Absorption spectroscopy of SiH2 near 640 nm

Article abstract of DOI:10.1063/1.476062

The A ¹B₁-X ¹A₁ absorption spectrum of SiH₂ has been observed using intracavity laser absorption spectroscopy with an equivalent path length of up to 13.0 km and the A ¹B₁(0, 0, 0)-X ¹A₁(0, 0, 0) band near 640 nm recorded for the first time. The silylene radical was generated in a continuous discharge in a flowing mixture of silane in argon, giving a concentration of the order of 10¹⁰ SiH₂/cm³. The spectrum spans the region between 15350 and 16100 cm^-1. Rotational transitions have been assigned to levels up to J = 16 and K_a = 9, with ΔK_a up to 5, ΔK_c up to 4. Perturbations have been detected in the spectrum, due to Renner-Teller and spin-orbit interactions between both electronic states and the ³B₁ state, predicted to be between them. However, the strength of the irregular perturbations affecting the rotational states of A ¹B₁(0,0,0) state is found to be much weaker than that affecting the other (0, v′₂, 0) levels previously studied. The analysis of the spectrum has allowed the determination of the rotational constants of the ¹B₁ (0,0,0) level, and a new estimation of those of the vibrational and electronic ground state. The geometry of the excited electronic level has also been determined for the first time from accurate experimental data. A change in the structure of this molecule takes place with this transition, the equilibrium angle opening from 92° to 122.4°, while the bond distance is reduced from 1.51 to 1.485 A.

Full text of DOI:10.1063/1.476062

Absorption spectroscopy of SiH 2 near 640 nm
Rafael Escribano and Alain Campargue
Citation: The Journal of Chemical Physics 108, 6249 (1998); doi: 10.1063/1.476062
View online: http://dx.doi.org/10.1063/1.476062
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/108/15?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Spectroscopy of diatomic ZrF and ZrCl: 760 – 555 nm
J. Chem. Phys. 135, 024308 (2011); 10.1063/1.3608055
Rotationally resolved spectroscopy of the A ̃ A 2 1 ← X ̃ B 2 1 transition of H 2 S + above the barrier to linearity
using the mass-analyzed threshold ionization photofragment excitation technique
J. Chem. Phys. 132, 124304 (2010); 10.1063/1.3358252
Absorption cross sections and correlation functions of the N H 2 A ̃ A 1 2 − X ̃ B 1 2 Renner–Teller system
J. Chem. Phys. 122, 234315 (2005); 10.1063/1.1929737
Persistence rewarded: Successful observation of the B̃ 2 Σ u + –X̃ 2 Π g electronic transition of jet-cooled BS 2
J. Chem. Phys. 120, 4258 (2004); 10.1063/1.1644097
Photodissociation spectroscopy of the Ca + –N 2 complex
J. Chem. Phys. 108, 2725 (1998); 10.1063/1.475700
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Thu, 27 Nov 2014 18:55:19

JOURNAL OF CHEMICAL PHYSICS
VOLUME 108, NUMBER 15
15 APRIL 1998
Absorption spectroscopy of SiH2 near 640 nm
Rafael Escribano
Instituto de Estructura de la Materia, CSIC, Serrano, 123, 28006 Madrid, Spain
Alain Campargue
´
`
´
Laboratoire de Spectrometrie Physique (UMR 5588), Universite Joseph Fourier de Grenoble, B.P. 87,
38402 Saint-Martin-d’Heres Cedex, France
͑Received 21 November 1997; accepted 13 January 1998͒
1
˜ ˜
The A B1 – X 1A1 absorption spectrum of SiH2 has been observed using intracavity laser
absorption spectroscopy with an equivalent path length of up to 13.0 km and the
1
B1(0, 0, 0)– X 1A1(0, 0, 0) band near 640 nm recorded for the first time. The silylene
˜
˜
A
radical was generated in a continuous discharge in a flowing mixture of silane in argon, giving a
concentration of the order of 1010 SiH2 /cm3. The spectrum spans the region between 15350 and
16100 cmϪ1. Rotational transitions have been assigned to levels up to Jϭ16 and Kaϭ9, with ⌬Ka
up to 5, ⌬Kc up to 4. Perturbations have been detected in the spectrum, due to Renner–Teller and
3
spin-orbit interactions between both electronic states and the B1 state, predicted to be between
them. However, the strength of the irregular perturbations affecting the rotational states of
1
˜
A B1͑0,0,0͒ state is found to be much weaker than that affecting the other (0, Ј, 0) levels
v2
previously studied. The analysis of the spectrum has allowed the determination of the rotational
1
constants of the B1 ͑0,0,0͒ level, and a new estimation of those of the vibrational and electronic
ground state. The geometry of the excited electronic level has also been determined for the first time
from accurate experimental data. A change in the structure of this molecule takes place with this
transition, the equilibrium angle opening from 92° to 122.4°, while the bond distance is reduced
from 1.51 to 1.485 Å. © 1998 American Institute of Physics. ͓S0021-9606͑98͒01815-7͔
I. INTRODUCTION
this transient species. SiH2 was first identified by Dubois 30
years ago7,8 by detection of its electronic absorption spec-
trum in the visible range. It was produced by photolysis of
phenylsilane and the spectrum was obtained with a path
length of 48 m and photographed at high dispersion. The
spectrum, which consists of a progression of several bands
separated by about 850 cmϪ1, shows a widely open rota-
tional structure due to the important change of geometry. It
Silylene, SiH2 , is an intermediate species of importance
in silicon hydride chemistry as it may play a role in the
deposition of amorphous silicon films by chemical vapor
deposition ͑CVD͒. The industrial applications of this very
reactive species have prompted several studies on its reactiv-
ity and dynamics.1–3
From the academic point of view, SiH2 is the simplest
polyatomic molecule containing silicon, it is isovalent with
CH2 , and its electronic spectrum presents a challenging
blend of Renner–Teller and spin-orbit couplings. It is also of
the right size to enable high level ab initio calculations.
In the ground electronic state (1A1), the equilibrium ge-
ometry is close to a rectangular triangle ͑␪eϭ92°͒, very simi-
lar to the SH2 molecule, with marked asymmetric top struc-
ture ͑asymmetry parameter ␬ϭ0.51͒. The next upper
was analyzed as an A 1B1(0, Ј, 0)ϪX 1A1(0, 0, 0)
˜
˜
v2
transition, each band of the progression corresponding to the
excitation of the bending vibration ( Јϭ1Ϫ7) in the A 1B1
˜
v2
electronic state. The interpretation of these spectra8,9 led to
the determination of the geometrical structure of the 1B1 and
1A1 states, but the rotational level structure of the excited
state was found to be affected by a large number of erratic
rotational perturbations. A rms observed-calculated deviation
of more than 1 cmϪ1 was obtained in the fit of the rotational
parameters of the excited states for the transitions to the
͑010͒, ͑020͒ and ͑030͒ states.8 More recently, Duxbury, Ali-
jah and Trieling4 have carried out an extensive study on
these previously published data,8,9 showing that these pertur-
bations arise from the combined effects of Renner–Teller
interaction between the two singlet states and spin-orbit cou-
electronic state is a 3B1 , estimated to be 7000 cmϪ1 above,
˜
although it has never been directly observed, as far as we are
aware. In the next upper state, A 1B1 , situated 15500 cmϪ1
˜
above the ground state, the bending angle opens up drasti-
cally to an equilibrium value of ϳ120°, giving a nearly pro-
late symmetric top structure (␬ϭϪ0.84͒. A schematic rep-
resentation of these two geometrical structures is presented
in Fig. 1. In both singlet states, large excitations of the bend-
ing vibration would bring the molecule near a linear configu-
ration, thus enabling a large Renner–Teller interaction,
which has been identified from the analysis of rovibronic
spectra and lifetimes of excited levels.4–6
pling with the lowest triplet a 3B1 . They also developed a
˜
model to treat these interactions simultaneously. The strong
4
rotational perturbations in the A 1B1 state have been related
˜
A number of spectroscopic studies has been devoted to
to the unusual radiative lifetimes recently measured by laser-
0021-9606/98/108(15)/6249/9/$15.00
6249
© 1998 American Institute of Physics
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Thu, 27 Nov 2014 18:55:19

6250
J. Chem. Phys., Vol. 108, No. 15, 15 April 1998
R. Escribano and A. Campargue
of the order of 108 SiH2/cm3 was achieved.22
First Dubois,8 and later Yamada et al.,10 determined the
ground state rotational constants of SiH2 , and hence the r0
structure, by making ground state combination differences
from their respective data. Based on the similar structure of
SiH2 and SH2 , Yamada et al. extended their calculation to
derive the equilibrium geometrical structure of silylene by
B
means of the rovibrational constants ␣i of the SH2 mol-
ecule, which had been reported before.10
Allen and Schaefer25 have performed the most compre-
hensive theoretical calculation on the geometrical structure,
force constants and vibrational spectra of SiH2 ͑as well as
SiH, SiH3 and SiH4) for all three electronic states mentioned
FIG. 1. Equilibrium geometry of SiH2 in the X 1A1 ͑full line͒ and A 1B1
͑dotted line͒ electronic states.
