Mask-less fabrication of areal density tunable magnetic nanoparticles with annealing of CoFeB/Cu bilayer thin film: Utilizing grain boundary diffusion

Summary form only given. This work demonstrates a mask-less fabrication method to form monodispersed magnetic nanopar-ticles from a continuous CoFeB/Cu bilayer thin film. The method manifests the capability of tuning the areal density of emerged nanoparticles by varying annealing temperature and under-layer Cu thickness. The formation of magnetic nanoparticles arises from the grain boundary diffusion of magnetic species into underneath Cu layer during the annealing process. The fabrication process started with DC magneto sputtering of CoFeB thin film onto Cu coated SiO2 substrate. The deposited continuous CoFeB/Cu bilayer film was post-annealed at high temperatures (500 or 620 °C) under ultra-high vacuum condition for hours to drive the diffusion of Co and Fe into Cu. The surface composition that was monitored by in-situ Auger electron spectroscopy (AES) at various stages of sample preparation indicates a Cu rich surface after annealing of a CoFeB/Cu bilayer film, verifying the diffusion of Co and Fe. Caused by the diffusion of Co and Fe into under-layer Cu, the magnetic species aggregated into nanoscale clusters at the interstitial sites of the concomitant grown Cu grains. The emergence of magnetic nanoparticles was illustrated by high-resolution atomic force microscopy (AFM) and magnetic force microscopy (MFM) in a chronological order in Fig. 1. The composition of the magnetic islands at Cu grain boundaries was examined by energy dispersive x-ray spectroscopy (EDS). The results indicate that Co and Fe are coalesced with their relative ratio remaining similar to the starting composition of the sputtering target, while boron is out diffused from the magnetic alloy matrices. Fig. 2 demonstrates the control of areal density of formed magnetic nanoparticles by tuning annealing temperature or under-layer Cu thickness. When the annealing temperatures remained constantly at 500 °C, the density was roughly 0.82+/-0.58 per square micron for thick Cu layer sample, 7.68+/-0.83 for the thin Cu, while the size of magnetic particle was about 256 nm+/-36 for the 280nm copper, compared with 187nm+/-44 for the 3nm Cu sample. In other words, while the density of magnetic particles varied by a factor of 8, the particle size varied by less than 60% over 2 orders of magnitude difference in Cu thickness. Therefore, the areal density of the nanoparticles is found to be strongly dependent on the Cu thickness, while the size of magnetic particles is only weakly dependent. Meanwhile, with an increment of annealing temperature from 500 to 620 °C for a thick Cu under-layer sample, the calculated areal density dropped dramatically from 0.82+/-0.58 to 0.24+/-0.13 per square micron while the size of obtained particles was almost identical to the low temperature sample. This feature reveals the significance of annealing temperature in determining the distribution of fabricated magnetic nanoparticles.