Rearrangement of dimethylcarbene to propene: Study by laser flash photolysis and ab initio molecular orbital theory

Article abstract of DOI:10.1021/ja9724598

Laser flash photolysis (Nd:YAG laser, 355 nm, 35 mJ, 150 ps) of dimethyldiazirine and dimethyldiazirine-d₆ produces dimethylcarbene (DMC) and dimethylcarbene-d₆ (DMC-d₆), respectively. The carbenes were trapped with pyridine to form ylides which absorb around 364 nm. It was possible to resolve the growth of the ylides as a function of pyridine concentration in Freon-113, α, α,α-trifluoromethylbenzene, and perfluorohexane as a function of temperature. The observed rate constant (k(obs)) of ylide formation was linearly dependent on the concentration of pyridine in all solvents and at all temperatures. From plots of kobs versus [pyridine] it was possible to extract values of k(pyr) (the absolute rate constant of reaction of the carbene with pyridine) and τ, the carbene lifetime in the absence of pyridine, and their associated Arrhenius parameters. In Freon-113 and α, α, α-trifluoromethylbenzene the carbenes decay both by rearrangement and by reaction with solvent. In perfluorohexane the carbene decay appears to be predominantly unimolecular. The experimental results are compared with ab initio molecular orbital calculations. The experimentally determined barrier to disappearance of DMC in perfluorohexane (2.56 ± 0.05 kcal/mol) is much smaller than that calculated (7.4 ± 2 kcal/mol) using ab initio molecular orbital theory. The Arrhenius parameters and isotope effects indicate that the rearrangement of DMC in perfluorohexane has a large component of quantum mechanical tunneling. The activation energy for the disappearance of DMC-d₆ in perfluorohexane (5.63 ± 0.03 kcal/mol) is consistent with calculations which indicate that QMT makes only a minor contribution to the deuterated system under the conditions of this study.

Full text of DOI:10.1021/ja9724598

4430
J. Am. Chem. Soc. 1998, 120, 4430-4438
Rearrangement of Dimethylcarbene to Propene: Study by Laser Flash
Photolysis and ab Initio Molecular Orbital Theory
Francis Ford,† Tetsuro Yuzawa,† Matthew S. Platz,*,† Stephan Matzinger,*,‡,1 and
Markus Fu1lscher*,§
Contribution from the Newman and Wolfrom Laboratory of Chemistry, The Ohio State UniVersity, 100 W.
18th AVenue, Columbus, Ohio 43210-1173, Institute for Physical Chemistry, UniVersity of Fribourg,
Pe´rolles, CH-1700 Fribourg, Switzerland, and Department of Theoretical Chemistry, Chemical Centre,
UniVersity of Lund, P.O. Box 124, 2-22100 Lund, Sweden
ReceiVed July 21, 1997. ReVised Manuscript ReceiVed March 10, 1998
Abstract: Laser flash photolysis (Nd:YAG laser, 355 nm, 35 mJ, 150 ps) of dimethyldiazirine and
dimethyldiazirine-d6 produces dimethylcarbene (DMC) and dimethylcarbene-d6 (DMC-d6), respectively. The
carbenes were trapped with pyridine to form ylides which absorb around 364 nm. It was possible to resolve
the growth of the ylides as a function of pyridine concentration in Freon-113, R,R,R-trifluoromethylbenzene,
and perfluorohexane as a function of temperature. The observed rate constant (kobs) of ylide formation was
linearly dependent on the concentration of pyridine in all solvents and at all temperatures. From plots of kobs
versus [pyridine] it was possible to extract values of kpyr (the absolute rate constant of reaction of the carbene
with pyridine) and τ, the carbene lifetime in the absence of pyridine, and their associated Arrhenius parameters.
In Freon-113 and R,R,R-trifluoromethylbenzene the carbenes decay both by rearrangement and by reaction
with solvent. In perfluorohexane the carbene decay appears to be predominantly unimolecular. The experimental
results are compared with ab initio molecular orbital calculations. The experimentally determined barrier to
disappearance of DMC in perfluorohexane (2.56 ( 0.05 kcal/mol) is much smaller than that calculated (7.4
( 2 kcal/mol) using ab initio molecular orbital theory. The Arrhenius parameters and isotope effects indicate
that the rearrangement of DMC in perfluorohexane has a large component of quantum mechanical tunneling.
The activation energy for the disappearance of DMC-d6 in perfluorohexane (5.63 ( 0.03 kcal/mol) is consistent
with calculations which indicate that QMT makes only a minor contribution to the deuterated system under
the conditions of this study.
I. Introduction
predicted to have a triplet ground state, but with a much smaller
singlet-triplet gap.9 Dimethylcarbene is predicted to have a
singlet ground state.10
Triplet carbenes are often compared to free radicals and
singlet carbenes are often compared to carbocations in terms
of their electronic structure and reactivity.8 The effects of
alkylation on singlet-triplet gaps are easily understood as a
consequence of the fact that alkyl groups afford more stabiliza-
tion to cations than to radicals.11
Methylene (CH2) is the simplest carbene, and methylcarbene
(MC) and dimethylcarbene (DMC) are the prototypical alkyl-
and dialkylcarbenes. Methylene has been studied extensively
by computational and experimental methods.2 Experiment and
theory are in good agreement as to the geometries of the singlet
and triplet states of CH2 and their energy separation.2,3 Among
the experimental methods which have been applied to the study
of methylene are gas-phase spectroscopy,2,4 solution-phase
CIDNP5 and nanosecond (ns) spectroscopy,6 low-temperature
matrix electron paramagnetic resonance (EPR) spectroscopy,7
and chemical trapping studies in the gas phase and in solution.8
Theory predicts that the triplet state of methylene is 9 kcal/
mol more stable than the singlet state.2,3 Methylcarbene is also
(6) (a) Turro, N. J.; Cha, Y.; Gould, I. R.; Padwa, A.; Gasdaska, J. R.;
Tomas, M. J. Org. Chem. 1985, 50, 4417. (b) Padwa, A.; Rosenthal, R. J.;
Dont, W.; Filho, P.; Turro, N. J. Hrovat, D. A. Gould, I. R. J. Org. Chem.
1984, 49, 3174.(c) Padwa, A.; Gasdaska, J. R.; Tomas, M.; Turro, N. J.;
Cha, Y.; Gould, I. R. J. Am. Chem. Soc. 1986, 108, 6739. (d) Turro, N. J.;
Cha, Y.; Gould, I. R. J. Am. Chem. Soc. 1987, 109, 2101.
(7) (a) Wasserman, E.; Kuck, V. J.; Hutton, R. S.; Anderson, E. D.;
Yager, W. A. J. Chem. Phys. 1971, 54, 4120. (b) Wasserman, E.; Hutton,
R. S. Acc. Chem. Res. 1977, 10, 27. (c) Trozzolo, A. M.; Wasserman, E. In
Carbenes; Moss, R. A., Jones, M., Jr., Eds.; Wiley: New York, 1975; Vol.
