Ultra Clean Processing of Semiconductor Surfaces XI

Paper Title Page

Authors: Quoc Toan Le, E. Kesters, T. Conard, H. Struyf, S. De Gendt
Abstract: In back-end of line (BEOL), the use of fluorocarbon-containing plasmas such as CF4 and C4F8 for patterning of low-k dielectrics with k-value ~2.3-2.5 can result in the presence of a highly fluorinated layer, deposited on the sidewalls and bottom of the trenches [1,. This polymer layer must be removed prior to subsequent processing steps to achieve good adhesion and coverage of materials (metals) deposited in the etched features. However, it is known that this type of fluorocarbon polymer is chemically inert to many existing wet clean solutions, including aqueous solutions such as fluoride ion-containing or highly alkaline solutions, and solvent mixtures [2]. Exposure of the polymer to UV irradiation (λ 200 nm) with doses 3 J/cm2 significantly modifies the polymer film, which results in substantial removal ability in a subsequent wet clean process. Polymer film modification was shown to be efficient either by using a narrow band single wavelength source with λ = 254 nm [ or by a broad band UV source with λ~200-300 nm [.
Authors: William R. Gemmill, Els Kesters, Quoc Toan Le
Abstract: Back end of the line processing requires removal of deposited polymers resulting from etch processes. These polymers typically exist on the whole of the pattern including the dielectric sidewalls and can be removed by wet cleans or a combination of wet cleans and plasma treatments. When a porous dielectric is present these residues cannot be efficiently removed using plasma or certain wet cleans without potentially damaging the underlying porous dielectric layer. Therefore there exists a need for a one-step wet clean that can completely remove the residues without damaging the porous dielectric. Previous work has shown that a combination of a UV pretreatment followed by a wet clean can remove these residues [1]. These residues are composed of CF, -CF2, and CF3 groups as described by X-ray photoelectron spectroscopy (XPS). In an effort to improve the manufacturing viability of such a process we have undertaken a study to develop a one-step wet clean for fluoropolymer removal. Utilizing a blanket checkerboard pattern with a model fluoropolymer deposited on a porous low-κ substrate we have demonstrated the one-step wet clean of the aforementioned fluoropolymer while maintaining compatibility with the pristine and etch processed porous low-k dielectric.
Authors: Roy Te Brake, D. Martin Knotter, Miranda Leenders–van Hees
Abstract: Fluoride contamination on aluminum bond pads is a major issue in the reliability of semiconductor products [. Fluoride contamination can lead to bond pad corrosion, bond pad staining [, (see Figure 1) and bad wire bond quality [. In the ITRS a maximum allowable fluoride concentration in the aluminum wafer process environment is listed [. However, it is not only the air concentration that determines the criticality, but also the exposure time, humidity, and if there is a fluoride source available that can maintain an increased air concentration of fluoride. Therefore, it is more valuable to know at what surface concentrations fluoride contamination on aluminum surfaces become an issue.
Authors: Emanuel I. Cooper, Rekha Rajaram, Makonnen Payne, Steven Lippy
Abstract: Titanium nitride (TiN) is widely used as a hard mask film protecting the inter-level dielectric (ILD) before metal or plating seed layer deposition steps. It is common practice to use a wet etch in order to remove residues formed during the ILD dry-etch step, and at the same time to remove some or all of the exposed TiN. From its thermochemical properties, it might be predicted that wet etching of TiN should be easy, since it is quite unstable with respect to both plain and oxidative hydrolysis. For example, in acidic solutions at 25°C [1, :
Authors: Y. Sun, J. Swerts, P. Verdonck, A. Maheshwari, J.L. Prado, S. de Feyter, S. Armini
Abstract: Self-assembled monolayers (SAMs) deposition is being recently explored to help sealing the pores of a k=2.0 material. In order to enable a covalent chemical low-k surface functionalization by SAMs, a hydroxyl groups density as high as 1 to 2.5 OH groups/nm2 is required. This surface modification must be carefully controlled to confine the k below 10%. In this paper, the effects of plasma temperature, time and power on the SAMs deposition and plasma-induced damage are investigated. The main findings are that there is always a trade-off between surface hydroxyl groups density and bulk damage. A thick modified layer allows the SAM molecules to penetrate inside the pores which results in a decreased porosity and an increased k value with respect to correspondent plasma-treated pristine substrates.
Authors: Samuel Suhard, Ihsan Simms, Ian H. Brown, Mizota Shogo, Kagawa Koji, Martine Claes, Thibault Buisson, Anne Jourdain, Gerald Beyer, Stefan De Gendt
Abstract: 3D stacked IC (3D-SIC) is one of the main approaches within 3D technologies as it is the most mature and economically viable technology and provides the highest through silicon via density. Enabling 3D SIC requires the modification of standard IC process flows by adding several process modules, such as through silicon via (TSV), wafer bonding and thinning, prior backside processing (e.g. Cu nails exposure), micro bumps formation for the front side and/or the backside and finally stacking to another chips [. Within the 3D-SIC technology developed at IMEC, surface control and preparation by means of wet clean and wet etch are essential steps notably in two key process modules: wafer thinning and micro bump formation.
