• 1:40 scaled physical model of a longwall coal mine to investigate mine ventilation and the formation of explosive gas zones, A

      Brune, Jürgen F.; Bogin, Gregory E.; Honorato Pinheiro Junior, Hamilton; Miller, Hugh B.; Kaunda, Rennie (Colorado School of Mines. Arthur Lakes Library, 2021)
      The formation of explosive gas zones (EGZs) is a critical problem in longwall coal mines. The investigation of how EGZs might form currently relies on Computational Fluid Dynamics (CFD) models, which have limitations around long solution times and availability of certain validation data in a longwall mine. Physical, scaled models are an alternative to investigating dynamic fluid behavior under complex scenarios that range from aircraft design to airway investigation. Scaled modeling requires extensive dimensional analysis evaluation to replicate complex airflow phenomena adequately. This work presents design and manufacturing considerations to build a 1:40 scaled version of a longwall coal mine to investigate mine ventilation strategies and the formation of EGZs. The physical model presented is the only known scaled model of a longwall coal mine built on a modular design and capable of simulating different ventilation strategies, longwall face advance, and shearer motion. The physical model has a mine-wide atmospheric monitoring system (AMS) capable of measuring airflow speed and gas concentrations in the air courses, longwall face, and gob region. Initial experimental data results prove the model to be consistent with CFD models and published ventilation data of longwall coal mines. The physical model captured the trends of flow leakage and simulated methane accumulation across the longwall face compared to published data of actual mines and CFD models.
    • 3D forward modeling and parametric inversion of induced polarization with electromagnetic coupling

      Swidinsky, Andrei; Donmez, Deniz; Li, Yaoguo; Krahenbuhl, Richard A. (Colorado School of Mines. Arthur Lakes Library, 2021)
      The induced polarization (IP) method is commonly used in mineral exploration. In particular, it has been successfully applied to explore for disseminated sulfide and/or porphyry deposits, which often have strong induced polarization responses. IP methods can provide information about the chargeability and resistivity distribution of the subsurface. However, these methods may underperform under certain geological conditions, such as in the presence of highly conductive or resistive cover. In the presence of resistive cover, the direct current (DC) signal cannot penetrate beneath this zone, and in the presence of conductive cover, a short circuit results. In such cases, the inductive response of the Earth can potentially contain information about the subsurface geological units due to complimentary physics to the DC resistivity method. This is because the current flow generated by electromagnetic (EM) induction can penetrate resistive features. However, EM coupling is generally considered to be noise and unwanted content in the DC/IP signal. In the traditional processing and interpretation of IP data, inductive effects are usually eliminated. Although removing the electromagnetic effect to increase the sensitivity to IP signal is the main focus of the mineral exploration industry, there is a possibility to extract more information about the subsurface by using EM induction. However, in order to handle this multi-parameter physical problem, traditional DC/IP modeling is deficient. Moreover, the conductivity model needs to be defined by a frequency or time-dependent complex conductivity. Throughout this thesis, we investigate the relationship between IP and EM by simulating grounded source time-domain DC/EM/IP data with a complex conductivity model called the stretched exponential with SimPEG and SimPEG EMIP. This method allows us to model IP relaxation in the time-domain instead of using the Cole-Cole model (CC) in the frequency-domain. Cole-Cole in the frequency-domain is a standard approach in the context of complex conductivity modeling. However, it is computationally expensive to apply in time-domain since Fourier transform is needed. With the stretched exponential, one can simulate a grounded source time-domain DC/EM/IP survey with complex conductivity more efficiently. We used different geoelectrical models such as a chargeable conductive block in a resistive subsurface, a chargeable resistive block in a conductive subsurface, a chargeable conductive block buried under a relatively conductive cover, a chargeable conductive block buried under a relatively resistive cover, and a 3D porphyry-mineral deposit model. By using the full DC/EM/IP step response, we calculated apparent resistivities by using the peak arrival time of the signal obtained from normalized pseudoimpulse response. Our results show that this method contains more information about the subsurface than the traditional DC/IP method. Furthermore, we apply a parametric inversion to a simple block model and show that grounded source DC/EM/IP data with complex conductivity can be inverted. However, a full 3d inversion is necessary to characterize this highly non-unique and difficult EM problem.
    • A treatise on the use of pseudopressure approach for the transient analysis of multiphase production from unconventional reservoirs

