The following is a list of the abstracts for papers which will be presented in SECOND INTERNATIONAL SYMPOSIUM ON PARTICLES ON SURFACES: DETECTION, ADHESION AND REMOVAL. The listing is alphabetical by presenting author. This list is updated continually to add abstracts as they become available and make appropriate corrections. This list may be conveniently searched by using the editor provided with most popular browsers (e.g. Microsoft Explorer, Netscape, ... etc.)
Resuspension of Particles in Turbulent Flows
Particle removal and re-entrainment in turbulent channel flows are studied. The instantaneous fluid velocity field is generated by the direct numerical simulation (DNS) of the Navier-Stokes equation via a pseudo-spectral method. Particle removal mechanisms in turbulent channel flows are examined and the effects of hydrodynamic forces, torques and the near-wall coherent vorticity are discussed. The particle resuspension rates are evaluated, and the results are compared with the model of Reeks. The particle equation of motion used includes the hydrodynamic, the Brownian, the shear-induced lift and the gravitational forces. An ensemble of 8192 particles is used for particle resuspension and the subsequent trajectory analyses. It is found that large size particles move away roughly perpendicular to the wall due to the action of the lift force. Small particles, however, follow the upward flows formed by the near wall eddies in the low speed streak regions. Thus, turbulent near wall vortical structures play an important role on small particle resuspension, while the lift is an important factor for re-entrainment of large particles. The simulation results suggests that small particles (with 0.023) primarily move away from the wall in the low speed streaks, while larger particles (with (780) are mostly removed in the high speed streaks.
Comparison of Various Wavelengths for Laser Assisted Particle Removal
(Abstract not yet available)
Analysis of Particles on Surfaces
the purpose of this paper is to direct the attention of contamination ane die defects caused by surface particles that are introduced from people, tools and other processes.
Particles on surfaces of tools, product (wafers), parts and other materials can be isolated using analytical equipment to determine probable causes of defects and yield losses. By using optical microscopes, boroscope with video tapes, black inspection wipers for sample collecting of contamination, Dryden QIII, SEM/EDX spectroscope, FTIR spectroscope, chemical analysis Air measuring laser sensors, one can identify particle contamination inside tools, on parts before installing in tools, reticles, robotic equipment, stackers, gloves and other items. Working with other analytical equipment such as TENCOR, JOEL and KLA to review product issues we can determine what type od defect and where they are located and by matching this with the process tool we can identify the probable cause of the defec
Sami B. Awad; Crest Ultrasonics; Trenton, NJ 08628
Ultrasonic Cleaning and Particle Removal
(Abstract not yet available)
A.S. Geller; Engineering Sciences Center, Sandia National Laboratories, Albuquerque, NM 87185
Particle Removal from Semiconductor Substrates Using the Plasmax Technology
Control of particulate contamination on semiconductor substrates becomes more difficult as the critical defect size shrinks with shrinking device sizes. PLASMAX, a vacuum based dry cleaning technology, uses a targeted combination of plasma, vibration, and gas flow forces to overcome particle binding energy to a surface and effect particle removal. Dynamic particle trajectory simulation was carried out to develop greater understanding of the particle adhesion mechanism as well as to advance optimization of the PLASMAX apparatus. Particle to surface separation values are calculated starting from equilibrium for a particle bound in a Lennard-Jones potential well. A sinusoidal motion is then imposed on the surface resulting in displacement of the bound particle. The model predicts separation values for the particle as it either oscillates in the well or is ejected from the surface. Predicted values of critical acceleration versus particle size are correlated with experimental cleaning efficiency data.
1) Department of Metallurgical and Materials Engineering, Michigan Technological University, Houghton, MI 49931, USA
2) Dura Pharmaceuticals, 7475 Lusk Boulevard, San Diego, CA 92121, USA
Measurements of Adhesion Forces in Pharmaceutical Powder-Polymer Systems by Atomic Force Microscopy
Atomic force microscopy (AFM) has been adapted to make direct measurements of the adhesion forces between pharmaceutical powder particles and polymeric surfaces. The measured adhesion (pull-off) force values are used to aid in the selection of materials for clinical usage. The AFM method is advantageous in the duality of information that can be provided. AFM can provide direct measures of the pull-off force required to remove a single powder particle from a polymer surface. The AFM is also capable of mapping the surface topography with a resolution on the nanometer scale. AFM can also be used for particle characterization since the pharmaceutical powder particles are irregular in shape. The study undertaken in this program demonstrates the difficulty and obstacles to be overcome when adapting the AFM technique to substrates with rough surface. It is established in this work that surface roughness of both powder particles and polymeric substrates has strong effect on a distribution of adhesion force values.
New Drying Method for Semiconductor Surfaces
A new method for drying critical surfaces is proposed. This method relies on water displacement from a surface by a stable interface formed by water and a non-aqueous, volatile fluid having a specific gravity greater than one. An important condition for maintaining a stable interface is that the mutual solubilities of water and the non-aqueous fluid be negligible. The drying mechanism involves slowly passing the two fluid interface in an upward direction across the surface plane positioned perpendicularly. Said interface is thermodynamically stable and as a result, capable of displacing aqueous films. The effectiveness of the proposed drying method is dependent on factors such us interfacial tensions among surface, water and the non-aqueous fluid as well as the speed of the interface across the surface. Whenever certain surface tension conditions are met, it is feasible to utilize the proposed technique for drying surfaces in applications where water evaporation is detrimental. This method can be used practically by designing a specially shaped container in which the drying action takes place so that the displaced water and the non-aqueous fluid are easily separated (by gravity) and collected for recycling and future use. Preferably, the non-aqueous fluid is highly volatile and free of detrimental contamination so that the final step occurs by simply evaporating the residual non-aqueous fluid from the surfaces.
