previous up next contents Previous: 2.3.1 Even Moments: Relaxation Up: 2.3 Closure Relations for Next: 2.3.3 Analytical Mobility and

2.3.2 Odd Moments: Mobilities

As the odd equilibrium values vanish, the macroscopic relaxation time approximation simplifies. To approximate the odd scattering terms we introduce the mobilities [GJK+04] in the odd equations. By setting

$\displaystyle \tau \hat{=} \frac{m^* \mu}{q}$ (2.56)

we get a formal correspondence between relaxation times and mobilities for parabolic bands. If $ \mu$ and $ q$ have the same sign, then $ \tau$ is positive.

With this correspondence we get:

$\displaystyle \langle O_{2i+1}\rangle = - \frac{\mu_{2i+1} m^{*}}{q} \bigg( \fr...
...3} \bigg( \bigg(1 + \frac{2i}{3}\bigg)M_{2i}\bigg) \bigg) \quad     (i=0,1,2)$ (2.57)

One possibility to model $ \mu$ is as a function depending on the the local temperature $ T_n(x)$. Hence the equations depend in a nonlinear way on $ M_2$ and $ M_0$. Information about the scattering term can be encoded into a fitting ansatz for the mobilities, for example in the form $ \mu_{\phi} (T) =
a_{\phi} T^{b_{\phi}}$ with fitting parameters $ a_{\phi},b_{\phi}$.

Alternatively, we can approximate the mobilities as functions of temperature and doping using results from bulk Monte Carlo simulations, see Section 2.3.4.

previous up next contents Previous: 2.3.1 Even Moments: Relaxation Up: 2.3 Closure Relations for Next: 2.3.3 Analytical Mobility and


R. Kosik: Numerical Challenges on the Road to NanoTCAD