Reactions of photogenerated fluorine atoms with molecules trapped in solid argon 4.* spectroscopic characteristics of β-c2h2f. radicals generated in reactions of mobile f atoms with c2h2 molecules tra

Article abstract of DOI:10.1023/A:1011352831808

Reactions of mobile fluorine atoms with C₂H₂, C₂D₂, and C₂HD molecules in solid argon were studied by ESR and IR spectroscopic techniques. Highly resolved ESR spectra of the stabilized radicals CHF=

Full text of DOI:10.1023/A:1011352831808

Russian Chemical Bulletin, International Edition, Vol. 50, No. 6, pp. 989—995, June, 2001
989
Reactions of photogenerated fluorine atoms with molecules
trapped in solid argon
•
4
.* Spectroscopic characteristics of β-C H F radicals generated in reactions
2
of mobile F atoms with C H molecules trapped in solid argon
2
2
2
E. Ya. Misochko,a« I. U. Goldschleger,a A. V. Akimov,a and C. A. Wight b
aInstitute of Problems of Chemical Physics, Russian Academy of Sciences,
1
42432 Chernogolovka, Moscow Region, Russian Federation.
Fax: +7 (096) 576 4009. E-mail: misochko@icp.ac.ru
bDepartment of Chemistry, University of Utah,
Salt Lake City, Utah 84112, USA.
Fax: +1 (801) 585 3207. E-mail: wight@chem.utah.edu
Reactions of mobile fluorine atoms with C H , C D , and C HD molecules in solid argon
2
2
2
2
2
were studied by ESR and IR spectroscopic techniques. Highly resolved ESR spectra of the
•
•
•
•
stabilized radicals CHF= CH, CDF= CD, CHF= CD, and CDF= CH were obtained for
the first time. Isotropic hyperfine constants on fluorine and proton nuclei were measured. It
•
was found that the radicals formed in the reaction F + C H correspond to the cis-β-C H F
2
2
2
2
isomer. A comparison of the measured HFC constants with the values calculated by modern
quantum-chemical methods allows the identification of the isomeric form of the radical,
whereas vibrational analysis of the IR absorption spectra gives unreliable results. The
•
calculation of the energy of the radical isomers predicts that cis-β-C H F is more stable than
2
2
•
–1
trans-β-C H F by ∼ 1.0 kJ mol
.
2
2
Key words: cryochemistry, ESR spectroscopy, IR spectroscopy, fluorine, acetylene,
radicals.
The reactions of F atoms with hydrocarbon mol-
ecules in the gas phase are well studied.2 The high
reactivity of radicals formed in primary reactions im-
pede their detection in the gas phase and determination
of their fundamental characteristics. For the spectro-
scopic study of radical products of chemical reactions,
the most successful is the matrix trapping method,
whose simplest variant is fast freezing of the gas phase
reaction products to cryogenic temperatures and sub-
sequent spectroscopic studies under steady-state con-
ditions.
ESR and IR absorption spectra of most free radicals
obtained by this method are commonly recognized.3
However, this method is not always appropriate for the
reliable identification and complete characterization of
stabilized intermediates. Difficulties in identification are
related to the presence of various products of secondary
chemical reactions in the gas phase and stabilization of
products in structural traps of the solid matrix resulting
in "matrix shifts" of IR absorption bands and impossibil-
ity of detection by standard equipment of low-frequency
vibrations (ν ≤ 300 cm–1) necessary for the reliable
identification of radical-molecular complexes and
various isomeric or conformation structures of inter-
mediates.
We have recently proposed a modification of the
method based on the high mobility of F atoms in solid
argon and combined use of ESR and IR spectros-
copy.1,4 The capability of F atoms of migrating in solid
argon at long distances5,6 provides its bringing together
with the isolated molecule of the second reactant, which
favors the accumulation of trapped radicals (in concen-
trations sufficient for detection) formed directly in the
solid phase. The potentialities of this approach are
described in previous publications.1 The advantages of
the combined use of two supplementing spectro-
scopic methods are shown for the reactions of the
F atom with the CH4,7,8 C H , NO,10 and NH3 1,4
9
2
4
molecules.
It is established that the full set of isotropic hyper-
fine coupling (IHFC) constants of radical products of
chemical reactions can be determined due to the high
resolution of ESR spectra of intermediates stabilized in
solid argon. A comparison of the measured IHFC con-
stants of the radicals and radical-molecular complexes
with the values calculated by the modern quantum-
chemical methods shows a good agreement with experi-
ments. The IHFC constants change strongly with insig-
1
nificant distortions of the equilibrium structure ; there-
fore, the ESR method is more reliable than IR spectros-
copy for the identification and establishment of the
spatial structure of stabilized radicals.
