exercises:2015_cecam_tutorial:neb
Differences
This shows you the differences between two versions of the page.
| Both sides previous revisionPrevious revisionNext revision | Previous revision | ||
| exercises:2015_cecam_tutorial:neb [2015/08/20 12:56] – add citations tmueller | exercises:2015_cecam_tutorial:neb [2020/08/21 10:15] (current) – external edit 127.0.0.1 | ||
|---|---|---|---|
| Line 2: | Line 2: | ||
| Problem: compute activation barrier for the last step (6 to 7) of the cyclodehydrogenation reaction CHP@Cu(111) -> TBC | Problem: compute activation barrier for the last step (6 to 7) of the cyclodehydrogenation reaction CHP@Cu(111) -> TBC | ||
| + | |||
| + | {{ : | ||
| + | |||
| + | * Original author: Carlo Pignedoli | ||
| + | * Complete source and output files: [[http:// | ||
| ===== Introduction ===== | ===== Introduction ===== | ||
| The system is composed by a molecule (56 atoms) adsorbed on a Cu(111) slab (~2900 atoms). | The system is composed by a molecule (56 atoms) adsorbed on a Cu(111) slab (~2900 atoms). | ||
| - | The molecule is treated within DFT (PM6[(Stewart2007> | + | The molecule is treated within DFT (PM6[(Stewart2007> |
| + | and the substrate with EAM[(Daw1983> | ||
| The substrate and the molecule are coupled through a par potential mimicking the van der Waals interaction and Pauli repulsion.[(Levi2007> | The substrate and the molecule are coupled through a par potential mimicking the van der Waals interaction and Pauli repulsion.[(Levi2007> | ||
| Line 17: | Line 23: | ||
| * and in the third section we define the parameters for the DFT part. | * and in the third section we define the parameters for the DFT part. | ||
| * The fourth and final section defines the target of the calculation(GEO_OPT, | * The fourth and final section defines the target of the calculation(GEO_OPT, | ||
| + | |||
| + | The files for this part of the exercise are found in the **'' | ||
| ==== First Section ==== | ==== First Section ==== | ||
| Line 79: | Line 87: | ||
| In this section we define the empirical potentials for the whole system: | In this section we define the empirical potentials for the whole system: | ||
| - | EAM will be used for the Cu-Cu interactions (parametrized in the file CP2KTUTORIAL/ | + | EAM will be used for the Cu-Cu interactions (parametrized in the file '' |
| - | a “C6/R^6 + pauli repulsion” will be defined for C-Cu and H-Cu and, finally, | + | a //C6/R^6 + pauli repulsion// will be defined for C-Cu and H-Cu and, finally, |
| NULL interactions will be defined for C-C, C-H and H-H since these will be obtained from DFT in the next section. | NULL interactions will be defined for C-C, C-H and H-H since these will be obtained from DFT in the next section. | ||
| The charge of every specie (zero in our example) as well as the interaction within each possible type-pair has to be defined. | The charge of every specie (zero in our example) as well as the interaction within each possible type-pair has to be defined. | ||
| Line 280: | Line 288: | ||
| ==== Tasks ==== | ==== Tasks ==== | ||
| - | - First of all obtain the equilibrium geometry for the initial and the final states (in the directories INI and FIN you will find the input files (geo.inp) the output files (out) and the trajectory of the geometry optimization (INI-pos-1.xyz and FIN-pos-1.xyz). Please note that the geometry optimization of the final state is constrained with respect to the position of the final H2 molecule. | + | - First of all obtain the equilibrium geometry for the initial and the final states (in the directories |
| - | - Perform a NEB simulation to find a //saddle point// between ini_eq.xyz and fin_eq.xyz. A linear guess for the path is not a good idea in this case; a clever guess can be obtained from a series of constrained geometry optimizations where the H-H distance is forced to vary from its initial value (more than 3A)to the H2 equilibrium distance. Such a guess is given, in the directory EX2, with the files ini_eq.xyz, due.xyz, | + | - Please note that the geometry optimization of the final state is constrained with respect to the position of the final H2 molecule. |
| + | - Perform a NEB simulation to find a //saddle point// between | ||
| + | - A linear guess for the path is not a good idea in this case; a clever guess can be obtained from a series of constrained geometry optimizations where the H-H distance is forced to vary from its initial value (more than 3 Å) to the H2 equilibrium distance. Such a guess is given, in the directory | ||
| The last section of the input described above is modified as follows: | The last section of the input described above is modified as follows: | ||
| Line 347: | Line 357: | ||
| <note important>'' | <note important>'' | ||
| + | {{ : | ||
| ===== Exercise 2: Metadynamics ===== | ===== Exercise 2: Metadynamics ===== | ||
