R.

Krishnamurthy I Computing Ligand Field Potentiah

and Ward B. Schaap'
Indiana Universitv 1 and Relative Energies of d Orbitals
Bloomington, 47401
I A simple, approach

I n recent years, the concepts and uses complicated to be presented to beginning students.
of the crystal field or ligand field theories have been In this article we attempt to bridge this gap by present-
recognized as an essential part of the training of ing a simple, general approach which permits the deriva-
chemists. The use of simple crystal field arguments in tion of the expressions for ligand field potentials and
the description of bonding, molecular geometry, mag- the quantitative evaluation of the energies of d orbitals
netism, spectra, and reactivity of transition metal (for the d' configuration) in a large variety of geometri-
complexes is becoming commonplace in teaching ma- Eal situations. These quantities are obtained from the
terials being written for lower division college students. potential expressions and from energies of the d orbitals
Effective use is made of pictorial representations of the derived for a minimum number of small primary ligand
boundary contours of the five real d orbitals, as illus- configurations. The primary groups are combined in
trated in Figure 1, which may have different energies building-block fashion to produce more complex struc-
and symmetry properties with respect to the arrange- tures. Similarly, the effects of the potential fields of
ment of ions or ligands in crystals or complexes. I n the primary groups on the energies of the individual d
orbitals are added algebraically to produce the effectof
the overall structure. The geometric aspects of such
combinations of primary groups are easily visualized;
yet the results for both the ligand field potentials and the
energies of the d orbitals are rigorously correct.
Background
The basic principles of crystal field theory were
originally developed by Bethe (1) in 1929. His classic
paper considered. the effects of the electrostatic po-
tential on a "free ion" which result when that ion is
incorporated in an electrical field of prescribed sym-
metry. He showed by means of group theoretical argu-
ments how the degenerate orbitals and atomic terms of
an ion of a given electronic configuration split when
Figvre 1. Illustration of h e boundary mntourr and tho angular de-
pendence of the ''real" d orbit01 wove functions in tetragono1 orientation. placed in a lattice of a particular symmetry. Because
the effects of the nearest-neighbor ions far outweigh the
the usual treatment, the splitting of the five d orbitals effects of more distant charges, subsequent workers
in an octahedral arrangement of negative charges into a applied the basic idea to coordination clusters in solu-
triply degenerate set, &,(d,,, d,,, d,,), and a doubly de- tion (B-6), that is, to central metal ions whose innermost
generate set, e,(d,..,~, d,3, as shown in Figure 2, is dis- coordination spheres contain specific arrangements of
cussed first, followed by an introduction of the concept ligands. These ideas have now developed into one of
of crystal field stabilization energy and a dis-
cussion of the inversion of the splitting pattern in
fields of tetrahedral geometry. A qualitative treat-
ment of the orbital splitting in fields of other sym-
metries, e.g., square planar, is also usually included.
Although the qualitative presentation of the crystal
field theory is effective and satisfying from the peda-
gogic point of view, a quantitative discussion of the rela-
tive d-orbital energies in fields of various symmetries
would be of even greater value, but unfortunately is too
This work was supported in part by the U S . Atomic Energy
Commission under Contract AT(l1-1)-256 (Document No.
COO-256-78). Presented in part before the Division of Inorganic
Chemistry at the 155th National Meeing of the American Chem-
ical Society, San Francisco, Calif., April, 1968, and before the
XI International Conference on Coordination Chemistry, Haifa,
Israel, September, 1968. Figure 2. Energy level diogroms showing ligand fleld splitling potterns
' Author to whom correspondenoe may be addressed. ford orbilalr in some common soordinotion environmenb.

Volume 46, Number 12, December 1969 / 799

 the principal theories of coordinat,iou complexes and are constituent arrangemenls, the potential of the larger
described in a numher of boolcs which treat the theory geometrical arrangement is the sum of t,he fields of its
and its applications from either a qualitative (7, 8) or a constituents. Thus, if the potential fields of a rela-
quantitat,ive (9-12) ~t~andpoint. tively small number of primary constituent groups of
Although the qualitat,ive effects of cryst,al or ligand specific geometry are known, the crystal fields of many
fields on the relative splittings of the d orbitals are larger, more complex arrangements can be easily ob-
easily and widely understood, the quantitative calcula- tained. For example, the potential field due to a regular
tion of ligand field effects is a matter of considerable octahedral arrangement of six ligands is precisely equal
mathematical complexity. The calculation involves to the sum of t,he fields arising from a square planar ar-
first setting up an expwssiou for the total crystal field rangement of four ligands in the X-Y plane and a linear
potential, V,.,., of suit,able symmetry stimulated a t the arrangement of two ligands along the Z axis, all equi-
central metal ion by a given geometrical disposition of distant from the central atom. Similarly, the po-
the ligends. The effect of this potential on the rl or- tential due to eight ligands a t the corners of a regular
bitals is then investigated and the problem takes the cube is equal to t,wice the field generat,ed by two tetra-
form known as the perturbation theory for degenerate hedral arrangements of four ligands each, in which each
systems. Using the expression for V,., as an operator, ligand lies a t the same distance from the central atom.
the matrix elements involving the d orbitals are eval- Although the relationships between the crystal field
uated. The resuking secular determinant for tthe potentials of simpler component sets of ligands and
pert,urbation energy is then solved in t,erms of the those of larger, more complex geometrical arrangements
radial parameters. The integrals involving the radial that can be obtained from them are obvious from an in-
part of the wave function, ( p , ( r ) ) , are not usually spection of the expressions for potential fields arising
solved, but instead are left as empirical paramet,ers. from various arr~ngement~s of the ligands (9), the
Their presence in the results does not detract from the implications of these relationships have not been
usefulness of the calculations of the relative splittings adequately realizcd. I n particular, to our knowledge
of the (1 orbitals due to various types of ligand arrange- it has not been pointed out previously that, subject
ments, but requires that. the results he expressed in to a few limitations arising from symmetry consider-
terms of a n empirically evaluated parameter such as ations, it is possible to obtain the Dq values of all the d
Dq, where Dq = (I/,) ( p , ( r ) ) , and where certain values orbitals in many complex configurations of ligands by a
for the ratio ( p , ( ~ . ) ) / ( p , ( r ) ) , expressed in terms of a mere summation of the Dq values of the correspond-
parameter p, are assumed. ing d orbital in each of several component sets of ligands.
The expression for the potential field a t the central The result is of import,ance to the practicing chemist;
metal ion due to a part,icular ligand geometry is ob- for if the Dq values of rl orhitals are known for a few
tained by summing the potentials due t,o the individual basic, primary component groups of ligands, the Dq
ligands making up the complex. This procedure im- values due to a new configuratiou derived from these
plies that, just as certain geometrical arrangements can components can be obtained a t once wit,hout having to
be considered t,o be const,ructcd from several smaller solve the fundamental quantum mechanical problem.

