Journal of Molecular Spectroscopy 206, 2732 (2001)

doi:10.1006/jmsp.2000.8285, available online at http://www.idealibrary.com on

Fourier Transform Emission Spectroscopy of CuCl
T. Parekunnel,

L. C. OBrien, T. L. Kellerman, T. Hirao,

M. Elhanine, and P. F. Bernath

,

Department of Chemistry, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1; Department of Chemistry, Southern Illinois University,
Edwardsville, Illinois 62026-1652; Laboratoire de Photophysique Mol eculaire du CNRS, B atiment 210, Universit e Paris Sud, 91405 Orsay,
France; and Department of Chemistry, University of Arizona, Tucson, Arizona 85721
Received July 26, 2000; in revised form November 28, 2000
The electronic spectra of CuCl were observed in the 18 000 cm
1
to 25 000 cm
1
spectral region using a Bruker IFS
120 HR Fourier transform spectrometer (FTS) and with the FTS associated with the McMath-Pierce Solar Telescope at Kitt
Peak. On the basis of ab initio calculations, the labels for the electronic states were revised, and the a
3

+
1
X
1

+
00 band,
the b
3

0
X
1

+
00, 10, and 01 bands, the b
3

1
X
1

+
00, 10, and 01 bands, the A
1
X
1

+
00, 10, and 02 bands,
and the B
1

+
X
1

+
00 and 10 bands were measured. Improved spectroscopic constants were obtained for the excited and
ground states. C 2001 Academic Press
INTRODUCTION
Asimple model for the electronic states of CuCl is to consider
the molecule as a Cu
+
(3d
10
) cation and a Cl

(3p
6
) anion,
and the molecular orbitals in terms of the atomic orbitals of
the individual ions. Using this model, the 3p orbitals on the
chlorine form a closed subshell and the lower energy electronic
transitions can be attributed to an electron being promoted from
one of the lled 3d orbitals to the empty 4s orbital on the copper
ion. When this happens, the electron leaves behind a d-hole
which results in three possible states: , , and . The state
depends on which orbital the electron was promoted from, d
z
2 ,
d
xz or yz
, or d
xy or x
2
y
2 for , , or , respectively. Each of the
resultant states can also be either a singlet or a triplet. Thus,
CuCl has a X
1

+
ground state and six low-lying excited states,
1

+
,
3

+
,
1
,
3
,
1
, and
3
.
The rst analyses of CuCl in 1927 (1) and 1938 (2) were vi-
brational analyses that located ve low-lying electronic states.
The ve band systems were the result of transitions with a com-
mon lower state. Since then, rotational analyses of the elec-
tronic emission bands have been performed (37), as well as
work in the millimeter and microwave regions of the spectrum
(8, 9). Work performed by Rao et al. (4, 5), as well as contri-
butions by Lagerqvist and Lazarava-Girsamoff (10) and Ahmed
and Barrow (11), led to the characterization of the six excited
electronic states, originally labeled as A
1
, B
1
, C
1

+
, D
1
,
E
1

+
, and F
1
states. However, the rst reference to a triplet
state was not made until 1984, when Balfour and Ramattributed
a 13 500 cm
1
band to a
3

1
transition by analogy with a sim-
ilar transition in CuF (12, 13). It now appears that this transition
Supplementary data for this article are available on IDEAL (http://www.
indealibrary.com) and as part of the Ohio State University Molecular Spec-
troscopy Archives (http://msa.lib.ohio-state.edu/jmsa hp.htm).
is between two highly excited electronic states (14). Ab initio
calculations were performed with new assignments for the elec-
tronic systems that included triplet states (15, 16). Delaval et al.
measured the radiative lifetimes of the low-lying electronic ex-
cited states and also suggested that some of the singlet states of
CuCl were in fact triplet states (17).
Recent work on CuCl includes microwave experiments in
the ground and rst excited vibrational levels (18, 19), laser
excitation spectroscopy of the 00 bands of the D
1
X
1

