Z. Physik 269, 245-252 (1974) 9 by Springer-Verlag 1974

Isotope Shift Measurements in the Atomic Spectrum of Lanthanum (LaI)* W. Fischer, H. Hfihnermann, and K. Mandrek Fachbereich Physik der Philipps-Universit~it Marburg Received April 29, 1974

Abstract. Using the reactor-produced lanthanum isotope l~7La, the highly enriched rare 138Tn57,~and the stable 1391-5 w a , the isotope shift has been measured in five lines of the La Ispectrum with the aid of a pressure-scanned Fabry-Pdrot interferometer. The isotope shift data obtained show surprisingly large specific mass effects, arising from configuration mixing of levels with a 4 f electron involved. The changes in mean square charge radius (r 2) of these nuclei extracted from the experimental isotope shift co n st an t s Cexp are compared with the corresponding values for the isotonic barium nuclei, where similar anomalies in the isotope shifts occur.

1. Introduction

2. Experimental Aspects

The measurement of isotope shifts (IS) in atomic spectra can give information on proton charge distribution as a function of neutron number [1]. In complicated spectra, however, the accuracy of this method is limited by the uncertainty in the calculation of electron densities at the nucleus, and the classification and the purity of the electronic levels.

2.1. Excitation of the Spectral Lines

Within the last years, considerable progress has been made in the study of the IS near the magic neutron number N= 82. Whereas precise measurements have been done in the case of xenon (Z=54) [2] cesium (Z= 55) [3], barium (Z= 56) [4], and cerium (Z= 58) [5] no measurements of the IS in the spectrum of lanthanum (Z=57) have been performed up to the present time. There exist two naturally occurring La isotopes: the stable 139La and the primordial radioactive 138La (half-life T~/2= 1.1 9 1011 a). Because of the very low natural isotopic abundance c~(138La)= 0.089% it is necessary to use enriched samples for IS measurements. Another well suited isotope is the long-lived reactor produced 137La with a half-life T1/2=6.104 a. 9 Partly reported at the "Frtihjahrstagung der D P G ' , Mtinchen 1973.

The lanthanum spectrum was excited in a hollow cathode discharge. Normally, neon, argon, krypton or xenon are the carrier gases. With helium, the spectrum of which has only few lines, no sufficient excitation could be obtained. Because disturbing lines of the other inert gases were perceptible especially when only a few micrograms of lanthanum are in the hollow cathode, the measurements in the line LaI, 2=6250 • had to be carried out with xenon. When xenon was used, the hollow cathode lamp could not as usual be cooled with liquid nitrogen in order to reduce the Doppler width of the spectral lines. In this case a cooling mixture of dry ice and methanol was utilized.

2.2. Recording of the Hyperfine Structure To isolate the spectral line under investigation, a plane grating monochromator was used. The hyperfine structure (hfs) was resolved with the aid of a pressure scanned Fabry-P6rot interferometer using Freon 12 as a scanning gas, the large refractive index of which permitted the interference orders to be scanned four times more in a single run than is the case for the

Z. Physik 269 (1974)

246

formerly used nitrogen [6]. A sensitive photomultiplier, cooled by a Peltier element in the case of the RCA type C 7164 R and the liquid nitrogen cooled RCAtype 931A, served to detect the red and green spectral lines respectively. The signals were then amplified by a dc-amplifier (Keithley-Picoammeter), smoothed with a RC-unit of time constants Zgc= 1 ... 2 s and then digitized by a digital voltmeter (type Solartron) and recorded on paper tape and chart recorder simultaneously. The interference orders were superposed by means of a C II C 90-40 computer. The line shape of the curves was assumed to be a convolution of a Gaussian with an Airy-function, disturbed by an exponential function because of the RC-unit. 2.3. Preparation o f the Hollow Cathodes

