Direct formation of CH2 (b 1B1) in the near-UV photodissociation of diazirine

Article abstract of DOI:10.1016/S0009-2614(98)00370-4

A prompt, long-lived, and red fluorescence has been observed in the photodissociation of the jet-cooled diazirine (H₂CN₂) excited at the origin (322.96 nm) of the S₁←S₀ transition. The fluorescence decays (~16 μs) and range (>575 nm) indicate that the emitting species is the electronically excited singlet methylene, CH₂ (b ¹B₁). The linear dependence of the fluorescence intensity on the photolysis laser power strongly supports that the fluorescing CH₂ (b ¹B₁) is produced directly from the first excited singlet state of diazirine. The observation that the fluorescence appears in the visible wavelength region indicates that the CH₂ (b ¹B₁) fragments are highly vibrationally excited.

Full text of DOI:10.1016/S0009-2614(98)00370-4

29 May 1998
Ž
.
Chemical Physics Letters 288 1998 828–832
1
˜
ž
/
Direct formation of CH2 b B1 in the near-UV photodissociation
of diazirine
)
Soon-Mi Lim, Taek-Soo Kim, Goo-Il Lim, Sang Kyu Kim, Young S. Choi
Department of Chemistry, Inha UniÕersity, Incheon 402-751, South Korea
Received 10 February 1998; in final form 24 March 1998
Abstract
A prompt, long-lived, and red fluorescence has been observed in the photodissociation of the jet-cooled diazirine
Ž
.
Ž
.
Ž
.
Ž
H2CN2 excited at the origin 322.96 nm of the S1§S0 transition. The fluorescence decays ;16 ms and range )575
1
˜
Ž
.
.
nm indicate that the emitting species is the electronically excited singlet methylene, CH2 b B1 . The linear dependence of
the fluorescence intensity on the photolysis laser power strongly supports that the fluorescing CH2 b B1 is produced
1
˜
Ž
.
directly from the first excited singlet state of diazirine. The observation that the fluorescence appears in the visible
1
˜
Ž
.
wavelength region indicates that the CH2 b B1 fragments are highly vibrationally excited. q 1998 Elsevier Science B.V.
All rights reserved.
1. Introduction
age of a highly strained three-membered ring. In
addition, it is small enough to be tractable with the
state-of-the-art theoretical methods. It seems like that
the explosive nature of diazirine has prevented it
from becoming the topic of recent experimental stud-
ies.
Ž
.
Diazirine H2CN2 , which contains a three-mem-
bered CN2 ring, has a strong p) §n transition in
Ž
.
the near-ultraviolet UV wavelength region, fol-
lowed by the decomposition into CH2 and N2 prod-
w x
ucts 1 . Due to its structural uniqueness and the
w
x
Recently Platz and coworkers 5,6 investigated
the photochemistry of several substituted diazirine
compounds in liquid solutions. They observed fluo-
rescence and found that its intensity decreases as the
temperature increases, showing that electronically
excited diazirines decay through non-radiative pro-
cesses with an activation barrier. In contrast to many
recent experimental studies for the substituted di-
strong absorption band, the photochemistry of di-
azirine has been both extensively and intensively
w x
w
x
studied 1 since its first preparation 2,3 .
However, diazirine has not drawn that much at-
w x
tention of recent state-resolved studies 4 on the
photodissociation dynamics despite that its photo-
chemistry brings some interesting dynamical fea-
tures; it produces easily detectable simple fragments
whose spectroscopic data are well documented and
the photochemical decomposition involves the break-
w
x
azirines, only a few theoretical studies 7–9 have
been reported. In 1978, Bigot et al. first studied
potential energy surfaces of the ground and low-lying
excited states with the SCF method using STO-3G
)
w x
basis set 7 . Later, Muller–Remmers and Jug inves-
Corresponding author. E-mail: yschoi@dragon.inha.ac.kr
¨
0009-2614r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S0009-2614 98 00370-4

(
)
S.-M. Lim et al.rChemical Physics Letters 288 1998 828–832
829
Ž
.
tigated the photochemical reaction mechanism of the
electronically excited diazirine with a SINDO1
color filter Schott OG-530 , and then detected with a
PMT. Two kinds of PMT were employed; for the
spectral distribution measurement, a red-sensitive
w x
method 8 . Yamamoto et al. has reported the most
Ž
.
