Binary Collision Approximation Code Research Articles

Hybrid simulation between molecular dynamics and binary collision approximation codes for hydrogen injection into carbon materials

Molecular dynamics (MD) simulation with modified Brenner’s reactive empirical bond order (REBO) potential is a powerful tool to investigate plasma wall interaction on divertor plates in a nuclear fusion device. However, the size of MD simulation box is generally set less than several nm because of the limits of a computer performance. To extend the size of the MD simulation, we develop a hybrid simulation code between MD code using REBO potential and binary collision approximation (BCA) code. Using the BCA code instead of computing all particles with a high kinetic energy for every step in the MD simulation, considerable computation time is saved. By demonstrating a hydrogen atom injection into a graphite by the hybrid simulation code, it is found that the hybrid simulation code works efficiently in a large simulation box.

Relevancy of displacement cascades features to the long term point defect cluster growth

Displacement cascades in iron have been generated by means of the MARLOWE binary collision approximation (BCA) code with primary knock-on atom (PKA) energies ranging from 5 to 100 keV. They serve as input for modelling long term evolution by means of the LAKIMOCA Object Kinetic Monte Carlo code. It is found that the size distributions of the fractions of vacancy and interstitial clustered in the long term are not significantly dependent on the PKA energy in this range. Since the subcascade formation, morphology and spatial extension, as well as the spatial correlations between primary point defect positions do depend on the PKA energy, it is concluded that the size distributions of clustered point defects fractions in the long term do not depend on these cascade features. In contrast, the size distributions of clustered point defect fractions in displacement cascades are found to be independent of the PKA energy while their spatial correlations strongly influence the cluster size distributions in the long term. The use of a mean field approximation in cluster growth kinetics predictions is thereby invalidated. Irradiation dose and dose-rate are also found to be determinant factors governing the long term evolution.

Ion-implant simulations: The effect of defect spatial correlation on damage accumulation

A predictive damage accumulation model, which takes into account different interdependent implant parameters, has been developed. The model assumes that the recrystallization rate of damage structures known as amorphous pockets (AP) is a function of its effective size, regardless of their spatial configuration. In the model, APs are three-dimensional agglomerates of interstitials (I) and vacancies (V), whose initial coordinates are generated by a binary collision approximation (BCA) code. This work addresses the importance of the spatial correlation of I's and V's in modeling damage accumulation and amorphization, by comparing simulations, whereby the initial coordinates of I and V are generated by BCA or randomly generated from the concentration distribution of an input damage profile. Low temperature implantations were simulated to avoid dynamic annealing in order to compare the initial damage morphology. For the same damage level, simulations by BCA resulted in ion mass dependent APs’ sizes, with lighter implant ions generating smaller APs’ sizes, implying more dilute damage compared with heavier ions. However, the ion mass dependent APs’ size effect was lost by loading the same damage profile and randomly positioning the I's and V's. Consequently, the damage morphology, as well as the annealing behaviour obtained by reading I, V damage profiles is substantially different from those obtained using the much more realistic cascades generated by BCA.

A model for ion-bombardment induced erosion enhancement with target temperature in liquid lithium

Sputtering does not vary strongly with target temperature for most solid materials. Sputtering yield measurements of liquid lithium self-sputtering for energies of 200–1000eV at oblique incidence, however, show an enhancement near an order of magnitude as the temperature of the liquid lithium target is increased from near its melting point at 473K up to about 690K (∼1.5Tm) after accounting for thermal evaporation. A new model that couples both the near-surface cascade of the hot liquid metal and the effect of multiple interaction mechanisms on the binding of the emitted particle explains this anomalous erosion enhancement with target temperature. The model, has been validated using a new hybrid computational tool named MD-TRIM (molecular dynamics in transport of ions in matter), at 473K and 653K and compared to experimental data. MD-TRIM consists of molecular dynamics and binary collision approximation (BCA) codes, TRIM (transport of ions in matter). The MD-TRIM code was designed to aid in understanding erosion enhancement mechanisms of lithium self-bombardment sputtering at low bombarding energies with a rise in target temperature. MD calculations alone do not show the sputtering enhancement with temperature rise due to an inadequate surface potential model.

