Skip to main content

Über dieses Buch

This volume documents the proceedings of the Second Symposium on Particles on Surfaces: Detection, Adhesion and Removal held as part of the 19th Annual Meeting of the Fine Particle Society in Santa Clara, California, July 20-25, 1988. The premier symposium on this topic was l organized in 1986 and has been properly chronicled . Based on the success of these two events and the high interest evinced by the technical community, we plan to regularly hold symposia on this topic on a biennial basis and the next one is slated for August 20-24, 1990 in San Diego, California. l As pointed out in the Preface to the first volume , the topic of particles on surfaces is of paramount importance in legion of technological areas. Particularly in the semiconductor device fabrication area, all signals indicate that the understanding of the behavior of particles on surfaces and their removal will attain heightened importance in the times to come. As the device dimensions are shrinking at an accelerated pace, so the benign particles of today will become the killer defects in the not too distant future. The tempo of research and development activity in the field of particles on surfaces is very high, and better and novel ways are continuously being devised to remove smaller and smaller particles.



Particle-Substrate Interaction, Deposition and Adhesion


The Adhesion of Small Particles to a Surface

In this paper the role of the force of adhesion in processes that govern deposition is discussed. The movement or transport of particles in gaseous systems is controlled by three basic processes:
  • arrival to a surface;
  • whether the particles bounce or stick upon impact; and
  • whether the particles are resuspended.
J. Reed

Application of Impact Adhesion Theory to Particle Kinetic Energy Loss Measurements

Although a number of models have been proposed to describe the adhesion of particles to surfaces, most theories are restricted to the case of the removal of precollected particles from a surface. This provides an estimate only of adhesion surface energy between the particle and substrate. In the dynamic case of an impact, it is necessary to consider the adhesion surface energy and the loss of impact kinetic energy as well as the interaction between the two mechanisms. Data from laser-Doppler measurements of kinetic energy loss were used to evaluate the few available impact-adhesion models from low impact velocities, where the major influence of adhesion surface energy is expected, up to high velocities, where plastic deformation dominates the impact.
Stephen Wall, Walter John, Simon L. Goren

Adhesion Force Due To a Meniscus in a Crossed-Fiber System

In certain circumstances, the adhesion force which holds two particles together can be due chiefly to a water meniscus that forms in the contact zone. The source of this meniscus is water vapor, which condenses from the surrounding atmosphere. We have measured the adhesion force due to a water meniscus between a pair of circularly cylindrical quartz fibers held in a crossed cylinder geometry. The experiment permitted many repeated measurements with the same fiber geometry and environmental conditions, as well as control of the geometry and environment. We report measurements of the adhesion force as a function of relative humidity and fiber size, compare our measurements to theory, comment on hysteresis observed on several time scales, and discuss statistics of our measurements. We also discuss the application of light-scattering techniques to the study of adsorbed surface layers.
Thomas M. Wentzel, William S. Bickel

Adhesion Induced Deformations between Particles and Substrates

Deformations of planar substrates and contacting particles arising from adhesion forces had been postulated long ago. We report direct observations of such deformations between submicron particles of polyvinylidene fluoride (PVF2) and a polyester—siloxane block copolymer or polished silicon substrates and between approximately 2–3 µm nickel particles and the same substrates. These observations were made using scanning electron microscopy (SEM). For the case of the silicon substrate, only the relatively soft particles of PVF2 were observed to deform. On the other hand, substantial deformations of the polyester—siloxane block copolymer were observed when contacted by either type of particle. The magnitude of the deformation of the polyester—siloxane substrate in contact with the PVF2 particles was calculated, assuming that the adhesion forces arise from van der Waals interactions and from the interfacial tension between the contacting surfaces. The cases of only Hertzian (elastic) and of both elastic and plastic deformation were considered. It was found that previously proposed models do not accurately predict the diameter of the contact area. However, the observed diameter was accurately estimated by combining the plastic response model of Krupp with the interfacial energy model proposed by Johnson, Kendall, and Roberts. The contributions of various experimental artifacts, such as joule heating, space charge effects resulting from the electron beam in the SEM, and the effects of conductive sputtered coatings on the observations, are also discussed.
L. P. DeMejo, D. S. Rimai, R. C. Bowen

