ACS Publications. Most Trusted. Most Cited. Most Read
Eleven NanoHUB Simulation Tools Using RASPA Software To Demonstrate Classical Atomistic Simulations of Fluids and Nanoporous Materials
More

OR SEARCH CITATIONS

My Activity
Publications

Figure 1Loading Img
Download Hi-Res ImageDownload to MS-PowerPointCite This:ACS Omega 2022, 7, 48, 44470-44484
  • Open Access
ADDITION/CORRECTION. This article has been corrected. View the notice.
Article

Eleven NanoHUB Simulation Tools Using RASPA Software To Demonstrate Classical Atomistic Simulations of Fluids and Nanoporous Materials
Click to copy article linkArticle link copied!

Open PDF

ACS Omega

Cite this: ACS Omega 2022, 7, 48, 44470–44484
Click to copy citationCitation copied!
https://doi.org/10.1021/acsomega.2c06978
Published November 23, 2022

Copyright © 2022 The Authors. Published by American Chemical Society. This publication is licensed under

CC-BY-NC-ND 4.0 .

Abstract

Click to copy section linkSection link copied!
High Resolution Image
Download MS PowerPoint Slide

Eleven interactive simulation tools were created on nanoHUB to help users learn how to perform classical atomistic simulations. These tools enable users to perform classical Monte Carlo and molecular dynamics simulations using RASPA software. These tools use comparatively small numbers of production cycles to keep the runtimes short, so that users will not be discouraged by long wait times to see results. Here, we show that these tools produce results of sufficient accuracy and reproducibility for learning purposes. The 11 tools developed were as follows: (1) calculation of the self-diffusion constant of gas molecules in metal–organic frameworks (MOFs), (2) gas adsorption in MOFs using the grand canonical ensemble, (3) Henry’s coefficient calculator for gas molecules in MOFs and a zeolite, (4) adsorption of a gas mixture in a MOF, (5) self-diffusion of a gas mixture in a MOF, (6) void fraction calculation for several MOFs and zeolites, (7) surface area calculation for several MOFs and zeolites, (8) calculation of radial distribution function and self-diffusion constant for several pure gases, (9) energy distribution of adsorption sites using a probe molecule in MOFs, (10) molecular dynamics simulation of pure fluids in the NPT ensemble, and (11) gas adsorption in MOFs using the Gibbs ensemble.

This publication is licensed under

CC-BY-NC-ND 4.0 .
  • cc licence
  • by licence
  • nc licence
  • nd licence
Copyright © 2022 The Authors. Published by American Chemical Society

Subjects

what are subjects

1. Introduction

Click to copy section linkSection link copied!
In this work, we developed and deployed tools on nanoHUB with capabilities to perform atomistic simulations and to output results to a graphical user interface (GUI) using RASPA software package on the backend. NanoHUB is an online platform for science and engineering originally funded by the National Science Foundation (NSF). It contains simulation tools and resources developed and contributed by members of the community. (1) It promotes the advancement of nanoscience and nanotechnology and professional collaboration. NanoHUB is a learning resource for both undergraduate and graduate students in science and engineering. (2) NanoHUB is a product of the Network for Computational Nanotechnology, which aids research efforts in the area of nanoscience and nanotechnology. NanoHUB can be accessed through the web portal nanohub.org. It contains over 500 simulation tools and has over 2500 citations in the scientific literature. (3) NanoHUB users run these simulation tools in their web browser as applets, while the actual calculations in these simulations are run transparently on the nanoHUB backend cloud Linux cluster.
RASPA software performs classical atomistic simulations of adsorption and diffusion of molecules in nanoporous materials such as carbon nanotubes, metal–organic frameworks (MOFs), and zeolites. (4) It can also be used to compute single-phase properties as well as phase equilibrium properties such as vapor–liquid equilibrium. It uses cutting-edge algorithms to perform molecular dynamics and Monte Carlo calculations in various thermodynamic ensembles. (4) Monte Carlo and molecular dynamics methods can be combined to form more advanced simulations for specialized applications such as gas adsorption in flexible MOFs. (5)
In this work, we created interactive nanoHUB educational modules that demonstrate basic inputs, outputs, and methods for classical atomistic simulations using RASPA software. These simulation tools should catalyze the interest of new students and researchers to study diffusion constants, adsorption isotherms, and fluid properties by educating them on how to use RASPA to perform these types of calculations. To the best of our knowledge, this work is the first time RASPA has been deployed within nanoHUB tools.
Our 11 new nanoHUB simulation modules do not involve installing any software on the user’s computer. They are web applets accessed and run through a normal web browser. The user enters all inputs by selecting options in a GUI that is located within a webpage on the nanoHUB.org website, and all of the simulation module’s outputs are returned to the user through this same GUI running as a web applet within the webpage. In effect, these simulation modules appear to the user as interactive webpages. The only requirements for running these simulation tools is that the user must have any web browser installed on their computer, and they must sign up for a free nanoHUB.org login username. These simulation tools do not require RASPA to be installed on the user’s computer because the RASPA simulations are run via the cloud on nanoHUB’s own computing cluster.
These 11 new nanoHUB simulation modules do not collect any information from the user or from the user’s computer, except which menu options have been selected by the user for the purposes of running the simulation. Separately, nanoHUB.org may collect certain user information as part of their privacy policy described at http://nanohub.org/legal/privacy. We reiterate that information is collected by and accessible to nanoHUB.org and not to us as the tool developers. As tool developers, we do not collect any information from or about the user or from the user’s computer.
During review of this article, we received an inquiry about whether it is possible to install these modules as a “standalone software”. The answer is that these modules are run as interactive webpages on the nanoHUB.org website. They are not a software program that could be installed on a user’s computer.
These 11 new nanoHUB simulation modules are intended as educational tools, not as research tools. They show users how to set up and perform these types of simulations and to interpret the results. To simplify both their use and testing, these simulation modules were deliberately constructed with limited menu options. Because these menu options are very limited, these simulation tools are unsuitable for scientific research calculations. Each simulation tool has an online documentation that includes a PowerPoint presentation that briefly summarizes the choices of inputs, the classical forcefields used, the simulation methods, and how to interpret the simulation results. The online documentation also includes an example RASPA input file for each tool; this helps each user learn how to set up a RASPA calculation in case they want to later set up their own independent RASPA simulations for research purposes. In this way, we hope to promote the wider use and understanding of classical atomistic calculations for simulating gas adsorption and diffusion in porous materials and for simulating fluids. We envision the following use cases.
Use case # 1: In a college course on computational chemistry, the instructor could perform in-class demonstrations using these tools. For example, the instructor could run an in-class molecular dynamics simulation using tool 1 or tool 5 to show students how to compute gas self-diffusivities in porous materials. This helps students learn four important aspects: (a) What are the required inputs for this calculation type? (b) What computational methods and thermodynamic ensemble (if any) are used? (c) What are the raw outputs from the calculation? and (d) How are these outputs analyzed to compute the desired properties? For example, graphical analysis is performed to extract the self-diffusivity from the mean-squared-displacement (MSD) versus time data. Furthermore, the instructor could assign homework problems that use one or more of these simulation tools.
A key benefit is that these tools can be used without requiring users to install RASPA software. When teaching a course, it can be challenging to require students to install new software. First, some students in the class might not own a personal computer and instead use various on-campus computer labs. Second, different operating systems and versions present challenges to software installation. All of these difficulties are avoided because these simulation tools do not require installing any software beyond a normal web browser that is already available on the user’s computer.
Variations of use case # 1 also include seminars, workshops, and short courses that last only 1–3 days rather than a full academic semester. These sometimes occur at conferences or other venues that bring together people from various industries, universities, and/or countries.
Use case # 2: A person can teach themselves the basics of classical molecular dynamics and Monte Carlo simulations by using these simulation tools. Many people who do not have access to a dedicated course on computational chemistry may wish to learn about the topic through self-study. Some of these users may not have access to high performance computing clusters. Because all of these simulations are run via the cloud on nanoHUB’s own computing cluster, a person who does not have access to high performance computing can still easily run them.
Use case # 3: These simulation tools can facilitate initial training of graduate students that will subsequently progress to doing research calculations using RASPA. In this case, the tools demonstrate example inputs, settings, outputs, and data analysis for common calculation types. For research calculations, graduate students can then run simulations for other molecules and materials using the standalone RASPA program, which does not involve our simulation tools in any way. Typically, the standalone RASPA program would be installed on a Linux cluster used by the research group. Research calculations should use more initialization, equilibration (for molecular dynamics), and production cycles than what is used in our simulation tools because the numbers of cycles are purposefully kept low in our simulation tools to allow them to complete faster for demonstration purposes.

2. Methods

Click to copy section linkSection link copied!

2.1. Classical Forcefields Used

Forcefields are functional forms with parameters used to specify the potential energy of a system of atoms. (6−8) The basic functional form of potential energy in molecular mechanics involves the bonded and non-bonded interactions which can be expressed mathematically as
U=Ubonded+Unonbonded
(1)
Ubonded=UR+Uθ+Uφ+Uω
(2)
Unonbonded=Uvdw+Uel
(3)
where UR is the bond stretch interaction, Uθ is the bond angle bending, Uφ is the dihedral angle torsion, Uω is the inversion term, Uvdw is the van der Waals energy (e.g., Lennard-Jones potential, which includes short-range exchange repulsion and long-range attractive dispersion interactions), and Uel is the electrostatic energy. (9) The bonded terms are used to specify interactions between atoms linked by a covalent bond, while the non-bonded interactions, on the other hand, consist of the electrostatic term and the van der Waals interaction.
To simplify computations, the non-bonded terms are often limited to pairwise interactions. The Lennard-Jones potential can be used as a model for the van der Waals interactions (10,11)
VLJ=4ε[(σr)12−(σr)6]
(4)
where ε is the depth of the potential well, σ is the finite distance at which the inter-particle potential is zero, and r is the distance between the two atoms or molecules. A typical potential energy curve for two interacting molecules is shown in Figure 1. rmin is the distance at which the potential reaches its minimum and is related to σ by
rmin=21/6σ
(5)

Figure 1

Figure 1. Lennard-Jones potential for argon. The energy unit is in Kelvin (used by RASPA); this can be multiplied by the Boltzmann constant to convert to Joules per particle.

High Resolution Image
Download MS PowerPoint Slide
Interactions between dissimilar non-bonded atoms require a combining rule for the part of the potential representing the van der Waals interaction. (12) For this project, the Lorentz–Berthelot (13,14) mixing rules were used, which assume the molecular size and energy parameters to be the arithmetic and geometric averages, respectively
σij=σii+σjj2
(6)
εij=εiiεjj
(7)
RASPA software comes with different forcefields built into the program. (4,15)Table 1 lists the different forcefields we used for various systems in the 11 simulation tools. For IRMOF-1 and IRMOF-16, the forcefield reported by Dubbeldam et al. was used keeping the framework rigid. (16) For the other MOFs, the GenericMOFs forcefield built into RASPA was used. (15) For the zeolites, the GenericZeolites forcefield built into RASPA was used. (15)
Table 1. List of Forcefields Used in the Different NanoHUB Tools
systemforcefieldflexiblereftools used in
ArLennard-JonesNa1, 4, 5, 11
H2(GenericMOFs)bN (21)1, 2, 9, 11
N2TraPPEN (20)1, 2, 8, 9, 11
O2TraPPEN (22)2
CO2TraPPEN (20)1, 2, 8, 9, 10, 11
CH4TraPPE-UAN (19)1, 2, 4, 5, 7–11
C2H6TraPPE-UAY (19)8, 9, 10
C3H8TraPPE-UAY (19)8, 10
C4H10TraPPE-UAY (19)8
n-pentaneTraPPE-UAY (19)3
n-hexaneTraPPE-UAY (19)3
n-heptaneTraPPE-UAY (19)3
n-octaneTraPPE-UAY (19)3
n-nonaneTraPPE-UAY (19)3
IRMOF-1DubbeldamNc (16)1–7, 9, 11
IRMOF-2(GenericMOFs)dN (9,23)7
IRMOF-3(GenericMOFs)dN (9,23)7
IRMOF-12(GenericMOFs)dN (9,23)7
IRMOF-16DubbeldamNc (16)1–3, 6, 7, 9, 11
ITQ-1GenericZeolitesNe6, 7
ITQ-3GenericZeolitesNe6, 7
ITQ-7GenericZeolitesNe6, 7
ITQ-12GenericZeolitesNe6, 7
ITQ-29GenericZeolitesNe6, 7
KFIGenericZeolitesNe6, 7
LTA4AGenericZeolitesNe6, 7
LTA5AGenericZeolitesNe6, 7
MFI_SIGenericZeolitesNe3
a

Lennard-Jones parameters for argon were set to σ = 3.34 Å and ε = 119.8 K to reproduce the well depth and rmin = 21/6 σ value for the dimer curve.

b

These parameters for the H2 molecule are part of the GenericMOFs forcefield collection in RASPA.

c

This forcefield included flexibility but it was not used.

d

IRMOF-2, IRMOF-3, and IRMOF-12 used generic atom types and parameters from the UFF and Dreiding forcefields, as compiled within the GenericMOFs forcefield collection in RASPA.

e

The GenericZeolites forcefield in RASPA uses Lennard-Jones parameter values collected from different sources: (a) for Si and O atoms, the GenericZeolites forcefield uses TraPPE-zeo (ref (24)) Lennard-Jones parameter values, (b) for Al atoms, the GenericZeolites forcefield used the same Lennard-Jones parameter values as for Si, (c) although these atoms are not present in any of the structures here, for Na and Ca atoms, the GenericZeolites forcefield uses UFF (ref (9)) Lennard-Jones parameter values.

We used Transferrable Potential for Phase Equilibria (TraPPE) forcefields to describe the molecules. TraPPE forcefields are popular and suitable for research and industrial applications because they have reasonably high accuracy for predicting thermophysical properties of various compounds across different physical states and compositions. (17−20) TraPPE forcefields are available in RASPA in two main flavors: TraPPE-United Atom (TraPPE-UA) model and TraPPE-Explicit Hydrogen (TraPPE-EH) model. The TraPPE-UA forcefield is a united atom representation for the alkyl groups; that is, each hydrogen atom is modeled implicitly in the same interaction site as the carbon atom it is bonded to. (19) The TraPPE-EH forcefield has a separate interaction site for each hydrogen atom. (17)

2.2. Monte Carlo, Molecular Dynamics, and Widom Insertion Simulations in RASPA Software

RASPA is a classical atomistic simulation software. (4,15) In this work, we used RASPA version 2 (i.e., RASPA2). (15) RASPA uses state-of-the-art algorithms to perform molecular dynamics and Monte Carlo calculations in various statistical ensembles. (4,25) Monte Carlo and molecular dynamics methods can be combined to form more advanced simulations for specialized applications such as adsorption on a swelling material, protein studies, lipids studies, and so on. (4) RASPA can perform simulations with flexible frameworks, (4,7,26,27) but to minimize the computational time, all frameworks in the nanoHUB tools described in this work were held rigid.
To perform a simulation in RASPA, certain parameters can be specified in the input file (15)

  • The simulation type (e.g., Monte Carlo or Molecular Dynamics)

  • The number of initialization, equilibration, and production cycles

  • The molecules and/or framework that make up the system to be simulated

  • The forcefields and mixing rules to be used

  • If a framework is included, whether framework flexibility is allowed or the framework is to be held rigid

  • For Monte Carlo calculations, the different kinds of trial moves and their probabilities

  • Keywords to specify various quantities to be computed and printed

  • Depending on the simulation type, the external temperature and/or pressure may need to be specified

  • The cutoff radius for interactions between atoms

RASPA2 uses the minimum image convention. The minimum image convention assumes that each atom type “A” in a unit cell only interacts with the closest image of each atom type “B”. (25) Also, interactions are neglected between atoms that are farther apart than the cutoff radius. (25) The cutoff radius must be less than half the shortest length of the unit cell.
Simulations that use a single probe particle test insertion (not a thermodynamic ensemble) do not need initialization cycles. This type of move is usually called Widom insertion. (4) Each Widom insertion move starts with no molecules present in the framework and then attempts the test insertion of a probe molecule. Even for this kind of test insertion, RASPA grows the internal configuration of flexible molecules using configurational bias Monte Carlo (CBMC). (4) In this work, the test insertion of a probe molecule was used to compute surface areas, void fractions, and Henry’s coefficients.
Monte Carlo algorithms use random sampling and inferential statistics to estimate the value of an unknown quantity. (25,28,29) In RASPA, one Monte Carlo step refers to a singular trial move (e.g., translation, rotation, swap, reinsertion, etc.), whereas a Monte Carlo cycle refers to the maximum between 20 move attempts and the number of particles that make up the system. (4,25) Each attempted Monte Carlo trial move has the following sequence. First, the type of move to attempt (i.e., translation, rotation, reinsertion, etc.) is randomly selected based on their assigned probabilities. Second, a trial move of the selected type will be prepared. For a translation or rotation move, the particle on which the move acts is randomly selected, and the direction of the translation or rotation is also randomly selected. If the move is a volume change, then a particle does not need to be selected, but the size of the volume change needs to be randomly selected. Third, the energy change of the system that would be caused by the trial move is computed. Fourth, acceptance criteria are used to determine whether to accept or reject the trial move. (28,29) If the trial move is accepted, then the modified structure which includes the trial move forms the next system configuration in the Markov chain. If the trial move is rejected, then the unmodified structure which does not include the trial move forms the next system configuration in the Markov chain; in other words, the new configuration retains the configuration of the previous move.
For a simple translation or a simple rotation of a molecule, the acceptance criterion is related to the Boltzmann factor exp[−ΔE/(kT)] computed for the attempted move. ΔE is the energy of the new configuration minus the energy of the old configuration, k is the Boltzmann constant, and T is the absolute temperature of the system. For a simple translation or a simple rotation of a molecule, the Boltzmann factor expresses the relative probability that the attempted move will be accepted. This can be implemented using the Metropolis–Hastings algorithm (30,31) in which a random number is generated from a uniform distribution between 0 and 1. If the Boltzmann factor is less than the random number, the move is rejected. If the Boltzmann factor is greater or equal to the random number, the move is accepted.
The acceptance criteria for more complicated moves (e.g., volume changes, swaps, reinsertions, etc.) is more complicated and depends on the type of biasing (if any) that is used. A detailed discussion of acceptance criteria for these moves can be found in the literature. (28,29)
Monte Carlo moves used by one or more of our 11 nanoHUB simulation tools included the following: (4,15,25)

  • Translation: This move stochastically displaces (translates) a random molecule. The internal configuration of the molecule remains unchanged.

  • Rotation: This move randomly selects a molecule and then randomly rotates it about the chosen starting bead. The internal configuration of the molecule remains unchanged. When a molecule is modeled as a sphere, then rotation is not needed. For example, rotation is not needed when running a Monte Carlo simulation of methane using the united-atom model, which models methane as a sphere.

  • Swap: This move randomly attempts to insert or delete a random molecule from the system. An insertion is attempted 50% of the time, and a deletion is attempted the other 50% of the time.

