[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-125101-en":3,"doc-seo-125101-105":30,"detail-sidebar-cat-0-en-105":95},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},125101,7971461740886,"Theodore","https://ap-avatar.wpscdn.com/davatar_3d24733baf745e90a7e4bdd5f77d97b2",8,"Research & Report","Machine Learning Intermolecular Potentials to Design Materials for Green Refrigeration","This project addresses the challenge of modeling interatomic interactions in the barocaloric material quinuclidinium hexafluorophosphate (Quin-PF6) through the development of a custom forcefield (FF). A neural network was trained on forcefield-derived learning signals to improve accuracy and adaptability, with the goal of enabling replacement of classical forcefields in molecular dynamics workflows. Training relied on trajectory data from molecular dynamics coupled with high-quality ab initio calculations, validated through comprehensive comparisons to pressure-dependent behavior. Results showed strong agreement between SAPT2 energies and optimized-forcefield predictions, while QUIN-PF6 modeling remained more complex.","Queen Mary University of London  \nMachine Learning Intermolecular Potentials to Design Materials for Green Refrigeration  \nA thesis submitted to the Queen Mary University of London for the degree of Master of Philosophy (MPhil)  \nBy  \nMiguel de Jesus Chavez Hernandez  \nStudent ID: [Redacted for confidentiality reasons]  \nSupervisor  \nProfessor Anthony E. Phillips School of Physical and Chemical Sciences Queen Mary University of London  \nSeptember, 2024  \nContents  \n1 Introduction 6  \n1.1 Motivations .......................................... 6  \n1.2 Aims .............................................. 7  \n1.3 Objectives ........................................... 7  \n2 Background 8  \n2.1 Materials and properties ................................... 8  \n2.1.1 Introduction to the barocaloric effect and plastic crystals ............ 8  \n2.1.2 Barocaloric Effect and its Thermodynamic Foundations ............. 9  \n2.1.3 Current Barocaloric Materials and the Relevance of Ionic Plastic Crystals ... 10  \n2.1.4 Promising ionic plastic crystals: Quin-PF6 .................... 11  \n2.2 Molecular Dynamics ..................................... 12  \n2.2.1 Introduction to molecular dynamics ........................ 12  \n2.2.2 Forcefields and intermolecular potentials ...................... 12  \n2.2.3 Equations of motion ................................. 14  \n2.2.4 Simulation parameters ................................ 16  \n2.2.5 System specifications and ensembles ........................ 16  \n2.3 Quantum Chemistry ..................................... 18  \n2.3.1 Introduction to quantum chemistry ......................... 18  \n2.3.2 Ab initio methods in quantum chemistry ..................... 19  \n2.3.3 Schrödinger equation ................................. 19  \n2.3.4 Hartree-Fock Method: the beginning ........................ 21  \n2.3.5 Density functional theory DFT ........................... 22  \n2.3.6 Symmetry adapted perturbation theory SAPT .................. 24  \n2.4 Machine Learning ....................................... 25  \n2.4.1 Introduction to Deep Learning ........................... 25  \n2.4.2 Data sources and preparation ............................ 26  \n2.4.3 Artificial Neural Networks (ANNs) ......................... 27  \n2.4.4 Structure of a Neural Network ........................... 28  \n3 Methodology 30  \n3.0.1 Forcefield and Potential Model for Quin-PF6 System ............... 30  \n3.1 Data Generation for forcefield Fitting ........................... 31  \n3.1.1 Forcefield Optimization ............................... 34  \n3.1.2 Application of the Optimized Forcefield in Molecular Dynamics ......... 35  \n3.1.3 Software for data analysis and visualisation .................... 36  \n3.1.4 Conversion to Descriptors and Neural Network Application ........... 36  \n4 Results and discussions 38  \n4.1 Force field Fitting and Energy Reproduction ........................ 38  \n4.1.1 PF6-PF6 Optimization ................................ 38  \n4.1.2 QUIN-QUIN Optimization ............................. 43  \n4.1.3 QUIN-PF6 Optimization .............................. 47  \n4.1.4 System simulation and phase transitions identification .............. 52  \n4.2 Neural Network Fitting for Energy Predictions ...................... 57  \n4.2.1 Feedforward Neural Network ............................ 58  \n4.2.2 Feedforward Neural Network with more layers ................... 59  \n4.2.3 Feedforward Neural Network with More Neurons and More Layers ....... 59  \n4.2.4 Feedforward Neural Network with AdamW Optimizer .............. 60  \n4.2.5 Feedforward Neural Network with AdamW Optimizer with More Neurons ... 61  \n4.2.6 Convolutional Neural Network ........................... 61  \n4.2.7 Long Short-Term Memory (LSTM) Network .................... 62  \n5 Conclusions and Further Work 64  \n5.1 Conclusions .......................................... 