Crystallography at SLS, SINQ and SNBL

J. Friso van der Veen, Swiss Light Source, PSI Villigen and ETH Zürich,
Jürg Schefer, Laboratory for Neutron Scattering, PSI Villigen

Each year, the United Nations’ Organization pays special attention to a field of great importance to society. Scientific topics were repeatedly in their focus: 2010 the biological diversity, 2011 chemistry, and next year, 2014, crystallography. The opening ceremony for the International Year of Crystallography¹ will be held on January 20 and 21, 2014 (IYCr2014) at the UNESCO headquarters in Paris. The International Union of Crystallography takes the lead in the organization of IYCr2014. The activities in this year complement the celebration in 2012 of the 100^th anniversary of the discovery of Bragg’s Law and pay tribute to the many Nobel prizes awarded to crystallographic research. Below we provide a selection of recent results obtained at the large facilities of Paul Scherrer Institute, PSI (SINQ and Swiss Light Source) and at the Swiss-Norwegian beamline of the European Synchrotron Radiation Facility, Grenoble, France.

1 Swiss activities for the UN year of crystallography

Fig. 1: Working principle of a Li⁺/Na⁺ battery.

Fig: 2. Na⁺ diffusion paths in Na_0.7CoO₂ . Left panels: Na-Na distances at different temperatures in one of the Na layers (distortions enlarged). Right panels: residual scattering between the Na ions as determined from Fourier differences. The Na1/Na2 labels indicate the positions of the Na⁺ ions in the layer. The paths chosen by the Na⁺ ions during the diffusion process appear as red spots. At low temperatures (T < 15 °C) no red spots are visible, indicating that the Na ions are fixed to their positions. Between 15 and 130°C they move only along one-dimensional paths corresponding to the shortest Na-Na distances (labeled as 's' in the figure). Above 130°C all Na-Na distances (labeled as 'm') become identical and the Na⁺ ions have enough energy to move within the whole layer [2].

1D to 2D Na ion diffusion linked to structural phase transitions

M. Medarde ¹, M. Mena ¹, J. L. Gavilano ¹, E. Pomjakushina ¹, J. Sugiyama ², K. Kamazawa ³, V. Yu. Pomjakushin ¹, D. Sheptyakov ¹, B. Batlogg ⁴, H. R. Ott ⁴, M. Månsson ¹, F. Juranyi ¹
¹ Paul Scherrer Institute, Villigen, ² Toyota Central R&D Labs, ³ Aichi, CROSS, Ibaraki, Japan, ⁴ ETH Zürich

Lithium ion batteries are highly efficient and provide electrical energy for laptops, mobile phones and lately also for a growing market of electric cars. One of the drawbacks of this technology is the low abundance of lithium on our planet (20 ppm in the Earth’s crust), which makes the material expensive. A possible alternative might be the replacement of lithium for sodium [1] – an element with similar chemical properties but much more abundant, both in the Earth’s crust (20’000) and in sea water (16’000 ppm). Sodium is bigger and heavier than lithium, resulting in larger batteries with reduced energy densities. However, this is not a significant drawback for stationary applications such as the storage of peak energy in solar panels or wind mills.

Charging and discharging of batteries occurs via Li⁺/Na⁺ ion migration in and out of the battery electrodes (Fig. 1). Therefore, in order to be able to develop the necessary sodium-based batteries, it is crucial to understand how sodium ions move within the relevant materials. An example of how crystallographic studies contribute to this fast growing, economically relevant field is provided by a recent investigation on the prospective cathode material Na_0.7CoO₂, built up from alternating atomic layers of cobalt oxide and sodium ions (Fig. 1 and [2]).

The experiments were carried out using the powder diffractometer HRPT at the Swiss spallation neutron source SINQ, Paul Scherrer Institute. The results visualized for the first time the paths along which Na ions move in this material (Fig. 2 and [2]). In addition, it was shown that these paths change considerably with temperature due to subtle changes in the crystal structure: below 15°C, the sodium ions can hardly move; between 15 and 130°C they move only along 1-dimensional zig-zag paths corresponding to the shortest Na-Na distances; at still higher temperatures (T > 130°C) all Na-Na distances become identical and the Na⁺ ions have sufficient energy to move within the whole layer. That is, the character of the Na motion changes abruptly from one to two-dimensional as the temperature increases above 130°C [2]. One may now consider ways of optimizing the material’s properties for energy storage, e.g., by slight modification of their structure or composition.

