Hydrogen sensor and method for detecting hydrogen

Abstract

A device (100) for detecting hydrogen. The device (100) comprises a hydrogen sensing element (101) and a magnetic sensor (102). The hydrogen sensing element (101) comprises a porous carbonaceous material which has an intrinsic magnetic property that changes in response to the physisorption of hydrogen. The magnetic sensor (102) is configured to measure the change in the magnetic property of the hydrogen sensing element in response to hydrogen physisorption with the hydrogen sensing element. Also disclosed is a method of detecting hydrogen gas in a gas composition. The method comprises contacting the gas composition with a porous carbonaceous material that has an intrinsic magnetic property that changes in response to hydrogen gas physisorption, detecting a change in the magnetic property of the porous carbonaceous material resulting from the hydrogen physisorption, and inferring that hydrogen is present from the change in the magnetic property.

Description

[0001] HYDROGEN SENSOR AND METHOD FOR DETECTING HYDROGEN

[0002] FIELD OF THE INVENTION

[0003] The present invention relates to a hydrogen sensing method and a hydrogen sensing apparatus.

[0004] BACKGROUND

[0005] Hydrogen gas (H2) possesses certain characteristics that make it important to accurately detect and monitor its presence. Hydrogen is an odourless, colourless and highly flammable gas, which can easily leak due to its small size and rapid diffusion even through metallic materials. It has a wide flammable concentration range of 4-75% in air and an explosive concentration range of 18-59%. Therefore, the potential for dangerous hydrogen mixture levels in the air from a leak or accidental release is high. Furthermore, the minimum ignition energy for hydrogen is relatively low, at 0.012mJ, which is approximately 16.5 times lower than methane. This means that there is a risk of even electrostatic discharge causing ignition of H2 gas mixtures.

[0006] There are a wide variety of industrial applications where accurate detection and monitoring of hydrogen concentration are crucial. For example, there is safety legislation in place relating to leak detection of hydrogen in industrial processes that use hydrogen. Additionally, measurements of hydrogen concentration are also necessary for various industrial processes such as measuring feedstock levels in ammonia and methanol production, determining product purity in synthesis gas (syngas) production and determining contaminant levels in steel or aluminium production.

[0007] Hydrogen sensors typically include a hydrogen sensing element that exhibits a measurable change in a physical property upon exposure to hydrogen. These hydrogen sensing elements form part of a sensing device that utilises a transduction mechanism to produce a changing signal in response to the change in the physical property of the hydrogen sensing element. Examples of sensing mechanisms include thermal conductivity sensing, resistive sensing, work function sensing, electrochemical sensing, catalytic combustion sensing, mechanical and acoustic sensing, optical detection and magnetic detection. Current hydrogen sensing elements typically fall into two categories: platinum group hydrogen sensing elements (e.g. palladium or platinum); and metal oxide based hydrogen sensing elements (e.g. SnCU, ZnO, SnFeCU and WO3). These hydrogen sensing elements rely on chemisorption of hydrogen onto the surface of the hydrogen sensing element, which causes the hydrogen (H2) to split into atomic hydrogen and diffuse into the atomic lattice of the hydrogen sensing element. Whereby the chemisorption of hydrogen into the lattice of the sensing element causes the changes in physical property (as outlined above) to be measured with additional apparatus in the sensing device. This absorption process can cause mechanical damage to the hydrogen sensing element, such as blistering or delamination, and can also require high operating temperatures (>180 degrees C) for the interaction mechanism to occur and result in meaningful hydrogen detection.

[0008] One of the most recent hydrogen detection methods is magnetic -based hydrogen detection. Magnetic-based hydrogen detection allows for measurements to be made without direct electrical contact, lowering the risk of sparking and subsequent explosions. Examples of sensing mechanisms for magnetic -based sensors include magnetization sensors, ferromagnetic resonance sensors and Kerr effect sensors. However, existing magnetic-based hydrogen sensors utilise platinum group or metal oxide hydrogen sensing elements as outlined above, thereby exhibiting the same disadvantages.

[0009] A method and apparatus for hydrogen sensing is desirable that addresses or mitigates at least some of these problems.

[0010] SUMMARY

[0011] A first aspect of the invention is a device for detecting hydrogen comprising: a hydrogen sensing element comprising a porous carbonaceous material which has an intrinsic magnetic property that changes in response to hydrogen physisorption; and a magnetic sensor configured to measure the change in the magnetic property of the hydrogen sensing element in response to hydrogen physisorption with the hydrogen sensing element.

[0012] The magnetic sensor may be configured to measure the change in the magnetic property of the carbonaceous material. The sensing element may comprise a spin glass material.

[0013] The magnetic property may comprise dilute, disordered or ordered magnetism.

[0014] The device may further comprise a magnet proximal to the hydrogen sensing element, The hydrogen sensing element may be positioned between the magnet and the magnetic sensor. The magnetic sensor may be configured to measure magnetic flux from the magnet through the hydrogen sensing element.

