 |
 |
 |
 |
 |
 |
||
 |
 |
 |
 |
 |
| 10 | 1/1 | 返回列表 |
| 查看: 3465 | 回復: 9 | ||||
[交流]
【求助】siesta計算中，如何進行BSSE修正？
|
|
即基組重疊誤差，全稱為Basis Set Superposition Error(BSSE)。對于雙分子A和B的相互作用能應該表述為E_interaction = E_AB - E_A - E_B + E_BSSE。第三項即為修正項。對于弱相互作用，E_BSSE所占E_interaction的比例往往不小，甚至超過它，如果不進行校正，可能正負號都不對。 這些也是從網上查到的，但是網上主要介紹的都是高斯算法中的BSSE修正，由于本人用siesta算，想麻煩各位高手多多賜教��！ |
回復此樓» 收錄本帖的淘帖專輯推薦
 atk&transiesta |
» 猜你喜歡
北科281學碩材料求調劑
已經有13人回復
-
289求調劑
已經有11人回復
-
網絡空間安全0839招調劑
已經有5人回復
-
材料277求調劑
已經有4人回復
-
274求調劑
已經有7人回復
-
085600 材料與化工 329分求調劑
已經有3人回復
-
302求調劑
已經有3人回復
-
312求調劑
已經有5人回復
-
329求調劑
已經有3人回復
-
【2026考研調劑】制藥工程 284分 求相關專業(yè)調劑名額
已經有8人回復
» 本主題相關商家推薦: (我也要在這里推廣)
» 本主題相關價值貼推薦，對您同樣有幫助:
- 如何萃取4-氯丁醇
已經有6人回復
- gaussian計算
已經有4人回復
- 急！ 關于Ms從Windows提交到Linux計算的問題？
已經有3人回復
- 請教：orca 中用CCSD(T)計算單點能出錯--提示內存不足 該怎么辦？
已經有10人回復
- 難道每次計算前都咬進行收斂性測試嗎？
已經有14人回復
- siesta 計算電子占據數(shù)的問題
已經有3人回復
- siesta計算dipole moment（電偶極矩）
已經有7人回復
- 關于SIESTA中如何計算功函數(shù)的一種做法探討
已經有8人回復
- siesta計算弱相互作用VDW
已經有17人回復
- 【求助】siesta進行ordern計算出錯求助
已經有4人回復
- 【求助】SIESTA 中ATOM生成贗勢測試后如何看贗勢的好壞
已經有4人回復
- 【求助】如何在siesta中設置ghost state原子的計算
已經有16人回復
- 【求助】BSSE計算后如何進行零點能校正？
已經有8人回復
ifmc1234
木蟲 (著名寫手)
- 應助: 18 (小學生)
- 金幣: 2797.9
- 散金: 6001
- 紅花: 14
- 帖子: 1761
- 在線: 572.6小時
- 蟲號: 825500
- 注冊: 2009-08-10
- 專業(yè): 理論和計算化學
★ ★ ★
tailor(金幣+2):謝謝參與
qasd(金幣+2):謝謝 2010-09-11 18:57:22
tailor(金幣+2):謝謝參與
qasd(金幣+2):謝謝 2010-09-11 18:57:22
|
不知道你算的是分子還是固體，建議不要用它來討論。你可在userguide里搜一下，我沒看有什么人用siesta討論過BSSE,在userguide里的27頁提到了BSSE。不過具體怎么做還不知道。說明書中只提到PAO.Basis里設置時，用Ghost原子可以處理BSSE。不過如果根據你上面的定義，我想可以這樣E_interaction = E_AB - E_A - E_B + Sum_E_BSSE,其中Sum_E_BSSE為A,B單個的BSSE之和。也就是說用鬼原子來處理BSSE，然后根據上面求的 |
2樓2010-09-06 21:29:23
zhangguangping
木蟲 (著名寫手)
- 1ST強帖: 23
- 應助: 71 (初中生)
- 貴賓: 0.031
- 金幣: 1973.1
- 散金: 9059
- 紅花: 79
- 帖子: 2681
- 在線: 2128.8小時
- 蟲號: 529624
- 注冊: 2008-03-20
- 性別: GG
- 專業(yè): 原子和分子物理
3樓2010-09-06 21:30:41
★
tailor(金幣+2):謝謝參與
tailor(金幣+2):謝謝參與
 |
4樓2010-09-06 21:41:19
請教老外，得到的回復
★
zxzj05(金幣+1):謝謝回帖交流 2010-09-25 22:34:07
zxzj05(金幣+1):謝謝回帖交流 2010-09-25 22:34:07
|
What we calculated is the Basis Set Superposition Error (BSSE) that is an error occurring when you use a finite basis set to study, for example, a chemical reaction. In a tipical reaction you calculate the electronic structure of the reactants separately, let us say A and B, and then the electronic structure of the product, A+B. What happens in this case is that you are using different basis sets to represent the reactants in each step of the calculation. When you calculate the isolated reactant A, you use the basis set of A, and the electrons of A will be arranged spatially on these provided orbitals. However, when you put both systems together in the A+B, the electrons of A will have now a greater spatial freedom, since they can now occupy orbitals coming from B. The same happens with B. You have different basis sets to distribute the electrons of A in the two situations. To correct this, you need to recalculate the energy of the isolated reactants (A and B) by considering the basis set of both reactants, i.e., the calculations of the isolated reactant A will now include the basis sets of B (without the system B itself). It needs to be done for each configuration that you take into account in the reaction of A+B. Now you will have the same basis set for the isolated reactants and the product. You can look for more information searching for "counterpoise correction" |
