WO2019207291A1

WO2019207291A1 - Sensors for determining volatile compounds based on electrical conductivity and cataluminescence measurements of metal oxide films

Info

Publication number: WO2019207291A1
Application number: PCT/GB2019/051125
Authority: WO
Inventors: Norman Ratcliffe; Ben DELACY COSTELLO; Timothy Cox; Richard Luxton; Janice Kiely
Original assignee: University Of The West Of England
Priority date: 2018-04-24
Filing date: 2019-04-23
Publication date: 2019-10-31
Also published as: GB201806635D0; GB2573118A

Abstract

The invention relates to the field of sensors for detecting volatile compounds, especially explosive materials. The sensors can also be used in medical settings for helping to diagnose diseases, by detecting volatile compounds in the breath or body fluids such as urine or saliva. The sensor of the present invention for detecting volatile compounds comprises: a film of metal oxide (1); a heater (3) configured to provide heat energy to the film of metal oxide, and/or a UV light configured to irradiate the film of metal oxide; an electrical conductivity meter (4) configured to measure the electrical conductivity of the film of metal oxide; and a photometer (7) configured to measure cataluminescence derived from an interaction between a volatile compound and the film of metal oxide.

Description

SENSORS FOR DETERMINING VOLATILE COMPOUNDS BASED ON ELECTRICAL CONDUCTIVITY AND CATALUMINESCENCE MEASUREMENTS OF METAL OXIDE FILMS

Field of Invention

The invention relates to the field of sensors for detecting volatile compounds, especially explosive materials. The sensors can also be used in medical settings for helping to diagnose diseases, by detecting volatile compounds in the breath or body fluids such as urine or saliva.

Background to the Invention

Gas sensors based on heated metal oxide semiconductor materials which are capable of reversibly altering their electrical resistance in the presence of volatile organic compounds (VOCs) are known. Such sensors may be used for the detection of non polar gaseous organic compounds such as hydrocarbons, although they are usually more sensitive to polar compounds such as ethanol. There are mature technologies available on the market, based on metal oxide semiconductors. These are predominantly focused towards toxic gas detection, environmental monitoring and flammable gas detection. Current available technology has sensitivities broadly in the range of 30 to 5000 ppm, depending on the compound being detected.

As an alternative to heating the sensor, ultraviolet (UV) light can be used to enhance the sensitivity of semiconducting materials towards volatile compounds, at ambient temperatures. For example, earlier work by some of the inventors, W02009/037289, discloses a gas sensor that works by irradiating a metal oxide material with UV light. The presence of a volatile compound can be detected by monitoring the change in the electrical conductivity of the material. See also Costello et al 2008 which discloses a UV activated sensor based on zinc oxide, for detecting acetone, acetaldehyde, and some hydrocarbons. Metal oxide-based sensors known in the art have been used to detect some volatile compounds including explosive components, see Sagar et al 2014 and Brudzewski et al 2012, but achieving the required sensitivity and selectivity is a barrier to their more widespread use, especially in situations where it is necessary to detect low levels of a particular volatile compound amongst an array of different volatile compounds. For example, in explosive detection the compounds of interest, such as the explosive material nitroglycerin, may be present at very low concentrations in air which may also contain many other interfering compounds, such as nail polish remover or alcoholic drinks. Other interferents may include, for example, acetone, ethanol, perfume and aftershave, cleaning products and water vapour.

Currently in the field of trace and vapour explosives detection more commonly used sensing systems are based on vapour detection. The vapour may be sampled directly from the environment or produced from the volatilisation of particulates. Suitable equipment includes, for example: (i) ion mobility spectrometer (e.g. Smiths Detection's Sabre); (ii) Thermo Electro Corporation's Defender combining Gas Chromatography with differential IMS; (iii) Scintrex's Chemilux systems based on luminol generated chemiluminescence; (iv) portable mass spectrometry based instruments e.g. Smiths Detection’s Guardion; (v) surface acoustic wave detectors combined with GC e.g. zNose and (vi) systems based on amplifying fluorescent polymers: FIDO range of detectors from FLIR. Although these instruments may be portable, they all constitute expensive and often bulky pieces of equipment. Some also contain radioactive sources, or rely on a supply of consumables. Accordingly they may not be suitable for widespread issue or deployment.

The present invention aims to address some of the limitations of prior art sensors, in particular by providing a sensor that has enhanced selectivity. This is particularly useful in the field of explosive detection. The sensors of the present invention can also provide good sensitivity, and be compact and inexpensive.

Summary of the Invention

According to a first aspect, the invention provides a sensor for detecting volatile compounds, the sensor comprising: a film of metal oxide; a heater configured to provide heat energy to the film of metal oxide, and/or a UV light configured to irradiate the film of metal oxide; an electrical conductivity meter configured to measure the electrical conductivity of the film of metal oxide; and a photometer configured to measure cataluminescence derived from an interaction between a volatile compound and the film of metal oxide. In contrast to prior art sensors, the sensor of the present invention can measure both electrical conductivity and cataluminescence i.e. has dual modality. The phenomenon of cataluminescence and its application to sensors is much less well known than the conductivity sensors discussed above, but has been reported previously, see for example CN102093890. However, it's believed that sensors that monitor both conductivity and cataluminescence are novel and have important advantages.

In particular, having dual modality allows the sensor to achieve enhanced selectivity, and can vastly reduce the level of false positive results. This is because the sensor can be designed so that the target compound, for example nitroglycerin in the case of explosive detection, can give a signal under both modalities with a well defined ratio of response between the change in resistance and the cataluminescence, which will allow it to be distinguished from interferents which may only give a signal under one modality, or may give signals in both modalities, but at a different ratio than the target compound. The sign of response is also important. While the cataluminescence is positive or absent, the change in resistance can be positive or negative, adding a further means of distinguishing different compounds. The specificity of the sensor can be optimised to the target relative to common interferents, in both modalities. Furthermore, while resistance may be easier to measure, a cataluminescence signal tends to recover much quicker than resistance after exposure to a target and is typically less sensitive to the presence of water vapour. Harnessing the advantages of both resistance and cataluminescence monitoring is key to the success of the invention.

