Atomic Screening Constants from SCF Functions. II.

Atoms with 37 to 86
Electrons
E. Clementi, D. L. Raimondi, and W. P. Reinhardt
Citation: The Journal of Chemical Physics 47, 1300 (1967); doi: 10.1063/1.1712084
View online: http://dx.doi.org/10.1063/1.1712084
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/47/4?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
On evaluating force constants from LCAOMOSCF electronic charge densities for diatomic molecules
J. Chem. Phys. 58, 381 (1973); 10.1063/1.1678934
Calculation of Atomic Force Constants from Electron Densities
J. Chem. Phys. 57, 4535 (1972); 10.1063/1.1678111
Ab Initio Calculation of Harmonic Force Constants. VII. Results from Some TwoElectron SCF
Wavefunctions
J. Chem. Phys. 57, 1364 (1972); 10.1063/1.1678408
Use of Atomic SCF Functions with Minimal Basis Sets
J. Chem. Phys. 49, 1987 (1968); 10.1063/1.1670353
Atomic Screening Constants from SCF Functions
J. Chem. Phys. 38, 2686 (1963); 10.1063/1.1733573

Reuse of AIP Publishing content is subject to the terms: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 65.110.254.159 On: Sat, 26
Mar 2016 04:50:34

THE JOURNAL OF CHEMICAL PHYSICS

VOLUME 41, NUMBER 4

15 AUGUST 1961

Atomic Screening Constants from SCF Functions. II. Atoms with 37 to 86 Electrons
E. CLEMENTI, D. L. RAIMONDI, AND W. P. REINHARDT*
IBM San Jose Research Laboratory, San Jose, California

(Received 10 April 1967)

Minimal basis-set atomic functions for the ground-state atoms from Rb(Z=37) to Rn(Z=86) are
presented. These functions are analyzed in order to obtain systematic data for the screening constants and
atomic radii following the work initiated by Slater.

1. INTRODUCTION

II, RESULTS

In a previous paperl (hereafter referred to as I),

self-consistent-field functions were computed with a
minimal basis set of Slater-type orbitals for atoms with
2 to 36 electrons. From the analysis of the optimized
orbital exponents, atomic screening constants were
derived for the 1s, 2s, 2p, 3s, 3p, 4s, 3d, and 4p electrons.
The screening constants were compared with those
proposed by Slater in his classical work2 and the new
"rules" were obtained. The previous work was motivated by the need for simple wavefunctions to complement the large number of highly accurate wavefunctions available for the first-, second-, and third-row
atoms. For the atoms of the fourth, fifth, and sixth rows
the need for simple wavefunctions is even more apparent, since little data is presently available. They are
particularly useful where preliminary computations are
desired or where high accuracy is either not feasible
with present techniques or not required. In this paper,
we give SCF minimal basis-set functions for atoms with
37 to 86 electrons or for the atoms from rubidium to
radon in the periodic table. In addition, the previous
results are combined with the present data in discussing
screening constants and orbital radii. In order to clarify
the results to be presented, a short review of the theory
follows.
The Slater-type basis set has the form
Xn,I,m=Nr n- l exp( -tn,l,mr) Y I,m(8, I{J),
(1)

A. Total Energies and Orbital Exponents

where n, l, and m are the usual quantum numbers, Y l m

is a spherical harmonic, N is a normalization factor, arid
t n,l.m is the orbital exponent related to the screening
constant Un,l.m and to the effective nuclear charge Zeff
by the relations
tn,I,m=Zeff/n= (Z-Un,l,m)/n.

(2)

It should be noted that we use integer values for "n"

whereas Slater used either integer values (in the firstand second-row elements) and fractional values in the
third and higher rows of elements.
* Present address: Harvard University, Chemistry Department,
Cambridge, Mass.
1 E. Clementi and D. L. Raimondi, J. Chern. Phys. 38, 2686
(1963) .
2 J. C. Slater, Phys. Rev. 36, 57 (1930).