+
and E
1

+
X
1

+
transitions (20, 21), and ab initio calculations
(2225).
EXPERIMENTAL DETAILS
To obtain an electronic emission spectrum of CuCl, a copper
hollowcathode lamp was used. The lamp consisted of a stainless
steel chamber as the anode and a copper hollow cathode with an
inner diameter of 1 cm.
The initial experiment was run with a small sample of Cu
metal inside the lamp, which was operated with HCl and He
gas. Emission from the CuCl from inside the cathode was fo-
cussed with a lens into a Bruker IFS 120 HR Fourier transform
spectrometer, modied to obtain double-sided interferograms,
located at the University of Waterloo. To increase the signal-
to-noise ratio, the experiment was improved. CuCl powder and
Cu metal were placed inside the lamp, which was then operated
with only He gas at a pressure of 5 Torr. The lamp was operated
at a current of 200 to 250 mA.
The spectrum was recorded in two spectral regions at a res-
olution of 0.02 cm
1
to cover the total spectral range from
18 000 to 26 500 cm
1
with a photomultiplier tube (PMT) detec-
tor and a visible quartz beamsplitter. Bandpass lters were in-
serted to reduce the effect of strong atomic lines on the spectrum.
27
0022-2852/ 01 $35.00
Copyright
C
2001 by Academic Press
All right s of reproduct ion in any form reserved.
28 PAREKUNNEL ET AL.
In the rst spectrum, spanning the region from 18 000 to 23 000
cm
1
, a 550-nmblue pass lter and a 450-nmred pass lter were
used. The second spectrum, spanning the region from 20 000 to
26 500 cm
1
, was recorded using a 450-nm blue pass lter and
a 400-nm red pass lter.
The excited CuCl molecules were also produced in a King-
type carbon tube furnace, charged with approximately 10 g of
CuCl powder. The tube was lled with 50 Torr of helium, then
heated to 1900

C, with a nal pressure of 400 torr. The CuCl

emission was focused onto the entrance aperture of the Fourier
transform spectrometer (FTS), located at the McMathPierce
Solar Observatory, Kitt Peak, Arizona. Twenty scans at a res-
olution of 0.021 cm
1
were co-added in 65 minutes of inte-
gration in 3rd-order alias. The spectral region from 22 200 to
33 000 cm
1
was recorded by the FTS, which was congured
with a CuSO
4
lter, a quartz beamsplitter, and two midrange
diode detectors. The carbon furnace gave spectra with higher
rotational excitation than the hollow cathode lamp.
CuCl SPECTRAL ANALYSIS
The CuCl spectra obtained from the Waterloo experiments
are shown in Figs. 1 and 2. Rovibronic line positions were mea-
sured for the a
3
Y

1
X
1
Y

00 band, the b
3
H
0
X
1
Y

00, 10,
and 01 bands, the b
3
H
1
X
1
Y

00, 10, and 01 bands, the

A
1
HX
1
Y

00, 10, and 02 bands, and the B

1
Y

X
1
Y

0
0, and 10 bands. Figure 3 shows the 00 band of the B
1
Y

X
1
Y

transition. For the more intense bands, strong unblended

lines were measured to an estimated precision of 0.005 cm
1
.
However, nearly all of the observed line positions were blended
FIG. 1. The emission spectrum of CuCl, spanning the spectral region from 18 500 to 22 500 cm
1
. a, b
0
, b
1
, A, and B indicate the a
3
Y

1
X
1
Y

, b
3
H
0
X
1
Y

,
b
3
H
1
X
1
Y

, A
1
HX
1
Y

, and B
1
Y

X
1
Y

transitions, respectively.
TABLE 1
The Low-Lying Electronic States of CuCl
and therefore the estimated precision varied from 0.02 to
0.04 cm
1
. This severe blending made the rotational analy-
sis difcult.
The electronic band systems were labeled based on the pres-
ence of a Q branch in the bands, as well as on comparisons
with previous assignments that are based on radiative lifetime
studies (26) and ab initio calculations (24). The selection rule
for O is LO = 0. 1. Those transitions that have a Q branch
have O = 1 in the excited state, while those without a Q branch
have O = 0

. The multiplicity and A values were then deter-

mined by the lifetime studies and the ab initio calculations.
The new labels for the low-lying electronic states are listed in
Table 1. The presence of triplet-to-singlet transitions, forbidden
by Hunds case (a) selection rules, suggests that there is consid-
erable mixing of the triplet and singlet states. For example, the
presence of a Q branch and substantial effective lambda dou-
bling in the a
3
Y

