Final Report Summary - SILICAT (New silicon-derived environmentally-benign Lewis-acid catalysts for asymmetric synthesis)
New silicon-derived environmentally-benign Lewis-acid catalysts for asymmetric synthesis
Brönsted and Lewis acid catalysed reactions are among the most important and practical transformations of organic molecules as they often offer unique reactivities and selectivities. Many key carbon-carbon bond forming reactions such as acylation, alkylation, polymerisation, addition and cycloaddition reactions which are of outstanding importance for the industrial production of pharmaceuticals, agrochemicals, fine and bulk chemicals are promoted by acid catalysts. However, these reactions often suffer from major drawbacks in particular when they are applied to the synthesis of large amounts of products. Turn-over frequencies are often low and, hence, the catalyst loading is high (sometimes a stoechiometric amount of the catalyst has to be used). These reactions often generate large amounts of salts and, consequently, work-up can be quite tedious. Also many Lewis acids are derived from toxic metals which have to be removed from the final products. Finally, many Lewis acids lack chemoselectivity and are thus incompatible with highly functionalised substrates. This introduces severe limitations in the use of these catalysts for the synthesis of complex highly functionalised molecules such as pharmaceuticals, agrochemicals and advanced materials. Not surprisingly a vast amount of work has been dedicated to this fundamental problem of catalysis and recently some elegant approaches have been proposed to overcome these problems.
The present project was based on the seminal discovery of Ghosez' and Mikami's groups that trialkylsilyltriflimide (Me3SiNTf2) acted as extremely powerful Lewis acid catalysts. They exhibit a strong Lewis acidities which could be tuned-up by varying the size of the alkyl groups on silicon and they showed a good tolerance for many acid-sensitive functional groups. An additional advantage of these trialkylsilyl imide catalysts was their easy removal from the reaction mixtures by simple extraction with water. Also these cheap and powerful catalysts were non-genotoxic and environmentally benign.
The main objectives of the project were the design and evaluation of powerful silicon triflimide catalysts which would allow for enantioselective reactions. An obvious choice was to introduce chiral alkyl substituents on the silicon atom. This was indeed successful with some derivatives of myrtenal. We prepared eight silylated triflimides derived from the terpene myrtenal and tested them but the enantiomeric excesses never exceeded 55 %. J. Leighton at Columbia University and B. List at the Max Planck Institüt, Mülheim respectively studied silicon chloride catalysts derived from chiral aminoalcohols or a trialkylsilyl sulfonimide catalyst derived from a chiral biaryl. They obtained excellent enantioselectivities but only for a limited number of applications since the catalytic activity was too low to allow for a general use. Thus, it became clear to us that a successful design of the 'ideal' silicon-derived catalyst would first required the understanding of the structural parameters controlling the catalytic activity. Accordingly, we modified our initial plans.
Our new strategy consisted in preparing a wide variety of silicon derivatives carrying:
(1) a very good leaving group;
(2) two ligands (mostly heterosubstituted) susceptible to be made chiral easily;
(3) an alkyl or aryl group.
Eighteen new compounds have been prepared. The Lewis acidity of these new compounds was tested by measuring their complexation with methyl acrylate or methyl crotonate. Then, the catalytic activity of these potential catalysts was tested on some model reactions (Diels-Alder, alkylation, aldol). As far as we can tell, this work represents the first systematic and comprehensive study of a 'structure-catalytic activity' relationship of these electrophilic silicon compounds. As expected, the presence of oxygen, nitrogen or carbene substituents on the silicon atom strongly influenced the catalytic activity of the silicon-derived Lewis acid. This study also led to the discovery of two catalytic species. The first species is derived from a tetracoordinated silicon atom and is tetrahedral. The second species is derived from a pentacoordinated silicon atom and is a bipyramid-trigonal molecule. This should offer even more opportunities for the design of efficient asymmetric catalysts.
The preparation and systematic study of these new Lewis acids took most of our time. However, in the last few months, we were able to prepare the very reactive triflimide analogs of the Leighton catalysts. As expected, they were much more active than Leighton's chlorides and could be used in 'catalytic' amounts in contrast to the Leighton's chlorides which have to be used in stoechiometric amounts. However, enantiomeric excesses on our model reactions were low. Further experimental and computational studies of the parameters controlling the facial selectivity of these reactions are definitely needed.
The fundamental knowledge acquired during these 18 months study should allow us to select and prepare classes of silicon-derrived Lewis acids which should ensure a satisfactory catalytic activity for a given reaction and the possibility of varying simply and practically the nature of the chiral information in the substituents. The availability of 'green', efficient and selective catalysts for reactions allowing to build efficiently complex, highly functionalised scaffolds for the development of new bioactive compounds is obviously of great importance. However, we believe that the results obtained in this Marie Curie project offer new perspectives for 'green' catalysis, a 'hot' area of fundamental importance for pharmaceutical, agrochemical and fine chemical industries. For exampIe, we found that our new catalysts were able to activate environmentally safe and non-genotoxic starting materials for alkylation reactions. It is very clear that more applications of these new classes of silicon-derived catalysts should be expected in the context of the development of sustainable chemical transformations both at the laboratory and the industrial level. Contacts have already been established with the representatives of chemical industries in Europe to evaluate the potential of these catalytic systems.
We believe that this concern about environmental safety should be shared by all researchers working in chemical synthesis. Working on a project like this was of great benefit for the education of a 21st century chemist. One drawback, however: the practical importance of the results is such that disclosure of the structure of these new catalysts has to wait for the filing of a patent. However, we are convinced that the game is really worth the candle.