Authors: Cass Shang, Taishih Maw, Fadi Coder
Abstract: In order to achieve high cleaning efficiency requirement for post Chemical Mechanical Polish (CMP) cleaning in Through Silicon Via (TSV) application due to the aggressive CMP process. More comprehensive wafer defect evaluation techniques are needed to understand the cleaning mechanisms and assist the formulation design process. In this paper, the CSX-T series chemistry is applied to the post CMP cleaning process of various wafer substrates commonly used in TSV integration schemes. The data collected by several techniques are analyzed in detail and compared to demonstrate how and when it can be used in new formulation screening process to ensure good cleaning performance.
Authors: Robert Mettin, P.E. Frommhold, X. Xi, F. Cegla, H. Okorn-Schmidt, A. Lippert, F. Holsteyns
Abstract: Applications of acoustic cavitation [ frequently suffer from certain random aspects (e.g., stochastic bubble nucleation events) as well as from its sensitivity to external parameters (like gas content in the liquid). This renders for example a prediction of bubble distributions in size and space still a difficult task. To improve this situation by a better understanding of the fundamentals, a "bottom-up" approach has recently been followed which tried to model collective bubble phenomena and bubble structures on the basis of single bubbles and their interaction [. If the behavior of individual bubbles can be well captured by the models, it is hoped to gain significant insight into a larger system of acoustically driven bubbles. Indeed, several aspects of multi-bubble systems and structures could be explained by single bubble dynamics, for instance by the inversion of the primary Bjerknes force in strong ultrasonic fields. Nevertheless, many details of bubble dynamics stay partly unclear, and considerable efforts are undertaken to improve our understanding and to optimize applications of acoustic bubbles.
Authors: F. Reuter, Robert Mettin, A. Lippert, F. Holsteyns, H. Okorn-Schmidt
Abstract: Ultrasonic cleaning is a well proven technique in many industrial, laboratory and even household applications. It is known that cavitation bubbles can induce fast microscale flows and thus are responsible for cleaning and even corrosion [1,2]. Nevertheless there are numerous effects that can have a potential role in cleaning processes, as the behavior of an acoustic bubble is very complex: radial oscillations, surface oscillations, leading sometimes to the disintegration of a bubble, collapses, rebounds and subsequently shockwaves, liquid jets and vortex flows can be observed. But as bubbles in sound fields typically appear in a random fashion and in complicated interactions, it is very hard to identify the processes and their effects with respect to cleaning. To isolate the various ongoing processes and to study them in detail, single cavitation bubbles and their interaction with a surface are examined in this work. The single bubbles are of sizes around 500 μm in radius and are produced by a pulsed laser that is focused into water, which allows creating bubbles of a repeatable size at a defined position.
Authors: Bong Kyun Kang, Ji Hyun Jeong, Min Su Kim, Hong Seong Sohn, Ahmed A. Busnaina, Jin Goo Park
Abstract: As the semiconductor manufacturing technology for ultra-high integration devices continue to shrink beyond 32 nm, stringent measures have to be taken to get damage free patterns during the cleaning process. The patterns are no longer cleaned with the megasonic (MS) irradiation in the advanced device node because of severe pattern damages caused by cleaning. Recently, several investigations are carried out to control the cavitation effects of megasonic to reduce the pattern damages. The mechanism of damage caused by an unstable acoustic bubble motion was mainly attributed to the high sound pressure associated with violent bubble collapse [1]. In order to characterize the dominant factors affecting the cavitation, MS cleaning was conducted with various dissolved gas concentrations in water. It was reported that the cavitation phenomena relating to particle removal efficiency (PRE) and pattern damage were considerably changed with the addition of a specific gas [2]. This changing behavior may be due to the difference in the physical properties of dissolved gases associated with acoustic bubble growth rate as a function of their concentration. In particular, cavitation effects induced during MS cleaning was controlled by adjusting the acoustic bubble growth rate. Also the change of bubble growth rate is well explained by both rectified diffusion for single bubble and bubble coalescence for multi-bubble, respectively. Similarly, it is well-known that surface active solute (SAS) in the ultrasound field plays an important role in controlling the cavitation effects. A detailed explanation of the acoustic bubble growth rate, cavitation threshold and their relationship with various types of SAS and concentration of biomedical and chemical reactions perspective have been reported elsewhere [3,4]. Their studies demonstrated that the change of cavitation effects depends not only on the chain length of alcohol in the solution but also on the physical properties such as surface tension and viscosity of SAS solutions.

Showing 31 to 40 of 75 Paper Titles