nextnano3 - Tutorial
next generation 3D nano device simulator
1D Tutorial
Valence band masses from k.p
Authors:
Stefan Birner, Michael
Povolotskyi
If you want to obtain the input files that are used within this tutorial, please
check if you can find them in the installation directory.
If you cannot find them, please submit a
Support Ticket.
-> 1DGaAs_vb_masses_from_kp.in
Valence band masses from k.p
-> 1DGaAs_vb_masses_from_kp.in
The energy dispersion of the valence band masses in a cubic crystal
has cubic symmetry.
Thus it is usually not a good approximation to take a spherical
effective mass tensor for the single-band Schrödinger equation for the
holes.
For that reason, usually a 6-band or 8-band k.p Hamiltonian should be
used.
However, there might be situations where the user wants to use the
single-band Schrödinger equation (e.g. to minimize CPU time) but wants
to use a local (i.e. grid point dependent) effective mass tensor that
takes into account the local strain tensor of each grid point.
This means that for each grid point, the bulk E(k) band structure
(which takes into account the local strain tensor) is used to calculate
the (strain dependent) components of the symmetric (3x3) effective mass
tensors of the heavy hole, light hole and split-off hole.
The relation of effective mass tensor to energy dispersion is as
follows:
(1/m)ij = 1/hbar2 d2
E / (dki dkj)
$numeric-control
!
...
valence-band-masses-from-kp = yes
! makes only sense for use with single-band Schrödinger equation
$output-1-band-schroedinger
...
effective-mass-tensor =
yes ! to output the
effective mass tensor compontents (1/m)ij
==> Schroedinger_1band/mass_tensor1D_vb001_qc001_sg001_deg001.dat
- heavy hole
==> Schroedinger_1band/mass_tensor1D_vb002_qc001_sg001_deg001.dat -
light hole
==> Schroedinger_1band/mass_tensor1D_vb003_qc001_sg001_deg001.dat -
split-off hole
Approximation scheme
From the energy dispersion function, E(kx,ky,kz),
the components of the symmetric (3x3) effective mass tensor are
calculated as follows:
mxx = hbar2 * dk2
/ (2*[-E(dk,0,0)
+ E(0,0,0)]) where 'E(0,0,0) ' is
the Gamma point.
mxy = hbar2 * dk2
/ [ E(dk,0,0) + E(0,dk,0) - E(0,0,0) - E(dk,dk,0)]
...
dk is chosen to be equal to: dk = 1d-4
GaAs
!Luttinger-parameters =
6.98d0 2.06d0 2.93d0 ! gamma1,
gamma2, gamma3
1.72d0 0.04d0
! kappa, q
6x6kp-parameters = -16.22d0
-3.86d0 -17.58d0 ! L,M,N [hbar^2/2m] (--> divide by
hbar^2/2m)
0.341d0
! deltasplit-off in [eV]
Heavy hole: components of the effective mass tensor
(1/m)xx = 2.86000 ==> mxx
= 0.34965 [m0]
(1/m)yy = 2.86000 ==> myy
= 0.34965 [m0]
(1/m)zz = 2.86000 ==> mzz
= 0.34965 [m0]
(1/m)xy = -1.35707 ==> mxy =
-0.73688 [m0]
(1/m)xz = -1.35707 ==> mxz =
-0.73688 [m0]
(1/m)yz = -1.35707 ==> myz =
-0.73688 [m0]
Light hole: components of the effective mass tensor
(1/m)xx = 11.1000 ==> mxx
= 0.09009 [m0]
(1/m)yy = 11.1000 ==> myy
= 0.09009 [m0]
(1/m)zz = 11.1000 ==> mzz
= 0.09009 [m0]
(1/m)xy = 1.35704 ==> mxy
= 0.73690 [m0]
(1/m)xz = 1.35704 ==> mxz
= 0.73690 [m0]
(1/m)yz = 1.35704 ==> myz
= 0.73690 [m0]
Split-off hole: components of the effective mass tensor
(1/m)xx = 6.98001 ==> mxx
= 0.14327 [m0]
(1/m)yy = 6.98001 ==> myy
= 0.14327 [m0]
(1/m)zz = 6.98001 ==> mzz
= 0.14327 [m0]
(1/m)xy = 0.000034 ==> mxy =
29128.7 [m0]
(1/m)xz = 0.000034 ==> mxz =
29128.7 [m0]
(1/m)yz = 0.000034 ==> myz =
29128.7 [m0]
These results can be compared with analytical formulas (e.g. I.
Vurgaftman et al., J. Appl. Phys. 89, 5815 (2001)):
- mhh[100] = 1 / ( gamma1
- 2
gamma2) = 1 / 2.86 = 0.34965
[m0]
- mlh [100] = 1 / ( gamma1 + 2 gamma2)
= 1 / 11.1 = 0.09009 [m0]
- mhh[110] = 2 / (2 gamma1
-
gamma2 - 3 gamma3) = 1 / 1.555
= 0.64309 [m0]
- mlh [110] = 2 / (2 gamma1
+
gamma2 + 3 gamma3) = 1 / 12.405 =
0.08061 [m0]
- mhh[111] = 1 / (gamma1
- 2
gamma3) = 1 / 1.12 = 0.89286 [m0]
- mlh [111] = 1 / (gamma1
+ 2
gamma3) = 1 / 12.84 = 0.07788 [m0]
Principal axes system
!+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++!
! ==> Leading to a diagonal effective mass tensor: m_xx = 0.145, m_yy = m_zz =
4.21
!+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++!
hkl-x-direction-zb = 1 1 1 ! [111]
hkl-y-direction-zb = 0 1 -1 ! [01-1]
!hkl-z-direction-zb = . . . ! [...] (calculated internally)
|