The metal oxide-based devices of the present invention have the potential to be miniaturised, relatively low cost, highly sensitive, with tuneable selectivity and capability for remote detection. These features will enable wider issue and allow the possibility of distributing a large number at strategic locations e.g. alongside surveillance cameras in an airport or public space.

According to a second aspect, the invention relates to a method of detecting a volatile compound in a gaseous environment, the method comprising the steps of: providing a sensor which includes a film of metal oxide; activating the film of metal oxide, by heating it, and/or by irradiating it with UV light; exposing the sensor to the gaseous environment; and monitoring the electrical conductivity of the film of metal oxide and cataluminescence derived from the interaction between the volatile compound and the film of metal oxide.

According to a third aspect, the invention relates to the use of a sensor according to the first aspect of the invention, to detect a volatile compound. The sensor of the present invention is particularly useful in security applications where the volatile compound is an explosive, explosive precursor, or a marking agent used in explosives, in medical settings where the volatile compound is a biomarker in the breath or in the headspace above a body fluid, in agricultural or retail settings where the volatile compound is a marker for ripeness or decay in food or crops, or to detect or monitor an environmental contaminant of the air or water.

Brief Description of the Drawings

Embodiments of the invention will now be described by way of example only with reference to the accompanying figures in which:

Figure 1 is a schematic view of a sensor according to a preferred embodiment of the invention, which is heat-activated;

Figure 2 is a schematic view of a sensor according to a preferred embodiment of the invention, which is UV-activated;

Figures 3 and 4 are schematic views of a sensor according to a preferred embodiment of the invention, showing the positioning of the electrical conductivity meter;

Figure 5 is a schematic view of a sensor according to a preferred embodiment of the invention showing the photometer arrangement;

Figure 6 shows the cataluminescence response of a doped and undoped zirconium oxide sensor according to a preferred embodiment of the invention;

Figure 7 shows the resistance response of doped and undoped zirconium oxide sensors according to preferred embodiments of the invention; Figure 8 shows the change in cataluminescence response of doped and undoped zirconium oxide sensors according to preferred embodiments of the invention;

Figure 9 shows the response of a sensor according to a preferred embodiment of the invention to alternate vapour inputs of l6ppb DMNB and laboratory air; Figure 10 shows the cataluminescence response of a europium doped zirconium oxide sensor according to preferred embodiments of the invention;

Figure 11 shows the resistance change response of a europium doped zirconium oxide sensor according to preferred embodiments of the invention;

Figure 12 shows the ratio between the cataluminescence response and resistance change response of a europium doped zirconium oxide sensor according to preferred embodiments of the invention;

Figure 13 shows the response of a zinc oxide sensor according to a preferred embodiment of the invention to (a) hydrogen peroxide and (b) ethanol;

Figure 14 shows the cataluminescence and change in resistance responses of an yttrium oxide sensor according to preferred embodiments of the invention;

Figure 15 shows the ratio between the cataluminescence and change in resistance responses of an yttrium oxide sensor according to preferred embodiments of the invention;

Figure 16 shows the cataluminescence and change in resistance responses of a tungsten oxide sensor according to preferred embodiments of the invention;

Figure 17 shows the cataluminescence response of a tungsten oxide sensor according to preferred embodiments of the invention;

Figure 18 shows the resistance change response of a tungsten oxide sensor according to preferred embodiments of the invention; Figure 19 shows the ratio between the cataluminescence response and resistance change response of a tungsten oxide sensor according to preferred embodiments of the invention.

Description

The invention relates to a sensor for detecting volatile compounds. By volatile compounds we mean compounds that undergo some degree of evaporation at room temperature, and exist partially (possibly in very small part) or entirely in the gaseous state at room temperature. This type of sensor is also typically referred to as a gas sensor.

The sensor comprises a film of metal oxide. As is known in the art, the film of metal oxide usually comprises what is known as a "thick film" of metal oxide, which is a continuous film usually of thickness in the order of 100 microns, but nanoparticular, nanoporous or nanostructured films are also possible.

Where nanoparticles are used they usually have an average particle size of 10 to 250 nm, preferably 20 to 100 nm, more preferably 40 to 60 nm, most preferably about 50 nm. Average particle size in this field is measured by Transmission Electron Microscopy (TEM), where the resulting image can be analysed to determine the particle size distribution.

The film of metal oxide is usually 0.001 to 0.5 mm thick, preferably 0.01 to 0.3 mm thick, most preferably about 0.2 mm thick, especially in the case of nanoparticular films. Continuous films are generally thinner than nanoparticular films, and could be just one micron, i.e. 0.001 mm, thick. The film is supported on a substrate, as discussed below.

The metal oxide can be selected depending on the volatile compound of interest, and is typically selected from EJPAC Groups 3, 4, 5, 6, 8, 9, 10, 11, 12, 13, 14 and 15 of the Periodic Table. Mixed metal oxides, i.e. two or more metal oxides, are often used in the present invention.

In one embodiment, the metal oxide is one that gives n-type conductivity. The inventors have discovered that the ability of a material to exhibit cataluminescence is often correlated to the ability of a material to exhibit photoluminescence. Accordingly, the skilled person will be able to select metal oxides that might be advantageously used in the invention by determining their photoluminescence properties.

In a preferred embodiment of the invention, the metal oxide is selected from one or more of the oxides of yttrium, aluminium, tin, indium, zinc, tungsten, chromium, titanium, nickel, cadmium, iron, niobium, manganese, gadolinium, vanadium, and zirconium. Examples include Y2O3, Ti0₂, Nb₂0₅, La₂03, Mn0₂, Gd₂03, V₂0₅, Zr0₂, SrZrCb, ZrEuO_x and ZrTbO_x.

