Path:
Periodical volume

Full text: Highlights Issue 2006

Berliner Elektronenspeicherring-Gesellschaft für Synchrotronstrahlung m.b.H.

Highlights 2006

Highlights 2006

Berliner Elektronenspeicherring-Gesellschaft für Synchrotronstrahlung m.b.H.
Member of the Leibniz Association

Published by: Berliner Elektronenspeicherring-Gesellschaft für Synchrotronstrahlung m.b.H. – BESSY Albert-Einstein-Straße 15 12489 Berlin, Germany phone +49 (0)30 / 6392 2999 fax +49 (0)30 / 6392 2990 www.bessy.de info@bessy.de Board of Directors: Prof. Dr. Dr. h.c. Wolfgang Eberhardt, Prof. Dr. Eberhard Jaeschke, Thomas Frederking Editors: Dr. Heike Henneken, Dr. Markus Sauerborn Layout: Annette Weber, Stitz & Betz GmbH, Berlin ISSN Number: 1611-6127

Cover Photo by Henrik Spohler / laif

Berliner Elektronenspeicherring-Gesellschaft für Synchrotronstrahlung m.b.H.
Member of the Leibniz Association

Introduction

Scientific Highlights

News & Events

Femtoslicing Special

Facility Repor t

User Pages

Introduction Scientific Highlights Atomic Physics Magnetism Surface Sciences Applied Sciences Material Sciences Life Sciences Metrology News & Events Femtoslicing Special Facility Report Maschine Operation Beamline Developments STARS shining bright the status of the BESSY FEL Project The Metrology Light Source of the PTB User Pages Operation Statistics Improvements for Users Beamlines Experimental Stations Board and Committees

04 06 08 10 16 22 26 30 34 36 42 48 49 50 52 53 54 55 56 58 60 62

Page

03

Introduction

Scientific Highlights

News & Events

Introduction

BESSY - Highlights 2006

Femtoslicing Special

Facility Repor t

User Pages

Dear BESSY-users and friends,
Diamonds, glass and STARS … This year’s Highlights appear to be shiny and sparkling. Researchers took a closer look at diamonds and glass and found that glass although breakable in our everyday experience is well suited for locking up toxic wastes in a stable, durable way for long time. Whereas blue diamonds are not only of jewellers’ interest. They appear to be superconductors behaving essentially like their metal cousins, such as niobium. Niobium is the basis of all linear accelerator (linac) modules for free electron lasers (FEL). Our FEL project received excellent marks in spring by the Wissenschaftsrat and we are eager now to build and operate the recommended two-stage cascaded FEL to demonstrate the feasibility of HGHG principle in the STARS project, the other sparkling issue. We are confident to achieve the goal since the successful operation of the Femtoslicing Source is already the first step of a cascaded FEL. In addition, recent measurements at BESSY showed for the first time that the accelerator modules can be adopted for continuous wave operation in a superconducting linac, which is also a major ingredient of our FEL project. Sadly, our long-time member of the scientific advisory board and supporter Neville Smith will not see the BESSY-FEL up and running. He deceased after a short suffering from cancer. We will miss his advice and his companionship. BESSY attracts many people with all different kinds of scientific and social background and we are pleased that we welcomed some 6,500 visitors during the year. Among them were visitors during the 'Lange Nacht der Wissenschaften', our Breakfast Physics during the summer holiday, delegates on a workshop on the use of Synchrotron Radiation in Arts and Archeology and our new minister of Science and Education Dr. Annette Schavan. We participated in the initative 'Germany – Land of Ideas' showing ideas and innovation during the entire year and presented BESSY at fairs, at the 'Highlights der Physik' and at political events. Free electron lasers have been in an one week focus during the FEL06 conference hosted by BESSY in collaboration with FZ Rossendorf in Berlin. Some 300 participants literally from all over the world discussed new developments and research opportunities during the four day conference. The 25th users meeting in December has seen a record number of participants and poster contributions in BESSY history, reflecting the growing interest in the use of synchrotron radiation and the tremendous variety of research topics. In 2006 all together seven prominent members of our user community and staff received calls for appointments as university professors based upon work carried out at BESSY. We would like to congratulate them and all our users for the exiting research they have been performing and we thank our staff members for their efforts and their never decreasing engagement. Enjoy reading the Highlights 2006.

Prof. Dr. Dr. h.c. Wolfgang Eberhardt

Prof. Dr. Eberhard Jaeschke

Thomas Frederking

Page

05

Introduction

Scientific Highlights

News & Events

Scientific Highlights

BESSY - Highlights 2006

Femtoslicing Special

Facility Repor t

User Pages

Atomic Physics

08

Magnetism

14

Surface Sciences

20

Applied Sciences

24

Material Sciences

28

Life Sciences

32

Metrology

34

Page

07

Introduction

Scientific Highlights

News & Events

Source: CartoonStock.com

Electrons take the dress, that fits the best: The photoionization of laser-excited Li atoms
D. Cubaynes1, E. Heinecke2, M. Meyer1, T. Richter2, F. Wuilleumier1, P. Zimmermann2

1 LIXAM, CNRS, Université Paris-Sud, France 2 Technische Universität Berlin

An atomic nucleus is enrobed with several layers of electronic orbitals. The inner orbitals are usually occupied by electrons, the outer ones empty. The particular shape and size of these 'dresses' is defined by the atomic potential, i.e. by the various interactions between the nucleus and the electrons. In case one of the electrons is removed from the atom by photoionization, the other electrons are most likely to stay in their orbitals. The electrons in the new ionic potential still fit to their old 'dresses'. However, in laser-excited Lithium we observed that this is not always true. Strong differences between the neutral and the ionic potential cause a mismatch of the orbitals in the neutral atom and the ion, in other words, the 'dress sizes' for the electrons change upon ionization. The electrons take the dress, that fits the best. As a result, in the photoionization of excited Lithium the electron emission is intimately connected to the promotion of another electron into a higher orbital. To investigate this effect we use the socalled pump-probe technique, i.e. the successive absorption of two photons by the atom (Fig. 1). The first photon, here from an intense laser, is used to promote one electron to a higher orbital. In this way the atom can be prepared in a specific well-defined excited state, in other words, an electron can be inserted into another orbital modifying thereby the intra-atomic interactions in a controlled way. The second photon, here synchrotron radiation from BESSY, is then used to ionize the atoms. The combination of laser and synchrotron radiation has proven to be ideally suited for these studies [1], since the intense narrow-band laser light provides an efficient, controlled excitation process and the high-energy, tunable synchrotron radiation allows observation of photoionization in a wide range of photon energies. The modifications to the electronic cloud introduced by the laser can be evaluated by observing and analyzing the resulting differences in the photoelectron spectrum. In the present study, we have investigated the photoionization of free Lithium atoms. Lithium has a very simple ground state configuration, Li 1s22s, composed from only three electrons. Atoms with few electrons

Fig. 1: Typical two-photon pump-probe excitation scheme.

exhibit a pronounced sensitivity to the displacement of one electron, making them an ideal testing ground for studies of electronic interactions. A high-resolution ring dye laser operating in a continuous wave mode at λ = 670 nm excited first the 2s electron to the 2p shell, i.e. more precisely to the Li*1s22p 2P3/2 state. A laser power of up to 500 mW (i.e. about 1018 photons/s) enabled us to excite about 20% of the Li atoms in the interaction volume. In addition, it was possible for the first time, to prepare the Li atoms in the 3p laser-excited state. This excitation requires laser photons in the more difficult to produce UV wavelength region. By frequency doubling in an external cavity up to 50 mW average power could be delivered at λ = 323 nm and about 1% of the Li atoms were excited to the 3p 2P3/2 state. In the experiment, the synchrotron radiation crosses a mixture of atoms in the ground and excited state (Fig. 2) and the photoelectron spectrum is composed of both contributions. Excellent energy resolution for the exciting photon beam, provided from the UE52-SGM beamline, as well as for the electron analysis (Scienta SES-2002) is therefore prerequisite for the success of the experiments. The relevant parts of the photoelectron spectra, recorded after the interaction of synchrotron photons of 100 eV with an effusive beam of Li atoms, are displayed in Figure 3. The spectrum in the lower part shows the photoionization of Li atoms in the ground state, i.e. with no laser light in the interaction volume. The highest intensity is found for the Li+1s2s configuration, giving rise to two main lines, 3S and 1S, as the two electrons can couple their spins in the same or opposite direction. But the removal of one charge in the ionization process results in a modification of the electronic orbitals with respect to the neutral atom. There is therefore a non-negligible probability that the actual position of the outer electron corresponds to a higher (less strongly bound) shell in the ion. This gives rise to the intensity in the Li+1s3s satellite lines at lower kinetic energies. Generally, the main lines exceed the satellites in intensity.

References:

[1] F. J. Wuilleumier, M. Meyer, J. Phys. B 39, R425 (2006). [2] Z. Felfli, S. Manson, Phys. Rev. Lett. 68, 1687 (1992). Financial support by the EC through EC-IA-SFS is acknowledged.

BESSY - Highlights 2006

Femtoslicing Special

Facility Repor t

User Pages

Fig. 3: Parts of the electron spectra recorded with 100 eV photon energy from the Li 1s22s ground state (bottom), the Li*1s22p (center) and Li*1s23p laser-excited state (top).

Fig. 2: Experimental geometry around the interaction volume. The bright white line in the center is produced from fluorescence of laser-excited Li atoms. The direction of the atomic Li beam is point is perpendicular to the drawing plane.

When the Li atoms are laser-excited to the Li*1s22p state (Fig. 3, middle part) before ionization, this clear situation changes. The main line, corresponding now to the Li+1s2p configuration, still comprises most of the intensity, but the relative intensity of the satellites, Li+1s3p in this case, is increased matching almost the main lines. A dramatic change, however, is observed when the 2s electron is laser-excited to the 3p shell (Fig. 3, upper part). The spectrum is now dominated by the Li+1s4p satellite, caused by the shake-up of the initial 3p electron to the higher 4p orbital. The main line is now much weaker than the satellite! It is the first time that such a strong alteration of electronic orbitals is observed in the direct photoionization process and only the preparation of the Li atom in the 1s23p laser-excited state provided the conditions to highlight this exceptional effect. The reason for the high satellite intensity is found in the vast mismatch of the 3p orbitals in the neutral and in the ionic state (Fig. 4). The ionization of the atom causes a spatial contraction of the wavefunctions making the 3p wavefunction in the laser-excited state very similar to the 4p wavefunction in the ion [2]. For atomic physics, the photoionization of laser-excited Li is a showcase to demonstrate fundamental principles of quantum

Fig. 4: Schematic representation of the excitation and ionization process.

mechanics, in particular the change of the radial distribution of electronic orbitals upon ionization. In a broader sense, it shows that it is not always possible to break-up a photoelectron spectrum into 'strong' main lines, arising from a direct, one-electron process, and 'weak' satellites, related to electron correlations. In more complex systems, the outer-shell of the atom can be modified by its environment, e.g. by other atoms or surfaces, and detailed knowledge of the electronic interaction is important.