detailed theoretical study on the photodissociation
side-view PMT HamamastuR928 was used and a
Y
w x
Ž
.
process of diazirine 9 . By using the MC-SCF
2 diameter head-on-type PMT HamamatsuH1161 ,
which provides much larger viewing area, was used
for the lifetime measurement. The time evolution of
the fluorescence signal was obtained by accumulat-
ing the PMT signal with a 300 MHz digital oscillo-
method with a 6-31G) basis set, they predicted that
the diazirine molecule excited to the S1 state decom-
poses into N2 q1CH2 via two different singlet paths
with little barriers.
Ž
.
Here, we synthesized the diazirine molecule and
scope LeCroy 9450A . The lifetime was deduced by
fitting the decay curve with a least-squares fitting
routine. Since the lifetime of the fluorescence was
investigated its photochemistry in the collisionless
1
1
Ž
.
condition. In an attempt to detect the CH2 a A
˜
1
Ž
.
photofragments, which was expected from the known
very long vide infra , it was not possible to obtain
the dispersed fluorescence spectrum with a mono-
chromator equipped with a PMT and a gated integra-
tor. Instead, the spectral distribution was roughly
estimated from the intensity change of the PMT
signal by varying the kind of the color filters in front
of the PMT.
w
x
photochemistry 1,7–9 , we observed a long-lived
and red fluorescence in the absence of the probe
1
1
˜
laser tuned to the b B §a A transition of the
˜
1
1
1CH2. A temporal and spectral characterization of
the fluorescence has led to the conclusion that the
1
˜
Ž
.
electronically excited CH2 b B1 is formed in one-
photon photodissociation of diazirine with the near-
UV light.
The shot-to-shot fluctuation of the excitation laser
pulses was monitored with a pyroelectric joule meter
Ž
.
MolectronP5-01 . The energy of the excitation pulse
was kept at approximately 1.0 mJrpulse and the
laser beam was mildly collimated to 1.0 mm in
diameter at the interaction region with a 1-m focal
length quartz lens. The optogalvanic spectrum of a
Ne hollow cathode lamp provided a frequency cali-
bration for the dye laser with an accuracy of "0.2
2. Experimental
The diazirine sample was synthesized following
w
x
the procedure reported previously 10 . The diazirine
gas was mixed with He gas in a 4-l stainless steel
cylinder. The diazirine was about 3% in the total
pressure of 6 atm. Diazirine in the mixture was fairly
stable and did not show any sign of decomposition
after the storage for 1 month at room temperature.
The mixture, kept at the pressure of 1 atm, was
expanded through a 0.5 mm diameter nozzle, which
was driven with a home-made driver. The output
cmy1
.
3. Results and discussion
As expected from the broad absorption spectrum
w
x
for the S1 §S0 transition 11 , no fluorescence from
Ž
Ž
.
beam from a pulsed dye laser Lambda Physik
the electronically excited diazirine S1 was ob-
served, which should appear in the UV or blue
.
SCANmate 2E pumped with a Q-switched Nd:YAG
Ž
.
w
x
laser Spectra-Physics GCR-150 was frequency-
visible region as in other diazirines 5,6,12,13 . In-
stead, a prompt and long-lived emission in the red
was observed when diazirine was exited at the origin
doubled in a KDP crystal. The phase-matching angle
of the crystal was set by an autotracker Inrad I
Ž
.
Ž
.
controlled by a personal computer. The second har-
monic output, which was separated from the funda-
mental with three dichroic mirrors, was directed
through the side arms, exciting the cooled diazirine
molecules in the jet expansions. The distance from
the nozzle to the excitation laser beam was 25 mm.
Fluorescence from the photofragments was collected
with a 2Y diameter fr2 quartz lens, filtered with a
band 322.96 nm of the S1 §S0 transition in colli-
sionless supersonic jet expansions.
