Coordination Compounds

Alfred Werner (Nobel Prize, 1913) worked out the structure of coordination compounds from a very simple experiment. Cobalt(III) chloride combines with ammonia to give a series of compounds of formula CoCl₃·xNH₃. When excess AgNO₃ is added to their solutions, the amount of AgCl precipitated is different for each: CoCl₃·6NH₃ (yellow) gives 3 mol AgCl, CoCl₃·5NH₃ (purple) gives 2 mol, CoCl₃·4NH₃ (green, and a violet form) gives 1 mol, and CoCl₃·3NH₃ gives none. Molar conductance falls in the same order. So only some of the chloride is free in solution; the rest is held tightly to the cobalt along with the ammonia.

Werner's postulates explain this. (i) In coordination compounds a metal shows two types of linkage, primary and secondary. (ii) The primary valences are normally ionisable and are satisfied by negative ions. (iii) The secondary valences are non-ionisable, are satisfied by neutral molecules or negative ions, and their number, the coordination number, is fixed for a given metal. (iv) The groups bound by secondary linkages have a definite spatial arrangement, which gives the complex its geometry. For cobalt the secondary valence is six, so the four compounds are [Co(NH₃)₆]Cl₃, [Co(NH₃)₅Cl]Cl₂, [Co(NH₃)₄Cl₂]Cl and [Co(NH₃)₃Cl₃]. Werner also proposed that coordination number 6 gives an octahedral arrangement and coordination number 4 gives tetrahedral or square planar.

The board likes the difference between a double salt and a complex. Both are formed from two or more stable compounds in stoichiometric ratio. A double salt such as carnallite KCl·MgCl₂·6H₂O, Mohr's salt FeSO₄·(NH₄)₂SO₄·6H₂O or potash alum K₂SO₄·Al₂(SO₄)₃·24H₂O dissociates completely into its simple ions in water. A complex such as K₄[Fe(CN)₆] does not give Fe²⁺ and CN⁻; the complex ion [Fe(CN)₆]⁴⁻ keeps its identity in solution.

A coordination entity is a central metal atom or ion bonded to a fixed number of ions or molecules, for example [CoCl₃(NH₃)₃], [Ni(CO)₄], [PtCl₂(NH₃)₂], [Fe(CN)₆]⁴⁻. The central atom or ion is the one to which the ligands are bound; it is a Lewis acid. Ligands are the ions or molecules bound to it; each ligand donates a lone pair, so it is a Lewis base.

Ligands are classified by the number of donor atoms they use. A unidentate ligand binds through one atom: Cl⁻, H₂O, NH₃. A didentate ligand binds through two: ethane-1,2-diamine (en, H₂NCH₂CH₂NH₂) and oxalate C₂O₄²⁻. A polydentate ligand binds through several: EDTA⁴⁻ is hexadentate, binding through two nitrogen and four oxygen atoms. When a di- or polydentate ligand uses two or more donor atoms on the same metal it forms a ring, called a chelate, and the ligand is a chelating ligand. The number of ligating groups is the denticity. Chelate complexes are more stable than those with the equivalent number of unidentate ligands; this extra stability is the chelate effect. An ambidentate ligand has two different donor atoms but uses only one at a time: NO₂⁻ binds through N (nitrito-N) or through O (nitrito-O); SCN⁻ binds through S (thiocyanato-S) or N (thiocyanato-N).

Coordination number is the number of ligand donor atoms directly bonded to the metal: 6 in [PtCl₆]²⁻, 4 in [Ni(NH₃)₄]²⁺, 6 in [Fe(C₂O₄)₃]³⁻ and [Co(en)₃]³⁺ because each didentate ligand counts twice. Only σ bonds are counted. The metal with its ligands inside the square bracket is the coordination sphere; the ions outside it are counter ions. The spatial arrangement of the ligand atoms around the metal is the coordination polyhedron, commonly octahedral, square planar or tetrahedral. The oxidation number of the central atom is the charge it would carry if all ligands were removed with their electron pairs; it is written in Roman numerals in parentheses, as in [Cu(CN)₄]³⁻, Cu(I). A homoleptic complex has only one kind of ligand, [Co(NH₃)₆]³⁺; a heteroleptic complex has more than one kind, [Co(NH₃)₄Cl₂]⁺.