˜
˜
above. They report also the estimated rovibrational constants
B
␣
for each state. Using these constants, they successfully
i
reduce the uncertainty in the geometrical structure found in
the determination of two parameters ͑re and ␪ ͒ from three
pairs of constants ͑Ae , Be and Ce , taken two by two͒. The
e
induced fluorescence ͑LIF͒ by Thoman, Steinfeld and
coworkers.5,6 In the 1A1 state, Yamada et al.10 studied the ␯
2
agreement between experimental and theoretical results is
band by means of an infrared diode laser spectrometer, thus
providing the only high resolution infrared data available for
this radical up to this moment.
1
1
very good in the ground A1 state, but less good in the B1
state, where the rotational constants of the Јϭ0 state had
v2
never been determined before. The ab initio potential energy
surfaces of the three low lying states of SiH2 have been
calculated by Kosecki and Gordon.26
Several authors have reported the LIF detection of si-
lylene produced by multiphoton or UV dissociation of silane
compounds.5,6,11–13 The LIF excitation and dispersed fluores-
cence spectra reported by Fukushima, Mayana and Obi13 led
to the determination of both the electronic transition moment
In the present work, we report the observation and rovi-
1
brational analysis of the A B1(0, 0, 0)ϪX 1A1(0, 0, 0)
transition, recorded by ICLAS between 15350 and 16100
cmϪ1. The only previous report concerning this transition is
the fluorescence excitation spectrum recorded at medium
resolution between 15540 and 15580 cmϪ1 and displayed
without rotational analysis in Fig. 4 of Ref. 6. This transition,
which is weaker than those previously studied, is particularly
important as it allows the determination of the geometry and
˜
˜
of the A B1 – X 1A1 transition (
͉
␮
͉
2ϭ0.26 e2 a20) and of
1
˜ ˜
e
the (0, Ј, 0)Ϫ(0, Љ, 0) Franck–Condon factors.
v2 v2
Most of the SiH2 detection experiments using
lasers5,6,11–21 have been performed on the A 1B1(0, 2, 0)
˜
ϪX 1A1(0, 0, 0) transition near 580 nm because of favor-
˜
able Franck–Condon access and of the availability of an ex-
cellent dye in the corresponding range ͑Rhodamine 6G͒. Ja-
sinski and coworkers14–16 used frequency modulation and
direct absorption techniques to monitor SiH2 by detecting
of the energy of the vibrationless level of the A 1B1 elec-
˜
tronic state. We have also determined the equilibrium re
structure of silylene in this singlet excited state. The accu-
racy of the line positions is about 0.01 cmϪ1, which is inter-
mediate between the high resolution infrared experiment of
Yamada et al.10 and the previous data from Dubois.7,8 We
also present a more comprehensive determination of the ro-
tational constants of the ͑0,0,0͒ level in the electronic ground
state, performed by combining all three sets of data with
appropriate weights in a single refinement process.
some rovibronic transitions of this band near 17250 cmϪ1
.
The same transition was observed by ICLAS19–21 and LIF17
in order to evaluate the SiH2 concentration produced in a RF
plasma20 or in reacting high temperature mixtures of
disilane.21 The estimation of the absolute concentration of
SiH2 was possible only by indirect comparative methods17,20
or by making an assumption on the reaction kinetic model
leading to the formation of SiH2.21 For in situ and real time
diagnostic of CVD plasma, absorption techniques are much
more suitable than LIF, resonantly enhanced multiphoton
ionization ͑REMPI͒ or four wave mixing. We have very
Another interest and motivation of the present study is
1
that the A B1(0, 0, 0)ϪX 1A1(0, 0, 0) transition near
˜
˜
640 nm corresponds to a range where diode lasers are avail-
able. Following the same procedure as described in our re-
recently22 obtained the Doppler limited absorption spectrum
1
B1(0, 2, 0)ϪX 1A1(0, 0, 0)
˜
A
˜
cent study of the
1
of the A B1(0, 2, 0)ϪX 1A1(0, 0, 0) transition using
two highly sensitive techniques equivalent to classical ab-
sorption: intracavity laser absorption spectroscopy
˜
˜
transition,22 it is then possible to use some specific rovi-
bronic transitions of this band for in situ and real time moni-
toring of SiH2 in CVD reactors using the newly developed
CW-CRDS technique.24
23
͑ICLAS͒ and the newly developed cw-cavity ring down
spectroscopy ͑CW-CRDS͒ technique.24 From the results of
Ref. 13 and using the rotational assignments given by
Dubois,8 it was possible to calculate the absolute line inten-
sity of some specific rovibronic transitions and then deduce
the SiH2 concentration from the measured absorption coeffi-
cient. To our knowledge, this study was the first direct and
absolute concentration measurement of SiH2 . A detectivity
II. EXPERIMENT
A. The discharge cell
The silylene radical was generated by a continuous dis-
charge ͑40 mA, 1 kV͒ in a slowly flowing mixture of silane
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Thu, 27 Nov 2014 18:55:19

J. Chem. Phys., Vol. 108, No. 15, 15 April 1998
R. Escribano and A. Campargue
6251
͑5%͒ in argon. The total pressure was about 1 Torr and gases
were pumped out with a mechanical pump. The plasma tube
͑80 cm length, 3.6 cm diameter͒ was fitted by Brewster angle
windows. The flow of argon entered the cell close to the two
windows while silane was injected through an inlet close to
the center of the cell. This arrangement prevents deposition
of silicon films on the Brewster windows as observed on the
walls of the cell. The distance between the two cylindrical
electrodes was 60 cm.
B. ICLAS spectrometer
The ICLAS experimental method has been described
before.23 Briefly, the ICLAS technique is based on the high
sensitivity of a broadband laser to intracavity losses namely
absorption. The absorption lines appear superimposed on the
broadband spectrum of the dye laser ͑DCM͒ pumped by an
argon laser. The spectrum dispersed by a high resolution
grating spectrograph ͑resolving power up to 800 000͒ is re-
corded by a 1024 photodiode array. Furthermore, the dye
spectrum is time resolved by using two synchronized
acousto-optic modulators and the generation time, tg , be-
tween the start of the laser and the recording of the spectrum
gives directly the equivalent path length:
FIG. 2. The ICLAS spectrum of SiH2 generated in a continuous discharge
͑40 mA, 1 kV͒ in a mixture of silane ͑5%͒ in argon. The total pressure was
about 1 Torr and the absorption equivalent path length was about 11 km
1
͑tgϭ60 ␮s͒. The band origin ␯ of the A B1(0, 0, 0)ϪX 1A1(0, 0, 0)
˜
˜
0
transition is indicated as well as some absorption lines from excited states of
Ar.
