Professional Documents
Culture Documents
Contents
1 Magnetic properties of free lanthanide ions: an overview
1.1 Electronic structure and Hunds rules . . . . . . . . . . . . . . . . . . .
1.2 Energetic effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3 Magnetic behaviour of Ln ions . . . . . . . . . . . . . . . . . . . . . . .
1.3.1 Ground multiplet thermally isolated (E(J, J 1) kB T ) . .
1.3.2 Ground multiplet thermally not isolated (E(J, J 1) kB T ) .
Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
3
3
3
4
4
5
6
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7
7
10
10
12
12
12
12
13
17
17
23
23
.
.
.
.
.
.
.
26
26
27
27
29
30
30
30
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
equation .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3.3.3 4f N systems . . . . . . . . . . . . . . . . . . .
3.3.4 Magnetic behaviour of Cs2 NaLnCl6 (overview)
cub
3.3.5 4f 2 system (Pr3+ , 3H4 : HLF
) . . . . . . . . .
2
3+ 3
cub
3.3.6 4f system (Pr , H4 : HLF + Hmag ) . . . . .
3.3.7 Prediction of LF splittings . . . . . . . . . . .
Problems . . . . . . . . . . . . . . . . . . . . . . . . .
Non-cubic LF . . . . . . . . . . . . . . . . . . . . . .
3.4.1 Ligand field operators . . . . . . . . . . . . .
cyl
3.4.2 4f 1 system (Ce3+ , 2F5/2 : HLF
) . . . . . . . . .
cyl
1
3+ 2
3.4.3 4f system (Ce , F5/2 : HLF + Hmag ) . . . .
Problems . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
35
37
39
41
42
52
53
53
53
54
56
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
56
56
59
59
63
64
64
65
67
69
70
70
71
72
72
73
73
75
76
78
79
80
82
3.4
.
.
.
.
.
.
.
.
.
.
.
5 Solutions
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
83
6 Appendix 1
87
6.1 Magnetic quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
6.2 Energy equivalents and conversion factors . . . . . . . . . . . . . . . . . . 88
7 Appendix 2: Matrix elements of operator equivalents
89
References
91
1.1
The 4f electrons of free Ln ions are influenced by nuclear attraction, interelectronic repulsion (Hee ) and spin-orbit coupling (HSO ), where Hee > HSO . To determine the free ion
ground multiplet 2S+1 LJ the following scheme is helpful (N: no. of 4f electrons):
Ln3+
Ce
Pr
Nd Pm
3
Sm
Eu
Gd
Tb
Dy Ho
Er
Tm
Yb
Lu
11
12
13
14
12 21
12
12 12
2 3
10
ms
+ 12 + 12 + 12
+ 12
+ 12
+ 12
+ 12
21
21
ml
+3 +2 +1
2 3
+3
+2 +1
ms,i = MS S).
P
2. For a given spin multiplicity, the term with highest L lies lowest in energy ( i ml,i =
ML L).
3. For less than half-filled subshells, the level with the lowest value of J lies lowest
(J = |L S|), while the highest J lies lowest when a subshell is more than half full
(J = L + S).
Examples:
1.2
Pr3+
Dy3+
[4f 2 ]:
[4f 9 ]:
S = 1,
S = 25 ,
L = 5,
L = 5,
J = 4;
J = 15
;
2
(3H4 )
(6H15/2 )
Energetic effects
The energy splittings of the electronic states of a magnetic system 1) are produced by Hee ,
HSO , the ligand field (HLF ), interatomic exchange interactions (Hex ), and the applied
magnetic field (Hmag ). The following scheme shows the order of Hee , HSO , and HLF .
3dN
4f N
Hee
>
HLF
>
HSO
weak field
HLF
>
Hee
>
HSO
strong field
HLF
Hee
>
HSO
intermediate field
Hee
>
HSO
>
HLF
(a)
Hee
>
HSO
HLF
(b)
(a) strong field lanthanide system ; (b) weak field lanthanide system
Table 1 gives approximate sizes of the corresponding energy splittings, including Hex
and Hmag .
1)
Effect
Electron-electron interaction
Ligand-field potential
System
3d, 4d, 5d
Hee
4f, 5f
HLF
4f
3d, 4d, 5d
HSO
4f, 5f
nd
Exchange interaction
Hex
4f
nd4f
Magnetic field
a)
b)
wavenumber/cm1 a)
3d > 4d > 5d 104
4f > 5f 104
b)
b)
Spin-orbit coupling
Energy equivalent
Hmag
b)
102
103
b)
b)
102
<1
< 10
0.5 (1 T)
Inspecting Table 1, two points are important for 4f ions: (i) Hex is at least two orders of
magnitude smaller than HLF (and HSO ). (ii) Hmag has to be chosen carefully on account
of comparable magnitude of Hex and Hmag . Therefore, in order to detect Hex , HLF should
be modelled most exactly and Hmag should be very weak.
1.3
The variation of m , (permanent magnetic moment corresponding with the Curie constant C), eff (effective Bohr magneton number2) ), m T , Mm as a function of T and B0
serves to characterize the low-lying electronic states of the magnetic systems.
1.3.1
J(J + 1) B
(except 4f 4 , 4f 5 , 4f 6
systems)
(1)
applies, where J is the total angular momentum of the ground multiplet and gJ the Lande
factor
gJ = 1 +
(2)
S, L, and J correspond to the total spin angular momentum, the total orbital angular
momentum and the total angular momentum, respectively, of the ground multiplet.
2)
Note that the empirical number eff has no connexion with the permanent moment except when
Curies law is obeyed [2].
A graph showing the variation eff vs. T is recommended for 4f/5f systems 3 ) . According to Kramers rule [4], ions with an odd number of 4f electrons (half integral values
of J) have always a degenerate ground state obeying the Curie law at low temperature
(regardless the symmetry of HLF , but presupposed that (i) solely the ligand field (LF)
ground state is occupied and (ii) cooperative magnetic effects are absent). For an even
number of f electrons, degenerate as well as singlet ground states can be found depending on the symmetry of HLF . A singlet ground state produces temperature independent
paramagnetism (TIP) at low temperature.
Going to 5f electron systems, HSO and HLF increase according to the larger effective
nuclear charge and according to the fact that 5f electrons are more accessible for ligands.
No simple approximation can be made. So, the only possibility to predict the magnetic
behaviour for 5f ions are computational methods.
1.3.2
General equation: m
where J
LS J(J + 1)
(2J + 1)J exp
2
kB T
NA 2B J=LS
= 0
(3)
L+S
P
3kB T
LS J(J + 1)
(2J + 1) exp
2
kB T
J=LS
kB T
and LS =
= gJ2 J(J + 1) + 2(gJ 1)(gJ 2)
LS
2S
L+S
P
The gJ values are given by eq. (2) (exception: gJ = L + 2 for J = 0). The term spin-orbit
coupling parameter LS is positive for ions with less than half-full subshell and negative
for ions with more than half-full subshell.
Example 1.1 4f 6 (Sm2+ , Eu3+ )
NA 2B 2
where
3kB T eff
kB T
27
kB T
2
1
144
eff = Z
exp
+
9
6k T
B
kB T
kB T
135
exp
+ 189 21
exp
15
+
2
2kB T
k T
B
kB T
5
5
1485
kB T
+ 405 27
exp
+
exp
33
3kB T
2
2kB T
2457
7
kB T
+
exp
and
39
2
2kB T
+ 5 exp
+ 7 exp
Z =
1 + 3 exp
6kB T
2kB T
kB T
5
5
7
+9 exp
+ 11 exp
+ 13 exp
3kB T
2kBT
2kB T
m = 0
3)
(4)
We recommend this plot or alternatively 2eff vs. T instead of the m T vs. T plot for two reasons:
(i) The number eff has the same numerical value in the SI and CGS-emu
system, (ii) eff is easily
p
compared with the lanthanides corresponding free-ion value eff = gJ J(J + 1) [5].
Ln3+
4f N
2S+1
LJ
4f a)
b)
gJ J gJ [J(J + 1)]1/2 exp
eff
gJ
La3+c) 4f 0
S0
Ce3+
4f 1
F5/2
625
6/7
15/7
2.535
2.32.5
Pr3+
4f 2
H4
758
4/5
16/5
3.578
3.43.6
Nd3+
4f 3
I9/2
3.618
3.43.5
Pm3+
4f 4
I4
1000
3/5
12/5
2.683
2.9d)
Sm3+
4f 5
H5/2
1157
2/7
5/7
0.845
1.6
Eu3+
4f 6
F0
1326
3.5
Gd3+
4f 7
S7/2
1450
7.937
7.87.9
Tb3+
4f 8
F6
1709
3/2
9.721
9.79.8
Dy3+
4f 9
H15/2 1932
4/3
10
10.646
10.210.6
Ho3+
4f 10
I8
2141
5/4
10
10.607
10.310.5
Er3+
4f 11
I15/2
2369
6/5
9.581
9.49.5
Tm3+
4f 12
H6
2628
7/6
7.561
7.5
Yb3+
4f 13
F7/2
2870
8/7
4.536
4.5
Lu3+b) 4f 14
S0
a)
The relation between 4f and LS of the Russell-Saunders ground term is given by LS =
(4f /2S), where (+) and () sign correspond to N 2l + 1 and N 2l + 1, respectively.
b)
Definition: m = 0 NA 2B 2eff /(3 kB T ).
c)
diamagnetic
d)
observed for Nd2+ compounds.
Problems
1. Verify the Russell-Saunders ground multiplets for Nd2+ and Er3+ given in Table 2.
2. Evaluate the susceptibility equation for a free p1 system perturbed by spin-orbit
coupling. (Hint: Apply eq. (3). Note that for the single-electron system LS .
Solutions are given in section 5.)
6
2.1
Functions
i
h
2 2
eV (r) (r) = E(r)
2me
(5)
For convenience the eigenfunctions (atomic orbitals) are given in spherical polar coordinates:
z
x = r sin cos
y = r sin sin
z = r cos
P(x,y,z)
r. cos q
q
r. sin q. cos f
.
f r sin q
r. sin q. sin f
(6)
r 2 = x2 + y 2 + z 2
cos = z/r
tan = y/x
(7)
Fig. 1:
Relation between cartesian coordinates and spherical polar coordinates
Yml l (, )
1 iml
e
2
The functions Yml l (, ) are spherical harmonics specified by the quantum numbers l
and ml (Table 3). Real functions (Table 4) are gained by linear combinations
1
2
1
i 2
[n,l,ml + n,l,ml ] =
[n,l,ml n,l,ml ] =
1
2
1
i 2
[n,l,ml + n,l,ml ] =
[n,l,ml n,l,ml ] =
1 Rn,l (r)l
ml ()
1
Rn,l (r)l
ml ()
1 Rn,l (r)l ()
ml
1 Rn,l (r)l ()
ml
cos ml
ml odd
sin ml
ml odd
(8)
cos ml
ml even
sin ml
ml even
For a complete description of the wave function, the spin has to be taken into consideration. Ignoring spin-orbit coupling, the total function (spin orbital) reads
(r, , ; ) = (r, , ) () where = 21 .
(9)
l ml
0
0
0
1
1
0
2 1
2
0
1
2
3
0
1
4 2
3
4
a)
Yml l (, ) a)
Yml l (x, y, z)
1 1/2
4
3 1/2
cos
4
3 1/2
sin ei
8
5 1/2
(3 cos2 1)
16
15 1/2
cos sin ei
8
15 1/2
sin2 ei2
32
7 1/2
(5 cos3 3 cos )
16
21 1/2
sin (5 cos2 1)ei
64
105 1/2
cos sin2 ei2
32
35 1/2
sin3 ei3
64
9 1/2
(35 cos4 30 cos2 + 3)
256
45 1/2
64
315 1/2
sin4 ei4
512
1 1/2
4
3 1/2 z
4
r
3 1/2 x iy
8
r
5 1/2 3z 2 r2
16
r2
15 1/2 z(x iy)
8
r2
15 1/2 (x iy)2
32
r2
7 1/2 z(5z 2 3r2 )
16
r3
21 1/2
(5z 2 r2)
(x iy)
64
r3
105 1/2 z(x iy)2
32
r3
35 1/2 (x iy)3
64
r3
9 1/2 (35z 4 30z 2r2 + 3r4)
256
r4
45 1/2
(7z 3 3zr2 )
(x iy)
64
r4
45 1/2
2
2
2 (7z r )
(x iy)
128
r4
315 1/2 z(x iy)3
64
r4
315 1/2 (x iy)4
512
r4
Phase factors correspond to the Condon-Shortley convention: 1 for odd positive ml and +1 otherwise.
Table 4: Real orthonormal linear combinations of the spherical harmonics Yml l (, ) for
l = 1, 2, 3.
function
3 1/2
cos
4
3 1/2
1
sin cos
4
3 1/2
sin sin
4
5 1/2
(3 cos2 1)
16
15 1/2
cos sin cos
4
15 1/2
2
cos sin sin
4
15 1/2
sin2 cos 2
16
15 1/2
sin2 sin 2
16
7 1/2
(5 cos3 3 cos )
16
21 1/2
sin (5 cos2 1) cos
32
21 1/2
sin (5 cos2 1) sin
32
105 1/2
3
cos sin2 sin 2
16
105 1/2
cos sin2 cos 2
16
35 1/2
sin3 cos 3
32
35 1/2
sin3 sin 3
32
designation
3 1/2 z
=
4
r
3 1/2 x
=
4
r
3 1/2 y
=
4
r
5 1/2 3z 2 r2
=
16
r2
15 1/2 xz
=
4
r2
15 1/2 yz
=
4
r2
15 1/2 x2 y 2
=
16
r2
15 1/2 xy
=
4
r2
7 1/2 z(5z 2 3r2)
=
16
r3
21 1/2 x(5z 2 r2 )
=
32
r3
21 1/2 y(5z 2 r2 )
=
32
r3
105 1/2 xyz
=
4
r3
105 1/2 z(x2 y 2 )
=
16
r3
35 1/2 x(x2 3y 2)
=
32
r3
35 1/2 y(3x2 y 2 )
=
32
r3
9
pz
px
py
dz 2
dxz
dyz
dx2 y2
dxy
fz 3
fxz 2
fyz 2
fxyz
fz(x2 y2)
fx(x2 3y2)
fy(3x2 y2 )
2.2
2.2.1
Angular momenta
Orbital angular momentum
(10)
r
p
Fig. 2: Definition of the angular momentum
l = lx i + ly j + lz k
= (ypz zpy )i + (zpx xpz )j + (xpy ypx )k
(11)
(12)
pq pq =
h
i q
(q = x, y, z;
i=
1)
lx = ypz zpy ;
ly = zpx xpz ;
lz = xpy ypx
lx = h
y z
; ly =
z x
; lz =
x
y
i
z
y
i
x
z
i
y
x
(13)
(14)
Commutation relations
[lx , ly ] = i
hlz ;
[ly , lz ] = i
hlx ;
[lz , lx ] = i
hly ,
(15)
(16)
l ml lz l ml
= ml h
l ml l ml
{z
}
|
1
10
= E , ( normalised eigenfunction of H)
generally H
Z
Z
d = E d = E
H
| {z }
1
Z
d h|H|i
H
matrix element (Dirac notation)
Application of lz :
lz 2 2
=
=
=
=
l2 l ml =
h
Y22 (, )
i
#
"
1/2
h
15
Rn,2 (r)
sin2 ei2
i
32
1/2
h
15
ei2
Rn,2 (r)
sin2
i 32
1/2
h
15
Rn,2 (r)i2
sin2 ei2 = 2
h22
i 32
l(l + 1)
h2 l ml
Rn,2 (r)
(17)
Shift operators:
l+ = lx + ily ;
l = lx ily .
(18)
Reverse operations:
ly = 1 (l+ l ).
lx = 1 (l+ + l );
2
2i
lz l ml
= ml h
l ml
l2 l ml
= l(l + 1) h
2 l ml
p
l l ml
l(l + 1) ml (ml 1) h
l ml 1 .
=
(19)
(20)
ml =+2
+1
1
2
11
2.2.2
Spin
ms = 12 :
Spin function: () | s ms i
ms = 12 :
s2 s ms = s(s + 1) h
2 s ms with s = 21
s ms with ms = 12
sz s ms = ms h
p
s(s + 1) ms (ms 1) h
s ms 1
s s ms =
(21)
2.3
Quantum-mechanical procedures
2.3.1
Operators
Postulate 1. The state of a system is fully described by the wavefunction (r1 , r2 , . . . , t).
