Two ways the same parts can differ
By now you can read a complex like a sentence: the brackets mark the coordination sphere, the donor atoms give the coordination number, and the geometry follows. But that reading skips a quiet possibility. Two compounds can share the exact same formula — same metal, same ligands, same counter-ions, same atom count down to the last hydrogen — and still be two different substances, with different colours, solubilities, and reactions. These are isomers, and coordination chemistry produces them in a variety that organic chemistry can only envy. The reason is structural: a metal centre offers many slots arranged in three dimensions, and ligands can occupy them, swap with counter-ions, or even bond through a different atom — each move generating a new compound from the same inventory of parts.
The whole zoo splits cleanly into two families, and the dividing question is simple: which atom is bonded to which? In a structural isomer, the connectivity itself differs — a chloride that was a ligand becomes a free counter-ion, or a nitrite ligand grabs the metal by its nitrogen in one isomer and by its oxygen in the other. Break the bonds and the atom-to-atom map is not the same. In a stereoisomer, the connectivity is identical — every atom is bonded to exactly the same neighbours — and only the arrangement in space differs: which slot points where. The two cautions to keep are that this is a model that ignores how fast the molecules might interconvert, and that the labels describe a snapshot of bonding, not the smeared, partly covalent reality the later crystal-field guides will refine.
Structural isomers: when the connectivity moves
[[structural-isomerism-complexes|Structural isomerism]] comes in four flavours that you can tell apart by asking what crossed the boundary of the square brackets — the line between Werner's inner sphere and the free counter-ions outside. The cleanest is ionization isomerism: two compounds swap which anion sits inside the coordination sphere and which floats outside as a counter-ion. The classic pair is [Co(NH3)5Br]SO4 and [Co(NH3)5SO4]Br. They have the identical formula, yet the first holds bromide tightly to the cobalt and lets sulfate roam free, while the second does the reverse. Drop silver nitrate into each: the first gives no precipitate of AgBr because the bromide is locked inside, but adding barium chloride throws down white BaSO4 from the free sulfate; the second does exactly the opposite. The formula is one thing; what is bonded is another.
A close cousin is hydrate (or solvate) isomerism, where water is the molecule that crosses the line — bound as a ligand in one isomer, sitting in the crystal as free water of crystallisation in the other. Chromium(III) chloride hexahydrate, formula CrCl3·6H2O, is the textbook trio: violet [Cr(H2O)6]Cl3 with all six waters bonded and three chlorides free; pale green [Cr(H2O)5Cl]Cl2·H2O with one chloride pulled into the sphere; and dark green [Cr(H2O)4Cl2]Cl·2H2O with two. Same bottle label, three genuinely different solids of three different colours, and you can again tell them apart by how many chlorides drop out as AgCl with silver. Third is coordination isomerism, possible only when both the cation and the anion of a salt are themselves complex ions: the ligands can be redistributed between the two metals. [Co(NH3)6][Cr(CN)6] and [Cr(NH3)6][Co(CN)6] are coordination isomers — the ammonias and cyanides simply trade which metal they surround.
The fourth and most subtle is linkage isomerism, and it needs a special kind of ligand to occur. Most ligands always reach in through the same donor atom — water through its oxygen, ammonia through its nitrogen. But an [[ambidentate-ligand|ambidentate ligand]] carries two different donor atoms and can bond through either one. The nitrite ion, NO2-, is the classic example: bind the metal through nitrogen and it is the yellow 'nitro' isomer, written -NO2; let it swing around and bind through an oxygen and it is the red 'nitrito' isomer, written -ONO. Jørgensen and Werner themselves fought over a pair of these cobalt complexes. Thiocyanate, SCN-, is another: soft metals tend to grab the soft sulfur end, hard metals the harder nitrogen end — a direct payoff of the hard-soft acid-base ideas from the earlier rung. Here the connectivity changes at a single atom, yet the two isomers can differ visibly in colour.
Stereoisomers: same bonds, different shape
Now keep every bond exactly where it is and only rearrange the slots in space. The first kind of [[geometric-isomerism-complexes|geometric isomerism]] is the cis–trans distinction, and the cleanest case is a square-planar complex of the type MA2B2 — a flat metal with two of one ligand and two of another at the four corners of a square. The two A ligands can sit on adjacent corners (90 degrees apart, the cis isomer) or on opposite corners (180 degrees apart, the trans isomer). These are not the same molecule rotated; no turn in space superimposes adjacent onto opposite. The textbook pair is [Pt(NH3)2Cl2]: cis has its two chlorides next to each other, trans has them across the square. Octahedral MA4B2 complexes do the same — the two B ligands either share an edge (cis) or sit at the poles (trans) — and the colours, dipole moments, and reactivities of the cis and trans forms differ measurably.