Writing the formula: the central atom is listed first, then the ligands in alphabetical order regardless of charge (polydentate and abbreviated ligands are also placed alphabetically, by the first letter of the abbreviation). The whole coordination entity goes in square brackets, polyatomic ligands in parentheses, no spaces between ligand and metal. The charge on a complex ion is written outside the bracket as a superscript with the number before the sign, [Co(CN)₆]³⁻; for a salt the cation is written first and the counter ion balances the charge.

Naming: (1) the cation is named first, then the anion, whichever is the complex. (2) Ligands are named in alphabetical order before the name of the metal, ignoring numerical prefixes. (3) Anionic ligands end in -o: chlorido, bromido, cyanido, hydroxido, oxalato, sulphato, carbonato, nitrito-N, nitrito-O, thiocyanato-S. Neutral ligands keep their name (methylamine, pyridine, ethane-1,2-diamine) except aqua for H₂O, ammine for NH₃, carbonyl for CO and nitrosyl for NO. (4) Numbers of ligands are shown by di, tri, tetra, penta, hexa; if the ligand name itself contains a numerical prefix or is complicated, use bis, tris, tetrakis with the ligand name in parentheses: bis(ethane-1,2-diamine), tris(oxalato) is also accepted as trioxalato. (5) The oxidation state of the metal follows its name in Roman numerals in parentheses. (6) If the complex is an anion the metal name ends in -ate: ferrate, cobaltate, nickelate, cuprate, chromate, platinate, zincate, argentate, aurate. In a cationic or neutral complex the metal name is unchanged.

Worked examples the board uses: [Cr(NH₃)₃(H₂O)₃]Cl₃ triamminetriaquachromium(III) chloride; [Co(H₂NCH₂CH₂NH₂)₃]₂(SO₄)₃ tris(ethane-1,2-diamine)cobalt(III) sulphate; [Ag(NH₃)₂][Ag(CN)₂] diamminesilver(I) dicyanidoargentate(I); K₃[Fe(CN)₆] potassium hexacyanidoferrate(III); [Ni(CO)₄] tetracarbonylnickel(0); [Pt(NH₃)₂Cl(NO₂)] diamminechloridonitrito-N-platinum(II); [Co(NH₃)₄Cl(NO₂)]Cl tetraamminechloridonitrito-N-cobalt(III) chloride; [Pt(NH₃)₂Cl(NH₂CH₃)]Cl diamminechlorido(methylamine)platinum(II) chloride; K₂[PdCl₄] potassium tetrachloridopalladate(II).

Isomers have the same formula but different arrangement of atoms. Stereoisomers have the same bonds but different spatial arrangement; structural isomers have different bonds. Stereoisomerism is of two kinds.

Geometrical isomerism arises in heteroleptic complexes because ligands can be next to each other (cis) or opposite (trans). In square planar [Ma₂b₂] there are cis and trans forms, the classic being [Pt(NH₃)₂Cl₂]. Square planar [Mabcd] gives three isomers, for example [Pt(NH₃)(Br)(Cl)(py)], because each ligand can be trans to any one of the other three. In octahedral [Ma₄b₂] there are cis and trans forms, [Co(NH₃)₄Cl₂]⁺; the same holds for [M(aa)₂b₂] with didentate aa, [CoCl₂(en)₂]⁺. Octahedral [Ma₃b₃] gives the facial (fac) isomer, three identical ligands on one triangular face, and the meridional (mer) isomer, three identical ligands around a meridian; [Co(NH₃)₃(NO₂)₃] is the NCERT example. Tetrahedral complexes show no geometrical isomerism because all four positions are adjacent to one another.

Optical isomerism: two isomers that are non-superimposable mirror images, called dextro (d) and laevo (l), rotate plane polarised light in opposite directions. It is common in octahedral complexes with didentate ligands. [Co(en)₃]³⁺ and [Cr(C₂O₄)₃]³⁻ are optically active. In [PtCl₂(en)₂]²⁺ and [CoCl₂(en)₂]⁺ only the cis isomer is optically active; the trans isomer has a plane of symmetry and is inactive. So [CoCl₂(en)₂]⁺ has three isomers in all: trans, and the d and l forms of cis. This three-isomer count is the most repeated drawing question from the chapter.

Linkage isomerism arises with an ambidentate ligand. The NO₂⁻ ligand in [Co(NH₃)₅(NO₂)]Cl₂ can bind through nitrogen, giving the yellow nitrito-N complex, or through oxygen, giving the red nitrito-O complex [Co(NH₃)₅(ONO)]Cl₂. SCN⁻ behaves the same way through S or N.