The Effect of External Noises on the Attachment of Particles to Solid Surfaces

A stochastic analysis of the reversible attachment of particulate material to solid surfaces in a batch system is presented. The effect of external noises on the transient behavior of the phenomenon is investigated through solving the stochastic differential equation describing the system. Two types of random sources have been considered: i) a random initial condition, and ii) a random operating environment which leads to a differential equation with random coefficients. The results of Monte Carlo simulation reveal that these random factors can have significant effect on the transient number of attached entities; their effect on its steady-state value, however, is negligible.
J. P. Hsu

Measurement of Detachable Submicrometer Particles and Surface Cleanliness of Clean Room Garments

The quality and cleanliness of clean room garments is an important issue to the contamination control industry. Garment cleanliness may be evaluated with techniques such as the Helmke Drum tumble test, and surface examinations using Multiple Internal Reflectance (MIR) and Energy Dispersive Spectroscopy (EDS). The author summarizes four years of experience with tumble testing of various clean room fabrics and monitoring the effectiveness of clean room laundry processing. Common patterns of tumble test results and fabric comparisons are reviewed. Measurements of salts, metals and organic surface contaminants on clean processed garments are also presented. Together these methods may be used to diagnose and monitor the clean wash process. Critical factors in maintaining the quality of freshly laundered garments are identified, and a comparison of clean room laundries is presented.
Mark B. Stutman

Accumulation of Particle Derived Ionic Contaminants on Electronic Equipment: Airborne Concentrations and Deposition Velocities

We previously reported deposition velocities for chloride, nitrate, sulfate, sodium, ammonium, potassium, magnesium, and calcium associated with fine and coarse particles at telephone company switching equipment locations in Wichita, Kansas and Lubbock, Texas. Preliminary data were also reported for a site in Newark, NJ. These results were based on comparisons of indoor concentrations, obtained using dichotomous samplers for collection and ion chromatography (IC) for analysis, with average annual surface accumulations that were obtained by collecting water extracts of zinc and aluminum structural surfaces and then analyzing by IC. In this paper we report the complete results for the Newark site and for a new site in Neenah, Wisconsin. The deposition velocities are based on average annual surface accumulations derived from approximately 500 extractions of zinc and aluminum surfaces as well as indoor concentrations measured for an annual cycle at weekly intervals at both the Newark and Neenah sites. The results demonstrate that deposition velocities for each of the major ions are similar, regardless of location, and can be used to predict surface accumulation rates and that variations in surface accumulations are not attributable to variations in airborne concentrations.
J. D. Sinclair, L. A. Psota-Kelty, C. J. Weschler, H. C. Shields

Adhesion of Ash Particles on Heat Transfer Surfaces in Coal Combustion Applications: Mechanisms and Implications

The useful lifetime of rotor blades and stator vanes in gas turbines burning pulverized coal may be limited by the build-up of thick, strongly-adherent, insulating multiphase deposits composed of captured ash particles and condensed liquid ‘glue’. The rate of growth of the deposit layer is governed by the competing dynamics of ash particle deposition and deposit erosion due to oncoming particles. A model for ash deposition developed previously has been extended here to account for deposit erosion due to particle impact. Sticking coefficient (ratio of captured to incident particle mass) predictions made using the present theory are compared against corresponding measurements in the DOE-METC combustion/deposition entrained reactor (CDER), a test facility designed to simulate the deposition characteristics of a cooled turbine blade surface. Inferences are drawn regarding the deposition mechanisms predominating at different gas and surface temperatures, and the role deposit erosion plays in determining the rate of deposit build-up.
R. Nagarajan

Factors Affecting Adhesion of Drug Particles to Surfaces in Pharmaceutical Systems

