W. Tsai

Plasma cleaned Si analyzed in situ by x‐ray photoelectron spectroscopy, secondary ion mass spectrometry, and actinometry

Silicon surfaces are cleaned in an electron cyclotron resonance excited hydrogen plasma and characterized by in situ x‐ray photoelectron spectroscopy and in situ static secondary ion‐mass spectrometry. Emission spectroscopy and actinometry are used to characterize the hydrogen plasma. Exposure to the plasma for 3 to 4 minutes without applying heat or bias to the substrate completely removes the native silicon oxide resulting in a hydrogen terminated surface that is resistant to reoxidation. Adventitious hydrocarbon, when present on the surface, is also completely removed by the plasma. A shift in the isotope ratios of silicon suggests that a clean 〈100〉 silicon surface is monohydride terminated, whereas a 〈111〉 silicon surface appears largely dihydride terminated. A depth profile of the silicon isotope ratios shows a temporal instability, which with the assignment of a H 1s state in the valence‐band spectra provides evidence that the hydrogen is concentrated at the surface and has not diffused deep into t...

Correlation of electrical resistivity and grain size in sputtered titanium films

Abstract Thin Ti films deposited onto thermal SiO2 by sputtering in Ar or Ne are analyzed accounting for electron scattering at both film and grain boundary surfaces. An intrinsic resistivity of 54 and 63 μΩ cm, and an electron mean free path of 18 and 15 nm is derived for films sputter deposited in Ar and Ne, respectively. For both gases, a grain boundary reflection coefficient of 0.17 is calculated, assuming pure specular electron reflection at the film surfaces. Resistivity lowering is shown to correlate directly with an increase in grain size. Sputtering in Ar results in larger grain size films that have a small 21 μΩ cm residual resistivity and a high 2800 ppm temperature coefficient of resistance. The film grain size is independent of the deposition rate and reduced when sputtered in Ne, or in Ar and an applied negative substrate bias. The Ti film texture is found to be independent of film thickness and the presence of a collimator but sensitive to an applied substrate bias.

Temperature dependence of the electrical resistivity of reactively sputtered TiN films

The electrical resistivity of reactively sputtered TiN films was measured as a function of film thickness. The effect of directionality of the sputtered atoms, substrate temperature, bias voltage, deposition rate, and film morphology on the electron conductivity in TiN films was studied. The combination of rapid deposition rate and high substrate temperature with bias‐collimated sputtering results in TiN films with the lowest resistivity, 45 μΩ cm, the largest temperature coefficient of resistance, 1355 ppm, and the highest superconducting transition temperature, 5.04 K. These films are characterized by small grains with mixed <111≳ and <200≳ orientation and reduced electron scattering with an estimated electron mean‐free path of 96 nm.

Experimental study and computer simulation of collimated sputtering of titanium thin films over topographical features

An experimental study and computer simulation of collimated sputtering of titanium (Ti) thin films onto submicrometer trenches is presented. The effect of square grid collimators with aspect ratios varying from 0.5 to 2 has been studied. Simulation of collimated sputtering involves the combination of the simulation of sputter distributions vapor transport model and the simulation by ballistic deposition film growth model. This combination is able to simulate the effect of collimation on the spatial and angular distributions of deposited atoms, and is able to predict the film coverage, deposition rate change, and microstructure of the deposited films. Both experimental and simulation results show that bottom coverage in trenches of aspect ratio 1.2 can be significantly improved from 50% to more than 80% using collimation. However, a penalty is paid by a corresponding decrease in deposition rate down to 15% of the uncollimated value. Additionally, the microstructure (grain size and orientation) is altered b...

Bombardment and gas rarefaction effects on the properties of sputtered Ti thin films

Abstract The effects of deposition pressure and sputtering power on the properties of Ti films have been investigated. Ti films of two different thickness ranges have been deposited by d.c. magnetron sputtering at nominally 50 °C substrate temperature with various deposition pressures and sputtering powers. For films 650 nm thick, the film density shows a drop of 20% and electrical conductivity decreases by a factor of 2 with increasing deposition pressures from 0.07 to 1.33 Pa at a sputtering power of 3 kW. However, at the higher power of 10kW, the film density and electrical conductivity are relatively unchanged with increasing deposition pressure. For thinner films of 65 nm, no significant effect of pressure on the film density and electrical conductivity is observed at either power level. These results can be explained by the effect of densification of film microstructure due to high energy of incident particles at low pressures. The application of high power sputtering, however, diminishes this effect by the rarefaction of the sputter gas through increased energy and momentum transfer.