The Removal of Submicron Alumina Particles from Post-CMP Silicon Wafers
Post-CMP cleaning is one of the most challenging cleaning applications in semiconductor manufacturing. Contact cleaning techniques (such as brush scrubbers) are effective and widely used in the industry. Current contact post-CMP cleans may cause scratching if not properly used. Noncontact cleaning techniques are widely used in FEOL applications but have not been widely used in post-CMP cleaning applications. There is a need for an effective post-CMP clean. The removal of submicron alumina particles (used in Chemical Mechanical Polishing slurry formulation) is accomplished using two different removal techniques. Alumina particles were deposited onto silicon oxide wafer by dipping the wafer in the CMP slurry and then spin-dried. Megasonic cleaning and brush cleaning techniques were used to remove the slurry residue. The removal of each technique (contact and non-contact) was evaluated as a function of the process parameters. The removal efficiency is a strong function of the cleaning parameters such temperature, cleaning time and megasonic power in the case of megasonic clean, and brush pressure, rotation speed and cleaning time in the case of brush cleaning. The results show that both cleaning techniques can be effective in removing slurry residue under certain conditions. Longer cleaning time and higher pressure had adverse effects on contact cleaning. Similarly, longer time and higher megasonic power had adverse effects on noncontact cleaning
The effect of Moisture on the Adhesion and Removal of Silica Particles on Silicon substrates
Particulate contamination has a detrimental effect on yield and product reliability. The adhesion forces of these particles are affected by the polishing, etching, rinsing and drying process that follow the particle deposition. In this study, silica particles were deposited onto thermal oxide silicon wafers and aged in different relative humidity. The results show that moisture has a significant effect on adhesion of silica particles on silicon oxide substrates. This is due to the formation of hydrogen bonds. This study introduces recent results that show the effect of high humidity aging on the adhesion and removal of silica particles from silicon wafers. Samples with deposited silica particles were aged in 45% and 95% relative humidity for one to six weeks and then they were removed. Results show that removal efficiency steadily declines with the increase of aging time. The experiments also show that particle contact radius increased dramatically to a size equal to the particle size. The formation of the adhesion bond which was measured to be more than two orders of magnitude more than the van der Waals force is induced by diffusion and reaction that took place during high humidity aging.
Thermoelastic Modeling for Submicron Particle Removal with Lasers
The removal of particles from elastic substrates has been an important practical problem in the electronics industry especially as the sizes of electronic units shrink. In recent years, there has been an interest in removing $submicron$ level particles from surfaces. The use of traditional surface cleaning methods, such as ultrasonically induced fluid flow, vibrational methods, centrifugal techniques, is limited to particles that require surface acceleration lower than 10^7 m/s^2. For the effective removal of sub-micron particles, a higher level surface acceleration is needed since the adhesion forces (mainly van der Waals force for dry surfaces) are related to the particle size and increase approximately linearly as the characteristic radius of small particles that are to be removed decreases. In current work, based on the generalized dynamic theory of thermoelasticity reported, a transfer matrix formulation including the second sound effect is developed for a thermoelastic layer. The transfer matrix for axisysmmetric wave propagation in a thermoelastic layer is obtained by adopting a double integral transform approach. The second sound effect is included to eliminate the thermal wave travelling with infinite velocity as predicted by the diffusion heat transfer model, and, consequently, the immediate arrival of waves. Using the current formulation and the periodic systems framework, a transfer function formulation for calculating the accelerations is developed for transient analysis. A double integral transform inversion method is used for transient response calculations. Acceleration levels, sufficient for submicron particle removal, are reported. Various processes may be responsible of surface acceleration components and particle removal, such as thermoelastic stresses, surface evaporation, and optical breakdown. In current work, only the surface acceleration due to transient thermoelastic wave propagation is under investigation.
The Use of High Frequency Ultrasonics for Effective Particle Removal from Solid Surfaces
(Abstract not yet available)
Particle Removal after Chemical-Mechanical Planarization
chemical-mechanical polishing (or planarization), CMP, removes unwanted material to make wafers and disks more smooth in various stages of their processing. CMP is the most rapidly growing process technology in integrated circuit manufacturing, but slurry particles not removed can cause serious defects in the product. Removal is done with a rinse, generally deionized water, and some mechanical action, generally contact with a rotating foam roller brush. The pressure at the contact will depend on the elastic modulus of the foam, which will depend, in turn, on the elastic modulus of the bulk material from which it is made., E', and the porosity (ie. The void fraction). There are two important length scales to consider in the fluid mechanics of the post CMP cleaning operation. One length scale is microns, roughly the size od the particles being removed and of the hydrodynamic boundary layer formed by the motion of product, fluid and roller brush. The other scale is centimeters, roughly the size of the procuct itself, where the "large scale" motion of the liquid is important for thoroughness of slurry removal. Brush-particle contact, fluid viscous shear and fluid impact pressure (proportional to v2) are likely to be the major contributors to slurry particle dislodgement. Prevention of backmixing of the slurry and the rinse water is desirable. These design issues and results presented elsewhere are discussed here.
1) Boehringer-Ingelheim Pharmaceuticals, 900 Ridgebury, Rd, Ridgefield, CT 06877-0368
2) Department of Chemical Engineering, Auburn University, Auburn, AL 36849-5127
Direct Adhesion Measurements Between Pharmaceutical Materials
The performance of a number of pharmaceutical products depends upon the dispersion of powders and the adhesion of powder to package materials or to delivery devices. Earlier work has discussed the implications for adhesion between relevant materials based upon various surface thermodynamic models of adhesion. Such work indicates that variable surface contamination is in significant part responsible for observed product variation. Little attention, however, has been given to the direct measurement of adhesion between relevant components. In this work the application of scanning probe microscopy and the colloidal probe method to the study of adhesion between pharmaceutical powders and other surfaces is discussed as are the implications for the use of surface engineering to achieve enhanced performance.
Simulation of Particle Adhesion
(Abstract not available yet)
Development of an In-process Particle Deposition Meter
During the last ten years increasingly stringent requirements have been set for the cleanliness of the production environment for contamination sensitive products. This is not only the case in the microelectronics industry, but also in the field of biotechnology and increasingly in the development of microelectromechanical systems. Not only the monitoring of airborne particles in the production environment is necessary for the correct manufacture of the products, but particularly the analysis of the deposited particles on the surfaces. Most of the measurement technologies for particle detection on surfaces are based on optical methods. All of them have the disadvantage that the particles have to be collected on product-similar surfaces in the vicinity of the product during manufacture. Afterwards they are analysed off-line. Normally the information concerning contaminating particles is available too late.
A measurement technology has been developed providing in situ and online information about deposited particles on a surface. Of interest are the number of particles, the differences between the number of particles per area in specific time intervals (particle flux density) and the particle diameter. Measurement technologies detecting the particle scattered light lead to fast non-contact particle detection and characterisation. The most important innovation is the use of a transparent material as deposition surface, which makes it possible to illuminate the particles and to detect the scattered light by the particles from the back of the surface. This makes it possible to detect in situ the deposited particles without disturbing the particle deposition. The basics, the design, the constructed prototype as well as its performance will be presented.
