Galvanic sensor

Abstract

A galvanic lead-free oxygen sensor comprises a barrier to the passage of anode metal ions between the anode and the sensing cathode. This reduces the deposition of anode metal compounds on the sensing cathode, extending sensor lifetime. The barrier may comprise an ion-exchange membrane. The barrier may comprise an anode metal ion binding agent, for example a cellulosic material. The anode may comprise copper.

Description

[0001] GALVANIC SENSOR

[0002] Field of the invention

[0003] The present invention relates to the field of galvanic sensors, especially lead-free galvanic sensors.

[0004] Background to the invention

[0005] It is known to provide long-life galvanic sensors to measure the concentration of a gas, including but not limited to oxygen. Galvanic sensors of this type typically use a high surface area lead anode, for example a lead wool anode, which is gradually consumed, in proportion to analyte (oxygen) concentration.

[0006] Due to the toxicity of lead, there are strong environmental, health and regulatory pressures to provide sensors which do not contain lead. In particular, the European Union Restriction of Hazardous Substances (RoHS) regulations require homogenous materials to comprise less than 0.1 % by mass of lead. Amperometric lead-free oxygen sensors are known but they are not drop-in replacements for galvanic sensors, require continuous power, cannot work indefinitely in anaerobic environments and have performance limitations in high carbon monoxide or carbon dioxide environments. Attempts have been made to replace lead in galvanic sensors with an alternative metal, but lead has superior properties and attempts to avoid the use of lead have lead to lower quality sensors. In particular, such sensors have had a much lower lifetime charge density, pe(C / mm3), than sensors with lead anodes, leading to a greatly reduced sensitivity for a given size and lifetime.

[0007] By the lifetime charge density of the electrochemical cell we refer to the total amount of charge, Qe, that the sensor can provide during its lifetime divided by the sensor cavity volume (C / mm3).

[0008] Lifetime charge density can be determined as follows. Lifetime output, Qe(C) is the product of sensitivity which is the sensor current, typically in ambient conditions, iamb (A), and the usable lifetime of the sensor, T (S).

[0009] Qe -amb l-

[0010] Then the lifetime charge density is given by the quotient of the lifetime output, Qe, of the electrochemical cell, and the sensor cavity volume, Vcav(mm3). The sensor cavity contains the electrodes, electrolyte and other parts in contact with the electrolyte.

[0011] Lifetime charge density is fundamentally limited by the electrochemistry and choice of materials in the sensor. At the present time, reliable galvanic sensors having a lifetime charge density approximating that of lead anode galvanic sensors are not commercially available.

[0012] Thus, there is an unmet need for an essentially lead-free galvanic sensor with a lifetime charge density, pewhich is comparable to that of galvanic sensors with lead anodes.

[0013] In this context, aspects of the invention seek to provide an improved essentially lead- free galvanic sensor. Summary of the invention

[0014] According to a first aspect of the invention there is provided a galvanic sensor comprising: a housing; a sensor cavity within the housing; a sensing cathode and an anode (which is typically porous) within the sensor cavity.

[0015] It may be that a barrier to the passage of anode metal ions between the anode and sensing cathode is provided between the anode and the sensing cathode. It may be that fluid communication between the anode and the sensing cathode, is only through a barrier to the passage of anode metal ions.

[0016] We have found that the provision of a barrier to the passage of anode metal ions from the anode to the sensing cathode, such as an ion exchange membrane and / or a size exclusion membrane substantially improves the longevity of the sensor.

[0017] The oxidation reaction which occurs at the anode during the normal operation of the sensor produces metal oxide. Generally, research in the field of galvanic sensors with lead-free anodes has focused on addressing the build-up of passivating layers of oxide, or other oxidation products of the anode metal, on the anode, and upon the increase in anode volume over time which could arise from this process, which may cause sensor rupture. The latter is a known problem and various solutions have been proposed.

[0018] However, we have found that deposition of compounds of the one or more anode metals on the sensing cathode leads to sensor failure and limits sensor lifetime.

[0019] The compounds typically comprise anode metal oxides. However, the compounds depend on the electrolyte and anode and may for example be anode metal carbonates or anode metal hydroxides, for example in the case of an oxygen sensor using an alkaline metal hydroxide as electrolyte. The anode metal compounds (e.g. oxides) have further passivation effects, for example on the anode but also other surfaces within the sensor, where they may for example interfere with electrical conductivity.

[0020] Accordingly, by providing a barrier to anode metal ions, or otherwise restricting the deposition of compounds of the one or more anode metals on the sensing cathode, the sensor has a longer reliable life than would otherwise be the case. This allows the use of anode metals other than lead.

[0021] Thus, the barrier to the passage of anode metal ions restricts the deposition of compounds of the one or more anode metals on the sensing surface. Restricting the deposition of compounds of the one or more anode metals on the sensing cathode, for example by providing a barrier to anode metal ions, therefore also enables the anode to form a larger fraction of the internal volume of the sensor. This is because if the anode volume was increased and electrolyte volume correspondingly decreased, for a given sensor cavity volume within the sensor, the concentration of anode metal ions in the electrolyte would increase for a given amount of oxidised anode (and therefore a given amount of analyte detected), which would in turn increase the rate of anode metal compound deposition. Thus, restriction of anode metal compound deposition allows a greater anode volume to electrolyte volume ratio, and therefore a greater anode volume and higher lifetime charge density, pe. This enables a good lifetime charge density, pewithout the use of lead in the anode.

[0022] The barrier to the passage of anode metal ions may be in contact with the sensing cathode.

[0023] It may be that the barrier to the passage of anode metal ions comprises a metal ion binding agent.

[0024] It may be that the barrier to the passage of anode metal ions comprises an ionexchange membrane which is permeable to electrolyte but restricts the passage of anode metal ions. It may be that the barrier to the passage of anode metal ions comprises a size exclusion membrane which is permeable to electrolyte but restricts the passage of anode metal ions.

[0025] It may be that the barrier to the passage of anode metal ions comprises a membrane which is permeable to electrolyte and is both an ion-exchange membrane and a size exclusion membrane and which restricts the passage of anode metal ions.

[0026] Typically, the barrier to anode metal ions is also a barrier to zero-valance anode metal species. Typically, the barrier to anode metal ions is also a barrier to anode metal species which are not ions. If any such species are formed in the electrolyte in use, from oxidised anode metal, it is advantageous for the barrier to anode metal ions to also act as a barrier to these species. For example, a size-exclusion membrane will block any species containing anode metal atoms larger than a cut-off size.

[0027] It may be that the barrier to the passage of anode metal ions is a viscoelastic electrolyte gel having a viscosity of at least 500 mPa.s, or at least 5,000 mPa.s, which has the effect of inhibiting diffusion of anode metal ions. Suitable viscoelastic gels may comprise additives such as poly acrylic acid (PAA), poly ethylene oxide (PEO) and poly vinyl alcohol (PVA), or inorganic additives such as fume silica or bentonite clay. These are compatible with alkaline electrolytes.

[0028] It may be that the barrier to the passage of anode metal ions is in contact with the sensing cathode. The barrier may be formed on the sensing cathode or vice versa.

[0029] It may be that there is liquid electrolyte between the barrier to the passage of anode metal ions and the sensing cathode, and liquid electrolyte between the barrier to the passage of anode metal ions and the anode. However, there is preferably not any path for liquid electrolyte around the barrier, between the anode and the sensing cathode.

[0030] Typically, there is a three-phase contact region where solid cathode catalyst, liquid electrolyte and analyte gas contact each other. Typically, the cathode catalyst comprises hydrophobic gas conducting regions, for example hydrophobic binder particles, which provide a path for analyte gas to contact both solid cathode particles and liquid electrolyte. The barrier to the passage of anode metal ions typically restricts the passage of anode metal ions into this three-phase contact region.

[0031] Typically, the barrier to the passage of anode metal ions is water permeable, for example a water permeable ion exchange membrane.

[0032] It may be that the sensor cavity comprises a first compartment and a second compartment. It may be that the anode is located in said first compartment and the sensing cathode is located in said second compartment. It may be that fluid communication between said first compartment and said second compartment is only via one or more apertures between the first and second compartment. It may be that the one or more apertures define a frame around a barrier to the passage of anode metal ions. It may be that the one or more apertures are occluded by a barrier to the passage of anode metal ions.

[0033] It may be that the sensor cavity comprises one or more further electrolyte permeable layers. It may be that the first and / or second compartment comprise one or more further electrolyte permeable layers. One or more of the electrolyte permeable layers may be the barrier. Typically, the one or more electrolyte permeable layers are wetted layers. One or more of the electrolyte permeable layers may be a separator. The one or more electrolyte permeable layers are typically wetted in use (by the electrolyte). The first and / or second compartment are typically wetted by electrolyte.

[0034] It may be that the sensor cavity comprises a perimeter seal, extending around the periphery of the sensor cavity to block the passage (e.g. diffusion) of anode metal ions from the anode to the sensing cathode except through the barrier.

[0035] It may be that the perimeter seal comprises a shim, which is typically a conductive metal shim, which forms a seal against an internal wall of the sensor cavity and comprises one or more apertures, wherein the one or more apertures are occluded by the barrier to the passage of anode metal ions.

[0036] It may be that the shim is electrically conductive and wherein a first electrically conductive connection is provided between an electrical connector of the galvanic sensor and the shim and a second electrically conductive connection is provided between the shim and the sensing electrode.

[0037] It may be that the conductive electrical connection between the shim and the sensing electrode comprises a noble metal region which is pressed against the sensing electrode.

[0038] The noble metal region is typically welded to or coated onto the conductive electrical connection (which may for example be formed of a non-noble metal wire). The noble metal region is typically not welded to (or coated on) the sensing electrode. The noble metal region may be welded or coated to be formed as a layer on the surface of the non-noble metal wire. We have found that using a noble metal region pressed against the sensing electrode is easy to construct and provided a reliable connection in the presence of some levels of anode metal compounds. The noble metal region may be a tab of a noble metal.

[0039] It may be that the barrier is compressed between the shim and the sensing cathode.

[0040] It may be that the barrier is a (porous) sheet of a polysaccharide, for example cellulose.

[0041] As well as reducing damage to the sensing cathode by deposition of anode metal compounds, the presence of a barrier, particularly an ion exchange membrane, sizeexclusion membrane, or anode metal ion binder, can improve tolerance to interfering gases such as nitrous oxide, and potentially also carbon monoxide or carbon dioxide to which sensors can be exposed in high concentrations in some end-use applications, such as automatic applications or medical or veterinary applications (e.g. anaesthetics). High carbon monoxide or carbon dioxide exposure can lead to formation of carbonates with an electrolyte. These carbonates have limited solubility and so can lead to rapid precipitation of anode metal carbonates on the sensing cathode, which is mitigated by the present invention.

[0042] The galvanic sensor is typically an oxygen sensor. The galvanic sensor is typically an essentially lead-free sensor. Typically, the one or more anode metals are not lead. Typically, the anode comprises less than 0.5% lead by mass. The anode may be essentially lead-free. The anode is typically porous. The anode may comprise copper. The anode may be a copper anode. The anode may comprise at least 25% or at least 50% copper by mass. The anode may comprise at least 95% copper by mass. The anode may be at least 99% copper by mass. The anode may be substantially entirely copper.

[0043] The anode may be bronze. The anode may comprise copper and tin, for example at least 25% copper and at least 5% tin by mass. It may be that at least 95% of the anode by mass ,or at least 99% of the anode by mass is formed of copper and tin.

[0044] The anode may comprise copper and nickel, for example at least 25% copper and at least 15% nickel by mass. It may be that at least 95% of the anode by mass, or at least 99% of the anode by mass is formed of copper and nickel.

