Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.issp.ac.ru/lhpp/PapersEfimchenko/8.pdf Äàòà èçìåíåíèÿ: Fri Mar 15 17:49:47 2013 Äàòà èíäåêñèðîâàíèÿ: Fri Feb 28 03:04:42 2014 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: ñ ð ð ñ ñ ï ï ð ï ï ñ ñ ñ ð ð ð ð ð ï ï ð ð ð ï ï ð

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier's archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright


Author's personal copy
Solid State Communications 154 (2013) 77 ­80

Contents lists available at SciVerse ScienceDirect

Solid State Communications
jo urnal homep age : www.elsevier.com/locate/ssc

Raman scattering study of a-MgH2 and g-MgH

2

M.A. Kuzovnikov a,n, V.S. Efimchenko a, E.V. Filatov a, A.A. Maksimov a, I.I. Tartakovskii a, A.J. Ramirez-Cuesta b
a b

Institute of Solid State Physics RAS, 2 Academician Ossipyan str., Chernogolovka, Moscow District 142432, Russian Federation ISIS Facility, Rutherford Appleton Laboratory, Chilton, Didcot, Oxon, OX11 0QX, UK

article info
Article history: Received 18 July 2012 Received in revised form 3 September 2012 Accepted 27 September 2012 by M. Grynberg Available online 3 October 2012 Keywords: A. Hydrides B. High pressure C. Raman scattering

abstract
Two modifications of MgH2 have been studied by Raman spectroscopy: a-MgH2 with the P42/mnm space group and g-MgH2 with the Pbcn space group. The latter was prepared from a-MgH2 by exposing it to a pressure of 5.6 GPa at 470 1C for 1.5 h. A comparison of the experimental spectra with ab-initio calculations allowed identification of a few phonon modes. A broad feature in the Raman spectrum of a-MgH2 in the range 1470­1790 cm þ 1 was identified as the B2g phonon mode. A Raman spectrum of g-MgH2 consists of five discernible peaks at 186 cm þ 1, 313 cm þ 1, 509 cm þ 1, 660 cm þ 1, 706 cm þ 1 and of three broad features in the regions 890­980 cm þ 1, 1010­1220 cm þ 1 and 1240­1430 cm þ 1. The peaks are identified as the 1Ag (186 cm þ 1), 3B3g (509 cm þ 1), 2Ag (660 cm þ 1) and 3B1g (706 cm þ 1) phonon modes. & 2012 Elsevier Ltd. All rights reserved.

1. Introduction MgH2 is a promising material for hydrogen storage. It has several polymorphs, the tetragonal a-MgH2 with the P42/mnm space group being the most common. Despite the high hydrogen capacity of 7.7 wt% of this compound, its industrial application as a hydrogen storage material is still rather limited due to the high decomposition temperature, slow kinetics of re-hydrogenation and irreversible oxidation in contact with oxygen or water vapor. The first two limiting factors stem from the relatively strong interaction between magnesium and hydrogen in the compound. Application of high pressure transforms a-MgH2 to more dense polymorphs: orthorhombic g, space group Pbcn [1]; cubic b, space group Pa-3 [2]; orthorhombic HP1, space group Pbc21 [3] (previously identified as d' with the Pbca space group [2]) and orthorhombic HP2, space group Pnma [3]. The tetragonal a-MgH2 phase has a rutile-type structure with 2 formula units per unit cell, and g-MgH2 has a a-PbO2 type structure with 4 formula units per unit cell. The atoms in the a and g phases have similar nearest environments: each magnesium atom is surrounded by an octahedron of hydrogen atoms, and each hydrogen atom belongs to three such octahedra (see Fig. 1). The g phase has a density that is 1.6% higher than a [1]. Other phases are considerably more dense and their synthesis requires the pressure higher than 10 GPa. Our

experiments suggest that the g phase becomes more stable than the a phase at pressures above 1.5 GPa at T ¼ 700 1C. Another way to transform the a phase into g is by ball milling [4]. For better understanding of hydrogen­magnesium interatomic forces a study of the lattice dynamics of different phases of MgH2 is needed. The best experimental methods to do this are inelastic neutron scattering, IR and Raman light scattering. Raman spectra of a-MgH2 were measured in Refs. [5] and [6] up to 1600 cm þ 1. In these studies, three peaks at 300, 950 and 1280 cm þ 1 were observed and identified as the B1g, Eg and A1g phonon modes, correspondingly. A Raman study of the g phase was carried out only in-situ in a diamond anvil cell in Ref. [3], and two features were observed at 200 and 650 cm þ 1. Theoretical calculations of the lattice dynamics and Raman frequencies of a-MgH2 were performed in Refs. [7] and [8]. The present paper reports on an experimental Raman scattering study of the a and g phases of MgH2. The experiment is supplemented with ab-initio calculations aimed at identifying the observed atomic vibration modes.

