Metal Hydride Heat Pumps for Thermal Management of Devices

Transcription

1 University of Nevada, Reno Metal Hydride Heat Pumps for Thermal Management of Devices A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Materials Science and Engineering by Anupam Kumar Dr. Dhanesh Chandra, Advisor December 2013

2 THE GRADUATE SCHOOL We recommend that the dissertation prepared under our supervision by ANUPAM KUMAR entitled Metal Hydride Heat Pumps For Thermal Management Of Devices be accepted in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Dhanesh Chandra, Advisor Alan Fuchs, Committee Member Dev Chidambaram, Committee Member Wen-Ming Chien, Committee Member Jaak J.K. Daemen, Graduate School Representative Marsha H. Read, Ph. D., Dean, Graduate School December, 2013

3 i ABSTRACT Metal hydrides have very high volumetric energy densities which makes metal hydride heat pumps (MHHPs) particularly efficient devices. Further, a metal hydride heat pump (MHHP) can be tuned to cover a wide range of temperatures. Most phase change materials (PCMs) have a fixed phase change temperature, but metal hydrides are more versatile since they undergo phase change at all temperatures by adjusting the hydrogen equilibrium pressure. Currently, there is knowledge of heat pumps of large sizes, but we do not have data for small/miniature heat pump devices for either heating or cooling. The motivation for this work was to develop MHHPs for small scale applications. In the present work, experimental studies have been carried out on heat pumps working with different hydride pairs, such as MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4, MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22, and LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4, where Mm represents Mischmetal. The issues addressed in this work are miniaturization of a MHHP, electronic control of heat pump hydrogen gas transfer, heat pump performance and its dependence on gas pressure ratio in a MHHP, and mitigation of parasitic losses by reducing the amount of excess H 2 in the system. The electronic control of MHHP was accomplished by using a proton exchange membrane (PEM) with catalyst, thus replacing the mechanical valve which separates the two hydrides in a conventional heat pump. Such a heat pump can be recharged just with the reversal of polarity of the applied voltage to the PEM thereby obviating the need to heat the low pressure side of a metal hydride heat pump for system recharge.

4 ii In a typical MHHP, the equilibrium plateau pressures of the hydrides (in two different cylinders) are used to control the non-equilibrium thermodynamics. Minimum temperatures in the range of +16 o C to -12 o C were obtained by varying the equilibrium pressure ratios (P 1 /P 2 ) of the two hydrides from to 150 respectively. It was found that the minimum temperature or the degree of cooling is logarithmically dependent on the pressure ratio. The minimum pressure ratio for optimum functioning of the heat pump was found to be P 1 /P 2 = 0.2. Also, the presence of excess H 2 is detrimental to heat pump performance, for example, reducing the excess amount of H 2 gas in the system from ~2.25 moles to ~0 moles led to the reduction of minimum temperature from 20.5 o C to 7 o C. A minimum temperature of 7 o C was obtained with the heat pump (LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 ) in the absence of an external heat load (and under thermal insulation). Experiments conducted on the same system with an external heat load (and without thermal insulation) showed a drop in temperature of the heat load from ~79 o C to ~46 o C by the metal hydride, a reduction by 33 o C. The advantage of this system is that it operates at sub-atmospheric pressures, which has implications for safety. The results of the current investigations are indicative of the utilization of MHHPs with enhanced features such as electronic control of hydrogen flow for small scale heating and cooling applications. Keywords: Metal Hydride Heat Pump (MHHP), Proton Exchange Membrane (PEM), MmNi 4.15 Fe 0.85, LaNi 4.78 Sn 0.22, LaNi 4.6 Sn 0.4, Hydrogen, Equilibrium Plateau Pressure

5 iii ACKNOWLEDGEMENTS First and foremost, I would like to thank Dr. Dhanesh Chandra for his constant inspiration, his fresh perspectives, and support during the course of this work without whose encouragement this work would not have been possible. I thank Dr. Alan Fuchs, Dr. Dev Chidambaram, and Dr. Wen-Ming Chien for sparing their valuable time to be part of the dissertation committee. My very special thanks to Dr. Jaak J.K. Daemen for acting as the graduate school representative. I extend my gratitude and special thanks to Mr. Frank Lynch, of Hydrogen Components, Inc. for providing the materials necessary for this work and also for the keen insights that he brought to these investigations. I would like to thank Mr. Daryl J. Nelson and Dr. Murli Tirumala of Intel Corp. for the inputs received for the design of the device. Finally, I would like to thank my friends for their encouragement in all my endeavors.

6 iv TABLE OF CONTENTS Abstract...i Acknowledgments...iii Table of Contents...iv List of Figures...vii List of Tables...xvi 1. Introduction Motivation Energy Storage Materials Energy Storage in Metal Hydrides Metal Hydride Heat Pump (MHHP) Thermodynamics MHHP working with a Permeation Membrane Experimental Details Pressure-Composition-Temperature (P-C-T) Measurements Metal Hydride Heat Pump Experiments...30

7 v 2.3 MHHP Experiments under dynamic heating Metal Hydride Heat Pipe Experiments Flow Control of Hydrogen through Polymer Membrane MHHP operated with a PEM Results and Discussions Introduction Pressure-Composition-Temperature (P-C-T) Isotherms Room-temperature isotherm of LaNi 4.6 Sn Room-temperature isotherm of LaNi 4.78 Sn Development of Metal Hydride Pair Materials Heat Pump - MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 pair (P 1 /P 2 = 150) Heat Pump - MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 (P 1 /P 2 = 30) Heat Pump - LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 pair (P 1 /P 2 = 5) Heat Pump - LaNi 4.6 Sn 0.4 /LaNi 4.78 Sn 0.22 pair (P 1 /P 2 = 0.2) Heat Pump - LaNi 4.78 Sn 0.22 /MmNi 4.15 Fe 0.85 pair (P 1 /P 2 = 0.03) Heat Pump - LaNi 4.6 Sn 0.4 / MmNi 4.15 Fe 0.85 pair (P 1 /P 2 = 0.006)...65

8 vi 3.4 Metal Hydride Cooling System (MHCS) under Dynamic Heating Dynamic Experiments on Heat Pumps with varying pressure ratios MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 pair (P 1 /P 2 = 150) MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 (P 1 /P 2 = 30) LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 pair (P 1 /P 2 = 5) LaNi 4.6 Sn 0.4 /LaNi 4.78 Sn 0.22 pair (P 1 /P 2 = 0.2) Heat Pipe Type Experiment with metal hydride, LaNi 4.78 Sn Flow Control of Hydrogen through Permeation Membrane Potentiostatic Experiments Permeation behavior at room-temperature and 45 o C at 1 V Galvanostatic Experiments Summary Tables Summary of Fuel Cell Experiments Electronic Control of a Metal Hydride Heat Pump (MHHP) Minimization of parasitic losses in the heat pump Hydrogen Transfer from Hydride A to B (Reversed heat pump)...107

9 vii 3.11 Heat Pump working with the same Metal Hydride on both sides H 2 transfer from MmNi 4.15 Fe 0.85 hydride to an empty cylinder Summary, Conclusions and Future Work Summary MHHPs without external heat load MHHPs with external heat load Conclusions Future Work Appendix References...139

10 viii LIST OF FIGURES Figure 1.1 Comparative state-of-the art volumetric and gravimetric storage capacities of hydrides...5 Figure 1.2 Gravimetric vs. Volumetric hydrogen storage in some metal hydrides...5 Figure 1.3 A typical Metal Hydride Heat Pump (MHHP)...6 Figure 1.4 Schematic of a Metal Hydride Heat Pump (MHHP)...7 Figure 1.5 Temperatures of B and A during discharge. The cylinders were not insulated during discharge. The starting temperature was 25 o C...8 Figure 1.6 Temperatures of B and A during discharge inside a freezer. The starting temperature was 0 o C...8 Figure 1.7 Regeneration of a Metal Hydride Heat Pump (MHHP)...9 Figure 1.8 Regeneration of a Metal Hydride Heat Pump (MHHP). The heat supplied to the side A reaches 170 o C...10 Figure 1.9 van't Hoff plot for a typical heat pump operation...13 Figure 1.10 Derivation of van't Hoff plot from the isotherms obtained at various temperatures...16 Figure 1.11 A compilation of the van t Hoff plots of selected elemental, classical, and complex hydrides...17 Figure 1.12 Isotherms for MmNi 4.15 Fe wt% Al at three different temperatures...19

11 ix Figure 1.13 Isotherms for MmNi 4.15 Fe 0.85 at room-temperature...20 Figure 1.14 Hydrogen pressure vs Hydrogen content in LaNi 4.78 Sn Figure 1.15 Hydrogen pressure vs Hydrogen content in LaNi 4.6 Sn Figure 1.16 van't Hoff plots of LaNi 4.6 Sn 0. and LaNi 4.8 Sn Figure 1.17 Gravimetric and Volumetric Energy Density variation with temperature of LaNi 4.78 Sn Figure 1.18 Schematic of a Metal Hydride Heat Pump (MHHP) operated with a permeation membrane...25 Figure 1.19 Schematic of the operation of a proton exchange membrane (PEM)...27 Figure 2.1 Schematic showing the apparatus for PCT measurements...28 Figure 2.2 Experimental set-up of a Metal Hydride Heat Pump (MHHP)...30 Figure 2.3 Schematic showing the experimental set-up of a MHHP...31 Figure 2.4 Schematic and apparatus of a MHHP operation under dynamic heating. The cold side B receives heat from the hot source. Side A acts as the radiator...34 Figure 2.5 Schematic and apparatus of a MHHP operation under dynamic heating. Side A receives heat from the hot source and the cold side B acts as the radiator...35 Figure 2.6 Experiment to illustrate "Heat Pipe" type phenomenon using solid-gas hydride reactions in LaNi 4.78 Sn Figure 2.7 Schematic of hydrogen flow through a PEM...38

12 x Figure 2.8 Schematic of MHHP operation with a PEM...40 Figure 2.9 Experimental set-up of the MHHP with the membrane assembly...41 Figure 3.1 van't Hoff plots of MmNi 4.15 Fe 0.85, LaNi 4.78 Sn 0.22, and LaNi 4.6 Sn Figure 3.2 Hydrogen charging data of LaNi 4.6 Sn Figure 3.3 Room-temperature isotherm of LaNi 4.6 Sn Figure 3.4 P-C isotherm of LaNi 4.6 Sn 0.4. at 23 o C...49 Figure 3.5 Hydrogen charging data of LaNi 4.78 Sn Figure 3.6 Room-temperature isotherm of LaNi 4.78 Sn 0.22 at 25 o C...52 Figure 3.7 P-C isotherm of LaNi 4.78 Sn 0.22 at 25 o C...53 Figure 3.8 Discharge curves for the MHHP working with MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 pair (P 1 /P 2 = 150)...56 Figure 3.9 Discharge curves for the MHHP working with MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 pair (P 1 /P 2 = 30)...57 Figure 3.10 Discharge curves for the MHHP working with LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 pair (P 1 /P 2 = 5)...59 Figure 3.11 Discharge curves for the MHHP working with LaNi 4.6 Sn 0.4 /LaNi 4.78 Sn 0.22 pair (P 1 /P 2 = 0.2)...62 Figure 3.12 Discharge curves for the MHHP working with LaNi 4.78 Sn 0.22 /MmNi 4.15 Fe 0.85 pair (P 1 /P 2 = 0.03)...64

13 xi Figure 3.13 Discharge curves for the MHHP working with LaNi 4.6 Sn 0.4 /MmNi 4.15 Fe 0.85 pair (P 1 /P 2 = 0.006)...66 Figure 3.14 Minimum temperature as a function of pressure ratio...68 Figure 3.15 Heat pump performance under dynamic heating of hydride B...71 Figure 3.16 Comparison of differential cooling rates with hydride at a reference temperature of 25 o C and under dynamic heating...72 Figure 3.17 Heat pump performance under dynamic heating of hydride A...73 Figure 3.18 Dynamic test on the MHHP working with MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 pair...75 Figure 3.19 Cooling and Heating Coefficients of a MHHP working with MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 pair...77 Figure 3.20 Dynamic test on the MHHP working with MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 pair...79 Figure 3.21 Cooling and Heating Coefficients of a MHHP working with MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 pair...80 Figure 3.22 Dynamic test on the MHHP working with LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 pair...82 Figure 3.23 Cooling and Heating Coefficients of a MHHP working with LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 pair...84 Figure 3.24 Dynamic test on the MHHP working with LaNi 4.6 Sn 0.4 /LaNi 4.78 Sn 0.22 pair...85

14 xii Figure 3.25 Cooling and Heating Coefficients of a MHHP working with LaNi 4.6 Sn 0.4 /LaNi 4.78 Sn 0.22 pair...87 Figure 3.26 Comparison of the Cooling and Heating Coefficients of the MHHPs...88 Figure 3.27 Experiment to illustrate Heat Pipe type phenomenon using solid gas hydride reactions in LaNi 4.78 Sn Figure 3.28 Hydrogen permeation through Nafion membrane using electromotive force (Potentiostatic) at room-temperature at constant voltages of 20 mv, 40 mv, 60 mv, 80 mv, and 100 mv...95 Figure 3.29 Hydrogen permeation through Nafion membrane using electromotive force (Potentiostatic) at room-temperature and 45 o C at a constant voltage of 1 V...96 Figure 3.30 Hydrogen permeation through Nafion membrane using electromotive force (Galvanostatic) at room-temperature at constant currents of 10 ma, 20 ma, 30 ma, 40 ma, and 50 ma...97 Figure 3.31 Comparison of heat pump operation operated with a valve and PEM Figure 3.32 Heat Pump operation with positive (+100 mv) and negative (-100 mv) voltage for 4 hours each Figure 3.33 Heat Pump with the two hydrides, LaNi 4.78 Sn 0.22 and MmNi 4.15 Fe 0.85 in A and B cylinders, respectively...106

15 xiii Figure 3.34 Reversed Heat Pump where H 2 transfer is from the low pressure hydride (A: LaNi 4.78 Sn 0.22 ) to the high pressure hydride (B: MmNi 4.15 Fe 0.85 ) Figure 3.35 Comparison of the temperature profiles obtained by using "regular" and "reversed" heat pump configurations Figure 3.36 Heat Pump with the same metal hydride (MmNi 4.15 Fe 0.85 ) on both A and B sides Figure 3.37 Comparison of minimum and maximum temperatures on the cold side for MmNi 4.15 Fe LaNi 4.78 Sn 0.22 and MmNi 4.15 Fe MmNi 4.15 Fe 0.85 pairs as a function of excess H 2 moles Figure 3.38 H 2 transfer from MmNi 4.15 Fe 0.85 hydride to an empty cylinder Figure 3.39 Comparison of minimum cold-side temperatures. Temperature vs. excess H 2 moles Figure 3.40 Minimum cold-side temperatures as a function of the ratio of plateau pressures of the participating hydrides Figure A-1 Hydrogen permeation through Nafion membrane using electromotive force (Potentiostatic) at 30 o C at constant voltages of 20 mv, 40 mv, and 100 mv Figure A-2 Hydrogen permeation through Nafion membrane using electromotive force (Potentiostatic) at 35 o C at constant voltages of 20 mv, 40 mv, and 100 mv...125

16 xiv Figure A-3 Hydrogen permeation through Nafion membrane using electromotive force (Potentiostatic) at 40 o C at constant voltages of 20 mv, 40 mv, and 100 mv Figure A-4 Hydrogen permeation through Nafion membrane using electromotive force (Potentiostatic) at 45 o C at constant voltages of 20 mv, 40 mv, 60 mv, 80 mv, and 100 mv Figure A-5 Hydrogen permeation through Nafion membrane using electromotive force (Potentiostatic) with and without humidifying the membrane Figure A-6 Hydrogen permeation through Nafion membrane using electromotive force (Galvanostatic) at 40 o C at constant currents of 10 ma, 20 ma, 30 ma, 40 ma, and 50 ma Figure A-7 Hydrogen permeation through Nafion membrane using electromotive force (Galvanostatic) at 45 o C at constant currents of 10 ma, 20 ma, 30 ma, and 40 ma Figure A-8 Hydrogen permeation through Nafion membrane using electromotive force (Galvanostatic) at 50 o C at constant currents of 10 ma, and 20 ma Figure A-9 Hydrogen permeation through Nafion membrane using electromotive force (Galvanostatic) at room-temperature and 50 o C at a constant current of 20 ma Figure A-10. Gravimetric versus Volumetric hydrogen storage capacities of Mg based hydrides Figure A-11. Gravimetric versus Volumetric hydrogen storage capacities of Sodium Alanates...136

17 xv Figure A-12. Gravimetric versus Volumetric hydrogen storage capacities of Li based hydrides Figure A-13. Gravimetric versus Volumetric H 2 storage capacities of intermetallic hydrides Figure A-14. Gravimetric and Volumetric energy densities of various metal hydrides Figure A-15. Gravimetric and Volumetric energy densities of intermetallic hydrides...138

18 xvi LIST OF TABLES Table 1.1 Comparison of important thermal properties of solid-solid, solid-liquid, and solid-gas thermal energy storage materials...4 Table 1.2 Thermodynamic properties of intermetallic metal hydrides for heat storage applications...18 Table 3.1 Summary of results obtained with the three hydride pairs...67 Table 3.2 Effect of the presence of excess H 2 moles on the temperature profile of the heat pump Table 3.3 Effect of excess H 2 moles on the temperature profile of the reversed heat pump Table 3.4 Effect of excess H 2 moles on the temperature profile a heat pump working with same metal hydride (MmNi 4.15 Fe 0.85 ) on both sides Table 3.5 Effect of excess H 2 moles on the temperature profile for hydrogen transfer from MmNi 4.15 Fe 0.85 to an empty cylinder Table 3.6 Summary of results obtained with MHHP with varying ratio of plateau pressures (P 1 /P 2 ), mass of metal hydride, and system volume Table A-1 Summary of results of Potentiostatic experiments at different temperatures Table A-2 Summary of Galvanostatic experiments for hydrogen permeation at different temperatures...134

19 Chapter 1 INTRODUCTION 1.1 Motivation Thermal energy storage materials store thermal energy by undergoing phase changes. The stored energy can be utilized for many purposes such as in heating or cooling devices. Thermal energy storage materials can be classified broadly into three types based on the phase transformation they undergo. These are (1) Solid-to-Solid Phase Change Materials (SSPCMs), (2) Solid-to-Liquid Phase Change Materials (SLPCMs), and (3) Solid-to-Gas Phase Change Materials (SGPCMs). Each material offers an advantage over the other. While most of the PCMs in vogue for cooling purposes are Solid-to-Solid Phase Change Materials (SSPCMs) and Solid-to-Liquid Phase Change Materials (SLPCMs) used mainly for their good gravimetric energy density, their volumetric energy density is not very good. This necessitates the exploration of novel materials which have high volumetric energy densities. Typically, energetic Solid-to-Solid Phase Change Materials (SSPCMs), such as Pentaglycerine (PG) have a maximum enthalpy of phase transition or storage capacity of x 10 6 J/Kg or 235 x 10 6 J/m 3. In order to enhance the storage properly 3-4 times, metal hydrides can be used which have high enthalpies, for example x 10 6 J/Kg or 1584 x 10 6 J/m 3 in the case of LaNi 4.78 Sn Thus, volumetrically ~3.7 times more energy can be stored in the classical metal hydrides. In modern (light weight) hydrides this is enhanced significantly.

