Coordination Compounds: Werner's Theory, Nomenclature and Isomerism

Alfred Werner, a Swiss chemist, spent three years in the 1890s puzzling over a strange set of results. Four compounds, all built from cobalt(III) chloride and ammonia in different ratios, gave four completely different answers to the same simple test. Dissolve one mole of CoCl3.6NH3 in water, add excess silver nitrate, and three moles of white AgCl precipitate out immediately. Do the same with CoCl3.5NH3, and only two moles of AgCl appear. With CoCl3.4NH3, just one mole precipitates, even though a third chloride ion is still sitting somewhere in the formula. Something was holding that third chloride back from reacting, and whatever it was, it wasn't obvious from the formula alone.

Werner's explanation, published in 1898, proposed that a metal ion has two distinct kinds of valence working at once. A primary valence is ionisable, satisfied by negative ions, and corresponds to the metal's ordinary oxidation state. A secondary valence is non-ionisable, fixed in number for a given metal, and satisfied by neutral molecules or negative ions bound directly to it, what we'd now call the coordination number. Crucially, Werner proposed that the atoms held by secondary valence occupy a fixed, definite spatial arrangement around the metal, a coordination polyhedron, most commonly octahedral, tetrahedral, or square planar. Everything inside that arrangement, written inside square brackets, forms a single entity that does not break apart in solution. Only what's left outside the brackets, the counter ions, ionises freely. In [Co(NH3)6]3+ 3Cl-, all six ammonias and the cobalt form one unbreakable coordination entity, while the three chlorides outside are free counter ions, so all three precipitate as AgCl.

Three terms do most of the work in describing any coordination compound. The coordination entity is the central metal atom or ion together with the fixed number of ions or molecules directly bonded to it, for example [CoCl3(NH3)3], where cobalt is surrounded by three chloride ions and three ammonia molecules. The central atom or ion, also called a Lewis acid in this context, is whichever atom or ion sits at the middle of that arrangement, accepting electron pairs from everything around it. The ligands are those ions or molecules bonded directly to the central atom, ranging from simple ions like Cl- to small molecules like H2O or NH3 to much larger molecules, even entire proteins.

Ligands differ in how many points of contact they make with the central metal. A ligand bound through a single donor atom, like Cl-, H2O, or NH3, is unidentate. One that binds through two donor atoms at once, like ethane-1,2-diamine (en) or the oxalate ion C2O42-, is didentate. A ligand with several donor atoms available, like N(CH2CH2NH2)3, is polydentate, and the number of donor atoms actually used to bind one metal ion is called its denticity. EDTA4-, ethylenediaminetetraacetate, is a particularly important example: a hexadentate ligand that can grip a single metal ion through two nitrogen atoms and four oxygen atoms all at once. When a di- or polydentate ligand wraps around and binds a single metal through two or more of its donor atoms simultaneously, the resulting ring-shaped structure is called a chelate complex, from the Greek for claw, and chelate complexes are consistently more stable than similar complexes built from unidentate ligands alone.

A small number of ligands have two different atoms that could each act as the donor, and which one actually bonds to the metal depends on the specific complex. These are called ambidentate ligands. The nitrite ion, NO2-, can bond to a metal either through its nitrogen atom or through one of its oxygen atoms. Jorgensen discovered exactly this behaviour in the complex [Co(NH3)5(NO2)]Cl2: bound through oxygen (as -ONO), the compound is red; bound through nitrogen (as -NO2), it's yellow, two genuinely different, isolable compounds from the same formula. The thiocyanate ion, SCN-, shows the same behaviour, bonding through sulphur (M-SCN) or through nitrogen (M-NCS).

