WO2016124874A1

WO2016124874A1 - Electrochemical sensor

Info

Publication number: WO2016124874A1
Application number: PCT/GB2015/050262
Authority: WO
Inventors: Jonathan METTERS; Dimitrios KAMPOURIS; Craig Banks
Original assignee: Kanichi Research Services Limited; Manchester Metropolitan University
Priority date: 2015-02-02
Filing date: 2015-02-02
Publication date: 2016-08-11

Abstract

This invention relates to an electrochemical sensor comprising o- phthalaldehyde and thioglycolic acid for detecting ammonia in breath, an apparatus comprising the sensor, and methods of use.

Description

Electrochemical Sensor

Field of invention The invention relates to an electrochemical sensor for ammonia detection, in particular, to a sensor detecting ammonia in the breath by indirect detection of an electroactive derivative of ammonia. An apparatus comprising the sensor, method of detecting ammonia, of making the sensor and a kit comprising a sensor and the use of the sensor are also disclosed.

At the time of writing this application, Craig Banks is an employee of Manchester Metropolitan University, Jonathan Metters is a PhD student at Manchester Metropolitan University, and Dimitrios Kampouris is an employee of Kanichi Research Services Ltd.

Background to the Invention

Ammonia is a gaseous analyte present in the breath of healthy individuals. It can, however, also be indicative of illness and has the potential to be used as a screening tool for a wide range of ailments. These include kidney disease, liver cirrhosis, or infections with microorganisms such as helicobacter pylori or Candida. An increase in the ammonia in the breath can also be a stress indicator. For these reasons it would be useful to provide a simple and rapid method of measuring the presence of this compound.

In a healthy individual, typical levels of ammonia in the breath are of the order of 800- 900 parts per billion. In unhealthy individuals this can increase to the order of 1-15 parts per million. As the differences between a healthy and unhealthy individual at the lower level of this range are small it is important that an accurate detection method is employed.

The sensing of ammonia in gaseous samples presents numerous potential problems, including the detection of the relatively low levels of ammonia present in the breath sample, the sometimes small difference in ammonia level between a healthy and unhealthy individual, the method of sampling (for instance whether mouth or nasal sampling preferred the need to select an appropriate electrode material entrapment solution and method of entrapment). Gaseous ammonia has been electrochemically sensed, for instance as described in Sensor's Journal, IEEE, 2010, 10, 49-53 where Guoma et al describes the detection monitoring of gases in exhaled human breath using molybdenum trioxide networks. However, the sensor showed a cross sensitivity to carbon monoxide and requires high temperature operation, which can cause the molybdenum trioxide to become semi- conducting. There therefore remains a need for a sensor which can be operated at lower temperature and in the presence of a variety of contaminants without reduction in accuracy.

In addition, the existing technologies are typically not easily transportable and/or require specialised training to operate. Conventional analysis techniques require bulky complex machinery, and many electrochemical sensors lack the stability or sensitivity necessary for this application.

As such, there remains a need for a simple, electrochemical sensor capable accurately detecting ammonia in low concentrations, and preferably at low temperature, whilst negating the problem of interference from other components of the breath. The invention is intended to overcome or ameliorate at least some aspects of this problem.

Summary of the Invention

Accordingly, in a first aspect of the invention, there is provided an electrochemical sensor for detecting ammonia in breath. Typically, the detection will comprise indirect detection of an electroactive derivative of ammonia. Ammonia itself is non- electroactive except at high overpotentials, and lacks sensitivity. As such, ammonia can only sensed directly under extreme conditions and using specialist electrochemical techniques. One method of addressing this is to convert the ammonia to thioisoindole using thioglycolic acid (mercaptoacetate) and o-pthalaldehyde. This combination of components are found in the well known OPA reagent. In the reaction, the thioglycolic acid forms a thioacetal complex with the o-pthalaldehyde and in subsequent ring forming reaction with ammonia thioisoindole is formed as shown in the reaction scheme below.

o-pihakklehyde (CPA) thioglycolic acid thioacetal

thioacetal Scheme 1 thioisoindole

The thioisoindole can act as a direct measure of the level of ammonia in the breath as one molecule of thioisoindole is formed per molecule of ammonia. This allows accurate determination of ammonia concentration. Further, the use of the OPA reagent, well known to react extremely reliably, ensures that all ammonia in the breath is trapped and indirectly detected. Scheme 2 illustrates the electrochemical reaction which occurs with the thioisoindole.

