SCM: ReaxFF Manual
Contents
1. Wikipedia also has some pages of interest e http en wikipedia org wiki Monte_Carlo_method e http en wikipedia org wiki Grand_canonical_ensemble It is important to note that this method heavily relies on random numbers and simulations are thus non repeatable in detail but should converge to the same answer About the Reaxff GCMC code The GCMC code for reaxff was originally developed by Thomas Senftle working as a Graduate Student at Penn State University under the supervision of Dr Adri van Duin The original version was a wrapper code that called an external executable to perform the reaxff minimization step and energy calculation and relied on file modification and parsing to steer the reaxff code and get the results back A rewrite of the code made by Hans van Schoot SCM in close collaboration with Thomas Senftle is now available in the ADF package The rewrite directly integrates into the ADF ReaxFF code solving performance issues of the original code by removing the calling overhead of the reaxff executable and the relatively slow file management It also merged several modifications of the original code to support the usage of whole molecules for Monte Carlo moves and supports the usage of multiple atom molecule types during the simulation Other improvements were made on the input options the accessible volume calculation the MC acceptance prefactor calculation and the writing of logfiles The relevant papers are2. Part of this forcefield is also published in L Z Zhang S V Zybin A C T van Duin S Dasgupta W A Goddard and E M Kober Carbon Cluster Formation during Thermal Decomposition of Octahydro 1 3 5 7 tetranitro 1 3 5 7 tetrazocine and 1 3 5 Triamino 2 4 6 trinitrobenzene High Explosives from ReaxFF Reactive Molecular Dynamics Simulations Journal of Physical Chemistry A 2009 113 10619 10640 http dx doi org 10 1021 3p901353a The parameters of the nitramine ReaxFF are based on a large number of ab initio QM calculations Over 40 reactions and over 1600 equilibrated molecules have been used they are designed to characterize the atomic interactions under various environments likely and unlikely high energy each atom can encounter The training set contains bond breaking and compression curves for all possible bonds angle and torsion bending data for all possible cases as well as crystal data Please see the supplimental material from Phys Rev Lett 2003 91 098301 for a detailed description of the parameterization of this force field Modeling the sorption dynamics of NaH using a reactive force field Journal of Chemical Physics 2008 128 164714 http dx doi org 10 1063 1 2908737 NiCH ff Ni C H O N S F Pt C1l JE Mueller A C T van Duin and W A Goddard III Development and Validation of ReaxFF Reactive Force Field for Hydrocarbon Chemistry Catalyzed by Nickel J Phys Chem C 2010 114 4939 4949 http dx doi or
3. The MCFF Optimizer uses a Monte Carlo approach to finding the best fit force field for the given training set It is based on the paper by E lype et al In the following sections the input and output of the MCFF Optimizer are described Input files In order to optimize a ReaxFF forcefield the files listed below must be present in the directory where the reaxff program is executed e jopt file containing a single text line with 4 on it written in the i3 format This will instruct reaxff to perform a Monte Carlo force field optimization e ffield the initial force field file e ffield_min minimum values for each force field value this file has the same format as ffield e ffield_max maximum values for each force field value this file has the same format as field e ffield_bool contains 1 0 or 0 0 as a flag whether the corresponding value is to be optimized or not this file has the same format as ffield e trainset in file with test values the same as in the original reaxff force field optimization see page 27 of the ReaxFF User Manual e geo file with test geometries in the BGF format the same as in the original reaxff force field optimization Geometries of different molecules in this file must be concatenated control in addition to general ReaxFF control parameters it also contains the MCFFOptimizer related ones explained below The mcffopt_water example in the examples reaxff directory demo
4. Scientific Computing amp Modelling ReaxFF Manual ADF Program System Release 2014 Scientific Computing amp Modelling NV Vrije Universiteit Theoretical Chemistry De Boelelaan 1083 1081 HV Amsterdam The Netherlands WWW www scm com E mail support scm com Copyright 1993 2014 SCM Vrije Universiteit Theoretical Chemistry Amsterdam The Netherlands All rights reserved Table of Contents RReax EF Marital css doses isecsexees sus nara a AEEA EA rir rA Eaa E Ea RA ES 1 IEEE o el EE 2 Preface inda A aA ba eva dew aca taps A a a 3 MCFFOptimizer Monte Carlo force field parameter optimizer ssssssssssenusrunserunsennnnennrnunnnnnnnnnnnnnnnnnnnnennn 4 liput OI 4 Control parameters ccececeseneeeeeeeeeeceeeeseneeeeeeeeseneesneeseseeaeeseeeeseaeseeneeesseneesnseeseeeeeseseeeesesneneneesesaneeens 4 Reese emsoemesuepegegessessteugeesesgegeeseegegeeguguEEENEE EENEG SEENEN SEENEN CERS ENEE 5 nu EE 5 Grand Canonical Monte Carlo GOMC icicccccccicccscccceteccecccatdaecececattacecocceretxeteccenesceeesecneceedaiesencasuedaateiueietecceds 7 General d giiesiasa cen taean ced tales ca sced twa ceduact iuaiea vaca aaa aeaa aeaa aauadness aoucd toast anaes Raa aR E 7 Lal o10 Sep eercrerre rrr correc a a er Peer Tere Pree Ter PreTrCePTTELecrer re a 7 OU oe aa E T A E 11 EE 13 included FORCOT E 15 Preface This manual describes additions and modifications by SCM MCFFOptimizer Monte Carlo force field parameter optimizer
