141102-3
Millimeter-wave optical polarization spectroscopy
J. Chem. Phys. 123, 141102 ͑2005͒
=1͓Q͑1͒,R͑1͔͒ and J =2͓P͑2͒,Q͑2͒,R͑2͔͒, Fig. 2͑a͒ shows
Љ
a negative-going transition at the expected position of the
R͑0͒ line. The sign of the extra signal indicates that the op-
tical transition out of J =0 causes a preferential depopula-
Љ
tion of the J =1 level relative to the J =2 level, which is
Љ
Љ
most likely due to collisional ͑J =0 hole filling from J =1͒
Љ
Љ
population transfer following optical excitation. The optical
pump can significantly alter the thermal population differ-
ence between the J =2 and J =1 levels because the CS
Љ
Љ
A 1⌸−X 1⌺+ ͑1-0͒ transition is fully allowed and the Franck-
Condon factor is 0.14.13 As a consequence, both negative and
positive mmODR signals are observed that are considerably
larger than the original absorption signal in the absence of
optical pumping.
The sensitivity of the mmODR experiments is limited
primarily by the large and fluctuating CS absorption back-
ground caused by instabilities in the CS number density pro-
duced in the pulsed-discharge supersonic expansion. This CS
production noise can be largely removed by using the
mmOPS technique. When the vertically polarized pump laser
is resonant with a CS electronic transition, it creates an an-
isotropic sample of CS molecules by preferentially pumping
certain MJ states. If one of the levels of the millimeter-wave
transition is in common with the pump-laser transition, then
the vertical and horizontal components of the millimeter-
wave probe are differentially absorbed, causing a polariza-
tion rotation of the millimeter waves. By using a set of
crossed polarizers, the unrotated millimeter-wave radiation
can be blocked, while only the rotated millimeter-wave ra-
diation strikes the detector. Thus, the polarization-detected
experiment can have a background level that is limited only
by the imperfect extinction of the polarizers and the birefrin-
gence of the millimeter-wave transmission optics. When the
polarizers are perfectly crossed, all mmOPS signals should
have the same sign because millimeter-wave radiation strikes
the detector only when the optical pump transition originates
from one of the linked levels of the millimeter-wave transi-
tion. The signal levels in this case are often too small to be
observed with the noise level of the current detector. As in
optical polarization-based experiments,17 however, the
signal-to-background ratio can be improved by imperfectly
crossing the polarizers to heterodyne the weak polarization-
rotation signal field with the much stronger nonrotated field.
Here, the spectra are collected at relatively large uncrossing
angles ͑ϳ5°͒ to obtain the best signal-to-background ratio.
Although the background level at 5° is approximately twice
the background level when the polarizers are perfectly
crossed, the cross term between the signal and carrier
millimeter-wave electric fields is much larger, which ulti-
mately increases the signal-to-background ratio. The hetero-
dyne term leads to both positive and negative features in the
mmOPS spectrum. In contrast to the mmODR spectra, how-
ever, the sign and intensity of the observed features contain
information about the rotational branch excited by the optical
field. A detailed theoretical analysis ͑in the nonperturbative
limit͒ of the expected intensities is currently in progress.18
The spectra in Fig. 2 have been normalized so that the
strongest transition in each spectrum has the same peak in-
tensity. The baseline in the region of 39 829 cm−1 is ex-
FIG. 3. mmOPS spectra of nominally spin-forbidden electronic transitions
to triplet states of CS, which borrow their intensity via spin-orbit interac-
tions with the A 1⌸ state. The millimeter-wave frequency is locked to the CS
X 1⌺+
͑
=0,J =2-1͒ rotational transition and the laser is scanned over the
v
Љ Љ
͑a͒ e 3⌺− −X 1⌺+ ͑2-0͒ and the ͑b͒ d 3⌬−X 1⌺+ ͑6-0͒ bands. The e 3⌺−
͑
v
Ј
=2͒ and the d 3⌬ ͑ =6͒ states have 17% and 1% nominal A 1⌸ character,
v
Ј
respectively, for the rotational levels accessed ͑Ref. 13͒. Each point is an
average of 20 laser shots.
panded to show that baseline noise is suppressed by a factor
of 4 for mmOPS relative to mmODR. The enhancement in
sensitivity ͑signal-to-background ratio͒ afforded by the
mmOPS technique has enabled us to record nominally spin-
forbidden optical transitions into triplet states of CS, as
shown in Fig. 3. Although the sensitivity enhancement of
mmOPS is largely due to the reduction of the CS production
noise present in the CS ground-state mmODR signal, it is
also improved because the millimeter-wave power incident
to the molecular beam can be increased without saturating
the millimeter-wave detector. In the absorption-detected
technique the power of the Gunn oscillator is attenuated by
approximately 15 dB to 600 W to avoid saturation of the
detector/preamplifier, while in the polarization-detected tech-
nique the full power of the oscillator ͑nominally 18 mW at
98 GHz͒ is incident to the molecular beam, but is attenuated
by the crossed polarizers to Ͻ1 mW on the detector.
To demonstrate the additional capabilities of the mmOPS
technique, we have measured a pure rotational transition in
one of the optically populated triplet states. Figure 4 shows
the mmOPS spectrum recorded with the laser populating the
J =1, N =1 level of the e 3⌺− ͑ =2͒ state. The trace is the
v
Ј
Ј
Ј
average of four scans, each recorded by stepping the Gunn
oscillator output in 50 kHz steps and averaging the bolom-
eter output for ten laser shots at each frequency. The center
frequency of the triplet rotational transition J =2, N =2
Ј
Ј
←J =1, N =1 is determined to be 76 229.027͑20͒ MHz.
Ј
Ј
The measured linewidth is 1.3 MHz, which is significantly
broader than the ground-state linewidth ͓ϳ240 kHz full
width at half maximum ͑FWHM͔͒. We attribute the addi-
tional linewidth to the radiative lifetime of the triplet state,
which we expect to be ϳ1.2 s, assuming that the lifetime is
dominated by the ϳ17% singlet character13 of the populated
“triplet” level and the 198 ns lifetime for the A 1⌸ ͑v=1͒
state.19
130.63.180.147 On: Mon, 24 Nov 2014 02:20:31