Scheme 2 As used herein, the term "breath" is intended to relate to some or all the mixture of components produced by exhalation from the lungs. This may include one or both of nasal breath or mouth breath and will generally have a gaseous component, for instance, air, the air often being depleted of oxygen and/or rich in carbon dioxide. Further, vapour and/or liquid components may be present, for instance water vapour or water itself. Further the vapour and/or liquid component may include dissolved substances. As such, the sensor of the invention is specifically intended to be able to detect ammonia in the "breath". The sensor may be a gas sensor, and/or it may detect the vapour or liquid component. Further, the sensor may detect "breath condensate", which as used herein refers to the liquid component of breath and/or the condensed vapour component. The sensor may additionally or alternatively be designed to solubilise the gaseous component at the point of detection, so that detection is in solution. The invention typically comprises an electrode assembly which comprises at least two electrodes, one of said electrodes being capable of detecting thioisoindole. The electrochemical sensor will often be in the form of an electrode assembly which comprises a plurality of electrodes or types of electrodes (by which we mean electrodes of different functions), typically at least two types of electrodes. The electrode assembly may comprise at least one working electrode capable of detecting thioisoindole, a counter electrode and a reference electrode. Even more typically, the electrodes are screen printed electrodes, often at least the working electrode will be a screen printed electrode. Often the working electrode will be selected from a carbon, platinum, gold, silver, oxides of these materials, compounds including these materials (for instance boron doped diamond), any nanofunctionalised derivatives of these materials (for instance carbon electrode with platinum nanoparticles), a screen-printed graphite electrode or a mercury drop electrode. In many examples the working electrode will be a carbon electrode. The carbon electrode may be a glassy carbon electrode, graphite electrode, diamond electrode, or pyrolytic carbon electrode, often it will be a graphite electrode, in many cases a screen-printed graphite electrode. The surface area of the working electrode will generally maximized, as this provides a greater and more rapid response, increasing the overall sensing speed. The working electrode may be a nanoelectrode (metallic or non-metallic) or a microelectrode and may comprise nanoparticles which have been, for instance, screen printed onto a substrate; this can have the advantage of offering an improved signal-to-noise ratio relative to macroelectrodes. The nanoparticulate material may be in the form of a sheet, fibre or powder which may be printed onto the substrate, powders are most often used as these are most easily suspended in solvents, and hence are easy to apply in the sensor construction process.

In some examples, the electrode could be derivatised with enzymes or other catalysts which can interact with thioisoindole. In such cases, the interaction of the thioisoindole with the catalyst causes a measurable change in the electrochemical properties of the catalyst, resulting in the production of a signal. Such systems would not require the application of a potential, as it would be the act of detection which would produce the electrochemical response.

It is advantageous to incorporate a reference electrode into the electrode assembly. The reference electrode may be any system which offers a stable predictable electrode potential and which is not sensitive to the analytes being detected. Often, the reference electrode will be selected from silver/silver chloride, hydrogen, saturated calomel, mercury/mercury chloride, lead/lead sulfate, copper/copper (II) sulfate, metallocene (such as ferrocene), quinone, anthracene, and palladium/hydrogen reference electrodes. Where compounds, such as metallocenes, quinone or anthracene are used, these may be substituted by electron withdrawing or donating groups as required to alter their electron transfer properties thereby ensuring that the electrochemical signals for the reference compound and the analytes in question are distinguishable. Often the selection will be from lead/lead sulfate, silver/silver chloride, copper/copper (II) sulfate and metallocene reference electrodes.

The counter electrode may be formed of any substance through which current can pass. However, the counter electrode will typically be selected from the types of electrode listed above with reference to the working electrode, most often the selection will be from carbon, platinum or gold. For screen printed applications, the counter electrode will typically be a carbon electrode.

Often one or more of the electrodes are nanoelectrodes, microelectrodes or combinations thereof. Arrays of the above may also be used, for instance microbands such as those described in US 6,790,341 can be incorporated into the systems of the invention. Nano and microelectrodes can offer a number of advantages over conventionally sized macroelectrodes in electrochemical experiments. These include their small size, which allows use in very small sample volumes or in vivo measurements, improved efficiency of mass transport, smaller double layer capacitance and cell time constant and low ohmic drop.

One or more of the electrodes may be screen printed, and this will often be the case because of ease of fabrication and low cost. Where one electrode is screen printed, often all of the electrodes present will be screen printed, and in these instances, printing will often be onto a single substrate. Often the screen printing will be onto a glass, ceramic or plastics substrate; for instance a plastics substrate may be polyester, polyvinyl chloride or polyvinyl carbonate. Often the substrate will be non-permeable, and it is generally desirable that the substrate be inert, at least to the components of the electrode and to any electrolyte present. As used herein the term "inert" is intended to encompass any material which does not react with the components of the sensor or apparatus with which it is in contact, or with the components of the sensor.

It can be advantageous if the substrate be flexible, so that a flexible electrode assembly is produced, this may be achieved by printing the electrode assembly onto a plastics film, such as a polyester film. A flexible electrode assembly can be of use as it can be moulded into a wider range of configurations, and hence housed in a wider range of sensor housings, than electrode assemblies which are printed onto rigid substrates.

Where the electrodes are screen-printed electrodes, these will be prepared from inks formulated such that when dried and/or cured they may carry electrons. As with non- screen printed electrodes for use in the invention, these materials may be conducting or semi-conducting as would be understood by the skilled reader. The sensor will typically have a limit of detection of the ammonia in the range 1 - 3 ppm, often around 1.5 - 2.5 ppm, often around 2 ppm. At higher levels, saturation of the signal can be prevented by dilution of the sample. This allows for the detection of ammonia at the levels typically observed in individuals with ailments where ammonia on the breath is a useful biomarker.

Often the sensor will further comprise an electrolyte. This could be potassium chloride or sodium chloride based. The electrolyte is generally present in aqueous solution at concentrations sufficiently high to ensure that electron flow within the sensor is not impeded (i.e. that resistance to electron flow is minimised). Nonaqueous electrolytes may also be used. Typical electrolyte concentrations would be in the range 0.1 - 10M, often 0.5 - 5M, often 1 - 2M. Contact between the electrodes can be provided by including a thin solution of electrolyte within the casing. In this context, by "thin" is meant a layer in the range 2 mm - 10 nm, 1mm - 10 μπι, 0.5 mm - 0.1 mm or combination thereof. It would be understood by the person skilled in the art that the thickness of the electrolyte layer will depend upon the precise design of the electrode assembly within the sensor. If desired, a liquid electrolyte may be contained within a suitable matrix, such as a thin strip of absorbent material, e.g. a gauze or mesh. Otherwise this may be provided in the form of a "free" electrolyte. Where the electrode materials themselves comprise a mesh or matrix of fibres (e.g. electrospun fibres), or otherwise comprise a suitable mesh or matrix-like layer or coating, these may aid in maintaining the electrolyte in contact with the active portion of the electrodes. Where the electrode material is encased, this may be provided in the form of a suitable mesh or matrix capable of retaining the electrolyte, this layer of coating is non-electroactive.

