Exploring molecular design space resulting in new paradigms for second-order nonlinear optics.

Since many years, researchers are investigating nonlinear optical active materials to obtain high values for the first hyperpolarizability. Most of the research is done on traditional D-pi-A structures. Improving the donor (D) and acceptor (A) capacity of such structures containing a conjugated pi-bridge gives often higher second-order nonlinear optical (NLO) responses. Another way of improving the first hyperpolarizability (beta) can be reached by elongating the conjugated path of the molecule. A lot of research is still going on to improve the D-pi-A structures, but in this work, the objective is to go beyond the D-pi-A paradigm and to explore the theory of a free electron on a helix.[1]A first step in that direction was to investigate molecules with a higher dimensionality. V-shaped molecules, molecules with a binaphthalenyl unit and chiral X-type pi-conjugated oligomers have been studied. During the research of the latter group, a direct through-space charge-transfer interaction between the p-type and n-type oligomers was uniquely revealed by a significant enhancement of the first hyperpolarizability values, which proved that hyper-Rayleigh scattering is a good tool to monitor the through-space transfer for this type of conjugated oligomer.Further extending the idea of higher dimensionalities and the theory of a free electron on a helix, led this work to the investigation of helicenes. Helicenes are seen as the best compromise between a long planar conjugation path and chiral out-of-plane molecules. However, by showing low values of the first hyperpolarizability and being hard to oxidize, the helicenes do not fulfill the expectations. As a consequence, the helicenes are abandoned as model compound for high second-order nonlinear optical responses. In a following step, more alternative geometries were explored, such as the (DA)n oligomeric system (D = 3,4-ethyleendioxythiophene, A = pyridine), which fulfills the noncentrosymmetry requirements in nonlinear optics in an alternative way. For every extra DA-unit, a higher value of the first hyperpolarizability is observed. Depolarization measurements of this series reveal that the compounds are forming a helical conformation, which is very interesting. After abandoning the helicenes, the helical conformation remains and is successful as a way of increasing the first hyperpolarizability. When dissolving the (a)chiral polyphenanthrene (Pa) in different solvents or solvent mixtures, different geometries can be obtained. The polymer adopts in a good or poor solvent respectively a disordered (random coil) conformation or a one-handed helical structure. The solvent conditions can be tuned and it is observed that the coil-helix transition is not a single-step process. The formation of achiral structure(s) with torsion angle precedes the helix conformation. For the coiled conformation, an unexpectedly high (higher than the apparent Kuzyk limit) value of the first hyperpolarizability is observed and thus helix formation nor intrinsic chirality are required. In a next part, more attention is paid to the free electron of the theory of Maki. In the same series, electrochemical switching is investigated together with protic and light switching. In this way, a three-way orthogonal switch for second-order nonlinear optics is for the first time investigated. The protic and light switch are based on the more traditional change in D-pi-A structures, while unfortunately the electrochemical switch was not effective enough to turn the first hyperpolarizability on and off. This shows that structures with a higher dimensionality are not necessarily more effective for switching than the one-dimensional D-pi-A structures being altered between noncentrosymmetry (D-pi-A) and centrosymmetry (A -pi-A or D-pi-D ).By reaching, for polymer Pa in the best solvent conditions, a high beta-value that breaks the apparent Kuzyk limit, this work clearly demonstrates that the objective to go beyond the D-pi-A paradigm to obtain high second-order NLO responses is realized.[1] Maki JJ, Persoons A : J. Chem. Phys. 1996, 104 (23), 9340-9348status: publishe