System and method for multi-gas detection at multiple operating temperatures

Abstract

Systems and methods for multi-gas detection using dielectric excitation of a single sensing material at multiple operating temperatures. By measuring dielectric excitation responses in the form of impedance, admittance, capacitance, and reactance of the gas sensing material, but preferably not in the form of resistance, less operating temperatures are used than those used by the same MOS gas sensing material configured to perform multi-gas discrimination based on resistance response alone, enabling multi-gas discrimination and improved discrimination. The gas sensor and gas sensing technology of the present disclosure enable improved response linearity, improved dynamic range, and reduced computational resource consumption with respect to multi-gas quantification compared to conventional resistance-based gas sensing methods. This embodiment unexpectedly demonstrates a MOS-based gas sensor capable of discriminating different gases using responses collected at at least two different operating temperatures, and this discrimination is superior in discrimination between different gases and baseline stability compared to the resistance responses of the same gas sensing material at three or more operating temperatures.

Description

Technical Field

[0001] Statement Regarding Research and Development Funded by the Federal Government This invention was made with government support under contract No. 75D30118C02617 awarded by the Centers for Disease Control and Prevention. The government has certain rights in this invention.

[0002] The subject matter disclosed herein generally relates to gas sensing, and more particularly, to gas sensing using metal oxide semiconductor (MOS) sensors.

Background Art

[0003] Metal oxide semiconductor (MOS) sensors can be operated by measuring a single sensor response or sensor output such as resistance. These MOS chemiresistors are popular due to their ability to non-selectively detect a number of gases by appropriately selecting the base semiconductor material and doping material. In such gas-responsive chemiresistors, a change in the resistance of the MOS sensing element is measured, and this change in resistance is proportional to the gas concentration in the fluid sample. However, the limited discrimination of MOS gas-responsive chemiresistors hinders the use of such sensors in certain multi-gas detection applications. Further, to achieve at least some degree of gas discrimination, conventional MOS-based gas sensors are required to perform resistance measurements at two or more operating temperatures, particularly at least two different operating temperatures, which shortens the lifespan of the sensor and may have an undesirable effect on electrical measurements. Further, for multi-gas analysis, conventional MOS gas sensors typically need to collect resistance measurements at multiple different operating temperatures (e.g., 3, 4, 5 or more temperatures), which results in a delay in operation waiting for each thermal cycle to complete, shortens the lifespan of the sensor, and may have an undesirable effect on electrical measurement values. Further, even when several different operating temperatures are utilized when performing such resistance measurements, conventional MOS sensors struggle to effectively discriminate between different gases in a fluid sample due to undesirable issues associated with resistance measurements such as baseline instability and non-linear response patterns.

Summary of the Invention

Problems to be Solved by the Invention

[0004] With the above in mind, the present embodiment relates to a system and method for multi-gas detection using dielectric excitation of a sensing material that uses at least two different operating temperatures. Embodiments of the gas sensors disclosed herein implement a metal oxide semiconductor sensing material that can be switched between different operating temperatures, and while the sensing material is exposed to a fluid sample, the dielectric excitation response of the sensing material is measured at each of these operating temperatures. The gas sensors and gas detection methods of the present disclosure unexpectedly provide desirable characteristics such as improved discrimination of multiple gases using fewer temperature switching steps compared to conventional MOS resistance-based gas sensors and gas detection methods. For example, by measuring the dielectric excitation response using a single gas sensing element that can be switched between two operating temperatures, embodiments of the present disclosure demonstrate superior multi-gas discrimination compared to other gas sensors and other gas detection methods that rely on resistance measurements collected at several (e.g., three or more) different operating temperatures.

Means for Solving the Problems

[0005] In one embodiment, a gas sensor system for multi-gas analysis of a fluid sample includes a gas sensing element configured to operate at a plurality of temperatures and configured to contact the fluid sample, a heating element coupled to the gas sensing element and configured to heat the gas sensing element, and a heater controller operably coupled to the heating element and configured to control the heating element to heat the gas sensing element to each of the plurality of temperatures while the gas sensing element is in contact with the fluid sample. The gas sensor system also includes a measurement circuit operably coupled to the gas sensing element and configured to cause dielectric excitation in the gas sensing element and measure the dielectric excitation response of the gas sensing element while the gas sensing element is heated to each of the plurality of temperatures and in contact with the fluid sample, and the measured dielectric excitation response improves discrimination between at least two gases in the fluid sample and improves response linearity to at least two gases in the fluid sample as compared to the resistance response of the gas sensing element when in contact with the fluid sample at each of the plurality of temperatures.

[0006] In one embodiment, a method of operating a gas sensor for multi-gas analysis of a fluid sample includes exposing a gas sensing material of the gas sensor to the fluid sample; measuring, by a measurement circuit of the gas sensor, a first set of dielectric excitation responses of the gas sensing material while the gas sensing material is heated to a first temperature and exposed to the fluid sample; measuring, by the measurement circuit of the gas sensor, a second set of dielectric excitation responses of the gas sensing material while the gas sensing material is heated to a second temperature and exposed to the fluid sample; and receiving, by an on-board data processor of the gas sensor, the first set of dielectric excitation responses of the gas sensing material at the first temperature and the second set of dielectric excitation responses of the gas sensing material at the second temperature. The method also includes decomposing at least two gases in the fluid sample by the on-board data processor based on at least a portion of the first set of dielectric excitation responses and at least a portion of the second set of dielectric excitation responses. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] These and other features, aspects, and advantages of the present invention will be better understood when the following detailed description is read with reference to the accompanying drawings. In the accompanying drawings, like reference numerals represent like parts throughout the drawings.

Figure 1

Figure 2

Figure 3

Figure 4A

Figure 4B

Figure 5A

Figure 5B

Figure 6

Figure 7

Figure 8A

Figure 8B

Figure 9

Figure 10

Figure 11A

Figure 11B

Figure 11C

Figure 12A

Figure 12B

Figure 13A

Figure 13B

Figure 13C

Figure 14A

Figure 14B

Figure 14C

Figure 15A

Figure 15B

Figure 15C

Figure 16A

Figure 16B

Figure 16C

Figure 17

Figure 18A

Figure 18B

[0008] The present embodiment relates to a system and method for multi-gas sensing using dielectric excitation of a single metal oxide semiconductor sensing material arranged as a single gas sensing element. Note that metal oxide semiconductor sensing materials are often abbreviated in the industry as metal oxide semiconductor (MOS) materials or semiconductor metal oxides (SMOX) or semiconductor metal oxides (MOX) materials. Conventional MOS-based gas sensors measure only the direct current (DC) resistance response under given excitation conditions. Measuring a single response for each sensor under given excitation conditions is known as a single output response or a single output readout, and the sensor is known as a single output sensor. To distinguish multiple gases in a fluid sample, conventional MOS-based gas sensors measure the DC resistance response at several different temperatures.

[0009] However, by measuring the dielectric excitation response of a gas sensing material, it is currently recognized that discrimination and resolution improvement of multi-gases can be achieved using a lower operating temperature than that used by the same MOS gas sensing material configured to perform multi-gas discrimination based only on resistance response. Further, the gas sensor and gas sensing technology of the present disclosure enable improved response linearity, improved dynamic range, and reduced computational resource consumption with respect to multi-gas quantification as compared to conventional resistance-based gas sensing methods. Further, by reducing the number of operation temperature switching events, the present embodiment enables a gas sensor having improved measurement quality and extended operation life. That is, when using a set of predetermined operating temperatures, the dielectric relaxation spectrum of the gas sensing material is affected differently by different gases, and such desired differences are more prominent compared to the resistance response of the same gas sensing material even when additional operating temperatures are used in conventional resistance-based gas sensing methods. Therefore, the present embodiment unexpectedly demonstrates a MOS-based gas sensor capable of distinguishing different gases using dielectric excitation responses collected using at least two different operating temperatures, and this discrimination is excellent in discrimination between different gases and baseline stability compared to the resistance responses of the same gas sensing material at three or more operating temperatures.

[0010] As described above, conventional MOS-based gas sensors typically measure the DC resistance response of a gas sensing material at several different operating temperatures when performing multi-gas analysis of a fluid sample. In contrast, this embodiment of the gas sensor includes a MOS-based gas sensing material that interacts with gases in a fluid sample at two or more operating temperatures and provides an excitation response to a specific dielectric excitation frequency, and these dielectric excitation responses of the gas sensing material are measured and analyzed to distinguish two or more gases in the fluid sample. As used herein, the terms "analyte", "gas to be analyzed", or "fluid to be analyzed" refer to the target component in the measured fluid. As used herein, the terms "interferent", "interfering gas", "interfering fluid" refer to any component in the measurement fluid that may have an undesirable effect on the accuracy and precision of the measurement of the analyte using the sensor.

[0011] With the above in mind, FIG. 1 is a schematic diagram of one embodiment of a gas sensor 10 for multi-gas analysis of a fluid sample according to the present technology. In different embodiments, the gas sensor 10 may be a wearable gas sensor, an ingestible gas sensor, or a tattoo-type gas sensor for personal (e.g., patient) monitoring. In a particular embodiment, the gas sensor 10 may be an industrial environment sensor, an asset monitoring sensor, an industrial process monitoring gas sensor, a consumer sensor, a transportation sensor, a security sensor, or any combination thereof.

