Laser flash photolysis studies on intra- and intermolecular reactions of some halocarbenes

Article abstract of DOI:10.1021/ja960176u

Laser flash photolysis studies on α-methylbenzylchlorodiazirine, n-propylchlorodiazirine, and isopropylchlorodiazirine have been carried out at low temperatures, ranging from 170 to 230 K. The present results, together with those previously obtained with benzylchlorodiazirine and methylchlorodiazirine, provide a direct measurement of the 'by-stander effect' of alkyl substituents on the C atom in the α position of the carbene center. Intermolecular reactions of α-methylbenzylchlorocarbene and tetramethylethylene in isooctane and methylcyclohexane have been measured but this reaction is shown to be too slow to measure by LFP in highly viscous decalin solvent.

Full text of DOI:10.1021/ja960176u

8098
J. Am. Chem. Soc. 1996, 118, 8098-8101
Laser Flash Photolysis Studies on Intra- and Intermolecular
Reactions of Some Halocarbenes
Michael T. H. Liu* and Roland Bonneau
Contribution from the Department of Chemistry, UniVersity of Prince Edward Island,
Charlottetown, Prince Edward Island, C1A 4P3 Canada, and URA 348, Laboratoire de Chimie
Physique A, UniVersite´ de Bordeaux I, 33405 Talence, France
ReceiVed January 18, 1996X
Abstract: Laser flash photolysis studies on R-methylbenzylchlorodiazirine, n-propylchlorodiazirine, and isopropy-
lchlorodiazirine have been carried out at low temperatures, ranging from 170 to 230 K. The present results, together
with those previously obtained with benzylchlorodiazirine and methylchlorodiazirine, provide a direct measurement
of the “by-stander effect” of alkyl substituents on the C atom in the R position of the carbene center. Intermolecular
reactions of R-methylbenzylchlorocarbene and tetramethylethylene in isooctane and methylcyclohexane have been
measured but this reaction is shown to be too slow to measure by LFP in highly viscous decalin solvent.
Introduction
Scheme 1
In recent years, absolute kinetics of 1,2-H migration in many
alkylhalocarbenes and dialkylcarbenes have been obtained by
the laser flash photolysis (LFP) technique with rate constants
ranging from 105 to 108 s-1. Developments in this research
area have been duly reviewed.1 In 1989, we reported the rate
constant for several alkylchlorocarbenes.2 The lifetime of some
carbenes was found to be too short at room temperature to be
measured with our experimental apparatus which has a response
time of approximately 4 ns. It should be possible to measure
these lifetimes accurately at lower temperatures, because the
1,2-H migration usually shows a non-negligible activation
energy. We now wish to report the results of LFP studies on
R-methylbenzylchlorodiazirine (1a), n-propylchlorodiazirine
(2a), and isopropylchlorodiazirine (3a). These results will be
compared to those previously obtained with benzylchlorocarbene
and methylchlorocarbene, thereby providing a direct measure-
ment of the “by-stander effect” of alkyl substituants on the C
atom in the R position of the carbene center, discussed by
Nickon.1b
4 ns. For low-temperature measurements, the cell holder was put in a
clear Dewar flask, where it was cooled down to -120 °C by a stream
of cold nitrogen gas. The temperature was measured with a thermo-
couple immersed in the solution, close to the region where the excitation
and analytical beams cross each other.
Experimental Section
Results
Diazirines 1a, 2a, and 3a were prepared by hypochlorite oxidation
of the corresponding amidines.3a The decomposition products of 1a,
2a, and 3a have been described previously.3b,c Tetramethylethylene
was acquired from Aldrich and used without purification. Pyridine
was distilled over CaH2. The laser flash photolysis setup used a
crossed-beam arrangement. The sample, in a 10 × 10 mm cell, was
excited at 355 nm by single light pulses (200 ps; 5-30 mJ) provided
by a frequency tripled mode-locked Nd-YAG laser (Quantel). The
detection system (pulsed Xe arc, monochromator, photmultiplier, and
Tektronix 7912 transient recorder) had a response time of approximately
In LFP experiments, every species that lacks a strongly
absorbing chromophore, i.e., one with a molar absorption
coefficient g103 M-1 cm-1, is invisible to laser experiment with
UV-vis detection. This difficulty was overcome by the
pyridinium ylide probe technique developed by Jackson et al.1f
The pyridinium probe technique may be used in two different
ways, relying on the measurement of either the kinetics or the
yield of the formation of the pyridinium ylide. In the Stern-
Volmer method, the amplitude of the ylide absorption is
measured as a function of pyridine, with excitation, temperature,
and other variables being constant. The amplitude of the ylide
* Address correspondence to this author at the University of Prince
Edward Island.
