Benzylchlorocarbene: Origins of Arrhenius Curvature in the Kinetics of the 1,2-H Shift Rearrangement

Article abstract of DOI:10.1021/jo972198k

Benzylchlorocarbene (1, BCC) was generated photochemically from benzylchlorodiazirine (2) in isooctane, methylcyclohexane (MCH), and tetrachloroethane (TCE) at temperatures from ～30 to -75°C. At -70°C in isooctane, the identified products included Z/E-β-chlorostyrenes 4 (46.6%), α-chlorostyrene 5 (2.4%), l,1-dichloro-2-phenylethane 6 (1.9%), a BCC-isooctane insertion product 8 (5.5%), carbene dimers 9 (3.8%), and azine 3 (30%). The significant incursion of intermolecular products 3, 8, and 9 implies that laser flash photolytic (LFP) kinetic data for the decay of BCC obtained at low temperature is biased and should not be employed in Arrhenius analyses. Accordingly, previously obtained curved Arrhenius correlations for BCC do not necessarily implicate , quantum mechanical tunneling (QMT) in the 1,2-H shift rearrangement of BCC to 4. Similarly in MCH, where BCC affords a solvent insertion product in ～44-53% yield, the curved Arrhenius correlation (Figure 1) cannot be readily interpreted. In polar solvents such as TCE, clean H-shift reactions of BCC are obtained even at -71°C; an Arrhenius correlation of LFP kinetic data is linear from 3 to -71°C (Figure 2), affording E_a = 3.2 kcal mol^-1 and log A = 10.0 s^-1. Therefore, QMT does not appear to play a major role in the 1,2-H shift rearrangement of BCC at ambient or near ambient temperature in solution.

Full text of DOI:10.1021/jo972198k

3010
J . Org. Chem. 1998, 63, 3010-3016
Ben zylch lor oca r ben e: Or igin s of Ar r h en iu s Cu r va tu r e in th e
Kin etics of th e 1,2-H Sh ift Rea r r a n gem en t
Dina C. Merrer,1 Robert A. Moss,*,1 Michael T. H. Liu,*,2 J . T. Banks,2-4 and Keith U. Ingold3
Department of Chemistry, Rutgers, The State University of New J ersey,
New Brunswick, New J ersey 08903, Department of Chemistry, University of Prince Edward Island,
Charlottetown, Prince Edward Island, Canada CIA 4P3, and Steacie Institute for Molecular Sciences,
National Research Council of Canada, Ottawa, Ontario, Canada KIA 0R6
Received December 3, 1997
Benzylchlorocarbene (1, BCC) was generated photochemically from benzylchlorodiazirine (2) in
isooctane, methylcyclohexane (MCH), and tetrachloroethane (TCE) at temperatures from ∼30 to
-75 °C. At -70 °C in isooctane, the identified products included Z/E-â-chlorostyrenes 4 (46.6%),
R-chlorostyrene 5 (2.4%), 1,1-dichloro-2-phenylethane 6 (1.9%), a BCC-isooctane insertion product
8 (5.5%), carbene dimers 9 (3.8%), and azine 3 (30%). The significant incursion of intermolecular
products 3, 8, and 9 implies that laser flash photolytic (LFP) kinetic data for the decay of BCC
obtained at low temperature is biased and should not be employed in Arrhenius analyses.
Accordingly, previously obtained curved Arrhenius correlations for BCC do not necessarily implicate
quantum mechanical tunneling (QMT) in the 1,2-H shift rearrangement of BCC to 4. Similarly in
MCH, where BCC affords a solvent insertion product in ∼44-53% yield, the curved Arrhenius
correlation (Figure 1) cannot be readily interpreted. In polar solvents such as TCE, clean H-shift
reactions of BCC are obtained even at -71 °C; an Arrhenius correlation of LFP kinetic data is
linear from 3 to -71 °C (Figure 2), affording Ea ) 3.2 kcal mol-1 and log A ) 10.0 s-1. Therefore,
QMT does not appear to play a major role in the 1,2-H shift rearrangement of BCC at ambient or
near ambient temperature in solution.
In tr od u ction
tational studies support the importance of tunneling in
the 1,2-H shift of MeCCl, helping to rationalize both the
negative ∆Sq and the unusual KIE temperature depen-
dence.9
Curvature in Arrhenius correlations usually signals
the active competition of 2 or more reaction channels
characterized by differing activation energies. Perhaps
the most interesting case is the incursion of quantum
mechanical tunneling (QMT) in a classically activated
process, where “the Arrhenius plot of log k against 1/T
is nonlinear, with the curvature tending toward zero
activation energy at low temperature”.5
Benzylchlorocarbene (BCC, 1), which is conceptually
Such a phenomenon appears to occur in the 1,2-H shift
of methylchlorocarbene (MeCCl) to vinyl chloride,6 which
suffers increasing intervention of QMT as the tempera-
ture decreases.6a Indeed, even at 25 °C, tunneling is
believed to account for >85% of the total rearrangement.6a
QMT is also suggested to account for two other unusual
findings in this reaction: the very negative entropy of
activation (-16.1 eu), and an increasing kH/kD for H(D)
shift with increasing reaction temperature (from 0.9 to
1.8 as T rises from 248 to 343 K).6a Negative entropies
of activation may simply reflect geometric constraints and
are rather common for intramolecular carbenic rear-
rangements,7 but the inverse temperature dependence
of the kinetic isotope effect (KIE) is unusual.8 Compu-
derived from MeCCl by the replacement of a hydrogen
atom by a phenyl group, can be readily generated from
benzylchlorodiazirine (2) and has been intensively stud-
ied since 1984.7,10,11 The results have raised questions
about the existence of carbene/alkene complexes,7,10-12 the
occurrence of excited diazirine rearrangements,6b,13 and
(8) Increasing KIE with increasing temperature requires a classical
reaction that has an appreciable KIE over the temperature range
investigated but that has a KIE that is reduced by the tunneling
correction.6a Normally, QMT is associated with very large KIEs.5
(9) Storer, J . W.; Houk, K. N. J . Am. Chem. Soc. 1993, 115, 10426.