Postulate 2. Observables are represented by operators chosen to satisfy the commutation
relation
(22)
qpq pq q = [
q , pq ] = i
h
(q = x, y, z; i = 1)
Example 2.1 Verification of the commutation relation
x = x
and px =
h
d
i dx
h
d
i dx
h
d(x)
d
h
px x =
+x
=
i dx
i
dx
h
(
xpx px x) = = i
h
i
xpx = x
12
quantity Da)
Ekin
1
3
Epot c)
a)
b)
c)
classical
p2x
me vx2
=
2
2me
2
p
2me
quantum-mechanical
2
h
d
h
2 d2
1
p2x
=
=
2m
2me i dx
2me dx2
e
2
2
2
h
2 2
h
2
+
+
2me x2 y 2 z 2
2me
eV (x)
eV (x) = eV (x)
eV (r)
eV (r) = eV (r)
b)
Dimension.
is the Nabla operator.
Valid for one electron with charge e in the potential V .
2.3.2
Perturbation theory
1. Non-degenerate states
(0) (0) = E (0) (0)
H
n
n
n
unperturbed system
(23)
find En , n
(24)
(25)
(26)
2
..
.
(1) )(1)
(H En )n = En n + (En H
n
..
.
(27)
(28)
(29)
(1)
(1) |n
En(1) = n|H
(30)
(1) |n
X m|H
(1)
(0)
n =
m
(0)
(0)
m6=n Em En
(31)
(1) |n 2
X m|H
En(2) =
(see Fig. 4)
(0)
(0)
m6=n Em En
(32)
E
Y2(0)
E 2(0)
Y2(0)
H12
(0)
E 1(0)
E 2(0)
E 2(0)
H22
Y1
E 2(0) H22
H21 H12
E 1(0)
E 2(0)
Y1(0)
E 1(0)
Y1(0)
H21
(0)
E 2(0)
E 1(0)
E 1(0)
H 11
Y2
E 1(0) H11
(a)
H12 H21
E 2(0)
E 1(0)
(c)
(b)
Fig. 4: Illustration of the possible effects of a perturbation on two nondegenerate levels; (a) 0th ,(b) 1st ,(c) 2nd order
14
2. Degenerate states
Eq. (30) (32) apply also in this case; additionally: determination of the correct zerothorder wavefunctions (see Fig. 5)
Example: doubly degenerate pair of states
(0) (0) = E (0) (0)
H
n
n,i
n,i
(0)
(i = 1, 2)
(33)
(0)
(0)
n = u1 n,1 + u2 n,2
(34)
The correct linear combinations are those which correspond to the perturbed functions
for 0.
Determination of u1 and u2 :
(35)
(0)
where Hij =
(0) (1) (0)
(1) |j
n,i H
n,j d = i|H
To find the non-trivial solutions of this pair of equations, the determinant of the coefficients
of the constants u1 and u2 must disappear:
H E (1)
H12
H11 , H22 : diagonal elements
n
11
(36)
(1) = 0;
H21
H22 En
H12 , H21 : off-diagonal elements
p
(1)
En(1,2) = H11 + H22 /2 (H11 H22 )2 /4 + |H12 |2
(1)
x(1,2) =
u1(1,2)
H12
=
(1)
u2(1,2)
H11 E
n(1,2)
Normalisation:
q
u
=
2(1,2)
x2(1,2) + 1
x(1,2)
x2(1,2) + 1
(1)
(0)
(0)
15
E
Y2(0)
Y2(0)
E 2(0)
E 2(0)
(0)
(0)
u 2(2) Y1,2
(0)
(0)
u 2(1) Y1,2
(0)
Y1(2) = u 1(2) Y1,1
(0)
(0)
Y1,1
, Y1,2
E 1(0)
E 2(0) H22
(0)
Y1(1) = u 1(1) Y1,1
E 2(0)
H22
E 1(0)
E 1(2)
(1)
H21(2) H1(2)2
(1)
E 1(0) E 1(2)
E 1(0)
E 1(0)
E 2(0)
(1)
H21(1) H1(1)2
(1)
(a)
E 1(0)
E 1(1)
E 1(0) E 1(1)
(a)
E 2(0)
(b)
E 2(0)
E 1(0)
(c)
Fig. 5: Illustration of the possible effects of a perturbation on a doubly degenerate ground state and a non-degenerate excited state; (a) 0th , (a) correct
0th , (b) 1st ,(c) 2nd order
16
2.4
Applications
Spin-orbit coupling (p1 , f 1 , f 13 )
2.4.1
where
(r) =
p1 system: l = 1, s = 12 , j1 = l s =
1
2
e 1 V (r)
.
2m2e c2 r r
and j2 = l + s =
(0)
3
2
n/cm-1
spin-orbit coupling
(4)
(38)
16973
1z
2
P3/2
17cm
16956
P1/2
(6)
1
(2)
off
S 1/ 2
-z
on
l = 589.8nm
589.2nm
Fig. 7:
Term scheme of the
sodium atom
(i = 1, 2, . . . , 6).
(39)
(0)
17
(40)
(0)
0 = u1 H11 E (1) + u2 H12 + + u6 H16
0 = u1 H21 + u2 H22 E (1) + + u6 H26
..
..
.
.
0 = u1 H61 + u2 H62 + + u6 H66 E (1)
(41)
(42)
(43)
1/2
hc n,l = h
(44)
1 1 .
2
hcn,l
m
m
l
s ml ms .
l
s
h
2
(45)
(46)
l
s = lz sz + 21 (l+ + l ) 12 (
s+ + s ) + 2i1 (l+ l ) 2i1 (
s+ s )
1
= lz sz + 4 (l+ s+ + l s+ + l+ s + l s l+ s+ + l s+ + l+ s l s )
= lz sz + 1 (l+ s + l s+ )
2
l
s ml ms = ml ms lz sz + 21 (l+ s + l s+ ) ml ms
ml ms
= ml ms lz sz ml ms
+ 21 ml ms l+ s ml ms
+ 12 ml ms l s+ ml ms .
(1)
n :
General hints to the evaluation of matrix elements mH
18
(47)
(48)
(1) n . This will result in a constant a multiplied by a wavefunction which
(i) Evaluate H
(1)
may or may not be the same as the original. For
present
us
the
let
assume H n = a n .
(ii) The result of (i) is then premultiplied
man .
by
m giving
(iii) Since a
is a constant
we have
= a m n and we are
thus left
m an
with the task of
evaluating m n . Provided m and n are orthonormalised, m n = 1 when m = n
but is zero otherwise.
On account of orthonormalised states
(49)
ml ms ml ms = ml ,ml ms ,ms ,
the integral is not zero when ml = ml and ms = ms . The wavefunctions are eigenfunctions
of lz und sz , so that the application of the operator products in eq. (48) on the wavefunction
on its right-hand side yields:
lz sz ml ms = ml ms h
2 ml ms
l+ s ml ms =
p
p
l(l + 1) ml (ml + 1) s(s + 1) ms (ms 1) h
2 ml + 1 ms 1
l s+ ml ms =
p
p
l(l + 1) ml (ml 1) s(s + 1) ms (ms + 1) h
2 ml 1 ms + 1
where s = 12 . For diagonal elements only lz sz may contribute, whereas for off-diagonal
elements only the step operators may account:
ml ms l s+ ml + 1 ms 1
ml ms l+ s ml 1 ms + 1
Matrix elements (46) which may contribute are restricted to the condition
ml + ms = ml + ms
(50)
1 12 l+ s 0 21
0 12 l+ s 1 12
1
1 21 l s+ 0 12 .
0 2 l s+ 1 12
ml ms
1
1 2
1 12
1
0 2
0 12
1 21
1 12
1
1 1
0
1
2
2
1
2
21
1
2
1
2
1
2
0 1
1 1
1 1
2
2
2
0
0
q
1
2
1
2
21
1
2
19
(51)
H11 =
H22 =
H33 =
H44 =
H55 =
H66 =
H23 =
H45 =
1 1
z 1 2 = 1 12 1 21 1 21 = 21
2 1 2 lz s
h
1
1 1 = 1 ( 1 ) = 1
l
s
z
z
2
2
2
2
h
2
0 12 lz sz 0 12 = 0 12 = 0
h
2
1
z 0 12 = 0 ( 12 ) = 0
2 0 2 lz s
h
1
1 1 = (1) 1 = 1
1
l
s
z
z
2
2
2
2
h
2
1 12 lz sz 1 12 = (1) ( 12 ) = 12
h
2
1 1
q1
1 1
1
1
1
l
s
0
=
= 2 = H32
2
1
1
2 2 +
2
2
2
2
h
2
q
1
1
1 1
1 2 = 2 2 1 = 12 = H54
2 0 2 2 l+ s
h
2 E (1)
2
= ( 1 E (1) )(E (1) ) 1 2 = 0
q
2
2
1
(1)
E
2
(1)
E(1) =
1
2
(1)
E(2) = .
(1)
(52)
(1)
u2(2)
= 2;
u2(2) = 23 ;
x(2) =
u3(2)
q
q1 1
1
2
3 = 3 1 2 + 3 0 2 .
u3(2) =
1
3
3
2
3
2
The results are given in Table 6.
20
(53)
(54)
(55)
l2 +
l
s + s
l + s2 =
l + s 2 = 2 .
(56)
If 2 acts on a quartet state function Q (1 , 2 , 4 , 6 ) and on a doublet state function
D (3 , 5 ), respectively, the result is
2 Q =
l2 + 2
l
s + s2 Q
2 ( 15
) Q = h
2 ( 32 )( 52 ) Q
= h
2 2 + 2 12 + 34 Q = h
4
2 D
= h
2 j2 (j2 + 1) Q
where j2 = 23
=
l2 + 2
l
s + s2 D
= h
2 2 2 1 + 34 D = h
2 ( 34 ) D = h
2 ( 21 )( 32 ) D
= h
2 j1 (j1 + 1) D
(57)
where j1 = 21 .
(58)
ml ms
1 1 12
q
q2 1
1
1
2
3 1 2 +
3 0 2
q
q2
1
1
0 1
4
+
1
3
2
3
2
6 1 21
q
q1 1
2
1
3 3 1 2 + 3 0 2
q
q1
2
1
0 1
5
1
3
2
3
2
j mj
mj = ml + ms j E (1)
3 3
2 2
3 1
2 2
3 1
2
2
3 3
2
2
1 1
2 2
1 1
2
2
3
2
1
2
21
3
2
1
2
1
2
23
1
2
21
Ln3+ [4f ]
ground multiplet
Ce3+ [4f 1 ]
Yb3+ [4f 13 ]
F5/2 (j = 5/2)
F7/2 (j = 7/2)
E
2Ce
32 Yb
excited multiplet
+ 23 Ce
+2Yb
F7/2 (j = 7/2)
F5/2 (j = 5/2)
Eigenfunctions are obtained with the help of vector coupling coefficients (Clebsch-Gordan
coefficients) (see Table 7).
21
m2 =
j1 +
1
2
j1
1
2
1
2
j1 + m + 21
2j + 1
s 1
j1 m + 12
2j1 + 1
m2 = 21
s
j1 m +
2j1 + 1
s
j1 + m +
2j1 + 1
1
2
1
2
a)
1 ; 6 52 52
2 ; 5 52 32
3 ; 4 52 12
1 ; 8 72 72
2 ; 7 72 52
3 ; 6 72 32
4 ; 5 72 12
(Ce3+ )
(Yb3+ )
ml ms
EJ
EJ
q
q1
1
6
7 3 2 7 2 12
q
q2
1
5
2
2
7 2 2 7 1 12
q
q
47 1 12 37 0 12
3 1
q 2
q6
1
1
3
+ 7 2 12
7
2
3
q
32
q5
2
1
1
2
2 2 + 7 1 2
q7
q4
3
1
1
+ 7 0 12
7
2
=
=
=
=
=
=
=
a)
Short form of the functions: the first symbol refers to the upper sign, the second
to the lower one.
The calculation of the coefficients is demonstrated for 52 25 (first line in Table 8).
Assignments: j = 5/2, m = mj = 5/2, j1 = l = 3, and m2 = ms = 1/2 (j2 = s = 1/2)
The roots of the lower row of Table 7 become
s
s
q
1
j1 + m + 2
3 + 52 21
m2 = 12 :
=
= 67
2j1 + 1
7
s
s
q
j1 m + 12
3 52 + 21
1
m2 = 2 :
=
= 17
2j1 + 1
7
Since the Condon-Shortley standard assignment is j1 s and j2 l the sign of the coefficients has to be changed according to the phase relation |jb ja jmi = (1)ja +jb j |ja jb jmi.
Finally, weq
obtain
5 5
q
= 6 3 1 + 1 2 + 1 .
2 2
22
Problems
4. Calculate the matrix elements hl, ml |lq |l, ml i (where q stands for z, +, ):
2
(a) h0, 0|lz |0, 0i, (b) h2, 2|l+ |2, 1i, (c) h2, 2|l+
|2, 0i, (d) h2, 0|l+l |2, 0i.
5. The 14 microstates |ml ms i of a f 1 system (l = 3, s = 12 ) yield under the influence
of spin-orbit coupling 14 eigenstates |jmj i which, apart from the states | 27 72 i,
are linear combinations of two microstates each. Use Table 7 to evaluate the vector
coupling coefficients for the coupled states | 52 12 i, | 25 12 i, | 27 32 i, and | 27 32 i. Control
your results with the entries of Table 8.
6. What levels (multiplets J) may arise from the terms (a) 1 S, (b) 2 P , (c) 3 P , (d) 3 D,
(e) 4 D? How many states (distinguished by the quantum number MJ ) belong to
each level?
2.4.2
n exp(En /kB T )
(En /B) exp(En /kB T )
n
n
P
= NA P
Mm = NA
exp(En /kB T )
exp(En /kB T )
n
(59)
(2)
n = En /B = Wn(1) 2BWn(2) . . .
P
(1)
(2)
(0)
[(Wn )2 /kB T 2Wn ] exp(Wn /kB T )
m = 0 NA n
P
(0)
exp(Wn /kB T )
(60)
23
(61)
(62)
(63)
(64)
(65)
(66)
(1)
z 1
Calculation of the off-diagonal element 2 H
5 5
(1)
1 = 7 5 e (lz + 2
s
)
2 H
z
z
2 2
2 2
q
q
Dq
q1
E
6
1
1
1
1
1
6
l
+
2
s
+
3
2
+
= e
3
2
+
z
7
2
7
2 z
7
2
7
2
e h
sz 3 12 3 12 lz + 2
sz 2 + 12
6 3 12 lz + 2
7
{z
}
{z
}
|
|
0
(3 1)
h
i
6
1
1
1
1
+ 6 2 + 2 lz + 2
B
sz 3 2 6 2 + 2 lz + 2
sz 2 + 2 =
7
|
|
{z
}
{z
}
0
(2 + 1)
h
(2)
5
i
2 2
(2)
W| 5 5 i
2 2
2
75
h
| e (lz + 2
sz )| 25 52 i
2 2
(0)
(0)
W7/2 W5/2
(6/49) 2B
12 2B
=
.
(7/2)
343
24
(67)
(0)
J MJ
Wn
5 5
2
2
5 3
2
2
2
5 1
2
2
7 7
2
2
7 5
2
2
3
7 3 2
2
2
7 1
2
2
(0)
Tab. 9: Energies Wn and Zeeman co(1)
(2)
efficients Wn , Wn of a free f 1 system
(gJ1 = 6/7 for 2F5/2 , gJ2 = 8/7 for 2F7/2 )
(0)
(1)
(1)
Wn /B
(2)
Wn /2B
(2)
where 2eff =
3kB T eff
144 12kB T
7
45 16kB T
+4
exp
+
3
7
49
7
49
2kB T
.
(68)
7
3 + 4 exp
2kB T
For Ce3+ the parameter in eq. (68) refers to the single-electron spin-orbit coupling constant while in the case of Yb3+ the term spin-orbit coupling constant LS = has to be
replaced for . (Notice that LS = /(2S).)
The absolute energy separation between 2F7/2 and 2F5/2 is 27 Yb = 72 2870 cm1
10 000 cm1 for the Yb3+ ion and 27 Ce = 27 625 cm1 2 200 cm1 for the Ce3+ -Ion.
In magnetochemical practice the higher lying multiplet is often ignored in eq. (68). There
are two possibilities for simplification:
1.) (4f 1 ) and LS (4f 13 ) respectively:
m (4f 1 ) = 0
2eff (4f 1 )
NA 2B 45
3kB T |{z}
7
2eff
45
=
=
7
m (4f 13 ) = 0
2
5
7
6
7
2
2
NA 2B 144
3kB T |{z}
7
2eff
2eff (4f 13 )
144
=
=
7
(69)
2
7
9
8
.
