At the heart of every gas detector
Gas detectors do ‘exactly what it says on the tin’ – they detect the presence, and in many cases levels, of gases.
Gas detection has always been closely linked with the protection of people and plant in industrial environments such as mines, confined spaces and oil and gas production facilities.
However this is no longer the complete picture of gas detection, with applications now including patient monitoring, exhaust gas analysis from vehicles, monitoring air quality in buildings, vehicles and aircraft and even fire detection.
With such diversity in applications it is no surprise that gas detectors come in almost every shape and size with prices ranging from less than 100 Euros to over 100,000 Euros. One common requirement for all gas detectors is they contain a gas sensor, or sensing technology. This article considers the different technologies that are used in gas sensors and explores why they are used in certain applications.
Gas sensor technologies
At the heart of every gas detector or detection system is a sensor or number of sensors that are responsible for the primary measurement of the target gas or vapour. Although often invisible to the user of the detector the sensors are the single most important component of the detector. Many different types of sensor have been developed over the years representing a very broad spectrum of technologies. These differ in many technical aspects and the prices can range from < 1 Euro to > 10,000 Euro.
Despite the ever-growing list of technologies gas detection is dominated by four technologies. These are:
- Electrochemical Gas Sensors
- Infrared Gas Sensors
- Catalytic Bead Gas Sensors
- Semiconductor Gas Sensors
Electrochemical gas sensors
Probably the best known technology for gas sensing and widely used in industrial safety / industrial hygiene applications.
We are all surrounded by electrochemistry in our daily lives from the batteries in our cars and mobile phones to the tiny batteries that power hearing aids. And in the future even in Fuel Cells that may one day power our cars and provide the power to our homes. Indeed it was as a spin off of this work in the 1960’s that the first of today’s generation of electrochemical gas sensors was developed. Electrochemical sensors have been developed to detect a wide range of toxic gases and oxygen and have the advantage of being relatively low cost, small in size, long life, reliable and stable. So what are they and how do they work?
Fig. 1 shows a schematic of an electrochemical oxygen sensor. It comprises a lead (Pb) electrode that is traditionally called the counter (or anode in this case) and a special PTFE supported catalyst layer that can react with oxygen called the sensing electrode (or cathode). Both of these electrodes are housed in a plastic body filled with a salt solution to act as an internal ionic conductor. The cathode electrode has a complicated role to play as it must allow gas to get to the catalyst from outside and also allow electrolyte to get to the catalyst but not to leak out through the gas access hole. This is achieved by using a thin (6 thou) porous PTFE tape as the supporting membrane. This material is very similar to ‘plumber’s tape’ and is porous to air but can stop liquid from passing through.
Figure 1 – Electrochemical oxygen sensor
In use oxygen flows through the gas access hole or ‘Diffusion Barrier’ to the sensing electrode where the catalyst promotes the following reduction reaction to occur:
O2 + 2H2O + 4e – ? 4OH-
The hydroxyl ions flow through the cell to the Lead anode where they react with the lead, liberating electrons that are consumed at the cathode:
2Pb + 4OH- ? 2PbO + 2H2O + 4e-
As soon as a connection is made externally between the sensing and counter electrodes the current will start to flow and by measuring the current, the flux of oxygen which is reaching the cathode may be determined. The overall cell reaction is thus:
2Pb + O2 ? 2PbO
The life of such a sensor is governed by the mass of lead inside and the rate at which it is used up. This in turn depends on the signal and the level of oxygen in the atmosphere under test. As detector manufacturers demand ever smaller sensors which therefore contain less lead the operating signals must be continually reduced if the target of a 2 year life is to be maintained
One further step is needed to turn this from an indicator into a quantitative sensor and that is to make the diffusion barrier sufficiently restrictive that it is the controlling factor in determining how much gas the sensor reacts with. By doing this much of the variability of the underlying chemistry is removed and the output of the sensor is then controlled by the physical characteristics of the diffusion barrier.
In practice there are two types of diffusion barrier in common use – capillary holes and solid membranes. Their properties are quite different and lead to very different sensor properties. Sensors with capillary holes measure the volume % concentration of a gas whereas those with solid membranes respond to the gas partial pressure. These two measurements are related by the fact that the concentration is the partial pressure divided by the total pressure.
