Correction of the Concentration Profile of the
Epitaxially Grown GaAs Thin Film Layers,
Measured by Electrolytic Capacitance-Voltage
Method
Antal Ürmös, Ákos Nemcsics
Óbuda University, Institute of Microelectronics and Technology, Tavaszmező utca 17, H-1084, Budapest, Hungary
E-mails: urmos.antal @ phd.uni-obuda.hu and nemcsics.akos@kvk.uni-obuda.hu
Abstract— In this work we are dealing with the solution of a
concentration profile measurement problem in the
epitaxially grown III/V type compound semiconductor
layers. The III/V based materials have specific electronic
and optical properties, so they have large significance in the
semiconductor device production. At these materials, the
epitaxial growth is the fundamental technological step.
During this epitaxial process is possible to change the
doping concentration in the grown layer. In most cases, the
electrolytic capacitance-voltage (ECV) technique serves for
the measurement of this concentration profile. This method
combines the conventional capacitance-voltage (CV)
measurement with the anodic oxidation, to avoid the
electrical break-down at the junction. The disadvantage of
this method is that the contact surface increases during the
measurement, because of the dissolution of the material.
This surface growth - in certain layer structure - causes
error in the measurement. In present work, we give a
computational solution for this problem by our software,
which we have developed in Matlab™. This algorithm
calculates and summarizes the wall capacitance and
subtract from the measured data, in each step. We
demonstrate the operation of the program on different
sample data series.
I. INTRODUCTION
The new applications of semiconductors (for instance
microwave and optical communications) directed the
researchers attention to the family of III/V compound
semiconductors (e.g. GaAs, InP etc.). Many favorable
properties (e.g.: direct band transition, higher charge
carrier mobility) allows us to make such electronic
devices, which we cannot prepare from silicon or we can
produce only with lower parameters.
In the technology of III/V-based compound
semiconductors, the epitaxial growth is one of the most
important step. Usually, in the silicon-based device
technology, the differently doped layers are made by
process of diffusion. This method is not applicable in case
of compound semiconductors, because of thermal
decomposition. During the epitaxial growth is possible to
change the doping concentration of the growing layer, so a
proposed concentration profile can be created. However
not only the dopant, but the components of the
semiconductors are also changeable. Hereby, the
heterostructures of these materials can be produced by this
method as well.
In the different semiconductor technologies, the various
qualifying measurements have very important role. These
studies give us a chance to verify, for instance how we
realized the required dopant distribution layer. For this
goal, in most cases we apply the CV measurement. The
disadvantage of this method is an electrical break-down of
the junction, when the critical field intensity is reached.
This effect can be avoided by ECV measurement, which
essence is, that the semiconductor is contacted with an
electrolyte, what is operated like a quasi-Schottky diode
[1]. The layer removal is achieved by anodic dissolution
of the semiconductor. Nevertheless there is another
problem, because the wall capacitance of the etched range
falsifies the measured results.
In this article we are looking for the solution of this
problem using our software, which we have developed in
Matlab™. This algorithm calculates and summarizes the
wall capacitance in each step and subtract from the
measured data. We demonstrate the operation of the
program on different sample data series. In this work, we
show the simulation results on a dataset, which contains
8000 measured sample data.
II. DESCRIPTION OF THE MEASURING TECHNIQUE
Because of the technological feedback, we need to
verify the epitaxially grown layer. One of the most
important parameter is the concentration profile. By this
property, we can satisfy, whether the realization of the
required dopant distribution is successful or not.
The most widely used method to measure the dopant
distribution is the CV measurement, because the depth
distribution of the free charge carriers is dependent from
the capacitance, which is strongly related to dopant
profile. Unfortunately, the measurement depth is limited,
because of the electrical break-down of the junction, when
the critical electrical field intensity is reached. At first - in
order to try to avoid this phenomenon – the researchers
etched the layer after the measurement, and measured the
CV characteristics again. The problem of this method, that
the measured results are hardly and inaccurately matched
to each other and this work is time-consuming.
In 1973, Ambridge and his colleagues suggested
another method. The technique is that the semiconductor
is contacted with an electrolyte, which operates as a quasi-
Schottky diode. This examination method is the ECV
measurement. Compared to other methods, there is an
CINTI 2012 • 13th IEEE International Symposium on Computational Intelligence and Informatics • 20–22 November, 2012 • Budapest, Hungary
137 978-1-4673-5206-2/12/$31.00 ©2012 IEEE