High-selectivity gas microsensor

Abstract

The invention relates to a MEMS (micro-electromechanical system) gas sensor combined with the architecture of a CMOS (complementary metal oxide semiconductor) image sensor, in which the conventional pixels are replaced with various two-dimensional (2D) gas-sensitive materials. This substitution significantly improves selectivity, allowing each gas to be identified accurately based on the specific reactions of the materials to the different compounds. The integration of CMOS technology with MEMS systems also enables fast and efficient data processing, while optimizing energy consumption and costs, making this solution ideal for gas detection in various environments.

Description

HIGH SELECTIVITY GAS MICRO-CAPPER FIELD OF INVENTION The present invention relates to a MEMS (Mechanical Microelectric System) gas sensor combined with a CMOS (Complementary Metal Oxide Semiconductor) image sensor architecture, where conventional pixels are replaced by various two-dimensional (2D) gas-sensitive materials. This substitution significantly improves selectivity, enabling the precise identification of each gas through the specific reactions of the materials to different compounds. The integration of CMOS technology with MEMS systems also allows for fast and efficient data processing, while optimizing energy consumption and costs, making this solution ideal for gas detection in diverse environments. EARLIER ART

[0001] MEMS gas sensors play a vital role in air quality monitoring, a crucial factor for individual well-being, health, and safety. By detecting harmful gases such as carbon monoxide, they help prevent the risk of poisoning and improve quality of life by reducing exposure to air pollutants. Good air quality is directly linked to a reduction in respiratory and cardiovascular diseases. However, despite their high sensitivity, these sensors have low selectivity, which complicates the precise identification of different gases present and limits their effectiveness in complex environments.

[0002] This weakness stems from the non-specific properties of the sensitive materials used, such as metal oxides. These materials react similarly to multiple gases, producing ambiguous electrical signals that make it difficult to accurately identify the detected gases. This lack of distinction arises from the adsorption and chemical reaction mechanisms of the gases on the surface of the sensor's sensitive layer, which are often common to different compounds. Improving the selectivity of these sensors is crucial for expanding their use in fields such as environmental monitoring, security, and medicine, where precise gas identification is essential.

[0003] Furthermore, some sensors can detect dangerous gases, such as carbon monoxide, as well as harmless ones, such as low concentrations of carbon dioxide. This inability to differentiate between gases can lead to false alarms, creating unnecessary confusion or panic. The level of danger depends on the concentration and type of gas detected, highlighting the importance of accurate identification. A sensor capable of recognizing each gas and its concentration would provide a clear picture of air quality, enabling more informed health and safety decisions.

[0004] Considerable efforts have been made to improve the selectivity of metal oxide gas sensors. New approaches include the engineering of specific nanostructures, such as the use of 2D materials and nanowires, which increase the sensing area and specific interactions with gases [1][2]. Advanced doping techniques, combining different metal oxides, also allow for modulation of sensitivity to targeted gases [3]. In addition, the integration of thermal modulation and data processing algorithms helps to differentiate sensor responses according to gases, thus contributing to better recognition and a reduction of cross-interference [4].

[0005] Despite advances, metal oxide sensors still struggle to accurately distinguish gases in complex environments containing multiple compounds [5]. Non-specific interactions between the sensitive materials and multiple gases result in electrical signals that are difficult to interpret, making gas distinction highly inaccurate. Furthermore, these sensors fail to provide accurate measurements of the concentration of a target gas in the presence of other interfering gases due to cross-effects and ambiguous responses. This lack of selectivity limits their effectiveness in applications requiring precise identification and reliable quantification of gases in complex mixtures.

[0006] Integrating a network of sensitive layers composed of various materials and dopings with the architecture of CMOS image sensors on the same membrane allows for improved scanning of the surrounding medium. Each sensor, integrated into the same chip and operating with the same microheating element, reacts differently to the gases present. Thanks to this configuration, approximately sixty sensors can operate in parallel, each providing a specific response based on its sensitive layer. The measurements collected by each layer are then sent sequentially to a processing unit integrated with the sensor, which analyzes and classifies the gases present in the medium while quantifying their concentrations, thus offering more precise and reliable detection. BRIEF DESCRIPTION OF THE INVENTION

[0007] This disclosure proposes a device to improve the selectivity, accuracy, and efficiency of gas sensors by integrating a CMOS image sensor architecture combined with two-dimensional (2D) sensitive materials. This device replaces traditional pixels with layers of gas-sensitive materials, enabling more precise detection and identification of gaseous compounds. This innovation increases gas selectivity in complex environments while optimizing data processing and reducing cross-interference. Gas sensors based on this CMOS architecture can be integrated with data processing electronics on a single chip, making them ideal for a variety of applications requiring compact, low-power devices.