1) Department of Metallurgical Engineering, University of Utah, Salt Lake city, UT 84112
2) Department of Chemical Engineering, University of Florida, Gainesville, FL 32611
Removal of Particles from Surfaces Using Surfactants
Post-CMP (Chemical-Mechanical Planarization) cleaning of silicon wafer products is of critical importance to the microelectronics industry. Removal of residual CMP polishing particles must be effective to prevent circuit failure and poor wafer yields. Surfactants have been shown to enhance dramatically the removal of residual particles following the CMP process. Basic principles and issues associated with the use of surfactants in enhancing particle removal will be discussed. Specific results for surfactant-enhanced particle removal from silicon, tungsten, and copper surfaces will be presented.
Cleaning, Rinsing and Drying Aspects in Post Cu Cmp Cleaning
In this work, the removal efficiency of slurry particles was investigated after Chemical Mechanical Polishing (CMP) of Cu interconnection lines used in IC manufacturing. Inspection of patterned Cu wafers after post CMP cleaning and drying in a spraytool showed preferential deposition of slurry particles onto different substrate materials (oxide, Cu) depending upon the nature of the slurry particles and the cleaning chemistry used during the post CMP cleaning step. In this work, this will be explained in terms of the electrostatic interactions during the cleaning step, and the importance of the subsequent rinsing and drying procedures will be demonstrated. More in detail, the alumina based slurry was investigated in view of a HF based post-CMP cleaning step. In this particular case, it was found that the slurry particles preferentially deposited onto the oxide fields, but that after a subsequent HF processing they ended up on the Cu fields. Contact angle measurements on the Cu lines showed that after this HF based processing, the Cu surface was hydrophobic. This indicates that the migration of alumina particles from the oxide fields to the Cu fields is due to the drying step, resulting from the mixed hydrophilic/hydrophobic surface on the patterned wafers.
Contamination Reduction in Loadlocks Through Pressure Control
Increased deposition of particles in loadlocks and reactor chambers undergoing purge or vent processes has long been observed, and theories to predict the rate of particle generation based on the rate of pressure change have been developed. However, pressure changes can lead to wafer contamination through pressure driven particle deposition as well as creation of additional particles. This presentation will describe a model for particle transport in a loadlock or reactor chamber with particles subject to gravitational and thermophoretic forces where the thermophoretic force is derived from the heating or cooling of the gas due to pressure changes. This model is solved both by means of a numerical solution of the complete governing equations and an analytic solution of a simplified model. The analytic solution provides insight into the effect of several parameters on the deposition of particles. Factors explored include the rate of vent or purge, the distance from the wafer to the loadlock surface, and type of gas used in the loadlock. Finally, suggestions for designing loadlocks with reduced wafer contamination are presented.
Study of Argon Aerosol Cleaning Method for Post-CMP Cleaning Application
Particle contamination on the wafer surfaces during processing is a major factor of device failure. It becomes more serious as the device size shrinks. One of the major sources of particles in a fabrication facility is CMP process that is widely used in the semiconductor industry for planarization. Liquid phase cleaning process is typical methods to clean residual slurry particles from the wafer surface after CMP process. However, it is limited to select an effective cleaning solution when metal layers are appeared on a wafer due to the corrosive nature of chemicals. In this study, we have evaluated an Ar aerosol cleaning method for post CMP cleaning. We have also compared the particle removal effect of Ar aerosol cleaning to conventional brush, SC1 and diluted HF cleaning processes applying post CMP cleaning step after an oxide layer is polished using an alkaline based silica slurry. We have studied the removal efficiencies of SiO2 slurry particles by the Ar aerosol for various process parameters. It has been confirmed that the Ar aerosol cleaning is more efficient than brush or conventional wet cleaning achieving more than 99% removal of particles when the flow ratio of Ar/N2 is 10 and the accelerating N2 flow rate is 90 l/min. We have also found that the Ar aerosol cleaning removes most of defects from the surface of sub-quarter micron devices. It is concluded that the Ar aerosol cleaning is applicable to post metal CMP process and is more effective than conventional cleaning processes.
Contamination Control Challenges for ULSI Manufacturing
As semiconductor devices continue to be highly integrated and their geometries shrink while both die and wafer sizes grow, not only particulate contaminants but also metallic and molecular (or organic) contaminants adsorbing onto the surface of silicon wafers will have an increasing detrimental impact on both the performance and yield of the semiconductor products. Ultraclean requirements for next and future generation ULSI manufacturing will be further tightened. This presentation will focus on detection and elimination of particulate, metallic, and organic contaminants on the wafers, with emphasis on the importance of rigorous wafer cleaning and preferable total contamination prevention systems development. Some paradigm shift in semiconductor manufacturing will also be discussed, which will create farther challenges and opportunities in contamination control engineering.
Adhesion of Charged Particles
Most particles in the form of dust or powder tend to be insulating. Furthermore, the particles typically acquire a charge through the natural phenomenon of triboelectricity. The degree of charging depends on the amount of rubbing or contact between the particle and other materials in transport or mixing processes. In the electrophotographic process utilized for copying or printing documents, the particles are deliberately charged to provide an electrostatic force for the imaging process. The purpose of this paper is to review the recent literature concerned with the adhesion of charged toner particles used in the electrophotographic process. A particular focus is placed on the evidence for the role of a nonuniform surface charge distribution in enhancing the electrostatic adhesion. Both centrifuge and electric field detachment measurements are reviewed along with model descriptions for the adhesion of charged particles that are either isolated or within a layer on a substrate. The collective literature suggests that the adhesion of triboelectrically charged toner particles as used in electrophotography is significantly enhanced by a nonuniform surface charge distribution.
Particle Removal and Metallic Contamination Control on CMP Oxide Film Surface Using Auriga EC System
In Integrated Circuit device manufacturing, Chemical Mechanical Planarization (CMP) is one of the most common processes used to planarize thin films. Slurries used in the CMP process consist of fine abrasive particles, which remove material from the surface of the substrate by chemical and mechanical action. The slurry-contaminated wafer is then cleaned in a following step to eliminate slurry residues. Cleaning is accomplished on the Auriga EC integrated tool using a double- sided brush scrubbing mechanism. Residual contaminants are removed by PVA brushes and dilute ammonium hydroxide solution. Generating like charges on the PVA brushes and slurry particles during the cleaning process results in repulsion that removes the vast majority of particles adhered to the wafer surface. Particles remaining after this initial cleaning step typically include metallic ions such as Fe, Mn, Zn, Cr, Ni, Ca, and K that originate from the CMP process consumables (i.e. slurry and polishing pad). These metallic ions may diffuse into or become embedded in the planarized film. These subsurface ionic contaminants are difficult to remove by scrubbing, but must be removed to ensure proper device performance and yield.