However, the computational results indicate that, above 200 K, the
classical rate exceeds that due to tunneling, and the KIE decreases
with increasing temperature.
(1) Rutgers University.
(2) University of Prince Edward Island.
(3) Steacie Institute for Molecular Sciences.
(4) Current address: Department of Chemistry, Lakehead Univer-
sity, Thunder Bay, Ontario, Canada P7B 5E1.
(5) Isaacs, N. S. Physical Organic Chemistry, 2nd ed.; Longman:
Essex, England, 1995; pp 304f.
(6) (a) Dix, E. J .; Herman, M. S.; Goodman, J . L. J . Am. Chem. Soc.
1993, 115, 10424. (b) La Villa, J . A.; Goodman, J . L. Tetrahedron Lett.
1990, 31, 5109. (c) La Villa, J . A.; Goodman, J . L. J . Am. Chem. Soc.
1989, 111, 6877.
(10) Tomioka, H.; Hayashi, N.; Izawa, Y.; Liu, M. T. H. J . Am. Chem.
Soc. 1984, 106, 454. Liu, M. T. H.; Chishti, N. H.; Tencer, M.; Tomioka,
H.; Izawa, Y. Tetrahedron 1984, 40, 887.
(11) Liu, M. T. H. Acc. Chem. Res. 1994, 27, 287.
(12) (a) Liu, M. T. H. Chem. Commun. 1985, 982. (b) Bonneau, R.;
Liu, M. T. H.; Kim, K. C.; Goodman, J . L. J . Am. Chem. Soc. 1996,
118, 3829.
(7) Moss, R. A. In Advances in Carbene Chemistry; Brinker, U. H.,
Ed.; J AI Press: Greenwich, CT, 1994; Vol. 1, pp 59f.
S0022-3263(97)02198-1 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/10/1998

Benzylchlorocarbene: Origins of Arrhenius Curvature
J . Org. Chem., Vol. 63, No. 9, 1998 3011
Ta ble 1. P r od u ct Distr ibu tion s for th e P h otolysis of 2 in Isoocta n ea
T (°C)
Z-4
E-4
5
6
8
9
3
10
unknown
22
14.7
(0.3)
12.0
(0.04)
11.6
(0.03)
10.5
74.8
(1.7)
66.6
(0.2)
56.0
(0.2)
36.1
(0.3)
6.48
(0.12)
13.0
(0.05)
4.87
(0.01)
2.40
(0.02)
1.15
(0.02)
2.19
(0.01)
3.59
(0.01)
5.47
(0.05)
1.64
(0.03)
5.15
(0.02)
13.7
(0.04)
30.3
(0.3)
1.19
(0.02)
1.04
0
-35
-70
(0.00)
5.91
(0.02)
1.91
4.40
(0.01)
b
3.79
(0.04)
(0.1)
(0.02)
a
Product distributions are taken from the averages of three capillary GC determinations as measured against a decane internal standard;
b
errors are in parentheses. Values are not calibrated for detector response. Two unknowns, 7.3 and 2.2%.
the intervention of QMT.6a,9,14,15 For BCC, as with
MeCCl, tunneling is a possibility because of curved
Arrhenius correlations that attend the kinetics of the 1,2-
H(D) shifts of BCC and its R,R-d2 analogue and because
of the abnormal KIE’s that accompany these rearrange-
ments.15
At 10-30 K in argon matrices, directly observed BCC
decays by rearrangement much more rapidly than an-
ticipated, based on extrapolations of 25 °C kinetic data.
with the isooctane solvent (although no product corre-
sponding to this reaction was reported;15 see below) or
to the incursion of QMT.
The QMT hypothesis would be strengthened if it could
be shown that products derived from intermolecular
reactions are absent. For example, formation of azine 3
by an intermolecular reaction of BCC with diazirine 2,
in competition with the intramolecular 1,2-H carbene
rearrangement, offers an alternative explanation for the
curved Arrhenius correlation: if the Ea for azine forma-
tion were significantly lower than that for the 1,2-H shift,
the Arrhenius correlation of the rate constants for
carbene decay would flatten out as the temperature
decreased. In the case of MeCCl, where azine is known
to form, this problem was addressed by “correcting” the
kinetics of carbene decay, at each temperature of mea-
surement, by extrapolation of the associated rate con-
stant to its value at [diazirine] ) 0.6a
Its R,R-d2 analogue is considerably more stable; kH/kD
≈
2000 could be estimated at 10 K. These results clearly
implicate QMT in the 1,2-H shift of BCC at a cryogenic
temperature,14 while the parallel behavior of MeCCl and
BCC (curved Arrhenius correlations, abnormal KIE’s) at
a higher temperature6a,15 suggests that tunneling could
also be important for BCC at higher temperatures. A
very careful examination of BCC chemistry is required
to evaluate this possibility.
What about BCC?