[0045] In some embodiments, the anode may comprise molybdenum. The anode may comprise at least 25% or at least 50% molybdenum by mass. The anode may comprise at least 95% molybdenum by mass. The anode may be at least 99% molybdenum by mass. The anode may be substantially entirely molybdenum.

[0046] It may be that the sensor comprises an electrical interface which is electrically connected to the anode by a conductor, and the conductor is fused to the anode by one or more of resistance welding, heat welding, chemical bonding, or conductive adhesive. This arrangement retains electrical conductivity as anode metal oxidises at the interface or as compounds deposit on the anode and conductor over time.

[0047] The sensing cathode is fluidically connected to the anode. The sensing cathode is electrically connected to the anode through a measurement circuit. The porous separator is part of the fluidic connection between sensing cathode and anode.

[0048] The anode is typically porous. The porosity increases the available surface area of the anode. The anode may be a wool or skein (of the anode metal). The anode may be a foam (of the anode metal). The anode may be sintered. The anode may be formed of sintered powder. It may be that the anode has a porosity in the range of 10% to 80%, or 20% to 60%, preferably 30 to 40%. By the porosity we refer to the fraction of the volume of the anode which is void. It can be determined from the dimensions of the anode, the mass of the anode and the known density of the anode metal.

[0049] It may be that the anode is formed of powder comprising particles having a mean diameter of less than 150 microns, or less than 125microns or less than 100 microns. The powder comprise particles sieved through a sieve pore size of no more than 150 microns, no more than 125 microns or no more than 100 microns, for example.

[0050] By the anode we refer to material which functions as anode in use (and so is oxidised over time due to the anode reaction) and not to parts which do not function as anode, for example, a support on which anode metal is supported, electrical connectors etc. Thus, the anode is formed of the material which is electrochemically active as an anode.

[0051] It may be that the housing of the galvanic sensor comprises a mass flow restriction member, typically a capillary, for regulating the diffusion of analyte to the sensing electrode. Typically, the reaction of the analyte at the sensing electrode is diffusionlimited. Thus, typically the flux of analyte gas (e.g. oxygen) into the sensor is limited by gas phase diffusion. In some embodiments, the mass flow restriction member is a membrane. In this case, typically the flux of analyte gas (e.g. oxygen) into the sensor is limited by the flux of analyte gas through the membrane.

[0052] Typically, the lifetime charge density, peof the electrochemical cell, comprising the anode, sensing cathode and electrolyte, is at least 0.15C / mm3or at least 0.25C / mm3or at least 0.5C / mm3.

[0053] As described above, it may be that the galvanic sensor is an oxygen sensor. In an oxygen sensor, oxygen molecules in gas contacting the sensing electrode undergo reduction to form hydroxide ions: O2 + 2H2O + 4e_40FT. In the case of an acidic electrolyte the reaction is O2 + 4H++ 4e_2H2O

[0054] It may be that the housing comprises a capillary for regulating the diffusion of analyte to the sensing electrode. The sensing cathode may for example be formed of gold, platinum, silver or a platinum group metal.

[0055] As described above, it may be that the barrier to the passage of anode metal ions comprises a metal ion binding agent. The presence of a metal ion binding agent is in itself useful to restrict the deposition of anode metal compounds on the sensing cathode. Accordingly, the invention extends in a second aspect to a galvanic sensor comprising a housing, the housing containing: a sensing cathode; an anode comprising one or more anode metals; an electrolyte; and a metal ion binding agent in contact with the electrolyte to bind ions of the one or more anode metals and thereby restrict the deposition of compounds of the one or more anode metals on the sensing cathode.

[0056] The metal ion binding agent removes anode metal ions from the electrolyte and / or otherwise restricts anode metal ions from reaching the sensing cathode (for example by restricting their diffusion to the sensing cathode), and thereby restricts the deposition of compounds of the one or more anode metals on the sensing cathode.

[0057] It may be that the metal ion binding agent binds with metal ions to form solid phase anode metal compounds (e.g. anode metal oxides), so that passivation of the anode or the sensing cathode by such compounds is minimised. The solid phase anode metal compounds may be formed in a binding layer, which may be in a wetted layer and / or separator. It may be that the metal ion binding agent binds with metal ions to form solid phase anode metal compounds (e.g. anode metal oxides) at a location minimising disruption to the electrical connections from electrodes to pins within the sensor. This sequesters the anode metal compounds without affecting the electrical operation of the sensor.

[0058] It may be that the metal ion binding agent also binds zero-valance anode metal species. It may be that the metal ion binding agent binds anode metal species which are not ions. If any such species are formed in the electrolyte from oxidised anode metal, it is advantageous for the metal ion binding agent to also bind these. Thus, typically the metal ion binding agent is an anode metal binding agent. The metal ion binding agent may be selected in dependence on the anode metal. It may be that the anode metal comprises copper and the metal ion binding agent binds with copper ions to form solid phase copper compounds. The metal ion binding agent may be a copper binding agent. As described above the anode may be a copper anode. The anode may comprise at least 25% or at least 50% copper by mass. The anode may comprise at least 95% copper by mass. The anode may be at least 99% copper by mass. The anode may be substantially entirely copper. The anode may be porous.

[0059] There are a number of different types of metal ion binding agent which are useful for this purpose.

[0060] The metal ion binding agent may be a metal ion immobilising agent. The metal ion immobilising agent may for example be a polysaccharide, for example a wetted layer comprising or formed of a polysaccharide. A polysaccharide or derivative of a polysaccharide may function as the metal ion immobilising agent.

[0061] The metal ion binding agent may be a metal ion complexing agent. The metal ion complexing agent may be a complex compound, or ligand. The complexing agent may be a saccharide, such as glucose, or a polysaccharide. A polysaccharide may function as the metal ion complexing agent.

[0062] The metal ion binding agent may be a gelling compound which forms metal compounds (e.g. metal oxides) containing gels with anode metal ions. The gelling compound may for example be carboxyalkyl cellulose or hydroxyalkyl cellulose. A polysaccharide may function as the gelling compound.

[0063] It may be that the metal ion binding agent is a polysaccharide. It may be that the metal ion binding agent is a derivative of a polysaccharide. It may be that the metal ion binding agent is cellulosic. It may be that the metal ion binding agent is cellulose or a derivative thereof.

[0064] The polysaccharide may be a structural polysaccharide, for example cellulose, hemicellulose, ethyl cellulose, chitin or amylopectin. The polysaccharide may be cellulosic. The polysaccharide may be a derivative of cellulose, for example, an alkyl derivative of cellulose, for example ethyl cellulose, a hydroxy alkyl derivative such as hydroxy ethyl cellulose, hydroxypropyl methylcellulose, a disaccharide such as sucrose, a carboxy alkyl derivative of cellulose, such as carboxy methyl cellulose. The polysaccharide may be a derivative of amylopectin, for example an alkyl, hydroxy alkyl, or carboxylate derivative of amylopectin.

[0065] We have found that polysaccharides, such as dextran and particularly structural polysaccharides, such as cellulose, and derivatives thereof, are highly effective at binding the ions of metals other than lead which are useful as anode metals. For example, with a copper-containing anode (e.g. copper or a copper alloy), polysaccharides effectively bind and concentrate copper.

[0066] It may be that the binding of anode metal ions to the metal ion binding agent (e.g. polysaccharide) leads to complexation of the metal ions.

[0067] It may be that the binding of anode metal ions to the metal ion binding agent (e.g. polysaccharide) leads to the formation of a gel comprising the metal ions.

[0068] It may be that the binding of anode metal ions to the metal ion binding agent (e.g. polysaccharide) leads to the immobilisation of the metal ions within the polysaccharide.

[0069] Typically, the housing defines a sensor cavity therein, the sensor cavity comprising the anode, sensing cathode, one or more wetted layers (e.g. of electrolyte permeable material), and the electrolyte.

[0070] By wetted we referred to a layer being permeated with electrolyte. It may be that the one or more wetted layers are damp with electrolyte. It may be that the one or more wetted layers are drenched with electrolyte. The electrolyte is typically aqueous. Typically, the one or more wetted layers are hydrophilic. The one or more wetted layers may be formed from one or more of glass fibre, polyamide, polyester or a polyolefin or other polymer known in the art which may be treated to increase hydrophilicity. The one or more wetted layers are typically porous. At least one wetted layer (optionally a plurality of wetted layers, for example all of the wetted layers) may comprise or be formed of the metal ion binding agent.

[0071] It may be that the amount of metal ion binding agent is at least 5mg per cm3of volume of the sensor cavity, at least 12mg per cm3of volume of the sensor cavity or at least 23 mg per cm3of volume of the sensor cavity.

[0072] It may be that the amount of metal ion binding agent is at least 5mg per ml of electrolyte, at least 12mg per ml of electrolyte or at least 23 mg per ml of electrolyte or at least 75 mg per ml of electrolyte.

[0073] It may be that the amount of metal ion binding agent is at least 1 mg per gram of anode metal, at least 3mg per gram of anode metal, or at least 6mg, per gram of anode metal.

[0074] It may be that the ratio of the mass of anode metal to the volume of the electrolyte is at least 2g of anode metal per 1 ml of electrolyte, at least 5g of anode metal per 1ml of electrolyte or at least 10g of anode metal per 1 ml of electrolyte.

[0075] The electrolyte may be an alkaline electrolyte. It may be that electrolyte is an alkaline metal hydroxide, for example 10 - 50 wt% of an alkaline metal hydroxide. The alkaline metal hydroxide may comprise one or more of lithium hydroxide, sodium hydroxide, potassium hydroxide, rubidium hydroxide and cesium hydroxide. It may be that the electrolyte is a quaternary ammonium hydroxide, for example 10 - 50 wt% quaternary ammonium hydroxide. Alternatively, the electrolyte may be an acid such as phosphoric acid, sulfuric acid, acetic acid, oxalic acid, citric acid and others.

[0076] Polysaccharides, such as cellulose, have very low solubility in alkaline electrolytes. In an alkaline electrolyte, electrolyte gradually permeates polysaccharides, such as cellulose, entering between strands of polysaccharides and thereby becoming immobilised. Polysaccharides can also chemically complex with anode metal ions, for example copper, and in time the polysaccharide (for example a separator of polysaccharide) will gradually break down releasing carboxylic acid derivative ligands and complexing agents. These mechanisms can work together to bind anode metal ions and thereby limit the build-up of anode metal compounds on the sensing cathode. Typically, the anode does not include significant quantities of lead. Typically, the anode has less than 0.5% lead by mass and preferably less than 0.1% lead by mass. Typically, the sensor is formed from less than 0.1 % lead by mass. Typically, each component of the sensor comprises less than 0.1% lead by mass. These are limits, if any lead is present. Thus, the sensor is typically essentially lead-free.

[0077] As used herein, essentially lead-free refers to having no significant amount of lead. Nevertheless, it will be understood in the industry that trace amount of lead can be present within guidelines. An essentially lead-free sensor may have an anode which comprises less than 0.5% of lead by mass, typically less than 0.1% of lead by mass. Similarly, an essentially lead-free sensor may have a sensing cathode which comprises less than 0.5% of lead by mass, typically less than 0.1% of lead by mass. The essentially lead-free sensor may as a whole comprise less than 0.5% of lead by mass, typically less than 0.1 % of lead by mass. The essentially lead-free sensor may have less than 0.5% of lead by mass, typically less than 0.1% of lead by mass in each homogenous component. Again, these are limits if any lead is present at all.

