WO2006102064A2

WO2006102064A2 - Gated gas sensor

Info

Publication number: WO2006102064A2
Application number: PCT/US2006/009686
Authority: WO
Inventors: James Novak; Prabhu Soundarrajan
Original assignee: Nano-Proprietary, Inc.
Priority date: 2005-03-18
Filing date: 2006-03-17
Publication date: 2006-09-28
Also published as: CA2599374A1; KR20070121761A; EP1864122A2; WO2006102064A3; WO2006102064B1; TW200706863A

Abstract

An apparatus for sensing an analyte gas is provided. The apparatus may include a signal amplifier that may include a thin film transistor that may include a semiconducting film that may include a metal oxide capable of chemical interaction with the analyte gas, such as carbon monoxide. The apparatus may be tuned for detecting the analyte gas by varying the gate voltage of the transistor.

Description

GATED GAS SENSOR

STATEMENTREGARDINGFEDERALLY SPONSORED RESEARCH OR

DEVELOPMENT [0001] The United States has certain rights in this invention pursuant to Grant

No. FA8650-05-M-6562 awarded by the Air Force.

TECHNICAL FIELD

[0002] The present invention relates in general to gas sensors. More particularly the present invention relates to gas sensors that operate by applying a gate voltage so as to tune detection of a current through a layer of a compound that is capable of chemical interaction with an analyte gas. Additionally, other methods could be used to tune the detection of an analyte, including, optical excitation, chemical dopants, surface chemical layers and combinations thereof. Additionally, the application of an external force, such as a gate bias, helps to eliminate or greatly reduce the requirement for a heated substrate surface. Still more particularly, the present invention relates to thin film gated metal oxide detectors adapted for detection of analyte gases, such a carbon monoxide.

BACKGROUND INFORMATION Heated Metal Oxide Sensors.

[0003] There is a large background of information regarding metal oxide chemical sensors. These have all been on heated substrates or measured by electrochemical detection. For example see Eranna, G. et al. "Oxide Materials for Development of Integrated Gas Sensors - A Comprehensive Review" Critical Reviews in Solids State and Materials Sciences, 29: 111-188, 2004 and the references therein.

[0004] It is desirable to provide sensors for detecting ambient levels of gases, particularly noxious gases. Heated metal oxide sensors are well studied in the literature. As a recent review, Eranna et al, (table 23, page 174, herein denoted Ref. 1) shows the range of gases that can be detected and the metal oxides that are sensitive to each gas. The inventors have demonstrated many improvements to the ability to sense gases using a metal oxide surface. [0005] Many metal oxides are semiconducting. This means there is an energy gap between the population of electrons called the valance band and the conduction band where these electrons can move through the material. This energy gap is commonly called the band gap and denoted as E_g. The metal oxide sensors take advantage of this semiconducting nature by promoting electrons from the valence band to the conduction band through heat. This thermal excitation of electrons can change the surface energy of the metal oxide promoting facile chemical reactions.

[0006] Metal oxides typically have an amorphous crystal structure. This means there will be individual crystalline grains but that there is no long-range order to the surface. The interface between each crystallite creates a grain boundary. Conduction through a metal oxide is limited by the energy barrier created at each grain boundary, hi addition to the change in surface energy heat changes the energy level of barriers created at these individual grain boundaries.

[0007] As one particular example, we will look at carbon monoxide (CO) detection, the current state of the art for detection and their limitations. Carbon monoxide poisoning presents a major problem in civilian and military sectors. It is estimated that more than 500 people accidentally die from carbon monoxide ("CO") poisoning each year in the United States, more than from any other poison, m addition, an estimated 10,000 people are treated annually for symptoms of CO exposure. While most of the household CO related incidents could be identified and treated, the situation is more critical in aircraft environments where a lack of suitable monitoring devices is available.

[0008] Carbon monoxide has about 210 times the affinity to bind to hemoglobin compared to oxygen. CO is an odorless, tasteless, colorless gas that causes hypaemic hypoxia wherein there is a reduced oxygen carrying capacity of the blood. Carbon monoxide in the blood creates carboxyhaemoglobin (COHb) which prevents oxygen uptake. At sea level, increased levels of COHb cause various symptoms ranging from headache to unconsciousness. At 200 ppm, CO at sea level causes a headache (equivalent to 15-20% COHb content in the body). At higher altitudes, the effects of CO poisoning and altitude hypoxia are cumulative, driving a need for a continuous low-level monitoring of sub-200 ppm levels of CO in aircraft cabins. [0009] Several metal oxide and electrochemical sensors have been operational in household CO detection alarms over the past decade, but none have had the precision to continuously and accurately measure lower ppm levels of CO. Continuous monitoring of carbon monoxide at 35 to 200 ppm levels presents a challenge to any commercially available CO detector technology. Continuous carbon monoxide monitoring is critical in the household, industrial and military sectors. At present, three technologies are used in the manufacture of carbon monoxide alarms. The advantages and disadvantages of each method are outlined below.

Heated metal oxide based detectors for carbon monoxide (CO):

[0010] Semiconductor based sensors use heated tin dioxide thin films on a ceramic substrate. CO is oxidized on the high temperature surface. The current increases as the tin dioxide is exposed to carbon monoxide. Microchip controlled electronics detect the change in current and will sound an alarm when levels of CO, as measured by the current, exceed a defined threshold. These sensors operate at high temperatures, greater than 400⁰C, contributing to high power consumption. This high temperature makes them susceptible to false signals generated by chemically similar analytes. The following is an advantage: inexpensive and easy to produce. The following are disadvantages: high power consumption, slow cycle time; oxygen contamination; susceptible to false positive signals; and requires heating to regenerate system. The above outlined technology remains insufficient to present a complete solution for the continuous detection of carbon monoxide in the lower ppm ranges.

