Moreover, at higher energy the 314 nm transition is that mainly
responsible for the converged higher b value of L4 compared to
L3, indeed the same state in L3 is characterized by a lower
transition dipole moment.
In conclusion L3 and L4 are particularly appealing because
of their notable values of EFISH mb1.097 that are likely due to
the absence of photorotamerism and photoisomerism pro-
cesses operating in the corresponding compounds L1 and L2
respectively. These properties together with the absence
of isolated double bonds ensure the astonishingly high
thermal stability of L4 that makes it a good candidate for
the preparation of materials characterized by high second
harmonic generation properties. We believe that, because of
their relevant technological interest, the results here reported
can constitute a springboard for the design of other NLO
chromophores of the same family such as 11-N,N-dibutyl-
amino-3-azapicene and 12-N,N-dibutylaminobenzo[m]-3-
azapicene with 5 and 6 rings respectively.
We thank deeply S. Righetto for EFISH measurements and
S. Proutiere for DSC study. This work was supported by
MIUR (FIRB RBNE033KMA) and CNR-INSTM
(PROMO 2006).
Fig. 3 Comparison between theoretical (blue line) and experimental
(red line) spectra of L3 and L4 with related computed hyperpolariz-
abilities through a SOS approach in the insets.
a weak HOMO–LUMO excitation at 361 nm, of p–p* character.
The transition giving the band at 332 nm is essentially from the
HOMO to the LUMO+1 and has charge-transfer character,
going from the donor dibutylamino part of the ligand to the
acceptor pyridine ring. The absorption band computed at 277 nm
is constituted by two main transitions of global p–p* character.
The experimental absorption spectrum of L4 has a quite similar
shape with respect to that of L3, even though it is red-shifted and
it shows the appearance of a band tail at ca.400 nm. We
computed bands at 353 nm and 312 nm to be compared to the
experimental absorption bands at 361 nm and 305 nm.
Notes and references
1 For example: Molecular Nonlinear Optics: Materials, Physics,
Devices, ed. J. Zyss, Academic Press, Boston, 1994.
2 (a) S. K. Yesodha, C. K. S. Pillai and N. Tsutsumi, Prog. Polym.
Sci., 2004, 29, 45; (b) L. R. Dalton, A. W. Harper, R. Ghosn,
W. H. Steier, M. Ziari, H. Fetterman, Y. Shi, R. V. Mustacich, A.
K.-Y. Jen and K. J. Shea, Chem. Mater., 1995, 7, 1060;
(c) K. D. Singer, E. Sohn, L. A. King, H. M. Gordon,
H. E. Katz and P. W. Dirk, J. Opt. Soc. Am. B, 1989, 6,
1339.
3 Photoreactive Organic Thin Films, ed. Z. Sekkat and W. Knoll,
Accademic Press, San Diego, California.
4 I. Ledoux and J. Zyss, Chem. Phys., 1982, 73, 203.
The band at 353 nm is a charge-transfer transition essentially of
HOMO–LUMO+1 character. We computed a weak lower
energy transition at 392 nm of charge transfer character,
corresponding to the HOMO–LUMO transition, related to the
low energy experimental feature and more intense than that of the
L3 ligand, consistently with the experimental spectra. To gain
insight into the electronic transitions responsible for the observed
quadratic hyperpolarizability b of L3 and L4, we performed a
Sum Over States (SOS) analysis.15 The results, reported as insets
of Fig. 3, show in both cases a converged positive value in
agreement with the experimental evidence. The ratio of the
computed b values of L3 and L4 is 2.2, which is in line with the
experimental increase of the EFISH b value observed passing
from L3 to L4. This increase is mainly due to the red-shifted
absorption spectrum of L4, related to the HOMO–LUMO gap
decrease. Indeed, according to the two-level model, b increases
quadratically with decreasing the energy of the electronic
transitions. From an in depth analysis of calculated SOS b values,
we noticed that the largest contributions to b are provided for
both L3 and L4 by ILCT transitions. Visualization of the charge
transfer processes originating the NLO response is reported in the
ESI.w We found that for L4 the lowest energy transition
considerably contributes to the calculated b, for L3 such excited
state has almost vanishing contribution to b, due to its
higher energy and considerably lower transition dipole moment.
5 D. Locatelli, S. Quici, S. Righetto, D. Roberto, F. Tessore,
G. J. Ashwell and M. Amiri, Prog. Solid State Chem., 2005, 33, 223.
6 (a) D. Roberto, R. Ugo, S. Bruni, E. Cariati, F. Cariati,
P. C. Fantucci, I. Invernizzi, S. Quici, I. Ledoux and J. Zyss,
Organometallics, 2000, 19, 1775; (b) F. Tessore, E. Cariati,
F. Cariati, D. Roberto, R. Ugo, P. Mussini, C. Zuccaccia and
A. Macchioni, ChemPhysChem, 2010, 11, 495.
7 G. Hubener, A. Lambacher and P. Fromherz, J. Phys. Chem. B,
2003, 107, 7896.
8 V. Calabrese, PhD thesis, University of Milan, Italy, 2008.
9 The absolute mb1.907 value decreases by increasing the concentra-
tion from 10À4 M to 10À3 M due to aggregation effects at higher
concentrations. Due to aggregation effects,
determined experimentally.
m could not be
10 E. A. Guggenheim, Trans. Faraday Soc., 1949, 45, 714.
11 V. Alain, M. Blanchard-Desce, I. Ledoux-Rak and J. Zyss, Chem.
Commun., 2000, 353.
12 D. M. Burland, R. D. Miller and C. A. Walsh, Chem. Rev., 1994,
94, 31.
13 M. J. Frisch, et al., Gaussian 03, Revision C.02, ESIw.
14 M. Cossi and V. Barone, J. Chem. Phys., 2001, 115, 4708.
15 This approach requires in principle the calculation of dipole matrix
elements between all possible couples of excited states (three level
terms), in addition to ground to excited states transition dipole
moments (two level terms).16 However two- and three-level terms
show approximately the same scaling with the number of excited
states as two-level terms, so that the latter can be used for a
semi-quantitative assessment of the contributions to the quadratic
hyperpolarizability.
16 D. R. Kanis, P. G. Lacroix, M. A. Ratner and T. J. Marks, J. Am.
Chem. Soc., 1994, 116, 10089.
c
8376 Chem. Commun., 2010, 46, 8374–8376
This journal is The Royal Society of Chemistry 2010