The Astrophysical Journal, 576:1108–1114, 2002 September 10
# 2002. The American Astronomical Society. All rights reserved. Printed in U.S.A.
ROTATIONAL REST FREQUENCIES FOR CrO (X 5Pr) AND CrN (X 4)
P. M. Sheridan, M. A. Brewster, and L. M. Ziurys
Department of Chemistry, Department of Astronomy, and Steward Observatory, 933 North Cherry Avenue, University of Arizona,
Tucson, AZ 85721; [email protected]
Received 2002 March 28; accepted 2002 May 15
ABSTRACT
and CrN (X 4) have been recorded using millimeter/subThe pure rotational spectra of CrO (X
millimeter-wave spectroscopy in the frequency range 248–636 GHz. These radicals were created by the reaction of chromium vapor, produced in a Broida-type oven, with N2O or NH3 under DC discharge conditions.
For CrO, 12 rotational transitions were recorded, in which up to five spin-orbit components were observed—
the first measurement of the pure rotational spectrum of a molecule in a 5P electronic state. Nine rotational
transitions were recorded for CrN; here all four spin components were measured in every transition. The data
obtained for both radicals were analyzed using an appropriate effective Hamiltonian. The resulting spectroscopic parameters of these species were determined to high precision, including those related to fine structure
and lambda doubling. This work will enable radioastronomical searches for CrN and CrO to be carried out.
CrO has already been observed in the optical spectra of stellar atmospheres.
Subject headings: ISM: molecules — line: identification — methods: laboratory — molecular data
5P )
r
meter transition frequencies with the necessary accuracy
for astronomical searches (<1 MHz). Therefore, interstellar studies of Cr-bearing molecules have been extremely
limited.
Two molecules of astrophysical interest are CrO and
CrN. Both species have been the subject of several previous
spectroscopic investigations. For example, Merer and coworkers measured the 5 þ X 5 , 5 D X 5 , and 5 X 5 electronic transitions of CrO with rotational resolution
(Hocking et al. 1980; Cheung, Zyrnicki, & Merer 1984;
Barnes et al. 1993). Their studies also involved further development of the specific Hamiltonians necessary to describe
these quintet electronic states. In addition, the permanent
electric dipole moment of CrO was measured using optical
stark spectroscopy (Steimle et al. 1989). The CrN radical
was first investigated by ab initio and density functional
theory (DFT) methods, which predicted a 4 ground electronic state (Harrison 1996; Andrews, Bare, & Chertihin
1997). This prediction was confirmed by the measurement
of the A 4 X 4 transition using laser-induced fluorescence (LIF) spectroscopy (Balfour et al. 1997). Furthermore, the permanent electric dipole moment was measured,
as well as fine and hyperfine parameters using data obtained
by pump/probe microwave-optical double-resonance
(PPMODR) techniques (Steimle, Robinson, & Goodridge
1999; Namiki & Steimle 1999).
In this paper we report the first measurements of the pure
rotational spectrum of both CrO and CrN in their respective
X 5Pr and X 4 ground electronic states. The purpose of
this work was to directly record rotational transitions of
these radicals for astronomical identification at millimeter/
submillimeter wavelengths. The molecules were studied
using direct-absorption methods in the frequency range
248–636 GHz. For both species, fine-structure splittings
were resolved, and in the case of CrO, lambda-doubling
interactions as well. These data have been analyzed to produce a revised set of spectroscopic parameters for these molecules. Here we present our results.
1. INTRODUCTION
Several small molecules containing metals (in the chemist’s sense) have been detected to date toward the circumstellar envelopes of late-type stars, in particular carbon-rich
objects. For example, NaCl and AlNC have been identified
in the expanding shell of the AGB star IRC +10216 (Cernicharo & Guélin 1987; Ziurys et al. 2002), and NaCN and
MgNC have been observed in the envelope of the post-AGB
object CRL 2688 (Highberger et al. 2001). These species
have all been identified in these sources on the basis of their
pure rotational spectrum, observed using millimeter-wave
telescopes. Crucial to these detections have been laboratory
measurements, in particular those at high spectral resolution (e.g., Robinson, Apponi, & Ziurys 1997).
