Room temperature absorption spectroscopy of GeH2 near 585 nm

Article abstract of DOI:10.1016/S0009-2614(99)01237-3

The absorption spectrum of the germylene radical produced in a continuous flow discharge of germane diluted in argon has been recorded in the 17 090-17 135 cm^-1 region. This range corresponds to the central part of the A¹B₁(0, 1, 0)-X¹A₁(0, 0, 0) rovibronic transition, which has been analysed for the first time. The spectrum was measured by Intracavity Laser Absorption Spectroscopy with an equivalent pathlength of 15 km. The experimental resolution and wavenumber accuracy are of the order of 0.06 and 0.015 cm^-1, respectively. Each rovibronic absorption is made up of five lines corresponding to the five germanium isotopomers appearing in a natural abundance sample. We have assigned around 25 transitions for each species, from which the rotational constants of the excited state and the band origin of the transition have been calculated. The equilibrium structure of the excited electronic state has been estimated with the help of ab initio data for some vib-rotational parameters. Good agreement is found with previous ab initio determinations of this structure.

Full text of DOI:10.1016/S0009-2614(99)01237-3

3
1 December 1999
Chemical Physics Letters 315 Ž1999. 397–404
www.elsevier.nlrlocatercplett
Room temperature absorption spectroscopy of GeH near 585 nm
2
Alain Campargue a,), Rafael Escribano b,1
a
Laboratoire de Spectrom e´ trie Physique (UMR 5588), UniÕersit e´ Joseph Fourier de Grenoble, B.P. 87,
3
8402 Saint-Martin d’H e` res Cedex, France
b
Instituto de Estructura de la Materia, CSIC, Serrano 123, 28006 Madrid, Spain
Received 14 September 1999; in final form 15 October 1999
Abstract
The absorption spectrum of the germylene radical produced in a continuous flow discharge of germane diluted in argon
y
1
1
has been recorded in the 17 090–17 135 cm
region. This range corresponds to the central part of the A˜ B Ž0, 1, 0.y
1
1
X˜ A Ž0, 0, 0. rovibronic transition, which has been analysed for the first time. The spectrum was measured by Intracavity
1
Laser Absorption Spectroscopy with an equivalent pathlength of 15 km. The experimental resolution and wavenumber
y
1
accuracy are of the order of 0.06 and 0.015 cm , respectively. Each rovibronic absorption is made up of five lines
corresponding to the five germanium isotopomers appearing in a natural abundance sample. We have assigned around 25
transitions for each species, from which the rotational constants of the excited state and the band origin of the transition have
been calculated. The equilibrium structure of the excited electronic state has been estimated with the help of ab initio data
for some vib-rotational parameters. Good agreement is found with previous ab initio determinations of this structure. q 1999
Elsevier Science B.V. All rights reserved.
1
. Introduction
desirable method to understand the gas phase and
surface chemistries in these CVD plasmas of indus-
trial interest. In the case of silylene, we have recently
shown w5,6x that Intracavity Laser Absorption Spec-
troscopy ŽICLAS. and CW-Cavity Ring Down Spec-
troscopy ŽCW-CRDS. can be applied to detect SiH2
Similarly to silylene ŽSiH ., germylene ŽGeH . is
2
2
believed to be an important intermediate species in
the production of hydrogenated amorphous films, by
plasma-induced deposition or laser assisted chemical
vapour deposition ŽCVD. w1x. Germylene has been
detected both in deposited thin layers and as an
absorption product on the germanium surface w2,3x.
in the gas phase and measure its concentration.
