Navigation Instructions

Use the dropdown menu to navigate to the desired combination of the following fields: Structure Type, Binding Site Composition, Adsorbate, and Result Type.

Structure Type

The two structure types explored for the catalyst are reflected here: a 79- atom nanoparticle (79 NP) and the bulk material (Slab).

Adsorbate

The two adsorbates explored for the fuel cell are reflected here: oxygen (O) for ORR and hydrogen (H) for HER.

Binding Site Composition

AAA/BBB and AAB/ABB

The tables are set up so that the primary metal (A) is on the x-axis and the secondary metal (B) is on the y-axis. The 3-atom hollow binding site composition can then be readas a concatenation of the respective metals (e.g. AAB would be 2 of metal A, 1 of metal B).

Catalytic Activity

The table reflects the catalytic activity for each site based on the volcano plots listed in the background info page, with a gradient key listed on the right. The green cells indicate overbinding, while the yellow indicates underbinding. A darker color indicates a more ideal activity level.
A tooltip appears when a cell is hovered over, displaying various information such as the observed and predicted binding energies in eV, the original binding site and alloy, and standard deviation of the observed binding energy.

Distance Moved

This data is used to display the average distance an adsorbate moved from its initial position in the 3-atom hollow site and if it moved to a bridge, top, or another hollow site. The color represents the mode of the site movements. For example, on the 79 NP AAB/ABB H Distance Moved table, the green AgPt indicates that a majority of the 5 runs optimized to the top site. Hovering over the square provides more information: the Pt top site was preferred 4 times, while the AgPt bridge site was preferred the other time. Thus, it can be reasonable to assume that the AgAgPt hollow site is unstable/less stable than the surrounding sites.

Graph

There is a dynamic graph located at the bottom of the webpage. By default, an ideal oxygen binding energy (OBE) line and ideal hydrogen binding energy (HBE) line are displayed. These are able to be added/removed by clicking the respective cell on the top right. The clear button will clear everything but the ideal lines on the graph.

When the graph result type is selected, a table of gray cells appears. These cells are able to be selected and unselected for both the binding energy and distance moved result types, which will also add data to the graph.

Binding Energy

Selecting a cell displays a line graph representing the respective average binding energies for each possible hollow site combination. The linearity of the graph reflects the tunability of the material. Upon hovering over, each data point displays the binding energy, accurate to 4 decimal points.

Distance Moved

Selecting a cell displays a bar graph representing the average adsorbate distance moved at each possible hollow site combination. This option is complementary to the binding energy graph. If, when analyzing tunability, an anomaly is observed, the distance moved graph can indicate whether the adsorbate moved sites or not.

NOTE: This webpage is currently incompatible with Microsoft Edge. Please switch to Google Chrome, Mozilla Firefox, or any other browser that supports JavaScript in order to view the rest of the database.

The code for the webpage can be found here.

Background Information

It is essential to promote alternative energy sources as our energy demands continue to increase. One such form of alternative energy is the hydrogen fuel cell, which converts hydrogen and oxygen gas into water and energy, but its viability is limited due to the expense of the necessary platinum catalyst ($847.77/oz, updated 7/18/19). Our research aims to replace platinum with a more economical catalyst via high-throughput computation and analysis of material candidates.

In the search for a suitable catalyst, binding energies, the energy released when the adsorbate attaches to the material, are explored. Oxygen is explored for the Oxygen Reduction Reaction (ORR) that occurs within the fuel cell to release energy and the Hydrogen Evolution Reaction (HER) that occurs outside the fuel cell to produce hydrogen gas from water. The equation used to calculate binding energy is displayed below.

Binding Energy = Energy_{Adsorbent + Adsorbate} - Energy_Adsorbent - ½ Energy_Adsorbate

There is an established trend between binding energy and catalytic activity for both ORR and HER, as displayed in the figures below. Evidently, platinum as a catalyst is near-ideal for both reactions. The idea in the bimetallic project is to combine two FCC metals in hopes of creating a more ideal or cheaper alternative. Eight FCC metals were selected for this project are Ni, Pd, Rh, Pt, Ir, Cu, Au, and Ag, allowing for 28 possible bimetallic combinations.

A computational approach was utilized due to the number of calculations that are required. Two structures were analyzed in 50:50 compositions: 79-atom nanoparticles arranged in a face-centered cubic (FCC) truncated octahedron and four-layer FCC(111) slabs with the bottom two layers frozen. Five randomly generated structures of each combination were generated to reflect the randomness of how these catalysts would be created for practical use. On each surface, four unique 3-atom hollow binding sites were explored with both oxygen and hydrogen adsorbates, as shown below. This results in 2240 calculations, not accounting for errors. Surrounding each hollow site are a few other possible sites: the 2-atom bridge site and 1-atom top site. On occasion, the adsorbate favored these alternate sites or even a neighboring hollow site as opposed to the designated hollow site. Thus, the distance moved statistic is measured to determine which structures observed this adsorbate movement, indicating instability at the hollow site. The hollow site is explored specifically because the ensemble effect dictates that the surrounding composition of the site generally reflects the overall composition of a structure. So when the adsorbate optimizes to a different binding site, this effect cannot be analyzed properly and affects a property called tunability. If a material exhibits tunability, its binding energy scales linearly with its overall composition, so it is important to know whether or not altering the structure will generate the expected results.

Along with tunability as a factor for predicting binding energies, a binding energy prediction algorithm is under development. The idea is to be able to predict binding energies of other metal combinations not explored in the bimetallic project without having to undergo the lengthy calculation process.