2, p 185.
(8) For reviews see: (a) Hine, J. DiValent Carbon; Ronald Press: New
York, 1964. (b) Bethell, D. AdV. Phys. Org. Chem. 1969, 7, 153. (c) Kirmse,
W. Carbene Chemistry, 2nd ed.; Academic Press: New York, 1971. (d)
Carbenes; Moss, R. A., Jones, M., Jr., Eds.; Wiley: New York, 1973; 1975;
Vol. I, II.
(9) (a) Schaefer, H. F., III; Ma, B. J. Am. Chem. Soc. 1994, 116, 3539.
(b) Evanseck, J D.; Houk, K. N. J. Phys. Chem. 1990, 94, 555. (c) Gallo,
M. M.; Schaefer, H. F., III. J. Phys. Chem. 1992, 96, 1515. (d) Khodabandeh,
S.; Carter, E. A. J. Phys. Chem. 1993, 97, 4360.
(10) (a) Matzinger, S.; Fu¨lscher, M. P. J. Phys. Chem. 1995, 99, 10747.
(b) Richards, C. A., Jr.; Kim, S.-J.; Yamaguchi, Y.; Schaefer, H. F., III. J.
Am. Chem. Soc. 1995, 117, 10104.
† The Ohio State University.
‡ University of Fribourg.
§ University of Lund.
(1) Present address: Department of Chemistry, Stanford University,
Stanford, CA 94305.
(2) Shavitt, I. Tetrahedron 1985, 41, 1531 and references therein.
(3) Leopold, D. G.; Murray, K. K.; Miller, A. E. S.; Lineberger, W. C.
J. Chem. Phys. 1984, 83, 4849.
(4) (a) Herzberg, G.; Shoosmith, J. Nature 1959, 183, 189.(b) Herzberg,
G. Proc. R. Soc. London 1961, A262, 291. (c) Herzberg, G.; Johns, J. W.
C. Proc. R. Soc. London 1966, A295, 107. (d) Herzberg, G.; Johns, J. W.
C. J. Chem. Phys. 1971, 54, 2276. (e) McKellar, A. R. W.; Bunker, P. R.;
Sears, T. J.; Evenson, K. M.; Saykally, R. J. Langhoff, S. R. J. Chem. Phys.
1983, 79, 5251 and references therein.
(5) (a) Roth, H. D. J. Am. Chem. Soc. 1971, 93, 1527. (b) Roth, H. D.
Acc. Chem. Res. 1977, 10, 85.
S0002-7863(97)02459-1 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/28/1998

Rearrangement of Dimethylcarbene to Propene
J. Am. Chem. Soc., Vol. 120, No. 18, 1998 4431
Experimental evidence for the existence of MC and DMC is
quite meager. There is as yet no direct observation of MC or
DMC in the gas phase or in cryogenic matrixes. Kramer and
Wright were able to trap MC in solution with phenylsilane, but
only in a very low yield (5%).12
Scheme 1
McMahon and Seburg were able to trap methylcarbene in a
cryogenic matrix with carbon monoxide to form a ketene, which
was readily characterized by IR spectroscopy.13
The paucity of experimental data concerning DMC has
historically been attributed to the expected ease of its rear-
rangement to propene (P)
hydrocarbon solvent was found to be limited, at least in part,
by rearrangement to propene because the lifetime of DMC-d6
was 3.2 times larger than that of DMC itself in a hydrocarbon
solvent at ambient temperature.15
Herein we are pleased to report new experimental studies of
DMC which allow the absolute measurement of both kpyr and τ
of DMC and DMC-d6 as a function of temperature in CF2-
ClCFCl2 (Freon-113), perfluorohexane, and R,R,R-trifluoro-
methylbenzene. These studies will demonstrate that the lifetime
of DMC is controlled by a mixture of rearrangement and
reaction with solvent in the case of Freon-113 and R,R,R-
trifluoromethylbenzene. The data in perfluorohexane are con-
sistent with unimolecular decay of the carbene. The rearrange-
ment of DMC is likely to proceed, at least in part, via quantum
mechanical tunneling (QMT) in perfluorohexane because of the
large isotope effects on Arrhenius parameters and because the
observed barrier to rearrangement of DMC is much smaller than
that predicted by calculations. The barrier to disappearance of
DMC-d6 in perfluorohexane is consistent with the calculated
barrier. Thus, any contribution of QMT to the kinetics of DMC-
d6 must be relatively minor.
a process which should be exothermic (on the basis of simple
bond additivity arguments)14 by more than 60 kcal/mol.
Nevertheless, we have previously trapped DMC with pyridine,
in solution, and used laser flash photolysis methods to deduce
that DMC has a lifetime of 20 ns in pentane at ambient
temperature.15 In this experiment (Scheme 1) dimethyldiazirine
(DMD) was irradiated with a pulse of 351 nm light from a XeF
excimer laser. Decomposition of the excited state of DMD
(1DMD*) leads to the formation of DMC which can, in principle,
react with solvent (RX, X ) H, hydrocarbons; X ) Cl, Freons),
rearrange (kR), or react with pyridine (kpyr) to form a UV-vis
active ylide (Y, Scheme 1).13
The spectrometer in use in that work had a time resolution
of ∼20 ns which prevented the direct measurement of the growth
of the ylide following the laser pulse.15 This limitation did not
permit the direct measurement of kpyr and the carbene lifetime
τ (τ ) 1/(kR + kRX[RX])). Instead, the lifetime was deduced
by measuring the yield of ylide (AY) as a function of pyridine
concentration. As expected15 plots of 1/AY versus 1/[pyr] were
linear. Dividing the intercept by the slope of this plot gave
kpyrτ, where τ is the lifetime of DMC in the absence of pyridine.
After assuming that DMC had a singlet ground state (which
was subsequently justified by theory)10 and that kpyr was 1 ×
109 M-1 s-1, a value of τ ) 20 ns in alkane solvent at ambient
temperature was deduced.15 The lifetime of DMC in the
II. Results
II.1. Laser Flash Photolysis Studies. The LFP system used
in this laboratory has been upgraded by incorporation of a
frequency-tripled YAG laser which delivers 150 ps pulses
(pretripling) at 355 nm with 30 mJ of energy, improved detection
circuitry, and a monochromator.16 The time resolution of the
spectrometer was determined by measuring the apparent growth
of triplet benzophenone, following a 355 nm laser pulse, at 525
nm. The apparent rise of the triplet benzophenone absorption
was <1 ns.