Heated molybdenum oxide (MoO₃) CO sensor.

[0011] Thus, research has been continuing in investigations of alternative sensors. For example, molybdenum oxide (MoO₃) thin films prepared by sol-gel and RF magnetron sputtering processes were previously employed in the development of CO sensors' as described in "Carbon Monoxide response of molybdenum oxide thin films deposited by different techniques," by E. Comini, G. Faglia, G. Sberveglieri, C. Cantalini, M. Passacantando, S.

Santucci, in Sensors and Actuators B 68, pp. 168-174 (2000), denoted herein Reference 2. The RF deposited films had a needle-like structure with longitudinal dimension ranging from

200-400 nm. The response was measured by applying a constant potential of 1 V to the sensing layer and registering the resistance with a picoammeter. This CO sensor operates as a chemiresistor. Figure 8 of Reference 2 shows the dynamic responses of a sol-gel sensor and an

RF sputtered sensor at 300° C to a square concentration pulse of 30 ppm CO. The current changes shown in Figure 8 of Reference 2 are the picoamp range. This range has the disadvantage that it tends to be almost impossible to record such weak output for a continuous monitor without the use of non-portable, highly sophisticated equipment. The sensor in Reference 2 was operated at 300° C. Heating a sensor substrate consumes much electrical power. This is a disadvantage for a portable device.

[0012] Further, research has been continuing in methods of depositing metal oxides. For example, the authors of "Size-selective electrodeposition of meso-scale metal particles: a general method," by H. Liu, F. Favier, K. Ng, M.P. Zach, R.M. Penner, in Electrochimica Acta

47 pp. 671-677 (2001), denoted herein Reference 3, demonstrated that monodisperse nanoparticles of molybdenum dioxide can be grown on a conductive surface using a pulsed voltammetric technique. Figure 5 of Reference 3 shows the scanning electron micrograph of molybdnum dioxide metal nanoparticles on graphite basal plane surfaces. As shown in that Figure, the nanoparticles have with an apparent size, as indicated by a 1 micrometer scale line, of 100-200 nm. It is also possible to oxidize an existing metal film.

[0013] Notwithstanding the above teachings, there is a strong requirement for an alternative technology to heated metal oxide sensors. In particular, there remains a need for gas sensors having low power requirements, broad environmental operating range, fast response time, high selectivity and high sensitivity.

SUMMARY OF THE INVENTION

[0014] The Inventors have discovered that any known metal oxide sensor that operates due to thermal activation can be preferably operated without a thermal excitation by application of alternative energy activation methods. Application of alternative energy methods provides similar changes in surface energy states and conductivity mechanisms of metal oxides. For example, application gate bias in a thin-film transistor (TFT) architecture eliminates the thermal requirements of described sensors. FIG. 1 shows how the gate bias can manipulate the metal oxide surface energy and change the energy barriers through a grain boundary. Additionally, the excitation methods could be optical, magnetic or combinations thereof. [0015] In some embodiments, the sensing mechanism of the above-mentioned metal oxide sensor is independent of operating temperature. In some such embodiments, the semiconducting channel is electronically manipulated. As an example, the sensor operates at - 6OF and 140F under the same gate bias.

[0016] The TFT architecture may include a semiconducting thin film that includes a compound capable of chemical interaction with an analyte gas. The compound is preferably a metal oxide. The chemical interaction is preferably electron transfer. Thus, according to an embodiment of the present invention, a gas sensor includes a metal oxide. The metal is preferably a transition metal, more preferably a Group 6B element, still more preferably molybdenum.

[0017] The Inventors have discovered that the built in signal amplification of a thin film transistor (TFT) architecture overcomes the problem of weak output in a sensor for detecting an analyte gas.

[0018] Further, in combination or alternatively, according to an embodiment of the present invention, a gas sensor operates with an electron transfer from an analyte gas. This electron transfer could be the result of a catalytic reaction on the surface of the sensor. The electron transfer may be electron donating. Alternatively, the electron transfer may be electron withdrawing. The analyte gas is preferably carbon monoxide.

[0019] Still further, in combination or alternatively, according to an embodiment of the present invention, a gas sensor includes a sensor architecture using a semiconductor metal oxide. In the sensor architecture, the metal oxide may be in the form of a thin film. For example, the metal oxide may be in the form of a nanoparticle or a network of nanoparticles. The sensor architecture preferably includes a thin film transistor architecture.

[0020] Yet further, in combination or alternatively, according to an embodiment of the present invention, a gas sensor includes at least one metal oxide nanoparticle made by a growth method. The thin film may grown in situ by first depositing a metal followed by its oxidation. The metal layer growth method may include electrochemical growth, sputtering, metal evaporation, solution processing, including sol gel and nanoparticle solutions. The metal layer deposition method could be combinations of the above mentioned techniques. After depositions of the metal layer, a post process oxidation is required to convert the metal to its corresponding metal oxide. This oxidation can be performed thermally, chemically, electrochemically or combinations thereof. Electrical conductivity through this metal oxide film may be dominated by grain boundary interfaces.

