Simulation of Nanodevices
Kinetic Monte Carlo Simulations
An alternative approach is based on kinetic Monte Carlo methods.
Here, the microscopic processes controlling the electrical properties of a given nanostructure are described in terms of rate equations which are in turn solved using Monte Carlo stochastics algorithms.
This approach has been applied to the analysis of transport across novel high-k dielectrics (e.g. ZrO2).
Ab-initio simulations can be coupled to the Monte Carlo, thus enhancing the reliability of the method.
Such tools have a high predictive content and can be used to design new materials and new structures.
The advances in nanotechnology have led to the introduction of several novel materials, nanostructures and nanodevices.
Simulations can greatly help in addressing their potentials in electronics and optoelectronic applications.
There is a variety of simulation tools which range from ab-initio tools capable of calculating the electronic properties of nanostructures exactly with no extra parameters to phenomenological approaches that can self-consistently simulated complex devices and circuits.
In such hyrerchy of approaches, the former can be very accurate but require huge computer resources, while the ladder are approximate but can run very fast.
One example of an atomistic calculation has been presented earlier in the contest of single molecule devices.
The more phenomenological Drift-Diffusion approach is actually the work-horse of the semiconductor industry.
We have used a commercial simulator (from Synopsys) for the study of organic devices (TFTs, photodetector and solar cells, as also already mentioned).
In addition we have simulated diodes based on ZnO. Such devices can be used as non linear elements in non-volatile memories based on cross-bar architectures.
By coupling the drift-diffusion code with a three-dimensional Poisson solver a self consistent simulation of the device (including leads) can be performed, which provides very useful information on the device operation.
Jegert G. et al. Monte Carlo simulation of leakage currents in TiN/ZrO2/TiN capacitors. IEEE Transactions on Electron Devices 58, 327 (2011). doi: http://dx.doi.org/10.1109/TED.2010.2090158
Arcari, M. et al. 2-D Finite-Element Modeling of ZnO Schottky Diodes With Large Ideality Factors. IEEE Transactions on Electron Devices 59, 2762 (2012). doi: http://dx.doi.org/DOI:10.1109/TED.2012.2207459
Laser Modeling and Simulation
The quantum cascade laser (QCL) is a novel type of compact semiconductor laser, which can emit light in the mid-infrared and terahertz range.
These frequency regimes are very attractive for many applications, such as spectroscopy and communications.
The development of reliable simulation tools is essential for a further improvement of these devices. For example, a major challenge is that terahertz QCLs up to now only work at cryogenic temperatures.
Based on the ensemble Monte-Carlo method, we develop highly optimized modeling tools to investigate the electron transport and optical properties of QCL structures, to identify detrimental effects in QCLs, and to design improved devices.
One milestone of our work has been the first coupled simulation of the electron transport and the optical cavity field in a QCL, allowing us to self-consistently model the actual lasing operation.
As an example, the simulated outcoupled laser power is shown below for a terahertz QCL.
Furthermore, the simulated energy resolved electron density is displayed for a highly efficient mid-infrared QCL.
Based on our simulations, we can systematically improve experimental QCL designs.
In this context, we are involved in a collaboration which has very recently resulted in a terahertz QCL with a world record operating temperature (~200 K).
Mátyás, A., Lugli, P., and Jirauschek, C. Photon-induced carrier transport in high efficiency midinfrared quantum cascade lasers. J. Appl. Phys. 110, 013108 (2011). doi: http://dx.doi.org/10.1063/1.3608116
Modeling Novel Devices and Architectures
Cross-bar Architectures Based on Semiconducting / Molecular Nanostructures
Cross-point architectures are the simplest functional nanodevices, possessing a very regular geometry and using only two-terminal passive devices.
This is why they hold a great promise in nanoelectronics, where fabrication challenges prohibit making a more complex, arbitrarily interconnected circuit.
Cross bar architectures can be of interest in memory technology for non volative resistive RAMs and can also find application in cryptography as highly secure physical unclonable functions.
Recently a lot of interest has been attributed to circuit architectures built from memristive crossbars whose nonlinear behavior resembles that of neuronal networks.
Circuit Modeling of Q-bits and Superconducting Microwave Resonators
Superconducting quantum devices are emerging as possible realization for quantum computers. We are designing microwave circuits to manipulate and read-out quantum bits.
We combine well-established engineering software packages for a finite-element solution of the electromagnetic field problem and develop "circuit models" to understand more about the coupling of electromagnetic fields and quantum systems.
Mátyás A. et al. Linear Circuit Models for On-Chip Quantum Electrodynamics. IEEE Transactions on Microwave Theory and Techniques 59, 65 (2011). doi: http://dx.doi.org/10.1109/TMTT.2010.2090406
Rührmair U. et al. Applications of High-Capacity Crossbar Memories in Cryptography. IEEE Transactions on Nanotechnology, 10, 489 (2011). doi: http://dx.doi.org/10.1109/TNANO.2010.2049367