Energy storage and generation from thermopower waves

Abstract

The nonlinear coupling between an exothermic chemical reaction and a nanowire or nanotube with large axial heat conduction guides a self-propagating thermal wave along the nano-conduit. The thermal conduit accelerates the wave by rapidly transporting energy to un-reacted fuel. The reaction wave induces what we term a thermopower wave, resulting in an electrical current in the same direction. At up to 7 W/g, peak power density is larger than that of many present micro-scale power sources (e.g. fuel cells, batteries) and even about seven times greater than commercial Li-ion batteries. Thermopower waves also tend to produce unipolar voltage pulses, although conventional thermoelectric theory predicts bipolar voltage. These waves also generate thermopower in excess of previous measurements in carbon nanotubes (CNTs) and therefore could increase figures of merit in a variety of thermoelectric materials. In this thesis, I have developed the theoretical framework to describe the thermal and chemical profiles of propagating reaction waves, and their electrical properties. My analysis yielded a new analytical solution for one-dimensional reaction and thermal diffusion systems with nth order kinetics that obviates many approximate or numerical approaches from the past 80 years. A generalized logistic. function describes the temperature and concentration profiles within the solid fuel and provides a solution for the wave velocity for a wide range of conditions. This approach offers new insight into such problems spanning several fields in science and engineering, including propulsion and self-propagating high temperature synthesis (SHS) of materials, as well as the dynamics of thermopower waves. Temperature and voltage measurements of thermopower waves on CNTs show that they can generate power as much as four times greater than predictions based on reference measurements of the Seebeck coefficient for static temperature gradients. We hypothesize that the excess thermopower stems from a chemical potential gradient across the CNTs. The fuel (e.g. picramide) adsorbs and dopes the CNTs ahead of the wave and desorbs and reacts behind the wave front. Furthermore, the excess thermopower depends on the mass of fuel added (relative to CNT mass), and the chemical potential difference matches the magnitude of the excess thermopower. Thus, a major conclusion of this thesis is that coupling to a chemical reaction can boost the performance of thermoelectric materials through differential doping. Thermopower waves can have well defined velocity oscillations for certain kinetic and thermal parameter values. Cyclotrimethylene-trinitramine (fuel) on multiwalled CNTs (conduit) system generates voltage oscillations of 400 to 5000 Hz. These frequencies agree with velocity oscillations predicted by my thermochemical model of the reaction wave, extended to include thermal transport within the conduits. Thermopower waves could thus find applications as new types of alternating current (AC) batteries and self-powered signal generators, which could easily be miniaturized. Microelectromechanical systems and sensors would benefit from thermopower wave generators to enable functions such as communications and acceleration that currently require large power packs. Additionally, the "self-discharge" rate of thermopower wave generators is extremely low in contrast to electrochemical storage, since their energy is stored in chemical bonds. Thermopower waves thus enable new energy storage devices and could exceed limitations of conventional thermoelectric devices.