Atmospheres with dangerously high O₂ may also occur, most commonly due to the release of relatively pure O₂ gas from certain types of welding or torch-cutting equipment, and from medical use of O₂. However, some chemical reactions can also produce oxygen-enriched environments. For example, hydrogen peroxide (H₂O₂) may decompose into water (H₂O) and O₂ gas, especially in the presence of a catalyst such as manganese dioxide or transition metals, as shown in this reaction: 2H₂O₂ → 2H₂O + O₂(g) Some organic peroxides and inorganic oxidizers might also decompose to produce O₂ gas, as with the heating of potassium chlorate (KClO3). Also, when H₂O₂ and sodium hypochlorite (NaOCl) are mixed, the two species will react vigorously to produce O₂ gas. A low O₂ reading obtained with a correctly calibrated and used O₂ sensor indicates that either O₂ has been consumed from the atmosphere tested or another gas has been added to dilute the O₂ normally present. Since the approximate atmospheric nitrogen/O₂ volumetric ratio is 4:1, the O₂ concentration will fall 1 percent by volume for each 5 percent total dilution. High-toxicity gases such as hydrogen sulfide (H₂S) can create an extremely dangerous atmosphere while only minimally diluting O₂ content. For example, the addition of sufficient H₂S to create a deadly 5,000 parts-per-million atmospheric concentration (0.5 percent H₂S by volume) would result in an O₂ decrease of only about 0.1 percent by volume. Such a minute decrease would not by itself be a cause for alarm. For this reason, an O₂ sensor is almost always used in a multi-gas instrument to allow detection of low or high O₂ concentrations as well as other important air contaminants at toxicologically relevant concentrations. Electrochemical methods are typically used to measure atmospheric O₂ concentrations. The most common methods rely on balanced oxidation-reduction reactions in which the total flow of electrical current produced is proportional to O₂ concentration. Electrical current may be expressed numerically as amperes; thus, these approaches are amperometric detection methods. As explained in a paper published in the April 2014 issue of the journal , two fundamental amperometric O₂ sensor approaches are possible: one uses a consumable lead anode in a galvanic fuel cell design, and the other involves an oxygen pump electrolytic mechanism. In a lead anode sensor, O₂ is reduced to hydroxyl ions at a cathode, in the presence of aqueous base: O₂ + 2H₂O + 4e- → 4OH- In the half-reaction that accompanies this reduction, a metallic lead anode is oxidized to lead oxide (PbO): 2Pb + 4OH- →2PbO + 2H₂O +4e- The overall balanced reaction may be written as: 2Pb + O₂ → 2PbO In the lead anode O₂ sensor, four electrons are generated for each O₂ molecule reduced at the cathode. This stoichiometric relationship allows the amount of O₂ present at the cathode to be quantified as electrical current proportional to O₂ concentration. PbO is created while metallic lead is consumed, so the anode has a finite lifetime determined by how much O2 enters the sensor. The reaction is spontaneous, requiring only the presence of O₂ at the cathode to generate current, and is typically catalyzed by platinum in the sensing electrode. The “oxygen pump” O₂ sensor avoids the use of lead, which is restricted in European markets due to hazardous substance regulations. In this type of sensor, four electrons are needed for each O₂ molecule that is reduced to water at a cathode by the following reaction: 4H+ + O2 + 4e- → 2H2O In the accompanying half-reaction, which occurs at the counter electrode (an anode), water is oxidized to produce O₂: 2H₂O → 4H+ + O₂ + 4e- The amount of current necessary to reduce the incoming O₂ at the sensing electrode is proportional to the O₂ concentration in the atmosphere being sampled. The consumption and production of O2, protons, water, and electrons is balanced. If only O₂ is considered, the net reaction is: O₂ → O₂ For each O₂ molecule that enters the sensor and is reduced at the cathode, another O₂ molecule is produced at the anode and then expelled. The external current applied to the sensor to accomplish this is proportional to the number of O₂ molecules reacted. This mechanism underlies the term “oxygen pump.” When comparing percent-by-volume to part-per-million concentrations, a 1 percent concentration is equal to 10,000 ppm. An electrochemical sensor designed to detect most toxic gases is typically only challenged with concentrations up to hundreds of ppm, while O₂ measurements must be made for hundreds of thousands of ppm. Thus, a means is needed to limit the ingress of O₂ into an amperometric sensor. Regardless of whether a sensor is a lead anode or oxygen pump type, two possible approaches may be used to limit O₂. The April 2014 paper describes these approaches: one involves the use of a thin capillary, while the other depends on a polymeric membrane to limit O₂ entry into a sensor. A capillary restrictor about the diameter of a human hair limits the flow of O₂ into a sensor, allowing for true readings of percent by volume with very little effect related to atmospheric pressure. In the membrane permeation-limiting approach, the O₂ reading is based on the difference between the partial pressure of O₂ in the air and within the sensor. This differential pressure varies according to total atmospheric pressure, even when O₂ content in