[0021] Still yet further, according to an embodiment of the present invention, a gas sensor includes a microprocessor.

[0022] Further, the present invention relates to gas sensors that operate by applying an external force to tune the sensor by changing the surface energy of the semiconducting layer and change the electrical transport through the semiconducting layer. Tuning the surface energy could be accomplished by applying a gate. Additionally, other methods could be used to tune the detection of an analyte, including, optical excitation, chemical dopants, surface chemical layers, magnetic fields and combinations thereof. Additionally, the application of an external force, such as a gate bias, helps to eliminate or greatly reduce the requirement for a heated substrate surface. [0023] Further still, according to an embodiment of the present invention, a gas sensor includes RF integration for remote sensing.

[0024] According to any one of the above-described embodiments of the present invention, the gas sensor may act as a dosimeter.

[0025] Thus, according to an embodiment of the present invention, a gas sensor for detecting an analyte gas may include a first contact, a second contact, a semiconducting layer, an insulating layer, a substrate, and a third contact. The third contact preferably serves as a gate contact. The gate contact may be the substrate. The semiconducting layer preferably includes a compound capable of chemical interaction with the analyte gas. The chemical interaction is preferably electron transfer. The insulating layer, the semiconducting layer, first contact, second contact, and third contact are preferably arranged in an architecture predetermined such that the sensor detects a variation in the level of the analyte gas (labeled e.g. CO) as a variation in a current between the first and second contacts occurring when an electron transfer event takes place between the compound and the analyte. hi this embodiment, CO is oxidized to carbon dioxide (CO₂). When this reaction takes place an electron is transferred from CO into the metal oxide surface. This electron transfer can be facilitated when a gate voltage is applied across the first contact and the third contact . The transport of the electron (or hole) created through this electron transfer reaction between the analyte and compound can be manipulated when a gate voltage is applied across the first contact and the third contact. Manipulation of the electron transport may be due to changing energy barrier heights at the grain boundary layers between nanoparticles. Application of a gate bias may reduce thermal requirements for the sensor and allow operation at room temperature and below without an external or internal heated substrate.

[0026] Further, according to an embodiment of the present invention, a gas sensor may include a signal amplifier that includes a thin film transistor that includes a semiconducting thin film that includes an oxide of molybdenum. The signal amplifier may be external. The signal amplifier may be the thin-film transistor sensor. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] For a more complete understanding of the present invention, and the advantages thereof, reference is now made to the following description, taken in conjunction with the accompanying drawings, in which: [0028] FIG. 1 illustrates tuning the surface energy for reaction and the grain boundary dependent electron transport by application of a gate bias;

[0029] FIG. 2 illustrates an embodiment of the present invention;

[0030] FIG. 3 is a plot illustrating the sensitivity based on internal amplification of a gas sensor according to an embodiment of the present invention; [0031] FIG. 4 illustrates the gated metal oxide sensor architecture;

[0032] FIG. 5 is a schematic diagram illustrating additional processing steps for making a gas sensor using the substrate as a third gate contact according to an embodiment of the present invention;

[0033] FIG. 6 is a schematic diagram illustrating a gas sensor for remote applications according to an embodiment of the present invention;

[0034] FIG. 7 is a plot of the response of an exemplary gas sensor according to an embodiment of the present invention; and

[0035] FIG. 8 is another plot of the response of an exemplary gas sensor according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0036] Referring now to FIG. 1, FIG. 1 denotes the tunability of the sensor. FIG. IA shows an exaggerated surface topography of a metal oxide between two contacts. The metal oxide and its contacts are present on top of an insulating layer that separates a third contact. The surface topography creates a grain boundary where conduction of an electron or hole is dominated by the energy barrier to get from one grain to an adjacent grain. Referring now to FIG. IB, we can see that there is an energy barrier associated with the grain boundaries depicted in FIG. IA. In this embodiment, we can tune the conductivity through the metal oxide layer. This is accomplished by changing the height of the energy barrier at each grain boundary. When a surface reaction takes place, for example the oxidation of CO to CO₂, the electron that is transferred is trapped near the reaction site between these energy barriers. Previous examples have shown that high heat can provide sufficient energy to transport that electron over the energy barrier. In this embodiment, the Inventors reduce the barrier height by application of an electrical bias to contact 3, thereby eliminating the requirement of heat. Referring to FIG. 1C, the Inventors show how the surface reaction energy can be tuned. In

FIG. 1C, the metal oxide is semiconducting and has a band gap. This band gap is the energy it takes to promote an electron from the valence band up into the conduction band. Previous examples have demonstrated that high heat is required to create a favorable distribution of electrons within the valence and conduction band for a surface reaction to take place, such as the oxidation of CO to CO₂. In this embodiment, the distribution of electrons witibdn the valence and conduction bands is tuned by application of a gate bias.

[0037] Referring now to FIG. 2, the gas sensor shown in FIG. 2 exemplifies a gas sensor for detecting an analyte gas that includes a first contact (labeled e.g. metal contact pad), a second contact, a semiconducting layer (labeled e.g. metal oxide nanoparticle layer), an insulating layer (labeled e.g. dielectric insulator), a substrate (labeled e.g. Si gate), and a third contact (not shown) contacting the substrate. The third contact preferably serves as a gate contact. The semiconducting layer preferably includes a compound capable of chemical interaction with the analyte gas. Still referring to FIG. 2, and as described further below, the insulating layer, the semiconducting layer, first contact, second contact, and third contact are preferably arranged in an architecture predetermined such that the sensor detects a variation in the level of the analyte gas (labeled e.g. CO) as a variation in conductivity between the first and second contacts occurring when a gate voltage is applied across the first contact and the third contact.