Dilute hydrofluoric acid (HF) scrub etching was used to remove the embedded metallic ions from the damaged sublayer of the substrate by removing a controlled amount of surface material. Metallic ion contamination levels were measured before and after the HF treatment, and showed a significant reduction in the post-treatment contamination levels.
The Auriga EC dry in/ dry out CMP system has been designed with HF-etching capability; this is an enhancement over the existing Auriga C tool design. Blanket 200mm TEOS wafers were processed through the Auriga EC with and without HF post-CMP cleaning; defectivity and metallic contamination data were compared for both groups. The results confirmed that the use of HF cleaning, made possible by upgrades to the equipment design, materials, and cleaning process of the Auriga EC, lowered the metallic ion contamination of the film surface. The data will be presented at the conference.
1. Entropic Systems, Inc., Woburn, MA 01801
2. The Boeing Company, Rocketdyne Propulsion & Power, Canoga Park, CA 91309
Reduction of Cleaning and Verification Times for an Aqueous Based Process by On-line Monitoring
A comprehensive study of on-line monitoring of an aqueous-based cleaning procedure has been performed. The goal of this study was to reduce of the cleaning, rinsing, and cleanliness verification times of complex rocket engine components. Commercially available in-line particle counters have been shown to accurately detect the presence or absence of particles of widely varying shapes (spherical particles, rigid fibers, and flexible fibers). Their ability to size these particles has been shown to be primarily a function of the particle cross-sectional area, with spherical particles being sized more accurately than elongated particles. A non-isokinetic sampling methodology has been developed to maximize the capture of suspended particles from a bulk cleaning circulation flow.
The rate of cleaning of previously contaminated surfaces immersed in a circulated cleaning solution was accurately monitored by In-situ particle counter analysis of a continuously sampled stream of this cleaning solution. The subsequent rinsing of residual ionic detergent from these surfaces by deionized water was accurately monitored by an in-line conductivity meter.
By using these on-line instruments, the total time required to clean, rinse, and verify the cleanliness of previously contaminated surfaces was repeatedly shown to be less than 1 hour, an order of magnitude reduction in the time previously thought to be necessary to achieve verifiably clean components. Although the focus of this study was the cleaning of component parts, the on-line monitoring methodologies developed should also be applicable to the cleaning-in-place (CIP) of complex process piping systems.
Spray Cleaning with Hydrofluorocarbon Solutions
Spray cleaning with an inert liquid (such as a hydrofluorocarbon or hydrofluoroether) is shown to be an effective means of removing contaminating particles from the surfaces of objects. This type of cleaning has produced surface cleanliness levels of better than class 200. It was also determined that liquid spraying with these fluids provides a suitable replacement for CFC-113 as a cleaning medium.
A variety of substrates, including silicon wafers, coated gold and aluminum coated mirrors, and zinc sulfide windows for FTIR (Fourier Transform Infra Red) spectroscopy test cells were spray cleaned to assess materials compatibility. No apparent deleterious effects were observed.
Recent Advances in Silicon Wafers' Surface Preparation for IC Manufacturing
(Abstract not yet available)
Adhesion of Small Particles and Methods for Their Removal
Submicron particles are a leading cause of failure of components and end products in widely diverse industries. These particles adhere very strongly to the surface and cannot be easily removed. Following a brief review of the adhesion of particles, methods for their removal will be described. In addition to established removal techniques such as aqueous cleaning, several non-aqueous precision cleaning techniques have been developed for effectively removing small particles and/or contaminant films. Many of these methods have been developed in response to the inherent limitations of aqueous cleaning for removing very small particles. These newer removal methods include plasma and carbon dioxide cleaning, supercritical fluid cleaning, ion beam etching and laser cleaning. Other innovative techniques, such as microcluster beams and supersonic jets, are in pre-commercialization or research and development stages. The advantages and limitations of these techniques for removal of small particles will be discussed.
Influence of Temperature and Dissolved Air on Megasonic Particle Removal
Abstract: An experimental study has been conducted to determine the influence of temperature, dissolved gas content, megasonic power and cleaning time on local and total particle removal efficiencies from silicon wafers. Results show that all four of these parameters are statistically significant when correlating particle removal. Cross correlations are also provided. Local cleaning efficiency data show variations on the wafer surface and the influence of local sound power intensity. Wafer surfaces are damaged under some combinations of operating parameters. Severe cavitation is thought to be the cause.
1) CFM Technologies Inc., 1336 Enterprise Dr., West Chester, PA 19380
2) Applied Materials, 3050 Bowers Avenue, MS 0104 Santa Clara, CA 95054
The Enhancement of Cleaning Particles by Adding Valtron SP2200 into SC1 Solution
SC1 (the mix of ID, H2O2 and NH4OH) solution has been used by CFM Full-Flow™ system for wet cleaning for a long time. Recently, it has been found that by adding Valtron SP2200 into the solution can enhance the performance of cleaning particles (with low count-in and low count-out) and also enhance the leaning ability of post photoresist striping. Valtron SP2200 is a water-soluble alkaline detergent with an unique blend of anionic and nonionic surfactants. It is speculated that Valtron SP2200 in SC1 solution helps preventing particles from depositing on wafer surfaces.
A SC1 process in this study is the following sequence, i.e., initial rinse, SC1 injecting, soaking, SC1 displacement by DI, DI rinse and IPA dry. With the design of a single, stationary and closed chamber system, the flow of DI water or chemicals enters the chamber from the bottom. Only the last step of IPA dry is IPA vapor entering from the top and pushing all of DI down to drain. Bare silicon wafers with low LPD counts (low point defect, below 400 at 0.09 µm or higher) or non-implanted photoresists wafers are used for this study.
This paper will present experimental conditions, data and conclusions of this study.
