Numerical simulation of thermophoretic particle decontamination in clean rooms

In order to develop strategies for minimizing deposition of contaminant particles of diameters ranging from 0.1 to 1.0 μm on a wafer, the effect of thermophoresis on a particle deposition velocity was numerically studied. The angle between wafer surface and direction of free-stream flow was introduced as a system parameter. Convection, diffusion, sedimentation, and thermophoresis were included as particle transport mechanisms. Similarity transform was applied to the model equations and obtained equations with dimensions reduced by one. The results suggest that it is possible to enhance the removal of particles of diameter ranging from 0.1 to 1.0 μm by heating with a temperature difference of 10–30°C between wafer surface and the air stream. If the filter of a clean room removes well around 0.1 μm sized particles, the free-slream velocity or flow angle should be increased for the effective removal of particle by thermophoresis, but, if the filter is efficient in removing particles around 1 μm, the free-stream velocity or flow angle should be decreased.     

Thermophoretic particle deposition efficiency in turbulent tube flow

This study investigated the thermophoretic particle deposition efficiency numerically. The critical trajectory was used to calculate thermophoretic particle deposition in turbulent tube flow. The numerical results obtained in turbulent flow regime in this study were validated by particle deposition efficiency measurements with monodisperse particles (particle diameter ranges from 0.038 to 0.498 μm) in a tube (1.18 m long, 0.43 cm i.d., stainless-steel tube). The theoretical predictions are found to fit the experimental data of Tsai et al. [Tsai, C. J., J. S. Lin, S. G. Aggarwal, and D. R. Chen, “Thermophoretic Deposition of Particles in Laminar and Turbulent Tube Flows,” Aerosol Sci. Technol., 38, 131 (2004)] very well in turbulent flows. In addition, an empirical expression has been developed to predict the thermophoretic deposition efficiency in turbulent tube flow.     

Annular flow configuration with high deposition efficiency for the experimental determination of thermophoretic diffusion coefficients

An experimental study was carried out to produce reliable data for the determination of the thermophoretic diffusion coefficient K_th of suspended oil particles in air, in the transition regime. An original device was used for the thermophoretic deposition efficiency measurement, involving a turbulent flow through a concentric tube annulus, with the inner tube cooled (5 °C) and the outer heated. Experimental parameters varied in particle diameter (0.039–5.13 μm), flow rate (150, 200, and 250 Nl min⁻¹, corresponding to Reynolds number in the range 5000–10 000) and hot wall temperature (65–125 °C). This configuration, based on three controlled temperatures (gas inlet, cold wall, hot wall), the so-called “3T”, permits an overall deposition efficiency enhancement compared to conventional “2T” penetration devices (hot gas flow in a cooled tube). In the 3T configuration, significant thermophoretic deposition efficiencies have been obtained (up to 27%), together with limited gas temperature axial variations, thus permitting a reliable determination of the thermophoretic diffusion coefficient K_th.An analytical model was developed for the prediction of the thermophoretic deposition efficiency, for a given value of the thermophoretic diffusion coefficient K_th. This model has been used, together with our measurement results, to derive the K_th experimental values, for a Knudsen number ranging from 0.01 to 3. These K_th values were compared with evaluations based on various models available in the literature. Although widely used, Talbot's model always provides K_th values higher than our experimental results in the transition regime. The most relevant model appears to be the one proposed by Beresnev and Chernyak, particularly for an energy accommodation slightly lower than one.     

Effects of air temperature and humidity on particle deposition

Particle deposition in a fully developed turbulent duct flow was studied. The random walk model of Lagrangian approach was used to predict the trajectories of 3000 particles with a density of 900 kg/m³. The effects of thermophoretic force and air humidity were also considered. The results were compared with the previous studies with a particle size range of 0.01–50 μm and air flow velocity of 5 m/s. The profile of dimensionless deposition velocity with relaxation time presents a V-shaped curve and the results are in good agreement with the previous studies.The effects of air temperature and humidity on particle deposition with a particle size of 1 μm were also investigated. The results show that thermophoretic force accelerates particle deposition onto the duct walls with increasing temperature difference between air flow and the duct wall surface. Meanwhile, it was found that particle deposition velocity increases with air humidity.     

氯含氢分析仪取样系统的设计

分析了氯舍氢分析仪的取样系统的特性,提出了6个设计要点：负压取样、样气干燥、设备及管路避光、管径设计、材质设计、尾气处理。设计的取样系统应用到实际生产中,运行平稳可靠,满足了安全生产的要求。     

A new approach to high-shear mixer granulation using positron emission particle tracking

Positron emission particle tracking (PEPT) can provide quantitative information on particle motion in a mixer. For the first time, the present study clarifies the flow patterns of particles and granules during the process in which granules are formed in high-shear mixer granulation in three dimensions using PEPT. A number of different regions with size-dependent flows account for the variations in frequency and velocity of collisions between differently sized granules in the kernel when segregation occurs in the granulator. These findings offer a better understanding of the product properties and process attributes of high-shear mixer granulation, which will be of direct benefit as pharmaceutical and other products become more complex.     

