New molecules may improve the performance of optical devices

The key element of any application that uses light, such as a DVD player or fiber optics, is the interaction of light with a material. We rely on such applications as we seek an ever faster Internet with larger data capacity, higher-density electronic storage media, and more effective cancer therapies. For example, an optical switch needs to be fast and consume little power, while an effective photodynamic cancer therapy requires specialized molecules that stick only to cancer cells and generate heat when exposed to laser light, zapping the disease without damaging healthy organs. Thus, increasing the efficiency of the interaction of light with the molecular building blocks of a material is important for a wide variety of applications. For over three decades, the efficiency of the best molecules remained a factor of 30 short of the fundamental limits.¹

In the past, a common approach to increasing the efficiency of interactions between light and matter was to make larger molecules and to pack more of them into the material. However, with this brute force approach, the efficiency never increased nearly as much as possible, so molecules never lived up to their full potential. The first step for breaking through this glass ceiling was taken by Zhou, Watkins, and myself when we used computer modeling to design the ultimate molecule.² We entered an arbitrary molecular shape into the computer, which calculated the switching efficiency per unit of molecular size. It then varied the shape while noting how the efficiency changed, in effect ‘tuning in’ to the ideal shape, as one would tune in to a radio station while listening for the best signal. The result was somewhat surprising, since it implied that the motions of the electrons in the molecule needed to be somewhat impeded. Shortly after the design criteria for the ideal molecule were established, I was part of an international collaboration with Pérez and Clays that reported on a new molecule with the required ‘speed bump’ to trip up the electrons. This new molecule broke through the long-standing glass ceiling with an efficiency that is 50% larger than the previously best molecules.³ Using the same paradigm, additional improvements—up to a factor of 20—are possible.

The strength of interaction between a molecule and light is quantified by the hyperpolarizability, which characterizes the probability that the molecule will mediate the merger of two photons into one. The second hyperpolarizability characterizes the merger of three photons, while higher-order terms represent the merger of even more photons. Applications such as electro-optic switching, frequency doubling, and non-contact chip testing rely on the hyperpolarizability, while all-optical switching, three-dimensional photolithography, and photodynamic cancer therapies rely on the second hyperpolarizability. Our computer simulations focused on optimizing the intrinsic hyperpolarizability, which divides out the effect of molecular size. The same technique can be applied to higher-order polarizabilities.

Figure 1 shows the initial smooth potential energy function— V(x)=tanh(x}—and the evolution of the potential energy function as it tunes in to the optimized response. After about 7000 iterations, the potential energy function develops dramatic wiggles that act to localize the various excited state wave functions. At this point, the intrinsic hyperpolarizability is more than 0.7, less than 30% away from the fundamental limit I calculated.⁴ Based on the observation of these wiggles, we suggested that a molecule with modulated conjugation should lead to the required bumpy potential. The inset to Figure 1 shows the proposed type of structure, which has alternating single and double bonds with nitrogens, carbons, and various ring structures. This variation in composition was proposed as a way to create the required bumps.

Using the concept of modulation of conjugation, we reported on a series of molecules that showed both smooth conjugation and modulated conjugations.³ Figure 2 shows a plot of the intrinsic hyperpolarizabilities of these molecules. In particular, molecule 4 has two identical rings, while molecule 7 has two rings of different compositions. These two different types of rings provide bumps of different sizes, which leads to modulation. The intrinsic hyperpolarizability of molecule 7 is 50% greater than the glass ceiling (i.e., the apparent limit characterizing the behavior of all previous molecules). Since this new record-breaking molecule is still about a factor of 20 below the fundamental limit, there is room for improvement.

The series of seven molecules, including number 7, were supplied by chemist Yuxia Zhao at the Chinese Academy of Sciences and measured in Belgium. The molecules currently have only lengthy, multisyllabic chemical names, and for that reason are not listed here.

To take full advantage of the concept of conjugation modulation, new molecules with many wiggles (i.e., speed bumps) will need to be synthesized and tested. If the theoretically possible improvement in intrinsic hyperpolarizability by a factor of 20 is realized, then green laser pointers (which use frequency doublers) can be made brighter, and optical switches can operate at lower voltage and power, thus making them more attractive in telecommunications systems. Subtler electric field variations can be detected with electro-optic sampling when testing new chip designs. By offering more efficient interactions between light and matter, the new supermolecules will enable these and other enhancements to current technologies ranging from data transmission to cancer treatments.