Thin films exhibit diverse morphologies involving clusters, facets, columnar structures, voids, and intricate grain boundary networks. Predictive models of thin film microstructure could be of great value in the design of devices, which require more exacting control of the film microstructure as the scale of the devices is reduced. We have developed a Monte Carlo model for polycrystalline thin films, which is capable of representing a wide range of conditions prevailing during deposition. In addition, the model can simulate time and length scales approaching those in actual device fabrication lines. The model was developed primarily for application to physical vapor deposition (sputtering, ionized PVD, MBE, etc.), but it also provides insight into deposition by other methods. We consider refractory materials such as Ta as well as those with higher diffusivities such as Cu and Al. The influences of energetic particles and clusters in the deposition flux are also discussed. Computer generated movies of the simulations will illustrate several of the film growth mechanisms.
Joint work with : G.H. Gilmer , D.L. Windt , F.H. Baumann , P. O'Sullivan , M. Djafari Rouhani 
1. Bell Labs, Lucent Technologies, Murray Hill, NJ 07974
2. LPST & LAAS-CNRS, Toulouse, France