OPTICAL MEASUREMENTS OF SURFACE PARTICLES FOR EVALUATING REMOVAL PERFORMANCE OF AIR-JET Min Tae Hong 1, Takehiro Tanaka 1, Yoon Changhyun 1, Sho Hirano 1, Shuji Fujii 1, and Koji Watanabe 2 1 Department of Mechanical and Environmental Informatics, Tokyo Institute of Technology, Tokyo, Japan 2 Hitachi Plant Technologies Ltd., Matsudo-shi, Chiba-Ken, Japan ABSTRACT The purpose of this study is to develop an optical measurement method of particles on a flat surface for evaluating particle removal performance of air-jet by nozzles in the cleanroom. Glass beads are selected for the test aerosol and they are adhered on test surfaces of Si wafer as witness plate. Authors developed new optical measuring system using Ar laser and surface particles are measured by the system and also a surface particle counter for Si wafer. The experimental survey is conducted for 14 types of nozzles and 3 different conditions. In this paper, standard nozzle and 3 better nozzles are selected to be reported. As the results, (1) the new measuring system can detect particles with.5 micro meters and larger range, (2) removal performances are estimated by size distribution and particles larger than 3.5 micro meters are removed over 9% by air-jet of the test nozzles. KEYWORDS Optical measurement, Removal performance, Surface particle, Air-jet nozzle, Flat panel display INTRODUCTION The cleanroom is widely used in both industrial and biological industries, such as electronics, flat panel display and medicine. Contamination control is important to keep good cleanroom performance and particles are still main contaminants in the factory of flat panel display industry. In the field of flat panel display industry, some process has a problem of contamination by particle larger than 5μm and particle removing by air-jet is utilizing in the process, the estimation has not been established yet. Accordingly, this study aims at proposing the optical measuring methods and evaluating the particle removal ability of the nozzle by witness plate of flat glass materials. In brief, an optical measuring system is proposed and its performance is measured by PSL (polystyrene Latex) particles. Si wafer is selected for the witness plate of flat glass panel in order to estimate the removal characteristics of air-jet nozzles. The experimental survey is conducted for 14 types of nozzles and 3 different conditions. In this paper, standard nozzle and 3 better nozzles are selected to be reported. Corresponding Author: Tel & Fax :+81-3-5734-2799 E-mail address: min.t.aa@m.titech.ac.jp
= Co-Cn Co THE NOZZLE TO ESTIMATE THE PERFORMANCE Figure 1 presents an example of the sectional diagram of nozzle A, which is currently being marketed among 14 types of nozzles used in estimating the performance. As indicated in the figure, the slit in the center of the nozzle push the clean air with the high pressure, and the target surface is passing below the nozzle. Adhered particles on the surface are removed in the supplied air and then to the vacuum exhausts. Totally 14 different features of nozzles and 3 conditions are tested. Table 1 represents the conditions and the number of tests by different nozzle models. Clean air pressure vacuum move Target surface After removing particles Before removing particles Figure 1. The outline of the removal system and sectional diagram of nozzle A Table 1. Conditions by different nozzle models Nozzle model Condition Gap(mm) Pressure Vacuum Pass speed # Number of (KPa) (KPa) (m/s) tests A a.6 14.7.2 5 b.6 14.7.1 3 a.6 16.65.2 2 G b 1. 14.65.2 2 c 1. 16.65.2 2 H a 1. 12, 5.65.2 3 I a 1. 14 1, 1.2 3 b 1. 14 2, 1.2 3 # As indicated in Fig 1, pass speed is that target surface is moving below the nozzle ESTIMATION METHOD FOR THE REMOVAL EFFICIENCY For the estimation of the particle removal ability for the nozzle performance, Surface particle concentrations are measured before and after passing below the nozzle and the removal rate is estimated as the removal efficiency. The removal efficiency η is calculated by equation (1). = Co - Cn Co (1) Where Co : Surface particles concentration before testing(number/ cm2 ) Cn : Surface particles concentration after testing(number/ cm2 )
