Troubleshooting for liquid particle counting
How to ensure that quality data is being obtained from optical particle counters (OPCs) in liquid applications
By Roger Carlone, Particle Measuring Systems
Ensuring that your liquid optical particle counter (OPC) is operating within its designed specification is critical to obtaining quality data. As data from OPCs becomes increasingly important to maintaining process control, it becomes more critical than ever for users to have confidence in the data being generated. OPC data is frequently used in qualifying process improvements, in control during chemical manufacturing, and in quality control for production of complex parts. It is not uncommon for users to question the results from their particle counters. Discussed in this article are common misunderstandings and other performance issues that would lead one to question their OPCs.
Users can take several easy steps to ensure that quality data is collected. All OPCs have defined and easily controlled specifications. These specifications are flow rate, background scatter, and concentration limit. Also important is the particle size distribution (PSD). Unless all of these parameters are correct, the data should be considered suspect. Even if one parameter is incorrect, the data quality has been compromised. This article summarizes these parameters and their effects on each other, and it also covers ultimately ensuring that quality data is achieved when analyzing ambient particle distributions in process chemicals and deionized (DI) water.
Normal particle distributions
Thirty years of particle counting experience shows that most ambient particle distributions in continuously filtered liquid systems follow a D-3 distribution (D = diameter) on total cumulative counts. Figure 1 shows a normal particle distribution. In extremely clean DI water systems, the distribution can be steeper (D-4), and in dirty systems, the distribution can be flatter (D-2). It is equally important that whatever the distribution is, it remains consistent from sample to sample.
When analyzing particle data, it is essential to ensure that you have a normal particle distribution. If not, it should immediately alert the user to a potential problem. Things that can lead to abnormal distributions or poor data are incorrect flow rate, excessive background scatter, and exceeding the OPC’s upper concentration limit. In addition, real particle events such as pump or valve failures, o-ring problems, damaged tubing, or other process upsets can cause abnormal particle distributions. This is why it is critical to ensure that quality OPC data is collected. Otherwise, one would not be able to identify a real particle excursion from a poorly performing OPC.
All OPCs have a flow rate specification. The specification is most
commonly listed as the set flow rate plus or minus a certain percentage; for example, 50 ml/min ??10 percent. The actual flow rate varies by model.
OPCs are calibrated at a specific flow rate. This means that real-world particles are intended to pass through the particle counter at the calibrated flow rate. If the proper flow rate is not maintained, both sizing and counting accuracy are compromised. As a particle travels through an OPC, it scatters light. The amount of scattered light is converted to an electrical signal, which is counted and measured to determine the particle concentration and size.
If the flow rate is set too high, the particles travel through the OPC too fast. Since the transit time is decreased, the electronics do not have sufficient time to fully integrate the signal, compromising sizing accuracy. Counting accuracy is compromised because the resulting signal is now too small to exceed the particle size threshold. Further complicating the issue is that the accompanying software is normalizing the data to the specified flow rate. Conversely, if the flow rate is too slow, the transit time is increased and the particle appears to be larger than its actual size. It is critical that the OPC be operated at the correct flow rate.
Table 1 shows the effects of changing the flow rate of an OPC. The specified flow rate is 50 ml/min.
Depending on the application and equipment used, there are three ways to control the flow rate through an OPC:
- Syringe samplers require the user to input the correct syringe volume and sample speed. This offers extremely accurate flow control.
- Online applications require a flow controller downstream of the OPC set to the proper flow rate. It is important to calibrate these flow controllers with a stopwatch and graduated cylinder. The accompanying software has no means to set the flow rate nor does it know if the OPC is being used at the correct flow rate.
- Compression samplers have a flow rate adjustment on the instrument to set the flow rate. In this case, the software will report the actual flow rate at the end of a sample.1 A detailed explanation of operating a compression sampler can be found in Particle Measuring Systems’ Application Note 56, Counting Challenge.
