Ti Film Thickness Research Articles

Temperature-Corrected Quartz-Crystal Microbalance for Real-Time Film Thickness Monitoring in Vacuum Evaporation

For a quartz-crystal microbalance system without thermocouples for monitoring thin film thickness, a computational method which corrects for the temperature variation in the quartz crystal is presented. This correction method eliminates the need for pipes and electrical wires, simplifying the conventional system. To avoid the use of electrical wires, both a laser diode and a microlens were used to transmit the signal of the change of resonant frequency of the quartz crystal. Using a 4.2 MHz resonant AT-cut quartz-crystal plate, the error in the measurement of Ti film thickness was <3.0% under the conditions of deposition time <230 s and the quartz-crystal temperature <160° C.

0.10 μm TiSi 2 technology utilizing nitrogen diffusion controlled RTA

The effects of Ti film thickness and rapid thermal annealing (RTA) ambient on the properties of narrow TiSi 2 lines were studied. Decreasing the Ti film thickness improves the critical line width of silicidation. Excessive nitridation suppresses the silicidation at Ti film thickness thinner than 30 nm in atmosphere N 2 RTA. Using a low-pressure N 2 RTA process (30 mTorr), a 0.10 μm line width with low resistivity was accomplished at the Ti film thickness of 20 nm without lateral overgrowth.

Rapid thermal nitridation of titanium on Cu/Si contact system

The TiN/TiSi 2 bilayer formed on n +- p junction by rapid thermal annealing (RTA) process was investigated. The various thicknesses of titanium film (10–80 nm) were sputtered on the n +Si substrate, then annealed by RTA at a temperature of 850°C for 30 s in a N 2 ambient (850°C/30 s/N 2). Junction leakage current densities of the n +- p diodes with TiN/TiSi 2 bilayers were about 10 nA/cm 2 for the initial Ti film thickness below 65 nm. The TiN/TiSi 2 bilayer was used as a diffusion barrier between Cu and n +Si. Leakage current density and specific contact resistance measurements were performed on the Cu/TiN/TiSi 2/ n +Si contact system after thermal stress on various temperatures (250–600°C). The thermal stability temperature increasing with initial Ti film thickness was found. The optimum diffusion barrier for Cu n + Si contact system was the sample with initial Ti film thickness 50 nm after RTA 850°C/30s/N 2. It possessed the TiN (20 nm)/TiSi 2 (100 nm) bilayer and the thermal stability temperature up to 475°C for the Cu n + Si contact system. We also displayed cross-section transmission electron microscopy and glancing angle incident X-ray diffraction data at various sintering temperatures and studied the failure mechanism.

Effect of dimension scaling on the nucleation of C54 TiSi2

The effect of the film thickness of C49 TiSi2 on the nucleation of C54 TiSi2 was investigated using X-ray diffraction and transmission electron microscopy. The C49 films of different thickness were prepared by isothermally annealing Ti/polycrystalline Si bilayers of varying Ti thicknesses (250, 550 and 1000 Å) in a high vacuum furnace. Nucleation studies were carried out in the temperature range from 640 to 740°C. It is found that the transformation temperature increases with decreasing initial C49 film thickness. The microstructure and morphology of the C49 phase depend upon the Ti film thickness and significantly influence the nucleation behavior of the C54 phase. An anomalous nucleation mode transition is observed for the C54 TiSi2; for thicker films (starting with 1000 Å Ti), the C54 TiSi2 nuclei predominantly form at normal grain boundaries (two-grain junctions) of the polycrystalline C49 phase, while the C54 phase nucleates at grain edges (three-grain junctions) of the C49 disilicides for samples starting with 250 and 550 Å Ti. The physical significance of these observations associated with the thickness scaling is discussed in terms of energy consideration of nucleation at different geometrical sites, nucleation site density and largely increased surface-to-volume ratio.

Applicability of TiN Adhesion Layer Formed by Nitridation of Sputtered Ti Film to Blanket CVD-W Contact Filling

TiN film formed by rapid thermal nitridation (RTN) of sputtered titanium (Ti) was used as the adhesion layer for chemical vapor deposition (CVD) blanket tungsten (BLK-W) in a very high-aspect-ratio contact filling process. In contact holes, both TiN and TiSi2 layers, which formed simultaneously during RTN become thin in proportion to the decrease of the Ti film thickness because of the increase of the contact aspect ratio. Therefore, contact resistance increase and barrier effect degradation of the TiN adhesion layer during subsequent BLK-W film deposition are of concern when employing RTN of sputtered Ti film as the formation process technique in VLSI's. This study focused on the applicability of a TiN adhesion layer formed by RTN of sputtered Ti film to BLK-W contact filling technology in very high-aspect-ratio contacts.

Titanium Diffusion into LiNbO3 Crystals. Crystal Surface Observations at the Diffusion Process.

Titanium diffused LiNbO3 crystals are widely used in integrated optics. This paper reports the observations of the LiNbO3 crystal surface under the diffusion of Ti. At the first stage of the diffusion process, Ti diffusion areas swelled up to about 3 times that of the initial metal Ti film thickness, and the surface roughness increases. On the progress of diffusion, the height and roughness of the Ti diffusion area decrease; finally the height becomes 1.5-2 times as small as that of the initial metal Ti thickness, and the surface roughness becomes as small as the initial condition. After the diffusion time calculated based on the experimental diffusion coefficie, the crystal surface becomes smooth.

