The atomic composition of asteroids is still uncertain presently - until we send a few robots to land there, drill and analyze, or shoot lasers or bullets at the surfacehttp://www.scienceforums.net/topic/76627-solar-thermal-rocket/page-2#entry807522http://www.scienceforums.net/topic/76627-solar-thermal-rocket/page-2#entry774819
Generally speaking, asteroids are supposed to be essentially rocks, with no hydrogen, no carbon, no nitrogen, no phosphorus - but metals and silicates yes. The presence of precious metals is very hypothetical, and stones and metals are not very attractive to make fuels.
Earth isn't very different by the way; it's just that the tiny C, N, H, P... amount is concentrated at the surface, available to life.
- A few of the outermost asteroids seem to contain much ice. They're considered comets instead (or will this distinction disappear when we better understand the history of the small bodies?)
- With this water, you can make oxygen and hydrogen for a rocket.
- Or you rocket can bring the hydrogen and produce locally the oxygen from stones (=silicates). Inconvenient but in principle feasible. Easier on Mars from atmospheric CO2.
- I checked a rocket that burns locally produced metals like Al and Li with local oxygen, and it's not efficient. Bringing hydrogen looks better.
- From a comet, an icy moon, a planetary ring, an icy asteroid - something with water - you can produce hydrogen to feed my sunheat engine
- Or you can feed my sunheat engine with water directly. Better for some missions
like bringing 400t water to Martian orbit from an asteroid with a single medium rocket launch (or 10,000t from one SLS launch).
Prospective plans at Nasa call it "in-situ propellant" (ISP, but not the Isp for specific impulse). While this would enable chemical propulsion for more ambitious missions, it's not very easy, and relies on unknown asteroid compositions - and the competition are Nasa's Vasimr engine and my sunheat engine, whose efficiency don't need in-situ propellants,