The simulation results of the temperature distribution in the growth area of graphene layers obtained by the method of thermal decomposition of the silicon carbide surface substrates in setup with induction heating are presented. The heating parametrs of the setup elements are calculated using the commercial package COMSOL Multiphysics taking into account the electrical, thermal and magnetic properties of the materials from which the growth plant elements are made. A numerical estimate of the heating inhomogeneity of silicon carbide plates over its area during the growth of graphene layers at a given temperature is given. It is shown that the lateral temperature distribution over the area of the plate has radial symmetry with decreasing values towards the center.

Graphene quality for use in biosensors was assessed in fabricated chips by a set of methods that includes atomic force microscopy (AFM), Raman spectroscopy, and low-frequency noise. It is shown that local areas of residues on the graphene surface, which arisen due to the interaction of graphene with a photoresist at the initial stage of chips development, leads to a spread of chips resistance (R) within 1-10 kΩ and to an increase in the root mean square (RMS) roughness up to 10 times that can significantly impair reproducibility of graphene parameters in biosensors chips. It was observed that the control of the photoresist residues after photolithography (PLG) by AFM and subsequent additional cleaning allow reducing the spread of R values in chips to 1–1.6 kΩ and obtaining RMS roughness similar to the roughness in pristine graphene before PLG. Monitoring of the spectral density of low-frequency voltage fluctuation (SU), which provides integral information about the defect system and quality of the material, makes it possible to identify chips with low graphene quality and with inhomogeneously distributed compressive stresses areas by the type of frequency dependence SU (f).

The quality of graphene intended for use in biosensors was assessed on manufactured chips using a set of methods including atomic force microscopy (AFM), Raman spectroscopy, and low-frequency noise investigation. It is shown that local areas of residues on the graphene surface, formed as a result of the interaction of graphene with a photoresist at the initial stage of chip development, led to a spread of chip resistance (R) in the range of 1–10 kOhm and to an increase in the root mean square (RMS) roughness up to 10 times, which can significantly worsen the reproducibility of the parameters of graphene chips for biosensor applications. It was observed that the control of the photoresist residues after photolithography (PLG) using AFM and subsequent additional cleaning reduced the spread of R values in chips to 1–1.6 kOhm and obtained an RMS roughness similar to the roughness in the graphene film before PLG. Monitoring of the spectral density of low-frequency voltage fluctuation (SU), which provides integral information about the system of defects and quality of the material, makes it possible to identify chips with low graphene quality and with inhomogeneously distributed areas of compressive stresses by the type of frequency dependence SU(f).