Yves Jean - Molecular Orbitals of Transition Metal Complexes (793957), страница 8
Текст из файла (страница 8)
This is the highestoccupied orbital on the ligand, and its energy is not very different fromthat of the d orbitals for most of the transition metals. Moreover, itsLigand orbitalsoverlap with a metal orbital (e.g. z2 , 1-34) is substantial since it is ahybrid orbital polarized towards the metal centre. The resulting interaction produces a bonding and an antibonding MO. If the former is doublyoccupied and the latter is empty, there is a σ bond between the metaland the ligand (σM−PR3 , or σM−CR3 , 1-34).SNR3 , PR3CR 3M–L1-33a1-33b1-34If we consider only the nonbonding orbital on ligands such as NR 3 ,PR 3 , or CR 3 , we are effectively supposing that the interactions of theother MO with the metal orbitals are much weaker than the σ interaction already described.
This hypothesis can be justified by analysingthe orbital structure of pyramidal AH3 molecules (Figure 1.4), wherewe see three bonding molecular orbitals (σA−−H ) that account for theA-H bonds, the nonbonding orbital thatisconcentratedon the central ∗ . The bonding orbitals ofatom, and three antibonding orbitals σA−Hthe ligand can interact with the metal orbitals. But they are far too lowin energy, since they are orbitals describing the σA−H bonds.
Moreover,they are partially distributed over the hydrogen atoms, that is, in thedirection away from the metal.For both these reasons (substantial energy gap and poor overlap), the interactions involving the bonding orbitals are weaker thanthose concerning the nonbonding orbital. In a similar way, interac∗orbitals and the metal centretions between the antibonding σA−Hare usually negligible for the description of the metal–ligand bond;these MO are at high energy and partially oriented away from themetal.3a12e2a1antibonding *A–HMOnonbonding MO1eFigure 1.4. Molecular orbitals for AH3pyramidal molecules (with the electronicoccupation appropriate for molecules witheight valence electrons, such as NH3 or PH3 ).bonding A–HMO1a1Setting the sceneCommentIt must, however, be noted that the antibonding orbitals can be stronglystabilized by very electronegative substituents, as, for example, in trifluorophosphine, PF3 .
In such a case, these orbitals can interact to a non-negligibleextent with the d orbitals of the metal centre. This is particularly importantfor the 2e orbitals (Figure 1.4) which can be involved in π-type interactions(see Chapter 3).1.5.2. Several orbitals: σ and π interactionsFor other ligands the situation is more complicated, since it is necessaryto take several orbitals into account to obtain a satisfactory descriptionof the bond with the metal. As a first example, we shall consider bentAH2 molecules.1.5.2.1. Ligands of the AH2 typeThe orbital structure of bent AH2 molecules (or more generally AR 2 ),see Figure 1.5, shows that there are two bonding orbitals, whichdescribe the σA−H bonds, the two corresponding antibonding orbitals∗, and at an intermediate energy level, two nonbonding molecularσA−Horbitals: the 2a1 orbital, which is a hybrid pointing in the directionaway from the hydrogen atoms, and the 1b1 orbital, which is a purep atomic orbital perpendicular to the molecular plane, (see Chapter 6,§ 6.5.2).
Depending on the nature of the atom A, these two orbitals cancontain one electron (BH2 , AlH2 ), two (CH2 , SiH2 ), three (NH2 , PH2 ),or four (OH2 , SH2 ).As in the previous example, the bonding and antibonding orbitalscan, in a first approximation, be neglected for the description ofthe metal–ligand interactions.
However, it is necessary to take bothnonbonding orbitals into account, as they are close in energy and bothcan lead to interactions with the metal centre. The 2a1 orbital plays the3a1antibonding *A–HMO2b21b12a1Figure 1.5. Molecular orbitals for AH2molecules (with the electronic occupationappropriate for molecules with six valenceelectrons, such as CH2 or SiH2 in theirlowest singlet state).nonbonding MO1b2bonding A–H1a1MOLigand orbitalsM–L1-35yzpxxzM–L1-36same role as the 2a1 orbital in AH3 molecules, and its interaction witha metal orbital (z2 , for example) leads to the formation of an MO thatcharacterizes a σ bond (1-35).The 1b1 orbital (px ) has the correct symmetry to interact with the xzorbital: the overlaps above and below the yz plane have the same sign, sothat the total overlap (the sum of the partial overlaps) is non-zero (1-36).This new metal–ligand interaction is said to be a π interaction, since theorbitals concerned share a common nodal plane (yz).