1
X
1
Y

system further suggests that mixing

occurs between the a
3
Y

1
and A
1
H states.
Copyright
C _
2001 by Academic Press
EMISSION SPECTROSCOPY OF CuCl 29
FIG. 2. The emission spectrum of CuCl, spanning the spectral region from 22 000 to 24 500 cm
1
. A and B indicate the A
1
HX
1
Y

and B
1
Y

X
1
Y

transitions, respectively.
The 1640 measured line positions for
63
Cu
35
Cl are provided
in the JMS depository. The signal-to-noise ratio for the strongest
lines was better than 25. The line positions were measured using
the PC Decomp program and then arranged into series of P, Q,
and R branches for each individual band using a color Loomis
Wood program.
FIG. 3. The 00 band of the B
1
Y

X
1
Y

transition.
A spectrum containing Ar and He lines was obtained for the
calibration of the CuCl lines. The He line positions were rst
calibrated using the Ar line positions from Ref. (27). The He
lines present in both the CuCl spectrum and the Ar and He
spectrum were then used to obtain a calibration factor for the
calibration of the CuCl line positions.
Copyright
C _
2001 by Academic Press
30 PAREKUNNEL ET AL.
TABLE 2
Band Constants for
63
Cu
35
Cl (in cm
1
)
a
xed.
b
q
Hv
= 4.18(22) 10
13
cm
1
.
c
q
Hv
= 4.18 10
13
cm
1
, held constant.
The data were then tted to the energy level expression
E
v. j
= T
v

B
v

q
v
2

J(J 1)

D
v

q
D
v
2

[J(J 1)]
2

H
v

q
H
v
2

[J(J 1)]
3
. [1]
where T
v
includes the electronic and vibrational energy, B
v
,
D
v
, and H
v
are the usual band constants, and q
v
and q
Dv
are
the lambda-doubling constants. Initial ts for individual bands
were performed by xing the ground electronic state constants
to literature values obtained from pure rotational transitions (8,
18, 19). Anal t was performed with all the data, including the
microwave data from the literature, and the excited and ground
state constants were allowed to oat. The constants obtained are
listed in Table 2.
The equilibrium molecular constants, listed in Table 3, were
determined using the expression for the energy of a transition
between vibrational levels characterized by : 1 and : of an
TABLE 3
Molecular Equilibrium Constants for
63
Cu
35
Cl (in cm
1
)
a

e
= 1.952(77) 10
6
cm
1
.
b
B
0
.
c
LG
1,2
.
d
r
0
.
anharmonic diatomic oscillator,
LG
:1,2
= G(: 1) G(:) =
e
2
e
x
e
2
e
x
e
: .
[2]
and the parametric expression describing the vibrational depen-
dence of the rotational constants,
B
:
= B
e

v
1
2

v
1
2

2
. [3]
For the b
3
H
0
, b
3
H
1
, A
1
H, and B
1
Y

states, LG
1,2
values are
shown in place of
e
, as only two vibrational levels were mea-
sured. Note that the band with L: ,= 0 are much more complex
than the 00 bands and there is the possibility of error in the
numbering of the P and R branches.
A comparison of the T
v
value obtained for the B
1
Y

X
1
Y

00 transitions with that fromRef. (21) shows agreement within

0.004 cm
1
, which is less than the linewidth obtained for our
Copyright
C _
2001 by Academic Press
EMISSION SPECTROSCOPY OF CuCl 31
FIG. 4. Energy level diagramof the low-lying electronic states of CuCl. The
left side shows the experimental values obtained in this work. The
3
L transition
energy was taken from Ref. (26). The right side shows recent theoretical results
(24).
CuCl spectra. The A
1
HX
1
Y

00 band of the
63
Cu
37
Cl iso-
topomer was also measured by the same group and T
0
was
determined for the dominant isotopomer,
63
Cu
35
Cl, by isotopic
scaling. The value obtained agrees with our determination of T
0
within 0.015 cm
1
.
Figure 4 shows the energy level diagram of the electronic
states of CuCl. The left side shows the T
v
values obtained during
this work. For comparison, the right side shows the most recent
theoretical values (24). While the theoretical results do not show
the spin-orbit splitting in the
3
H state, as observed in our exper-
iment, there is good agreement between the T
v
values within
500 cm
1
. The
3
Lstate shown in Fig. 4 was not measured in this
work. The value shownwas takenfromRef. (26) andincludedfor
completeness.
The effective A-doubling constant, q, for the a
3
Y

1
state was
determined with good precision; 4.059(29) 10
3
cm
1
, and
the previous positive value (6) is clearly erroneous. In the a
state, the A-doubling effect is mainly due to the spinspin
interaction in the triplet state, which is in fact a second order
spinorbit effect (28). By using Hunds case (a) basis functions
and ignoring centrifugal distortion terms, the rotational energy
levels for the a
3
Y