Laser Removal of Particles from Solid Surfaces
(Abstract not yet available)
1) AEG of IDEA, LLC, 3715 Felda Street, Cocoa, Florida 32926
2) NASA M/S MM-G2, KSC, FL 32899
Automated Particle Fallout Monitor
Traditionally, measurement of the particle fallout contamination levels in a facility is accomplished by manually counting particles on a collection plate under a microscope. This process is tedious, time-consuming and prone to human error. This paper presents an automated in-situ approach to these measurements using an Automated Particle Fallout Monitor (APFM). The APFM is a quantitative (it counts particles) particle fallout monitor which measures the size and number of particles that are collected on the surfaces of a room in which the monitor is placed. The APFM can measure particles as small as 5 microns in diameter and calculate their contribution to percent area coverage. The instrument correctly processes irregularly shaped particles as well as fibers and provides a measure of the cleanliness of a room according to Mil Standard 1246. The instrument consists of a processor unit and associated sensor heads. A sensor head collects particles on a sample plate that rotates under the view of CMOS image sensors. The image surface is magnified via optics that yield a resolution such that particles from 5 to 750 microns in size can be resolved. Image data is passed to the processor unit over a USB link. Up to six sensor heads can be attached to each USB port on the processor, thus allowing for multiple sensor heads to be spatially distributed throughout a facility. The processor unit contains a miniaturized Pentium-based computer that processes, analyzes, and archives image and particle data, and controls movement of the sample plates in the sensor heads. The system had been developed under a cooperative agreement with NASA to commercialize a KSC-developed automated fallout monitor that benefits both NASA and private industry.
Klaus Wolke; STEAG Microtech, Carl-Benz strasse 10 D-72124 Pliezhauzen, Germany
The IMEC-Clean: A clean for advanced CMOS manufacturing
The IMEC-Clean concept has been introduced in 1993  and investigated in more detail over the past years [2,3,4,5]. This cleaning concept is based on a two step cleaning approach. The first step uses an oxidant to remove the organics from the surface and to form a uniform oxide over the wafer. The second step uses a HF based chemistry to remove the oxide. In this step the particles are lift off and the metals are dissolved . After this step the surface is passivated again in a mixture to grow a clean chemical oxide and the wafers are dried. In this paper we will give an overview of the basic principles how this imec-clean is removing particles. Furthermore data measured using the wet bench in the prototyping line of IMEC will be presented showing the robustness of a process designed from first principles. The clean was introduced at IMEC for the 0.35m CMOS process and its use is succesfully extended in the 180nm and 130nm CMOS modules presently developed at IMEC.
*permanent address: Texas Instruments, Kilby Center, Dallas, Texas 75265, USA
M. Meuris et al., 3rd Int Symp. Cleaning Technol. In Semic. Dev. Manuf., 184th ECS Symp., Oct 1993.
 F. Tardif et al., Proceed. Of UCPSS94, p. 309.
 M. Meuris et al., 23rd Symp. ULSI Ultra Clean Technology, Nov 1994, Japan.
 F. Tardif et al., Proceed. Of UCPSS96, p. 175.
 K. Wolke et al., Proceed. of UCPSS96, p. 199.
1) Department of Physics, University of Konstanz, Fach M676, 78457 Konstanz, Germany
2) Institut für Angewandte Physik, Johannes-Kepler-Universität Linz, 4040 Linz, Austria
3) Instituto de Optica, CSIC, Serrano 121, 28006 Madrid, Spain
Laser Cleaning for Particle Removal from Silicon Wafers
We studied the removal of submicron particles from silicon wafers by the "steam laser cleaning (SLC)" and "dry laser cleaning (DLC)" processes that are currently investigated as new promising cleaning technologies for complementing traditional methods in industrial applications. For SLC a thin liquid layer (e.g. a water-alcohol mixture) is condensed onto the substrate, and is subsequently evaporated by irradiating the surface with a short laser pulse. The DLC process, on the other hand, only relies on the laser pulse, without application of a vapor jet. Using well-characterized monodisperse polymer, silica and alumina particles with diameters down to 60 nm we have systematically investigated the efficiency of the two processes. Additionally the influence of light wavelength and laser pulse duration from the nanosecond to the femtosecond range was studied. Our results demonstrate that for the gentle cleaning of silicon wafers SLC is a very efficient method and is superior to DLC. This is due to lower cleaning thresholds in laser fluence for SLC compared to DLC and the possibility of serious surface damage in DLC caused by field enhancement under the contaminants, an effect that has been only rarely taken into account so far in laser cleaning studies.
Seong-Ho Yoo; Particle Technology Laboratory, Department of Mechanical Engineering, 111 Church Street S.E., Minneapolis, MN 55455
Benjamin Y.H. Liu; Particle Technology Laboratory and Environmental Division, Department of Mechanical Engineering, 130-A Mechanical Engineering, 111 Church Street S.E., Minnapolis, MN 55455
Particle Deposition Technique for Minimization of Wafer Aging Effects for Cleaning Evaluation
A new method of test wafer preparation for wafer-cleaning-system evaluation is presented. The method involves "dry" deposition of process particles on the wafer surface, using a Differential Mobility Analyzer (DMA) based instrument. The deposition process allows precise control of particle number, size and deposition location. Conventional test-wafer preparation techniques such as particle deposition by "dipping" or by a "nebulizer" suffer from the limitation that the particle adhesion to the wafer surface increases quite significantly with time, i.e., the wafer "ages". This causes large variations in the results of cleaning system evaluation experiments performed on different days or under different conditions. The "dry" deposition method overcomes this limitation. Wafer aging is reduced considerably, as indicated by cleaning tests involving silicon nitride and tungsten particles. The DMA-based test-wafer preparation technique represents a significant step toward the standardization of the methodology of cleaning system performance evaluation.
*Arizona State University, Department of Mechanical and Aerospace Engineering, Tempe, AZ 85287-6106
Modeling of Scattering from Wafer Contaminants using Two Hybrid Versions of the Discrete-Dipole Approximation
Evaluation of light scattering from wafer surfaces is a method used extensively by the semiconductor manufacturing industry to detect the presence of wafer contaminants. Differences in light scattering signatures between pre-cleaning and post-cleaning of wafers can be used to determine the success of contaminant removal. Electromagnetic models are tools to predict the light scattering from illuminated surface features, and have been used for semiconductor applications. The discrete-dipole approximation (DDA) is one method used for light scattering modeling. When modeling a system of scattering features which are optically far apart, the full form DDA method can be computationally intensive in CPU time and memory storage. Large scattering systems may become impractical to model by the DDA method. To address this problem, we propose two hybrid versions of the DDA method which reduce computational time and memory requirements while maintaining accuracy. The hybrid methods compute the surface reflected fields between interacting dipoles of separate features by the Reflection Coefficient Approximation rather than the Sommerfeld Integrals used by the full-form DDA. This approximation reduces the CPU time and memory required to compute the dipole-dipole interaction. The two methods are: (1) the segmentation method, which considers dipole-dipole interaction between each dipole of the interacting features; and (2) the collected dipole method, which uses a single "collected" dipole to represent the scattering feature when considering feature-feature reflection interaction.
