Rubidium Hydride: An Exceptional Dehydrogenation Catalyst for the

Article pubs.acs.org/JPCC

Rubidium Hydride: An Exceptional Dehydrogenation Catalyst for the Lithium Amide/Magnesium Hydride System Tolulope Durojaiye, Jalaal Hayes, and Andrew Goudy* Department of Chemistry, Delaware State University, 1200 North DuPont Highway, Dover, Delaware 19901, United States ABSTRACT: The 2LiNH2/MgH2 system has been identified as an attractive system for hydrogen storage, but suitable catalysts are needed to reduce the hydrogen desorption temperature and improve the rates of reaction. One of the most effective catalysts for lowering the desorption temperature and improving the desorption rate from this system has been KH. In this work a new catalytic additive, rubidium hydride (RbH), was synthesized, and its effects on hydrogen desorption from the 2LiNH2/MgH2 system were studied and compared to those of KH. Temperature-programmed desorption measurements showed that the addition of approximately 3 mol % RbH lowered the desorption temperature of the system by 94 °C, which is somewhat better than KH. Desorption enthalpies for the catalyzed samples were found to be approximately 42 kJ/mol, which is significantly lower than the 65 kJ/mol that was found for the uncatalyzed mixture. The hydrogen desorption rate of the RbH-doped sample was found to be approximately twice as fast as the KH-doped sample and about 60 times faster than the uncatalyzed sample, making RbH one of the best catalytic additives to date for the 2LiNH2/MgH2 system. Modeling studies were done using two approaches, and both indicated that diffusion controlled the rate of hydrogen desorption from 2LiNH2/ MgH2 in the two-phase plateau region.

1. INTRODUCTION In recent years, complex metal hydrides have been widely studied for hydrogen storage applications because of their relatively high hydrogen-holding capacities.1−3 Systems of interest include the alanates,4,5 borohydrides,6,7 and amides.8,9 However, various scientific and technological barriers need to be overcome to optimize the potential of these materials. Problems such as high desorption temperatures, slow kinetics, poor reversibility due to loss of ammonia or diborane, and the formation of stable intermediates have been identified with many of these systems.10 Further investigations have shown that it is often convenient to study mixtures of these compounds to obtain more favorable thermodynamic conditions.11 Some binary mixtures that have been studied include LiNH2/MgH2, Mg(NH2)2/2LiH, LiBH4/CaH2, LiBH4/MgH2, Mg(BH4)2/Ca(BH4)2, LiBH4/LiNH2, LiNH2/LiH, and MgH2/ LiH.12−21 The US DOE Metal Hydride Center of Excellence (MHCoE) has described the 2LiNH2/MgH2 system as an important “near term” system for hydrogen storage that would be a good candidate for engineering subsystem testing. This is because of its good long-term cycling behavior and higher hydrogen capacity than NaAlH4.22 The mechanism of its hydrogen release has been investigated by several groups, and the general reaction scheme for the 2LiNH2/MgH2 system is reported as follows: When ball milled, the starting material (2LiNH2 + MgH2) undergoes a reaction to form (Mg(NH2)2 + 2LiH). This mixture can release hydrogen to form (Li2Mg(NH)2 + 2H2).15,22−24 The equations that represent these processes are MgH2 + 2LiNH 2 → Mg(NH 2)2 + 2LiH © 2013 American Chemical Society

Mg(NH 2)2 + 2LiH ↔ Li 2Mg(NH)2 + 2H 2

(2)

Recently, more work has been focused on lowering the desorption temperature and improving the kinetics of the 2LiNH2/MgH2 system with effective catalysts. Any potential catalyst would have to be soluble in the reaction region of this system to be effective. Several additives such as NaOH,25 V, V2O5, VCl3,26 Si, and Al27 have been tried. However, perhaps the most effective catalyst to date has been KH. Wang et al.15 reported that doping this system with approximately 3 mol % of KH greatly improved the desorption and absorption kinetics of Mg(NH2)2/2LiH. Luo et al.22 also verified this. Durojaiye et al. were able to compare the desorption kinetics of this system to others using constant-pressure thermodynamic driving forces.14 Since KH has been found to be such an effective catalyst for the Mg(NH2)2/2LiH system, it was of interest to determine if RbH might also serve as a suitable catalyst. Both rubidium and potassium are in the same group on the periodic table, and thus they have similar properties. They both have the same electronegativity, and they are known to form stable ternary amides with magnesium.28 Thus it seemed likely that RbH should also have excellent catalytic properties. In this study, the introduction of rubidium into the system was accomplished by replacing a portion of the LiH in eq 2 with RbH.