Table I lists atomic number (Z), atomic symbol,

outer-shell configuration, electronic state, and the total
TABLE I. Total energy for neutral atoms from Rb through Xe.
Z

Atom

37
38
39
40
41
41
42
42
43
44
44
45
45
46
46
47
47
48
49
50
51
52
53
54

Rb
Sr

Configuration

Zr
Nb
NbMo
MoTc
Ru
RuRh
RhPd
PdAg
AgCd
In
Sn
Sb
Te
I

Xe

5s(1)
5s(1)

5s(1)
5s(1)
5s(0)

5s(1)

5s(1)
5s(2)
4d(1)
4d(2)
4d(3)
4d(4)
4d(4)
4d(5)
4d(5)
4d(6)
4d(7)
4d(7)
4d(8)
4d(8)
4d(1O)
4d(9)
4d(10)

4d(1O)
5p(1)
5p(2)
5p(3)
5p(4)
5p(5)
5p(6)

State

Energy

2S
'S

-2930.6932
-3123.7176
-3324.7806
-3531.3152
-3745.4783
-3745.2448
-3967.0362
-3966.3422
-4196.0510
-4432.3593
-4432.1106
-4676.2624
-4676.0290
-4927.8059
-4927.1726
-5187.0706
-5186.8939
-5454.1909
-5729.0988
-6011.6722
-6302.0043
-6600.0386
-6905.9462
-7219.7921

2D
3F
4F
6D
5D

7S
6S
aD
5F
4F
4F
3F

'S

2D

2S
'S

2p
3p

4S
3p
2p

'S

- The filling up of the electronic shells is not exactly consecutive for the
ground states of the fourth-row atoms. Therefore. for the fourth-row atoms
we have given two states: the lowest one and that which one would obtain
if the filling up of the shell would be regular.

energy for atoms from rubidium to xenon. Table II

lists the optimized orbital exponents for the same group
of atoms. Tables III and IV similarly apply to the atoms
from cesium to radon. It is noted that the 4f electrons
are not well described by a single Slater-type orbital
when the number of 4f electrons is small. This is attributed to the extreme broadness of the 4f orbital. Considerable trouble was involved in obtaining results for

1300

Reuse of AIP Publishing content is subject to the terms: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 65.110.254.159 On: Sat, 26
Mar 2016 04:50:34

1301

ATOMIC SCREENING CONSTANTS. II

TABLE II. Tabulation of optimized orbital exponents for neutral Rb through neutral Xe in their ground state.

Z
37
38
39
40
41
41
42
42
43
44
44
45
45
46
46
47
47
48
49
50
51
52
53
54

Atom

Is

36.2078
Rb
Sr
37.1911
Y
38.1756
39.1590
Zr
40.1423
Nb
Nb a 40.1423
Mo 41.1256
Mo 41.1256
Tc
42.1090
43.0923
Ru
Ru 43.0923
44.0756
Rh
Rh- 44.0756
Pd
45.0589
Pd- 45.0589
Ag
46.0423
Ag- 46.0423
47.0256
Cd
In
48.0097
48.9920
Sn
49.9744
Sb
Te
50.9568
I
51.9391
52.9215
Xe

2$

2p

3$

3p

4s

3d

4p

5s

13.5784
13.9509
14.3111
14.6869
15.0626
15.0626
15.4384
15.4384
15.8141
16.1899
16.1899
16.5656
16.5773
16.9414
16.9414
17.3171
17.3171
17.6929
18.0618
18.4297
18.7977
19.1656
19.5335
19.9015

16.5194
17.0152
17.5016
17.9964
18.4911
18.4911
18.9859
18.9859
19.4704
19.9754
19.9754
20.4702
20.4702
20.9650
20.9650
21.4597
21.4597
21.9545
22.4490
22.9427
23.4363
23.9300
24.4237
24.9173

7.2809
7.5546
7.8505
8.1205
8.3905
8.3905
8.6605
8.6605
8.9304
9.2004
9.2004
9.4704
9.4795
9.7404
9.7404
10.0104
10.0104
10.2804
10.5436
10.8066
11.0697
11.3327
11.5958
11.8588

7.1011
7.3892
7.6975
7.9485
8.2184
8.2052
8.5030
8.4912
8.7947
9.0844
9.0737
9.3724
9.3848
9.6616
9.6732
9.9476
9.9362
10.2305
10.5069
10.7844
11.0613
11.3363
11.6138
11.8892

