Enhanced Spin Depolarization and Storage Time in a Rb Vapor

Abstract

The experiment of measuring the spin depolarized time and light storage time in a Rb vapor under different conditions is performed. Typically, these measurements are accomplished in three different containers: atoms in a bare glass cell, atoms in a buffer gas cell, and atoms in a tetracontane (C40H82) coating cell. The increasing depolarization and storage times are observed in both buffer gas cell and tetracontane coating cell. In the latter case, the storage time greater than 400 µs is obtained. PACS: 42. 50. Gy, 03. 67. −a Using light to manipulate the spin states in atomic ensembles lies at the heart of many quantum optics effects. [1] Recently, mapping a polarization state of light onto an atomic spin state via electromagnetically induced transparency (EIT) [2] has been proposed [3−6] and demonstrated. [7−9] Storage and recovery of light information in an atomic ensemble is an interesting topic in the field of quantum information because pho-tons are excellent long distance carriers of information, and quantum gates can be realized with atoms. [1,10,11] Light storage is implemented by the technique of ultraslow light group velocity and EIT effect, [12−14] which are both caused by the resonant or near res-onant interaction between light and a Λ-type atom (Fig. 1(a)). Typically, we use a control light to make an opaque medium become transparent near atomic resonance. An orthogonal polarized weak optical field (the "signal filed") can then propagate without much absorption but with a substantially reduced group ve-locity. At the same time a mixture of signal light field and collective atomic excitation of spin transitions, which are known as dark-state polariton, can be gen-erated. As we smoothly turn off the control field, the signal field is converted to a purely atomic spin-wave polariton and the quantum state of the signal field is then transferred to the atoms. By turning on the con-trol field again after a time interval, the stored signal field can be released out. The storage of arbitrary polarized lights with a good fidelity in the atomic ensembles has also been demonstrated recently. [15] This leads to a possible way to use current technology to photonic quantum mem-ory or quantum repeater. However, such light stor-age process requires high quality state preparation and minimal depolarization and decoherence of the spin state of the atomic ensemble. Depolarization and decoherence are generally caused by the collisions of the target atoms with the cell walls or with other atoms and by external fields such as the leakage of the Earth's magnetic field. [16] In warm atomic vapor cells, spin state lifetimes are often limited by the atom-wall collisions, which thermalize internal atomic states and destroy spin co-herence. To avoid such spin states deteriorate, filling the cell with noble buffer gas or coating the cell with paraffin material is usually used. The buffer gas keeps the atoms away from the walls of the cell such that the spin depolarization and decoherence are suppressed. Coating the walls of cell with a paraffin derivative allows atoms to undergo many wall-collisions with-out losing their spin coherence and increases spin life-times. In the previous papers, many concerns have been taken for the novel coating methods which can significantly improve the depolarization time and re-duce the group velocity. [17,18] Paraffin-coated alkali va-por cells have been successfully used to demonstrate spin squeezing, entanglement of atomic ensembles and quantum memory for continuous quantum variables, and are also used for high-precision atomic clocks and magnetometers. In this Letter, we report a comparison study to measure the spin depolarized time and light storage time in a Rb vapor at different conditions, i.e., atoms in a bare glass cell, atoms in a buffer gas cell, and atoms in a tetracontane coating cell. We find that the depolarization and storage times are increased in buffer gas cell and tetracontane coating cell. The ex-perimental results also show that coating the cell has significant impact on the light storage time. Three different Rb cells are assembled as follows. First, a thoroughly cleaned and dried glass cell sys-tem is constructed. The cell system composes of two main parts, a cylindrical glass cell, which is named as the working cell, and a rubidium source chamber.

