T

Fionn Fitzmaurice
[email protected]
8.12.2009
H
Z
E
E
E
E
F
M
F
A
E
N
C

T




























School of Physics, Trinity College Dublin

The Zeeman Effect

Fionn Fitzmaurice
[email protected]
08.12.2009






Abstract

The Zeeman effect, a phenomenon in quantum mechanics, was studied in the course
of this experiment. It is the lifting of degeneracy of spectral lines through the
application of a magnetic field.
In this discourse, the wavelength of the cadmium red line was determined to
be
0
= 649.7 nm and the Bohr magneton determined to be
B
= 9.274 10
-24
JT
-1
.
The split spectral lines were discovered to be linearly polarised in the transverse
orientation of the magnetic field and circularly polarised in the longitudinal
orientation.



























Introduction

Pieter Zeeman (25 May 1865 9 October 1943) shared the Nobel Prise in 1902 for
his discovery of the effect whos name lends itself to this lab report. The Zeeman
Effect is the splitting of otherwise degenerate spectral lines in the presence of a
magnetic field. This occurs via interaction between the external magnetic field and the
atomic magnetic moment. The atomic magnetic moment is related to the total angular
momentum, which is given by the vector
operator equation

J
i
= L
i
+ S
i

where the total orbital angular momentum

L

= r p

where L

, r and p are position and linear

momentum operators,

r = r , V =
i

p .

The operator L

obeys the canonical commutation relation for angular momentum

[L
i
, L
j
] = i
ijk
L
k


where
ijk
is the antisymmetric Levi-Civita tensor.
S
i
is the total spin angular momentum, for which the angular momentum
canonical commutation relation still holds. This criterion also holds for the total
angular momentum. Both the orbital and spin operators have eigenvalue equations

L
2
| =
2
l (l + 1) |

S
2
| =
2
s (s + 1) |

for some eigenfunction |. The value of l may only be integer, while s may take on
integer and half-integer values.


Part 1. To determine the wavelength of the cadmium red line

From this point on, errors will be denoted by the lower case rather than to avoid a
clash of notation.
The objective was to measure the wavelength of the cadmium red line in zero
magnetic field, this wavelength denoted by
0
.


|
|
.
|

\
|
=
2 2
2
2
1 2
f n
r
nd m
m

A visualisation of the Levi-
Civita tensor, where red
represents 1, blue; 1 and
empty cubes represent 0.

Therefore,
0
can be calculated by measuring the ring radius, r
m
, as a function of the
ring order, m, since n, d, f are known quantities.
The magnet and lamp assembly was rotated such that its orientation was
appropriate, with the two electromagnets facing the detector, rather than in line with it.
The lenses L
1
and L
2
were adjusted in order to properly focus the image. This was the
effect of moving the two lenses. Once the ring pattern was sharp, uniformly bright and
centered, the eyepiece was replaced by 2048 pixel CCD. 2048 pixels is the maximum
resolution for the CCD and was therefore the resolution used during the experiment,
as it provided the highest level of accuracy.
r
m
was determined for the first 10 rings using computer software in tandem
with the CCD. This was possible as the software gives a value for the pixel number,
which I will denote N. This can be related to the separation, x, of two points from
the pixel number difference, N, via the simple relation

x = N 14 m.

Results and Analysis

The values of r
m
2
and m were plotted against each other, see graph 1.1 r
m
2
vs. m
appended. The slope of this graph,

s
L
= 5.3247 10
-6


and


0
2
=
L
s
nf
d


and therefore, using the values for d, n, f, we arrive at


0
3
) 0225 . 0 ( 457 . 1
) 10 4 (
=

L
s


0
= 649.7 nm

The accepted value is
0
= 643.85 nm. This is well within experimental error, as these
values differ by a factor of 1.01.


Part 2. The transverse normal Zeeman effect

The objective of experiment 2 was to measure the wavelength shift, of the outer
triplet lines and show the relation B, to determine the value of the Bohr
magneton
B
and to determine the polarisation state of each of the triplet lines.


2 2
f n
r rA
=
A

This shows that can be calculated by measuring r and r, since , n, f are known
quantities.


hc
B
B

2
0
= A

From this we can therefore determine experimentally a
value for
B
.
The magnet and lamp assembly was left in its
position as per experiment 1. The ring pattern was
viewed through the eyepiece as the field was increased to
7 A, which was not exceeded. At high currents the
electromagnetic coils heat and this changes their
resistance. In order to compensate for this and keep the
current (and by extension field) constant, the voltage
must increase in tandem. This can be set to happen automatically by setting the power
supply to constant current mode. This will only operate effectively at currents
below 7 A, hence the operational upper limit.
The eyepiece was then replaced with the VideoCom CCD and the fringes
sharpened. Intensity distributions for I = 4, 5, 6, 7 A were recorded, as were the
corresponding values of r and r, up to about 6 rings.
Finally the polarisation state of each spectral line was determined using the
linear polariser. This was done by looking through the eyepiece while rotating the
polariser by a known angle and noting the change in intensity in each line.

Results and Analysis

The magnetic field varies with current, but this variation is non-linear, because of
remnant fields. For I = 0, B 0. The following table contains values for I and B.


I (A) B (mT) for I B (mT) for I
4 394 407
5 485 493
6 552 555
7 594 594

The values for r, r and rr are tabulated below for different values of I.

