Faraday filtering of Raman light in an isotopically pure alkali metal vapor

Abstract

We present an application of the Faraday effect to produce a narrow band atomic filter in an alkali metal vapor. In our experiment two Raman beams separated in frequency by the ground state hyperfine splitting in 87Rb are produced using an EOM and then filtered using the Faraday effect in an isotopically pure 85Rb thermal vapor. An experimental transmission spectra for the filter is presented along with a theoretical calculation. The performance of the filter is then demonstrated and characterized using a Fabry-Perot etalon. For a temperature of 70 degrees and a longitudinal magnetic field of 80 G a suppression to -18 dB is achieved, limited by the quality of the polarizers. Comment: 3 pages, 3 figures

Full text

Page 1

arXiv:0904.4142v1 [physics.atom-ph] 27 Apr 2009
Faraday filtering of Raman light in an isotopically pure alkali metal vapor
Richard P. Abel∗, Ulrich Raitzsch, Paul Siddons, Ifan G. Hughes and Charles S. Adams
Department of Physics, Durham University,
Rochester Building, South Road,
Durham DH1 3LE, United Kingdom
∗Corresponding author: r.p.abel@dur.ac.uk
We present an application of the Faraday effect to produce a narrow band atomic filter in an
alkali metal vapor. In our experiment two Raman beams separated in frequency by the ground
state hyperfine splitting in87Rb are produced using an EOM and then filtered using the Faraday
effect in an isotopically pure85Rb thermal vapor. An experimental transmission spectra for the
filter is presented along with a theoretical calculation. The performance of the filter is then demon-
strated and characterized using a Fabry-Perot etalon. For a temperature of 70◦C and a longitudinal
magnetic field of 80 G a suppression to -18 dB is achieved, limited by the quality of the polarizers.
In many atomic physics experiments one requires phase
coherent laser light at frequencies separated by the
ground state hyperfine splitting; examples include stim-
ulated Raman transitions [1]; coherent population trap-
ping [2]; Λ-system and N-system electromagnetically in-
duced transparency (EIT) [3, 4] and mesoscopic Rydberg
gates [5]. Alternatively, Raman light is also produced in
experiments involving Raman scattering processes [6]. In
many cases the two components of the Raman light are
of unequal intensity and are not spatially separated. We
therefore require a filter that will separate the two fre-
quencies into separate beams, producing two sources of
light suitable for subsequent applications.
A narrow band atomic filter can be realized using the
Faraday effect where a longitudinal magnetic field in-
duces a circular birefringence in the medium [7]. Atomic
filters exploiting birefringence were first introduced and
demonstrated by¨Ohman [8]. This principle was devel-
oped into the Faraday anomalous-dispersion optical fil-
ter (FADOF), which has been demonstrated in Cs [9], Rb
[10, 11, 12] and Na [13]. Similarly, the induced-dichroism
excited atomic line (IDEAL) filter, which operates with-
out a magnetic field, has been demonstrated in K [14].
More recently, atomic filters have been produced using
absorption in a thermal vapor cell [6] and velocity selec-
tive optical pumping in an atomic vapor [15]. Narrow
band atomic filters have applications in free space laser
communications [11], atmospheric measurements using
LIDAR [13, 16], ocean temperature profiling using LI-
DAR [12] and the generation of narrow band quantum-
noise-limited light [9]. In our experiment we use an iso-
topically pure cell such that we can exploit the Faraday
effect in85Rb to filter Raman light resonant with87Rb.
The Faraday effect is observed when a magnetic field
is applied parallel to the direction of light propagation,
causing initially linearly polarized light to be rotated by
an angle, θ, given by:
θ = V BL, (1)
where V is the Verdet constant, B the magnitude of the
applied magnetic field and L the length of the medium.
The Verdet constant is dependent on the properties of
the medium, the wavelength of the light and the temper-
ature. Typical commercial Faraday isolators employ a
Terbium Gallium garnet crystal with a Verdet constant
of 134 Rad T−1m−1at 632 nm [17].
The polarization rotation is produced by an induced
circular birefringence in the medium [18] which produces
a relative phase shift, ∆φ, between circular field compo-
nents given by:
∆φ =ωL
c(n+− n−), (2)
where n+and n−are the refractive index of the circular
components driving σ+and σ−transitions, respectively.
FIG. 1: (a) Schematic experimental setup. An EOM is used to
create sidebands at ν±νmwith νm = 6.8 GHz producing light
corresponding to the Raman transitions in87Rb as shown in
(b). The components at ν and ν +νm are separated using the
Faraday effect in a heated85Rb vapor cell within a solenoid.
The input light is initially horizontally polarized using a po-
larizing beam splitting (PBS) cube. Following the solenoid
the different frequency components are separated using a sec-
ond PBS cube. The insets show schematics of Fabry-Perot
etalon signals.