[1] B. L. Ellis and L. F. Nazar, Current Opinion in Solid State and Materials Science 16, 168–177 (2012)
[2] M. Medarde, M. Mena, J. L. Gavilano, E. Pomjakushina, J. Sugiyama, K. Kamazawa, V. Yu. Pomjakushin, D. Sheptyakov, B. Batlogg, H. R. Ott, M. Månsson and F. Juranyi, Physical Review Letters 110, 266401 (2013)

Inversion centers and ferroelectric materials

Fig. 1. Schematic of the magnetic structure of TbMnO₃ at (a) T=35 K and (b) T=15 K, projected onto the b-c plane. Filled arrows indicate direction and magnitude of Mn moments. The longitudinally modulated phase (a) has a point of inversion while the spiral phase (b) does not. Hence, electric polarization indicated by the unfilled arrow is allowed in (b) but not in (a). Data were collected on the TriCS diffractometer at SINQ [3].

M. Kenzelmann ¹, A. B. Harris ², S. Jonas ³, C. Broholm ^3,4, J. Schefer ¹, S. B. Kim ⁵, C. L. Zhang ⁵, S.-W. Cheong ⁵, O. P. Vajk ⁴, J. W. Lynn ⁴
¹ Paul Scherrer Institute, Villigen, Switzerland, ² University of Pennsylvania, ³ J. Hopkins University, Baltimore, ⁴ NIST, Gaithersburg, ⁵ Rutgers University, Piscataway, USA.

Materials with ferroelectric and magnetically ordered ground states have been known for more than forty years [1]. For a long time it was thought that such materials are quite rare, because one of the dominant mechanisms for ferroelectricity relies on empty d-shells on the transition metal ions, which naturally precludes magnetism. In the past decade, substantial progress has been made towards the identification of new magnetically ordered ferroelectrics, also called multiferroics. Different mechanisms have been identified such as the lone-pair activity of Bi in BiFeO₃, geometric effects in hexagonal rare-earth manganites and barium fluorides, and magnetically induced ferroelectricity. While in the first two mechamisms the onset of ferroelectricity and magnetic order occurs at vastly different temperatures, magnetically induced ferroelectricity features coupled magneto-electric phase transitions.

One of the very first materials for which magnetically induced ferroelectricity was demonstrated, is TbMnO₃ [2]: experiments using TriCS at PSI show that the magnetic structure is an incommensurate spiral that breaks inversion symmetry and thus generates ferroelectric polarization [3]. There are two magnetic phases: the high-temperature magnetic structure is described by only one irreducible representation, and has inversion symmetry. The low-temperature magnetic structure is described by two irreducible representations and breaks inversion symmetry precisely in a way that allows for ferroelectricity along the c-axis. This identification has been an important step in the understanding of this type of materials.

[1] G. A. Smolenskii and I. E. Chupis, Phys. Usp. 25 (7) (1982).
[2] T. Kimura, T. Goto, H. Shintani, K. Ishizaka, T. Arima, Y. Tokura, Nature 426, 55 (2003).
[3] M. Kenzelmann, A. B. Harris, S. Jonas, C. Broholm, J. Schefer, S. B. Kim, C. L. Zhang, S.-W. Cheong, O. P. Vajk and J. W. Lynn, Phys. Rev. Lett. 95, 087206 (2005).

Protein micro-crystallography at the SLS

Fig. 1: Top left: Vertical beam projections obtained by integrating the intensity at the beam spot in the horizontal direction. Top right: Horizontal beam projections. Bottom row from left to right: Full beam, and beams through the 30 and 10 µm apertures, imaged by the on-axis sample microscope on a YAG:Ce scintillator screen.

Fig. 2. Crystal structure of activated beta-2 adrenergic receptor in complex with Gs. The receptor is colored red, Gα green, Gβ cyan and Gγ yellow. T4 Lysozyme and nanobody, which were used to facilitate crystallization, are omitted for clarity. Rendered from PDB entry 3SN6.

M. Wang, Paul Scherrer Institute, Villigen

During the last decade protein crystallography has been revolutionized at third-generation synchrotron radiation facilities by the development of undulator beamlines with a high-degree of automation. The recent upgrade of protein crystallography beamline X10SA at the SLS has successfully delivered a 10 × 10 μm² sized X-ray beam (Fig. 1) with a high flux density by combining beam shaping apertures and a smaller undulator gap. Sub-micron precision in both sample rotation and translation has been reached with a new micro-diffractometer. In addition, the highly sensitive and noise-free pixel array detector PILATUS 6M has been upgraded to reach a 25 Hz frame rate. The micro-beam, precision sample stage and faster X-ray detector have enabled development of a new fast 2D scanning routine (grid-scan) for identifying the best diffracting parts of crystals and locating micro-crystals in a high-throughput manner.

A micro-beam with grid-scan capability is essential for membrane protein crystallography because crystals of membrane proteins are often very small (<10 microns) and often impossible to visualize due to the lipidic cubic phase (LCP), which turns opaque upon cryo-freezing. The LCP crystallization methods have been applied successfully in crystallizing G-protein coupled receptors (GPCR) and their complexes (Figure 2) – a large family of membrane proteins, which are the targets of one third of available therapeutic drugs on the market. The micro-crystallography developments at the SLS have recently enabled two international pharmaceutical companies to determine their very first membrane protein structures with crystals grown in LCP.

Crystallography of nanoparticles

Fig. 1. Fe₂O_x NP oxidation and magnetism. Above: Number- and mass-based size distributions for the sample A1. Solid lines depict the size-dependent lattice parameter (purple curve) and sof of Fe(oct) (green curve). NPs consist of a Fe₂O₃ core and an oxidized Fe₃O₄ (maghemite) shell. From ref. [3]

Fig. 2. Crystallization of hydroxyapatite in the presence of citrate. Above: XRPD pattern fit as an ensemble of NPs. From this, the size and shape distributions of several samples at different maturation times have been determined. Below: schematic of the symmetry-breaking crystallization process of hydroxyapatite from an amorphous precursor in the presence of citrate. From ref. [4].

A. Cervellino, N. Casati, Paul Scherrer Institute, Villigen

Structural features of nanoparticles (NPs) are investigated at the Materials Science beamline of the Swiss Light Source (SLS). Having recently been upgraded, this beamline delivers undulator radiation for X-ray powder diffraction (XRPD) and surface diffraction experiments in a continuous energy range from 5 to 35 keV [1]. The powder diffraction station is fitted with highly efficient Mythen II detectors [2], having capability for simultaneous (time-resolved) SAXS-WAXS experiments. In particular, so-called total scattering methods have been applied at this station to investigate structural features of NPs with unprecedented accuracy. For example, the structure of Fe₂O_x (x=2.67–3) NPs, prepared as magnetite and partly oxidized under different conditions, has been determined. It was shown that they consist of a magnetite core with a shell of partly ordered magnetite and possibly – if an amorphous hydroxide precursor was used – a residual surface layer of amorphous hydroxide. The relative depth of the oxidized layer shows a dependence on particle size (Fig. 1). Magnetic properties associated to these NP systems have also been measured and correlated with the determined structural features [3].

Another example is the crystallization of calcium phosphate (hydroxyapatite) [4]. Starting from an amorphous precursor, the role of citrate anions in defining the shape of crystallizing apatite NPs was studied. Their shape breaks the hexagonal symmetry of the crystal lattice and this type of symmetry breaking is crucial to understanding the biological process of bone formation.

[1] P. R. Willmott et al., J. Synchrotron Rad. (2013). 20, 667-682.
[2] A. Bergamaschi et al., J. Synchrotron Rad. (2010). 17, 653-668.
[3] R. Frison et al., Chem. Mater., 2013, in press.
[4] J. M. Delgado-López et al, Adv. Func. Mater. (2013). DOI: 10.1002/adfm.201302075.

Fig. 1: Layout of the SNBL diffractometer with the new PILATUS detector from DECTRIS/Baden, Switzerland.

Fig. 2: Diffuse X-ray scattering patterns around the (0,0,-2) and (0,1,-2) reciprocal lattice points in lead zirconate at 550 K. Rows 1 and 3 show the experimental patterns, rows 2 and 4 the theoretical ones [2].