20 2 Metal hydrides fall in the category of Solid-to-Gas Phase Change Materials (SGPCMs) which have many applications such as batteries, hydrogen storage, thermal energy storage, heating, and cooling. They have the additional advantage of being environment friendly. Usually, metal hydrides are used for cooling purposes on a large scale and have been successfully used for cooling of space telescopes [1][2][3], refrigeration, and airconditioning [4][5]. In fact, a temperature as low as K can be reached (where LaNi 4.78 Sn 0.22 is used for hydrogen storage) for cooling of telescopes [1] where very large amount of metal hydride is used at a pressure of 48 bar [3]. They have never been used for small scale cooling purposes at atmospheric or sub-atmospheric pressures, partly because of high initial cost of the alloys and partly because of design issues. Nonetheless, they are promising for small scale applications as well as they are compact. Most phase change materials (PCMs) have a fixed phase change temperature but metal hydrides are more versatile since they undergo phase change at all temperatures by adjusting the hydrogen equilibrium pressure. This fact can be exploited to tailor metal hydrides to develop heat storage devices suitable for a wide range of applications. The exploration of PCMs with high volumetric energy densities for small scale cooling or heating purposes is the primary motivation behind this work. 1.2 Energy Storage Materials As discussed in section 1.1, thermal energy storage materials can be classified into three types based on the type of phase transformation they undergo on the application of thermal energy: (1) Solid-to-Solid Phase Change Materials (SSPCMs), (2) Solid-to- Liquid Phase Change Materials (SLPCMs), and (3) Solid-to-Gas Phase Change Materials

21 3 (SGPCMs). Excellent examples of a Solid-to-Gas Phase Change Materials (SSPCMs) are Polyalcohols such as Pentaerythritol (PE), Pentaglycerine (PG), Neopentylglycol (NPG), and tris(hydroxymethyl)amino methane (TRIS). Salts, for example, Glauber and Hydrated Eutectic are examples of Solid-to-Liquid Phase Change Materials (SLPCMs). Metal hydrides are examples of Solid-to-Gas Phase Change Materials (SGPCMs). A comparison of the important thermal properties of metal hydrides, SSPCMs, and other materials is shown in Table Energy Storage in Metal Hydrides Contemporary hydrides are classified as (A) Classical Metal Hydrides and (B) Complex Hydrides. The classical hydrides have been thoroughly investigated and some have been perfected to operate under repeated cycling. The complex hydride field is rather new and has proliferated in this century. The group at the University of Nevada, Reno has been working on both types of hydrides for quite some time now. Typical storage capacities are listed in Figure 1.1; taken from the Swiss group (Zuttel et. al [38] ) which were also presented at the IPHE/IEA meeting in Lucca, Italy. The volumetric vs. gravimetric hydrogen storage capacities of some hydrides are shown in Figure 1.2. The thermal energy densities along with the gravimetric and volumetric uptake of different metal, intermetallic, and complex hydrides is given in Appendix I.

22 4 Table 1.1 Comparison of important thermal properties of solid-solid, solid-liquid, and solid-gas thermal energy storage materials (Chandra et al. [15-37] ). Substance Phase Change Temp. ( o C) Density (kg/m 3 ) Enthalpy MJ/kg MJ/m 3 SOLID GAS TRANSFORMATIONS La 0.9 Gd 0.1 Ni 5 H x +G LaNi 5.2 H x +G LaNi 4.8 Sn 0.22 H x +G LaNi 4.75 In 0.27 H x +G SOLID LIQUID TRANSFORMATIONS Ice L Eutectic Salt L (Hydrated) Glauber Salt L n-hexadecane L SOLID SOLID TRANSFORMATIONS Pentaerythritol (PE) Pentaglycerine (PG) Neopentyglycol (NPG) Neopentylalcohol (NPA) 2-amino 2-methyl 1, propanediol (AMPL) tris(hydroxymethyl)am inomethane (TRIS)* 50%TRIS mol.%PE* 75%TRIS mol.%PE* *New Results

23 5 Figure 1.1 Comparative state-of-the art volumetric and gravimetric hydrogen storage capacities of hydrides (Courtesy of Zuttel et al. [38] ) Gravimetric and Volumetric Capacities of Hydrides Mg(BH 4 ) Intermetallic Hydrides MmNi 5 Al(BH 4 ) 2 Volumetric H2 Density (kg H2 m 3 ) LaNi 5 LaNi 4.8 Sn 0.2 ZrFe 1.5 Cr 0.5 MmNi 4.5 Al 0.5 MmNi 3.5 Co 0.7 Al 0.8 MmNi 4.15 Fe 0.85 LaNi 4.25 Al 0.75 TiFe 0.8 Ni 0.2 MgH 2 Mg(NH 2 ) 2 Ca(NH 2 ) 2 NaAlH4 KBH 4 LiNH 2 NaBH 4 LiAlH 4 LiBH NaNH KNH Gravimetric H 2 Density (mass%) Figure 1.2 Gravimetric vs. Volumetric hydrogen storage in some metal hydrides.

24 6 1.4 Metal Hydride Heat Pump (MHHP) The functionality of the heat storage device is derived by the use two metal hydrides in a heat pump mode called a Metal Hydride Heat Pump (MHHP). Metal Hydride Heat Pump (MHHP) consists of two cylinders containing metal hydrides with different affinities for hydrogen gas. A typical MHHP is shown in Figure 1.3 and consists of a high-pressure (HP) and a low-pressure (LP) hydride that charge and discharge by a thermodynamic drive for energy storage. The high-pressure (HP) hydride is the low-temperature (LT) hydride usually referred to as the cold side. Similarly, the low-pressure (LP) hydride is the high-temperature (HT) hydride usually referred to as the hot side. Pressure Transducer Valve H 2 Hydride Reactor Figure 1.3 A typical Metal Hydride Heat Pump (MHHP).

25 7 H 2 Figure 1.4 Schematic of a Metal Hydride Heat Pump (MHHP). Figure 1.4 shows the schematic of a Metal Hydride Heat Pump (MHHP). It consists of a low-pressure, high-temperature (LP-HT) metal hydride A (the hot side) and a highpressure, low-temperature (HP-LT) metal hydride B (the cold side) separated by a mechanical valve. The pressure transducers are mounted on top of both cylinders for pressure recording. A fully charged MHHP is at ambient temperature until the valve is opened. Initially, the high-pressure, cold side is charged with H 2 gas and the alloy is in a hydrided state whereas the low-pressure, hot side is devoid of any H 2. Then hydrogen is desorbed from the hydride in cylinder B and transferred to the alloy in cylinder A. The reaction in B is endothermic. On the other hand, the absorption in A is exothermic. The temperature of hydride B decreases whereas that of A increases. Figure 1.5 shows the results of such a process when the cylinders are not insulated, discharged at 25 o C. A discharge process inside a freezer (at 0 o C) is shown in Figure 1.6.

26 8 Figure 1.5 Temperatures of B and A during discharge. The minimum temperature (Hydride B) is 0 o C and the maximum temperature (Hydride A) is 65 o C. The cylinders were not insulated during discharge. The starting temperature was 25 o C. (Previous Data) Figure 1.6 Temperatures of B and A during discharge inside a freezer. The minimum temperature (Hydride B) is -23 o C and the maximum temperature (Hydride A) is 35 o C. The cylinders were not insulated during discharge. The starting temperature was 0 o C. (Previous Data) Thus, metal hydrides act as "on-demand heat storage" media with very small amount of hydrogen gas present in the system. The system may be recharged by waste heat via a

27 9 heat exchanger while absorbing heat from a device. Figure 1.7 shows such a recharging process where a silicone rubber heater is wrapped around the side A to push hydrogen back into side B. The result of such a regeneration process is shown in Figure 1.8. Figure 1.7 Regeneration of a Metal Hydride Heat Pump (MHHP). A silicone rubber heater is wrapped around the side A and hydrogen is pushed back to side B. One of the best possible ways to make use of waste heat (also known as low grade heat) is to use it for heating or cooling applications. Metal Hydride Heat Pumps (MHHPs) are highly effective for this purpose because of the high enthalpy of metal hydrides and the environmental friendliness of the working fluid of heat pumps (H 2 gas).

28 10 Figure 1.8 Regeneration of a Metal Hydride Heat Pump (MHHP). It is noticeable that the temperature of the cold side B remains at around room-temperature while absorption. The heat supplied to the side A reaches 170 o C. (Previous Data) As mentioned already, Metal-Hydride/Hydrogen (MH/H 2 ) (Solid-Gas) systems can be used in a particular type of chemical heat pump where the exothermic nature of hydride formation and the endothermic nature of dehydrogenation can be used for heating or cooling applications. A Metal Hydride Heat Pump (MHHP), when used for cooling purposes is known as a Metal Hydride Cooling System (MHCS). Metal Hydride Cooling Systems (MHCSs) are different from conventional cooling systems such as heat pipes where cooling is performed based upon the liquid to gas phase transformation of the working fluid. In MHHPs, the thermodynamic interaction (endothermic and exothermic reactions) between

29 11 a pair of alloys and H 2 gas is responsible for achieving either cooling or heating. The chemical reactions for hydride formation and dehydrogenation can be written as follows: M 1 H 2 2 MH Q (1.1) 1 MH Q H 2 2 M (1.2) In addition to being environmentally clean and safe, Metal Hydride Cooling Systems (MHCSs) offer various advantages over conventional cooling systems such as their tunability over wide ranges of temperatures and pressures which makes it possible to tailor these devices for specific cooling needs. The pressure and temperature characteristics of these alloys can be changed at will just by a fractional change in the alloy composition. But their biggest advantage lies in their very high enthalpies of formation. Metal hydride heat pumps (MHHP) for heating and cooling applications have been reported in literature [39-54]. The efficiency of these heat pumps can further be enhanced by employing some simple methods to optimize heat transfer characteristics of the metal hydrides. The simulation of heat and mass transfer characteristics of hydride beds has been performed by many groups [56-58]. Though the MHHP works pretty well even if only metal hydride powders are used in the reactor beds, the system efficiency can be increased by using adding Oxygen-free high thermal conductivity copper (OFHC Cu) which increases the effective thermal conductivity of the hydride beds. The efficiency of a heat pump can also be increased by increasing the hydrogen supply per unit volume of the alloy. A simple technique that can be used in order to achieve this is to use a hollow

30 12 pipe which runs vertically all through the alloy sample for hydrogen supply. These techniques not only enhance the performance of heat pumps but also minimize the cost of the system as less alloy is needed to achieve the desired results. [54] These techniques become more important especially in miniaturized samples where one can expect more parasitic heat losses. But we have limited the current experiments to using only hydride powders in small cylindrical reactors without employing any of these techniques. 1.5 Thermodynamics The processes of hydrogen transfer across the two reactors and the corresponding hydriding and dehydriding characteristics of the alloys in the two reactors have been schematically represented in the ln P vs 1/T plot of Figure 1.9. It can be seen from Figure 1.9 that the energy transfer via the hydrogen gas occurs at a high pressure ln P, in between the reactors. The thermal energy excursions in between the reactors takes place via the H 2 gas. The enthalpies and entropies of hydriding and dehydriding of the alloys are governed by the van't Hoff equation: f f H S ln P (1.3) RT R The complete operation cycle occurs in the following four processes: a-b. The LP-HT hydride AH x receives heat from the source (Q 1 ) and the valve is kept closed. Since dehydriding is an endothermic process, the temperature of A is reduced.

31 13 van't Hoff Plot for High and Low Pressure Hydrides H 2 c Heat Intput (Q 3 ) ln P d Heat Output (Q 4 ) a High Pressure Hydride Heat Output (Q 2 ) Heat from Source (Q 1 ) H 2 b Low Pressure Hydride 1/T Figure 1.9 van't Hoff plot for a typical heat pump operation. Then the valve is opened and H 2 gas is transferred to alloy B whereby hydride BH x is formed and its temperature increases. b-c. As a result of the exothermic reaction between alloy HP-LT alloy B and hydrogen gas, heat is produced (heat output Q 2 ) which is lost to the surroundings. (Valve is kept open).

32 14 c-d. This is the regeneration process where the hydride BH x is heated (heat input Q 3 ) or the valve is opened to form hydride AH x resulting in the increase of temperature of side A (Valve is still open). The exothermic reaction in A results in the production of heat (heat output Q 2 ). Valve is closed to allow pressure to build up in the reactor. d-a. The valve is then opened. Pressure at side A decreases resulting in the increase in temperature. A typical isotherm of a reversible hydride referred to as Pressure-Composition (P-C) isotherm measured at a particular temperature is shown in Figure 3 (a). The number of hydrogen to metal atoms (H/M) is represented on the x-axis whereas the pressures are represented on the y-axis. When hydrogen is initially introduced into the crystal lattice, a solid-solution (α-phase) is formed[15]. Here, hydrogen absorption is controlled by the dissociative chemisorption given by: H 2H 2 (1.4) The initial addition of H2 is governed by Sievert's Law [15]: H 0.5 kp (1.5) M where k is a constant dependent on temperature and P is the pressure. The crystal lattice becomes saturated with H-atoms and a second phase, designated as β-phase, begins to appear [15]. As the transformation from α to β increases at nearly constant pressure, a flat plateau region begins to appear and is governed by the equations:

33 15 M y H 2 2 MH y (α-phase) MH x y 2 y H 2 MH x (β-phase) (1.6) The actual behavior in most cases is somewhat different, where a sloping plateau appears as the lattice expansions corresponding to different equilibrium pressures are different. The slope of the plateau is given by the expression: (ln P) Slope (1.7) ( H / M ) Figure 1.10 depicts the generation of van't Hoff plot. A typical Pressure-Composition (P- C) isotherm is shown in Figure 1.10 (a). Pressure-Composition (P-C) isotherms at different temperatures are collected as shown in Figure 1.10 (b). The mid-point of the plateau pressures at different temperatures are plotted against 1/T in order to get the van't Hoff plot. The slope of the van't Hoff is equal to H/R and the enthalpy of a particular metal hydride can be easily calculated by calculating the slope of the van't Hoff.

34 16 Figure 1.10 (a) An isotherm showing a sloped plateau for hydrogen absorption and desorption isotherms. Hysteresis between absorption and desorption isotherms is also shown. (b) The effect of temperature on the isotherm plateau pressure and phase transitions regions from α α + β β are shown. (c) The van't Hoff plot derived from the isotherms obtained at various temperatures; whose slope yields the enthalpy of hydriding. Chandra et al. [15] A compilation of the van t Hoff plots of selected elemental, classical, and complex hydrides is shown in Figure The boxed area (red) represents the desired temperature and pressure range of operation for vehicular applications. The boxed area (blue) represents the desired temperature and pressure range of operation for thermal energy storage applications. Important thermodynamic properties of metal hydrides suuch as enthalpies of formation/dehydrogenation, Gibbs Free Energy, and entropies are given in Table 1.2.

35 17 Figure 1.11 A compilation of the van t Hoff plots of some metal hydrides. The boxed area (blue) represents the desired temperature and pressure range of operation for thermal energy storage applications. Adapted from hhtp://hydpark.ca.sandia.gov. [59]

36 18 Table 1.2 Thermodynamic properties of intermetallic metal hydrides for heat storage applications. T=298 K T=298 K Enthalpy ΔH o T (kj/mol H2) Gibbs Free Energy ΔG o T (kj/mol H2) Entropy ΔS o T (kj/mol H2.K) LaNi5.2 + H2» LaNi5H6 (A) LaNi5H6» LaNi5.2 + H2 (D) La0.9Gd0.1Ni5 + H2» La0.9Gd0.1Ni5H6 (A) La0.9Gd0.1Ni5H6» La0.9Gd0.1Ni5 + H2 (D) LaNi4.8Sn H2» LaNi4.8Sn0.22H6 (A) LaNi4.8Sn0.22H6» LaNi4.8Sn H2 (D) LaNi4.75In H2» LaNi4.75In0.27H6 (A) LaNi4.75In0.27H6» LaNi4.75In H2 (D) In the current research six pair of heat pumps with differing ratios of plateau pressures in between the HP-LT and LP-HT hydrides have been investigated. The hydride pairs are MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 (ratio of plateau pressures = 150) MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 (ratio of plateau pressures = 30) LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 (ratio of plateau pressures = 5) LaNi 4.6 Sn 0.4 /LaNi 4.78 Sn 0.22 (ratio of plateau pressures = 0.2) LaNi 4.78 Sn 0.22 /MmNi 4.15 Fe 0.85 (ratio of plateau pressures = 0.03) LaNi 4.6 Sn 0.4 / MmNi 4.15 Fe 0.85 (ratio of plateau pressures = 0.006)

37 19 The Pressure-Composition isotherms for MmNi 4.15 Fe wt% Al, MmNi 4.15 Fe 0.85, LaNi 4.78 Sn 0.22, and LaNi 4.6 Sn 0.4 are shown in Figure 1.12, Figure 1.13, Figure 1.14, and Figure 1.15 respectively. Figure 1.12 Isotherms for MmNi 4.15 Fe wt% Al at three different temperatures (18 o C, 42 o C, and 65 o C). Ron et al. [60] It can be noticed by comparing figures 1.12 and 1.13 that the addition of 18 wt% Al to MmNi 4.15 Fe 0.85 considerably brings down the plateau pressure of MmNi 4.15 Fe 0.85 from 15 bar to 5 bar. In general, the addition of elements like Sn, Al, Mn, and Fe brings down the plateau pressures of these alloys because of the elimination of stacking faults from the crystal lattice as a result of which the absorption or desorption capacity increases.