There is always a small amount of background noise or scattered light in an OPC that originates from the laser interacting with the optical path. The scatter from the capillary walls or optical windows can be measured. Normally, the amount of scatter from the capillary walls is minimal and has no effect on the data. As the capillary becomes contaminated or, in severe cases, damaged, the amount of scatter increases. In other designs, the optical windows from which the sample is viewed can become contaminated. Eventually, the amount of scatter coming from contaminated optics equals the amount of scatter from the smallest detectable particle size. In most cases, simple cleaning of the capillary can resolve this issue.2
There are two ways to recognize whether this is occurring. First, the accompanying software reports the amount of background scatter. Each OPC has a specification for this value. Once this value has been exceeded, the collected data should be considered suspect.
Additionally, if the amount of scatter from contamination equals the amount of scatter from the smallest detectable particle size, there will be an excessive number of counts in the first channel. One will often see hundreds to thousands of counts in only the first channel on known clean DI water, with zero to little counts in the other size channels, as shown in Fig. 2.
Users often ask if they can ignore the first channel. The answer is “no” for two reasons. First, the OPC electronics become busy processing the noise signal, which prevents it from accurately processing real particle signals. Second, there is a constant amount of light scatter hitting the detector. The OPC is designed to measure total cumulative counts from the smallest detectable size and greater. The electronics and signal processing are not designed to handle such situations, resulting in compromised data across all channels.
Upper concentration limit
All OPCs have an upper concentration limit. This is the maximum allowable concentration of particles that the particle counter should be exposed to. This is typically expressed as <10 percent coincidence loss at 10,000 particles/ml total cumulative counts. This means that at 10,000 particles/ml, up to 10 percent of the particles are not being counted.
Optical coincidence is defined as when more than one particle passes through the viewing volume of the OPC simultaneously. As the concentration of particles starts to reach the upper limit, optical coincidence begins to occur. Particles can pass side by side and appear larger, or smaller particles can hide behind bigger ones. The end result is that as more and more particles pass through the OPC, the OPC loses the ability to distinguish one particle from another. As this occurs, the distribution of particles changes and typically appears as if there is the same number of counts in the first several particle channels of the OPC as seen in Fig. 3.
Once you exceed the concentration limit, there is no guarantee of the actual counts or that two instruments will agree with each other. Coincidence loss is neither predictable nor necessarily a linear response as the concentration increases.
In severe cases, if the particle concentration is extremely excessive, the OPC will saturate. When saturation occurs, the electronics are completely overloaded, so they no longer have the ability to respond to individual particles. The particle distribution will actually invert and show fewer counts in the first channel as seen in Fig. 4. In such cases, the OPC can actually read zero counts.
A phenomenon that can occur in OPCs with less than 0.1 µm sensitivity is false counts due to cosmic rays. Cosmic rays are normally occurring, natural events. When a cosmic ray hits the photodetector, it will register as a particle. The rate of cosmic rays hitting the detector varies by location and environmental factors but generally occurs about once per minute, resulting in less than 1 count/ml. It is also common for these counts to appear in the second channel of the OPC, which in most cases is 0.1 µm.
In normal applications this is not significant. However, most DI water systems–and even some process chemicals used in semiconductor manufacturing–have normal concentrations of particles less than 1 to 2 counts/ml. If the OPC is otherwise within specification, and the normal concentration of particles is less than 1 count/ml, the resulting distribution of particles can appear as in Fig. 5. It is important to note that most OPCs have a zero count spec of 1 to 4 counts/ml. This means that anything below this limit should be considered noise.
Proper operation and adherence to published specifications of an optical particle counter are critical to obtaining quality data. Without quality data, one cannot make informed decisions regarding the control of his or her processes.
A few simple steps can ensure that quality data can and is being collected. First, one should look for a normal particle distribution. This is a good initial indicator of quality data. Even with a normal distribution, one should check that the OPC is operating with the correct flow rate and within the published specifications of background scatter and particle concentration.
Roger Carlone is an applications engineer at Particle Measuring Systems (www.pmeasuring.com) based in Boulder, CO.
- App Note 56, A Counting Challenge.
- App Note 55, LiQuilaz Maintenance: Capillary Cleaning Procedure.