CP2K Open Source Molecular Dynamics

User Tools

Site Tools

Trace:
exercises:2015_cecam_tutorial:mtd1

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
exercises:2015_cecam_tutorial:mtd1 [2015/08/19 12:56] – fix file links tmuellerexercises:2015_cecam_tutorial:mtd1 [2020/08/21 10:15] (current) – external edit 127.0.0.1
Line 3: Line 3:
 Problem: Dissociation reaction of nitric acid on graphene and atomic rearrangements of a <chem>Si6H8</chem> cluster described using coordination numbers Problem: Dissociation reaction of nitric acid on graphene and atomic rearrangements of a <chem>Si6H8</chem> cluster described using coordination numbers
  
-======= Introduction =======+  * Original author: Marcella Iannuzzi 
 +  * Complete source and output files: [[http://cp2k.org/static/exercises/2015_cecam_tutorial/MTD1.tar.xz|MTD1.tar.xz]] 
 + 
 +===== Introduction =====
  
 For this tutorial some input and output files are given in order to present a complete procedure to solve the given problem. Some hints are also given to help in the analysis of the results. In order to be able to run these examples, some paths need to be correctly set in the input files (i.e. set the variables ''LIBPATH'', ''XYZPATH'', ''RUNPATH''). The coordinates are always read from xyz files. All the coordinate files needed for these exercises are collected in ''XYZ'', whereas ''LIB_TOOLS'' contains the PP, basis sets and DFTB parameter file. For this tutorial some input and output files are given in order to present a complete procedure to solve the given problem. Some hints are also given to help in the analysis of the results. In order to be able to run these examples, some paths need to be correctly set in the input files (i.e. set the variables ''LIBPATH'', ''XYZPATH'', ''RUNPATH''). The coordinates are always read from xyz files. All the coordinate files needed for these exercises are collected in ''XYZ'', whereas ''LIB_TOOLS'' contains the PP, basis sets and DFTB parameter file.
Line 16: Line 19:
   * Metadynamics simulation aimed at observing atomic rearrangements of the cluster bay changing the coordination of both Si and H species.                                                                                                               * Metadynamics simulation aimed at observing atomic rearrangements of the cluster bay changing the coordination of both Si and H species.                                                                                                            
                                                                                                                                                                                                                                                                      
-======= First task: dynamics of two HNO3 molecules over a graphene sheet =======                                                 +===== First task: dynamics of two HNO3 molecules over a graphene sheet =====                                                 
                                                                                                                                                                                                                                                                      
 The examples on this system are in the directory ''GR_2HNO3''. The goal is to simulate the dissociation of the <chem>HNO3</chem> molecules with formation of products like <chem>H2O</chem> and/or <chem>NO</chem> or <chem>NO2</chem> fragments. These reaction can occur in gas phase. The examples on this system are in the directory ''GR_2HNO3''. The goal is to simulate the dissociation of the <chem>HNO3</chem> molecules with formation of products like <chem>H2O</chem> and/or <chem>NO</chem> or <chem>NO2</chem> fragments. These reaction can occur in gas phase.
Line 194: Line 197:
 By plotting the CV as recorded along the short MD trajectory (3 ps), the amplitude of the equilibrium fluctuations can be evaluated and then used to set up the size of the Gaussian hills that build up the biasing potential. The first CV fluctuates close to zero, with fluctuations smaller than 0.2. The second is around 2.8. The fluctuations are smaller due to the stiffness of the three NO bonds. The coordination of H to C is also typically zero, but it can change a lot when the molecules approach the layer, even if there is no dissociation of H and no binding to C. This indicates that this variable is difficult to control and might turn out to be tricky to use it to distinguish among different states of the reaction process. The point to plane distance shows quite large fluctuations and it is clearly not suited to distinguish a specific state along the reaction path. Moreover, its minima, when the two molecules are closer to the layer, correspond to the maxima of the third CV, i.e. the CN of H to C. At least before dissociation, the information that this variable provide is redundant. It might be interesting to run again this preliminary simulation after modifying the definition of the CV. For example, by changing the two exponents or even the reference distance of the CN, the range of the function can be made shorter or longer. It is maybe important to remind that the function defining the CV must have a gradient different from zero to affect the behavior of the system in a MTD run. Namely, the MTD force term affecting the dynamics of the atoms involved in the definition of the CV is proportional to the gradient of the CV function. By plotting the CV as recorded along the short MD trajectory (3 ps), the amplitude of the equilibrium fluctuations can be evaluated and then used to set up the size of the Gaussian hills that build up the biasing potential. The first CV fluctuates close to zero, with fluctuations smaller than 0.2. The second is around 2.8. The fluctuations are smaller due to the stiffness of the three NO bonds. The coordination of H to C is also typically zero, but it can change a lot when the molecules approach the layer, even if there is no dissociation of H and no binding to C. This indicates that this variable is difficult to control and might turn out to be tricky to use it to distinguish among different states of the reaction process. The point to plane distance shows quite large fluctuations and it is clearly not suited to distinguish a specific state along the reaction path. Moreover, its minima, when the two molecules are closer to the layer, correspond to the maxima of the third CV, i.e. the CN of H to C. At least before dissociation, the information that this variable provide is redundant. It might be interesting to run again this preliminary simulation after modifying the definition of the CV. For example, by changing the two exponents or even the reference distance of the CN, the range of the function can be made shorter or longer. It is maybe important to remind that the function defining the CV must have a gradient different from zero to affect the behavior of the system in a MTD run. Namely, the MTD force term affecting the dynamics of the atoms involved in the definition of the CV is proportional to the gradient of the CV function.
  