2. EXPERIMENTAL SECTION All materials used for this research were obtained from Sigma Aldrich. The LiNH2 was 95% pure, and the MgH2 was of Received: January 28, 2013 Revised: March 8, 2013 Published: March 11, 2013

(1) 6554

dx.doi.org/10.1021/jp400961k | J. Phys. Chem. C 2013, 117, 6554−6560

The Journal of Physical Chemistry C

Article

hydrogen storage grade, 98% pure. The RbH was prepared using a modified mechanical alloying method previously described by Elansari et al.29 The method consisted of placing rubidium metal (99.6% pure) into a stainless steel vial and then pressurizing it with hydrogen. The vial was then placed into a Fritsch pulverizette 7 planetary mill and subjected to milling at various rates. The vial had to be repressurized at various stages. At the conclusion of the ball milling, the vial was placed in an oven at 120 °C for 1 h. A white powder was formed at the end of the incubation period. Mixtures of catalyzed 1.9LiNH2/ 1.1MgH2 were also prepared by a mechanical alloying procedure. In this case the 1.9LiNH2/1.1MgH2 mixtures along with 3.3 mol % of either KH or RbH were separately placed in 65 mL stainless steel milling pots containing ∼20 g balls (two 12 mm and four 6 mm balls). The mixtures were mechanically milled in a SPEX 8000D high-energy mixer/mill for 2 h under an argon atmosphere. X-ray characterizations of all the materials used in this research were carried out using a Panalytical X’Pert Pro MPD X-ray Diffractometer model PW3040 Pro. Analyses were done using copper K-alpha radiation. The instrument was equipped with an accelerator detector that allowed for rapid analyses. Samples were covered with a polyimide film (Kapton) for protection from air and moisture outside of the glovebox during analyses. Kapton has very high transmittance to X-rays such that no adjustment was required. Temperature-programmed desorption (TPD) and pressure composition isotherm (PCI) analyses were done on the asmilled materials using a Gas Reaction Controller-PCI unit. This instrument was manufactured by the Advanced Materials Corporation. TPD measurements were carried out between 30 and 230/300 °C for the various mixtures at 4 °C/min. Data for isotherm temperatures ranging from 190 to 230 °C were collected using a Lab View based software program. The data obtained were used to calculate the enthalpy of each system from van’t Hoff plots. Differential thermal analyses (DTA) were carried out under an argon atmosphere using a Perkin-Elmer TGA/DTA Instruments. This instrument was used to determine the thermostability of the investigated Li−Mg−N−H systems discussed in this paper. DTAs were carried out between 30 and 230/300 °C for various mixtures at 4 °C/min. Data obtained from the DTA curves were used to calculate activation energies from Kissinger plots. Desorption kinetics measurements were carried out in the plateau region of the isotherms at 210 °C. For easy comparison, the kinetics was run at the same temperature and thermodynamic driving force. To achieve the same thermodynamic driving force for all of the experiments, the same ratio of plateau pressure to applied pressure was applied to each sample. Further details about this procedure have been described by Goudy and co-workers.16−18

Figure 1. (A) XRD patterns for the 2LiNH2/MgH2 mixture catalyzed by RbH. The diffraction pattern in (a) is for the RbH catalyst, (b) is the pattern for the catalyzed as-milled 2LiNH2/MgH2 mixture, (c) is the pattern for the dehydrided 2LiNH2/MgH2 mixture, and (d) is the pattern for the rehydrided 2LiNH2/MgH2 mixture. (B) XRD pattern for the 2LiNH2/MgH2 mixture catalyzed by KH. The diffraction pattern in (a) is for the KH catalyst, (b) is the pattern for the catalyzed as-milled 2LiNH2/MgH2 mixture, (c) is the pattern for the dehydrided 2LiNH2/MgH2 mixture, and (d) is the pattern for the rehydrided 2LiNH2/MgH2 mixture.