7
2
2
ground
multiplet
3.1
f 1 f 13
2
F 5 2F 7
2
f 4 f 10
5
I4 5I8
f 3 f 11
4
I 9 4I 15
f 2 f 12
3
H4 3H6
f5 f9
6
H 5 6H 15
2
f6 f8
7
F0 7F6
f7
(71)
S7
X
+k
X
Aqk r k Cqk (, ),
where
(72)
k=0 q=k
(R) k
= e(1)
C (, ) dR
Rk+1 q
{z
}
|
geometrical coordination factor
Aqk
and
Cqk
|
1/2
4
=
Yqk (, )
2k + 1
{z
}
Racah tensor
for f 1 system
cub
H
LF
z
}|
{
h
h
p
p
i
i
6
0 6
4
0 4
4
4
6
6
= A4 r C0 + 5/14 C4 + C4 +A6 r C0 7/2 C4 + C4 , where
{z
}
|
(73)
for d1 system
C04 =
4
C4
=
C06 =
6
C4
1
(35 cos4 30 cos2 + 3)
8
1
70 sin4 ei4
16
1
(231 cos6 315 cos4 + 105 cos2
16
3
14 sin4 (11 cos2 1) ei4
32
5)
LF n,l,m (r, , )d
n,l,m
(r, , ) H
l
l
(74)
p
4
Ymll (, ) [C04 + 5/14(C44 + C4
A04
Rn,l (r) r 4 Rn,l (r) r 2 dr
)] Yml (, ) sin d d
l
{z
} 0 0
|0
radial integral < r 4 >
26
h
h
i
p
p
i
4
cub
4
6
4
4
6
6
6
(75)
With the help of integrals of the type hYml l |Cqk |Yml i [3], listed in Table 10 (p electrons)
l
and Table 13 (f electrons), LF effects can be evaluated. The k values in eq. (72) comprise
terms with k 2l and k even. The term with k = 0 has spherical symmetry, produces
no splitting and is therefore omitted. The relevant k values for LF terms are k = 2 (p
electrons), k = 2 and 4 (d electrons), k = 2, 4, and 6 (f electrons). Notice that the
operators Cqk act only on the orbital part of a wave function.
Introduction: p1 system
3.2
3.2.1
cyl
Cylindrical LF: HLF
+ HSO
(76)
With the help of the matrix components of the electrostatic interactions, shown in
Table 10, the 3 3 matrix (77) is obtained. Notice that only the red numbers corresponding to c2 are relevant. Since the operator C02 does not change ml , matrix elements
< ml | C02 | ml > with ml 6= ml are automatically zero.
Table 10: Matrix components of electrostatic interactions, ck (lml , l ml ) for l = l =
ml
ml
c0
5 c2 a)
1 1 +1 1
1
0 0 + 3
0
0 +1 +2
1 1 0 6
a)
ml
3 3 matrix of operator (76):
|1i
h 1 | 51 B02
h 1 |
h0|
| 1 i | 0 i
15 B02
(77)
2 2
B
5 0
Results: (i) All off-diagonal matrix elements are zero; (ii) LF operator eq. (76) splits
the threefold degenerate p1 system (in the absence of spin-orbit coupling) into an orbital
doublet | 1i with energy ELF = B02 /5 and a singlet | 0 i with energy ELF = 2B02 /5.
4)
27
Example 3.2 The p1 system under the action of spin-orbit coupling and a cylindrical
ligand field (cf. spin-orbit matrix (51))
+H
cyl = (r)
H
l
s + B02 C02
SO
LF
ml ms
1 21
0 21
1 12
12
1
2
1
2
1 1
2
B02 /5
(78)
1
0
0 1
2
2
q
1
21 B02 /5
2
q
1
2B02 /5
2
1 12
12 B02 /5
1
2
B02 /5
1/5
2/5
1 1
= 2 0 12 lz sz 0 12 + B02 0C02 0
= 2B02 /5
{z
}
| {z } | 2 {z 2 }
h
|
0
H55 =
H66 =
1/5
1 1 1
1
1 1 + B02 1C02 1
2 2 = 2 B02 /5
1
l
s
z
z
2
2
2
|
{z
}
{z
}|
{z
}
h
|
2 /2
h
H23 =
2/5
h2 /2
1 1
0
=
l
s
2 2
2
{z
}
h
2 |
2/2
H45
1/5
H44
1
2
1
2
1 1
= 2 1 12 lz sz 1 12 + B02 1C02 1
= 12 B02 /5
| {z } | 2 {z 2 }
{z
}
h
|
h2 /2
H33
1 1
2
1 1 1
1 1
2
+
B
= B02 /5
l
s
1
1
z
z
0 1 C0 1
2
}
| {z } | 2 {z 2 } 2
h
2 | 2 {z
2 /2
h
H22
2B02 /5
1 21
H11 =
1 1
1/5
1
2
= H32
q1
1
1 1
= 2 0 2 2 l+ s 1 2 = 2 = H54
{z
}
h
|
2/2
2
2 H23
=
=
tan 2 =
H22 H33
/2 + 3B02 /5
1
2
H23 = H32 = / 2
(79)
3 2
B /
5 0
cyl
Result: The combined effect of HSO + HLF
splits the sixfold degenerate spin orbital
states into three doublets. The composition of the wave functions depends on the ratio
B02 /. No simple closed expressions can be given for the functions.
3.2.2
or
Orthorhombic LF: HLF
+ HSO
HLF = B0 C0 + B2 C2 + C2
where C2 = 38 sin2 exp(2i)
ml
|1i
h1|
15 B02
h 1 |
6
5
| 1 i
6
5
(80)
|0i
B22
15 B02
B22
2 2
B
5 0
h0|
Determination of the correct zeroth order functions by application of scheme eq. (37):
2 6 B22 /5
B02 /5 + B02 /5
6 B22 /5
0
2 = 90 = 45 ; cot 45 = 1
Result: Neglecting spin, the orthorhombic ligand field splits the threefold degenerate
orbital states into three singlets.
Example 3.4 The p1 system under the action of spin-orbit coupling and an orthorhombic
ligand field
2
+H
or = (r)
H
l
s + B02 C02 + B22 (C22 + C2
)
SO
LF
(81)
The matrix elements of this operator are stated in scheme (82). The E (1) s in the
diagonal elements are omitted and also the multiplicative factor hc for the spin-orbit
coupling matrix elements which are usually given in the energy equivalent cm1 . The
empty off-diagonal elements are automatically zero. The diagonalisation of the 3 3
29
blocks, the energies and the composition of the wave functions must be calculated with
the aid of a computer.
ml ms
1 21
0 12
12
1 21
1
2
1
2
B02 /5
0
6B22 /5
0 1
2
0
2B02 /5
q
1
1 1
2
6B22 /5
q
1
3.3.1
12 B02 /5
12
0 21
1
2
B02 /5
6B22 /5
1 12
3.3
1 1
1
0
1 1
6B22 /5
1
2
2B02 /5
0
(82)
0
1
2
B02 /5
Cubic LF
Group theoretical tools I [9, 8]
Character tables of the point groups serve Tab. 11: Character table of the symmetry
to assign the LF states with respect to symgroups O and Td
metry and degeration. For the cubic sysO E 8 C3 3 C2 6 C4 6 C2
tems O and Td Table 11 is relevant. It is
applicable to Ln ions with integer J. The Td E 8 C3 3 C2 6 S4 6 d
degree of degeneration of the LF states is re- A1 a) 1
1
1
1
1 1 b)
sponsible for the orbital contribution of the A
1
1
1 1 1 2
2
lanthanides magnetic moment: The singlet
E
2 1
2
0
0 3
states A1 (1 ) and A2 (2 ) as well as the dou0 1
1 1 4
blet state E(3 ) are non-magnetic, while the T1 3
0 1 1
1 5
triplet states T1 (4 ) and T2 (5 ) are mag- T2 3
netic. For half-integer J all states are on
a)
Mulliken nomenclature.
principle magnetic; for cubic LF character
b)
Bethe nomenclature.
cub
4f 1 system (Ce3+ , 2F : HSO + HLF
)
Ce3+ is a half-integer system. So, O , Td are relevant. Under the action of the LF operator
q
q
n
o
o
n
6
4
4
6
6
6
4
cub
1
4
5
7
+
B0 C0 2 [C4 + C4 ] ,
HLF (4f ) = B0 C0 + 14 [C4 + C4 ]
|
{z
}
irrelevant for restricted basis 2F5/2
the free ion multiplets split according to 2F5/2 G (8 ) + E (7 ) and 2F7/2 G (8 )
+ E (7 ) + E (6 ). If HSO is distinctly stronger than HLF it is convenient to consider
30
4 C3
4 C23
3 C2
3 C34
3 C4
6 C2
4 C23 R 4 C3 R 3 C2 R 3 C34 R 3 C4 R 6 C2 R
4 C3
4 C23
3 C2
3 S34
3 S4
6 d
Td
A1 (1 )a)
A2 (2 )
E(3 )
T1 (4 )
T2 (5 )
E (6 )b)
2 2
4 4
E (7 )
2 2
G (8 )
a)
b)
4 C23 R
4 C3 R 3 C2 R
1
1
3 S34 R
3 S4 R 6 d R
1
0
0
1 5 : integer J; 1 3 : non-magnetic.
6 8 : half-integer J; magnetic.
HLF acting only on the ground multiplet 2F5/2 [ G (8 ) + E (7 )]. This approximation
cub can be restricted to the fourth degree term (k = 4), since
has the advantage that H
LF
the corresponding LF matrix elements of the sixth degree term are zero. Using Table 8
cub in scheme (83) are obtained.
and Table 13 the matrix elements of H
LF
Normally, the spin-orbit coupled states |J MJ i are used as a basis for the multiplet.
In the following examples, however, we write them as |ML MS i basis with the advantage
cub on the orbital function |M i becomes obvious.
that the action of H
LF
L
Table 13: Matrix components of electrostatic interactions, ck (lml , l ml ) for l =
ml
c0
3 3 +1
3 2
3 1
2 2 +1
2 1
2
0
0
1 1 +1
15 c2 a) 33 c4 a)
5
+5
10
0
+ 15
20
+3
+3
30
+ 54
63
429
5
c6 a)
ml
ml
c0
+ 7
28
+ 84
7
+6
+ 32 105
3 + 224
+1
15
0 +1
15 c2
+ 2
33 c4
+ 15
+4
+6
0
3 3
3 1
2 1
3 2
2 2
1 1
0
+ 42
0
+ 70
0
14
24 40
429
5
c6
+ 350
+20
924
+ 462
210
+ 504
378
420
The numerical factor is the denominator for all ck values of a column. (Notice that only
the red terms are relevant in Example 3.5.)
a)
31
cub 5
LF
| 2 i of matrix (83)
Example 3.5 Evaluation of the diagonal element H11 = h 25 |H
H11 =
1
7
cub
6 | 3, 1 i + | 2, 1 i
6 h 3, 12 | + h 2, 12 | H
LF
2
2
(i) Integration with respect to the spin, (ii) insert of ck (lml , l ml ) values 5) :
cub | 3 i + h 2 |H
cub | 2 i =
H11 = 17 6 h 3 |H
LF
LF
h
i
4
6
6
4
1
6
4
+
h
2
|C
|
2
i
+
B
6
h
3
|C
|
3
i
+
h
2
|C
|
2
i
=
B
6
h
3
|C
|
3
i
0
0
0
0
7
| {z0 }
|
{z
}
| {z0 }
c4 (33,33)=3/33
1
21
B04 .
c4 (22,22)=7/33
3 5 3 1 1
4f 1 (cub) 52
2
2
2
2
2
5
1
5
2
21
21
3 5
3
2 21
21
5
1
5
2
21
21
3
3
5
21
2
21
1
2
2
21
1
2
2
21
(83)
cub | 3 i
Example 3.6 Evaluation of the off-diagonal element h 25 |H
LF
2
H12 =
1
7
H12 =
1
7
cub
2 | 1, 1 i + 5 | 2, 1 i
6 h3, 12 | + h2, 21 | H
LF
2
2
cub | 1 i + 5 h 2 |H
cub | 2 i =
12 h 3 |H
LF
LF
h q
i
4
4
1
5
4
B
12
h
3
|C
|
1
i
+
5
h
2
|C
|
2
i
0
4
4
7
14
| {z
| {z
}
}
42/33
70/33
q
i
h
=
72 B06
12h 3 |C46 | 1i + 5 h 2 |C46 | 2 i
{z
}
|
5
21
B04
The HLF matrix (83) has two identical 1 1 and two identical 2 2 blocks. The former
two correspond to the |MJ i states | 21 i and | 12 i yielding the first-order correction to the
energy E (1) = ELF = (2/21) B04. The energy of the other states is obtained by applying
eqn. (36):
1 4
5
B0 E
E1 = 2 B04
B04
21
21
21
=
0
=
3
5
E2 = 4 B 4 .
B04
21
B04 E
0
21
21
32
ELF
2
B4
21 0
|G , i
c)
4
21
B04
G 8 ; E 7 .
For the states in short cf. ref. [6].
c)
Phase
P for Kramers doublets:
|i = J,M CJ,M |JM i
P
|i = J,M CJ,M
(1)JM |J M i.
a)
b)
Magnetic susceptibility
Zeeman operator (eq. (66)):
mag,z = g J B
H
e J z z
where e = e/(2me ),
gJ = 6/7
Application of the z component of the Zeeman operator onto the six symmetry-adapted
linear combinations of the ground multiplet 2F5/2 (Table 14) yields the matrix elements
in Table 15 (in units of gJ B Bz and = ELF (G ) ELF (E ) = (6/21) B04).
(1)
(0)
Wn
(0)
(1)
(2)
|G, i
|G , i
i
|E M
Diagonal element:
mag,z |G i
H11 = hG |H
5 5
= ( 6 )( 2 ) + ( 16 )( 23 ) gJ B Bz = ( 11
)gJ B Bz
6
Off-diagonal element:
mag |E i
H15 = hG |H
q q
i
hq q z
1
5 5
1
5
3
2 5
(
)
(
)
g
B
=
(
) gJ B Bz .
=
J B z
6
6 2
6
6
2
3
33
Wn
gJ B
(2)
Wn
gJ2 2B
11/6 +
20
9
1/2
5/6
20
9
(0)
(1)
(2)
Substitution of Wn , Wn , and Wn into the Van Vleck eq. (60) yields the magnetic
susceptibility of the 4f 1 system in cubic ligand fields:
NA 2B 2
where
3kB T eff
40
130
40
2 25
gJ
+
kB T +
kB T exp
12 3
12
3
kB T
=
.
1 + 2 exp
kB T
m = 0
2eff
(84)
1
Fig. 8 shows the eff T and the 1
(LF ground state
m T behaviour for = +605 cm
4
1
1
4
E (7 ), B0 = 2 119 cm , oct.) and = 605 cm (G (8 ), B0 = 2 119 cm1 , tetrah.,
cube).
a
c
2.0
500
-3
5
c-1
mol /10 mol m
2.4
meff
1.6
1.2
0.8
0.4
0.0
400
f
300
200
b
100
0
1
Fig. 8: Ce3+ in cubic LF; eff T (a,c,e) and 1
oct.,
m T diagrams (b,d,f); = 605 cm
1
(e,f), = 605 cm tetrah., cube, (c,d); straight lines (a,b) correspond to the free ion.
Discussion
1. The ligand field effect produces a distinct deviation from free ion magnetic behaviour, except Ln ions with 4f 7 configuration (Eu2+ , Gd3+ ).
2. At very low temperature the 1
m vs. T curves for systems with an odd number of
4f electrons (half-integral J) become nearly straight lines running into the origin
(Curie-like behaviour, explicable with eq. (84)), provided that cooperative magnetic
effects are absent.
3. In the case of integral J the low-temperatur magnetic behaviour is more complicated
(see Section 3.3.4).
4. Warning: The upper region of the 1
m versus T curves does not follow the Curie
7
or Curie-Weiss law (except 4f ions). To compare the magnetic behaviour of Ln
ions in a crystal with that of free ions, the eff versus T plot is suited.
34
(85)
6. Susceptibility equations of the type (84) are not only suited for molecular and ionic
systems, but also for intermetallics with stable Ln valence. The 6s and 5d valence
electrons are delocalised to a high degree and yield normally only a small TIP.
7. Ce, Yb, Eu are candidates for mixed-valence behaviour (Ce3+/4+ , Yb2+/3+ , Eu2+/3+ ).
Paramagnetism of Ce3+ , Yb3+ , U5+ ions in an octahedral LF
Figs. 9 and 10 exhibit the typical paramagnetic behaviour of octahedrally coordinated
Ce(III), Yb(III) and U(V) compounds as 1
m T and eff T plots calculated with program
CONDON [12], using the parameter values (in cm1 ):
Ce(III): B04 = 2 119, B06 = 261, = 623, see Tab. 16 [13];
Yb(III): B04 = 1 471, B06 = 0, = 2 903 [13];
U(V): B04 = 23 100, B06 = 3 750, = 2 200 [14].
The 4f 13 ion with ground multiplet 2 F7/2 is the strongest paramagnet, the 5f 1 ion (2 F5/2 )
the weakest one. The lower eff value of the latter compared to eff of Ce(III) is the
consequence of the much stronger LF effect.
