AmericanChemicalSociety.com

The two students, Anna Brockway '12 and Alex Vargo '12, who started work on calculating the lattice thermal conductivity of nanowires withdefects during Summer 2009, continued their work during 2009-2010 academic year.  By May 2010, they had worked out all the remainingdetails necessary for determining the thermal conductivity using our chosen reverse-non-Equilibrium molecular dynamics mechanism, and had obtained the initial data on thermal conductivities as a function of nanowire radius, which agreed with past experiments and theoreticalcalculations.  Both students did external REU projects over the past summer; in their absence I was able to resolve some of the remaining data-processing issues.  During the 2010-2011 academic year they are both continuing work on this project and are currently performing the remaining calculations needed to finish our first study on the thermal conductivities of the four elementary stacking fault types for wurtzite nanowires.  In addition, several potential collaborators are interested in applying this methodology to computing the thermal conductivity of organic thermoelectric materials; from our perspective this only requires using a different force field, as the rest of our procedure is completely general.

I incorporated a simplified version of these molecular dynamics simulations  into the Fall 2010 Junior-level "Superlab" course. During the past academic year, I have also incorporated readings and exercises on a simplified version of the 1-dimensional band-structure calculations into the Junior-level Quantum Chemistry course, and during the current academic year I am incorporating a segment on thermoelectricity in the Junior-level Thermodynamics course; this has been helpful as it has forced me to consider chemical analogies to thermoelectrics, which led to one of the spin-off project below.  In addition, I trained two additional students, Ethan Glor '11 and Samuel Blau '12in the use of planewave-pseudopotential density functional calculations of solids.  The initial applications were to solid compounds that they had both synthesized in the laboratory of Prof. Alexander Norquist at Haverford, those projects are currently nearing completion.  Following this training, during the next year, Mr. Blau plans to take on the electronic-structure portion of the thermoelectric modeling project.

In addition, this study of thermoelectrics has resulted in two spin-off projects:  (1) By studying the thermoelectric effect in general terms, I came to the conclusion that one could consider a"thermoelectrochemical" effect, in which temperature gradients are used to produce "separation" work (i.e., Gibbs energy of mixing) instead of electrical work, if one has energy selective transmission, such as quantum tunneling.  Based on this insight, I studied the feasibility of using an existing, chemically-synthesized porous graphene membrane to separate Helium from natural gas.  Interestingly, quantum mechanical tunneling plays a significant role, even at roomtemperature.  Using the mass-dependence of the tunneling probability, one can achieve up to an 8% relative concentration enhancement of 3He/4He under steady state conditions, using just a temperature gradient to drive the system. (2) The simple pi-electron model used to develop and test our thermoelectric calculations was subsequently used to study multiple exciton generation in graphene nanostructures.  The student working on this project studied over 800 structures and identified optimal structures for which one high energy photon can yield an average of 1.3-1.8 excitons (depending on whether one uses an upper or lower bound for the exciton cooling rate).  Closely related polyaromatic hydrocarbons have recently been used for dye-sensitized solar cells; using our structures instead would lead to increased photocurrent in the resulting solar cells, boosting the overall efficiency.