hydrided mixtures, respectively. A comparison of the patterns shows some significant differences in the dehydrided and hydrided samples. The pattern for the as-milled sample in Figure 1A(b) contains peaks for the reactants MgH2 and 2LiNH2. Upon complete dehydrogenation, the pattern in Figure 1A(c) contains only peaks for the expected dehydrogenated product, Li2Mg(NH)2. This product has an orthorhombic structure, and it agrees very well with that reported by Hu and Fichtner.30 No RbH peaks appear in this XRD pattern. However, upon rehydrogenation, the pattern in Figure 1A(d) is obtained. No peaks for the initial reactants MgH2 and 2LiNH2 are seen. This XRD pattern contains peaks for (Mg(NH2)2 + LiH), as indicated in eq 2 along with peaks for the RbH catalyst. This confirms that the reaction in eq 2 occurs reversibly. Since RbH was reformed this indicates that it was not consumed during the reaction and that it is truly behaving as a catalyst. The behavior displayed in Figure 1A for the RbHcatalyzed 2LiNH2/MgH2 system has also been observed for the KH-catalyzed system by Markmaitree et al., Luo et al., and Durojaiye et al.21,23,14 Figure 1B contains XRD patterns for the KH-catalyzed 2LiNH2/MgH2 mixtures. Figure 1B(a) shows the XRD pattern for the KH catalyst; Figure 1B(b) contains a pattern for the “as-milled” sample; while Figures 1B(c) and

3. RESULTS AND DISCUSSION 3.1. X-ray Diffraction (XRD) Analysis. The 2LiNH2/ MgH2 system reacts with hydrogen according to the reactions described in eqs 1 and 2. XRD measurements were done on the hydrided and dehydrided mixtures, with and without RbH catalyst, to confirm that the catalyzed 2LiNH2/MgH2 system could absorb and release hydrogen reversibly. Figure 1A(a) shows the XRD pattern for the RbH catalyst; Figure 1A(b) contains a pattern for the “as-milled” sample; while Figures 1A(c) and 1A(d) contain patterns for the dehydrided and 6555



Article

1B(d) contain patterns for the dehydrided and hydrided mixtures, respectively. A comparison of the patterns shows that the as-milled sample in Figure 1B(b) contains peaks for the reactants MgH2 and 2LiNH2. Upon complete dehydrogenation, the pattern in Figure 1B(c) contains only peaks for the expected dehydrogenated product, Li2Mg(NH)2. This is the same behavior that was observed for the RbH-catalyzed mixture. However, unlike the dehydrogenated RbH-catalyzed mixture, a peak at 27° for KH does appear in this XRD pattern for the dehydrogenated state. The most likely reason for the observed difference is the fact that the dehydrogenations were done at 210 °C. RbH decomposes at 170 °C, and therefore none of its peaks appear in the dehydrogenated pattern. However, KH does not decompose until 400 °C, and thus it has a peak in the dehydrogenated pattern. Upon rehydrogenation, the pattern in Figure 1B(d) is obtained. No peaks for the initial reactants MgH2 and 2LiNH2 are seen. This XRD pattern contains peaks for (Mg(NH2)2 + LiH), as indicated in eq 2 along with peaks for the KH catalyst. This is the same behavior that was observed for the RbH-catalyzed mixture. 3.2. Temperature-Programmed Desorption (TPD) Analysis. It has already been established that KH is among the most effective catalysts for lowering the desorption temperature of the 2LiNH2/MgH2 system. Therefore, TPD measurements were performed on this system using RbH and KH catalysts to compare their effectiveness. The samples were heated from 30 to 250/350 °C at 4 °C/min. The results are shown in Figure 2. When reporting TPD results, it is a common

Table 1. Summary of Desorption Parameters for Thermodynamics and Kinetics Results Onset Temp. (°C) Desorp. Temp., Td (°C) Desorp. ΔH (kJ/mol) T90 (min) Ea (kJ/mol) Pm at 210 °C (atm) “m” values

KH

RbH

uncatalyzed

75 146 42.0 ± 0.12 62 87.0 ± 0.7 46.1 0.587

76 143 42.7 ± 0.03 27 86.8 ± 0.1 48.3 0.652

109 237 65.8 ± 0.04 1600 119.0 ± 1.7 34.2 0.647

3.3. Pressure Composition Isotherm Analysis and van’t Hoff Plots. Thermodynamic stabilities were determined from pressure composition temperature (PCT) isotherm measurements that were carried out on the catalyzed and uncatalyzed mixtures. Measurements were made in the 200− 230 °C range. Figure 3A contains the PCT isotherms for the

