Date of Award

Spring 5-15-2016

Author's School

School of Engineering & Applied Science

Author's Department

Energy, Environmental & Chemical Engineering

Degree Name

Doctor of Philosophy (PhD)

Degree Type



New sources of energy that are environmentally friendly, cost-effective, and renewable are essential if we are to combat the effects of global climate change. Two of these sources are solar photovoltaic (PV) cells to convert sunlight into electricity and thermoelectric (TE) devices to convert heat to electricity. To be practical on a large scale, the properties (e.g. electrical conductivity, band gap, Seebeck coefficient, etc) of the the underlying materials must be improved significantly through judicious control of structure and composition. Significant understanding of materials properties is required to design and engineer new high-performing materials. First principles calculations using density functional theory (DFT) help us to predict materials behavior, with reasonable accuracy, before they are even made.

We first calculate the energy and density of states of electrons, the frequency of lattice vibrations (i.e. phonons), dielectric constants and deformation potential. We then use this as an input to the model that we have developed: ab initio model for calculating the mobility and Seebeck coefficient using Boltzmann transport equation (BTE), aMoBT. We solve the BTE via Rode's iterative method considering both elastic and inelastic mechanisms such as ionized impurity and polar optical phonon scattering mechanisms respectively to calculate the electrical mobility and conductivity and Seebeck coefficient of the semiconductor with minimum to no reliance on any experimental data. We have tested and validated aMoBT by predicting the electronic properties of GaAs, InN, Cu- and Al-doped ZnS all of which agree very well with experimental measurements. Without fitting to experimental data, aMoBT enables us to understand how the structure and composition of a given material exhibits certain properties, and facilitates screening of new materials without the need for costly and time-consuming experimental synthesis and characterization.

In addition to aMoBT, we have developed and adapted methods for ab initio thermodynamic calculations to understand the formation of neutral and charged point defects in semiconductors and their effect on the conduction mechanism; for example, we identified the negatively charged Zn vacancies in thermoelectric zinc antimonide (Zn4Sb3) as the reason for its native p-type behavior and the difficulty to n-dope it. Furthermore, we found that Cu and Al are feasible candidates to dope ZnS p- and n-type respectively to make promising transparent conducting materials (TCM) which then can be used in solar cells, LCD, touch screens, transparent devices, etc. These type of calculations are essential for predicting the behavior of real materials in presence of all the impurities and defects. Finally, combining these methods implemented as automated computer codes enables us to perform high-throughput screening of materials properties for design and discovery of new materials for specific applications. We screened 75 binary and ternary oxides to find, via aMoBT, the most conductive n-type and p-type oxides with reasonably large band gap as high performing TCM. We can use the same principles used in this work to predict semiconducotrs behavior for a broad range of applications.


English (en)


Cynthia Palghat S. Lo Ramachandran

Committee Members

Parag Banerjee, Pratim Biswas, Elijah Thimsen, Li Yang,


Permanent URL: http://dx.doi.org/10.7936/K7J38QTQ