LIGAND FIELD THEORY AND ITS APPLICATIONSP262R AD LIGAND FIELD THEORY AND ITS APPLICATIONS BRIANN.FIGGIS University of Western Australia MICHAEL A. HITCHMAN @® WILEY-VCH New York / Chichester / Weinheim / Brisbane / Singapore / TorontoThis book is printed on acid-free paper. @ Copyright © 2000 by Wiley-VCH. All rights reserved, Published simultaneously in Canada. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except as permitted under Sections 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4744, Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc,, 605 Third Avenue, New York, NY 10158-0012, (212) 850-6011, fax (212) 850-6008, E-Mail: PERMREQ @ WILEY.COM. For ordering and customer servi I 1-800-CALL-WILEY. Library of Congress Cataloguing in Publication Data: Figgis, Brian N. Ligand field theory and its applications / Brian N. Figgis, Michael A. Hitchman. pcm Includes index. ISBN 0-471-31776-4 (cloth: alk. paper) 1. Ligand field theory. 1. Hitchman, Michael A. Il. Title. (QD475.F54 2000 99-28986 S41.2/242-de21 Printed in the United States of America, 1098765432‘Toour wives, Jane and Annette, and children, Benjamin, Honor, Martin, Peter, and Ruth, who said it would never get done!SPECIAL TOPICS IN INORGANIC CHEMISTRY INTRODUCTION TO THE SERIES This text represents the first in a series of one-volume introductions to major areas of inorganic chemistry written by leaders in the field. Inorganic chemistry covers a variety of diverse substances including molecular, coordination, organometallic, and nonmolecular compounds as well as special materials such as metallobiomolecules, semiconductors, ceramics, and minerals. The great structural diversity of inorganic compounds makes them vitally important as industrial feedstocks, fine chemicals, catalysts, and advanced materials. Inorganic compounds such as metalloenzymes also play a key role in life processes. This series will provide valuable, concise graduate texts for use in survey courses covering diverse areas of inorganic chemistry. Department of Chemistry R. Bruce KING, Series Editor University of Georgia Athens, Georgia vilCONTENTS PREFACE 1 INTRODUCTION 1 al ee a . - 1.6 _ The Concept of a Ligand Field / 1 The Scope of Ligand Field Theory / 4 The d and Other Orbitals / 5 The Symmetry Properties of Molecules and Wavefunctions / 14 1.4.1 The Molecular Point Groups / 14 1.4.2 The Representations of Wavefunctions / 15 1.43. Typical Applications in Ligand Field Theory / 16 Qualitative Demonstration of the Ligand Field Effect / 17 The Physical Properties Affected by Ligand Field Theory / 21 1.6.1 Thermochemical Properties and Geometric Distortions / 21 1.6.2 Spectral Properties / 22 1.6.3 Magnetic Properties / 23 Crystal Fields and Ligand Fields / 24 QUANTITATIVE BASIS OF CRYSTAL FIELDS 7 a ee) Crystal Field Theory / 27 ‘The Octahedral Crystal Field Potential / 30x CONTENTS 2.3. The Effect of Vgc, on the d Wavefunctions / 33 2.4 The Evaluation of A / 38 2.5 The Tetrahedral and Cubic Potentials / 40 2.6 Naming the Real d Orbitals / 41 2.7 Potentials for Lower Symmetries / 41 2.7.1 Tetragonally Distorted Octahedron / 41 2.7.2 Trigonally Distorted Octahedron / 44 2.8 Other Parameterization Schemes / 45 2.8.1 The A and B Coefficients / 46 2.8.2 The Parameters Ds and Dt / 47 2.9 Limitations of Crystal Field Theory: Ligand Field Theory / 48 2.10 f Orbitals and the Crystal Field Potential / 50 THE ANGULAR OVERLAP MODEL 53 3.1 Basis of the Angular Overlap Model (AOM) / 53 3.1.1 Simple MO Picture of the Bonding in Transition-Metal Complexes / 54 3.1.2 Derivation of AOM Parameters Using the Wolfsberg—Helmholtz Approximation / 58 3.1.3. Derivation of the d Orbital Energies and Wavefunctions in a Complex Using the AOM / 60 3.2. AOM Expressions for Complexes of Various Symmetries / 61 3.2.1 Octahedral Complexes / 61 3.2.2. Tetragonally Distorted Octahedral, Planar, and Linear Complexes: The Importance of d-s Mixing / 64 3.2.3 Tetrahedral, Distorted Tetrahedral, and Square-Based Pyramidal Complexes / 67 Trigonal Bipyramidal Complexes / 68 Variation of AOM Parameters with Bond Distance / 69 Typical AOM Parameters / 70 3.3. Extensions of the AOM for Some Polyatomic Ligands / 73 3.3.1 Phase-Coupled Ligators / 73 3.3.2 Off-Axis Bonds and Interactions with Nonbonding Electron Pairs / 74 3.4 Approximations in the Derivation of