Sep 22 2010
Over 80% of all products manufactured today in the chemical and pharmaceutical industries require the use of catalysts.
Catalysts are materials which themselves are not consumed within chemical reactions, but which serve to accelerate those reactions and set them on course to create the desired products. To date, the search for the optimal catalyst has been akin to hunting for a needle in a haystack, and is mostly driven by intuition and happy accident. However, in order to accelerate the discovery of optimal catalysts, the working group led by Prof. Bernhard Breit, Internal Senior Fellow of the FRIAS School of Soft Matter Research, has now developed an entirely new concept which allows these catalysts to be found much more easily than before. The chemists have demonstrated this new technique using the enantioselective hydrogenation of alkenes as a model, and their results have recently been published in the online edition of "Nature Chemistry".
The new process for generating and identifying catalysts uses a combinatorial approach where catalyst libraries are produced by simply mixing complementary components. In this case, catalytically active rhodium(I) centres are modified with phosphine ligands which bond to them. Here, precisely two phosphine ligands always bond to one rhodium centre, a process that the Freiburg chemists have been able to ensure by designing special phosphine ligands. Similar to the adenine (A) - thymine (T) base pair in DNA, these ligands can form complementary hydrogen bonds with one another.
By mixing twelve phosphine ligands with ten different complementary phosphine ligands and a metal salt, 120 self-assembling, defined molecular catalysts will form without any additional synthesis steps. In order to identify the most active and the most selective candidates from this catalyst library, the new principle of iterative deconvolution (unravelling) was developed. For this purpose, the entire library is divided into sub-libraries. These sub-libraries of catalysts now compete against one another in each test reaction, with activity and, in this case, enantioselectivity(*) as the competition criteria. In the next step, the researchers focused exclusively on the sub-library which had achieved the best result in this competition (it must contain the most active and most selective catalysts), and then re-divided this sub-library into smaller units.
These libraries then competed again in the same test reaction. This process was continued until the best individual catalysts were identified. Using this method, it was possible to identify excellent catalysts from a library of 120 by means of 17 individual experiments for each different class of substrate. This approach is clearly superior to the classic procedure in which all 120 catalysts must be tested in parallel (therefore entailing 120 experiments) and a corresponding number of reaction analyses performed. The approach is universally applicable and should be transferable to many problems associated with chemical and biochemical catalysis.