Most preferably the metal oxide is selected from an oxide of yttrium, tungsten, zinc and/or zirconium, including europium doped zirconium oxide, and terbium doped zirconium oxide. Metal oxide nanoparticles are commercially available and can be selected, and then deposited on a substrate to make the sensor.

It can be advantageous for the metal oxide to comprise a dopant, preferably at a level of 1 to 10% by mole fraction, most preferably 4 to 6 % by mole fraction. One reason for this is that doping can reduce the resistance of the metal oxide, making it easier to monitor. For example, europium doped zirconium oxide has a lower resistance than zirconium oxide. In this way metal oxides that have been used for cataluminescence sensors, but where the resistance would be too high for a resistance sensor, can be modified by means of a dopant to make the metal oxide suitable for the present invention dual modality sensor. Another reason is that doping can lead to enhancement of the cataluminescent signal. An example is given below, of how doping of zirconium dioxide with rare earth metals (europium and terbium) leads to enhanced cataluminescence and reduced film resistance. For other metal oxides, such as yttrium, zinc and tungsten doping is not required, as demonstrated below.

Preferred doped metal oxides are zirconium oxide doped with europium or terbium.

The sensor of the present invention comprises a heater configured to provide heat energy to the film of metal oxide, and/or a ETV light configured to irradiate the film of metal oxide. The purpose of the heater or light is to activate the film of metal oxide, either by heating it, or by irradiating it with UV light. Suitable heaters and UV lights are known in the art to the skilled person.

In one embodiment, a platinum resistance heater is used. Changes in the resistance of the platinum heater are used to give a direct read out of the temperature. The film is usually heated to 100 to 500 °C, preferably 200 to 400 °C. Different effects can be seen at different temperatures. The skilled person can, by empirical means, determine which temperatures gives the best sensitivity/selectivity combination for different volatile compounds, so can design the sensor to best detect the particular volatile compound of interest. In a preferred embodiment of the invention, the sensor is activated by heat, and comprises a heater rather than a UV light.

In the alternative embodiment, where the sensor is activated by UV light, the UV light is usually a low power UV-LED. Typically the ultraviolet light preferably has a wavelength in the range 360-400nm, that is close to (at the high end of the wavelength range) visible light. Conveniently, UV-LEDs may be used as the source of activating UV light for the sensor, since they emit in the preferred range and, being small, enable the invention to be performed with portable apparatus. The skilled person can easily select appropriate UV light for the material being used. Beneficially, certain UV- LEDs which can focus the emitted UV onto the surface of the film via an integral lens may be used. The UV light is usually positioned close to the film.

An advantage of the use of UV light as an activator of the metal oxide film, is that UV activation can take place at ambient room temperature. Accordingly, no heating of the metal oxide material is required and the overall energy requirement is thus kept to a minimum.

In the unactivated state, the metal oxide film is not optimally sensitive to volatile compounds. It is well known that the metal oxides develop optimum sensitivity to volatile compounds during activation. Most prior art sensors involve activation with heat. Some involve activation by irradiation with ultraviolet light. Before activation, the metal oxide usually has a high electrical resistance and is insensitive to high concentrations of various volatile organic compounds, high concentrations in this context meaning in general concentrations of greater than lOOOppm, whereas, on or following activation, typically the electrical resistance initially falls substantially and thereafter reaches a steady state value, after which the material is activated and exhibits higher sensitivity to volatile compounds. The heating or irradiation may be continuous over the period of activation, or pulsed for the purpose of saving power.

In one embodiment, the sensor is operated with pulsed heating and/or UV light activation. This can lead to advantageous sensing results. For example, as the sensor has a high surface area, it can act as a pre-concentrator (gas collector) when operated at low temperature, and then be heated to higher temperature in a pulsed arrangement. This can lead to enhancement of cataluminescence signal. Enhancement factors in the cataluminescence signal of greater than ten have been achieved for pulsed operation as compared with the static temperature operation.

In order to detect a volatile compound, the sensor of the present invention comprises both an electrical conductivity meter configured to measure the electrical conductivity of the film of metal oxide and a photometer configured to measure cataluminescence derived from an interaction between a volatile compound and the film of metal oxide. Accordingly, the method of the invention involves monitoring both the electrical conductivity and cataluminescence. This is believed to be a key difference from the prior art sensors which detect either conductivity or cataluminescence but not both.

In other words, the films of the present invention are activated either by (i) heating or by (ii) illuminating with ultraviolet light, or (iii) both. When the compound of interest, i.e. the target, comes into contact with the activated metal oxide surface, two processes of interest may occur in the presence of the target: (a) generation of light (cataluminescence) and/or (b) a change in the electrical conductivity of the film (the electrical resistance, or ability to pass electrical current when a voltage is applied to the film). By measuring the detailed changes in conductivity of the metal oxide film and the properties of the emitted light (cataluminescence) of sensor, it is possible to measure the concentration of a particular, pre-selected, volatile compound (the target) with high specificity and sensitivity.

Suitable electrical conductivity meters are known to the person skilled in the art. The sensor including the film of metal oxide is connected to a voltage source, for example up to 12 volts, 9 volts being typical, and the change in current in response to the presence of volatile compounds is monitored for example via a scanner and electrometer and connected through a suitable interface to a computer. A programmable electrometer such as a Keithley model 617 electrometer is suitable for current measurements. The temperature and humidity data may also be recorded.

In the case of UV activation, the intensity of the UV light may be monitored using photodiodes. The photodiode can be used to measure and thereby stabilise the LED output, via a feedback control loop.

In one embodiment, the sensor devices are based on a micro-hotplate device. On the sensor side are interdigitated gold electrodes on a substrate. Gold pads are connected to the interdigitated electrodes. These pads would have gold wires bonded to them. The substrate can then be mounted on to a transistor outline header package, commonly referred to as a TO header in order to package it. The gold wires are connected to the pins of the TO header. The pins are then, in turn, connected to the electrometer/voltage source. The electrodes can, for example, have a gold track width of l50pm and electrode gaps of lOOpm. The electrodes can be on the front of a substrate of dimension 3x3mm. The reverse side can include a heater, such as a platinum resistance heater. The temperature can also be deduced from a measurement of the resistance of the platinum heater.

