Observation of several new electronic transitions of the SrOH free radical

Article abstract of DOI:10.1063/1.479065

We report four new electronic transitions of the SrOH free radical, the C ²Π-X ²Σ⁺, D ²Σ⁺-X ²Σ⁺, E ²Σ⁺-X ²Σ⁺, and F ²Π-X ²Σ⁺ transitions. SrOH was prepared in a supersonic jet by laser ablation and spectra were recorded using laser-induced fluorescence. The C ²Π-X ²Σ⁺ excitation spectrum shows complex vibronic structure which is attributed, at least in part, to Renner-Teller activity in the excited electronic state. This is supported by dispersed fluorescence spectra which show substantial bending mode activity in the emission from several different excited vibronic levels. It is suggested that the prominence of nominally forbidden vibrational features arises from a large change in permanent electric dipole moment between the X and C states. In turn, this suggests that the C ²Π state of SrOH is the analogue of the reverse-polarized ² Π states known for the alkaline-earth monohalides, i.e., the highest occupied π orbital points towards the O atom. The D ²Σ⁺-X ²Σ⁺, E ²Σ⁺-X ²Σ⁺, and F ²Π-X ²Σ⁺ spectra are much simpler than the C-X system, being dominated by regular structure in the Sr-O stretching vibration.

Full text of DOI:10.1063/1.479065

Observation of several new electronic transitions of the SrOH free radical
Matthew S. Beardah and Andrew M. Ellis
Citation: The Journal of Chemical Physics 110, 11244 (1999); doi: 10.1063/1.479065
View online: http://dx.doi.org/10.1063/1.479065
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/110/23?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
An experimental and ab initio study of the electronic spectrum of the jet-cooled F2BO free radical
J. Chem. Phys. 140, 164302 (2014); 10.1063/1.4871010
The electronic spectrum of the fluoroborane free radical. II. Analysis of laser-induced fluorescence and single
vibronic level emission spectra
J. Chem. Phys. 130, 164310 (2009); 10.1063/1.3122031
The electronic spectrum of the fluoroborane free radical. I. Theoretical calculation of the vibronic energy levels of
the ground and first excited electronic states
J. Chem. Phys. 130, 164309 (2009); 10.1063/1.3122008
Observation of a new phosphorus-containing reactive intermediate: Electronic spectroscopy and excited-state
dynamics of the HPBr free radical
J. Chem. Phys. 123, 144304 (2005); 10.1063/1.2055227
The laser-induced fluorescence spectrum, Renner–Teller effect, and molecular quantum beats in the à 2 Π i -X̃ 2
Π i transition of the jet-cooled HCCSe free radical
J. Chem. Phys. 121, 5801 (2004); 10.1063/1.1786924
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:
28.248.155.225 On: Sun, 23 Nov 2014 13:11:11
1

JOURNAL OF CHEMICAL PHYSICS
VOLUME 110, NUMBER 23
15 JUNE 1999
Observation of several new electronic transitions of the SrOH free radical
Matthew S. Beardah and Andrew M. Ellisa)
Department of Chemistry, University of Leicester, University Road, Leicester, United Kingdom
͑
Received 20 January 1999; accepted 23 March 1999͒
We report four new electronic transitions of the SrOH free radical, the C 2
˜
͟
–X ͚ ,
˜
2
ϩ
D ͚ –X ͚ , E ͚ –X ͚ , and F 2͟ –X ͚ transitions. SrOH was prepared in a
˜
2
ϩ
˜
2
ϩ
˜
2
ϩ
˜
2
ϩ
˜
˜
2
ϩ
supersonic jet by laser ablation and spectra were recorded using laser-induced fluorescence. The
˜
C
2͟ –X ͚ excitation spectrum shows complex vibronic structure which is attributed, at least in
˜
2
ϩ
part, to Renner–Teller activity in the excited electronic state. This is supported by dispersed
fluorescence spectra which show substantial bending mode activity in the emission from several
different excited vibronic levels. It is suggested that the prominence of nominally forbidden
˜
vibrational features arises from a large change in permanent electric dipole moment between the X
and C states. In turn, this suggests that the C 2͟ state of SrOH is the analogue of the
˜
˜
‘
‘reverse-polarized’’2
orbital points towards the O atom. The D ͚ –X ͚ , E ͚ –X ͚ , and F 2͟ –X ͚
͟ states known for the alkaline-earth monohalides, i.e., the highest occupied
˜
2
ϩ
˜
2
ϩ
˜
2
ϩ
˜
2
ϩ
˜
˜
2
ϩ
␲
˜
˜
spectra are much simpler than the C–X system, being dominated by regular structure in the Sr–O
stretching vibration. © 1999 American Institute of Physics. ͓S0021-9606͑99͒01023-5͔
I. INTRODUCTION
˜
2
ϩ
The four highest known states of CaOH, the D ͚ ,
˜
2
ϩ
˜
˜
Polyatomic alkaline-earth metal-containing free radicals
have received considerable attention from spectroscopists in
recent years. These molecules are particularly attractive be-
cause they are relatively easily observed using laser-induced
fluorescence ͑LIF͒ spectroscopy. A large number of these
ported. The D, E, and F states were observed by Pereira and
Levy using both LIF and mass-resolved resonantly enhanced
multiphonon ionization spectroscopies.18 Vibrational struc-
ture was resolved for each of these states together with some
radicals have been investigated, ranging from triatomics,
˜ ˜
partially resolved rotational structure in the D–X system.
such as metal monohydroxides1
–20
and monoisocya-
through to larger species, such as metal
˜
˜
The symmetries of the D and E states were readily estab-
2
1–23
nides,
2
ϩ
˜
24–28
lished as ͚ . However, the F state, whose symmetry was
monocyclopentadienyls.
However, it is alkaline-earth
not determined, proved to be more interesting, showing evi-
monohydroxides that have been the most widely studied,
mainly because they are relatively simple to prepare and usu-
ally yield spectra which are reasonably straightforward to
analyze.
CaOH has been the most extensively investigated. To
date, eight electronic states of CaOH have been identified.
dence for a bent equilibrium geometry. The G 2
˜
͟
state of
CaOH was reported by Hailey, Jarman, and Bernath from an
˜ ˜
optical–optical double resonance experiment using the B–X
20
transition as the first resonant step. As in the four lowest
states of CaOH, the unpaired electron seems to be largely
The four lowest states, X ͚ , A 2͟, B ͚ , and C ⌬,
˜
2
ϩ
˜
˜
2
ϩ
˜
2
˜ ˜
˜
metal localized in the D, E, and G states ͑the situation is less
have been characterized in several LIF studies2
,4,11
and the
˜
clear for the F state͒. Given that the ligand is largely a spec-
ground state has also been studied in detail using millimeter
tator in these electronic transitions, with the possible excep-
wave spectroscopy.29 The A 2
˜
͟ and C ⌬ states are espe-
˜
2
˜
tion of the F state, this suggests that comparable new metal-
cially interesting because of the possibility of Renner–Teller
activity in these states. In fact detailed studies have shown
that the Renner–Teller effect in these states is small.1
Given that the unpaired electron, which is the source of the
net orbital angular momentum, is thought to be in metal-
localized orbitals polarized away from the OH group in the
localized transitions might be observed for many other
alkaline-earth containing free radicals.
Following this line of reasoning, we have recently re-
˜
1,12,16
ported two new electronic states of BaOH, the C 2
͟ and
D ͚ states.30 In the current work our investigations have
been extended to the SrOH radical. Four new electronic sys-
tems have been identified in the ultraviolet region. Three of
˜
2
ϩ
˜
A
˜
2͟ and C ⌬ states, this finding is not surprising. After
2
all, the remoteness of the unpaired electron from the Ca–
O–H bending motion, which is dominated by movement of
the H atom, does not lend itself to efficient coupling of or-
bital and vibrational angular momenta.
these, the D ͚ –X ͚ , E ͚ –X ͚ , and F 2͟ –X
͚
˜
2
ϩ
˜
2
ϩ
˜
2
ϩ
˜
2
ϩ
˜
˜
2
ϩ
systems, show simple vibrational structure. However,
˜
the C 2
˜
2
ϩ
͟
–X ͚ transition gives extensive and complex vi-
bronic structure and yields evidence of significant Renner–
Teller activity. In this article we present excitation and dis-
a͒Electronic mail: andrew.ellis@le.ac.uk
0
021-9606/99/110(23)/11244/11/$15.00
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:
28.248.155.225 On: Sun, 23 Nov 2014 13:11:11
11244
© 1999 American Institute of Physics
1