[0015] The magnet may be a first magnet, and the device may further comprise a second magnet. The magnetic sensor may comprise a first probe configured to measure a magnetic flux from the first magnet through the hydrogen sensing element and a second probe configured to measure a magnetic flux from the second magnet that does not pass through the hydrogen sensing element.

[0016] The device may comprise a reference material between the second magnet and the second probe. The magnetic properties of the reference material may be independent of hydrogen adsorption of the reference material.

[0017] Any one of the hydrogen sensing element or the reference material is adhesively coupled to the respective first or second magnets.

[0018] At least one of the hydrogen sensing element or the reference material may be composed of a thin film deposited on the respective first or second magnets.

[0019] The magnetic sensor may comprise one or more hall effect sensors.

[0020] The device may further comprise a sealed housing. The magnetic sensor may be within the sealed housing. The hydrogen sensing element may be external to the sealed housing.

[0021] The porous carbonaceous magnetic material may be (or comprise at least 90%, 95%, 97%, 98%, or 99% by mass) carbon. The porous carbon of the hydrogen sensing element may comprise at least 20% sp3bonds, expressed as the number of sp3bonds divided by the sum of the number of sp2and sp3bonds.

[0022] At least 20% of the pores of the porous carbon of the hydrogen sensing element may have a pore diameter of less than 5 nm.

[0023] The porous carbonaceous material may be at least one of (e.g. all of, or at least two of): macroporous, mesoporous and microporous.

[0024] The porous carbonaceous material is preferably microporous.

[0025] The device may comprise a controller, arranged to determine a concentration of hydrogen from the amount of the change of the magnetic property of the hydrogen sensing element.

[0026] The device may comprise a display configured to indicate the determined hydrogen concentration.

[0027] The device may comprise an indicator configured to provide an alert when the determined hydrogen concentration exceeds a threshold hydrogen concentration.

[0028] According to a second aspect, there is provided a method of detecting hydrogen gas in a gas composition, comprising: contacting the gas composition with a porous carbonaceous material that has an intrinsic magnetic property that changes in response to hydrogen gas physisorption; detecting a change in the magnetic property of the porous carbonaceous material resulting from the hydrogen gas physisorption; inferring that hydrogen is present from the change in the magnetic property.

[0029] The method may further comprise estimating a concentration of hydrogen from the amount of the change in the magnetic property.

[0030] The method of the second aspect may comprise using the device of the first aspect, including any of the optional features thereof. The device of the first aspect may comprise any of the features of the second aspect. The device of the first aspect and the method of the second aspect may comprise any of the features described with reference to the example embodiments.

[0031] DETAILED DESCRIPTION

[0032] Embodiments of the invention will be described, purely by way of example, with reference to the accompanying drawings, in which:

[0033] Figure 1 is a schematic diagram of an example device for measuring hydrogen concentration;

[0034] Figure 2 is a schematic diagram of a further example device for measuring hydrogen concentration;

[0035] Figures 3 and 3b illustrate example spin orientations in a ferromagnetic material 300 and a carbonaceous material 400;

[0036] Figures 5a-b illustrate example graphs of molar magnetisation against applied magnetic field and molar magnetic susceptibility against temperature for a porous carbon sample in a vacuum;

[0037] Figures 6a-b illustrate example graphs of molar magnetisation against applied magnetic field and molar magnetic susceptibility against temperature for a porous carbon sample in a hydrogen atmosphere;

[0038] Figures 7a-7b illustrate example difference plots for molar magnetisation against applied magnetic field and molar magnetic susceptibility against temperature using the data from the plots of Figures 5a-b and 6a-b; and

[0039] Figures 8a-b illustrate example graphs of measured magnetic field against pressure for a hydrogen sensing device similar to a hydrogen sensing device of Figure 2.

[0040] Figure 9 is a schematic diagram of a further example device for measuring hydrogen concentration

[0041] Figure 1 is a schematic diagram of an example proof of principle device 100 for measuring hydrogen concentration. The device 100 comprises a hydrogen sensing element 101 comprising porous carbon; and a magnetic sensor 102 configured to measure a change in magnetic properties of the hydrogen sensing element 101 in response to hydrogen physisorbed on the hydrogen sensing element 101. The device 100 further comprises a magnet 103 proximal to the hydrogen sensing element 101, wherein the hydrogen sensing element 101 is positioned between the magnet 103 and the magnetic sensor 102. In the example of the device 100, the magnetic sensor 102 is configured to measure a change in magnetic flux from the magnet 103 though the hydrogen sensing element 101 in response to hydrogen physisorbed on the porous carbon of the hydrogen sensing element 101. This change in magnetic flux results from a change in the magnetic permeability (or magnetic susceptibility) of the porous carbon of the hydrogen sensing element 101. The magnet 103 increases the magnetic flux concentration through the hydrogen sensing element 101 so that when hydrogen is physisorbed on the hydrogen sensing element 101, the magnitude of the change in magnetic flux is greater, thereby increasing the sensitivity of the device 100. In principle, it may be possible to detect the change in magnetic permeability resulting from hydrogen physisorption without the presence of a magnet (for example, based on flux concentration of the ambient magnetic field of the Earth).

