Dealing with pH Dependence
Amperometric sensors are widely used to measure free chlorine in drinking water. But not all amperometric chlorine methods are equivalent. Some require treating the sample with chemicals before measurement, others require no pretreatment but need an auxiliary pH sensor, and one recently introduced method requires neither reagents nor a pH measurement.
Why are there so many ways of making the same measurement? Simply put, manufacturers have found different ways of coping with a certain condition: the response of an amperometric free chlorine sensor depends on pH levels. This is a real problem. If the pH of the sample changes from the value it had when the sensor was calibrated, the chlorine reading will be in error. The errors can be quite large—as much as 20% per unit change in pH. The error has two sources: the chemistry of free chlorine itself and the operating principle of the sensor.
Depends on pH
Free chlorine is not a single compound. It is a mixture of hypochlorous acid (HOCl) and hypochlorite ion (OCl-). As Figure 1 shows, the relative amount of each depends on pH. At pH 5.5, free chlorine is predominantly HOCl. As pH increases, the proportion of HOCl decreases and the proportion of OCl- increases. Because free chlorine is the sum of HOCl and OCl-, pH changes do not affect the concentration of free chlorine, only the relative amounts of HOCl and OCl-.
Amperometric chlorine sensors are electrochemical devices. A voltage applied to a metal electrode called the cathode causes an electrochemical reaction in which HOCl molecules combine with electrons and are destroyed.
Because the concentration of HOCl at the cathode is zero, HOCl in the bulk solution diffuses to the cathode where it, too, is destroyed. The continuous destruction of HOCl requires a constant flow of electrons. Thus, the sensor produces a current directly proportional to the diffusion rate. In turn,the diffusion rate is proportional to the concentration. It is important to understand the sensor current arises from the reaction of HOCl, not OCl-, at the cathode.
A profound effect
A glance at Figure 1 suggests that pH changes can have a profound effect on the sensor current. Between pH 7.0 and 8.5 (the typical range for drinking water) the percentage of free chlorine present as HOCl drops from 75% to 10%.
Thus, the current would be expected to drop by a factor of 7.5. In fact, as shown in Figure 2, the actual drop in current is less. The current is higher than expected because of the dynamic equilibrium between HOCl and OCl-. Loss of HOCl at the cathode upsets the equilibrium, causing OCl- to form HOCl. The additional HOCl then reacts at the cathode, increasing the current. Nevertheless, the current is still a strong function of pH. There are four ways to address pH dependence of free chlorine sensors:
- Restrict applications to those in which the pH varies no more than about ±0.1. Although a few water sources have fairly constant pH, the majority have variable pH. A free chlorine method that fails to account for pH variability is too limited to be widely practical.
- Treat the sample with acid (e.g., vinegar) to lower the pH to a value where free chlorine exists only as HOCl. This is a sound technique and is used by many manufac- turers. The method has drawbacks, though. It requires a sample conditioning system, which increases initial costs. It also uses reagents that have to be managed and replaced at regular intervals, leading to ongoing operating costs.
- Measure the pH of the sample and use the pH to calculate a correction factor. Several manufacturers take this approach. It has the advantage of being reagent- free but has the disadvantage of requiring a pH sensor. In addition, no pH correction algorithm is perfect. If the sensor response does not perfectly match the correction equation, the corrected reading will be in error.
- Adjust the pH and convert free chlorine to HOCl inside the sensor. At least one manufacturer uses this recently developed method. Internal pH correction has two big advantages: it requires no reagent and no auxiliary pH sensor.
Figure 3 compares the operation of a free chlorine sensor with internal pH correction (B) to a standard free chlorine sensor (A). Because pH adjustment occurs inside the sensor, the method is suitable only for membrane-covered sensors. The membrane used is hydrophilic; that is, any water-soluble substance, including HOCl and OCl-, can pass through it.
In another scenario, an acid solution floods the space between the cathode and membrane, lowering the pH and converting OCl- to HOCl.
Because free chlorine has been converted to HOCl, the sensor current is practically independent of sample pH. To provide the longest possible time between refills, a large reservoir containing slurry of a slightly soluble organic acid is incorporated into the sensor. The acid is adequate to provide pH adjustment in samples containing as much as 300 ppm (as CaCO3) total alkalinity.
Figure 4 shows how well the internal pH correction works. The change in sensor response with unit change in pH is about 4%.
A statement from the Puerto Rico Aqueduct and Sewer Authority (PRASA) about their experience with a reagentless chlorine system employing internal pH correction said, “With the installation of the system, PRASA improved the reliability of a critical variable in the safe distribution of drinking water; non-compliance violations have stopped. Maintenance frequency has been reduced from 12 to three times per year and associated labor costs have been reduced by 75%. Avoidance of chemical reagents saves an additional $760 per system per year. PRASA has also approved the system for their east, north and metro regions.”
Over the years, manufacturers of amperometric free chlorine sensors have devised several ways of dealing with the pH dependence issue. The most recent method, adjusting pH within the sensor, eliminates much of the expense and potential errors associated with earlier methods.