[0012] In the embodiment shown in FIG. 1, the gas sensor 10 generally includes at least one gas sensing element 12, a control circuit 14, and one or more output devices 16. Each gas sensing element 12 includes a substrate 18 on which sensing electrodes 20 are disposed, and a gas sensing material 22 (e.g., a properly formulated metal oxide semiconductor material (MOS) applied to form a gas sensing film) disposed on the substrate 18 between the sensing electrodes 20. In certain embodiments, the gas sensor 10 may include a plurality of gas sensing elements 12 (e.g., an array of gas sensing elements 12), such as gas sensing elements having different MOS-based gas sensing materials 22. In certain embodiments, there may be three or more sensing electrodes 20, and the sensing electrodes 20 may include a plurality of meshed sensing electrodes. It will be appreciated that the gas sensing material 22 is generally applied on the electrodes 20 to form a gas sensing film such that dielectric excitation of the gas sensing material 22 and measurement of the dielectric excitation response of the gas sensing material 22 are performed via the electrodes 20.

[0013] Furthermore, a resistive heating element 24 is disposed on the surface of the substrate 18 opposite the gas sensing material 22 and is designed to heat the gas sensing material 22 to an appropriate operating temperature during multi-gas analysis of the fluid sample 26. In certain embodiments, the heating element 24 may be disposed on the surface of the substrate 18 opposite the gas sensing material 22, but in other embodiments, the heating element 24 may be disposed on the same surface of the substrate 18 as the gas sensing material 22. In embodiments having a plurality of gas sensing elements 12, in certain cases, two or more gas sensing materials 22 may be applied to a common substrate to form a plurality of gas sensing elements 12 on a common substrate 18 that can be heated by a single heating element, but in other cases, each gas sensing material 22 may be disposed on each substrate 18 having its own heating element 24. Furthermore, in certain embodiments, the heating element 24 may be integrated with the substrate 18 as a monolithic structure.

[0014] During operation of the gas sensor 10, the gas sensing material 22 of the gas sensing element 12 is heated to each of a plurality of different operating temperatures when the gas sensing material 22 is exposed to a fluid sample 26 that may contain two or more gases (e.g., one analyte gas and one interfering gas, at least two analyte gases). Thus, the control circuit 14 of the illustrated gas sensor 10 includes a heater controller 28 that is electrically connected to the heating element and configured to control the heating element 24 to achieve each of the different operating temperatures. For example, in one embodiment of the gas sensor 10 designed to measure the excitation response of the gas sensing material 22 at a first operating temperature and a second operating temperature, the heater controller 28 provides a first voltage to the heating element 24 to heat the gas sensing element 12 to the first operating temperature when a first set of dielectric excitation responses of the gas sensing material 22 is being measured, and provides a second voltage to the heating element 24 to heat the gas sensing element 12 to the second operating temperature when a second set of dielectric excitation responses of the gas sensing material 22 is being measured. For example, in certain embodiments, the operating temperature of the gas sensing material 22 may be 30°C to 1000°C, 50°C to 900°C, or 80°C to 600°C.

[0015] In the illustrated embodiment, the sensing electrode 20 of the gas sensing element 12 is electrically coupled to the measurement circuit 32 of the control circuit 14 of the gas sensor 10. The measurement circuit 32 is designed to cause at least dielectric (AC) excitation of the gas sensing material 22 at a preselected frequency and measure the dielectric response (e.g., impedance response) of the gas sensing material 22 to these excitations. In certain embodiments, the measurement circuit 32 may further be capable of (or designed to) cause direct current (DC) excitation of the gas sensing material 22 and measure the DC response (e.g., resistance response) of the gas sensing material 22 to these excitations. In certain embodiments, the measurement circuit 32 may measure both the AC response and the DC response of the gas sensing material 22. However, in certain embodiments, the measurement circuit 32 may be designed to cause only dielectric excitation of the gas sensing material 22 and measure only the dielectric response of the gas sensing material 22.

[0016] As used herein, "dielectric excitation" of a MOS sensing material refers to AC excitation of the MOS sensing material at the shoulder of its dielectric relaxation region. As used herein, "impedance" is a non-limiting term for any electrical response of the sensing system to an alternating current (AC) applied to the gas sensing material 22. It will be understood that such responses may be measured as different electrical properties in different embodiments. Non-limiting examples of these electrical responses of the gas sensing material 22 to an alternating current include impedance, the real part of the impedance, the imaginary part of the impedance, admittance, reactance, susceptance, and the like. In this specification, an example of the response is shown as impedance, but other electrical responses of the gas sensing material 22 to AC excitation may be generated as well. In one embodiment, the electrical response of the gas sensing material 22 can be monitored at the gas-modulated high-frequency shoulder of the dielectric relaxation peak of the sensing material. In one embodiment, the electrical response of the sensing system can be monitored at the gas-modulated low-frequency shoulder of the dielectric relaxation peak of the sensing material.

[0017] The gas sensor 10 may represent one or more different versions of the multi-gas detection system described herein. In one or more embodiments, the measurement circuit 32 may include a resistor-capacitor (RC) electrical circuit including one or more resistor (R) components and capacitor (C) components, and these components may be electronically changed by the controller circuit 14 by the presence of one or more target analysis target gases. In one or more embodiments, the measurement circuit 32 can perform dielectric excitation and impedance measurement at one or more different frequencies or with one or more different RC configurations of the measurement circuit 32. For example, the measurement circuit 32 of the gas sensor 10 can measure the impedance response of the gas sensing material 22 at different frequencies, with different resistances of the RC electrical circuit of the measurement circuit 32, with different capacitances of the RC electrical circuit of the measurement circuit 32, or with any combination of two or more of these. The measurement circuit 32 causes excitation and enables measurement of the response of the sensing element to the gas. The measurement circuit is not designed to be affected by the concentration of the gas being measured. Instead, only the gas sensing element 12 is designed to be predictably affected by the concentration of the gas being measured.

[0018] The control circuit 14 of the illustrated gas sensor 10 includes a data processing unit 34 (also referred to herein as a data processing circuit) communicatively coupled to the measurement circuit 32 to receive the excitation response measured by the measurement circuit 32. The data processing unit 34 includes an on-board data processor 36 and includes a memory 38 that stores a gas analysis model 40, such as a gas classification model 42, a gas quantification model 44, or any combination thereof. These gas analysis models 40 are generally mathematical models that store the relationship between an excitation response (e.g., a dielectric excitation response) and a specific classification or concentration of a gas in a fluid sample. For example, the gas classification model 42 can store the relationship between the excitation response of the gas sensing material 22 at a specific operating temperature and a specific classification of the gas, while the gas quantification model 44 can store the relationship between the excitation response of the gas sensing material 22 at a specific operating temperature and a specific concentration of the gas. In certain embodiments, the gas analysis model 40 can include one or more coefficients having values determined experimentally and stored in the memory 38. In some embodiments, the number of target gases to be analyzed determined by the gas classification model 42, or the gas quantification model 44, or any combination thereof, for the illustrated gas sensor 10 can range from one target gas to 50 target gases.

[0019] A gas sensor element having two or more responses or outputs is called a multivariate gas sensor element. To analyze the output from the multivariate gas sensor element, multivariate data processing principles are applied. By applying the multivariate data processing principles, the diversity of the responses of the multivariate sensor to different gases can be quantified. A multivariate transfer function can be constructed to quantify different gases. The constructed multivariate transfer function can be implemented to quantify different gases in new measurement data from this multivariate gas sensor element. Non-limiting examples of multivariate data processing principles include methods for performing gas classification / cluster analysis and quantification. Classification / cluster analysis can be performed to accurately determine the type of gas to be analyzed. Quantification can be performed to accurately determine the concentration of the gas to be analyzed. Examples of classification / cluster analysis algorithms include, but are not limited to, principal component analysis (PCA), hierarchical cluster analysis (HCA), independent component analysis (ICA), linear discriminant analysis (LDA), and support vector machine (SVM) algorithms. Non-limiting examples of methods for performing analyte quantification to determine the concentration of a specific gas to be analyzed include principal component regression (PCR), independent component regression (ICR), non-linear regression analysis (NRA), discriminant function analysis (DFA), or artificial neural network analysis (ANN). In certain aspects of the subject matter of the invention described herein, a quantification algorithm can follow a classification algorithm.

[0020] As described below, the on-board data processor 36 receives the excitation responses measured by the measurement circuit 32, selects specific excitation responses (e.g., dielectric excitation responses) for analysis, provides these excitation responses as inputs to one or more of the stored gas analysis models 40, and the gas analysis model 40 returns an output that decomposes or differentiates two or more gases in the fluid sample 26. As used herein, "decomposing" two or more gases in a fluid sample, "providing resolution" between two or more gases in a fluid sample, "differentiating" two or more gases in a fluid sample, or "providing discrimination" between two or more gases in a fluid sample means determining a respective classification for each of the gases in the fluid sample, determining the respective concentration of each of the gases in the fluid sample, or determining both the respective classification and the respective concentration of each of the gases in the fluid sample. As used herein, "classifying" or "determining a classification" of a gas means determining the exact chemical identity of the gas (e.g., ethanol, acetone, hydrogen, carbon monoxide, methane, toluene, benzene), or determining the chemical class to which each gas belongs (e.g., hydrocarbons, oxides, sulfides, ketones, aromatic hydrocarbons, etc.).

