Formation of molecular iodine by photodissociation of CF 2 I 2 at 193 nm

Article abstract of DOI:10.1016/S0009-2614(00)00844-7

Difluorodiiodomethane CF₂I₂ has been studied by photofragment translational energy spectroscopy at 193 nm. The analysis of the TOF spectra revealed three decay pathways: a three-body decay CF₂I₂+hν→CF₂+I+I (75%) and two distinctly different molecular decays CF₂I₂+hν→CF₂+I₂ (21%) followed by a complete secondary decay of I₂ to I+I, and CF₂I₂+hν→CF₂+I₂ (4%) producing stable iodine molecules. The latter involves electronic product excitation consistent with CF₂ in its first excited singlet state. The measured anisotropic recoil distributions are in agreement with these processes and indicate absorption at 193 nm to prepare mainly an electronic state of A₁ symmetry.

Full text of DOI:10.1016/S0009-2614(00)00844-7

25 August 2000
Ž
.
Chemical Physics Letters 326 2000 366–374
www.elsevier.nlrlocatercplett
Formation of molecular iodine by photodissociation
of CF2 I2 at 193 nm
Heiner A. Scheld, Alan Furlan, J. Robert Huber)
Physikalisch-Chemisches Institut der UniÕersitat Zurich, Winterthurerstrasse 190, CH-8057 Zurich, Switzerland
¨
¨
¨
Received 25 May 2000
Abstract
Difluorodiiodomethane CF2I2 has been studied by photofragment translational energy spectroscopy at 193 nm. The
Ž
.
analysis of the TOF spectra revealed three decay pathways: a three-body decay CF2I2qhn
CF2qIqI 75% and two
CF2qI2 21% followed by a complete secondary decay of I2 to IqI,
CF2qI2 4% producing stable iodine molecules. The latter involves electronic product excitation
Ž
.
distinctly different molecular decays CF2I2qhn
Ž
.
and CF2I2qhn
consistent with CF2 in its first excited singlet state. The measured anisotropic recoil distributions are in agreement with these
processes and indicate absorption at 193 nm to prepare mainly an electronic state of A1 symmetry. q 2000 Elsevier Science
B.V. All rights reserved.
1. Introduction
decay produces also iodine in the spin-orbit excited
2
state P1r2
,
The rich photodissociation of difluorodi-
iodomethane CF2 I2 is distinctly excitation wave-
length-dependent. The changing efficiency of the
different competing decay channels has been studied
in some detail between 350 and 248 nm using
photofragment translational energy spectroscopy
CF2 I2 qhn CF2 IqI1r 2
2
Ž .
where D E I1r 2 yI3r 2 s7600 cmy1. At 304 nm
Ž
.
three competing decays become active which include
Ž .
channel 2 , the sequential three-body decay
CF2 I2 qhn CF2 IqI3r 2 CF2 qI3r 2 qI3r 2
Ž
. w
x
PTS 1,2 or the method of resonance enhanced
3
Ž .
Ž
.
multiphoton ionization time-of-flight REMPI-TOF
w x
w x
and a concerted three-body decay 1
CF2 I2 qhn CF2 qI3r 2 qI3r 2
3 . With an excitation energy corresponding to 351
nm the dissociation was found to be governed by the
4a
Ž
.
w x
radical or two-body decay 2 :
with the iodine atoms released in the ground state.
Using excitation at 248 nm the dissociation has been
shown to proceed exclusively via two competing
concerted three-body decays, the dominant one being
CF2 I2 qhn CF2 IqI3r 2
1
Ž .
with iodine emerging in the electronic ground state
2 P3r2. Increasing the energy to 337 nm this two-body
CF2 I2 qhn CF2 qI1r 2 qI3r 2
4b
Ž
.
)
Ž
.
and the minor one corresponding to decay 4a . Thus
with increasing excitation energy an apparently
Corresponding author. Fax: q41-1-6356838; e-mail:
jrhuber@pci.unizh.ch
0009-2614r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S0009-2614 00 00844-7