X Abstract published in AdVance ACS Abstracts, July 1, 1996.
(1) (a) Schaefer, H. F. Acc. Chem. Res. 1979, 12, 288. (b) Nickon, A.
Acc. Chem. Res. 1993, 26, 84. (c) Liu, M. T. H. Acc. Chem. Res. 1994, 27,
287. (d) Moss, R. A. AdVances in Carbene Chemistry; Brinker, U., Ed.;
JAI: Greenwich, CT, 1994; Vol. 1. (e) Moss, R. A. Pure Appl. Chem. 1995,
67, 741. (f) Jackson, J. E.; Platz, M. S. AdVances in Carbene Chemistry;
Brinker, U., Ed.; JAI: Greenwich, CT, 1994; Vol. 1.
absorption, OD, is proportional to Φ ) ky[pyridine]/(k1
+
ky[pyridine]). With P as the proportional factor, the yield is
OD ) PΦ, and 1/OD ) 1/P + k1/Pky[pyridine]. LFP of 1a, in
isooctane at 25 °C, in the presence of pyridine gives the
pyridinium ylide absorption, at 370 nm. The plot of the
reciprocal of the ylide absorption, 1/OD vs 1/[pyridine], yielded
a straight line. The ratio slope/intercept of this line led to k1/ky
) 7.7 × 10-2 M, where k1 is the rate constant for disappearance
of the R-methylbenzylchlorocarbene (1b), in the absence of
(2) Bonneau, R.; Liu, M. T. H.; Rayez, M. T. J. Am. Chem. Soc. 1989,
111, 5973.
(3) (a) Graham, W. H. J. Am. Chem. Soc. 1965, 87, 4396. (b) Frey, H.
M.; Liu, M. T. H. J. Chem. Soc. A 1970, 1916. (c) Tomioka, H.; Hayashi,
N.; Izawa, Y.; Liu, M. T. H. J. Am. Chem. Soc. 1984, 106, 454.
S0002-7863(96)00176-X CCC: $12.00 © 1996 American Chemical Society

Intra- and Intermolecular Reactions of Halocarbenes
J. Am. Chem. Soc., Vol. 118, No. 34, 1996 8099
Figure 1. Arrhenius plot for 1,2-H shift in C6H5(CH3)CH-C¨ -Cl
obtained by measuring the rate of decay of carbene in isooctane at 305
nm.
Figure 2. Arrhenius plot for C6H5(CH3)CH-C¨ -Cl in decalin at 305
nm: 9, no TME; 2, [TME] ) 1.0 M.
Table 1. Temperature Dependence for C6H5(CH3)CH-C¨ -Cl in
Various Solvents
temp (°C)
τ(ns)
temp (°C)
τ(ns)
temp (°C)
τ(ns)
Isooctane
-43
-46
-45.5
-55
-64
-68
10
12
12.5
15.7
19
-72
-74
-77
-79
-83
-85
26.6
29.8
33
33
43
-87
-91
50
58
77
98
101
-94
-102
-105
24
44
Methylcyclohexane
-68
-75
-76
-78
-88
27
-89
55
67
95
91
-106
-110
-116
-119
120
141
185
222
39.7
37.4
41
-93
Figure 3. Quenching of C6H5(CH3)CH-C¨ -Cl by TME in isooctane
as a function of temperature: b, kobs ) k1 + kq[TME]. Solid line
calculated from A and Ea in isooctane: 2, kq[TME] with [TME] )
0.21 M.
-104
-105
48
Decalin
-67
-74
-76
-78
24
30
35
35
-88
-94
-95
55
71
80
-100
-101
-112
100
102
182
pyridine, and ky is the rate constant for the reaction of 1b with
pyridine. Under the conditions that the reaction of 1b with 1a
is not significant because the lifetime for 1b is short and the
concentration of 1a is low, k1 can be taken as the rate constant
for the 1,2-H shift in 1b. Assuming that ky ≈ 7.5 × 109 M-1
s-1, by similarity to the value measured with benzylchlorocar-
bene,4 one gets k1 ≈ 5.8 × 108 s-1, i.e., a lifetime <2 ns for 1b
at 25 °C in the absence of pyridine, too short to be measured
with our experimental setup.