[0038] Disclosed herein, as an exemplary gas sensor, is a miniaturized, low power, rapid responsive CO sensor based on metal oxide nanoparticle networks applied to thin-film transistor ("TFT") architecture. Certain metal oxides are n-type semiconductors that show an increase in conductance due to the transfer of electrons resulting from oxidation or reduction of an analyte gas. The change in conductance is proportional to the concentration of the analyte gas. The nanoparticle network approach provides considerable improvement in sensitivity and selectivity over the most successful commercial technology of CO detectors based on metal oxide films due to the following reasons. The nanostructured interaction of a metal oxide nanoparticle network provides higher sensitivity compared to commercial detectors. Operation in lower ppm ranges (e.g., 0-200 ppm) is made possible, which cannot be achieved in thick tin oxide thin film based commercial detectors. Further, the metal oxide nanoparticle network can be refreshed by ambient oxygen. Still further, the TFT design enables increased sensitivity due to the built-in gain. The gain comes from a non-linear current vs. voltage curve, characteristic of semiconductors. Yet further, salient characteristics of the proposed device that present an improvement over existing technologies include the following advantages: built-in gain through TFT architecture; on-chip design and integration; fast response and continuous monitoring; built in refresh through chemistry and gate voltage; quantitative response; and time integrated response for cumulative exposure. The TFT architecture can be made using CMOS processing for highly parallel and low-cost manufacturing. [0039] The ability to apply a gate bias can control the sensitivity of the device. Previous work by Fan et al. [Fan, Z. et al. "ZnO nanowires field-effect transistor and oxygen sensing property" Appl. Phys. Lett. 2004, (85) 24, 5923-5925, herein denoted Ref. 4] demonstrated an oxygen sensor made from a zinc oxide (ZnO) nanowire field effect transistor. This work showed that the device sensitivity to oxygen changed as a function of gate bias. This device, however, operates with fundamental differences with respect to this invention. The oxygen in this device is physically adsorbed to the ZnO surface. The oxygen adatom (adsorbed atom) is electronegative and removes electron density from the semiconductor changing the distribution of current carriers. This change in carrier concentration is the same thing that happens when a gate bias is applied to any semiconductor and is by definition a change in current by field effect. This previous devices operates as a ChemFET. Their device will not operate in ambient environmental conditions due to surface saturations at normal oxygen concentrations. It is not a sensor, but rather a physical change in transistor response due to a change in environment.

[0040] The present invention operates via chemical reactions at a metal oxide surface. These chemical reactions can add (through oxidation) or remove (through reduction) electrons from the semiconducting layer of the sensor. The change in number of electrons will change the current through the device. The removal of electrons will always cause a change in current even if the material is a poor semiconductor whereas oxygen adsorption changing electron density or distribution will not. [0041] Previously, Dalin [Dalin, J. "Fabrication and characterization of a novel MOSFET gas sensor" Final Thesis at Linkopings Institute of Technology, Fraunhofer Institute for Physical Measurement Techniques, Frieburg, Germany, 6-5-2002, LiTH-ISY-EX-3184, herein denoted Ref. 5] showed that a gate bias can modulate the current through a tin oxide (SnO₂) gas sensor operated at 200⁰C or 28O⁰C. This is a heated sensor. While the current level through the sensor changed, the sensitivity did not. For example, Ref. 5, Figure 6.5 shows the actual current values to different concentrations of CO at variable gate bias. This figure shows a noticeable change in current of the sensor response at each gate bias. In Ref. 5, Figure 6.6, the response of a heated SnO₂ sensor to CO is plotted as initial current over actual current, hi this figure, there is a measurable, but small change in sensitivity as a function of gate bias. If one inspects the change in response, one can see that the baseline of the response changes nearly the same as the exposed response. This indicates that the gate bias changes the current level through the sensor but does not increase the sensitivity. The sensor of Ref. 5 did not operate at lower temperatures.

[0042] The gated metal oxide sensor of the present invention does not require heat for operation. This gated metal oxide sensor operates from -6O⁰C to greater than 100⁰C. It demonstrates an increased response at lower temperatures. Technical approach - chemistry:

[0043] Metal oxide chemistry is a driving force for this invention to sense an electron transfer from a surface reaction of an analyte gas. The analyte gases include, but are not limited to, carbon monoxide (CO) and other electron donating and or electron accepting species. The success of the sensor requires a maximized semiconductor response in the metal oxide. Any metal oxide thin film/nanoparticle system is deemed suitable with the present invention, but a more specifically transitional metal oxide such as molybdenum oxide (MoO₃) is used due to its unique properties towards CO. Metal oxides exist in several forms. For example, molybdenum oxide could be MoO, MoO₂, MoO₃ depending on the oxidation state of the metal. The invention cites MoO₃ as a specific example, but the invention is applicable to other oxides of molybdenum. MoO₃ is an n-type semiconductor that will oxidize CO through electron transfer, which causes a measurable change in resistance. Molybdenum trioxide contains molybdenum in its hexavalent state. Hexavalent molybdenum has no electrons in its 4d orbitals. As a result, oxidation of carbon monoxide involves an electron transfer of an electron from CO to Mo⁺⁶. Following this initial electron transfer step, several reaction pathways are possible for the subsequent oxidation of carbon monoxide to carbon dioxide.

