pimiПриборы и методы измеренийDevices and Methods of Measurements2220-95062414-0473BNTU10.21122/2220-9506-2026-17-1-77-86pimi-1034Research ArticleМетоды измерений, контроля, диагностикиMethods of measurements, monitoring, diagnosticsМашинное обучение в расшифровке начального взаимодействия зонда атомно-силового микроскопа с поверхностьюMachine Learning in Understanding the Initial Interaction of the Atomic Force Microscope Probe with the SurfaceМорозовИ. А.MorozovI. A.

Адрес для переписки:Институт механики сплошных сред УО РАН,ул. Академика Королёва, 1, г. Пермь 614013,Россияimorozov@icmm.ru

Address for correspondence:Institute of Continuous Media Mechanics, Ural Branch of the Russian Academy of Sciences,Academician Korolev str., 1, Perm 614013,Russia imorozov@icmm.ru

imorozov@icmm.ruИнститут механики сплошных сред Уральского отделения Российской академии наукРоссияInstitute of Continuous Media Mechanics Ural Branch of the Russian Academy of SciencesRussian Federation2026010420261717786Copyright © Морозов И.А., 20262026Морозов И.А.Morozov I.A.Данная работа распространяется под лицензией Creative Commons Attribution 4.0.This work is licensed under a Creative Commons Attribution 4.0 License.https://pimi.bntu.by/jour/article/view/1034

Достоверное определение начала контакта зонда атомно-силового микроскопа с поверхностью в силовых измерениях необходимо для изучения структурных и физико-механических свойств материалов. В экспериментах в воздушной среде переход от бесконтактного взаимодействия к контактному сопровождается резким ускоренным движением острия к поверхности. В случае деформируемой поверхности высокая скорость изгиба кантилевера на участке бесконтактного взаимодействия, с одной стороны, и относительно низкая частота записи данных атомно-силового микроскопа, с другой стороны, не позволяют определить начало контакта из эксперимента, опираясь на характерные точки силовой кривой. Для решения этой проблемы предлагается использовать алгоритмы машинного обучения, «натренированные» на решении множества модельных задач. Взаимодействие острия с поверхностью моделировали гармоническим осциллятором, варьируя параметры зонда, материала и динамические условия эксперимента. Использование разработанных алгоритмов продемонстрировано на примере обработки результатов индентирования полиэтилена. Полученные контактные отклонения не совпадают с имеющимися точками экспериментальных кривых.

Reliable determination of the tip-surface contact in atomic force microscopy measurements is necessary for structural and physico-mechanical analysis of the surface properties. Transition from non-contact interaction to contact one during experiments in air environment is accompanied by a rapid jump of the tip to the surface. High velocity of the tip movement in the area of non-contact interaction and the relatively low rate of the atomic force microscopy data capture do not allow determining of the onset of contact from the points of the force curve especially in the case of a deformable surface. The proposed solution is to use machine learning algorithms trained on model results. The interaction of the tip with the surface was modeled using a harmonic oscillator varying parameters of the probe, the material, and the experiment. As a result deflection of the probe in the moment of contact is predicted using input experimental parameters. Use of the developed algorithms is demonstrated by treating the results of the indentation of polyethylene. The obtained contact deflections are significantly differ from the available points of the experimental curves.