Ionic liquids may also be used as electrolyte materials and for this purpose may be encased within a suitable inert substrate which serves to retain the liquid in contact with the electrodes. Such materials are either sufficiently porous to retain the liquid or can be granulated to provide a material with suitable pores. Zeolites and clays are one example of commercially available materials for this purpose. Other support materials include metal oxides such as titanium oxide, aluminium oxide, zirconium oxide, silicon dioxide and mixtures thereof such as silica-alumina. Ionic liquids contain essentially only ions and are salts with relatively low melting points, e.g. below 100°C. A wide range of ionic liquids or mixtures thereof may be used in the sensor of the invention. Suitable room temperature ionic liquids generally consist of bulky organic cations such as l-alkyl-3-methylimidazolium, 1-alkylpyridinium, N-methyl-N- alkylpyrrolidinium, ammonium cations (e.g. tetraalkyl ammonium) and phosphonium cations (e.g. tetraalkylphosphonium). A wide range of anions may be employed from simple halide ions (e.g. F, CI, Br) to inorganic anions such as tetrafluorob orate and hexafluorophosphate and to larger organic anions such as triflate or tosylate. Ionic liquids comprising a l-alkyl-3-methylimidazolium cation are often used in which the alkyl group is preferably Ci-6 alkyl, e.g. Ci alkyl. l-butyl-3-methylimidazolium (bmim) is a particularly commonly used cation which may be present in the form of the tetrafluorob orate salt, [bmim] [BF4], or the hexafluorophosphate salt, [bmim] [PF₆]. An advantage of using ionic liquids is that these can be applied to the electrodes in a very thin layer or coating (e.g. of the order of nanometres) which provides a sensor with a fast response time (the thickness of the layer determines the rate at which the solution comes into contact with the working electrode).

In some cases, a solid electrolyte precursor may be provided which contacts the electrodes. On contact with the gaseous stream of exhaled breath, the water vapour present in the breath serves to hydrate the precursor to form an electrolyte. Examples of such materials include water-absorbing polymers such as super absorbent polyacrylate polymers (SAPs) which are cross-linked. One example of such a polymer material is sodium polyacrylate. Suitable electrolyte precursors also include gels or gel-like materials.

The electrolyte may additionally comprise a buffer, often an alkaline buffer (such as a boric acid buffer), to ensure that the pH within the sensor remains stable. It has been found, that the formation of the thioisoindole is favoured in alkaline solution, and so an alkaline buffer will often be used. By this we mean that the buffer solution will typically be a solution having a pH of greater than 7 which resists a change in pH on acidification, basification or dilution. Typically, the buffer solution used in the invention are in the range of, 8 to 13, , and most typically in the range 10 to 11. The buffer solution is not especially limited to any particular buffer solution and may be selected from any having the desired pH known to the skilled person, although often a boric acid buffer will be used because it is particularly stable with o-phthalaldehyde and thioglycolic acid (the OPA reagent). The electrolyte may further include an internal reference material, an electrochemically active compound present in a known concentration. The presence of the internal reference facilitates the determination of the concentration of the biomarkers present as it allows comparison of the signal obtained from the biomarker with that of the internal reference. The internal reference may be a compound such as a metallocene, anthracene or quinone as described above, or an alternative electroactive compound. The internal reference could be in solution as described above, or, alternatively, the internal reference could be immobilised within the electrode assembly. In a second aspect of the invention, there is provided an apparatus comprising: the electrochemical sensor of the first aspect of the invention;

a housing for the electrochemical sensor, the housing containing o-phthalaldehyde and thioglycolic acid; and

a receiver configured to allow the passage of breath into the housing

The sensor of the invention is inexpensive to prepare, and can therefore be discarded after a limited or single number of uses if appropriate, thereby avoiding problems of contamination or poisoning of the electrode or the need to recalibrate the sensor between uses. The removal of the need to recalibrate the sensor reduces operating costs and the level of expertise required to use the sensor. As a result, often the apparatus will be single or limited use.

A semi-permeable membrane may be employed within the sensor to screen out cationic and/or anionic species which may be encountered in the sample and thus prevent these from penetrating to the working electrode. This membrane will typically comprise corresponding anionic and/or cationic groups capable of binding to such undesired species and can be specifically tailored to screen out unwanted gases such as sulphur compounds. The semi-permeable membrane may, for example, be cellulose acetate or a conventional dialysis membrane. Other suitable membrane materials include nafion, polyvinyl sulphonate, carboxymethylcellulose, diethylaminoethylcellulose, polylysine and sulphonated polymers. The sensor will generally comprise a protective casing in which the electrode assembly is provided. Where an electrolyte is provided in the form of a liquid, this casing will also serve to retain this liquid without leakage, i.e. it will provide a suitable seal. The precise shape and dimensions of the casing are not critical to the operation or performance of the sensor provided that this is of sufficient capacity to enable the gas to be sampled by the electrode assembly. Often the electrolyte will comprise an alkaline buffer to promote capture of the ammonia through formation of the thioisoindole.

