OPTICAL PUMPING OF RUBIDIUMphyweb.physics.nus.edu.sg/~L3000/Level3manuals... · electron spin,...
Transcript of OPTICAL PUMPING OF RUBIDIUMphyweb.physics.nus.edu.sg/~L3000/Level3manuals... · electron spin,...
Page 1 of 24
OPTICAL PUMPING OF RUBIDIUM
1. INTRODUCTION
Optical pumping [123] is a process used in high frequency spectroscopy,
developed by A. Kastler who was awarded the Nobel Prize for this in 1966.
In this experiment, anomalous Zeeman transitions in the ground state of 87Rb and 85Rb with σ+- and σ--circular-polarized pumping light are observed
and measured.
2. THEORY
a) Energy level scheme
Rubidium is an alkali metal in the first main group of the periodic table.
The low energy states of this element can be described very well using
Paschen notation, i.e. 2S+1LJ.
The rubidium atom consists of a spherically symmetrical atomic residue with
orbital spin 0, and one optically-active electron with an orbital angular
momentum 0,1,2... and an electron spin of ½. The fine structure of the energy states is illustrated in Fig. 1. It is caused by the half-integral
electron spin, which leads to a multiplicity of 2, i.e. a doublet system.
The ground state is an S-state; the orbital angular momentum here is L = 0.
The spin-orbit coupling results in a total angular momentum quantum number
J = ½. Due to this coupling, the first excited level with L = 1 splits up into a 2P1/2 and a
2P3/2 state. Both states can be easily excited in a gas
discharge. During the transitions from the first excited states 2P1/2 and 2P3/2 into the ground state
2S1/2, the doublet D1 and D2 characteristic of
all alkali atoms is emitted. For rubidium, the wavelengths of the
transitions are 794.8 nm (D1 line) and 780 nm (D2 line).
The hyperfine interaction, caused by the coupling of the orbital angular
momentum J with the nuclear spin I, leads to the splitting of the ground
state and the excited states. The additional coupling provides hyperfine
levels with a total angular momentum F = I ± J. For 87Rb with a nuclear
spin I = 3/2, the ground state 2S1/2 and the first excited state 2P1/2 split
up into two hyperfine levels with the quantum numbers F = 1 and F = 2.
Page 2 of 24
Compared with the transition frequency of 4 x 1014 Hz between 2P1/2 and 2S1/2, the resulting splitting of the ground state and the excited states
is much smaller. For the ground state, this hyperfine splitting of
6.8 x 109 Hz is approx. 5 powers of 10 smaller than the fine structure
splitting.
Fig. 1 Scheme of energy levels for 87Rb
In the magnetic field, an additional Zeeman splitting (Fig. 1) into 2F + 1
sub-levels respectively is obtained. For magnetic fields of approx. 1 mT,
the transition frequency between neighbouring Zeeman levels of a hyperfine
state is 8 x 106 Hz, i.e. another 3 powers of 10 smaller than the hyperfine
splitting. The energy respective frequency relationships between the
individual states are of particular significance in understanding optical
pumping.
Fine structure splitting
Hyperfine structure splitting
Zeeman splitting
Page 3 of 24
b) Optical pumping
The process can be explained in more detail using the energy level scheme
of 87Rb (Fig. 1) as a reference:
The transitions from 2S1/2 to 2P1/2 and 2P3/2 are electrical dipole
transitions. They are only possible if the selection rules ∆mF = 0 or
∆mF = ±1 have been fulfilled.
The transitions between the Zeeman levels are detected using a method
discovered by A. Kastler in 1950 which will be described in the
following [1]. The D1 line emitted by the rubidium lamp displays such a
high degree of Doppler broadening that it can be used to induce all
permissible transitions between the various Zeeman levels of the 2S1/2 and 2P1/2 states. If an absorption cell located in a weak magnetic field and
filled with rubidium vapour is irradiated with the σ+ circularly-polarized
component of the D1 line, the absorption taking place in the cell excites
the various Zeeman levels which are higher by ∆mF = +1. However, the
excited states decay spontaneously to the ground state and re-emit π, σ+ or σ- light in all spatial directions, in accordance with the ∆mF = 0 or
∆mF = ±1 selection principle.
