Article
Engineering, Chemical
Hannes Weinhold, Klaus Wekenborg, Jurgen Rarey
Summary: In this study, a new method is proposed to improve the accuracy of conventional group contribution models for specialty chemical products, by using a local estimator that leverages data on similar molecules, resulting in reduced errors in melting and boiling point temperature and vapor pressure estimation. This enables faster and more accurate development of the downstream production process.
INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH
(2023)
Article
Chemistry, Physical
Nanthinie Thangarajoo, Syed Ali Ammar Taqvi, Pranesh Matheswaran, Khairiraihanna Johari, Mohd Hilmi Noh
Summary: The Group Contribution (GC) method is used to estimate the thermodynamics properties of a compound by developing a predictive model for the infinite dilution activity coefficient (IDAC) of ethanol and propanol in ionic liquids. The GC model, using the van't Hoff model and multiple linear regression (MLR), showed good performance in predicting IDAC values as compared to experimental values. The structural characteristics of ionic liquids play a significant role in their interaction with alcohol solutes, leading to variations in IDAC values.
JOURNAL OF MOLECULAR LIQUIDS
(2021)
Article
Chemistry, Physical
Ahmadreza Roosta, Reza Haghbakhsh, Ana Rita C. Duarte, Sona Raeissi
Summary: In this study, hybrid machine learning models were developed to predict the viscosity of DESs using the group contribution concept and artificial neural network and support vector machine algorithms. The models can accurately determine the viscosity of DESs.
JOURNAL OF MOLECULAR LIQUIDS
(2023)
Article
Engineering, Chemical
Yijia Sun, Nikolaos Sahinidis
Summary: This paper presents a new method for identifying functional groups in molecular structures and constructing group contribution models to predict properties. The models showed good predictive power and were embedded in a molecular design framework to design organosilicon coolants systems, outperforming current commercial and nonorganosilicon coolants.
INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH
(2021)
Article
Chemistry, Multidisciplinary
Wan Ren, Song Fan, Peng Changjun, Liu Honglai
Summary: The group contribution method (GCM) was established to predict the infinite dilution molar conductivity model of unconventional cations, providing insights into the influence of different groups on the conductivity and the relationship between temperature and conductivity. The results showed that GCM accurately predicted the molar conductivity with a low average absolute relative deviation, proving to be a simple and reliable method for studying unconventional ions.
CHEMICAL JOURNAL OF CHINESE UNIVERSITIES-CHINESE
(2021)
Article
Chemistry, Physical
Calen J. Leverant, Jacob A. Harvey, Todd M. Alam, Jeffery A. Greathouse
Summary: Predicting the diffusion coefficient of fluids under nanoconfinement is crucial for various applications, and a machine learning model trained on a subset of MD data showed good predictive ability for Lennard-Jones fluids in pores. By using MD simulations and an artificial neural network model, the study presented a new approach for predicting fluid diffusion coefficients.
JOURNAL OF PHYSICAL CHEMISTRY C
(2021)
Article
Engineering, Chemical
Tian Tan, Hongye Cheng, Guzhong Chen, Zhen Song, Zhiwen Qi
Summary: Accurate prediction of infinite dilution activity coefficient (gamma(infinity)) is crucial for phase equilibria and process design. This study proposes a new method based on neural collaborative filtering (NCF) to fill in the gamma(infinity) matrix, and the experimental results show that it outperforms traditional models and previous machine learning models. The completed matrix can also be used for solvent screening and parameter extension.
Article
Biochemical Research Methods
Hao Li, Yirui Wu, Hexuan Hu, Hu Lu, Qian Huang, Shaohua Wan
Summary: Deep learning has made significant progress in medical image analysis, but its lack of interpretability may lead to high-risk misdiagnosis. To address this, we propose Representation Group-Disentangling Network (RGD-Net) to explain the feature extraction and decision-making process of deep learning in medical analysis, in order to correct biased results.
Article
Chemistry, Physical
Joshua P. Allers, Jane Keth, Todd M. Alam
Summary: Artificial neural networks were developed to predict the self-diffusion constants in binary fluid mixtures. The presence of strong hydrogen bonding molecules and other fluid properties were considered in the input features to the neural networks, resulting in an accurate and generalized model. The importance of critical properties and self-association energies on model performance was also investigated.