Coordination isomerism arises when both cation and anion are complexes and the ligands are interchanged between the two metals: [Co(NH₃)₆][Cr(CN)₆] and [Cr(NH₃)₆][Co(CN)₆].

Ionisation isomerism arises when a counter ion is itself a potential ligand and can swap places with a ligand inside the sphere: [Co(NH₃)₅(SO₄)]Br and [Co(NH₃)₅Br]SO₄. They give different ions in solution, and that is the test the board asks for: [Co(NH₃)₅Br]SO₄ gives a white precipitate of BaSO₄ with BaCl₂ but nothing with AgNO₃; [Co(NH₃)₅(SO₄)]Br gives a pale yellow precipitate of AgBr with AgNO₃ but nothing with BaCl₂. For [Co(NH₃)₅Cl]SO₄ and [Co(NH₃)₅(SO₄)]Cl the same logic works with BaCl₂ and AgNO₃.

Solvate (hydrate) isomerism is ionisation isomerism where the ligand that swaps is the solvent, usually water. CrCl₃·6H₂O exists as violet [Cr(H₂O)₆]Cl₃, grey-green [Cr(H₂O)₅Cl]Cl₂·H₂O and dark green [Cr(H₂O)₄Cl₂]Cl·2H₂O. They are distinguished by the moles of AgCl precipitated per mole (3, 2 and 1) and by the number of water molecules lost over concentrated H₂SO₄ (0, 1 and 2 respectively). A common 3-mark question gives one of these observations and asks for the structural formula and IUPAC name.

According to VBT the metal atom or ion, under the influence of ligands, uses its (n−1)d, ns, np or ns, np, nd orbitals to form a set of equivalent hybrid orbitals, which overlap with the filled orbitals of the ligands. The hybridisation fixes the geometry: coordination number 4 gives sp³ (tetrahedral) or dsp² (square planar); 5 gives sp³d (trigonal bipyramidal); 6 gives d²sp³ (octahedral, inner orbital, uses the inner (n−1)d orbitals) or sp³d² (octahedral, outer orbital, uses the outer nd orbitals). Strong field ligands such as CN⁻, CO and NH₃ (for Co³⁺) force pairing of the metal d electrons so that inner d orbitals become vacant; weak field ligands such as F⁻, Cl⁻ and H₂O do not.

The examples to reproduce with orbital diagrams: [Co(NH₃)₆]³⁺, Co³⁺ is 3d⁶; NH₃ pairs the electrons into three 3d orbitals, two 3d, one 4s and three 4p form d²sp³: inner orbital, diamagnetic. [CoF₆]³⁻: no pairing, sp³d², outer orbital, four unpaired electrons, paramagnetic. [Ni(CN)₄]²⁻, Ni²⁺ is 3d⁸; CN⁻ pairs the electrons into four 3d orbitals, leaving one 3d empty for dsp²: square planar, diamagnetic. [NiCl₄]²⁻: no pairing, sp³, tetrahedral, two unpaired electrons, paramagnetic. [Ni(CO)₄]: Ni(0) is 3d⁸4s²; CO pushes the 4s electrons into 3d giving 3d¹⁰, sp³, tetrahedral, diamagnetic. [Fe(CN)₆]⁴⁻, Fe²⁺ d⁶, d²sp³, diamagnetic; [Fe(CN)₆]³⁻, Fe³⁺ d⁵, d²sp³, one unpaired electron, weakly paramagnetic; [FeF₆]³⁻ and [Fe(H₂O)₆]³⁺, sp³d², five unpaired electrons, strongly paramagnetic. [Ni(NH₃)₆]²⁺ is outer orbital (sp³d²) because d⁸ cannot free two inner d orbitals even with pairing.

Magnetic moment is measured and compared with μ = √(n(n+2)) BM. For d¹, d², d³ the number of unpaired electrons is the same in the free ion and the octahedral complex. For d⁴ to d⁶ the result depends on the ligand: [Mn(CN)₆]³⁻ has two unpaired electrons while [MnCl₆]³⁻ has four; [Fe(CN)₆]³⁻ has one while [FeF₆]³⁻ has five; [CoF₆]³⁻ has four while [Co(C₂O₄)₃]³⁻ is diamagnetic. VBT explains this as inner versus outer orbital complexes.