These measurements performed in an argon–5%
silane DC discharge used some Doppler-limited ab-
The infrared detection, by matrix-isolation technique,
of germylene produced in deposition plasmas has
been reported by Lloret et al. w4x. However, in situ
1
1
˜
˜
sorption lines of the A B §X A electronic transi-
1
1
tion in the visible range. However, the extraction of
the absorber concentration from the measurement of
a single rovibronic absorption line requires both a
preliminary rotational analysis of the band under
consideration and a knowledge of the electronic tran-
sition moment w6x. Very similar experiments were
carried by Kawasaki et al. w7x who monitored the
detection of gas phase germylene would be the most
)
Corresponding author. Fax: q33-476514544; e-mail:
alain.campargue@ujf-grenoble.fr
1
E-mail: imtre28@fresno.csic.es
0
009-2614r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.
PII: S0009-2614Ž99.01237-3

3
98
A. Campargue, R. EscribanorChemical Physics Letters 315 (1999) 397–404
spatial distribution of the absolute GeH density in a
not rotationally assigned. to GeH from the charac-
2
2
P
RF plasma by ICLAS of the P Ž1. transition of the
teristic intensity variation of their components, due
to the five germanium isotopomers. These transitions
will be rotationally assigned in the forthcoming anal-
ysis.
1
1
1
y1
˜
Ž
.
˜
Ž
.
A B 0, 0, 0 yX A 0, 0, 0 band at 16 312 cm
However, no details were given in this report about
.
1
1
the procedure followed to obtain the absolute con-
centration of GeH w7x.
For quantitative in situ and real time optical diag-
nostics of CVD plasmas, absorption techniques are
much more suitable than LIF. For this purpose, the
rovibrational assignment of the absorption spectrum
2
As far as we know, a very limited number of
studies has been devoted to the spectroscopy of
germylene in the gas phase w8–12x. All these previ-
1
X
2
˜
Ž
.
Ž
ous reports focused on the A B 0, n , 0 y
recorded at room temperature i.e. close to real CVD
1
X
1
n
2
˜
Ž
.
Ž
.
.
X A 0, 0, 0 system noted 2 hereafter , consisting
conditions is a necessary prerequisite to monitor the
1
0
of a progression of bands between 615 and 485 nm
GeH concentration from a single line. In this con-
2
corresponding to the excitation of the bending vibra-
text, we present below the first rotational analysis of
X
tion Žn s0–6.. Saito, Obi and Fukushima w8,10,11x
a vibronic band of GeH recorded at room tempera-
2
2
1
0
y1
reported the laser-induced fluorescence ŽLIF. spec-
trum of jet cooled germylene produced by photolysis
of phenylgermane and measured the fluorescence
ture, namely the 2 band near 17 100 cm
.
p
lifetimes from the P Ž1. transition. Karolczak et al.
2. Experimental
1
w12x generated germylene by applying an electric
discharge at the exit of a supersonic expansion of a
The experimental set up is essentially the same as
mixture of GeH4 in argon. They observed a suffi-
the one used for our SiH studies w5,6x. The germy-
2
0
cient number of rovibronic transitions of the 0 band
lene radical was generated by a continuous discharge
Žabout 640 V. in a slowly flowing mixture Ž35 sccm.
of germane Ž5%. in argon. The total pressure was
0
to derive the rotational constants and then the r0
structures of the ground and excited states. They also
determined the upper state fundamentals, n and n ,
about 1 torr. The flow of argon entered the plasma
tube close to the two Brewster angle windows to
prevent deposition of germanium. The discharge cell
was inserted into the dye ŽRhodamine 590. laser
cavity of our ICLAS spectrometer Žsee Ref. w14x for
1
2
p
and studied the isotope effect from the P Ž1. lines of
the 20 progression. Another important result of this
1
X
2
n
work w12x is that, as a consequence of J-dependent
predissociation process, only fluorescence from low
J value rotational states could be detected. Indeed,
more details on the principle and experimental ar-
rangements of ICLAS.. The distance between the
p
only the P Ž1. line reaching the 0 excited rota-
1
00
X
2
n
tional level could be recorded for the 20 bands with
n )0, while only fluorescence from low angular
momentum states could be detected for the 0 band.