  • Reinsertion: “A full reinsertion move for the current component.” (15) “The ‘reinsertion’ move removes a randomly selected molecule and reinserts it at a random position. For rigid molecules, it uses orientational biasing, and for chains, the molecule is fully regrown (the internal configuration is modified).” (25) “Multiple first beads are chosen, and one of these is selected according to its Boltzmann weight. The remaining part of the molecule is grown using biasing. This move is very useful, and often necessary, to change the internal configuration of flexible molecules.” (15)

  • Volume change: This move isotopically changes the volume of the simulation box.

  • Gibbs ensemble calculations use two boxes. (28,29) In these computations, a “Gibbs swap” moves a molecule from one box to the other; a new internal configuration is regrown for the molecule using CBMC in the second box.

In RASPA, every Monte Carlo simulation is first initialized at the beginning of the simulation. The initialization cycles warm up the Markov chain. (4,15,30) Data are collected during the production cycles that follow the initialization cycles. Monte Carlo methods can be applied to a variety of molecular studies such as adsorption calculations, (32) vapor–liquid equilibrium calculations, (33) chemical reactions, (34) and so on.
Molecular dynamics calculates the time-dependent behavior of a system. The trajectories of atoms and molecules are calculated numerically by solving Newton’s equations of motion for interacting particles. (28,29,35) The potential energy and interaction forces between atoms and molecules are calculated using the forcefield. The timestep is the time interval between adjacent configurations. A 0.5–1.0 femtosecond timestep is appropriate for molecules containing light moveable atoms; a somewhat larger timestep can be used for rigid molecules or if all of the atoms are heavy or if the forcefield is coarse-grained. (36) (We used a timestep of 0.5 femtosecond per MD cycle in tools # 1, 5, 8, and 10.) Three types of cycles are used in a molecular dynamics calculation: initialization, equilibration, and production. In RASPA, initialization cycles are performed at the beginning using Monte Carlo moves to get a statistically relevant configuration of particle positions. Then, equilibration cycles are performed using molecular dynamics moves to get a statistically relevant configuration of particle velocities and positions. Finally, data are collected during the production cycles. Figure 2 shows the types of cycles involved in Monte Carlo and molecular dynamics simulations.

Figure 2

Figure 2. Monte Carlo and molecular dynamics cycles.

High Resolution Image
Download MS PowerPoint Slide
Molecular dynamics can be used to study diffusion coefficients (both self-diffusion and collective diffusion), (37,38) interactions between different molecules, (39) material characteristics, (40) protein folding, (41) and so on. In this work, we computed self-diffusion constants rather than collective (also called transport or Fickian) diffusion constants. (38)
Common thermodynamic ensembles used for Monte Carlo and molecular dynamics include the following: (25,28,29)

  • Microcanonical ensemble (NVE)─This statistical ensemble holds the number of particles N, the volume V, and the energy E constant during the course of the simulation.

  • Canonical ensemble (NVT)─This ensemble holds the number of particles N, the volume V, and the average temperature T constant.

  • Grand Canonical ensemble (μVT)─This ensemble holds the chemical potential μ, the volume V, and the average temperature T constant.

  • Isobaric–isothermal ensemble (NPT)─This ensemble holds the number of particles N, the average pressure P, and the average temperature T constant.

  • Isoenthalpic–isobaric ensemble (NPH)─This ensemble holds the number of particles N, the average pressure P, and the enthalpy H constant.

2.3. Creating NanoHUB Tools That Run RASPA Simulations

Figure 3 depicts the process flow diagram of nanoHUB tool creation. The nanoHUB infrastructure used to develop tools in this project include the workspace, Rappture toolkit, middleware, and repository.

Figure 3

Figure 3. Flow diagram illustrating the process to create a new nanoHUB tool.

High Resolution Image
Download MS PowerPoint Slide
As shown in Figure 4, the workspace is an in-browser Linux command line terminal that provides access to computational resources on the Network for Computational Nanotechnology. This workspace can be used to develop code, test, compile, and debug new simulation tools before they are ready to be deployed on nanoHUB. A new developer of nanoHUB simulation tools must submit a special request to the support team to get permission to access this workspace. The nanoHUB support team also installed the RASPA software on the backend computing cluster.

Figure 4

Figure 4. NanoHUB workspace.

High Resolution Image
Download MS PowerPoint Slide
Rappture is an acronym for “rapid application infrastructure”. The Rappture toolkit was deployed inside the workspace by typing “rappture -builder” (without quotes) at the Linux command line prompt. Rappture facilitates nanoHUB tool development using drag and drop functionality. It makes tool development a quick and easy process by allowing users to drag and drop from the “Object Types,” as shown in Figure 5, to the “Tool Interface” input and output. This allows the developer to add input fields such as dropdown menus, radio buttons, text boxes, and so forth and output fields such as text boxes, graphs, and so forth to the GUI. This was used to develop the GUI for each tool.

Figure 5

Figure 5. Rappture builder GUI.

High Resolution Image
Download MS PowerPoint Slide
Developing a GUI is the first half of building a simulation tool; the other half is writing code within the simulator to access inputs, run calculations, and generate results. Rappture generates two skeleton files to which code is manually added by the developer. The first generated file is tool.xml. This is an extensible markup language file that contains all the input and output fields selected and predefined in the Rappture builder during the process of creating the GUI. The other file generated by Rappture builder is “main.ext”; where “ext” is the extension of the file based on the chosen programming language (e.g., main.py for Python). This file acts as the main program in the simulator. This file contains commands that retrieve the input data from the GUI, generates the RASPA input file, submits the calculation in RASPA, extracts the results from the RASPA output files, and finally sends back the results to the GUI. After these two files (i.e., tool.xml and main.py) are auto generated, the developer opens each of these two files in a text editor and manually adds the desired code lines to perform the different tasks they want. For this project, the main.py script creates the RASPA input file titled “simulation.input” based on the user input specified in the GUI. Then main.py submits the RASPA job and extracts the relevant data from the RASPA output files and returns these to the GUI. Umeh’s master’s thesis contains several examples of the tool.xml, main.py, and simulation.input files. (42)
After making necessary changes to both the tool.xml and main.py files, the tool is ready to be registered on nanoHUB. Tool registration requires that the developer provide information on the tool name, repository host [e.g., Subversion (SVN) or Github], whether the tool will be public or private, licensing options, and so forth. It is recommended that SVN be used as the repository host for developers working on nanoHUB. SVN tracks and saves all changes made to each file and can revert to a previous version if so desired.
Once the tool is registered, it automatically triggers the creation of a repository for the tool. Figure 6 shows the TortoiseSVN repository browser. After the repository is created, the developer must manually upload the main.py and tool.xml files to the repository.

Figure 6

Figure 6. TortoiseSVN repository browser.

High Resolution Image
Download MS PowerPoint Slide
In the newly created repository, the middleware folder contains a file named “invoke” that acts as the interface between the tool.xml and main.py files and the RASPA software. Middleware is the software layer that lies between the operating system and the applications on each side of a distributed computing system in a network. To set up the invoke file, all we had to specify was the simulation tool name (e.g., rdf for the radial distribution function tool) and the RASPA software pathname, and the nanoHUB middleware handles all the interfacing details. For further details, see the master’s thesis of Umeh. (42) This invoke file must be made executable.
After the source code has been uploaded and the middleware file edited, the tool is ready to be installed, tested, and published. Clicking the hyperlink “I’ve committed new code. Please install the latest version for testing and approval.” on the nanoHUB tool page alerts support and lets them know that the tool is ready for installation. After the tool is installed, tests should be performed to ensure that it runs as expected. If there are needs for changes to the tool, the changes should be made to the source code on the developer’s local repository on their PC and the source code should be recommitted to the repository and the new code installed. This cycle should be repeated until the developer is satisfied with the performance of the tool. Finally, once the tool is working to the satisfaction of the developer, the developer can approve the tool for publishing.
In the tools developed, nanoHUB typically saves the results of the simulation in a cache such that if a simulation is later rerun with exactly the same input options as before, it will display the already computed results rather than redo the same calculation. Because molecular dynamics and Monto Carlo simulations in RASPA use a random number generator, we desire that a new set of randomly generated numbers be used for each replicate run in RASPA. To ensure that a new set of randomly generated numbers is always used rather than displaying nanoHUB’s cached results of a prior calculation, we included a line of code in main.py that purges the simulation tool’s output cache at the beginning of each nanoHUB simulation.

2.4. Computing a Molecule’s Self-Diffusivity

Tools # 1, 5, and 8 calculate and display the self-diffusion constant Ds using Einstein’s equation (29)
MSD=6Dst+C
(8)
where C is the fitted intercept, and MSD is the mean-squared displacement of a molecule during time t. eq 8 describes the relationship between MSD and time in the diffusive regime.
As shown in Figure 7, a plot of MSD versus time can be separated into the following regimes. (38,43) When the time elapsed is much shorter than the mean free time between particle collisions, this kind motion is called ballistic transport (aka “ballistic regime”). In the ballistic regime, log[MSD] versus log[t] has a slope of approximately 2 because each individual particle (e.g., individual molecule) tends to travel at approximately constant speed between collisions, even though the speeds of different particles will be different. The diffusive regime occurs when the time elapsed is much longer than the mean free time between particle collisions. In this case, the trajectory of each particle (i.e., each molecule) resembles a random walk (e.g., Brownian motion) that was studied by Einstein to derive eq 8. Between the ballistic and diffusive regimes is a transition regime in which the time elapsed is approximately similar to the mean free time between collisions.

Figure 7

Figure 7. Left panel: A log–log plot of mean-squared displacement versus time that shows the separate ballistic, transition, and diffusive regimes. Right panel: A linear–linear plot of mean-squared displacement versus time using the starting and ending times selected by the algorithm that extracts the self-diffusion constant Ds from a data subset within the diffusive regime.

High Resolution Image
Download MS PowerPoint Slide
To compute Ds, we used the following algorithm to find a segment of the MSD versus time curve within the diffusive regime. For each time tj, a moving average slope of log[MSD] versus log[t] was computed as
avg_slopej=log[MSD(j+10)]−log[MSDj]log[t(j+10)]−log[tj]
(9)
This average slope starts with value 2.00 in the ballistic regime. As tj increases into the transition regime, avg_slopej changes until it reaches a value of ∼1 near the start of the diffusive regime. Now, the cutoff of how close avg_slopej must be to 1.0 to mark the start of the diffusive regime is a judgment call. In this work, the following heuristic algorithm was used to define the start of a curve segment for fitting within the diffusive regime:
1.

Loop over j = 1, 2 to (N-20), where N is the total number of datapoints. (The 20 points at the end are reserved to ensure that there will always be a large enough curve segment remaining for us to perform an adequate fit to extract the Ds.) During each iteration of this loop, if avg_slopej >1.25, then set k = j.

2.

Ds is computed as

Ds=log[MSD(k+15)]−log[MSD(k+5)]log[t(k+15)]−log[t(k+5)]
(10)
When N is sufficiently large, step # 1 above sets k equal to the last point within the transition regime for which avg_slopej > 1.25 such that avg_slope(k+1) ⩽ 1.25. However, if N is relatively small, then k may equal N-20, which is less than ideal but still ensures that enough datapoints are reserved to perform a fit to compute Ds. Step # 2 above then performs the fit starting with index (k + 5) and continuing to index (k + 15). The rational is that once avg_slopej drops below 1.25 [i.e., index value (k + 1)], then adding +4 to the index [i.e., (k + 5)] should have the effect of choosing a starting point with a slope near 1. For example, Figure 7 shows the first of three runs at the San Diego Supercomputing Center that was used to compute the self-diffusivity of N2 in IRMOF-16 at 298 K, as listed in Table 2. Using this algorithm, the selected start time for the fit of eq 8 had avg_slope = 1.10, which is sufficiently close to 1 to give a meaningful Ds.
Table 2. Self-Diffusion of N2 in IRMOF-16 at 298 K Was Computed Using Tool 1
 initialization cyclesequilibration cyclesproduction cyclesself-diffusion (Å2/picosecond)
nanoHUB5000500030,00011 ± 2
SDSC50,00050,000250,00011.1 ± 0.6

3. Simulation Tool Descriptions

Click to copy section linkSection link copied!
Note: The graphs of molecules versus number of cycles in tools # 2, 4, and 11 were prepared as follows. The production cycles were divided into five blocks of equal sizes, and the average number of molecules in each phase was computed by RASPA for each block. For each phase, this gives a total of five data points that correspond to the average number of molecules during 0–20% (aka block 1), 20–40% (aka block 2), 40–60% (aka block 3), 60–80% (aka block 4), and 80–100% (aka block 5) of the production cycles. For the mixed gas adsorption tool 4, separate molecules versus cycles plots are computed and displayed for methane and argon.

3.1. Gas Diffusion Coefficient in MOFs (Tool 1)

This tool calculates the self-diffusion constant of gas molecules through a MOF using molecular dynamics in the NVT ensemble. Gases used for this tool include argon, hydrogen (H2), nitrogen (N2), methane, and carbon dioxide. MOFs used for this tool are IRMOF-1 and IRMOF-16. The simulation temperature is 298 K. The user selects the gas and the MOF from drop lists. Sixteen of the chosen molecules are included in the simulation box along with the chosen MOF. To minimize the effects of periodic boundary conditions on the computed self-diffusion constant, the simulation box contains 27 (i.e., 3 × 3 × 3) MOF unit cells. The tool outputs a graph of the MSD versus time (t) in both linear–linear and log–log scales. This tool uses the algorithm described in Section 2.4 to calculate the gas molecule’s self-diffusion constant Ds. The value of Ds and the start and end times for the linear fit to compute Ds are displayed to the user. Figure 8 shows screenshots for this gas diffusion calculation tool.

Figure 8

Figure 8. Screenshots of Tool 1 for calculating the self-diffusion constant of a gas in a MOF.

High Resolution Image
Download MS PowerPoint Slide

3.2. Gas Adsorption Calculator (Tool 2)

This tool calculates the average absolute and excess adsorption of gas molecules in a MOF using Monte Carlo in the grand canonical ensemble (μVT). Gases used for this tool include hydrogen (H2), nitrogen (N2), oxygen (O2), methane, and carbon dioxide. MOFs used for this tool are IRMOF-1 and IRMOF-16. The user selects the gas and the MOF from drop lists. The user types in the temperature (in K) and pressure (in Pa) for the simulation. The tool outputs the average absolute adsorption, the average excess adsorption, and a graph of molecules vs cycles. This graph of molecules versus cycles helps demonstrate that the calculation has reached equilibrium. Figure 9 shows screenshots of this tool.

Figure 9

Figure 9. Screenshots of Tool 2 for calculating the absolute and excess adsorption of a gas in a MOF.

High Resolution Image
Download MS PowerPoint Slide

3.3. Henry’s Coefficients Simulator (Tool 3)

This tool uses Widom insertion to simulate Henry’s coefficients of n-alkane (n-pentane to n-nonane) for several nanoporous materials. To run a simulation, users are required to select one of the following nanoporous materials from a drop list: IRMOF-1, IRMOF-16, or the zeolite MFI_SI. The temperature used for the simulation is 573 K. Using a pre-calculated ideal gas Rosenbluth weight for each molecule, the tool outputs the Henry’s coefficients for the above-mentioned n-alkanes, as depicted in Figure 10. The procedure this tool uses to compute the Henry’s coefficient is analogous to an example described in the RASPA 2.0 manual. (15)

Figure 10

Figure 10. Screenshot of Tool 3 for calculating the Henry’s coefficients of n-alkanes in porous materials.

High Resolution Image
Download MS PowerPoint Slide

3.4. Mixed Gas Adsorption Calculator (Tool 4)

This tool simulates the adsorption of mixed gases (argon and methane) in IRMOF-1 using a Monte Carlo algorithm with the grand canonical ensemble (μVT). The GUI prompts the user to select the composition of the gas mixture from a drop list and then to enter the simulation temperature and pressure. Compositions available are (10 mol % argon + 90% methane) to (90 mol % argon + 10% methane) in 10% increments. This tool outputs the absolute and excess adsorption for the argon and methane components. The GUI also plots molecules versus the number of cycles separately for argon and methane to demonstrate how closely the calculation has approached equilibrium. Figure 11 shows the developed tool interface for this mixed adsorption calculation.

Figure 11

Figure 11. Screenshots of Tool 4 for calculating adsorption of a gas mixture in a porous material.

High Resolution Image
Download MS PowerPoint Slide

3.5. Mixed Gas Diffusion Calculator (Tool 5)

This tool simulates the self-diffusion of gas mixtures through a MOF using molecular dynamics with the NVT ensemble. To minimize the effects of periodic boundary conditions on the computed self-diffusion constants, the simulation box contains 27 (i.e., 3 × 3 × 3) MOF unit cells. The gases used in this tool are argon and methane, and the MOF used is IRMOF-1. The GUI prompts the user to select the composition of the gas mixture from a drop list and then to enter the simulation temperature. Compositions available are (10 mol % argon + 90% methane) to (90 mol % argon + 10% methane) in 10% increments. The tool outputs a graph of the MSD versus time (t) in both linear–linear and log–log scales separately for argon and methane. This tool applies the algorithm described in Section 2.4 to separately compute Ds for methane and argon. The Ds values are displayed to the user along with the start and end times for the linear fits used to compute these Ds values. Figure 12 depicts the developed tool interface for mixed gas diffusion calculation.

Figure 12

Figure 12. Screenshots of Tool 5 for calculating diffusion of a gas mixture in a porous material.

High Resolution Image
Download MS PowerPoint Slide

3.6. Void Fraction Calculator (Tool 6)

The void fraction calculator tool simulates the pore volume fraction of nanoporous materials. The nanoporous materials being calculated include both MOFs and zeolites: IRMOF-1, IRMOF-16, ITQ-1, ITQ-3, ITQ-7, ITQ-12, ITQ-29, KFI, LTA4A, and LTA5A. The tool requires the user to select (from a drop list) the material for which they want to calculate the void fraction. The void fraction is computed using Widom insertion of a helium atom, as described by Dubbeldam et al. (4)Figure 13 shows the developed tool interface for void fraction calculation.

Figure 13

Figure 13. Screenshot of Tool 6 for calculating the void fraction of a porous material.

High Resolution Image
Download MS PowerPoint Slide

3.7. Surface Area Calculator (Tool 7)

This tool simulates the surface area of nanoporous materials. Users have the options to select the type of material and the probe distance (either σ or rmin as illustrated in Figure 1). The materials used in this simulation include both MOFs and zeolites: IRMOF-1, IRMOF-2, IRMOF-3, IRMOF-12, IRMOF-16, ITQ-1, ITQ-3, ITQ-7, ITQ-12, ITQ-29, KFI, LTA4A, and LTA5A. This tool simulates surface area of the material using a single methane probe molecule at 298 K. Figure 14 shows the developed tool interface for surface area calculation.

Figure 14

Figure 14. Screenshot of Tool 7 for calculating the surface area of a porous material.

High Resolution Image
Download MS PowerPoint Slide

3.8. Radial Distribution Function Calculator (Tool 8)

Figure 15 shows the developed tool interface for radial distribution function calculation. The GUI allows the user to select (from a drop list) which gas is to be used for the simulation: methane, ethane, propane, butane, nitrogen (N2), or carbon dioxide. The GUI also allows the user to enter the simulation temperature. The simulation box is 30 Å × 30 Å × 30 Å and contains 100 molecules. Using molecular dynamics with the NVT ensemble, the tool performs the simulation and outputs a graph of the radial distribution function. It also outputs the density and self-diffusion graphs (i.e., MSD versus time for both linear–linear and log–log scales). This tool uses the algorithm described in Section 2.4 above to calculate the gas molecule’s self-diffusion constant Ds. The value of Ds and the start and end times for the linear fit to compute Ds are displayed to the user.