[0008] This disclosure presents a high-selectivity micro gas sensor based on an integrated CMOS architecture, designed to improve the detection accuracy and efficiency of gas sensors. It is suitable for a variety of applications, such as air quality monitoring, medicine, electronic noses for odor identification in intelligent robots (AI), smart agriculture, mining, and security systems. This sensor uses 2D sensitive materials instead of conventional pixels. This approach enables more accurate gas identification in complex environments by optimizing data processing and reducing cross-interference. Thanks to this advanced architecture, the sensor offers enhanced detection performance while maintaining low power consumption and high reliability. BRIEF DESCRIPTION OF THE FIGURES

[0009] In all drawings, similar reference numbers designate the same object or action, unless the context indicates otherwise. The sizes and relative positions of objects in the drawings are not necessarily drawn to scale.

[0010] Figure 1: Architecture of the 2D sensitive material network adopted in the high selectivity gas microsensor.

[0011] Figure 2: Top view of the high selectivity gas micro sensor chip before packaging.

[0012] Figure 3: Cross-sectional view of the high-selectivity gas micro-sensor chip. DETAILED DESCRIPTION OF THE INVENTION

[0013] In the following description, certain aspects are discussed in detail for a clear understanding of the disclosed invention. It should be noted that the well-known structures and manufacturing methods of MEMS and CMOS sensors have not been addressed.

[0014] The high-selectivity gas microsensor based on CMOS and MEMS architecture with 2D materials represents a major advancement in the field of gas detection. By replacing traditional pixels with gas-sensitive 2D materials, this innovative device significantly improves accuracy and selectivity. The integration of MEMS technology enables optimal thermal management through a micro-heating element, while the CMOS architecture ensures efficient integration of data processing circuits, allowing for rapid and precise analysis of detected gases. This combination optimizes detection while reducing power consumption and cross-interference. Furthermore, this innovation can open new perspectives as a reliable solution for applications such as electronic noses integrated into intelligent robots, enabling odor differentiation and more advanced interaction with the environment.

[0016] This miniaturized sensor, based on 2D materials and a CMOS architecture, can also be integrated into smartphones to add a new sensory dimension: odor detection. While current smartphones can already touch (touchscreen), see (camera), and hear (microphone), they are unable to smell. This sensor will allow smartphones to detect ambient air quality, monitor odors related to food safety, such as checking if meat is spoiled, or even distinguish between different varieties of coffee. Integrating this technology into smartphones would thus offer a useful new feature in the fields of health, safety, and food.

[0017] Figure 1 illustrates the internal architecture of the high-selectivity gas microsensor 100, based on a CMOS architecture. Inspired by CMOS image sensors, this structure uses detection units 1 similar to "pixels," each unit measuring the resistance variations of the metal oxide layer during interaction with gases. Instead of the photodiodes used in image sensors, these are replaced by classification detection elements 8, made of differently doped 2D materials, whose resistance varies according to the nature and concentration of the detected gases. This change in resistance is then converted into an electrical signal via a readout circuit, similar to that of image sensors, where a pixel amplifier 7 amplifies the voltage generated by the resistance variation.Furthermore, a line 2 decoder and a column 3 decoder, along with their column 29 and line 9 switches, allow the data from each classification 8 detection element to be read in a multiplexed manner, thus providing high-resolution detection while minimizing circuit and energy consumption. The efficiency of this approach relies on the variable sensitivity of the 2D materials and the optimization of the readout circuit.

[0018] In this architecture, a column amplifier 4 is added to the columns to provide additional amplification, compensating for signal losses that can occur during data transmission. This second amplifier also plays a key role in signal uniformity by compensating for output variations between the different detection units 1. Furthermore, it helps reduce noise due to high temperature and interference that can occur during signal transmission from detection unit 1 to the column bus.

[0019] An analog-to-digital converter (ADC) 5 is integrated into this configuration at the column level to convert the amplified analog signal into a digital signal. This process is essential for transforming the voltage variation, induced by the change in resistance in the presence of gas, into a numerical value corresponding to the detected gas concentration. Finally, a digital control unit 6 is added to improve measurement quality by applying noise reduction and defect correction algorithms. It also enables data compression and encoding to minimize the required storage space, while managing the transfer and communication of information between the sensor and the processor for efficient processing.