[0021] In certain embodiments, the memory 38 may be integrated with the on-board data processor 36. In certain embodiments, the on-board data processor 36 is a multi-core processor. For example, in some embodiments, the on-board data processor 36 is a multi-core processor on a single integrated circuit having two or more distinct processing units (also called cores) each of which reads and executes program instructions. In certain embodiments, the multi-core processor may include only a single central processing unit (CPU) and a plurality of additional cores. In embodiments where the on-board data processor 36 is a multi-core processor, different gas analysis models and / or different signal processing algorithms can be independently executed by different cores to reduce the power consumption of the data processing unit 34 and / or the gas sensor 10.

[0022] In the illustrated embodiment, the gas sensor 10 includes one or more output devices 16. In certain embodiments, the output device 16 includes one or more display devices 46 configured to present information regarding multi-gas analysis, such as the classification and / or concentration of two or more gases in the fluid sample 26. In some embodiments, other output devices 16 may include alarms 49, such as visual alarms (e.g., light-emitting diodes (LEDs)), audible alarms (e.g., speakers), and / or tactile alarms (e.g., tactile feedback devices). In certain embodiments, the output device 16 includes one or more communication devices 48 (e.g., wired communication interfaces, wireless communication interfaces) that enable the gas sensor 10 to communicate with other computing systems, such as a desktop computer, a mobile computing device (e.g., laptop, smartphone), a remote server (e.g., Internet server, cloud server), or other sensors of a multi-sensor monitoring system (e.g., gas sensors, temperature sensors, vibration sensors, health monitors). For example, in certain embodiments, information determined by the on-board data processor 36 regarding the discrimination of two or more gases in the fluid sample 26 may be provided to an external computing system that functions as a controller for a mesh of sensors including the gas sensor 10. In some embodiments, the gas sensor 10 can additionally or alternatively use the communication device 48 to provide measurements of the excitation response to an external computing system, such that the external computing system can use these measurements to calculate one or more coefficient values for one or more of the gas analysis models and return these coefficient values to the gas sensor 10 for storage in the memory 38.

[0023] Furthermore, the illustrated gas sensor 10 includes a battery 50 that is electrically coupled to supply power to various components of the gas sensor 10, including a control circuit 14 and an output device 16. It will be appreciated that the battery 50 should have a capacity suitable for powering all components of the gas sensor 10. For example, this can include heating the gas sensing material 22, causing dielectric excitation in the gas sensing material 22, measuring the dielectric excitation response of the gas sensing material 22, analyzing the measured dielectric excitation response to distinguish two or more gases in a fluid sample, and presenting the results of the analysis via a suitable output device 16. In certain embodiments, the battery 50 can have a capacity sufficient to operate the gas sensor 10 for at least 10 hours. In some embodiments, the battery 50 can have a battery capacity of 1 milliampere-hour (mAh) to 500 mAh, or 1 mAh to 200 mAh, or 1 mAh to 100 mAh. In certain embodiments, such as those where the gas sensor 10 is designed to be particularly thin (e.g., for ingestible or tattoo-style embodiments of the gas sensor 10), the battery 50 can have a thickness of less than about 5 millimeters (mm). In some embodiments, all of the components of the gas sensor 10 may be coupled to or at least partially disposed within a package or housing suitable for a particular gas sensing application. For example, in the case of personal monitoring applications, the packaging of the gas sensor 10 can be made of a biocompatible polymer that can be worn externally, injected subcutaneously, or ingested to perform multi-gas analysis on an individual or patient.

[0024] The gas sensor 10 may be a wearable device that can be worn or moved by an operator from one location to another. The gas sensor 10 may be disposed within or be an integrated part of a helmet, hat, gloves, or other clothing item. For example, the gas sensor 10 may be held within a wearable or non-wearable transportable object such as the frame of military or industrial eyewear, a wearable pulse oximeter, a safety vest or harness, clothing items, a mobile device (e.g., a mobile phone, tablet, etc.). The wearable device may be integrated into the fabric of clothing, disposed in clothing such as a pocket, be in the form of a wristband, or be worn on the wrist or other limb. The wearable device may be worn by a subject such as a human, animal, or robot, or may be removably coupled or integrated with an item worn by the subject (e.g., a shirt, pants, safety vest, safety protective clothing, glasses, hat, helmet, hearing device, etc.), or may be movable such that the sensor can be moved between different positions and may be fixed or substantially fixed, or any alternative device. The wearable device may be worn or otherwise carried by various subjects or individuals such as, but not limited to, soldiers, medical professionals, athletes, system operators, students, and other active or inactive individuals. Optionally, the wearable sensing system may be coupled, integrated, and disposed on an asset such as a mobile system, a fixed system, such as a drone. The wearable system may be disposed in items such as a helmet, a pocket (e.g., of a shirt, pants, bag, etc.), gloves worn by a subject, a wristband, an ear covering, etc., and may be directly attached or otherwise coupled to the subject or asset, such as around the wrist, ankle, etc. The wearable device can be manufactured using manufacturing techniques based on complementary metal oxide semiconductor electronic circuits, flexible electronic circuits, flexible hybrid electronic circuits, and other known techniques for providing conformal and flexible designs, implementations, and uses.Optionally, the gas sensor 10 may be a fixed device, may be independently movable (e.g., removable from an operator and capable of moving independently of the operator), may be something that floats in the air, etc.

[0025] The gas sensor 10 may be in the form of a fluid container having a controlled volume, may be in the form of a container, or may be in the form of an open area such as indoor equipment (e.g., a room, hall, house, school, hospital, confined space, etc.) or outdoor equipment (e.g., a stadium, gas production site, filling station, gasoline filling station, hydrogen filling station, compressed natural gas filling station, liquefied natural gas filling station, gas distribution site, fuel distribution site, coast, forest, city, urban environment, marine environment, etc.) and may contact the fluid 26. In one embodiment, the gas sensor 10 may enable continuous monitoring of the fluid 26 in a reservoir or flow path. In one or more embodiments, the gas sensor 10 may be an impedance gas sensor, an electromagnetic sensor, an electronic sensor, a hybrid sensor, or another type of sensor. Optionally, the gas sensor 10 may be part of a sensor array.

[0026] The fluid 26 may be a gas, a liquid, a gas-liquid mixture, a solid, particles or particulate matter, etc., and contains one or more gases to be analyzed therein. In another embodiment, the fluid 26 may be a gas or fuel such as a hydrocarbon fuel. An example of the fluid 26 is natural gas or hydrogen gas supplied to a power system (e.g., a vehicle, an aircraft engine, or a stationary generator set) for consumption. Other examples of such fluid 26 include gasoline, diesel fuel, jet fuel or kerosene, biofuel, petroleum diesel-biodiesel fuel blend, natural gas (liquefied or compressed), and fuel oil. Another example of the fluid 26 is indoor or outdoor ambient air. Another example of the fluid 26 is air in industrial, residential, military, construction, urban, and any other known locations. Another example of the fluid 26 is ambient air containing relatively low concentrations of benzene, naphthalene, carbon monoxide, ozone, formaldehyde, nitrogen dioxide, sulfur dioxide, ammonia, hydrofluoric acid, hydrochloric acid, phosphine, ethylene oxide, carbon dioxide, hydrogen sulfide, chemical weapons, e.g., nerve gas, blister agent, blood agent, and choking agent, hydrocarbons and / or other contaminants. Other examples of the fluid 26 are disinfectants such as alcohol, aldehyde, chlorine dioxide, hydrogen peroxide, etc. Another example of the fluid 26 is ambient air containing relatively low, medium, and high concentrations of combustible or flammable gases, e.g., methane, ethane, propane, butane, hydrogen, and / or other gases. Another example of the fluid 26 is at least one gas dissolved in an industrial liquid such as transformer oil, bioprocess medium, fermentation medium, wastewater, etc. Another example of the fluid 26 is at least one gas dissolved in a consumer liquid such as milk, non-alcoholic beverage, alcoholic beverage, cosmetics, etc. Another example of the fluid 26 is at least one gas (e.g., biomarker) dissolved in a body fluid such as blood, sweat, tears, saliva, urine, etc.

