METHOD FOR MEASURING THE GAS-SENSITIVE CHARACTERISTICS OF SEMICONDUCTOR SENSORS BY DYNAMIC RESPONSE PARAMETERS

Abstract

Detection of dangerous gases and vapors is relevant both at work and in everyday life. Sensor
elements based on semiconductor structures are highly sensitive to gases of a very different
nature. To increase the accuracy of measurements, calibration and the speed of concentration
determination, special methods are required, such as signal processing in various ways. The study
uses gas sensors based on silicon–carbon films that are sensitive to a whole set of gases. In the
first part of the article, general problems and a method for solving them are considered, which
makes it possible to increase the selectivity of semiconductor gas sensors. The analysis of dynamic
parameters, such as the first and second derivatives of the response curves, as well as the analysis
of the Elovich equation is carried out. Such calibration dependences, constructed from the extremes
of the derivatives and the slope coefficients of the Elovich equation, show high linearity.
As a promising solution, it is proposed to use a set of calibration lines to determine the concentration
of the target gas. The developed method allows, using dynamic response parameters, to determine
the gas type and its concentration using a single sensor, also to increase the accuracy of
measurements, along with a reduction in detection time. The experimental results of data processing
with the determination of the gas type and its concentration using the described method
are presented. For a sensor based on silicon-carbon materials, the developed method and algorithms
made it possible to carry out measurements using a single sensor element for a set of gases
(NO2, CO, SO2) with different concentrations. At the same time, the smallest relative error did not
exceed 3.6% for SO2, 2.7% for NO2, 2% for CO. The distinctive features of the developed method
are the use of several calibration lines in a multidimensional space, rather than one, as well as an
original algorithmic signal processing methods.