LFP of DMD in Freon-113 fails to produce a detectable UV-
vis active transient between 300 and 700 nm. Similar results
are obtained upon LFP of pyridine in Freon-113. LFP (355
nm, 150 ps, 30 mJ, Nd:YAG) (Figure 1) of DMD and pyridine
produces the same transient spectrum of ylide Y (Figure 1)
observed previously with excimer laser pulses (351 nm, 55 mJ,
17 ns).15 With the 150 ps pulsed laser it was possible to resolve
the growth of the ylide (Figure 2). The growth is adequately
(11) Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin,
R. D.; Mallard, W. G. J. Phys. Chem. Ref. Data 1988, 17, Suppl. 1.
(12) Kramer, K. A. W.; Wright, A. N. Tetrahedron Lett. 1962, 1095.
(13) Seburg, R. A.; McMahon, R. J. J. Am. Chem. Soc. 1992, 114, 7183.
(14) (a) Benson, S. W. Thermochemical Kinetics; Wiley: New York,
1965. (b) Griller, D.; Kanabus-Kaminska, J. M. In Handbook of Organic
Photochemistry; Scaiano, J. C., Ed.; CRC Press: Boca Raton, FL, 1989;
Vol. II, p 359.
(15) (a) Modarelli, D. A.; Platz, M. S. J. Am. Chem. Soc. 1991, 113,
8985. (b) Modarelli, D. A.; Platz, M. S. J. Am. Chem. Soc. 1993, 115, 470.
(c) Modarelli, D. A.; Morgan, S.; Platz, M. S. J. Am. Chem. Soc. 1992,
114, 7034. (d) Platz, M. S.; Modarelli, D. A.; Morgan, S.; White, W. R.;
Mullins, M.; Celebi, S.; Toscano, J. P. Prog. React. Kinet. 1994, 19, 93.
(e) Ge, C.-S.; Jang, E. G.; Jefferson, E.; Liu, W.; Moss, R. A.; Wlostowska,
J.; Xue, S. J. Chem Soc., Chem. Commun. 1994, 1479.
(16) Gritsan, N. P.; Zhai, H. B.; Yuzawa, T.; Karweik, D.; Brooke, J.;
Platz, M. S. J. Phys. Chem. A 1997, 101, 2833.

4432 J. Am. Chem. Soc., Vol. 120, No. 18, 1998
Ford et al.
Figure 1. Transient spectra of ylide Y produced by LFP of DMD in
Freon-113 at ambient temperature.
Figure 3. kobs versus pyridine concentration obtained by LFP of DMD
in R,R,R-trifluoromethylbenzene as a function of temperature.
Figure 2. Formation of ylide (Y) monitored at 370 nm in R,R,R-
trifluoromethylbenzene in the presence of 0.03 M pyridine at 243 K.
Figure 4. kobs versus pyridine concentration obtained by LFP of DMD-
d6 in R,R,R-trifluoromethylbenzene as a function of temperature.
described by an exponential function with observed rate constant
(kobs), which is related to the absolute rate constants of Scheme
1 by eq 1.17
to obtain absolute kinetic data. These solutions were then
continuously irradiated with 350 nm radiation from a Ray-O-
Net reactor. The products of photolysis were then characterized
by gas chromatography-mass spectrometry to establish the
importance of the kRX[RX] term to kT in the solvents of interest.
Photolysis of DMD in Freon-113 led to the formation of a
mixture of products. One reaction product has a molecular ion
and fragmentation pattern consistent with that of 1, formed by
formal C-Cl bond insertion. This product likely derives from
reaction of DMC with solvent via a radical pair intermediate.
kobs ) kR + kRX[RX] + kpyr[pyr]
(1)
As predicted by eq 1, a plot of kobs versus pyridine
concentration is linear (Figures 3-5) with slope kpyr and
intercept (kR + kRX[RX] ) 1/τ). Results were obtained with
DMD and DMD-d6 over a temperature range of 243-273 K
and in R,R,R-trifluoromethylbenzene and perfluorohexane sol-
vent. Arrhenius treatment of the 1/τ data is presented in Figures
6-8 and summarized in Table 1. The values of kpyr show little
systematic variation with temperature, consistent with Ea ≈ 0
for this process, as observed with other carbenes.18
II.2. Reactions of Dimethylcarbene with Solvent. Di-
methyldiazirine (DMD) was condensed into the solvents used
(17) Scaiano, J. C. Acc. Chem. Res. 1982, 15, 252.
(18) (a) Jackson, J. E.; Soundararajan, N.; Platz, M. S. Tetrahedron Lett.
1989, 30, 1335. (b) Moss, R. A.; Turro, N. J. In Kinetics and Spectroscopy
of Carbenes and Biradicals; Platz, M. S., Ed.; Plenum: New York, 1990;
p 213.
Products with molecular ions and fragmentation patterns
consistent with the formation of DMC-solvent adducts (e.g.,
2) were detected upon photolysis of DMD in R,R,R-trifluoro-
methylbenzene. We cannot rule out the possibility that DMC

Rearrangement of Dimethylcarbene to Propene
J. Am. Chem. Soc., Vol. 120, No. 18, 1998 4433
Table 1. Activation Parameters of the Reactions of DMC and DMC-d6 Determined from the Intercepts (kR + kRX[RX]) of Plots of kobs versus
Pyridine Concentration
DMC
DMC-D6
solvent
Ea (kcal/mol)
∆Sq (298 K) (cal mol-1 K-1
)
log A
Ea (kcal/mol)
∆Sq (298 K) (cal mol-1 K-1
)
log A
Cl2CFCClF2
C6H5CF3
C6F14
3.11 ( 0.05
2.90 ( 0.05
2.56 ( 0.05
-9.6 ( 0.7
-11.0 ( 0.7
-12.8 ( 0.7
10.7 ( 0.3
10.4 ( 0.3
10.0 ( 0.3
3.36 ( 0.05
3.85 ( 0.05
5.63 ( 0.03
-9.6 ( 0.7
-8.2 ( 0.7
-2.7 ( 0.7
10.7 ( 0.3
11.0 ( 0.3
12.2 ( 0.3
Figure 7. Arrhenius plots of the 1/τ data obtained in R,R,R-
trifluoromethylbenzene.
Figure 5. kobs versus pyridine concentration obtained by LFP of DMD
and DMD-d6 in Freon-113 at 278 K.
Figure 6. Arrhenius plots of the 1/τ data obtained in Freon-113.
Figure 8. Arrhenius plots of the 1/τ data obtained in perfluorohexane.
and fragmentation patterns are available as Supporting Informa-
tion. Suffice it to say that the kRX[RX] term for the disappear-
ance of DMC in Freon-113 and R,R,R-trifluoromethylbenzene
is not trivial.
