Skip to content
shaoyulu edited this page Jan 3, 2018 · 1 revision

Welcome to the OLED-GA_v2 wiki!

OLED-GA Program (USER’S GUIDE)

1. INTRODUCTION

This Reference Manual contains descriptions of all input, output and execution features of OLED-GA program. OLED-GA is both an evolutionary search and its computer program implementation, to perform electronic structure calculations of organic molecules to predict the high-performance multiple OLED device. Its main characteristics are:

• It contains different kinds of selection methods, such as, roulette wheel selection (RWS), the stochastic universal sampling (SUS), the linear rank selection (L_Rank), the exponential rank selection (E_Rank), the tournament selection (Tournament).

• It connects semi-empirical scientific program (MOPAC) to perform electronic structure calculations and geometric optimization for all candidates.

• The 3D structure of each potential molecule is generated by using OpenBabel

• It is written in C++

It routinely provides:

• Total energy and orbital energies.

• Select best OLED structures for both HOST and transporting layer.

• Varity of selection methods.

• Controllable GA parameters, such as, mutation rate, crossover rate, population size... etc.

• Statistical analysis of OLED score for each fragment.

2. COMPILATION

Type the command make to generate the executable “OLEDsearch.exe”. Then selected one of the script file to proceed the GA evolutionary search

./OLEDcleanrun: This script will clean the exist database and clean the previous calculated results than run an initial OLED-GA search

./OLEDrunrestart: This script will keep the exist database and moved the previous calculated results to the directory (OLED-GA/OLEDdata)

3. MOLECULAR FRAGMENT LIBRARY

OLED-GA/lib/Seed: Setup the seeds in this directory for building the molecules

OLED-GA/lib/Reg: Setup the functional fragments for the 1st and 2nd molecular segment in this directory for building the molecules

OLED-GA/lib/Cap: Setup the capping fragments to form the final molecular segment in this directory for building the molecules

4. DETAILE DESCRIPTION OF PROGRAM OPTIONS

Here follows a description of the variables that you can define in OLED-GA input file. There are two essential input files (saved in /OLED-GA/inp) need to be prepared before OLED-GA search. (ga.inp and mopac.inp)

4.1 GA evolutionary search parameters (OLED-GA/inp/ga.inp)

All the initial parameters for OLED-GA evolutionary search are divided into several parts. Complex input data structures are called blocks and are placed between “$block label” and a “$block label END”. There are three divided blocks in OLED-GA program, that are, $S-Generator, $Fitness, $GA.

$S-Generator block

Mode (integer): It specifies the type of GA scientific search. #1 is for OLED, #2 is
for Transition metal catalytic search, #3 is for chain-like molecule.

Default value: There is no default. But only #1 is work for this version.

Layers (integer): It specifies the how many molecular segments for OLED molecular generation. Ex. #3 means the structure will be generated from a seed and connect 3 molecular segments. Default value: There is a limitation for building layers. Minimum is 3 and maximum is 4.

Library (string): It defined the directory of molecular building blocks. Default value: There is no default.

Best_Restart (logical): If it set to “on” then the program will read the database and select top 1% candidates from the file “data” for generating the 1st GA pool.
Default value: There is no default.

Reference_Electrode (logical): It defined the idealized energy levels for the transport of holes and electrons across the device. Default value: There is no default.

JobType (logical): It defined the type of job script files (chose SLURM or PBS).
Default value: There is no default.

WriteFragScore (logical): If it set to “on” then the program will count the number of fragments with appearing in a setup FragScoreThreshold and write evaluate the fitness score of fragments for both transporting layer and HOST in output file.

Default value: There is no default.

FragScoreThreshold (double): It defined the minimum value for counting the number of fragments from database. Default value: There is no default.

$Fitness block

Reference (string): It defined the reference structure file for TM catalytic search. Default value: There is no default.

Def_File (string): It defined the essential criteria for TM catalytic search. Default value: There is no default.

$GA block

Population_Size (integer): It defined the size of population for each GA evolutionary process. (the population size only for even number) Default value: There is no default.

Termination (integer): It defined the maximum number of GA evolutionary process. Default value: There is no default.

CrossoverRate (double): It defined the possibility of a pair of genes to do the crossover operation. Use: The value should setup from 0 to 1. Default value: There is no default.

MutationRate (double): It defined the possibility of a gene to do the mutation operation. Use: The value should setup from 0 to 1. Default value: There is no default.

$Selection (double): Several selection methods for a GA process.

• RWS: roulette wheel selection

• SUS: stochastic universal sampling

• E_Rank: exponential rank selection

• L_Rank: linear rank selection

• B_Tournament: tournament selection

For the ranking method, it is essential to set up the rate L_Rank and base of E_Rank.

$Selection End

4.2 MOPAC calculation parameters (OLED-GA/inp/mopac.inp)

Method (string): It defined the Hamiltonian for MOPAC calculation Gnorm (double): Exit when gradient norm drops below n kcal/mol/Angstrom Unrestricted (logical): If it set “on” then use restricted Hartree-Fock methods

5. PROGRAM OUTPUT

5.1 Main output file (OLED-GA/log)

Log: This is a main output file, which contains all the information of each gene during GA evolutionary process (such as, gene string codes and their corresponding probabilities, frontier orbital energy levels and scores)

5.2 Structure Results (OLED-GA/Scratch/mm)

mm/ETL_#: The potential ETL structures for each generation will be automatically generated by using OpenBabel and saved into those directories. # is the number of generation.

mm/HTL_#: The potential HTL structures for each generation will be automatically generated by using OpenBabel and saved into those directories. # is the number of generation.

5.3 Calculated MOPAC Results (OLED-GA/Scratch/mopac)

mopac/ETL_#: The MOPAC calculated results of potential ETL structures for each generation are saved into those directories. # is the number of generation.

mopac/HTL_#: The MOPAC calculated results of potential HTL structures for each generation are saved into those directories. # is the number of generation.

5.4 Other output Data for OLED-GA

Best_ETL, Best_HTL: This output list the best candidate OLED in each GA generation process.

Bestforfar_ETL, Bestsofar_HTL: This output list the top best candidate OLED so far in each GA generation process.

OLEDdata: All the structures for both ETL and HTL with the fitness score higher than 900 will be saved into the directory (OLED-GA/).

Data: All the potential candidates for both ETL and HTL will be saved in this output file as a database.