Hello,
A change in allocation has been requested:
Requester: rrankin (Rees Rankin)
Project: DFT_TM_Catal_Screen
Title: DFT Screening Investigation on Catalytic and Electrocatalytic Phenomena of Transition Metals
Description: We intend to perform these calculations to study multiple important reactions in the context of fundamental catalytic as well as electrocatalytic applications. Specifically, we began in FY2009 by studying the reduction of hydrogen peroxide (H2O2) on these surfaces. The applications of studying this reaction were numerous. It is well known that H2O2 historically has not been synthesized efficiently via direct conversion of H2 and O2. Recent work in the literature suggests metallic alloy nanoparticles can create new chemistry in the selective synthesis of H2O2 from H2 and O2. After completing our initial work (flat, and stepped surfaces of 'pure' single component TM species) on this reaction in FY2009, we are ready to move forward quickly with screening of alloys/skins for this reaction, and to extend our now-developed methods to look at related oxygen reduction (ORR) reactions and the interesting reactions involved with the electrochemical reduction of CO2
(CDR) towards more economically and environmentally beneficial target species. Additionally, by building on the knowledge generated from such fundamental reactions, it will be much more feasible to predict the performance of novel catalysts in other reactions which involve the same or similar reactants and intermediates but different products, or conversely, differing reactants with similar intermediates and products. Our work will continue to supplement the computational materials-by-design database being developed.
All calculations will be performed using the Vienna ab initio Simulation Program (VASP). This program uses a periodic plane-wave representation of the electronic wavefunctions of materials with a self-consistent minimization/solution under the Density Functional Theory (DFT) formalism. Recent improvements in the parallelization of VASP combined with the improved hardware capabilities in the LCRC Fusion system (compared to Jazz) allow for reasonably efficient near-linear scaling for jobs running on up to 80 cores (10 nodes). Superlinear scaling can now even be witnessed for small jobs in a single compute node with appropriate cache sizes. Additional calculations in the future may utilize the Dacapo DFT code, which performs similarly to VASP in the above regards. GPAW, the successor to Dacapo, may become a viable option within FY 2010 as well; GPAW is inherently designed to scale well beyond Dacapo or Vasp in terms of parallelization efficiency on large HPC systems such a
s the BlueGene.
Preliminary calculations for this work have already been performed on the HPC Carbon cluster available at the CNM (ANL), and the JAZZ system at LCRC (ANL) in FY2009, and are continuing into FY2010. However, the resources available at CNM-ANL Carbon cluster are currently significantly oversubscribed and will likely remain so even with the upcoming hardware expansion; small scale jobs can actually sit in the submission queue for time intervals longer than their execution related wall time. An allocation of 650,000 core hours on LCRC-Fusion would ostensibly allow for approximately 1600 calculation-sets (1 calculation-set here can be envisioned as scanning several related materials (such as differently structured surface alloys of the same stoichiometry) +adsorbate over many adsorption sites, or scanning a single material surface +adsorption site over many adsorbates, or as a series of coupled calculations to determine the transition state/activation energy between 2 local minim
a). As our initial work has determined the best single component Transition Metals relevant for H2O2 selective synthesis, we are now ready to significantly augment the work by screening through vast combinations of alloy/skins/etc derived from these materials. Additionally, we will extend our work to look at related reactions on these surfaces such as the electro-reduction of CO2, which is far more complicated and will hence require more calculations to initially understand. The ability to augment our knowledge base with the incorporation of this many additional calculation-sets would be paramount in developing the detailed understanding necessary to move forward in a purely computational catalysis by design (or broader materials by design) project. As a note, we used our FY2009 Allocation of 96000 hours effectively in ~6 months, and we propose to do effectively ~20x the number of total calculations in FY2010 now that we have a detailed initial understanding of our method
s and some basic insights into fundamental reaction mechanisms/TM species we wish to study in vastly greater detail. The improved efficiency of these calculations with respect to Fusion's higher core clock speeds and improved intra-node and inter-node communication should make this feasible under the requested allocation.
Current: undetermined amount
Justification: As a brief summary of typical results, Vasp 4.6 or Vasp 5.x+ with appropriate compilation, Infiniband, and modern processor cores scales well (near linear) out to approximately 80 cores or 10 (Fusion) nodes for most typical calculations on these similar HPC systems. VASP scaling does fall significantly in the regime of calculations with severely large file I/O overhead--however, this is not expected to apply to the jobs performed in the proposed work. I am not certain how to attach a file to this web template (I am using Firefox and see no options for a file upload?) I can certainly provide, via email if necessary, some parallelization benchmarks datafiles or references for VASP calculations on similar modern HPC installations (such as EMSL @ PNL, or the Carbon cluster at ANL's CNM). I expect most jobs submitted would use between 32-64 cores (4-8 Fusion nodes)--a regime where again, VASP is expected to work as well, or better, on Fusion than it does on
similar HPC systems where I use the same level of resources and parallelization is acceptably good.
Requested: 240000
A specific reason has been given:
The current project is proceeding well and manuscripts are being prepared. However we are also currently investigating more detailed kinetics of the system than initially expected. Thus we would like to perform more NEB calculations than initially budgeted in our proposal. The current request for additional allocation time should cover the requisite new NEB calculations.
This needs to be approved and the final allocation amount decided upon.
Thank You,
The LCRC Accounts System