Compositional variations in Ti‐W films sputtered over topographical features

To study compositional variations in Ti‐W films sputtered from Ti‐W alloy targets and deposited over topographical features, the sticking coefficients and angular distributions of sputtered flux arriving at the substrate for titanium (Ti) and tungsten (W) atoms have been investigated by an overhang structure, pinhole experiment, and simulation package. The simulation involves the combination of a vapor transport model which is able to model the angular distributions for the respective materials, and a film growth model which is able to predict the compositional variation of the deposited Ti‐W films over topographical features. Experimentally, it was found that the sticking coefficients of Ti and W are both very close to unity for the conditions considered. However, the angular distributions of these two materials are quite different due to their different transport properties through the sputter gas. For the Ti‐W films, the compositional variations calculated using the simulated angular distributions agre...

Hydrogenating silicon dioxide in an electron cyclotron plasma

The hydrogenating effect of a low‐temperature, electron cyclotron resonance excited H2 plasma on the surface chemistry of thermal SiO2 films is analyzed in situ by x‐ray photoemission spectroscopy and static secondary ion mass spectrometry. Hydrogenation with this nominal 10 eV proton flux results in Si‐(O4), H‐Si‐(O3), (H2)‐Si‐(O2), (H2)‐Si‐O, and H‐Si‐(Si3) bonding states to the complete exclusion of Si—OH bond formation. A simple thermodynamic argument accounts for the exclusivity of Si—H bonds terminating the outermost (O3)‐Si‐O‐Si‐(O3) network of a thick SiOx<2 film, thereby transforming what is normally a hydrophilic surface into one that is hydrophobic.

Effect of electron cyclotron resonance H/sup +/, Ne/sup +/, Ar/sup +/, and Xe/sup +/ plasma precleaning on titanium silicide formation

A vacuum integrated cluster tool process incorporating electron cyclotron resonance plasma cleaning, Ti sputter deposition, and rapid thermal annealing in N/sub 2/ is used to form a TiN/sub x 150 eV Ne/sup +/ promotes silicidation, thereby minimizing nitride thickness. The effects of precleaning are significant as the activation energy for TiSi/sub y/ formation is reduced from 1.8 eV characteristic of a BOE cleaned surface to 1.2 eV on Si etched with 250 eV Ne/sup +/. Mechanistically, the silicide kinetics are shown to be inhibited by the presence of a thin amorphous layer that is formed only when cleaning Si with Ar/sup +/ and Xe/sup +/ with the effect that both knock-on oxygen atoms and implanted noble gas atoms trapped within the amorphous layer retard the requisite solid-phase epitaxial regrowth kinetics. Recrystallizing the amorphous Si surface prior to metallization appears to restore the near-normal silicide kinetics that is characteristic of Ne/sup +/ cleaning. >

Study of silicon surfaces bombarded with noble gas ions in an electron cyclotron resonance plasma

The effect of electron cyclotron resonance plasma cleaning (100)Si surfaces with low energy Ne, Ar, and Xe ions is studied by in situ x‐ray photoelectron spectroscopy, high‐resolution cross‐sectional transmission electron microscopy, atomic force microscopy, and measurement of both the single‐surface reflectance and the modulated optical reflectance. Photoemission spectra show that complete removal of all surface contaminants with oxygen atoms knocked‐in from the original native Si oxide and noble gas atoms implanted into the Si substrate are a common characteristic of this plasma cleaning process. The oxygen concentration appears to decrease with ion energy for all three ions, whereas the noble gas concentration is inversely proportional to the ion mass and essentially independent of the ion energy. This low energy ion bombardment sputters the surface causing only point defects with Ne ions, formation of a continuous thin amorphous overlayer with either Ar or Xe ions, and in addition, occasional subsurfa...

Simulation of Uniformity and Lifetime Effects in Collimated Sputtering

Collimated sputtering has been successful in providing good contact barriers for sub-half micron contacts with aspect ratios of 3 and greater. This approach does present drawbacks however, particularly in terms of reduced deposition rates and degraded film uniformity. The flux collected on the collimator leads to closing off of the cells, further reducing deposition rate on the wafer and limiting the life of the collimator. This paper demonstrates simulation of the filling of the collimator with different system configurations and pressures using the SIMSPUD vapor transport and SIMBAD thin film growth simulators. The model can determine collimator filling uniformity, blanket film uniformity, angular distribution of collimated sputter flux, and lifetime of the collimator. Given the target erosion profile, system geometry, and deposition rate, collimator lifetime can be predicted. The model indicates that for a 300 mm diameter source a drop in operating pressure from 0.67 Pa to 0.27 Pa has little effect on collimator life in terms of kWh, while increasing collimator life in terms of wafers by about 50%. The increase in the number of wafers processed comes at the expense of a small loss of uniformity.