Finally, no adducts derived from reaction with DMC with
solvent were observed upon photolysis of DMD in perfluoro-
hexane. There is no evidence that the kRX[RX] term is
significant in perfluorohexane, and it will be neglected.
forms a transient π complex with this aromatic solvent as
postulated by Goodman,20 Moss,21 and Jones.22 The GC traces
(20) Khan, M. I.; Goodman, J. L. J. Am. Chem. Soc. 1993, 117, 6635.
(21) Moss, R. A.; Yan, S.; Krogh-Jesperson, K. J. Am. Chem. Soc. 1998,
120, 1088.
(22) Jones, M., Jr.; Thamattoor, D. M.; Ruck, R. T. Kyushu International
Symposium on Physical Organic Chemistry, Kyushu, Japan, Dec 3, 1997.
(19) (a) Barcus, R. L.; Hadel, L. M.; Johnston, L. J.; Platz, M. S.; Savino,
T. G.; Scaiano, J. C. J. Am. Chem. Soc. 1986, 108, 3928. (b) Platz, M. S.;
Maloney, V. M. In Kinetics and Spectroscopy of Carbenes and Biradicals;
Platz, M. S., Ed.; Plenum: New York, 1990; p 239.

4434 J. Am. Chem. Soc., Vol. 120, No. 18, 1998
Ford et al.
Azine (A) was formed on photolysis of DMD in all solvents.
Azine can be formed by several mechanisms including the
reaction of a carbene with its diazirine precursor. This reaction
is very fast and complicates analysis of the kinetics of
benzylchlorocarbene.23
It seems unlikely that azine is formed by reaction of DMD
with DMC because the lifetime of DMC is very short and does
not change upon doubling the concentration of precursor. We
believe that DMC rearranges to its isomeric diazo compound
and that this species forms azine by a non-carbenic mechanism.
Figure 9. Calculated structures of the reactants, transition states, and
products of the reactions dimethylcarbene f propene (A), (cyano-
methyl)methylcarbene f 1-cyanopropene (B), and ethylmethylcarbene
f 2-butene (C).
the nonbonded electrons localized at the carbenoid center. For
propene the active space includes the valence π space.
All multireference calculations have been performed using
the MOLCAS-3 program systems.33 Gaussian 9434 was used
to optimize the geometries, and the CCSD(T) calculations were
performed with MOLPRO96.35
III. Theory
III.1. Results. III.1.1. Geometries. Recently, some of us
(S.M., M.F.) have reported the singlet-triplet energy separation
of methylated carbenes.10a This was the starting point of the
present investigation. As the rearrangement reaction begins,
one hydrogen atom moves in a plane that is almost perpendicular
to the plane formed by the three carbon atoms, while the
remaining two hydrogen atoms on the same methyl group swing
very rapidly into a nearly planar position. Moreover, as the
π-bond is formed, we observe a contraction of the C-C bond
distance. In contrast, the “spectator” methyl group behaves
almost perfectly like a rigid body. It rotates and bends by
approximately 16° and 7°, respectively, such that the methyl
group is placed in a conformation close to that in propene. At
the transition state, the lengths of the breaking and forming C-H
bonds are 125.1 and 132.6 pm, respectively. The length of the
preformed C-C double bond is 140.2 pm. Focusing on the
second half of the reaction, the moving hydrogen atom rotates
into the plane of the carbon atoms to form the new C-H bond.
This movement is followed by a further shortening of the
forming C-C double bond and additional, rather small adjust-
ments of the positions of the other atoms. The structures of
the reactants, transitions states, and products are shown in Figure
9.
To complement the experimental work we studied the 1,2-
H-shift reaction of dimethylcarbene, as well as the cyano- and
methyl-substituted analogues, by ab initio quantum chemical
methods. To optimize the geometries of the reactants, products,
and transition states, we used the second-order Møller-Plesset
(MP2) method24 in combination with analytic derivative methods
and 6-31G(2p,d) basis sets.25,26 Barrier heights and exother-
micities were then determined by single-point calculations using
the CCSD(T) method27,28 and atomic natural orbital (ANO) type
basis sets29 contracted to (4s3p2d) and (3s2p) functions for
carbon, nitrogen, and hydrogen, respectively.
To validate our approach, we performed calculations using
the complete active space (CAS)SCF method30 followed by
second-order multireference configuration interaction (SD-
MRCI),31 including Davidson’s correction,32 and multireference
second-order perturbation calculations, the so-called CASPT2
approach.33 Due to the strong mixing of the orbitals describing
the C-H and C-C bonds, the choice of an active space for the
CASSCF method was particularly difficult, and it was necessary
to use the smallest meaningful active space, two electrons in
two orbitals. Thus, the CASSCF wave function of singlet
dimethylcarbene as well as that of the transition state structure
included the ...(σ)2(π)0 and ...(σ)0(π)2 configurations correlating
To optimize the geometries of the substituted systems, i.e.,
(cyanomethyl)methylcarbene and methylethylcarbene, the struc-
tures of dimethylcarbene, the transition state, and propene were
used as templates to which the methyl or cyano group was
attached. These structures were then used as starting points
for geometry optimizations. The structures of all possible
rotamers were optimized, but only the structures of lowest
(23) Merrer, D. C.; Moss, R. A.; Liu, M. T. H.; Banks, J. T.; Ingold, K.
U. J. Org. Chem., submitted for publication.
(24) Møller, C.; Plesset, M. S. Phys. ReV. 1934, 46, 618.
(25) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28, 213.
(26) Dill, J. D.; Pople, J. A. J. Chem. Phys. 1975, 62, 2921.
(27) Raghavachari, K.; Trucks, G. W.; Polpe, J. A.; Head-Gordon, M.
Chem. Phys. Lett. 1989, 157, 479.
(28) Jayatilaka, D.; Lee, T. J. Chem. Phys. Lett. 1992, 199, 211.
(29) Widmark, P.-O.; Malmqvist, P.-A.; Roos, B. O. Theor. Chim. Acta
1990, 77, 291.
(30) Roos, B. O. In Lecture Notes in Quantum Chemistry; Roos, B. O.,
Ed.; Springer-Verlag: Berlin, 1992.
(31) Siegbahn, P. E. M. J. Chem. Phys. 1980, 72, 1647.
(32) Langhoff, S. R.; Davidson, E. R. Int. J. Quantum Chem. 1974, 8,
61.
(33) Andersson, K.; Roos, B. O. In Modern Electron Structure Theory;
Yarkony, R., Ed.; World Scientific Publishing: New York, 1994; Vol. 1.
Andersson, K.; Fu¨lscher, M. P.; Karlstro¨m, G.; Lindh, R.; Malmqvist, P-A° .;
Olsen, J.; Roos, B. O.; Sadlej, A. J.; Blomberg, M. R. A.; Siegbahn, P. E.
M.; Kello¨, V.; Noga, J.; Urban, M.; Widmark, P.-O. MOLCAS Version 3,
Department of Theoretical Chemistry, Chemical Centre, University of Lund,
P.O. Box 124, S-22100 Lund, Sweden, 1994.