J. Chem. Phys., Vol. 110, No. 23, 15 June 1999
M. S. Beardah and A. M. Ellis
11245
persed fluorescence spectra of the four new electronic
systems of SrOH. Furthermore, we discuss the character of
tiple shot averaging was achieved by home-written software
routines.
the C 2͟ state and the origin of its Renner–Teller effect.
˜
III. RESULTS AND DISCUSSION
The 25 500–34 000 cmϪ1 region was scanned in a
search for new SrOH bands. Many intense peaks were ob-
served and at least four different electronic transitions were
identified within this region. A variety of evidence supports
the assignment of these transitions to SrOH, but the simplest
means of firmly establishing the carrier was to record dis-
persed fluorescence spectra. When applied to each prominent
vibronic transition the dispersed fluorescence spectra, several
of which are presented later, showed vibrational structure
characteristic of the ground state of SrOH.
II. EXPERIMENT
SrOH was prepared in the gas phase by laser ablation of
strontium metal. An intense and stable laser-induced fluores-
cence signal for SrOH was obtained by seeding hydrogen
peroxide, or water, into a helium carrier gas. The LIF signal
could also be obtained with a comparable intensity with no
precursor present in the carrier gas, though in this case the
shot-to-shot stability of the signal was much less reproduc-
ible. In view of these observations, it seems likely that the
bulk of the SrOH comes from the laser ablation of strontium
hydroxide on the surface of the strontium target, rather than
from a gas phase reaction between strontium atoms and
added reagent.
˜
2
˜
2
؉
A. The C ͟–X ͚ system
1. Overview
An LIF excitation spectrum covering 27 200–28 000
cmϪ1 is shown in Fig. 1. A large number of bands appears in
Laser ablation occurred in an aluminum fixture attached
to the faceplate of a pulsed valve ͑General Valve series 9͒.
This fixture contained a 2-mm-diam channel along its central
axis to allow passage of the seeded carrier gas mixture. A
stagnation pressure of 12 bar was typically used behind the
pulsed valve. Strontium metal was hammered into a brass
sample holder and filed to prepare a flat, even surface for
ablation. The brass holder was inserted into the fixture,
around 15 mm from the faceplate of the pulsed valve, so that
the strontium surface was protruding slightly into the central
channel. A 10 mm extension piece was attached to the end of
the fixture, thereby increasing the length of the pre-
expansion, postablation mixing zone to approximately 20
mm. In a typical experiment 20 mJ pulses from an ArF ex-
cimer laser ͑Lambda Physik, model Compex 100͒ were used
to ablate the strontium. The excimer laser beam was softly
focused onto the metal surface, and firing was timed to co-
incide with the arrival of a gas pulse from the pulsed valve.
The resulting gas mixture then expands into a Roots-pumped
the spectrum, with several notable strong and medium inten-
sity bands appearing alongside many weaker ones. As dis-
cussed later, most of these bands do not seem to fit into any
obvious pattern. The wave numbers of strong and medium
intensity bands are collected in Table I.
To make some sense of this and other SrOH spectra, a
useful starting point is to compare the relative positions of
electronic transitions of SrOH with those of its isoelectronic
analogue, SrF. The known electronic transitions of SrF are
predominantly metal localized. Consequently, one would ex-
pect the electronic spectra of SrF and SrOH to have some
similarities. The electronic spectroscopy of SrF is well
31
documented and the relative positions and symmetries of
its electronic states are illustrated in Fig. 2. Prior to the
present work, the only reported excited states of SrOH were
the A 2
˜
˜
͟
and B ͚ states.1,3,15 Notice that the A 2
˜
2
ϩ
˜
͟
and
B ͚ states of SrOH and the A 2
2
ϩ
͟ and B ͚ states of SrF
2
ϩ
have, as expected, similar energies. There are other similari-
ties. For example, the spin-orbit coupling constants in the
Ϫ4
vacuum chamber ͑base pressure of ϳ10 mbar͒ to form a
supersonic jet.
first excited states of both molecules are comparable, being
Ϫ1
2
81 and 263 cm for SrF and SrOH, respectively. Also, the
The expanding jet was crossed by the beam from a
pulsed dye laser ͑Lambda Physik Scanmate 2E, pumped by a
Continuum Surelite Nd:YAG laser͒ about 25 mm down-
stream of the fixture. The optimum SrOH LIF signal stability
was achieved by synchronizing the dye laser pulse with the
trailing edge of the gas pulse. In the current work the dye
laser was operated without an intracavity etalon, resulting in
Sr–O and Sr–F stretching vibrations have similar frequen-
cies in the ground and first two excited electronic states. All
of these observations are consistent with the nonbonding
character of the unpaired electron.
If the similarity between SrF and SrOH also holds for
˜
˜
higher electronic states, the C–X system of SrOH should be
Ϫ1
2
2
ϩ
a linewidth of ϳ0.2 cm . Laser excitation spectra were ob-
a ͟ – ͚ transition with an origin in the region of 27 400
cm . Furthermore, comparison with SrF would lead us to
expect a spin-orbit coupling constant of ϳ60 cm in the
Ϫ1
tained by directing the fluorescence onto a photomultiplier
tube ͑PMT͒ via a f/1.3 collection lens. For all excitation ex-
periments a 500 nm long-pass filter was placed between the
lens and PMT to reduce scattered light interference from
both the dye and excimer lasers. Dispersed fluorescence
spectra were obtained by passing the unfiltered fluorescence
into a 0.27 m monochromator ͑Acton Research͒ equipped
with a PMT detector.
Ϫ1
˜
C
2͟ state.
2. Electronic origin
The strongest bands in Fig. 1 appear at 27 291 and
Ϫ1
2
7 315 cm and have nearly equal intensities. Given their
˜
˜
proximity to the predicted origin of the C–X system, it is
possible that these form spin-orbit pair in the
˜
2͟ –X ͚ manifold of SrOH. If correct, this would im-
C
All PMT signals were amplified ͑SRS 545 preamplifier͒,
digitized by a CAMAC-based transient recorder ͑LeCroy
a
˜
2
ϩ
2262͒, and transferred to a PC. Gated integration and mul-
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:
128.248.155.225 On: Sun, 23 Nov 2014 13:11:11