38 20 Figure 1.13 Isotherms for MmNi 4.15 Fe 0.85 at room-temperature (Courtesy of Frank Lynch, HCI) It should be noted here that Mm (in MmNi 4.15 Fe 0.85 ) is the symbol used for an alloy Mischmetal, which is composed on the rare earth elements Lanthanum (La), Cerium (Ce), Neodymium (Nd), and Praseodymium (Pr). In a typical Mischmetal (Mm) alloy, Lanthanum (La), Cerium (Ce), Neodymium (Nd), and Praseodymium (Pr) are present 25%, 48%, 17%, and 5% by weight.

39 21 Figure 1.14 Hydrogen pressure vs Hydrogen content in LaNi 4.78 Sn 0.22 at 25 o C, 42 o C, and 93 o C. Chandra et al. [16] Figure 1.15 Hydrogen pressure vs Hydrogen content in LaNi 4.6 Sn 0.4 (pointer: red arrow) [61]

40 22 The plateau pressures of MmNi 4.15 Fe 0.85, LaNi 4.78 Sn 0.22, and LaNi 4.6 Sn 0.4 at roomtemperature are 15 bar, 0.5 bar, and 0.1 bar respectively. The van't Hoff plots for LaNi 4.78 Sn 0.22 and LaNi 4.6 Sn 0.4 are shown in Figure The van't Hoff plot of MmNi 4.15 Fe 0.85 has already been shown in Figure Figure 1.16 van't Hoff plots of LaNi 4.6 Sn 0.4 (circled red) and LaNi 4.8 Sn 0.2 (circled blue). [61] Two performance criteria can be used to quantify the performance of a heat pump: cooling coefficient and heating coefficient. Each of them is a good numerical measure of a heat pump's performance. The cooling coefficient can be defined as the temperature reduction incurred due to heat absorption by the cold side of a heat pump per unit input

41 23 temperature on the cold side. Similarly, the heating coefficient can be defined as the temperature amplification on the hot side of a heat pump per unit input temperature on the cold side. Mathematically, C COOL TB 1 (1.8) T Source( B) T A CHEAT (1.9) TSource(B) The variation of gravimetric and volumetric enthalpies LaNi 4.78 Sn 0.22 with temperature is shown in Figure Figure 1.17 (a) shows the gravimetric enthalpy variation with temperature whereas Figure 1.17 (b) shows the volumetric enthalpy variation with temperature. It can be seen from the figures that enthalpy in both cases increases with temperature. The gravimetric energy density varies linearly from 70 J/g to 500 J/g in the temperature range of o C whereas the volumetric energy density varies linearly from 600 J/cc to 4200 J/cc in the same temperature range.

42 Gravimetric Energy Density vs Temperature (LaNi 4.78 Sn 0.22 ) Gravimetric Energy Density (J/g) (a) Temperature ( o C) Volumetric Energy Density vs Temperature (LaNi 4.78 Sn 0.22 ) 4000 Volumetric Energy Density (J/cc) (b) Temperature ( o C) Figure 1.17 Gravimetric and Volumetric Energy Density variation with temperature of LaNi 4.78 Sn The plots were derived using Kirchhoff's relation: H f H i C p.( T f Ti ) (1.10) where H f is the desired enthalpy, H i is the known enthalpy at a certain temperature. T f is the temperature at which enthalpy is required and T i the temperature at which enthalpy

43 25 is known experimentally or otherwise. The enthalpy of LaNi 4.78 Sn 0.22 at 20 o C is J/mol. The molecular weight is 445 g/mol. The specific heat capacity, C p of LaNi 4.78 Sn 0.22 is J/g/K and its density is 8.4 g/cc. These values are used in equation 11 to give the plots of Figure Metal Hydride Heat Pump (MHHP) operating with a Permeation Membrane A Metal Hydride Heat Pump (MHHP) operates with a mechanical valve separating the two reactors which has to be operated manually. This can be replaced with an electronically controlled permeation membrane so that the need for manual operation is obviated. A schematic of such a set up is shown in Figure Figure 1.18 Schematic of a Metal Hydride Heat Pump (MHHP) operated with a permeation membrane. The replacement of the mechanical valve with an electronically controlled permeation membrane offers the following advantages: The heat pump need not be operated manually.

44 26 The heating and cooling can be controlled by controlling the flow rate of H 2 gas in between the reactors. The flow of H 2 gas can be reversed so that the MHHP can be regenerated without heating the low pressure hydride by reversing the applied voltage to the permeation membrane. A proton exchange membrane (PEM) can be used as the permeation membrane. PEMs work by the application of a voltage/current at the anode. The applied voltage breaks up the H 2 molecule at the anode into protons and electrons. The protons pass through the PEM and reach the cathode. The electrons travel through the external circuit and reach the cathode where they combine with the protons to form hydrogen gas. Anode is also known as the Working Electrode and cathode is also known as the Counter Electrode (CE). The higher the applied voltage, the higher is the permeation rate of H 2 gas through the membrane, in essence higher H 2 flow is achieved rate across the two terminals. A schematic of H 2 flow through a PEM is shown in Figure The reactions that take place at the anode and cathode are: H 2 2 2H e (Half-cell, Anode) (1.11) 2H 2e H 2 (Half-cell, Cathode) (1.12) H 2 ( Anode) H 2 ( Cathode) (Overall) (1.13)

45 Figure 1.19 Schematic of the operation of a proton exchange membrane (PEM). 27

46 28 Chapter 2 EXPERIMENTAL 2.1 Pressure-Composition-Temperature (PCT) Measurements The schematic of the apparatus for measuring the Pressure-Composition-Temperature (PCT) isotherm is shown in Figure 2.1. It consists of a known calibration volume (40 cc), the reactor containing metal hydride, pressure transducer, and a tank containing H 2 for supply to the reactor. The pressure transducer and a thermocouple (K-type) connect the reactor to the data acquisition system for pressure and temperature measurements. Valves V 0, V 1, V 2, and V 3 connect the various system parts to each other via stainless steel pipes. Figure 2.1 Schematic showing the apparatus for PCT measurements.

47 29 The PCT measurement takes place in the following manner: Initially, there is no H 2 in the system except the tank (H 2 supply). The whole system is under vacuum and at room-temperature all the valves are closed except V 2. Valve V 0 is opened keeping V 2 open other valves closed and H 2 at a known pressure is allowed to fill the calibration volume (40 cc). Valve V 0 is closed. Valve V 1 is then opened keeping the other valves closed. H 2 gas enters the reactor via V 2 containing the intermetallic alloy (which initially is devoid of any H 2 ) and as a result an exothermic reaction takes place inside the reactor. The reactor is allowed to come back to room-temperature. When the temperature of the reactor reaches room-temperature, valve V 1 is closed. Valve V 0 is opened to allow the calibration volume to be filled with fresh H 2 gas. The whole process is repeated for a number of cycles till the alloy becomes saturated with H 2. The equilibrium pressure (the pressure at room-temperature) for each cycle is plotted against time or H/M in order to get the PCT isotherm.

48 Metal Hydride Heat Pump (MHHP) Experiments The experimental set-up for the MHHP is shown in Figure 2.2. The apparatus consists of two metal hydrides (B and A) in two cylinders separated by a mechanical valve. Two pressure transducers are mounted on top to record pressure in each cylinder. A calibration volume (40 cc) is connected to the high pressure hydride B for keeping track of the amount of H 2 absorbed in B during charging. Two thermocouples (K-type) connect the reactors to the data acquisition system temperature measurements. Each reactor contains g of metal hydride in addition to 4.75 g of teflon. Teflon is added to the alloys in order to provide fluidity to the metal hydrides. Figure 2.2 Experimental set-up of a Metal Hydride Heat Pump (MHHP)

49 31 The schematic of the MHHP is shown in Figure 2.3. The tank supplies H 2 to the highpressure hydride B for charging via the calibration volume. A mechanical pump is connected at the end of the system to remove hydrogen from the alloys at the end of an experiment or to remove excess hydrogen. Valves V 0, V 1, V 2, V 3, V 4,and V 5 connect the various system parts to each other via stainless steel pipes. Figure 2.3 Schematic showing the experimental set-up of a MHHP.

50 32 The operation of the MHHP consists of charging and discharge cycles. The charging process consists of the following steps: Initially, there is no H 2 in the system including the alloys B and A. The whole system is under vacuum and at room-temperature all the valves are closed except V 2. Valve V 0 is opened keeping V 2 open and all other valves closed and H 2 at a known pressure is allowed to fill the calibration volume (40 cc). Valve V 0 is closed. Valve V 1 is then opened keeping the other valves closed (V 2 is open). H 2 gas enters the reactor via V 2 containing the high-pressure alloy B (which initially is devoid of any H 2 ) and as a result an exothermic reaction takes place inside the reactor B. Reactor B is allowed to come back to room-temperature. When the temperature of the reactor reaches room-temperature, valve V 1 is closed. Valve V 0 is opened to allow the calibration volume to be filled with fresh H 2 gas. The whole process is repeated for a number of cycles till B becomes saturated with H 2. This is determined by the absence of any pressure drop in the reactor B.

51 33 After metal hydride B is fully charged, the discharge process is carried out in the following manner: Valves V 2 and V 4 are open and all other valves closed. Both the reactors are insulated. Valve V 3 is opened. H 2 gas from cylinder B reaches cylinder A. As a result, an exothermic reaction takes place in A and an endothermic one in B. Pressure in both the cylinders decreases because of the absorption of H 2 gas by hydride A. The system is kept running until temperature in both the reactors reaches roomtemperature. 2.3 Metal Hydride Heat Pump (MHHP) Experiments under dynamic heating In order to determine the efficacy of a heat pump for cooling applications, experiments under dynamic heating conditions need to be performed when the MHHP is in a fully charged condition. The heat source can be experimentally simulated by using a heater such as a flexible silicone rubber heater. The experimental set-up is shown in Figure 2.4. Side B contains the high-pressure, low-temperature (HP-LT) hydride MmNi 4.15 Fe It consists of 111 g of the alloy in addition to 12 g of TFE powder which acts as the "flow aid" for the metal hydride. Side A contains the low-pressure, high-temperature (LP-HT) hydride LaNi 4.78 Sn It consists of 140 g of the alloy without any TFE powder.

52 34 Figure 2.4 Schematic and apparatus of a MHHP operation under dynamic heating. The cold side B receives heat from the hot source. Side A acts as the radiator. The experimental procedure is as follows: Side B is fully charged with hydrogen at a pressure of 350 psi. Most of the hydrogen is on side B (inside MmNi 4.15 Fe 0.85 and over it) and a negligible amount is on side A. The valve separating the cylinders is closed. The system is allowed to come to room-temperature. A very thin polyimide tape is put around cylinder B on top of which a thermocouple is placed. Another layer of polyimide tape is put around the cylinder such that it covers the thermocouple. A flexible silicone rubber heater is wrapped around side B. The valve is opened. Simultaneously, side B is gradually heated with the heater from room-temperature to 80 o C over 2 hours at a heating rate of 0.5 o C/min in order to simulate the heat received from a hot source.

53 35 The temperatures on both sides of the cylinders are recorded. In order to measure the source temperature, all the H 2 from side B is removed after which the valve is closed. The heat pump is allowed to come back to room-temperature and the cylinder is heated at a temperature ramp of 0.5 o C/min for 2 hours. The recorded temperature gives the source temperature. In a similar manner, the cooling obtained when side A receives heat from the hot source is measured. Figure 2.5 depicts the schematic of the process. Figure 2.5 Schematic and apparatus of a MHHP operation under dynamic heating. Side A receives heat from the hot source and the cold side B acts as the radiator. The complete experimental procedure is as follows: Side A is fully charged with hydrogen at a pressure of 10 psi. Most of the hydrogen is on side A (inside LaNi 4.78 Sn 0.22 and over it) and a negligible amount is on side B. The valve separating the cylinders is closed.

54 36 The system is allowed to come to room-temperature. A very thin polyimide tape is put around cylinder A on top of which a thermocouple is placed. Another layer of polyimide tape is put around the cylinder such that it covers the thermocouple. A flexible silicone rubber heater is wrapped around side A. The valve is opened. Simultaneously, side A is gradually heated with the heater from room-temperature to 80 o C over 2 hours at a heating rate of 0.5 o C/min in order to simulate the heat received from a heat source. The temperatures on both sides of the cylinders are recorded. The source temperature is measured in a manner similar to the previous case. All the H 2 from side A is removed after which the valve is closed. The heat pump is allowed to come back to room-temperature and the cylinder is heated at a temperature ramp of 0.5 o C/min for 2 hours. The temperature is recorded in order to give the source temperature. 2.4 Metal Hydride Heat Pipe Experiments An experiment was performed to determine if a heat-pipe like behavior can be observed by using a single hydride in the system; in this case LaNi 4.78 Sn 0.22 was loaded in one cylinder of 2.5 cc volume. The amount of alloy in the cylinder is 7.23 g. When the cylinder with the charged hydride is heated, the furnace is in raised position as shown in

55 37 Figure 2.6. An over pressure of 2.4 bar is introduced (the valve shown in the Figure 2.6 is open at all times mimicking a heat pipe. The furnace temperature was set to 100 o C via Watlow PID controller in up position, when the heating cycle began which resulted in release of hydrogen from the hydride thus increasing the pressure. After one hour the furnace was lowered, and pressure began to decrease due to phase transformations (reabsorption of hydrogen in the alloy). This behavior of the hydride is similar to those of a water-steam type heat pipe. Furnace ON Furnace OFF Cylinder containing LaNi 4.78 Sn Figure 2.6 Experiment to illustrate "Heat Pipe" type phenomenon using solid-gas hydride reactions in LaNi 4.78 Sn Flow control of Hydrogen through Polymer Membrane Hydrogen permeation experiments through a proton exchange membrane (PEM) were performed using a simple fuel cell set up and applied a potential/current to the anode/cathode of the membrane. In this case experiments were performed independent

56 38 of the hydride heat pump in order ascertain the permeation behavior of the membrane at different temperatures (22 o C to 50 o C), voltages (Potentiostatic) and currents (Galvanostatic). A schematic of H 2 flow through a PEM is shown in Figure 2.7. Figure 2.7 Schematic of hydrogen flow through a PEM. In the first set of these experiments, potentiostatic experiments were performed in which hydrogen permeation characteristics were investigated by using a Nafion membrane (in fuel cell assembly) at room-temperature and constant voltages of 20 mv, 40 mv, 60 mv, 80 mv, and 100 mv. More experiments were performed at 30, 35, 40, and 45 o C and at different voltages ranging from 20 to 1000mV. A finite amount of moisture was added in all the experiments performed. However, to observe if humidification had an impact on the permeation, one experiment was performed without humidifying the membrane.

57 39 In the next set of experiments (Galvanostatic), the current was kept constant and the voltage varied using a Nafion membrane in a fuel cell assembly at room-temperature and constant currents of 10 ma, 20 ma, 30 ma, 40 ma, and 50 ma. Additional experiments were performed at 40, 45, and 50 o C and at different currents ranging from 10 to 50 ma. 2.6 Metal Hydride Heat Pump (MHHP) operated with a PEM As mentioned in the introduction, the mechanical valve in a MHHP can be replaced by a Proton Exchange Membrane (PEM). The schematic of the experimental apparatus is shown in Figure 2.8. It can be observed from the figure that the fuel cell bypasses the valve V 3 in such a way that hydrogen gas when released from hydride B passes through the fuel cell to reach hydride A instead of V 3. In addition, a moisture trap in connected to the cylinder B to prevent any moisture from reaching the hydrides. With the goal of device miniaturization, the mass of alloy in each cylinder was drastically reduced such that each reactor contains 7.65 g of alloy. Reactor B contains the high-pressure hydride MmNi 4.15 Fe 0.85 and reactor A contains the low-pressure hydride LaNi 4.78 Sn The successful operation of the device takes place in the following steps: Hydride B is fully charged with hydrogen from the H 2 tank at a pressure of 395 psi. Hydride A is devoid of any hydrogen. Valve V 2 is closed. Excess hydrogen in the system is removed through the mechanical pump.

58 40 A limited amount of moisture is added to the membrane by opening the valve V 8 after which it is closed. Valves V 6 and V 7 are opened and an open circuit (OC) voltage is applied across the membrane. Figure 2.8 Schematic of MHHP operation with a PEM. Then a potential of 100 mv is applied across the PEM simultaneously opening the valve V 2. This applied potential is maintained for 4 h. During this time an endothermic reaction takes place in B whereas an exothermic reaction takes place in A.