| + | |||
| + | The files for this part of the exercise are found in the **'' | ||
| ==== Part 1 ==== | ==== Part 1 ==== | ||
| Obtain two restart files for independent configurations of the system at the geometry of the initial state thermalized at 450 K. | Obtain two restart files for independent configurations of the system at the geometry of the initial state thermalized at 450 K. | ||
| - | In the directory EQUIL the input file “inp” is used to run 3000 MD steps of MD in the NVT ensamble with a CSVR thermostat (a longer simulation should be used) two restart files obtained during the simulation (EQUIL-1_2000.restart and EQUIL-1_3000.restart) will be used to run a Metadynamics simulation with two WALKERS. | + | In the directory |
| The trajectory of the MD should be also used to define the shape of the gaussians added in the metadynamics run. | The trajectory of the MD should be also used to define the shape of the gaussians added in the metadynamics run. | ||
| - | Trajectories are printed in dcd format (EQUIL-pos-1.dcd use, e.g. “vmd s.xyz EQUIL-pos-1.dcd” to visualize the trajectory). | + | Trajectories are printed in dcd format ('' |
| The '' | The '' | ||
| Line 398: | Line 411: | ||
| ==== Part 2 ==== | ==== Part 2 ==== | ||
| - | The two restart files are copied in the directory '' | + | The two restart files are copied in the directory '' |
| - | In the directory also the geometry for the whole system in the initial configuration as well as the geometry of the molecular fragment are included (files | + | In the directory also the geometry for the whole system in the initial configuration as well as the geometry of the molecular fragment are included (files |
| We perform a metadynamics simulation with two collective variables: the sum of C-H distances (CV1) and the H-H distance (CV2). | We perform a metadynamics simulation with two collective variables: the sum of C-H distances (CV1) and the H-H distance (CV2). | ||
| Walls have to be introduced to limit the range of variability of the two CVs. | Walls have to be introduced to limit the range of variability of the two CVs. | ||
| + | |||
| + | {{ : | ||
| To define the collective variables introduce a '' | To define the collective variables introduce a '' | ||
| Line 571: | Line 586: | ||
| </ | </ | ||
| - | Here we declare that we are using two walkers, that the input belongs to the walker with ID 1, that the data relative to the different walkers are stored in the files containe din the directoty | + | Here we declare that we are using two walkers, that the input belongs to the walker with ID 1, that the data relative to the different walkers are stored in the files contained in the directory '' |
| - | We need two independent directories (WALK1 and WALK2 in the example) belonging to each WALKER, each directory will contain an input in which the WALKER ID is modified (1 and 2). | + | We need two independent directories ('' |
| The calculation will run as a FARMING: we need a common input that specifies that we are running two cp2k problems: | The calculation will run as a FARMING: we need a common input that specifies that we are running two cp2k problems: | ||
| - | The inp file for the calculation (inp in the directory EX1) will be: | + | The input file for the calculation ('' |
| < | < | ||
| Line 599: | Line 614: | ||
| We declare that the cpus used for the calculation belongs to two groups of calculations, | We declare that the cpus used for the calculation belongs to two groups of calculations, | ||
| - | one job has to run in the directory | + | one job has to run in the directory |
| - | and a second job has to run in the directory WALK2. | + | and a second job has to run in the directory |
| - | If we run the calculation with 8 cpus, 4 will be devoted to the JOB in directory WALK1 and 4 to WALK2. | + | If we run the calculation with 8 cpus, 4 will be devoted to the JOB in directory |
| - | Each group of 10 cpus will be then partitioned according to partition instruction contained in each inp file (GROUP_PARTITION 1 3). | + | Each group of 10 cpus will be then partitioned according to partition instruction contained in each '' |
| During execution you can monitor the evolution of the CVs of each walker from the files | During execution you can monitor the evolution of the CVs of each walker from the files | ||
| - | WALK1/ | + | '' |
| e.g. | e.g. | ||
| Line 619: | Line 634: | ||
| fes.popt -cp2k -mathlab -file META-META-1.restart -ndim 2 -ndw 1 2 -out fes_12.dat | fes.popt -cp2k -mathlab -file META-META-1.restart -ndim 2 -ndw 1 2 -out fes_12.dat | ||
| </ | </ | ||
| + | |||
| + | {{ : | ||
| + | ===== References ===== | ||
| + | |||
| + | ~~REFNOTES~~ | ||
exercises/2015_cecam_tutorial/neb.1440075402.txt.gz · Last modified: 2020/08/21 10:14 (external edit)