REVIEWS OF MODERN Puvsrcs
NUMBER 2 APRIL, 1958
Isotope Shift Studies of Nuclei
Peren Busx, Jn fr Technische Kernphysit, Technische Hochschule Darmstadt, Germony
ax
Hans Kovrnaaav, Hrsts Pipsikalicies Fst, Universitat Heidelberg, Gomany
1, SUMMARY OF EXPERIMENTAL RESULTS
‘HE, information about size and angular shape of
atomic nuclei that is available from isotope shifts
in the optical spectra of heavy elements can be con-
ddensed into one experimental number for each pair of
isotopes. This is due to the fact that the shifts at-
tributable to the field effec of valence s electrons are—
within the theoretical framework and experimental
accuracy available at present—proportional to the
charge densities of these electrons at the nucleus. An
isotope shift constant?* Crap can be defined, therefore, as
Cop AT o/(We(0)rau?/Z), ay
where AT, isthe level shift due to an s electron, ¥u(0)
is the unrelativistie wave function of that electron at
the origin, ax is the Bohr radius, and Z the atomic
number. The quantity ¥."(0) can be evaluated either
from the magnetic hyperfine structure due to the $
electron, or from the level scheme of the respective
atom orion. For elements with more than two isotopes
is usually convenient to measure Cay for one selected
pair, and relative isotope shifts for the others.
‘Tsotope shift constants have so far been evaluated for
the region 36<2<88 only. Figure I shows the data for
isotopie pairs with an even number of neutrons. Apart
from a factor 61, considered later, the ordinate isthe
ratio Cay/Ca. Cu is the isotope shift constant caleu-
lated for a standard model of spherical, homogeneously
charged, incompressible nuclel with radius ro propor-
tional to 41, used only for the purpose of comparison.
‘The ratio Co/Cuy shows how the quantity
cae frei, ®
with
a2 (a=e/he), @
increases in going from mass number Ay to z= Art,
compared with the change
5(0™ jaa (2o/20-43)(6A/A)ra* a
for the standard model, For nuclei which are not
spherically symmetric, the nuclear charge density ep
will be a function of angle as well as radius.
‘The strong variations with neutron number of the
quantity plotted in Fig. 1 can be accounted for in
LB Bee and H. Koperiann, 2. Phys 126 34 100)
2 P. Bris and H. Koptermann, Festschr- Akad. Wiss. GBttngen,
MathPiys KL, 17 931),
terms of changes in nuclear deformations, The isotope
shifts, therefore, provide a sensitive measure of the
change of intrinsic quadrupole moments of nuclei, The
data show furthermore that addition of neutrons to a
nucleus changes the charge radius less than given by the
A¥ law, This can be understood by assuming a nuclear
compressibility
Several recent reviews!® exist of the theory and
interpretation of isotope shifts, to which the reader is
referted for a detailed discussion. A complete literature
survey up to July, 1956, is given by Mack and Arroe*
‘The published compilation of experimental isotope shift
constants, however, is out of date now, Therefore, we
present in Table I the numerical values and the litera-
ture used in the preparation of Fig. 1.
With the gross features of nuclear electric radii now
reliably known from other sources, the isotope shifts
should provide the most sensitive measure available of
the change of nuclear charge distribution from one
isotope to another. There are, unfortunately, several
dificulties involved in obtaining the values of &(7")
from the spectroscopic measurements.