[0042] Magnetic susceptibility is generally denoted by , by Xm for molar magnetic susceptibility and Xv f°rvolume magnetic susceptibility. Magnetic susceptibility is the ratio of magnetization M (magnetic moment per unit volume) to the applied magnetic field intensity H. The volumetric magnetic susceptibility Xv is adimensionless property and is defined as:

[0043] M =XvH

[0044] The volume magnetic susceptibility is related to the magnetic permeability by:

[0045] / I = jUo(l + Zv)

[0046] The molar magnetic susceptibility Xm may be defined as:

[0047] M Xm ~ Xv P

[0048] Where M is molar mass (in kg / mol) and p is density (kg / m3). A mass magnetic susceptibility Xg may be defined as: The CGS system is frequently used in magnetics. The respective susceptibilities are multiplied by 4TT to give the corresponding SI quantities:

[0049] In vacuum, magnetic flux density B (measured in tesla in SI) is related to magnetic field H (measured in A / m in SI) by the vacuum permeability n0

[0050] Using porous carbon for the hydrogen sensing element 101 has a number of technical advantages over existing hydrogen sensors. Porous carbon is orders of magnitude cheaper than the raw materials used in platinum group hydrogen sensing elements and metal oxide based hydrogen sensing elements, decreasing the cost of the overall hydrogen sensing device 100. Porous carbon is also more readily available than the rare metals required in platinum group hydrogen sensing elements and metal oxide hydrogen sensing elements, meaning the hydrogen sensing device 100 can be produced in greater volume without being affected by supply chain issues. For example, the porous carbon of the hydrogen sensing element 101 could be produced from waste organic material.

[0051] The porous carbon of the hydrogen sensing element 101 employs a physisorption mechanism instead of chemisorption, improving the speed of detection and the longevity of the hydrogen sensing device 100. The chemisorption mechanism used in platinum group and metal oxide hydrogen sensing elements causes swelling and shrinking, ultimately resulting in blistering and delamination of hydrogen sensing elements over time, harming sensor stability and longevity. Furthermore, platinum group based sensor elements are vulnerable to poisoning by sulphur based compounds. For the porous carbon of the hydrogen sensing element 101, the chemical reaction pathway of sulphur with carbon is a much higher energy pathway than that of platinum group metals so the reaction is less likely to occur. Therefore, using a physisorption mechanism instead of chemisorption also decreases service intervals for the hydrogen sensing device 100. Furthermore, compared to platinum group hydrogen sensing elements and metal oxidebased hydrogen sensing elements, the hydrogen sensing element 101 has a better temperature range, being capable of measuring hydrogen concentration in temperatures ranging from 2 K to at least 250K and measurements are likely to be possible up to temperatures of at least 350K . Metal oxide based hydrogen sensing elements can only operate at temperature ranges of between 180 - 650 °C due to their redox reaction mechanism and they therefore also typically require a heating element. This increases the power consumption of devices with metal oxide based hydrogen sensing elements.

[0052] In order to demonstrate the transduction mechanism, hydrogen 105 was dosed into sensing area 104 using a dosing apparatus. The dosing apparatus was configured to dose and evacuate gases from the sensing area 104 at pressures ranging from 104Pa to 105Pa. Sensing area 104 may be considered a first atmospheric environment, contained by enclosure 106. The magnetic sensor 102 is positioned external to the enclosure 106 and is therefore positioned in a second atmospheric environment. Hydrogen sensing element

[0053] 101 and magnet 103 are positioned within enclosure 106. Having the magnetic sensor

[0054] 102 positioned external to the enclosure allows contactless signal transduction: no electronics (or electrical signals) are required in the sensing area 104. This may improve safety if, for example, the sensing area 104 is an explosive environment. This may be particularly important if the hydrogen sensing device 100, 200 is designed to comply with ATEX directives. Contactless signal transduction lowers the additional level of engineering required for designing an electrical device safe for explosive environments (such as those governed by ATEX directives) and reduces the need for complex electrical contact fabrication methods. For example, the electronics (comprising the magnetic sensor 102) may be provided in a sealed enclosure, and the hydrogen sensing element 101 arranged for contact with the gas environment (whether this is flammable, explosive or otherwise).

[0055] The porous carbon of the hydrogen sensing element 101 may be formed into a monolithic element or a thin film (e.g. a layer of material with a thickness ranging from 100 nm to 100 pm). The hydrogen sensing element may be mesoporous (i.e. containing pores with diameters between 2nm and 50nm) and / or microporous (having pores less than 2nm in diameter) and / or macroporous (having pores with diameters greater than 50nm). Pore size may be determined by gas porosimetry (e.g. for pore sizes ranging from <lnm to ~100nm) and / or by mercury intrusion porosimetry (e.g. for pore sies ranging from ~5nm to >lmm).

