The s-metals have a strong affinity for hard F, O, and N ligands in solution, and they form strong complexes with them. This is especially true when the ligands are polydentate, that is they have more than one electron pair donor site. Examples are diphosphate, P₂O₇^4-, and EDTA^4-, ethylenediaminetetraacetate. Monodentate ligands are only weakly bound due to the weak coulombic interactions and lack of covalent bond formation.

P2O74-	EDTA4-

Other ligands, with even more donor groups, form complexes which are even more stable than those with mono- or polydentate ligands of the type above. Crown ethers form compounds which survive in non-aqueous solutions, and bicyclic cryptate ligands form complexes which can survive even in aqueous solution.

18-crown-6	2.2.1 cryptate	2.2.2 cryptate

The cryptate ligands trap the metal cation in the central cavity, and so the stability of these compounds is very sensitive to the size of the cation.

The smaller 2.2.1 crypt-ligand forms more stable compounds with Li⁺ and Na⁺, whereas the larger 2.2.2 crypt-ligand forms more stable compounds with K⁺ and Rb⁺. 18-crown-6 is strongly size selective for K⁺.

The increased stability of these polydentate ligand complexes is due to the chelate effect: that is that complexes containing chelate (polydentate) rings are more stable than the system which is as similar as possible but contains fewer or no rings. The extra stability is the chelate effect.

The Chelate effect
(en = H2NCH2CH2NH2)
log K = 10.6	ΔH = -54 kJmol-1	ΔS = +23 JK-1mol-1
log K = 7.7	ΔH = -46 kJmol-1	ΔS = -8.4 JK-1mol-1

It can be seen that the chelate effect is largely entropic in nature: in both cases there are two Cu-N bonds formed and two Cu-O bonds broken, and so the reaction enthalpy is similar, but the big difference is that in the first case, we start with 2 molecules and end up with 3 in solution, whereas in the second the number of molecules in solution is conserved. The first situation is more entropically favourable. The formation of strong complexes with the EDTA ligand is a good example of the chelate effect.

Another result of this is known as the macrocyclic effect: a macrocyclic ligand complex is more stable than its open chain analogue. This is due to the preorganisation of the ligand towards complexation. Much of the work needed to be done to get the ligands in the correct orientation for complexation is already achieved when the ligand is macrocyclic, and hence the energetics of formation are more favourable than with open chain analogues, which have to reorient themselves for complexation, which takes energy. This effect can be seen in the reactions on the crypt-ligands and crown ethers.

One result of the stability of the complexes with these cyclic ligands is the fact that in non-aqueous solutions, the alkali metals may dissolve to give a solutions containing the alkalide ion, M^-. These solutions have absorption spectra which depend on the nature of the ion, suggesting the absorption comes from the transfer of charge from the alkalide ion to the solvent.

The Na⁺ ion is trapped in the cavity of the 2,2,1-cryptand. In the crystal, the sodide ion, Na^-, is held in cavities  and has a radius larger than that of I^-.