Hello,
A yearly allocation for the LCRC cluster has been requested with the
following updated information:
Submitter/PI: John J. Low
Project Name: Raney_Nickel
Division: CLS
Project title: A Molecular Model of Raney Ni
Associated funding: None
Other Systems: None
Science: Raney nickel is a nanostructured catalyst which is used in a variety of industrial processes. Due to its catalytic activity at room temperature, it is used for reduction of benzene, conversion of nitro compounds to amines, and for desulfurization of thioacetals to hydrocarbons. It has a characteristic porous, amorphous structure. The initial form of Raney Ni is a NiAl alloy which is transformed into the active form by leaching out aluminum which leads to the characteristic porous structure of the final catalyst. The porosity depends on the initial composition of the NiAl alloy which is typically prepared by melting nickel together with aluminum and subsequent quenching. The quenching process induces the formation of different NiAl phases which react differently with the caustic and thus affect the pore formation in the final catalyst. Experimental studies indicate that the sponge-type structure of Raney Nickel grains is formed by a collection of crystallites adher
ing together . Since the structure of both the precursor alloy and the active form of Raney Nickel is determining the catalytic activity and performance, it is crucial to gain a detailed understanding of the structural properties at the atomistic level. The goal of this project is to leverage molecular modeling to provide critical insights into how the composition of the precursor alloy affects the porosity and thus activity of Raney Nickel.
The initial NiAl compositions can range from 40 to 70 wt% Ni 2, , . Freel at al.2 determined the mean pore diameter to be in a range between 2.5 to 10 nm. The basis for approaching the structure of the active form are the characterization studies of Raney Nickel2, , , by Anderson’s group and others which yield a qualitative picture of Raney Nickel. The void volume of these particles ranges from 0.05 to 0.15 cm3/g (1). X-ray diffraction3 and X-ray Absorption Spectra4 show that local structure of Raney Ni is similar to FCC Ni. Fouilloux3 gave a qualitative description of Raney Nickel as a spongy material composed of 100 nm nano-agglomerates of smaller than 2.5 to 15 nm nanocrystals. Several modeling studies of grain boundaries use molecular dynamics to generate structures. Van Swygenhoven and Caro used a stochastic procedure that starts with randomly placed seed particles with the rest of the space filled with a Voroni construction. Keblinski randomly placed seed particles
in a liquid metal and then quenched the melt to generate a model crystal with grain boundaries. These procedures yield a model of a solid composed of nanocrystalline grains.
These studies focus on modeling the structure of the final catalyst. However, a critical question is, how the initial alloy composition influences the final catalyst performance. The aim of the our work is (a) to demonstrate how a porous structure of Raney Ni can be modeled with random pore distribution while taking the initial precursor composition into account and (b) to investigate to what extent the activity of the final catalyst depends on the composition of the precursor alloy.
Project description: In the previous phases of the project, molecular dynamics (MD) simulations have been performed on fusion and blues using the LAMMPS software package in order to analyze the influence of the initial composition of the alloy on the structural properties of the precursor. Model structures for four different compositions (40/50/60/70 wt% Ni) have been created using Scienomics MAPS platform by randomly replacing Ni and Al in the corresponding fractions in a NiAl3 supercell. In order to ensure a statistically relevant sampling, five initial starting configurations have been generated for each of these model structures. The structures were equilibrated at 2000 K and then cooled down to room temperature over 1 ns in a similar way as described by Noya et al. For one configuration of each composition, an additional NPT simulation of 1 ns at room temperature was performed to check the convergence of the density and the cell size. The analysis of theses simulatio
ns showed that the composition of the precursor has a significant influence on the density and cell parameters, while the configuration has only a very minor impact. In general, the density, cell parameters, total energy and volume over the simulation time show a consistent behavior.
