Coordination Compounds: Fundamental Concepts

Coordination chemistry, founded on exactly the concepts above, underlies enormous swaths of modern chemistry and biology. In medicine, cisplatin, cis-[PtCl₂(NH₃)₂] — a simple square-planar Pt(II) coordination compound with CN = 4 — remains one of the most widely used chemotherapy agents, its mechanism depending entirely on which chloride ligands are labile (replaceable, inside the coordination sphere) versus the ammine ligands that remain inert and are essential for binding DNA guanine bases after aquation. In biology, the iron-porphyrin coordination sphere in Hemoglobin and Oxygen Transport reversibly adds O₂ as a sixth ligand to five-coordinate heme iron, precisely the CN = 5 → 6 transition that coordination theory predicts and quantifies; cobalt's coordination sphere in vitamin B₁₂ features an unusual Co–C coordinate bond, one of the few organometallic bonds in biology. In industry, homogeneous catalysis depends on metals cycling between different coordination numbers and oxidation states — Wilkinson's catalyst (Example 2 above) and the Rh- and Ir-based catalysts of the Monsanto and Cativa acetic-acid processes both operate by ligand association/dissociation steps that change coordination number by exactly one unit, a direct extension of the CN bookkeeping introduced here. In materials science, coordination polymers and metal–organic frameworks (MOFs) are literally extended coordination spheres, in which multitopic (multidentate) organic ligands bridge many metal centers into three-dimensional, porous, and often catalytically or gas-storage-active lattices. Analytical chemistry exploits selective, kinetically stable coordination complexes (EDTA titrations, colorimetric metal indicators) built on precisely the same charge-balance and denticity logic developed above. Every one of these applications begins from the same starting definitions: a central metal Lewis acid, a set of ligand Lewis bases, a fixed coordination sphere, and a coordinate covalent bond holding it all together — concepts that extend naturally into Ligands and Denticity and Coordination Nomenclature.

:::quiz[In [Pt(NH₃)₄]Cl₂, what is the charge of the coordination sphere (complex ion) and how many free Cl⁻ ions exist per formula unit?]

• +2 charge, 2 free Cl⁻ ✓
• 0 charge, 2 free Cl⁻
• +2 charge, 0 free Cl⁻
• −2 charge, 2 free Cl⁻

:::quiz[Werner's resolution of [Co(en)₃]Cl₃ into optically active isomers was historically significant because it]

• proved that ethylenediamine itself is chiral
• was the first resolution of any optically active compound in chemistry
• demonstrated three-dimensional octahedral geometry around the metal even with no carbon stereocenter present ✓
• showed that cobalt complexes never exhibit isomerism

:::faq[Is the coordinate bond in a complex like [Co(NH₃)₆]³⁺ fundamentally different from an ordinary covalent bond in an organic molecule?] No — once formed, it is populated and behaves like any other two-electron covalent bond; the only distinguishing feature is the formal electron origin before bonding (both electrons nominally from the ligand lone pair rather than one from each atom). This is confirmed by a simple two-orbital MO treatment: the bonding molecular orbital that results is doubly occupied regardless of whether the two electrons started as "1 + 1" (ordinary covalent) or "2 + 0" (coordinate/dative), and bond strengths for metal-ligand coordinate bonds are entirely comparable to typical single covalent bond enthalpies elsewhere in chemistry.

Ask the tutor: In [Pt(en)₂Cl₂]Cl₂, what is the oxidation state and coordination number of platinum, and how many free chloride ions should precipitate instantly with AgNO₃?