[0078] Nevertheless, although the invention is of especial benefit in enabling lead-free sensors, the invention may also be useful with a lead anode, where it leads to some reduction in lead use through extending lead anode sensor lifetime and so it may be that the anode is a lead anode.

[0079] It may be that the amount of polysaccharide (e.g. cellulose) is one or more of: (i) at least 5mg per cm3of volume of the sensor cavity or at least 12mg per cm3of volume of the sensor cavity or at least 23mg per cm3of volume of the sensor cavity; (ii) at least 5mg per ml of electrolyte, at least 12mg per ml of electrolyte, or at least 23mg per ml of electrolyte; and (iii) at least 1mg per gram of anode, at least 3mg per gram of anode or at least 6mg per gram of anode.

[0080] The metal ion binding agent (e.g. the polysaccharide) may be formed as a coating, for example on the sensing cathode or another sensor component.

[0081] The metal ion binding agent (e.g. the polysaccharide) may be dissolved in the electrolyte. The metal ion binding agent (e.g. the polysaccharide) may be formed as solid particles in the electrolyte.

[0082] The metal ion binding agent (e.g. the polysaccharide) may be formed as a layer, for example between the anode and the sensing cathode.

[0083] It may be that the sensor comprises one or more separators, which are permeated by electrolyte and separate the anode from the sensing electrode and which comprise the polysaccharide. It may be that the one or more wetted layers are separators.

[0084] Typically, the separator is electrically insulating. Typically, the separator is permeable by the electrolyte. The separator may be porous to electrolyte. Typically the separator is a sheet of (typically porous) material. Typically, the porous material is or is coated with the metal ion binding agent. For example, one or more separators may be or be coated with a polysaccharide.

[0085] It may be that the only path for electrolyte to diffuse between the anode and the sensing cathode is through one or more separators. The sensing cathode and one or more separators may be formed as a stack of layers. The stack of layers may be sealed around their periphery so that anode metal ions cannot diffuse to the sensing cathode without passing through the one or more separators.

[0086] At least one separator (e.g. wetted layer) may be made from a material such as a polysaccharide, such as cellulose, or a polysulfone, such as polyethersulfone, commonly referred to as PES, or Zirfon (a polyphenylene sulfide).

[0087] The one or more separators (e.g. wetted layers) may comprise one or more separators that do not comprise metal ion binding agent.

[0088] The one or more separators (e.g. wetted layer) may comprise one or more separators which are electrolyte permeable but which restrict (e.g. block) the passage of anode metal ions. For example, the one or more separators may comprise an ion exchange membrane selected to restrict (e.g. block) the passage of anode metal ions. The one or more separators may comprise a size-exclusion membrane selected to restrict (e.g. block) the passage of anode metal ions. Ions within the electrolyte may travel through the electrolyte between sensing cathode and anode. The ions may travel through the at least one separator, which is typically wetted by electrolyte.

[0089] The separator (e.g. wetted layer) impedes the passage of metal ions. The separator thereby prevents metal ions which may originate from the anode, for example an oxidised metal produced during normal operation of the sensor, from travelling to the sensing cathode. Deposition or precipitation of anode metal compounds (e.g. oxides) on the sensing cathode is thereby reduced.

[0090] The sensor may comprise a plurality of separators, formed as layers, such that the anode and sensing cathode are separated by a plurality of layers of separator (e.g. wetted layers). This provides multiple barriers to the passage of metal ions from the anode to the sensing cathode.

[0091] It may be that the anode is copper or a copper alloy comprising at least 25% copper by mass, or at least 50% copper by mass, or at least 75% copper by mass, or at least 90% copper by mass, or at least 95% copper by mass, or at least 99% copper by mass. It may be that the metal ion binding agent binds ions of copper to thereby restrict the deposition of compounds of copper (for example, copper oxides) on the sensing cathode.

[0092] The sensor may comprise a vent through which gas may enter into or egress from the sensor cavity.

[0093] The housing may comprise an inlet in a first (upper) external surface. The first external surface may form a first end of the sensor. The first external surface may be part of the second sensor portion. The housing may comprise a second external surface, opposite the first external surface. The second external surface may form a second end of the sensor. The second external surface may be part of the first sensor portion. The second external surface may be a base. The second external surface typically comprises conductors for electrical communication between the sensor and a circuit board. The housing may comprise one or more third external surfaces, between the first and second external surfaces. The one or more third external surfaces may be side surfaces of the sensor. The housing may be cylindrical (in the form of a cylinder), with the first and second external surfaces being opposite circular surfaces and the third external surface being the curved surface of the cylinder, between the first and second ends.

[0094] The vent may extend through the third external surface. The vent may extend into the anode compartment. The vent may extend into the first compartment (described below). The vent may extent through the wall of the first sensor portion. The vent may extend to a side of the anode.

[0095] Because the vent extends to a side of the anode, and not to its base, the vent is less likely to be obscured by electrolyte than if it were located on the base of the sensor, underneath the anode. It therefore provides the role of allowing gas pressure equalisation, for example when there is a change in temperature, without forcing gas through the inlet and sensing cathode where it would distort the sensor reading, and it does so more reliably than if the vent was located on the base of the sensor, underneath the anode.

[0096] The sensor cavity typically comprises a gas permeable membrane which covers the vent opening into the sensor cavity. The membrane is typically impermeable to liquid electrolyte, although it may be permeable to water vapour. This allows gas egress and ingress while restricting the loss of electrolyte. It may be that the third external surface is curved and the vent extends from the curved third external surface to a vent opening in a flat section of the inner wall of the sensing cavity to which the membrane is attached, covering the vent opening. This facilitates a liquid tight seal between the membrane and the inner wall of the sensor cavity, across the vent opening, despite the curved exterior.

[0097] The invention extends in a third aspect to a galvanic oxygen sensor comprising: a housing, a sensor cavity within the housing; a sensing cathode and an anode within the sensor cavity; and a vent through which gas may enter into or egress from the sensor cavity, wherein the housing comprises an inlet in a first external surface, the housing further comprising a second external surface, opposite the first surface, and at least one third external surface, wherein the vent extends through the third external surface to a side of the anode. The galvanic oxygen sensor of the third aspect may be a galvanic oxygen sensor according to the first aspect. The galvanic oxygen sensor of the third aspect may be a galvanic oxygen sensor according to the second aspect.

[0098] The invention extends in a fourth aspect to a galvanic oxygen sensor comprising a housing, the housing containing: a sensing cathode; an anode comprising one or more anode metals, the anode comprising at least 25% molybdenum by mass; and an electrolyte.

[0099] It may be that the anode is porous. It may be that the anode comprises less than 0.5% lead by mass.

[0100] It may be that the anode is molybdenum or a molybdenum alloy comprising at least 25% molybdenum by mass, or at least 50% molybdenum by mass, or at least 75% molybdenum by mass, or at least 90% molybdenum by mass, or at least 95% molybdenum by mass, or at least 99% molybdenum by mass. It may be that the metal ion binding agent binds ions of molybdenum to thereby restrict the deposition of compounds of molybdenum on the sensing cathode.

[0101] It may be that a barrier to the passage of anode metal ions between the anode and the sensing cathode is provided between the anode and the sensing cathode. It may be that the housing contains a metal ion binding agent in contact with the electrolyte to bind ions of the one or more anode metals and thereby restrict the deposition of compounds of the one or more anode metals on the sensing surface.

[0102] The galvanic oxygen sensor of the fourth aspect may be a galvanic oxygen sensor according to the first aspect. The galvanic oxygen sensor of the fourth aspect may be a galvanic oxygen sensor according to the second aspect. The galvanic oxygen sensor of the fourth aspect may be a galvanic oxygen sensor according to the third aspect.

[0103] The invention extends in a fifth aspect to a method of forming a galvanic sensor according to the first, second, third and / or fourth aspect, comprising the step of forming a first sensor portion comprising a portion of the housing, with the anode therein, and a second sensor portion comprising a portion of the housing, with the sensing electrode therein, fitting the perimeter seal to the second sensor portion and then attaching the second sensor portion to the first sensor portion.

[0104] When forming the second sensor portion, the method comprising inserting into the portion of the housing separator layers, typically polysaccharide (e.g. cellulose) separator layers, and a conducting shim strip which wraps around the separator layers thereby compressing the separator layers and further providing an electrical path between the sensing cathode on one side of the separator layers and the other side of the separator layers. The conducting shim strip is typically electrically attached to a shim prior to insertion, whereby the shim is conducting and acts as the perimeter seal. Once the sensor is assembled, typically the separator layers are wetted by electrolyte and become wetted layers. One or more of the separator layers may function as the barrier to the passage of anode metal ions.

[0105] Features described in relation to any aspect of the invention are optional features of each aspect of the invention. In particular, features described above in connection with the first aspect are also optional features of the second aspect, and vice versa.

[0106] Description of the Drawings

[0107] An example embodiment of the present invention will now be illustrated with reference to the following Figures in which:

[0108] Figure 1 (a) is a first embodiment of a sensor of the present disclosure; Figure 1 (b) is a schematic illustration of the shim separating the cathode compartment and the anode compartment; Figure 1(c) is an expanded view of the inner layered structure of an alternative embodiment of the sensor of the present disclosure; Figure 1(d) illustrates the anode and cathode compartments within the sensor.

[0109] Figure 2 shows a cellulose layer from a dismantled oxygen sensor after exposure to ambient atmosphere for 29 days, with 1 ,4mA current, by when 3500 C of anode had been used. Figure 3 shows response functions from capillary-type sensors according to the present disclosure. Each of Figure 3(a) and Figure 3(b) shows the output response of three fresh sensors to a pulse of oxygen. Figure 3(c) shows the aging over two months of a sensor output. Figures 3(d) and (e) show the effect of the quantity of cellulose in a sensor on the T90 response over 150 days. The x-axis is time in days and the y-axis is current in pA (3(a), 3(b), 3(c)) and T90 response in seconds (3(d), 3(e)).

[0110] Figure 4 is a cross-sectional view of sensor identifying where electrical connections within the sensor structure are welded.

[0111] Figure 5 is a series of graphs of sensor output showing the effects of welding internal electrical connections. The x-axis is time in days and the y-axis is current in pA. Multiple experiments are shown with separate lines.

[0112] Figure 6 is an expanded view of the inner layered structure of an alternative embodiment of the sensor of the present disclosure.

[0113] Figure 7 is a further embodiment of the sensor of the present disclosure in (a) cross-sectional form; and (b) in an expanded view illustrating the inner layered structure.

[0114] Figure 8 is an embodiment of a galvanic gas sensor in (a) cross-sectional and (b) in expanded view.

[0115] Figure 9 shows the evolution of the T90 response (y-axis) of the galvanic sensor over time (x-axis, days) for multiple repeats of each of 9(a) a copper anode, 9(b) a copper-nickel anode, 9(c) a molybdenum anode and 9(d) a lead anode.

[0116] Figure 10 is a further embodiment of a sensor with an ion exchange membrane in cross-section.

[0117] Figure 11 is a further embodiment of a sensor with an ion exchange membrane in expanded view.

[0118] Figure 12 shows first and second views of the inside of the sensor housing. Figure 13 shows a series of experimental results with different anode compositions. The x-axis is time in days and the y-axis is T90 response in seconds.

[0119] Figure 14 shows the response of the sensor to a change in temperature and the effect of different vent positions. The x-axis is time in hours and the y-axis is output in pA.

[0120] Figure 15 shows the variation in time (x-axis, days) with accumulated charge (left hand y-axis, C) and the output (right hand axis, %) of sensors according to the present disclosure with (a) an anode made of 150pm particle size, and (b) an anode made of 100pm particle size.

[0121] Figure 16 shows a porous anode made from sintered molybdenum (a) unmagnified, and (b) at a magnification factor of 210.