160
5+
3+
Yb
120
meff
-3
6
c-1
mol / ( 10 mol m )
200
80
Ce3+
40
Ce 3+
3+
Yb
0
100
200
T/K
300
0
0
400
Fig. 9: Typical 1
m T diagrams of compounds with Ce(III), Yb(III) and U(V) in an
octahedral ligand field.
3.3.3
100
200
T/K
300
5+
400
Fig. 10: Typical eff T diagrams of compounds with Ce(III), Yb(III) and U(V) in an
octahedral ligand field.
4f N systems
cub reads
For many-electron lanthanide systems H
LF
cub
LF
H
(4f N ) = B04
N n
X
i=1
C04 (i)
5
14
[C44 (i)
4
C4
(i)]
35
B06
N n
q
o
X
6
C06 (i) 72 [C46 (i) + C4
(i)] (86)
i=1
Coupling schemes: The 4f states are controlled by Hee , HSO (see Fig. 11) and the
ligand field 6) (HLF ), where the order of energetic effects is Hee > HSO > HLF (except for
the 4f 4 , 4f 5 , 4f 6 systems where HSO HLF )7) . To describe the electronic situation, three
coupling schemes are distinguished, depending on the relative strength of HLF and on the
desired accuracy in LF parameter determination [15].
0
1
2
15/2
5
13/2
11/2
E / 10 cm
-1
3
2
7/2
0
J=4
3+
Pr
J= 5/2
9/2
3+
Sm
J=0
Eu
3+
J=6
Tb 3+
Fig. 11: Energies of ground state multiplets of selected Ln3+ ions (kB T =
208.6 cm1 at 300 K)
Russell-Saunders coupling scheme (LS coupling scheme)
Including spin-orbit coupling in 1st order, S, L, and J are good quantum numbers
in this coupling scheme, i. e., this model requires multiplets 2S+1LJ that are (i)
well separated in energy and (ii) only slightly influenced by HLF . To estimate LF
parameters it is sufficient to study the action of HLF solely on the ground multiplet
2S+1
LJ [16] (except the 4f 4 , 4f 5 , 4f 6 systems).
Intermediate coupling scheme
The intermediate coupling scheme is applicable, if S and L are no longer good
quantum numbers on account of the competing perturbations by Hee and HSO while
J remains a good quantum number. This model is adequate, if the ligand field
splitting is not too strong, i. e., in the case of the weak-field lanthanide system 8) .
Examples: lanthanide chlorides and bromides (Cl , Br ).
J-mixing scheme
The J-mixing coupling scheme is favourable for strong-field lanthanide systems. If
J-mixing, caused by HLF , is considered, a 2S+1LJ multiplet may have contributions
from terms with other J values, that is, J is no longer a good quantum number.
Examples: lanthanide oxides and fluorides (O2 , F ).
6)
For lanthanides the term crystal field is often used instead of ligand field.
For actinides in high oxidation states there is no energetic order, that is, Hee HSO HLF .
8)
The terms weak field and strong field here have another meaning than for transition metal
compounds. For a strong-field transition metal system, HLF is a stronger interaction than Hee whereas
for lanthanides HLF is always weaker than Hee .
7)
36
3.3.4
On the basis of spectroscopic data of the series Cs2 NaLnCl6 9) [13] (see Tab. 16) the magnetic behaviour is calculated with program CONDON 10) , developed by Schilder and
Lueken [12]. The eff T data for the representative metal ions Pr3+ [4f 2 ], Nd3+ [4f 3 ],
Sm3+ [4f 5 ] and Eu3+ [4f 6 ] are compared with those of the free ions.
Cl
Cs
Ln
Na
Fig. 12: Cs2 NaLnCl6 : unit cell of the elpasolite type
Table 16: LF parameters B04 and B06 of octahedrally coordinated Ln3+ ions in Cs2 NaLnCl6
Ln3+
B04 a)
B06 a)
Ce3+
2119
261
Pr3+
1938
290
3+
Nd
1966
258
3+
b)
Sm
(1671)
(228) b)
Eu3+
2055
308
3+
Gd
1776
136
a)
b)
Ln3+
Tb3+
Dy3+
Ho3+
Er3+
Tm3+
Yb3+
B04
1624
1614
1593
1492
1498
1471
B06
150
148
171
163
159
[0]
Values in cm1 .
Values for Cs2 NaYCl6 : Sm3+ .
The chemical environment of the Ln ions, compared to the one of d ions, has only a
minor effect on the 4f electrons. The LF effect produces splittings of HLF 102 cm1
9)
37
4.0
240
200
3.0
2.0
120
80
150
c-1mol /105mol m-3
160
meff
5
-3
c -1
mol /10 mol m
200
1.0
40
0
0.0
100 200 300 400
T/K
100
50
0
0
3.6
3.2
2.8
2.4
2.0
1.6
1.2
0.8
0.4
0.0
100 200 300 400
meff
Eu3+ : In the complete temperature range the magnetic properties of Eu3+ compounds
are determined by the TIP on account of the non-magnetic ground multiplet with J = 0
(see Fig. 14, right). The effect of a cubic ligand field is negligible.
T/K
3+
Fig. 13: 1
(left) and Nd3+ (right), calculated with the
m T and eff T diagrams for Pr
spectroscopically determined data of Cs2 NaPrCl6 and Cs2 NaNdCl6 , respectively (solid
lines: full basis; dottet lines: ground multiplet only; dashed lines: free ions)
38
4.0
8
d
meff
2.0
b
a
1.0
5
4
300
200
T/K
100
3
2
c
d
0.0
0
6
-7
-3
c-1
mol /10 mol m
3.0
400
200
T/K
100
300
400
3+
Fig. 14: eff T (left) and 1
and Eu3+ , calculated with
m T diagrams (right) for Sm
3+
the spectroscopically determined data of Cs2 NaYCl6 :Sm (a) and Cs2 NaEuCl6 (d), respectively; free Sm3+ ion (b), free Eu3+ ion (c).
cub
4f 2 system (Pr3+, 3H4 : HLF
)
3.3.5
cub reads:
For a 4f 2 system H
LF
cub (4f 2 ) = B 4
H
LF
0
2 n
X
C04 (i) +
i=1
(88)
2 n
q
o
o
X
4
4
6
6
6
6
5
7
[C
(i)
+
C
(i)]
+
B
C
(i)
[C
(i)
+
C
(i)]
4
4
0
0
4
4
14
2
i=1
Compared to the two multiplets of the 4f 1 (Ce3+ ) free ion, the 4f 2 (Pr3+ ) free ion has
13 multiplets and, in consequence, exhibits a more complicated LF splitting pattern (see
Table 17 and ref. [13]).
Table 17: Spectroscopically determined levels of the Pr3+ ion (Oh , octahedron)
in Cs2 NaPrCl6 . The energy increases from top to bottom and from left to right.
3
H5
A1 [1 ](0 a) ) T1
(1) b)
T1 [4 ](236)
T2
T2
E[3 ](422)
A1
T2 [5 ](701)
T1
H4
(2)
F2
E c)
T1
(1)
T2
H6
G4
D2
P0
P1
P2
A1
T2
A1
T1
A1
T2
A1
T2
T1
T1
A2
A1
T1
T2
T2
T2
A2
F3
A2 [2 ]
F4
T2
(2)
T2
(1)
(2)
T1
a)
I6
E
b)
2300 cm1;
39
c)
S0
According to the procedure outlined in 3.3.2, only the ground multiplet 3H4 will be
considered for the 4f 2 ion, that is, one starts with the |JMJ i basis consisting of the
2J + 1 = 9 states |4 4i, |4 3i, . . . , |4 4i. The 4f 2 situation is much more complicated
than the 4f 1 one for the following reasons: In order to apply the entries of Table 13,
the |JMJ i functions have to be decomposed into the microstates of the two electrons
(cf. ref. [7], page 120). Subsequently, the operator eq. (88) must be applied. Then the
respective LF matrix elements have to be summed up and collected in a 9 9 matrix.
After diagonalisation of this matrix, the symmetry adapted functions and their energies
are obtained as a function of the LF parameters B04 and B06 .
These calculations would be time-consuming and must be worked out by a computer,
using, for example, the program CONDON [12] which is at our disposal. In general, this
program uses the complete set of basis functions, but can be restricted to the ground
multiplet11) . The results, obtained by CONDON with the restricted basis, are presented
in Table 18. A splitting in a singlet (A1 ), a doublet (E) and two triplets (T1 , T2 ) is
obtained and the entries of Table 17 are confirmed.
Tab. 18: Functions and energies after perturbation of the 3H4 (4f 2 )
ground multiplet by a cubic ligand field
functions |MJ i (J = 4)
q
1
14
|
0
i
+
5
|
4
i
+
5
|
4
i
24
q
1
10
|
0
i
+
7
|
4
i
+
7
|
4
i
24
q
1
(| 2 i + | 2 i)
2
q
q
18 | 3 i 78 | 1 i
q
1
(| 4 i | 4 i)
2
q
q
1
7
|
3
i
+
| 1 i
8
8
q
q
7
| 3 i 18 | 1 i
8
q
1
(| 2 i | 2 i)
2
q
q
78 | 3 i + 18 | 1 i
a)
b)
A1 1 , E 3 , T1 4 , T2 5 [6].
b4 = (15/2)J B04 , b6 = (315/4)J B06 ;
11)
ia) )
|M
ELF
b)
|A1 a1 i
28 b4 80 b6
|Ei
4 b4 + 64 b6
|Ei
|T1 1i
|T1 0i
14 b4 + 4 b6
|T1 1i
|T2 1i
|T2 0i
26 b4 20 b6
|T2 1i
4
,
J = 5 445
J =
272
4 459 455 .
As an alternative, a rather simple method is available to determine LF effects for cubic lanthanide
systems, if the description of the LF effect within the ground multiplet is sufficient. In this method an
operator is applied that consists of so-called operator-equivalents instead of the Racah tensors. We will
come back to its practice in 3.3.7 and in Appendix 2.
40
3.3.6
cub
4f 2 system (Pr3+, 3H4 : HLF
+ Hmag )
Applying the Zeeman operator eq. (66) on the eigenfunctions, the matrix elements (89)
are obtained (in units of gJ B Bz ; gJ = 4/5).
(89)
A1
A1
a1
a1
1
T1
0
1
T1
20
3
28
3
1
2
28
3
7
2
0
21
20
T2
q3
0
q
0
q
T2
7
2
7
2
5
2
7
2
(0)
52
(1)
(2)
(1)
(2)
(0)
Wn a)
|A1 a1 i
28 b4 80 b6
|Ei
4 b4 + 64 b6
|Ei
|T1 1i
14 b4 + 4 b6
|T1 0i
|T2 1i
|T2 0i
a)
26 b4 20 b6
(2)
Wn b)
gJ2 2B
(20/3)
0
(T1 ; A1 )
(28/3)
0
(T1 ; E)
4
0
(T2 ; E)
(7/4)
1/2
(T2 ; T1 )
(20/3)
(28/3)
0
+
+
(T1 ; A1 ) (T1 ; E)
(7/4)
5/2
+
(T2 ; T1 )
4
0
+
(T2 ; E)
Wn
gJ B
(0)
41
(0)
(1)
(2)
Setting Wn , Wn and Wn into the Van Vleck eq. (60), the susceptibility equation
(0)
for the Pr3+ ion in cubic ligand fields is developed (where X Wn ()/kB T ):
NA 2B 2
m (Pr3+ ) = 0
2
(T2 ; T1 ) (T2 ; E)
1
exp(XA1 ) + 2 exp(XE ) + 3 exp(XT1 ) + 3 exp(XT2 )
(90)
3+
Fig. 13 (left) exhibits calculated 1
ion in an
m T and eff T diagrams for the Pr
4
1
6
1
octahedral ligand field where B0 = 1938 cm and B0 = 290 cm . Dotted lines refer to
calculations with the reduced basis 3H4 (eq. (90)), whereas the solid lines correspond to
calculations with the complete basis (91 microstates). Differences in m values between
the two models amount to 3 %. In addition, the expected magnetic behaviour of the free
ion is shown for comparison (dashed lines).
3.3.7
The magnetic behaviour of Ln ions depends on position and charge of the ligands, both
controlling sign and amount of the LF parameters Bqk . If one accepts the qualitative
validity of the point charge electrostatic model (PCEM) for the geometrical coordination
factor Aqk (defined in eq. (72)), the sign of the corresponding Bqk = Aqk < r k >, that is,
the sign of B04 and B06 in the case of cubic coordination polyhedra (octahedron, cube,
tetrahedron), can be predicted reliably12) . We consider the point charges Qe 13) at the
vertices of regular polyhedra (octahedron, tetrahedron, cube) with the distance R to the
centre of the polyhedron, and calculate A04 and A06 .
Geometrical coordination factors
A04 For the octahedron the integration in eq. (72) reduces to a summation of the six
ligands with (j , j ) coordinates (0, 0), (, 0), (/2, 0), (/2, /2), (/2, ), (/2, 3/2):
A04 (oct.)
6
6
Qe2 X 4
Qe2 X 1
7 Qe2
4
2
= 5
C0 (j , j ) = 5
,
(35
cos
30
cos
+
3)
=
j
j
R j=1
R j=1 8
2 R5
(91)
For negatively charged ligands (Q > 0) which coordinate the Ln ion octahedrally, A04 (oct.)
is positive and consequently also A04 < r 4 > = B04 , whereas the A04 s for tetrahedron and
cube are negative (angle coordinates for tetrahedron: (, /4), (, 5/4), ( , 3/4),
12)
42
( , 7/4), where cos = 1/ 3; additionally for the cube (, 3/4), (, 7/4), (
, /4), ( , 5/4)):
A04 (tetr.) =
14 Qe2
;
9 R5
A04 (cube) =
28 Qe2
9 R5
(92)
A06
A06 (oct.)
6
Qe2 X 6
C (j , j ),
=
R7 j=1 0
where
C06 (j , j )
4
13
1/2
Y06 (j , j )
6
Qe2 X 1
3 Qe2
6
4
2
=
(231
cos
315
cos
+
105
cos
5)
=
j
j
j
R7 j=1 16
4 R7
A06 (tetr.) =
8 Qe2
;
9 R7
A06 (cube) =
16 Qe2
9 R7
(93)
(94)
For negatively charged ligands (Q > 0), A06 is positive, regardless the polyhedron.
Having fixed the sign of the LF parameters B04 and B06 , the possible LF ground states
of the Ln ion under investigation can be determined. If the program CONDON is at
our disposal, we need only the values of B04 and B06 as an input to obtain energy and
composition of the LF states. Otherwise, we can benifit from tables of eigenvectors and
diagrams of energy eigenvalues, expressed as a function of the ratio between the fourth
cub , that are published for all J-manifolds of the lanthanides
and sixth degree terms of H
LF
by Lea, Leask, Wolf [16]. The immediate application of these diagrams and tables is
the prediction of spectroscopic energy level schemes and possible g-factors for lanthanide
ions in cubic coordination. We start with introductory remarks concerning the operator
equivalent technique.
Operator equivalents: To describe the fundamental magnetic behaviour of Ln ions
in a solid, a model is sufficient that, according to the Russell-Saunders coupling scheme,
LF into consideratakes solely the ground multiplet 2S+1LJ and its perturbation by H
tion (exceptions: Sm3+ [4f 5 ], Eu3+ [4f 6 ]). The restriction of the basis functions allows
|J M i by application of the soa simple calculation of the matrix elements hJ MJ |H
LF
J
called operator-equivalent technique. Using this technique, the calculation of the integrals
hli ml1 |Cqk (i)|li ml1 i entailing a troublesome decoding of the |J MJ i basis with regard to
the microstates is unnecessary.
The operator equivalent technique is founded on the fact that the operators Cqk =
p
4/(2k + 1) Yqk , written in cartesian coordinates, can be replaced by operators Jx , Jy ,
qk are
Jz of the total angular momentum [17] 14) . By this means operator equivalents O
designed whose matrix elements are proportional to the corresponding matrix elements of
LF . The restriction to the ground multiplet has the advantage that the matrix elements
H
cub |JM i are easily evaluated: Given the wave functions in the |JM i basis, the
hJMJ |H
LF
J
J
cub on the orbital part of the functions is taken into consideration by a simple
action of H
LF
cub (eq. (72)) are
factor (Stevens factor [18]). If x, y, z in the Racah tensors Cqk (x, y, z) of H
LF
Originally, the operator equivalents are related to the spherical harmonics Yqk [18, 21]. Conversion
factors are published in ref. [22]. Details are given in the Final remarks, p. 49.
43
k are obtained, whose matrix elements are proportional to the matrix elements of H
cub .
O
q
LF
Simple examples of operator equivalents are:
x2 y 2 (Jx2 Jy2 ),
xy ( 21 )(Jx Jy + Jy Jx ).
