Technical field

The invention relates to a crystalline nanostructured semiconductor material for the detection of carbon dioxide in a large interval of concentrations. Said crystalline nanostructured semiconductor material is based on indium oxide doped with sodium (Na: ln2O3).

CO2 detection devices are also the subject of invention that use the crystalline nanostructured semiconductor material of the invention.

More specifically, the invention relates to a crystalline nanostructured material based on indium oxide doped with sodium (Na: ln20a) for the detection of aeriform carbon dioxide.

The crystalline nanostructured material of the invention is typically produced in powder to be used as a functional material to obtain sensitive porous films in order to produce solid state sensors. The sensitive film consisting of nanostructured material based on Na: ln2O3 produced according to the invention changes its properties (for instance the conductance) when the carbon dioxide present in the environment is adsorbed on its surface.

The peculiarities that make this material suitable for detecting carbon dioxide are the high reactivity of its surface, by virtue of the presence of sodium, towards this gas and its nanostructured form that ensures a large active surface.

Known art

Carbon dioxide (CO2) is a climate changing gas (climate-change emission), which produces climate change, and which mostly contributes to heating the planet. Climatic databases and simulators monitor climate change due to the anthropic CO2 emission and provide for the future one. According to data collected since 1880, the average temperature of the planet has increased by about 1 .1 °C with strong peaks in some areas (e.g. +5°C at the north pole in the last century), accelerating important transformations of the ecosystem (ice dissolution, raising and acidification of the oceans, loss of biodiversity, desertification) and making extreme phenomena (winds, snow, heat waves) increasingly frequent and acute. With the 2015 Paris agreement, countries around the world have committed themselves to limiting the temperature of global warming to 2°C, compared to pre-industrial levels. To achieve this, the European Union through the European Green Deal (Com/2019/640 Final) has defined new extremely ambitious energy and climatic objectives, which require the reduction of climate changing gases (Greenhouse Gases, GHG) to 55% by 2030 and climatic neutrality in 2050.

Carbon dioxide, in addition to being a climate changing gas, can represent a danger to human health. Exposure to CO2 concentrations greater than 2000 ppm for a prolonged time can lead to headache, sleepiness, concentration difficulties and nausea [1], symptoms attributed in 1983 by the World Health Organization to the Sick Building Syndrome (SBS) [2], Given the toxicity of CO2 at high concentrations, the average exposure of a healthy operator during a work shift of eight hours (TLV- TWA) must not exceed 0.5% (5000 ppm).

Therefore, CO2 monitoring on a wide range of concentrations is of great interest for the control of air quality in an external environment and for safety in internal and confined environments.

The most accurate methods for measuring CO2 concentration are gas chromatography and mass spectroscopy, which, however, need specialized personnel, large and expensive instruments and therefore are not suitable for in situ measures. To date, the most used direct monitoring systems are based on infrared detection principles (IR), as infrared non- dispersive sensors (NDIR). The advantages of these sensors are their precision, resolution and robustness. The disadvantages are the dimensions, the high energy consumption and the relatively high cost. Therefore, alternative sensors are sought that combine precision, resolution and robustness with small size, low costs and consumption, characteristics that would make them more attractive for loT (Internet of Things) networks.

Solid state sensors are very compact devices and can be used for the detection of gases such as CO, H2, NOx, SO2, CH4, C3H8 and volatile organic compounds (VOC). Depending on the sensing mechanism, they can be classified in solid electrolyte, catalytic combustion or semiconductor sensors [3].

(i) Solid electrolyte sensors can monitor relatively low concentrations of different types of gases, but they are short-lived and moreover advanced techniques are requested to obtain selective devices.

(ii) The catalytic combustion sensors are simple and use low-cost technology, but require oxygen to work, easily saturate the response and can be poisoned by chlorinated or sulfur and silicone compounds.

(iii) The semiconductor sensors are mechanically robust, they guarantee long-term stability and work well even in wet conditions but have a non-linear response as a function of the analyte concentration and can be poorly selective [4],

Cheap and sufficiently sensitive solid-state sensors for CO2 detection operating applications are not currently available due to the inert nature of this analyte. However, the semiconductor chemo resistive sensors are the best candidates because they offer a wider interval of structural variables for the engineering of materials to specific applications [5].

In chemo resistive sensors, a nanostructured material film is arranged on an inert support between two electrodes. This semiconductor layer adsorbs selectively the gases of interest present in the atmosphere to which it is exposed and varies in its conductance. This makes quantitative analysis possible in a large range of concentrations, by CO2 adsorption on a special semiconductor layer.

The nanostructured semiconductor metal oxides (MOS) represent an interesting class of semiconductors, since their properties can be engineered, modifying the chemical composition of the metal oxide, the crystalline structure, the size and shape of the nanostructures. For this reason, they were the precursors of numerous technological progresses and are widely used in applications such as catalysis, energy accumulation (batteries, supercapacitors) and energy collection (fuel cells, photovoltaics, hydrogen production by water photolysis), as well as for gas detection [6]. The MOSs can be distinguished according to their composition in binary oxides, such as TiO2, GeO2, Cr2O3, Mn2O3, NiO, CuO, CdO, CeOa, MgO, BaO, ln20a, WOa, V2O3, Fe2O3, Nb2O5, M0O3, C03O4, Ta2O5, La2O3, Nd2O3 and in complex metal oxides, such as Cdln2O4, NiTa2O6, COTa2O6, CuTa2O6, BaSnO3, LnFeO3 [7], These can also be doped or decorated, introducing in the crystalline structure atoms not present in the original semiconductor, in the first case, or by creating catalyst agglomerations on the surface, in the second case. A substance with the same chemical composition can present different structural characteristics. These states include the amorphous state, the glassy state, the nanocrystalline state, the polycrystalline state and the single crystalline state. These affect the stability, the type and quantity of active sites on the surface and therefore on the sensitivity and selectivity of the material [8].

In addition, the morphology of metal oxides can be adjusted to obtain nano-structuring along one or more dimensions (D): OD, 1 D, 2D, 3D structures include for instance nanoparticles, nanowires, nanotubes and nanosheets. The use of films consisting of nanostructured material represents a significant improvement with respect to the compact layers (both thin and thick) of semiconductor, as it increases the surface-volume ratio that is used for gas chemisorption. Each modification in terms of chemical composition, crystalline structure and morphology affects the chemical-physical properties of the material, making the object new compared to what was previously studied and with peculiar characteristics and potential.

The materials for CO2 detection found in the literature, and their qualitative limits, are shown in Table 1 . Table 1: List of materials for detecting CO2 and their limits.

