Conformality and Composition of Films Deposited at Low Pressures
- PDF / 2,169,785 Bytes
- 12 Pages / 414.72 x 648 pts Page_size
- 55 Downloads / 216 Views
number implies that collisions between species in the gas phase are much less likely than collisions between gas phase species and the surface which defines the feature. In this context, the area which connects the source volume to the feature should be considered to help define the feature. This transport regime is termed free molecular or ballistic transport. To date, modeling and simulation efforts for feature scale transport and reaction have focused on solving species (material) balances. The majority of studies have focused on predicting film profile evolution; however, a few have considered the possibility of composition profiles [4-6]. Other film properties are not currently amenable to first principles modeling and simulation; however, see Refs. 7 and 8. In this paper, I first review the ballistic transport and reaction model valid for low pressure deposition (and etch) processes inside micron scale features on patterned wafers. I then discuss how studies which combine experiments and simulations are being used to help understand deposition processes. The processes discussed are; 1) film profile evolution during LPCVD of SiO 2 from TEOS, 2) film profile evolution during plasma enhanced deposition of Si0 2 from TEOS/oxygen mixtures, and 3) composition variations in the sputter deposition (PVD) of titanium-tungsten films.
BALLISTIC TRANSPORT AND REACTION MODEL More complete descriptions of the BTRM can be found in publications by Cale and coworkers [9-11], Islam-Raja et al. [12], Singh et al. [13] and Hsieh and Joshi [14]. The BTRM is based on species balances for the various species which exist in the deposition system, both in the gas phase and on the surface of the evolving film. The basic assumptions of the BTRM are: 1. The frequency of collisions between gas phase species is small relative to the frequency of collisions between gas phase species and surfaces; i.e., intra-feature transport is by free molecular flow. 2. Deposition occurs by heterogeneous reactions between gas phase species and the evolving film surface. 3. The film grows slowly relative to the redistribution of fluxes to the feature surfaces caused by film evolution. The third assumption is supported by comparing molecular speeds with the speed of the evolving film profile. The second assumption does not imply that homogeneous reactions are not important to film deposition. Homogeneous reactions are not significant inside features; however, they might be important in the reactor volume. Homogeneous reactions can indeed lead to precursors which then react on the surface. Other assumptions regarding the BTRM, as used for the studies discussed in this paper, are discussed below. It is important to note that from the point of view of transport and reaction based modeling of deposition processes in features, there is no intrinsic difference between the LPCVD, PECVD and PVD systems used as examples in this paper. The assumptions regarding species transport and surface interactions differ. In all cases, the open end of the feature is exposed t
Data Loading...
