24小时热门版块排行榜    

查看: 2748  |  回复: 17
【有奖交流】积极回复本帖子,参与交流,就有机会分得作者 shijing879 的 9 个金币 ,回帖就立即获得 1 个金币,每人有 1 次机会

shijing879

新虫 (初入文坛)


[交流] BSE+GW计算光学吸收谱(激子结合能)

大家好,请问大家谁有BSE+GW计算光学吸收谱(激子结合能)的INCAR设置步骤,KPOINTS和注意事项,要求能计算成功,我要计算的是二维材料,官网和网上的一些教程就不要推荐了,我已经试过!
回复此楼

» 猜你喜欢

» 本主题相关商家推荐: (我也要在这里推广)

» 抢金币啦!回帖就可以得到:

查看全部散金贴

已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖
回帖支持 ( 显示支持度最高的前 50 名 )

wangyikeco

木虫 (正式写手)



shijing879(金币+1): 谢谢参与
VASP sucks!

Goal: account for excitonic effects in the frequency dependent dielectric function
• GW0+BSE: Solve the Bethe-Salpeter-Equation on top of GW0
WORKFLOW of a GW0+BSE calculation
The workflow of the GW0+BSE calculation in job.sh consists of five consecutive steps:
Step 1: a “standard" DFT groundstate calculation
Step 2: compute additional DFT “virtual” orbitals (empty states):
• Needs the WAVECAR from Step 1.
Step 3: the GW0 calculation:
• Needs the WAVECAR and WAVEDER files from Step 2.
Step 4: (optional) use LOPTICS=.TRUE. to plot the IPA dielectric function using
the GW0-QP energies instead of DFT eigenenergies
• Needs the WAVECAR file from Step 3.
Step 5: the BSE calculation:
• Needs WAVECAR file from Step 3 and WAVEDER file from Step 2.
6楼2023-08-28 10:16:03
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

wangyikeco

木虫 (正式写手)


引用回帖:
7楼: originally posted by shijing879 at 2023-08-28 15:48:51
这些我都知道,请问是否有每一步具体的incar,kpoints, thanks in advance!...

Step 1: a “standard” DFT groundstate calculation
KPOINTS (e.g. KPOINTS.6):
6x6x6
0
G
6 6 6
0 0 0
INCAR (copy INCAR.DFT to INCAR):
PREC = Normal
ENCUT = 250
Default
Set cutoff energy to 250 eV
ISMEAR = 0
SIGMA = 0.01
Gaussian smearing
Set small smearing width
KPAR = 2 Additional parallelization
EDIFF = 1.E-8 Set tight convergence criterium
8楼2023-08-29 02:56:10
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

wangyikeco

木虫 (正式写手)


•Step 2: compute additional DFT “virtual” orbitals (empty states)
INCAR (copy INCAR.DIAG to INCAR):
PREC = Normal ; ENCUT = 250
ALGO = Exact
NELM = 1
Exact diagonalisation of the Hamiltonian
Only 1 electronic step
ISMEAR = 0 ; SIGMA = 0.01
NBANDS = 128 Keep 128 bands after diagonalisation
LOPTICS = .TRUE.
LPEAD = .TRUE.
Compute derivative of the orbitals w.r.t. the
Bloch wave vector (stored in the WAVEDER file)
using the PEAD formalism
OMEGAMAX = 40

N.B.: This calculation needs the orbitals (WAVECAR file) written in Step 1.
9楼2023-08-29 02:58:01
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

wangyikeco

木虫 (正式写手)


Step 3: RPA quasiparticles: single-shot GW (G0W0)
INCAR (copy INCAR.GW0 INCAR):

PREC = Normal ; ENCUT = 250
ALGO = GW0 ; NELM = 1 Request G0W0 calculation
ISMEAR = 0 ; SIGMA = 0.01
ENCUTGW = 150
NOMEGA = 50
OMEAGTL = 280
PRECFOCK = Normal
Eenergy cutoff for response functions
Number of points for frequency integration
NBANDS = 128 Use the same #-of-bands as in Step 2., otherwise the
WAVEDER file can not be read correctly
NBANDSGW = 12 Compute QP energies only for the first 12 bands
KPAR = 2
LWAVE = .TRUE. Write WAVECAR (contains QP-energies)

N.B.: This calculation needs the orbitals (WAVECAR file) , and the derivative of
the orbitals w.r.t. the Bloch wave vectors (WAVEDER file) written in Step 2.
10楼2023-08-29 03:03:57
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

wangyikeco

木虫 (正式写手)


