Lazo Trkulja

Purdue University

Project Title: Yttria-Stabilized Zirconia Pellet Optimization and Characterization for Solid Oxide Fuel Cells
Advisor: Prof. Elliott Slamovich

Introduction

Solid oxide fuel cells (SOFC’s) are an efficient method of generating electricity from fuels such as hydrogen gas. SOFC’s consist of three parts which are the cathode, electrolyte, and anode. The cathode and anode are the positive and negative electrodes, respectively, and the electrolyte is a dense ceramic (usually yttria-stabilized zirconia) where oxygen ions from the cathode pass through it in order to get to the anode. The connection between the anode and the electrolyte is the focus of the research because a smooth anode surface can allow for a uniform application of the electrolyte.

Project Objectives

  • Optimize current processing methods to obtain the best quality YSZ pellet
  • Observe how sample quality varies with binder composition
  • Apply optimized procedure to make Ni/YSZ pellets that will be used as SOFC anodes

Approach

The approach began by initially making three sets of pellets with binder contents of 2%, 4%, and 10% by volume of solid. YSZ powder, binder (Duramax HA-8), dispersant (Duramax D-3005), and de-ionized water were mixed using a ball mill. After drying, the powders were uniaxially pressed at 68.7 MPa and held for one minute. Before pressing, stearic acid dissolved in isopropyl alcohol is applied to the die walls to reduce friction. After pressing the samples are bisqued at 950 °C for 2 hours to increase their strength. Bisquing is considered the early stages of sintering where the powder particles only begin to join together to form a larger particle.

Flaws on pellet surfaces were examined using an optical microscope. Quantitative surface roughness data was gathered using a profilometer, a device that operates by running a stylus across the pellet surface. The device output is used to determine the average roughness. Nine scans were taken on each side of a pellet, thereby providing a good statistics regarding pellet roughness.

Another quantitative method was to examine the edge removals of the pellets and measure those that were significant. These were edge defects that had penetration points into the pellet larger than others observed near the edge. From this data the average significant edge removal length was calculated as well as the number of the significant edge defects observed per pellet.

Findings

The average roughness data showed that the 2% and 4% binder sets were close relative to surface quality. Due to higher roughness values, the 10% binder set pellets were not considered as possible final candidates. After measuring the significant edge removals, the 4% binder set had a lower average removal length than the 2% set. However, the 2% binder set had pellets that had a lower number of edge removals. This created a trade-off for choosing between which set of pellets has the best results. The 4% set could become the better set if the pressing process is carefully monitored. Defects can occur more on the pellet if there is friction present in the die walls, cap, and punch. Uniform application of the stearic acid can reduce this. Another important aspect is the handling of the pellet after pressing since it can create edge removals when taking it from the die. Adjusting the pellet processing methods can determine whether or not the pellet is of good quality or not suited for use in SOFC anodes.

Fig. 1. Example of poor surface quality
(10% Binder Set).
Fig. 2. Example of good surface quality
(4% Binder Set).

Contact me: lazo@purdue.edu