Laser flash photolysis study of alkylhalocarbenes generated from non-nitrogenous precursors

Article abstract of DOI:10.1021/jp972464y

Previously we have studied the photochemistry of alkylchlorodiazirines using laser flash photolysis methods and the pyridine ylide technique (White et al. J. Org. Chem. 1992, 17, 2841). The yield of trappable carbene was found to be a sensitive function of the alkylcarbene structure. This was attributed to ring opening of the diazirine to form a biradical which could suffer rearrangement, in addition to fragmentation to a carbene. It was also possible to explain the data in terms of a carbene-pyridine complex which can either rearrange or collapse to an ylide. In this work, alkylchlorocarbenes are generated from phenanthridenes. The results do not rule out the intermediacy of a carbene-pyridine complex; however, if this species is formed, it does not rearrange to chloroalkene in competition with collapse to form an ylide. This study reduces the urgency with which to postulate carbene-olefin complexes (Tomioka et al. J. Am. Chem. Soc. 1984, 106, 454).

Full text of DOI:10.1021/jp972464y

J. Phys. Chem. A 1998, 102, 1507-1513
1507
Laser Flash Photolysis Study of Alkylhalocarbenes Generated from Non-Nitrogenous
Precursors
Marc Robert, Igor Likhotvorik, and Matthew S. Platz*
Newman and Wolfrom Laboratory of Chemistry, The Ohio State UniVersity, 100 W. 18th AVenue,
Columbus, Ohio 43210-1173
Sarah C. Abbot, Mary M. Kirchhoff, and Richard Johnson*
Department of Chemistry, UniVersity of New Hampshire, Durham, New Hampshire 03824
ReceiVed: July 28, 1997; In Final Form: October 20, 1997
Previously we have studied the photochemistry of alkylchlorodiazirines using laser flash photolysis methods
and the pyridine ylide technique (White et al. J. Org. Chem. 1992, 17, 2841). The yield of trappable carbene
was found to be a sensitive function of the alkylcarbene structure. This was attributed to ring opening of the
diazirine to form a biradical which could suffer rearrangement, in addition to fragmentation to a carbene. It
was also possible to explain the data in terms of a carbene-pyridine complex which can either rearrange or
collapse to an ylide. In this work, alkylchlorocarbenes are generated from phenanthridenes. The results do
not rule out the intermediacy of a carbene-pyridine complex; however, if this species is formed, it does not
rearrange to chloroalkene in competition with collapse to form an ylide. This study reduces the urgency with
which to postulate carbene-olefin complexes (Tomioka et al. J. Am. Chem. Soc. 1984, 106, 454).
I. Introduction
SCHEME 1
Tomioka, Liu, Bonneau (TLB), and co-workers have exten-
sively studied the decomposition of benzylchlorodiazirine (1).1
In alkane solvents, the major decomposition products are (E)
and (Z) â-chlorostyrenes 2, apparently derived from benzyl-
chlorocarbene by 1,2 migration of hydrogen (Scheme 1).
In the presence of a carbene trap such as tetramethylethylene
a cyclopropane adduct 3 is formed. These results can be
explained using the simple mechanism pictured in Scheme 1.
A plot of the yield of 3/2 versus [TME] is not linear.
Furthermore the E/Z ratio of â-chlorostyrene products varies
as a function of [TME]. Both of these observations are
inconsistent with Scheme 1.
As noted by the original investigators, the data require that
there must be at least two pathways that form â-chlorostyrenes.
The second route to â-chlorostyrenes was associated with a
carbene-olefin complex (COC) which can collapse to form
adduct 3, or rearrange to 2, but with an E/Z signature different
from that of a free carbene (Scheme 2). Calculations do not,
however, support the intermediacy of a carbene-alkene com-
plex. The COC is not calculated to be a minimum on the
potential energy surface.2
It is important to note that the need for two pathways that
can form stable products when diazirines are decomposed was
first noted over 30 years ago. This conclusion followed from
the discovery that different modes of diazirine decomposition
lead to very different product mixtures.3
Several groups have studied the photochemistry of alkyl-
chlorodiazirines using the laser flash photolysis (LFP)-pyridine
ylide technique.1,4 In this experiment, an alkylchlorodiazirine
is excited with a pulse from an excimer laser (XeF, 351 nm).
The photochemically prepared carbene is trapped by pyridine
to form an ylide (e.g., 4, Scheme 1). These types of ylides are
easily detected because of their intense absorption and relatively
long lifetimes.4 In our work, the yield of ylide (Ay) was
measured optically as a function of pyridine concentration. The
yield grows smoothly as [pyridine] increases from zero and then
reaches a maximum (Ay∞). At these high concentrations of
pyridine, we assumed that every carbene produced in the laser
pulse was intercepted by pyridine. Under these conditions ylide
formation is much faster than all other possible processes that
can consume the carbene.5 Values of Ay∞ were determined for
a series of diazirines 5.
