Features of the Implementation of Methods for a Comprehensive Study of Properties of Thermoelectric Materials Roman Dunets1, Bogdan Dzundza1,2, Liliia Turovska3, Nazariy Senkiv2 1 Lviv Polytechnic National University, Bandera str., 12, Lviv, 79013, Ukraine 2 Vasyl Stefanyk PreCarpathian National University, Shevchenko str., 57, Ivano-Frankivsk, 76018, Ukraine 3 Ivano-Frankivsk National Medical University, Halytska str. 2, Ivano-Frankivsk, 76018, Ukraine Abstract Methods for the implementation of software and hardware tools for comprehensive nondestructive research of thermoelectric parameters of semiconductors have been analyzed and adapted. An information-measuring system has been developed, in which, due to a combination of different research methods, it is possible to perform the whole complex of thermoelectric measurements in one technological cycle and on one sample of typical configuration, in particular, thermo-EMF, electrical conductivity, Hall coefficient, magnetoresistance, Nernst-Ettingshausen coefficient, thermal conductivity, and thermoelectric figure of merit. The use of digital algorithms for filtering and processing the received data made it possible to obtain important parameters that are difficult to measure directly, in particular, the mobility and concentration of charge carriers, parameters of near-surface layers, to reconstruct the profiles of the distribution of thermoelectric parameters over the thickness. An important advantage of these methods is the absence of the need for accurate measurements of heat fluxes, which greatly simplifies and reduces the time for conducting experimental studies. Keywords 2 Thermoelectric properties, information-measuring systems, measurement methods, computer tools, signal processing, defects identification 1. Introduction specialization of thermoelectric research. Therefore, an urgent task is the adaptation of methods and the development of tools both for the Thermoelectric materials are becoming more study of the main thermoelectric parameters of widespread as simple and reliable energy semiconductor materials and for express methods converters, but their efficiency is still quite low. for determining the operating characteristics of Therefore, a large number of studies are carried thermoelectric energy conversion modules. out aimed at increasing the efficiency of thermoelectric materials. Such studies require measurements of electrical conductivity, Seebeck 2. Selection and adaptation of coefficient, thermal conductivity, which, when measurement methods using classical techniques, is a rather laborious task, since samples of various configurations and When studying thermoelectric and accurate measurement of heat fluxes are required. photoelectric properties, it becomes necessary to A large number of universal tools for laboratory measure a sufficiently large number of quantities research have been developed, but their effective of different nature (electrical conductivity, use is not always possible due to the narrow ISIT 2021: II International Scientific and Practical Conference «Intellectual Systems and Information Technologies», September 13–19, 2021, Odesa, Ukraine EMAIL: roman.b.dunets@lpnu.ua (A. 1); bohdan.dzundza@pnu.edu.ua (A. 2); lturovska@ifnmu.edu.ua (A. 3); senabo33@gmail.com (A. 4) ORCID: 0000-0002-3325-7908 (A. 1); 0000-0002-6657-5347 (A. 2); 0000-0002-3530-7518 (A. 3) ©️ 2021 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). CEUR Workshop Proceedings (CEUR-WS.org) Seebeck coefficient, thermal conductivity, analysis of the "mobility spectrum" [1]. This efficiency, etc.) depending on various factors method consists in the fact that the components of (temperature, film thickness, production the conductivity tensor are represented in the form parameters, type of substrate), which makes such of integral equations depending on the experiments quite laborious. concentration and mobility of charge carriers. Since the preparation of a sample makes up Such methods require a decrease in the most of the labor costs in thermoelectric research, measurement error of the Hall voltage and it is urgent to develop an automated computer magnetoresistance, since in the presence of noise system that will allow combining direct and in the experimental data it is difficult to correctly indirect research methods on one sample. In determine the search area for the parameters. particular, the implementation of the Hall For a more visual analysis, we have methods will make it possible to study the constructed a model of the measuring channel, in galvanomagnetic, temperature, thickness and time particular, Fig. 1 shows a model of the measuring dependences of the properties, in particular, to channel in the study of the galvanomagnetic determine the electrical conductivity, the properties of the material, taking into account the concentration of charge carriers, the Seebeck influence of parasitic effects, such as EMF of coefficient, and magnetoresistance. nonequipotentiality, EMF of magnetoresistive Despite the rather slow processes in the effect, thermo-EMF ΣUi(B,I,T), distortions implementation of Hall methods, a number of caused by electronic nodes and inaccuracies of difficulties arise associated with low signal levels actuators G(B,I,T), as well as the influence of and the large influence of parasitic effects. external noise and interference η(B,I,T). Spectral methods of analysis are