Figure 2. TPD profiles for catalyzed and uncatalyzed 2LiNH2/MgH2 mixtures. Data were collected at a scan rate of at 4 °C/min.

practice to report the onset temperatures of the dehydrogenation process, but onset temperatures can be misleading since they are the temperature at which dehydrogenation just begins to occur. This temperature is often much lower than that during which the bulk of the hydrogen is released. Therefore a decision was made to compare desorption temperatures that correspond to the inflection point of the TPD curve. The results in Figure 2 show that the desorption temperatures, Td, are in the order: RbH ≤ KH ≪ Uncatalyzed. The onset temperatures and desorption temperatures for these materials are reported in Table 1. The fact that RbH- and KH-catalyzed mixtures have much lower desorption temperatures than the uncatalyzed mixture is a bit surprising. If the RbH and KH are truly behaving as catalysts, they should not have any effect on the desorption temperature of this system. Therefore it was useful to determine the thermodynamic stabilities of these mixtures.

Figure 3. (A) Desorption PCT isotherms for the RbH-doped 2LiNH2/MgH2 mixture. (B) Desorption PCT isotherms for the uncatalyzed 2LiNH2/MgH2 mixture.

RbH-doped mixture, and Figure 3B contains the PCT isotherms for the uncatalyzed mixture. The isotherms show a well-defined plateau region at each temperature. The plateau pressures for the RbH-doped mixture are higher than those for the uncatalyzed mixture under the same temperature conditions. The plateau pressures were used to construct van’t Hoff plots, and the slopes were used to determine the ΔH values for the desorption reactions. van’t Hoff plots for the RbH-doped, KH-doped, and uncatalyzed samples are shown in Figure 4. The enthalpy values obtained for the different mixtures are shown in Table 1. The desorption enthalpy values 6556



Article

Figure 5. Differential thermal analysis plots for a 2LiNH2/MgH2 mixture catalyzed by RbH. The scan rate units (dpm) denote degrees per minute. Figure 4. van’t Hoff plots for hydrogen desorption from the catalyzed and uncatalyzed 2LiNH2/MgH2 mixtures.

the DTA curves. The slopes of these plots, shown in Figure 6, were used to determine the activation energies of each mixture.

obtained for the RbH-doped mixture (42.7 kJ/mol) and the KH-doped mixture (42.0 kJ/mol) are similar in magnitude. This is to be expected since the hydrogen desorption temperatures of both of these mixtures are nearly the same. These enthalpies are significantly lower than that (65 kJ/mol) found for the uncatalyzed mixture. This is also expected since desorption of hydrogen from the uncatalyzed mixture occurs at a much higher temperature than those for the catalyzed mixtures. It should also be noted that the desorption enthalpy for the uncatalyzed mixture is in the same range as the 53.4 kJ/ mol that was reported based on DFT calculations.34 Some other investigators have reported desorption enthalpies in the 40−44 kJ/mol range,15,22,31 but these are not consistent with our findings. Our findings can be explained based in part on the fact that Rb and K have the same Pauling electronegativity value, 0.82. Apparently both of these electropositive elements have nearly equal destabilizing effect on the N−H bond. This destabilization results in significant reductions in the desorption temperature and enthalpy. Thus, we see that RbH and KH are behaving as more than just catalysts. Perhaps a better term would be “catalytic additive”. 3.4. Differential Thermal Analysis (DTA) and Kissinger Plots. Differential thermal analysis curves (DTA) were carried out to better understand the effects of the catalytic additives on the dehydrogenation of LiNH2/MgH2 mixtures. The DTAs were carried out for the catalyzed and uncatalyzed mixtures at different heating rates between 4 and 20 °C per minute. Figure 5 shows the DTA curves for the RbH-doped mixture at different heating rates. The plots show that the desorption temperatures increase with increasing scan rates. These scans are typical of those that were obtained for the other mixtures. There is a relationship between the scan rate and the activation energy that is described by the Kissinger equation.32 ⎛ β ⎞ E ⎛ 1 ⎞ ln⎜ 2 ⎟ = − a ⎜ ⎟ + FKAS(α) R ⎝ Tmax ⎠ ⎝ Tmax ⎠

Figure 6. Kissinger plots for catalyzed and uncatalyzed 2LiNH2/MgH2 mixtures. β represents the scan rate.