11246
J. Chem. Phys., Vol. 110, No. 23, 15 June 1999
M. S. Beardah and A. M. Ellis
FIG. 1. Laser excitation spectrum
showing the C 2
˜
͟
ϩ
–X 2
˜
͚
ϩ
system and
–X 2
˜
͚ ͚ system of
part of the D 2
˜
ϩ
SrOH. This spectrum is a composite of
scans using two different laser dyes;
band intensities have not been cor-
rected for variations in dye laser out-
put. Symbols above certain peaks refer
ˆ
to the following: labels those transi-
tions for which dispersed fluorescence
spectra are given in Fig. 5;* identifies
˜
˜
0
0
a peak thought to be the D–X 0 tran-
sition.
ply a spin-orbit coupling constant of 24 cmϪ1 in the C 2͟
˜
region. Consequently, to try and obtain firmer assignments of
the 27 291/27 315 cm pair, we have ͑i͒ recorded dispersed
fluorescence spectra and ͑ii͒ investigated the rotational struc-
ture.
Ϫ1
Ϫ1
state, a value considerably smaller than the ϳ60 cm sug-
gested above in the analogy with SrF. One way of checking
this assignment is to search for other spin-orbit doublets
showing the same characteristic splitting. Scans to lower
wave numbers revealed no such pairings. On the high wave
number side there are several possible candidates, each dis-
Given the high intensities of the bands at 27 291 and
Ϫ1
2
7 315 cm , the most likely assignment is that they form
the origin of the C 2
͟ –X ͚ system. The dispersed fluo-
˜
˜
2
ϩ
Ϫ1
playing a splitting of ϳ30 cm . However, some of these
doublets may result from the accidental proximity of very
Ϫ1
different bands, especially in the dense 27 700–27 800 cm
TABLE I. Major vibronic bands in the 27 290–28 500 cmϪ1 region.
FIG. 2. Comparison of the energies of electronic states of SrF and SrOH.
The energies of SrF states were taken from Huber and Herzberg ͑Ref. 31͒
while those of the A 2
˜
͟
and B 2
˜
͚
ϩ
states of SrOH are from Refs. 2 and 4.
The energies of the four highest states of SrOH shown in this diagram have
been determined in the present work.
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:
28.248.155.225 On: Sun, 23 Nov 2014 13:11:11
1

J. Chem. Phys., Vol. 110, No. 23, 15 June 1999
M. S. Beardah and A. M. Ellis
11247
FIG. 3. Dispersed fluorescence spectra obtained by
˜
C
2͟1/2–X 2͚ 0 and ͑b͒
˜
˜
ϩ
0
pumping the ͑a͒
C
0
2͟3/2
–X 2͚ 0 transitions. The small shift in the rela-
˜
ϩ
0
0
tive positions of the bands in the two spectra, indicated
by dashed lines, is due to the excited state spin-orbit
Ϫ1
splitting ͑ϳ24 cm ͒.
rescence data are consistent with this assignment. Apart from
a shift in band positions due to excited state spin-orbit split-
ting, virtually identical dispersed fluorescence spectra are ex-
pected on excitation of these two transitions. As shown in
Fig. 3, this is precisely what is observed. The dominant band
is due to emission at the pump wavelength, as expected for
excitation of the origin of a transition involving movement of
a primarily nonbonding electron. However, there are also
structure in the experimental spectrum gives excited state
Ϫ1
rotational constants of BЈϭ0.2564Ϯ0.0009 cm and AЈ
Ϫ1
ϭ24.52Ϯ0.04 cm
.
3. Higher vibronic bands
Although the C 2͟1/2–X ͚ and C 2͟3/2–X ͚
˜
˜
2
ϩ
˜
˜
2
ϩ
0
0
0
transitions are relatively easy to assign, the remaining bands
Ϫ1
other much weaker features, including peaks at 368Ϯ7 cm
Ϫ1
and 724Ϯ7 cm to the red of the respective pump positions.
3
Ϫ1
Brazier and Bernath reported values of 361 and 727 cm
,
respectively, for the fundamental and first overtone of the
˜
2
ϩ
Sr–O–H bending mode (␯ ) in the X ͚ ground electronic
2
state of SrOH. These assignments were subsequently con-
firmed and refined in a much higher resolution dispersed
17
fluorescence study by Presunka and Coxon. We therefore
Ϫ1
Ϫ1
assign the 368Ϯ7 cm and 724Ϯ7 cm bands in Fig. 3 to
these bending mode transitions. We also observe very weak
Ϫ1
Ϫ1
peaks at 520Ϯ10 cm and 1046Ϯ7 cm in the spectrum.
Ϫ1
These correlate well with the values of 528 and 1048 cm
obtained by Brazier and Bernath for ␯ and 2␯ , where ␯ is
3
3
3
the Sr–O stretch, in the electronic ground state. The peaks at
Ϫ1
8
91 and 1230 cm are assigned to the ␯ ϩ␯3 and 2␯2
2
ϩ␯ combination bands.
3
Expanded views of the 27 291 and 27 315 cmϪ1 bands
are shown in Fig. 4, together with simulations of the rota-
Ϫ1
tional structure. The dye laser linewidth of 0.2 cm pre-
vented full rotational resolution, but some structure is re-
solved in parts of the bands and is readily assigned. The
rotational structure is entirely consistent with the assignment
2
2
ϩ
2
to a
͟ – ͚ transition in which the ͟ state undergoes
32
Hund’s case ͑a͒ coupling. At the present resolution four
branches should be observed for each band and all four can
be seen in Fig. 4 ͑although only the two outer branches are
specifically identified͒. Using the known rotational constant
of the X ͚ state of SrOH (BЉϭ0.249 20 cmϪ1͒,3 the ob-
served rotational structure can be simulated by varying the
rotational and spin-orbit coupling constants in the excited
state together with the ground state rotational temperature. A
least squares fit of the simulated spectrum to the assigned
FIG. 4. ͑a͒ Rotational contours of the bands centered at 27 291 and 27 315
cmϪ1. These bands are assigned to the C 2͟1/2–X 2
˜
˜
ϩ
˜
and C 2͟3/2–
͚
˜
2
ϩ
˜
2
ϩ
0
X
͚ 00 transitions, respectively. The coarse 3B, B, ϪB, Ϫ3B branches
2
2
ϩ
expected for a
͓͟case ͑a͒– ͚ transition ͑Ref. 31͒ are easily identified,
although individual rotational components are only resolved for the 3B and
2
2
ϩ
Ϫ3B branches. ͑b͒ Simulation assuming a
͟ – ͚ transition, a rotational
Ϫ1
Ϫ1
temperature of 15 K, BЉϭ0.2492 cm , BЈϭ0.2564 cm and AЈϭ24.52
Ϫ1
cm
.
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:
28.248.155.225 On: Sun, 23 Nov 2014 13:11:11
1

