Differences

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

--- probe_particle_model [2017/01/26 23:00]
krejcio
+++ probe_particle_model [2017/01/26 23:01]
@@ Line 1: / Line 1: @@
-===== Introduction =====
-Probe particle model code is supposed to simulate AFM ( and to some extent STM and inelastic-STM) images obtained with tip modified by an atom or small molecule (such as Xe, CO, CH4, H2). In all cases the actual particle (atom or molecule) is replaced by a spherical model particle and classical potential (Lenard-Jones) is used for description of Pauli repulsion and Van der Waals attraction.
-New code is written in C/Python and can operate in framework of Lennard-Jones forces as well as electrostatic forces, if necessary.
-{{:ptcda_df.png|}}
-===== Older Fortran Version =====
-[[Older documentation]]
-===== New Version =====
-The documentation is for "Master" branch, that can be downloaded from [[http://github.com/ProkopHapala/ProbeParticleModel/]] or by running:
-  git clone http://github.com/ProkopHapala/ProbeParticleModel/
-in your terminal.
-It should be working for developers "dev" branch, too.
-It can be downloaded from [[http://github.com/ProkopHapala/ProbeParticleModel/tree/dev]] or simply from terminal by running this commands:
-  git clone http://github.com/ProkopHapala/ProbeParticleModel/
-  cd ProbeParticleModel
-  git checkout dev
-===== Software requirements =====
-python-2.7; python-2.7-numpy(-1.8); gcc(-4.8); python-2.7-matplotlib(-1.3)  - for plotting of figures;
-== Metacentrum ==
-  module add python27-modules-intel
-  module add scipy-0.17.1-py2.7.10
-And don't forget to add:
-  import matplotlib as mpl;  mpl.use('Agg');
-on the 3rd line of __plot_results.py__, if it is not allready in the script and if you want to use it on a cluster. This makes it run without Xserver (e.g. on supercomputer) # see http://stackoverflow.com/questions/4931376/generating-matplotlib-graphs-without-a-running-x-server
-===== Hartree potential creation =====
-[[DFT inputs]]
-===== Inputs =====
-== Geometry ==
-The geometry of the sample can be read from *.bas or *.xyz files, which has to have **xyz extension**! Here we show part of the geometry file from example attached to the model:
-  C       0.165   0.165   0.165   -0.0132352941
-  C       2.2987  1.39372 0.16643 -0.0132352941
-  C       4.4293  2.62743 0.16468 -0.0132352941
-  C       6.56208 3.85683 0.16431 -0.0132352941
-  C       8.69719 5.08888 0.16545 -0.0132352941
-  ...
-Which is in format:
-Number of atoms
-optional line
-Symbol/or/Z-of-element x y z charge-optional
-The Lennard-Jones (L-J) potential needed for calculations can be created by running:
-  python PATH_TO_YOUR_PROBE_PARTICLE_MODEL/generateLJFF.py -i YOUR_INPUT_FILE.xyz
-If charges are also in the input file add "-q" flag to create electrostatic field.
-Optionally, the geometry can be also read from *.xsf or *.cube file, too. The command for creation of L-J force field then looks like:
-  python PATH_TO_YOUR_PROBE_PARTICLE_MODEL/generateLJFF.py -i YOUR_INPUT_FILE.xsf
-or
-  python PATH_TO_YOUR_PROBE_PARTICLE_MODEL/generateLJFF.py -i YOUR_INPUT_FILE.cube
-x, y and z components of L-J forces are stored in __LJFF_x.xsf__, __LJFF_y.xsf__ and __LJFF_z.xsf__ files, respectively. These files that can be viewed e.g. via XCrySDen (http://www.xcrysden.org/) or VESTA (http://jp-minerals.org/vesta/en/).
-== Hartree potential ==
-If an electrostatic Hartree potential is obtained from some DFT calculations, it can be read *.xsf or *.cube files. The electrostatic force field is created by running:
-  python PATH_TO_YOUR_PROBE_PARTICLE_MODEL/generateLJFF.py -i YOUR_INPUT_FILE
-If default parameters are used, than you have monopole represented by an Gaussian cloud of charge with its FWHM of 0.7 Ǎ. The monopole can be changed to non-tilting dipoles or quadrupoles by adding flag: -t type, where type ∈ {s,px,py,pz,dx2,dy2,dz2,dxy,dxz,dyz}; s stands for monopole (default), p for dipoles, d for quadrupoles. The FWHM of the Gaussian cloud can be changed by adding flag: -s FWHM.
-== params.ini ==
-This files contains all important information about the scan and informations for creation of important forcefields. Here we show an example of it:
-  probeType       8                               # atom type of ProbeParticle (to choose L-J potential ),e.g. 8 for CO, 54 for Xe
-  charge          0.0                             # effective charge of probe particle [e]
-  stiffness       0.20 0.20 20.00                 # [N/m] harmonic spring potential (x,y,R) components, x,y is bending stiffnes, R particle-tip bond-length stiffnes,
-  r0Probe         0.0 0.0  4.00                   # [Å] equilibirum position of probe particle (x,y,R) components, R is bond length, x,y introduce tip asymmetry
