There is a clear trend in the market towards replacing polarographic with optical dissolved oxygen. The shift in preference has many sources including: simplified maintenance, elimination of polarization, lower cost of ownership, viability in single use applications, and more. When evaluating the switch to optical, engineers must consider if there will be any impact on the current process due to the change in measurement principal.
The short answer is that under many circumstances optical can be used as a direct replacement for polarographic but there are circumstances outline below where difference can be observed.
Does an optical DO sensor trend match a polarographic DO sensor?
Yes, both sensor measure the partial pressure of oxygen in solution so when inserted into the same process the measurement will show great correlation under typical operating conditions.
The graph on the right shows a head to head comparison of optical and polarographic dissolved oxygen during a seed fermentation over 16 hours.
What conditions cause a difference between optical and polarographic measurements?
Accurate dissolved oxygen measurement relies on a proper calibration. During a process conditions may occur that cause the sensor to drift from the initial calibration value. Since the DO sensor is used to control the process, it is possible that these shifts would go unnoticed until a reference sensor is used that is not impacted by the process conditions. Below are a few used cases to consider:
- Pressure spikes in the process, resulting in an offset
- Poisoning of the sensor from CO₂ or other gases generated by the process
- Chemical attack from solvents in the process
How do pressure spikes impact DO measurement?
Polarographic sensors can be sensitive to fluctuating process pressure. Pressure spikes or vacuum conditions can cause unreinforced sections of the membrane to shift or deform. This change affects the diffusion rate of oxygen across the membrane, resulting in a measurement offset. Optical sensors have a rigid construction that is not impacted by transient changes in pressure.
The graph on the left shows a head-to-head comparison of optical and polarographic dissolved oxygen. After about 3 hours, a pressure spike occurred from a valve that opened and closed. After the spike, both measurements show similar trends, but with a difference of approximately 6% air saturation. A product calibration was used to correct this error in the polarographic sensor.
How does CO₂ impact dissolved oxygen control?
A byproduct of cellular respiration is CO₂ gas. In solution, this gas has the ability to diffuse across the membrane of a polarographic sensor. Once inside the sensor, the electrolyte is acidified by carboxylic acid. This changes the pH of the electrolyte and ultimately causes a drift in measurement. Since the DO sensor is being used to maintain oxygen at a set point, this results in the oxygen set-point drifting upwards as long as the process runs. This effect becomes more evident when comparing to an optical sensor that does not have electrolyte and is not impacted by CO₂.
The graph below shows optical and polarographic DO sensors that were trended over a 38 day period. After approximately 4-5 days, it seems that the optical DO sensor (blue trace) is drifting upwards when compared to the polarographic DO sensor (green trace). In this test, the controlling sensor was the polarographic sensor. As the sensor began to drift due to the acid gas, it called for more and more oxygen from the mass flow controller to maintain the established set-point. As can be seen by the optical DO trace, the oxygen concentration is drifting higher during the run.
To prove that the issue is related to drift of the polarographic sensor and not the optical sensor, the nA reading in air was recorded under the same conditions prior to the run and after completion. The difference in nA output correlated to the observed drift.
Can optical be used when solvents are present?
One established shortcoming of optical dissolved oxygen sensors is their usability in applications with organic solvents.
Some sensor designs utilize a PTFE membrane to limit the exposure of the luminophore to incompatible solvents. This has proven to be very effective in biological processes. For applications with higher solvent concentrations, there is a risk that the luminophore is leached from the sensor cap. To combat this the luminophore can be immobilized. While effective, there can still be difficulty with calibrating the sensor to be accurate in mixed solvents.
In these applications, a polarographic sensor may be the best solution. Optical solutions may still be an option, but it will depend on the specifics of the process.