11248
J. Chem. Phys., Vol. 110, No. 23, 15 June 1999
M. S. Beardah and A. M. Ellis
in Fig. 1 are more challenging. The excitation spectrum is
remarkably complex for a transition which is thought to be
dominated by promotion of a largely nonbonding electron. In
fact it will be seen later that this simple picture of the elec-
tronic transition needs to be modified. To aid the assignment
process we have followed the same approach as employed
for the origin, i.e., have made use of information from rota-
tional contours and from dispersed fluorescence spectra. Un-
fortunately, as will be seen, in most cases the latter do not
give clear-cut information about vibrational assignments.
The rotational structure has proved more valuable, allowing
us to identify either parallel or perpendicular character in
most bands and therefore providing important information on
upper and lower state vibronic symmetries.
The first band of substantial intensity on the high wave
number side of the C 2͟ –X ͚ origin is centered at
2
˜
˜
2
ϩ
Ϫ1
7 422 cm . Even at low resolution ͑see Fig. 1͒ this band
appears broader than others in the spectrum and clearly
arises from more than one transition. In fact an expanded
view reveals that there are three contributing bands. Conges-
tion makes it difficult to establish the vibronic symmetries in
Ϫ1
all three but the strongest ͑at 27 429 cm ͒ has the appear-
2
2
ϩ
ance of a
sponding dispersed fluorescence spectrum, shown in Fig.
͑a͒, is remarkably complicated. Extensive vibrational struc-
͓͟case ͑a͔͒– ͚ vibronic transition. The corre-
5
ture attributable to both the Sr–O stretching and the Sr–O–H
bending vibrations is seen. Notice also that there is some
emission to the blue of the laser wavelength, indicating at
least a partial contribution from a hot/sequence transition.
The next two bands, centered at 27 522 and 27 588
cmϪ1, can be identified as parallel bands from their rotational
FIG. 5. Dispersed fluorescence spectra obtained by laser excitation of ͑a͒
Ϫ1
Ϫ1
Ϫ1
27 422 cm , ͑b͒ 27 522 cm , and ͑c͒ 27 588 cm transitions. The arrows
indicate the laser wave number. Modes ␯2 and ␯3 are the Sr–O–H bend and
the Sr–O stretch, respectively. The askerisks in ͑a͒ indicate bands that are
presumably contributions from excitation of an underlying hot or sequence
band transition.
2
2
structure ͑see Fig. 6͒ and are assigned to ͚– ͚ transitions
on account of their missing Q branches. The corresponding
dispersed fluorescence spectra are shown in Figs 5͑b͒ and
Ϫ1
Ϫ1
5
͑c͒. Like the 27 422 cm peak, the 27 522 cm peak also
the Sr–O stretching and Sr–O–H bending vibrational
modes.
shows some emission to the blue of the pump wavelength,
indicating that it is a hot band. However, the dispersed fluo-
rescence spectrum is otherwise completely different, the
single dominant feature to the red of the pump position aris-
ing from single quantum excitation of the Sr–O–H bending
mode. The dispersed fluorescence spectrum obtained by ex-
˜
˜
4
. Bands to the red of the C–X origin
Ϫ1
citing the 27 588 cm transition is different again, showing
Three bands with significant intensities have been seen
activity in both the Sr–O stretch and the bending mode. Un-
˜
˜
Ϫ1
on the low wave number side of the C–X origin. These
like the 27 522 cm transition, excitation of the 27 588
Ϫ1
bands, which are not shown in Fig. 1, were found at 26 931,
cm transition does not yield any emission to the blue of the
Ϫ1
2
6 569, and 26 395 cm . The observation of emission to the
laser wavelength.
Ϫ1
blue of the laser wavelength in the dispersed fluorescence
The 27 695–27 900 cm region contains at least ten
Ϫ1
spectrum has been used to show that the 26 931 cm feature
bands with substantial intensities. The conclusions from the
rotational contours are summarized in Table I. The first three
bands in this group are parallel bands and are attributed to
is a hot band. No blueshifted emission could be convincingly
observed in the dispersed fluorescence of the other two tran-
sitions, so their assignment as hot bands is not confirmed.
However, it should be noted that all three bands are, within
experimental error, at the positions expected for the
2
2
͚
– ͚ transitions. However, most of the remaining bands
2
2
ϩ
are consistent with
͓͟case ͑a͔͒– ͚ transitions. The infor-
mation from dispersed fluorescence spectra is less specific.
Nevertheless, there is one general and important conclusion
which emerges from the dispersed fluorescence spectra for
2͟1/2–X ͚ 2 , 2 , and 2 3 transitions. Thus, it seems
˜
2
ϩ
0
0
0 0
˜
C
1 2 1 1
˜
˜
likely that all three are hot bands in the C–X system. Notice,
Ϫ1
however, that the spin-orbit partners, the corresponding tran-
all transitions in the 27 700–27 920 cm region. Although
sitions in the C 2͟3/2–X ͚ manifold, were not observed.
˜
˜
2
ϩ
the details differ from one excited state to another, each dis-
persed fluorescence spectrum shows some activity in both
We are unable to explain this unusual result at present.
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:
128.248.155.225 On: Sun, 23 Nov 2014 13:11:11

J. Chem. Phys., Vol. 110, No. 23, 15 June 1999
M. S. Beardah and A. M. Ellis
11249
sequence that can be seen in low resolution spectra is that
some transitions which are nominally forbidden in the
Franck–Condon limit may gain intensity through Renner–
Teller coupling.33 For SrOH this means that odd quantum
number changes of the bending vibration become partially
allowed. These transitions are very weak in the
˜
A
2͟ –X ͚ system of SrOH but, as detailed in the final
˜
2
ϩ
3
section of this paper, there are good reasons to expect these
bands to be much stronger in the C 2
͟ –X ͚ system.
˜
˜
2
ϩ
Assuming substantial Renner–Teller activity in the
˜
C
2͟ state, we can begin to tackle band assignments. How-
ever, we must emphasize at the outset that the assignments
made below are provisional and represent our best explana-
˜
˜
tion for some of the bands in the C–X spectrum. A complete
explanation for all of the structure seen in the spectrum cur-
rently eludes us.
In terms of vibrational modes we can restrict our atten-
tion to the Sr–O stretch (␯ ) and the Sr–O–H bend (␯ )
3
2
since the O–H stretching frequency will be too high to be
responsible for any structure in Fig. 1. Since our earlier
analysis shows that the rotational constant increases on exci-
tation from the X ͚ state to the C 2
˜
2
ϩ
˜
͟
state, it is reason-
able to assume that the Sr–O stretching frequency also in-
Ϫ1
creases. Given that the ground state value is 528 cm , an
Ϫ1
excited state value in the range 550–600 cm is likely in
view of the modest increase in rotational constant. Conse-
quently, using these arguments, we would expect a pair of
FIG. 6. Rotational contours of the bands centered at ͑a͒ 27 522 and ͑b͒
Ϫ1
2
͚
7 588 cm . The lower spectrum in each case is a simulation assuming a
2
ϩ
2
ϩ
Ϫ1
Ϫ1
– ͚ transition (BЉϭ0.249 cm , BЈϭ0.258 cm , and Tϭ15 K͒.
bands arising from excitation of the two spin-orbit compo-
nents of the C 2
˜
͟
–X ͚ 3 transition appearing within, or
˜
2
ϩ
1
0
Ϫ1
close to, the 27 850–27 900 cm region. These perpendicu-
lar bands should show rotational structure characteristic of a
͓case ͑a͔͒– ͚ transition. There is a clear candidate for
one of the spin-orbit components, the band at 27 873 cm
˜
˜
5
. Complexity of the C–X spectrum: Evidence for
Renner–Teller Coupling
2
2
ϩ
͟
The excitation spectrum attributed to the C 2͟ –X ͚
˜
˜
2
ϩ
Ϫ1
,
transition of SrOH is characterized by a number of unusual
features: ͑i͒ a relatively large number of vibronic bands, ͑ii͒
both ͚–͚ and ͟–͚ vibronic transitions, and ͑iii͒ no obvious
regular vibrational progressions in the excitation spectrum.
Furthermore, the pronounced activity seen in the bending
vibrational mode in most of the dispersed fluorescence spec-
tra was also surprising. These observations clearly merit an
explanation.
The observation of substantial structure in the bending
mode would not be expected in the Franck–Condon limit if
the molecule is linear in both the ground and excited state. In
particular, transitions to vibrational states involving odd
changes in the bending mode quantum number are Franck–
Condon forbidden. Two scenarios would relax this con-
straint. One possibility is that SrOH is bent in the excited
state ͑SrOH is known to be linear in the ground electronic
state͒. However, this can be ruled out because the observa-
tion of a substantial excited state spin-orbit splitting for the
but there is no obvious spin-orbit partner. There are, how-
ever, several other perpendicular bands near this region ͑see
Table I͒ which complicates the picture.
A plausible source of some of these additional perpen-
dicular bands is that they arise from double quantum excita-
tion of the bending mode. A 2
͟ electronic state undergoing a
combination of Renner–Teller activity and spin-orbit cou-
pling will produce six vibronic components with different
energies, four 2
͟
and two ⌽, corresponding to the ͑020͒
2
34
2
vibrational manifold. Transitions to the ⌽ levels from the
ground electronic state will be optically forbidden, but all
four 2
͟ components are accessible. Consequently, by invok-
2
1
ing the 2 and 3 transitions, we would expect a total of six
0
0
perpendicular bands. Five perpendicular bands have been
Ϫ1
found between 27 750 and 27 890 cm and there may well
be at least one more, although this cannot be firmly identified
at present due to band congestion. If these bands arise from
2
0
1
0
the 2 and 3 transitions then the proximity of the 2␯ and
2
˜
˜
C–X origin shows that a linear equilibrium geometry is pre-
␯3 vibrational levels will inevitably give rise to Fermi reso-
served.
nance. This may explain why, if our general assignment is
2
We are therefore led to the alternative explanation,
namely that electronic-vibrational mixing occurs. In a
correct, the 2 bands are relatively strong, since intensity can
0
2
1
͟
be ‘‘stolen’’ from the 30 transition. Fermi resonance will
state the electronic orbital angular momentum may couple
with vibrational angular momentum from degenerate bend-
ing modes, giving rise to the Renner–Teller effect. One con-
also perturb the ͑020͒/͑001͒ vibronic energy level structure
and this in turn may explain why the perpendicular transi-
Ϫ1
tions in the 27 750–27 890 cm region do not seem to fit
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:
128.248.155.225 On: Sun, 23 Nov 2014 13:11:11

