Networking Offers Wealth of Practical Experience
HAROLD D. FITCH
A few days ago I had lunch with Jack Hedman, who is a long time friend and contamination control compatriot. He has over 23 years hands-on experience in contamination control and trouble-shooting actual manufacturing applications. This month, we`ll share some of Jack`s key observations about contamination control. In upcoming columns we will delve into some actual case histories that will, hopefully, solve some of your own line problems or help you avoid future problems.
Welcome, Jack
Jack worked in quality control for nearly 20 years, then he spent another six years in site defect-density engineering. He is a well-respected contamination control practitioner whose accomplishments include presenting regularly at the annual corporate contamination-control seminar, contributing regularly to the corporate defect engineering newsletter, and auditing various sites throughout the corporation.
He worked extensively across several manufacturing lines, interfaced through purchasing with most of the measurement equipment manufacturers and set up internal programs for testing, calibration, repair, installation, and utilization of line testing instruments, such as gas and liquid particle counters.
Jack`s background and experience can be categorized into four areas of interest that we will cover in the future: 1) Air Flow Problems; 2) Chemical Monitoring; 3) Particle Counters; and 4) Continuous Environmental Monitoring Systems.
Jack`s major observations about contamination control include several points that I will detail below.
Contamination control requires good detective work, attention to minute details, and careful follow-up. A good example is the effort involved in ensuring proper airflow in a cleanroom situation. First, detective work is needed while measuring the actual airflow in the room, workstations, and above the product. It involves determining the actual flow pattern, the airflow velocity, temperature variations, and humidity conditions throughout the room. Then, a combination of past experience and airflow modeling ability is required to determine what changes have to be made to achieve the proper airflow pattern.
Once the changes are pinpointed, attention to every little detail ensures that velocities are correct, surface geometries are smooth and continuous, and no protrusions or interruptions to proper flow patterns are present. Unexpected leaks or stray air currents can disrupt the best design.
When you achieve the proper performance, a careful follow-up ensures personnel are trained to handle the conditions that must be met and necessary steps are taken to meet them. A well-designed room can easily grow out of control if not used properly. Some of the common problems that can occur are overcrowding personnel, changing flow patterns with storage of product or supplies, or changing a flow caused by a maintenance person removing a ceiling tile and not replacing it.
In the long-run, it is usually much more cost effective to understand, solve, and eliminate or control a problem than to apply a Band-Aid solution. An example is the situation we ran into when trying to extend the bath life of buffered hydrofluoric (HF) acid. When continuous recirculation and filtration were added to our HF baths, the early efforts removed the current buffering agent and made the etching reaction unstable. A simpler solution might have been to stick with the old method of changing baths after a few hundred wafers and not worry about it any further. This, of course, would not have been the proper course of action from the point of view of chemical conservation or improvement in etching control. In this case, the engineers pursued the problem and worked with suppliers to find a new buffering agent that was not removed by the filtration process.
This is fairly self explanatory but it can be seen in action by the use of a line contamination control index. In this case, the line personnel are taught to audit their line for violations. They keep a performance chart posted on violations over time. With proper management reinforcement, this approach has always brought the number of violations down in actual line situations and the line personnel are always proud of the results.
Continuous monitoring–using in-place, dedicated particle counters–is a key contributor to improving cleanroom performance. Jack, for example, had been responsible for pulling together a line monitoring air-particle counting program for his work site for several years. Over time, it had expanded from a couple dozen workstations and an occasional room count to several hundred workstations and numerous room counts–over hundreds of thousands of square feet of cleanroom space.
The monitoring was accomplished by utilizing portable particle counters and several quality operators on a continual basis over three shifts, seven days a week. The trouble-shooting activities, particle counter testing, calibration and repair, required almost all of Jack`s time. He worked with the various suppliers and trained the site instrument services department to do testing, calibration, and repair on-site. This was necessary for rapid turnaround time. Because the particle counters were repaired and calibrated at the manufacturer`s facility they were often out of calibration by the time they returned to the plant.
Because of this huge workload, any single workstation or room location was usually only monitored for a few minutes every two to three months. The introduction of dedicated, continuous monitoring particle sensors to critical workstations and room locations, improved on this situation in several ways.
First, it provided much more complete historical performance data on various locations. Major changes became readily apparent and time trends showed up as well. Looking at the data over 24-hour periods, showed the effect of shift changes, coffee and meal breaks, and events such as cleaning activities. Major disruptions could be tracked accurately. It could be determined if disruptions were continuous or cyclic in nature and the time of occurrence could be pinpointed. This made it much easier to track and eliminate the causes. Fitting continuous monitors with alarms also ensured that people were notified as soon as something went out of control.
Second, continuous monitoring reduced the manpower required for routine testing so more effort could be applied to advanced trouble-shooting and performance enhancement. It also required less engineering support for the day-to-day monitoring, therefore engineering could be applied to a site-wide continuous environmental monitoring system.
Third, the advancement in room and workstation control allowed more time to be applied to the more critical areas of tool and process control. n
Harold D. Fitch is president of Future Resource Development, a consulting firm (Burlington, VT) which specializes in cleanroom education and problem-solving.