(34) Gaussian 94, Revision B.1: 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.; E. S. Replogle,
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,
Inc., Pittsburgh, PA, 1995.
(35) MOLPRO is a package of ab initio programs written by H.-J. Werner
and P. J. Knowles, with contributions from J. Almlof, R. D. Amos, M. J.
O. Deegan, A. T. Elbert, C. Hampel, W. Meyer, K. Peterson, R. Pitzer, A.
J. Stone, P. R. Taylor, and R. Lindh, Universitiy of Birmingham, 1996.

Rearrangement of Dimethylcarbene to Propene
J. Am. Chem. Soc., Vol. 120, No. 18, 1998 4435
Table 4. Mulliken’s Population Analysis
Table 2. Bond Distances (pm) and Bond Angles (deg)a
b
b
c
d
e
f
g
h
i
j
C3a
C1a
H32a
Dimethylcarbene
0.25 -0.50
Transition State Structure
∑
H31 + H33c
Md
Reactants
Al 148.0 148.0 110.3 98.0 85.5 110.6
Bm 147.5 149.4 109.2 97.1 83.6 110.6
Cn 148.2 147.5 110.6 98.6 83.5 111.2
98.0 85.5
98.5 93.3
94.4 91.4
-0.87
-0.53
-0.50
0.12
0.56
-0.06
-0.03
-0.07
-0.02
-0.02
0.10
-0.45
0.48
0.40
Transition Structure
Propene
-0.33
A
B
C
150.9 140.2 112.3 105.1 69.2 125.1 132.6 59.6 77.7
150.4 140.9 111.5 104.4 70.4 125.2 134.1 60.2 78.8
151.1 140.8 112.1 104.9 70.6 124.1 134.0 60.4 82.6
0.19
a Cf. footnote a in Table 2. b Sum of Mulliken charges at the centers
of C3, C1, and H32. c Sum of Mulliken charges for the hydrogen atoms
attached to C3 but not taking active part in the 1,2-H-shift reaction.
d Sum of Mulliken charges for the spectator methyl group.
Products
150.1 133.8 124.4 111.1 59.4
149.6 134.5 125.3 110.4 59.0
150.3 134.2 127.4 110.6 59.1
A
B
C
Mulliken charges are known to be strongly biased by the choice
of basis sets. Nevertheless, if we analyze only relatiVe charge
differences during the course of the reaction, meaningful insight
can be gained. Table 4 lists the partial charges at the carbenoid
center, the hydrogen-donating carbon atom, the moving hydro-
gen, the “spectator” methyl group, and the remaining two
hydrogen atoms.
The carbenoid C-atom has a partial positive charge in
dimethylcarbene. During the hydrogen atom migration process
negative charge flows from the hydrogen-donating carbon atom
to the carbenoid center and the migrating proton. At the
transition state the charge transfer to the carbenoid center is
already completed. For the second half of the reaction a small
back flow of positive charge occurs predominantly from the
migrating hydrogen to the terminal, double-bonded carbon atom
in propene. Charge development in the transition state is
consistent with our finding that the lifetime of DMC is
considerably shortened in polar solvents relative to nonpolar
solvents.15
During the 1,2-H-shift reaction in dimethylcarbene a propene-
like structure is achieved at an early stage while the migrating
hydrogen atom remains in a position favoring the hyperconju-
gative interaction. The electron-withdrawing cyano group
destabilizes the positive charge developing on the carbon R to
the carbene center and reduces the amount of negative charge
which flows to the carbene center. Thus, the charge population
of the carbenoid center is reduced in (cyanomethyl)methylcar-
bene, leading to an increase of the reaction barrier, in accord
with the principle of maximum overlap. In contrast, attaching
a weak electron donor, such as a methyl group, to the hydrogen-
donating carbon lowers the energy barrier of the reaction because
the substituent cooperates with the charge transfer to the
carbenoid center.
III.2. Validation, Error Estimates. In general, one expects
the MP2 method to fail to properly describe bond formation.
To validate the present approach, we also determined MCSCF
wave functions for the 1,2-H-shift reaction of dimethylcarbene
using the CASSCF method which was complemented by SD-
MRCI and CASPT2 calculations. Table 5 summarizes the
structures of the CASSCF and SD-MRCI wave functions.
a For convenience, we label the carbenoid center C1 and the carbon
atom from which the hydrogen atom is removed C3. C2 denotes the
carbon atom of the spectator group. The migrating hydrogen atom is
labeled H32. Finally, H22 denotes the hydrogen atom attached to C2
and oriented almost perpendicular to the central plane. Because the
cyano and methyl substituents can in good approximation be described
as rigid bodies, we did not tabulate all geometrical parameters. The
full set of coordinates can be obtained from the authors upon request.
b Bond distance C1-C2. c Bond distance C1-C3. d Bond angle C2-
C1-C3. e Bond angle C1-C2-H22. f Dihedral angle C3-C1-C2-
i
H22. g Bond distance C3-H32. h Bond distance C1-H32. Bond angle
C1-C3-H32. k Dihedral angle C2-C1-C3-H32. l Reaction dimeth-
ylcarbene f propene. m Reaction (cyanomethyl)methylcarbene f cy-
anopropene. n Reaction ethylmethylcarbene f butene.
Table 3. Barrier Heights and Exothermicities (kcal/mol)
system
∆Ea
ZVPEb
∆Ec
ZPVEb
Ad
Be
Cf
8.5
10.1
5.5
-1.1
-1.3
-1.2
-68.6
-74.6
-68.9
2.4
2.4
2.3
a Barrier height. b Zero-point vibrational energy correction. c Exo-
thermicity. d Reaction dimethylcarbene f propene. e Reaction (cya-
nomethyl)methylcarbene f cyanopropene. f Reaction ethylmethylcar-
bene f butene.
energy are reported because they were used to determine the
energetics of the reaction.
Table 2 summarizes the bond distances and angles, focusing
on the region of interest. We find that the effect of the cyano
and methyl substituents on the geometry of the central region
of the reactants, transition states, and products is rather small.
III.2.2. Reaction Barrier Heights and Exothermicities.
The computed barrier heights and exothermicities of the
rearrangement reactions are collected in Table 3 which also
includes the zero-point vibrational energy corrections (ZVPE).
The latter have been computed at the MP2 level of theory using
the harmonic approximation.
Our best theoretical estimate for the barrier height in the 1,2-
H-shift reaction in dimethylcarbene is 7.4 kcal/mol obtained at
the CCSD(T)/ANO+ZPVE level of theory. The overall exo-
thermicity for the rearrangement of dimethylcarbene to propene
was calculated to be 66.2 kcal/mol. For (cyanomethyl)-
methylcarbene the barrier height of the reaction and the overall
exothermicity increase by 1.4 and 6.0 kcal/mol, respectively.