Tetrahedral complexes, by contrast, show no cis–trans isomerism at all: in a tetrahedron every corner is adjacent to every other corner, so there is no 'opposite' position to distinguish — a useful reminder that geometry, not just formula, decides what isomers are even possible. For octahedral complexes of the type MA3B3 — three of one ligand and three of another — a second flavour of geometric isomerism appears: facial (fac) versus meridional (mer). In the fac isomer the three identical ligands occupy one triangular face of the octahedron, all mutually cis, capping one side. In the mer isomer they string out around a meridian — a great circle through the metal — with two of them trans to each other and one off to the side. Picture three corners of a single face lit up versus three corners tracing a bent line from pole to equator to pole.
octahedral MA3B3 (positions: top, bottom, and 4 around the equator)
fac (facial) mer (meridional)
A A
| |
B - M - A <- one face B - M - A <- A's trace a meridian:
| is all A | top-A, equator-A, and
B (+ A below) B (+ B below) one more A in plane
three A on one triangular two A trans to each other,
face, all mutually cis the third A cis to bothOptical isomers: a complex and its mirror
The most striking stereoisomers are mirror images that cannot be superimposed — like your left and right hands. A molecule that is non-superimposable on its mirror image is chiral, and the two mirror-image forms are [[optical-isomerism-complexes|optical isomers]] or enantiomers. They share every ordinary property — same melting point, same colour, same solubility — and differ in just one measurable way: they rotate the plane of polarised light by equal amounts in opposite directions. (As the symmetry rung put it, a molecule is chiral exactly when it has no mirror plane and no centre of inversion.) This is the very phenomenon Werner exploited in 1911: by resolving a fully inorganic, carbon-free complex into its two mirror-image forms, he proved that the six positions really splay out as an octahedron, because only a three-dimensional arrangement can be handed.
Chirality in complexes most often comes from chelating ligands — the [[inorg-chelate|chelate]] rings of the previous guide. A complex like [Co(en)3]3+, with three ethylenediamine (en) ligands each biting the cobalt at two adjacent sites, wraps three rings around the octahedron in a propeller-like twist. That propeller can turn left-handed or right-handed, and the two windings are non-superimposable mirror images, labelled the lambda and delta forms. A complex with two chelate rings and two other ligands, like cis-[Co(en)2Cl2]+, is likewise chiral — but only the cis form: the trans isomer happens to gain a mirror plane and so is achiral. This is the satisfying payoff of stacking the ideas: whether a complex is optically active depends jointly on its geometry, its chelation, and even its cis–trans arrangement.
Why a rearrangement can change everything
Isomerism is not a bookkeeping curiosity; it can decide whether a compound works or fails. The sharpest example is the anticancer drug cisplatin, which is exactly the cis isomer of [Pt(NH3)2Cl2] you met above. Cis-platin's two chlorides sit on adjacent corners, the right distance apart to let the platinum bridge two neighbouring sites on a strand of DNA — a binding that kinks the DNA and stops a cancer cell dividing. The trans isomer, transplatin, has the identical formula and the identical atoms bonded to the identical platinum, yet its chlorides point in opposite directions, cannot reach the same pair of DNA sites, and is clinically useless as a drug. Same parts, one geometric flip, the difference between a life-saving medicine and an inert compound.
The same lesson runs through biology and materials. Living systems are built almost entirely from one handedness of their molecules, so a chiral metal complex meant to interact with a protein or with DNA may bind well as one enantiomer and barely at all as its mirror image — the binding pocket, itself chiral, can tell left from right by touch. Geometric isomers differ in measurable physical properties too: the cis isomer of a square-planar complex generally has a nonzero [[inorg-dipole-moment|dipole moment]] because its like ligands are bunched to one side, while the trans isomer's symmetric arrangement often cancels to zero — a fact you can confirm in the lab and use to assign which isomer you are holding. Colour, magnetic behaviour, and rate of reaction can all shift between isomers because they all trace back to the precise three-dimensional environment of the metal's d electrons, the very thing the next rung on crystal field theory will dissect.
Two honest closing cautions. First, whether two isomers can actually be kept in separate bottles depends on how fast they interconvert — a thermodynamic question of which is more stable is entirely separate from the kinetic question of whether they swap too quickly to isolate. Many tetrahedral and some octahedral complexes scramble their ligands in fractions of a second, so the isomers exist on paper but never in a flask; others, like Werner's cobalt(III) complexes, are so sluggish that the isomers sit unchanged for years. Stability and lability are independent ideas, and a later guide makes that distinction central. Second, the neat geometric pictures here are idealised: real octahedra are often slightly distorted, and the crystal-field guides ahead will show you why some are squashed or stretched on purpose. The isomer labels remain a faithful and powerful map — just remember they are a map, not the smeared electronic territory itself.