3.0970
3.3611
3.5659
3.7254
3.9090
3.8207
4.1037
4.0241
4.2996
4.4876
4.4140
4.6571
4.6454
4.8568
4.7465
5.0398
4.9662
5.2173
5.4403
5.6645
5.8859
6.1021
6.3243
6.5432

7.2264
7.5754
8.4657
8.5223
8.7847
8.7490
9.1097
9.0761
9.4510
9.7981
9.7863
10.1478
10.1350
10.4989
10.4837
10.8503
10.8466
11.2023
11.5594
11.9139
12.2666
12.6131
12.9669
13.3156

2.7202
2.9830
3.1864
3.3650
3.5652
3.5211
3.7622
3.7442
3.9528
4.1291
4.1087
4.3236
4.2849
4.5010
4.4308
4.6777
4.6406
4.8528
5.0922
5.3163
5.5453
5.7805
6.0074
6.2393

0.9969
1. 2141
1.2512
1.2891
1.3392
1.1842
1.3952
1.2212
1.4453
1.4905
1.2969
1.5286
1.3279
1.5675
1.6057
1.3511
1.6384
1.9023
2.1257
2.3222
2.5076
2.6807
2.8436

4d

5p

3.9896
3.2679
3.0796
2.8094
3.1110
2.8481
3.2205
3.3470
3.2032
3.4937
3.3606
3.6476
3.4044
3.8064
3.6907
3.9692
4.2354
4.4925
4.7436
4.9900
5.2335
5.4733

1.6940
1.8204
1.9989
2.1617
2.3223
2.4849

These atoms have configurations corresponding to atoms labeled "au of Table I.

TABLE III. Total energy for neutral atoms from Cs through Rn

(in atomic units) .

Atom

55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86

Cs
Ba
La
Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Hf
Ta
W
Re
Os
Ir
Pt
Au
Hg
Tl
Pb
Bi
Po
As
Rn

Configuration

State

Energy

4/(1)

2S
IS
2F

4f(2)
4f(3)

4J

7541.3432
7870.6740
- 8207.7052
- 8552.4491
- 8905.3989
- 9267.0675
- 9637.0680
-10015.325
10402.457
-10797.853
-11 202.707
-11 615.016
-12038.572
-12467.338
-12907.267
-13 356.155
-12816.567
-14285.319
-14763.336
-15250.513
15746.921
-16252.392
-16767.183
-17291.316
-17824.845
-18367.869
-18920.276
-19481.953
-20052.992
-20633.317
-21223.079
-21822.323

6$(1)
6s(2)

3[

4/(4)

6[

4f(5)

GH

4/(6)
4/(7)

7F
8S
7F

4f(8)

4/(9)
4/(10)
4/(11)
4/(12)
4f(13)
4f(14}
5d(1)
5d(2)
5d(3)
5d(4)
5d(5)
5d(6)
Sd(7)
5d(8)
5d(9)
5d(10)
6p(l)
6p(2)
6p(3)
6p(4)
6p(5)
6p(6)

6H
5[

4J
3[

2F
IS
2D

3F
4F
5D

'S
6D

'F
SF
2D

IS
2p
ap

4S
.p
2p
IS

the atoms from lathanum (Z =57) to promethium

(Z = 60). In fact, for La, Pr, Nd, and Pm 2 STO's were
required to get convergency. In Table IV these cases are
apparent by the use of a second 4f STO (labeled 41').
In each case the first 4j STO given is the dominant one.
In the case of La and Ce the extremely small orbital
exponent indicates the diffusiveness of the 4f electron,
since the orbital exponent is inversely proportional to
the radius of maximum charge density.
B. Screening Constants and Orbital Radii
The screening constants reported in F were presented
with the following qualifications. If one wishes to
identify the u's with the f's via Eq. (2), then one should
restrict himself only to those orbitals which have no
other orbital of the same symmetry in the atom. For
example, if one considers the neon atom, the correspondence between f and u is unambiguous for the 2p
but not for the 1s and 2s orbitals. The reason is simply
that the complete basis set of a given symmetry is used
in building the orbitals of that symmetry.
In other words, as noted in I, the coupling of the
vectors for the basis set of a given symmetry prevents
an objective division of various contributions to the
screening constants. The problem becomes clearly more
acute the larger the basis set of a given symmetry. In
Fig. 1, the screening constants calculated from Eq. (2)
are plotted versus atomic number for the atoms from
helium to radon. Certainly, discontinuities should exist
where new orbitals begin to fill, and these are evident.
But a detailed discussion of these results is clouded by
the ambiguities discussed above. There is a simple way
by which these ambiguities may be in part removed. Let
us consider again the neon atom. In the minimal basis