Much information has already been acquired. Ther-
molysis or photolysis of 2 in CCl4 or benzene yielded Z-
and E-â-chlorostyrenes, attributed to the 1,2-H shift of
BCC.10 Initial laser flash photolytic (LFP) kinetic studies
from 0 to 31 °C, carried out either by direct observation
of the BCC or by the pyridine ylide method,16 gave linear
Arrhenius correlations and activation parameters of Ea
) 4.83 ( 0.24 kcal mol-1, log A ) 11.28 ( 0.20 s-1, or Ea
) 4.5 ( 0.3 kcal mol-1, log A ) 11.1 ( 0.2 s-1, respec-
tively.17 At 24 °C, k(1,2-H) was 5.4 × 107 s-1. Compa-
rable data were later obtained for para-substituted BCC,
where the substituents included CF3, Cl, Me, and MeO.18
The unusual Arrhenius behavior of MeCCl6 induced
Liu and Bonneau to extend the temperature range15 of
their initial17 BCC kinetic studies. The linearity observed
over the 0-31 °C temperature range17 disappeared when
the decay kinetics of BCC and BCC-d2 were studied in
isooctane from -80 to 60 °C.15 For BCC, the rate
constant first decreased with decreasing temperature and
then leveled off at -40 °C; with BCC-d2, leveling occurred
at -10 °C. The limiting rates of carbene decay converged
at 1 × 107 s-1 for both carbenes15 and as with MeCCl,6a
the KIE for BCC/BCC-d2 appeared to increase with
increasing temperature (viz., kH/kD ) 0.87 at -50 °C and
2.62 at 30 °C).15 The Arrhenius curvature was attributed
either to competitive intermolecular reactions of BCC
A reinvestigation showed that photolytically generated
BCC did yield azine 3 by reaction with 2; in neat
diazirine, 40% of Z/E-3, 40% of the 1,2-H shift products,
Z/E-4, 10% of the 1,2-Ph shift product 5, and 10% of the
carbene-HCl addition product 6 were formed.19 The rate
constant for azine formation was determined by LFP to
be (3.6 ( 0.4) × 108 M-1 s-1, but this reaction was
considered to be important only at high [2]: under the
LFP conditions employed in the Arrhenius studies of
BCC, azine 3 was not reported, and it was not observed
when 10 mM 2 was decomposed to 20% conversion at -80
°C in isooctane.15
On the other hand, a recent study of the kinetics and
activation parameters associated with the 1,2-rearrange-
ments of the BCC analogue mesitylmethylchlorocarbene
7 (and 7-R,R-d2) revealed curved Arrhenius correlations
that were definitely associated with competitive azine
formation.20 At a concentration of 20 mM diazirine
precursor, Arrhenius correlations for 7 and 7-d2 were
curved, and significant azine formation was observed at
-35 and -70 °C. However, when the diazirine concen-
tration was reduced to 10 mM, the Arrhenius correlation
remained linear to -70 °C.20 Moreover, with relevance
to the present report, a preliminary reinvestigation of the
2/BCC system revealed that up to 30% of azine 3 formed
when 15 mM 2 was photolytically decomposed in isooc-
tane at -70 °C, suggesting that competitive azine forma-
(13) (a) White, W. R. III; Platz, M. S. J . Org. Chem. 1992, 57, 2841.
(b) Modarelli, D. A.; Morgan, S.; Platz, M. S. J . Am. Chem. Soc. 1992,
114, 7034.
(14) Wierlacher, S.; Sander, W.; Liu, M. T. H. J . Am. Chem. Soc.
1993, 115, 8943.
(15) Liu, M. T. H.; Bonneau, R.; Wierlacher, S.; Sander, W. J .
Photochem. Photobiol., A: Chem. 1994, 84, 133.
(16) J ackson, J . E.; Soundararajan, N.; Platz, M. S.; Liu, M. T. H.
J . Am. Chem. Soc. 1988, 110, 5595.
(17) Liu, M. T. H.; Bonneau, R. J . Am. Chem. Soc. 1990, 112, 3915.
(18) Liu, M. T. H.; Bonneau, R. J . Am. Chem. Soc. 1992, 114, 3604.
(19) Liu, M. T. H.; Chapman, R. G.; Bonneau, R. J . Photochem.
Photobiol., A: Chem. 1992, 63, 115.
(20) Moss, R. A.; Merrer, D. C. Chem. Commun. 1997, 617.

3012 J . Org. Chem., Vol. 63, No. 9, 1998
Merrer et al.
tion was at least a contributing cause of the Arrhenius
curvature attending the rearrangement of BCC.20
Herein, we provide a full account of our joint reinves-
tigation of the products and the kinetics for the reactions
of BCC as generated from benzylchlorodiazirine. In-
cluded are new studies of solvent and temperature
dependence.
Resu lts
P r od u ct Stu d ies. Solutions of 2 in isooctane contain-
ing a decane internal standard (0.014 M, A344 ) 0.7) were
photolyzed in a Rayonet reactor (350 nm) until all the
diazirine was destroyed (UV). The photolyses were
conducted at 22, 0, -35, or -70 °C, and the product
solutions were analyzed by capillary GC and GC-MS.
Table 1 summarizes the product distributions.
In addition to the H-shift products, â-chlorostyrenes
Z-4 and E-4, the Ph-shift R-chlorostyrene 5, and the HCl
addition product 6, we also observed a BCC-isooctane
insertion product 8, BCC dimer(s) 9, the azine 3, and
F igu r e 1. Arrhenius correlations for BCC in MCH: ([) 8.5
mM 2, (2) 18 mM 2-d2.
centrations because of insufficiently intense carbene
signals during the LFP experiments.
At first glance, the results appear consistent with the
isooctane findings, where azine formation is a principal
cause of Arrhenius curvature. Azine 3 also forms in these
MCH experiments as evidenced by UV examination of
spent photolysis mixtures, in which the broad azine
absorption can be seen between 300 and 330 nm and is
enhanced at -80 °C, relative to the 22 °C photolyzates.