[0020] The heating microelement 11 is powered via the positive pin 15 and the negative pin 14, as shown in Figure 2. The voltage is regulated by a driver 10, allowing precise temperature control at the membrane 27. This heating microelement 11 is equipped with heat dissipation plates 12, which ensure uniform heat distribution, particularly at the center of the membrane 27. Metal oxide nanowires, serving as quantification detector elements 18, are deposited above the area occupied by the heating microelement 11 and heat dissipation plates 12, playing a role in gas quantification due to their large detection surface area. Around the heating microelement 11, detection units 1 using 2D materials, doped differently to improve selectivity, are also integrated onto the membrane 27.This configuration increases selectivity while maintaining a wide sensitivity range in gas detection.

[0021] The sensor 200 chip is also equipped with two communication pins following the protocol l 2 C, namely the bidirectional data line (SDA) 16 and the bidirectional synchronization clock line (SCL) 17. These pins enable the transmission of data collected by the classification detection elements 8 and the quantization detection element 18. The collected information is sent from the digital control unit 6 to a processor or microcontroller for further processing. This communication protocol ensures fast and reliable data transmission, while reducing the complexity of the connections required between the electronic components.

[0022] The device is designed with a 200 µm diameter tungsten heating element 11 embedded in a 700 µm diameter circular dielectric membrane 27. A thin silicon pn diode 21 for temperature measurement is embedded in the membrane beneath the heating element 11. A field-effect transistor (NFET) driver 10 is also embedded on the chip to drive the heating element 11. The chip size is 1200 µm x 1200 µm. The device can be fabricated using a 1.0 µm silicon insulator (SOI) CMOS process, followed by deep ion reactive etching (DRIE) on the back side to form a thin membrane 27.

[0023] The CMOS process used for this device is a high-temperature process that uses tungsten metallization instead of aluminum for the interconnects. Due to its high melting point, thermal stability, and high electrical resistivity, tungsten enables efficient and reliable heating at elevated temperatures. It also offers better resistance to corrosion and thermal deformation than aluminum, thus ensuring durability and long-term performance in extreme environments.

[0024] The SOI 19 wafer used can be a 6-inch wafer, comprising a buried 1 µm layer of silicon dioxide 22 and a thin 0.25 µm layer of silicon. This wafer then undergoes standard CMOS processing steps, such as ion implantation, gate oxide formation, and polysilicon deposition. The silicon dioxide 22 is deposited as the dielectric material, followed by alternating layers of tungsten interconnects 23 and a second layer of silicon dioxide 28. A tungsten interconnect layer forms the heating microelement 11 and heat dissipation plates 12.The thin silicon layer forms the temperature sensing pn diode 21, the NFET control transistor for the heating microelement 11, and the nMOS / pMOS transistors 24 that make up the electronic circuits, including the pixel amplifiers 7 and the line switches 9 manufactured at the gas sensing area 25, as well as the column amplifier 4, the ADC 5, the line decoder 2, the column decoder 3, the column switches 29, and the digital control unit 6 manufactured at the integrated circuit area 26.

[0025] A top layer of tungsten 20 is added to form the electrodes of the sensing layers and the power supply pins 15-14 and communication pins 16-17. Before the deposition of the sensing layers, such as nanowires and 2D materials, a silicon nitride layer 13 is applied for passivation. A DRIE etching step on the back of the wafer then defines the membrane area 27. The total thickness of this membrane 27 is approximately 5 µm, composed mainly of silicon dioxide 22, with a silicon nitride passivation layer 13 with a thickness of 0.5 µm. The tungsten layer is 300 nm thick. Figure 3 illustrates the structure of the high-selectivity gas microsensor 300 with its various components according to the embodiment of this disclosure.

[0026] These 2D materials, such as molybdenum disulfide (MoS2), titanium dioxide (TiO2), or zinc oxide (ZnO), offer a large active surface area, increased sensitivity, and improved selectivity for gas detection. Their low operating temperature, below 150 °C, allows for optimal integration with the electronic circuits embedded in the gas detection zone 25, while ensuring high performance and reduced power consumption. The two-dimensional structure enables better interaction with gas molecules, while their electronic properties, such as a narrow band gap and high charge carrier mobility, optimize detection even at low concentrations.

[0027] It should be noted that by modifying the type of 2D materials and dopants used, the sensitivity of metal oxide semiconductors can be adjusted to target specific gases. Each combination of materials and dopants influences the electronic interactions between the surface and the gas molecules, thus modifying the sensor's response. This ability to adjust sensitivity allows for gas differentiation and improved selectivity for a variety of sensing applications.