Practically, the possibility of having CO2 chemo resistive sensors that are sensitive, in a range of concentrations useful for indoor and outdoor applications, can still be unsatisfied, which can be used at low temperature and sufficiently selective to work in wet conditions. In addition, the semiconductor materials so far studied in the literature provide for complex and relatively expensive synthesis with respect to the preparation techniques for the materials to date on the market in this field, in addition to requesting, in some cases, specific treatments for the regeneration of the sensible film. ln2O3 is an n-type metal oxide semiconductor that can be synthesized in nanostructured form and cubic crystalline phase. The material is suitable for different applications, such as catalysis, photovoltaic devices, solar cells, phono detectors, thermo-reflecting windows and gas sensors [18- 20],

Recently ln2O3 has been studied in the development of electrochemical processes for CO2 reuse. In particular, it has proven to be an efficient catalyst for CO2 hydrogenation to methanol [21 ], Although the state of the art in this field suggests a high reactivity of its superficial sites towards CO2, pure ln2O3 has shown that it has a very low response for this analyte [14], In fact, the results of the research conducted by the inventors have been unsatisfactory, as their sensitivity did not overtake the best performance reported in literature with other MOS.

Always in the electrochemical field, it has been discovered that alkalis, such as potassium and sodium, are promoters for CO2 hydrogenation, as they facilitate its adsorption on the surface of the catalyst [22, 23]. To date, sodium has been added as a promoter in different catalysts such as those based on iron [22-24], iron-manganese [25], iron-cerium [26], iron-zinc [27], cobalt [28, 29], and FeCoCuAI [30].

However, in the sensor field, the recent materials doped with alkaline metals, such as Na-doped ZnO, have not shown a high sensitivity to this analyte, which discouraged the pursuit of this teaching in the field of research for solid state sensors.

The patent documents JPH0611473A (B2) and CN108061779 (A) describe the use of an indium oxide as a sensitive material for gases monitoring.

JPH0611473 (B2) reports a material based on ln2O3 as an active ingredient of a CO2 sensor and the addition of CaO, SrO, BaO and La2Oa as sensitizers. The additives are dissolved to form an aqueous solution of nitrate or acetate, with which the ln2O3 sensor is impregnated after being thermally treated at 500°C. However, the commercial interest in these materials appears to have failed due to their modest response.

CN108061779 (A) relates to a monitoring system of the internal air using a group of sensors for the detection of carbon monoxide, formaldehyde, carbon dioxide, nitrogen dioxide and VOC. In particular, the document describes a sensor based on an indium oxide and zinc oxide mixture loaded with potassium particles to monitor nitrogen dioxide. Therefore, it is stated that potassium is loaded on the surface, forming clusters, and does not enter the reticular structure of the ln2O3/ZnO mixture as a doping agent. Therefore, K is used for a different method of sensitization, which has unequal effects on the chemical and electronic characteristics of the material [31].

Therefore, none of these documents describes a crystalline nanostructured material based on indium oxide doped with sodium (Na: ln20a) for the detection of wide-ranging carbon dioxide and in real conditions (in dry or damp air).

Patent documents EP0948671 B1 and DE3604594A1 describe indium oxides variously doped with alkaline metals. However, these materials differ in morphological and/or crystallographic structural properties, which have an important impact on their characteristics in the sensor field.

EP0948671 B1 is related to a method for the preparation of metal chalcogenides similar to fullerenes intercalated with metals. In particular, the document describes the synthesis of different metal oxides through the technique described in example 12, wherein an amorphous nanostructured powder (50-60 nm) is obtained. Therefore, it can be deduced that ln20a doped with Na as reported in example 17 has an amorphous structure with dimensions in the range of nm. The aforementioned material differs from this invention in the structure, which appears amorphous and non-crystalline. Although it is possible to obtain stable amorphic structures at room temperature [32], these become unstable at higher temperatures [8], especially if maintained for a prolonged time, as is also required for the stable functioning of a sensitive material. In fact, a non-homogeneous crystallization of the material (for instance the formation of a polycrystalline structure) occurs, increasing the scattering centers and modifying the mobility of the charge carriers [32], The coalescence of the grains that alters the specific surface of the material and, ultimately, its sensitivity occurs. Therefore, the use of nanoparticles of amorphous ln2O3 doped with Na for the realization of sensitive chemo resistive films that are thermo-activated would be in devices with electrical characteristics subject to deterioration and/or with non-reproducible responses. Therefore, the amorphous semiconductor is unsuitable as a sensitive material for a thermo-activated sensor. On the contrary, Na:ln2O3 with nanocrystalline cubic structure that is the subject of the invention has optimal properties for sensor manufacturing, including a high surface area (between 20 and 50 m ² g’ ¹ ) due to the small size of the crystallites and the stability of both structural and electrophysical properties.

DE3604594A1 is related to thin multilayer film sensors. These are provided with a basic layer and at least an additional activation layer, all applied on a substrate obtained by immersion (dip-coating). With this technique it is possible to obtain Na doped ln2O3 layers with a nanometric thickness (for instance 5 nm). These layers can be defined as compact, i.e., the interaction with gas occurs only on the geometric surface (bi- dimensional) of the sensitive film and the electrical conduction takes place in a direction parallel to the maximum effect of banding bands [8,33]. On the contrary, the material subject of the invention presents itself as a Na:ln2O3 crystalline nanostructured powder, with which it is possible to create nanostructured porous layers. The porous layers obtained according to the present invention make their volume accessible to the gases, guaranteeing a high surface-volume ratio and a greater number of active sites obtained from the exposed faces of the crystallites. These characteristics allow a better sensitivity of the devices and will be better detailed in the following. In addition, the sensors described in the DE3604594A1 document are mainly used for detecting traces of gas or vapors, such as H2, O2, O3, H2S, hydrocarbons, NH3, NO2, ketones and aldehydes, but not CO2.

If not specifically excluded in the detailed description that follows, what is described in this chapter is to be considered as an integral part of the detailed description.

Summary of the invention

This invention aims to solve the problems highlighted above and not solved by known art.

Therefore, it is an object of the present invention a nanostructured semiconductor material for the detection of carbon dioxide in a large interval of concentrations. Said nanostructured material is based on an indium oxide doped with sodium (Na:ln2O3). This material is prepared to present a highly reactive surface towards CO2, a necessary condition for a sensitive chemo resistive film.

Another object of the invention is the production process of said nanostructured semiconductor material.

Another object of the invention are the CO2 detection devices that use said nanostructured semiconductor material. These detection devices are sensors suitable for the determination in the air of CO2 levels wherein the sensitive element is or comprises or consists of a nanostructured semiconductor material for the detection of carbon dioxide in a large interval of concentrations and in dry and damp air conditions. Said nanostructured material is based on an indium oxide doped with sodium (Na:ln2O3).

Further object of this invention is an oxide based on indium doped with sodium of formula Na:ln2O3 having chemo-resistance that can be used to create a gas sensor for the detection of CO2 in real environmental conditions such as those in dry air and in damp air.

Further objects, purposes and advantages will be evident from the following detailed description of the invention.

Brief description of the figures Figure 1 : schematic representation of the sensor, which highlights and defines the components according to the list in the "detailed description" paragraph.

Figure 2: scheme of the reaction mechanism on the surface of a type n semiconductor and influence of redox reactions on the inter-grain barrier potential in the air (a) and after exposure to carbon dioxide (b).

Figure 3: a) response of Na:ln2O3 sensor to 500 ppm of CO2 in synthetic air, as a function of the operating temperature; b) calibration curve for CO2 detection at a sensor working temperature equal to 200°C, in dry synthetic air.