1
state are given by the approximate formulae
E(O = 1

) = BJ(J 1)
2
3
. [4]
E(O = 0

) = BJ(J 1) q J(J 1) 2B 2
4
3
. [5]
E(O = 1

) = E(O = 1

) q J(J 1). [6]

In this approximation, B, , and are the rotational, spin
rotation, and spinspin constants, respectively, and the spinspin
constant is assumed to be large ( 2BJ). The value of q is
then given by the formula
q =
( 2B)
2
2 2B
. [7]
FromEq. [7] (ignoring the spinrotation constant), the spinspin
constant is estimated to be 14 cm
1
. The a
3
Y

state, expected
in the region 28 cm
1
below the a
3
Y

1
state, was not observed.
The A-doubling splitting in a
1
Hstate is usually explained by
a simple pure precession approximation (28):
q
v
=

v
/
2[B
1
Y. v
/
[B(r)L

[ A
1
H. v)]
2
E
A.v
/ E
B.v

2l(l 1)B
2
v
E
A.v
E
B.v
. [8]
For l = 2, however, the calculated q
0
is 3.1 10
3
cm
1
,
which is three times larger than the observed value,
1.1191(26) 10
3
cm
1
.
A similar discrepancy was also seen in the b
3
H
1
state. In this
case, the splitting between e and f levels is represented by the
formula (29)
E
f
E
e
=

4B

( p

2q

)
A

J(J 1). [9]

where the constants denoted with asterisks indicate true
molecular constants for the
3
H state obtained with Hunds case
(a) basis functions and the N
2
Hamiltonian dened in Ref. (30).
Assuming that the spinorbit and spinspin coupling constants,
A

and

, are similar to those of CuF, 413 cm

1
and 18 cm
1
,
respectively(31), the A-doublingconstant is expectedtobe close
to q

in Eq. [9], and thus,

q
v
q

v
/
2[a
3
Y. v
/
[B(r)L

[b
3
H. v)]
2
E
b.v
E
a.v
/

v
/
2l(l 1)[a
3
Y. v
/
[B(r)[b
3
H. v)]
2
E
b.v
E
a.v
/
[10]
within the pure precession approximation. The value of q
0
(q

)
from this formula was calculated as 2.3 10
3
cm
1
, which
is three times smaller than the observed value of 7.008(17)
10
3
cm
1
. Moreover, higher order A-doubling terms were re-
quired in the t, which is unlikely for a heavy molecule in a
normal
1
H state. There is probably a similar effect in CuF, in
which the ef splitting is not proportional to J(J 1) (32, 33)
because the b
3
H
0
state is located only 55 cm
1
higher than the
b
3
H
1
state (34).
The simple-minded pure precession mixing is compromised
primarily by conguration mixing and spinorbit coupling. A
more sophisticated model that includes these effects has been
developed by Delaval et al. (35). Their predictions for the qs
Copyright
C _
2001 by Academic Press
32 PAREKUNNEL ET AL.
are 5.0 10
3
, 1.04 10
3
, and 2.1 10
3
cm
1
for the
a
3
Y