Bonding Parameters / 76 3.5. Advantages of the AOM Compared to the Electrostatic Crystal Field Theory / 78 3.6 Calculations of Electronic Spectra and Magnetic Properties Using Computer Programs Based on the AOM / 80CONTENTS xi 4 THE ORIGIN AND CALCULATION OF A 83 41 co 7 44 45 46 Calculations Based on Electrostatic Interactions / 83 One-Electron Molecular Orbital Calculations / 84 All-Electron Molecular Orbital Calculations / 87 Symmetries Lower Than Cubic / 89 f Electron Systems / 89 Real Electron Density Distribution / 90 5 ENERGY LEVELS OF TRANSITION METAL IONS 92 a — 53 54 5.5 Introduction / 92 Free Transition Ions / 93 Free Ion Terms / 95 5.3.1 d'/ 98 53.2 d?/ 99 5.3.3 d°-d? / 101 Term Wavefunctions / 103 Spin-Orbit Coupling / 107 6 EFFECT OF LIGAND FIELDS ON THE ENERGY LEVELS OF TRANSITION IONS 12 6.1 65 6.6 iy ‘The Effect of a Cubic Ligand Field on S and P Terms / 112 6.1.1 S Terms / 113 6.1.2 P Terms / 113 The Effect of a Cubic Ligand Field on D Terms / 115 The Effect of a Cubic Ligand Field on F Terms / 117 The Effect of a Cubic Ligand Field on G, H, and I Terms / 121 Strong-Field Configurations / 121 ‘Transition from Weak to Strong Ligand Fields / 122 Correlation Diagrams / 126 6.7.1. d',d°: Any Stereochemistry / 127 6.7.2 d? (Oy), d® (Tq) / 127 6.7.3 d? (Tq), d® (Oy) / 127 6.7.4 d® (Op), d? (Ta) / 127 6.7.5 d°(Tq),d7 (Oy) / 127 6.7.6 d* (Op), d° (Ta) / 128 6.7.7 d* (Ta), d° Oy) / 129 6.7.8 d°(O,and T,) / 130xii CONTENTS. 6.8 Tanabe-Sugano Diagrams / 131 6.9 Spin-Pairing Energies / 141 7 INFLUENCE OF THE d CONFIGURATION ON THE GEOMETRY AND STABILITY OF COMPLEXES 145 7.1 Dependence of the Geometry of a Complex on Its d Configuration / 146 7.1.1 Nondegenerate Electronic States / 146 7.1.2 Degenerate States: The Jahn-Teller Effect / 148 7.1.3. Bond Length Differences Between High- and Low-spin Complexes / 160 7.1.4 Variation of Bond Lengths on Crossing the Transition Series / 163 7.2. Dependence of the Stability of a Complex on Its d Configuration / 166 7.2.1 Thermodynamic Effects / 167 7.2.2 Kinetic Effects / 176 8 THE ELECTRONIC SPECTRA OF COMPLEXES 79 8.1 Important Features of Electronic Spectra / 179 8.1.1 Band Intensities / 179 8.1.2 Band Energies / 189 8.1.3 Band Widths and Shapes / 195 8.1.4 Effect of Temperature on Electronic Bands / 201 8.2 Characteristic Spectra of Complexes of First-Row Transition Tons / 204 8.2.1 Hexaaqua Complexes of First-Row Ions / 204 8.2.2. Tetrahedral and Planar Complexes of First-Row Transition Ions / 211 8.3. Typical Spectra of Second- and Third-Row Transition Ions / 214 8.4 The Spectrochemical and Nephelauxetic Series / 215 8.4.1 The Spectrochemical Series / 215 8.4.2 The Nephelauxetic Series / 218 8.5 Charge Transfer Spectra / 221 8.6 Luminescence Spectra / 224 9 MAGNETIC PROPERTIES OF COMPLEX IONS 228 9.1 The Theory of Magnetic Susceptibility / 228 9.1.1 General / 2289.2 9.3 9.4 96 97 98 9.9 9.10 911 CONTENTS xl 9.1.2 Paramagnetism / 232 9.1.3 Quantum Mechanical Treatment of Paramagnetic Susceptibilities / 233 The Magnetic Properties of Free Ions / 237 9.2.1 The First-Order Zeeman Effect / 237 9.2.2 The Second-Order Zeeman Effect / 238 9.2.3 States ~kT / 239 9.2.4 States His (1.3) Third transition series : Hip ~ Her © His (1.4) Lanthanides : Her 2 His > Hie (1.5) Actinides : Herz Hys © Wir (1.6) Hyp takes into account the presence of the surrounding ligand molecules, whether treated as point charges, as in the crystal field model, or as at least partly covalently bound entities, as in ligand field methods. It will be discussed first in some detail, because it forms the whole reason for the presentation of this chapter. Hp takes into account the repulsions within the d or f electron set, including the application of the constraints of the Pauli exclusion principle and electron spin. It consists of two components — the Coulomb repulsion and the exchange terms. This subject will be introduced only to the extent necessary for the application of the Hyp term to relevant physical properties (Chapter 6). Hys takes into account the formal coupling between the spin and the angular momenta of the d or f electrons. In fact, it arises from the same basis as Hx and is treated along with that term for the same purposes (Chapter 5). 