Most likely, these pathways involve rapid, free radical chain steps, which translate to a fast, responsive sensor. [0044] The invention includes, but is not limited to, molybdenum oxide nanoparticles for CO detection. Molybdenum oxide presents certain unique properties suitable for the present invention. Other metal oxides may be used with solid state sensor design, but none have equivalent properties to molybdenum oxide. The two other metals in Group 6B are Chromium (Cr) and Tungsten (W). They have a similar chemistry to molybdenum oxide. The top of the period (CrO₃) will be more reactive. This reactivity comes at a cost. The more reactive species will create a more stable product increasing the difficulty of reversing the reaction, i.e., refreshing of the sensor. The increased reactivity will also reduce the sensors selectivity. Conversely, the bottom of the row (WO₃) will be less reactive reducing sensitivity but more easily reversed. Molybdenum oxide has the highest reactivity combined with the greatest ease of reversibility. Metal oxides outside Group 6B have not demonstrated the required sensitivity or selectivity towards CO. This includes tin dioxide.

[0045] Chemical interferences to the proposed sensor systems, such as water vapor, are not problematic. Other possible contaminants in a vehicle or in an industrial or household environment such as carbon dioxide, nitrogen dioxide or saturated hydrocarbons are not expected to interfere with this sensor system. They will not bind to MoO₃ through an electron transfer, and will therefore not produce a signal. It is possible to design a metal oxide material that would be specific and selective to a particular analyte. At the same time, it is possible to design a metal oxide material that will exclude sensitivity of a particular analyte. Additionally, a gate bias applied to a metal oxide material could enhance selectivity toward chemically similar analytes.

[0046] The use of spherical nanoparticle films has several advantages compared with nanoparticles of other morphologies. Spherical nanoparticles have an increased percentage of active surface atoms (diameters ranging from 5-300 nm). The atoms in the middle of the particle, called the "bulk," do not contribute electronically to any reactions or binding events.

When a reaction or binding event takes place, these surface atoms make a greater contribution to the overall electronic structure of the nanoparticle. This increased contribution translates directly into an increased signal. Quasi-spherical nanoparticles have similar surface to bulk ratios. This would include a nano-"bump" on a substrate or small grains of material on a surface. Additionally, very thin-films of metal oxides will show increased sensitivity due to high surface to bulk ratios. Maximum sensitivity should occur at a thickness near the Debye- length for the particular semiconducting material; however, overall conductivity must also be considered. TFT sensor design:

[0047] Due to the semiconducting properties of MoO₃, the invention employs a thin-film transistor (TFT) architecture to maximize the signal output of the CO sensor. Molybdenum Oxide is an n-type semiconducting material. This means that conduction through the material can be manipulated by a third terminal contact commonly called a gate. Because it is an n-type semiconductor, the resistance will decrease as we move to a positive gate voltage. An electron transfer from the oxidation of CO to CO₂ will increase the number of electrons in the MoO₃ film, therefore increasing the number of carriers and also the current through the device. This is effectively the same as applying a positive gate voltage. [0048] Referring again to FIG. 2, FIG. 2 illustrates a metal oxide semiconductor thin film transistor for CO detection. An electron transfer takes place when CO is oxidized on the surface of the metal oxide. This surface reaction changes the carrier concentration of the n-type semiconductor and is measured as a change in current.

[0049] The semiconducting nature of the MoO₃ contributes to an increased sensitivity in the sensor. The TFT architecture has inherent gain, which is to say that a very small change in gate voltage (for example an electron transfer caused by an oxidation of CO) creates change in current. An electron transfer event from an oxidation of a gas like CO into the metal oxide nanoparticle framework is depicted in FIG. 1. As seen in FIG. 1, this electron transfer event causes a change in current that can be up to several orders of magnitude depending on the slope of the current vs. gate voltage curve. This curve is based on an n-type semiconducting material. The invention can be manipulated for a reducible gas by making the semiconducting channel p-type. In this type of device, the surface would provide an electron to allow an oxidized chemical species to be reduced. Changing from n-type to p-type would require a change in the chemistry of the metal oxide by changing the base metal, alloys of other metals or doping with small amounts of additional materials.

[0050] Referring now to FIG. 2, FIG. 2 illustrates Current Vs Gate Voltage for an n-type semi-conductor showing the high inherent change in current (ΔI) by changing a small gate voltage (ΔV) due to the electron transfer from CO.

[0051] As seen in FIG. 2, movement from point A to point B requires only a small change in gate voltage. Thus a small change in voltage results in a large change (nearly three orders of magnitude) in current, thereby increasing the inherent (ΔV/ΔI). This signal amplification allows for unmatched levels of sensitivity, otherwise not achievable in a chemresistor transduction. In addition, manipulation of the gate voltage allows the sensor to operate in the most sensitive area of the I-V curve, i.e., where the slope, and therefore the sensitivity of the device is the steepest.

[0052] The TFT architecture may be processed by a method that includes electrochemically depositing MoO₃ nanoparticles on a conductive substrate as described below. Further, the metal oxide may be deposited by a solution method. Still further, the metal oxide may be grown by oxidizing a thin metal film or metal nanoparticle film.

