Controlling The Purity of Purified Water

Pindar wrote 2,500 years ago that “of all things water is best”. He then diverted to write about gold, sport and music but my interest today is in water and, specifically, the purified water in your laboratory. Are you sure that it is the best? Is it what you are expecting? Is it fit for purpose? Are you confident that the impurity levels are as low as you need them to be?

Any real confidence can only be based on the use of a water purification system designed to achieve at least the purity required and designed to allow effective monitoring of the product water purity.  Don’t all water purification systems do that? Unfortunately, they do not. Although most use similar purification technologies, the use of reliable components made of the right materials and good design principles are not universal.

Confidence through Monitoring or Design?

Unavoidably, both are essential for real confidence.

Monitoring alone might be enough if all the potential impurities could be monitored in real time but, as summarised in Table 1 , only resistivity (a marker for ionic impurities) and TOC (an indicator of organic contamination) can be monitored rapidly enough. Testing for bio-active species, including bacteria and endotoxins, is carried out off-line. On-line particle counting is far too slow and expensive. For these types of impurities good design is essential to minimise the risks of contamination and reliance on monitoring.

Design Features

Pure water is like a vacuum. A vacuum can only be maintained by pumping out the residual gases and, similarly, ultrapure water can only be maintained at its highest purity by on-going purification. This can only be achieved by preventing impurities from entering the purified water during storage, by repeatedly recirculating it through the best purification technologies and sanitising when needed.

Recirculation can produce dramatic improvements in bacterial contamination as illustrated below. Initially sterile, purified water was stored in two sterile 25 liter reservoirs fitted with protective filters. In both cases water was dispensed regularly and replaced with sterile water.  One reservoir was kept without recirculation and sampled asceptically. The water in the second reservoir was recirculated intermittently through a UV chamber and a pack of ion-exchange resin. Samples were taken from the reservoir and after the recirculation loop. Bacterial levels in the static reservoir were high ranging from 4 to over 1000 CFU/ml.  The counts in the recirculated reservoir were an average of 2.1 CFU/ml, highlighting the much lower load on any final treatment or filter. The best results were obtained from samples after the purification technologies – typically 0.1 CFU/ml or less.

Monitoring

There are various aspects of monitoring which are not normally discussed by manufacturers of water purification equipment.

1. The monitoring should be of the product water itself.
2. When dispensing water, the monitor should provide “real-time” data before the water volume is completely dispensed so that any problems with its purity can be detected before the water is used.
3. The monitoring should be as representative of the volume of water used as possible.

Product Water Monitoring

As specified by CLSI and USP  (and what clearly makes sense) the water purity should be monitored as close to the dispense point as possible and, definitely, after all major purification technologies. The monitors should also be located so that they are part of any water recirculation loop ensuring that they are kept clean and will respond rapidly when water is dispensed. The only technologies accepted as point-of-use (POU) devices are micro-filters which serve to prevent back contamination by bacteria. They are essentially physical barriers made from high purity materials and have been found not to degrade the purity of the water as long as they are changed or autoclaved regularly.

The practice of fitting only micro-filters as point-of-use devices in laboratory water purification systems had become the norm until recently when a manufacturer decided to fit a range of POU chemical treatments on the outlets of their dispensers, long after any monitoring. This has to be viewed as a retrograde step. Such POU treatments introduce a large volume of reactive and high surface area media after all monitoring.  In these situations there is no control of product water purity and this problem is compounded by the lack of any water recirculation through these POU treatments to keep them clean. Build up of contamination over time is inevitable and it is highly doubtful if this would meet CLSI, ASTM and USP standards in view of the lack of definition of water purity.

Timely and Representative Monitoring

Regarding the timeliness of monitoring data, it is, to say the least, highly desirable that any lowering of water purity is evident while water is being dispensed. Once the operator has taken water from the purifier they are highly unlikely to return some minutes later to check that the water is still good! This is especially true, as they have no reason to believe that the data they saw, and possibly logged, is only historic and not necessarily related to the water they have taken.

Fortunately, resistivity detectors respond very quickly to any major change in ionic contamination and any such changes could be apparent during dispense.

However, this is not the case for TOC. For laboratory water purification systems, TOC monitors are based on the same principle: oxidization of organic molecules produces acids and carbon dioxide. These species are conducting and are detected by an increase in the conductivity of the water. Different monitors approach the generation and measurement of this conductivity in various ways. Some take a slow sample stream of water from near the dispense point which they oxidize and measure the change in conductivity. This can be quite rapid but is a function of the speed of the side-stream and the sample is also limited to the volume measured. One manufacturer rinses water through a cell, takes a small sample and oxidizes it over a period of several minutes before reporting the result. In this case, the result is from a very small sample and the resultant delay means that results are not available during dispense.  The ELGA TOC monitor uses conductivity changes produced by the UV oxidation system which purifies the water to monitor the resultant product water TOC. This enables an extremely rapid response based on the entire water being dispensed.

It is important that the TOC monitoring is truly continuous and does not miss any change in contamination level that could ruin an analysis. The response of the monitor in ELGA’s PURELAB Ultra (in dark blue) to a sudden change in TOC (in red) is shown below. The response of the TOC monitor closely mirrors the TOC of the water dispensed.

Conclusion

Reliable water purity relies on good system design to minimise contamination and to enable effective purity monitoring. Rapid monitoring of key parameters after all new purification technologies is essential for confidence in your water purity. As Pindar, doubtless, would have said, if he had been alive today, “of all reagents in the laboratory, pure water that can be shown to be fit for purpose is the best”. Or, put another way, it is strongly recommended that you consider whether the purity shown on your purifier is really the purity of the water you are dispensing. Not as pithy but, nonetheless, very important.