BEGIN:VCALENDAR VERSION:2.0 PRODID:-//CERN//INDICO//EN BEGIN:VEVENT SUMMARY:Influence of the variable cross section on the domain wall dynamic s in curved ferromagnetic nanostripes DTSTART;VALUE=DATE-TIME:20240117T105000Z DTEND;VALUE=DATE-TIME:20240117T111000Z DTSTAMP;VALUE=DATE-TIME:20260419T013511Z UID:indico-contribution-63-266@indico.bitp.kiev.ua DESCRIPTION:Speakers: Dmytro Karakuts (Taras Shevchenko National Universit y of Kyiv\, 01601 Kyiv\, Ukraine)\nCurvilinear magnetism is an emerging fi eld which connects low-dimensional magnetic architectures with their geome trical properties to describe resulting novel phenomena. Nanoscale curved magnetic stripes are a crucial building block in the race for high-density and ultrafast magnetic memory\, and can substantially augment the perform ance of spintronic devices [1\,2].\n\nHere\, we aim to theoretically and n umerically study how changes in the cross-section area of the curved ferro magnetic nanostripe affect its magnetic structure. Namely\, we show that t ailoring of the gradient of the cross-section area can be used to control the motion of the domain wall. Although\, the magnetic response of the cur ved nanostripe is already well-studied\, in all previous studies cross sec tion of the nanowire was assumed to be constant. Only recently\, a new mic romagnetic framework emerged that allows the exploration of the changes in the magnetization dynamics caused by the cross-section area gradient [3]. \n\nHere\, we use the collective variable approach by adapting the q-Ф mo del to stripes with varying cross section. The description of the domain w all motion in ferromagnetic wires is investigated by considering the tempo ral evolution of the magnetization vector based on the Landau–Lifshitz –Gilbert equation. To derive the specific equations of motion for the co llective variables\, the Lagrange-Rayleight formalism is used\, where the dependency on the spatial variation of the cross-section area is factored in both the Lagrangian and the dissipative function.\n\nThe resulting equa tions of the domain wall motion are studied for three specific geometries: a straight stripe\, an arc of a circle and an Euler spiral. For the case of a straight stripe with absent curvature\, we show that the domain wall velocity is mainly determined by the cross-section gradient. For the case of constant curvature\, the asymptotic domain wall velocity is shown to ha ve two components: the first component is identical to the speed of the do main wall in the straight wire\, while the second one describes the coupli ng between the curvature and the cross-section gradient. For the case of t he wire\, whose curvature grows at a constant rate\, the asymptotic domain wall velocity has two contributions. The first one is the same as the vel ocity in the straight wire\, however\, the second one differs from the pre vious case and is determined solely by the rate of curvature growth. All a nalytical predictions are well­confirmed by the NMAG micromagnetic simula tions [4].\n\nThe gradient of the cross-section area of the curved ferroma gnetic nanostripe creates an internal driving force\, which in turn induce s the domain wall motion and can explain the recent experimental results [ 5]. We have shown how the dynamics of the domain wall motion can be modifi ed through tailoring the cross-section area of the nanostripe. Although we showcased our approach only for three specific geometries\, it is general and can be applied to a wide class of geometries.\n\n1. Curvilinear Micro magnetism: From Fundamentals to Applications\, edited by D. Makarov and D. Sheka\, Topics in Applied Physics Vol. 146 (Springer\, Cham\, 2022).\n2. D. D. Sheka\, O. V. Pylypovskyi\, O. M. Volkov\, K. V. Yershov\, V. P. Kra vchuk\, and D. Makarov\, Small 18\, 2105219 (2022).\n3. K. V. Yershov\, D. D. Sheka\, Phys. Rev. B 107\, L100415 (2023).\n4. T. Fischbacher\, M. Fra nchin\, G. Bordignon and H. Fangohr\, IEEE Transactions on Magnetics\, Vol . 43\, No. 6\, (2007).\n5. L. Skoric et al. ACS Nano 16\, 8860 (2022).