Given the observation of sodium-, magnesium-, aluminum-, and even potassium-bearing molecules in circumstellar gas, it is possible to contemplate species containing other
metals. One such metal is chromium, which is thought to be
primarily produced in explosive silicon burning in supernovae (Woosley & Weaver 1995). Nonetheless, it has a relatively large cosmic abundance compared to some of the
other early iron group metals. For example, titanium has an
abundance Ti=H 8:5 108 , while the Cr/H ratio is
4:8 107 (Savage & Sembach 1996). The solar abundance
of chromium is therefore only about a factor of 5 less than
that of sodium and aluminum. Indeed, CrO is a known contributor to the electronic spectra found in stellar atmospheres (e.g., Davis 1947). However, thus far, chromiumbearing molecules have not been observed in circumstellar
gas primarily because there has been a dearth of high-resolution laboratory data for such small, astrophysically relevant species. Although several possible chromium
compounds have been investigated via their electronic transitions at optical wavelengths (Barnes, Hajigeorgiou, &
Merer 1993; Balfour, Qian, & Zhou 1997), these studies
have not had sufficient resolution to determine spectroscopic constants that can reproduce millimeter/submilli1108
ROTATIONAL REST FREQUENCIES
2. EXPERIMENTAL
The pure rotational spectra of CrO and CrN were measured using one of the millimeter/submillimeter-wave spectrometers of the Ziurys group (Ziurys et al. 1994). The
instrument consists of a set of phase-locked Gunn oscillators and Schottky diode multipliers that produce radiation
over the frequency range 65–650 GHz, a double-walled,
water-cooled, steel reaction chamber containing a Broidatype oven, and an InSb bolometer detector cooled to
approximately 4 K. The radiation is launched from the
source as a Gaussian beam, directed into the reaction cell
via a set of offset ellipsoidal mirrors, and passed into the
reaction chamber through a polystyrene window. At the
rear of the cell, a rooftop reflector rotates the plane of polarization of the radiation by 90 and propagates it back
through the cell and optics into the detector. Phase-sensitive
detection is employed using FM source modulation.
CrO was produced by the reaction of chromium metal
vapor with N2O. The vapor was created in a high-temperature Broida-type oven, packed with alumina and zirconia
insulation to achieve the necessary high temperatures.
Approximately 10–15 mtorr each of N2O and argon, the
carrier gas used to entrain the metal vapor, were introduced
into the reaction chamber through the bottom of the oven.
A DC discharge was not required to produce CrO; however,
use of a discharge enhanced the synthesis of the molecule
(0.01 A at 200 V). CrN was created by an identical method,
except that NH3 was substituted for N2O and a discharge
was required with higher current (0.7 A).
For both molecules, rotational transitions were recorded
over the frequency range 248–636 GHz. Line widths ranged
from 600 kHz at 248 GHz to 1700 kHz at 636 GHz. Final
frequency measurements were obtained by fitting Gaussian
curves to the line profiles. These profiles consist of averages
of one to six scan pairs, 5 MHz in coverage, with an equal
number increasing and decreasing in frequency.
3. RESULTS
The transition frequencies measured for CrO are presented in Table 1; in all, 12 rotational were recorded. The
ground state of CrO is 5Pr, and therefore the spectra are
complicated by the presence of L x S coupling, which splits
every rotational transition into five spin-orbit components,
labeled by the quantum number . This quantum number is
the sum of the projection of the electron orbital angular
momentum along the internuclear axis, , and the projection of the electron spin angular momentum, , i.e.,
¼ þ , as is appropriate for a Hund’s case (a) coupling
scheme. Since jj ¼ 1 and S ¼ 2 in a 5P state, takes on
the values 1, 0, 1, 2, and 3; hence, five sublevels are created.
In addition, there can be further splittings in every component due to lambda-doubling interactions, which are
potentially present in every degenerate electronic state. This
effect can split each level into doublets, which are labeled
by the parity notation e and f.
As shown in Table 1, both spin-orbit and lambda-doubling interactions were observed in the data recorded for
CrO. In six of the 12 rotational transitions, all five spin-orbit
components were observed and their frequencies measured.
