Efficient hybrid algorithm for the dynamic creation of wormlike chains in solutions, brushes, melts and glasses

Abstract

We present an updated version of a program that had been published earlier in this journal. The program executes an algorithm for the creation and relaxation of large, dense or diluted homogeneous many particle systems made of wormlike, finite extendable, semiflexible multibead chains and – optionally – solvent particles, which repulse each other. The key feature is its efficiency, its output has been proven to serve as an excellent basis for any subsequent off-lattice computer simulation. The application allows to choose (i) the bead number density or packing fraction, temperature, chain length, system size, concentration, (ii) the interaction potentials, hence the local features such as bond length and bending rigidity of the chains, and (iii) the degree of pre-relaxation, parametrized and expressed through a minimum intermolecular distance. The monodisperse polymers are represented by chains of monomer coordinates in 3D space. During the dynamical two-step process of sample creation the initially (Monte Carlo step 1) predicted global characteristics of the molecular conformations remain as unaffected as possible (during molecular dynamics step 2) and the potential energy and the entropy production are relaxing towards their minima. The potentials, the distribution of bond lengths, the integration time step and temperature are smoothly controlled during the creation/relaxation process until they finally approach their prescribed or physical values. The quality of the algorithm is by its nature independent of concentration, system size or degree of polymerization; the CPU speed is quite independent of the latter quantity and linear in the system size. Chains tethered to a surface (dry polymer brushes) can be generated as well. New version program summaryProgram Title: GenPolProgram Files doi:http://dx.doi.org/10.17632/5kgz4bmt2h.1Licensing provisions: MITProgramming language: Fortran, PerlJournal reference of previous version: Comput. Phys. Commun. 118 (1999) 278.Does the new version supersede the previous version?: The basic methods remained unchanged. The new version runs on modern compilers, and the caller script had been adapted to enter parameters more conveniently.Reasons for the new version: Mainly compatibility issues.Summary of revisions: original k-shell script replaced by Perl script. Outdated Fortran commands replaced by state-of-the-art commands. Resulting configuration is Z1-formatted so that it can directly be used by CPC program Z1 (ADVG).Nature of problem: The problem is to place and relax flexible, semiflexible or stiff, tethered or free model polymer chains within a finite volume with periodic boundaries such that (i) the configurational statistics is obtained from a microscopic potential which determines the local and – affected by concentration and excluded volume – global conformational features of the system in a ‘physical’ way. The resulting configuration obeys a minimum distance criterion.Solution method: In a first step, according to the chosen system parameters, a mixture of phantom and excluded volume chains plus solvent particles is placed into the (finite and final) simulation box by a Monte Carlo algorithm. A subsequent molecular dynamics algorithm solves Newtons equations of motion during the relaxation phase, while the strength of the repulsive and attractive forces, the temperature and integration time step control interact with each other by a global optimization procedure which minimizes the CPU request for the goals (i) the minimum distance between particles is reaching (finally above) its lower limit and (ii) the changes in both local and global conformational properties – as determined by the Monte Carlo procedure – are kept at a very low level. The algorithm interrupts the relaxation process when a break off condition (actually the minimum distance criterion) is fulfilled.Additional comments including restrictions and unusual features: None, except that the machine must provide the needed main memory, proportional to the number of particles. Installation instructions and script syntax are described in the README.txt file accompanying the new version.