III. RESULTS AND DISCUSSION
The ICLAS spectrum of SiH2 was recorded between
15350 and 16100 cmϪ1. Figure 2 shows an overview of the
spectrum between 15350 and 15800 cmϪ1. The excellent
agreement between our ICLAS spectrum22 and the
l ϭ l/L͒ct ,
͑
eq
g
where (l/L) is the ratio of the length of the cell to the optical
length of the dye laser cavity.
spectrum reported by Dubois8 for the A 1B1(0, 2, 0)
˜
ϪX 1A1(0, 0, 0) transition, left no doubt about the assign-
˜
In the present study, we have inserted the plasma tube in
the dye laser cavity and used generation times ranging be-
tween 20 and 70 ␮s ͑absorption equivalent path lengths be-
tween 3.7 and 13.0 km͒. The occupation ratio of the cavity
by the cell was 62%.
In order to reduce the number of elementary spectra
needed to cover the whole band ͑750 cmϪ1), we did not use
in this work our spectrograph in its higher resolution con-
figuration. As a consequence, the observed linewidth ͓half
width at half-maximum ͑HWHM͒ of about 0.05 cmϪ1͔ is
mainly determined by the apparatus function of the spec-
trograph, the Doppler linewidth ͑HWHM of about 0.017
cmϪ1) being almost negligible.
ment of this spectrum to the SiH2 species and was confirmed
by ground state combination differences.
A. Assignments
The central region of the spectrum provided a clue to
start the assignment process. This band is of c type, with
selection rules ⌬Ka odd, ⌬Kc even ͑see Fig. 1 for the defi-
nition of the axis labels͒. However, we use symmetric top
formalism, with the K subindex representing KЉa , to label
those series for which there is no possibility of confusion.
Figure 3 presents the spectrum of this central region, show-
ing also a computer simulation of the spectrum. The pQ1 and
The wave number calibration was achieved by simulta-
neous recording of the iodine absorption spectrum.27 We es-
timate the wave number accuracy in the SiH2 line positions
to be about 0.01 cmϪ1. We have checked the quality of our
wave number calibration on some absorption lines from ex-
cited states of Ar which appear in our spectra, and from some
rovibronic transitions of NH2 which are observed near 15900
28
cmϪ1
.
This species is probably produced in the discharge
from N2 present as a sample impurity.
Our previous study of the
1B1(0, 2, 0)
˜
A
22
ϪX 1A1(0, 0, 0) band allowed us to calculate the SiH2
concentration in the discharge cell. We obtained a value of
the order of 1010 SiH2/cm3 and estimated our detection limit
to be about 108 SiH2/cm3. From the Franck–Condon factors
˜
reported in Ref. 13, we calculate that the intensity
1
of the presently studied A B1(0, 0, 0)ϪX 1A1(0, 0, 0)
˜
˜
transition is 22% of that of the
ϪX 1A1(0, 0, 0) transition.
1B1(0, 2, 0)
˜
A
FIG. 3. Comparison of the ICLAS spectrum of SiH2 with the simulated
spectrum obtained using the rotational constants of Table II.
˜
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Thu, 27 Nov 2014 18:55:19

6252
J. Chem. Phys., Vol. 108, No. 15, 15 April 1998
R. Escribano and A. Campargue
TABLE I. Energy of the perturbed ͑J KЈ KЈ) rotational states of the
rQ0 branches stand out clearly. Note that in the present case,
it is safe to use this notation because of the lack of asymme-
try splitting when KЈa or KaЉ are equal to zero. The J assign-
ment for each of these two branches was facilitated by the
spin statistical weights of the rotational levels of this radical:
KaЉϩKcЉ odd lines have three times as much weight as KaЉ
ϩKcЉ even. Thus, consecutive lines in the above-mentioned
series have alternating weights, those with J even being the
Ј
a
c
1B1(0,0,0) state, and energy shift ⌬ ͑in 10Ϫ2 cmϪ1) due to the pertur-
˜
A
bation.
⌬
J
KaЈ
KcЈ
EЈ (cm-1)
͑10Ϫ2 cmϪ1
)
Na
Ј
v
3
4
5
3
4
7
8
3
5
8
4
6
8
4
5
7
6
7
6
8
5
6
8
5
6
7
1
1
1
2
2
2
2
2
2
2
3
3
3
3
3
3
4
4
4
4
5
5
5
5
5
5
3
4
5
1
2
5
6
2
4
7
1
3
5
2
3
5
2
3
3
5
0
1
3
1
2
3
15610.33
15642.34
15682.75
15655.63
15691.13
15851.58
15924.12
15655.74
15732.75
15912.90
15914.98
15853.22
15984.83
15758.19
15801.02
15914.37
15946.46
16005.38
15946.05
16076.27
16008.55
16061.19
16192.36
16008.53
16061.23
16122.53
9
Ϫ15
Ϫ6
15
2
3
3
4
5
3
3
3
4
2
4
5
2
4
4
5
5
2
2
3
2
3
4
3
2
4
p
stronger in Q1 , while those with J odd are the stronger in
rQ0 . This figure can be compared to Fig. 3 in Ref. 16, where
the central region of the 020-000 band is shown. Both figures
are quite similar, the main differences being that in our spec-
37
Ϫ53
Ϫ59
60
Ϫ12
14
r
trum the Q0 branch is folded over at a lower J value, and
that the bandhead of Q1 is less complex.
To ensure our initial assignments, we searched for tran-
sitions reaching the same excited levels in order to provide
the appropriate ground state combination differences. We
p
14
Ϫ59
Ϫ52
38
Ϫ31
Ϫ16
53
50
13
Ϫ23
Ϫ12
20
p
p
p
thus found the P1 and R1 series to accompany Q1 , and
the pP2 series ͑with KЈc ϭKcЉϭ1͒ to confirm Q0 . Once a
r
few initial assignments were thus settled, we proceeded in a
cautious way, looking gradually for series with higher Ka
values, accepting only those transitions which we could
check by combination differences ͑unless the assignment
was indisputable͒, and refining the excited state rotational
constants after each assignment step. We have assigned tran-
sitions up to Jϭ16 and Kaϭ9, with ⌬Ka up to 5, ⌬Kc up to
4.
68
Ϫ14
26
16
The only previously available parameters for the excited
state of this transition came from a graphical extrapolation
aNumber of observed transitions used to calculate the upper state energy
level.
carried out by Dubois8 from his data on higher Ј states. We
v2
used Dubois’ parameters in our initial prediction, which gave
a reasonable agreement with the observation, apart from a
shift of 14 cmϪ1 in the band origin wave number, which was
immediately noticed.
and an approximate value of the perturbation induced dis-
placement. Whereas in some cases the perturbation presents
the typical look of a 2 by 2 interaction, in other cases the
pattern is much more complex, which would indicate that
several levels are interacting simultaneously. An example of
the former is given by the 313 and 414 levels, which are
pushed down ͑to lower energy͒ and pushed up ͑to higher
energy͒, respectively, while the 515 level is pushed up with a
smaller displacement. Besides, the 313←303 transition, dis-
played in Fig. 3, appears not only displaced from its pre-
dicted position, but with an intensity loss of almost 30% of
its expected value. This intensity must have been transferred
to a transition to the perturber level. Indeed, the comparison
of the experimental and calculated spectra ͑Fig. 3͒ shows
that a number of lines carrying significant intensity remain
unassigned.