[0056] Typically, platinum group-based hydrogen sensing elements and metal oxide-based hydrogen sensing elements require hydrogen sensing elements in the form of wire or continuous thin films due to the need for complex electrical contacts, reducing their surface area. A hydrogen sensing element 101 in powder form or a porous element has a larger surface area than a wire or a thin film, increasing its sensitivity.

[0057] The magnet 103 is preferably a permanent magnet. An electromagnet may be used, but this may remove a number of potential advantages of magnetic sensing because it requires electrical power. To avoid these disadvantages, if the magnet 103 is an electromagnet, the magnet 103 may be positioned in a sealed housing (e.g. with the magnetic sensor 102) rather than in contact with the gas to be sensed thereby removing any electronics from contact with the sensing area 104.

[0058] The magnetic sensor 102 may be a hall effect sensor, optical magnetometer, Overhauser magnetometer, radio or microwave frequency excitation and pick up coils, SQUID or any other suitable magnetic sensing device. The magnetic sensor 102 will typically measure magnetic flux B (in tesla or gauss) or magnetic field H (in A / m or Oe). The magnetic susceptibility or magnetic permeability of the hydrogen sensing element 101 may be inferred from the output of the magnetic sensor, which is directly affected by hydrogen physisorption with the hydrogen sensing element 101.

[0059] Figure 2 is a schematic diagram of a further example device 200 for measuring hydrogen concentration. The device 200 comprises the hydrogen sensing element 101 and magnet 103 of the device 100 of Figure 1 , wherein the magnet 103 is now a first magnet 103. The device 200 further comprises a dual probe magnetic sensor 202, a second magnet 203 adjacent to the first magnet 103, wherein the dual probe magnetic sensor 202 is configured for differential sensing. A first probe of the magnetic sensor 202 measures a magnetic field that does not change with hydrogen concentration, and the second probe of the magnetic sensor 202 measures a magnetic field that is responsive to hydrogen physisorbed on the hydrogen sensing element 101. The device 200 comprises a reference material 201 proximal to the second magnet 203, wherein the magnetic properties of the reference material 201 are independent of hydrogen adsorption of the reference material 201. The magnetic dual probe magnetic sensor 202 is configured to measure the magnetic flux from the second magnet 203 through the reference material 201 along with the magnetic flux from the first magnet 103 through the hydrogen sensing element.

[0060] The reference material 201 may be composed of a polymer. In some embodiments the reference material may be omitted, and an air gap used instead. A differential measurement provides for common mode rejection of any changes in background magnetic field, improving the accuracy and reliability of the hydrogen sensing device 200 (see Figures 8a and 8b).

[0061] Sensing area 204 may be contained in pipe 205 (for example). The pipe 205 may form part of a larger industrial process where monitoring hydrogen concentration is necessary, such as for leak detection purposes or for measuring feedstock levels. The first environment of sensing area 204 may be classified as an ATEX environment. In some embodiments, the magnetic sensor 202 is housed in a sealed container along with any supporting electronics (e.g. a power source, readout circuit and display), while the sensing element 101 is exposed to the atmosphere in which hydrogen is to be detected.

[0062] In the example of sensing in pipe 205, the dual probe magnetic sensor 202 may be positioned external to the pipe 205. The hydrogen sensing element 101, first magnet 103, reference material 201 and second magnet 203 are positioned on the other side of the sealing wall of the pipe 205 and are therefore positioned within a different atmospheric environment.

[0063] The magnet 203 may be a permanent magnet or an electromagnet. If the magnet 203 is an electromagnet, the magnet 203 may also be positioned external to the pipe 205 in the first atmospheric environment, again removing the need for electronics in the sensing area 204. The hydrogen sensing element 101 or the reference material 201 may be adhesively coupled to the respective first or second magnets 103, 203. Alternatively, the hydrogen sensing element 101 or the reference material 201 may be composed of a thin film deposited on the respective first or second magnets 103, 203. At 0.1 bar (10 000 Pa) in an atmosphere of 100 % hydrogen, a change in mass susceptibility of +10pemu / molc-Oe (471* 1011m3.mol ') was measured for the device 200. A response time to complete a physisorption and desorption cycle of hydrogen from the hydrogen sensing element 101 was estimated at less than 2 seconds. This represents a significant improvement over palladium based hydrogen sensing elements which have a response time ranging between 2-60 seconds.

[0064] As mentioned above, in relation to Figures 1 and 2, the hydrogen sensing element 101 comprises porous carbon. Porous carbon is a type of activated carbon with a high specific surface area (surface area per unit mass) resulting from its high porosity. The porous carbon of the hydrogen sensing element 101 may be prepared by pyrolysis of a carbonaceous feedstock. This process is known as carbonisation. The carbonaceous feedstock may undergo low-temperature hydrothermal carbonization (180-280 °C) or high-temperature carbonization under an inert atmosphere (Nitrogen or Argon) at 300 - 900 °C to produce a carbonized-char material with some porosity.