Table 1. Potential Field Expressions and Relative d-Orbital Energies far

Three Primary Geometric Configurations
Coordination number Relative energy of d orbital in units of Dp for p = 2.0
and configuration of (Values for p = 1.0 in parentheses)
primary group Potential field expressian"J d,, d,n., d=v d., a,=
I . C.N. 1 = v1 = *?*PW + x1~ z ( ~ ) +~ 23P4(r)z40]
1
0
5.14
(3.43)
-3.14
(-1.43)
-3.14
(-1.43)
0.57
(-0.285)
0.57
(-0.285)
One ligand or charge
located on the Z axis

11. C.N. = 2 -2.14 6.14 1.14 -2.57 -2.57

T~~ ]igands at right V9 = GET (-0.43) (4.43) (-0.57) (-1.71.5) (-1.715)
angles in the A'-Y
plane (on the X and
Y axes)
1
zPd(r)(: +
4%
z40 Tzr+O]

111. C . N . = 4 P (Dp values as a foncbion of 8 )

Four ligands in a
staggered, tetra-
hedml-like arranee-
ment, two shove &d
two below the X-Y
plane; each a t avari-
able angle 6 with re-
snect to the =tZaxes
I p 1 . 1 1i d x i . . . r I ., I h : .t i f I '1 . The V . I ~ ~ U I I1- ~ ~ 1 1 1 %P,
s ell.. R I V ~ ~ c r q lwil iri ~ ll~
I I I I I I i I A . I I .I . I~W~ 1111,I W
I>VP Cionvtiwa: i i 1 1 1 ~1.131 11 1 h 1 9 1 \ w r n l ~ r. x -
ort.rsit,lr i n h!trrt~<n.~c
III~I:~~IOI). L I + ui 1111-1.01R1 (,I, ~t.t.atl\.
" ia~.~lil.~~l .r C
\ OIIIV11 .~I e ~ ~ : ~ r i +emiW . ~ C W I I(I(
. C.I f. < r f u ~ ~ h it xv~ l 3 n n -
lion of the terms and symbols b e d .
"he negative sign preceding t,he last term for C.N. = 4 applies to orientation A in which the ligsnd projections in the X-Y plane lie
a t 45" between the ztX and i l ' a x e s ; the positive sign applies to orientation B in which the projections lie on t,he i X and +I' axes.
The Dp v a h m ai the d orbitals given in the table are for orientation A. In orientation B, the energies of the d,2-,2 and d,, orbitals are
interchanged, while the energies of the other orbitals are unaffected.

800 I journal of Chemical Education

 we have used the values for p = 2 in the examples in
this section. Some recent experimental results, on the
other hand, indicate that p = 1.0 may be a better choice
for this parameter in many cases (15, 14). We have
therefore included the numerical Dq values for both
values of p.

, 3 Additivity of the Potential Fields and

Figure 3. lllurtration d the three prirnwy ligond groups, ML, MLz, and
DqValuer of Primary Groups
MI4, showng ligand locations relative to the central metal ion and the
coordinate axes.
The potential fields and relat,ive d-orbital energies
given in Table 1 may be combined in a simple, direct
manner to give the expressions fgr potent.ial fields and
Illustrated in Figure 3 are three primary groups of d-orbit,al energies for larger, more complex arrangements.
ligands from which almost all the geometrical structures For example, a linear molecule with a coordination
possible for coordinat,ion complexes can he constructed. number of two, which can be considered to be oriented
I n Table 1 are given the expressions for the crystal field along the Z axis, would have a pote~it~ial field equal to
potentials generated a t the central metal ion by the 2V, and horbital energies corresponding to twice the
three primary component ligand geometries to- values shown for each d orbital in the top row of Tahle 1.
gether with the Dq values for each of the d orbitals in Similarly, a square planar arrangement of four ligands
these fields, assuming a dl electronic configuration and in the X-Y plane would creat,c a potential field a t the
two different values for the radial integral parameter p. central metal ion equal to t,wice that of a single right-
The combination of the component groups to give angle group, i.e., 2V2,and would give rise to relative (1-
larger, more complex geometries is then discussed and orbital energies equal to twice the values given for a
illustrated. single right angle primary group in the X-Y plane, i.e.,
Discussion
Primary Groups
The arrangement6 of ligands designated "primary
for 2 lignnds atIIOD
inX-Yplane 1 2.14
x2.14
6.14 1.14 -2.57
6.14 1.14 -2.57
-2.57
-2.57
Sum (for 4 ligands;
groups" in Table I represent the smallest arrays of aqnare planar) -428 12.28 2 . 2 8 - 5 . 1 4 -5.14
charges whose potential field expressions contain only
certaiu second and fourth order harmonic terms and The sum of the potential fields (ind t,he d-orbital
which possess t,he property of direct additivity of fields energies) of two right angle groups in the X-Y plane
to produce new fields which also contain only similar corresponds to the field produced by superimposing the
harmonic terms. A primary group is needed for ligands two groups to produce double charges a t an angle of 90'.
occupying axial positions, for ligands in planar positions Because of the presence of a center of symmetry in
and for ligands not occupying axial or planar positions. cach d orbital, the effect of a doublc charge located on
To encompass all the structural arrangements common one of the two positive or one of the two negative lobes
in inorganic complexes, only three primary confignra- of a d orbital is t,he same as the effect of two single
tions are necessary. These are illustrated in Figurc 3 charges separated a t 180" and located one each on
and are (1) one charge located along the Z axis, (2) t.wo opposite positive or negative lobes. The summation is
charges forming a right angle in the X-Y plane, e.g., therefore valid for a square planar arrangement of four
placed on t,he + X and +Y axes, and (3) four charges ligands and leads to the proper splitting pattern of the
(2 orbitals (Fig. 2).
located a t pseudo-tetrahedral positions. The last
group applies t o charges not placed on the Z axis Larger and more complicated structures can he
or in the X-Y plane, but a t angles between 0" and 90' built up in a n analogous way. For example, the
with respect to the Z axis or to the X-Y plane. I t con- familiar octahedral army of charges can be formed
sists of two ligands above the X-Y plane, both a t an by combining two groups along the Z axis with two
angle 10 with respect to the Z axis, and another similar sets of right angle groups in the X-Y plane, i.e., by add-
pair of ligands instaggered positions below theX-Y plane ing the two groups along the Z axis to a square planar
making an angle a-8 with the +Z axis. N o k that if structure. I n this case V,,, = 2V1 2V2 and t.hc +
B = 54"44', which is one half a tetrahedral angle, the relative d-orbital energies would be summed as follows
third primary group becomes a regular tetrahedron. &orhila1 energies d,* d,a-,a dzv d,, d,,
Forother values of 10,this group will be a distorted tetra- for 1 ligand on 3.14 -3.14 0.57 0.57
hedron. The most common coordination configurations Z axis 14 -3.14 -3.14 0.57 0.57
can all he formed from only the first two primary 6.14 1 . 1 4 - 2 . 5 7 -2.57
90' in X-Y -2.11 6.14 1 . 1 4 -2.fi7 - 2 . 5 7
groups, but the third is needed to form structures re-
lated to tetrahedra, such as a cube, an antiprism, and
an icosahedron. Sum (far6ligands; 6.00 6.00 -4.00 -4.00 -4.00
octahedron)
The numerical values for the relativc energies of the
d orbitals given in Tahle 1 are calculated for two differ- The values obtained by the summation correspond
ent values of the parameter p, i.e., for p = 2.0 and p = to the known Dq values for the triply degenerate h, and
1.0. Up to the present time, most discussions of d- the doubly degenerate e, sets of orbitals in a regular
orbital energies in crystal fields have assumed that p = octahedral configuration of ligands. If only one ligand
2.0 (see, for example, reference (S)), and for this reason on the Z axis were used instead of the two in t,he above

Volume 46, Number 12, December 1969 / 801

 Table 2. Symmetry Representations of the Real 3d Orbitals as o
Function of Group Symmetry"
Symmetry
6r point^
Coord. group Location Symmetry representation for orbital'
number deswnatian Configuration (example) of ligands d*, d d,.
Linear +Z m+(a,) s(ed Xed
fZ + 8. 8.
+X, + Y a, bt b2
X-Y plane as e' e'
Riaht an& + x , + Y , +z
Trhngul& pyramid
Tetrahedron
Distorted tetrahedron
Plane square X-Y plane
Plane pentagon X-Y plane
Square pyramid
Trigonal bipyramid
Octahedron
Pentagonal pyramid
Plane hexagon X-Y plxne
Triangular prism
Pentaeonal biovramid
~ e x a & n a lpy&nid
Plane heptagon X-Y plane
Cube
Antiprism (twisted cube)
Hexagonal biovrmid
X-Y plane