2. Experimental details 2.1. X-ray powder diffraction A flat sample holder was made of a silicon single crystal and had a hole 7 mm in diameter and 0.5 mm in height encircled with an O-ring of vacuum grease. The powder samples were placed in

n

Corresponding author. Fax: × 7 4965228160. E-mail address: kuz@issp.ac.ru (M.A. Kuzovnikov).

0038-1098/$ - see front matter & 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ssc.2012.09.022


Author's personal copy
78 M.A. Kuzovnikov et al. / Solid State Communications 154 (2013) 77 ­80

copper capsule were electrically isolated from each other using mica sheets. The synthesis of g-MgH2 was performed at 5.6 GPa and 470 1C for 90 min. The high pressure chamber was further cooled to room temperature and disassembled. The sample was extracted out of the copper capsule under liquid nitrogen. The extracted sample was loaded to a hermetical vial, removed from the liquid nitrogen and transported to the Ar glove box, where it was placed onto the X-ray sample holder. The diffraction pattern of the sample is shown at the bottom of Fig. 2. The sample was found to contain 82 wt% g-MgH2, 14 wt% a-MgH2 and 3 wt% MgO. 2.3. Raman scattering study
Fig. 1. Crystal structures of a- and g-MgH2. The octahedra of H atoms surrounding Mg atoms are shown.


wt.% 86 12 2

exp exp-calc R p=13.9 R wp =18.42

Intensity, arb. units

MgH2 Mg(OH)2 Mg

0


wt.% 80 MgH2 14 MgH2 5 MgO

exp exp-calc R p =11.41 R wp=15.16

0 25 30 35 40 45 50 2, deg. 55 60 65

The MgH2 powder samples were placed into hermetically sealed glass cells to avoid any contact with air. The samples were irradiated with a 488-nm Ar × laser line (laser power $ 10 mW) at room temperature using a backscattering geometry. The laser beam was focused to a spot of approximately 2 mm in size, while the samples consisted of particles with the characteristic sizes of about 10 mm. That allowed to select various particles of the sample. The spectra were recorded by a Dilor XY-500 spectrometer equipped with a micro-Raman setup and a Peltier cooled chargecoupled device (CCD) detector. Several spectral ranges were covered, and the corresponding spectra were glued using standard procedures. The recorded spectra consisted of a Raman signal and luminescence background, which intensity increased with the frequency. Luminescence background was subtracted from the spectrum of g-MgH2 shown at the bottom of Fig. 3. In the case of a-MgH2, the baseline was not subtracted from the spectrum shown at the top of Fig. 3.

Fig. 2. (Colour Online) X-ray powder diffraction patterns of a-MgH2 (top) and g-MgH2 (bottom) samples. Room temperature, Cu Ka radiation. Black curves represent the experimental patterns, and the differences between experimental and calculated patterns are shown by red curves. Residual functions are qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi P P P ÏIobs þIcalc ÷2 P Rp Ï%÷ ¼ 9Iobs þIcalc 9= Iobs and Rwp Ï%÷ ¼ = Iobs . Iobs

3. Calculations A factor group analysis shows that a-MgH2 has the following phonon modes at the center of the Brillouin zone: þ à ÏA2u × Eu ÷acoustic × ÏA2u × 3Eu ÷IR × A1g × B1g × B2g × Eg Raman þ à × A2g × 2B1u silent The Eu and Eg phonon modes are doubly degenerate. The analysis for g-MgH2 gives the following phonon modes: ÏB1u × B2u × B3u ÷acoustic × Ï4B1u × 3B2u × 4B3u ÷IR þ à × 4Ag × 5B1g × 4B2g × 5B3g Raman × Ï4Au ÷silent All optic modes are non-degenerate in the case of the g phase. Due to the presense of inversion center in both a and g phases, the Raman and infrared activites are incompatible with each other for any phonon mode. The vibrational properties of MgH2 in both polymorphic phases were calculated by density-functional theory and the plane-wave pseudopotential method as implemented in the CASTEP code [11,12]. Pseudopotentials were of the optimized norm-conserving variety with a plane-wave cutoff of 990 eV. Calculations were performed under the PBE approximation to exchange and correlation. Brillouin-zone integration was performed according to the Monkhorst­Pack scheme with a 4 á 4 á 3 for a-MgH2 and 4 á 4 á 3 for g-MgH2, which gave convergence of all modes to a precision of better than 3 cm þ 1. Density functional perturbation theory was used to calculate the dynamical matrices [11]. Raman intensities were calculated using the method as described in [13]. The results of the calculation are shown in Fig. 3. To distinguish between different phonon modes with the same character

this hole and hermetically sealed with a mylar film in a glove box under an Ar atmosphere. The samples were examined with a Siemens D500 powder X-ray diffractometer using CuKa radiation. The obtained X-ray patterns were analyzed using the POWDERCELL2.4 software [9]. The VESTA software [10] was used for rendering the crystal structures. 2.2. High pressure­High temperature synthesis MgH2 powder was purchased from Aldrich. An X-ray diffraction pattern of the initial MgH2 powder is shown at the top of Fig. 2. The powder contained 91 wt% a-MgH2, 7 wt% Mg(OH)2 and 2 wt% Mg. A copper capsule 9 mm in diameter and 9 mm in height was filled with 0.2 g MgH2 powder under an Ar atmosphere, covered with a copper disc and hermetically sealed with a lead plug pressed inside the capsule. The capsule was then placed in the working zone of a toroid-type high pressure chamber. The pressure in the chamber was calibrated against the oil pressure in a hydraulic press using well-defined phase transitions in bismuth and thallium determined with a four-point electroresistance method. The accuracy of the pressure measurement was 7 0.2 GPa. The heater was made of 14 nichrome wires encompassing copper capsule. The temperature during the high pressure experiment was controlled by a chromel-alumel thermocouple with an accuracy of 7 10 1C. The heater, thermocouple and


Author's personal copy
M.A. Kuzovnikov et al. / Solid State Communications 154 (2013) 77 ­80 79

5 B
1g

4 Raman intensity, s
-1

3

2 E 1
g

A1g

B2g

0 0 5 4 3 Raman intensity, s 2 1 0 -1 -2 -3 -4 -5 0 500 1000 Raman shift, cm 1500
-1 -1

500

1000 Raman shift, cm

1500
-1



The Raman spectrum of g-MgH2 (the bottom of Fig. 3) consists of intense narrow peaks at 186 and 660 cm þ 1, which were earlier seen in Ref. [3]; weak peaks at 509 cm þ 1, 313 cm þ 1 and 706 cm þ 1 (listed in the order of decreasing intensity), and three broad bands in the 890­980 cm þ 1, 1010­1220 cm þ 1 and 1240­1430 cm þ 1 ranges. Several phonon modes are predicted by our calculations to be near 186 cm þ 1 peak, the most intense being 1Ag. In the neighborhood of the 660 and 706 cm þ 1 peaks, there are two modes in the calculation, 2Ag and 3B1g, the first one being much more intense. This enables us to do the assignment for these peaks. The 509 cm þ 1 peak can be assigned to the 3B3g mode, because this is the only mode in its vicinity. There are two modes, 1B2g and more intense 2B3g, near the 313 cm þ 1 peak. However, we cannot assign this peak to 2B3g mode with certainty, because the B1g mode of a-MgH2 contributes to the same energy range (the top of Fig. 3) and the sample was contaminated with a-MgH2. An inspection of the broad bands in the spectrum shows that their positions are the same as in the Raman spectrum of a-MgH2, which could be evidence of contamination. The profile of the 890­980 cm þ 1 band, however, cannot be explained in this way. It has a two-peak shape, the position of the higher peak being the same as that of the Eg mode in a-MgH2.Takinginto account that there is a 3Ag mode with a comparable intensity in the vicinity, one can speculate whether or not the lower peak at 915 cm þ 1 represents this phonon mode.

5. Conclusion

1Ag 1B1g 2Ag 2B1g 1B3g 3B3g 1B2g 2B3g

4B3g 3B1g 4B1g 3B 4Ag 2g 5B1g 2B2g 3Ag 5B3g 4B2g

Raman spectra of a-MgH2 and g-MgH2 have experimentally been studied and compared with ab-initio calculations that allowed us to assign the phonon modes to some experimental peaks. Despite the similar nearest-neighbor environment of the Mg and H atoms in a-MgH2 and g-MgH2, the Raman spectra of these phases are very different because of the different crystal symmetries. A work is in progress to study the interatomic interactions and phonon properties in these modifications of MgH2 with inelastic neutron scattering, which is a symmetry insensitive method in contrast to Raman scattering.

Fig. 3. (Colour Online) Raman spectra of a-MgH2 (top) and g-MgH2 (bottom). The experimental spectra are plotted by the blue curves and the calculated spectra by the red ones. The strong luminescence background rising with the frequency was subtracted from the experimental spectrum of g-MgH2; the experimental spectrum of a-MgH2 is shown without modifications. The calculated spectrum of g-MgH2 is plotted downwards.

Acknowledgements This work was supported by the Programme ``The Matter under High Energy Density'' of the Russian Academy of Sciences.

in the case of g-MgH2, we label them with a number prefix in the order of the increase in the calculated Raman shift.

Appendix. Supporting information Supplementary data associ ated with th is article c a n be found in the online version at http://dx.doi.org/10.1016/j.ssc. 2012.09.022.

4. Results and discussion Animations demonstrating the atomic movements in the phonon modes considered in this Section are presented as Supplementary material. The basic features of the Raman spectrum of a-MgH2 (the top of Fig. 3) are the same as in Ref. [6]: the peak at 310 cm þ 1 is narrow and the peaks at 945 and 1270 cm þ 1 are broad and the 1270 cm þ 1 peak has a long shoulder towards lower Raman shifts. The origin of such a broadening is unclear. In addition to these previously known features, a broad band in the 1470­1790 cm þ 1 range is seen in the spectrum. Its intensity is of the same order of magnitude as the intensities of other features. Our first-principle calculation predicts a B2g phonon mode in this range with the Raman scattering intensity similar to that of other modes. On these grounds, we assign this feature to the B2g phonon mode.

References
¨ [1] M. Bortz, B. Bertheville, G. Bottger, K. Yvon, J. Alloys Compd. 287 (1999) L4­L6. ° [2] P. Vajeeston, P. Ravindran, B.C. Hauback, H. Fjellvag, A. Kjekshus, S. Furuseth, M. Hanfland, Phys. Rev. B 73 (2006) 224102. [3] T. Moriwaki, Yu. Akahama, H. Kawamura, S. Nakano, K. Takemura, J. Phys. Soc. Jpn. 75 (7) (2006) 074603. [4] H.G. Schimmel, M.R. Johnson, G.J. Kearley, A.J. Ramirez-Cuesta, J. Huot, F.M. Mulder, J. Alloys Compd. 393 (2005) 1­4. [5] J.R. Santisteban, G.J. Cuello, J. Dawidowski, A. Fainstein, H.A. Peretti, A. Ivanov, F.J. Bermejo, Phys. Rev. B 62 (2000) 37. [6] H.G. Schimmel, M.R. Johnson, G.J. Kearley, A.J. Ramirez-Cuesta, J. Huot, F.M. Mulder, Mater. Sci. Eng. B108 (2004) 38­41. [7] N. Ohba, K. Miwa, T. Noritake, A. Fukumoto, Phys. Rev. B 70 (2004) 035102. [8] J. Lasave, F. Dominguez, S. Koval, M.G. Stachiotti, R.L. Migoni, J. Phys. Condens. Matter 17 (2005) 7133­7141.


Author's personal copy
80 M.A. Kuzovnikov et al. / Solid State Communications 154 (2013) 77 ­80

[9] W. Kraus, G. Nolze, J. Appl. Crystallogr. 29 (1996) 301­303. [10] K. Momma, F. Izumi, J. Appl. Crystallogr. 41 (2008) 653­658. [11] S.J. Clark, M.D. Segall, C.J. Pickard, P.J. Hasnip, M.J. Probert, K. Refson, M.C. Payne, Z. Kristallogr. 220 (5-6) (2005) 567.

[12] K. Refson, S.J. Clark, P.R. Tulip, Phys. Rev. B 73 (2006) 155114. [13] R. Gremaud, Z. Lodziana, P. Hug, B. Willenberg, A.-M. Racu, J. Schoenes, ¨ A.J. Ramirez-Cuesta, S.J. Clark, K. Refson, A. Zuttel, A. Borgschulte, Phys. Rev. B 80 (2009) 100301. (R).