59 41 The polarity of the applied voltage is reversed such that the applied voltage is now -100 mv. This reversed voltage is maintained for 4 h. During this latter period of the experiment, an endothermic reaction takes place in A and an exothermic reaction takes place in B. Different values of pressure, temperature, and current are recorded against time. A photograph of the experimental set-up is shown in Figure 2.9. Figure 2.9 Experimental set-up of the MHHP with the membrane assembly

60 42 Chapter 3 RESULTS AND DISCUSSIONS 3.1 Introduction The results obtained during the course of the development of a metal hydride heat pump (MHHP) suitable for small-scale cooling/heating have been presented in the following manner: First, selection of materials to be used for the metal hydride heat pump will be presented. These include the room-temperature Pressure-Composition (P-C) isotherms of LaNi 4.6 Sn 0.4 and LaNi 4.78 Sn After evaluation of classical as well as complex hydrides the following metal hydride pairs were selected for the optimal functioning of the cooling system in a heat pump set-up: (1) MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 (ratio of plateau pressures = 150), (2) MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 (ratio of plateau pressures = 30), and (3) LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 (ratio of plateau pressures = 5). The individual plateau pressures of MmNi 4.15 Fe 0.85, LaNi 4.78 Sn 0.22, and LaNi 4.6 Sn 0.4 are 15 bar, 0.5 bar, and 0.1 bar respectively. A comparison of the van't Hoff plots of the three hydrides has been presented in Figure 3.1 (the figure is not to scale). The heat pump results have been shown for the reverse mode of operation next. Here H 2 transfer takes place from the low-pressure hydride to the high-pressure

61 43 hydride. The performance of this 'reversed heat pump' is not good because the ratio of the plateau pressures of the participating hydrides is less than 1. The results of the 'static tests' using the pair of hydrides with differing ratio of plateau pressures in the manner described above have been presented next. The results show the effect of the ratio of plateau pressures on the amount of cooling obtained. Then the results of the 'dynamic tests' on the heat pump working with MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 pair have been presented where both the high pressure and low pressure sides were heated up to a temperature of ~80 o C simulating the high phase activity of a hot source over a period of 2 hours. LaNi 4.78 Sn 0.22 Heat Pump 3 15 bar MmNi 4.15 Fe 0.85 ln P 0.5 bar Heat Pump 2 LaNi 4.6 Sn 0.4 Room-temperature 0.1 bar 1/T Heat Pump 1 Figure 3.1 van't Hoff plots of MmNi 4.15 Sn 0.85, LaNi 4.78 Sn 0.22, and LaNi 4.6 Sn 0.4 (Figure not to scale)

62 44 We have also developed heat pipes using single hydride systems (of LaNi 4.78 Sn 0.22 and MmNi 4.15 Fe 0.85 ) with metal hydride in one cylinder and an empty cylinder for hydrogen capture on the other side. In section 3.6, solid gas and the subsequent gas solid phase transformation have been shown with LaNi 4.78 Sn The cooling (drop in temperature) obtained with solid gas phase transformation of MmNi 4.15 Fe 0.85 has been shown in section The results from Hydrogen flow control by using a proton exchange membrane (PEM) that allowed electronic control of flow of hydrogen in forward and reverse directions will be presented. These experiments were performed under the action of applied electromotive forces in two modes: Potentiostatic and Galvanostatic. The potentiostatic experiments were performed where the applied voltage was varied in the range of 20 mv to 1000 mv. For the galvanostatic experiments the applied value of current was varied in the range of 10 ma to 50 ma. It was found that within the range of applied voltages and currents, hydrogen permeation was better in the case of Potentiostatic experiments rather than Galvanostatic electromotive forces. Next, in order to observe the performance of a metal hydride heat pump (MHHP) under electronic control, experiments were performed where the mechanical valve separating the two hydride reactors was bypassed with an electronically controlled proton exchange membrane (PEM). The drop in temperature on the high pressure side was recorded and the heat pump was recharged by reversal of polarity of the

63 45 applied voltage thus obviating the need to heat the low pressure hydride for system recharge. Mitigation of parasitic losses occurring in the system was then attempted. Here, the cooling obtained was studied as a function of the excess H 2 moles present in the system. The amount of excess H 2 in the heat pump was reduced from ~2.25 moles to ~0 moles and it was observed that the lowest temperatures obtained on the high pressure side reduced significantly upon H 2 mole reduction. The heat pump results with varying excess H 2 moles have been shown next for the reverse mode of operation. In addition, results of a heat pump using the same metal hydride, MmN i4.15 F e0.85, on both sides (ratio of plateau pressures = 1) are shown for reference. Next, the performance of heat pumps working with differing ratio of plateau pressures (0.03 to 150) in between the participating hydrides has been presented. One of the motives of the present work is the development of a cooling device suitable for commercial purposes. A step in this direction is to reduce the temperature on the radiator side in addition to active cooling of the hot source. Experiments on different solid-solid phase change materials (SSPCMs) were performed in an effort to determine the phase change temperature and the corresponding heat absorbed at these temperatures. The SSPCMs results are presented in the end.

64 Pressure-Composition-Temperature (PCT) Isotherms Room-temperature isotherm of LaNi 4.6 Sn 0.4 measured from Heat Pump Reactor In order to determine the equilibrium pressures at which a particular hydride absorbs or desorbs most of the hydrogen introduced into a cylinder, we need the number of moles of hydrogen atoms absorbed in the metal alloy as a function of pressure. In Chapter 1, Figure 1.14, we show the equilibrium pressure as function of H/M. from "Hydrogen isotherms for LaNi 4.6 M 0.4 alloys where M=group 4A elements by S Luo, Ted B Flanagan, R.C Bowman Jr." [61]. However, we needed to ascertain if the alloy that we have used for experiments conform to that obtained by others. In this experiment we used one of the cylinders from the heat pump and connected it to calibration volume for the hydrogen supply to the alloy. The charging data of LaNi 4.6 Sn 0.4 is shown in Figure 3.2. It was found that the hydrogen absorption saturation was achieved approximately after 11 cycles; noting that a fixed calibration volume of 40 cc at a pressure of 250 psi was used for each cycle. We obtained (1) Pressure vs. Time curves shown in Figure 3.2 (a) and (2) the corresponding Temperature vs. Time curves in Figure 3.2 (b). We used 0.06 moles of H-atoms per cycle. It can be observed from Figure 3.2 (a) that as soon as the alloy is exposed to hydrogen from the calibration volume, it starts to absorb hydrogen and as a result the pressure in the chamber decreases. At the same time the temperature of the alloy increases during H 2 absorption. Gradually the temperature reaches roomtemperature at which points the pressure in the chamber equilibrates. Fresh H 2 is introduced in each cycle and ultimately the alloy saturates for a particular pressure after

65 47 11 cycles. The equilibrium pressure (the pressure at the end of each cycle) is plotted against time to give the isotherm at room-temperature as shown in Figure 3.3. Figure 3.3 indicates a very flat plateau for LaNi 4.6 Sn 0.4. This is advantageous for heat pumping applications because it shows that the hydride can absorb much hydrogen at a given pressure. More importantly, the plateau occurs at a low pressure of 0.1 bar which makes this hydride suitable in applications where high pressures are undesirable. It should be noted here that the plateau pressure is the average of the pressures of the curve of the "α + β" phase (the phase at which most of the hydrogen absorption occurs) of the alloy shown in Figure The pressure-composition (P-C) isotherm of LaNi 4.6 Sn 0.4 at 23 o C in terms of hydrogen atom to metal ratio is shown in Figure 3.4. It can be noted from the figure that LaNi 4.6 Sn 0.4 the maximum H/M ratio is 6.16.

66 48 14 Pressure vs Time Charging LaNi 4.6 Sn Pressure (bar) Cycle 1 Cycle 2 Cycle 3 Cycle 4 Cycle 5 Cycle 6 Cycle 7 Cycle 8 Cycle 9 Cycle 10 Cycle 11 2 Temperature ( o C) Time (s) Temperature vs Time Charging LaNi 4.6 Sn (a) Cycle 1 Cycle 2 Cycle 3 Cycle 4 Cycle 5 Cycle 6 Cycle 7 Cycle 8 Cycle 9 Cycle 10 Cycle 11 (b) Time (s) Figure 3.2 Hydrogen charging data of LaNi 4.6 Sn 0.4. (a) Pressure vs. Time (b) Temperature vs. Time. Charging was performed for 11 cycles from a fixed calibration volume of 40 cc at a pressure of 250 psi for each cycle.

67 Room-Temperature isotherm for LaNi 4.6 Sn P (bar) bar Time (s) Figure 3.3 Room-temperature isotherm of LaNi 4.6 Sn P-C-T isotherm of LaNi 4.6 Sn 0.4 at 23 o C P (bar) H/M Figure 3.4 P-C isotherm of LaNi 4.6 Sn 0.4 at 23 o C.

68 Room-temperature isotherm of LaNi 4.78 Sn 0.22 In a manner similar to LaNi 4.6 Sn 0.4, the isotherm for LaNi 4.78 Sn 0.22 was determined. The charging data for LaNi 4.78 Sn 0.22 is shown in Figure 3.5. Charging was performed for 14 cycles from a fixed calibration volume of 40 cc at a pressure of 250 psi for each cycle. Again 0.06 moles of H-atoms were introduced into the alloy during each cycle. The Pressure vs. Time curves are shown in Figure 3.5 (a) whereas the corresponding Temperature vs. Time curves are shown in 3.5 (b). As in the case of LaNi 4.6 Sn 0.4 the equilibrium pressure (the pressure at the end of each cycle) is plotted against time to give the room-temperature isotherm of LaNi 4.78 Sn 0.22 shown in Figure 3.6. It can be observed from the plot that the plateau does not show any slope and occurs at a pressure of 0.5 bar. The pressure-composition (P-C) isotherm of LaNi 4.78 Sn 0.22 at 25 o C in terms of hydrogen atom to metal ratio is shown in Figure 3.7. It can be noted from the figure that LaNi 4.6 Sn 0.4 the maximum H/M ratio is 6.65.

69 51 16 Charging Pressure vs Time LaNi 4.78 Sn Pressure (bar) Temperature ( o C) Cycle 1 Cycle 2 Cycle 3 Cycle 4 Cycle 5 Cycle 6 Cycle 7 Cycle 8 Cycle 9 Cycle 10 Cycle 11 Cycle 12 Cycle 13 Cycle Time (s) Temperature vs Time 11 (a) Cycle 1 Cycle 2 Cycle 3 Cycle 4 Cycle 5 Cycle 6 Cycle 7 Cycle 8 Cycle 9 Cycle 10 Cycle 11 Cycle 12 Cycle 13 Cycle 14 Charging LaNi 4.78 Sn (b) Time (s) Figure 3.5 Hydrogen charging data of LaNi 4.78 Sn (a) Pressure vs. Time (b) Temperature vs. Time. Charging was performed for 14 cycles from a fixed calibration volume of 40 cc at a pressure of 250 psi for each cycle.

70 Room-Temperature isotherm for LaNi 4.78 Sn P (bar) bar Time (s) Figure 3.6 Room-temperature isotherm of LaNi 4.78 Sn Development of Metal Hydride Pair Materials The metal hydrides used in heat pumping applications should possess the following characteristics: The metal hydrides should have high volumetric and gravimetric energy densities. They should be meta-stable. They should have a long cycle-life. The hydrogen absorption and desorption should be reversible.

71 53 The two metal hydrides to be used in the heat pump should have sufficient difference in their respective plateau pressures to provide the necessary 'thermodynamic drive' to obtain the required temperature drop. 10 P-C-T isotherm of LaNi 4.78 Sn 0.22 at 25 o C P (bar) H/M Figure 3.7 P-C isotherm of LaNi 4.78 Sn 0.22 at 25 o C. Certain Mg- and Li- based complex hydrides have very high volumetric and gravimetric energy densities compared with the classical intermetallic hydrides. But they are not meta-stable such that they could be used for heat pumping applications. Also, they do not show phase transitions and reversibility around room-temperature and their reaction

72 54 kinetics is slow. Keeping in mind, these limitations of the complex metal hydrides, LaNi 4.78 Sn 0.22, LaNi 4.6 Sn 0.4, and MmNi 4.15 Fe 0.85 were chosen as working hydrides for the heat pump in spite of their lower gravimetric and volumetric energy densities. Among the AB 5 intermetallic alloys, LaNi 4.78 Sn 0.22, LaNi 4.6 Sn 0.4, and MmNi 4.15 Fe 0.85 were chosen because of the long cycle life of the alloys. Intensive work had been carried out by researchers such as Dhanesh Chandra, Frank Lynch, and Bob Bowman Jr. [16] on perfecting these alloys for a more widespread use during the initial stages of the intermetallics development. In each of the following six experiments presented in this section, the weight of the hydrides on either side was g and a "flow-aid" (TFE powder) of 4.75 g was added to each hydride in order to prevent the walls of the reactor from cracking. This is because without any "flow-aid" these intermetallic hydrides expand to about 20% of their initial volume (dehydrided-to-hydrided state) Heat Pump working with MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 pair (P 1 /P 2 = 150) The experiment was performed with the MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 pair where MmNi 4.15 Fe 0.85 acts as the high pressure hydride and LaNi 4.6 Sn 0.4 as the low pressure hydride. The ratio of the plateau pressures of these two hydrides is 150. MmNi 4.15 Fe 0.85 (B) was fully charged directly from the H 2 tank at a pressure of 18 bar and LaNi 4.6 Sn 0.4 (A) was devoid of any H 2 at the beginning of the experiment. The pressure on the cold side was ~18 bar and that on the hot side was ~0 bar at the beginning of the discharge experiment as seen from Figure 3.8 (a).

73 55 The starting temperature was 26 o C. Both sides were insulated for preventing heat loss to the environment. As soon as the valve is opened, H 2 flows from B to A; MmNi 4.15 Fe 0.85 starts to desorb H 2 whereas LaNi 4.6 Sn 0.4 starts to absorb H 2. The pressure in both the reactors equalizes as soon as the valve is opened. The experiment was performed for ~18 hours. Since MmNi 4.15 Fe 0.85 desorbs H 2 and LaNi 4.6 Sn 0.4 absorbs H 2, there is an endothermic reaction in B and the temperature drops from 26 o C to -12 o C as soon as the valve is opened and the temperature on the LaNi 4.6 Sn 0.4 side rises from 26 o C to 100 o C as a result of the exothermic reaction in A. This is evidenced from the Temperature vs. Time plot of Figure 3.8 (b). Heat is absorbed into B (cold) and rejected to A (hot) for about 4 hours after which the temperatures on both sides start to approach room-temperature. The maximum drop in temperature is 38 o C. Further, it can be observed from the Pressure vs. Time plot of Figure 3.8 (a) that the H 2 pressure decreases linearly from 2.5 bar to 0.5 bar in 2.25 hours. Thereafter the device performs at sub-atmospheric pressure.

74 56 Pressure (bar) Pressure vs Time MmNi 4.15 Fe 0.85 (HP-LT) B (High Pressure, Low-Temperature) LaNi 4.6 Sn 0.4 MmNi4.15Sn0.85 LaNi4.6Sn0.4 MmNi 4.15 Fe LaNi 4.6 Sn 0.4 (LP-HT) A (Low Pressure, High-Temperature) Discharge (a) Time (s) Temperature vs Time MmNi4.15Sn0.85 LaNi4.6Sn0.4 LaNi 4.6 Sn 0.4 (LP-HT) A (Low Pressure, High-Temperature) Temperature ( o C) Min. Temp. ~ -12 o C Max. Temp. ~ 100 o C 0-20 MmNi 4.15 Fe 0.85 (HP-LT) Discharge B (High Pressure, Low-Temperature) (b) Time (s) Figure 3.8 Discharge curves for the MHHP working with MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 pair. (a) Pressure vs. Time: There is an instantaneous decrease in pressure when the valve is opened. Thereafter the device performs at gradually decreasing pressure from 2.5 bar to 0.5 bar for about 2.25 hours. For the remainder of the time the device performs at near 0.4 bar. (b) Temperature vs. Time: Active cooling was maintained for ~4 hours. The minimum temperature on the cold side containing MmNi 4.15 Fe 0.85 side is -12 o C and the maximum drop in temperature is ~38 o C.

75 Heat Pump working with MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 pair (P 1 /P 2 = 30) 18 Pressure vs Time MmNi4.15Sn0.85 LaNi4.78Sn0.22 MmNi Fe 0.85 (HP-LT) B (High Pressure, Low-Temperature) 14 Pressure (bar) LaNi 4.78 Sn 0.22 MmNi 4.15 Fe LaNi 4.78 Sn 0.22 (LP-HT) A (Low Pressure, High-Temperature) Discharge (a) Time (s) Temperature vs Time MmNi4.15Sn0.85 LaNi4.78Sn0.22 LaNi 4.78 Sn 0.22 (LP-HT) A (Low Pressure, High-Temperature) Temperature ( o C) Min. Temp. ~ -3 o C Max. Temp. ~ 77 o C 0 MmNi 4.15 Fe 0.85 (HP-LT) B (High Pressure, Low-Temperature) Discharge (b) Time (s) Figure 3.9 Discharge curves for the MHHP working with MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 pair. (a) Pressure vs. Time: There is an instantaneous decrease in pressure when the valve is opened. Thereafter the device performs at gradually decreasing pressure from 3.5 bar to 1 bar for about 4 hours. For the remainder of the time the device performs at near 1 bar. (b) Temperature vs. Time: Active cooling was maintained for ~5 hours. The minimum temperature on the cold side containing MmNi 4.15 Fe 0.85 side is -3 o C and the maximum drop in temperature is ~30 o C.

76 58 The experiment was performed with the MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 pair where MmNi 4.15 Fe 0.85 acts as the high pressure hydride and LaNi 4.78 Sn 0.22 as the low pressure hydride. The ratio of the plateau pressures of these two hydrides is 30. MmNi 4.15 Fe 0.85 (B) was fully charged directly from the H 2 tank at a pressure of 18 bar and LaNi 4.78 Sn 0.22 (A) was devoid of any H 2 at the beginning of the experiment. The pressure on the cold side was ~18 bar and that on the hot side was ~0 bar at the beginning of the discharge experiment as seen from Figure 3.9 (a). The starting temperature was 27 o C. Both sides were insulated for preventing heat loss to the environment. As soon as the valve is opened, H 2 flows from B to A; MmNi 4.15 Fe 0.85 starts to desorb H 2 whereas LaNi 4.78 Sn 0.22 starts to absorb H 2. The pressure in both the reactors equalizes as soon as the valve is opened. The experiment was performed for ~18 hours. Since MmNi 4.15 Fe 0.85 desorbs H 2 and LaNi 4.78 Sn 0.22 absorbs H 2, there is an endothermic reaction in B and the temperature drops from 27 o C to -3 o C as soon as the valve is opened and the temperature on the LaNi 4.78 Sn 0.22 side rises from 27 o C to 77 o C as a result of the exothermic reaction in A. This is evidenced from the Temperature vs. Time plot of Figure 3.9 (b). Heat is absorbed into B (cold) and rejected to A (hot) for about 5 hours after which the temperatures on both sides start to approach room-temperature. The maximum drop in temperature is 30 o C. Further, it can be observed from the Pressure vs. Time plot of Figure 3.9 (a) that the H 2 pressure decreases linearly from 3.5 bar to 1 bar in 4 hours. Thereafter the device performs at sub-atmospheric pressure.

77 Heat Pump working with LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 pair (P 1 /P 2 = 5) Pressure vs Time LaNi 4.78 Sn 0.22 (HP-LT) B (High Pressure, Low-Temperature) LaNi4.78Sn0.22 LaNi4.6Sn Pressure (bar) LaNi 4.6 Sn 0.4 LaNi 4.6 Sn 0.4 (LP-HT) Discharge A (Low Pressure, High-Temperature) (a) Time (s) Temperature vs Time LaNi 4.6 Sn 0.4 (LP-HT) A (Low Pressure, High-Temperature) H 2 LaNi 4.78 Sn 0.22 LaNi4.78Sn0.22 LaNi4.6Sn0.4 Temperature ( o C) LaNi 4.78 Sn 0.22 (HP-LT) B (High Pressure, Low-Temperature) Time (s) Min. Temp. ~ 7 o C Max. Temp. ~ 45 o C Discharge (b) Figure 3.10 Discharge curves for the MHHP working with LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 pair. (a) Pressure vs. Time: The devices performs at sub-atmospheric pressure for the entire cooling period. (b) Temperature vs. Time: Active cooling was maintained for ~6. The minimum temperature on the cold side containing LaNi 4.78 Sn 0.22 side is 7 o C and the maximum drop in temperature is ~13 o C.

