Single chemistry cleaning may serve as a potential substitute for the conventionally used RCA cleaning sequence to meet progressively stringent requirements during semiconductor fabrication. The current study involves stability determination of aromatic complexing agents (CAs) such as catechol, 8hqsa and pyridinone-type compounds employed in 1/4/20 APM and related cleaning mixtures (like 1.65/1/5 NC and TPM) at 35degreesC and 50degreesC. The CA concentration was monitored as a function of UV absorption after periodic sampling from corresponding solutions. CA degradation was assumed to follow a linear or an exponential decay representing rate equation laws of zeroth and first or pseudo-first order, respectively, and the lifetimes of the CAs (t(1/2)) deduced accordingly. Compound 'X' was the most stable of all CAs under investigation. All others were less stable by a factor of 3-5. The stability of CAs may be discriminated according to their radical scavenging capability and where applicable the oxidation potentials of their tautomers.

As dimensions scale down and government, regulations are becoming stricter, the industry is moving to dilute single step cleaning, preferably with,low etching and re-usable chemicals. In combination with a single tank tool this provides a high throughput process that can be run on a low footprint tool, economically translated: a low Cost of Ownership. This article shows that APM+ -a chelating agent modified APM- run in a Single Tank Tool, is a good alternative for traditional cleaning sequences. APM+ is able to remove metals without risk for redeposition, without altering other APM properties like particle neutrality and removal. Electrical data show that a yield, of almost 100 % is easily attained, compared to 0 % for the same chelating agent free APM solution with added metal contaminants, proving the possibilities of APM+ as an alternative for future cleaning.

Wet chemical etching using hot phosphoric acid at high temperature is commonly used for the removal of nitride films during silicon wafer processing. The nitride-to-oxide etch selectivity using H3PO4 at 160 degreesC typically amounts to 50:1. As rather high over-etch times are employed to ensure complete removal of the nitride films and because thinner and/or faster etchable oxide films are used for future generations of technologies, a more reproducible and higher etch selectivity is demanded. In this paper it will be demonstrated that the nitride-to-oxide etch selectivity can be significantly improved by using a Si-based additive in the bath.

We have studied the particle removal efficiency of HF-based cleaning mixtures used to clean wafer surfaces during semiconductor manufacturing. SiO2, Si3N4, and metallic oxide (Al2O3, TiO2) particles can be easily removed from silicon wafers using a HF-based clean, whereas the removal of metallic particles and especially Si and polymeric particles is much more difficult. This is explained in terms of surface hydrophobicity effects. For thermal oxide wafer substrates, a low removal efficiency is observed for the positively charged Si3N4 and Al2O3 particles. This has been explained previously by redeposition of the particles from the carry-over layer during the final rinse [R. Vos, I. Cornelissen, M. Meuris, P. Mertens, and M. Heyns, in Cleaning Technology in Semiconductor Device Manufacturing VI, J. Ruzyllo, T. Hattori, and R. E. Novak, Editors, PV 99-36, p. 461, The Electrochemical Society Proceedings Series, Pennington, NJ (1999)]. Surfactants are found to increase the removal of Si and polymeric particles from silicon substrates. This is attributed to the elimination of hydrophobic attraction forces. In addition, the surfactant is also successful in preventing the particles from redepositing during the final rinse treatment, because during the rinse, both the particle and the substrate have the same surface charge. (C) 2001 The Electrochemical Society.

A basic understanding of the interactions between a particle and the wafer substrate during cleaning is presented and the usefulness of surfactants to improve the overall particle performance for dHF-cleaning mixtures is described.

The corrosion of aluminum during isopropylalcohol (IPA)-vapor drying and during Marangoni drying is investigated and specifications for the cleanliness of the IPA used during both drying processes are determined. Therefore, a controlled contamination with some important contaminants which can be present in the IPA (i.e., water, n-propanol and the corrosive anions Cl-, F-, SO42- and PO43-) is performed. The amount of corrosion is determined electrically by measuring the resistance of thin aluminum lines.When ultra-pure IPA is used, both an IPA-vapor dry and a Marangoni dry step are found to have no effect. When water is added to the IPA, no corrosion is detected even when water concentrations as high as 13% are spiked to the IPA. As far as the corrosive anions that were evaluated are concerned, only chloride is found to be corrosive. Corrosion during IPA-vapor drying is observed at lower Cl- concentrations compared to Marangoni drying. Moreover, water is found to inhibit the corrosion caused by exposure of aluminum lines to chloride containing IPA-vapors.

The H2O2 in NH4OH/H2O2/H2O cleaning mixture is known to be subject to decomposition [1]. This decomposition not only results in a gradual change of the composition of the cleaning mixture but has also been correlated with degradation of the silicon wafer surface integrity, particularly if made hydrophobic by an HF treatment [1-5]. Several experimental studies on the kinetics of the decomposition of H2O2 in SCl-like cleaning solutions have been performed [3-6].

In this study the impact of temperature and metal contamination on the stability of hydrogen peroxide in the two most common wet chemical cleaning mixtures for wafer process operations has been investigated. The stability of the caustic mixture (NH4OH/H2O2/H2O) was found to be very sensitive to certain metallic contaminations in the sub-ppb range, while the stability of the acid mixture (HCl/H2O2/H2O) is mainly influenced by non metallic, anionic components of the solution itself. We observed also a strong oscillating behaviour of the rate of the oxygen gas evolution caused by the decomposition of H2O2. Furthermore it was found, that the oxygen gas bubbles, formed by the decomposition of hydrogen peroxide cause a certain kind of micro-roughness on the silicon surface through a micro masking mechanism. In a series of experiments we could prove that this kind of surface roughness has a significant impact on the integrity of thin gate oxides.