RCA Cleaning Research Articles

Effect of RCA cleaning on the surface chemistry of glass and polysilicon films as studied by ToF-SIMS and XPS

The surface chemistry of an alkaline-earth boroaluminosilicate glass is changed by contact with chemical solutions. The present study shows that RCA cleaning creates a silica-rich surface on the glass. This altered surface can be removed by hydrofluoric acid etching. During the RCA cleaning process, glass components can be transferred to a polysilicon film placed in the same alkaline solution. The acidic solution in turn removes most of the contamination from the polysilicon. © 1998 John Wiley & Sons, Ltd.

Materials science communication effects of rca clean-up procedures on the formation of roughened poly-si electrodes for high-density drams' capacitors

A novel structure, the phosphorus-implanted poly-Si films treated with phosphoric acid (H 3PO 4) and cleaned by standard RCA cleanup procedures, has been demonstrated as the bottom electrodes of DRAMs' stacked capacitors. After the H 3PO 4 treatment and RCA cleaning process, micro-island structures are formed on the poly-Si surface of the storage electrodes. The NH 4OH + H 2O 2 + H 2O (SC-1) solution in the RCA cleaning procedures is the main component to change the porous surfaces and engraved structures, formed by H 3PO 4 treatment, into micro-islands. Although the electrical properties are slightly degraded compared to the control samples, but they fulfill the requirements of high-density DRAMs, resulting from the increment of cell capacitance.

SI3N4 Particle Removal Efficiency Study

ABSTRACTControlling particle contamination in wafer cleaning is important to reduce defect density and improve device performance and yield. In this study, a screening experiment was employed to evaluate particle removal efficiency among different cleanings, including FSI BCLN, bench rinse and dry only, bench SC1/megasonic only, bench RCA cleaning, and bench RCA-based cleaning. To optimize particle removal efficiency in RCA-based cleaning, a design of experiment (DOE) has been done to investigate the impact of SC1/megasonic cleaning on Si3N4 particle removal efficiency. Bath temperature, megasonic power, and solution chemistry of SCI bath were evaluated. The removal efficiency in relations to particle sizes was also investigated

Megasonic Irradiation Induced Chemical Reaction in the Solution for Silicon Wafer Cleaning

ABSTRACTWe have already proposed chemical vapor free and room temperature wet cleaning (UCT cleaning) for Si substrate surface, instead of conventional RCA cleaning, which consumes very little amount of chemicals and ultrapure water compared to that of RCA cleaning. This new wet cleaning has been developed by scientifically understanding contaminants removal mechanism as follows;(1) Particles can be removed by simultaneously satisfying following two conditions,(a) particles and substructure surface must have same polarity of zeta potential in the cleaning solution to exhibit repulsive electric coulomb force with each other.(b) adhered particles must be lifted off from substrate surface by slight etching to make electric repulsive force greater than van der Waals force,(2) Redox potential of the cleaning solution must be larger than a critical value to enable removal of electrons from adhered metals in order to dissolve them into the cleaning solution as positive ions and to decompose adhered organic molecules to CO2, H2O etc.Megasonic irradiation is very essential in UCT cleaning, particularly to remove particles by lifting them off from substrate surface. In this study, chemical reaction in the cleaning solution induced by megasonic irradiation is mainly discussed. Irradiation of ultrasonic having frequencies higher than a few hundred KHz (Megasonic) to ultrapure water has been confirmed to be able to decompose H2O molecules to H radicals and OH radicals. Decomposition efficiency of H20 molecules is strongly dependent on remaining gas components and their volume in the ultrapure water. When the remaining gas volume is decreased to less than 0.2 ∼0.3ppm, H2O molecules decomposition to H radicals and OH radicals has not been observed. Generated OH radicals have been confirmed to produce H2O2, and NH4+, NO2−, NO3− ions by reacting. with remaining N2 gas. Thus the cleaning capability of the cleaning solutions can be controlled by irradiating megasonic.

