SiH4 - Revision history

Davide at 22:06, 19 November 2023

2023-11-19T22:06:37Z

← Older revision		Revision as of 22:06, 19 November 2023
Line 1:		Line 1:
	'''WARNIGN: This page is presently under construction'''		'''WARNIGN: This page is presently under construction'''

			==The material==
	[[File:SiH4.jpg\|300px\|frameless\|right\|alt=SiH4 molecule\|SiH4 molecule]]		[[File:SiH4.jpg\|300px\|frameless\|right\|alt=SiH4 molecule\|SiH4 molecule]]
	~~==The material==~~
	SiH4: common name Silane		SiH4: common name Silane
	Supercell 35.0 [a.u.]		Supercell 35.0 [a.u.]

Davide: /* BSE with coupling terms */

2023-11-14T16:22:00Z

BSE with coupling terms

← Older revision		Revision as of 16:22, 14 November 2023
Line 201:		Line 201:
	you will immediately notice that the size of the BS Kernel increased from 136 to 272 due to the coupling terms.		you will immediately notice that the size of the BS Kernel increased from 136 to 272 due to the coupling terms.
	Moreover the full BS matrix is now not Hermitian, but also in this case it is possible to use the Haydock technique thanks to a recent generalization to pseudo-Hermitian matrices[6]. Now we compare the result with the previous one:		Moreover the full BS matrix is now not Hermitian, but also in this case it is possible to use the Haydock technique thanks to a recent generalization to pseudo-Hermitian matrices[6]. Now we compare the result with the previous one:
			[[File:SiH4 bse cpl.png\|700px\|frameless\|center\|alt=SiH4 bse cpl dielectric function]]

	As you can see including coupling terms improve the agreement with the first ant third peak, but the second one remains unchanged. Why BSE is not able to reproduce the experimental result?		As you can see including coupling terms improve the agreement with the first ant third peak, but the second one remains unchanged. Why BSE is not able to reproduce the experimental result?

Davide: /* Bethe-Salpeter Equation (BSE) */ plot inserted

2023-11-14T16:20:32Z

Bethe-Salpeter Equation (BSE): plot inserted

← Older revision		Revision as of 16:20, 14 November 2023
Line 186:		Line 186:

	Finally we can compare the resulting BS optical absorption with the experimental results and the independent particle(IP) approximation. As you can see the excitonic effects strongly modify the spectra respect to the IP, but the peaks are still not enough close to the experimental ones.		Finally we can compare the resulting BS optical absorption with the experimental results and the independent particle(IP) approximation. As you can see the excitonic effects strongly modify the spectra respect to the IP, but the peaks are still not enough close to the experimental ones.
			[[File:SiH4 bse.png\|700px\|frameless\|center\|alt=SiH4 bse dielectric function]]

	Now we will try to improve the outcome of our calculations including coupling terms in the BSE.		Now we will try to improve the outcome of our calculations including coupling terms in the BSE.

	==BSE with coupling terms==		==BSE with coupling terms==
	06_bse (yambo -o b -k sex -y h)		06_bse (yambo -o b -k sex -y h)

Davide: Added few plots

2023-11-14T16:18:30Z

Added few plots

← Older revision		Revision as of 16:18, 14 November 2023
Line 67:		Line 67:
	<---> [04] Game Over & Game summary		<---> [04] Game Over & Game summary

	If now you edit the r-01_setup you will find a lot of useless () information about your system like a short summary of the electronic structure previously calculated at DFT level, and some details about atomic structure and the cell, for instance the HOMO-LUMO gap		If now you edit the r-01_setup you will find a lot of useless information about your system like a short summary of the electronic structure previously calculated at DFT level, and some details about atomic structure and the cell, for instance the HOMO-LUMO gap

	States summary : Full Metallic Empty		States summary : Full Metallic Empty
Line 93:		Line 93:

	Then you can edit the input file changing the values of NGsBlkXd and FxcGRLc variables, for instance to 400 or more RL. Now we can compare the result of our calculations. The experimental result [1] denotes a rather broad spectrum where we can recognize three peaks located at: 8.8 eV, 9.7eV, 10.7eV. If we plot the obtained outputs:		Then you can edit the input file changing the values of NGsBlkXd and FxcGRLc variables, for instance to 400 or more RL. Now we can compare the result of our calculations. The experimental result [1] denotes a rather broad spectrum where we can recognize three peaks located at: 8.8 eV, 9.7eV, 10.7eV. If we plot the obtained outputs:
			[[File:SiH4 alda gspace.png\|700px\|frameless\|center\|alt=SiH4 alda gspace dielectric function\|SiH4 alda gspace]]

	The comparison with the experiment clearly shows that the agreement with the experiment is really poor. Most importantly, it shows that the absorption intensity is really small (if you enlarge the energy range you would observe that the spectrum here is hundreds of time too small). Such instability is due to a numerical instability of the TDDFT Dyson equation, with ALDA kernel, when applied to low-dimensional electronic systems, and evaluated in reciprocal space. The instability is due to the presence of large regions with vanishing density, as in large super-cells. In a region of the pace with vanishing density we have that fxcHEG (n) diverges so that the evaluation of the term involving the fxc in the Dyson equation cannot be directly calculated in reciprocal space, because the f ALDA is not well defined.		The comparison with the experiment clearly shows that the agreement with the experiment is really poor. Most importantly, it shows that the absorption intensity is really small (if you enlarge the energy range you would observe that the spectrum here is hundreds of time too small). Such instability is due to a numerical instability of the TDDFT Dyson equation, with ALDA kernel, when applied to low-dimensional electronic systems, and evaluated in reciprocal space. The instability is due to the presence of large regions with vanishing density, as in large super-cells. In a region of the pace with vanishing density we have that fxcHEG (n) diverges so that the evaluation of the term involving the fxc in the Dyson equation cannot be directly calculated in reciprocal space, because the f ALDA is not well defined.
Line 132:		Line 133:

	Plotting the output we get:		Plotting the output we get:
			[[File:Sih4 alda rspace.png\|700px\|frameless\|center\|alt=SiH4 alda rspace dielectric function]]

	First, we note that, the inclusion of the coupling part in the matrix has a rather strong effects. When the Tamm-Dancoff approximation is removed we get the correct number of peaks in the energy range of interest, but unfortunately, even if now we get a stable result, it looks that ALDA does not reproduce the position of the experimental peaks. Note that such a failure is not general for finite system, as the ALDA is commonly used for computation of optical spectra of molecules and clusters [2]. The SiH4 molecule turns out to be a very challenging molecule as described in [2], where several functionals are tested displaying quite dissimilar results.		First, we note that, the inclusion of the coupling part in the matrix has a rather strong effects. When the Tamm-Dancoff approximation is removed we get the correct number of peaks in the energy range of interest, but unfortunately, even if now we get a stable result, it looks that ALDA does not reproduce the position of the experimental peaks. Note that such a failure is not general for finite system, as the ALDA is commonly used for computation of optical spectra of molecules and clusters [2]. The SiH4 molecule turns out to be a very challenging molecule as described in [2], where several functionals are tested displaying quite dissimilar results.

Davide: Improving a bit the layout

2023-11-14T16:08:51Z

Improving a bit the layout

Show changes

Davide: First version of the page created

2023-11-14T15:56:53Z

First version of the page created

Show changes

Davide: Creating empty page

2023-11-14T15:47:47Z

Creating empty page

New page