Page 2

2
The relative phase shift has a spectral dependance which
can be expressed as:
d(∆φ)
dω
=L
c(n+
g− n−
g). (3)
The relative phase shift causes a rotation of the polar-
ization of initially linearly polarized light by an amount
θ = ∆φ/2 (see Methods section in [18]). By exploit-
ing the Faraday effect in an isotopically pure Rb ther-
mal vapor it is possible to design a system with a fre-
quency dependent polarization rotation. Light that is
resonant with the lower ground state hyperfine to ex-
cited state transition in one isotope, but is in the ab-
sorption wings of the filter isotope, will be transmitted
unchanged. Whereas light resonant with the upper hy-
perfine to excited state transition experiences a rotation
in polarization equal to π/2, and therefore can be sepa-
rated using a polarizing beam splitting (PBS) cube, thus
realizing a narrow band atomic filter for Raman light.
A schematic experimental setup of the Faraday fil-
ter is shown in Figure 1 (a). Light from a diode laser
(Toptica DLPRO) is frequency stabilized to the
5S1/2(F = 2) → 5P3/2(F′= 3) transition using mod-
ulation transfer spectroscopy [19] before passing through
an EOM (New Focus 4851 driven by a Aglient E8267D
signal generator) modulated at νm = 6.8 GHz (corre-
sponding to the ground state hyperfine-splitting in87Rb
as shown in Figure 1 (b)). The EOM generates sidebands
87Rb
?
sult of a theoretical calculation. The solid red vertical line
shows the frequency of the carrier and the dashed red line the
frequency of the sideband, ν + νm, generated by the EOM.
Zero detuning corresponds to the weighted center of the D2
line.
6
?
4
?
2
Detuning (GHz)
0
24
6
0.0
0.2
0.4
0.6
0.8
1.0
Tv
0.0
0.2
0.4
0.6
0.8
1.0
Th
FIG. 2: Transmission spectra at both output ports of the
Faraday filter. The upper panel corresponds to the transmit-
ted port of the filter with the bottom panel corresponding to
the reflected port. The temperature of the filter is approxi-
mately 70◦C and the magnetic field is 80 G. The blue line
represents the experimental data and the black line is the re-
at ±νmwith an amplitude of 1%. The beam is then lin-
early polarized by a PBS cube and propagated through a
heated isotoptically pure85Rb vapor cell (Triad Technol-
ogy, TT-Rb85-75-V-Q) within a solenoid 28 cm in length
with a field uniformity better than 1% over the length of
cell. The Rb vapor cell contains a mixture of Rb isotopes
in a ratio of85Rb :87Rb equal to 99.9:0.1. The current in
the solenoid produces both a homogenous magnetic field
along the length of the cell and Ohmic heating, with ad-
ditional temperature control provided by a Peltier effect
heater. Following the vapor cell the light is analysed us-
ing a second PBS cube, which generates the two output
ports of the filter.
The filter output as a function of the laser frequency
for the Rb D2 line is shown in Figure 2. The detuning
axis is scaled so that zero corresponds to the weighted
center of the D2line. The upper panel shows the trans-
mission of light with a horizontal polarization, Th, and
the lower panel light with a vertical polarization, Tv. The
experimental data are represented by the blue line and
the black line is a theoretical calculation, which is seen
to be in good agreement. The calculation is based on the
theoretical model presented in [20] and the polarization
rotation found using equation 2 from [18].
netic field dependence is calculated using a full matrix
diagonalization as in [12]. The solid red vertical line in-
dicates the position of light at the carrier frequency and
The mag-
?
4
?
2
0
2
4
68 10
Detuning (GHz)
?
20
?
15
?
10
?
5
0
(c)
?
20
?
15
?
10
?
5
0
Intensity (dB)
(b)
?
20
?
15
?
10
?
5
0
(a)
FIG. 3: Fabry-Perot etalon signals (a) with the filter turned
off, (b) at the transmitted port of the polarizer (Th) and (c)
at the reflected port (Th). In (a) both carrier and sideband
are present, in (b) the carrier is suppressed to approximately
-18 dB and in (c) the sideband is suppressed to the noise level
leaving only the carrier.