The mechanism of adhesion of drug powders on the surface of polymer coated glass beads was studied. Glass beads (500 µm) were air suspension coated using a 5% hydroxypropylmethylcellulose phthalate solution in dichloromethane: methanol solvent (50:50). Specific particle size fractions of several sulphonamide powders were prepared by fluid energy milling and sonic sifting; distributions were characterized by laser diffraction. Average charge to mass ratios on particle detachment were measured using an air stream Faraday cage. Total adhesion was determined by a centrifugal method and characterized by an S50 value, i.e. the speed required to detach 50% of the drug particles. Electrical interactions probably caused by triboelectrification during the preparation of the interactive system contributed significantly to the initial drug adhesion. Adhesion decreased with time and was well correlated with the charge to mass ratio decrease. The rate of decrease in adhesion increased with increasing relative humidity. Adhesion did not occur uniformly over the polymer surface with local multilayer adsorption occurring when the drug concentration was increased. When the interactive systems were prepared at higher relative humidity conditions (60% RH), adhesion between the drug particles and surface was significantly decreased; non-electrical interactions contributed to the adhesion process under these conditions.
P. J. Stewart

Particle Detection, Analysis and Characterization


The Role of Infrared and Raman Microspectroscopies in the Characterization of Particles on Surfaces

The nondestructive identification of particles on surfaces can be accomplished quickly and definitively using vibrational microspectroscopy. Both Fourier-transform infrared (FT-IR) and Raman microspectroscopies are vibrational spectroscopy methods which are now commercially available. These two techniques arfe strongly complementary for several reasons. Both methods detect vibrational energies between atoms in molecules. However, the quantum mechanical selection rules differ such that infrared spectroscopy is more sensitive to asymmetric molecular distortions and Raman spectroscopy to symmetric modes. This means that weak bands in infrared spectra are often strong in Raman spectra and vice versa. Infrared spectroscopy directly measures absorption of incident infrared radiation. However, Raman spectroscopy observes the same vibrational frequencies by measuring the differences in energy between incident and scattered visible light. This means that infrared microspectroscopy has a diffraction limitation of about 10 micrometers while the diffraction limit for Raman microspectroscopy is near 1 micrometer, so Raman microspectroscopy can be used for smaller particles.
Kenneth J. Ward, David R. Tallant

Infrared and Raman Microspectroscopy: An Overview of their Use in the Identification of Microscopic Particulates

The recent coupling of infrared and Raman spectrometers with microscopes has provided the analyst with two very powerful new tools for identification of microscopic particulates. Unlike other analytical tools, these microspectroscopic techniques yield molecular, as opposed to elemental, information. Both infrared and Raman microspectroscopy are reviewed with respect to their historical development and their respective advantages, limitations and capabilities. The two methods are then discussed in terms of their complementarity in solving chemical identification problems, and it is shown that both methods are often needed for complete sample characterization. Finally, their application to the identification of microscopic particulates is illustrated by two case studies: identification of solder flux residue contamination and characterization of various dust particles.
Patricia L. Lang, Andre’ J. Sommer, J. E. Katon

Particle Identification By Auger Electron Spectroscopy

The use of Auger Electron Spectroscopy (AES) as a method for particle characterization has grown in importance, particularly as semiconductor geometries have become smaller. The surface sensitivity and small analysis size of AES makes it the technique of choice for the characterization of sub-micron particles because X-ray techniques approach the limit of their spatial and depth resolution. AES provides particle identification with a minimum of interference from the surrounding matrix. The analysis and identification of particles found during the processing of electronic circuits is discussed. A comparison of surface and near-surface techniques for applicability to sub-micron particle analysis is also made.
Kenneth D. Bomben, William F. Stickle

Identification and Characterization of Nonmetallic Particulate Contamination Removed from Aerospace Components