Diffuse scattering: Structural information between the Bragg spots

Ph. Pattison ^1,2, V. Dmitriev ², ¹ EPF Lausanne, Switzerland, ² SNBL@ESRF, Grenoble, France

Both Norway and Switzerland have relatively large and exceptionally active scientific communities using X-ray diffraction and absorption. To overcome the limited supply of synchrotron beamtime available at large-scale facilities, the Swiss and Norwegian scientists formed in 1990 a consortium and obtained access to a bending magnet port at ESRF with two branch lines: one dedicated to single-crystal diffraction and the other to powder diffraction, EXAFS and topography. Since the start of operation, more than one thousand publications have appeared using data from the SNBL, making it one of the most successful beamlines at ESRF.

While initially the emphasis was on the investigation of pharmaceutically relevant materials, the main focus of activity is now oriented towards energy-related research. Instrumentation development targeted at in-situ experiments has also been an important aspect of the beamline work in recent years, particularly for studying catalytic reactions and hydrogen storage. A recent example [1] illustrates the use of combined powder diffraction and x-ray spectroscopy to investigate the concept of inverse sigma transformation of a zeolite framework to generate a new structure by removal of a layer of framework atoms. For single-crystal experiments, users have exploited either the characteristics of a large area image plate detector or of a CCD detector mounted on a multi-axis diffractometer. Very recently, the image plate system has been replaced by a hybrid pixel array detector of the latest generation (a PILATUS 2M detector supplied by Dectris Ltd, Baden) mounted on a very flexible and versatile diffraction platform. The diffractometer with the PILATUS detectors is shown in Fig. 1. One of the scientific areas for which the new setup is particularly well suited is the investigation of diffuse X-ray scattering in single crystals. For example, a group led by Alexander Tagantsev from the Ceramics Laboratory of EPFL has investigated the lattice dynamics of antiferroelectric lead zirconate using inelastic and diffuse X-ray scattering techniques and Brillouin light scattering [2]. The diffuse X-ray scattering patterns together with the modelled distribution are shown in Fig. 2. The results show that the antiferroelectric state is a ‘missed’ incommensurate phase and that the paraelectric to antiferroelectric phase transition is driven by the softening of a single lattice mode via flexoelectric coupling. These findings resolve the mystery of the origin of antiferroelectricity in lead zirconate and suggest an approach to the treatment of complex phase transitions in ferroics.

The combination of high brilliance and high x-ray energies provided by the bending magnet source at the ESRF also allows the users of SNBL to investigate the effects of high pressure with diamond anvil cells. A most unexpected and puzzling phenomenon that can be observed under high pressure, is negative linear compressibility. In a collaboration between the beamline staff and the University of Oxford [3] it was possible to reveal that the molecular framework material Zn[Au(CN)₂]₂ exhibits the most extreme and persistent negative linear compressibility behavior yet reported: under increasing hydrostatic pressure its crystal structure expands in one direction at a rate that is an order of magnitude greater than the typical contraction observed for common engineering materials. In these and similar studies, it is the combination of the excellent characteristics of the source and the high performance of the new generation of pixel detectors that opens up many exciting avenues of research in the fields of solid state physics and crystal chemistry using synchrotron radiation.

[1] E. Verheyen et al., Nature Materials, 11, 1059–1064 (2012)
[2] A. K. Tagantsev et al., Nature Communications, 4:2229 (2013)
[3] Andrew B. Cairns et al., Nature Materials, 12, 212–216 (2013)

Outlook

The complexity of materials requires a combination of probes and studies over a wide range of length, energy, and time scales in order to be able to investigate their structural, electronic and magnetic properties. While these studies are of fundamental interest and increase our understanding of the complex correlations and interactions that are at work in such systems, materials are essential for technological applications ranging from IT devices to energy harvesting and storage. Of equal importance is macromolecular crystallography and pharmaceutical research. Swiss scientists are in the unique position to profit from the availability of large facilities delivering neutron, photon and muon beams for their research. Under construction is the free electron laser SwissFEL, which will deliver ultrashort X-ray pulses for flash crystallography.

[Released: January 2014]