*
For Part 3, see Ref. 1.
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 6, pp. 947—953, June, 2001.
066-5285/01/5006-989 $25.00 © 2001 Plenum Publishing Corporation
1

9
90
Russ.Chem.Bull., Int.Ed., Vol. 50, No. 6, June, 2001
Misochko et al.
•
early 1990s.17 Based on these calculations, we can
This work is devoted to the study of β-C H F
2
2
•
radicals formed in reactions of mobile F atoms with
acetylene molecules. According to reaction (1), the first
reaction channel predominates in the gas phase.11,12
conclude that cis-β-C H F is more stable than the
2
2
•
trans-isomer; the spin density distribution in β-C H F
2
2
(responsible for "internal" hyperfine magnetic interac-
tions) differs from that in C H3•; the IHFC constants
2
•
β-C H F
of the β-C2H2F• isomers differ significantly, and the
2
2
•
•
[β-C H F ]*
13,15
C H + F
(1)
measured values
are closer to those calculated for
2
2
2 2
•
•
cis-β-C H F ; the frequencies of IR vibrations of the
C
CH + HF
2 2
radical isomers differ insignificantly (the differences are
comparable with the calculation accuracy) and, there-
fore, the experimental spectrum16 cannot unambigu-
ously be assigned to one of these isomers.
•
The β-C H F radical can exist in two isomeric forms,
2
2
depending on the position of the F atom toward the
orbital of an unpaired electron
Thus, the results of calculations certainly show
that the conclusions by the authors of experimental
works13,15,16 are dubious. To be convinced in a reliabil-
ity of the calculated data, we preliminary performed
analogous calculations using the improved quantum-
chemical methods and completely confirmed all con-
clusions presented above. Taking into account afore-
said, we carried out the experimental studies aimed at
F
F
H
•
C
C
C
C
•
H
H
H
cis
trans
The results of matrix studies of isomers of this
radical obtained in 1970—1980 and recent quantum-
chemical data are contradictory. This problem can be
solved only by the combination of kinetic IR and ESR
measurements with quantum-chemical calculations and
taking into account the experience of our previous
works on the determination of possibilities of each
spectroscopic method in the identification of inter-
mediates.
The first attempts to obtain β-C D F radicals were
made by the photolysis of solid Ar—F —C D mixtures
at 4.2 K.13 Although the authors of this work could not
detect well-resolved spectra of the stabilized radicals,
they determined the IHFC constants on nuclei of the F
atom (aF = 6.9 mT) and one of the D atoms
obtaining resolved ESR and IR absorption spectra of
•
the C H F radicals and reliably determining the char-
2
2
acteristics of the stabilized isomeric form based on the
comparison of the spectroscopic data with calculation
results.
Experimental
•
2
2
The procedure of preparation of samples has previously
been described in detail.8 Samples were prepared by the com-
2
2
2
bined condensation in vacuo of the Ar—C H and Ar—F2 gas
2
2
mixtures on a support cooled to 15 K (a sapphire rod in ESR
experiments and CsI in IR studies). Samples with the relative
molar content of reactant molecules from 3•10–3 to 2.5•10–4
(
aD = 0.6 mT). Taking into account the known IHFC
were studied. The samples with Ar : F2 : C H = 3000 : 1 : 1
2
2
constants of isomers of the C H3 radical,14 the ob-
•
were studied in most detail; their thickness being ∼ 100 µm.
ESR spectra were measured using a flow-type cryostat with
a mobile helium chamber. The thermoregulation system pro-
vided the stability of the sample temperature of at least 0.1 K in
the range of 15—40 K. IR absorption spectra were recorded on
a Mattson Model RS/10000 Fourier spectrometer with a spec-
tral resolution of 0.5 cm–1 in the 4000—500 cm–1 range. In
these experiments, a closed-cycle helium refrigerator with a
minimum working temperature of 13 K was used. In some
experiments, we checked that the radiation of the spectrometer
globar had no effect on chemical reactions in samples.
2
•
served ESR spectrum was ascribed to the trans-β-C H F
radical. The better resolved spectra of the β-C H F
radical in the SF matrix1 were also attributed to the
2
2
•
2
2
5
6
•
trans-β-C H F isomer. This assignment was argued, in
2
2
particular, by the semiempirical INDO calculations,
according to which the trans-isomer should be more
stable than the cis-isomer.