      Ozkan, E.; Al Ali, Mohammed M.; Firincioglu, Tuba; Kazemi, Hossein (Colorado School of Mines. Arthur Lakes Library, 2021)
      This thesis presents a formal discussion of the use of the pseudopressure approach to analyze multiphase pressure- and rate-transient data of fractured wells in unconventional reservoirs. Among several options for the treatment of the multiphase flow problem, pseudopressure approach is an attractive semi-analytical choice for pressure- and rate-transient analysis as it reduces the problem to a single-phase flow analogy and enables the use of standard well-test analysis tools and techniques. However, the pseudopressure approach is inherently approximate and its success is strongly dependent on the validity of the assumptions made. Therefore, the appropriate use of the pseudopressure approach requires good documentation and understanding of its underlying assumptions. Most introductions of the pseudopressure approach for multiphase flow in naturally fractured media are based on numerical observations, extensions from known cases, or even by heuristic arguments, which do not permit explicitly delineating the parametric relations and associated assumptions. Therefore, in this thesis, both numerical and analytical approaches are used to document the assumptions and limits of application of the pseudopressure approach for multiphase flow in unconventional reservoirs. For ease of discussion, the problem is reduced to two-phase (oil and gas) linear (1D) flow in a naturally fractured porous medium. The dual-porosity idealization with transient interporosity flow option is used to represent the naturally fractured reservoir. To emulate the properties of unconventional reservoirs, the matrix is assumed to be very tight and the majority of the pore and pore-throat sizes are in the range of tens of nanometers. Due to the extreme pore proximity, the effect of capillary pressure in the matrix system is taken into consideration. The numerical model used in this research is constructed with COZSim, which is a specialty black-oil simulator developed to include the effect of pore proximity on phase behavior. The numerical model is verified under single-phase flow conditions by comparison to the analytical model in the Topaze software of Kappa. In this study, an analytical derivation of the total flow equation in terms of pseudopressure is also presented for two-phase flow in a linear dual-porosity medium. The analytical model provides the basis of the pseudopressure definition and verifies the conditions under which the single-phase flow analogy can be used to analyze multiphase flow responses. An example case is generated by the numerical model and the results in terms of pseudopressure are compared with the corresponding single-phase responses. Comparison of the numerical results with the single-phase analog and the analytical derivations indicate that a single definition of pseudopressure is not possible for all times. The problem is caused by the capillary pressure difference between the matrix and fracture systems. While neglecting the effect of capillary pressure in pseudopressure definition causes deviation from single-phase results at late-times, inclusion of capillary pressure provides a match at late times but causes deviation at early times. The derivation of the analytical model indicates that single-phase definitions of the dual-porosity parameters (storativity and transmissibility) can be extended to multiphase flow as a diffusivity-weighted sum of the individual-phase parameters. The analytical model also provides early-, intermediate-, and late-time asymptotic solutions to be used in the straight-line analysis. The asymptotic solutions reveal the explicit form of the effective properties estimated from straight-line analysis. Finally, a synthetic example is analyzed by using the producing gas-oil ratio (GOR) to generate the saturation profile. Differences are observed between the produced GOR saturation predictions and the simulated saturations. This indicates, for dual-porosity reservoir models, the produced GOR method may not be a suitable option to predict the saturation profile considering the sharp pressure and saturation difference between the matrix and fracture due to the capillary pressure. Therefore, the produced GOR method may need to be enhanced in order to be applied for dual-porosity reservoirs which requires a further research.
    • Acceleration of alloy design and manufacturing via machine learning and automated optimization

      Zhang, Xiaoli; Liu, Sen; Stebner, Aaron P.; King, Jeffrey C.; Kappes, Branden Bernard; Amin-Ahmadi, Behnam (Colorado School of Mines. Arthur Lakes Library, 2021)
      Unveiling the composition – process – structure – property (CPSP) relations is challenging and a perpetual pursuit for alloy design and manufacturing processes. The applications of machine learning (ML) or artificial intelligence (AI) techniques to these research fields are emerging and expected to be a common practice in the future. The key behind this expectation is those unknown domains can be accurately estimated by learning from sufficient training examples. However, on the one hand, the experimental data are usually insufficient, highly sparse, and used as private resources, which poses a significant challenge in applying ML/AI approaches to real-world engineering problems. Alloy design or metal additive manufacturing (AM) are typical examples and often conduct limited amounts of high-cost experiments. The samples produced have high CPSP relations in terms of multiple length/time scales, multiple physics fields, and the “curse of dimensionality” in statistical learning. On the other hand, with the advancement of characterization instruments, a wealth of data is generated, such as microscopy and X-ray diffraction images that far outpace domain knowledge can interpret and analyze in real-time.Based on these challenges, the research of this thesis explores the state-of-the-art ML/AI techniques to assist materials manufacturing, property design, process optimization, and characterization. Specifically, the methods in the thesis automate the microscopy image/data analysis and make ML with limited experiments data to speed up scientific discovery and improve modeling performance. The results highlight the possibility or potential that a large fraction of manual work in the alloy design and additive manufacturing processes will be assisted with ML/AI-based automation systems. The methods developed will benefit a broad range of materials fields and enable high-throughput materials design and manufacturing.
    • Adapting conditional simulation using circulant embedding for irregularly spaced data and exploring its limits

