pbdR available on h2p cluster

Dear R users of CRC,

Packages from pbdR (programming in big data with R) are now available to use over the h2p Omnipath cluster. The packages will allow R users to write MPI enabled parallel code in R. An example code and slurm job definition are available at “/ihome/crc/how_to_run/pbdr” on the cluster. To know more about pbdR see [1,2]. Do not hesitate to contact us should you face difficulties with using these packages.

[1] https://www.hpcwire.com/off-the-wire/ornl-researchers-bridge-gap-r-hpc-communities
[2] https://rbigdata.github.io

Computational investigation of CO2 electroreduction on tin oxide and predictions of Ti, V, Nb and Zr dopants for improved catalysis

The Journal of Matetrials Chemistry A has just published their issue highlighting their 2017 “rising stars” and one of those individuals invited to submit an article was our very own ChemE Assistant Professor John Keith.  The link to John’s article is here.

Journal of Materials Chemistry A is proud to present this themed issue highlighting 2017’s rising stars of materials chemistry research. This issue gathers the very best work from materials chemists in the early stages of their independent career. Each contributor was recommended by experts in their fields as carrying out work with the potential to influence future directions in materials chemistry. Congratulations to all of those who feature on their important work so far in the field of materials energy and sustainability.

H2P GPU Charging Scheme

Dear H2P GPU Users,

If you are not using the GPU cluster on H2P you can safely ignore this. We have enabled the full charging scheme on the GPU cluster. Try crc-sinfo.py to see the new partitions:

  1. gtx1080 (default): Charging 1 SU per card hour
  2. titanx: Charging 3 SUs per card hour
  3. k40: Charging 6 SUs per card hour

Check out H2P Service Units and Queue Information for more details.

Thank you!

The CRC Team

1-minute survey on workshop topics

Dear Users,

We are writing to you to request your participation in a brief survey on workshop topics for Fall 2017-Spring 2018. We would like to get feedback about your suggested topics. Your responses to this survey will help us evaluate the effectiveness of our proposed workshops so that we can design better topics. The survey is brief and it takes less 1-minute to complete. Please submit your entry before June 30.

CRC Team.

Vice Provost of Research announced a new PSC initiative

Dear Pitt Faculty,

I want to make you aware of a new initiative at the Pittsburgh Supercomputing Center (PSC) for our faculty.  This could be a great opportunity for those faculty whose research includes High Performance Computing.

Earlier this year PSC completed a technical upgrade of Bridges, its unique high-performance computing (HPC) system. Bridges enables applications that have not traditionally used HPC, integrates HPC with Big Data and artificial intelligence, and helps researchers facing challenges in Big Data to work more intuitively.

PSC is now making Bridges available at no charge to CMU and Pitt faculty through a simple, easy process so you can discover how Bridges can facilitate your research.

To learn more on how Bridges can be a good fit for your research, and how to get started with this exciting new initiative go to https://www.psc.edu/prcinitiative.

Pittsburgh Supercomputing Center is a joint effort of Carnegie Mellon University and the University of Pittsburgh. PSC provides university, government and industrial researchers with access to several of the most powerful systems for high-performance computing, communications and data storage available to scientists and engineers nationwide for unclassified research.

 

Best,

Mark S. Redfern, PhD

Vice Provost for Research

William Kepler-Whiteford Professor of Bioengineering

University of Pittsburgh

Wednesday April 12: On-The-Fly Heuristic Reordering Approach to Deterministic Optimization For Qualitative Chemical Property Prediction

Speaker

B. Christopher Rinderspacher
Army Research Laboratory

Title

On-The-Fly Heuristic Reordering Approach to Deterministic Optimization For Qualitative Chemical Property Prediction