[0122] Figure 17 shows the output response of 56 day old galvanic sensors having a porous anode made of sintered molybdenum to a pulse of oxygen, with (a) 5 Viledon polyamide wetting layers; (b) 5 Viledon polyamide wetting layers and a layer of FAS- 30 ion-exchange membrane and (c) 5 layers of cellulose. The x-axis is time in seconds and the y-axis is output current in pA.

[0123] Figure 18 shows output performance over time plots for galvanic sensor with a porous anode made of sintered molybdenum. As with Figure 17 for (a) 5 Viledon polyamide wetting layers; (b) 5 Viledon polyamide wetting layers and a layer of FAS- 30 ion-exchange membrane and (c) 5 layers of cellulose. The x-axis is time in days and the y-axis is T90 response in seconds.

[0124] Figure 19 is a method of forming a galvanic sensor.

[0125] Figure 20A presents the response to N2O as an interfering gas for a galvanic oxygen sensor comprising a copper anode, having (a) 6 layers of cellulose or (b) no layers of cellulose and 1 layer of FAS-30 ion-exchange membrane and Figure 20B presents the response to N2O as an interfering gas for a galvanic oxygen sensor comprising a copper anode, having (a) 6 layers of cellulose or (b) no layers of cellulose and 1 layer of piperion ion-exchange membrane.

[0126] Detailed Description of an Example Embodiment

[0127] Figure 1 (a) is a schematic illustration of a first embodiment of a galvanic lead-free oxygen sensor 100. The sensor is housed in housing 110 which defines the sensor cavity therein and contains a sensing cathode 120, an anode 130 and an electrolyte. The sensing cathode is mounted on a PTFE support 127. The electrolyte provides the conductive path within the sensor between sensing cathode 120 and anode 130. A first 140 and a second connecting pin 145 extend from the bottom of the sensor housing, enabling an electrical connection to be established from outside the sensor housing 110 to each of the cathode 120 and the anode 130 respectively. Each of the connecting pins acts as an electrical interface.

[0128] Gas to be detected by the sensor passes a dust filter 150 and enters the housing via capillary 152 located at the top of the sensor body. In an alternative embodiment, the entrance to the capillary from the external atmosphere may be covered by a bulk flow disk which has the effect of damping sudden fluctuations in the pressure external of the sensor. In either case, the sensor is diffusion limited.

[0129] A diffuser 154 is positioned at the sensor side of the capillary 152. The diffuser 154 serves to spread out gas molecules entering the sensor through the capillary 152 over as large an area as possible of the sensing cathode 120.

[0130] Oxygen molecules present in the gas entering the sensor are reduced at the sensing cathode 120 producing hydroxide ions. The balancing reaction at the anode 130 is an oxidation of metal atoms of the anode.

[0131] The sensing cathode 120 catalyses the reduction of oxygen molecules. In the embodiment illustrated in Figure 1 (a) the sensing cathode 120 is made of platinum. The sensing cathode may be made from a less expensive material than platinum and may have a thin platinum surface layer to catalyse oxygen gas reduction. Other possible catalysts include gold, silver and platinum group metals. The catalyst is typically deposited on a PTFE support. The anode of the embodiment illustrated in Figure 1 (a) is made of sintered copper. That is the anode is manufactured from copper particles (commercial copper powder) which have been compressed together and subjected to a heat treatment. This manufacturing process results in a porous anode. In an example, the copper particles may have a particle size of <100pm and be sintered in a non-oxidising atmosphere at a temperature of between 500 and 1085 °C giving an anode with a porosity of 20 to 60% (preferably 30 to 40%).

[0132] Direct electrical contact between the sensing cathode 120 and the anode 130 is prevented by the presence of an electrically insulating body 131 , or separator, between cathode 120 and anode 130. In the embodiment illustrated in Figure 1(a), a shim 125 is also present between cathode 120 and anode 130. The shim 125 divides the inner space of the housing into an anode compartment 138 (functioning as the first compartment), in which the anode 130 resides, and a cathode compartment 128 (functioning as the second compartment), in which the sensing cathode 120 resides. The shim 125 of this embodiment has an annular form, that is it is in the form of a disk with a central hole 126. The shim 125 is made of metal. In other words, the shim 125 has the form of a metallic washer. The shim is therefore electrically conducting. A first conducting connection 118 is present between the shim 125 and the sensing cathode 120. In the embodiment illustrated in Figure 1(a) the first conducting connection is composed of a shim strip 118 electrically connected at one of its ends with the shim. The other end of the first conducting connection is electrically connected to the sensing cathode through a platinum tag 119 welded to the first conducting connection 118 of, or of itself welded to, the shim 125 (see Figure 1 (b)). The anode compartment 138 is within the dashed boundary lines indicated in Figure 1 (d); the cathode compartment 128 is within the dashed boundary lines indicated in Figure 1(d). Each of the anode compartment 138 and the cathode compartment is bounded by the internal walls of the sensor housing and the shim.

[0133] A second conducting connection 142 is present between the shim 125 and the first connecting pin 140. In this way the shim 125 provides part of the conducting path between the sensing cathode 120 and the first connecting pin 140. The second conducting connection 142 is electrically connected at one end to the shim 125 and electrically connected at its second end to the first connecting pin 140. The second conducting connection 142 extends down the inside of the sensor housing 110, along a side of the anode 130. An insulator layer 144 is positioned between the anode 130 and the second conducting connection 142 to ensure that they are kept electrically isolated from each other. In the embodiment illustrated in Figure 1(a), the insulator layer 144 is in the form of a wick band. As well as providing a function of isolating the anode 130 from the second conducting connection 142, the wick band 144 wicks electrolyte and thus helps to distribute electrolyte throughout the internal space of the sensor.

[0134] An anode connector 164 provides the electrical connection between the anode 130 and the second connecting pin 145, enabling the anode to be connected electrically from outside the sensor housing 110.

[0135] The central hole 126 of the shim provides a fluidic connection between the anode compartment 138 and the cathode compartment 128. The shim 125 is mounted within the sensor housing 110 in such a way that the only fluidic connection between the anode compartment 138 and the cathode compartment 128 is through the central hole 126 of the shim 125. The outer edge of the shim abuts the inner wall 112 of the housing, creating a seal at the outer edge. Liquid moving between the anode compartment 138 and the cathode compartment 128 is required to pass through the central hole 126 of the shim 125. The seal between the outer edge of the shim and the inner wall 112 of the sensor housing 110 prevents electrolyte, or any fluid, from passing around the outer edge of the shim 125, and so functions as the perimeter seal.

[0136] The top surface of the anode is the surface closest to the shim. In the embodiment illustrated in Figure 1(a), three separator layers 132, 134, 136 are positioned between the top surface of the anode 130 and the shim 125. The shim also functions to compress separator layers 132, 134, 136 against the top surface of the anode 130. A glass wool wetting filter 122 and an additional separator layer 124 are positioned between the shim 125 and the sensing cathode 120 and compressed by the shim. Each of these layers 132, 134, 136, 122, 124 is permeable to the sensor electrolyte and is electrically insulating. Due to the mounting of the shim 125 and the seal between the outer edge of the shim and the inner wall 112 of the sensor housing 110 electrolyte passing between the anode compartment 138 and the cathode compartment 128 is forced to pass through the separator layers. In the illustrated embodiment each of these separator layers is made of cellulose.

[0137] A further separator layer 148, also made of cellulose, is positioned below the anode 130. Between the further separator layer 148 and the bottom of the anode 130 is a 0.5mm thick stainless steel mesh 156 spacer sandwiched between the wick band insulator layer and the bottom of the anode.

[0138] A vent capillary 158 is present in the bottom wall of the sensor housing. The vent capillary 158 enables excess gas pressure built up within the sensor housing to escape into the surrounding atmosphere (or conversely, excess external pressure to equalise with the inner gas space). A porous membrane 162 covers the entrance to the vent capillary within the sensor housing 110, thereby preventing unwanted release of electrolyte from inside the sensor.

[0139] As is apparent from Figure 1(a), internally the sensor 100 has a layered structure. The layered structure of the sensor 100 is set out in expanded form in Figure 1 (c). In the manufactured sensor 100, as ready for use, the layers are tightly packed together within the housing. In a first configuration of the embodiment illustrated in Figure 1 (a) and (c), each of separator layers 124, 132, 134 and 136 is made of cellulose. The cellulose separator layers function as the wetted layers. The electrolyte is potassium hydroxide.

[0140] As has been mentioned above, the internal construction of the sensor is such that liquid moving between the anode compartment 138 and the cathode compartment 128 is constrained to pass through the central hole 126 of the shim 125. For sensors which have been operated over an accelerated lifetime, the present inventors have observed that where sensors have a copper anode, the cellulose layer neighbouring the shim 125 accumulates a blue deposit 129. An example of a cellulose layer in such a sensor which has been opened up is shown in Figure 2. The centrally positioned blue deposit 129 is clearly visible. Such a blue colouring is characteristic of oxidised copper. Without wanting to get tied down to a particular theory, it is apparent that oxidised copper atoms become entrapped in the cellulose separator. The origin of the oxidised copper atoms can only be the copper present in the anode, which is oxidised during operation of the senor. The oxidised copper atoms pass into the electrolyte from where they are deposited on or within the cellulose separator membrane. Thus, copper compounds are immobilised where they are not in contact with the anode, sensing cathode, or electrical connections to the anode and sensing cathode.

[0141] The performance of the sensor of the present disclosure is illustrated in Figure 3. Figure 3(a) and 3(b) each shows the response of three capillary-type sensors, i.e. the gas to be measured passes through a capillary before contacting the sensing cathode. The response shown is to a sudden exposure of 20.9% of oxygen at around 300 s and then a cessation of the oxygen exposure at around 600 s. The sensors of Figure 3(a) each have one layer of cellulose between the shim and the sensing cathode and two layers of cellulose between the shim and the sintered copper porous anode (total 55.9mg cellulose; cellulose per unit volume 23.8 mg / cm3; cellulose / anode mass ratio 6.1 mg / g; cellulose per electrolyte volume 79.9mg / ml; anode mass / electrolyte volume 13.1g / ml. The sensors of Figure 3(b) each have one layer of cellulose between the shim and the sensing cathode, three layers of cellulose between the shim and the porous anode, and one layer of cellulose between the base of the anode and the inside of the sensor housing (total 90.9mg cellulose; cellulose per unit volume 38.8 mg / cm3; cellulose / anode mass ratio 9.9mg / g; cellulose per electrolyte volume 129.9mg / ml; anode mass / electrolyte volume 13.1g / ml).

[0142] The age evolution of the sensors is shown in Figure 3(c) which shows how the output response to 20.9% of oxygen varies over more than 60 days. The sensors demonstrate what is essentially a constant output over 60 days.