(Poor signal-to-noise ratio in the spectra or gaps in the multiplier coverage prevented observation of all five sublevels in
every transition.) Lambda-doubling interactions were also
1109
resolved in all spin-orbit components except for the ¼ 3
sublevel. (The e and f parity labels were assigned assuming
that the dominant perturber is the A 5+ state; Bauschlicher, Nelin, & Bagus 1985.) For the ¼ 0 sublevel, the
splitting was the largest (500–700 MHz) and was found to
increase with increasing J quantum number. For the ¼ 1,
2, and 3 components the lambda doubling successively
decreased in magnitude, although for the individual sublevels there was an increase in the splitting with J. The opposite effect was observed for the ¼ 1 sublevel, where the
separation decreased with J, such that it was only a few
MHz at the J ¼ 19
18 transition.
Figure 1 shows a stick spectrum of the J ¼ 16
15 rotational transition, which covers the range 485–517 GHz.
The experimentally observed intensities are also shown in
the diagram, as are the lambda-doubling splittings. The
total splitting of the sublevels in frequency space in this
transition is rather large—almost 32 GHz, which is about
twice the rotational constant. However, the splitting
between the components is fairly regular. The spin-orbit
component lying lowest in frequency corresponds to the one
that is lowest in energy, ¼ 1. About 9 GHz higher in frequency is the ¼ 0 sublevel, followed by ¼ 1, 2, and 3.
Because each successive spin-orbit component lies higher in
energy by 63 cm1, the intensity slowly decreases for
> 0. (The ¼ 3 line is stronger than the ¼ 1 and 2 features only because there is no lambda doubling; thus, this
line gains a factor of 2 in intensity.)
Figure 2 displays representative spectra of CrO near 486
GHz. Here the ¼ 1 sublevel of the J ¼ 16
15 transition is shown, along with the ¼ 3 component of the
J ¼ 15
14 transition. The effects of lambda doubling are
evident in the ¼ 1 line, which is split into doublets. In
the ¼ 3 feature, such effects are negligible and it appears
as a single line, increasing its intensity relative to the other
component.
Table 2 presents the measured transition frequencies for
CrN in its ground 4 electronic state. Data for nine rotational transitions were recorded over the frequency range
294–636 GHz. The spectra are complicated because of the
Fig. 1.—Stick spectrum of the J ¼ 16
15 transition of CrO. All five
spin-orbit components are shown with their approximate measured intensities. Their regular but wide (7–9 GHz) spacing is apparent. The ¼ 0
and 1 spin-orbit components are split by lambda-type doubling into two
well-separated lines, while this effect is smaller in the ¼ 1 and 2 components. The ¼ 3 spin-orbit component appears as the strongest feature
since the lambda doublets are collapsed into a single line.
1110
SHERIDAN, BREWSTER, & ZIURYS
TABLE 1
Measured Rotational Transition Frequencies for
CrO (X 5Pr)a
J þ1
J
8 7 ............
9 8 ............
10 9 ..........
11 10 ........
12 11 ........
13 12 ........
14 13 ........
15 14 ........
Parity
obs
obs calc
0
0
1
1
2
2
3
3
1
1
0
0
1
1
2
2
3
3
1
1
0
0
1
1
1
1
0
0
1
1
2
2
3
3
1
1
0
0
1
2
2
3
3
1
1
0
0
1
1
2
2
3
3
1
1
2
2
3
3
1
1
0
0
1
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
247,587.324
248,019.651
252,162.958
251,811.237
255,710.397
255,701.467
259,184.975
259,184.975
273,231.590
273,181.301
278,534.967
279,009.559
283,659.746
283,274.706
287,649.794
287,637.182
291,551.121
291,551.121
303,607.735
303,556.262
309,482.214
309,995.245
315,148.286
314,733.448
333,988.891
333,937.489
340,428.850
340,976.285
346,627.973
346,186.991
351,502.035
351,479.606
356,248.848
356,248.848
364,375.388
364,325.292
371,374.703
371,952.340
378,097.986
383,412.907
383,384.496
388,578.080b
388,578.080b
394,767.534
394,720.047
402,319.707
402,923.057
409,557.772
409,076.184
415,312.743
415,277.240
420,892.769b
420,892.769b
441,006.513
440,510.592
447,200.388
447,156.920
453,191.756b
453,191.765b
455,569.497
455,531.352
464,205.828
464,847.262
472,443.662
0.035
0.024
0.089
0.084
0.028
0.015
0.008
0.020
0.001
0.007
0.012
0.031
0.030
0.025
0.021
0.002
0.032
0.020
0.004
0.006
0.033
0.011
0.015
0.031
<0.000
0.011
0.026
0.006
0.032
0.022
0.019
0.036
0.069
0.070
0.007
0.018
0.015
0.004
0.007
0.100
0.009
0.129
0.083
0.014
0.030
0.038
0.004
0.043
0.008
0.013
0.044
0.180
0.132
0.024
0.064
0.081
0.028
0.279
0.165
0.005
0.020
0.003
<0.000
0.013
Vol. 576
TABLE 1—Continued
J þ1
J
15 14 ........
