The plasma chemistry of an expanding thermal plasma in argon/hydrogen mixtures in interaction with silane is discussed. Based on Langmuir probe measurements and (appearance potential) mass spectrometry, it is argued that under optimal conditions in terms of the a-Si:H deposited the fluxes reaching the growing film mainly consist of and atomic hydrogen. The surface chemistry, hydrogen incorporation, growth rate and dependence on substrate temperature are discussed and the results obtained are compared with growth models as proposed by Matsuda, Gallagher and Perrin. It is argued that although atomic hydrogen plays no role in the modification of the bulk properties, atomic hydrogen is probably responsible for the creation of dangling bond sites by hydrogen abstraction. Possible consequences for the substrate temperature dependence of the growth rate are discussed.