.
.
12 I*
We have used the Schoenfliess symbols to designate the various symmetry groups or point groups. For an explanation of these sym-
bols see the books by F. A. Cotton (16) or by H. Jaffee and M. Orchin (I@).
The ~vmmetrvrenreaent,at,ionaf"r the various d orbitals fallow the notation orooosed bv R. S. Mulliken. These svmbols are also ex-

a: The orbital (actually, the wave function of the orbital) is symmetric about the principal rotation axis, i.e., does not change sign
upon rotation of 2w/n about the n-fold rotation axis of the group.
b: The orbital is antisymmetric-changes sign upon n-fold rotstion about principal axis.
e: A set of two degenerate orbitals (one transforms into the other merely by a rotation or reflection allowed by the overall group sym-
metry).
1: A set of three degenerate orbitals.
g: A set of four degenerate orbitals.
h: A set of five degenerate orbitsls.
(For s. linear molecule with a C, rotation axis, the symbols c,T,a,+, etc., replace a, el,, eh ex, etc. All these states are doubly degenerate
except o,which is non-degenerate, and correspond in symmetry to the s, p, d , f,. . .orbltsls, respectively. There is no symmetry property
of linear molecules related to b. The superscript + or - signs in the case of linear molecules indicate the even or odd character, re-
spectively, of the representation with respect to c, reflections.)
Superscript primes, ' and ",are used to indicate if the orbital is even or odd, respectively, upon reflection through a horizontal pplne of
symmetry perpendicular to the principal axis (greater than 2-fold) in geometries without a center of symmetry. (Primes are used for
example with D,o groups when n is add.)
Subscript g (gerade): the structure possesses a eenter of symmetry or of inversion and the orbital does not change sign upon inversion.
(The d orbitals are always gerade if a center of inversion is present.)
Subscript u (ungerade): the orbital changes sign upon inversion through the inversion center.
Subscripts 1 and B are used with the symbols a and b to denote whether the orbital is even or odd (does not or does qhange gign),
respectively, with respect to a two-fold rotation axis perpendicular to the principal axis, or with respect to a vertlcal plane if that 1s the
only symmedry plane present.
Subscripts 1, 2, 3 etc., are used with the symbol e to distinguish among possible degenerate pairs; specifically, if the wave functions of
the orbitals transform ss paws (cos 4, sin $1, {cos26, sin 261, (eos 39, sin 341, respectively, upon n-fold rotation about the principal axis.
Subscripts I and B are attached to t to indicate if the orbitals arc even or odd upon rotation or reflection about the various axes of
regular polyhedra.

example, the sums would have corresponded to the Dq or ligand fields arising from various geometric config-
values for a square pyramid of coordination number five. urations is presented in Table 2. A brief explanation of
the symmetry symbols used is made in the footnote to
Symmetry Considerations
the table.
The direct combination of Dq values of the primary Included in Table 2 are symmetry groups and repre-
groups listed in Table 1 may not always give Dq values sentations with respect to the Z axis as the principal
for all the d orbitals which agree with the values ob- rotation axis (axis of highest symmetry). For these
tained by solution of the quantum mechanical problem groups the symmetry representations of the individual
for the overall geometrical configuration of interest. d orbitals in the character table of that symmetry
I n such cases symmetry differences exist between the group indicate the degeneracies (in symmetry and i n
structure actually formed by direct combination of the energy) among the set of real d orbitals depicted in
primary groups and the structure of equivalent co- Figure 1. This correspondence arises because the set of
ordination number for which Dq values are desired. If real d orbitals used here is tetragonally oriented, i.e.,
these symmetry differences are taken into account, the it is chosen so that the unique d orbital (d,. in this case)
results can be made concordant. A listing of the sym- falls along the Z or fourfold axis.%
metry designations of each of the five d orbitals in ionic The total electrostatic effect of any array of ions or

802 / lournol o f Chemicol Education

ligands lying along a particular axis or in a given plane The desired planar hexagonal arrangement is shown as
on all the d orhitals of the central metal ion is inde- array B.
pendent of the geometrical arrangement of the ions or For array A the combination of the three right angle
ligands in that array. Nevertheless, to make the groups in the manner indicated has t,he following effects
effect of the total field applicnblc to a particular ge- on the energies of the d orbitals
ometry or orientation of ligands, it, may be necessary to
divide the effect of the total field in a manner consistent
with the symmetry properties of the d orbitals in the in S ~ Yplane
geometric configurat,ion of interest and to make d (mi~ltiolvbv three) X3 X3 X3 X3 X3
orbitals of like symmetry degenerate, i.e., of identical for 6 ligands s t 00' in
energy. This is done simply by averaging the total
effect of the field over the degenerate orhitals, which
X-Y plane -6.42
-
18.42 3.42 -7.71 -7.71

amounts to combining and averaging the Dq values of

the individual orbitals of the degenerate set. The six charges in the X - Y plane arranged as shown
As an example, consider the formation of a planar in array A have the same effect on the energies (Dq
hexagonal array of ligands from three primary right values) of the orhitals not in the X-Y plane (d,,, (l,,, d,,)
angle groups of t.wo ligands each in the X-Y plane. I t as the hexagonal planar arrangement, array B. The
is known from prior solution of the quantum mechanical orientation of the chnrgcs in array A affects the relative
problem (8) that the relative energies of the d energies of the d,j-,. and d,, orhitals, which have differ-
orbitals in a potential field produced by six ligands ent orientations in the X-Y plane. The total effect of
arranged in the form of a plane hexagon are (in Dq units the six charges on the energies of both the d+,. and
for p = 2)&, -6.42; rl,>-,., 10.92; d,,, 10.92; d,,, -7.70; clCyorbitals in array A is (in Dq units) 18.42 3.43 = +
and d,,, -7.70. Reference to Table 2 and to the relative 21.54, which i~ cqual to the sum of the Dq values for
energy values of the orbitals shows that the d , a ~ , and
~
d,, orbitals are degenerate (e2,) and that the ds, and dr,
orbit.als also form a degenerate pair (el,) in a planar
Far configurations in which the % axis is not the principal
hexagonal field. Addit,ion of three primary right angle axis of rutation, the usual symmetry rropresentations will not, be
groups consisting of two ligands each, all in the X-Y valid for t,he set of real d orbitals used here. For example, the
plane, amounts to t,he superposition of t,hese groups to primary gronp M X . which lies in the XY plnrre is usually con-
give the configuration shown as array A in Figure 4. sidered lo possess C3, symmetry, hut not with respect to the Z
axis as the axis of rotation. Actually, the relat,ive energies and
degeneracies of the d orbirds in the field of the MXl primary
group are the same as those far the square planar Drh group.
This equality arises because the d orbitals have a center of sym-
metry and any group may he reflecled lhrough the origin (located
a t the meLd nlom) without nlterirrg the relalive energies of the d
orhitals. Thns as far as the d orbitals are ~:oncerned,the effect of
any ligand group will be the same as that of s larger group con-
sisting of all t,he original ligands plus all additional ligand points
generated by the inversion through the otigin. Accordingly, it,
can be confirmed from Tables 1, 2, and 3, for example, that the
splitting patterns of the d ot.bitals are identical in the cases of
Figure 4. Formation of a planar hexagonal array d ligmds from three (a) either one or t,wo lipands lying on the Z axis, (b) the M X ,
primary right angle group. of the two ligands eoch in the X-Y plane. primary group and the sqrlsre planar ( D d group, ( c ) the plannr
A, superposition of the three groups by addition. 8, the desired hexagonal MXI group (Dab) and the planar M X s gt.oup (De), and (d) the
arrangement. tclrahedmn and the cube.