The activation energies for KH- and RbH-doped mixtures are 87 and 86.8 kJ/mol, respectively, which is much smaller than the activation energy (119 kJ/mol) for the uncatalyzed mixture. The similar activation energies for the two catalyzed mixtures can be correlated with the electronegativity values of Rb and K in the same way as was done for desorption temperatures. Table 1 shows the activation energy values of all the mixtures. 3.5. Kinetics and Modeling Studies. Fast reaction rate is just as important as low desorption temperatures according to the DOE requirements for hydrogen storage systems. Therefore comparisons were made of the desorption kinetics of the catalyzed and uncatalyzed systems. Kinetics measurements were carried out for each mixture in the two-phase plateau region at 210 °C using constant pressure thermodynamics forces. Constant pressure thermodynamic force is achieved by keeping the ratio of the plateau pressure to the applied pressure constant for all measurements. The ratio has been defined as the N-Value. In these experiments, an N-Value of 10 was applied to all samples. A summary of the plateau pressures at 210 °C is shown in Table 1. Figure 7 contains the desorption kinetics plots of 2LiNH2/MgH2 mixtures with and without RbH and KH catalysts. It can be seen that the order of reaction

(3)

where Tmax = the temperature at maximum reaction rate; β = heating rate; Ea = the activation energy; α = the fraction of transformation; FKAS(α) is a function of the fraction of transformation; and R = the gas constant. Kissinger plots for various mixtures were constructed from the data obtained from 6557



Article

Figure 8. Modeling plots for the RbH-doped mixture. The experimental data fit the diffusion-controlled model during the first 70% of the reaction.

Figure 7. Desorption kinetics plots for catalyzed and uncatalyzed 2LiNH2/MgH2 mixtures. Kinetics measurements were done at 210 °C and at an N-value of 10.

kinetic rates is RbH > KH ≫ Uncatalyzed. The RbH-doped mixture desorbs hydrogen about twice as fast as the KH-doped and about 60 times as fast as the uncatalyzed mixture. The times taken by the various reactions to reach 90% completion (T90) are shown in Table 1. The fact that the RbH-doped mixture desorbs hydrogen twice as fast as the KH-doped mixture was unexpected based on the fact that both Rb and K have the same electronegativity. Therefore kinetics modeling studies were conducted to determine if there was a plausible explanation for this. Smith and Goudy performed kinetics modeling studies on the LaNi5−xCox hydride system (x = 0, 1, 2, and 3) and used two theoretical equations based on the shrinking core model to model the system.33 These same equations were used to model the 2LiNH2/MgH2 system. They are as follows t = 1 − (1 − XB)1/3 (4) τ where

τ=

ρB R bksCAg

t = 1 − 3(1 − XB)2/3 + 2(1 − XB) τ where τ=

Figure 9. Modeling plots for the KH-doped mixture. The experimental data fit the diffusion-controlled model during the first 60% of the reaction.

(5)

ρB R2 6bDeCAg

where t is the time at a specific point in the reaction, and XB is the fraction of the metal reacted. R is the initial radius of the hydride particles; “b” is a stoichiometric coefficient of the metal; CAg is the gas phase concentration of reactant; De is the effective diffusivity of hydrogen atoms in the hydride; ρB is the density of the metal hydride; and ks is a rate constant. Equation 4 is based on a shrinking core model with chemical reaction at the phase boundary controlling the rate, whereas eq 5 is based on a process in which diffusion controls the reaction rate. Figures 8−10 show kinetic modeling results for the RbHdoped, KH-doped, and uncatalyzed mixtures, respectively. Equations 4 and 5 were fitted to the kinetic data in Figure 7 for each of the sample mixtures to determine which kinetic model best describes the reactions in this study. To determine

Figure 10. Modeling plots for the uncatalyzed mixture. The experimental data fit the diffusion-controlled model during the first 70% of the reaction.

the theoretical curves, it was first necessary to determine a value for the constant τ. This was accomplished through a series of statistical data analyses to determine a value which resulted in the smallest standard deviation between the experimental and 6558



