21816
J. Phys. Chem. C 2010, 114, 21816–21823
Pressure Effect on the 2NaH + MgB2 Hydrogen Absorption Reaction Claudio Pistidda,*,†,‡ Sebastiano Garroni,†,§ Christian Bonatto Minella,† Francesco Dolci,⊥ Torben R. Jensen,| Pau Nolis,∇ Ulrike Bo¨senberg,† Yngve Cerenius,# Wiebke Lohstroh,‡ Maximilian Fichtner,‡ Maria Dolores Baro´,§ Ru¨diger Bormann,† and Martin Dornheim† Institute of Materials Research, Materials Technology, GKSS Research Centre, Geesthacht GmbH, Max-Planck-Strasse 1, D-21502 Geesthacht, Germany, Institute of Nanotechnology, Karlsruhe Institute of Technology, Postfach 3640, 76021 Karlsruhe, Germany, Departament de Fı´sica, and SerVei de Ressona`ncia Magne`tica Nuclear (SeRMN), UniVersitat Auto`noma de Barcelona, 08193 Bellaterra, Spain, Center for Materials Crystallography, iNANO, and Department of Chemistry, Aarhus UniVersity, Langelandsgade 140, DK-8000, Denmark, Institute for Energy, DG Joint Research Centre, European Commission, P.O. Box 2, 1755 ZG Petten, The Netherlands, and MAX-lab, Lund UniVersity, S-22100 Lund, Sweden ReceiVed: August 5, 2010; ReVised Manuscript ReceiVed: October 6, 2010
The hydrogen absorption mechanism of the 2NaH + MgB2 system has been investigated in detail. Depending on the applied hydrogen pressure, different intermediate phases are observed. In the case of absorption measurements performed under 50 bar of hydrogen pressure, NaBH4 is found not to be formed directly. Instead, first an unknown phase is formed, followed upon further heating by the formation of NaMgH3 and a NaH-NaBH4 molten salt mixture; only at the end after heating to 380 °C do the reflections of the crystalline NaBH4 appear. In contrast, measurements performed at lower hydrogen pressure (5 bar of H2), but under the same temperature conditions, demonstrate that the synthesis of NaBH4 is possible without occurrence of the unknown phase and of NaMgH3. This indicates that the reaction path can be tuned by the applied hydrogen pressure. The formation of a NaH-NaBH4 molten salt mixture is observed also for the measurement performed under 5 bar of hydrogen pressure with the formation of free Mg. However, under this pressure condition the formation of crystalline NaBH4 is observed only during cooling at 367 °C. For none of the applied experimental conditions has it been possible to achieve the theoretical gravimetric hydrogen capacity of 7.8 wt %. Introduction The recent discovery of the unique kinetic property of MgB21-5 in facilitating the hydrogenation of light metal complex borohydrides at moderate pressure and temperature conditions has raised new prospects for the development of innovative high capacity-low enthalpy hydrogen storage materials. These new composite materials consist of a binary light metal hydride (like LiH, NaH, and CaH2) and MgB2 and can be hydrogenated according to the following general reaction
2/xMHx + MgB2 + 4H2 f 2/xM(BH4)x+MgH2
(1) where M stands for an alkaline or an alkaline earth metal (x ) 1, 2). The required temperature and pressure conditions for reaction 1 are generally much more moderate than the respective reaction using elemental boron.1-3 This finding suggests that MgB2 is kinetically superior to pure boron for hydrogen absorption reactions. However, the underlying reaction mechanisms are not yet fully understood. Among the hydrogen * To whom correspondence should be addressed. E-mail: claudio.pistidda@ gkss.de. † GKSS Research Centre. ‡ Karlsruhe Institute of Technology. § Departament de Fı´sica, Universitat Auto`noma de Barcelona. | Aarhus University. ⊥ DG Joint Research Centre. # Lund University. ∇ Servei de Ressona`ncia Magne`tica Nuclear (SeRMN), Universitat Auto`noma de Barcelona.
storage materials listed above, recently 2NaH-MgB2 composites have received soaring attention.6-8 The system has a theoretical gravimetric hydrogen capacity of 7.8 wt %, with an overall reaction enthalpy of 62 kJ mol-1 H2 resulting in an equilibrium pressure of 1 bar at 350 °C.6 However, although this system had been recently investigated in several works, little is known about the hydrogenation process. In fact, although in their recent works Garroni et al.,6 Mao et al.7 and Czujko et al.8 discuss the multistep nature of the 2NaH-MgB2 absorption reaction and the impossibility to achieve the expected theoretical hydrogen capacity, the reasons which lay behind these behaviors are not investigated in detail. With the aim of clarifying the reaction mechanisms involved in the formation of complex borohydrides, we investigated the hydrogenation reaction of 2NaH + MgB2 by powder X-ray diffraction (PXD), in situ synchrotron radiation powder X-ray diffraction (SR-PXD), high pressure titration (volumetric measurement), high pressure differential scanning calorimetric techniques (HP-DSC), and solid state magic angle spinning nuclear magnetic resonance spectroscopy (MAS NMR). Experimental Details NaH (95% purity) and MgB2 (99.99% purity) were purchased from Sigma-Aldrich and Alfa Aesar, respectively. NaH and the MgB2 were charged into a hardened steel vial and milled for 1 h in a Spex 8000 ball mill, with a ball to powder ratio of 10:1. Handling and milling were performed in a dedicated glovebox under a continuously purified argon atmosphere. Powder X-ray diffraction analyses (PXD) were carried out using a Siemens D5000 X-ray diffractometer using Cu KR radiation. The powder was spread onto a silicon single crystal and sealed
10.1021/jp107363q 2010 American Chemical Society Published on Web 11/15/2010
Pressure Effect on the 2NaH + MgB2 Reaction
J. Phys. Chem. C, Vol. 114, No. 49, 2010 21817
Figure 1. Absorption kinetics of as-milled 2NaH + MgB2 measured in a Sievert’s-type apparatus. The samples were heated under 50, 25, and 5 bar of hydrogen pressure from RT to 400 °C (curves A, B, and C) using a heating rate of 3 °C/min. (Left) Complete measurements. (Right) First 3 h of absorption.
