That's what they do on CSI, but that's not how it works in real life. I'm not a chemical engineer, so I can't explain why this doesn't work, but I've proposed this approach and various variations on it to some engineers for other purposes and got all sorts of explanations why it wouldn't work which I didn't understand; but the underlying concept was that you can't just take a sample of whatever and deconstruct and reconstruct it at will.