16 15 ........
17 16 ........
18 17 ........
19 18 ........
a
b
Parity
obs
obs calc
1
2
2
3
3
1
1
0
0
1
1
2
2
3
3
1
1
0
0
1
1
2
2
2
2
1
1
0
0
1
1
2
2
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
e
f
471,937.663
479,074.839
479,022.766
485,474.018b
485,474.018b
485,979.644
485,948.120
495,146.531
495,800.196
503,868.636
503,356.501
510,935.551
510,873.877
517,738.574b
517,738.574b
516,396.106
516,372.421
526,085.209
526,746.600
535,280.830
534,766.663
542,781.637
542,709.564
574,612.008
574,528.989
577,247.575
577,243.515
587,955.228
588,619.119
598,064.947
597,558.428
606,426.279
606,331.313
0.042
0.033
0.041
0.413
0.202
0.022
0.003
0.012
0.016
0.017
0.024
0.024
0.004
0.501
0.329
0.042
0.005
0.009
0.019
0.015
0.037
0.054
0.025
0.072
0.031
0.024
0.025
0.086
0.037
0.062
0.049
0.014
0.075
In MHz.
Not included in fit.
presence of three unpaired electrons in this molecule. The
resultant spin angular momentum couples with the molecular frame rotation, indicated by quantum number N, to produce fine structure, labeled by J, where J ¼ N þ S. Four
fine-structure components are therefore generated per rotational level, and all four were recorded in every transition.
The separation between the fine-structure components was
found to decrease with increasing N, such that the total separation is 6 GHz in the N ¼ 8
7 rotational transition,
decreasing to 1.6 GHz in the N ¼ 17
16 lines.
In Figure 3 a typical spectrum for CrN is shown, illustrating the fine-structure splittings. Here all four spin components of the N ¼ 13
12 rotational transition near 484–
487 GHz are displayed, necessitating a frequency gap in the
spectrum. The fine-structure lines are not evenly spaced,
and the ¼ 3=2 component (J ¼ 14:5
13:5) lies lower in
frequency than that corresponding to ¼ 12 (J ¼
13:5
12:5). These two components actually shift relative
to each other as a function of N.
A diagram illustrating the shift of the fine-structure components is presented in Figure 4. Here stick figures of the
N¼9
8, 12
11, and 17
16 transitions are shown.
For the N ¼ 9
8 transition, the ¼ 3=2 component lies
almost 0.5 GHz lower in frequency relative to the ¼ 12 line,
but the separation successively narrows such that at the
N ¼ 17
16 transition, it actually lies at higher frequency.
No. 2, 2002
ROTATIONAL REST FREQUENCIES
1111
TABLE 2
Measured Rotational Transition Frequencies for
CrN (X 4)a
N þ1
N
8 7 ..............
9 8 ..............
10 9 ............
11 10 ..........
Fig. 2.—Spectrum showing the ¼ 1 component of the J ¼ 16
15
transition and the ¼ 3 line in the J ¼ 15
14 transition of CrO (X 5Pr),
measured in the laboratory near 486 GHz. The lambda doubling is seen in
the ¼ 1 spin-orbit component, which is split into two separate features.
The lambda-type doubling in the ¼ 3 component is not resolved and
therefore appears as a single line. This spectrum is a composite of six scans,
each 100 MHz in frequency coverage, and acquired in 1 minute.
The other components are moving closer to each other as
well, such that at very high N, a nicely spaced quartet would
be expected per transition, similar to KC (Xin & Ziurys
1999). The change from an irregular quartet to a regular one
occurs because of a term in the energy eigenvalues roughly
proportional to 2/N, which becomes progressively less
important as N increases (see Sheridan et al. 2002 for
details). At this point, the constant governing the separation
of the spin components is , the spin-rotation parameter.
