Evidence for β-chlorocarbenium and β-bromocarbenium ions

Article abstract of DOI:10.1021/ol070673n

Isotopic perturbation of degenerate equilibrium is used to determine whether tetramethylethylenechloronium and tetramethylethylenebromonium ions are closed 1,2-bridged structures or rapid equilibria of open β-halocarbenium ions. The observed ¹³C NMR isotope shifts are consistent with a combination of large equilibrium shifts and small upfield intrinsic shifts. The presence of equilibrium shifts in both halonium ions indicates that these ions are not closed 1,2-bridged structures. Rather, they are best represented by equilibria of β-halocarbenium ions.

Full text of DOI:10.1021/ol070673n

ORGANIC
LETTERS
2
007
Vol. 9, No. 12
317-2320
Evidence for
â-Chlorocarbenium and
2
â
-Bromocarbenium Ions
Brian K. Ohta,* Ruth E. Hough, and Jeffrey W. Schubert
Department of Chemistry, VillanoVa UniVersity, VillanoVa, PennsylVania 19085
brian.ohta@VillanoVa.edu
Received March 19, 2007
ABSTRACT
Isotopic perturbation of degenerate equilibrium is used to determine whether tetramethylethylenechloronium and tetramethylethylenebromonium
13
ions are closed 1,2-bridged structures or rapid equilibria of open
with a combination of large equilibrium shifts and small upfield intrinsic shifts. The presence of equilibrium shifts in both halonium ions
indicates that these ions are not closed 1,2-bridged structures. Rather, they are best represented by equilibria of -halocarbenium ions.
â-halocarbenium ions. The observed C NMR isotope shifts are consistent
â
A fundamental question in the study of 1,2-bridged halonium
ions is whether they exist as closed structures with equal
C-X bond orders (a) or as rapid equilibria of open
structures is isotopic perturbation of degenerate equilibrium.4
Servis and Domenick applied the method to 1-d and
concluded that the ion is a closed structure (1a). We applied
the method to 2-d and now report it is an equilibrium of
â-chlorocarbenium ions (2b). Additionally, we provide
evidence for 1,2-methyl shifts in both 1-d and 2-d . On the
basis of this new evidence, we reexamined Servis’ work and
report 1-d is also an equilibrium of â-bromocarbenium ions
1b).
As its name suggests, isotopic perturbation of degenerate
equilibrium uses equilibrium isotope effects to alter dynami-
6
5
1
â-halocarbenium structures with unequal bond orders (b).
3
6
3
Scheme 1. Closed Halonium Ion (a) and Open
â-Halocarbenium Ions (b)
6
(
1
3
cally equivalent C NMR chemical shifts. The method
requires molecules with an asymmetric distribution of
13
deuteria. The observed C chemical shift difference between
On the basis of chemical shifts and carbon-fluorine
the labeled and nonlabeled molecules (∆obs ) δC(D) - δC(H)
)
coupling constants, Olah et al. showed that the tetrameth-
can indicate the presence of an equilibrium. In both closed
2
ylethylenefluoronium ion is an open â-fluorocarbenium ion.
and open structures, the deuteria impart intrinsic isotope shifts
A similar conclusion for the analogous bromonium (1) and
chloronium (2) ions could not be made on the basis of
chemical shifts.3 A direct method for resolving these
n
(
o
∆ , where n is the number of bonds between the carbon
and deuterium). These shifts are typically small, upfield
(
3) Olah, G. A.; Westerman, P. W.; Melby, E. G.; Mo, Y. K. J. Am.
(
1) (a) Olah, G. A. Halonium Ions; Wiley-Interscience: New York, 1975;
pp 113-116 and references therein. (b) Ruasse, M.-F. Acc. Chem. Res.
990, 23, 87. (c) Bellucci, G.; Bianchini, R.; Chiappe, C.; Brown, R. S.;
Slebolcak-Tilk, H. J. Am. Chem. Soc. 1991, 113, 8012.
2) (a) Olah, G. A.; Bollinger, J. M. J. Am. Chem. Soc. 1967, 89, 4744.
b) Olah, G. A.; Prakash, G. K. S.; Krishnamurthy, V. V. J. Org. Chem.
983, 48, 5516.
Chem. Soc. 1974, 96, 3565.
(4) (a) Saunders, M.; Telkowski, L.; Kates, M. R. J. Am. Chem. Soc.
1977, 99, 8070. (b) Siehl, H.-U. AdV. Phys. Org. Chem. 1987, 23, 63. (c)
Forsyth, D. A. Isotopes in Organic Chemistry; Buncel, E., Lee, C. C., Eds;
Elsevier: Amsterdam, The Netherlands, 1987; Vol. 6, Chapter 1. (d) Perrin,
C. L.; Ohta, B. K. Bioorg. Chem. 2002, 30, 3.
1
(
(
1
(5) Servis, K. L.; Domenick, R. L. J. Am. Chem. Soc. 1985, 107, 7186.
1
0.1021/ol070673n CCC: $37.00
© 2007 American Chemical Society
Published on Web 05/08/2007

