... subsystem2.1
In particular, this subsystem may be represented by one particle.



























































... domain2.2
Here $ \vec{r}$ and $ \vec{p}=\hbar\vec{k}$ denote all coordinates and quasi-momenta of the subsystem. For an electron they are just $ (x,y,z)$ and $ (p_{x},p_{y},p_{z})$ (in Cartesian coordinates).



























































... function2.3
Note that due to the fact that $ \frac{\partial(\vec {r},\vec {k})}{\partial(\vec {r},\vec {k}_{c})}=1$ we can consider the phase space distribution as a function of the variables $ (\vec {r},\vec {k}_{c})$ in the phase space of the variables which are canonically conjugated through the Hamiltonian (2.17). This will be assumed in the following.



























































... other2.4
For particles this means the absence of scattering between them.



























































... space2.5
Here $ d\vec {r}_{1}d\vec {k}_{1}$ and $ d\vec {r}_{2}d\vec {k}_{2}$ are elements of phase spaces of the two parts.



























































... system2.6
In solids $ \mu$ is the Fermi energy denoted as $ \epsilon_{f}$.



























































... function2.7
The Boltzmann distribution is valid when all $ \langle n_{k} \rangle\ll 1$. Physically it corresponds to a dilute system.



























































... as2.8
The factor $ 1/(2\pi)^{3}$ comes from the fact that in the quasi-classical case the phase space volume $ (2\pi\hbar)^{3}$ corresponds to two quantum states of a particle. These states differ by spin orientation.



























































... as2.9
The spin of a phonon is taken to be zero. Thus the pre-factor $ g=2s+1=1$. Since the phonon number is not fixed and determined by the equilibrium condition, the chemical potential of the phonon gas is equal to zero: $ \mu=0$.



























































... processes2.10
This of course overestimates the efficiency of scattering needed to restore the equilibrium state.



























































... quasi-momentum2.11
In the semiclassical approach collisions are considered as instantaneous events taking place at a given point in real space. Thus the semiclassical transport model only considers changes of the non-equilibrium distribution function which happen during time intervals longer in comparison with the collision duration and at distances longer than the collision domain.



























































... space2.12
The phase space is considered in the sense of canonical conjugate variables.



























































... function2.13
The fact that the equilibrium distribution is considered is shown by the absence of the time among the arguments.



























































... equation2.14
In steady state there is no any explicit time dependence. Thus $ \frac{\partial f}{\partial t}=0$.



























































... motion2.15
This can be easily seen using the Hamiltonian form of the semiclassical equations of motion.



























































... subsystem2.16
The other six independent integrals of motion are the components of momentum of the subsystem and components of its angular momentum.



























































... operator2.17
The alternative symbol often used for the collision integral is $ \mathbf{St}f$.



























































... follows2.18
For the sake of simplicity the scattering within one band and without spin flipping is considered here.



























































... zero2.19
This can easily be checked substituting (2.30) where $ \epsilon(\vec{k})$ is replaced with $ \epsilon(\vec {k})+U(\vec {r})$, $ U(\vec {r})$ is the potential energy.



























































... periodicity2.20
It should be noted that the periodicity is only a simplification. Real crystals never have an ideal periodicity due to impurities, thermal vibrations and so forth.



























































... branches2.21
It is assumed that the discrete levels are continuous functions of the parameter $ \vec{k}$.



























































... follows2.22
Summation is understood under repeated indeces



























































... rank2.23
It should be noted that $ m_{\alpha\beta}^{-1}>0$ for a minimum, and $ m_{\alpha\beta}^{-1}<0$ for a maximum.



























































... energy2.24
The surface in the first Brillouin zone.



























































... respectively2.25
For example for Si $ X$ valleys along $ z$ axes $ m_{x}=m_{y}=m_{t}$, $ m_{z}=m_{t}$.



























































... quasi-particles2.26
For example electron-plasmon scattering.



























































... general2.27
In the sense that the form of its eigenstate cannot be determined in contrary to the ideal crystal case where Bloch's theorem exists.



























































... equation2.28
$ \hat{H}$ denotes the full Hamiltonian: $ \hat{H}=\hat{H}_{0}+\hat{H}_\mathrm{int}$.



























































... side2.29
Matrix elements $ \langle s^{'}\vert H_\mathrm{int}\vert s^{ are considered as quantities of first order.



























































...WF_decomp)2.30
The case when $ s^{'}\ne s$ is considered. The probability $ \vert a_{s}(t)\vert^{2}$ is obtained from (2.89).



























































... spectrum2.31
Quasi-momentum components and phonon quasi-wave vectors are considered as discrete quantities.



























































... limit2.32
Since it is assumed that an interaction decays much faster in comparison with the free-flight time, the error caused by the long interaction time limit is negligible.



























































... amplitude2.33
It is also referred to as polarization.



























































... gas2.34
Similarly to phonons plasmons represent bosons.



























































... field2.35
The same is valid for two positively charged particles.



























































... charge2.36
In our case the charge density of two ions.



























































... other2.37
This is valid for a rather weak potential $ V_{t}$.



























































... theory2.38
Sometimes it is called the Random Phase Approximation (RPA).



























































... degeneracy3.1
The original degeneracy is related to the symmetry of the unstrained Si and Ge crystals.



























































... principle3.2
That is the statement about the radius of the atomic forces. This radius must be considered to be zero in the macroscopic theory.



























































... condition3.3
The integration variables must be $ x^{'}$. However, due to weak strain the derivative in (3.22) differs from the derivative with respect to $ x^{'}$ by higher order terms.



























































... two3.4
Moment of force is a vector. Its components can be expressed as $ M_{i}=\frac{1}{2}e_{ijk}M_{jk}$, where $ e$ is the unit antisymmetric tensor of rank three and $ M_{jk}$ is the antisymmetric tensor of rank two which is integrand in (3.24).



























































... field4.1
The quasi-momentum and electric field are assumed to be transformed to the Herring-Vogt space to spherical constant energy surfaces instead of ellipsoidal ones which are the case for Si and Ge.



























































... system4.2
This is just an integral form of the Boltzmann kinetic equation discussed later in this chapter.



























































... events4.3
In general a Monte Carlo trajectory is not identical with a real trajectory and depends on a specific Monte Carlo algorithm.



























































... rewritten4.4
This is the standard form from the theory of integral equations.



























































...\space 4.5
For inhomogeneous case it stands for $ (\vec {k},\vec {r},t)$ that is a point in the seven dimensional space.



























































... series4.6
Also known as the Neumann series [72].



























































... space4.7
It should be noted that in our case this is the relative number of carriers determined by the perturbation $ f_{1}(\vec{k},t)$.



























































... conservation4.8
Or from the incompressibility of the phase space liquid.



























































... particlesA.1
The symbol $ \vert...)$ stands for the tensor product of the single-particle states.