Volume 46, Number 12, December 1969 / 803

these orbitals in the hexagonal planar field. To make tained simply by adding the Dq valucs for one group
the sum of the d-orbital energies for the three primary located along the Z axis (first primary group). Sim-
groups in array A applicable to the planar hexagonal ilarly, the Dqvalues for all the polygonal bipyramids are
arrangement in array B, it is necessary only to average obtained by adding twice the Dq values for a single
the total effectof the field on the d+,s and d,, orbitals. ligand on the Z axis to the values for the plane polygon.
This serves to make them a degenerate pair, consistent The potential field expressions for the polygon-based
with the symmetry properties demanded by a plane pyramids and the polygonal bipyramids arc obtained
hexagonal field. The average energies are (19.42 + by analogous combinations.
3.42)/2 = 10.92 for each orbital, which are precisely All of the geometric configurations of ligands dis-
the values noted above for these orbitals in a hexagonal cussed above consist of ligands lying in the X-Y plane
field. The use of orbital symmetries in calculating Dq and/or along the Z axis and can be built from only the
values for various ligand fields is illustrated further in first two primary groups listed in Table 1. Other struc-
the examples that follow. tures exist in coordination compounds, however, in
which some of the ligands are not axial or planar and
Dq Values in Ligond Fields of Vorious Geomefries which cannot be constructed with only the first two
Application of the procedures discussed above primary groups. Examples of such structures include
enables one to calculate with ease the relative the tetrahedron, the cube, the various prisms and anti-
energies of d orbitals in ligand fields representing a prisms and the icosahedron. I n order to calculate the
variety of geometric configurations. For example, the ligand field potentials and the relative energies of the
Dq values of d orbitals in a planar hexagonal field cal- d orbitals in fields resulting from these configurations,
culated above can be used in the calculation of the an additional primary group must he used.
energies in a planar trigonal field of coordination number The third primary group is pictured in Figure 3 and
three. Because the degeneracies of the d orbitals are consists of four ligands in a tetrahedron-like arrange-
similar in planar hexagonal and trigonal fields, the Dq ment, two of which are above the X-Y plane and two
values in the latter field are simply one-half their magni- of which are below the X-Y plane in staggered positions.
tudes in the hexagonal field. (Two orientations, A and B, are defined for this
primary group. See footnote to Table 1.) All four
d-orbilal energies ligands are equidistant from the central atom and form
( P = 2) d,l d d*, d., d, angles of either 9, or a-8with the +Z (polar) axis. Be-
for 6 ligmds (plane
hexagon) -6.42 10.92 10.92 -7.70 -7.70
cause the relative energies of the various d orbitals in
symmetry in hex- Dq units vary with the polar angle p, a value of the
agonal field (Deb) a,. eso es. el. el. angle that the ligands make with the Z axis must be
symmetry in trig- specified before the third primary group can be used.
onal field ( D s ~ ) a, 0' c' eV er I n a tetrahedron, cube or square antiprism, for example,
for 3 ligands (plane
triangle) + 2 -3.21 5.46 5.46 -3.85 -3.85 p equals one half the regular tetrahedral angle, or
54"44'. In a trigonal prism, p may equal 45', while
The Dq values of the d orbitals in a planar pentagonal in an icosahedron or in a pentagonal antiprism formed
field can he calculated from at least three differentcom- by removing the two apical groups from a regular
binations: (1) adding the Dq values for a trigonal icosahedron, ,9 equals GZ028'. (Although some of these
plane group of coordination numher three (derived angles appear strangc when expressed in degrees, the
above) to the Dq values for the primary right angle cosines of these particular angles are t h e reciprocal
group of coordination number two; (2) adding the Dq square roots of small integers, i.e., 1 / 4 2 (4s0), 1 / 4 3
values for four primary right angle groups, each of (54'44'), 1/2 (GO0) and 1/45 (GZ02S').)
C.N. = 2, and subtracting from this sum the Dq values The addition of this third primary group enables the
for the trigonal plane; and (3) multiplying the Dq values relative d-orbital energies to be calculated for most of
for a primary right angle group by 5/2 to take into remaining structures possible for coordination com-
account the increase in the number of ligands from two pounds. For example, taking p to be equal to 54"44',
to five. I n each case the symmetry properties of the d which is one half the regular tetrahedral angle, one can
orbitals in the planar pentagonal field (D,,) must he calculate the Dq values for both a cube and a square
considered in determining the final values for each antiprism (a twisted, staggered cube), both of which
orbital. We illustrate below the third possibility. have a coordination number of 8, from those for two
d-orbital mergies d*z d,%-,n dzv d,, dw regular tetrahedra.
for 2 ligmds s t 90°, d-orbital energies da
X-Y plane -2.14 6.14 1.14 -2.57 -2.57
for a. reg. tetrahedron,
(multiply by 512) X5/2 X512 X5/2 X512 X5/2
ML4 -2.67
for 5 ligands in (3rd primary group,
X-Y plane -5.35
- -
15.35 2.85
sum = 18.20
-6.42 -6.42 6 = 54" 44')
(multiply by 2) XZ
symmetry in Dsn far 8 ligands, M L -5.34
field a,) PZ' ea' el" elu symmetry in tetra-
for 5 ligands a s plane hedral Td field E
pentagon -5.35 9.10 9.10 -6.42 -6.42 symmetry in cubic Oh
field eg
Using the values of the relative d-orbital energies in symmetry in anti-
the trigonal, square, pentagonal, and hexagonal planar prism's D4dfield a,
fields, all of which are calculated above, the Dq values for 8 ligands in anti-
for the respective polygon-based pyramids can be oh- prism -5.34

