The rapid detection of point-source contamination locations within the cleanroom can have significant benefits for defect reduction efforts. Depending upon the volume of production and the cost of the product, financial implications of delayed detection can quickly become a serious problem. A multi-dimensional strategy utilizing real-time point of use and mobile monitoring strategies can efficiently detect and locate sources of contamination within any critical manufacturing environment.
All airborne molecular contamination (AMC) monitoring technologies require three similar conceptual steps in order to accomplish an analysis and provide a concentration value representative of the contaminants present in the environment. These steps are: sample collection, sample analysis, and results reporting.
In the cleanroom, a beneficial feature of any monitoring tool is real-time data. With monitoring technologies that utilize a capture- then- analyze strategy, such as impinger plus ion chromatography, it is impossible achieve real-time results. Wait times between sample collection and delivery of results are typically measured in days, if not weeks. The likelihood of the contamination control engineer being able to correlate the result back to a specific, often episodic event in the manufacturing environment is very low. When out-of-specification or unexpected results are obtained, having the ability to correlate the result back to an event or even a process-specific event is paramount if meaningful troubleshooting and corrective action steps are intended.
Another available sampling technique utilizes a multipoint manifold for sample collection and subsequent chemical detection. The key benefit here is that analyzer equipment costs can be spread over multiple sample locations, reducing cost per sample location. One key trade-off is invisibility of concentration changes while not sampling.
Because of the amount of time it takes a manifold system to purge and respond to variable concentrations as it moves from one sample to the next, the interval spent monitoring each location is typically very short. As an example, consider that a 30-point manifold may only capture a truly representative sample at each location for 6 minutes per day, with the remaining time spent either purging that sample line or sampling from other locations.1
A third monitoring approach often used in critical manufacturing environments is real time, point-of-use monitoring. The relative low cost per point of a point- of- use analyzer has enabled widespread adoption in AMC monitoring programs. Point-of-use analysis provides continuous data and ensures that an AMC contamination event in the vicinity of the analyzer will be detected quickly, usually within one minute of the event. Software can alert the contamination control engineer to the event, who can then take action.
When the engineer does take action, often one of the first steps taken is to employ an additional or referee analysis to verify that the result is real and not the product of an unrelated transient condition within the space, or a problem with the analyzer reporting an out-of-specification result. Depending upon the measurement technology in use at the facility, deployment of this referee measurement is often a challenge. The primary limitations of the collect-then-analyze and the multipoint manifold measuring techniques are delay and fixed locations respectively. In the case of the analyzer used, the key limitation is the time to establish system stability. Depending on the state that the analyzer was stored (powered versus unpowered), the time to stability can vary from two hours to overnight. The detection system used for this study solved this problem by providing a self-powered mobile utilities platform from which multiple point-of-use analyzers can be deployed as quickly as a mobile cart can be moved to the problem location. The self-powered feature was especially useful to quickly verify a concentration currently being reported by another analyzer and to check around tools and other point source contamination locations within the manufacturing environment for suspected sources of AMC.
Continuous data, measured using an AirSentry II AMC monitoring system from Particle Measuring Systems from within a clean manufacturing environment, is displayed in Figure 1. The time from midnight to 7 AM represents typical background amines concentrations (1). The step variation is due to Tool A in the vicinity that is known to emit low concentrations of amines on a periodic cycle. At the end of the shift, Tool A is powered down, and a typical dip in the reported amine concentration is noted during shift change (2). During the following shift (3), construction activity in the vicinity of the sensor is noted to increase the background amines concentration; this is most likely due to increased ammonia emissions from the additional personnel.
At approximately 16:30, (4) a notable spike in the reported amines concentration alerted the contamination control engineers of a nearby contamination event. At this point, a mobile referee unit was deployed to verify the measurement and map the concentration in the general vicinity of the detected event.
Figure 2 shows three maps of the vicinity of the event. The color indicates the relative amines concentration: green through red indicating increasing concentrations. The first map (A) displays standard monitoring locations that are used for historical spatial concentration analysis. At these locations power is available and the mobile system is connected to it to prolong the internal uninterruptable power supply (UPS). Map B displays concentrations typical of background concentrations in the manufacturing environment. Map C clearly shows the concentration gradients in the vicinity of the point source. This information was used by the contamination control engineer to identify a leaking process control valve in the tool identified by the blue four point star in Map C. As shown in figure 1, after the source of the leak was repaired (5), the point of use unit reported amines concentrations returning to normal background levels (6).
In this example, a multi-dimensional monitoring approach utilizing both point-of-use monitoring and an AMC detection system enabled a significant point source contamination event to be verified, identified, and isolated within a two-hour window.
The AMC detection system provided a self-powered, mobile utilities platform capable of supporting up to three point-of-use ion mobility spectrometer analyzers for up to 60 minutes without the need for a connection to building electrical power. Where quick verification of an out-of-specification result is required in clean manufacturing environments, or where a contamination control engineer has a need to sniff for AMC in difficult-to-reach areas near and around tools, this mobile approach enables accurate, reliable AMC measurements as quickly as the instrument can be moved from one location to another within the facility. With this ease of portability coupled with zero warm-up or clear-down time for the different analyzer chemistries, contamination control engineers have a powerful verification tool that substantially increases the probability of correlating a high reading to a source of AMC in critical clean manufacturing environments.
- Rowley, Steven. “In AMC Monitoring, Simple Designs Go a Long Way.” 29 Feb. 2008. Particle Measuring Systems. 26 Apr. 2012.
Mark Berdovich is currently Applications Engineer for the Americas for Particle Measuring Systems, Inc. Mr. Berdovich has 17 years’ experience in cleanroom and process microcontamination monitoring, design, and construction, and has a B.S. degree in Mechanical Engineering from the University of Illinois at Urbana-Champaign. He may be reached at firstname.lastname@example.org.
Thomas Pietrykowski is currently a Product Marketing Engineer for Particle Measuring Systems, Inc. Mr. Pietrykowski has 11 years’ experience in analytical instrumentation research and development, and holds a B.S. degree in Environmental Chemistry from Lake Superior State University in Sault Ste. Marie, MI. He may be reached at email@example.com.