Engineering topics
Doan, Hieu A.
Publications and source records attributed to Doan, Hieu A..
Acidity-Governed Rules in the Electrochemical Performance of Fluorinated Benzenes for High-Voltage Lithium Metal Batteries
Judicious selection of the optimal fluorobenzene (FB) as a nonsolvating cosolvent for lithium metal batteries (LMBs) is reported. For this work, we found the key correlation between FB structures and cycling stabilities of cells: increased fluorine substitution of FBs results in higher anodic stability but at the expense of reduced reductive stability, and FBs containing three or more fluorine atoms exhibit insufficient anodic stability in the electrolyte system comprised of fluoroethylene carbonate (FEC) and ethyl methyl carbonate (EMC). More importantly, FBs with higher acidity (lower pK a ) due to protons located between two adjacent fluorine atoms tend to be more susceptible to side reactions during cycling. Our results indicate that difluorobenzenes with no “acidic” proton (DFB2 and DFB4) have emerged as the optimal choice with the desired redox stability in high-voltage LMBs. Nuclear magnetic resonance and X-ray photoelectron spectroscopy confirmed these findings, providing guidance for selecting the most suitable FB variants as nonsolvating cosolvents for high-voltage LMBs.
An integrated high-throughput robotic platform and active learning approach for accelerated discovery of optimal electrolyte formulations
Solubility of redox-active molecules is an important determining factor of the energy density in redox flow batteries. However, the advancement of electrolyte materials discovery has been constrained by the absence of extensive experimental solubility datasets, which are crucial for leveraging data-driven methodologies. In this study, we design and investigate a highly automated workflow that synergizes a high-throughput experimentation platform with a state-of-the-art active learning algorithm to significantly enhance the solubility of redox-active molecules in organic solvents. Our platform identifies multiple solvents that achieve a remarkable solubility threshold exceeding 6.20 M for the archetype redox-active molecule, 2,1,3-benzothiadiazole, from a comprehensive library of more than 2000 potential solvents. Significantly, our integrated strategy necessitates solubility assessments for fewer than 10% of these candidates, underscoring the efficiency of our approach. Our results also show that binary solvent mixtures, particularly those incorporating 1,4-dioxane, are instrumental in boosting the solubility of 2,1,3-benzothiadiazole. Beyond designing an efficient workflow for developing high-performance redox flow batteries, our machine learning-guided high-throughput robotic platform presents a robust and general approach for expedited discovery of functional materials.
Active Learning Guided Computational Discovery of Plant-Based Redoxmers for Organic Nonaqueous Redox Flow Batteries
Organic nonaqueous redox flow batteries (ONRFBs) are promising energy storage devices due to their scalability and reliance on sourceable materials. However, finding suitable redox-active organic molecules (redoxmers) for these batteries remains a challenge. Using plant-based compounds as precursors for these redoxmers can decrease their costs and environmental toxicity. In this computational study, flavonoid molecules have been examined as potential redoxmers for ONRFBs. Flavone and isoflavone derivatives were selected as catholyte (positive charge carrier) and anolyte (negative charge carrier) molecules, respectively. To drive their redox potentials to the opposite extremes, in silico derivatization was performed using a novel algorithm to generate a library of > 40000 candidate molecules that penalizes overly complex structures. A multiobjective Bayesian optimization based active learning algorithm was then used to identify best redoxmer candidates in these search spaces. Furthermore, our study provides methodologies for molecular design and optimization of natural scaffolds and highlights the need of incorporating expert chemistry awareness of the natural products and the basic rules of synthetic chemistry in machine learning.
Trimer Quinoxalines as Organic Cathode Materials for Lithium-Ion Batteries
Due to synthetic ease, high redox potentials, low solubility in polar electrolytes, and good electric conductivity of their semiconducting crystals, discotic quinoxaline trimers (3Q) have been considered as possible candidates for 4 V organic cathodes in lithium-ion batteries. To assess their feasibility as such materials, several 3Q derivatives have been synthesized and tested in half-cells. In voltage limited cycling tests at 1.2–3.9 V vs Li/Li + , the specific discharge capacities of 40–180 mAh g -1 were obtained at a rate of 1 C, and multiple lithiation of 3Q and its derivatives was observed during discharge. However, the obtained discharge capacity was only a fraction of the theoretical capacity expected for reversible six-electron reduction; there was also rapid capacity fade. Our spectroscopic studies indicate the reversible three-electron lithiation at 2 V vs Li/Li + , but suggest instability of more highly discharged states. Finally, our conclusion is that while the 3Q derivatives combine several traits that are desirable in an organic cathode material (including negligible solubility, capacity for multiple charging, and near-100% coulombic efficiency), these materials are still impractical to use.