∞
It was determined that the values of Ay were strongly
dependent on the strength of the C-H bond alpha to the
diazirine moiety. These data can be explained, following TLB,
S1089-5639(97)02464-X CCC: $15.00 © 1998 American Chemical Society
Published on Web 02/07/1998

1508 J. Phys. Chem. A, Vol. 102, No. 9, 1998
Robert et al.
SCHEME 2
SCHEME 3
state. This pathway has been dubbed “rearrangement in excited
states” or RIES.1 RIES is the more economical mechanism, as
photolysis of a diazirine necessarily produces an excited state.
Furthermore, it is our intuition that donation of a lone pair of
electrons into the empty p orbital of a carbene as in a pyridine-
carbene complex would provide kinetic as well as thermody-
namic stabilization and reduce the rate of 1,2 hydrogen shift
relative to a free intermediate.8 Thus it has been our view that
a 1,2 hydrogen shift in the putative complex should not be
competitive with the collapse of a complex to the eventual stable
product.
SCHEME 4
We also noted that it was possible to explain the TLB results
by isomerization of benzylchlorodiazirine to diazo compound
7.5,6 Decomposition of 7 by ionic pathways can lead to
â-chlorostyrenes 2 with a unique E/Z signature and eliminate
the need to invoke carbene-olefin complexes (Scheme 4) to
explain the data.
Thus, it seems clear that non-nitrogenous precursors of
alkylchlorocarbenes would be extremely useful in clarifying and
distinguishing the different mechanistic views that have been
proposed.
Phenanthridene compounds have been used successfully to
study chloro-9 and dichlorocarbene10 and will be shown here
to be useful sources of alkylchlorocarbenes. Related compounds
provide a useful source of dibromocarbene.11
as the result of the formation of a carbene-pyridine complex
6, which can partition between ylide formation and 1,2
rearrangement of hydrogen.
Nevertheless, we identified the diazirine excited state as the
branching point (Scheme 3).5,6 In this proposed mechanism the
diazirine excited-state C-N bond cleaves to form a biradical.7
The biradical partitions between carbene formation and isomer-
ization to alkene in concert with nitrogen extrusion. Thus, we
have associated the second pathway with the biradical/excited

Alkylhalocarbenes from Non-Nitrogenous Precursors
J. Phys. Chem. A, Vol. 102, No. 9, 1998 1509
SCHEME 5
1a,9b-dihydrocyclopropa[1]phenanthrene (1.50 g, 5.8 mmol) in
THF and ether (25 mL, 3:2 v/v) was cooled to -78 °C.
n-Butyllithium in hexanes (11.0 mmol) was added dropwise,
whereupon the mixture became dark forest green. After 1 h,
methyl iodide (1.1 mL, 18.0 mmol) was added. The reaction
was kept at -78 °C for an additional 1 h and then allowed to
warm slowly to room temperature and stirred overnight.
Quenching with brine, ether extraction, and concentration
afforded a yellow solid. The crude product was purified by
rotary chromatography on silica gel, eluted with 2:100 v/v ether/
hexane. The resulting white solid was recrystallized twice from
hexane to give white crystals assigned as 8c (0.73 g, 53%). Data
included: mp 116-118 °C; 1H NMR (360 MHz, CDCl3 δ 1.86
(s 3H), δ 2.65 (s 2H), δ 7.14-7.34 (m, 6H), δ 7.98 (d 2H, J )
7.7 Hz); 13C NMR δ 24.7, 28.3, 32.3, 122.6, 127.1, 127.5, 130.2,
130.6, 131.5; IR (KBr) 3008, 2960, 2926, 1487, 1447 cm-1
.
Anal. Calcd for Cl16H13Cl: C, 79.82; H, 5.45. Found: C,
80.09; H, 5.77.
1-exo-d3-Methyl-1-endo-chloro-1a,9b-dihydrocyclopropa-
[l]phenanthrene (8d). This was prepared as above, but using
deuterated methyl iodide (1 mL, 16.0 mmol, 99.5+ atom %
D). Workup and rotary chromatography afforded a white solid,
which was recrystallized from hexane to give white fluffy
crystals of 8d (0.77 g, 55%): mp 128-130 °C; 1H NMR (360
MHz, CDCl3) δ 2.73 (s, 2H), δ 7.26-7.39 (m, 6H), δ 8.02 (d,
2H); high-resolution MS calcd for C16D3H10Cl 245.0865, found
245.0862.
It is possible, of course, that these precursors fragment to
form biradicals (path b, Scheme 5) in competition with the
concerted cleavage of two bonds to form carbene and phenan-
threne (path a, Scheme 5). However, there are two important
differences between the biradicals of Schemes 3 and 5. It seems
very unlikely, on thermodynamic grounds, that biradical 9 will
fragment to carbene and phenanthrene (path d) or mimic the
free carbene and fragment to phenanthrene plus vinyl chloride
(path e), which corresponds to the key postulate of Scheme 3
(hydrogen migration in concert with fragmentation).12 The
likely fate of biradical 9 should be cyclization to reform starting
material (path c) or 1,2 hydrogen shift to form a 9-substituted
phenanthrene (10, path f).