атомно-силовая микроскопияиндентированиеконтактмашинное обучениеatomic force microscopyindentationcontactmachine learningРабота выполнена при поддержке Российского научного фонда и Министерства образования и науки Пермского края (соглашение № С-26/1774 от 12.05.2025): грант № 25-21-20086.The work was supported by the Russian Science Foundation and the Ministry of Education and Science of the Perm Region (agreement No. C-26/1774 dated May 12, 2025): grant No. 25-21-20086.ReferencesWang D, Fujinami S, Nakajima K, Nishi T. True Surface Topography and Nanomechanical Mapping Measurements on Block Copolymers with Atomic Force Microscopy. Macromolecules. 2010;43(7):3169-3172. DOI: 10.1021/ma9028695Wang D, Fujinami S, Nakajima K, Nishi T. True Surface Topography and Nanomechanical Mapping Measurements on Block Copolymers with Atomic Force Microscopy. Macromolecules. 2010;43(7):3169-3172. DOI: 10.1021/ma9028695Dehnert M, Spitzner E-C, Beckert F, Friedrich C, Magerle R. Subsurface Imaging of Functionalized and Polymer-Grafted Graphene Oxide. Macromolecules. 2016;49(19):7415-7425. DOI: 10.1021/acs.macromol.6b01519Dehnert M, Spitzner E-C, Beckert F, Friedrich C, Magerle R. Subsurface Imaging of Functionalized and Polymer-Grafted Graphene Oxide. Macromolecules. 2016;49(19):7415-7425. DOI: 10.1021/acs.macromol.6b01519Morozov IA. Subsurface AFM Study of Inhomogeneous Polymeric Materials. Journal of Applied Polymer Science. 2025;142(11):e56611. DOI: 10.1002/app.56611Morozov IA. Subsurface AFM Study of Inhomogeneous Polymeric Materials. Journal of Applied Polymer Science. 2025;142(11):e56611. DOI: 10.1002/app.56611Stühn L, Fritschen A, Choy J, Dehnert M, Dietz C. Nanomechanical sub-surface mapping of living biological cells by force microscopy. Nanoscale. 2019;11(27):1308913097. DOI: 10.1039/C9NR03497HStühn L, Fritschen A, Choy J, Dehnert M, Dietz C. Nanomechanical sub-surface mapping of living biological cells by force microscopy. Nanoscale. 2019;11(27):1308913097. DOI: 10.1039/C9NR03497HGisbert VG, Garcia R. Fast and high-resolution mapping of van der Waals forces of 2D materials interfaces with bimodal AFM. Nanoscale. 2023;15(47):1919619202. DOI: 10.1039/D3NR05274EGisbert VG, Garcia R. Fast and high-resolution mapping of van der Waals forces of 2D materials interfaces with bimodal AFM. Nanoscale. 2023;15(47):1919619202. DOI: 10.1039/D3NR05274EMorozov IA, Izumov RI. Influence of experimental conditions on apparent AFM tip-surface contact in air. Ultramicroscopy. 2025;273:114148. DOI: 10.1016/j.ultramic.2025.114148Morozov IA, Izumov RI. Influence of experimental conditions on apparent AFM tip-surface contact in air. Ultramicroscopy. 2025;273:114148. DOI: 10.1016/j.ultramic.2025.114148Cao Y, Yang D, Soboyejoy W. Nanoindentation Method for Determining the Initial Contact and Adhesion Characteristics of Soft Polydimethylsiloxane. J. Mater. Res. Springer Science and Business Media LLC. 2005;20(8):2004-2011. DOI: 10.1557/jmr.2005.0256Cao Y, Yang D, Soboyejoy W. Nanoindentation Method for Determining the Initial Contact and Adhesion Characteristics of Soft Polydimethylsiloxane. J. Mater. Res. Springer Science and Business Media LLC. 2005;20(8):2004-2011. DOI: 10.1557/jmr.2005.0256Garcia M, Schulze KD, O’Bryan CS, Bhattacharjee T, Sawyer WG, Angelini TE. Eliminating the surface location from soft matter contact mechanics measurements. Tribology – Materials, Surfaces & Interfaces. 2017;11(4):187-192. DOI: 10.1080/17515831.2017.1397908Garcia M, Schulze KD, O’Bryan CS, Bhattacharjee T, Sawyer WG, Angelini TE. Eliminating the surface location from soft matter contact mechanics measurements. Tribology – Materials, Surfaces & Interfaces. 2017;11(4):187-192. DOI: 10.1080/17515831.2017.1397908Yang C-W, Chen C-H, Ding R-F, Liao H-S, Hwang I-S. Multiparametric characterization of heterogeneous soft materials using contact point detectionbased atomic force microscopy. Applied Surface Science. 2020;522:146423. DOI: 10.1016/j.apsusc.2020.146423Yang C-W, Chen C-H, Ding R-F, Liao H-S, Hwang I-S. Multiparametric characterization of heterogeneous soft materials using contact point detectionbased atomic force microscopy. Applied Surface Science. 2020;522:146423. DOI: 10.1016/j.apsusc.2020.146423Hadjiiski L, Linnemann R, Stopka M, Oesterschulze E, Rangelow I, Kassing R. Application of neural networks to a scanning probe microscopy system. Thin Solid Films. 1995;264(2):291-297. DOI: 10.1016/0040-6090(95)05851-6Hadjiiski L, Linnemann R, Stopka M, Oesterschulze E, Rangelow I, Kassing R. Application of neural networks to a scanning probe microscopy system. Thin Solid Films. 1995;264(2):291-297. DOI: 10.1016/0040-6090(95)05851-6Liu Y, Sun Q, Lu W, Wang H, Sun Y, Wang Z, Lu X, Zeng K. General Resolution Enhancement Method in Atomic Force Microscopy Using Deep Learning. Advanced Theory and Simulations. 