2.3.1. Preparation of the Hollow Cathodes with Natural Lanthanum Natural lanthanum (isotopic abundance ci(139La)= 99.911%) with a chemical purity of 99.5 % is supplied in metallic form from FLUKA Laboratories, BuchsSwitzerland. The impurities consist chiefly of praseodymium and cerium and did not disturb the investigated lanthanum lines. About 10 mg of metallic lanthanum were put into the hollow cathode of 3 mm diameter and 15 mm depth. The hollow cathode itself consisted of very pure copper or aluminium. With one filling the lanthanum spectrum could be excited for at least 100 hours. 2.3.2. Production and Preparation of 137La The radioactive isotope 137La can be produced in a reactor by the following reaction 136 ssCe(n,

7) 137Ce 58 e 1 ~ t

.....

+ 137157~i:t" pture

As starting material lanthanum free natural cerium oxide was used which contains only 0.193 % 136CeO2. This sample had been irradiated in the Karlsruhe reactor for several months and the produced 137La had then been separated chemically from the surplus cerium [-7]*. The ~37La-sample was converted into chloride by dissolving the oxide in hydrochloric acid and then put into the interior of a hollow cathode lamp and evaporated in a drying chamber at 80 ~ Preparatory experiments have shown that in this case a hollow cathode made of copper has to be preferred to an aluminium one because of excitation and longtime stability of the spectral lines. A mass spectrometric determination of this sample, which H. Ihle, KFA Jtilich, carried out after the experiments had been * E. Gerdau and H.Winkler from II. Physikalisches Staatsinstitut in Hamburg have placed about 100 gg of 137LazO3 at our disposal.

concluded, yielded the values: Cy(137La)=76.4(3)% and cf(139La)= 23.6 (3) %. 2.3.3. Filling of the Hollow Cathode with 138La The optical investigation of 138La requires a high enrichment of this isotope because of the low natural isotopic abundance of c~(138La) = 0.089 %. Preparatory experiments with low enriched material from Oak Ridge National Laboratories and theoretical considerations showed that at least an enrichment of 30 % is necessary. The enrichment was performed with the electromagnetic isotope separator SIDONIE II at Jtilich [8]. In order to avoid difficulties with the chemical reprocessing, the separated 138La was deposited directly in the hollow cathodes. Earlier investigations of Martin and Walcher on separated silver and copper [9], and Hiihnermann and Wagner on cesium [10] have shown that the hollow cathodes prepared in such a way gave an intense and stable excitation of the spectral lines even with low quantities of deposited material. The ion beam of the Jiilich separator has at the collector a width of about 1 mm and length of 30 mm. Therefore five hollow cathodes were placed one upon another and filled simultaneously. The hollow cathodes were made of aluminium and had a cavity of 3 mm ~ . In order to prevent self sputtering during the separation the beam was reduced to a height of 2mm by a molybdenum diaphragm. The total duration of the separation was 20 hours, by which time about 10 l-tg of 138La had been collected in each hollow cathode. The feed material for the ion source was natural anhydrous LaC13 [-11]. A mass spectrometric analysis of the deposited lanthanum in the hollow cathodes performed after the optical experiments yielded cs(38La)= 69.0(5)%.

3. Measurements of the Isotope Shifts The spectrum of La I was classified by Russell and Meggers in 1932 with spectroscopic methods [12]. Recently, some odd levels of the configuration 5 d 6 s 4f have been reclassified by combining spectroscopic data with thermodynamic ones [,13]. The hyperfine structure of numerous lines in the La I spectrum has been carefully investigated by Liihrs [14] with the aid of a Fabry-P6rot interferometer. Some of the Aand B-values of the electronic levels have been remeasured by the highly accurate methods of radiofrequency spectroscopy. The hfs splitting values of the ground states 139La 5 d 6s z a ~ D5/2,3/2 have been measured by Ting [15] and all levels of the configuration 5d26s of 139LaI below 9000cm -1 by

W. Fischer et al.: Isotope Shift Measurements

247

Iexp.