Thus, with precursors of the type 8, there should be no second
pathway to phenanthrene or to the products of carbene rear-
rangement (e.g., 11). Furthermore, there is also no possibility
of the rearrangement of 8 to a diazo compound and there seems
no facile route to form a species capable of ionic reactions that
will form 11.
Herein we are pleased to report laser flash photolysis studies
of phenanthridene precursors of alkylchlorocarbenes. The
results of this study do not argue for or against the existence of
a carbene-pyridine complex. However, our experimental data
rule out the intermediacy of carbene-pyridine complexes (e.g.,
6) which partition between ylide formation and 1,2 hydrogen
migration. This does not discredit the postulation of a carbene-
olefin complex1 which can migrate hydrogen or form cyclo-
propane but does reduce the urgency with which to embrace
this concept.
1-exo-d3-Ethyl-1-endo-chloro-1a,9b-dihydrocyclopropa[l]-
phenanthrene (8e). 1,1-Dichloro-1a,9b-dihydrocyclopropa[l]-
phenanthrene (1.50 g, 5.7 mmol) in THF and ether (25 mL, 3:2
v/v) was cooled to - 78 °C and treated with n-butyllithium
(11.0 mmol). After 1 h, bromoethane (1.3 mL, 17.0 mmol)
was added dropwise. The reaction was kept at -78 °C for an
additional 1 h and then allowed to warm slowly to room
temperature with stirring overnight. Workup as above followed
by rotary chromatography on silica gel (1:39 v/v ethyl acetate/
hexane) afforded a white solid which was recrystallized twice
from hexanes to give white crystals (0.60 g, 41%): mp 122-
123 °C. 1H NMR (360 MHz, CDCl3) δ 1.22 (t, 3H, J ) 7.2
Hz), δ 1.95 (q, 2H, J ) 7.2 Hz), δ 0.67 (s, 2H), δ 7.19-7.35
(m, 6H), δ 8.00 (δ, 2H, J) 7.7); 13C NMR δ 11.1, 31.8, 35.0,
44.9, 122.8, 127.3, 127.8, 130.4, 130.9, 131.8; high-resolution
MS calcd for C17H15Cl 254.0862, found 254.0850.
1-exo-Benzyl-1-endo-chloro-1a,9b-dihydrocyclopropa[l]-
phenanthrene (8f). As described above, 1,1-dichloro-1a,9b-
dihydrocyclopropa[l]phenanthrene10 (1.48 g, 5.7 mmol) in THF/
ether (25 mL) was treated with n-butyllithium (11.0 mmol).
After 1 h, benzyl bromide (2.1 mL, 18.0 mmol) was added.
The reaction was kept at -78 °C for an additional 1 h and then
allowed to warm slowly to room temperature with stirring
overnight. Workup as above gave a yellow solid which was
purified by rotary chromatography on silica gel and eluted with
1:4 v/v dichloromethane/hexane. The resulting white solid was
recrystallized twice from hexane to give white needles (0.55 g,
1
31%). mp 123.5-125.5 °C; H NMR (360 MHz, CDCl3) δ
2.93 (s, 2H), δ 3.39 (s, 2H), δ 7.27-7.41 (m, 11H), δ 8.02 (δ,
2H, J ) 7.5 Hz); 13C NMR δ 31.2, 43.0, 46.6, 122.7, 126.9,
127.2, 127.7, 128.4, 129.3, 130.3, 130.4, 131.8, 137.6; Anal.
Calcd for C22H17Cl: C, 83.39; H, 5.42. Found: C, 83.79; H,
5.34.
II. Experimental Section
All commercial materials were used without further purifica-
tion. Reactions were performed under an argon atmosphere.
THF was distilled from sodium benzophenone ketyl immediately
before use. Melting points are uncorrected. Exact mass was
determined using a Kratos MS-30 instrument.
1-exo-Isopropyl-1-chloro-1a,9b-dihydrocyclopropa[l]phenan-
threne (8g). To a solution of 1,1-dichloro-1a,9b-dihydro-
cyclopropa[l]phenanthrene13 (0.579 g, 2.22 mmol) in 4 mL of
dry THF was added 2 M isopropylmagnesium chloride in THF
Synthesis. 1-exo-methyl-1-endo-chloro-1a,9b-dihydro-
cyclopropa[l]phenanthrene (8c). A solution of 1,1-dichloro-

1510 J. Phys. Chem. A, Vol. 102, No. 9, 1998
Robert et al.