804 / Jovrnol of Chemicol Education

It is interesting to note that when four of the eight addition of the D q values corresponding to two groups
ligands of the cube are rotated 45" to producc the anti- added on the Z axis gives 0.00 for all the d orbitals,
prism, the energies of the d,., d,, and d,, orbitals are which is known to be their value in the field of a reg-
unchanged, while the d,,-,, and d,, orbitals become ular icosahedral array of ligands.
degenerate and equal to a meanvalue (-5.34 3.56)/2+ The Dqvalues of rl orbitals in a variety of ligand fields,
= -0.89 D q units. all of which can be calculated from the D q values of the
The over-all effect of a potential field due to a six- three primary groups given in Table 1, are presented in
coordinate trigonal prismatic structure is equal to a/z Table 3. I n each case the D q values of the d orbitals
the effect of the potential field of the four-coordinate for each primary component group are added algebraic-
third primary group, chosen with an appropriate value ally as the primary groups are added or removed in
of 0 . The energies of the individual d orbitals may building-block fashion to produce the desired over-all
then be calcuated with reference to their symmetries. structure. The summed values are then averaged over
A particularly interesting pair of related structures is the degenerate orbitals of the particular structure to
the icosahedron (ML12) and the pentagonal antiprism give the final D q values. The D q values for all struc-
IML,,)
~ --.which can be derived from it by removal of the tures are calculated for both p = 1 and p = 2 .
two apical groups. It is known that the d orbitals ex-
hibit five-fold (h,) degeneracy in an icosahedral field, Dq Volues for Substituted Complexes, M X i Y i . . .
which therefore acts as if it were spherically symmetrical The calculation of D q values described above applies
with respect to t,he d orbitals. Because the orhit,als are to complexes containing only one type of ligand, though
not split, their relative energies are all the same and equal of diiering coordination number and geometrical con-
to 0.00 D q units. Removal of the two apical groups, figuration. The approach can, in addition, be extended
which may be considered to lie on the Z axis, leaves the to most substituted complexes containing two and even
pentagonal antiprism (0 = 6 2 " 2 S 1 ) . Subtraction of three or four different types of ligands, pvovided only
the D q values associated with the two apical groups that each type of ligand i s added in units covvespomling to
(twice the D q values of t,he first primary group of Table one or move p ~ i m a v yg~.oups, or in units deviuable fvom
1 ) from the D q values of thc icosahedral field gives the them.
D q values of the pentagonal antiprism. In the examples below, the valuc p = 1 is used be-
~~. cause of incrcasing evidence that this value for t,his
d-orbilal
rnzrgics h* d d,, d,, dvs parameter gives better agrecmcnt with experimental
for 12 ligands
results than does p = 2 .
(icosahedron) 0.00 0.00 0.00 0.00 0.00 A{X6Y. Consider for example the mono-substituted
less 2 ligands an six-coordinate complex, M X S Y . This complex can hc
saxis ( p = 2 ) -10.28 +6.28 +6.28 -1.14 -1.14 considered to be composed of two sets of right-angle
for 10 ligands primary groups, M X 2 in the X-Y plane, plus one S
(pentxg. anti- ligand and one Y ligand added along the Z axis. In this
prism) -10.28 +6 28 +6 28 - 1 14 - 1 14 case, it is assumed that ligands X and Y may have differ-
symmetry in D5a ent ligand field strengths which must be taken into
field asl el. CZS em %
account. Alternatively, we can conceive of t,he com-
The relative energies of the d orbitals in the field due to plex MX6Y being formed by removing one of t,he two
a pentagonal antiprism are idcntical to those of a linear X ligands that lie along thc Z axis of a regular octa-
complex of coordination number two, ML2, except that hedral complex, M X 6 , and then adding back one Y
the signs are reversed and the order inverted. The ligand along the Z axis. The two alternative ap-
field of the pentagonal antiprism with 0 = 6 2 " 2 S f may proaches give the same D q values for all the d orbitals.
therefore he characterized as one of (inverted) linear The combinations of the D q values of thc orbitals for
symmetry. Making use of the "electron-hole formal- the latter approach are shown bclow.

d-orbital energies ( p = 1) dss dz*q d," dzs dw

for a regular octahedron, MXn 6 . 0 0 Dqx 6 . 0 0 Dqx - 4 . 0 0 Dqx -4.00Dqx -4.00Dqx
loss Dq far one X group ( Z axis) -3.43Dqx +1.43Dqx + I 4 3 Dqx + O 28.5 Dqx + O 283 Dqx
plus Dp for one Y group ( Z axis) 3.43 Dqu - 1 . 4 3 Dqy -1.43Dqy -0.283 Dqr -0.285 Dqu
Algebraic sum ( f o rM X s Y ) 2.57 Dqx 7 . 4 3 Dqx -2.57 Dqx -3.716 Dqx -3.715 Dqx
+3.43 D ~ Y -1.43Dqu -1.43Dqu -0.285 Dqy - 0 . 2 8 5 Dqy
Symmetry ( C d a, bl bs e e

ism," it is predicted that the splitting pattern of a dl ion A schematic energy-level diagram showing the rela-
in a linear complex, ML2,will be identical to that of a do tive energies of the various d orbitals is given in loigure
ion in a pentagonal antiprismatic ficld. 5 for the parent octahedral complex, M X s , and for the
The relative energies of the d orbitals in the pen- mono-substituted complex M X J Y for the cases in which
tagonal antiprismatic and icosahedral structures could D q , > D q x and D q x > D q , ( D q , = 1.25 D q x and D q y
also have been calculated from those of thc first and = 0.75 D q x ) . The splitting of both the e, and the t,,
third primary groups. The total ligand field effect of levels of the parent octahedral complex depends on the
the ten ligands in the pentragonal antiprism is equal to difference in ligand field strengths of ligands X and Y .
6/2 the effect of the four ligands in the third primary I t is significant that the energy (10 D q x ) of the
group with 0 = 62'28'. After adjustment of the D q principal d-d transition of the regular octahedral com-
values to make them consistent with DSd symmetry, plex, MX6, is retained by one of the transitions from the

Volume 4 6 , Number 7 2, December 7 9 6 9 / 805

lowest energy level (ground state for d' configuration) of
the substituted complex M X S Y if the ligand field
strength (proportional to D q ) of the substituting Y
ligand is greater than that of the X ligand bcing replaced
from the parent complex. For this case, the three d-d
transitions from the d1 ground state in order of in-

(d,,
(d.,
(d,.
--
creasing energy are
d,,,d,.) v~ = 1.145 ( D m
d , ~ ~ ;YI ) = 10 Dqx
-t d,a) us = 10 Dqx +
- Dqx)
4.86 (Dqy - Dqx)
Usually, the difference in ligand field strengths, DqY
- Dq,, will be much smaller than 10 Dqx, so that the
energy of the first transition will be small and this band
will lie in the near-infrared region of the spectrum.
The second and third transitions, which usually lie in Figure 5. Energy level diagromr showing further splitting of the prrrent
the visible region of the spectrum, will be separated in octahedral levels upon rvbrtitvtion of one ligand in o sir-coordinate som-
plex. Energy levels ore drawn os.uming p = 1 ond illustrate splitting
energy by an amount related to the difference in ligand far a 25% difference in ligand neid strengths, i.e., Dqy = 1.25 Dqx
field strengths of the two types of ligands, i.e., by ond Dqy = 0.75 Dqx. Thevertical energy scale ir given in "nib of Dqx.
4.86 (Dqy - Dqx). If Dqy = 1.25 Dqx, for example,
the frequencies of the three bands expressed in terms Trans-MX4Y2. The trans-disubstituted complex
of Dpx will be vl = 0.29 Dgx, vz = 10 Dqx and v3 = 11.22 M X 4 Y 2may be considered to be built up from two right-
Dqx. These latter values correspond to the transitions angle M X z groups in the X-Y plane and two Y ligands
illustrated in Figure 5. added along the Z axis or, alternatively, by removing the
If the minimum difference in frequency for two transi- two X ligands which lie along the Z axis of the octa-
tions to show observable splitting is assumed to be 2500 hedral complex M X 6 and then replacing them with two
cm-1 for solution spectra, it can be shown that the Y ligands. For the latter procedure the D q values are
minimum difference in the ligand
- field uarameters, 10 combined as follows
d-orbital energies (p = 1) dd d,.-,* dw d., dv
for a regular octahedron, MX6 6.OODqx 6.00Dqx - 4 . 0 0 Dqx -4.OODqx -4.00Dqx
less Dq for two X groups ( Z axis) -6.86 Dyx +2.86Dqx +2.86 Dqx +0.57 Dqx +0.57 Dqx
plus Dq for two Y groups (Z axis) +6.86 Dqy -2.86Dqy - 2 . 8 6 Dqy -0.57 &Y -0.57 Dqy
Algebrilic Sum (far trans-MX,Y*) -0.86 Dqx 8 86Dqx -1.14Dqx -3.43Dqx -3.43 Dqx
+ 6 . 8 6 Dqu - 2 . 8 6 Dqy -2.86 Dqy -0.57 Dqy - 0 . 5 7 Dqy
Symmetry ( D d at. bl. bw e, e,

Dqx and 10 Dq,, for a monosubstituted d1 complex I t can be seen from these orbital energies that, if
> Dqr) to exhibit split bands is approximately D q Y > DYX,

