l2.1il6 VERIFICATION OF ION IMPLANTATION MODELS BY MONTE CARLO SIMULATIONS

Moonte Carlo simulations are perfectly suit.ed to check the validity of simiple models. We investi-gate 3 models: First, we show that ID models for thhe imiplantation into multilayer targets give reasonable results only if the stopping powers of mask and bulk material are similar. Second, we discuss the construction of 2D poinit responses from ID profiles. Third, we show that the method of superposing poi it responses at mnask cdges may fail in some cases. 1.lNTRODUCTION The Monte Carlo method is known to be the most powerful tool for ttie slinulatioi of i0 iinplarital-ion. Analytical models, however, require much less CPU! times and allow easy consideration of experimental data. The latter is particularly importallt because Monte Carlo simulations usually assumie amorphous targets so that they do not always yield correct pro-files for implantations into crystalline targets l1j. As simple nodels are usually based on physical con-siderations and Monte Carlo sirnulatio is take physics most accurately into account (apart From the assump-tion of amorphous targets), Monte Carlo simulatLions are perfectly suited to check the validity of these sim-pLe models. In particular, we will investigate in tI is paper ID models for the irplantattion into multiayer targets (Chapter 2), the construction of 21) point re-sponses from ID profiles (Chapter 3), and the imeti od of superposing point responses to obtain dopant dis-tributions near mask edges (Chapter 4). Our Monte Carlo progranm is, fromn a physical point of view, sirnilar to the well kinown programn THINlM 121. One mayor difference of our code is thal we evaluatI scattering angles by interpolation in a precornpu ted table. The 2D simulationrs iave been performed with a code vhich allows arbitrary geonetries. Both features are described in Ref. 131