especially sensitive to signal noise, in particular, methods of Figure 1: Model of the measuring channel in the study of the galvanomagnetic properties of the material. As a result, the measured Hall voltage will A special cryostat design, where the sample is consist of a useful signal and noise, according to clamped between massive copper plates [2], and the expression the measurements are carried out at a minimum 𝑈𝐻(𝐵, 𝐼, 𝑇) = 𝑈𝐻 + (1) sufficient current, at which there is no noticeable +𝑆𝑖(𝐵, 𝐼, 𝑇)𝐺(𝐵, 𝐼, 𝑇) + 𝜂(𝐵, 𝐼, 𝑇), heating of the sample, the power released must not where UH(B,I,T) is the Hall voltage signal with exceed a few mW, has been developed to deal both linear and nonlinear distortion. A similar with the occurrence of uncontrolled temperature situation will be in the study of other gradients. Low currents lead to a decrease in the thermoelectric and photoelectric properties useful signal, but they can effectively deal with associated with voltage measurement. the uncontrolled heating of the sample. In thermoelectric materials, parasitic effects The automatic compensation circuitry has that make the main contribution to the error been designed to eliminate the error associated include the EMF of nonequipotentiality, with the nonequipotentiality voltage. In the thermoelectric and thermomagnetic effects, absence of a magnetic field and a given current which, even with a small change in the flowing through the sample, the temperature gradient, significantly distort the nonequipotentiality voltage is measured by the result due to the large coefficient of thermo-EMF. analog-to-digital converter ADC and compensated using the compensation circuit on magnetoresistance do not change with time, but the operational amplifier, which is guided by the with a change in the magnitude of the magnetic voltage from the DAC. induction, the speed of the experiment, and In the developed hardware-software complex, therefore the maximum frequency of the several stages of data filtering are performed, in measured signal, can be controlled. This system particular, at the first stage, the signal is amplified, can significantly reduce the noise component in brought to the ADC range and passed through the the signal (Fig. 2). hardware low-pass filter. After digitizing the Both modeling of the effect of noise on the signal, a median filter is applied to the data results and experimental studies on real test obtained from the ADC. This filter gets rid of the samples have been carried out to study the random spikes associated with the operation of the effectiveness of noise control and the effect of ADC. The third step is to apply a digital low-pass digital filtering on the parameter determination filter to the entire measured relationship. To error. As a result of using the developed system, reduce the noise component in the measured data, it was possible to significantly reduce the error in including the quantization noise level, a digital determining the concentration and mobility of low-pass filter with a finite impulse response charge carriers, in particular in the presence of based on the Blackman weight function is used several types of charge carriers, the error in [3]. The use of a low-pass filter is due to the fact determining the concentration and mobility of that the measured signals change at a rather low heavy holes decreased by 4 times, light holes and frequency, tenths and hundredths of hertz. electrons – by 1.5-2 times. Considering that the Hall voltage and a b Figure 2: Dependence of the Hall voltage on the magnetic field: noisy (a) and after digital filtering (b). The design of the cryostat provides for the measure thermal conductivity, heat capacity, presence of a gradient heater and a differential thermoelectric figure of merit, and other thermocouple to determine the thermo-EMF quantities. The combination of direct and indirect coefficient and the Nernst-Ettingshausen methods makes it possible to determine electrical coefficient by direct methods. Along with direct conductivity, carrier concentration, Seebeck methods, indirect methods based on the modified coefficient, Nernst-Ettingshausen coefficient, Harman method [4] and impedance spectroscopy magnetoresistance, thermal conductivity, as well [5,6] are implemented in the measuring complex as thermoelectric figure of merit, and to carry out for the complete characterization of express diagnostics of finished thermoelectric thermoelectric material and automated express energy conversion modules on one sample in one diagnostics of thermoelectric elements. These technological cycle. methods are indirect measurement methods and are favorably distinguished by a short experiment 3. Hardware and software tools for time and do not require complex and laborious measurements of heat fluxes through the sample. research implementation This combination of methods made it possible to solve two main problems of classical methods, The set of methods for complex nondestructive namely, the need for accurate measurement of research of thermoelectric parameters of heat fluxes through the sample and the need to semiconductors determines the characteristics of prepare samples of various configurations to the system that it implements, namely, the amount of input information, the speed of information operating systems. When constructing a system receipt and the processing time of the input for complex