Both measurands are valid with partial pressure being favoured in medical applications as it more closely relates to the way oxygen is absorbed in the lungs, whilst concentration has long been preferred in industrial safety / industrial hygiene applications as is does not change if weather patterns cause atmospheric pressure to move up or down. Capillary sensors also have the benefit of much lower temperature coefficients that are usually small enough to ignore, but suffer from transient signals when pressure is suddenly increased or decreased which must be carefully damped by the sensor designer and instrument manufacturer if false alarms are to be avoided.
Detecting gases such as CO, H2S, SO2, NO, NO2, Cl2, NH3, HCN, HCl, PH3 and some others can also be done electrochemically. To detect these gases a catalytic electrode is required that can catalyse the oxidation or reduction of the gas in question. In general these catalysts are platinum group metals or alloys and the electrolyte is no longer a salt solution (as in oxygen sensors) but a strong acid. A typical example is hydrogen sulphide which can be oxidised on a platinum sensing electrode as follows:
H2S + 4H2O ? H2SO4 + 8H+ + 8e-
The counter will then need to undergo a reduction reaction to consume these electrons and would use atmospheric oxygen in a similar way to the oxygen sensor above, however in acid solution the reaction would be:
2O2 + 8H+ + 8e- ? 4H2O
This would be a 2-electrode sensor; just like the oxygen sensor however in this case the performance might be restricted. A better design is to incorporate an extra reference electrode (fig. 2) to measure a stable internal potential and use this in a special operating circuit (potentiostat) to stabilise the operating potential of the sensing electrode. This is called a 3-electrode sensor and forms the majority of the sensors used in industrial safety gas detectors today.
Sensors based on these principles should last a long time as there is no consumption of any materials during the sensing process, however there will always be some degradation of the catalyst occurring over time and a typical life expectancy is 3-5 years. As with oxygen sensors the trend is towards smaller sensors and the challenge for smaller sensors is to maintain sufficient electrolyte inside a very small housing to cope with use at extremes of humidity.
Sensors with a sulphuric acid electrolyte will tend to dry out when operated in very dry environments and to absorb water in very humid ones. Sufficient space must be left and sufficient electrolyte used to accommodate these swings. This means small sensors must have small signals to reduce the rate of water loss/gain and this is a considerable challenge.
Catalytic bead gas (pellistor) sensors
In a number of applications, especially industrial, the presence of explosive gases and vapours can present a significant risk to individuals and property. Hardly a week passes without a report appearing in the press of some disaster caused by an explosion in a mine or at a chemical plant around the work. Gas detectors are increasingly being used to warn of dangerous levels of explosive gases and vapours and these devices rely on catalytic bead sensors (pellistors) to detect these gases.
Detection of a potential explosive atmosphere has traditionally made use of the catalytic bead or pellistor sensor. This is essentially a platinum coil heated to 400°C to 500°C and coated with a ceramic catalyst (fig. 4) designed to promote the combustion of any flammable gases present. The heat released from the combustion process causes the bead temperature to rise and the resistance to increase.
Figure 4 – A pellistor / catalytic bead
It is normal to operate the pellistor as a pair of devices, one of which has been deliberately inactivated so it is not affected by the presence of the combustible gas, and to place them in a Wheatstone bridge circuit so that the small resistance changes may be measured. Since the mode of operation is actually to try to burn any combustible gas present these devices are placed in a flameproof housing, behind a flame arrestor in order to be safe to use in a potentially hazardous atmosphere. They are normally supplied as matched pairs in an approved housing with all the necessary safety certifications.
Pellistors are a good match for the application because they measure the property of interest directly i.e. they respond to the combustible nature of the gas. The sensitivity to different gases varies, however when expressed relative to the LEL of the gas, many are quite similar. Thus it is possible to have a very general combustible hazard warning device for a wide range of gases. Pellistors have two main drawbacks – the power used and the potential for poisoning or inhibition by certain materials such as sulphur and silicon containing molecules.