However, product studies by GC-MS reveal that the
BCC-MCH photolyses are complicated in a way different
from that of the isooctane studies: in MCH, azine is a
minor contaminant, but BCC insertion into the MCH
becomes a major problem. At 25 °C, with [2] ) 18 mM,
GC-MS examination of the photolyzate reveals 15.3%
of Z-4, 69.0% of E-4, 1.2% of azine 3, and 14.4% of a
BCC-MCH product (m/e 236, 238). At -75 °C, with [2]
) 18 mM, BCC-MCH increases to 44%, while Z/E-4
drops to only 23% of the product mixture. Azine 3 (4.5%)
is a minor constituent, but two unknowns comprise 29%
of the product mixture. The results are no better at [2]
) 8.5 mM at -75 °C, when BCC-MCH constitutes 53%
of the products, but alkenes 4 are only 24%. Azine is
reduced to 2.6%, but unknowns contribute 21%.
Clearly, the low-temperature LFP data of Figure 1 do
not mainly reflect carbenic decay via the 1,2-H shift, and
they are not very useful for further analysis. The facile
low-temperature reaction of BCC with MCH, however,
is of considerable interest and will be further examined.
Stu d ies in Tetr a ch lor oeth a n e. Earlier experiments
demonstrated that the 1,2-H rearrangements of BCC (or
BCC-d2) exhibited linear Arrhenius behavior in solvent
CHCl3 from 60 to -55 °C; for BCC the Arrhenius
parameters were Ea ) 3.6 kcal mol-1 and log A ) 10.4.15
The change from curved to linear Arrhenius correlation
as the solvent is altered from isooctane to chloroform has
been associated with a more efficient 1,2-H shift (relative
to intermolecular reactions in the more polar solvent).23
Such an effect is very marked with benzylfluorocarbene
(PhCH2CF, BFC) photolytically generated from ben-
(traces of) benzyl chloride 10. Products 4-6 were identi-
fied by comparisons of GC retention times and NMR
spectra to those of authentic or previously reported
samples.19,21 Products 8, 9, and 10 were assigned by GC-
MS; the structure of 8 is uncertain, although it probably
corresponds to BCC insertion into the tertiary C-H bond
of isooctane. Azine 3 was prepared by a “preparative”
photolysis of 2 in isooctane (A344 ≈ 1.0) with purification
by silica gel chromatography. The assignment of the
structure rests on NMR, MS, and an exact mass deter-
mination.
Table 1 reveals that the product mixture derived from
BCC becomes more complex with decreasing reaction
temperature. Notable are the now documented appear-
ances of the isooctane insertion product 8, carbene dimer
9, and azine 3. Their contributions are greatest in the
-70 °C photolysis, where azine constitutes 30% of the
product. These trends parallel our observations with
carbene 7.20
Rea ction s in Meth ylcycloh exa n e. Photolyses of
diazirine 2 were also carried out in methylcyclohexane
(MCH) for comparison to the studies in isooctane. BCC
and BCC-R,R-d2 were photolytically generated from the
diazirines at [2] ) 18 or 8.5 mM. Carbene decay was
monitored directly at 310 nm as a function of tempera-
ture (20 to -78 °C) in LFP experiments; the results are
shown as Arrhenius correlations in Figure 1.
Curvature is very apparent in the 18 mM BCC-d2
correlation, whereas it appears to be more gentle in the
8.5 mM BCC correlation, with an onset at lower temper-
ature.22 It was impractical to use lower diazirine con-
(23) Polar solvents stabilize the “polar” 1,2-H shift transition state:
see ref 13b, and Sugiyama, M. J .; Celebi, S.; Platz, M. S. J . Am. Chem.
Soc. 1992, 114, 966. The more negative ∆Sq observed in the polar
solvents may reflect enhanced solvation of the H-shift transition state.
(21) Liu, M. T. H.; Subramanian, R. J . Org. Chem. 1985, 50, 3218.
(22) The results at least superficially resemble those obtained15 with
BCC and BCC-d2 using 10 mM [2] in isooctane.

Benzylchlorocarbene: Origins of Arrhenius Curvature
J . Org. Chem., Vol. 63, No. 9, 1998 3013
behavior occurs for BCC, together with simplification of
its chemistry. Parallel experiments indicate that analo-
gous results can be obtained in other, nonhalogenated
polar solvents, such as methyl benzoate (ꢀ ) 6.59)25 or
ethyl acetate (ꢀ ) 6.11).25
Discu ssion
Curved Arrhenius correlations for 1,2-H shift carbenic
rearrangements have been reported for MeCCl,6 MesCH2-
CCl (7),20 and BCC (1).15 For temperatures down to at
least -70 °C, it is clear that the curvature for 7 is mostly
due to competition between (intramolecular) 1,2-H rear-
rangement and (intermolecular) azine formation.20 The
latter process is effectively temperature independent and
must have a near zero activation energy,26 so that it
becomes increasingly important at low temperature.
For BCC, the product studies of Table 1 clearly reveal
the presence of azine 3, solvent insertion product 8, and
carbene dimer 9 in a total yield of ∼40% at -70 °C and
in lesser but still significant quantities at higher tem-
peratures. Again, curvature in the Arrhenius correla-
tions obtained in isooctane15 or in MCH (Figure 1) can
be attributed to competitive intermolecular reactions,
with solvent insertion of particular significance in MCH.
A reviewer has suggested that we “correct” the BCC
Arrhenius correlation for the incursion of azine 3 using
the measured rate constant for the reaction of BCC and
diazirine 2.26 When this is done, the curvature is
“straightened” and its onset is moved to lower temper-
ature (i.e., from -45 °C before correction to -60 °C after
correction). This is in keeping with our suggestion that,
as with carbene 7,20 much or all of the Arrhenius
curvature exhibited by BCC above -70 °C is associated
with competing intermolecular reactions. Nevertheless,
the reviewer notes that bimolecular reactions (such as
azine formation) and tunneling could have similar acti-
vation parameters and might occur together, making
their differentiation very difficult.
