Remote substituent effects on phenylchlorocarbene C-H insertion reactions

Article abstract of DOI:10.1021/ol990161b

Formula presented Phenylchlorocarbene inserts into the tertiary CH bonds of 1,3-dimethyladamantane and 1-X-adamantanes (X = H, OMe COOMe, Cl, CN). There is a good correlation between the relative rate constants for these insertions and the inductive subst

Full text of DOI:10.1021/ol990161b

ORGANIC
LETTERS
1999
Vol. 1, No. 5
819-822
Remote Substituent Effects on
Phenylchlorocarbene C−H Insertion
Reactions
Robert A. Moss* and Shunqi Yan
Department of Chemistry, Rutgers, The State UniVersity of New Jersey,
New Brunswick, New Jersey 08903
moss@rutchem.rutgers.edu
Received July 12, 1999
ABSTRACT
Phenylchlorocarbene inserts into the tertiary CH bonds of 1,3-dimethyladamantane and 1-X-adamantanes (X ) H, OMe, COOMe, Cl, CN). There
is a good correlation between the relative rate constants for these insertions and the inductive substituent constants, σI(X), with G ) −1.5.
Absolute rate constants for the insertions range from 2.5 × 105 M-1 s-1 for Ad-Me2 to 2.5 × 104 M-1 s-1 for Ad-CN. B3LYP/6-31G* calculations
give a good account of reactivity in these systems.
We recently described the absolute kinetics of phenylchloro-
carbene (PhCCl) insertion reactions with a variety of
substrates.1 In benzene solution, for example, PhCCl inserted
into a tertiary CH of adamantane (Ad-H; 1a) with k ) 1.26
Hammett study of CCl2 insertions into the tertiary CH of
substituted cumenes gave F ) -1.19 (σ) or -0.89 (σ+),
consistent with the development of δ+ on the cumyl carbon
at the transition state.2a Analogous insertions into the Si-H
bonds of silacumenes revealed F values ranging from -0.974
to -0.63.5
With the cumyl and silacumyl substrates, both inductive
and resonance effects of the substituents act upon the reaction
center. We imagined that a study of the remote substituent
effects attending the efficient1,6 carbene insertions into
bridgehead adamantane C-H bonds could “isolate” the
inductive effect, providing a direct estimate of its importance.
Qualitative results for CCl2 insertions (at C3-H) into
1-substituted adamantanes (1) demonstrate that electron-
withdrawing substituents deactivate the substrate and lower
product yields.6 Here, we present quantitative data for the
related PhCCl insertions: both kinetics and computational
studies are in good accord with generalized TS 2 and
implicate a surprisingly large inductive effect operating on
these C-H insertion reactions.
× 105 M-1 s-1 (per H atom) and with Ea ) 3.2 kcal/mol
and ∆Sq ) -24 eu.1 The transition state (TS) for carbene
C-H insertion can be schematically represented as 2, where
a hydride-like migration between the substrate and carbenic
carbon atoms engenders charge separation, with positive
charge on the former and negative charge on the latter.2,3
Insertions of (for example) CCl2 into benzylic or R-ether
C-H bonds are readily interpretable with TS 2.2,3 Thus, a
(4) Watanabe, H.; Ohsawa, N.; Sudo, T.; Hirakata, K.; Nagai, Y. J.
Organomet. Chem. 1977, 128, 27.
(1) Moss, R. A.; Yan, S. Tetrahedron Lett. 1998, 39, 9381.
(2) (a) Seyferth, D.; Cheng, Y. M. J. Am. Chem. Soc. 1973, 95, 6763.
(b) Seyferth, D.; Mai, V. M.; Gordon, M. E. J. Org. Chem. 1970, 35, 1993.
(3) Pascal, R. A., Jr.; Mischke, S. J. Org. Chem. 1991, 56, 6954.
(5) Seyferth, D.; Damrauer, R.; Mui, J. Y.-P.; Jula, T. F. J. Am. Chem.
Soc. 1968, 90, 2944.
(6) Tabushi, I.; Yoshida, Z.; Takahashi, N. J. Am. Chem. Soc. 1970, 92,
6670.
10.1021/ol990161b CCC: $18.00 © 1999 American Chemical Society
Published on Web 08/13/1999