The housing may be fabricated from a plastics material, metal, glass or other material such as a paper product. Where the housing is intended to be a multi-use component of the apparatus, it will often be formed from a material selected from plastics, metals or glass; however, paper products may also be used as may a combination of the above. Where the housing is intended to be a single use component of the apparatus, it will often be formed from a plastics material or a paper product, such as cardboard. It will generally be the case that the housing be formed from a rigid material, often this material will be inert, or non-reactive to the components in exhaled breath.

The housing may be configured in a wide variety of ways so that breath passing from the receiver into the housing is held in contact with the sensor for sufficient time to be analysed. As will be understood, the precise conformation of the housing may be modified as appropriate to provide an ergonomic product.

Typically the receiver will be a mouthpiece, although it may also be a member configured to receive breath exhaled through the nose. The receiver may be fabricated from a wide range of materials, such as plastics, metals, glass or paper products. Where the receiver is a multi-use component, it will often be formed from metals, glass or plastics materials, most often plastics materials as these may easily and inexpensively be moulded to provide a receiver which is ergonomic and safe to use. For instance, a receiver with smooth contours may be beneficial as it would fit well with the mouth or nose and there would be no risk of harm to the user from sharp edges. Alternatively, where the receiver is intended to be a disposable component of the apparatus, whether alone as the only disposable component of the apparatus or with the sensor and housing as the disposable unit described above, it will generally be fabricated from a paper product, such as cardboard.

Typically, a reader is connected to the sensor for displaying results produced by the electrochemical sensor. The reader may be detachable and this will be the case where the apparatus is a limited use or more typically a single use unit. In such cases, the apparatus will be reversibly connectable to the reader. Limited use refers to a number of uses fewer than the reader. When a sample is administered into the chamber of the device, many (if not substantially all) interferent ions precipitate out of solution and, although they will be less visible to the sensor, will remain in the chamber as contaminants. As such, having a detachable and readily disposable apparatus which can be replaced prior to each measurement ensures that each sample is not contaminated.

Further, the single use unit may be connected to the reader such that once removed any particular single use unit combination may not be replaced, hence preventing inadvertent reuse of a soiled unit. Alternatively, the single use unit may include a fuse component containing a material which degrades under conditions of controlled passage of charge to break the electrical contact and prevent re-use of the single use unit. Connection methods which may be used to connect the limited or single use unit to the reader include, reversible clip techniques such as snap fit configurations, adhesive, friction fit, screw fit, hook and loop fixings, the use of screw or tack fixings or combinations thereof. The reader may be connected to the sensor using any of a wide range of conventional techniques known to the skilled person. Connection will typically be reversible to allow for disposal of the limited use or single use unit. Reversibility of the connection between the reader and allows for the reader to be retained when the apparatus is discarded, which may be advantageous as the reader will typically be a significantly higher cost item than the reader, will readily be capable of multiple uses and is not prone to contamination in the same manner as the apparatus.

The reader will often be a conventional product, typically the reader will be electronic and will operate using software which can convert the current output from the sensor into a results output which can be interpreted by an operator, or by the user themselves, and utilised in performing a diagnosis. Additionally or alternatively, the reader may include software which allows the reader to offer guidance as to the diagnosis. For instance, the reader may be programmed to offer a percentage likelihood that Candida is present, or a simple "yes/no" indicator. The software may be configured to measure the magnitude of the current, particularly where cyclic voltammetry is used. A change over a threshold value would be indicative of the presence of the biomarkers. Chronoamperometry may also be used, in which case the change in current with time is measured. In some embodiments, chronoamperometry is desirable as it is easier to monitor than cyclic voltammetry.

In a third aspect of the invention, there is provided a method of electrochemically detecting ammonia in breath comprising the steps of: a. reacting o-phthalaldehyde with thioglycolic acid to produce a

thioacetal intermediate;

b. reacting the thioacetal with ammonia in breath to form thioisoindole; c. optionally applying a potential; and

d. measuring a current generated.

As described above by reacting the ammonia with o-pthalaldehyde and thioglycolic acid to produce first the thioacetal intermediate, and then the thioisoindole, the presence and concentration of ammonia can be indirectly measured using this electroactive species. This provides for a reliable determination of the ammonia present in the breath. As one molecule of thioisoindole is formed per molecule of ammonia in this reaction, the method may comprise a further step of calculating the concentration of the ammonia. It is generally preferred that thioisoindole is formed alkaline pH as this encourages the reaction to proceed to completion more rapidly by encouraging the formation of the thioacetal. The classic reaction conditions for the OPA reagent are a ratio of thioglycolic:o-pthalaldehyde acid of around 2: 1 to ensure that only the single thioacetal is formed and not a diacetal as the formation of the diacetal would hinder the ring-forming reaction with ammonia. Therefore, the ratio of o-pthalaldehyde:thioglycolic acid used in the invention will typically be in the range 0.5: 1 to 5: 1. However, it has been found that a higher ratio offers a more rapid reaction rate and so often the ratio of o-pthalaldehyde:thioglycolic acid will be in the range 3.5: 1 to 4.5: 1, often around 4.

The method of this aspect of the invention may be used to diagnose a variety of illnesses including kidney disease, liver cirrhosis, stress, infection with fungi such as Candida, or infection with bacterium such as helicobacter pylori. The breath may be the breath of a human or mammal, although typically the sensor and apparatus of the invention are intended for human use. In most examples a potential will be applied in step c) as part of the detection method of the invention; however, this step is optional as where the electrode has been functionalised with catalytic materials which can interact with the biomarkers and generate a change in potential as a result of this interaction, no potential need be applied. This has energy saving benefits.