Manufacturers are constantly improving poison resistance and today’s portable safety sensors are an order of magnitude less susceptible to poisoning than 10 years ago. Power consumption is still an issue having decreased by only a factor of about 5 since the very first commercial use. Nevertheless clever packaging has reduced the size dramatically as is shown in figure 5.
Infra-red spectroscopy is a well established and documented analytical technique and the use of characteristic parts of a molecule’s IR spectrum has long been used as the basis of non-dispersive infra-red (NDIR) gas sensors.
Non-dispersive infra-red sensors measure the absorption of infra-red radiation over a frequency range. The spectra (fig. 6) shows some of the gases of interest and their infra-red absorption peaks.
By choosing a single frequency it would be possible to uniquely identify many different compounds. However, current low cost systems use broad-band sources and detectors with some form of wavelength filter to select the spectral region of interest. In this case it is possible to detect a range of similar compounds such as butane /propane/ pentane although with differing sensitivity.
These sensors differ significantly from the pellistors in that they are looking directly at the amount of a particular gas present and not at its combustible nature. However because they are based on a physical property of the molecules they are not subject to poisoning and have the benefit of being fail-safe against a number of fault conditions.
The technology is capable of being packaged into the industry standard 20 x 16mm cylinder size (fig. 7) and supplied as a safety approved device for use in potentially hazardous areas. Currently the power consumption is fairly high for a portable gas detector, but much work is under way to try to find useable solid-state sources which would offer significantly lower power systems.
Semiconductor gas sensors
Semiconductor gas sensors are used widely in both domestic and industrial applications where the measurement of the exact gas concentration is unnecessary and the cost of the sensor needs to be very low.
The discovery in 1953 that adsorption of a gas onto the surface of a metal oxide semiconductor produced a large change in its electrical resistance signalled the advent of mixed metal oxide semiconductor sensor (MMOS) technology. The effect is commercially exploited for only a few oxides due to the requirement for a unique combination of resistivity, magnitude of resistance change in gas (sensitivity) and humidity effects. Amongst the oxides that are used as MMOS sensors are Cr2TiO3, WO3 and SnO2.
The resistance change is caused by a loss or a gain of surface electrons as a result of adsorbed oxygen reacting with the target gas. If the oxide is an n-type, there is either a donation (reducing gas) or subtraction (oxidising gas) of electrons from the conduction band. The result is that n-type oxides increase their resistance when oxidising gases such as NO2, O3 are present while reducing gases such as CO, CH4, EtOH lead to a reduction in resistance.
The converse is true for p-type oxides, such as Cr2TiO3, where electron exchange due to gas interaction leads either to a rise (oxidising gas) or a reduction (reducing gas) in electron holes in the valence band. This then translates into corresponding changes in electrical resistance.
Since the change in electrical resistance in the sensing oxide is caused by a surface reaction, it is advantageous to maximise the surface area to intensify the response to gas. Accordingly, commercial gas sensors take the form of highly porous oxide layers, which are either printed down or deposited onto alumina chips. The electrodes are usually co-planar and located at the oxide/chip interface. A heater track is also present usually on the back side of the chip to ensure the sensors run “hot”. This is a necessary requirement as both the interference from humidity is minimised and the speed of response is increased.
MMOS sensors do not normally discriminate between different target gases. As such, considerable care is taken to ensure the microstructure of the oxide, its thickness and its running temperature are optimised to improve selectivity. In addition, selectivity is further enhanced through the use of catalytic additives to the oxide, protective coatings and activated-carbon filters. Figure 8 shows a typical MMOS semiconductor sensor.
Despite such diversity in technology for gas detection the choice is often defined by the requirements of the application. The following five applications demonstrate how the choice of sensor technology is influenced by a combination of the technical requirements and the needs of the gas detector manufacturer. Whilst this list of applications is far from comprehensive it does represent some of the major applications of gas detection in both industrial and domestic environments.
In this application both portable and fixed-point gas detectors are used to protect people and property from the risks posed by toxic and flammable gases as well as the depletion in levels of oxygen. This application is regulated by health and safety legislation and the gases detected have occupational exposure levels set by each country.
