Time-evolution algorithm: Difference between revisions
No edit summary |
|||
Line 1: | Line 1: | ||
The [[Dielectric_properties|macroscopic dielectric function]], <math>\epsilon_{ij}(\omega)</math>, measures how a given dielectric medium reacts when subject to an external electric field. | The [[Dielectric_properties|macroscopic dielectric function]], <math>\epsilon_{ij}(\omega)</math>, measures how a given dielectric medium reacts when subject to an external electric field. From <math>\epsilon_{ij}(\omega)</math> one can extract several optical properties such as absorption, optical conductivity, reflectance. However, it is important that the interacting electrons and holes are taken into account. This makes the evaluation of the macroscopic dielectric function more involved, since it goes beyond the single-particle level, working either at the [[Bethe-Salpeter equations|Bethe-Salpeter]] or [[Time-dependent density-functional theory calculations|time-dependent density-functional theory]] level. | ||
== | Within VASP, users can select two different methods for how <math>\epsilon_{ij}(\omega)</math> is computed. The first is based on the eigendecomposition of the electron-hole Hamiltonian, <math>H^\mathrm{exc}</math>. It allows for the evaluation of <math>\epsilon_{ij}(\omega)</math> by first obtaining the eigenvalues and eigenvectors of <math>H^\mathrm{exc}</math> and it is based on the [[Bethe-Salpeter equations|Bethe-Salpeter equation]] or the [[Time-dependent density-functional theory calculations|Casida equation]]. The second method transforms the mathematical expression of <math>\epsilon_{ij}(\omega)</math> into a time-dependent integral. By propagating in time the dipolar moments and then applying a Fourier transform, it can compute <math>\epsilon_{ij}(\omega)</math>. | ||
The starting point is that one can re-write <math>\epsilon_{ij}(\omega)</math> as a time-dependent integral{{cite|schmidt:prb:2003}} | |||
The advantage of the later method in comparison to the former is related to the cost, with the time-dependent integral being <math>O(N^2)</math>, while the eigendecomposition has a cost of <math>O(N^3)</math>, where <math>N</math> is the rank of <math>H^\mathrm{exc}</math>. This means that for very large numbers of bands or k-points, the time-dependent formalism is cheaper than the eigendecomposition method. | |||
Below we present a brief description of the method, from its theoretical support to how calculations should be performed, with the relevant approximations needed in the two-particle Hamiltonian. | |||
==The macroscopic-dielectric function as a time-dependent integral== | |||
The starting point is that one can re-write <math>\epsilon_{ij}(\omega)</math> as a time-dependent integral{{cite|schmidt:prb:2003}}. It starts from its expression, given by | |||
::<math> | ::<math> | ||
\ | \epsilon^M(\omega)=1+\frac{4 \pi}{\Omega_0} \sum_\lambda\left|\sum_{c v \mathbf{k}} \mu_{c v \mathbf{k}} A_{c v \mathbf{k}}^\lambda\right|^2\left[\frac{1}{\omega+E_\lambda+\mathrm{i} \eta}-\frac{1}{\omega-E_\lambda+\mathrm{i} \eta}\right] | ||
\sum_{c | |||
</math>, | </math>, | ||
where <math>\mu_{v c \mathbf{k}}^j=\frac{\left\langle c \mathbf{k}\left|v_j\right| v \mathbf{k}\right\rangle}{\varepsilon_c(\mathbf{k})-\varepsilon_v(\mathbf{k})}</math> is | where <math>\mu_{v c \mathbf{k}}^j=\frac{\left\langle c \mathbf{k}\left|v_j\right| v \mathbf{k}\right\rangle}{\varepsilon_c(\mathbf{k})-\varepsilon_v(\mathbf{k})}</math> is the dipolar moment associated to the the conduction <math>c</math> and valence band <math>v</math>, and k-point <math>k</math>, <math>\lambda</math> is the index of the eigenstate of <math>H^\mathrm{exc}</math>, with <math>A^\lambda</math> and <math>E_\lambda</math> being the associated eigenvector and eigenvalue. This equation can be brought into operational form, | ||
::<math> | ::<math> | ||
\ | \epsilon^M(\omega)=1+\frac{4 \pi}{\Omega_0}\left\langle\mu\left|\left[\frac{1}{\omega+\mathrm{i} \eta+\hat{H}^{\mathrm{exc}}}-\frac{1}{\omega+\mathrm{i} \eta-\hat{H}^{\mathrm{exc}}}\right]\right| \mu\right\rangle | ||
</math> | </math> | ||
by using the spectral decomposition <math>\left[\hat{H}^{\mathrm{exc}}-\omega\right]^{-1}=\sum_\lambda \frac{\left|A_\lambda\right\rangle\left\langle A_\lambda\right|}{E_\lambda-\omega}</math>. Then, one can bring the new expression of <math>\epsilon(\omega)</math> into a time-dependent integral, by using | |||
::<math> | |||
\frac{1}{\omega+\mathrm{i} \eta-\hat{H}^{\mathrm{exc}}}|\mu\rangle=-\mathrm{i} \int_0^{\infty} e^{-\mathrm{i}\left(\omega-\hat{H}^{\mathrm{exc}}+\mathrm{i} \eta\right) t}|\mu\rangle=-\mathrm{i} \int_0^{\infty} e^{-\mathrm{i}(\omega+\mathrm{i} \eta) t} e^{\mathrm{i} \hat{H}^{\mathrm{exc}}t}|\mu\rangle | |||
</math>, | |||
and recognising that <math>e^{\mathrm{i} \hat{H}^{\mathrm{exc}}t}|\mu\rangle = |\xi(t)\rangle</math> is the exponential form of a time-dependent equation. These considerations allow the expression of <math>\epsilon_{ij}(\omega)</math> to be written as | |||