in the glovebox with an airtight hood of Kapton foil. However, due to an imperfect air isolation of the samples, the material partly oxidized during the PXD measurements and NaOH was observed in all the samples (Figure 2 below). In situ synchrotron radiation powder X-ray diffraction (SRPXD) measurements were performed at the MAX II Synchrotron, at beamline I711 in the research laboratory MAX-lab, Lund, Sweden. The selected wavelength was in the range of 1.06476-1.09719 Å for the various measurements. A special sample holder designed for in situ monitoring of solid/gas reactions was utilized.9-11 All the raw SR diffraction data were elaborated and converted to powder patterns by the use of the FIT2D program.12 Thermodynamic investigations during hydrogen absorption were performed by HPDSC Netzsch DSC 204 HP Phoenix. The HPDSC measurements were carried out at the constant pressure of 5, 25, and 50 bar of hydrogen pressure, respectively, with a heating rate of 5 °C/min. The HPDSC apparatus was placed in a dedicated glovebox under a continuously purified argon atmosphere. Volumetric measurements were performed using a Sievert’s-type apparatus (Hera, Quebec, Canada). The solid state nuclear magnetic resonance (MAS NMR) studies were performed using a Bruker Avance 400 MHz spectrometer with a wide bore 9.4 T magnet and employing a boron-free Bruker 4 mm CPMAS probe. All the spectra were acquired using a spinning speed of 12 kHz. The spectral frequencies were 128.33 MHz for 11B nucleus and 105.85 for the 23Na nucleus, and the NMR chemical shifts are reported in parts per million (ppm) externally referenced to BF3 · O(CH2CH3)2 and NaCl, respectively. Results Figure 1 shows the hydrogen absorption kinetics of as-milled 2NaH + MgB2 under 50, 25, and 5 bar of hydrogen pressure (curves A, B, C), measured from RT to 400 °C, using a heating rate of 3 °C/min, and kept in isothermal conditions at 400 °C for several hours. Curve A shows the first hydrogen absorption measurement performed at 50 bar. Hydrogenation starts at roughly 250 °C and is proceeding in two separate steps, which corresponds to a hydrogen uptake of about 0.6 and 3.2 wt %. The total amount of absorbed hydrogen reaches 3.8 wt %. However, hydrogen uptake stopped after ∼1 h. The absorption curve of as-milled 2NaH + MgB2 measured under 25 bar of hydrogen pressure (Figure 1, curve B) clearly indicates again a multistep hydrogen absorption uptake. The first step starts at
around 300 °C. The amount of hydrogen stored in this stage is equal to 0.7 wt %. Upon further heating, a second hydrogen absorption step starts roughly at 350 °C and continues at 400 °C. The measurement was stopped after 63 h, when the total amount of hydrogen charged in the system reached 6.2 wt %. Considering that the measurement was performed at a hydrogen pressure much lower than the previous experiment, the amount of hydrogen stored in the system is rather surprising. A further absorption measurement was performed at a pressure of 5 bar (Figure 1, curve C). The hydrogen uptake starts at roughly 330 °C and then continues at 400 °C. After 45 h, an amount of hydrogen equal to 3.8 wt % was stored into the system. Different from the first two absorption measurements, the volumetric measurement performed under 5 bar hydrogen pressure (curve C) does not show the first marked absorption step. Surprisingly the application of higher hydrogen pressure led to a decreased hydrogen uptake. Considering that the theoretical gravimetric capacity for the expected absorption reaction (reaction 1 with M ) Na, x ) 1) is equal to 7.8 wt %, the volumetric measurements clearly show that a complete hydrogenation reaction of the system 2NaH + MgB2 could not be achieved at any of the conditions described above. As the origins of the different absorption behaviors are not yet understood, XRD measurements of the absorbed materials were performed. The diffraction patterns of the material after hydrogen absorption at 400 °C under 50, 25, and 5 bar of hydrogen pressure (respectively patterns A, B, and C) are shown in Figure 2. The sample hydrogenated at 50 bar (pattern A) still contains the starting materials NaH and MgB2. In addition, NaBH4, NaMgH3, and free Mg are found. Note that there are no traces of MgH2 in the XRD data, although under these conditions MgH2 is known to be the thermodynamically stable phase, not Mg.13 The diffraction pattern of the material obtained after hydrogen absorption under 25 bar of hydrogen pressure (pattern B) shows exactly the same reflections observed for the material synthesized at a pressure of 50 bar (pattern A). Significant differences can be noticed in the respective diffraction phase intensities. In fact, the ratios of peak intensities of NaH and NaMgH3 reflections over those of NaBH4 are inferior for the material obtained at a pressure of 25 bar compared to the material prepared under 50 bar. For example, the intensity ratio of the peaks NaH (111) at 31.60 2θ angle and NaMgH3 (200) at 32.81 2θ angle to the NaBH4 peak (220) at 41.26 2θ angle changes respectively from 1.49 and 1.31 for the material charged at 50 bar to 0.42 and 1.17 for the material charged at 25 bar of H2 pressure.
