Have you heard the story of 42.zip? It’s supposedly a compressed zip file that is only around 42 kB, but when fully uncompressed yields around 4500 TB of data. It’s an old file and there are many newer files which exceed even this limit. Yet even this is small on the scale of the internet. How much information is stored on there? Enough that they added new SI prefixes to denote 10²⁷ and 10³⁰ bytes, because they saw that 10²⁴ bytes (10¹⁵ GB) would be too small for digital information in the near future.

With that in mind, how can this data be stored? Current data centers use hard disk drives due to their low cost per byte of storage, but apparently cooling of storage systems accounts for a large portion of energy consumption in a data center. The generation of heat is attributed to ‘electron mobility and electrical resistance’ found in current storage solutions, a problem which could be avoided using ‘spin-waves’.

Spin-waves or magnons are to magnetic fields what phonons are to lattice vibrations, or photons to electromagnetic fields (they propagate as a disturbance in the magnetic field). Current manufacturing techniques are limited in their ability to generate structures for guiding spin-waves with wavelengths less than 100 nm. Block copolymers and polyoxometalates could possibly solve this by forming an orderly array of magnetically active material.

“Basil, I have to ask: do you actually understand what’s going on here?” Honestly not, dear reader. But it looks like the engineers* got the manufacturing process working, so I’ll report on it as I understand it.

The story starts with spin coating of the block copolymer PS-b-P4VP unto the substrate, either a silicon wafer or a NiFe permalloy layer atop a silicon wafer. The copolymer is deposited as a semi-ordered arrangement of spherical micelles. An aqueous solution of a polyoxometalates, with the shorthand designation ‘Fe₃₀W₇₂‘ was poured on the material, which resulted in binding of the Fe₃₀W₇₂ particles into the micelle core.

When they probed the material to gauge its properties, the authors found that the Fe₃₀W₇₂ did not exhibit magnetism at room temperatures, but exhibited antiferromagnetism at very low temperatures. There was also some interaction between the NiFe layer and the Fe₃₀W₇₂ that was not seen in the NiFe layer with just the block copolymer micelles, which was attributed to some form of ‘exchange bias’.

Based on my limited research, exchange bias is important for ensuring reliability in current hard disk drives and other magnetic storage media. It looks like they were able to find a new way to make a magnetic array, and discover something unusual about the interaction between the Fe₃₀W₇₂ and the permalloy layer. I’m not sure what course would go deep into this area (try CHEMMAT 725: Functional Materials?). Read the article or contact the authors if you’d like to learn more.

– Basil

Article link:

Nanoscale Magnetic Arrays through Block Copolymer Templating of Polyoxometalates
doi.org/10.1021/acs.nanolett.3c03825