Figure 4: Na:ln2O3 sensor response at 500 ppm of CO2 in dry synthetic air at room working temperature (about 25°C).

Figure 5: a) conductance variation after injection of 500 ppm of CO2 at different percentages of relative humidity (RH%); b) influence of RH% on the sensor response at 500 ppm of carbon dioxide.

Figure 6: SEM image of Na:ln2O3 nanoparticles which make up the synthesized powder.

Figure 7: XRD pattern for Na:ln2O3.

Figure 8: a) HR-TEM images of the nanoparticles; b) SAED patterns of Na:ln2O3.

Figure 9: high resolution XPS spectra of a) In3d, b) O1 s, c) Na1 s of Na:ln2O3 powders (dashed line), `n2O3 and NaNO3 (continuous line).

Figure 10: optical characterization by UV-visible spectroscopy of the Na:ln2O3 powder samples (dashed line) and ln2O3 (continuous line).

Figure 11 : repeatability test for Na:ln2O3 sensor at 4 cycles as a function of time. Each cycle includes a period of 2 h of stabilization of the dried synthetic air flow followed by a period of 30 min of a measuring in flow of a mixture of dry synthetic air with 1200 or 400 ppm of CO2. The working temperature of the sensor was equal to 200°C.

Figure 12: response of the sensors of pure ZnO and Na:ZnO, at the operating temperature of (a) 25°C and (b) 200°C, at 1000 ppm of CO2 in dry synthetic air.

Detailed description

As part of this invention with the term "nanostructured material" we mean a material characterized by an internal (crystallite) and superficial (morphology) nanostructure, wherein at least one of the dimensions is in the order of the nanometric scale (1 ÷ 100 nm). The advantages in the use of nanostructured materials for gas detection derive mainly from the large surface-volume ratio, the high specific surface area (20-50 m ² g ^-1 ) and the greatest number of superficial active sites obtained from the exposed faces of the crystals. Thanks to the high surface area, the nanostructured material of the invention has a much higher amount of surface atoms available to the analyte than the non-nanostructured materials. Consequently, the gas sensors based on the nanostructured material of the invention can offer better performance than devices based on compact layers [34],

As part of this invention with the term "compact layer" are defined all those layers whose interaction with gas occurs only on the geometric (bi- dimensional) surface of the film.

As part of this invention with the term "porous layer" are defined all those layers that make their volume accessible to the gases, guaranteeing a high surface-volume ratio and a greater number of active sites obtained from the exposed faces of the crystals.

As part of this invention with the term " substitutional doping" we mean a type of doping wherein the semiconductor atoms are replaced by foreign atoms [31 ].

As part of this invention with the term "interstitial doping" we mean a type of doping wherein foreign atoms, generally light elements, occupy the interstitial sites within the semiconductor lattice [31 ].

As part of this invention with the term "crystallinity of the material" we mean the structural degree of a solid. In a crystal, the arrangement of the atoms or molecules is consistent and periodic.

As part of this invention with the term "unitary cell" we mean the smallest group of atoms or ions which, repeating itself at regular intervals in three dimensions, forms the lattice of the crystalline structure.

As part of this invention with the term "reticular parameters" we mean the reticular dimension of the unitary cell along the three axes (X, Y, Z) and the angle between these axes. In particular, in this invention, the term "reticular parameter a" means the reticular dimension of the cubic unitary cell. The definition of the reticular parameters allows the calculation of the volume of the unitary cell.

As part of this invention with the term "crystallite/s" we mean or are meant the aggregates of monocrystals held together by defective links and with the term "size of crystallites" we mean the average size of the crystalline sub-micrometric domains. This dimension is estimated from the diffractogram of powder through Scherer's law. For instance, when the size of spheroid crystallites is less than 80 nm, it has been shown that sensitivity is directly proportional to 1/D, wherein D is the diameter of the crystallites. The increase in the surface area allows the preadsorbed oxygen to adsorb and desorb a greater quantity of analyte gases. Therefore, the detection response improves thanks to the increase in the crystallization of the material [35]. For instance, as part of this invention, the size of the diameter of the crystallites is about 10 nm.

As part of this invention with the term "micro strain" of crystallites, is meant the internal mechanical stress (strain) due to long-range random deformations, in the various crystalline directions, within each crystallite.

As part of this invention with the term "frit" we mean a powder mixture of silica-based glass oxides loaded with alkaline-earth oxides or with oxides of the IV Group, commonly used in the preparation of enamels for the ceramic industry. Added to the organic precursors, it allows better mechanical resistance of the sensitive film and adhesion to the substrate.

As part of this invention the wording "straight band gap " means the case wherein the crystalline moment of electrons and gaps is the same both in the conduction band and in the valence band; therefore, an electron can directly release a photon. The state of minimal energy in the conduction band and the state of maximum energy in the valence band are each characterized by a certain moment of crystalline (K-vector) in the Brillouin area.

The material band gap of the material can be defined through the Tauc Plot method, thus calculating and adapting the absorption data of the UV-visible spectrum with respect to direct transition energy: ahv = ED (hv - E _g ) ^1/2

Wherein a is the optical absorption coefficient, hv is the photon energy, E _g is the straight band gap and ED is a constant [36]. By tracing the plot of (a v) ² as a function of the photon energy and extrapolating the linear portion of the curve up to absorption equal to zero, the values of the straight band gap (E _g ) of the investigated material are obtained.

The invention relates to a Na:ln2O3 based nanostructured material which is obtained by doping the Na ⁺ ion inside the crystalline lattice of ln2O3. This doping (which can be interstitial or substitutional) affects the structural characteristics of the material, modifying the catalytic properties towards the surface reactions of certain gases, such as CO2. The material can be synthesized through sol-gel technique, co-precipitation, electrochemical deposition, hydrothermal synthesis, spray pyrolysis, magnetron coating (magnetron sputter coater) and irradiation with beam of aggregate gas ions (gas cluster ion beam irradiation) [37], Among those listed, all within the reach of the technical expert in the field, the sol-gel and co-precipitation methods are the simplest and most economic to be achieved, so only the sol-gel synthesis will be exemplified below.

First of all, the indium precursor, preferably indium nitrate, but also indium isopropoxide or indium acetate, is dissolved in deionized water (DI). Useful condition for the synthesis of the material of the invention is that the indium precursor is soluble in aqueous solution. Subsequently, a basic aqueous solution obtained from the sodium-based precursor, preferably NaOH, is added to the previously prepared solution. The resulting suspension is maintained under stirring and heating at a temperature of 50 ÷70°C. The addition of NaOH makes basic the solution containing the indium precursor and induces the hydrolysis reaction of ln ³⁺ , which precipitates in the form of indium hydroxide powder (ln(OH)3). This reaction is favored by heating the solution at a temperature of 50 ÷70°C. In addition, during the growth of the colloid, the sodium is partially introduced into the inorganic network. Na:ln(OH)3 is therefore obtained.

The concentrations of indium and sodium precursor are preferably in molar ratio 1 : 35 ÷70, preferably 1 :35, preferably 1 :40, preferably 1 :50 and preferably 1 :70, for instance 0.10 M indium nitrate and 0.5 M NaOH solutions can be used. This ratio allows to obtain a basic pH necessary for the precipitation of ln(OH)3 and the incorporation of Na ⁺ in its lattice.