J (k = 4)
J (k = 6)
For cubic symmetry only terms with k = 4 and 6 are relevant, so that the multiplet under
consideration is generally characterised by the two figures J and J 15) . The two following
equations show the operator and its operator equivalent:
cub (4f N ) = B 4
H
LF
0
N n
N n
q
q
o
o
X
X
4
4
4
6
6
6
6
5
7
C0 (i) + 14 [C4 (i) + C4 (i)] + B0
C0 (i) 2 [C4 (i) + C4 (i)]
i=1
i=1
q
q
i
h
i
h
6
5
4 + O
4
6 7 O
6 + O
6
cub = B 4 O
4 +
O
+
B
O
H
4
4
0 J
0
4
4
LF
0 J
0
14
2
(95)
C04 (i)
4 = J
= J O
0
1
8
i=1
N
X
4
4 = J
C4
(i) = J O
4
i=1
N
X
6 =
C06 (i) = J O
0
J
i=1
o
n
4
2
2
2
1
16
n
70
1
2
J4
(96)
525 J(J + 1) + 294]J2z 5 J 3 (J + 1)3 +40 J 2 (J + 1)2 60 J(J + 1)
N
q
o
X
n
6
6
63
1
2z J(J + 1) 38]J4 + J4 [. . .]
C4 (i) = J O4 = J 128
[11
J
(97)
4
i=1
k |JM i that are necessary to set up HLF matrices are availThe matrix elements hJMJ |O
q
J
able and collected in the Tables 35 38 on pages 89 90. In the following three examples
the handling of operator equivalents is shown.
Example 3.7 Application of the operator-equivalent method to 2F5/2 of Ce3+ [4f 1 ]
The HLF operator equivalent for the cubic 4f 1 system reads
q
i
h
5
4 + O
4
cub = B 4 O
4 +
O
H
4
4
LF
0 J
0
14
(98)
Inspecting the Tables 35 and 36 the relevant data for the basis | 52 MJ i are:
4| 1 i =
h 12 |O
15
0
2
3
3 4
h 2 |O0 | 2 i = 45/2
4| 5 i =
h 52 |O
15/2
0
2
3
5 4
h 2 |O4| 2 i = 15 14/2.
15)
(99)
Exceptions are the multiplets 2F5/2 (Ce3+ [4f 1 ]) and 6H5/2 (Sm3+ [4f 5 ]) with J = 5/2, where only
is relevant.
the fourth degree term (k = 4) in H
LF
44
5
2
7
2
o
2
315
15/2
2
On the basis of the operator-equivalent matrix elements (99) and J = 315
(Example 3.8)
cub
2
the HLF matrix within the basis F5/2 can be set up. It is identical to matrix (83).
14 b4 + 4 b6
0
70 (b4 6 b6 )
where b4 =
0
15
J B04 ,
3
2
70 (b4 6 b6 )
18 b4 20 b6
4
0
70 (b4 6 b6 )
and b6 =
1
70 (b4 6 b6 )
14 b4 + 4 b6
315
4
J B06
2
(100)
2
3 21 b4 17 b6
7 (5 b4 + 3 b6 )
1
7 (5 b4 + 3 b6 )
9 b4 + b6
2
11 b4 + 22 b6 15 b4 + 42 b6
2
15 b4 + 42 b6 11 b4 + 22 b6
cub matrix yields the energies and eigenfunctions given in Table 21.
Evaluation of the H
LF
The splitting in a singlet (A1 ), a doublet (E) and two triplets (T1 , T2 ) is confirmed.
45
qk in the basis
Tab. 20: Matrix elements of the operator equivalents O
2
|JMJ i (J = 4) for 4f and cubic symmetry (see Appendix 2)
qk |M i
hMJ |O
J
0k | 0 i
h 0 |O
k=6
qk |M i
hMJ |O
J
135 1575
0k | 4 i
h4 |O
k=4
k| 1 i
h1 |O
0
135
2
315
4
k| 2 i
h2 |O
0
165
2
3465
2
0k | 3 i
h3 |O
315
2
5355
4
k=4
k=6
105
315
k | 2 i 45 70 945 14
h2 |O
4
2
2
k | 1 i 105 10 945 2
h3 |O
4
2
4
k
4
h4 |O
|0i
105
945 5
Tab. 21: Functions and energies after perturbation of the 3H4 (4f 2 )
multiplet by a cubic ligand field
functions |MJ i (J = 4)
q
1
14
|
0
i
+
5
|
4
i
+
5
|
4
i
q 24
1
10 | 0 i + 7 | 4 i + 7 | 4 i
24
q
1
(| 2 i + | 2 i)
2
q
q
1
8 | 3 i 78 | 1 i
q
1
(| 4 i | 4 i)
q2
q
1
7
|
3
i
+
| 1 i
q8
q8
7
| 3 i 18 | 1 i
q8
1
(| 2 i | 2 i)
2
q
q
7
8 | 3 i + 18 | 1 i
a)
b)
ia) )
|M
ELF
b)
|A1 a1 i
28 b4 80 b6
|Ei
4 b4 + 64 b6
|Ei
|T1 1i
|T1 0i
14 b4 + 4 b6
|T1 1i
|T2 1i
|T2 0i
26 b4 20 b6
|T2 1i
Level tuning [16]: The ground state of an Ln ion in a cubic LF depends on J B04 and
J B06 . We want to obtain information about the succession of ELF of the LF states as a
function of the ratio J B04 /J B06 . Using the operator equivalent matrix elements in Tables
35 38, the (2J + 1) (2J + 1) matrix can in principle be drawn up. The tables contain
factors common to all the matrix elements, F (4) and F (6). These factors are separated
out in order to keep the energy values in the same numerical range for all ratios of the
fourth and sixth degree terms. We start out from eq. (95)
q
q
i
i
h
h
6 6
cub
7
4
6
6
44 + O
4
04 + 5 O
LF
O
+
O
O
+
B
H
= J B04 O
4
4
0
J
0
14
2
|
|
{z
}
{z
}
4
6
b
b
O
O
46
b 4 + B6 O
b6
= J B04 O
J 0
b6
b4
O
O
+ J B06 F (6)
.
= J B04 F (4)
F (4)
F (6)
(101)
In order to cover all possible values of the ratio between the fourth and sixth degree terms
a mixing parameter x has been introduced which is governed by the relative magnitude
of B04 and B06 . Furthermore, a factor W has been defined that scales ELF :
J B04 F (4) = W x,
| {z }
b4
where
1 < x < +1
(102)
b6
Solving both eqs. (102) for W , an equation is obtained that covers all possible values of
the ratio between B04 and B06 . It follows that
J B04
x
F (6)
,
=
6
J B0
1 |x| F (4)
(103)
so that (J B04 /J B06 ) = 0 for x = 0, while (J B04 /J B06 ) = for x = 1. In consequence, eq. (101) can be rewritten as
"
!
!#
b4
b6
O
O
cub = W x
H
+ (1 |x|)
(104)
LF
F (4)
F (6)
For each J-manifold x is allowed to take the values in the range 0 1. The energy
expressions for the LF states as a function of B04 F (4) = b4 and B06 F (4) = b6 are
replaced by W x and W (1 |x|), respecticely. Table 22 exhibits the results for the 4f 2
system and Fig. 15 displays the corresponding diagram ELF /W versus x.
Tab. 22: Energies ELF of the 4f 2 system in cubic ligand fields as
a function of W and x as well as b4 and b6 , see eqs. (102)
ELF /W
ELF
x = 0 x = 1
A1 28 b4 80 b6
80
28
4 b4 + 64 b6
64
T1 14 b4 + 4 b6
14
20
26
Diagrams like Figs. 15 and 16, displaying LF energy eigenvalues ELF /W as a function of
the LF parameters, are available for all J-manifolds of the lanthanides ground multiplets
[16]. In examples 3.10 and 3.11 the diagrams are used to determine the LF ground state for
Pr3+ [4f 2 ] and Nd3+ [4f 3 ] in cubic LF. In example 3.12 the construction of the x ELF /W
diagram for Pr3+ is demonstrated.
47
60
E (G3)
40
40
20
T1 (G4)
20
(2)
(2)
( G8 )
E (G6)
x
T2(G5)
-40
A1(G1)
-1
X
-20
-40
-60
-1
Pr3+ [4f 2 ]
(1)
(1)
( G8 )
1
(J =
B04 =
17)
The result of the PCEM calculation is supported by spectroscopic investigations, showing that
1 938 cm1 and B06 = 290 cm1 (see Table 16).
B04 = 1 966 cm1, B06 = 258 cm1 (see Table 16).
48
The x(ELF /W ) diagram in Fig. 16 shows that for the situation x > 0 and W < 018)
either the quartet G or the doublet E is the ground state. Thus, ignoring interionic spinspin couplings, an octahedrally coordinated Nd3+ ion should exhibit Curie paramagnetism
regardless T . Fig. 13 (right) exhibits the magnetic behaviour.
Note, that the diagrams in Figs. 15 and 16 differ distinctly: While for the Pr3+ ion
exclusively straight lines result, one straight line and two curves are observed for the
Nd3+ ion (see Remarks on page 49 for further details).
Nd3+ : The signs of W and x for a tetrahedral/cubical LF:
Since both W and x are negative (according to J , J , B04 < 0 and B06 > 0, eq. (103)) the
ground state is the quartet G .
Final remarks
According to todays standard practice, we use the operator equivalents that are
related to the Racah tensors Cqk [17] while Stevens definition relates to the spherical harmonics Yqk [18, 19]. Since the article of Lea, Leask, Wolf [16] is based on
Stevens definition Yqk , some hints to the conversion factors between both derivations
is informative with respect to the application of the LLW diagrams:
(i) The LLW diagrams can be applied directly, since the conversion factor for both
W and x is 1. In other words, the values O4 /F (4) (based on Yqk ) in equation
b4 /F (4) (based on C k ) in eq. (104) are identical and the
(5) of ref. [16] and O
q
same is true for the sixth degree terms (k = 6).
(ii) Conversion
(see eq. (96) and (97)):
p
(iii) The LF parameters, used by LLW, B04 (LLW) and B06 (LLW) [16], include the
Stevens factors J and J , while in todays standard practice, used by us,
both are separated:
B04 (LLW) = J B04 ,
B06 (LLW) = J B06
Within the x(ELF /W ) diagrams there are two types of i curves to be considered:
(i) curves that are straight lines and (ii) curves that deviate more or less from
straight lines.
(i) A state i whose energy eigenvalue is a linear function of x, occurs once only in
the decomposition of the J-manifold. For example, in the cubic Pr3+ systems
with ground multiplet J = 4 each split term A1 (1 ), E(3 ), T1 (4 ), T2 (5 )
occurs once only. So, the diagram exhibits only straight lines. The state
functions have fixed compositions (see Table 21) independent of B04 and B06 .
Other examples are ions in cubic LFs with J = 52 (E (7 ), G (8 )) and 27 (E
(6 ), E (7 ), G (8 )).
18)
According to the negative sign of W , one has to mirror the x(ELF /W ) diagram at the x axis with
the consequence that the ELF /W axis shows downwards for W < 0.
49
Tab. 23: Ground states of Ln3+ ions in cubic ligand fields [16]
Ln3+ [4f N ]
2S+1
LJ
Ce3+ [4f 1 ]
2
F5/2
Pr3+ [4f 2 ]
Nd3+ [4f 3 ]
H4
I9/2
Sm3+ [4f 5 ]
6
H5/2
Tb3+ [4f 8 ]
7
F6
2
35
52
2 475
7
1 089
13
315
1
99
2
315
4
5 445
136
467 181
26
10 395
2
16 335
272
4 459 455
1 615
42 513 471
1
891 891
F (4)a)
15
2
15
2
21
2
15
2
15
2
F (6)
315 b)
4
315
945
2
[O]c)
+ d)
E [ +1
]
A1 [ +
]
G , E [
]
+
+
E [ +1
]
A2 , A1 [
]
[T ]
G [ +1
]
+
A1 , T2 [ +
]
G [
]
G [ +1
]
A2 , E [
]
+
Ln3+ [4f N ]
2S+1
LJ
Dy3+ [4f 9 ] Ho3+ [4f 10 ] Er3+ [4f 11 ] Tm3+ [4f 12 ] Yb3+ [4f 13 ]
6
H15/2
I8
I15/2
H6
F7/2
2
315
1
450
4
1 575
1
99
2
63
8
135 135
1
30 030
2
45 045
8
49 005
2
1 155
4
3 864 861
5
3 864 861
8
3 864 861
5
891 891
4
27 027
F (4)
15
2
105
2
15
2
15
2
15
2
F (6)
3 465
4
3 465
4
3 465
4
945
2
315 b)
4
[O]
]
E , E [ +
]
E, A1 [ +
]
E , G [ +
+
]
A2 , A1 [
+
]
E [
[T ]
E , G [ +
]
+
E, T2 [
]
E , E [ +
]
A2 , E [
]
+
E , E [ +
]
+
F (4) and F (6) are equivalent to F (4) and F (6) [16] and consider the different
qk and Oqk .
definitions of the operator equivalents O
b)
Note that there is a misprint in the corresponding Table 4.39 of ref. [7].
c)
Indices O and T identify the expected LF ground state for octahedral and
tetrahedral (cubal) symmetry and negatively charged ligands (A1 1 ; A2 2 ; E
3 ; T1 4 ; T2 5 ; E 6 ; E 7 ; G 8 ).
d)
Combination of signs for the W and x parameter W
x .
a)
50
(ii) Systems with J > 4 decompose as well into states which occur more than once
and deviate from straight lines. Taking J = 29 (Nd3+ ) as an example, decom(1)
(2)
cub
position by HLF
takes place into the states E (6 ), G (8 ), G (8 ). Here
(1)
(2)
the energy of the states G (8 ) and G (8 ) depends in a more complicated
way on the ratio B04 /B06 and consequently on x (see Fig. 16)19) . This is the
consequence of the fact that the functions of both states transform according
to the same irreducible representation of the cubic group. The state E (6 ),
however, occurs once only. So, its energy is a linear function of x.
The multiplets of Ln3+ ions with the same J, e. g., the two pairs Tb3+ , Tm3+ (J = 6)
and Dy3+ , Er3+ (J = 15/2), respectively, are each reduced in cubic ligand fields into
the same irreducible representations.
Exact values for LF parameters cannot be calculated reliably. However, the sign
of Bqk values can often be predicted. Consequently, the signs of x and W can be
deduced, leading to the determination of the ground state.
Systems with non-integer J have exclusively magnetic states (showing Curie paramagnetism 20) ), while in the systems with integer J ligand field ground states may be
produced which are nonmagnetic, for example A1 (1 ), A2 (2 ) and E(3 ), yielding
temperature independent paramagnetism (TIP).
The x(ELF /W ) diagrams are helpful not only for cubic systems but also for coordination polyhedra that dont deviate greatly from cubic symmetry, that is, have
cubic pseudo symmetry.
Example 3.12 : Energy of the LF states of Pr3+ in Cs2 NaPrCl6
To verify Fig. 15 we use Pr3+ in Cs2 NaPrCl6 as an example with B04 = 1 938 cm1 and
B06 = 290 cm1 (Table 16). With the help of the entries in Table 23 as well as eqs. (102)
and (103), the following data result:
4
15
4
b4 = J F (4)B0 = W x =
1938 cm1 = 10.68 cm1
5 445
2
315
272
6
51
Tab. 24: Calculated energies ELF of the LF states of Pr3+ in Cs2 NaPrCl6 on
the basis of spectroscopically determined values B04 = 1938 and B06 = 290 cm1
[13].
ELF,calc. /cm1
a)
ELF,obs. b)
A1 (1 )
410
T1 (4 )
144
266
236
E(3 )
46
456
422
T2 (5 )
250 660c)
701
a)
b)
c)
Problems
7. Determine the crystal field overall splitting (CFOS) for the Dy3+ ion in Cs2 NaDyCl6
on the basis of Fig. 35. Apply eqs. (102) and (103) as well as the entries in Table 23
and take the B04 and B06 values for Dy from Table 16. (Solutions are specified in
section 5.)
52
3.4
3.4.1
Non-cubic LF
Ligand field operators (single electron systems)
(105)
tet = B 2 C 2 + B 4 C 4 + B 4 C 4 + C 4 + B 6 C 6 + B 6 C 6 + C 6
H
LF
0 0
0 0
4
4
4
0 0
4
4
4
(106)
(107)
tri
4
6
6
LF
H
= B02 C02 +B04 C04 +B34 C3
C34 +B06 C06 +B36 C3
C36 +B66 C6
+ C66 (108)
For d electron systems only terms with k 4 are relevant. The number of terms for f(d)
systems with cylindrical, hexagonal, tetragonal, trigonal symmetry is 3(2), 4(2), 5(3),
operators (including the symmetry Ih ) is given
6(3), respectively. A complete list of H
LF
in ref. [15].