The irradiating, circularly-polarized light effects a polarization of the
atomic vapour in the absorption cell. This can be interpreted as follows:
during the process of absorption, the polarized light transmits angular
momentum to the rubidium atoms. The rubidium vapour is polarized and thus
magnetized macroscopically.
Without optical irradiation, the difference between the population numbers
of the various Zeeman levels in the ground state is infinitesimally small,
due to its low energy spacing in thermal equilibrium. However, the
irradiation with σ+ light results in a strong deviation from the thermal
equilibrium population. In other words, the population distribution changes
as a result of optical pumping. The F = 2, mF = +2 level has the largest
population probability — represented in Fig. 1 by bolder print — as the
angular momentum selection principle prevents excitation from this level to
higher ones. A Zeeman state with the mF quantum number 3 does not exist in
the first excited state. In other words, the irradiation of the absorption
cell with the D1 σ+ line generates a new population distribution in the
Page 4 of 24
equilibrium between optical pumping and the simultaneous relaxation
processes. These are caused by depolarizing collisions of rubidium atoms
with the walls of the vessel, or by interatomic collisions. In the pumping
equilibrium, the population probabilities decrease in the order
mF = 2, 1, 0, -1, -2 in the F = 2 state, and in the order mF = -1, 0, +1 in
the F = 1 state.
In the case of the magnetic fields used here, the position of the energy
states can be calculated with the Breit-Rabi formula [4,5]
For 21
IF ±=
( )( )
2
1
2FFINF 1I2
m41
2mg
1I22m,F
ξ+ξ
++
∆Ε±Βµ+
+∆Ε
−=Ε
where Β∆Ε
µ−µ=ξ NIBJ gg
(I)
F = Total angular momentum
I = Nuclear spin
J = Angular momentum of the electron shell
mF = Magnetic quantum number of the total angular momentum F
gI = g-factor of the nucleus
gJ = g-factor of the electron shell
∆Ε = Hyperfine structure spacing
µB = Bohr magneton
µN = Nuclear magneton
B = Magnetic flux density
When irradiated with σ+ pumping light, the Zeeman levels within a hyperfine
state which have positive quantum numbers mF become enriched at the expense
of the levels with negative quantum numbers. The result is a population
which deviates from the thermal equilibrium population. In the ground state 87Rb the level F = 2, mF = +F has the greatest population.
When irradiated with a linearly polarized alternating magnetic field of the
correct frequency, more transitions take place from a higher Zeeman level
mF to the next lowest level with mF - 1 than in the other direction. With
Page 5 of 24
σ- pumping light, the situation is just the opposite, as Zeeman levels with
negative quantum numbers predominate.
The frequency f of these transitions is always
( )h
g1mmf IN
FFΒµ
±=−↔
( )
ξ+ξ
+
−+−
ξ+ξ
++
∆Ε+
2
1
2F2
1
2F
1I2
1m41
1I2
m41
h2 (II)
The result is an average frequency spacing of approx. 8 MHz mT-1 between
neighbouring Zeeman levels in the ground state 2S1/2 for 87Rb. This means
that in a magnetic field of 1 mT, irradiation with a high frequency of
8 MHz can induce transitions between ground state levels. In the case of
the magnetic fields used here, the energy spacing ∆EF = hf between the
Zeeman levels in the ground state of 87Rb amounts to 3 x 10-8 eV. For the
induced emission, the transition of the atom between two Zeeman levels with
∆mF = 1 is accompanied by the emission of the inducing HF quantum as well
as one having the same energy. In view of the very small number of atoms
available at a vapour pressure of 10-5 mbar, the total energy released in
this process is so low that a direct detection is not possible; a higher
vapour pressure leads to a broadening of the lines due to collisions. If
the HF field were observed directly, the signals would be masked entirely
by noise. The process developed by A. Kastler requires that, for every act
of induced emission, the population equilibrium between optical pumping and
relaxation be changed by 1 atom. This allows the absorption of an
additional pumping photon with an energy ∆Eopt = 1.5 eV. By observing this
absorption directly, one thus achieves an amplification of
1.5/(3x10-8) = 5 x 107, which is obtained without incurring.