In contrast, the effect of substituting a hydrogen atom by a
methyl group on the energetics is opposite in sign; i.e., the
reaction barrier is lowered by 3.1 kcal/mol, while the exother-
micity increases by only 0.2 kcal/mol. This is consistent with
our earlier report that the lifetime of ethylmethylcarbene is about
10 times shorter than that of DMC, under comparable condi-
tions.15
By inspection we found that all CASSCF wave functions were
dominated by the Hartree-Fock determinant. The second most
important contribution for dimethylcarbene and the transition
state is the ...(σ)0(π)2 doubly excited configuration. For propene
the ...(π)0(π*)2 excited configuration contributes with a weight
of 4-5% to the total wave function. All remaining CI
coefficients were small. Table 5 also includes the weight of
the CASSCF reference function in the SD-MRCI and CASPT2
wave functions. They were very similar for both methods and
all species. Finally, the last column of Table 5 reports the value
of the so-called T1-diagnostic for the CCSD(T) wave function.
To obtain a qualitative understanding of the substitution
effects, we computed Mulliken populations for the rearrange-
ment of dimethylcarbene to propene at the MRCI level.

4436 J. Am. Chem. Soc., Vol. 120, No. 18, 1998
Table 5. Structure of the Wave Functionsa
Ford et al.
and Houk,38 in agreement with the recently reported value of
5.2 kcal/mol by Sulzbach et al.39
wMRCIb wCASPT2c
T1
d
CASSCF
MRCI
Dimethylcarbene
IV. Discussion
(σ)2(π)0 96.8%
(σ)0(π)2 3.2%
(σ)2(π)0 85.5%
(σ)0(π)2 2.3%
87.9
87.2
86.2
87.3
0.0133
0.0144
0.0104
It is clear from Table 1 that the experimentally determined
activation energies for the disappearance of DMC are much
smaller than the value predicted by theory for the 1,2-hydrogen
migration reaction of this carbene.
Transition State Structure
(σ)2(π)0 98.1%
(σ)0(π)2 1.9%
(σ)2(π)0 86.2%
(σ)0(π)2 1.2%
87.5
87.9
Propene
In Freon-113 the activation energies observed for the disap-
pearance of DMC and DMC-d6 are the same within experimental
error (3.3-3.4 kcal/mol; see Table 1). The lack of any isotope
effect is consistent with the GC-MS study which demonstrated
that DMC reacts with Freon-113. There is no conflict with
theory with the barrier measured in Freon-113, as DMC does
not disappear in this solvent by a unimolecular reaction.
In R,R,R-trifluoromethylbenzene solvent, an isotope effect
of 0.6 kcal/mol on the barrier to the disappearance of DMC is
observed. This demonstrates that the lifetime of DMC is
controlled, in part, by rearrangement to propylene. Nevertheless,
photolysis of DMD in this solvent leads to the formation of
dimethylcarbene-R,R,R-trifluoromethylbenzene adducts, de-
tected by GC-MS. Thus, the measured activation barrier in
this solvent cannot be associated with intramolecular hydrogen
migration although the isotope effects indicate there is more
rearrangement chemistry in R,R,R-trifluoromethylbenzene than
in Freon-113.
(π)2(π*)0 95.8% (π)2(π*)0 85.3%
(π)0(π*)2 4.1% (π)0(π*)2 2.7%
a The contributions of the individual configurations are given as
relative weights. b Weight of the CASSCF reference function in the
SD-MRCI wave function. c Weight of the CASSCF reference function
in the CASPT2 wave function. d Value of the T1-diagnostic of the
CCSD(T) wave functions.
In summary, all these indicators support the initial assumption
that the MP2 method is a suitable and economic choice to
compute the geometries of the systems of this work and confirm
earlier conclusions.9a However, there are indicators that the
MP2 method may overestimate hyperconjugative effects.34,35
Another factor affecting the accuracy of the calculated
reaction barriers and exothermicities is the selection of the
correlation method (cf. Table 6). As the present systems are
reasonably well described by a single determinant reference
function, there is no doubt that the CCSD(T) method provides
the most accurate results. Comparing the relative energies
computed at various levels of theory, we observe that the
CASPT2, SD-MRCI, and CCSD(T) methods all predict the
barrier height to be within less than 0.7 kcal/mol. The
correspondence is somewhat worse, about 2 kcal/mol, for the
exothermicity. As the difference is largest for the CCSD(T)
method we may conclude that triple and higher order excitations
are needed to achieve a balanced correlation treatment, in
particular with respect to the partially occupied π* orbital at
the carbenoid center. We also note that the relative MP2
energies appear to be shifted to lower values by 2 kcal/mol.
This effect may be due to the overestimation of the hypercon-
jugation effect which appears to be strongest in dimethylcarbene.
There are a number of additional factors affecting the
accuracy of the present calculations. Among these is the zero
point vibrational energy correction which may be too large due
to the harmonic approximation. Another aspect is the saturation
of the one-particle basis space.
To summarize, the predicted barrier heights and exothermici-
ties are computed with an accuracy of about 2 kcal/mol. The
geometries are believed to be the largest source of error leading
to an under- and overestimation of barrier height and exother-
micity, respectively.
The rearrangement of dimethylcarbene to propene has been
studied previously by Evanseck and Houk38 who employed the
MP2 method and somewhat smaller basis sets (6-31G*).
Evanseck and Houk reported that the lowest energy structure
of dimethylcarbene is of Cs symmetry, and their best theoretical
estimate for the activation energy of the 1,2-H-shift reaction
was 4.7 kcal/mol. This value is 2.4 kcal/mol smaller than our
best estimate for two reasons: The lowest energy structure of
singlet dimethylcarbene is of C2 symmetry and is 0.98 kcal/
mol lower in energy than the Cs structure favored by Evanseck
It is possible that DMC forms a complex20-22,40 with an
aromatic solvent. Intuition suggests that complexation of DMC
would lead to kinetic stabilization and to raising the barrier to
rearrangment, a view supported by the calculations of Hadad
and Sulzbach.39 Our data do not argue for or against complex-
ation of DMC with R,R,R-trifluoromethylbenzene, but com-
plexation does not explain the small barrier measured in
C6H5CF3.
DMC is less likely to react with or complex to perfluoro-
hexane than to R,R,R-trifluoromethylbenzene. Nevertheless, the
enthalpic barrier to disappearance of DMC in perfluorohexane
is actually smaller than that measured in the presumably more
reactive solvents. However, the enthalpic advantage is offset
by a more unfavorable entropic term. Thus, the rate of
disappearance of DMC, in the absence of pyridine and at 298
K, is slowest in perfluorohexane and is fastest in Freon-113, as
deduced from the Arrhenius parameters. This is consistent with
increasing reaction of DMC with solvent upon changing the
medium from C6F14 to C6H5CF3 to CF2ClCFCl2.