TEST SURFACE AND PARTICLES Test surface Flat glass panel could be considered to be suitable for the estimation of nozzles used in flat panel industry. But for the estimation of removals, witness plates are used because cleanliness of flat glass panel was not so good and its fluctuation was also high level. Therefore, the silicon wafer (6inch side mirror wafer, Mitsubishi material Ltd.) is utilized as the representative surface of the estimation target. They maintained the high quality as under 1 particles over.2μm adhered on the surface at the initial condition. Test particles PSL particles are firstly used to find the detection limit on size of surface particles for the developed optical system. But PSL particles were not suitable to test nozzles because of interaction by Si surface and moisture occurred with particle generation. Therefore, glass beads (MBP 1-1, APPIE, Japan) are used as the test particles. For the particle emission, standard particle generation device (nanometer, JSR Inc.) is utilized for PSL particles and the ejector developed by our research team is utilized for the glass beads. Figure 2 shows the particle deposition system by using the self-made ejector. inducement Pure air P Pump Mixing air Exhaust air Air contain with particles Si-wafer Glass beads Figure 2. Set up the particle deposition system by using the ejector
MEASURING METHODS OF THE SURFACE PARTICLE CONCENTRATION Developed optical system Figure 3 represents the outline of the developed optical measuring system using an Ar laser and a camera. Ar Laser shoots the test surface with flat beam and the scattered light of particle is detected by a camera at the forward position. After photographing, the picture is processed by software of Magical IP (Magical art. Inc) to find the surface number concentration. By using PSL particles, the detection limit of this system is measured as about.5μm size. For the nozzle test, the removal efficiency is calculated from the surface concentrations before and after passing below the test nozzle. F : 3mm Target area 45mmX15mm Optical lens F : mm 45mm Optical fiber Cylindrical lens Laser generation device Figure 3. Outline of the optical measuring system as measuring equipment As Si-wafer is used for witness plate, a surface particle counter can be used in this study. Table 2 represents the outline of the counter which is commercially used. The measurements are taken in the size range of.75~1.5μm, 1.5~2.5μm, 2.5~3.5μm and over 3.5μm. Surface concentrations are calculated by the area of wafer and the detected number of particles. For the nozzle test, the removal efficiency is also calculated. Table 2. Operation condition and equipments Model Type Subject Wafer Size Thickness LS Silicon Wafer 4, 5, 6, 8 inch Approx..4 -.75mm A picture of the equipment
RESULTS Results by the developed optical measuring system Figure 4 represents the image measured by optical measuring method and the adhered particle image at the sampling area. The sampling area is cut from the original image by using Magical IP and the number of particles is obtained. Then surface concentration is calculated by the number and the cut area. Figure 5 shows the results on removal efficiency (maximum, average and minimum value) of test nozzles at each condition. Table 3 represents the results as the case at the highest removal efficiency in each nozzle. It includes nozzle type, condition of nozzle, measured data at before and after condition (image, count and concentration) and removal efficiency. Removal efficiencies of G-a and I-a represents the higher levels than others as approximately 9%. 1 9. 89.9 Removal efficiency(% 81.2.6 71.5 73.3 7.2 57.5 79.5 68.9 85.6 83.2 77.1 76.3 68.6 69.3.4 73.3 61.8 86.5 84.2 85.3 77. 69.2 Figure 4. Image of the photographing by proposed optical measuring method a b a b c a a b G H A I Nozzle model Figure 5. Results of the removal rate classified by nozzle Table 3. Results by the optical measurement methods Model Nozzle A (standard) Nozzle G Nozzle H Nozzle I Gap(mm) Pressure Vacuum Pass (KPa) (KPa) speed(m/s).6 14.7.2 Before Results Count(ea), Concentration(ea/ cm2 ) 3 5.4 64 9.5 Removal efficiency (%) 81.2 Before After 1. 14 1..2 Count(ea), Concentration(ea/ cm2 ) 1.7 14 2.1 Removal efficiency (%) 9. Before After 12, 1..65.2 5 Count(ea), Concentration(ea/ cm2 ) 367 54.4 72 1.7 Removal efficiency (%).4 Before After 1. 1. 14.2 1. Count(ea), Concentration(ea/ cm2 ) 415 61.5 42 6.2 Removal efficiency (%) 89.9 After