[0027] In certain embodiments, fluid 26 can include an analyte gas that is a toxic industrial material or a toxic industrial chemical. A non-limiting list of exemplary toxic industrial materials and chemicals includes, but is not limited to, ammonia, arsine, boron trichloride, boron trifluoride, carbon disulfide, chlorine, diborane, ethylene oxide, fluorine, formaldehyde, hydrogen bromide, hydrogen chloride, hydrogen cyanide, hydrogen fluoride, hydrogen sulfide, nitric acid (fuming), phosgene, phosphorus trichloride, sulfur dioxide, sulfuric acid, and tungsten hexafluoride. In certain embodiments, fluid 26 can include an analyte gas that is a toxic substance with a moderate hazard index. A non-limiting list of exemplary toxic substances with a moderate hazard index includes, but is not limited to, acetone cyanohydrin, acrolein, acrylonitrile, allyl alcohol, allylamine, allyl chloroformate, boron tribromide, carbon monoxide, carbonyl sulfide, chloroacetone, chloroacetonitrile, chlorosulfonic acid, diketene, 1,2-dimethylhydrazine, ethylene dibromide, hydrogen selenide, methanesulfonyl chloride, methyl bromide, methyl chloroformate, chlorosilane, methylhydrazine, isocyanate, methyl mercaptan, nitrogen dioxide, phosphine, phosphorus oxychloride, phosphorus pentafluoride, selenium hexafluoride, silicon tetrafluoride, stibine, sulfur trioxide, sulfuryl chloride, sulfuryl fluoride, tellurium hexafluoride, n-octyl mercaptan, titanium tetrachloride, trichloroacetyl chloride, and trifluoroacetyl chloride.

[0028] In certain embodiments, fluid 26 may include the analyte gas, which is a low-hazard-index toxic substance. A non-limiting list of exemplary low-hazard-index toxic substances includes, but is not limited to, allyl isothiocyanate, arsenic trichloride, bromine, bromine chloride, bromine pentafluoride, bromine trifluoride, carbonyl fluoride, chlorine pentafluoride, chlorine trifluoride, chloroacetaldehyde, chloroacetyl chloride, crotonaldehyde, cyanogen chloride, dimethyl sulfate, diphenylmethane-4,4'-diisocyanate, ethyl chloroformate, ethyl chlorothioformate, ethyl phosphonous dichloride, ethyl phosphonous dichloride, ethyleneimine, hexachlorocyclopentadiene, hydrogen iodide, pentacarbonyl iron, isobutyl chloroformate, isopropyl chloroformate, isopropyl isocyanate, n-butyl chloroformate, n-butyl isocyanate, nitric oxide, n-propyl chloroformate, parathion, perchloromethyl mercaptan, sec-butyl chloroformate, tert-butyl isocyanate, tetraethyl lead, tetraethyl pyrophosphate, tetramethyl lead, toluene 2,4-diisocyanate, and toluene 2,6-diisocyanate.

[0029] In certain embodiments, fluid 26 may include the analyte gas, which is an indoor pollutant. A non-limiting list of exemplary indoor pollutants includes, but is not limited to, acetaldehyde, formaldehyde, 1,3-butadiene, benzene, chloroform, methylene chloride, 1,4-dichlorobenzene, perchloroethylene, trichloroethylene, naphthalene, and polycyclic aromatic compounds. In certain embodiments, fluid 26 may include the analyte gas, which is an outdoor pollutant. A non-limiting list of exemplary outdoor pollutants includes, but is not limited to, ozone, nitrogen dioxide, sulfur dioxide, and carbon monoxide.

[0030] Embodiments of the gas sensor 10 have the ability to distinguish different concentrations of gases in the fluid 26. For example, the gas sensor 10 can distinguish the gas to be analyzed at regulated vapor exposure limits established by different organizations. In certain embodiments, the gas sensor 10 can break down the gas to be analyzed below the Permissible Exposure Limit (PEL). In some embodiments, the gas sensor 10 can break down the gas to be analyzed below the Threshold Limit Value - Short Term Exposure Limit (TLV - STEL). In some embodiments, the gas sensor 10 can break down the gas to be analyzed below the Threshold Limit Value - Time Weighted Average (TLV - TWA). In some embodiments, the gas sensor 10 can break down the gas to be analyzed below the Immediately Dangerous to Life or Health (IDLH). In certain embodiments, the gas sensor 10 can break down the gas to be analyzed below and above the Lower Explosive Limit (LEL). In certain embodiments, the gas sensor 10 may be capable of breaking down gases at concentrations less than 5%, less than 100 parts per million (ppm), less than 100 parts per billion (ppb), less than 100 parts per trillion (ppt).

[0031] FIG. 2 is a flow diagram showing one embodiment of a process 70 by which the gas sensor 10 performs multi - gas analysis of a fluid sample 26. Process 70 begins with exposing the gas - sensing material 22 of the gas - sensing element 12 to a fluid sample having at least two gases (e.g., at least two gases to be analyzed, at least one gas to be analyzed and one interfering gas) (block 72). For example, the entire gas sensor 10, or only the gas - sensing material 22 of the gas sensor 10, may be exposed to the fluid sample. Process 70 includes heating the gas - sensing material 22 to a first operating temperature using the heater controller 28 and the heating element 24 (block 74). Typically, the gas - sensing material 22 is heated before, during, and after being exposed to the fluid sample 26.

[0032] When the gas sensing material 22 is exposed to the fluid sample 26 and heated to a first operating temperature, process 70 causes the measurement circuit 32 to induce dielectric excitation in the gas sensing material 22 operating at the first operating temperature using at least two preselected frequencies (block 76), and then proceeds to measure the dielectric excitation response (e.g., impedance response) of the gas sensing material 22. In certain embodiments, the measurement circuit 32 may further apply DC excitation to the gas sensing material 22 and measure the DC excitation response (e.g., resistance response) of the gas sensing material 22 at the first temperature. However, in some embodiments, the measurement circuit 32 may measure only the dielectric excitation response of the gas sensing material 22 when in contact with the fluid sample 26 at the first operating temperature. After measuring at least the dielectric response of the gas sensing material 22 at the first operating temperature, process 70 continues by heating the gas sensing material 22 to a second operating temperature using the heater controller 28 and the heating element 24 (block 78).

[0033] When the gas sensing material 22 is heated to a second operating temperature while still exposed to the fluid sample 26, process 70 causes the measurement circuit 32 to induce dielectric excitation in the gas sensing material 22 operating at the second operating temperature using at least two preselected frequencies (block 80), and then proceeds to measure the dielectric excitation response (e.g., impedance response) of the gas sensing material 22. In certain embodiments, the measurement circuit 32 may further apply DC excitation to the gas sensing material 22 and measure the DC excitation response (e.g., resistance response) of the gas sensing material 22 at the second temperature. However, in some embodiments, the measurement circuit 32 may measure only the dielectric excitation response of the gas sensing material 22 when in contact with the fluid sample 26 at the second operating temperature. In other embodiments, such as multi-gas analysis involving three or more gases, process 70 may include any suitable number of additional steps in which the gas sensing material 22 is heated to another (e.g., third, fourth, fifth, etc.) operating temperature, and it should be noted that the measurement circuit 32 measures at least the dielectric excitation response of the gas sensing material 22, as generally indicated by arrow 82 in FIG. 2.

[0034] Conventionally, the MOS gas sensor 10 measures the DC resistance response of the MOS-based sensing element 12 and uses the power-law relationship between the measured resistance and the gas concentration to associate the measured DC resistance response with the concentration of the gas. Such a DC resistance response from the MOS gas sensor 10 may be provided as a signal output in the form of an analog signal (e.g., to a user). Depending on the design of the analog circuit, the analog signal from the MOS gas sensor 10 may represent linear resistance, logarithmic resistance, or conductivity. Alternatively, the DC resistance response from the conventional MOS-based gas sensor 10 may be provided as a signal output in the form of a digitized DC resistance response signal. Depending on the design of the analog / digital circuit, the digital signal from the MOS gas sensor 10 may correlate with linear resistance, logarithmic resistance, or conductivity. The digital signal from the MOS gas sensor 10 that correlates with its DC resistance response can be provided (e.g., to a user) by any digital communication protocol, such as I2C (Inter-Integrated Circuit), alternatively known as IIC, and any other communication protocol.

[0035] In the illustrated embodiment, process 70 continues with the on-board data processor 36 of the gas sensor 10 performing on-board data analysis of the dielectric excitation response measured at each of the operating temperatures based on at least one of the stored gas analysis models 40 to provide real-time discrimination or resolution of the gases in the fluid sample that is not affected by ambient conditions (e.g., ambient temperature, ambient humidity) (block 84). In certain embodiments where the DC excitation response is also measured by the measurement circuit 32, the on-board data processor 36 can also provide the DC excitation response as an input to at least one of the stored gas analysis models 40 when decomposing the gases in the fluid sample. In this context, "real-time" means that the on-board data processor 36 of the gas sensor 10 can locally and rapidly decompose or discriminate the gases in the fluid sample without the need for the measured excitation response to be provided to an external computing system for processing.

[0036] In the embodiment of process 70 shown in FIG. 2, after decomposing the gas in fluid sample 26, gas sensor 10 can use one or more output devices 16 to provide an output of each classification 86 of the gas in the fluid sample, each concentration 88 of the gas in the fluid sample, or both (block 90). For example, one or more output devices 16 of gas sensor 10 may present or display each classification 86 and / or each concentration 88 of the gas in fluid sample 26. In certain embodiments, gas sensor 10 may provide each classification 86 and / or each concentration 88 of the gas to an external computing system via one or more suitable communication devices 48 (e.g., a wireless communication interface) of gas sensor 10. In certain embodiments, gas sensor 10 may use one or more output devices 16 to output each alarm 49 of the presence of gas in the fluid sample that exceeds a particular predetermined threshold level stored in memory 38 of gas sensor 10. Further, in certain embodiments, process 70 can proceed by returning to block 74 and repeating the remaining steps of process 70 for a predetermined time or a predetermined number of cycles, as indicated by arrow 92.