Full text

Page 1

CHIN. PHYS. LETT. Vol.26,No.11(2009)114211
Enhanced Spin Depolarization and Storage Time in a Rb Vapor∗
QI Yue-Rong(???)1,2, GAO Hong(??)1,2∗∗, ZHANG Shou-Gang(???)1
1National Time Service Center, Chinese Academy of Sciences, PO Box 18, Xi’an 710600
2MOE Key Laboratory for Nonequilibrium Synthesis and Modulation of Condensed Matter, Xi’an Jiaotong University,
Xi’an 710049
(Received 23 July 2009)
The experiment of measuring the spin depolarized time and light storage time in a Rb vapor under different
conditions is performed. Typically, these measurements are accomplished in three different containers: atoms in
a bare glass cell, atoms in a buffer gas cell, and atoms in a tetracontane (C40H82) coating cell. The increasing
depolarization and storage times are observed in both buffer gas cell and tetracontane coating cell. In the latter
case, the storage time greater than 400µs is obtained.
PACS: 42.50.Gy, 03.67.−a
Using light to manipulate the spin states in atomic
ensembles lies at the heart of many quantum optics
effects.[1]Recently, mapping a polarization state of
light onto an atomic spin state via electromagnetically
induced transparency (EIT)[2]has been proposed[3−6]
and demonstrated.[7−9]Storage and recovery of light
information in an atomic ensemble is an interesting
topic in the field of quantum information because pho-
tons are excellent long distance carriers of information,
and quantum gates can be realized with atoms.[1,10,11]
Light storage is implemented by the technique of
ultraslow light group velocity and EIT effect,[12−14]
which are both caused by the resonant or near res-
onant interaction between light and a Λ-type atom
(Fig.1(a)). Typically, we use a control light to make
an opaque medium become transparent near atomic
resonance. An orthogonal polarized weak optical field
(the “signal filed”) can then propagate without much
absorption but with a substantially reduced group ve-
locity. At the same time a mixture of signal light field
and collective atomic excitation of spin transitions,
which are known as dark-state polariton, can be gen-
erated. As we smoothly turn off the control field, the
signal field is converted to a purely atomic spin-wave
polariton and the quantum state of the signal field is
then transferred to the atoms. By turning on the con-
trol field again after a time interval, the stored signal
field can be released out.
The storage of arbitrary polarized lights with a
good fidelity in the atomic ensembles has also been
demonstrated recently.[15]This leads to a possible way
to use current technology to photonic quantum mem-
ory or quantum repeater. However, such light stor-
age process requires high quality state preparation
and minimal depolarization and decoherence of the
spin state of the atomic ensemble.
and decoherence are generally caused by the collisions
Depolarization
of the target atoms with the cell walls or with other
atoms and by external fields such as the leakage of the
Earth’s magnetic field.[16]
In warm atomic vapor cells, spin state lifetimes
are often limited by the atom-wall collisions, which
thermalize internal atomic states and destroy spin co-
herence. To avoid such spin states deteriorate, filling
the cell with noble buffer gas or coating the cell with
paraffin material is usually used. The buffer gas keeps
the atoms away from the walls of the cell such that the
spin depolarization and decoherence are suppressed.
Coating the walls of cell with a paraffin derivative
allows atoms to undergo many wall-collisions with-
out losing their spin coherence and increases spin life-
times. In the previous papers, many concerns have
been taken for the novel coating methods which can
significantly improve the depolarization time and re-
duce the group velocity.[17,18]Paraffin-coated alkali va-
por cells have been successfully used to demonstrate
spin squeezing, entanglement of atomic ensembles and
quantum memory for continuous quantum variables,
and are also used for high-precision atomic clocks and
magnetometers.
In this Letter, we report a comparison study to
measure the spin depolarized time and light storage
time in a Rb vapor at different conditions, i.e., atoms
in a bare glass cell, atoms in a buffer gas cell, and
atoms in a tetracontane coating cell. We find that
the depolarization and storage times are increased in
buffer gas cell and tetracontane coating cell. The ex-
perimental results also show that coating the cell has
significant impact on the light storage time.
Three different Rb cells are assembled as follows.
First, a thoroughly cleaned and dried glass cell sys-
tem is constructed. The cell system composes of two
main parts, a cylindrical glass cell, which is named
as the working cell, and a rubidium source chamber.
∗Supported by the National Natural Science Foundation of China under No 10834007, and the Key Program of Chinese Academy
of Sciences (KJCX2-SW-T12)
∗∗Email: honggao@mail.xjtu.edu.cn
c ?2009 Chinese Physical Society and IOP Publishing Ltd
114211-1

