Sensing and sensitivity: Computational chemistry of graphene‐based sensors

Abstract

Highly efficient, tunable, biocompatible, and environmentally friendly electrochemical sensors featuring graphene‐based materials pose a formidable challenge for computational chemistry. In silico rationalization, optimization and, ultimately, prediction of their performance requires exploring a vast structural space of potential surface‐analyte complexes, further complicated by the presence of various defects and functionalities within the infinite graphene lattice. This immense number of systems and their periodic nature greatly limit the choice of computational tools applicable at a reasonable cost. An alternative approach using finite nanoflake models opens the doors to many more advanced and accurate electronic structure methods, while sacrificing the realism of representation. Locating the surface‐analyte complex is followed by an in‐depth in silico analysis of its energetic and electronic properties using, for example, energy decomposition schemes, as well as simulation of the signal, for example, a zero‐bias transmission spectra or a current–voltage curve, by means of the nonequilibrium Green's function method. These and other properties are examined in the context of a sensor's selectivity, sensitivity, and limit of detection with an aim to establish design principles for future devices. Herein, we analyze the advantages and limitations of diverse computational chemistry methods used at each of these steps in simulating graphene‐based electrochemical sensors. We present outstanding challenges toward predictive models and sketch possible solutions involving such contemporary techniques as multiscale simulations and high‐throughput screening. This article is categorized under: Structure and Mechanism > Computational Materials Science Electronic Structure Theory > Density Functional Theory Electronic Structure Theory > Ab Initio Electronic Structure Methods

Recent Advances in ZnO-Based Carbon Monoxide Sensors: Role of Doping

“…Undoubtedly, such drawbacks can be overcome if sensor materials are modeled using first-principle or ab initio methods. In this context, first-principle simulations based on the density functional theory (DFT) are essential for explaining and understanding experimental results at the molecular level, or for predicting novel gas sensors [ 85 , 86 , 87 , 88 ]. DFT-based simulations provide relevant critical information for designing novel gas sensors.…”

The Law of Attraction: Relating Computed Energetics of Physisorption with Performance of Graphene-Based Sensors for Nitroaromatic Contaminants

“…The exceptional electrochemical 13 and mechanical properties 14 of various graphene-based materials (GBMs, Figure 1B) make them particularly attractive to produce cheap, robust, and highly sensitive sensors. 15 Two-dimensional GBM sensors enable the detection of a broad range of compounds, from amino acids to metal cations; 16 such devices are also used for the subsequent removal of diverse aromatic contaminants from water. [17][18][19] In 2014, reduced graphene oxide functionalised with 1,3,6,8-pyrenetetrasulfonic acid sodium salt and palladium nanoparticles was used to manufacture a sensor with a low limit of quantification of NAC explosives, 20 however, this functionalised substrate is not economically viable for large-scale production.…”

Assessing in Silico Models and Methods for Non-Covalent Interactions between Graphene and Small Molecules

“…[4][5][6][7][8] Development and optimisation of graphene-based gas sensors, which typically operate via (non-)covalent interactions of adsorbates with the graphene surface, greatly benefit from theoretical insights into the strengths and nature of these interactions. 9 Examples of the properties studied in silico include adsorption geometries, energies, and charge transfer of small molecules (H2O, NO, NO2, NH3, and CO) adsorbed on graphene, 10 selectivity of NH3 detection with graphene nanoribbons, 11 and the role of surface defects in the adsorption of CO2 and CO on graphene. 12 In these studies, graphene and its derivatives were modelled as periodic systems, however, an infinite graphene sheet can instead be represented by a finite molecular fragment.…”

“…13 Choosing an appropriate model of the adsorbate-surface complex in conjunction with an electronic structure theory method, which affords an efficient sampling of adsorption geometries for a given adsorbate concentration regime, is a key to accurately simulate graphene-based gas sensors. 9 Periodic representation of the surface can be advantageously free of defects and edge effects; yet, on an ab initio level it is usually feasible only at the local density or generalised gradient approximations (LDA and GGA, respectively) of density functional theory (DFT), which cannot describe dispersive interactions without empirical corrections or non-local functional extensions. While more high-level density functionals, periodic second-order perturbation theory (MP2), random phase approximation (RPA), and the GW approach are available and able to address the aforementioned limitations of LDA and GGA DFT, they generally come at a prohibitively high computational cost.…”

“…While more high-level density functionals, periodic second-order perturbation theory (MP2), random phase approximation (RPA), and the GW approach are available and able to address the aforementioned limitations of LDA and GGA DFT, they generally come at a prohibitively high computational cost. 9 This is particularly the case for adsorption in the low coverage regime, where large surface slab models are required in addition to large vacuum gaps along the normal directions of the surfaces. Finite cluster models, on the other hand, can be treated with a broad spectrum of wavefunction-based methods, as well as density functionals from the higher rungs of Jacob's ladder.…”