-
(Dq, the trans-disubstituted complex of a metal
5000 cm-l. For smaller differences in the values of thc ion with a d1 electronic configuration will again have one
ligand field parameters, the d-d absorption band in the component of a spin allowed band (d,, d+,>) at the
visible region will not be distinctly split, but will be same frequency as the spin-allowed d-d band of its
broadened (in comparison to the d-d band of the parent parent octahedral complex M X 6 , i.e., at 10 Dqx. The
octahedral complex). The two components of the band lowest energy transition (26,) should have just twice the
will have a mean frequency of 7.57 Dqx +
2.43 D q y and, frequency of the lowest energy band of the mono-
if they are of equal intensity, the band center will have substituted complex (6,) and the splitting (energy differ-
shifted in frequency by an amount corresponding to ence) between the two higher energy transitions of the
25% of the difference in frequency between the bands of trans-disubstituted complex (26,) should be exactly
the two octahedral parent complexes M X 6 and M Y B . twice the splitting of the mono-substituted complex
If Dqx > Dqy, the lowest energy level becomes doubly (&), as might he expected intuitively. The energy
degenerate and none of the transitions from this ground levels for the trans-complex are shown in Figure 6.
state retains the energy of the d-d transition of the 1,2,3-MX3Y3, The 1,2,3- or cis-trisubstituted com-
parent complex M X B (see Fig. .5). The extent of the plex may he considered to be composed of one right-
splitting of the band in the visible region will be the angle Am2group and one right angle MY2 group, both
same as in the case that Dqy > Dqx, but now the mean in the X-Y plane, together with one X ligand and one Y
energy of the two components is 8.72 D q x +
1.00 Dq,, ligand along the Z axis. The D q values for the d
which represents a slight decrease in energy from the orbitals for such a combination of primary groups are
transition in the parent complex that occurs at 10 as follows
Dpx.

&orbital mergies ( p = 1) d.8 dz%-v, d=r d,, dv

far two X ligmds at 90' in X-Y plane -0.43 Dqx 4 . 4 3 Dqx -0.57 Dqx -1.715 Dqx -1.715Dqx
for one X ligand on Z axis 3.43Dqx -1.43%~ -1.43Dqx -0285Dq.x -0.285Dyx
for two Y ligands at 90' in X-Y plane -0.43 Dqy 4 43 Dqy -0.57 Dqy -1715Dqu -1.715Dqy
for one Y ligand on %: axis 3.43Dqu -1.43Dqr -1.43Dqy -0.285Dqu -0285Dqy
Algebraic sum of 1,2,3-MXzY8 3 . 0 0 Dqx 3.00 Dqx -2.OODqx - 2 . 0 0 Dqx -2.00 Dqx
+3.00Dq~ +3.00Dqy -2.OODqy - 2 . 0 0 Dqy -2.00 Dqy

806 / Journol of Chemicol Educofion

Figure 6. Ligmd Geld rplilting pattern* predicted for ihe series of sir.coordinde substituted eamplexer between MXI and MYa. The diagram assume*
d l electronic configuration, p = 1 and Dqu = 1.25 D q x ; the vertical scale i l drown in "nib of Dqx; referr to splitting of parent e, level, and 61 to
gplitting of parent h, level upon single substitution.

The energy level diagram predicted for 1,2,3-MXaYs MXsY is related to its octahedral parent MXs. That
is included in Figure 6. On the basis of its spectrnm is, if a ligand X of higher field strength is substituted on
and energy levels, this particular complex can be said to cis-MX3Yato give cis-MXIY,, the spectral relations and
possess "pseudo-cubic" symmetry. The energy of each splitting patterns are the same as the case in which a
d orbital is the same as it would he if the complex con- ligand of higher field stength (Y) substitutes on MX6 to
tained six identical ligands of the average field strength, give MXaY. In both cases, one of the transitions in
i.e., of a field strength equal to the average of the field the visible region remains a t the energy of the single
strengths of ligands X and Y: Dqave = (Dqx D~Y). + d-d transition of the parent "cubic" complex while the
I n terms of Dq,,, the energy of the degenerate pair of e other is displaced toward higher energy and is separated
orbitals is 6.00 Dq,,., while the energy of degenerate tz in energy from the 10 D q transition by 4.86 ( D ~ Y-
set is -4.00 Dq,,,. The d-d transition for a d 1 complex D q x ) . In case DqY > D y x , the complex cis-MX2Y4
of the formula 1,2,3-MXaY8is therefore not split in the behaves as a singly-substituted derivative of the pseudo-
trigonal (C3J ligand field and its energy is equal to 10 cubic complex MX,Y,. This is the case illustrated in
Dq.,.(= 5 Dqx + 5 D q u ) . I n both these character- Figure 6.
istics, the complex MX3Y3resembles a regular octahedral The fact that the splitting of d-d spectral bands in
complex (17). cis-disubstituted complexes is one-half that expected
Cis-MX,Y,. The cis-disubstituted complex MX4Y2 for trans-disubstituted complexes is well known and has
(as well as cis-M&YJ is closely related to the cis- been used as one of the criteria for distinguishing be-
trisubstituted complex, as explained below. I t can be tween cis- and trans-isomers on the basis of their spectra
considered to be made up of one primary right angle (17,lS).
MY2 group and one primary iMXz groups, both in the Cis-MX2Y4, Trans-MXzY4, MXY5. The energy
X-Y plane, and two X ligands along the Z axis. The levels for the more highly substituted complexes can be
energies of the d orbitals for this combination of ligands derived in an' analogous manner and show splitting
are summed below and the corresponding energy level patterns similar to those in the examples cited above.
diagram is included in Figure 6 for the case D ~ Y> Energy level diagrams for the series of substituted
DUX.
A~~

&orbital energies dd d&v2 dzv d,, duz

for M X , (90') in X-Y plane - 0 . 4 3 Dqx 4.43Dpx -057Dpx -1.715Dqx -1.715Dpx
for two X ligands on Z axis 6 . 8 6 Dyx - 2 . 8 6 Dpx -2.86 Dqx -0.57 Dpx -0.57 Dqx
for M Y . (90")in X-Y plane - 0 . 4 3 Dqu 4.43Dpy -0.57Dpu - 1 7 1 5 Dyy -171.5 Dqy
6 . 4 3 Dpx 1.57Dpx -343Dqx -2.285Dyx -2.285Dqx
Sum (for cis-MXdYa) - 0 . 4 3 Dqy + 4 4 3 Dqu -0.57 Dqu -1.715Dqy -1.715Dpy

