We have investigated plasma-surface interactions with molecular dynamics (MD) simulations. It, however, is high cost computation and is limited to simulations for materials of nanometer order. In order to overcome the limitation, a complementary model based on binary collision approximation (BCA) can be established. We employed a BCA-based simulation code ACAT and extended to handle any structure involving crystalline and amorphous. The extended code, named "ACaT", stores all positions of projectile and target atoms and velocities of recoil atoms, so it can be combined with the MD code. It also holds the potential to reproduce channeling phenomena. Thus it is expected to be useful for evaluation of channeling effects.
In nuclear fusion devices plasmas of hydrogen isotopes exist and contact material surfaces.
The “divertor configuration” is employed in order to control impurities produced by impacts of plasmas to the surface and to reduce the heat load to plasma facing materials. In the configuration divertor plates, whose potential constituent includes carbon and tungsten, are installed. Understanding of the divertor physics and designing an appropriate configuration is essential for the establishment of nuclear fusion reactor. These require knowledge on plasmasurface interactions (PSI). Thus we have investigated interactions between hydrogen atoms and carbon material, such as graphite, with molecular dynamics (MD) simulations. 1,2 The molecular dynamics simulation code solves equations of motion for all particles under the modified Brenner reactive empirical bond order (REBO) potential. 1 Evaluation of the REBO potential requires consideration of effects from multiple particles, and thus it is generally high cost computation. This limits the applicable material scale length of the MD simulation to about order of nanometer and the energy range to about order of kilo electron volt.
In order to investigate PSI with numerical simulation we have to overcome the limitation contained by MD simulation. A complementary model can be established based on binary collision approximation (BCA). There exist intensive PSI related works with binary collision based monte carlo simulations. [3][4][5][6][7][8] The binary collision approximation simplifies interactions between material elements and reduces them to the sequence of the binary collisions. A benefit of the model is that it is rather simple and requires less computing resources than MD model. It, however, holds a limit of application on lower energy region.
We think that a hybrid simulation of molecular dynamics and binary collision approximation is promising. In the integrated simulation code the part of binary collision approximation covers higher energy region, and the part of molecular dynamics governs lower energy region and solves only the vicinity of the projectile and recoil atoms. The threshold energy between the two regions should be determined according to a validity condition of the binary collision approximation, which is roughly estimated as 200 eV.
In this paper we extend an existing binary-collision-approximation-based simulation code ACAT 5 in order to combine MD simulation code and BCA based one.
We employ a binary-collision-approximation (BCA) based simulation code ACAT. The ACAT code was developed to simulate atomic collisions in an amorphous target within the framework of the binary collision approximation. Projectile particles are traced through binary collisions. Target particle, with which projectile collides, is randomly distributed in each unit cell whose size R 0 = N -1/3 , where N is the number density of the target material. In terms of randomly distributed target particle, this code employs the monte carlo method and aims for the atomic collisions in amorphous target.
Figure 1 depicts the trajectory of two particles interacting according to a conservative central repulsive force. The scattering angle in the center-of-mass system (CM-system) is
where,
b is the impact parameter, E r = E 0 m 1 /(m 1 + m 2 ) is the relative kinetic energy, E 0 is the incident kinetic energy of the projectile, V (r) is the interatomic potential, r 0 is the solution of g(r) = 0, m 1 and m 2 are the mass of the projectile and the target atom, respectively.
The trajectories of particles are approximated as the asymptotes of them in the laboratory system (L-system). So they consists of linkage of straight-line segments. The starting point of the projectile and the recoil atom after a collision is given by ∆x 1 and ∆x 2 , which are the shifts from the initial position of the target atom shown in Fig. 1:
where
and the mass ratio
As the interatomic potential V (r), the Moliere approximation to the Thomas-Fermi potential 9 is employed:
where a is the screening length, and Z 1 and Z 2 are the atomic numbers of the projectile and the target atom, respectively.
The procedure of searching the collision partner is as follows: We define a unit vector e p in the direction of a moving projectile and a step length ∆x. Here, the notation “projectile” means not only the incident particle but any recoil atom. The position of the projectile in
different from that of the initial position R (0) , a target atom is produced in the new unit cell by use of four random numbers. Three random numbers are for location and one is for the kind of target atom. This target atom is the partner in the next collision.
The impact parameter b is given as
where R A is the position of the target atom. Figure 2 depicts the situation described above.
We can obtain the trajectory of particles by the eqs. 1-8.
Three parameters are defined: bulk binding energy E B , displacement energy E d , and mi
This content is AI-processed based on open access ArXiv data.