A "cool" way to remove hydrogen...and possibly a faster way to grow better crystals?

In growth of essentially every compound material such as GaN, one element always diffuses faster than the other(s) at the growth front. To grow good-quality materials, even the most sluggish element has to be sufficiently mobile, forcing materials growers to go to higher growth temperatures. Higher growth temperatures, in turn, are acompanied with undesirable drawbacks: the element(s) with intrinsically higher mobilities would desorb from the surface (higher mobility means weaker binding to the surface and easier desorption), the interfaces would be rougher because of more intermixing, etc.

To fight this fundamental limitation in growth science imposed by nature's disparity in adatom mobility, it was suggested recently that one could try to selectively inject energy by sending in photons with the right frequency so that only the slugguish adatoms can absorb them and get vibrationally excited. Once excited, these adatoms see a less corrugated energy landscape, and become more mobile. As a result, the intrinsic disparity in mobility is weakened, and better quality crystals could be expected even at much lower growth temperatures (Wu, Cohen, Feldman, & Zhang, Appl. Phys. Lett. 84, 2175 (2004)).

The validity of SEAD (selectively enhanced adatom diffusion/desorption) was verified experimentally using the important prototype system of hydrogen and deuteron desorption from a silicon surface (Liu, Feldman, Tolk, Zhang, & Cohen, Science 312, 1028 (2006)): When light of frequency corresponding to the H-Si stretching mode was shed on a Si(111) surface covered with a mixure of H and D, only H can be desorbed effectively, at ROOM temperature! In contrast, in typical growth of Si crystals via chemical vapor desorption, one has to go to ~800K in order to effectively desorb H and expose the Si surface for the incoming Si atoms to join. One therefore can dream of growing better quality Si (and other) crystals much faster and at much lower tempereatures by fully exploiting the SEAD idea to induce the right chemical processes at the growth front (see, e.g.: Tully, Science 312, 1004 (2006)).