Article

theoretical data. Figure 8, for the RbH-doped mixture, shows three curves. One is an experimental curve taken from Figure 7; a second curve was calculated from eq 4; and a third curve was calculated from eq 5. As shown in Figure 8, the data generated from the model with diffusion controlling the overall rate fit the experimental data better than the data generated from the model with chemical reaction controlling the overall reaction rate. This is true only during the first 70% of the reaction. During the last 30% of the reaction, neither model was a good fit with the experimental data. The reason for this is that during the first 70% of the reaction desorption occurred along the plateau region where the pressure remained relatively constant. Therefore the N-Value was constant, and thus the thermodynamic driving force was constant. When the reaction reached the left edge of the plateau region the pressure decreased sharply, and the thermodynamic driving force decreased. At this point the kinetics slowed more rapidly than either model predicted. The modeling results in Figure 9 for the KHcatalyzed mixture and in Figure 10 for the uncatalyzed sample showed similar results. In all cases, the data generated from the model with diffusion controlling the overall rate fit the experimental data better than the data generated from the model with chemical reaction controlling reaction rate. In all the models, diffusion controlled the rates only along the twophase plateau region. The fact that diffusion is the ratecontrolling process could be used to explain why the desorption rate in the RbH-doped mixture is faster than the KH-doped mixture. The atomic radius of Rb is 248 pm, whereas the radius of K is only 227 pm. Since Rb is larger, it would cause a slight expansion of the lattice, thereby allowing the diffusing species to move faster through the lattice. To confirm that diffusion controls the reaction rates, a second modeling technique for comparing the kinetics of solidstate reactions was applied. This method, which is based on the classical equation for analysis of nucleation and growth, was developed by Hancock and Sharp.34 The classical equation for nucleation and growth processes can be written as F = 1 − exp( −Bt m)

(6)

−ln ln(1 − F ) = ln B + m ln t

(7)

Figure 11. Modeling plots for catalyzed and uncatalyzed LiNH2/ MgH2 mixtures based on the method of Hancock and Sharp. “F” represents the reacted fraction.

4. CONCLUSIONS The results of this study show that the new catalytic additive, RbH, is significantly better than KH for enhancing hydrogen desorption from the 2LiNH2/MgH2 system. The RbH and KH additives have nearly the same capability for lowering the hydrogen desorption temperature of the 2LiNH2/MgH2 mixture. This is believed to be largely due to the fact that both Rb and K have the same electronegativity and thus similar destabilizing effect on the N−H bond. The fact that the desorption enthalpy for hydrogen release from the uncatalyzed mixture was much higher than that from the catalyzed mixtures is consistent with these findings. The kinetics results show that the RbH-doped mixture releases hydrogen over twice as fast as the KH-doped mixture and approximately 60 times faster than the uncatalyzed sample. The fact that the activation energy enthalpy for hydrogen release from the uncatalyzed mixture was much higher than that from the catalyzed mixtures is consistent with these findings. The modeling studies using the shrinking core model and a method developed by Hancock and Sharp both indicated that the reaction rates are controlled by diffusion in the two-phase plateau region. The faster kinetics in the RbHdoped mixture is attributed to the fact that Rb has a larger radius and is thereby able to expand the lattice, thus facilitating the diffusion process.

or

where F is the reacted fraction; B is a constant; and “m” is a constant that can vary according to the geometry of the system. In this method, plots of −ln ln(1 − F) vs ln t can be constructed and used to determine the rate-controlling process. The slope “m” is an indication of the rate-controlling process. In the analysis, “m” values in the 0.54−0.62 range denote diffusion-controlled reactions. Values in the 1.07−1.11 range indicate a phase boundary controlled process. This type of analysis was applied to the systems used in the current study. Figure 11 contains plots of ln ln(1 − F) vs ln t for the doped and uncatalyzed systems studied. In each case the analysis was done with reacted fraction values in the 0.15−0.50 range (or 15−50%). The values of “m” obtained for each mixture are tabulated in Table 1. The data show that the values for all the systems correspond most closely to those expected for diffusion-controlled reactions. Thus, two entirely different approaches have both indicated that diffusion is the ratecontrolling process in the 2LiNH2/MgH2 system.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: 302-857-6534. Fax: 302-857-6539. E-mail: agoudy@ desu.edu. Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS This research was financially supported by the U.S. Department of Energy and the U.S. Department of Transportation. REFERENCES

