Time-Dependent Quantum Chemistry: From Structure to Dynamics.
The advent of Attosecond Chemistry with all its potential for new developments in spectroscopic and dynamic molecular imaging techniques requires powerful computational tools for the simulations of time-resolved quantum molecular dynamics.
In this lecture, we will review important concepts of time-independent Quantum Chemistry that have survived the passage to these time-dependent problems. We will describe adaptations that are required of these concepts to fit the context of Quantum Dynamics. This review will focus in particular on techniques adapted from multi configuration approaches of Electronic Structure Theory to yield powerful electronic and/or nuclear wavepacket propagation methods. Illustrations include recent contributions to delineate the marks of electron correlation in intense field ionization, as manifested in its scalar and vectorial observables.
Computational peculiarities of the time-dependent approaches will be highlighted and discussed in relation to more general challenges in quantum and computational chemistry.
Design of energy materials through defects and doping
Functional materials such as those for energy applications are often structurally and chemically complex, where defects and impurities can be vital or fatal to the materials’ performance. A detailed understanding of defect physics in these materials is thus required for explaining, predicting, and optimizing their properties, and for designing materials with improved performance. With recent advances in electronic-structure methods, first-principles calculations for defects and impurities have become an important tool in providing such an understanding.
In this lecture, I focus my discussion on defect physics vis-à-vis functional properties in energy materials. Specific examples will be taken from recent work on complex oxides for battery electrodes.
Through these examples, I will illustrate how state-of-the-art defect calculations can serve as a study of materials’ response to interventions–done on purpose and in a systematic and well-controlled manner–at the electronic and atomic level, and how such a study can provide a fundamental understanding of the materials and guidelines for rational materials design.
Computational guided design of metal oxide catalysts in biomass conversion
Institute of High-Performance Computing (IHPC), A*STAR (Agency for Science, Technology and Research), 1 Fusionopolis Way #16-16 Connexis, 138632 Singapore
There are strong incentives to produce chemicals and fuels from renewable biomass resources to reduce the reliance on diminishing fossil resources . Lignocellulosic biomass is expected to be the most promising renewable feedstock, owing to its wide abundance, high energy content, and sustainability. Heterogeneous catalysts, in particularly transition metal oxides (TMO), have proven to be active catalysts for the selective conversion of cellulose/glucose. In this presentation, taking CuO as a model TMO, three representative applications of metal oxide catalysts in biomass conversion are presented, including 1) the selective oxidation of cellulose to gluconic acid , 2) the oxidation of glycerol to dicarboxylic acids  and 3) unconventional selective sono-oxidation of cellulose to glucuronic acid . An integrated experimental and theoretical approach is used in those studies to demonstrate the unique activity of surface lattice oxygen on the TMO surface in activating the C-H bonds and acting as oxidize agents in oxidation reactions. The detailed reaction energy profile analysis for the conversion of different biomass substrates on TMO provides insightful understand on the activity of TMO and opens new avenues for designing efficient catalyst in this field.
Reference: Q. T. Trinh, B. K. Chethana, S. H. Mushrif, The Journal of Physical Chemistry C 2015, 119, 17137-17145.  P. N. Amaniampong, Q. T. Trinh, B. Wang, A. Borgna, Y. Yang, S. H. Mushrif, Angewandte Chemie International Edition 2015, 54, 8928-8933.  P. N. Amaniampong, Q. T. Trinh, J. J. Varghese, R. Behling, S. Valange, S. H. Mushrif, F. Jérôme, Green Chemistry 2018, 20, 2730-2741.  P. N. Amaniampong, Q. T. Trinh, K. De Oliveira Vigier, D. Q. Dao, N. H. Tran, Y. Wang, M. P. Sherburne, F. Jérôme, Journal of the American Chemical Society 2019, 141, 14772-14779.