Quantum mechanical tunneling (QMT) can be invoked in
hydrogen transfer reactions which overcome small apparent
barriers. Indeed, Dix, Herman and Goodman42 have proposed
that the 1,2-rearrangement of hydrogen in methylchlorocarbene,
in solution, has a contribution of quantum mechanical tunneling.
Their data have been fit by theory,43 which validates their
conclusions. Tunneling “reduces” the barrier from 10.9 to 7.7
kcal/mol in methylchlorocarbene and by 2.2-2.6 kcal/mol in
(39) Sulzbach, H. M.; Platz, M. S.; Schaefer, H. F., III; Hadad, C. M. J.
Am. Chem. Soc. 1997, 119, 5682.
(40) (a) Liu, M. T. H.; Bonneau, R. J. J. Phys. Chem. 1989, 93, 7298.
(b) Bonneau, R.; Liu, M. T. H.; Kim, K. C.; Goodman, J. L. J. Am. Chem.
Soc. 1996, 118, 3829. (c) Liu, M. T. H. J. Phys. Org. Chem. 1993, 6, 696.
(d) Liu, M. T. H.; Bonneau, R. J. Am. Chem. Soc. 1990, 112, 3916. (e)
Houk, K. N.; Rondan, N. G.; Mareda, J. Tetrahedron 1985, 41, 1555. (f)
Houk, K. N.; Rondan, N. G.; Mareda, J. J. Am. Chem. Soc. 1984, 106,
4291. (g) Blake, J. F.; Wierschke, S. G.; Jorgensen, W. L. J. Am. Chem.
Soc., 1989, 111, 1919.
(41) This is supported by ab initio calculations. Hadad, C. M.; Sulzbach,
H. M. private communication.
(42) Dix, E. J.; Herman, M. S.; Goodman, J. L. J. Am. Chem. Soc. 1993,
115, 10424.
(36) Raghavachari, K.; Whiteside, R. A.; Pople, J. A.; Schleyer, P. V.
R. J. Am. Chem. Soc. 1981, 103, 5649.
(37) Schleyer, P. V. R.; Koch, W.; Liu, B.; Fleischer, U. J. Chem. Soc.,
Chem. Commun. 1989, 1098.
(38) Evanseck, J. D.; Houk, K. N. J. Am. Chem. Soc. 1990, 112, 9148.

Rearrangement of Dimethylcarbene to Propene
J. Am. Chem. Soc., Vol. 120, No. 18, 1998 4437
Table 6. Total and Relative Energiesa
RHF
MP2
CASSCF
CASPT2
SD-MRCI
CCSD(T)
Dimethylcarbene
-117.020226
-117.459037
-117.039055
-117.465449
-117.510107
-117.522636
Transition State Structure
-116.995464
(15.54)
-117.450119
-117.010491
-117.452929
-117.496473
-117.509073
(5.60)
(17.92)
(7.85)
(8.55)
(8.51)
Propene
-117.118444
(-66.63)
-117.575217
(-72.90)
-117.145366
(-66.71)
-117.577827
(-70.52)
-117.622222
(-70.35)
-117.631995
(-68.62)
a The total energies are given in hartrees whereas the relative energies are given in kcal/mol and in parentheses.
cis- and trans-1,2-dichloroethylidene, respectively.43 QMT
explains the inter- and intramolecular transfer of hydrogen in
carbenes prepared in low-temperature matrixes.44 Typically, the
case for QMT in matrix reactions at low temperature is much
stronger than that which can be made at higher temperature in
fluid solution because the measured kinetic isotope effects are
very large at low temperature.
VI. Experimental Section
Laser Flash Photolysis Studies. Stock solutions of the 3,3-
dimethyl-3H-diazirine and 3,3-bis(trideuteriomethyl)-3H-diazirines were
prepared for each solvent of interest with an optical density of 0.35 (
0.20 at 355 nm for lifetime and kinetic experiments. The solvents
utilized were Freon-113, R,R,R-tifluoromethylbenzene, and perfluoro-
hexane. A constant volume (1.5 mL) of the stock solution was added
to Suprasil quartz fluorescence-free static cells and a varying amount
of pyridine until the pyridine concentration reached approximately 1.12
M. A special sample holder for variable temperature work was used
to avoid scattering the laser beam, and the temperature of the setup
was regulated with a NESLAB RTE-110 variable temperature control-
ler.
The LFP apparatus16 consists of a Continuum PY62C-10 Nd:YAG
laser (355 nm, 30 mJ, 2 ns). Samples were deoxygenated by bubbling
with dry, oxygen free argon for 2 min. Each of the samples was placed
in the appropriate sample cell fitted with a rubber septum. The sample
cells were irradiated with Nd:YAG laser pulses that impinged on the
sample at a right angle to a 150 W Xe arc lamp fitted with an aspherab
beam collimator. The monitoring beam was focused on the slit of an
Oriel monochromator, selected for the wavelength of interest, with both
front and rear slits set between 0.2 and 0.4 mm. Signals were obtained
with a photomultiplier tube detector and were digitized by a Tektronix
7912 A/D transient digitizer. The entire apparatus is controlled by a
Macintosh IIx microcomputer which was used for storage of the
digitized data. Data analysis was performed with IgorPro PPC (version
3.02) by Wavemetrics.
Preparation of 3,3-Dimethyl-3H-diazirine.45 Ammonia was con-
densed in a solution of acetone (4.00 g, 0.07 mol) in methanol (20
mL) at -35 °C. Hydroxylamine-O-sulfonic acid was added (slowly)
over a period of 45 min, and the reaction was allowed to stir for 4 h
at -35 °C. The temperature was slowly raised to room temperature
and the ammonia allowed to evaporate. The resulting solution was
filtered, and the undissolved ammonium salts were washed with 2 ×
50 mL of methanol. The methanol washes were collected, and the
solvent was evaporated, leaving a residue which was dissolved in
chloroform (50 mL), filtered, and dried over anhydrous sodium sulfate.
Chloroform was evaporated, yielding an oil which solidified (mp 38-
40 °C) upon standing.
Crude diaziridine (3.00 g, 0.04 mole) was dissolved in water (25
mL) in a modified three-necked flask. A tube (∼15 cm) containing
KOH and Drierite is attached to the side port of the reaction flask and
the other side connected to a trap prepurged with argon-containing trap
solvent (Freon-113, R,R,R-trifluoromethylbenzene, or perfluorohexane).
Argon was circulated from the reaction flask exiting through the trap.