Contact: Michael Meyer michael.meyer lixam.u-psud.fr Page

09

Introduction

Scientific Highlights

News & Events

Two magnetic ions join forces in multiferroics
E. Dudzik, R. Feyerherm, O. Prokhnenko, N. Aliouane, D. Argyriou

Hahn-Meitner-Institut Berlin

Unlike magnets, which occupy a familar place both on the fridge and inside the computer, ferroelectrics are mainly known to experts. But the combination of ferroelectricity and magnetism has the potential to change our computers in future. It could eventually lead to ultra-high density non-volatile memory devices, capable of retaining data during power loss and with very short boot-up times. In magnets microscopic magnetic moments appear spontaneously, which can be switched back and forth by an applied magnetic field, and thus can be used for data storage. Ferroelectric materials behave in an analogous fashion: they have a spontaneous electric polarisation which can be switched by an applied electric field. The name 'ferroelectric' does not mean that these materials contain iron – it was coined because their behaviour is so similar to that of a ferromagnet. In a small group of materials – called multiferroics - ferromagnetism and ferroelectricity are closely entwined: their magnetisation can be switched by an electric field, and an applied magnetic field will switch their electric polarisation [1,2]. This multifunctional mix makes it possible to build a wide range of complex electronic devices, for example storage media where data can be written by applying an electric field (eliminating the need for a high writing current in a purely magnetic device), and read magnetically. The orthorhombic rare earth manganites are one such group of materials that show a complex interplay of magnetism and electric polarisation. They contain two different magnetic ions, manganese and a rare earth metal, whose spins show several stages of magnetic ordering [3]. The following discussion is limited to the rare earth metals Terbium and Dysprosium. At room temperature these materials are paramagnetic. The first magnetic phase transition occurs at around 40 K, where the Mn moments order in a sinusoidal spin density wave whose period changes with temperature. At a second transition the Mn order locks into a helicoidal spin density wave in the bc plane of the crystal lattice. The rare earth ions order magnetically at temperatures below 10 K.

Both DyMnO3 and TbMnO3 show a spontaneous electric polarization below the intermediate lock-in transition (see Fig. 1, above). In TbMnO3 the electric polarisation sets in below 28 K, and after an initial rapid increase rises gradually towards lower temperatures. The DyMnO3 polarisation on the other hand increases rapidly below the lock-in temperature at about 16 K, reaches a maximum value which is double the Tb polarisation, and then drops to TbMnO3 levels at about 5 K. The Dy polarisation also shows a marked hysteresis at low temperatures. The electric polarisation in TbMnO3 has been explained theoretically [4] on the basis of neutron diffraction measurements. Below the lock-in the helicoidal ordering of the Mn moments in the bc plane breaks both time reversal and spatial inversion symmetry and leads to a polarisation along the c direction (see Fig. 2). Our experiment was motivated by the question why the temperature dependence and the magnitude of the DyMnO3 polarisation are so different from that of TbMnO3. Dy is practically opaque for neutrons, so X-ray resonant scattering was used to study the magnetism. The experiments were carried

References: [1] N. A. Hill, J. Phys. Chem. B 104, 6694 (2000). [2] W. Eerenstein et al., Nature 442, 759 (2006). [3] T. Goto et al., Phys. Rev. Lett. 92, 25721 (2004). [4] M. Mostovoy, Phys. Rev. Lett. 96, 067601 (2006). [5] R. Feyerherm, Phys. Rev. B 73, 180401(R), (2006). [6] O. Prokhnenko et al., Phys. Rev. Lett., in press

Fig. 1: Above: electric polarisation in TbMnO3 and DyMnO3 (after Goto et. al. [3]). Below: Magnetic diffraction peak intensities in DyMnO3.

BESSY - Highlights 2006

Femtoslicing Special

Facility Repor t

User Pages

out at the new high-energy MAGS beamline operated by the Hahn-Meitner-Institute at BESSY. Samples were cooled in a closed-cycle cryostat mounted on a six-circle diffractometer. Measurements were done both at the Dy L3-edge at 7,790 eV and non-resonant at 12,398 eV. X-ray resonant scattering clearly shows the three stages of magnetic ordering (Fig. 3). Below the Neel temperature of 40 K, the Mn orders in a sinusoidal spin wave with a propagation vector τMn=0.385 b* in reciprocal space. Although the diffraction peaks from the Mn magnetic order could not be observed directly, we found second order diffraction peaks which are due to a structural distortion associated with the Mn order at positions k ± 2τMn (Fig. 3, red inset). These second order peaks appear, because the magnetic order distorts the crystal unit cell. This distortion has half the period of the magnetic ordering (Fig. 4). These peaks persist below the lock-in transition at 16 K. At the lock-in, a second set of diffraction peaks appear at positions k ± τMn (Fig. 3, orange inset), which are resonant at the Dy L3 edge. Linear polarization analysis shows that these peaks are of magnetic origin. This shows that in the lock-in regime, where the spontaneous electric polarization appears, the Dy moments have ordered with the same periodicity as the Mn. At temperatures below 4.35 K the Dy moments order independently of the Mn. This leads to the collapse of the magnetic diffraction peaks at k ± τMn and the appearance of magnetic peaks with a period of τDy = 0.5 b* (Fig. 3, blue inset). As in the case for the Mn, these magnetic diffraction peaks are found to be associated with a second order structural distortion. These peaks should appear exactly at integer order positions, but instead are found at peak positions of q = 0.905 b*. It appears that competition with the persisting Mn order shifts both the Dy and the Mn second order structural peaks slightly from their expected positions towards one another [5]. The transition from the independent Dy magnetic order to the induced order above the Dy Neel temperature shows a strong hysteresis (Fig. 2 below). This hysteresis coincides in size and temperature dependence with the hysteresis in the Dy electric polarisation shown in the upper part Fig. 2. Therefore we argue that in the lock-in regime, where the Dy moments order with the same periodicity as the Mn, the Dy moments contribute to the

Fig. 2: Helicoidal magnetic order can induce an electric field in a crystal that is perpendicular to the magnetic propagation vector.

Fig. 3: Electric polarisation in DyMnO3 shown with the X-ray diffraction data for high, intermediate and low temperatures (red, orange and blues insets, respectively).

Fig. 4: A magnetic modulation can distort the crystal structure. When the period of this structural distortion is half that of the magnetic modulation, it leads to second order diffraction peaks.

electric polarisation. Thus the increase in polarisation above the Tb value shown in Fig. 2 between 4.35 and 16 K appears to be due to a contribution from the Dy moments. Surprisingly, the two magnetic moments situated at completely different positions in the unit cell combine to induce an electric polarisation as if they simply sum up [6]. This idea is supported by the fact that in TbMnO3 in the ferroelectric phase the Tb moments order along a different direction from the Mn, and so cannot contribute to the electric polarisation.

Contact: Esther Dudzik dudzik hmi.de

Page

11

Introduction

Scientific Highlights

News & Events

Source: Alastair Rae

Melting the antiferromagnetic ice sheet
J. Goedkoop1, S. Konings1, Ch. Schüßler-Langeheine2, H. Ott2, E. Schierle3, E. Weschke3

1 University of Amsterdam, Netherlands 2 Universität zu Köln 3 Freie Universität Berlin

Polar bears and physicist share a strong interest in phase transitions. Polar bears take the real world irregularities that cause their ice flow to break up in fixed frozen and more dynamic molten parts rather stoically. For physicists however, who like to think in terms of ideal ordered systems, real world disorder is just bad news. Pet theories like critical scaling, which describe the divergence of the size of fluctuations at phase transitions, break down in the presence of disorder. Take for instance thin films, where structural disorder can never be avoided. As a result, magnetic phase transitions in such films are difficult to understand, a situation that is quite disturbing in view of the importance of such films in science and technology. To oversee the battle between order and fluctuations in phase transitions and to understand the role of structural disorder in this fight requires a tool that covers a huge range of length- and timescales. While the length scale of fluctuations can be readily probed by conventional diffraction experiments, the observation of the corresponding fluctuation times is much more challenging. As we will show here, coherent soft X-rays turn out to provide a remarkably powerful looking glass for this purpose.

References: [1] H. Ott et al., Phys. Rev. B 74, 094412 (2006). [2] B.J. Berne and R. Pecora, Dynamic Light Scattering with Applications, John Wiley (1976). [3] G. Grübel and F. Zontone, J. Alloy. Compd. 362, 3 (2004). [4] D.O. Riese et al., Phys. Rev. Lett. 85, 5460 (2000). [5] I. Sikharulidze et al., Phys. Rev. Lett. 88, 115503 (2002). [6] http://www.science.uva. nl/research/cmp/goedkoop/group/ docs/fluctuations/fluc2.html. [7] Stephen D. Kevan, Personal communication. Acknowledgements: The authors are indebted to H. Zabel for making excellent samples. Funded by EU, the DFG through SFB 608 and FOM/NWO (Netherlands).

As a test case, we studied an epitaxiallygrown ultrathin 11-monolayer holmium film displaying a helical magnetic order with a periodicity along the surface normal [1]. This order disappears in a broad transition around TN=76 K. In the experiment the coherence of the light of the beamline (U49-2 PGM 1 and UE46 PGM) is obtained by a spatial filter consisting of two pinholes (Fig. 1). The periodic helical magnetic structure produces strong magnetic diffraction peaks, when the photon energy is tuned to the M5 resonance of holmium. One these peaks we intercept with a CCD camera. In this way we can observe the melting of the magnetization, much like polar bears watch the ice melt under their feet. Unlike the polar bears however, we can regulate the temperature in our experiment within 10 mK and in this way control the magnetic correlation length ξ, the typical length scale over which the sample is magnetically ordered. With incoherent light we would observe one smooth diffraction peak, with a width that is inversely proportional to the correlation length of the ordered regions. As expected, we find that the closer we approach the transition temperature TN, the smaller is the scattered intensity (Fig. 2). The correlation length also decreases, although in a way that is quite different from standard models. What is happening here? In order to find out we make our light coherent and study the fluctuations in the time domain. When coherent light is used, the magnetic peak breaks up in a myriad of speckles, which form the diffraction pattern of the magnetic structure of the illuminated spot. Any changes in this structure are immediately reflected in changes in the speckle pattern. By recording the speckle-pattern as a function of time and performing a time correlation analysis of the speckle intensities it is possible to extract the correlation time of the fluctuations.