      Nychka, Douglas; Bandyopadhyay, Soutir; Bailey, Maggie; Ryan, Jennifer K.; Thompson, Eric (Colorado School of Mines. Arthur Lakes Library, 2021)
      For both scenario and real earthquakes, it is important to estimate the resulting shaking intensity in the near-epicentral region. These estimates can be constrained with ground motion recordings from real earthquakes, as is currently done in near-real-time by the United States Geological Survey (USGS) ShakeMap software. For scenario earthquakes, the ground motions are only estimated by the application of empirical models. The ground motion estimates are expressed as maps of the mean and standard deviation of the estimated intensity, but the consequences of the shaking on society and the built environment depend critically on the spatially correlated aleatory variability of the ground motions. This variability has been accounted for by generating random, spatially correlated realizations of the ground motions. Methods for generating these realizations, however, are computationally demanding, especially when the estimates are conditioned on numerous observed intensity values. In this presentation, a new and approximate conditional simulation approach is applied for use in ShakeMap. This approach, termed circulant embedding (CE), builds on a fast method for simulating Gaussian processes. However, standard CE is restricted to simulating stationary Gaussian processes (possibly anisotropic) on regularly spaced grids. It is also known that if the range parameter of a spatial process is large relative to the domain, this method fails. In this work we explore new algorithms that adapt CE for (a) irregularly spaced data points, and (b) methods for working with large range parameters in order for CE to be widely applicable. It is found that one method provides better accuracy and efficiency, and the solution to failure of CE results in manageable error for an exponential covariance function. However this error increases for larger shape parameters. We also illustrate the computational efficiency of this approach relative to previous methods. These ideas are illustrated with ground motion intensity measures and also validated through simulation studies.
    • Advanced methods for bastnaesite concentrate leaching

      Anderson, Corby G.; O'Kelley, Brock; Colligan, Grant T.; Taylor, Patrick R. (Colorado School of Mines. Arthur Lakes Library, 2021)
      With the world’s rapid advancement of technology, the demand and need for the materials that make up that technology has exploded. An important group of materials needed in this rapid advancement are the rare earth elements (REE) used in countless necessary applications. One source of rare earths found in the United States is bastnaesite, a rare earth bearing fluorocarbonate, mined at the Mountain Pass Mine in California. To increase production, it has been essential to optimize existing processes and create new ones to exploit current reserves. This research program was run to expand the understanding of the bastnaesite leaching system. A novel single stage hydrochloric leach system was created to optimize the rare earth extraction from bastnaesite. Typically, this process has utilized a two-stage leach system involving a high temperature hydrochloric acid leach followed by a caustic crack.A series of single stage leach experiments were run by reacting locked cycle REE flotation concentrate with hydrochloric acid. The results of these single stage tests showed rare earth recoveries significantly higher than reported recoveries of the leach used in the historic two stage caustic crack process. These results also indicated a decrease in the amount of reagents needed to extract rare earths. This novel single stage leach was than assessed for economic performance compared to the historic two stage caustic crack process using a cashflow value added model. This economic analysis showed that the single stage leach had better economic performance showing higher created economic value and lower total costs over the lifetime of the process. This economic analysis also identified that the driving force on profitability is the rare earth leaching recovery. In summary, the results of this research a novel single stage hydrochloric leach was developed and shown to recover rare earth elements from concentrated bastnaesite ore significantly higher than the leach of the historic two stage caustic crack process, while using less reagents, and creating more economic value.
    • Advances in spatial frequency modulation imaging: spatio-spectral encoding

      Squier, Jeff A.; Czerski, John; Brice, Craig Alan, 1975-; Adams, Daniel; Durfee, Charles G. (Colorado School of Mines. Arthur Lakes Library, 2021)
      In the following thesis I present my work advancing the state of the art in spatial frequency modulation imaging (SPIFI). There are many imaging scenarios in which traditional segmented detectors are impractical. Segmented detectors become prohibitively expensive outside of commonly used wavelength ranges and are susceptible to scattering. In such cases, single element detectors provide access to a larger range of wavelengths and may remove scattering ambiguity; however, they require a method for coupling spatial information into temporal signals. SPIFI couples spatial information into temporal signals from a single element detector by modulating the spatial intensity distribution of light illuminating the sample. This work extends the utility of SPIFI by developing simple optimization algorithms for the modulation pattern, developing a fiber deliverable SPIFI system, two multi-dimensional SPIFI systems, and a single shot SPIFI system. The thesis is organized into six chapters. I begin with an introduction to single element imaging and SPIFI. In chapter two I describe how to optimize the SPIFI modulation pattern for various constraints such as the manufacturing resolution or the numeric aperture of an optical system. Chapter 3 describes the technique I developed for fiber deliverable SPIFI imaging: wavelength domain SPIFI. By modulating the spectrum of the illumination beam, wavelength domain SPIFI facilitates remote delivery via optical fiber or free space transmission. I provide theoretical analysis of wavelength domain SPIFI along with experimental validation of the technique and its compatibility with fiber delivery. Wavelength domain SPIFI can be combined with SPIFI along the perpendicular transverse dimension. In chapter 4 I present experimental realization of a two dimensional SPIFI system with spatial and wavelength modulation. I also demonstrate a scan free SPIFI system where the wavelength encoded axis is sampled with a linear CMOS detector. Chapter 5 presents initial results from wavelength multiplexed single shot SPIFI. This technique provides a valuable illumination scheme for phenomena that occur faster than the scan time of traditional SPIFI. I end the thesis by describing future work related to these techniques and presenting some concluding statements.
    • Advancing cyanobacterial production of sustainable chemical feedstocks