air remains constant at 20.9 percent by volume. While the sensor output actually measures the partial pressure of O₂, the detector readout is typically converted into percent by volume. Most O₂ sensors produced for the industrial hygiene and safety market employ a capillary-limiting design. The capillary-limited O₂ sensor is most commonly used in handheld multi-gas direct-reading instruments. This design provides for a true percent-by-volume measurement as long as nitrogen (N₂) is the major component of both the atmosphere the instument is being used to analyze and of the instrument’s calibration gas. This requirement is based on the interaction between O₂ molecules and those of the gas into which the O₂ molecules are mixed as they diffuse through a capillary. Matrix gases with lower molecular weight than N₂ will increase the diffusion rate of O₂ molecules through a capillary inlet and will lead to inaccurate (high) sensor readings. The presence of high concentrations of matrix gases with a higher molecular weight than N₂, such as carbon dioxide (CO₂), will lead to decreased O₂ readings. Also, high levels of CO₂ can lower electrolyte pH, which may reduce the concentration of hydroxyl ions that are needed to support the electrochemistry occurring at the sensor electrodes. Minor fluctuations in temperature and barometric pressure will not change the capillary-limited O₂ sensor reading. However, in general, large pressure changes can cause erroneous readings as the sensor pressure slowly equalizes with the atmosphere through the capillary. In an O₂ sensor where diffusion of atmospheric gases is limited by a membrane, entry of O₂ into the sensor is driven by the O₂ partial pressure differential across the membrane. This results in a partial pressure reading, which is usually corrected to percent O₂ by volume in the detector’s electronics. When a sensor is calibrated with gas containing a given percent O₂ by volume at a particular atmospheric pressure, subsequent changes in both atmospheric O₂ concentration or in the barometric pressure can change the sensor readings. At lower barometric pressure (that is, at a higher altitude), any gas is less dense, and for a fixed volume of gas mixture at higher pressure, each gas present will have a higher partial pressure than the mixture with identical percent-by-volume composition at a lower total pressure. This is the reason that the partial pressure of O₂ in a “normal” atmosphere (20.9 percent O₂) is lower when atmospheric pressure is lower, and higher when atmospheric pressure increases. A capillary-limited O₂ sensor will always tend to read 20.9 percent O₂ in an atmosphere with normal O₂ content despite gradual changes in altitude or barometric pressure. As altitude increases (and total pressure decreases), a permeation-limited O2 sensor reading will decrease even with constant exposure to 20.9 percent O₂ by volume. A permeation-limited O₂ sensor is desirable for monitoring O₂ content in gases used in underwater diving, where partial pressure and matrix effects of diluent gases are of concern. For example, helium is often used to replace N2 in breathing air for deep diving where N₂ narcosis is possible. A capillary-limited O₂ sensor will not be accurate when changing from an N₂/O₂ mixture to one composed of helium/O₂ due to the matrix effect noted previously. A permeation-limited O₂ sensor would also be desirable in any other situation where light or heavy gases may be encountered—for example, where liquid helium (lower molecular weight compared to air) or CO₂ (greater molecular weight compared to air) are used. Sensors can fail for a variety of reasons. Once all the available surface area of lead anode O₂ sensor is converted to PbO, the sensor will no longer measure O₂. Electrolyte leakage due to physical damage can also cause a sensor to fail, as can desiccation resulting from storage in a hot, dry environment. Blockage of a capillary pore or a permeation membrane by liquids or dirt will also cause failure. Freezing an oxygen sensor can cause the sensor to burst, which may occur with sensors stored at extremely low temperatures. The stoichiometric conversion of metallic lead to PbO results in expansion of the anode material as an O₂ sensor ages; lead anode sensor designs provide space within the sensor to account for this. An older lead anode that has undergone substantial conversion from metallic lead to PbO will be more susceptible to rupture from frozen electrolyte than a new sensor exposed to similar temperatures. When a sensor is ruptured due to freezing, very high O₂ readings are likely as the diffusion-limiting capillary or permeation membrane has been circumvented and O₂ is flowing into the sensor unimpeded. The concern in this situation is that the caustic electrolyte could contact the sensor circuit board and other internal circuitry. Also, electrolyte neutralized by exposure to high levels of CO₂ can be a source of O₂ sensor damage. The technology used in real-time detection instruments has shown remarkable progress in recent years. Oxygen sensors are an essential tool for protecting workers in confined spaces. Just as knowledge of hazards is vital for industrial hygienists, knowledge of how these instruments work is equally important. Acknowledgment: The author thanks Chris Wrenn and Lee Monteith for their contributions to developing the material discussed in this article.