Page 2

CHIN. PHYS. LETT. Vol.26,No.11(2009)114211
The glass cell system is connected to a vacuum system
through a vacuum valve and is pumped to a vacuum
pressure at about 10−7torr. Then the vacuum valve
is closed, and a pure rubidium 87 isotope ampoule in
the source chamber is broken by an iron harmer and
a magnets. Rubidium is transferred to the cylindrical
working cell by using a heat gun. The bare rubidium
cell is directly made by sealing the working cell from
vacuum at this stage. While the buffer gas cell is made
by injecting the 5 torr Nitrogen into the working cell
before sealing.
(a)
e>
g>
g>
CL
SL
(b)
(c)
(d)
DL
λ/4
Rb cellPD
C
PB
PL
BS
M
DL
AOM
Iris
PC
Rb cell
PBS
PD1
PD2
M
II
Glass valve
I
Fig. 1. Schematic of the atomic energy level and exper-
imental setup.(a) A Λ-type atom for EIT effect.
Experimental configuration to measure the spin depolar-
ized time. (c). Experimental configuration to measure the
light storage time. (d) The schematic structure of tetra-
cantone coating cell. Regions I and II are the working cell
and Rb storage cell, respectively. |e?, excited state; |g1?
and |g2?, ground states; CL, control light; SL, signal light;
DL, diode laser; PL, pump light; PB, probe beam; BS,
beam splitter; M, mirror; PD, photo detector; C, optical
chopper; PC, Pockel’s cell; PBS, polarized beam splitter.
(b)
The coating cell is assembled more complex than
the above two cells. First, we have a similar glass cell
system, while the working cell now is connected to a
long glass tube with a glass valve on it. We drop some
tetracontane into the working cell and pump the whole
glass cell system to about 10−7torr. Then the glass
valve is closed. We warm up the working cell region
until all the tetracontane is melt. Keep this status for
several hours such that the inner wall of the working
cell is well coated by tetracontane. When this step
is finished, the working cell is cooled down to room
temperature and back to vacuum again by gradually
switching on the glass valve. After this stage, the glass
and vacuum valves are closed. Repeat the bare cell as-
sembling process we produce the coating cell with Rb
stored in a storage cell (Fig.1(d)).
In all our experiment, we maintain the Rb cell
temperature of 70–80◦C, corresponding to an atomic
density of about 1011–1012/cm3. For the coating cell,
we keep the working cell temperature slightly higher
(80–90◦C) than that of storage cell in order to avoid
Rb being stack and attached on the working cell win-
dows. Therefore, we first warm up the working cell to
a target temperature, then open the glass valve and
warm up the storage cell to a temperature slightly
lower than the working cell target temperature. Re-
verse the above processes to recycle the Rb to the
storage cell again when the experiment is finished.
The first experiment we perform is to measure the
depolarized time. Figure 1(b) shows the experimental
setup. An extended cavity diode laser with 300kHz
bandwidth is used as the light source. After passing
through a quarter wave plate, an original linearly po-
larized light is converted to the circular one. Then the
laser beam is split to two parts with the same circu-
lar polarization. The main beam carrying most of the
laser power severs as a pump light and polarizes the
Rb atoms in its path. The weaker probe beam, which
is not strong enough to noticeably affect the polar-
ization of the cell, is absorbed by unpolarized atoms
in its path. By polarizing the atom with the pump
beam and then measuring the polarization over time
with the probe beam, spin depolarized time can be
determined. The laser frequency is adjusted to the
D1 transition of87Rb (λ = 794.987nm), i.e., 5S1/2,
F = 2 → 5P1/2, F = 1, which is checked by observing
the fluorescence or absorption spectrum. The laser
beam diameter is about 5mm with an output power
of 8mW.
?
?
?
Fig. 2. Experimental results of measuring spin depolar-
ized time T1. (a) Bare rubidium cell, T1 = 10µs. (b)
Buffer gas filling cell, T1= 300µs. (c) Tetracantone coat-
ing cell, T1= 3.21ms.
Figure 2 shows the experimental results of (a) a
bare rubidium cell, (b) a buffer gas filling cell, and (c)
the tetracantone coating cell, respectively. We fit the
experimental results with the exponential decay and
obtain the lifetime as 10µs for bare cell, 300µs for
buffer gas cell and 3.2ms for coating cell. The exper-
114211-2