Step 4: (optional) plot IPA dielectric function using G0W0QP-energies
INCAR (copy INCAR.NONE INCAR):
ALGO = Nothing ; NELM = 1 Do nothing except read WAVECAR and
proceed to post-processing
NBANDS = 128
LWAVE = .FALSE. Do not write WAVECAR
LOPTICS = .TRUE. ; LPEAD = .TRUE. Compute IPA dielectric function
11楼2023-08-29 03:07:17
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

wangyikeco

木虫 (正式写手)


Step 5: the BSE calculation
INCAR (copy INCAR.BSE INCAR):
PREC = Normal ; ENCUT = 250
ALGO = BSE
ANTIRES = 0
Request BSE calculation
Use the Tamm-Dancoff approximation
ISMEAR = 0 ; SIGMA = 0.01
ENCUTGW = 150 Energy cutoff for response functions
NBANDS = 128 Use the same #-of-bands as in Step 2., otherwise the WAVEDER
file can not be read correctly
NBANDSO = 4
NBANDSV = 8
Setup BSE matrix for NBANDSO HOMOs and
NBANDSV LUMOs
OMEGAMAX = 20
PRECFOCK = Normal

N.B.: This calculation needs the orbitals (WAVECAR file) of Step 3, and the derivative
of the orbitals w.r.t. the Bloch wave vectors (WAVEDER file) written in Step 2.
12楼2023-08-29 03:10:12
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

shijing879

新虫 (初入文坛)


引用回帖:
6楼: Originally posted by wangyikeco at 2023-08-28 10:16:03
VASP sucks!

Goal: account for excitonic effects in the frequency dependent dielectric function
• GW0+BSE: Solve the Bethe-Salpeter-Equation on top of GW0
WORKFLOW of a GW0+BSE calculation
...

这些我都知道,请问是否有每一步具体的INCAR,KPOINTS, Thanks in advance!
7楼2023-08-28 15:48:51
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

shijing879

新虫 (初入文坛)


引用回帖:
12楼: Originally posted by wangyikeco at 2023-08-29 03:10:12
Step 5: the BSE calculation
INCAR (copy INCAR.BSE INCAR):
PREC = Normal ; ENCUT = 250
ALGO = BSE
ANTIRES = 0
Request BSE calculation
Use the Tamm-Dancoff approximation
ISMEAR = 0 ; SIGMA = 0.0 ...

这些我也知道,都是官网上面的,我已经按照这个算过了,没成功。请问你有没自己亲身计算成功的一些算例。
13楼2023-08-29 15:19:18
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

wangyikeco

木虫 (正式写手)


引用回帖:
13楼: Originally posted by shijing879 at 2023-08-29 15:19:18
这些我也知道,都是官网上面的,我已经按照这个算过了,没成功。请问你有没自己亲身计算成功的一些算例。...

What  version of VASP are you doing?
14楼2023-08-30 09:32:08
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

shijing879

新虫 (初入文坛)


引用回帖:
14楼: Originally posted by wangyikeco at 2023-08-30 09:32:08
What  version of VASP are you doing?...

我用的是vasp6.3.2
15楼2023-08-31 14:55:35
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

wangyikeco

木虫 (正式写手)


引用回帖:
15楼: Originally posted by shijing879 at 2023-08-31 14:55:35
我用的是vasp6.3.2...

我的老版本上似乎没问题
你升级到最新的6.4.2
或者把你的script发给我, zrli3 AT hotmail dot com
16楼2023-09-03 02:23:57
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

理想的美好

新虫 (初入文坛)



shijing879(金币+1): 谢谢参与
您好,楼主,我想请教您如何用 GW+BSE 来计算激子结合能,同时给它画成能带形式
BSE+GW计算光学吸收谱(激子结合能)



发自小木虫Android客户端
17楼2024-11-19 18:37:07
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖
简单回复
2023-08-27 15:44   回复  
shijing879(金币+1): 谢谢参与
toppaper3楼
2023-08-27 16:11   回复  
shijing879(金币+1): 谢谢参与
发自小木虫Android客户端
tzynew4楼
2023-08-27 16:53   回复  
shijing879(金币+1): 谢谢参与
2023-08-27 19:53   回复  
shijing879(金币+1): 谢谢参与
XG-WUST18楼
2024-11-24 07:11   回复  
相关版块跳转 我要订阅楼主 shijing879 的主题更新
提示: 如果您在30分钟内回复过其他散金贴,则可能无法领取此贴金币
普通表情 高级回复 (可上传附件)
信息提示
请填处理意见