The Omnidirectional Dielectric Mirror
December 22, 2008


I read an article on CNN last week about the recent invention of a “perfect mirror”. The article, naturally, is short on science so I found some of the original papers on the subject (here, and here, if you’re interested) and a press release from MIT to provide some context. From what I understand of it, this looks like a very significant advance.

I would recommend reading the press release first – it explains that there are two types of mirrors. Metallic mirrors (e.g., car mirrors) reflect light omnidirectionally over a wide range of wavelengths, but they absorb a few percent of the light they reflect, so they don’t work well for reflecting really intense light like lasers. Dielectric mirrors can reflect light nearly perfectly (so they can reflect laser light), but usually only one specific wavelength of light and only from one specific direction.

In 1998, Yoel Fink (then a graduate student at MIT and now a professor), figured out how to stack dielectric mirrors so they can reflect light omnidirectionally and over a wide range of wavelengths. The trick seems to be simply choosing the right combination of properties of the layers of mirror so that the math works out to produce a significant wavelength region of omnidirectional reflectivity. And everybody had just assumed this was impossible, so nobody had really looked for it before.

I should point out the evolution of the applications for this mirror. The research was funded by DARPA, the Defense Advanced Research Projects Agency, which is an awesome government program that funds far-fetched future technology and makes it reality (the Internet used to be DARPAnet, for example). This research was intended for military use with laser (!) weapons, to provide portable laser weapons for our troops or to reflect enemy laser fire (to protect a satellite or missile, say). In the 1998 press release, the scientists suggested putting the mirrors in paint for insulating buildings. And now, the most significant application seems to be medical: the mirrors can be used to make fiber-optic laser guides, so surgeons are now free to use their powerful lasers like ultra-sharp scalpels. They’ve even introduced laser pens! Nice work, Fink et al. 1998. And just goes to show how useful it is to fund pure research.