[0143] The impact of the number of cellulose layers on the response speed of capillary-type sensors according to the present disclosure is illustrated in Figure 3(d). The plots show the evolution over circa 150 days of the T90 response, i.e. the amount of time required for the sensor output to reach 90% of its maximum, for (i) a capillary-type sensor with no cellulose layer, instead there were layers of polyamide (total Omg cellulose; cellulose per unit volume 0 mg / cm3; cellulose / anode mass ratio 0 mg / g; cellulose per electrolyte volume 0 mg / ml; anode mass / electrolyte volume 13.1g / ml); (ii) a capillarytype sensor with one cellulose layer, that being in the cathode compartment next to the shim (total cellulose 20.9mg; cellulose per unit volume 8.9 mg / cm3; cellulose / anode mass ratio 2.3mg / g; cellulose per electrolyte volume 29.9mg / ml; anode mass / electrolyte volume 13.1g / ml); (iii) a capillary-type sensor with one cellulose layer in the cathode compartment next to the shim, and two cellulose layers in the anode compartment neighbouring each other and in contact with the anode (total cellulose 55.9mg; cellulose per unit volume 23.8 mg / cm3; cellulose / anode mass ratio 6.1 mg / g; cellulose per electrolyte volume 79.9mg / ml; anode mass / electrolyte volume 13.1g / ml); and (iv) a capillary-type sensor with two cellulose layers in the cathode compartment, and three cellulose layers in the anode compartment (total cellulose 94.3mg; total cellulose per unit volume 40.2 mg / cm3; cellulose / anode mass ratio 10.3 mg / g; cellulose per electrolyte volume 134.7mg / ml; anode mass / electrolyte volume 13.1g / ml). Results are shown for three examples of each sensor type. The data in Figure 3(d) clearly illustrate that the number of cellulose layers in a sensor has a direct influence on the retainment of sensor response time as the sensor ages, and that a sensor with more cellulose layers retains its speed of response better than a sensor with fewer cellulose layers. Each cellulose layer has around 17 mg of cellulose. The sintered anode has a mass of circa 9 g and is created out of copper particles which are 100 pm or less, creating an anode with a porosity of 34%. The volume of the sensor cavity is 2344 mm3. The T90 response of the oxygen sensors at 100 days as a function of the quantity of cellulose present within the sensor is plotted in Figure 3(e), showing effectively a linear response.

[0144] In the sensor of the present disclosure the inventors have found that the working lifetime of the sensor can be optimised by utilising welding between internal conductors where electrical continuity is required for sensor operation. The sensor elements between which a welded connection is made are shown in the example sensor 300 illustrated in Figure 4. Shim strip 318 which provides the electrical connection between the sensing cathode 320 and shim 325 through platinum tag 319, is welded at one end to platinum tag 319, and welded at the other end to shim 325. Second conducting connection 342, providing the electrical connection between shim 325 and the sensing cathode 320 and first connecting pin 340, is welded at the first shim end to the shim 325, and also welded at the second connecting pin end to connecting pin 340. Anode connector 364 connecting electrically anode 330 to second connecting pin 345 is welded at the anode end to the anode 330, and at the connecting pin end is welded to the connecting pin 345. Welding the connections at these points has been found to improve the lifetime of the sensor. In the absence of a welded connection, oxide can build up on the surface of one or more of the metals. This can result in a loss of electrical contact. Welding two metallic surfaces together creates a bond between the two metals which is not susceptible to being destroyed through oxidation.

[0145] The positive effect that welding internal connections has on the lifetime of the sensor is illustrated in the graphs of Figure 5. Each of the graphs of Figure 5 shows the output of the sensor over a period of circa 70 days. Figure 5(a) illustrates sensor output for a series of identical oxygen sensors which have no welding between the shim strip 318 and shim 325. Figure 5(b) similarly illustrates sensor output for a series of identical oxygen sensors in which the anode connector 364 is not welded to the anode 330. The graphs in Figures 5(a) and 5(b) are to be compared to Figure 5(d) which shows output graphs for a series of identical sensors where shim strip 318 is welded to shim 325 and anode connector 364 is welded to the anode 330. It is apparent that the welding has a positive effect on the sensor lifetime. Figure 5(c) shows a further series of output plots for a group of sensors which do not include platinum strip 319 which is normally welded to the shim strip 318 and connects to sensing cathode 320. It is apparent from Figure 5(c), particularly when compared to Figure 5(d), that the presence of platinum strip 319 improves the performance of the sensor. Platinum is highly resistant to oxidation, suggesting that the deterioration in the sensor lifetime observed if platinum is left out of the sensor structure is due to oxidation induced breakdown on electrical contacts within the sensor.

[0146] An alternative embodiment to the sensor is illustrated in Figure 6. Although this alternative embodiment is similar in many ways to what is described elsewhere in this specification, the separator layer 168 in contact with the sensing cathode 120 (on the side of the cathode opposite to that where the diffuser is positioned) is made from an ion exchange membrane 168, specifically an anion exchange membrane (in this example Fumasep FAM brand anion exchange membrane, Fumasep is a trade mark) which allows passage of anions, particularly hydroxide anions, but which does not allow the passage of anode metal ions.

[0147] A further type of membrane property which may be utilised in the present sensor is that of size exclusion, which only allows ions to pass if they are smaller than a particular threshold. For example, a size exclusion membrane may be chosen to exclude anode metal ions, e.g. Cu2+ions, by size. In this way anode metal ions originating from oxidation of the anode can be retained within the anode compartment, while allowing the passage of OH' ions (in the case of an alkaline electrolyte). Some membranes may function as both ion-exchange membranes and size-exclusion membranes.

[0148] In a second configuration of the embodiment, the separator layer 124 contacting the shim 125 in the cathode compartment 128 is a size selective membrane.

[0149] In a further alternative, each of the separator layers may comprise an ion exchange membrane.

[0150] Figure 7 illustrates a further embodiment of the present invention. The sensor 200 illustrated in Figure 7 has a partial pressure membrane 252 positioned between the diffuser 254 and the sensing cathode 220 at the entrance where gas to be detected enters the sensor. The sensing cathode is mounted on a PTFE mount 227. The sensor 200 does not possess a vent capillary. Otherwise, the embodiment illustrated in Figure 7(a) shares many the features with the sensor described earlier.

[0151] Gas to be detected by the sensor passes a dust filter 250 and enters the sensor housing 210 via partial pressure membrane 252 located at the top of the sensor body.

[0152] Diffuser 254 is positioned at the sensor side of partial pressure membrane 252. Diffuser 254 causes gas molecules entering the sensor to spread out over as large an area as possible of the external surface of the partial pressure membrane 252.

[0153] Oxygen molecules present in the gas entering the sensor are reduced at the sensing cathode 220 producing hydroxide ions. The balancing reaction at the anode 230 is an oxidation of metal atoms of the anode.

[0154] The sensing cathode 220 catalyses the reduction of oxygen molecules. In the embodiment illustrated in Figure 7(a) the sensing cathode 220 is made of platinum. The sensing cathode may be made from a less expensive material than platinum and may have a thin platinum surface layer to catalyse oxygen gas reduction. Other possible catalysts include gold, silver and platinum group metals.

[0155] The anode of the embodiment illustrated in Figure 7(a) is made of sintered copper, prepared as described with reference to Figure 1 above, resulting in a porous anode. Direct electrical contact between the sensing cathode 220 and the anode 230 is prevented by the presence of an electrically insulating body 231 , or separator, between cathode 220 and anode 230. In the embodiment illustrated in Figure 7(a), a shim 225 is also present between cathode 220 and anode 230. The shim 225 divides the inner space of the housing into an anode compartment 238 (the first compartment), in which the anode 230 resides, and a cathode compartment (the second compartment) 228, in which the sensing cathode 220 resides. The shim 225 of this embodiment has an annular form, that is it is in the form of a disk with a central hole 226. The shim 225 is made of metal. In other words, the shim 225 has the form of a metallic washer. A first conducting connection 218 is present between the shim 225 and the sensing cathode 220. In the embodiment illustrated in Figure 7(a) the first conducting connection 218 is composed of a shim strip 218 electrically connected at one of its ends with the shim. The other end of the first conducting connection is electrically connected to the sensing cathode through a platinum tag 219 welded to the shim 218 (see Figure 1 (b)).

[0156] A second conducting connection 242 is present between the shim 225 and the first connecting pin 240. In this way the shim 225 provides part of the conducting path between the sensing cathode 220 and the first connecting pin 240. The second conducting connection 242 is electrically connected at one end to the shim 225 and electrically connected at its second end to the first connecting pin 240. The second conducting connection 242 extends down the inside of the sensor housing 210, along a side of the anode 230. An insulator layer 244 is positioned between the anode 230 and the second conducting connection 242 to ensure that they are kept electrically isolated from each other. In the embodiment illustrated in Figure 7(a), the insulator layer 244 is in the form of a wick band. As well as providing a function of isolating the anode 230 from the second conducting connection 242, the wick band 244 wicks electrolyte and thus helps to distribute electrolyte throughout the internal space of the sensor.

[0157] An anode connector 264 provides the electrical connection between the anode 230 and the second connecting pin 245, enabling the anode to be connected electrically from outside the sensor housing 210. The central hole 226 of the shim provides a fluidic connection between the anode compartment 238 and the cathode compartment 228. The shim 225 is mounted within the sensor housing 210 in such a way that the only fluidic connection between the anode compartment 238 and the cathode compartment 228 is through the central hole 226 of the shim 225. The outer edge of the shim abuts the inner wall 212 of the housing, creating a seal at the outer edge. Liquid moving between the anode compartment 238 and the cathode compartment 228 is required to pass through the central hole 226 of the shim 225. The seal between the outer edge of the shim and the inner wall 212 of the sensor housing 210 prevents electrolyte, or any fluid, from passing around the outer edge of the shim 225.

[0158] The top surface of the anode is the surface closest to the shim. In the embodiment illustrated in Figure 7(a), three separator layers 232, 234, 236 are positioned between the top surface of the anode 230 and the shim 225. Two additional separator layers 222, 224 are positioned between the shim 225 and the sensing cathode 220. Each of these separator layers 232, 234, 236, 222, 224 is permeable to the sensor electrolyte and is electrically insulating. Due to the mounting of the shim 225 and the seal between the outer edge of the shim and the inner wall 212 of the sensor housing 210, electrolyte passing between the anode compartment 238 and the cathode compartment 228 is forced to pass through the separator layers. In the illustrated embodiment each of these separator layers is made of cellulose.

[0159] A further separator layer 248, also made of cellulose, is positioned below the anode 230. Between the further separator layer 248 and the bottom of the anode 230 is a mesh 256 sandwiched between the wick band insulator layer and the bottom of the anode, which functions as a spacer.

[0160] As is apparent from Figure 7(a), internally the sensor 200 has a layered structure. The layered structure of the sensor 200 is set out in expanded form in Figure 7(b). In the sensor 200 which is ready for use, the layers are tightly packed together within the housing. Cathode catalyst 220 is deposited on the PTFE membrane which is welded to the housing to form a liquid seal. In a first configuration of the embodiment illustrated in Figure 7(a) and (b), each of separator layers 222, 224, 232, 234 and 236 is made of cellulose. (Grade 1 Whatman. Whatman is a trade mark) The electrolyte is potassium hydroxide. In a second configuration of the embodiment illustrated in Figure 7(a) and (b), the separator layer 224 contacting the shim 225 in the cathode compartment 228 is a size selective membrane.

[0161] In an alternative embodiment which is not illustrated, the separator layer 268 in contact with the sensing cathode 220 (on the side of the cathode opposite to that where the diffuser is positioned) is made from an ion exchange membrane 268, specifically an anion exchange membrane (in this example Fumasep FAM brand anion exchange membrane, Fumasep is a trade mark) which allows passage of anions (small anions and also small cations to some degree), particularly hydroxide anions, but which does not allow the passage of anode metal ions.

[0162] An embodiment of a galvanic sensor 300 is illustrated in cross-section in Figure 8(a) and in an expanded view in Figure 8(b).

