LIFE SCIENCES - Self-powered oxygen sensor operates in space
A self-contained optical oxygen sensor that uses a stable radioluminescent light source recently traded the polluted waters of the Iowa City Wastewater Treatment Facility for the relatively clean environs of the Space Shuttle Columbia's bioreactor. Developed by Mark Arnold and Han Chuang at the University of Iowa (Iowa City, IA), the sensor detects oxygen flowing through a capillary tube. Unlike previous devices, however, this chemical sensor does not foul under extended periods of operation, require constant recalibration, or need electricity for its optical source.
The highly reactive element oxygen is crucial to many chemical processes, not the least of which is cellular survival. Traditional oxygen-sensing techniques depend on Clark electrodes. These devices use a platinum cathode and silver/silver chloride anode separated by a gas-permeable membrane. As oxygen diffuses across the membrane and makes contact with the cathode, it generates an electric current.
Although these sensors have excellent response times, they require frequent calibration because of membrane fouling, which can alter the rate of oxygen diffusion, reducing its permeable surface area and subsequently altering the relationship between oxygen content in the solution and the electric current. A secondary-albeit slight-drawback to this system is that it consumes oxygen as it makes a measurement. Consumption can be problematic in closed systems with limited oxygen budgets.
Arnold and Chuang's solution does not consume oxygen. Rather, oxygen serves as a quenching agent for a fluorescently active ruthenium compound (4,7-diphenyl-1, 10-phenanthroline) ruthenium(II) chloride (Ru(dpp)3). A glass capillary 2 cm long with an internal diameter of 1.5 mm is coated with a mixture of the ruthenium solution, a silicon adhesive, and titanium oxide. After this layer dries, a layer of black silicon (silicon with black carbon) is applied to isolate the optically active ruthenium layer from the sample.
The ruthenium layer naturally fluoresces when excited by 450-nm light. In the past, devices from lasers to lamps and LEDs have acted as light sources for ruthenium oxygen sensors; however, these bright sources require external power and lock-in amplification to improve their signal-to-noise ratio (S/R). While these bright sources tend to provide a high S/R-assuming the system compensates for source fluctuation-they also hasten photodecomposition of fluorescent materials.
Tritium gas powers the radioluminescent (RL) light source of the oxygen sensor, providing a low-power stable optical source to excite the ruthenium coating. The miniature unit sans electronics (inset) was tested in an effluent wastewater plant, collecting oxygen level data for three weeks without any significant change in sensitivity.
To make a low-maintenance, stable device that has lower power requirements, the researchers in the Optical Science and Technology department at the University of Iowa used a radioluminescent source from MB-Microtec (North Tonawanda, NY), at the heart of which is a glass capsule with 3 mCi of tritium gas contained in a blue phosphor (see figure). As the tritium breaks down, it excites the phosphor, which then emits at 450 nm, the excitation peak for the ruthenium solution.
Both the capillary sensor and light source are glued to a bandpass filter with a 90-nm (full-width, half-maximum) bandwidth centered at 602 nm with 64% transmission at the peak wavelength. (The bandpass filter transmits ruthenium's fluorescence, which is centered at 610 nm, while blocking the excitation signal from the radioluminescent source.) Source, sensor, and filter are then glued on the reception window of a Hamamatsu silicon-photodiode detector (Bridgewater, NJ).
Signal solution
As oxygen passes through the black silicon layer, it diffuses through the ruthenium layer, quenching its fluorescence. Therefore, in the absence of oxygen, the fluorescence is high; as more oxygen flows through the capillary, the optical signal drops. According to Arnold, the relationship between the oxygen content and optical output is nonlinear, however, because of heterogeneity in the polymer matrix, stray light, and nonquenchable luminescence. Therefore, Arnold uses a transform to linearize the oxygen level data.
Because of the low excitation power from the radioluminescent source, Arnold said, the signal must be amplified and filtered electronically. With this design, the output signal tends to lose sensitivity as the oxygen level reaches saturation. However, with oxygen levels at 100%, the sensor can detect changes as small as 30 ppb.
As the sensor faced its first space flight, Arnold and Chuang remained confident about its ability to operate unattended for long periods of time. The sensing layer had already survived a three-week trial submerged 2 ft into effluent at a wastewater treatment plant with no discernable change in performance. Following the successful shuttle mission in July, the onsite team said test cells inside the bioreactor unit grew as expected, and there was no trace of contamination from the oxygen sensor.
R. WINN HARDIN is a science writer based in Atlantic Beach, FL; [email protected].