Sampling calibration of ion implantation profiles in crystalline silicon from 0.1 to 300 keV using Monte Carlo simulations

In this work, we study previously published Pearson models in amorphous silicon and present an improved Pearson IV model of ion implantation as a function of implant energy and crystal orientation for use in crystalline silicon. The first 4 moments of the Pearson IV distribution have been extracted from impurity profiles obtained from the Binary Collision Approximation (BCA) code, Crystal TRIM for a wide energy range 0.1–300 keV at varying tilts and rotations. By comparisons with experimental data, we show that certain amounts of channelling always occur in crystalline targets and the analytical Pearson technique should be replaced by a more robust method. We propose an alternative model based on sampling calibration of profiles and present implant tables that has been assimilated in the process simulator DIOS. Two-dimensional impurity profiles can be subsequently generated from these one-dimensional profiles when the lateral standard deviation is specified.

Ion implantation range distributions in silicon carbide

The first to fourth order distribution moments, Rp, ΔRp, γ, and β, of 152 single energy H1, H2, Li7, B11, N14, O16, Al27, P31, Ga69, and As75 implantations into silicon carbide (SiC) have been assembled. Fifty of these implantations have been performed and analyzed in the present study while the remaining implantation data was compiled from the literature. For ions with a limited amount of experimental data, additional implantations were simulated using a recently developed binary collision approximation code for crystalline materials. Least squares fits of analytical functions to the distribution moments versus implantation energy provide the base for an empirical ion implantation simulator using Pearson frequency functions.

Range Distributions of Implanted Ions in Silicon Carbide

The first four distribution moments (R-P, DeltaR(p), gamma and beta) of 113 experimental single energy ion implantations into silicon carbide (SiC) have been assembled to form the base for an empirical ion implantation simulator using Pearson frequency functions. The studied ions are H-1, H-2, Li-7, B-11, N-14, O-11, Al-27, P-31, and As-75, and the implantation energies range from 0.5 keV to 4 MeV. Thirty-nine of these implantations have been implanted, measured, and analyzed in the present study while the remaining implantation data were gathered from the literature. Furthermore, 16 additional implantations were simulated - using a new Binary Collision Approximation (BCA) code for crystalline materials - to fill up the missing energies for ions with limited experimental data. The extracted range data are presented as tabulated fitting constants to analytical functions of the distribution moments versus implantation energy.

Depth profiles obtained from simulation of GaAs bombarded with noble gas ions using MARLOWE

Ion implantation is widely used in integrated circuit manufacturing, and is expected to continue well into the future. In addition, particle bombardment is used to tailor the properties of semiconductor materials and devices. A reliable description of as-implanted depth and vacancy profiles and resulting damage is needed for technological development. The binary collision approximation (BCA) methods are a time economical tool to predict these profiles. During this investigation the BCA code, MARLOWE, was used with a Ziegler, Biersack and Littmark (ZBL) interatomic potential and Firsov’s model for electronic stopping. For each ion investigated (He, Ne, Ar and Kr) the depth profile and vacancy distributions created by the bombardment were calculated at different energies. From these results it is clear that the mass and energy of the ion used to bombard GaAs strongly influences the depth profile and vacancy distribution.

Atomistic simulation of ion implantation and its application in Si technology

Atomistic computer simulations based on the binary collision approximation (BCA) are very well suited to predict the dependence of as-implanted dopant profiles on implant parameters like energy, dose and direction of incidence as well as on the arrangement of oxide, poly-Si and other materials on the single-crystalline Si substrate. In particular channeling effects, the enhanced dechanneling due to accumulation of radiation defects during ion bombardment and due to pre-existing ion-beam-induced defects can be simulated in a reasonable manner. The BCA code Crystal-TRIM was successfully integrated into 1D and 2D process simulators for the Si technology. The application of the trajectory splitting algorithm and the lateral duplication method ensures a high computational efficiency.

