Article
Chemistry, Multidisciplinary
Yeji Kim, Yoonho Jeong, Jihoo Kim, Eok Kyun Lee, Won June Kim, Insung S. Choi
Summary: Most graph neural networks in deep-learning chemistry are based on the 2D representation of molecules, but our proposed model, MolNet, incorporates 3D non-bond information and bond-strength information, achieving superior performance in classification and regression tasks.
CHEMISTRY-AN ASIAN JOURNAL
(2022)
Article
Biochemistry & Molecular Biology
Tiago Janela, Kosuke Takeuchi, Juergen Bajorath
Summary: Fingerprint representations of chemical structure are widely used in chemoinformatics and computational medicinal chemistry, with a distinction between 2D and 3D FPs. 2D FPs are particularly popular for similarity searching and compound classification, showcasing robustness and comparable performance to 3D FPs. A new 2D FP called CSFP has been developed with fewer structural features, demonstrating high performance in similarity searching and machine learning, challenging the belief that more structural information is necessary for establishing molecular similarity relationships.
Article
Chemistry, Physical
Qun-Chao Ding, Chun-Sheng Jia, Chao-Wen Wang, Xiao-Long Peng, Jian-Yi Liu, Lie -Hui Zhang, Rui Jiang, Su -Yang Zhu, Hua Yuan, Hao-Xuan Tang
Summary: We propose a unified non-fitting formulization representation of the Gibbs free energy for diatomic substances, which correlates temperature and pressure with four molecular constants. To validate the model, we use it to predict Gibbs free energy and isobaric heat capacity values for 36 diatomic substances across a wide temperature range. The average absolute deviations (AADs) between the predicted values and NIST data are within 0.0145% and 6.7765% for Gibbs free energy and isobaric heat capacity, respectively, for the 36 substances.
JOURNAL OF MOLECULAR LIQUIDS
(2023)
Article
Engineering, Electrical & Electronic
Provakar Paul, Arabinda Mallick, Tapas Majumdar
Summary: The molecular password system is based on commercial fluorescent dyes and allows encryption of passwords with identification of each character through fluorescent readouts. The system supports all 44 (256) four-character passwords from four chemical inputs, with potential for increasing password length for more complex applications.
ACS APPLIED ELECTRONIC MATERIALS
(2021)
Article
Biochemistry & Molecular Biology
Sam De Craemer, Karen Driesen, Bart Ghesquiere
Summary: Tracer metabolomics is a powerful technology for studying disease-inflicted metabolic mechanisms. However, the interpretation of tracer metabolomics results can be challenging, especially for less experienced researchers. This research proposes a structured approach using intuitive visualization concepts and a new opensource tool to help researchers interpret their own datasets.
Article
Chemistry, Multidisciplinary
Angana Borbora, Manideepa Dhar, Arpita Shome, Nishanta Barman, Swaraj Roy, Uttam Manna
Summary: Bio-inspired underwater superoleophobicity is suitable for a variety of applications, but achieving both mechanical durability and optical transparency is challenging.
ADVANCED FUNCTIONAL MATERIALS
(2023)
Article
Engineering, Electrical & Electronic
LiangLiang Liu, Zijuan Tang, Chenyang He, Serhiy Korposh, Shuqin Lou, Stephen P. Morgan
Summary: The optical fibre ammonia gas sensor demonstrated in the study utilizes a functionalized suspended-core fibre and a ratiometric method for signal processing, showing a capability of detecting ammonia at ppb levels with a fast response time and reusability after treatment.
JOURNAL OF LIGHTWAVE TECHNOLOGY
(2021)
Article
Engineering, Environmental
Chung-Seop Lee, Sujin Guo, Hojung Rho, Juliana Levi, Sergi Garcia-Segura, Michael S. Wong, Jorge Gardea-Torresdey, Paul Westerhoff
Summary: The study found that the use of different model solutions led to gradual changes in surface composition, morphology, and catalytic activity of the catalyst, providing insights into potential strategies for improving long-term catalyst operations.
ENVIRONMENTAL SCIENCE & TECHNOLOGY
(2021)
Article
Infectious Diseases
Liqiang Li, Xiang Guo, Xiaoqing Zhang, Lingzhai Zhao, Li Li, Yuji Wang, Tian Xie, Qingqing Yin, Qinlong Jing, Tian Hu, Ziyao Li, Rangke Wu, Wei Zhao, Sherman Xuegang Xin, Benyun Shi, Jiming Liu, Shang Xia, Zhiqiang Peng, Zhicong Yang, Fuchun Zhang, Xiao-Guang Chen, Xiaohong Zhou
Summary: This study establishes a global genotyping framework for dengue virus (DENV) using phylogenetics, population genetics, and other methods. It identifies global patterns of DENV-1 and demonstrates increasing cross-transmission trends. The framework can be used for accurate surveillance and warning, and is important for preventing the global spread of dengue.