Location

Wednesday, April 12th
1-2pm
307 Eberly Hall

Abstract

Chemical optimization and design affords current and future researchers the ability to harness the potential of chemical space towards the fast and efficient discovery of novel materials. In the present work, a multi-constraint deterministic optimization technique based on the on-the-fly heuristic reordering of chemical subspace has been developed and used towards materials discovery in several systems of interest. The competitive advantage of the deterministic optimization method results from the combination of fast computational techniques and innovative design algorithms which allow for intelligent screening of a large number of chemical compounds within a reasonable computational time.  A family of search algorithms has been used to approach the problem of navigating chemical subspace, including general base line search and general base gradient local search techniques. Because ideal ordering of the chemical subspace is not known, measures must be taken to ensure that the space is being sampled properly. There are several strategies utilized in this work to ensure this requirement is being satisfied. First, optimization algorithms based on heuristic reordering of the chemical subspace are developed to assist and direct the optimization procedure. The heuristic reordering algorithms play an essential role in optimization efficiency and each heuristic scheme has been uniquely developed as an effort to further enhance the subspace sampling. Each of these algorithms have been benchmarked and tested for their performance with respect to candidate structure discovery. In addition, to further combat the potential subspace sampling partiality, a binary entropic, enhanced sampling approach has been employed. This technique allows for generation and searching of structures which are chemically maximally different from the local best candidate structure. This is advantageous because a larger breadth of space is able to be sampled in an equally efficient manner. This method has been applied to several systems of interest including high-hyperpolarizability materials, energetic materials and optically switchable materials. Detailed analysis was performed for each of these systems and qualitative structure property relationships were determined.

NSF CAREER Awards to Four CRC-Affiliated Faculty

Four CRC-affiliated faculty were awarded the National Science Foundation (NSF) CAREER Award this funding cycle. The CAREER Award is the NSF’s most prestigious award for early career-development for teacher-scholars, and has a reviewing and selection process that is one of the most competitive within NSF. The Pitt CRC faculty who received the CAREER Award this cycle are John Keith, Peng Liu, Giannis Mpourmpakis and Christopher Wilmer. Remarkably, three of the faculty, Keith, Mpourmpakis and Wilmer, all come from a single department within the Swanson School of Engineering (Chemical & Petroleum Engineering). All four faculty rely heavily on CRC resources to perform their research and all contributed startup funds to access the high-levels of computer time and support required to make their research a success. The role that CRC plays in facilitating their research is illustrated by this quote from Dr. Mpourmpakis: “CRC has been very instrumental in accelerating our research both in terms of available computational resources and support from experienced personnel.”

Descriptions of each of the CAREER research projects are given below.

John A. Keith, Assistant Professor and Inaugural R.K. Mellon Faculty Fellow in Energy SusChEM: Unlocking local solvation environments for energetically efficient hydrogenations with quantum chemistry (#1653392)

John Keith’s proposal “Unlocking local solvation environments for energetically efficient hydrogenations with quantum chemistry” was recently selected for an NSF CAREER award. The project addresses the production of carbon-neutral liquid fuels via electrocatalytic reduction of the greenhouse gas carbon dioxide (CO2) to methanol. Specifically, the study seeks to improve the efficiency and selectivity of current solvent-based electrochemical processes by advancing understanding of how aqueous electrolytes participate in the overall reaction mechanisms at the atomic scale. The research will be coupled with educational thrusts that engage students in grades 8-12 in learning about renewable energy catalysis and computational chemistry.

In the figure shown above, A) overlaid Pourbaix diagrams for an N-doped graphene ribbon (gray/purple) and carbonic acid (solid lines). B) QC calculated ΔE values along the reaction pathway for the hydride transfer reaction: 2H2O + BH4– + CO2 →H3O+ +BH3OH– + HCO2–. PBE data corresponding to minimum energy pathways for a) explicit solvent + counter ion, b) continuum solvent only, c) 1st solvent shell + continuum solvent, d) counter ion and continuum solvent, e) 1st solvent shell + counter ion + continuum solvent.


Peng Liu, Department of Chemistry
Computational Studies of Transition Metal Catalyzed Reactions in Organic Synthesis (#1654122 )

In this CAREER project funded by the Chemical Structure, Dynamic & Mechanism-B Program (CSDM-B) of the Chemistry Division, Professor Peng Liu of the Department of Chemistry at the University of Pittsburgh is developing new strategies to use computational tools to investigate mechanisms and effects of ancillary ligands in transition-metal-catalyzed reactions of unactivated starting materials, such as C-C and C-H bonds, and unactivated olefins. The goal of this research is to reveal the fundamental reactivity rules of common organometallic intermediates in these transformations and to develop new models to interpret ligand effects on reactivity and selectivity. This proposal’s educational and outreach plan aims to maximize the power of computations to enhance learning of organic chemistry concepts and to facilitate synthetic organic chemistry research. Professor Liu’s team will develop virtual reality (VR) software and educational materials to visualize three-dimensional molecular structures and reaction mechanism videos in an interactive and immersive environment.