5樓2010-09-10 19:36:33
zhangguangping
木蟲 (著名寫手)
- 1ST強帖: 23
- 應助: 71 (初中生)
- 貴賓: 0.031
- 金幣: 1973.1
- 散金: 9059
- 紅花: 79
- 帖子: 2681
- 在線: 2128.8小時
- 蟲號: 529624
- 注冊: 2008-03-20
- 性別: GG
- 專業(yè): 原子和分子物理
6樓2010-09-11 10:02:01
夕陽西下
金蟲 (著名寫手)
- 應助: 138 (高中生)
- 貴賓: 0.041
- 金幣: 9247.1
- 散金: 73
- 紅花: 28
- 沙發(fā): 1
- 帖子: 1268
- 在線: 1164.4小時
- 蟲號: 220750
- 注冊: 2006-03-20
- 專業(yè): 凝聚態(tài)物性 II ：電子結構
- 管轄: 物理
7樓2010-09-11 12:08:57
cenwanglai
榮譽版主 (知名作家)
老和山猥瑣派九段
- 1ST強帖: 5
- 應助: 46 (小學生)
- 貴賓: 8.842
- 金幣: 7440.4
- 散金: 9654
- 紅花: 118
- 沙發(fā): 23
- 帖子: 5306
- 在線: 1961.4小時
- 蟲號: 537452
- 注冊: 2008-04-01
- 性別: GG
- 專業(yè): 理論和計算化學
- 管轄: 仿真模擬
8樓2010-09-25 16:50:46
skyscorpioa
新蟲 (初入文壇)
- 應助: 0 (幼兒園)
- 金幣: 529.9
- 散金: 12
- 紅花: 1
- 帖子: 38
- 在線: 76.6小時
- 蟲號: 1089120
- 注冊: 2010-09-04
- 專業(yè): 凝聚態(tài)物性 II ：電子結構
9樓2010-11-20 02:43:11
valenhou001
至尊木蟲 (職業(yè)作家)
- 1ST強帖: 13
- 應助: 241 (大學生)
- 金幣: 25701.7
- 散金: 602
- 紅花: 166
- 帖子: 3782
- 在線: 873.8小時
- 蟲號: 1007127
- 注冊: 2010-04-27
- 專業(yè): 凝聚態(tài)物性 II ：電子結構
★
tailor(金幣+2): 謝謝參與
tailor(金幣+2): 謝謝參與
|
這里有很詳細的步驟： https://www.mail-archive.com/siesta-l@uam.es/msg02916.html 拷貝如下： Re: Re: [SIESTA-L] about basis-set superposition error (BSSE) Abraham Hmiel Tue, 29 Mar 2011 21:51:07 -0700 "Should I set a ghost atom between the molecule and the nanotube ?" No, this is what you must do: You've already got the relaxed structure of the nanowire + adsorbed molecule system, right? Let's call this system "A" and its total energy is E-A You need 6 more calculations to complete the puzzle this is pretty much what you should do: 1 calculation with the same k-point grid, mesh grid, and cell size as system A, but only for the nanotube, fully relaxed to a similar tolerance as system A. Find its total energy. Call this E-NT. 1 calculation with the same k-point grid, mesh grid, and cell size as system A, but only for the adsorbed molecule, fully relaxed to a similar tolerance as system A. Find its total energy. Call this E-AD 1 single-point calculation (0 CG steps) of the nanowire in the relaxed geometry of system A (use the final .XV file or .xyz file or whatever you want but remove the adsorbed molecule). Find its total energy. Call this E-noghost-NT 1 single-point calculation (0 CG steps) of the adsorbed molecule in the relaxed geometry of system A (use the final .XV file or .xyz file or whatever you want but remove the adsorbed molecule). Find its total energy. Call this E-noghost-AD 1 single-point calculation (0 CG steps) of the nanowire in the relaxed geometry of system A, except replace any adsorbate chemical species with ghost atoms. If you have species in the adsorbate that are present in the nanowire, for example, a simulation