Werner originally postulated that coordination compounds can be formed because the central atoms carry the capacity to form secondary, or coordinate, bonds, in addition to the normal, or valence, bonds. A more complete description of coordinate bonding, in terms of electron pairs, became possible in the 1920s, following the introduction of the concept that all covalent bonds consist of electron pairs shared between atoms, an idea advanced chiefly by the American physical chemist Gilbert N. Lewis. In Lewis’s formulation, when both electrons are contributed by one of the atoms, as in the boron-nitrogen bond formed when the substance boron trifluoride (BF3) combines with ammonia, the bond is called a coordinate bond:

Coordination Compound: boron-nitrogen bond formed when the substance boron trifluoride combines with ammonia. The bond is called a coordinate bond.

In Lewis’s formulas, the valence (or bonding) electrons are indicated by dots, with each pair of dots between two atomic symbols representing a bond between the corresponding atoms.

Following Lewis’s ideas, the suggestion was made that the bonds between metals and ligands were of this same type, with the ligands acting as electron donors and the metal ions as electron acceptors. This suggestion provided the first electronic interpretation of bonding in coordination compounds. The coordination reaction between silver ions and ammonia illustrates the resemblance of coordination compounds to the situation in the boron-nitrogen compound. According to this view, the metal ion can be regarded as a so-called Lewis acid and the ligands as Lewis bases:

Coordination Compound: coordination reaction between silver ions and ammonia

A coordinate bond may also be denoted by an arrow pointing from the donor to the acceptor.

Geometry

Many coordination compounds have distinct geometric structures. Two common forms are the square planar, in which four ligands are arranged at the corners of a hypothetical square around the central metal atom, and the octahedral, in which six ligands are arranged, four in a plane and one each above and below the plane. Altering the position of the ligands relative to one another can produce different compounds with the same chemical formula. Thus, a cobalt ion linked to two chloride ions and four molecules of ammonia can occur in both green and violet forms according to how the six ligands are placed. Replacing a ligand also can affect the colour. A cobalt ion linked to six ammonia molecules is yellow. Replacing one of the ammonia molecules with a water molecule turns it rose red. Replacing all six ammonia molecules with water molecules turns it purple.

Among the essential properties of coordination compounds are the number and arrangement of the ligands attached to the central metal atom or ion—that is, the coordination number and the coordination geometry, respectively. The coordination number of a particular complex is determined by the relative sizes of the metal atom and the ligands, by spatial (steric) constraints governing the shapes (conformations) of polydentate ligands, and by electronic factors, most notably the electronic configuration of the metal ion. Although coordination numbers from 1 to 16 are known, those below 3 and above 8 are rare. Possible structures and examples of species for the various coordination numbers are as follows: three, trigonal planar ([Au {P(C6H5)3}3]+; four, tetrahedral ([CoCl4]2−) or square planar ([PtCl4]2−); five, trigonal bipyramid ([CuCl5]>}]3−) or square pyramid (VO(acetylacetonate)2); six, octahedral ([Co(NO2)6]3−) or trigonal prismatic ([Re {S2C2(C6H5)2}3]); seven, pentagonal bipyramid (Na5[Mo(CN)7].10H2O), capped trigonal prism (cation in [Ca(H2O)7]2[Cd6Cl16(H2O)2].H2O), or capped octahedron (cation in [Mo(CNC6H5)7][PF6]2); eight, square antiprism or dodecahedron ([Zr(acetylacetonate)4]; and nine, capped square antiprism (La(NH3)9]3+) or tricapped trigonal prism ([ReH9]2−).

Coordination numbers are also affected by the 18-electron rule (sometimes called the noble gas rule), which states that coordination compounds in which the total number of valence electrons approaches but does not exceed 18 (the number of electrons in the valence shells of the noble gases) are most stable. The stabilities of 18-electron valence shells are also reflected in the coordination numbers of the stable mononuclear carbonyls of different metals that have oxidation number 0—e.g., tetracarbonylnickel, pentacarbonyliron, and hexacarbonylchromium (each of which has a valence shell of 18).

