Vol 54, No 3 (2025)

Scientific novelty of this work lies in the application of well-established methods (Feret diameter method and equivalent diameter method) in a new domain – the design and development of microelectronic components. Accurate measurements of metal grain sizes and voids in TSV structures are critically important for improving the reliability and performance of micro- and nanoelectronic devices. Manual methods for analyzing the morphological characteristics of materials are time-consuming and prone to subjective errors. This paper presents an automated approach for calculating grain sizes based on the processing of scanning electron microscope (SEM) images. The methodology incorporates two approaches for calculating the average grain size: the Feret diameter method and the equivalent circle method. The correlation between the results of these methods confirms the validity of the segmentation and the high accuracy of the analysis. Experimental studies demonstrate that the proposed methodology effectively identifies grains and voids, even in low-contrast and noisy images. The results confirm the versatility, high accuracy, and reproducibility of the method, as well as its potential for integration into quality control and microelectronic system design processes. The automation of analysis significantly reduces the human factor, shortens data processing time, and opens up new opportunities for optimizing the manufacturing processes of micro- and nanoelectronic devices.

Impulse neural networks, suitable for hardware implementation based on memristors, are very promising for robotics due to their energy efficiency. However, reinforcement learning algorithms using such networks remain poorly understood. One of the key motivations for using memristors as network weights is, in addition to energy efficiency, their ability to learn (change conductivity) in real time by superimposing voltage pulses from pre- and postsynaptic signals. The article presents the results of numerical modeling of a spiking neural network (SNN) with memristive synaptic connections, which approximately solves the optimal control problem using trace variables for weight changes, allowing one to approach reinforcement learning on a true time scale. The fundamental possibility of such training in the task of holding a pole on a moving platform is shown, a comparison of various reward functions is given, and assumptions are made about ways to increase the effectiveness of this approach.

The analysis of upset occurrence in very large-scale integrated circuits (VLSI) under the influence of pulsed ionizing radiation of various nature is carried out. Upsets in VLSI under gamma or electron pulses, first of all, are determined the effects of rail span collapse, due to volume ionization of semiconductor structures. The features of upsets occurrence due to the non-stationary latchup and several possible competing effects are analyzed. Non-stationary surface radiation effects and fast annealing of radiation defects can lead, mainly, to temporary parametric failures, which depend on the dose rate. Single event effects under the influence of pulsed beams of neutrons, protons or ions are considered.

The results of a study of the temperature characteristics of a simple current mirror on high-voltage SOI nLDMOS transistors with a large drift area with topological norms of 0.5 microns in an extended range of external temperatures are discussed. The characteristics of a simple current mirror at temperatures of –60, 25, 125 °C have been experimentally studied. A mathematical model of a high-voltage nLDMOS transistor with a large DRIFT region has been developed for static operation in the field of high drain voltages and a wide range of ambient temperatures. Based on the results of experimental and numerical studies, a temperature range has been established in which the transfer characteristic of the current mirror retains linearity. It is 300 °C from –110 to 190 °C in the control voltage range from 25 to 55 V. In the same temperature range, the transmission coefficient (specularity) depends linearly on the input current level. Based on the data obtained, the conditions for determining the SOA of a simple current mirror on SOI LDMOS transistors are formulated.

Using spectral and probe diagnostic methods for the radical composition and electronic component of the removed plasma of an RF discharge in a mixture of 50 %Ar/Сl₂ /O₂, low-energy (E_i ~80 eV) etching of a nanometer-thick Ru film was studied depending on pressure, RF power, and relative content of Сl₂ /O₂. With a 10–30 percent chlorine content in the plasma, a wide maximum of the Ru etching rate is observed. In a plasma of this composition, using an array of amorphous silicon nanoconuses as a mask, vertical nanoconded Ru structures with a height of 35 nm and a distance between them of 10–20 nm were obtained. The mechanism of Ru etching in plasma of 50 %Ar/Сl₂ /O₂ is discussed.