Substituted Silylium Ions in Solution

To test these predictions, boryl substituted silylium ions were investigated by Ottosson and co-workers. [46] Hyperconjugation between Si+ and boryl substituents is more stabilizing than in the case of silyl substituents since Si-B 

These observations led to the question whether the work on silylium cations in solution is still justified and whether one should continue to publish research results on silylium cations in the top ranking journals of chemistry. Despite of the many failures, it was in particular Lambert who continued to work on the problem ignoring all criticism and trying to get a solution with better and better chemical tools. Independent of Lambert, Cremer and co-workers carried on with both ab initio and DFT calculations to solve the question of blocking the Si+ center of a silylium cation in a more efficient way against solvent attacks. The break through in the computational work was supported by two developments which on first sight had little to do with the silylium cation problem. 

Of course, one can calculate many molecules that will never be experimentally investigated because of insurmountable problems in connection with their synthesis. Exactly, this seemed to be the case for the trimesitylsilylium cation Mes3Si+ (61). Attempts to generate 61 via reaction (7) 

The 29Si NMR chemical shift of 61 in benzene solution was measured to be 225.5 ppm, which suggested that Si possesses a considerable positive charge. [135] Lambert and Zhao showed further that the 829Si value did hardly change (A 1 ppm) when other aromatic solvents were used, which caused the authors to conclude that interactions between ion 61 and solvent molecules are weak and a free silylium cation had been generated. [135] 

Of course, a NMR chemical shift value of 225 ppm can be interpreted in different ways. One can assume that 61 represents a free silylium cation in solution that because of considerable -conjugation between 3pn(Si) orbital and the three phenyl rings is internally stabilized and, therefore, does not interact with solvent molecules. The question in this case would be to which extent 11-conjugation has reduced the silylium cation character and how much positive charge is spread over the C framework. It could be that 61 represents more a carbenium ion in the same way as the silaguanidinium ion represents more an ammonium ion (see section 4.1). 

---