(456i) Design of Porous Material Based Electronic Nose for Gas Sensing: Impact of Adsorbent Equilibrium and Kinetics

The global gas sensor market has been expanding steadily over the years due to the ever increasing use of sensors in indoor, industrial, medical and automotive sectors. Most gas sensing devices can be classified into two groups, i.e. sensors whose sensing mechanism depends on electrical variation (e.g. impedance, resistance, or capacitance) or sensors whose sensing mechanism depends on change of a given property (e.g. luminescence, mass change, acoustic change). Irrespective of the sensing mechanism, a good sensor is one that exhibits excellent selectivity, sensitivity, stability, reusability, and response time.

Metal-organic frameworks (MOFs), a class of porous materials, has shown great promise in gas sensing applications. Their tunable chemical and structural features can lead to a very high selectivity toward a target gas, which is beneficial for gas sensing applications. Additionally, MOFs are excellent adsorbents at both ambient and elevated pressure and temperature conditions.¹ These favorable features might not always be satisfied by commercial metal oxide or polymeric sensors. Nevertheless, a single MOF cannot detect multiple gases in a mixture. Instead, an electronic nose or sensor array, composed of several materials that exhibit preferential selectivity toward a given gas in the mixture, has to be constructed and the gas composition can be resolved by analyzing the response of all the constituent materials in the array.

Given the large number of candidate materials, screening MOFs that can be coated on a sensor array is challenging. Recent computational studies have reported methodologies to quantify the performance of a sensor array and to screen a set of materials that would allow a sensor array to exhibit the best performance.^2,3 In these studies, a gravimetric sensor array was used, where the sensing mechanism is based on estimating the gas composition by monitoring the change in mass of the material due to gas sorption. These studies provide valuable insights and a framework to screen materials, when the gas is at equilibrium or reaches equilibrium instantaneously with the materials in the sensor array. However, both equilibrium and kinetic characteristics dictate the gas sorption on any given material. The amount of gas adsorbed will determine the sensitivity and the rate at which the gas is adsorbed on the material will determine the response time of a sensor.

The overarching goal of this work is two fold. First, to systematically highlight the impact of equilibrium characteristics on the sensor array response and on the gas composition estimation and to propose a simple graphical approach to rapidly screen gravimetric sensor array materials. Second, to highlight the need to incorporate gas sorption kinetic characteristics and design variables (i.e. size of the device and flow rate) to estimate gas composition and to screen gravimetric sensor array materials. To address this aim, we developed a computational test bench that incorporates a mathematical model of the array to simulate its response when exposed to a gas mixture and an optimization routine that estimates the corresponding gas composition. Our studies account for both equilibrium and kinetic characteristics of the materials in the array. We used hypothetical materials that exhibit adsorption behavior similar to MOFs to keep the study as general as possible. Therefore, the framework and the conclusions drawn thereof can be easily translated to real materials.

We first simulated several case studies accounting only for the equilibrium characteristics of the materials. The mathematical model for the sensor array in these studies makes use of the adsorption isotherm of the hypothetical materials. These studies led to several key learnings. First, one must incorporate more than one material to resolve a multicomponent gas mixture and include material balance constraints when estimating the gas composition. Second, one can exploit the shape of the material response as a function of gas composition to identify regions of gas compositions that would facilitate resolving the composition of the mixture to high accuracy. Finally, one can utilize the shape of the material response to rapidly screen materials that can be incorporated into a sensor array.

Then, we developed a detailed mathematical model by incorporating both the equilibrium and kinetic characteristics of the materials. We modeled the sensor device with the materials as a continuous stirred tank reactor⁴ and described the mass transfer in these materials using the linear driving force model. The system of equations was integrated over time and coupled with an optimizer to obtain a time-resolved evolution of the sensor response and the gas composition, respectively. Subsequently, we conducted several case studies. First, we looked at a hypothetical array that does not exhibit instantaneous equilibrium. We estimated gas compositions by accounting for either equilibrium or both equilibrium and kinetics. The study, as expected, highlighted the inability of the purely equilibrium approach to resolve the gas mixture. Second, we performed a parametric study on three factors, namely the gas flow rate, the size of the device, and the kinetics of adsorption, to elucidate the impact of these on the optimal sensing performance of the array and the consequences of not incorporating these effects.

To conclude, we present here the first study that aims to systemically evaluate the effect of adsorption equilibrium and kinetics on the performance of a gravimetric sensor array. This work stresses the importance of including a detailed sensor model, which is often not considered, to better reflect reality, both for material screening and for composition estimation purposes.

References

1. Zhou, H.-C.; Long, J. R.; Yaghi, O. M. Chem. Rev. 2012, 112 (2), 673–674.
2. Gustafson, J. A.; Wilmer, C. E. ACS Sensors 2019, 4 (6), 1586–1593.
3. Sousa, R.; Simon, C. M. ACS Sensors 2020, 5 (12), 4035–4047.
4. Brandani, S.; Mangano, E. Adsorption 2020, 1, 3.