Fig. 1: Sketch of the coherent diffraction experiment and the helical magnetic structure of holmium. The length scale over which the magnetic structure is ordered in the plane of the sample is depicted by the correlation length .

BESSY - Highlights 2006

Femtoslicing Special

Facility Repor t

User Pages

� (arb. units)

�� � (nm)

Fig. 2: Temperature dependence of the correlation length  and intensity I. The fluctuations could be measured in the time domain up to 70 K.

This technique is called Photon Correlation Spectroscopy [2] which is well known from laser light scattering where, among other things, it is used to study diffusion in colloidal systems. With the availability of brilliant X-ray sources, this laser technique has started to make inroads into the X-ray regime [3], where the fluctuations can be measured to much smaller correlation lengths. So far, only soft condensed matter systems have been probed such as colloids [4] or smectic films [5]. Here we use the technique to follow correlations in a hard condensed matter system, namely the magnetic fluctuations in our film. Fig. 3 shows one of the speckle patterns with an exposure time of 4 seconds together with snapshots of the 3-hour speckle movies, some of which can be viewed online [6]. From these it is clear that at 50 K the magnetic structure is fixed, but that at higher temperatures fluctuations appear that speed up dramatically. Moreover, when all images in a movie taken at a fixed temperature are averaged, we find something remarkable. At 50 K, the intensity profile has a contrast close to a single speckle pattern of 4 s exposure, confirming a frozen domain structure (Fig. 4). At 70 K the intensity profile blurs out, but still some fixed speckles remain. This proves that the magnetic fluctuations are highly constrained by the defect structure of the film. In more technical terms, the system is said to be non-ergodic. Although there are some important differences, this transition is analogous to the melting of ice, in which the correlation length of the ordered ice phase (e.g. the ice floes) will be reduced when the temperature increases. The limited coherent flux binds us at present to the ultra slow regime, between 4 seconds and 3 hours. From an intensity correlation analysis we find that this time regime corresponds to the fluctuation times in our sample at temperatures below roughly 70 K. This is the first time such magnetic correlations have been followed not only in the space but simultaneously in the time

Fig. 3: Top: speckle pattern at 50 K. Below: examples of 4 s exposures acquired at the indicated time in the area marked in the top panel at 50 K (top rows) and at 70 K (bottom rows), showing the fluctuations in the speckle pattern that reflect the fluctuations in the magnetic structure. At 50 K the speckle pattern is almost completely static but is fluctuating rapidly at 70 K (6 K below the film’s ordering temperature).

Fig. 4: Intensity profile obtained by averaging over 2,200 speckle patterns acquired over about three hours.

domain. The movies undoubtedly show that the strange behavior of the correlation length is not only due to the reduced dimensionality of the film, but also due to real world defects. The challenge now is to determine the temperature-dependent ratio of the dynamic to static parts of the sample in this magnetic system. Meanwhile beckoning on the horizon are the plethora of correlated electron systems showing stripes, charge and spin density waves or other long range order effects which may or may not be dynamic in nature. With the 1,000 times higher coherent flux that can be expected at optimized beamlines [7] or the 109 times higher flux provided by X-FELs, the future for this technique looks excellent.

Contact: Jeroen Goedkoop Goedkoop science.uva.nl Page

13

Introduction

Scientific Highlights

News & Events

Source: Juan Salmotal Franco

Filling the seats: Atomic-scale roughness of antiferromagnetic films
W. Kuch1, L. I. Chelaru2, F. Offi2, J. Wang2, M. Kotsugi2, J. Kirschner2

1 Freie Universität Berlin 2 Max-Planck-Institut für Mikrostrukturphysik, Halle

References: [1] G. A. Prinz et al., Science 282, 1660 (1998). [2] S. A. Wolf et al., Science 294, 1488 (2001). [3] W. Kuch et al., Nature Mater. 5, 128 (2006). [4] J. Stöhr et al., Science 259, 658 (1993). [5] W. Kuch et al., Phys. Rev. Lett. 92, 017201 (2004). [6] J. C. Slonczewski, Phys. Rev. Lett. 67, 3172 (1991). [7] F. Offi et al., Phys. Rev. B 66, 064419 (2002) Acknowledgments: We thank B. Zada and W. Mahler of the UE56-2 beamlines for their help. Supported by the BMBF

Saturday afternoon: The first supporters arrive and fill the stand, but this stadium has peculiar rules. On the orange seats women and men are seated in a strict alternating order, while the blue stand has to be filled with either only men or only women. The gender of the supporters in the blue seats has to be selected such that at the interface between the orange and blue seats preferably couples of men and women are placed. This makes the filling of the seats an interesting problem, which very much depends on the details of how the border line between orange and blue exactly runs. This is pretty much the situation of the interaction at the interface in a stack of ferro- and antiferromagnetic films. These multilayers are important ingredients of many devices like hard disk read heads, magnetic sensors, or magnetic random access memories [1,2]. Their widespread use in commercial applications is, however, not paralleled by a detailed fundamental understanding of the interaction between antiferromagnetic and ferromagnetic materials. While in ferromagnets all the atomic magnetic moments are pointing in the same direction (i.e. are all male or female in the stand), the magnetic order in antiferromagnets is more complex. In antiferromagnets the atomic moments align such that the total moment vanishes if averaged over a few neighbor atoms. In the simplest ones every other atom has its spin pointing into the opposite direction. To achieve understanding of the magnetic interaction between an antiferromagnetic and a ferromagnetic material, it is thus vital to characterize and control the interface structure and roughness on the atomic level. We used the technique of magnetic domain imaging by x-ray magnetic circular dichroism photoelectron emission microscopy (XMCDPEEM) [4] to study the dependence of the magnetic interface coupling on thickness and layer filling of both the ferromagnetic and antiferromagnetic films [3]. The absorption of circularly polarized synchrotron radiation from the elliptic undulator beamline UE56-2 PGM 2 was used to obtain magnified images of the sample by means of the emitted secondary electrons. In these images, regions of different intensity represent different magnetization directions. The XMCD

effect occurs only in resonance at elemental absorption lines, thus magnetic domain images of different layers at different depths of the single-crystal heterostructures can be acquired separately if they contain different elements. Crossed-wedge samples, in which the thickness of a ferromagnetic and an antiferromagnetic layer vary along perpendicular directions, were prepared, and allow the simultaneous visualization of the magnetic coupling as function of two different layer thicknesses. Fig. 1 shows an example of a trilayer on a Cu(001) substrate in which antiferromagnetic FeMn is sandwiched by two ferromagnetic layers, a Co layer at the bottom, and a Co/Ni hybrid layer at the top. Panel (a) shows a sketch of the wedge geometry. The thickness of the bottom Co layer (blue) increases from left to right up to 8 atomic monolayers (ML), and then stays constant. The thickness of the antiferromagnetic FeMn layer (red) varies from bottom to top. Panel (b) shows the magnetic domain image of the as-grown Co bottom layer, panels (c) and (d) the domain images obtained at the Co and Ni absorption resonances, respectively, after deposition of the complete structure. Because both the bottom and top ferromagnetic layers contain Co, image (c) is a superposition of the magnetic domain patterns (b) and (d), while panel (d) represents the top layer only. Alternating regions of parallel and antiparallel coupling across the FeMn layer are indicated by couples of parallel and antiparallel arrows in (c). They alternate with a 2-ML period as a function of FeMn thickness, but also exhibit an interesting saw-tooth-like behavior on the thickness of the bottom Co layer. The latter represents the dependence on the interface morphology. It is modulated by the thickness, and hence the atomic layer filling, of the bottom Co layer. From these measurements and supporting magneto-optical Kerr effect experiments, the following picture could be deduced [3]: First, to have a significant magnetic interaction between the ferromagnetic and the antiferromagnetic layers, steps of single atom height at the interface are required. Perfectly flat regions do not contribute. This follows from

BESSY - Highlights 2006

Femtoslicing Special

Facility Repor t

User Pages

the absence of 90° coupling, which would otherwise be expected at FeMn thicknesses close to n + 0.5 ML thicknesses with n an integer number [6]. Also the influence of the Co bottom layer thickness on the sign of the coupling is not compatible with flat regions being the dominant source of the antiferromagnet/ferromagnet coupling. Second, the coupling is higher if these monatomic steps are laterally confined at small islands. Larger islands or elongated steps mediate a weaker coupling. The strength of the coupling can only be deduced from the Kerr measurements. The observed saw-tooth pattern is surprising. The interface roughness has maxima at around 50% atomic layer fillings and decreases towards both sides, for higher and lower fillings, leading to the sine-like oscillations typically observed in the diffracted electron intensity [7]. In the present case, however, a 20% filling is completely different from an 80% filling, although the number of monatomic steps and thus the roughness may be equal. This leads to the third conclusion, that the amount of antiferromagnetic material needed to complete the outmost atomic layer of the ferromagnets is not contributing to the sign of the coupling. The coupling is mediated by uncompensated spins of the antiferromagnet at monatomic step edges at the interface. A sketch of a possible interface spin configuration at such a monatomic step edge is shown in Fig. 2. Black and gray bullets with arrows represent atomic moments of next-level atomic planes of the non-collinear antiferromagnetic spin structure [5] of FeMn. Ellipses at the step edges indicate regions in which the antiferromagnetic spins do not cancel, but follow the magnetization direction MFM of the ferromagnetic layer. These uncompensated atomic moments are responsible for the magnetic coupling to the ferromagnet. Our results indicate that, in general, the interface coupling can be enhanced by the controlled incorporation of atomic-level roughness features with small lateral size. This would be like designing the border line between the orange and blue seats in the stadium in a particular rough zig-zag shape such as to provide a very clear-cut decision about the gender of the supporters in the blue stand. With the forthcoming advent of atomic-scale manipulation in nanotechnology, this may be a feasible way to controllably modify the coupling strength in ferromagnetic–antiferromagnetic systems.