[0065] The porosity of the carbon of the hydrogen sensing element 101 may be increased by a chemical or physical activation procedure, increasing the disorder of the material. The physical activation procedure may include physically activating the carbonised char material in a high-temperature steam or carbon dioxide atmosphere (800-1400 °C). The chemical activation procedure may include impregnating the carbonised char with bases or salts and heating them at high temperatures (800-1400 °C). Examples of bases or salts used in the chemical activation procedure include one of or a mixture of sodium hydroxide, zinc chloride, potassium hydroxide, potassium carbonate or urea . Alternatively, the steps of carbonisation and activation may be performed simultaneously by impregnating the carbonaceous feedstock with a chemical agent. The chemical agent may be, for example, zinc chloride, phosphoric acid or sulphuric acid.

[0066] The porous carbon of the hydrogen sensing element 101 may also undergo one or more post-processing treatments including further heating (1000-1800 °C) under vacuum or in an inert atmosphere, or washing in acids such as hydrochloric acid or phosphoric acid. Alternatively, the porous carbon of the sensing element 101 may also be produced by chemical vapour deposition of amorphous carbons via magnetron sputtering or pulsed laser deposition using carbon sources such as acetylene or methane.

[0067] The carbonaceous feedstock may be synthetic or organic waste. In the example hydrogen sensing element 101 used in proof of principle experiments, the porous carbon is prepared from a phenolic resin-based feedstock in a carbon dioxide atmosphere at a temperature of 1173 K. Similar activated carbon may be produced with other processes . For example, the porous carbon of the hydrogen sensing element 101 may also be prepared from other carbon-based feedstocks such as petroleum-pitch, coal or lignite. The porous carbon of the hydrogen sensing element 101 can also be derived from sustainable sources such as biowaste feedstock or a waste woven fabric feedstock. The biowaste feedstock may include but is not limited to: rice husk, flax, hemp, coconut, eucalyptus chippings and other lignin -predominant sources. The waste woven fabric feedstock may include but is not limited to: rayon, cotton and nylon.

[0068] At least 30% of the pores of the porous carbon of the hydrogen sensing element 101 may have a pore width of less than 2 nm. According to International Union of Pure and Applied Chemistry (IUPAC) classifications, a micropore corresponds to a pore width of less than 2 nm. Materials in which 30% of the pores have a pore width of less than 2nm are clearly microporous and may also be described as nanoporous (on the basis they have a significant fraction (e.g. >10%) of pores that are lOnm or smaller) . A high proportion of small pores in the material (e.g. >20% with pore width / diameter of less than 5nm) both increases surface area and increases the disorder of the carbon, contributing to its dilute magnetic properties.

[0069] The specific surface area of the porous carbon of the hydrogen sensing element 101 is approximately 1400 m2 / g (and is preferably at least 700 m2 / g, or at least 1000 m2 / g). The porous carbon used in the example hydrogen sensing element 101 was TE7 (obtained from Mast Carbon International), with an sp2: sp3bond ratio of 7:3. A relatively high proportion of sp3bonds may contribute to the dilute magnetic properties of the material. The porous carbon of the hydrogen sensing element 101 may comprise at least 20% sp3bonds, expressed as the number of sp3bonds divided by the sum of the number of sp2and sp3bonds. The Raman spectra ID / IG ratio of the porous carbon of the hydrogen sensing element 101 is approximately 1 (and is preferably greater than 0.9). This further indicates a high degree of disorder within the material.

[0070] Typically, graphitic carbon-based materials exhibit diamagnetic behaviour. However, a hydrogen sensing element such as the hydrogen sensing element 101 comprising porous carbon has a crystallographically disordered or amorphous structure. Magnetism can be induced in such structures by vacancies, adatoms, edge defects, graphite sheet curvature, the sp2: sp3bond ratio and functionalization of the edges. The amorphous nature of the porous carbon of the hydrogen sensing element 101 results in an intrinsic paramagnetic response in the hydrogen sensing element 101. A possible mechanism for this is the trapping of an sp2orbital between three sp3orbitals resulting in a localised unpaired electron (and thus a localised magnetic moment) in the porous carbon of the hydrogen sensing element 101. An additional mechanism may be via a mismatch in homogeneity of adjacent graphitic zigzag edges resulting in the formation of magnetic moments located at the edges of said graphitic planes, also known as spin polarised edge states. Both of the above mechanisms are the result of random defects in the structure of the porous carbon of the hydrogen sensing element 101. The number of defects (and therefore the number of localised magnetic moments) in the structure of the porous carbon of the hydrogen sensing element 101 is small compared to the total number of carbon atoms of the hydrogen sensing element 101 and randomly distributed within the porous carbon of the hydrogen sensing element 101. This results in localised, magnetic moments randomly dispersed within the structure of the porous carbon of the hydrogen sensing element 101. The hydrogen sensing element 101 can be said to be a magnetically dilute. Magnetic exchange interactions can exist between the dispersed magnetic moments in the porous carbon structure. At the microscale these may include ferromagnetic, ferrimagnetic or antiferromagnetic exchange interactions between spins localised on a graphitic plane, dependent upon local defects in the structure. Across the macroscale of the highly disordered microporous carbon structure, magnetic exchange interactions between planes, grains may also occur resulting in competing antiferromagnetic and ferromagnetic exchange interactions between the domains, which may result in magnetically disordered or spin glass behaviour.