[0028] This invention offers several notable advantages. By replacing the pixels of the CMOS image sensor with 70 differently doped 2D materials, it enables highly selective gas detection. The heating microelement 11 with heat dissipation plates 12 ensures a uniform temperature for the quantification detector element 18, thus optimizing its sensitivity, while the classification detector elements 8, distributed around the heating microelement 11, operate at different temperatures due to the thermal gradient. This modifies the material characteristics, allowing each material to react differently to gases depending on its optimal temperature. This configuration enables the highly accurate classification of several types of gases, while also quantifying their concentration, through the processing of the collected data. INDUSTRIAL APPLICATION The high-selectivity gas microsensor can be applied in various industrial sectors, including air quality monitoring, wearable devices, and smartphones, where the sensor could be used to detect toxic or harmful gases. In the food industry, it can monitor the freshness of foods such as meat or dairy products by identifying volatile compounds. It can also be integrated into industrial systems to monitor gases in hazardous environments, or into robots equipped with "electronic nose" capabilities to distinguish odors, thus improving safety and diagnostics. REFERENCES [1] Parichenko, Alexandra, et al. "Recent advances in technologies toward the development of 2D materials-based electronic noses." TrAC Trends in Analytical Chemistry (2023): 117185. [2] Domènech-Gil, Guillem, et al. "Highly sensitive SnO2 nanowire network gas sensors." Sensors and Actuators B: Chemical 383 (2023): 133545. [3] Abdelkarem, Khaled, et al. "Design of high-sensitivity La-doped ZnO sensors for CO2 gas detection at room temperature." Scientific Reports 13.1 (2023): 18398. [4] Fan, Ya-Han, et al. "Gas Identification Algorithm Based on Dynamic Response Analysis of Metal Oxide Sensors Under Temperature Modulation." IEEE Access (2024). [5] Dutta, Taposhree, et al. "Road Map of Semiconductor Metal-Oxide-Based Sensors: A Review." Sensors 23.15 (2023): 6849.

Claims

DEMANDS 1. A device for the qualitative and quantitative detection and identification of gases, consisting of a high-selectivity gas microsensor (300), characterized in that it comprises: • a gas detection zone (25) comprising: a micro heating element (11) integrated into a membrane (27) associated with heat dissipation plates (12) to maintain a uniform distribution of heat; a quantification detector element (18) positioned above the heating microelement (11) for measuring the concentration of gases; detection units (1) distributed around the heating microelement (11) on the membrane (27), configured to operate at different temperatures by taking advantage of the thermal gradient and comprising o a pixel amplifier (7) to amplify the voltage resulting from the resistance variations of sensitive materials induced by the presence of gas, o in-line switches (9), o and classification detection elements (8) composed of 2D materials for the classification of detected gases; and • an integrated circuit area (26) comprising: a driver (10), a digital control unit (6), an analog-to-digital converter "ADC" (5), column amplifier (4), a line decoder (2), a column decoder (3), a positive pin (15), a negative pin (14), an SDA pin (16), an SCL pin (17), and column switches (29).

2. The device according to claim 1, characterized in that the 2D materials are selected from molybdenum disulfide (MoS2), titanium dioxide (TiO2) or zinc oxide (ZnO).

3. The device according to claims 1 and 2, characterized in that the heating microelement (11) with the heat dissipation plates (12) ensures a uniform temperature for the quantification detector element (18) in order to quantify the detected gases, while the classification detection elements (8), arranged around this microelement, operate at varying temperatures due to the thermal gradient, thus modifying the characteristics of the materials so that each reacts differently to the gases according to its optimal temperature.

4. The device according to at least one of claims 1 to 3, characterized in that the line decoder (2) and the column decoder (3), together with their column switches (29) and line switches (9), allow the data from each classification detection element (8) to be read in a multiplexed manner, thus providing high-resolution detection while minimizing circuit and energy consumption.

5. The device according to at least one of claims 1 to 4, characterized in that the digital control unit (6) applies noise reduction and defect correction algorithms to improve measurement accuracy, while managing data transfer and compression.

6. The device according to at least one of claims 1 to 5, characterized in that the heating microelement (11) and the interconnections (23) are made of tungsten to ensure thermal and electrical stability and high resistance to extreme environments.

7. The device according to at least one of claims 1 to 6, characterized in that it can be integrated into portable electronic devices, such as smartphones or smart robots, for applications of gas detection, odor recognition, or air quality control.

Images