78 60 The experiment was performed with the LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 pair where LaNi 4.78 Sn 0.22 acts as the high pressure hydride and LaNi 4.6 Sn 0.4 as the low pressure hydride. The ratio of the plateau pressures of these two hydrides is 5. The charging data for LaNi 4.78 Sn 0.22 has already been shown in Figure 3.5. Charging was performed for 14 cycles from a fixed calibration volume of 40 cc at a pressure of 250 psi for each cycle. LaNi 4.78 Sn 0.22 (B) was completely charged with H 2 and LaNi 4.6 Sn 0.4 (A) devoid of any H 2 at the beginning of the discharge experiment. The pressure on the cold side was ~16.5 bar and that on the hot side was ~0 bar at the beginning of the discharge experiment as seen from Figure 3.10 (a). The starting temperature is 20 o C. Both sides were insulated for preventing heat loss to the environment. As soon as the valve is opened, H 2 flows from B to A; LaNi 4.78 Sn 0.22 starts to desorb H 2 whereas LaNi 4.6 Sn 0.4 starts to absorb H 2. The pressure in both the reactors equalizes over time. The experiment was performed for ~18 hours. Since LaNi 4.78 Sn 0.22 desorbs H 2 and LaNi 4.6 Sn 0.4 absorb H 2, there is an endothermic reaction in B and the temperature drops from 20 o C to 7 o C almost as soon as the valve is opened and the temperature on the LaNi 4.6 Sn 0.4 side rises from 20 o C to 45 o C as a result of the exothermic reaction in A as shown in the Temperature vs. Time plot of Figure 3.10 (b). Heat is absorbed into B (cold) and rejected to A (hot) for about 6 hours after which the temperatures on both sides start to approach room-temperature. The maximum drop in temperature is 13 o C. Further, it can be observed from the Pressure vs. Time plot of Figure 3.10 (a) that the H 2 pressure stayed well below 1 bar throughout the active cooling period and this fact is very

79 61 encouraging that a hot device could be cooled even if the Metal Hydride Cooling System (MHCS) performs under sub-atmospheric pressures Heat Pump working with LaNi 4.6 Sn 0.4 /LaNi 4.78 Sn 0.22 (P 1 /P 2 = 0.2) The experiment was performed with the LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 pair where LaNi 4.78 Sn 0.22 acts as the low pressure hydride and LaNi 4.6 Sn 0.4 as the high pressure hydride. The ratio of the plateau pressures of these two hydrides is 0.2. LaNi 4.6 Sn 0.4 (A) was completely charged with H 2 and LaNi 4.78 Sn 0.22 (B) devoid of any H 2 at the beginning of the discharge experiment. The starting pressure was ~17.75 bar at the beginning of the discharge experiment as seen from Figure 3.10 (a). The starting temperature is 16.7 o C. Both sides were insulated for preventing heat loss to the environment. As soon as the valve is opened, H 2 flows from A to B; LaNi 4.6 Sn 0.4 starts to desorb H 2 whereas LaNi 4.78 Sn 0.22 starts to absorb H 2. The pressure in both the reactors equalizes over time. The experiment was performed for ~18 hours. Since LaNi 4.6 Sn 0.4 desorbs H 2 and LaNi 4.6 Sn 0.4 absorb H 2, there is an endothermic reaction in A and the temperature drops from 16.7 o C to 5 o C almost as soon as the valve is opened and the temperature on the LaNi 4.6 Sn 0.4 side rises from 16.7 o C to 36 o C as a result of the exothermic reaction in B as seen from the Temperature vs. Time plot of Figure 3.11 (b). Heat is absorbed into A and rejected to B (hot) for about 5.5 hours after which the temperatures on both sides start to approach room-temperature. The maximum drop in temperature is 11.5 o C.

80 Figure 3.11 Discharge curves for the MHHP working with LaNi 4.6 Sn 0.4 /LaNi 4.78 Sn 0.22 pair. (a) Pressure vs. Time: The devices performs at sub-atmospheric pressure for the entire cooling period. (b) Temperature vs. Time: Active cooling was maintained for ~5.5 h. The minimum temperature on the LaNi 4.6 Sn 0.4 side is 5 o C and the maximum drop in temperature is ~11.5 o C. 62

81 63 Further, it can be observed from the Pressure vs. Time plot of Figure 3.11 (a) that the H 2 pressure stayed well below 1 bar throughout the active cooling period and this fact is very encouraging because it means that a hot device could be cooled even if the Metal Hydride Cooling System (MHCS) performs under sub-atmospheric pressures Heat Pump working with LaNi 4.78 Sn 0.22 /MmNi 4.15 Fe 0.85 (P 1 /P 2 = 0.03) The experiment was performed with the LaNi 4.78 Sn 0.22 /MmNi 4.15 Fe 0.85 pair where MmNi 4.15 Fe 0.85 acts as the low pressure hydride and LaNi 4.78 Sn 0.22 as the high pressure hydride. The ratio of the plateau pressures of these two hydrides is LaNi 4.78 Sn 0.22 (A) was fully charged directly from the H 2 tank at a pressure of 17 bar and MmNi 4.15 Fe 0.85 (B) was devoid of any H 2 at the beginning of the experiment. The starting temperature was 20 o C. Both sides were insulated for preventing heat loss to the environment. As soon as the valve is opened, H 2 flows from A to B; LaNi 4.78 Sn 0.22 starts to desorb H 2 whereas MmNi 4.15 Fe 0.85 starts to absorb H 2. The pressure in both the reactors equalizes as soon as the valve is opened. The experiment was performed for ~18 hours. Since LaNi 4.78 Sn 0.22 desorbs H 2 and MmNi 4.15 Fe 0.85 absorbs H 2, there is an endothermic reaction in A and the temperature drops from 20 o C to 14 o C as soon as the valve is opened and the temperature on the MmNi 4.15 Fe 0.85 side rises from 20 o C to 29 o C as a result of the exothermic reaction in B. as shown in the Temperature vs. Time plot of Figure 3.12 (b). Heat is absorbed into A and rejected to B for about 2.7 hours after which the temperatures on both sides start to approach room-temperature. The maximum drop in temperature is 5.5 o C. Further, it can be observed from the Pressure vs. Time plot of Figure 3.12 (a) that the H 2 pressure is maintained for almost the entire experiment.

82 Figure 3.12 Discharge curves for the MHHP working with LaNi 4.78 Sn 0.22 /MmNi 4.15 Fe 0.85 pair. (a) Pressure vs. Time: There is an instantaneous decrease in pressure when the valve is opened. Thereafter the device performs at atmospheric pressure. (b) Temperature vs. Time: Active cooling was maintained for ~2.7 hours. The minimum temperature on the side containing LaNi 4.78 Sn 0.22 is 14 o C and the maximum drop in temperature is ~5.5C. 64

83 Heat Pump working with LaNi 4.6 Sn 0.4 /MmNi 4.15 Fe 0.85 pair (P 1 /P 2 = 0.006) The experiment was performed with the LaNi 4.6 Sn 0.4 /MmNi 4.15 Fe 0.85 pair where MmNi 4.15 Fe 0.85 acts as the low pressure hydride and LaNi 4.6 Sn 0.4 as the high pressure hydride. The ratio of the plateau pressures of these two hydrides is LaNi 4.6 Sn 0.4 (A) was fully charged directly from the H 2 tank at a pressure of 18 bar and MmNi 4.15 Fe 0.85 (A) was devoid of any H 2 at the beginning of the experiment. The starting temperature was 21.7 o C. Both sides were insulated for preventing heat loss to the environment. As soon as the valve is opened, H 2 flows from A to B; LaNi 4.6 Sn 0.4 starts to desorb H 2 whereas MmNi 4.15 Fe 0.85 starts to absorb H 2. The pressure in both the reactors equalizes as soon as the valve is opened. The experiment was performed for ~18 hours. Since LaNi 4.6 Sn 0.4 desorbs H 2 and MmNi 4.15 Fe 0.85 absorbs H 2, there is an endothermic reaction in A and the temperature drops from 21.7 o C to 15.7 o C as soon as the valve is opened and the temperature on the MmNi 4.15 Fe 0.85 side rises from 21.7 o C to 31 o C as a result of the exothermic reaction in A as seen from the Temperature vs. Time plot of Figure 3.13(b). Heat is absorbed into A and rejected to B (hot) for about 2.7 hours after which the temperatures on both sides start to approach room-temperature. The maximum drop in temperature is 6 o C. The device performs at a pressure of ~3 bar for the maximum length of time.

84 Figure 3.13 Discharge curves for the MHHP working with LaNi 4.6 Sn 0.4 /MmNi 4.15 Fe 0.85 pair. (a) Pressure vs. Time: There is an instantaneous decrease in pressure when the valve is opened. Thereafter the device performs at a pressure of 3 bar. (b) Temperature vs. Time: Active cooling was maintained for ~2.7 hours. The minimum temperature on the side containing LaNi 4.6 Sn 0.4 side is 15.7 o C and the maximum drop in temperature is ~6 o C. 66

85 67 Some interesting conclusions can be drawn by comparison of the heat pump experiments done with higher pressure ratios with the lower pressure ratio heat pumps. The temperature drop attained with higher pressure ratio heat pumps is more compared to the lower pressure ratio heat pumps because of the larger thermodynamic drive. The duration of cooling increases with decreasing pressure ratio for P 1 /P 2 > 1 (suggesting faster kinetics), whereas for P 1 /P 2 < 1 the duration of cooling decreases with decreasing pressure ratios (suggesting slower kinetics). More interestingly, the optimum pressure ratio for obtaining any significant cooling is 0.2. The results obtained with the six hydride pairs have been summarized in Table 3.1 Table 3.1 Summary of results obtained with the six hydride pairs Hydride Pair P 1 /P 2 Min. T Max. T Drop Cooling Duration Max. T MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn o C 38 o C 4 hours 100 o C MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn o C 30 o C 5 hours 77 o C LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn o C 13 o C 6 hours 45 o C LaNi 4.6 Sn 0.4 /LaNi 4.78 Sn o C 11.5 o C 5.5 hours 36 o C LaNi 4.78 Sn 0.22 /MmNi 4.15 Fe o C 5.5 o C 2.7 hours 29 o C LaNi 4.6 Sn 0.4 /MmNi 4.15 Fe o C 6 o C 2.7 hours 31 o C The minimum temperatures obtained on the cold side as a function of the pressure ratio have been plotted in Figure 3.14.

86 68 20 Minimum Temperature as a function of pressure ratio T (min) ( o C) P1/P2 Figure 3.14 Minimum temperature as a function of the pressure ratio The minimum temperatures (in o C) are related with the pressure ratios by the following equation: P1 T min. 2.45ln( ) (3.1) P Metal Hydride Cooling System (MHCS) performance under dynamic heating In order to determine the efficacy of a heat pump for cooling applications, experiments under dynamic heating conditions were performed when the MHHP is in a fully charged condition. Side B contains the high-pressure, low-temperature (HP-LT) hydride MmNi 4.15 Fe It consists of 111 g of the alloy in addition to 12 g of TFE powder which acts as the "flow aid" for the metal hydride. Side A contains the low-pressure, high-

87 69 temperature (LP-HT) hydride LaNi 4.78 Sn It consists of 140 g of the alloy without any TFE powder. A standard metal hydride heat pump was charged (i.e. the metal hydride B was full charged with hydrogen and isolated, by a valve, from the hydride A which was fully discharged, and was in an alloy form; schematic of the device is shown in the inset of Figure 3.15 (a). Once the valve is opened the temperature of hydride B decreased down to ~0 o C, and the temperature of the hydride A instantly increased to ~64 o C as can be seen from Figure 3.15 (a). The changes in temperature of A and B are due endothermic desorption of hydrogen in B, and exothermic charging of Alloy in A to form a hydride which is a well-established fact. The heat pump was left in open with no insulation, what may be referred to as static condition. Our goal is to study the operation of the heat pump for small scale cooling or heating, so we are attempting to demonstrate that using the same device shown in Figure 3.15 (a). Similar experiment to the above was performed, except in this case we heated the hydride in B, with a strip heater, acting as a virtual hot device with dynamic heating, and observed the temperature observed temperature excursions, as shown in Figure 3.15 (b) in black color plot (B). As the hydrogen was being released from B it was concurrently absorbed in A and resulted a graph shown in blue color (A) in Figure 3.15 (b). Before conducting this dynamic experiment, a dry run with the same alloy (in B) without any hydrogen was performed.

88 70 The temperature on the cylinder B was recorded and is shown in Figure 3.15 (b) in red colored symbols, and is called reference temperature. It is clear that with system under ambient surrounding (and no external heating), the B side temperature is ~2.5 o C as shown in Figure 3.15 (a), black curve. Whereas in the second case with dynamic heating, the temperature after 30 minutes was 37 o C in B, as compared the reference temperature plot in red symbols at ~70.5 o C. Figure 3.16 shows a T vs. time plotted to show to show difference between constant reference temperature (~25 o C) as compared to dynamic heating (red symbols). The T up to 50 minutes is greater in the case of dynamic heating. This approach affords choices for configuration of hydrides for cooling devices. In another experiment, Hydride A was heated and the temperature excursions were observed. Figure 3.17 shows dynamic heating of hydride A. The temperature profiles of A and B are shown in blue and black, respectively. The reference temperature plot is shown in red symbols. The cooling effect was not as good as in the case shown in dynamic heating of hydride B in Figure 3.15 (b). It is interesting to note that the radiator temperature of hydride B is very low in this case, ~21 o C.

89 71 A (a) Starting T=25 o C (No external heating) B B A Source Source (Reference Temperature - No hydrogen in the cylinder) (b) Temperature ( o C) MmNi 4.15 Fe 0.85 (B - HP, Cold Side) Source CHIP 20 LaNi 4.78 Sn 0.22 (A - LP, Radiator Side) 10 Starting T=18 o C with dynamic heating Temperature vs Time Time (s) Figure 3.15 (a).heat pump temperature profiles of hydride in A and B with surroundings at near 25 o C. It can be seen that the temperature dropped down to ~2.5 o C in 30 minutes in B hydride. In another experiment (b), heat pump with dynamic heating of B hydride shows temperatures excursions of A(blue) and B(black). An additional reference temperature plot (red colored symbols) that shows a steady temperature increase up to ~80 o C in ~2 hours; this was just an alloy with no hydrogen interactions. It can be seen that the hydride brings down the temperature of the source significantly as compared to the reference plot. The vessel with hydride A is considered to be a radiator.

90 72 0 Comparison of Cooling rates 5 10 del T ( o C) Reference temp. Constant at ~25 o C Reference temperature changing Time (min.) Ref. T= 25 o C Ref. T=18 to 80 o C Figure 3.16 Comparison of differential cooling rates with hydride at a reference temperature of 25 o C and under dynamic heating. 3.5 Dynamic Tests on Heat Pumps with varying pressure ratios In a manner discussed in section 3.4, dynamic tests were performed on heat pumps working with (1) MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 (ratio of plateau pressures = 150), (2) MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 (ratio of plateau pressures = 30), (3) LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 (ratio of plateau pressures = 5), and (4) LaNi 4.6 Sn 0.4 /LaNi 4.78 Sn 0.22 (ratio of plateau pressures = 0.2) pairs. In each of the three experiments discussed in this section, the weight of the hydrides on either side was g with a "flow-aid" (TFE powder) of 4.75 g in each of the cylinders.

91 Source (Reference Temp.) Temperature vs Time B A Source Temperature ( o C) LaNi 4.78 Sn 0.22 (A - LP, H ot Side) CHIP Source MmNi 4.15 Fe 0.85 (B - H P, C old Side) Figure 3.17 Heat pump temperature profiles of hydride in A and B with dynamic heating of A hydride that shows temperatures excursions of A (blue) and B (black). An additional reference temperature plot (red colored symbols) that shows a steady temperature increase up to ~80 o C in ~2 hours; this was just alloy A with no hydrogen interactions. It can be seen that the cooling effect is not as good as compared to heating the hydride B. However, the radiator temperature (B) is very low, ~21 o C. The cooling and heating coefficients for each of the heat pumps were calculated in accordance with the following equations: C COOL TB 1 (3.2) T Source( B) T A CHEAT (3.3) TSource(B)

92 74 Three different source temperatures were recorded on all the three hydride cylinders devoid of hydrogen gas. The source temperatures have been shown in blue whereas the cold side and hot side temperatures have been plotted in black and red respectively. The high pressure, cold side (B) was fully charged directly from the H 2 tank at a pressure of 17.5 bar and the low pressure, hot side (A) was devoid of any H 2 at the beginning of the experiment. The pressure on the cold side was ~17.5 bar and that on the hot side was ~0 bar at the beginning of the discharge experiment. A silicone rubber heater was wrapped around the high pressure, cold side (B) which acts as a virtual hot device and the heat pump was left in open with no thermal insulation. The strip heater was heated to ~80 o C starting from room-temperature over a period of 2 hours Dynamic Test on Heat Pump working with MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 pair (P 1 /P 2 = 150) The experiment was performed with the MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 pair where MmNi 4.15 Fe 0.85 acts as the high pressure hydride and LaNi 4.6 Sn 0.4 as the low pressure hydride. The ratio of the plateau pressures of these two hydrides is 150.

93 75 Figure 3.18 Dynamic test on the MHHP working with MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 pair. (a) Pressure vs. Time: There is an instantaneous drop in pressure from 17.5 bar to 2.25 bar when the valve is opened. However, due to heating on side B, the pressure increases from 2.25 bar to ~5 bar. After about 1 hour, the pressure stabilizes at 1 (b) Temperature vs. Time: The metal hydride MmNi 4.15 Fe 0.85 brings down the temperature of the hot source from 20.5 o C to -5 o C, a reduction of 25.5 o C. Active cooling was maintained for 17 minutes. The hot side reaches maximum temperature of 93 o C. Source (MmNi 4.15 Fe 0.85 ) is the MmNi 4.15 Fe 0.85 temperature in the absence of hydrogen in the cylinder which experimentally simulates the hot source temperature.