The IR absorption spectra of the reaction products
frozen in solid argon in the Ar—NF —C H gas mix-
tures have first been presented in Ref. 16. Although the
obtained spectra contain bands of products of various
reactions, the authors succeeded to isolate the IR spec-
trum of the β-C H F radical. Based on the conclu-
sions, the authors attributed the observed spectrum to
the trans-isomer. The analysis of normal vibrations of
the trans-β-C H F isomer and comparison of the cal-
culated vibrational spectrum with the measured IR fre-
quencies favor this assignment.
3
2
2
Fluorine atoms were generated in photolysis of F2 mol-
ecules by UV lasers with the wavelengths λ = 355 nm (radiation
of the third harmonic of an Nd : YAG laser with a frequency of
1
0 Hz in IR measurements) and λ = 337 nm (radiation of an
•
2
2
N2 laser with a frequency of 1000 Hz in ESR measurements).
The average radiation power did not exceed 10 mW cm–2. To
separate reactions of photogenerated "hot" and diffusing "ther-
mal" F atoms, photolysis of F2 was carried out at 15 K.
Diffusion of F atoms in solid argon is thermally activated at
temperatures above 20 K. The reaction of diffusing thermal
atoms was performed as follows: after the end of photolysis, the
sample was annealed for 3—5 min at 23 K, then the tempera-
ture was decreased to 15 K, and the IR absorption spectrum
was recorded. These cycles were repeated 10—12 times to the
complete cessation of chemical transformations.
1
3
•
2
2
The first reliable calculations of spectroscopic char-
•
acteristics of the C H F radicals by nonempirical meth-
ods (for example, CCSD) and in the framework of the
density functional theory (NLSD) were performed in
2
2

Reactions of photogenerated fluorine atoms
Russ.Chem.Bull., Int.Ed., Vol. 50, No. 6, June, 2001
991
According to ESR measurements, the prepared
Ar—C H —F samples contain no radical. UV photolysis fol-
in the samples with a relative content of the reactants of
∼ 3•10–4 (Refs. 7 and 9).
2
2
2
lowed by annealing of the Ar—F2 sample at 23 K results in the
Subsequent annealing of the samples at 23 K results
in the appearance and growth of new bands at 643, 698,
–
1
appearance of a narrow band at 1490 cm in the IR spectra,
which is attributed to the FO2• radicals. In similar ESR
experiments, the spectrum of this radical is also observed.
Molecular oxygen is the main impurity in gaseous fluorine used
and is always present in the studied samples in a relative molar
concentration of ∼ 10–4—10–5. Since F atoms diffusing in solid
85, 1070, 1263, 1631, 3026, and 3193 cm–1 (see Fig. 1,
7
spectrum 3; Fig. 2 presenting the IR spectrum of the
Ar—F2—C2D2 sample after annealing). The increase in
their intensity is accompanied by a decrease in bands of
the isolated C H (or C D ) molecules. Simultaneously
argon react efficiently with O2 molecules at temperatures above
2
2
2
2
•
2
0 K to form stabilized FO2 radicals,18 we used this reaction
•
the intensity of the narrow band of the FO2 radicals
increases (Fig. 3). The linear plots of the increase in the
intensity of new bands vs. decrease in the intensity of
as a test for monitoring reactions of thermal F atoms.
Results and Discussion
the bands of C H assert that a series of eight IR bands
2
2
is attributed to the primary products of the reactions of
diffusing F atoms with acetylene molecules.
IR absorption spectra of Ar—F —C H (C D )
2
2
2
2
2
samples. The IR spectrum of the starting sample
According to reaction (1), both isomers of the
Ar : F2 : C H2 = 3000 : 1 : 1 (Fig. 1, spectrum 1)
2
•
•
β-C H F radical, as well as the C H radical and HF
2
2
2
contains broad lines of acetylene molecules with maxima
at 736.5, 1334.5, and 3288.5 (3302.0) cm–1. The short
photolysis (∼ 5 min) at 15 K increases rapidly the intense
molecule, can be its primary products. The absence of
an intense absorption band of the HF molecule in the
region of 4000—3000 cm–1 and a characteristic band of
the C2H• radical in solid argon at 1845 cm–1 imply
bands in the regions of a stretching vibration of C—F
16
1058, 1129, and 1153 cm–1) and a molecular vibration
(
•
of H—F (3756 and 3900 cm–1) (see Fig. 1, spectrum 2).
