Nanoscale optical frequency doublers

Recent advances in blue-light-based photonic devices such as next generation data storage blocks or fully integrated opto-chips have renewed significant interest for the development of small and efficient light conversion units. In this context, a nanoscale integrated frequency doubler would allow the use of cheap, power-saving and brilliant near-infrared laser sources —such as easily integrable photonic band gap vertical cavity surface emitting lasers—to obtain intense and coherent blue light sources.

Highly efficient frequency doubling, however, requires that the hyperpolarizabilities of the constituent molecular building blocks be optimized. The use of doubling elements with dimensions significantly smaller than the wavelength of the doubled light fortunately makes phase matching issues obsolete. But aggregating molecules into nanoelements still imposes restrictions on the doubling efficiency related to the aggregation state, which can be single crystalline, polycrystalline, or amorphous.

We have demonstrated that it is possible to grow well-oriented crystalline, needle-like aggregates (nanofibers) on template substrates such as muscovite mica using organic molecular beam epitaxy (see Figure 1).¹ These nanofibers consist of molecules oriented nearly parallel to the substrate surface and perpendicular to the long axis of the fibers. Their dimensions are a few hundred nanometers in width, a few tenths of nanometers in height and several hundred micrometers in length. In terms of frequency doubling, an important feature of this class of nanoaggregates is their crystalline perfection. This is due to the epitaxial growth process that allows them to adopt the optimum configuration for the given molecular constituents. Hence, they also exhibit optimal optoelectronic properties, only limited by the molecular building blocks themselves.

However, such ‘perfect’ epitaxial growth imposes strong boundary conditions on the choice of substrate and adsorbate. In fact, for several years, it seemed that only para-hexaphenylene (p6P) molecules could fulfill these conditions. Although their high quantum efficiency makes these molecules very attractive for photonic applications, their hyperpolarizability is very low since they lack donor and acceptor groups. More promising nanofibers are based on the non-symmetric chemical functionalization of a para-quaterphenylene (p4P) block incorporating electron push-and-pull groups, such as methoxy and amino groups.² The p4P block is required to fulfill the epitaxial growth criteria.

Figure 2 shows the optical emission spectra of different functionalized molecules upon infrared excitation. The p6P spectrum consists solely of two-photon luminescence (TPL), while that of the methoxy and amino functionalized p4P (MONHP4) only consists of second-harmonic generated (SHG) emission. If chlorine is used instead of the amino groups, both SHG and TPL spectral signatures are obtained. Molecular tailoring clearly paves the way to enhanced optoelectronic flexibility.

Figure 3 shows a plot of SHG intensity vs. excitation power with the expected quadratic dependence. It can also be seen that the conversion efficiency of infrared light can be significantly increased if the fibers are functionalized with appropriate end groups.

Such SH-optimized fibers can easily be imaged in the light of their emission. Figure 4(a) shows the SH image of MOCNP4 nanofibers acquired with a laser scanning microscope. In contrast to fluorescence images, the SH images display a different intensity distribution along the fiber surface when imaged in reflection and transmission, as shown in Figure 4(c). This is due to the fibers having different interfaces with air and support substrate, resulting in different local SGH enhancement factors,³ as illustrated in Figure 4(b).

In conclusion, we have successfully synthesized nonlinear optically active functionalized molecules, growing nanofibers that emit true SHG blue light following excitation with near-infrared femtosecond light pulses. Our approach yields nanofibers in which the elements are multiplied by a factor of millions per square centimeter without losing their optimized properties. Since individual nanofibers or arrays can easily be transferred⁴ or embedded, important advances in working devices that incoporate nanoscale nonlinear optical elements are anticipated.