Page 6 of 24
c) Experiment arrangement
Fig. 2 is an overview diagram of the experimental set-up. Fig. 3
schematically illustrates the individual optical and magnetic components
and the optical path.
Fig. 2: Overview diagram of the entire experimental setup; position specifications are measured from the left edge of the optical riders
1 Rubidium high frequency lamp 2 Lens 3 Interference filter 4 Linear polarization filter 5 λ/4 plate 6 Lens 7 Helmholtz coil 8 Absorption chamber 9 Absorption cell 10 HF coils 11 Lens 12 Silicon photodetector
Fig. 3: Schematic representation of the experiment set-up
Page 7 of 24
A high-frequency rubidium lamp (1) serves as the light source. A quartz
cell filled with 1 to 2 mbars of Argon and containing a small amount of
rubidium is located in the electromagnetic field of a resonant circuit coil
of an HF transmitter (60 MHz, 5W) which is integrated in the lamp housing.
In order to prevent an excessive rate of rise of the Rb vapour pressure
during HF excitation and thus the spontaneous inversion of the D1 line, the
lamp temperature is maintained at approx. 130 °C by means of an
electronically controlled thermostat. For this, the lamp is in thermal
contact with the electrically heated base plate. A temperature sensor and a
two-point controller in the supply device control the temperature of this
plate and thereby the flow of heat between the lamp and the lamp housing.
The lamp is located at the focus of the lens (2).
An interference filter (3), a linear-polarization filter (4) and a
λ/4 plate (5) are used to separate the σ+ circularly-polarized component of
the D1 line from the emission spectrum of the rubidium lamp. The D1 σ+
light thus generated is focused on to the absorption cell (9) with a lens
(6) having a short focal length. The absorption cell (9) also contains
rubidium. It is located in a water bath, the temperature of which is
maintained at approx. 70 °C with a circulation thermostat. At this
temperature, the rubidium vapour pressure in the absorption cell settles at
approx. 10-5 mbar. The absorption chamber (8), together with the absorption
cell (9) is located in a pair of Helmholtz coils (7), so that the magnetic
field necessary for the Zeeman separation can be generated. The axis of the
magnetic field coincides with the optical axis of the experiment set-up.
The transmitted light is focused onto a silicon photodetector (12) by means
of another lens (11) with a short focal length. The detector signal is
proportional to the intensity of the radiation passing through the
absorption cell and is fed to the Y input of an oscillograph, an XY
recorder or an interface, via a downstream current-to-voltage converter. HF
coils (10) are positioned on both sides of the absorption cell. A function
generator serves as the HF source. Its frequency is swept around by a
gradually rising, saw-tooth voltage. This voltage is simultaneously used
for the X-deflection of the registering system. As the HF coils are
oriented perpendicularly with respect to the Helmholtz coil field, the
linearly polarized HF coil field generates the circularly-polarized HF
field necessary for the Zeeman transitions. The linearly polarized HF field
can be divided into one right-handed and one left-handed circularly-
Page 8 of 24
polarized field; the component with the correct direction of rotation
(corresponding to ∆mF = ±1) induces transitions in the rubidium vapour of
the absorption cell.