5. total cell volume occupied volume 4 vacc needed volume specified accessible volum vacc occupied volume 435 0 vacc lif ivol 2 or ivol 4 specify Vacc in angrstoms 3 0 0 vvacu if ivol 0 specify non accessible vacuum volume Vvacu in angrstoms 3 0 25 ivlim volume change limit value between between 0 and 1 Vnew 1 ivlim V1 J resopt write restart files 0 no l yes 1 resfrq frequency of writing restart files MC code only writes files if the MC move is accepted 0 debug print debug output if set to 1 print even more debug output when set to 2 5 nmols Number of MC molecule types must match the number of molecule blocks l Molecule Speci This part i that follow fic Data C2H2 example s fixed format We need cmpot on line 1 possibly followed by the noinsr on line two and forced to be ended with nmatom on line 2 or 3 nmatom lines of coordinates the coordinates are FIXED FORMAT 2chars symbol after decimal 1 space 75 00 cmpot 4 nmatom 12 180480000000000 0 E 13 124731000000000 0 G 11 349475000000000 0 H 13 957314000000000 0 H Molecule Specific Data 75 00 cmpot chemical potential of molecule 6 nmatom number of atoms in molecule 13 989222000000000 0 405391000000000 13 316784000000000 0 399646000000000 i 11 494513000000000 0 461837000000000 11 335219000000000 0 353577000000000 i 13 811701000000000 0 340224000000000 13 97056
6. Phys Chem A 2012 116 49 pp 12163 12174 http dx doi org 10 1021 jp308507x Not all cross terms between the two forcefield files are defined which might cause problems if the system has for example C Cu interactions AuSCH_2011 ff Au S C H T T Jarvi A C T van Duin K Nordlund and W A Goddard III Development of Interatomic ReaxFF Potentials for Au S C H Systems J Phys Chem A 115 10315 10322 2011 http dx doi org 10 1021 4p201496x AuSCH_2013 f Au S C H Gyun Tack Bae and Christine M Aikens Improved ReaxFF Force Field Parameters for Au S C H Systems Journal of Physical Chemistry A 2013 117 40 10438 10446 http dx doi org 10 1021 3jp405992m Based upon T T Jarvi A C T van Duin K Nordlund and W A Goddard Development of interatomic ReaxFF potentials for Au S C H systems Journal of Physical Chemistry C 115 2011 10315 10322 yields improvements for bond bending potential energy surfaces aimes to agree with DFT geometries of small clusters and gold thiolate nanoparticles PDMSDecomp ff C H O Si K Chenoweth S Cheung A C T van Duin W A Goddard III and E M Kober Simulations on the Thermal Decomposition of a Poly dimethylsiloxane Polymer Using the ReaxFF Reactive Force Field J Am Chem Soc 2005 127 19 pp 7192 7202 http dx doi org 10 1021 ja050980t Specialized forcefield designed to investigate the failure of the poly dimethylsiloxane polymer PDMS at high temp
7. columns followed by a couple of spaces at least 1 followed by the 6 character keyword The order is free except for the nmols keyword which should be the last one The nmols keyword signals the parser that the next section of control_MC will define X new MC Molecule Types This is an example for the control_MC file GCMC control file example 0 iensmb select MC ensemble 0 Mu NVT with fixed volume 1 Mu NPT with variable volume 5000 niter number of MC iterations to do this simulation 0 nstart start the iteration counter with an offset usefull for restarts to avoid double files 300 0 mctemp Temperature of the simulation affects acceptance rate for steps that increase the energy 00 mcpres NPT pressure in GPa set to zero for incompressible solid systems unless at very high pressures 3 0 rmaxpl Max radius for atom placement on insert displace move Te rminpl Min radius for atom placement on insert displace move 2000 nmctry Maximum number of trials allowed when inserting or moving a molecule If the l rmaxpl and rminpl variables are very strict this number needs to be large T igcfac include GC prefactor in probabilities 0 no 1 yes 0 ivol select MC volume calculation technique 0 vvacu needed volume total volume occupied volume specified vacuum volume vvacu UP ols volume total cell volume 2 vacc needed volume specified accessible volum vacc ne e volume
8. the free energy and E_diss is the dissociation energy of the O2 molecule The no insert noinsr keyword The noinsr keyword tells the GCMC code to keep the number of molecules atoms of this type fixed It will thus disable Insert Delete moves on this type meaning it can only do a displacement move or volume change move if the iensmb keyword is set to 1 the control file The control file is a regular reaxff control file and it influences the minimization step after an MC trial move Because of this only a small number of the reaxff keywords are used during the GCMC simulation An example of the control file some of the parameters that influence the minimization step in the GCMC code 1 icentr Put the center of mass at the center of the cube 1 igeofo O xyz input geometry 1 Biograf input geometry 2 xmol input geometry 2 50000 endmm End point criterium for MM energy minimisation 500 imaxit aximum number of iterations 0 icelop Optimize cell parameters O no Lues 1 00050 celopt Cell parameter change 0 imaxmo In this case 0 POLAK RIBIERE Conj Grad method 1 Limited memory BFGS method The code has been tested with various imaxit and endmm values the other options have not been fully tested Other reaxff keywords might also influence the minimization procedure but those are best left to their default settings the ffield file 10 The ffield file should be an normal reaxff forcefield file as des
9. 1000000000 0 453325000000000 Molecule Specific Data H20 Example chemical potential of molecule number of atoms in molecule 421696000000000 376902000000000 459560000000000 335843000000000 C2H4 example 24d 15 1x A2 X Vr Z followed by 15 24 wide 316689000000000 568360000000000 988208000000000 101000000000000 000150000000000 885795000000000 970612000000000 129581000000000 084000000000000 756236000000000 75 00 cmpot chemical potential of molecule 1 noinsr setting this to 1 disables insert deletion moves If it is set to 1 for all types th nsamble becomes NVT NPT 3 nmatom number of atoms in molecule 39 996720000000000 40 747660000000000 40 512210000000000 H 40 000210000000000 39 999520000000000 39 934730000000000 O 40 000030000000000 39 259880000000000 40 523700000000000 H Molecule Specific Data H2 Example 75 00 cmpot chemical potential of molecule 2 nmatom number of atoms in molecule 5 025812000000000 0 0000000000000000 0 000000000000000 H 5 774188000000000 0 0000000000000000 0 000000000000000 H Singl atom Molecule Specific Data 75 00 cmpot 1 nmatom xampl chemical potential of molecule number of atoms in molecule 0 000000000000000 0 0000000000000000 0 000000000000000 H The Molecule Specific Data blocks define the molecules or atoms that can be inserted moved deleted with the MC code The atoms named here should of c