We suggest that QMT plays a relatively minor role
in the BCC 1,2-H shift at temperatures above -70 °C in
solution. At lower temperatures, we may expect an
incursion of QMT. This would not be surprising because
QMT is known to govern the 1,2-H shift of BCC in argon
matrices at 30 K.14 What is yet unknown is the contri-
bution of QMT to the BCC H-shift between -243 and -70
°C. At what temperature does its contribution equal or
exceed that of the classical rearrangement? Is QMT
important beyond cryogenic conditions? It will be dif-
ficult to answer these questions experimentally because
of the competitive intermolecular reactions encountered
at low temperatures, but LFP studies in high-melting
matrices or in viscous nonpolar fluids may provide new
information.
F igu r e 2. Arrhenius correlation for BCC in TCE ([2] ) 22
mM).
zylfluorodiazirine. In hydrocarbon solvents at 25 °C, this
carbene’s H-shift products are accompanied by 20-38%
of azine, whereas in the polar solvent 1,1,2,2-tetrachlo-
roethane (TCE), azine formation is reduced to 6%.24
The kinetics of BFC rearrangement have been deter-
mined in TCE,24 so for purposes of comparison, the BCC
rearrangement was examined in this solvent. The prod-
uct distribution, in contrast to that of the isooctane or
MCH cases, is uncomplicated. Samples of diazirine 2
(∼22 mM) were completely photolyzed in the Rayonet
reactor (λ ) 350 nm) at 22, 0, -35, or -70 °C. Products
were analyzed by capillary GC against an internal decane
standard. Only Z- and E-4 were found at each temper-
ature, with Z/E distributions ranging from 18:82 (-70
°C) to 22:78 (22 °C). Even at -70 °C, no azine or other
additional products could be found.
LFP kinetic studies of the BCC f 4 H-shift in TCE
utilized the pyridine ylide method,16 which afforded a
stronger signal than direct observation of BCC. After the
351-nm laser flash, growth of the BCC-pyridine ylide
signal was followed at 370 nm in TCE solutions contain-
ing 22 mM 2 and 1.4-9.0 mM pyridine. Extrapolation
of the linear correlation of kobs for ylide formation vs
[pyridine] to [pyridine] ) 0 then afforded a Y intercept
which, in view of the very clean BCC f 4 rearrangement,
could be taken as kre for the 1,2-H shift in TCE. At 0 °C,
for instance, kre was 2.67 × 107 s-1. Repetition of these
determinations at 11 temperatures from 3.2 to -71.4 °C
afforded 11 rate constants which appear as an Arrhenius
correlation in Figure 2. In contrast to the curvature
observed in isooctane15 or MCH (see Figure 1), good
linearity (r ) 0.988) is obtained in TCE (as in CHCl315).
The Arrhenius parameters in TCE are Ea ) 3.2 kcal
mol-1 and log A ) 10.0 s-1, corresponding to ∆Sq ) -14.5
eu (at 25 °C), which compare quite well with the earlier
determinations in CHCl3.15
For BCC, two remaining observations, a significantly
negative ∆Sq for the 1,2-H shift and a KIE that increases
with increasing temperature, require comment. A nega-
tive activation entropy is common in 1,2-H (and 1,2-C)
carbenic rearrangements7 and can be associated with the
unusual geometries of these rearrangements that engen-
In both TCE (ꢀ ) 8.20)25 and CHCl3 (ꢀ ) 4.81),25 the
intramolecular hydride shift is potentiated, effectively
shutting off alternative intermolecular chemistry (e.g.,
azine formation, solvent insertion). Linear Arrhenius
(26) For example, kq for azine formation from BCC and 2 is (3.6 (
0.4) × 108 M-1 s-1 in isooctane at 25 °C19 and (3.15 ( 0.3) × 108 M-1
s-1 in MCH at -90 °C (this work). The rate constants were obtained
by LFP, following the rate of decay of BCC as a function of [2].
(27) Moss, R. A.; Ho, G.-J .; Shen. S.; Krogh-J espersen, K. J . Am.
Chem. Soc. 1990, 112, 1638. Modarelli, D. A.; Platz, M. S. J . Am. Chem.
Soc. 1993, 115, 470.
(24) Moss, R. A.; Maksimovic, L.; Merrer, D. C. Tetrahedron Lett.
1997, 38, 7049.
(25) Dean, J . A., Ed. Lange’s Handbook of Chemistry, 14th ed.;
McGraw-Hill: New York, 1992; p 5.91-5.125.

3014 J . Org. Chem., Vol. 63, No. 9, 1998
Merrer et al.
Ta ble 2. Ar r h en iu s P a r a m eter s for Rea r r a n gem en t of
BCC a n d BF Ca
der reduced transmission coefficients across the activa-
tion barriers.27 The unusual temperature dependence of
the KIE may be suspect at lower temperatures because
the kH and kD values were obtained by interpolations on
smooth curves drawn through experimental rate con-
stants that are now known to have been “contaminated”
by intermolecular reactions. However, increasing KIEs
(2.00-2.62) from 0 to 30 °C remain unexplained and are
analogous to the KIEs observed with MeCCl, where the
unusual temperature dependence has been ascribed to
QMT.6a,9
Presently, MeCCl is the only known carbene for which
tunneling at ambient temperatures is considered to make
a major contribution to the 1,2-H shift.6a,9 A priori, one
would expect that the phenyl “substitution” that converts
MeCCl to BCC should lower the Ea for the classical 1,2-H
shift of BCC, relative to MeCCl, and also lower the
temperature at which QMT would come to dominate the
classical H-shift mechanism. Ab initio computations28,29
clearly display the anticipated ∆Ea, but experimental
results for MeCCl6a and BCC (see Table 2) indicate a
smaller ∆Ea between these carbenes.