In-Situ Chemical Concentration control cor Wafer Wet Cleaning

ABSTRACTThe continuously increasing integration of today's advanced semiconductors requires increasingly tight process control in the IC manufacturing steps. This paper demonstrates the use of conductivity sensors to monitor and control the chemical concentrations of RCA cleaning and HF etching solutions. Electrodeless conductivity sensors were used to monitor and regulate the concentration of these process chemicals. A linear relationship between the conductivity of the solution and the chemical concentration was obtained within the range studied. A chemical monitoring and concentration scheme (ICE-1™) was developed. Different concentrations of RCA and HF solutions were investigated. Results show that these techniques are suitable for monitoring and controlling the concentration of chemicals in the process tanks for better process control. These techniques provide a lower cost of ownership of the process due to longer bath lives and the use of dilute chemicals.

Total Room Temperature Wet Cleaning for Si Substrate Surface

An ultraclean wafer surface is crucial for high quality processing in Si technologies. Cleaning of the Si wafer surface has been accomplished by RCA wet cleaning for the past quarter century, where there exists high temperature processes consisting of , and treatment. Thus, RCA cleaning requires a large number of processing steps, resulting in the consumption of a huge volume of liquid chemicals and ultrapure water, and simultaneously consuming a large volume of clean air exhaust to suppress chemical vapor from getting into the clean room. Total room temperature wet cleaning consisting of five cleaning steps has been developed for Si wafer surfaces, where consumption volume of liquid chemicals and ultrapure water has been reduced less than 1% and 5%, respectively, compared to that of RCA cleaning. The newly developed cleaning technology has been confirmed to contribute to future simplified and low cost manufacturing of ultralarge scale integrated devices.

Thickness of the [formula omitted] interface and composition of silicon oxide thin films: effect of wafer cleaning procedures

We determined the areal density of Si atoms constituting the oxide-silicon interface and the stoichiometry of ultra-thin silicon oxide films, thermally grown on Si(001) in dry 18O 2 atmospheres, using the channeling of α-particles along the 〈001〉 axis of the Si substrates associated with grazing angle detection of the scattered particles. The amount of 18O atoms in the films was determined independently using the 18O(p,α) 15N nuclear reaction at 730 keV. The Si wafers were submitted to different cleaning procedures before oxidation in 18O 2, namely: standard RCA cleaning, HF etching followed by a rinse in ethanol and rapid thermal cleaning (RTC) under high vacuum. The stoichiometry of all oxide films having thicknesses between 2 and 13 nm could be fitted assuming a ratio O Si = 2 , that is, the films were constituted by silicon dioxide. By comparing the results for samples cleaned in different ways, however, we noticed a pronounced change in the number of atoms in the non-registered Si layers at the SiO 2 Si interface and so in the thickness of these interfaces.

Ultrahigh vacuum system for atomic-scale planarization of 6 inch Si(001) substrate

Ultrahigh vacuum (UHV) system equipped with aluminum radiation shields has been developed in order to obtain atomically flat Si(001) surfaces in a large area. Aluminum radiation shields were employed to suppress the temperature rise of chamber walls. The UHV system has enabled a 6 inch substrate to be heated up to 700°C under a pressure of less than 10 −8 Pa range. The surface structure of the substrate was confirmed by the reflection high energy electron diffraction (RHEED), when the 6 inch substrate was heated at 600°C in UHV after a modified RCA cleaning, the RHEED pattern indicated oxide free surface consisted of2 × 1structure. The roughness was evaluated from numerous observations by cross-sectional transmission electron microscopy (XTEM). Root-mean-square roughness of the Si surface was less than 0.1 nm for almost all over the 6 inch substrate. This implies that the heating in this system gives rise to the atomic-scale planarization of the 6 inch substrate.

Reliability Evaluation of Manufacturing Processes for Bipolar and MOS Devices on Silicon‐on‐Diamond Materials

The chemical stability of polycrystalline diamond subjected to common silicon device integrated circuit process steps has been investigated using Raman spectroscopy, scanning electron microscopy, I‐V, and capacitance‐voltage. Wet chemical process steps such as RCA cleaning, KOH etching, and aluminum etching were performed without significant degradation of the diamond. No contamination, as judged from electrical data, of reference metal oxide semiconductor capacitors from the diamond was seen during furnace treatments. Diamond was found to withstand annealing at 950°C without electrical degradation. Patterning of diamond was demonstrated and utilized in manufacturing of test structures. Methods have been found to protect the diamond at temperatures up to 1100°C during thermal oxidation of adjacent silicon. Capping layers consisting of silicon nitride, silicon nitride on top of chemically vapor deposited oxide and polysilicon were found to protect diamond during thermal oxidation. A silicon‐on‐diamond wafer process based on process steps investigated in this paper is proposed.