The breath tested using the electrochemical sensor according to the first aspect of the invention will typically comprise one or more interferents. These interferents may include oxygen, carbon monoxide, carbon dioxide, methane, nitric oxide, nitrogen, hydrogen sulphide, nitrous oxide, acetone, isoprene and combinations thereof. It has been found, that using the sensor of the first aspect of the invention, ammonia can be detected despite the presence of interferents.

cThe electrochemical potential applied between the electrodes is in the range of -2.0V The electrochemical potential applied between the electrodes is in the range of -2.0V to 2.0V, often -1.0 to 1.0V and the invention is typically capable of detecting and measuring current between the ranges of -5.0xl0^"9A and 1.0xlO^"4A. The application of a potential within these parameters leads to the favourable oxidation of thioisoindole. This transition causes a measurable current typically within the above mentioned values which corresponds to this transition. As such, the concentration of ammonia can be calculated as it is proportional to the current flowing from the one electron oxidation of thioisoindole. The time taken to determine a concentration of ammonia is usually in the range of 5 seconds to 5 minutes, often 30 to 120 seconds.

In a fourth aspect of the invention, there is provided a method of manufacturing an electrochemical sensor according to the first aspect of the invention, the method comprising the steps of: a) applying a first ink to a substrate;

b) drying the first ink to form a working electrode;

c) applying a second ink to the substrate;

d) drying the second ink to form a reference electrode; and

e) creating electrical connections.

The application of the first and second inks to the substrate may independently be by a method selected from ink jet printing, pad printing, screen printing, sputter coating, chemical vapour deposition, electrochemical plating, laser jet printing, dipping, brush application, spray application or a combination of these. The method selected must be capable of applying a film of ink. It is desirable that the inks be applied in an even layer, and for this reason ink jet printing, laser jet printing and/or screen printing are preferred because these printing techniques offer a reliable, yet rapid, method of ink deposition.

It is generally preferred that the electrochemical sensor/electrode assembly be formed by screen printing, although as described above, other types of electrodes may be used. However, the use of screen printing provides a greater flexibility of sensor design, and a high level of control over the precise arrangement of electrodes relative to one another. In addition, screen printed electrode assemblies are physically small, aiding the miniaturisation of the sensor, and allowing the production of an apparatus which is small and hence portable. This is of particular importance where the apparatus is intended to be a single-use apparatus, or where the chamber and sensor form a disposable unit, as the operator may wish to carry many of these units with them at one time. The apparatus of the invention is intended to be small enough to fit into a pocket. Finally, as screen printed electrodes are very simple, they are inexpensive to produce, and may be manufactured quickly and in high volume relative to other types of electrode.

Once the first ink has been applied, the ink is dried, often cured. Drying/curing may be at a temperature in the range 50°C - 150°C, often in the range 55°C - 100°C or 60°C - 80°C. Depending upon the ink formulation, each ink may be dried or cured at a different, or the same temperature, for differing or the same periods of time. Drying/thermal curing periods are typically of the order of minutes rather than hours, for instance, drying/curing may occur over a time period in the range 10 - 60 minutes, often 20 - 45 minutes, often around 30 minutes, for instance 25 - 35 minutes. The heat is supplied by conventional means, for instance in a fan oven, although other methods may be used as would be known to the person skilled in the art.

In a final step of sensor formation, electrical connections are created. This is often performed by application of a third ink which is also dried and/or cured as described above.

In some examples, the application of the initial step may additionally form a counter electrode. It can be advantageous to form the counter electrode in a single step with the working electrode as this is more efficient, and costs are reduced. This is particularly so where the working and counter electrode are formed from the same material.

In order to improve the effectiveness of the sensor, the gap between the working and counter electrode and any reference electrode and the counter electrode, can be kept to a minimum.

Typically, each electrode will comprise a single layer of the desired electrode material. However, multi-layered electrodes may also be employed in which each layer is applied to the substrate material using the same or different coating techniques. For example, the working electrode may comprise a layer of graphite particles coated with particles of an electrocatalytic metal such as platinum. Such particles may be held in place by a polymeric binder or otherwise linked to the surface of the graphite layer using known chemical linking agents. Alternatively, the nanoparticles may be dispersed in a coating on the electrode. Those portions of the electrodes which are not intended to come into contact with the sample may be provided with an inert coating in order to improve the electrical insulation of the electrodes. This coating will generally comprise an insulating dielectric layer which leaves exposed only the active portions of the electrodes. As used herein, the term "inert" is intended to encompass any material which does not react with the components of the sensor or apparatus with which it is in contact, in particular, the inert substrate should not interact with components of the solution, or with components of the sensor.

In a fifth aspect of the invention there is provided the use of a sensor according to the first aspect of the invention, in the detection of ammonia. Typically, the use will be in the quantitative detection of ammonia.

In a sixth aspect of the invention there is provided a breath testing kit comprising the electrochemical sensor according to the first aspect of the invention. This kit may be disposable, or multi-use. The kit may comprise the sensor and housing alone, with a receiver, or typically the apparatus comprising the sensor will be used in combination with a reader and optionally instructions for connection to the reader and use. As such, the kit may additionally comprise instructions for use. Alternatively, the kit may comprise the electrochemical sensor of the first aspect of the invention, a housing containing o-phthaladehyde and thioglycolic acid, a receiver and or optionally a buffer which may also contain instructions for use. The instructions for use may be simple assembly instructions, they may include instructions for taking the concentration measurement, they may include guide levels for concentrations which are indicative of the presence of malaise in the patient, or a combination of these. Such a kit would provide the user and operator with a portable, rapid, easy to use point-of-use tool for assessing whether the concentration of ammonia in the breath was within expected parameters. Unless otherwise stated each of the integers described in the invention may be used in combination with any other integer as would be understood by the person skilled in the art. Further, although all aspects of the invention preferably "comprise" the features described in relation to that aspect, it is specifically envisaged that they may "consist" or "consist essentially" of those features outlined in the claims.

