When is pure water not pure?
Continuous TOC monitoring

Organic contamination of purified water can lead to significant problems in general analyses and process chemistry. Total organic carbon is the only practical measure of organic impurities, but its lack of specificity both as to the nature and quality of compounds present means that highly accurate measurements are not helpful. Much more important is the ability to monitor it continuously.

Organic contaminants—even at very low levels—can pose substantial problems in the lab. They can interfere with processes, lead to false results, and affect analyses. Even ultrapure water can still contain small, but significant, quantities of organic contamination. As a result it is important to monitor the levels of total organic carbon (TOC) so that problems can be anticipated. Some of the potential adverse effects of organic contamination are summarised in Table 1.
Contaminats
Ultrapure water is generally produced by treating a potable water supply by a
succession of different techniques designed to remove the various contaminants, whether bacteria, organic, inorganic or solid particulates. Organic contaminants are both naturally occurring and manmade. Natural contaminants include a complex mixture of fulvic and humic acids that result from the decomposition of leaves and grasses, and also come from the peat or marshland that the water may have
flowed through. Bacteria and other living
creatures are frequently present in water, with their by-products, such as endotoxins, providing further organic contamination.
Man-made organic contaminants can come from domestic and industrial waste, including pesticides, herbicides and fertilisers, as well as detergents, oils and solvents. And while the majority of these are removed by
purification to make the water fit to drink, further organic impurities are all too often introduced, such as the plasticisers that can leach out of plastic pipes or storage tanks, and other organics created by reaction with the ozone or chlorine being used to purify the water. The techniques used to create ultrapure water include microfiltration, reverse osmosis (RO), ion exchange, absorption and UV photo-oxidation in various combinations. While these techniques remove the majority of contaminants, they can leave behind small amounts of organic impurities.
Total organic carbon (TOC)
TOC is now the established measure of the levels of organic contaminants in water. It is generally sub-divided into:
-particulate organic carbon (POC),
-dissolved organic carbon (DOC) and
-volatile organic carbon (VOC).
These distinctions are largely irrelevant to measurements in ultrapure water, as filtration will have removed the particulates, and on-line TOC measurement techniques do not distinguish between the other two categories. Much more important is the relationship between TOC and the equivalent concentrations of the various organic compounds that are likely to be present in purified water. Some examples are given in Table 2.
As the proportion of carbon in likely contaminants ranges from around 10% to over 75%, the actual amount of unidentified contaminants that gives a TOC of 10 ppb can vary widely. It might be a mixture of
25 ppb urea and 50 ppb chloroform, or
6.6 ppb phenol and 9.6 ppb ethanol—a much lower absolute level of contamination, because of the higher proportion of carbon in these molecules.
Yet, despite the lack of absolute information it gives, measuring TOC is still a worthwhile procedure. It remains the best technique we have that indicates the presence of organic impurities—regardless of precise composition of the contaminant, it will still show that it is there, giving a degree of confidence that contamination is below a certain level. For example, if the measured TOC is 10 ppb, it indicates that the maximum level of organic impurities is around 100 ppb. It could, of course, be much lower, perhaps as low as 15 ppb, if the impurities contain a higher proportion of carbon. Because of this near order of magnitude difference between the upper and lower levels of absolute contamination, highly precise measurement of TOC is unnecessary; being able to judge
whether the water contains 10 ppb or
11 ppb TOC is a waste of time because of the inherent inaccuracy of its prediction of absolute levels. The significance of small changes depends on what impurity has caused the change, and whether it will actually interfere with the application the water is being used for.
TOC measurement is a useful way of determining whether ultrapure water contains organic contaminants and whether the amounts of contaminants are changing significantly, but highly accurate
measurements serve little purpose. What is needed is a monitor that is sufficiently sensitive to pick up low TOC levels and is on-line with a fast response time. Ideally, it should be built into the water purifier
for convenience, and have low running costs.

TOC monitors
Large industrial water purification plant has been equipped with sophisticated, expensive TOC monitors for many years. However, these are far too big and costly to make them a practical option for inclusion with every single lab water purifier. Other disadvantages include a delay in response and high running costs. Smaller, more practical TOC monitors have been available for lab purifiers for a decade now, with the first monitor from Elga LabWater being followed over the years by monitors with different designs from other companies. All built-in monitors take advantage of the same physical effect: the fact that, when irradiated with high energy UV light from a low pressure mercury lamp, organic impurities in the water are oxidised. This oxidation produces ionised species such as acids, and ultimately the carbon can be converted to carbon dioxide. These species conduct electricity, and the conductivity of the water rises as a
result. The change in conductivity can
be measured, to give an indication of
the amount of carbon present in the
water.
Monitoring procedure
Most of the available on-line lab monitors are essentially scaled down versions of the older industrial monitors, and retain many of their disadvantages. Additionally, their robustness and performance specifications have been reduced to cut manufacturing costs. These monitors are connected to a side-stream from the pure water recirculation loop just before it is dispensed. The water is first flushed through the reactor cell for a fixed time, after which the flow is stopped to allow the oxidation reaction to run to completion. The sample is the volume held in the oxidising/measuring cell—typically a millilitre or less. In one system, the measurements are made in the same cell, with the final value being reported after the estimated end of oxidation. Other systems employ a fixed oxidation time followed by separate conductivity measurements. Either way, there is a time delay between the sample being taken and the TOC value being displayed.
Neither sampling nor analysis are continuous. The resulting time lag introduces a delay of at least three minutes, and maybe as many as nine. This means contaminated water could already have passed into the lab application and, as the monitor is not continuous, it can all too easily allow a spike of contaminant to pass through unnoticed. The only way to avoid these problems is to use a system that is truly continuous.
Continuous TOC monitoring
In the Elga LabWater system, the TOC monitor uses the 185nm UV chamber
that is already fitted in the equipment as part of the purification process. The increased conductivity of the water can be measured on-line, giving continuous, almost instantaneous readings of TOC
levels.
As an example, the Elga LabWater system was compared to one of the side-stream alternatives. The other TOC monitor was connected to a modified Purelab Ultra water purifier, and 3 ml injections
of a 100 ppm solution of the solvent
methyl ethyl ketone were made into the feedwater, at four different points in the
other monitor’s measuring cycle. Similar test injections were made in a standard Purelab Ultra with Elga LabWater’s own monitor. In addition to noting the readings
on the TOC monitors, the TOC of the dispensed water was also measured continuously.
The results of the test are shown in the graphs on this page. These graphs clearly show that the Elga monitor detected each of the four injections. The ability of the side-stream monitor to pinpoint the methyl ethyl ketone contamination was dependent on precisely where in the cycle the injection was made. If it coincided exactly with the measurement point, then it was fully detected. In all other circumstances the small sample taken means that only a small proportion—or even none—of the solvent contamination was detected. In all events the time lag would make it all too easy for a user to use contaminated water without realising it and, worse, believing that all was well. And that contamination could be severe.
Conclusion
The potential problems that organic contamination can cause in the lab are substantial. Analytical results can be wrong, reproducibility can be ruined, and media can rapidly become fouled. So it is vital to know if the water is contaminated before it is used. TOC is currently the only practical way of monitoring organic impurities, but due to its non-specificity it can only act as a guide that contamination has occurred. Much more important is the ability to tell quickly that something is amiss. Highly accurate TOC measurements are an unnecessary and expensive luxury of little benefit. What is essential is the ability to tell extremely rapidly that there is a problem, and to be reassured that no contamination is able to slip through unnoticed. Only truly continuous monitoring is able to provide this peace of mind.