coordination chemistry

three types of structural/constitutional isomerism

hydrate: ligand interchanged/replaced with water of hydration - e.g. [Cr(H₂O)₆]Cl₃ - [Cr(H₂O)₅]Cl₂.H₂O - etc

ionisation: interchange of anionic ligands - e.g. [Co(NH₃)₅SO₄]Br - [Co(NH₃)₅Br]SO₄

linkage: where the same ligand can bind through different atoms (has more than one possible coordination mode - ambidentate) - e.g. nitro -NO₂ vs nitrito -ONO; thiocyanate

what dominates hard-hard and soft-soft interactions?

hard-hard is ionic bonding - but dominated by ΔS - as they are well solvated (by water), when they bond, the release of water molecules is an entropic driving force

soft-soft is covalent bonding - and dominated by ΔH - orbital overlap determined by HOMO-LUMO gap

β_n vs K_n

β_n is the overall formation constant, = [ML_n] / [M][L]ⁿ

K_n is the stepwise formation constant, = [ML_n] / [ML_n-1][L]

β_n = K₁K₂…K_n (consider denominators and numerators of successive K_ns)

usual trend in successive stepwise stability constants

K₁ > K₂ > … > K_n

• statistical phenomenon: decreasing number of available sites as ligands complex

• steric hindrance: if the complexing ligand is larger than the ligand displaced

• Coulombic interactions: if the complexing ligand is more charged than the ligand displaced

K_n trend for Cu(II) / NH₃

abnormality is K₅ and K₆ very small due to Jahn-Teller distortion; weak bonding at the axial 5th and 6th coordination sites

K_n trend for Cd(II) / I^-

abnormality is K₃ large (> K₂) - due to stereochemical change:

CdI₂(OH₂)₄ → [CdI₃(OH₂)]^-

octahedral → tetrahedral due to steric factors (Cd²⁺ group 12, I^- large)

loss of 3 H₂O - entropically favoured

K_n trend for Fe(II) / phen

abnormality is K₃ large (> K₂) due to spin state change

[Fe(phen)₂]²⁺ is still high spin but → [Fe(phen)₃]²⁺ is low spin; big increase in LFSE stabilising (d⁶ !)

Mn²⁺ < Fe²⁺ < Co²⁺ < Ni²⁺ < Cu²⁺ > Zn²⁺ trend in (log)K₍₁₎

Irving-Williams series; insensitive to choice of ligand

• across the series Z_eff increases (because …)

  • (increased Lewis acidity;) 3d AOs are lower in energy and are a better match for ligand AOs

• LFSE stabilisation

  • increases across the period to Ni²⁺ and is zero at Zn²⁺

• but Cu²⁺ largest?

  • Jahn-Teller distortion - stronger bonding of equatorial ligands increases the value of K (definitely for K₁)

Irving-Williams for M²⁺ / en

normal for K₁ and K₂ (so K₁ > K₂ > K₃) but for Cu²⁺, K₃ is (very) low due to rigid en

rigid bidentate ligand prevents significant J-T distortion for Cu(en)₃²⁺ (fine for Cu(en)₂²⁺ as can bind in the equatorial plane)

naming of cryptands? try drawing [2.1.1], [2.2.1], [2.2.2], [3.2.2]. which ion best fits [2.2.2]?

K⁺ for [2.2.2] cryptand

note also K⁺ best for 18-crown-6

entropic driving force for binding for all: release of solvation (H₂O) molecules

factors affecting binding affinity

size match (geometric complementarity):

• if too small, bonds are too long (on average - maybe some are lost entirely if metal does not sit in the centre) are therefore weaker

• if too large, cage has to distort and so strain in the system is not enthalpically favourable

chelate effect: enhanced stability of chelated complexes compared with less chelated or non-chelated complexes

factors contributing to the chelate effect

• mainly entropic: chelation results in a net increase in the number of independent molecules in solution

• enthalpic contributions:

  • polar donor groups are already covalently linked together in the chelating ligand

    • no increase in lp-lp repulsion, compared to lp being brought closer together upon binding for a non-chelating ligand

  • chelating ligand usually a better donor (more basic) - due to electron donating alkyl groups

• greater effective concentration of chelating ligands

  • once one donor group binds, it becomes more probable that another donor group of the chelating ligand binds as it is held in close proximity to the metal ion. (diagram…)

  • is obviously kinetic but also thermodynamic with an entropic effect - probability - can be considered in terms of microstates…

effect of chelate ring size

chelate effect generally decreases with increasing ring size

except for 5- vs 6- membered chelate rings:

• bite size is larger in a 5-membered ring compared to 6-membered (draw hexagon to demonstrate)

  • the bond length between donor atom and metal is longer in a 5-ring vs 6-. ideal 6- rings (with their smaller bond lengths) have less strain (cf. chair cyclohexane), but for ions which are too large then this does not hold.