,i .~ r! !i i:r r,r~ft

t'~'[ii

!

;

tOOmK

,

J

l

.~!~i!!1i !

',.._

il

*

I,

9

~

v

Childs and Goodman [16], who used the atomicbeam-magnetic-resonance technique. The level crossing method has been used by Hese [17, 18] for the determination of the splitting values of four levels of the configuration 5d6s6p of L a I (see also [-19]). However, the half of the A- and B-values of 139La necessary for the evaluation of the five lines considered in this paper have not yet been measured with radiofrequency methods. A determination of these values was necessary and was performed with a photoelectrically recording Fabry-P6rot interferometer. Other necessary parameters for the evaluation of the IS from the observed hyperfine structures are the spins and nuclear moments of the three isotopes which are considered here. The unknown spins and moments have been determined and are published in preceding papers [20, 21]. The isotopes ~37La and 139La have the same nuclear spin I = 7/2 and nearly the same magnetic dipole moment; therefore, the hyperfine structure of spectral lines of both isotopes is quite similar. Since the isotope shift is less than the line width the spectrum of each isotope was excited in a different hollow cathode discharge. Two discharges were run simultaneously, and a movement of a reflecting prism enabled light from the hollow cathodes to enter the optical system in alternate orders of interference of the Fabry-P6rot interferometer, which was scanned continuously throughout. The hollow cathodes were of the same type and run under as nearly identical conditions as possible in order to avoid possibly occurring relative pressure and Stark-effect shifts [22]. Nevertheless, control measurements for the IS of 139La against 139La disclosed systematic errors up to 1 inK,

,J

,..

Fig. 1. Determination of the isotope shift of 137La-139La in the line La I, 2 = 6 250 ~ with the aid of two hollow cathode discharges. The hfs pattern on the right was recorded with natural 139La, whereas the pattern on the left has been detected from a 13VLasample. The component on the far left side is the main component of the 139La-structure shifted twice the free spectral range of the interferometer

which have to be taken into account in the IS measurements of 137,139La discussed later on. Fig. 1 demonstrates the measurement of the IS in the line La I 2 = 6 250 h. The hfs pattern on the right was recorded with natural 139La, whereas the pattern on the left has been detected from the 137La_sample with the dc-amplifier made 30 times more sensitive. The dark current of the photomultiplier raises therefore the base line. The component on the far left side is the main component of the 139La-structure shifted twice the free spectral range of the interferometer. In the case of the laSLa sample all the measurements could be done with one hollow cathode as the hfs pattern showed well separated components of 138La and 139La. This is due to the different spins (I(138La)=5) and magnetic moments, which make these IS measurements more reliable. Fig. 2 shows the measured hfs pattern (A) of the transition 138'139La 2=6250/~. The hatched hfspattern (B), however, results from an independent measurement with natural 139La in order to give an impression of the hfs-positions for the two isotopes and of the high enrichment attained. The IS in the LaI-spectrum could be measured in five different lines. The wavelengths, the corresponding upper and lower electronic levels and the shifts are listed in Table 1. The IS values vary greatly and show a large odd-even staggering. This means that the isotopic positions v(N) of an isotope with an odd number N of neutrons is shifted towards the position of the lighter isotope with N-1 neutrons. In the case of xenon [2], cesium [3], and barium [4], as well as in the case of lanthanum, this effect is so pronounced that a reversion of the isotope positions can occur.

248

Z. Physik 269 (t974)

A

P". .: i 9 [

r., ", ,.

. . . .

/i

A

/ / ,!