Substrates included adamantane (Ad-H, 1a) and its 1-X
derivatives Ad-OMe (1b), Ad-COOMe (1c), Ad-Cl (1d), and
Ad-CN (1e), as well as the readily available (Aldrich) 1,3-
dimethyladamantane (Ad-Me2; 3). For preparative and
product identification purposes, PhCCl was generated either
thermally (100-110 °C, 2-6 h) or photochemically (Ad-
Cl, λ > 320 nm, 30 min) from phenylchlorodiazirine7 in
concentrated benzene solution (A374 > 3.0). The PhCCl
reacted with 18-37 mmol of Ad-X in 2-6 mL of benzene
to afford products 4a-e and 5, which were purified by silica
gel chromatography and characterized by their NMR and
GC-MS spectra, as well as elemental analysis or high-
resolution mass spectrometry.
in order to demonstrate the internal consistency of the relative
reactivities. Appropriate relative rate constants were aver-
aged; the results, relative to Ad-H, are displayed in Table 1,
where average deviations in krel are better than (5%.
Table 1. Kinetics of PhCCl Insertion into the Adamantyl
Tertiary CH Bonda
substrate
krel (tot)b krel (per H)c 105kabs, M-1 s-1 d
σIe
Ad-Me2 (3)
Ad-H (1a )
Ad-OMe (1b)
Ad-COOMe (1c) 0.315
Ad-Cl (1d )
Ad-CN (1e)
0.995
1.000g
0.420
1.99
1.00g
0.56
0.42
0.26
0.20
2.51
1.26h
0.70
0.53
0.33
0.25
-0.02f
0.00
0.30
0.32
0.47
0.57
0.197
0.150
a In benzene at 23 °C. b Reactivity relative to Ad-H, not statistically
corrected. c Per-bond reactivity, relative to a single Ad-H tertiary CH bond.
d Absolute rate constant, per C-H bond, derived from the absolute rate
constant for PhCCl + Ad-H,1 and krel (per CH). e Inductive substituent
constant for X in Ad-X.10a f σI(Me) ) -0.01; for 2 Me’s we take σI )
-0.02. g Standard substrate. h See ref 1.
Yields of insertion products ranged from 96% (5) and 90%
(1a,b) to 60% (4d) and 35% (4e).8 Carbene dimer and azine
byproducts increased from 1% and 3% yields with substrate
3 to 15% and 25% with 1d, and even higher with le.8 Clearly,
electron-withdrawing substituents adversely affect the reac-
tivity of Ad-H.6
The “spread” in krel is a factor of 10 from the most (Ad-
Me2) to the least reactive substrate (Ad-CN), while the
corresponding absolute rate constants for C-H insertion
range from 2.5 × 105 to 2.5 × 104 M-1 s-1 (per bridgehead
H atom). There is a clear correlation between the reactivity
of a substrate’s tertiary CH bond and the electronic character
of its substituent, X, as quantitated by the relation between
