Barium ferrite thick disks have been fabricated for incorporation in self-biased microwave devices. The process uses barium ferrite nanopowder mixed with epoxy. The mix is placed on alumina substrates to fabricate 500 μm thick disks of BaM to achieve good magnetic properties such as magnetic saturations from 2,000 to 2,500 G and coercivities fom 3800 to 4000 Oe. In addition, the barium ferrite thick disks have been incorporated into microwave devices and successfully deposited into copper contact lines for testing and use.

Each barium ferrite thick disk is made from a combination of commercially available barium ferrite nanopowder and 30-minute epoxy with a ratio of 80/20, respectively. The barium ferrite nanopowder is weighed and placed into the mixing container; the epoxy is then weighed and added to the mix. To help with the mixing process, 1 ml of acetone is added to the mix. This first part of the mixing is done for two minutes and the goal is to incorporate all the material together to form a dough-like compound. The dough is mixed by hand for another two minutes to achieve a uniform, homogeneous composition. A small piece of dough is placed in a mechanical press containing 1" magnets.

The mixed material is placed in the substrate and carefully pressed, then another piece of the mixed substance is added on top. The press is closed carefully. Each press contains two 1" magnets, with one on top and one on bottom. The function of the magnets is to align the magnetic particles of the mixed substance so that their c-axes are uniformly parallel and oriented. The samples are kept in the high magnetic field for 48 hours, and then released and cured at room temperature for another 48 hours.

After the curing process is completed, each sample is polished using a precision lapping/polishing machine. The polishing process is done in several steps with the lapping machine spinning at 250 rpm:

  1. Samples are polished with 320 waterproof sandpaper to remove the excess material.
  2. After all the excess material is removed, a polishing cloth is placed in the polishing machine. Three different polishing slurries are sequentially used — 3.0 micron, 0.5 micron, and a glass polish 0.05 micron. Each of the slurries is used for six minutes for each side.
  3. After polishing, the samples are placed on a hot plate at 450°C for 24 hours to dry.

During the drying step, some gaps develop between the ferrite disk and substrate. To remedy this problem, an epoxy filler is prepared and placed into the gaps using a thin blade; this particular step requires extreme care. It is important that the epoxy not be applied inside the via holes that are used as part of the electrical circuit. Once all the excess material is removed, the samples are placed again on the hot plate for 24 hours for curing.

Once the samples are cured, each one is then cleaned with acetone and alcohol. They are subsequently dried with compressed air. To achieve a uniform surface for copper patterning, each sample is placed in the spin coater and coated with a UV adhesive. The spin process lasts about one minute at a rate of 4,000 rpm. Samples are then exposed in a 30mW/cm2 UV light for 15 minutes and placed on a hot plate at 50°C for 12 hours. The same process is performed on both sides of the substrate.

The substrate is placed in a vacuum chamber with the target, and an inert gas, such as argon, is introduced at low pressure, typically in the range of 10-100m Torr. Gas plasma is created by using a power source, which causes the gas to ionize. The ions are accelerated towards the surface of the target causing atoms of the source material to break off from the target, which then vaporize and condense on all surfaces including the substrate. Difference in pressure and voltage typically relate to the manner in which the ion bombardment of the target is realized.

The ferrite samples are characterized using a vibrating sample magnetometer (VSM). Two measurements are taken: one at a 0-degree angle (out of plane) and one at a 90-degree angle (in plane). Because the hexagonal substrates cannot be placed on the VSM to be measured, circular samples are made at the same time from the same ferrite compound as the original samples. The 80/20 ratio had the best overall results. A 90/10 ratio was also tested, but the mixing process was difficult and the uniformity of the compound was compromised.

This work was done by Jeffrey L. Young of the University of Idaho for the Office of Naval Research.