Abstract This Project addresses steps towards developing a new type of thermoelectric power generation technique, and will function as gateway research to aid eventual invention and production of a revolutionary lightweight deployable solar power panel for space applications. The conventional method for generating power in space has come to be the deployment of fairly simple photovoltaic solar panels. Photovoltaic and thermoelectric cells are both brittle and heavy, but photovoltaic currently utilize cheaper materials, thus they are more commonly used. The research to be performed will test coatings and thermoelectric cells specifically for use in space environments, particularly once flexible low cost thermoelectric semiconductors have been developed. Old Dominion University Page 1
Table of Contents Sections Abstract page 1 Introduction page 3 Proposed Approach page 5 Organization page 8 Cost considerations page 9 Summary page 10 Appendix A page 11 Figures and Tables Figure 1-1 page 3 Figure 1-2 page 3 Figure 1-3 page 4 Figure 1-4 page 5 Figure 1-5 page 6 Figure 1-6 page 7 Figure A-1 page 11 Appendix B page 12 Old Dominion University Page 2
The thermoelectric effect is the direct conversion of a finite temperature difference, to an electronic potential difference, or voltage. Figure 1-1 shows the basic thermodynamic diagram. For more background information on the function of thermoelectric devices see Appendix A. Introduction Figure 1-1 Thermodynamic diagram The proposed experiments are to fabricate and test different materials under variable light and temperature conditions and to collect data on temperature, voltage and current output of the thermoelectric in question. Our objective is to produce a model, and show that collected data will be useful in future development of a thermoelectric cell design for harvesting the suns energy in space. Thermoelectric power generation in space shows great potential at being a vast improvement over current photovoltaic methods. Firstly, solar energy can be absorbed and turned into heat very easily, and finding an optimal configuration of materials for efficiently absorbing all wavelengths the Sun outputs is a matter of experimentation we plan to begin. Using photovoltaic cells, only parts of the visible spectrum of 390nm 750nm (see figure 1-2) can be used, and any emissivity in the infrared (greater than 750nm) is lost to the surroundings. Figure 1-2 Solar spectrum at sea level and outside the atmosphere by wavelength [1] Old Dominion University Page 3
Secondly, a flat plate in space facing directly towards the sun at a distance of 1 au will experience a temperature of 140 C on the sunny side, and -100 C on the dark side. Since Voltage directly correlates with temperature difference in a thermoelectric apparatus, and power directly with voltage, this 240 C temperature difference suggests thermoelectric devices would be very well suited for a space environment. Thirdly, many materials commonly used in coatings of photovoltaic cells could also be used in thermoelectric cells, which would suggest the industry could be easily capable of switching production from PV to thermoelectric cells cheaply. In order to take full advantage of what thermoelectric power generation has to offer, two coating layers are needed on the sunny side of the device. A material with a very high solar absorptivity on all wavelengths (2) is desired to contact the top surface of the thermoelectric semiconductors (3, 4). A perfect "blackbody" material would be ideal; however a blackbody also has perfect emissivity, and would reradiate heat in the infrared spectrum. To heavily reduce re-radiation, a material with a very low thermal emissivity can be placed on top of the absorbing layer (1). The material on the dark side must be of high emissivity overall in order to quickly radiate the rejected heat to the space behind it (3). Refer to the Figure 1-3. Figure 1-3 Diagram of a thermoelectric couple in a space environment (not to scale) Old Dominion University Page 4
Proposed Approach One major aspect of the proposed experiments is solar simulation, this will be done with an Oriel Solar simulator using a Xenon Light source and multiple filtering devices in order to closely approximate the light output of the sun (see Figure 1-4). Materials can first be tested for their ability to convert the solar simulators rays into heat, this will be the first step of our experiments. Zinc oxide, aluminum oxide, and some polymer composites will be briefly tested for low emissivity, as they already have uses in photovoltaic cells. Graphite is among the most promising high absorptivity and high emissivity materials to be tested. Figure 1-4 Cutaway showing the function of the Oriel Solar Simulator [2] The second aspect is actual fabrication of test solar power panels. The dimensions of the n-type and p-type semiconductors are set, as they come in 1x1mm blocks. Refer to Figure 1-3: Layer (1) requires low thermal emissivity, Zinc oxide, aluminum oxide, or a polymer composite will be used, and each one will be tested. Layer (2) would ideally absorb every wavelength for heat conversion at 100% efficiency, graphite is the likely candidate for this purpose. The thickness of these layers is among the variables with which we can tweak the outcome of the experiments. Layer (3) and Layer (4) are the n-type and p-type semiconductors, bismuth telluride and antimony telluride respectively will be used for these layers. The n-type and p-type semiconductors, while expensive, are readily available, and come in small blocks measuring 1x1 millimeters. Lastly, graphite will also be used in Layer (5). Old Dominion University Page 5
Custom thermoelectric panels must be constructed in lab to achieve the desired flexibility for testing. Since our research is not actually utilizing the flexible semiconductors which will hopefully one day be readily available, the major obstacle preventing the use of modern thermoelectric panels for this testing is a hard ceramic plate layer manufactured onto the semiconductors. The pre-manufactured panels are, consequently, hard to dismantle, and do not allow for the variability required for us to conduct our tests. It is also true that in a real space environment, the materials currently used as a base substrate would decay, the polyamide we will be using would be a functional choice for space applications. These custom thermo-electric panels are created simply by wiring many thermoelectric couples in electrical series, and thermo-dynamically in parallel (see Figure 1-5). An Aluminum conductor adhered to the semiconductors with a silver epoxy will be used to fill out the arrangement. This will have to be done in the lab, utilizing fine manipulation tools under a microscope, or perhaps innovation a faster method.[3] Figure 1-5 Thermoelectric module showing the direction of charge flow and wiring of individual thermoelectric couples. [3] Old Dominion University Page 6