two cylindrical electrodes was 60 cm leading to an
occupation ratio of the laser cavity by the absorber
of about 50%. A generation time of 80 ms, corre-
sponding to an absorption equivalent pathlength of
15 km was adopted for the recording. The wavenum-
ber calibration was performed by simultaneous
recording of the iodine absorption spectrum w15x. The
X
2
0
0
Similar conclusions were previously derived by Obi,
Fukushima and Saito w11x who proposed a second
1
1
order predissociation mechanism ŽA B ™X A ™
1
1
3
3
a B . leading to GeŽ P.qH , to account for their
1
2
observations.
wavenumber accuracy in the GeH line positions is
2
y1
In the case of non- or weakly-fluorescent transi-
tions, high sensitive absorption spectroscopy is the
ideal alternative to LIF. Very recently, Becerra et al.
w9x and Alexander et al. w13x studied the germylene
estimated to be 0.015 cm
as checked for some
NH rovibronic transitions w16x which appear super-
2
imposed on the GeH2 spectrum Žsee Fig. 1.. The
NH2 radical is produced from N probably present
2
kinetics at room temperature by monitoring the time
dependence of the concentration of GeH2 produced
by a flash photolysis of phenylgermane. They ob-
served by direct multipass absorption, two rovibronic
in the discharge as a result of a small leak in the
discharge cell. The spectrum, recorded between
17 000 and 17 250 cm , appeared to be congested
y1
and highly dependent of the discharge conditions,
1
U
2
transitions of the 2 band, which were attributed Žbut
with many lines which could not be attributed to N ,
0

A. Campargue, R. EscribanorChemical Physics Letters 315 (1999) 397–404
399
r
Q Ž J. branches. The symmetric top formalism can
0
safely be used here, as both of these transitions are
made up of one K component only. It was soon
found that the central region of this band presents a
0
very similar structure to that of the corresponding 00
band of SiH w5x, whereas every feature is five-fold
2
in the present case. This was sometimes of help in
the assignment, but sometimes it greatly complicated
the structure. Fig. 1 shows the observed spectrum
Žabove. and a computer simulation Žbelow. including
p
all five isotopomers. Transitions labelled Q Ž4. or
1
p
Q Ž6. are very good examples of the former case;
the region at the head of the Q branch, where
Q Ž8. is also present, is on the other hand quite
1
r
0
p
1
Fig. 1. Upper: Observed ICLAS spectrum in the 17105–17123
complicated indeed. The presence of NH2 lines,
labelled with a cross in the figure, adds some further
y1
cm
region, of natural abundance GeH . The rotational assign-
2
ment is indicated for the more intense lines, corresponding to the
7
4
complexity to the problem.
Ge isotopomer. Lower: predicted spectra, with the weight of the
y1
The lines near 17 111.5 cm
observed by Be-
spectrum of each isotopomer scaled by its corresponding abun-
y1
w x
w x
dance ratio. A Gaussian line profile Ž0.06 cm
FWHM. was
cerra et al. 9 and Alexander et al. 13 were in
perfect wavenumber agreement with our observation,
attached to each transition of the simulated spectrum. The lines
marked by q are due to NH2 rovibronic transitions w16x.
p
and are assigned to the Q Ž6. Ge isotopomer struc-
ture. The line at 17 118.67 cm , used by Becerra et
1
y1
al. to monitor some experiments, coincides with one
U
r
NH , GeH or Ar . As a consequence, we were
of our observations assigned by us to Q Ž7. of the
2
2
0
7
4
obliged to restrict our analysis to the 17 090–17 135
spectral region, as described below, which
corresponds to the central region of the 2 band.