of H2O on a hydrogen-passivated SiNW (or C in methane on a CNT), then copy the H.psf file to a new file like H_ghost.psf and then create a new chemical species H_ghost with atomic number -1 and a different atomic species index. Replace any index of the H in the adsorbate with the new index, and introduce a new basis set for H_ghost that is identical to the one you used for H (except for the label H_ghost). Find its total energy. Call this E-ghost-NT 1 single-point calculation (0 CG steps) of the adsorbed molecule in the relaxed geometry of system A, except replace any nanowire chemical species with ghost atoms and follow the procedure above if you have any species in the adsorbate that are also present in the nanowire. Find its total energy. Call this E-ghost-AD The counterpoise correction is: (E-ghost-AD - E-noghost-AD + E-ghost-NT - E-noghost-NT) call this E-CC. It should be a fraction of an eV, have a negative sign and very sensitive to the adsorption site geometry. The BSSE should _reduce_ the adsorption energy... and the energy of adsorption is: E-NT + E-AD - E-A + E-CC And that is how you do the counterpoise correction with SIESTA. Best, Abraham Hmiel Katherine Belz Groves Graduate Fellow in Nanoscience, Xue Group The College of Nanoscale Science and Engineering at SUNY Albany "Clouds are not spheres, mountains are not cones, coastlines are not circles, and bark is not smooth, nor does lightning travel in a straight line." - Benoit Mandelbrot On Wed, Mar 30, 2011 at 12:09 AM, yf liu <liuyf1...@gmail.com> wrote: > Dear Herbert Fruchtl: > thank you very much for your reply. I have read the manual carefully, > but find little message about the ghost atoms. the new question is: How can > I use the ghost atom to correct the binding energy etc. ? For example, I > want correct the calculation about the molecule adsorption on the nanotube. > Should I set a ghost atom between the molecule and the nanotube ? > > looking forward your reply. > > 2011/3/20 Wei Hu <gyrw4...@mail.ustc.edu.cn> > >> I am sorry, I am a freshman about the siesta. The input is listed in >> the following. >> My calculations about the binding energy do not meet the experimental >> results,so,I have to check the ghost atoms calculations.Now,The impact is >> acceptable. >> Another problem is how to get the chemical potential or total energy of >> free C or N atom. Does it need to calculate the the ghost atoms affected by >> the supercell? >> >> SystemName C62N_ghost >> SystemLabel C62N_ghost >> NumberOfSpecies 3 >> >> %block ChemicalSpeciesLabel >> 1 6 C >> 2 7 N >> 3 -6 Cg >> %endblock ChemicalSpeciesLabel >> >> %block PS.lmax >> C 1 >> %endblock PS.lmax >> >> %include coord.fdf >> >> PAO.BasisSize DZP >> >> #SolutionMethod dm_on >> >> >> SolutionMethod diagon >> >> MeshCutoff 200.0000000 Ry >> >> #MD.TypeOfRun Broyden >> #MD.TypeOfRun CG >> #MD.NumCGsteps 500 >> >> WriteForces >> >> MD.MaxForceTol 0.04 eV/Ang >> #DM.UseSaveDM T >> >> MaxSCFIterations 100 >> DM.MixingWeight 0.1 >> DM.NumberPulay 6 >> #DM.MixingWeight 0.25 >> #DM.NumberPulay 0 >> >> >> SpinPolarized .true. >> #FixSpin .true. >> #TotalSpin 2.0 >> >> WriteMullikenPop 1 >> >> NetCharge -1.0 >> >> %block kgrid_Monkhorst_Pack >> 2 0 0 0.0 >> 0 2 0 0.0 >> 0 0 2 0.0 >> %endblock kgrid_Monkhorst_Pack >> >> >> >> > -----Original E-mail----- >> > From: "Herbert Fruchtl" <herbert.fruc...