The solar simulator, alongside temperature, current, and voltage probes attached to each thermoelectric panel we fabricate will provide us with data and insight into the efficiency of the proposed models for power generation. Figure 1-6 shows our experimental setup. Tests at both room temperature and simulated extraterrestrial conditions will be executed. In extraterrestrial conditions, the top layer of the thermoelectric is expected to experience an ideal temperature of 140 C, and the dark side of the panel to experience -100 C ideal ambient. To replicate these conditions, the top surface will be exposed to the solar simulator, and the bottom can be cooled using one of several methods including liquid nitrogen or other cold circulatory cooling techniques. Solar Simulator Organization Figure 1-6 Picture of the essential equipment in performing the thermoelectric tests The temperature, voltage and current data will be collected using respective thermal cameras and voltmeters fed into a computer running Labview software. Ultimately final results should give insight into the feasibility of Solar driven thermoelectric power generation in space. Old Dominion University Page 7
Organization The experiments timeline is divided into three main sections; our progress to between May to June, June to August, the fall. Until a new xenon light bulb is delivered for use in the solar simulator by the end of May, different coating materials for use for the fabrication of the thermoelectric solar cells will be the focus of testing. For the top layer of the thermoelectric solar cell, we will be studying different coating material with low emissivity like zinc oxide and aluminum oxide. For the mid and bottom layer of the thermoelectric solar cell we will be studying different materials with high emissivity like graphite. We will also experiment with different types of metals to use as conductors such as aluminum and silver paste. During the months June and August we will primarily focus on the fabrication of the different combinations to have tested in the ready solar simulator that will be provided to us to use. Because our team consist of two people, all task performed will require us to work cooperatively in the fabrication of the thermoelectric solar cells, as well as the testing of our solar cell using the solar simulator. During the end of the summer we will have an additional team member who currently work at NIA, who would assist us on the simulation of our different combinations of solar cells. Old Dominion University Page 8
Cost Considerations Heat dissipation in space is limited to radiation, as no fluid exists in space to convect. Consequently the primary use of thermoelectric devices in space at the present date is to cool electronic devices. The one exception to this is on the radioisotope thermoelectric generators like the one used on the Voyager probe [4]. The reason for this is, when compared with photovoltaic cells, thermoelectric panels, until now, have not promised enough improvement to be worth the added cost stemming from the expensive Bismuth Telluride and Antimony Telluride, thus thermoelectric devices are not currently utilized for power generation applications when sunlight is available. Many materials used in thermoelectric devices are extremely cheap and readily available, among these are Zinc Oxide, Aluminum Oxide, and highly spectrally absorptive carbon based materials such the graphite used in the proposed experiments. The Polyamide material to be used as the basis of our fabrication process is also relatively inexpensive. Promising research from multiple scientific establishments is investigating the ability to use polymer based nanocomposite semiconducting material for the n-type and p-type thermoelectric materials. By testing the best materials and technology available today, the proposed experiments can help provide guideline data for which specifications of future designs utilizing future materials can be based on. Upon the invention of new materials, relatively easy measurement of zt values can be used in conjunction with these experimental results to accomplish these guidelines. Replacing the rigid, brittle, heavy semiconductors with equally or even slightly higher priced polymers would yield heavy enough improvements in weight, deployability and flexibility of thermoelectric solar power generating models to make them more appealing than existing photovoltaic designs. Lastly, manufacturability of a new thermoelectric panel will not change drastically from previous methods for both thermoelectric and photovoltaic models, so it can be conjectured that there will be very little increase in cost due to manufacturing considerations. Old Dominion University Page 9
Summary The Seebeck effect and governing theories would suggest that the wide temperature gradient an object in space can experience would make thermometric solar power generation ideal in space. The proposed experiments aim to test the performance of materials at efficiently converting solar energy into heat, then to electricity. The research will function as a gateway for key future materials to replace current ones in order to achieve the creation of a revolutionary, lightweight, deployable solar power panel. Old Dominion University Page 10
Appendix A The Thermoelectric Effect The thermoelectric effect is the direct conversion of a finite temperature difference, to an electronic potential difference, or voltage. Thermoelectric effects occur due to the presence and mobility of charge carriers in semiconductors, and their ability to carry heat and charge simultaneously. Two types of thermoelectric materials are necessary to form a thermoelectric couple: n-type, which contains free electrons, and p-type, which contains free holes (see figure below). Voltage (V) is dependent upon the Seebeck coefficient (α) and the temperature difference (ΔT). [3] V = α T Current (I) is dependent upon the heat flow across the thermoelectric (Q), the temperature of the material (T) and the Seebeck coefficient (α). I = Q αt The efficiency of a thermoelectric material can be written as z = α2 ρk or zt = α2 T ρk Note that the second equation is simply the first equation multiplied by temperature. Since materials exhibit different conductivities at different temperatures, the (z) value will change accordingly. Hence, the efficiency is multiplied by T for accurate comparison. Figure A-1 simple thermoelectric couple. Old Dominion University Page 11
Appendix B References [1] Oriel Product Training, solar simulation Newport. Stratford, CT. <www.newport.com/oriel> [2] Jeffrey Snyder and Eric S Toberer. Complex Thermoelectric materials. Nature materials Vol 7 February 2008. (pg 105-112). California Institute of Technology [3]NASA Jet Propulsion Laboratory Voyager Interstellar missions. <http://voyager.jpl.nasa.gov/spacecraft/instruments_rtg.html> Old Dominion University Page 12