Reuse of AIP Publishing content is subject to the terms: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 65.110.254.159 On: Sat, 26
Mar 2016 04:50:34

.......
CN

T ABLE IV. Tabulation of optimized orbital exponents for neutral Cs through neutral Rn in their ground state.
Z
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86

Atom
Cs
Ba
La
Ce
Pr
Nd

Pm
8m
Eu
Gd

Tb
Dy

Ho
Er
Tm
Yb

Lu
Hf
Ta
W
Re
Os

IiPt
Au
Hg
TI
Pb
Bi
Po
As
Rn

Is
53.9043
54.8861
55.8683
56.8481
57.8306
58.8132
59.7958
60.7783
61.7609
62.7435
63.7261
64.7086
65.6912
66.6737
67.6563
68.6389
69.6195
70.6016
71. 5837
72.5657
73.5478
74.5299
75.5119
76.4940
77.4761
78.4581
79.4409
80.4195
81.3982
82.3768
83.3554
84.3341

2s
20.2558
20.6234
20.9767
21.3700
21. 7310
22.1081
22.4852
22.8674
23.2354
23.6085
23.9861
24.3547
24.7278
25.1008
25.4739
25.8470
26.2249
26.5949
26.9649
27.3349
27.7049
28.0749
28.4449
28.8149
29.1849
29.5547
29.8421
30.2150
30.5880
30.9609
31.3338
31.7068

2p
25.4098
25.9048
26.3978
26.8912
27.3847
27.8783
28.3719
28.8655
29.3590
29.8527
30.3462
30.8398
31.3334
31.8270
32.3206
32.8142
33.3055
33.7994
34.2932
34.7871
35.2810
35.7749
36.2688
36.7627
37.2566
37.7505
38.2431
38.7383
39.2335
39.7286
40.2238
40.7190

3s

3p

4s

3d

4p

12.1258
12.3852
12.6477
12.8864
13.1670
13.4476
13.7282
14.0088
14.2894
14.5699
14.8505
15.1311
15.4117
15.6923
15.9728
16.2534
16.5115
16.7705
17.0305
17.2900
17.5495
17.8091
18.0686
18.3281
18.5876
18.8471
19.1397
19.3841
19.6285
19.8729
20.1173
20.3617

12.1926
12.4388
12.7132
12.9865
13.2748
13.5630
13.8513
14.1395
14.4280
14.7164
15.0049
15.2935
15.5818
15.8703
16.1587
16.4455
16.7221
16.9944
17.2668
17.5392
17.8115
18.0839
18.3563
18.6287
18.9010
19.1734
19.4555
19.7165
19.9774
20.2383
20.4992
20.7602

6.7606
6.9800
7.1991
7.4200
7.5833
7.7466
7.9099
8.0731
8.2170
8.3610
8.5050
8.6480
8.8280
9.0580
9.2844
9.3794
9.5673
9.7443
9.9397
10.1397
10.3391
10.5238
10.7120
10.9097
11.1033
11.3112
11.5197
11.7232
11.9268
12.1304
12.3339
12.5375

13.6602
14.0081
14.3534
14.6951
15.0508
15.3856
15.6994
16.0763
16'.4176
16. 75J)0
17.0995
17.4433
17.7823
\8.1201
18.4581
18.7989
19.1396
19.4766
19.8137
20.1508
20.4849
20.8249
21.1620
21.4991
21. 8361
22.1732
22.5114
22.8489
23.1805
23.5240
23.8615
24.1991

6.4644
6.7008
6.9266
7.1516
7.2642
7.5035
7.6558
7.7720
7.9687
8.1617
8.3497
8.4565
8.6407
8.7773
8.9970
9.1005
9.2976
9.4824
9.6837
9.8871
10.0933
10.2860
10.4785
10.6826
10.8867
11.1015
11.3042
11.5084
11.7126
11.9168
12.1210
12.3253