B. Perturbations
Even for low Ka transitions, we soon found that large
perturbations were present in the spectrum. Pairs of strong
lines, which gave good ground state combination differences,
were poorly reproduced in a fit to the excited state param-
eters, with observed-calculated deviations up to a few tenths
of a wave number. The perturbations had a very localized
nature, affecting just some J levels in a given Ka ,Kc series.
Already Dubois8 had found perturbations in his spectra of
higher Ј vibrational states, which were later interpreted by
v2
Dubois, Duxbury and Dixon9 and Duxbury, Alijah and
The rotational structure of the ͑0,0,0͒ state studied here
Trieling.4 Figures 3 and 4 of this last study illustrate the
is much less perturbed than that of the ͑0, Ј,0͒ states ob-
v2
effect of these perturbations for several K values in Јϭ2
while Table IX gives observed and calculated energies for
v2
served by Dubois. For comparison, we have performed a fit
of the rotational constants of the ͑0,0,0͒ state to reproduce
the whole set of assigned transitions, including those given
in Table I, whose upper state is clearly perturbed. The stan-
dard deviation of this fit was 0.18 cmϪ1, to be compared to
the values from 2 to 8 cmϪ1 obtained for the standard devia-
several rotational levels for Ј up to 8. These authors devel-
v2
oped a model to describe the Renner–Teller interactions and
could explain the relatively smaller value of the perturba-
tions for Јϭ1 to 4, and their increasing effect for higher Ј
v2
v2
values. They concluded that the local perturbations observed
Ј
Ј
v
tion of fits to Dubois’ data for ͑0, ,0͒ levels, with 2ϭ1 to
v2
for Јϭ2, 3 and 4 must be due to spin-orbit coupling be-
5.4,8 The maximum individual perturbation shift observed in
our data is 0.6 cmϪ1 for the 322 level, where deviations of
several wave numbers were detected in the study of higher
bending levels. Duxbury, Alijah and Trieling4 concluded that
v2
1
3
tween the B1 state and the B1 state.
We present in Table I a summary of the perturbations
observed in our spectra, indicating the rotational assignment
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Thu, 27 Nov 2014 18:55:19

J. Chem. Phys., Vol. 108, No. 15, 15 April 1998
R. Escribano and A. Campargue
6253
TABLE II. The rovibrational parameters of the ͑ 3) ϭ͑0,0,0͒ level
,
,
v
v1 v2
the local perturbations ͑and the anomalous radiative life-
times͒ in the 1B1 state of SiH2 were due to indirect spin-orbit
of the X 1A1 and A 1B1 electronic states of SiH2 . The ground state pa-
rameters were fitted from the data of Yamada et al. ͑Ref. 10͒, Dubois ͑Ref.
8͒ and our present data.a Parameters without uncertainties were constrained
˜
˜
1
3
coupling between the B1 and B1 states, mediated by the
Renner–Teller coupling between 1A1 and 1B1 . According to
these authors, a direct coupling of the 1B1 with either 3B1 or
1A1 could not reproduce the observed erratic perturbations:
the direct spin-orbit coupling between the 1B1 and 3B1 states
would be very weak because of the strong similarity of the
bending potential energy curves of these two states; and the
to 0. All values in cmϪ1
.
1A1
1B1
˜
A
˜
X
␯
0
15547.7730͑93͒
18.3241͑21͒
4.89951͑87͒
0
A
B
C
␦
␦
⌬
⌬
⌬
⌽
⌽
⌽
⌽
8.09617͑25͒
7.02262͑25͒
3.70058͑21͒
1
1
3.76611͑31͒
direct Renner–Teller coupling between A1 and B1 would
0.20238͑89͒ϫ10Ϫ3
0.412͑28͒ϫ10Ϫ4
0
J
produce smooth perturbations for Ј less than 5 and not ir-
v2
0
K
regular perturbations as observed.4 In the frame of this
model, we believe that the decrease of the strength of the
0.4341͑22͒ϫ10Ϫ3
0.895͑53͒ϫ10Ϫ4
Ϫ0.337͑63͒ϫ10Ϫ3
0.3516͑14͒ϫ10Ϫ1
0
J
Ϫ0.14013͑89͒ϫ10Ϫ2
JK
K
perturbations in ͑0,0,0͒ compared to the other ͑0, Ј ,0͒
0.22995͑86͒ϫ10Ϫ2
v2
0
0
0
0
0
states, must be attributed to the weakening of the Renner–
J
Ϫ0.399͑70͒ϫ10Ϫ5
0.235͑12͒ϫ 10Ϫ4
0.1843͑28͒ϫ 10Ϫ3
Ϫ0.816͑19͒ϫ10Ϫ6
0.038
JK
KJ
K
Teller coupling when Ј decreases, which consequently soft-
v2
ens the effect of the indirect spin-orbit coupling with 3B1 . It
would be interesting to check if the Hamiltonian describing
the Renner–Teller coupling given in Ref. 4 should reproduce
quantitatively the large decrease ͑about one order of magni-
tude͒ in the strength of the perturbation going from ͑0,1,0͒ to
͑0,0,0͒.
LK
rms
0.045
aThe total number of data included in the fit is 1315 corresponding to 54
transitions of Ref. 10 affected with weight 10, 988 data of Ref. 8 with a
weight 0.1 and 273 transitions of our work with weight 1. With these data,
851 ground state combination differences were formed.
Since we have been unable to identify transitions to the
‘‘dark’’ band ͑or bands͒, we cannot add more to the inter-
pretation of the perturbations at this stage. Our present re-
sults could perhaps be used in the future, if more data be-
Table II. We found that the best results were obtained when
up to quartic centrifugal distortion constants only were taken,
and even in this case the off-diagonal ␦K was not well deter-
mined and we fixed it to zero. This choice is in contrast with
that of Yamada et al.,10 which used a much larger parameter
set for a limited number of combination differences. They
needed to release up to ten constants in order to reproduce
their very accurate data within experimental accuracy. We
have preferred not to force such a reproducibility of the
highly accurate data, with the aim of reproducing a very
large set of data with a limited number of well defined, cor-
relation free parameters.
Our rotational constants A, B and C are closer to those
of Dubois8 than to those of Yamada et al.,10 because of the
effect of the larger data set we used. The quartic centrifugal
distortion constants are very close to those of Yamada et al.,
indicating that they are well determined by the data. Dubois
3
come available, especially from the B1 state, to enable a
more comprehensive understanding of these perturbations.
C. Vibronic parameters
Although the perturbed data cannot be used to estimate
the rovibrational parameters of the excited state within the
standard Watson rotational model, they can still provide in-
formation on the parameters of the ground state, via the ap-
propriate ground state combination differences. Thus, we
have at our disposal data from three sources ͑Refs. 8, 10 and
this work͒, with different degrees of accuracy, which can be
used together in the most comprehensive fit of the ground
state constants of silylene.
We first checked the data from Dubois relative to the
͑010͒, ͑020͒ and ͑030͒ levels,8 band by band, in order to
eliminate eventual missassignments or misprints in the pub-
lication of the tables. To these we added our data from this
used the ␶
parameters instead of the ⌬’s and ␦’s more
abcd
frequently used today, which makes the comparison less
straightforward. Although our results are not very different
from those previously published, we believe that the param-
eters of Table II provide a reliable and accurate representa-
tion of the rotational constants of SiH2 in the electronic and
vibrational ground state.
work, and those of Yamada et al.10 on the ␯ fundamental.