The sensor of the present invention includes a photometer, to measure cataluminescence. By cataluminescence, we mean specific luminescence derived from an interaction between a volatile compound and the film of metal oxide. The cataluminescence measured is derived from the interaction between the volatile compound being detected, and the metal oxide film. Is it unclear to researchers whether the cataluminescence originates from the volatile compound, the metal oxide or, where present, a dopant within the metal oxide film.

Photometers have been used less often in the art than electrometers. Where they have been used it has been in conjunction with heated metal oxide sensors, but not with UV light activated sensors. The concept of measuring cataluminescence from a UV- activated metal oxide sensor is novel. Where UV irradiation is used to activate the metal oxide film, on exposure to the volatile compounds of interest, light will be emitted, at a different frequency to that with which it is irradiated. There are two methods of operation, the first in which filters are used between the photometer and the sensor, to block out the activating UV. In the second the UV light is turned off and light emission is then measured. This is possible due to an after-glow period, where the light emission process is able to occur for a short while after the UV is turned off. The skilled person can easily select a suitable photometer.

In one embodiment, to measure the photon flux due to cataluminescence, a small photomultiplier tube (PMT) was used in photon counting mode. The output from the PMT is a series of voltage pulses (e.g. 3 V, 10 nanoseconds width) where each pulse corresponds to a single photon at the input to the PMT. A pulse counter then counts the number of voltage pulses in a set period, e.g. 0.5 seconds. The cataluminescence signal as a function of time is then the number of pulses counted in each time period. Specific photomultipliers that can be used include (i) A Hamamatsu H7828 series photon counting head, and (ii) A Hamamatsu silicon photomultiplier module: C13366-3050GD featuring a 3 x 3 mm detection area with thermoelectric cooling.

The spectral (i.e. wavelength) distribution of the emitted cataluminescence can be determined by using an EG and G 1460 Optical Multichannel Analyser. This features a Jarrell Ash quarter meter polychromator combined with a cooled diode array detection system.

For the electrical conductivity monitoring, the metal oxide is activated by heating or UV irradiation, with the intensity set to an optimum value. Voltage is then applied to the sensor and the current monitored over time. Once the current has stabilised, the sensor can be exposed to the target volatile compound or compounds, the presence of which is detected by virtue of a change in current. In the present invention, for the first time, cataluminescence emission data can be considered in addition to the conductivity data, to improve sensitivity and selectivity of the sensor.

Regarding processing of the cataluminescence and resistance data, conventional machine learning approaches can be used to create a profile for different target compounds and interferents in a laboratory, against which real life samples can be compared. Pattern recognition can be used to tune the sensors and improve specificity.

In particular, in the laboratory the response of a multimodal sensor to the input of a known concentration profile of the target compound can be determined. Ideally, one form of test input would be a top hat function, a second form could be a delta function of concentration. The response can then be recorded, which would involve measuring (i) the resistance change and (ii) light output at different stages including (a) before the input i.e. to establish the baseline, (b) during the input, and (c) during the recovery phase after the target has been removed. The sensor response to this known input concentration of a target can then characterised by specific response parameters, for example: the sensitivity function (i.e. response per concentration unit); the sign of the sensitivity function (i.e. an increase or decrease in response); the input response time (e.g. related to the rise time, via the first derivative); and the recovery response time (e.g. related to the recovery time via the first derivative). For a recovery which is exponential in time, analysis might equate to plotting log(response) vs l/t and determining the slope. Such a set of parameters could be used to characterise the sensor response in each of the two modalities. Information from each mode can then be combined to produce a much more accurate profile than with one mode alone. The ratio of cataluminescence response to resistance response is a powerful tool in the analysis.

The film of metal oxide is supported on a substrate. The supporting substrate is often ceramic, but can be any suitable material, and can include an electrode which is part of the electrical conductivity meter, for example can include interdigitated electrodes. The metal oxide can be applied to the surface of the substrate as a finely-ground dispersion in a suitable liquid carrier, forming a material having the consistency of a paste. Having been applied to the substrate surface, the paste is allowed to dry in air. However, the metal oxide may in other instances comprise a continuous film or layer applied to a suitable supporting substrate in any suitable way. For example the metal oxide particles maybe be applied to the substrate by screen printing or inkjet printing.

In one embodiment of the invention, an array of sensors is used. The array can, for example, comprise 2 to 100 sensors, preferably 5 to 50 sensors. The reason for this, is that different metal oxide film type at different temperatures, can give a different response to different volatile compounds. For example, a range of temperatures would typically be 100 - 500°C, preferably 200 to 400°C, more preferably at least 250°C with different metal oxides, for example tin oxide, zinc oxide, mixed tin oxide / zinc oxide, tungsten oxide and zirconium oxide. Most targets and interferents give a response (usually a decrease in resistance) on all of the sensor types. The response varies for each volatile compound / metal oxide / temperature combination. Thus, by using an array consisting of sensors of different materials and temperatures, a finger print response characteristic of a particular compound could be produced. This is the equivalent of the olfactory process in nature, i.e. an electronic nose.

In addition, or as an alternative to an array of sensors, the present invention relates to a kit comprising a sensor or sensor array according to the invention, and another type of analytic device, such as a gas chromatograph. For example, in order to enhance the qualitative analytical capabilities of the invention, the method may involve the use of a gas chromatography column in series with the metal oxide sensor, whereby the chromatography column will provide the means to distinguish between and identify different volatile organic compounds the presence of which in more non-specific terms has been identified by the metal oxide. MEMS (Micro-Electro-Mechanical Systems) technology can be used to ensure the sensor is miniaturised, and therefore as widely deployable as possible.