Figure 15

Figure 15. Screenshots of Tool 8 for calculating the radial distribution function of a gas.

High Resolution Image
Download MS PowerPoint Slide

3.9. Adsorption Energy Calculator (Tool 9)

This tool calculates the histogram of adsorption energies onto different sites within a MOF using Monte Carlo moves with one molecule per simulation box. No adsorbate–adsorbate interactions appear. This tool requires the user to select the kind of material to be used for the simulation, the gas molecule, and the temperature. Figure 16 shows the developed tool interface for adsorption energy calculation. The MOFs used for this tool include IRMOF-1 and IRMOF-16. The gas molecules are methane, ethane, hydrogen (H2), nitrogen (N2), and carbon dioxide. The tool outputs the energy histogram of the gas molecule as well as the average adsorption energy.

Figure 16

Figure 16. Screenshot of Tool 9 for calculating the adsorption energy histogram.

High Resolution Image
Download MS PowerPoint Slide

3.10. NPT Simulator (Tool 10)

This tool uses molecular dynamics with the isobaric–isothermal ensemble (NPT) to simulate the properties of gas molecules. To run a simulation, users are required to select the molecule from a drop list and enter the simulation temperature and pressure. The molecules used in the tool include methane, ethane, propane, and carbon dioxide. Fifty molecules of the user-selected compound are simulated inside the simulation box. Figure 17 shows the interface for this NPT simulation tool. The tool outputs the density and total energy of the gas at the given conditions.

Figure 17

Figure 17. Screenshot of Tool 10 for calculating the density and total energy of a gas in the NPT ensemble.

High Resolution Image
Download MS PowerPoint Slide

3.11. Gibbs Adsorption Simulator (Tool 11)

This tool simulates the adsorption of gas molecules onto a MOF. The difference between adsorption simulation using the Gibbs ensemble versus using the grand canonical ensemble (μVT) is that Gibbs ensemble adsorption uses two boxes for the simulation. The framework is contained in one box, and a gas phase is in the other box. The simulation contains a total of 100 molecules, and these can move between the two boxes. The gas-phase box can change volume. This simulation computes absolute adsorption using the forcefield without requiring a fugacity coefficient or an equation of state (Adsorption calculations using the grand canonical ensemble require a fugacity coefficient that is not computed self-consistently from the forcefield). The MOFs used for this tool are IRMOF-1 and IRMOF-16. The gas molecules include methane, argon, hydrogen (H2), nitrogen (N2), and carbon dioxide. This tool is similar to the grand canonical adsorption tool in terms of input parameters; that is, the user specifies a temperature and a pressure. The tool outputs the average absolute adsorption of the gas molecule, as shown in Figure 18. This tool also outputs a graph of molecules versus cycles in one box to show how closely the calculation has reached equilibrium.

Figure 18

Figure 18. Screenshots of Tool 11 for calculating gas adsorption in a MOF using the Gibbs ensemble.

High Resolution Image
Download MS PowerPoint Slide

4. Results and Discussion

Click to copy section linkSection link copied!

4.1. Convergence Tests

Each simulation tool was systematically tested for convergence. First, the menu options to be tested were randomly selected. For tools that had multiple choices of material, one material was selected at random. For tools that had choices of temperature, pressure, and/or composition as input, a random selection was made. Each of these nanoHUB tools has fixed numbers of initialization, equilibration, and/or production cycles that are used every time the tool runs. The numbers of these cycles are listed in Tables 2345678910.
Table 3. Absolute and Excess Adsorption of CO2 in IRMOF-1 at 308 K and 10 Bar Was Computed Using Tool 2 in the Grand Canonical Ensemblea
   tool 2 (grand canonical ensemble)tool 11 (Gibbs ensemble)
 initialization cyclesproduction cyclesabsolute adsorption (mol/kg framework)excess adsorption (mol/kg framework)absolute adsorption (mol/kg framework)
nanoHUB500010,0008.7 ± 0.28.0 ± 0.28.5 ± 0.2
SDSC25,00050,0008.56 ± 0.057.91 ± 0.058.85 ± 0.01
a

For comparison, the last column lists the absolute adsorption from Tool 11 in the Gibbs ensemble.

Table 4. Henrys’ Coefficients for n-Pentane to n-Nonane in MFI_SI Zeolite at 573 K Was Computed Using Tool 3a
 production cyclesC5C6C7C8C9
nanoHUB500030.2 ± 0.760 ± 1123 ± 1243 ± 10446 ± 13
SDSC25,00028.8 ± 0.257.83 ± 0.04116 ± 1223 ± 6410 ± 9
a

Units are 10–7 mol/(kg Pa).

Table 5. Absolute and Excess Adsorption of 40% Argon and 60% Methane Mixture at 385 K and 9 Bar Was Computed Using Tool 4
 initialization cyclesproduction cyclesargon absolute adsorption (mol/kg framework)argon excess adsorption (mol/kg framework)methane absolute adsorption (mol/kg framework)methane excess adsorption (mol/kg framework)
nanoHUB500010,0000.366 ± 0.0020.203 ± 0.0020.8849 ± 0.00130.6414 ± 0.0013
SDSC25,00050,0000.3654 ± 0.00050.203 ± 0.0010.883 ± 0.0020.640 ± 0.002
Table 6. Self-Diffusion Constant of 20% Argon and 80% Methane Mixture in IRMOF-1 at 303 K Was Computed Using Tool 5
 initialization cyclesequilibration cyclesproduction cyclesargon self-diffusion (Å2/picosecond)methane self-diffusion (Å2/picosecond)
nanoHUB50,00050,000100,0002.4 ± 0.93.7 ± 0.3
SDSC50,00050,000250,0003.4 ± 0.53.2 ± 0.5
Table 7. Void Fraction (Tool 6 for 300 K) and Surface Area (Tool 7 Using Sigma Probe Radius) of ITQ-29 Zeolite
 production cyclesvoid fractionsurface area (m2/g)
nanoHUB20000.4623 ± 0.0013758.9 ± 0.2
SDSC10,0000.4634 ± 0.0003759.26 ± 0.06
Table 8. Self-Diffusion Constant of Methane at 328 K and a Density of 98.66 kg/m3 Was Computed Using Tool 8
 initialization cyclesequilibration cyclesproduction cycleself-diffusion (Å2/ps)
nanoHUB50,00050,000100,00014.4 ± 0.2
SDSC50,00050,000250,00014.8 ± 0.7
Table 9. Adsorption Energy of CO2 in IRMOF-1 at 342 K Was Computed Using Tool 9
 initialization cyclesproduction cyclesadsorption energy (K)
nanoHUB500010,000–1173 ± 2
SDSC25,00050,000–1168 ± 2
Table 10. Properties of Methane at 350 K and 9 Bar Was Computed in the NPT Ensemble Using Tool 10
 initialization cyclesequilibration cyclesproduction cycledensity (kg/m3)system energy (K)
nanoHUB5000500010,0005.11 ± 0.14–428 ± 20
SDSC25,00025,00050,0005.01 ± 0.07–447 ± 4
Three simulations were performed for each tool, and the mean and standard deviation for the main numeric results were computed. All three simulations used the same simulation conditions, except a different seed value was used to initialize the random number generator. The same conditions were run on clusters at the San Diego Supercomputing Center (SDSC) using a larger number of cycles and different random seed values to test for numerical convergence. As shown in Tables 2345678910, results with the larger number of cycles were similar to those computed using our nanoHUB tools. This demonstrates a reasonable level of convergence. Table 11 lists the times required to run each nanoHUB tool.
Table 11. Time Required to Run Each Nanohub Toola
toolrun time (min)toolrun time (min)
1160 ± 307120.7 ± 0.2
224.6 ± 0.788 ± 2
332 ± 196.5 ± 0.4
433.0 ± 0.2104.77 ± 0.01
5153 ± 21128.0 ± 0.5
60.676 ± 0.008  
a

The average and standard deviation of three runs is listed.

4.2. Comparing Simulations to Experiments

Here, we compared the experimental results from various literature sources to the simulated results obtained from two nanoHUB tools. These tools are gas adsorption calculator (tool 2) and surface area calculator (tool 7).
Table 12 compares experimentally extracted and calculated surface areas for IRMOF-2 and IRMOF-3. As shown in Table 12, surface areas computed using the “sigma” and “rmin” probe distances were larger than experimentally extracted Brunauer–Emmett–Teller (BET) surface areas and smaller than surface areas computed using the Connolly method with the probe distance set equal to the N2 kinetic diameter.
Table 12. Comparison between Experimentally Extracted and Calculated Surface Areas for IRMOF-2 and IRMOF-3a
materialarea (m2/g) Langmuirbarea (m2/g) BETbarea (m2/g) (N2 kinetic diam)barea (m2/g) (sigma)area (m2/g) (rmin)
IRMOF-22544172227802684.4 ± 1.02349.4 ± 0.5
IRMOF-33062244636133296.4 ± 0.92834.0 ± 1.3
a

Sigma and rmin are surface areas calculated using two different probe distances. The calculations used 2000 production cycles.

b

From ref (44).

Finally, Figure 19 compares calculated and experimental adsorption isotherms of nitrogen (N2), methane, and carbon dioxide gases adsorbed in IRMOF-1 at 298 K. These simulated adsorption isotherms predicted slightly more adsorption than the experimental isotherms. This could be due to imperfections in the forcefield or due to impurities (e.g., residual molecules) or defects in the MOF crystals that partially block adsorption sites.

Figure 19

Figure 19. Comparison of calculated to experimental isotherms of N2, methane, and CO2 gases adsorbed in the MOF IRMOF-1. The black curves are the calculated isotherms. The orange curves are the N2 (ref (45)), methane (ref (46)), and CO2 (ref (47)) experimental isotherms.

High Resolution Image
Download MS PowerPoint Slide

5. Conclusions

Click to copy section linkSection link copied!
Eleven simulation modules were successfully developed and published on nanoHUB. These tools were created to teach users how to use classical atomistic simulations to compute the following: (i) gas adsorption and diffusion in porous materials, (ii) radial distribution functions and self-diffusivity of pure fluids, and (iii) surface areas and void fractions of porous materials. These modules included a GUI on the front end to accept user inputs and to display results back to the user. Middleware communicated information between the GUI front end and the RASPA software that performed the simulations. These tools are accessible worldwide through an online web browser without requiring the user to install any software. These simulations run on nanoHUB’s computing cluster without requiring the user to supply computational resources. Each simulation tool has an associated set of PowerPoint slides that briefly explain how the tool works. An example RASPA input file is also provided for each tool. This arrangement allows users to quickly learn basic simulation principles.
In this article, we also reported tests demonstrating these tools give useful results. We report runtimes, which were purposefully kept short to facilitate easier learning. We show the computational precision is reasonable and compares favorably to longer runs that included more production cycles. For a couple of the tools, computed results were compared to published experimental data. As shown in Table 12, surface areas computed using the “sigma” and “rmin” probe distances were larger than experimentally extracted BET surface areas and smaller than surface areas computed using the Connolly method with the probe distance set equal to the N2 kinetic diameter. For the adsorption of N2, methane, and CO2 in IRMOF-1 at 298 K shown in Figure 19, we compared results from the gas adsorption simulation tool to previously reported experimental results. In these cases, the simulated adsorption isotherms predicted slightly more adsorption than the experimental isotherms; this could be due to imperfections in the forcefield or due to impurities (e.g., residual molecules) or defects in the MOF crystals that partially block adsorption sites.

Author Information

Click to copy section linkSection link copied!
  • Corresponding Author
  • Author
    • Julian C. Umeh - Chemical and Materials Engineering, New Mexico State University, Jett Hall 268, Las Cruces, New Mexico88003-3805, United States;  Orcidhttps://orcid.org/0000-0002-7132-8033
  • Author Contributions

    J.C.U. programmed the 11 nanoHUB simulation tools. T.A.M. supervised the project. Both authors designed the study, performed simulations, analyzed data, and wrote the manuscript.

  • Notes
    The authors declare no competing financial interest.

Acknowledgments

Click to copy section linkSection link copied!

National Science Foundation (NSF) CAREER Award DMR-1555376 provided financial support. Supercomputing resources were provided by the Extreme Science and Engineering Discovery Environment (XSEDE). (48) XSEDE is funded by NSF Grant ACI-1548562. XSEDE project Grant TG-CTS100027 provided allocations on the Comet and Expanse clusters at the San Diego Supercomputing Center (SDSC). Special thanks to the nanoHUB support team for answering our many questions about setting up tools and for their technical assistance.

References

Click to copy section linkSection link copied!

This article references 48 other publications.