INFECTIOUS DISEASES OF POVERTY
(2022)
Review
Biochemistry & Molecular Biology
Colin P. S. Kruse, Sarah E. Wyatt
Summary: This article reviews the response of plants to gravity, focusing on the differential protein and gene regulation induced by the absence of gravity stimulus aboard the International Space Station. It proposes an integrated theoretical schematic of gravity response, with nitric oxide and cysteine S-nitrosation playing a central role. The proposed model is supported by the evolutionary conservation of regulatory amino acids within protein components and may have implications for studying various environmental and stress responses.
Article
Chemistry, Multidisciplinary
Elena-Luiza Epure, Iuliana Stoica, Raluca Marinica Albu, Camelia Hulubei, Andreea Irina Barzic
Summary: The orientation of liquid crystal displays is affected by the surface anisotropy created through a novel strategy of combining rubbing with cloth and stretching via pressing. The chemical structure of the polymer and order of surface treatment steps impact the orientation of chains, resulting in competitive factors affecting interactions with nematic liquid crystals and ultimately influencing the reliability and uniformity of display devices.
Article
Engineering, Electrical & Electronic
Syed Ghause Ibrahim
Summary: In the emerging field of research, the focus is on nanomaterials which preserve the properties of a good material with advantages in size and cost. The research work focuses on nanocrystalline manganese sulfide thin films grown using spray pyrolysis technique on glass substrates, showing n-type conductivity and a resistivity value of 6.232 x 10^6 ohm cm.
JOURNAL OF MATERIALS SCIENCE-MATERIALS IN ELECTRONICS
(2021)
Article
Chemistry, Multidisciplinary
Yoonho Jeong, Jihoo Kim, Yeji Kim, Insung S. Choi
Summary: This work proposes the use of I-N - A instead of the normalized form I-N + A in graph convolutional networks for deep-learning chemistry. The performance of the GCN model with D-1/2(I-N - A)D-1/2 is comparable to the vanilla GCN using (D) over tilde (-1/2)(I-N + A)(D) over tilde in various chemistry datasets. The use of I-N - A provides more chemically intuitive information on bond properties in molecules. This work emphasizes the necessity of addressing molecular-representation problems in deep-learning chemistry from unprecedented angles.
BULLETIN OF THE KOREAN CHEMICAL SOCIETY
(2022)
Article
Chemistry, Physical
Guan-Hong Liu, Qun-Chao Ding, Chao-Wen Wang, Chun-Sheng Jia
Summary: In this study, a non-fitting formula for predicting Gibbs free energy of five compounds was developed. The formula only requires knowledge of molecular mass and experimental constant data, without the need for fitting a large amount of experimental data. The predicted values of Gibbs free energy had low deviations from the experimental data.
JOURNAL OF MOLECULAR STRUCTURE
(2023)
Article
Engineering, Mechanical
Ji Lin, Panpan Zhu, Chuanshuai Tian, Haofei Zhou, Rui Xiao
Summary: This work reveals an abnormal rate-dependent stress relaxation response of glassy polymers, where a higher initial loading rate leads to a more-activated structure state of the polymers. By introducing the effective temperature model and Adam-Gibbs model, a micro-mechanical model is developed to capture this abnormal behavior successfully.
EXTREME MECHANICS LETTERS
(2022)
Article
Chemistry, Multidisciplinary
Christoph Plett, Stefan Grimme
Summary: Modeling intermolecular interactions of complex non-covalent structures is important in chemistry. To facilitate this, we introduce an automated computational interaction site screening workflow (aISS) that generates reasonable geometries and saves computational time.
ANGEWANDTE CHEMIE-INTERNATIONAL EDITION
(2023)
Article
Chemistry, Inorganic & Nuclear
Hui Zhu, Zheng-Wang Qu, Stefan Grimme, Philipp C. C. Brehm, Rainer Streubel
Summary: Recently, we synthesized phosphane W(CO)(5) complexes with a P-O-TEMP ligand motif as stable precursors for thermally accessible phosphanoxyl complex radicals with a P-O group. By using dispersion-corrected DFT calculations, we explored W(CO)(5) and Fe(CO)(4) phosphane complexes with P-E-TEMP ligands (E = O, S, NMe, and PMe) to obtain thermally accessible radicals with a P-E motif. In addition, we disclosed a more general SET oxidation approach to synthesize P-E radicals via anionic precursors. Furthermore, we studied the tendencies for self-trapping and prototropic reactions of such radical complexes for the first time.
