12.07.2015 Views

Ab initio molecular dynamics: Theory and Implementation

Ab initio molecular dynamics: Theory and Implementation

Ab initio molecular dynamics: Theory and Implementation

SHOW MORE
SHOW LESS
  • No tags were found...

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

where ɛ j <strong>and</strong> ɛ i are the eigenvalues of occupied <strong>and</strong> unoccupied orbitals, respectively;see Eq. (26) in Ref. 467 for the case where both orbitals are occupied ones.It can be seen from Fig. 2 that the harmonic approximation works faithfully ascompared to the exact spectrum; see Ref. 471 <strong>and</strong> Sect. IV.A in Ref. 472 for a moregeneral analysis of the associated equations of motion. Since this is in particulartrue for the lowest frequency ωemin , the h<strong>and</strong>y analytic estimate for the lowestpossible electronic frequencyω mine∝(Egapµ) 1/2, (53)shows that this frequency increases like the square root of the electronic energydifference E gap between the lowest unoccupied <strong>and</strong> the highest occupied orbital.On the other h<strong>and</strong> it increases similarly for a decreasing fictitious mass parameterµ.ω maxnIn order to guarantee the adiabatic separation, the frequency difference ωemin −should be large, see Sect. 3.3 in Ref. 513 for a similar argument. But boththe highest phonon frequency ωnmax <strong>and</strong> the energy gap E gap are quantities that adictated by the physics of the system. Whence, the only parameter in our h<strong>and</strong>sto control adiabatic separation is the fictitious mass, which is therefore also called“adiabaticity parameter”. However, decreasing µ not only shifts the electronicspectrum upwards on the frequency scale, but also stretches the entire frequencyspectrum according to Eq. (52). This leads to an increase of the maximum frequencyaccording toω maxe∝(Ecutµ) 1/2, (54)where E cut is the largest kinetic energy in an expansion of the wavefunction interms of a plane wave basis set, see Sect. 3.1.3.At this place a limitation to decrease µ arbitrarily kicks in due to the maximumlength of the <strong>molecular</strong> <strong>dynamics</strong> time step ∆t max that can be used. The time stepis inversely proportional to the highest frequency in the system, which is ωemax <strong>and</strong>thus the relation( ) 1/2 µ∆t max ∝(55)E cutgoverns the largest time step that is possible. As a consequence, Car–Parrinellosimulators have to find their way between Scylla <strong>and</strong> Charybdis <strong>and</strong> have to makea compromise on the control parameter µ; typical values for large–gap systems areµ = 500–1500 a.u. together with a time step of about 5–10 a.u. (0.12–0.24 fs).Recently, an algorithm was devised that optimizes µ during a particular simulationgiven a fixed accuracy criterion 87 . Note that a poor man’s way to keep the timestep large <strong>and</strong> still increase µ in order to satisfy adiabaticity is to choose heaviernuclear masses. That depresses the largest phonon or vibrational frequency ωnmaxof the nuclei (at the cost of renormalizing all dynamical quantities in the sense ofclassical isotope effects).20

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!