  1. 1
    Madhavan, K.; Zentner, L.; Farnsworth, V.; Shivarajapura, S.; Zentner, M.; Denny, N.; Klimeck, G. nanoHUB.org: cloud-based services for nanoscale modeling, simulation, and education. Nanotechnol. Rev. 2013, 2, 107– 117,  DOI: 10.1515/ntrev-2012-0043
    Google Scholar
    There is no corresponding record for this reference.
  2. 2
    Magana, A. J.; Brophy, S. P.; Bodner, G. M. An exploratory study of engineering and science students’ perceptions of nanoHUB.org simulations. Int. J. Eng. Educ. 2012, 28, 1019– 1032
    Google Scholar
    There is no corresponding record for this reference.
  3. 3
    Nanohub.org usage statistics. https://nanohub.org/usage, (accessed March 1, 2022).
    Google Scholar
    There is no corresponding record for this reference.
  4. 4
    Dubbeldam, D.; Calero, S.; Ellis, D. E.; Snurr, R. Q. RASPA: molecular simulation software for adsorption and diffusion in flexible nanoporous materials. Mol. Simul. 2016, 42, 81– 101,  DOI: 10.1080/08927022.2015.1010082
    Google Scholar
    4
    RASPA: molecular simulation software for adsorption and diffusion in flexible nanoporous materials
    Dubbeldam, David; Calero, Sofia; Ellis, Donald E.; Snurr, Randall Q.
    Molecular Simulation (2016), 42 (2), 81-101CODEN: MOSIEA; ISSN:0892-7022. (Taylor & Francis Ltd.)
    A new software package, RASPA, for simulating adsorption and diffusion of mols. in flexible nanoporous materials is presented. The code implements the latest state-of-the-art algorithms for mol. dynamics and Monte Carlo (MC) in various ensembles including symplectic/measure-preserving integrators, Ewald summation, configurational-bias MC, continuous fractional component MC, reactive MC and Baker's minimisation. We show example applications of RASPA in computing coexistence properties, adsorption isotherms for single and multiple components, self- and collective diffusivities, reaction systems and visualisation. The software is released under the GNU General Public License.
  5. 5
    Rogge, S. M. J.; Goeminne, R.; Demuynck, R.; Gutiérrez-Sevillano, J. J.; Vandenbrande, S.; Vanduyfhuys, L.; Waroquier, M.; Verstraelen, T.; Van Speybroeck, V. Modeling gas adsorption in flexible metal-organic frameworks via hybrid Monte Carlo/molecular dynamics schemes. Adv. Theory Simul. 2019, 2, 1800177  DOI: 10.1002/adts.201800177
    Google Scholar
    5
    Modeling Gas Adsorption in Flexible Metal-Organic Frameworks via Hybrid Monte Carlo/Molecular Dynamics Schemes
    Rogge, Sven M. J.; Goeminne, Ruben; Demuynck, Ruben; Gutierrez-Sevillano, Juan Jose; Vandenbrande, Steven; Vanduyfhuys, Louis; Waroquier, Michel; Verstraelen, Toon; Van Speybroeck, Veronique
    Advanced Theory and Simulations (2019), 2 (4), 1800177CODEN: ATSDCW; ISSN:2513-0390. (Wiley-VCH Verlag GmbH & Co. KGaA)
    Herein, a hybrid Monte Carlo (MC)/mol. dynamics (MD) simulation protocol that properly accounts for the extraordinary structural flexibility of metal-org. frameworks (MOFs) is developed and validated. This is vital to accurately predict gas adsorption isotherms and guest-induced flexibility of these materials. First, the performance of three recent models to predict adsorption isotherms and flexibility in MOFs is critically investigated. While these methods succeed in providing qual. insight in the gas adsorption process in MOFs, their accuracy remains limited as the intrinsic flexibility of these materials is very hard to account for. To overcome this challenge, a hybrid MC/MD simulation protocol that is specifically designed to handle the flexibility of the adsorbent, including the shape flexibility, is introduced, thereby unifying the strengths of the previous models. It is demonstrated that the application of this new protocol to the adsorption of neon, argon, xenon, methane, and carbon dioxide in MIL-53(Al), a prototypical flexible MOF, substantially decreases the inaccuracy of the obtained adsorption isotherms and predicted guest-induced flexibility. As a result, this method is ideally suited to rationalize the adsorption performance of flexible nanoporous materials at the mol. level, paving the way for the conscious design of MOFs as industrial adsorbents.
  6. 6
    Lin, L. C.; Lee, K.; Gagliardi, L.; Neaton, J. B.; Smit, B. Force-field development from electronic structure calculations with periodic boundary conditions: applications to gaseous adsorption and transport in metal-organic frameworks. J. Chem. Theory Comput. 2014, 10, 1477– 1488,  DOI: 10.1021/ct500094w
    Google Scholar
    6
    Force-Field Development from Electronic Structure Calculations with Periodic Boundary Conditions: Applications to Gaseous Adsorption and Transport in Metal-Organic Frameworks
    Lin, Li-Chiang; Lee, Kyuho; Gagliardi, Laura; Neaton, Jeffrey B.; Smit, Berend
    Journal of Chemical Theory and Computation (2014), 10 (4), 1477-1488CODEN: JCTCCE; ISSN:1549-9618. (American Chemical Society)
    A systematic and efficient methodol. to derive accurate (nonpolarizable) force fields from periodic d. functional theory (DFT) calcns. for use in classical mol. simulations has been developed. The methodol. requires reduced computation cost compared with other conventional ways. The whole process is performed self-consistently in a fully periodic system. The force fields derived by using this methodol. accurately predict the CO2 and H2O adsorption isotherms inside Mg-MOF-74, and is transferable to Zn-MOF-74; by replacing the Mg-CO2 interactions with the corresponding Zn-CO2 interactions, we obtain an accurate prediction of the corresponding isotherm. This methodol. was used to address the effect of water on the sepn. of flue gases in these materials. In general, the mixt. isotherms of CO2 and H2O calcd. with these derived force fields show a significant redn. in CO2 uptake with the existence of trace amts. of water vapor. The effect of water, however, was found to be quant. different between Mg- and Zn-MOF-74.
  7. 7
    Dubbeldam, D.; Walton, K. S.; Vlugt, T. J. H.; Calero, S. Design, parameterization, and implementation of atomic force fields for adsorption in nanoporous materials. Adv. Theory Simul. 2019, 2, 1900135  DOI: 10.1002/adts.201900135
    Google Scholar
    7
    Design, Parameterization, and Implementation of Atomic Force Fields for Adsorption in Nanoporous Materials
    Dubbeldam, David; Walton, Krista S.; Vlugt, Thijs J. H.; Calero, Sofia
    Advanced Theory and Simulations (2019), 2 (11), 1900135CODEN: ATSDCW; ISSN:2513-0390. (Wiley-VCH Verlag GmbH & Co. KGaA)
    A review. Mol. simulations are an excellent tool to study adsorption and diffusion in nanoporous materials. Examples of nanoporous materials are zeolites, carbon nanotubes, clays, metal-org. frameworks (MOFs), covalent org. frameworks (COFs) and zeolitic imidazolate frameworks (ZIFs). The mol. confinement these materials offer has been exploited in adsorption and catalysis for almost 50 years. Mol. simulations have provided understanding of the underlying shape selectivity, and adsorption and diffusion effects. Much of the reliability of the modeling predictions depends on the accuracy and transferability of the force field. However, flexibility and the chem. and structural diversity of MOFs add significant challenges for engineering force fields that are able to reproduce exptl. obsd. structural and dynamic properties. Recent developments in design, parameterization, and implementation of force fields for MOFs and zeolites are reviewed.
  8. 8
    Kaplan, I. G. Intermolecular Interactions: Physical Picture, Computational Methods and Model Potentials; John Wiley & Sons: West Sussex, U.K., 2006.
    Google Scholar
    There is no corresponding record for this reference.
  9. 9
    Rappe, A. K.; Casewit, C. J.; Colwell, K. S.; Goddard, W. A.; Skiff, W. M. UFF, a full periodic-table force-field for molecular mechanics and molecular-dynamics simulations. J. Am. Chem. Soc. 1992, 114, 10024– 10035,  DOI: 10.1021/ja00051a040
    Google Scholar
    9
    UFF, a full periodic table force field for molecular mechanics and molecular dynamics simulations
    Rappe, A. K.; Casewit, C. J.; Colwell, K. S.; Goddard, W. A., III; Skiff, W. M.
    Journal of the American Chemical Society (1992), 114 (25), 10024-35CODEN: JACSAT; ISSN:0002-7863.
    A new mol. mechanics force field, the Universal force field (UFF), is described wherein the force field parameters are estd. using general rules based only on the element, its hybridization and its connectivity. The force field functional forms, parameters, and generating formulas for the full periodic table are presented.
  10. 10
    Walters, E. T.; Mohebifar, M.; Johnson, E. R.; Rowley, C. N. Evaluating the London dispersion coefficients of protein force fields using the exchange-hole dipole moment model. J. Phys. Chem. B 2018, 122, 6690– 6701,  DOI: 10.1021/acs.jpcb.8b02814
    Google Scholar
    10
    Evaluating the London Dispersion Coefficients of Protein Force Fields Using the Exchange-Hole Dipole Moment Model
    Walters, Evan T.; Mohebifar, Mohamad; Johnson, Erin R.; Rowley, Christopher N.
    Journal of Physical Chemistry B (2018), 122 (26), 6690-6701CODEN: JPCBFK; ISSN:1520-5207. (American Chemical Society)
    London dispersion is one of the fundamental interactions involved in protein folding and dynamics. The popular CHARMM36, Amber ff14sb, and OPLS-AA force fields represent these interactions through the C6/r6 term of the Lennard-Jones potential, where the C6 parameters are assigned empirically. Here, dispersion coeffs. of these three force fields are shown to be roughly 50% larger than values calcd. using the quantum-mech.-derived exchange-hole dipole moment (XDM) model. The CHARMM36 and Amber OL15 force fields for nucleic acids also exhibit this trend. The hydration energies of the side-chain models were calcd. using REMD-TI for the CHARMM36, Amber ff14sb, and OPLS-AA force fields. These force fields predict side-chain hydration energies that are in generally good agreement with the exptl. values, which suggests the total strength of aq. dispersion interactions is correct, despite C6 coeffs. that are considerably larger than XDM predicts. An anal. expression for the dispersion hydration energy using XDM coeffs. shows that higher-order dispersion terms (i.e., C8 and C10) account for roughly 37.5% of the hydration energy of methane. This suggests that the C6 dispersion coeffs. used in contemporary force fields are elevated to account for the neglected higher-order terms.
  11. 11
    Jones, J. E. On the determination of molecular fields - II From the equation of state of a gas. Proc. R. Soc. London, Ser. A 1924, 106, 463– 477,  DOI: 10.1098/rspa.1924.0082
    Google Scholar
    11
    The determination of molecular fields (II) From the equation of state of a gas
    Jones, J. R.
    Proceedings of the Royal Society of London, Series A: Mathematical, Physical and Engineering Sciences (1924), 106 (), 463-77CODEN: PRLAAZ; ISSN:1364-5021.
    The virial coeffs. are evaluated for an equation of state of gases for the case of a mol. with attractive and repelling forces, as described above. The values obtained for the consts. of these forces for A agree fairly well with those calcd. from the viscosity data. In an appendix, recalcn. on the basis of more recent data on the isotherms of A give better agreement, pointing to 14 1/3 as the best value for the index of the repulsive force.
  12. 12
    Boda, D.; Henderson, D. The effects of deviations from Lorentz–Berthelot rules on the properties of a simple mixture. Mol. Phys. 2008, 106, 2367– 2370,  DOI: 10.1080/00268970802471137
    Google Scholar
    12
    The effects of deviations from Lorentz-Berthelot rules on the properties of a simple mixture
    Boda, Dezso; Henderson, Douglas
    Molecular Physics (2008), 106 (20), 2367-2370CODEN: MOPHAM; ISSN:0026-8976. (Taylor & Francis Ltd.)
    Generally, the parameters in the interaction potential between like mols. in a mixt. can be detd. in a relatively straightforward manner from the properties of the pure components. However, the detn. of the parameters in the interaction potential between the unlike pairs in the mixts. is more difficult. As a result, these parameters are usually estd. from avs. of the like parameters. The most common recipes are the Lorentz-Berthelot mixing rules, where the energy and mol. size parameters are presumed to be geometric and arithmetic avs., resp. There have been some studies of the consequences of deviations from the energy rule but almost no studies of the consequences of deviations from the size rule. Here, we study the effects of deviations from both rules on the radial distribution functions of a simple mixt. We find from simulations that, for this mixt., the effect of deviations from the energy rule on the radial distribution function are rather small but that the effect of deviations from the size rule can be significant and are interesting.
  13. 13
    Berthelot, D. Sur le mélange des gaz. Compt. Rendus 1898, 126, 1703– 1706
    Google Scholar
    13
    On mixtures of gases
    Berthelot, D.
    Comptes Rendus Hebdomadaires des Seances de l'Academie des Sciences (1898), 126 (), 1703CODEN: COREAF ISSN:.
    The author proposes a formula of the van der Waals type for a mixture of two gases and shows how the constants A and B for the mixture can be calculated from the corresponding constants for the single gases.
  14. 14
    Lorentz, H. A. Ueber die anwendung des satzes vom virial in der kinetischen theorie der gase. Ann. Phys. 1881, 248, 127– 136,  DOI: 10.1002/andp.18812480110
    Google Scholar
    There is no corresponding record for this reference.
  15. 15
    Dubbeldam, D.; Calero, S.; Vlugt, T. J. H.; Ellis, D. E.; Snurr, R. Q. RASPA 2.0: Molecular Software Package for Adsorption and Diffusion in (Flexible) Nanoporous Materials , program manual, March 2015, pp 1− 145.
    Google Scholar
    There is no corresponding record for this reference.
  16. 16
    Dubbeldam, D.; Walton, K. S.; Ellis, D. E.; Snurr, R. Q. Exceptional negative thermal expansion in isoreticular metal-organic frameworks. Angew. Chem., Int. Ed. 2007, 46, 4496– 4499,  DOI: 10.1002/anie.200700218
    Google Scholar
    16
    Exceptional negative thermal expansion in isoreticular metal-organic frameworks
    Dubbeldam, David; Walton, Krista S.; Ellis, Donald E.; Snurr, Randall Q.
    Angewandte Chemie, International Edition (2007), 46 (24), 4496-4499CODEN: ACIEF5; ISSN:1433-7851. (Wiley-VCH Verlag GmbH & Co. KGaA)
    A new model for flexible frameworks is used to simulate the structures and adsorption properties of isoreticular metal-org. frameworks (IRMOFs), such as IRMOF-1. The structural simulations suggest that the IRMOFs have neg. thermal-expansion coeffs. over their full temp. ranges of stability.
  17. 17
    Chen, B.; Siepmann, J. I. Transferable potentials for phase equilibria. 3. Explicit-hydrogen description of normal alkanes. J. Phys. Chem. B 1999, 103, 5370– 5379,  DOI: 10.1021/jp990822m
    Google Scholar
    17
    Transferable potentials for phase equilibria. 3. Explicit-hydrogen description of normal alkanes
    Chen, Bin; Siepmann, J. Ilja
    Journal of Physical Chemistry B (1999), 103 (25), 5370-5379CODEN: JPCBFK; ISSN:1089-5647. (American Chemical Society)
    Motivated by shortcomings of the available united-atom models for alkanes, a new explicit-H model for n-alkanes (TraPPE-EH, transferable potentials for phase equil.-explicit H) is developed from fitting to 1-component fluid-phase properties. In addn. to Lennard-Jones sites on C atoms, this model utilizes Lennard-Jones sites on the centers of C-H bonds. Configurational-bias Monte Carlo simulations in the Gibbs and canonical ensembles were carried out to calc. the 1-component vapor-liq. phase equil. for CH4 to n-dodecane, to det. the phase diagram of supercrit. C2H6 and n-heptane mixts., to obtain the Gibbs free energies of transfer for n-pentane (I) and n-hexane between He vapor and n-heptane liq. phases, and to study the high-pressure region of the equation of state for I and n-decane. The explicit-H representation, with its more faithful description of the mol. shape of alkanes, allows us to find a set of Lennard-Jones parameters that yields significantly better agreement with expt. for 1- and multicomponent phase equil. than our united-atom alkane model, but the price is higher computational cost.
  18. 18
    Eggimann, B. L.; Sunnarborg, A. J.; Stern, H. D.; Bliss, A. P.; Siepmann, J. I. An online parameter and property database for the TraPPE force field. Mol. Simul. 2014, 40, 101– 105,  DOI: 10.1080/08927022.2013.842994
    Google Scholar
    18
    An online parameter and property database for the TraPPE force field
    Eggimann, Becky L.; Sunnarborg, Amara J.; Stern, Hudson D.; Bliss, Andrew P.; Siepmann, J. Ilja
    Molecular Simulation (2014), 40 (1-3), 101-105CODEN: MOSIEA; ISSN:0892-7022. (Taylor & Francis Ltd.)
    A review. The transferable potentials for phase equil. (TraPPE) force field aims to be accurate, computationally efficient and applicable to a wide range of chem. compds., state points and thermophys. properties. When new users wish to implement TraPPE models into their chosen simulation program, they face several obstacles: the TraPPE models are dispersed over many sep. publications and misinterpretations of the primary literature are possible; the TraPPE force field makes specific choices for std. conventions that may require non-trivial code modifications for some simulation software. Therefore, the TraPPE developers report here a resource website and online searchable parameter and property database designed to provide new and experienced users with tools for successful implementation and validation (http://www.chem.umn.edu/groups/siepmann/trappe/).
  19. 19
    Martin, M. G.; Siepmann, J. I. Transferable potentials for phase equilibria. 1. United-atom description of n-alkanes. J. Phys. Chem. B 1998, 102, 2569– 2577,  DOI: 10.1021/jp972543+
    Google Scholar
    19
    Transferable Potentials for Phase Equilibria. 1. United-Atom Description of n-Alkanes
    Martin, Marcus G.; Siepmann, J. Ilja
    Journal of Physical Chemistry B (1998), 102 (14), 2569-2577CODEN: JPCBFK; ISSN:1089-5647. (American Chemical Society)
    A new set of united-atom Lennard-Jones interaction parameters for n-alkanes is proposed from fitting to crit. temps. and satd. liq. densities. Configurational-bias Monte Carlo simulations in the Gibbs ensemble were carried out to det. the vapor-liq. coexistence curves for methane to dodecane using three united-atom force fields: OPLS [Jorgensen, et al. J. Am. Chem. Soc. 1984, 106, 813], SKS [Siepmann, et al. Nature 1993, 365, 330], and TraPPE. Std. specific densities and the high-pressure equation-of-state for the transferable potentials for phase equil. (TraPPE) model were studied by simulations in the isobaric-isothermal and canonical ensembles, resp. It is found that one set of Me and methylene parameters is sufficient to accurately describe the fluid phases of all n-alkanes with two or more carbon atoms. Whereas other n-alkane force fields employ Me groups that are either equal or larger in size than the methylene groups, it is demonstrated here that using a smaller Me group yields a better fit to the set of exptl. data. As should be expected from an effective pair potential, the new parameters do not reproduce exptl. second virial coeffs. Satd. vapor pressures and densities show small, but systematic deviation from the exptl. data.
  20. 20
    Potoff, J. J.; Siepmann, J. I. Vapor–liquid equilibria of mixtures containing alkanes, carbon dioxide, and nitrogen. AIChE J. 2001, 47, 1676– 1682,  DOI: 10.1002/aic.690470719
    Google Scholar
    20
    Vapor-liquid equilibria of mixtures containing alkanes, carbon dioxide, and nitrogen
    Potoff, Jeffrey J.; Siepmann, J. Ilja
    AIChE Journal (2001), 47 (7), 1676-1682CODEN: AICEAC; ISSN:0001-1541. (American Institute of Chemical Engineers)
    New force fields for carbon dioxide and nitrogen are introduced that quant. reproduce the vapor-liq. equil. (VLE) of the neat systems and their mixts. with alkanes. In addn. to the usual VLE calcns. for pure CO2 and N2, calcns. of the binary mixts. with propane were used in the force-field development to achieve a good balance between dispersive and electrostatic (quadrupole-quadrupole) interactions. The transferability of the force fields was then assessed from calcns. of the VLE for the binary mixts. with n-hexane, the binary mixt. of CO2/N2, and the ternary mixt. of CO2/N2/propane. The VLE calcns. were carried out using configurational-bias Monte Carlo simulations in either the grand canonical ensemble with histogram-reweighting or in the Gibbs ensemble.
  21. 21
    Darkrim, F.; Levesque, D. Monte Carlo simulations of hydrogen adsorption in single-walled carbon nanotubes. J. Chem. Phys. 1998, 109, 4981– 4984,  DOI: 10.1063/1.477109
    Google Scholar
    21
    Monte Carlo simulations of hydrogen adsorption in single-walled carbon nanotubes
    Darkrim, Farida; Levesque, Dominique
    Journal of Chemical Physics (1998), 109 (12), 4981-4984CODEN: JCPSA6; ISSN:0021-9606. (American Institute of Physics)
    Within the framework of a study on the properties of carbon nanotubes, a promising new material, we performed numerical simulation of hydrogen adsorption at room temp. in single-walled nanotubes. The structure of this material is favorable to the adsorption phenomenon because of the narrow size distribution of the nanotube diams., which have dimensions on the order of the range of the carbon attractive interaction. We discuss the influence of the single-walled carbon nanotube diams. on the relative arrangement of carbon atoms and hydrogen mols. within an array of parallel single-walled carbon nanotubes. We also studied the influence on adsorption of the distance between the nearest-neighbor nanotubes.
  22. 22
    Zhang, L.; Siepmann, J. I. Direct calculation of Henry’s law constants from Gibbs ensemble Monte Carlo simulations: nitrogen, oxygen, carbon dioxide and methane in ethanol. Theor. Chem. Acc. 2006, 115, 391– 397,  DOI: 10.1007/s00214-005-0073-1
    Google Scholar
    22
    Direct calculation of Henry's law constants from Gibbs ensemble Monte Carlo simulations: nitrogen, oxygen, carbon dioxide and methane in ethanol
    Zhang, Ling; Siepmann, J. Ilja
    Theoretical Chemistry Accounts (2006), 115 (5), 391-397CODEN: TCACFW; ISSN:1432-881X. (Springer GmbH)
    Configurational-bias Monte Carlo simulations in the Gibbs ensemble were used to calc. Henry's law consts., Ostwald solubilities, and Gibbs free energies of transfer for oxygen, nitrogen, methane, and carbon dioxide in ethanol at 323 and 373 K. These three soly. descriptors can be expressed as functions of mech. properties that are directly observable in the Gibbs ensemble approach, thereby allowing for very precise detn. of the descriptors. Addnl., the Henry's law consts. of multiple solutes can be computed from a single simulation. Most of the simulations were carried out for systems contg. 1000 solvent and up to 8 solute mols., and further simulations using either 500 or 2000 solvent mols. point to negligible system size effects. A comparison with exptl. data shows that the united-atom version of the transferable potential for phase equil. force field yields Henry's law consts. that reproduce well the differences between the four solutes and the changes upon increase of the temp.
  23. 23
    Mayo, S. L.; Olafson, B. D.; Goddard, W. A. DREIDING - A generic force field for molecular simulations. J. Phys. Chem. 1990, 94, 8897– 8909,  DOI: 10.1021/j100389a010
    Google Scholar
    23
    DREIDING: a generic force field for molecular simulations
    Mayo, Stephen L.; Olafson, Barry D.; Goddard, William A., III
    Journal of Physical Chemistry (1990), 94 (26), 8897-909CODEN: JPCHAX; ISSN:0022-3654.
    The parameters are given for a new generic force field, DREIDING, that is useful for predicting structures and dynamics of org., biol., and main-group inorg. mols. The philosophy in DREIDING is to use general force consts. and geometry parameters based on simple hybridization considerations rather than individual force consts. and geometric parameters that depend on the particular combination of atoms involved in the bond, angle, or torsion terms. Thus all bond distances are derived from at. radii, and there is only one force const. each for bonds, angles, and inversions and only six different values for torsional barriers. Parameters are defined for all possible combinations of atoms and new atoms can be added to the force field rather simply. The parameters are given for the "nonmetallic" main-group elements (B, C, N, O, F columns for the C, Si, Ge, and Sn rows) plus H and a few metals (Na, Ca, Zn, Fe). The accuracy of the DREIDING force field is tested by comparing with (i) 76 accurately detd. crystal structures of org. compds. involving H, C, N, O, F, P, S, Cl, and Br, (ii) rotational barriers of a no. of mols., and (iii) relative conformational energies and barriers of a no. of mols. There is excellent agreement.