10. An example of the Elog file Iteration Naccepted Volume MC Energy RxFFEnergy 0 1 15625 00 3098 88 3179 88 2 2 15625 00 3107 92 3269 92 12 3 3 15625 00 3 13013 TL 4 4 15625 00 3160 05 3484 05 6 5 15625 00 3169 77 3493 77 SE 6 15625 00 3200 13 3605 13 reaxout kf This is the binary logfile generated by the GCMC code Its contents can be viewed with the KFBrowser utility or it can be loaded into ADFMovie to view the geometries in the file Only the data of succesful MC moves is written to this file Code Details Overview The GCMC code will perform niter control_MC file option Grand Canonical Monte Carlo trial moves and accept or reject them based on the Energy produced by the ReaxFF minimization step of the trial geometry The Monte Carlo algorithm will always accept a step if it results in a decrease of the energy and accept steps that go up in energy with a probability This section will give some details about how the code works MC Moves Insert Delete Move Volume The GCMC code currently supports 4 types of MC Moves Insert Delete Move displace Volume change The first three moves always change a whole molecule of the system as defined in the control_MC file a molecule can of course contain only a single atom Every MC iteration selects one MC Molecule Type from the defined molecules in control_MC at random followed by a random MC Move unless there are no molecules of the type in the system in tha
11. Thomas P Senftle Randall J Meyer Michael J Janik and Adri C T van Duin Development of a ReaxFF potential for Pd O and application to palladium oxide formation J Chem Phys 139 044109 2013 Thomas P Senftle Adri C T van Duin Michael J Janik Determining in situ phases of a nanoparticle catalyst via grand canonical Monte Carlo simulations with the ReaxFF potential Catalysis Communications volume 52 5 July 2014 Pages 72 77 Input Overview The GCMC code in ADF ReaxFF needs the following input files to run e control_MC The GCMC control file which holds MC settings and the atoms molecules to insert move delete e control A reaxff control file in which only a small number of parameters is of interest e ffield A standard reaxff forcefield file geo A geometry file preferably in biograph format code not yet tested with xyz iopt Text file that should only contain a 5 without the quotes insertData_MC optional file Table used when restarting GCMC simulations Also the current version of the GCMC code can only run in serial so please set the NSCM environment variable to 1 type export NSCM 1 without quotes in the shell before starting a GCMC reaxff run the control_MC file Lines in the control_MC file that start with or will be ignored so those can be used for comments Empty lines are also ignored so feel free to leave some in the file Lines with keywords should have their value in the first 8
12. cribed in the reaxff documentation by A van Duin visit the documentation section on the SCM website to obtain this document the geo file The GCMC code has been tested with biograph input files but other input formats might work The details of this file are also described in the original reaxff documentation by van Duin the iopt file The iopt file is a text file with a single digit inside thatselects the execution mode of the reaxff code To run the GCMC code this file should contain a 5 without the quotes the insertData_MC file The GCMC code can insert multiple atom molecule types in a single simulation so it needs to keep track of what atom belongs to which insert This information is automatically stored and updated when insertion deletion moving of atoms or molecules during the simulation but is by default unknown of the atoms of the starting geometry The GCMC code will therefore by default not modify the atoms in the original input in the MC trial moves keep in mind that they can move around during the minimization step the insertData_MC file can be used to tell the GCMC code what atoms in the geo file belong to which molecule An example of the insertData_MC file atomNumber MCInsMolType MCInsertNmbr 30 1 ii 40 2 1 46 2 1 47 ii 2 48 1 3 This example specifies 4 molecules atoms that are modifiable by the GCMC code belonging to 2 different GCMC molecules atoms that are defined in the control_MC file The fir
13. d dissociations angle and dihedral distortions and reactions between hydrocarbons and vanadium oxide clusters In addition the training set contains charge distributions for small vanadium oxide clusters and the stabilities of condensed phase systems including V205 VO2 and V203 in addition to metallic V V0 ZnOH ff Zn O H D Raymand A C T van Duin M Baudin K Hermannson A reactive force field ReaxFF for zinc oxide Surface Science 2008 602 1020 1031 http dx doi org 10 1016 j susc 2007 12 023 updated version published by D Raymand A C T van Duin D Spangberg W A Goddard K Hermansson Water adsorption on stepped ZnO surfaces from MD simulation Surface Science 2010 604 9 10 741 752 http dx doi org 10 1016 j susc 2009 12 012 Based on QM calculations for Zn s ZnO s and Zn hydroxide clusters Zn OH 2 and O ZnOH 2 ReaxFF parameters were generated for Zn O and Zn Zn bond energies and for Zn O Zn O Zn O O Zn Zn and Zn O H valence angle energies QM calculations were performed for the four crystal polymorphs of the wurtzite zincblende rocksalt and caesium chloride structures the structures are also referred to as h ZnS c ZnS NaCl and CsCl respectively Al H20 ff Al H 0 M Russo R Li M Mench and A C T van Duin Molecular Dynamic Simulation of Aluminum Water Reactions Using the ReaxFF Reactive Force Field International Journal of Hydrogen Energy 36 2011 5828 5835 http dx doi org 10 1016 j
