Surface and bulk passivation of multicrystalline silicon solar cells by ...

84
Jdark(V) = J01i.exp{(-eV/kT) - 1 } + J02i{exp (-eV/2kT)-1 }
The saturation currents J01 and J02 can be written in standard forms for a Ρ/N
junction. The total current, J, is given by:
J = Jphi - Jdarki(V)
where Jphi and Jdarki (V) are the photogenerated and the dark-current densities,
respectively, and i corresponds to either a defect-free cell element or a cell element with
defects [115, 116]. The values of Jphi, J0ii, and J02i can be estimated from experimental
measurements. For example, we select one cell and make an estimate of J ph values for
defect-free cells and cells with defects based on LBIC (long wavelength) responses and
cell I- V plots. However, J01 and Joe cannot be determined from the cell itself. A library of
J01 and J02 values for a variety of materials and for different defect densities is available.
It uses a diode array technique that has been described in the literature [117]. Edgepassivated,
mesa diode arrays are fabricated on wafers and their electrical characteristics
are probed. The device characteristics and their defect data are compiled and used as
input in the model. The output of the model generates terminal I- V characteristics of the
total cell and spatial distribution of cell voltages and currents for any terminal voltage.
These sets of data result in excellent agreement between calculated and actual terminal
characteristics of the large-area cell (as seen in next section). It should be pointed out that
the network model assumes no internal carrier transport—the communication between
the devices occurs via a highly conducting emitter region and the bus bar.
5.4.2 Experimental Approach
The major objective of the experimental work is to fabricate solar cells on wafers of
known distribution of defect clusters and compare the cell characteristics with theoretical