2
Considering the different accuracy of each data set, we gave
relative weights of 0.1, 1 and 10 to individual transitions
belonging to each set, respectively. A few lines that were
assigned as blends of two or more SiH2 transitions were
given a weight of one tenth of that of the pure unblended
lines ͑0.1 for our data͒. The program was designed to search
for ground state combination differences within each set.
Weights for the combination differences were taken as the
mean of the individual weights of each transition. Altogether
we had a set of 851 combinations of non-null weight, from
which we refined the ground state parameters. We performed
several fits for different choices of parameters. Our aim was
to obtain a set of parameters giving a reasonable fit of all
data, being well determined in the fit, and which were not
highly correlated among them. Our final choice is shown in
Once these parameters were fixed, we aimed to deter-
mine a set of rotational parameters for the vibrational ground
1
state of the B1 electronic state. In order to do this, we ex-
tracted from the data set used for the determination of the
ground state parameters, those transitions which were clearly
perturbed, and which could not be reproduced without the
adequate model to treat the spin-orbit or Renner–Teller per-
turbations. It may be somewhat subjective to decide the
threshold within which a line can be labeled as perturbed or
not. Looking for a more objective way, we decided to carry
out a refinement process in which lines whose observed
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Thu, 27 Nov 2014 18:55:19

6254
J. Chem. Phys., Vol. 108, No. 15, 15 April 1998
R. Escribano and A. Campargue
1
TABLE III. Rotational assignments and wave numbers of the A B1(0,0,0)–X 1A1(0,0,0) rovibronic transitions of SiH2 ͑obs.-calc.͒ values are given in
10Ϫ3 cmϪ1. The lines marked by an asterisk or two asterisks were given a weight 0.1 and 0, respectively, in the fit of the upper state rotational constants. A
weight of unity was attributed to the other lines. The last column gives the calculated line intensities in arbitrary units.
˜
˜
o.-c.
13
Ϫ6
0
Ϫ47
Ϫ507
4
54
80
Ϫ7
Ϫ6
37
10
14
57
22
2
39
Ϫ32
33
Ϫ53
11
Ϫ23
Ϫ140
17
Ϫ573
Ϫ1
Ϫ132
Ϫ142
Ϫ53
Ϫ19
4
Int.
83
219
74
67
24
38
94
o.-c.
Ϫ13
20
404
Ϫ1
Ϫ5
631
Ϫ4
14
500
5
41
Ϫ76
Ϫ2
Ϫ174
16
Ϫ4
28
Int.
537
317
220
1595
56
J
J
Ј
Ј
J
␴ (cmϪ1
)
J
␴ (cmϪ1
)
KaЈ
KcЈ
Љ
KaЉ
KcЉ
KaЈ
KcЈ
Љ
KaЉ
KcЉ
6
6
8
9
7
7
8
8
4
6
7
4
6
8
5
8
5
3
7
5
4
4
7
6
6
6
4
5
2
3
3
5
4
6
2
2
8
2
3
1
6
3
5
1
1
7
6
8
7
0
5
6
7
5
8
2
4
6
9
2
0
1
1
1
2
1
0
1
0
1
1
1
0
0
1
1
0
0
1
1
1
0
3
0
3
1
1
2
0
0
1
0
2
3
0
1
1
1
2
0
0
2
3
1
1
2
3
2
4
0
0
1
1
1
1
2
1
1
1
2
1
3
0
0
0
0
1
6
5
7
9
5
6
8
8
4
5
7
3
6
8
4
7
5
3
7
5
3
4
5
6
3
5
4
4
2
3
2
5
2
4
2
1
7
2
1
1
6
2
3
0
1
5
3
6
4
0
5
5
6
4
7
0
3
5
8
1
9
1
3
2
4
1
10
7
7
9
3
4
2
2
3
2
1
2
3
2
2
4
1
3
2
4
1
3
2
2
2
1
4
3
4
4
2
3
3
1
2
3
3
4
1
2
4
2
3
1
3
3
4
2
2
3
2
5
5
1
3
2
2
2
2
3
2
4
2
3
2
4
1
1
1
1
2
4
3
7
9
5
6
8
8
2
5
7
1
6
6
4
5
5
1
5
5
3
4
5
4
3
3
4
4
0
3
2
3
2
4
2
1
3
2
1
1
2
2
3
0
1
5
5
2
4
0
1
5
6
4
7
0
3
1
8
1
9
1
3
2
4
1
10
15367.674
15374.060
15401.606
15404.825
15408.342**
15415.069
15415.821*
15418.854*
15427.650
15429.728
15433.050
15437.441
15441.980
15444.796
15445.894
15446.031
15455.390*
15456.900
15462.209
15462.660
15463.878
15469.144
15470.983**
15471.248
15473.113**
15473.706*
15478.153**
15478.886**
15481.948
15483.700
15484.176
15484.570
15485.466**
15495.460
15499.471
15505.664
15509.271
15510.789
15513.185**
15516.515
15519.317
15519.953**
15520.555**
15525.385
15526.707
15527.056**
15528.772**
15529.642**
15530.148
15532.605
15532.925
15534.214
15534.498
15535.36*
15535.838
15537.585
15537.887*
15537.887*
15538.149
15539.077
15541.123
15541.631**
15543.495
15543.699
15544.068
15544.592
15544.753**
5
2
4
6
5
0
1
3
0
3
1
0
0
4
0
1
1
1
1
1
0
1
1
2
1
0
1
1
2
2
0
0
1
1
0
2
0
0
0
1
0
4
2
0
0
1
1
0
1
2
1
1
0
2
1
2
1
1
5
1
1
0
1
2
3
3
1
1
1
2
3
1
5
1
2
6
2
11
7
8
2
9
6
5
7
4
8
10
9
3
2
10
11
2
11
9
10
12
2
12
1
4
7
3
5
13
13
6
3
6
5
2
5
6
6
1
2
4
1
2
2
1
1
5
1
0
0
0
0
0
1
0
0
3
0
1
0
0
1
1
1
1
0
0
1
1
1
1
1
0
1
5
1
1
1
0
2
1
2
1
2
0
1
3
2
3
2
2
6
2
2
1
0
1
4
4
2
0
2
3
2
0
5
1
2
6
4
11
7
8
2
9
6
5
7
4
8
10
9
3
2
10
11
2
11
9
10
12
0
12
1
2
7
1
3
13
13
4
3
6
15545.264
15546.312
15546.598**
15546.852
15547.745
15548.632**
15548.632
15550.518
15551.940**
15552.408
15552.570*
15552.570*
15552.835
15553.079**
15553.527
15554.322
15554.526*
15554.526**
15555.340**
15555.791
15556.302*
15556.302*
15557.275
15557.569**
15557.670**
15558.322
15558.525
15558.973*
15559.105*
15559.186*
15559.308**
15559.492
15559.692
15560.400*
15560.828
15560.950*
15560.950**
15561.426
15562.373
15562.536
15562.882
15563.922
15564.126
15564.269
15564.493*
15564.963*
15564.983**
15564.983*
15565.515**
15566.538
15566.682**
15566.682*
15566.682*
15567.092**
15567.092
15567.092**
15567.441
15567.441
15568.571**
15569.253
15570.093*
15570.093*
15570.943
15571.505
15573.352**
15574.635
15578.675
10
8
8
9
9
5
7
8
5
7
8
6
8
6
4
7
6
5
5
8
6
7
6
5
6
3
4
4
5
5
7
3
3
7
3
4
2
5
4
6
2
2
7
7
7
8
1
4
6
7
5
8
3
4
5
9
3
10
5
3
2
4
1
11
12
7
12
7
134
495
1300
206
358
637
2032
1694
669
475
840
1142
1791
185
292
208
445
641
120
253
443
192
150
669
304
214
103
75
34
8
6
9
6
5
7
4
8
0
9