that the excited [β-C2H2F ]* radical is not decomposed
in the solid matrix. Each of the isomers of the
β-C2H2F• radical possesses nine normal vibrations, and
the strongest stretching vibrations of C—F and C=C
should be comparable in intensity (Table 1). However,
each spectral range contains only one strong absorption
band (see Fig. 1, spectrum 3, and Fig. 2) and, hence,
the observed eight bands belong to one isomer. There-
These bands have previously be assigned to CHF=CHF
and C HF in the complex with HF.1
9,20
Such molecular
2
products can be formed in the cage reaction during
photolysis of complexes of the reactants
CHF CHF
hν
[
C H + F2]
[C H F ]*
(2)
2
2
2 2 2
•
fore, one of the isomers of the β-C H F radical is
2
2
FC CH + HF
the predominant product of the reaction of diffusing
F atoms.
The maximum decrease in the intensity of the IR
bands of acetylene during prolonged photolysis is ∼ 2%,
which corresponds to the fraction of binary complexes
Note that the annealing of the samples containing
stabilized radicals for 72 h at 30 K does not notice-
C H
2
2
C H
2
2
C H
2
2
0
0
.16
.04
1
2
C HF
C HF
2
2
C H F
2
2
2
*
(
C=C)
FO2•
0
.04
*
*
*
*
(C—F)
*
*
*
3
4
000
3500
3000
1500
1000
ν/cm–1
Fig. 1. IR absorption spectrum of the sample Ar : F2 : C H = 3000 : 1 : 1 after deposition (1), differential IR absorption spectrum
2
2
before and after photolysis at 15 K (2), differential spectrum of the irradiated at 15 K sample before and after annealing at 24 K (3),
•
and spectra recorded at 15 K (IR bands of the β-C H F radical are marked by asterisks).
2
2

9
92
Russ.Chem.Bull., Int.Ed., Vol. 50, No. 6, June, 2001
Misochko et al.
*
1
2
(
C=C)
FO2•
*
(
C—F)
*
0
.04
*
*
*
*
C D
2
2
C D
3
2
2
FO2•
aHα
aHβ
aF
314
3
11
Fig. 4. ESR spectra of samples after photolysis at 15 K and
annealing at 24 K: Ar : F2 : Ñ H = 200 : 1 : 1 (1),
317
320
H/mT
2
500
2000
1500
1000
ν/cm–1
Fig. 2. Differential spectrum of the irradiated at 15 K sample
Ar : F2 : C D = 3000 : 1 : 1; C H -d , 98% D) before and
after annealing at 24 K (the spectrum was recorded at 15 K;
2
2
(
2
2
2
2
2
Kr : F2 : Ñ H = 3000 : 1 : 1 (2), and Ar : F2 : Ñ H =
2
000 : 1 : 1 (3) (spectra were recorded at 30 K). The character-
istic splittings in HFS of the CHF= CH and FO are shown.
2
2
2
3
•
IR bands of β-C D F are marked by asterisks).
•
•
2
2
2
I/rel. units
ably change the spectra. Therefore, thermoactivated
cis-trans-isomerization does not occur at this tem-
perature.
1
0
0
0
0
.0
.8
.6
.4
.2
•
ESR spectrum of β-C H F radicals. The ESR spec-
2
2
trum of the sample Ar : F2 : C H = 3000 : 1 : 1
2
2
recorded at 30 K after UV photolysis at 15 K
and subsequent annealing at 23 K (Fig. 4, spec-
trum 3) consists of eight narrow isotropic lines
1
2
(
linewidth ≤0.06 mT) with equal integral intensities.
Despite different linewidths, all of them are the compo-
•
0
0.2
0.4
0.6
0.8
IFO •2 /rel. units
nents of the ESR spectrum of one radical (β-C H F ).
2
2
Analysis of the spectrum distinguishes three hyperfine
doublet splittings a = 6.50, a = 3.86, and a = 0.25 mT.
The a and a constants are close to those known for the
1 2
•
Fig. 3. Intensity (I) of the IR band (1) of the β-C H F radical
2
2
–
1
1
2
3
(
(
1631 cm ) and "consumption" (2) of the band of C H
2
2
–
1
•
736.5 cm ) vs. intensity of the band of the FO
radi-
2
1
3,15
–
1
2-fluorovinyl radical.
