Alfredo Cardenas, Postdoctoral Research Associate
Cornell University
Postdoctoral Research Associate in the Department of Computer Science at Cornell. Ph. D. of Chemistry earned at the University of Pittsburgh in 2000. Master and BS degrees at the Simon Bolivar University in Venezuela (1993 and 1988, respectively).
My main interest is Theoretical Physical Chemistry. This field of Chemistry applies physical and mathematical tools on complex atomic and molecular systems to understand and predict the properties of matter and light. Theorists in this field spend great deal of their time developing mathematical models and computer simulations to test those models.
My specific research interests have been varied. I have been involved
in studies of the statistical mechanics properties of liquid and vapour
fluids, non-linear optical processes, quantum dynamics of multi-dimensional
systems with non-adiabatic coupling between potential energy surfaces,
and semiclassical dynamics "on the fly" applied to linear spectroscopy
of condensed state systems.
Recently, I have developed models using 3D
lattice relaxation algorithm to study ion transport through channel proteins
embedded in lipid bilayers to understand the biophysical properties
of these systems.
Here at Cornell, I am working in algorithms to study long-time dynamics in
biomolecular systems.
Hobbies & Interests
The use of programming and computer software to simulate the physical reality using theoretical models. I build my own computer programs from scratch or develop existing codes to perform these simulations and also use different applications available commercially.I like reading, and as far as sports I like baseball, soccer and football.
Some of my research projects
| Gaussian Wavepacket Dynamics-Path
Integral (GWD-PI) to study the quantum dynamics of multidimensional systems
with curve crossing effects.
Photochemical processes involving molecules adsorbed on solid surfaces are of great experimental and theoretical interest. Upon illumination with light of an appropriate frequency, interactions of the adsorbate-surface system with photons can result in desorption, dissociation and other chemical reactions. These processes often occur on nonadiabatically coupled excited potential energy surfaces, which makes theoretical treatment of the relevant quantum dynamics difficult. We have developed an efficient and accurate method for computing the quantum dynamics of a molecule which photodesorbs from a solid surface at zero temperature in systems characterized by strong nonadiabatic coupling between several excited state potential energy surfaces. We have studied experimentally relevant spectroscopic properties for a model of a diatomic molecule adsorbed to a collinear crystal. To evolve the wavepackets on the two coupled potential energy surfaces we used a GWD-PI algorithm. This method enables us to compute the nuclear coordinate wavepacket evolution on either of the two surfaces at any time. We have shown that spectral signatures associated with the vibrational states of the desorbed molecule can be computed with reasonable accuracy for direct photodesorption dynamics on nonadiabatically coupled potential surfaces. We have also identified an important property of the GWD/PI technique for this class of problems that enables fast calculation of spectroscopic properties describing the dynamics of the desorbed particle, namely that only a few paths through electronic state space dominate the transition amplitudes of interest. We will use this ``most important paths'' property of the GWD/PI algorithm to extend the applicability of this model to more complicated systems. Two outstanding issues that we are addressing are the effects of finite temperature preparation of the initial system (i.e., considering temperatures greater than zero) and inclusion of non-collinear aspects of the desorption process, e.g. rotational motion of the desorbed molecule. |
On the Fly calculation of electronic
spectroscopy in condensed phase media.
We have developed a technique that determines the nuclear quantum dynamics using Gaussian Wavepackets Dynamics methods and potential energy surface information at the nuclei positions calculated ``on the fly'' (i.e., the information is updated at every time step during the nuclear evolution) using a semi-empirical electronic structure program. We have used this technique to study the influence of solvent interaction over the nuclear dynamics of solutes and obtained good agreement between our results and spectroscopic measurements for the systems studied, namely trans-octatetraene (a polyene) dissolved in chloroform and methanol. We have used QCFF/PI to compute the potential energy surfaces of the polyene and first derivatives at the position of the nuclei at each time step during the nuclear propagation. To simulate the dynamics of the solvent molecules we use a commercial classical molecular dynamics program. The interaction potential between solvent and solute molecules is considered to be electrostatic and Lennart-Jones types. |
through the Gramicidin A channel using three-dimensional
Poisson-Nernst-Planck Theory
Theoretical treatments of ion transport through channel proteins are important because they provide an understanding of the channel ionic selectivity and mechanisms involved in this dynamical process. Recently a lattice relaxation algorithm was developed to solve the Poison-Nernst-Planck (PNP) equations that govern the electrodiffusion dynamics of the charges through the channel. In this model the mobile ions are represented as a continuous charge density and the protein channel is described by discretizing it on a three-dimensional grid. This model was used to study ion transport through gramicidin A dimer, a peptide that forms highly selective positive ion channels across lipid membranes. Good agreement with several experimental measurements of current as a function of voltage applied across the channel was obtained.
We will use the PNP theory to study the effects of fixed charges and interfacial dipoles in the lipid membrane surface on ion conductance through the gramicidin channel. Using the lattice relaxation algorithm we will map the peptide and the charged lipid membrane onto a three-dimensional grid and solve the PNP equations numerically. By studying how charges and dipoles on the membrane surface affect the current-voltage curve, we will gain new insights into the mechanisms of ion transport in these biophysical systems.
Gramicidin channel with surrounding charges
Cornell University
Department of Computer Science
Ithaca, NY 14853
alfredo@cs.cornell.edu
http://www.cs.cornell.edu/people/alfredo