3.4.2
cyl
4f 1 system (Ce3+ , 2F5/2 : HLF
)
In the following calculations the basis functions are restricted to the multiplet 2F5/2 23) .
cyl (2F5/2 ) = B 2 C 2 + B 4 C 4
Ligand field operator: H
0 0
0 0
LF
(109)
Operator (109), acting on 2F5/2 , produces solely diagonal elements (q = 0). The matrix
elements hMJ |C0k |MJ i for k = 4 are known from matrix (83). The data for k = 2 are
obtained with the help of Table 8 and Table 13:
h21 |C02 | 12 i =
8
;
35
h32 |C02 | 32 i =
2
;
35
h52 |C02 | 52 i = 10
.
35
4
= 41 35 sin3 cos ei3 ;
C02 = 12 3 cos2 1 ; C3
1
1
6
6
105 sin3 cos 11 cos2 3 ei3 ; C6
= 32
231 sin6 ei6 .
C3
= 16
cyl
22)
HLF applies also for D4d .
23)
cyl and H
hex are identical.
Note that under this restriction the effects of H
LF
LF
21)
53
|MJ i
cyl
Table 25: Splitting of the 2 F5/2 [4f 1 ] multiplet by H
LF
(0)
En,LF Wn
| 21 ia)
8
B2
35 0
2
B4
21 0
| 23 i
2
B2
35 0
3
B4
21 0
| 25 i
B2 +
10
35 0
1
B4
21 0
a)
Wavefunctions adapted to
the magnetic field in x-direction:
| i = 12 | 21 i + | 12 i .
cyl
4f 1 system (Ce3+ , 2F5/2 : HLF
+ Hmag )
where
3kB T effk
1
9 25
1
2
3
+ +
exp
exp
4
kB T
4
4
kB T
2
= gJ
,
1
2
exp
+ 1 + exp
kB T
kB T
(110)
mk = 0
2effk
and
(0)
(0)
(0)
(0)
1 = W1/2 W3/2 =
6
B2
35 0
gJ =
6
7
5
B4
21 0
B2 +
2 = W5/2 W3/2 = 12
35 0
4
B4.
21 0
cyl , H
magk (z) and H
mag (x);
Table 26: Splitting of the 2 F5/2 [4f 1 ] multiplet by H
LF
(0)
(1)
(1)
(2)
energies ELF Wn , Zeeman coefficients Wn,z and Wn,x , Wn,x
|MJ i
(0)
En,LF Wn
(1)
(1)
(2)
Wn,z /gJ B
Wn,x /gJ B
Wn,x /gJ2 2B a)
| 12 ib)
8
B2
35 0
2
B4
21 0
21
32
2/1
| 32 i
2
B2
35 0
3
B4
21 0
23
5/(42 ) 2/1
| 52 i
10
B2 +
35 0
1
B4
21 0
25
5/(42 )
a)
b)
(0)
(0)
(0)
(0)
5
6
2
B02 + 21
B04 ;
2 = W5/2 W3/2 = 12
1 = W1/2 W3/2 = 35
35 B0 +
mag : | i = 1 | 1 i + | 1 i .
Wave functions adapted to H
2
2
2
54
4
4
21 B0
;
|
i
=
i
+
|
i
.
=
0;
W
=
n(1,2)
2 J B
2
2
2
2 gJ B Wn(1)
The H matrix reads (in units of gJ B Bx ):
5
2
5
2
3
2
3
2
5 5 3 3
+
2
2
2
2
5
2
5
2
5
2
3
2
5
2
23
(0)
(1)
(2)
In Table 26, column 4 and 5 the results are collected. Substituting Wn , Wn,x and Wn,x
in eqn. (60) yields the expression for m :
NA 2B 2
where
(111)
3kB T eff
9 4kB T
1
5kB T
4
5
2
2
= gJ 3
exp
+
kB T
+
exp
4
1
kB T
1 22
22
kB T
1
2
1
+ 1 + exp
exp
kB T
kB T
m = 0
2eff
and
(0)
(0)
(0)
(0)
1 = W1/2 W3/2 =
6
B2
35 0
5
B4
21 0
2 = W5/2 W3/2 = 12
B2 +
35 0
4
B 4.
21 0
(112)
Problems
8. In which respect does the magnetic susceptibility of a mononuclear lanthanide systems with Ln point symmetry (i) Oh , (ii) Dh, (iii) D2h differ?I
4.1
Heitler-London model of H2
r1 2
1
ra 2
rb 1
rb2
ra1
a
ra b
Valence bond ansatz: construction of products with orbital configuration a b using the
four spin orbitals
a
=
=
=
=
(113)
(114)
(115)
(116)
(S1 = S2 = 1/2)
2 = (S
1 + S
2 )2 and Sz = Sz1 + Sz2 with S = 1
D3 and D4 are eigenfunctions | S MS i of S
and MS = 1 and 1 (Spin triplet functions | 1 1 i and | 1 1 i), respectively, while linear
combinations of D1 and D2 yield functions with MS = 0:
D1 + D2 = | 1 0 i (spin triplet function);
D1 D2 = | 0 0 i (spin singlet function)
1 = D1 D2
|00i
2 = D3
|11i
3 = D1 + D2
|10i
56
4 = D4
| 1 1 i
a (1) b (1)d1 =
a (2) b (2)d2 6= 0
(1)(2)
q
1
= Nu [a (1)b(2) a (2)b (1)]
[(1)(2) + (2)(1)]
3
2
|
{z
}
4
anti
(1)(2)
|
{z
}
sym
2
where Ng = (2 + 2Sab
)
1/2
2
and Nu = (2 2Sab
)
1/2
anti
i (1, 2) = i (2, 1)
orbital sym (g), spin function anti
orbital anti (u), spin function sym
Symmetry of the orbital forces a distinct multiplicity of the spin function on account
of the Pauli principle
Evaluation of the energy E(S) and E(T ) of the singlet and triplet states
h
2 2
e2
e2
e2
e2
e2
h
2 2
+
H=
(1)
(2)
2me
r
rb1
2me
r
rb2
r12
{z a1
}|
{z a2
}
|
h(1)
h(2)
1 g i = 2(h + hab Sab ) + Jab + Kab
E(S) = h1 g1 |H|
1
2
1 + Sab
3 u i = 2(h hab Sab ) + Jab Kab
E(T ) = h3 u2 |H|
2
2
1 Sab
where
57
(117)
(118)
h =
=
hab =
Jab =
Kab =
E
(one-centre
a (i) h(i) a (i)
D
E
b (i) h(i)
one-electron integral)
b (i)
E
D
(transfer or hopping integral)
a (i) h(i) b (i)
2
a (1)b(2) e /r12 a (1)b (2)
(Coulomb integral)
2
(Exchange integral).
a (1)b(2) e /r12 a (2)b (1)
g1
E
1 g
H 1
E(S) =
D
2
2
(119)
Singlet-triplet splitting:
E(T, S) = E(T ) E(S)
2
2Kab 4hab Sab + 2Sab
(2h + Jab )
Application of the Heitler-London model to dinuclear complexes having S1 = S2 =
centres
1
2
58
4.2
Heisenberg operator
The interaction between the unpaired electrons of centre 1 with the ones of centre 2 are
phenomenologically described by an apparent spin-spin coupling, whose magnitude and
sign are given by the spin-spin coupling parameter (exchange parameter) Jex :
ex = 2Jex S
1 S
2
H
where
2Jex = E(T, S)
(120)
,
2J
S
2Jex S
J
ex
ex
1
2
}
| 2{z }
| {z
E(S)
E(T )
(121)
4.3
ex = 2Jex
H
i S
j
S
(123)
i<j
ex = 2Jex S
1 S
2 = 2Jex Sz1 Sz2 + Sx1 Sx2 + Sy1 Sy2
H
h
i
= 2Jex Sz1 Sz2 + 1 S+1 S2 + S1 S+2
2
(124)
Basis: spin functions in the form | MS MS i where the first MS refers to electron 1 and
the second to electron 2
H-Matrix:
1
1
2 2
MS1 MS2
1 1
Jex /2
2 2
1 1
2 2
1 1
2
2
1 1
2 2
1
2 2
1 1 1 1
2
2
2
2
Jex /2
Jex
Jex
Jex /2
Jex /2
59
(125)
2Jex 12 21 12 S1 S+2 12 12 = Jex (1)(1) = Jex
Results:
2 S +1
E
30 J
20 J
12 J
6J
2J
0
11
S
5
Fig. 18: Relative energies and multiplicities of the spin states of a dinuclear Fe3+ complex
(S = 52 ); for Cu2+ (S = 21 ) only the two lowest levels are relevant, while for Gd3+ (S = 72 ) the
two levels with S = 6 (E = |42Jex |) and S = 7 (E = |56Jex |) have to be added.
60
10
MS
/ 10 m mol
-7
S=1 4
-1
E(T,S)
-1
S= 0 0
E / cm
-1
10
1
2
-2
3
2
B0 / T
15
10
T /K
20
25
1
2
(126)
0 0
1 0
1 1
S MS 1 1
1 1 gB Bz
0
1 0
1 1
gB Bz
0 0
0
Matrix elements:
(1)
|11i = g Bz W
h11|H
B
Mz
|11i = gB
(1)
|1 1i = g Bz W
h1 1|H
B
Mz
|11i = gB
(1)
(1)
(0)
The remaining matrix elements (Zeeman coefficients) are zero. W|11i , W|11i , WS
(0)
WT
E(S), and
= E(T ) are substituted into the Van Vleck equation. After dividing by
2, the Bleaney-Bowers expression (m per centre) is obtained, here extended by 0 .
1
1
2Jex
NA 2B g 2
1 + exp
+ 0 , only applicable to a (Fig. 20)
(127)
m = 0
3kB T
3
kB T
61
e
c
0.8
-3
7
-1
cmol
/ ( 10 mol m )
1.0
d
0.6
b
0.4
0.2
0.0
200
100
300
100
200
300
400
T/K
2.5
e
2.0
d
c
1.5
meff
a
b
c
d
e
400
T/K
10
9
8
7
6
5
4
3
2
1
0
1.0
0.5
0.0
0
100
200
300
Fig. 21 a: m T diagram
Fig. 21 b: 1
m T diagram
Fig. 21 c: eff T diagram
400
T/K
0 NA 2B g 2
n
3kB T
)
S (S + 1)(2S + 1)(S ) exp E(S
kB T
P
E(S )
1
2
(128)
3d (Oh ), 3d (Td )
4f 7
4.4
A1
A2
S7/2
3d (Oh )
A2
where S =
1
2
[Cu2+ ],
1,
3
2
[Cr3+ ],
7
2
X
N
N
2
e2 X
h
i + V (ri ) +
(ri)
li si
+
H =
2m
r
e
ij
i>j
i=1
i=1
{z
} | {z } |
{z
}
|
(0)
SO
ee
H
H
) N
(H
k
N
X
XX
X
k
Bqk Cq
(i) + (1)q Cqk (i)
B0k C0k (i) +
B (
li + 2
si ) B0
+
+
N
X
q=1
i=1 k=0
ex = 2Jex
H
N X
N
X
{z
LF
H
sk
sl
|i=1
{z
Hmag
k=1 l=1
With regard to the LF symmetry of the lanthanide our choice is cubic and cylindrical for
two reasons: (i) the former is isotropic and the latter anisotropic, (ii) the number of LF
parameters is low [15] (cf. scheme (129)) 24 ) .
cubic
point group
no. of Bqk s
Bqk
24)
Oh
Td
2
B04 (+B44 )
B06 (+B46 )
D6h , D3h
D6 , C6v
4
B02 , B04
B06 , B66
trigonal
D4h , D4
D3d , D3
C4v , D2d
C3v
5
6
B02 , B04 , B44 B02 , B04 , B34
B06 , B46
B06 , B36 , B66
(129)
For the calculations we use only B04 as LF parameter, except for Ce3+ (cyl) where B02 is considered.
63
Cubic (Oh , Td )
N
X
cub =
H
LF
i=1
n
n
p
p
4
o
6
o
B04 C04 (i)+ 5/14 C4
(i) + C44 (i) + B06 C06 (i) 7/2 C4
(i) + C46 (i)
B44 =
fixed relations:
p
B46 = 7/2B06
p
5/14 B04 ,
Cylindrical (Dh )
cyl =
H
LF
N
X
i=1
Symmetry options
symmetry
cub
cyl
Ln
Bqk
Bqk
B04 (+B44 )
Ce (cub)
1 100
Ce (cyl)
coordination polyhedron
B0k > 0
octahedron
linear
Nd,Sm,Tb,Dy
B0k < 0
cube, tetrahedron
square antiprism
(cub, cyl)
4.5
4.5.1
Gd3+ Gd3+
2 000
S1 , S2 = 27 , n = 2:
(130)
12
-1
Jex = 0.5 cm
Jex = 0
10
-1
Jex = -0.5 cm
eff
8
6
4
2
0
0
50
100
150
200
250
300
T/K
64
5
1
a)
Occasionally, the selected |Jex |
values are larger than realistic values in
order to inspect more distinctly the Hex
effect (cf. Tab. 1).
600
Jex a)
4.5.2
We saw in 3.3.1 that on the basis of group theory it is possible to assign the LF split
terms of Ln ions. Symmetry aspects are also useful to assign the splitting produced by
cub (eq. (86)) + H
iso (eq. (120)) on the product functions of the
the combined effect of H
LF
ex
two interacting metal ions. The following examination is restricted to the simple case of
homodinuclear Ln systems equipped with orbital contributions of the magnetic moments.
iso in singlets, doublets and triplets.
The product states split under the action of H
ex
The split pattern is obtained in two steps:
application of ab (R) = a (R)b (R) , where i is the character and R is a symmetry operation, given in the headlines of Table 28.
Tab. 28: Character table of the symmetry groups O and Td
O
Td
1 (A1 )a)
2 (A2 )
3 (E)
4 (T1 )
5 (T2 )
1 1 b)
2 2
3 3
4 4
5 5
1 2
a)
b)
E 8 C3 3 C2 6 C4 6 C2
E 8 C3 3 C2 6 S4 6 d
1
1
1
1
1
1
1
1 1 1
2 1
2
0
0
3
0 1
1 1
3
0 1 1
1
1
1
1
1
1
1
1
1
1
1
4
1
4
0
0
9
0
1
1
1
9
0
1
1
1
1
1
1 1 1
1
1
1 2 3
1 3 4 5
1 3 4 5
2
a (R) and b (R) are the characters of the irreducible representations a and b ,
while ab (R) is the character of the (in general reducible) product state.
Example: The character system of the product functions ab
3 resulting from the
single-ion states a3 and b3 , 3 3 , is given in line 8 of Table 28.
P
application of ai = h1 R i (R) (R) to determine the irreducible parts of ab
3 ;
ai : frequency of i in ; h: order of the group;
i (R): character of the ith irreducible representation
i for the group element R;
P
(R): character of of the group element R;
R : sum over all group elements.
1
24 (1
1
24 (1
1
24 (1
a1
a2
a3
a4
= a5 = 0;
1 4 + 8 1 1 + 3 1 4 + 6 1 0 + 6 1 0) = 1
1 4 + 8 1 1 + 3 1 4 6 1 0 6 1 0) = 1
2 4 + 8 (1) 1 + 3 2 4 + 6 0 0 + 6 0 0) = 1
=
3 3 1 2 3
For metal ions with half-integer J three types of cubic states exists: the doublets 6 (E ),
7 (E ), and the quartet 8 (G ) (cf. Tab. 29).
Tab. 29: Character table of the double groups O and Td
Td
1 (A1 )a)
2 (A2 )
3 (E)
4 (T1 )
5 (T2 )
6 (E )b)
7 (E )
8 (G)
1
1
2
3
3
2
2
4
4 C3
4 C23
3 C2
3 C4
3 C34
6 C2
4 C23 R 4 C3 R 3 C2 R 3 C34 R 3 C4 R 6 C2 R
4 C3
4 C23
3 C2
3 S4
3 S34
6 d
R
2
3
4 C3 R 4 C3 R 3 C2 R 3 S4 R 3 S4 R 6 d R
R
1
1
2
3
3
2
2
4
1
1
1
0
0
1
1
1
1
1
1
0
0
1
1
1
1
1
2
1
1
0
0
0
1
1
0
1
1
2
0
1
1
0
1
1
2
0
1
1
0
1
1
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
1
1
1
1
0
0
0
0
1
1
0
0
1
1
1
1
0
0
2
2
0
2
0
0
2
0
2
0
0
0
0
0
2
2
0
2
0
0
2
0
2
0
0
0
0
0
0
0
0
0
0
0
4 6 d) 6 6
5 6
6 6
2 6
2 2
3 6
4 4
4 7
6 6
5 7
6 6
2 7
2 2
3 7
4 4
2 8
4 4
3 8
8 8
4 8
12 12
5 8
12 12
1
1
1
1
1
1
6 6 c) 4
7 7
4
8 8
16
6 7
4
6 8
8
7 8
8
a)
b)
c)
d)
4
4
16
4
8
8
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
irreducible representations
of the coupled states
1 4
1 4
1 2 3 24 25
2 5
3 4 5
3 4 5
6 8
7 8
7
8
7 8
6 8
6
8
8
6 7 8
6 7 2 8
6 7 2 8
1 5 : integer J; 1 3 : non-magnetic.