94 76 Figure 3.18 shows the results of the dynamic test. There is an instantaneous drop in pressure from 17.5 bar to 2.25 bar as soon as the valve is opened as seen from Figure 3.18 (a). However, the pressure soon increases from 2.25 bar to ~5 bar because of the enhanced rate of release of hydrogen from the metal hydride owing to dynamic heating. The pressure stabilizes at 1 bar after a period of 1 hour. The temperature vs. time plot of Figure 3.18 (b) shows that the metal hydride MmNi 4.15 Fe 0.85 brings down the temperature of the hot source from 20 o C to -5 o C resulting in a temperature reduction of 25 o C. Active cooling is maintained for 17 minutes after which the source and cold side temperature curves merge. The hot side reaches maximum temperature of 93 o C because of a very high pressure ratio of 150. The quantification of the heat pump's performance has been done in terms of cooling and heating coefficients depicted in Figures 3.19 (a) and 3.19 (b) respectively. The cooling coefficient increases from 0 to 1.25 as soon as the valve is opened. Thereafter, it gradually decreases from 1.25 to 0.1 because of the gradually decreasing hydrogen desorption capacity of the high pressure hydride, MmNi 4.15 Fe 0.85 as shown in Figure 3.19 (a). The cooling coefficient gradually approaches zero afterwards. The heating coefficient increases from 0 to 4.3 initially on opening the valve. Thereafter, it gradually decreases to 0.5 because of the increasing saturation of the low pressure hydride (LaNi 4.6 Sn 0.4 ) with hydrogen absorption which results in reduced hydrogen absorption capacity as seen from Figure 3.19 (b).

95 77 Figure 3.19 Cooling and Heating Coefficients of a MHHP working with MmNi 4.15 Fe 0.85 /LaNi 4.6 Sn 0.4 pair. (a) Cooling Coefficient vs. Time: The cooling coefficient increases from 0 to 1.25 as soon as the valve is opened. Thereafter, it gradually decreases from 1.25 to 0.1 because of the gradually decreasing hydrogen desorption capacity of the high pressure hydride, MmNi 4.15 Fe The cooling coefficient gradually approaches zero afterwards. (b) Heating Coefficient vs. Time: The heating coefficient increases from 0 to 4.3 initially on opening the valve. Thereafter, it gradually decreases to 0.5 because of the increasing saturation of the low pressure hydride (LaNi 4.6 Sn 0.4 ) with hydrogen absorption which results in reduced hydrogen absorption capacity.

96 Dynamic Test on Heat Pump working with MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 pair (P 1 /P 2 = 30) The experiment was performed with the MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 pair where MmNi 4.15 Fe 0.85 acts as the high pressure hydride and LaNi 4.6 Sn 0.4 as the low pressure hydride. The ratio of the plateau pressures of these two hydrides is 30. There is an instantaneous drop in pressure from 17.5 bar to 3.25 bar as soon as the valve is opened as seen from Figure 3.20 (a). However, the pressure soon increases from 3.25 bar to ~7 bar because of the enhanced rate of release of hydrogen from the metal hydride owing to dynamic heating. The pressure stabilizes at 1 bar after a period of 1 hour. The temperature vs. time plot of Figure 3.20 (b) shows that the metal hydride MmNi 4.15 Fe 0.85 brings down the temperature of the hot source from 21 o C to 0 o C resulting in a temperature reduction of 21 o C. Active cooling is maintained for 20 minutes after which the source and cold side temperature curves merge. The hot side reaches maximum temperature of 79 o C because of a sufficiently high pressure ratio of 30. The cooling coefficient increases from 0 to 1 as soon as the valve is opened as shown in Figure Thereafter, it gradually decreases from 1 to 0 because of the gradually decreasing hydrogen desorption capacity of the high pressure hydride, MmNi 4.15 Fe The heating coefficient increases from 0 to 3.5 initially on opening the valve. Thereafter, it gradually decreases to 0.25 because of the increasing saturation of the low pressure hydride (LaNi 4.78 Sn 0.22 ) with hydrogen absorption which results in reduced hydrogen absorption capacity as seen from Figure 3.21 (b).

97 79 Figure 3.20 Dynamic test on the MHHP working with MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 pair. (a) Pressure vs. Time: There is an instantaneous drop in pressure from 17.5 bar to 3.25 bar when the valve is opened. However, due to heating on side B, the pressure increases from 3.25 bar to ~7 bar. After about 1 hour, the pressure stabilizes at 1 bar. (b) Temperature vs. Time: The metal hydride MmNi 4.15 Fe 0.85 brings down the temperature of the hot source from 21 o C to 0 o C, a reduction of 21 o C. Active cooling was maintained for ~20 minutes. The hot side reaches maximum temperature of 79 o C. Source (MmNi 4.15 Fe 0.85 ) is the MmNi 4.15 Fe 0.85 temperature in the absence of hydrogen in the cylinder which experimentally simulates the hot source temperature.

98 80 Figure 3.21 Cooling and Heating Coefficients of a MHHP working with MmNi 4.15 Fe 0.85 /LaNi 4.78 Sn 0.22 pair. (a) Cooling Coefficient vs. Time: The cooling coefficient increases from 0 to 1 as soon as the valve is opened. Thereafter, it gradually decreases from 1 to 0 because of the gradually decreasing hydrogen desorption capacity of the high pressure hydride, MmNi 4.15 Fe (b) Heating Coefficient vs. Time: The heating coefficient increases from 0 to 3.5 initially on opening the valve. Thereafter, it gradually decreases to 0.25 because of the increasing saturation of the low pressure hydride (LaNi 4.6 Sn 0.4 ) with hydrogen absorption which results in reduced hydrogen absorption capacity.

99 Dynamic Test on Heat Pump working with LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 pair (P 1 /P 2 = 5) The experiment was performed with the LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 pair where LaNi 4.78 Sn 0.22 acts as the high pressure hydride and LaNi 4.6 Sn 0.4 as the low pressure hydride. The ratio of the plateau pressures of these two hydrides is 5. As seen from Figure 3.22 (a), there is an instantaneous drop in pressure from 17.5 bar to 1 bar as soon as the valve is opened. However, the pressure soon increases from 1 bar to ~3 bar in about 2 hours because of the enhanced rate of release of hydrogen from the metal hydride owing to dynamic heating. The temperature vs. time plot of Figure 3.22 (b) shows that the metal hydride LaNi 4.78 Sn 0.22 brings down the temperature of the hot source from 20 o C to 10.5 o C resulting in a temperature reduction of 9.5 o C initially. Active cooling is maintained for 100 minutes (1.67 hours) after which the source and cold side temperature curves merge. Enhanced cooling is maintained for almost the entire length of the heat pump operation. For example, at about 40 minutes of the heat pump operation the source temperature is 79 o C whereas the high pressure metal hydride is at 46 o C, which means a reduction in temperature of a hot device by 33 o C. The maximum hot side temperature is 55 o C because of a lower pressure ratio of 5.

100 Figure 3.22 Dynamic test on the MHHP working with LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 pair. (a) Pressure vs. Time: There is an instantaneous drop in pressure from 17.5 bar to 1 bar when the valve is opened. However, due to heating on side B, the pressure increases from 1 bar to ~3 bar in 2 hours. (b) Temperature vs. Time: The metal hydride LaNi 4.78 Sn 0.22 brings down the temperature of the hot source from 20 o C to 10.5 o C, a reduction of 9.5 o C. At about 40 minutes of heat pump operation, the source temperature is brought down by 33 o C by LaNi 4.78 Sn 0.22 from 79 o C to 46 o C. Active cooling was maintained for 100 minutes (1.67 hours). The hot side reaches maximum temperature of 55 o C. Source (LaNi 4.78 Sn 0.22 ) is the LaNi 4.78 Sn 0.22 temperature in the absence of hydrogen in the cylinder which experimentally simulates the hot source temperature. 82

101 83 The heat pump's performance in terms of the cooling and heating coefficients has been shown in Figures 3.23 (a) and (b) respectively. The cooling coefficient increases from 0 to 0.54 as soon as the valve is opened as shown in Figure 3.23 (a). It then suddenly decreases to After ~10 minutes it starts to increase and reaches 0.45 in the next 10 minutes. The value is maintained at ~0.45 for the next 10 minutes. An almost linear drop in observed for the rest of the period of heat pump operation. The behavior of the heating coefficient is not unlike those of the previous two heat pumps. heating coefficient increases from 0 to 2.1 initially on opening the valve. Afterwards, it gradually decreases to 0.3 because of the increasing saturation of the low pressure hydride (LaNi 4.6 Sn 0.4 ) with hydrogen absorption which results in reduced hydrogen absorption capacity as can be observed from Figure 3.23 (b) Dynamic Test on Heat Pump working with LaNi 4.6 Sn 0.4 /LaNi 4.78 Sn 0.22 pair (P 1 /P 2 = 0.2) The experiment was performed with the LaNi 4.6 Sn 0.4 /LaNi 4.78 Sn 0.22 pair where LaNi 4.78 Sn 0.22 acts as the low pressure hydride and LaNi 4.6 Sn 0.4 as the high pressure hydride. The ratio of the plateau pressures of these two hydrides is 0.2.

102 84 Figure 3.23 Cooling and Heating Coefficients of a MHHP working with LaNi 4.78 Sn 0.22 /LaNi 4.6 Sn 0.4 pair. (a) Cooling Coefficient vs. Time: The cooling coefficient increases from 0 to 0.54 as soon as the valve is opened. Thereafter, it suddenly decreases to After ~10 minutes it starts to increase and reaches 0.45 in the next 10 minutes. The value is maintained at ~0.45 for the next 10 minutes. An almost linear drop in observed for the rest of the period of heat pump operation. (b) Heating Coefficient vs. Time: The heating coefficient increases from 0 to 2.1 initially on opening the valve. Afterwards, it gradually decreases to 0.3 because of the increasing saturation of the low pressure hydride (LaNi 4.6 Sn 0.4 ) with hydrogen absorption which results in reduced hydrogen absorption capacity, not unlike the previous two heat pumps.

103 Figure 3.24 Dynamic test on the MHHP working with LaNi 4.6 Sn 0.4 /LaNi 4.78 Sn 0.22 pair. (a) Pressure vs. Time: There is an instantaneous drop in pressure from 17.5 bar to ~1 bar when the valve is opened and the pressure in the cylinders is maintained at that value for the rest of the experiment. (b) Temperature vs. Time: The metal hydride MmNi 4.15 Fe 0.85 brings down the temperature of the hot source from 20 o C to 12 o C, a reduction of 8 o C. Active cooling was maintained for more than 2 hours. The hot side reaches a maximum temperature of 41 o C. Source (LaNi 4.6 Sn 0.4 ) is the LaNi 4.6 Sn 0.4 temperature in the absence of hydrogen in the cylinder which experimentally simulates the hot source temperature. 85

104 86 Figure 3.24 shows the results of the dynamic test. There is an instantaneous drop in pressure from 17.5 bar to 1 bar as soon as the valve is opened as seen from Figure 3.24 (a). The pressure in both the cylinders is maintained at that value for the rest of the experiment. The temperature vs. time plot of Figure 3.24 (b) shows that the metal hydride LaNi 4.6 Sn 0.4 brings down the temperature of the hot source from 20 o C to 12 o C resulting in a temperature reduction of 8 o C. Active cooling is maintained for more than two hours. The hot side reaches maximum temperature of 41 o C because of a low gas pressure ratio of 0.2. The quantification of the heat pump's performance has been done in terms of cooling and heating coefficients depicted in Figures 3.25 (a) and 3.25 (b) respectively. The cooling coefficient increases from 0 to 0.4 as soon as the valve is opened. A sudden drop from 0.4 to 0.15 is observed for a very short period followed by an increase in the value to 0.27 in about 12.5 minutes. Thereafter, it gradually decreases from 0.27 to 0.7 because of the gradually decreasing hydrogen desorption capacity of LaNi 4.6 Sn 0.4. The heating coefficient increases from 0 to 2.1 initially on opening the valve. Thereafter, it gradually decreases to 0.4 because of the increasing saturation of the low pressure hydride (LaNi 4.78 Sn 0.22 ) with hydrogen absorption which results in reduced hydrogen absorption capacity as seen from Figure 3.25 (b).

105 87 Figure 3.25 Cooling and Heating Coefficients of a MHHP working with LaNi 4.6 Sn 0.4 /LaNi 4.78 Sn 0.22 pair. (a) Cooling Coefficient vs. Time: The cooling coefficient increases from 0 to 0.4 as soon as the valve is opened. A sudden drop from 0.4 to 0.15 is observed for a very short period followed by an increase in the value to 0.27 in about 12.5 minutes. Thereafter, it gradually decreases from 0.27 to 0.7 because of the gradually decreasing hydrogen desorption capacity of LaNi 4.6 Sn 0.4. (b) Heating Coefficient vs. Time: The heating coefficient increases from 0 to 2.1 initially on opening the valve. Thereafter, it gradually decreases to 0.4 because of the increasing saturation of the low pressure hydride (LaNi 4.6 Sn 0.4 ) with hydrogen absorption which results in reduced hydrogen absorption capacity.

106 88 Figure 3.26 Comparison of the Cooling and Heating Coefficients of the MHHPs. A comparison of the cooling and heating coefficients of the three heat pumps is shown in Figure Figure 3.26 (a) is a plot of cooling coefficient vs. time and Figure 3.26 (b) is a plot of the heating coefficient vs. time. Both the cooling and heating coefficients decrease with decrease with decreasing equilibrium gas pressure ratios.

107 Heat Pipe Type Experiment with Metal Hydride - LaNi 4.78 Sn 0.22 Pressure (bar) H 2 Desoprtion from Hydride Heating Cycle Pressure vs Time (LaNi 4.78 Sn 0.22 ) H 2 Desorption from Hydride Automatic Reabsorption H 2 Desoprtion of H 2 to re-form from Hydride Hydride Cooling Cycle Heating Cycle Cooling Cycle Time (s) Temperature ( o C) A Temperature vs Time (LaNi 4.78 Sn 0.22 ) Temperature of the chip is represented the Furnace Temp. Furnace (ON) Forced to mainatin this temp. Heating Cycle (Hydride Temp.) B C D E F Heating Cycle Cooling Cycle Furnace On Furnace Off ~7 grams of Material Furnace (OFF). Cooling Cycle(Hydride Temp.) Time (s) G Dipiction of Slighly modifed vant' Hoff plot to Correlate with the data.in the Abvie Figure Heating D Pressure (bar) B C F E Cooling A G T( o C) Figure 3.27 Experiment to illustrate Heat Pipe type phenomenon using solid gas hydride reactions inlani 4.78 Sn (a) Effect of Heating of Hydride in a single cylinder; Pressure vs. Time plot showing the increase in pressure in the system when the furnace is turned on during the heating cycle. Here the hydrogen gas is released from the hydride thus increasing the pressure inside the reactor. As soon as the furnace is removed (Furnace OFF) during the cooling cycle, all the H 2 that had been released during the heating cycle is absorbed back by the alloy. In essence, the single hydride system works as a heat pipe. Figure (b) Temperature vs. Time (Green curve) of the Furnace temperature when it is ON, and the dark blue curve represents the Furnace temperature when it is OFF. The hydride profile is shown below, black symbols for heating cycle and red symbols for cooling cycle. (c) Pressure vs. temperature diagram depicting the heating and cooling cycle (Figure not to scale).

108 90 An experiment was performed to determine if a heat-pipe like behavior can be observed by using a single hydride in the system; in this case LaNi 4.78 Sn 0.22 was loaded in one cylinder of 2.5 cc volume. A photograph of this is sown in Figure 3.27 (bottom). When the cylinder with the charged hydride is heated, the furnace is in raised position. An over pressure of 2.4 bar was introduced (the valve shown in the Figure 3.27 (bottom) was open at all times mimicking a heat pipe. The furnace temperature was set to 100 o C via Watlow PID controller in up position, when the heating cycle began which resulted in release of hydrogen from the hydride thus increasing the pressure as indicated in the pressure vs. time plot (Figure 3.27 top left). After one hour the furnace was lowered, and pressure began to decrease due to phase transformations (reabsorption of hydrogen in the alloy), as shown in Figure 3.27 (top left. This behavior of the hydride is similar to those of a water-steam type heat pipe. Heating the cylinder also changes the hydride temperature profile. Keeping in mind the furnace temperature is set at 100 o C, (green symbols show the heating profile and blue symbols the cooling, as shown in Figure 3.27 (top right). The temperature of the hydride is much lower than the (nearly constant) furnace temperature of 100 o C. The cylinder was not in contact with the furnace inner surface. We have marked the temperature profile of the hydride with letters, A thorough G in Figure 3.27 (top right). At the point D we removed the furnace from the hydride (shown in down position in the figure). In Figure 3.27 (bottom right not to scale) a profile of pressure vs. temperature loop is shown that corresponds to the heating and cooling curves. In essence, the single hydride system works as a heat pipe.

