There are a lot of different types of quantum dots. Some are colloidal (dots in a liquid) - others are buried or built into materials. The fluorescent dots that you are familiar with are the colloidal ones; some are made of CdSe, ZnSe, etc, and being in a liquid medium, of course they are injectable and can be used as biological fluorescent markers. In terms of color of light emitted, the bulk material emits at some characteristic color. With QDs, as you change their size, the light emitted changes color, even though you're using the same material. Larger dots emit at a longer wavelength (redder), smaller dots at higher wavelengths (bluer).

The other type of quantum dots, the ones with photovoltaic/electronic applications, tend to be dots that are buried or grown into another solid material. The "dots" that this researcher has created are of this type - basically it seems he's managed to create individual silicon atoms on a surface that have dangling bonds in a sea of non-dangling-bond Si. The fact that the dangling bond Si atoms are far-separated from each other makes them maintain their atomic energy levels instead of having their energy levels develop into bands, as what happens in typical crystalline material. It seems like these dots were developed for quantum computing purposes and are concerned with the wave functions of the electrons, as opposed to light emission and band gap energies.

As someone who works with typical quantum dots, I find Wolkow's research interesting, but I wouldn't necessarily call what he's created a "quantum dot." Usually we are concerned with the bandgap shifting that is possible by changing the size of the dot. As I interpret his paper, it seems he's managed to create individual dangling-bond Si atoms surrounded by Si terminated by H. These dangling-bond states *handwaving explanation* seem to remain with quantized energy states instead of acting like the bulk material they're surrounded by, which have energy bands. It seems like he's interested more in electron tunneling effects for quantum computing rather than bandgap size manipulation. Maybe I'm just old-fashioned in how I think about quantum dots :p