However, the last splitting a3
cals (1490 cm ) formed during annealing of the sample
Ar : F2 : Ñ H = 3000 : 1 : 1 at 24 K.
was not observed in previous works.1
3,15
2
2
Table 1. Calculated (B3LYP/6-311++G(3df,2p)) in the harmonic ap-
proximation frequencies of IR vibrations (ν) and IHFC constants (a) of
•
isomers of the CHF= CH radical compared to the experimental values
Vibration
Calculation
Experiment
This work
trans
cis
Ref. 16
3178
ν/cm–1 (Irel (%))
3308 (0.13)
3122 (0.04)
1678 (0.73)
1291 (0.05)
1093 (1.00)
734 (0.47)
479 (0.04)
827 (0.08)
645 (0.42)
a/mT
1
2
3
4
5
6
7
1
2
a´ (CH )
3307 (0.09)
3184 (0.01)
1664 (0.48)
1305 (0.01)
1024 (1.00)
779 (0.08)
433 (0.00)
883 (0.27)
661 (0.28)
3193 (0.13)
3026 (0.04)
1631 (0.71)
1263 (0.11)
1070 (1.00)
698 (0.38)
α
a´ (CH )
β
a´ (CC)
1623
1211
1066
678
462
785
a´ (H CC)
β
a´ (CF)
a´ (H CC)
α
a´ (FCC)
a″ (wag)
a″ (twist)
785 (0.12)
643 (0.40)
631
aF
aHβ
aHα
10.120
1.69
0.14
6.69
3.81
0.17
6.50
3.86
0.25
Note. Relative intensities of vibrations are given in parentheses.

Reactions of photogenerated fluorine atoms
Russ.Chem.Bull., Int.Ed., Vol. 50, No. 6, June, 2001
993
The IHFC constant with the α-proton is close to
zero,21 therefore, the absence of splitting on α-H in the
ESR spectra can be related to their insufficient resolu-
tion. Figure 4 shows the obtained ESR spectra of the
curve 1) contains three series of lines. The first, prevail-
•
ing group of the β-C D F radical contains six lines:
2
2
a doublet (a1 = 6.50 mT) of triplets 1 : 1 : 1
(a = 0.59 mT). The second series consists of four lines
2
•
•
β-C H F radicals in the argon matrix with a high
(CHF= CD): a doublet (a = 6.50 mT) of doublets
2
2
1
•
concentration of C H (see Fig. 4, spectrum 1) and in
(a = 3.86 mT). In the third series (CDF= CH) includ-
2
2
2
the krypton matrix (spectrum 2) with magnetic nuclei
the nuclear spin I of 83Kr is equal to 9/2). In the first
ing 12 lines (six of them are disguised by more intense
lines of another series), we can isolate two splittings on
nonequivalent nuclei with the spin I = 1/2 (a1 = 6.50
(
case, most of the C H2 molecules exist as dimers (and
2
•
more complex clusters). Therefore, the β-C H F radi-
and a = 0.25 mT) and one on the nucleus with the spin
2
2
3
cals that formed are bound to the adjacent C H mol-
I = 1 (a = 0.59 mT).
2
2
2
ecules, resulting in the line broadening due to the
incomplete averaging of the HFC anisotropy because of
the retarded motion of the radicals. In the second case,
the ESR lines are strongly broadened due to the interac-
tion of an unpaired electron of the radical with mag-
netic nuclei of the matrix. A comparison of spectra 1
and 2 with spectrum 3 (see Fig. 4) shows that a high
spectral resolution is achieved only for the argon matrix
For isotopic substitution H/D hyperfine splittings on
nuclei of D atoms should be lower by γ /γD ≈ 6.5 times
H
according to the gyromagnetic ratio of H to D, and that
on the 19F nucleus should remain unchanged. There-
fore, the splitting a = 6.50 mT, which does not change
1
in all spectra, should be assigned to the nucleus of the
1
9
F atom. Doublet splitting a = 3.86 mT of the
2
•
•
CHF= CH and CHF= CD radicals and the triplet
a = 0.59 mT of the CDF= CD and CDF= CH radi-
•
•
(
with non-magnetic nuclei) when stabilized radicals are
2
ideally isolated. This enables the observation of HFC
with a splitting of 0.25 mT.
cals correlates with the β-proton. The remaining doublet
•
•
splitting a = 0.25 mT of the CHF= CH and CDF= CH
3
In order to assign the observed hyperfine splitting of
the ESR spectrum, we carried out experiments with
radicals should be attributed to α-H. The corresponding
•
•
triplet splitting of the CDF= CD and CHF= CD radi-
cals is lower than the linewidth and, hence, is not
observed. Thus, in these experiments we succeeded for
the first time to obtain the resolved ESR spectrum of
mixtures enriched in C D by 75%, C HD by 22%, and
2
2
2
C H by 3%. The addition of the F atom to the C D
2
2
2
2
•
molecule results in the formation of the CDF= CD
•
radical. If the F atom is added to the C HD molecule,
the β-C H F radical and determine all three IHFC
2
2
2
•
•
the CHF= CD and CDF= CH radicals can be formed.
