I am doing research in computational chemistry. Bascially I have studied 1 reaction the entire semester. If you introduce Li2 (lithium) to biphenylene with ultrasound, the ring opens up to give you a biphenyl molecule. What I have to do is use a computer to find the mechanism of this reaction. I tell the computer to make a potential energy surface of a structure that I suspect may be a ground state structure. The computer then optimizes the geometry of the structure to obtain a minimum energy. The computer does this by basically "walking" along the potential energy surface and stops when it gets to a minimum point. When the minimum point is found the computer then does frequency calculations. If there are imaginary eigenvalues (which comes from a bunch of quantum theory stuff) then the strucuture that was minimized is some sort of transition state structure which I am not interested in. My goal is to keep optimizing structures until I find a structure that gives frequency calculations resulting in all real numbers. This means that the structure optimized is indeed a ground state structure. Next, when I find two ground state structures, I do another calculation, known as a QST2 calc to try to link those two structures together. If after I find the structure that is inbetween those 2 ground state strucutures results in 1 imaginary frequency, then it is likely that that structure is a transition state from ground state structure A to ground state structure B. You can kind of get the idea of what I am doing. I am basically taking a bunch of peices of a "movie" of a reaction and trying to put them together in the right order--simply by using a computer, no lab work at all. When the movie is finally complete we can get a pretty good idea of how the reaction proceeds from start to finish. It is actually quite interesting stuff. I don't know how the computer actually does these calculations though. How the program works goes way over my head into theoretical physics and chemistry.