User Tools

Site Tools


exercises:2017_ethz_mmm:qmmm

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision
Previous revision
Next revisionBoth sides next revision
exercises:2017_ethz_mmm:qmmm [2017/06/02 02:28] dpasseroneexercises:2017_ethz_mmm:qmmm [2017/06/02 03:43] dpasserone
Line 2: Line 2:
 === (exercise by Matthew Watkins, University college, London) === === (exercise by Matthew Watkins, University college, London) ===
 In this exercise you will validate the mixed quamtum/classical model for a KCl slab.  In this exercise you will validate the mixed quamtum/classical model for a KCl slab. 
-Hte : [[doi>10.1021%2Fja505936b]].+The present exercise is referring to the following paper: [[doi>10.1002/jcc.23904]].
  
 <note tip>  <note tip> 
-  * In the first part of the exercise you will consider the optimized configuration (already in the directory) and study the pure electronic adsorption energy, namely the difference between the total energy of the surface-molecule system and the energy of the molecule alone and surface alone **in the same geometry as the surface-molecule system minimum structure**. This will allow to show the binding pattern of the electronic density+  * You will optimize the geometry of a KCL slab with the same arrangement as the one depicted below 
-  * In the second part, you will optimize the surface and the molecule separately; this will allow to compute the total adsorption energy.+{{ :exercises:2017_ethz_mmm:screen_shot_2017-06-02_at_05.21.07.png?400 |}} 
 +  For simplicity, you will consider three layers. 
 +  You will compare the calculation performed with the full QM, one layer QM and two MM, two layer QM and one MM.  
 +  * In particular you will compare the band gaps and the density of states.
 </note>  </note> 
  
  
- {{ :exercises:2017_ethz_mmm:master.img-002.jpg?nolink&600 |}} 
  
  
-===== 1. Task: Familiarize yourself  ===== +===== 1. Task: Prepare the input files  ===== 
-The coordinates of the optimized configuration are provided to you as ''S_M.opt.xyz'' (S stands for "Substrate", M for "Molecule", opt for "optimized")Visualize the geometry with VMD and familiarize yourself with the system+The file input.inp contains the partition of the QM and MM regions by specifying the atom index. Check in the ''kcl.xyz'' how the indexes are distributedThe slab orthogonal coordinate is the ''y''.  
- +<note important>   
-===== 2Task: Bond induced density differences ===== +   *  Make three copies of **input.inp** and call them **qm_1l.inp**, **qm_2l.inp** and **qm.inp**.  
-Compute the density difference induced by the adsorption bonding +   * **qm_1l.inp** should have ''PROJECT KCl_1'' and one QM layerModify accordingly the MM_INDEX lines by looking first at the ''kcl.xyz'' coordinate files. 
-For this you will have to run three separate energy calculationsusing the *.ene.inp files.  +   * **qm_2l.inp** should have ''PROJECT KCl_2'' and two QM layersModify accordingly the MM_INDEX lines by looking first at the ''kcl.xyz'' coordinate files. 
-  - combined system  (file ''S_M.opt.xyz'') +   * **qm.inp** should have ''PROJECT KCl_QM'' and the full QM treatment. For thisit is sufficient to change ''QMMM'' at the beginning to ''QS''. Change also the input coordinates from kcl.xyz to kcl_opt.xyz: these are the already optimized coordinates for the full QM treatment. In this way you will spare time.
-  - lone acetylene (file ''M.S_M.xyz'') +
-  - lone slab (file ''S.S_M.xyz''+
- +
-In order to output the electronic densities as cube filesyour input file has to contain the following snipped: +
-<code> +
-&DFT +
-  &PRINT +
-    &E_DENSITY_CUBE +
-    &END E_DENSITY_CUBE +
-  &END +
-&END DFT +
-</code> +
- +
-<note tip> +
-The calculations involving the slab should be run on at least 16 cores with ''qsub run -v INP=prefix''. Check the  ''run'' file for the number of nodes.+
 </note> </note>
  