14. d by changing parameters in the istop file present in the calculation directory The file is read every 10 iterations The parameters are explained below StopKey replace 0 with 1 to stop the calculation e Beta the current B value corresponding to the mcbeta control parameter Command one of NONE WALK JUMPWALK MINIMIZE JUMPMINI e NONE no change in the procedure e WALK switch to Monte Carlo steps the default e MINIMIZE switch to gradient free minimization of the latest accepted force field e JUMPWALK take the best force field so far and start the Monte Carlo procedure from there e JUMPMINI take the best force field so far and minimize it ScaleFactor change the current step size corresponding to the mcstep control parameter ActiveParameterFraction set the fraction of force field parameters changed at each step mcacpf deltaBeta change the current value of the mcdbet control parameter BetaScaling change the current value of the mcbsca control parameter ScaleParSpace change the current value of the mcscps control parameter Grand Canonical Monte Carlo GCMC General info About Monte Carlo the Grand Canonical Ensamble It is best to read a bit about Monte Carlo and ensambles before working with the GCMC code Almost every book or review text on molecular simulations will do for example Frenkel D Smit B Understanding molecular simulation from algorithms to applications Academic Press 2002 672 p
15. describing angular distortions in a set of small AB related AB H3N BH3 molecules These data were used to derive the initial ReaxFF angular parameters The training set was extended with reaction barriers for key reaction steps such as H2 release from AB dimerization of H2B NH2 and reaction energies associated with H2 release from AB and with AB oxidation AuCSOH ff Au C S 0O H J A Keith D Fantauzzi T Jacob and A C T van Duin Reactive forcefield for simulating gold surfaces and nanoparticles Physical Review B 2010 81 235404 1 235404 8 http dx doi org 10 1103 PhysRevB 81 235404 The original Au Au parameters were extended by three publications Au O K Joshi A C T van Duin and T Jacob Development of a ReaxFF description of gold oxides and initial application to cold welding of partially oxidized gold surfaces Journal of Materials Chemistry 20 2010 10431 10437 http dx doi org 10 1039 COJMO01556C Au C S H T T Jarvi A C T van Duin K Nordlund and W A Goddard Development of interatomic ReaxFF potentials for Au S C H systems Journal of Physical Chemistry C 115 2011 10315 10322 http dx doi org 10 1021 jp201496x C O H S Rahaman O van Duin A C T Goddard W A III and Doren D J Development of a ReaxFF reactive force field for glycine and application to solvent effect and tautomerization Journal of Physical Chemistry B 115 2011 249 261 http dx doi org 10 1021 jp108642r Th
16. e Journal of Chemical Physics 139 044109 2013 http dx doi org 10 1063 1 4815820 used for studying Oxidation states of Pd nanoparticles surfaces and bulk configurations with a GCMC method T P Senftle M J Janik and A C T van Duin 19 A ReaxFF Investigation of Hydride Formation in Palladium Nanoclusters via Monte Carlo and Molecular Dynamics Simulations The Journal of Physical Chemistry C 2014 118 9 pp 4967 4981 http dx doi org 10 1021 jp411015a used in combination with a GCMC method Xue Qing Zhang E Iype S V Nedea A P J Jansen B M Szyja E J M Hensen and R A van Santen Site Stability on Cobalt Nanoparticles A Molecular Dynamics ReaxFF Reactive Force Field Study The Journal of Physical Chemistry C 2014 118 13 pp 6882 6886 http dx doi org 10 1021 jp500053u forcefield was generated using a recently developed Monte Carlo algorithm with simulated annealing CHONSMgPNaCuCl ff C H O N S Mg P Na Cu Cl Susanna Monti Cui Li and Vincenzo Carravetta Reactive Dynamics Simulation of Monolayer and Multilayer Adsorption of Glycine on Cu 110 J Phys Chem C 2013 117 10 pp 5221 5228 http dx doi org 10 1021 jp312828d Reactive MD force field for amino acids on copper CHOSMONiLiBFPN f C H O S Mo Ni Li B F P N Md M Islama V S Bryantsevb A C T van Duin ReaxFF Reactive Force Field Simulations on the Influence of Teflon on Electrolyte Decomposition during Li SWCNT Anode Discharge in Lithi
17. e forcefield does not include Au N parameters CHO ff C H O Hydrocarbon oxidation K Chenoweth A C T van Duin W A Goddard ReaxFF Reactive Force Field for Molecular Dynamics Simulations of Hydrocarbon Oxidation J Phys Chem A 2008 112 1040 1053 http dx doi org 10 1021 jp709896w To obtain the H C O compound data required to extend the hydrocarbon training set DFT calculations were performed on the following systems a dissociation energies for various bonds containing carbon oxygen and hydrogen The ground state structure was obtained through full geometry optimization Dissociation curves were calculated by constraining only the bond length of interest and re optimization of the remaining internal coordinates Optimization was also performed for the various angles and torsions associated with C H O interactions HCONSB ff H C O N S B M R Weismiller A C T van Duin J Lee and R A Yetter ReaxFF Reactive Force Field Development and Applications for Molecular Dynamics Simulations of Ammonia Borane Dehydrogenation and Combustion J Phys Chem A 2010 114 5485 5492 http dx doi org 10 1021 jp100136c The parameters in this forcefield were extended improved by two other publications A M Kamat A C T van Duin and A Yakovlev Molecular Dynamics Simulations of Laser Induced Incandescence of Soot Using an Extended ReaxFF Reactive Force Field Journal of Physical Chemistry A 2010 114 12561 1257 http d