log krel (per H) and σI(X), the inductive substituent constant10
(see Figure 1). Here, F ) -1.50, consistent with TS 2, where
The absolute rate constant for PhCCl insertion into 1a is
5.02 × 105 M-1 s-1 (in benzene9 at 23 °C), corresponding
to 1.26 × 105 M-1 s-1 per bridgehead C-H bond.1 We
determined relatiVe rates for C-H insertions with substrates
1b-e and 3 by competition reactions in which an insuf-
ficiency of PhCCl was allowed to compete with ∼50-fold
excesses of substrate pairs. PhCCl was generated by pho-
tolysis (λ > 320 nm) of the diazirine in benzene solution at
23 °C for 20 min, after which no diazirine was detectable
by UV spectroscopy. Molar ratios of the derived insertion
products, 4a-e and 5, were determined by capillary GC on
a CP-Sil-5CB column, using a calibrated flame ionization
detector and electronic integration. Relative rate constants
were calculated from kA/kB ) (SB/SA)(PA/PB), where SB/SA
is the initial molar ratio of substrates and PA/PB is the molar
product ratio. Competitions were performed in duplicate, with
GC analyses in triplicate, and the reproducibilities of kA/kB
were better than (4%.
Relative reactivities were determined for the following
substrate pairs (kA/kB shown in parentheses): Ad-Cl/Ad-H
(0.205), Ad-Me2/Ad-Cl (5.12), Ad-OMe/Ad-H (0.44), Ad-
COOMe/Ad-Cl (1.62), Ad-CN/Ad-H (0.150). Additional
competitions were carried out between Ad-Me2/Ad-Cl, Ad-
OMe/Ad-Me2, Ad-COOMe/Ad-OMe, and Ad-Cl/Ad-COOMe
Figure 1. Correlation of log k/k0 with σI for insertions of PhCCl
into the tertiary CH bonds of substrates 3 and 1a-e. See text for
discussion.
(7) Graham, W. H. J. Am. Chem. Soc. 1965, 87, 4396.
(8) Product 4e, from Ad-CN, was accompanied by 15% of PhCCl dimer
and 50% of (PhCCldN)2 azine. 4e could not be separated from excess 1e
by chromatography and was identified in situ by GC-MS.
(9) The insertion is 2-3 times slower in benzene than in pentane; benzene
“complexes” PhCCl and modulates its reactivity: Moss, R. A.; Yan, S.;
Krogh-Jespersen, K. J. Am. Chem. Soc. 1998, 120, 1088. Krogh-Jespersen,
K.; Yan, S.; Moss, R. A. J. Am. Chem. Soc. 1999, 121, 6269.
an electron-withdrawing substituent destabilizes δ+ on the
Ad carbon atom and slows the insertion reaction, while an
(10) (a) Charton, M. Prog. Phys. Org. Chem. 1981, 13, 119. (b) Hansch,
C.; Leo, A.; Taft, R. W. Chem. ReV. 1991, 91, 165.
820
Org. Lett., Vol. 1, No. 5, 1999