-To process the cube files we are going to use the [[tools:cubecruncher | cubecruncher]] tool. It is part of CP2K and is in your exercise directory. +===== 2. Task: Run the jobs  =====
-<code> +
-you@eulerX ~$ ./cubecruncher.x -i S_M-ELECTRON_DENSITY-1_0.cube -subtract S-ELECTRON_DENSITY-1_0.cube -o tmp.cube +
-you@eulerX ~$ ./cubecruncher.x -i tmp.cube -subtract M-ELECTRON_DENSITY-1_0.cube -o Delta_ads.cube +
-</code> +
- +
- +
-The generated cube file is not aligned with the simulation cell. Center the cube file with the cubecruncher.x tool: +
-<code> +
-you@eulerX ~$ ./cubecruncher.x -center geo -i Delta_ads.cube -o Delta_ads-centered.cube +
-</code> +
- +
-You can visualize the resulting file ''delta_ads-centered.cube'' with VMD. This has been covered in a [[reaction_energy_2017| previous exercise]]. +
- +
-What you get should look similar to this:  +
-{{ dye_tio_bonding_density.png?300 |}} +
- +
-===== 3. Task: Bonding energies  ===== +
-Compute the binding energy: +
- +
-\[ E_\text{binding}=\sum E_\text{products} - \sum E_\text{reactants} \] +
- +
-For this you will need the energy values of three systems: +
-  - lone acetylene molecule (run geometry optimization, use energy of last step) +
-  - lone  slab  (you can use the already geometry optimized coordinates from ''S.opt.xyz'' at the end of the exercise) +
-  - combined system adsorbed (can be reused from previous task) +
 <note important> <note important>
-You can not reuse the energy values for the lone sub-systems from the previous taskSince the unbound subsystems might relax into different geometrythey have to be geometry optimized first+  * Run the jobs by giving the command: ''qsub run -v INP=qm_1l'' and similarly for the other input files. 
 +  * You will also get cube files for hartree potential and electronic density. They can be examined with vmd. 
 +  * For each job, ''*pos*xyz'' optimization file is producedas well as two PDOS files, one for the species 1 (K), the other for the species 2 (Cl).
 </note> </note>
 +
  
  
 ===== Questions ===== ===== Questions =====
-  * Sketch briefly the geometry of the molecule **when adsorbed** and **in the gas phase**. +  
-  * Report the system energy for the bonded system, lone slab, and lone molecule. +
-  * Can you estimate the contribution due to the geometry relaxation?  +
-  * Briefly report the bond induced density difference on the system.+
  
 ===== Required Files ===== ===== Required Files =====
-<note tip> When you are dealing with big systems and multiple atomic species, the input can be simplified by splitting it into multiple files. We are going to use separate files for the coordinates, the basis-setsand the pseudo-potentials. All these files should reside in the same directory as the main input file.</note>+<note tip> When you are dealing with complex job structure, the input can be simplified by splitting it into multiple files. We are going to use separate files for the coordinates, the QM part, the MM part. All these files should reside in the same directory as the main input file.</note>
  
 <note warning> <note warning>
-The provided files are all in the directory ''/home/psd/Exercise_9''Change the name of the xyz file accordingly in the input files.+The provided files are all in the directory ''/home/psd/Exercise_13''When you create new input files with different parameters, remember to change the name of the **PROJECT** as well.
 </note> </note>
  
Line 114: Line 75:
     #but should be treated as parameters in general     #but should be treated as parameters in general
     #fit to some physical property     #fit to some physical property
-    &MM_KIND K +
-      RADIUS 1.52 +
-    &END MM_KIND+
     &MM_KIND Cl     &MM_KIND Cl
-      RADIUS 1.67 
-    &END MM_KIND 
-    #define the model 
-    &QM_KIND K 
-      MM_INDEX 25..32 41..48 
-    &END QM_KIND 
-        &MM_KIND Cl 
       RADIUS 1.67       RADIUS 1.67
     &END MM_KIND     &END MM_KIND
exercises/2017_ethz_mmm/qmmm.txt · Last modified: 2020/08/21 10:15 by 127.0.0.1