Differences in energy calculation results for single molecules

[Soma]

Structural optimization calculations of ibuprofen were performed based on the reference literature, but there were discrepancies in the values（mine:-116.8 Hartree，reference:-656.2 Hartree）. I want to know the reason. I am sharing the condition of DFT calculation below.

【my condition】

Software:CP2K

Functionals:PBE

Basis Functions:DZVP-MOLOPT-SR-GTH

Solvent effect:None

SCF Convergence:1.0E-4 Hartree

Energy Convergence:1.0E-14 Hartree

Maximum number of iteration for SCF calculation:400

Periodic boundary:None

Molecule modeling soft:Avogadro

【Reference】

Software:Materials Studio(v7.0)

Functionals:GGA (with PW91)

Basis Functions:The double numeric polarization (DNP)

Solvent effect:water (COSMO)

SCF Convergence:1.0E-4 Hartree

Energy Convergence:1.0E-14 Hartree

Maximum number of iteration for SCF calculation:unknown

Periodic boundary:unknown

Molecule modeling soft:unknown

I understand that different conditions can lead to different results, but I wonder  that the results are so different.

[Frederick Stein]

Dear Soma,

I guess that you are comparing total energies. They are only comparable between codes if you use consider the same number of electrons with a similar description of you nucleai. In case of CP2K, you are probably using GTH pseudopotentials, i.e. you do not take the core electrons explicitly into account whereas you are probably running all-electron calculations with materials studio. This explains the different orders of magnitudes between your energies. In general, you will compare relative energies such as atomization energies, cohesive energies etc.

Best,

Frederick