Nuclear Magnetic Resonance Studies of Hydrogen Storage Materials

By: Dr. Robert C. Bowman, Jr. and Dr. Son-Jong Hwang, Material Matters Volume 2 Article 2


California Institute of Technology


Solid-state nuclear magnetic resonance (NMR) methods have been used to characterize metal hydrides and other hydrogen storage materials for over fifty years. Until recently, attention was usually focused on assessing structural properties, electronic parameters, and diffusion behavior of the hydride phases of metals and alloys using mostly transient NMR techniques or low-resolution spectroscopy.1-3 This interest was stimulated by the excellent resonance properties of the three common hydrogen isotopes (i.e., 1H, 2D, and 3T) and some other host nuclei (e.g., 23Na, 45Sc, 51V, 89Y, 93Nb, and 139La), which allowed detailed evaluation of local interactions between hydrogen and metal atoms, thus complementing diffraction and thermochemical measurements. In particular, the NMR relaxation times are extremely useful to assess diffusion processes over very wide ranges of hydrogen mobility in crystalline and amorphous phases.1-3

In order to achieve the challenging weight and volume goals set for the hydrogen storage systems for automotive applications by the U.S. Department of Energy,4 only the lightest elements (i.e., Li, B, C, N, Na, Mg, Al, Si) can be considered as a basis for such systems. The development and evaluation of hydride-based materials combining various light metal hydrides and transition metal catalysts are the focus of research efforts by numerous international research groups. By implementing advanced solid-state NMR techniques,5 such as Magic Angle Spinning (MAS), and multi-quantum (MQ) MAS NMR in addition to traditional measurements of nuclear relaxation times,1-3 hydride phases of light elements can be investigated more efficiently. Solid-state NMR techniques also offer a better insight into complicated relationships between various processes accompanying the formation of hydride phases and their transformations, including reaction kinetics, reversibility, and the role of catalysts.

To illustrate the current state of solid-state NMR research in the area of hydrogen storage, we will discuss some recent examples typical for the field.

back to top

Aluminum-Hydrogen Systems

With hydrogen content of about 10 wt%, alane (i.e., AlH3) and alkali metal alanates (i.e., LiAlH4 and NaAlH4) are being widely investigated as potential hydrogen storage materials. Although some additives such as Ti-compounds can substantially enhance the reaction rates for alanates 6 and alane based materials,7 significant issues remain with respect to the reversibility and stability of these hydride phases. Increasingly, solid-state NMR techniques are being used to evaluate phase compositions and transformations in alanates8-15 and alane.16-18
Figure 1 presents static and 1H MAS NMR spectra of a-AlH3 obtained at the Caltech Solid-State NMR Facility.

Figure 1. Static and 1H MAS NMR spectra of a phase aluminum hydride (a- AlH3).

Figure 1. Static and 1H MAS NMR spectra of a phase aluminum hydride (a- AlH3) obtained at the resonance frequency of 500.23 MHz. (bottom) static (0 kHz) d1= 300 s, (middle) MAS (14.5 kHz), d1=300 s, (top) MAS (35 kHz), d1=300 s, where d1 is the repetition delay time during signal averaging.

Static NMR spectra of solids often reveal featureless broadening due to anisotropic interactions between atoms in the solid-state. Such interactions include dipole-dipole couplings, chemical shift anisotropies and quadrupole couplings for nuclei with I.1/2. High-speed (i.e., . 5 kHz) rotation of a sample aligned at the specific “magic angle” of 54.7° with respect to the direction of the external magnetic field averages out most of these broadening sources to yield highly resolved narrow and isotropic lines.5 In Figure 1, the very broad component in the static spectrum (0 kHz) is from the immobile protons in the hydride phase while a sharper peak is attributed17 to the molecules of gaseous hydrogen trapped in the solid (~4% of the total hydrogen content). Although MAS narrows the broad peak, intense residual spinning side bands are still observed in the spectrum even at the very high rotation speed of 35 kHz. These side bands are the result of very strong dipolar interactions that cannot be removed completely.

A series of spinning side bands are also observed in the 27Al MAS NMR spectra shown in Figure 2.

Figure 2. 27Al MAS NMR spectra of a and g phases of AlH3, NaAlH4 and Na3AlH6 obtained at the resonance frequency 130.35 MHz with strong 1H decoupling.

Figure 2. 27Al MAS NMR spectra of a and g phases of AlH3, NaAlH4 and Na3AlH6 obtained at the resonance frequency 130.35 MHz with strong 1H decoupling.

They arise from various quadrupolar interactions5 in a and g-AlH3, or other aluminum-based hydrides. However, while the intrinsic small range of chemical shifts for protons along with large residual dipolar interactions limits the information that can be obtained from even fast-spinning 1H MAS NMR spectra, the 27Al MAS NMR is capable of providing an additional insight into the local symmetry and bonding in the material. The isotropic chemical shifts (diso) and quadrupolar parameters5 CQ and h for several different AlH3 and alanate phases are summarized in Table 1. It is easy to see that 27Al MAS NMR technique allows for a clear and distinct identification of tetrahedral and octahedral coordination sites in various types of aluminumbased hydride materials.

