Elevated concentrations of oxygen are used in many chemical processes to boost the yield without significant cost increases. The oxygen enrichment for chemical processes typically ranges from 80% to 100%. Oxygen supplied to the chemical processing industry is either generated cryogenically, or, for moderate to low-level consumption, generated by pressure swing absorption and/or a vacuum-pressure adsorption method.
Paramagnetic-Type Oxygen-Sensing Devices
Both oxygen producers and users have historically relied on paramagnetic-type oxygen sensing devices for measuring oxygen purity. These sensors offer highly accurate results, especially at the suppressed ranges of 90% to 100% oxygen. They are, however, expensive, and require extensive maintenance in terms of frequent calibration. Furthermore, paramagnetic-type sensors are very sensitive to changes in gas flow rate; to the presence of even small counts of fine particulates; and to moisture, temperature, pressure, and mechanical vibrations. Analyzers incorporating these sensors therefore require calibration on almost a daily basis.
There always has been an interest in low-cost, low-maintenance, and more versatile ways to measure oxygen purity. Galvanic-type micro-fuel cells, which are specific to oxygen, satisfy these requirements. In addition, they are very simple to operate. These devices have not been used to measure elevated levels of oxygen (>90%), however, because their signal output drifts with time. An understanding of the reasons for this drift requires a familiarity with the construction and operation of a micro-fuel cell.
The basic components of a typical micro-fuel cell are shown in Figure 1. Under normal operating conditions, oxygen diffuses through a Teflon membrane, dissolves in a very thin layer of electrolyte between the sensing electrode and the membrane, and then diffuses to the sensing surface where it is reduced. Oxygen reduction is facilitated by simultaneous oxidation of an anode material, typically lead or cadmium.
Cell Reactions. Cell reactions are described as follows:
O2 + 2H2O + 4e 4OH cathode reaction
2Pb + 4OH 2PbO + 2H2O +4e anode reaction
The overall cell reaction is as follows:
O2 + 2Pb 2PbO
Electrons released by oxidation of the anode are consumed during oxygen reduction and flow through an external circuit. Measurement of electron flow (current) is indicative of the concentration of oxygen. For a micro-fuel cell to function properly on a continuous basis, it is necessary that:
• The amount of the anode material is sufficient to support the oxygen reduction over several months to a few years.
• The concentration of hydroxyl ions at and near the electrode surface is sufficient at all times.
• The concentration of PbO at and near the electrode surface does not build up to a level that can cause blockage of the oxygen reduction reaction.
Such requirements pose certain limitations on the practical design of the micro-fuel cell in terms of physical size, the amount of lead/cadmium used as the anode, and the thickness of the Teflon membrane. A typical micro-fuel cell is ~1.25 in. dia. and up to 1 in. long. The membrane thickness is <1 mil, which produces a 90% F.S. response time of <15 s.
With oxygen concentration in the low percent to high ppm levels, anode material of about 10—15 g provides enough surface area for the reaction to proceed. Also, the reactant and the reaction product move into and out of the thin layer of electrolyte very efficiently. As a result, the sensor provides a very stable signal output (see Figure 2).
With elevated oxygen concentration (>90%), the hydroxyl ions are used up rapidly and their concentration near the electrode surface decreases. At the same time, the concentration of PbO at and near the electrode surface builds up and eventually reaches a level where PbO starts precipitating and covers the active sites on the sensing electrode (see Figure 3).
Both of these factors cause the micro-fuel cell’s signal output to slowly drop with time. One common method of reducing the rate of production of the reaction product is to use a thick membrane to decrease the amount of oxygen reaching the electrode surface. This approach produces a micro-fuel cell with very stable signal output, but the response time to changes in oxygen concentration increases so much (as high as 60 s) that the device becomes impractical for use on process streams.
New Micro-Fuel Cell Technology
Analytical Industries Inc., through proprietary methods, has been able to control the rate of production of the reaction product without sacrificing the response time of the micro-fuel cell. A typical example of the new cell with a small physical size (1.25 in. by 0.75 in.) offers a 90% F.S. response time of <13 s, with an expected life in 100% oxygen of up to two years. This cell was tested over 14 months in 100% oxygen and has been found very stable. The tests were conducted by placing the cell in a temperature chamber kept at 85°F, ±1°F. The ambient pressure was monitored with a Motorola barometric pressure transducer and the cell output was compensated for ambient pressure variations. The sample flow rate was set at 0.1 lpm (the cell output was insensitive to changes in sample gas flow of up to 1.0 lpm) with the sample vented to the atmosphere via a 1/4 in. tube to minimize the backpressure on the cell.
With the micro-fuel cell in a temperature-controlled chamber and the output compensated for ambient pressure variations, it was possible to measure oxygen in the suppressed ranges of 90%—100% oxygen with less than ±1% of F.S. (±0.1% oxygen) accuracy. The calibration was checked periodically and found to be within ±1% of F.S. over a 14 month period.
The output of a micro-fuel cell when exposed to 99.5% oxygen over four weeks is plotted in Figure 4. The maximum variation in the signal is less than ±0.1% oxygen over a 24 hr period and is due mainly to the variation in ambient temperature. Over a six month period, the maximum drift in the signal output was less than ±0.5% oxygen.
These innovations in micro-fuel cell technology offer a very simple yet reliable and inexpensive alternative to the paramagnetic technique for measuring oxygen purity at suppressed ranges. Advanced Instruments Inc. has incorporated one of these cells in its Model GPR-31 oxygen analyzer. The analyzer exhibits excellent stability in 100% oxygen over extended periods of time. The analyzer is insensitive to minor changes in the rate of flow of the sample gas and unaffected by particulates and moisture. The calibration interval could be extended from one month to six months. The analyzer costs significantly less than paramagnetic analyzers yet offers excellent stability and sensitivity, fast response, long life, inexpensive sensor replacement, and very minimal maintenance.
About the author
Mohammad Razaq, Ph.D., is Vice President, Engineering, Advanced Instruments Inc., 2855 Metropolitan Pl., Pomona, CA 91767; 909-392-6900, fax 909-392-3665, [email protected], www.aii1.com, www.aii2.com.