as are the splittings in all four Tb(EDTA)(CA) emission spectra. As
Stark splittings of the emission bands report on the composition
and geometry of the lanthanide coordination sphere, the data sug-
gest that these CAs all bind to Tb^3 þin the same fashion for a given
Tb(ligand)binarycomplex.Theexcitationspectraindicatethatthe
Cat moiety of each analyte is dianionic, and therefore bidentate
coordination via deprotonated hydroxyl moieties is very likely. In
all cases, the intensity was greater for DO2A complexes than those
containingEDTA.Wethereforeconcludethat,althoughbothancil-
lary ligands are capable of forming ternary complexes with CAs,
[Tb(DO2A)]þis the more effective sensor.
The stability provided by the receptor ligand is integral to the
success of this assay. CAs can only bind strongly to cations when
fully deprotonated, which means working at high pH, and
lanthanides only remain soluble at high pH with the aid of a
helper ligand. The use of the ancillary ligand allows these two
components—the analyte and the lanthanide—to exist under the
same conditions, making this otherwise unwieldy assay sensitive,
rapid, and straightforward.
Similar effects are seen with the inclusion of the DO2A helper
ligand in the previously described dipicolinate system. A pH
dependence study conducted over a range from 6.1 to 10.4
indicates that with the DO2A ligand bound, the lanthanide does
not precipitate and the [Ln(DO2A)(DPA)]–complex is stable over
the entire pH range for all lanthanides studied( 92 ). In contrast,
the Ln(DPA)þcomplexes form Ln(DPA) 33 – at high pH, indicating
precipitation of some of the lanthanide as Ln(OH) 3. This further
validates the stabilizing effect of the DO2A ancillary ligand,
allowing the bacterial spore detection assay to be performed over
a much wider range of conditions.
These pH dependence studies have also demonstrated an
important point in terms of using lanthanide ions and complexes
as sensors. In every dipicolinate system, the luminescence inten-
sity of the Ln(DPA)þcomplex varied significantly with pH, due
largely to the precipitation of Ln(OH) 3 and the resulting forma-
tion of more strongly luminescent Ln(DPA)n species, where
n¼2 or 3. There was no longer a direct correlation between lumi-
nescence intensity and dipicolinate concentration, and the num-
ber of bacterial spores could not be quantified. Such intensity
variations are undoubtedly a problem with other analyte sys-
tems as well, as any bi- or tridentate analyte can form
multimeric Ln^3 þ complexes. With the receptor ligand, fortu-
nately, the change in luminescence intensity with pH is no more
than 5% for dipicolinate, and therefore the bacterial spore con-
centration can be determined directly from emission intensity.
26 MORGAN L. CABLEet al.