14.3. THE d AND OTHER ORBITALS At the heart of crystal field theory, and of ligand field theory in general, lie the properties of the set of five d orbitals. The following discussion summarizes6 INTRODUCTION Figure 1.1. Spherical polar coordinates in relationship to Cartesian coordinates. these properties and the properties of the hydrogen-like wavefunctions in general (4). The hydrogen-like wavefunctions of a free ion (by a free ion is meant an ion not subject to any external influences) are eigenfunctions of the Hamiltonian operator: Hy = —(h?/87?m)V? — Zeee?/r (1.7) where Ze is the effective nuclear charge. Referred to a set of spherical polar coordinates (Fig. 1.1), the wavefunctions are given by: Vrim = Ru VP (1.8) where R,v is the radial and Y/" the angular component of the wavefunction. Here, nis the principal quantum number, / the quantum number specifying the orbital angular momentum of the electron, and m the quantum number specifying the component of this angular momentum in the z direction. For the present, electron spin is not considered. » is a positive integer, / takes on integral values from 0 to (n—1), and m takes on integral values from —I to /. In Figure 1.1, it may help to recognize that in referrence to the surface of the earth, # corresponds to latitude, with zero at the North Pole, and ¢ to longitude, with zero at Greenwich. The element of volume dr is given by: dr= sin®-d0-dd (1.9) The function r?R3,, with Ry, taken to be real (as is always the case) describes the time-averaged probability of finding an electron in the specified orbital at aTHE d ANDOTHERORBITALS «7 distance r from the nucleus. The factor r? occurs here simply because the volume of a shell of thickness dr is proportional to r?. For reasons that are given later, it is of little use to employ radial distribution functions taken from free ion data in actual ligand field calculations. Thus we do not consider the details of these functions further than to note that, by definition, the relationships in Eq. 1.10 hold; i.e., the entire set of hydrogen-like wavefunctions form an ortho- normal set: [RoReuvTve' ar = 8(n,n/sh,U'sm, m!) (1.10) This equality means that the integral is zero unless m = n‘;1 = I' and m =m’. Some properties of the spherical harmonics Y/" are outlined in Appendix 1, and the functions up to order six are listed in Appendix 2. Rather than talking about electrons described by wavefunctions, it is conventional to say that electrons occupy orbitals. These orbitals are specified by the foregoing angular parts of the wavefunctions and can be represented pictorially. Such diagrams are given in Figure 1.2 as electron densities corresponding to the s, p, d, and f orbitals. The plots show regions of space in which there is a high probability of finding the electron and have axial symmetry about the z axis. Considering the wavelike properties of the electron, the amplitude may have a positive ( +) or negative (—) phase. The functions Y? are essentially concentrated along the z axis, the pairs Y}*" form rings around the z axis, which look identical. They differ only in that one member of the pair may be regarded as corresponding to clockwise rotation of the electron cloud about the z axis and the other member to couter-clockwise rotation. The pair with |m| = is always located with maxima in the xy plane. ‘Actual calculations are usually performed with these solutions to the Schrédinger equation. On the other hand, it is often more convenient, for the purposes of visualizing the relationships among the different orbitals, to express them in alternative forms that are explicit expressions of the Cartesian coordinates. In these alternative forms, the wavefunctions for the orbitals do not contain e’”* and consequently may be said to be the real orbitals. The real orbitals are simple linear combinations of those we gave above for cylindrical polar coordinates; the combinations are taken to relocate i. To combine the imaginary forms of the orbitals into the real forms we note that (de Moivre’s theorem): etim? — cos (mo) + isin (md) (1.11) and that x=r-sind-cosd, y=r-sind-sing, z= r-cos8 (1.12)Figure 1.2. The angular dependence of the electron density distribution corresponding to some of the s, p, d, and f functions that are solutions of the Schrédinger equation.THEA ANDOTHERORBTTALS = We take, for example, the combination (2) +(-2) (1.13) where (m) is an abbreviation for Y‘},, that