First project period 04/2006 - 09/2010

Project No. Project Name Responsible
G2 Integrated Modeling and Simulation of Deposition and Planarization Processes for Interconnect Systems Prof. Dr. Thomas OTTO
Dr. Hermann WOLF

Description

Further application of advanced deposition processes for interconnection systems in high performance integrated circuits requires support by modeling and simulation [G2-1].

The first co-operation in this field between Fudan and Chemnitz University started in 1996. An important result of the joint project "Process simulation of new materials" [CHN 345-96] was the development of basic modules for an extendible, integrated, in-house simulation tool T2 used for process, equipment, and interconnect simulation.

Precise knowledge of feature topography after barrier and seed layer deposition is an important prerequisite for the simulation of subsequent ECD and CMP steps. Severe problems can arise from overhangs caused by isotropic components of particle flow and from incomplete coverage by the seed layer. Thus, for a variety of advanced deposition processes (IPVD, ALD, MOCVD, PECVD, ...) process development has to be supported by topography simula-tion. Therefore, advanced tools for three-dimensional simulations on feature scale are needed. For selected processes, reactor scale simulation can be performed to establish the link be-tween process conditions and topography. Simulation of IPVD is currently performed using the simulation tool T2 which has been improved further within the scope of the nationally funded projects MOSAIK and SKALAR. Simulation comprises the description of barrier deposition under special consideration of minimum overhang generation and of topography evolution during seed layer deposition under special consideration of complete coverage.

Simulation of ballistic particle transport in rarefied gases is based on the Monte Carlo approach. Although sophisticated models exist for the description of particle interaction in the gas phase, application to semiconductor processing remains difficult. Main reasons are the complex mutual influencing of charged particles and fields in and around the plasma region and the incomplete knowledge about the interactions between particles and solid surfaces. The properties and results of such surface reactions (adsorption and reflection probabilities, yields, dynamic state of the leaving species) can be calculated using molecular dynamics. This approach of modeling the interactions on atomic scale is based on the approximation of quantum interactions by classical ones. Instead of solving the Schrödinger equation, a semi-empirical model of classical interaction between atoms is established and then the Newton equations are solved. The construction of suitable interaction potentials remains an issue for a lot of materials to be applied in future metallization systems.

In this context, cooperation with the German subprojects G1, G3 (Gessner) is planned with respect to processing and simulation of entire interconnect systems. Cooperation with G6 (Richter) and G5 (Lang) will be beneficial with respect to the materials properties necessary for simulation. Further cooperative work is planned with the Chinese subproject F3 (Qu Xin-ping).

References

[G2-1] ITRS, 2004 Update, Semiconductor Industry Association, San Jose, CA, 2004.