Molecular orbital modeling of aqueous organosilicon complexes: implications for silica biomineralization

Researchers recently have proposed that hypercoordinate Si-organic complexes can form in biologically relevant fluids, and they have reported the first evidence for a transient organosilicon complex generated within the life cycle of an organism (Kinrade et al., 2001b, 2002). These interpretations are based upon peak assignments of ²⁹Si NMR spectra that invoke Si-polyol (e.g., Si-sorbitol) complexes with Si in five- and six-fold coordination states. However, ab initio analyses of the proposed organosilicon structures do not reproduce the experimentally observed ²⁹Si NMR chemical shifts (Sahai and Tossell 2001, 2002 and this work). In place of the originally proposed structures, we have modeled one of the observed δ²⁹Si values with a 5-fold Si-disorbitol complex involving 5-membered ring configurations (i.e., Si-O-C-C-O). The calculated δ²⁹Si value of this new structure closely matches the observed δ²⁹Si peaks in the −100 to −102 ppm range. Likewise, ²⁹Si NMR peaks near −144 ppm were well fit by a model complex in which a 6-fold Si was complexed to three sorbitol molecules in 5-membered ring configurations. The ability to reproduce the observed NMR peaks using molecular orbital calculations provides support for the controversial role of hypercoordinate organosilicon species in the uptake and transport of silica by biological systems. The existence of such complexes in turn may explain other puzzles in Si biogeochemistry, such as the persistence of dissolved silica in concentrated biological fluids, the biofractionation of Si isotopes, and fractionation of Ge from Si.