[0053] The present inventors contemplate growing nanoparticles with two parallel approaches using, for example, a Gamry Potentiostat with PC interface. In approach I the process includes direct deposition of the nanoparticles on a conductive substrate, followed by transforming a surface portion of the conductive substrate into an insulating layer. In approach I, the two step process includes indirect deposition of the nanoparticles on a conducting substrate and removal of the nanoparticles from the conducting substrate followed by deposition of the nanoparticles on an insulating substrate. While the indirect and direct deposition are described by way of example as electrochemical growth it will be understood that alternative deposition methods known in the art are contemplated, for example sputtering, thermal evaporation, electron beam evaporation, and the like. Further, in accordance with the deposition method, initial deposition may occur on any suitable surface, selecting from among conductors, insulators, and semiconductors. Processing Steps:

[0054] 1. Approach I: Electrochemical growth of metal oxide nanostructures:

[0055] Referring now to FIG. 3, FIG. 3 illustrates electrochemical processing steps to create a MoO₃ TFT-based CO sensor. Step A will electrochemically deposit the nanoparticle film on a Si substrate. Step B will thermally grow a Gate oxide. [0056] Still referring to FIG. 3, the present inventors will start with a conductive silicon substrate and directly grow the nanoparticles until a conductive film is achieved arriving at step B. The silicon substrate will eventually serve as a global back gate. The electrochemical fabrication of the metal oxide nanostructure will be carried out by a two step procedure involving a nucleation step (applying a high negative voltage for a short period of time <10sec) and a prolonged growth step (up to 10 minutes) at lower negative voltages in an aqueous/organic metal oxide solution. The electrochemical fabrication may be carried out in a constant voltage mode (chronoamperometry) or in a constant current mode (chronopotentiometry).

[0057] The electrochemical growth will be followed by a thermal oxidation to grow silicon oxide under the molybdenum oxide layer (step B). This 1000° C thermal step will serve two functions. First, it will produce a high quality, thermal silicon oxide gate insulator, and avoid the alignment issues of photolithography. Second, the thermal layer will anneal the semi conducting molybdenum oxide film to increase conductivity. Higher conductivity will lessen the need for complex electronics to eliminate noise when measuring low current signals. As an alternative to direct electroplating the metal oxide, a second parallel approach is also feasible. [0058] 2. Approach II: Indirect electrochemical growth of metal oxide nanostructures:

[0059] The present invention will involve a procedure, wherein the molybdenum oxide will be nucleated and grown onto a conductive substrate (e.g., freshly cleaved graphite surface) followed by removal of the nanoparticles from the conductive substrate and collection in a solution phase. This will allow us to solution deposit the nanoparticles onto any insulating substrate. We will create conductive nanoparticle films using drop casting or Langmuir-

Blogett solution based techniques. Both of these techniques are known to deposit continuous nanoparticle films. These techniques are depicted in FIG. 4. Once the molybdenum oxide films are formed, we will move forward into final device processing.

[0060] Referring again to FIG. 4, FIG. 4 illustrates indirect electrochemical processing steps to create a MoO₃ TFT-based CO sensor. Step A will electrochemically deposit the nanoparticle film on a conductive (e.g., graphite) substrate followed by a harvest step wherein the nanoparticles are dispersed in a liquid suspension. The nanoparticles could then be deposited by as a dropcast or a Langmuir Blodgett film on an insulating substrate.

[0061] 3. Approach III: metal deposition followed by oxidation. [0062] The present invention can involve a procedure, wherein the molybdenum is deposited onto the substrate. This metal layer could be deposited as a liquid suspension of metal nanoparticles, a thermal evaporation, sputtering, electron beam evaporation or other technique known in the art. This metal layer will be deposited onto an insulating substrate followed by oxidation to the metal oxide. The metal layer could be oxidized by heating in an oxygen containing environment. Controlling the temperature and time of oxidation will control the conductivity of the metal layer. Metals can oxidize at a temperature between 100⁰C and 1400⁰C. Metals can also be oxidized chemically or electrochemically. Once this layer of metal oxide is formed we can move on to final processing.

[0063] 4. Final Photolithography process steps and device assembly:

[0064] Referring now to FIG. 5, FIG. 5 illustrates photolithography processing steps for the MoO₃ CO sensor.

[0065] Upon completion of Approaches I, II, or III, several additional steps will be necessary to complete the devices. Referring again to FIG. 5, first metal contact electrodes creating the source and drain will be evaporated onto the MoO₃ film using standard photolithography and lift-off (step C). These contacts will be used to measure current or resistance through the active area of the device. Next, the active area of the device will be patterned using photolithography as shown in Step D. The photoresist will create a protective film for the subsequent reactive ion etching. This etch step will remove the molybdenum oxide from unwanted areas on the chip, prevent cross-talk between adjacent sensors, and define the dimension of the actual end device. The final step will be to open a hole in the insulating silicon oxide followed by metallization to allow a gate contact with the underlying substrate

(shown in Step E). Following completion of processing a wafer it will be cleaved into individual sensor elements. The sensors will then be mounted to a multi-pin header using a suitable epoxy. The three contacts (source, drain and gate) will then be wire bonded to a header pin using packaging methods known to the one skilled in art. Microprocessor Development and device integration:

[0066] The microprocessor developed as a part of this invention will have the ability to manipulate the gate voltage, measure current through our CO sensor, compute a CO concentration, and drive a digital display and output to an alarm. The device box will likely contain both a piezo-based audible alarm and an LED based visual alarm. The microprocessor will be required to run more than one input channel. An inactive reference sensor will be incorporated in the device to cancel aging and temperature drift. The reference channel will be measured along with the active sensor during each sampling cycle. Data samples are averaged to filter noise and converted to CO concentration levels. The on-board LCD display will be updated every 20 seconds or less. [0067] Referring now to FIG. 6, FIG. 6 illustrates radiofrequency (RF) integration for remote applications. [0068] The same architecture will be fully compatible with a low-power RF link for remote readout. Referring again to FIG. 6, FIG. 6 shows a possible arrangement. The software capability would handle normal "I'm alive" and battery status reporting, as well as change-of- measurement readout. An approximate duty cycle of 0.1% activity is assumed with reporting of CO changes occurring at approximately 10 second intervals. This holds total power consumption to approximately 2 mW, occurring at the 0.1% interval. This low average power consumption enables long battery life.

[0069] The present invention will be more easily and fully understood by the following example. The example is representative of a gas sensor in accordance with one embodiment of the present invention.

EXAMPLE

[0070] A sensor was prepared by growing molybdenum trioxide, as an exemplary sensing compound, in a thin film arranged as part of a thin film transistor architecture. The growth was via electron beam evaporation of molybdenum, followed by thermal oxidation of molybdenum. The structure of the film was characterized using a scanning electron microscope (SEM). The film had a nanoparticle structure. The deposited metal film had a thickness of less than 20 nm.

[0071] FIG. 7 shows the response of the sensor to a continuous flow of carbon monoxide at 50 ppm inside a sealed chamber. The sensor was operated at room temperature. The response was measured as the normalized ratio of R, the resistance in the presence of carbon monoxide, to R₀, to the resistance in the absence of carbon monoxide, as a function of time. The response was determined for two different values of gate voltage, +5 V, and -5 V. The lines marked ON and OFF indicate the dose of CO being turned on and off. The upper curve (-5Vg) shows a response to CO. The lower curve (+5Vg) shows no response to CO. The results demonstrate that the response of the sensor may be tuned by varying the gate voltage.

[0072] FIG. 8 shows the response of the same sensor to a continuous stream of carbon monoxide at increasing levels of concentration of 2ppm, 5ppm, 10 ppm, and 20 ppm. The sensor was operated at room temperature. The response was measured as the normalized ratio of R, the resistance in the presence of carbon monoxide, to R₀, to the resistance in the absence of carbon monoxide, as a function of time. The results demonstrate the sensitive response of the sensor to low levels of gas. The response to CO is linear. [0073] These above-described results further demonstrate that the requirement of heating the sensor is eliminated so as to operate at room temperature (22 degrees C) by using a gate bias.

[0074] The present inventors have discovered that operation of at temperatures lower than room temperature is also possible, for example -60 degrees F. Thus, the sensor may operate at atmospheric temperatures encountered from ground level to up to 40,000 feet, and thus is adapted for use in an air plane or other high altitude application. Thus, a method of operating the sensor may include adjusting the gate voltage according to the temperature.

[0075] The present inventors have further discovered that the gate bias can be tuned to different analyte gases at a wide range of concentrations.

[0076] Thus, a method of operating a sensor according to an embodiment of the present invention may include tuning any one or combination of the gate bias and the sensing compound so as to select the analyte.

[0077] Although the present invention and its advantages has been described in detail, it should be understood that various changes substitutions and modifications can be made herein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims

Claims:

1. A sensor for detecting the presence of an analyte gas, wherein the sensor functions by tuning the surface energy of a semiconducting layer and modulating the conductivity of a semiconducting layer, the sensor comprising: a substrate; an insulating layer; a semiconducting layer separated from the substrate by the insulating layer; a first contact contacting the semiconducting layer; a second contact contacting the semiconducting layer; and a third contact contacting the substrate; wherein the semiconducting layer comprises a compound capable of chemical interaction with the analyte gas; wherein the third contact is a conductive layer on the substrate separated from the semiconducting layer by an insulator; and wherein the insulating layer, the semiconducting layer, first contact, second contact, and third contact are arranged in an architecture predetermined such that the sensor detects a variation in the level of the analyte gas as a variation in a current between the first and second contacts occurring when a voltage is applied across the first contact and the third contact.

2. The sensor according to claim 1 , wherein the compound comprises an oxide of a metal.

3. The sensor according to claim 2, wherein the metal comprises a transition metal.

4. The sensor according to claim 3, wherein metal comprises a mixture of metals

5. The sensor according to claim 3, wherein mixtures are secondary, tertiary, quadranary mixtures, and wherein additional metals act as dopants to change orbital energy levels.

6. The sensor according the claim 3, wherein the transition metal comprises a Group 6B element.

7. The sensor according to claim 6, wherein the Group 6B element comprises molybdenum oxide.

8. The sensor according to claim 1, wherein the chemical interaction comprises electron transfer.

9. The sensor according to claim 8, wherein the electron transfer comprises electron donation by the analyte gas to the compound.

10. The sensor according to claim 8, wherein the electron transfer comprises electron withdrawal by the analyte gas from the compound.

11. The sensor according to claim 1, wherein the analyte gas comprises carbon monoxide.

12. The sensor according to claim 1, wherein the semiconducting layer comprises a thin film.

13. The sensor according to claim 12, wherein the thin film comprises a plurality of nanoparticles comprising the compound.