!.i' ~:>.. t

f 9

J.A..~ I

Y/!~-

'.2X!

e~U

B

/

V

i

Fig. 2. Hyperfine structure of the line La I, 2 = 6 2 5 0 A measured with an enriched ~38La sample (curve A). The hatched pattern (curve B) results from an independent measurement with natural 139La in order to give an 9impression of the hfs-positions of the two isotopes

100 rnK

Table 1. Measured isotope shifts Av,s of some spectral lines of LaI. (The normal mass effect AVN~s is also listed.) The errors given here are only statistical ones and do not include possibly occurring systematic deviations (1 m K = 10- 3 c m - ' ) Wavelength

Transition

AVNMs/AN. m K

I3SAa39hs/mK

137A139V~s/mK

5177

5d26py4D~

0.55

-8.0(3)

-4.9(8)

5 234

5 d 6 s 4 f a GO/2_ 5 d 2 6 s a 4F9/2

0.54

- 5.2 (3)

+ 6.6 (8)

5 455

5 d 6 s 6 p y 2D~ - 5 d 6 s z a 2D s/2

0.52

- 13.3 (4)

- 4.2 (6)

6 250

5 d z 6 p z 4G01/2 - 5 d 2 6 s a 4F9/2

0.45

- 8.4 (3)

- 2.9 (6)

6266

5 d Z 6 p z 2 H u0 / 2 - 5 d

0.45

-4.5(3)

+5.6(6)

2

6saZGg/2

4. Theoretical Aspects of Isotope Shift The isotope shifts A hs in a spectral line of the frequency v between two isotopes (1) and (2) can be expressed as a sum of three terms [30]: the normal mass shift A VNMs and the specific shift A VsMs due to the different nuclear masses and the volume or field shift A Vvs due to changes in the energy of the electrons in the electrostatic field of the nucleus when the number of the neutrons is altered. Whereas the normal mass shift (NMS) can easily be evaluated (see e.g. [1]) and amounts in our cases to about 0 . 5 m K - A N the specific mass shift (SMS) cannot be determined so simply (AN: difference of the neutron numbers). It can however be calculated by the so-called "Vinti-integrals" from one electron wave functions [23], With the aid of Hartree-Fock wave functions several authors have calculated the specific mass shift in various elements. The results, however, give the observed effects mostly only qualitatively [24-26]. Since the normal and the specific mass shift are also proportional to 1/A z (where A is the mass number of the considered isotope) both effects should be small compared to the field shift in the case of heavy atoms. It has, nevertheless, been

shown that the specific shift is up to ten times larger than tile normal mass shift when atoms with partly filled 3 d- or 4f-electron shells are considered [27]. Because the measured shift is composed of normal and specific mass shifts and field shift (FS), it is not so easy to distinguish the two latter parts. This may complicate a comparison of measured and calculated specific mass shifts. The field shift A Vvs can be reduced into two factors:

Avvs=F. C

(i)

where the electronic factor F = A l0 (0)[z ~ (a~/Z) depends mainly on the change of the electronic density A 10(0)] 2 at the nucleus, whereas the so-called isotope shift constant C depends only on nuclear properties. The value of A [O(0)l 2 can be calculated with the help of Hartree-Fock approximations (see e.g. [34]). The IS constant C can be evaluated by solving Dirac's equations for the electron moving in the electric field of the nucleus. Bodmer [28] has shown that in a sufficient approximation C depends only on the mean square radius of the nuclear charge distribution ( r 2 ) and on differences 6
W. Fischer et aI.: Isotope Shift Measurements

249

with Z < 80, Fradkin [29J gives the following approximation

)2

r(l+2-a)

an

A v1 F1 AN--AVsMs'l--AVsMs'2" Fz

\ (r 2) ]"

In this case a = ] f l - - Z z e2. The other constants have the usual meaning. By introducing a uniformly charged sphere of the radius R. with the same total charge and value for ( r 2) as the true charge distribution one obtains [5~1/2 ~t.2)l/2. u-,3,

R_

(3)

Assuming now a simple A1/a-law for the dependence of the radius Ru on the mass number A, Ru, std = ro " A 1/3

(with ro = 1.2 fm)

