0198506961.pdf

6.4 Measurement of hyperfine structure 113

Even isotopes (no HFS)

Lines

from

even

isotopes

0 10 20 30

Transmitted

intensity

Frequency (GHz)

(a)

(b)

Fig. 6.11(a) The 5s6s^3 S 1 level in the

odd isotopes of cadmium (^111 Cd and

(^113) Cd) has a large hyperfine structure;
theF =3/2 hyperfine level lies be-
lowF =1/2 because the nuclearg-
factor is negative and similar in size
for both odd isotopes. (The total mag-
netic field created by the unpaired s-
electrons at the nucleus is anti-parallel
toJ, as in the ground state of hydro-
gen and the alkalis.) (b) An exper-
imental trace of the hyperfine struc-
ture for the 5s5p^3 P 0 –5s6s^3 S 1 line
at a wavelength of 468 nm obtained
with a pressure-scanned Fabry–Perot
́etalon (as in Fig. 1.7). There are three
peaks in each order of the ́etalon: a
large peak from the isotopes with no
nuclear spinI= 0 and hence no hyper-
fine structure (generally isotopes with
an even number of nucleons so there are
no unpaired spins within the nucleus);
and two smaller peaks whose separa-
tion equals the hyperfine splitting of
the 5s6s^3 S 1 level. (Data from the Ox-
ford Physics Teaching Laboratory prac-
tical course; further details are in Lewis
(1977).)
can be checked by verifying that the observed peaks have the expected
displacements from their centre of gravity (which is approximately at
the position of the even isotopes).^3333 This is not straightforward in this ex-
ample because of the overlap of lines,
but this could be done using a curve-
fitting program on a computer. Addi-
tionally, there is a sum rule for the in-
tensities as in fine structure.
It can be seen from Fig. 6.11 that optical spectroscopy is not generally
suitable for measuring the Zeeman effect of hyperfine structure because
spectral lines have a Doppler broadening that can be comparable with
the hyperfine splitting. Thus in the low-field regime (μBB<A)the
Zeeman splitting is too small to be resolved. To show this quantitatively