12 11 ..........
13 12 ..........
14 13 ..........
16 15 ..........
17 16 ..........
4. ANALYSIS
5P
r
The data for the X
state of CrO were analyzed using
an effective Hamiltonian in a case (a) basis, which consists
of five basic interactions:
^ rot þ H
^ so þ H
^ ss þ H
^ sr þ H
^ ld :
^ eff ¼ H
H
a
J þ1
6.5
7.5
8.5
9.5
7.5
8.5
9.5
10.5
8.5
9.5
10.5
11.5
9.5
10.5
11.5
12.5
10.5
11.5
12.5
13.5
11.5
12.5
13.5
14.5
12.5
13.5
14.5
15.5
14.5
15.5
16.5
17.5
15.5
16.5
17.5
18.5
J
5.5
6.5
7.5
8.5
6.5
7.5
8.5
9.5
7.5
8.5
9.5
10.5
8.5
9.5
10.5
11.5
9.5
10.5
11.5
12.5
10.5
11.5
12.5
13.5
11.5
12.5
13.5
14.5
13.5
14.5
15.5
16.5
14.5
15.5
16.5
17.5
obs
obs calc
294,381.337
297,596.375
301,753.358
300,509.832
332,982.177
335,238.997
338,587.895
337,733.576
371,150.475
372,810.113
375,573.833
374,983.139
409,057.722
410,328.221
412,655.303
412,249.292
446,798.240
447,805.158
449,798.189
449,525.296
484,426.450
485,248.703
486,980.640
486,806.095
521,974.902
522,664.049
524,188.137
524,087.936
596,906.399
597,422.959
598,641.015
598,642.991
634,310.973
634,770.366
635,874.043
635,911.493
0.008
0.011
0.010
0.009
0.025
0.006
0.005
0.024
0.008
0.010
0.008
0.008
0.001
0.001
0.001
0.009
0.030
0.011
0.019
<0.000
0.007
0.007
0.024
0.033
<0.000
0.009
0.013
0.011
0.010
0.008
0.062
0.068
0.019
0.015
0.007
0.025
In MHz.
ð1Þ
The first term deals with molecular frame rotation, the next
three with spin-orbit, spin-spin, and spin-rotation couplings, and the final term with lambda doubling. Centrifugal
distortion corrections are included in each interaction.
^ so , H
^ ss , and H
^ sr can be found else^ rot , H
Standard forms of H
where (Brown et al. 1979; Barnes et al. 1993). The spin-orbit
Hamiltonian also involves the coupling between this term
and the spin-spin interaction (Brown et al. 1981), characterized by the constant , i.e.,
3S 2 1
ð3Þ
2
^
:
ð2Þ
Hso ¼ Lz Sz Sz 5
The lambda-doubling Hamiltonian, which includes centrifugal distortion, takes on the form (Barnes et al. 1993)
2
1
2
^ ld ¼ 1 ðo þ p þ qÞ Sþ
2 ðp þ 2qÞðJþ Sþ þ J S Þ
H
þ S
2
2
2 x 2
þ 12 q Jþ2 þ J2 þ 12 ðo þ p þ qÞD Sþ
þ S
R
2
2
1
1
x
2 ðp þ 2qÞD ðJþSþ þ J S Þ R þ 2 qD Jþ þ J2 x R2 :
ð3Þ
Three constants (o, p, and q) are thus required to describe
lambda doubling in 5P states.
For CrN, the Hamiltonian used to analyze the data consists of rotational, spin-rotation, and spin-spin terms (Nelis,
Brown, & Evenson 1990):
^ rot þ H
^ ss þ H
^ sr :
^ eff ¼ H
H
ð4Þ
Included in the spin-rotation coupling is the third-order correction to this interaction, which in a case (b) basis is best
expressed in tensor form:
10
ð3Þ
^
ð5Þ
Hsr ¼ pffiffiffi s T3 L2 ; N x T3 ðS; S; SÞ :
6
This term is thought to be necessary to describe the spinrotation coupling in states with quartet multiplicity or
higher (Hougen 1962).