nondestructive research of information. thermoelectric parameters of semiconductors, it is The process of measuring the parameters of the necessary to determine the solution of which sample determines the maximum amount of input problems will be implemented in hardware, and information and the maximum rate of its entry into which ones – by specialized and universal tools. the system. The largest amount of input The criterion for this can be the comparison of the information will be in the implementation of given time for solving the problem and the time galvanomagnetic methods, and the maximum data for solving it by hardware or software and input speed will be in the implementation of the hardware tools, which is given in [7]. impedance spectroscopy and will be no more than Taking this into account, the system that 20х106 counts/s. ensures the implementation of the set of these The minimum processing time of the input methods is implemented as a three-level information is determined by the process of specialized computer system (Fig. 3), optimized controlling the operating conditions of the test to obtain the maximum number of parameters that sample, that is, the formation of control signals to fully characterize the sample without destroying it the actuators, and the shortest time between two and without the need to change the configuration control signals will be when the Harman pulse of the sample for various research methods. This method is implemented and will be 0.1 μs. concept, combined with modular structuring That is, the system that ensures the techniques, makes it possible to design an implementation of the set of these methods will information-measuring and control system that not only be information-measuring, but also can be easily upgraded or expanded. At the lower control at the same time. On the other hand, the level, the actuators are controlled to create the implementation of the system should optimally necessary conditions for the experiment; sensor combine the known approaches to the polling, analog processing and filtering of signals construction of computer systems – purely are carried out. At the middle level, digital signal hardware implementation and software and processing, control signal generation, and self- hardware implementation. The hardware diagnostics are performed. Depending on the implementation provides maximum performance, required performance, this level can be but it requires redesigning each time in case of implemented both on a single microcontroller and changes in the system's algorithm. The hardware on an FPGA or their combination. The upper level and software implementation as a whole has a is implemented in software on a standard personal lower performance, but its operation can only be computer, which made it possible to develop a reprogrammed when the operation algorithm is convenient graphical interface for control and changed. In turn, software and hardware tools are visualization of results. In addition, the transfer of divided into universal and specialized – all calculations and simulations to the PC makes microprocessor tools that work without operating it easy to expand the capabilities of software data systems, and microcomputer tools that run under processing without interfering with the hardware. Figure 3: General concept of a specialized computer system for comprehensive thermoelectric and photoelectric research. Guided by the principles of modular parameters by analyzing the scatter of the structuring, all the implemented methods for received data, which reduces the probability of studying the properties of semiconductors are receiving incorrect data and outliers. combined into groups and implemented as For automated processing and visualization of separate subsystems. the results, a software analytical module has been In particular, the subsystem for studying the developed for the application of physical and galvanomagnetic properties of semiconductors mathematical models to determine the main implements the classical methods of Hall thermoelectric parameters, including those that measurements in constant magnetic fields, cannot be measured directly, such as the measurements of thermo-EMF, electrical concentration and mobility of charge carriers, and conductivity, Hall coefficient, magnetoresistance, the reconstruction of the profiles of these and Nernst-Ettingshausen coefficient. parameters over the sample thickness [9,10]. The subsystem for express studies of Approximation and fitting were carried out by the thermoelectric elements implements pulse least squares method using the algorithm for methods and, for sufficiently thin film samples, minimizing functions of many variables by the requires high frequencies and, accordingly, high Nelder-Mead deformed polyhedron method. The speed of the system for generating and processing analytical module also provides automatic signals. In addition, a highly stable source of both decoding of spectrograms and determination of DC and AC of 10 μA to 500 mA, with a frequency thermoelectric parameters that are difficult to of up to 2 MHz [7], and voltage measurement measure directly, for example, thermal from 1 μV up to 1 V with a resolution of 12 bits, conductivity and thermoelectric figure of merit, as up to 100 Mps with noise filtering and taking into well as automatic diagnostics and defects account errors from parasitic physical processes