Calibrating Flowmeters for Use in Measuring Contamination in High-Purity Process Liquids
David B. Blackford and Donald C. Grant, CT Associates, Inc.
Norman Neher and Jeffrey S. Franz, The Futurestar Corporation
The cleaning and etching processes used to manufacture semiconductor microcircuits require numerous steps employing a variety of hazardous liquid process chemicals. Many of these steps are adversely affected by the high levels of particulate contamination typically found in liquid chemicals. Particle contamination in these chemicals is usually measured using optical particle counters that report particle concentration as the number of particles per unit volume (traditionally, particles per milliliter or per liter). The accuracy of the particle concentration depends on accurately measuring the flow of the liquid chemical through the particle counter. Float-type variable-area flowmeters are often used to measure the flow rate. Because of the corrosive nature of semiconductor process liquids, the flowmeter body and float must be made of Teflon PFA or another inert material.
Ideally, since a flowmeter’s calibration depends on the physical properties of the liquid flowing through it, each meter should be calibrated in each process liquid. However, flowmeter manufacturers typically supply a calibration for water only, so the user must recalibrate the flowmeter for each specific liquid chemical. Such recalibrations are hazardous, time-consuming, and expensive. The purpose of the study reported herein was to derive a correlation that allows a calibration curve to be developed for a particular flowmeter if the density and viscosity of the liquid are known. This study does not attempt to mathematically solve the complex boundary-layer problem in variable-area flowmeters using fundamental fluid mechanics principles. Due to the wide variety of flowmeter and float designs, any attempt to derive a complete solution would require too many assumptions and approximations to be of practical use.
CTA publication #20 Micro, 13(6):53-62, 1995