The most widely used sensors for this application are electrochemical sensors and pellistors. The accuracy, resolution and stability of electrochemical sensors make them ideal for this demanding application and, importantly, as they do not require any significant power they are ideal for use in battery powered portable gas detectors.
Although infra-red sensors for detecting flammable gases are used the sensor of choice in the majority of instances remains the pellistor primarily for its broad spectrum sensitivity to flammable gases.
Both electrochemical sensors and pellistors are small and are ideal for integration into portable gas detectors. Even so, the size of the sensors has been shrinking over the last 10 years so enabling gas detector manufacturers to make smaller and smaller instruments.
By testing the composition of gases in the flue of a boiler or furnace it is possible to check for and adjust combustion efficiency, ensure gaseous emissions remain within permitted environmental levels and importantly to ensure the device is operating safely and not producing dangerously high levels of carbon monoxide.
Gas detector manufacturers have developed a range of portable instruments that are used for regular (though not continuous) testing of flue gases. Again electrochemical sensors offer the best technical solutions for this application. Their performance meets the very exacting requirements of emissions standards and their relatively small size allows them to be packaged into small gas detectors.
Many countries have legislation limiting gaseous emissions from vehicle engines and also requiring vehicles to be tested regularly to ensure compliance. To meet this requirement a range of exhaust gas analysers have been developed that are used within garage workshops and test centres. Depending on the country, the gases that are measured are oxygen, hydrocarbons, carbon dioxide, carbon monoxide and nitric oxide. CO, CO2 and hydrocarbons are routinely measured using infra-red sensors whilst O2 and NO are measured with electrochemical sensors.
Car manufacturers are becoming increasingly concerned with the quality of air inside vehicles and a number are introducing sensors to monitor air quality and control the recirculation if the quality of the air outside the vehicle deteriorates. This application requires a simple, long-life sensor that is able to detect changes from the “norm” in CO and NOx levels that is also very low cost. For this application semiconductor sensors have proved the ideal solution.
It is necessary to monitor the concentration of gases, primarily oxygen and anaesthesia agents administered to patients during surgery. To achieve this the anaesthesia workstations are generally equipped with a gas detection module.
Various technologies of gas sensors are used. Oxygen in the “breathing circuit” is measured by both electrochemical and paramagnetic sensors. For this application electrochemical oxygen sensors use solid membrane diffusion barriers and hence measure the partial pressure of oxygen rather than the % volume concentration.
Infra-red technology is widely used to monitor levels of anaesthesia agents and expired carbon dioxide (capnogrophy).
By far the largest single application of gas sensors in the home is in carbon monoxide alarms with over 10 million sold each year. This application requires sensors that are long life and reliable but do not need the precision of sensors used in industrial hygiene applications. Three different types of sensor are used in CO alarms: semiconductors, electrochemical and biomimetic. Semiconductors are relatively high power devices and are confined to use in mains powered alarms whereas low power electrochemical sensors are widely used in battery operated alarms.
Demand Control Ventilation (DCV) is a growing application in HVAC where ventilation in rooms or indoor spaces is controlled in line with its occupancy. Carbon dioxide levels are used as an indicator of occupancy and are measured with infra-red sensors mounted either directly on the wall or in ventilation ducts.
Clearly there are many other applications where gas sensors are used, for example oxygen sensing in internal combustion engines with lambda probes (zirconium oxide electrochemical sensors), and as such those mentioned in this article should not be considered as an exhaustive list. The same applies for gas sensor technology as well.
So what of the future? Will the same type of technologies be used in 10 years time or will they have been replaced by new technologies? The next 10 years will probably not see a revolution in the technologies used to sense gases but more of an evolution as existing technology continues to be developed to meet the changing needs of the gas detector manufacturers. It is likely that infra-red based sensors will become more widely used, particularly with the development of low cost solid-state sources and a reduction in the cost of sensors. Gases will always be present in industrial, commercial, environmental and domestic environments and the need to monitor these for reasons of safety and comfort will always be with us, as will gas sensors.
Published: 10th Jan 2005 in Health and Safety International