(
)
H.A. Scheld et al.rChemical Physics Letters 326 2000 366–374
367
smooth transition from a single two-body to a se-
quential three-body decay and finally to a concerted
and 608, where the scattering angle Q represents the
angle between the molecular beam and the detection
axis. For Q-308 the background signal from the
parent molecule was subtracted on a shot-to-shot
basis by leaving every second molecular beam pulse
unphotolyzed. The TOF distributions shown below
were corrected for the ion flight time through the
mass filter. The anisotropy measurements were car-
ried out at six different polarization angles between
the electric field vector and the detection axis. The
Ž
.
three-body decay has been observed. Reactions 4a
Ž
.
and 4b were deemed to be concerted based on the
analysis of the kinetic energy distributions of the
Ž
Ž .
fragments and their high recoil anisotropy b I s
Ž
.
. w x
q1.1; b CF2 sy0.8 at 248 nm 1 as well as on
w x
femtosecond pump-probe measurements 4 which
have shown CF2 and I atoms to be formed with
t;100 fs and no indication of an intermediate
species CF2 I. The molecular elimination channel
Ž
laser light was linearly polarized polarization degree
.
of 92"1% by directing the laser beam through a
CF2 I2 qhÕ CF2 qI2
5
Ž .
stacked-plates polarizer followed by a zeroth-order
half-wave retarder. A measurement was repeated
several times to minimize the influence of long-term
drifts and the runs were corrected for the laser power
fluctuations.
was not found to be active except in the femtosecond
excitation experiments. Under these conditions and
w
x
applying photoion and photoelectron detection 4,5 ,
the product species CF2 , I and I2 were detected with
the latter in an appreciable amount. In view of
CF2 I2 was obtained by the fluorination of te-
w x
previous results on CH2 I2 6 it was then speculated
Ž .
traiodomethane with mercury II fluoride and puri-
that not impurities but multiphoton-absorption pro-
cesses leading to high electronic states and subse-
quent dissociation to CF2 and I2 are the origin of the
formation of the stable iodine molecules. Therefore,
with the present experiments we attempt to provide
clear evidence for I2 production from higher elec-
tronic states of CF2 I2 by using the PTS method and
low-intensity nanosecond-pulse excitation at 193 nm.
In parallel the dissociation channels competing with
the molecular decay were elucidated.
fied by multiple distillation at reduced pressure, until
w x
the product showed a sharp boiling point of 368C 8 .
The contamination of CF2 I2 by I2 , as manifested by
a pale reddish color of the fresh sample, was kept
minimal by adding sodium thiosulfate to the liquid.
The I2 concentration in the sample was monitored
after the PTS measurements and found to be -0.4%
based on the characteristic I2 absorption bands
around 400 and )500 nm.
Expansions of premixed gas samples of 2% CF2 I2
seeded in helium carrier gas at a total stagnation
pressure of 400 mbar were used throughout this
2. Experimental
Ž .
study. The molecular beam velocity distribution f Õ
2
2
2
w
Ž
.
x
The experiments were carried out with
a
;Õ exp y ÕyÕS ra was determined before
photofragment translational energy spectrometer
and after each measurement using a chopper wheel
synchronized with the pulsed valve. To check and to
avoid cluster formation of CF2 I2 in the molecular
beam and multiphoton processes, we recorded TOF
spectra with varying CF2 I2 concentrations between
1% and 10% and with the laser power from 5 to 250
mJrpulse.
w x
which has been described in detail elsewhere 7 . It
consists of a rotatable, pulsed molecular beam source,
a 34.5-cm-long drift tube and a quadrupole mass
spectrometer for mass filtering and detection. The
experiments were performed at 193 nm with an ArF
Ž
.
excimer laser Lambda Physik Compex 102 oper-
ated at 20 Hz. The photolysis laser beam crossed the
molecular beam at a distance of 65 mm from the
piezoelectric pulsed nozzle. The laser beam was
focused to a spot size of 6=2 mm2 at the intersec-
tion with the molecular beam, providing a laser
fluence between 5 and 250 mJrcm2.
3. Results and analysis
Photofragments were collected at mre s 31
q
q
q
Ž
.
Ž
.
Ž
.
Ž .
The time-of-flight TOF fragment distributions
CF , mres50 CF2 , mres127 I and mre
q
Ž
.
were recorded at scattering angles Qs98, 188, 308
s254 I2 . The formation of additional energeti-

