Reducing Errors in Process pH Reference Electrodes
In today's industrial process, pH has become a universal measurement, bringing dramatic improvements to the quality of the product produced. As the applications have multiplied, however, so have the measurement difficulties, and while pH control instrumentation is reliable and sophisticated, the demands made on the wet-end measuring devices have not always been met.
The performance of a typical process control pH sensor is governed by two primary factors: the ability of the reference half-cell to produce a stable and constant voltage independent of the process conditions, and the electrical impedance of the pH measuring bulb. The impedance can be controlled and stabilized over extended periods of time in difficult applications, but the problem of long-term stability of the reference electrode or half-call remains. It is not unusual for reference electrode errors to exceed the voltage being measured by the pH electrode, so reduction or elimination of these potentials is of vital importance if an accurate and reproducible sensor is to result.
A reference half-cell is composed of the various components shown in Figure 1. The purpose of the reference electrode is to establish an electrochemical half-cell where a solid phase is in electrochemical equilibrium with a conductive solution of constant composition. This solution, or electrolyte, makes connection to the process by means of a salt bridge and liquid junction. The boundary between the salt bridge and the process is the source of the liquid junction potentials or errors examined here.
Traditionally, industrial reference electrodes have used "e;flowing"e; liquid junctions, primarily because of the selection of junction materials. Such materials as asbestos fibers, ceramic wicks, and the like, have a small pore size and very low flow rates that are not easily controlled (Figure 2). During the manufacturing process, most early junctions were subjected to the inconsistent heats and pressures of hand fabrication, both of which affect the ultimate flow rate of the finished reference junction. This usually resulted in reduction of flow rates with time, lack of measurement reproducibility, and the need to pressurize the reference cell to maintain an outward flow of electrolyte. The desire to establish positive flow of the salt bridge solution was related to the theory that the heteroionic boundary between the salt bridge solution and the process must be at the outside surface of the liquid injunction in order to minimize junction potential errors and to alleviate clogging of the junction.
We now know that junction potentials are caused by charge separation due to the different ionic mobilities of the various ions in the electrolyte and in the process. By using junctions with small pore size, and inducing outward flow, one is actually forcing a charge separation to occur. The small pores act as a filter, allowing smaller, more mobile ions to pass through the pores, while the larger ions tend to remain behind the junction "e;screen."e; Eventually, electrostatic effects balance the applied pressure and an equilibrium is reached, but this remains balanced only as long as the internal and external pressures and temperatures, as well as the flow rate of the process stream, all remain constant.
To minimize these problems, salt bridge solutions with an anion and cation of similar size are chosen, so there is little difference due to ionic mobility in the steady-state. As soon as flow is induced through the injunction, however, charge separation occurs, and errors are evident. As the smaller ions arrive at the process side of the junction, further spurious potentials arise due to the differing ionic strengths of the two solutions. Ionic mobilities and activity coefficients are temperature dependent, introducing further errors if the temperature of the process varies. Calculations based on the Henderson equation suggest that these temperature errors are negligible for process excursions of 10oC or less. Unfortunately, in deriving his equation, which describes junction potential behavior, Henderson assumes that ionic mobilities are independent of concentration and also that activity coefficients are unity. We now know that these assumptions are invalid, and that both of these expressions contribute to the junction potential phenomenon.
Obviously, the key to reducing or eliminating junction potentials is to design a reference cell in which the electrolyte is ionically balanced, the liquid junction pore size is sufficiently large to diminish charge separation, and the heteroionic boundary is allowed to be established by osmotic pressure. Increasing the surface are of the liquid junction will further reduce the errors. With these design criteria satisfied, the only charge separation that can occur will be caused by the different ionic strengths and activity coefficients found in the process stream itself.
Problem solved? Not entirely, because now we have a junction with quite large-pore size and surface area, which will allow excessive electrolyte flow, reducing the electrode life and allowing salt egress to contaminate or dilute the process. The next step must be to control the electrolyte flow rate. The flow rate is proportional to differential pressure, pore size, junction area, and electrolyte viscosity, and we have already established the junction porosity and surface area. Now we need to induce a negative pressure, or increase viscosity of the electrolyte. As we have seen, applying pressure to achieve flow has its share of problems, and creating a vacuum to impede flow would lead to similar difficulties. By modifying the solution viscosity, we can control the flow rate with great accuracy, as well as reduce the mobility of all ions and increase electrode life. Reducing the ionic convection effect also reduces the charge separation phenomenon due to the slowing of the ionic mobility.
The liquid junction material selected should have suitable pore size to negate charge separation, be impervious to most process solutions, and be reproducible. Porous Teflon junctions satisfy all these requirements, in addition to being non-clogging, and allowing the addition of detergent and antifreeze agents that are not compatible with conventional ceramic junctions (see Figure 3). A further important advantage in choosing porous Teflon as the junction material, and therefore tends to have a higher charge transmission capability, allowing greater freedom of movement of the charged ion pairs.