Abstract
We present an ab initio modeling study of steady-state and time-dependent charge transport in hole-only devices of the amorphous molecular semiconductor α–NPD [N,N ′ −Di(1–naphthyl)−N,N ′ −diphenyl−(1,1 ′ −biphenyl)−4,4 ′ −diamine]
α–NPD [N,N′-Di(1–naphthyl)-N,N′-diphenyl-(1,1′-biphenyl)-4,4′-diamine]
. The study is based on the microscopic information obtained from atomistic simulations of the morphology and density functional theory calculations of the molecular hole energies, reorganization energies, and transfer integrals. Using stochastic approaches, the microscopic information obtained in simulation boxes at a length scale of ∼10 nm is expanded and employed in one-dimensional (1D) and three-dimensional (3D) master-equation modeling of the charge transport at the device scale of ∼100 nm. Without any fit parameter, predicted current density-voltage and impedance spectroscopy data obtained with the 3D modeling are in very good agreement with measured data on devices with different α-NPD layer thicknesses in a wide range of temperatures, bias voltages, and frequencies. Similarly good results are obtained with the computationally much more efficient 1D modeling after optimizing a hopping prefactor
α–NPD [N,N′-Di(1–naphthyl)-N,N′-diphenyl-(1,1′-biphenyl)-4,4′-diamine]
. The study is based on the microscopic information obtained from atomistic simulations of the morphology and density functional theory calculations of the molecular hole energies, reorganization energies, and transfer integrals. Using stochastic approaches, the microscopic information obtained in simulation boxes at a length scale of ∼10 nm is expanded and employed in one-dimensional (1D) and three-dimensional (3D) master-equation modeling of the charge transport at the device scale of ∼100 nm. Without any fit parameter, predicted current density-voltage and impedance spectroscopy data obtained with the 3D modeling are in very good agreement with measured data on devices with different α-NPD layer thicknesses in a wide range of temperatures, bias voltages, and frequencies. Similarly good results are obtained with the computationally much more efficient 1D modeling after optimizing a hopping prefactor
Original language | English |
---|---|
Article number | 2433301 |
Pages (from-to) | 1-4 |
Journal | Applied Physics Letters |
Volume | 109 |
Issue number | 24 |
DOIs | |
Publication status | Published - Dec 2016 |