Design improvements on rotary valve particle feeders used for obtaining suspended airflows

The back repulsion of low density particles, sticking of particles in feeder valve and interlocking and cohesive archs of particles in feeder hopper have been widely observed problems of a conventional rotary valve feeder used for feeding of granular particles into an airflow line. In this study, these problems of the rotary valve feeders were isolated by means of novel modifications to its conventional design. Results of the modifications showed that the modified feeder can be sensitively used to feed granular particles with mass feeding rates ranging between 5±0.15 and 85±0.85 g/s without encountering problems experienced with conventional rotary valves.     

Influence of particle size distribution on rheology and particle packing of silica-based suspensions

The effect of particle size, particle size distribution and milling time on the rheological behaviour and particle packing of silica suspensions was investigated using slurries containing total solids loading of 46 vol.%. Three silica powders with different average particle sizes (2.2, 6.5 and 19 μm), derived from dry milling of sand, and a colloidal fumed silica powder with 0.07 μm were used. Different proportions of colloidal fumed silica powder were added to each of the coarser silica powders and the mixtures were ball-milled for different time periods. The influence of these factors and of the particle size ratio on the rheological behaviour of the suspensions and densities of green slip cast bodies was studied.The results show that the flow properties of slips are strongly influenced by the particle size distribution. The viscosity of suspensions increases with the addition of fine particles, imposing some practical limitations in terms of volume fraction of fines that can be added. On the other hand, increasing the size ratio enhanced the shear thinning character of the suspensions, while decreasing the size ratio led to an accentuation of the shear thickening behaviour. For all mixed suspensions, green densities increased with increasing milling time, due to size reduction of silica powders and a more efficient deagglomeration of fumed silica. Increasing amounts of fumed silica led to a first increase of particle packing up to a maximum, followed by a decreasing trend for further additions. Good relationships could be observed between rheological results and packing densities.     

Collection efficiencies and particle size distributions from sampling cyclones — comparison of recent theories with experimental data

The grade collection efficiencies and particle size distributions from Stairmand-type cyclones are compared with predictions from four theories, viz. the models of Leith and Licht (1972), Dietz (1981) Mothes and Loffler (1988) and Li and Wang (1989). It is shown that the theory of Mothes and Loffler, which recogizes different flow regions within a cyclone, coupled with a realistic estimate of the particles turbulent diffusivity, is capable of providing good estimates of both the grade-efficiency curves and particle size distributions of the cyclone catch and/or outlet dusts. The lack of adequate theories or empirical formulae to estimate the effective turbulent diffusivity under cyclone flow seems the major factor hindering the use of the Mothes and Loffler theory for predictive purposes. High loadings and agglomeration of fines, if present, are responsible for strong departures from theoretical predictions.     

Evaluation of an in situ sampling probe for its accuracy in determining particle size distributions from flames

An in situ sampling probe for flames is evaluated for any biases with respect to size distributions in a given size and concentration range. Ultrafine metal particles (Pd, Pt) suspended in argon gas are added to a laminar diffusion flame, burning with argon-diluted methane. Samples of combustion aerosol are extracted from the flame via a small orifice and immediately quenched with nitrogen. The size distribution of the metal particles, recorded at various sampling positions in the flame, is found to be but little affected by the flame and the sampling procedure. The distribution of carbonaceous soot particles from an unseeded flame at a fixed sampling position is not shifted upon seeding but coexists next to the peak from the metallic flame additive. It is concluded that droplets of volatile combustion products, formed in the sampling system through homogeneous nucleation, do not significantly contribute to the soot particle signal. An investigation of the photoelectric yield of the samples implies significant changes of the surface properties of the metal particles during the combustion process. This seeding method offers a new path to study chemical processes in the flame, especially heterogeneous and catalytic reactions of flame gases with particles.     

Effect of particle concentration and turbidity on particle characterization using digital holography

In this paper, the effect of particle concentration and turbidity on the performance of inline transmission based digital holography (DH) for particle characterization is studied. The results are analyzed based on the two metrics, i.e., detection efficiency and mean size of the particle population. Two sample depths are analyzed to quantify the effect of particle concentration. Considering 50% detection efficiency as a threshold, it is concluded that DH works well up to 0.1% (v/v) and 0.2% (v/v) particle concentrations for 10 mm and 5 mm sample depths, respectively. From the turbidity tests, it is found that DH works well up to 150 nephelometric turbidity unit (NTU) turbidity level for 10 mm sample depth.     