3 ML Co

0–20 ML FeMn 0–8 ML Co
(a)

15 ML Ni Cu(001)
16
Co bottom layer

(b)

FeMn thickness (ML)

10

12

14

8 16

10

12

14

8 16

10

12

14

[010]
h�
Co

10 µm
(c)

Ni

(d)

8

4

5

8 7 6 Co thickness (ML)

Fig. 1: (a): Geometry of a crossed double-wedge sample. Antiferromagnetic FeMn of varying thickness is sandwiched between ferromagnetic layers consisting of Co at the bottom and Co/Ni at the top. (b): Domain pattern of the Co bottom layer. Bright and dark regions correspond to magnetization direction down and up, respectively. (c): Element-selective domain image of the complete sandwich structure, acquired at the Co L3 edge. Bright and dark regions result from a superposition of magnetization directions of Co in the bottom and top layer, indicated by couples of arrows. (d): Element-selective domain image of the top layer, acquired at the Ni L3 edge. Bright and dark regions correspond to magnetization direction down and up, respectively.

Fig. 2: Sketch of a possible interface spin structure of the antiferromagnetic FeMn layer. Uncompensated atomic moments at monatomic step edges (ellipses) mediate the magnetic coupling to an adjacent ferromagnetic layer.

MFM
Contact: Wolfgang Kuch kuch physik.fu-berlin.de

Page

15

Introduction

Scientific Highlights

News & Events

Peer pressure among electrons: How localized and itinerant states interact
Kai Godehusen

Yu. S. Dedkov1, M. Fonin2, Yu. Kucherenko3, S. Molodtsov1, U. Rüdiger2, C. Laubschat1

1 Technische Universität Dresden 2 Universität Konstanz, 3 Institute for Metal Physics, Kiev, Ukraine

Sometimes the world of atoms behaves very human: Magnetism is a collective phenomenon where the spins of electrons are oriented in the same direction. This is possible in a solid state where all atoms share the outermost electrons to build a band. In contrast to this concept are rare earth (RE) atoms, whose outermost 4ƒ shell is actually very close to the core and therefore more like in a single atom [1]. The question arises, whether an individual (4ƒ) electron can change the behavior of the group (band magnetism) and vice versa? Talking about people, the reaction could be manifold, i.e. a certain glance is enough to cause a reaction. For the atoms one has to consider electrons which hop between the localized (4ƒ) orbital and the valence band (group). The magnitude of the hopping probability depends on the hybridization matrix element and the density of valence-band states. Both quantities are usually large in RE transition metal systems. Here, the magnetic properties depend not only on the localized 4ƒ moments, but also on itinerant d states that form exchange-split bands with different densities of states for different spin orientations [2]. One would expect, that the hopping probabilities and, thus, the magnetic properties of the 4ƒ states vary also as a function of spin orientation. Although this idea is rather transparent, such an effect has not been discussed to our knowledge in the literature so far, even so it might have severe consequences for the understanding of magnetic anomalies in these compounds. Ce, the first element of the RE series, is subject of strongest interactions between 4ƒ and valence-band states. Here we report on a spin- and angle-resolved resonant photoemission (PE) study of an ordered Ce monolayer on Fe(110) (inset in Fig. 1) at the Ce 4d -->4ƒ absorption threshold. The experiments were performed with a hemispherical PHOIBOS 150 electron-energy analyzer combined with a 25 kV mini-Mott spin detector using synchrotron radiation from UE112 PGM 1 beamline of BESSY. Application of resonant PE is necessary for a proper discrimination ofthe Ce 4ƒ and valence-band emissions.

References: [1] H. R. Kirchmayr and C. A. Poldy, in Handbook on the Physics and Chemistry of Rare Earths, vol. 2 (North-Holloand, Amsterdam, 1978). [2] O. Gunnarsson and K. Schönhammer, Phys. Rev. Lett. 50, 604 (1983). [3] E. Weschke et al., Phys. Rev B 44, 8304 (1991) and references therein. [4] S. Danzenbächer et al., Phys. Rev. B 72, 033104 (2005). [5] D. V. Vyalikh et al., Phys. Rev. Lett. 96, 026404 (2006). Acknowledgements: This work was funded by the DFG, SFB 463, Projects TP B4 and TP B16, and the BMBF.

PE spectra were taken on- and off-resonance at 121 eV and 112 eV photon energies, respectively, in normal emission geometry. In order to extract the pure Ce 4ƒ contributions from these spectra, the off-resonance data were subtracted from the on-resonance spectra after proper normalization of the intensities. The resulting spin-resolved 4ƒ spectra are shown in the upper part of Fig. 2 together with the corresponding spin polarization P (inset) defined as P = (N ↑−N ↓) / (N ↑ + N ↓), where N ↑ and N ↓ denote the intensities of the majority (spin up) and minority (spin down) channels, respectively. The spectra reveal a double-peak structure of the Ce 4ƒ emission consisting of a main maximum at 2.2 eV corresponding to the ionization peak expected from the unhybridized 4ƒ1 ground state and a hybridization peak at the Fermi level, EF. From the weak intensity of the latter relative to the ionization-peak signal, a weak hybridization similar to the one in γ-Ce may be concluded as it is expected for a Ce surface layer [3]. Its intensity with respect to the ionization peak is, thus, a measure of the magnitude of the hopping probability and is expected to vary with spin orientation if the interaction depends on the latter. The most important observation is, however, that the intensity of the hybridization peak is larger for the minority- than for the majority-spin component (Fig. 2) indicating larger 4ƒ hybridization of the former. The spin polarization of both, the ionization and the hybridization peaks, gives a negative sign in P revealing that the preferred orientation of the Ce 4ƒ spins is opposite to the magnetization direction of the Fe layers. In order to simulate the observed variation of 4ƒ hybridization as a function of spin orientation we used a simplified approach to the periodic Anderson model (PAM) that was recently successfully applied to explain the angle-resolved photoemission spectra of CePd3 and Ce/W(110) [4, 5]. In this approach double occupation of the 4ƒ states is ignored (on-site ƒ-ƒ Coulomb interaction energy, Uƒƒ → ∞) and electron momentum conservation upon hybridization is assumed. For the hybridization matrix element we used calculated ƒ-projected local expansion coefficients of the Bloch functions around the

BESSY - Highlights 2006

Femtoslicing Special

Facility Repor t

User Pages

Fig. 1: Results of spin-polarized band structure calculations for an ordered monolayer of La on top of Fe(110) surface. Inset shows the diffraction image for the Ce/Fe(110) system and the structural model obtained after simulations of LEED.

rare-earth sites. Respective energy distributions of these coefficients at the Γ point were taken from a band-structure calculation of the system, in which Ce was replaced by La in order to exclude contributions of localized Ce 4ƒ orbitals (Fig. 1). Since these ƒ-contributions are formed by linear combinations of wave functions of the neighboring atoms (mainly Fe 3d) penetrating into the La atomic spheres, they reflect to some extent the energy and spin distribution of the latter and are rather different for majority- and minority-spin electrons. On the basis of these data spin-resolved Ce 4ƒ PE spectra were calculated using the energy of the unhybridized 4ƒ state and a hybridization constant as the only adjustable parameters. The results are presented in Fig. 2 (lower panel): The energy distribution of the PE intensity agrees well with that of the experimental spectra (Fig. 2, upper part). The minority-spin spectrum reveals high intensity of the hybridization peak due to the large density of the 4ƒ minority-spin contributions close to EF. On the contrary, no hybridization peak is obtained for the majority spin-direction due to an energy gap at the Fermi level for the respective valence-band states (Fig. 1). This particular result deviates from the experiment where a reduced but finite hybridization peak was observed. The deviation may be ascribed to the finite angle resolution of the experiment that samples also regions in reciprocal space where majority-spin bands cross EF. The calculated spin polarization (Fig. 2, inset in the lower part) reproduces qualitatively the energy dependence of the measured polarization. Particularly good agreement is obtained for points where the spin polarization changes its sign.

Fig. 2: Spin-resolved experimental (upper part) and calculated (lower part) Ce 4ƒ emission for Ce/Fe(110). Total, majority- and minority-spin photoelectron intensities are shown by solid circles, open and solid triangles, respectively. The insets show the corresponding spin polarization P.

Contact: Yu. S. Dedkov dedkov physik.phy.tu-dresden.de Page

17

Introduction

Scientific Highlights

News & Events

On or Off? Identifying isomerisation in molecular switches
P. M. Schmidt1, R. Püttner2, C. Kolczewski1, K. Hermann1, K. Horn1, T. U. Kampen1 The phenomenal rate of increase in the integration density on semiconductor micro-chips is driven by advances in optical lithography, that is, the process used for patterning in microelectronic device fabrication. Although the introduction of shorter-wavelength light sources and resolution enhancement techniques helps to maintain the current rate of device miniaturization, at some time in the near future this top-down approach can no longer attain the required feature sizes. Among several alternative techniques for the future, promising results have been obtained by the use of molecular systems [1]. Here, the advantages are easy fabrication and the possibility to shape and fine-tune organic molecules into desired structures by organic synthesis. Molecules with an extended π-electron system are especially promising candidates. Intermolecular interactions of the π-systems favor a stacking of molecules with their aromatic planes parallel to each other. In such highly ordered thin films, overlap between π-orbitals takes place only in one direction leading to an anisotropy of optical and physical properties, i.e. charge carrier mobility or optical absorption [2, 3]. A further advantage of organic materials is that the smallest building block, the molecule itself, represents an entity with a certain functionalization. For example, a molecule with two different possible conformations incorporated into one-dimensional 'molecular wires' would ideally serve as a 'molecular switch' (Fig. 1) The simplest model switch of this sort is the stilbene molecule (1,2-diphenylethylene) for which the phenyl groups can be arranged in two conformations, on the same side or on opposite sides around the central C=C bridge, i.e. to cis- and trans-stilbene, respectively (Fig. 1). This isomerisation can be triggered by UV irradiation as already shown by experiment and explained theoretically for the gas phase and in solution [5, 6, 7, 8]. Our goal is to find a spectroscopic fingerprint for the two isomers on the surface corresponding to the 'on' or 'off' position of this molecular nano-switch (Fig. 1). For molecular switching to be useful in microelectronics the switching pathway has to work on surfaces, i.e. in a constrained situation. However, the bonding with the substrate surface must not hinder the switching process. This is possible for adsorbed stilbene since bonding to a surface via the central C=C double bond seems to be most favorable with the phenyl rings allowed to rotate freely. To identify changes in the molecule due to the adsorption on the surface, we have taken Near Edge X-Ray Absorption Fine Structure Spectroscopy (NEXAFS) spectra from the molecule in the gas phase (i.e. prior to adsorption) and in the adsorbed phase on the surface accompanied by theoretical calculations (Fig. 2.). NEXAFS is an element-specific and surface sensitive method. Using anglular-dependent measurements the geometry of the adsorbed molecules can be determined. Thus, σ*-orbitals of the molecular backbone and π*-orbitals can be probed selectively. NEXAFS spectra of trans- and cis-stilbene in the gas phase (Fig. 3), show a prominent π*-resonance between 284 and 286 eV photon energy and several smaller σ*resonances at higher photon energies [12]. The σ*-resonances are found to be quite comparable (with only minor differences) due to the isomers possessing the same backbone. The π*-resonance, on the other hand, shows significant differences. While the cis-isomer displays a single sharp peak, the π*-resonance of trans-stilbene is split. Theoretical studies obtained by transition potential calculations with corrections for the ionization potential and relativistic effects