The process makes it possible to observe the absorption in terms of its
dependence on the HF irradiation.
d) Observation of the pump signal
Without HF irradiation, an equilibrium is reached in the absorption cell
while pumping with σ+ light; in the ground state 2S1/2, F = 2, the Zeeman
level with mF = +2 is the most populated. The absorption of the cell is
constant and thus so is the light intensity observed in transmission. If
the circular polarization of the pumping light is now quickly changed from
σ+ to σ-, light is absorbed more strongly. With the irradiation from σ-
light, absorption transitions with ∆mF = -1 are possible, and a new
equilibrium settles in, where now the state 2S1/2, F = 2, mF = -2 is the
most populated. The circular polarization of the pumping light radiation
(σ+ or σ-) relative to the magnetic field orientation is decisive in
determining whether ∆mF = +1 or ∆mF = -1 transitions occur. Thus when
right-handed circularly-polarized light is used for irradiation parallel to
the magnetic field, it produces ∆mF = +1 transitions, but for irradiation
antiparallel to the field it causes ∆mF = -1 transitions. Instead of the
quick switchover from σ+ to σ- pumping light, which is experimentally
difficult to realize, a direction changeover of the magnetic field can be
used. This change in the magnetic field direction can be easily carried out
by changing the polarity of the Helmholtz coils with a switch. The signals
observed when this is done (Fig. 4) are a clear indication that the
absorption cell is being pumped optically.
Fig.4 Above: voltage across the Helmholtz coil pair with respect to time Below: Pumping light signal with respect to time
Page 9 of 24
e) HF transitions between Zeeman levels in the ground state
Two processes can be used to observe the transitions between Zeeman levels
in the ground state. In one case, the magnetic field can be slowly changed
at constant frequency. With increasing magnetic field strength the energy
spacing between neighbouring Zeeman levels increases. If this spacing is
exactly E(F,mF) - E(F,mF-1) = hf0 and if f0 is the frequency of the
irradiating high frequency, induced transitions between the corresponding
Zeeman levels take place and the absorption for the pumping light changes.
If the HF irradiation is strong enough, population equilibrium of the
corresponding Zeeman levels is forced and, as a result, a deviation from
pumping equilibrium.
In the other case, the frequency of the irradiating HF field can be slowly
varied at constant magnetic field strength. Here, the absorption changes
each time ∆E = hf reaches the spacing between neighbouring Zeeman levels.
We will make use of the latter method here.
The spacing between neighbouring Zeeman levels in the hyperfine state F = 2
increases monotonically in the order mF = +2 to +1, +1 to 0, 0 to –1,
-1 to –2 and in the F = 1 state from mF = -1 to 0, 0 to +1. As a result of
this, the individual absorption signals appear in the corresponding order
with increasing frequency of the HF irradiation. The relative position
between the absorption transitions of the F = 2 state and those of the
F = 1 state is dependent on the strength of the applied magnetic field. The
intensity of the absorption lines is determined by the population
difference between the Zeeman levels involved each time. When σ+ light is
used for pumping, the transition F = 2, mF = +2 to +1 appears with the
greatest intensity (Fig. 5). Here we are dealing with a special case of the
anomalous Zeeman effect. If the polarization is changed, i.e. σ- light is
radiated into the absorption cell instead of σ+ light, or the magnetic
field direction of the Helmholtz coils is reversed while the polarization
is retained, the transition F = 2, mF = -2 to –1 appears with the greatest
intensity.
Page 10 of 24
Fig. 5 Absorption spectrum 87Rb with σ+ light
f) Zeeman splitting in 85Rb
A further special feature of the anomalous Zeeman effect is displayed in
the multiplicity of the Zeeman energy levels in isotopes with different
nuclear spins. In natural rubidium the isotopes 87Rb and 85Rb occur in the
ratio 72:28. While 87Rb has nuclear spin 3/2, 85Rb has a nuclear spin 5/2.
The corresponding angular momentum quantum numbers of the hyperfine levels
are F = 3/2 + 1/2 = 2 and F = 3/2 – 1/2 = 1 for 87Rb and F = 3 and F = 2
for 85Rb. Due to the different nuclear g factors the hyperfine splitting is
6.8 x 109 Hz for 87Rb and 3.1 x 109 Hz for 85Rb.