Limitations of VBT: it involves a number of assumptions; it gives no quantitative interpretation of magnetic data; it does not explain the colour of complexes; it gives no quantitative interpretation of thermodynamic or kinetic stability; it cannot predict whether a 4-coordinate complex will be tetrahedral or square planar; and it does not distinguish weak and strong field ligands.

CFT treats the metal–ligand bond as purely ionic: the ligands are point charges (anions) or point dipoles (polar molecules). In a free ion the five d orbitals are degenerate. When ligands approach, the electrons in orbitals pointing at the ligands are repelled more than those in orbitals pointing between them, so the degeneracy is lifted. This is crystal field splitting.

Octahedral field: six ligands lie on the axes. The d_x²−y² and d_z² orbitals point at the ligands and are raised in energy; they form the e_g set. The d_xy, d_yz and d_zx orbitals point between the ligands and are lowered; they form the t_2g set. The gap is Δ_o, the crystal field splitting energy. The average energy is preserved, so t_2g lies 0.4Δ_o below and e_g 0.6Δ_o above the barycentre. Δ_o depends on the field strength of the ligand and on the charge on the metal ion. Ligands arranged in order of increasing field strength give the spectrochemical series: I⁻ < Br⁻ < S²⁻ < SCN⁻ < Cl⁻ < F⁻ < OH⁻ < C₂O₄²⁻ < H₂O < NCS⁻ < edta⁴⁻ < NH₃ < en < CN⁻ < CO. It is an experimental series based on absorption spectra.

Filling: d¹, d², d³ go into t_2g one each, by Hund's rule. For d⁴ the fourth electron has a choice: if Δ_o < P (pairing energy), it goes into e_g, giving t_2g³e_g¹, a high spin complex with a weak field ligand; if Δ_o > P, it pairs in t_2g, giving t_2g⁴e_g⁰, a low spin complex with a strong field ligand. The same choice arises for d⁵, d⁶ and d⁷. So [Mn(H₂O)₆]²⁺ (d⁵, weak field) is t_2g³e_g² with five unpaired electrons, while [Mn(CN)₆]⁴⁻ is t_2g⁵e_g⁰ with one. [Co(NH₃)₆]³⁺ (d⁶) is t_2g⁶, diamagnetic; [CoF₆]³⁻ is t_2g⁴e_g² with four unpaired electrons.

Tetrahedral field: four ligands lie between the axes, so the splitting is reversed, the e set (d_x²−y², d_z²) lies lower and t₂ higher, and it is smaller: Δ_t = (4/9)Δ_o. Because Δ_t is usually less than the pairing energy, tetrahedral complexes are almost always high spin; low spin tetrahedral complexes are rarely observed.

A complex is coloured because it absorbs light of one colour from the visible region and transmits the rest; the colour we see is the complementary colour of the light absorbed. In CFT the absorbed energy promotes an electron from t_2g to e_g, a d–d transition, and the energy needed is Δ_o, which for most complexes falls in the visible range. The NCERT example is [Ti(H₂O)₆]³⁺, a d¹ ion: it absorbs light of about 500 nm in the blue-green region to move its single electron from t_2g to e_g, and the transmitted light gives the complex its violet colour. Some values from the NCERT table: [CoCl(NH₃)₅]²⁺ absorbs 535 nm (yellow) and looks violet; [Co(NH₃)₅(H₂O)]³⁺ absorbs 500 nm (blue-green) and looks red; [Co(NH₃)₆]³⁺ absorbs 475 nm (blue) and looks yellow-orange; [Cu(H₂O)₄]²⁺ absorbs 600 nm (red) and looks blue.

Three consequences are examined. First, if there is no ligand there is no splitting and no colour: anhydrous CuSO₄ is white, CuSO₄·5H₂O is blue. Second, if the d subshell is empty or full there is no d–d transition: [Sc(H₂O)₆]³⁺ (d⁰) and Zn²⁺ complexes (d¹⁰) are colourless. Third, the ligand changes the colour because it changes Δ_o: [Ni(H₂O)₆]²⁺ is green, and adding ethane-1,2-diamine step by step gives [Ni(H₂O)₄(en)]²⁺ pale blue, [Ni(H₂O)₂(en)₂]²⁺ blue-purple and [Ni(en)₃]²⁺ violet. A stronger ligand means larger Δ_o, higher energy absorbed, shorter wavelength absorbed. This is why the wavelength of absorption follows [Ni(NO₂)₆]⁴⁻ < [Ni(NH₃)₆]²⁺ < [Ni(H₂O)₆]²⁺. The colour of gemstones is the same effect in a lattice: ruby is Al₂O₃ with about 0.5 to 1 per cent Cr³⁺ replacing Al³⁺ in octahedral sites; emerald is Cr³⁺ in beryl, Be₃Al₂Si₆O₁₈.