Ge isotopomer.
y1
cm
Altogether we have assigned around 25 transitions
1
0
for every isotopomer, as collected in Table 1. The
p
r
assignment of the Q Ž10. and Q Ž9. lines, both
1
0
fairly strong and close in frequency, was crucial to
complete the whole set of assignments, extended up
to Js13. Sometimes lines of the less abundant
species were overlapped by more intense absorp-
tions, or even in the limit of the noise level for the
weaker features, which made the corresponding as-
signments dubious, and are missing in the table. We
3
. Results
3
.1. RoÕibronic analysis
Several features were found with the characteris-
7
0
tic pattern for the five germanium isotopomers Ž Ge
7
2
73
74
p
2
0.5%, Ge 27.4%, Ge 7.8%, Ge 36.5% and
include in this table the P Ž1. observations of Karol-
1
7
6
Ge 7.8%., depicted in fig. 1 of Karolczak et al.
czak et al. w12x, which were used in the refinement
process with the same weight as our ICLAS data,
since both experiments have a similar accuracy.
In order to carry out the analysis of the spectrum,
we first had to derive rotational constants for the
ground state of all five isotopomers. Karolczak et al.
w12x, or in fig. 1 of Becerra et al. w9x. A prediction of
the spectrum using the vibrational band origin of this
band, and the rotational constants for the 00 band,
both from Ref. w12x, enabled us to start the first
assignments, based on intensity considerations and
approximate wavenumber positions. The alternating
spin statistical weights of the rotational levels also
0
0
0
74
w12x list their observations of the 0 band for H2 ,
and also the ground state constants they derive for
this species, but they do not provide any information
on the other isotopomers. They also give their esti-
mation of the r0 molecular structure, calculated by
averaging over the three results that can be achieved
Y
a
Y
c
facilitated the assignments: K qK odd levels have
Y
a
Y
c
a spin weight three times larger than K qK even
levels. This band is of c type, and the central part of
p
the spectrum should be dominated by the Q Ž J. and
1

Table 1
y1
1
1
Observed wavenumbers, in cm , for the A˜ B Ž0,1,0.y X˜ A Ž0,0,0. transition of GeH
1
1
2
X
X
a
X
c
Y
Y
Y
76
74
73
72
70
J
K
K
J
Ka Kc
Ge
Ge
Ge
Ge
Ge
103 Žo–c.
103 Žo–c.
103 Žo–c.
103 Žo–c.
103 Žo–c.
3.1
obs.
obs
obs
obs
obs
0
2
3
1
4
5
6
7
8
4
5
3
6
7
2
9
8
1
9
0
0
1
1
2
2
3
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
0
1
1
1
0
1
0
1
0
1
0
0
1
2
3
1
4
5
6
7
8
4
5
3
6
7
2
9
8
1
9
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
1
0
0
0
1
0
1
0
1
0
1
0
2
3
1
4
5
6
7
8
4
5
3
6
7
2
9
8
1
9
17094.693a
17105.306
17105.573
17105.864
17106.792
17108.695
17111.065
17113.724
17116.550
17116.712
17116.712
17117.130
17117.285
17118.422
17118.422
17119.434
y0.7 17094.978a
26.2 17105.573
y1.4 17095.122a
12.9 17105.707
2.5 17106.071
y8.9 17106.381
y10.9 17107.198
y11.6 17109.079
y6.4 17111.435
11.6 17114.082
2.2 17095.280
9.4 17105.864
w75.5x 17106.165
w67.6x 17106.458
1.7 17107.332
