THURSTON, G. D. & SPENGLER, J. D. 1985. A Quantitative Assessment of Source Contributions to Inhalable Particulate Matter Pollution in Metropolitan Boston. Atmospheric Environment (1967), 19, 9-25.
from my experience and following the reply from Mark, you have to do a regression analysis with respect to the PM mass. With only the variance you may have an idea on how much important is your factor to explain your database. But this is not necessarily linked to its weight in terms of mass concentration = source contribution.
PCA would give you qualitative (source factors) rather than quantitative results (percentage contribution). You should perhaps try to use EPA PMF which is window base and easy to run. If you put the sample mass in the ambient data file with a large uncertainty (4 times of the mass itself, recommended by Prof. Hopke) then you can have the contributions in percent or in absolute value (mass) directly.