OK. You have known values for the standard cure you built, and you say you have good R-squared. That means you have areas for your results. How will you compare your unknown samples to your known?
Consider: the 10 ng gives an area of 1000, the 5 ng gives an area of 500, now what? If your area of the breath sample is 500 or 1000, you'll know exactly what the answers is. But that's unlikely to happen.
There's more work than this, depending on your level. But you should be able to start here.