The 18-electron rule applies particularly to covalent complexes, such as the cyanides, carbonyls, and phosphines. For more ionic (also called outer-orbital) complexes, such as fluoro or aqua complexes, electronic factors are less important in determining coordination numbers, and configurations corresponding to more than 18 valence electrons are not uncommon. Several nickel(+2) complexes, for example—including the hexafluoro, hexaaqua, and hexaammine complexes—each have 20 valence electrons.

Any one metal ion tends to have the same coordination number in different complexes—e.g., generally six for chromium(+3)—but this is not invariably so. Differences in coordination number may result from differences in the sizes of the ligands; for example, the iron(+3) ion is able to accommodate six fluoride ions in the hexafluoro complex [FeF6]>]3− but only four of the larger chloride ions in the tetrachloro complex [FeCl4]. In some cases, a metal ion and a ligand form two or more complexes with different coordination numbers—e.g., tetracyanonickelate [Ni(CN)4]>]2− and pentacyanonickelate [Ni(CN)5]>]3−, both of which contain Ni in the +2 oxidation state.

Isomerism

Coordination compounds often exist as isomers—i.e., as compounds with the same chemical composition but different structural formulas. Many different kinds of isomerism occur among coordination compounds. The following are some of the more common types.

Cis-trans isomerism

Cis-trans (geometric) isomers of coordination compounds differ from one another only in the manner in which the ligands are distributed spatially; for example, in the isomeric pair of diamminedichloroplatinum compounds Coordination Compound: isomeric pair of diamminedichloroplatinum compoundsthe two ammonia molecules and the two chlorine atoms are situated next to one another in one isomer, called the cis (Latin for “on this side”) isomer, and across from one another in the other, the trans (Latin for “on the other side”) isomer. A similar relationship exists between the cis and trans forms of the tetraamminedichlorocobalt(1+) ion:

Coordination Compound: the cis and trans forms of the tetraamminedichlorocobalt(1+) ion.

Enantiomers and diastereomers

So-called optical isomers (or enantiomers) have the ability to rotate plane-polarized light in opposite directions. Enantiomers exist when the molecules of the substances are mirror images but are not superimposable upon one another. In coordination compounds, enantiomers can arise either from the presence of an asymmetric ligand, such as one isomer of the amino acid, alanine (aminopropionic acid), Coordination Compound: an isomer of the amino acid, alanine (aminopropionic acid)or from an asymmetric arrangement of the ligands. Familiar examples of the latter variety are octahedral complexes carrying three didentate ligands, such as ethylenediamine, NH2CH2CH2NH2. The two enantiomers corresponding to such a complex are depicted by the structures below.

Coordination Compound: Structural formulas for the two enantiomers corresponding to an octahedral complex.

The ethylenediamine ligands above are indicated by a curved line between the symbols for the nitrogen atoms.

Diastereomers, on the other hand, are not superimposable and also are not mirror images. Using AB as an example of a chelating ligand, in which the symbol AB implies that the two ends of the chelate are different, there are six possible isomers of a complex cis-[M(AB)2X2]. For example, AB might correspond to alanine [CH3CH(NH2)C(O)O], where both N and O are attached to the metal. Alternatively, AB could represent a ligand such as propylenenediamine, [NH2CH2C(CH3)HNH2], where the two ends of the molecule are distinguished by the fact that one of the Hs on a C is substituted with a methyl (CH3) group.

Ionization isomerism

Certain isomeric pairs occur that differ only in that two ionic groups exchange positions within (and without) the primary coordination sphere. These are called ionization isomers and are exemplified by the two compounds, pentaamminebromocobalt sulfate, [CoBr(NH3)5]SO4, and pentaamminesulfatocobalt bromide, [Co(SO4)(NH3)5]Br. In the former the bromide ion is coordinated to the cobalt(3+) ion, and the sulfate ion is outside the coordination sphere; in the latter the sulfate ion occurs within the coordination sphere, and the bromide ion is outside it.

