next up previous contents
Next: 3.5 Velocity Saturation Up: 3.4 Carrier Mobility Previous: 3.4.4 High-Field Mobility for

Subsections


3.4.5 High-Field Mobility for HD Equations


The deviation from the ohmic low-field mobility is modeled as a function of the carrier temperature, $T_\nu$, after Hänsch [180].
$\displaystyle \mu^{\mathrm{LIST}}_\nu$ $\textstyle =$ $\displaystyle \frac{\mu^{\mathrm{LIS}}_{\nu}}{1 + \alpha_{\nu}\cdot\left(T_{\nu}-
T_{{\mathrm{L}}}\right)}$ (3.129)
$\displaystyle \alpha_{\nu}$ $\textstyle =$ $\displaystyle \frac{3\cdot \mathrm{k_B}\cdot \mu^{\mathrm{LIS}}_{\nu}}{2\cdot \mathrm{q}\cdot
\tau_{\epsilon} \cdot \left(v^{\mathrm{sat}}_{\nu}\right)^2}$ (3.130)

$\tau_{\epsilon}$ denotes the energy relaxation times and $v^{\mathrm{sat}}_{\nu}$ are the saturation velocities calculated respectively in (3.138) and in (3.134).

For some III-V semiconductor materials a multi-valley mobility model can be used. In the following model a weighted mean is calculated from the low-field mobilities of the $\Gamma $- and $\mathrm {L}$-electrons [181].

$\displaystyle \mu^{\mathrm{LIT}}_n$ $\textstyle =$ $\displaystyle \frac{\mu^{\mathrm{LI}}_{n,\Gamma}+\mu^{\mathrm{LI}}_{n,\mathrm {L}}\cdot P_{\mathrm{L}}}{1+P_{\mathrm{L}}}$ (3.131)
$\displaystyle P_{\mathrm{L}}$ $\textstyle =$ $\displaystyle 4\cdot\left(\frac{m_{\mathrm{L}}}{m_\Gamma}\right)^{3/2}
\cdot\exp\left(-\frac{E_{C,{\mathrm{L}}}-E_{C,\Gamma}}{\mathrm{k_B}\cdot T_n}\right)$ (3.132)

$P_{\mathrm {L}}$ denotes the ratio of the $\Gamma $ and $\mathrm {L}$-valley populations [86]. The valley mobilities ${\mu^{\mathrm{LI}}_{n,\Gamma}}$ and ${\mu^{\mathrm{LI}}_{n,\mathrm {L}}}$ account for impurity scattering after (3.97), and are constant with respect to the lattice temperature.

Table 3.29: Parameter values for the two-valley HD mobility model
Parameter GaAs Unit
$\mu^{\mathrm{L}}_{n,\Gamma}$ 8000 cm$^2$/Vs
$\mu^{\mathrm{L}}_{n,\mathrm {L}}$ 400 cm$^2$/Vs
$\mu^{\mathrm{min}}_{n}$ 1500 cm$^2$/Vs
$C^{\mathrm{ref}}_{n}$ 1.426e+17 cm$^{-3}$
$\alpha_{n}$ 0.5385  
$\tau_{n,\Gamma}$ 1.2 ps
$\tau_{n,\mathrm {L}}$ 0.6 ps
$v^{\mathrm{sat}}_{n,\Gamma}$ 2.5e5 m/s
$v^{\mathrm{sat}}_{n,\mathrm {L}}$ 0.9e5 m/s


3.4.5.1 Semiconductor Alloys

In the case of alloy materialsthe model employs the low-field mobilities $\mu^{\mathrm{LI}}$ of the basic materials (A and B) and combines them by a harmonic mean.
\begin{displaymath}
\frac{1}{\mu^{\mathrm{AB}}} =\frac{1-x}{\mu^{\mathrm{A}}}+\frac{x}{\mu^{\mathrm{B}}}+
\frac{\left(1-x\right)\cdot x}{C_\mu}
\end{displaymath} (3.133)

$C_\mu$ is referred to as nonlinear or bowing parameter.

Table 3.30: Parameter values for mobility model for alloy materials
Material $C_{\mu,n}$ [cm$^2$/Vs] $C_{\mu,p}$ [cm$^2$/Vs]
SiGe 1e6 -100
AlGaAs 180 1e6
InGaAs 1e6 1e6
InAlAs 1e6 1e6
InAsP 1e6 1e6
GaAsP 1e6 1e6
InGaP 1e6 1e6


For calculation of the high field mobilities no additional parameters need to be specified for the model. The respective interpolations between the basic materials are carried out in the models for the saturation velocities and, in the case of HD simulation, of the energy relaxation times.


next up previous contents
Next: 3.5 Velocity Saturation Up: 3.4 Carrier Mobility Previous: 3.4.4 High-Field Mobility for
Vassil Palankovski
2001-02-28