y9.0 17109.215
y8.4 17111.578
5.6 17114.211
20.7 17117.030
wy9.2x 17117.183
wy72.3x 17117.183
y44.7 17117.703
y0.1 17117.839
y1.4 17118.941
wy8.1x 17118.941
y12.8 17119.893
w17120.334x 17120.488
w17120.924x 17121.031
48.4 17122.394
y20.7 17122.850
w19.5x 17124.703
w17125.844x 17125.906
7.2 17127.284
w17129.073x 17129.135
8.0 17095.589a
8.9 17106.160
11.7 17106.381
y10.0 17106.792
y18.6 17107.630
y19.0 17109.510
0.3 17111.852
12.3 17114.485
50.0 17117.286
w33.4x 17117.534
y40.3 17117.534
17.3 17118.025
y4.9 17118.117
y10.0 17119.251
w31.9x 17119.251
w24.7x 17120.162
2
3
1
4
5
6
7
8
4
5
3
6
7
2
9
8
1
9
y6.1
b
b
wy16.6x 17105.864
wy78.3x
b
wy26.6x 17106.165
10.9
y20.8
b
y12.8 17107.055
y16.2 17108.949
y14.9 17111.312
0.4 17113.965
wy18.5x
y14.5
3.2
b
28.6 17116.782
w68.4x 17116.895
36.5 17116.888
28.2
w70.7x
11.3
w33.9x 17117.023
y41.8 17117.023
y1.0 17116.895
wy54.0x 17117.398
wy45.0x 17117.534
b
2.1 17117.534
12.4
y26.2 17117.706
y12.1
w28.2x
w3.3x
19.9
2.8
y16.2
wy41.6x
y40.4
w111.6x
w71.5x
y11.9 17118.671
0.8 17118.803
w11.6x 17118.671
w53.8x 17118.803
4.1 17119.748
11.4
28.7
16.1 17119.643
y10.1 17120.221
y28.7 17120.752
33.7 17122.142
y9.8 17122.619
7.4 17124.424
w17125.495x 17125.673
w17126.702x 17126.905
y44.4 17128.916
18.8 17129.808
b
b
17119.958
17120.488
17121.917
y0.3 17120.752
b
14.8 17121.349
y13.0 17122.619
y4.5 17123.087
36.1 17125.030
13.6 17122.296
b
1
1
1
1
1
1
1
10 10
10 10
11 11
11 11
12 12
12 12
13 13
10 17122.394
10 17124.146
11
y3.9 17122.727
39.4 17124.522
y37.9
y46.9 17126.302
w47.8x 17127.502
wy50.9x 17129.390
11
y40.0 17127.076
y8.8
9.3
wy86.5x
12 17128.671
12 17129.568
13 17132.000
23.8 17129.902
7.3
y20.7 17130.105
14.0
w17130.362x
wy111.8x
wy64.3x 17132.294
w17132.461x 17132.489
wy86.8x 17132.779
y3
y1
The rotational assignment is indicated, with the observed minus calculated deviation, in 10 cm units. Heavily blended lines, or absorptions coincident with NH2 lines, were
not included in the fit, and their deviation is listed within brackets. Predicted wavenumbers for very weak transitions, not found in the observations, are also given in brackets.
a
Ref. w12x.
b
Blended lines assigned to more than one transition are given half weight in the fit.

Table 2
y1
1
1
Rotational constants of all five GeH2 isotopomers, in cm , for the A˜ B Ž0,1,0.y X˜ A Ž0,0,0. transition
1
1
7
6
74
73
72
70
Ge
Ge
Ge
Ge
Ge
Y
6.99764 Ž212.a
A
6.99262
6.53331
3.33080
7.00018
6.53331
3.33254
7.00262
6.53331
3.33310
7.00826
6.53331
3.33439
Y
Y
a
B
C
n
6.53330 Ž136.
a
3.33195 Ž50.
17108.2196 Ž140.
18.74150 Ž1667.
4.58821 Ž407.
3.40518 Ž179.
0
y7.250 Ž694.
20
17108.5104 Ž160.
18.65388 Ž2206.
4.58555 Ž493.
3.40762 Ž221.
10.22 Ž325.
17108.6533 Ž122.
18.71439 Ž1558.
4.58604 Ž382.
3.40678 Ž173.
0
y7.728 Ž656.
20
17108.8079 Ž141.
18.64895 Ž2142.
4.59095 Ž495.
3.40489 Ž237.
11.14 Ž440.
17109.1274 Ž93.
18.68115 Ž1156.
4.59124 Ž382.
3.40482 Ž176.
0
y9.580 Ž701.
17
X
A
X
X
B
C
X
y6
D =10
J
X
JK
y3
D
=10
y8.354 Ž782.
21
y10.406 Ž830.