DALTON TRANSACTIONS
(2023)
Article
Chemistry, Physical
A. McSloy, G. Fan, W. Sun, C. Hoelzer, M. Friede, S. Ehlert, N. -E. Schuette, S. Grimme, T. Frauenheim, B. Aradi
Summary: Tight-binding approaches, such as Density Functional Tight-Binding (DFTB) and extended tight-binding schemes, enable efficient quantum mechanical simulations of large systems and long-time scales. They are derived from ab initio density functional theory with pragmatic approximations and empirical terms, striking a balance between speed and accuracy. Machine learning techniques can enhance their accuracy by tuning empirical parameters, especially when incorporating local environment information. The Tight-Binding Machine Learning Toolkit is an open-source framework that facilitates the implementation of these combined approaches, offering potential for various applications and demonstrating the functionality of the toolkit.
JOURNAL OF CHEMICAL PHYSICS
(2023)
Article
Chemistry, Physical
Stefan Grimme, Marcel Mueller, Andreas Hansen
Summary: This article describes a new tight-binding electronic structure Hamiltonian that achieves high accuracy and speedup in calculations using a deeply contracted valence double-zeta basis set. The method performs well for difficult systems and bonding situations, and can be used for machine learning applications or as a basis for accurate high-speed DFT methods.
JOURNAL OF CHEMICAL PHYSICS
(2023)
Editorial Material
Chemistry, Multidisciplinary
Zahid Hussain, Yong-an Luo, Yile Wu, Zheng-wang Qu, Stefan Grimme, Douglas W. Stephan
Summary: Interest in main group chemistry related to the Haber-Bosch process is less studied compared to transition metal species. In this study, steric demands in (tBuO(2)CN)(2) hinder the initial interaction of B(C6F5)(3) with nitrogen, leading to the loss of methylpropene and CO2 and the formation of diazene (N2H2) borane adduct 1 and the analogous hydrazine (N2H4) adduct 2. These species react with basic phosphines to form anions of 3 and 5, which contain N2H and N2H3 fragments and serve as metal-free models of plausible intermediates in N-2 reduction.
JOURNAL OF THE AMERICAN CHEMICAL SOCIETY
(2023)
Article
Chemistry, Multidisciplinary
Marcel Bamberg, Thomas Gasevic, Michael Bolte, Alexander Virovets, Hans-Wolfram Lerner, Stefan Grimme, Markus Bursch, Matthias Wagner
Summary: Silafulleranes with endohedral Cl- ions are a unique class of silicon clusters and host-guest complexes. In this study, regioselective derivatization reactions were performed on a specific siladodecahedrane, resulting in the formation of perhydrogenated and non-pyrophoric hydridosilanes. The influence of solvent on the reaction outcomes was also investigated.
JOURNAL OF THE AMERICAN CHEMICAL SOCIETY
(2023)
Article
Chemistry, Physical
Marcel Stahn, Sebastian Ehlert, Stefan Grimme
Summary: We have developed a new method, named CPCM-X, to accurately account for solvation effects in semiempirical quantum mechanics. This method combines the computationally efficient ddCOSMO model with established solvation models like COSMO-RS and SMD, and improves the accuracy of solvation free energies by up to 40% for large supramolecular association reactions. The proposed method is more accurate and more than 2 orders of magnitude faster than the current methods in the xtb program package.
JOURNAL OF PHYSICAL CHEMISTRY A
(2023)
Article
Chemistry, Multidisciplinary
Dipendu Mandal, Zheng-Wang Qu, Stefan Grimme, Douglas W. Stephan
Summary: Cyclopropenium cations with electron deficient substituents exhibit Lewis acidity despite the presence of π-electrons. The chloride and electron affinities are analyzed computationally and experimentally, respectively. These cations form traditional Lewis acid-base adducts with PPh3, while sterically hindered phosphines lead to frustrated Lewis pairs that participate in FLP additions. Depending on the basicity of the phosphine, addition to alkynes or alkyne deprotonation can occur. In both cases, new C-C bonds are formed, thus expanding the application of FLP chemistry to these delocalized π-cations.
