I would use the 100 million to investigate possible ways of mitigating the effects of such a crash. As the actual causes would be essentially unpredictable I'd pass on that effort and concentrate on what we could do to establish a degree of continuity of core areas of knowledge: maths, medicine, geography etc.
One option would be to build bunkers full of useful books in remote locations (like the gene bank in the Svalbard). However, a better way might be to build a long lasting organisation set apart from our culture - very much like secular versions of medieval monasteries or the mathic concents in Neal Stephenson’s excellent Anathem.