Determining the Dynamic Efficiency of Cleanroom Wipers for Removal of Liquids and Particles from Surfaces
At least three factors are vitally important in the removal of liquid and particles from a critical surface using a wiper, whether liquid is added deliber-ately to the surface for the purpose of cleaning or whether it is present merely as the result of a spill.
First and foremost is the dynamic efficiency with which a wiper is capable of sorbing liquids. Second, there is the number of particles already present in the spill (or on the surface being wiped) and the extent to which those particles are removed during the wiping process. Third, there is the very real concern regarding particles and fibers which the wiper itself may leave behind on the surface being wiped.
In this paper, new procedures are described for investigating these factors. Data are presented which demonstrate that cleanroom wipers made from fabrics that have an excep-tional ability to "wipe the surface dry" leave the wiped surface cleaner than those which do not, since the residual contamination resulting from a spill lies suspended in the liquid phase left behind on the wiped surface. The over-whelm-ing conclusion from these data is that wipe-dry is not merely a desirable feature in a cleanroom wiping material from a housekeeping point of view, but is a critical feature in wiping up spills of dirty liquids and, by extension, in the removal of particles from surfaces.
The inherent cleanliness of wiping materials is an import-ant characteristic in and of itself because wipers can contribute particles to clean environments in many ways. However, the results suggest that this "static" characteristic is of less importance in selecting wiping materials for removing liquids from surfaces than is the ability of those same wiping materials to wipe surfaces dry.
Santa Ana, CA 92705
The Surface Topography Challenge is Met by the Single Wafer Megasonic
(Abstract not yet available)
Adhesion Measurements to Aid the Formulation of Dry Powder Inhalation
Drug delivery to the lungs is an important approach to the treatment of lung diseases. Interactive powder mixtures containing the drug in micronised form adhered to coarse grade carrier particles provide one method. During inhalation the drug particles must be re-suspended in the respiratory air stream to be able to pass through the lower airways, whereas the coarse carrier particles are usually swallowed. Adhesion forces must be overcome during this resuspension step. If particles are not re-suspended, they will stick to the carrier particles and will be swallowed. However, from resuspension models it is known that there is a complex relationship between resuspension, force of adhesion and particulate properties.
Experiments were carried out to measure the adhesion force between drug and carrier particles in interactive powder mixtures. The properties of the carrier particles and the mixing procedure were varied. Powder resuspension of the mixtures was assessed by determination of the amount of particles detached from the carrier particles and the aerodynamic particle size of the re-suspended drug at an air flow rate of 60 l/min. An increase in median adhesion force resulted in a reduced total amount of re-suspended drug particles, but size and surface roughness of the carrier particles also influenced the aerodynamic particle diameter. A mathematical resuspension model was developed which was found to be able to predict the aerodynamic particle size and the total amount of drug re-suspended from the adhesion forces measured, the size of the carrier particles, and a measured value for the coefficient of static friction.
Removal of Fibrin Coated Particles from Surfaces
Arterial thrombi consist mainly of platelets and fibrin polymer, which is the biopolymer produced in the final step of the coagulation cascade. The objective of this study is to investigate the role of the fibrin polymer network on the resistance of a clot to fragmentation under shear flow. To this end, a fibrin clot model which consists of monodispersed latex beads coated with fibrin polymer that adhere to a glass plate also coated with fibrin polymer is considered as a representative model of native clots. A shear flow chamber was especially designed to determine the adhesion force due to specific fibrin/fibrin interactions by measuring the wall shear stress required to remove the particles from the plate. A complete series of experiments has been performed to investigate three different physiological cases associated to three different surface coated conditions which are a fibrin monolayer, a fibrin bilayer and a fibrin bilayer degraded by a physiological fibrinolytic protein .. A balance of forces and torque on particles, taking into account the elastic behavior of fibrin/fibrin interactions is used to interprete the results in terms of net adhesion force due to fibrin polymer.
Adhesion and Removal of Particles from Charged Surfaces under Humidity Controlled Air Stream
The fouling and the subsequent cleaning of fibrous media are encountered in various industrial and domestic applications. This work focuses on the adhesion and the removal mechanisms of individual mineral particles on a fiber surface of a fibrous medium under air stream. To this end, an experimental apparatus using an air shear-flow cell to estimate the global adhesion force of particles by measuring the wall shear stress required to remove them from a flat plate has been developed. The model mineral particles are spherical monodispersed silica beads of 19 micrometers diameter. The model fiber surface is a conductive plate made of glass covered by a nanometer sized tin oxide layer. In this fundamental study, the curvature of the fiber surface is neglcted, assuming that the particle radius is much smaller than the curvature radius of the surface. The effect of humidity, surface wettability and surface charge on the value of the critical wall shear stress necessary to remove the particles are analyzed. The results are interpreted in terms of global adhesion force using a balance of forces and torque on the particle and adopting the dry friction assumption. The existence of capillary effects and of a lift electrostatic force is discussed.
Total Reflection and Grazing Emission X-ray Fluorescence Spectrometry -An Attempt to Measure the Spatial Distribution of Contaminants on Silicon Wafer Surfaces.
Total Reflection X-ray Fluorescence spectrometry (TXRF) has proven to be a useful tool in the semiconductor industry where it is employed for the determination of metallic contaminants on silicon wafer surfaces. Additionally, we will show that TXRF can be used to yield information about the elevation of particles above the wafer surface, if the incident angle is varied. The technique is based on the existence of a standing wave field arising from the interference between the incident and the reflected primary X-rays. The intensity of the field at a certain distance from the surface varies strongly with the incident angle of the primary beam. Similar effects occur in an inverse geometry called "Grazing Emission X-ray Fluorescence" (GEXRF) spectrometry. An experimental set up for grazing emission measurements will be presented. In contrast to TXRF, which uses energy dispersive detectors and is therefore better suited to the heavier elements, GEXRF is equipped with wavelength-dispersive detection and is aimed at the low-Z elements. In order to increase the intensity and improve the spectral quality of the primary radiation, a Laser Plasma source is projected which is specifically designed as an X-ray source for the Grazing Emission X-ray fluorescence spectrometry of wafer surfaces.