-======= Second task: Metadynamics of the dissociation of HNO3 over a graphene sheet =======+===== Second task: Metadynamics of the dissociation of HNO3 over a graphene sheet =====
  
 The presented MTD run employs as CV only the three CN described above. The related input file is ''[[http://cp2k.org/static/exercises/2015_cecam_tutorial/MTD1/GR_2HNO3/gr2hno3_mtd_3cv_p1.inp|GR_2HNO3/gr2hno3_mtd_3cv_p1.inp]]'' and the output is stored in ''DFTB_MTD_3CV''. The ''[[inp>MOTION/FREE_ENERGY/METADYN]]'' input section has been modified to activate the MTD algorithm. The presented MTD run employs as CV only the three CN described above. The related input file is ''[[http://cp2k.org/static/exercises/2015_cecam_tutorial/MTD1/GR_2HNO3/gr2hno3_mtd_3cv_p1.inp|GR_2HNO3/gr2hno3_mtd_3cv_p1.inp]]'' and the output is stored in ''DFTB_MTD_3CV''. The ''[[inp>MOTION/FREE_ENERGY/METADYN]]'' input section has been modified to activate the MTD algorithm.
Line 243: Line 246:
 Other quantities that can be monitored from the ''[[inp>MOTION/FREE_ENERGY/METADYN/PRINT/COLVAR]]'' output, beside the instantaneous values of the three CVs (2nd, 3rd, and 4th col.) are : the instantaneous gradient of the bias potential computed with respect to CV (5th, 6th, 7th col.), the gradients wit respect to the CVs of wall potentials, if present, (8th,9th,10th col), the instantaneous value of the bias potential, and the instantaneous values of the wall potentials. Other quantities that can be monitored from the ''[[inp>MOTION/FREE_ENERGY/METADYN/PRINT/COLVAR]]'' output, beside the instantaneous values of the three CVs (2nd, 3rd, and 4th col.) are : the instantaneous gradient of the bias potential computed with respect to CV (5th, 6th, 7th col.), the gradients wit respect to the CVs of wall potentials, if present, (8th,9th,10th col), the instantaneous value of the bias potential, and the instantaneous values of the wall potentials.
  
-======= Third task: dynamics of Si6H8 =======+===== Third task: dynamics of Si6H8 =====
  