Using these respective Hamiltonians, the two data sets
were analyzed using a least-squares fitting routine. The
spectroscopic parameters determined from these fits are
given in Tables 3 (CrO) and 4 (CrN). As the tables show, all
1112
SHERIDAN, BREWSTER, & ZIURYS
Vol. 576
Fig. 3.—Laboratory spectrum of the N ¼ 13
12 rotational transition
of CrN (X 4) near 484–487 GHz. There is a frequency gap in the data of
1.4 GHz. Here all four fine-structure components, indicated by quantum
number J, are visible. (Their relative intensities are not all the same only
because of variations in production efficiency.) The line corresponding to
¼ 3=2 (J ¼ 14:5
13:5) lies lower in frequency relative to the ¼ 12 feature (J ¼ 13:5
12:5). The spectrum is a composite of 16 scans, each 100
MHz in frequency width, and recorded in 1 minute.
constants used in the analysis are well determined, and the
rms values of the fits are 39 (CrO) and 23 kHz (CrN).
In the case of CrO, the spectroscopic constants established from optical LIF data are included in Table 3 for
comparison. These parameters were taken from Barnes et
al. (1993), who studied the A0 5 D X 5 systems using LIF
and combined their result with Fourier transform infrared
emission spectra. As shown, the millimeter-wave constants
established in this work are in good agreement with those of
Barnes et al. (1993) but improve the accuracy of the rotational, lambda-doubling, and some of the fine-structure values. There are some differences as well. For example, in the
current study, the centrifugal distortion terms qD and D
were not found to improve the overall rms of the fit and were
therefore not included in the final analysis. On the other
hand, the higher order spin-spin interaction term, , was
found to be necessary for a good fit. It should also be noted
that the constants and AD are highly correlated with each
other, and consequently it is difficult to establish their values
accurately in the analysis (Brown et al. 1979). Finally, rotational transitions with J > 10 originating in the ¼ 3 spinorbit components were not included in the least-squares fit.
Lambda doubling was not resolved in this component, and
at high J, where this effect was substantial in other sublevels, large residuals were generated (500 kHz) for these
features. Despite the exclusion of these transitions, 84
individual lines were included in the final analysis.
For CrN, the millimeter-wave constants determined in
this study are in agreement with the LIF/PPMODR values,
which are given in Table 4 (Namiki & Steimle 1999). There
appears to be a factor of 2 difference and a sign change in
the value of the spin-rotation interaction parameter s
between the two fits, however. On the other hand, the analysis done in this work employs three spin-rotation parameters (, D, and s), while the previous study only used and
s. Moreover, the data set from the previous work is not
Fig. 4.—Stick spectra of three separate rotational transitions of CrN,
illustrating the progression of the fine-structure splittings. In the top panel,
the spin components of the N ¼ 9
8 rotational transition have a total
separation of over 6 GHz, with the ¼ 3=2 line lying lower in frequency
11 transition, the fine-structhan its ¼ 12 counterpart. In the N ¼ 12
ture lines are distributed over only 3 GHz and the ¼ 3=2 and ¼ 12 spin
components lie closer in frequency. Finally, in the N ¼ 17
16 transition,
the fine-structure quartet has collapsed to a total separation of 1.6 GHz.
Now the ¼ 12 component is lower in frequency than the ¼ 3=2 line.
nearly as extensive. Some variations are therefore expected.
In addition, the matrix elements for the third-order spinrotation term have some subtle differences, depending on
the basis set, as discussed by Adam et al. (1994). Here those
of Nelis et al. (1990) were used, in a case (b) basis. Namiki &
Steimle (1999) modified those of Adam et al. (1994), which
utilized a case (a) basis set. (The theory and interpretation
behind the s term are not well developed because it is
required only for high spin states, of which there are few
examples studied at high resolution.)
5. DISCUSSION
The primary result of this investigation is that very accurate rotational rest frequencies have been directly measured
for CrO and CrN in their ground electronic states in the submillimeter region. (The complicated ground states exhibited
by these two radicals make direct measurements desirable.)
The spectroscopic constants of both these radicals have
additionally been refined. Moreover, this work is the first
time a molecule in a 5P ground state has been studied by
No. 2, 2002
ROTATIONAL REST FREQUENCIES
TABLE 3
Rotational Constants Determined for CrO (X 5Pr)
Parameter
Millimeter Wavea
B ....................
