Intra- and intermolecular diastereoselectivity of 5-hydroxy-2-adamantylidene [23]

Full text of DOI:10.1021/ja001123m

7430
J. Am. Chem. Soc. 2000, 122, 7430-7431
Scheme 1: Reactions of 5-Hydroxyadamantylidene (2)
Intra- and Intermolecular Diastereoselectivity of
5-Hydroxy-2-adamantylidene†
Michael M. Bobek and Udo H. Brinker*
Institut fu¨r Organische Chemie, UniVersita¨t Wien
Wa¨hringer Strasse 38, A-1090 Wien, Austria
ReceiVed March 30, 2000
ReVised Manuscript ReceiVed June 13, 2000
The rationale of the diastereoselectivity in the reactions of
alleged sterically unbiased molecules has been reviewed exten-
sively, but remains a controversial issue.1 An explanation of this
phenomenon was brought forth by Cieplak2 who invoked hyper-
conjugative donation of electrons into the σ*-orbital of the
incipient bond formed. Other authors criticized this approach3
because of its neglect of direct electrostatic interactions.4 Another
explanation is based on the distortion of the geometry of the
precursor that could induce a steric preference for the reagent’s
attack.5 Although various types of reactions with substituted
adamantanes have been studied,6 virtually nothing is known about
the behavior of the corresponding carbenes.7 The geometry of
these highly reactive intermediates should closely resemble the
activated complexes of their reactions.8 Therefore, the investiga-
tion of carbene reactions should allow general conclusions about
the nature of the observed diastereoselectivity.
To this end, the intra- and intermolecular insertion reactions
of 5-hydroxy-2-adamantylidene (2) were studied in the gas phase
and in solution. 2 appears to be an ideal system for the following
reasons. Direct steric interactions with the approaching reagent
should be negligible, because the substituent at C-5 is located
practically opposite of the divalent carbon. Moreover, the rigidity
of the adamantane skeleton prohibits conformational changes that
would influence the course of the reaction.
Carbene 2 was generated thermally and photochemically from
2-azi-5-hydroxyadamantane 1,9 which was prepared from the
corresponding ketone.10 In general, carbenes can be generated by
photolysis of diazirines via two competing pathways.11 They can
be formed by the extrusion of nitrogen either directly from the
excited state diazirine or indirectly from the linear diazo
compound.12 Calculations13 and experimental results14 suggest a
singlet ground state for adamantylidene.
In polar protic solvents such as methanol, however, the diazo
compound is readily protonated leading to the corresponding
carbocation.15 This carbeniumion reacts with methanol to yield
methyl ethers. The possible formation of an interfering carbocation
from the diazo compound must be inhibited, if only carbene
reactions are to be studied.15
The intramolecular 1,3 C-H insertion products 3 and 4
(Scheme 1) were formed quantitatively by vacuum pyrolysis in
a 90:10 ratio in favor of the anti insertion product 3 (Table I).16,17
Diazirine 1 was photolyzed18 in cyclohexane to yield products 7
and 8. For the intermolecular C-H insertion reaction, the ratio
was reversed to 89:11 in favor of the syn-substituted (Figure 1)
product 7. The photolysis of 1 in methanol was carried out without
fumaronitrile (FN) and in the presence of at least a one 100-fold
excess of FN. In pure methanol, the O-H insertion occurs in
high yields with a ratio of 74:26 in favor of the syn-substituted
product 5.19 However, in the presence of FN, a potent 1,3-
dipolarophile, this ratio is increased to 85:15. The combined yield
of O-H insertion products 5 and 6 dropped from 89 to 58%,20
independent of the concentration of FN chosen. Since the diazo
compound is completely scavenged in a 0.5 M FN solution, the
higher ratio of 5 to 6 (85:15) can be exclusively attributed to
carbene insertions.
In addition, an ab initio geometry optimization of the singlet
ground state of 2 was performed.21 The calculated structure shows
a deviation of 7.4° away from the O-H group of the carbene
carbon from the H1-C1-C3-H3 plane. Even when choosing a
starting geometry with C2 bent toward the O-H substituent by
about 20°, the carbene carbon passes the H1-C1-C3-H3 plane
and settles in the described conformation.
Analysis of the orbitals of this conformation reveals that the
unoccupied p-orbital (LUMO) of the singlet carbene participates
in the C1-C8 and C3-C10 bonds (Figure 2).
(15) Kirmse, W.; Meinert, T. J. Chem. Soc., Chem. Commun. 1994, 1065;
Kirmse, W.; Meinert, T.; Modarelli, D. A.; Platz, M. S. J. Am. Chem. Soc.
1993, 115, 8918.
(16) Teager, D. S.; Murray, R. K., Jr. J. Org. Chem. 1993, 58, 5548.
(17) An analogous behavior was found for 5-chloroadamantylidene and
5-methyladamantylidene. See: Hirsl-Starcevic, S.; Majerski, Z. J. Org. Chem.