Picture a bare transition-metal ion in the gas phase, say Co³⁺. Its electron configuration is [Ar]3d⁶: six electrons in five d orbitals, and — crucially — completely empty 4s and 4p orbitals, plus (in the valence-bond/crystal-field picture used for six-coordination) two more empty d orbitals become available as hybrid acceptor orbitals. This ion is a potent Lewis acid: small, highly charged, with low-lying empty orbitals that are strongly attracted to electron density. Surround it with six ammonia molecules, each carrying a nonbonding lone pair on nitrogen. Each NH₃ lone pair points its electron density directly at the metal, forming a two-electron, two-center bond in which both electrons come from nitrogen — this is precisely the coordinate (dative) bond of Lewis acid–base theory extended to metals. The six Co–N bonds that result are geometrically fixed: they point along ±x, ±y, ±z, giving octahedral geometry, because this arrangement lets six identical ligands maximize their mutual separation (minimizing electron-pair repulsion) while symmetrically filling the metal's six accepting orbital lobes (one s + three p + two d, i.e., d²sp³ hybrids in valence-bond language, or equivalently the eg-symmetry and a1g/t1u metal orbitals in molecular-orbital language). This fixed geometric unit, [Co(NH₃)₆]³⁺, is the coordination sphere. It carries an overall charge of +3 (Co³⁺ plus six neutral NH₃). In the solid salt [Co(NH₃)₆]Cl₃, the three chloride ions balance that charge but sit outside the coordination sphere, in the crystal lattice, held only by ordinary ionic (electrostatic) forces — not coordinate bonds. That is why they behave as free Cl⁻ in solution, immediately precipitated by Ag⁺, while the six NH₃ ligands remain firmly attached to cobalt and are not detected as free NH₃ by any simple chemical test. Contrast this with [CoCl(NH₃)₅]Cl₂. Now one chloride has moved inside the coordination sphere, directly bonded to cobalt, and only two chloride ions remain as free counter-ions. Conductivity measurements and precipitation tests immediately distinguish these two compounds even though both have the empirical formula "CoCl₃·5NH₃" (for [CoCl(NH₃)₅]Cl₂) or "CoCl₃·6NH₃" (for [Co(NH₃)₆]Cl₃) — this was exactly the kind of puzzle that ordinary 19th-century "double salt" thinking could not resolve, and that Werner's coordination theory explained by distinguishing the two kinds of bonding.

:::svg[The coordination sphere: metal, ligands, and counter-ions in [Co(NH₃)₆]Cl₃]

::: The bonding picture generalizes. Any species with an accessible lone pair or π system — halide ions, water, ammonia, cyanide, carbon monoxide, phosphines, alkenes — can act as a ligand toward a metal center with empty or partially filled orbitals of appropriate symmetry and energy. Ligand donation is not restricted to simple lone pairs: CO donates through a σ lone pair on carbon and accepts electron density back from filled metal d-orbitals into its empty π* orbital (synergic "back-bonding," central to organometallic chemistry, e.g. in Ni(CO)₄ or Fe(CO)₅). But the foundational, oldest, and pedagogically clearest case — Werner's ammine and aqua complexes — is a straightforward σ-donor coordinate bond, and it is the model on which the whole vocabulary of coordination chemistry is built. A second essential picture is geometric. The coordination number (CN) — the number of donor atoms directly attached to the metal — dictates a preferred geometry through a balance of electron-pair repulsion (VSEPR-like reasoning, though ligand sterics and metal d-orbital occupancy matter more here than for main-group VSEPR) and crystal-field/ligand-field stabilization. CN = 6 strongly favors octahedral; CN = 4 splits between tetrahedral (favored for d⁰, d⁵ high-spin, d¹⁰, and for bulky ligands) and square planar (favored for d⁸ ions such as Pt²⁺, Pd²⁺, Ni²⁺ with strong-field ligands, where crystal-field stabilization strongly favors the square-planar arrangement); CN = 2 gives linear geometry (classic for d¹⁰ Ag⁺ complexes like [Ag(NH₃)₂]⁺); CN = 5 gives trigonal bipyramidal or square pyramidal; CN = 7, 8, and 9 occur for larger metal ions, especially lanthanides and actinides.

The essential relationships of coordination chemistry can be stated compactly. First, the charge-balance relationship that fixes the overall charge of the coordination sphere (the complex ion):

where: $Q_{\text{complex}}$ is the net charge on the coordination sphere (complex ion), $z_{\text{metal}}$ is the oxidation state of the central metal, and $q_i$ is the charge of the $i$-th ligand (0 for a neutral ligand such as NH₃ or H₂O, −1 for Cl⁻, CN⁻, OH⁻, etc.). Reading: the metal's oxidation state and each ligand's own charge simply add up; this is nothing more than bookkeeping of formal charges, but it is the bookkeeping that lets you predict whether a given complex is cationic, anionic, or neutral, and hence how many counter-ions are needed for electroneutrality in the solid salt. Second, the electroneutrality condition for the whole salt:

where: $n$ is the number of counter-ions of charge $q_{\text{counter-ion}}$ required to balance the complex ion's charge, so that the bulk salt is electrically neutral. Reading: this is exactly ordinary ionic-compound stoichiometry (as in NaCl or CaCl₂), except that one of the "ions" is now a polyatomic coordination sphere rather than a simple monatomic ion. Third, Werner's coordination-number/geometry correspondence, stated as a rule rather than an equation — but quantifiable through the number of ligand-metal σ bonds:

where: CN is the coordination number (total donor-atom attachments to the metal), and the denticity of ligand $i$ is the number of donor atoms that one ligand molecule uses to bind the same metal center (1 for monodentate NH₃ or Cl⁻, 2 for bidentate ethylenediamine, etc. — developed fully elsewhere). Reading: coordination number is not simply "how many ligands," but "how many donor-atom bonds"; a single bidentate ligand contributes 2 to CN even though it is one molecule. This distinction is essential once chelating ligands are introduced.

Example 1 — Foundational: identifying the coordination sphere and predicting ionic behavior. Consider the salt [Cr(H₂O)₄Cl₂]Cl, a well-known green isomer of "chromium(III) chloride hexahydrate." Step 1: Identify metal, ligands, and coordination sphere. The bracket contains Cr³⁺ as the central metal, four H₂O and two Cl⁻ as ligands directly bonded to it. Coordination number: 4 (from H₂O) + 2 (from Cl⁻ ligands) = 6, consistent with octahedral geometry expected for Cr³⁺ (a d³ ion with strong preference for octahedral CN = 6). Step 2: Compute the charge of the complex ion using $Q_{\text{complex}} = z_{\text{metal}} + \sum q_i$.

Step 3: Apply electroneutrality to find the number of counter-ions. With $Q_{\text{complex}} = +1$ and counter-ion Cl⁻ ($q = -1$):

So exactly one Cl⁻ lies outside the bracket — matching the given formula [Cr(H₂O)₄Cl₂]Cl. Step 4: Predict experimental behavior. Only the one "free" chloride (outside the bracket) precipitates instantly with AgNO₃; the two chloride ligands inside the coordination sphere are covalently bound to Cr³⁺ and do not give an immediate precipitate (they can eventually exchange, but slowly, via ligand substitution kinetics — a separate topic from instantaneous ionic precipitation). Conductivity should correspond to 2 ions per formula unit ([Cr(H₂O)₄Cl₂]⁺ + Cl⁻), distinctly lower than for the "chromium(III) hexahydrate" formula [Cr(H₂O)₆]Cl₃, which would show 4 ions and 3 equivalents of instantly precipitated AgCl. Limiting-case check: if instead all six ligands were H₂O ([Cr(H₂O)₆]Cl₃), the same formula gives $Q = 3 + 6(0) = +3$, requiring $n = 3$ free Cl⁻ — reproducing the well-known violet isomer that gives 3 equivalents of AgCl immediately, exactly the opposite extreme of the green isomer computed above. The method correctly discriminates the two known hydration isomers of "CrCl₃·6H₂O" from the same charge-balance formula.

Example 2 — Applied: coordination number and geometry in a catalytically relevant complex. Wilkinson's catalyst, RhCl(PPh₃)₃, is a square-planar Rh(I) hydrogenation catalyst. Determine its coordination number, the oxidation state of rhodium, and rationalize its geometry. Step 1: Assign oxidation state. Treat Cl as Cl⁻ (charge −1) and PPh₃ as a neutral two-electron donor (charge 0, phosphorus lone pair donor, directly analogous to NH₃). For the neutral molecule, $Q_{\text{complex}} = 0$ (there is no counter-ion; the whole molecule is neutral), so:

Rhodium is Rh(I), i.e. formally $d^8$ (Rh has atomic configuration [Kr]4d⁸5s¹, so Rh⁺ is [Kr]4d⁸). Step 2: Determine coordination number. Four donor atoms are directly bonded: one Cl and three P (each PPh₃ is monodentate, denticity 1). By $\text{CN} = \sum(\text{denticity})$:

Step 3: Rationalize geometry. For CN = 4, a $d^8$ configuration strongly favors square-planar geometry over tetrahedral, because in an octahedral-derived crystal-field splitting, removing the two ligands along one axis to reach a square-planar limit drops one d-orbital ($d_{z^2}$-derived) sharply in energy and leaves the highest-energy orbital ($d_{x^2-y^2}$) empty and very high — for $d^8$, this configuration lets all eight d-electrons occupy the four lower-lying orbitals, leaving the highest orbital empty and unpopulated. This large crystal-field stabilization for $d^8$ square-planar complexes (the same effect that stabilizes Pt(II), Pd(II), Ni(II) with strong-field ligands) is why Rh(I), Pt(II), and related $d^8$ ions overwhelmingly adopt square-planar rather than tetrahedral geometry, in contrast to $d^0$ or $d^{10}$ CN=4 complexes, which show no such electronic preference and typically go tetrahedral instead (sterics permitting). Limiting-case check: if we instead examine [NiCl₄]²⁻, a $d^8$ system but with the weak-field Cl⁻ ligand and a small crystal-field splitting, the CFSE advantage of square-planar over tetrahedral is not large enough to overcome the electron-pair-repulsion (steric/electrostatic) preference for tetrahedral, so [NiCl4]2− is tetrahedral rather than square planar — showing that the $d^8$ square-planar preference above is real but is a tendency, not an absolute law; ligand field strength also matters and is treated fully in crystal-field theory.

Example 3 — Exam/competition level: deducing structure and CN from combined analytical and conductivity data. A cobalt(III) chloride–ammonia complex has the empirical formula CoCl₃·4NH₃. Aqueous solution conductivity corresponds to 2 ions per formula unit, and treatment with excess AgNO₃(aq) at room temperature immediately precipitates exactly 1 mole of AgCl per mole of complex. A separate sample, when boiled with excess NaOH, releases no ammonia gas (ruling out any hydrolyzable NH₄⁺ present as a simple ammonium salt contaminant). Determine the formula of the coordination sphere, the oxidation state of Co, its coordination number, and predict whether geometric isomers should exist. Step 1: Use the AgCl precipitation data to count free Cl⁻. 1 mole of AgCl precipitated per mole of complex means exactly 1 Cl⁻ is a free counter-ion, so 2 of the total 3 chlorides must be ligands (bound inside the coordination sphere, non-ionizable on this timescale). Step 2: Use the conductivity data as a cross-check. 2 ions per formula unit is consistent with a 1:1 electrolyte: one complex cation of charge +1 and one free Cl⁻. This matches Step 1 exactly (1 free Cl⁻ implies the complex cation has charge +1, and a 1:1 salt gives exactly 2 ions). Step 3: Assign oxidation state using charge balance. The complex cation is [CoCl₂(NH₃)₄]⁺. Using $Q_{\text{complex}} = z_{\text{Co}} + \sum q_i$:

Cobalt is Co(III), $d^6$, exactly as expected (consistent with the parent [Co(NH3)6]3+ series and with Co(III)'s well-known kinetic inertness, which is exactly why two of the three chlorides survive as non-ionizing ligands rather than exchanging or dissociating on the timescale of the AgNO₃ test). Step 4: Determine coordination number. Ligands directly bound: 4 NH₃ (monodentate) + 2 Cl⁻ (monodentate) = 6 donor atoms, so:

consistent with the strong preference of Co³⁺ ($d^6$, octahedral crystal-field stabilization is maximal for low-spin $d^6$) for six-coordinate octahedral geometry. Step 5: Predict isomerism. An octahedral MX₂Y₄ complex (here M = Co, X = Cl, Y = NH₃) has two possible geometric arrangements of the two Cl ligands: cis (Cl–Co–Cl angle 90°, the two chlorides on adjacent vertices) and trans (Cl–Co–Cl angle 180°, the two chlorides on opposite vertices). Both isomers of [CoCl₂(NH₃)₄]Cl are known experimentally (the trans isomer is green, the cis isomer is violet/purple), confirming octahedral CN = 6 geometry rather than, say, a trigonal-prismatic or other alternative arrangement, and this cis/trans pair was in fact one of Werner's key pieces of evidence for octahedral (rather than, e.g., hexagonal-planar) geometry around six-coordinate metal centers — a hexagonal-planar arrangement would have allowed three distinct MX₂Y₄ isomers, not two, so the experimental observation of exactly two isomers rules it out. Limiting-case check: if instead all 3 chlorides had precipitated instantly (3 mol AgCl per mole complex) and conductivity corresponded to 4 ions per formula unit, the analysis above would force $z_{\text{Co}}=+3$ with CN = 6 filled entirely by 6 NH₃, i.e. [Co(NH3)6]Cl3 — the original hexammine, with no possibility of cis/trans isomerism at all (all six ligands identical), correctly recovering the simplest member of this series as the boundary case.