  24. 24
    Bai, P.; Tsapatsis, M.; Siepmann, J. I. TraPPE-zeo: Transferable potentials for phase equilibria force field for all-silica zeolites. J. Phys. Chem. C 2013, 117, 24375– 24387,  DOI: 10.1021/jp4074224
    Google Scholar
    24
    TraPPE-zeo: Transferable Potentials for Phase Equilibria Force Field for All-Silica Zeolites
    Bai, Peng; Tsapatsis, Michael; Siepmann, J. Ilja
    Journal of Physical Chemistry C (2013), 117 (46), 24375-24387CODEN: JPCCCK; ISSN:1932-7447. (American Chemical Society)
    The transferable potentials for phase equil. (TraPPE) force field is extended to all-silica zeolites. This novel force field is parametrized to match the exptl. adsorption isotherms of n-heptane, propane, carbon dioxide, and ethanol with the Lennard-Jones parameters for sorbate-framework interactions detd. in a consistent manner using the Lorentz-Berthelot combining rules as for other parts of the TraPPE force field. The TraPPE-zeo force field allows for accurate predictions for both adsorption and diffusion of alkanes, alcs., carbon dioxide, and water over a wide range of pressures and temps. To achieve transferability to a wider range of mol. types, ranging from nonpolar to dipolar and hydrogen-bonding compds., Lennard-Jones interaction sites and partial charges are placed at both the oxygen and the silicon atoms of the zeolite lattice, which allows for a better balance of dispersive and 1st-order electrostatic interactions than is achievable with the Lennard-Jones potential used only for the oxygen atoms. The use of the Lorentz-Berthelot combining rules for unlike interactions makes the TraPPE-zeo force field applicable to any sorbate as long as the relevant TraPPE sorbate-sorbate parameters are available. The TraPPE-zeo force field allows for greatly improved predictive power compared to force fields that explicitly tabulate the individual cross-interaction parameters.
  25. 25
    Dubbeldam, D.; Torres-Knoop, A.; Walton, K. S. On the inner workings of Monte Carlo codes. Mol. Simul. 2013, 39, 1253– 1292,  DOI: 10.1080/08927022.2013.819102
    Google Scholar
    25
    On the inner workings of Monte Carlo codes
    Dubbeldam, David; Torres-Knoop, Ariana; Walton, Krista S.
    Molecular Simulation (2013), 39 (14-15), 1253-1292CODEN: MOSIEA; ISSN:0892-7022. (Taylor & Francis Ltd.)
    We review state-of-the-art Monte Carlo (MC) techniques for computing fluid coexistence properties (Gibbs simulations) and adsorption simulations in nanoporous materials such as zeolites and metal-org. frameworks. Conventional MC is discussed and compared to advanced techniques such as reactive MC, configurational-bias Monte Carlo and continuous fractional MC. The latter technique overcomes the problem of low insertion probabilities in open systems. Other modern methods are (hyper-)parallel tempering, Wang-Landau sampling and nested sampling. Details on the techniques and acceptance rules as well as to what systems these techniques can be applied are provided. We highlight consistency tests to help validate and debug MC codes.
  26. 26
    Heinen, J.; Burtch, N. C.; Walton, K. S.; Dubbeldam, D. Flexible force field parameterization through fitting on the ab initio-derived elastic tensor. J. Chem. Theory Comput. 2017, 13, 3722– 3730,  DOI: 10.1021/acs.jctc.7b00310
    Google Scholar
    26
    Flexible Force Field Parameterization through Fitting on the Ab Initio-Derived Elastic Tensor
    Heinen, Jurn; Burtch, Nicholas C.; Walton, Krista S.; Dubbeldam, David
    Journal of Chemical Theory and Computation (2017), 13 (8), 3722-3730CODEN: JCTCCE; ISSN:1549-9618. (American Chemical Society)
    Constructing functional forms and their corresponding force field parameters for the metal-linker interface of metal-org. frameworks is challenging. We propose to fit these parameters on the elastic tensor, computed from ab initio d. functional theory calcns. The advantage of this top-down approach is that it becomes evident if functional forms are missing when components of the elastic tensor are off. As a proof-of-concept, a new flexible force field for MIL-47(V) is derived. Neg. thermal expansion is obsd. and framework flexibility has a negligible effect on adsorption and transport properties for small guest mols. We believe this force field parameterization approach can serve as a useful tool for developing accurate flexible force field models that capture the correct mech. behavior of the full periodic structure.
  27. 27
    Heinen, J.; Dubbeldam, D. On flexible force fields for metal-organic frameworks: Recent developments and future prospects. Wiley Interdiscip. Rev.: Comput. Mol. Sci. 2018, 8, e1363  DOI: 10.1002/wcms.1363
    Google Scholar
    There is no corresponding record for this reference.
  28. 28
    Allen, M. P.; Tildesley, D. J. Computer Simulation of Liquids, 2nd ed.; Oxford University Press: Oxford, U.K., 2017.
    Google Scholar
    There is no corresponding record for this reference.
  29. 29
    Frenkel, D.; Smit, B. Understanding Molecular Simulation: From Algorithms to Applications, 2nd ed.; Academic Press: San Diego, California, 2002.
    Google Scholar
    There is no corresponding record for this reference.
  30. 30
    Hastings, W. K. Monte Carlo sampling methods using Markov chains and their applications. Biometrika 1970, 57, 97– 109,  DOI: 10.1093/biomet/57.1.97
    Google Scholar
    There is no corresponding record for this reference.
  31. 31
    Metropolis, N.; Rosenbluth, A. W.; Rosenbluth, M. N.; Teller, A. H.; Teller, E. Equation of state calculations by fast computing machines. J. Chem. Phys. 1953, 21, 1087– 1092,  DOI: 10.1063/1.1699114
    Google Scholar
    31
    Equation-of-state calculations by fast computing machines
    Metropolis, Nicholas; Rosenbluth, Arianna W.; Rosenbluth, Marshall N.; Teller, Augusta H.; Teller, Edward
    Journal of Chemical Physics (1953), 21 (), 1087-92CODEN: JCPSA6; ISSN:0021-9606.
    A general method, suitable for fast computing machines, is described for investigating such properties as equations of state for substances consisting of interacting individual mols. The method consists of a modified Monte Carlo integration over configuration space. Results for the 2-dimensional rigid-sphere system were obtained on the Los Alamos MANIAC. These results were compared to the free-vol. equation of state and to a 4-term virial coeff. expansion.
  32. 32
    Hirotani, A.; Mizukami, K.; Miura, R.; Takaba, H.; Miya, T.; Fahmi, A.; Stirling, A.; Kubo, M.; Miyamoto, A. Grand canonical Monte Carlo simulation of the adsorption of CO2 on silicalite and NaZSM-5. Appl. Surf. Sci. 1997, 120, 81– 84,  DOI: 10.1016/S0169-4332(97)00222-5
    Google Scholar
    32
    Grand canonical Monte Carlo simulation of the adsorption of CO2 on silicalite and NaZSM-5
    Hirotani, Akiyasu; Mizukami, Koichi; Miura, Ryuji; Takaba, Hiromitsu; Miya, Takeshi; Fahmi, Adil; Stirling, Andras; Kubo, Momoji; Miyamoto, Akira
    Applied Surface Science (1997), 120 (1/2), 81-84CODEN: ASUSEE; ISSN:0169-4332. (Elsevier)
    The adsorption of carbon dioxide in silicalite and NaZSM-5 zeolite has been studied using new Monte Carlo software. In this program, sodium cations and framework are movable during the simulation. The calcd. adsorption isotherms are in good agreement with the exptl. results. The energy distribution of carbon dioxide over silicalite and NaZSM-5 shows that the increase of the adsorption energy for NaZSM-5 is mainly due the elec. field generated by sodium cations.
  33. 33
    Gao, G.; Wang, W. Gibbs Ensemble Monte Carlo simulation of binary vapor-liquid equilibria for CFC alternatives. Fluid Phase Equilib. 1997, 130, 157– 166,  DOI: 10.1016/S0378-3812(96)03195-0
    Google Scholar
    33
    Gibbs Ensemble Monte Carlo simulation of binary vapor-liquid equilibria for CFC alternatives
    Gao, Guangtu; Wang, Wenchuan
    Fluid Phase Equilibria (1997), 130 (1-2), 157-166CODEN: FPEQDT; ISSN:0378-3812. (Elsevier)
    The Gibbs Ensemble Monte Carlo (GEMC) simulation method has been used for vapor-liq. equil. calcns. of the binary systems R22-R142b and R22-R152a, which are considered to be promising mixt. refrigerants for replacing R12. All the mols. have been treated in terms of an effective Lennard-Jones potential, with temp. dependent parameters regressed by fitting vapor pressures and satd. liq. densities at various temps. for pure substances of interest. Good agreement between exptl. and calcd. data from the vol.-translation Peng-Robinson equation of state and the simulated results, including the compns., molar volumes for both the vapor and liq. phases, and heats of vaporization, indicates that the GEMC method can be applied to the description of phase behavior for these CFC alternative binary systems.
  34. 34
    Slepoy, A.; Thompson, A. P.; Plimpton, S. J. A constant-time kinetic Monte Carlo algorithm for simulation of large biochemical reaction networks. J. Chem. Phys. 2008, 20, 205101  DOI: 10.1063/1.2919546
    Google Scholar
    There is no corresponding record for this reference.
  35. 35
    Dove, M. T. An introduction to atomistic simulation methods. Seminarios de la SEM 2008, 4, 7– 37
    Google Scholar
    There is no corresponding record for this reference.
  36. 36
    Braun, E.; Gilmer, J.; Mayes, H. B.; Mobley, D. L.; Monroe, J. I.; Prasad, S.; Zuckerman, D. M. Best practices for foundations in molecular simulations. Living J. Comput. Mol. Sci. 2019, 1, 5957  DOI: 10.33011/livecoms.1.1.5957
    Google Scholar
    There is no corresponding record for this reference.
  37. 37
    Abouelnasr, M. K. F.; Smit, B. Diffusion in confinement: kinetic simulations of self- and collective diffusion behavior of adsorbed gases. Phys. Chem. Chem. Phys. 2012, 14, 11600– 11609,  DOI: 10.1039/c2cp41147d
    Google Scholar
    37
    Diffusion in confinement: kinetic simulations of self- and collective diffusion behavior of adsorbed gases
    Abouelnasr, Mahmoud K. F.; Smit, Berend
    Physical Chemistry Chemical Physics (2012), 14 (33), 11600-11609CODEN: PPCPFQ; ISSN:1463-9076. (Royal Society of Chemistry)
    The self- and collective-diffusion behaviors of adsorbed methane, helium, and isobutane in zeolite frameworks LTA, MFI, AFI, and SAS were examd. at various concns. using a range of mol. simulation techniques including Mol. Dynamics (MD), Monte Carlo (MC), Bennett-Chandler (BC), and kinetic Monte Carlo (kMC). This paper has three main results. (1) A novel model for the process of adsorbate movement between two large cages was created, allowing the formulation of a mixing rule for the re-crossing coeff. between two cages of unequal loading. The predictions from this mixing rule were found to agree quant. with explicit simulations. (2) A new approach to the dynamically cor. Transition State Theory method to anal. calc. self-diffusion properties was developed, explicitly accounting for nanoscale fluctuations in concn. This approach was demonstrated to quant. agree with previous methods, but is uniquely suited to be adapted to a kMC simulation that can simulate the collective-diffusion behavior. (3) While at low and moderate loadings the self- and collective-diffusion behaviors in LTA are obsd. to coincide, at higher concns. they diverge. A change in the adsorbate packing scheme was shown to cause this divergence, a trait which is replicated in a kMC simulation that explicitly models this behavior. These phenomena were further investigated for isobutane in zeolite MFI, where MD results showed a sepn. in self- and collective- diffusion behavior that was reproduced with kMC simulations.
  38. 38
    Skoulidas, A. I.; Sholl, D. S. Self-diffusion and transport diffusion of light gases in metal-organic framework materials assessed using molecular dynamics simulations. J. Phys. Chem. B 2005, 109, 15760– 15768,  DOI: 10.1021/jp051771y
    Google Scholar
    38
    Self-Diffusion and Transport Diffusion of Light Gases in Metal-Organic Framework Materials Assessed Using Molecular Dynamics Simulations
    Skoulidas, Anastasios I.; Sholl, David S.
    Journal of Physical Chemistry B (2005), 109 (33), 15760-15768CODEN: JPCBFK; ISSN:1520-6106. (American Chemical Society)
    Metal-org. framework (MOF) materials pose an interesting alternative to more traditional nanoporous materials for a variety of sepn. processes. Sepn. processes involving nanoporous materials can be controlled by either adsorption equil., diffusive transport rates, or a combination of these factors. Adsorption equil. was studied for a variety of gases in MOFs, but almost nothing is currently known about mol. diffusion rates in MOFs. The authors used equil. mol. dynamics (MD) to probe the self-diffusion and transport diffusion of a no. of small gas species in several MOFs as a function of pore loading at room temp. Specifically, the authors have studied Ar, CH4, CO2, N2, and H2 diffusion in MOF-5. The diffusion of Ar in MOF-2, MOF-3, and Cu-BTC was assessed in a similar manner. Results greatly expand the range of MOFs for which data describing mol. diffusion is available. The authors discuss the prospects for exploiting mol. transport properties in MOFs in practical sepn. processes and the future role of MD simulations in screening families of MOFs for these processes.
  39. 39
    Werder, T.; Walther, J. H.; Jaffe, R. L.; Halicioglu, T.; Koumoutsakos, P. On the water-carbon interaction for use in molecular dynamics simulations of graphite and carbon nanotubes. J. Phys. Chem. B 2003, 107, 1345– 1352,  DOI: 10.1021/jp0268112
    Google Scholar
    39
    On the Water-Carbon Interaction for Use in Molecular Dynamics Simulations of Graphite and Carbon Nanotubes
    Werder, T.; Walther, J. H.; Jaffe, R. L.; Halicioglu, T.; Koumoutsakos, P.
    Journal of Physical Chemistry B (2003), 107 (6), 1345-1352CODEN: JPCBFK; ISSN:1520-6106. (American Chemical Society)
    A systematic mol. dynamics study shows that the contact angle of a H2O droplet on graphite changes significantly as a function of the H2O-C interaction energy. Together with the observation that a linear relation can be established between the contact angle and the H2O monomer binding energy on graphite, a new route to calibrate interaction potential parameters is presented. Through a variation of the droplet size at 1000-17,500 H2O mols., the authors det. the line tension to be pos. and ∼2 × 10-10 J/m. To recover a macroscopic contact angle of 86°, a H2O monomer binding energy of -6.33 kJ mol-1 is required, which is obtained by applying a C-oxygen Lennard-Jones potential with the parameters εCO = 0.392 kJ mol-1 and σCO = 3.19 Å. For this new H2O-C interaction potential, the authors present d. profiles and hydrogen bond distributions for a H2O droplet on graphite.
  40. 40
    Henry, A. S.; Chen, G. Spectral phonon transport properties of silicon based on molecular dynamics simulations and lattice dynamics. J. Comput. Theor. Nanosci. 2008, 5, 141– 152,  DOI: 10.1166/jctn.2008.2454
    Google Scholar
    40
    Spectral phonon transport properties of silicon based on molecular dynamics simulations and lattice dynamics
    Henry, Asegun S.; Chen, Gang
    Journal of Computational and Theoretical Nanoscience (2008), 5 (2), 141-152CODEN: JCTNAB; ISSN:1546-1955. (American Scientific Publishers)
    Although the thermal cond. of silicon has been studied before, current estns. for the phonon mean free paths have not provided full explanation of the strong size effects exptl. obsd. for various silicon micro and nanostructures. Since phonon relaxation time models are mostly semi-empirical, the mean free paths cannot be detd. reliably and questions remain as to which polarizations, frequencies, and wavelengths are dominant heat carriers. Here we have used a combination of equil. mol. dynamics simulations and lattice dynamics calcns. to fully detail the spectral dependence of phonon transport properties in bulk silicon. By considering the frequency dependence of the sp. heat, group velocities, and mean free paths, we address these unresolved questions and examine the errors assocd. with isotropic and frequency averaged approxns. Simulation details, such as the convergence of results on the simulation time and extn. of phonon transport properties in different crystallog. directions, are also discussed.
  41. 41
    Karplus, M.; McCammon, J. A. Molecular dynamics simulations of biomolecules. Nat. Struct. Biol. 2002, 9, 646– 652,  DOI: 10.1038/nsb0902-646
    Google Scholar
    41
    Molecular dynamics simulations of biomolecules
    Karplus, Martin; McCammon, J. Andrew
    Nature Structural Biology (2002), 9 (9), 646-652CODEN: NSBIEW; ISSN:1072-8368. (Nature Publishing Group)
    A review. Mol. dynamics simulations are important tools for understanding the phys. basis of the structure and function of biol. macromols. The early view of proteins as relatively rigid structures has been replaced by a dynamic model in which the internal motions and resulting conformational changes play an essential role in their function. This review presents a brief description of the origin and early uses of biomol. simulations. It then outlines some recent studies that illustrate the utility of such simulations and closes with a discussion of their ever-increasing potential for contributing to biol.
  42. 42
    Umeh, J. C. Development of Simulation Tools on Nanuhub as Learning Modules. Master’s thesis; New Mexico State University: Las Cruces, NM, 2019.
    Google Scholar
    There is no corresponding record for this reference.
  43. 43
    Humbert, M. T.; Zhang, Y.; Maginn, E. J. PyLAT: Python LAMMPS analysis tools. J. Chem. Inf. Model. 2019, 59, 1301– 1305,  DOI: 10.1021/acs.jcim.9b00066
    Google Scholar
    43
    PyLAT: Python LAMMPS Analysis Tools
    Humbert, Michael T.; Zhang, Yong; Maginn, Edward J.
    Journal of Chemical Information and Modeling (2019), 59 (4), 1301-1305CODEN: JCISD8; ISSN:1549-9596. (American Chemical Society)
    Reproducibility and accuracy have become increasingly important issues for the mol. simulation community. The availability of validated open-source postprocessing tools to analyze simulation trajectories and compute properties is key to helping researchers conduct more accurate and reproducible simulations. Here the authors describe a suite of open-source Python-based postprocessing routines the authors have developed called PyLAT. PyLAT is compatible with the popular mol. dynamics package LAMMPS and enables users to compute viscosities, self-diffusivities, ionic conductivities, mol. or ion pair lifetimes, dielec. consts., and radial distribution functions using best-practice methods.
  44. 44
    Rowsell, J. L. C.; Yaghi, O. M. Effects of functionalization, catenation, and variation of the metal oxide and organic linking units on the low-pressure hydrogen adsorption properties of metal-organic frameworks. J. Am. Chem. Soc. 2006, 128, 1304– 1315,  DOI: 10.1021/ja056639q
    Google Scholar
    44
    Effects of Functionalization, Catenation, and Variation of the Metal Oxide and Organic Linking Units on the Low-Pressure Hydrogen Adsorption Properties of Metal-Organic Frameworks
    Rowsell, Jesse L. C.; Yaghi, Omar M.
    Journal of the American Chemical Society (2006), 128 (4), 1304-1315CODEN: JACSAT; ISSN:0002-7863. (American Chemical Society)
    The dihydrogen adsorption isotherms of eight metal-org. frameworks (MOFs), measured at 77 K up to a pressure of 1 atm, were examd. for correlations with their structural features. All materials display approx. Type I isotherms with no hysteresis, and satn. was not reached for any of the materials under these conditions. Among the six iso-reticular MOFs (IRMOFs) studied, the catenated materials exhibit the largest capacities on a molar basis, up to 9.8 H2 per formula unit. The addn. of functional groups (-Br, -NH2, -C2H4-) to the phenylene links of IRMOF-1 (MOF-5), or their replacement with thieno[3,2-b]thiophene moieties in IRMOF-20, altered the adsorption behavior by a minor amt. despite large variations in the pore vols. of the resulting materials. But replacement of the metal oxide units with those contg. coordinatively unsatd. metal sites resulted in greater H2 uptake. The enhanced affinities of these materials, MOF-74 and HKUST-1, were further demonstrated by calcn. of the isosteric heats of adsorption, which were larger across much of the range of coverage examd., compared to those of representative IRMOFs. Probably under low-loading conditions, the H2 adsorption behavior of MOFs can be improved by imparting larger charge gradients on the metal oxide units and adjusting the link metrics to constrict the pore dimensions; however, a large pore vol. is still a prerequisite feature.
  45. 45
    Siberio-Perez, D. Y.; Wong-Foy, A. G.; Yaghi, O. M.; Matzger, A. J. Raman spectroscopic investigation of CH4 and N2 adsorption in metal-organic frameworks. Chem. Mater. 2007, 19, 3681– 3685,  DOI: 10.1021/cm070542g
    Google Scholar
    45
    Raman Spectroscopic Investigation of CH4 and N2 Adsorption in Metal-Organic Frameworks
    Siberio-Perez, Diana Y.; Wong-Foy, Antek G.; Yaghi, Omar M.; Matzger, Adam J.
    Chemistry of Materials (2007), 19 (15), 3681-3685CODEN: CMATEX; ISSN:0897-4756. (American Chemical Society)
    The adsorption behavior of CH4 and N2 (298 K, 30 bar) in a series of isoreticular metal-org. frameworks (IRMOFs) was investigated by Raman spectroscopy. For CH4, the ν1 vibrational mode shifted to lower frequency by 7.6, 8.4, 11.0, 10.3, and 10.1 cm-1 from 2917 cm-1 when adsorbed to IRMOF-1, -6, -8, -11, and -18, resp. Along this same series, the adsorbed N2 stretch exhibited smaller shifts of 2.7, 3.1, 4.2, 4.1, and 3.7 cm-1. These shifts arise because of interactions within the framework pores, and not with the outer crystal surface. In all cases, Raman spectra at pressures up to 30 bar showed that satn. of the sorption sites does not occur. The obsd. shifts of the vibrational modes for each gas indicate different chem. environments within different IRMOFs, pointing to the important role the linkers play in the adsorption of gases.
  46. 46
    Duren, T.; Sarkisov, L.; Yaghi, O. M.; Snurr, R. Q. Design of new materials for methane storage. Langmuir 2004, 20, 2683– 2689,  DOI: 10.1021/la0355500
    Google Scholar
    46
    Design of new materials for methane storage
    Duren Tina; Sarkisov Lev; Yaghi Omar M; Snurr Randall Q
    Langmuir : the ACS journal of surfaces and colloids (2004), 20 (7), 2683-9 ISSN:0743-7463.
    One of the strategic goals of the modern automobile manufacturing industry is to replace gasoline and diesel with alternative fuels such as natural gas. In this report, we elucidate the desired characteristics of an optimal adsorbent for gas storage. The U.S. Department of Energy has outlined several requirements that adsorbents must fulfill for natural gas to become economically viable, with a key criterion being the amount adsorbed at 35 bar. We explore the adsorption characteristics of novel metal-organic materials (IRMOFs and molecular squares) and contrast them with the characteristics of two zeolites, MCM-41, and different carbon nanotubes. Using molecular simulations, we uncover the complex interplay of the factors influencing methane adsorption, especially the surface area, the capacity or free volume, the strength of the energetic interaction, and the pore size distribution. We also explain the extraordinary adsorption properties of IRMOF materials and propose new, not yet synthesized IRMOF structures with adsorption characteristics that are predicted to exceed the best experimental results to date by up to 36%.
  47. 47
    Walton, K. S.; Millward, A. R.; Dubbeldam, D.; Frost, H.; Low, J. J.; Yaghi, O. M.; Snurr, R. Q. Understanding inflection and steps in carbon dioxide adsorption isotherms in metal-organic frameworks. J. Am. Chem. Soc. 2007, 130, 406– 407,  DOI: 10.1021/ja076595g
    Google Scholar
    There is no corresponding record for this reference.
  48. 48
    Towns, J.; Cockerill, T.; Dahan, M.; Foster, I.; Gaither, K.; Grimshaw, A.; Hazlewood, V.; Lathrop, S.; Lifka, D.; Peterson, G. D.; Roskies, R.; Scott, J. R.; Wilkins-Diehr, N. XSEDE: accelerating scientific discovery. Comput. Sci. Eng. 2014, 16, 62– 74,  DOI: 10.1109/MCSE.2014.80
    Google Scholar
    There is no corresponding record for this reference.

