Chemical sensors Flashcards
Chemical sensors
The focus is on the detection of gaseous species, in particular combustible gases such as CO, H2, alcohols, propane and other hydrocarbons. There are two main alternative approaches:
- Semiconducting metal oxides: A chemical reaction between oxygen and the combustible gas occurs at the surface of the solid, changing the resistance of the solid. To make the resistance sensitive to such a chemical activity, we must select oxides with special forms or properties, and with special additives.
- FETs: MOSFET with chemical activity changing the potential at the “gate”. It is a modified MOSFET: it has built-in amplification which leads to high sensitivity.
Metal oxide sensors
A main example is Figaro’s Taguchi sensor, and they have in general several problems such as poor selectivity usually, but for many applications are well compensated for by the low cost of the sensor and by its sensitivity in the detection of combustible gases. The operation is based on the decrease of resistance of a layer of powdered SnO2 if a combustible gas is present in the ambient atmosphere.
Operating principle: in air, oxygen adsorbs on the surface, dissociates to form O-, with the electron exctracted from the semiconductor. This electron excitation tends to increase the resitance (assuming an n-type semiconductor). In the presence of a combustible gas such as hydrogen, it reacts with the adsorbed O- to form water and the electron is re-injected into the semiconductor, tending to decrease the resistance (CO for example reacts giving CO2). A competition results between oxygen removing electrons and the combustible gas restoring these electrons. So the steady-state value of the resistance depends on the concentration of the combustible gas. Reactions take place over a well-defined threshold temperature (e.g. 200-400 °C usually, but it depends on the gas investigated and on the form and material used as metal-oxide semiconductor).
Typical metal-oxide semiconductors used are: SnO2 , ZnO, TiO2 , WO3. A catalyst is usually provided to accelerate the reaction rate so that the response of the sensor is rapid.
Metal oxide sensors: the role of the catalyst
Metal-oxide gas sensors need a catalyst deposited on the surface (finely distributed over the particle surface to maximize metal-oxide/ gas interaction) of the semiconductor to accelerate the reaction and increase the sensitivity. A catalyst is a material that increases the rate of chemical reactions without itself changing (it lowers the activation energy of the reaction).
Considering the reduction phase, without the catalyst we would have the reactions:
H2 –> 2H ()
O2 –> 2O ()
2H + O –> H2O ()
Equations () require a huge input of energy and will not happen at moderate temperature. The energy is regained in (**), but the reaction is stopped kinetically by the “activation energy” of ().
On the other hand with Pt (or Pd) available as a catalyst we have:
H2+ 2Pt –> 2Pt-H ()
O2+ 2Pt –> 2Pt-O ()
2Pt-H + Pt-O –> 3Pt + H2O ()
In this case (*) require very little energy input. A hydrogen molecule adsorbed on the Pt surface readjusts its bonds easily to form Pt-H groups.
The Taguchi sensor
Gold electrodes are deposited on a small ceramic tube (3 mm long, 1.5 mm diameter). A paste of semiconducting powder is applied on the outside of the tube: it is a complex preparation procedure so exact reproducibility cannot be expected. A heater element (Pt coil) in the center of the tube brings the oxide coating to the desired operating temperature. EVOLUTION: "THICK FILM“ structures where the powder is deposited on a flat substrate by a technique such as silk screening or screen printing (a kind of printing technique). Silk-screened RuO2 or evaporated Pt (more common) are possible heater materials. Industrial applications: •Gas monitoring systems •Gas leak detectors in factories •Analysis equipment •Fermentation control •Fire and toxic gas detector Domestic applications: •Gas leak alarm •Ventilation control •Cooking for microvawe oven •Humidity control
2D IMPLEMENTATION
The 2D implementation consists in an intergration: micromachined metal oxide sensor and a thin membrane of low thermal conductivity (for isolation between active area and substrate). The overall thickness of the film should be as small as possible to maximize sensitivity.
The advantages are:
• Thin thermally insulated films/membranes can reduce the thermal budget and result in faster response
• Thin film heaters (evaporated Pt) to replace the heater coil in Taguchi sensors
• Fast heating/cooling of thin film heaters so new temperature programmed sensing modes of operation
• Low power consumption (30- 150 mW against 200 mW - 1 W of screen printed devices)
• Signal processing electronic on the same chip so lower costs per unit
• New processing steps for metal oxide deposition resulting in nm scale grain size for improved sensitivity
Microhotplate technology
The integration of SnO2 films on microhotplate substrates:
– Chemical Vapor deposition of SnO2
– Deposition temperature of 500°C
– Deposition pressure of 1 atm
Huge grains and non-uniform grain distribution result in irreproducible sensing characteristics of microhotplates.