[0071] A hydrogen sensing element 101 composed of pure graphite would not function in accordance with the present invention because of its highly diamagnetic nature. The porous carbon of the hydrogen sensing element 101 may be poisoned by a chemical reaction with a highly oxidising material within proximity to the magnetic sites / edges. Examples of highly oxidising materials include nitrates, bromine or hydrogen peroxide.

[0072] Figures 3 and 4 illustrate example spin orientations in a ferromagnetic material 300 and a carbonaceous material 400 (representative of the material of the hydrogen sensing element 101). The ferromagnetic material 300 comprises a plurality of ferromagnetic bonds 302a-c connected between a first plurality of atoms. Ferromagnetic bonds 302a- c tend to orient the spin of an atom of the first plurality of atoms in a first spin state 301a-c.

[0073] The carbonaceous material 400 comprises a plurality of ferromagnetic bonds 402a -b and antiferromagnetic bonds 403a connected between a second plurality of atoms. Ferromagnetic bonds 402a-b tend to orient the spin of an atom of the first plurality of atoms in a first spin state 401a-b. Antiferromagnetic bonds tend to orient the spin of an atom of the second plurality of atoms in a second spin state 404a-b. The combination of ferromagnetic bonds 402a-b and antiferromagnetic bonds 403a in the carbonaceous material 400 results in residual frustration in lower energy states: there is no spin configuration of the first spin state 401a-b and the second spin state 404a-b that satisfies all of the ferromagnetic bonds 402a-b and antiferromagnetic bonds 403a. This is illustrated by frustrated spin pair 405. Instead, a plurality of disordered spin configurations of the first spin state 401a-b and the second spin state 404a-b exist at equilibrium, resulting in disordered magnetic exchange interactions .

[0074] The porous carbon of the hydrogen sensing element 101 may be a carbonaceous material 400, as depicted in Figure 4. A portion of the randomly dispersed dilute magnetic moments of the porous carbon of the hydrogen sensing element 101 may couple together either ferromagnetically or antiferromagnetically, resulting in localised regions of ferromagnetic coupling (ferromagnetic bonds 402a-b) and antiferromagnetic coupling (antiferromagnetic bonds 403a). Therefore, the porous carbon of the hydrogen sensing element 101 in the example embodiment exhibits intrinsic dilute paramagentism with disordered magnetic exchange interactions .

[0075] When hydrogen (specifically ortho -hydrogen which possesses a nuclear magnetic moment) is physisorbed within the pores of the porous carbon of the hydrogen sensing element 101 , the hydrogen interacts with the dilute magnetic moments in the porous carbon. This may provide the required energy for spins such as frustrated spin pair 405 to transition from a first spin state 401b to a second spin state 404b (or vice versa), altering the overall magnetisation of the porous carbon of the hydrogen sensing element 101.

[0076] Figures 5a-b illustrate example graphs 510, 520 of molar magnetisation against applied magnetic field and molar magnetic susceptibility against temperature for a porous carbon sample according to an embodiment, suitable for use as the porous carbon of the hydrogen sensing element 101. The example porous carbon sample in the measurements described herein is a TE7 carbon, produced by carbonisation of phenolic resin at 1173 K in a carbon dioxide atmosphere. ICP-OES analysis of the carbon shows that negligible amounts of metal is present (parts per billion or less). In other embodiments some metal may be present, for example, up to 5% by mass. Magnetic properties of any included metal may interact with the magnetic properties of the carbon to modify the response of the magnetic property of the sensing element.

[0077] Measurements were taken in a vacuum. The magnetic response of the sample holder for the porous carbon sample has been removed as a background subtraction from graphs 510, 520. Magnetisation data for graphs 510, 520 were determined by DC SQUID magnetometer measurements and Levenberg-Marquet fitting.

[0078] Magnetisation traces 511a-g correspond with measurements taken at 10K, 30K, 35K, 60K, 77K, 100K and 200K respectively and exhibit a negative gradient for magnetisation against applied magnetic field. There is a small paramagnetic contribution visible at low temperatures in magnetisation trace 511a.

