data between different
play

Data Between Different Instruments Mike Naggar MGN International - PowerPoint PPT Presentation

Correlating Particle Counter Data Between Different Instruments Mike Naggar MGN International Inc. May 2018 ULTRAPUREMICRO2018.COM 1 Purpose Who / Where: Anyone that uses a particle counter system: filter and membrane manufacturers,


  1. Correlating Particle Counter Data Between Different Instruments Mike Naggar MGN International Inc. May 2018 ULTRAPUREMICRO2018.COM 1

  2. Purpose • Who / Where: • Anyone that uses a particle counter system: filter and membrane manufacturers, container and fitting makers, UPW and facilities, chemical and photo-chemical suppliers, semiconductor Fabs, contamination control, parts cleaning, QA/QC, etc. • What: • How to correlate and compare data between different particle counter models. • When: • Before, during, and after the switch to a different particle counter model, design, or technology. • Comparing particle counters with different specifications. • Why: • Traceability and correlation back to full exposure (100%) light scattering particle counter. • Establish correlation of new instrument to historical data: instrument qualification, define/update control limits, product specifications, utilization of historical SPC data. ULTRAPUREMICRO2018.COM 2

  3. Outline • Basics of Light Scattering Particle Counter • Challenges of Detecting Smaller Particle Sizes • Counting Efficiency, Flowrate, and Effective Flowrate (View Volume) • Example Design Changes to Improve Signal Intensity • Counting Efficiency Across the Dynamic Range • How to Normalize Data for Comparison • Example of Data Normalization Across Different Models • Calculation examples and data graphs • Measurement precision / repeatability / stability • Correlation overlap across different models (same manufacturer) • Conclusions / Summary ULTRAPUREMICRO2018.COM 3

  4. Science of Light Scattering Particle Counter     2          1 1 P cos dP cos 2 1          i a b        1   ( 1)  sin d      1     2          1 1 P cos dP cos 2 1          i b a        2   ( 1)  sin d      1                  ' ' m m m      p p p a                   ' ' m m m     p p p     • Mie’s Theory of light scattering              ' ' m m m      p p p b                   ' ' m m m     p p p 1 cos   P  : Legendre polynomial   ,  : Bessel function   • Rayleigh’s Theory of light scattering ULTRAPUREMICRO2018.COM 4

  5. Design of Light Scattering Particle Counter 1.0 m m Signal (voltage) 0.5 m m 0.3 m m Time • 3 Core Components • Light Source (Laser) • Detection Zone / Flowcell • Detector ULTRAPUREMICRO2018.COM 5

  6. Challenges of Detecting Smaller Particles • Exponential decrease of particle signal as physical size decreases • Scatter Intensity ∝ Particle Size ^6 • Ex: If particle size decreases by half, the scatter intensity drops to 1/64 • Elimination of background noise to improve accuracy / repeatability • Electronic, chemical matrix, polymers, surfactants, contamination, etc. • Efforts to increase particle signal also increases background noise • Separation between particle signal and background noise (S/N ratio) • Need to increase particle signal while simultaneously decrease background noise ULTRAPUREMICRO2018.COM 6

  7. Counting Efficiency, Actual Flowrate, Effective Flowrate • Counting Efficiency • The percentage of total sample volume that is measurable at a particular particle size • 100% counting efficiency – 100% of the sample liquid passing through the sensor is measurable • 1% counting efficiency – 1% of the sample liquid passing through the sensor is measurable • May vary from size channel to size channel and instrument to instrument. • Dependent on technology, design, and engineering of each instrument • Independent of the actual flowrate • Actual Flowrate / Actual Volume • The total sample flowrate or total sample volume through the sensor • 10 mL/min into the sensor • 1,000 mL/min into the sensor • Independent of counting efficiency • Effective Flowrate / View Volume • Effective Flowrate = (Actual Flowrate)(Counting Efficiency) • View Volume = (Actual Volume)(Counting Efficiency) • 10 mL/min at 100% efficiency = 10 mL/min effective flowrate; or 10 mL View Volume in 1 min. • 1,000 mL/min at 1% efficiency = 10 mL/min effective flowrate; or 10 mL View Volume in 1 min. ULTRAPUREMICRO2018.COM 7

  8. Design Example to Increase Particle Signal • Below designs and specifications are based on liquid particle counters from RION CO., LTD. (Japan) 0.2µm 0.1µm 0.05µm 0.03µm ~ Signal 1 1 1 Compared to N/A 4,096 87,791 64 0.2µm Laser 780 nm 830 nm 532 nm 532 nm Wavelength Laser Output 40 mW 200 mW 500 mW 800 mW Power Counting 100% 70% 10% 5% Efficiency Data 100 in 100 in 100 in 100 in Example Reads 100 Reads 70 Reads 10 Reads 5 ULTRAPUREMICRO2018.COM 8

  9. Counting Efficiency Across the Dynamic Range • Depending on the design, there could be different counting efficiencies at different particle sizes across the dynamic range • Flat / even laser intensity distribution = same / similar counting efficiencies • Distributed / uneven laser intensity distribution = different counting efficiencies • Consult the particle counter manufacturer for the counting efficiencies at the particle sizes for comparison ULTRAPUREMICRO2018.COM 9 * Courtesy of RION CO., LTD. (Japan)