Control of size, morphology and distribution of grains can result in reproducible sensing characteristics. Nanoparticle Engineering involves using metals as seed layers to aid the initial nucleation and growth of SnO2. The goal was to control grain size to result in uniform morphology.
Selectivity in metal-oxide sensors
The ultimate objective for metal-oxide sensors would be a series of sensors each of which would respond to only one gas to have high selectivity. Of course it is difficult to obtain (Pd for example will catalyze the oxidation of CO, H2 and all organic molecules) but to induce some selectivity, we have several parameters available:
-the catalyst
-the temperature
-“filters” that restrict the presence of some gases on the sensor
-even the particle size of the metal-oxide sensor can affect the selectivity
TEMPERATURE
Temperature is important because some gases (e.g. alcohols or CO) are easier to oxidize than others (e.g. CH4 and alkanes): a low working temperature induces selectivity towards alcohols or CO and a high temperature induces selectivity towards CH4. Because of this temperature selectivity, a time-varying temperature is sometimes used, to attempt to show a spectrum of gases. The easily oxidizable gases can be analysed at low temperature (1st sampling), then an exchange of ambient gas is performed (2nd sampling) and temperature is raised to make measurement of gases such as alkanes more accurate.
FILTERS
Another approach to selectivity is represented by FILTERS. For example an SiO2 thin film on SnO2 can be permeated only be small molecules such as H2. Ultrafine SnO2 rejects methanol. A carbon cloth has been used to prevent highly reactive or large molecules from reaching the sensor.
CATALYSTS
Different combinations of catalysts can change the selectivity significantly. The theory behind their use is not entirely clear at the present. Their use is, to a great extent, empirical.
ARRAY OF SENSORS
An approach to selective gas sensing is to provide an array of sensors, each one with a significantly different response spectrum for various gases. For example, one sensor could be chosen to change resistance by a factor of 2 in 10 ppm H2 and by a factor of 4 in 100 ppm CO. Another sensor could be chosen such that it changes resistance by a factor of 4 in 100 ppm H2 and 2 in 100 ppm CO. Measurement of a H2/CO mixture by the two sensors would, in principle, provide the complete analysis. If, for instance, we can perform a calibration of the different sensors array elements with respect to the different gases potentially present in a specific atmosphere we will extract the different αx,Gy , that is to say the sensitivities of the array element x to the gas Gy.
Metal oxide sensors: long-term drift
After a SnO2/Pd-based sensor has been held at room temperature for a time and then heated to operating temperature, it indicates the presence of gases. A high conductivity spike is observed and it occurs whether or not a combustible gas is in the ambient atmosphere. This high conductivity spike almost vanishes after a few minutes, but a residual excess conductivity is retained for days or weeks after storage at room temperature. If one wants a reasonably accurate measure of a low combustible gas partial pressure, it is necessary to “burn-in” the sensor for several days until the initial spike has decayed to a value low enough to be considered a negligible background. Manufacturers burn-in the sensors for 3 to 7 days before they calibrate it and recommend a similar time for the user before they use the resulting calibration.
The probable reason for this spike is the adsorption of combustible gases from the atmosphere while the sensor is stored at room temperature. Room temperature is too low to support catalytic oxidation of the contaminating gases. When the sensor is heated to operating temperature the pre-adsorbed gases react with the surface oxygen ions, injecting electrons, and giving a false signal.
Fet devices
Field effect transistors (FET) can be sensitive to some gases or ions if the gate is exposed (by properly modifying the gate).
The most studied is the FET with Pd gate for H2 detection: H2 dissolves in the Pd gate and affects the gate voltage (and consequently the resulting IDS or VGS). Such a FET, together with other chemical sensing FETs, can be termed a ChemFET or GasFET.
H2 is catalytically split up in H+ at the Pd surface, is dissolved in the Pd, moves at the Pd/SiO2 interface and forms a dipole layer: there is a change of the work function difference between the metal and the SiO2 and a change of the gate voltage.
LEAK DETECTION
Hydrogen as a tracer gas can be used for locating leaks or for leak detection. In the case of locating a leak, the diluted Hydrogen is simply injected into the test object and a Hand Probe connected to a Hydrogen Leak Detector is used to search for leaks.
The Detector indicates with an audio signal that increases the closer you get to the point of leak.
ION SENSING
A MOSFET is prepared but without the metal gate. The gate voltage (to produce the channel) is applied to the reference electrode, then the concentration of H+ ions (acid solution) or OH-ions (basic solution), modulates the gate potential at the solution/insulator interface and consequently the channel conductivity.
The reference electrode can be a problem (stability, poisoning of the solution or by the solution, dimensions, …).
