Things like this are part of why I think the requirement for 10 thousand year containment of the materials at nuclear waste sites is a silly one:
http://environment.newscientist.com/channel/earth/mg19726396.200-pacman-molecule-chews-up-uranium-contamination.html?feedId=online-news_rss20 Note: this isn't even true nanotechnology. In fact, it's almost a classical-chemical approach to removing the contaminant, though it does rely on the molecule's shape to some extent.