JOURNAL OF PHYSICAL CHEMISTRY B
(2022)
Article
Mathematics
Jan Chvalina, Bedrich Smetana, Jana Vyroubalova
Summary: This paper discusses the construction of infinite cyclic groups of differential neurons and their basic properties, which are modifications of artificial neurons analogous to linear ordinary differential operators.
Article
Chemistry, Physical
Joshua P. Allers, Fernando H. Garzon, Todd M. Alam
Summary: Artificial neural networks (ANNs) were developed to accurately predict self-diffusion constants for different phases. The log transformation of diffusion values improved prediction accuracy, while the density of the compound was found to have a significant impact on the predictions.
PHYSICAL CHEMISTRY CHEMICAL PHYSICS
(2021)
Article
Thermodynamics
Yingxue Fu, Yuqiu Chen, Chuntao Zhang, Yang Lei, Xinyan Liu
Summary: A database was established to collect surface tension data for IL-H2O hybrid systems, and an ANN-GC model was proposed to predict the surface tension. The model showed reliable predictions for both the hybrid system and pure IL systems.
FLUID PHASE EQUILIBRIA
(2022)
Article
Biochemistry & Molecular Biology
Lei Xu, Shourun Pan, Leiming Xia, Zhen Li
Summary: In this paper, we combined the SALSTM and GAT methods to explore the feature information of molecules from sequences and graphs. By using SALSTM to obtain embedded atoms through SMILES strings and combining them with graph node features, we fed them into GAT to extract the global molecular representation. Data augmentation was added to enlarge the training dataset and improve the model performance. The fusion of attention layers from both models enhanced the interpretability of the model and highlighted key atoms. Comparisons with other graph-based and sequence-based methods showed that our method achieved high prediction accuracy and good generalizability for multiple datasets.
Article
Engineering, Chemical
Fangyou Yan, Dongdong Cao, Xiaojie Feng, Jialiang Xiong, Qiang Wang, Qingzhu Jia, Shuqian Xia
Summary: The atomic connectivity group contribution (ACGC) method is developed for the first time to predict critical properties of organic compounds. A new group defining method called atomic adjacent group (AAG) method is proposed based on the relationship between core atom and its adjacent atoms. Shape factors (SFs) and atomic connectivity factors (ACFs) are introduced to address the challenge of dealing with isomers. ACGC models using a general formula are established, achieving high accuracy in predicting critical temperature, pressure, and volume for various organic compounds.
CHEMICAL ENGINEERING SCIENCE
(2023)
Article
Mathematics
Gabriel Larotonda, Jose Luna
Summary: This article extends several results for the structure group of a real Jordan algebra V to the setting of infinite dimensional JB-algebras. It is proved that the structure group Str(V), the cone preserving group G(Omega), and the automorphism group Aut(V) of the algebra V are embedded Banach-Lie groups of GL(V), and that each of the inclusions Aut(V) ⊆ G(Omega) ⊆ Str(V) are embedded Banach-Lie subgroups. A full description of the components of Str(V) using cones, isotopes, and central projections is given. These results are applied to V = B(H)sa, the special JB-algebra of self-adjoint operators on an infinite dimensional complex Hilbert space, to describe the groups Str(V), G(Omega), Aut(V), their Banach-Lie algebras, and their connected components. It is shown that the action of the unitary group of H on Aut(V) has smooth local cross sections, thus Aut(V) is a smooth principal bundle over the unitary group, with the structure group S1.