The present invention also relates to a method of detecting a volatile compound in a gaseous environment, the method comprising the steps of: providing a sensor which includes a film of metal oxide; activating the film of metal oxide, by heating it, and/or by irradiating it with ETV light; exposing the sensor to the gaseous environment; and monitoring the electrical conductivity of the film of metal oxide and cataluminescence from the film of metal oxide. The film of metal oxide is preferably activated by heating to 100 to 500 °C, preferably 200 to 400 °C.

Similarly, the present invention also relates to the use of the sensor of the invention, to detect a volatile compound. The sensor, and the method of the invention, are particularly useful in security applications, wherein the volatile compound to be detected is an explosive or explosive precursor or a marking agent, preferably wherein the volatile compound is hydrogen peroxide, nitroglycerin (NG), ethylene glycol dinitrate (EGDN), 2,3- dimethyl-2,3-dinitrobutane (DMNB), 2,4 dinitrotoluene (2,4 DNT) and/or Triacetone triperoxide (TATP).

However, the sensor and the method of the invention can also be used in the medical arena. In this case, the volatile compound would be a biomarker in the breath or in the headspace above a body fluid, preferably wherein the volatile compound is acetone, acetaldehyde, pentane or other hydrocarbons, ammonia, hydrogen sulphide, hydrogen cyanide, nitric oxide, isoprene, hydrogen peroxide, methane and/or hydrogen. The method according to the present invention has particular application in terms of medical diagnostics, being able to detect compounds such as acetone, acetaldehyde and pentane when in the vapour state, which compounds are known to be potential indicators of disease or of well-being and are present in body fluids including breath and vapours emitted, for example, by stool, urine, blood and sputum. For example, increased levels of acetone in the breath are indicative of uncontrolled diabetes mellitus (ketoacidosis). Increased levels of pentane and other hydrocarbons in the breath have been linked to increased oxidative stress in diseased such as cancer, arthritis and during myocardial infarction.

Further, the sensor and method of the invention could be used in agriculture or retail to detect a volatile compound which is a marker for ripeness or decay in food or crops, preferably wherein the volatile compound is a Cl to C6 alcohol, a sulphide such as dimethyl disulphide, an aldehyde, a ketone, or ethylene. Rotting potatoes give off large concentrations of Cl to C6 alcohols and sulphides such as dimethyl disulphide. Fungally contaminated wheat gives off a range of aldehydes, ketones and alcohols. Ethylene is a marker of ripeness in fruit etc.

Finally, it is envisaged that the sensor and method of the invention could be used in environmental applications, where the volatile compound is an environmental contaminant of the air or water, preferably wherein the volatile compound is methane, ozone, hydrogen sulphide, carbon monoxide, petroleum, paraffin jet fuel, hydrogen, ammonia, sulphur dioxide and/or NOx.

Embodiments of the invention will now be described by way of example with reference to the accompanying figures. Figure 1 is a schematic of the metal oxide sensor system according to a preferred embodiment of the invention, showing the dual sensing modalities.

The metal oxide film 1 is supported on a ceramic substrate 2 held at an elevated temperature (T) via a heater 3 on the rear face of the substrate. In this case the metal oxide film is of nanoparticles and is approximately 0.2 mm thick. In modality 1 : the change in resistance, AR, of the film is monitored by electrical conductivity meter 4. In modality 2: light emission (cataluminescence) 6 generated by the interaction of the target volatile compound 5 with the heated metal oxide film 1 is measured by photometer 7.

In Figure 2, the metal oxide film 1 is also a film of nanoparticles, and is approximately 0.2 mm thick. In this embodiment, the film is supported on a ceramic substrate 2, with a further substrate 8 at ambient temperature, and is activated by illumination with ultraviolet light 9, e.g. at 400 nm. In the context of UV activation the change in resistance of the metal oxide film, AR, is called Modality 3, and is monitored with electrical conductivity meter 4. In addition, the light emission (cataluminescence) 6 generated by the interaction of the target volatile compound 5 with the UV illuminated metal oxide film 1 is measured - Modality 4 - by photometer, 7.

As shown in Figure 1, Modalities 1 and 2 have been combined in a single sensor thereby offering multimodal sensing. Similarly, Figure 2 shows Modalities 3 and 4 combined in a single sensor. The target volatile compound 5 interacts with the film resulting in a change in the resistance of the film and/or in the emission of light (cataluminescence) 6. Multimodal sensing offers the prospect of greatly enhanced selectivity of detection relative to that of a single modality sensor.

Figure 3 is a schematic which shows a top view of the metal oxide sensor showing the interdigitated electrodes 10 with gaps between them 11 on the ceramic substrate 8. Figure 4 shows a top view of the sensor in Figure 3, but with a nanoparticle metal oxide film 1 is deposited onto the electrodes. The resistance of the film, R, is monitored by electric conductivity meter 4 as a volatile target compound is introduced to the sensor surface.

The cataluminescence detection process is shown in figure 5. The key sensor component is a metal oxide film 1 of typical thickness 0.2 mm, composed of nanoparticles of a metal oxide of typical diameter 50 nm. The film is heated typically to temperatures in the range 100 - 500 °C or activated by irradiation with UV light. It is preferable that the sensor is incorporated into a flow system 12 so that air containing target volatile compounds and potential interferents may be introduced for testing purposes. The air flows in 13 and out 14 of the flow system 12 are illustrated. An optical filter, 16, for example a short pass filter allowing transmission of light with a wavelength of less than 600 nm, selectively passes cataluminescence 6 whilst rejecting the majority of the black body radiation 15 which arises from the heated film. The cataluminescence 6 is then detected by a photometer 7, such as a photomultiplier that is operated in photon counting mode. The sensor also has an electrical conductivity meter, 4.