[0037] Figure 3 is a graph showing an exemplary impedance spectrum 100. In impedance spectroscopy, measurements of the real part Z’ and the imaginary part Z’’ of the impedance are performed over a wide range of frequencies to determine the shape of the impedance spectrum 100 of the gas sensor 10. As shown, the impedance spectrum includes two curves, each representing a part of the impedance response of the gas sensor 10 over a wide range of frequencies and determining the shape of the impedance spectrum. In particular, the first curve 102 represents the real part (Z’) of the impedance of the gas sensor 10, while the second curve 104 represents the imaginary part (Z’’) of the impedance of the gas sensor 10 measured over a wide range of frequencies. Different from broadband impedance spectroscopy measurements, dielectric excitation measurements are performed over a specific frequency range by following the front (high frequency or low frequency) shoulders of the dielectric relaxation regions obtained from their impedance measurements when MOS materials (n-type or p-type, respectively) are exposed to various gas concentrations.

[0038] In this embodiment, the measurement circuit 32 is an impedance detector that measures the dielectric excitation response of the gas sensor 10 in two or more frequency ranges 106, 108, or includes an impedance detector (which may or may not be disposed in the “dielectric relaxation region” of the gas sensor 10). For example, in a particular embodiment, each dielectric excitation response measured by the measurement circuit 32 may include a combination (e.g., sum, difference, any other mathematical expression) of a value (e.g., a real part impedance value) from the first curve 102 and a value (e.g., an imaginary part impedance value) from the second curve 104, both selected from the frequency ranges 106, 108. Alternatively, in some embodiments, each dielectric excitation response measured by the measurement circuit 32 may include a combination (e.g., sum, difference, any other mathematical expression) of a value (e.g., a real part impedance value Z’) from the first curve 102 and a value (e.g., an imaginary part impedance value Z’’) from the second curve 104, both selected from the frequency ranges 106, 108 or other frequency ranges.

[0039] The selection of the frequency ranges 106, 108 may depend on the type of the gas detection element 12 of the gas sensor 10. For example, in relation to the gas detection element 12, the selection of the frequency ranges 106, 108 may depend on the type of the MOS detection material, such as n-type or p-type, or a combination of n-type and p-type MOS detection materials, and the type of the gas for measurement, such as a reducing gas or an oxidizing gas. As a result, either the high-frequency shoulder region or the low-frequency shoulder region may be selected for measurement. For example, the response in the frequency ranges 106, 108 may include data indicating the sensor response to the gas in the fluid sample. Then, the identity and / or concentration of the gas in the fluid sample may be determined based on the sensor responses in the first and second frequency ranges 106, 108.

[0040] Alternatively, in some embodiments, each dielectric excitation response measured by the measurement circuit 32 may include a combination (e.g., sum, difference, any other mathematical expression) of values (e.g., real part impedance values) from the first curve 102 and values (e.g., imaginary part impedance values) from the second curve 104, both selected from the frequency ranges 106, 108 or other frequency ranges. The selection of the frequency ranges 106, 108 may depend on the type of the MOS detection material, such as n-type or p-type, or a combination of n-type and p-type MOS detection materials, and the type of the gas for measurement, such as a reducing gas or an oxidizing gas. As a result, either the high-frequency shoulder region or the low-frequency shoulder region may be selected. For example, the response in the frequency ranges 106, 108 may include data indicating the sensor response to the gas in the fluid sample. Then, the identity and / or concentration of the gas in the fluid sample may be determined based on the sensor responses in the first and second frequency ranges 106, 108.

[0041] In this embodiment, it will be appreciated that the dielectric excitation measurements shown in FIG. 3 are performed using at least two different operating temperatures. In the gas sensing element 12 of the present disclosure, temperature switching events still occur during multi-gas analysis, but it is now recognized that the number of temperature switching events (also referred to herein as temperature steps) is reduced compared to the case where the same gas sensing element is used only for resistance-based multi-gas analysis. It will be appreciated that reducing the number of temperature steps in multi-gas analysis has many advantages. For example, in certain embodiments, the gas sensor 10 utilizes fewer temperature steps for multi-gas analysis than a conventional resistance-based MOS gas sensor, so the measurement circuit 32 spends less time waiting for the gas sensing material 22 to reach a specific temperature after switching the operating temperature. As a result, faster and more efficient operation is achieved compared to conventional MOS-based gas sensors, as well as more stable and reproducible results. Furthermore, it is now recognized that the continuous temperature cycling of conventional MOS-based gas sensors undesirably degrades the heating elements and limits the operating life of these sensors. Therefore, by reducing the number of temperature steps, the gas sensor 10 of the present disclosure has been found to exhibit an improved operating life or lifespan compared to conventional resistance-based MOS gas sensors. Additionally, it has been found that reducing the number of temperature steps improves the quality of electrical measurements by reducing unwanted noise and hysteresis associated with temperature switching. Furthermore, for the gas sensor 10 of the present disclosure, the dielectric excitation response at a specific frequency of the sensor dielectric relaxation spectrum provides superior multi-gas discrimination and baseline stability compared to the resistance response of the same gas sensing element 12, even when a larger number of operating temperatures are used.

[0042] Experimental Example 1 To further demonstrate the excellent performance of the technology of the present disclosure, a first set of experiments was conducted to compare the capabilities of conventional resistance measurements and dielectric excitation measurements with respect to distinguishing between two different target gases in a fluid sample. In the first set of experiments, all measurements were performed using metal salt-doped tin oxide (SnO2) as the gas sensing material 22. The measurements were performed by the measurement circuit 32 as either a DC excitation response (e.g., resistance measurement) or a dielectric excitation response (e.g., impedance measurement) selected from the high-frequency shoulder regions 106 or 108 of the corresponding impedance spectrum 100 after dielectric excitation. Each excitation response measurement was performed using heater element voltages from a heater controller of 1.6 volts (V) and 2.4 volts (V). In the embodiment of the gas sensor 10 used in these experiments, a heater element voltage of 1.6 V corresponded to an operating temperature of approximately 250 degrees Celsius (°C), while a heater element voltage of 2.4 V corresponded to an operating temperature of approximately 350 °C.

[0043] As shown in FIGS. 4A and 4B, for the first set of experiments, the gas sensor 10 was exposed to two different target gases at four different concentrations, as well as binary mixtures of the two target gases at eight different ratios. The first target gas is carbon monoxide (CO), and the four different concentrations of the first target gas are 160 parts per million (ppm), 320 ppm, 480 ppm, and 640 ppm, which are shown as 0.2, 0.4, 0.6, and 0.8 in FIGS. 4A and 4B, respectively. The second target gas is methane gas (CH4), and the four different concentrations of the second target gas are 800 ppm, 1600 ppm, 2400 ppm, and 3200 ppm, which are shown as 0.2, 0.4, 0.6, and 0.8 in FIGS. 4A and 4B, respectively. Thus, in the experiment shown in FIG. 4, the gas sensor 10 was exposed to (A) each of the four concentrations of the first target gas while resistance responses and impedance responses were being collected at each operating temperature, (B) each of the four concentrations of the second target gas while resistance responses and impedance responses were being collected at each operating temperature, and then (C) each of the eight binary mixtures of the first and second target gases while resistance responses and impedance responses were being collected at each operating temperature.

[0044] Principal component analysis (PCA) was applied to analyze the resistance responses (e.g., DC excitation responses) of the experiment described in FIG. 4 to examine the ability of the gas sensor 10 to distinguish between the two target gases using two operating temperatures. For each sensor state (two target gases at different concentrations and blank), three data points were extracted from the raw dynamic responses at the steady state or maximum signal change of the sensor response. FIGS. 5A and 5B show the results of PCA as score plots of the first two principal components (PC1 vs. PC2). More specifically, FIG. 5A is a score plot 120 showing the results of PCA in the analysis of the linear resistance response, and FIG. 5B is a score plot 122 showing the results of PCA in the analysis of the logarithmic resistance response.

[0045] As shown by the score plot 120 of FIG. 5A, the analysis of the linear resistance response by PCA indicates that the desired discrimination of multi-gases with sufficient gas response linearity and baseline stability cannot be obtained from the resistance measurements. For example, as shown in the figure, the measured DC excitation response did not correlate well with the concentrations of the first and second target gases for analysis. Further, for the binary mixture, the first and second target gases for analysis were not sufficiently discriminated (e.g., distinguished, identified, resolved) from each other. Further, in FIG. 5A, the baseline instability in the raw resistance response was significantly prominent in the PCA score plot as the spread between those data points.

[0046] As shown by the score plot 122 of FIG. 5B, the analysis of the logarithmic resistance response by PCA shows some improvement compared to the analysis of the linear resistance response of FIG. 5A. For example, in FIG. 5B, the DC excitation response measured during exposure to the second target gas for analysis correlated more strongly with the concentration of the second target gas for analysis than that observed in FIG. 5A, but the correlation showed substantial non-linearity. Further, although baseline drift was still observed in FIG. 5B, the baseline instability was not as prominent as that observed in FIG. 5A. However, in the score plot 122 of FIG. 5B, the binary mixture of the first and second target gases for analysis still was not sufficiently discriminated or resolved from each other. Therefore, in the analysis of the logarithmic resistance response, the desired discrimination of multi-gases with the desired gas response linearity and baseline stability still could not be obtained.