Interpretation of sputter depth profiles by mixing simulations

The interpretation of sputter depth profiles can be simplified by use of computer simulations. Distortions caused by mixing effects and distortions caused by the information depth of the analytical method have to be distinguished. Atomic mixing and the information depth distort the depth profile simultaneously. Therefore, it is necessary to take into consideration a superposition of both distortion effects. The sputtering of a GaAs/A1As multilayer has been calculated on a personal computer with the binary collision approximation code T-DYN by Biersack and with an own layer model. A new computer code LAMBDA has been used for the investigation of the influence of the AES information depth in addition to atomic mixing and preferential sputtering. A comparison of the calculated and the measured depth profile explains the observed effects. Therefore conclusions can be drawn about the original elemental distribution in the sample from the measured depth profile.

Round robin computer simulation of ion transmission through crystalline layers

Round robin computer simulations were performed by 11 groups using 6 different molecular dynamics codes and 6 different binary collision approximation codes. The process simulated is the transmission of 0.2 keV, 0.5 keV, and 1.0 keV B atoms through 9 monolayers of <001⪢Si and the transmission of 1.0 keV Ar atoms through 5 monolayers of <001⪢Cu. In all cases the energy distribution and the angular distribution of the transmitted atoms have been calculated with and without taking into account the interaction between the target atoms. The results of the simulations with the different molecular dynamics codes are in good agreement. The binary collision approximation results are compared with each other and with the molecular dynamics data. Deviations are discussed. It is concluded that the binary collision concept is still applicable for such low energies if simultaneous collisions are taken into account.

Round robin computer simulation of ion transmission through crystalline layers

Round robin computer simulations were performed by 11 groups using 6 different molecular dynamics codes and 6 different binary collision approximation codes. The process simulated is the transmission of 0.2 keV, 0.5 keV, and 1.0 keV B atoms through 9 monolayers of <001⪢Si and the transmission of 1.0 keV Ar atoms through 5 monolayers of <001⪢Cu. In all cases the energy distribution and the angular distribution of the transmitted atoms have been calculated with and without taking into account the interaction between the target atoms. The results of the simulations with the different molecular dynamics codes are in good agreement. The binary collision approximation results are compared with each other and with the molecular dynamics data. Deviations are discussed. It is concluded that the binary collision concept is still applicable for such low energies if simultaneous collisions are taken into account.

Low energy ion implantation simulation using a modified binary collision approximation code

We have developed a modified version of the binary collision approximation code MARLOWE with the aim of accurately predict the implanted dopant profiles, even in the low energy case. In our simulations, instead of using the asymptotic path of the projectile, the path of the colliding particles is carefully evaluated for each collision. In order to avoid the computation time penalty of such calculation, the data of the outgoing particles is tabulated previously for a large set of collision events, and an interpolation is done during the main Monte Carlo simulation. The simulated profiles show a better agreement with the experimental data for low energy implants, while the accuracy for the high energy implants is maintained.

Calibrating a multi-model approach to defect production in high energy collision cascades

A multi-model approach to simulating defect production processes at the atomic scale is described that incorporates molecular dynamics (MD), binary collision approximation (BCA) calculations and stochastic annealing simulations. The central hypothesis is that the simple, fast computer codes capable of simulating large numbers of high energy cascades (e.g., BCA codes) can be made to yield the correct defect configurations when their parameters are calibrated using the results of the more physically realistic MD simulations. The calibration procedure is investigated using results of MD simulations of 25 keV cascades in copper. The configurations of point defects are extracted from the MD cascade simulations at the end of the collisional phase, thus providing information similar to that obtained with a binary collision model. The MD collisional phase defect configurations are used as input to the ALSOME annealing simulation code, and values of the ALSOME quenching parameters are determined that yield the best fit to the post-quenching defect configurations of the MD simulations.

Crystal-binary collision simulation of atomic collisions and dynamic damage buildup in crystalline silicon

Abstract Channeling profiles of 5–200 keVB+, 100–500 keV P+ and 40–300 keV As+ implanted near and axes in crystal silicon are calculated using the binary collision approximation code CRYSTAL. Calculated profiles are compared to carefully chosen experimental ones. Also, high dose phosphorus profiles with damage accumulation are compared to published experimental data.

Time development of cascades by the binary collision approximation code

To link a molecular dynamic calculation to binary collision approximation codes to explore high energy cascade damage, time between consecutive collisions is introduced into the binary collision MARLOWE code. Calculated results for gold by the modified code show formation of sub-cascades and their spatial and time overlapping, which can affect formation of defect clusters.