3
4
8
7
9
6
5
7
4
8
187
181
148
223
181
70
98
75
80
58
10
9
3
4
101
370
38
16
9
64
286
488
318
187
92
150
80
125
431
77
10
11
2
11
10
11
12
2
10
11
2
11
10
11
12
1
32
402
472
16
Ϫ13
34
12
1
12
1
40
16
141
243
29
4
8
3
5
3
8
2
4
157
Ϫ5
27
Ϫ9
5
40
101
64
11
385
Ϫ73
Ϫ22
Ϫ4
Ϫ56
51
132
Ϫ10
Ϫ83
595
Ϫ303
19
527
105
552
823
159
180
230
192
90
723
482
246
568
32
13
13
6
13
13
5
100
303
159
109
802
37
194
65
92
6
7
7
7
7
6
7
5
Ϫ89
Ϫ77
1
14
14
7
14
14
6
8
5
5
7
15
15
1
6
6
7
4
3
8
4
16
16
4
0
1
3
0
2
5
14
14
6
14
14
4
6
3
5
5
15
15
1
4
4
5
2
3
6
2
16
16
4
0
1
1
0
0
3
27
8
6
6
8
7
5
6
7
Ϫ2
Ϫ12
Ϫ21
Ϫ32
Ϫ121
52
126
Ϫ10
Ϫ617
24
Ϫ3
657
Ϫ42
Ϫ150
66
Ϫ44
Ϫ136
Ϫ42
Ϫ48
36
79
427
314
177
121
40
52
36
229
90
192
42
23
33
72
24
1024
262
817
549
157
102
132
182
671
10
Ϫ578
Ϫ582
Ϫ579
16
Ϫ9
47
Ϫ1
52
29
6
21
Ϫ9
42
11
151
167
184
434
30
15
15
3
15
15
3
6
8
7
5
8
8
4
5
7
6
4
9
7
3
495
144
181
365
799
555
153
100
268
239
543
447
1053
1526
221
61
16
16
5
16
16
5
28
18
3
3
4
1
3
7
9
2
4
4
3
0
2
6
9
1
10
4
139
Ϫ30
Ϫ37
Ϫ25
Ϫ7
283
3
2
4
1
24
38
Ϫ601
7
7
1
7
1
11
16
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Thu, 27 Nov 2014 18:55:19

J. Chem. Phys., Vol. 108, No. 15, 15 April 1998
R. Escribano and A. Campargue
6255
TABLE III. ͑Continued.͒
o.-c.
589
Ϫ119
25
Ϫ2
126
527
55
Ϫ157
68
45
Ϫ78
10
51
371
Ϫ77
198
369
25
Ϫ302
42
132
716
363
154
Ϫ55
Ϫ53
Ϫ53
Ϫ5
Int.
o.-c.
Int.
J
J
Ј
Ј
J
␴ (cmϪ1
)
J
␴ (cmϪ1
)
KaЈ
KcЈ
Љ
KaЉ
KcЉ
KaЈ
KcЈ
Љ
KaЉ
KcЉ
3
8
3
5
8
6
2
7
7
5
5
4
2
4
6
3
4
5
5
2
3
8
4
7
4
4
9
6
4
3
8
8
8
6
7
6
3
4
5
7
6
6
6
8
6
7
4
5
8
4
6
3
3
7
8
5
7
4
5
8
5
8
5
6
8
6
7
7
4
5
2
3
1
4
2
4
2
3
2
4
2
1
2
2
3
2
2
1
3
2
2
5
2
5
4
4
3
1
3
2
2
3
4
3
3
3
3
3
2
4
5
5
2
1
3
2
3
3
4
3
3
3
3
3
4
5
4
3
5
6
3
3
4
3
4
4
3
4
3
4
2
6
2
1
7
2
1
5
6
2
4
3
0
2
4
1
2
4
3
0
1
3
2
3
0
1
7
5
2
2
6
5
5
3
5
3
1
1
4
4
1
2
5
7
3
6
2
3
4
1
4
0
1
5
4
0
3
2
1
2
3
6
2
4
5
2
5
3
1
1
3
8
2
6
7
7
2
7
6
6
4
3
2
3
6
3
4
4
5
1
2
9
3
8
5
5
8
5
4
2
8
8
8
6
6
5
3
4
4
7
7
7
5
7
5
6
4
5
8
3
6
2
2
7
7
6
7
3
6
9
4
8
5
5
7
6
6
6
3
5
1
2
0
5
3
3
1
2
3
5
3
0
1
3
2
1
1
0
2
1
1
4
1
6
5
5
4
0
2
1
1
2
3
2
4
4
2
2
1
3
6
6
1
0
2
1
0
0
3
2
0
2
2
0
5
6
3
2
6
7
2
0
3
2
5
3
2
5
0
3
2
6
2
1
5
4
1
5
4
2
2
3
2
0
4
3
4
4
3
0
1
5
2
3
0
1
5
5
2
2
8
7
5
5
3
1
1
3
4
4
1
2
5
7
3
6
4
5
6
1
6
0
1
7
2
0
5
2
1
2
3
8
2
4
3
4
5
1
3
3
15581.571**
15581.571**
15587.355
722
88
181
144
54
7
4
8
6
9
7
7
6
7
5
6
7
8
8
5
9
7
5
4
4
6
6
7
9
4
5
8
6
6
8
8
5
7
6
6
4
4
4
3
4
4
6
4
6
4
4
5
5
3
4
4
4
4
4
4
6
6
5
7
4
4
4
5
4
7
7
5
5
5
5
6
6
7
7
5
5
6
5
5
6
5
6
6
8
8
6
6
6
6
6
7
6
7
6
6
6
7
6
7
6
7
9
9
7
4
0
5
3
6
4
1
3
2
2
2
3
3
5
1
6
4
2
0
1
0
1
2
3
0
1
4
1
2
1
2
1
3
1
2
4
2
1
0
0
1
2
3
2
4
3
4
2
0
1
1
2
0
1
2
1
1
3
0
4
2
3
2
2
3
3
0
1
1
6
4
8
5
9
7
8
6
8
5
5
7
8
7
4
8
6
4
3
3
7
7
7
5
3
1
0
1
1
7
1
7
1
3
4
4
0
3
3
3
3
3
3
7
7
4
8
1
1
1
4
1
8
8
2
4
4
4
7
5
8
8
4
4
5
2
2
3
2
5
5
9
9
5
5
5
5
3
6
3
4
3
3
3
4
1
6
1
6
10
10
6
2
2
7
5
8
6
1
5
2
4
2
3
5
7
1
6
4
2
0
1
0
1
4
3
2
3
8
3
4
1
2
3
3
1
2
2
4
1
0
0
1
2
5
4
6
5
4
2
0
1
1
2
0
1
4
3
3
5
2
6
4
7
6
2
7
3
0
1
1
15704.058
15714.369*
15722.635**
15723.147**
15723.393
15723.773
15725.369
15726.115**
15726.264
15729.597
15740.788**
15741.977**
15748.869**
15751.047**
15751.276
15752.408
15753.042
15757.958
15758.475
15760.029
15760.933
15761.029
15763.839*
15772.044*
15775.970
15778.059
15779.879
15780.686**
15782.284**
15812.697
15813.082
15818.611**
15841.627**
15844.825**
15849.625**
15850.006**
15850.680
15851.087
15851.087
15854.842**
15855.676**
15863.206
15867.818**
15868.236
15869.732*
15902.140**
15929.127
15932.893
15951.359
15951.359
15951.909
15954.890
15964.678
15965.131
15970.001
15977.321
15979.563
15984.676
15990.796
16003.394
16003.874
16011.691**
16044.334
16049.526
16052.237*
16056.147
16059.664
16059.664
16070.766
6
Ϫ16
Ϫ271
Ϫ621
Ϫ6
13
33
151
54
68
532
154
701
Ϫ518
16
167
199
105
253
265
336
72
102
229
215
259
111
208
147
188
81
333
860
936
358
379
127
80
15587.890
15588.405**
15589.044**
15589.332
15589.931**
15590.695
15590.922
15597.008
15597.970
15599.349
15599.473**
15599.843
15599.843**
15601.886**
15610.324
161
108
361
204
512
192
375
300
245
151
178
646
89
37
17
49
15611.131**
15614.142
485
667
243
93
Ϫ42
Ϫ42
Ϫ26
Ϫ15
Ϫ53
36
Ϫ39
12
50
15615.750**
15616.185**
15616.957**
15617.480**
15618.296
15618.715
15618.999
15624.143
15625.289**
15625.831**
15627.398**
15631.216**
15633.043**
15633.230**
15633.870**
15636.858**
15639.267*
15642.107**
15643.500**
15649.468
15650.034**
15651.778**
15654.397
15655.606
15663.305**
15666.296
15668.975**
15670.986**
15674.139
15674.287**
15675.290
15677.455
15679.912
15680.582**
15682.100
15682.712**
15682.897**
15682.897**
15682.897**
15684.124
15685.148**
15687.368
15688.093
532
202
690
236
149
197
127
685
295
495
40
691
236
319
194
524
478
187
320
119
122
113
123
271
61
258
348
636
96
280
983
280
189
180
139
300
546
89
10
3
4
8
6
5
9
9
5
6
5
5
9
8
8
8
4
4
7
7
6
9
6
8
7
9
9
6
6
5
5
7
8
6
9
5
8
6
10
7
8
8
8
10
10
7
62
180
127
95
219
410
131
45
197
518
Ϫ1
325
606
Ϫ581
Ϫ525
Ϫ205
Ϫ605
Ϫ191
Ϫ607
Ϫ31
149
Ϫ41
Ϫ162
164
198
277
366
Ϫ49
18
62
417
502
239
64
123
242
80
312
1013
72
279
121
129
172
111
144
141
47
143
544
821
284
251
68
74
66
85
146
142
57
57
36
20
174
25
75
332
10
8
7
7
5
5
7
8
7
9
7
9
8
8
8
7
7
6
6
8
8
7
9
6
9
7
10
8
9
9
9
9
9
8
Ϫ133
13
188
55
Ϫ116
Ϫ131
32
684
Ϫ43
5
168
Ϫ26
Ϫ82
Ϫ7