(
negative), and attenuate as n increases.4c,6 If the ions are
closed (1a-d and 2a-d in Scheme 2), intrinsic isotope shifts
6
3
6 3
Scheme 2. Closed and Open Structure for 1-d and 2-d
Figure 1. The quaternary carbon region of the 13C NMR spectrum
of a 2:1 mixture of 2-d and 2-d at -60 °C.
3
0
on the quaternary carbons are expected to be small upfield
with a 10 mm probe. The samples were prepared in NMR
tube inserts (Wilmad) and acetone-d was placed in the outer
tube to act as both a lock solvent and chemical shift reference
(29.8 ppm).
Figure 1 shows the quaternary carbon region of the 13C
2
3
∆
o
and ∆
However if the ions are equilibria of open â-halocarbenium
ions (1b-d and 2b-d ), the deuteria impart an equilibrium
isotope shift (∆eq) in addition to ∆ . The ∆eq arises from
o
.
6
6
3
o
â-deuterium isotope effects on the carbenium center, a
spectrum for a 2:1 mixture of 2-d and 2-d . There are two
3
0
destabilizing effect that causes the equilibrium to favor the
isotopomer with deuteria distal to the carbenium center.
Thus, the time averaged 13C NMR signal for the quaternary
carbon distal to the deuteria should appear downfield while
the proximal carbon signal should appear upfield. The key
feature of equilibrating systems is the appearance of a large
signals for 2-d separated by 3.32 ppm. The lone signal for
3
7
2-d resides between the 2-d peaks and acts as an internal
0
3
reference to calculate ∆ . For the quaternary carbons of
obs
2-d , the ∆ are +1.42 and -1.90 ppm.
3
obs
3
The two ∆obs values of 2-d can be understood in the
framework of an isotopic perturbation of an equilibrium
involving open â-chlorocarbenium ions. The shifts are
consistent with small downfield intrinsic shifts in combina-
tion with large equilibrium shifts that move the signals away
from the 2-d signal. The downfield ∆ corresponds to the
∆
eq in which symmetrically related carbon signals are
displaced by equal magnitudes and in opposite directions.
In summary, isotopic perturbation of degenerate equilib-
rium predicts that the 13C NMR spectrum of halonium ions
will be qualitatively different from an equilibrium mixture
of â-halocarbenium ions. For example, the quaternary
0
obs
quaternary carbon distal to the deuteria and is shifted by
3
3( ∆ + ∆ ) ) +1.42 ppm. The upfield ∆ corresponds
o
eq
obs
2
carbons of 2a-d
3
should exhibit small upfield (negative) ∆
, the nonlabeled chemical shift reference:
o
to the carbon proximal to the deuteria and is shifted by 3( ∆
- ∆ ) ) -1.90 ppm. The absolute values of ∆ , ∆ , and
o
2
3
relative to 2a-d
0
eq
o
o
3
2
∆
obs ) 3( ∆
o
) and ∆obs ) 3( ∆
o
3
). For 2b-d , the ∆eq will
∆
cannot be calculated from the data; however, the sum
eq
2
3
dominate and the carbon proximal to the deuteria will appear
3( ∆ + ∆ ) ) -0.48 ppm is consistent with typical two-
and three-bond intrinsic shifts.
o
o
2
4c,6
far upfield of 2b-d
0
: ∆obs ) 3( ∆
o
- ∆eq). The signal of the
carbon distal to the deuteria will appear far downfield of
The ∆obs do not change between -60 and -80 °C.
Although others have pointed out that the absence of a
temperature dependence of ∆obs is inconsistent with the
proposal of an equilibrium, it is possible that the depen-
dence is small and could go undetected in the observed
3
2
b-d
upfield and downfield shifts for the quaternary carbons of
-d indicates the presence of ∆eq and supports open
0
: ∆obs ) 3( ∆ + ∆eq). Thus, the presence of the large
o
2
3
5,9
â-chlorocarbenium ions.
The deuterated precursors to halonium ions 1 and 2 were
6
temperature range. In the solvolysis of isopropyl-d halides
8
prepared by known procedures. The deuterium content was
and sulfonates, Leffek et al. observed a â-secondary duet-
erium isotope effect that did not change over a 30 °C
>
95% by NMR. Halonium ions were prepared under stable
2a
10
ion conditions (SbF
5
/SO
2
at -60 °C). Spectra of 1-d
to 2-d were obtained on a
Varian Mercury 300 NMR spectrometer (75 MHz C) fitted
6 3
, 2-d ,
temperature range. Assuming the equilibrium isotope effect
is due to the enthalpic contribution to the free energy, a lower
bound of ∆H for the temperature-dependent regime can be
determined. On the basis of estimates of chemical shifts in
2-d
0
, and a 2:1 mixture of 2-d
3
0
13
(
6) Hansen, P. E. Annu. Rep. NMR Spectrosc. 1983, 15, 106.
2
b and the limit of detection for shift changes in the NMR,
(7) DeFrees, D. J.; Taagepera, M.; Levi, B. A.; Pollack, S. K.;
the lower bound on the absolute value of ∆H is estimated
Summerhays, K. D.; Taft, R. W.; Wolfsberg, M.; Hehre, W. J. J. Am. Chem.
Soc. 1979, 101, 5532.
1
1
to be 6.7 cal/mol.
(8) Deuterium was incorporated into the precursor of 1-d6 by treating
acetone-d6 with isopropylmagnesium bromide. Subsequent dehydration and
bromine addition yielded the appropriately labeled precursor. Deuterium
was incorporated into the precursor of 2-d3 by base-catalyzed exchange of
pinacolone in D2O/MeOD. Treatment of pinacolone-d3 with PCl5 yielded
the appropriately labeled precursor.
(9) Forsyth, D. A.; Botkin, J. H.; Puckace, J. S.; Servis, K. L.; Domenick,
R. L. J. Am. Chem. Soc. 1987, 109, 7270.
(10) Leffek, K. T.; Robertson, R. E.; Sugamori, S. E. Can. J. Chem.
1961, 39, 1989.
2318
Org. Lett., Vol. 9, No. 12, 2007