D....................
....................
AD ..................
A....................
ðo þ p þ qÞ .....
ðp þ 2qÞ..........
q.....................
ðo þ p þ qÞD ...
ðp þ 2qÞD .......
qD...................
....................
D ..................
.....................
D...................
h.....................
rms.................
15722.0298(15)
0.0218464(46)
315(213)
4.4(3.6)
1894999(408)
998.3(8.8)
206.55(46)
1.03(13)
0.071(21)
0.00254(17)
...
34407(361)
0.052(13)
2302.0(9.4)
...
4.67(94)
0.039
Opticalb
15722.0(3)
0.0219(1)
363(12)
3.60(24)
1895690(20)
959(6)
209(1)
1.3(1)
0.093(12)
0.0024(12)
0.00011(8)
34683(15)
0.066(12)
2240(0)
0.27(4)
...
42
a
In MHz; errors are 3 and apply to last quoted digits.
From Barnes et al. 1993; values originally quoted in
cm1; errors are 3 .
b
pure rotational spectroscopy. Thus, it serves as a test of
angular momentum coupling theory.
Both CrN and CrO exhibit high spin states, which arise
from unusual electron configurations. The primary valence
electron configuration of the ground state of CrN is
82349112 (Balfour et al. 1997) and of the ground state of
CrO is 8234911241 (Barnes et al. 1993). In these configurations, the 8 and 3 are bonding molecular orbitals
formed primarily from the 3d and 3d atomic orbitals on
chromium and the 2p and 2p atomic orbitals from oxygen
or nitrogen, with the 4 and 10 as the corresponding
antibonding ones. The 9 and 1 molecular orbitals arise
predominately from the 4s and 3d atomic orbitals of chromium, respectively, and they both are almost completely
nonbonding.
If the 4 orbital is actually antibonding, addition of an
electron to this site should increase the bond length of CrO
relative to CrN. Indeed, this increase does occur. From the
millimeter-wave data, the r0 bond lengths of the CrN and
CrO radicals were determined to be 1.5652 and 1.6213 Å,
respectively, resulting in a bond lengthening of 0.0561 Å for
the oxide compound. Such an increase appears to generally
occur for 3d transition metal oxides relative to the nitrides,
independent of the filling of orbitals. For example, VN has a
1113
TABLE 4
Rotational Constants Determined for CrN (X 4)
Parameter
Millimeter Wavea
PPMODR/Opticala, b
B ................
D................
................
D ..............
s ...............
................
D ..............
rms.............
18702.9055(14)
0.0318543(32)
209.261(47)
0.000291(64)
0.1328(85)
78281.97(58)
0.21744(56)
0.023
18702.952(33)
0.01558(75)
209.523(55)
...
0.225(22)
78281.32(21)
...
...
a
In MHz; errors are 3 .
From Namiki & Steimle 1999; merged fit of PPMODR and
LIF data, excluding hyperfine terms.
b
bond length of 1.566 Å (Balfour et al. 1993), while that of
VO is 1.592 Å (Merer 1989). A very similar difference is also
seen in FeN versus FeO (P. M. Sheridan, T. Hirano, & L.
M. Ziurys 2002, in preparation; Allen, Ziurys, & Brown
1996) and TiN versus TiO (Namiki et al. 1998). This general
result most likely occurs because the 2p atomic orbitals of
nitrogen are nearly 24,000 cm1 higher in energy than those
of oxygen. Thus, there is a smaller energy gap between the
nitrogen orbitals and the 3d atomic orbitals of the metal,
which enables a stronger bond to form in the nitride as
compared to the corresponding oxide.
Chromium oxide is most likely to be found in O-rich circumstellar envelopes such as OH 231.8. Searching for rotational lines in the ¼ 1 component, which lies lowest in
energy and exhibits small lambda doubling, is likely the best
approach. CrN, on the other hand, may be detectable in
shells of carbon-rich stars, including IRC +10216 or CRL
2688. Metal cyanide species have already been observed in
these objects, and a nitride compound may also be feasible,
given the presence of SiN in IRC +10216 (Turner 1992).
