A theoretical study of HCN adsorption and width effect on co-doped armchair graphene nanoribbon

[Display omitted] •Adsorption energy of hydrogen cyanide (HCN) gas is investigated by DFT theory.•Adsorption of HCN gas on ArGNR was enhanced by introducing defects and co-doping.•Energy gap of ArGNRs is inversely proportional to ribbon width.•Co-doping effects transforms ArGNRs from semiconductor to metallic behaviour.•Structural and electronic properties of ArGNRs were changed by doping and defects. Hydrogen cyanide (HCN) is one of the hazardous gas and requires sensitive detection material. In the present work stable structures, adsorption analysis, charge transfer and electronic properties of defective, pristine, Phosphorus doped, Boron doped and Boron Phosphorus co-doped ArGNRs are studied using first principles calculations combined with Density functional theory (DFT). In two models of ArGNRs with N = 6 and N = 8 atoms along width, it is observed that pristine ArGNR is not a good sensing material for HCN molecule however, on introducing defects and dopants its sensitivity drastically increases and results in chemisorption for all models of ArGNRs. Different doping site optimizations for all models of ArGNR are observed which reveal that defective, Phosphorous doped, Boron-Phosphorus co-doped ArGNR with width N = 6 shows higher negative value of adsorption energy with adsorption energy values are −5.81 eV, −1.27 eV and −1.06 eV. Band nature predicts that for BP-ArGNR with N = 6 shows large reduction as compared to co-doped system of N = 8 model system. A remarkable charge transfer occurs between the HCN molecule and BP-ArGNR as proved by the electronic charge densities. DOS analysis predicts that the defective, phosphorus doped and boron-phosphorus co-doped ArGNRs are good sensing materials with intense peaks after adsorption of cyanide molecule as compared to pristine ArGNR for N = 6 model system. This implies that defective, Phosphorus and B, P co-doped ArGNR with N = 6 has high sensitivity to cyanide molecule as compared to N = 8 model system. Results of our simulated work are not only helpful to evaluate sensing mechanism but also provide a potential sensor material detection of HCN.