\n\n https://indico.bitp.kiev.ua/event/11/contributions/266/ LOCATION:BITP & Zoom URL:https://indico.bitp.kiev.ua/event/11/contributions/266/ END:VEVENT BEGIN:VEVENT SUMMARY:Nonlinear wave phenomena in Josephson junctions DTSTART;VALUE=DATE-TIME:20240117T100500Z DTEND;VALUE=DATE-TIME:20240117T105000Z DTSTAMP;VALUE=DATE-TIME:20260419T013511Z UID:indico-contribution-63-287@indico.bitp.kiev.ua DESCRIPTION:Speakers: Yaroslav Zolotaryuk (Bogolyubov Institute for Theore tical Physics of the National Academy of Sciences of Ukraine)\nhttps://ind ico.bitp.kiev.ua/event/11/contributions/287/ LOCATION:BITP & Zoom URL:https://indico.bitp.kiev.ua/event/11/contributions/287/ END:VEVENT BEGIN:VEVENT SUMMARY:Spatial coherence of compact sources with respect to X-ray phase c ontrast imaging formation DTSTART;VALUE=DATE-TIME:20240117T092500Z DTEND;VALUE=DATE-TIME:20240117T094500Z DTSTAMP;VALUE=DATE-TIME:20260419T013511Z UID:indico-contribution-63-277@indico.bitp.kiev.ua DESCRIPTION:Speakers: Artur Ovcharenko (Institute of Applied Physics of th e NAS of Ukraine)\nSignificant successes and important results of research on synchrotron radiation sources raised a new urgent question\, namely th e development of bright compact radiation sources with coherence parameter s that allow the implementation of phase-informative methods. This is due to the growing demand for the use of new research methods based on X-ray r adiation in engineering\, materials science and medicine. The study concer ns the topical issue of implementation of diffraction research methods on compact radiation sources [1]. Taking into account phase information helps to increase the sensitivity and informativeness of the obtained images\, making their use more effective in a wide range of research and diagnostic tasks.\n\nThe aim is to determine the research capabilities of various ty pes of installations in solving specific applied problems\, as well as to develop applied calculation methods for planning and conducting X-ray phas e contrast experiments. X-ray phase contrast arises due to the difference in the phase of X-rays when they pass through objects with different refra ctive or absorption indices. One of the approaches for modeling X-ray phas e contrast is the Fresnel-Kirchhoff scalar diffraction theory [2-4]. This study presents a method for calculating phase-contrast images of test obje cts. A two-stage algorithm for calculating the X-ray phase contrast image from sources of different shapes and sizes has been developed. The evoluti on of the wave front along the source-object and object-screen path is tak en into account. The source-object and object-screen distances at which it is possible to observe phase-contrast images in the Fresnel region have b een found. Simulated images of test objects were obtained in the form of i ntensity contrast on the screen and its phase profile. As the calculations show\, increasing the thickness of certain sections of the test object le ads to changes in the phase distribution and intensity of the resulting im age. This is manifested in the form of the appearance of intensity maxima at the edges of the object\, where there is a fast change in its thickness \, as well as minima\, where the thickness is the largest\, which is expla ined by a greater deviation of X-rays from their original direction.\n\nTh e results of the research will be useful for the development of methods fo r studying the internal structure of materials\, as well as for solving pr oblems in medical diagnostics\, where X-ray phase contrast can help distin guish tissues with different physical properties\, for example\, in the st udy of soft tissues and organs.\n\n1. L. Quenot\, S. Bohic and E. Brun\, A ppl. Sci.\, 12(19)\, 9539\; (2022)\nhttps://doi.org/10.3390/app12199539\n2 . D. Paganin Coherent x-ray optics. Oxford University Press\, 2013\, 424 p p.\n3. V. Kolobrodov\, G. Timchyk\, Applied Diffraction Optics: Textbook\, K.: NTUU "KPI"\, 2014\, 312 pp.\n4. A. Pil-Ali et al.\, Proc. SPIE 11595\ , Medical Imaging 2021: Physics of Medical Imaging\, 115951N (15 February 2021)\; doi: 10.1117/12.2582225\n\nhttps://indico.bitp.kiev.ua/event/11/co ntributions/277/ LOCATION:BITP & Zoom URL:https://indico.bitp.kiev.ua/event/11/contributions/277/ END:VEVENT BEGIN:VEVENT SUMMARY:The Mutual Influence of Production Rate of Point Defects and Tempe rature on the Effects of Radiation-Induced Segregation in the Fe-20Cr-8Ni Alloy DTSTART;VALUE=DATE-TIME:20240117T090500Z DTEND;VALUE=DATE-TIME:20240117T092500Z DTSTAMP;VALUE=DATE-TIME:20260419T013511Z UID:indico-contribution-63-274@indico.bitp.kiev.ua DESCRIPTION:Speakers: Roman Skorokhod (Institute of Applied Physics of Nat ional Academy of Sciences of Ukraine\, Sumy\, Ukraine)\nIn this paper\, a computer simulation of radiation-induced segregation in the Fe-20Cr-8Ni al loy was carried out within the framework of the modified inverse Kirkendal l model [1-3]. The dependences of the concentrations of Cr and Ni atoms on the surface of the sample on the rate of production of point defects and temperature\, as well as the concentration profiles of Cr and Ni atoms and point defects\, were calculated. The influence of the rate of production of point defects and temperature on the concentration profiles of componen ts and point defects is analyzed in detail. It is shown that the effect of the increase in the rate of production of point defects on the concentrat ion profiles of the alloy components can be compensated with a sufficientl y high accuracy by a relatively small increase in temperature (fig. 1). Th is effect is explained by the competition of two mechanisms: (i) increasin g the irradiation of the alloy leads to a narrowing of the concentration p rofiles\; (ii) an increase in temperature leads to a broadening of concent ration profiles.