Hi Wei
Thanks for the paper. I didnt know that about siesta well done so! I
will read the paper :)
It is gpl ?
Concerning the accuracy of Siesta I think yes, you can get the same
accord of a planewave code. There are several people in the Siesta
community working on oxide with good results indeed. The only constrain
i think is the dimension of the basis you use....but since I never use
Siesta for oxides I really cannot comment on that.
Maybe you can have a look at some of the papers already done with siesta
on oxides.
btw, thousand of atoms? it is an entire device! tough job!
Keep on!
I think in the end i cannot add more to what has been already said.
I wish you all the best for the project!
ciao
federico.
On 04/04/2014 08:11 PM, Wei Hu wrote:
Hi Federico,
Thanks very much for your answer. I just started doing this work.
We want to focus on metallic lanthanide metal oxide without considering the
localized electrons (d or f) via GW and HSE06 methods.
The system may be very large with thousands of atoms, and can not deal with the
GW and HSE06 methods.
How about the atomic orbitals describe these systems compared with plane wave
basis set?
According to Philippe, tuning the parameters (Such as increasing basis set) in
SIESTA may obtain similar accurate results to VASP?
By the way, the HSE06 functional has already implemented in SIESTA by our
research team. See Ref: JCP 135, 034110 (2011).
Best,
Wei
-----Original E-mail-----
From: "Federico Iori" <[email protected]>
Sent Time: 2014-4-4 21:27:06
To: [email protected]
Cc:
Subject: Re: [SIESTA-L] SIESTA calculation of transition metal or lanthanide
metal oxide
Hi.
if you want to proper describe correctly the charge density and
wavefunction, that means describing localized electrons (d or f), of
transition metal oxides you have to go beyond the local density
approximation of DFT (this holds even for GGA).
Otherwise if u want to describe an insulating phase of such compound in
LDA or GGA you will get instead a metallic density of state.
You will not be able to open correctly a gap, to correctly reproduce the
occupation of such localized orbitals and their correct energy ordering
for example. (this holds for several paradigmatic TMO)
Also in the metallic phase you will get metal, correct, but with an
almost incorrect redistribution of spectral weight in the density of state.
To better describe the electronic structure of such compound you have
two chances:
1. to use a hybrids functional scheme with screened range separation
(liek HSE06) to have a first proper description of the insulating and
metallic electronic structure + G0W0 calculation on top of this to
better describe the screening of your system
2. to face the problem directly using many body perturbation theory. In
the specific the approach could be to use a SelfConstistent GW scheme or
CoHSEX + G0W0.
have a look at
Phys. Rev. B 85, 115129 – Published 28 March 2012
Phys. Rev. Lett. 99, 266402 – Published 28 December 2007
for example.
Coming ti your question: in the gpl version of Siesta still there is no
GW scheme implemented. If I am not wrong neither HSE06 are implemented.
VASP and pwscf have both.
VASP all in one. PWSCF have HSE06 and is optimally interfaced with YAMBO
a code to perform many body calculations.
Hope to have been clear.
Ciao
federico
On 04/03/2014 07:07 PM, Wei Hu wrote:
Hi,
How about the SIESTA calculation of transition metal or lanthanide metal oxide
compared with VASP or pwscf? Including computing time and accuracy.
I have found few works about this respect. Why?
Best,
Wei
--
Dr. Wei Hu
Postdoc
Computational Research Division
Lawrence Berkeley National Laboratory
Berkeley, CA 94720, USA
Phone: +1-510-541-6034
E-mail: whu@@lbl.gov
Ph.D
Department of Chemical Physics
Hefei National Laboratory for Physical Sciences at the Microscale
University of Science and Technology of China
Hefei, Anhui 230026, P.R.China
Phone: +86-551-63606428
Mobile: +86-159 0560 4040
E-mail: [email protected]
--
Dr. Federico Iori, PhD!
Dipartimento di Scienze e Metodi per l' Ingegneria (DISMI)
Universita' di Modena e Reggio Emilia
Via Amendola,2
Pad. Morselli
42122 Reggio Emilia (Italy)
--
Dr. Wei Hu
Postdoc
Computational Research Division
Lawrence Berkeley National Laboratory
Berkeley, CA 94720, USA
Phone: +1-510-541-6034
E-mail: whu@@lbl.gov
Ph.D
Department of Chemical Physics
Hefei National Laboratory for Physical Sciences at the Microscale
University of Science and Technology of China
Hefei, Anhui 230026, P.R.China
Phone: +86-551-63606428
Mobile: +86-159 0560 4040
E-mail: [email protected]
--
Dr. Federico Iori, PhD!
Dipartimento di Scienze e Metodi per l' Ingegneria (DISMI)
Universita' di Modena e Reggio Emilia
Via Amendola,2
Pad. Morselli
42122 Reggio Emilia (Italy)