The suspension based on In and Na forms instantly and can be washed several times with alcohols and water, such as isopropanol, ethanol and DI preferably using a centrifuge. The resulting precipitate can be dried at a temperature in an interval preferably between 100°C and 250°C. Subsequently the powder of nanostructured Na:ln(OH)3 is subjected to a thermal cycle of 300 ÷500°C to obtain crystalline Na:ln2O3.

The crystalline Na:ln2O3 based nanostructured material can be used for the creation of a porous film, applied as a layer on a gas sensor device. Deposition techniques such as serigraphy, spray-coating, spin-coating, dip-coating, or a combination of these techniques, allow to maintain the morphology of the powder, whose nanostructures are arranged closely. In this configuration, meso-macroporous films are obtained with an unordered form of the pores [38]. The thermal treatment at high temperatures, such as those indicated subsequently, allows to stabilize the porous network and increase the degree of interconnection between the nanostructures to improve the electrical conductance of the film.

According to an embodiment of the invention, the film of crystalline Na:ln2O3 based nanostructured material can be obtained by mixing Na:ln2O3 powder with organic precursors both impregnating (a-terpineol, butyl carbitol or other glycol ethers) and binders (ethyl cellulose, acrylic resin, polyvinylpyrrolidone) and with the addition of an inorganic adjuvant for the sintering of the powder (silica or frit) [39]. The inorganic precursor allows better mechanical resistance of the sensitive film and strengthens adhesion to the substrate. The organic precursors can be included in a total amount from 50% to 80% by mass, according to the desired consistency of the desired compound for an optimal deposition, while glassy oxides are in a percentage between 0.5 and 1 %.

The mixture of powder and solvent thus obtained is deeply stirred to obtain a dough; therefore, it is preferably applied through screen printing technique in the form of a film on the sensor. The favorite coating method is serigraphy, but any other method of film formation can be used, as mentioned below. The film obtained at 250°C for 30 min and then at 450°C for 3h is then dried. The high temperature allows the elimination of the organic solvent, used for the preparation of the dough, and the dissolution of the inorganic adjuvant to have a good adhesion of the sensitive film to the substrate. The thickness of the film of crystalline Na:ln2O3 based nanostructured material deposited according to the embodiment of the invention described can vary between 10 and 30 pm.

The film of crystalline Na:ln2O3 based nanostructured material can be deposited by means of techniques such as serigraphy, spray-coating, spin-coating, dip-coating, or a combination of these techniques, on supports of different nature and composition. For instance, to name a few and without being limiting, the supports can be: alumina substrates, microworked silicon membranes and quartz membranes, flexible plastic substrates or paper.

In Figure 1 is represented the schematization of the device including the sensitive film of crystalline Na:ln2O3 based nanostructured material deposited above two electrodes, to which a potential difference is applied. This film can be thermo-activated by a heater placed on the back of the inert substrate, in order to thermally excite electrons to the conduction band from energy donor levels nearby, thus increasing the number of charge carriers. The thermo-activation also makes reversible the detection process of the gaseous analyte on the surface of the sensitive film. When the sensitive film interacts with the gases present in the surrounding environment, such as oxygen or CO2, through redox reactions, the number of charge carriers changes and there is a variation of the resistance and therefore of the conductance of the film. In fact, the conductance of an n type semiconductor, such as the crystalline nanostructured Na:ln2O3, is proportional to the number of charge carriers (electrons) [40].

The chemo-adsorption reactions of oxygen molecules on the surface of the film based on crystalline nanostructured Na:ln2O3 in the form of oxyanions (O2; O’ and O ² ) lead to the establishment of an inter-grain barrier potential that decreases the conductance of the film. On the other hand, when the CO2 molecules react with the adsorbed oxyanions on the surface of the semiconductor, the barrier potential decreases as schematized in figure 2, and the conductance of Na:ln2O3 increases.

Figure 3 summarizes the electrical characterization of the sensor schematized in figure 1 . The responses were calculated as reported in literature [41 , 42],

As shown in Figure 3(a), the optimal working temperature of the sensor is between 200 and 250°C but can reveal 500 ppm of CO2 even at lower temperatures; therefore, the device is characterized by reduced energy consumption. The calibration curve (response vs. CO2 concentration) in Figure 3(b) demonstrates a high sensitivity in a large range of analyte concentrations (250 ÷5000 ppm).

In fact, the sensors, which are the subject of the invention, have truly peculiar and innovative characteristics compared to the previous MOS reported in the literature, including the ability to transduce even at room temperature (Figure 4). In the temperature interval from 25 up to 150°C, the conductance of the crystalline Na:ln2O3 based nanostructured sensor in the presence of CO2 decreases, unlike the interval 150-450°C wherein it increases, index of a different surface interaction of CO2 with the sensitive film [43].

In addition, the device's response also remains high in wet conditions, as highlighted by the variation of the conductance at different percentages of relative humidity (RH%) in Figure 5.

List of components represented in Figure 1 :

(1 ) Porous sensitive film of crystalline nanostructured Na:ln2O3

(2) Substrate, for instance made of alumina (3) Au electrodes on the alumina substrate front

(4) Heater, for instance in Pt on the back of the alumina substrate

(5) Dielectric coverage on the back of the substrate

(6) Au wires to contact the alumina substrate to the support (7) TO- 39 ("Transistor Outline")

(7) TO-39 support, standardized for a through hole metal device used for integrated circuits

(8) Pin pair to connect the electrodes (3) interdigitated to an electronic board (not shown)

(9) Pin pair to connect the heater to the electronic board (not shown)

Industrial application:

The gas detection based on the chemo resistive effect that occurs in nanostructured semiconductors has been widely studied over the years for the numerous opportunities that this technology is able to investigate, from the environment to well-being and from health to industrial processes. According to the new report of gas sensors market [44], this should be evaluated at 2.1 billion USD by 2027, with an annual growth rate of approximately 8.9%.

The material subject of this invention is a sensor that can be used in different application areas. In addition to those previously mentioned in the state of art such as, monitoring of air quality in closed or confined environments, both private and public, the invention can be exploited in the following sectors: sustainable agriculture, food-packaging, incubators for biological sciences, monitoring of CO2 segregation and conversion, monitoring of the status of a pack of lithium batteries (support for the battery management system, BMS), refrigeration cells, transport of food, beverages, fruit and vegetables, fermentation and birrification, ecological measurements such as breathing soil and CO2 measurement in the environment, also through the implementation of sensors on public transport.

In particular, an atmosphere with controlled CO2 concentrations is fundamental in the greenhouses for carbonate fertilization (usually concentrations vary from 700 to 1200 ppm) [45] and in the packaging for the conservation of fruit and vegetables (up to 25%) [46].

Recently, as reported by Peng and Himenez (2021 ), CO2 could be used as an indicator for the regulation of internal ventilation and, in turn, of the probability of infection from pathogens (e.g., Sars-Cov-2) through aerosol transmission [47],

The applications mentioned above show the need for CO2 monitoring through devices based on functional materials, such as the subject of the invention, in a large concentration interval (250-5000 ppm) with a good sensitivity.