Linkage isomerism

Isomerism also results when a given ligand is joined to the central atom through different atoms of the ligand. Such isomerism is called linkage isomerism. A pair of linkage isomers are the ions [Co(NO2)(NH3)5]2+and [Co(ONO)(NH3)5]2+, in which the anionic ligand is joined to the cobalt atom through nitrogen or oxygen, as shown by designating it with the formulas NO2(nitro) and ONO(nitrito), respectively. Another example of this variety of isomerism is given by the pair of ions [Co(CN)5(NCS)]3− and [Co(CN)5(SCN)]3−, in which an isothiocyanate (NCS) and a thiocyanate group (SCN) are bonded to the cobalt(3+) ion through a nitrogen or sulfur atom, respectively.

Coordination isomerism

Ionic coordination compounds that contain complex cations and anions can exist as isomers if the ligands associated with the two metal atoms are exchanged, as in the pair of compounds, hexaamminecobalt(3+) hexacyanochromate(3–), [Co(NH3)6][Cr(CN)6], and hexaamminechromium(3+) hexacyanocobaltate(3–), [Cr(NH3)6][Co(CN)6]. Such compounds are called coordination isomers, as are the isomeric pairs obtained by redistributing the ligands between the two metal atoms, as in the doubly coordinated pair, tetraammineplatinum(2+) hexachloroplatinate(2–), [Pt(NH3)4][PtCl6], and tetraamminedichloroplatinum(2+) tetrachloroplatinate(2–), [PtCl2(NH3)4][PtCl4].

Ligand isomerism

Isomeric coordination compounds are known in which the overall isomerism results from isomerism solely within the ligand groups. An example of such isomerism is shown by the ions, bis(1,3-diaminopropane)platinum(2+) and bis(1,2-diaminopropane)platinum(2+),

Coordination Compound: the ions bis(1,3-diaminopropane)platinum(2+) and bis(1,2-diaminopropane)platinum(2+)

Bonding theories

Valence bond theory

Several theories currently are used to interpret bonding in coordination compounds. In the valence bond (VB) theory, proposed in large part by the American scientists Linus Pauling and John C. Slater, bonding is accounted for in terms of hybridized orbitals of the metal ion, which is assumed to possess a particular number of vacant orbitals available for coordinate bonding that equals its coordination number. (See the article chemical bonding for a discussion of the theories of chemical bonding.) Each ligand donates an electron pair to form a coordinate-covalent bond, which is formed by the overlap of an unoccupied orbital of the metal ion and a filled orbital of a ligand. The configuration of the complex depends on the type and number of orbitals involved in the hybridization—e.g., sp (linear), sp3 (tetrahedral), dsp2 (square planar), and d2sp3 (octahedral), in which the superscripts denote the number of orbitals of a particular type. In many cases, the number of unpaired electrons, as determined by magnetic susceptibility measurements, agrees with the theoretical prediction. The theory was modified in 1952 by the Canadian-born American Nobel chemistry laureate Henry Taube, who distinguished between inner orbital complexes (d2sp3) and outer orbital complexes (sp3d2) to account for discrepancies between octahedral complexes. The main defect of the simple VB theory lies in its failure to include the antibonding molecular orbitals produced during complex formation. Thus, it fails to offer an explanation for the striking colours of many complexes, which arise from their selective absorption of light of only certain wavelengths. From the early 1930s through the early 1950s, VB theory was used to interpret almost all coordination phenomena, for it gave simple answers to the questions of geometry and magnetic susceptibility with which chemists of that time were concerned.

Crystal field theory

Considerable success in understanding certain coordination compounds also has been achieved by treating them as examples of simple ionic or electrostatic bonding. The German theoretical physicist Walther Kossel’s ionic model of 1916 was revitalized and developed by the American physicists Hans Bethe and John H. Van Vleck into the crystal field theory (CFT) of coordination, used by physicists as early as the 1930s but not generally accepted by chemists until the 1950s. This view attributes the bonding in coordination compounds to electrostatic forces between the positively charged metal ions and negatively charged ligands—or, in the case of neutral ligands (e.g., water and ammonia), to charge separations (dipoles) that appear within the molecules. Although this approach meets with considerable success for complexes of metal ions with small electronegative ligands, such as fluoride or chloride ions or water molecules, it breaks down for ligands of low polarity (charge separation), such as carbon monoxide. It also requires modification to explain why the spectral (light-absorption) and magnetic properties of coordinated metal ions generally differ from those of the free ions and why, for a given metal ion, these properties depend on the nature of the ligands.