Silylium cation

U. Pidun, M. Stahl, and G. Frenking, Chem. Eur. J., 2, 869 (1996). The Silaguanidinium Cation and the Search for a Stable Silylium Cation in Condensed Phases. 

This occurs because the Si-H bond in free silanes is already polarized in the sense Si8+ Hs and coordination to an electrophilic M increases the positive charge on Si. This favors its effective elimination as a silylium cation, R3Si +, a powerful electrophile that can abstract OH from trace water to give R3SiOH and also extract fluoride from counteranions such as SbF6 [Eq. (28)] and even the BArf anion (104). As initially proposed by Crabtree (103) and supported by calculations discussed below, the cleavage in most cases is likely to be a concerted process, i.e., the nucleophile attacks the bound Si [Eq. (29)]. 

It is justified to ask why so much attention has been paid to a problem that may be only interesting within Si chemistry. Why did the silylium cation problem lead to dozens of investigations, publications, and several review articles just within the last years The answer to this question has to be given in three parts. First of all, there is of course the question whether a silyl cation chemistry can be established in solution phase in a similar way as this was done in the case of carbocations. There is a general chemical interest to see how... 

Finally, there is another aspect of the research on silylium cations. The question whether silylium cations are free or coordinated in solution requires to determine the structure of a solvated molecule. Presently, there is no experimental method available that can fulfil this task in a detailed and satisfactory manner. Questions concerning the geometry of the solvated ion, the number of solvent molecules or counterions in contact with the ion, the type of solvent-solute interactions, etc. cannot be answered directly by experiment. The information that is available on silylium ions in solution stems almost exclusively from NMR spectroscopy in form of chemical shift measurements. It is also possible to get additional information from X-ray structural analysis of ion-solvent complexes in the crystal state, however, the assumptions made to extrapolate from the solid state to solution phase can be as large as those used to extrapolate from the gas phase to solution phase. 

Quantum chemical calculations can provide a direct answer to this question and show whether Pauling s arguments are correct. Accordingly, we will discuss the silylium cation problem by focusing on the contribution that Quantum Chemistry can provide in this case. First, we will describe the quantum chemical methods needed for this purpose. Accordingly, Section 2 of this work is devoted to a discussion of the NMR/ab initio/IGLO method and its extension to density functional theory (DFT), namely the NMR/DFT/IGLO method. 

Figure 2. Dependence of 9Si shift on the Si-C(benzene) bond length in the Wheland c-complex (Me3Si)3Si (CgH (,)+ involving the tris-(trimethylsilyl)silylium cation and benzene. (HF-IGLO/[7s6p2d/5s4pld/3slp] calculations from Ref. 45)...

DFT/IGLO NMR chemical shifts, 3prc(Si) populations and Si charges of silylium cations 1 - 4. a... 

A silylium ion Y3Si+ with Y = NR2, OR, etc. possesses partial SiY double bonds (Scheme 3). For example, the carbocation analogues of ions 5 and 6 in Scheme 3 are considered to possess just little trivalent carbenium ion character. [87a] An amino substituent completely changes the nature of the LUMO of a silylium cation. For the silaguanidinium cation, (NH2)3Si+ (5), the next-LUMO rather than the LUMO itself possesses the expected 3pjt(Si) character. Moreover, the energies of both LUMO and next-LUMO are high, and therefore, its electrophilicity is much lower than that of an alkylsilylium cation. 

Clearly, 1) and 2) depend on each other and both define the electrophilicity of a silylium cation, which determines the chemical nature of silylium cations. 

One can argue that for the application of 1) and 2) only trends in orbital nature and orbital populations matter so that any quantum chemical method (HF or DFT) or any way of calculating charges (Mulliken or natural orbital populations) suffices for this purpose as long as it is applied in a consistent way. One can further argue that the silylium cation character of a given silyl ion has only to be determined for the gas phase. Once this has been done, other molecular properties of the ion in question can be calculated and compared with the corresponding measured values for either gas or solution phase. In this way, a purely theoretical definition of the silylium ion character could be adjusted and extended to measurable quantities. 

The approach just described can also be applied to carbenium ions R3C+, its usefulness can be tested, and at the same time a better understanding of the electronic nature of silylium cations can be provided by comparing them with the corresponding carbenium ions. For this purpose, we will shortly discuss in the next section the nature of carbenium ions in solution. 

When a silylium cation is generated from a neutral silyl compound in solution, then both reactant and product will be solvated to some extent, i.e. solvation of a silylium ion does not necessarily take place as a consequence of its generation from a neutral silyl compound in solution. Often solvation facilitates dissociation of a neutral silyl compound into silylium cation and counterion. In this case, part of the solvation shell may be carried along with the developing silylium ion in form of a solvent complex R3Si(S)n+ that possesses completely different properties than a silylium ion. One has to check in such a case whether it is still justified to speak of dissociation into silylium ion and counterion. 

A solvated molecule is rather difficult to describe since there can be specific interactions with one or more solvent molecules apart from nonspecific interactions with one or more solvent shells. Although the investigation of specific solvent complexes provides a description of major solvent effects it is far from giving an exact picture of a solvated molecule. This can only be obtained by excessive solvent modelling using Monte Carlo methods which becomes too costly when carried out for a larger number of silylium cations in different solvents. 

The solvent molecule has to be sterically demanding to make the attack of a (sterically also shielded) silylium cation more difficult. 

In the following, we will discuss the possibility of generating nearly free silylium ions under the conditions of external solvation where special emphasis is laid on stepwise introducing conditions (1) - (6) to see which effects are most important for the realization of a free silylium cation in solution. 

In the way it became clear that at least one or two of the criteria (1) - (6) had to be fulfilled to expect any success in the quest for free silylium cations in solution, one systematically started to use solvents with weak nucleophilic character as well as weakly coordinating counterions such as 33 or 35 shown in Scheme 5. The lead in this research was taken by Lambert, but Reed made also several important contributions and, therefore, one can say that both Lambert and Reed pushed forward the issue of silylium cations in solution despite scepticism and criticism on the usefulness of such work for general chemistry. Benzene and toluene were chosen as the ideal solvents and in 1993 Lambert and Zhang reported on trialkyl substituted silyl cations in aromatic solvents at the presence of TPFPB as a counterion. [7] From measured 29Si NMR chemical shifts (80 - 110 ppm), Lambert and Zhang concluded that they had obtained R3Si+ (R = Me, Et, Pr, Me3Si) ions with reduced electrophilic interactions with solvent or anion and, therefore, with nearly free cationic structure. 

---