electron-releasing substituent has the opposite effect. For
comparison, correlation of σI(X) with krel for bridgehead H
abstractions from 1 by the trichloromethyl radical at 40 °C
gives F ) -0.90.11 Charge separation (in the sense of 2) is
less pronounced in the electrophilic radical abstraction by
CCl3 than in the PhCCl insertion.
distances rather insensitive to substrate identity: for 3
through 1e, r2 varies from 1.374 to 1.387 Å, with bonding
between H and the carbenic carbon well established.
Distances r1 (2.497-2.489 Å) and r4 (2.491-2.474 Å) are
also computed to be relatively substituent insensitive. The
TS for the Ad-H/PhCCl insertion is illustrated in Figure 2.
Although there is appreciable scatter for the X ) Me and
H points in Figure 1, the overall correlation is of good
quality, with r ) 0.97, significant at the 99% confidence
level. The apparent scatter may be largely due to uncertainty
in σI(Me), given as -0.01 ((0.02).10a Using a more negative
value (e.g. -0.0510b) improves the correlation (r ) 0.986)
and marginally trims the F value to -1.40.
The magnitude of F (∼-1.5) is rather large for “remote”
substituents acting on the δ+ of 2 only by an inductive effect;
recall that for CCl2 insertion into cumenes, where para
substituents interact with the reaction center by both reso-
nance and inductive effects, F ranges only from -0.89 to
-1.19.2a Substantial contribution of a resonance component,
originating at X, to the stabilization of δ+ in TS 2 (illustrated
for X-Ad+ by the “frangomeric”12 resonance structures 6) is
Figure 2. B3LYP/6-31G* transition state for insertion of PhCCl
into the tertiary CH bond of substrate 1a. Uninvolved H atoms are
omitted for clarity.
As shown in Table 2, considerable positive charge is
induced on the adamantyl carbon of TS 7, with comparable
negative charge buildup on the carbenic carbon (NBO
analysis). These charges are slightly sensitive to the Ad
substituents; both positive and negative charges decrease as
the TS moves toward product with increasing electron-
withdrawing character of the substituents. Substituent sen-
sitivity is more strongly displayed by the computed activation
energies (Table 2), which (uncorrected for zero point energy)
range from 8.55 (Ad-Me2) to 9.94 kcal/mol (Ad-CN).
Although these values are too high,1,16 their trend with
substituent variation is appropriate. Indeed, there is a
reasonable correlation (r ) 0.93) between the computed Ea’s
and σI(X), with F ) 1.86; electron-withdrawing substituents
destabilize the positive charge on the migration origin of 7,
raising both TS and activation energies.
unlikely, because the krel of Ad-OMe is well-correlated by
σI(OMe); cf. Figure 1. Note that the electronic effect of X,
here termed an “inductive effect,” may be propagated through
space, through the intervening σ bonds, or by a combination
of modalities; partition of the transmission mode cannot be
easily accomplished.13
The B3LYP hybrid density functional method, as con-
tained in the Gaussian 94 package,14 affords reasonable
transition states and energetic trends for PhCCl/C-H inser-
tion reactions.1 B3LYP/6-31G* transition states, represented
by 7, were located for the reactions of PhCCl with 1a-e
Table 2. Computational Resultsa
substrate
Ea, kcal/molb
δ+ c
δ- d
Ad-Me2 (3)
Ad-H (1a )
Ad-OMe (1b)
Ad-COOMe (1c)
Ad-Cl (1d )
8.55
9.10
9.30
9.46
9.64
9.94
0.202
0.197
0.195
0.193
0.185
0.184
0.190
0.189
0.188
0.188
0.183
0.183
and 3, and confirmed by frequency calculations that gave
only one negative frequency. Transition state 7 is “late,” with
the migrant hydride closer to the PhCCl terminus than the
adamantyl origin.15 Thus, r3 < r2 for all substrates, with these
Ad-CN (1e)
a B3LYP/6-31G*. b Uncorrected for zero-point energy. c Positive charge
on adamantyl carbon in TS 7. d Negative charge on carbenic carbon in TS
7.
(11) Owens, P. H.; Gleicher, G. J.; Smith, L. M., Jr. J. Am. Chem. Soc.
1968, 90, 4122.
(12) Grob, C. A.; Fischer, W.; Katayama, H. Tetrahedron Lett. 1976,
2183 and references therein.
(13) Exner, O.; Charton, M.; Galkin, V. J. Phys. Org. Chem. 1999, 12,
289 and accompanying papers.
(14) GAUSSIAN 94, Revision E.2; Gaussian, Inc., Pittsburgh, PA, 1995.
The B3LYP/6-31G* calculations thus provide a service-
able model for the PhCCl insertion reactions, sensitively
Org. Lett., Vol. 1, No. 5, 1999
821

reproducing the observed substituent effects on carbenic
reactivity. At the same time, the adamantane system furnishes
an excellent scaffold on which to probe the electronics of
this archetypal carbene reaction.
Acknowledgment. We are grateful to Professor Marvin
Charton for helpful discussions and to the National Science
Foundation for financial support. Access to the computational
resources of the Center for Computational Neuroscience at
Rutgers University (Newark) is much appreciated.
(15) A similar observation was made for the computed PhCCl/cyclo-
hexane TS.1
(16) The computed Ea for the PhCCl/Ad-H insertion is 9.1 kcal/mol,
whereas the experimental value is 3.2 kcal/mol.
OL990161B
822
Org. Lett., Vol. 1, No. 5, 1999