Cited By

Click to copy section linkSection link copied!
Citation Statements
Explore this article's citation statements on scite.ai
powered by  Scite

This article is cited by 5 publications.

  1. Guozhao Liu, Hongyu Wang, Xiaolong Yang, Guanwen Su, Hongyuan Wei, Leping Dang. Insight on Crystal Morphology of ε-CL-20 in Binary Green Solvents: From Molecular Simulations to Experiments. Industrial & Engineering Chemistry Research 2025, 64 (3) , 1762-1773. https://doi.org/10.1021/acs.iecr.4c04400
  2. Felipe L. Oliveira, Binquan Luan, Pierre M. Esteves, Mathias Steiner, Rodrigo Neumann Barros Ferreira. pyMSER─An Open-Source Library for Automatic Equilibration Detection in Molecular Simulations. Journal of Chemical Theory and Computation 2024, 20 (19) , 8559-8568. https://doi.org/10.1021/acs.jctc.4c00417
  3. Hongyu Wang, Guozhao Liu, Guanwen Su, Hongyuan Wei, Leping Dang. Design of solvent systems for preparation of ε-CL-20 crystals with high sphericity assisted by molecular simulation. CrystEngComm 2025, 27 (4) , 547-558. https://doi.org/10.1039/D4CE01181C
  4. Yutong Liu, Yawen Dong, Hua Wu. Comprehensive overview of machine learning applications in MOFs: from modeling processes to latest applications and design classifications. Journal of Materials Chemistry A 2025, 13 (4) , 2403-2440. https://doi.org/10.1039/D4TA06740A
  5. Haixia He, Xiaoxi Guo, Yameng Wan, Jingwen Zhang, Xinyu Niu, Ruiai Wang, Qing Liu. Solubility, thermodynamic investigation and molecular dynamics simulation of guanidine hydrochloride in four binary solvents. Journal of Molecular Liquids 2023, 389 , 122902. https://doi.org/10.1016/j.molliq.2023.122902
Download PDF
Go to ACS Omega
Get e-Alerts
Get e-Alerts

ACS Omega

Cite this: ACS Omega 2022, 7, 48, 44470–44484
Click to copy citationCitation copied!
https://doi.org/10.1021/acsomega.2c06978
Published November 23, 2022

Copyright © 2022 The Authors. Published by American Chemical Society. This publication is licensed under

CC-BY-NC-ND 4.0 .

Article Views

2325

Altmetric

-

Citations

5
Learn about these metrics

Article Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to reflect usage leading up to the last few days.

Citations are the number of other articles citing this article, calculated by Crossref and updated daily. Find more information about Crossref citation counts.

The Altmetric Attention Score is a quantitative measure of the attention that a research article has received online. Clicking on the donut icon will load a page at altmetric.com with additional details about the score and the social media presence for the given article. Find more information on the Altmetric Attention Score and how the score is calculated.

  • Abstract

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 1

    Figure 1. Lennard-Jones potential for argon. The energy unit is in Kelvin (used by RASPA); this can be multiplied by the Boltzmann constant to convert to Joules per particle.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 2

    Figure 2. Monte Carlo and molecular dynamics cycles.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 3

    Figure 3. Flow diagram illustrating the process to create a new nanoHUB tool.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 4

    Figure 4. NanoHUB workspace.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 5

    Figure 5. Rappture builder GUI.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 6

    Figure 6. TortoiseSVN repository browser.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 7

    Figure 7. Left panel: A log–log plot of mean-squared displacement versus time that shows the separate ballistic, transition, and diffusive regimes. Right panel: A linear–linear plot of mean-squared displacement versus time using the starting and ending times selected by the algorithm that extracts the self-diffusion constant Ds from a data subset within the diffusive regime.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 8

    Figure 8. Screenshots of Tool 1 for calculating the self-diffusion constant of a gas in a MOF.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 9

    Figure 9. Screenshots of Tool 2 for calculating the absolute and excess adsorption of a gas in a MOF.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 10

    Figure 10. Screenshot of Tool 3 for calculating the Henry’s coefficients of n-alkanes in porous materials.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 11

    Figure 11. Screenshots of Tool 4 for calculating adsorption of a gas mixture in a porous material.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 12

    Figure 12. Screenshots of Tool 5 for calculating diffusion of a gas mixture in a porous material.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 13

    Figure 13. Screenshot of Tool 6 for calculating the void fraction of a porous material.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 14

    Figure 14. Screenshot of Tool 7 for calculating the surface area of a porous material.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 15

    Figure 15. Screenshots of Tool 8 for calculating the radial distribution function of a gas.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 16

    Figure 16. Screenshot of Tool 9 for calculating the adsorption energy histogram.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 17

    Figure 17. Screenshot of Tool 10 for calculating the density and total energy of a gas in the NPT ensemble.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 18

    Figure 18. Screenshots of Tool 11 for calculating gas adsorption in a MOF using the Gibbs ensemble.

    High Resolution Image
    Download MS PowerPoint Slide

    Figure 19

    Figure 19. Comparison of calculated to experimental isotherms of N2, methane, and CO2 gases adsorbed in the MOF IRMOF-1. The black curves are the calculated isotherms. The orange curves are the N2 (ref (45)), methane (ref (46)), and CO2 (ref (47)) experimental isotherms.

    High Resolution Image
    Download MS PowerPoint Slide

  • References

    This article references 48 other publications.