21818
J. Phys. Chem. C, Vol. 114, No. 49, 2010
Figure 2. PXD patterns of 2NaH + MgB2 after absorption under 50 (A), 25 (B), and 5 bar (C) of hydrogen pressure at 400 °C. PXD data measured at RT using λ ) 1.541 84 Å.
Figure 3. HP-DSC traces of the 2NaH + MgB2 hydrogen absorption reactions, measured at 50 (A), 25 (B), and 5 bar of hydrogen pressure (C) from RT to 400 °C and subsequently cooled using a heating/cooling rate of 5 °C/min.
The application of just 5 bar leads to quite different results. Pattern C (Figure 2) shows the formation of only NaBH4 and the presence of free Mg together with the starting reactants. NaMgH3 could not be detected among the final absorption products. Although the simultaneous presence of unreacted materials, together with NaMgH3 and free Mg, well justifies the unachieved theoretical capacity, the reason for the formation of NaMgH3 and free Mg is an issue that still remains to be addressed. A common observation for all the above-described absorption experiments is that the material after absorption seems to “wet” the inner walls of the sample holder. For this reason we assume that one or more absorption reaction steps occurs via molten phases. In order to visualize the sequence of events taking place during absorption under various conditions, HP-DSC analyses were performed. Figure 3 shows the HP-DSC traces recorded under 50 (A), 25 (B), and 5 bar of hydrogen pressure (C) measured from room temperature to 400 °C and then cooled to room temperature (constant heating/cooling rate 5 °C/min). The measurement under 50 bar of H2 (curve A) shows during heating the presence of three main peaks: two exothermic events with respective onsets at 270 and 353 °C and one endothermic peak at 330 °C. During cooling, two exothermic peaks with onset
Pistidda et al.
Figure 4. Series of SR-PXD patterns of the 2NaH + MgB2 system heated under 50 bar of hydrogen pressure from RT to 400 °C and cooled to 240 °C (5 °C/min, λ ) 1.072 Å).
temperature of 367 and 316 °C are observed. Curve B in Figure 3, measured under 25 bar of H2, similarly to the measurement performed under 50 bar of hydrogen pressure (Figure 3, curve A), shows a main exothermic peak with an onset temperature of 284 °C and a small endothermic signal at 330 °C. It should be noticed that at the lower pressure the onset of the absorption process is shifted to higher temperatures. As also observed for HP-DSC measurement carried out at 50 bar, the cooling period is characterized by the presence of two exothermic signals with onsets at 366 and 316 °C. The HP-DSC trace measured under 5 bar of hydrogen pressure (Figure 3, curve C) shows, during heating, two exothermic signals with onset temperature at 290 and 320 °C. During the cooling period only one single exothermic event with onset temperature at 367 °C can be detected, which is in contrast to the measurements performed at 50 and 25 bar of H2 respectively, where two exothermic events (at roughly 367 and 316 °C) are observed. In order to clarify the undergoing hydrogenation mechanisms occurring at the various reaction conditions, in situ SR-PXD patterns were measured. The experimental conditions for H2 absorption matched the conditions applied for the HP-DSC measurements. Figure 4 shows the measurement carried out at 50 bar of H2 pressure, in scanning temperature from RT to 400 °C and then cooled to 240 °C, both with a heating/cooling rate of 5 °C/min. The phases in the starting material are NaH and MgB2. Upon heating, due to thermal expansion all peaks shift continuously toward lower 2θ angles. At roughly 280 °C, the formation of an unknown phase with major reflections at 14.36, 16.59, 19.54, 23.54, and 27.79 2θ angle (λ ) 1.072 Å) is observed. This phase is found to be stable up to 325 °C, and then the diffraction peaks disappear. Moreover, at roughly 330 °C, the formation of the NaMgH3 starts. This is accompanied by a significant decrease of the NaH diffracted intensity. Formation of NaMgH3 continues up to a temperature of 350 °C. At 380 °C, reflections of crystalline NaBH4 appear and continuously grow until the final temperature 400 °C is reached. The cooling period is characterized by two main events, which take place at 370 and 320 °C, respectively. At roughly 370 °C, the intensity of NaBH4 and NaH peaks quickly rise, and later at about 320 °C, the peaks related to the unknown phase observed during the heating period reappear. In a second experiment, the starting materials 2NaH + MgB2 have been heated to 400 °C in 50 bar of H2 and then kept at this temperature. This measurement simulates the absorption
Pressure Effect on the 2NaH + MgB2 Reaction
Figure 5. Series of SR-PXD patterns (λ ) 1.097 19 Å) of the 2NaH + MgB2 system heated (5 °C/min) under 50 bar of hydrogen pressure from RT to 400 °C and kept under isothermal conditions.
