Often, when asked about my work by people not familiar with (theoretical) chemistry, I find it hard to explain in simple words what I was doing during my PhD thesis. So I decided to put up this little paragraph. Please note, I did not unravel the switching mechanism of azobenzene. I just use it as an illustrating example. Hopefully it helps giving those interested in my research a more vivid picture.

Beforehand, please note: The actual research results can be found in my list of publications.

**First off: What are we interested in?**

We are interested in a class of molecules that can exist in (two) different structures. Ideally, the structures differ in their physical properties, for example their color. If the structures can be interconverted by laser light, these molecules are called photoswitchable molecules. The most prominent photoswitchable molecule is called azobenzene (try and google it, there’s tons of pictures and publications on azobenzene!) and is shown in the picture below.

Azobenzene exists in two forms: The permanent stable one is depicted on the left side and labeled (1). It’s called *trans-*azobenzene. Once *trans-*azobenzene is hit by UV laser light (2), it changes shape and turns into it’s second form called *cis*-azobenzene (3). The photoswitching process can be used for all kinds of fancy photoswitchable materials, for example polymer films that bend with UV light and relax with blue light, materials that change their color or even as motor units in nano-machines. Obviously, we know azobenzene is changing it’s shape when hit by UV-light, so what’s the bit we don’t know? What we don’t know is this: what exactly does happen between step (2) and step (3)?

**How do we find out what is going on between step (2) and step (3)?**

Well, there’s actually a lot of ways how to find out what is going on, but -you guessed it – i will try to explain the theoretical approach that we use. The approach is labeled FOCI-AM1 surface-hopping, or – more general – FOCI-AM1 semi-classical dynamics. The “classical dynamics” in FOCI-AM1 semi-classical dynamics refers to Newtons equations of motion. These are the equations that describe the movement of planets, the path taken by a football or the path taken by the famous apple that hit Newton’s head. They do also describe the movement of a ball attached to a spring that goes up and down once the ball is pulled down and released. Imagine, the atoms that make up the azobenzene molecule are little balls connected with springs as shown in the figure below.

By solving Newton’s equations of motion we can simulate how the molecule is moving and vibrating, i.e. it’s dynamics. That’s nice, but there’s a problem with applying the model to photoswitchable molecules. The problem is: the approximation breaks down completely if the molecule is hit by light. When the molecule is hit by light strange things are happening. In the classical picture of balls and springs it looks something like this: The strengths of the springs start changing constantly with time as indicated in the middle of the above figure. In order to account for this, we use an equation called the Schrödinger equation. The Schrödinger equation is the basic equation of quantum mechanics and it is used, because the springs actually consist of electrons that form chemical bonds. Electrons are elementary particles and must be treated quantum mechanically. For larger molecules like azobenzene, the Schrödinger equation has to be solved approximately with the help of computers. This can only be done by introducing some kind of approximation. The approximation we use for solving the Schrödinger equation is called FOCI-AM1. So this is why it is called FOCI-AM1 semi-classical dynamics: We solve Newton’s equations of motion (“classical dynamics”), alongside we approximately solve the Schrödinger equation (“FOCI-AM1”) to update the strenghts of the springs all the time.

The below video illustrates what the results look like. Note, it comes from a ‘real’ FOCI-AM1 semi-classical calculation. The simulated time is 1 picosecond or 0.000000000001 seconds…

FOCI-AM1 semi-classical calculations were carried out with a MOPAC development version kindly provided by the theoretical chemistry group of Prof. M. Persico, University of Pisa. Graphics were made using xyz2svg by Sascha ‘Sushi’ Frick and VMD – Visual Molecular Dynamics.