22
y3
rms=10
7
4
Constants for the initial state have been derived from the observations of Karolczak et al. w12x for the Ge isotopomer, and scaled for the other species according to their
moments of inertia. Constants of the excited state have been fitted from the vib-rotational transitions of Table 1. Uncertainties in parentheses are one standard error in units of the
least significant digit. The root mean square Žrms. deviation of each fit is also given.
a
Rotational constant derived from the fit of the ground state energy levels deduced from the observations of Karolczak et al. w12x by constraining the distorsion constants to zero.

4
02
A. Campargue, R. EscribanorChemical Physics Letters 315 (1999) 397–404
by deriving two molecular parameters, r0 and u0,
from three data, the A , B and C constants. To
deviation, of the order of the wavenumber accuracy
of the experimental measurements. The resulting pa-
rameters for all five species are listed in Table 2. The
shift in the vibrational origin is regular and of the
0
0
0
calculate the rotational constants of the other four
7
4
species, we scaled the values of the Ge isotopomer,
using as scaling factors the corresponding ratios of
the moments of inertia, obtained with the r and u
y1
order of 0.30 cm for a change of two units in the
isotopic mass. This is the only clear trend that is
detected in the refined parameters of the different
isotopomers. The band origins are estimated with
0
0
parameters of Ref. w12x. These scaling factors are of
course only approximate, since the inertial defect is
ignored in such calculation, but we believe that this
approximation results in smaller errors than using the
same ground state constants for all five species.
y1
good accuracy Ž;0.014 cm ., whereas DJ is the
least well determined parameter. As expected, there
is a very large variation in the A rotational constant
with respect to the ground electronic state value,
corresponding to the large increase in the equilib-
Besides, it is interesting to remark that the r struc-
0
ture of Karolczak et al. w12x is very similar to the
equilibrium structure predicted in the theoretical
rium angle that takes place upon excitation of the
1
works of Bunker, Phillips and Buenker w17x Ž1.591 A˚
B state. This change in the geometrical structure
1
and 91.48., Balasubramanian w18x Ž1.587 A˚ and 91.58.
and Mineva et al. w19x Ž1.598 A˚ and 91.88, from their
local spin density calculation.. It should be noted
that a substitution of one germanium isotope by
another does not induce a change in the B rotational
also induces a reduction of the B constant, whereas
the moment of inertia with respect to the axis per-
pendicular to the plane, and consequently the C
constant, is little affected. The A constant is also
y1
increased, by ;2.25 cm , with respect to the value
constant within this approximation, and thus all five
estimated by Karolczak et al. w12x for the ground
Y
1
B values are equal. The rotational constants derived
vibrational state of this excited electronic state B .
0
1
X
in this way are given in Table 2.
The variation of B with the vibrational excitation of
The analysis then proceeded in the standard way.
We chose Watson’s A-reduction, in the I representa-
tion. We tried to release only a limited number of
parameters, namely the rotational constants and the
vibrational band origin, but we found it necessary to
also release at least one centrifugal distortion con-
the bending coordinate is also positive, but small,
;0.06 cm X , which is similar but of opposite sign
to that of C .
r
y1
The DJK parameter takes a fairly large value, of
y2
y1
the order of 10
cm . This may be due to the
presence of some hidden perturbation among the
data, even for the low Ka levels involved in the
observed transitions. We found a similar behaviour
stant, DJK , and in some cases D as well, in order to
J
reproduce the data with a satisfactory mean square
Table 3
B
Vibrational dependence of the rotational constants Ža parameters., equilibrium values of the rotational constants derived from them Žall in
i
y1
1
74
cm ., and equilibrium geometrical structure of the B1 electronic state of H2
a1B
a2B
a3B
Eq. values
re
ue
Ae
Be
Ce
0.3061
0.0773
0.0618
y2.2389
y0.0674
0.0525
0.4775
0.0610
0.0503
15.6874
4.5537
3.5424
Ž A , B .
