High-Throughput Materials Explorations Strategy

Introduction

Fuel cells typically use platinum electrocatalysts for the oxidation of hydrogen at the anode and the reduction of oxygen at the cathode. However, platinum poses several challenges to the widespread use of fuel cells. Impurities in hydrogen fuel (carbon monoxide and sulfur compounds) can poison the anode. The reaction rate of oxygen reduction is slow, causing an energy loss (called overpotential) of at least 30% of the theoretical value. Platinum is also costly.
The anode electrocatalysts recently developed at CFCI are based on binary ordered intermetallic compounds. We are continuously searching for new electrocatalyst materials with improved properties over platinum. Due to the vast number of possible binary and ternary intermetallic compounds, the complete exploration of all materials would require many years. To do so in a reasonable amount of time, we are using a combinatorial approach to synthesize many different compositions on a single sample and measure their electrocatalytic properties.

Synthesizing Composition Spreads

Figure 1. One of two vacuum chambers used to sputter thin film composition spreads.

Figure 2. Three sputter guns arranged in an equilateral triangle. Each gun is loaded with a different element for deposition on a 3-inch substrate (not shown in picture).

We use on-axis co-sputtering of two or three elements inside a vacuum chamber (Fig. 1) to create thin films of composition spreads. We position three sputter guns at the vertices of an equilateral triangle (Fig. 2), each one loaded with a different disk (called the target) of a particular element. The vacuum chamber lid holds a 3-inch diameter substrate 1.5 inches from the guns, centered on the gun triangle. The concentration of atoms from each sputter gun is greatest at points on the substrate closest to the gun axis and lower at points farther away. As both PtBi and PtPb intermetallic compounds exhibit good electrocatalytic activity, we are currently using Pt, Bi and Pb targets to create composition spreads. The graph below (Fig. 3) shows the variation of atom concentration as a function of distance from the gun axis for Pt, Bi, and Pb targets. When we sputter these elements simultaneously, the ratio of each one varies with position on the substrate, creating a vast number of compositions.

Figure 3. The graph shows how the concentration of Pt, Bi, and Pb atoms vary on the substrate as a function of distance from the gun axis.

If we take these compositions and plot them on a Pt-Bi-Pb ternary phase diagram (Fig. 4), we see that we will obtain approximately 60% of the possible compositions in a single sample.

Figure 4. All the compositions inside the points shown on the ternary phase diagram will be synthesized on a single Pt-Bi-Pb thin film composition spread.

We have recently introduced a new deposition system (Fig. 5) capable of
depositing pseudo-quaternary composition spreads (films containing 4
elements). The new system is also equipped to measure deposition profiles
to automatically map data onto composition plots such as that seen in
Figure 4. Our numerical modeling of resputtering in codeposition is
important in this mapping.

For more information on the sputtering system, please see:

"Getter sputtering system for high-throughput fabrication of composition
spreads" John M. Gregoire, R. B. van Dover, Jing Jin, Francis J. DiSalvo, and
Héctor D. Abruña. Rev. Sci. Instrum. 78, 072212 (2007).

Figure 5:

Schematic of the new sputtering system, capable of synthesizing quaternary (4-element) composition spreads.

Finding Regions of Electrocatalytic Activity

To complement the rapid composition spread synthesis, we need an in-parallel method for testing the thin film that will be able to identify regions of high electrocatalytic activity. In an attempt to do so, we are performing thermal imaging of the sputtered electrode sample during electrochemical measurements. The sample is placed in an electrochemical cell that allows a thin film of electrolyte solution to cover the electrode. We then vary the substrate potential and measure the resulting current. The infrared camera is mounted above the cell, shown in the picture below (Fig. 6). Regional increases in temperature on the sample are expected to indicate the location of compositions with higher electrochemical activity.

Figure 6. The thermal imaging system, used for parallel screening of compositions for electrochemical activity.

Once high activity regions are roughly identified, we refine our analysis using scanning electrochemical microscopy (SECM). We place the sample in an electrochemical cell and immerse it in an acidic electrolyte. A tip is held close to the composition spread and can raster over a selected region, as shown in the diagram below (Fig. 7).

Figure 7. A diagram of hydrogen oxidation activity measurement on the SECM.

The tip is held at a sufficiently negative potential so as to reduce protons to hydrogen. If the composition below the tip is a good electrocatalyst, then it will oxidize the hydrogen back to protons at a high rate, producing relatively high current through the tip. From the scan we can produce a graph of current as a function of position and identify the location of good electrocatalysts on the composition spread. We can also test the samples for oxygen reduction (Fig. 8) and formic acid oxidation (Fig. 9).

Figure 8. Oxygen reduction activity measurement on the SECM.

Figure 9. Formic acid oxidation activity measurement on the SECM.

Characterization of Active Regions

Once we have identified the location of a good electrocatalyst on our composition spread, we need to determine its characteristics in order to synthesize and test the material in bulk. We can obtain its composition with microprobe and Rutherford back-scattering (RBS). We can observe the surface texture and crystal grain size using scanning electron microscopy (SEM). X-ray diffraction in a general area diffraction detection system (GADDS) reveals the crystal structure.

Webpage Detail coming soon ...