The spectrum of this sample after low-temperature
UV photolysis followed by annealing at 23 K (Fig. 5,
constants.
It is known that small molecules and radicals in
matrices of inert gases rotate almost freely (see, e.g.,
Ref. 22). If radicals rotate non-coherently, changes in
the linewidths and intensities observed in the ESR
spectra are explained by incompletely averaged aniso-
tropic magnetic interactions. The proposed model of
rotation diffusion23 based on the Redfield equations24
allows the relation of the linewidths of the HFC compo-
nents in the ESR spectrum to the structural parameters
of the radical, such as the electron density distribution.
In the fast rotation limit, the expression for the linewidth
takes the following form:
1
FO•2
2
F
−1
2,0
I
T2− =T
1
+
A + B m + C mi2
+
E m m τR
,
(3)
G
∑i
i
i
∑i
i
∑
ij
i
j
J
H
i< j
K
•
CDF= CD
where m and m are the z components of nuclear spins
•
i
j
CDF= CH
for the ith and jth nuclei; τ is the time of rotation
correlation, and A, B, C, and E are determined by the
anisotropic parts of the HFC tensor and g factor.
•
CHF= CD
3
11
314
317
320
H/mT
•
Since the ESR spectrum of the CHF= CH radical is
Fig. 5. ESR spectrum (1) of the sample Ar : F2 : (Ñ D (75%)
symmetric relative to the center and nuclear spins of all
magnetic nuclei are equal to 1/2, only the cross term
E m m determines the different linewidths. In spec-
2
2
+
Ñ HD (22%) + C H (3%)) = 3000 : 1 : 1 after photolysis at
2 2 2
1
5 K and annealing at 24 K; the spectrum was recorded at 30
K; calculated ESR spectrum (2) of the sample containing
ij
i
j
trum 3 (see Fig. 4) this is most pronounced for pairs of
lines for which the m m product for the F and α-H
•
•
•
radicals CDF= CD (78%), CHF= CD (11%), and CDF= CH
11%); IHFC constants and linewidths are taken from the
experimental spectra. The characteristic splittings in HFS of
i
j
(
nuclei changes its sign. The E coefficients are obtained
ij
•
•
•
•
by multiplying the HFC tensors of various nuclei.
23,25
the CDF= CD, CDF= CH, CHF= CD, and FO
are shown.
radicals
2
The estimation of the rotation correlation time by for-

9
94
Russ.Chem.Bull., Int.Ed., Vol. 50, No. 6, June, 2001
Misochko et al.
•
Table 2. Calculated geometric parameters of cis- and trans-isomers of the CHF= CH radical
Parameter
cis-Isomer
trans-Isomer
B3LYP/
B3LYP/
QCISD/
B3LYP/
B3LYP/
QCISD/
6
-311++G(d,p) 6-311++G(3df,2p) 6-311++G(d,p)
6-311++G(d,p) 6-311++G(3df,2p) 6-311++G(d,p)
R
R
/Å
/Å
1.074
1.088
1.356
1.303
1.073
1.087
1.349
1.300
1.079
1.089
1.349
1.316
1.074
1.084
1.369
1.302
1.073
1.082
1.361
1.299
1.079
1.085
1.360
1.315
CHα
CHβ
RCF/Å
RCC/Å
H CC/deg
139.9
139.5
136.0
144.5
144.1
140.1
α
H CC/deg
FCC/deg
125.6
122.9
125.2
123.2
125.2
122.5
125.9
122.8
125.6
122.9
125.8
122.1
β
mula (3) taking into account the observed difference in
widths ∆T2–1 ∼ 0.01 mT at 30 K gives τR ∼ 10–9 s.