18. eded if you want to restart your calculation from an accepted MC step The table contains 1 values for atoms that were in the original input and did not get a manually assigned MCInsert Molecule Type and MC Insert Number the GCMC code will not modify these atoms during the MC steps Also see the section on insertData_MC file MCstats The MCstats file is a logfile that contains the statistics of the MC simultaion The GCMC code writes a single line to it after every MC step containing the number of Tried MC moves tried Accepted MC moves accept Rejected MC moves reject Accepted Insertion Deletion Moving Volume change MC moves addAcc delAcc mvAcc volAcc Rejected Insertion Deletion Moving Volume change MC moves addRej delRej mvRej volRej An example of the MCstats file tried accept reject addAcc delAcc mvAcc volAcc addRej delRej mvRej volRej 0 1 0 1 0 0 0 0 0 0 0 1 1 0 0 0 0 0 1 0 2 2 0 0 0 0 0 2 0 3 3 0 0 0 0 1 2 0 4 4 0 0 0 1 2 0 5 5 0 0 0 1 3 0 6 6 0 0 0 2 3 0 7 7 0 0 0 3 3 0 8 8 0 0 0 3 4 0 9 9 0 0 0 4 4 0 10 1 10 1 0 0 0 4 5 0 11 2 10 2 0 0 0 4 5 0 12 3 10 3 0 0 0 4 5 0 Elog The Elog file contains the Volume and energies of the accepted MC steps The energies in this logfile are the pure ReaxFF energy of the system RxFFEnergy and the MC corrected energy which is used in determining if the step should be accepted or not see the section on calculating energies for details
19. eratures and pressures and in the presence of various additives TiOCHNC1 ff C H O N S Mg P Na Ti C1 F S Y Kim A C T van Duin and J D Kubicki Molecular dynamics simulations of the interactions between TiO02 nanoparticles and water with Na and Cl methanol and formic acid using a reactive force field Journal of Materials Research Volume Issue 03 2013 pp 513 520 http dx doi org 10 1557 jmr 2012 367 used for simulating Ti02 both rutile and anatase nanoparticles with water methanol and formic acid The force field was validated by comparing water dissociative adsorption percentage and bond length between Na O with density functional theory DFT and experimental results PtCH ff C H Pt C F Sanz Navarro P Astrand De Chen M Ronning A C T van Duin T Jacob and W A Goddard III Molecular Dynamics Simulations of the Interactions between Platinum Clusters and Carbon Platelets J Phys Chem A 112 1392 1402 2008 http dx doi org 10 1021 jp074806y BaYZrCHO ff C H O Ba Zr Y A C T van Duin B V Merinov S S Jang and W A Goddard III ReaxFF Reactive Force Field for Solid Oxide Fuel Cell Systems with Application to Oxygen Ion Transport in Yttria Stabilized Zirconia J Phys Chem A 112 3133 3140 2008 http dx doi org 10 1021 jp076775c CHONSSiPtZrNiCuCo ff C H O N S Si Pt Zr Ni Cu Co K D Nielson A C T van Duin J Oxgaard W Q Deng and W A Goddard III Development of the ReaxFF Reactive Force F
20. g 10 1021 4p9035056 16 SiOH ff Si O H J C Fogarty H M Aktulga A Y Grama A C T van Duin S A Pandit A reactive molecular dynamics simulation of the silica water interface J Chem Phys 2010 132 174704 http dx doi org 10 1063 1 3407433 This force field was trained to model the interaction of water at the SiO2 surface with specific emphasis on proton transfer reactions Updated parameters were fitted for all Si O H bond angle and torsion interactions as well in addition to the dissociation of a water molecule from a single Si OH 4 molecule and reaction energies for the polymerization of Si OH 4 SiC ff Si C O H N S D Newsome D Sengupta H Foroutan M F Russo and A C T van Duin Oxidation of Silicon Carbide by 02 and H20 A ReaxFF Reactive Molecular Dynamics Study Part I Journal of Physical Chemistry 2012 116 16111 16121 http dx doi org 10 1021 jp306391p The included forcefield is based on the Newsome reference with slightly improved parameters by van Duin et al VOCH ff V O C H K Chenoweth A C T van Duin P Persson M J Cheng J Oxgaard W A Goddard Development and Application of a ReaxFF Reactive Force Field for Oxidative Dehydrogenation on Vanadium Oxide Catalysts J Phys Chem C 2008 112 14645 14654 http dx doi org 10 1021 jp802134x The ReaxFF force field parameters have been fit to a large quantum mechanics QM training set containing over 700 structures and energetics related to bon
21. h step 1 0 A value of mcbsca lt 1 0 has an effect similar to positive mcdbet mcacpf probability to vary a variable at each step 0 2 To avoid taking very large steps only some of the variables are varied at each step selected randomly merxdd number of steps to divide the parameter range between ffield_min and ffield_max into 100 e mcstep initial max step size in units of range mcrxdd where range difference between ffield_max and ffield_min values 1 0 memxst maximum allowed value of max step size 100 e mcescps factor to scale max step size to satisfy acceptance tolerance 1 1 e mctart target acceptance rate percent 30 0 mcemart max acceptance rate percent 70 0 e memini if not 0 minimize the best force field parameter set after so many iterations 0 The optimization is performed only if the best set has changed since the previous minimization The minimization employed here is gradient free and relatively slow so it should not be used too frequently e replic number of replicas to try at each step 1 At each step replic Monte Carlo steps are done and the one with the lowest error is selected for the next iteration Relation between mcrxdd mcstep mcmxst and mcscps The allowed range for each parameter is divided into mcrxdd steps At each optimization step a number of force field parameters is changed by the Ax mcstep e Xmax Xmin merxdd where Xmax and Xmin a