JOURNAL OF ALGEBRA
(2023)
Article
Thermodynamics
Mikael Mannisto, Petri Uusi-Kyyny, Dominique Richon, Ville Alopaeus
JOURNAL OF CHEMICAL AND ENGINEERING DATA
(2017)
Article
Thermodynamics
Deresh Ramjugernath, Alain Valtz, Dominique Richon, Mark D. Williams-Wynn, Christophe Coquelet
JOURNAL OF CHEMICAL AND ENGINEERING DATA
(2017)
Article
Thermodynamics
Muhammad Saad Qureshi, Tom Le Nedelec, Hernando Guerrero-Amaya, Petri Uusi-Kyyny, Dominique Richon, Ville Alopaeus
JOURNAL OF CHEMICAL THERMODYNAMICS
(2017)
Article
Thermodynamics
Farhad Gharagheizi, Poorandokht Ilani-Kashkouli, Ronald C. Hedden
Article
Thermodynamics
Deresh Ramjugernath, Alain Valtz, Dominique Richon, Mark D. Williams-Wynn, Christophe Coquelet
JOURNAL OF CHEMICAL AND ENGINEERING DATA
(2018)
Article
Thermodynamics
Paul J. Antonick, Zhichao Hu, Yoshiko Fujita, David W. Reed, Gaurav Das, Lili Wu, Radha Shivaramaiah, Paul Kim, Ali Eslamimanesh, Malgorzata M. Lencka, Yongqin Jiao, Andrzej Anderko, Alexandra Navrotsky, Richard E. Riman
JOURNAL OF CHEMICAL THERMODYNAMICS
(2019)
Article
Thermodynamics
Gaurav Das, Malgorzata M. Lencka, Ali Eslamimanesh, Peiming Wang, Andrzej Anderko, Richard E. Riman, Alexandra Navrotsky
JOURNAL OF CHEMICAL THERMODYNAMICS
(2019)
Article
Engineering, Chemical
Mohamed K. Hadj-Kali, Mamoun Althuluth, Salim Mokraoui, Irfan Wazeer, Emad Ali, Dominique Richon
CHEMICAL ENGINEERING COMMUNICATIONS
(2020)
Article
Engineering, Chemical
Petri Uusi-Kyyny, Muhammad Saad Qureshi, Juha-Pekka Pokki, Ville Alopaeus, Dominique Richon
INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH
(2019)
Article
Thermodynamics
Jouni Touronen, Mikael Mannisto, Dominique Richon, Petri Uusi-Kyyny, Ville Alopaeus
JOURNAL OF CHEMICAL AND ENGINEERING DATA
(2020)
Article
Engineering, Chemical
Farhad Gharagheizi, David S. Sholl
Summary: This paper systematically evaluates the accuracy of the ideal adsorbed solution theory (IAST) for gas adsorption using a large collection of binary experimental data, which will be valuable for future efforts to test or develop mixing theories that improve upon FAST. This analysis includes data from 63 gas mixtures of 37 different molecular species and 174 different adsorbents, making it the most systematic evaluation to date of the accuracy of IAST for gas adsorption.
INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH
(2022)
Article
Construction & Building Technology
Xiaofei Zhao, Minhaz Ur Rahman, Tharanga Dissanayaka, Farhad Gharagheizi, Carla Lacerda, Sanjaya Senadheera, Ronald C. Hedden, Gordon F. Christopher
Summary: The study investigates the effect of adding PP/PE copolymer modified asphalt binder on the ability to resist rutting in high temperatures. Results show that the PP/PE copolymer can significantly increase the rutting factor of asphalt binder within a specific temperature range, effectively reducing the risk of pavement rutting.
CASE STUDIES IN CONSTRUCTION MATERIALS
(2021)
Article
Nanoscience & Nanotechnology
Farhad Gharagheizi, Zhenzi Yu, David S. Shoff
Summary: A collection of over 20,000 experimentally derived crystal structures for metal-organic frameworks (MOFs) without two- or three-dimensional covalently bonded networks has been developed. These structures are obtained from materials available at the Cambridge Crystallographic Data Centre. More than 12,000 of these structures have been verified to be solvent-free and in exact agreement with the stoichiometry of the synthesized materials. The study also reveals that there are more than 2,000 1D MOFs with a zero band gap, and that adsorbate-induced deformation plays a significant role in determining adsorption isotherms in these materials.
ACS APPLIED MATERIALS & INTERFACES
(2022)
Article
Thermodynamics
Hassan Pahlavanzadeh, Ali Eslamimanesh, Amirreza Ghavi
Summary: In this study, the effects of TBPB and THF on the phase equilibrium of CO2 hydrates were investigated. The results showed that the presence of TBPB and THF in the system had a significant impact on the hydrate phase equilibrium conditions.
JOURNAL OF CHEMICAL AND ENGINEERING DATA
(2022)
Article
Chemistry, Physical
Jifeng Sun, Farhad Gharagheizi, Hanjun Fang, Peter I. Ravikovitch, David S. Sholl
Summary: This study combines structure screening and high-throughput DFT calculations to investigate the adsorption of O2 and N2 in materials, identifying multiple materials that selectively bind O2 over N2 and suggesting design rules for tuning the O2 and N2 affinities in these materials.
JOURNAL OF PHYSICAL CHEMISTRY C
(2022)