Park Knowledge Portal

Abstract

By integrating a high-performance magnetic field generator with Park Systems’ research AFMs, in-situ characterization of magnetic materials/systems behavior at the nanoscale in the presence of both in-plane and out-of-plane magnetic fields is greatly simplified and improved. In this app-note, Artificial Spin Ice (ASI) lattice’s magnetic pole orientations are measured at various field strengths to determine the field necessary to induce partial and complete pole switching.

Introduction

Magnetic Force Microscopy (MFM) has long been used by researchers to characterize the magnetic properties of materials. This technique uses a magnetized probe and a two-pass method to decouple topography and magnetic signal contributions as seen in Figure 1. Of particular interest to many researchers studying magnetic materials are the changes they undergo when an external magnetic field is applied. While many experimental setups have been used to induce in-plane or out-of-plane magnetic fields ^1-5, the research space has lacked a flexible and commercially available magnetic field generator (MFG) option that can produce both in-plane and out-of-plane magnetic fields with high accuracy. In collaboration with Caylar, Park Systems has developed an field-switchable MFG (F-s MFG) that can be integrated with the FX and NX series of research AFMs to provide industry leading ease-of-use and control of magnetic field strength and orientation. The new F-s MFG allows for the advanced characterization of thin films with perpendicular magnetic anisotropy ^6-7, magnetic nanowires, and Artificial Spin Ice (ASI) lattices ⁸, providing insights into magnetic domain walls, vortices, skyrmions, magnetic bubbles, and switching behavior in the presence of a varying external field. This app-note will focus on ASI lattice response to an external field.

ASI lattices are ferromagnetic nanostructures usually composed of nano islands (NI’s) which are considered as macrospin elements that magnetically interact and create a geometrically frustrated ground state since the competing pairwise magnetic dipolar interactions between NI’s cannot be satisfied simultaneously ^8-11. These systems provide an excellent platform for the study of many nanomagnetic
phenomena due to the ability to adjust their magnetic properties via NI size, array geometry, and material. ASI lattices can be used to study fundamental magnetic physics but are increasingly under investigation as promising 2D magnonic conduits with reconfigurable properties for use in spin-wave computing systems ⁸.

Figure 1. A schematic representation of the dual pass lift mode MFM technique. A: (1) The AFM probe tracks the surface in non-contact mode to obtain the topography of the sample. (2) In the second pass, the AFM probe tracks the previously collected
topography line with a user defined “Lift Height”. This allows the magnetic force to dominate the tip-sample interaction, and the resulting amplitude or phase signal can provide magnetic property information. B: Example of the height and MFM information obtained during a line scan of the sample shown in “A”

Experimental

Sample Preparation

A bilayer electron beam resist, consisting of copolymer and polymethyl methacrylate (PMMA), was applied to a high-resistance silicon substrate via spin coating. The resist was then baked at 170°C for two minutes. Many artificial spin ice (ASI) array configurations were subsequently patterned using a 100 keV electron beam lithography system.

The exposed resist was developed in a solution of methyl-isobutyl-ketone (MIBK) diluted with isopropyl alcohol in a 1:1 ratio, followed by a rinse in isopropyl alcohol. Following development, a 25 nm thick layer of N_i81F_e19 (Permalloy) with a 3 nm Al capping layer was deposited using electron beam evaporation at a base pressure of 10 ^-6 mBar. The liftoff process involved immersing the processed silicon wafer in N-methyl-2-pyrrolidone (NMP) solution, which removed the resist while leaving only the nanostructure arrays on the substrate.

This application note specifically focuses on a 45° offset square ASI array “R” where each nanomagnet measures approximately 110 nm in width and 810 nm in length and covers a total area of 625 μm². Although numerous nanomagnet arrays on periodic and quasicrystal lattices were fabricated, this configuration is highlighted for detailed analysis due to its size, ease of measurement, and expected switching properties.

Figure 2. A: Lithography design layout, configuration “R” outlined in blue.
B-E: SEM images of configuration “R” at increasing magnifications, all scale bars are inset.

Materials

The data presented in this app note were collected on a Park Systems NX20 AFM in combination with Field-switchable MFG. The F-s MFG, however, is also compatible with the Park NX10, FX40, and FX200 AFMs. All images were acquired with the dual-pass lift mode MFM technique (outlined in Fig. 1) using a PPP-MFMR (NanoSensors) probe attached using the clip-type chip carrier to avoid stray
magnetic field contributions.

Measurements

The ASI lattice sample was placed on the F-s MFG configured for in-plane field generation and exposed to a -1,000 G field perpendicular to the fast-scan (X-axis) direction. This served to create a reference state for future measurements as initial experiments showed that the magnetic polarity was fully switched by ± 250 G. After resetting the reference state with a -1,000 G field, the sample was exposed to
increasing positive fields from +125 G to +220 G in order to fully characterize the pole switching behavior. All images used the same scan parameters (scan size = 5x5 μm, scan rate = 0.7 Hz, X pixels = 512, Y pixels = 256, Lift height = 20 nm).

Degree

Figure 3. MFM phase data collected at various in-plane magnetic field strengths. A: 155 G, all pairs in the initial state. B: 170 G, most pairs in a disordered partially switched state. C: 205 G, all pairs in fully switched state. The red box outlines the same pair in each image to more easily observe the changes.

Results and Discussion

Representative MFM phase images are shown in Fig. 2. While images were taken at 5 G intervals beginning at 125 G, it is observed that no switching occurred until a field of 160 G at which point some lattices exhibited partially switched states. The percentage of partially switched states increases from 160 G to 175 G (Fig. 2B) while the percentage of lattice pairs in the initial state correspondingly decreases.
By 175 G, no lattice pairs remain in the initial state and a small number have begun to fully switch (Fig. 2C). The percentage of fully switched lattice pairs then gradually increases from 180 G to 205 G when all lattice pairs have switched polarity from their initial state. The full data summary is presented in Figure 3 and a brief video of the results can be found here: 150-210 G MFM Video.mp4

Figure 4. Summary of observed switching behavior with increase in the in-plane field strength.

Figure 5. Representative lattice interface state configurations

The interface states shown in Fig 5 A-E arise from the geometrically constrained fields present in the ASI lattice. These frustrated fields at the interface region generate competing forces that cannot be satisfied simultaneously leading to the non-conformal and unstable states observed during this experiment prior to the complete pole reversal above 205 G.

Conclusions

In summary, we successfully imaged the magnetization state of the sample in the hysteretic regime. Our results demonstrate that artificial spin ices (ASIs) undergo rapid avalanches in the presence of high-energy vertices. The small switching regime (60 G) suggests minimal lattice defects, such as width variations and missing islands, introduced during sample fabrication. Furthermore, we show that
MFM images can be reliably obtained under an applied magnetic field by integrating a high-performance magnetic field generator with the Park Systems NX and FX series. This capability enables imaging of magnetic reservoir states during minor loops, which has applications benefiting the neuromorphic computing research community as well as the field of nano magnetics as a whole.