11250
J. Chem. Phys., Vol. 110, No. 23, 15 June 1999
M. S. Beardah and A. M. Ellis
into any obvious pattern. Because of this, and because of the
limited resolution in our experiments, we have not attempted
27 522 cmϪ1 band is
a
hot band instead of the
˜
2
˜
2
ϩ
C(010)␮ ͚–X ͚ ͑000͒ transition. It is perhaps conceiv-
specific assignments of individual bands in the 27 695–
able that there are two transitions contributing to the 27 522
cmϪ1 band, the main ␮ ͚–X ͚ transition and a weak
underlying hot band. However, we must admit that the ex-
panded view of the band in Fig. 4͑b͒ provides no convincing
evidence in support of this suggestion.
Ϫ1
2
7 900 cm region.
2
2
ϩ
˜
Although we have not made specific band assignments,
we are in a position to extract some limited information on
the proposed Renner–Teller effect in the C 2͟ state. The
˜
2
four
͟ vibronic components in the ͑020͒ manifold will split
into two groups of energy levels, the ␮ ͟1/2,3/2 and
͟1/2,3/2 levels. In the absence of Fermi resonance the
splitting between the ␮ and the ␬ levels is given approxi-
mately by an effective spin-orbit coupling constant, A␯*
To conclude this section, we have presented arguments
which provide possible assignments for several bands in the
˜ ˜
2
2
34
␬
C–X region. However, there are other bands for which no
assignments have been made. The Renner–Teller effect has
been employed to explain part of the complexity in the ex-
citation spectrum but this is clearly not the whole story. As
will be discussed in Sec. III B, there is good reason to be-
,
,K
2
where34
2
2
2
2
2
2
A* ,Kϭ
ͱ
A ϩ⑀ ␻ ͓͑␯ ϩ1͒ ϪK ͔.
͑1͒
Ϫ1
␯
2
lieve that one of the bands below 28 000 cm is part of the
2
˜
2
ϩ
˜
2
ϩ
D ͚ –X ͚ system. However, to explain the remaining
unassigned bands it seems that at least one additional, and as
yet unidentified, electronic system of SrOH must overlap this
region.
In Eq. ͑1͒ K is the quantum number for projection of the
vibronic angular momentum along the internuclear axis
2
͑Kϭ1 for
͟
vibronic components͒, A is the true spin-orbit
coupling constant ͑i.e., in the absence of vibronic interac-
tion͒, ⑀ is the Renner parameter, and ␻2 is the harmonic
frequency of the bending vibration. The separation between
B. The D ͚ –X 2͚؉ system
˜
2
؉
˜
Ϫ1
the 27 756 and 27 873 cm bands, which are the outer
2
ϩ
2
2
ϩ
Analogy with SrF would lead us to expect a ͚ elec-
tronic state of SrOH lying slightly above the C 2͟ state. We
have found the band system due to excitation to this new
members of the group of observed
͟ – ͚ bands in the
2
0
1
˜
2
/3 region, provides an upper limit for A*. Using
Aϭ24.52 cmϪ1, which was extracted from the C–X 0 tran-
sition, Eq. ͑1͒ yields ͉⑀␻2͉р40 cm . Furthermore, the mid-
point of this cluster of bands gives a crude estimate of 255
cm for ␻ . This is a much lower frequency for the bending
mode than in the ground electronic state (␯ ϭ361 cm ͒.
Having extracted information from the 2 /3 region, we
now turn our attention to the 2 region. If the parameters
extracted above are reasonable, then the 2 transition should
be centered at ϳ27 560 cmϪ1. The C 2͟ ͑010͒ vibronic
manifold will be composed of two ⌬ and two ͚ vibronic
components of which only the latter are optically accessible
from the zero point level of the ground electronic state. Ac-
0
0
0
˜ ˜
Ϫ1
2͚ϩ state, the D ͚ state. Figure 7 shows the key region.
A series of bands showing no evidence of spin-orbit structure
but forming a clear vibrational progression have been iden-
2
ϩ
˜
Ϫ1
2
Ϫ1
˜ ˜
tified in Fig. 7. These are assigned to the D ͚ –X ͚
2
ϩ
2
ϩ
2
2
0
1
0
transition. The vibrational progression, with an interval of
1
0
Ϫ1
ϳ630 cm , is clearly due to the Sr–O stretch since the O–H
1
0
stretch and the Sr–O–H bending modes are expected to have
much higher and lower frequencies, respectively. The Sr–O
˜
2
2
˜
2
ϩ
stretching frequency in the D ͚ state is approximately
0% higher than in the ground state, showing that the Sr–O
2
bond is significantly stronger in the excited state.
2
cording to Eq. ͑1͒, the separation between the two ͚ com-
Although it is straightforward to identify the Sr–O
stretching progression, it is not easy to locate the origin of
Ϫ1
Ϫ1
ponents will be р84 cm using
␧␻2
͉ ͉р40 cm . In other
words, two vibronic transitions are expected, the
the D–X system. The strongest band lies at 28 317 cmϪ1,
˜
˜
C(010)␮ ͚–X ͚ ͑000͒ transition at ϳ27 520 cmϪ1 and
˜
2
˜
2
ϩ
but extrapolation to the red yields another possible candidate,
the C(010)␬ ͚–X ͚ ͑000͒ transition at ϳ27 600 cmϪ1
˜
2
˜
2
ϩ
.
Ϫ1
the band at 27 698 cm . This band, whose rotational con-
2
ϩ
2
ϩ
Plausible candidates for these transitions can be found in
tour shows that it arises from a ͚ – ͚ transition, lies
˜ ˜
Ϫ1
the excitation spectrum. The band at 27 588 cm has rota-
amongst the densest group of bands in the C–X system.
tional structure consistent with a ͚– ͚ vibronic transition.
Furthermore, the emission spectrum from the upper vibronic
level shows prominent activity in the ground state bending
mode ͓see Fig. 5͑c͔͒, albeit in combination with one quantum
There is therefore a strong possibility that the corresponding
˜
2
ϩ
excited state, if it is associated with the D ͚ manifold,
2
ϩ
will be perturbed by levels of ͚ vibronic symmetry in the
˜
C
2͟ manifold.
Ϫ1
in the Sr-O stretch. Thus it seems likely that the 27 588 cm
˜
˜
2
˜
2
ϩ
Several of the C–X bands in this region have probable
˜
band is due to the C(010)␬ ͚–X ͚ ͑000͒ transition.
2͚ϩ
upper state vibronic symmetries. In the absence of vi-
There is a band at almost exactly the position expected for
2
bronic coupling the dispersed fluorescence spectrum would
help confirm the assignment, since no significant bending
the transition to the ␮ ͚ component, namely the band at
Ϫ1
2
7 522 cm . It too shows rotational structure consistent
˜
2
ϩ
˜
2
ϩ
mode activity would be expected for the D ͚ –X ͚
with a ͚– ͚ transition. It also shows very prominent excita-
tion in a single quantum of the bending mode in the dis-
persed fluorescence spectrum ͓see Fig. 5͑b͔͒. Unfortunately,
the dispersed fluorescence spectrum also registers emission
to the blue of the laser wavelength, suggesting that the
transition. The dispersed fluorescence spectra arising from
Ϫ1
excitation of the 28 956 and 29 590 cm transitions, which
˜
˜
are rather distant from the C–X manifold, show precisely
this scenario, namely a progression in only the Sr–O stretch.
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:
128.248.155.225 On: Sun, 23 Nov 2014 13:11:11

