Enhanced formation of methane hydrates via graphene oxide: Machine learning insights from molecular dynamics simulations

Lin, Yanwen; Hao, Yongchao; Shi, Qiao; Xu, Yihua; Song, Zixuan; Zhou, Ziyue; Fu, Yuequn; Zhang, Zhisen; Wu, Jianyang

doi:10.1016/j.energy.2023.130080

Cited by 12 publications

(4 citation statements)

References 67 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the second stage, the extracted latent vectors are used as input to train a model using an LSTM network. The LSTM [19] is an improved version of recurrent neural networks, which is capable of learning longterm dependencies in sequential data. Very recently, LSTM has started to be applied in combination with MD [18,20,21] and phase-field simulations [27] for tasks such as microstructure formation, garnering attention in the field of materials science.…”

Section: Time Evolution Prediction Of Physical Properties Based On Lstmmentioning

confidence: 99%

“…In addition to the aforementioned wide range of ML applications, very recently, new attempts have been made to predict future structures and properties of materials at times not covered by the MD simulations using generative models and recurrent neural networks [15][16][17][18][19][20]. For example, MD-GAN (MD-Generative Adversarial Network) [15,16] is a method that utilizes a generative model to predict long-term dynamics using short-term data from MD simulations.…”

Section: Introductionmentioning

confidence: 99%

“…Moreover, a long short-term memory (LSTM) [17] neural networkbased ML model has been developed to effectively predict the complex dynamic growth of methane hydrates in terms of clathrate cages and relevant parameters using short-time MD information [18]. This is followed by the further development of a new model based on the eXtreme Gradient Boosting (XGBoost) ML algorithm [19]. Moreover, the authors have developed a novel method to predict multi-atom cooperative phenomena at atomic scale based on a deep generative model in combination with recurrent neural networks [20].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Predicting long-term trends in physical properties from short-term molecular dynamics simulations using long short-term memory

Noda,

Shibuta

2024

J. Phys.: Condens. Matter

View full text Add to dashboard Cite

This study proposes a novel long short-term memory (LSTM)-based model for predicting future physical properties based on partial data of molecular dynamics (MD) simulation. It extracts latent vectors from atomic coordinates of MD simulations using graph convolutional network, utilizes LSTM to learn temporal trends in latent vectors and make one-step-ahead predictions of physical properties through fully connected layers. Validating with MD simulations of Ni solid-liquid systems, the model achieved accurate one-step-ahead prediction for time variation of the potential energy during solidification and melting processes using residual connections. Recursive use of predicted values enabled long-term prediction from just the first 20 snapshots of the MD simulation. The prediction has captured the feature of potential energy bending at low temperatures, which represents completion of solidification, despite that the MD data in short time do not have such a bending characteristic. Remarkably, for long-time prediction over 900 ps, the computation time was reduced to 1/700th of a full MD simulation of the same duration. This approach has shown the potential to significantly reduce computational cost for prediction of physical properties by efficiently utilizing the data of MD simulation.

show abstract

Section: Time Evolution Prediction Of Physical Properties Based On Lstmmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Predicting long-term trends in physical properties from short-term molecular dynamics simulations using long short-term memory

Noda,

Shibuta

2024

J. Phys.: Condens. Matter

View full text Add to dashboard Cite

show abstract

“…More recently, particle-type additives have been highlighted as KHPs, which provide heterogeneous hydrate nucleation sites and allow efficient heat removal owing to their high thermal conductivities. Metal nanoparticles (silver and copper) , and nanocarbons (carbon nanotubes and graphene) , have been tested as KHPs. The carbon materials play a seeding role in the formation of water structures such as gas hydrate and ice nuclei, inducing the formation of gas hydrates .…”

Section: Introductionmentioning

confidence: 99%

Promoting CH₄ Hydrate Formation Kinetics by Utilizing Synergistic Promotion Effects of Graphite Particles with Sodium Dodecyl Sulfate

Kim,

2024

Energy Fuels

View full text Add to dashboard Cite

Solidified natural gas in the form of hydrates, ice-like crystalline compounds formed with water and gas, has advantages of high storage capacity, low explosion risk, and environment friendliness. However, practical applications of hydrates have been hindered by the uncertainty of nucleation and slow formation kinetics required for reliable and economic operation. Kinetic hydrate promoters (KHPs) such as surfactants and amino acids have been studied to enhance hydrate formation kinetics, but the rate and amount of gas uptake upon hydrate formation are still insufficient. Recently, graphite particles (GPs) were proposed as KHPs providing heterogeneous hydrate nucleation sites and allowing efficient heat removal for rapid hydrate growth. However, the combined use of GPs with conventional KHPs to achieve superior hydrate formation kinetics has not yet been investigated. Here, we report the synergistic promotion effect of GPs with sodium dodecyl sulfate (SDS) on CH4 hydrate formation kinetics. Under nonstirred conditions, while the use of GPs alone did not lead to hydrate formation, adding GPs to the systems with SDS and L-tryptophan (l-Trp) greatly increased the CH4 hydrate formation rate. The combined use of GPs and SDS rather than l-Trp achieved the fastest kinetics by enhanced mass transfer with the spontaneous formation and falling of GPs + hydrate aggregates. Under stirred conditions, the extensive surface area of GPs provides abundant heterogeneous nucleation sites and the combined use of GPs and SDS therefore resulted in a 350% increase in CH4 gas uptake rate upon hydrate formation compared to the pure water system. The synergistic effect observed between GPs and SDS was hypothesized by the reduced energy barrier for hydrate nucleation, supported by hydrate onset temperature and surface tension measurement data. GPs will be applicable to practical applications of hydrates by further revealing their role in hydrate formation and maximizing the synergistic promotion effects with other KHPs.

show abstract

Effects of shear loading rate on tetrahydrofuran Hydrate Adhesion strength for enhanced flow assurance

Lin,

Yan,

Zhang

et al. 2024

Engineering Fracture Mechanics

View full text Add to dashboard Cite

Enhanced formation of methane hydrates via graphene oxide: Machine learning insights from molecular dynamics simulations

Cited by 12 publications

References 67 publications

Predicting long-term trends in physical properties from short-term molecular dynamics simulations using long short-term memory

Predicting long-term trends in physical properties from short-term molecular dynamics simulations using long short-term memory

Promoting CH₄ Hydrate Formation Kinetics by Utilizing Synergistic Promotion Effects of Graphite Particles with Sodium Dodecyl Sulfate

Effects of shear loading rate on tetrahydrofuran Hydrate Adhesion strength for enhanced flow assurance

Contact Info

Product

Resources

About

Enhanced formation of methane hydrates via graphene oxide: Machine learning insights from molecular dynamics simulations

Cited by 12 publications

References 67 publications

Predicting long-term trends in physical properties from short-term molecular dynamics simulations using long short-term memory

Predicting long-term trends in physical properties from short-term molecular dynamics simulations using long short-term memory

Promoting CH4 Hydrate Formation Kinetics by Utilizing Synergistic Promotion Effects of Graphite Particles with Sodium Dodecyl Sulfate

Effects of shear loading rate on tetrahydrofuran Hydrate Adhesion strength for enhanced flow assurance

Contact Info

Product

Resources

About

Promoting CH₄ Hydrate Formation Kinetics by Utilizing Synergistic Promotion Effects of Graphite Particles with Sodium Dodecyl Sulfate