[0079] The traces shown in graph 520 correspond with field cooled with data collected on warming (FCW), zero field cooling and with magnetic field strengths of 250, 1000 and 5500 Oe (with 1 Oe = 103 / 4TT A.m '). The differences between FCW and ZFC curves are relatively small, so curves 521a, 521b, 521c correspond with the different field strengths of 250, 1000 and 5500 respectively. Magnetic susceptibility traces 521a-c exhibit a negative magnetic susceptibility across the entire temperature range , with an increasing magnetic susceptibility at low temperatures. A negative gradient for magnetisation against magnetic field and a negative magnetic susceptibility is consistent with the response of a chiefly diamagnetic material. The increase in magnetic susceptibility at low temperatures for graphs 510, 520 is indicative of an intrinsic dilute paramagnetic contribution of the material. Therefore, graphs 510, 520 suggest that that the porous carbon of the hydrogen sensing element 101 is predominantly non-magnetic in the absence of a hydrogen atmosphere.

[0080] Figures 6a-b illustrate example graphs 610, 620 of molar magnetisation against applied magnetic field and molar magnetic susceptibility against temperature for a porous carbon sample such as the porous carbon of the hydrogen sensing element 101. Measurements were taken in the presence of a hydrogen atmosphere. Similar to the graphs 510, 520 of Figures 5a-b, the magnetic response of the sample holder for the porous carbon sample removed as a background subtraction from graphs 610, 620, and magnetisation data for graphs 610, 620 was determined by DC SQUID magnetometer measurements and Levenberg-Marquet fitting.

[0081] Magnetisation traces 611a-h correspond with measurements taken at 10K, 15K, 30K, 35K, 60K, 77K, 100K and 200K respectively and exhibit an increase in gradient for magnetisation against applied magnetic field compared to magnetisation traces 511a-g. Traces measured at lower temperatures such as magnetisation traces 611a-b exhibit a positive gradient. Magnetic susceptibility traces 621a-c (again corresponding with field strengths of 250, 1000 and 5500 respectively) exhibit an increased magnetic susceptibility across the entire temperature range compared to magnetic susceptibility traces 521 a-c, with lower temperature measurements (at 10K and 15K) yielding a positive magnetic susceptibility and higher temperature measurements yielding a negative magnetic susceptibility. A positive gradient for magnetisation against magnetic field and a positive magnetic susceptibility is consistent with the response of a material in a paramagnetic or other magnetic state. Therefore, graphs 610, 620 illustrate that that upon physisorption of hydrogen within the pores of the porous carbon of the hydrogen sensing element 101 , a significant change in magnetisation against applied magnetic field and magnetic susceptibility against temperature is observed. This significant change is unexpected because hydrogen is diamagnetic and so should not cause an increase in the paramagnetic behaviour of the porous carbon of the hydrogen sensing element 101. Figures 7a-b illustrate difference plots 710, 720 for molar magnetisation against applied magnetic field and molar magnetic susceptibility against temperature for a porous carbon sample such as the porous carbon of the hydrogen sensing element 101. Difference plots 710, 720 are derived from the difference between magnetisation graphs 510, 610 and magnetic susceptibility graphs 520, 620 respectively. Difference plots 710, 720 illustrate the magnitude of the change in magnetisation against applied magnetic field and magnetic susceptibility against temperature upon physisorption of hydrogen within the pores of the porous carbon of the hydrogen sensing element 101. All magnetisation traces 711a-g (corresponding with 10K, 30 K, 35K, 60K, 77K, 100K and 200K respectively) exhibit a positive gradient and all magnetic susceptibility difference traces 721a-c exhibit a positive magnetic susceptibility difference.

[0082] Figures 8a and 8b illustrate example graphs 810, 820 for measured magnetic field against pressure for a proof of principle hydrogen sensing device similar to hydrogen sensing device 200. In the proof of principle device, a reference glass vial was used to provide a background measurement and a sensing glass vial includes a carbon hydrogen sensing element. The reference glass vial does not include a hydrogen sensing element, but includes a permanent magnet that is the same as that in the sensing glass vial. Identical Hall sensors were used to detect a magnetic field under each of the reference and sensing glass vials. Traces 811, 821 represent measurements in a 100% N2 atmosphere and traces 812, 822 represent measurements in a 100% H2 atmosphere. Traces 811, 812 represent a direct measurement of the change in magnetic properties of the hydrogen sensing element 101 (without subtraction of the measurement from the reference glass vial). In traces 821, 822, measurements from the reference glass vial are removed as a background subtraction from the respective traces 811, 812. Line 823 intersects with measurements made at atmospheric pressure. Comparing traces 821 and 822, a difference in magnetic field of approximately 20 mT is observed at atmospheric pressure between measurements taken with 100% N2 and 100% H2. It is expected that a response function can readily be determined for calibration of a hydrogen sensor employing this sort of sensing approach.