Ea,
kcal log A, ∆Sq,
entry carbene
conditions
mol-1
s-1
eu
ref
1
2
3
4
5
BCC
BCC
BCC
BCC
BFC
isooctane, 31 to 0 °Cb
isooctane, 60 to -42 °Cd
CHCl3, 60 to -55 °C
TCE, 3 to -71 °C
4.8
5.8
3.6
3.2
3.2
11.3
11.9
10.4 -12.9
10.0 -14.5
9.5 -17.2
-8.9
-6.2
c
e
e
f
TCE, 66 to -68 °C
g
a
b
At 298 K. Estimated errors are (0.5 kcal mol-1 in Ea. Data
are for the direct observation of BCC. c Reference 17. Corrected
for reaction with the solvent, assuming k ) 9 × 106 s-1 for this
reaction. The main intermolecular reaction, however, is azine
d
formation, see above. e Reference 15. f This work. Reference 24;
g
the activation parameters are means of three determinations, with
errors of (10% in Ea and (2.4% in log A.
In TCE, however, the azine yield was reduced to 6%,
while variable temperature LFP studies afforded the
activation parameters shown in Table 2 (entry 5).
Surprisingly, the activation parameters for the 1,2-H
shifts of BCC and BFC in TCE differ by only 2.7 eu in
∆Sq, favoring ∼4-5-fold faster rearrangement of the
chlorocarbene.24 This modest difference suffices to pro-
duce ∼50% of azine from BFC at -50 °C, even in TCE,24
whereas no azine is detected in comparable BCC reac-
tions. The cautionary message is that major contamina-
tion of intramolecular carbenic rearrangements by in-
termolecular side reactions can be induced by very small
changes in the activation parameters. Careful attention
to experimental factors such as precursor concentration,
solvent, and reaction temperature are essential to obtain-
ing uncomplicated product mixtures and interpretable
kinetic data.
The activation energy for the 1,2-H shift of BCC has
now been determined several times in isooctane, CHCl3,
and TCE; the data are collected in Table 2, entries 1-4.
Given the influence of competing intermolecular reactions
in isooctane and the uncertainties attending correction
of the data (entry 2), it is difficult to obtain a precise set
of activation parameters. The value obtained in isooctane
at 0-31 °C, where intermolecular reactions are minimal,
is Ea ) 4.8 kcal mol-1 17
It is the best hydrocarbon
.
solvent value currently available, though it does not differ
experimentally from the value for MeCCl (Ea ) 4.9 ( 0.5
kcal mol-1).6c Under these conditions, the activation
entropy for the BCC 1,2-H shift is moderately negative
at ∼-9 eu (∆Sq ∼ -16 eu for MeCCl).6c
Con clu sion s
In polar solvents such as CHCl3 or TCE (entries 3 or
4, Table 2), the activation energy for 1,2-H migration is
reduced by ∼1-1.5 kcal mol-1, presumably due to sta-
bilization of the polar transition state 11 by the polar
The Arrhenius curvature observed for the rate con-
stants of the 1,2-H rearrangement of BCC in hydrocarbon
solvents is mainly due to competitive intermolecular
chemistry (azine formation, solvent insertion) down to a
temperature of -60 to -70 °C. In this regard, the
chemistry of BCC resembles that of its mesityl analogue
7.20 In polar solvents (CHCl3, TCE), the products formed
from BCC are solely the intramolecular H-shift isomeric
â-chlorostyrenes 4; no intermolecular products are en-
countered, even at -70 °C. Rate constants and activation
parameters were determined for the BCC 1,2-H shift in
several solvents: the activation energies are ∼3.2 kcal
mol-1 in the polar solvents and ∼4.8 kcal mol-1 in the
hydrocarbon solvent; the entropy of activation is negative,
ranging from ∼-6 to -14 eu, depending on the solvent,
conditions of measurement, and corrections applied to the
data. QMT does not appear to play a major role in the
1,2-H shift rearrangement of BCC at ambient or near
ambient temperatures in solution. At temperatures
below -70 °C, however, QMT may contribute signifi-
cantly; by -243 °C, its influence is dominant.14
solvent.23 These small decreases in the Ea for BCC
rearrangement appear to be dramatically reflected in the
product composition, which is remarkably clean in the
polar solvents. In contrast to the complexity caused by
the incursion of intermolecular reactions in isooctane or
15
MCH, in CHCl3 or TCE, only the intramolecular BCC
H-shift products Z- and E-4 are observed.
A similar sensitivity to solvent was observed with
benzylfluorocarbene (BFC), where 25 °C photolysis of the
corresponding diazirine (11.6 mM) in various hydrocar-
bon solvents gave 20-38% of the azine in addition to 62-
80% of the expected â-fluorostyrene H-shift products.24
Exp er im en ta l Section
(28) Calculated (B3LYP/6-31G*, with ZPE corrections) values of Ea
Gen er a l. Proton NMR spectra were determined on a
Varian VXR-200 spectrometer (200 MHz). Chemical shifts are
reported as ppm (δ) relative to TMS. GC-MS data were
obtained on a Hewlett-Packard (HP) 5890 GC (column: HP5)
equipped with an HP 5971 mass selective detector. High-
resolution mass spectra (HRMS) were recorded on a VG ZAB-T
double-focusing mass spectrometer. Capillary GC analyses
for 1,2-H shifts of MeCCl and BCC are 12.9 and 6.8 kcal mol-1
,
respectively,29 although these values are clearly too high; e.g., the
6c
experimental value of Ea for the 1,2-H shift of MeCCl is ∼5 kcal mol-1
.