Analyses of HF/NH4F buffer-treated Si(111) surfaces using XPS, REM and SIMS

The effect of different cleaning procedures on Si(111) wafers has been studied. A three-step cleaning process was used. The first two steps (thermal oxidation followed by RCA cleaning) were common to all samples. The final step involved rinsing in one of a set of HF/NH4F buffer solutions with a wide range ofpH values. Three different surface techniques were used for characterizing the chemical condition and morphology of the treated surfaces: XPS (X-ray Photoemission Spectroscopy), REM (Reflection Electron Microscopy) and SIMS (Secondary-Ion Mass Spectroscopy). It has been found that thepH value of an HF solution does significantly affect the etching rate and morphology of the Si(111) surface: For the same type of solution, the smaller thepH value, the higher the etching rate. Basic solutions withpH values larger than eight have a much weaker etching effect on the surface, which is contradictory to some previous reports. The most effective solutions for the etching of the Si(111) surface are the solutions of HF buffered by NH4F, with thepH in the range of 2–6. REM images indicate that the surface morphology after etching in the HF solution is strongly affected by the length of the etching time: Overetching will roughen the surface. The SIMS data show that water rinsing in air during the cleaning process does speed up oxidation, but it is necessary to use water to clean off the residuals from the HF solutions.

Metal contamination of silicon wafers induced by reactive ion etching plasmas and its behavior upon subsequent cleaning procedures

Reactive ion etching (RIE) causes several types of damage in fabrication of integrated circuits. In this article the metallic contamination content of silicon wafer surfaces after performing RIE plasmas has been measured by total reflection x-ray fluorescence. The dependence of the contamination content on different electrode materials was investigated and could be shown to be mainly a function of the electrode material, not as much as of the rest of the material of which the reactor is built. In order to remove the metals found on the wafer surface after RIE, different cleaning cycles were applied to the wafers and it could be shown that the standard RCA cleaning, followed by a HF dip, works well for most contaminants, with the exception of Cu and Ni. In particular, it was found that Ni deposited onto a silicon surface as a consequence of a RIE type plasma cannot be removed by a standard RCA clean. There are strong indications that Ni is not deposited in the form of atoms, but that a stable nickel silicide is formed.

A Simple Chemical Treatment for Preventing Thermal Bubbles in Silicon Wafer Bonding

A periodic acid aqueous solution has been shown to remove thermally unstable hydrocarbons from silicon surfaces resulting in residual hydrocarbon concentrations which are much lower than those after RCA cleaning. Bonded pairs of silicon wafers cleaned in the periodic acid solution prior to bonding generate no bubbles after annealing at any temperature. The powerful oxidizing reaction of periodic acid solution with hydrocarbons is believed to be responsible for the performance. Periodic acid solution is environmentally friendly, hazard‐free, and compatible to mainstream semiconductor technology.

Influence of prebonding cleaning on the electrical properties of the buried oxide of bond-and-etchback silicon-on-insulator materials

Three different groups of metal-oxide-semiconductor devices were manufactured of bond-and-etchback silicon-on-insulator wafers where the buried oxide functioned as the gate dielectric. The groups differed in the procedure used to clean the surfaces prior to bonding and in the location of the bonded interface. The surfaces were cleaned using either the standard RCA cleaning procedure without HF dip or by rinsing in de-ionized water only. The location of the bonded interface was in the buried oxide or at its interface toward a silicon wafer. The RCA-cleaned devices with the bonded interface within the buried oxide were found to degrade severely under bias temperature stress. This degradation was evident from both oxide charging and an increase in the density of states at the Si/SiO2 interface for negative gate biases. For positive biases the most prominent effect was lateral nonuniform charging of the oxide. The lateral nonuniformities might be connected to voids formed by ammonia desorption during postbonding annealing. Devices rinsed in de-ionized water prior to bonding and devices with a homogeneous oxide showed only slight degradation after bias temperature stress. Electron injection by internal photoemission showed that the buried oxides contained electron traps with capture cross sections corresponding to Coulomb attractive traps. The different processing conditions did not affect the trap cross section but influenced the trap density Nt. Devices with the bonded interface within the buried oxide had Nt≊1×1011 cm−2 in the case of RCA cleaning and Nt≊4×1010 cm−2 for de-ionized water rinsing. The devices with a homogeneous oxide had Nt≊4×109 cm−2. Electron trapping in the Coulomb attractive traps was accompanied by a corresponding increase in the density of states at the thermally grown Si/SiO2 interface for devices with a bonded buried oxide. SIMS investigations revealed a correlation between the degradation upon stress and hydrogen concentration in the devices with bonded oxides.

