Dissertation
From first-year chemistry to materials design: the versatility of computational chemistry
University of Iowa
Doctor of Philosophy (PhD), University of Iowa
Autumn 2021
DOI: 10.17077/etd.006274
Abstract
Computational chemistry is an ever-growing field, which utilizes first-principles methods to gain a greater understanding of the world, and is currently used in materials design, agriculture, and water remediation. While computational chemistry has foundations in quantum mechanics, the underlying trends and behaviors observed can be related to material taught in first-year courses. Many programs of study require students to take these foundational courses, but few of these first-year courses encompass the computational aspect of chemistry. To adapt the curriculum to meet the growing demands of data analytics in numerous fields, it is imperative to develop resources for the introduction of computational chemistry early in student careers. This may be done by utilizing a commonly used computational approach and tying it to topics in the first-year sequence of courses. To demonstrate the relationships between first-year coursework and materials design, the behavior of atomic orbitals, molecular orbitals, and periodic trends are related to corrective terms used in predictive material design. To accomplish this, we first investigate the use of an open-source code in teaching electron configurations, the Aufbau principle, and ionization energies to first-year students. We show that use of this code provides students with an interactive, hands-on approach and permits them to develop an enhanced understanding of concepts, rather than relying on rote memorization techniques. Additionally, it provides students with valuable data analysis skills, which are appreciated in the workplace.
Periodic trends were then further studied through use of the widely used first-principles computational approach, density functional theory (DFT). DFT has been used to simulate an array of systems from atomic to solids utilizing assumptions based on the electron density. The introduction of DFT has significantly decreased the complexity of describing a many-atom system. However, DFT is not entirely free from potential errors when dealing with transition metal oxide systems, which involve significant interactions between the metallic and nonmetallic atoms. DFT has been known to over-delocalize the electrons belonging to the metallic species, and at times, results in the underestimation of band gaps. One option to address this deficiency is to apply a corrective potential barrier to delocalization, which is known as the Hubbard U. We look at a systematic derivation of Hubbard U values for three distinct crystal structures using the GBRV set of ultrasoft pseudopotentials. We relate the derived U values to periodic trends taught in first-year coursework, and determine how U can vary with oxidation state and in which situations it may be acceptable to use a single U value for multiple oxidation states.
Lastly, the DFT + thermodynamics approach is analyzed to determine how the model can be adapted to work for an array of systems to provide greater understanding of the energetics driving reactions. This model builds connects between fundamental thermodynamics including Gibbs free energy and the Nernst equation, and computational methodology. Applications of the model for substitution, dissolution, and dissolution in the presence of media other than water are examined, and the accuracy of the model in these situations analyzed. I discuss how this approach has been applied to nanomaterials relating to biology, engineering, recycling and agriculture to enhance our understanding on how reactions occur. By incorporating the data from these studies into classroom problem-solving activities, students are able to build connections between what they’re learning and how it relates to their fields of studies.
Details
- Title: Subtitle
- From first-year chemistry to materials design: the versatility of computational chemistry
- Creators
- Irene K. Metz
- Contributors
- Sara E. Mason (Advisor)Christopher M. Cheatum (Committee Member)Renee S. Cole (Committee Member)Nicole M. Becker (Committee Member)Tori Z. Forbes (Committee Member)
- Resource Type
- Dissertation
- Degree Awarded
- Doctor of Philosophy (PhD), University of Iowa
- Degree in
- Chemistry
- Date degree season
- Autumn 2021
- DOI
- 10.17077/etd.006274
- Publisher
- University of Iowa
- Number of pages
- xviii, 333 pages
- Copyright
- Copyright 2021 Irene K. Metz
- Language
- English
- Description illustrations
- color illustrations
- Description bibliographic
- Includes bibliographical references.
- Public Abstract (ETD)
In this thesis, the relationship between first-year chemistry topics and current computational research methodology is discussed. As computational chemistry and related fields of data analytics are rapidly growing, it is important to ensure the curricula students are exposed to in their first-year courses is encompassing and comprehensive. Once the theory behind the methods is discussed, means of communicating the theory is reviewed. This resulted in the enhancing the readability of tutorials for an open-source software, and designing a laboratory for first-year chemistry students that exposes them to computational chemistry. This activity also addresses an area of concern for faculty: fundamental quantum mechanics. Later, building on material from first-year chemistry, periodic trends and their applications in deriving corrective terms and understanding observations in data are also addressed. Finally, a review of the relationship between Gibbs free energy, the Nernst equation, and their relationship with computational chemistry is addressed to determine how first-year chemistry and computational chemistry be used to guide a greater understanding the energetics of various chemical reactions for various fields of study.
- Academic Unit
- Chemistry
- Record Identifier
- 9984210444502771
Metrics
13 File views/ downloads
45 Record Views