    1. 1
      Madhavan, K.; Zentner, L.; Farnsworth, V.; Shivarajapura, S.; Zentner, M.; Denny, N.; Klimeck, G. nanoHUB.org: cloud-based services for nanoscale modeling, simulation, and education. Nanotechnol. Rev. 2013, 2, 107– 117,  DOI: 10.1515/ntrev-2012-0043
      There is no corresponding record for this reference.
    2. 2
      Magana, A. J.; Brophy, S. P.; Bodner, G. M. An exploratory study of engineering and science students’ perceptions of nanoHUB.org simulations. Int. J. Eng. Educ. 2012, 28, 1019– 1032
      There is no corresponding record for this reference.
    3. 3
      Nanohub.org usage statistics. https://nanohub.org/usage, (accessed March 1, 2022).
      There is no corresponding record for this reference.
    4. 4
      Dubbeldam, D.; Calero, S.; Ellis, D. E.; Snurr, R. Q. RASPA: molecular simulation software for adsorption and diffusion in flexible nanoporous materials. Mol. Simul. 2016, 42, 81– 101,  DOI: 10.1080/08927022.2015.1010082
      4
      RASPA: molecular simulation software for adsorption and diffusion in flexible nanoporous materials
      Dubbeldam, David; Calero, Sofia; Ellis, Donald E.; Snurr, Randall Q.
      Molecular Simulation (2016), 42 (2), 81-101CODEN: MOSIEA; ISSN:0892-7022. (Taylor & Francis Ltd.)
      A new software package, RASPA, for simulating adsorption and diffusion of mols. in flexible nanoporous materials is presented. The code implements the latest state-of-the-art algorithms for mol. dynamics and Monte Carlo (MC) in various ensembles including symplectic/measure-preserving integrators, Ewald summation, configurational-bias MC, continuous fractional component MC, reactive MC and Baker's minimisation. We show example applications of RASPA in computing coexistence properties, adsorption isotherms for single and multiple components, self- and collective diffusivities, reaction systems and visualisation. The software is released under the GNU General Public License.
    5. 5
      Rogge, S. M. J.; Goeminne, R.; Demuynck, R.; Gutiérrez-Sevillano, J. J.; Vandenbrande, S.; Vanduyfhuys, L.; Waroquier, M.; Verstraelen, T.; Van Speybroeck, V. Modeling gas adsorption in flexible metal-organic frameworks via hybrid Monte Carlo/molecular dynamics schemes. Adv. Theory Simul. 2019, 2, 1800177  DOI: 10.1002/adts.201800177
      5
      Modeling Gas Adsorption in Flexible Metal-Organic Frameworks via Hybrid Monte Carlo/Molecular Dynamics Schemes
      Rogge, Sven M. J.; Goeminne, Ruben; Demuynck, Ruben; Gutierrez-Sevillano, Juan Jose; Vandenbrande, Steven; Vanduyfhuys, Louis; Waroquier, Michel; Verstraelen, Toon; Van Speybroeck, Veronique
      Advanced Theory and Simulations (2019), 2 (4), 1800177CODEN: ATSDCW; ISSN:2513-0390. (Wiley-VCH Verlag GmbH & Co. KGaA)
      Herein, a hybrid Monte Carlo (MC)/mol. dynamics (MD) simulation protocol that properly accounts for the extraordinary structural flexibility of metal-org. frameworks (MOFs) is developed and validated. This is vital to accurately predict gas adsorption isotherms and guest-induced flexibility of these materials. First, the performance of three recent models to predict adsorption isotherms and flexibility in MOFs is critically investigated. While these methods succeed in providing qual. insight in the gas adsorption process in MOFs, their accuracy remains limited as the intrinsic flexibility of these materials is very hard to account for. To overcome this challenge, a hybrid MC/MD simulation protocol that is specifically designed to handle the flexibility of the adsorbent, including the shape flexibility, is introduced, thereby unifying the strengths of the previous models. It is demonstrated that the application of this new protocol to the adsorption of neon, argon, xenon, methane, and carbon dioxide in MIL-53(Al), a prototypical flexible MOF, substantially decreases the inaccuracy of the obtained adsorption isotherms and predicted guest-induced flexibility. As a result, this method is ideally suited to rationalize the adsorption performance of flexible nanoporous materials at the mol. level, paving the way for the conscious design of MOFs as industrial adsorbents.
    6. 6
      Lin, L. C.; Lee, K.; Gagliardi, L.; Neaton, J. B.; Smit, B. Force-field development from electronic structure calculations with periodic boundary conditions: applications to gaseous adsorption and transport in metal-organic frameworks. J. Chem. Theory Comput. 2014, 10, 1477– 1488,  DOI: 10.1021/ct500094w
      6
      Force-Field Development from Electronic Structure Calculations with Periodic Boundary Conditions: Applications to Gaseous Adsorption and Transport in Metal-Organic Frameworks
      Lin, Li-Chiang; Lee, Kyuho; Gagliardi, Laura; Neaton, Jeffrey B.; Smit, Berend
      Journal of Chemical Theory and Computation (2014), 10 (4), 1477-1488CODEN: JCTCCE; ISSN:1549-9618. (American Chemical Society)
      A systematic and efficient methodol. to derive accurate (nonpolarizable) force fields from periodic d. functional theory (DFT) calcns. for use in classical mol. simulations has been developed. The methodol. requires reduced computation cost compared with other conventional ways. The whole process is performed self-consistently in a fully periodic system. The force fields derived by using this methodol. accurately predict the CO2 and H2O adsorption isotherms inside Mg-MOF-74, and is transferable to Zn-MOF-74; by replacing the Mg-CO2 interactions with the corresponding Zn-CO2 interactions, we obtain an accurate prediction of the corresponding isotherm. This methodol. was used to address the effect of water on the sepn. of flue gases in these materials. In general, the mixt. isotherms of CO2 and H2O calcd. with these derived force fields show a significant redn. in CO2 uptake with the existence of trace amts. of water vapor. The effect of water, however, was found to be quant. different between Mg- and Zn-MOF-74.
    7. 7
      Dubbeldam, D.; Walton, K. S.; Vlugt, T. J. H.; Calero, S. Design, parameterization, and implementation of atomic force fields for adsorption in nanoporous materials. Adv. Theory Simul. 2019, 2, 1900135  DOI: 10.1002/adts.201900135
      7
      Design, Parameterization, and Implementation of Atomic Force Fields for Adsorption in Nanoporous Materials
      Dubbeldam, David; Walton, Krista S.; Vlugt, Thijs J. H.; Calero, Sofia
      Advanced Theory and Simulations (2019), 2 (11), 1900135CODEN: ATSDCW; ISSN:2513-0390. (Wiley-VCH Verlag GmbH & Co. KGaA)
      A review. Mol. simulations are an excellent tool to study adsorption and diffusion in nanoporous materials. Examples of nanoporous materials are zeolites, carbon nanotubes, clays, metal-org. frameworks (MOFs), covalent org. frameworks (COFs) and zeolitic imidazolate frameworks (ZIFs). The mol. confinement these materials offer has been exploited in adsorption and catalysis for almost 50 years. Mol. simulations have provided understanding of the underlying shape selectivity, and adsorption and diffusion effects. Much of the reliability of the modeling predictions depends on the accuracy and transferability of the force field. However, flexibility and the chem. and structural diversity of MOFs add significant challenges for engineering force fields that are able to reproduce exptl. obsd. structural and dynamic properties. Recent developments in design, parameterization, and implementation of force fields for MOFs and zeolites are reviewed.
    8. 8
      Kaplan, I. G. Intermolecular Interactions: Physical Picture, Computational Methods and Model Potentials; John Wiley & Sons: West Sussex, U.K., 2006.
      There is no corresponding record for this reference.
    9. 9
      Rappe, A. K.; Casewit, C. J.; Colwell, K. S.; Goddard, W. A.; Skiff, W. M. UFF, a full periodic-table force-field for molecular mechanics and molecular-dynamics simulations. J. Am. Chem. Soc. 1992, 114, 10024– 10035,  DOI: 10.1021/ja00051a040
      9
      UFF, a full periodic table force field for molecular mechanics and molecular dynamics simulations
      Rappe, A. K.; Casewit, C. J.; Colwell, K. S.; Goddard, W. A., III; Skiff, W. M.
      Journal of the American Chemical Society (1992), 114 (25), 10024-35CODEN: JACSAT; ISSN:0002-7863.
      A new mol. mechanics force field, the Universal force field (UFF), is described wherein the force field parameters are estd. using general rules based only on the element, its hybridization and its connectivity. The force field functional forms, parameters, and generating formulas for the full periodic table are presented.
    10. 10
      Walters, E. T.; Mohebifar, M.; Johnson, E. R.; Rowley, C. N. Evaluating the London dispersion coefficients of protein force fields using the exchange-hole dipole moment model. J. Phys. Chem. B 2018, 122, 6690– 6701,  DOI: 10.1021/acs.jpcb.8b02814
      10
      Evaluating the London Dispersion Coefficients of Protein Force Fields Using the Exchange-Hole Dipole Moment Model
      Walters, Evan T.; Mohebifar, Mohamad; Johnson, Erin R.; Rowley, Christopher N.
      Journal of Physical Chemistry B (2018), 122 (26), 6690-6701CODEN: JPCBFK; ISSN:1520-5207. (American Chemical Society)
      London dispersion is one of the fundamental interactions involved in protein folding and dynamics. The popular CHARMM36, Amber ff14sb, and OPLS-AA force fields represent these interactions through the C6/r6 term of the Lennard-Jones potential, where the C6 parameters are assigned empirically. Here, dispersion coeffs. of these three force fields are shown to be roughly 50% larger than values calcd. using the quantum-mech.-derived exchange-hole dipole moment (XDM) model. The CHARMM36 and Amber OL15 force fields for nucleic acids also exhibit this trend. The hydration energies of the side-chain models were calcd. using REMD-TI for the CHARMM36, Amber ff14sb, and OPLS-AA force fields. These force fields predict side-chain hydration energies that are in generally good agreement with the exptl. values, which suggests the total strength of aq. dispersion interactions is correct, despite C6 coeffs. that are considerably larger than XDM predicts. An anal. expression for the dispersion hydration energy using XDM coeffs. shows that higher-order dispersion terms (i.e., C8 and C10) account for roughly 37.5% of the hydration energy of methane. This suggests that the C6 dispersion coeffs. used in contemporary force fields are elevated to account for the neglected higher-order terms.
    11. 11
      Jones, J. E. On the determination of molecular fields - II From the equation of state of a gas. Proc. R. Soc. London, Ser. A 1924, 106, 463– 477,  DOI: 10.1098/rspa.1924.0082
      11
      The determination of molecular fields (II) From the equation of state of a gas
      Jones, J. R.
      Proceedings of the Royal Society of London, Series A: Mathematical, Physical and Engineering Sciences (1924), 106 (), 463-77CODEN: PRLAAZ; ISSN:1364-5021.
      The virial coeffs. are evaluated for an equation of state of gases for the case of a mol. with attractive and repelling forces, as described above. The values obtained for the consts. of these forces for A agree fairly well with those calcd. from the viscosity data. In an appendix, recalcn. on the basis of more recent data on the isotherms of A give better agreement, pointing to 14 1/3 as the best value for the index of the repulsive force.
    12. 12
      Boda, D.; Henderson, D. The effects of deviations from Lorentz–Berthelot rules on the properties of a simple mixture. Mol. Phys. 2008, 106, 2367– 2370,  DOI: 10.1080/00268970802471137
      12
      The effects of deviations from Lorentz-Berthelot rules on the properties of a simple mixture
      Boda, Dezso; Henderson, Douglas
      Molecular Physics (2008), 106 (20), 2367-2370CODEN: MOPHAM; ISSN:0026-8976. (Taylor & Francis Ltd.)
      Generally, the parameters in the interaction potential between like mols. in a mixt. can be detd. in a relatively straightforward manner from the properties of the pure components. However, the detn. of the parameters in the interaction potential between the unlike pairs in the mixts. is more difficult. As a result, these parameters are usually estd. from avs. of the like parameters. The most common recipes are the Lorentz-Berthelot mixing rules, where the energy and mol. size parameters are presumed to be geometric and arithmetic avs., resp. There have been some studies of the consequences of deviations from the energy rule but almost no studies of the consequences of deviations from the size rule. Here, we study the effects of deviations from both rules on the radial distribution functions of a simple mixt. We find from simulations that, for this mixt., the effect of deviations from the energy rule on the radial distribution function are rather small but that the effect of deviations from the size rule can be significant and are interesting.
    13. 13
      Berthelot, D. Sur le mélange des gaz. Compt. Rendus 1898, 126, 1703– 1706
      13
      On mixtures of gases
      Berthelot, D.
      Comptes Rendus Hebdomadaires des Seances de l'Academie des Sciences (1898), 126 (), 1703CODEN: COREAF ISSN:.
      The author proposes a formula of the van der Waals type for a mixture of two gases and shows how the constants A and B for the mixture can be calculated from the corresponding constants for the single gases.
    14. 14
      Lorentz, H. A. Ueber die anwendung des satzes vom virial in der kinetischen theorie der gase. Ann. Phys. 1881, 248, 127– 136,  DOI: 10.1002/andp.18812480110
      There is no corresponding record for this reference.
    15. 15
      Dubbeldam, D.; Calero, S.; Vlugt, T. J. H.; Ellis, D. E.; Snurr, R. Q. RASPA 2.0: Molecular Software Package for Adsorption and Diffusion in (Flexible) Nanoporous Materials , program manual, March 2015, pp 1− 145.
      There is no corresponding record for this reference.
    16. 16
      Dubbeldam, D.; Walton, K. S.; Ellis, D. E.; Snurr, R. Q. Exceptional negative thermal expansion in isoreticular metal-organic frameworks. Angew. Chem., Int. Ed. 2007, 46, 4496– 4499,  DOI: 10.1002/anie.200700218
      16
      Exceptional negative thermal expansion in isoreticular metal-organic frameworks
      Dubbeldam, David; Walton, Krista S.; Ellis, Donald E.; Snurr, Randall Q.
      Angewandte Chemie, International Edition (2007), 46 (24), 4496-4499CODEN: ACIEF5; ISSN:1433-7851. (Wiley-VCH Verlag GmbH & Co. KGaA)
      A new model for flexible frameworks is used to simulate the structures and adsorption properties of isoreticular metal-org. frameworks (IRMOFs), such as IRMOF-1. The structural simulations suggest that the IRMOFs have neg. thermal-expansion coeffs. over their full temp. ranges of stability.
    17. 17
      Chen, B.; Siepmann, J. I. Transferable potentials for phase equilibria. 3. Explicit-hydrogen description of normal alkanes. J. Phys. Chem. B 1999, 103, 5370– 5379,  DOI: 10.1021/jp990822m
      17
      Transferable potentials for phase equilibria. 3. Explicit-hydrogen description of normal alkanes
      Chen, Bin; Siepmann, J. Ilja
      Journal of Physical Chemistry B (1999), 103 (25), 5370-5379CODEN: JPCBFK; ISSN:1089-5647. (American Chemical Society)
      Motivated by shortcomings of the available united-atom models for alkanes, a new explicit-H model for n-alkanes (TraPPE-EH, transferable potentials for phase equil.-explicit H) is developed from fitting to 1-component fluid-phase properties. In addn. to Lennard-Jones sites on C atoms, this model utilizes Lennard-Jones sites on the centers of C-H bonds. Configurational-bias Monte Carlo simulations in the Gibbs and canonical ensembles were carried out to calc. the 1-component vapor-liq. phase equil. for CH4 to n-dodecane, to det. the phase diagram of supercrit. C2H6 and n-heptane mixts., to obtain the Gibbs free energies of transfer for n-pentane (I) and n-hexane between He vapor and n-heptane liq. phases, and to study the high-pressure region of the equation of state for I and n-decane. The explicit-H representation, with its more faithful description of the mol. shape of alkanes, allows us to find a set of Lennard-Jones parameters that yields significantly better agreement with expt. for 1- and multicomponent phase equil. than our united-atom alkane model, but the price is higher computational cost.
    18. 18
      Eggimann, B. L.; Sunnarborg, A. J.; Stern, H. D.; Bliss, A. P.; Siepmann, J. I. An online parameter and property database for the TraPPE force field. Mol. Simul. 2014, 40, 101– 105,  DOI: 10.1080/08927022.2013.842994
      18
      An online parameter and property database for the TraPPE force field
      Eggimann, Becky L.; Sunnarborg, Amara J.; Stern, Hudson D.; Bliss, Andrew P.; Siepmann, J. Ilja
      Molecular Simulation (2014), 40 (1-3), 101-105CODEN: MOSIEA; ISSN:0892-7022. (Taylor & Francis Ltd.)
      A review. The transferable potentials for phase equil. (TraPPE) force field aims to be accurate, computationally efficient and applicable to a wide range of chem. compds., state points and thermophys. properties. When new users wish to implement TraPPE models into their chosen simulation program, they face several obstacles: the TraPPE models are dispersed over many sep. publications and misinterpretations of the primary literature are possible; the TraPPE force field makes specific choices for std. conventions that may require non-trivial code modifications for some simulation software. Therefore, the TraPPE developers report here a resource website and online searchable parameter and property database designed to provide new and experienced users with tools for successful implementation and validation (http://www.chem.umn.edu/groups/siepmann/trappe/).
    19. 19
      Martin, M. G.; Siepmann, J. I. Transferable potentials for phase equilibria. 1. United-atom description of n-alkanes. J. Phys. Chem. B 1998, 102, 2569– 2577,  DOI: 10.1021/jp972543+
      19
      Transferable Potentials for Phase Equilibria. 1. United-Atom Description of n-Alkanes
      Martin, Marcus G.; Siepmann, J. Ilja
      Journal of Physical Chemistry B (1998), 102 (14), 2569-2577CODEN: JPCBFK; ISSN:1089-5647. (American Chemical Society)
      A new set of united-atom Lennard-Jones interaction parameters for n-alkanes is proposed from fitting to crit. temps. and satd. liq. densities. Configurational-bias Monte Carlo simulations in the Gibbs ensemble were carried out to det. the vapor-liq. coexistence curves for methane to dodecane using three united-atom force fields: OPLS [Jorgensen, et al. J. Am. Chem. Soc. 1984, 106, 813], SKS [Siepmann, et al. Nature 1993, 365, 330], and TraPPE. Std. specific densities and the high-pressure equation-of-state for the transferable potentials for phase equil. (TraPPE) model were studied by simulations in the isobaric-isothermal and canonical ensembles, resp. It is found that one set of Me and methylene parameters is sufficient to accurately describe the fluid phases of all n-alkanes with two or more carbon atoms. Whereas other n-alkane force fields employ Me groups that are either equal or larger in size than the methylene groups, it is demonstrated here that using a smaller Me group yields a better fit to the set of exptl. data. As should be expected from an effective pair potential, the new parameters do not reproduce exptl. second virial coeffs. Satd. vapor pressures and densities show small, but systematic deviation from the exptl. data.
    20. 20
      Potoff, J. J.; Siepmann, J. I. Vapor–liquid equilibria of mixtures containing alkanes, carbon dioxide, and nitrogen. AIChE J. 2001, 47, 1676– 1682,  DOI: 10.1002/aic.690470719
      20
      Vapor-liquid equilibria of mixtures containing alkanes, carbon dioxide, and nitrogen
      Potoff, Jeffrey J.; Siepmann, J. Ilja
      AIChE Journal (2001), 47 (7), 1676-1682CODEN: AICEAC; ISSN:0001-1541. (American Institute of Chemical Engineers)
      New force fields for carbon dioxide and nitrogen are introduced that quant. reproduce the vapor-liq. equil. (VLE) of the neat systems and their mixts. with alkanes. In addn. to the usual VLE calcns. for pure CO2 and N2, calcns. of the binary mixts. with propane were used in the force-field development to achieve a good balance between dispersive and electrostatic (quadrupole-quadrupole) interactions. The transferability of the force fields was then assessed from calcns. of the VLE for the binary mixts. with n-hexane, the binary mixt. of CO2/N2, and the ternary mixt. of CO2/N2/propane. The VLE calcns. were carried out using configurational-bias Monte Carlo simulations in either the grand canonical ensemble with histogram-reweighting or in the Gibbs ensemble.
    21. 21
      Darkrim, F.; Levesque, D. Monte Carlo simulations of hydrogen adsorption in single-walled carbon nanotubes. J. Chem. Phys. 1998, 109, 4981– 4984,  DOI: 10.1063/1.477109
      21
      Monte Carlo simulations of hydrogen adsorption in single-walled carbon nanotubes
      Darkrim, Farida; Levesque, Dominique
      Journal of Chemical Physics (1998), 109 (12), 4981-4984CODEN: JCPSA6; ISSN:0021-9606. (American Institute of Physics)
      Within the framework of a study on the properties of carbon nanotubes, a promising new material, we performed numerical simulation of hydrogen adsorption at room temp. in single-walled nanotubes. The structure of this material is favorable to the adsorption phenomenon because of the narrow size distribution of the nanotube diams., which have dimensions on the order of the range of the carbon attractive interaction. We discuss the influence of the single-walled carbon nanotube diams. on the relative arrangement of carbon atoms and hydrogen mols. within an array of parallel single-walled carbon nanotubes. We also studied the influence on adsorption of the distance between the nearest-neighbor nanotubes.
    22. 22
      Zhang, L.; Siepmann, J. I. Direct calculation of Henry’s law constants from Gibbs ensemble Monte Carlo simulations: nitrogen, oxygen, carbon dioxide and methane in ethanol. Theor. Chem. Acc. 2006, 115, 391– 397,  DOI: 10.1007/s00214-005-0073-1
      22
      Direct calculation of Henry's law constants from Gibbs ensemble Monte Carlo simulations: nitrogen, oxygen, carbon dioxide and methane in ethanol
      Zhang, Ling; Siepmann, J. Ilja
      Theoretical Chemistry Accounts (2006), 115 (5), 391-397CODEN: TCACFW; ISSN:1432-881X. (Springer GmbH)
      Configurational-bias Monte Carlo simulations in the Gibbs ensemble were used to calc. Henry's law consts., Ostwald solubilities, and Gibbs free energies of transfer for oxygen, nitrogen, methane, and carbon dioxide in ethanol at 323 and 373 K. These three soly. descriptors can be expressed as functions of mech. properties that are directly observable in the Gibbs ensemble approach, thereby allowing for very precise detn. of the descriptors. Addnl., the Henry's law consts. of multiple solutes can be computed from a single simulation. Most of the simulations were carried out for systems contg. 1000 solvent and up to 8 solute mols., and further simulations using either 500 or 2000 solvent mols. point to negligible system size effects. A comparison with exptl. data shows that the united-atom version of the transferable potential for phase equil. force field yields Henry's law consts. that reproduce well the differences between the four solutes and the changes upon increase of the temp.
    23. 23
      Mayo, S. L.; Olafson, B. D.; Goddard, W. A. DREIDING - A generic force field for molecular simulations. J. Phys. Chem. 1990, 94, 8897– 8909,  DOI: 10.1021/j100389a010
      23
      DREIDING: a generic force field for molecular simulations
      Mayo, Stephen L.; Olafson, Barry D.; Goddard, William A., III
      Journal of Physical Chemistry (1990), 94 (26), 8897-909CODEN: JPCHAX; ISSN:0022-3654.
      The parameters are given for a new generic force field, DREIDING, that is useful for predicting structures and dynamics of org., biol., and main-group inorg. mols. The philosophy in DREIDING is to use general force consts. and geometry parameters based on simple hybridization considerations rather than individual force consts. and geometric parameters that depend on the particular combination of atoms involved in the bond, angle, or torsion terms. Thus all bond distances are derived from at. radii, and there is only one force const. each for bonds, angles, and inversions and only six different values for torsional barriers. Parameters are defined for all possible combinations of atoms and new atoms can be added to the force field rather simply. The parameters are given for the "nonmetallic" main-group elements (B, C, N, O, F columns for the C, Si, Ge, and Sn rows) plus H and a few metals (Na, Ca, Zn, Fe). The accuracy of the DREIDING force field is tested by comparing with (i) 76 accurately detd. crystal structures of org. compds. involving H, C, N, O, F, P, S, Cl, and Br, (ii) rotational barriers of a no. of mols., and (iii) relative conformational energies and barriers of a no. of mols. There is excellent agreement.
