Fig. 2 Variation of the fluorescence intensity ratio at 615 nm as a
function of molar molybdate fractions at pH 5.7 (average of two
independent titrations). The inset shows the corresponding emission
spectra. Buffer 2,6-lutidine, λexc = 460 nm.
Fig. 4 pH dependence of the relative emission intensity of solutions
containing: a) 1, 1 ϩ molybdate (0.5 equiv.), Fe3ϩ (0.3 equiv.), Cu2ϩ
(0.3 equiv.) or Zn2ϩ (0.5 equiv.); b) 1 and 1 ϩ the indicated anions
(0.5 equiv. each). λexc = 460 nm, λem = 615 nm.
In a further step, the selectivity of 1 was investigated by titrat-
ing the acidic sensor solution with standard base in the presence
of potentially interfering cations or anions. The emission inten-
sity vs. pH-profiles obtained in the presence of metal cations,
such as Fe3ϩ, Cu2ϩ or Zn2ϩ superimpose well on that observed
for the titration of 1 alone (Fig. 4a). The lack of interactions
may be rationalised by the electrostatic repulsion between the
Fig. 3 Time-resolved emission decay profiles for a) 1 and b) 1 ϩ
Na2MoO4 (1 : 1), λexc = 355 nm, λem = 615 nm.
metal cations and the positively charged sensor molecule.
transfer an electron to a proximate photogenerated RuIII-
3Ϫ
Remarkably, even the addition of oxoanions, such as PO4
,
(bipy Ϫ) moiety, leading to luminescence quenching.10
ؒ
3Ϫ
2Ϫ
ReO4Ϫ, VO4 and WO4 did not alter the pH-profile sig-
nificantly (Fig. 4b). This demonstrates that selective binding
and signalling of molybdate can also be achieved by taking
advantage of the characteristic co-ordination chemistry of
molybdenum.
Analogous pH-dependent experiments were carried out with
solutions also containing 0.5 equiv. of molybdate. As is evident
from Fig. 1, the emission intensity of 1 decreases at significantly
2Ϫ
lower pH in the presence of MoO4 than in its absence.
This quenching effect can be related to the co-ordination of
molybdenum to the catecholate unit of 1. The end point of the
titration of 1 with molybdate in buffered solution at pH 6.7 is
consistent with a molybdenum : catecholate 1 : 2 complex (Fig.
2). The fact that the emission intensity of the non-complexing
derivative 2 is not affected by the presence of molybdate (Fig. 1)
further confirms that the quenching observed in the case of 1 is
caused by molybdenum binding and is not due to a diffusion
controlled, intermolecular process.
Fluorescence measurements were also performed at low-
temperature. Aqueous acetonitrile solutions containing a) 1
and b) 1 and 0.5 equiv. of molybdate were adjusted to pH 9 and
pH 7, respectively, to ensure that substantial quenching occurs.
Cooling the solutions to 77 K gave glasses and largely restored
the emission of the signalling unit in both cases. Such lumin-
escence revival indicates that the quenching is due to electron
transfer (eT) and not energy transfer (ET) processes.14
A preliminary time-resolved emission study of 1 and its
molybdenum complex was conducted using 355 nm Nd : YAG
excitation (10 ns pulsewidth, Fig. 3). The decay of the MLCT
excited state of 1 was fitted adequately by a single-exponential
function, giving a lifetime of 710 30 ns, which is in accord-
ance with the lifetimes typically observed for [Ru(bipy)3]2ϩ
complexes. In the presence of molybdate, the decay profiles
were found to fit to dual-exponential functions. The longer life-
time corresponds to free 1, while the shorter lifetime (ca. 30 ns)
corresponds to the molybdenum complex of 1. The proportion
of the two functions altered with molybdate concentration.
According to the pH-profiles obtained, the highest selectivity
for molybdate can be achieved between pH 4.8 and pH 5.8. In
fact, by using a buffered solution of 1 at pH 5.7, no significant
luminescence quenching was observed on addition of Fe(),
see supplementary information (ESI†) (or vanadate), while the
subsequent addition of molybdate to the same solution resulted
in a significant intensity decrease (like that shown in Fig. 2).
In summary, the biomimetic modular system 1 represents a
promising prototype of a new class of luminescent sensor
reagents for molybdate. Under pH control, 1 selectively
2Ϫ
binds MoO4 and signals its presence and concentration by
a decrease in emission intensity. The selectivity is based on
a combination of metal–ligand bonding and electrostatic
interactions. Interesting applications should be possible by
modification of the prototype, e.g. through an improvement
of its water solubility or its immobilisation at the tip of a
fibre-optic probe. Studies directed towards the determination of
stability constants and the improvement of the water com-
patibility of the system are underway.
We thank Dr. Jared D. Lewis for help with the laser experi-
ments and the University of York Research Priming Fund for
financial support.
Notes and references
‡ 1H-NMR (500 MHz, 313 K, DMSO-d6) δ 1.55 (s, broad, 4H,
CH2(CH2)2CH2), 2.48 (s, 3H, CH3), 3.21 (m, broad, 2H, CH2CH2NH),
2328
J. Chem. Soc., Dalton Trans., 2001, 2327–2329
Jedner, Stephanie B.
