EViews.com

how can i estimate a VECM when one of my variables is not integrated of the same order as the others?
what is the list of things i have to check to ensure my model is properly specified before estimating the VECM?

It's not possible. First you've got to run a cointegration test on your stationary variables and a condition to run that test is that your stationary variables are integrated of the same order, otherwise there is no cointegration risk whatsoever. No cointegration, no VECM.