1
, b
3
H
1
, and A
1
H states, respectively, to be compared with
the observed values (Table 2) of 4.06 10
3
, 0.70 10
3
and 1.12 10
3
cm
1
. This model is obviously much bet-
ter than pure precession but still deviates somewhat from
experiment.
CONCLUSIONS
The emission spectra of CuCl have been investigated at high
resolution in the region from 18 000 to 25 000 cm
1
using a
Fourier transform spectrometer. The bands observed in this re-
gion have been assigned to a number of transitions involving
electronic states with recently revised labels. Rotational analysis
of these transitions has been carried out and improved spectro-
scopic constants have been obtained for the excited and ground
electronic states.
ACKNOWLEDGMENTS
This work was supported by the Natural Sciences and Engineering Research
Council of Canada. We also acknowledge the Petroleum Research Fund and the
Killam Foundation for partial support. LCO gratefully acknowledges support
from the National Science Foundation, NSF-CHE-9753254. ME acknowledges
support fromCNRS/GDRof Physico-Chimie des Grains et Molecules Interstel-
laires.
REFERENCES
1. R. Ritschl, Z. Phys. 42, 172210 (1927).
2. S. Bloomenthal, Phys. Rev. 54, 498497 (1938).
3. J. Terrien, Ann. Phys. NY 9, 477538 (1938).
4. P. R. Rao and J. K. Brody, J. Chem. Phys. 35, 776787 (1961).
5. P. R. Rao, R. K. Asundi, and J. K. Brody, Can. J. Phys. 40, 412422,
423430, and 14431456 (1962).
6. G. P. Mishra, S. B. Rai, and K. N. Upadhya, Can. J. Phys. 59, 289297
(1981).
7. P. M. R. Rao, Indian J. Pure Appl. Phys. 18, 200203 (1980).
8. E. L. Manson, F. C. De Lucia, and W. Gordy, J. Chem. Phys. 62, 10401043
(1975).
9. E. Tiemann and J. Hoeft, Z. Naturforsch. 32a, 14771479 (1977).
10. A. Lagerqvist and V. Lazarava-Girsamoff, Ark. Fys. 20, 543553 (1961).
11. F. Ahmed and R. F. Barrow, J. Phys. B Ato. Mol. Phys. 8, L362L363
(1975).
12. W. J. Balfour and R. S. Ram, J. Phys. B Ato. Mol. Phys. 17, L19L21
(1984).
13. W. J. Balfour and R. S. Ram, J. Phys. B Ato. Mol. Phys. 16, L163L166
(1983).
14. L. C. OBrien, H. Cao, and J. J. OBrien, J. Mol. Spectrosc. 199, 100108
(2000).
15. N. W. Winter and D. L. Huestis, Chem. Phys. Lett. 133, 311316 (1987).
16. M. T. Nguyen, M. A. McGinn, and N. J. Fitzpatrick, J. Chem. Soc. Faraday
Trans. 2 82, 14271443 (1986).
17. J. M. Delaval, Y. Lefebvre, H. Bocquet, P. Bernage, and P. Niay, Chem.
Phys. 111, 129136 (1987).
18. R. J. Low, T. D. Varberg, J. P. Connelly, A. R. Auty, B. J. Howard, and
J. M. Brown, J. Mol. Spectrosc. 161, 499510 (1993).
19. K. D. Hensel, C. Styger, W. Jager, A. J. Merer, and M. C. Gerry, J. Chem.
Phys. 99, 33203328 (1993).
20. A. Bath and E. Tiemann, Chem. Phys. Lett. 175, 8486 (1990).
21. I. Burghardt, L. R. Zink, D. A. Fletcher, J. M. Brown, and I. R. Beattie,
Mol. Phys. 67, 14011417 (1989).
22. A. Ramirez-Solis andJ. P. Daudey, J. Phys. BAto. Mol. Phys. 23, 22772291
(1990).
23. A. Ramirez-Solis, Phys. Rev. A. 47, 15101513 (1993).
24. C. Sousa, W. A. De Jong, R. Broer, and W. C. Nieuwpoort, J. Chem. Phys.
106, 71627169 (1997).
25. C. Sousa, W. A. De Jong, R. Broer, and W. C. Nieuwpoort, Mol. Phys. 92,
677686 (1997).
26. J. M. Delaval, J. Schamps, A. Ramirez-Solis, and J. P. Daudey, J. Chem.
Phys. 97, 65886592 (1992).
27. G. Norl en, Phys. Scripta, 8, 249268 (1973).
28. H. Lefebvre-Brion and R. W. Field, Perturbations in the Spectra of Di-
atomic Molecules. Academic Press, Orlando, FL., 1986.
29. J. M. Brown and A. J. Merer, J. Mol. Spectrosc. 74, 488494 (1979).
30. J. M. Brown, E. A. Colbourn, J. K. G. Watson, and F. D. Wayne, J. Mol.
Spectrosc. 74, 294318 (1979).
31. C. R. Brazier, J. M. Brown, andT. C. Steimle, J. Mol. Spectrosc. 97, 449453
(1983).
32. F. Ahmed, R. F. Barrow, A. H. Chojnicki, C. Dufour, and J. Schamps,
J. Phys. B 15, 38013818 (1982).
33. T. C. Steimle, C. R. Brazier, and J. M. Brown, J. Mol. Spectrosc. 110, 3952
(1985).
34. J. M. Delaval, J. Schamps, and C. Dufour, J. Mol. Spectrosc. 137, 268277
(1989).
35. J. M. Delaval, J. Schamps, and F. L. Sefyani, Phys. Scrip. 49, 404407
(1994).
Copyright
C _
2001 by Academic Press