14. The sensor according to claim 13, wherein the nanoparticles are spherical.

15. The sensor according to claim 13, wherein the nanoparticles are non spherical.

16. The sensor according to claim 2, wherein the continuous metal oxide has a rough surface where topography defines grain boundaries

17. The sensor according to claim 1, wherein gate bias affects conductivity through the semiconducting layer.

18. The sensor according to claim 16, wherein gate bias affects energy barriers between grain boundaries.

19. The sensor according to claim 13, wherein the nanoparticles have homogeneous size.

20. The sensor according to claim 1 , wherein the architecture comprises a thin film transistor architecture.

21. The sensor according to claim 1, wherein the sensor is made by a method comprising growing nanoparticles comprising the compound.

22. The sensor according to claim 12, wherein the growth of the semiconducting layer comprises electrodeposition.

23. The sensor according to claim 12, wherein the growth of the semiconducting layer comprises sputtering.

24. The sensor according to claim 12, wherein the growth of the semiconducting layer occurs directly on the insulating layer or a conducting precursor of the insulating layer.

25. The sensor according to claim 12, wherein the growth of the semiconducting layer occurs indirectly away from a surface of insulating substrate or a precursor of the insulating substrate.

26. The method according to claim 25, wherein the method further comprises applying the nanoparticles to the surface.

27. The sensor according to claim 1, wherein the sensor is made by a method comprising deposition of a metal thin-film followed by transformation to its metal oxide.

28. The sensor according to claim 27, wherein the transformation to metal oxide is made by thermal annealing.

29. The sensor according to claim 27, wherein the transformation to metal oxide is made by chemical reaction.

30. The sensor according to claim 27, wherein the transformation to metal oxide is made by electrochemical reaction.

31. The sensor according to claim 21 , wherein the method further comprises thermally annealing the nanoparticles.

32. The gas sensor according to claim 1, wherein the gas sensor further comprises a microprocessor adapted for manipulating the voltage.

33. The gas sensor according to claim 1, wherein the gas sensor further comprises a radiofrequency link adapted for remote readout of the variations in current.

34. The gas sensor according to claim 1, wherein the gas sensor is adapted to act as a dosimeter.

35. The sensor according to claim 2, wherein the sensing event takes place without any thermal requirement.

36. The sensor according to claim 2, wherein the sensor operates independent of temperature and is a function of gate voltage.

37. The sensor according to claim 1, wherein the sensor can be operated at variable conductivity levels.

38. The sensor according the claim 37 where the electrical current levels are below one (1) nano amp.

39. A method for detecting the presence of an analyte gas using a semiconducting layer, wherein the semiconducting layer is modified in a manner selected from the group consisting of (a) tuning the surface energy of the semiconducting layer, (b) modulating the conductivity of the semiconducting layer, and (c) combinations thereof.

40. The method according to claim 39, wherein the analyte gas is CO.

41. The method according to claim 39, wherein the semiconducting layer has a surface topography.

42. The method according to claim 41, wherein the surface topography creates a grain boundary.

43. The method according to claim 42, wherein there is an energy barrier at the grain boundary.

44. The method according to claim 43, wherein the semiconducting layer is part of a sensor.

45. The method according to claim 39, wherein the semiconducting layer is tuned by application of a bias through a third contact.

46. The method according to claim 43, wherein the energy barrier is tuned by radiation

47. The method according to claim 46, wherein the radiation is light

48. The method according to claim 43, wherein the energy barrier is tuned by a magnetic field.

49. The method according to claim 43, wherein the energy barrier is tuned by changing the semiconducting material of which the semiconducting layer is comprised.

50. The method according to claim 44, wherein the semiconducting layer is a different metal oxide.

51. The method according to claim 44, wherein the semiconducting layer is a mixture of metal oxides.

52. The method according to claim 39, wherein the surface energy tuned by promoting electrons from one energy band to another.

53. The method according to claim 52, wherein the energy bands are molecular orbitals.

54. The method according to claim 39, wherein the surface energy of the semiconducting layer is tuned by application of a gate bias through the third contact.

55. The method according to claim 54, wherein the sensitivity is dependent on tuning surface energy.

56. The method according to claim 39, wherein the surface energy of the semiconducting layer is tuned by exposure to radiation.

57. The method according to claim 39, wherein the surface energy of the semiconducting layer is tuned by a magnetic field.

58. The method according to claim 43, wherein the energy barrier is tuned by changing the semiconducting material.

59. The method according to claim 58, wherein the semiconducting material is a different metal oxide.

60. The method according to claim 58, wherein the semiconducting material is a mixture of metal oxides.

61. The method according to claim 39, wherein the semiconducting layer is part of a sensor that is sensitive to multiple analytes by manipulation of a gate bias.

62. The method according to claim 61, wherein the sensor is selective to individual analytes by application of a specific gate bias.

63. A gas sensor, the gas sensor comprising a signal amplifier, the signal amplifier comprising a thin film transistor, the thin film transistor comprising a semiconducting thin film of an oxide of molybdenum, wherein the gas sensor functions by modifying the semiconducting thin film in a manner selected from the group consisting of (a) tuning the surface energy of the semiconducting thin film, (b) modulating the conductivity of the semiconducting thin film, and (c) combinations thereof.

64. A method for sensing an analyte gas, wherein the method employs the use of the sensor of claim 1.