5s
3.0889
3.3239
3.5622
3.7827
3.5226
3.7485
3.7671
3.6498
3.7180
3.7764
3.8341
3.8608
3.9152
3.9436
4.0074
4.0300
4.1910
4.3666
4.5387
4.7083
4.8714
5.0190
5.1691
5.3176
5.4655
5.6222
5.8244
6.0263
6.2058
6.4046
6.5867
6.7786

4d

5p

6s

4/

5.7096
5.9460
6.1813
6.4152
6.5743
6.7023
6.9350
7.0599
7.2352
7.4084
7.5775
7.7545
7.9179
8.0678
8.2360
8.3974
8.8223
8.8810
9.0810
9.2933
9.5136
9.7145
9.9343
10.1575
10.3820
10.6170
10.8472
11.0799
11.3098
11.9168
11.7624
11.9857

2.7302
2.9601
3.1792
3.3931
3.0567
3.3922
3.2828
3.2562
3.3110
3.3528
3.3922
3.4254
3.4678
3.4944
3.5456
3.5663
3.7360
3.9170
4.0947
4.2651
4.4288
4.5820
4.7322
4.8839
5.0340
5.1934
5.4177
5.6060
5.8042
6.0049
6.2080
6.3942

1.0605
1.2625
1.5520
1.7994
1.2911
1.5511
1.5659
1.3353
1.3536
1.3691
1.3834
1.3906
1.4065
1.4127
1.4307
1.4322
1.4674
1.5274
1.5875
1.6424
1.6860
1.7205
1.7611
1. 7919
1.8230
1.8589
2.1366
2.3500
2.5400
2.7218
2.8833
3.0540

0.3400
0.4190
5.2752
5.5665
5.7835
5.8829
6.0800
6.2534
6.4662
6.6340
6.8674
6.9946
7.1585
7.3580
7.7328
8.0524
8.3676
8.6777
8.9812
9.2882
9.5862
9.8765
10.1624
10.4402
10.7169
10.9922
11.2673
11.5396
11.8101
12.0828

4/'

5d

6p
(')

t-<
tr1
~
tr1

3.5230
0.5000
0.5800
0.6500

....>-3
:;0

;;.-

....
~
0

Z
t:I

4.0226
3.3239
3.2736
3.3484
3.4766
3.5994
3.7392
3.8815
4.0253
4.1712
4.4050
4.6304
4.8488
5.0608
5.2678
5.4706

.....
;;.Z
t:I
:;0

tr1

....
Z

::r:
2.0423
2.0655
2.2233
2.3701
2.5272
2.6793

Reuse of AIP Publishing content is subject to the terms: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 65.110.254.159 On: Sat, 26 Mar 2016 04:50:34

;;.-

:;0

t:I
>-3

ATOMIC SCREENING CONSTANTS. II

70

j 6P

68

65

1303

66
64
62
60

58
56

);;;:=
~J!

54
52
50
48

46
fz 44
~
(f)
z 42
0
u 40 <!)

38

36 --

('4f

5p
55

5p

5d

cr 34

(f)

32 -

55

30
2826
4p

24-

.......

~~

20
18
16

/~;::/::

/y

4d

22

4f

4d

45
3d

14

35
3p
25

3d

12
10
35

3p

6
2p

25

2p

I s

Is

0
0

10

20

30

40

50

60

70

80

90

Z ---.

FIG. 1. The screening constants IT calculated from Eq. (2) are plotted vs atomic number Z.

set-SCF formalism, we use one is STO, one 2s STO,

and one 2p STO. The radial part of the SCF orbital,
f(r) , is characterized by the value of r for which rf(r)
has an extrema or a node. If the minimal basis set is
well optimized the extrema and nodes are unambiguously determined by plotting the radial functions.

