Predicting Single-Substance Phase Diagrams: A Kernel Approach on Graph Representations of Molecules

2021 ◽  
Author(s):  
Yan Xiang ◽  
Yu-Hang Tang ◽  
Hongyi Liu ◽  
Guang Lin ◽  
Huai Sun
Keyword(s):  
Phase Diagrams ◽  
Graph Representations ◽  
Single Substance ◽  
Kernel Approach

scholarly journals Predicting Single-Substance Phase Diagrams: A Kernel Approach on Graph Representations of Molecules

2021 ◽  
Author(s):  
Yan Xiang ◽  
Yu-Hang Tang ◽  
Hongyi Liu ◽  
Guang Lin ◽  
Huai Sun
Keyword(s):  
Covariance Function ◽  
Molecular Graph ◽  
Gaussian Process Regression ◽  
Graph Representation ◽  
Graph Kernel ◽  
Graph Representations ◽  
Kernel Machine ◽  
Pure Compounds ◽  
Kernel Approach ◽  
Pure Substances

<p>This work presents a Gaussian process regression (GPR) model on top of a novel graph representation of chemical molecules that predicts thermodynamic properties of pure substances in single, double, and triple phases. A transferable molecular graph representation is proposed as the input for a marginalized graph kernel, which is the major component of the covariance function in our GPR models. Radial basis function kernels of temperature and pressure are also incorporated into the covariance function when necessary. We predicted three types of representative properties of pure substances in single, double, and triple phases, i.e., critical temperature, vapor-liquid equilibrium (VLE) density, and pressure-temperature density. The data is collected from Knovel Data Analysis Beta: NIST ThermoDynamics Pure Compounds. The accuracy of the models is nearly identical to the precision of the experimental measurements. Moreover, the reliability of our predictions can be quantified on a per-sample basis using the posterior uncertainty of the GPR model. We compare our model against Morgan fingerprints and a graph neural network to further demonstrate the advantage of the proposed method. The marginalized graph kernel is computed using GraphDot package at <a href="https://github.com/yhtang/GraphDot">https://github.com/yhtang/GraphDot</a>. All codes used in this work can be found at <a href="https://github.com/Xiangyan93/Chem-Graph-Kernel-Machine">https://github.com/Xiangyan93/Chem-Graph-Kernel-Machine</a>.</p>

scholarly journals Predicting Single-Substance Phase Diagrams: A Kernel Approach on Graph Representations of Molecules

2021 ◽  
Author(s):  
Yan Xiang ◽  
Yu-Hang Tang ◽  
Hongyi Liu ◽  
Guang Lin ◽  
Huai Sun
Keyword(s):  
Covariance Function ◽  
Molecular Graph ◽  
Gaussian Process Regression ◽  
Graph Representation ◽  
Graph Kernel ◽  
Graph Representations ◽  
Kernel Machine ◽  
Pure Compounds ◽  
Kernel Approach ◽  
Pure Substances

<p>This work presents a Gaussian process regression (GPR) model on top of a novel graph representation of chemical molecules that predicts thermodynamic properties of pure substances in single, double, and triple phases. A transferable molecular graph representation is proposed as the input for a marginalized graph kernel, which is the major component of the covariance function in our GPR models. Radial basis function kernels of temperature and pressure are also incorporated into the covariance function when necessary. We predicted three types of representative properties of pure substances in single, double, and triple phases, i.e., critical temperature, vapor-liquid equilibrium (VLE) density, and pressure-temperature density. The data is collected from Knovel Data Analysis Beta: NIST ThermoDynamics Pure Compounds. The accuracy of the models is nearly identical to the precision of the experimental measurements. Moreover, the reliability of our predictions can be quantified on a per-sample basis using the posterior uncertainty of the GPR model. We compare our model against Morgan fingerprints and a graph neural network to further demonstrate the advantage of the proposed method. The marginalized graph kernel is computed using GraphDot package at <a href="https://github.com/yhtang/GraphDot">https://github.com/yhtang/GraphDot</a>. All codes used in this work can be found at <a href="https://github.com/Xiangyan93/Chem-Graph-Kernel-Machine">https://github.com/Xiangyan93/Chem-Graph-Kernel-Machine</a>.</p>