6 8 : half-integer J; magnetic.
Direct products where both metal ions have half-integer J or S.
Direct products where one metal has integer J and the other has half-integer J or S.
66
4.5.3
67
Remark to Fig. 26: The product states 8 8 are split by Hex into 1 (singlet (1))
2 (singlet (1)) 3 (doublet (2)) 2 4 (triplets (3)) 2 5 (triplets (3)). For Jex > 0, the
sequence of split terms is given with increasing energy by (3) (3) (3) (1) (3) (2)
(1); for Jex < 0, this order is reversed. (Remember that only the triplets are magnetic.)
Ce3+ Ce3+ (cyl)
310
(2)
| 1/2 i
(2)
| 3/2 i
ELF /cm1
0
(2)
| 5/2 i
68
4.5.4
For B04 > 0 (oct. LF; W < 0, x > 0; cf. Ex. 3.11)
the LF ground state is 6 while for B04 < 0 (cub/
(2)
tet) 8 is lowest in energy (see Fig. 28).
B04 < 0
B04 > 0
4.0
-1
-1
Jex = 1 cm
Jex = 1 cm
Jex = 0
Jex = 0
-1
Jex = -1 cm
-1
Jex = -1 cm
3.5
3.5
3.0
eff
eff
3.0
2.5
2.5
2.0
2.0
1.5
1.5
0
10
20
30
40
50
10
20
30
40
50
T/K
T/K
Results: Fig. 30: Only a little increase of eff is observed after switching on Hex ;
Fig. 31: For the ferromagnetically coupled Nd-Nd the peak of eff at 1 K is remarkable.
69
4.5.5
0.40
-1
Jex = 1 cm
-1
Jex = 1 cm
Jex = 0
0.7
Jex = 0
0.35
-1
Jex = -1 cm
-1
0.30
0.6
0.25
0.5
0.20
0.4
eff
eff
Jex = -1 cm
0.15
0.3
0.10
0.2
0.05
0.1
0.0
0.00
0
10
12
14
T/K
10
12
14
T/K
7 7 1 4
8 8 1 2 3 24 25
Fig. 33: eff vs. T plot of an exchange-cou-
Fig. 32: eff vs. T plot of an exchange-coupled SmSm unit in oct. LF (LF ground state
7 ). For Jex > 0 the eff values are only little
enhanced; eff 0 for Jex < 0 .
4.5.6
B04 > 0
TIP at T 10 K
B04 < 0
70
4.5.7
B04 > 0
B04 < 0
(1)
(3)
12
10
10
-1
-1
Jex = 1 cm
Jex = 1 cm
Jex = 0
Jex = 0
8
-1
-1
Jex = -1 cm
eff
eff
Jex = -1 cm
0
0
20
40
60
80
100
120
140
20
40
60
80
100
120
140
T/K
T/K
Fig. 36: eff versus T plot of an exchangecoupled DyDy unit in an octahedral LF (B04 =
2 000 cm1 ; LF ground state 6 ).
Results: For LF ground state 6 (Fig. 36) Jex > 0 enhances eff in the whole temperature
range while for Jex < 0 eff diminishes. For LF ground state 8 (Fig. 37) it is remarkable
that for Jex > 0 eff drops steeply from 11.3 to 2.2 far below eff = 8.9 of the uncoupled
Dy-Dy pair, caused by a very weak magnet contribution of the ground triplet.
71
4.5.8
(2)
420 | 11/2 i
(2)
404 | 9/2 i
..
..
..
.
.
.
(2)
315 | 7/2 i
(2)
305 | 13/2 i
..
..
..
.
.
.
(2)
206 | 5/2 i
..
..
..
.
.
.
(2)
114 | 3/2 i
ELF /cm1
(2)
62
| 1/2 i
(2)
| 15/2 i
4.5.9
eff
eff
see 4.6.5
see 4.6.5
3+
Ln
free ion
cub
cyl
a)
Jex /cm1
+1
+5
+5
Ce3+
Nd3+
Sm3+
Tb3+
Dy3+
2.54
3.62
0.84
9.72
10.64
+/
2.3
3.8
0.48
TIP
11.2
++
b)
b)
++
12.7
24.0
5.2
At T = 2 K.
b)
72
4.6
4.6.1
ML MS
J MJ
q
q
1
3 23 1 12 + 13 0 12
2
5
1
2
q
q1
1 1
1
2
1
3 0 12
3
2
2
2
MJ = ML + MS J
1
2
1
2
P1/2
21
ex can be
Since the calculations are restricted to the ground multiplet J = 1/2, H
applied in the form
ex = 2Jex (g 1) J
H
S
=
2(g
1)J
J
S
+
J
S
+
J
S
ex
z z
x x
y y
J
| J {z }
De Gennes
i
h
1
(131)
= 2(gJ 1)Jex Jz Sz + 2 J+ S + J S+
with the advantage, that for the p system the basis functions |J MJ i can be used.
1 1 1
J
2 2
6 ex
1 1
1
2 2
J
16 Jex
3 ex
1 1
1
2
J
16 Jex
2
3 ex
1 1
1
J
2 2
6 ex
(132)
2( 31 )Jex 12 21 Jz Sz 12 21 = ( 23 )Jex 12
25)
1
2
= 16 Jex
The complete set of basis functions of 2 P1/2 and 2 P3/2 is given Table 6.
73
74
4.6.2
Fig. 39: eff vs. T plot of a ferromagnetically coupled and an uncoupled Gd3+ Cu2+ unit as a
function of the applied field: Jex = 5 cm1 (solid lines), Jex = 0 (dotted lines); gGd = gCu = 2;
B0 = 0.01 (a), 0.1 (b), 1.0 T (c).
The solid lines in Fig. 39 refer to a heterodinuclear Gd3+ Cu2+ complex (SGd = 7/2;
SCu = 1/2) of ferromagnetically coupled spins with the exchange parameter Jex = 5 cm1 .
Solely for B0 = 0.01 T (a) the expected plateau at eff = 8.94 is obtained at T
2 K except for T 0.2 K. In order to adjust Jex by fitting procedures to corresponding
experimental data, it is justified to apply a susceptibility equation valid for B0 0 (Van
Vleck approach). For measurements with the stronger field 0.1 T (b) and all the more 1 T
(c) deviations are obvious, which are not adequately described by the simple model. They
could be misinterpreted, however, as an additional antiferromagnetic intermolecular spinspin coupling. A similar problem arises if a corresponding uncoupled system is considered
(Fig. 39, dotted lines). A straight line parallel to the T axis is obtained at eff = 8.13 for
B0 0 (curve a) while increasing deviations from this reference line to lower values are
again obvious with increasing field (b, c), falsely interpreted as an intramolecular spinspin coupling of antiferromagnetic nature. Hence, either the measurements are carried
out at very weak applied fields or the extended susceptibility formula is used as in the
given model calculations.
75
4.6.3
After discussion of the p1 s1 case under the influence of HSO + Hex , we pass to the f 1 s1
system perturbed by HSO + HLF + Hex . To clarify the computational procedures, the lowtemperature magnetic behaviour of a system consisting of a 4f 1 ion, exchange-coupled
to a pure spin system with S = 12 , is described with restricted basis functions. After
oct
switching on HSO and HLF
acting on the Ce3+ ion, only the ground doublet 7 (E ) is
considered for spin-spin coupling with the 2 S1/2 system (represented here by Cu2+ [3d9 ]).
The procedure is outlined by three steps:
4f 1
SO
F5/2
LF
7 + 8
2 5
7 6
H oct
Hex
F5/2 + 2 F7/2
The wave functions of the Ce3+ doublet 7 are given in Tab. 32. The state is assumed to
be thermally isolated (similar to 2 P1/2 of the ps molecule (see section 4.6.1) on account
of HSO > HLF (oct) Hex ) and exchange-coupled to the S = 12 pure spin system.
Table 32: Doublet wavefunctions of 2 F5/2
perturbed by a cubic ligand field
i
|M
|MJ i
q E q
E
E
1 5
5
3
=
E
6
2
6
2
q
E q E
E
1
52 56 23
=
E
6
1
5
E , 2
Jex
42
1
5
5
E , 2
J
21
Jex
42 ex
1
5
5
21
E , 2
Jex
J
42 ex
1
5
E , 2
42
Jex
(133)
5
2( 71 )Jex E , 21 Jz Sz E , 12 = ( 72 ) ( 61 )( 52 )( 12 ) + ( 56 )( 23 )( 12 ) Jex = 42
Jex
Results:
15
Esinglet = 42
Jex
5
Etriplet = 42
Jex
10
E = Etriplet Esinglet = 21
Jex
Simulation calculations: For Jex > 0, the triplet state 5 is the ground state while for
Jex < 0 the singlet 2 has the lowest energy (as is predicted for the restricted basis).
76
Figs. 40 and 41 exhibit the corresponding eff vs. T plots and Figs. 42 and 43 the 1
m vs.
26)
T plots . Inspecting Fig. 40 it is remarkable that at T > 5 K the antiferromagnetic spinspin coupling yields higher eff values than the ferromagnetic spin-spin coupling. This
finding is the consequence of the fact that corresponding to Hunds third rule J is L S
and that for a spin-spin coupling of antiferromagnetic nature the spin parts of the JS
dinuclear unit cancel to a certain degree, so that the orbital part is enhanced. As opposed
to that, the spin-spin coupling of ferromagnetic nature leads to a further weakening of
the orbital part of the 4f 1 configuration. The fact that the eff values of the uncoupled
CeCu unit at T < 2 K are larger than the eff values of both coupled systems is the
consequence of the reduction of orbital parts according to the exchange splitting.
3.0
1.0
-1
Jex = 10 cm
Jex = 0
-1
2.5
Jex = -10 cm
0.8
2.0
-1
Jex = 10 cm
eff
eff
0.6
-1
Jex = -10 cm
1.5
0.4
1.0
0.2
0.0
0.0
0.5
0
10
20
30
40
50
0.5
1.5
2.0
T/K
T/K
Jex = 10 cm-1
Jex = 0
Jex = -10 cm-1
-3
2.0
1.0
m / 10 mol m
Notice that, apart from the lowest temperatures, the spin-spin coupling produces
small parallel shifts of the 1
m T curve of
the uncoupled system (Fig. 42, black line).
The shifts are described by
1
= 1
m
m MF
1.5
-1
1.0
0.5
where (1
m ) corresponds with the shifted
curve and MF is the molecular field parameter.
0.0
0
50
100
150
200
250
300
T/K
26)
The data presented in Figs. 40 43 have been calculated with the full basis set of 14 2 = 28
functions (using the computer program CONDON).
77
Fig. 43 displays that at T < 5 K the blue curve (Jex < 0) tends to temperature
independent paramagnetism (TIP), whereas the red curve (Jex > 0) exhibits Curie paramagnetism.
1.4
-1
Jex = 10 cm
Jex = 0
1.2
-1
Jex = -10 cm
0.8
m / 10 mol m
-3
1.0
-1
0.6
0.4
0.2
0.0
0
T/K
..
.
(2)
316
ELF /cm1
7
0
..
.
3
..
.
(3)
321
ELF /cm1
(1)
(3)
Jex > 0
4/5
..
.
(3)
(1)
Jex < 0
5
2
1 vs. T plot of an exchange-coupled Ce3+ (oct)Cu2+ (S = 1 ) unit, showing the
Fig. 43: m
2
crossing of the red and blue curve around 0.2 K and the energy splitting of the two lowest states
5 and 2 due to Hex .
In Fig. 43 the black reference curve of the uncoupled system follows Curies law in
a good approximation on account of the thermally isolated states 7 (Ce) and 6 (Cu).
nd
order Zeeman
Both states show linear 1
m T behaviour (apart from a very small 2
contribution of 7 ).
4.6.4
Figs. 44 and 45 exhibit the eff vs. T plots of the heterodinuclear species Nd(S = 21 )
(Cu2+ ) and Nd(S = 23 ) (Cr3+ ), respectively, in cubic LF. Inspecting both Figures it
is remarkable that, similar to the Ce-Cu system (Figs. 40 43) the antiferromagnetic
spin-spin coupling leads to higher eff values than the ferromagnetic coupling on account
78
of the fact that the residual component of the spin-spin coupling strengthens the orbital
part of the lanthanides moment.
4.5
4.0
3.5
-1
Jex = 5 cm
Jex = 0
3.0
-1
Jex = -5 cm
eff
2.5
2.0
1.5
1.0
0.5
0.0
0
50
100
150
200
250
300
T/K
Fig. 44: eff vs. T diagram of an exchangecoupled Nd3+ Cu2+ unit with Nd in a cubic
LF (B04 = 2 000 cm1 ) and Cu regarded as a
pure spin system S = 21 .
Fig. 45: eff vs. T diagram of an exchangecoupled Nd3+ Cr3+ unit with Nd in a cubic
LF (B04 = 2 000 cm1 ) and Cr regarded as a
pure spin system S = 32 .
4.6.5
12
-1
Jex = 5 cm
Jex = 0
-1
Jex = -5 cm
eff
10
4
0
10
20
30
40
50
T/K
79
eff
k
eff
k
eff
eff
eff
eff
Gd3+ V3+
4.6.6
The 4f ion (Gd3+ , 8 S7/2 ) is regarded as a pure spin system, unaffected by ligands, zerofield splitting ignored. The electronic situation of the transition metal ion (for example
V3+ ) depends on HLF and HSO , demonstrated in Figure 48.
3
4
2
eff
mol
-1
/ 10
mol m
-3
0
0
100
200
300
T/K
Fig. 48: eff versus T and 1
m versus T diagrams showing the predicted mag3+
2
netic behaviour of V [3d ] ion in octahedral (solid lines) and orthorhombic surrounding (dashed lines); parameter values [6]: Racah parameters B = 861 cm1 ,
C = 4 165 cm1 , = 209 cm1 , B04 = 25 000 cm1 for the octahedral coordination,
B02 = 34 000, B22 = 16 000, B04 = 3 500, B24 = 1 800, B44 = 2 300 cm1 for
orthorhombic symmetry.
While the octahedral system shows between 100 K and 300 K nearly temperature independent paramagnetism, the orthorhombic one approximates to a Curie paramagnet with
only slightly varying eff . The two situations differ distinctly in the orbital contributions
to the magnetic moment. The expected magnetic properties of the V3+ Gd3+ unit is
given in Figures 49 and 50).
If V3+ is octahedrally coordinated, eff in the uncoupled situation (Jex = 0) is temperature dependent owing to the loss of magnetic activity of vanadium with decreasing
temperature (see the solid reference line in Figure 49, eff = 8.37 (300 K) 8.05 (50 K)).
When T approaches zero solely the magnetism of Gd3+ contributes. The uncoupled system
with vanadium in orthorhombic symmetry, however, exhibits a temperature independent
eff , except for T 25 K. This is the consequence of the nearly total quenching of orbital
contributions of vanadiums valence electrons owing to the low-symmetric ligand field. In
consequence, the transition metal ion behaves in a wide temperature range as a pure spin
system S = 1.
80
10
eff
7
J
ex
-1
[cm ]
10
5
[T]
0
2
0
-2
-5
-10
100
50
100
200
150
300
T / K
Fig. 49: eff versus T diagram (B0 = 0) showing the predicted magnetic behaviour of an exchange-coupled 3d2 (oct)4f 7 species where Jex =
0, 2, 5, 10 cm1 ; insert: model calculations for Jex = 2 cm1 and B0 = 0,
1, and 5 T.
10
eff
10
Jex [cm
-1
10
B0 [T]
-2
-5
-10
100
50
200
100
150
300
T/K
Fig. 50: eff versus T diagram (B0 = 0) showing the predicted magnetic behaviour of an exchange-coupled 3d2 (rhomb)4f 7 species where Jex =
0, 2, 5, 10 cm1 ; insert: model calculations for Jex = 2 cm1 and B0 = 0,
1, and 5 T.
81
Problems
9. The Bleaney-Bowers formula, eq. (127), approaches for high temperature the CurieWeiss law m = C/(T ). Determine the relation between
Jex and . Is the result
P
in agreement with the relation = (2S(S + 1)/3kB ) i zi Jex ?
10. What magnetic behaviour is obtained, if in the Bleaney-Bowers formula, eq. (127),
Jex is set to 0?
11. What are the electronic (4f N ) and the LF symmetry relevant preconditions for building up a homodinuclear Ln molecule with large eff and high magnetic anisotropy?
12. Magnetochemical results are often presented as m T , 1
m T or eff T diagrams.
What type of diagram is suited for (a) Curie paramagnetism, (b) intramolecular
ferromagnetic interactions, (c) intramolekular antiferromagnetic interactions, (d)
diamagnetic behaviour, (e) TUP behaviour?