Ϫ23
47
42
Ϫ21
15
Ϫ51
Ϫ8
64
242
Ϫ43
Ϫ50
Ϫ596
47
397
Ϫ314
75
168
Ϫ69
Ϫ5
Ϫ9
Ϫ167
29
Ϫ128
498
330
78
Ϫ48
2
Ϫ135
Ϫ82
Ϫ314
Ϫ80
74
Ϫ55
Ϫ230
569
701
81
206
161
61
434
313
68
265
120
53
Ϫ91
Ϫ23
Ϫ17
Ϫ40
49
15689.324
15691.211**
15692.656**
15694.438**
15701.578**
15702.195**
15702.968
Ϫ132
498
173
6
Ϫ1
Ϫ58
27
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Thu, 27 Nov 2014 18:55:19

6256
J. Chem. Phys., Vol. 108, No. 15, 15 April 1998
R. Escribano and A. Campargue
TABLE IV. Rotational constants ͑in cmϪ1) and geometrical structures for
cmϪ1 for the bending wave number in A 1B1. This value is
˜
the X 1A1 and A 1B1 states of SiH2 . The rotational constants at equilib-
˜
˜
fairly different from the ab initio result of Allen and Schaefer
͑Table 18 of Ref. 25͒, 888 cmϪ1. The spacing between con-
secutive bending states as measured by Dubois is quite ir-
regular, probably because of the Renner–Teller perturba-
tions, which has refrained us from attempting to evaluate
anharmonicity constants.
B
rium were calculated using the ␣i rovibrational constants computed by
Allen and Schaefer ͑Ref. 25͒ ͑set A in Table 15 and Table 20 for X 1A1 and
˜
1B1, respectively͒. The re ͑Si-H͒ and ␪ ͑H-Si-H͒ values were derived
˜
A
e
from the three pairs of rotational constants and are given in Å and degrees,
respectively.
1A1
Be
1B1
Be
˜
X
˜
A
We present in Table III a list of all observed and as-
signed transitions, with the observed-calculated deviations
and an indication of the lines that were left out of the fit,
because of the perturbations. It is hoped that these line posi-
tions will help future investigations aimed, for example, at
monitoring silylene in plasmas. We add also in this Table a
calculated value of the intensity of each line, which is nec-
essary for concentration measurements of silylene ͑see Ref.
22͒. As another example of the importance of the inclusion
of intensities in such tables, we would mention the difficulty
that we sometimes had in identifying NH2 lines in our spec-
trum; their wave number was given in Ref. 28, but since no
intensities were given in that reference, it was difficult at
times to ascertain whether a weak line could be due to an
NH2 absorption or to an unassigned silylene line.
Ae
Ce
Ae
Ce
8.09813 7.04917 3.77198 17.71856 4.94926 3.84201
re
␪
re
␪
e
e
(A,B)
1.51446
1.51377
1.51382
1.51402
91.9807
92.0348
91.9336
91.9830
1.48184
1.48725
1.48688
1.48532
122.6231
121.8658
122.8359
122.4416
(B,C)
(A,C)
Average
Dubois
ab initio
1.51408a
1.51477b
91.990a
92.42b
1.4871c
121.83c
122.53d
1.48328d
aValues obtained in Ref. 25 ͑Set A of Table 15͒ from the ground state
rotational constants given by Dubois in Ref. 8.
bB2 2R CISD ab initio calculations ͑Table 10 of Ref. 25͒.
cValues obtained in Ref. 25 ͑Table 20͒ from the extrapolated values of the
rotational constants given by Dubois in Ref. 8.
dB2 CISD ab initio calculations ͑Table 18 of Ref. 25͒.
D. Equilibrium geometry
With the new rotational constants for the vibrational
ground states derived in this work, it is possible to recalcu-
late the equilibrium value of the rotational constants, and
hence the equilibrium geometry of the molecule in both elec-
minus calculated value was larger than 2.5 times the standard
deviation of the previous cycle, would be excluded from the
present iteration. The next choice was again that of the set of
parameters to be released in the fit. After several trials, we
finally chose a set which included basically quartic and sex-
tic centrifugal distortion constants, with the exclusion of ␦K
and HJ, which were poorly determined, and the inclusion of
LK, which was necessary to reproduce the higher K transi-
tions: 99,0←1010,0 and 99,1←1010,1 . The parameters result-
ing from this procedure are quoted also in Table II, with their
fitting uncertainty in parentheses.
1
tronic states involved. In the A1 state, the rotational con-
stants A0 ,B0 and C0 calculated here are very close to those
of Dubois,8 which makes the value of the inertial defect of
the ͑0,0,0͒ vibrational state of this work to be coincident with
the values observed and calculated by Dubois8 ͑0.0073 and
0.0074 uÅ2, respectively͒. We have derived the equilibrium
values of the rotational constants Ae ,Be and Ce by adding
the calculated values of the vibrational dependence of the
rotational constants ͑the ␣iB parameters͒ calculated by Allen
and Schaefer.25 From Ae , Be and Ce it is possible to estimate
the equilibrium geometry of SiH2 , given by the two param-
eters re and ␪e , either by taking two rotational constants at a
time, or by making a mean average of all three values. Our
results are collected in Table IV, together with those derived
by Allen and Schaefer,25 from the experimental values of
Dubois8 using the same procedure, and from their own ab
initio calculations. Although all three sets are in good agree-
ment, our values are closer to those of the experimental work
of Dubois.8
These values are the first experimental determination of
the parameters of the ͑0,0,0͒ vibrational level of the 1B1
electronic state. Dubois8 had estimated these parameters by
extrapolation from his data on higher bending states, and
Duxbury, Alijah and Trieling4 had obtained calculated values
for these parameters through their Renner–Teller model.