Preliminary attempts to estimate ∆H by calculating the
ZPE’s of each isotopomer of 2b-d indicate that these
3
calculations are not trivial. In vacuo, geometry optimization
of an initially open structure yields a closed halonium ion.
Because a stationary point that represents an open structure
is a requirement for the frequency calculations, a discrete
solvent model was used to obtain an optimized structure for
2
2b. With the introduction of a single SO molecule, geometry
optimization of the 2b-SO complex at the MP2/6-31G* level
2
generates stationary points corresponding to open structures.
However, the frequency calculations are sensitive to the
solvent configuration in the 2b-SO complex. Our initial
2
estimates indicate that the calculation method is not precise
enough to determine whether ∆H is below the lower bound;
the estimated range of ∆H is -9.6 to 28.7 cal/mol.
Figure 2. The methyl carbon region of the 13C NMR spectrum of
1-d at -60 °C.
6
13
The C NMR spectra of 2-d
C indicate an interesting dynamic behavior. The line shapes
of the quaternary signals sharpen at low temperature. This
result indicates that the methyl-d group migrates between
3
taken between -60 and -80
°
6 3
occur in 1-d . As mentioned above, methyl shifts in 2-d
generate identical isotopomers and, in the slow exchange
regime, will not affect the analysis of isotope shifts. This is
3
quaternary carbons on the NMR time scale. The methyl
signals show similar behavior but because the chemical shift
difference between interchanging methyl groups is smaller,
the full coalescence behavior can be observed within the
temperature range. At -80 °C, two methyl signals appear
6
not the case for 1-d . Methyl shifts will generate a mixture
of isotopomers with geminal and vicinal orientations of the
methyl-d groups (Scheme 3). Additionally, the symmetric
3
displacement of deuteria for the isotopomer with vicinal
methyl-d groups causes the equilibrium to be degenerate
3
and result from different intrinsic shifts for the methyl-d
0
and therefore these isotopomers will exhibit ∆ only.
o
3
4
o o
groups that are geminal ( ∆ ) or vicinal ( ∆ ) to the methyl-
13
Figure 2 shows the methyl carbon region of the C NMR
d
3
group. At -60 °C, a single signal appears in the spectrum.
6
spectrum of 1-d . The presence of two signals is inconsistent
A likely mechanism that explains these data consists of a
with a single isotopomer containing geminal methyl-d3
series of 1,2-methyl shifts (Scheme 3). Similar 1,2-methyl
groups. Rather, the data are consistent with a mixture of
isotopomers with vicinal and geminal methyl-d
Thus, 1,2-methyl shifts occur in 1 and the sample of 1-d
should be considered a mixture when analyzing the quater-
3
groups.
6
6 3
Scheme 3. Methyl Shifts in 1-d and 2-d
1
3
nary carbon region of the C NMR.
The quaternary carbon region of the NMR spectrum of
1
6
-d is qualitatively similar to Figure 1 but with the central
13
signal nearly equidistant from the outer signals. The ∆obs
of the two downfield signals are consistent with those
5
previously reported by Servis, +1.50 and -0.30 ppm. The
additional signal appears upfield of the others (∆obs ) -2.11
ppm).
These signals can be understood in the framework of an
isotopic perturbation of equilibrium on a mixture of isoto-
pomers. The central signal is assigned to isotopomers with
shifts were observed by Olah et al. in the isomerization of
-butenebromonium ion at -40 °C12 and as a proposed
2
2
b
mechanism for equilibrating â-fluorocarbenium ions. The
,2-methyl shifts do not affect the analysis of the equilibrium
vicinal methyl-d
3
groups which show ∆ only. The ∆obs for
o
1
both quaternary carbons of this isotopomer are identical:
of â-chlorocarbenium ions because the ∆eq appears at low
temperatures when the methyl migration is slow. Addition-
2
3
3
( ∆
typical two- and three-bond intrinsic shifts. The outer signals
are assigned to the geminal methyl-d isotopomer and are
o
+ ∆ ) ) -0.30 ppm. This sum is consistent with
o
ally, methyl-d
3 3
shifts in 2-d generate identical asymmetri-
3
cally labeled isotopomers.
consistent with a combination of intrinsic and equilibrium
Considering our evidence for â-chlorocarbenium ions and
the presence of 1,2-methyl shifts, we reexamined the work
of Servis and Domenick to determine whether methyl shifts
2
3
isotope shifts: 6( ∆
)
o
- ∆eq) ) -2.11 ppm and 6( ∆
o
+ ∆eq
o
)
2
3
+1.50 ppm. The sum of intrinsic shifts, 6( ∆
o
+ ∆ ) )
-0.61 ppm, agrees with the value determined from the
vicinal methyl-d
3
isotopomer.
(11) The equilibrium isotope shift of 2b-d3 can be calculated from
equations found in ref 4b. Chemicals shifts of 2b were calculated by using
B3LYP/6-311G* GIAO calculations of the 2b-SO2 complex (vide infra).
The limit of detection for the difference in the observed isotope shifts was
estimated to be 0.1 ppm.
(13) Additional signals appear in the spectrum which result from the
loss of deuterium label. These signals can also be understood in the
framework of 1,2-methyl shifts and isotopic perturbation of equilibrium of
1-d5, -d4, -d3, -d2, and -d1. A signal for the nonlabeled isotopolog 1-d0 is
also present and was used as the internal chemical shift reference.
(12) Olah, G. A.; Bollinger, J. M.; Brinich, J. M. J Am. Chem. Soc. 1968,
9
0, 2587.
Org. Lett., Vol. 9, No. 12, 2007
2319

