I have to do multiple imputation where the substitutions separately for stress and control groups. Is split file the best function to use? Also how to have all 5 imputed data sets as one single data file which to be exported to excel/csv data.

SPLIT FILES will work fine, but with only two groups, you could also just use a SELECT IF. MI actually uses the SPLIT FILES mechanism, creating an imputation variable named Imputation_.

When you do the MI, you have a choice of creating a new dataset or writing directly to a sav file. Either way, you have a single data object. Of course, if you are going to do something in Excel, you need to take account of the fact that the data actually appear n+1 times (original data plus n imputed sets).