Example 1 - Synthesis and testing of zirconium oxide sensors

Synthesis of doped zirconium oxide

The synthesis of zirconium oxide nanoparticles doped with the rare earths terbium and europium was conducted in house by the applicant using a method based, as far as possible (because some of the conditions were not specified), on Zhang et al , Analytica Chimica Acta, vol 535, pl45-l52 (2005). The method is summarised as follows. 2.5ml 0.02M Europium nitrate or Terbium nitrate solution was added to lOml 0.1M Zirconium nitrate solutions. 0.1M Ammonium nitrate solution was titrated in with rapid mixing and pH was tested. When the pH was greater than 8, the product was filtered, washed with water and air dried. The next step was to calcine in the furnace at 600 °C for 3 hours. In both methods (terbium and europium doped zirconium oxide) lOOmg of each product was produced, representing approximately a 70% yield. The level of europium or terbium was around 5% by mole fraction. Synthesis of doped and undoped zirconium oxide sensors

In all sensors (undoped, europium doped and terbium doped), the sensor was made as follows. For the doped sensors, the material made as per the method above was used as the metal oxide. For the undoped material, the metal oxide used was zirconium (IV) oxide Zr0₂ - nanopowder <l00nm particle size - sourced from Sigma- Aldrich.

The metal oxide material (0.2 g) was placed in an agate pestle and mortar and sufficient water (ca. 0.4 g) was added to produce a thick slurry and the mixture was ground for 10-15 min to eliminate any air. The sensor was prepared by drop coating the oxide paste (5 microlitres, ca. 5 mg) onto the substrate (2.5 x 2.5 mm² alumina tile) - so that it completely covered the interdigitated gold electrodes printed on the obverse side of the alumina tile. (The platinum heater is printed on the reverse side). This was then allowed to dry at room temperature ca. 22 °C for 12 h prior to use.

Testing of the metal oxide sensors

There are three distinct testing scenarios which have been used: testing scenario 1 - Exposing the sensor to high concentrations of volatile compounds in a static system; testing scenario 2 - Exposing the sensors to low concentrations of volatile compounds in a stream of dry air; and testing scenario 3 - Exposing the sensors to headspace concentrations of volatile compounds. These are explained in detail as follows.

Testing Scenario 1: The sensor was housed in a light tight chamber (machined aluminium, circa 250 ml volume). The chamber is furnished with input and output tubes of internal diameter 3 mm. Test samples are input to the sensor chamber via the input tubing. In the experiment, a small volume, e.g. lml sample, of a high concentration of the volatile compound is injected directly into the sensor chamber. The sample is allowed to disperse by diffusion, i.e. there is no additional gas flow imposed.

Testing Scenario 2: The sensor was housed in a light tight chamber

(machined aluminium, circa 100 ml volume). In the experiment, a steady flow of dry air at a flow rate of 100 ml/min, from a pressurised cylinder, was passed over the sensor. The sensor is allowed to reach a steady state prior to testing. The test vapours are then injected into the input flow of dry air through a septum.

Testing Scenario 3: The sensor was housed in a light tight chamber (machined aluminium, circa 100 ml volume), air was drawn continuously through the chamber and across the sensor at a rate of 100 ml/min using a KNF-Neuberger micro diaphragm pump. During the experiments, small pure sealed samples of the target volatile compounds (explosives) were used. The samples were opened to expose the headspace and immediately the inlet tubing to the sensor system was placed into the headspace above the vial (distance of circa 5cm above the liquid/solid sample) and the headspace was sampled for 1 minute. The sensor system was then moved to clean air and allowed to equilibrate until the original baseline value (resistance and/or light emission) was restored. The system was then tested to the next sample using the same method.

Cataluminescence response of doped vs undoped zirconium oxide The doped zirconium dioxide materials were found to have some highly desirable properties as compared to the undoped zirconium dioxide materials, as follows.

Heated undoped zirconium dioxide sensors give a strong cataluminescence response, but in many cases the resistance of the film is too high to measure. The addition of europium enhances the cataluminescence signal for some volatile compounds whilst also reducing the film resistance into a measurable range. Thus the europium doped zirconium material offers the possibility of multimodal sensing (resistance and cataluminescence) which is more difficult for the undoped material.

The spectrum (wavelength dependence) of the cataluminescence is significantly changed when the rare earth is added to the zirconium oxide. The emission spectrum is found to have structure which is related to the sharp atomic emission lines from the rare earth atoms in the zirconium oxide matrix.

This is shown in Figure 6, for sensor operation at 350 °C, where the top spectrum is for cataluminescence from the undoped zirconium oxide in the presence of acetone as determined using Testing Scenario 1. The concentration of acetone was about 2 parts per thousand at the sensor. The spectral distribution of the cataluminescence is determined by sampling the generated cataluminescence, via a glass viewing portal, directly into an EG and G 1460 Optical Multichannel Analyser System.

The middle spectrum is for cataluminescence from the europium doped material (5% by mole fraction) in the presence of acetone at a concentration around 10 parts per thousand at the sensor. In contrast to the broad structureless spectrum for the undoped zirconia case, sharp peaks are present corresponding to emission from the europium dopant atoms.

The lower spectrum is for cataluminescence from the europium doped material for TATP of concentration around 2ppm at the sensor. Again, peaks correspond to emission from the europium dopant atoms by the bottom spectra.

One advantage for the europium doped material is that a narrower optical band pass filter can be used to select the cataluminescence. This will lead to an improvement in signal to noise as the background due to black body emission can be minimised. Multimodal responses of doped and undoped zirconium oxide sensor to various volatile compounds

The (i) undoped, (ii) europium doped, and (iii) terbium doped sensors were also tested, using Testing Scenario 2, to show the resistance change (figure 7) and cataluminescence change (figure 8) for a range of different potential target volatile compounds and potential interferents, which were air, water, hydrogen peroxide, 2- nitrotoluene (labelled 2NT/5 in the figures), TATP (triacetone triperoxide), EGDN (ethylene glycol dinitrate), ethanol and acetone.