In addition, the availability of a CO2 sensor that combines operation at low temperature and a negligible influence of humidity, allows its implementation in contexts mentioned by favoring their commercial diffusion and intervention capillarity. Finally, the easy integrability of sensors as economic and low energy portable devices favors applications in the loT field.

The crystalline nanostructured Na:ln2O3 material has a good sensitivity to CO2 in a large interval of concentrations, can also be synthesized through simple techniques, for instance sol-gel or coprecipitation synthesis, and uses sodium as a promoter of the indium oxide, that is, a particularly cheap catalyst.

In summary, the characteristics of the material subject of this invention make the technology represented by solid state sensors based on MOS materials competitive compared to the current one based on NDIR optical devices.

Porous films based on crystalline nanostructured Na:ln2O3, subject of this invention are, aware of the inventors, the best materials for the detection of CO2 in terms of intensity of the response and sensitivity (response for analyte concentration unit), in the interval of interest for applications in external, internal and confined environments. As demonstrated in the description of the invention operation, the response at a concentration of interest remains high also in the wet atmosphere, a fundamental prerequisite for the use of a sensor outside the laboratory activity.

The low production and maintenance cost favors the marketing and compatibility with the productive skills of the industrial manufacture. In addition, the production procedures are simple and do not cause significant damage to the environment. In fact, the methods of preparation of the material subject of invention are of low environmental impact and the miniaturization of the devices, together with the modularity of the electronic components, reduces the sensor disposal costs.

Finally, the sensor that can be made according to the invention allows better control of the legislative provisions in force regarding the monitoring of carbon dioxide and the consequent mitigation of the effects related to it.

The following examples are provided to illustrate the invention and are not to be considered limiting of the relative scope.

EXAMPLES

Materials

The indium (III) nitrate hydrate (99.9%) (ln(NO3)3 x 5H2O) was purchased by Sigma-Aldrich, USA. Sodium hydroxide anhydrous pellets (NaOH) and propan-2-ol (C38sO) were purchased by Carlo Erba Reagents SAS. The distilled water (DI) was prepared by Millipore water purification system.

Synthesis, film deposition and characterization of NadnsOs

Na:ln2O3 crystalline nanostructured powder has been synthesized by sol-gel technique. First of all, 0.1 M ln(NO3)3 x 5H2O was dissolved in 60 ml of DI water. Subsequently, 1.2 g of NaOH were added to the aforementioned precursor solution and stirred for 40 minutes at 70°C. The suspension of ln(OH)3 was washed with propanol and DI several times using a centrifuge at 5000 rpm for 2 minutes. The white precipitate was dried at 100°C for 4 hours and subsequently at 200°C for 2 hours. The dried powder was thermally treated at 450°C for 3 hours in the room air. Na:ln2O3 crystalline nanostructured powder has been ground in an agate mortar for 1 minute, to decrease its lumpiness due to the aggregation of nanostructures through weak bonds in the thermal treatment phase. Subsequently, the Na:ln2O3 powder was mixed with a-terpineol, ethyl cellulose and frit to form a homogeneous dough. The organic precursors have been included in a total amount of 80% by mass, while glassy oxides are in a 0.5% percentage. The resulting composite was screen-seated on alumina substrates with golden electrodes interdigitated on the front and a platinum heater on the rear side to thermo-activate the sensitive layer of the device. The printed film was calcinated at 450°C for 3 hours in the air. The thickness of the film thus obtained is about 15 pm. Finally, the substrate is assembled by linking the four contacts to a TO-39 support using gold wires with a diameter of 0.06 mm, by means of thermocompression.

The morphology of the material obtained has been studied by scan electronic microscopy (SEM) using a Zeiss Leo 1530 FEG microscope.

The collection of X-ray diffraction data (XRD) was performed with a BROKE D8 Advance Da Vinci diffractometer working in Bragg-Brentano's geometry and equipped with an X-ray tube with Cu anode, Ni filter to suppress the Cu component and a LynxEye XE silicon tape detector (angle interval covered by the detector = 2.585° 20) set to discriminate the CuKa1,2 radiation.

The powder was loaded in a 2 mm deep cavity in a sample holder in poly(Methyl methacrylate) and scanned in continuous mode from 5 to 90° 20, with a pace of 0.02° 20 and a time of count of 2s per pace. The qualitative analysis of the collected patterns phase was performed by means of the Bruker AXS EVA software (V.6.0.0.7). The collected XRPD patterns have been modeled through the Rietveld approach of the fundamental parameters (TOPAS v.5.0, Bruker).

The high-resolution transmission electronic microscopy (HR-TEM) was performed via Philips TECNAI F20 ST microscope, equipped with accessory for micro-analysis of dispersion energy. The sample was suspended in isopropanol and treated with ultrasounds. A few drops have been deposited and evaporated on a molybdenum grid. The images were acquired in phase contrast mode.

The X-ray photo-electron spectroscopy measures (XPS) were performed using a Kratos AXIS UltraDLD (Kratos Analytical, Manchester, United Kingdom) device equipped with a hemispherical analyzer and a monochromatic source of X-ray Al Ka (1486.6 eV), in spectroscopy mode. For the measures, the powders of ln20a and Na:ln2O3 were deposited on carbon tape, placed above a silicon support. The samples were analyzed using a 0° take-off angle between the normal surface of the sample and the analyzer axis, corresponding to a sampling depth of about 10 nm. The investigation was recorded in the energy interval 1300,-5 eV to identify the elements on the surface. For each sample, the high-resolution scans of the core levels of In 3d, O 1 s and Na 1 s have been collected. The XPS quantification was performed using the sensitivity factors of the instrument, associating them with high resolution scans. The spectra were aligned by setting the peak of the core level of the C 1 s at 285 eV. All XPS data were analyzed using the software described in [48].

The analysis of optical absorption was performed using a JASCO V- 670 double ray spectrophotometer. The instrument is equipped with a deuterium lamp (190-350 nm) and a halogen lamp (330-2700 nm). The measurement was carried out in the 200-800 nm wavelength range, with a 1 nm sampling interval. In order to carry out the analysis, the powder was dispersed in 2-propanol and subjected to ultrasonic treatment for 30 minutes. The powder band gap has been calculated through the Tauc Plot method, that is, calculating and adapting the absorption data of the nano powders compared to direct transition energy: ahv = ED (hv - E _g ) ^1/2 wherein a is the optical absorption coefficient, hv is the energy of the photon, is the direct band gap and ED is a constant [36]. By plotting the graph of (a v) ² according to the energy of photons and extrapolating the linear portion of the curve up to absorption equal to zero, the values of the direct band gap (E _g ) of the investigated materials are obtained.

Gas detection measures The detection properties of the Na:ln2O3 film have been tested in a sealed room with a flow of 500 seem gas flow (standard cubic centimeters per minute).

