Probably I am done with proving the error estimates

Ironically, I have established a conditional stability of the used convexity splitting scheme with Fourier spectral method in space for my non-homogeneous parabolic Bi-harmonic PDE, which was proved to be unconditionally stable with respect to time step.
  Don't know whether it will be discarded or approved with special privileges.