In silico discovery of a new class of anolyte redoxmers for non-aqueous redox flow batteries
A novel simplicity-biased molecular structure generation method and Bayesian optimization based active learning have been used to discover a new class of heterocyclic molecules for organic non-aqueous redox flow batteries.
Computational Screening of Supported Metal Oxide Nanoclusters for Methane Activation: Insights into Homolytic versus Heterolytic C–H Bond Dissociation
Since its discovery in zeolites, the [CuOCu] 2+ motif has played an important role in our understanding of selective methane activation over supported metal oxide nanoclusters. Although there are two known C-H bond dissoci-ation mechanisms, namely homolytic and heterolytic cleavage, most computational studies on optimizing metal oxide nanoclusters for improved methane activation reactivity have focused only on the homolytic mechanism. In this work, both mechanisms were examined for a set of 21 mixed metal oxide complexes of the form of [M 1 OM 2 ] 2+ (M 1 , M 2 = Mn, Fe, Co, Ni, Cu, Zn). Except for pure copper, heterolytic cleavage was found to be the dominant C-H bond activation pathway for all systems. Furthermore, mixed systems including [CuOMn] 2+ , [CuONi] 2+ , and [CuOZn] 2+ are predicted to possess similar methane activation activity as pure [CuOCu] 2+ . Furthermore, these results suggest that both homolytic and heterolytic mechanisms should be considered in computing methane activation energies on supported metal oxide nanoclusters.
Accelerating the evaluation of crucial descriptors for catalyst screening via message passing neural network
Graph neural networks developed for adsorption energy prediction on molybdenum carbide catalysts provide a significant acceleration over density functional theory calculations.
Mesolytic cleavage of homobenzylic ethers for programmable end-of-life function in redoxmers
Irreparable chemical damage to redox-active monomers, oligomers, and polymers (i.e. redoxmers) limits the lifetime of energy storage devices (e.g. redox flow batteries) by causing electrode and membrane fouling, as well as irreversible capacity loss. To predictably restore device and materials performance, it is desirable to add programmed destruction capabilities into these damage-prone materials. Here, we report the use of triggerable retrograde reactions to achieve programmable end-of-life function in redoxmers. Retrograde reactions are intended to break up damaged redoxmer materials into more soluble small-molecule constituents to prevent irreparably fouling battery interfaces. We investigated the redox-triggered mesolytic cleavage of homobenzylic ethers (HBEs) for this purpose. Combining experimental methods with simulations, we probed the influence of ring substituents on the programmable bond-scission behavior of HBEs. Variation of para-substituents on the HBEs allowed us to modify the oxidation potential, degradation pathway, and electrochemical mechanisms of the generated products. Given its ideal oxidation potential, we selected the para-methoxy-based HBE as the cleavage scaffold and integrated it with dialkoxybenzene, TEMPO, viologen, and para-nitrobenzene redox centers for compatibility, bulk deconstruction, and electrode defouling tests. Upon applying a high oxidation potential, the redox-active pendant is cleaved from the backbone via HBE bonds; of the four redox centers tested, three of them remained fully redox active. Viologen-appended redox active polymers with HBE linkers were cleaved to fully separate the redox centers from the backbone. Polymer-filmed electrodes were partially restored following HBE oxidation defouling. Here these studies highlight the co-design of new materials functions for developing sustainable energy storage materials.
Experimental Protocols for Studying Organic Non-aqueous Redox Flow Batteries
We report that Redox flow batteries (RFBs) are promising devices for grid-scale energy storage due to the decoupling of power and energy, which can be independently scaled by the electrode area and storage tank size, respectively. To date, only aqueous RFBs, such as the vanadium RFB, have been implemented commercially. Nevertheless, the limited energy densities and high-cost materials may preclude their wider market penetration. Organic non-aqueous redox-flow batteries (O-NRFBs), which utilize redox-active organic molecules (ROMs), have been offered as an attractive alternative. Many possible advantages include the use of earth-abundant elements (C, H, N, O, S, F), wherein the ROMs can be prepared from low-cost and sustainable materials. Additionally, a large variety of electroactive moieties are accessible as building blocks, providing a synthetic platform to tune the properties of ROMs through rational design. As a consequence, the development of novel ROMs has attracted researchers with diverse backgrounds, prompting remarkable innovations in the past decade. However, consistency in experimental protocols (e.g., electrochemical methods, cycling stability, experimental conditions) has not coincided with this uptick in research, leading to sometimes ambiguous and incomparable results. This is further convoluted by the complexity innate to battery development, such as cell design, detection, and characterization of reactions and active components. The O-NRFB application imposes stringent requirements on the physical or/and chemical processes involved in electrochemical cycling. Yet such considerations are often overlooked. Further, while quantum mechanical calculations provide a convenient tool for evaluating and selecting ROM candidates, this simulation is not always performed. Thus, in this Energy Focus, we detail a means to standardize experimental protocols for studies of O-NRFBs and suggest practices to facilitate fundamental understanding and development of ROMs.
