Film stoichiometry

Plasma-deposited siUcon nitride contains large amounts of hydrogen, typically in the range of 20—25 atomic % H, and has polymer-like properties. The electrical resistivity of the film depends on the deposition temperature, the film stoichiometry, and the amounts of hydrogen and oxygen in the film. 

The run-to-run deposition zone temperature (350-425 °C) was the first parameter varied. Temperature variation affected film stoichiometry [Fig. 6.25(a)] and crystalline orientation [Fig. 6.25(b)], while not significantly affecting the deposition rate [Fig. 6.25(c)], From Fig. 6.25(a), we can see that the films were closest to stoichiometry when deposited at 395 °C. Cu-to-In ratios ranged from... 

The next parameter varied was the location of the susceptor within the deposition zone. Variations of the susceptor location did not affect film stoichiometry, but they did influence the morphologies of the films. When the susceptor was moved toward the evaporation zone, denser, smoother, and shinier looking films were obtained from a reduced concentration boundary layer (less diffusion-limited) toward the evaporation zone. 

Variation of the deposition zone temperature affected the film stoichiometry and crystalline structure while not significantly affecting the deposition rate. A deposition zone temperature was optimized at 395 °C to produce (112)-oriented films without any detectable secondary phases. The susceptor location within the furnace did not affect the stoichiometry of deposited films, but it did alter morphology. Moving the susceptor toward the evaporation zone... 

Figure 3 Film stoichiometry as a function of WF6 flow rate.7...

The second study was done in a cold-wall reactor12-13 using the same reactants. The reactor was a single-wafer system, similar to the tube reactor of Figure 18 in Chapter 2, with the wafer heated by an electrical resistance heater in the pedestal. In this case, the sublimator was operated at 88°C with a 10 seem flow of H2. The influence of SiH4 flow rate on the film stoichiometry and resistivity (after anneal) are shown in Figure 11. 

The only other plasma-enhanced CVD film that has seen wide use in integrated circuit manufacture is the plasma oxide film. We say "so-called" because it is not truly Si02, but rather SiOxNyHz. In fact, it is just this ability to modify the film stoichiometry that makes these films so valuable. Many of the film characteristics change depending on this stoichiometry, so it allows a freedom to alter film characteristics that is not possible with thermally-grown films. 

The ellipsometric technique described earlier has the unique feature that the index of refraction can be determined independently of the film thickness. Then, knowledge of this index can be used to infer the chemical composition of a film. For example, thin silicon dioxide films have an index of 1.46, while silicon nitride films have a value of 2.0 typically. Now, when either of these films are deposited by PECVD techniques, their stoichiometry can vary depending on deposition conditions. It turns out that this variation in stoichiometry can be related to the measured refractive index. Accordingly, measurements of the refractive index can be used as an approximate guide to film stoichiometry. 

The most commonly used methods for the preparation of ultrathin oxide films are (1) direct oxidation of the parent metal surface, (2) preferential oxidation of one metal of choice from a suitable binary alloy, and (3) simultaneous deposition and oxidation of a metal on a refractory metal substrate. The detailed procedures for (1) and (2) are discussed elsewhere [7,56,57] procedure (3) is discussed here in detail. Preparation of a model thin-film oxide on a refractory metal substrate (such as Mo, Re, or Ta) is usually carried out by vapor-depositing the parent metal in an oxygen environment. These substrate refractory metals are typically cleaned by repeated cycles of Ar sputtering followed by high-temperature annealing and oxygen treatment. The choice of substrate is critical because film stoichiometry and crystallinity depend on lattice mismatch and other interfacial properties. Thin films of several oxides have been prepared in our laboratories and are discussed below. 

The dependence of upon film composition for CVD-derived YBCO films using Y(dpm)3, Ba(dpm)2, and Cu(dpm)2 precursors under 1.25 Torr oxygen partial pressure with Ts = 750°C is shown in Figure 2-21 [146], Elemental determination by X-ray spectroscopy (EDS) was used to assay film stoichiometry. The dashed circle in Figure 2-21 corresponds to superconducting films, with the smaller solid circle indicating the range... 

Thus film stoichiometry would be maintained in the deposit. Several oxides also behave in this way, for example, B2O3, GeO, and SnO. 

A necessary condition for cBN phase formation in thin films is that boron and nitrogen are incorporated in a nearly 1 1-ratio. Based on highly accurate composition measurement using neutron depth profiling, Hackenberger et al. [41] investigated the relationship between film stoichiometry and the phases present in the film. From these measurements, together with an analysis of data from the literature, they found evidence that film stoichiometry is one of the factors that stabilize the cubic phase. 

ToF-ERD is a powerful technique for profiling multilayered, multielemental samples, and thin films as may be found in the microelectronic industry, to provide information about the film stoichiometry, homogeneity, surface, and interfacial properties, necessary to engineer the film to the desired functional properties. 

Metal alloys, such as Al u, Co r or Ni-Cr, can generally be evaporated directly from a single heated source. If two constituents of the alloy evaporate at different rates causing the composition to change in the melt, two different sources held at different temperatures may be employed to ensure uniform deposition. Unlike metals and alloys, inorganic compounds evaporate in such a way that the vapor composition is usually different from that of the source. The resulting molecular structure causes the film stoichiometry to be different from that of the source. High purity films of virtually all materials can be deposited in vacuum by means of electron beam evaporation. 

The refractive index of silicon nitride and silicon carbonitride thin films deposited at various temperatures is shown in Fig. 9. The re active index of silicon nitride has been well characterized and is generally reported to be between 1.8 and 2.1 [18]. The scatter in the reported values is largely attributed to variations in film stoichiometry and methods of deposition. The presence of impurities such as hydrogen, oxygen and firee silicon in particular, may also account for some of the r orted divergences. The measured refractive index of silicon nitride increased with deposition temperature from 1.82 to 1.95. Minor fluctuations in the atomic ratio of Si—to—N, that can be seen from the AES analysis (Kg. 7) may be responsible for the observed dependence of the refractive index on the deposition temperature. The refractive index for silicon carbonitride similarly ranged from 1.68 to 1.94. 

Gonzalo, J., Afonso, C.N., and Perrtere, f. (1996) Influence of laser energy density on the plasma expansion dynamics and film stoichiometry during laser ablation of BiSrCaCuO. J. Appl Phys., 79, 8042. 

Wicklein, S., Sambri, A., Amoruso, S., Wang, X., Bruzzese, R., Koehl, A., and Dittmann, R. (2012) Pulsed laser ablation of complex oxides the role of congruent ablation and preferential scattering for the film stoichiometry. Appl Phys. Lett, 101, 131601. 

---