Effect of surfaces Hydrophobicity on Particle Detachment
We have measured interactions between hydrophilic and hydrophobic surfaces in an aqueous medium at various pH and ionic strengths as well as in some organic solvents using atomic force microscopy and analyzed them in terms of particle adhesion and detachment from surfaces. In hydrophilic systems (phi-phi) the forces observed were found to be well described by DLVO theory at large separation distances. Very long-range hydrophobic forces were not observed in hydrophilic-hydrophobic (phi-pho) systems. Nevertheless the jump into contact was found to occur at distances greater that those predicted by just van der Waals attraction. The interaction between two hydrophobic surfaces (pho-pho) was dominated by the long-range attraction due to hydrophobic forces. This interaction was found to be sensitive to the type of substrate as well as to the pH and electrolyte concentration. The pull-off force for pho-pho interactions was found to be an order of magnitude larger than for phi-pho systems which was an order of magnitude larger than for phi-phi systems.
Measured pull-off forces showed poor reproducibility. However, average values showed clear trends and were used to estimate interfacial energies or work of adhesion for all systems studied by means of the Derjaguin approximation. These values were compared to those calculated by the surface tension component theory using the acid-base approach. Good qualitative agreement was obtained giving support for the usefulness of this approach in estimating interfacial energies between surfaces in liquid media.
Hydrodynamic detachment experiments conducted with the different systems showed that particle detachment was much more difficult in pho-pho systems. A comparison of the measured adhesion force with hydrodynamic detachment experiments showed good qualitative agreement.
Surface Energy Characteristics and Adhesion Measurements under Different Mechanical and Environmental Conditions of Toner Particles
The performance of many industrial processes depends highly on the degree of tuning of the adhesion behavior between substrates and particles. In this paper we demonstrate how the surface energy characteristics of toner powder (for xerographic applications) and its adhesion behavior on various substrates, such as silicon and rubber, can be investigated. Surface energy characteristics were investigated by using the toner particles as a stationary phase in inverse gas chromatography (IGC) experiments. As an attempt to simulate real conditions, adhesion between a single toner particle and a surface was measured by atomic force microscopy (AFM) force curve experiments with toner particles attached to AFM cantilevers. A custom data acquisition system and analysis software was used to determine the statistical distribution of pull-off forces of an attached toner particle and a surface as a function of substrate velocity, humidity, temperature, and materials properties of the probe and substrate. The results of this investigation provide an empirical relationship between the parameters of interest and the adhesion of the toner particle.
Robert Sherman; Applied Surface Technologies, 15 Hawthorne Dr., New
Providence, NJ 07974
Recent Developments in CO2 Snow Cleaning
CO2 Snow Cleaning has been established as an environmentally friendly alternative surface cleaning process. CO2 Snow Cleaning can remove particles and organic residues from surfaces non destructively and without residues. Past data has generally established the limit of 0.1 microns as the smallest particle removed and organic contamination removal down to a monolayer. We will present new data to support these trends and also show particle removal down to smaller sizes. The recent trends in CO2 Snow Cleaning have been directed towards application developments along with process control environments, automation and manufacturing integration. We will discuss some recent concepts and trends in these area.
Applying Surface Analysis Techniques for Particle Identification
Surface analysis techniques such as X-ray Photoelectron Spectroscopy (XPS), Auger Electron Spectroscopy (AES) and Time-of-Flight Secondary Ion Mass Spectrometry (TOF SIMS) are routinely used in materials' and process characterization as well as tools for failure analysis. Very often the identification of particles is of great importance in various manufacturing and semiconductor technologies in order to determine the source of the particulate contamination, which lead to product failures. XPS provides chemical and elemental identification of particles, with its major limitation being limited spatial resolution of about 10 microns. Auger electron spectroscopy with its high surface sensitivity and small analysis volume makes it the technique of choice for the characterization of sub-micrometer inorganic particles. In the case of examining small particles in an organic matrix TOF SIMS is the technique of choice.
In today's complex manufacturing processing environment all of these surface analysis techniques are commonly used for solving problems related to particles. Analysis and identification of particles found during manufacturing processing by the application of these surface analysis techniques will be discussed.
Microcluster Source Operation For Precision Cleaning of Surfaces
(Abstract not yet available)
A New Approach for Particle Removal Based on a Monte Carlo Simulation of Electrostatic Interactions
We can propose a paper on a new approach for particle removal which go further the DLVO theory.This theory is based on a Monte Carlo simulation procedure where the particle/substrate system is represented as two infinite and parallel planes with uniform surface charge density. The influence of cleaning parameters such as pH, dielectric constant of media, bath temperature, valence, size and concentration of ions contained in the solution are practically studied under well controlled conditions and compared with simulation data. Results show that the conditions which favor particle removal are high surface charge density, dielectric constant, temperature and ionic strength and a low counter-ion valence and small counter-ion radius.
Influence of Plasma Treatment on the Wettability of Composite Membranes
(Abstract not yet available)
Note on the Effect of Roughness on the Particles' Adhesion and Removal
The adhesion of the particles on the wall is due to the attractive forces between them while the removal of the particles from the wall is related to the perturbation forces which act on the particles. All the removal particles practices point out that the difficulty to remove fine particles increases with time, effect that is known as aging. The aging indicates that the attractive forces between the fine particles and the wall surface surety increase.
All the attractive interactions that determinate the adhesion state between the particles and the wall depends on their distance which is known only when smooth surfaces have been assumed for the particles and the wall. Usually to establish the particles - wall attractive force, smooth particles with smooth surface have assumed.
The use of the SEM allows to determine the morphology of particles of the order of micron but not their micro - roughness. The use of STM has been shown that also apparently smooth surface has a micro - roughness with characteristic dimension of the order of nm. So it is sound to assume that that also fine particles show a similar micro - roughness which has a high effect on the attractive forces.
The micro - roughness of the surface of particles and of the wall, where the particles adhere, imposes to apply new suitable tools which consider the micro - roughness effect on the attractive forces and on the increasing of these forces with time. In this work some tools have been proposed to obtain relationships between roughness and attractive forces, as well as, roughness and aging of the particles.
Advantages of High Power Megasonic Cleaning
(Abstract not yet available)
Particles at 90 Nm : Can Wet Cleaning Still Do the Job ?