Table 1. The peak shifts and quadrupole coupling parameters for 27Al spectra from AlH3, the Li and Na alanate phases, and Al metal.

Table 1. The peak shifts and quadrupole coupling parameters for 27Al spectra from AlH3, the Li and Na alanate phases, and Al metal.

aCQ is the quadrupole coupling is the asymmetry parameter.

27Al MAS NMR also allows detection of aluminum metal and its oxide phases in aluminum-based hydrides during their processing and monitoring both chemical processes and contaminations.8-18 For example, oxide phases14 in a-AlH3 can be clearly identified in the spectrum shown in Figure 3.

Figure 3. Two-Dimensional (2D) MQMAS 27Al spectrum of a-AlH3 (wr=35 kHz) shows two weaker peaks at ~40 ppm and ~65 ppm from 5-fold and 6-fold Al-O sites in addition to the peak at 6 ppm from the hydride.

Figure 3. Two-Dimensional (2D) MQMAS 27Al spectrum of a-AlH3 (wr=35 kHz) shows two weaker peaks at ~40 ppm and ~65 ppm from 5-fold and 6-fold Al-O sites in addition to the peak at 6 ppm from the hydride.

NMR analysis of quadrupole nuclei such as 27Al, 23Na, 11B, etc. can benefit from the MQMAS NMR technique developed by Frydman19 and others20,21 for obtaining highly resolved isotropic line components free of quadrupole interactions. The MQMAS method is especially appropriate for quadrupole nuclei because it eliminates the second order quadrupole interactions, which cannot be removed by the MAS NMR alone. Figure 3 shows an example of 27Al 2D MQMAS NMR spectrum that reveals the presence of oxide phases, which the corresponding 1D 27Al MAS NMR spectrum alone cannot identify unambiguously. It is worth noting that recently Herberg, et al.14 have successfully used 27Al MQMAS to evaluate Al-oxide phases in Ti-doped sodium alanate.

back to top

MAS NMR Spectroscopy of Deuterides

Information about the specific locations and symmetries of hydrogen in a hydrogen storage material is highly desirable for determining the detailed structures and dynamic behavior of the hydride phase. Unfortunately, due to narrow ranges of chemical shifts and extensive residual dipolar proton-proton interactions, 1H MAS NMR is not always a best solution for such investigations. However, dipolar broadening as well as quadrupolar contributions can be significantly reduced or even eliminated for the deuterons (2D) in deuteride phases, which makes it possible to observe local symmetries and multiple site occupancies directly.22-24

Figure 4 presents the 2D MAS NMR spectra of two different ZrNiDx phases with two distinctly different lattice sites for each phase.23 These NMR results are strongly supported by neutron powder diffraction studies of the b-ZrNiD0.88,25 which, contrary to the prior assignment, also reveal two locations for D-atoms. This allows for a better understanding23 of diffusion processes in both phases of ZrNiD0.88.

Figure 4. Rigid-lattice 2D MAS NMR spectra (fo=76.79 MHz) for the g-ZrNiD2.99 and the b-ZrNiD0.

Figure 4. Rigid-lattice 2D MAS NMR spectra (fo=76.79 MHz) for the g-ZrNiD2.99 (upper) and the b-ZrNiD0.88 (lower). Spinning sidebands are not shown in both spectra. The fits to these spectra yield two distinct resonances for each phase that correspond to the indicated D sites as described in Ref. 23.

In another example, Adolphi and co-workers have used 2D MAS NMR techniques to differentiate the occupancy and mobility of deuterons in the tetrahedral and octahedral sites in YDx phases and showed that this approach can be used on different classes of deuterides.22,24

back to top


A few examples showing the versatility of multinuclear and multidimensional NMR to evaluate the structure and behavior of metal hydrides were presented in this short review. Additionally, nuclear relaxation times have been used1-3 to characterize diffusion processes and mechanisms for many classes of metal hydrogen systems. Recent studies, describing diffusion assessments by means of relaxation times measurements, include reports on LaNi5Hx,24 Mg-based hydrides,24 NaAlH4,26 and the ZrNiHx/ZrNiDx phases.27 Thus, various NMR approaches can be applied to various hydrogen storage media being investigated now and coming into the picture in the future. Immediate candidates include metal alanates, borohydrides, metal amides, and destabilized lithium and magnesium hydrides.28 The combinations of MAS NMR and MQMAS NMR techniques should provide much greater insights into the local structures, bonding, and dynamics of these materials.