Whereas the extrapolation8 was not accurate for the A con-
stant ͑17.75 compared to the experimental value of 18.32
cmϪ1͒, which changes largely from one bending state to the
next, the extrapolated results for B and C were quite good
͑4.9 and 3.8 compared to 4.899 and 3.766 cmϪ1, respec-
tively͒, as these constants are much less sensitive to the
bending excitation. On the other hand, the Renner–Teller
model4 gives a value for the A constant ͑18.30 cm Ϫ1͒ which
is very close to the experimental one, while the B and C
parameters are calculated with less precision ͑4.722 and
3.651 cmϪ1, respectively͒. Duxbury, Alijah and Trieling4
also estimated the vibrational band origin of this transition at
1
For the B1 state, the inertial defect in the ͑0,0,0͒ vibra-
tional level calculated from our rotational constants is fairly
high: 0.115 uÅ2. This may be a consequence of the pertur-
bations still remaining in our data, although this number may
be comparatively normal for the electronic configuration of
the molecule in this state, where the bonds have a weaker
nature. We have transformed the rotational constants into
p
15546 cm Ϫ1 from the LIF excitation spectra of the P1(1)
B
rotational line observed by Fukushima, Mayana and Obi.13
Their value agrees with the present band origin within 1.7
cmϪ1. From our band origin and that of the ͑0,1,0͒ state from
Dubois ͑Table IV of Ref. 8͒, we obtain a value of 856.53
their equilibrium values by means of the ␣i parameters cal-
culated by Allen and Schaefer25 for this electronic level. The
geometrical parameters obtained in this way are also given in
Table IV. It is interesting to note that our values are closer
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Thu, 27 Nov 2014 18:55:19

J. Chem. Phys., Vol. 108, No. 15, 15 April 1998
R. Escribano and A. Campargue
6257
now to the theoretical results of Allen and Schaefer than to
those of Dubois, which were in fact estimated by extrapola-
tion from observations of higher bending states. Allen and
Schaefer questioned the reason of the less good agreement
between their theoretical results and the experimental ones
for this electronic state. Our results seem now to confirm the
high quality of their calculation.
3 J. R. Doyle, D. A. Doughty, and A. Gallagher, J. Appl. Phys. 68, 4375
͑1990͒.
4 G. Duxbury, A. Alijah, and R. R. Trieling, J. Chem. Phys. 98, 811 ͑1993͒.
5 J. W. Thoman and J. I. Steinfeld, Chem. Phys. Lett. 124, 35 ͑1986͒.
6 J. W. Thoman, J. I. Steinfeld, R. I. McKay, and A. E. W. Knight, J. Chem.
Phys. 86, 5909 ͑1987͒.
7 I. Dubois, G. Herzberg, and R. D. Verma, J. Chem. Phys. 47, 4262 ͑1967͒.
8 I. Dubois, Can. J. Phys. 46, 2485 ͑1968͒; Depository for unpublished data,
National Science Library, Ottawa, Canada ͑1968͒.
9 I. Dubois, G. Duxbury, and R. N. Dixon, J. Chem. Soc. Faraday Trans. 2
71, 799 ͑1975͒.
IV. CONCLUSION
10 C. Yamada, H. Kanomori, E. Hirota, N. Nishiwaki, N. Itabashi, K. Kato,
and G. Goto, J. Chem. Phys. 91, 4582 ͑1989͒.
The new experimental measurements presented here al-
low a more accurate determination of the energy and geo-
metrical structure of the 1B1 electronic state of silylene.
These results complement previous observations of excited
bending levels in this electronic state.4,8 The transitions ob-
served here could be used for monitoring this species in sili-
con plasmas. Local perturbations affecting the spectra are
probably due to crossings with rovibrational levels of the
3B1 electronic state, which has not yet been experimentally
observed. The detection of silylene in the triplet state would
be of great interest to clarify the energy level structure of this
radical.
11 G. Inoue and M. Suzuki, Chem. Phys. Lett. 105, 651 ͑1984͒; 122, 361
͑1985͒.
12 D. M. Rayner, R. P. Steer, P. A. Hackett, C. L. Wilson, and P. John,
Chem. Phys. Lett. 123, 449 ͑1986͒.
13 M. Fukushima, S. Mayana, and K. Obi, J. Chem. Phys. 96, 44 ͑1992͒.
14 J. M. Jasinski, E. A. Whittaker, G. C. Bjorklund, R. W. Dreyfus, R. D.
Estes, and R. E. Walkup, Appl. Phys. Lett. 44, 1155 ͑1984͒.
15 J. M. Jasinski, J. Phys. Chem. 90, 555 ͑1986͒.
16 J. M. Jasinski and J. O. Chu, J. Chem. Phys. 88, 1678 ͑1988͒.
17 A. Kono, N. Koike, K. Okuda, and T. Goto, Jpn. J. Appl. Phys., Part 2 32,
543 ͑1993͒.
18 H. Nomura, K. Akimoto, A. Kono, and T. Goto, J. Phys. D 28, 1977
͑1995͒.
19 J. J. O’Brien and G. H. Atkinson, Chem. Phys. Lett. 130, 321 ͑1986͒.
20 K. Tachibana, T. Shirafuji, and Y. Matsui, Jpn. J. Appl. Phys., Part 1 31,
2588 ͑1992͒.
ACKNOWLEDGMENTS
21 M. W. Markus and P. Roth, J. Quant. Spectrosc. Radiat. Transf. 52, 783
͑1994͒.
The expert help of Ludovic Biennier ͑Grenoble͒ was ap-
preciated for the recording of the ICLAS spectrum. Daniele
Romanini and Nader Sadeghi ͑Grenoble͒ are thanked for
helpful discussions and for their contribution in the design-
ing of the intracavity discharge cell. R.E.’s work is supported
by the Spanish DGES, Project No. PB96-0881.
22 A. Campargue, D. Romanini, and N. Sadeghi, J. Phys. D ͑accepted͒.
23 A. Campargue, F. Stoeckel, and M. Chenevier, Spectrochim. Acta Rev.
13, 69 ͑1990͒.
24 D. Romanini, A. A. Kachanov, N. Sadeghi, and F. Stoeckel, Chem. Phys.
Lett. 264, 316 ͑1997͒.
25 W. D. Allen and H. F. Schaefer III, Chem. Phys. 108, 243 ͑1986͒.
26 S. Koseki and M. S. Gordon, J. Mol. Spectrosc. 123, 392 ͑1987͒.
S. Gerstenkorn and P. Luc, Atlas du spectre d’absorption de la molecule
27
1 A. Matsuda and N. Hata, Glow-discharge Hydrogenated Amorphous Sili-
con, edited by K. Tanaka ͑KTK, Tokyo, 1989͒.
2 T. Shirafuji, K. Tachibana, and Y. Matsui, Jpn. J. Appl. Phys., Part 1 34,
4239 ͑1995͒.
´
d’iode ͑Edition du CNRS, Paris, 1978͒.
28 S. C. Ross, F. W. Birss, M. Vervloet, and D. A. Ramsay, J. Mol. Spec-
trosc. 129, 436 ͑1988͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Thu, 27 Nov 2014 18:55:19