A direct sampling particle filter from approximate conditional density function supported on constrained state space

Constraints on the state vector must be taken into account in the state estimation problem. Recently, acceptance/rejection and projection methods are proposed in the particle filter framework for constraining the particles. A weighted least squares formulation is used for constraining samples in unscented and ensemble Kalman filters. In this paper, direct sampling from an approximate conditional probability density function (pdf) is proposed. It is obtained by approximating the a priori pdf as a Gaussian. The support of the conditional density is a subset of the intersection of two supports, the 3-sigma bounds of the priori Gaussian and the constrained state space. A direct sampling algorithm is proposed for handling linear and nonlinear equality and inequality constraints. The algorithm uses the constrained mode for nonlinear constraints.     

Modeling of Particle Removal using Non-contact Brush Scrubbing in Post-CMP Cleaning Processes

Particle removal using non-contact brush scrubbing for post-CMP (Chemical Mechanical Planarization) cleaning is investigated analytically. The removal of Si O ₂ and A l ₂ O ₃ particles adhered onto Si O ₂ film coated on the wafer surface are considered. The cleaning fluid (H ₂ O/N H ₄ OH = 1:25 and 1:200) flowing between the brush and wafer surface is treated as a thin-film fluid flow. The flow field details and its effect on the drag force acting on the adhered particles are discussed. In addition to the drag force, the electrical double layer (EDL) and thermophoretic force effects on particle removal are also considered. It was found that the dominant force in achieving particle removal using a rolling mechanism is the drag force. The EDL and thermophoretic forces have an insignificant effect on particle removal. Based on the results from this study, particles of submicron size can be removed from a wafer surface using higher brush rotation speed and pure deionized (DI) water as the cleaning fluid.     

Unbiased estimation for sequential multinomial sampling plans

For sequential multinomial sampling, a sufficient condition for the stopping rule to be closed is obtained. And, by the application of the Rao-Blackwell method,an unbiased estimator based on the sufficient statistics is given for some functions of unknown parameters. The results are applied to some sequential stopping rules.     

Experimental study of particle trajectory in electrostatics powder coating process

Experimental studies of particle trajectory were conducted in an isolated Plexiglass coating booth. Polyester particles were injected with a Norsdon® electrostatics spray gun with a fixed distance from the gun to a grounded plate. Using a Dantec® Particle Dynamic Analyzer, the particle velocity and size distribution were measured simultaneously. The experimental results showed that the effect of electrostatics charging on the particle trajectory is strong in the close vicinity of the target, but can be neglected at locations away from the target. The influence on particle size and velocity profile due to the electric field between the charged particles and grounded target was weakened as more particles deposited on the target. When no charge is introduced on the coating powder, particle segregation is observed for particles larger than 100 μm. Particle gravitational settlement is noticed even near the gun tip. However, particle charging largely eliminates the segregation at a gun-to-target distance of 25 cm and helps break agglomerates formed in the spray system. The gun-to-target velocities of larger particles exhibited noticeable deviation from those of the flow field as the grounded target was approached. The study revealed that the onset of electrostatic coating is an important period that can affect the transfer efficiency and film thickness.     

新型全柱状采样器研究与推广应用

由于煤炭是大宗产品,动力用煤煤量较大,供销频繁。电煤紧张时,煤炭供需双方质量纠纷不断,现行机械化采样器是按照标准煤炭产品设计的,在贵州省使用时存在很大的局限性,极不适应煤炭市场复杂的质量状况。企业使用中存在着很多问题,分层装车、不同品质的煤混装等造成购煤企业不能评价煤炭的真实质量,热值虚高,经济损失较大。根据现行机械化采样器存在的问题,提出了采用全柱状、全深度、全断面、全方位采取子样,子样量大且全部破碎缩分,解决了物料不均匀情况下的采制样难题,检验结果取得较好的准确性。     

Design space description through adaptive sampling and symbolic computation

Design space definition is one of the key parts in pharmaceutical research and development. In this article, we propose a novel solution strategy to explicitly describe the design space without recourse decisions. First, to smooth the boundary, the Kreisselmeier–Steinhauser (KS) function is applied to aggregate all inequality constraints. Next, for creating a surrogate polynomial model of the KS function, we focus on finding sampling points on the boundary of KS space. After performing Latin hypercube sampling (LHS), two methods are presented to efficiently expand the boundary points, that is, line projection to the boundary through any two feasible LHS points and perturbation around the adaptive sampling points. Finally, a symbolic computation method, cylindrical algebraic decomposition, is applied to transform the surrogate model into a series of explicit and triangular subsystems, which can be converted to describe the KS space. Two case studies show the efficiency of the proposed algorithm.