Method for identification of magnetic and mechanical features with nanoscale resolution

Yi-De Liou^1*, Yi-Hsin Weng¹, Jan-Chi Yang^1,2, Yi-Chun Chen^1,2

¹Department of Physics, National Cheng Kung University, Tainan City, Taiwan
²Center for Quantum Frontiers of Research & Technology (QFort), National Cheng Kung University, Tainan City, Taiwan

* Presenter:Yi-De Liou, email:leave710584@gmail.com

Magnetic random access memory (MRAM) that utilizes the spin-transfer torque (STT) effect has become a promising candidate for the emerging nonvolatile memory technologies due to its fast write/read speed, low power operation, long-endurance, and especially high scalability. Traditional magnetic force microscopy (MFM) offers a high resolution of ~50 nm for clearly imaging the magnetic domain structures of MRAM devices but now fails to provide a full visualization of the STT-MRAM devices when their lateral bit size scaling down towards 10 nm. In this study, we propose a new technique for magnetic stray field sensing based on a combination of PeakForce tapping (PFT) mode and principal component analysis (PCA), which not only provides an enhanced spatial resolution of magnetic imaging but also has several advantages over traditional MFM. PFT mode is utilized to periodically drive a magnetic probe tip to approach, touch, and then withdraw from the surface of magnetic films. During the process, the magnetic probe naturally oscillates and forms an instantaneous ringing signal, which has a slighter inertial characteristic and therefore potentially exhibits more magnetic sensitivity than that of traditional MFM. PCA, an unsupervised machine learning method, is employed to perform dimensionality reduction on the ringing signals collected within the sampling area. We found both the value and sign of the force acting on the magnetic probe during PFT tapping can be unambiguously extracted from the hidden signals, forming a high quality and quantitative magnetic domain image near the sample surface. Besides, information related to mechanical properties including DM modulus, surface deformation, and adhesion can be simultaneously obtained after the PCA transform. Our proposed method for magnetic measurement with simultaneous visualization of associated mechanical properties offers an efficient, convenient, and quantitative route to image magnetic nanostructures with high spatial resolution at the ambient operating environment, yielding a widely applicable approach for the development of STT-MRAM and associated magnetic devices.

Keywords: magnetic domain imaging, atomic force microscopy (AFM), magnetic force microscopy (MFM), PeakForce tapping (PFT) mode, principal component analysis (PCA)