Determination of phosphorus distribution in the region of a SiO2−Si interface by substoichiometric analysis

A simplified method for the substoichiometric analysis of phosphorus has been developed and applied to determine the concentration distribution of phosphorus in the region of a SiO2−Si interface in order to explain why phosphorus is lost from the ion-implanted silicon surface throughout the oxidation and oxide removal processes. It is revealed that phosphorus piles up on the SiO2 side at the interface by the thermal oxidation of silicon surface and is removed with the oxide by wet etching and with the resulting silicon by RCA cleaning. This results in a total loss of ion-implanted phosphorus of 3.5%.

Hydrogen-terminated Si(100) surfaces investigated by reflectance anisotropy spectroscopy

Boron-doped Si(100) wafers (1 × 10 15 cm −3) with different off-angle orientations (0, 4.5 and 6°) towards [011] were prepared with a modified RCA cleaning and a 5% HF dip. The resulting hydrogen-covered surface was examined either in an ultrahigh vacuum (UHV) or a metallo-organic vapour-phase epitaxy (MOVPE) environment by reflectance anisotropy spectroscopy (RAS) and in the UHV chamber also by low energy electron diffraction and Auger spectroscopy. Under UHV conditions the (1 × 1) (at room temperature) dihydride and the double-domain (2 × 1, 1 × 2) (at higher temperatures) monohydride reconstruction of the hydrogen-terminated surface were observed. Independent of the surface reconstruction the RAS spectrum always exhibited a peak at 3.45 eV with a half-width of 100 meV. The height of this peak increases with increasing misorientation towards [011] and vanishes for on-axis wafers. By annealing up to 560 °C the RAS signal nearly disappears in correlation with the desorption of the hydrogen. Under MOVPE conditions (H 2 carrier gas at 100 mbar), reflectance anisotropy spectra similar to those measured under UHV conditions are obtained. Simulating the first step of GaAs on Si growth by adding AsH 3 to the carrier gas 560 °C (above the AsH 3 decomposition temperature) it is found that As is deposited on the surface. The corresponding reflectance anisotropy spectrum with a double-peak structure resembles that found after depositing 1 monolayer of arsenic under molecular beam epitaxy conditions. Reflection high energy electron diffraction showed in the latter case a double-domain (2 × 1) reconstruction. The data are discussed in terms of stepped surfaces with single or double steps. While in the former the RAS signals from different terraces cancel, they add up in the case of double steps and produce a RAS signal. The appearance of the RAS signal seems to be furthermore correlated with the presence of SiH bonds or SiAs bonds.

Effects of predeposition HF/NH4F treatments on the electrical properties of SiO2/Si structures formed by low-temperature plasma-assisted oxidation and deposition processes

Si(100) and (111) wafers were prepared by a standard RCA cleaning process followed by a treatment in HF/NH4F solutions with different pH values. SiO2/Si structures were formed on these surfaces by a two-step low-temperature, 200–300 °C, plasma-assisted oxidation/deposition process. The SiO2/Si(111) interface, subjected to a predeposition rinse in a 40 wt % NH4F solution, displayed a midgap interface trap density Dit of approximately 5×1010 cm−2 eV−1, which is significantly lower than is generally observed on thermally oxidized Si(111). The Dit values increased systematically up to about 2×1011 cm−2 eV−1 as the pH of the HF/NH4F treatment was decreased. For Si(100) wafers, the absolute level, and the energy distribution of Dit within the Si band gap, were essentially independent of the pH of the HF/NH4F treatment. The micromorphology and perfection of the H passivation of the Si surfaces after the HF/NH4F treatment were characterized by Auger electron spectroscopy and low-energy electron diffraction, and were correlated with the electrical properties of the resulting SiO2/Si structure. It was observed for the first time that the Si(100) surface partially reconstructs into a 2×1 structure after a rinse in 40 wt % NH4F that follows a traditional RCA cleaning treatment.