13. To reliably characterise magnetic properties measurements at different field strength
are essential. What is the reason?
14. A frequent mistake in magnetochemical investigations is the application of too strong
magnetic fields. Why may this be unfavourable?
15. To evaluate the paramagnetic part of the susceptibility of a compound with macrocyclic ligands, the problem may occur that the incremental method for the diamagnetic correction is not as precise as necessary. What is to be done?
16. For a polynuclear complex one observes at high temperature Curie-Weiss behaviour
with > 0 and in the low-temperature region field-dependent susceptibilities. What
magnetic collective effect can be expected?
17. With decreasing temperature the eff data of a homodinuclear compound increase
weakly and then, after passing a maximum at low temperature, steeply drop. (i)
What is the reason for this behaviour? (ii) What model (susceptibility expression)
should be tried to simulate the behaviour?
18. You notice that the paramagnetic properties of a homodinuclear molecular compound is not satisfactorily described with the corresponding eq. (128). What simple
extensions of the model are in principle possible?
82
Solutions
1. Verify the Russell-Saunders ground multiplets for Nd2+ and Er3+ given in Table 2.
Nd2+ [4f 4 ]: L = 6, S = 2, J = 4 = 5 I4 .
Er3+ [4f 11 ]: L = 6, S = 3/2, J = 15/2 = 5 I15/2 .
2. Evaluate the susceptibility equation for a free p1 system perturbed by spin-orbit
coupling. (Hint: Apply eq. (3); note that for the single-electron system LS .)
3
1
+ 21 32 2
2
2
3
= 23
g1/2 = 1 +
1
2
5 2 1 2 3
3
+
2
= 43
g3/2 = 1 + 2 2 3 2 5 2
2 2 2
2
k T
2
1
3
2
B
+
2
2
= 31 + 8 k9BT
1/2 = 32
2
2
3
3
2 3 5
3/2 = 43
+ 2 34 1 43 2 kBT = 20
4 k9BT
2
2
3
Z
NA 2B
, where
m = 0
3k T
N
B
8 kB T
1
Z = 2 3 + 9 exp 8 k3BT + 4 20
3
NA 2B
3kB T
4 kB T
9
exp 815
kB T
1 8 kB T
3
40 8 kB T
exp
+
+
3
9
3
9
2 kB T
3
1 + 2 exp
2kB T
(134)
To verify eq. (134), it is recommended to develop the susceptibility equation for the
borderline case 0 (use
p the approximation exp x 1 x). For an uncoupled
1
system the relation = L(L + 1) + 4S(S + 1) B holds,
so that for the p system
(L = 1, S = 1/2) the permanent magnetic moment = 5 B and the susceptibility
equation m = 0 5NA 2B /(3kB T ) must be the results.
3. Use the spherical harmonics in Table 3 and construct the real functions dx2 y2 and
dxy with the help of eqs. (8). Verify the results with the data given in Table 4.
1
2
[n,l,ml + n,l,ml ] =
[n,l,ml
hq
15
32
(cos
2
+
i
sin
2
cos
2
+
i
sin
2)
n,l,ml ] = i1 2
32
q
15
=
sin2 sin 2 = dxy
16
=
1
2
1
2
15
16
83
4. Calculate the matrix elements hl, ml |lq |l, ml i (where q stands for z, +, ):
2
(a) h0, 0|lz |0, 0i, (b) h2, 2|l+ |2, 1i, (c) h2, 2|l+
|2, 0i, (d) h2, 0|l+l |2, 0i.
(Hint: Apply eqs. (20))
| {z }
1
2
2
2
h2, 2|l+ |2, 1i = 2 6 h
h2, 2|2, 2i = 2 6 h
h2, 0|l+|2, 1i = 6 + (1 + 1) h
6h
h2, 0|2, 0i = 6 h
(d) h2, 0|l+ l |2, 0i = 6 h
2
| {z }
1
6. What levels (multiplets J) may arise from the terms (a) 1 S, (b) 2 P , (c) 3 P , (d) 3 D,
(e) 4 D? How many states (distinguished by the quantum number MJ ) belong to
each level?
(a) 1 S: J = 0 (1 state)
(b) 2 P : J = 12 (2), J = 32 (4)
(c) 3 P : J = 0 (1), J = 1 (3),
(d) 3 D: J = 1 (3), J = 2 (5),
(e) 4 D: J = 21 (2), J = 32 (4),
J = 2 (5)
J = 3 (7)
J = 52 (6),
J=
7
2
(8)
7. Determine the crystal field overall splitting (CFOS) for the Dy3+ ion in Cs2 NaDyCl6
on the basis of Fig. 35. Apply eqs. (102) and (103) as well as the entries in Table 23
and take the B04 and B06 values for Dy from Table 16.
J F (4)B04
27)
8
= Wx =
135 135
15
2
According to the phase convention one has to change the sign of a and b (cf. Example 2.3).
84
J F (6)B06
= W (1 |x|) =
x
= 5.401
1 |x|
4
3 864 861
3465
4
x = 0.8438,
1
(exp x 1 x for x 0)
1
2Jex
1 + exp
3
kB T
2 2
0 NA 2B g 2
0 NA 2B g 2
3
1
NA B g
=
=
0
2Jex
4
2Jex
Jex
3kB
4
T
T 4kB
3kB T
3 3kB T
T
2kB
X
Jex
Jex
;
[2S(S + 1)/3kB ]
zi Jex,i =
for S = 12 , i = 1, zi = 1
=
2kB
2k
B
i
10. What magnetic behaviour is obtained, if in the Bleaney-Bowers formula, eq. (127),
Jex is set to 0?
NA 2B g 2
m = 0
3kB T
1
NA 2B g 2
ex =0
J
m =
2Jex
1
3kB T
1 + exp
3
kB T
3
4
Curie law
11. What are the electronic (4f N ) and LF symmetrical preconditions for building up a
homodinuclear Ln molecule with large eff and high magnetic anisotropy?
(i) Ln from the second half of the Ln series (on account of free ion ground states
with J = L + S);
(ii) Ln with half-integer J (nonmagnetic LF ground states of the Ln ion are ruled
out);
(iii) Ln with large S and large L;
(iv) anisotropic LF, for example unique axis anisotropy (cylindrical, hexagonal,
tetragonal)
85
(b) eff T
86
Appendix 1
6.1
Magnetic quantities
0
B
H
Quantity
SI
CGS-emu
Permeability
of vacuum
Magnetic
induction
Magnetic field
strength
0 =
4 107 H/m b
B = 0 (H + M )
T = Vs/m2
H
A/m
B0 = 0 H
T = Vs/m2
M
A/m
m = MV
Am2 = J/T
1
B = e
h/(2me )
Am2
= M /c
Am2 /kg
Mm = M M/e
Am2 /mol
a = Mm /(NA B )
Am2
1
M = H
1
g = /
m3 /kg
m = M/
m3 /mol
[3kB /0 NA 2B ]1/2 [m T ]1/2
1
B0
Magnetic field
Magnetization
Magnetic
dipole moment
m/B
B
Mm
a
Bohr magneton
specific magnetic
dipole momentd
Molar
magnetizationf
Atomic magnetic
dipole momentg
a /B
g
m
eff
Magnetic volume
susceptibility
Magnetic mass
susceptibility
Molar magnetic
susceptibility
Effective Bohr
magneton number [2]
B = H (ir) + 4M
G
Oe
87
104 T/G
103
(A m1 )/Oe
4
104 T/G
103 (A/m)/G
m = MV
G cm3
1
B = e
h/(2me )
Gcm3
= M /
Gcm3 /g
Mm = M M/
Gcm3 /mol
a = Mm /(NA B )
Gcm3
1
M = (ir) H (ir)
1
(ir)
g = (ir) /
cm3 /g
(ir)
m = (ir) M/
cm3 /mol
[3kB /0 NA 2B ]1/2 [m T ]1/2
1
Multiplicative factor applied to the value in CGS-emu units to obtain the value in SI units.
H = Henry; H/m = Vs/(Am).
c
= mass density.
d
Specific magnetization; s = specific saturation magnetization.
e
M = molar mass.
f
s
Mm
= molar saturation magnetization.
g s
a = atomic magnetic saturation moment.
b
Factora
10
1
Am2 /Gcm3
Am2 /kg
Gcm3 /g
1
4
4 m3 /kg
103 cm3 /g
4/106 m3 /cm3
1
The legal SI units are not generally accepted. The CGS-emu system is still widely
used in magnetochemistry. Therefore, use magnetic quantities which are independent of the two systems, e. g., eff or 2eff instead of m T .
Do not mix the systems, e. g., use G (Gauss) instead of T (Tesla) in the CGS-emu
system.
Use B0 = 0 H in graphical representations (conversion factor 104 T/Oe).
6.2
1J
1 eV
1 s1
1 cm1
1K
1T
1 kJ/mol
s1
cm1
eV
1.0218
10
1018
1
19
1033
2.1799
10
14
6.62607 4.13567 1
1034
1022
8.6554
10
1015
10
10
1011
10
1022
1.6045
10
1011
1023
1020
104
101
1011
1013
1.72760 9.64853
10
102
105
1010
101
24
10
1.48874 8.31451
10
10
kJ/mol
10
10
103
5.58494
103
102
1012
88
101
102
102
Tab. 35:
F (4)a)
5
2
7
2
9
2
11
2
13
2
15
2
15/2
12
32
72
92
11
2
13
2
15
2
3 13
15/2
21/2
18
15
28
15/2
108
15/2
189 129
0
12 13
63 13
2
18
105/2
15/2
84
64
3/2
105/2
5
2
7
2
9
2
11
2
13
2
15
2
31
22
18
33
27
92 132
11
33
77
143
91
14
21
273
7 8
54
96
66
99
24
39
39
13
52
Tab. 36:
9 11
756 621
15/2
a)
3 17
36
hMJ ||MJ i
25
4 [20]
Matrix elements of the operator equivalents O
4
F (44)
h 25 || 32 i h 72 || 21 i
3 35/2
5
3 35/2
5 3
35
p
21 5/2
5 3
5 2
3 35
35
3 105
p
3 35/2
42 5
35 6
p
3 35/2
42 15 10 231
hMJ ||MJ 4i
h 92 || 12 i
p
p
p
p
p
35/2
35/2
35/2
35/2
35/2
3 2
5 21
15 22
15 77
h 11
|| 3 i
2 2
h 13
|| 5 i h 15
|| 7 i
2 2
2 2
165
15 11
5 429
715
5005
1365
h2|| 2i h3|| 1i
h4||0i
h5||1i
h6||2i h7||3i h8||4i
15
5 7
70
35
5 42
3 70
210
70
21 10 15 14
5 66
3 55
126
70 3 5 462 15 33 5 143
1001
89
6 [20]
Matrix elements of the operator equivalents O
0
Tab. 37:
F (6)
5
2
7
2
9
2
11
2
13
2
15
2
23
52
72
29
11
2
13
2
15
2
20
315/4
315
945/2
135
3465/4
200
11 31
11
227 11 319
143
87
59
12
36
29 48
15
200 125
20
945/4
3465/4 120
40
40
22 17
20
22
85
39 117
43
50
197
176
93
128
55
22
55 286
65
65
143
78 169 104
6 [20]
Matrix elements of the operator equivalents O
4
Tab. 38:
25 185
945/2
25
10 11
0
315/2
45
25
315/4
5
2
7
2
9
2
11
2
13
2
15
2
75
hMJ ||MJ i
12
F (64)
h 25 || 32 i
0
0
45 14/4
7 3
315 2/4
16 3
135 7/2
63
135 14/2 56 5
495 14/4 42 15
135 14/4
45 14/4
135 14/4
45 14/4
495 14/4
h2|| 2i
14
168
168
1260
252
h 72 || 21 i
3 35
6 2
105
21 6
6 231
hMJ ||MJ 4i
h 29 || 12 i
h 11
|| 3 i
2 2
h 13
|| 5 i
2 2
h 15
|| 7 i
2 2
30 2
13 21
13 22
3 77
7 165
46 11
7 429
6 715
3 5005
5 1365
h3|| 1i
h4||0i
h5||1i
h6||2i
h7||3i
h8||4i
7
2 70
13 42
12 70 15 210
35 10
8 14 21 66
28 55
90
References
[1] A. F. Williams, A Theoretical Approach to Inorganic Chemistry, Springer, Berlin
(1979).
[2] J. H. Van Vleck, Electric and Magnetic Susceptibilities, Oxford University Press,
Oxford 1932.
[3] E. U. Condon, G. H. Shortley, The Theory of Atomic Spectra, University Press,
Cambridge 1970; p. 179.
[4] A. Abragam, B. Bleaney, Electron Paramagnetic Resonance of Transition Ions,
Clarendon Press, Oxford 1970; p. 289.
[5] (a) S. Hatscher, H. Schilder, H. Lueken, W. Urland: Practical Guide to Measurements and Interpretation of Magnetic Properties (IUPAC Technical Report), Pure
Appl. Chem. 2005, 77, 497 - 511; (b) German version: Angew. Chem. 2006, 118,
8283 - 8240.
[6] J. S. Griffith, The Theory of Transition-Metal Ions, Cambridge University Press,
Cambridge, 1971.
[7] H. Lueken, Magnetochemie, Teubner, Stuttgart, Leipzig 1999.
[8] H. L. Schlafer, G. Gliemann, Einf
uhrung in die Ligandenfeldtheorie, Akad. Verlagsges., Wiesbaden 1980; engl. edition: Basic Principles of Ligand Field Theory,
Wiley-Interscience, New York 1969.
[9] P. W. Atkins, R. S. Friedman, Molecular Quantum Mechanics, 4th ed., Oxford
University Press, Oxford 2005.
[10] H. Bethe, Termaufspaltung in Kristallen, Ann. Phys. 1929, 5, 133 - 208; further
publication of this article in: Symmetry in chemical theory, Benchmark papers in
inorganic chemistry, vol. 4, ed. J. P. Fackler, Jr., Dowden, Hutchinson, Ross, Inc.
1973 (ISBN: 0-87933-018-X).
[11] C. J. Ballhausen, Ligand Field Theory, McGraw-Hill, New York 1962.
[12] H. Schilder, H. Lueken, Computerized magnetic studies on d, f, dd, ff, and dS,
fS systems under varying ligand and magnetic fields J. Magn. Magn. Mater. 2004,
281, 17- 26.
[13] P. A. Tanner, V. V. Ravi Kanth Kumar, C. K. Jayasankar, M. F. Reid, Analysis
of spectral data and comparative energy level parametrizations for Ln3+ in cubic
elpasolite crystals, J. Alloys Comp. 1994, 215, 349 - 370.
[14] M. J. Reisfeld, G. A. Crosby, Analysis of the absorption spectrum of cesium uranium(V) hexafluoride, Inorg. Chem. 1965, 4, 65 - 70.
[15] C. Gorller-Walrand, K. Binnemans, Rationalization of Crystal-Field Parametrization, in K. A. Gschneidner, Jr., L. Eyring, Handbook on the Physics and Chemistry
of Rare Earths, Elsevier, Amsterdam 1996, vol. 23, ch. 155, p. 121 - 283.
91
[16] K. R. Lea, M. J. M. Leask, W. P. Wolf: The Raising of Angular Momentum Degeneracy of f-Electronic Terms by Cubic Crystal Fields, J. Phys. Chem. Solids, 1962,
23, 1381 - 1405.
[17] D. Smith, J. H. M. Thornley, The use of operator equivalents, Proc. Phys. Soc. 1966,
89, 779.
[18] K. W. H. Stevens, Matrix elements and operator equivalents connected with the
magnetic properties of rare earth ions, Proc. Phys. Soc. A [London] 1952, 65, 209.
[19] R. J. Elliott, K. W. H. Stevens, Proc. Roy. Soc. A [London] 1953, 218, 553.
[20] R. J. Birgeneau, Tables of matrix elements of Racah tensor operator equivalents,
Can. J. Phys. 1967, 45, 3761 - 3771.
[21] M. Hutchings, Point-charge calculations of energy levels of magnetic ions in crystalline electric fields, Solid State Phys. 1964, 16, 227.
[22] A. J. Kassman, Relationships between the coefficients of the tensor operator and
operator equivalent methods, J. Chem. Phys. 1970, 53, 4118 - 4119.
[23] R. J. Elliott, K. W. H. Stevens, The Theory of magnetic resonance experiments on
salts of the rare earths Proc. Phys. Soc. 1953, A 218, 553.
[24] H. Lueken, M. Meier, G. Klessen, W. Bronger, J. Fleischhauer, J. Less-Common
Met. 1979, 63, P35 - P44.
[25] A. Schroder, R. van den Berg, H. von Lohneysen, H. Lueken, Solid state commun.
1988, 65, 99 - 101.
[26] C. A. Morrison, R. P. Leavitt, and D. E. Wortman, J. Chem. Phys. 1980, 73, 2580
- 2598.
92