Therefore, one can obtain two sets of s-'s. The first set
contains the r's which correspond to the optimized
minimal basis set. The second set contains the r's which
correspond to STO's, not used in the computation, but
which have the same rmax as given by the first set. Since
the minimal basis set of STO's gives an orthogonal set

Reuse of AIP Publishing content is subject to the terms: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 65.110.254.159 On: Sat, 26
Mar 2016 04:50:34

1304

CLEMENTI, RAIMONDI, AND REINHARDT

0.8

0.6
;:;

.;

""E

0.4

0.2

5.8

0.0

65

60

70

z __80

90

5.4
55

5.0
45

4.6

...,

4.2

:::J

.!i.

)(

"

~E

3.8
35
3.4
25
3.0

~ 6p

2.8
4p
2.4

\ \ 6p

3p

65

2.0
1.8

2p

1.4

1.0

5d

5p

.8
.6

:4

.2j__~~~=;~~~~~~~~~~~~~~~~~~~~I:N~SE~R~T:-__-1

.0

10

20

30

40

50

FIG. 2. Radii of maximum charge density

of SCF orbitals and since the STO's are nodeless functions, the only rmax of rj(r) which should be considered
is that corresponding to the main and outermost lobe
from the nucleus. This presents no ambiguity since the
main lobe is distinct from the orthogonality lobes of the
SCF orbitals. For example, using again the case of the

60
70
Z---..
rmax

80

90

are plotted vs atomic number Z.

neon atom, the minimal basis set is

1s(SCF) =f(rlB) Y oo = 0.99735x (1s) +O.01092x(2s) ,
2s(SCF) =j(r28) Yoo= -O.25499x(1s) +1.02937x(2s) ,
2p(SCF) =f(2p) Y lm =X2p,

(3)

Reuse of AIP Publishing content is subject to the terms: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 65.110.254.159 On: Sat, 26
Mar 2016 04:50:34

1305

ATOMIC SCREENING CONSTANTS. II

r-He
0.31

Li

Be

1.67

1.12

B
0.87

C
0.67

0
0.48

0.56

Ne

0.42

0.38

Na

Mg

Al

Si

CI

Ar

1. 90

1.45

1.18

1.11

0.98

0.88

0.79

0.71

K
2.43

Ca

Sc

Ti

1.94

1.84

1.76

Rb
2.65

Sr

Y
2.12

Zr

Nb

Mo

2.19

2.06

1. 98

1.90

Hf
2.08

Ta

Re

Os

Ir

Pt

Au

Eg

TI

Pb

Bi

Po

At

Rn

2.00

1.93

1. 88

1. 85

1. 80

1. 77

1.74

1.71

1.56

1.54

1.43

1. 35

1. 27

1. 20

Cs

Ba

La

2.98

2.53

(6.22)

V
1.71

Cr

Mn

Fe

Co

Ni

Cu

Zn

Ga

Ge

As

Se

Br

KT

1.66

1.61

1.56

1.52

1.49

1:45

1. 42

1.36

1.25

1.14

1.03

0.94

0.88

r..:.,c

Ru

Ag

Cd

In

Sn

Sb

Te

Xe

1.78

Rh
1. 73

Pd

1.83

1.69

1.65

1.61

1.56

1.45

1. 33

1.23

1.15

1. 08

I Ce I Pr I Nd I Pm I Sm I Eu I Gd I Tb I:
(5.05)

2.47

2.06

2.05,2.38

2.31

2.33

2.25

Dy
2.28

I go I Er I Tm
2.26

2.26

2.22

I I I
Yb
2.22

Lu
2.17

FIG. 3. Atomic radii (Angstroms), determined from minimal-basis-set SCF functions.

where x(ls), x(2s), and x(2p) are normalized STO's

with orbital's exponents equal to 9.6421,2.8793, and
2.8792, respectively. From rf(r)max one would obtain
the following set of f's:
f(ls) =9.6339,
S-'(2s) =2.7933,

(4)

f(2p) =2.8792,

which is different than the set used in the computation

of the SCF function. Since we are using minimal basis
sets, the difference will not be too large for those cases

where the number of STO's of the same symmetry is

small. However, the differences will be considerably
greater for atoms with large numbers of electrons.
In addition to the usefulness of the orbital radii
maxima for the screening constant analysis, these data
are interesting in themselves for consideration of atomic
radii. Certainly, there is a correlation between atomic
radii and the radius of maximum charge density in the
outermost shell of the atom. In a recent paper, Slater3
has found a good correlation between his empirically
determined radii and radii calculated from the wavefunctions of Liberman, Waber, and Cromer. 4 5 Thus, in
Fig. 2 we give radii of maximum charge density for each