To establish a simulation protocol for modeling the porous structure of Raney Ni, we have focused on one structure with the initial alloy composition of 50 wt % Al and 50 wt % Ni which corresponds to the commonly used composition of the precursor. After quenching to room temperature, Aluminum was removed. For studying the influence of the remaining amount of Al on the pore formation of the final catalyst, two percentages of remaining Al have been considered: (a) 0% and (b) 5 % Al were left for modeling the active form of the catalyst. These structures were then optimized using 5000 steps Steepest Descent. A short 200 ps NVT simulation was performed from which the structures of 10 snapshots were extracted. These structures were optimized and used as starting point for a 10 ns NVT MD simulations at 300 K. In order to ensure a statistically relevant sampling, initial velocities were randomly chosen for each system. Finally, a 20 ns NPT simulation was performed to equilibrate the
structures under physically meaningful conditions. In order to study the influence of the system size on the size of the created pores, a 2x2x2 supercell was created after quenching for both the 0% and 5% Al structure containing 163,840 atoms and 181,856 atoms, respectively. For both systems, first a 10 ns NVT simulation was performed followed by a 10 ns NPT simulation.
The MD simulations and final structures were characterized using the MAPS platform11 for studying the evolution of the density, free volume, and free surface and the Zeo++ software , for analyzing the pore size and pore size distribution. The structural properties of our modeled structures are in very good agreement with experimental data. In particular, an average value of 4.1 ± 0.4 g/cm3 was obtained for the density and of 43 ± 5 Å for the maximum pore size, which is both well within the experimental range of 3.5 - 7 g/cm3 and 20 - 100 Å, respectively. The simulations also indicate that the presence of a small percentage of Aluminum stabilizes the pores as the comparison of the evolution of the pore size and the pore size distribution of the final 0% and 5% Al structures suggests. The analysis of the 2x2x2 supercell structures showed the same trend with respect to the structural characteristics corroborating the validity of our simulation approach .
After having established a simulation protocol, we applied the workflow on three other compositions (40/60/70 wt% Ni) of the initial alloy precursor for investigating the influence of the composition on the pore formation. For each initial alloy composition, we performed three sets of calculations with different percentages (0%, 5%, 10%) of aluminum content remaining in the final catalyst. A 10 ns NVT simulation was performed for inducing the pore formation followed by a 15 ns NPT simulation for relaxing the porous structure. The results for the three new initial alloy compositions were consistent with results previously obtained for the initial 50 wt% Al alloy model. The simulations revealed that the density of the final structure and the largest pore size diameter is indeed influenced through both the initial and final composition. The simulations further confirmed the assumption that aluminum has a stabilizing influence on the porosity. Moreover, simulations on a 2x2x2 su
percell with 10% Al content in the final structure showed consistent results with the original cell and were in line with simulations on supercells with 0 and 5% Al, respectively.
In the next phase, we plan to study three aspects:
Firstly, we plan to study the pore formation using heterogeneous starting materials. Experimentally, the quenching process induces the formation of different NiAl phases. Apart from NiAl3 which we were using as initial alloy material, Ni2Al3 and an eutectic Al–Ni Al3 phase exist in the precursor. Since the leaching rates for the various NiAl phases are different, it can be assumed that the porosity will not be uniformly distributed in the final catalyst. Therefore, we plan to extend our studies to more complex models and take different NiAl phases into account. For this purpose, a suitably large simulation box of Ni2Al3 will be generated comparable in size to the NiAl3 model. For the initial alloy precursor, the Ni:Al ratio will be adjusted to 50 wt% Al corresponding to the commercially most widely used precursor. The Ni2Al3-based model will be subjected to a set of molecular dynamics simulations following the simulation protocol as described above. Three Al fractions (0%,
5%, 10%) for the final form of the catalyst will be studied. The evolution of the pore formation and structural characteristics such as the pore size and density will be analyzed and compared with our previous results for the NiAl3-based system. In order to investigate how different NiAl phases influence each other during pore formation, we will model a NiAl3/Ni2Al3 interface involving highly CPU-demanding MD simulations. Structures will be build using MAPS12 and the simulations will be performed using LAMMPS8. The trajectory analysis will be done using MAPS and Zeo++14,15.