1 Fritz-Haber-Institut der Max-Planck-Gesellschaft, Berlin 2 Freie Universität Berlin

References: [1] R.L. Carroll, et al., Angew. Chem. Int. Ed., 41, 4378 (2002). [2] M. Friedrich et al., J. Phys.Condes. Matter, 15, S2699 (2003). [3] T.U. Kampen et al., Appl. Surf. Sci., 212-213, 501 (2003). [5] V.D. Vachev, et al., Chem. Phys. Lett. 215, 306 (1993). [6] R.J. Sension, et al., J. Chem. Phys., 98, 6291 (1993). [7] J.S. Baskin, et al., J. Phys. Chem., 100, 11920 (1996). [8] W. Fuß, et al., Angew. Chem. Int. Ed., 43, 4178 (2004). [9] P.M. Schmidt, et al., Surf. Sci. (2007), doi:10.1016/ j.susc.2007.01.044 [10] The program package StoBe is a modified version of the DFT-LCGTO program package DeMon, originally developed by A. St.-Amant and D. Salahub (University of Montreal), with extensions by K. Hermann and L.G.M. Pettersson. Acknowledgements: Supported by the SONS-program (MOL-VIC) of the European Science Foundation (ESF).

Fig. 1: Basic concept of an optically activated molecular switch.

BESSY - Highlights 2006

Femtoslicing Special

Facility Repor t

User Pages

Fig. 2: Calculated NEXAFS spectrum of trans-stilbene showing the contribution of each carbon to the overall spectrum. For some peaks the corresponding final states orbitals are shown.

using the StoBe code [13] confirm all spectral features observed in the experiment. In particular, the also π*-resonances are very well reproduced by the calculations; this can be used to clearly identify both isomers. For higher photon energies the calculations predict stronger signal intensities than found in the experiment. This may be due to the use of stronger broadening in the calculated intensities above the ionization threshold. Figure 4 shows NEXAFS spectra recorded for a monolayer of cis- and trans-stilbene. The intensity curves to the right refer to the π*-resonance for different photon incidence angles. The key features distinguishing between the isomers in the gas-phase spectra are also visible in the spectra of the adsorbed phase. The trans-isomer displays a split π*-resonance while that of the cis-isomer is sharp. Thus cis- and trans-stilbene adsorbed on Si(100) can be distinguished. The π*-resonance signal near 285 eV photon energy shows an asymmetric line shape which is different for the two isomers. The intensities of the π*-resonances upon changing the incidence angle of the monochromatized synchrotron light has been used to determine the orientation of the molecules with respect to the substrate surface. As shown in Figure 4 the intensity of π*-resonance of trans-stilbene increases with increasing incidence angle. From this qualitative behavior we can already conclude that this isomer is lying almost flat on the surface. The angular dependence of the π*-resonance of cis-stilbene, on the other hand, shows a different behavior. The intensity is high for normal incidence and decreases only slightly with increasing incidence angle. This angular dependence is also obtained in our theoretical studies on the most stable adsorption geometries of the isomers. The two structure models of this detailed analysis are shown in Figure 4, where both isomers are found to bond with the silicon dimer at the substrate surface via their central carbon double bond.

Fig. 3: Experimental and theoretical NEXAFS spectra of trans- and cis-stilbene in the gas phase.

Fig. 4: Experimental NEXAFS spectra for trans- and cis-stilbene adsorbed on Si(100). The angular dependence of the *-resonance is a consequence of the adsorption geometry of the respective isomer.

Our results demonstrate, that clear differences between the π*-resonance can serve as a fingerprint and permit a clear distinction between the cis- and trans-isomers, i.e. the 'on' or 'off' state of the molecular switch, which is essential for the detection and characterization of the isomerisation at the surface.

Contact: Philipp M. Schmidt philipp fhi-berlin.mpg.de

Page

19

Introduction

Scientific Highlights

News & Events

Source: Michael Peters

Charge transport in plastic electronics
M. Häming1, S. Hame1, J. Ziroff1, T. Boecking3, O. Seitz2, G. Gavrilla4, D. Batchelor1, A. Schöll1, F. Reinert1, E. Umbach1

1 Universität Würzburg 2 Weizmann Institute of Science, Israel 3 The University of New South Wales, Australia 4 TU Chemnitz

Organic electronic devices found their way into everyday live, as cheap, flexible, and energy saving LEDs and displays in mp3 players or mobile phones. However, for more sophisticated applications of organic semiconductors such as high efficiency plastic solar cells, organic lasers, or in nano-electronics based on single molecules, it is essential to gain a much better understanding of the electronic structure on a molecular level. Organic semiconductors often exhibit similar conduction mechanisms to inorganic semiconductors (e.g. charge-transfer complexes) including the presence of a hole and electron conduction layer and a band gap. In organic LEDs (OLEDs) both layers need to interact and hence exchange charge efficiently in order to emit light. In solar cells on the other hand photons induce charges in the semiconductor which have to be removed or transported very efficiently in the material to create an electrical current. How does charge occur and how does it move in organic materials? Common belief is that an electron or hole resides on one molecule ('localized') and hops from there to a neighbouring molecule ('hopping transport'). It carries a polarization cloud and perhaps a molecular distortion ('polaron') with it. The consequences of this picture are low mobility and significant energy differences between charge carriers at a surface as compared to the bulk. Actually, staying in this picture a charge should even avoid a surface for energetic reasons. Also the exciton binding energy ('electronhole interaction') or the energy required to separate charges in photovoltaics should strongly depend on whether the charges are localized within one or delocalized over several molecules. To tackle this important issue we create a strongly localized hole in the C 1s core level and measure the ensuing charge redistribution (screening) of the systems valence electrons. However, as the polarization cannot be measured directly, we compare the screening difference between surface and bulk molecules by comparing the binding energies of their photoelectrons. Since surface molecules have fewer neighbours than

bulk molecules the screening by polarization should be lower – and hence the binding energy higher – at the surface than in the bulk. This effect is known to be in the range of 0.3 - 0.4 eV [1]. As C60 is a good organic semiconductor, we chose it as one model system [2]. We prepared thin C60 films in-situ by organic molecular beam deposition in UHV. The probing depth of photoelectron spectroscopy was varied by tuning the electron emission angle and especially the photon energy that the mean free path of the emitted electrons was either minimal or maximal resulting in a surface or bulk sensitive experiment, respectively. With very high resolution we find for C60 no difference in the C 1s peak energies between surface and bulk (Fig. 2a).

Fig. 1: Side view at a 'ball and stick' model of alkyl chains adsorbed on a Si(111) surface. [6]

References: [1] E. V. Tsiper, et al., Chemical Physics Letters 360, 47 (2002). [2] B. Casu et al., PRL (in preparation) [3] A. Salomon et al., Phys. Rev. Lett. 95, 266807 (2005). [4] O. Seitz et al., Langmuir 22, 6915 (2006). [5] Ch. Zubrägel et al., Chem. Phys. Lett. 238, 308 (1995). [6] L. Segev et al., Phys. Rev. B 74, 165352 (2006). [7] A. Salomon et al., Adv. Mater., 19, 445 (2007). Acknowledgements: Supported by the BMBF. This work is part of a collaboration with the groups of D. Cahen and L. Kronik (Weizmann Inst.) and with T. Boecking (UNSW)

Fig. 2: A comparison between surface and bulk sensitive signals using C1s peaks of Alq3 (top) and C60 (bottom) shows that there is hardly any polarization difference. This leads to the conclusion that the screening charge is delocalized over several molecules.

BESSY - Highlights 2006

Femtoslicing Special

Facility Repor t

User Pages

For the very sharp C 1s signal of C60 the SCLS can be quantified to be lower than about 10 meV. This leads to the conclusion that polarization hardly contributes to screening in contrast to the statements in the literature [1]. Similar results are obtained for Alq3 (Tris-[8-hydroxyquinoline] Aluminium), a common component in the manufacture of OLEDs, which – unlike C60 – grows in amorphous structures (Fig. 2b). That implies that in contrast to the general picture charge is delocalised due to intermolecular interaction even in amorphous films in a way that the effective charge is spread over several molecules. This challenges the present models for polarization screening, charge transport between molecules (inter-molecular transport), and exciton binding energies in organic materials. In addition to the understanding of intermolecular transport, it is also essential to understand the charge transport within a molecule (intra-molecular transport), for the development of plastic electronics. We chose long-chain alkyl molecules which are self-assembled on a silicon surface (Fig. 1) to study such intra-molecular transport. Earlier highly reproducible transport measurements [3, 4] through such monolayers revealed that these materials have quite good transport properties in spite of their large electronic gap [6]. This excited our interest to study their intramolecular electronic band structure, which is expected to be simply due to a 1D Brillouin zone [5]. DFT calculations also suggested an influence of the interface on the charge transport [6]. We studied samples with various alkyl chain lengths (C12, C14, C16, C18) on Si(111) substrates, which have been prepared at the Weizmann Institute (Israel) and at the University of New South Wales (Australia) [3, 4]. First we utilised NEXAFS spectroscopy at the C-K edge with p- or s-polarized light to study the molecular alignment and orientation. The spectra show two resonances which are assigned to σ*(C-H) and σ*(C-C) bonds at 287 and 292 eV, respectively (Fig. 3). The evaluation of the polarization dependence revealed a 40° tilt angle of the alkyl chains with respect to the surface normal which is in the range found in the DFT-calculations [6] Varying the photon energy in the valence photoemission spectra (Fig. 4) leads to a change in momentum k of the emitted electrons along the surface normal and thus allows the study of dispersion along the alkyl chains. Indeed, we observe significant intramolecular dispersion, i.e. a 'molecular band structure'. Some of the features between 5 and 10 eV which originate from C2p orbitals

p-pol.

