where is the dispersion relation. If we expand around the center of wavepacket
we find
where
is the phase velocity, and is the group velocity.
So
This is the so-called Ehrenfest theorem.
That is to say, if we talk about a wavepacket in terms of its center of mass and center of momentum, we may recover the classical dynamics. This is the key idea of semiclassical dynamics.
We want to apply this trick to Bloch electrons, so we need wavepacket that is localized in both real and k-space
Normalization
center of mass:
center of momentum:
What is ? need position operator in the Bloch representation (布洛赫表象)
In the last step, we have performed an integration by parts (分部积分). We now introduce a quantity call the Berry connection
So the center of the wavepacket is
that is simply stating that in the Bloch representation
The Bloch wavefunctions have gauge freedom. That is,
is still the same Bloch state. Under this change of basis, we find
which ensures that the center of a wavepacket is invariant.
2. Semiclassical equations of motion
In an external electromagnetic potential for uniform magnetic and electric fields
is the orbital moment from the self-rotation of a wavepacket.
Hereafter, we will drop subscript . Consider
For a uniform magnetic field,
So this leads to the EOM
The novel term in the Lagrangian is , with which we have
where the pseudovector is defined via
is called the Berry curvature of Bloch electrons, whose contribution to the velocity is called the anomalous velocity (反常速度).
3. Anomalous Hall effect
Under an electric field, the anomalous velocity can cause a transverse current, even without an external magnetic field. This is called the anomalous Hall effect (反常霍尔效应).
Historically, there had been an debate over the origin of the anomalous Hall effect (AHE). If the anomalous Hall current comes from the Berry curvature of the Bloch bands, as described above, it is called the intrinsic AHE since it depends on the band structure and not on scattering. Correspondingly, the AHE conducitivtiy is
Take a two-dimensional system as an example,
Early experimental data2 suggest . However, there had been a long debate, as at finite temperatures inelastic scattering by bosons (phonon or magnon) makes a complete microscopic theory difficult. Other origins of the AHE have been suggested. These theories are based certain asymmetry in impurity scattering, such as side jump and skew scattering, which are called extrinic mechanisms. We will not go into details of these extrinsic mechanism, interested readers can consult Nagaosa's review.3
How do we measure the AHE conductivity? People have noted that the Hall effect in ferromagnetic and nonmagnetic metals are very different. In nonmagnetic metals, the Hall current comes the Lorentz force. In a ferromagnet, is linear in applied magnetic field when is small, but then saturates (饱和) at high fields. Empirically, peopel have found
References and notes
1Ming-Che Chang and Qian Niu. Phys. Rev. B53, 7010 (1996).↩