A New Two-Step Plasma-Assisted Surface Cleaningoxidation and Film-Deposition Process Sequence for The Formation of Si(100)/SiO2 Interfaces with Low Densities of Interfacial Traps

ABSTRACTDevice quality SiO2 films have been deposited by Remote Plasma Enhanced Chemical Vapor Deposition (RPECVD). The substrate surface cleaning, both ex-situ and insitu, influence the density of trapping states at the Si/SiO2 interface. Pre-deposition exposure to plasma-generated atomic-H significantly reduces hydrocarbon and 0 contamination, but damages the Si surface leading to high interface trap densities (Dit) in the Si band-gap. Pre-deposition exposure to plasma generated O-atoms, removes the residual hydrocarbons, and oxidizes the Si substrate while maintaining an atomically smooth interface. Metal Oxide Semiconductor (MOS) structure fabricated by RPECVD have midgap Dit values of ∼1-3×1010cm−2eV−1 for 0-plasma cleaned surface. Final rinse after RCA cleaning in HF solutions with different pH values affects surface morphology; however surface roughening by atomic-H is always the determinant contributor to high values of Dit.

Etching characteristics of Si1−xGex alloy in ammoniac wet cleaning

Etching characteristics of Si1−xGex alloys in ammoniac wet cleaning (RCA cleaning) were examined. The etching rate of Si1−xGex became larger with increasing Ge ratio (X). Temperature dependence of the etching rate was studied and the etching rate was large at high temperatures. However, no obvious difference was observed in the temperature dependence of Si1−xGex etching rate at different Ge ratio (X). A surface morhology degradation after RCA cleaning was observed at high Ge ratio (X). A stoichiometry change of Si1−xGex surface after RCA cleaning was observed by x-ray photoelectron spectroscopy (XPS). The etching rate increase and the surface morphology degradation are thought to be due to the rapid etching of Ge atoms at the top surface layer.

Characterization of Si Layers Deposited on (100) Si Substrates by Plasma CVD and Its Application to Si HBTs

Studies have been carried out on Si layers deposited on (100) Si substrates by plasma CVD, and on heterojunction devices fabricated with the layers. These Si layers vary from microcrystalline to epitaxial structures depending on substrate surface cleaning and deposition temperature conditions. The characteristics of Si HBTs using these layers as emitters depend greatly on the structures of the layers. Very thin interfacial oxides formed by RCA cleaning are effective in growing homogeneous microcrystalline emitters. Hetero-junction effects exemplified by high current gains at room and low temperatures demonstrated the usefulness of the microcrystalline emitters for HBTs and future BiCMOS technology.

The influence of an aluminum overlayer on the interaction of tungsten films with silicon substrates

The interactions of tungsten films with silicon and aluminum during annealing treatments of W/Si and Al/W/Si structures have been investigated using sheet resistance measurements, nuclear reaction analyses, x-ray diffraction techniques, Rutherford backscattering, and Auger electron spectroscopy. W(H) and W(Si) films, 100–200 nm thick, were deposited via the H2 reduction of WF6 on undoped single-crystal Si wafers and the Si reduction of WF6 on As implanted Si wafers, respectively. Prior to the metal deposition, As implanted Si wafers were cleaned by dipping in NH4OH:H2O2 followed by HCl:H2O2 at 80 °C for 20 min (RCA cleaning treatment). The oxygen content at the metal–Si interface in W(Si)/As implanted Si structures can reach 3 at. % whereas in W(H) /mono-Si structures, the metal–Si interface is oxygen free. In W/Si samples, the formation temperature of WSi2 was found to be dependent on the interfacial oxygen concentration. For oxygen-free W/Si structures, the silicidation reaction occurred at 625 °C while for structures containing interfacial oxygen atoms, this reaction was observed above 800 °C. In Al/W/Si structures, the intermetallic compound WAl12 was formed by annealing at 450 °C for 90 min. Tungsten atoms diffused in the aluminum overlayer and islandlike structures of WAl12 surrounded by unreacted Al appeared at the surface of samples. Furthermore, the formation of WSi2 was observed in the Al/W/Si structures annealed at a temperature in the range of 550–600 °C regardless of the oxygen concentration at the W–Si interface. The mechanisms of WAl12 and WSi2 formation are analyzed and the role of interfacial oxygen atoms in the silicidation reaction is demonstrated. The origin of the diminution in thermal stability of W–Si interfaces in Al/W/Si structures is also discussed in this paper.