Majority carrier electrical properties in TCO determined by the method of four
coefficients utilising a new AC method to determine the Nernst coefficient.
Ian Forbes
1
, Eric Don *
2
1
Northumbria Photovoltaics Applications Centre, Northumbria University, Ellison Building,
Newcastle upon Tyne, NE1 8ST, UK
2
SemiMetrics Ltd. PO Box 36, KINGS LANGLEY, WD4 9WB, UK
* Corresponding authorE-mail:
Recently, the rapidly, expanding market demand for transparent conducting layers has
created the need to utilise more abundant and less toxic materials driving investigations of
new TCO window layers such as those based on ZnO, (ZnO:Al, ZnO:Al:Mg).
A “figure of merit” is often used to compare TCO performance, it includes the carrier
mobility as one of the properties which must be optimised. The carrier mobility in turn is
determined by the effective mass of the carrier; a property inherent to the material that is
intrinsic to the crystal structure, and the mean time between scattering; a property determined
by extrinsic factors such as defects and impurities.
Young et al have published details of a method (named
the method of four coefficients)
, for
determining these properties at a single temperature by measurement of Resistivity, Hall
coefficient, Seebeck coefficient and the Nernst coefficient on each sample in one
measurement system.
In practice we have found that the measurement of the Nernst coefficient is extremely
difficult and errors in this coefficient are major factor limiting the applicably of this method.
In this paper we report details of a measurement system for a measuring the four coefficients
which incorporates a new method for determining values of the Nernst coefficient with
improved accuracy and repeatability.
We report the use of our method of four coefficients (4C) to determine the majority carrier
electrical properties which are related to the process and film morphology e.g. grain boundary
scattering, defect scattering and also the determination of parameters related to the
fundamental semiconductor properties, for example the majority-carrier effective mass,
which would determine the upper limit of the TCO material performance.
We find that for degenerately doped TCOs, the grain boundaries play a minor role in majority
carrier transport and we can thus readily separate defect scattering mechanisms from
material- structure-limited fundamental transport parameters.
We describe our new implementation of the 4C measurement and the steps taken to correlate
our measurements with other laboratories and to introduce traceability to National Standards
from NIST.
PS2 17
-211-