Page 3

CHIN. PHYS. LETT. Vol.26,No.11(2009)114211
imental results show that coating the glass wall has a
significant impact on the depolarized time, which re-
veals that the collision of atoms with the cell wall is
an important factor relative to the depolarized time.
The second experiment we perform is the light
storage experiment. Figure 1(c) shows the experi-
mental setup. A linear polarized laser beam is rapidly
switched by an acoustic-optical modulator (AOM) and
iris. A fast Pockel’s cell slightly rotates the linearly po-
larized laser light to create a weak pulse, which is the
signal field. The polarization of the weak pulse is per-
pendicular to that of the remaining light, which serves
as the control field. Both the signal and control light
are then sent into the Rb cell to accomplish the light
storage experiment. A linear polarized light is used to
keep the experiment simple. Other polarized lights,
especially circular light, are found to have no effect
the light storage time. A PBS and two photodiodes
(PD) comprise the detection system. For the light
storage experiment, the time sequence of the control
and signal fields is the same as in Refs.[7,8,15].
?
?
?
Fig. 3.
time Ts. (a) Bare rubidium cell, Ts = 3.89µs. (b) Buffer
gas filling cell, Ts= 46.6µs. (c) Tetracantone coating cell,
Ts= 398.6µs.
In order to complete the light storage experiment,
one should first check the EIT effect. We carefully
adjust the laser frequency until EIT occurs. This can
be checked when the signal light is totally absorbed
by blocking the control light. Figure 3 shows the light
storage experimental results in three different cells,
Experimental results of measuring light storage
where (a) is bare cell, (b) is buffer gas filling cell and
(c) is tetracontane coating cell. The solid lines are the
exponential decay fit to the experimental data and the
fitting results give the storage time about 3.89µs for
bare cell, 46.6µs for buffer gas cell and 398µs for the
coating cell, respectively. Corresponding to the depo-
larized time measurement results, the storage time is
decreased almost by one order. This is mainly because
the light storage time is not only governed by the de-
polarization time, but also by the decoherence time.
Indeed decoherence time plays an important role in
the light storage experiment since the retrieval signal
is coherently added when control light is switched on
again. Unlike the depolarized time, the signal exists
as long as atoms are still on the target energy level,
we do not have any light storage signal if these atoms
loss the phase coherence.
In summary, we have performed the experiment to
measure the spin depolarized time and light storage
time in three different Rb vapor cells, and found that
depolarized and storage times are increased both in
buffer gas cell and tetracontane coating cell. In the
latter case, we obtain the storage time greater than
400µs.
References
[1] Hielsen M A and Chuang I L 2000 Quantum Computation
and Quantum Information (Cambridge: Cambridge Uni-
versity)
[2] Harris S E 1997 Phys. Today 50 36
[3] Fleischhauer M and Lukin M D 2000 Phys. Rev. Lett. 84
5094
[4] Csesznegi J R and Grobe R 1997 Phys. Rev. Lett. 79 3162
[5] Kocharovskaya L, Rostovtsev Y and Scully M O 2001 Phys.
Rev. Lett. 86 628
[6] Fleischhauer M and Lukin M D 2002 Phys. Rev. A 65
022314
[7] Phillips D F, Fleischhauer A, Mair A, Walsworth R L and
Lukin M D 2001 Phys. Rev. Lett. 86 783
[8] Liu C, Dutton Z, Behroozi C H and Hau L V 2001 Nature
409 490
[9] Zibrov A S, Matsko A B, Kocharovshaya O, Rostovtsev Y
V, Welch G R and Scully M O 2002 Phys. Rev. Lett. 88
103601
[10] Qi R, Yu X L, Li Z B, Liu W M 2009 Phys. Rev. Lett. 102
185301
[11] Ji A C, Liu W M, Song J L, Zhou F 2008 Phys. Rev. Lett.
101 010402
[12] Zhao Y, Zhao J, Xiao L, Yin W and Jia S 2004 Chin. Phys.
Lett. 21 76
[13] Ye C 2004 Chin. Phys. Lett. 21 853
[14] Zhou B, Du C and Li S 2004 Chin. Phys. Lett. 21 856
[15] Gao H, Rosenberry M and Batelaan H 2003 Phys. Rev. A
67 053807
[16] Suter D 1997 The Physics of Laser-Atom Interactions
(Cambridge: Cambridge University) chap 5
[17] Phillips D F, Boca A and Walsworth R L 1999 Evaporative
Coating of Rb Maser Cells (unpublished)
[18] Klein M, Novikova I, Phillips D F and Walsworth R L 2006
J. Mod. Opt. 53 2583
114211-3