Transmission electron microscope studies in special ceramics

1978 ◽  
Vol 36 (1) ◽  
pp. 268-269
Author(s):  
A. Zangvil ◽  
L.J. Gauckler ◽  
G. Schneider ◽  
M. Rühle
Keyword(s):  
Phase Diagrams ◽  
Crystal Structures ◽  
Thin Foil ◽  
Limited Extent ◽  
Complex Materials ◽  
X Ray Diffraction ◽  
X Ray ◽  
Transmission Electron ◽  
Electron Microscopes ◽  
Lattice Resolution

The use of high temperature special ceramics which are usually complex materials based on oxides, nitrides, carbides and borides of silicon and aluminum, is critically dependent on their thermomechanical and other physical properties. The investigations of the phase diagrams, crystal structures and microstructural features are essential for better understanding of the macro-properties. Phase diagrams and crystal structures have been studied mainly by X-ray diffraction (XRD). Transmission electron microscopy (TEM) has contributed to this field to a very limited extent; it has been used more extensively in the study of microstructure, phase transformations and lattice defects. Often only TEM can give solutions to numerous problems in the above fields, since the various phases exist in extremely fine grains and subgrain structures; single crystals of appreciable size are often not available. Examples with some of our experimental results from two multicomponent systems are presented here. The standard ion thinning technique was used for the preparation of thin foil samples, which were then investigated with JEOL 200A and Siemens ELMISKOP 102 (for the lattice resolution work) electron microscopes.

Phase diagrams of antiferromagnetic Z(q) models

1993 ◽  
Vol 3 (12) ◽  
pp. 2397-2409
Author(s):  
A. Benyoussef ◽  
L. Laanaït ◽  
M. Loulidi
Keyword(s):  
Phase Diagrams

Study of the causes for wax deposition under the operating conditions of main oil pipelines

2020 ◽  
Vol 10 (6) ◽  
pp. 610-619
Author(s):  
Rustam Z. Sunagatullin ◽  
◽  
Rinat M. Karimov ◽  
Radmir R. Tashbulatov ◽  
Boris N. Mastobaev ◽  
...  
Keyword(s):  
Phase Diagrams ◽  
Total Content ◽  
Operating Conditions ◽  
Wax Deposition ◽  
Oil Saturation ◽  
Oil Pipelines ◽  
Thermobaric Conditions ◽  
Wall Zone ◽  
Significant Factors ◽  
Commercial Oil

The results of investigations of the main causes and the most significant factors of intensification of paraffin deposition in main oil pipelines are presented. A comprehensive analysis of the composition and properties of commercial oils and their sediments was carried out, according to which phase diagrams of equilibrium of oil dispersed systems were obtained using the example of commercial oils from Bashkir fields. Based on the phase diagrams, a curve of wax oil saturation was constructed, the analysis of which confirms that the existing thermobaric conditions during the operation of main oil pipelines do not allow transporting oil without the risk of waxing. It was noted a special influence of the value of the temperature gradient in the near-wall zone and the imbalance of the ratio of high-molecular oil components in commercial batches formed in the process of joint pumping on the intensity of waxing of sections of oil pipelines complicated by deposits, which was confirmed by statistical data on the frequency of pigging. The regularities obtained in this way are proposed to be used as an express method for predicting complications associated with intensive waxing of main oil pipelines. In order to quickly assess the risks of waxing of sections of main oil pipelines, an indicator is introduced that characterizes the ratio of the content of solid paraffins to the total content of resins and asphaltenes of oil, called the criterion of instability of a commercial oil batch.

Sign in / Sign up

Export Citation Format

Share Document