      Boyle, Nanette R.; Newman, Darrian M.; Ramey, C. Josh; Krebs, Melissa D. (Colorado School of Mines. Arthur Lakes Library, 2021)
      The temperature on earth is rising. Studies show that the average surface temperature is already 1°C higher than in pre-industrial times. This rise in temperature is directly linked to anthropogenic carbon emissions and will only continue to rise if immediate action is not taken. Work to improve sustainability in our society has already begun, with vast improvements in solar, wind, geothermal, and other forms of sustainable energy production being made every day. Fossil fuel dependency is pumping carbon dioxide into our atmosphere at unsustainable rates, and cyanobacteria present a feasible means to combat emissions and close the carbon loop for a more sustainable future. Cyanobacteria are prokaryotic, photosynthetic microorganisms with the capability to produce a wide range of chemicals, from biofuels to plastics to sugar to rubber, all while utilizing minimal resources. Only sunlight, trace minerals, unarable land, and sub-optimal water such as ocean water are needed for growth and production of valuable specialty and commodity chemicals. With a higher photosynthetic eciency and genetic tractability than plants and no competition for arable land or fresh water, cyanobacteria have the potential to be pivotal in the ght against climate change. In this thesis, I characterize the central metabolism of a biofuel producing strain of cyanobacteria to gain a better understanding of how to manipulate cyanobacterial metabolism for increased production of not only biofuels but a range of valuable products. I probe pathways of primary and secondary metabolism using metabolic engineering techniques in an eort to glean useful information and help demystify the complex metabolic network of these complex organisms. Additionally I improve and validate existing tools needed to accelerate the design-build-test-learn cycle and establish cyanobacteria as a commercially viable sustainable producer of valuable products.
    • Adversarial machine learning in computer vision: attacks and defenses on machine learning models

      Yue, Chuan; Qin, Yi; Camp, Tracy; Han, Qi; Belviranli, Mehmet; Mohagheghi, Salman (Colorado School of Mines. Arthur Lakes Library, 2021)
      Machine learning models, including neural networks, have gained great popularity in recent years. Deep neural networks are able to directly learn from raw data and can outperform traditional machine learning models. As a result, they have been increasingly used in a variety of application domains such as image classification, natural language processing, and malware detection. However, deep neural networks are demonstrated to be vulnerable to adversarial examples at the test time. Adversarial examples are malicious inputs generated from the legitimate inputs by adding small perturbations in order to fool machine learning models to misclassify. We mainly aim to answer two research questions in this thesis: How are machine learning models vulnerable to adversarial examples? How can we better defend against the adversarial examples? We first improve the effectiveness of adversarial training by designing an experimental framework to study Method-Based Ensemble Adversarial Training (MBEAT) and Round Gap Of Adversarial Training (RGOAT). We then demonstrate the strong distinguishability of adversarial examples and design a simple yet effective approach called defensive distinction under the formulation of multi-label classification to protect against adversarial examples. We also propose fuzzing-based hard-label black-box attacks against machine learning models. We design an AdvFuzzer to explore multiple paths between a source image and a guidance image, and design a LocalFuzzer to explore the nearby space around a given input for identifying potential adversarial examples. Lastly, we propose a key-based input transformation defense to defend against adversarial examples.
    • Analog and digital adiabatic quantum annealing with oscillating transverse fields

      Kapit, Eliot; Tang, Zhijie; Han, Qi; Eley, Serena; Gong, Zhexuan (Colorado School of Mines. Arthur Lakes Library, 2021)
      This thesis investigates both analog Quantum Annealing and digital Quantum Annealing with oscillating transverse field in solving hard optimization problems. In the first part, we consider a range of unconventional modifications to Quantum Annealing (QA), applied to an artificial trial problem with continuously tunable difficulty. In this problem, inspired by “transverse field chaos” in larger systems, classical and quantum methods are steered toward a false local minimum. To go from this local minimum to the global minimum, all N spins must flip, making this problem exponentially difficult to solve. We numerically study this problem by using a variety of new methods from the literature: inhomogeneous driving, adding transverse couplers, and other types of coherent oscillations in the transverse field terms (collectively known as RFQA). We show that all these methods improve the scaling of the time to solution (relative to the standard uniform sweep evolution) in at least some regimes. Comparison of these methods could help identify promising paths towards a demonstrable quantum speedup over classical algorithms in solving some realistic problems with near-term quantum annealing hardware. In the second part of the thesis, we explore a digitized version of RFQA inspired by the performance of RFQA in analog quantum computing. The digitization of Quantum Annealing is a combination of analog Quantum Annealing(QA) and Quantum Approximate Optimization Algorithm (QAOA). Digitized-QA can be applied to full-scale, fault-tolerant quantum algorithms. We apply the digitized version of RFQA and QA to various trial problems using classical numerical simulation and show that digitized-RFQA is a potentially promising tool in solving hard optimization problems and can be a new tool to complement QAOA and traditional digitized-QA. In the third part of the thesis, for the preparation of an experiment, we investigate the 1D TFIM(transverse field ising model) and show that RFQA-D is able to accelerating the N-spin tunneling, the acceleration suggests that RFQA has the potential of mitigating the cost of minor embedding overhead.
    • Analysis of watershed parameters controlling turbidity following the West Fork Complex fire

      Singha, Kamini; Hall, Nicholas; Hogue, Terri S.; Ebel, Brian (Colorado School of Mines. Arthur Lakes Library, 2021)
      In this thesis, I explore watershed characteristics associated with increased turbidity following wildfires, with the goal of developing relations between turbidity and slope aspect, soil type, slope, and vegetation health from the 2013 West Fork Complex Fire (WFC) in southwest Colorado, USA. Turbidity, precipitation, and stream discharge were previously measured from May to September in 2015 and 2016 in seven watersheds, four burned and three unburned. I then characterized slope, slope aspect, soil type, vegetation, precipitation, and burn severity for each of the seven watersheds, as well as inside the burned areas of the four burned watersheds. I used turbidity as a proxy for sedimentation in each watershed as the dependent variable.Results indicate that from July to September of both 2015 and 2016, burned watersheds had larger spikes in turbidity following precipitation events than unburned watersheds. Higher burn severity and poor vegetation recovery were associated with the strongest positive correlations between total storm volume and turbidity responses. Enhanced Vegetation Index (EVI) was not consistently able to predict which watersheds would experience elevated turbidity following precipitation events; similarly, watershed slope and aspect alone were not able to predict which watersheds would experience larger turbidity responses to precipitation events. Despite seeing significant differences from July to September, during runoff from May to June, there were no significant differences in turbidity between burned and unburned watersheds. These results indicate that drivers of sedimentation in these burned watersheds, for example erosive soils, were more susceptible to precipitation than snowmelt due to exposure to rain splash and the higher intensity precipitation associated with convective rainfall. We also note that if summers continue to extend in the western U.S., burned watersheds with steep slopes, poor vegetation recovery, and erosive soils may experience longer periods of increased sedimentation. Results from this work provide insight on characteristics that determine vulnerability to sedimentation after a wildfire and can help guide land managers in developing effective mitigation strategies.
    • Applied mineral economics: an analysis on depletion, taxation and exploration

      Eggert, Roderick G.; Castillo, Emilio; Lange, Ian; Gilbert, Benjamin; Smith, Nicole M. (Colorado School of Mines. Arthur Lakes Library, 2021)
      This dissertation focuses on the empirical analysis of mineral economics and policy. The economic analysis and discussion around minerals mostly rely on theoretical models and normative principles. Empirical evidence is not as abundant as theoretical models, creating a gap in the discipline. This study contributes to the literature of mineral economics in three different areas that are relevant to support a data-driven policy debate - mineral depletion, taxation and the exploration of new deposits. First, mineral depletion broadly involves two diverging positions. These positions are reconciled through a methodology to assess future availability, which is applied to copper resources. Second, the expected impact of mineral royalties on exploration decisions has been addressed in theoretical models. However, an analysis of the impact of the Chilean profit-based royalties on early-stage mineral exploration indicates that the tax changes did not greatly affect average exploration decisions. Smaller companies, though, are more susceptible than major companies to changes in mineral taxation. Third, understanding what drives mineral discoveries is a major issue for future resource availability. The analysis on copper and gold discoveries indicates that increasing grassroots exploration budgets do not appear to be associated with more discoveries. Nevertheless, institutions do play a role in discoveries as more and larger discoveries are more likely to occur in countries with stronger protection of property rights.
    • Area-selective atomic layer deposition of dielectrics for semiconductor manufacturing

      Agarwal, Sumit; Xu, Wanxing; Ciobanu, Cristian V.; Wolden, Colin Andrew; Samaniuk, Joseph R.; Hausmann, Dennis (Colorado School of Mines. Arthur Lakes Library, 2021)
      As feature dimensions of state-of-the-art semiconductor devices approach the sub-5 nm node, device fabrication based on conventional top-down patterning technique is becoming ever more challenging. Area-selective atomic layer deposition (ALD) offers the advantage of eliminating lithographic patterning through selective growth, allowing for bottom-up fabrication of advanced nanopatterning of features with atomic-scale accuracy. In this work, we have primarily focused on developing area-selective ALD processes for dielectrics guided by optical diagnostic tools including in situ attenuated total reflection Fourier transform infrared spectroscopy and ellipsometry. We first tested a published claim that inherently area-selective ALD of ZrO2 on SiO2 versus Cu could be enabled using C2H5OH as the reducing agent for the Cu, the nongrowth surface, and as the oxygen source for ALD. Surprisingly, and contrary to the previous reports in the literature, no ZrO2 film growth was obtained using Zr[N(CH3)(C2H5)]4 and C2H5OH. However, we showed that using the H2O-C2H5OH mixture as the oxygen source led to ZrO2 film growth, but did not provide selective growth of ZrO2 on SiO2 versus Cu, which was again contrary to the literature. Also, so far, only a few claims of inherently area-selective ALD were published in the literature: this study shows that inherently area-selective ALD is not practical, suggesting that we need a more robust approach, such as site blocking. Secondly, in area-selective ALD based on site blocking, the first part of study focuses on understanding the site blocking ability of aminosilanes on SiO2 during area-selective ALD of ZrO2 and Al2O3. We developed strategies to effectively functionalize the SiO2 through the vapor phase, which was more compatible with the current semiconductor manufacturing than solution-based functionalization method. We showed that ALD of ZrO2 was blocked on aminosilane-functionalized SiO2 for only a few ALD cycles, and identified that the loss of selectivity was due to the strong coordination interaction between the Zr atom in Zr precursors and O atom in Si–O–Si present on aminosilane-functionalized SiO2. This strong interaction led to residual physisorbed molecules on the nongrowth surface, which contributed to the initiation of growth. Our focus then shifted to ALD of Al2O3, which was also technologically relevant. Using a lower reactivity, heteroleptic Al precursor, (CH3)2AlOCH(CH3)2 (DMAI), ALD of Al2O3 was delayed on vapor-functionalized SiO2 with small-molecule aminosilanes for up to ~30 ALD cycles, or ~3.5 nm with a selectivity of ~0.9, which was similar to solution-based functionalization. Furthermore, an effective approach was demonstrated to significantly extend the growth inhibition by lowering the DMAI precursor dose. In this part of study, we find that ideal ALD precursors, which are designed for high reactivity and high GPC, may not be ideal for area-selective ALD. This implies that we need precursor selection and design, such as replacing Al(CH3)3 with DMAI in this study. Finally, area-selective ALD of dielectric on chemically similar growth and nongrowth surfaces is much more challenging than that on chemically dissimilar growth and nongrowth surfaces. We demonstrated that ~2.7 nm of Al2O3 was selectively grown on plasma-deposited SiNx versus SiO2 at a selectivity of ~0.9 using the same functionalization method that was developed in the last part of this study. We showed that plasma-deposited SiNx surface was not fully saturated with aminosilanes, which provided a sufficient number of reactive surface sites for ALD of Al2O3. However, after exposure to ambient, a longer nucleation delay on SiNx surface during the ALD of Al2O3 was observed due to the higher surface coverage of aminosilanes, indicating the need to remove the oxynitride prior to functionalization with inhibitors. In this study, we find that fully passivating the nongrowth surface is necessary for achieving growth inhibition, but ALD can initiate on a partially passivated growth surface.
    • Assembly and propulsion of colloidal particles under combined electric and magnetic fields

      Wu, Ning; Haque, Md. Ashraful; Samaniuk, Joseph R.; Petruska, Andrew J. (Colorado School of Mines. Arthur Lakes Library, 2021)
      Colloids are reminiscent of atoms and molecules with larger dimensions. Under external fields, they tend to get organized in a hierarchical order that shares the same fundamental aspect of functional materials and living organisms. The primary goal of this thesis is to explore the combined impacts of electric and magnetic fields on colloidal assembly and actuation. We first report a strategy combining a planar magnetic field with a one-dimensional electric field to assemble and actuate linear chains made of paramagnetic microspheres. While a horizontal chain lying on the substrate is symmetric fore and aft and does not translate, a two-dimensional magnetic field can tilt the chain with an angle relative to the substrate. A superimposed alternating-current electric field leads to the propulsion of tilted chains along the substrate due to an unbalanced electrohydrodynamic flow. Using the magnetic field for steering and electric field for driving, we reveal a new propulsion mechanism that breaks the symmetry of hydrodynamic flow by manipulating the orientation of a microscopic object. We have also applied electric and magnetic fields orthogonally to magnetic microspheres, which allow us to independently control the magnitude and direction of dipolar attraction and repulsion between the particles. As a result, we obtain various kinds of well-aligned hierarchical structures based on the building blocks of microspheres and colloidal clusters of trimers, tetramers, heptamers, and nonamers. Our results demonstrate the potential in using combined fields to make diversified types of highly aligned structures for applications in high strength composites, optical materials, and battery electrodes.
    • Assessing ecohydrologic land management strategies and water supply impacts in the western U.S.

      Hogue, Terri S.; Kurzweil, Jake Robert; Smith, Steven M.; Singha, Kamini; Zhou, Wendy (Colorado School of Mines. Arthur Lakes Library, 2021)
      As the human population continues to grow and climate change progresses, stressors on ecohydrologic systems are exacerbated. Groundwater-dependent ecosystems such as freshwater springs are ecological hotspots supporting highly diverse habitats yet are understudied and, in most places, eradicated via human development. Our forested mountain systems and subsequent water supplies are also facing shifts in their structure, functionality, and succession. With the intensity and frequency of forest fires increasing and snowpack declining in the western United States, a common question has become how can we reduce forest fire risk while increasing watersheds efficiency at generating water supplies? The link between these two topics is a need for management through an ecohydrologic lens that provides a holistic approach to preserving our natural resources. Using Mt. Tamalpais, located in Marin County in northern California, we evaluated the effectiveness of an adapted springs ecosystem assessment protocol (A-SEAP) at characterizing springs geophysical properties and identifying aggregate areas of concern. The second portion of this dissertation evaluated the hydrologic response to forest fire mitigation in the Ashland watershed of southern Oregon. The third portion of this work included modeling scenarios of the hydrologic response to forest fire mitigation strategies in order to inform land managers on thresholds needed to increase surface runoff from vegetation removal. A-SEAP demonstrated that most springs in this region have fair ecohydrologic integrity relative to other spring studies, with a mean value of 3.62 out of 6. A-SEAP indicators identified springs located in Mt. Tamalpais State Park as lower ecohydrologic integrity than springs in the protected Marin Municipal Water District. A-SEAP indicators also successfully identified springs in need of restoration that fit the goals of the local land management agencies. In Ashland, despite cumulatively treating roughly 15% and 20% of the basins, these treatments only represent a decrease in canopy cover at the watershed scale of 3% and 4%, in the West and East sub-basins in the Ashland watershed, respectively. We found that in the post-treatment period, the West and East basins experienced an average annual water yield decline of 26% and 24%, respectively, with 66% (West) and 72% (East) of the changes in water yield attributed to annual variations in precipitation. Our modeling efforts in the Ashland basin demonstrated that treatments with at least 25% of the watershed treated with a basal area reduction exceeding 25% led to a significant increase in annual water yield over low, average, and above average water years. We also found that treatment intensity plays a larger role than area treated. A scenario with 25% of the watershed experiencing a 75% reduction in basal area had significantly higher annual water yields as compared to a treatment of 75% of the watershed with a 25% reduction in basal area. This dissertation provides methodologies and insights into how to best manage landscapes undergoing a change from both anthropogenic and natural stressors. This novel research presents a framework to develop holistic, ecohydrologic land management strategies that ensure climate resilient ecosystems and water supplies for our future generations.
    • Assessing the environmental, economic, and social sustainability of guar and guayule cultivated in the southwest United States

      Landis, Amy E.; Mealing, VeeAnder S.; McCray, John E.; Munakata Marr, Junko; Smith, Jessica, 1980- (Colorado School of Mines. Arthur Lakes Library, 2021)
      Since the mid 20th century, energy consumption has been the main driver of human-induced climate change, rapidly warming the planet’s atmosphere, oceans, and land. These trends have led to growing efforts to develop more sustainable feedstocks to meet increasing energy demands. Guar (Cyamopsis tetragonoloba) and guayule (Parthenium argentatum) are potential feedstocks for renewable fuels; these crops have highly valuable co-products that can enhance their economic viability. Guar and guayule are draught tolerant crops that can be cultivated in the Southwest US. Guar bagasse can be utilized to produce bioenergy, and its co-products include guar gum, a highly valuable chemical used in the food industry as an emulsifier and in the hydraulic fracturing industry as a friction reducer. Guayule bagasse can also be utilized to produce bioenergy, and its valuable co-products include natural rubber and resin, the latter of which can be used to make adhesives, coatings, antifungal agents, and fuels. To ensure holistic sustainability of these emerging agricultural feedstocks, an array of sustainability tools must be utilized in tandem. This thesis quantifies the three pillars of sustainability - environmental, economic, and social – for guar and guayule and assesses their potential to be integrated into the southwestern bio-economy. Life cycle analysis (LCA) was used to quantify the environmental impacts of guar and guayule, highlighting that irrigation is the main driver of agricultural impacts and thus provides the biggest opportunity for improved agricultural sustainability. Techno-economic analysis (TEA) quantifies the economic impact of guayule rubber production showing irrigation as the main agricultural contributor of economic impacts, while the largest processing contributor is capital/loans. This research is unique in its LCA and TEA approach, where data was collected from field trials and interviews of farmers and researchers were utilized to determine data quality and likelihood. This process produces more robust LCA and TEA results that are more representative of commercial farming practices. Existing social assessment tools were reviewed to determine the most appropriate approach for evaluating social sustainability. Ultimately, social sustainability was assessed by conducting focus groups of guar and guayule farmers and experts to identify social challenges and opportunities. This research aids in providing a clear path for the holistic sustainable development of a bio-economy in the Southwest US for guar and guayule.
    • Assessing the impact of water reuse on water quality in the Los Angeles River

      Hogue, Terri S.; Hennon, Victoria; Wolfand, Jordyn; Rust, Ashley J.; Navarre-Sitchler, Alexis K. (Colorado School of Mines. Arthur Lakes Library, 2021)
      Water reuse through water reclamation plants (WRPs) is an approach used by Los Angeles to reduce dependence on imported water. Three WRPs in the Los Angeles River (LAR) watershed, located in southern California, contribute up to 70% of the river discharge during dry weather periods. Increasing water reuse will lead to flow reductions in-stream, which could affect water quality. The purpose of this research is to identify the impact of water reuse on loads and concentrations of copper, zinc, lead, total suspended solids (TSS), and total dissolved solids (TDS) in the LAR. First, an empirical water quality data analysis was accomplished by collecting concentration data from monitoring programs and evaluating the relationships between pollutants and discharge. A water quality module was then added to a calibrated EPA-SWMM hydrologic model for water years 2011 to 2017 on a daily time-step, which was developed in prior work. Land use characteristics, event mean concentrations, urban baseflow concentrations, and WRP effluent data were added to the model. The model was then calibrated at four locations along the river mainstem. Water reuse scenarios were applied to the calibrated model and the simulated daily loads and concentrations were assessed using water quality regulatory compliance. Results show that TSS and metals are strongly correlated with each other and that discharge is a moderate predictor of pollutant concentration. Median daily loads for all pollutants generally decrease under scenarios of varying WRP flow reduction, but median daily concentrations either increase or decrease depending on the major sources for each pollutant. Copper, zinc, and lead regulatory compliance are not greatly impacted by WRP flow reduction, but TDS compliance is degraded. The results from this study will help inform regional water managers of potential impacts of flow reductions on water quality in the LAR and help aid decisions on water recycling policies.
    • Assessing the utilization of remote sensing and GIS techniques for flood studies and land use/land cover analysis through case studies in Nigeria and the USA