109 Hydrogen permeation through Polymer Membrane for Flow control of H Potentiostatic Experiments Experiments were performed using a simple fuel cell set up and applied a potential to the anode/cathode of the membrane. In this case we are performing experiments independent of the hydride heat pump in order ascertain the permeation behavior of the membrane at different temperatures (22 o C to 50 o C), voltages (Potentiostatic) and currents (Galvanostatic). Subsequent experiments, in the near future, we will interface the membrane system with the heat pump; effectively replacing the valve between the two hydride container of the heat pump. In the first set of these experiments, we performed Potentiostatic experiments in which hydrogen permeation characteristics were investigated by using a Nafion membrane (in fuel cell assembly) at room-temperature and constant voltages of 20 mv, 40 mv, 60 mv, 80 mv, and 100 mv. Figure 3.28 (a) shows a pressure (P 2 ) at the permeate side, referred to as counter electrode (CE), vs. Time. One can observe an increase in pressure at different rates, depending on the voltages used. It was observed that at higher applied voltages (100 mv) the pressure increase was rapid compared to lower voltages (between mv). It can also be observed that the nearly all the hydrogen on the high pressure side indicated in the Figure 3.28 (a) is transferred (permeated) to the empty side in 12.5 minutes at 100 mv. In the Figure 3.28 (b) we show Pressure (P 1 ) (WE) vs. Time plot for the corresponding high pressure side transferring the hydrogen, this side is also referred to as working electrode (WE). Figure 3.28 (c) shows the current vs. Time plot and the maximum current generated at 100 mv is 819 ma at which the permeation rate is

110 92 highest. Figures 3.28 (d) and (e) describe the pressure change on either side with the change in current at different voltages. In order to reverse the direction of hydrogen flow (so as to recharge the hydride in an actual system), we performed experiments that yielded Pressure vs. Time curves (Figure 3.28f) for the WE and CE sides at 100 mv, and -100 mv. The voltage was reversed after 90 minutes. To verify if hydrogen flow can be stopped when desired, we performed open circuit (OC) potential experiments that showed no hydrogen transport on either side, in other words pressures at either side remained the same (represented in the midsection between minutes at open-circuit (OC) in Figure 3.28f). The reversal of voltage yielded reverse hydrogen transport in the system. The corresponding current vs. time profile is shown in Figure 3.28g. To conclude, it was found that even low voltages applied to permeating membrane with catalyst allows significant amount of hydrogen transport at controllable rate. Thus, in a heat-pump with the membrane assembly incorporated we can control the hydrogen flow to the hydrides reversibly by application low voltages. More experiments were performed at 30, 35, 40, and 45 o C and at different voltages ranging from 20 to 1000 mv. These results are shown in Figures A-1 to A-4 in the Appendix. It should noted that there a finite amount of moisture added in all the experiments performed. However, to observe if the humidification had an impact on the permeation, we performed one experiment without humidifying the membrane that shows the characteristics of adding moisture to the membrane in Figure A-5 (Appendix).

111 Comparison of Permeation behavior at room temperature and 45 o C at 1V. In order to compare the hydrogen transport properties of the membrane at constant applied voltage of 1000 mv, we plotted P 1 and P 2 vs. time in Figures 3.29 (a) and (b), respectively. Comparative graphs of hydrogen permeation through the Nafion membrane using Potentiostatic experiments at room-temperature and 45 o C at a constant voltage of 1000 mv have been plotted. We found all the hydrogen is transported to the permeate side at room temperature in minutes at room-temperature, and in minutes at 45 o C at 1 V (Figure 3.29a and b). Figure 3.29c and d show corresponding current v. time changes; the maximum value of current at room-temperature is 715 ma and at 45 o C is 751 ma. Figure 3.29d and e show pressures (P 1 and P 2 ) at WE and CE vs. current respectively at these two temperatures. It can be seen that there is not much difference in the hydrogen permeation through these membrane at higher temperatures and voltages. So we propose using the lower voltages (100 mv) and room temperature for the heat pumps. It should be noted that a temperature of the system does not affect the hydrogen permeation significantly Galvanostatic Experiments In the first set of Galvanostatic experiments, we kept the current constant and varied the voltage using a Nafion membrane fuel cell assembly at room-temperature and constant currents of 10 ma, 10 ma, 30 ma, 40 ma, and 50mA. Figure 3.30 (a) shows a pressure (P 2 ) at the permeate side, referred to as counter electrode (CE), vs. Time. One can observe an increase in pressure at different rates, depending on the currents used. It was

112 94 observed that at higher applied currents (50 ma) the pressure increase was slightly more than observed at lower currents. However, the permeation was extremely low as compared the Potentiostatic experiments. Also, in none of the Galvanostatic experiments complete transport of the hydrogen from reservoir to the other was observed. Referring to Figures 3.30 (a) to (c) it can be observed that higher values of applied current yielded higher hydrogen transport at a relatively faster rate. The corresponding change in pressure on either side with the voltage generated can be seen in Figure 3.30 (d) and (e). In order to check if reversing the current affects the direction of hydrogen flow, we performed experiments that yielded pressure vs. time curves (Figure 3.30f) for the WE and CE sides at 50 ma, and -50 mv; the current was reversed after 90 minutes. We verified that open circuit does not transport the hydrogen. The reversal of current yielded corresponding voltage vs. time profile is shown in Figure 3.30g. To conclude, we found it was Galvanostatic experiments were not so affective for the hydrogen flow control at the low currents, although there is transport of hydrogen, it is very low compared to the Potentiostatic method. It is possible that increasing the area of the membrane or on applying significantly higher currents we achieve higher hydrogen transfer. Details of other Galvanostatic experiments at 40, 45 and 50 o C are shown in Figures A-6 to A-9 (Appendix) Summary Tables We have compiled the results of the permeation experiments in Tables A-1 and A-2 (Appendix) that show Potentiostatic and Galvanostatic results.

113 P2 (bar) Pressure vs Time (Counter Electrode) at RT 100 mv 80 mv 60 mv 40 mv 20 mv HP WE P1 LP CE P2 20 mv 40 mv 60 mv 80 mv 100 mv Time (s) (a) P2 (bar) Pressure vs Current (Counter Electrode) at RT 60 mv 80 mv 100 mv 20 mv 40 mv mv 60 mv 80 mv 20 mv 100 mv (d) Current (A) Pressure (bar) Potentiostatic Pressure vs Time at RT Forward H + Transport Forward H + Transport OC Reverse H + Transport Reverse H + Transport OC Reversal of Voltage P1 P Time (s) Working Electrode Counter Electrode (f) P1 (bar) Pressure vs Time (Working Electrode) at RT 20 mv 40 mv 60 mv 80 mv 100 mv P1 (bar) Pressure vs Current (Working Electrode) at RT 20 mv 40 mv 60 mv 80 mv 100 mv 20 mv 40 mv Current (A) 0.8 Current vs Time at RT 100 mv -100 mv Forward H + Transport 0.0 OC -0.2 Reverse H + Transport mv 100 mv 40 mv 60 mv 20 mv Time (s) (b) 60 mv mv 100 mv (e) Current (A) (g) Time (s) Current (A) Potentiostatic Curves at Room-Temperature 100 mv 80 mv 60 mv 40 mv 20 mv 20 mv 40 mv 60 mv 80 mv 100 mv Time (s) (c) Figure Hydrogen permeation through the Nafion membrane using electromotive force (Potentiostatic) at room-temperature at constant voltages of 20 mv, 40 mv, 60 mv, 80 mv, and 100 mv. (a) Pressure (CE) vs Time (b) Pressure (WE) vs Time: It can be observed that nearly all the hydrogen is transported to the permeate side in min at 100 mv (c) Current vs Time: The maximum value of current at 100 mv is 819 ma (d) Pressure (CE) vs Current (e) Pressure (WE) vs Current (f) Pressure vs Time curves for the WE and CE sides at 100 mv, OC, and -100 mv. (g) Current vs Time profile at 100 mv and -100 mv with open circuit (OC) potential applied at 60 min. The voltage was reversed at 90 min.

114 C Pressure vs Time (Counter Electrode) at 1 V 2.0 Pressure vs Time (Working Electrode) at 1 V RT 45 0 C 0.8 Potentiostatic Curves at 1 V RT 45 0 C 1.5 Room-Temperature P2 (bar) 1.0 P1 (bar) 1.0 Current (A) HP WE P1 LP CE P2 RT 45 0 C (a) Room-Temperature 45 0 C (b) Room-Temperature 45 0 C 0.0 (c) Time (s) Time (s) Time (s) 2.0 Pressure vs Current (Counter Electrode) at 1 V 2.0 Pressure vs Current (Working Electrode) at 1 V RT 45 0 C P2 (bar) 1.0 Room-Temperature 45 0 C P1 (bar) C Room-Temperature RT 45 0 C Figure Current (A) (d) Current (A) (e) Figure 3.29 Hydrogen H permeation through the Nafion membrane using electromotive force (Potentiostatic) at room-temperature and 45 o 2 C at a constant voltage of 1 V. (a) Pressure (CE) vs Time (b) Pressure (WE) vs Time: Almost all the hydrogen is transported to the permeate side in min at roomtemperature and in min at 45 o C at 1 V (c) Current vs Time: The maximum value of current at room-temperature is 715 ma and at 45 o C is 751 ma (d) Pressure (CE) vs Current (e) Pressure (WE) vs Current

115 97 P2 (bar) Pressure vs Time (Counter Electrode) at RT HP WE P1 LP CE P2 50 ma 40 ma 30 ma 20 ma P2 (bar) Pressure vs Voltage (Counter Electrode) at RT 10 ma 20 ma 30 ma 40 ma 50 ma 20 ma 30 ma 50 ma 40 ma Pressure (bar) Galvanostatic Pressure vs Time at RT Working Electrode Forward H + Transport Reverse H + Transport OC Reversal of Current OC P1 P ma (a) Time (s) ma (d) Voltage (V) 0.5 Forward H + Transport Reverse H + Transport Counter Electrode (f) Time (s) Pressure vs Time (Working Electrode) at RT Pressure vs Voltage (Working Electrode) at RT Forward H + Transport Voltage vs Time at RT OC Reverse H + Transport ma ma -0.4 P1 (bar) ma 20 ma 30 ma 40 ma 50 ma 20 ma 30 ma 40 ma 50 ma (b) P1 (bar) ma 20 ma 30 ma 40 ma 50 ma 20 ma 30 ma 50 ma 40 ma (e) Voltage (V) ma -50 ma (g) Time (s) Voltage (V) Time (s) Voltage (V) Galvanostatic Curves at Room-Temperature 50 ma 40 ma 30 ma 20 ma ma 20 ma 30 ma ma 50 ma (c) Time (s) 10 ma Figure Hydrogen permeation through the Nafion membrane using electromotive force (Galvanostatic) at room-temperature at constant currents of 10 ma, 20 ma, 30 ma, 40 ma, and 50 ma. (a) Pressure (CE) vs Time (b) Pressure (WE) vs Time: It can be observed that in galvanostatic scan all the hydrogen is not transported to the permeate side with the maximum amount (0.58 bar) being transported at 50 ma. (c) Voltage vs Time: The maximum value of voltage at 50 ma is -8 mv. (d) Pressure (CE) vs Voltage (e) Pressure (WE) vs Voltage (f) Pressure vs Time curves for the WE and CE sides at 50 ma, OC, and -50 ma. (g) Voltage vs Time profile at 50 ma and -50 ma with open circuit (OC) applied between 60 min and 90 min. The current was reversed at 90 min.

116 Summary of Fuel Cell Experiments The hydrogen transport was optimum by using the fuel cell membrane set up at 100 mv and room temperature. Although much higher applied voltages and temperatures may yield different results we have enough hydrogen transport with these parameters. It was also observed that Galvanostatic experiments were not so affective for the hydrogen flow control at the low currents, although there is transport of hydrogen, it is very low compared to the Potentiostatic method. It is possible that increasing the area of the membrane or on applying significantly higher currents we achieve higher hydrogen transfer. Details of other Galvanostatic experiments at 40, 45 and 50 o C are shown in Figures A-6 to A-9 (Appendix). 3.8 Electronic Control of Flow of hydrogen from Hydride B to A and vice versa Typically, heat pumps use mechanical valves to transfer hydrogen gas from one chamber to another. However, it is advantageous to use a membrane that can be electronically controlled. The main advantage is for recharging the system electronically by changing polarity. We are currently using a conventional fuel for this purpose which can generate power, as well. The temperature profiles of the heat pump operated by a mechanically valved system, as compared to an electronically controlled system using a membrane (Figure 3.31). There are two plots for hydrides, A and B shown in Figure 3.31 in which the cold side (B- black symbols) shows ~10.5 o C after ~ 4 minutes; while the low pressure side hydride (B-red color symbols) records a maximum temperature of 24 o C; with slight

117 99 increase in temperature subsequently. Now, let us compare the temperatures over 4 minutes with the mechanical valve; the minimum temperature is 7 o C. It should be noted that transport of hydrogen through the membrane is always slower as compared to a mechanical valve. But the valve operated heat pumps suffer with the disadvantage that the H 2 flow in between the hydrides is not controlled or reversible. This has implications when we want to recharge the high pressure hydride in B without heating the low pressure hydride in A by a simple voltage reversal. To the best of our knowledge electronic heat pump operation with a permeation membrane has never been thought of and attempted before. Also, cooling with such small amount of hydride has not been done before. Another advantage of using PEM instead of a mechanical valve seems to be the low hot side temperature obtained (only 27 o C compared to 40 o C). The pressure versus time plot is shown in Figure 3.31b. The associated current generated in the circuit (under an applied voltage of 100 mv) as a function of time is shown in Figure 3.31 (c). The current generated drops over time because the pressure differential in between the reactors decreases over time as the PEM pumps more and more H 2 from the cold side to the hot side. Now we will discuss reversal of hydrogen flow by electronic control via a PEM membrane in the system which is not possible in the mechanical valve system. Figure 3.31 (a) shows the pressure profiles of A and B hydrides as a function of time, using the PEM membrane. It should be noted that the overpressure in the both the hydride vessels is nearly zero. Hydrogen is in solid state in the hydride B and cylinder A there is only the alloy.

118 Valve Temperature vs Time Temperature ( o C) A (LP, Hot Side) LaNi 4.78 Sn 0.22 PEM B (HP, Cold Side) MmNi 4.15 Fe 0.85 PEM T min. (PEM) = 10.5 o C T min. (Valve) = 7 o C Valve MmNi 4.15 Fe 0.85 Pressure vs Time B (HP, Cold Side) PEM Time (s) T max. (PEM) = 27 o C T max. (Valve) = 40 o C W t. of alloy on each side = 7.65 g PEM Valve Valve PEM PEM (a) PEM PEM Valve Valve Pressure (bar) Valve PEM Valve A (LP, Hot Side) LaNi 4.78 Sn 0.22 W t. of alloy on each side = 7.65 g Time (s) Potentiostatic Scan 100 mv (b) PEM Current (A) Time (s) (c) Figure 3.31 (a) Comparison of heat pump operation operated with a valve and PEM (a). The minimum temperature obtained on the cold side with the proton exchange membrane (PEM) is 10.5 o C, whereas that obtained with mechanical valve is 7 o C. The maximum temperature with PEM is ~27 o C as compared to the valve operation of ~40 o C. (b) shows pressure vs. Time plot for the same experiment. (c) Current vs Time: Current generated in the circuit (under an applied voltage of 100 mv as a function of time.

119 101 Pressure (bar) Pressure vs Time MmNi 4.15 Fe B (HP, Cold Side) +100mV Mm@+100mV 00m 0mVV 00mV0mV 0mV -100mV 1 LaNi Sn 0.22 A (LP, H ot Side) (a) Tim e (s) m V 0.6 Current (A) PEM +100mV -100mV Temperature ( o C) Exothermic Tim e (s) Temperature vs Time A (LP, Hot Side) LaNi Sn m V Endothermic MmNi Fe 0.85 B (HP, Cold Side) Endothermic Tim e (s) -100mV (b) Exothermic Mm@+100mV 0mV Mm@-100mV mv (c) Figure 3.32 Heat Pump operation with positive (+100 mv) and negative (-100 mv) voltage for 4 hours each. Figure (a) shows +100mV profile on the left side. After 4 hours the polarity was reversed that resulted in changing the flow of hydrogen in the opposite direction. This portion of the heat pump operation is shown on the right. Figure (b) shows the changes in current corresponding to the profiles in Figure (a). Figure (c) The minimum temperature obtained on the cold side B (MmNi 4.15 Fe 0.85 ) with +100 mv is 10.5 o C, whereas on the hot side (A) it is only 27 o C.

120 102 Operation with +100mV: As soon as +100 mv is applied to the PEM membrane hydrogen is transported from cylinder B to A. Keeping in mind that the two hydrides are separated by the membrane, it can be seen that there is endothermic reaction in the hydride B, due to release of hydrogen, leading to low temperature of ~10.5 o C (Figure 3.32 (c)). The pressure in the cylinder B momentarily increases to ~4.5 bar; tracking the temperature in cylinder A in which hydride is being formed there is exothermic reaction taking place. Operation with 100 mv: when the voltage is reversed the flow of hydrogen reverses and the endothermic reactions occur in the cylinder A and exothermic reactions in cylinder B. This demonstrates the feasibility of using a PEM membrane in a heat pump. It can be seen that the pressure of the hydride A drops initially effect of change in polarity that reverses the H 2 transport. The temperature profiles of the two hydrides also change when the voltage is reversed. The minimum temperature obtained on the cold side (MmNi 4.15 Fe 0.85 ) with +100 mv applied voltage is 10.5 o C whereas on the hot side (LaNi 4.78 Sn 0.22 ) it is only 27 o C. The polarity was reversed after 4 hours. As more and more H 2 starts to go back to the cold side on the reversal of polarity, the temperature on the cold side (MmNi 4.15 Fe 0.85 ) starts to increase whereas a drop in temperature of the hot side (LaNi 4.78 Sn 0.22 ) is observed. Figure 3.32 (a) shows the associated pressure versus time profile. It can be observed that on the reversal of polarity the H 2 is pushed back to the high pressure (MmNi 4.85 Fe 0.15 ) side and as a consequence the pressure starts to drop on the low pressure (LaNi 4.78 Sn 0.22 ) side. The current generated in the circuit as a function of time is shown in Figure 3.32 (c).