(1) Mauron, P.; Buchter, F.; Friedrichs, O.; Remhof, A.; Bielmann, M.; Zwicky, C. N.; Zuttel, A. Stability and Reversibility of LiBH4. J. Phys. Chem. B 2008, 112, 906−910. 6559



Article

(2) Zhou, Y.; Liu, Y.; Wu, W.; Zhang, Y.; Gao, M.; Pan, H. Improved Hydrogen Storage Properties of LiBH4 Destabilized by in Situ Formation of MgH2 and LaH3. J. Phys. Chem. C 2012, 116, 1588− 1595. (3) Jiang, Y.; Liu, B. H. Dehydrogenation Kinetics of 2LiBH4 + MgH2 Enhanced by Hydrogen Back Pressure and a CuCl2 Catalyst. J. Alloys Compd. 2011, 509, 9055−9059. (4) Zaluska, A.; Zaluski, L.; Strom-Olsen, J. O. Sodium Alanates for Reversible Hydrogen Storage. J. Alloys Compd. 2000, 298, 299−308. (5) Yang, H.; Ojo, A.; Ogaro, P.; Goudy, A. J. Hydriding and Dehydriding Kinetics of Sodium Alanate at Constant Pressure Thermodynamics Driving Force. J. Phys. Chem. C 2009, 113, 14512−14517. (6) Fang, Z.; Kang, X.; Wang, P. Improved Hydrogen Storage Properties of LiBH4 by Mechanical Milling with Carbon Additives. Int. J. Hydrogen Energy 2010, 35, 8247−8257. (7) Cerny, R.; Severa, G.; Ravnsbak, D.; Filinchuk, Y.; Anna, V.; Hagemann, H.; Haase, D.; Jensen, C.; Jensen, T. A Novel ScandiumBased Borohydride. J. Phys. Chem. C 2010, 114, 1357−1364. (8) Pinkerton, F. Decomposition Kinetics of Lithium Amide for Hydrogen Storage Materials. J. Alloys Compd. 2005, 400, 76−82. (9) Meisner, G.; Pinkerton, F.; Meyer, M.; Balogh, M.; Kundrat, M. Study of the Lithium−Nitrogen−Hydrogen System. J. Alloys Compd. 2005, 404, 24−26. (10) Jain, I.; Jain, P.; Jain, A. Novel Hydrogen Storage Materials: A Review of Lightweight Complex Hydrides. J. Alloys Compd. 2010, 503, 303−309. (11) Barison, S.; Agresti, F.; Russo, S.; Maddalena, A.; Palade, P.; Principi, G.; Torzo, G. A Study of the LiNH2−MgH2 System for Solid State Hydrogen Storage. J. Alloys Compd. 2008, 459, 343−347. (12) Luo, W. (LiNH2−MgH2): A Viable Hydrogen Storage System. J. Alloys Compd. 2004, 381, 284−287. (13) Liu, Y.; Zhong, K.; Gao, M.; Wang, J.; Pan, H.; Wang, Q. Hydrogen Storage in a LiNH2−MgH2 (1:1) System. Chem. Mater. 2008, 20, 3921−3927. (14) Durojaiye, T.; Goudy, A. Desorption Kinetics of Lithium Amide/Magnesium Hydride Systems at Constant Pressure Thermodynamic Driving Forces. Int. J. Hydrogen Energy 2012, 37, 3298−3304. (15) Wang, J.; Liu, T.; Wu, G.; Li, W.; Liu, Y.; Araujo, C.; Scheicher, R.; Blomqvist, A.; Ahuja, R.; Xiong, A.; Xiong, Z.; Yang, P.; Gao, M.; Pan, H.; Chen, P. Potassium-Modified Mg(NH2)2/2LiH System for Hydrogen Storage. Angew. Chem., Int. Ed. 2009, 48, 5828−5832. (16) Ibikunle, A.; Goudy, A.; Yang, H. Effects of Ti-Based Additives on the Hydrogen Storage Properties of a LiBH4/CaH2 Destabilized System. J. Alloys Compd. 2009, 475, 110−115. (17) Sabitu, S.; Goudy, A. Dehydrogenation Kinetics and Modeling Studies of 2LiBH4+MgH2 Enhanced by NbF5 Catalyst. J. Phys. Chem. C 2012, 116, 13545−13550. (18) Ibikunle, A.; Goudy, A. Kinetics and Modeling Study of a Mg(BH4)2/Ca(BH4)2 Destabilized System. Int. J. Hydrogen Energy 2012, 37, 12420−12424. (19) Aoki, M.; Miwa, K.; Noritake, T.; Kitahara, G.; Nakamori, Y.; Orimo, S.; Towata, S. Destabilization of LiBH4 by Mixing with LiNH2. Appl. Phys. A: Mater. Sci. Process. 2005, 80, 1409−1412. (20) Osborn, W.; Markmaitree, T.; Shaw, L.; Hu, J.-Z.; Kwak, J.; Yang, Z. Low Temperature Milling of the LiNH2+LiH Hydrogen Storage System. Int. J. Hydrogen Energy 2009, 34, 4331−4339. (21) Markmaitree, T.; Osborn, W.; Shaw, L. Comparisons Between MgH 2 and LiH Containing Systems for Hydrogen Storage Applications. Int. J. Hydrogen Energy 2008, 33, 3915−3924. (22) Luo, W.; Stavilla, V.; Klebanoff, L. New Insights into the Mechanism of Activation and Hydrogen Absorption of (2LiNH2− MgH2). Int. J. Hydrogen Energy 2012, 37, 6646−6652. (23) Luo, W.; Sickafoose, S. Thermodynamic and Structural Characterization of the Mg−Li−N−H Hydrogen Storage System. J. Alloys Compd. 2006, 407, 274−281. (24) Rijissenbeek, J.; Gao, Y.; Hanson, J.; Huang, Q.; Jones, C.; Toby, B. Crystal Structure Determination and Reaction Pathway of Amide−Hydride Mixtures. J. Alloys Compd. 2008, 454, 233−244.