(
)
368
H.A. Scheld et al.rChemical Physics Letters 326 2000 366–374
Ž
.
cally accessible products F, CFI, and IF was
checked, but at none of the corresponding ion masses
was a signal found.
The solid line in Fig. 1 represents the best fit
w x
obtained by the forward convolution procedure 9 of
Ž
.
the kinetic energy distribution P ET for the stable
The TOF signal recorded at mres254 I2 ,
fragment pairs CF2qI2 emerging from the molecular
Ž .
Ž .
Ž
.
which is shown in Fig. 1, stems exclusively from
molecular iodine created by photodissociation of
CF2 I2. Other possible sources of I2 were carefully
examined, in particular CF2 I2 cluster formation
which was found to be absent under our experimen-
tal conditions. Moreover, the slight contamination of
the parent beam by I2 cannot contribute to the
photofragment signal because we recorded the TOF
spectra in the difference mode i.e. every second
molecular beam pulse was left unphotolyzed and
subtracted from the photoproduct signal.
channel 5 . The corresponding P ET is displayed
as dotted line in Fig. 2a. The kinetic energy distribu-
tion is centered at only ;25 kJrmol and is surpris-
ingly narrow with regard to the high available energy
Eavl of the fragments CF2qI2 which is given by
Eavl shnqEint yD0 CF yI
6
Ž .
Ž
.
2
2
where hns619 kJrmol and the internal energy Eint
of CF2 I2 could be neglected due to efficient cooling
Ž
.
Fig. 1. TOF distribution of I2 fragments mres254 measured at the scattering angle 98 after unpolarized excitation of CF2 I2 at 193 nm.
Ž
.
The solid line was calculated using the translational energy distribution P ET shown as a dotted line in Fig. 2a.

(
)
H.A. Scheld et al.rChemical Physics Letters 326 2000 366–374
369
Ž
.
A different distribution P ET is expected for ground
state fragments:
CF2 I2 qhn CF2 qI2
groundstateswith P Eb
7b
Ž
Ž
T ..
Ž
.
This decay is necessarily followed by a sponta-
neous secondary dissociation of the I2 fragments
I2
IqI. The TOF signals recorded at the ion
masses of CFq and CF2q were found to be identical
except for a shift of f5 ms due to the different ion
flight times. Hence the TOF signals of CFq, shown
in Fig. 3, stem from CF2 fragments cracked in the
ionizer. These fragments can be generated by differ-
ent mechanisms, namely as counterfragments to sta-
ble I2 , as counterfragments to unstable I2 , or by a
synchronous or sequential 3-body decay to CF2 qI
qI. The signal contributions from CF2 formed as
Ž
.
counterfragments to stable I2 Fig. 1 were calcu-
Ž
.
Ž
.
lated using the P ET of Fig. 2a dotted line and
are displayed as dotted lines in Fig. 3. This compo-
nent obviously accounts only for a small fraction of
the total signal; most CF2 fragments possess consid-
erably higher kinetic energies. While the assignment
of the Iq2 signal is unambiguous, we can only pro-
vide a tentative interpretation of the fast CF2 signal.
We first consider the possible sources of the
fastest CF2 fragments. The analysis revealed that
both a sequential and a synchronous three-body de-
cay are unable to account for the fastest CF2 species
due to energetic constraints. The fastest CF2 which
can be generated by a synchronous three-body decay
Ž
.
Fig. 2. Center-of-mass translational energy distributions P ET
for the three photodissociation channels identified for CF2 I2 at
Ž .
193 nm. Both curves displayed in panel
a correspond to a
molecular decay to CF2 q I2. While the solid line represents the
decay to unstable I2 which further decays to IqI before reaching
the detector, the dotted line yielded the best fit of the TOF
Ž .
b corre-
distribution of the decay to stable I2 products. Panel
sponds to a synchronous three-body decay to CF2 qIqI. The
arrow marks the available energy of the fragments CF2 qIqI.
Ž
and assuming the angle between the I atoms to be
.
fixed at 1208 , are indicated by the onset of the
dashed curves in Fig. 3 at 80 ms flight time for
fs98. A sequential decay releases even slower CF2
fragments. The only possibility for a greater momen-
tum to be imparted to CF2 at a fixed total kinetic
energy is achieved by a reduction of the angle
between the two departing I atoms which occurs if
in the supersonic expansion. The dissociation energy
Ž
.
Ž
D0 CF2-I2 s116 kJrmol is derived from D0 I-
.
w x
Ž
.
CF2 I s 215 " 8 kJrmol 2 , D0 I-CF2 ( 50
w x
Ž .
w x
kJrmol 2 and D0 I2 s149 kJrmol 10 . Accord-
Ž
the two I atoms depart as I2 followed in a second
ingly the resulting Eavl s 503 kJrmol is much
.
step by dissociation to IqI . The three-body decay
becomes then a two-body decay with a strongly
increased CF2 momentum. The thin solid lines in
Fig. 3 were calculated assuming this decay mode and
Ž
.
higher than the dissociation limit of I2 149 kJrmol .
The stability of nascent I2 fragments requires there-
fore CF2 andror I2 to be formed in electronically
excited states:
Ž
.
Ž
.
applying the P ET in Fig. 2a solid line . The
remaining signal is attributed to a synchronous
CF2 I2 qhn
CF2 qI2 excitedstateŽs. with P Ea
T ..
7a
Ž
Ž
Ž
.
Ž .
three-body decay according to reaction 3 . The

