| Author: | Stefano |
|---|---|
| category: | Computational Chemistry |
Of the many things I posted, I never had the chance to write something about my direct scientific activity. Recently I worked on optical properties of polyenes. A paper has been published recently on Journal of Chemical Physics. Another one is submitted right now, and a third is in preparation.
Molecules interact with light. This should come to no surprise, as (for example) the whole world of colors depends on this effect. The most trivial interaction between the light and a molecule is normally absorption of one photon followed by emission of the same photon. Nothing changes in terms of energy of the photon. A given wavelength enters, the same wavelength leaves. The molecule is unchanged and unaffected by the event, except for a brief excitation of the electrons cloud. This event is typical linear optics behavior.
However, if we increase the photon density enough, two photons can be absorbed at the same time by the same molecule. After this event takes place, the molecule has many choices to return to electronic ground state. One is to re-emit two photons again. Nothing changes, still a linear optics effect.
However, the molecule could also release a single photon, whose frequency is derived by the total amount of energy provided by the two original photons. In other words, a photon of frequency twice the original will be emitted. This is called Second Harmonic Generation and if you have seen a green laser pointer, you directly experienced this effect: the green laser diode is in reality a very strong infrared laser diode covered with a crystal of potassium titanyl phosphate. The infrared photons are absorbed and doubled by the crystal into the green wavelength we see. Exactly the same phenomenon is used with blue lasers used for Blu-ray technology.
Frequency doubling is just one of the non-linear optics effects. There are many more, and they are promising to develop optoelectronic devices, where we control light with an electric field. Unfortunately, in most cases the intensity of non-linear effects is very, very small. You need a very strong laser emission to reach a sufficient photon density for the phenomenon to be appreciable (or used for pratical, non purely instrumental purposes). Moreover, we are using inorganic crystals at the moment, but crystals are fragile, tend to degrade as they are under thermal and optical stress, and they cost a lot. As a consequence, research focused on organic compounds, carbon-based molecules able to produce sizable non-linear optics effects at a fraction of the price, better mechanical properties, and less degradation. Having a polymer able to produce strong non-linear effects would dramatically reduce the cost and increase the life of these devices.
Exploring the chemical space of organic compounds in the "wet lab" is demanding, polluting, and in some cases unachievable, so we simulate the lab on a computer, running a computational method that predicts the intensity of the non-linear effects of a molecular structure of our choice. Our task is therefore to produce a lot of molecules, feed them into this computational machinery, get the evaluation of the non-linear behavior, and try to spot some rules to guide us in maximizing the characteristics we are interested in. As it frequently happens, there's no "perfect compound". Instead, there's a good trade-off, but we are interested in devising rules, not finding the perfect molecule with a brute force approach (which is unachievable, there are simply too many compounds possible out of carbon, hydrogen, nitrogen, oxygen and sulphur: infinite).
My work was focused on polyenes. It's a nice class of compounds with a nice single-double bond alternation. This allows "almost free" flow of electrons through this kind of molecular wire.
We know that non-linear optics properties are influenced by
- The length of the polyene chain. Longer chains give higher values, with a behavior which can be approximated as a power law for short chains.
- The substituents groups we put at the ends of the chain (marked as black dots in the picture). Different groups produce different molecules, and therefore, different non-linear properties.
We explored what happened to the non-linear properties as we increase the length of the chain, and at the same time, we include different combinations of end-caps substituents groups. We chose four critically important substituents: two electron donors, one strong (NH2), one weak (OH), and two electron acceptors, one strong (NO2) and one weak (CN). In addition, we used the neutral substituent H. Results were very interesting, and in some cases unexpected. Among many other things, we found that the presence of two substituents can be approximated, in some cases, as a simple addition of two single substitutions, meaning that for certain lengths of the chain, the interaction of the two groups vanishes and they behave as they are isolated.
We also found that the presence of these groups distorts the molecule from its linear, rod-like shape to a C-shaped or S-shaped chain, depending on their nature. This is rather remarkable finding, as there was no computational report for this and very scarce experimental report only on a similar class of compounds. The shape of the molecule has both an effect on the non-linear properties, and on how the polymer crystallizes (depending how good is the packing of the various chains). A new paper on the Journal of Physical Chemistry A has just been accepted on these findings, and it will be published as soon as the editorial process is performed.
So I have something to celebrate tonight. I think I'll go out for a nice sushi!