It leads to the formation of two MO, one bonding, πM−L , represented schematically in 1-36,∗. This interaction is particularly importantand one antibonding, πM−Lfor the description of the metal–ligand bond when two electrons are concerned: the bonding molecular orbital (1-36) is then doubly occupied butthe antibonding orbital is empty. A π-type interaction therefore adds toand reinforces the σ interaction (1-35), giving a double-bond characterto the metal–ligand bond. It should be noted that there is an importantdifference between the σ (1-35) and π (1-36) interactions concerningthe change in the overlap when there is a rotation about the M-L bond:the σ overlap does not change, as it possesses cylindrical symmetrywith respect to the bond direction, but the π overlap is eliminated by arotation of 90◦ .CommentThis is exactly analogous to what happens in an organic molecule suchas ethylene (Figure 1.3), in which there is a σCC bond for any relativeorientation of the two methylene groups, whereas the existence of a πCCbond depends on the two groups being coplanar.1.5.2.2.
AH ligandsFollowing the reasoning developed in the previous paragraph, threemolecular orbitals need to be considered for an A-H (or A-R) ligand(see Figure 1.6): the nonbonding orbital 2σ , analogous to the 2a1 orbitalin AH3 and AH2 ligands, which allows a metal–ligand σ bond to beformed, and the two degenerate π orbitals (px and py ) which may beinvolved in π interactions with the metal centre. Depending on thenature of A, these three orbitals may contain two electrons (BH, AlH),three (CH, SiH), four (NH, SH), five (OH, SH), or six (FH, ClH).1.5.2.3.
Monoatomic ligands AWith the exception of the ligand H, for which there is only a singlevalence orbital 1sH to consider (a σ interaction), one must treat all thevalence s and p orbitals on a monoatomic ligand A. Now the s orbital isusually much lower in energy than the d orbitals on the metal, especiallySetting the scene3antibonding *A–H MO1nonbonding MO2Figure 1.6. Molecular orbitals for AHmolecules (with the electronic occupationappropriate for molecules with four valenceelectrons, such as BH or AIH in their lowestsinglet state).1bonding A–H MOif A is a fairly electronegative element. In this case, one can thereforeneglect the interaction of the s orbital (which, if it is doubly occupied,therefore describes a lone pair localized on A), and only consider thethree p orbitals on the ligand, which are higher in energy and thereforecloser to the metal d orbitals.
The one which points towards the metalcentre (pz , 1-37) is used for the σ interaction, and the two orbitals whoseaxis of revolution is perpendicular to the bond (px and py ) can lead to πinteractions (1-38).z2pz1-37xzyzpxpy1-38CommentIn a more sophisticated model of the σ orbitals, one can suppose that pzand s are mixed, to form two s–p hybrid orbitals, one pointing towardsthe metal to form the σ bond, and the other in the opposite direction,so as to describe a lone pair of σ type on A.
This new approach does notfundamentally change anything in the simplified description given above.The preceding examples show that one must always consider theatomic or molecular orbital on the ligand that allows a σ bond to beformed with the metal. This orbital may be a nonbonding s orbital (H),a hybrid s–p orbital (AH3 , AH2 , and AH molecules), or a p orbital whichpoints towards the metal centre (A atoms).
When the atom bound tothe metal also possesses nonbonding p orbitals perpendicular to themetal–ligand bond (AH2 , AH, A), it is also necessary to consider them,since they lead to π -type interactions with the metal orbitals.Even though it is an approximation to neglect the other MO on theligand, this usually leads to an acceptable description of the bond withLigand orbitalsthe metal centre. It is particularly appropriate when the neglected ligandorbitals are very low in energy and the antibonding orbitals very high.These conditions are usually satisfied when the orbitals are involved inthe σ bonds of the ligand. However, when the ligand possesses one ormore π bonds, its π bonding and π ∗ antibonding molecular orbitalsmust usually be taken into account.1.5.2.4.
Ligands with a π system: the example of CO* COCCOCOFigure 1.7. Electronic stucture of CO (threehighest occupied and two lowest emptyorbitals).12A more detailed analysis will enable usto show that one can, in a first approximation,reduce this number of orbitals to three: theσ orbital and the two π ∗ orbitals (Chapter 3, §3.2.2).When the ligand possesses a π bond involving the atom bound to themetal (η1 coordination), this leads to the presence of a π bonding andπ ∗ antibonding orbital on the ligand.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