(a) Uncertainties Attributable to Cix
Calculations should include effects of the intrinsic
magnetic moment of the electron and of a possible
polarization of nuclei by atomic electrons, as has been
pointed out by Breit? and co-workers,
(b) Uncertainties Attributable to Cexy
‘The reliable experimental determination of AT,
¥4(0), and of the mass dependent isotope shifts present
other problems, which Breit? was the first to discuss,
Tn Sec. 2 we sketch how Carp has been derived in a
typical case, The element Gd is chosen both because
of its complex spectrum and its position in the interest=
ing region of strongly deformed nuclei
3K, W. Ford and D. L, Hil, Ann, Rev, Nuclear Si. §, 25
«385 ee
“TAR. Stganoy and Yu. P.Dontsov, Uspekhi Fs, Nauk 55,
315 (1955). ws
Jp Mack and H. Aros, Ann. Rev. Nuclear Sci, 6, 117
a3)
HL Kopfermann, Kerwnomene (Akadeninche Vrlgsgesl
sch Franke, 1380); Nadar Moments (Reet Pres Tae
Noe Yor 938)
Ser Bre Opt Soe. Am. 47,446 1950)
aie et aa hae iin within Aton
Nl, Euelopaioof Posie (pingerVerag, Bet, 197),
gE, hg ts of Poss Cosogervere
*G. Brit, Phys, Rev. 42, 48 (1932),
st7S18
P, BRIX AND H. KOPFERMANN
» T T T
bserved isotope shift
Stondord ‘sotope shill Cn
(Gm. with Re =(20:10-%em)
2s +
20 — |
| |
15 —
re |
a5 : SAY 7
oe | {
ae eee
™ ‘|
° s |
“ 50 60 70 20 90 m0 «0S«ROSCBO
—
Fic. 1, Ratio of observed iotope shit to standard shift for uniform density nuclel
‘number i, and ae placed at the higher N value for each pai of isotopes Pll ees indi
‘Rb ave vugh estimates only
2, TYPICAL EVALUATION OF Cup FOR A
‘COMPLEX SPECTRUM
‘The spectra of Gd and Gd 1 have been very care-
fully analyzed by Russell? who assigned electron con-
figurations to many levels of the neutral and singly
ionized atom. With samples of natural isotopic com-
position the isotope shifts are most easily measured
for the pair Gd"™.Ga, The numberof lines studied by
diferent authors" # amounts to about 120, From these
the shifts of about $0 Gar levels and 35 Gd 11 levels
have been deduced. A’ discussion of perturbations
finally leads to the shifts of presumably unperturbed
configurations presented in Fig. 2
Al data of Fig. 2 can be described as follows: a shift
of about 74X10" emis assigned to the 6s electron
in the configuration 4,[6s of Ga 1. The other valence
electrons reduce this isotope shift by reducing the 6s
wave function at the nucleus; these screening effects
amount to ~20% for a 8d oF Gs electron and ~10%
for a 6p electron. A comparison with screening effets
HHL N, Rusel J. Opt. Soc, Am, 40, $80 (1950),
NP! Bris and HD, Engler, Z. Paya 188,862 (1952),
BS" Suway J. Phys, Soe Japan 8, $77 (193).
Bie, 2. Phyl 183, 519 (1952)
MD. Bris and KIL, Lindenberge, 2. Physik U1, 1 1989)
RA), Points are plotted sesus neutron
ce Ua Zeal The vals piven for Sand
of the valence electrons in Hg,*#* included in Fig. 2,
shows good numerical agreement. It is concluded that
Setecrn éscatetn
cap am ap cae
a a ee ee
Bl LS] Le] [ee]
I
&} (#] Le] on
T 4
e YY x
soe? a
eS em eeeecon
2 Top shit of GG and sre eta. The
eens ah are even nthe Squares, with ao ats
Stblgned to levels of S86p and 5d6p, respect ‘are pre:
fmably unperturbed. Corrections for tonfguraion iteration
have been made, i neceary Prom the ration of the sie the
scr ctor ven atthe en ofthe ennatog line have
bjeen obfalned, ‘The corresponding factors for Hg are shown fh
brackets. The broken lines indeate an extraplation for the
Minsereened™ Ge electron in 47s. The quantum numbers Ste
always 5d 6, and 6p.
