Exhaust gas selective sensing using printed RF gas sensors – CARDIF

Functionalized inks are obtained by LCPO team through the modification of commercially available polyethyleneimines (PEI) by control of the composition ratio of primary (1°), secondary (2°) and tertiary (3°) amines within the PEI moiety. Another explored pathway is to modify OH groups with fluorinated silane molecules in order to render the PEI more hydrophobic, as first sensor characterization sessions showed a moisture sensitivity that may interfere with the targeted gas. Finally, in a more systematic way, LCPO is developing linear PEI that could be more versatile in terms of structures, functionalities and hydrophobicity features.
Different RF structures were designed and fabricated by XLIM depending on the active material to be deposited on the sensors. The sensors work on the principle of varying electrical properties such as the permittivity and conductivity of the sensitive material in the presence of target species. This variation affects the propagation of electromagnetic waves, which results in a change in the response of the resonator in amplitude and / or in its resonant frequency. As PEDOT-PSS-MWCNT, a conductive material, was used at the beginning of the project, a half-wave resonator terminated at its ends by stubs printed with PEDOT-PSS-MWCNT ink was dimensioned to optimize the sensitivity of the structure incorporating the sensitive material on the maxima of electric field. For PEI-based materials that are dielelectric, an interdigitated structure covered with sensitive material was dimensioned in order to optimize its sensitivity in the frequency ranges targeted in the project.
For characterization of the sensors in a controlled environment, humidity was considered at first as it is the main potential interferent in the environment. For that, IMS has implemented a suitable test bench consisting of a climatic chamber to generate different concentrations of relative humidity and temperatures, a 4-port vector network analyzer (VNA Keysight E5080a) to measure the sensor scattering parameters (Sij), an Arduino-driven commercial temperature and humidity sensor (SHT85), and a computer to perform data acquisition and processing. The sensor response was recorded for relative humidity ranges of 30 to 70% RH.
For characterization of the sensors in real-life environment, during the first period of the project Uni Eiffel has prepared a state of the art regarding existing applications of the sensors developed in CARDIF to urban pollution monitoring, which will allow to later fine-tune CARDIF sensors specifications toward a reduced number of relevant use cases.
In parallel, Uni Eiffel and IMS have developed an initial set up for in-situ calibration of CARDIF sensors in Sense-City facility, and deployed the first – preliminary - sensors prepared by IMS over several weeks.

Regarding functionalized inks, 5 materials of interest have been screened and deposited on sensors so far: PEDOT-PSS-MWCNT, EB-PEI, Commercial PEI (reference material), Linear PEI-COPh and HB-PEI-Silane. Moisture sensitivity being one of the materials characteristics to attenuate, the fluorinated PEI HB-PEI-Silane shows notably a contact angle with water droplets of 106° vs 11.6° for the commercial PEI.
As for sensor characterization, the results with the PEDOT-PSS-MWCNT-based device highlighted a good repeatability and a high sensitivity to humidity, with a saturation from 48 %RH or less – thus humidity variations may affect mainly the baseline in lab conditions with drying nitrogen carrier flow, rather than in most of real conditions. The EB PEI-based device (3.2 GHz, EB-PEI of thickness 100 µm deposited by ISORG) highlighted a sensitivity of -1.325 MHz /% RH and a good reversibility and reproducibility, supporting the further development of a mathematical model by Uni Eiffel in order to deduce the effects of ambient humidity as an interferent. Additional sensors based on other preselected active materials are under characterization to this day.
In preparation of collecting sensors’ data at IMS lab and at Sense-City, Uni Eiffel has developed a Bayesian algorithm enabling to determine the most likely calibration models (including multi-parameter and nonlinear ones), then to use this calibration to predict the concentration of the target gas based on sensors data.

On the material functionalization side, future work will consist in controlling the content in 2°/3° amines of linear PEIs and in enhancing hydrophobicity of modified PEIs.
After characterization of the functionalized materials sensitivity to humidity, the next step will be to focus on a first set of two harmful vapors: NO2 and SO2.
Finally, once data will have been collected from sensors tested at IMS and Sense-City, the algorithm developed by Uni Eiffel will be tested and optimized in order to calibrate the sensors for detection of these target gases.

2 papers have been published so far:
• J. George et al., «Inkjet-Printed RF Gas Sensors based on conductive nanomaterials for VOCs monitoring,« 2021 IEEE MTT-S International Microwave Symposium (IMS), 2021, pp. 93-96, doi: 10.1109/IMS19712.2021.9574856.
• J. George et al., «A new dual RF sensor in gas detection and humidity influence,« 2020 IEEE SENSORS, 2020, pp. 1-4, doi: 10.1109/SENSORS47125.2020.9278775.
No patent has been file to this day.

Exposure to toxic gases or volatile organic compounds (VOCs) affects safety and public health. More than 6.5 million deaths per year worldwide are attributed to environmental pollution (indoor and outdoor air quality). With the Internet of Things, large-scale deployment of gas sensors becomes accessible to monitor and analyze the environment of individuals, in public or private spaces. This market, driven in particular by the construction, medical industry or consumer applications markets, requires low-energy, low-cost monitoring devices. Portable sensors alone represent a market estimated at 3 billion units in 2025, of which more than 30% are emerging communicating sensors, including chemical sensors, growing exponentially over the next 10 years. In these new markets, research and development of innovative tools is becoming an exciting new field for electronics.
In this context, we propose to address some of the issues related to monitoring the quality of polluted air related to exhaust gases due, for instance, to transports and industrial activity. This has led to specifications in particular for certain harmful vapors such as nitrogen dioxide (NO2), sulfur dioxide (SO2) and ozone (O3) as highlighted in the French regulations. The objective of CARDIF is to respond in particular to the challenge of selectivity, identified as a bottleneck limiting the contribution of conventional metrology in observing or diagnostic systems in a heterogeneous medium. This will be done using functionalized polymers with specific groups.
The current innovative sensors generally suffer from high consumption and / or bulky instrumentation due to low frequency operation, and are mainly based on expensive solutions. As part of the CARDIF project, we propose another type of sensors based on microwave transducers operating at ambient temperature. In addition to being a passive device and therefore not consuming power, they could also operate wirelessly. They are thus suitable for networking and high-frequency communications, usable for real-time detection and providing directly exploitable information. In addition, because of its planar structure, the device can be manufactured on a flexible substrate by low-cost printing technologies.
This multidisciplinary study is made possible thanks to the close collaboration between two industrial (ISORG, Efficacity) and three academic partners (LCPO, IMS, XLIM). Partners have the experience to meet these scientific and technical challenges. To our knowledge, no internationally referenced work has focused on such printed RF gas sensors with optimized selectivity, meeting the following key features:
1 - real-time monitoring of the quality of the outside air,
2 - high sensitivity at room temperature and therefore low energy consumption,
3 - the selective detection of NO2, SO2 and O3 at levels of a few ppb to ppm using functionalized polymers,
4 - low cost manufacturing processes based on collective printing technologies,
5 - new autonomous and wireless solutions, operating in real time thanks to the passive microwave transduction,
6 - outdoor tests following a realistic deployment scenario.
By this way, CARDIF clearly responds to the priorities of the call, in particular the development of sensors for environmental monitoring (smart monitoring)