Discovery of Energy Storage Molecular Materials Using Quantum Chemistry-Guided Multiobjective Bayesian Optimization
Redox flow batteries (RFBs) are a promising technology for stationary energy storage applications due to their flexible design, scalability, and low cost. In RFBs, energy is carried in flowable redoxactive materials (redoxmers) which are stored externally and pumped to the cell during operation. Further improvements in the energy density of RFBs necessitates redoxmer designs with wider redox potential windows and higher solubility. Additionally, designing redoxmers with a fluorescence-enabled self-reporting functionality allows monitoring of the state of health of RFBs. To accelerate the discovery of redoxmers with desired properties, state-of-the- art machine learning ( ML) methods, such as multiobjective Bayesian optimization (MBO), are useful. Here, we first employed density functional theory calculations to generate a database of reduction potentials, solvation free energies, and absorption wavelengths for 1400 redoxmer molecules based on a 2,1,3-benzothiadiazole (BzNSN) core structure. From the computed properties, we identified 22 Pareto-optimal molecules that represent best trade-off among all of the desired properties. We further utilized these data to develop and benchmark an MBO approach to identify candidates quickly and efficiently with multiple targeted properties. With MBO, optimal candidates from the 1400-molecule data set can be identified at least 15 times more efficiently compared to the brute force or random selection approach. Importantly, we utilized this approach for discovering promising redoxmers from an unseen database of 1 million BzNSN-based molecules, where we discovered 16 new Pareto-optimal molecules with significant improvements in properties over the initial 1400 molecules. Furthermore, we anticipate that this active learning technique is general and can be utilized for the discovery of any class of functional materials that satisfies multiple desired property criteria.
TEMPO allegro: liquid catholyte redoxmers for nonaqueous redox flow batteries
Redoxmers are organic active molecules storing energy in redox flow batteries (RFBs). Liquid redoxmers represent an extreme scenario where maximum concentration may be achieved by minimizing supporting solvents, thus maximizing the energy density of RFBs. Herein, a series of (2,2,6,6-tetramethylpiperidin-1-yl)oxyl (TEMPO)-based high potential (catholyte) liquid redoxmers, TEMPO-EG1, TEMPO-EG2, and TEMPO-EG3, were developed by incorporating polyethylene glycol (PEG) chains. Such modifications not only afford dramatic physical changes from solid to liquid and full miscibility in acetonitrile, but also impact the redox behavior. DFT calculations indicate that the incorporated PEG chains impact the charge distribution, which may account for the electrochemical changes. Importantly, compared to our previous liquid catholytes, the new redoxmers exhibit lower viscosity, which is desired for enhancing high concentration cycling performance. Furthermore, by using a hybrid flow cell, TEMPO-EG1 demonstrated more than 70% capacity retention over 100 cycles at 0.1 M and 66% capacity retention at 0.5 M, affording excellent cyclability at various concentrations. The study exemplifies how molecular engineering tuned the rheological properties of redoxmers, such as viscosity, to improve the high concentration cycling performance of RFBs, which may represent a promising avenue for a high energy density and low-cost flow battery system.