Figure 6. (I) 11B{1H} MAS (12kHz) single pulse NMR spectrum of as-milled 2NaH + MgB2, (II) as-milled 2NaH + MgB2 heated from room temperature up to a final temperature of 300 °C under a pressure of 50 bar of hydrogen, and (III) as-milled 2NaH + MgB2 heated from room temperature up to a final temperature of 400 °C and then kept at 400 °C for 2 h under a pressure of 50 bar of hydrogen.
process performed in the Sievert’s-type apparatus (Figure 5, wavelength ) 1.097 19 Å). The sequence of events taking place during the heating period exactly traces out those abovedescribed for Figure 4 until the point when NaBH4 forms. At 400 °C the NaH disappearance is followed by the appearance of an amorphous background at 18.69 2θ angle. Although it starts to form at roughly 350 °C, it becomes clearly visible only in the isothermal period at 400 °C. Most likely this molten phase was formed also in the previous SR-PXD analysis (Figure 4), but the short time spent above 350 °C did not lead to its significant formation. The isothermal period at 400 °C is characterized by the complete disappearing of the NaH diffraction peaks and the growth of the NaBH4 phase, which takes place within the first 30 min of the isothermal period. It must be noticed that NaBH4 formation is followed by the simultaneous appearance of free Mg. Due to the possible formation of amorphous compounds, the employment of the MAS NMR technique in the study of the absorption process is necessary. For this reason several specimens with different hydrogenation degrees were prepared and characterized. In all the MAS NMR analysis here reported (Figures 6, 7, 10, and 11), the main peak spinning sidebands are indicated with the sign *. The single pulse 11B{1H} and 23Na{1H} MAS NMR analyses are shown in Figures 6 and 7, respectively. The 11B{1H}
J. Phys. Chem. C, Vol. 114, No. 49, 2010 21819
Figure 7. (I) 23Na{1H} MAS (12kHz) single pulse NMR spectrum of as-milled 2NaH + MgB2, (II) as-milled 2NaH + MgB2 heated from room temperature up to a final temperature of 300 °C under a pressure of 50 bar of hydrogen, and (III) as-milled 2NaH + MgB2 heated from room temperature up to a final temperature of 400 °C and then kept at 400 °C for 2 h under a pressure of 50 bar of hydrogen.
spectrum collected for the as-milled material (Figure 6I) shows at 96.51 ppm a peak due to the presence of MgB2. Moreover, a small signal at -42.00 ppm is detectable (Figure 6I). The presence of this last signal (at -42.00 ppm), although suggesting a partial formation of NaBH4 already during milling, does not find confirmation in the 23Na{1H} MAS NMR analysis of Figure 7I. In fact, despite the signal observed at -42.00 ppm in Figure 6I, the 23Na{1H} spectrum of the as-milled 2NaH + MgB2 (Figure 7I) shows only the signal related to the NaH (10.84 ppm). Therefore, the signal at -42 ppm is not due to the formation of NaBH4. The investigation of the phase which generates the signal at -42.00 ppm (Figure 6I) is still in progress. A first hydrogen-charged specimen was prepared by heating the as-milled material from room temperature up to a final temperature of 300 °C (heating rate 5 °C/min), under a pressure of 50 bar of hydrogen, and subsequently cooling it down to room temperature. According to the HP-DSC analysis of Figure 3A and the SR-PXD data shown in Figure 4, this sample is expected to contain the observed unknown phase together with the starting reactants. The presence of an additional phase among the starting reactant is confirmed by both 11B{1H} and 23Na{1H} MAS NMR analysis (Figures 6II and 7II, respectively). The 11 B{1H} spectrum of Figure 6II shows an intensity increment of the already present peak at -42 ppm (peak at -42.29 ppm), in addition to the MgB2 signal at 96.51 ppm. Moreover, the 23 Na{1H} spectrum of Figure 7II shows, differently from the as-milled material (Figure 7II), the presence of two additional signals at -11.71 ppm and at -15.85 ppm. The last sample was prepared by heating the as-milled material from room temperature to a final temperature of 400 °C (heating rate 5 °C/min) under 50 bar of hydrogen, and then kept at 400 °C for 2 h (Figures 6III and 7III). The region of negative shift of the 11 B{1H} NMR spectrum in Figure 6III clearly shows at -42.29 ppm the peak relative to presence of the NaBH4, plus an additional signal at -15.97 ppm. Due to the overlapping with the NaBH4 signal at -15.97 ppm, the NaMgH3 signal (at -17.90 ppm) is not visible in the 23Na{1H} spectrum of Figure 7III. The analysis of the region of positive shifts reveals, in addition to the peak of the remaining MgB2 at 96.51 ppm, three more signals: a broad peak at 3.09 ppm (overlapped with the spinning sideband of MgB2) and two sharp peaks (probably related to the same B-containing compound) at 6.16 and 18.20 ppm.
21820
J. Phys. Chem. C, Vol. 114, No. 49, 2010
Figure 8. 11B{1H} and 1H decoupled 11B MAS (12kHz) single pulse NMR spectra of as-milled 2NaH + MgB2 heated from room temperature up to a final temperature of 400 °C and then kept at 400 °C for 2 h under a pressure of 50 bar of hydrogen (black line and blue line, respectively).