In order to identify the species giving the prompt
fluorescence, the temporal decay of the total emis-
sion was measured. Fig. 1 shows the decay curve
obtained with an OG 530 color filter in front of the
PMT, which transmits most of the fluorescence while
blocking the scattered light of the excitation. The

(
)
830
S.-M. Lim et al.rChemical Physics Letters 288 1998 828–832
1
˜
Ž
.
b B1 in a collisionless condition. The lifetime
strongly depends on the excited quantum state and
Ž
. Ž
.
varies from 4 to 10 ms for the 0,14,0 , 0,15,0 ,
Ž
.
Ž
.
w x
0,16,0 and 0,17,0 vibrational levels 14 . In gen-
eral, the lower vibrational levels tend to have the
longer lifetimes. Therefore, the observed lifetime of
1
˜
Ž
.
16 ms is consistent with that of CH2 b B1 in the
case where most of the fragments are populated in
Ž
.
Ž
.
the 0,14,0 andror in the lower levels vide infra .
The red spectral distribution of the fluorescence is
also in a good agreement with the emission spectrum
1
˜
Ž
. w x
of the laser-excited CH2 b B1 15 and that of the
1
˜
Ž
.
photochemically generated CH2 b B1 in the 193
nm photodissociation of ketene 16 .
There are two possible mechanisms for the gener-
ation of CH2 b B1 in the photodissociation of
CH2 N2: one is the direct formation of CH2 b B1
from the one-photon excited CH2 N2 S1 and the
other from the highly excited CH2 N2 Sn, n02 by
Fig. 1. The temporal decay of the prompt and red fluorescence
observed in the photodissociation of diazirine in the near-UV
w
x
Ž
.
wavelength 322.96 nm . The lifetime was estimated to be 16 ms.
Some deviation from the single-exponential decay is noticeable,
which is expected for the species with long lifetimes; a large
portion of the emitting species travels out of the viewing zone.
1
˜
Ž
.
1
˜
Ž
.
Ž
.
Ž
.
decay curve follows a single-exponential function in
general but shows some deviation at the long delay
time. The deviation is likely due to the fact that the
emitting species travels out of the viewing zone of
the PMT with a rate faster than the inverse of its
lifetime. A least-squares fit to the experiment gives a
lifetime of 16 ms with an uncertainty of "0.2 ms.
The second key information required for the identifi-
cation of the emitting species is the spectral distribu-
tion of the emission. As stated in Section 2, we
failed to disperse the emission with a monochro-
mator and our current detection system. As an alter-
native, we measured the intensity of the PMT signal
with varying the cut-off wavelength by using several
long-pass color filters in front of the PMT. The
dispersed spectrum was then roughly estimated from
the intensity variation with the cut-off wavelength
two- or three-photon absorption.
CH N S qhn™CH N S
2 Ž
.
2 Ž
1
.
2
0
2
1
1
™CH2 b B qN X Sq
1
˜
Ž .
Ž
.
2 ž /
1
g
CH N S qnhn ns2 or 3
2 Ž .
.
Ž
2
0
1
1
™CH N S ™CH b B qN X Sq
˜
2 Ž
.
2 Ž
.
2 ž
/
2
n
1
g
2
Ž .
The simplest way to distinguish these two mecha-
Ž
.
change Fig. 2 . The broken line is the estimated
spectral distribution of the emission. In spite of the
low resolution of this method, it is clear that the
emission is most intense at the wavelengths longer
than 575 nm. It should be mentioned here that the
Ž
.
spectral sensitivity of the PMT HamamatsuR928
significantly decreases at the wavelength longer than
750 nm and thus the intensities at these wavelengths
Fig. 2. The spectral distribution of the prompt emission in the
photodissociation of diazirine in the near-UV wavelength. The
broken line shows an estimated spectrum and the procedure to
are underestimated.
The fluorescence lifetimes of CH2 b B1 have
been measured by many groups, but exact times are
1
˜
Ž
.
obtain this spectrum is described in the text. The spectrum is very
w
x
1
still controversial 14 . Recently, Moore and cowork-
ers have carefully measured the lifetimes of CH2
˜
Ž
.
similar to that of the CH2 b B1 fragments generated in the 193
w
x
nm photodissociation of ketene 16 .

