Healthy Indoors Magazine
Issue link: https://hi.healthyindoors.com/i/1076124
Healthy Indoors | 29 cess and respond much more quickly to events, saving huge sums in many cases. The early instruments were very crude, incorporating only discrete logic chips (no controllers or onboard proces- sors) they would simply have manually calibrated thresholds that a technician would adjust, that would drive binary count- ers and then periodically shift these counts to small numeric displays before resetting themselves and restarting. There was no logging, no control, and very few features. These products were a commercial reaction to a pressing need and a number of vendors sprang up to supply these and meet the growing demand. There were no standards as to how this should be done, with each company developing their own solutions based on this basic principle. Over time features like logging, printing, external com- munication interfaces, more channels, smaller channels, etc. were added in response to client demands and in at- tempt to differentiate products. This took a few decades and over that time these instrumenst had huge impacts on production yields and quickly became absolute requirements in every semiconductor cleanroom. An enormous appetite for semiconductors had created a huge need for these in- struments. Eventually these products would find their way into all industrial cleanrooms (i.e. disk industry, life sci- ences, etc.). Of course, it only makes sense to measure what you're trying to control and the more important that control is, the more critical the mea- surement becomes. explosion of semiconductors in the 1960-1970s revolution- ized the world, circuitry that prior to this was large, bulky and drew loads of power was being quickly replaced by min- iature solid-state circuits produced at fractions of the cost. The miniaturization though meant the very small geometries in these circuits could be compromised by airborne (or liquid borne) particulates inadvertently deposited on the surface of the circuitry during the manufacturing process. Particulates above a specific size (dictated by the process geometries of the devices being produced) could cause shorts or opens within these circuits causing them to malfunction or fail. Since many of these circuits were manufactured simultane- ously (on the same silicon wafer) even short-lived air quality issues could destroy large numbers of these devices. In order to ensure that particles above a specific thresh- old size were kept to an absolute minimum, semiconduc- tor manufacturers built large, expensive cleanrooms with extensive filtration to attempt to ensure that the air in their manufacturing environments was kept as free of particu- lates above that threshold size as possible. Of course, installing filtration was no guarantee that this would indeed be the case. There are lots of ways that filters can fail and that clean environments can be compromised, so it was an ongoing battle to ensure that the environment was kept within the required tolerances for the process in question. The major tool that was used to ensure this was the optical particle counter. Simplistically, these instruments sample the air in an en- vironment by passing it carefully through a chamber where a light (typically a laser) shines a ribbon of light through which the air passes. If the air is perfectly clean it pass- es through the ribbon without scattering any of that light. However, when particles are present they scatter light as they pass through the beam, the amount of light scattered being proportional to the size of individual particles. A de- tector would measure the scattered light and update one or more internal counters based on the amount of light for that particle. So, in this way an instrument could count all the particles passing through an instrument and report on how many particles were seen during a sample period, sorting them in various size bins (or channels). The manufacturer could then look at these readings and determine whether the quantity of particulates above their specified thresh- old was acceptable. Having this informa- tion in real-time (as opposed to capturing a sample of air and sending it off for analysis) allowed them to more tightly control their pro-