In the absence of pyridine, at 230 K, LFP of diazirine 1a in
isooctane produces a transient species absorbing in the range
of between 280 and 310 nm, assigned to 1b, because of its
similarity to benzylchlorocarbene.4 The lifetime of this transient,
monitored at 305 nm, is ≈10 ns at 230 K. This lifetime was
measured as a function of temperature, from 170 to 230 K, in
isooctane (Figure 1), methylcyclohexane (MCH), and decalin
(mixture of cis and trans isomers). The results obtained in these
solvents are similar (see Table 1). The corresponding Arrhenius
plots are good straight lines yielding the following kinetic
parameters:
Ea ) 2.85 ( 0.06, 2.46 ( 0.15, and 3.00 ( 0.06 kcal/mol
and log(A) ) 10.66 ( 0.07, 10.17 ( 0.18, and 10.80 ( 0.07 in
isooctane, MCH, and decalin, respectively. The lifetime of 1b
was then measured as a function of temperature, in the same
range (160-230 K) and in the same solvents, in the presence
of tetramethylethylene (TME). Under these conditions, 1/τ )
k1 + kq[TME] where kq is the rate constant for reaction of 1b
with TME. In decalin (Figure 2) the carbene lifetime is not
affected by the addition of TME up to 1 M, whereas it is
Figure 4. Quenching of C6H5(CH3)CH-C¨ -Cl by TME in methyl-
cyclohexane as a function of temperature: b, kobs ) k1 + kq[TME].
Solid line calculated from A and Ea in methylcyclohexane: 2, kq[TME]
with [TME] ) 0.4 M.
substantially shortened in isooctane (Figure 3) with [TME] )
0.21 M and in MCH (Figure 4) with [TME] ) 0.4 M. The
value of kq as a function of temperature was obtained by the
difference between the reciprocal values of τ at various
temperatures and the value of k1 calculated at the same
temperatures from the Arrhenius parameters listed above. This
gives values of kq around 7 × 107 M-1 s-1 in isooctane and 6
× 107 M-1 s-1 in MCH, but <5 × 106 M-1 s-1 in decalin.
These values of kq are nearly independent of the temperature;
the Arrhenius plots give “apparent activation energies” equal
to 0.66 ( 0.10 kcal/mol in MCH and -0.17 ( 0.11 kcal/mol
in isooctane with preexponential factors respectively equal to
(4.5 ( 2.5) × 108 and (5.5 ( 3) × 107 M-1 s-1
.
In the case of n-propylchlorodiazirine (2a), the LFP does not
produce any observable transient which can be assigned to
n-propylchlorocarbene (2b), but this “invisible” carbene may
be studied by using the pyridinium ylide probe method.1f The
kinetic analysis of the growth of the absorption of the pyridinium
ylide, at 370 nm, gives 1/τgrowth ) kobs ) k1 + ky[pyridine].
(4) Liu, M. T. H.; Bonneau, R. J. Am. Chem. Soc. 1990, 112, 3915.

8100 J. Am. Chem. Soc., Vol. 118, No. 34, 1996
Liu and Bonneau
Figure 7. Plot of log(k1) vs 1/T for 2b in isooctane. The bottom line
is log (k1 - 7 × 106) vs 1/T.
rocarbene,6 is due to a process with a low or a zero activation
energy, which leads to the disappearance of the carbene, either
by an intramolecular 1,2-H shift by tunneling6 or by a reaction
of the carbene with the solvent.5 Assuming that the rate
constant, kx, of this enigmatic process, whatever its real nature
may be, is equal to 7 × 106 s-1 over the whole temperature
range considered here, the rate constant, ki, of the “normal”
1,2-H migration in 2b is given by ki ) k1 - kx ) (k1 - 7 ×
106) s-1. The plot of log(ki) vs 1/T is then a good straight line,
which gives the following kinetic parameters: log(A) ) 10.44
( 0.05 and Ea ) 3.64 ( 0.05 kcal/mol.
The 1,2-H migration in the isopropylchlorocarbene (3b),
produced by laser photolysis of 3a, seems to be very fast, even
at a low temperature, and no accurate measurement of the rate
constant of this reaction could be made. The lifetime of 3b,
estimated from the kinetics of the growth of the pyridinium ylide
absorption, is less than 20 ns at -90 °C. But this corresponds
to τgrowth, which is close to the response time of our detection
system because large amounts of pyridine have to be used. At
low amounts of [pyridine], the low yield of conversion carbene
w ylide (due to the short lifetime of carbene 3b), coupled with
the low yield of the production of the carbene upon photolysis
of 3a (previously noticed by Platz7), results in a small signal
and a poor signal/noise ratio, so that τgrowth cannot be determined
with accuracy even if τgrowth, at low temperature, would be long
enough to be measurable.