J. Chem. Phys., Vol. 110, No. 23, 15 June 1999
M. S. Beardah and A. M. Ellis
11251
FIG. 7. Laser excitation spectrum
˜
D
2͚ϩ –X 2͚ϩ and
˜
showing the
˜
E
2͚ϩ –X 2
˜
͚
ϩ
transitions of SrOH.
In both cases the only significant vi-
brational structure is a progression in
the Sr–O stretching mode. This spec-
trum was obtained by joining scans of
three different laser dyes, hence the
uneven background signal.
cmϪ1. As for the D ͚ state, this is a significant increase
˜
2
ϩ
However, dispersed fluorescence spectra obtained by pump-
ing the 28 317 and 27 698 cm transitions both show more
Ϫ1
Ϫ1
from the 535 cm frequency observed for the ground state.
complex vibrational structure with substantial activity in
both the Sr–O stretch and the bending mode. Clearly the
dispersed fluorescence spectra do not provide a definitive
Comparable findings are found for the metal–ligand stretch
in the alkaline earth monohalides, with all electronic states
above the C 2͟ state showing an increase in frequency com-
˜
Ϫ1
assignment. Consequently, even though the 27 698 cm
31
pared with the ground state. This is presumably due to an
0
band is identified as the 0 transition in Table I, it is not clear
0
increase in ionic bonding as the unpaired electron gains more
Rydberg character.
Ϫ1
whether it or the 28 317 cm
band is the actual
˜
2
ϩ
˜
2
ϩ
D ͚ –X ͚ origin.
C. The E 2͚ –X 2͚؉ system
˜
؉
˜
D. The F 2͟ –X 2͚؉ system
˜
˜
˜
2
ϩ
˜
2
ϩ
In addition to the D ͚ –X ͚ system, Fig. 7 also
shows another new system of SrOH. If the electronic state
ordering for SrF is assumed for SrOH, the next highest op-
tically accessible excited state of SrOH would be the ana-
Scans to higher wave numbers revealed yet another new
electronic transition of SrOH. The spectrum, shown in Fig. 8,
Ϫ1
is dominated by the pair of peaks at 32 974 and 32 996 cm
.
These appear to form a spin-orbit doublet which, if correct,
2
Ϫ1 31
Ϫ1
logue of the E
͟ state of SrF (T ϭ31 600 cm ͒. The
͟ state of SrF has a spin-orbit coupling constant of ϳ85
cm and so a substantial spin-orbit splitting should also be
seen for SrOH. However, no evidence of any spin-orbit split-
ting is seen for this new system in Fig. 7. Consequently, the
upper electronic state is assigned as another ͚ state, pre-
sumably the analogue of the F ͚ state of SrF (T ϭ32 820
cm ͒. This assignment is confirmed by the rotational con-
tours of individual bands, which show that they arise from
e
would imply an excited state spin-orbit splitting of 22 cm
.
2
E
This assignment is supported by the rotational contours,
Ϫ1
2
2
which have the profiles expected for a
͟
– ͚ transition in
which the 2
͟
state undergoes Hund’s case ͑a͒ coupling.
There is one other doublet in the spectrum with this splitting,
2
ϩ
Ϫ1
namely the two weak peaks at 33 567 and 33 587 cm
,
2
ϩ
e
which most probably arise from excitation of the Sr–O
Ϫ1 31
Ϫ1
stretch, giving an excited state frequency of 592 cm for
this mode. However, this assignment should be viewed as a
tentative one given the relatively low intensities of these fea-
tures.
2
ϩ
2
ϩ
͚
– ͚ transitions. To be consistent with the labels used
for the other excited states, the new excited state of SrOH is
˜
2
ϩ
The upper state is assigned as a 2
designated as the E ͚ state.
͟
state, which we label
˜
2
ϩ
˜
2
ϩ
as the F 2
smaller than in the C 2
approximately midway between that of the E
states of SrF. However, the latter is known to possess a much
˜
Unlike the D ͚ –X ͚ system, the origin of the
͟
state. The spin-orbit splitting is small, even
˜
2
ϩ
˜
2
ϩ
˜
state. The F 2
˜
͟
E ͚ –X ͚ system is easy to locate. The band at 29 990
͟
state of SrOH is
Ϫ1
0
0
2
͟ ͟
and G 2
cm corresponds to the 0 transition, since it is intense and
no bands are found at lower wave numbers which fit into the
clear Sr–O stretching progression seen on the high wave
Ϫ1 35
smaller spin-orbit coupling constant, Aϭ22.94 cm
,
a
number side. Consequently, the bands at 30 610, 31 222,
value almost identical with that reported here for SrOH.
Ϫ1
31 834, and 32 439 cm are, respectively, assigned to one,
Consequently, it seems likely that the F 2͟ state of SrOH is
˜
two, three, and four quanta in the Sr–O stretch. This gives an
the analogue of the G 2
͟
state of SrF. This would leave the
SrOH analogue of the E ͟ state of SrF unaccounted for.
˜
2
ϩ
2
E ͚ excited state Sr–O stretching frequency of ϳ612
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:
128.248.155.225 On: Sun, 23 Nov 2014 13:11:11

