Ensuring the right result

Validating cleanrooms using particle measurement technology can be problematical, but it is posssible to obtain reliable results. Leander Mölter of Palas puts the technology under the microscope

The degree of cleanliness in cleanrooms depends on the extent of the purity, precision or high quality of the goods produced. There are different classes of cleanliness defined according to international conventions. In practice, a cleanroom has to comply with the requirements of these classes and this compliance must be proven according to basic technical rules.

The following requirements are part of generally acknowledged basic rules for quality control in measurement technology (see DIN standard 1319, ISO standard 9000, EN standard 45000): • The measurement uncertainty of each measurement device must be known • Each measurement result has to include an indication of the measurement uncertainty The indication of uncertainty of the whole measurement process is required. However, due to a lack of practical methods, several common mistakes that might occur during the evaluation of measurement uncertainty in particle measurement technology have not been considered at all, or only in a very approximate way. For example: • Uncertainties during the production of test aerosols • Particle losses during the transport of test aerosols • Sampling mistakes • Uncertainties during the calibration of optical particle counters So why are cleanroom counters, particle generators and dilutions systems used for particle technological acceptance tests and validation of cleanroom plants? The reasons are explained in Figures 1 and 2. Fig. 1 illustrates a simple cleanroom which shows how a particle technological validation of cleanrooms works. There is a pre-filter (EU 10 quality) with an assumed penetration of 15% within a particle range from 0.1 to 0.3µm. The final filter (EU 14 quality) is chosen with an efficiency grade of 99.995%; i. e. penetration is 5 x 10-5. Fig. 2 shows the range of particle concentrations for cleanroom classifications. The aim of such a classification is to determine the concentration of the surrounding air or of a test aerosol. Usually, the particle concentration is measured in particles/cm3, particles/m3 or, if the measurement is carried out with a cleanroom counter, in particles/ft3. 1m3 = 106cm3 1ft3 = 28.400cm3 1m3= 35.21ft3 (see Table 1) Fig. 2 shows that the lower measurement limitation of 0.1µm for OPC1 and 0.2µm for OPC2 influences the necessary dilution factor for the upstream measurement considerably.

Why dilution systems? The following example illustrates why it is necessary to use dilution systems when validating a cleanroom. In this example, the outside-air concentration according to Junge serves as the orientation value (see Fig. 2). The real concentration of outside air might be considerably higher or lower, depending on the surroundings (e.g. higher near a cement plant, lower on the top of a mountain). Particle concentration and particle size distribution of the outside air might also differ at different times of the day and on different days within a week or a month. Fig. 2 shows that the aerosol needs to be diluted for measurements carried out with OPC1, a clean-room counter which only measures 7 particles/cm3 or 7 mio particles/m3. The lower measurements range of OPC1 is 0.1µm. A measurement without coincidence error can only be achieved with this counter if, according to Junge, the outside-air concentration is diluted by a factor of 1000. The values of any OPC can be added to Table 2 in order to find out the necessary dilution factor. Fig. 1 shows the scheme of a cleanroom plant, including particle technological components. For the following calculations, we assume an external air concentration as measured with OPC1, which is 7000 particles/cm3. The pre-filter is a filter of quality EU 10 and 85% of the particles within a range from 0.1 up to 0.3µm are separated by this pre-filter. Thus, the particle concentration behind the pre-filter (or before the final filter) is 1,050 particles/cm3. In order to make OPC1 measure correctly, the dilution factor must be determined and the outside air must be diluted accordingly.

Dilution Factor = Cup = 1,050 = 150 Cdown 7 Cup = concentration upstream Cdown = concentration downstream

Without dilution air, mixing air or a dilution system, the upstream measurement in this case would be incorrect. The maximum measurable concentration of the cleanroom counter used would be exceeded by a factor of 150 and consequently, the upstream concentrations value would be much too low. Thus, the filter would be classified as much worse than it really is, which is why it is essential to dilute the upstream concentration when using a cleanroom counter. If the goal of a measurement is to determine the fractional efficiency at one spot of the final filter, loaded with the aforementioned ambient air concentration, the volume flow of the cleanroom counter must be defined, e.g. to 1 ft3/min. This has been found to be the only way to calculate the number of particles that can be measured within one minute behind the final filter.

Why aerosol generators? In the following example, we assume the above-mentioned value of 1,050 particles/cm3 is the concentration: 1,050 particles/cm3 x 28,400 cm3/ft3 = 29,820,000 particles/ft3 upstream

The separation efficiency of an EU16 filter is 99.99995% or a penetration of 5 x 10-7. In order to find out the number of particles per minute to be found behind the final filter with this upstream concentration, the concentration must be multiplied by the penetration of the final filter. The upstream concentration is calculated as follows:

Cup x penetration = Cup x P = 15/ft3

Cup = concentration upstream (P = penetration)

Approximately 15 particles/min are measured downstream at VOPC = 28.4 l/min. Such a low particle concentration is not suitable for a statistically exact measurement (see Poisson distribution, VDI guideline 2083). So to keep the downstream measurement time short and to receive statistically exact results, high-concentrated test aerosols are used to generate a defined upstream concentration. If a cleanroom counter with a volume flow of 0.1ft3/min or 0.01ft3/min were used in the mentioned case, only 1.5 particles/min or 0.15 particles/min would have been measured. That means, if the measurement set-up remains the same and the OPC is replaced by several OPCs with different flow volumes (e.g. OPC with 1ft3/min. replaced by OPC with 0.1ft3/min or 0.01ft3/min), the duration of the downstream measurement must be extended by factor 10 or 100 in order to receive results that are statistically correct. Due to this statistical uncertainty it was internationally agreed that an artificial aerosol with special characteristics must be used for acceptance tests and validations of cleanrooms. The economical and reproducible acceptance test is made possible only by using such test aerosols. To determine the fractional grade efficiency of HEPA and ULPA filters, a high-concentrated test aerosol with known particle size distribution and concentration must be used. Typical concentrations of test aerosols are within the range from 106 to 1010cm-3 and the user of a generator that produces such particles needs to be informed about the measurement criteria used to characterise the generator. It is vital to note that in order to avoid coincidence errors, coagulation problems or unnecessary contamination of the filter, the upstream concentration should not be chosen at random. It is, however, important that the upstream concentration is detectable by a commercially available measuring device. It is also important to know that, according to EN standard 1822, page 3, a test of filter media must be done within the so-called MPPS-range (MPPS = Most Penetration Particle Size). A test aerosol that serves for validation and acceptance tests of cleanrooms does not need to fulfil this requirement. However, a test aerosol with particles smaller than 2µm has specific advantages. Another essential fact to remember is that the upstream concentration is not distributed constantly over the surface of the final filter. In Fig. 1, the upstream concentration is lower on the right side above the final filter than it is on the left side.

Conclusion To reliably validate a cleanroom, it is necessary to use a cleanroom counter, an aerosol generator and one or several dilution systems. There are different characteristics of a cleanroom counter which influence its measurement results, such as resolving power, sizing accuracy, border-zone-error, coincidence error, counting efficiency, possible ambiguities and counting error rate. Unfortunately, definite information about these important characteristics is not given by manufacturers of the counters. In order to minimise this problem, it is very important to use well-defined aerosol generators and dilution systems. Thus, there remains only one instrument in the process with a certain uncertainty. It is important, therefore, that dilution systems and aerosol generators fulfil this requirement and are supplied together with a calibration certificate. Only if these instruments are unambiguously characterised is it possible to carry out reproducible measurements and to validate cleanrooms reliably.