Figure 5. Plot of 1/τgrowth or kobs vs temperature in isooctane for three
concentrations of pyridine (b, 1.55 mM; 0, 4.0 mM; 9, 6.2 mM)
obtained by measuring the growth of the pyridinium ylide at 370 nm.
The bottom curve (O) gives the values of k1 for 11 selected tempera-
tures. INSET: Plot of kobs vs [pyridine] with the slopes equal to ky and
intercepts equal to k1.
Discussion
At room temperature, the above measurements yielded
lifetimes of around 2 ns for 1b and 17 ns for 2b, to be compared
respectively to 18 ns for benzylchlorocarbene4 and 330 ns for
methylchlorocarbene.8 Since, in the absence of reactants,
reaction pathways other than isomerization by 1,2-H migration
are negligible, the reciprocal lifetime k1 is equal to ki. In both
cases, the rate constant for 1,2-H migration, ki, is strongly
increased by alkyl substitution of the C atom in the R-position
of the carbene center. This effect, named the “by-stander effect”
by Nickon,1b is well-known. Whereas previous reports on this
effect were founded on relative measurements of the rates of
the rearrangement and addition reactions by products analysis,
we now have a direct measurement of the enhancement of the
rates of rearrangements. Clearly, this enhancement is not due
to an increase of the frequency factor. The A factors measured
in isooctane, 1010.44 s-1 for 2b and 1010.66 s-1 for 1b, are even
slightly lower than the A factor measured for the benzylchlo-
Figure 6. Arrhenius plot for the formation of pyridinium ylide by the
reaction of n-propylchlorocarbene with pyridine.
The values of kobs were therefore measured in isooctane, as a
function of temperature from 170 to 270 K, for three concentra-
tions of pyridine (1.55, 4.0, and 6.2 mM). Since the measure-
ments were not made at the same temperatures for the three
concentrations of pyridine, a smooth curve representing the
variations of 1/τgrowth vs T was drawn through each of the three
sets of points (kobs, T) corresponding to a given value of
[pyridine] (see Figure 5). The values of 1/τgrowth were then
measured on these curves every 10 K, yielding the values of k1
and ky for 11 values of the temperature, ranging from 170 to
270K.
The Arrhenius plot (Figure 6) of log(ky) vs 1/T is a good
straight line and gives log10(A) ) 10.5 and an apparent
activation energy, Ea(meas) ) 1.22 kcal/mol. In contrast, as
shown on Figure 7, the plot of log(k1) vs 1/T is approximately
linear on the upper part of the temperature range but shows a
pronounced curvature in the low-temperature region. This
phenomenon, already observed for benzyl-5 and methylchlo-
(5) Liu, M. T. H.; Bonneau, R.; Wierlacher, S.; Sander, W. J. Photochem.
Photobiol. A: Chem. 1994, 84, 133.
(6) (a) Dix, E. J.; Herman, M. S.; Goodman, J. L. J. Am. Chem. Soc.
1993, 115, 10424. (b) Storer, J. W.; Houk, K. N. J. Am. Chem. Soc. 1993,
115, 10426.
(7) White, W. R.; Platz, M. S. J. Org. Chem. 1992, 57, 2841.
(8) Liu, M. T. H.; Bonneau, R. J. Am. Chem. Soc. 1989, 111, 6873.

Intra- and Intermolecular Reactions of Halocarbenes
J. Am. Chem. Soc., Vol. 118, No. 34, 1996 8101
rocarbene, 1011 s-1: in fact, the A factor is expected to be doubly
smaller for 1b than for benzylchlorocarbene because of the
number of H on the C in the R position. The enhancement in
rate constants is due to a decrease of the activation energy. For
benzylchlorocarbene5 Ea drops from 5.7 kcal/mol to 2.85, 2.46,
or 3.00 kcal/mol (mean value, 2.77) upon substitution of one H
by a methyl. Similarly, Ea changes from 4.9 kcal/mol in
methylchlorocarbene9 to 3.64 kcal/mol for 2b.