@st-andrews.ac.uk> >> > Sent Time: 2011-3-18 20:06:46 >> > To: siesta-l@uam.es >> > Cc: >> > Subject: Re: [SIESTA-L] about basis-set superposition error (BSSE) >> > >> > From the energy part of the output alone we can't tell if the input was >> > correct. You see a smallish difference and lower energy with the ghost, >> which >> > one would expect. >> > >> > Total energies are meaningless, and of course adding a few basis >> functions won't >> > change the total by a lot. You need to compare differences (like a >> binding >> > energy, as you indicated in your original post). >> > >> > In the example outputs below, the difference (in absolute energies) is >> 0.09 eV. >> > That's 2 kcal/mol or 9 kJ/mol. Depending on the interaction you are >> looking at, >> > this may or may not be negligible. >> > >> > Herbert >> > >> > On 03/18/2011 02:00 AM, Wei Hu wrote: >> > > Dear,the ghost atoms calculations seem to little effects on the total >> energy from my results. Is there any wrong? >> > > >> > > My results about a nitrogen-vacancy center in bulk diamond C62N >> (2*2*2): >> > > >> > > The results with ghost atoms: >> > > >> > > siesta: Program's energy decomposition (eV): >> > > siesta: Eions = 16680.576136 >> > > siesta: Ena = 3653.202863 >> > > siesta: Ekin = 7163.588024 >> > > siesta: Enl = -820.400928 >> > > siesta: DEna = -182.458272 >> > > siesta: DUscf = 15.575306 >> > > siesta: DUext = 0.000000 >> > > siesta: Exc = -3054.613637 >> > > siesta: eta*DQ = 0.000000 >> > > siesta: Emadel = 0.000000 >> > > siesta: Ekinion = 0.000000 >> > > siesta: Eharris = -9905.679557 >> > > siesta: Etot = -9905.682781 >> > > siesta: FreeEng = -9905.682783 >> > > >> > > siesta: Final energy (eV): >> > > siesta: Kinetic = 7163.588024 >> > > siesta: Hartree = 894.608032 >> > > siesta: Ext. field = 0.000000 >> > > siesta: Exch.-corr. = -3054.613637 >> > > siesta: Ion-electron = -8393.758885 >> > > siesta: Ion-ion = -6515.506314 >> > > siesta: Ekinion = 0.000000 >> > > siesta: Total = -9905.682781 >> > > >> > > The results without ghost atoms: >> > > >> > > siesta: Program's energy decomposition (eV): >> > > siesta: Eions = 16680.576136 >> > > siesta: Ena = 3653.202863 >> > > siesta: Ekin = 7164.385562 >> > > siesta: Enl = -820.554764 >> > > siesta: DEna = -182.943454 >> > > siesta: DUscf = 15.589130 >> > > siesta: DUext = 0.000000 >> > > siesta: Exc = -3054.694636 >> > > siesta: eta*DQ = 0.000000 >> > > siesta: Emadel = 0.000000 >> > > siesta: Ekinion = 0.000000 >> > > siesta: Eharris = -9905.587400 >> > > siesta: Etot = -9905.591435 >> > > siesta: FreeEng = -9905.591437 >> > > >> > > siesta: Final energy (eV): >> > > siesta: Kinetic = 7164.385562 >> > > siesta: Hartree = 894.861106 >> > > siesta: Ext. field = 0.000000 >> > > siesta: Exch.