The computational time needed for performing these simulations and analyzing the trajectories will be considerable: The model structures contain more than 65,000 atoms and in total four 25 ns MD simulations (3 percentages, 10 ns NVT + 15 ns NPT and 1 interface,10 ns NVT + 15 ns NPT) need to be performed. According to the LAMMPS Cu benchmarks , 100 steps for a system size of 32,000 atoms would take 0.3 s (fixed CPU time) on 512 cores. A 25 ns trajectory (25,000,000 steps) for 32,000 atoms would take 10,700 core-hours, a 65,000 atom system about 21,000 core-hours and a 130,000 atom system about 45,000 core-hours, respectively. 109,000 core-hours will be needed for the MD simulations of the various model structures.
The simulation protocol is generally applicable for modeling nanoporous materials. In the last years, these types of materials have been attracting much attention in the area of energy storage. To overcome the dependence on fossil fuels and because of the necessity to move towards renewable and sustainable energy resources, efficient energy storage systems become more and more important. Due to demanding requirements of industry and consumers on competitively viable energy storage technologies, it is crucial to understand and know the underlying storage mechanisms. Molecular modeling techniques can provide pivotal insights into energy storage-related processes such as the oxygen reduction reaction in nanoporous fuel cell catalysts or the chemical processes in carbon-based electrochemical double layer capacitors . Ab initio or density functional theory (DFT) approaches, respectively, are most suited to simulate chemical reactions, as these methods are able to describe the ele
ctronic structure. Because of the inhomogeneity of these amorphous, nanoporous structures, large fragments are needed to take structural influences on the reactions into account. For studying such systems sizes the computational effort will be significant and is beyond the scope of the current work. In the spirit of a multiscale approach, future efforts will combine molecular dynamics simulations for modeling nanoporous energy storage materials and DFT to study the chemical reactions involved. Preliminary studies on suitable model systems, e.g. PtNi alloys or carbon nanotubes using QuantumEspresso or CP2K, will focus on elaborating suitable parameters by systematic convergence studies.
Secondly, we intend to study the leaching process using adaptive kinetic Monte Carlo (AKMC) simulations. AKMC has some important advantages compared to molecular dynamics: AKMC can be used to model realistically sized agglomerates (100 nm) and time scales (hours). No a priori knowledge of the reaction mechanism is required, rare and unexpected configurations can be sampled. AKMC simulations thus allow not only to model the structure of the final catalyst, but provides also critical insights into the kinetics of the leaching reaction. The program package EON will be used for this purpose. Preliminary studies have shown that the same EAM potential as used in the MD simulations can be employed with the AKMC method implemented in EON. Once suitable parameters have been established, production runs will be performed for two MD-equilibrated alloy models: The NiAl3- and Ni2Al3-based alloy model containing 50 wt% Al obtained from the preceding MD simulations. In this way, we gain i
nformation how the NiAl phase affects the kinetics of leaching process. We estimate that 100,000 core hours will be required for the AKMC effort.
Thirdly, we aim at modeling the catalytic reaction of benzene hydrogenation on Raney Ni. Initial studies have been performed using a modeled structure of Raney Ni with 0% Al based on an alloy initially containing 50 wt% Al. As a first step towards modeling the whole reaction, the adsorption of benzene on Raney Ni was studied using the reactive molecular dynamics as implemented in LAMMPS. For comparison, similar simulations have been performed for two clean Ni surfaces. The preliminary results indeed suggest a faster adsorption on the catalytic Raney Ni system. However, more advanced long-term simulations need to be performed. Additionally the application of ReaxFF on the catalytic effect of Raney Nickel will be studied. For this purpose, an improved development version of the hydrocarbon ReaxFF23 obtained from Ari van Duins group will be used. Initial test will include simulations of benzene/H2 mixture on different Ni surfaces. If they are successful, they will be followed b
y simulation of Benzene/H2 mixture in the Raney Nickel catalyst.
The computational time needed for ReaxFF simulation is much more important than for usual MD simulations since that, for an accurate representation of the system the simulation timestep needs to be 10 time smaller than the usual one. The computational time for performing a 5 ns ReaxFF simulation is therefore expected to be about 25,000 core-hours. We estimate that a total of 100,000 core-hours will therefore be required for the ReaxFF part.