This, however, means that for the same magnetic field the Zeeman splitting
is of different magnitude for the two isotopes. While the splitting for 87Rb in the ground state has an average value of 8 MHz for a magnetic field
of 1 mT, for 85Rb and the same field the corresponding value is 5 MHz.
Both pumping light source and absorption cell contain the natural isotope
mixture. By setting the appropriate frequency on the function generator the
Zeeman structures of the two isotopes can be observed separately. The
hyperfine states with F = 2 and F = 1 for 87Rb split into 2F + 1 = 5 or
2F + 1 = 3 Zeeman levels respectively, while the splitting for 85Rb has 7
Page 11 of 24
or 5 levels respectively. Correspondingly 4 + 2 lines occur in the 87Rb
absorption spectrum (Fig. 5) and 6 + 4 lines in the 85Rb spectrum (Fig. 6).
Fig. 6 Absorption spectrum 85Rb with σ+ light
3. PROCEDURE
a) Observation of the pumping signal
i) Optical adjustment of the experiment setup and warming up the system
take several hours. Ask the laboratory officer in charge of the
experiment to prepare the apparatus in the morning so that you can
perform the experiment in the afternoon the same day.
Settings:
Operation device for optical pumping:
Supply voltage: U = 19.0 V
Temperature setting: θ ≈ 118 °C
Circulation thermostat:
Temperature setting: 65 °C
Page 12 of 24
ii) Measuring
▬ Switch on the oscilloscope’s storage mode by pressing and holding
the oscilloscope STOR. ON/HOLD pushbutton for about 1 second.
▬ Set the Helmholtz coil current Icoil ≈ 0.8 A.
▬ Press and hold the SINGLE pushbutton for about 1 second to switch to
SINGLE mode.
▬ Start oscilloscope storage mode by briefly pressing the oscilloscope
button RES. (RESET)
▬ Toggle the two-way switch back and forth.
Settings:
Oscilloscope:
Operating mode: Storage mode
Trigger: Auto
Y-deflecting, Channel II: 200 mV/DIV. (DC)
Timebase: 1 s/DIV
Polarization Filter
Angle: 0°
Quarter wavelength panel:
Angle: +45° or -45°
I/U Converter:
Toggle switch: DC
Measuring example:
Fig. 7 Transmitted intensity as a function of time for multiple rapid reversals of the Zeeman magnetic field.
Page 13 of 24
b) Observation and measurement of Zeeman splitting in 87Rb
i) Finding the absorption signal
▬ For 87Rb the frequencies of the Zeeman transitions in the ground
state in a Helmholtz field of 1.2 mT (coil current 740 mA) are
around 8 MHz. When the setup is optimally adjusted, the absorption
signal reaches an amplitude of at least 40 mV:
▬ Set the polarization filter to 0° and the quarter wavelength panel
to +45° or -45°.
▬ Set the desired operating mode and frequency range on the function
generator.
▬ Start the function generator by pressing the button labelled MANUAL.
▬ Vary the Helmholtz coil current until the maximum (negative)
absorption signal appears on the oscilloscope.
▬ If necessary, set the toggle switch of the I/U converter to AC or
use the offset potentiometer of the I/U converter to bring the
signal back to the middle of the oscilloscope screen.
▬ Maximize the absorption signal by changing the operating parameters
of the rubidium high-frequency lamp.
Settings:
Polarization filter:
Angle 0°
Quarter wavelength panel:
Angle: +45° or -45°
I/U converter:
Toggle switch: DC
Oscilloscope:
Operating mode: X-Y mode
Channel I: 500 mV/DIV. (DC)
Channel II: 20 mV/DIV. (DC)
Page 14 of 24
Function generator:
Function: : (sine)
Amplitude: Middle position
Attenuation: 20 dB
DC-offset: 0 V (DC button pressed)
Sweep button: Pressed
Mode*: ‘Cu
Stop*: 8.5 MHz
Start*: 7.5 MHz
Period*: ca. 100ms (fast sweep)
* Press button and set desired value with knob
ii) Measuring
Preparation
Finding the signal:
▬ Operate the oscilloscope in the X-Y mode by pressing button DUAL for
about 1s until XY appears on the screen.