(4)

one finally arrives at a theoretical IS constant C~ta which can be compared with an experimental one. Since A VFS.exp_ Cexp

6 (Ri)~xp

3 (r2)exp

ZJ YFS,std

2 3 (R,)std

3 (r2)~t d

Cstd

electronic factors F1/F2 in the two lines, whereas the intercepts A v yield information on the SMS. We get for example for the intercept on the (1)-axis

(5)

the influence of the field shift can also be discussed in terms of Cexp/C~ta [30]. With the aid of the so-called King-graphs [31, 30] one can easily obtain information on different SMS and FS of two or more spectral lines. The measured shifts in one line (1), corrected for NMS, are plotted against the corresponding shifts in another line (2), first dividing each by the difference of the neutron numbers AN. The slope of the straight line gives the ratio of the

(6)

5. Interpretation of the Measured Isotope Shifts The evaluation of the SMS is especially simple in a case where the SMS in the reference line (2) is taken to be equal to zero; then the intercept A Vl yields the SMS in line (1) directly. Plotting the IS values of Table 1 in the way described above, using the line, 2 = 5 1 7 7 A as a reference line, one obtains four lines, the slopes of which deviate from unity. Since at all transitions one 6s-electron is "jumping", according to the level assignment in Table 1, the change of the electronic densities at the nucleus should be the same in all the lines, and the slopes of all the King-lines should therefore be equal to unity. Furthermore, the lines do not go through origin and therefore reveal large and different SMS. Since the intercepts obtained differ so strongly, no reasonable assumptions on the SMS can be made. It has, however, been shown in several cases that Kingplots of the IS can be drawn even in two different elements, for example for the shifts of barium and cesium [4]. In Fig. 3 the IS-values for lanthanum have been protracted in the way described above against the IS values of the isotonic nuclei in the BaII-spectrum, which is a one-electron spectrum where the SMS

AVis ( B a l l ) / r n K . A N /~,) I

1

-15

-I0

/(~,, //~/csJ

/(ff///(2J

I

-5

Vls / m K AN

-,o

Fig. 3. King-graphs of some LaI-lines referred to the line BaII, 2=4934A. (1: 2=5177A, 2: 2=5234A, 3: 2=5455/~, 4: )~=6250A,

5:2=6266/~)