Brönsted and Lewis acid catalysed reactions are among the most important and practical transformations of organic molecules as they often offer unique reactivities and selectivities. Many key carbon-carbon bond forming reactions such as acylation, alkylation, polymerisation, addition and cycloaddition reactions which are of outstanding importance for the industrial production of pharmaceuticals, agrochemicals, fine and bulk chemicals are promoted by acid catalysts. However, these reactions often suffer from major drawbacks in particular when they are applied to the synthesis of large amounts of products. Turn-over frequencies are often low and, hence, the catalyst loading is high (sometimes a stoechiometric amount of the catalyst has to be used). These reactions often generate large amounts of salts and, consequently, work-up can be quite tedious. Also many Lewis acids are derived from toxic metals which have to be removed from the final products. Finally, many Lewis acids lack chemoselectivity and are thus incompatible with highly functionalised substrates. This introduces severe limitations in the use of these catalysts for the synthesis of complex highly functionalised molecules such as pharmaceuticals, agrochemicals and advanced materials. Not surprisingly a vast amount of work has been dedicated to this fundamental problem of catalysis and recently some elegant approaches have been proposed to overcome these problems.
The present project was based on the seminal discovery of Ghosez' and Mikami's groups that trialkylsilyltriflimide (Me3SiNTf2) acted as extremely powerful Lewis acid catalysts. They exhibit a strong Lewis acidities which could be tuned-up by varying the size of the alkyl groups on silicon and they showed a good tolerance for many acid-sensitive functional groups. An additional advantage of these trialkylsilyl imide catalysts was their easy removal from the reaction mixtures by simple extraction with water. Also these cheap and powerful catalysts were non-genotoxic and environmentally benign.
The main objectives of the project were the design and evaluation of powerful silicon triflimide catalysts which would allow for enantioselective reactions. An obvious choice was to introduce chiral alkyl substituents on the silicon atom. This was indeed successful with some derivatives of myrtenal. We prepared eight silylated triflimides derived from the terpene myrtenal and tested them but the enantiomeric excesses never exceeded 55 %. J. Leighton at Columbia University and B. List at the Max Planck Institüt, Mülheim respectively studied silicon chloride catalysts derived from chiral aminoalcohols or a trialkylsilyl sulfonimide catalyst derived from a chiral biaryl. They obtained excellent enantioselectivities but only for a limited number of applications since the catalytic activity was too low to allow for a general use. Thus, it became clear to us that a successful design of the 'ideal' silicon-derived catalyst would first required the understanding of the structural parameters controlling the catalytic activity. Accordingly, we modified our initial plans.
Our new strategy consisted in preparing a wide variety of silicon derivatives carrying:
(1) a very good leaving group;
(2) two ligands (mostly heterosubstituted) susceptible to be made chiral easily;
(3) an alkyl or aryl group.
Eighteen new compounds have been prepared. The Lewis acidity of these new compounds was tested by measuring their complexation with methyl acrylate or methyl crotonate. Then, the catalytic activity of these potential catalysts was tested on some model reactions (Diels-Alder, alkylation, aldol). As far as we can tell, this work represents the first systematic and comprehensive study of a 'structure-catalytic activity' relationship of these electrophilic silicon compounds. As expected, the presence of oxygen, nitrogen or carbene substituents on the silicon atom strongly influenced the catalytic activity of the silicon-derived Lewis acid. This study also led to the discovery of two catalytic species. The first species is derived from a tetracoordinated silicon atom and is tetrahedral. The second species is derived from a pentacoordinated silicon atom and is a bipyramid-trigonal molecule. This should offer even more opportunities for the design of efficient asymmetric catalysts.
The preparation and systematic study of these new Lewis acids took most of our time. However, in the last few months, we were able to prepare the very reactive triflimide analogs of the Leighton catalysts. As expected, they were much more active than Leighton's chlorides and could be used in 'catalytic' amounts in contrast to the Leighton's chlorides which have to be used in stoechiometric amounts. However, enantiomeric excesses on our model reactions were low. Further experimental and computational studies of the parameters controlling the facial selectivity of these reactions are definitely needed.
The fundamental knowledge acquired during these 18 months study should allow us to select and prepare classes of silicon-derrived Lewis acids which should ensure a satisfactory catalytic activity for a given reaction and the possibility of varying simply and practically the nature of the chiral information in the substituents. The availability of 'green', efficient and selective catalysts for reactions allowing to build efficiently complex, highly functionalised scaffolds for the development of new bioactive compounds is obviously of great importance. However, we believe that the results obtained in this Marie Curie project offer new perspectives for 'green' catalysis, a 'hot' area of fundamental importance for pharmaceutical, agrochemical and fine chemical industries. For exampIe, we found that our new catalysts were able to activate environmentally safe and non-genotoxic starting materials for alkylation reactions. It is very clear that more applications of these new classes of silicon-derived catalysts should be expected in the context of the development of sustainable chemical transformations both at the laboratory and the industrial level. Contacts have already been established with the representatives of chemical industries in Europe to evaluate the potential of these catalytic systems.
We believe that this concern about environmental safety should be shared by all researchers working in chemical synthesis. Working on a project like this was of great benefit for the education of a 21st century chemist. One drawback, however: the practical importance of the results is such that disclosure of the structure of these new catalysts has to wait for the filing of a patent. However, we are convinced that the game is really worth the candle.
