The ability to measure the cleanliness of objects of various shapes and materials is important in maintaining a comprehensive contamination control program. This allows a Contamination Control Engineer (CCE) to qualify the effectiveness of different parts or cleaning vendors to make direct cleanliness comparisons of different commodity items (gloves, wipes, ect.), and to evaluate the performance of alternative cleaning recipes on material cleaners.

LSI Logic worked with Particle Measuring Systems (PMS) to develop a machine that provides a repeatable, sensitive testing method. This machine, the Surfex™, uses a DI water recirculating filtration tank, coupled with ultrasonic transducers and a liquid particle counter. The tank is housed in an ISO Class 2 minienvironment.

This article discusses the characterization methodology used to determine the optimal process settings for various materials to be measured, and provides two examples of normal use investigations.

The performance of the machine needed to be characterized before it could be put into use as a regular metrology tool. Ultrasonic power settings and frequencies had to be determined for maximum particle removal effectiveness, coupled with a minimum of surface damage. Recipes needed to be developed for varying materials (aluminum, stainless steel, and plastic). Additionally, a method for verifying the quality of the data needed to be developed.

The machine (Figure 1) includes a DI water recirculating filtration system, an ultrasonic tank with ultrasonic transducers, a liquid particle counter, and an ISO Class 2 minienvironment. The tank is designed to accommodate 1.0 cu ft sized objects. A secondary rack is provided to suspend smaller vessels into the tank to decrease particle dilution levels for small objects. The transducers are capable of running at various power settings and frequencies. The liquid particle counter is capable of measuring particles from 0.2 µm to 2.0 µm. The DI water is particle-counted before and after the ultrasonic-enhanced extraction. The change in particle counts is attributed to the tested object.