Fig. 3: Grazing (70°) incidence NEXAFS spectra of a Si/C14H29 sample, recorded in the partial electron yield mode with p- and s-polarization of the incident light. Evaluation of the polarization dependence results in a tilt angle of 40°.
70∞

s-pol.

hw

show a rather strong energy dependence on momentum. These electronic bands are more than 2 eV wide, and hence arise from strong intra-molecular interaction. Another interesting feature at 0 - 4.5 eV can be assigned to C-Si interface states by comparison with DFT-calculations [6]; these states may play a critical role in charge injection [7]. In the bulk of alkyl crystals where the interface has no influence this regime is part of the electronic gap. Taking together the results from earlier transport experiments, DFT calculations, and the present spectroscopic data it can be derived that charge transport along the chain is strongly influenced by molecular states induced by the interface. That means that the properties of the electronic states within the molecule, including band dispersion, depend on both the strength of intra-molecular interaction and on the interface bond. Thus, electronic transport through the molecules can be controlled by chemisorption at the interface.

Fig. 4: Normal emission PES spectra of a Si/C18H37 sample recorded with p-polarization of the incident light. The photon energy was tuned from 30 to 195 eV in 5 eV steps, and the spectra were normalized to the maximum of the C2p band. The Si-C interface states between 0 eV and 4.5 eV, play a critical role in charge injection and transport. In the bulk of alkyl crystals where the interface has no influence this regime is part of the electronic gap.

Contact: Marc Häming Marc.Haeming physik.uni-wuerzburg.de Page

21

Introduction

Scientific Highlights

News & Events

How safe is toxic waste in a glass cage?
F. Pinakidou, M. Katsikini, E.C. Paloura

Aristotle University of Thessaloniki, Greece

Fig. 1: Glass (left) and glass-ceramic (right) waste contaminated products. The glasses have a shining, dark color while the vitroceramics have a graded color due to surface crystallization.

Glass is very fragile – an annoying experience everyone made who dropped a full bottle. Why would one use glass of all things to store toxic waste? The secret is that the glass is not only the cage, but the waste itself is transformed into glass (Fig. 1). The process is called vitrification. It is an alternative to the common incineration, where the residues of the combustion are stored in tanks or disposed in landfilling areas where they can generate groundwater contamination and weathering problems. Making them to glass is a rather simple and cost effective method for the management of Fe and Pb containing toxic wastes that originate from petroleum storage tanks. The process produces glass and vitroceramic materials which can homogeneously incorporate into their matrix various toxic elements [1]. The vitrification process basically consists of mixing and co-melting the toxic waste with silica and soda and quenching of the melt to form solid materials. The final products can be freely disposed, or used for construction and even decorative applications. [2,3]. Nevertheless the question arises: How durable are these glasses? Here we focus on the determination of the nanostructure and the long term stability of vitrified Fe- and Pb-rich waste-containing stabilized products by means of X-ray absorption fine structure (XAFS) measurements [4]. XAFS is an atom selective, non-destructive characterization technique which provides information on the coordination environment (nearest neighbor distances and coordination numbers) of the absorbing atom. The capabilities of XAFS can be further extended by using capillary optics, which reduce the size of the X-ray beam down to 1-5 µm and thus enhance the spatial resolution of the experiment. The use of the capillary optics also permits the two dimensional mapping of the distribution of various elements in the samples using X-ray fluorescence (XRF) spectroscopy. Thus, by applying the XAFS technique in pre-selected spots of a sample (on the basis of the XRF maps), we can determine the effect of the inhomogeneous distribution of certain elements on their nanostructure.

A very interesting issue that needs to be addressed is the fatigue of waste glasses. Annealing at high temperatures promotes the evolution of fatigue-related effects and thus mimics the effect of long term exposure of the products to the environment. In addition to that, annealing is used for the production of glass-ceramic materials which exhibit exceptional mechanical properties compared to their glassy counterparts. The XRF maps of two solid glasses that contain 60 wt% waste, subjected to annealing at Tann = 600 and 800°C are shown in Fig. 2a and b respectively, while that of an as-casted solid product containing 70 wt% waste is shown in Fig. 2c. It is clearly seen that regions with high and low Fe concentration are grown in the bulk of the samples while the distribution of Pb is homogeneous. Therefore, since Pb is depleted from the Fe-rich inclusions it can be proposed that the Fe-rich inclusions are mainly formed by Fe-oxides. µ-XAFS spectra were recorded from the regions with high (H-) and low (L-) Fe concentration of the annealed glass to determine whether the inhomogeneous distribution of Fe affects the nanostructure of the Fe atoms. The corresponding Fourier transforms (FT), which have clearly different profiles, are shown in Fig. 3. The only peak that is resolved in the spectra from the L-regions indicates that these regions are glassy [5, 6]. On the contrary, the structure that appears in the distance range 2-4 Å in the FT of the spectrum from the H-region is a signature of crystallinity [7, 8]. Indeed, fitting the spectrum from the high-Fe region reveals the presence of magnetoplumbite (PbFe12O19) microcrystallites embedded into a glassy matrix. Hence, annealing at high Tann destroys the homogeneity of the glass: Fe and Pb belong to both the glassy matrix and to crystalline inclusions. In order to determine the percentage of the Fe atoms that belong to the glassy and crystalline phases we recorded XAFS spectra with the capillary optics removed. In this case the beam size is larger and thus the collected information is averaged over a larger sample volume. The FTs of the Fe-XAFS spectra recorded from the two annealed samples are shown in Fig. 3.

References [1] K.E. Haugsten et al., Waste Manage., 20 167 (2000). [2] G. Scarinci et al., J. Eur. Ceram. Soc., 20 2485 (2000). [3] G.H. Zheng et al.,, Environmental Progress, 15 283 (1996). [4] 'EXAFS: Basics, Principles & Data Analysis', B.K. Teo, Springer-Verlag, Berlin (1986). [5] F. Pinakidou et al., J. NonCryst. Solids, 352, 2933 (2006). [6] F. Pinakidou et al., J. NonCryst. Solids, 351, 2474 (2005). [7] F. Pinakidou et al., Nucl. Instr. Meth. B, 246, 238 (2006). [8] F. Pinakidou et al., Nucl. Instr. Meth. B, 246, 170 (2006). [9] 'Crystal Structures' vol. 2, R.W.G. Wyckoff, John Wiley & Sons, New York (1964).

BESSY - Highlights 2006

Femtoslicing Special

Facility Repor t

User Pages

a) Fig. 2: µ-XRF maps of two stabilized industrial glasses subjected to thermal treatment at (a) 600 and (b) 800oC, respectively. The maps picture the inhomogeneous distribution of Fe into the waste products. (c) µ-XRF maps of a product containing 70wt% waste showing the distribution of both Fe and Pb. Notice the Pb is homogeneously distributed in the same area where Fe forms Fe-rich islands.

b)

The fitting was performed assuming that a fraction of the Fe atoms is bonded in tetrahedra in the glass [6] while the rest occupies octahedral sites in microcrystallites [9]. It is revealed that the increase in the annealing temperature does not affect the extent of crystallinity, i.e. the same percentage of Fe (~80 at%) belongs to the glass. After annealing at Tann = 600°C, only magnetoplumbite crystallites are formed, which render the material safe for the stabilization of Pb. However upon increasing the Tann both Fe-rich and mixed Pb and Fe-oxide microcrystallites are formed. This has a detrimental effect on the stability of the product since the former crystalline regions grow with the depletion of Fe from the later, enabling in this way Pb to escape from the material. Thus, the type of microcrystallites grown as a result of thermal treatment affects significantly the stability of the product. To conclude, the chemical stability of the vitrified solid products is strongly correlated to the structural role of both Pb and Fe. The parameters that need to be monitored during the vitrification process are the concentration of the toxic waste, the processing conditions and the chemical composition of the final products. It is disclosed that the production of vitreous materials is possible when the waste content is less than 65 wt%. Process dependent inhomogeneities in the distribution of the heavy, toxic elements and the type of crystalline phases affect their long term stability. Proper exploitation of our results may lead to safer disposal of Fe- and Pb-rich industrial waste.

c)

Fig. 3: Fourier transforms (FT) of the Fe-K µ-XAFS spectra recorded from the H- and L-regions of the glass with 60 wt% waste, annealed at 600oC and of the Fe-K XAFS spectra of the same glass after annealing at the same temperatures. The experimental curve and the fitting are shown in thin and thick solid lines, respectively.

Contact: Eleni Paloura paloura auth.gr

Page

23

Introduction

Scientific Highlights

News & Events

Source: Bill Adler, www.billadler.web

Package induced deformation of integrated circuits
R. Tilgner1, P. Alpern1, B.R. Müller2, A. Lange2, M. Harwardt2, M.P. Hentschel2

1 Infineon Technologies, Munich 2 Bundesanstalt für Materialforschung und -prüfung (BAM), Berlin

Microelectronics makes live easier and safer. In modern automobiles micro electronic power devices (Fig. 1) are used to control e.g. the window-lift motors, mirror adjustments, automated locking systems, and the footwell lighting. Besides the elevated comfort the safety is also increased, because the driver can concentrate on driving and is not distracted too much by opening or closing the car windows. But the increasing implementation of car electronics makes the vehicle also vulnerable to microelectronics failure. During his “life” a car is usually subjected to all kinds of environmental changes like differences in temperature and humidity or mechanical stress, therefore, electronic components need to have a reasonable robustness in order to match the longevity car. A major reason for microelectronics failure is due to the way these components are fabricated (Fig. 1): Within a microelectronic component containing an integrated circuit (IC) the interconnect structures establish a link between switching entities of the silicon

chip and the component’s outer pins designated to become soldered to a printed circuit board (PCB). The mechanical interaction between package and chip induces deformations which may become starting points of mechanical flaws [1-3], and hence bears the risk of passivation cracking over the metal lines and therefore the failure of the device because of corrosion. The main reason for deformation of the chip is the layered structure of the overall system chip and package. To spread the heat from the chip switching currents of ~1A the silicon is soldered at 328°C by lead on a copper heat slug (the use of lead is regulated by ROHS [4]). Cooling down to room temperature the shrinking copper will distort the thermally more inert silicon chip. In a similar way a mold compound is used to fix and to cover the construction. It is processed at 175°C as a liquid which shrinks and hardens after cooling down. As these interactions of different materials and temperatures influence the stability of the device, modern quality and reliability development ask for the degree of these interactions. The knowledge of the real deformation of the crystal is essential for calculations and predictions concerning nanofracture mechanics and hence to develop a device package with minimized IC deformation. To visualize the real deformation of the crystal we developed a new inspection method at the materials research beamline (BAMline) [5]. The method – Synchrotron Laue Contrast Radiography (SLCR) – is based on the coherent scattering of monochromatic synchrotron radiation by the net planes of the silicon single crystal of the electronic device (Fig. 2). X-rays will be deflected by a single crystal if d (the spacing between the planes in the atomic lattice of the single crystal), λ (the wavelength of the synchrotron beam) and Θ (the angle between the incident beam and the scattering planes) fulfil the Bragg condition nλ = 2d·sinΘ. For a perfectly flat single

Fig. 1: Micro electronic power device for switching currents of up to 1 A.