(
)
S.-M. Lim et al.rChemical Physics Letters 288 1998 828–832
831
there is a well-defined reaction pathway leading the
system toward the singlet diazomethane conical in-
tersection exclusively. It is interesting to note that
1
˜
Ž
.
the formation of CH2 b B1 has been postulated
1
before on the photochemical reactions of CH2 to
explain the higher reactivity of photochemically gen-
erated carbene compared to the thermally generated
w
x
carbene 17 . Our observation suggests that this pos-
tulate may be true, but more data for other diazirines
would be needed to confirm this postulate.
The observation that the emission spectrum of the
1
˜
Ž
.
CH2 b B1 fragments appears in the visible region
1
˜
Ž
.
indicates that the CH2 b B1 fragments are pro-
1
˜
Ž
.
Fig. 3. The laser power dependence of the CH2 b B1 fluores-
duced with a significant vibrational excitation. The
cence intensity. The fluorescence intensity linearly depends upon
1
1
˜
Ž
. Ž
.
0,14,0 – 0,0,0, band of the b B la A transition
˜
1
1
1
˜
Ž
.
the laser power, confirming that the formation of CH2 b B1 is
the one-photon process.
spans from 588 to 615 nm for the rotational levels up
w
x
to Js8 14 . Hence the short wavelength limit of
575 nm of the emission indicates that the CH2
1
˜
Ž
Ž
.
b B1 fragments are populated at least up to the
0,14,0 level. Such a high vibrational energy content
.
nisms is to measure the laser power dependence of
the fluorescence signal intensity. The first Scheme
might come from the large difference in the H–C–H
1
˜
Ž .
Ž
.
Ž
.
1 would give a linear dependence of the fluores-
bond angles of the CH2 N2 S1 and CH2 b B1
w x
w x
cence on the laser power, while the second Scheme
Ž .
radicals, about 1258 9 and 1808 18 , respectively.
That is, as the dissociation proceeds, the bending
potential experiences a strong torque along the bend-
ing angle, and this gives highly vibrational excitation
of the fragment. A quantitative measurement of the
vibrational energy content requires a high-resolution
dispersed fluorescence spectrum, which is under
progress with a photon counting detection system in
this group, and thus detailed discussion on the dy-
namical implication of the vibrational energy excita-
2 would give a quadratic or cubic dependence with
two or three photons absorbed, respectively. As
Ž
.
shown in Fig. 3, the ln fluorescence intensity y
Ž
.
ln laser power plot gives a slope of 0.93, and this
Ž .
strongly supports the Scheme 1 . That is, the forma-
1
˜
Ž
.
tion of CH2 b B1 is from the one-photon process.
Another possibility is the excitation of the CH2
1
Ž
.
a A formed in one-photon photodissociation of
˜
1
diazirine to the c 1A state by the photolysis laser
˜
1
light. Even in this case, however, the power depen-
dence should be quadratic, which is contrary to the
observation. Furthermore no fluorescence has been
tion is deferred to the forthcoming paper.
1
Ž
.
The existence of the CH2 a A fragment has
˜
1
1
Ž
.
been also checked since energetically CH2 a A
can be surely formed in the photodissociation of
Ž .
diazirine. We tried to detect the CH2 a A frag-
˜
1
ment by exciting its 0,15,0 – 0,0,0 band with a
probe laser light, but have failed because it was
impossible to distinguish the emission due to laser-
induced fluorescence by the probe laser from the
prompt emission obtained with the photolysis laser
˜
1
reported from the c 1A state. It is convincing, there-
˜
1
1
fore, that the electronically excited singlet methylene
1
˜
Ž
.
Ž
. Ž
.
radical, CH2 b B1 is directly formed in the decom-
position of CH2 N2 excited to its first singlet excited
Ž
.
state S1 .
1
˜
Ž
.
The significant generation of CH2 b B1 in the
photodissociation of diazirine in the near-UV has not
w
x
been expected in the former studies 1,7–9 . In a
only. Furthermore, the broad quantum state distribu-
1
w x
Ž
.
recent ab initio study, Yamamoto et al. 9 have
tion of the CH2 a A fragments would give very
˜
1
mentioned the possibility of the existence of the
reaction pathway from CH2 N2 S1 to CH2 b B1
and N2 products. However, they concluded that the
contribution of this channel would be minor because
weak signal in the laser-induced fluorescence tech-
nique, in which only a single quantum state is probed
at a given excitation wavelength. Therefore, this
observation does not exclude the non-negligible yield
1
˜
Ž
Ž
.