Further, in the discussion of the invention, unless stated to the contrary, the disclosure of alternative values for the upper or lower limit of the permitted range of a parameter, is to be construed as an implied statement that each intermediate value of said parameter, lying between the smaller and greater of the alternatives, is itself also disclosed as a possible value for the parameter.

In addition, unless otherwise stated, all numerical values appearing in this application are to be understood as being modified by the term "about".

Brief description of the drawings

The invention will now be described by way of example only, with reference to the following drawings.

Figure 1 is a linear sweep voltammogram showing the signal strength of three different electrodes graphite screen printed electrode (black dashed line); glassy carbon electrode (black line); boron-doped diamond electrode (grey dashed) together with a blank scan (grey and black dashed line). Conditions: 8.5 ppm of Ammonia (Ammonium Hydroxide solution), pH 10 boric acid buffer, OPA and thioglycolic acid solution using differing working electrodes: Scan rate: 50 mVs^"1;

Figure 2 is a linear sweep voltammogram showing the effect of time on ammonia detection. Conditions: 8.5 ppm of Ammonia (Ammonium Hydroxide solution), pH 10 boric acid buffer, OPA and thioglycolic acid solution deposited upon a screen printed graphite electrode surface with measurements being recorded at extended periods of time. Scan rate: 50 mVs^"1;

Figure 3 shows the effect of the OPA composition. Conditions: Ammonia (Ammonium Hydroxide solution) added to a pH 10 boric acid buffer and OPA solution over the range of 1.7 - 17 ppm, using a single screen printed electrode. Error bars are also included N = 3;

Figure 4 is a linear sweep voltammogram showing the effect of pH on the system. Conditions: 8.5 ppm of Ammonia (Ammonium Hydroxide solution), pH 10 boric acid buffer, OPA and thioglycolic acid solution deposited upon a screen printed graphite electrode surface with measurements being recorded different pH values. Scan rate: 50 mVs^"1;

Figure 5 shows the effect of pH on peak potential (Ep), as determined from the results of Figure 4;

Figure 6 is a linear sweep voltammogram recorded at differing ratios of o- pthalaldehyde (P) and thioglycolic acid (M): Conditions: 8.5 ppm of Ammonia (Ammonium Hydroxide solution), pH 10 boric acid buffer, OPA and thioglycolic acid solution deposited upon a screen printed graphite electrode surface. Scan rate: 50 mVs^"1;

Figure 7 is a linear sweep voltammograms recorded in the presence of different buffers: boric acid buffer (black line); phosphate buffer (grey line); and bicarbonate buffer (black dashed line). Conditions: 8.5 ppm of Ammonia (Ammonium Hydroxide solution), pH 1 1.24 boric acid, OPA and thioglycolic acid solution deposited upon a screen printed graphite electrode surface. Scan rate: 50 mVs^"1;

Figure 8 shows the effect of ammonia addition to peak current. Conditions: Ammonia (Ammonium Hydroxide solution) added over range 1.7 - 17 ppm, pH 1 1.24 boric acid buffer, OPA and thioglycolic acid solution deposited upon a screen printed graphite electrode surface. Scan rate: 50 mVs^"1. Error bars are also included N = 3;

Figure 9 shows the effect of ammonia addition to peak current. Conditions: Ammonia (Ammonium Hydroxide solution) added over range 0.43 - 8.6 ppm, pH 1 1.24 boric acid buffer, OPA and thioglycolic acid solution deposited upon a screen printed graphite electrode surface. Scan rate: 50 mVs^"1. Error bars are also included N = 3; Figure 10 is an overlay of Figure 8 and Figure 9;

Figure 1 1 is a linear sweep voltammogram showing signal detection with ammonia gas. Conditions: Ammonia gas bubbled through solution (100 ppm in N₂) for 30 seconds, pH 1 1.24 boric acid buffer, OPA and thioglycolic acid solution deposited upon a 3mm screen printed graphite electrode surface. Scan rate: 50 mVs^"1;

Figure 12 shows the change in current over time in the experiment of Figure 1 1 ; and Figure 13 is a linear sweep voltammogram showing the compatibility of the thioisoindole detection method with the presence of a lead (II) internal standard. Conditions: 3.4 ppm ammonia (Ammonium Hydroxide solution), pH 1 1.24 boric acid buffer, OP A and thiogly colic acid solution deposited upon a 3 mm screen printed graphite electrode surface. Scan rate: 50 mVs^"1; Deposition potential: -1.2 V; deposition time: 120 seconds.

Examples Test Methods

Reagent: OPA reagent - o-phthaldehyde with thioglycolic acid

Voltammetric measurements were carried out using a μ-Autolab III (Eco Chemie, The Netherlands) potentiostat/galvanostat and controlled by Autolab GPES software version 4.9 for Windows XP.

