of 37 with triethylamine in refluxing
hexafluoroisopropanol (HFIP) gave the dienone 38 in 59% yield
along with a small amount (9%) of 39 (eq 1).7 The stereochem-
Table 2. Retro-Nazarov Reaction of Selected Bromoketones
istry of 37 was inferred from its synthesis, which involved
addition of diphenylcuprate to the corresponding bromoenone.
Data for this compoud (see Supporting Information) indicated
a single isomer was present. Conrotatory ring opening of an
entry
reactant
R
product
yield (%)
1
2
3
4
6
27
29
31
33
35
phenyl
28
30
32
34
36
91
65
82
75
low
p-methoxy phenyl
vinyl
2-methylpropenyl
methyl
(7) When this reaction was conducted in trifluoroethanol in triethylamine, 38
and the corresponding Michael adduct were produced in a 1:1 ratio (crude
NMR) in a combined yield of 60%.
9
10956 J. AM. CHEM. SOC. VOL. 126, NO. 35, 2004

The Retro-Nazarov Reaction
A R T I C L E S
oxyallylic cation generated from 37 would thus be expected to
produce 38 as the all E isomer, as shown.
However, removal of the remaining methyl group completely
stopped retro-Nazarov ring opening. Bromoketone 40 gave only
elimination product upon treatment with base under standard
reaction conditions. No other reaction conditions afforded retro-
Nazarov product.
Finally, the reaction of 41 with triethylamine in trifluoro-
ethanol afforded the retro-Nazarov product 42 in 51% yield
(eq 2). This compound arose from retro-Nazarov reaction and
In summary, we have computed transition structures and
reaction energies for the retro-Nazarov reaction of a variety of
cyclopentenyl oxyallylic cations. Our experimental findings give
greater meaning to the computations and suggest that reactions
with computed gas-phase barriers of near 10 kcal/mol will not
proceed effectively along the retro-Nazarov pathway. The results
shown in Scheme 2 are especially interesting in this context, as
we anticipate that high torquoselectivity should be observed in
the retro-Nazarov reaction given the relatively large energy
difference between the two conrotations. Efforts to study this
are in progress and will be reported in due course.
subsequent Michael addition of trifluoroethanol to the resulting
dienone. The yield of this reaction could be improved through
the use of tetramethylpiperidine as base, which produced 42 in
61% yield. In the best case to date for this reaction, heating 41
in the presense of TMP in HFIP afforded both 43 and 44 in a
2:1 ratio in a combined yield of 77% (eq 3). Other reaction
conditions afforded slightly different ratios of products, but it
is clear that a single alkoxy substituent is sufficient to enable
the retro-Nazarov reaction.
Acknowledgment. This work was supported by the National
Science Foundation, to whom we are grateful.
Supporting Information Available: x,y,z-Coordinates and
absolute energies for all computed species, experimental
1
procedures, and copies of H and 13C NMR. This material is
available free of charge via the Internet at http://pubs.acs.org.
JA048942H
9
J. AM. CHEM. SOC. VOL. 126, NO. 35, 2004 10957