    24. 24
      Bai, P.; Tsapatsis, M.; Siepmann, J. I. TraPPE-zeo: Transferable potentials for phase equilibria force field for all-silica zeolites. J. Phys. Chem. C 2013, 117, 24375– 24387,  DOI: 10.1021/jp4074224
      24
      TraPPE-zeo: Transferable Potentials for Phase Equilibria Force Field for All-Silica Zeolites
      Bai, Peng; Tsapatsis, Michael; Siepmann, J. Ilja
      Journal of Physical Chemistry C (2013), 117 (46), 24375-24387CODEN: JPCCCK; ISSN:1932-7447. (American Chemical Society)
      The transferable potentials for phase equil. (TraPPE) force field is extended to all-silica zeolites. This novel force field is parametrized to match the exptl. adsorption isotherms of n-heptane, propane, carbon dioxide, and ethanol with the Lennard-Jones parameters for sorbate-framework interactions detd. in a consistent manner using the Lorentz-Berthelot combining rules as for other parts of the TraPPE force field. The TraPPE-zeo force field allows for accurate predictions for both adsorption and diffusion of alkanes, alcs., carbon dioxide, and water over a wide range of pressures and temps. To achieve transferability to a wider range of mol. types, ranging from nonpolar to dipolar and hydrogen-bonding compds., Lennard-Jones interaction sites and partial charges are placed at both the oxygen and the silicon atoms of the zeolite lattice, which allows for a better balance of dispersive and 1st-order electrostatic interactions than is achievable with the Lennard-Jones potential used only for the oxygen atoms. The use of the Lorentz-Berthelot combining rules for unlike interactions makes the TraPPE-zeo force field applicable to any sorbate as long as the relevant TraPPE sorbate-sorbate parameters are available. The TraPPE-zeo force field allows for greatly improved predictive power compared to force fields that explicitly tabulate the individual cross-interaction parameters.
    25. 25
      Dubbeldam, D.; Torres-Knoop, A.; Walton, K. S. On the inner workings of Monte Carlo codes. Mol. Simul. 2013, 39, 1253– 1292,  DOI: 10.1080/08927022.2013.819102
      25
      On the inner workings of Monte Carlo codes
      Dubbeldam, David; Torres-Knoop, Ariana; Walton, Krista S.
      Molecular Simulation (2013), 39 (14-15), 1253-1292CODEN: MOSIEA; ISSN:0892-7022. (Taylor & Francis Ltd.)
      We review state-of-the-art Monte Carlo (MC) techniques for computing fluid coexistence properties (Gibbs simulations) and adsorption simulations in nanoporous materials such as zeolites and metal-org. frameworks. Conventional MC is discussed and compared to advanced techniques such as reactive MC, configurational-bias Monte Carlo and continuous fractional MC. The latter technique overcomes the problem of low insertion probabilities in open systems. Other modern methods are (hyper-)parallel tempering, Wang-Landau sampling and nested sampling. Details on the techniques and acceptance rules as well as to what systems these techniques can be applied are provided. We highlight consistency tests to help validate and debug MC codes.
    26. 26
      Heinen, J.; Burtch, N. C.; Walton, K. S.; Dubbeldam, D. Flexible force field parameterization through fitting on the ab initio-derived elastic tensor. J. Chem. Theory Comput. 2017, 13, 3722– 3730,  DOI: 10.1021/acs.jctc.7b00310
      26
      Flexible Force Field Parameterization through Fitting on the Ab Initio-Derived Elastic Tensor
      Heinen, Jurn; Burtch, Nicholas C.; Walton, Krista S.; Dubbeldam, David
      Journal of Chemical Theory and Computation (2017), 13 (8), 3722-3730CODEN: JCTCCE; ISSN:1549-9618. (American Chemical Society)
      Constructing functional forms and their corresponding force field parameters for the metal-linker interface of metal-org. frameworks is challenging. We propose to fit these parameters on the elastic tensor, computed from ab initio d. functional theory calcns. The advantage of this top-down approach is that it becomes evident if functional forms are missing when components of the elastic tensor are off. As a proof-of-concept, a new flexible force field for MIL-47(V) is derived. Neg. thermal expansion is obsd. and framework flexibility has a negligible effect on adsorption and transport properties for small guest mols. We believe this force field parameterization approach can serve as a useful tool for developing accurate flexible force field models that capture the correct mech. behavior of the full periodic structure.
    27. 27
      Heinen, J.; Dubbeldam, D. On flexible force fields for metal-organic frameworks: Recent developments and future prospects. Wiley Interdiscip. Rev.: Comput. Mol. Sci. 2018, 8, e1363  DOI: 10.1002/wcms.1363
      There is no corresponding record for this reference.
    28. 28
      Allen, M. P.; Tildesley, D. J. Computer Simulation of Liquids, 2nd ed.; Oxford University Press: Oxford, U.K., 2017.
      There is no corresponding record for this reference.
    29. 29
      Frenkel, D.; Smit, B. Understanding Molecular Simulation: From Algorithms to Applications, 2nd ed.; Academic Press: San Diego, California, 2002.
      There is no corresponding record for this reference.
    30. 30
      Hastings, W. K. Monte Carlo sampling methods using Markov chains and their applications. Biometrika 1970, 57, 97– 109,  DOI: 10.1093/biomet/57.1.97
      There is no corresponding record for this reference.
    31. 31
      Metropolis, N.; Rosenbluth, A. W.; Rosenbluth, M. N.; Teller, A. H.; Teller, E. Equation of state calculations by fast computing machines. J. Chem. Phys. 1953, 21, 1087– 1092,  DOI: 10.1063/1.1699114
      31
      Equation-of-state calculations by fast computing machines
      Metropolis, Nicholas; Rosenbluth, Arianna W.; Rosenbluth, Marshall N.; Teller, Augusta H.; Teller, Edward
      Journal of Chemical Physics (1953), 21 (), 1087-92CODEN: JCPSA6; ISSN:0021-9606.
      A general method, suitable for fast computing machines, is described for investigating such properties as equations of state for substances consisting of interacting individual mols. The method consists of a modified Monte Carlo integration over configuration space. Results for the 2-dimensional rigid-sphere system were obtained on the Los Alamos MANIAC. These results were compared to the free-vol. equation of state and to a 4-term virial coeff. expansion.
    32. 32
      Hirotani, A.; Mizukami, K.; Miura, R.; Takaba, H.; Miya, T.; Fahmi, A.; Stirling, A.; Kubo, M.; Miyamoto, A. Grand canonical Monte Carlo simulation of the adsorption of CO2 on silicalite and NaZSM-5. Appl. Surf. Sci. 1997, 120, 81– 84,  DOI: 10.1016/S0169-4332(97)00222-5
      32
      Grand canonical Monte Carlo simulation of the adsorption of CO2 on silicalite and NaZSM-5
      Hirotani, Akiyasu; Mizukami, Koichi; Miura, Ryuji; Takaba, Hiromitsu; Miya, Takeshi; Fahmi, Adil; Stirling, Andras; Kubo, Momoji; Miyamoto, Akira
      Applied Surface Science (1997), 120 (1/2), 81-84CODEN: ASUSEE; ISSN:0169-4332. (Elsevier)
      The adsorption of carbon dioxide in silicalite and NaZSM-5 zeolite has been studied using new Monte Carlo software. In this program, sodium cations and framework are movable during the simulation. The calcd. adsorption isotherms are in good agreement with the exptl. results. The energy distribution of carbon dioxide over silicalite and NaZSM-5 shows that the increase of the adsorption energy for NaZSM-5 is mainly due the elec. field generated by sodium cations.
    33. 33
      Gao, G.; Wang, W. Gibbs Ensemble Monte Carlo simulation of binary vapor-liquid equilibria for CFC alternatives. Fluid Phase Equilib. 1997, 130, 157– 166,  DOI: 10.1016/S0378-3812(96)03195-0
      33
      Gibbs Ensemble Monte Carlo simulation of binary vapor-liquid equilibria for CFC alternatives
      Gao, Guangtu; Wang, Wenchuan
      Fluid Phase Equilibria (1997), 130 (1-2), 157-166CODEN: FPEQDT; ISSN:0378-3812. (Elsevier)
      The Gibbs Ensemble Monte Carlo (GEMC) simulation method has been used for vapor-liq. equil. calcns. of the binary systems R22-R142b and R22-R152a, which are considered to be promising mixt. refrigerants for replacing R12. All the mols. have been treated in terms of an effective Lennard-Jones potential, with temp. dependent parameters regressed by fitting vapor pressures and satd. liq. densities at various temps. for pure substances of interest. Good agreement between exptl. and calcd. data from the vol.-translation Peng-Robinson equation of state and the simulated results, including the compns., molar volumes for both the vapor and liq. phases, and heats of vaporization, indicates that the GEMC method can be applied to the description of phase behavior for these CFC alternative binary systems.
    34. 34
      Slepoy, A.; Thompson, A. P.; Plimpton, S. J. A constant-time kinetic Monte Carlo algorithm for simulation of large biochemical reaction networks. J. Chem. Phys. 2008, 20, 205101  DOI: 10.1063/1.2919546
      There is no corresponding record for this reference.
    35. 35
      Dove, M. T. An introduction to atomistic simulation methods. Seminarios de la SEM 2008, 4, 7– 37
      There is no corresponding record for this reference.
    36. 36
      Braun, E.; Gilmer, J.; Mayes, H. B.; Mobley, D. L.; Monroe, J. I.; Prasad, S.; Zuckerman, D. M. Best practices for foundations in molecular simulations. Living J. Comput. Mol. Sci. 2019, 1, 5957  DOI: 10.33011/livecoms.1.1.5957
      There is no corresponding record for this reference.
    37. 37
      Abouelnasr, M. K. F.; Smit, B. Diffusion in confinement: kinetic simulations of self- and collective diffusion behavior of adsorbed gases. Phys. Chem. Chem. Phys. 2012, 14, 11600– 11609,  DOI: 10.1039/c2cp41147d
      37
      Diffusion in confinement: kinetic simulations of self- and collective diffusion behavior of adsorbed gases
      Abouelnasr, Mahmoud K. F.; Smit, Berend
      Physical Chemistry Chemical Physics (2012), 14 (33), 11600-11609CODEN: PPCPFQ; ISSN:1463-9076. (Royal Society of Chemistry)
      The self- and collective-diffusion behaviors of adsorbed methane, helium, and isobutane in zeolite frameworks LTA, MFI, AFI, and SAS were examd. at various concns. using a range of mol. simulation techniques including Mol. Dynamics (MD), Monte Carlo (MC), Bennett-Chandler (BC), and kinetic Monte Carlo (kMC). This paper has three main results. (1) A novel model for the process of adsorbate movement between two large cages was created, allowing the formulation of a mixing rule for the re-crossing coeff. between two cages of unequal loading. The predictions from this mixing rule were found to agree quant. with explicit simulations. (2) A new approach to the dynamically cor. Transition State Theory method to anal. calc. self-diffusion properties was developed, explicitly accounting for nanoscale fluctuations in concn. This approach was demonstrated to quant. agree with previous methods, but is uniquely suited to be adapted to a kMC simulation that can simulate the collective-diffusion behavior. (3) While at low and moderate loadings the self- and collective-diffusion behaviors in LTA are obsd. to coincide, at higher concns. they diverge. A change in the adsorbate packing scheme was shown to cause this divergence, a trait which is replicated in a kMC simulation that explicitly models this behavior. These phenomena were further investigated for isobutane in zeolite MFI, where MD results showed a sepn. in self- and collective- diffusion behavior that was reproduced with kMC simulations.
    38. 38
      Skoulidas, A. I.; Sholl, D. S. Self-diffusion and transport diffusion of light gases in metal-organic framework materials assessed using molecular dynamics simulations. J. Phys. Chem. B 2005, 109, 15760– 15768,  DOI: 10.1021/jp051771y
      38
      Self-Diffusion and Transport Diffusion of Light Gases in Metal-Organic Framework Materials Assessed Using Molecular Dynamics Simulations
      Skoulidas, Anastasios I.; Sholl, David S.
      Journal of Physical Chemistry B (2005), 109 (33), 15760-15768CODEN: JPCBFK; ISSN:1520-6106. (American Chemical Society)
      Metal-org. framework (MOF) materials pose an interesting alternative to more traditional nanoporous materials for a variety of sepn. processes. Sepn. processes involving nanoporous materials can be controlled by either adsorption equil., diffusive transport rates, or a combination of these factors. Adsorption equil. was studied for a variety of gases in MOFs, but almost nothing is currently known about mol. diffusion rates in MOFs. The authors used equil. mol. dynamics (MD) to probe the self-diffusion and transport diffusion of a no. of small gas species in several MOFs as a function of pore loading at room temp. Specifically, the authors have studied Ar, CH4, CO2, N2, and H2 diffusion in MOF-5. The diffusion of Ar in MOF-2, MOF-3, and Cu-BTC was assessed in a similar manner. Results greatly expand the range of MOFs for which data describing mol. diffusion is available. The authors discuss the prospects for exploiting mol. transport properties in MOFs in practical sepn. processes and the future role of MD simulations in screening families of MOFs for these processes.
    39. 39
      Werder, T.; Walther, J. H.; Jaffe, R. L.; Halicioglu, T.; Koumoutsakos, P. On the water-carbon interaction for use in molecular dynamics simulations of graphite and carbon nanotubes. J. Phys. Chem. B 2003, 107, 1345– 1352,  DOI: 10.1021/jp0268112
      39
      On the Water-Carbon Interaction for Use in Molecular Dynamics Simulations of Graphite and Carbon Nanotubes
      Werder, T.; Walther, J. H.; Jaffe, R. L.; Halicioglu, T.; Koumoutsakos, P.
      Journal of Physical Chemistry B (2003), 107 (6), 1345-1352CODEN: JPCBFK; ISSN:1520-6106. (American Chemical Society)
      A systematic mol. dynamics study shows that the contact angle of a H2O droplet on graphite changes significantly as a function of the H2O-C interaction energy. Together with the observation that a linear relation can be established between the contact angle and the H2O monomer binding energy on graphite, a new route to calibrate interaction potential parameters is presented. Through a variation of the droplet size at 1000-17,500 H2O mols., the authors det. the line tension to be pos. and ∼2 × 10-10 J/m. To recover a macroscopic contact angle of 86°, a H2O monomer binding energy of -6.33 kJ mol-1 is required, which is obtained by applying a C-oxygen Lennard-Jones potential with the parameters εCO = 0.392 kJ mol-1 and σCO = 3.19 Å. For this new H2O-C interaction potential, the authors present d. profiles and hydrogen bond distributions for a H2O droplet on graphite.
    40. 40
      Henry, A. S.; Chen, G. Spectral phonon transport properties of silicon based on molecular dynamics simulations and lattice dynamics. J. Comput. Theor. Nanosci. 2008, 5, 141– 152,  DOI: 10.1166/jctn.2008.2454
      40
      Spectral phonon transport properties of silicon based on molecular dynamics simulations and lattice dynamics
      Henry, Asegun S.; Chen, Gang
      Journal of Computational and Theoretical Nanoscience (2008), 5 (2), 141-152CODEN: JCTNAB; ISSN:1546-1955. (American Scientific Publishers)
      Although the thermal cond. of silicon has been studied before, current estns. for the phonon mean free paths have not provided full explanation of the strong size effects exptl. obsd. for various silicon micro and nanostructures. Since phonon relaxation time models are mostly semi-empirical, the mean free paths cannot be detd. reliably and questions remain as to which polarizations, frequencies, and wavelengths are dominant heat carriers. Here we have used a combination of equil. mol. dynamics simulations and lattice dynamics calcns. to fully detail the spectral dependence of phonon transport properties in bulk silicon. By considering the frequency dependence of the sp. heat, group velocities, and mean free paths, we address these unresolved questions and examine the errors assocd. with isotropic and frequency averaged approxns. Simulation details, such as the convergence of results on the simulation time and extn. of phonon transport properties in different crystallog. directions, are also discussed.
    41. 41
      Karplus, M.; McCammon, J. A. Molecular dynamics simulations of biomolecules. Nat. Struct. Biol. 2002, 9, 646– 652,  DOI: 10.1038/nsb0902-646
      41
      Molecular dynamics simulations of biomolecules
      Karplus, Martin; McCammon, J. Andrew
      Nature Structural Biology (2002), 9 (9), 646-652CODEN: NSBIEW; ISSN:1072-8368. (Nature Publishing Group)
      A review. Mol. dynamics simulations are important tools for understanding the phys. basis of the structure and function of biol. macromols. The early view of proteins as relatively rigid structures has been replaced by a dynamic model in which the internal motions and resulting conformational changes play an essential role in their function. This review presents a brief description of the origin and early uses of biomol. simulations. It then outlines some recent studies that illustrate the utility of such simulations and closes with a discussion of their ever-increasing potential for contributing to biol.
    42. 42
      Umeh, J. C. Development of Simulation Tools on Nanuhub as Learning Modules. Master’s thesis; New Mexico State University: Las Cruces, NM, 2019.
      There is no corresponding record for this reference.
    43. 43
      Humbert, M. T.; Zhang, Y.; Maginn, E. J. PyLAT: Python LAMMPS analysis tools. J. Chem. Inf. Model. 2019, 59, 1301– 1305,  DOI: 10.1021/acs.jcim.9b00066
      43
      PyLAT: Python LAMMPS Analysis Tools
      Humbert, Michael T.; Zhang, Yong; Maginn, Edward J.
      Journal of Chemical Information and Modeling (2019), 59 (4), 1301-1305CODEN: JCISD8; ISSN:1549-9596. (American Chemical Society)
      Reproducibility and accuracy have become increasingly important issues for the mol. simulation community. The availability of validated open-source postprocessing tools to analyze simulation trajectories and compute properties is key to helping researchers conduct more accurate and reproducible simulations. Here the authors describe a suite of open-source Python-based postprocessing routines the authors have developed called PyLAT. PyLAT is compatible with the popular mol. dynamics package LAMMPS and enables users to compute viscosities, self-diffusivities, ionic conductivities, mol. or ion pair lifetimes, dielec. consts., and radial distribution functions using best-practice methods.
    44. 44
      Rowsell, J. L. C.; Yaghi, O. M. Effects of functionalization, catenation, and variation of the metal oxide and organic linking units on the low-pressure hydrogen adsorption properties of metal-organic frameworks. J. Am. Chem. Soc. 2006, 128, 1304– 1315,  DOI: 10.1021/ja056639q
      44
      Effects of Functionalization, Catenation, and Variation of the Metal Oxide and Organic Linking Units on the Low-Pressure Hydrogen Adsorption Properties of Metal-Organic Frameworks
      Rowsell, Jesse L. C.; Yaghi, Omar M.
      Journal of the American Chemical Society (2006), 128 (4), 1304-1315CODEN: JACSAT; ISSN:0002-7863. (American Chemical Society)
      The dihydrogen adsorption isotherms of eight metal-org. frameworks (MOFs), measured at 77 K up to a pressure of 1 atm, were examd. for correlations with their structural features. All materials display approx. Type I isotherms with no hysteresis, and satn. was not reached for any of the materials under these conditions. Among the six iso-reticular MOFs (IRMOFs) studied, the catenated materials exhibit the largest capacities on a molar basis, up to 9.8 H2 per formula unit. The addn. of functional groups (-Br, -NH2, -C2H4-) to the phenylene links of IRMOF-1 (MOF-5), or their replacement with thieno[3,2-b]thiophene moieties in IRMOF-20, altered the adsorption behavior by a minor amt. despite large variations in the pore vols. of the resulting materials. But replacement of the metal oxide units with those contg. coordinatively unsatd. metal sites resulted in greater H2 uptake. The enhanced affinities of these materials, MOF-74 and HKUST-1, were further demonstrated by calcn. of the isosteric heats of adsorption, which were larger across much of the range of coverage examd., compared to those of representative IRMOFs. Probably under low-loading conditions, the H2 adsorption behavior of MOFs can be improved by imparting larger charge gradients on the metal oxide units and adjusting the link metrics to constrict the pore dimensions; however, a large pore vol. is still a prerequisite feature.
    45. 45
      Siberio-Perez, D. Y.; Wong-Foy, A. G.; Yaghi, O. M.; Matzger, A. J. Raman spectroscopic investigation of CH4 and N2 adsorption in metal-organic frameworks. Chem. Mater. 2007, 19, 3681– 3685,  DOI: 10.1021/cm070542g
      45
      Raman Spectroscopic Investigation of CH4 and N2 Adsorption in Metal-Organic Frameworks
      Siberio-Perez, Diana Y.; Wong-Foy, Antek G.; Yaghi, Omar M.; Matzger, Adam J.
      Chemistry of Materials (2007), 19 (15), 3681-3685CODEN: CMATEX; ISSN:0897-4756. (American Chemical Society)
      The adsorption behavior of CH4 and N2 (298 K, 30 bar) in a series of isoreticular metal-org. frameworks (IRMOFs) was investigated by Raman spectroscopy. For CH4, the ν1 vibrational mode shifted to lower frequency by 7.6, 8.4, 11.0, 10.3, and 10.1 cm-1 from 2917 cm-1 when adsorbed to IRMOF-1, -6, -8, -11, and -18, resp. Along this same series, the adsorbed N2 stretch exhibited smaller shifts of 2.7, 3.1, 4.2, 4.1, and 3.7 cm-1. These shifts arise because of interactions within the framework pores, and not with the outer crystal surface. In all cases, Raman spectra at pressures up to 30 bar showed that satn. of the sorption sites does not occur. The obsd. shifts of the vibrational modes for each gas indicate different chem. environments within different IRMOFs, pointing to the important role the linkers play in the adsorption of gases.
    46. 46
      Duren, T.; Sarkisov, L.; Yaghi, O. M.; Snurr, R. Q. Design of new materials for methane storage. Langmuir 2004, 20, 2683– 2689,  DOI: 10.1021/la0355500
      46
      Design of new materials for methane storage
      Duren Tina; Sarkisov Lev; Yaghi Omar M; Snurr Randall Q
      Langmuir : the ACS journal of surfaces and colloids (2004), 20 (7), 2683-9 ISSN:0743-7463.
      One of the strategic goals of the modern automobile manufacturing industry is to replace gasoline and diesel with alternative fuels such as natural gas. In this report, we elucidate the desired characteristics of an optimal adsorbent for gas storage. The U.S. Department of Energy has outlined several requirements that adsorbents must fulfill for natural gas to become economically viable, with a key criterion being the amount adsorbed at 35 bar. We explore the adsorption characteristics of novel metal-organic materials (IRMOFs and molecular squares) and contrast them with the characteristics of two zeolites, MCM-41, and different carbon nanotubes. Using molecular simulations, we uncover the complex interplay of the factors influencing methane adsorption, especially the surface area, the capacity or free volume, the strength of the energetic interaction, and the pore size distribution. We also explain the extraordinary adsorption properties of IRMOF materials and propose new, not yet synthesized IRMOF structures with adsorption characteristics that are predicted to exceed the best experimental results to date by up to 36%.
    47. 47
      Walton, K. S.; Millward, A. R.; Dubbeldam, D.; Frost, H.; Low, J. J.; Yaghi, O. M.; Snurr, R. Q. Understanding inflection and steps in carbon dioxide adsorption isotherms in metal-organic frameworks. J. Am. Chem. Soc. 2007, 130, 406– 407,  DOI: 10.1021/ja076595g
      There is no corresponding record for this reference.
    48. 48
      Towns, J.; Cockerill, T.; Dahan, M.; Foster, I.; Gaither, K.; Grimshaw, A.; Hazlewood, V.; Lathrop, S.; Lifka, D.; Peterson, G. D.; Roskies, R.; Scott, J. R.; Wilkins-Diehr, N. XSEDE: accelerating scientific discovery. Comput. Sci. Eng. 2014, 16, 62– 74,  DOI: 10.1109/MCSE.2014.80
      There is no corresponding record for this reference.