[0047] For comparison with resistance measurements and analysis, PCA was also applied to analyze the dielectric excitation responses (e.g., impedance responses) of the first set of experiments described in FIG. 4, and to examine the ability of gas sensor 10 to distinguish between two target gases using two operating temperatures. FIG. 6 is a score plot 130 showing the PCA results of the first two principal components (PC1 vs. PC2) of the dielectric excitation responses of gas sensing material 22 when operating at two operating temperatures. In FIG. 6, score plot 130 shows a clear distinction between the two target gases including the binary mixture, and shows a high degree of linearity in the correlation between the dielectric excitation response and the concentration of the target gas. For example, in the score plot 130 of FIG. 6, the dielectric excitation responses of the blanks form a dense cluster, and each of the dielectric excitation responses of the target gases forms a relatively linear trend that coincides with the blank samples. Thus, unlike the DC excitation responses shown in FIGS. 5A and 5B, the score plot 130 of FIG. 6 unexpectedly shows a strong correlation or proportional relationship with a high degree of linearity between the dielectric excitation response of gas sensing material 22 and the respective concentrations of the two target gases when two operating temperatures are used.

[0048] Experimental Example 2 To further demonstrate the excellent performance of the technology of the present disclosure, a second set of experiments was conducted to compare the capabilities of conventional resistance measurements and dielectric excitation measurements for distinguishing between three different target gases in a fluid sample at two operating temperatures. In the second set of experiments, all measurements were performed using metal salt-doped tin oxide (SnO2) as gas sensing material 22. The measurements were performed by measurement circuit 32 as either a DC excitation response (e.g., resistance measurement) or a dielectric excitation response (e.g., impedance measurement) selected from the high frequency shoulder regions 106 or 108 of the corresponding impedance spectrum 100 after dielectric excitation. Each excitation response measurement was performed at a heating element voltage of 6.5V (corresponding to the first operating temperature) and also at a heating element voltage of 7.5V (corresponding to the second operating temperature). For the embodiment of gas sensor 10 used in these experiments, a heating element voltage of 6.5V corresponds to an operating temperature of approximately 250 degrees Celsius (° C.), while a heating element voltage of 7.5V corresponds to an operating temperature of approximately 285° C.

[0049] For the second set of experiments, gas sensor 10 was exposed to three different target gases at each of four different concentrations. The first target gas was toluene vapor, and the four different concentrations of the first target gas were 8 ppm, 16 ppm, 24 ppm, and 32 ppm. The second target gas was acetone vapor, and the four different concentrations of the second target gas were 8 ppm, 16 ppm, 24 ppm, and 32 ppm. The third target gas was benzene vapor, and the four different concentrations of the third target gas were 8 ppm, 16 ppm, 24 ppm, and 32 ppm. Thus, in the second set of experiments, gas sensor 10 was exposed to each of the four concentrations of the first target gas while resistance responses and impedance responses were being collected at each operating temperature, and then (B) to each of the four concentrations of the second target gas while resistance responses and impedance responses were being collected at each operating temperature.

[0050] PCA was applied to analyze the resistance responses (e.g., DC excitation responses) of the second set of experiments to examine the ability of gas sensor 10 to distinguish three target gases using two operating temperatures. For each sensor state (two target gases at different concentrations and blank), three data points were extracted from the raw dynamic responses at the steady state or maximum signal change of the sensor response. FIG. 7 is a score plot 140 showing the results of PCA of the first two principal components (PC1 vs. PC2) for the analysis of the logarithmic resistance responses. The analysis of the logarithmic resistance responses by PCA shows the distinction between the three gases, which have some significant non-linearity in the measured DC excitation responses. Also, in the score plot 140 of FIG. 7, the baseline instability in the raw resistance responses was prominent in the PCA score plot as the spread between the data points of the baseline (blank) responses.

[0051] For comparison with resistance measurement and analysis, PCA was applied to analyze the dielectric excitation responses (e.g., impedance responses) of the second set of experiments and examine the ability of the gas sensor 10 to distinguish three target gases using two operating temperatures. FIG. 8A is a score plot 150 showing the PCA results of the first two PCs (PC1 vs. PC2) of the dielectric excitation responses of the gas sensing material 22 when operating at two operating temperatures, and FIG. 8B is a score plot 150 showing the PCA results of the first three PCs (PC1 vs. PC2 vs. PC3) of the dielectric excitation responses of the gas sensing material 22 when operating at two operating temperatures. In FIGS. 8A and 8B, the score plots 150 and 152 show a clear distinction between the three target gases and a high degree of linearity in the correlation between the dielectric excitation responses and the concentrations of the target gases. For example, in FIGS. 8A and 8B, the responses to the blank form a dense cluster, and each of the responses to the target gases forms a relatively linear trend that coincides with the blank sample. Thus, unlike the PCA of the DC excitation responses shown in FIGS. 5A and 5B, the score plots 150, 152 in FIGS. 8A and 8B unexpectedly show a strong correlation or proportional relationship with a high degree of linearity between the dielectric excitation responses of the gas sensing material 22 and the respective concentrations of the two target gases when two operating temperatures are used.

[0052] For further comparison between the DC excitation response measurements and the dielectric excitation response measurements, hierarchical cluster analysis (HCA) was also applied to analyze the DC excitation responses (e.g., resistance responses) and dielectric excitation responses (e.g., impedance responses) for the second set of experiments and examine the ability of the gas sensor 10 to distinguish three target gases using two operating temperatures. For the HCA analysis, a dendrogram of the data with autoscaling preprocessing was generated using the K-nearest neighbor algorithm with two PCs for analysis and the Mahalanobis distance metric. The K-nearest neighbor algorithm is a nonparametric method used for classification and regression. The Mahalanobis distance is an effective multivariate distance metric for measuring the distance between a specific data point and the distribution of data points.

[0053] FIG. 9 is an HCA plot 160 showing the results of HCA for the logarithmic resistance response from the second set of experiments. As shown by the HCA plot 160 of FIG. 9, the blank sample and the third analyte gas are well resolved from the other analytes, while the first and second analyte gases are not well resolved. For comparison, FIG. 10 is an HCA plot 162 showing the results of HCA for the dielectric excitation response from the second set of experiments. As shown by the HCA plot 162 of FIG. 10, all three of the analyte gases are desirably resolved from each other and also from the blank sample.

[0054] Experimental Example 3 To further demonstrate the excellent performance of the technology of the present disclosure, a third set of experiments was conducted to compare the capabilities of conventional resistance measurements and dielectric excitation measurements for distinguishing between three different analyte gases in a fluid sample at three operating temperatures. In the third set of experiments, all measurements were performed using metal salt-doped tin oxide (SnO2) as the gas sensing material 22. The measurements were performed by the measurement circuit 32 as either a DC excitation response (e.g., resistance measurement) or a dielectric excitation response (e.g., impedance measurement) selected from the high-frequency shoulder region 106 or 108 of the corresponding impedance spectrum 100 after dielectric excitation. Each excitation response measurement was performed at a heating element voltage of 5 V (corresponding to the first operating temperature), at 6 V (corresponding to the second operating temperature), and at 7 V (corresponding to the second operating temperature). For the embodiment of the gas sensor 10 used in these experiments, a heating element voltage of 5 V corresponds to an operating temperature of approximately 200 degrees Celsius (°C), a heating element voltage of 6 V corresponds to an operating temperature of approximately 230 °C, and a heating element voltage of 7 V corresponds to an operating temperature of approximately 265 °C.

[0055] For the third set of experiments, gas sensor 10 was exposed to three different target gases at each of three different concentrations. The first target gas was toluene vapor, and the three different concentrations of the first target gas were 6.25 ppm, 12.5 ppm, and 18.75 ppm. The second target gas was acetone vapor, and the three different concentrations of the second target gas were 6.25 ppm, 12.5 ppm, and 18.75 ppm. The third target gas was benzene vapor, and the three different concentrations of the third target gas were 6.25 ppm, 12.5 ppm, and 18.75 ppm. Thus, in the third set of experiments, gas sensor 10 was (A) heated to the first operating temperature, (B) exposed to a blank sample while resistance response and impedance response were measured, (C) exposed to each of the three concentrations of the first target gas while resistance response and impedance response were measured, (D) exposed to each of the three concentrations of the second target gas while resistance response and impedance response were measured, (E) exposed to each of the three concentrations of the third target gas while resistance response and impedance response were measured, and then (F) exposed to a blank sample while resistance response and impedance response were measured. Then, while steps (B) to (F) were repeated, gas sensor 10 was heated to the second operating temperature, and then again while steps (B) to (F) were repeated, gas sensor 10 was heated to the third operating temperature. For each sensor state (two target gases at different concentrations and blank), three data points were extracted from the raw dynamic response at the steady state of the sensor response or the maximum signal change.

