Nanoparticles are small chunks of matter whose sizes are typically <100 nanometers. To give you a sense of how tiny these particles can be: we are talking about sizes that are roughly one-thousandth the breadth of your hair! Do you think you can see such tiny particles using your eye? Usually study of such tiny particles require microscopes that are much better than the ones you have used in middle and high school. Microscopes that you used in your high school biology lab are called optical microscopes. Study of nanoparticles often require electron microscopes, which are much more powerful, and hence capable of probing much smaller chunks of matter.

Gallium nitride particles made using ammonothermal methods are a few microns large (about one-tenth to one-hundredth the breadth of your hair). In principle, these particles can be broken down to give nanoparticles that are about 20-30 nm. The idea we had was to use rare earth doped GaN made using ammonothermal methods (remember: these materials are excellent light emitters) as the starting material to make nanoparticles. A popular technique for making nanoparticles from these micron-sized particles is essentially by "chopping" them repeatedly. The advantage of such chopping techniques (formally called "nanoscission") is that it is easily scalable, and hence commercially viable. However we found that such chopping methods resulted in loss of light emitting property; however it did result in particles of the right size regime. We found that nanoscission techniques, which are very successful and hence popular for making catalysts on an industrial scale, very often fail when it comes to optical materials.

In order to circumvent the problem, we turned to another popular industrial process called ball milling. It turned out that ball milling could retain the emissive properties of RE:Ga, as opposed to nanoscission techniques. This might be because nanoscission techniques results in flatter particles, with much higher surface area per unit mass of the materials. In general, these extra surfaces provide opportunities for excitons (Coumbically bound electron-hole pairs, generated when you sufficiently excite the optical material) to recombine, without transferring their energy to the light emitting ion (rare earth ion, in this case). Hence we found that physical particle size reduction of optical materials is better done, when one uses the ball-mill approach, as opposed to the nanoscission method.

We were eventually able to engineer the ball-mill process to achieve Er doping of GaN. Er doping in GaN is particularly tricky, since Er levels get easily quenched in GaN matrix. For example, while making Er:GaN using ammonothermal synthesis, the slightest amount of oxygen in the reaction chamber can completely quench the Er transitions. However we found that samples made using our ball-mill approach gave Er emissions that were much more stable that Er:GaN obtained using chemical methods!

Please note that these physical methods have yields approaching 100%, which make these techniques viable for large scale industrial production of nano-GaN based phosphors.

Ref: Journal of Crystal Growth Vol. 311, 4402-4407 (2009).

MRS Proceedings http://dx.doi.org/10.1557/PROC-1202-I09-12