::<math> | |||
\epsilon_{ij}(\omega)=\delta_{ij}-\frac{4\pi e^2}{\Omega}\int_0^{\infty} \mathrm{d} t | |||
\sum_{c,v,\mathbf{k}}\left(\langle\mu^j_{cv\mathbf{k}}| \xi^i_{cv\mathbf{k}}(t)\rangle+ \mathrm{c.c.}\right) e^{-\mathrm i(\omega-\mathrm i \delta) t} | |||
</math>, | |||
The fundamental aspect behind this transformation is that the new, time-dependent vector <math>\left.\mid \xi^j(t)\right\rangle</math> follows the equation | |||
::<math> | ::<math> | ||
\mathrm i \frac{\mathrm d}{\mathrm d t}\left|\xi^j(t)\right\rangle=\hat{H}(t)\left|\xi^j(t)\right\rangle, | |||
</math> | </math> | ||
with the initial vector elements given by <math>\left|\xi^j(0)\right\rangle=\left|\mu^j\right\rangle</math>. It is worthy to note that <math>H(t)</math> is split into two parts, <math>H(t) = H_0 + V(t)</math>, where <math>H_0</math> is the ground-state Hamiltonian, and <math>V(t)</math> is the time-dependent perturbation. | |||
==The delta-like perturbation== | |||
explain how the system is perturbed with the delta-potential | |||
==The many-body terms in the hamiltonian== | |||
===Independent-particle approximation=== | |||
===Hartree exchange potential=== | |||
===Screened two-particle interaction=== | |||
====Exchange-correlation effects from time-dependent density functional theory==== | |||
====Ladder diagrams from many-body perturbation theory==== | |||
explain what different components are included in H (LFXC,LHARTREE,LADDER, or none) | |||
==Step-by-step instructions== | |||
===Step 1: Ground state with extra empty states=== | |||
===Step 2: Time-evolution run=== | |||
====Setting up the time-step==== | |||
====Choosing the direction of perturbation==== | |||
== | ==Comparison to other methods== | ||
===Bethe-Salpeter equation=== | |||
===Casida equation=== | |||
== | ==Related tags and articles== | ||
==References== | |||
<!---[[Category:Linear response]] | <!---[[Category:Linear response]] |
Revision as of 16:12, 24 June 2024
The macroscopic dielectric function, , measures how a given dielectric medium reacts when subject to an external electric field. From one can extract several optical properties such as absorption, optical conductivity, reflectance. However, it is important that the interacting electrons and holes are taken into account. This makes the evaluation of the macroscopic dielectric function more involved, since it goes beyond the single-particle level, working either at the Bethe-Salpeter or time-dependent density-functional theory level.
Within VASP, users can select two different methods for how is computed. The first is based on the eigendecomposition of the electron-hole Hamiltonian, . It allows for the evaluation of by first obtaining the eigenvalues and eigenvectors of and it is based on the Bethe-Salpeter equation or the Casida equation. The second method transforms the mathematical expression of into a time-dependent integral. By propagating in time the dipolar moments and then applying a Fourier transform, it can compute .
The advantage of the later method in comparison to the former is related to the cost, with the time-dependent integral being , while the eigendecomposition has a cost of , where is the rank of . This means that for very large numbers of bands or k-points, the time-dependent formalism is cheaper than the eigendecomposition method.
Below we present a brief description of the method, from its theoretical support to how calculations should be performed, with the relevant approximations needed in the two-particle Hamiltonian.
The macroscopic-dielectric function as a time-dependent integral
The starting point is that one can re-write as a time-dependent integral[1]. It starts from its expression, given by
- ,
where is the dipolar moment associated to the the conduction and valence band , and k-point , is the index of the eigenstate of , with and being the associated eigenvector and eigenvalue. This equation can be brought into operational form,
by using the spectral decomposition . Then, one can bring the new expression of into a time-dependent integral, by using
- ,
and recognising that is the exponential form of a time-dependent equation. These considerations allow the expression of to be written as
- ,
The fundamental aspect behind this transformation is that the new, time-dependent vector follows the equation
with the initial vector elements given by . It is worthy to note that is split into two parts, , where is the ground-state Hamiltonian, and is the time-dependent perturbation.
The delta-like perturbation
explain how the system is perturbed with the delta-potential
The many-body terms in the hamiltonian
Independent-particle approximation
Hartree exchange potential
Screened two-particle interaction
Exchange-correlation effects from time-dependent density functional theory
Ladder diagrams from many-body perturbation theory
explain what different components are included in H (LFXC,LHARTREE,LADDER, or none)