TABLE 1: Absorption Maxima of Ylides 4 in Isooctane
(4.5 mL, 9 mmol) at room temperature. After 21.5 h the mixture
was cooled to -78 °C, and a 1.5 M chlorine solution in
tetrachloromethane (6 mL, 9 mmol) was added over 10 min
with intensive stirring. The mixture was allowed to warm to
room temperature and concentrated on a rotory evaporator. Flash
chromatography on alumina (hexanes) afforded 71 mg of
product contaminated with 1-isopropyl and 1,1-diisopropyl-1a,-
9b-dihydrocyclopropa[l]phenanthrenes. This solid was recrys-
tallized from pentane to yield 24 mg of pure white crystalline
compound with mp 155-158 °C. Another crop of crystals was
obtained by prep TLC of the mother liquor on silica gel using
hexanes as eluent. The combined yield was 44 mg (7.4%): 1H
NMR (CDCl3) δ 1.21 (d, J ) 6 Hz, 6H), 1.55 (m, 1H), 2.70 (s,
2H), 7.23-7.42 (m, 6H), 7.92-8.03 (m, 2H); 13C NMR (CDCl3)
δ 18.87, 31.98, 39.34, 48.94, 122.61, 127.00, 127.56, 130.04,
130.66, 131.76; HRMS calcd for C18H1735Cl 268.1019, found
268.1021.
1-exo-tert-Butyl-1-chloro-1a,9b-dihydrocyclopropa[l]phenan-
threne (8h). To a well-stirred suspension of copper(I) chloride
(1 g, 10.1 mmol) in 20 mL of THF was added 1.5 M
tert-butyllithium solution in pentane (14 mL, 21 mmol)
dropwise at -50 °C. The mixture was allowed to warm to 0
°C, stirred for 5 min, and then cooled to -50 °C. A solution
of 1,1-dichloro-1a,9b-dihydrocyclopropa[l]phenanthrene13 (0.524
g, 2 mmol) in 6 mL of dry THF was added. The mixture was
stirred for 0.5 h at -50 °C and then at -20 °C for 2.5 h. After
this period the mixture was cooled to -78° C in a dry ice/
acetone bath, and a 1.5 M chlorine solution in tetrachlo-
romethane (15 mL, 22.5 mmol) was added over 5 min, followed
by 5 mL of brine and 3 mL of 10% sodium metabisulfite
solution. The organic phase was separated, washed with brine,
and concentrated in vacuo. Flash chromatography on silica gel
(hexane) yielded 106 mg (18.6%) of the white crystalline com-
pound: mp 145-149 °C after recrystallization from carbon
tetrachloride-hexane; 1H NMR (CDCl3) δ 1.20 (s, 9H), 2.91 (s,
2H), 7.19-7.38 (m, 6H), 7.97-8.08 (m, 2H); 13C NMR (CDCl3)
δ 27.20, 27.80, 36.32, 51.76, 122.61, 127.00, 127.55, 130.09,
131.20, 132.20; HRMS calcd for C19H1935Cl 282.1176, found
282.1181.
Photolysis of 1-exo-Methyl-1-endo-chloro-1a,9b-dihydro-
cyclopropa [l]phenanthrene (8c) with Cyclohexene. Precur-
sor 8c (11.0 mg) was placed into each of two quartz test tubes
with cylohexene (5 mL, freshly distilled) and hexane (15 mL),
and the tubes were closed and purged with nitrogen for 10 min.
One tube was irradiated for 30 min and the second for 1 h at
254 nm in a Rayonet photoreactor. Aliquots were analyzed by
NMR, with diphenylmethane added as internal standard. The
conversion of starting material to phenanthrene was determined
to be 38% for the sample irradiated for 30 min and 58% for the
sample irradiated for 1 h. Gas chromatographic analysis of
aliquots used 1,2,3,4 tetrahydonaphthalene as internal standard
and revealed the formation of the expected cyclopropane adduct.
The responses of the endo- and exo-methylchloronorcaranes
were calibrated with authentic samples. The yields of endo-
and exo-chloromethylnorcaranes were determined to be 83.4%
(endo/exo 1.0:1.9) for the 30-min photolysis and 77.5% (endo/
exo 1:2.3) for the 1-h photolysis, on the basis of reacted starting
material.
Produced by LFP of a Phenanthridene Precursor, Unless
Otherwise Noted
R
λmax (nm)
H
373.5
Cl
385-390
CD3
CH3
CH2Ph
CH2CH3
C(CH3)3
C(CH3)2
360
360.5 (360a)
369.5 (379a)
368
376 (376a)
370
a Values obtained from diazirine precursors.
pulsed 150-W He/Xe arc lamp. Transient spectra were recorded
using an EG and G Princeton Applied Research model 1460
Optical Multichannel Analyzer (OMA).
Stock solutions of precursors 8 were typically prepared with
an optical density ∼0.5 at 308 nm. A constant volume of the
stock solution was added to Suprasil quartz cuvettes. To each
cuvette was added varying amounts of pyridine. Solvent was
added to each cuvette to maintain a constant volume of sample
throughout the experiment. Samples were degassed by purging
with dry, oxygen-free argon for 3-4 min.