Examination of the energy levels shows that the octahedral complexes between MX6and MY6 are shown
splitting of both the lower and the upper sets of energy in Figure 6. The levels are drawn to illustrate the
levels is the same as it is in the case of the mono- specific case in which D ~ Y= 1.25 D q x . The vertical
substituted complex, MXsY. I n fact, if D q x > D q u , scale is given in units of D y x . The only substituted
cis-MX4Yzis related to the pseudo-cubic trisuhstituted complex missing from the series is trans-MX~YI,to
complex 1,2,3-MX3Y3 in exactly the same way as which this approach cannot be applied because this
Volume 46, Number 12, December 1969 / 807
 isomer cannot be constructed from the primary ligand d1 system would be the same as those shown, but the
groups. diagrams for the d9system would be inverted.)
The value of the Dq parameter is usually not pre-
cisely known, though tables giving approximate values
have been published (12). Dq values vary with the
The prediction of the energy level diagrams for d metal ion, the ligand, the valence state, the coordination
orbitals by the procedures described in this paper is number, the type of structure, and the metal-ligand dis-
useful in understanding and interpreting the d-d spec- tance. I t is usually preferable to evaluate Dq by fitting
tral transitions of d1 and dycomplexes in the visible and one or more observed transitions to a calculated ex-
ultraviolet spectral regions, as well as those of d4 and CP pression and then to use this value to calculate the re-
complexes in their high spm (weak field) configurations. maining transitions. For the case being considered, Dq
As examples, quantitative comparisons will be made can be calculated by equating the expressions for the
of calculated and observed frequencies of d9 copper highest energy transition, i.e.,
chloride complexes with two different structures.
The crystal structure, spectrum and magnetic proper-
ties of CszCuC14have been investigated by Ferguson (19)
and others (13, 14, 20). The CuClp2- portion of the Using this value for Dq the other transitions are cal-
complex has a distorted (flattened) tetrahedral struc- culated
ture (D2J in which the average value of tho polar
angle 6 is about 62'. Spectral transitions are ob-
served at 9050 cm-', 7900 cm-I and -5200 (4800,5550)
em-'. For these complexes, the value p = 1 appears The agreement of the calculated and experimental
to be the better choice for this parameter. values is good in this example.
From Table 1, the relative energies of the d orbitals As a second example, the spectral transitions pre-
in Dq units in the field of the four-coordinate third dicted for the five-coordinate C u C I P ion will he cal-
primary group at B = 54'44' (tetrahedron) and at 13 = culatcd using the value of Dq evaluated above. I n the
62"28' (compressed tetrahedron) are as shown in Figure crystal [Co(NH,),] [CuCl,] the chloride ions are known
7, assuming p = 1. The distortion present in the to lie in a slightly distorted trigonal bipyramidal ar-
rangement around the cupric ion and the d-d spectral
transitions are observed at 8200 and 10,400 cm-' (IS,
1 4 ) According to Table 3, the relative energies of the
d orbitals in Dq units in increasing order are -3.14
(d,, and d,,), 0.035 (d,s-,n and d,,) and 6.21 (d,.) for a
regular trigonal bipyramid and for p = 1 (19). Placing
the nine d electrons in these orbitals leaves one un-
paired electron and the "electron hole" in the d,%orbital
(ground state). I n the states of higher energy, the
electron hole first jumps to the degenerate d,*-,. and
d,, orbitals and then to the degenerate d,, and d,, set.
The energies of these transitions are seen to be 6.175 Dq
(YI) and 9.35 Dq (vr). Using the value 1280 cm-' for
Dq, we calculated vl = 6.175 X 1280 = 7900 cm-' and
vz = 9.35 X 1280 = 12,000 cm-l. Although one of the
calculated values is too high and the other too low,
Figure 7. Energy level diogrom of the d orbitals in Dq vnih in the Reld they agree reasonably well with the observed values,
of fie four-coordinate third primary group in o tetrahedroi ond o com-
preaed tetmhedral field. especially in view of the distortion of the structure and
the change in coordination number between the two
examples, which may change the metal-ligand bond
CuClrz- structure causes appreciable splitting of both distances.
the 1, and e levels predicted for the regular tetrahedral The fit between the calculated and observed energies
structure. Because all the orbitals except the upper- of the transitions can be improved if the changes in
most are occupied in a dy complex, the transition of metal-ligand bond distances and the distortion from a
lowest energy that can occur is one in which an electron regular trigonal bipyramidal configuration are both
from either the d,, or d,, orbitals jumps up to the d,, taken into account. A recent study of the crystal
level, causing the vacancy or "electron hole" to move structure of [ C O ( N H ~ ) ~ ] [ C U(21)
C ~ ~has
] verified the
down in the opposite direction. I n the transition of presence of appreciable axial compression in this com-
higher energy, the hole would jump from its ground plex in that the three equatori$l chloride ions have a
(d,,) level to the &2-u2 orbital, while in the highest Cu-Cl bond distance of 2.39 A, while the two axis1
energy transition, the hole would move from its ground chloride ions have a shorter bond distance of 2.30 A.
state to the d,. level. The energies of these transitions If it is assumed in accord with the ionic crystal field
are seen from the diagram to be 4.11, 6.40, and 7.08 Dq model that, for small changes in bond distances, Dq
units, respectively, and are designated by arrows. values vary in proportion to I/+, where T is the metal-
(The order of the energy states of the dy complex are to-ligand distance, then the effectiveDq values for both
just the reverse of those of the d1 system. If the dia- the equatorial and axial chloride ligands in CUCI,~-can
grams above were drawn in terms of the energies of the be calculated from the value of 1280 cm-' established
atomic states instead of the orbitals, the diagrams for the in the previous example for chloride ligands in Cs2CuC14.
808 / lournol o f Chemicol Educofion
I n this latter complex, the average Cu-C1 bond dis- The frequency of the other transition is then predicted
tance is 2.22 A (22), so that the corrected Dq values are to be
Equatorid: Dq,, = 1280 X (2.22/2.39)5 = 885 cm-'
Axid: Dq., = 1280 X (2.22/2.30)' = 1070 cm-'
Dq.,+ 3.26 Dq., = 10.91
X 995 = 10.900 cm-'
The effects of the distortion due to the presence of
two different metal-to-ligand bond distances can be
calculated by the approach discussed previously for This value, obtained by evaluating the Dq parameter
substituted complexes. One can obtain the Dq values from data for the same complex, is in much better agree-
of the d orbitals in the desired axially substituted com- ment with the experimental frequency than the others.
plex IMX3X2' by simply adding the Dq values given in The above examples demonstrate that the calculation
Table 3 for a plane triangular group (MX3) with those of the frequencies of spectral transitions of complexes
given for the linear (axial) MX,'group, keeping in mind in terms of experimentally evaluated parameters is not
that the Dq values for the two types of ligands X and only easy to perform, but is also surprisingly successful.
X' are different, i.e., Dqx = Dq., and Dq,, = Dq,,. The errors and approximations inherent in the simple
Using the tabulated values for p = 1, it is seen that the point-charge crystal field model are largely washed out
relative energies of the d orbitals in the distorted if Dq and p are treated as empirical parameters to he
trigonal bipyramid structure are as shown in Figure 8. evaluated by fitting to experimental data. Calcula-
tions of crystal field effects in an absolute sense, on the
other hand, have not been very successful with the
simple crystal field model (23).
Other Applications. The Dq values and relative
energies of the d orbitals of many other complexes con-
taining more than one type of ligand can be derived in a
manner analogous to that illustrated above for suh-
stituted six-coordinate complexes. The method is
applicable to other geometrical configurations and
coordination numbers and is subject to the limitation
mentioned previously that each type of ligand be added
only in units of primary groups. For six-coordinate
complexes, the most highly substituted species that can
Figure 8. Energy level diogrorn of the d orbitals in tho distorted trigond be treated by the simple approach presented here is the
bipyrmmid structure. complex with four different types of ligands MX2Y2AB,
where the two A' ligands and the two I' ligands occupy