Laser-Beam Direct Writing of TiO2 Channels for Fabrication of Ti:LiNbO3 Waveguides

A novel photoresist-free process used for fabrication of Ti:LiNbO3 waveguides is proposed and demonstrated. In this process, smooth waveguide patterning of TiO2 channels can be performed by laser-beam (LB) direct writing at up to 0.2 mm/s, where Ti film deposited on LiNbO3 is thermally oxidized in air by the focused Ar+ laser beam. LB writing has sufficiently high accuracy for waveguide patterning because the Ti film thickness is usually below 0.1 µm. Unlike the previously reported resist-free processes, the process demonstrated here does not require any specific tools or materials, and can provide Ti:LiNbO3 waveguides more easily. It was also confirmed that there are no marked differences in guiding properties between two Ti:LiNbO3 waveguides fabricated by the proposed resist-free process and the standard process based on waveguide patterning in photoresist as thick as 1 µm.

An Auger and electron energy-loss study of reactions at the Ti-SiO2 interface

In this paper, we use Auger electron spectroscopy and electron energy-loss spectroscopy in order to investigate the room-temperature formation of the Ti/SiO2 interface and its reactivity upon annealing. We perform anneals in the temperature range 200–900 °C for which the phenomena involved are discussed with respect to the Si oxide and Ti film thicknesses.

Process optimization and scale-up of a rapid thermal processing system using design of experiments

Process optimization and scale-up of a rapid thermal processing system was done for TiSi2 films using design of experiments. The conventional two-step TiSi2 process was studied and an experimental design conducted with ten process variables. The screening experiment used to test process variables for main effects, indicated that only three of the ten process variables—low temperature anneal, time of anneal, and Ti film thickness—affected the measured film properties with greater than 95% significance. A second D-optimal design experiment was used to study the affect of three process variables on the film properties of TiSi2. This design studied the main effects as well as quadratic interactions of the process variables on the measured responses of the TiSi2 film. The experimental results predicted an optimized TiSi2 film with 0.60% uniformity and 0.96 Ω/sq sheet resistance. These predicted results were confirmed by device lots. The process capability index of the transferred process is 3.12.

Design parameters of Ti:LiNbO/sub 3/ waveguides for optimum acoustooptic interaction efficiency

A computer model of optical and surface acoustic wave modes in titanium indiffused Y-cut LiNbO/sub 3/ guiding structures is used to produce several graphs for the design of guided-wave acoustooptic devices with optimum interaction characteristics. The design concentrates on waveguide fabrication parameters in order to achieve the optimum acoustic frequency and the acoustooptic interaction bandwidth. The optimum acoustic frequency for a given value of Ti film thickness and diffusion depth is defined as the frequency causing the maximum value of the overlap integral.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

High-accuracy characterization of titanium films for LiNbO/sub 3/ guided wave devices by optical densitometry

An experimental study of optical densitometry as a technique for accurate determination of Ti film thickness for integrated optical devices on LiNbO/sub 3/ is discussed. Densitometric data, obtained with a simple instrument for photographic applications, are compared to conventional stylus profilometric measurements and to microgravimetric and Rutherford backscatter measurements; excellent linearity and a sensitivity of about 3 AA are evidenced, as well as capability of absolute calibration in terms of metal atoms per unit surface area, over the entire range of Ti thickness commonly used in integrated optics. The possibility of using this very simple, rapid, and nondestructive method for systematic sample testing, online deposition monitoring, and utilizing optimum coating area is also presented. >

Arsenic out-diffusion during TiSi2 formation

The behavior of implanted As atoms during high-temperature (800–900 °C) TiSi2 formation was investigated by helium backscattering. A significant As loss was observed during TiSi2 formation by vacuum furnace annealing. The main factors influencing the As loss were the initial As profiles and Ti film thickness. Most of the As loss was caused by rapid As out-diffusion through the TiSi2 layer; the remaining As atoms were only observed in the Si substrate after TiSi2 formation. As long as the As distribution was deeper than the TiSi2 layer, the loss of As atoms was kept to a minimum. Careful selection of the As implantation energy and the Ti film thickness can achieve high As concentration at the TiSi2/Si interface.

Crosstalk characteristics of lens-like Ti-LiNbO 3 intersecting channel waveguides

Crosstalk characteristics of intersecting channel waveguides are presented. Full diffusion times of various Ti film thicknesses are measured first. The concentration distributions in Ti-LiNbO3 crosschannel waveguides are then calculated and compared with the measured distribution. The lens-like medium is the crossover region of crosstalk between Δn and 2Δn crosschannel waveguide.

Ti diffusion in Ti : LiNbO3 planar and channel optical waveguides

Optical mode measurements have been used to determine the anisotropic diffusion coefficient and surface index changes for planar Ti : LiNbO3 waveguides at 0.633 μm. Measured values for the diffusion coefficient at 1000 °C are 𝒟y=9.4×10−13 and 𝒟z=1.4×10−12 cm2/sec. The ratio of the extraordinary to the ordinary surface index change (Δne/Δn0) was found to vary between 1.3 and 1.8, depending on diffusion temperature and Ti film thickness. Ti and Li concentration profiles were measured by secondary-ion mass spectrometry. Additional high concentration peaks, 0.2–0.3 μm wide for both ions, were found superimposed on otherwise well-behaved diffusion profiles. This observation is interpreted to result from a tendency towards Li-Ti-O compound (Li2TiO3 or Li2Ti3O7) formation at the diffusion temperature in a dilute mixture with LiNbO3. Such compound formation has the effect of impeding the Ti diffusion into the LiNbO3 substrate. Electron microprobe measurements were used to measure lateral diffusion from channel waveguides. Under nominally identical conditions, lateral diffusion coefficients varying by as much as a factor of 3 have been observed. The minimum lateral diffusion coefficients observed at 1000 °C were 𝒟y=9.7×10−13 and 𝒟z=1.4×10−12 cm2/sec.