Infnite yields of pyridinium ylides 4 were obtained by
adjusting the concentration of precursors 8 so that the absorbance
of each solution closely match ((1% variation over the whole
set of starting compounds) at 308 nm. Pyridine concentration
was adjusted to be above the saturation point. Several transient
spectra were recorded at λmax for each ylide; the experiment
was repeated several times, and average results are presented
in Table 3 and Figure 4. The reproductibility of the Ay∞ values
was (10%.
Isooctane (Aldrich, 99.8%), cyclohexane (Aldrich, HPLC
grade), and cyclooctane (Aldrich, 99+%) were used as received.
Pyridine was distilled over CaH2 and stored over KOH pellets.
Chemical Analysis of Mixtures Formed on Photolysis of
Precursors 8. Solutions of precursors 8 (concentration ≈5 mM)
were prepared in isooctane in the presence of a standard
(dimethylterephthalate, concentration ≈5 mM). After degassing
and sealing the samples in quartz cuvettes, each was photolyzed
in a Rayonet photoreactor fitted with 300-nm bulbs. The
samples were irradiated for 50 min and analyzed by capillary
GC (Perkin-Elmer 8500 GC equipped with a flame ionization
detector). Phenanthrene was identified by comparison with an
authentic sample and was in all cases the main product of the
photolysis.
GC/MSD mass spectral analyses were performed on a HP-
5970B Mass Spectrometer, HP-5890 capillary Gas Chromato-
graph, and the 1,2 hydrogen shift from biradical intermediate 9
to rearranged structure 10 (Scheme 5) was analyzed. Besides
phenanthrene, another peak was detected after photolysis, the
area of the second product peak increases when going from
methylchlorocarbene to ethylchlorocarbene and finally to iso-
propylchlorocarbene. GC/MS spectra revealed that this peak
accounts for a product that has the same fragmentation pattern
as the rearranged product 10. Phenanthridene precursor shows
5-10% decomposition upon GC injection, which varies with
structure. We thus concluded that under irradiation the excited
state of the precursor partitions between carbene formation (and
concomitantly phenanthrene) and opens to a biradical that
rearranges to a substituted phenanthrene product. With meth-
ylchlorocarbene and ethylchlorocarbene, the chemical yields of
phenanthrene are high and the rearrangement pathway is minor.
With isopropylchlorocarbene, the amount of rearranged product
is close to the amount of phenanthrene produced (at least in
III. Results
Laser Flash Photolysis. The LFP system used at The Ohio
State University has been described previously.14 Samples were
exposed to the pulses of a 308-nm laser (XeCl, 17 nm, 150 mJ,
Lambda Physik LPX100 excimer laser) at right angles to a

Alkylhalocarbenes from Non-Nitrogenous Precursors
J. Phys. Chem. A, Vol. 102, No. 9, 1998 1511
TABLE 2: Dichlorocarbene and Alkylchlorocarbene Lifetimes in Various Solvents and Absolute Rate Constants of Their
Reaction with Pyridine. The Data Were Obained Using Precursors 8, Except Where Noted Otherwise
Cl-C¨ -Cl
Cl-C¨ -CH3
Cl-C¨ -CH2CH3
kpyr
(M-1 s-1
kpyr
τ
(ns)
kpyr
τ
(ns)
τ
(ns)
solvent
(M-1 s-1
)
(M-1 s-1
)
)
isooctane
cyclohexane
cyclooctane
7 × 109
3.9 × 109
∼10000
∼1250
7.6 × 109
3.8 × 109 c
340 (330a)
195 (10
∼40
8.5 × 109 b
5.3 × 109 d
4.2 × 109 e
7 (10c)
∼7
4.75 109
∼4.5
a Value obtained from diazirine precursor (see ref 15). b Average of previous values obtained with various alkylchlorocarbenes (see ref 15).
c Value obtained from diazirine precursor (see ref 15). d The value of kpyr in this solvent is not known. It will be assumed kpyr[in cyclohexane] )
kpyr [in isooctane]/1.6 by analogy with the variation of kpyr between these solvents observed with CMeCl. See ref 15. e The value of kpyr in this
solvent is not known. It will be assumed kpyr[in cyclooctane] ) kpyr [in isooctane]/2 by analogy with the variation of kpyr between these solvents
observed with CCl2 and CCBr2. See ref 15.
TABLE 3: Yields of Pyridine Ylides Obtained from
Diazirines and the Yields of Pyridine Ylides and
Phenanthrene Obtained from Phenanthridene Precursors.
All Values in a Column Are Relative to Methyl ≡ 1.0
substituent
R in 8
diazirinesa
phenanthridenesb
normalized yield
of phenanthrenec
Ay∞
Ay∞
tert-butyl
CD3
1.9
1.09
1
0.75
0.55
0.63
0.8-0.75
0.97 (0.97a)
1 (1a)
1
CH3
1
0.85
0.77
CH2Ph
CH2CH3
H
Cl
isopropyl
0.9 (0.85a)
0.65 (0.66a)
0.65
0.55-0.6
0.2
0.75
0.5
0.04
a Values obtained in methylene chloride. b Values obtained in
isooctane. c Yields of phenanthrene measured by GC after 50 min of
photolysis at 300 nm.
terms of peak area), indicating that both pathways are ap-
proximately equally favored.
