Cross-referenced combinatorial libraries for the discovery of metal-complexing ligands: library deconvolution by LC-MS

N-Acylthioureas are excellent ligands for a variety of heavy metals, but th eir metal selectivity is highly dependent on the precise nature of the subs tituents present. In this paper we show how combinatorial chemistry techniq ues can be used to establish relative affinities for copper within a mixtur e of 100 such thioureas. Following a straightforward synthesis, and copper extraction using standard liquid-liquid extraction techniques, LC-MS was us ed to identify the ligands which bind most strongly to the copper ions. Amo ng the 100 ligands XC(O)N(Z)C(S)NHY, the most important substituent is the Y group bound to the NH: only aromatic Y substituents give strong binding t o copper. The acyl X substituents are invariably aromatic, and an electron- rich X group is best; the affinity for copper seems to be less dependent on the Z substituent, although a large group such as benzyl disfavours copper binding. The five ligands from the library which bind copper most strongly have been clearly identified by a series of experiments: they all have aro matic groups in the Y position, but the X and Z substituents can be more va ried. This is a very convincing demonstration of the power of combinatorial methods: to have found the same information by conventional methods would have required a lengthy and repetitive series of syntheses and investigatio ns. In addition, our results give some preliminary evidence for synergistic binding of two different ligands, but this requires further investigation.