Although most likely the peak observed at 3.09 ppm is due to the formation of amorphous boron, due to the low signals proportion (less than 1% of total boron signal) an assignment for the peaks at -15.97, 6.16, and 18.20 ppm is rather difficult. For this reason, a direct spectral sensitivity comparison between equally recorded, proton-decoupled boron experiment 11B{1H} and proton-coupled boron experiment 11B was performed. This analysis allows distinguishing between species which contain boron atoms strongly coupled to protons (directly bonded) and those which are not. Several strategies are known for performing heteronuclear decoupling. Herein, for CPD (composite pulse decoupling) we employed the TPPM technique (two pulse phase modulation). Figure 8 shows the 11B{1H} MAS NMR spectra of as-milled 2NaH + MgB2 hydrogenated at 50 bar and 400 °C with proton CDP (blue line) and without proton CDP (black line). Clearly, the proton decoupling leads to an enhancement of the signal relative to the [BH4]- anion at -42.29 ppm and of its spinning sidebands at 144.31, 61.62, and -135.33 ppm. Different behavior is observed for the signals at 18.20, 6.16, 3.09, and -15.97 ppm (Figure 8). For these signals the application of the proton CDP does not influence the signal intensity. This behavior is due to the fact that these species contain B-atoms which are not strongly coupled to hydrogen atoms. This suggests the formation of species without B-H bonds. Spectrum III in Figure 7 shows the 23Na{1H} NMR analysis of the material hydrogenated at 400 °C and 50 bar of H2 pressure. At 10.84 ppm the peak corresponding to the remaining NaH and at -15.85 ppm the signal of the Na contained in the NaBH4 are detected. Note that the peak related to the presence of the unknown phase at -11.71 ppm (Figure 7II) completely disappeared. A further broad signal is observed at 2.32 ppm (Figure 7III). Most likely this last signal is related to the peaks observed at 18.20, 6.16, and -15.97 ppm in the 11 B{1H} NMR analysis of Figure 6, spectrum III. In order to investigate the reaction mechanisms involved in the hydrogenation at low pressure, a series of SR-PXD patterns were collected at 5 bar of hydrogen pressure heating up the sample from room temperature to 400 °C and then keeping it isothermally at this temperature for 3 h before cooling down to 60 °C (Figure 9, constant heating/cooling rate 5 °C/min). The phases in the starting material are again NaH and MgB2. In contrast to the measurements at elevated pressures, nothing happens during the heating until 320 °C is reached. At this
Pistidda et al.
Figure 9. Series of SR-PXD patterns (λ ) 1.072 Å) of the 2NaH + MgB2 system heated (5 °C/min) under 5 bar of hydrogen pressure from RT to 400 °C and kept under isothermal condition for 3 h before cooling down to 60 °C.
Figure 10. (I) 23Na{1H} MAS (12kHz) single pulse NMR spectrum of as-milled 2NaH + MgB2 and (II) as-milled 2NaH + MgB2 heated from room temperature up to a final temperature of 400 °C and then kept at 400 °C for 40 h under a pressure of 5 bar of hydrogen.
temperature the NaH and MgB2 peaks start to decrease and simultaneously the signal of an amorphous background formation is observed at about 19.50 2θ angle, which stays constant up to 400 °C. Within the first 10 min of the isothermal period at 400 °C, the complete disappearance of the NaH reflections together with the rise of free Mg peaks is observed. No significant changes are observed for the remaining part of the isothermal period. During the cooling, at roughly 370 °C the instantaneous formation of NaBH4 and partial reformation of NaH coupled with the vanishing of the amorphous background are observed. No further changes occur during cooling until the measurement was stopped at 60 °C. Like for the absorption measurement performed at 50 bar of hydrogen pressure, also the hydrogen absorption process under 5 bar was investigated by MAS NMR technique (Figures 10 and 11). Although the 23Na{1H} and the 11B{1H} NMR spectra for the as-milled material were previously discussed, they are here presented again for comparison purpose (Figures 10I and 11I). In the 23Na{1H} NMR analysis of the hydrogenated material (Figure 10II), the signals of the unreacted NaH at 10.88 ppm, and of the formed NaBH4 at -15.65 ppm, are clearly visible. The 11B{1H} MAS NMR spectrum (Figure 11II) shows
Pressure Effect on the 2NaH + MgB2 Reaction
Figure 11. (I) 11B{1H} MAS (12kHz) single pulse NMR spectrum of as-milled 2NaH + MgB2 and (II) as-milled 2NaH + MgB2 heated from room temperature up to a final temperature of 400 °C and then kept at 400 °C for 40 h under a pressure of 5 bar of hydrogen.