[0083] Figure 9 is a schematic diagram of a further example device 900 for measuring hydrogen concentration. The device 900 comprises the hydrogen sensing element 101, magnetic sensor (e.g. Hall sensor) 102 and magnet 103 of the device 100 of Figure 1. The magnetic field generated by the magnet 103 is represented by field lines 904a-d. Field lines 904a-d penetrate the hydrogen sensing element 101 and hall probe 102. The hall probe 102 is situated within the housing 905 whereas the sensing element 101 and magnet 103 are situated external to the housing 905. Spacers 902a-b are connected between the hydrogen sensing element 101 and the housing 905. This increases the exposed surface area of the hydrogen sensing element 101 , thereby increasing its speed of response.

[0084] The housing 905 further comprises a controller 906 connected to a display 903. The controller 906 is arranged to determine a concentration of hydrogen from the amount of change of the magnetic property of the hydrogen sensing element 101. The controller 906 may use pre-configured calibration data loaded onto its memory to determine hydrogen concentration. The display 903 is configured to indicate the determined hydrogen concentration. The housing 905 further comprises an indicator 901 configured to provide an alert when the determined hydrogen concentration exceeds a threshold hydrogen concentration.

[0085] The housing 905 forms an atmospheric seal between the external environment and the components situated within the housing 905. This may be particularly important if the hydrogen sensing device 900 is designed to comply with ATEX directives. The electrical components of the hydrogen sensing device 900 (i.e. the magnetic sensor 102, indicator 901, display 903 and controller 906) are provided in the sealed enclosure of the housing 905. The hydrogen sensing element 101 can then arranged for contact with the gas environment (whether this is flammable, explosive or otherwise).

[0086] Although specific examples have been described, the skilled person will appreciate that variations are possible, within the scope of the invention, which should be determined with reference to the accompanying claims.

Claims

CLAIMS1. A device for detecting hydrogen comprising: a hydrogen sensing element comprising a porous carbonaceous material which has an intrinsic magnetic property that changes in response to the physisorption of hydrogen; and a magnetic sensor configured to measure the change in the magnetic property of the hydrogen sensing element in response to hydrogen physisorption with the hydrogen sensing element.

2. The device of claim 1, further comprising a magnet proximal to the hydrogen sensing element, the hydrogen sensing element positioned between the magnet and the magnetic sensor, wherein the magnetic sensor is configured to measure magnetic flux from the magnet through the hydrogen sensing element.

3. The device of claim 2, wherein the magnet is a first magnet, the device further comprising a second magnet, wherein the magnetic sensor comprises a first probe configured to measure a magnetic flux from the first magnet through the hydrogen sensing element and a second probe configured to measure a magnetic flux from the second magnet that does not pass through the hydrogen sensing element.

4. The device of claim 3, the device further comprising a reference material between the second magnet and the second probe, wherein the magnetic properties of the reference material are independent of hydrogen adsorption of the reference material.

5. The device of any one of claim 2 to claim 4, wherein at least one of the hydrogen sensing element or the reference material is adhesively coupled to the respective first or second magnets.

6. The device of any one of claim 2 to claim 5, wherein any one of the hydrogen sensing element or the reference material is composed of a thin film deposited on the respective first or second magnets.

7. The device of any preceding claim, wherein the magnetic sensor comprises one or more hall effect sensors.

8. The device of any preceding claim, further comprising a sealed housing, wherein the magnetic sensor is within the sealed housing, and the hydrogen sensing element is external to the sealed housing.

9. The device of any preceding claim wherein the sensing element comprises at least 90% carbon.

10. The device of any preceding claim, wherein the porous carbonaceous material of the hydrogen sensing element comprises at least 20% sp3bonds, expressed as the number of sp3bonds divided by the sum of the number of sp2and sp3bonds.

11. The device of any preceding claim, wherein at least 20% of the pores of the porous carbonaceous material of the hydrogen sensing element have a pore diameter of less than 5 nm.

12. The device of any preceding claim, wherein the porous carbonaceous material is at least one of: macroporous, mesoporous and microporous.

13. The device of any preceding claim, wherein the porous carbonaceous material is microporous.

14. The device of any preceding claim, further comprising a controller, arranged to determine a concentration of hydrogen from the amount of the change of the magnetic property of the hydrogen sensing element.

15. The device of claim 14, further comprising a display configured to indicate the determined hydrogen concentration.

16. The device of any one of claim 14 or 15, further comprising an indicator configured to provide an alert when the determined hydrogen concentration exceeds a threshold hydrogen concentration.

17. A method of detecting hydrogen gas in a gas composition, comprising: contacting the gas composition with a porous carbonaceous material that has an intrinsic magnetic property that changes in response to hydrogen gas physisorption;detecting a change in the magnetic property of the porous carbonaceous material resulting from the hydrogen physisorption; inferring that hydrogen is present from the change in the magnetic property.

18. The method of claim 17, further comprising estimating a concentration of hydrogen from the amount of the change in the magnetic property.

Images