TABLE V. Comparison of Roo .. for rare-gas atoms determined by using SCF functions of varying basis-set dimension (in atomic units).
Atom
He
He
He
Ne
Ne
Ne
Ar
Ar
Ar
Kr
Kr
Kr
Xe
Xe

Basis
set
AI
B6
C6
Al
Be

Al
B6
C6
AI
B6
C6
A
B"

Is

2s

2p

3s

3p

4s

3d

4p

5s

4d

5p

0.593
0.568
0.569
0.104
0.103
0.103
0.057
0.057
0.057
0.028
0.028
0.028
0.019
0.018

0.716
0.687
0.683
0.346
0.338
0.339
0.162
0.158
0.158
0.106
0.102

0.695
0.623
0.635
0.286
0.280
0.280
0.126
0.126
0.125
0.081
0.080

1. 23
1.18
1.18
0.489
0.476
0.475
0.301
0.291

1.35
1.30
1.30
0.471
0.469
0.467
0.279
0.279

1.46
1.37
1.38
0.700
0.687

0.436
0.408
0.412
0.227
0.225

1.66
1.56
1.57
0.710
0.705

1.81
1. 74

0.760
0.752

2.04
1.96

A, Single zeta.
B, Double zeta.
C, Hartree-Fock.
" Preliminary results of Clementi.

J. C. Slater, J. Chern. Phys. 41, 3199 (1964).

(D. Liberman, J. T. Waber, and D. T. Cromer, Phys. Rev. 137, A27 (1965).
Ii J. Waber and D. T. Cromer, J. Chern. Phys. 42, 4116 (1965) j see in addition, C. Froese, J. Chem. Phys.45, 1417 (1966).

Reuse of AIP Publishing content is subject to the terms: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 65.110.254.159 On: Sat, 26
Mar 2016 04:50:34

1306

CLEMENTI, RAIMONDI,

AND REINHARDT

70

f6 P

68

65

66
64
62
60

5d

58
5p
55

56
54
52
50

-b

48
46
44

45
4p

42
en
z 40
0

4d

I-

I-

(.)
<!)

38

z
z

36

W
0:

34

en

32

(.)

4f
35

3p

30
28
26

25

4p

24
22
20
18
16
14

3d

12
10
3s

8
6

---

25

4
2p

Is

15

0
0

10

2p

20

30

40

50

60
70
Z-.

FIG. 4. The screening constants u' calculated from

of the orbitals for atoms from helium to radon, and in

Fig. 3 the atomic radii in angstroms are given for
helium to radon as calculated from the minimal-basisset wavefunctions described above. Apart from some
peculiarities, notably the behavior of the 6s orbital as
the 4j shell is being filled, the variations with increasing

"mn

80

90

are plotted vs atomic number Z.

atomic number are relatively smooth. The contraction

of the d and f orbitals is certainly evident in the Fig. 2
insert. Due to the qualitative nature of these simple
wavefunctions, a detailed discussion of these results is
postponed until a similar analysis is made on the more
accurate wavefunctions which are available for atoms

Reuse of AIP Publishing content is subject to the terms: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 65.110.254.159 On: Sat, 26
Mar 2016 04:50:34

ATOMIC

SCREENING

from helium to xenon. This is exemplified in Table V

where we have compared orbital radii for the rare gas
atoms (helium, neon, argon, krypton, and xenon) as
determined from wavefunctions of varying basis set
dimensions.6 In the outer shells, large deviations are
evident.
In Fig. 4 we have plotted the screening constants
determined from the orbital radii by the technique
described above. Although the trends are similar, these
data are not clouded by the coupling of the vectors of
orbitals of the same symmetry. One difference which
stands out is the screening of the 4s, 4p, 4d, and 4f
6

E. Clementi, IBM J. Res. and Develop. Supp\. 9, 2 (1965).

CONSTANTS.