The presence of ∆eq in 2-d
3
and 1-d
6
is consistent with
Acknowledgment. We thank Dr. Timothy Dudley and
Dr. Walter Boyko for their time and expertise. We thank
the Department of Chemistry and the College of Liberal Arts
and Sciences at Villanova University for financial support
of this research. We are also grateful to the Merck Summer
Undergraduate Research Fellowship for providing financial
support for Jeffrey Schubert.
the proposal that these ions are best described as equilibria
of â-halocarbenium ions. To our knowledge, this is the first
direct evidence to support â-chlorocarbenium and â-bro-
mocarbenium ions as low-energy intermediates on the
potential energy surface. These results contradict the gener-
ally accepted notion that closed chloronium and bromonium
ion structures are solely responsible for stereospecific halogen
addition to alkenes. The presence of equilibrium shifts does
not rule out the existenece of halonium ions. However, they
do indicate that â-halocarbenium ions represent a significant
contribution to the overall structure and, if present, halonium
ions are in equilibrium with â-halocarbenium ions. Future
experimental and theoretical investigations will determine
if our conclusions are general and a more detailed report on
these and other halonium ions is being prepared.
Supporting Information Available: NMR spectra of 2-d
at -50 to -80 °C and the quaternary carbon region of 1-d
3
6
.
This material is available free of charge via the Internet at
http://pubs.acs.org.
OL070673N
2320
Org. Lett., Vol. 9, No. 12, 2007