Theoretical Models for Interpreting the Luminescence of Silver Clusters Embedded in Zeolites
Ngo Tuan Cuong
Center for Computational Science and Faculty of Chemistry
Hanoi National University of Education, Cau Giay, Ha Noi, Vietnam
We present in this talk our theoretical models with the aim to elucidate a number of recent spectroscopic observations, in particular the structure-to-function relationship in the fully Ag-exchanged luminescent FAUX and FAUY zeolites featuring different Si/Al ratio and Ag content. Our experimental collaborators have used their X-ray-excited optical luminescence (XEOL) to directly detect the extended X-ray absorption fine structure (EXAFS) and unravel the detailed structures of the emissive few-atom silver clusters (Agn). Time-resolved spectroscopy emphasized the nature of the main emission at the µs lifetimes observed in each topology of the Ag cluster isomers. Using DFT computations we propose some simple but unified models to interpret these phenomena.
Representation of hidden chemistry with machine learning
1 Faculty of Computer Science, Phenikaa University, Yen Nghia, Ha Dong Dist., Hanoi
2 Phenikaa Institute for Advanced Study (PIAS), Phenikaa University, Yen Nghia, Ha Dong Dist., Hanoi
Human beings have always paid significant attention to learning nature’s “game” by observing natural phenomena (data) and making models (theory) to predict future outcomes. In this respect, we have observed the vast diversity of nature and unified different natural phenomena in a small set of fundamental variables or laws. This consideration of science is strongly related to the field of data mining, which is developed to discover hidden knowledge and build predictive models. Recently, the increasing volume of available experimental and quantum computational materials databases, together with the development of machine learning techniques, especially deep neural networks, has provided new opportunities for developing techniques that help researchers accelerate the discovery and comprehension of new materials. By using machine learning algorithms, hidden information on materials, including patterns, features, chemical rules, and physical laws, can be automatically discovered from both first-principles-calculated data and experimental data. In this talk, we present our recent studies on applying deep neural networks on solid materials: (1) the representation of potential energy surfaces; (2) material stability assessment; (3) local structure formability and similarity measurement; (4) material prediction. We focus on magnetic materials based on bimetal alloys of lanthanide metal and transition-metal (LAT) and LAT alloys including a light element X, which may be B, C, N, or O (LATX). Our studies demonstrate that machine learning models can well represent total energy, formation energy, and thermodynamical stability in comparison with DFT calculations. We also show that machine learning models, especially deep learning models, can be trained to recognize materials or local structures which are synthesizable.
- Tien-Lam Pham, Hiori Kino, Kiyoyuki Terakura, Takashi Miyake, Koji Tsuda, Ichigaku Takigawa, Hieu-Chi Dam “Machine learning reveals orbital interaction in materials”, (2017), Science and technology of advanced materials 18 (1), 756-765.
- Tien-Lam Pham, Nguyen-Duong Nguyen, Van-Doan Nguyen, Hiori Kino, Takashi Miyake, Hieu-Chi Dam, “Learning structure-property relationship in crystalline materials: A study of lanthanide–transition metal alloys”, (2018), The Journal of chemical physics 148 (20), 204106.
- Tien-Lam Pham, Duong-Nguyen Nguyen, Minh-Quyet Ha, Hiori Kino, Takashi Miyatake, Hieu-Chi Dam, “Explainable machine learning for materials discovery: predicting the potentially formable Nd–Fe–B crystal structures and extracting the structure–stability relationship”, (2020) IUCrJ.
- Tien-Cuong Nguyen, Van-Quyen Nguyen, Van-Linh Ngo, Quang-Khoat Than, Tien-Lam Pham, “Learning Hidden Chemistry with Deep Neural Networks”, (2021) Com. Mat. Sci., 2000, 110784.
