Have you checked the DECHEMA data book series for experimental VLE data? This is a voluminous set of books that lists all experimental data reported in the open literature that have been measured up to some date preceding the publication. These books are expensive, but most respectable libraries carry them. Also, they include VLE regression results for the more common activity coefficient models.
I feel almost certain that the binary mixtures in your system have been measured. Once you get the data, you would re-regress the binary interaction parameters (in Aspen Plus, for your chosen activity coefficient model - I would assume you're considering Wilson, NRTL, or UNIQUAC), since the vapor pressures might not quite agree with those used to regress the data in the DECHEMA books. The multicomponent VLE calculations are based solely on the regressed binary VLE parameters.
If, by some bizarre chance, the data for the constituent binaries is not available, all hope is not lost. You could use the UNIFAC group contribution method to estimate the missing VLE. This is available as an option in Aspen Plus; however, as you should know, UNIFAC estimates of infinite dilution activity coefficients can be off by 30% or more and, therefore, your column simulation may be significantly in error.
