A computer simulation was developed to simulate electron-beam-induced deposition (EBID). Simulated growth produced high-aspect-ratio, nanoscale pillar structures by simulating a stationary Gaussian electron beam. The simulator stores in memory the spatial and temporal coordinates of deposited atoms in addition to the type of electron, either primary (PE), back-scattered (BSE), or secondary (SE), that induced its deposition. The results provided in this paper apply to tungsten pillar growth by EBID on a tungsten substrate from WF(6) precursor, although the simulation may be applied to any substrate-precursor set. The details of the simulation are described including the Monte Carlo electron-solid interaction simulation used to generate scattered electron trajectories and SE generation, the probability of molecular dissociation of the precursor gas when an electron traverses the surface, and the gas dynamics which control the surface coverage of the WF(6) precursor on the substrate and pillar surface. In this paper, three specific studies are compared: the effects of beam energy, mass transport versus reaction-rate-limited growth, and the effects of surface diffusion on the EBID process.

The electron-beam-induced deposition of silicon oxide from tetraethyorthosilicate and tungsten from tungsten hexafluoride is simulated via a Monte Carlo simulation. Pseudo one-dimensional nanopillars are grown using comparable electron-beam parameters and a comparison of the vertical and lateral growth rate and the pillar morphology is correlated to the precursor and deposited material parameters. The primary and secondary electrons (type I) are found to dominate the vertical growth rate and the lateral growth rate is dominated by forward and secondary electrons (type II). The resolution and morphology of the nanopillars are affected by the effective electron interaction volume and the resultant surface coverage of the precursor species in the effective electron interaction region. Finally, the simulated results are compared to previously reported experimental results.

Fundamental proximity effects for electron beam induced deposition processes on nonflat surfaces were studied experimentally and via simulation. Two specific effects were elucidated and exploited to considerably increase the volumetric growth rate of this nanoscale direct write method: (1) increasing the scanning electron pitch to the scale of the lateral electron straggle increased the volumetric growth rate by 250% by enhancing the effective forward scattered, backscattered, and secondary electron coefficients as well as by strong recollection effects of adjacent features; and (2) strategic patterning sequences are introduced to reduce precursor depletion effects which increase volumetric growth rates by more than 90%, demonstrating the strong influence of patterning parameters on the final performance of this powerful direct write technique.

The effects that adsorbed precursor surface diffusion has on electron beam induced deposition are explored via a three-dimensional Monte Carlo simulation. Initially the growth rate and resolution are compared for a common set of deposition conditions with a variable surface diffusion coefficient ranging from 0 to 1 × 10(-8) cm(2) s(-1). The growth rate and resolution are shown to both be enhanced as the growth changes from a mass transport limited regime to a reaction rate limited regime. The complex interplay between the vertical growth rate, the lateral growth rate, the interaction volume and the adsorbed and diffused precursor species are discussed. A second scenario is also simulated in which only gas diffused from a constant source at the perimeter of the simulation boundary is assumed (no gas phase adsorption). At low diffusion coefficients, the diffusing gas is consumed by secondary and backscattered electrons and experimentally observed ring-like structures are generated. At higher diffusion coefficients, the diffusion length is sufficient for the precursor atoms to diffuse to the center (and up the pillar sidewalls) to generate nanowires.