The performance and reliability of today’s highly advanced aircraft and space hardware have become increasingly dependent upon the elimination or very tight control of particulate contamination. The presence of particles on optical surfaces of communication components can cause beam scattering, as well as destruction of optical surface coatings, thereby resulting in significant power losses. Particulate contamination within critical tolerance aircraft hardware components can degrade system performance. These particles can originate from either external sources during assembly or from internal materials of construction. Identification and characterization of particulate material is essential in determining its source. Once the source is identified, corrective steps can be initiated to reduce or eliminate the particulate contamination. This paper will describe Fourier Transform Infared Spectroscopy and Probe Mass Spectrometry techniques which have been employed in the analysis of particulate contamination to increase performance and reliability of today’s aerospace products.
C. E. Wilson, D. A. Scheer

Surface Particle Inspection Plans in Semiconductor Manufacturing

The review of two existing inspection policies for estimation of particle densities has shown that the policy which reduces the number of inspected wafers leads to a biased density estimator. In the present paper, a class of sequential inspection polcies with unbiased density estimators is proposed. These policies reduce the number of inspected wafers. In many practical situations the variances of density estimators for these sequential policies are smaller than the variances of the estimators for the standard existing policies.
Zinovy Fichtenholz, Leon L. Pesotchinsky

Particulate Generation and Detection on Surfaces

This paper describes a tool that was developed to evaluate particulate generation of materials. The tool provides realtime data of both airborne and settling particulates in a consistent and repeatable manner by inputting a continuous and controlled stress on the material in a controlled test chamber environment. Sample test results on several cleanroom materials are presented to demonstrate the use of the tool.
Marvin Fein

Identification of Small Particles on a Silicon Wafer

An SEM with EDX is usually used to analyze particles on silicon wafer surfaces, but since the particles are so few and so small, much time must be spent searching for them. In the present report, a combined method is described for particle detection and analysis. First, a surface particle counter (Tencor Surfscan 4500) is used to locate particle positions. The coordinates of each particle are then converted to SEM coordinates, and each particle observed and analysed. Using this new method, the time spent for full particle analysis of a wafer is greatly reduced.
N. Fujino, S. Miyazaki, M. Takeshita, H. Horie, S. Sumita, T. Shiraiwa

The Evaluation of PWA and SMA Cleanliness Levels for “In-Line” Defluxing by High Performance Liquid Chromatography

Recently surface mount technology (SMT) has gained popularity. Surface mount assemblies (SMA’s) are more electrically efficient, sophisticated in function and optimized with respect to board real estate. Boards containing SMD’s (surface mount devices) require solder paste screening, surface component attachment by infrared or vapor phase reflow processing as well as conventional processing if they contain both SMD’s and through hole components (THC’s).
R. F. Klima, J. K. Bonner

Particle Prevention and Implications


Wear Resistant Coatings Reduce Particulate Contamination in a Magnetic Disk Drive

A test was performed to evaluate the particulate contamination coming from components in a magnetic disk drive. The 300-series stainless steel components undergo unlubricated sliding motion and are susceptible to adhesive wear and fretting. To reduce the wear debris, the components were coated with six different hard coatings: electroless nickel, chromium, Nedox®, titanium nitride, amorphous diamond, and iron boride. Hardware was tested in a specially made chamber that permitted collection of the airborne particles on nucleopore filters as well as real-time particle counts. Particle size distribution and elemental composition were provided by automated image analysis using SEM and EDX.
Wendy Jones, John McDowell, Walter Prater, Garvin Stone

Particle Reduction on Silicon Wafers as a Result of Isopropyl Alcohol Vapor Displacement drying after Wet Processing

The presence of particles on the surfaces of semiconductor wafers is known to negatively affect device yields. As linewidths continue to shrink, the inherent limitations in many of the commonly used methods for drying wafers after wet processing have created the need for alternatives which offer better particle control. In this paper the use of IPA vapor displacement drying was examined as a means to dry wafers. Wafers were immersed in water in an enclosed chamber, and IPA vapor was then released into the top of the chamber while the water was pumped out through the bottom. Physical observation of drop formation on the wafers indicated both temperature and drain rate are important factors in effective drying with this method. When a commercial apparatus applying these principles was used to dry wafers in a Class 10 clean room, fewer than two particles greater than 0.5μm in diameter were routinely added to l00mm wafers, confirming the practical applicability of the findings.
Joan W. Koppenbrink, Christopher F. McConnell, Alan E. Walter