Quantum-chemical calculations of energy and spec-
Table 3. Differences between energies ignoring (∆E0) and tak-
ing into account (∆E0 + ∆ZPE) energies of zero point vibra-
•
tions of trans- and cis-isomers of the CHF= CH radical
(
J mol–1
)
•
troscopic parameters of isomers of the β-C H F radi-
2
2
cal. The calculations by the Gaussian 98 program pack-
age26 showed that the methods of the density functional
Method
∆E0
∆E0 + ∆ZPE
theory reliably reproduce both the structural and spec-
troscopic parameters of free radicals.17,27 The calcula-
tion accuracy is comparable and sometimes exceeds that
inherent in non-empirical post-Hartree—Fock methods
B3LYP/6-311++G(d,p)
B3LYP/6-311++G(2d,2p)
B3LYP/6-311++G(3df,2p)
QCISD/6-311++G(d,p)
G1
–95
10
859
678
1100
1200
336
600
1232
1174
1351
1489
(
for example, the cluster decomposition method CCSD
G2
or configuration interaction method QCISD), although
much less calculations are required.
We used the method of the density functional theory
B3LYP with the 6-311++G(d,p) and 6-311++G(3df,2p)
basis sets. Experimental data are lacking for the struc-
IHFC constants of the cis-isomer virtually coincide with
the measured values. Third, our data3,4,10 and results of
other works17,28 show that this method reproduces the
experimental IHFC constants of free radicals with an
accuracy of at least 10%. Thus, the comparison of the
calculated and measured spectroscopic parameters of
•
ture of the β-C H F radical, therefore, we used the
2
2
QCISD/6-311++G(d,p) method to control the validity
of its prediction. It follows from Table 2 that all meth-
ods give almost the same structures of isomers. The best
agreement with the testing calculation was achieved in
the B3LYP variant with the 6-311++G(3df,2p) ba-
sis set.
•
the β-C H F radical asserts that the cis-isomer of
2
•
2
β-C H F is the main stabilized product of the reac-
2
2
tions of the F atoms with the C H molecules in solid
2
2
The results of calculations of the spectroscopic pa-
argon.
•
rameters of isomers of the β-C H F radical are pre-
To reveal the reason for the stabilization of only one
isomer of β-C H F , let us compare their calculated
energies. Table 3 contains the differences between the
energies ignoring (∆E ) and taking into account
(∆E0 + ∆ZPE) energies of zero vibrations of the
trans- and cis-isomers of the CHF= CH radical. The
results show that with increasing the level of calcula-
tions the cis-isomer becomes more energetically favor-
able. The maximum difference between energies of the
isomers reaches ∼ 1.0 kJ mol–1 (∼ 1.5 kJ mol–1 taking
into account zero point vibrations). Although this dif-
ference is comparable with the accuracy of calculations
of energies of the isomers, the observed tendency cer-
tainly indicates a high stability of the cis-isomer.
2
2
•
sented in Table 1. Despite that the calculated frequen-
cies for the cis-isomer are closer to the experimental
values than those for the trans-isomer, it is impossible to
assign unambiguously the experimental IR spectrum to
one of them on the basis of vibration analysis. One of
the reasons is the absence of anharmonic corrections in
the calculation of vibration frequencies and the pres-
ence of "matrix" shifts of the IR bands comparable to
the difference in the characteristic vibration frequencies
of the isomers. Therefore, even analysis of the IR
spectra taking into account isotopic (H/D) substitution
does not allow a reliable identification of the stabilized
isomeric form.
2
2
0
•
At the same time, the ESR spectrum can unambigu-
ously be assigned by the comparison of experimental
and calculated IHFC frequencies. First, the calculations
predict that the HFC constants of two isomers dif-
fer very strongly (aF and aHβ differ by more than
The calculated barrier of cis-trans-isomerization was
9.0 kJ mol–1. Therefore, the equilibrium concentration
corresponding to the trans-isomer at temperatures below
30 K is very low. The experiment confirms qualitatively
the conclusions based on the calculations. Since at 30 K
the equilibrium concentration of one isomer is at least
1
.5 and 2 times, respectively). Second, the calculated

Reactions of photogenerated fluorine atoms
Russ.Chem.Bull., Int.Ed., Vol. 50, No. 6, June, 2001
995
∼
20-fold higher than that of the second isomer, the
14. E. L. Cochran, F. J. Adrian, and V. A. Bowers, J. Chem.
Phys., 1964, 40, 213.
difference in their enthalpies should exceed 0.7 kJ mol–1.
1
5. M. Shiotani, Y. Nagata, and J. Sohma, J. Phys. Chem.,
982, 86, 4131.