-  PBC             True                            # Periodic boundary conditions ? [ True/False ]
-  gridN           240 240 200                     # Grid division around each cell axis; Not necessary - if it is not here a 0.1 division is applied
-  gridA   23.10994667   0.000000      0.000000    # lattice vector of the cell; If geometry is read from .xsf or .cube
-  gridB   11.55497333  20.01380089    0.000000    # !!!! If geometry is read from *.xyz, but electrostatics from .xsf or .cube than gridN and gridA/B/C has
-  gridC    0.000000     0.000000    20.000000     # to be in agreement with the harteee potential. Also the shift vector in the cube file has to be 0.0 0.0 0.0 !!!!
-  scanMin     00.0   00.0    10.0                 # start of scanning (x,y,z) (z should be something like: the highest atom of the sample + R + 2.5)
-  scanMax     30.0   22.0    15.0                 # end of scanning (x,y,z)
-  scanStep    0.1   0.1    0.10                   # steps of scan (dx, dy, dz)
-  Amplitude       1.0                             # [Å] oscillation amplitude for conversion Fz->df
-If you want to make a scan for different probe, you have to change the probeType in __params.ini__ and to recompute L-J forces.
-**The number of grid divisions in *.xsf files is enlarged by one in each direction. Therefore, gridN have to be numbers of cubicles in *.xsf file reduced by one.**
-===== Simulating AFM =====
-With having L-J and electrostatic forces made by generating scripts, you can try to run an AFM scan via:
-  python PATH_TO_YOUR_PROBE_PARTICLE_MODEL/relaxed_scan.py
-which run the scan with charge Q and lateral stiffness K written in __params.ini__. Please note, that electrostatic forces are necessary only if Q ≠ 0.0.  The result - Fz force acting on the tip  -- is saved in __Q?.??K?.??__ directory as an __OutFz.xsf__ file.
-The df results for constant height scans can be plotted by:
-  python PATH_TO_YOUR_PROBE_PARTICLE_MODEL/plot_results.py  --df
-that plot results for oscillation written in __params.ini__. The results - __df_???.png__ maps for each tip height - are in directory: __Q?.??K?.??/Amp?.??__
-optionally WSxM files for each height can be written by putting flag - -WSxM behind the plotting command. Outputs are __df_???.xyz__ files with WSxM head and x y df data. If - -save_df flag is applied the df data are stored in __df.xsf__ file.
-Flag - -atoms used simultaneously with - -df puts positions of atoms of the sample saved in __input_plot.xyz__ into __df_???.png__. The flag - -bonds ads into the maps also lines connecting close-by atoms.
-If a flag - -pos is applied for both commands (scanning & plotting) than xy positions of the relaxing Probe Particle (PP) are shown in __xy_???.png__ as a red dots, while the gray scale on the background maps represent z position of the PP (brighter - higher).
-===== Tests =====
-Examples of df simulation is already in the downloaded/cloned repository in folder __examples/__. You should try to run these examples before going to your own stuff, in order to see that the code is working on your machine.
-===== Scans with different charge (Q), lateral stiffness (K) or oscillation amplitude (A) =====
-A scan with different charge (Q) and/or lateral stiffness (K) than those written in __params.ini__ can be calculated via running (be aware, that for Q ≠ 0.0, you need to have precalculated electrostatic forces):
-  python PATH_TO_YOUR_PROBE_PARTICLE_MODEL/relaxed_scan.py -q (Q) -k (K)
-It will create a new folder __Q?.??K?.??/__ where will be stored the results from the new run. A df map for wanted oscillation amplitude (A) can be then created by:
-  python PATH_TO_YOUR_PROBE_PARTICLE_MODEL/plot_results.py --df -q (Q) -k (K) -a (A)
-The results will be saved in subfolder __Amp?.??/__.
-Also scans over ranges of (Q),(K) and (A) can be done, via:
-  python PATH_TO_YOUR_PROBE_PARTICLE_MODEL/relaxed_scan.py --krange min max nK  --qrange min max nQ
-  python PATH_TO_YOUR_PROBE_PARTICLE_MODEL/plot_results.py --df --krange min max nK  --qrange min max nQ --arange min max nA
-===== References =====
-Prokop Hapala, Georgy Kichin, Christian Wagner, F. Stefan Tautz, Ruslan Temirov, and Pavel Jelínek, Mechanism of high-resolution STM/AFM imaging with functionalized tips, Phys. Rev. B 90, 085421 – http://journals.aps.org/prb/abstract/10.1103/PhysRevB.90.085421
-Prokop Hapala, Ruslan Temirov, F. Stefan Tautz, and Pavel Jelínek, Origin of High-Resolution IETS-STM Images of Organic Molecules with Functionalized Tips, Phys. Rev. Lett. 113, 226101 – http://journals.aps.org/prl/abstract/10.1103/PhysRevLett.113.226101

Nanosurf Lab

User Tools

Site Tools

Differences

Page Tools