Small curvature tunneling corrections reduce the classical, calculated
Ea for MeCCl to 8.8 kcal mol-1
(29) Keating, A. E.; Houk, K. N. Private communication, August 4,
1997.
.

Benzylchlorocarbene: Origins of Arrhenius Curvature
J . Org. Chem., Vol. 63, No. 9, 1998 3015
were acquired on a Varian 3700 GC with a flame ionization
detector (injector 200 °C, detector 290 °C) connected to a
Varian 4270 integrator. For all analytical GC data, we used
a 25 m × 0.25 mm × 0.25 µm Chrompack CP-Sil 5CB
(Chrompack, Inc., Raritan, NJ ), chemically bonded 100%
dimethyl polysiloxane column. Radial preparative TLC was
conducted on a Chromatotron with a silica gel rotor (2 mm
thickness, Analtech, Inc., Newark, DE). Photolyses were
performed in a Rayonet photochemical reactor equipped with
16 RPR-3500 bulbs (λ ) 350 nm). Constant-temperature
photolyses utilized a Pyrex Dewar flask and a methanol-dry
ice bath. UV spectra were obtained on an HP 8451 diode array
spectrophotometer.
All chemicals were purchased from Aldrich Chemical Co.
unless otherwise noted. Hydrogen chloride and ammonia
gases were purchased from Matheson Gas Products, E. Ru-
therford, NJ . Pyridine was dried by distillation from CaH2
and stored over molecular sieves. sym-Tetrachloroethane was
dried over CaCl2. All other chemicals were used as received.
A 10% decane solution in isooctane was used as an internal
standard in all thermolyses and photolyses unless otherwise
noted.
LFP experiments in TCE utilized a Lambda Physik EMG
101 XeF excimer laser. The laser produced pulses at 351 nm
with a 14-ns duration at an average power of 35-50 mJ /
pulse.30 Data were processed with Igor Pro 2.0 software
(Wavemetrics, Inc.) on a Macintosh IIsi personal computer.
Suprasil quartz cuvettes (1 × 1 cm) were used to hold the
samples. For variable-temperature experiments, the sample
cell was equilibrated in an ethanol or an ethanol-dry ice bath
of the appropriate temperature before insertion in the sample
holder. The temperature of the sample at the instant of the
laser flash was measured by an indwelling thermocouple
connected to a digital thermometer. LFP experiments in MCH
utilized 355-nm light (third harmonic) from a Nd:YAG laser.
This system has also been described previously.31
3-Ch lor o-3-ben zyld ia zir in e (2). The diazirine was pre-
pared by hypochlorite oxidation32 of benzylamidinium hydro-
chloride,33 mp 145-148 °C. A mixture of 7.0 g of LiCl and 3.0
g (0.018 mol) of benzylamidinium hydrochloride in 100 mL of
DMSO and 100 mL of pentane was placed in a 1-L, three-
necked round-bottom flask fitted with a dropping funnel,
mechanical stirrer, and a thermometer. Aqueous NaOCl
solution (150-200 mL; “pool chlorine”, 11% OCl-) saturated
with NaCl was added with stirring to the mixture; the
temperature was maintained at 35-40 °C. During the reac-
tion, an additional 50 mL of pentane was added. After
addition of the NaOCl solution was completed, the mixture
was stirred at room temperature for 30 min. The pentane
layer was washed with 3 × 1000 mL of water and dried over
CaCl2. The diazirine solution was purified by chromatography
on silica gel (pentane eluent) to afford diazirine 2, λmax ) 344,
346, 356 nm.
The structure of E-4 was confirmed by comparison of its
NMR spectrum with the previous report21 and by GC-MS. 1H
NMR (CDCl3): 6.64 and 6.85 (AB q, J AB ) 13.7 Hz, 2H); 7.31
(s, 5H). GC-MS: m/e (rel intensity) 138, 140 (65:22, M+, M+
+ 2), 103 (base peak, M+ - 35).
P h otolyses of Dia zir in e 2 in Isoocta n e. Solutions of 2
in isooctane (A344 ) 0.7, 2.0 mL) were contained in sealed
quartz cuvettes and photolyzed at 22, 0, -35, or -70 °C until
all the diazirine had been destroyed in each sample (20-40
min). The products were analyzed by capillary GC and GC-
MS. Products 5 and 6 were identified by NMR and GC
comparisons with authentic samples (see below). Product 8
(insertion of BCC into a C-H bond of isooctane) and product
9 (dimer of BCC) were identified by GC-MS. Azine 3 was
prepared as described below and purified by chromatography
on silica gel.
r-Ch lor ostyr en e (5). A 100-mL, three-necked round-
bottom flask was fitted with a rubber septum, a condenser
topped with an N2 inlet, and a thermometer. Into the
N2-purged system was placed 11.0 g (0.053 mol) of PCl5. To
this was slowly added with stirring 9.0 mL (0.075 mol) of
acetophenone. The reaction mixture was heated at 80 °C for
70 min. 1H NMR and GC-MS of the crude product revealed
>90% of 5. The NMR spectrum (CDCl3) compared well with
the previously reported data:34 5.52 and 5.77 (AB q, J AB ) 1.6
Hz, 2H); 7.35 (m, 5H). GC-MS: m/e (rel intensity) 138, 140
(46:16, M+, M+ + 2), 103 (base peak, M+ - 35).
1,1-Dich lor o-2-p h en yleth a n e (6). A solution of benzyl-
chlorodiazirine (2) in 1.0 M HCl/ether (A342 ≈ 2.0, 10 mL) was
placed in a screw-top Pyrex pressure tube and photolyzed for
3 h at room temperature in the Rayonet reactor. The reaction
was complete when the yellow reaction solution turned color-
less. Excess HCl was removed by vacuum with a water
aspirator, and dichloride 6 was isolated by radial preparative
TLC on silica gel (pentane eluent). Fraction 1 contained
â-chlorostyrenes (4); fraction 2 contained the desired product.