In this way, we can gain a comprehensive understanding of the catalytic system: Information about the influence of the NiAl phase on the porosity of the final catalyst and the kinetic control of the leaching process will be obtained through MD and AKMC simulations, while information about the catalytic process will be gained through reactive force field simulations.
Overall, 310,000 core-hours will be needed over a period of 12 months for performing the in silico experiments for the project.
Industry partnership: About Scienomics
Scienomics was established in 2004 and with selected partnerships with the best-of-the-breed (e.g.: Sandia National Labs, Max Planck Institute, Fraunhofer Institute, University of Illinois, Demokritos, University of Shanghai), Scienomics’s MAPS platform offers a unique and powerful blend of multiscale and multiparadigm modeling and simulation modules.
With high quality science and industrial solution oriented applied research, Scienomics has garnered collaborative projects with companies in the areas of energy and green chemistry, alternate fuels, catalysis, polymers, biodefense, auto exhaust, etc.
Biography of Relevant Members
John J. Low, Principal Computational Science Specialist at Argonne National Laboratory has a Ph.D. in Chemistry from the California Institute of Technology. After graduation John worked at UOP LLC, a Honeywell Company. His work at UOP focused on modeling and characterization of catalysts and adsorbents used in the petrochemical industry, hydrogen storage and carbon sequestration. John supports the computational chemistry applications at high performance computers and performing research on Li-Ion batteries, converting biomass to fuels, and nuclear materials at Argonne. He is expert in quantum chemistry, molecular dynamics and Monte Carlo methods on high performance computers.
Dr. Robin Chaudret, Research Scientist at Scienomics, has a PhD in Computational Chemistry from the Université Pierre et Marie Curie of Paris. He has been doing research in the fields of organic, organometallic and bio chemistry. He applied and developed various tools to analyze and simulate the structure and reactivity of different systems. His ability to perform multiscale simulations (Quantum, Classical and QM/MM) allows him to cover a broad range of research areas such as homogeneous and heterogeneous catalysis, biochemistry or bioinspired chemistry, surfaces, etc. During his PhD and postdoc experiences he developed several collaborations with Paris or Montpelier Universities in France and Duke University in North Carolina and organized and participated to different meetings and published several papers.
Dr. Sabine Schweizer, Senior Scientist at Scienomics has a Ph. D. in Theoretical Chemistry from University of Tuebingen, Germany. Her work experience is in the area of applying quantum chemical methods to large and small molecules. Her experience is particularly useful in the areas of homogenous and heterogenous catalysis, where energetics and dynamics play an important role in the reactions. She has experience working with multiscale hybrid methods combining quantum chemical and classical methods which makes her particularly versatile in applying Scienomics’ multiscale, multiparadigm MAPS platform to study hard materials such as surfaces, coatings, semiconductors, solar cells, etc.
Dr. Lalitha Subramanian, Chief Scientific Officer and VP of Services has 18 years of experience providing insight into chemical systems and processes that are of critical interest to industry. Following her Ph. D. in Chemistry, her post-doctoral work was with Prof. Roald Hoffmann (Nobel Laureate) at Cornell University. She has been a leading architect of solutions in the areas of alternate energy, catalysis, materials design, and process optimization. She has worked on numerous proprietary projects for diverse companies and in this pursuit; she has delivered product enhancements, process optimization and cost savings to her customers. Lalitha maintains a broad range of partnerships in oil & gas, chemical, personal care, materials, semiconductors, automotive, aerospace, and pharmaceutical industries. She has co-authored a book on software techniques used in Materials Science published by CRC Press, 2005. She continues to present invited lectures and has numerous scientifi
c publications.
Project URL:
Current FY Hours Used: undetermined amount
New FY Requested allocation: 301000
Q1: 75250
Q2: 75250
Q3: 75250
Q4: 75250
Justification:
Storage requirements:
Thank You,
The LCRC Accounts System
