Why does chemistry beat the curse of dimensionality?

In a typical chemical reaction, the goal is to find the best conditions to get the highest yield. Combinations of reagents, catalysts, temperatures and times are tested to find the optimal conditions. Finding the best combinations should require testing the majority of the possible combinations, but typically only 5–10 combinations are tested to avoid the ‘curse of dimensionality’ –  where the number of possible experiments grows exponentially with the number of variables. 

Now, US scientists reporting in Chemical Science say that, contrary to this reasoning, experience shows that synthesis and property optimisations are far easier to achieve than the curse of dimensionality suggests. They put forward the ‘OptiChem’ theory, concluding that the most efficient method is to change all important variables, performed with automated high-throughput synthesis.

Intrigued? Download the Edge article by Herschel Rabitz and colleagues to learn more.

Digg This
Reddit This
Stumble Now!
Share on Facebook
Bookmark this on Delicious
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)