2019;2(2):1800137. DOI: 10.1002/adts.201800137Liu Y, Sun Q, Lu W, Wang H, Sun Y, Wang Z, Lu X, Zeng K. General Resolution Enhancement Method in Atomic Force Microscopy Using Deep Learning. Advanced Theory and Simulations. 2019;2(2):1800137. DOI: 10.1002/adts.201800137Li M, Rieck J, Noheda B, Roerdink JBTM, Wilkinson MHF. Stripe noise removal in conductive atomic force microscopy. Sci Rep. 2024;14(1):3931. DOI: 10.1038/s41598-024-54094-wLi M, Rieck J, Noheda B, Roerdink JBTM, Wilkinson MHF. Stripe noise removal in conductive atomic force microscopy. Sci Rep. 2024;14(1):3931. DOI: 10.1038/s41598-024-54094-wNartova AV, Mashukov MYu, Astakhov RR, Kudinov VYu, Matveev AV, Okunev AG. Particle Recognition on Transmission Electron Microscopy Images Using Computer Vision and Deep Learning for Catalytic Applications. Catalysts. 2022;12(2):135. DOI: 10.3390/catal12020135Nartova AV, Mashukov MYu, Astakhov RR, Kudinov VYu, Matveev AV, Okunev AG. Particle Recognition on Transmission Electron Microscopy Images Using Computer Vision and Deep Learning for Catalytic Applications. Catalysts. 2022;12(2):135. DOI: 10.3390/catal12020135Anantatamukala A, Krishna KVM, Dahotre NB. Generative adversarial networks assisted machine learning based automated quantification of grain size from scanning electron microscope back scatter images. Materials Characterization. 2023;206:113396. DOI: 10.1016/j.matchar.2023.113396Anantatamukala A, Krishna KVM, Dahotre NB. Generative adversarial networks assisted machine learning based automated quantification of grain size from scanning electron microscope back scatter images. Materials Characterization. 2023;206:113396. DOI: 10.1016/j.matchar.2023.113396Giergiel M, Zapotoczny B, Czyzynska-Cichon I, Konior J, Szymonski M. AFM image analysis of porous structures by means of neural networks. Biomedical Signal Processing and Control. 2022;71:103097. DOI: 10.1016/j.bspc.2021.103097Giergiel M, Zapotoczny B, Czyzynska-Cichon I, Konior J, Szymonski M. AFM image analysis of porous structures by means of neural networks. Biomedical Signal Processing and Control. 2022;71:103097. DOI: 10.1016/j.bspc.2021.103097Kim Y, Gu GH, Asghari-Rad P, Noh J, Rho J, Seo MH, Kim HS. Novel deep learning approach for practical applications of indentation. Materials Today Advances. 2022;13:100207. DOI: 10.1016/j.mtadv.2022.100207Kim Y, Gu GH, Asghari-Rad P, Noh J, Rho J, Seo MH, Kim HS. Novel deep learning approach for practical applications of indentation. Materials Today Advances. 2022;13:100207. DOI: 10.1016/j.mtadv.2022.100207Weber A, Vivanco M. dM, Toca-Herrera JL. Application of self-organizing maps to AFM-based viscoelastic characterization of breast cancer cell mechanics. Sci Rep. 2023;13(1):3087. DOI: 10.1038/s41598-023-30156-3Weber A, Vivanco M. dM, Toca-Herrera JL. Application of self-organizing maps to AFM-based viscoelastic characterization of breast cancer cell mechanics. Sci Rep. 2023;13(1):3087. DOI: 10.1038/s41598-023-30156-3Sotres J, Boyd H, Gonzalez-Martinez JF. Locating critical events in AFM force measurements by means of one-dimensional convolutional neural networks. Sci Rep. 2022;12(1):12995. DOI: 10.1038/s41598-022-17124-zSotres J, Boyd H, Gonzalez-Martinez JF. Locating critical events in AFM force measurements by means of one-dimensional convolutional neural networks. Sci Rep. 2022;12(1):12995. DOI: 10.1038/s41598-022-17124-zJeong K, Lee H, Kwon OM, Jung J, Kwon D, Han HN. Prediction of uniaxial tensile flow using finite element-based indentation and optimized artificial neural networks. Materials & Design. 2020;196:109104. DOI: 10.1016/j.matdes.2020.109104Jeong K, Lee H, Kwon OM, Jung J, Kwon D, Han HN. Prediction of uniaxial tensile flow using finite element-based indentation and optimized artificial neural networks. Materials & Design. 2020;196:109104. DOI: 10.1016/j.matdes.2020.109104Bao M, Yang H. Squeeze film air damping in MEMS. Sensors and Actuators A: Physical. 2007;136(1):3-27. DOI: 10.1016/j.sna.2007.01.008Bao M, Yang H. Squeeze film air damping in MEMS. Sensors and Actuators A: Physical. 2007;136(1):3-27. DOI: 10.1016/j.sna.2007.01.008Clifford CA, Seah MP. The determination of atomic force microscope cantilever spring constants via dimensional methods for nanomechanical analysis. Nanotechnology. 2005;16(9):1666-1680. DOI: 10.1088/0957-4484/16/9/044Clifford CA, Seah MP. The determination of atomic force microscope cantilever spring constants via dimensional methods for nanomechanical analysis. Nanotechnology. 2005;16(9):1666-1680. DOI: 10.1088/0957-4484/16/9/044Payam AF, Morelli A, Lemoine P. Multiparametric analytical quantification of materials at nanoscale in tapping force microscopy. Applied Surface Science. 2021;536:147698. DOI: 10.1016/j.apsusc.2020.147698Payam AF, Morelli A, Lemoine P. Multiparametric analytical quantification of materials at nanoscale in tapping force microscopy. Applied Surface Science. 2021;536:147698. DOI: 10.1016/j.apsusc.2020.147698

The authors declare that there are no conflicts of interest present.