is for a d wavefunction, so that (2) stands for the wavefunction R,2Y3. (2) = Raa: (15/16)"? - (2m)? sin? @- (cos26 +isin2¢) (1.14) (=2) =Rra- (15/16)"? - (2n)""- sin? @- (cos 24 — isin2¢) (1.15) (2) + (2) = 2Rya - (15/16)" - 20)" - sin? @- cos 24 = Rp - (15/16)? - (27)? - sin? 0 - (cos? ¢ — sin? ¢) = 2Ryp = (15/16)"? « (2m)? « (x? — y?) /r? x (x? — y?) (1.16) In the last step, all the factors other than functions of the Cartesian coordi- nates are omitted. Finally, it is necessary to make sure that the new orbital is normalized. To normalize a wavefunction VY we form the new wavefunc- tion: w/(] vwary* (17) Now, f@+ carte +a = fier + (2)*(=2) + (—2)"(2) + (-2)"(-2)]Jdr =14+0+041 =2 (1.18) Hence the real wavefunction, which is called d,2_,2, is: das yt = (2) + (29/4 2) + (29°12) + (2) 49" = 212) + (-2)] (1.19)40 wTRODUCTION By making other suitable combinations the five d wavefunctions in their real forms are obtained as: da = (0)(dq is really dig2,2/3)) (1.20) dy. = 12-71) + (-)] (1.21) dy, = 2°'7((1) — (-D] (1.22) day = -12-Y?((2) — (-2)] (1.23) dye_y2 = 277((2) + (-2)] (1.24) Exactly the same procedure may be adopted with p and f orbitals (the s orbital is real in form anyway). For example, the p orbitals, with p,, the p wavefunction specified by m, are: P.=Po (1.25) py = -12-"(p, — py) (1.26) p,=2'7(p) +p) (1.27) In Figure 1.3 the real forms of some of the p, d, and f orbitals are illustrated. The signs on different lobes of an orbital in these diagrams indicate the relative phases of the wavefunctions in that part of space. Of course, the probability density, from the square of the wavefunctions, is always positive. The shapes of the f orbitals are discussed more fully by King (5). The orthonormal properties of the real d wavefunctions are readily demonstrated. For example, [esetetr = forttay = (0) fa) = ener =3 f= C=) + -(-Djer =4i-0-0+3) =a (1.28) and Jeitenar =i 212) - (-2))-(@) + (-2)18r i fie + (2) + (2) -(-2) = (-2) - (2) = (-2)- (-2)]dr {1+0-0-]] 2 i 2 0 (1.29)THEGANDOTHERORBTTALS 11 PLP, fy¢sy2 - 372) Figure 1.3. The angular dependence of some functions representing real forms of s, p, d, and f orbitals. The difference in behavior under the operator for the z component of orbital angular momentum (J) exhibited by the d wavefunctions in the two forms is important and is a result that is required later. To find the observable value of orbital angular momentum that corresponds to a wavefunction 1, we have to evaluate the quantity (6): Jeevar (130) The actual form of [, is: 1, = (h/2m)(8/86) (1.31)2 INTRODUCTION Thus 1,(m) = m(h/2n)(m) (1.32) For example: 1,(2) = 2(h/2n)(2) (1.33) ‘We have, then: Jorn'setomar = [my (a/2mmimar = (6/2) fom omar Ce) = (h/2n)m and lo tent) = om" /2m)m"n! yar ae =0 Specifically, Jerrnerer = +2(h/2m) (1.36) [e2rnc2nr =o (1.37) which show that the wavefunctions (2) and (-2) are associated, respectively, with 2 and —2(h/2z7) units of orbital angular momentum in the z direction but that there is no cross-term (“off-diagonal element”) of I, between them. On the other hand, when the real forms of the orbitals are used: Jetstederde = [212 = 291-4102) - (-2)}¢r = 5 [l@-4@ - @) 14-2) - (2) 12.0) + (-2) -L(—2)]dr = 3 [Pe +202)-(-2) -2(-2)-) + 2(-2) - (~2)j(h/2n)dr = (h/2n)(2+0-0-2) =0 (1.38)THEd AND OTHERORBTALS 13 And Jostede ear = if21e) = (-2)] -L[(2) + (-2)]ar =$[I@ .@ — @) 14-2) - -2)-40) + (—2) -1,(-2)|dr = 5 Pe -@ - 20). (-2) -2-2)-@) +2(-2) - (—2)](h/2m)dr = 5 (h/2m2 —0-0+2) = 2i(h/2n) (1.39) Similarly, [orttuar 7 Josetsdndr =0 (1.40) Jeretatr =h/2n (1.41) In these real orbitals, all the orbital angular momentum in the z direction is associated with the cross-term between different members. Of course, for dz, because m=0 [ontatndr(er) Jes Iadr=0 (1.42) ‘The foregoing relationships have the following significance: The z component of orbital angular momentum, say, is associated with a torus of electron distribution about the z axis. It is zero if the electron distribution maximizes along that axis. In the orbitals specified by (m) such a distribution exists, whereas in those of the real type it does not, These latter orbitals have no z component of orbital angular momentum by themselves; however, because a rotation by 27/8 about the z axis transforms d,2_,2 into d,,, and a rotation by 27/4 transforms d,_ into dy., orbital angular