I = 4 A
r (m) r (m) rr (m
2
)
0.003332 0.00039 1.29948 10
-6

0.005782 0.000225 1.30095 10
-6

0.007462 0.00018 1.34316 10
-6

0.008806 0.00015 1.3209 10
-6

0.009954 0.0001275 1.26914 10
-6


I = 5 A
r (m) r (m) rr (m
2
)
0.003374 0.0004875 1.64483 10
-6

A photograph P.
Zeeman took of his
eponymous effect.
0.00581 0.000285 1.65585 10
-6

0.007462 0.0002175 1.62299 10
-6

0.008806 0.0001875 1.65113 10
-6

0.009982 0.000165 1.64703 10
-6


I = 6 A
r (m) r (m) rr (m
2
)
0.003374 0.00054 1.82196 10
-6

0.005782 0.0003075 1.77797 10
-6

0.007462 0.0002325 1.73492 10
-6

0.008806 0.0002025 1.78322 10
-6

0.009982 0.000165 1.64703 10
-6


I = 7 A
r (m) r (m) rr (m
2
)
0.003374 0.0006 2.0244 10
-6

0.00581 0.000345 2.00445 10
-6

0.007462 0.000255 1.90281 10
-6

0.008806 0.0002175 1.91531 10
-6

0.009982 0.00021 2.09622 10
-6

r = 10
-5
m, r = 10
-5
m
rr = 2r 10
-5
m
2


The mean value of rr for each current is tabulated below:

I (A) Average (rr) (m
2
)
4 1.30673 10
-6

5 1.64436 10
-6

6 1.75302 10
-6

7 1.98864 10
-6

I = 0.01 A
rr = 2r 10
-5
m
2


The graph of these averages versus the magnetic field was plotted, see graph 2.1 rr
vs. B appended, showing a direct proportionality between and B. The slope of this
graph is

s
L
= 3 10
-6


and


B
L
f n
hc s

=
0
2 2



B
= 1.92 10
-23
JT
-1
1 10
-6
JT
-1


The recognised value of the Bohr magneton is
B
= 9.274 10
-24
JT
-1
. The
experimental value deviates from the expected value by a factor of 2.07.
The polarisation state for each of the triplet lines was examined. It was found
that the interior and exterior rings had the same polarisation as each other, observed
when the polariser had rotated 0 radians. The ring between these two rings was found
to be polarised by radians with respect to the other rings, i.e. the polariser was
rotated by in order to achieve a maximum in intensity for this ring. The quarter-
wave plate was not required to make these judgements.


Part 3. Longitudinal normal Zeeman effect

The objectives of experiment 3 were to determine the change, , in , to compare
this with the change in wavelength for the transverse Zeeman effect explored in
experiment 2 and to once again determine the polarisation state of the spectral lines,
this time a doublet.
The magnet and lamp assembly was rotated by such that the light was
viewed in the direction of the field. The current was set to 5 A and 7 A and in each
case the wavelength shift in the doublet lines was measured and compared with the
transverse orientation.
The polarisation state of each of the doublet lines was also determined, using a
linear polariser and a quarter-wave plate.

Results and Analysis

At I = 5 A and I = 7 A, the following values were measured in a similar manner as
before.

I (A) Average (rr) (m
2
)
5 2.25 10
-6

7 1.63 10
-6

I = 0.01 A
rr = 2r 10
-5
m
2


As before, from


2 2
f n
r rA
=
A

we can determine the value of , which can be compared to the value of in the
transverse case. In each case, the values of rr should remain roughly constant, to
within experimental error. Unfortunately, this is not the case, particularly with the
results in part 3. I am unable to explain this large deviation; I believe it must boil
down to human error.
Taking the average of the average of the values of rr in part 2, the value of
which, denoted
2
= 1.67 10
-6
m
2
, and discounting the value of rr obtained for
I = 5 A in part 3 due to the obvious inaccuracy of the value, such that for part 3,

3
= 1.63 10
-6
m
2
, the ratio of which being the dimensionless quantity

. 98 . 0
2
3
=

Therefore the values of in each case are in 98 % agreement.
Upon introducing the linear polariser, the intensity of both the lines reduced.
However, upon rotation, this intensity remained unchanged and uniform between the
lines, indicating that the lines were not linearly polarised.
The quarter-wave plate was then placed between the source and the linear
polariser. In this setup, with both polarisers in their initial positions (i.e. zero rotation),
the exterior ring exhibited a maximum intensity, whilst the interior exhibited a
minimum.
Upon rotating either polariser (the quarter-wave plate or the linear polariser)
by , this relation was reversed such that the inner ring exhibited maximum
intensity and the exterior, a maximum. This implies that this doublet is circularly
polarised and the polarisation states for each ring are orthogonal to each other.


Conclusion

The longitudinal and transverse Zeeman effect was observed in this experiment. It
was found that the splitting of energy levels was directly proportional to the intensity
of the magnetic field. The wavelength of the cadmium red line was experimentally
measured, as was the value of the Bohr magneton. Polarisation states for the split
spectral lines were also determined experimentally for different orientations of the
magnetic field, and this orientation was seen to determine the type of polarisation
occurring.























References

1) Zeeman, P. (11 February 1897). The Effect of Magnetisation on the Nature of Light
Emitted by a Substance. Nature 55: 347

2) Spencer, J. Brookes (1970). On the Varieties of Nineteenth-Century Magneto-Optical
Discovery. Isis 61 (1): 3451.