[0163] The housing 310 of the sensor is made up of a first housing section 316 and a second housing section 314 which connect together to form the housing. A capillary 352 is provided in the second housing section 314, which can also be considered as the lid of the sensor, to allow gas to be measured to enter into the sensor. Having passed through the entrance capillary, gas to be measured then interacts with diffuser 354 which distributes the gas over its own extent, thus spreading the flow of gas molecules over a large surface area of the sensing cathode, or sensing electrode, 320. The sensing cathode is formed as a platinum / graphite powder, mixed with a gas permeable, hydrophobic PTFE powder along with dispersant and thickening agents, which is deposit onto a gas permeable, hydrophobic PTFE support membrane and baked. The PTFE powder sinters to bind the mixture together. The additives burn off or are harmless residues. The cathode therefore comprises a network of hydrophilic pores and channels, creating many regions of three-phase-interface where electrolyte, electrode and gas are all present together allowing for efficient gas reactions to take place.

[0164] The diffuser 354 does not extend to the side wall of the cavity and an adhesive ring 329 surrounds the periphery of the diffuser 354 filling in the space to the cavity side wall. In an alternative sensor construction, instead of a capillary inlet, a diffusion membrane is utilised to limit the flow of gas to be measured into the sensor. A diffusion membrane construction is suited for partial pressure measurements. The diffusion membrane may be positioned directly after the diffuser 354 in the direction of flow of gas into the sensor.

[0165] The cavity within the housing 310 is largely occupied by the anode 330. The anode is a sintered copper anode. A cellulose disc 356 is sandwiched between the base of the anode 330 and the base of the cavity in the first housing section 316. The cavity of the sensor housing, created when first housing section 316 and second housing section 314 are connected together, is divided into a first compartment (in the first housing section) and a second compartment (in the second housing section). The first compartment is separated from the second compartment by a shim washer 325. The shim washer 325 is arranged within the sensor housing in such a way that the only way for liquid to pass between the two compartments is through the central hole of the washer. When the sensor is assembled by heat welding, the shim washer is trapped and compressed to form a hermetic seal between the cavity wall and the outer edge of the shim washer, the seal preventing the passage of liquid electrolyte between the first compartment and the second compartment around the edge of the shim.

[0166] A wick 344 extends from the base of the anode 330, up the side of the anode to the top surface of the anode. The wick is of polyethersulfone (PES) and facilitates transport of electrolyte within the cavity.

[0167] Sandwiched between the top surface of the anode 330 and the shim washer 325, and located in the first compartment, is a plurality of layers each formed by a cellulose disc. The illustrated embodiment has four cellulose discs 332, 334, 336, 338 layered one upon the other. One end of the PES wick 344 is sandwiched between the cellulose disc 332 which is in direct contact with the anode 330, and the next cellulose disc 334. A tightly packed arrangement of anode and cellulose between the base of the housing and the shim defining the second compartment is what determines the number of discs present.

[0168] The second compartment houses the working electrode 320, otherwise called the sensing cathode 320. In the embodiment illustrated in Figure 8, the sensing cathode 320 is located centrally on the main axis of the sensor and extends to the side wall of the cavity. On the side of the sensing cathode 320 which is opposite to the side facing the entrance capillary 352 and in contact with the sensing cathode 320 is a glass wool wetted layer 368 and layered cellulose discs 324, 323. Sandwiched between this series of cellulose discs and shim 325 is a disc of polyethersulfone (PES) 322. A metallic tab 318 provides an electrical connection between shim washer 325 and sensing cathode 320.

[0169] The embodiment of the sensor illustrated in Fig. 8 includes a vent capillary 358 in the first housing section 316. The vent is in the side wall of the first housing section. The vent capillary 358 provides an exhaust for excess gas which may build up within the sensor, or alternatively a gas inlet should ambient gas pressure outside of the sensor exceed internal gas pressure. A porous vent membrane 362 is positioned over the vent capillary entrance within the sensor. This has the effect, firstly, of preventing liquid electrolyte within the sensor from exiting through the vent, while the capillary itself mitigates the loss of electrolyte through evaporation. The porous vent membrane 362 is attached to the cavity wall inside the sensor housing by welding. The inside of the sensor housing contains a flat surface 359, in the middle of which is the entrance to the vent capillary 358, and on which the porous vent membrane 362 is positioned. Figures 12(a) and 12(b) show two inside views of the sensor housing. In Figure 12(a) the flat surface 459 for receiving the vent membrane is seen. In Figure 12(b) the porous membrane 462 is shown in place. The flat surface 459 provides a stable surface on which the vent membrane 462 can be mounted. The flat surface 459 facilitates the welding of the vent membrane 462 onto the mounting surface and provides for a stable final arrangement minimising strain on the vent membrane.

[0170] An effect of the cellulose layers in the sensor is demonstrated in the data shown in Figure 9. Figure 9 plots the progression of the T90 response of a galvanic sensor over time. The T90 response is the time it takes for the sensor to reach 90% of a new, stable reading after a sudden change in the measured gas concentration. The different graphs in Figure 9 each represent a different anode material.

[0171] Figure 9(a) shows data from a sensor with a copper anode. Within the figure are three solid lines for which the T90 response lengthens with time. The sensor used to obtain this data included zero cellulose discs (total cellulose: Omg; cellulose per unit volume (of the sensor cavity) 0 mg / cm3; cellulose / anode mass ratio 0 mg / g; cellulose per electrolyte volume 0 mg / ml; anode mass / electrolyte volume 14.5 g / ml). Polyamide (Viledon) discs were used instead as wetting discs. The two dashed lines reflect a sensor with a more stable T90 response over time. These data were recorded using a sensor with three layers of cellulose (total cellulose 55.9 mg; cellulose per unit volume 23.8 mg / cm3; cellulose / anode mass ratio 5.9 mg / g; cellulose per electrolyte volume 86 mg / ml; anode mass / electrolyte volume 14.5 g / ml). Finally, the three lines which include square data points show a clear improvement in the T90 response of the sensor over time. These data were obtained from a sensor with five layers of cellulose (total cellulose 90.9 mg; cellulose per unit volume 38.8 mg / cm3; cellulose / anode mass ratio 9.7 mg / g; cellulose per electrolyte volume 139.8 mg / ml; anode mass / electrolyte volume 14.5 g / ml).

[0172] Figure 9(b) presents similar plots showing the T90 evolution of a sensor according to the present disclosure with time. The sensor has an anode of copper-nickel. The three lines without explicitly shown data points correspond to measurements taken using a sensor with no cellulose layers (Omg cellulose). Polyamide (Viledon) discs were used instead as wetting discs. The three lines with square shaped data points correspond to measurements taken using a sensor with five cellulose layers between the anode and the sensing cathode (90.9mg cellulose; cellulose per unit volume 38.8 mg / cm3; cellulose / anode mass ratio 9.7 mg / g; cellulose per electrolyte volume 139.8 mg / ml; anode mass / electrolyte volume 14.5 g / ml). The data demonstrate a positive effect on the longevity as determined by the T90 response.

[0173] Figure 9(c) presents T90 evolution data for a sensor according to the present disclosure, this time with an anode of molybdenum. The solid lines represent data which was recorded from a sensor with no cellulose layers (Omg cellulose). Polyamide (Viledon) discs were used instead as wetting discs. The dashed line (and square data points) represents measurements taken using a sensor with five cellulose layers (90.9mg cellulose; cellulose per unit volume 38.8 mg / cm3; cellulose / anode mass ratio 12.2 mg / g; cellulose per electrolyte volume 139.8 mg / ml; anode mass / electrolyte volume 11.4 g / ml).

[0174] Finally Figure 9(d) presents T90 evolution data for a galvanic sensor comprising a lead anode. The two solid lines shown without explicitly emphasised data points were recorded using sensors comprising zero cellulose layers (Omg cellulose). Polyamide (Viledon) discs were used instead as wetting discs. The three dashed lines were recorded using sensors comprising three cellulose layers. The three solid lines shown with square data points were recorded using sensors comprising five cellulose layers (90.9mg cellulose; cellulose per unit volume 38.8 mg / cm3; cellulose / anode mass ratio 7.9 mg / g; cellulose per electrolyte volume 139.8 mg / ml; anode mass / electrolyte volume 17.7 g / ml).

[0175] The data presented in Figure 9 clearly demonstrate that the presence of cellulose within the galvanic sensor as herein disclosed has the technical effect of improving the longevity of the sensor as determined by its T90 response. The degree to which this technical effect manifests itself is dependent on details of the sensor, such as the material of the anode.

[0176] The sensor illustrated in Figure 8 incorporates cellulose in a layer or in a plurality of layers to restrict the passage of ions between the sensing cathode and the anode. In particular, the passage of anode metal ions to the sensing cathode is restricted. The arrival at the sensing cathode of anode metal ions, which may be created during operation of the sensor, can impair the function of the sensing cathode, thereby affecting the performance of the sensor.

[0177] Although in the above examples, the sensor is divided into two compartments, it is also useful to include metal binding agent, such as cellulose layers, in sensors which are not divided so clearly into two compartments as they anyway bind anode metal ions and restrict the passage of anode metal ions to the sensing cathode. Still, the use of two compartments with a barrier to anode metal ions passing between the compartments except through a metal binding agent, such as a cellulose layer, is advantageous.

[0178] A further embodiment of a sensor 400 is illustrated in cross-section in Figure 10. The sensor 400 is shown in an exploded view in Figure 11. Figure 11 shows the same sensor illustrated in Figure 10, but in an expanded view. In this sensor, an ion exchange membrane (I EM) is used to restrict the passage of ions between the sensing cathode and the anode, particularly the passage of anode metal ions to the sensing cathode.

[0179] The housing 410 of the sensor is made up of a first housing section 416 and a second housing section 414 which connect together to form the housing. A capillary 452 is provided in the second housing section 414, which can also be considered as the lid of the sensor, to allow gas to be measured to enter into the sensor. Having passed through the entrance capillary, gas to be measured then interacts with diffuser 454 which distributes the gas over its own extent, thus spreading the flow of gas molecules over a large surface area of the sensing cathode, or sensing electrode, 420. An adhesive ring 429 surrounds the periphery of the diffuser 454, which does not extend to the side wall of the cavity. In an alternative sensor construction, instead of a capillary inlet, a diffusion membrane is utilised to limit the flow of gas to be measured into the sensor. A diffusion membrane construction is suited for partial pressure measurements. The diffusion membrane may be positioned directly after the diffuser 454 in the direction of flow of gas into the sensor.

[0180] The cavity within the housing 410 is largely occupied by the anode 430. The anode is a sintered copper anode. A disc of polyamide filter material 456 is sandwiched between the base of the anode 430 and the base of the cavity in the first housing section 416. The polyamide filter material used was obtained from Viledon (Viledon is a trade mark of Freudenberg GmbH). The cavity of the sensor housing is divided into a first compartment and a second compartment. The first compartment is separated from the second compartment by a shim washer 425. The shim washer 425 is arranged within the sensor housing in such a way that the only way for liquid to pass between the two compartments is through the centre of the washer. When the sensor is assembled by heat welding, the shim washer is trapped and compressed to form a hermetic seal between the cavity wall and the outer edge of the shim washer, the seal preventing the passage of liquid electrolyte between the first compartment and the second compartment around the edge of the shim.

[0181] A wick 444 extends from the base of the anode 430, up the side of the anode to the top surface of the anode. The wick is of polyethersulfone (PES) and facilitates transport of electrolyte within the cavity. The base of the sensor has a plurality of electrical connection pins 445.

[0182] Sandwiched between the top surface of the anode 430 and the shim washer 425, and located in the first compartment, are a plurality of polyamide wetting layers 455a, 455b, 455c (Viledon). These wetting layers are hydrophilic and porous. In contrast to cellulose or other barrier layers they do not interact with the dissolved anode metal ions.