momentum is to be associated with these two pairs of orbitals. The association of orbital angular momentum with such a rotation is conditional on the fact that the two orbitals that transform into each other under the rotation are degenerate. If the degeneracy is removed, the associated orbital angular momentum is lost, because the two orbitals cannot be rotated into each other without an energy change.14 iTRooucTION 1.4 THE SYMMETRY PROPERTIES OF MOLECULES AND WAVEFUNCTIONS The idea that a molecule is “highly symmetric” or of “low symmetry” is familiar in a qualitative sense to all chemists (7). At a more rigorous level, the symmetry of a molecule is described mathematically using group theory, which provides a powerful way of simplifying many chemical problems. It also forms the basis of a convenient notation system; the most widely used symbols are those of Schénflies. Indeed, group theory is now so important that it is included in many general textbooks; thus, the way in which symmetry properties are described using group theory is not covered here. A brief outline of the terminology is, however, appropriate. 1.41 The Molecular Point Groups The symmetry properties of a complex are introduced formally through the symmetry elements it exhibits. These symmetry elements include mirror planes, an inversion center, and rotation axes. To identify these, a Cartesian coordinate system may be defined at an appropriate point in the molecule, and the effect of the symmetry operations associated with each symmetry element—reflection, inversion, and rotation—may be explored. The resulting set of symmetry elements allows the molecule to be assigned a point group, so called because the operations are applied with respect to a fixed point in the molecule (the symmetry elements relating the atoms in a crystal unit cell include also translations and define a space group). Most coordination complexes exhibit a symmetry that, to at least a first approximation, conforms to that of one of the so-called Platonic solids, The commonest geometry involves six equivalent metal-ligand bonds at right angles, for instance, as in the CrF3~ ion. The Cartesian axes are drawn to coincide with the metal—ligand bonds, and the complex possesses a high degree of symmetry, having not only an inversion center but also many mirror planes and twofold, threefold, and fourfold rotation axes. The solid formed by joining the ligand atoms is the eight-faced octahedron shown in Figure 1.4a, and the complex belongs to the point group O,, The most common geometry involving four metal—ligand bonds conforms to the symmetry properties of the four-faced tetrahedron; and a complex of this kind, ¢.g., the CoCI2~ ion, belongs to the point group Ty. In this case, the Cartesian axes bisect the metal-ligand bond angles (Fig. 1.4b), and the complex does not have an inversion center. The axis system is chosen in the manner that allows the symmetry of the complex to be described most easily. By convention, z is the axis associated with the largest number of rotational symmetry elements. The point groups O;, and Ty both have cubic symmetry in that, like a regular cube, the directions x, y, and z are equivalent. This is not true for a square planar complex involving four equivalent metal—ligand bond lengths at right angles; that belongs to the point group Day, where the unique symmetry axis zis normal to the plane of the complex, with the equivalent x and y axes passing along the metal-ligand bond directions.‘THE SYMMETRY PROPERTIES OF MOLECULES AND WAVEFUNCTIONS 15 F-| > F fa) F ci cr Figure 1.4. The relationship between the geometry of an octahedral (a) and a tetrahedral (b) complex with those of the Platonic solids. Chemists often use the terminology describing symmetry quite loosely. For instance, the complex Cr(H,0)3* is generally referred to as an octahedral complex of O, symmetry, even though this is strictly t tue only if the hydrogen atoms are ignored. A complex such as Co(NH3)sCI** will also be described as octahedral, unless group theory is being used to interpret some property where the use of the true point group (neglecting the hydrogen atoms) Cay is necessary, for instance the absorption of polarized light as discussed in Section 8.1.1.4. 