1.5441
1.5412
1.5413
122.73
123.13
122.61
e
e
ŽB , C .
e
e
ŽC , A .
e
e
Average
1.5422 Ž13.
122.82 Ž20.
a
ab initio , SCF
1.544
1.566
1.553
123.4
123.2
122.1
a
ab initio , CI
b
ab initio
The a 2B parameters are experimental values Žfrom this work and Ref. w12x., and the a and a 3B parameters are taken from the calculation
B
1
of Allen and Schaeffer w20x Žtable 20. for SiH . The r ŽGe–H. and u ŽH–Ge–H. values have been derived from the three pairs of rotational
2
e
e
constants and are given in A˚ and degrees, respectively. Comparison with previous ab initio works is also shown.
a
Ref. w21x.
Ref. w18x.
b

A. Campargue, R. EscribanorChemical Physics Letters 315 (1999) 397–404
403
in the spectrum of SiH w5x, where the 31,3, 41,4 and
et al. w21x or Balasubramanian w18x. Interestingly, the
SCF calculation of Barthelat et al. better predicts the
bond distance, whereas the CI calculation gives a
better result for the bond angle.
2
51,5
levels of the excited state were shifted by a local
y1
perturbation by as much as 0.15 cm . The perturba-
tion, if present, is smaller in the germylene radical.
The centrifugal distortion parameter becomes in this
case an effectiÕe constant, taking an abnormally
large value in order to accommodate the perturba-
tion. The predictive character of the refined parame-
ters for higher values of the rotational quantum
The uncertainty in the average values intrinsic to
the averaging process is more than five times smaller
now that in the calculation of Karolczak et al. w12x,
based on a r0 structure. However, the uncertainty
induced by the use of SiH values instead of the
2
7
4
B
B
numbers, especially K , may be limited.
unavailable GeH2 results for the a1 and a3
a
The lower part of Fig. 1 shows the predicted
spectra in this region, with the weight of the spec-
trum of each isotopomer scaled by its corresponding
natural abundance.
parameters, is difficult to ascertain. In any case, we
estimate a reduction of around 0.04 A in the bond
˚
distance and an increase of around 328 in the equilib-
1
rium angle, with respect to the structure of the A
1
electronic state. The corresponding increments for
˚
3
.2. Equilibrium geometry
the SiH molecule were ;0.03 A and 30.58, respec-
2
tively.
With the values of the rotational constants of the
1
n s1 level of the B electronic state obtained in
2
1
this work, and those for the n s0 level reported by
4. Conclusion
2
Karolczak et al. w12x, it is possible to calculate the
vibrational dependence of the rotational constants
We have observed and analyzed new spectral data
of the germylene radical in the visible region, which
augments the scarce observations previously avail-
able for this species. The analysis of the data allows
the determination of the rotational constants of all
B
with the bending mode, given by the a2 parameters
Žwe use here B as superscript to designate either of
the rotational constants, A, B, or C.. An estimation
of the equilibrium structure of this electronic state
B
requires the knowledge of the values of the a and
five germanium isotopomers in the n s1 level of
1
2
B
1
a3 parameters. As far as we are aware, there has
the B electronic state. From these constants and
1
been no previous experimental or theoretical deter-
mination of these parameters. One possible approxi-
mate approach to this problem is to use the values of
the SiH2 molecule for the n1 and n3 vibrations
calculated by Allen and Schaeffer w20x, based on the
those of the n s0 level determined previously, we
2
have carried out an improved estimation of the geo-
metrical structure of the excited electronic state. The
results confirm the ab initio predictions for this
structure.
similarity of the spectroscopic properties of these
molecules. We gather in Table 3 the values of the
Some of the lines observed here had been de-
tected and used to monitor GeH2 reactions before.
We now provide an assignment for these transitions,
and a set of new observations which can be used for
the same purpose in the future.