In the case of 3b, with the two methyl substituents on the
R-C, the activation energy for rearrangement must be even
lower, probably around 2 kcal/mol (this would give a 10-ns
lifetime at -90 °C if the A factor is 1010.3, slightly lower than
the A factor for 2b). Therefore, lowering the temperature does
not significantly increase the lifetime of 3b.
When considering the rate constants of a bimolecular reaction,
such as for the reaction of 1b with TME, kq, and for the
formation of the pyridinium ylide of 2b, ky, diffusion must be
taken into account. This explains why we could not detect any
measurable decrease of the lifetime of 1b upon addition of TME
in decalin, a viscous solvent in the considered temperature range,
since η increases from ≈1 P at 210 K to ≈100 P at 170 K.10
The rate constant for diffusion is usually given at kdiff ) 8RT/
3000η, and the values of η as a function of T may be obtained
from the literature.10-12 The plots of log(T/η) vs 1/T are nearly
straight lines in the range 210-170 K, yielding an “apparent
activation energy” due to diffusion, Ea(diff), of around 2.6 and
3.4 kcal/mol in isooctane and MCH, respectively. Thus the
activation energy of the reaction between 1b and TME, Ea(react)
) Ea(meas) - Ea(diff), would be (-0.17 - 2.6) ) -2.77 in
isooctane and (0.66 - 3.4) ) -2.74 kcal/mol in MCH.
The correction term, Ea(diff), may be somewhat overestimated
by this method. If the formation of the pyridinium ylide has
virtually no activation energy, since the measured rate constants
are close to a diffusion controlled limit, one would expect that
Ea(meas) ) Ea(diff) for this reaction. However, the values of
Ea(meas) for the reaction of pyridine with 2b and phenylchlo-
rocarbene13 in isooctane are 1.22 and 1.5 kcal/mol, respectively,
significantly lower than the 2.6 kcal/mol obtained from the plot
of log(T/η) vs 1/T. Yet, even if the values of Ea(diff) obtained
from the plots of log (T/η) vs 1/T were indeed overestimated
by ≈1.4 kcal/mol, the values of Ea for the reaction between the
2b and TME would still be negative, -1.35 kcal/mol in
isooctane as well as in MCH. Negative values have already
been reported for the activation energy of the reaction of
phenylchlorocarbene14 and benzylchlorocarbenes15 with TME.
They have been interpreted by the existence of a “carbene-olefin
complex” (COC)14,15 on the way to cyclopropanation or,
alternatively, by a control of the reaction by entropic terms.16
Theoretical chemistry16 has denied the existence of a COC, as
a minimum on the energy surface connecting the system
“carbene + olefin” to “cyclopropane adduct”.
During the last few years, there has been much controversy
between those favoring one or the other of these inter-
pretations.1c,d,3 It still remains true that the analysis of products
obtained by photolysis and thermolysis of diazirines17 in the
presence of olefins indicates an “excess” of rearrangement
products, which can be explained only by a rearrangement
occurring within a COC. The actual value of the probability
for the COC to give cyclopropane instead of the rearrangement
products depends not only on the diazirine but also on the olefin
used as reactant. Steric factors as well as electron densities at
the reactive centers on the carbene and the olefin may be
decisive.
Acknowledgment. M. T. H. Liu thanks the NSERC of
Canada for financial support.
JA960176U
(13) Jackson, J. E.; Soundararajan, N.; Platz, M. S.; Doyle, M. P.; Liu,
M. T. H. Tetrahedron Lett. 1989, 30, 1335.
(14) Turro, N. J.; Lehr, G. F.; Butcher, J. A., Jr.; Moss, R. A.; Guo, W.
J. Am. Chem. Soc. 1982, 104, 1754.
(15) Liu, M. T. H.; Bonneau, R. J. Am. Chem. Soc. 1992, 114, 3604.
(16) Houk, K. N.; Rondan, N. G. J. Am. Chem. Soc. 1984, 106, 4293.
(17) Bonneau, R.; Liu, M. T. H.; Kim, K. C.; Goodman, J. L. J. Am.
Chem. Soc. 1996, 118, 3829.
(9) LaVilla, J. A.; Goodman, J. L. J. Am. Chem. Soc. 1989, 111, 6877.
(10) Von Salis, G. A.; Labhart H. J. Phys. Chem. 1968, 72, 752.
(11) Riddick J. A.; Bunger, W. Tech. Chem. (N. Y.) 1970, 2, 95.
(12) Murov, S. L. Handbook of Photochemistry; M. Dekker Inc., New
York, 1973.