a peak at 96.61 ppm, due to the presence of unreacted MgB2, and a strong signal at -42.76 ppm which clearly confirms the formation of NaBH4. Different from the material hydrogenated under 50 bar, the application of a pressure of 5 bar seems not to lead to the formation of phases different from NaBH4. Discussion The results presented above demonstrate that the hydrogen absorption process for the 2NaH + MgB2 system is a multistep reaction dependent on the hydrogen pressure under which the measurement is performed. Besides the formation of NaBH4, the formation of several intermediate phases is observed under 50, 25, and 5 bar of H2. In order to give a more exact description of the events taking place under the investigated conditions, a cross-check between the in situ SR-PXD analysis, HP-DSC measurements, and MAS NMR studies was conducted. Due to its features, the hydrogen absorption reaction of the system 2NaH + MgB2 performed under a pressure of 50 bar (Figure 1, curve A) can be divided in two different parts. The first part includes the first absorption step from an amount of stored hydrogen equal to 0 wt % up to 0.6 wt %. Then the second part goes from 0.6 wt % until the measurement stops. Regarding the first part of the absorption reaction, the comparison of the in situ SR-PXD measurement in Figure 4 with the HP-DSC trace A in Figure 3 allows to attribute the first absorption step to the formation of the observed unknown phase and NaMgH3 (Figure 4). Based on the MAS NMR analysis of Figures 6II and 7II, it is possible to claim that the unknown phase contains sodium (peaks at -11.71 and -15.85 ppm in Figure 7II) and boron bonded with hydrogen atoms in the form of [BH4]- anions (increased peak at -42.29 ppm in Figure 6II). In addition, the comparison between the MAS NMR analyses performed on the materials completely hydrogenated at 50 and 5 bar clearly shows that along with NaMgH3 several others phases are formed. Although these phases are not detectable in the in situ SR-PXD measurements (Figures 4 and 5), their presence is recognizable in Figure 7III (signal at -11.71 ppm) and Figure 6III (signals at 3.09, 6.16, 18.20, and -15.97 ppm). Among these phases, the presence of amorphous boron is strongly suggested by the presence of the broad peak with the maximum at 3 ppm in the MAS NMR trace of Figure 6III. In addition, the CDP analysis (Figure 8) points out that the other phases formed together with NaMgH3 do not contain hydrogen.
J. Phys. Chem. C, Vol. 114, No. 49, 2010 21821 Analyzing the second part of the 2NaH + MgB2 absorption reaction, it is possible to relate the endothermic peak of the HP-DSC curve with the maximum at 330 °C to the disappearing of the unknown phase (Figures 4 and 5). The exothermic signal starting at 353 °C (Figure 3, curve A) is due to the simultaneous formation of NaBH4 and free Mg (Figures 4 and 5). Earlier in this work, we claimed that most likely one or more reaction steps could take place in the liquid state. This assumption is further confirmed by the appearance of the amorphous background in the SR-PXD (Figure 5) at roughly 350 °C. Trace A in Figure 3 shows during the cooling period the presence of two exothermic peaks at 367 and 316 °C. These two signals are related to the NaBH4-NaH precipitation and recrystallization of the unknown phase, respectively (Figure 4). The formation of a molten NaBH4-NaH phase explains the reason why although NaBH4 already forms at 353 °C it is not visible until the temperature reaches 380 °C (Figure 4). The absence of the unknown phase signals in the MAS NMR analysis in Figures 6III and 7III indicates that this phase is progressively consumed or it decomposes when the material is kept under isothermal conditions at 400 °C The amount of hydrogen stored in the system under 25 bar of hydrogen pressure sensibly differs from that stored under 50 bar. However, based on the HP-DSC trace (Figure 2, trace B) and the PXD analysis (Figure 3, pattern B) we assume a common reaction path for both of them. The absorption measurement performed under 5 bar does not lead to the formation of the unknown phase and NaMgH3. This suggests that the appearance of these two phases strongly depends on the applied hydrogen pressure. Comparing the calorimetric analysis of Figure 3, curve C to the in situ SRPXD measurements of Figure 9, it is possible to assign the exothermic signal at 320 °C (Figure 3, curve C) to the formation of the amorphous background (Figure 9). In addition, the exothermic peak observed during cooling at 367 °C (Figure 3, curve C) has to be related to the simultaneous precipitation of NaBH4 and NaH (Figure 9). It must be noticed that, differently from the HP-DSC analysis carried out at higher pressure (50 and 25 bar), the measurement performed at 5 bar shows upon cooling a single exothermic peak at 316 °C only. This can be explained by the fact that the unknown phases are not formed under these pressure conditions. The molten phase, which so far has been always observed in almost all measurements, presumably comprises a molten mixture of NaH and NaBH4 since both these phases reappear in the SRPXD analysis of Figure 9, once the temperature decreases below 370 °C. Moreover, further experiments performed on a 1:1 mixture of NaH and NaBH4 show a melting point at 383 °C (under 5 and 50 bar of H2 pressure). This temperature is lower than the melting points of both single compounds (Tm(NaH) ) 400 °C14 and Tm(NaBH4) ) 505 °C15). This experiment confirms a reaction which involves only NaH and NaBH4. It follows from the above discussion that for the measurements performed under 50 and 5 bar of H2 pressure, the observed 2NaH + MgB2 hydrogen absorption reaction can be resumed by the schemes shown in Figures 12A and 12B, respectively. A peculiarity in all the absorption measurements in this work is the lack of MgH2 even though its formation is thermodynamically favored in the experiments at 50 and 25 bar of H2 and it has been observed by Barkhordarian et al.3 and Mao et al.7 However, in these experiments NaMgH3 and free Mg are found instead. Recently Ikeda et al.16 reported that NaMgH3 can be synthesized by mechanical milling of NaH and MgH2 mixture, at ambient temperature. Considering that our experiments were carried out at
21822
J. Phys. Chem. C, Vol. 114, No. 49, 2010
Pistidda et al.
Figure 12. Scheme of the 2NaH + MgB2 absorption reaction performed under 50 and 5 bar of H2 pressure.