Synthetic air (20% O2 and 80% N2) and target gas from certified cylinders (N5.0 purity level) have been mixed and flowed through mass flow regulators. The relative humidity and the temperature inside the test chamber were controlled by a Honeywell HIH-4000 commercial sensor. The test chamber was placed in a climatic chamber to maintain an internal temperature around 25°C. The sensors were maintained at their optimal work temperature, identified after an appropriate calibration in an interval between 150 and 450°C, under a continuous flow of synthetic air, until the thermodynamic stationary state is achieved. The sensor response was defined as in [41 , 42],

The CO2 daily concentrations were chosen based on the values indicated by the National Institute for Safety and Health at work (NIOSH). The response and recovery times have been calculated respectively as the time needed to reach 90% of the constant response and time to restore 90% of the basic level. To study the influence of humidity on the response to CO2, the sensor has been exposed to water vapor, making part of the total flow of synthetic air flow through a bubbler filled with distilled water.

Results and discussion

The SEM analysis of Na:ln2O3 powder obtained through the sol-gel synthesis method has shown particles characterized by spherical morphology with a diameter of 20 ± 5 nm (Figure 6). The refinement of the XRD pattern (Figure 7) shows that the sample is crystalline, with a cubic structure strongly oriented along the plane (222). The absence of peaks of additional phases confirms the incorporation of the Na ⁺ ions in the ln2O3 lattice. In the Na:ln2O3 sample the diffraction peaks move towards Bragg’s angles lower than those of the pure ln2O3, indicating a variation of the interplanar distances, in particular the reticular parameter 'a' and therefore the volume of the unitary cell increases, as shown in Table 2. In addition to a difference in the cell parameter 'a', there is also a different value of micro strain, typical of structural changes, since it provides indications on the existence of long-range random deformations in the various crystalline directions. The introduction of Na in the lattice promotes a reduction in the micro strain (eo) of the crystallites and therefore an increase in the crystallinity of the material, as well as producing an increase in the volume (V) of the unitary cell of at least 0.05% compared to the volume of the pure ln2O3 cell.

Table 2: main reticular characteristics of the samples analyzed, including the phase, the space group, the size of the reticular parameter "a" and volume of unitary cell, size ( RD) and micro strain of crystallites (eo).

The structure of the nanoparticles of ln2O3 doped with sodium was further investigated using the HR-TEM, as shown in Figure 8. From Figure 8 (a) it emerges that the surface consists of nanoparticles of size between 10-20 nm. In particular, the particles show a rounded shape, with evident crystalline facets. In fact, the fringes of the lattice are visible with high magnification (see 2 nm scale insert). The particles are crystalline and therefore the presence of amorphic or secondary phases segregated on the margins is not observed. The SAED pattern (Figure 8 (b)) shows the interplanar distances typical of the cubic phase of ln2O3, corresponding to the crystalline plans (222), (400) and (440). In addition, the STEM-EDX analysis confirmed the presence of 3 elements, O, Na and In, in addition to the Mo element coming from the grid that supports the sample. Therefore, the SEM, XRD and TEM techniques confirm that the material subject of the invention is a Na:ln2O3 nanostructured and with homogeneous crystallinity (there are no amorphic regions). Thanks to its nanocrystalline cubic structure Na:ln2O3 subject of the invention has optimal properties for sensor, including a high surface area (between 20 and 50 m ² g ¹ ) due to the small size of the crystallites and the stability of both structural and electrophysical properties.

The presence of Na in Na:ln2O3and the nature of its bond have been confirmed by high resolution XPS investigations. For this purpose, three areas of the XPS spectrum were examined:

- the In 3d region (440-455 eV, Figure 9 (a)), wherein the dubbed peaks corresponding to 3ds/2 and 3da/2 of the spectrum were observed, relating to the ln-0 link in ln2O3 and Na:ln2O3 samples;

- the O 1 s region (526-535 eV, Figure 9 (b)), wherein it is observed that in both ln2O3 and Na:ln2O3 samples the peak can be deconvoluted in two peaks, corresponding to ln-0 and to In-OH surface hydroxyl groups;

- the Na 1 s region (1066-1075 eV, Figure 9 (c)), wherein the Na:ln2O3 sample was compared with NaNOa salt.

Table 3 reports the values of the binding energies of the peaks revealed for each of the two samples.

A variation of the binding energies of the peaks of In 3d and O 1 s can be observed, which are placed at lower values for the Na naOa sample. The shifting of these peaks, characteristic of the reticular bonds, indicates a different chemical environment of In and O in the two powders, due to the incorporation of Na in the InaOa lattice [49].

In Figure 9 (c) the peak corresponding to Na in the Na naOa sample is observed, which confirms the success of the insertion of sodium in InaOa with the sol-gel method.

It is observed that the energy of the peak of Na 1 s in the Na:lnaOa powder is similar to that of the NaNOa salt, therefore the oxidation state +1 of the sodium is confirmed, which is linked to the O ² ’, instead of ln ³⁺ , changing the electrical properties of the material [50]. The limited shifting of the peak of Na:ln2O3 compared to that of the NaNO3 salt is due to the variation of the chemical environment and not to a different oxidation state.

Table 3: values of binding energies (eV) of the corresponding analyzed 5 samples.

The optical properties of ln2O3 nano powders were characterized by UV-visible spectroscopy (Figure 10). Specifically, the optical absorption of the powder has been investigated to verify how the insertion of sodium in o the structure impacted the direct band gap of ln2O3. Through this analysis, a direct band gap of 3.60 eV for ln2O3 was observed, that’s very close to the value found in other experimental works [50]. By adding Na, it is possible to observe a decrease in the band gap of the material to a value of 3.46 eV. This figure represents a confirmation of the presence of Na as 5 a doper in the ln2O3 powder. In fact, if Na in the sample was present as a metallic Na°, dispersed together with the ln2O3 nano powder, the shifting of the direct band gap of the semiconductor would not be observed, which instead happens when a doper (in this case Na ⁺ ), enters to be part of the semiconductor structure. 0 The optimal response to the gas of the crystalline nanostructured Na:ln2O3 film was determined by measuring the change of conductance before and after the injection of 500 ppm of carbon dioxide, according to the previously described methods, at different temperatures of the sensor operating in the interval 150-450°C.

As shown in Figure 3 (a), the optimal working temperature is in the 200-250°C interval. In this work, an operating temperature of 200°C was chosen to meet the low energy consumption request.

The sensitivity to carbon dioxide has been studied by measuring the conductance of the Na:ln2O3 film when exposed to 250, 500, 1000, 2000, 3500 and 5000 ppm of CO2 in dry air. This concentration interval has been selected to demonstrate sensor performance in a wide range of potential real applications. As can be seen in Figure 3 (b), the sensor response gradually increased by 4.33 (250 ppm) to 12.70 (5000 ppm). Therefore, sensitivity is suitable for both external and internal applications.

It is underlined that at the same operating temperature (200°C), the response to 1000 ppm of CO2 of the nanostructured crystalline Na:ln2O3 based sensor is 10 times higher than the response of the device based on pure nanostructured crystalline ln2O3.