Observation of the four spin components within a single
rotational transition should be sufficient evidence for identification of this radical, given the large frequency separation
of these lines. Both CrO and CrN also have relatively large
dipole moments as well (3.88 D, Steimle et al. 1989; 2.31 D,
Steimle et al. 1999), which furthers the possibility of detection in interstellar/circumstellar gas.
This research was supported by NSF grant AST-9820576 and NASA grant NAG 5-1033. The authors wish to
thank John Brown for use of his Hamiltonian code.
REFERENCES
Brown, J. M., Milton, D. J., Watson, J. K. G., Zare, R. N., Albritton,
Adam, A. G., Azuma, Y., Barry, J. A., Merer, A. J., Sassenberg, U.,
D. L., Horani, M., & Rostas, J. 1981, J. Mol. Spectrosc., 90, 139
Schröder, J. O., Cheval, G., & Féménias, J. L. 1994, J. Chem. Phys., 100,
Cernicharo, J., & Guélin, M. 1987, A&A, 183, L10
6240
Cheung, A. S.-C., Zyrnicki, W., & Merer, A. J. 1984, J. Mol. Spectrosc.,
Allen, M. D., Ziurys, L. M., & Brown, J. M. 1996, Chem. Phys. Lett., 257,
104, 315
130
Davis, D. N. 1947, ApJ, 106, 28
Andrews, L., Bare, W. D., & Chertihin, G. V. 1997, J. Phys. Chem., 101,
Harrison, J. F. 1996, J. Phys. Chem., 100, 3513
8417
Highberger, J. L., Savage, C. S., Bieging, J. H., & Ziurys, L. M. 2001, ApJ,
Balfour, W. J., Merer, A. J., Niki, H., Simard, B., & Hackett, P. A. 1993, J.
562, 790
Chem. Phys., 99, 3288
Hocking, W. H., Merer, A. J., Milton, D. J., Jones, W. E., & KrishnaBalfour, W. J., Qian, C. X. W., & Zhou, C. 1997, J. Chem. Phys., 106,
murty, G. 1980, Canadian J. Phys., 58, 516
4383
Hougen, J. T. 1962, Canadian J. Phys., 40, 598
Barnes, M., Hajigeorgiou, P. G., & Merer, A. J. 1993, J. Mol. Spectrosc.,
Merer, A. J. 1989, Annu. Rev. Phys. Chem., 40, 407
160, 289
Namiki, K. C., Saito, S., Robinson, J. S., & Steimle, T. C. 1998, J. Mol.
Bauschlicher, C. W., Jr., Nelin, C. J., & Bagus, P. S. 1985, J. Chem. Phys.,
Spectrosc., 191, 176
82, 3265
Namiki, K. C., & Steimle, T. C. 1999, J. Chem. Phys., 111, 6385
Brown, J. M., Colbourn, E. A., Watson, J. K. G., & Wayne, F. D. 1979, J.
Nelis, T., Brown, J. M., & Evenson, K. M. 1990, J. Chem. Phys., 92, 4067
Mol. Spectrosc., 74, 294
1114
SHERIDAN, BREWSTER, & ZIURYS
Robinson, J. S., Apponi, A. J., & Ziurys, L. M. 1997, Chem. Phys. Lett.,
278, 1
Savage, B. D., & Sembach, K. R. 1996, ARA&A, 34, 279
Sheridan, P. M., Xin, J., Ziurys, L. M., Beaton, S. A., Kermode, S. M., &
Brown, J. M. 2002, J. Chem. Phys., 116, 5544
Steimle, T. C., Nachman, D. F., Shirley, J. E., Bauschlicher, C. W., Jr., &
Langhoff, S. R. 1989, J. Chem. Phys., 91, 2049
Steimle, T. C., Robinson, J. S., & Goodridge, D. 1999, J. Chem. Phys., 110,
881
Turner, B. E. 1992, ApJ, 388, L35
Woosley, S. E., & Weaver, T. A. 1995, ApJS, 101, 181
Xin, J., & Ziurys, L. M. 1999, J. Chem. Phys., 110, 4797
Ziurys, L. M., Barclay, W. L., Jr., Anderson, M. A., Fletcher, D. A., &
Lamb, J. W. 1994, Rev. Sci. Instrum., 65, 1517
Ziurys, L. M., Savage, C., Highberger, J. L., Apponi, A. J., Guélin, M., &
Cernicharo, J. 2002, ApJ, 564, L45