[0183] The second compartment houses the working electrode 420. In the illustrated embodiment, the sensing cathode 420 is located centrally on the main axis of the sensor and extends to the side wall of the cavity. The sensing cathode 420 is formed as described above on a PFTE porous support membrane. On the side of the working electrode which is opposite to the side facing the entrance capillary 452 and in contact with the working electrode is an ion exchange membrane 468. Sandwiched between the ion exchange membrane 468 and shim 425 is a second plurality of wetting layers 427a (glass wool), 427b (polyamide). A metallic tab 418 provides an electrical connection between shim washer 425 and sensing cathode 420.

[0184] The embodiment of the sensor illustrated in Fig. 10 includes a vent capillary 458 in the first housing section 416. Vent capillary 458 provides a path for gas exchange between the interior 464 of the first housing section and the exterior of the sensor, to allow pressure equilibration. A porous vent membrane 462 is positioned over the capillary entrance within the sensor. This has the effect of preventing liquid electrolyte within the sensor from exiting through the vent, whereas the capillary restricts moisture loss by gas phase evaporation. The porous vent membrane 462 is attached to the cavity wall inside the sensor housing by welding. The inside of the sensor housing contains a flat surface 459, in the middle of which is the entrance to the vent capillary 458, and on which the porous vent membrane 462 is positioned. Figures 12(a) and 12(b) show two inside views of the sensor housing. In Figure 12(a) the flat surface 459 for receiving the vent membrane is seen. In Figure 12(b) the porous membrane 462 is shown in place. The flat surface 459 provides a stable surface on which the membrane can be mounted. The flat surface 459 facilitates the welding of the membrane 462 onto the mounting surface and provides for a stable final arrangement minimising strain on the membrane.

[0185] Examples of commercially available ion exchange membranes (lEMs) which have been tested in sensors include PiperlON exchange membrane and FAS anion exchange membrane. For each membrane tested, the presence of an ion exchange membrane was seen to improve the T90 response of the sensor, T90 being the time it takes for the sensor to reach 90% of a new, stable reading after a sudden change in the measured gas concentration. Figure 13(a) compares the T90 response of two sensors in which only wetting layers (polyamide and glass wool) are present between the sensing cathode 420 and the shim 425 (round data points and solid lines) to two sensors which include an FAS-30 ion exchange membrane 468 in contact with the sensing cathode 420 as described above (square data points and dashed lines). The anode 430 is of sintered copper. Similar improved T90 responses were recorded from other sensor systems. The results shown in Figure 13(b) are from a similar sensor system to Figure 13(a), with a sintered copper anode and wetting layers of polyamide (Viledon) and glass wool material. A PiperlON exchange membrane is in place instead of a FAS-30 membrane. The presence of the PiperlON exchange membrane preserves the T90 sensor response (triangular data points and dashed lines) over a sensor without an ion exchange membrane (round data points and solid lines).

[0186] The results plotted in Figure 13(c) and Figure 13(d) are all from sensors comprising a bronze anode. The solid lines join data obtained from sensors with wetting layers (polyamide and glass wool) and no ion exchange membrane between the anode and the sensing cathode. The dashed lines join data points from sensors which comprises an ion exchange membrane as per the arrangement of Figure 10. The data of the dashed lines in Figure 13(c) come from sensors comprising a FAS-30 anion exchange membrane. The data of the dashed lines in Figure 13(d) come from sensors comprising a PiperlON anion exchange membrane.

[0187] The data shown in Figure 13(e) and Figure 13(f) were obtained from sensors with a copper-nickel anode. The solid lines in both figures is from data taken using sensors without an ion exchange membrane and comprising only polyamide and glass wool wetting layers. The data producing the dashed lines in Figure 13(e) are taken from sensors with a FAS-30 ion exchange membrane in place. The data producing the dashed lines in Figure 13(f) are taken from sensors with a PiperlON ion exchange membrane in place.

[0188] A final example of the positive effect of an ion exchange membrane on the T90 response of a galvanic sensor is shown in Figure 13(g). The figure compares data recorded using sensors without an ion exchange membrane (solid lines) with a sensor comprising a FAS-30 ion exchange membrane (dashed lines). The anode is molybdenum. All the data in the seven graphs shown in Figure 13 tell a consistent story, viz. that the presence of an ion exchange membrane in a galvanic sensor serves to preserve the T90 response of the sensor compared to a sensor which comprises only wetting layers and in which no ion exchange membrane is used.

[0189] We hypothesize that the T90 improvement which the presence of the ion exchange membrane in the sensor brings is due to a protecting effect on the sensing cathode 420. All fluid flow between the first compartment in the sensor housing and the second compartment must pass through the empty centre of the shim washer 425. We propose that the ion exchange membrane 468 traps or otherwise prevents ion species originating from the anode 430, which are possibly created during normal operation of the sensor, from reaching the sensing cathode 420 and poisoning reaction sites. In particular, the ion exchange membrane 468 blocks the passage of anode metal ions to the sensing cathode. This is at least in part due to the membrane functioning as a size exclusion membrane.

[0190] The position of the vent capillary of each of the sensor incorporating cellulose and the sensor incorporating an ion exchange membrane presented above is in the cylindrical side wall of the sensor housing. The present inventors have discovered that the position of the vent capillary is an important consideration in the design of the sensor.

[0191] A change of ambient temperature can result in a pressure drop within the sensor or a build-up of air pockets within the sensor. This can result in bubbles of gas travelling through the wetted layers, which generate glitches or spikes in the sensor output. The presence of a vent can be helpful in avoiding a build-up of or a drop in pressure within the sensor.

[0192] Figure 14 shows sensor output as the ambient temperature is changed from 20°C to - 30°C over circa 40 minutes. The temperature is shown as the dashed line. These are sensors with five layers of cellulose and with a copper anode (total cellulose 90.9 mg; cellulose per unit volume 38.8 mg / cm3; cellulose / anode mass ratio 9.7 mg / g; cellulose per electrolyte volume 139.8 mg / ml; anode mass / electrolyte volume 14.5 g / ml) In Figure 14, line (a), with the open square data points, is a plot of sensor output from a sensor without a vent. A strong perturbation to the sensor output as a result of the temperature change is clear. Lines (b) are from sensors with a capillary vent on the base of the sensor, i.e. on the sensor wall which is opposite to the wall with the entrance capillary, 352, 452. The effect of the change in the ambient temperature is subdued compared to the sensor without a vent. Finally, line (c) is data from a sensor with a vent in the side wall, as per the sensor illustrated in Fig. 8 and that of Fig. 10. This configuration demonstrates the best response to the change in the ambient temperature.

[0193] The inventors believe that positioning the vent in the side wall may provide a benefit as the vent is closer to the sensing cathode where an air pocket is more likely to occur than on the bottom. A further advantage of having the vent exit in a side wall of the sensor is that any humidity which may be expelled from the sensor is not directed to the printed circuit board (PCB) on which the electronics operating the sensor are to be found. In this way, the sensor electronics are protected from damage.

[0194] It is also possible that a vent in the side wall has better access to the ambient atmosphere, although the details of how the sensor is mounted can also have an effect in this regard. The inventors observe that when no vent is present in a galvanic sensor, the oxidation of the anode progresses from the side closest to the cathode to the far side where the anode is welded to the current collector. Similarly, oxygen entering the sensor through the vent is likely to chemically cause local copper oxidation. If the vent is positioned close to the anode weld to the current collector, the aforementioned local oxidation may impact the weld. Positioning the vent on the side would mitigate this effect.

[0195] The present invention also provides a method of manufacturing a sensor. The method comprises forming a first sensor portion by inserting an anode into a first portion of the housing. The method comprises fitting the shim to the second sensor portion with the cathode. The shim has an annular form, that is it is in the form of a disk with a central hole. The outer edge of the shim creates a tight fit to the inside wall of the portion of the sensor housing. The central hole of the shim provides a fluidic connection between sensor cavities created within the sensor housing by the presence of the shim. Once the shim has been fitted, the method comprises attaching the first sensor portion (comprising a first portion of the housing and the anode) to the second sensor portion (comprising a second portion of the housing and the sensing electrode).

[0196] Figure 15(a) shows the accumulated charge and the output of three sensors according to the present disclosure over a period of 80 days, each with a copper anode of 34% porosity and made of particles of up to 150pm size. Figure 15(b) shows equivalent data for three sensors according to the present disclosure, each with a copper anode of 34 % porosity, but made of particles of up to 100pm size. The solid lines show accumulated charge (left hand scale) and the dashed lines show output signal as a percentage (right hand scale). In each case there were five layers of cellulose and the amount of cellulose was total cellulose 90.9 mg; cellulose per unit volume 38.8 mg / cm3; cellulose / anode mass ratio 9.7 mg / g; cellulose per electrolyte volume 139.8 mg / ml; anode mass / electrolyte volume 14.5 g / ml. The decline in sensor performance overtime is more visible in the output data than in the accumulated charge data. It is apparent from the data of Figure 15 that the sensor with the anode formed from particles of smaller size has a performance which is improved over the larger size because the drop off in output signal is delayed further in time, showing a large usable lifetime.

[0197] A further embodiment of a galvanic sensor as otherwise herein disclosed comprises an anode made of sintered molybdenum. The anode is manufactured by compressing together particles in the range of 100-200 pm, with a maximum particle size of 250 pm. The resulting porosity of the anode is 33%. Figure 16(a) shows an image of the anode. The surface structure of the anode is more apparent in Figure 16(b) which is taken at a magnification of x210.

[0198] The performance of sintered molybdenum anode galvanic oxygen sensors is illustrated in Figures 17 and 18. The Figure 17 plots shows the response of 56 day old sensors to a pulse of oxygen. The Figure 18 plots shows the evolution of the response to an unchanging atmosphere of 20.9% oxygen over 70 days.

[0199] In Figure 17 and 18, the sensors of traces (a) have 5 Viledon polyamide wetting layers and so do not have metal ion binding agent nor ion exchange membrane or any other barrier to anode metal ions. The sensors of traces (b) have 5 Viledon polyamide wetting layers and a layer of FAS-30 ion-exchange membrane and so show the performance of molybdenum anode devices with an ion-exchange membrane. The sensors of traces (c) have 5 layers of cellulose and show the performance of molybdenum anode devices with cellulose acting as a metal ion binding agent. The results in each of these figures are from sensors with a capillary entrance.

[0200] As can be seen in Figure 17, when the sensors are suddenly exposed at 300 s to 20.9% (molar percent) oxygen, the sensor output jumps. When the oxygen is removed from the atmosphere sampled by the sensor at 600 s, the sensor response drops immediately. However, for those with only wetting layers, (Viledon and glass wool), the response is weaker. Figure 18 shows that the performance of the sensors holds for at least 70 days for the sensors with FAS anion exchange membrane or cellulose. However, for those with only wetting layers, the performance deteriorates after 42 days.

[0201] Thus, the sensors using molybdenum as anode material are useful as galvanic oxygen sensors, especially with a metal ion binding agent or ion exchange membrane or other barrier to anode metal ions.

[0202] Figure 19 illustrates a method of manufacture 500 of the sensor. In a series of steps the first sensor portion is manufactured. Manufacture of the first sensor portion of this example includes mounting connecting pins to the first portion of the housing 510 and welding the vent membrane 362 to the first portion; attaching electrical connectors to the connecting pins internal to the sensor housing 520; inserting a cellulose layer into the first portion of the housing 530; inserting an anode into the first portion of the housing and electrically connecting the anode to a connecting pin 540 and at the same time inserting an insulator into the first portion of the housing to ensure that an undesired electrical short connection is not created inside the first portion of the housing 550; inserting cellulose separator layers into the first portion of the housing covering the anode 560.