11252
J. Chem. Phys., Vol. 110, No. 23, 15 June 1999
M. S. Beardah and A. M. Ellis
FIG. 8. Laser excitation spectrum
showing the F 2
˜
͟
˜
–X 2
͚
ϩ
transition
of SrOH. A fall off in laser dye effi-
ciency at the low wave number end of
the spectrum distorts the relative in-
tensities of the two origin sub-bands.
At the high wave number end of the
spectrum a strong atomic strontium
transition is seen.
E. Character of excited electronic states of SrOH
However, it would appear at first sight that the situation
for the C 2
˜
͟
state is quite different given the far more com-
The D ͚ , E ͚ , and F 2͟ states of SrOH have
˜
2
ϩ
˜
2
ϩ
˜
plex LIF spectrum. To try and understand this, analogy be-
tween the alkaline-earth monohalides and the monohydrox-
ides is, once again, a good starting point. In the alkaline-
much in common with the A 2
˜
͟ and B ͚ states. Transi-
˜
2
ϩ
tions to these states from the ground electronic state yield
simple LIF spectra in which the only significantly Franck–
Condon active mode is the Sr–O stretching vibration. The
Sr–O stretching frequencies ͑see Table II͒ are larger in the
earth monohalides there is strong evidence that the unpaired
electron in the ␲ orbitals in the A 2
͟
and C
2
͟
states have
3
6–38
similar metal atomic orbital characters.
is also an important difference. In order for the two ␲ orbit-
However, there
˜
˜
˜
˜ ˜
˜
D, E, and F states than in the X, A, and B states but the
increase is modest (Ͻ20%͒ and explicable in terms of in-
creased ionic character in the higher states. It is therefore
reasonable to assume that transitions between these states
involve movement of the unpaired electron between different
nonbonding metal-centered orbitals.
als to be orthogonal they must possess opposite polariza-
tions. In the A 2
͟
state the ␲ orbital is known to be polar-
ized away from the ligand. Consequently, the ␲ orbital in the
2
36–40
C
͟
state should be polarized towards the ligand,
it will be ‘‘reverse polarized.’’
Making the reasonable assumption that the same process
i.e.,
occurs in the alkaline-earth monohydroxides, the C 2
˜
͟ state
TABLE II. Spectroscopic constants ͑in cmϪ1͒ of new SrOH electronic
of SrOH should also be reverse polarized. In the alkaline-
earth monohalides the reverse polarized state possesses a
much larger electric dipole moment than the other known
electronic states. This is because the reverse polarization pre-
vents the buildup of significant unpaired electron density on
the opposite side of the metal from the ligand. In the other
known states this lobe of negative charge goes some way
towards canceling the dipole moment generated by the polar
˜
states.
˜
C
2͟
T00
27 303
A
B
⑀␻2
␻2
24.52Ϯ0.04a
0.2564Ϯ0.0009
р40b
a
͉
͉
ϳ255b
D
2͚ϩ
T00
27 698a
630
˜
␯Sr–O
metal ligand bond. In the reverse-polarized C 2
͟
state of
˜
2͚ϩ
E
T00
29 990
612
SrOH we would therefore expect a relatively large electric
dipole moment. This in turn has important implications for
␯Sr–O
˜
˜
˜
2͟
the spectroscopy of the C–X system.
F
T00
A
32 985
22
Bolman and Brown33 have described a mechanism by
which vibronically induced parallel bands can gain intensity.
This model was developed for transitions between
͓592͔c
␯Sr–O
2
͟ and
a
These values are for the C 2͟ zero point level. The errors quoted represent
one standard deviation and were derived from a least squares fit to rota-
tional structure in the C 2͟ –X ͚ 0 band.
See text for details.
˜
2
2
͚
states in a linear triatomic molecule. The ͟ state can
2
gain some ͚ character through Renner–Teller interaction
when the bending vibration is excited, and this in turn con-
fers a nonzero probability to transitions which involve single
˜
˜
2
ϩ
0
0
b
c
This value is based on a tentative band 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:
128.248.155.225 On: Sun, 23 Nov 2014 13:11:11

J. Chem. Phys., Vol. 110, No. 23, 15 June 1999
M. S. Beardah and A. M. Ellis
11253
quantum excitation of the bending mode. According to Bol-
bending mode structure seen in dispersed fluorescence spec-
tra which would be forbidden in the Franck–Condon limit
for a linear molecule. This bending mode structure was rela-
tively weak, indicating that any deviation from linearity in
man and Brown, the intensity of these nominally forbidden
1
0
bands, which includes the 2 band of SrOH, is proportional
to the square of the difference in dipole moments between
the upper and lower electronic states.
˜
˜
the F state is modest. The suspicion must be that the F state
of CaOH correlates with the reverse-polarized C 2
͟
state of
Given the proximity of electronic states surrounding the
˜
C
2͟ state, several of which have ͚ character, the two-state
2
CaF, a correlation supported by their similar electronic ener-
Ϫ1
˜
˜
gies (T00 is 30 215 and 30 217 cm for CaOH and CaF,
model is likely to be a severe oversimplification for the C–X
system of SrOH. However, as pointed out by Bolman and
Brown, the same qualitative arguments will apply to a mul-
respectively͒. We speculate that the non-linear geometry re-
sults from increased metal–ligand overlap due to the reverse-
polarized character of the excited state. A greater degree of
covalency is expected for CaOH than for SrOH or BaOH and
this may be sufficiently large in the case of CaOH to favor a
nonlinear equilibrium geometry.
2
tistate model, the quantitative difference being that each ͚
state will contribute a dipole difference term weighted by an
appropriate coefficient describing the mixing with the C 2͟
˜
state. Since all nearby 2͚ states are expected to possess
much smaller dipole moments than the C 2͟ state, this
˜
IV. CONCLUSIONS
mechanism could give substantial intensity to transitions in-
˜
˜
volving ⌬␯ ϭϮ1 in the C–X system. In fact a combination
Laser-induced fluorescence spectra of several new elec-
tronic transitions of the SrOH free radical have been re-
ported. Four transitions have been firmly identified of which
2
38,39,41,42
of experimental and theoretical data
alkaline-earth monohalides and monohydroxides indicates
that the electric dipole moment of the C 2͟ state of SrOH
will be close to the limit expected for Sr –OH and will be
some three to four times larger than in most nearby elec-
tronic states ͑including the A 2
would expect, for example, the Franck–Condon forbidden
␯ ϭϮ1 transitions in the C 2͟ –X ͚ system to be at
for various
˜
three, the D ͚ –X ͚ , E ͚ –X ͚ , and F 2͟
˜
˜
2
ϩ
˜
2
ϩ
˜
2
ϩ
˜
2
ϩ
˜
–
ϩ
2
ϩ
X ͚ systems, yield relatively simple spectra dominated by
progressions in the Sr–O stretching vibration.
˜
͟
state͒. On this basis we
However, far more complex spectra are observed for the
˜
2͟ –X ͚ system. The origin of this system has been
˜
2
ϩ
C
˜
˜
2
ϩ
⌬
established, but higher frequency bands do not fit into regu-
lar vibrational progressions and their assignment is uncer-
tain. Laser excitation of several of these transitions yield
marked Sr–O–H bending mode activity in dispersed fluores-
cence spectra. One source of this complexity is thought to be
˜
2
least an order of magnitude more intense than in the
˜
A
˜
2͟ –X ͚ system. Since such transitions are seen with
2
ϩ
˜
˜
substantial intensities in the C–X excitation spectrum and in
several dispersed fluorescence spectra, these observations
support the suggestion that the C 2
a substantial Renner–Teller effect in the C 2
͟ state. The
˜
͟
state is reverse polar-
˜
2͟ state of SrOH is the analogue of the reverse-polarized
states known for the alkaline-
C
ized.
͑
or back-polarized͒ excited 2
͟
To close Sec. III E we consider the effect the reverse-
polarized character of the C 2
˜
͟ state is likely to have on the
earth monohalides.
All of the new SrOH transitions yield strong LIF spectra
and so all should be excellent candidates for higher resolu-
tion work. Such studies will be essential for providing a
complete band assignment. Fully rotationally resolved spec-
tra should also yield further insight into the character of the
˜
Renner parameter. Since the bending vibration will be domi-
nated by motion of the hydrogen atom, the coupling between
vibrational and angular momenta should be weak if the un-
paired electron is strongly metal localized. On these grounds
˜
one would expect a larger Renner parameter for the C 2
͟
C
2͟ state, especially for clarifying the extent of Renner–
state than for the A 2
˜
͟
state because the polarization of the
Teller coupling. It is hoped that the present work will stimu-
late interest in this problem.
highest occupied molecular orbital in the former favors
greater delocalization of the unpaired electron density onto
the ligand. Comparison of the reported Renner parameter for
the A 2͟ state (␧ϭϪ0.0791)15 with the limits we have de-
ACKNOWLEDGMENTS
˜
˜
duced in this work for the C 2
with this idea. Nevertheless, it seems likely that the unpaired
electron is still predominantly metal localized in the C 2
state.
The authors acknowledge the financial support of the
UK Engineering and Physical Sciences Research Council for
the award of a studentship to one of the authors ͑M.S.B.͒.
M.S.B. is also grateful to the Society of Chemical Industry
for the award of a Messel Scholarship. The program
SPECSIM,43 which was used to obtain rotational structure
simulations, was kindly provided by Terry A. Miller.
͟
state (␧р0.16) is consistent
˜
͟
F. Comparison with other MOH radicals
It is interesting that we have apparently observed reverse
2
polarized
͟
excited electronic states of both SrOH and
1
J. Nakagawa, R. F. Wormsbecher, and D. O. Harris, J. Mol. Spectrosc. 97,
BaOH given that no analogous state of CaOH was identified
in the work of Pereira and Levy.18 However, Pereira and
Levy did report an intriguing new electronic state of CaOH,
37 ͑1983͒.
2
R. C. Hilborn, Z. Qingshi, and D. O. Harris, J. Mol. Spectrosc. 97, 73
͑
1983͒.
3
4
C. R. Brazier and P. F. Bernath, J. Mol. Spectrosc. 114, 163 ͑1985͒.
P. F. Bernath and S. Kinsey-Neilsen, Chem. Phys. Lett. 105, 663 ͑1984͒.
˜
the F state, which was judged to be nonlinear on the basis of
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:
28.248.155.225 On: Sun, 23 Nov 2014 13:11:11
1