64  \n5.2 Further Work ......................................... 65  \nAbstract  \nThis project addresses the challenge o","cbCaikUZvnPnJdwi","https://ap.wps.com/l/cbCaikUZvnPnJdwi","pdf",3517043,1,72,"English","en",105,"# Introduction\n## Motivations\n## Aims\n## Objectives\n# Background\n## Materials and properties\n## Molecular Dynamics\n## Quantum Chemistry\n## Machine Learning\n# Methodology\n## Forcefield and Potential Model for Quin-PF6 System\n## Data Generation for forcefield Fitting\n# Results and discussions\n## Force field Fitting and Energy Reproduction\n## Neural Network Fitting for Energy Predictions\n# Conclusions and Further Work\n## Conclusions\n## Further Work","[{\"question\":\"What is the main goal of the thesis?\",\"answer\":\"To model interatomic interactions in Quin-PF6 for barocaloric applications by developing a custom forcefield and training neural networks to improve energy predictions in molecular dynamics simulations.\"},{\"question\":\"How is the custom forcefield trained and validated?\",\"answer\":\"The forcefield is fitted using high-quality data derived from ab initio calculations, with training signals coming from molecular dynamics trajectory data, then validated via comparisons of simulation results against reference behavior under pressure changes.\"},{\"question\":\"What do the results show about energy prediction accuracy?\",\"answer\":\"The optimized forcefield demonstrates strong correlation with energies calculated using SAPT2, capturing complex intermolecular interactions and pressure-dependent energy variations.\"},{\"question\":\"What challenges remain for future work?\",\"answer\":\"Accurately modeling the more complex QUIN-PF6 system, improving molecular interaction representations via advanced descriptors, expanding and refining the dataset, and exploring hybrid models combining machine learning with traditional physical approaches.\"}]","Machine Learning Intermolecular Potentials to Design Materials for Green Refrigeration | PDF",1785896654,181,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":90,"head_meta":92,"extra_data":94,"updated_unix":28},"machine-learning-intermolecular-potentials-to-design-materials-for-green-refrigeration","",{"@graph":36,"@context":89},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/machine-learning-intermolecular-potentials-to-design-materials-for-green-refrigeration/125101/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-05",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81,85],{"name":72,"@type":73,"acceptedAnswer":74},"What is the main goal of the thesis?","Question",{"text":75,"@type":76},"To model interatomic interactions in Quin-PF6 for barocaloric applications by developing a custom forcefield and training neural networks to improve energy predictions in molecular dynamics simulations.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How is the custom forcefield trained and validated?",{"text":80,"@type":76},"The forcefield is fitted using high-quality data derived from ab initio calculations, with training signals coming from molecular dynamics trajectory data, then validated via comparisons of simulation results against reference behavior under pressure changes.",{"name":82,"@type":73,"acceptedAnswer":83},"What do the results show about energy prediction accuracy?",{"text":84,"@type":76},"The optimized forcefield demonstrates strong correlation with energies calculated using SAPT2, capturing complex intermolecular interactions and pressure-dependent energy variations.",{"name":86,"@type":73,"acceptedAnswer":87},"What challenges remain for future work?",{"text":88,"@type":76},"Accurately modeling the more complex QUIN-PF6 system, improving molecular interaction representations via advanced descriptors, expanding and refining the dataset, and exploring hybrid models combining machine learning with traditional physical approaches.","https://schema.org",{"og:url":52,"og:type":91,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":93,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":96},[97,101,105,109,114,119,124,127,132,135,139],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":106,"show_sort_weight":107,"slug":108},"Exam",70,"exam",{"id":110,"doc_module":4,"doc_module_name":46,"category_name":111,"show_sort_weight":112,"slug":113},5,"Comic",60,"comic",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},6,"Technology",50,"technology",{"id":120,"doc_module":4,"doc_module_name":46,"category_name":121,"show_sort_weight":122,"slug":123},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":125,"slug":126},30,"research-report",{"id":128,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":130,"slug":131},9,"Religion & Spirituality",20,"religion-spirituality",{"id":130,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":130,"slug":134},"World Cup","world-cup",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":136,"slug":138},10,"Lifestyle","lifestyle",{"id":140,"doc_module":4,"doc_module_name":46,"category_name":141,"show_sort_weight":110,"slug":142},19,"General","general"]