Fabrication of the screen-printed electrodes Screen-printed carbon macroelectrodes (SPEs) which have a 3 mm diameter working electrode were fabricated in-house with appropriate stencil designs using a DEK 248 screen printing machine (DEK, Weymouth, UK). For the fabrication of the screen-printed sensors, firstly, a carbon-graphite ink formulation (Product Code: C2000802P2; Gwent Electronic Materials Ltd, UK) was screen-printed onto a polyester (Autostat, 250 micron thickness) flexible film. This layer was cured in a fan oven at 60 degrees for 30 minutes. Next a silver/silver chloride reference electrode was included by screen printing Ag/AgCl paste (Product Code: C2040308D2; Gwent Electronic Materials Ltd, UK) onto the polyester substrates. Finally, a dielectric paste (Product Code: D2070423D5; Gwent Electronic Materials Ltd, UK) was then printed onto the polyester substrate to cover the connections. After curing at 60 degrees for 30 minutes the screen-printed electrodes are ready to be used. The reproducibility of the batch of screen-printed sensors were found to correspond to 0.76 % RSD using the Ru(NH₃)^2+/3+ redox probe in 1 M KC1. Selection of electrode material for detection of thioisoindole

Three electrode materials were tested, glassy carbon, boron-doped diamond and a screen-printed graphite electrode for their response to thioisoindole formed from ammonia (ammonia concentration of 8.5 ppm). The results are shown in figure 1.

Inspection of figure 2 reveals that the optimum and most desirable voltammetric response was achieved utilising the screen-printed graphite electrode. Such observations offer further advantages with the screen-printed sensor enabling the housing of all three essential components (working, counter and reference electrode) upon a single unit, rather than the requirement of bulky individual electrodes as is a necessity with the remaining electrode materials. Further to this, the screen-printed sensor also allows for the potential incorporation of the reagent materials within the electrode itself further simplifying the analytical procedure.

The effect of time

As the reaction of o-phthaldehyde with thioglycolic acid and then ammonia to form thioisoindole is a CE reaction, it is important that the role played by the effect of the reaction has upon the subsequent electrochemical measurement. To do so an aliquot of a solution containing ammonia and the OPA reagent was deposited upon the surface of a screen printed graphite electrode. Linear sweep voltammetric measurements were then undertaken at given time periods.

From figure 2 it was determined that no sufficient change in the observed voltammetric peak current occurred as a result of time indicating that the reaction occurring between the OPA reagent and the ammonia is a rapid one. As a result it was decided that further tests would be made after an initial period of 30 seconds after the addition of ammonia.

The effect of concentration The detection limit and linear range offered through the utilisation of the initial OPA regent towards the sensing of ammonia was determined. Ammonia (ammonium hydroxide solution) was added over the range of 1.7 - 17 ppm to a boric acid buffer of pH 10 containing the OPA reagent. As is shown in figure 3, an initial linear increase in the observed voltammetric peak current is observed until an ammonia concentration of - 13.6 ppm at which point a deviation from linearity is clear with the occurrence of an apparent plateau in peak current with relation to ammonia concentration. Current flow was observed as low at 1.7 ppm, indicating that ammonia detection can be achieved using these systems at levels well within those observed in human breath.

Solution pH

The optimum pH for the formation of the electroactive product was determined. This was done using a set ammonia concentration of 8.5 ppm in solution with the pH of the boric acid solution containing the OPA reagent being modified.

Inspection of figure 4 reveals that the oxidation of the reactant occurs most easily at alkaline pH, in particular at high alkaline pH such as between 11 and 12. The optimum measured pH was 11.24. Additionally, a greater voltammetric peak current is also noted at this solution pH. Consequently subsequent electrochemical measurements were carried out at a solution pH of 11.24.

Using figure 5 it is possible to calculate the number of electrons and protons involved in the electrochemical reaction. It was determined that the reaction being studied involved 1 proton and 1 electron. o-phthalaldehyde and thioglycolic acid ratio

The OPA reagent comprises two chemicals o-pthalaldehyde and thioglycolic acid. The optimum reported ratio of the two for a reaction with an amine is one of 2: 1 for thioglycolic acid : o-phthalaldehyde. To further examine this relationship differing ratios of the two chemicals were utilised with an addition of 8.5 ppm ammonia in solution being made to each and the resulting voltammetry observed. Varying ratios of the two chemicals were utilised, as is shown in figure 6. Additionally on OP A reagent was made utilising the commonly used ratio but in a greatly increased concentration of the two chemicals. The resultant voltammograms unexpectedly highlight that the use of more o-phthalaldehyde than thioglycolic acid is preferable for this application, with the optimum ratio being 4: 1 for o-pthalaldehyde : thioglycolic acid, as this yielded the optimum voltammetric peak current.

Buffer selection A range of buffers were tested as shown in figure 7. Boric acid, bicarbonate and phosphate buffers were tested. A boric acid buffer is seen to offer the most desirable voltammetry.

Optimised calibration plot

After determining the optimum pH and ratio of o-pthalaldehyde and thioglycolic acid additions of ammonia (ammonium hydroxide solution) over the range of 1.7 to 17 ppm were once more added to a boric acid buffer of pH 11.24 containing the OPA reagent. As is shown in figure 8, an initial linear increase in the voltammetric peak current is observed until an ammonia concentration of - 11.9 ppm at which point a deviation from linearity is clear with the occurrence of an apparent plateau in peak current with relation to ammonia concentration. Current flowed with concentrations as low as 1.7 ppm. To further assess the absolute limits of resolution towards the sensing of ammonia utilising linear sweep voltammetry and the graphite screen printed electrodes additions of ammonia (ammonium hydroxide solution) over the range of 0.43 to 8.6 ppm were made to a boric acid buffer of pH 11.24 containing the OPA reagent. As is clear in figure 9, an initial linear increase in the observed voltammetric peak current is observed until an ammonia concentration of ~ 6.02 ppm at which point a deviation from linearity is clear with the occurrence of an apparent plateau in peak current with relation to ammonia concentration. When comparing both the high concentration range with the low concentration range (figures 8 and 9) as is depicted in figure 10, it is clear that the same behaviour/trend is noted over each concentration range. Such observations would indicate an important role/change occurring at the working electrode. It can be determined that such observations are not in relation to the OPA reagent and corresponding ring closure reaction being saturated or complete as if such was the case the given trend would occur at a particular ammonia concentration.