121 Minimization of parasitic losses in the heat pump In order for the heat pump to perform at its true potential it is necessary to minimize the parasitic losses occurring in the system. In a heat pump, parasitic losses can result from two sources: the reactor bodies (steel) and from H 2 gas which is the working fluid. Mathematically, the parasitic losses can be represented as m.c p.dt, where m is the mass of the body of the heat or the mass of H 2 gas, c p is the heat capacity, and dt the change in temperature incurred. The heat capacity of the containers plus plumbing (steel) is 0.47 J/g.K and that of H 2 gas at room-temperature is J/g.K which is rather high (in fact 30 times higher than steel). The system volume and its mass are very high because of addition of the fuel cell in the system. The volume of system accessories is much higher compared to the hydride reactors. In summary, Volume of reactors, V reactor = 2.5 cc Volume on HP-LT side, V HP-LT = 252 cc, and Volume on LP-HT side, V LP-HT = 103 cc. Volume of accessories, V accessories (HP-LT) = cc (This volume is associated with the fuel cell plumbing that has a moisture trap in addition to fuel cell pipes.) Volume of accessories, V accessories (LP-HT) = cc (We do not have a moisture trap but other plumbing). For best results, hydrogen should be contained in near zero volume. Since the volume of accessories is 40 to 100 times the volume of the reactor, the HP-LT (MmNi 4.85 Fe 0.15 )

122 104 alloy does not go down in temperature in accordance with its potential. As soon as H 2 leaves the HP-LT (MmNi 4.85 Fe 0.15 ) hydride to the other side, H 2 from the moisture trap and the fuel cell pipe enters the hydride to increase its temperature. Moreover, V HP-LT > V LP-HT. Actually, V HP-LT = 2.45*V LP-HT. The HP-LT (MmNi 4.85 Fe 0.15 ) side will go down in temperaturemuch more if V HP-LT < V LP-HT or even if V HP-LT = V LP-HT, because in that case a significantly lower amount of H 2 has to be gotten rid of from the HP-LT (MmNi 4.85 Fe 0.15 ) side and pushed to the LP-HT (LaNi 4.78 Sn 0.22 ) side. But we did in fact carry out experiments (for the first time to the best of our knowledge) to minimize the losses occurring due to hydrogen gas mass. As expected, the minimum temperatures obtained on the cold side (7 o C o C) increased with increasing H 2 moles (0 to 2.25). Actually, the presence of extra H 2 over the alloy prevents the H 2 contained in the lattice of the hydride molecules to escape more freely. Therefore, the endothermic effect (drop in temperature) is minimized. It is not just a question of minimizing heat losses through the walls of system accessories, but more importantly of removal of excess H 2 in the system. A corollary of this discussion is that operating conditions (the presence of excess moles of H 2 in the system in particular) play an important role in heat pump performance. The results of the variation of excess H 2 moles in the system are shown in Figure 3.33 (a). There were 10 experiments performed with varying amount of H 2 moles (from 0 to 2.25). At zero excess H 2 moles we were able to attain lowest temperature of 7 o C with the heat pump. With increasing number of moles the temperature on both cold (7 o C o C)

123 105 and hot (40 o C - 60 o C) sides increases, and the associated pressure vs. time data is shown in Figure 3.33 (b). The minimum temperatures obtained as a function of operating pressures have been shown in Figure 3.33 (c). The results have been summarized in Table 3.2. The mass of the hydrides in each of the cylinders is 7.65 g. Table 3.2 Effect of the presence of excess H 2 moles on the temperature profile of the heat pump Excess H 2 moles Minimum Cold Side (B) temperature ( o C) Maximum Hot Side (A) temperature ( o C)

124 106 Temperature ( o C) Temperature vs Time LaNi 4.78 Sn 0.22 A (LP, Hot Side) A ( H 2 moles) Initial Temperature - RT B (0 H 2 mole) A (0 H 2 mole) B (0.25 H 2 mole) A (0.25 H 2 mole) B (0.50 H 2 mole) A (0.50 H 2 mole) B (0.75 H 2 mole) A (0.75 H 2 mole) B (1.00 H 2 mole) A (1.00 H 2 mole) B (1.25 H 2 moles) A (1.25 H 2 moles) B (1.50 H 2 moles) A (1.50 H 2 moles) B (1.75 H 2 moles) A (1.75 H 2 moles) B (2.00 H 2 moles) A (2.00 H 2 moles) B (2.25 H 2 moles) A (2.25 H 2 moles) 10 B (HP, Cold Side) B ( H 2 moles) MmNi 4.15 Fe 0.85 Wt. of alloy on each side = 7.65 g (a) Time (s) 16 Pressure vs Time 14 MmNi 4.15 Fe 0.85 B (HP, Cold Side) Wt. of alloy on each side = 7.65 g B (0 H 2 mole) A (0 H 2 mole) B (0.25 H 2 mole) A (0.25 H 2 mole) B (0.50 H 2 mole) H 2 moles A (0.50 H 2 mole) B (0.75 H 2 mole) A (0.75 H 2 mole) B (1.00 H 2 mole) Pressure (bar) H 2 moles 1.75 H 2 moles 1.5 H 2 moles 1 H 2 mole A (1.00 H 2 mole) B (1.25 H 2 moles) A (1.25 H 2 moles) B (1.50 H 2 moles) A (1.50 H 2 moles) B (1.75 H 2 moles) A (1.75 H 2 moles) B (2.00 H 2 moles) A (2.25 H 2 moles) B (2.50 H 2 moles) A (2.50 H 2 moles) H 2 mole 0.5 H 2 mole 0.25 H 2 mole LaNi 4.78 Sn 0.22 A (LP, Hot Side) 0 H 2 mole Time (s) Minimum Cold Side Temperature as a function of excess H 2 moles (b) 18 B (min) ( o C) B (HP, Cold Side) MmNi 4.15 Fe Weight of alloy = 7.65 g Volume of reactor, V reactor = 2.5 cc Volume on HP-LT side, V HP-LT = 252 cc (c) n (H 2 ) mole Figure 3.33 Heat Pump with the two hydrides, LaNi 4.78 Sn 0.22 and MmNi 4.15 Fe 0.85 in A and B cylinders, respectively. (a) Temperature vs Time (b) Pressure vs. Time (c) Minimum Temperature vs. number of excess. At 0 excess moles, for example, the high pressure (B) side showed a minimum temperature of 7 o C, whereas the low pressure side (A) reached ~60 o C at 2.25 excess moles.

125 Hydrogen Transfer from Hydride A to B (Reversed heat pump) A typical heat pump has the hydride B fully charged; in this case we have fully charged the hydride A to see the effects on the temperature and pressure profiles. In particular we were interested in minimizing the temperature obtained on the radiator side (in this case, hydride B), i.e., H 2 transfers from the hydride A to B (instead of from B to A). The same operating conditions as the previous experiments whose results are shown in Figure 3.31 were retained. It can be observed from Figure 3.34 (a) that the maximum temperature on the radiator side (side B) does not go above 27.5 o C (3.5 o C above room-temperature for this set of experiments) at any of the excess H 2 moles. But the minimum temperature on the source side (side A) is not very low (with the minimum being 22 o C). Figure 3.34 (b) shows the pressure versus time profile. The maximum cold side temperatures as a function of H 2 over-pressure has been plotted in Figure 3.34 (c). Table 3.3 summarizes the minimum and maximum temperatures attained during the operation of the heat pump. This is expected since the ratio of plateau pressures is 0.03 which is very low. A comparison of the temperature profile obtained from the regular heat pump (in which hydrogen flows from hydride B to A) and reversed heat pump (in which hydrogen flows from hydride A to B) is shown in Figure We would like to point out that the maximum radiator temperature (T max = 27 o C) is low in the reversed heat pump, as compared the regular heat pump (T max.=40 o C). However, the regular heat pump provided better cooling (T min = 7 o C) and the reversed heat pump shows T min. =22 o C. A comparison of the temperature profiles of the regular and reversed heat pumps with 0 excess moles is shown in Figure 3.35.

126 108 Pressure (bar) Temperature ( o C) Temperature vs Time B (HP, Cold Side) MmNi 4.15 Fe B ( H 2 moles) Initial Temperature - RT A ( H LaNi 4.78 Sn 2 moles) A (LP, Hot Side) H 2 transport from Hot to Cold Side (a) LaNi Sn A (LP, Hot Side) MmNi 4.15 Fe Time (s) Pressure vs Time 2.25 H 2 moles 2 H 2 moles 1.75 H 2 moles 1.5 H 2 moles 1.25 H 2 moles 1 H 2 mole 0.75 H 2 mole 0.5 H 2 mole 0.25 H 2 mole 0 H 2 mole B ( 0 H 2 mole) A ( 0 H 2 mole) B ( 0.25 H 2 mole) A ( 0.25 H 2 mole) B ( 0.50 H 2 mole) A ( 0.50 H 2 mole) B ( 0.75 H 2 mole) A ( 0.75 H 2 mole) B ( 1.00 H 2 mole) A ( 1.00 H 2 mole) B ( 1.25 H 2 moles) A ( 1.25 H 2 moles) B ( 1.50 H moles) 2 A ( 1.50 H moles) 2 B ( 1.75 H moles) 2 A ( 1.75 H moles) 2 B ( 2.00 H moles) 2 A ( 2.00 H moles) 2 B ( 2.25 H moles) 2 A ( 2.25 H moles) 2 B ( 0 H 2 mole) A ( 0 H 2 mole) B ( 0.25 H 2 mole) A ( 0.25 H 2 mole) B ( 0.50 H 2 mole) A ( 0.50 H 2 mole) B ( 0.75 H 2 mole) A ( 0.75 H 2 mole) B ( 1.00 H 2 mole) A ( 1.00 H 2 mole) B ( 1.25 H 2 moles) A ( 1.25 H 2 moles) B ( 1.50 H 2 moles) A ( 1.50 H 2 moles) B ( 1.75 H 2 moles) A ( 1.75 H 2 moles) B ( 2.00 H 2 moles) A ( 2.00 H 2 moles) B ( 2.25 H 2 moles) A ( 2.25 H 2 moles) B (HP, Cold Side) H 2 transport from Hot to Cold Side Time (s) Maximum Cold Side Temperature as a function of excess H 2 moles (b) B (max) ( o C) H 2 MmNi 4.15 Fe B (HP, Cold Side) (c) n (H 2 ) mole Figure 3.34 Reversed Heat Pump: The object of this experiment is to reduce the radiator side temperature. H 2 transfer is from the low pressure hydride (A: LaNi 4.78 Sn 0.22 ) to the high pressure hydride (B: MmNi 4.15 Fe 0.85 ). (a) Temperature vs. Time: Maximum temperature on the radiator side (side B) does not go above 27.5 o C (3.5 o C above room-temperature for this set of experiments). (b) Pressure vs. Time (c) Minimum Temperature vs. number of excess H 2 moles.

127 109 Table 3.3 Effect excess H 2 moles on the temperature profile of the reversed heat pump. Excess H 2 moles Maximum Cold Side (B) temperature ( o C) Minimum Hot Side (A) temperature ( o C) Temperature ( o C) Temperature vs Time B (Regular Heat Pump) Hydrogen Transfer: "B" to "A" A (Reversed Heat Pump) Hydrogen Transfer: "A" to "B" A (Reversed Heat Pump) Hydrogen Transfer: "A" to "B" B: MmNi 4.15 Fe 0.85 A: LaNi 4.78 Sn 0.22 Initial Temperature - RT 10 5 B (Regular Heat Pump) Hydrogen Transfer: "B" to "A" Time (s) Figure 3.35 Comparison of the temperature profiles obtained by using "regular" and "reversed" heat pump configurations.

128 Effect of excess H 2 moles on the temperature profiles using a Heat Pump with same Metal Hydride on Both Sides (P 1 /P 2 =1) It would be interesting to investigate the performance of a heat pump when the ratio of plateau pressures is 1. Therefore, both sides of the heat pump containers were filled with the same alloy (MmNi 4.15 Fe 0.85 ). The idea was to use the hysteresis in the hydrides as the difference of van t Hoff pressures at constant temperature. The temperature profiles for the pressure H 2 moles are recorded in Figure As observed before, the zero excess moles condition yielded the lowest temperature (T min =8 o C at 0 H 2 moles, and T min.= 21 o C at 1.75 H 2 moles. Figure 3.36 (b) shows the corresponding pressure profiles for the data in figure 3.36 (a). Figure 3.36 (c) summarizes the T max and T min at different pressures on the radiator and source side. The results for the experiments for mole variation have been summarized in Table 3.4. The MmNi 4.15 Fe MmNi 4.15 Fe 0.85 pair as opposed to MmNi 4.15 Fe LaNi 4.78 Sn 0.22 pair affords the obvious advantage of low radiator temperature as expected. The cooling obtained with the MmNi 4.15 Fe MmNi 4.15 Fe 0.85 pair (ratio of plateau pressures = 1) is more than that obtained with the LaNi 4.78 Sn MmNi 4.15 Fe 0.85 pair (ratio of plateau pressures = 0.03) but less than that obtained with the MmNi 4.15 Fe LaNi 4.78 Sn 0.22 pair (ratio of plateau pressures = 30) because cooling is directly proportional to the ratio of plateau pressures.

129 111 Table 3.4 Effect excess H 2 moles on the temperature profile a heat pump working with same metal hydride (MmNi 4.15 Fe 0.85 ) on both sides. Excess H 2 moles Maximum Cold Side (B) temperature ( o C) Minimum Hot Side (A) temperature ( o C) It can be observed from Figure 3.37 (a) that the heat pump working with MmNi 4.15 Fe LaNi 4.78 Sn 0.22 pair affords the best results (maximum temperature drops) when cooling of a hot source is required whereas Figure 3.37 (b) reveals that the heat pump working with MmNi 4.15 Fe MmNi 4.15 Fe 0.85 pair offers the best results when very low radiator temperature is required.

130 B (LP, Hot Side) Temperature vs Time B ( H 2 moles Temperature ( o C) Initial Temperature - RT B ( moles) B (HP, Cold Side) Mm Ni 4.15 Fe 0.85 Wt. of alloy on each side = 7.65 g 0 H 2 mol e 0 H 2 mol e 0.25 H 2 mole 0.25 H 2 mole 0.5 H 2 mole 0.5 H 2 mole 0.75 H 2 mole 0.75 H 2 mole 1.25 H 2 moles 1.25 H 2 moles 1.50 H 2 moles 1.50 H 2 moles 1.75 H 2 moles 1.75 H 2 moles MmNi Fe MmNi 4.15 Fe 0.85 Pair (a) Time (s) Pressure (bar) Pressure vs Time 0 H 2 mole 0 H 2 mole 0.25 H 2 mole 0.25 H 2 mole 0.5 H 2 mole 0.5 H 2 mole 0.75 H 2 mole 0.75 H 2 mole 1.25 H 2 moles 1.25 H 2 moles 1.50 H 2 moles 1.50 H 2 moles 1.75 H 2 moles 1.75 H 2 moles Wt. of alloy on each side = 7.65 g MmNi 4.15 Fe MmNi Fe 0.85 Pair (b) Time (s) Minimum and M aximum Tempera tures as a function of ex cess H 2 moles 25 B (LP, Hot Side) Mm Ni Fe T 15 MmNi 4.15 Fe B (HP, Cold Side) 10 B (min) B (max) MmNi 4.15 Fe MmNi 4.15 Fe 0.85 Pair (c ) n (H2) Figure 3.36 Heat Pump with the same metal hydride (MmNi 4.15 Fe 0.85 ) on both A and B sides. (a) Temperature vs. Time: Low temperatures on both high and low pressure sides were obtained. At 0 excess H 2 moles, for example, the high pressure side reached 8 o C whereas the low pressure side reached 25.5 o C. (b) Pressure vs. Time (c) Temperature vs. Pressure: The minimum and maximum temperatures on both sides as a function of pressure.

131 113 A comparison of the maximum and minimum temperatures on the cold and hot sides for the heat pumps working with MmNi 4.15 Fe LaNi 4.78 Sn 0.22 and MmNi 4.15 Fe MmNi 4.15 Fe 0.85 pairs as a function of excess H 2 moles in the system is shown in Figure Comparison of Minimum Temperatures as a function of H 2 moles MmNi Fe MmNi Fe 0.85 Pair 18 B (min) ( o C) MmNi 4.15 Fe LaNi Sn Pair 6 (a) n (H 2 ) 60 Comparison of Maximum Temperatures as a function of H 2 moles A (max) ( o C) MmNi Fe LaNi 4.78 Sn 0.22 Pair MmNi Fe MmNi Fe 0.85 Pair 25 (b) n (H 2 ) Figure 3.37 (a) Comparison of minimum temperatures on the cold side for MmNi 4.15 Fe LaNi 4.78 Sn 0.22 and MmNi 4.15 Fe MmNi 4.15 Fe 0.85 pairs as a function of excess H 2 moles. (b) Comparison of maximum temperatures on the hot side for MmNi 4.15 Fe LaNi 4.78 Sn 0.22 and MmNi 4.15 Fe MmNi 4.15 Fe 0.85 pairs as a function of excess H 2 moles.

132 H 2 transfer from MmNi 4.15 Fe 0.85 hydride to an empty cylinder (Heat Pipe) In order to determine the temperature drop in a "metal hydride heat pipe," series of experiments were done at different amounts of excess H 2 moles in the system where H 2 transfer took place from MmNi 4.15 Fe 0.85 hydride to an empty cylinder. The minimum temperatures obtained were above those obtained in the previous set of experiments. This is expected since there is no alloy present on the other side to absorb the released H 2. The temperature-time profile has been shown in Figure 3.38 (a). Figure 3.38 (b) shows the pressure versus time profile. The minimum hydride temperatures as a function of excess H 2 moles has been plotted in Figure 3.38 (b). The results have been summarized in Table 3.5. A comparison of the three approaches as a function of pressure has been shown in Figure The lowest temperatures were obtained for the MmNi 4.15 Fe LaNi 4.78 Sn 0.22 pair, followed by MmNi 4.15 Fe MmNi 4.15 Fe 0.85 pair, and finally by H 2 transfer from MmNi 4.15 Fe 0.85 to an empty cylinder. It is important to note here that the temperature differences become significant only at higher H 2 moles, whereas at low amount of hydrogen present in the system, the performance of any of the three systems is good. Table 3.5 Effect excess H 2 moles on the temperature profile for hydrogen transfer from MmNi 4.15 Fe 0.85 to an empty cylinder. Excess H 2 moles Minimum Temperature ( o C)

133 Temperature vs Time 0 H 2 mole 0.25 H 2 mole 0.50 H 2 mole 0.75 H 2 mole 1.00 H 2 mole Temperature ( o C) H 2 mole 0.75 H 2 mole 0.5 H 2 mole 0.25 H 2 mole 0 H 2 mole MmNi 4.15 Fe 0.85 to Em pty Cylinder Wt. of alloy = 7.65 g Time (s) (a) Pressure vs Time 1 H 2 mole 0.75 H 2 mole 0.5 H 2 mole Pressure (bar) H 2 mole 0 H 2 mole 2 0 MmNi 4.15 Fe 0.85 to Empty Cylinder Wt. of alloy = 7.65 g Time (s) (b) 18 Minimum Temperatures as a function of H 2 moles B (min) ( o C) n (H 2 ) (c) Figure 3.38 H 2 transfer from MmNi 4.15 Fe 0.85 hydride to an empty cylinder. (a) Temperature vs. Time: The hydride reached a minimum temperature of 9.5 o C at 0 excess moles of hydrogen. The minimum temperatures obtained were above those obtained in the previous set of experiments. This is expected since there is no alloy present on the other side to absorb the released H 2. (b) Pressure vs. Time (c) Minimum temperature as a function of excess moles of hydrogen.