If you only need small amounts of sample you can use droplet microfluidics to make monodisperse oil-in water, or water-in-oil emulsions. With a small chip you can achieve 5 micron diameters. (this is the smallest microfluidic droplets I know of: http://pubs.acs.org/doi/abs/10.1021/nn401661d)
Membrane Emulsification can help you to get uniform emulsion droplets. The amount is higher than microfluid. The instrument is available in the market.
Thank You Martin, Titus, Louxiang, We managed to make emulsion with micelles (oil + surfactant droplets) even up to 20 um, but the size distribution was very broad. We filtered them via the membrane filter, got better but still plenty of small stuff. We are now trying membrane emulsification in our own design apparatus. Looks promising. Best, Tomek
I would use a high shear mixer to form a crude emulsion and the pass through a membrane which has holes in it rather than a mesh. By varying the hole size you ought to get to the size you require.
A small point refer to them as emulsion droplets not micelles, Micelles are aggregates of surfactant and would be around 10 nm in size