(
)
370
H.A. Scheld et al.rChemical Physics Letters 326 2000 366–374
Fig. 3. TOF distributions of CFq fragments mres31 measured at Qs98, 308 and 608 after unpolarized excitation of CF2 I2 at 193 nm.
Ž
.
The thin solid lines and the dotted lines are the signal contributions from CF2 formed in coincidence with unstable and stable I2 ,
Ž
.
respectively. They were obtained by forward convolution of the P ET s shown in Fig. 2a. The contribution from the synchronous
Ž
.
three-body decay, shown by dashed lines, was obtained from the P ET in Fig. 2b. The bold solid lines display the sum of all contributions.
three-body contribution, indicated by the dashed lines
in Fig. 3, was calculated with the kinetic energy
distribution of Fig. 2b and an assumed geometry of
the excited CF2 I2 molecule for which the two I
constitute a major fraction of the total signal at this
mass, as depicted by the dashed lines in Fig. 4. At
Ž
.
the large scattering angle of 608 bottom of Fig. 4
all other contributions become negligible. The signal
contribution from cracking of stable I2 and from the
secondary decay of unstable I2 are shown by dotted
w x
atoms enclose an angle of 1208 1 .
The spectra recorded at mass mres127 are
displayed in Fig. 4 and serve as confirmation for our
previous assignments including a three-body and two
molecular decay channels one to stable and one to
unstable I2. The signal contributions from the syn-
chronous three-body decay, calculated by momen-
tum-matching with the corresponding CF2 signal,
Ž
.
and thin solid lines, respectively. The primary P ET
of the unstable I2 covers a range of considerably
higher kinetic energies compared to the stable I2
fragments. However, the corresponding internal en-
Ž
.
ergies 190–490 kJrmol are still above the I2 disso-
ciation threshold and therefore all primary I2 frag-

(
)
H.A. Scheld et al.rChemical Physics Letters 326 2000 366–374
371
q
Ž
.
Fig. 4. Unpolarized TOF spectra of mres127 I at Qs98, 308 and 608. As in Fig. 3 the three signal contributions are shown by thin
solid, dotted and dashed lines, while the sum of all the contributions is shown by bold solid lines.
w
x
ments are expected to undergo spontaneous sec-
ondary dissociation. The secondary decay of these I2
species was calculated using the procedure given in
I2 in the ionizer 14 . With a pure I2 source we
found equal amounts of signals at mres127 and
254, implying that about 33% of the Iq2 ions are
cracked and detected as Iq. Taking this into account
we estimate that about 75"10% of the CF2 I2
molecules undergo a synchronous three-body decay,
and 25"10% of the molecules dissociate to CF2 qI2
w
x
Refs. 11–13 and assuming a symmetric distribution
of the kinetic energy of the IqI pair centered at 41
w
x
kJrmol FWHMs12 kJrmol together with a for-
ward–backward symmetric angular distribution of
the fragments. The thin solid lines in Fig. 4 represent
the signal contribution from these I2 fragments un-
dergoing secondary decay.
Ž .
5 . For the latter about 20% of the I2 molecules are
stable or remain stable during the flight time from
the reaction zone to the detector ;200 ms, cf.
Fig. 1 , while 80% originating from a different chan-
Ž
.
The branching between the synchronous three-
Ž .
Ž .
body decay 4 , and the molecular decays 5a and
nel undergo a secondary decay to IqI.
Ž
.
5b was estimated by comparing the relative weights
TOF spectra with linearly polarized laser light
of the three contributions in the TOF spectra of CFq
were recorded at mres31 CF and mres127
q
Ž
.
q
q
Ž
.
Ž
.
Ž
I
.
at Qs188 for six different polarization angles
Fig. 3 and I
Fig. 4 . Independent experiments
were carried out to determine the cracking pattern of
´. The anisotropy parameters b shown in Fig. 5