Ammonia gas

The above tests show the viability of the test method for ammonia in solutions. However, for a breath sensor, it is necessary to determine the transferability of the protocol into the detection of gaseous ammonia. Once more the optimised OPA reagent was utilised with ammonia gas (100 ppm in N₂) being bubbled into the reagent with the resultant linear sweep voltammogram depicted in figure 11.

The protocol utilised for the sensing of ammonia in solution is readily transferable towards the monitoring of ammonia gas. As such the effect of bubbling time upon the observed voltammetric peak current was determined over a period of 5 minutes.

Interference study

The potential interferents present within breath were studied with the effect of their presence upon the standard voltammetric response observed for the OPA reagent. This was done to ensure that any potentially electrochemically active constituents did not exhibit a voltammetric peak in the same potential region at that of the thioisoindole, and to verify that the potential interferents do not partake in the ring closure reaction with the OPA reagent and as such give 'false positive' electrochemical responses. The gaseous interferents were bubbled into a solution of OPA reagent of pH 1 1.24 for 5 minutes individually. A new OPA reagent solution and screen printed graphite electrode was utilised for each gas. Table 1 describes the effects of the gas upon the observed electrochemical response, whereby it was determined that the gases trialled did not alter the voltammetric profile. Table 1: The findings arising from electrochemical measurements obtained after the bubbling of prescribed interferents into a solution of OP A reagent.

Further to this both acetone and isoprene; both of which can be present in breath samples were studied in to determine their effect and/or reaction with the OPA reagent. High concentrations of each (5 mM) were added to separate OPA reagents with the consequent voltammetric response observed. It was determined that both acetone and isoprene did not interfere or interact with the electrochemical procedure at hand.

As can be seen, the test protocol gives an unadulterated and accurate response in the presence of the common interferants, clearly showing that it can be viably used as a breath testing method.

Utilisation of an internal standard

In line with the potential requirement for the inclusions and/or utilisation of an internal standard it was decided to determine the potential interference or overlap of the desired internal standard lead (II). As is depicted in figure 13 excellent separation is observed between the peak potentials for the stripping of lead (II) and the product from the OPA reagent and NH3 reaction. It should be appreciated that the sensors, apparatus, methods and uses of the invention are capable of being incorporated in the form of a variety of embodiments, only a few of which have been illustrated and described above.

Claims

1. An electrochemical sensor for detecting ammonia in breath.

2. A sensor according to claim 1, wherein the detection comprises indirect detection of an electroactive derivative of ammonia.

3. A sensor according to claim 2, wherein the electroactive derivative of ammonia is thioisoindole.

4. A sensor according to any preceding claim, comprising an electrode assembly which comprises at least two electrodes, one of said electrodes being capable of detecting thioisoindole.

5. A sensor according to claim 4, wherein the electrode assembly comprises at least one working electrode capable of detecting thioisoindole, a counter electrode and a reference electrode.

6. A sensor according to claim 5, wherein the working electrode is selected from a glassy carbon electrode, a boron doped diamond electrode and a screen-printed electrode.

7. A sensor according to claim 6, wherein the working electrode is a screen-printed graphite electrode.

8. A sensor according to any one of claims 4 to 7, wherein one or more of the electrodes are metallic or non-metallic nanoelectrodes, microelectrodes or combinations thereof.

9. A sensor according to any preceding claim, additionally comprising an internal reference material.

10. A sensor according to any preceding claim, further comprising a boric acid buffer.

11. An apparatus comprising:

the electrochemical sensor according to any preceding claim;

a receiver configured to allow the passage of breath into the housing

12. An apparatus according to claim 11, wherein the housing additionally contains an alkaline buffer.

13. An apparatus according to claim 11 or claim 12, further comprising a reader connected to the sensor for displaying results produced by the electrochemical sensor.

14. An apparatus according to claim 13, wherein the apparatus forms a single use unit which is reversibly connectable to the reader.

15. A method of electrochemically detecting ammonia in breath comprising the steps of:

reacting o-phthalaldehyde with thioglycolic acid to produce a thioacetal intermediate;

reacting the thioacetal with ammonia in breath to form thioisoindole; applying a potential; and

measuring a current generated.

16. A method according to claim 15 further comprising a step of calculating the concentration of ammonia.

17. A method according to claim 15 or claim 16, wherein the thioisoindole is formed at alkaline pH.

18. A method according to any of claims 15 to 17, wherein a ratio of o-phthalaldehyde : thioglycolic acid is in the range 3.5: 1 to 4.5: 1.

19. A method of manufacturing an electrochemical sensor according to any of claims 1 to 10 comprising the steps of:

f) applying a first ink to a substrate;

g) drying the first ink to form a working electrode;

h) applying a second ink to the substrate;

i) drying the second ink to form a reference electrode; and

J) creating electrical connections.

20. A method according to claim 19, wherein the electrochemical sensor is formed by screen-printing.

21. Use of a sensor according to any of claims 1 to 10, in the quantitative detection of ammonia.

22. A kit comprising the sensor according to claims 1 to 10, for detecting ammonia in breath.

23. A kit according to claim 22, further comprising a reader, a housing, a receiver or combinations thereof.

24. An electrochemical sensor, apparatus and kit substantially as described herein with reference to the drawings.