identification of thermoelectric elements. has been implemented. Also, based on the obtained data and the adaptive algorithm, which 4. Conclusions compares not only the absolute values, but also their deviations from those typical for a given The analysis has been carried out, methods for series, defects identification of the studied element with the determination of the probable studying the thermoelectric properties of type of defect is provided [7]. semiconductors have been selected and adapted, The generation and synchronous detection of and a computer system has been implemented, signals with a frequency of up to 2 MHz and their which makes it possible to obtain all the necessary mathematical processing based on fast Fourier parameters of the test sample in one technological transforms to determine the amplitude and phase cycle on a sample of one configuration by nondestructive methods, which several times shift between them have been implemented for the reduces the time for preparing and conducting an subsystem for studying thermoelectric properties based on impedance spectroscopy [8]. The main experiment. task of fast digitalization of an analog signal is A special design of a cryostat and a sample solved using a high-speed analog-to-digital holder, an algorithm and a circuit for converter, for example, AD9643 (Analog compensation of nonequipotential voltage have Devices, USA), which has two independent high- been developed, which makes it possible to quality sample-hold devices and a built-in minimize parasitic effects during Hall reference voltage source. Generation of a signal of measurements. It is shown that the use of digital the required frequency, filtering and mathematical signal filtering algorithms to effectively reduce processing of data have been implemented on the the noise component in the measured data has FPGA, and a 32-bit microcontroller has been made it possible to reduce the error in determining chosen for signal generation, control of switching the concentration and mobility of charge carriers and operation of operational amplifiers, and by 2-4 times. communication with a computer [7,8]. The use of indirect methods for studying For analyzing the quality of contacts and self- thermoelectric properties made it possible to diagnostics, algorithms and a subsystem for avoid the need to measure heat fluxes, to automated analysis of the ohmicity of contacts implement algorithms for fast diagnostics of have been developed, in particular, for analysing thermoelements and to reduce the error in the I–V curve for linearity by software methods, measuring thermoelectric figure of merit by 1.5-2 times. for detecting breakage and instability of contact A decrease in the laboriousness of processing impedance spectroscopy of thermoelectric the obtained data has been achieved by energy converters, ISTCMTM 81/4 (2020) developing software tools for automated data 18–24. doi:10.23939/istcmtm2020.04.018 preprocessing in accordance with physical and [9] R. Dunets, B. Dzundza, M. Deichakivskyi, mathematical models that describe thermoelectric Methods of computer tools development for properties and make it possible to determine measuring and analysis of electrical parameters that cannot be measured directly. properties of semiconductor films, Eastern- European Journal of Enterprise Technologies 5. References 103, 1/9 (2020) 32–38 doi:10.15587/1729- 4061.2020.195253 [10] Y.P. Saliy, B.S. Dzundza, I.S. Bylina, O.B. [1] Gang Du, J.R. Lindemuth, Characterizing Kostyuk, The influence of the technological Multi-Carrier Devices with Quantitative factors of obtaining on the surface Mobility Spectrum Analysis and Variable morphology and electrical properties of the Field Hall Measurements, Japanese Journal PbTe films doped Bi, Journal of Nano- and of Applied Physics 41 (2002) 1055–1064. Electronic Physics 8/2 (2016) 02045 doi: doi:10.1143/JJAP.41.1055. 10.21272/jnep.8(2).02045. [2] B. Dzundza, L. Nykyruy, T. Parashchuk, E. Ivakin, Y. Yavorsky, Z. Dashevsky, Transport and thermoelectric performance of n-type PbTe films, Physica B: Physics of Condensed Matter 588 (2020) 412178–1–8. doi:10.1016/j.physb.2020.412178 [3] A.V. Trifanov, D.Yu. Protasov, V.Ya. Kostyuchenko, The noise suppression in results of automatic hall measurements by using low-frequency digital-data filter, Izvestiya vuz. Fizika 56 (2013) 134–136 (in Russian). [4] B. Beltrán-Pitarch, J. Prado-Gonjal, A.V. Powell, Experimental conditions required for accurate measurements of electrical resistivity, thermal conductivity, and dimensionless figure of merit (ZT) using Harman and impedance spectroscopy methods, Journal of Applied Physics 125/2 (2019) 025111. doi: 10.1063/1.5077071. [5] J. De Boor, E. Müller, Data analysis for Seebeck coefficient measurements, Review of scientific instruments 84 (2013) P. 065102–1–9. doi:10.1063/1.4807697. [6] Jorge Garcıa-Canadas, Gao Min, Impedance spectroscopy models for the complete characterization of thermoelectric materials, Ournal of Applied Physics 116 (2014) 174510-1-7. doi: 10.1063/1.4901213 [7] R. Dunets, B. Dzundza, L. Turovska, M. Pavlyuk, Features for the design of a specialized information-measuring system for the study of thermoelectric properties of semiconductors, Eastern-European Journal of Enterprise Technologies 110/2 (2021) 23– 31. doi:10.15587/1729-4061.2021.227135 [8] R. Dunets, B. Dzundza, O Kostyuk, Specialized software and hardware for