This project aims to address two basic challenges in performing computational studies on transition-metal-catalysis: 1) the lack of mechanistic understandings in many recently developed catalytic systems, and 2) the complexities in analyzing and rationalizing computational data, in particular, the origin of ligand effects. The proposed research will investigate novel reaction pathways involving the activated organometallic intermediates formed after the C-H and C-C bond cleavage steps, and elucidate the effects of ligands, directing groups, substituents, ring strain, and norbornene and Lewis acid co-catalysts. To systematically characterize the origin of ligand effects on reactivity and selectivity, a ligand-substrate interaction model will be developed. This model uses energy decomposition analysis (EDA) methods to dissect the through-space ligand-substrate interactions into chemically meaningful terms, including steric repulsion, polarization, charge transfer, and dispersion. The insights obtained from the proposed ligand-substrate interaction model will be used to develop of a catalyst screening methodology for transition-metal-catalysts.


Giannis (Yanni) Mpourmpakis, Assistant Professor
Designing synthesizable, ligand-protected bimetallic nanoparticles and modernizing engineering curriculum through computational nanoscience (#1652694)

“The goal of this project is to develop a novel open-access computational framework for predicting the growth mechanisms and morphologies of ligand-protected metal nanoparticles (NPs).
With NPs impacting numerous fields of science and technology, from energy to medicine to the environment, there is a critical need to determine the growth mechanisms of ligand-protected metal NPs and predict NP morphologies that can be synthesized in the laboratory. Although metal nanoparticles (NPs) of different sizes and shapes can be synthesized by colloidal chemistry methods, advances towards controlling NP morphology have been based largely on trial and error experimentation, which is often tedious and costly. The proposed computational framework will employ novel first-principles-based structure-property relationships accounting for structure sensitivity and metal composition. The integration of research and education efforts will focus on modernizing the traditional Chemical Thermodynamics course by introducing animation modules based on cutting-edge nanotechnology examples. Outreach activities are planned through a nanoscale-inspired interactive computer game to engage high school students, including underrepresented minorities, into pursuing STEM careers and increase awareness about the importance of the field of nanotechnology.

The proposed research project will combine Density Functional Theory methods with Monte Carlo and Molecular Dynamics simulations, Machine Learning, and scientific computing to develop a novel, open-access computational framework, applicable to the design of ligand-protected NPs. This framework will generate a library of crystal structures and electronic properties of thermodynamically stable, thiolate-protected, Au-based bimetallic NPs, across a range of heterometals and particle morphologies, all under realistic experimental conditions. The proposed work aims to advance current theories on NP stabilization, which are based on simplified, electron counting rules. The proposed computational framework will enable rational design of ligand-protected NPs. It will also elucidate NP growth steps that are experimentally intractable, thus accelerating nanomaterials discovery. The research findings will be made available online for experimental verification.”


Christopher Wilmer, Assistant Professor
Fundamental limits of physical adsorption in porous materials (#1653375) 

“The research objective of this proposal is to further our understanding of the range of physically accessible adsorption behavior in porous materials, and in so doing determine theoretical efficiency limits on important adsorption-related processes
, such as post-combustion carbon capture. The PI will use classical molecular modeling to simulate adsorption in randomly generated porous materials, called pseudomaterials, where the constraint that the materials be energetically stable is relaxed. Since the limits of adsorption in pseudomaterials will necessarily be higher than in real materials, determining the limits of pseudomaterials will also determine the limits for real materials. This approach will be used to establish a rigorous theoretical upper limit on the efficiency of a membrane-based post-combustion carbon capture process, which is considered one of the most promising technologies for mitigating climate change due to fossil fuel-based power plant emissions.”
“The educational objective of this proposal is to further public understanding of gas separations processes at the molecular level, especially as they pertain to carbon capture technologies relevant to climate change. The PI will (1) leverage his past success in creating award-winning scientific movies to develop a sequence of five educational movies on the fundamental physics, and applications, of gas adsorption (one for each year of the grant period), (2) teach lessons on adsorption to high school students interested in STEM careers as part of the University of Pittsburgh’s INVESTING NOW program, and (3) mentor undergraduates from underrepresented groups via the University of Pittsburgh’s EXCEL program. To increase their effectiveness and scale of impact, the educational movies will be created in consultation with science film-making professionals at Untamed Science, who will also help disseminate the movies online.”