[0056] Figures 11A to 11C show the logarithmic resistance response of the gas sensor 10 for a third set of experiments. More specifically, FIG. 11A is a graph 170 showing the logarithmic resistance response of the gas sensor 10 at a first operating temperature, FIG. 11B is a graph 172 showing the logarithmic resistance response of the gas sensor 10 at a second operating temperature, and FIG. 11C is a graph 174 showing the logarithmic resistance response of the gas sensor 10 at a third operating temperature. The logarithmic resistance responses in FIGS. 11A to 11C show that the logarithmic resistance response has a specific pattern at different operating temperatures. In the illustrated example, the magnitude from the minimum value to the maximum value of the logarithmic resistance response is at a heating element voltage of 5V (first operating temperature), and the pattern is the third analyte gas (benzene), then the first analyte gas (toluene), then the second analyte gas (acetone). At a heating element voltage of 6V (second operating temperature), the pattern is the third analyte gas (benzene), then the second analyte gas (acetone), then the first analyte gas (toluene). At a heating element voltage of 7V (third operating temperature), the pattern is the second analyte gas (acetone), then the third analyte gas (benzene), then the first analyte gas (toluene). However, even using a logarithmic scale, the resistance response is non-linear. It will be understood that such non-linearity significantly reduces the ability to distinguish and quantify different analytes. Furthermore, it is currently recognized that such non-linearity imposes a significant and undesirable burden on the calibration of the gas sensor 10.

[0057] PCA was applied to analyze the resistance responses (e.g., DC excitation responses) of the third set of experiments to examine the ability of gas sensor 10 to distinguish three target gases using three operating temperatures. FIG. 12A is a score plot 180 showing the results of PCA for the first two principal components (PC1 vs. PC2) in the analysis of the logarithmic resistance responses, and FIG. 12B is a score plot 182 showing the results of PCA for the first three principal components (PC1 vs. PC2 vs. PC3) in the analysis of the logarithmic resistance responses. Both FIG. 12A and FIG. 12B show the desired discrimination between the three target gases, but both plots also show significant and undesirable gas response non-linearity and non-monotonic behavior. FIGS. 13A - 13C show the contributions of PC1, PC2, and PC3, respectively, to the response patterns of gas sensor 10 by resistance readout.

[0058] For example, as shown by plot 190 in FIG. 13A, the resistance response of gas sensor 10 monotonically follows the concentrations of the three vapors only for PC1. In plot 192 of FIG. 13B, the resistance responses for the first target gas (toluene) and the second target gas (acetone) are non-monotonic, and only the resistance response for the third target gas (benzene) is monotonic in PC2. In plot 194 of FIG. 13C, the responses for all three analytes are non-monotonic with respect to PC3. It is currently recognized that such non-monotonic responses of gas sensor 10 significantly degrade the accuracy of discrimination between analytes, the accuracy of quantification of different analytes, and the ability to quantify different mixtures of analytes.

[0059] Figures 14A to 14C show the dielectric excitation response of the gas sensor 10 for the third set of experiments. More specifically, the graph 200 in FIG. 14A shows the dielectric excitation response of the gas sensor 10 at the first operating temperature, the graph 202 in FIG. 14B shows the dielectric excitation response of the gas sensor 10 at the second operating temperature, and the graph 204 in FIG. 14C shows the dielectric excitation response of the gas sensor 10 at the third operating temperature. The dielectric excitation responses in FIGS. 14A to 14C have specific patterns at different operating temperatures. In the illustrated example, the magnitude from the minimum value to the maximum value of the inductive excitation response is at a heating element voltage of 5V (first operating temperature), and the pattern is the third analyte gas (benzene), then the first analyte gas (toluene), then the second analyte gas (acetone); at a heating element voltage of 6V (second operating temperature), the pattern is the third analyte gas (benzene), then the second analyte gas (acetone), then the first analyte gas (toluene); at a heating element voltage of 7V (third operating temperature), the pattern is the second analyte gas (acetone), then the third analyte gas (benzene), then the first analyte gas (toluene). Therefore, in the embodiment of the gas sensor 10 used in the third set of experiments, this pattern regarding the magnitude of the dielectric excitation response at different operating temperatures coincides with the pattern for the DC excitation response. However, unlike the resistance responses in FIGS. 11A to 11C, the dielectric excitation responses in FIGS. 14A to 14C have a high degree of linearity or are completely linear. This high degree of linearity is currently recognized to significantly improve the ability of the gas sensor 10 to distinguish different analytes and quantify the concentrations of different analytes. Furthermore, this high degree of linearity reduces the calibration burden of the gas sensor 10.

[0060] The response linearity of the gas sensor 10 as measured by the logarithmic resistance response (e.g., FIGS. 11A to 11C) or the dielectric excitation response (e.g., FIGS. 14A to 14C) was determined as the coefficient of determination R of the linear fit between the gas concentration and the sensor response. 2 For the logarithmic resistance response at three voltages of 5V, 6V, and 7V, R of the first analyte gas 2are 0.787, 0.829, and 0.945 respectively, and R for the second gas to be analyzed 2 are 0.751, 0.811, and 0.913 respectively, and R for the third gas to be analyzed 2 were 0.891, 0.907, and 0.985 respectively. For the dielectric excitation responses at three voltages of 5V, 6V, and 7V, R for the first gas to be analyzed 2 are 0.999, 0.996, and 0.992 respectively, and R for the second gas to be analyzed 2 are 0.994, 0.986, and 0.970 respectively, and R for the third gas to be analyzed 2 were 0.993, 0.994, and 0.991 respectively. Therefore, the response linearity of gas sensor 10 was consistently higher when measured as the dielectric excitation response (e.g., FIGS. 14A to 14C) compared to when measured as the logarithmic resistance response (e.g., FIGS. 11A to 11C) of gas sensor 10.

[0061] PCA was applied to analyze the dielectric excitation responses of the third set of experiments to examine the ability of gas sensor 10 to distinguish three target gases using three operating temperatures. FIG. 15A is a score plot 210 showing the results of PCA for the first two principal components (PC1 vs. PC2) in the analysis of the dielectric excitation responses, FIG. 15B is a score plot 212 showing the results of PCA for the second and third principal components (PC2 vs. PC3) in the analysis of the dielectric excitation responses, and FIG. 15C is a score plot 214 showing the results of PCA for the first three principal components (PC1 vs. PC2 vs. PC3) in the analysis of the dielectric excitation responses. The score plots in FIGS. 15A-15C all show the discrimination between the three target gases with high response linearity. FIGS. 16A-16C show the contribution of each of PC1, PC2, and PC3 to the response pattern of gas sensor 10 by dielectric excitation readout (response). In each of the plots 220, 222, and 224 in FIGS. 16A-16C, the response of gas sensor 10 linearly and monotonically and desirably follows the concentrations of the three analytes for all three PCs. Such a monotonic behavior of gas sensor 10 facilitates the accurate discrimination between analytes and the ability for the accurate quantification of different analytes and the quantification of different target mixtures.

[0062] When using the dielectric excitation response against the resistance response with the same MOS gas sensing material, a further comparison of the discrimination and linearity of multi-gases that can be achieved at lower operating temperatures was performed. In FIGS. 12A and 12B, the resistance response was analyzed by applying PCA, and the ability of the gas sensor 10 to distinguish three target gases was examined using the resistance response and three operating temperatures related to 5V, 6V, and 7V of the heating element voltage. The contributions of PC1, PC2, and PC3 were 91.99%, 7.73%, and 0.28%, respectively, corresponding to 100% of the variation captured by the three PCs. At higher PCs, the contribution is more and the dispersion of the sensor is greater. The dispersion or dimensionality of the sensor response is the sensor ability to provide independent outputs generated by the sensor. FIGS. 12A and 12B show that the three gases were discriminated, but there are two important points. First, non-linear responses were observed for all three gases. Second, the contribution to PC1 of 91.99% was relatively large, and when operating at three temperatures, there were not many other PCs left to support the 2D or 3D dispersion of the gas sensor response.

[0063] In contrast, only two operating temperatures were used to examine the ability of the gas sensor 10 to distinguish three target gases using the dielectric excitation response for the third set of experiments. The two operating temperatures were related to 5V and 6V of the heating element voltage. FIG. 17 is a score plot 230 showing the results of PCA for the first two principal components (PC1 vs. PC2) in the analysis of the dielectric excitation response. The contributing PC1 was 81.51% and the contributing PC2 was 18.49%, corresponding to 100% of the variation captured by the two PCs. FIG. 17 shows that the three gases were distinguished by three important points. First, linear responses were observed for all three gases. Second, the contribution to PC1 of 81.51% enabled a substantial value of PC2, which was 18.49%. Third, the contributing PC2 of 18.49% due to the dielectric excitation response of the MOS gas sensor was 2.4 times larger compared to the contributing PC2 of 7.73% due to the resistance response obtained at more levels of operating temperature. This such larger value of PC2 achieved with the dielectric excitation response is important in the discrimination of multi-gases.

[0064] Figures 18A and 18B show the respective contributions of PC1 and PC2 to the response pattern of gas sensor 10 by dielectric excitation response and at two operating temperatures. In each of plots 240 and 242 of Figures 18A and 18B, the response of gas sensor 10 desirably follows linearly and monotonically with the concentration of three analytes for all two PCs. Such a monotonic and linear behavior of gas sensor 10 facilitates an accurate distinction between analytes and promotes the ability for accurate quantification of different analytes and quantification of different analyte mixtures.