250

Z. Physik 269 (1974)

The results of Wilson for the interesting levels are listed in Table 3, where the main admixtures are given. Since the electron density l0 (0)l2 in the configuration 5 d 6 s 4 f is even lower than in 5 d 2 6p where no 6 sWavelength F(,~) A VsMs electron is present, admixtures of levels with 5 d 6s 4f 2//~ F(2 = 5 177 A) mK. AN to the upper level will increase the F-value. In the case 5 177 1.00 0.3 (7) of cerium Champeau [5, 36] has shown experimentally 5 234 1.53 (20) 7.7 (7) and theoretically that the screening effect of the 4fi 5 455 2.02 (20) 3.9 (5) electron is important for the field effect. 6 250 1.25 (20) 2.2 (4) For the line 2 = 5 4 5 5 A one obtains about twice the 6 266 1.32 (20) 6.6 (4) value for A lO (0) l2 as is the case for the line 2 = 5 177 ~. This may be partly due to the impurity of the upper (purity only 45~o, 35~o adshould be small compared with the NMS. Under the level 5d6s6pye2D~ mixture of 5d 2 6p x 20o/2 ). Because of the large adassumption that the SMS in BaII is negligible [32], mixture of 5 d 2 6p (and probably 5 d 6 s 4f) the electron the SMS in the measured LaI-lines shall be derived density for this level is depressed. For the purpose of from these King-plots. In Table 2 the values for these SMS so obtained are listed together with the ratios of discussing the ratio of 10(0)] 2 in the levels 5 d 2 6S and the electronic factors arrived at from a similar diagram 5 d 6 s 2 one has to consider the screening effects of one for La lines, where the La-line 2 = 5 177A has been 5d-electron and the mutual screening of the two 6sused as a reference line, as mentioned above. As no electrons. Measurements in the HgI- and HgII-spectra appreciable SMS seems to occur only in the line LaI have shown that two 6s-electrons have only 1.6 times 2 = 5 1 7 7 A the forthcoming evaluations of experi- the electron density of one 6s-electron [1]. Hartreemental isotope shift constants Cexp and changes in Fock calculations and experimental investigations of Champeau [5, 36] in the cerium spectrum have shown (r 2) are based on measurements in this line. The deviation of the slopes of King-lines from unity that-referred to CeII 4 f 5 d Z - i n the analogue concan be explained by configuration mixing of the upper figuration CeI 4 f 5 d 6s 2 one obtains exactly twice the and lower levels and screening effects of the elec- field effect than in Ce I 4 f 5 d 2 6 s. Finally, considering the admixtures in the upper and lower levels of the trons [33]. As the LaI-spectrum is a rather complicated one, line 2 = 5455 A one can explain the ratio for where the energetic positions of most of the levels with the configurations 5 d 6 s 4 f and 6 s 2 4 f have not A [~/(0)12455/A ]1]/(0)l 2177 ~ 2 . yet been determined experimentally, reliable calculations of the eigenstates are only available for the Ab initio evaluations of the specific mass shifts with low lying even-parity levels of the configurations the aid of Hartree-Fock methods by Bauche [27] 5d 2 6s and 5d6s 2 [34]. For some of the higher lying yielded AVsMs=15mK.AN, when a 4f-electron is odd-parity levels of 5d26p and 5d6s6p exist only participating in the transition. The measurements in the spectra of cerium, neodymium and samarium very preliminary calculations [35]. Table 2. Results of the ratios of the electronic factors F/F' and the specific mass shifts A Vs• s from the King-plots of Fig. 3 assuming that A VsMs(Ba II) = 0

Table 3. Purities and admixtures of the La I-levels of the investigated spectral lines according to the calculations of Wilson [34, 35] Wavelength 2/A

Level assignation upper/lower

5 177

5 d 2 6p y 4D~

5234

Purity

68

5 d 2 6 s a 4F7/2

,~ 100

5 d 6 s 4 f 2G~ 5d 2 6sa4F9/2

,.~ 100~

5455

5d6s 6p y 2D~ 5d6s 2a 2D5/2

45 84Yo

6 250

5 d 2 6 p z 4GOi/2 5d26sa4F9/2

84 ~, lO0 ~o

6266

5 d 2 6 p z 2H~ /z 5d 26sa2G9/z

58 Yo 92~

Predominant admixtures

18 ~o (5d6s4f4D~

35 ~ (5 d z 6p x ZD~ 13 ~ (5 d z 6s b 2D5/2) 15 ~ (5d6s4f4G~ 18 ~ (5d6s4fz 2H~ 7 ~ (5d 3 2G9/2)

+ 10~oo(5d6s4f4H~

w. Fischer et al.: Isotope Shift Measurements

251

confirm this result qualitatively. The experimental contribution of the @electron, however, appears to be between one half and two thirds of the theoretical values. Considering first the two lines 2 = 6 2 5 0 / ~ and 2 = 6266A, one discovers that the value for the SMS (Table 2) arrived at are roughly proportional to the admixtures of levels of the configuration 5 d 6s 4 f as given by Wilson. In the case of the line 2 = 5 455 A not all admixing levels are given there. Supposing that the missing levels belong to configurations with a 4felectron, then the SMS in this line can also be explained. The upper level of the line 2 = 5 2 3 4 A , formerly 0 assigned as lvsz, has been reclassified by Brewer [13] with the aid of thermodynamical methods as 5d6s4f2G~ . But here again configuration interaction seems to affect the upper level heavily. The SMS obtained is much smaller than the theoretical value from Bauche [27]. Unfortunately, this simple theory fails in the case of the line 2 = 5 1 7 7 ~ where large admixtures of the configuration 5 d 6 s 4 f to the upper level should cause a large SMS in contradiction to the results from the King-graphs. Besides the influence of the 4f-electron on the SMS the 5 d-electron can also cause considerable contributions [33]. A settled evaluation, however, will only be possible with a new investigation and classification of the LaI-spectrum as a number of levels have not yet been found experimentally. The energetic positions of these levels are necessary parameters for the calculation of the eigenvectors. This work is in progress at Orsay [37].