-corr. = -3054.694636 >> > > siesta: Ion-electron = -8394.637153 >> > > siesta: Ion-ion = -6515.506314 >> > > siesta: Ekinion = 0.000000 >> > > siesta: Total = -9905.591435 >> > > >> > > >> > > >> > > >> > >> -----Original E-mail----- >> > >> From: "Herbert Fruchtl"<herbert.fruc...@st-andrews.ac.uk> >> > >> Sent Time: 2011-3-17 22:36:53 >> > >> To: siesta-l@uam.es >> > >> Cc: >> > >> Subject: Re: [SIESTA-L] about basis-set superposition error (BSSE) >> > >> >> > >> You can calculate counterpoise-corrected energies by specifying ghost >> atoms. >> > >> There is no way of doing counterpoise-corrected geometry >> optimisations in SIESTA. >> > >> >> > >> Herbert >> > >> >> > >> On 03/17/2011 01:57 PM, yf liu wrote: >> > >>> Hello everybody, >> > >>> >> > >>> whether the basis-set superposition error (BSSE) is included in >> siesta >> > >>> calculation. when i calculate the molecule adsorption on carbon >> nanotube, i >> > >>> found the adsorption energy is some higher than the well know value. >> how can i >> > >>> add the BSSE in siesta calculation? >> > >>> >> > >>> yours >> > >>> yufeng liu >> > >>> >> > >> >> > >> -- >> > >> Herbert Fruchtl >> > >> Senior Scientific Computing Officer >> > >> School of Chemistry, School of Mathematics and Statistics >> > >> University of St Andrews >> > >> -- >> > >> The University of St Andrews is a charity registered in Scotland: >> > >> No SC013532 >> > > >> > >> > -- >> > Herbert Fruchtl >> > Senior Scientific Computing Officer >> > School of Chemistry, School of Mathematics and Statistics >> > University of St Andrews >> > -- >> > The University of St Andrews is a charity registered in Scotland: >> > No SC013532 >> >> > |
10樓2014-12-02 09:25:05
相關版塊跳轉
我要訂閱樓主 tailor 的主題更新
| 10 | 1/1 | 返回列表 |
| 最具人氣熱帖推薦 [查看全部] | 作者 | 回/看 | 最后發(fā)表 | |
|---|---|---|---|---|
|
[考研] 北科281學碩材料求調劑 +11 | tcxiaoxx 2026-03-20 | 13/650 |
|
|---|---|---|---|---|
|
[考研] 網絡空間安全0839招調劑 +3 | w320357296 2026-03-25 | 5/250 |
|
|
|
[考研]
 085602 289分求調劑
+7
|
WWW西西弗斯 2026-03-24 | 7/350 |
|
|
|
[考研]
材料與化工考研調劑
+9
|
孅華 2026-03-22 | 9/450 |
|
|
|
[考研] 求調劑 +3 | 李李不服輸 2026-03-25 | 3/150 |
|
|
|
[考研] 生物學學碩求調劑 +7 | 小羊睡著了? 2026-03-23 | 10/500 |
|
|
|
[考研] 材料學求調劑 +6 | Stella_Yao 2026-03-20 | 6/300 |
|
|
|
[考研] 085601求調劑總分293英一數(shù)二 +3 | 鋼鐵大炮 2026-03-24 | 3/150 |
|
|
|
[考研] 0854AI CV方向招收調劑 +3 | 章小魚567 2026-03-23 | 3/150 |
|
|
|
[考研] 307求調劑 +5 | 超級伊昂大王 2026-03-24 | 5/250 |
|
|
|
[考研]
 一志愿東華大學化學070300，求調劑
+7
|
2117205181 2026-03-21 | 8/400 |
|
|
|
[基金申請] 山東省面上項目限額評審 +4 | 石瑞0426 2026-03-19 | 4/200 |
|
|
|
[考研] 材料求調劑 +5 | @taotao 2026-03-21 | 5/250 |
|
|
|
[考研] 求調劑 +4 | 要好好無聊 2026-03-21 | 4/200 |
|
|
|
[考研] 0703化學調劑 +4 | 妮妮ninicgb 2026-03-21 | 4/200 |
|
|
|
[考研] 332求調劑 +3 | 鳳凰院丁真 2026-03-20 | 3/150 |
|
|
|
[考研] 321求調劑 +9 | 何潤采123 2026-03-18 | 11/550 |
|
|
|
[考研] 一志愿蘇州大學材料求調劑，總分315（英一） +5 | sbdksD 2026-03-19 | 5/250 |
|
|
|
[考研] 一志愿西安交通大學 學碩 354求調劑211或者雙一流 +3 | 我想要讀研究生 2026-03-20 | 3/150 |
|
|
|
[考研] 本科鄭州大學物理學院，一志愿華科070200學碩，346求調劑 +4 | 我不是一根蔥 2026-03-18 | 4/200 |
|