J, Blas, thesis, Paris (1957),ISOTOPE SHIFT STUDIES
Tame 1
519
Toto sie commune Cn (uth = AVKL20XI0-% cm, an Cag The fro exp i she text cstomaly
The Cox values not given directly can be obtained from the
Telative sits,
oa rs a Hue
oa, mG iiss woe tate
* “om Re 75 185-187 io
om bay ht
sar a he ” 130830 elie)
alee 1m ipee tat yy
gay a be m0 gag toate
wiass 10D bea ae
ig nw Pare
100-40 s ‘Pb 82 204-206 124 524 O89 a belee
} MBM) 2 ae BR EB owe fh
BST nahied prams ‘edt a nec. ae aie they i, ois
Gn Roribiceniian le TRE AE GUO By as nan sae.
Ee Seva cae nhe merce adigaytone ah. Ne
"PRESET I ungpeg tyke on
Tee Nekowski Z Physik 14, 284 (1955),
ays, IE. Krebs and H. Neluown, 2, Puyalk 148, 849 (1980)
ee Sal ae ath Bay Ech teks ba hg
Hee eee eee anno, qui bar anl eal NAL Bue a. a0
AE Bs, rar 99,623 955, ze Begs ohl's "ates {Opos e Sasas ,a29 aos,
“MI Kelly snd JH, Sitherland, Can. j, Phys, 34 $21 (1956), SEG Fonts, J Ont, S90 An, 44,88 (198,
PRES OTERO EOE os TEER UO dah ces aan,
ZW. R. Hindmarsh and #1, Kahn, Proc. Phys. Soc. (London) A68, 433 J. K.'Holmes and F- MeCtana Boe. Aim. 47,297 (1957).
oo Coma ass Rr RA ON
all observed Gd isotope shifts can be understood on the
basis of a nuclear field effect and that, therefore, Eq,
(1) is applicable.
Only empirical methods'* were available at first to
evaluate ¥.'(0) for the 56s and the Sd6r configura
tions, With these the isotope shift constants 134% 10
mr and 120X10~ cm, respectively, were obtained
from the indetined values of Fig. 2, Recently the
magnetic hyperfine structure splitting constant a, of
Gal for 546, a8 well a8 the ratio of the magnetic
‘P. Brix and H, Prank, Z, Physik 127,289 (1950)
moments (Gd"™)/u(Eu"") have been measured.”."*
‘These data in connection with the isotope shift con-
stant of Eu give ~108X10 cm as a third value, not
dependent upon an empirical evaluation of y(0) for
Gd, and in agreement with the formerly published
average
Coxo(Ga**— Ga) = (125-420) 10-* em,
‘The factor 8 indicates that a further uncertainty exists
due to a possible contribution to the isotope shift of
ED. B. Speck,
Rev, 101, 1725 (1986)
BW, Low, Phys
fev. 103, 1309 (1956).520
the inner closed shells of the Gd atom: the wave func-
tions of the 1, 2s, «++ shells could possibly be
affected by changes in electron configuration of the
valence electrons. Crawford and Schawlow'* and later
Humbach® have tried to calculate and for Hg the
result seems to be 8=1. For Gd rr the only estimate
available is from the isoelectronic Eu t spectrum, where
the isotope shift of 6p, electrons has been used" to
get the empirical estimate $= 1.
‘Mass dependent effects have been neglected. ‘They
are assumed to be unimportant for Gd for theoretical
reasons, but no quantitative values can be calculated
at present. The fact that the nonequidistant relative
3 ML F. Crawford and A. L, Schawlow, Phys. Rev. 76, 1310
(198)
SW. Humbach, 2, Physk 188, S89 (1952); 141,59 (1955)
P, BRIX AND H, KOPFERMANN
tope shifts of Ga" and other rare earth elements
are constant within experimental error is the most con-
clusive argument against appreciable mass effects.
‘We believe that the error given above for 8Cxp is
conservative and that Cap can be assumed to give a
fir estimate of Cup with an accuracy of about 20%,
This uncertainty is typical for all isotope shift con-
stants and should not be overlooked. The relative
shifts, on the other hand, are usually much more
accurate, with uncertainties of a few percent or less in
favorable cases,
We are indebted to Dr. G. Naldeke for his help in
bringing Table I up to date.
' Kopfermann, Krlger, and Steuel, Ann. Physik (6), 20, 258
95h