14.2 The Representations of Wavefunctions Features like the Cartesian axes and, most important, wavefunctions are conveniently classified by considering their behavior with respect to the various symmetry operations of the point group of the molecule. The traces, or characters (sums of diagonal elements), of the matrices used to represent the symmetry operations of the point group mathematically (irreducible representa- tions) are used to generate labels that summarize the possible ways in which the features of interest conform to the symmetry elements of the point group. The symmetry properties of the irreducible representations are conveniently presented in the character table of the point group. Even when the symmetry is not directly relevant, it is common to refer to wavefunctions by their irreducible representations. This notation is useful because it incorporates considerable information. For first-row transition ions, a good approximation is made by considering the spin and orbital parts of the wavefunctions separately (this aspect is discussed in some detail in chapter 5). The symmetry then refers just to the orbital part of each wavefunction. For common point groups such as Oy, Tg, and Dap, the wavefunctions all transform as A, B, E, or T, where the symbol denotes not only the symmetry of the wavefunction with respect to rotations but also the number16 INTRODUCTION of orbital wavefunctions of that type. A and B refer to a single orbital wavefunction (a nondegenerate set); E, to a pair of orbital wavefunctions (a doubly degenerate set) and T to three wavefunctions of equal energy (a triply degenerate set). When appropriate, the behavior with respect to rotations or reflections is distinguished by a numerical subscript. When the point group is centrosymmetric, each wavefunction falls into one of two subgroups, depending on whether it changes sign on inversion through the center. The representations that are “even” carry the subscript g, whereas those that are “uneven” carry the subscript u (from the German gerade and ungerade). Generally, capital letters are used when referring to an energy state, and small letters are used for orbitals, or sets of orbitals. Often, Greek letters are used for vibrational, as opposed to electronic, wavefunctions. As an illustration of this terminology, we note that the d orbitals of the complex Cr(H,0)%* are split into two sets: a lower energy tog set and a higher energy e, set. When describing the energy levels resulting from the interaction of the three valence d electrons with one another and with the water ligands, we say the lowest level is an Agg state and the first excited state is a Ty state. This terminology refers just to the orbital component of the states; the designation of the spin component is discussed in chapter 5. The totally symmetric vibration of the complex has a1, symmetry. 14.3 Typical Applications in Ligand Field Theory The power of group theory, as applied to the symmetry properties of transition metal complexes, lies in its ability to answer certain chemically relevant questions very simply. For instance, by simply inspecting the representations of the d orbitals in the character table for the point group O,, it may be seen that the day, dz, and dye orbitals transform as tg, while the d,:_y2, d.: pair of orbitals transform as ¢,. The fact that the d orbitals split into two sets in a molecule with O, symmetry may be derived just from their symmetry labels alone, without the need for any calculation. It must be stressed, however, that group theory by itself gives no indication of the magnitude, or even the sign of the energy splitting between the sets. It provides a qualitative, but not a quantitative deduction. Like the octahedron, the cube belongs to the group Os, but we shall see it produces a ligand field splitting of opposite sign (Section 2.5). Group theory also provides a powerful means of telling whether certain integrals can be nonzero. This will be the case only if the direct product of the relevant wavefunctions and operator gives the totally symmetric representation of the point group of the complex. Thus the probability that an octahedral complex can undergo a transition in which an electron is excited from the tog orbitals to the e, orbitals by absorbing a photon of light may be investigated by inspecting the result of the direct product: lag % thy X eg (1.43)