$15,000/yr ACM SIGHPC/Intel Computational & Data Science Fellowships

ACM SIGHPC and Intel have partnered to create Computational and Data Science Fellowships, a 5-year program to increase the diversity of students pursuing graduate degrees in data science and computational science. Specifically targeted at women or students from racial/ethnic backgrounds that have not traditionally participated in the computing field, the program is open to students pursuing degrees at institutions anywhere in the world. Here are the important dates:

Submissions open: March 15
Submissions close: April 30
Winners announced: by July 31

For more information on qualification criteria and application submission, please visit SIGHPC website.

ARC 2017 Poster Session Winners

CRC would like to acknowledge and congratulate our poster session winners from our recent Advancing Research through Computing symposium!  Two winners were selected from more than 20 submissions.  Each winner received a $500 travel stipend.  More information about each winner below.

Michael Taylor

BioI received my B.S. in Chemical Engineering from University of Nebraska-Lincoln, followed by a year of process engineering work experience at Cargill in Nebraska before coming to the University of Pittsburgh to work on my PhD. I am pursing my PhD in the Department of Chemical Engineering at the University of Pittsburgh under the direction of Prof. Giannis Mpourmpakis. My primary focus is on the stability and synthesis of colloidal “magic-number” nanoclusters, while I am have on several side-projects on topics ranging from kidney stone growth modifiers to catalytic applications of nanomaterials.

Poster Title: “Stability and Prediction of Thiolated Metal Nanoclusters”

Poster Abstract: Thiolate-protected metal nanoclusters (NCs) of the form Mn(SR)m, exhibiting “magic numbers” in their stability, and several of their corresponding atomically-precise structures have been identified. These NCs show promising applications in fields such as drug delivery, bio-imaging, and catalysis. However, understanding what drives NC stability remains limited. Herein we introduce a first-principles derived structure-stability model of thiolate-protected NCs that rationalizes their experimental stability. Utilizing this model we report the successful reproduction of a known NC core structure, taking a step towards a general methodology for NCs structure prediction. Ultimately, our model aids in accelerating the discovery of atomically precise, highly-stable, colloidal NCs.


Jon Ruffley 

Bio: I received my B.S. in Chemical Engineering from The Ohio State University, where I did undergraduate research in the organic synthesis lab of Dr. Noel Paul. I am currently working toward my PhD at University of Pittsburgh under the guidance of Dr. J. Karl Johnson. My areas of study include plasmonic properties of non-noble metal nanoparticles, and the application of quantum chemistry methods to the design of metal-organic frameworks.

Poster Title: Hybrid Stratified MOF-Plasmonic Nanoparticle Materials for Detection and Destruction of Chemical Agents

Abstract: There is a pressing need for methods capable of rapidly detecting and destroying chemical warfare agents. We seek to develop a detailed understanding of the fundamental properties of multifunctional plasmonic hybrid nanomaterials for sorption, transport, photodetection, and photocatalytic degradation of target chemical species. We use continuum models to predict extinction spectra of CuSe nanoparticles, quantum mechanical methods to identify promising functional groups to bind target species in metal-organic frameworks (MOFs), and Monte Carlo methods to perform pure fluid adsorption isotherms in candidate MOFs. We have accurately predicted the extinction spectra of noble metal nanoparticles. Initial ab initio calculations have been used to identify functional groups capable of binding target chemical warfare agent simulants. We have calculated isotherms of small gas molecules in MOFs and compared our results with experiments, as a step toward computing isotherms of chemical warfare agent simulants.