Figure 13. Absorption kinetic of the material 2NaH + MgB2 as-milled measured in a Sievert’s-type apparatus. The measurement was performed at 450 °C under 20 bar of hydrogen pressure. Thick line shows hydrogen uptake, while thin dotted line refers to the temperature of the powder bed.
relatively high temperatures, a possible explanation for the missing MgH2 formation can be given by the fact that as soon as MgH2 is formed it reacts with the present NaH to form NaMgH3. In order to prove this assumption, an absorption measurement was performed at conditions of hydrogen pressure and temperature at which the MgH2 formation is thermodynamically impossible, but where NaMgH3 is stable. Based on the van’t Hoff equation the equilibrium H2 pressure of MgH2 was calculated to be 47.46 bar at 450 °C (assuming an enthalpy of formation ∆HF equal to -74.4 KJ mol-113 and entropy of formation ∆SF of -135 J mol-1 K-113 for the range of temperatures between 314 and 576 °C). NaMgH3 was found to be thermally stable up to 500 °C at a hydrogen pressure of 10 bar,17 and consequently we performed a hydrogen absorption measurement under isothermal condition at 450 °C and 20 bar of hydrogen pressure of the as-milled 2NaH + MgB2 mixture (Figure 13). The initial phase of the absorption reaction was characterized by a fast hydrogen uptake until the hydrogen stored in the system reached a value of 2.5 wt %, and then the absorption continued reaching asymptotically a total amount of stored hydrogen equal to 5.45 wt % after 17.5 h (note that since at these experimental conditions MgH2 should not be formed, a maximal H2 capacity of 6 wt % is expected). During cooling, at a temperature of 360 °C the hydrogen absorption starts again charging a further 0.1 wt % of hydrogen. The PXD analysis of the material as above synthesized is shown in Figure 14. The diffraction pattern shows the presence of NaBH4, Mg, NaH, MgB2, and only of a
Figure 14. PXD patterns of the 2NaH + MgB2 after absorption under 20 bar of hydrogen pressure at 450 °C (λ ) 1.541 84 Å).
tiny amount of NaMgH3. We suppose that this NaMgH3 is only formed during cooling, where the system crosses conditions of temperature and pressure favorable for MgH2 formation. (At 360 °C, peq(MgH2) ) 8.44 bar of H2, assuming an enthalpy of formation ∆HF equal to -74.4 KJ mol-113 and entropy of formation ∆SF of -135 J mol-1K-1 13) These results underline the supposition that NaMgH3 is formed by reaction between MgH2 and NaH; moreover, they also explain the reason why it has never been possible to observe MgH2 among the final absorption products. In addition, the above discussion justifies the lack of NaMgH3 formation in the hydrogen absorption measurement performed at 5 bar of hydrogen pressure. In fact, peq(MgH2) ) 5 bar at 350 °C; thus, MgH2 formation was possible only in a short range of temperature (and time), which very likely was not enough to lead to a significant formation of MgH2. Consequently, since MgH2 is not formed, it is not possible to observe NaMgH3 among the final products, even if it is known to be stable under these temperature and hydrogen pressure conditions.17 On the basis of the combined Sievert’s and ex situ X-ray diffraction measurements, it is possible to state that the formation of NaMgH3 has a detrimental effect on the absorption reaction: on one hand it removes NaH from the system, which is not available for forming NaBH4, and on the other hand it appears to slow down and finally hamper the reaction process. Recently Shane et al.18 demonstrated that atomic hydrogen can easily diffuse into the NaMgH3 structure; however, the formation of such hydride
Pressure Effect on the 2NaH + MgB2 Reaction
J. Phys. Chem. C, Vol. 114, No. 49, 2010 21823 of the molten salt mixture is still not yet understood, and further work is needed to elucidate it.
Figure 15. Series of SR-PXD patterns of the 2NaH + MgB2 system heated at 50 bar of hydrogen pressure from RT to 300 °C and kept under isothermal condition (5 °C/min, λ ) 1.097 19 Å).