 The data file for this example are in ''SI6_CLU''. In this case, a small Si cluster of 6 Si atoms saturated by 8 H atoms is studied. Si clusters show different arrangements. The equilibrium structure should be such that Si atoms keep the preferred tetrahedral coordination. In the presence of H saturating the dangling Si bonds, the structure can be open, like the chair structure that is used here as starting conformation. By loosing H atoms, through the formation of molecular hydrogen, the cluster undergoes some rearrangement. The structure should become more compact in order to saturate the Si coordination shell. The data file for this example are in ''SI6_CLU''. In this case, a small Si cluster of 6 Si atoms saturated by 8 H atoms is studied. Si clusters show different arrangements. The equilibrium structure should be such that Si atoms keep the preferred tetrahedral coordination. In the presence of H saturating the dangling Si bonds, the structure can be open, like the chair structure that is used here as starting conformation. By loosing H atoms, through the formation of molecular hydrogen, the cluster undergoes some rearrangement. The structure should become more compact in order to saturate the Si coordination shell.
Line 296: Line 299:
 The dynamics of the CVs along the two simulations, with and without Lagrangian MTD scheme, is equivalent. The CV and the auxiliary variables, in the Lagrangian MTD simulation, closely follow each other, which points to a strong enough coupling (could be also a bei looser). The masses assigned to the auxiliary variables seem not to affect significantly the time evolution at equilibrium, i.e. the inertia effect is quite small, also because the temperature of the auxiliary variables is also set at 300 K. In order to slowdown the oscillations of the three CN, and sample better the accessible configurations at each point of the in the CV space, the parameters to be tuned are then the mass and the temperature of the auxiliary variables. In the input ''[[http://cp2k.org/static/exercises/2015_cecam_tutorial/MTD1/SI6_CLU/si6_clu_mtd_l_h0_p2.inp|SI6_CLU/si6_clu_mtd_l_h0_p2.inp]]'', the definition of the Si-Si CN has been slightly changed, and the mass of the auxiliary variables has been increased. The effect f these changes can be investigated by running this input and comparing the results with the previous results (both the input can be run at a lower level of theory, just to explore the effects of the different parameters on the dynamics of the CV). A third input is proposed, where the MTD temperature is reduced to 100 K, ''[[http://cp2k.org/static/exercises/2015_cecam_tutorial/MTD1/SI6_CLU/si6_clu_mtd_l_h0_p3.inp|SI6_CLU/si6_clu_mtd_l_h0_p3.inp]]''. The dynamics of the CVs along the two simulations, with and without Lagrangian MTD scheme, is equivalent. The CV and the auxiliary variables, in the Lagrangian MTD simulation, closely follow each other, which points to a strong enough coupling (could be also a bei looser). The masses assigned to the auxiliary variables seem not to affect significantly the time evolution at equilibrium, i.e. the inertia effect is quite small, also because the temperature of the auxiliary variables is also set at 300 K. In order to slowdown the oscillations of the three CN, and sample better the accessible configurations at each point of the in the CV space, the parameters to be tuned are then the mass and the temperature of the auxiliary variables. In the input ''[[http://cp2k.org/static/exercises/2015_cecam_tutorial/MTD1/SI6_CLU/si6_clu_mtd_l_h0_p2.inp|SI6_CLU/si6_clu_mtd_l_h0_p2.inp]]'', the definition of the Si-Si CN has been slightly changed, and the mass of the auxiliary variables has been increased. The effect f these changes can be investigated by running this input and comparing the results with the previous results (both the input can be run at a lower level of theory, just to explore the effects of the different parameters on the dynamics of the CV). A third input is proposed, where the MTD temperature is reduced to 100 K, ''[[http://cp2k.org/static/exercises/2015_cecam_tutorial/MTD1/SI6_CLU/si6_clu_mtd_l_h0_p3.inp|SI6_CLU/si6_clu_mtd_l_h0_p3.inp]]''.
  
-======= Fourth task: Lagrangian MTD of the atomic rearrangement of Si6H8 =======+===== Fourth task: Lagrangian MTD of the atomic rearrangement of Si6H8 =====
  
 ''MTD_L_P2'' contains the output of the MTD run performed with the parameters tested by running ''[[http://cp2k.org/static/exercises/2015_cecam_tutorial/MTD1/SI6_CLU/si6_clu_mtd_l_h0_p2.inp|SI6_CLU/si6_clu_mtd_l_h0_p2.inp]]''. The corresponding input is ''[[http://cp2k.org/static/exercises/2015_cecam_tutorial/MTD1/SI6_CLU/si6_clu_mtd_l_p2.inp|SI6_CLU/si6_clu_mtd_l_p2.inp]]''. The ''[[inp>MOTION/FREE_ENERGY/METADYN/METAVAR#SCALE]]'' parameter is the same for the three variables, since the fluctuations of the three CN are going to be quite similar. The collocation rate is every 100 MD steps, which is quite often, but reasonable, also because the hill size is not too big (about 1 Kcal/mol for the height). ''MTD_L_P2'' contains the output of the MTD run performed with the parameters tested by running ''[[http://cp2k.org/static/exercises/2015_cecam_tutorial/MTD1/SI6_CLU/si6_clu_mtd_l_h0_p2.inp|SI6_CLU/si6_clu_mtd_l_h0_p2.inp]]''. The corresponding input is ''[[http://cp2k.org/static/exercises/2015_cecam_tutorial/MTD1/SI6_CLU/si6_clu_mtd_l_p2.inp|SI6_CLU/si6_clu_mtd_l_p2.inp]]''. The ''[[inp>MOTION/FREE_ENERGY/METADYN/METAVAR#SCALE]]'' parameter is the same for the three variables, since the fluctuations of the three CN are going to be quite similar. The collocation rate is every 100 MD steps, which is quite often, but reasonable, also because the hill size is not too big (about 1 Kcal/mol for the height).
exercises/2015_cecam_tutorial/mtd1.1439988977.txt.gz · Last modified: 2020/08/21 10:14 (external edit)