This old post serves as the Video Companion to this video. This reaction was done some time before the video was shot, so some conditions here were a bit different (e.g. the ignition method).
This one was one of my favorite experiments to date. The silicon dioxide I used was common beach sand. I picked it up off the ground from Panama City Beach, ground it down to a finer powder in a mortar and pestile, and mixed it into thermite. This composition is extremely hard to ignite, so I added sulfur as well. This sets up a helper reaction between sulfur and aluminum that burns hot enough to sustain the rest of the thermite. The ratio I used was 9:12:10 SiO2:S:Al. This experiment was done when I was trying out different ignition methods, so here I used "thermite ignition mixture" from www.unitednuclear.com (my favorite science supplier), which was ignited with an M-80 fuse.
Two reactions take place here.
Sulfur helper reaction: 3S + 2Al -> Al2S3
Actual thermite: 3SiO2 + 4Al -> 2Al2O3 + 3Si
After the reaction, I broke apart the slag and found a number of pieces of pure silicon metal. These were caked in aluminum oxide and aluminum sulfide, and smelled awful. I treated the pieces with hydrochloric acid to remove this, since silicon is very resistant to acids. This worked beautifully and gave me some excellent samples for my element collection, shown in the last photo. The best part of all this was that I made metal from sand!
You can see the blue flame once the reaction got started, caused by burning sulfur (the smell was less pleasant). This reaction burned for a much longer time than the other compositions I've done, and much hotter as silicon's melting point is 2577 degrees Fahrenheit!