electrons. Because of the contraction of the 4f, 4d, and

4p orbitals the screening constants for these orbitals
should in the limit of high Z be less than the screening
constant for the 4s orbital. Considering radon, in Fig. 1
the order of increasing values of the screening constants
are 4s, 4p, 4f, and 4d. But in Fig. 4, we have similarly
4f, 4d, 4p, and 4s which is the correct description. Again
a more detailed analysis is postponed until this analysis
can be applied to more accurate wavefunctions.
So that all the data is available to the reader, tables of
orbital radii for the atoms from helium through radon
are available upon request to the authors.7
7 E. Clementi, D. L. Raimondi, and W. P. Reinhardt, IBM Res.
Rept. RJ-431.

VOLUME 47, NUMBER 4

THE JOURNAL OF CHEMICAL PHYSICS

1307

II

15 AUGUST 1967

Temperature Dependence, Orientation Correlation, and Molecular Fields in

Second-Harmonic Light Scattering from Liquids and Gases
D. L.

WEINBERG*

Lincoln Laboratory,t Massachusetts Institute of Technology, Lexington, Massachusetts

(Received 3 January 1967)
The properties of second-harmonic light scattering in fluids are derived in terms of the molecular polarizabilities, orientation correlation among molecules, and the molecular field Fa. Measurements of scattering
intensity in CCl4 and water, from about 10 to 60C, are reported. The variations with temperature, and
therefore apparently the effects of changes of orientation correlation, are slight. The influence of preferred
orientation in CCI 4, of the type deduced from x-ray diffraction, is calculated. The third-rank molecular
polarizability tensor consists of a fourth-rank tensor contracted with Fa, in addition to the term previously
known, which is independent of Fa. With Fa arising from molecular multipole moments, the new term
explains some measured depolarizations in liquids, and accounts for the spectral line, but not the background, observed in methane gas. In the case of molecular dipole moments, the new term can cause spectral
narrowing.

1. INTRODUCTION
Second-harmonic light scattering has been observed
in liquids and fused quartzl and methane gas. 2 When
an optical electric field of angular frequency w is incident
on matter, Rayleigh scattering arises from local, e.g.,
molecular, induced electric dipoles, (neglecting higherorder induced moments) radiating independently of
each other at the fundamental frequency w. Secondharmonic scattering is related, but the nonlinear response at 2w is pertinent. Raman scattering differs
from both, in that the scattered light differs from a
harmonic of w by a natural frequency of the medium.
Finally there are effects in which the local dipoles
oscillate coherently, i.e., in definite phase relations with

* Present address: NASA Electronics Research Center, Cambridge, Mass.

t Operated with support from the U.S. Air Force.
1 R. W. Terhune, P. D. Maker, and C. M. Savage, Phys. Rev.
Letters 14, 681 (1965).
2 P. D. Maker, Physics of Quantum Electronics, P. L. Kelley,
B. Lax, and P. E. Tannenwald, Eds. (McGraw-Hill Book Co.,
New York, 1966), p. 60.

each other, such as harmonic generation and stimulated

eIffiSSlOn.

The formulation which Buckingham and Stephen3

(BS) applied to Rayleigh scattering is used. Let the
dipole moment of an isolated molecule in an electric
field Ga be
Pa=j.la+cxa,sC/l+Ma/lP/lG'Y+ha/l'YoGflGrGo+"".

(1)

This can be applied to a molecule in a fluid, where

Ga=Ea+la+Fa, if the fields are sufficiently homogeneous. Ea is the optical field in the medium, Ea+la
is the local optical field, and Fa is the field with no
incident light beam. (The present I a+F a is the Fa of
BS.) Since 2w is well below the frequencies of the
strong absorption bands in the cases to be considered,
the polarizability tensors are approximately symmetrical in all pairs of indices. They will be assumed to be
exactly so. Table I relates the elements of f3a/l'f and of
'Ya/l'Yo for some molecules of interest here.
3 A. D. Buckingham and M. J. Stephen, Trans. Faraday Soc.
53, 884 (1957), BS.

Reuse of AIP Publishing content is subject to the terms: https://publishing.aip.org/authors/rights-and-permissions. Downloaded to IP: 65.110.254.159 On: Sat, 26
Mar 2016 04:50:34