Implications of Particle Contamination for Thin Film Growth

Particles can serve as nucleation sites for the growth of defects in metal and semiconductor thin films deposited by evaporation, sputtering, and Molecular Beam Epitaxy (MBE). For each of these deposition technologies, examples of particle-induced defects are given. Future work is outlined with an emphasis on MBE.
Arye Shapiro, Charles M. Falco

Implications of Particulate Contamination in E-Beam Lithography

Experiments were made to compare chromium spot and pinhole defects on finished photomasks written with Manufacturing Electron Beam Lithography Systems (MEBES®). The experiments were performed while adding particulate contamination naturally during blank manufacture, packaging, storage, and by the writing process. In addition, particulate contamination that was added purposely before or after exposure was compared to chromium spot and pinhole defects found on finished experimental plates.
Robert L. Dean

Particle Removal


Enhanced Removal of Sub-micron Particles from Surfaces by High Molecular Weight Fluorocarbon Surfactant Solutions

A novel patented method of removing unwanted sub-micron particles from solid surfaces is described, A key component of this approach is the use of solutions of high molecular weight fluorinated surfactants (M.W.> 2,000 Daltons) in an inert perfluorinated carrier liquid to effect particle removal. The unique characteristics of the high molecular weight surfactants used in these solutions are that they:
can result in the formation of a relatively thick solvated shell around the particles, which promotes the detachment of solid particles from a solid substrate, and
are reversibly adsorbed on teh solid surfaces being treated, and that adsorption is controlled by controlling their concentration in the liquid phase, and which can be made as low as desired.
Robert Kaiser

Comparison of Freon with Water Cleaning Processes for Disk-drive Parts

This paper documents the evaluation of three parts-cleaning processes for disk-drive parts: Freon TF/ultrasonics, water/ultrasonics, and water spray systems. Particulate, organic, and ionic contaminants, along with moisture residue, were measured and compared. The data showed that a water cleaning system was a feasible replacement for a Freon cleaning system for disk-drive parts.
Ming Ko

Ultrasonic and Hydrodynamic Techniques for Particle Removal from Silicon Wafers

Ultrasonic and hydrodynamic cleaning techniques were used to remove particulate contaminants from bare silicon wafers and wafers with surface oxide layers. The variation of cleaning efficiency with particle size, composition and cleaning solvent was quantified. In comparison to submicrometer polystyrene and glass particles, silicon dust was found to be a particularly difficult contaminant to clean. The RCA standard cleaning solution (SC-1), which is a blend of 5 H2O : 1 NH4OH: 1 H2O2, when used in the ultrasonic or hydrodynamic cleaner, was less effective than DI-water in removing silicon particles, especially at low cleaning energies. Both types of cleaning systems resulted in particle removal efficiencies above 80% for submicrometer particles, the ultrasonic technique being more sensitive to particle and wafer surface composition.
V. B. Menon, L. D. Michaels, R. P. Donovan, D. S. Ensor

New Sonic Cleaning Technology for Particle Removal from Semiconductor Surfaces

Several sonic cleaning technologies are currently used in the semiconductor industry. An overview of these technologies is presented, comparing their cleaning mechanisms, applications and suitability. Using an analogy to photography, we then discuss a “reciprocity principle” in megasonic cleaning and describe an experiment we performed to test this hypothesis. Equivalent cleaning was observed for the same average acoustic power, independent of peak intensity. In addition, cleaning was approximately dependent on total energy exposure. Implications for the design of megasonic cleaning equipment are discussed.
D. M. Berg, T. Grimsley, P. Hammond, C. T. Sorenson


Weitere Informationen