[0065] In contrast, Figures 13A - 13C show the respective contributions of PC1, PC2, and PC3 to the response pattern of gas sensor 10 by resistance response at three operating temperatures. The resistance response of gas sensor 10 follows monotonically with the concentration of three vapors only for PC1, as shown by plot 190 in Figure 13A. Such a monotonic response to the vapor gradually deteriorates into a non - monotonic response for higher PCs. In plot 192 of PC2 in Figure 13B, the resistance responses to the first and second analyte gases are non - monotonic, while only the resistance response to the third analyte gas is monotonic. In plot 194 of Figure 13C, the responses to all three analytes are non - monotonic with respect to PC3. Therefore, by measuring the dielectric excitation response of the gas sensing material, an improvement in multi - gas discrimination can be achieved using fewer operating temperatures than those used by the same MOS gas sensing material configured to perform multi - gas discrimination based only on the resistance response.

[0066] The technical effects of the present invention include enabling the detection of multi-gases using a smaller number of temperature cycles as compared to conventional resistance-based gas detection. That is, by measuring the dielectric excitation response of the gas sensing material, a smaller operating temperature is used than that used by the same MOS gas sensing material configured to perform the discrimination and decomposition of multi-gases based only on the resistance response, and an improvement in the discrimination of multi-gases can be achieved. Further, the gas sensor and gas detection technology of the present disclosure enable improved response linearity, improved dynamic range, and reduced computational resource consumption with respect to multi-gas quantification as compared to conventional resistance-based gas detection methods. Further, by reducing the number of operation temperature switching events, the present embodiment enables a gas sensor having improved measurement quality and improved operating life. That is, when using a set of predetermined operating temperatures, the dielectric relaxation spectrum of the gas sensing material is affected differently by different gases, and such a desired difference is more prominent compared to the resistance response of the same gas sensing material even when additional operating temperatures are used. Therefore, the present embodiment unexpectedly demonstrates a MOS-based gas sensor capable of distinguishing different gases using responses collected using at least two different operating temperatures, and this discrimination is excellent in the discrimination between different gases and baseline stability compared to the resistance responses of the same gas sensing material at three or more operating temperatures.

[0067] This specification uses examples to disclose the invention, including the best mode, and to enable the practice of the invention, including the making and using of any devices or systems and the performing of any incorporated methods, by those of ordinary skill in the art. The patentable scope of the invention is defined by the claims and may include other examples that occur to those of ordinary skill in the art. Such other examples are intended to be within the scope of the claims if they have structural elements that do not differ from the literal language of the claims, or if they include equivalent structural elements that do not differ substantially from the literal language of the claims.

Claims

**Claim 1** A gas sensor system for multi-gas analysis of a fluid sample, comprising: a gas sensing element configured to operate at a plurality of temperatures and to contact the fluid sample; a heating element coupled to the gas sensing element and configured to heat the gas sensing element; a heater controller operably coupled to the heating element and configured to control the heating element to heat the gas sensing element to each of the plurality of temperatures while the gas sensing element is in contact with the fluid sample; a measurement circuit operably coupled to the gas sensing element and configured to cause dielectric excitation in the gas sensing element while the gas sensing element is heated to each of the plurality of temperatures and in contact with the fluid sample, and to measure a dielectric excitation response of the gas sensing element, wherein the measured dielectric excitation response, when compared with a resistance response of the gas sensing element when in contact with the fluid sample at each of the plurality of temperatures, improves discrimination between at least two gases in the fluid sample and improves response linearity to the at least two gases in the fluid sample; a gas sensor system. **Claim 2** An on-board data processor communicably coupled to the measurement circuit, configured to receive, from the measurement circuit, the dielectric excitation response of the gas sensing element at each of the plurality of temperatures while the gas sensing element is in contact with the fluid sample, and to select at least two of the dielectric excitation responses of the gas sensing element at each of the plurality of temperatures to discriminate between the at least two gases in the fluid sample The gas sensor system according to claim 1, comprising. **Claim 3** The gas sensor system according to claim 2, wherein the at least two selected dielectric excitation responses are impedance responses of the gas sensing element at each of the plurality of temperatures. **Claim 4** The gas sensor system according to claim 2, wherein the at least two selected dielectric excitation responses are not resistance responses. **Claim 5** The on-board data processor is configured to discriminate between the at least two gases by determining a classification, a concentration, or a combination thereof of each of the at least two gases in the fluid sample based on the at least two selected dielectric excitation responses of the gas sensing element at each of the plurality of temperatures. The gas sensor system according to claim 2. **Claim 6** The gas sensor system according to claim 1, wherein the measurement circuit is configured to measure the dielectric excitation response of the gas sensing element at a plurality of preselected frequencies at each of the plurality of temperatures.

7. The gas sensor system according to claim 6, wherein the measurement circuit measures the dielectric excitation response of the gas sensing element that correlates with the concentration of a specific gas among a plurality of gases in the fluid sample at each of the preselected frequencies and at each of the plurality of temperatures.

8. The gas sensing element is a substrate, wherein the heating element is coupled to the substrate, and a single sensing material disposed on the substrate and configured to contact the fluid sample, and an electrode coupled to the single sensing material and electrically coupled to the measurement circuit, the electrode being configured to apply the dielectric excitation caused by the measurement circuit to the single sensing material, and the measurement circuit being configured to measure the dielectric excitation response of the single sensing material at each of the plurality of temperatures via the electrode. The gas sensor system according to claim 1.

9. The gas sensor system according to claim 8, wherein the single sensing material is a semiconductor metal oxide material.

10. The gas sensor system according to claim 1, wherein the plurality of temperatures includes at least two different temperatures.

11. The gas sensor system according to claim 1, wherein the at least two gases in the fluid sample include at least two gases to be analyzed or at least one gas to be analyzed and at least one interfering gas.

12. A method of operating a gas sensor for multi-gas analysis of a fluid sample, comprising: exposing a gas sensing material of the gas sensor to the fluid sample; measuring, by a measurement circuit of the gas sensor, a first set of dielectric excitation responses of the gas sensing material while the gas sensing material is heated to a first temperature and exposed to the fluid sample; measuring, by the measurement circuit of the gas sensor, a second set of dielectric excitation responses of the gas sensing material while the gas sensing material is heated to a second temperature and exposed to the fluid sample; receiving, by an on-board data processor of the gas sensor, the first set of dielectric excitation responses of the gas sensing material at the first temperature and the second set of dielectric excitation responses of the gas sensing material at the second temperature, and Based on at least a part of the first set of the dielectric excitation responses and at least a part of the second set of the dielectric excitation responses, decomposing at least two gases in the fluid sample by the on-board data processor A method comprising.

13. The discrimination of the at least two gases in the fluid sample based on at least a part of the first set of the dielectric excitation responses and at least a part of the second set of the dielectric excitation responses is superior to the discrimination of the at least two gases in the fluid sample that is possible using the resistance response of the gas sensing material while in contact with the fluid sample at each of the first and second temperatures. The method according to claim 12

14. Before measuring the first set of the dielectric excitation responses, applying a first voltage to the heater element of the gas sensor by the heater controller of the gas sensor to heat the gas sensing material to the first temperature, and Before measuring the second set of the dielectric excitation responses, applying a second voltage to the heater element by the heater controller of the gas sensor to heat the gas sensing material to the second temperature, wherein the first voltage and the second voltage are different, The method according to claim 12

15. Measuring the first set of the dielectric excitation responses includes Measuring, by the measurement circuit, a first dielectric excitation response of the gas sensing material at a first frequency related to the first temperature and a concentration of a first gas in the fluid sample, and Measuring, by the measurement circuit, a second dielectric excitation response of the gas sensing material at a second frequency related to the first temperature and a concentration of a second gas in the fluid sample The method according to claim 12

16. Measuring the second set of the dielectric excitation responses includes Measuring, by the measurement circuit, a third dielectric excitation response of the gas sensing material at the first frequency related to the second temperature and the concentration of the first gas in the fluid sample, and Measuring, by the measurement circuit, a fourth dielectric excitation response of the gas sensing material at the second frequency related to the second temperature and the concentration of the second gas in the fluid sample The method according to claim 15

17. The decomposing includes Selecting, by the on-board data processor, at least two of the first set of the dielectric excitation responses of the gas sensing material at the first temperature and at least two of the second set of the dielectric excitation responses of the gas sensing material at the second temperature, and determining the respective concentrations of at least two gases in the fluid sample based on the at least two selected dielectric excitation responses of the first set of dielectric responses and the at least two selected dielectric excitation responses of the second set of dielectric responses The method according to claim 12, comprising:

18. The method according to claim 17, wherein the at least two selected dielectric excitation responses of the first set of dielectric responses and the at least two selected dielectric excitation responses of the second set of dielectric responses are impedance responses.

19. The method according to claim 18, wherein the at least two selected dielectric excitation responses of the first set of dielectric responses and the at least two selected dielectric excitation responses of the second set of dielectric responses do not include a resistance response.

Images