Laser flash photolysis (308-nm, XeCl excimer laser) of 8a-h
fails to produce UV-vis active transient intermediates. How-
ever, LFP of 8a-h in the presence of pyridine leads to intensely
absorbing transient species (Figure 1, Table 1). These transients
are attributed to ylides 4a-g. This conclusion is supported by
the fact that the transient spectra are quite similar to those
obtained from the corresponding diazirine precursors. Further-
more, the formation of phenanthrene in good yields was
demonstrated by gas chromatographic analysis of solutions of
8 exposed to continuous 300-nm radiation.
In certain cases it was possible to resolve the formation of
the ylide. In these cases, the ylide was formed in an exponential
process, which yielded an observed rate constant, kobs, upon
analysis.
Figure 1. Transient spectra of some selected pyridinium-alkylchlo-
rocarbene ylides produced upon LFP of 8 (A, methylchlorocarbene;
B, benzylchlorocarbene; C, isopropylchlorocarbene) in isooctane
containing pyridine (A, 0.12 M; B, 0.6 M; C, 0.6 M) at 295 K. The
spectra were recorded over a window of 400 ns, approximately 1.2 µs
after laser pulse.
constant of reaction between the carbene and pyridine and the
lifetime of the carbene (Figure 2B).
The optical yield (Ay) as a function of pyridine concentration
was also measured. In the absence of pyridine, Ay is close to
zero and increases steadily as the concentration of pyridine is
increased (Figure 3A). At high concentrations of pyridine, the
yield of ylide is saturated (Ay∞); every carbene produced in the
laser pulse is then trapped prior to reaction with solvent. The
The observed rate constant kobs is the sum of kpyr[pyridine]
and k0, the pseudo first-order rate constant associated with the
sum of all processes consuming the carbene in the absence of
pyridine. As predicated, the plots of kobs versus [pyridine] are
found to be linear, thereby allowing us to extract the rate

1512 J. Phys. Chem. A, Vol. 102, No. 9, 1998
Robert et al.
Figure 2. (A) Rate of formation of ylide 4b, following LFP in
cyclohexane containing 0.124 M pyridine at 295 K. (B) Plot of kobs of
ylide 4b versus [pyridine].
quantum yield of ylide formation (φy) and the optical yield of
ylide are given in eq 1, where φc is the quantum yield of carbene
formation, kpyr is the rate constant of quenching by pyridine,
and k0 is a pseudo-first-order rate constant that is the sum of
all the processes consuming the carbene in the absence of
pyridine.4,5
Figure 3. (A) Yield of ylide 4c as a function of [pyridine] in
cyclooctane at 295 K. (B) Double-reciprocal treatment of the data of
part A.
φckpyr[pyridine]
∞
φy )
and Ay ) φyAy
(1)
(2)
k0 + kpyr[pyridine]
kτ
1
1
1
)
+
∞ kpyr[pyridine]
∞
Ay
φcAy
φcAy
Thus a plot of 1/Ay versus 1/[pyridine] is predicted and found
to be linear (Figure 3B). Division of the intercept by the slope
gives the ratio (kpyr/k0), which is equal to kpyrτ, where τ is the
lifetime of carbene in the absence of pyridine. After assuming
values of kpyr (by comparison to similar carbenes with known
rate constants1,4) it was possible to deduce lifetime values for
the short-lived (τ < 30 ns) carbenes.
The lifetimes of methylchlorocarbene and ethylchlorocarbene
in isooctane determined in this work are in good agreement with
previous results (Table 2).1,15 There is a tendency for the
lifetimes to decrease as solvent viscosity increases. Upon
changing from isooctane (viscosity equal to 0.49) to cyclooctane
(viscosity equal to 2.2),16 the methylchlorocarbene lifetime
decreases by a factor of approximately 8.5. Dichlorocarbene
is a very long-lived carbene.10 Its lifetime is close to 10 µs in
isooctane and is shortened by a factor of approximately 8.5 in
the more viscous solvent, cyclooctane. It is difficult to interpret
this result because the lifetime of long-lived carbenes can be
Figure 4. Plot of the yield of phenanthrene versus the yield of ylide
4 obtained from precursors 8b-h [(b) Cl, (c) CH3, (d) CD3, (e) CH2-
CH3, (f) CH2Ph, (g) CH(CH3)2, (h) C(CH3)3].
controlled by reaction with trace amounts of water and oxygen,
whose concentration can vary with solvent. Ethylchlorocarbene
is an extremely short-lived carbene.15 The variation of its
lifetime with solvent viscosity is small and may be due to
experimental error and variation in the value of the rate constant
of reaction with pyridine with solvent.