back to top


  1. Cotts, R. M. in Hydrogen in Metals I: Basic Properties; Alefeld, G.; J. Volkl (Eds.); Springer: Berlin-Heidelberg, 1978; p. 227.
  2. Richter, D.; Hempelmann, R.; Bowman, Jr., R. C. in Hydrogen in Intermetallic Compounds II – Surfaces and Dynamic Properties, Applications, Schlapbach, L. (Ed.); Springer-Verlag: Berlin, 1992; p. 97.
  3. Barnes, R. G. in Hydrogen in Metals III: Properties and Applications; Wipf, H. (Ed.); Springer-Verlag: Berlin; 1997; p. 93.
  4. Chandra, D.; Reilly, J. J.; Chellappa, R. Journal of Metals (JOM) 2006, 58, 26.
  5. MacKenzie, K. J. D.; Smith, M. E. Multinuclear Solid- State NMR of Inorganic Materials; Pergamon: Amsterdam, 2002.
  6. Bogdanovicˇ, B.; Sandrock, G. MRS Bulletin 2002, 27, 712.
  7. Graetz, J.; Reilly, J. J. J. Phys. Chem. B 2005, 109, 22181.
  8. Tarasov, V. P.; Bakum, S. I.; Privalov, V. I.;. Muravlev, Yu. B; Samoilenko, A. A. Russ. J. Inorg. Chem. 1996, 41, 1104.
  9. Tarasov, V. P.; Bakum, S. I.; Kuznetsova, S. F. Russ. J. Inorg. Chem. 1997, 42, 694.
  10. Balema, V. P.; Wiench, J. W.; Dennis, K. W.; Pruski, M.; Pecharsky, V. K. J. Alloys Compd. 2001, 329, 108.
  11. Bogdanovicˇ, B.; Felderhoff, M.; Germann, M.; Hartel, M.; Pommerin, A.; Schuth, F.; Weidenthaler, C.; Zibrowius, B. J. Alloys Compd. 2003, 350, 246.
  12. Wiench, J. W.; Balema, V. P.; Pecharsky, V. K.; Pruski, M. J. Solid State Chem. 2004, 177, 648.
  13. Majzoub, E. H.; Herberg, J. L.; Stumpf, R.; Spangler, S.; Maxwell, R. S. J. Alloys Compd. 2005, 394, 265.
  14. Herberg, J. L.; Maxwell, R. S.; Majzoub, E. H. J. Alloys Compd. 2006, 417, 39.
  15. Mamatha, M.; Bogdanovicˇ, B.; Felderhoff, M.; Pommerin, A.; Schmidt, W.; Schuth, F.; Weidenthaler, C. J. Alloys Compd. 2006, 407, 78.
  16. Zogal, O. J.; Stalinski, B.; Idziak, S. Z. Physk. Chem. N.F. 1985, 145, 167.
  17. Hwang, S.-J.; Bowman, R. C.; Graetz, J. Reilly, J. J. In: Wang, J. C. F.; Tumas, W.; Rougier, A.; Heben, M. J.; Akiba, E. (Eds), Hydrogen Storage Materials, Mater. Res. Soc. Symp. Proc. Warrendale, PA, 2006, Vol 927E, 0927-EE03-03.
  18. Hwang, S.-J.; Bowman, Jr., R. C.; Graetz, J.; Reilly, J. J.; Langley, W.; Jensen, C. M. J. Alloys Compd. 2007, in press.
  19. Frydman, L.; Harwood, J. S. J. Am. Chem. Soc. 1995, 117, 5367.
  20. Amoureux, J. P.; Fernandez, C. Solid State NMR 1998, 10, 211.
  21. Amoureus, J. P.; Pruski, M. In: Grant, D. M; Harris, R. K. (Eds) Encyclopedia of NMR, John Wiley & Sons: Chichester, UK, 2002, 9, 226.
  22. Adolphi, N. L.; Balbach, J. J.; Conradi, M. S.; Markert, J. T.; Cotts, R. M.; Vajda, P. Phys. Rev. B 1996, 53, 15054.
  23. Bowman, Jr., R. C.; Adolphi, N. L.; Hwang, S.-J.; Kulleck, J. G.; Udovic, T. J.; Huang, Q.; Wu, H. Phys. Rev. B 2006, 74, 184109.
  24. Conradi, M. S.; Mendenhall, M. P.; Ivancic, T. M.;. Carl, E. A; Browning, C. D.; Notten, P. H. L.; Kalisvaart, W. P.; Magusin, P. C. M. M.; Bowman, Jr., R. C.; Hwang, S.-J.; Adolphi, N. L. J. Alloys Compd. 2007 (in Press).
  25. Wu, H.; Zhou, W.; Udovic, T. J.; Rush, J. J.; Yildirim, T.; Huang, Q.; Bowman, Jr., R. C. Phys. Rev. B 2007, 75, 064105.
  26. Majer, G.; Stanik, E.; Valiente Banuet, L. E.; Grinberg, F.; Kircher, O.; Fichtner, M. J. Alloys Compd. 2005, 738, 404.
  27. Browning, C. D.; Ivancic, T. M.; Bowman Jr., R. C.; Conradi, M. S. Phys. Rev. B 2006, 73, 134113.
  28. Vajo, J. J.; Mertens, F.; Ahn, C. C.; Bowman, Jr., R. C.; and Fultz, B. J. Phys. Chem. B 2004, 108, 13977.

back to top

Related Links