      Zhou, Wendy; Idowu, Dorcas; Düzgün, H. Sebnem; Santi, Paul M. (Paul Michael), 1964-; Singha, Kamini (Colorado School of Mines. Arthur Lakes Library, 2021)
      Globally, there has been a rise in geologic hazards such as flooding. A rise, which has often been attributed to climate change. First world countries especially the United States of America have well-structured ground and space-based flood monitoring systems through which data are obtained to provide real-time flood prediction and warnings to stakeholders. However, developing regions of the world mostly suffer the devastating effect of flooding due to a lack of adequate flood monitoring systems to provide accurate flood warnings. Land-use changes associated with an increase in impervious surface resulting from vegetation loss and/or replacement of flood plains and wetland with pavements are known to increase flood intensity. Hence, knowing how land cover changes affect floods in an area is therefore crucial to mitigating it and so is knowing the flood hazard level for these areas. In this study, we evaluated the effectiveness of the Gravity Recovery and Climate Experiment data (GRACE) storage-based Flood Potential Index (FPI) at correctly predicting floods in Nigeria with a focus on its efficacy at predicting past floods in the country. A newly derived Water Budget-based FPI was assessed and compared to the GRACE-FPI in terms of its capability to predict floods in the Mississippi River basin in the USA. Finally, the influence of changes in LULC on flooding was assessed for Lagos State using satellite datasets.
    • Atomistic simulations of polarization switching in ferroelectric materials

      Asle Zaeem, Mohsen; Nobarani, Hamed; Brennecka, Geoffrey; Tucker, Garritt J.; Ciobanu, Cristian V. (Colorado School of Mines. Arthur Lakes Library, 2021)
      Domain polarization switching in ferroelectric materials plays an essential role in determining their properties and performance for applications in capacitors, transducers, sensors, actuators, and memory devices. Understanding the intrinsic behavior of ferroelectrics during domain polarization switching caused by an electric field is challenging due to the very fast switching process of domains and the nano/micro scale size of domains. In this Ph.D. research, an atomistic-scale computational framework is developed to study the domain switching process of bismuth ferrite (BiFeO3) samples in nanoscale. An in-house second nearest-neighbor modified embedded atom method plus charge equilibration interatomic potential for BiFeO3 is used for the atomistic simulations. We performed a series of molecular dynamics simulations to study the effects of magnitude and duration of an applied electric field, domain size and specimen thickness on the domain polarization switching. The switching map with respect to the magnitude and duration of the electric field was obtained for BiFeO3 sample possessing 180° domain walls, showing that 180° switching occurred when a relatively small electric field was applied for a long enough time; on the other hand, a relatively large electric field was needed when the duration of the applied electric field was short. Also, we studied the sample size effect, and we did not observe a significant change in the switching process when the surface size increased, however, by increasing the sample thickness, the time to reach the complete 180° switching increased. In addition, we studied the effects of defects such as vacancies and dislocations, as well as applied tensile strain on the switching process. Oxygen vacancies and Schottky defects led to a localized reduction in the polarization. Overall, dislocations deteriorated the expected increase in the polarization magnitude with the electric field. Unlike the monodomain with almost the same polarization, the polarization calculated for a structure with dislocations was nonhomogeneous. Large localized polarization was observed when the stress field between the dislocation cores exceeded the MD-calculated threshold of 4 GPa. Moreover, the hysteresis loop for BiFeO3 sample was obtained, containing essential information about the ferroelectric material, such as remanent polarization and saturation points. This work demonstrated the capabilities of atomistic simulations for better understanding the effects of different factors on the switching behaviors of ferroelectric materials, which can contribute to the design of ferroelectrics with enhanced or controllable switching performance.
    • Batteries and thermal energy storage: an analytical method for sizing and analyzing potential synergies in hybrid systems

      Tabares-Velasco, Paulo Cesar; Woods, Jason; Brandt, Matthew; Braun, Robert J.; Porter, Jason M. (Colorado School of Mines. Arthur Lakes Library, 2021)
      Electric-grid generation capacity is sized to meet the peak load, typically in the summer when building cooling loads are high and is underutilized for much of the year. Electric utilities often pass on the costs associated with these peaking generators to building owners through demand charges. Building owners can minimize these demand charges by shifting energy use away from peak periods with behind the meter storage. This storage can include batteries, which can directly shift the metered load, or thermal energy storage, which can shift thermal-driven electric loads like air conditioning. However, there is a lack of research on how best to combine battery and thermal energy storage. In this study, an analytical sizing method is developed for a hybrid system, calculating the potential demand reduction and annualized cost savings for different combinations of thermal and battery energy storage sizes. It is shown that adding batteries to a thermal energy storage system increases the total system’s load shaving potential. This is particularly true when the building has onsite PV generation or electric vehicle charging, which add significant variability to the load shape. It is also shown that for a given total storage size, selecting a higher fraction of thermal energy storage can significantly lower the cycling of the battery, and therefore extend the battery life. This, combined with the expected lower first cost of thermal energy storage materials compared to batteries, shows that hybrid energy storage systems can outperform a standalone battery or standalone thermal storage system.