Quantum chemical studies on lowcoordinated group 14-16 compounds have been performed. This thesis focuses particularly on silenes influenced by reverse Si^δ-=C^δ+ bond polarization. Hypercoordinated carbon compounds are also studied.

The geometries from calculations with several common computationally inexpensive methods have been tested against high level CCSD/cc-pVTZ geometries for a series of substituted silenes. Hybrid HF/DFT methods performed best among the inexpensive methods tested for silenes.

Heavy alkenes strongly influenced by reverse polarization are found to have less exothermic dimerization energies for both head-to-head and head-to-tail dimerizations, and to have higher activation energies for water addition than naturally polarized heavy alkenes.

We also investigated solvated lithium, magnesium and potassium silenolates and found that lithium and magnesium ions coordinate preferably to O, giving their SiC bond some double bond character.

Reverse polarized 2-siloxy-, 2-thiosiloxy-, and 2-(N-sila-N-methyl)-silenes could according to calculations be formed thermolytically from the corresponding tetrasilanes as transient species. It was, however, found that silenes highly influenced by π-conjugative reverse polarization have low barriers for the back-reaction, and thus these silenes are more difficult to form as stable species than naturally polarized silenes.

It is also found that conjugated 1-siladienes, formed by electrocyclic ring-opening of 1-silacyclobut-2-enes, which are highly influenced by π-conjugative reverse polarization, have higher barriers for electrocyclization back to starting material than naturally polarized 1-siladienes.

It is found that CHe₅⁴⁺, CHe₆⁴⁺, CNe₅⁴⁺, and CNe₆⁴⁺ are the closest carbon analogs of SiH₅^-, SiH₆^2-, SiF₅^- and SiF₆^2-, respectively. However, due to their exothermic dissociation reaction, these very high-lying local minima will be impossible to reach experimentally.