In fact, despite the conductance in the air of Na:ln2O3 film is less than that of the pure ln2O3film by an order of magnitude it should be noted that, while the latter in the presence of CO2 shows a conductance in the same order of magnitude as the one measured In the air (10 ^-7 S), instead the doping with sodium significantly increases the electrical activity of the sensitive film, bringing its conductance from 10’ ⁸ S in the air to 10’ ⁷ S in the presence of 1000 ppm of CO2.

Figure 1 1 shows the repeatability of dynamic responses to 400 ppm (average outdoor value) and 1200 ppm (average indoor value) of carbon dioxide in four cycles. Each cycle includes a period of 2 h of stabilization of the dried synthetic air flow followed by a period of 30 min of a measuring in flow of a mixture of dry synthetic air with 1200 or 400 ppm of CO2. The working temperature of the sensor was equal to 200°C. The measures were carried out in the camera test, previously described in the "Gas detection measures" section, wherein the temperature of the chamber and the relative humidity were about 25°C and 2 RH% respectively. The sensor response and recovery times, for CO2 concentrations of 400 and 1200 ppm respectively, were determined in the order of 5 and 3 minutes and in the order of 9 and 23 minutes. It is observed that these reaction timing can be compared with those of other consolidated sensitive materials [41 -42, 51 ]. These parameters depend on the size and geometry of the chamber, the speed of the gas flow [52] and the relatively low operating temperature. The conductance in the presence of CO2 remains stable during the gas delivery, allowing a punctual calibration of the sensor (response of the device according to the concentration of the analyte). In fact, the variation of the conductance considered after the sensor response time is about 2% during the delivery of 400 ppm of CO2 and about 3% during the delivery of 1200 ppm of CO2.

The relative humidity has remained one of the critical challenges for the real application of the MOS gas sensors. Figure 5 (a) shows that the conductance of the crystalline nanostructured Na:ln2O3 film in the presence of humidity (150 nS at 17 RH%) is increased by 3 times compared to the conductance in dry air (50 nS at 3 RH%), since H2O in the vapor phase reacts as a reducing gas, covering the surface with hydroxylic -OH groups.

The Na:ln2O3 sensor has been exposed to 500 ppm of carbon dioxide at different levels of relative humidity in the 3-64 RH% interval to verify its influence on selectivity to the analyte of interest. Although the conductance is increased by H2O in the vapor phase, the presence of -OH groups decreases the active sites available for interaction with CO2, this involves a decrease in the response (Figure 5 (b)). This occurs only at low RH% values. In fact, in the interval between 3 and 17 RH% the response decreases by a half (from 6 to 3), while between 17 and 64 RH% remains practically constant. This experimental evidence demonstrates the usefulness of the Na:ln2O3 sensor for real applications that cannot be left aside from the presence of humidity.

According to the inventors’ knowledge, the use of sodium in the context of solid-state sensor for the detection of CO2 has only one precedent reported in [17]: the aforementioned article refers to a completely different material: it is, therefore, the doping of ZnO with sodium; the experimental conditions of this work are completely different. In fact, Mohamed A. Basyooni et al. conducted the electrical measures in a nitrogen atmosphere and not in the air, thus bypassing the preadsorption mechanism of oxygen on the surface of the film, an unavoidable condition in the real applications of a gas sensor and fundamental in the interpretation of the sensing mechanism; in the measurement system CO2 is injected inside the test chambers with a flow of 20-70 seem, which corresponds to a concentration interval between 2x10 ⁵ and 7x10 ⁵ ppm. This CO2 concentration is significantly high and differs from real applications; it can also be observed that the responses obtained at much lower CO2 concentrations by the sensor presented in the work subject of this invention are at least 13 times greater than those published in [17], the resistance of the sensitive film does not stabilize throughout the measurement period in the presence of CO2. This behavior does not allow punctual sensor calibration.

In order to evaluate the properties of crystalline nanostructured pure ZnO and ZnO doped with Na (Na:ZnO) as functional materials for CO2 sensors in the air (real conditions), these materials have been synthesized and electrically characterized in the measurement system previously described in the paragraph “Examples; synthesis, film deposition and characterization of Na:ln2O3". Figure 12 represents the response at 1000 ppm of CO2 of pure ZnO and Na:ZnO sensors at the operational temperatures of (a) 25°C and (b) 200°C. It is observed that only at the operating temperature of 200°C the sensors detect the analyte. The doping with Na increases the response of the pure ZnO sensor, which however remains considerably lower than that of the Na:ln2O3 sensor subject of the invention. In addition, both measures reported by Mohamed A. Basyooni et al., the one in N2 atmosphere, and the one in Figure 12 (b) in the air do not reach a situation of electrical stability necessary for the definition of the response value at the fixed CO2 concentration. This condition represents a fundamental requirement to define the applicability of the device in contexts where its calibration is required. Therefore, both the known art, and the new comparative examples, advise against pursuing this teaching, since the sensor based on zinc oxide doped with sodium does not prove to be stable in the presence of CO2, not even at high concentrations.

According to the inventors’ knowledge, the use of indium oxide in the context of solid-state sensor for the detection of CO2 has only one precedent reported in [14]: o the aforementioned article shows that pure ln2O3 is not suitable for being a CO2 sensor because the response to the analyte is low. o the authors therefore propose a loading of ln2O3 with CaO, which makes the device sensitive, even if with low responses, in a large range of concentrations (500-5000 ppm). However, the sensitivity characteristics are not maintained over time, indeed they are completely inhibited. In fact, the sensor requires high temperature cycles in wet conditions to regenerate the material.

Although known art would advise against doping with both Na and nanostructured ln2O3 for the aforementioned reasons, the Na:ln2O3 crystalline nanostructured sensor has proven to detect CO2 with a high response value both in the dry and damp air, and in a large range of concentrations. In addition, the conductance of the film is stable during the gas delivery, allowing a punctual calibration of the sensor.

To further evaluate the performance of Na:ln2O3 crystalline nanostructured sensors, in Table 4, synthesis information, working temperature, response vs. concentration of other MOS used as CO2 sensors are reported. Table 4: Summary of gas sensors for the detection of CO2 based on MOS.

Notes:

* Value not explicitly declared in the study, but approximated from a graph.

** Indicative value, given that the sensor has not reached a stationary state.

Wherein S is the sensor response; R _gas is gas resistance; Rair is resistance in the air; G _gas corresponds to the conductance to the gas and Gair to the conductance in the air. This invention has been described above in relation to exemplary implementations which, however, do not intend to be limiting the scope of this invention, and any change and modification within the reach of the skilled in the art can be made without leaving the scope of application of the present invention as defined by the claims.

There are possible variations to the non-limiting described example, without however leaving the scope of protection of this invention, including all the equivalent realizations for a skilled technician, to the content of the claims. From the description above the skilled technician is able to obtain the object of the invention without having to introduce further details.