(
)
372
H.A. Scheld et al.rChemical Physics Letters 326 2000 366–374
Fig. 5. Anisotropy parameter b measured at mres127 and evaluated for different flight time intervals. The flight time intervals are shown
as horizontal lines together with the error bars.
.
0.1 at short flight times and decreases toward a
longer flight times Fig. 5b .
were obtained from a fit of the angle-dependent
signal intensities integrated over selected time inter-
Ž
.
Ž .
vals. The signal-to-noise ratio of the mres254 I2
spectra was insufficient for a b-determination. For I
Ž
.
atoms mres127 the anisotropy parameter shows
4. Discussion
a negative value of y0.35"0.1 over the entire
range of the TOF spectrum Fig. 5a . The b value
determined for the CF2 fragments is greatest 1.2"
Ž
.
The present set of TOF data shown in Figs. 1, 3
and 5 is quantitatively reproduced by an analysis
Ž

(
)
H.A. Scheld et al.rChemical Physics Letters 326 2000 366–374
373
Ž .
which includes the three-body decay 4 and the two
distinct molecular decays 5a and 5b . The different
origins of these molecular decays are most clearly
formed in a synchronous three-body decay, and from
an initial I–C–I angle of 1208, is expected to exhibit
Ž
.
Ž .
Ž .
bsy0.25. Our experimental values, b I sy0.35
Ž
.
Ž
.
Ž
evident by their different P ET displayed in
Fig. 2a. The formation of CF2 qI2 in their electronic
"0.15 and b CF2 s1.2"0.1 for the fastest CF2
fragments , are in qualitative agreement with these
.
Ž
.
ground states 5b results in high internal energies of
the fragments and as a consequence I2 becomes
subject to spontaneous secondary decay to IqI.
expectations. It is thus reasonable to assume that
excitation at 193 nm mainly involves preparation of
a totally symmetric excited state which also allows
dissociation along a molecular decay channel, in
contrast to states with a nodal plane between the two
I atoms such as B1 states encountered at the 248 nm
Ž
.
However along the competing reaction path 5a ,
leading to electronic excitation of either CF2 or I2 , a
sufficiently high fraction of Eavl is channeled into
electronic excitation so that the I2 fragments remain
stable at least on the time scale of the experiment.
After subtracting the measured kinetic energies of
w
x
Ž
.
photolysis 1–3 . The reduction of b CF2 from the
limiting value of 2.0 is most likely caused by an
admixture of another electronic state of different
symmetry at 193 nm excitation since a reduction due
Ž
.
the fragments CF2 qI2 stable , as given in Fig. 2a
from Eavl, the internal energy including electronic
excitation is found to be in the range 459–503
kJrmol. The possible excited product states are
w
x
to a long excited state lifetime 17 is in contradic-
tion to the femtosecond experiments where all pho-
toinduced C–I bond scissions were found to be very
1
3
w
x
w
x
therefore the B1 or B1 states of CF2 15 , or the
fast 4,5 .
The surprising result that stable I2 is formed after
irradiation of CF2 I2 at 193 nm, but not at longer
3
metastable Pq0u state of I2 with energies of 446,
237, and 482 kJrmol above the electronic ground
states, respectively. Considering further the energet-
ics and lifetimes of these states we favor an assign-
w
x
wavelenths 1–3 , is attributed to excitation of an A1
electronic state and the formation of CF2 fragments
in the first excited singlet state which absorbs most
of the large available energy. At longer wavelengths
the molecule is prepared in a state of B1 symmetry
characterized by a repulsive potential between the
two I atoms thus preventing I2 formation. More
specifically, providing retention of C2 Õ symmetry
Ž1
.
ment for which the first singlet state of CF2 B1 is
Ž
.
formed in the molecular decay 5a . In this case the
internal energies of the I2 molecules are (45
kJrmol and thus not sufficient for a spontaneous
secondary decay to I atoms D0 s149 kJrmol 10 .
Based on this analysis and interpretations the TOF
spectra allow us to estimate the yields of the three
decay channels which are 75"10% for the syn-
Ž
w x.
Ž
.
cf. Fig. 1 during dissociation, a B1 state does not
correlate with stable and energetically accessible
states of I2. On the other hand the synchronous
three-body decay is supposed to proceed from a B1
state in analogy to this decay channel at 248 nm
Ž .
chronous three-body decay 4 , 21"10% for the
decay to unstable I2 with subsequent secondary de-
Ž
.
cay 5b , and about 4% for the formation to stable I2
Ž
.
w
x
5a .
1–3 . At 193 nm this state is then populated either
by direct B1 A1 excitation or via curve crossings
w x
18 from the initially prepared A1 state.
The recoil anisotropies of the fragments CF2 and I
Ž
.
Fig. 5 can provide useful information in support of
the mechanistic picture gained from the spectral
analysis. This, however, requires knowledge of the
excited state of CF2 I2 prepared with 193 nm excita-
tion. In analogy to the findings of the closely related
compound CH2 I2 we assume this state to be of A1
Acknowledgements
w
x
symmetry 16 . Thus an A1
A1 electronic transi-
tion implies that the transition dipole moment lies
perpendicular to the lines connecting the I atoms and
the F atoms. In the case of a prompt dissociation of
CF2 I2 an anisotropy with bs2.0 is then expected
for the CF2 and I2 fragments, while the I fragments
This work was supported by the Swiss National
Science Foundation and the Alfred Werner Legat.
The authors thank Rolf Pfister for the synthesis of
difluorodiiodomethane and Dr. Robert T. Carter for
critically reading the manuscript.