▬ Switch off the oscilloscope storage mode.
▬ If necessary, switch off the sensitivity of oscilloscope channel I.
▬ Set the toggle switch of the I/U converter to DC.
▬ Set the start and stop frequencies on the function generator
(fA = 9.0 MHz, fE = 9.2 MHz)
▬ Switch the function generator to period 100 ms (fast sweep).
▬ Start the function generator by pressing the button labelled MANUAL.
▬ Set the Helmholtz coil current I ≈ 0.791 A and vary it until an
absorption signal can be seen on the oscilloscope screen.
Page 15 of 24
Oscilloscope storage mode:
▬ Switch on the storage mode of the oscilloscope.
▬ Press the Start button of the function generator.
▬ Adjust the horizontal deflection of the oscilloscope with the knob
Y-Pos I to xA = 1.0 scale division.
▬ Press the Stop button of the function generator.
▬ Press and hold CH I – VAR pushbutton until the VAR-LED located above
the VOLTS/DIV. knob is lit.
▬ Adjust the horizontal deflection of the oscilloscope with the
channel I VOLTS/DIV. knob to xE = 9.0 scale divisions.
Fine adjustment:
▬ Start the function generator by pressing the button labelled MANUAL.
▬ Switch the function generator to period 10 s (slow sweep).
▬ Switch the vertical deflection of the oscilloscope to sensitive.
▬ Turn the quarter wavelength panel back and forth between +45° and
-45° and check whether all lines of the absorption spectrum appear
on the oscilloscope screen.
▬ If necessary, readjust the Helmholtz coil current I or the start
frequency fA and stop frequency fE accordingly.
Procedure
Note: at σ+-polarization the absorption line with the lowest frequency
is the most intense.
First run:
▬ Set the quarter wavelength panel to σ+-polarization.
▬ Start the function generator by pressing the button labelled MANUAL.
▬ Wait until the absorption spectrum has been completely recorded.
Page 16 of 24
▬ Stop recording of the absorption spectrum by pressing the
oscilloscope button HOLD.
▬ Determine the position x of the absorption lines on the oscilloscope
screen.
▬ Determine the amplitude U of the absorption lines.
▬ Check the start frequency fA and the stop frequency fE.
▬ Check the Helmholtz coil current I.
Second run:
▬ Set the quarter wavelength panel to σ--polarization.
▬ Start the function generator by pressing the button labelled MANUAL.
▬ Wait until the absorption spectrum has been completely recorded.
▬ Stop recording of the absorption spectrum by pressing the
oscilloscope button HOLD.
▬ Determine the position x of the absorption lines on the oscilloscope
screen.
▬ Determine the amplitude U of the absorption lines.
▬ Check the start frequency fA and the stop frequency fE.
▬ Check the Helmholtz coil current I.
▬ Tabulate x and U values in Table 1.
Setting:
Oscilloscope:
Operating mode: X-Y mode
Storage mode
Trigger: Auto
Channel I: >500 mV/DIV. (DC)
Channel II: 10 mV/DIV. (DC)
Recording range: 1.0 division – 9.0 divisions
Timebase: 200 S/s
Page 17 of 24
Function generator:
Stop: 9.2 MHz
Start: 9.05 MHz
Period: 10 s (slow sweep)
Amplitude: 12½ o’clock
Attenuation: 20 dB
Measuring example:
Fig. 8 Absorption spectrum 87Rb with σ+ light (top) and σ- light (bottom).