The coordination number of a metal ion is the number of ligand donor atoms directly bonded to it, counting only sigma bonds, never pi bonds. In [PtCl6]2-, platinum's coordination number is 6, one for each of the six chloride ligands. In [Co(en)3]3+, cobalt's coordination number is also 6, but from only three ligand molecules, since ethane-1,2-diamine is didentate and contributes two donor atoms each. The coordination sphere is the central atom together with its ligands, everything written inside the square brackets; whatever ionisable groups sit outside those brackets are counter ions. And the oxidation number of the central atom is defined as the charge it would carry if every ligand were removed along with the electron pairs it had been sharing, written as a Roman numeral immediately after the coordination entity's name, as in cobalt(III) or platinum(II).

Coordination compound formulas follow a strict, unambiguous set of rules. The central atom is always listed first. The ligands then follow in alphabetical order, regardless of each ligand's own charge, and polydentate or abbreviated ligands are alphabetised the same way, by their own first letter. The entire coordination entity, whether it carries a charge or not, sits inside square brackets, with any polyatomic ligand's own formula enclosed in parentheses. No spaces appear anywhere inside the coordination sphere. And if a charged coordination entity is written without its counter ion, that charge appears outside the bracket as a right superscript, the number written before the sign, as in [Co(CN)6]3-.

Naming reverses part of the formula-writing logic. The cation is always named first, whether the overall complex is a cation or an anion. Ligands are named in alphabetical order before the central atom's own name, exactly the opposite order from how they're written in the formula. Anionic ligands take the ending '-o' (chlorido, cyanido), while neutral ligands mostly keep special names, aqua for water, ammine for ammonia, carbonyl for CO. Numerical prefixes (di, tri, tetra) count simple ligands, while bis, tris, tetrakis are used instead when the ligand's own name already contains a numerical term, with that ligand's name placed in parentheses. If the complex ion is a cation, the metal keeps its ordinary name; if it's an anion, the metal's name takes the suffix '-ate' instead. Consider [Cr(NH3)3(H2O)3]Cl3: named triamminetriaquachromium(III) chloride. Ammine is alphabetised before aqua, three chlorides outside the bracket mean the complex carries a +3 charge, and since every ligand here is neutral, chromium's own oxidation state must also be +3.

Isomers share an identical chemical formula but differ in how their atoms are arranged, and that difference alone can change their physical and chemical properties. Geometric isomerism arises in heteroleptic complexes from the different possible spatial arrangements of ligands. In a square planar complex of formula [MX2L2], the two X ligands can sit adjacent to each other, the cis isomer, or directly opposite each other, the trans isomer. The same possibility exists in octahedral complexes of formula [MX2L4], and when three donor atoms of one kind occupy adjacent corners of the octahedron rather than two opposite pairs, a related pattern called the facial (fac) isomer appears, versus the meridional (mer) isomer when they run around the octahedron's middle. Tetrahedral complexes, by contrast, never show this kind of isomerism, since every ligand position around a tetrahedral centre sits in an identical relationship to every other position.

Optical isomers are mirror images of each other that can never be superimposed, no matter how they're rotated, called enantiomers, and molecules or ions with this property are described as chiral. The two forms are labelled dextro (d) and laevo (l), for the direction each one rotates the plane of polarised light passing through it in a polarimeter. Optical isomerism turns up commonly in octahedral complexes built from didentate ligands, such as [Co(en)3]3+. Four further kinds of isomerism arise not from spatial arrangement but from genuinely different bonds. Linkage isomerism comes from an ambidentate ligand bonding through a different atom, as already seen with nitrite. Coordination isomerism arises when ligands swap between the cationic and anionic parts of a salt built from two different metals, as in [Co(NH3)6][Cr(CN)6] versus [Cr(NH3)6][Co(CN)6]. Ionisation isomerism happens when the counter ion itself is a potential ligand that trades places with one already inside the bracket, as in [Co(NH3)5Br]SO4 versus [Co(NH3)5SO4]Br, each giving a different precipitate test with barium or silver ions. And solvate isomerism, best known as hydrate isomerism when water is the solvent involved, differs only in whether a solvent molecule sits bonded directly to the metal or free in the crystal lattice, as in violet [Cr(H2O)6]Cl3 against grey-green [Cr(H2O)5Cl]Cl2.H2O.