1H NMR data matched the data previously reported by Liu.21
1H NMR (CDCl3): 3.46 (d, J ) 6.42 Hz, 2H); 5.83 (t, J ) 6.43
Hz, 1H); 7.31 (s, 5H). GC-MS: m/e (rel intensity) 174, 176,
178 (9.3:5.3:1.3, M+, M+ + 2, M+ + 4), 159, 161, 163 (47:29:
5.3, M+ - 15), 91 (base peak, M+ - 83).
P r od u cts 8 a n d 9. Chromatographic fractions from the
isolation of azine 3, which contained isooctane insertion
product 8 and dimer 9 were analyzed by GC-MS: 8 m/e (rel
intensity) 252 (1.3, M+), 145 (6.9, M+ - 107), 131 (5.0, M+
-
121), 113 (20, M+ - 139), 91 (36.3, M+ - 161), 57 (base peak,
M+ - 195). 9 m/e (rel intensity) 276, 278, 280 (44.7:29.8:5.1,
M+, M+ + 2, M+ + 4), 205 (44.7, M+ - 71), 163 (88.1, M+
113), 91 (base peak, M+ - 185).
-
Azin e 3.19 Several 4.0-mL samples of 2 in isooctane (A344
≈ 1.0) were photolyzed at -70 °C for 20 min each. After
photolysis, the product mixtures were combined, and crude 3
was purified by column chromatography on silica gel (pentane
eluent). 1H NMR (CDCl3): 3.97 (s, 4H); 7.34 (s, 10H). GC-
Th er m olysis of Dia zir in e 2 in Isoocta n e. A solution of
2 in isooctane (A344 ) 0.65) was placed in a screw-top Pyrex
pressure tube and heated in the dark in an oil bath at 105 °C
for 5.5 h. Capillary GC analysis revealed the hydride-shift
products Z- and E-4 (> 98% total yield); the Z/E ratio was 1:7.
The identity of Z-4 was confirmed by comparison of its NMR
spectrum with the reported spectrum21 and by GC-MS. 1H
NMR (CDCl3): 6.28 and 6.65 (AB q, J AB ) 8.4 Hz, 2H); 7.31
(s, 5H). GC-MS: m/e (rel intensity) 138, 140 (80:24, M+, M+
+ 2), 103 (base peak, M+ - 35).
MS: m/e (rel intensity) 304, 306, 308 (10.6:7.5:2.5, M+, M+
+
2, M+ + 4)], 269 (13.8, M+ - 35), 116 (base peak, M+ - 188),
91 (90, M+ - 213). HRMS (EI): exact mass for C16H14N2Cl2
calcd, m/e 304.053 404; observed, 304.053 367.
P h otolyses of Diazir in e 2 in Meth ylcycloh exan e. These
experiments were conducted at [2] ) 18 or 8.5 mM in MCH at
25 or -75 °C in a manner analogous to that of the isooctane
photolyses (see above). The product distributions are discussed
in the Results section. The BCC-MCH product was identified
by GC-MS: m/e (rel intensity) 236, 238 (13:4.3, M+, M+ + 2);
128 (10, M+ - 108); 97 (base peak, M+ - 139); 91 (55, M+
145); 55 (41, M+ - 181).
-
(30) For a description of this installation, see Moss, R. A.; Shen, S.;
Hadel, L. M.; Kmiecik-Lawrynowicz, G.; Wlostowska, J .; Krogh-
J espersen, K. J . Am. Chem. Soc. 1987, 109, 4341 and Moss, R. A.; Xue,
S.; Liu, W.; Krogh-J espersen, K. J . Am. Chem. Soc. 1996, 118, 12588.
(31) Kazanis, S.; Azarami, A.; J ohnston, L. J . J . Phys. Chem. 1991,
95, 4430.
(32) Graham, W. H. J . Am. Chem. Soc. 1965, 87, 4396.
(33) The amidine was prepared by Pinner’s method from phenylac-
etonitrile, as described by Dox, A. W.; Whitmore, F. C. In Organic
Syntheses, 2nd ed.; Gilman, H., Blatt, A. H., Ed.; Wiley: New York,
1941; Collect. Vol. I., pp 5f.
LF P Stu d ies in Tetr a ch lor oeth a n e. Solutions of 2 in
sym-tetrachloroethane (A346 ) 1.0) containing (1.4-9.0) × 10-3
M of pyridine (2.0 mL total) were irradiated with an XeF
excimer laser (λ ) 351 nm) at selected temperatures over a
range of -71 to +3 °C. The formation of the BCC-pyridine
ylide was monitored at 370 nm.
(34) Vo-Quang Yen, M. Ann. Chim. 1962, 7, 785.

3016 J . Org. Chem., Vol. 63, No. 9, 1998
Merrer et al.
For product studies, samples of 2 in TCE (A346 ) 1.0, 2.0
mL) were placed in sealed quartz cuvettes and irradiated in
the Rayonet reactor at -70, -35, 0, and 22 °C for 40 min.
Capillary GC analysis revealed only â-chlorostyrenes Z- and
E-4 as the products in all cases.
and Ms. A. E. Keating for helpful discussions. The
authors at Rutgers thank the National Science Founda-
tion for financial support. The Canadian authors are
grateful to the Natural Sciences and Engineering Re-
search Council of Canada.
Ack n ow led gm en t. Dedicated to the memory of Dr.
I. D. R. Stevens. We are grateful to Professor K. N. Houk
J O972198K