1982, 47, 2520.
(18) Hanovia 700-W medium-pressure mercury arc lamp doped with FeI2;
Pyrex filter; 2 h photolysis time at 12-15 °C.
† Carbene Rearrangements. 54. For Part 53, see: Wagner, R. A.; Weber,
J.; Brinker, U. H. Chem. Lett. 2000, 246.
(1) See the special issue: Chem. ReV. 1999, 99, 1067.
(2) Cieplak, A. S. J. Am. Chem. Soc. 1981, 103, 4540.
(3) Frenking, G.; Ko¨hler, K. F.; Reetz, M. T. Angew. Chem., Int. Ed. Engl.
1991, 30, 1146.
(4) Wu, Y.-D.; Tucker, J. A.; Houk, K. N. J. Am. Chem. Soc. 1991, 113,
5018; Paddon-Row, M. N.; Wu, Y.-D.; Houk, K. N. J. Am. Chem. Soc. 1992,
114, 10638; Adcock, W.; Cotton, J.; Trout, N. A. J. Org. Chem. 1994, 59,
1867.
(5) Gung, B. W.; Wolf, M. A. J. Org. Chem. 1996, 61, 232; Gung, B. W.;
Wolf, M. A.; Mareska, D. A.; Karipides, A. J. Org. Chem. 1994, 59, 4899.
(6) Kaselj, M.; Chung, W.-S.; le Noble, W. J. Chem. ReV. 1999, 99, 1387.
(7) Kaselj, M.; le Noble, W. J.J. Org. Chem. 1996, 61, 4157; ref 17.
(8) Hammond, G. S. J. Am. Chem. Soc. 1955, 77, 334.
(9) Yurchenko, A. G.; Isaev, S. D.; Novoselov, E. F. J. Org. Chem. USSR
1984, 20, 201.
(10) Bobek, M. M.; Brinker, U. H. Synth. Commun. 1999, 29, 3221.
(11) Liu, M. T. H.; Stevens, I. D. R. In Chemistry of Diazirines; Liu, M.
T. H., Ed., CRC: Boca Raton, 1987; Vol. I, p 112.
(12) Bonneau, R.; Liu, M. T. H. J. Am. Chem. Soc. 1996, 118, 7229.
(13) Shustov, G. V.; Liu, M. T. H. Can. J. Chem. 1998, 76, 851.
(14) Bally, T.; Matzinger, S.; Truttmann, L.; Platz, M. S.; Morgan, S.
Angew. Chem., Int. Ed. Engl. 1994, 33, 1964; Morgan, S.; Jackson, J. E.;
Platz, M. S. J. Am. Chem. Soc. 1991, 113, 2782.
(19) 5 and 6 are also formed by solvolysis of the corresponding tosylates.
Grob, C. A.; Wang, G.; Yang, C. Tetrahedron Lett. 1987, 28, 1247.
(20) Fumaronitrile traps the diazo compound in a 1,3-dipolar cycloaddition,
preventing the formation of a carbocation via the diazonium ion route (also
see ref 15). In addition to 5 and 6, several products probably stemming from
the 1,3-dipolar addition to the diazo compound and decomposition products
thereof could be detected. For the photochemical decomposition of 1,3-dipolar
addition products see: Sander, W.; Wrobel, R.; Komnick, P.; Rademacher,
P.; Muchall, H. M.; Quast, H. Eur. J. Org. Chem. 2000, 91.
(21) RHF/6-31G*, Fletcher-Reeves conjugate gradient, Hyperchem 5,
Hypercube Inc.: Gainesville, FL.
10.1021/ja001123m CCC: $19.00 © 2000 American Chemical Society
Published on Web 07/14/2000

Communications to the Editor
J. Am. Chem. Soc., Vol. 122, No. 30, 2000 7431
Table 1: Ratios of Diastereomeric Products Formed from 1
reactant seem very unlikely. The insertion of a carbene into C-H
bonds is a concerted reaction of the singlet state.24 This is in
contrast to the polar mechanism O-H insertion.25 One would
expect a higher selectivity for the charged intermediate involved
in the O-H insertion. The comparable diastereoselectivity
observed in the O-H (85:15) and C-H (89:11) insertion reactions
strongly argues against direct electrostatic control of the reagent’s
orientation. Although a hyperconjugative stabilization of the σ*-
orbital of the new C-C or C-O bonds formed in the transition
state cannot be ruled out, it is not necessary to invoke Cieplak’s
concept2 to explain the results of the intermolecular reactions. It
is not clear, however, whether the Cieplak approach is applicable
also for the intramolecular case. If hyperconjugation with the
newly formed C-H bond is considered, the experimental results
are in agreement with Cieplak’s model. On the other hand,
consideration of the C-C bond formed contradicts the experi-
mental findings.
product
distributionb
reaction conditions
yielda %
pyrolysis
photolysis
photolysis
3
90
4
10
gas phase (neat)
>99
7
89
8
11
5 mM in cyclohexane
5 mM in MeOH
92
5
74
85
83
6
26
15
17
89c
57
58
5 mM in MeOH (0.5 M fumaronitrile)
5 mM in MeOH (1.0 M fumaronitrile)
a Combined yield of insertion products, determined by GC. b Error:
(2%, determined by GC and NMR spectroscopy. c 1-Hydroxyada-
mantane, 3 and 4 are formed in small amounts.