IPhoto
0.2 µA 9.10 9.15
9.10 9.15 f [MHz]
f [MHz]
1 2
3
4
5
6
2
3 4 5 6
1
Page 18 of 24
No. x [Div] U [mV] No. x [Div] U [mV]
1 1
2 2
3 3
4 4
5 5
6 6
Table 1: Position x, amplitude U of the 87Rb absorption lines.
left side: Measurement with σ+ light, I = __________ A
right side: Measurement with σ- light, I = __________ A
iii) Evaluation
Determining the nuclear spin I of 87Rb
The irradiated alternating magnetic field makes possible transitions
with the selection rules ∆F = 0, ±1 and ∆mF = ±1. The available
frequency range only permits ∆F = 0. Thus, between the nuclear spin I
and the number of transitions n there exists the relationship
n = 2F+ + 2F- = 4I (III)
We can observe six lines. According to equation (III), the nuclear
spin of 87Rb thus has the value I=3/2.
Determining the transition frequencies.
The position x on the oscilloscope screen allows us to determine the
frequencies fM of the absorption lines using the formula
( )AE
AAEAM xx
xxffff
−
−−+= (IV)
xA = Start of scale (set with Y-POS. I)
xE = End of scale (set with channel I VOLTS/DIV. knob)
fA = Start frequency
fE = Stop frequency
Summarize the results in Table 2 as the quantities fM(σ+) and fM(σ
-).
The standard deviation for these quantities is σ = 0.002 MHz.
Page 19 of 24
No. fM(σ+) [MHz] fM(σ-) [MHz] f [MHz] F mF(1)↔mF(2)
1 2 2 ↔ 1
2 2 1 ↔ 0
3 2 0 ↔-1
4 1 1 ↔ 0
5 2 -1↔-2
6 1 0 ↔-1
Table 2: 87Rb transition frequencies at B = __________ mT
Assignment of the transitions
The following numerical values can be found in the literature:
µN = 5.05083(4) x 10-27 JT-1 [9]
µB = 9.27410(7) x 10-24 JT-1 [9]
gJ = 2.002332(2) [6]
h = 6.6260(5) x 10-34 Js [9]
∆E/h = 6834.682614(1) MHz [8]
gI = -1.827231(2) [7]
Using (II), we obtain within the hyperfine state F = 2 four transitions
which have frequencies
( ) ( )( )
ξ+ξ−+−ξ+ξ++
×−==
212F
212F
2F
1m1m1MHz34.3417
mTB
MHz01393.0f
(V)
and within the hyperfine state F = 1 two transitions with the
frequencies
( ) ( )( )
ξ+ξ−+−ξ+ξ++
×==
212F
212F
1F
1m1m1MHz34.3417
mTB
MHz01393.0f
(VI)
where mTB
10)4(10243.4 3 ××=ξ −
Page 20 of 24
B ≈ 1.64 x I(A) mT
Start with the above approximate B value. Vary B in equations (V) and
(VI) until the measured transition frequencies are reproduced
mathematically. Record the value of B and compare the measured
transition frequencies fM(σ+) and fM(σ
-) with the calculated transition
frequencies f in Table 2.
When changing from σ+- to σ--pumping light, the amplitude U of the
absorption lines also change. Compare the intensities of the
transitions belonging to hyperfine state F = 2 and F = 1 in Table 3.