.

(
)
832
S.-M. Lim et al.rChemical Physics Letters 288 1998 828–832
1
Ž
.
of the CH2 a A fragment. An ingenious method
References
˜
1
1
Ž
.
which can simultaneously measure the CH2 a A
˜
1
w x
1
Ž
.
H.M. Frey, Adv. Photochem. 4 1966 225, and references
cited therein.
1
˜
Ž
.
and CH2 b B1 fragments should be devised to give
an unambiguous answer to this question.
In conclusion, we experimentally found that the
electronically excited singlet methylene, CH2 b B1
w x
Ž .
E. Schmitz, R. Ohme, Tetrahedron Lett. 1961 612.
S.R. Paulsen, Angew. Chem. 72 1960 781.
H. Sato, Photodissociation of Simple Molecules in the Gas
2
w x
3
Ž
.
1
w x
4
˜
Ž
.
Ž
.
Phase Bunshin Publ., Tokyo, 1992 .
is generated in the photodissociation of diazirine
w x
5
D.A. Modarelli, S. Morgan, M.S. Platz, J. Am. Chem. Soc.
Ž
.
H2CN2 by the S1 §S0 transition. The long life-
Ž
.
114 1992 7034.
time and red spectral range of the fluorescence
w x
6
Ž
.
D.A. Modarelli, M.S. Platz, J. Am. Chem. Soc. 115 1993
470.
1
˜
Ž
.
strongly suggest that the CH2 b B1 fragment is
highly vibrationally excited. High-resolution dis-
persed spectrum of the fluorescence from CH2
b B1 fragments, as well as measurements of dy-
namical quantities such as the translational and inter-
nal energy contents of N2 fragments or branching
w x
7
B. Bigot, R. Ponec, A. Sevin, A. Devaquet, J. Am. Chem.
Soc. 100 1978 6575.
Ž
.
w x
8
Ž
.
P.L. Muller-Remmers, K. Jug, J. Am. Chem. Soc. 107 1985
¨
7275.
1
˜
Ž
.
w x
9
N. Yamamoto, F. Bernando, A. Bottoni, M. Olivucci, M.A.
Robb, S. Wisley, J. Am. Chem. Soc. 116 1994 2064.
Ž
.
1
1
w
w
x
Ž
.
10 R. Ohme, E. Schmitz, Chem. Ber. 97 1964 297.
˜
Ž
Ž
.
.
ratios of the CH2 a A and CH2 b B1 channels
˜
1
x
Ž
.
11 L.C. Robertson, J.A. Merritt, J. Mol. Spectrosc. 19 1966
would be helpful for the better understanding of the
dissociation dynamics of diazirine in the S1 state.
Work is underway along this direction.
372.
w
w
w
w
x
12 T.-S. Kim, Y.S. Choi, I. Kwak, J. Photochem. Photobiol. A.
Ž
.
Chem. 108 1997 123.
x
13 T.-S. Kim, S.K. Kim, Y.S. Choi, I. Kwak, J. Chem. Phys.
Ž
.
107 1997 8719.
x
14 I. Garcia-Moreno, E.R. Lovejoy, C.B. Moore, J. Chem. Phys.
Acknowledgements
Ž
.
98 1993 873.
x
15 D. Feldmann, K. Meier, R. Schmiedl, K.H. Welge, Chem.
This work was supported by the Ministry of
Education of Korea through the 1996–1997 Basic
Science Research Institute Program Project No.
Ž
.
Phys. Lett. 60 1978 30.
w
w
x
Ž
.
16 J. Ruiz, M. Martin, Chem. Phys. Lett. 226 1994 300.
x
Ž
.
17 K.-T. Chang, H. Shechter, J. Am. Chem. Soc. 101 1979
Ž
5082.
.
BSRI-96-3432 and the Korean Science and Engi-
w
x
18 G. Herzberg, Molecular Spectra and Molecular Structure III.
Ž
.
neering Foundation Project No. 95-0501-04-01-3 .
The authors gratefully acknowledge the support of
the 1996 Inha University Research Fund.
Electronic Spectra and Electronic Structure of Polyatomic
Ž
.
Molecules Van Nostrand Reinhold, New York, 1996 p.
583.