1
This work was financially supported by the Russian
Foundation for Basic Research (Project No. 98-03-
3
1
1
6. M. E. Jacox, Chem. Phys., 1980, 53, 307.
7. V. Barone, C. Adamo, and N. Russo, Chem. Phys. Lett.,
3175) and the National Scientific Foundation of the
1
993, 212, 5.
USA (Grant CHE-9970032).
18. E. Ya. Misochko, A. V. Akimov, and C. A. Wight, Chem.
Phys. Lett., 1997, 274, 23.
1
9. C. Zetsch, Ph. D. Dissertation, Georg-August University,
Göttingen, Germany, 1971.
References
2
0. G. R. Hunt and M. K. Wilson, J. Chem. Phys., 1961,
3
4, 1301.
1
. E. Ya. Misochko, I. U. Goldshleger, and A. V. Akimov, Izv.
Akad. Nauk, Ser. Khim., 2000, 832 [Russ. Chem. Bull., Int.
Ed., 2000, 49, 829].
21. H. M. McConnel and J. Strathdee, Mol. Phys., 1959, 2, 129.
22. M. E. Jacox, in Inert Gases, Ed. M. L. Klein, Springer
Series in Chemical Physics, 1984, 34.
2
3
4
. W. E. Jones and E. G. Skolnik, Chem. Rev., 1976, 76, 563.
. M. E. Jacox, Rev. Chem. Intermed., 1985, 6, 77.
. E. Ya. Misochko, I. U. Goldschleger, A. V. Akimov, and
C. A. Wight, Low Temp. Phys., 2000, 26, 981.
23. J. H. Freed and G. K. Fraenkel, J. Chem. Phys., 1963,
39, 326.
24. A. G. Redfield, IBM Res. Develop., 1957, 1, 19.
25. J. H. Freed and G. Fraenkel, J. Chem. Phys., 1963, 40, 1815.
26. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski,
J. A. Montgomery, Jr., R. E. Stratmann, J. C. Burant,
S. Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin,
M. C. Strain, O. Farkas, J. Tomasi, V. Barone,
M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo,
S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala,
Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck,
K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz,
B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz,
I. Komaromi, R. Gomperts, R. L. Martin, D. J. Fox,
T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
C. Gonzalez, M. Challacombe, P.M.W. Gill, B. Johnson,
W. Chen, M. W. Wong, J. L. Andres, C. Gonzalez,
M. Head-Gordon, E. S. Replogle, and J. A. Pople,
GAUSSIAN 98, Revision A.3, Gaussian, Inc., Pittsburgh
(PA), 1998.
5
6
7
. J. Feld, H. Kunttu, and V. A. Apkarian, J. Chem. Phys.,
1
990, 92, 1009.
. E. Ya. Misochko, A. V. Akimov, and C. A. Wight, Chem.
Phys. Lett., 1998, 293, 547.
. E. Ya. Misochko, V. A. Benderskii, A. U. Goldschleger,
A. V. Akimov, A. V. Benderskii, and C. A. Wight, J. Chem.
Phys., 1997, 106, 3146.
. E. Ya. Misochko, V. A. Benderskii, A. U. Goldshleger,
A. V. Akimov, A. V. Benderskii, and C. A. Wight, Izv.
Akad. Nauk, Ser. Khim., 1997, 709 [Russ. Chem. Bull.,
8
1
997, 46, 678 (Engl. Trans.)].
9
. E. Ya. Misochko, A. V. Benderskii, and C. A. Wight,
J. Phys. Chem., 1996, 100, 4496.
1
1
0. E. Ya. Misochko, V. A. Akimov, I. U. Goldschleger, A. I.
Boldyrev, and C. A. Wight, J. Am. Chem. Soc., 1999,
1
21, 405.
1. G. A. Kapralova, T. P. Nagornaya, and A. M. Chaikin,
Kinet. Katal., 1970, 11, 809 [Kinet. Catal., 1970, 11, 667
27. F. De Proft, J. M. L. Martin, and P. Geerlings, Chem.
Phys. Lett., 1996, 250, 393.
28. V. Barone, J. Chem. Phys., 1994, 101, 6834.
(
Engl. Transl.)].
2. R. L. Williams and F. S. Rowland, J. Phys. Chem., 1973,
7, 301.
1
1
7
3. E. L. Cochran, F. J. Adrian, and V. A. Bowers, J. Phys.
Chem., 1970, 74, 2083.
Received October 6, 2000;
in revised form February 23, 2001