- Van-Quyen Nguyen, Viet-Cuong Nguyen, Tien-Cuong Nguyen, Tien-Lam Pham, Pairwise interactions for Potential energy surfaces and Atomic forces with Deep Neural network, (2021), arXiv:2111.05603
Computation-Experiment Correlation in the Inhibitability of Plant-based Compounds and Protein-related Structures
Bùi Quang Thành,1 Phan Tứ Quý,2 and Nguyễn Thị Ái Nhung1
1Department of Chemistry, University of Sciences, Hue University, Hue, Vietnam
2Department of Natural Sciences & Technology, Tay Nguyen University, Buon Ma Thuot, Vietnam
In pharmaceutical development, conventional pre-clinical research including drug design and laboratory studies, often lasts many years in order to arrive at noticeable candidates for promising treatment. Such a drawback puts heavy pressure on the efforts of workers in the field to tackle highly contagious and mutable diseases. An efficient way of designing drugs is the harnessing of computer-based power. In particular, a variety of computational approaches are available for use in the field of drug design, such as DFT calculations for structural verification, docking simulations for inhibitability, QSARIS analysis for prediction of physicochemical properties, ADMET for regression of pharmacokinetics and pharmacology, etc.
On the other hand, corresponding experiments can also be valuable for the design of new compounds, including various spectroscopic methods for chemical determination, bioassays for antibiotic activity, etc. If a solid computation-experiment correlation could be established to a certain degree of reliability, the knowledge jointly acquired could be considered as a significant contribution to the search for new compounds and/or drugs.
In this presentation, we would briefly review our attempts to look for evidence of the theory-experiment correlation, which is primarily focused on the plant-based compounds and protein-related structures.
Keywords: Computation-experiment correlation; DFT; Docking, QSARIS; ADMET
Surface-Enhanced Raman Spectroscopy of Pesticide Chlorpyrifos Adsorbed on Silver Nanoparticles: Experimental and Theoretical Studies
Thi Chinh Ngo, Ph.D.*
Duy Tan Computational Chemistry Lab (DTC2), Institute of Research and Development,
Duy Tan University, Da Nang, Viet Nam
Surface-enhanced Raman spectroscopy (SERS) in which the Raman intensity of a molecule can be measured on nanostructured metallic surfaces (i.e., silver, gold, and copper), is of great use for the detection of persistent contamination in, among others, agricultural products. SERS experiments and quantum chemical calculations on the interactions of chlorpyrifos (CPF) which is an intensively used pesticide, with a roughed silver nanoparticle surface were thoroughly investigated to study the inherent chemical mechanism. Ligand–cluster interaction geometries show that the CPF molecule is mainly adsorbed on the silver surface via both S atom and pyridine ring involving a covalent Ag···S coordination as well as van der Waals physisorption. Raman vibrational modes of CPF are centered at 474, 632, 678, 1277 and 1551 cm–1 characterizing the P-O-C bending, P=S stretching, Cl-ring mode and pyridine ring stretching, respectively, which are all enhanced when CPF is adsorbed on a silver surface.
The concentration-dependent effect of CPF on silver substrates has been reproduced for the first time by coordinating 2 and 3 CPF molecules on an Ag20 silver cluster model simulated by DFT computations. Intensities of characteristic peaks of CPF as shown in calculated SERS spectra are increased by 2 and 3 times with respect to those of the CPF–Ag20 complex which indicates positive influence of high analyst concentration on SERS signal.
This observation can be explained by the electron donating effect of CPF upon adsorption. The latter donates electron from its lone pair on S and Cl atoms, p electron on S=P bond to silver atoms on surface, and then the positive charge of silver surface is displaced to the CPF moiety via Ag···S and Ag···Cl contacts. Information obtained from the adsorption of CPF on silver by SERS is helpful to understand the mechanism of adsorption process involving chlorpyrifos ligand coordinated on silver nanoparticle surfaces. It also contributes to design field detection methods for rapid screening and monitoring of pesticides in environment or agricultural products by using portable detection systems such as paper-based or fiber-based SERS sensors.
