Organic chemistry is a visual science. That’s what trips up so many people. I’ve known plenty of smart, intellectual scientists who loved inorganic chemistry with its straight, rigid structures and defined bond angles, but they crashed and burned when it came time to learn about organic chemistry. Organic chemistry is entirely three dimensional. Where atoms are in space plays a key role in reactivity. Where the electrons are situated places a huge role in reactivity. Without a strong visual capability to cope with imagining a molecule in three-dimensions, organic chemistry becomes extremely difficult, even impossible.
In the not so distant past, the only way to cope with this was a model kit. Small colored plastic spheres would plug together with plastic “bonds” – almost like Legos, or a construction set – to make a reasonably correct three dimensional view of the compound in question. You could then look at the model, and determine for yourself some of its properties. Was it extremely difficult to force the atoms and bonds into the shape you needed? Did you have to force and push, straining to connect things? Does the finished model look like it might spontaneously fall apart at any moment? In that case, you might predict that the molecule isn’t very stable. As it is on the large scale, so it is on the molecular scale, and in the laboratory you might find that the molecule isn’t very stable and falls apart whenever you try to create it.
Does the three dimensional model show that a particular group of atoms is far apart from the rest of the molecule, on opposite sides of the molecule even? In that case, you might predict that it would be very difficult for that part of the molecule to interact and react with the rest of the molecule – a so-called intramolecular reaction. That may be a desired quality, and it may not be a desired quality – many “ring closing” reactions depend on the ability of the molecule to contort and flip around, so that two separate ends of the same molecule come into very close proximity.
It was this type of crude approximation that could be performed with the plastic models of old. Relatively recently, user friendly and powerful computer software packages have been released that computerize this entire process, and give not only much more detailed data, but more accurate data as well. This was made possible entirely by the rapid increase in computer processing speed over the past three decades. One of the most powerful software suites currently available is Spartan 08. While it is expensive (several thousand dollars for a single license), it is an incredibly useful piece of software, and well worth the investment.
The process of using Spartan is fairly simple. Using the mouse and keyboard and a user interface, the chemist draws a molecular structure on their screen. Using point and click, the user then selects the properties that they are interested in predicting, along with the amount of precision they’ll be happy with. While it is theoretically possible to get incredibly precise information about any molecule of interest, in practice you find that larger molecules and extra precision carry a heavy computational burden. Predicting simple properties about a small molecule is trivial; learning detailed, precise properties about a complex molecule could take weeks, even months. Therefore one of the most important steps is learning how much accuracy you are happy with. The calculation is submitted, and the computer goes to work – a completely automated process; you just leave the program running, and come back to it when everything is complete.
The image I’ve attached to this article shows just a very simple example of this type of technology. The molecule is called “phenol” – it and its derivatives are common bacteriacides and are the foundation of an entire class of organic chemicals. The molecule took about ten seconds to input using Spartans user interface, and the calculation – which was performed on a small molecule, and at a reasonably low level of theory – was complete in two seconds on my home desktop. The results are shown in the image, which is fitting given organic chemistry and it’s three-dimensional aspect. The UI allows you to freely zoom in and out, and flip / rotate the image in three dimensions.
You can instantly glance at the molecules image and gain valuable information about the molecule. Just by looking at it, you can see that all the sides of the hexagon ring are equal in length. This is an important principle of organic chemistry called “aromaticity”, and is vital to understanding the chemical reactivity of phenol. If the compound wasn’t “aromatic”, it would have alternating long and short bond lengths. The colors also tell a valuable story. Red areas contain a lot of electron density, meaning that more electron-type character surround those atoms as compared to blue areas. Notice the red area is centered around one atom – in this case, an oxygen. This makes sense, as from the Periodic table we can see that oxygen has more unbound pairs of electrons than carbon, and oxygen is also more “electronegative” – it has a higher greed for electrons than carbon, drawing electron density in like a magnet.
This has unfortunate effects for the hydrogen attached to the oxygen, as you can see based on the intense blue color surrounding it. This is because the electrons that make up the oxygen-hydrogen bond are not equally shared between the two atoms; oxygen is more electronegative, and sucks up available electron density, leaving little to surround the hydrogen atom. This has the effect of making that particular hydrogen quite acidic, as atoms that lack sufficient electron shielding are good targets for bases.
So, just by looking at the colors and the shapes, a chemist can instantly gain important chemical information about the molecule. All of this is generated entirely by the computer and requires only seconds. While most molecules are larger than phenol and therefore require longer computational times, the same principles for deciphering the image hold true. This type of technology – where chemical compounds can be “screened” before they’re even synthesized in the laboratory – is vital for new drug discovery and for the advances we have become accustomed to in recent years. An engineer tells the chemist what type of property is needed; the chemist translates that into his own type of terminology, and envisions a type of structure which might be beneficial; these structures are evaluated in a modeling program like Spartan 08, and the most promising candidates from that step are then synthesized in the laboratory and tested. Molecular modeling saves the chemist not hours, not weeks, but years of work.
Spartan ’08 is an incredibly powerful piece of software. While on the surface it’s very simple – it shows a structure in three dimensions and then uses chemical and mathematical formulas to predict the molecular properties – the ramifications are profound. Chemists can spend their time and expertise investigating only a handful of compounds, instead of making thousands in the lab only to discard 95+% of them as useless. As a result, the scientific process is vastly quickened, to the benefit of us all.
For more information about computational modeling, or to order your own copy of Spartan 08, check out: http://www.wavefun.com/