quantum number no and dnJdn:

glo(a 4F7/2)= 1.5948,

dnffdn = 1.06.

On the assumption that in the line 2 = 5 1 7 7 • no SMS is present, irrespective of Wilson's analysis, we have then:

NAN' IriS - - SAN' VNMS Cexp(N, N') ~-

(8)

0.261

These values are given in column 2 of Table4 for lanthanum. Since there remains an undeterminable uncertainty in the estimations of SMS and possible systematic errors no error bars can be given at the present. The values for 6(r2)exp obtained by the Eqs. (2)-(5) are given in column 3. The corresponding values for the isotonic barium nuclei fit well with the data of lanthanum. Thompson [38] has measured absolute values of ( r 2) for the isotonic nuclei 138Ba, 139La, and lZ~ with the aid of muonic atoms. The radii of these nuclei were found to increase about 1.63 times faster than according to the relation R . = r o . A 1/3, i.e. 8RffSZ= l.63 dRffdA. The A1/3-1aw holds true for the entire periodic system of the elements with sufficient accuracy. Following Bodmer [28] we have

d R , _ 8eu d g dA 8N dA

8R, dZ 8Z dA

(9)

(7)

Considering the nuclei along the bottom of the valley of stability we get in the region of the mass number A = 139 the values dN/dA~lO/17 and dZ/dA~7/17. Putting these figures into Eq. (9) one obtains 8R./SN 0.6. dRffdA. This value of the so-called "isotope shift discrepancy" r/= 0.6 is confirmed by measurements in barium [-32]. Adopting this value for the lanthanum nuclei one can calculate 3 (r2)theo = ~ 6 (r2)std (column 5 of Table4). The differences between ~(r2)exp and 6 (r:)th~o may be attributed to different deformations. Since the values for 6 ( r 2) are nearly the same for the isotonic barium nuclei, one should also expect similar deformations as well.

Russell and Meggers [12] give for the level La 5d26sa4F7/2 the following values for the effective

The authors are very much indebted to Prof. E. Gerdau and Dr. H. Winkler, II. Pbysikalisches Staatsinstitut Hamburg, who placed

6. Determination of 6 @2) Assuming that the FS occurs only in one of the two electronic levels and that exactly one s-electron is jumping, Cexp can be evaluated with the aid of FermiSegr6's formula:

NAN' VFS C~xp(N, N ' ) = ZJ (dnJdn) n2 3 9

Table 4. Experimentalisotope shift constants Cexpof lanthanum, changes in the mean square charge radius 6 (r2)oxpof lanthanum and of the isotonic barium nuclei [4] and theoretical values 3 (r2)theofor lanthanum

N,N'

Ce,:v(La)/mK

6(r2)~xr,(La)/fm 2

6(r2)~xp(Ba)/fm 2

6(r2)thoo(La)/fm z

81, 82 80, 82

32.8 23.0

0.067 0.047

0.066 0.049

0.067 0.134

252 some 100 ~tg of 13VLa at their disposal; to Dr. H. Ihle, Kernforschungsanlage Jiilich, for performing the mass spectrometric analyses, and to Dr. H. Ihle and Dr. R.Wagner, Kernforschungsanlage Jiilich, who did the electromagnetic isotope separation. One of us (K.M.) acknowledges the financial support of the Deutsche Forscbungsgemeinschaft.