\n\nFig. 1. Concentration profiles of Cr and Ni in the Fe- 20Cr-8Ni alloy under irradiation.\n\n1. G.S. Was\, *Fundamentals of Radiat ion Materials Science. Metals and Alloys* (2nd ed.) (New York: Springer: 2 017).\n2. R.V. Skorokhod\, O.M. Buhay\, V.M. Bilyk\, V.L. Denysenko and O. V. Koropov\, *East European Journal of Physics*\, **5**\, No. 1: 61–69 ( 2018) (in Ukrainian).\n3. R.V. Skorokhod and O.V. Koropov\, *Metallofiz. N oveishie Tekhnol.*\, **44**\, No. 6: 691—711 (2022) (in Ukrainian).\n\nh ttps://indico.bitp.kiev.ua/event/11/contributions/274/ LOCATION:BITP & Zoom URL:https://indico.bitp.kiev.ua/event/11/contributions/274/ END:VEVENT BEGIN:VEVENT SUMMARY:Revealing inaccuracies of the GFN1-xTB semi-empirical method with machine-learning toolset DTSTART;VALUE=DATE-TIME:20240117T084500Z DTEND;VALUE=DATE-TIME:20240117T090500Z DTSTAMP;VALUE=DATE-TIME:20260419T013511Z UID:indico-contribution-63-275@indico.bitp.kiev.ua DESCRIPTION:Speakers: Andrii Terets (1) Faculty of Physics\, Taras Shevche nko National University of Kyiv. 2) Chuiko Institute of Surface Chemistry \, NAS of Ukraine\,)\nThe change in conformation during biological functio ning is a characteristic feature of biomolecules. The presence of various conformers complicates their physics-based modeling\, as the properties of each conformer need to be determined separately. Moreover\, the relative energies of the conformers become an additional quantity of interest becau se they determine the probability of each conformation in a real system.\n \nAlthough ab initio quantum-mechanical methods are the most accurate for determining the physical properties of molecules\, their computational cos ts become prohibitive when applied to biomolecules with a large number of atoms and/or a large number of conformations. Semi-empirical quantum mecha nical methods\, particularly those based on the tight-binding approximatio n in density functional theory (DFTB)\, significantly reduce computational requirements by utilizing parameterized models for the elements of the Fo ck or Kohn-Sham matrices. However\, due to such simplifications\, the resu lting method can show low accuracy in some cases\, despite performing accu rately in other cases.\n\nIn this work\, we demonstrate that machine learn ing (ML) methods such as graph neural networks and ensemble models can be employed to predict errors in determining the relative energies of conform ers by the semi-empirical method GFN1-xTB from the DFTB family\, based sol ely on the molecular structural formula as input.\n\nThe developed ML ense mble model achieves an accuracy of about 80% in determining whether a give n molecule is 'challenging' for GFN1-xTB. This approach can be used to fin d the molecules that most vividly highlight the shortcomings of the physic al model underlying the GFN1-xTB method. We demonstrate that for the 190 m olecules selected using the developed model\, the average error in relativ e energies of their conformers obtained by the GFN1-xTB method is 4.2 kcal /mol\, in contrast to 1.8 kcal/mol when the same number of test molecules are selected randomly. This indicates that the ML toolset indeed allows fo r the identification of challenging molecules and can therefore be useful in improving the approximations utilized by semi-empirical methods.\n\nhtt ps://indico.bitp.kiev.ua/event/11/contributions/275/ LOCATION:BITP & Zoom URL:https://indico.bitp.kiev.ua/event/11/contributions/275/ END:VEVENT BEGIN:VEVENT SUMMARY:Application of machine learning approaches to the generation of or ganic compounds and prediction of their properties DTSTART;VALUE=DATE-TIME:20240117T080000Z DTEND;VALUE=DATE-TIME:20240117T084500Z DTSTAMP;VALUE=DATE-TIME:20260419T013511Z UID:indico-contribution-63-288@indico.bitp.kiev.ua DESCRIPTION:Speakers: Maksym Druchok (Institute for Condensed Matter Physi cs of the National Academy of Sciences of Ukraine)\nhttps://indico.bitp.ki ev.ua/event/11/contributions/288/ LOCATION:BITP & Zoom URL:https://indico.bitp.kiev.ua/event/11/contributions/288/ END:VEVENT END:VCALENDAR