Figure 7 shows the high sensitivity of europium doped sensor to TATP, and the differential response to EGDN, where the resistance increased rather than decreased. Figure 8 shows that the doped materials exhibit a better sensitivity and selectivity to TATP vs ethanol and acetone.

Multimodal response of zirconium oxide sensor to DMNB and hydrogen peroxide. The undoped zirconium oxide sensor produced as above was also tested, using Testing Scenario 2, with 2, 3 -dimethyl-2, 3 -dinitrobutane (DMNB) and hydrogen peroxide. The response of the zirconium oxide sensor to alternate input pulses of DMNB (l6ppb) and laboratory air is shown in figure 9. The sensor was activated by heating to and being maintained at 300 °C.

As can be seen in Figure 9, the cataluminescence response, i.e. increase in light output, (lower trace) is above the noise level for l6ppb DMNB. In contrast, no significant cataluminescence response is observed for the lab air input. The high background levels of light signal are due to residual black body thermal emission.

In contrast there are significant resistance decreases (upper traces) for both DMNB (l6ppb - 35%) and for lab air (-36%) challenges. The combination of the two sensor modalities (cataluminescence and resistance) may thus be used to differentiate between inputs of DMNB and lab air. This also shows that the cataluminescence response is much less reactive to water vapour than the resistance response.

In more detail, Figure 9 shows the response of the heated zirconium oxide sensor at 300 °C to alternate vapour inputs of l6ppb DMNB and laboratory air at the sensor. Upper trace: resistance of the sensor; lower trace: cataluminescence as measured by a photomultiplier. The sensor responds within a few seconds of inputting the vapour challenge to the system.

Multimodal response of the europium doped zirconium oxide sensor to NG, MEKP, and TATP

In a further experiment, the response of the europium doped zirconium oxide sensor to lab samples of explosive compounds was tested using Testing Scenario 3. The sensor is made and tested as set out above. The sensor was activated by heating to and being maintained at 300 °C. Figure 10 shows the cataluminescence response to NG (nitrogylcerin) on the left, MEKP (methyl ethyl ketone peroxide) in the middle and TATP (triacetone triperoxide) on the right. Figure 11 is similar but instead of cataluminescence, shows the resistance change to NG on the left, MEKP in the middle and TATP on the right. Figure 12 illustrates the ratio of light emission to resistance change for the three volatiles in the same order. The ratios are around 3000 for NG, 13700 for MEKP and 8200 for TATP. Having two values by using the dual modes as in the present invention allows ratios to be calculated, which greatly aids the characterisation of different compounds, and improves the selectivity and specificity of the sensors compared to prior art single mode sensors. Example 2 - Multimodal response of a zinc oxide sensor to hydrogen peroxide and ethanol

A zinc oxide sensor was made by the method described above, i.e. the oxide material (0.2 g) was placed in an agate pestle and mortar and sufficient water (ca. 0.4 g) was added to produce a thick slurry and the mixture was ground for 10-15 min to eliminate any air. The sensor was prepared by drop coating the oxide paste (5 microlitres, ca. 5mg) onto the substrate (2.5 x 2.5 mm² alumina tile) - so that it completely covered the interdigitated gold electrodes printed on the obverse side of the alumina tile. (The platinum heater is printed on the reverse side). This was then allowed to dry at room temperature ca. 22 °C for 12 h prior to use. The oxide material used in this case was Zinc (II) Oxide ZnO - nanopowder <l00nm particle size - sourced from Sigma- Aldrich.

The zinc sensor underwent testing according to Testing Scenario 2. The results are shown in figure 13 for the selective response of hydrogen peroxide headspace (a) as compared with ethanol (b). In more detail, Figure 13 shows the response for the multimodal zinc oxide sensor, activated by heating at 350 °C, when exposed to (a) hydrogen peroxide and (b) ethanol. In both cases, the upper trace is the resistance response and the lower trace is the cataluminescence response. As can be clearly seen, hydrogen peroxide gives an increase in resistance whereas ethanol gives a decrease in resistance. Both materials give a good cataluminescence response.

This multimodal sensor can thus differentiate between hydrogen peroxide headspace and ethanol, e.g. as found in the headspace of many fragrances.

Example 3 - Multimodal response of an yttrium oxide sensor to various target volatile compounds. An yttrium oxide sensor was made as per the method in Example 2, using yttrium (III) oxide Y2O3 - nanopowder <50 nm particle size - sourced from Sigma-Aldrich as the metal oxide material.

The sensor was then tested according to the Testing Scenario 2 with target volatile compounds which were 1 - 4NT (4-nitrotoluene), 2 - 2NT (2-nitrotoluene), 3 - H2O2 (hydrogen peroxide), 4 - NG (nitroglycerin), 5 - Chex (cyclohexanone) and 6 - NH3 (ammonia). The numbers 1 - 6 relate to Figures 14 and 15.

Figure 14 shows that all the volatile compounds tested gave a resistance and light emission response. Figure 15 gives the ratio of light emission to change in resistance and shows that the ratio of the two responses is different for each volatile compound.

Example 4 - Multimodal response of a tungsten oxide sensor to various target volatile compounds.

A tungsten oxide sensor was made as per the method in Example 2, using Tungsten (VI) oxide WO3 - nanopowder <l00nm particle size - sourced from Sigma-Aldrich as the oxide material .

The sensor was then tested according to the Testing Scenario 2 at 400 °C with target volatile compounds which were 4-NT, EGDN, H2O2, and interferents which were window and glass cleaner, Brasso, ethanol headspace and acetone headspace.

As shown in Figure 16, in this case the explosive target volatile compounds gave a change in resistance, interestingly with the change in resistance for EGDN being positive, but no cataluminescence response for 4-NT, EGDN. TATP and the interferents gave a response for both resistance and cataluminescence.