REFERENCES

[1] https://www.cdc.gov/niosh/npq/npqd0103.html

[2] C. A. Redlich et aL, Vol. 349, 9057, (1997) 1013-1016. https://doi.org/10.1016/S0140-6736(96)07220-0

[3] A. M. Azad et al 1992 J. Electrochem. Soc. 139 3690. https://doi.org/10.1149/1 .2069145

[4] S. Karthikeyan et al. 2015 J. Environ. Nanotechnol. 4(3). https://doi.org/10.13074/jent.2015.12.153163

[5] S. Mulmi, V. Thangadurai 2020 J. Electrochem. Soc. 167 037567. https://doi.orq/10.1149/1945-7111 /ab67a9

[6] G.Korotcenkov, Advances in Metal Oxides and Their Composites for Emerging Applications (2022) ISBN: 9780323857055

[7] G. Korotcenkov Nanomaterials 2020, 10(7), 1392. https://doi.org/10.3390/nano10071392

[8] G. Korotcenkov Materials Science and Engineering R 61 (2008) 1-39 https://doi.org/10.1016/i.mser.2008.02.001

[9] S. Keerthana et al. J. Alloys Compd. 897 (2022) 162988. https://doi.org/10.1016/i.iallcom.2021.162988

[10] N. Rajesh, et al, Ceram. Int. 41 (2015) 14766-14772. https://doi.org/10.1016/i.ceramint.2O15.07.208

[11] S. Joshi, S.J. et al. ACS Appl. Mater. Interfaces. 9 (2017) 27014- 27026. https://doi.org/10.1021 /acsami.7b07051

[12] C.A. Zito, T.M. et al. ACS Appl. Mater. Interfaces. 12 (2020) 17745- 17751. https://doi.org/10.1021 /acsami.OcOI 641

[13] I. Djerdj, A. et al. Chem. Mater. 21 (2009) 5375-5381. https://doi.orq/10.1021 /cm9013392

[14] A. Prim et al. Adv. Funct. Mater. 17 (2007) 2957-2963. https://doi.orq/10.1002/adfm.200601072

[15] V. Manikandan et al. RSC Adv. 10 (2020) 3796-3804. https://doi.org/10.1039/C9RA09579A

[16] Y. Xia et al. Sens. Actuators B Chem. 357 (2022) 131359. https://doi.orq/10.1016/i.snb.2022.131359

[17] M. Basyooni et al. Sci Rep 7, 41716 (2017). https://doi.org/10.1038/srep41716

[18] O. Bierwagen 2015 Semicond. Sci. TechnoL 30 024001. https://doi.org/10.1088/0268-1242/30/2/024001

[19] Y.-G. Yan et al. Crystal Growth & Design 2007, 7, 5, 940-943. https://doi.org/10.1016/i.mssp.2O13.09.026

[20] Y.-K. Lv et al. ACS Appl. Mater. Interfaces 2020, 12, 30, 34245- 34253. https://doi.org/10.1021 /acsami.0c03369

[21] T. Herranz et al. Applied Catalysis A: General 311 (2006) 66-75 67. https://doi.org/10.1021 /acscatal.0c03665

[22] B. Liang et al. ACS Sustainable Chem. Eng. 2019, 7, 1 , 925-932. https://doi.org/10.1021/acssuschemeng.8b04538

[23] V. L. Mejia-Trejo Chem. Mater. 2008, 20, 22, 7171-7176. https://doi.org/10.1021 /cm802132t

[24] C. Wei et al, Applied Surface Science 525 (2020) 146622. https://doi.org/10.1016/i.apsusc.2020.146622

[25] T. Herranz et al. Applied Catalysis A: General 311 (2006) 66-75. https://doi.org/10.1016/i.apcata.2006.06.007

[26] Z. Zhang et al Chemical Engineering Journal 428 (2022) 131388. https://doi.org/10.1016/j.cej.2O21 .131388

[27] W. Tu et al. Applied Catalysis B: Environmental 298 (2021 ) 120567. https://doi.org/10.1016/j.apcatb.2O21 .120567

[28] S. Zhang et al. Applied Catalysis B: Environmental 293 (2021 ) 120207. https://doi.org/10.1016/j.apcatb.2O21 .120207

[29] L. Liu et al. Applied Catalysis B: Environmental 235 (2018) 186-196. https://doi.org/10.1016/j.apcatb.2O18.04.060

[30] Y. Zheng et al Catalysts 2021 , 11 , 735. https://doi.org/10.3390/catal11060735

[31] D. Degler et al. ACS Sens. (2019), 4, 9, 2228-2249. https://doi.org/10.1021 /acssensors.9b00975

[32] D. Bruce Buchholz Chem.Mater.2014, 26, 5401 -5411 https://doi.org/10.1021 /cm502689x

[33] N. Barsan et al. Journal of Electroceramics, 7, 143-167, 2001 https://doi.org/10.1023/A:1014405811371

[34] J. Zhang Adv. Mater. 2016, 28, 795-831 https://doi.org/10.1002/adma.201503825

[35] Z. M. Seeley Materials Science and Engineering B 164 (2009) 38-43 https://doi.org/10.1016/i.mseb.2009.06.009

[36] P. Makula et aL, J. Phys. Chem.Lett.2018,9,6814-6817. https://doi.org/10.1021/acs.ipclett.8b02892

[37] X. Peng et al. Cell Reports Physical Science 2, 100436, May 19, (2021 ). https:doi.org/10.1016/j.xcrp.2O21 .100436

[38] M. Tiemann Chem. Eur. J. 2007, 13, 8376 - 8388 https://doi.org/10.1002/chem.200700927

[39] V. Guidi et al., “Printed films: Materials science and applications in sensors, electronics and photonics” B. in WOODHEAD PUBLISHING SERIES IN ELECTRONIC AND OPTICAL MATERIALS (2012) pp.278- 334

[40] Y. Xiong et aL, Sens. Actuators B Chem. 241 (2017) 725-734. https://doi.org/10.1016/i.snb.2O16.10.143

[41] E. Spagnoli et aL Sens. Actuators B Chem. 347 (2021 ) 130593. https://doi.org/10.1016/i.snb.2O21 .130593

[42] A. Rossi et aL Sensors 2023, 23, 6291 https://doi.org/10.3390/s23146291

[43] G. Korotcenkov et aL, Thin Solid Films 515 (2007) 3987-3996 https://doi.org/10.1016/i.tsf.2006.09.044

[44] https://www.marketsandmarkets.com/PressReleases/gas-sensor.a sp

[45] https://edepot.wur.nl/42427

[46] M.N. Anwar et al Journal of Environmental Management 226 (2018) 131-144. https:doi.org/10.1016/i.jenvman.2O18.08.009

[47] Z. Peng et aL Environ. Sci. TechnoL Lett. (2021 ), 8, 5, 392-397. https://doi.org/10.1021/acs.estlett.1 c00183

[48] G. Speranza and R. Canteri. SoftwareX 10 (2019) 100282. https://doi.org/10.1016/i.softx.2019.100282

[49] I. Singh et aL RSC Adv. 7 (2017) 54053-54062. https://doi.org/10.1039/C7RA10108B

[50] Granqvist, C.G. AppL Phys. A57, 19-24 (1993). https://doi.org/10.1007/BF00331211

[51] M. Valt et aL ACS AppL Mater. Interfaces. 13 (2021 ). https://doi.orq/10.1021/acsami.1 c10763

[52] A. Gaiardo et al. Sens. Actuators B Chem. 305 (2020) 127485. https://doi.orq/10.1016/i.snb.2O19.127485