In principle it might be possible to generate SK parameters for SiO2 and HfO2 separately, but I doubt one could get something that is even remotely accurate for the interface between them. DFTB might be an option, but the effort to fit parameters will be large. My recommendation would be to go with a cheap DFT-LCAO basis set like SingleZetaPolarized. Now, you don't write what properties you intend to compute, but obviously if the band gap is of importance, you don't want to rely just on GGA, but HSE will be very time-consuming, so for that I would recommend the TB09 MetaGGA approach.