Page 3

3
the dashed red line shows the sideband frequency, ν+νm.
The temperature of the filter is approximately 70◦C and
the magnetic field generated by the solenoid is 80 G. The
measurement is taken in the ‘weak probe’ regime [21]. At
the carrier frequency, Tv is approximately 70% whereas
This close to zero. At the sideband frequency, ν + νm,
the contrary is the case, with Th being approximately
90% and Tv is less than 1%. This produces the desired
separation of light at frequencies 6.8 GHz apart without
a large portion of the light being absorbed by the vapor
cell. The Verdet constant for the filter can be found us-
ing equation 1. With a magnetic field of 80 G the Verdet
constant V = 2600 rad T−1m−1, which is more than an
order of magnitude larger than typical Faraday isolators.
In Figure 3 the performance of the filter is demon-
strated by using a Fabry-Perot etalon with a free spec-
tral range ∼ 745 MHz to analyse the light (a) with the
magnetic field switched off, (b) at the transmitted port
of the polariser (Th) and (c) at the reflected port (Tv).
The detuning axis is calibrated by taking the separation
of the sidebands and carrier as being equal to the EOM
modulation frequency. The second sideband, appearing
at a detuning of -6.8 GHz, is omitted for clarity as it has
no relevance to the Raman resonance. In Figure 3 (a)
the two components of the input beam are clearly visible
with the signal dominated by the carrier. In Figure 3 (b)
the carrier frequency is suppressed to approximately -18
dB, leaving the sideband unchanged. In Figure 3 (c) the
sidebands are suppressed to a level of -19 dB leaving only
the carrier. The suppression of the carrier is limited by
the extinction ratio of the polarization optics used.
In summary, we have demonstrated the filtering of light
separated in frequency by the ground state hyperfine
splitting in87Rb using the Faraday effect in an isotopi-
cally pure85Rb thermal vapor. This technique allows the
generation of phase stable beams suitable for driving Ra-
man transitions between the two hyperfine states or the
filtering of light generated by a Raman scattering pro-
cess. The filter has been shown to result in a suppression
of the carrier frequency to approximately -18 dB. We
expect this work to be applicable to other alkali metal
atom experiments where narrow band atomic filtering is
required.
The authors are grateful to S. L. Cornish for the loan of
equipment. We thank the EPSRC for financial support.
[1] M. Kasevich and S. Chu, Phys. Rev. Lett. 67, 181 (1991).
[2] E. Arimondo, Prog. Opt. 35, 257 (1996).
[3] M. G. Bason, A. K. Mohapatra, K. J. Weatherill and C.
S. Adams, Phys. Rev. A 77, 032305 (2008).
[4] M. G. Bason, A. K. Mohapatra, K. J. Weatherill and C.
S. Adams, J. Phys. B: At. Mol. Opt. Phys. 42, 075503
(2009).
[5] M. M¨ uller, I. Lesanovsky, H. Weimer, H. P. Bchler, P.
Zoller, arXiv:0811.1155v1 (2008) accepted to appear in
Phys. Rev. Lett.
[6] S. Manz, T. Fernholz, J. Schmiedmayer and Jian-Wei
Pan, Phys. Rev. A 75, 040101 (2007).
[7] M. Faraday, Phil. Trans. Royal Soc. 136, 1 (1846).
[8] Y.¨Ohman, Stockholm Observ. Ann. 19, 3 (1956).
[9] J. Menders, K. Benson, S. H. Bloom, C. S. Liu and E.
Korevaar, Opt. Lett. 16, 846 (1991).
[10] D. J. Dick and T. M. Shay, Opt. Lett. 16, 867 (1991).
[11] T. Junxiong, W. Qingji, L. Yimin, Z. Liang, G. Jianhua,
D. Minghao, K. Jiankun and Z. Lemin, Appl. Opt. 34,
2619 (1995).
[12] A. Popescu, K. Schorstein and T. Walther, Appl. Phys.
B 79, 955 (2004).
[13] H. Chen, C. Y. She, P. Searcy and E Korevaar, Opt. Lett.
18, 1019 (1993).
[14] S. K. Gayen, R. I. Billmers, V. M. Contarino, M. F.
Squicciarini, W. J. Scharpf, Guangning Yang, P. R. Her-
czfeld and D. M. Allocca, Opt. Lett. 20, 1427 (1995).
[15] A. Cer` e, V. Parigi, M Abad, F. Wolfgramm, A. Predo-
jevi´ c and M. W. Mitchell, Opt. Lett. 34, 1012 (2009).
[16] C. S. Gardner, Appl. Opt. 43, 4941 (2004).
[17] A. Balbin Villaverde, D. A. Donatti and D. G. Bozinis,
J. Phys. C: Solid State Phys. 11, 495 (1978).
[18] P. Siddons, N. C. Bell, C. Yifei, C. S. Adams, I. G.
Hughes, Nat. Photon. 3, 225 (2009).
[19] D. J. McCarron, S. A. King and S. L. Cornish Meas. Sci.
Technol. 19, 105601 (2008).
[20] P. Siddons, C. S. Adams, C. Ge and I. G Hughes, J. Phys.
B: At. Mol. Opt. Phys. 41, 155004 (2008).
[21] B. E. Sherlock and I. G. Hughes, Am. J. Phys. 77, 111
(2009).