In a further experiment, the cataluminescence (Figure 17) and % resistance change (Figure 18) for the same sensor (tungsten oxide sensor operated at 400 °C) was measured, using Testing Scenario 3, for four target volatile compounds, NG, EGDN, MEKP and TATP, shown in order from left to right in Figures 17 and 18. As can be seen, a strong cataluminescence response is given for NG, MEKP and TATP, but none for EGDN. A resistance change is given for all the compounds, but interestingly the sign of the resistance change is positive for EGDN but negative for NG, MEKP and TATP. Figure 19 shows the ratio of cataluminescence to resistance change, which is around 70 for NG, 0 for EGDN (due to the lack of cataluminescence response), 164 for MEKP and 104 for TATP. Accordingly, as explained in theory, and demonstrated in the examples, the combination of resistance and cataluminescence monitoring opens up a new world of possibilities for enhancing the selectivity, and sensitivity of sensors for volatile compounds.

References BPJ de Lacy Costello, RJ Ewen, NM Ratcliffe, M Richards. Sensors and Actuators B. Highly sensitive room temperature sensors based on the ETV-LED activation of zinc oxide nanoparticles. Chemical 134 (2), 945-952 (2008)

M Sagar at al. Metal oxide semiconductor based thin film sensor for nitro aromatic explosive detection. International Conference on Convergence of Technology, p575 (2014).

K Brudzewski el al. Metal oxide sensor arrays for detection of explosives at sub-parts- per million concentration levels by differential electronic nose. Sensors and Actuators B, Volume 161, p 528-533 (2012).

Claims

1. A sensor for detecting volatile compounds, the sensor comprising:

a film of metal oxide;

a heater configured to provide heat energy to the film of metal oxide, and/or a UV light configured to irradiate the film of metal oxide;

an electrical conductivity meter configured to measure the electrical conductivity of the film of metal oxide; and

a photometer configured to measure cataluminescence derived from an interaction between a volatile compound and the film of metal oxide.

2. A sensor for detecting volatile compounds according to claim 1, wherein the film of metal oxide comprises nanoparticles of metal oxide.

3. A sensor for detecting volatile compounds according to claim 2, wherein the nanoparticles have average particle size of 20 to 100 nm, preferably 40 to 60 nm, most preferably about 50 nm.

4. A sensor for detecting volatile compounds according to any preceding claim, wherein the film of metal oxide is 0.001 to 0.5 mm thick, preferably 0.01 to 0.3 mm thick, most preferably about 0.2 mm thick.

5. A sensor for detecting volatile compounds according to any preceding claim, wherein the metal oxide comprises a dopant, preferably wherein the metal oxide comprises 1 to 10 % by mole fraction, most preferably 4 to 6 % by mole fraction dopant.

6. A sensor for detecting volatile compounds according to any preceding claim, wherein the metal oxide is selected from oxides of metals in the IUPAC Groups 3, 4, 5, 6, 8, 9, 10,11,12, 13, 14 and 15 of the Periodic Table, preferably the metal oxide is selected from one or more of the oxides of yttrium, aluminium, tin, indium, zinc, tungsten, chromium, titanium, nickel, cadmium, iron, niobium, manganese, gadolinium, vanadium, and zirconium, more preferably the metal oxide is selected from an oxide of yttrium, tungsten, zinc, and/or zirconium.

7. A sensor for detecting volatile compounds according to any preceding claim, wherein the metal oxide is zirconium dioxide doped with europium or terbium.

8. A sensor for detecting volatile compounds according to any preceding claim, wherein the sensor comprises a heater and not a UV light.

9. A sensor for detecting volatile compounds according to any of claims 1 to 7, wherein the sensor comprises a UV light and not a heater, preferably wherein there is a filter between the photometer and the sensor.

10. An array of sensors for detecting volatile compounds according to any preceding claim, wherein the array comprises 2 to 100 sensors, preferably 5 to 50 sensors.

11. A kit comprising a sensor or sensor array for detecting volatile compounds according to any preceding claim, and a gas chromatograph.

12. A method of detecting a volatile compound in a gaseous environment, the method comprising the steps of:

providing a sensor which includes a film of metal oxide;

activating the film of metal oxide, by heating it, and/or by irradiating it with UV light;

exposing the sensor to the gaseous environment; and

monitoring the electrical conductivity of the film of metal oxide and cataluminescence derived from the interaction between the volatile compound and the film of metal oxide.

13. The method of detecting a volatile compound in a gaseous environment according to claim 12, wherein the film of metal oxide is activated by heating to 100 to 500°C, preferably 200 to 400°C.

14. Use of a sensor to detect a volatile compound, the sensor comprising:

a film of metal oxide;

a photometer configured to measure cataluminescence derived from an interaction between the volatile compound and the film of metal oxide.

15. Use according to claim 14, wherein the volatile compound is an explosive or explosive precursor, preferably wherein the volatile compound is hydrogen peroxide, nitroglycerin (NG), ethylene glycol dinitrate (EGDN), 2, 3 -dimethyl-2, 3 -dinitrobutane (DMNB), 2,4 dinitrotoluene (2,4 DNT) and/or triacetone triperoxide (TATP).

16. Use according to claim 14, wherein the volatile compound is a biomarker in the breath or in the headspace above a body fluid, preferably wherein the volatile compound is acetone, acetaldehyde pentane, ammonia, hydrogen sulphide, hydrogen cyanide, nitric oxide, isoprene, hydrogen peroxide, methane and/or hydrogen.

17. Use according to claim 14, wherein the volatile compound is a marker for ripeness or decay in food or crops, preferably wherein the volatile compound is a Cl - C6 alcohol or a sulphide such as dimethyl disulphide, an aldehyde, a ketone, or ethylene.

18. Use according to claim 14, wherein the volatile compound is an environmental contaminant of the air or water, preferably wherein the volatile compound is methane, ozone, hydrogen sulphide, carbon monoxide, petroleum, paraffin jet fuel, hydrogen, ammonia, sulphur dioxide and/or NOx.