Mechanism of "turn-on" fluorescent sensors for mercury(II) in solution and its implications for ligand design

The tendency of a Hg(II) ion to strongly quench fluorescence of potential fluorescent sensors is explored. Fluorescence measurements show the expected order of the chelation-enhanced fluorescence (CHEF) effect of Zn(II) > Cd(II) >> Hg(II) ~ Cu(II), which is interpreted as (1) unpaired electrons causing the weak CHEF effect for Cu(II) and (2) the order Zn(II) > Cd(II) >> Hg(II) reflecting the "heavy atom" effect, which may be due to increasing spin-orbit coupling constants (ζ) for Zn(II) < Cd(II) << Hg(II). The structures of mercury(II) complexes of N-(9-anthracenylmethyl)-N-(2-pyridinylmethyl)-2-pyridinemethanamine (ADPA) are reported. [Hg(ADPA)Cl(2)HgCl(2)] (1) has one Hg(II) held by two bridging chlorides, while the other Hg(II) is coordinated to the ADPA ligand. The latter Hg(II) has a nearest π contact of 3.215 Å with a C atom from the anthracenyl group, which falls in the range of reported Hg-C π contacts with aromatic groups. This contact may be important in quenching the fluorescence of the Hg(II)/ADPA complex. A density functional theory study shows that the Hg-C interaction is strong enough to prevent a simple HOMO → LUMO transition of the fluorophore. In fact, the S(0) → S(1) and S(2) transitions in the Hg(II)/ADPA complex have significant charge-transfer character to mercury. An important aspect of the coordination geometry of Hg(II) is illustrated by 1, where Hg(II) tends to form a few (often only two) short bonds to the more covalently binding donor atoms present, with much longer bonds to other donor atoms. The Hg-N bonds to the two pyridyl N-donor atoms of ADPA in 1 are relatively short at 2.212(8) and 2.224(8) Å, while that to the central saturated N-donor atom of ADPA is long at 2.603(8) Å. The latter long Hg-N bond may allow a photoinduced electron-transfer (PET) effect, quenching the fluorescence of the anthracenyl fluorophore. The structure of [Hg(ADPA)Br(2)] (2) reflects the more covalent binding of the two bromine ligands compared to the clorine ligands of 1, with much longer Hg-C contacts with the anthracenyl fluorophore and a Hg-N contact with the saturated N atom of ADPA of 2.917 Å. The latter long Hg-N contact is related to the nearly negligible fluorescence of the ADPA complex in the presence of added Br(-). The addition of extra ligands to the Hg(II)/ADPA complex produces a weak increase in the fluorescence intensity for OH(-) ~ Cl(-) >> Br(-) > I(-), which is discussed in terms of an increasing PET effect, and to collisional quenching. The ligand design principles for generating turn-on sensors for mercury suggested by this work are discussed.