22. ield for Describing Transition Metal Catalyzed Reactions with Application to the Initial Stages of the Catalytic Formation of 18 Carbon Nanotubes J Phys Chem A 109 493 499 2005 http dx doi org 10 1021 jp046244d Glycine ff C H O N O Rahaman A C T van Duin W A Goddard III and D J Doren Development of a ReaxFF Reactive Force Field for Glycine and Application to Solvent Effect and Tautomerization J Phys Chem B 115 249 261 2011 http dx doi org 10 1021 jp108642r SiONH ff C H O N Si S A D Kulkarni D G Truhlar S G Srinivasan A C T van Duin P Norman and T E Schwartzentruber Oxygen Interactions with Silica Surfaces Coupled Cluster and Density Functional Investigation and the Development of a New ReaxFF Potential J Phys Chem C 2013 117 1 pp 258 269 http dx doi org 10 1021 jp3086649 Aimed at oxygen interactions with realistic silica surfaces CHONPt ff C H O N Pt M F Russo D Bedrov S Singhai A C T van Duin Combustion of 1 5 Dinitrobiuret DNB in the Presence of Nitric Acid Using ReaxFF Molecular Dynamics Simulations J Phys Chem A 2013 117 38 pp 9216 9223 http dx doi org 10 1021 jp403511q specialized combustion forcefield for 1 5 dinitrobiuret DNB and nitric acid CHOFe ff C H O Fe C1 Si A1 Chenyu Zou A C T Van Duin Investigation of Complex Iron Surface Catalytic Chemistry Using the ReaxFF Reactive Force Field Method JOM December 2012 Volume 64 Issue 12 p
23. ijhydene 2011 02 035 CaSiAlO ff C H O Fe C1 Si Al Ca M C Pitman and A C T van Duin Dynamics of Confined Reactive Water in Smectite Clay Zeolite Composites J Am Chem Soc 2012 134 6 3042 3053 http dx doi org 10 1021 ja208894m dispersion CHONSSi lg ff C H O N S Si L Liu Y Liu S V Zybin H Sun and W A Goddard III ReaxFF lg Correction of the ReaxFF Reactive Force Field for London Dispersion with Applications to the Equations of State for Energetic Materials The Journal of Physical Chemistry A 2011 115 40 11016 11022 http dx doi org 10 1021 jp201599t This forcefield adds London dispersion correction terms to reaxFF and is optimized for the energetic materials RDX PETN TATB and NM plus graphite polyethylene solid carbon dioxide and solid N2 using the low temperature crystal structures to determine the lg correction parameters 17 CHOFeAINiCuS ff C H O Fe A1 Ni Cu S O Rahaman A C T van Duin W A Goddard III and D J Doren Development of a ReaxFF reactive force field for glycine and application to solvent effect and tautomerization Journal of Physical Chemistry B 115 2011 249 261 http dx doi org 10 1021 jp204894m C O H parameters only The Cu Fe Al Ni parameters are from Y K Shin H Kwak C Zou A V Vasenkov and A C T van Duin Development and Validation of a ReaxFF Reactive Force Field for Fe Al Ni Alloys Molecular Dynamics Study of Elastic Constants Diffusion and Segregation J
24. nstrates the use of the MCFFOptimizer Note however that this example is not physically meaningful For example many atomic force field parameters are allowed to vary in a very broad range which is the same for all elements In practice you will want to set the range for each element separately Control parameters The following control parameters are related to MCFFOptimizer The default value for each parameter is given in parentheses e mcffit number of MC iterations 10000 Since the Monte Carlo method does not have any notion of convergence the optimization is stopped after mcffit iterations e mcbeta initial MC beta parameter in the acceptance probability calculation P exp B AE 1 0 Here AE is a difference in the error function between the current and the best step so far If the current step is the best it is always accepted Otherwise the acceptance probability is calculated using the formula above and it is calculated with a random number from the 0 1 range The optimal value of beta depends on values of the error function e mcedbet simulated annealing increase the beta parameter by this value at each step 0 0 A positive mcdbet value means that the probability to take a step that increases the error function will decrease over time This has the same effect as decreasing the temperature in the classical molecular Monte Carlo method mcbsca simulated annealing divide the beta parameter by this value at eac
25. ourse be in the forcefield files and the coordinates should form a reasonable structure The MC code uses these coordinates during the insertion step by giving them a random rotation followed by a random translation to generate a random position of the molecule inside the box Currently there is no check to make sure the molecule stays inside the boundries of the box the code only checks that the rmaxpl rminpl values are satisfied If you plan on inserting large molecules make sure there is enough room in the rmaxpl value otherwise the code will stop with an error message The chemical potential cmpot keyword The cmpot keyword sets the chemical potential of the molecule or atom reservoir and is employed when calculating the Boltzmann accept reject criteria after a MC move is executed This value can be derived from first principles using statistical mechanics or equivalently it can be determined from thermochemical tables available in literature sources For example the proper chemical potential for a GCMC simulation in which single oxygen atoms are exchanged with a reservoir of O2 gas should equal 1 2 the chemical potential of O2 at the temperature and pressure of the reservoir cmpot Mu_O T P 1 2 Mu_O2 T P 1 2 Mu_ref T P_ref kT Log P Pref E_diss where the reference chemical potential Mu_ref T P_ref is the experimentally determined chemical potential of O2 at T and Pref kT Log P Pref is the pressure correction to