Identification of Active Metal Carbide and Nitride Catalytic Facets for Hydrodeoxygenation Reactions
The catalytic hydrodeoxygenation (HDO) reaction is of considerable interest for biomass conversion to valuable chemicals and fuels, where one of the critical bottlenecks is the lack of cost-effective and efficient catalysts. To discover cost-efficient catalysts for the HDO reaction, we employed a density functional theory-based hierarchical catalyst design strategy based on catalytic descriptors, reaction energy profiles, and microkinetic modeling (MKM). We focused on the carbide and nitride catalyst space, for which we calculated 121 catalyst surfaces of Mo 2 C, MoC, Mo 2 N, W 2 C, NbC, VC, VN, and NbN catalysts. Based on the computed surface energies, reaction energies of oxygen removal, carbon binding strength, and the surface area of nanoparticles, the likely active facets are the Mo 2 C(111), MoC(011), VN(100), Mo 2 N(001), Mo 2 N(011), and Mo 2 N(100) surfaces. Further, detailed energy profiles were obtained, and MKM was performed for a model reaction (glycolaldehyde + 2H 2 . ethylene + 2H 2 O) on the Mo 2 C(111), VN(100), and MoC(100) surfaces. Based on the computed volcano map obtained from MKM, the predicted active facets for this HDO reaction are the Mo 2 C(111), MoC(011), VN(011), Mo 2 N(001), Mo 2 N(011), and Mo 2 N(100) surfaces. Additionally, none of the carbide and nitride catalyst surfaces are located in the optimal catalytic activity part. Therefore, it is essential to modify the catalyst via adding dopants or alloying to improve the catalytic activity. Catalytic modifications that can destabilize the surface adsorption of O*/H 2 O* and decrease the energy barriers of O-H bond formation are recommended to facilitate the HDO on the carbide and nitride catalysts. These a priori investigations provide guidelines for future low-cost HDO catalyst development.
Molecular Structure and Electron Affinity of Metal-Solvent Complexes: Insights from Density Functional Theory Simulations
A molecular level understanding of the structure and energetics of the monovalent and divalent metal ion complexes is of great importance for development of next-generation batteries. Here, Density Functional Theory (DFT) simulations at the ωb97xD/6-31 + G(d,p) level of theory are performed to investigate the interaction of metal ions (Li + , Na + , K + , Mg 2+ , Ca 2+ , Zn 2+ ) with 26 organic solvent molecules. The reduction energetics (electron affinity and reduction potential) and structural responses of the solvent molecules and the molecular complexes are discussed. The DFT calculations are carried out to investigate the structure, energetics and electron affinities of chelated complexes of water (H 2 O), tetrahydrofuran (THF) and di-methoxy ethane (DME) solvent molecules. Additionally, ab initio dynamic simulations (AIMD) at 298 K using atom centered density matrix propagation (ADMP) formalism are performed to understand the spontaneous structure formation upon electron attachment of the metal ion-solvent complexes. The ADMP simulations indicate the decomposition of Mg + -(DME) 3 complex via cleavage of C–O bond of one of the three DME molecules indicating irreversible decomposition of DME in the presence of the Mg + radical. We believe that the data collected as part of this investigation serves as a library of fundamental knowledge towards a deeper understanding of the electrode-electrolyte interfacial reactions.
Quantum Chemistry-Informed Active Learning to Accelerate the Design and Discovery of Sustainable Energy Storage Materials
Here we employed Density Functional Theory (DFT) to compute oxidation potentials of 1,400 homobenzylic ether molecules to search for the ideal sustainable redoxmer design. The generated data were used to construct an active learning model based on Bayesian optimization (BO) that targets candidates with desired oxidation potentials utilizing only a minimal number of DFT calculations. The active learning model demonstrated not only significant efficiency improvement over the random selection approach but also robust capability in identifying desired candidates in an untested set of 112,000 homobenzylic ether molecules. Our findings highlight the efficacy of quantum chemistry-informed active learning to accelerate the discovery of materials with desired properties from a vast chemical space.
Effect of Ni Dopant on Furan Activation over Mo 2 C Surface: Insights from First-Principles-Based Microkinetic Modeling
Here, the role of Ni dopant in a molybdenum carbide (Mo 2 C) catalyst for butadiene and propyne production from furan is investigated by using a Mo 2 C(001) surface model and first-principles-based microkinetic modeling (MKM). A reaction mechanism including 18 reaction steps is generated to compute the turnover frequency (TOF) of butadiene and propyne from furan gas. The butadiene production is dominant with the TOF being 3-6 orders of magnitude higher than the TOF of propyne on a bare Mo 2 C surface due to an active C-O bond scission reaction compared to C-C bond cleavage reaction. The exothermic binding of ring-opened furan (CHCHCHCHO*) inhibits the activity of a bare Mo 2 C surface for butadiene and propyne production. Calculations predict that Mo 2 C surface modification with a Ni dopant energetically destabilizes the adsorption of CHCHCHCHO*, which facilitates the formation of butadiene and propyne. In the presence of Ni dopants on a Mo 2 C(001) surface, the selectivity toward the butadiene formation is decreased compared to propyne due to relatively favorable C-C bond scission as opposed to the C-O bond scission. This first-principles-based study provides mechanistic and kinetic insights into the role of dopants in cost-efficient Mo 2 C catalysts during the vapor phase conversion of furan to gaseous products.