References 1. Kopfermann, H.: Kernmomente. Frankfurt/Main 1956 2. Fischer, W., Hfihnermann, H., Kr6mer, G., Schgfer, H.-J.: Third Conference on Atomic Spectroscopy, Reading 1971 3. HiJhnermann, H.: Thesis Marburg 1967 4. Hiihnermann, H.: Habilitationsschrift Marburg 1971 5. Champeau, R.-J.: Physica 62, 209 (1972) 6. Mtiller, G., Skrok, D., Winkler, R.: Optik 32, 410 (1971) 7. Gerdau, E., Winkler, H., Roggenbuck, A.: Private communication 8. Camplan, J., Meunier, R., Sarrouy, J.L.: Nucl. Instr. Meth. 84, 37 (1970) 9. Martin, R, Walcher, W.: Marburger Sitzungsberichte 75, 5 (1952) 10. Htihnermann, H., Wagner, H.: Phys. Lett. 21, 303 (1966) 11. Wagner, R.: Conference on Electromagnetic Mass Separation. Sk/Svde 1973 12. Russell, H.N., Meggers, W.F.: National Bureau of Standards. J. Research IX, 1932 13. Brewer, L.: J. Opt. Soc. Am. 61, 1101 (1971) 14. Ltihrs, G.: Z. Physik 141, 486 (1955) 15. Ting, Yu.: Phys. Rev. 108, 295 (1957) 16. Childs, W.J., Goodman, L.S.: Phys. Rev. A3, 25 (1971) 17. Hese, A.: Z. Physik 236, 42 (1970) 18. Hese, A.: Ann. Physik (7) 25, 299 (1970)

Z. Physik 269 (1974) t9. Fischer, W., Htihnermann, H., Mandrek, K.: Z. Physik 248, 53 (1971) 20. Fischer, W., Hiihnermann, H., Mandrek, K.: Z. Physik 254, 127 (1972) 21. Fischer, W., Hiihnermann, H., Mandrek, K., Ihle, H.: Phys. Lett. 40B, 87 (1972) 22. Fischer, W., Hiihnermann, H.: J. Physique 30, C-l, 59 (1969) 23. Vinti, J. P.: Phys. Rev. 56, 1120 (1939) 24. Bauche, J.: Comptes Rendus B 263, 685 (1966) 25. Elbel, M., Hiihnermann, H.: J. Physique 30, C1-41 (1969) 26. Bauche, J., Crubellier, A.: J. Physique 31, 429 (1970) 27. Bauche, J.: Fourth Conference on Atomic Spectroscopy. Amsterdam 1972 28. Bodmer, A.R.: Nucl. Phys. 9, 371 (1958) 29. Fradkin, E.E.: Sov. Phys. JETP 15, 550 (1962) 30. Stacey, D.N.: Rep. Progr. Phys. 29, 171 (1966) 31. King, W.H.: J. Opt. Soc. Am. 53, 638 (1963) 32. Fischer, W., Hartmann, M., Htihnermann, H., Vogg, H.: Z. Physik 267, 209 (1974) 33. King, W.H., Steudel, A., Wilson, M.: Z. Physik 265, 207 (1973) 34. Wilson, M.: Phys. Rev. A3, 45 (1971) 35.' Wilson, M.: Private communication 36. Champeau, R.-J.: Physica 62, 225 (1972) 37. BenAhmed, Z., Blaise, J, Verg6s, J., Wyart, J.F.: Fifth Conference on Atomic Spectroscopy, Lund 1973 38. Thompson, A.Ch.: Thesis Pittsburgh 1969 Prof. Dr. W. Fischer Prof. Dr. H. Htihnermann Dr. K. Mandrek Fachbereich Physik der Philipps-Universit~it D-3550 Marburg/Lahn Renthof 5 Federal Republic of Germany