| AVS 54th International Symposium | |
| Electronic Materials and Processing | Wednesday Sessions |
| Session EM-WeM |
| Session: | Contacts, Interfaces and Defects in Semiconductors |
| Presenter: | R.L. Opila, University of Delaware |
| Authors: | M.Z. Burrows, University of Delaware U.K. Das, University of Delaware R.L. Opila, University of Delaware R.W. Birkmire, University of Delaware |
| Correspondent: | Click to Email |
The search for an ideal surface passivation layer of crystalline silicon (c-Si) to be employed in the silicon heterojunction photovoltaic device has garnered much attention. The leading candidate is a few nanometers thick amorphous silicon (a-Si:H) film. This is due to the fact high open circuit voltages above 700mV, key to 20% power conversion efficiencies, are only possible with low surface recombination velocities at the passivated c-Si / a-Si interface. Our approach involves a concentrated effort to link the deposition parameters to the thin a-Si:H material properties as revealed with Fourier transform infrared spectroscopy (FTIR) and match observed changes in H bonding to passivation quality as determined by effective minority carrier lifetime measurements. Reported dependencies of film surface passivation quality on substrate preparation, orientation, and deposition temperature have been extended in this work to include H to SiH4 dilution ratio and post-deposition annealing. Marked differences have been observed with carrier lifetimes ranging from few microseconds to few milliseconds. A simple yet extremely sensitive FTIR procedure based on Brewster angle transmission measurements enables the probing of films of just 5-10nm thickness. By cataloguing the changes in H content and bonding environment as hydrogen dilution or annealing conditions were varied a comprehensive picture of material quality as related to passivation quality has emerged. Simple avoidance of the growth regimes that lead to epitaxial growth of Si on the c-Si substrate produces decent lifetimes on the order of 500μsec can be achieved. However this often entails harsh deposition conditions that lead to defective films of primarily bulk SiH2 bonding. Subsequent low temperature anneals, presumed to only relax the amorphous lattice, are shown to not only cause unexpected bulk hydrogen evolution but also involve various complex reactions. Annealing in atmosphere changes surface SiH2 to chainlike (SiH2)n as well as oxygen back-bonding to create surface oxides like SiH(O3). Annealing in vacuum causes minimal surface SiH2 disturbance but rather surface SiH and SiH3 evolution. Finally it is concluded that the best passivation layer consists of primarily well ordered mono-hydride bonding in purely amorphous phase. Sub-optimal amorphous phase films can be improved by post-deposition anneal.