• large ions favour 5-membered rings, small ions favour 6-membered rings. Pb²⁺ vs Cu²⁺

origins of the macrocyclic effect

enhanced stability of macrocyclic ligand complexes compared with open-chain analogues

mainly enthalpic effect:

• size complementarity - stronger M-L bonds

• solvation - macrocycles are usually less solvated than open-chain analogues (diagram?), so fewer ligand-solvent bonds to break before binding to metal

• ligand pre-organisation: favourable conformational enthalpy is already present in the synthesised molecule, which overcomes repulsive forces between polar donor groups (e.g. lp-lp repulsion)

  • no increase in repulsion (strain) upon binding to the metal ion, vs open-chain which does have increase in e.g. lp-lp repulsion

entropic effect:

• pre-organisation: conformational entropy - macrocycles lose fewer degrees of freedom upon complexation as they are inherently less flexible

• solvation effects are harder to analyse, as although solvent molecules released from metal ion, they are also released from the centre of the cavity

  • choice of solvent (how polar?) can effect binding affinity - how strong are the ion-solvent, ligand-solvent bonds that need to be broken?

kinetics of macrocyclic effect

K is much greater for macrocycles than open-chain

• but this is due to (very) low dissociation rate for macrocycles

  • mutliple coordinate bonds must be broken at once; for acyclic can do one at a time (see below) due to greater conformational flexibility

• formation rate usually a bit faster for acyclic

• Eigen-Wilkins: form an outer-sphere complex first, then exchange one S for L (reversible!)

• any individual step could be rate determining.

  • for macrocycles, usually due to strain/high energy conformation required to put donor atom in the right position; free rotation in acyclic can avoid this

how would you promote macrocycle dissociation?

compete with the metal, e.g. use H⁺ to compete with M for N based ligands, EDTA…

anion coordination chemistry?

often bigger and more charge diffuse; need bigger receptor ligands

geometries of anions vary; spherical to linear to ??

• e.g. elongated cryptand-like (protonated) macrocycle prefers linear azide

preorganisation, macrocycles helps, as with cations…

synthesis of macrocycles… how?

template effect:

• (e.g.) metal-ligand interactions used to pre-organise components into the desired geometry for reaction

• especially useful when desired reaction conformation is not kinetically and/or thermodynamically favoured

• thermodynamic and kinetic; see other flashcard

problems:

• demetallation can be difficult: may need to add competing ligand (EDTA, CN) or protonate (Ns) or redox…

• hard to identify correct M

if templating not possible, use high dilution synthesis (dropwise addition of reactants over many hours)

• to promote intramolecular cyclisation over oligomerisation

thermodynamic and kinetic template effects

thermodynamic: metal ion stabilises (part of) the RHS of the equilibrium, driving it

• in this case Ni²⁺ back bonding in to C=N π*, making it less electrophilic

kinetic: coordination of precursor to macrocycle holds reactive groups in close proximity, and in the correct geometry

beware [Co(en)₂Cl₂]⁺

cis/trans isomerism - but cis isomer also has optical isomers

Schiff base condensation (diketone with diamine)

size of metal ion template can determine which product is formed [1+1] vs [2+2]

S₃-macrocycle synthesis

MeCN a weakly coordinating ligand

Mo(0) favours octahedral coordination…?

point of the templating is to allow the S^- (which are S-Mo) close to each other (would otherwise repel…) then can form the macrocycle with the dibromo, rather than oligo

high dilution synthesis of cryptands

high dilution x 2

why? metal ions will not template well to the bis acid chlorides (electrophiles…); use high dilution

Co(en)³⁺ with ammonia and formaldehyde

good LFSE of Co³⁺ in O_h geometry; then that geometry allows capping vs oligomerisation

the point is that by holding in the right geometry, rate of intra > rate of intermolecular - kinetic template

can also happen with Fe³⁺ with triamine + diester

Pd²⁺, Fe²⁺ templating

Pd²⁺ square planar, can form ‘squares’ with en + 4,4’-bipy

Fe²⁺ octahedral, can form tetrahedral cage

catenane synthesis… how? (bis phenols + I-C-(C-O-C)₄-C-I)

‘catenane’ = chain

need T_d coordination: Cu(I) is used - Cu(MeCN)₄⁺ - MeCN weakly binding

to demetallate ^-CN used (could use EDTA perhaps)

the reaction has the two ligands (bis phenols) orthogonal (T_d coordination)

then with the deprotonated phenols S_N2 on I-C-(C-O-C)₄-C-I

[2]-rotaxane, Schiff base chemistry; both components have 3 Ns

so we need Oh coordination since 6-coordinate

use first row TMs; Co(II) especially good

always need to demetallate

[2]-rotaxane; not yet macrocycle (dangling alkenes) has 3 Ns, axle has only 1N; 3 N component has dangling alkenes

use Pd(II) - e.g. Pd(OAc)₂ - for square planar geometry - coordinate to 3N component

(the fourth coordination orginally from solvent, e.g. MeCN) then the axle coordinates

then need to use Grubbs catalyst (Ru based carbene) to clip - two alkenes into one, lose ethene - ring closing metathesis

then add competing ligand to demetallate (not specified, ^-CN perhaps for square planar)

catenane; components have only one donor atom each (N), have dangling alkenes

need a linear template - use Au(I) (d¹⁰)

this forms the orthogonal precursor complex (much like the T_d case)

then clip alkenes with Grubbs catalyst (Ru based carbene) - lose ethene, ring closing metathesis

need to demetallate - e.g. H⁺ for N ligands

can reduce the alkene to alkane with H₂, Pd/C