Fundamental Studies of TiN Film Growth by CVD From Ti ( N M e 2 ) 4 and Ammonia
- PDF / 1,362,238 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 98 Downloads / 159 Views
FUNDAMENTAL STUDIES OF TiN FILM GROWTH BY CVD FROM Ti(NMe,) 4 AND AMMONIA J. A. PRYBYLA, C.-M. CHIANG, and L.H. DUBOIS AT&T Bell Laboratories, Murray Hill, New Jersey 07974
ABSTRACT TiN films were grown by CVD from Ti(NMe 2) 4 and ammonia using a novel gas delivery system which allowed the sample to be kept in high vacuum while the reactants were mixed at elevated pressures. The object was to study fundamental chemistries and growth properties. We obtained clean (99.8%, Isotec; ND3 , 99%, Cambridge) within the gas injector described above. The titanium precursor was transported to the injector by the carrier gas He (99.999%, Airco). The precursor was held at - 60 0 C. Mass flow controllers were used to control the rate of flow of both the He carrier gas and the ammonia. Mat. Res. Soc. Symp. Proc. Vol. 282. @1993 Materials Research Society
288
RESULTS AND DISCUSSION Clean TiN Films: Growth Parameters and Process Latitude The conditions which lead to the CVD growth of stoichiometric (Ti:N = 1.15+/- 0.1), low-contamination TiN films were: substrate temperature of 300°C, source-line pressure of ,-,50-300 torr, growth-chamber pressure of 10- 4 torr, and flowrate ratio of -,3:1 for [pure ammonia] : [He-carrier gas plus Ti(NMe2) 4]. The concentration of Ti (NMe2)4 in the source is unknown but may be quite low since the vapor pressure of Ti(NMe 2)4 at 60 0 C is -1.5 torr.[10] Growth rates varied between -•50 and 10021 per minute. A representative RBS trace for a film grown under these conditions is shown in Figure 1. The Ti-to-N ratio here is near unity while the carbon impurity level has been reduced to below 5%. (We have obtained films with as little as 3% carbon.) In general we found that the growth of low-contamination, near-stoichiometric films is relatively insensitive to the reaction conditions. Specifically, for the film properties measured, we obtained the same results for growth temperatures between 250 and 3500C, and for a large range of NH 3 : Ti(NMe 2)4 ratios where the only requirement seems to be that the ammonia is in great excess. When the ammonia concentration gets too high, the film growth rate slows markedly. The same is true when the substrate temperature gets too low. Good films have been grown with a source pressure in the range 50-300 torr. The prediction of high process latitude that these results suggest is very important and gives some promise for this processing scheme to become manufacturable. We note that while in situ AES showed little or no oxygen present in our films, RBS analysis showed a non-negligible amount of this contaminant (sometimes as high 15 atomic percent). Since oxygen does not appear in either of the precursor gases, we must conclude that the oxygen is absorbed into the film upon exposure to air.[11] Because this oxygen absorption occurs to a greater extent than for PVD-grown films, our CVD films must be more porous. Isotopic Substitution Experiments:
5
1
NH3 and ND 3
A major object of the present investigation was to probe the reaction mechanism of TiN film growth by
Data Loading...