might affect the diffusion of species different from H. A likely scenario is that NaMgH3 forms around MgB2 particles, segregating them and consequently disturbing the borohydride formation by blocking any direct contact between NaH and MgB2, with a mechanism mirroring the one explained by Stander19 and Zdhanov et al.20 for the MgH2 formation. This would explain the reason why absorption at lower H2 pressures (where NaMgH3 formation is minimized) yields higher amounts of hydrogen uptake. Moreover, the observed formation of amorphous boron (and of the other unidentified phases) along with NaMgH3 represents a severe issue for the reversibility of the system. In fact, upon cycling the accumulation of unreactive amorphous boron would lead to a drastic reduction of the overall hydrogen storage capacity of the system. Therefore, in order to avoid this undesired capacity contraction, the absorption process has to be performed at temperature and hydrogen pressure conditions under which MgH2 and consequently NaMgH3 are not stable. A last issue that must be addressed in this work is the presence of free Mg among the final absorption products. As observed in Figure 5, the appearance of free Mg with the consequent stop of the growth of NaMgH3 phase follows the formation of the NaH/NaBH4 molten phase at 350 °C. In order to understand if the appearance of free Mg is related to the formation of this liquid phase, we performed an in situ SR-PXD measurement at 50 bar of H2 restricting the temperature to 300 °C (Figure 15). In fact at this temperature all the reactants and products should remain in the solid phase. As already observed for the previous SR-PXD measurement performed under 50 bar of pressure, the formation of the unknown phase is found to take place at roughly 270 °C. No further phase formation is observed until the temperature reaches 300 °C. After reaching the constant temperature of 300 °C, the formation of the NaMgH3 starts, followed later by the formation of NaBH4. Due to the low temperature, the formation of the amorphous background cannot be observed, hence confirming an “all solid state” reaction. Any of the free Mg reflection cannot be observed. This clearly demonstrates that the NaH/NaBH4 molten phase represents a barrier for the hydrogen diffusion into the system. However, the hydrogen isolating mechanism
Conclusions In this work the hydrogen absorption reaction mechanism of the system 2NaH + MgB2 was deeply characterized as a function of the applied hydrogen pressure. As a main result it was observed that the reaction paths strongly depend on the applied conditions. In case of hydrogen pressures of 50 and 25 bar, the formation of NaBH4 was preceded by the formation of an unknown hydride phase, NaMgH3, amorphous boron, and unidentified B-containing phases, whereas the application of a hydrogen pressure of 5 bar led to the direct formation of NaBH4. A common finding for all the absorption experiments was the formation of a NaH-NaBH4 molten phase at temperatures higher than 350 °C. For the NaMgH3 and the NaH-NaBH4 molten phases, boundary conditions and the effect of their formation on the absorption reaction were also discussed. Although for the experimental conditions investigated in this work the complete conversion of 2NaH + MgB2 into 2NaBH4 + MgH2 could not be achieved, the possibility to reversibly store up to 6 wt % of hydrogen, under temperature and pressure conditions which do not allow the NaMgH3 formation, was reported. Acknowledgment. This work was supported by the European Community under MRTN-Contract “Complex Solid State Reactions for Energy Efficient Hydrogen Storage” (MRTN-CT2006-035366). We are grateful to the Carlsberg Foundation and the Danish Research Council for Nature and Universe (Danscatt). We thank the Servei de Ressonancia Magnetica Nuclear RMN at UAB for their technical assistance. Financial support from 2009-SGR-1292 is also acknowledged. M.D.B. was partially supported by an ICREA ACADEMIA award. References and Notes (1) Barkhordarian, G.; Klassen, T.; Bormann, R. WO 2006/063627 A1. (2) Dornheim, M.; Eigen, N.; Barkhordarian, G.; Klassen, T.; Bormann, R. AdV. Eng. Mater. 2006, 8, 377. (3) Barkhordarian, G.; Klassen, T.; Dornheim, M.; Bormann, R. J. Alloys Compd. 2007, 440, L18. (4) Vajo, J. J.; Skeith, S. L.; Mertens, F. J. Phys. Chem. B 2005, 109, 3719. (5) Vajo, J. J.; Mertens, F. O.; Skeith, S. L.; Balogh, M. P. WO 2005/ 097671. (6) Garroni, S.; Pistidda, C.; Brunelli, M.; Vaughan, G. B. M.; Surinach, S.; Baro, M. D. Scr. Mater. 2009, 60 (12), 1129. (7) Mao, J. F.; Yu, X. B.; Guo, Z. P.; Liu, H. K.; Wu, Z.; Ni, J. J. Alloys Compd. 2009, 479, 619. (8) Czujko, T.; Varin, R. A.; Wronski, Z.; Zaranski, Z.; Durejko, T. J. Alloys Compd. 2007, 427, 291. (9) Clausen, B. S.; Steffensen, G.; Fabius, B.; Villadsen, J.; Feidenhansl, R.; Topsoe, H. J. Catal. 1991, 132, 524. (10) Rodriguez, J. A.; Hanson, J. C.; Frenkel, A. I.; Kim, J. Y.; Perez, M. J. Am. Chem. Soc. 2002, 124, 346. (11) Bosenberg, U.; Doppiu, S.; Mosegaard, L.; Barkhordarian, G.; Eigen, N.; Borgschulte, A.; Jensen, T. R.; Cerenius, Y.; Gutfleisch, O.; Klassen, T.; Dornheim, M.; Bormann, R. Acta Mater. 2007, 55, 3951. (12) http://www.esrf.eu/computing/scientific/FIT2D/. (13) Zeng, K.; Klassen, T.; Oelerich, W.; Bormann, R. Int. J. Hydrogen Energy 1999, 24, 989. (14) San-Martin, A.; Manchester, F. D. J. Phase Equilib. 1990, 11, 287. (15) Stasinevich, D. S.; Egorenko, G. A. Russ. J. Inorg. Chem. 1968, 13, 341. (16) Ikeda, K.; Nakamori, Y.; Orimo, S. Acta Mater. 2005, 53, 3453. (17) Ikeda, K.; Kogure, Y.; Nakamori, Y.; Orimo, S. Scr. Mater. 2005, 53, 319. (18) Shane, D. T.; Corey, R. L.; Bowman, R. C.; Zidan, R.; Stowe, A. C.; Hwang, S. J.; Kim, C.; Conradi, M. S. J. Phys. Chem. C 2009, 113, 18414. (19) Stander, C. M. Z. Phys. Chem. (Frankfurt) 1977, 104, 229. (20) Zdhanov, V. P.; Krozer, A.; Kasemo, B. Phys. ReV. B 1993, 47, 11044.
JP107363Q