No. U(σ+)
[mv]
U(σ-)
[mV]
mF(1)↔mF(2) No. U(σ+)
[mV]
U(σ-)
[mV]
mF(1)↔mF(2)
1 2 ↔ 1 4 1 ↔ 0
2 1 ↔ 0 6 0 ↔-1
3 0 ↔-1
5 -1↔-2
Table 3: 87Rb transition intensities at B = __________ mT. left side: F = 2, right side: F = 1
Page 21 of 24
c) Observation and measurement of Zeeman splitting in 85Rb
Repeat all procedures in (b) with the following settings:
Helmholtz coil current I = 0.838 A
Oscilloscope:
Operating mode: X-Y mode
Storage mode
Trigger: Auto
Channel I: >500 mV/DIV. (DC)
Channel II: 10 mV/DIV. (DC)
Recording range: 1.0 division — 9.0 divisions
Timebase: 200 S/s
Function generator:
Stop: 6.55 MHz
Start: 6.35 MHz
Period: 10 s (slow sweep)
Amplitude: 12½ o’clock
Attenuation: 20 dB
Measuring example:
Fig. 9 Absorption spectrum 85Rb with σ+ light (top)
and σ- light (bottom)
1
2
3 4
5 6 7 8
9 10
1 2 3
4 5
6
7
8
9
10
0.2 µA
6.40 6.45 6.50 f [MHz]
6.40 6.45 6.50 f [MHz]
IPhoto
Page 22 of 24
No. x [Div] U [mV] No. x [Div] U [mV]
1 1
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
Table 4: Position x, amplitude U of the 85Rb absorption lines.
left side: Measurement with σ+ light, I = __________ A
right side: Measurement with σ- light, I = __________ A
Assignment of the transitions
The following numerical values can be found in the literature:
µN = 5.05083(4) x 10-27
JT-1 [9]
µB = 9.27410(7) x 10-24
JT-1 [9]
gJ = 2.002332(2) [6]
h = 6.6260(5) x 10-34
Js [9]
∆E/h = 3085.732439(5) MHz [8]
gI = -5.39155(2) [7]
Using (II), we obtain within the hyperfine state F = 3 six transitions
which have frequencies
( )
ξ+ξ−+−
ξ+ξ++
×−==
212
F
212
F
3F
1m32
1m32
1MHz866.1542
mTB
MHz004199.0f
Page 23 of 24
and within the hyperfine state F = 2 four transitions with the
frequencies
( )
ξ+ξ−+−
ξ+ξ++
×==
212
F
212
F
2F
1m32
1m32
1MHz866.1542
mTB
MHz004199.0f
where mTB
10)9(08341.9 3 ××=ξ −
No. fM(σ+) [MHz] fM(σ-) [MHz] f [MHz] F mF(1)↔mF(2)
1 3 3 ↔ 2
2 3 2 ↔ 1
3 2 2 ↔ 1
4 3 1 ↔ 0
5 2 1 ↔ 0
6 3 0 ↔-1
7 2 0 ↔-1
8 3 -1↔-2
9 2 -1↔-2
10 3 -2↔-3
Table 5: 85Rb transition frequencies at B = __________ mT
No. U(σ+)
[mV]
U(σ-)
[mV]
mF(1)↔mF(2) No. U(σ+)
[mV]
U(σ-)
[mV]
mF(1)↔mF(2)
1 3 ↔ 2 3 2 ↔ 1
2 2 ↔ 1 5 1 ↔ 0
4 1 ↔ 0 7 0 ↔-1
6 0 ↔-1 9 -1↔-2
8 -1↔-2
10 -2↔-3
Table 6: 85Rb transition intensities at B = __________ mT.
left side: F = 3, right side: F = 2
Page 24 of 24
Literature
[1] A. Kastler: Journal de Physique, 11 (1950) 255
[2] H.Kopfermann: Über optisches Pumpen an Gasen, Sitzungsberlchte der
Heidelberger Akademie der Wissensschaften, Jahrgang 1960, 3.
Abhandlung
[3] J. Recht, W. Klein: LH-Contact 1 (1991) pp. 8-11
[4] G. Breit, I. Rabi: Phys. Rev. 38 (1931) 2002
[5] Kopfermann, H.: Kernmomente, 2nd ed., pp. 28ff
[6] S. Penselin: Z. Physik 200 (1967) 467
[7] C.W. White et al.: Phys. Rev. 174 (1968) 23
[8] G. H. Fuller et al. : Nuclear Data Tables A5 (1969) 523
[9] B. N. Taylor et al.: Rev. Mod. Phys. 41 (1969) 375