CHEMICAL COMMUNICATIONS
(2023)
Article
Chemistry, Physical
Christoph Plett, Abylay Katbashev, Sebastian Ehlert, Stefan Grimme, Markus Bursch
Summary: Efficient quantum chemical methods are required for the computational treatment of large molecular structures. The ONIOM multi-layer scheme has been successfully implemented into the xtb program package and applied to transition-metal complexes with hundreds of atoms. By combining density functional theory, semi-empirical, and force-field methods in the ONIOM framework, computational costs can be significantly reduced without sacrificing accuracy.
PHYSICAL CHEMISTRY CHEMICAL PHYSICS
(2023)
Article
Chemistry, Multidisciplinary
Tunga Salthammer, Uwe Hohm, Marcel Stahn, Stefan Grimme
Summary: Proton transfer reaction mass spectrometry (PTR-MS) is an important tool for indoor related sciences, allowing online monitoring and identification of substance mixtures in the gas phase. Kinetic laws and collision theory are used to quantify and calculate reaction rate constants under different conditions in the reaction chamber. The dipole moment and polarizability of organic compounds frequently found in indoor air were determined using advanced quantum mechanical methods, and their applicability in PTR-MS measurements was critically discussed.
Article
Chemistry, Multidisciplinary
Marcel Bamberg, Thomas Gasevic, Michael Bolte, Alexander Virovets, Hans-Wolfram Lerner, Stefan Grimme, Markus Bursch, Matthias Wagner
Summary: Treating perhydrogenated silafullerane [nBu(4)N][Cl@Si-20(SiH3)(12)H-8] with BBr3 resulted in partially and exhaustively brominated clusters, [nBu(4)N][Cl@Si-20(SiBr2H)(12)Br-8] and [nBu(4)N][Cl@Si-20(SiBr3)(12)Br-8], respectively. Full bromination increased steric strain on the cluster surface, enabling regioselective derivatization of the Si-32 framework under mild conditions. Partial Br/H exchange on [nBu(4)N][Cl@Si-20(SiBr2H)(12)Br-8] with iBu(2)AlH yielded [nBu(4)N][Cl@Si-20(SiH3)(12)Br-8].
CHEMICAL COMMUNICATIONS
(2023)
Article
Chemistry, Multidisciplinary
Zahid Hussain, Yong-An Luo, Yile Wu, Zheng-Wang Qu, Stefan Grimme, Douglas W. Stephan
Summary: The reactions of (tBuO(2)CN)(2) with FLPs were investigated. B(C6F5)(3) interacts with the carbonyl oxygen atoms and causes the loss of CH2=CMe2. However, in the presence of basic donors, the protons are intercepted and form salts [Hbase](2) [((C6F5)(3)BO2CN)(2)]. In the presence of (o-Tol)(3)P, a proton transfers to the diazo-N atom and forms (o-Tol)(3)PN(CO(2)tBu)NHB(C6F5)(3). Further addition of B(C6F5)(3) prompts the loss of olefin CH2=CMe2 and CO2, resulting in the formation of (o-Tol)(3)PNHNHB(C6F5)(3). DFT calculations provide insights into the reaction mechanisms and confirm the nature of 5 as a FLP reduction product of a diazene fragment.
CHEMICAL COMMUNICATIONS
(2023)
Article
Chemistry, Inorganic & Nuclear
Vaibhav Bedi, Dipendu Mandal, Zahid Hussain, Shi-Ming Chen, Yile Wu, Zheng-Wang Qu, Stefan Grimme, Douglas W. Stephan
Summary: The reaction of (tBuO(2)CN)(2) with 9-BBN leads to the formation of a bicyclic heterocyclic compound, while its reactions with BF3 or [Et3Si][B(C6F5)(4)] result in the isolation of different compounds. Computational studies reveal that the steric and electronic properties of the Lewis acid are important in the formation of one of the compounds.
DALTON TRANSACTIONS
(2024)
Article
Chemistry, Multidisciplinary
Christoph Plett, Stefan Grimme, Andreas Hansen
Summary: The computational prediction of molecular properties relies on reliable benchmark sets, however currently available sets do not account for solvent effects. This study introduces a solvated benchmark set with explicit water molecules, and provides accurate reference values for conformational energies of biomolecules in aqueous solution.
JOURNAL OF COMPUTATIONAL CHEMISTRY
(2023)
Article
Chemistry, Physical
Marvin Friede, Sebastian Ehlert, Stefan Grimme, Jan-Michael Mewes
Summary: This study investigates the interdependency between dispersion correction parameters and the range-separation parameter omega in large molecules. The results show that some functionals are strongly affected by omega values, leading to overbinding and poor performance. Strategies to mitigate these issues are discussed, providing insights for future improvements.
JOURNAL OF CHEMICAL THEORY AND COMPUTATION
(2023)