  10. Counting Efficiency Across the Dynamic Range • Need to know the counting efficiency across the dynamic range • Consult the particle counter manufacturer for the counting efficiencies at the particle sizes for comparison • Ideally, the counting efficiencies of a particle counter is substantially the same across the dynamic range Effective Flowrate 0.05 0.2 Particle Sizes ULTRAPUREMICRO2018.COM 10

  11. How to Normalize Data for Comparison • Good Practices for Test Planning and Experiment Setup • Minimize variables • Batch sampling is preferred (control and consistency in sample) • Inline sampling can be done, but will require very precise particle injection system (SEMI-C-77-0912) • Measure samples from same bottle • Use same inlet / outlet lines • Use same flow control • Minimize disturbance to the sample and lines • Minimize contamination • Pre-rinse, flush, and baseline setup • Clean environment (cleanroom, laminar flow hood, clean zone systems) • Gloves, masks, etc. • Avoid touching the liquid contacting portions of the inlet tubing • Minimize the time between sampling • Run all tests in one setting • Poly dispersed sample solution • Ex: Drop(s) of city water into bottle of filtered DIW • Measures all channels at once with one sample ULTRAPUREMICRO2018.COM 11

  12. How to Normalize Data for Comparison • What You Need: • Raw data • Counting efficiencies at each particle size for comparison • Actual flowrate • Measurement time • Determine Sample Volume: 𝑛𝑀 Sample Volume 𝑛𝑀 = Actual Flowrate × Measurement Time (𝑢) 𝑢 • How to Normalize Data for Comparison (to 100%) 𝐷𝑝𝑣𝑜𝑢𝑡 Normalized = Raw Data 𝐷𝑝𝑣𝑜𝑢𝑡 ÷ SampleVolume 𝑛𝑀 ÷ Counting Efficiency (𝑣𝑜𝑗𝑢𝑚𝑓𝑡𝑡) 𝑛𝑀 • How to Normalize Data for Comparison (from unit A to unit B with different counting efficiency) • This method is used to minimize the normalization factor(s) used and/or when both A & B have low counting efficiency 𝐷𝑝𝑣𝑜𝑢𝑡 = Raw Data A 𝐷𝑝𝑣𝑜𝑢𝑡 Counting Efficiency B Normalized to B × 𝑛𝑀 𝑛𝑀 Counting Efficiency A ULTRAPUREMICRO2018.COM 12

  13. How to Normalize Data for Comparison: Examples 𝑛𝑀 Sample Volume 𝑛𝑀 = Actual Flowrate × Measurement Time (𝑢) 𝑢 𝐷𝑝𝑣𝑜𝑢𝑡 Normalized = Raw Data 𝐷𝑝𝑣𝑜𝑢𝑡 ÷ SampleVolume 𝑛𝑀 ÷ Counting Efficiency (𝑣𝑜𝑗𝑢𝑚𝑓𝑡𝑡) 𝑛𝑀 Particle Raw Counting Actual Meas. Sample Volume Normalized Data Size Data Efficiency Flowrate Time 100 𝑑𝑝𝑣𝑜𝑢𝑡 10 𝑛𝑀 100 𝑛𝑗𝑜 × 1𝑛𝑗𝑜 = 10 𝑛𝑀 10 𝑛𝑀 = 14.3 𝑑𝑝𝑣𝑜𝑢𝑡/𝑛𝑀 0.1um 70% (0.7) 10mL/min 1 min counts 0.7 7 𝑑𝑝𝑣𝑜𝑢𝑡 10 𝑛𝑀 7 𝑛𝑗𝑜 × 1𝑛𝑗𝑜 = 10 𝑛𝑀 10 𝑛𝑀 0.1um 5% (0.05) 10mL/min 1 min = 14.0 𝑑𝑝𝑣𝑜𝑢𝑡/𝑛𝑀 counts 0.05 * Consult the particle counter manufacturer for counting efficiency ULTRAPUREMICRO2018.COM 13

  14. How to Normalize Data for Comparison: Examples ULTRAPUREMICRO2018.COM 14

  15. How to Normalize Data for Comparison: Examples ULTRAPUREMICRO2018.COM 15

  16. How to Normalize Data for Comparison: Examples ULTRAPUREMICRO2018.COM 16

  17. How to Normalize Data for Comparison: Examples ULTRAPUREMICRO2018.COM 17

  18. How to Normalize Data for Comparison: Examples DIW DIW 1000000 100 Unit A (1%) 0.05μm Number of Particles ( N/10mL ) Particle Numbers (/mL) 100000 0.1μm 10 0.1μm 10000 0.15μm 1 0.2μm 1000 0.3μm 0.1 100 0.5μm Unit B (70%) 0.01 0 4 8 12 16 20 10 0.01 0.1 1 Time (hours) Particle Diameter ( m m ) Unit A (1%) Unit B (70%) ULTRAPUREMICRO2018.COM 18

Recommend


More recommend