(25) Liang, C.; Liu, Y.; Wei, Z.; Jiang, Y.; Wu, F.; Gao, M.; Pan, H. Enhanced Dehydrogenation/Hydrogenation Kinetics of the Mg(NH2)2−2LiH System with NaOH Additive. Int. J. Hydrogen Energy 2011, 36, 2137−2144. (26) Shahi, R.; Yadav, T.; Shaz, M.; Srivastva, O. Studies on Dehydrogenation Characteristics of Mg(NH2)2/LiH Mixture Admixed with Vanadium and Vanadium Based Catalysts (V, V2O5 and VCl3). Int. J. Hydrogen Energy 2010, 35, 238−246. (27) Nayebossadri, S. Kinetic Rate-Limiting Steps in Dehydrogenation of Li−N−H and Li−Mg−N−H Systems − Effects of Elemental Si and Al. Int. J. Hydrogen Energy 2011, 36, 8335−8343. (28) Jacobs, H.; Birkenbeul, J.; Kochelkorn, J. Darstellung und Eigenschaften der Amidomagnesate des Kaliums und Rubidiums K2[Mg(NH2)4]− und Rb2[Mg(NH2)4]−Verbindungen mit Isolierten [Mg(NH2)4]2‑-Tetraedern. J. Less-Common Met. 1984, 97, 205−214. (29) Elansari, L.; Antoine, L.; Janot, R.; Gachon, J.; Kuntz, J.; Guerard, D. Preparation of Alkali Metal Hydrides by Mechanical Alloying. J. Alloys Compd. 2001, 329, L5−L8. (30) Hu, J.; Fichtner, M. Formation and Stability of Ternary Imides in the Li−Mg−N−H Hydrogen Storage System. Chem. Mater. 2009, 21, 3485−3490. (31) Yang, J.; Sudik, A.; Wolverton, C. Activation of Hydrogen Storage Materials in the Li−Mg−N−H System: Effect on Storage Properties. J. Alloys Compd. 2007, 430, 334−338. (32) Kissinger, H. E. Reaction Kinetics in Differential Thermal Analysis. Anal. Chem. 1957, 29, 1702−1706. (33) Smith, G.; Goudy, A. J. Thermodynamics, Kinetics and Modeling Studies of the LaNi5−xCox Hydride System. J. Alloys Compd. 2001, 316, 93−98. (34) Hancock, J. D.; Sharp, J. H. Method of Comparing Solid-State Kinetic Data and Its Application to the Decomposition of Kaolinite, Brucite, and BaCO3. J. Am. Ceram. Soc. 1972, 55, 74−77.

6560


Rubidium Hydride: An Exceptional Dehydrogenation Catalyst for the

Recommend Documents