Plastics are used all over the modern world from toothbrushes to fighter jets, because they have a very wide range of physical properties. This is partly because there are lots of different plastics with different chemistries, but two plastics with the same chemistry but different structures can have very different properties. Plastics are polymers, they are made up of repeating chains, petrol, wax and polythene are all just different lengths of the same type of hydrocarbon chain. Plus the chains don't have to be straight, they can be branched, and the branches can be of different lengths, all of which affect the strength stiffness and other properties of the plastic.
This is a huge advantage, but it is also a problem, as normally plastics for a particular application are produced essentially by trial and error, testing lots of different conditions, amounts of catalyst etc, until they produce the right properties. This is particularly a problem if we are to replace conventional plastics with biodegradable or plant derived plastics.
Daniel Read and colleagues from Leeds and Durham universities may have made this process much more predictable. They have developed computer model which will take the conditions of a polythene reaction chamber and predict the degree of branching, and the structure of that branching.
They then feed this data into a second computer model which predicts the physical properties, from the molecular structure by predicting the amount the molecules can wriggle past each other, in the middle of the chains, and how easily the ends of the chains can move, and they have now got it to work very well for polythene.
This means that they could already predict the conditions you need to produce a polythene with particular properties, and potentially what happens if you blend different polythenes as you recycle them, and by changing a few numbers in their simulation they could adapt it for new systems like bio-polymers making their development much faster and more efficient.