B
ai parameters and the equilibrium values of the
rotational constants obtained therefrom. From Ae,
B and C , we have derived the equilibrium geomet-
e
e
1
rical structure of the B electronic state, taking the
1
parameters two by two in order to determine r and
e
ue. The results of these calculations are also listed in
Table 3, with the mean average of all three values,
and previous ab initio results w18,21x. It can be seen
that: first, the geometrical parameters take very simi-
lar values in all three estimations, which indicates a
small inertial defect error propagated in the analysis;
and second, the average parameters are very close to
the ab initio equilibrium structure of either Barthelat
Acknowledgements
We wish to thank R. Saint-Loup of the University
of Burgundy for providing us the germane sample.
The continuous interest of N. Sadeghi ŽLSP, Greno-
ble. is also acknowledged. This research is supported
by the Integrated Action HF1998-0070 between the
French and Spanish Ministries of Education. R.E.’s

4
04
A. Campargue, R. EscribanorChemical Physics Letters 315 (1999) 397–404
w8x K. Saito, K. Obi, Chem. Phys. Lett. 215 Ž1993. 193.
w9x R. Becerra, S.E. Boganov, M.P. Egorov, O.M. Nefedov, R.
Walsh, Chem. Phys. Lett. 260 Ž1996. 433.
w10x K. Saito, K. Obi, Chem. Phys. 187 Ž1994. 381.
w11x K. Obi, M. Fukushima, K. Saito, Appl. Surf. Sci. 79r80
Ž1994. 465.
research has been carried out with financial support
from DGES of the Spanish Ministry of Education,
Project PB96-0881.
w12x J. Karolczak, W.W. Harper, R.S. Grev, D.J. Clouthier, J.
Chem. Phys. 103 Ž1995. 2839.
w13x U.N. Alexander, N.A. Trout, K.D. King, W.D. Lawrance,
Chem. Phys. Lett. 299 Ž1999. 291.
References
w1x T. Motooka, J.E. Greene, J. Appl. Phys. 59 Ž1986. 2015.
w2x W. Du, L.A. Keeling, C.M. Greenlief, J. Vac. Sci. Technol.
A 12 Ž1994. 2281.
w14x M. Herman, J. Lievin, J. Vander Auwera, A. Campargue,
´
Adv. Chem. Phys. 108 Ž1999. 1.
w15x S. Gerstenkorn, P. Luc, Atlas du spectre d’absorption de la
w3x J.E. Crowell, G. Lu, J. Electron Spectrosc. Related Phenom.
molecule d’iode, Editions du CNRS, Paris, 1978.
´
5
4 Ž1990. 1045.
w16x S.C. Ross, F.W. Birss, M. Vervloet, D.A. Ramsay, J. Mol.
w4x A. Lloret, M. Oria, B. S e´ oudi, L. Abouaf-Marguin, Chem.
Spectrosc. 129 Ž1988. 436.
Phys. Lett. 215 Ž1993. 193.
w17x P.R. Bunker, R.A. Phillips, R.J. Buenker, Chem. Phys. Lett.
w5x R. Escribano, A. Campargue, J. Chem. Phys. 108 Ž1998.
110 Ž1984. 351.
6
249.
w18x K. Balasubramanian, J. Chem. Phys. 89 Ž1986. 5731.
w19x T. Mineva, N. Russo, E. Sicilia, M. Toscano, Int. J. Quant.
Chem. 56 Ž1995. 669.
w6x A. Campargue, R. Romanini, N. Sadeghi, J. Phys. D: Appl.
Phys. 31 Ž1998. 1168.
w7x H. Kawasaki, J. Kida, K. Sakamoto, T. Fukuzawa, M. Shi-
ratani, Y. Wanatabe, Japanese J. of Appl. Phys. 37 Ž1998. Pt.
w20x W.D. Allen, H.F. Schaeffer III, Chem. Phys. 108 Ž1986. 243.
w21x J.-C. Barthelat, B.S. Roch, G. Trinquier, J. Satg e´ , J. Am.
Chem. Soc. 102 Ž1980. 4080.
2
, no. 4B.