26. p 1426 1437 http dx doi org 10 1007 s11837 012 0463 5 only the parameters for Fe and crossterms differ from the CHOA1Si ff forcefield CHOA1Si ff C H O Fe C1 Si A1 F Castro Marcanoa A C T van Duin Comparison of thermal and catalytic cracking of l heptene from ReaxFF reactive molecular dynamics simulations Combustion and Flame Volume 160 Issue 4 April 2013 Pages 766 775 http dx doi org 10 1016 4 combustflame 2012 12 007 only the parameters for Fe and crossterms differ from the CHOFe ff forcefield CHOLi ff C H O N S Mg P Na Li D Bedrov G D Smith A C T van Duin Reactions of Singly Reduced Ethylene Carbonate in Lithium Battery Electrolytes A Molecular Dynamics Simulation Study Using the ReaxFF Journal of Physical Chemistry A 2012 116 11 pp 2978 2985 http dx doi org 10 1021 4p210345b specifically generated for simulating Lithium battery electrolytes must be used in combination with the MOLCHARGE keyword to set a charge restraint on Li and CO3 SiOA1Li ff H O Si A1 Li B Narayanan A C T van Duin B B Kappes I E Reimanis and C V Ciobanu A reactive force field for lithium aluminum silicates with applications to eucryptite phases Modelling and Simulation in Materials Science and Engineering 2012 20 015002 http dx doi org 10 1088 0965 0393 20 1 015002 T P Senftle R J Meyer M J Janik and A C T van Duin Development of a ReaxFF potential for Pd O and application to palladium oxide formation Th
27. re values of this parameter from ffield_max and ffield_min respectively When performing optimization the program keeps track of the average acceptance rate and adjusts mcstep up or down by the mcscps factor to keep the acceptance rate close to mctart If the acceptance rate is too low the step size is decreased to allow searching for a smaller parameter space The mcstep value can never be larger than mcmxst It should be noted that the value of the MC step size and thus all the parameters discussed in this section applies to all force field parameters to the same extent which means that it is very important to select the min and max parameter values very carefully The rule of thumb here is that the range should be as small as possible covering only the physically meaningful values Results Main results of the MCFFOptimizer are saved in the following files e ffield_best force field file corresponding to the lowest error value e ffield_last the most recent accepted force filed e MCFFOptimizer log summary of iterations including the error function value number of changed and bounded force field parameters cumulative number of accepted and rejected steps at each step Also the current MC parameters such as the B value and the acceptance rate are shown as well as the elapsed time in seconds e fort 99 the error function breakdown for the latest step Run time control The progress of the force field optimization can be controlle
28. rected using the following formula E MC _Corr E_reaxff_last_accept Pressure 0 1439 newV oldV 1 0 beta ninsertedMols log newVavail oldVavail where newVavail and oldVavail are calculated from the MC available volume see the section calculating volumes Calculating volumes 13 The GCMC code can calculate the available volume in a couple of different ways depending on the ivol setting in control_MC e ivol 0 volume total volume occupied volume specified vacuum volume vvacu ivol 1 volume total cell volume ivol 2 volume specified accessible volume vacc ivol 3 volume total cell volume occupied volume e ivol 4 volume specified accessible volume vacc occupied volume Where the occupied volume is calculated by summing up the volumes of the atoms in the geo file that are not specified to be part of an MC type molecule The volume of an atom is calculated using the average of the covalent atomic radius and the vd Waals radius of the atom which are found in the reaxff forcefield file ffield the vacc and vvacu options can be specified in the control_MC file to get a more accurate available volume Acceptance criteria An MC move is always accepted if the reaxff energy is lower than the corrected MC energy of the last accepted MC move or if the energy increase is small enough If the new energy is higher the code generates a random number between 0 and 1 and accepts the mo
29. st molecule in the control _MC file should thus consist of a single atom if this doesn t match the code will most likely crash It was inserted three times atom 30 47 and 48 The second molecule has two atoms and was inserted once The atoms do not have a fixed order and not all atoms have to be defined If an atom is not appointed to a certain MCInsMolType and MClInsertNmbr if will simply not be modified during the MC moves The insertData_MCXXXXxXx files generated by the restart option of the code can be directly used as valid insertData_MC files just remove the digits from the filename and replace the geo file with the corresponding geo _MCXXXXXX file Output Overview The GCMC code writes a couple of output files each described in this section It also produces a number of reaxff output files and some of these are described in the original reaxff documentation by van Duin Keep in mind that these files might not provide a complete or correct picture of the simulation as they could also contain data originating from rejected MC trial moves geo_MCXXXXXX This file is generated every X accepted MC moves and contains the current geometry of the system in biograph format X is set with the resfrq keyword in the control_MC file insertData_MCXXXXXX 11 This file contains a table of all the atoms in the system with their MC Molecule Type and MC Insert Number This data can be used to map atoms to an inserted molecule and is ne
30. t case it will do an insert move The Insert and Displace move MC Moves will generate a random rotation and position for the molecule and then check if the random positions are within the RminPI and RmaxPI boundaries this means no atom in the molecule can be closer to any atom currently in the system than RminPl and it should be within RmaxPI distance to an atom in the system If the conditions are not satisfied a new set of coordinates is generated and the code checks again This is repeated a maximum number of nmctry times before stopping with an error The volume change is controlled by the ivlim variable in control_MC The ivlim sets the volume change limit and it should be a value between between 0 and 1 The new volume will be calulated like this Vnew 1 ivlim Vold Calculating energies Because the GCMC simulation adds and deletes atoms or molecules during the runtime it cannot directly compare the ReaxFF energies for the MC acceptance criteria inserting a molecule will usually lower the total energy of the system causing the MC to always accept it and always reject a deletion To balance this out the GCMC code calculates a corrected MC energy to compare the trial reaxFF energy with consisting of the previously accepted ReaxFF energy the chemical potential cmpot in control MC for the inserted molecule or the ReaxFF energy the chemical potential for the deleted molecule The volume change energy is also cor