Alkylhalocarbenes from Non-Nitrogenous Precursors
J. Phys. Chem. A, Vol. 102, No. 9, 1998 1513
∞
Relative values of Ay are collected in Table 3 along with
as a function of alkene as in the original study involving
diazirines.1 This work is in progress and will be reported
shortly.17
the values obtained previously with the diazirine precursors and
the yield of phenanthrene obtained upon continuous photolysis
of 8a-g at 300 nm. In each column, the yield of the appropriate
quantity obtained with precursors of chloromethylcarbene is
defined as unity. Comparisons within a vertical column are
meaningful. Comparisons across a horizontal row are useful
only in a qualitative sense.
A plot of the yield of phenanthrene versus the yield of
pyridine ylides 4b-h in isooctane is shown in Figure 4. The
plot is linear with a slope of 0.65.
Acknowledgment. Support of this work by the National
Science Foundation (CHE-8814950, The Ohio State University,
and CHE-9122141 and CHE-9616388, University of New
Hampshire) is gratefully acknowledged.
References and Notes
(1) (a) Tomioka, H.; Hayashi, N.; Izawa, Y.; Liu, M. T. H. J. Am.
Chem. Soc. 1984, 106, 6, 454. (b) Liu, M. T. H.; Soundararajan, N.; Paike,
N.; Subramanian, R. J. Org. Chem. 1987, 52, 2, 4223. (c) Liu, M. T. H.;
Bonneau, R. J. Am. Chem. Soc. 1990, 112, 3915. (d) Liu, M. T. H.; Chishti,
N. H.; Tencer, M.; Tomioka, H.; Izawa, Y. Tetrahedron 1984, 40, 887. (e)
Liu, M. T. H.; Suresh, R. J. Org. Chem. 1989, 54, 486. (f) Liu, M. T. H.;
Bonneau, R.; Wierlacher, S.; Sander, W. J. Photochem. Photobiol. A: Chem.
1994, 84, 133. (g) Wierlacher, S.; Sander, W.; Liu, M. T. H. J. Am. Chem.
Soc. 1993, 115, 8943. (h) Bonneau, R.; Liu, M. T. H. J. Photochem.
Photobio. A: Chem. 1992, 68, 97. (i) Liu, M. T. H. Acc. Chem. Res. 1994
27, 287. (j) Bonneau, R.; Liu, M. T. H. J. Am. Chem. Soc. 1996, 118, 8098.
(k) Bonneau, R.; Liu, M. T. H. J. Am. Chem. Soc. 1996, 118, 8, 3829.
(2) Houk, K. N.; Rondan, N. G.; Marada, J. Tetrahedron 1985, 41,
1555.
(3) (a) Frey, H. M. J. Chem. Soc. 1962, 2293. (b) Kibby, C. L.;
Kistiakowsky, G. B. J. Phys. Chem. 1966, 70, 126. (c) Frey, H. M.; Stevens,
I. D. R. Proc. Chem. Soc. 1962, 79. (d) Frey, H. M.; Stevens, I. D. R. J.
Chem. Soc. 1962, 3865. (e) Mansoor, A. M.; Stevens, I. D. R. Tetrahedron
Lett. 1966, 1733. (f) Frey, H. M.; Stevens, I. D. R. J. Am. Chem. Soc. 1965,
87, 1. (g) Frey, H. M. J. Am. Chem. Soc. 1962, 84, 2647. (h) Frey, H. M.
AdV. Photochem. 1964, 4, 225. (i) Friedman, L.; Shechter, H. J. Am. Chem.
Soc. 1959, 81, 5512. (j) Frey, H. M.; Stevens, I. D. R. J. Chem. Soc. 1965,
3101. (k) Chang, K.-T.; Shechter, H. J. Am. Chem. Soc. 1979, 101, 5082.
(4) (a) Moss, R. A. Pure Appl. Chem. 1995, 67, 741. (b) Moss, R. A.;
Turro, M.-J. In In Kinetics and Spectroscopy of Carbenes and Biradicals;
Platz, M. S., Ed.; Plenum: New York, 1990; p 213. (c) Platz, M. S.; White,
W. R., III; Modarelli, D. A.; Celebi, S. A. Res. Chem. Intermed. 1994,
175-193. (d) Chidester, W.; Modarelli, D. A.; White, W. R., III; Whitt, D.
E.; Platz, M. S. J. Phys. Org. Chem. 1994, 7, 24-27. (e) Jackson, J. E.;
Platz, M. S. In Laser Flash Photolysis Studies of Ylide Forming Reactions
of Carbenes; Brinker, U., Ed.; AdV. Carbene Chem. 1994, 1, 89-160, JAI
Press: Greenwich, CT.
(5) (a) White, W. R., III; Platz, M. S. J. Org. Chem. 1992, 17, 2841.
(b) Platz, M. S. AdV. Carbene Chem. II, in press. (c) Platz, M. S.; Modarelli,
D. A.; Morgan, S.; White, W. R.; Mullins, M.; Celebi, S. J.; Toscano, J.
P.; Huang, H. Prog. React. Kinet. 1994, 19, 93.
(6) The quantum yield of diazo formation from simple alkyl diazirines
has recently been determined. See: Bonneau, R.; Liu, M. T. H. J. Am.