[0203] In a second series of steps the second sensor portion is manufactured. Manufacture of the second sensor portion of this example includes mounting the sensing cathode in a second sensor portion 505; inserting a shim strip into the second sensor portion, whereby the shim strip is conductively attached to a shim, and mounting cellulose separator layers so that the shim strip wraps around the separator layers, sandwiching the separator layers between the sensing cathode and the shim 515; fitting the shim with the separator layers as perimeter seal to the second sensor portion, thereby compressing the separator layers and also forming by the fitting of the shim a fluid tight seal between the outer edge of the shim and the innerwall in the second sensor portion 525. An I EM layer may also be mounted as an alternative as appropriate.

[0204] In a third step 590 attaching the first sensor portion to the second sensor portion.

[0205] We have shown that the use of an anode metal-ion binder as a barrier for anode metal ions improves T90 efficiency, and therefore the operational lifetime of the sensor. Cellulosic binders work especially well with copper containing anodes and provide some improvement to existing lead sensors. The use of an anion exchange membrane, particularly one functioning as a size exclusion membrane, also functions well as a barrier for anode metal ions and is effective with a wide range of anode metals. In each case, we hypothesize that the technical benefits arise from reducing the deposition of anode metal compounds on the sensing cathode.

[0206] We have also found that the anode metal-ion binder and anion exchange membrane can be effective at reducing the effect of certain interferent gases. For example, Figure 20A presents the response to N2O as an interfering gas for a galvanic oxygen sensor comprising a copper anode, having (a) 6 layers of cellulose (total 99.7mg of cellulose) or (b) no layers of cellulose (Omg of cellulose) and 1 layer of FAS-30 ion-exchange membrane and Figure 20B presents the response to N2O as an interfering gas for a galvanic oxygen sensor comprising a copper anode, having (a) 6 layers of cellulose (total 99.7mg of cellulose) or (b) no layers of cellulose (Omg of cellulose) and 1 layer of Piperion ion-exchange membrane. In both experiments the sensors are exposed to nitrogen for 20 min. before being exposed to a mixture of 16.5% (molar percent) N2O in nitrogen for 5 min. A clear sensor response is observed when sensors do not comprise an ion exchange membrane. The use of a ion exchange membrane is necessary here to get the desired immunity to N2O. The sensors used in Figure 20a and Figure 20b are constructed according to Figure 8 with the exception of the sensors for which the cellulose layers have been swapped for Viledon layers and an ion exchange membrane.

[0207] Throughout the description and claims of this specification, the words “comprise” and “contain” and variations of them mean “including but not limited to”, and they are not intended to and do not exclude other components, integers, or steps. Throughout the description and claims of this specification, the singular encompasses the plural unless the context otherwise requires. In particular, where the indefinite article is used, the specification is to be understood as contemplating plurality as well as singularity, unless the context requires otherwise.

[0208] Features, integers, characteristics, or groups described in conjunction with a particular aspect, embodiment, or example of the invention are to be understood to be applicable to any other aspect, embodiment or example described herein unless incompatible therewith. All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and / or all of the steps of any method or process so disclosed, may be combined in any combination, except combinations where at least some of such features and / or steps are mutually exclusive. The invention is not restricted to the details of any foregoing embodiments. The invention extends to any novel one, or any novel combination, of the features disclosed in this specification (including any accompanying claims, abstract and drawings), or to any novel one, or any novel combination, of the steps of any method or process so disclosed.

Claims

Claims1 . A galvanic oxygen sensor comprising: a housing; a sensor cavity within the housing; a sensing cathode and a porous anode within the sensor cavity; wherein a barrier to the passage of anode metal ions between the anode and cathode is provided between the anode and the sensing cathode, and wherein the barrier to the passage of anode metal ions comprises an ionexchange membrane which is permeable to electrolyte but restricts the passage of anode metal ions, or wherein the barrier to the passage of anode metal ions comprises a size exclusion membrane which is permeable to electrolyte but restricts the passage of anode metal ions, or wherein the barrier to the passage of anode metal ions comprises a membrane which is permeable to electrolyte and is both an ion-exchange membrane and a size exclusion membrane and which restricts the passage of anode metal ions.

2. A galvanic oxygen sensor according to claim 1 , whereby fluid communication between the anode and the sensing cathode, is only through the barrier to the passage of anode metal ions.

3. A galvanic sensor according to claim 1 or claim 2, where the anode comprises copper, and less than 0.5% lead by mass.

4. A galvanic sensor according to claim 3, wherein the anode comprises at least 25% or at least 75% copper by mass.

5. A galvanic sensor according to any one preceding claim, wherein the sensor cavity comprises a first compartment and a second compartment, the anode located in the first compartment, the sensing cathode located in the second compartment, fluid communication between the first compartment and the second compartment is only via one or more apertures between the first and second compartment; and wherein the one or more apertures are occluded by the barrier to the passage of anode metal ions.

6. A galvanic sensor according to any one preceding claim, comprising a perimeter seal, extending around the periphery of the sensor cavity to block the passage of anode metal ions from the anode to the sensing cathode except through the barrier.

7. A galvanic sensor according to any one preceding claim, wherein the perimeter seal comprises a shim, which forms a seal against an internal wall of the sensor cavity and comprises one or more apertures, wherein the one or more apertures are occluded by the barrier to the passage of anode metal ions.

8. A galvanic sensor according to claim 7, wherein the shim is electrically conductive and wherein a first electrically conductive connection is provided between an electrical connector of the galvanic sensor and the shim and a second electrically conductive connection is provided between the shim and the sensing cathode.

9. A galvanic sensor according to claim 8, wherein the conductive electrical connection between the shim and the sensing cathode comprises a noble metal region which is pressed against the sensing cathode.

10. A galvanic sensor according to any one of claims 7 to 9, wherein the barrier is compressed between the shim and the sensing cathode.

11. A galvanic sensor according to any one preceding claim, further comprising a metal ion binding agent.

12. A galvanic sensor according to claim 11 , wherein the metal ion binding agent is a porous sheet of a polysaccharide, for example cellulose.

13. A method of forming a galvanic sensor according to any one of claims 6 to 12, comprising the step of forming a first sensor portion comprising a first portion of the housing, with the anode therein, and a second sensor portion comprising a second portion of the housing, with the sensing cathode therein, fitting theperimeter seal to the second sensor portion and then attaching the first sensor portion to the second sensor portion.

14. A galvanic oxygen sensor comprising a housing, the housing containing: a sensing cathode; an anode comprising one or more anode metals, wherein the anode is porous and comprises at least 25% copper by mass and less than 0.5% lead by mass; an electrolyte; and a metal ion binding agent in contact with the electrolyte to bind ions of the one or more anode metals and thereby restrict the deposition of compounds of the one or more anode metals on the sensing cathode.

15. A galvanic sensor of claim 14, where the anode comprises at least 95% copper by mass.

16. A galvanic sensor according to claim 14 or claim 15, wherein the metal ion binding agent is a metal ion immobilising agent.

17. A galvanic sensor according to any one of claims 14 to 16, wherein the metal ion binding agent is a gelling compound which forms metal compound containing gels with anode metal ions.

18. A galvanic sensor according to any one of claims 14 to 17, wherein the metal ion binding agent binds with metal ions to form solid phase anode metal compounds.

19. A galvanic sensor according to any one of claims 14 to 18, wherein the metal ion binding agent binds with copper ions to form solid phase anode metal compounds.

20. A galvanic sensor according to any one of claims 14 to 19, wherein the metal ion binding agent is a polysaccharide or derivative thereof21 . A galvanic sensor according to any one of claims 14 to 20, wherein the metal ion binding agent is cellulosic.

22. A galvanic sensor according to any one of claims 14 to 21 , wherein the housing defines a sensor cavity comprising the anode, the sensing cathode, one or more wetted layers, and the electrolyte, wherein the amount of metal ion binding agent is at least 5mg per cm3of volume of the sensor cavity.

23. A galvanic sensor according to any one of claims 14 to 22, wherein the housing defines a sensor cavity comprising the anode, the sensing cathode, one or more wetted layers, and the electrolyte, wherein the amount of metal ion binding agent is at least 5mg per ml of electrolyte.

24. A galvanic sensor according to any one of claims 14 to 23, wherein the housing defines a sensor cavity comprising the anode, the sensing cathode, one or more wetted layers, and the electrolyte, wherein the amount of metal ion binding agent is at least 1mg per gram of anode metal.

25. A galvanic sensor according to any one of claims 14 to 24, wherein the ratio of the mass of anode metal to the volume of the electrolyte is at least 2 g of anode metal per 1 ml of electrolyte.

26. A galvanic sensor according to any one of claims 22 to 25, wherein the at least one wetted layer comprises or is formed of the metal ion binding agent.

27. A galvanic sensor according to any one of claims 14 to 26, wherein the housing comprises a mass flow restriction member, such that the flux of analyte gas into the sensor is limited by gas phase diffusion, wherein the lifetime charge density of the electrochemical cell, defined as the total amount of charge that the sensor can provide during its lifetime divided by the sensor cavity volume, is at least 0.5 C / mm3.

28. A galvanic sensor according to any one of claims 14 to 27, wherein the housing comprises a membrane such that the flux of analyte gas into the sensor is limited by flux of analyte gas through the membrane material and where the lifetime charge density of the electrochemical cell, defined as the total amountof charge that the sensor can provide during its lifetime divided by the sensor cavity volume, is at least 0.15 C / mm3.

29. A galvanic sensor according to any one of claims 14 to 28, wherein the electrolyte is 10 - 50 wt% of an alkaline metal hydroxide comprising one or more of a lithium hydroxide, sodium hydroxide, potassium hydroxide, rubidium hydroxide, and cesium hydroxide, or the electrolyte is 10 - 50 wt% quaternary ammonium hydroxide.

30. A galvanic sensor according to any one of claims 14 to 29, wherein the sensor comprises an electrical interface which is electrically connected to the anode by a conductor, and the conductor is fused to the anode by one or more of resistance welding, heat welding, chemical bonding, or conductive adhesive.31 . A galvanic oxygen sensor according to any one preceding claim, wherein the anode is formed from particles having a mean diameter of less than 150 microns.

32. A galvanic oxygen sensor comprising a housing, the housing containing: a sensing cathode; an anode comprising one or more anode metals, wherein the anode comprises at least 25% molybdenum by mass; and an electrolyte.

33. A galvanic oxygen sensor according to claim 32, wherein the anode is porous and comprises less than 0.5% lead by mass, and wherein a barrier to the passage of anode metal ions between the anode and sensing cathode is provided between the anode and the sensing cathode.

34. A galvanic oxygen sensor comprising: a housing, a sensor cavity within the housing; a sensing cathode and an anode within the sensor cavity; and a vent through which gas may enter into or egress from the sensor cavity, wherein the housing comprises an inlet in a first external surface, the housing further comprising a second external surface, opposite the first external surface, and at least one third external surface which is curved and the vent extends fromthe curved third external surface to a vent opening in a flat section of an inner wall of the sensing cavity to which a membrane is attached, covering the vent opening.

35. A galvanic oxygen sensor according to claim 34, wherein the housing is cylindrical, with the first and second external surfaces being opposite circular surfaces, forming first and second ends of the sensor, and the third external surface being the curved surface of the cylinder, between the first and second ends.

36. A galvanic oxygen sensor comprising: a housing; a sensor cavity within the housing; a sensing cathode and a porous anode within the sensor cavity; wherein a barrier to the passage of anode metal ions between the anode and cathode is provided between the anode and the sensing cathode, wherein the barrier to the passage of anode metal ions comprises a viscoelastic gel electrolyte with viscosity of at least 500 mPa.s which has the effect of inhibiting diffusion of anode metal ions.

Images