The reaction flask was cooled to 0 °C, and a solution of potassium
permanganate (6.48 g, 0.04 mol) in 75 mL of 10% aqueous NaOH
was added over a period of 30 min. The volatile diazirine was eluted
with the carrier gas into the trap cooled to -78 °C and identified by
its characteristic UV spectrum (λmax ) 365 nm) and 1H NMR (CDCl3)
δ (ppm) 0.99 (CH3, 6H, s).
The isotope effects observed on the lifetime of DMC in
perfluorohexane are large and dramatic (∆EaD - ∆EaH ) 3.18
kcal/mol, AD/AH ) 158) and cannot be explained by transition
structure theory. Thus, we conclude that the decay of DMC
and DMC-d6 in perfluorohexane is intramolecular and that the
rearrangement of DMC proceeds by a quantum mechanical
tunneling mechanism in this solvent. The barrier to disappear-
ance of DMC-d6 in perfluorohexane is close to that predicted
by theory. Therefore, the decay of DMC under these conditions
may not necessarily contain a component of QMT and is not
altered by complexation with perfluorohexane.
V. Conclusions
Laser flash photolysis (Nd:YAG laser, 355 nm, 35 mJ, 150
ps) of dimethyldiazirine and dimethyldiazirine-d6 produces
dimethylcarbene (DMC) and dimethylcarbene-d6 (DMC-d6),
respectively. The carbenes were trapped with pyridine to form
ylides which absorb near 364 nm. It was possible to resolve
the growth of the ylides as a function of pyridine concentration
in Freon-113, R,R,R-trifluoromethylbenzene, and perfluoro-
hexane as a function of temperature. The observed rate constant
(kobs) to ylide formation was linearly dependent on the concen-
tration of pyridine. From plots of kobs versus pyridine concen-
tration it was possible to extract values of kpyr (absolute rate
constant of reaction of the carbene with pyridine) and τ, the
carbene lifetime, and their associated Arrhenius parameters. In
Freon-113 the carbenes decay by rearrangement and by reaction
with solvent. In perfluorohexane the carbene decay appears to
be predominantly unimolecular. The Arrhenius parameters
indicate that the rearrangement of hydrogen has a strong
tunneling component. The fact that the observed barrier to
rearrangement of DMC is much smaller (2.56 kcal/mol) than
that predicted by ab initio theory (7.4 ( 2 kcal/mol) also points
to the involvement of tunneling in the rearrangement of DMC.
The barrier to rearrangement of deuterium in DMC-d6 in
perfluorohexane (5.6 kcal/mol) is comparable to that predicted
by theory. Tunneling and complexation with solvent may not
be important in the deuterated system in perfluorohexane.
Preparation of 3,3-Bis(trideuteriomethyl)-3H-diazirine. A solu-
tion of acetone-d6 (4.35 g, 0.068 mol) and ammonia-d3 (10 mL, 0.34
(43) (a) Storer, J. W.; Houk, K. N. J. Am. Chem. Soc. 1993, 115, 10420.
(b) Keating, A. E.; Garcia-Garibay, M. A.; Houk, K. N. J. Am. Chem. Soc.
1997, 119, 10805.
(44) (a) Liu, M. T. H.; Bonneau, R.; Wierlacher, S., Sander, W. J.
Photochem. Photobiol., A 1994, 84, 133. (b) Wierlacher, S., Sander, W.;
Liu, M. T. H. J. Am. Chem. Soc. 1993, 115, 8943.
(45) (a) Schmitz, E. In Chemistry of Diazirines; Liu, M. T. H., Ed.; CRC
Press: Boca Raton, FL, 1987; Vol. I, p 57. (b) Schmitz, E.; Ohme, R. Chem.
Ber. 1961, 94.

4438 J. Am. Chem. Soc., Vol. 120, No. 18, 1998
Ford et al.
mol) was prepared at -35 °C under an argon atmosphere in anhydrous
methanol-d. To this solution was added tert-butylhypochlorite (14.80
g, 0.136 mol) over a period of 1.5 h, and the resulting solution was
stirred continuously for an additional 4 h. The temperature was slowly
raised to room temperature and the ammonia allowed to evaporate.
The resulting solution was filtered, and the undissolved ammonium
salts were washed with 2 × 50 mL of methanol. The methanol washes
were collected, and the solvent was evaporated, leaving a residue which
was dissolved in chloroform (50 mL), filtered, and dried over anhydrous
sodium sulfate. Chloroform was evaporated, yielding an oil. No further
purification was attempted at this point.
quartz cuvettes with the solutions degassed by bubbling with Ar before
photolysis using a Rayonet apparatus. The diazirine was photolyzed
at 350 nm for 24 h at 5 °C.
GC-MSD mass spectral analyses were performed on a HP-5970B
mass spectrometer, 5965A infrared detector, with a capillary direct
interface connected to a HP-5890 gas chromatograph with a fused silica
capillary column cross-linked with 5% phenyl methyl silicone (column
i.d. 0.25 nm, column length 15 m). GC traces and fragmentation patterns
in Freon-113, R,R,R-trifluoromethylbenzene, and perfluorohexane are
provided as Supporting Information.
Crude diaziridine (1.50 g, 0.02 mol) was dissolved in D2O (25 mL)
in a modified three-necked flask. A tube (∼15 cm) containing KOH
and Drierite was attached to the side port of the reaction flask and the
other side connected to a trap prepurged with argon-containing trap
solvent (Freon-113, R,R,R-trifluorotoluene, or perfluorohexane). Argon
was circulated from the reaction flask exiting through the trap. The
reaction flask was cooled to 0 °C, and a solution of potassium
permanganate (3.24 g, 0.02 mol) in 25 mL of 10% aqueous NaOH
was added over a period of 30 min. The volatile diazirine was eluted
with the carrier gas into the trap cooled to -78 °C and identified by
its characteristic UV spectrum (λmax ) 364 nm) and 1H NMR (CDCl3)
δ (ppm) 1.033 ((CD3)2), with % D ) 100.
Acknowledgment. Support of this work by the NSF is
greatly appreciated (Grant CHE-9613861). We are also indebted
to Professors Carpenter and Moss for useful conversations about
kinetic isotope effects and to Professors Moss and Hadad for
allowing us to quote their results prior to publication. This paper
is dedicated to Professors Maitland Jones and Shelton Bank on
the occasions of their 60th and 65th birthdays, respectively.
Supporting Information Available: Results of GC-MS
analysis on the photolysis of dimethyldiazirine as in Freon-113
and R,R,R-trifluoromethylbenzene (7 pages, print/PDF). See
any current masthead page for ordering information and Web
access instructions.
GC-MS Analysis. 3,3-Dimethyldiazirine was condensed into
Freon-113, R,R,R-trifluoromethylbenzene, or perfluorohexane and the
concentration adjusted to give ∼0.10 mol of diazirine (determined
spectrophotometrically, ꢀ360 ) 1200). Photolyses were conducted in
JA9724598