IV.A.1m Preparation and Reactions of Complex Hydrides for Hydrogen Storage: Metal Borohydrides and Aluminum Hydrides TABLE 1. Progress towards Meeting Technical Targets for Reversible Gilbert M. Brown (Primary Contact), Hydrogen Storage Douglas A. Knight, Claudia J. Rawn, Characteristic/ Units 2010/2015 ORNL Antonio F. Moreira Dos Santos, System DOE Targets/ observed Theoretical Joachim H. Schneibel, and Frederick V. Sloop, Jr. System wt% (kg H2/kg 6%/9% − Gravimetric system x 100) Oak Ridge National Laboratory (ORNL) Capacity P.O. Box 2008, MS-6119 Oak Ridge, TN 37831-6119 Al(BH4)3 − 16.92% 14.5% Phone: (865) 576-2756; Fax: (865) 574-4939 Al(BH ) (NH ) − 17.19% 15.2% E-mail: [email protected] 4 3 3 2 Al(BH4)3(NH3)3 − 17.26% 16.0% DOE Technology Development Manager: Mg(BH ) (NH ) − 16.0% 14.6% Ned Stetson 4 2 3 2 Phone: (202) 586-9995; Fax: (202) 586-9811 E-mail: [email protected] Accomplishments Project Start Date: April 1, 2005 Progress was made toward preparing new materials Project End Date: September 30, 2010 and understanding the mechanism of hydrogen release from metal borohydrides: • Initiated a study of ammine metal borohydride (AMBH) complexes: synthesis, structure, and Objectives potential as hydrogen storage materials. • Mg- and Ca-AMBH were prepared via a solvent-free ORNL is conducting research to develop the synthesis route. chemistry for a reversible hydrogen storage system based on complex hydrides, chosen mostly from the • The solvent free reaction of anhydrous ammonia with Al(BH ) gives the product Al(BH ) (NH ) ; borohydrides, amine borohydrides, amides/imides, alane 4 3 4 3 3 2 or the alanates of the light elements in the periodic table reaction of ammonia with aluminum borohydride in that will achieve the DOE/FreedomCAR performance toluene and ether solvents gives different products. targets for 2010. • Investigated ionic liquids as a reaction medium for - regenerating alane. The AlH4 anion appears to react with the imidazolium cation of the ionic liquid. Technical Barriers Further work on ionic liquids will be discontinued. This project addresses the following technical barriers from the Hydrogen Storage section of the G G G G G Hydrogen, Fuel Cells and Infrastructure Technologies Program Multi-Year Research, Development and Introduction Demonstration Plan: (A) System Weight and Volume The focus of work at ORNL is the development of new materials, synthetic methods for materials, and (P) Lack of Understanding of Hydrogen Chemisorption studies of chemical reactions and processes that will lead and Physisorption to materials that will achieve the DOE/FreedomCAR performance targets for 2010. ORNL generally Technical Targets develops synthetic methods based on methodology that offers the most potential for scaling to kilogram ORNL is investigating high wt% materials, and and greater lot size. Synthetic methods are being efforts are directed toward making the release and developed and reactions are being studied for two uptake of hydrogen reversible. Percentages in Table 1 types of target materials for hydrogen storage: complex are for H2 release only. anionic materials (Metal Hydride Center of Excellence [MHCoE] Project B) and amide/imide (M-N-H) systems (MHCoE Project C). For both types of materials ORNL is seeking (a) methods to scale up synthesis of FY 2009 Annual Progress Report 467 DOE Hydrogen Program IV.A Hydrogen Storage / Metal Hydride CoE Gilbert M. Brown – Oak Ridge National Laboratory known materials or new materials identified by MHCoE The Mg-AMBH is a quite promising material partners (b) development of new materials. ORNL that was previously reported by Soloveichik, et al. [1]. is also collaborating with scientists at Brookhaven Our innovation is the preparation of this compound National Laboratory in MHCoE Project D, the alane by the direct (solvent free) reaction of anhydrous focus group, with an emphasis on using the methods ammonia with magnesium borohydride. This material of solution inorganic and organometallic chemistry in exhibits a remarkably low temperature for desorption an effort to develop or improve processes to regenerate compared to its parent metal borohydride, and our alane. work shows it undergoes a fairly clean hydrogen desorption reaction, lacking any trace of by-products Approach such as ammonia or diborane as observed by mass spectrometry. The quantity of hydrogen produced Our research involves materials from both known in the desorption indicates the ammonia is serving synthesis as well as those from newly designed synthetic as a source of hydrogen as well as the borohydride. procedures. The primary method used to follow the Interestingly, although Mg(BH4)2 was found to undergo progress of reactions which release hydrogen is by a completely solid state desorption, the Mg-AMBH temperature programmed pressure measurements melts just prior to the H2 desorption (~97°C). The solid where gaseous reaction products other than hydrogen magnesium:nitrogen:boron product following desorption are analyzed using a mass spectrometer or by infrared of the hydrogen is amorphous to XRD and remains spectroscopy. Hydrogen