Limitations of CFT. (i) It treats ligands as point charges, so anionic ligands should give the largest splitting; in fact they lie at the low end of the spectrochemical series. (ii) It ignores the covalent character of the metal–ligand bond, which is real and explains why CO and CN⁻ are such strong field ligands.

Homoleptic carbonyls, complexes containing only carbonyl ligands, are formed by most transition metals. They have simple, well defined shapes: [Ni(CO)₄] is tetrahedral, [Fe(CO)₅] trigonal bipyramidal, [Cr(CO)₆] octahedral. The binuclear carbonyls are [Mn₂(CO)₁₀], made of two square pyramidal Mn(CO)₅ units joined by a Mn–Mn bond, and [Co₂(CO)₈], which has a Co–Co bond bridged by two CO groups in addition to the six terminal CO groups. The metal is in oxidation state zero in all of these.

The metal–carbon bond has both σ and π character. The σ bond is formed by donation of the lone pair on the carbon atom of CO into a vacant orbital of the metal. The π bond is formed by donation of a pair of electrons from a filled d orbital of the metal into the vacant antibonding π* orbital of carbon monoxide. The metal-to-ligand π bonding creates a synergic effect: electron density pushed back into the ligand strengthens the metal–CO bond, and the extra negative charge on the metal from σ donation is relieved. This is why CO, a poor Lewis base towards ordinary acids, binds transition metals so strongly and sits at the top of the spectrochemical series.

The board asks two things here: draw and describe the synergic bonding (a diagram with the σ arrow from C to M and the π arrow from M d orbital to the π* of CO), and state the shapes of [Ni(CO)₄], [Fe(CO)₅] and [Cr(CO)₆]. The oxidation state of Ni in [Ni(CO)₄] is zero and its hybridisation is sp³, with the 4s electrons pushed into the 3d subshell so the complex is diamagnetic. Metal carbonyls are also the basis of the Mond process for refining nickel: impure nickel with CO at 330–350 K gives volatile [Ni(CO)₄], which decomposes at 450–470 K to pure nickel.

The board asks for one example under each heading, so learn them as a list. In qualitative and quantitative analysis: metal ions are detected and estimated through coloured complexes; Ni²⁺ gives a red precipitate with dimethylglyoxime (DMG); Cu²⁺ gives deep blue [Cu(NH₃)₄]²⁺; the hardness of water is estimated by titrating Ca²⁺ and Mg²⁺ against EDTA, and the two can be estimated separately because their EDTA complexes have different stability constants.

In metallurgy: silver and gold are extracted by leaching with cyanide in the presence of air, 4Au + 8CN⁻ + 2H₂O + O₂ → 4[Au(CN)₂]⁻ + 4OH⁻, and the metal is recovered with zinc, 2[Au(CN)₂]⁻ + Zn → 2Au + [Zn(CN)₄]²⁻. Nickel is purified by the Mond process through [Ni(CO)₄]. In catalysis: Wilkinson's catalyst, [(Ph₃P)₃RhCl], is used for hydrogenation of alkenes. In biology: chlorophyll (the pigment of photosynthesis) is a magnesium complex, haemoglobin (the oxygen carrier of blood) is an iron complex, and vitamin B₁₂ (cyanocobalamine, the anti-pernicious anaemia factor) is a cobalt complex; enzymes such as carboxypeptidase A and carbonic anhydrase are metal complexes too.

In medicine: cisplatin, cis-[Pt(NH₃)₂Cl₂], is used in cancer chemotherapy; EDTA is used in the treatment of lead poisoning by chelation; excess copper and iron are removed by the chelating ligands D-penicillamine and desferrioxamine B respectively. In electroplating: articles are coated smoothly with silver and gold from baths of the cyanide complexes [Ag(CN)₂]⁻ and [Au(CN)₂]⁻ rather than from simple salts. In photography: after the film is developed, hypo (sodium thiosulphate) dissolves the unreacted AgBr as the soluble complex [Ag(S₂O₃)₂]³⁻, which fixes the image.