(
)
374
H.A. Scheld et al.rChemical Physics Letters 326 2000 366–374
w x
9
R.K. Sparks, K. Shobatake, L.R. Carlson, Y.T. Lee, J. Chem.
References
Ž
.
Phys. 75 1981 3838.
w x
1
w
x
E.A.J. Wannenmacher, P. Felder, J.R. Huber, J. Chem. Phys.
10 K.P. Huber, G. Herzberg, Molecular Spectra and Molecular
Structure IV. Constants of Diatomic Molecules, Van Nos-
trand, New York, 1979.
Ž
.
95 1991 986.
w x
2
Ž
.
G. Baum, P. Felder, J.R. Huber, J. Chem. Phys. 98 1993
1999.
w
x
11 X. Zhao, G.M. Nathanson, Y.T. Lee, Acta Physico-Chim.
w x
3
Ž
.
K. Bergmann, R.T. Carter, G.E. Hall, J.R. Huber, J. Chem.
Phys. 109 1998 474.
Sinica 8 1992 70.
Ž
.
w
x
Ž
.
12 E.J. Hintsa, X. Zhao, Y.T. Lee, J. Chem. Phys. 92 1990
w x
4
W. Radloff, P. Farmanara, V. Stert, E. Schreiber, J.R. Huber,
Chem. Phys. Lett. 291 1998 173.
2280.
Ž
.
w
w
w
x
13 P. Felder, Habilitation, University of Zurich, 1993.
¨
w x
5
x
Ž
.
P. Farmanara, V. Stert, W. Radloff, E. Schreiber, J.R. Huber,
Soc. Opt. Quant. Electr. Lasers’98 1999 .
14 A. Furlan, G.E. Hall, J. Chem. Phys. 109 1998 10390.
Ž
.
x
15 T.K. Minton, P. Felder, R.C. Scales, J.R. Huber, Chem.
w x
6
Ž
.
U. Marvet, E.J. Brown, M. Dantus, Phys. Chem. Chem.
Phys. Lett. 164 1989 113.
Ž
.
w
x
Phys. 2 2000 885, and refs. therein.
16 M. Kawasaki, S.J. Lee, R. Bersohn, J. Chem. Phys. 63
w x
Ž
.
Ž
.
7
w x
8
P. Felder, Chem. Phys. 143 1990 141.
S. Elsheimer, J.W.R. Dolbier, M. Murla, K. Seppelt, G.
Paprott, J. Org. Chem. 49 1984 205.
1975 809.
w
w
x
Ž
.
17 S. Yang, R. Bersohn, J. Chem. Phys. 61 1974 4400.
Ž
.
x
Ž
.
18 M. Klessinger, Angew. Chem. Int Ed. Engl. 43 1995 549.