Inserting the values of Dq,, and Dqa, calculated above cis positions in the X-Y plane and the ligands A and B
gives the relative energies of the orbitals shown a t the lie at trans positions on the Z axis. In the case of
right of Figure 8. From these values we obtain substituted four-coordinate complexes, only the cis-
disubstituted square planar complex, MX2Y2,can he
constructed from the primary groups.
The possibility of calculating the Dq values of d or-
bitals in various real or hypothetical configurations of
These values are in better relative agreement with the ligands has applications also in the area of kinetics
experimental frequencies than are the values calculated (8). Because crystal field stabilization energies can he
above for the regular structure with no axial compres- easily computed for many postulated configurations, it
sion. The correction for the distortion has made the is possible to compare the stabilization energy of a
difference in the frequencies of the two transitions given complex in its ground state with the energies of
approximately the same as that of the observed transi- possible reaction intermediates of either lower or higher
tions (-2400 versus 2200 em-'). The fact that both coordination number, thus affording a comparison of the
contributions of the ligand fields of the possible inter-
calculated frequencies are somewhat low indicates that
the calculation of Dq values from the value obtained for mediate structures to the activation energy (24).
CszCuC14is not precise. Apparently, the Dq values for
both the axial and equatorial ligands should be slightly
larger, though the ratio of their values, (2.39/2.30)5 = Limitations
1.21, must be approximately correct.
The method described here for obtaining the Dq
As an alternative, it is also possible to apply the
values of d orbitals in fields generated by various num-
relation Dqa, = 1.21 Dq,,, which corrects for the distor-
bers and configurations of ligands is based on the simple
tion from a regular structure, and then to solve for the
crystal field theory as applied to a dl electronic config-
value of Dq,, by equating the calculated expression for uration. Complicating factors such as the Jahn-Teller
the energy of one transition to the observed frequency,
effect (25) and spin-orbit coupling (9-12) have not been
as was done in the first example.
considered. For the d1 configuration there are no
I n the example being considered
interelectronic repulsions such as arise in the presence
F, = 7.65 Dq,, + 0.58 Dp, = 8.23 Dq, = 8200 om-' of two or~moreelectrons, so that the splitting of the
and energy levels is determined only by the geometry and
strength of the ligand or crystal field. The d' electronic
Dp, = 995 cm-I configuration is thus the simplest system for assessing
Volume 46, Number 12, December 1969 / 809
the effects of ligand fields. Unfortunately, relatively few tivity of Dq values of d orbitals are identical to those
complexes with the d l electronic configuration have been derived from complete perturbation treatments ap-
prepared and studied, and relatively few spectra have plied t o the over-all structures. Relative d-orbital en-
been recorded. ergies are calculated for d' complexes with almost all
More exam~lesare known of comvlexes with a d gcon- known structures and with coordination numbers raug-
figuration, together with their spedra. The approach ing from one to twelve. The approach is also applicable
described for the d l system is directly applicable to ds to d9 complexes, to weak-field d4 and dB complexes, to
complexes, but the order of the energies of the states many distorted structures, and to most substituted com-
will he inverted. In this case, however, the approxi- plexes, providing that the substituting ligands are
mations should not he as good as in the d' case because added in units corresponding to the primary groups or in
the neglected spin-orbit coupling constants may become units derivable from them.
significant factors. The d' energy levels should also
he applicable to d4 and d6 complexes in their weak field
(high spin) configurations. Literature Cited
The tetragonally oriented basis set of d orbitals used (1) BETHE, H., Ann. Physik, 3, 133 (1929). (An English
(Fig. 1) is compatible with ligand fields in which the Z translation of this paper is available from Consultants
axis is the principal axis of symmet,ry and is (in effect) Bureau, Inc., New York, N. Y.)
at least a four-fold rotation axis. I n such fields the (2) VANVLECK,J., "The Theory of Electric and Paramagnetic
real d orbitals do not mix, hut retain a direct one-to-one Susceptibilities," Oxford Univ. Press, 1952.
(3) SCHLAPP, R., and PENNY, W., Phw. Re"., 42,666 (1932).
correspondence with the roots of the perturbation (4) ILSE,F., A N D HIRTMAN, H., Z. physlk Chem., 197, 239
energy equation obtained from only the diagonal (19.51).
matrix elements. In a few ligand fields, such as those of (5) ORGEL,L. E., J. Chem. Soe., 4756 (1952).
C2. (f go0), C X a (SOo),
f Dz,, and Dad symmetry, a differ- (6) NYHOLM, R. S., Qzm-1. Rev., 7, 377 (1953).
ent linear combination of d orbitals should be used since (7) ORGEL,L. E., "An Introduction to Transition Metal
the presence of off-diagonal elements in the potential
Chemistrv: Lieand Field Thoorv." .. Methuen and Ca..
Ltd., ~ o i d o n fi60.
,
fields of these point groups serves to mix the d orhitals. BASOLO,F., A N D PEARSON, R., "Mechanisms of Inorganio
I n these groups for which the additivity of Dq values Reactionn," John Wiley & Sons, Inc., New York, 1958,
does not apply, the ligand positions of the original Chapt. 2. (2nd Ed., 1967).
GRIFFITH,J. S., "The Theory of Transition Metal Ions,"
group plus all positions generated by the allowed in- Cambridge Univ. Press, Cambridge, 1961.
version through the origin do not give rise to a rotation BALLHAUSI:N, C. J., "Introduction to Ligand Field Theory,"
axis of at least fourth order. In some other small McGraw-Hill Book Co., New York, 1962.
groups to which the additivity relations do apply, such J#RGENSEN, C. K., "Absorption Spectra and Chemical
Banding in Complexes," Pergamon Press, Lid., London,
as C,,, C2,(9Oo),C8,(900),D2&and D3,, the ligand points 1962.
of the original group, together with thc points generated FIGGIS, B. N., "Introduction bo Ligand Fields," John Wiley
by inversion through tbc origin, always result in a & Sam, Inc., New York, 1966.
structure in which the Z axis is the principal n-fold HATFIELD, W., A N D PIPRR,T. S.,Inorg. Chem., 3,841 (1964).
rotation axis and n > 4. I n these cases all the terms DAY,P., Proc. Chem. Sac., 18 (1964).
COTTON,F. A,, "Chemical Applications of Group Theory,"
in the potential field expressions which can give rise to Interscienee (division of John Wilev & Sons Inc.),
off-diagonal matrix elements vanish, and the d orbitals New York, 1963.
of the defined set do not mix hut remain discrete so that (16) JAFFE, H., A N D ORCHIN,M., "Symmetry in Chemistry,"
the effects of the ligands on their energies are additive. John Wiley & Sons, Inc., New York, 1965.
(17) KRISANAMURTHY, R., SCHAAP, W. B., AND PERUMAREDDI,
Summary J., Inorg. Chem., 6, 1338 (1967).
(18) L ~ w r s J.,
, A N D WILKINS,R., (Editom),"Modern Coordina-
A simple approach has been presented for calculating tion Chemisiry," Interscienee (division of John Wiley &
the relative energies of d orbitals in crystal fields of a Sans, Inc.), New York, 1960. (See especially chapters
4 and 6).
large variety of geometrical configurations. The (19) FsncusoN, J., J . Chcm. Phgs., 40, 3406 (1960).
method is based on the use of three small, primary (20) FIGGIS,B., GERLOCH, &I., L&\!-Is, J., I N D SLIDE, R., 1.
groups of ligands which serve as building blocks in the Chem. Sac., 2028 (1968).
construction of larger, more complex structures. As (21) BERNAL, I., ELLIOTT,N., LAL.\NCETTG, R., A N D BRENNIN,
the building blocks are combined to form the more com- T., "Progress in Coordination Chcmistry," (Editor:
M. CIIS) Elsevier, Amsterdam, 1968, p. 518.
plex structures, the resultant energies of the d orbitals (22)
. . HELMHOLE, . R.,. J . Am. Chem. SOC.,74, 1176
. L... A N D I<RUH,
are obtained by adding algebraically the Dq values (1952).
associated with the ligand fields of the primary groups. (23) Onom, L. E., "Proceedings, Welch Faundat,ion Conference
Changes in orbital degeneracies and energies arising on Chemical Research. VI Topics in Modern Inorganic
Chemistry," IIauston, Texas, 1963, Chapt. IV.
from symmetry differences in structures of equivalent (24) SF'ESS, S., PERUMAR~DDI, J. R., A N D ADAMSON, A , J . Am.
coordination number are easily taken into account. Chem. Sac., 90, 6fi26 (1968).
The results obtained by use of this principle of addi- (25) VANVI.ECK,J . IT., J . Chem. Phys., 7, 61, 72 (1939).

810 / Journal of Chemical Education