Microstructural Engineering of Hydroxyapatite Membranes for Fuel Cell Applications

Keith Savino

University of Rochester
Department of Chemical Engineering

OPT 407: Practical Electron Microscopy
Spring 2009
Final Project

Background and Introduction
Fuel Cell Basics

Experimental Procedutre
Growth of Hydroxyapatite

Results and Discussion
Micrographs of Growth
Micrographs with Different Techniques
Composition Analysis
Electron Flight Simulator


Background and Introduction top

Fuel Cell Basics top

Rising energy costs, limited energy supplies, and increasing pollution are the fundamental issues that must be considered in the cost analysis of any energy source. These problems for a range of fuel sources can be reduced since a fuel cell can operate with a variety of fuel sources by altering its membrane. Fuel cells offer an improvement over existing technology such as the internal combustion engine by converting the fuel directly to energy. Rather than combusting the fuel and using the heat to produce energy, fuel cells use an electrochemical process that extracts energy directly from the fuel. These electrochemical reactions are much more efficient and are capable of converting more of the fuel to usable energy. The reactions are also cleaner and have very few side reactions.

A standard fuel cell diagram is shown in Figure 1 below. The diagram shows hydrogen and oxygen from the air as the fuel. Instead of hydrogen, other fuel sources could be methanol, ethanol, or biomass. As shown, the hydrogen enters on the anode side of the fuel cell and is converted into protons and electrons. This reaction occurs because of the catalyst at the surface of the electrolyte, also called the membrane. The catalyst is typically pure platinum or a platinum alloy. The protons travel through the membrane, while the electrons are forced around to an external circuit since the membrane is an insulator. The electrons produce energy in the external circuit, and then meet the proton and oxygen at the cathode to produce the byproduct. In this case, the only byproduct is water. The cathode also uses a catalyst of either pure platinum or a platinum alloy.

Figure 1. Standard fuel cell diagram.

One of the major hurdles to commercializing the fuel cell is its high cost. One of the largest costs is the catalyst. Numerous approaches are taken to reduce the cost of the catalyst. One method is to find new catalysts. This has proven difficult because many synthesized catalysts are complicated to make, just as expensive, or lack the necessary performance. Another technique being studied is to optimize the location of catalyst and increase its surface area through the use of nanoparticles. While this is a useful approach no matter what catalyst is used, it has not been enough to make the fuel cell financially viable.

Another approach is to operate the fuel cell at a higher temperature. Proton exchange membrane fuel cells (PEMFC) currently operate around 80C because their membranes need to be hydrated for satisfactory proton conduction. This temperature is low enough to make the kinetics of the fuel cell reaction very slow unless a sufficient amount of a very active catalyst is used. The catalyst at these temperatures is also more sensitive to poisoning from impurities in the fuel. On the opposite spectrum is the solid oxide fuel cell (SOFC). They operate between 600C and 800C, allowing the use of non-precious metal catalysts due to the high temperatures. However, the materials to contain the fuel cell need to be special heat resistant ceramics, increasing the overall costs.

Hydroxyapatite top

The chemical formula for hydroxyapatite is Ca10(PO4)6(OH)2. It has a density of 3.156 g/cm^3 and a molecular weight of 502.31. Hydroxyapatite makes up 70% of bone and is the main mineral of dental enamel and dentin. Much research has been done on hydroxyapatite as a coating on prosthetic implants to allow better integration between the implant and natural bone.

Hydroxyapatite is a material that conducts protons as low as 200C. This property could be used in the production of an intermediate fuel cell. Operating between 200C to 300C would provide the advantages of the high and low temperature fuel cell. In this temperature range, a non-precious catalyst could still be used, but the temperature would not require expensive heat resistant ceramics. A membrane of hydroxyapatite or some hybrid material is a potential means for producing an intermediate fuel cell.

Experimental Procedure top

Growth of Hydroxyapatite Membrane top

The hydroxyapatite membrane fuel cell would have the same design as the standard fuel cell in Figure 1, but with different materials. Figure 2 below outlines the major steps in creating a hydroxyapatite membrane fuel cell. In our design, we first grow a thin layer (~20m) of palladium catalyst onto a porous stainless steel (PSS) substrate. The (PSS) substrate acts as support for the palladium catalyst. Then a seed layer of hydroxyapatite is coated onto the palladium catalyst. This seed layer allows for a controlled growth which we call the secondary growth. This secondary growth grows hydroxyapatite perpendicular to the substrate, allowing for a straight and direct path for proton conduction. The next step is a tertiary growth, which grows the crystals together. It is necessary to have a uniform distribution of hydroxyapatite to increase proton conduction and to prevent fuel from crossing over.

Figure 2. Schematic of hydroxyapatite growth.