References: [1] P. Alpern et al., IEEE Trans. CPMT A17(3), 583 (1994); Erratum in 18, 862 (1995). [2] P. Alpern et al., VTE 10(1), 10 (1998). [3] Z. Suo, Comprehensive Structural Integrity (I. Milne, R.O. Ritchie, B. Karihaloo, Editors-inChief), Elsevier 2003 [4] Official Journal of the European Union, L37/19-23, Annex, point 7, 2003 [5] Müller, B.R. et al., Nucl. Inst. and Meth. in Phys. Res. A, 467/468, 703 (2001).

Fig. 2: Illustration of the pseudo absorption of single crystals. Left: X-rays are deflected across the whole flat single crystal due to diffraction. No Xrays arrive at the detector (the radiograph would be dark). Right: for curved crystals the Bragg condition is only fulfilled in local parts. This leads to dark bands in the radiograph (local pseudo absorption, see Fig. 2, right). The position of the dark band can be moved by rotating the crystal.

BESSY - Highlights 2006

Femtoslicing Special

Facility Repor t

User Pages

Fig. 3: Radiograph of the bonding wires, (left) without and (right) with dark bands, respectively

crystal the condition is fulfilled across the entire crystal simultaneously (Fig. 2, left). Thus the whole radiograph becomes dark, but not because of absorption (pseudo-absorption). In flexural loaded single crystals the scattering planes are distorted. Thus the Bragg condition is only fulfilled in some local areas of the crystal (Fig. 2, right). This leads to more ore less small dark bands in the radiograph. The shape and the width of the local pseudoabsorption bands depend on the shape and the radius of curvature of the flexural loaded crystal, respectively. To explore the final shape of the whole chip the microelectronic device is placed in front of a CCD camera system (pixel size 14.4 µm x 14.4 µm) and irradiated by a parallel and monochromatic beam (55 keV). The surface normal of the chip is parallel to the beam axis and the beam cross-section covers the entire chip. The measurement is performed by rotating the device around an axis perpendicular to the beam while recording in parallel the radiographs of the device. In most cases the CCD camera displays the radiograph of the absorbing gold bonding wires (Fig. 3, left). But for specific orientations of the device dark bands across the silicon crystal become visible (Fig. 3, right). By rotating the curved crystal (angle scan range 0.11°, step width 0.001°) the dark bands are moving across the radiograph and assemble a series of dark band positions (isoclines) across the whole crystal (aggregated in a 2D plot in Fig. 4). The isoclines density represents the local radius of curvature of the crystal, typically, ranging between 5 m and 50 m (the higher the density the smaller the radius of curvature). From the 2D isoclines plot we derived a 3D surface profile of the flexural loaded Silicon crystal with a very high resolution of about ±1 nm in height (Fig. 5). A 3D surface profile of a chip just after packaging is shown in Fig. 5, left. As mentioned, the temperature dependent shrinkage of the heat slug which is situated beneath the profile exceeds that of the chip, and the mold compound leading to a smooth convex shape of the chip. The amplitude between the minimal and maximal height is about 1,500 nm, which may

Fig.4: Aggregation of all dark band positions (isoclines) during rotation (angle range 0.11°, step width 0.001°). In the background the radiograph of the chip is shown. In the centre part of the radiograph a lead void is visible by the brighter grey coloured area.

not harm the functionality of the electronic device. A second device was measured after a stress treatment of 1,000 temperature cycles from −55°C to +150°C. The shape has changed to a more irregular, 'magic-carpetlike' profile with a reduced height amplitude of about 300 nm (Fig. 5, right), which is possibly due to a partial loss of the adhesive strength between the package and the chip caused by degradation e.g. of the solder during stress treatment. But what are the detection limits for more minute mechanical influences of the package. We found a void within the lead solder from X-ray absorption radiography. The SLCR measurement also discovered the mechanical effect of the gold wire bonds and the lead solder void on the surface of the silicon crystal. Finally, even higher order effects such as interaction between void and bonding wires have been detected. From these results SLCR can be considered as a new powerful inspection method to evaluate even minute degradation processes in electronic devices. This is an essential for any prognostic reliability work in the future.

Fig. 5: 3D surface profiles (lateral resolution: 14.4 µm x 14.4 µm, height resolution: ±1 nm) deduced from the isoclines plots (Fig. 4). The Cu heat slug is situated beneath the profile. Left; chip just after packaging, right; chip after a stress treatment of 1,000 temperature cycles from −55°C to +150°C.

Contact: Bernd R. Müller bernd.mueller bam.de

Page

25

Introduction

Scientific Highlights

News & Events

© Books Illustrated, Illustration by Susan Scott

Soon to become physicists’ best friend? Blue diamond superconductivity in the light of terahertz spectroscopy
M. Ortolani1, U. Schade1, L. Baldassarre2, S. Lupi2, P. Calvani2, Y. Takano3, M. Nagao3, T. Takenouchi4, H. Kawarada4
1 BESSY 2 Università di Roma 'La Sapienza', Italy 3 National Institute for Materials Science, Tsukuba, Japan 4 Waseda University, Tokyo, Japan

Diamonds are transparent, lustrous, sparkling, and as we all know 'the girls best friends'. Even more attractive and expensive are the very rare natural 'blue diamonds'. For a scientist diamond is carbon, where each atom is covalently bonded to four other carbon atoms in a compact crystal structure to generate the largest insulating gap. This means that diamond is transparent to infrared, visible and ultraviolet light up to 5.5 eV and that diamond does not conduct electricity. Blue diamonds contain a small amounts of boron impurities. Being a p-type semiconductor blue diamonds can weakly conduct an electrical current if the boron concentration reaches 0.5% (Fig. 1) [1]. At higher boron contents it held a surprise in hand, which might make blue diamonds soon to be also 'physicists' best friend'! Nowadays, diamond crystals and films can be produced in a laboratory (even if the beauty and the size of natural gems cannot be equalled) with a technique called microwave-assisted chemical vapour deposition (MW-CVD). Blue diamond has also been produced in several labs. Due to its properties boron-doped diamond is a possible candidate for substituting silicon in the future generation computer chips. However, the way to the 'diamond computer' is still long: n-type diamond is not easily available yet and the behaviour of charge carriers travelling in a diamond crystal still needs to be understood. In 2004, it came as a surprise for the scientific community that weakly conducting boron-doped diamond is even a superconductor at liquid-helium temperature (4.2 K) for boron-concentrations above 3%. Diamond films grown at the NIMS in Tsukuba showed a Tc as high as 11 K. Superconductivity is a quantum state of matter where electrical currents are free to run without friction (zero electrical resistance), found in some metals such as tin, aluminium, lead, and niob, below a critical temperature TC of few K. The Bardeen, Cooper and Schrieffer (BCS) theory explains superconductivity in metals with the pairing of electrons via the electron-phonon interaction. The ground state below Tc is a 'Cooper

pair' condensate, and the first excited state is the breaking of a Cooper pair, requiring an energy 2Δ. In the case of weak BCS interaction, all the above described quantities are related to each other by: 2Δ = 3.53 kBTc ≈ hω0 exp(-1/λ) (1) where ω0 is an Einstein phonon frequency, and λ<< 1 is a coefficient describing the strength of the electron-phonon interaction. Not all superconductors are well described by the BCS theory, which holds for metals. Highly-doped semiconductors like the high-Tc copper oxides also have a superconductivity state, where 2Δ seems uncorrelated with a Tc ranging up to 160 K. The mechanism for superconductivity in the high-Tc copper oxides is still unknown; it is then legitimate to ask ourselves if we can describe boron-doped diamond within the BCS theory, or not. At University of Rome we measured by Fourier-transform infrared spectroscopy (FT-IR) the reflectivity R (Fig. 2a) of a MW-CVD-grown boron-doped diamond film, produced at NIMS with a Tc of 6 K. The sample surface had a size of 2.5 x 2.5 mm, one of the largest to be produced with the necessary homogeneity in the boron-content. Through Kramers-Kronig analysis, the presence of electron-phonon interaction is demonstrated by the abrupt increase for E = hω > 130 meV of the scat-

References: [1] Y. Takano et al., Appl. Phys. Lett. 85, 2851-2853 (2004). [2] L.Boeri et al., Phys. Rev. Lett. 93, 237002 (2004). [2] M. Abo-Bakr et al., Phys. Rev. Lett. 90, 094801 (2003). [4] M. Ortolani et al., Phys. Rev. Lett. 97, 097002 (2006). [5] T. Yokoya et al., Nature 438, 647-650 (2005). Acknowledgements: Supported by the EU via the IASFS funds. We thank J.S. Lee for valuable help and P. Kuske, G. Wustefeld and J. Feikes for tuning the storage ring during the “low-alpha” shifts.

Fig. 1: Resistivity curves show that B-doped diamond is a hole-doped semiconductor. For high enough B concentration, it also becomes a superconductor below 4 - 8 K.