Results of the Measurement by the surface particle counter and the comparison Figure 6-A to Figure 6-I shows the results of removal efficiency for nozzle A, G, H, I respectively. They included the nozzle conditions and the efficiency values as maximum, average, minimum measured in particle size range by surface particle counter as compared to the results of developed system. Removal efficiency has a tendency to become higher size by size; it becomes greater than 9% over 3.5μm size but about % at the size of.75~1.5μm size. Table 4 represents results of the average removal efficiency by particle sizes and nozzles. It was also shown the SD (standard deviation) classified by conditions. It was shown that 94.8% is the highest in the nozzle G, condition c at the size over 3.5μm. Efficiency by the optical measurement method is almost same level of the value between 2.5μm and 3.5μm in the surface particle counter. 1 94.2 9.2 93.7 89.5 93. 81.2.6 75.7 88. Removlal efficiency% 73.3 71.5 7.2 57.5 35.7 35.3 29.7 26.9 49.5 48. 44.8 45.9 43.3 38.9 75.7 74.5 69.5 73.9.2 26. 11.4 a b a b a b a b a b.75-1.5 1.5-2.5 2.5-3.5 Over 3.5 Optical method Figure 6-A. Comparative results between the two measuring methods in the nozzle A 1 removal efficiency(%) 9. 85.6 83.2 79.5 77.1 76.3 68.9 68.6 69.3 53.4 52.2 5.2 5.5 51.4 47.6 5.5 47.7 45. 77.4 74.1 76.7 71.5 69.9 76.1 69.1 68.9 68.2 92.1 93.5 94.9 92. 92.8 94.8 94.8 91.9 92.1 29.5 28.5 26.6 28.9 24.6 28.4 17.6 22.6 6.7 a b c a b c a b c a b c a b c.75-1.5 1.5-2.5 2.5-3.5 Over 3.5 Optical method Figure 6-G. Comparative results between the two measuring methods in the nozzle G
73.3 1 93.1.4 92.3 9.9 removal efficiency(%) 73.3 61.8 41.1 68.2 67.7 66.7 27.9.6 39.9 25. 21.5 Optical method.75~1.5 1.5~2.5 2.5~3.5 over 3.5 particle size(um) Figure 6-H. comparative results between the two measuring methods in the nozzle H 1 95.5 95.6 89.9 86.5 85.3 84.2 77. 79.4 73. 78. 94.5 93.4 94. 92.1 removal efficiency(%) 69.2 46.7 44.9 53.2 5.5 48.4 72. 7.8 76.6 42.4 32.8 29.1 27.8 26.2 23.5 23.5 a b a b a b a b a b.75-1.5 1.5-2.5 2.5-3.5 Over 3.5 Optical method Figure 6-I. comparative results between the two measuring methods in the nozzle I Table 4. Results of the average removal efficiency (%) by particle sizes and nozzles Nozzle Condition Optical method.75~1.5 μm 1.5~2.5 μm 2.5~3.5 μm over 3.5 μm A a 71.5±8.9 26.9±9.5 44.2±4.1 66.5±4.7 89.5±.9 b 73.3±2.7 29.7±2.7 45.9±2.4 74.5±1. 93.7±.6 a 79.5±15. 17.6±15 5.5±4. 71.5±3.7 92.±.1 G b 77.1±12. 24.6±2.9 47.6±3.7 69.1±1.1 92.8±4.9 c 76.3±1. 28.9±.8 51.4±1.2 76.7±.9 94.8±1.4 H a 73.3±1. 25.±3.2.6±.6 67.7±.8 92.3±1.2 I a 86.5±3. 27.8±4.7 44.9±2.3 72.±1.1 94.5±.9 b 77.±8.1 26.2±2.8 5.5±2.5 78.±1.4 93.4±1.9
CONCLUSIONS The experimental survey is conducted for the estimation of air-jet nozzles for removing surface particles on flat surface. Measuring system is discussed and removal efficiency for 14 types of nozzle and 3 conditions are evaluated and the following results are obtained. -It was possible to visualize the particles up to.5μm size as using the developed optical measuring system. -The estimation of nozzle performance by the optical measuring system represented that the removal efficiency is 75% on average and the highest levels is 9% in G-a and I-a. - For the measurement as witness plate using the surface particle counter, it was possible to compare the removal efficiency by the particle size. Removal efficiency has a tendency to become higher size by size; it becomes greater than 9% over 3.5μm size but about % at the size of.75~1.5μm size. -Efficiency by the optical measurement method is almost same level of the value between 2.5μm and 3.5μm in the surface particle counter. In this study, witness plate is used for the estimation for nozzle performance. The developed optical measurement system could be applicable for the flat glass panel. Therefore, the authors have a plan to study the limitation for measuring the particle on glass surface by the system. REFERENCES 1. Kim kyang young (1989) A Study on surface contamination by telemeter a floating corpuscle in the clean room Tokyo institute of technology, dissertation. 2. William C. Hinds (1982) Aerosol Technology 3. Hodkinson, J. R.(1966) The Optical Measurement of Aerosols Aerosol science, C.N.Davies, Academic Press, New York. 4. Cooke, D.D. and Kerker, M. (1975) Response Calculations for Light Scattering Aerosol Particle Counters Appl. Optics, 14, 734-739