11254
J. Chem. Phys., Vol. 110, No. 23, 15 June 1999
M. S. Beardah and A. M. Ellis
5
27
28
29
W. T. M. L. Fernando, M. Douay, and P. F. Bernath, J. Mol. Spectrosc.
44, 344 ͑1990͒.
E. S. J. Robles, A. M. Ellis, and T. A. Miller, J. Am. Chem. Soc. 114,
7171 ͑1992͒.
E. S. J. Robles, A. M. Ellis, and T. A. Miller, J. Phys. Chem. 96, 8791
1
6
C. J. Whitham, B. Soep, J.-P. Vistocot, and A. Keller, J. Chem. Phys. 93,
91 ͑1990͒.
9
͑
1992͒.
7
8
T. Gustavsson, C. Alcaraz, J. Berlande, J. Cuvellier, J.-M. Mestdagh, P.
Meynadier, P. de Pujo, O. Sublemontier, and J.-P. Visticot, J. Chem. Phys.
L. M. Ziurys, W. L. Barclay, and M. A. Anderson, Ap. J. 384, L63 ͑1992͒.
S. J. Pooley, M. S. Beardah, and A. M. Ellis, J. Electron Spectrosc. Relat.
Phenom. 97, 77 ͑1998͒.
K. P. Huber and G. Herzberg, Molecular Spectra and Molecular Struc-
ture. IV. Constants of Diatomic Molecules ͑Van Nostrand Reinhold, New
York, 1979͒.
G. Herzberg, Molecular Spectra and Molecular Structure. I. Spectra of
Diatomic Molecules ͑Krieger, Malabar, Florida, 1989͒.
P. S. H. Bolman and J. M. Brown, Chem. Phys. Lett. 21, 213 ͑1973͒.
G. Herzberg, Molecular Spectra and Molecular Structure. III. Electronic
Spectra and Electronic Structure of Polyatomic Molecules ͑Van Nostrand
Reinhold, New York, 1966͒.
C. Nitsch, J. O. Schr o¨ der, and W. E. Ernst, Chem. Phys. Lett. 148, 130
͑1988͒.
L. Klynning and H. Martin, Phys. Scr. 24, 33 ͑1981͒.
S. F. Rice, H. Martin, and R. W. Field, J. Chem. Phys. 82, 5023 ͑1985͒.
P. Bundgen, B. Engels, and S. D. Peyerimhoff, Chem. Phys. Lett. 176,
407 ͑1991͒.
30
1
45, 210 ͑1991͒.
R. A. Hailey, C. N. Jarman, W. T. M. Fernando, and P. F. Bernath, J. Mol.
Spectrosc. 147, 40 ͑1991͒.
J. A. Coxon, M. Li, and P. I. Presunka, J. Mol. Spectrosc. 150, 33 ͑1991͒.
T. C. Steimle, D. A. Fletcher, K. Y. Jung, and C. T. Scurlock, J. Chem.
Phys. 96, 2556 ͑1992͒.
C. N. Jarman and P. F. Bernath, J. Chem. Phys. 97, 1711 ͑1992͒.
M. Li and J. A. Coxon, J. Chem. Phys. 97, 8961 ͑1992͒.
J. A. Coxon, M. Li, and P. I. Presunka, Mol. Phys. 76, 1463 ͑1992͒.
P. I. Presunka and J. A. Coxon, Can. J. Chem. 71, 1689 ͑1993͒.
P. I. Presunka and J. A. Coxon, J. Chem. Phys. 101, 201 ͑1994͒.
M. Li and J. A. Coxon, J. Chem. Phys. 102, 2663 ͑1995͒.
P. I. Presunka and J. A. Coxon, Chem. Phys. 190, 97 ͑1995͒.
R. Pereira and D. H. Levy, J. Chem. Phys. 105, 9733 ͑1996͒.
C. Zhao, P. G. Hajigeorgiou, P. F. Bernath, and J. W. Hepburn, J. Mol.
Spectrosc. 176, 268 ͑1996͒.
31
9
0
1
32
1
1
1
1
1
1
1
1
1
1
2
3
4
5
6
7
8
9
33
34
35
36
37
2
0
38
R. A. Hailey, C. Jarman, and P. F. Bernath, J. Chem. Phys. 107, 669
͑1997͒.
2
2
1
2
M. Douay and P. F. Bernath, Chem. Phys. Lett. 174, 230 ͑1990͒.
T. C. Steimle, D. A. Fletcher, K. Y. Jung, and C. T. Scurlock, J. Chem.
Phys. 97, 2909 ͑1992͒.
C. T. Scurlock, T. C. Steimle, R. D. Suenram, and F. J. Lovas, J. Chem.
Phys. 100, 3497 ͑1994͒.
39
40
¨
W. E. Ernst and J. Kandler, Phys. Rev. A 39, 1575 ͑1989͒.
W. E. Ernst, J. K a¨ ndler, C. Noda, J. S. McKillop, and R. N. Zare, J. Chem.
Phys. 85, 3735 ͑1986͒.
T. T o¨ rring, W. E. Ernst, and J. K a¨ ndler, J. Chem. Phys. 90, 4927 ͑1989͒.
A. R. Allouche, G. Wannous, and M. Aubert-Fr e´ con, Chem. Phys. 170, 11
2
3
4
4
1
2
2
2
4
5
L. C. O’Brien and P. F. Bernath, J. Am. Chem. Soc. 108, 5017 ͑1986͒.
A. M. Ellis, E. S. J. Robles, and T. A. Miller, J. Chem. Phys. 94, 1752
͑1993͒.
43
S. Panov, X. Q. Tan, and T. A. Miller, Proceedings of the 48th Ohio State
University International Symposium on Molecular Spectroscopy, 1993,
paper RA06.
͑1991͒.
2
6
E. S. J. Robles, A. M. Ellis, and T. A. Miller, J. Phys. Chem. 96, 3247
1992͒.
͑
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:
28.248.155.225 On: Sun, 23 Nov 2014 13:11:11
1