Results and Discussion top

Micrographs of Growth top

Figures 3 through 5 below show micrographs of hydroxyapatite growth at each step. All micrographs were taken with either InLens or SE2 detectors and were uncoated. Coating was not possible for the first two sets of micrographs since the crystals were still being grown, causing some charging. The last set of micrographs were not coated for comparison purposes to other imaging techniques.

Figure 3. These two micrographs show the addition of a hydroxyapatite seed layer onto the palladium coated porous stainless steel substrate. The micrograph on the left shows a darker reagion where the crystals were not seeded. The figure on the right is a close shot of the submicron seed crystals.


Figure 4. These micrographs show hydroxyapatite crystals after secondary growth. This step grows hydroxyapatite crystals perpendicular to the substrate.


Figure 5. These micrographs show the hydroxyapatite crystals after the tertiary growth. This step grows the crystals together.


Micrographs with Different Techniques top

The above micrographs can be considered a baseline for two reasons. From the research standpoint, this is how the crystals should appear. The crystals are aligned parallel to each other and are perpendicular to the substrate. They are tightly packed with very few spaces between crystals. From a microscopy standpoint, these micrographs are as standard as can be. They do not have any coating and use the basic SE2 or InLens detectors. The following micrographs are either coated or uncoated with variable pressure or backscattered electron detection. The micrographs were also intentionally chosen for their aesthetic appeal, not necessarily their research value. In fact, many of the micrographs show deviations from desired research results. I hypothesize the cause of these deviations to be from my frequent use of the same three samples and not being as careful with them as I should have been.

Figure 6. The micrograph on the left shows crystals after the secondary growth that were uncoated using the SE2 detector. These crystals were not parallel to each other, but provided for a nice micrograph. The micrograph on the right shows crystals on the edge of my sample using a back scatter detector. The cyrstals often grew in this ball shape around the edge of my sample, allowing for more interesting micrographs.

Figure 7. Both of these micrographs use the variable pressure mode. The one on the left is coated, while the one on the right is uncoated.

Figure 8. More micrographs of uncoated variable pressure micrographs. These micrographs show the quality that can be obtained without the need for coating, which is useful for my seed layer and secondary growth steps.


Composition Analysis top

Figures 9 through 11 show three techniques for analyzing my sample. The spectra and quantified compositional data in Figures 9 and 10 respectively show that the main elements were oxygen, phosphorous, and calcium, the three elements of hydroxyapatite. The compositional maps in Figure 11 show a uniform distribution of these elements throughout the sample.

Figure 9. Spectra of grown hydroxyapatite crystals.


Figure 10. Quantified compositional data from hydroxyapatite crystals.


Figure 11. Composition maps of hydroxyapatite. The original micrograph is shown for reference purposes. The yellow map is calcium, the purple map is oxygen, and the green map is phosphorus. The orignial micrograph shows a few spots where crystal growth was not uniform. I oringinally thought this may be due to a contaminant or some compositional flaw. The maps show, however, that they are hydroxyapatite, just grown incorrectly.


Electron Flight Simulator top

Four electron flight simulations are shown in Figure 12. The simulations show the interaction volume of hydroxyapatite for 5, 10, 15, and 20 kV. The micrographs show an expected trend in that the penetration depth is increased with increasing beam strength. This information is particularly useful for my sample. My sample charges easily, especially when uncoated. These maps show how much beam would be needed for a given thickness of my sample. Due to my samples being quite thin, between 1 and 10 microns depending on growth conditions, I do not want a penetration depth too deep. Especially for compositional analysis, I do not want the penetration depth to be below the crystals.

Figure 12. Electron flight simulation for beam strengths between 5 and 20 kV.

Conclusion top

Summary top

Hydroxyapatite crystals were grown on a palladium coated porous stainless steel substrate. Their aligned and tight growth provides a potential means for producing an intermediate fuel cell. A variety of electron microscopy techniques were used and compared to analyze my samples. These techniques include comparing the difference between coated vs. uncoated, SE2, InLens, and backscattered detectors, variable pressure, EDAX spectra, maps, and quantification, and Electron Flight Simulator.

Acknowledgements top

My sincere gratitude goes out to Brian McIntyre who was patient enough to bestow his vast knowledge on electron microscopy to a student like myself who had never even seen an SEM prior to this class. I would also like to thank Andreas Liapis for teaching many practical skills during labs. Finally, I would like to thank Dongxia Liu for teaching me how to grow aligned hydroxyapatite crystals.

References top

1. http://en.wikipedia.org/wiki/Hydroxyapatite
2. Http://www.p2sustainabilitylibrary.mil/issues/emergeoct2005/fuelcell.jpg.
3. Liu, Dongxia, and Matthew Z. Yates. "Microstructural Engineering of Hydroxyapatite Membranes to Enhance Proton Conductivity."



Please enter any comments, criticisms, questions, etc. below.

Your name:

Email address: