that of the covalent characteristics of the pyrazolate fragment,
resulting in significant red-shift from the corresponding 1a–e
ligands.
2d and 2e exhibited strong, single fluorescence maximized at
460 and 430 nm, respectively in THF (see Fig. 2 for 2e). Similar
results were observed in the solid crystal. Conversely, while the
solid crystal of 2a–c showed only one emission centered at 382,
360 and 370 nm, respectively, dual emission was observed in
solution, consisting of a normal emission band (2a: 375, 2b: 365
and 2c: 374 nm, the F1 band) and an anomalously large Stokes
shifted emission (2a: 505, 2b: 488 and 2c: 512 nm, the F2 band)
in THF. The excitation spectra for both F1 and F2 bands are
identical, which are also effectively the same as the absorption
profile, excluding its origin from traces of impurity. In contrast
to the nearly solvent independent F1 emission frequency, the
fluorescence peak frequency for the F2 band was linearly
proportional to solvent polarity (Df).† Accordingly, a change of
dipole moment of ~ 6.6 Debye between ground and excited
states was deduced.
Fig. 3 Temperature-dependent emission spectra of 2a in 2MTHF at
298–150 K. Insert: The plot for ln(kot) versus the reciprocal of tem-
peratures.
Further insight into the correlation of dual emission proper-
ties was gained from the dynamic studies. As shown in Table 1
the lifetime of the F1 band for 2a was fitted to be ~ 0.53 ns (c2
= 1.02) at 298 K, while the F2 band is apparently composed of
rise and decay components that were fitted to be 0.52 ns and
18.3 ns, respectively (c2 = 1.05). The rise time of the F2 band,
within experimental error, is identical with the decay time of the
F1 band, supporting a precursor-successor type of relaxation
mechanism. As shown in Fig. 3, the F1/F2 ratio for 2a increased
upon decreasing the temperature. The reaction rate monitored
by the decay dynamics of the F1 band or equivalently the rise
dynamics of the F2 band revealed significant temperature
dependence (see insert of Fig. 3). Similar dual emission and
precursor-successor related dynamics were also obtained for 2b
and 2c and the results are listed in Table 1.
The above results can plausibly be rationalized by a
mechanism incorporating a photoinduced electron transfer (ET)
process from the phenyl moiety to the pyrazolate ligand. The
separations from the center of two phenyl fragments to the
center of pyridyl and pyrazolate ring systems are estimated to be
4.6–4.7 Å and 4.9–5.0 Å, respectively. These lengths are
significantly shorter than the typical distance ( ~ 7 Å) that
allows the occurrence of through-space electron transfer. The
reduction peak potential of 5-(2-pyridyl) pyrazolate complexes
2a–b and 2d–e occurred at 21.67, 21.72, 21.76 and 21.86 V,
respectively in CH3CN. The decrease of reduction potential
correlates well with the trend of electron withdrawing properties
of substituents at the pyrazolate ligand. The result, in combi-
taion with the increase of the absorption energy gap being in the
order of 2a ~ 2b ~ 2c > 2d,2e, qualitatively rationalizes the
occurrence of ET in 2a–c. To further support for the phenyl ring
acting as an electron donor in the proposed ET mechanism we
also synthesized 2f in which the absorption gap is similar to 2a,
whereas the phenyl ring is replaced by a perfluorophenyl moiety
to increase the relative oxidation potential. As a result, ET
process is prohibited in 2f, as indicated by a unique, normal
emission band maximum at 382 nm (see Fig. 2).
The drastic difference in photophysics between single crystal
(single, normal emission) and solution phase (dual emission) in
2a–c is intriguing. The logarithm plot of the ET rate6 vs. 1/T is
sufficiently linear (insert of Fig. 3), from which a barrier of 1.31
kcal mol21 and a frequency factor of 1.88 3 1010 s21 were
deduced for 2a. The Ea value obtained for 2a–c (see Table 1) is
on a similar magnitude as the viscosity barrier ( ~ 1.82 kcal
mol21) in 2-methyltetrahydrofuran (2MTHF), indicating that
certain large amplitude motions may couple with the ET
process. Both X-ray and AM1 approaches indicate that the
phenyl rings are nearly orthogonal to the (2-pyridyl) pyrazolate
ligand in 2a–f, of which the configuration may prohibit the ET
process.‡ It is thus tentatively proposed that in solution phase
the ET mechanism may incorporate rotation of the phenyl ring
to an optimum oreintation so that a through-space ET reaction
can take place.
Due to the straightforward syntheses of boron complexes it is
feasible to adjust the D/A strength so that the degrees of charge-
transfer interactions and consequently the luminescence effi-
ciency can be fine-tuned. An ideal boron-ligands system for
devices may exhibit both LE and CT emissions covering an
entire white-light region. Work on other systems such as
imidazoles and triazoles is currently in progress.
Notes and references
crystallographic data in .cif or other electronic format.
1 (a) Q. Wu, M. Esteghamatian, N.-X. Hu, Z. Popovic, G. Enright, Y. Tao,
M. D’Iorio and S. Wang, Chem. Mater., 2000, 12, 79; (b) Y. Li, Y. Liu,
W. Bu, J. Guo and Y. Wang, Chem. Commun., 2000, 1551; Y. Liu, J.
Guo, H. Zhang and Y. Wang, Angew. Chem., Int. Ed., 2002, 41, 182.
2 For example, see:(a) T. D. James, K. R. A. S. Sandanayake and S.
Shinkai, Angew. Chem., Int. Ed., 1996, 35, 1910; (b) K. Rurack, M.
Kollmannsberger and J. Daub, Angew. Chem., Int. Ed., 2001, 40, 385; (c)
N. DiCesare and J. R. Lakowicz, J. Phys. Chem., 2001, 105, 6834; (d) W.
Yang, H. He and D. G. Drueckhammer, Angew. Chem., Int. Ed., 2001, 40,
1714; (e) J. Killoran, L. Allen, J. F. Gallagher, W. M. Gallagher and D.
F. O’Shea, Chem. Commun., 2002, 1862.
3 (a) A. Satake and T. Nakata, J. Am. Chem. Soc., 1998, 120, 10391; (b) W.
R. Thiel and J. Eppinger, Chem. Eur. J., 1997, 3, 696.
4 (a) S.-F. Liu, Q. Wu, H. L. Schmider, H. Aziz, N.-X. Hu, Z. Popovic and
S. Wang, J. Am. Chem. Soc., 2000, 122, 3671; (b) Q. Liu, M. S. Mudadu,
H. Schmider, R. Thummel, Y. Tao and S. Wang, Organometallics, 2002,
21, 4743.
5 For example, see:(a) U. Mitschke and P. Bauerle, J. Mater. Chem., 2000,
10, 1471; (b) A. J. Campbell, D. D. C. Bradley, T. Virgili, D. G. Lidzey
and H. Antoniadis, Appl. Phys. Lett., 2001, 79, 3872.
6 ket was deduced by kobs 2 knr. The kobs value for 2f was taken as knr due
to its non-ET process. Note lnketT1/2 vs. 1/T is also sufficiently linear
mainly due to the much larger ket value.
Fig. 2 Emission spectra of (a) 2a (-0-), (b) 2b (-5-), (c) 2c (-:-), (d) 2e
(-!-) and (e) 2f (-8-) in THF.
CHEM. COMMUN., 2003, 2628–2629
2629