Spectroscopic Determination of Trace Contaminants in High-Purity Oxygen
- Created on Thursday, 01 August 2013
A glow discharge emission system is used to detect and quantify trace amounts of argon in pure oxygen.
Oxygen used for extravehicular activities (EVAs) must be free of contaminants because a difference in a few tenths of a percent of argon or nitrogen content can mean significant reduction in available EVA time. These inert gases build up in the extravehicular mobility unit because they are not metabolized or scrubbed from the atmosphere. A prototype optical emission technique capable of detecting argon and nitrogen below 0.1% in oxygen has been developed. This instrument uses a glow discharge in reduced-pressure gas to produce atomic emission from the species present. Because the atomic emission lines from oxygen, nitrogen, and argon are discrete, and in many cases well-separated, trace amounts of argon and nitrogen can be detected in the ultraviolet and visible spectrum. This is a straightforward, direct measurement of the target contaminants, and may lend itself to a device capable of on-orbit verification of oxygen purity.
A glow discharge is a plasma formed in a low-pressure (1 to 10 Torr) gas cell between two electrodes. Depending on the configuration, voltages ranging from 200 V and above are required to sustain the discharge. In the discharge region, the gas is ionized and a certain population is in the excited state. Light is produced by the transitions from the excited states formed in the plasma to the ground state. The spectrum consists of discrete, narrow emission lines for the atomic species, and broader peaks that may appear as a manifold for molecular species such as O2 and N2, the wavelengths and intensities of which are a characteristic of each atom. The oxygen emission is dominated by two peaks at 777 and 844 nm.
For testing, a quartz capillary tube with stainless steel end fittings forms the glow discharge tube. The sample gas is introduced into the glow discharge cell using an adjustable vacuum leak valve. From the glow discharge cell, the sample gas passes a vacuum gauge, the downstream valve, and then the vacuum pump. During operation, the pressure in the glow discharge cell is maintained between 0.5 and 10 Torr using the adjustable leak valve and the downstream valve. Light from the discharge is collected by a lens and coupled to a UVvisible fiber-optic cable. This cable directs the light from the glow discharge into a spectrometer. The spectrometer detects in the 200- to 850-nm region with a spectral resolution of 1.5 nm using a 25-μm entrance slit. The spectrometer is connected to a data acquisition computer via a USB cable. For this work, Ocean Optics’ SpectraSuite® software was used for the data acquisition, setting parameters such as wavelength range, integration times, and scans to average.
For a peak to be at the detection limit, it must be recognizable as a peak, be resolved from other peaks, and have a peak intensity three times the standard deviation of background noise in the region of the peak. The peak corresponding to 0.01% argon is just above the baseline of pure oxygen (see Figure 2), and the signal-to-noise ratio is 2.6, indicating the detection limit is between 0.05 and 0.01%.
This work represents a proof-of-concept investigation into using a glow discharge emission system to detect and quantitate trace amounts of argon in pure oxygen. A similar analysis will need to be done for nitrogen. Optimization of experimental parameters such as operating pressure, discharge current, voltage, and spectrometer integration time needs to be further investigated. A redesigned discharge cell that will use a lower-voltage DC power supply with a higher discharge current is being designed to provide a spectrally brighter, lower-noise glow discharge.
This work was done by Steven Hornung of Johnson Space Center. MSC-25116