QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: qweasel on February 8, 2017, 17:34

Title: Why k-point density is given in angstroms?
Post by: qweasel on February 8, 2017, 17:34

The unit for "Density" of k-point sampling in density of states calculation is given in angstroms (default value=2).

• What does it mean? Is it N*a/2pi? Why not just unitless N?
• So larger value means higher k resolution (dk=2pi/Na)?

Title: Re: Why k-point density is given in angstroms?
Post by: Petr Khomyakov on February 8, 2017, 18:36

1. You have an alternative option of choosing the k-point grid with 'Sampling' instead of 'Density' to set the number of k-points (N_A, N_B, N_C) in the Brillouin zone along the A, B, C reciprocal lattice vectors. 

2. Higher k-point grid density means that more k-points is used to sample the Brillouin zone.

Title: Re: Why k-point density is given in angstroms?
Post by: Daniele Stradi on February 9, 2017, 13:53

The density is in Angstrom because it is a density in reciprocal space.

In real space, a density for 1-D system would have units of Angstrom**-1. However, in reciprocal space, this quantity becomes (Angstrom**-1)**-1 = Angstrom.