Chem. Soc. 1996, 118, 7229.
(7) (a) Bigot, B.; Ponec, R.; Sevin, A.; Devaquet, A. J. Am. Chem.
Soc. 1978, 100, 6573. (b) Mu¨ller-Remmers, P. L.; Jug, K.; J. Am. Chem.
Soc. 1985, 107, 7275. (c) Yamamoto, N.; Bernardi, F.; Bottoni, A.; Olivucci,
M.; Robb, M. A.; Wilsey, S. J. Am. Chem. Soc. 1978, 116, 2064.
(8) Intuition is supported by the calculations of Sulzbach and Hadad
(unpublished research at The Ohio State University).
IV. Discussion
The yield of ylide 4b-h varies by a factor of 47.5 from the
tert-butylchloro- to isopropylchlorocarbene systems using di-
azirine precursors. This large spread can be attributed to
hydrogen migration in a carbene-pyridine complex 6 or to the
RIES mechanism (Scheme 3).
The carbene-pyridine complex mechanism predicts that the
relative yields of ylide will be independent of precursor. As
shown in Table 3, this clearly is not the case. The yield of
pyridine ylides derived from precursor 8 varies by only a factor
of 3. The dependence of the yield (Ay∞)) of carbene from
precursor 8 on the strength of the C-H bond alpha to the
carbene center is much less sensitive than that realized from
the diazirine precursors. In fact, the tert-butyl system, which
lacks a migratable hydrogen alpha to the carbene center, gives
roughly twice as much carbene, per laser flash, as does methyl,
which has such a hydrogen when using the diazirine as the
precursor. Yet in the phenanthridene series, the methyl
compound 8c produces more carbene, per laser pulse, than does
the tert-butyl analogue 8h. The data demonstrate that carbene-
pyridine complexes are simply not responsible for the variation
in the yield of ylides realized in LFP experiments. If such
complexes are formed, they must collapse cleanly to form the
ylide.
The RIES mechanism (Scheme 3) predicts that the yield of
ylide will correlate with the yield of phenanthrene (precursors
8b-h) or nitrogen (diazirines). This correlation is not possible
to study with diazirines because nitrogen is a reaction product
of both decay routes of the biradical. In the phenanthridene
series, phenanthrene should be formed in a stoichiometric yield
in pathways that produce a carbene. Indeed, as shown in Figure
4, the yield of phenanthrene does correlate with the yield of
carbene. However, the correlation is not perfect. A perfect
correlation would have a slope of 1.0. The deviation may be
due to the back reaction of carbene with phenanthrene, to reform
precursor, the efficiency of which is a function of carbene
structure. The deviation may also be due to the aforementioned
decomposition of certain precursors during GC analysis.
This study provides no evidence for or against the existence
of a carbene-pyridine complex. However, this work provides
conclusive evidence against the existence of carbene-pyridine
complexes that can rearrange to alkenes in competition with
collapse to form ylides. This work does not discredit the
mechanism of Scheme 2, in which a carbene-olefin complex
partitions between rearrangement and the formation of cyclo-
propanes. However, it certainly reduces the urgency to embrace
this concept.
(9) Robert, M.; Toscano, J. P.; Platz, M. S.; Abbot, S. C.; Kirchhoff,
M. M.; Johnson, R. P. J. Phys. Chem. 1996, 100, 18426.
(10) Chateauneuf, J. E.; Johnson, R. P.; Kirchhoff, M. M. J. Am. Chem.
Soc. 1990, 112, 3217.
(11) Hartwig, J. F.; Jones, M., Jr.; Moss, R. A.; Lawrynowicz, W.
Tetrahedron Lett. 1986, 27, 5907.
(12) To our knowledge trimethylene type biradicals are not known to
thermally fragment to form carbenes and alkenes. See: Gajewski, J. J.
Hydrocarbon Thermal Isomerization; Academic Press: New York, 1981,
Chapter 3.
(13) Halton, B.; Officer, D. L. Aust. J. Chem. 1983, 36, 1167.
(14) Gritsan, N. P.; Zhai, H. B.; Yuzawa, T.; Karweik, D.; Brooke, J.;
Platz, M. S. J. Phys. Chem. A 1997, 101, 2833.
(15) (a) LaVilla, J. A.; Goodman, J. C. J. Am. Chem. Soc. 1989, 111,
6877. (b) Liu, M. T. H.; Bonneau, R. J. Am. Chem. Soc. 1990, 118, 8098.
(c) Liu, M. T. H.; Bonneau, R. J. Am. Chem. Soc. 1992, 114, 3604.
(16) Handbook of Chemistry and Physics, 1991-1992, 72nd ed.; Lide,
D. R., ed., CRC Press: Boca Raton, FL, 156-160.
A definitive proof or disproof of the carbene-olefin complex
postulate is available using 8f and repeating the product analysis
(17) Nigam, M.; Platz, M. S. Unpublished research at The Ohio State
University.