31. tems J Phys Chem A 2010 114 6298 6307 http dx doi org 10 1021 jp101332k The Cl parameters where published by Rahaman A C T van Duin V S Bryantsev JE Mueller S D Solares W A Goddard III and D J Doren Development of a ReaxFF Reactive Force Field for Aqueous Chloride and Copper Chloride J Phys Chem A 114 2010 3556 3568 http dx doi org 10 1021 3p9090415 The initial force field parameters for the Fe Fe parameters were taken from an earlier force field development project on bulk iron metal based on DFT calculations on antiferromagnetic BCC and FCC The ReaxFF parameters have not been published yet however the DFT data can be found in ref 31 of the above mentioned manuscript The O H parameters were taken from the ReaxFF bulk water description The Fe Fe and O H parameters were kept fixed to these initial values whereas the Fe O parameters were reoptimized against the quantum mechanical results presented in the above mentioned manuscript Detailed information on the force field parameters is given in the supporting information of the above mentioned manuscript HE ff C H O N RDX High Energy O Ojwang R Van Santen G J Kramer A C T van Duin and W A Goddard Zhang A C T van Duin S V Zybin and W A Goddard Thermal Decomposition of Hydrazines from Reactive Dynamics Using the ReaxFF Reactive Force Field Journal of Physical Chemistry B 2009 113 10770 10778 http dx doi org 10 1021 jp900194da
32. um Sulfur Batteries J Electrochem Soc 2014 volume 161 issue 8 E3009 E3014 http dx doi org 10 1149 2 005408jes forcefield for Electrochemistry in Li S batteries CHONSSiNaFZr ff C H O N S Si Na F Zr A Rahnamoun and A C T van Duin Reactive Molecular Dynamics Simulation on the Disintegration of Kapton POSS Polyimide Amorphous Silica and Teflon during Atomic Oxygen Impact Using the Reaxff Reactive Force Field Method J Phys Chem A 2014 118 15 pp 2780 2787 http dx doi org 10 1021 jp4121029 comments in the forcefield file interactions with water and Na Fogarty et al JCP 2010 with glycine C H F parameters Si F bond offdiag angle parameters Si S dummy parameters S O H parameters Yun 2012 Oct8 H F bond offdiag Jan14 2013 Joon Jan31 added Zr O H C TiClOH ff C H O N S Mg P Na Ti Cl1 F Sung Yup Kim and A C T van Duin Simulation of Titanium Metal Titanium Dioxide Etching with Chlorine and Hydrogen Chloride Gases Using the ReaxFF Reactive Force Field J Phys Chem A 2013 117 27 pp 5655 5663 http dx doi org 10 1021 3p4031943 adaptation evolution of the TiOCHNC1 ff forcefield by Kim S Y et al See also e Included Forcefields development version 20
33. ve if the random number is bigger than prob preFactor exp Beta deltaE Where the prefactor is calculated for insert and delete moves using the deBroglie wavelength of the inserted molecules the number of inserted molecules and the available MC volume of the system 14 Included Forcefields Description of ReaxFF force fields Disclaimer Using these force fields for systems they have not been explicitly trained against may produce unrealistic results Please see the full manuscripts for more detailed information The force field files used by the SCM version of ReaxFF are compatible with those used by the original ReaxFF code So if you have the force field information from somewhere else you can just use it save it in a text file and select it in ADFinput via the Other option Ap EE H O N B Ammonia Borane M R Weismiller A C T van Duin J Lee R A Yetter ReaxFF Reactive Force Field Development and Applications for Molecular Dynamics Simulations of Ammonia Borane Dehydrogenation and Combustion J Phys Chem A 2010 114 5485 5492 http dx doi org 10 1021 jp100136c QM data were generated describing the single and if relevant double and triple bond dissociation for all B N O H combinations These data were used to derive initial ReaxFF bond parameters and all calculations were performed using DFT with the B3LYP functional and the Pople 6 311G basis set The training set was then extended with QM data
34. x doi org 10 1021 jp1080302 F Castro Marcano A M Kamat M F Russo A C T van Duin and J P Mathews Combustion of an Illinois No 6 Coal Char Simulated Using an Atomistic Char Representation and the ReaxFF Reactive Force Field 15 Combustion and Flame 2012 159 23273 1285 http dx doi org 10 1016 4 combustflame 2011 10 022 The C H O parameters are the same as in the CHO forcefield with added S C S H and S O descriptions This force field was used in Castro et al Combustion and Flame 2011 The Boron and Nitrogen parameters are based on but not identical to the parameters used in Weismiller et al JPC A 2010 CuCl H20 ff Cu C1 H 0 O Rahaman A C T van Duin V S Bryantsev J E Mueller S D Solares W A Goddard III and D J Doren Development of a ReaxFF Reactive Force Field for Aqueous Chloride and Copper Chloride J Phys Chem A 114 2010 3556 3568 http dx doi org 10 1021 3p9090415 This forcefield is an extension of A C T van Duin V S Bryantsev M S Diallo W A Goddard O Rahaman D J Doren D Raymand and K Hermansson Development and validation of a ReaxFF reactive force field for Cu cation water interactions and copper metal metal oxide metal hydroxide condensed phases Journal of Physical Chemistry A 2010 114 9507 9514 http dx doi org 10 1021 jp102272z FeOCHC1 f Fe 0 C H C1 M Aryanpour A C T van Duin J D Kubicki Development of a Reactive Force Field for Iron Oxyhydroxide Sys
Download Pdf Manuals
Related Search