The calculations show the singlet ground state of carbene 2 to
be most stabilized when bent 7.4° away from the O-H substituent.
According to the principle of least motion,26 the intramolecular
C-H insertion should preferentially occur into the C-H bonds
proximal to the reactive center. This agrees well with the
experimental findings; the major product 3 stems from an insertion
into the C8-H or C10-H bond. The predominant intermolecular
O-H and C-H insertion products 5 and 7 result from an approach
of the reagent via the more exposed syn face of adamantylidene
2. The bulkier part of the reacting molecule can be better
accommodated there. Both, the intra- and intermolecular selectiv-
ity can, therefore, best be explained by the distortion of the
geometry of carbene 2. EFOE analysis of 2 shows, that the plane-
divided accessible space (PDAS) of 266.5 au3 of the syn-side is
much larger than the PDAS of 196.2 au3 of the anti-side of the
molecule.27 These results emphasize that the observed diastereo-
selectivity is caused by changes of the diastereofacial accessibility
already present in the ground state. Therefore, in the adamantyl-
idene system, the outcome of a reaction is much more governed
by steric factors than expected so far.6 More calculations and
experiments including electron-donating substituents, are in
progress to fully address these questions.
Figure 1. Possible sides for the attack on 2.
Figure 2. HOMO (left) and LUMO (right) of 2.
Acknowledgment. We are indebted to the Fonds zur Fo¨rderung der
wissenschaftlichen Forschung in O¨ sterreich (Project P12533-CHE) for
financial support, to Dr. C. Klein for granting access to computer facilities
and to Professor H. Kalchhauser for recording the NMR spectra of 7.
Inspiring discussions with M. G. Rosenberg are highly appreciated. We
are grateful to Professor S. Tomoda, University of Tokyo, for performing
the EFOE analysis and for helpful discussions.
It is generally accepted that an electron-withdrawing substituent
in the 5-position of the adamantane skeleton can reduce the
electron density from the C1-C9 and the C3-C4 bonds.6 This is
due to hyperconjugation with the antibonding σ*-orbital of the
C-O bond and the -I effect of the O-H group. As a
consequence, the empty p-orbital of C2 preferentially interacts
with the more electron rich C1-C8 and C3-C10 bonds. This
hyperconjugative interaction can account for the observed devia-
tion of the geometry of the divalent carbon at C2. This conclusion
is corroborated by calculations of unsubstituted adamantylidene13
and low-temperature single-crystal X-ray structures of 2-adaman-
tyl cations.22 These electronically related compounds show a
similar geometrical distortion caused by hyperconjugation.22
Single-crystal X-ray analysis of carbene precursor 1 reveals no
significant distortion of C2 and the diazirine ring.23 Therefore,
the observed selectivity cannot be attributed to 1.
Supporting Information Available: Analytical data (NMR, MS, IR)
of compounds 3 to 8 and structural data of 2 (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
JA001123M
(24) Jones, W. M.; Brinker, U. H. In Pericyclic Reactions; Marchand, A.
P., Lehr, R. E., Eds.; Academic: New York, 1977; Vol. I, p 109.
(25) Kirmse, W. In AdVances in Carbene Chemistry; Brinker, U. H., Ed.;
JAI: Greenwich, 1994; Vol. I, p 1.
(26) Hine, J. J. Org. Chem. 1966, 31, 1236.
(27) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.; Johnson,
B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G. A.;
Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski, V. G.;
Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.; Nanayakkara,
A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.; Wong, M. W.;
Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Binkley,
J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-Gordon, M.; Gonzalez,
C.; Pople, J. A. Gaussian 94; Gaussian, Inc.: Pittsburgh, PA, 1995. Tomoda,
S., private communication: HF/6-31G*(d). EFOE density of LUMO: syn:
3.1403%, anti: 1.5494%; EFOE Density of HOMO: syn: 1.2625%, anti:
1.7965%; for an overview of the method see: Tomoda, S. Chem. ReV. 1999,
99, 1243; Tomoda, S.; Senju, T. Tetrahedron 1999, 55, 5303.
Now one can begin to answer the question about the cause of
the observed diastereoselectivity. In the intermolecular reactions,
direct electrostatic interactions through space with the approaching
(22) Laube, T. Acc. Chem. Res. 1995, 28, 399.
(23) Cambridge Crystallographic Data Centre, CCDC-137063. Bobek, M.
M.; Giester, G.; Ka¨hlig, H.; Brinker, U. H. Tetrahedron Lett. 2000, 41, in
press.