The International Technology Roadmap for Semiconductors (ITRS) calls for a need to control the particle density on silicon wafers down to 0.064/cm2 for particles largers than 90 nm for a 180 nm technology. This is being introduced by most major semiconductor manufactureres right now. By the year 2005, we will have to control particles of sizes larger than 50 nm. We have measured particles at 90 nm on silicon wafers for different wet chemistries and characterized the particle removal. Models to explain the experimental results will be presented. We found that improved filtering and the use of surfactants to control 90 nm particles is necessary. Finally, the use of ozonated chemistries for particle control and the characterization of particles after resist strip will be presented.
Mertens and M. Heyns; IMEC, Kapeldreef 75 B-3001 Leuven, Belgium
Klaus Wolke; STEAG Microtech, Carl-Benz strasse 10 D-72124 Pliezhauzen, Germany
Removal of Particles Using Dhf-based Cleaning Recipes
Submicron particles on a wafer surface can have a detrimental effect on the device yield and with shrinking dimensions of IC structures, this effect becomes more and more important. Therefore, there is a growing need to have cleaning mixtures with optimal particle removal performance in order to control the density of these particles on silicon wafers. In this paper, it will be described how dHF-based cleaning recipes have to be optimized in order to have a good particle removal efficiency and at the same time to minimize the (re-)deposition of particles. For this purpose, both the composition of the HF-bath and the following rinse have to be optimized in order to obtain a good removal for all kinds of particles and substrates.
Effect of Surface Roughness on van der Waals and Electrostatic Contributions to Particle Adhesion
Models for calculating the van der Waals and electrostatic interaction energies between two electrically charged particles of arbitrary shape and roughness have been developed. The van der Waals model uses a rigorous integration of the intermolecular pair interaction to calculate the interparticle van der Waals attraction. Retardation effects, which become important at gap widths larger than approximately 10 nm, can also be included. The electrostatic interaction between the particles is solved using a boundary element approach in which the particle surfaces are first discretized using flat, triangular elements. A key feature of this model is the use of a generalized discretization algorithm that allows representing essentially any desired shape. The relative size of the various elements is controlled using an element density function that minimizes the total number of elements required while ensuring that the elements in the near-contact region between the particles are small enough to capture the local particle shape as well as the change in surface properties upon approach of the two particles. In this talk, the effects of three different types of roughness, namely bumps, pits, and surface waves, on the interaction energy profile will be presented.
Identification of Particles for Ink-Jet Cartridge Monitoring
(Abstract not yet available)
Effect of Megasonics coupled with SC-1 Process Parameters on Particle Removal on 300mm Silicon Wafers
The effect of megasonic use, bath temperature, and NH4OH:H2O2:H2O ratio were studied to determine an effective means of particle removal on a bare 300mm silicon wafer. This is one of the first studies to look at this process in the 300mm platform. Two types of megasonics were used in testing, a divergent lens megasonic and a focused beam megasonic. Experimental results show overall removal efficiencies greater than 98% of predeposited nitride particles for both megasonics when coupled with optimized temperature and chemistry. These higher removal efficiencies were found when the use of a dilute SC-1 chemistry was implemented. The methods of removal were studied for each megasonic, and the particle removal efficiencies per bin size analyzed. Particles in the bin size that were added in the deposition process were removed with removal efficiencies greater than 98% as well. The effects of megasonic power, chemistry and bath temperature on surface roughness were also studied and analyzed in house by use of atomic force microscopy. These results are currently being analyzed and may not be presented, due to effects not related to the SC-1 process on surface roughness data.
Curie-Sklodowska Sq. 3, 20-031 Lublin, POLAND
Particle-Particle Interaction Through Detachment Experiments: Relevance to Surface Contamination
To study the stability of a solid-liquid dispersion system we measured the destruction time of the sediment column structure. The method of the measurements is described in our papers published elsewhere and it was presented during the 22nd Annual Meeting of the Adhesion society. The destruction time of the sediment column structure for different size fraction of coal grains was measured in two homologous series of liquids (n-alkane and n-alcohol). We found the destruction time of the coal sediment column structure increased nonlinearly with increasing density and viscosity of the studied liquid. The linear relationship was obtained for the reciprocal of the destruction time as a function of the density of the liquids and the free energy of interaction between two coal grains separated by a thin film of the liquid. By extrapolation of the linear relationship between the reciprocal of the destruction time and cohesion work of the liquid we determined the cohesion work of n-alkane or alcohol in which the coal column sediment structure was stable. The discussion of the results showed that the detachment of the coal grains from one another by disruption of the thin film of the liquid and the changes of the destruction time resulted from equilibrium of the inertial and interfacial forces.
Advanced Laser Technologies for Cleaning Applications
(Abstract not yet available)
1) Xaar Jet AB, Box 516, S-17526 Jarfalla, Sweden
2) Fachhochschule Kaiserslautern / Zweibruecken, Amerikastrasse 1,
D-66482 Zweibruecken, Germany
3) Universitaet Freiburg, Mikrosystemtechnik, Georg-Koehler Allee 103,
D-79085 Freiburg, Germany
Laser Cleaning of Silicon Membrane Stencil Masks
Liquid film enhanced laser cleaning (termed 'LF laser cleaning' in the following) with XeCl- and KrF-excimer lasers was earlier demonstrated to remove sub-micrometer particles from silicon wafer surfaces . This technique is presently investigated for cleaning of ion-beam lithography (IPL) masks from particulate contamination. These are silicon membrane masks of 3 µm thickness, containing the pattern to be printed as stencil holes of dimensions down to 400nm.
Based on earlier experience we concentrate on 'liquid film enhanced' LF laser cleaning with short pulsed UV lasers. This technique uses a thick liquid film condensed onto the surface to be claening just prior to the laser pulse. The liquid film will then explode with high pressure, and expel the particles from the surface
Our experimental results showed sharp laser thresholds for efficient removal of e.g. Al2=3 particles (size range 0.2 - 2µm) from silicon wafers for XeCl-excimer laser pulses of 300 mJ/cm2 . With 32 pulses at 450 mJ/cm2 we achieved 99.9% removal of particles. Similar results we obtained with other particles like SiO2, or Si.
Cleaning of silicon membrane masks was also possible, and particles could be removed out of the trenches of the stencil mask.
Experimental results with short pulse Er:YAG lasers at 3 different wavelengths around the water absorption peak were not as successful as those with the excimer laser. Analysis and interpretation will be presented at the conference.
 A. C. Tam, W. P. Leung, W. Zapka, W. Ziemlich; J. Appl. Phys. 71(7), 1992, 3515
Particle Adhesion and Removal in CMP
(Abstract not yet available)