desorption and uptake are unidentified. To assist in the identification of the 15 investigated in a traditional Sieverts apparatus. Fourier residue, we obtained N-labeled ammonia and prepared 15 transform infrared spectroscopy, X-ray diffraction a sample of Mg(BH4)2( NH3)2. This sample was given (XRD), and elemental analysis by inductively coupled to our collaborators at the Jet Propulsion Laboratory and plasma-atomic emission spectroscopy as well as solution Caltech for analysis of the reaction product residue by 15 nuclear magnetic resonance (NMR) are available high resolution N-magic angle spinning-NMR (MAS- at ORNL to characterize solid reaction products, NMR). This experiment is still in progress. and Raman and solid NMR are available through The temperature programmed desorption of the collaboration with MHCoE partners. Structural Ca-AMBH yielded disappointing results. Although determination using high resolution synchrotron- this material was easily synthesized (as either a mono based XRD is available through internal collaboration or diammine species) via the solvent-free method, it at ORNL. Each material will initially be examined was found that this material loses the NH3 at a much in a dehydrogenization study, both with and without lower temperature than that for hydrogen desorption. catalysts. Those materials found to be suitable hydrogen Several studies with a Sievert’s apparatus proved that storage material candidates will then be examined for the material easily transformed back into Ca(BH ) regenerative (hydrogen absorption) capability. From 4 2 prior to H2 desorption, and then simply followed the these results, chemical (and-or procedural) adjustment usual desorption path previously shown for solvent free to the most promising of the materials will be made in calcium borohydride. On a positive note, we found order to obtain the optimal in hydrogen de/absorption that even the solvated Ca(BH4)2-tetrahydrofuran (THF) capacity. (as purchased from Aldrich) can be easily transformed into Ca(BH4)2-2NH3 and then subsequently converted Results to the solvent-free Ca(BH4)2 as the ammonia is released. This reaction could be potentially useful for The majority of research at ORNL during the preparing Ca(BH4)2 in the pores of a carbon scaffold. review period involved a study of AMBH, obtained The Ca(BH4)2 could be deposited in the pores in THF from the reaction of ammonia with a metal borohydride. solution, the THF removed by formation of the NH3 Although other researchers are investigating these adduct, and the NH3 removed by thermal treatment. materials as well, a unique feature in our study is our The ammonia adduct of LiBH4 was also observed to use of a solvent-free reaction system in which anhydrous evolve ammonia prior to loss of hydrogen from the ammonia is directly reacted with a metal borohydride. borohydride [2]. Our study has included the reaction of ammonia with Aluminum borohydride was found to readily react Mg, Ca, and Al borohydrides, and for all three metals with anhydrous ammonia in a 1:2 ratio in the absence we can easily form the AMBH as a completely solvent- of solvent to form Al(BH ) • 2NH . The formation of free product by direct reaction. This ammination of 4 3 3 this Al-AMBH resulted in a greatly increased material the metal borohydrides was also shown to provide very stability relative to the parent metal borohydride. While important and desirable material properties as well as an the volatile Al(BH ) (a liquid at room temperature) is increase in the amount of hydrogen desorbed in the case 4 3 explosive in air, its ammine counterpart is a white solid of Mg and Al borohydrides. that only slowly decomposes in air and gently bubbles hydrogen as water is added. This same increase of DOE Hydrogen Program 468 FY 2009 Annual Progress Report Gilbert M. Brown – Oak Ridge National Laboratory IV.A Hydrogen Storage / Metal Hydride CoE material stability has been noted with the ammination evolution of gas which included some diborane. One of the magnesium and calcium borohydrides as well. of the reaction products seems to be identical to the The H2 desorption study of this Al ammine metal product formed in ether solution borohydride demonstrate that the added ammonia is The product formed in toluene has the structure actually contributing to the wt% hydrogen desorbed shown in Figure 2 in which the Al centers have trigonal from the material as was observed for the Mg-AMBH. bipyramidal geometry with axial ammonia groups The temperature programmed desorption data is shown coordinated to the Al(III) sites. A view of the unit cell in Figure 1. Mass spectrometry analysis of the gas shows an additional ammonia group loosely associate desorbed from the Al-AMBH has revealed trace amounts with one of the borohydrides. We presume that the of ammonia, diborane, and borazine, B H N H , 3 3 3 3 structure of the solvent-free product Al(BH ) (NH ) has suggesting that the desorption reaction involves reaction 4 3 3 2 coordination about the Al site analogous to that shown of BH and NH components, facilitated by the Al(III), 3 3 in Figure 2.