Omicron Variant Binds to Human Cells More Strongly than SARS-CoV-2 Wild Type
Hoang Linh Nguyen1,2,3,*, Nguyen Quoc Thai1,4, and Mai Suan Li5,*
1Life Science Lab, Institute for Computational Science and Technology, Quang Trung, Software City, Tan Chanh Hiep Ward, District 12, Ho Chi Minh City, Vietnam
2Ho Chi Minh City University of Technology (HCMUT), Ho Chi Minh City 700000, Vietnam
3Vietnam National University, Ho Chi Minh City 700000, Vietnam
4Dong Thap University, 783 Pham Huu Lau Street, Ward 6, Cao Lanh City, Dong Thap, Vietnam
5Institute of Physics, Polish Academy of Sciences, al. Lotnikow 32/46, 02-668, Warsaw, Poland
The emergence of the variant Omicron (B.1.1.529) of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) aggravates the covid-19 pandemic due to its very contagious ability. The high infection rate may be due to the high binding affinity of Omicron to human cells, but both experimental and computational studies have yielded conflicting results on this issue.
Some studies showed that the Omicron variant binds to human angiotensin-converting enzyme 2 (hACE2) more strongly than wild type (WT), but other studies reported comparable binding affinities. To shed light on this crucial problem, we calculate the binding free energy of the receptor binding domain (RBD) of the WT and Omicron spike protein to hACE2 using all-atom molecular dynamics simulation and molecular mechanics Poisson-Boltzmann surface area (MM-PBSA) method.
We show that Omicron binds to human cells more strongly than WT due to increased RBD charge, which enhances electrostatic interaction with negatively charged hACE2. N440K, T478K, E484A, Q493R and Q498R mutations in RBD have been found to play a critical role in the stability of the RBD-hACE2 complex. The effect of homogeneous and heterogeneous models of glycans coating the viral RBD and the peptidyl domain (PD) of hACE2 is examined. Although the total binding free energy is not sensitive to the glycan model, the distribution of per-residue interaction energies depends on it.
Development of electronic structure methods towards applications in complex molecules and materials
The theoretical description of the electronic structure of complex molecules and materials faces two main challenges. First, the size of simulated systems needs to be large enough to remove the artificial effects caused by the finite size. Second, many-body effects need to be described accurately. While high-level methods can tackle the latter, their high computational costs limit them to small-size simulated systems. On the other hand, while low-cost methods like density functional theory can deal with large-size systems, they are usually unable to provide a satisfactory accuracy for systems with strong many-body effects. Therefore, methods that can balance cost and accuracy are highly desirable.
In this talk, I will discuss my perspective on the development of electronic structure methods for complex molecules. I will also present some electronic-structure methods that we are currently developing in our lab: self-consistent perturbation theory, hybrid quantum-classical framework, and neural-network quantum states.
Gold nanostructures as delivery systems and detectors of pramipexole drug
Pramipexole (PPX) is a well-known drug used in the treatment of Parkinson’s disease and restless legs syndrome. To probe its delivery and detection using gold nanoparticles, we carry out a theoretical investigation on the pramipexole–Au cluster interactions. Three gold AuN clusters with sizes N = 6, 8 and 20 are used as reactant models to simulate the metallic nanostructured surfaces. Quantum chemical computations are performed in both gas phase and aqueous environments using density functional theory (DFT) with the PBE functional and the cc-pVDZ-PP/cc-pVTZ basis set.
The PPX drug is mainly adsorbed on gold clusters via its nitrogen atom of the thiazole ring with binding energies of ca. 22 to 28 kcal/mol in vacuum and ca. 18 to 24 kcal/mol in aqueous solution. Surface-enhanced Raman scattering (SERS) of PPX adsorbed on the Au surfaces and its desorption process are also examined. A chemical enhancement mechanism for SERS procedure is again established in view of the formation of nonconventional hydrogen interactions Au/H–N. The binding of PPX to a gold cluster is expected to be reversible and triggered by the presence of cysteine residues in protein matrices or lower-shifted alteration of the environment’s pH.