BESSY - Highlights 2006

Femtoslicing Special

Facility Repor t

User Pages

tering rate Γ(ω) (Fig. 2b) and a peak at hω = 130 meV in the electron-phonon interaction spectrum α2F(ω), indeed corresponding to a Raman-active phonon of diamond [2]. However, to check Eq. 1 we still needed to determine the optical gap 2Δ. In principle, this can also be done with the FT-IR technique, since R(T2 k direct beam

MPG MPG BESSY PTB PTB

6T-WLS 0.4T-LFD U49-1

Litho 13° PGM

20 - 1,900 85 - 1,600 85 - 1,600 85 - 1,600 120 - 1,700 60 - 1,300 715 / 786 / 861 60 - 1,300 ~ 250 - ~ 600 170 - 1,800 20 - 700 5 - 180 4.5 k - 17.7 k 4.5 k - 17.7 k 4.5 k - 17.7 k 85 - 1,600 85 - 1,600

PTB BESSY BESSY Uni Wü HMI BESSY BESSY FZJ BESSY BESSY BESSY BESSY FUB BESSY BESSY BTUC

UE49 UE52

PGM a/b SGM PGM

D. Batchelor (Uni Wü) Ch. Jung Th. Schmidt (Uni Wü) D. Schmitz (HMI) H. Rossner (HMI) T. Kachel J.-S. Schmidt A. Firsov S. Cramm (FZJ) Ch. Jung O. Schwarzkopf R. Follath U. Müller M. Fuchs (FUB) U. Müller O. Schwarzkopf P. Hoffmann (BTUC) J.-S. Schmid W. Braun M. Mast F. Eggenstein St. Dorausch F. Senf Ch. Stamm I. Packe

UE46 UE56-1

PGM PGM a/b ZP-mono SGM

U41 U125-1 7T-WLS-2

TXM/STXM PGM PGM 1 PGM 2 a/b BL 14.1 BL 14.2 BL 14.3

BESSY/Uni Gö P. Guttmann (Uni Gö) G. Schneider

U49-2

PGM 1 PGM 2

BESSY - Highlights 2007

Femtoslicing Special

Facility Repor t

User Pages

Dipole Beamlines at BESSY (January 2007)
Further information on beamlines can be found on the ’user info pages’ of the BESSY website
PTB PhysikalischTechnische Bundesanstalt HUB HumboldtUniversität zu Berlin FUB Freie Universität Berlin Uni P Universität Potsdam BAM Bundesanstalt für Materialforschung und -prüfung

Monochromator Energy Range (eV)
Litho EUV KMC-1 IR stations ISISS Litho 1m-NIM 1 1m-NIM 2 KMC PGM HE-SGM Optics-BL KMC 2 PGM 3 3m-NIM EDR TGM 4 TGM 7 PGM-RD-BL CP-NIM
��
�

Operated by
BESSY BESSY BESSY FHI BESSY PTB PTB PTB PTB PTB CRG BESSY BESSY BESSY BESSY Uni P BESSY HMI Uni Pet BESSY

Contact Persons
H.-U. Scheunemann M. Gorgoi U. Schade M. Hävecker (FHI) A. Knop-Gericke (FHI) B. Löchel H.-U. Scheunemann M. Richter (PTB) R. Thornagel (PTB) R. Thornagel (PTB) M. Krumrey (PTB) F. Scholze (PTB) A. Lippitz (BAM) F. Senf A. Erko T. Kachel G. Reichardt W. Leitenberger (Uni P) K. Godehusen C. Pettenkofer (HMI) W. Bremsteller (HMI) Y. Dedkov (Uni DD) S. Molodtsov (Uni DD) F. Schäfers M. Mertin I. Packe F. Eggenstein I. Packe A. Erko M. Mast O. Schwarzkopf T. Blume M. Bednarzik H. Köhrich H. Köhrich M. Mertin

95 1.7 k - 10 k Thz - 0.5 80 - 2,000 direct beam 3 - 35 3 - 35 white beam 1.7 k - 10 k 30 - 1,800 200 -700 20 - 2,000 4.5 k - 15 k 20 - 1,900 4 - 35 2 k - 12 k 8 - 120 8 - 120 30 - 1,500 4 - 35
��
�

HMI Hahn-MeitnerInstitut Uni DD Universität Dresden Uni Pet University St. Petersburg, Russia

��

�

��

�

���

�

���

�

��

��

������������������������������������������������������������

��

��

���� �����
��
��

��

������������� ����� �����������

��

�

��

��

��

��

��

��

���������� ���������������������������� �

Rapid prototyping for rapid pulses: Highly flexible mirror holders for the fs-laser line of the Femtoslicing source.

��

�

��

�

��

�

��

�

� ���

� ���

�������������������

Page

59

Introduction

Scientific Highlights

News & Events

Experimental Stations
LIQUIDROME is an experimental station dedicated for soft X-ray NEXAFS investigations on liquid and degassing sample. The experiments can be carried out under various conditions: • pressure can be varied between 10-6 mbar and 1,000 mbar of inert atmosphere (typically Helium) • temperature can be varied between -50°C and +100°C
LIQUIDROME cell for volatile liquids

• electric and magnetic fields can be applied. A window located on a nozzle close to the region of interaction separates the synchrotron radiation from the working conditions inside the chamber. Different window materials like Si3Ni4, Aluminum or Polyimide are available and can be changed in-situ. The fluorescence yield is recorded via a GaAsP-diode. In addition, the electron yield can also be measured if required. Non-volatile liquids can be studied in a jet, with a jet speed between 100 ml and 1,000 ml per minute. For volatile liquids, a sealed cell is available, where the fluid circulates behind a Si3Ni4-window (see Fig). A setup to measure stop-flow chemical reaction kinetics can be implemented. The chamber is quite spacious to allow for special detectors or sample environments to be inserted by users. For the data analysis, a bundle of software packages is available to calculate the electronic structure of the sample under study. Using the software codes StoBe, FEFF8, Multiplet and Gaussian03, NEXAFS or EXAFS spectra can be derived on the basis of structural models. Furthermore, information on molecular energies and structure or on bond and reaction energies can be retrieved (see graph).

ISISS (Innovative Station for In Situ Spectroscopy), a project of BESSY and the Inorganic Chemistry department of the FHI (Fritz-Haber-Institut der Max-Planck-Gesellschaft), is dedicated to study the interaction of surfaces with their environment (i.e. gas/ solid interfaces) under conditions equal to or close to reality allowing a detailed characterisation of catalytic relevant materials under realistic conditions to push forward rational catalyst design. The PGM design offers the possibility to easily adapt the beamline to the users need (e.g. high flux, high resolution, or high purity of the X-rays). It delivers photons in the energy range between 80 eV and 2,000 eV. The monochromator holds two gratings (600 l/mm, 1,200 l/mm). Fast photon energy scans are feasible (“Quick NEXAFS”) due to the implementation of a continuous driving mode of the monochromator. A typical spot size at sample position is around 150 µm (horizontal) x 100 µm (vertical, depending on exit slit size). The high-pressure X-ray photoelectron spectroscopy (XPS) end-station of the FHI is equipped with differentially pumped electrostatic lens system attached to a modified hemispherical electron analyser (PHOIBOS 150). This set-up allows both in situ XPS and X-ray absorption spectroscopy measurements (total electron and Auger electron yield mode) in the presence of a reactive gas (p ~1 mbar). The ambient gas phase can be analysed simultaneously to the spectroscopic characterisation of the sample by electron impact mass spectrometers (QMS) and a proton transfer reaction mass spectrometer (PTR-MS). Heating of the sample up to approx. 1,000 K is provided by a NIR-laser from the rear. The station is equipped with a sample transfer system that allows fast exchange of samples via a load lock. The experimental end-station is placed inside a ventilated lab with safety sensors for H2 and hydrocarbons allowing the use of hazardous substances. The lab provides up to eight high purity grade gases as well as extra space for sample preparation and chemicals storage. Regular operation of ISISS will start in the 2nd semester of 2007.

NEXAFS spectra of Na in ethanol and water (E.F. Aziz et al., Phys. Rev. B, 73, 075120 (2006))

Contact LIQUIDROME: Emad Flear Aziz Bekhit emad.aziz bessy.de

Contact ISISS: Axel Knop-Gericke knop fhi-berlin.mpg.de Michael Hävecker mh fhi-berlin.mpg.de

ISISS endstation

BESSY - Highlights 2007

Femtoslicing Special

Facility Repor t

User Pages

Permanent experimental stations
Experiment
THz spectroscopy IR ellipsometry
Ind BioSR BioSR, Ind

Contact
holldack bessy.de schade bessy.de hinrichs isas-berlin.de baumgaer rz.uni-potsdam.de antje.vollmer bessy.de weinelt mbi-berlin.de hermann.duerr bessy.de s.borysenko ifw-dresden.de
Ind

Location
IR IR IR 3m-NIM optics beamline UE112 PGM 1 UE56-1 PGM 1 UE112 PGM 2b U49-2 PGM 2 UE49 PGM a UE49 PGM b UE46 PGM ISISS PGM optics beamline U41-XM 6T-WLS, DIP 06 KMC 1 U125-2 KMC 7T-WLS-1 KMC 2 7T-WLS-2 7T-WLS-1 7T-MPW 7T-MPW

IR-spectroscopy and -microscopyBioSR, Ind UVIS - protein circular dichroism spectroscopy Two-Photon-Photoemission Experiment fs-slicing facility 1 -ARPES ultra high resolution photoemission
3

SURICAT - photoelectron and absorption spectroscopy

SAMIC - spectroscopy and microscopy integrating chamber SPEEM - spin resolved photoemission microscopyplanned

patrick.hoffmann tu-cottbus.de florian.kronast bessy.de thomas.schmidt physik.uni-wuerzburg.de schmitz hmi.de knop fhi-berlin.mpg.de schaefers bessy.de guttmann bessy.de loechel bessy.de mihaela.gorgoi bessy.de braun pdi-berlin.de zizak bessy.de erko bessy.de umue bessy.de

SMART - spectro-microscope with highest spatial resolution 7T high-field end station ISISS in-situ catalysis beamline Reflectometry
Ind Inds

XM - X-ray MicroscopyBioSR, Ind X-ray lithographyInd HIKE- Hard X-ray high kinetic energy photoelectron spectroscopy X-ray diffraction during MBE µSpot micro-XANES, -EXAFS, -fluorescenceBioSR Diffraction, XANES, EXAFS Protein crystallography
BioSR, Ind Ind

BAMline - nondestructive testing in analytical chemistry EDDI - energy dispersive diffraction MagS - resonant magnetic scattering

heinrich.riesemeier bam.de genzel hmi.de feyerherm hmi.de

Variable experimental stations
Experiment
HIRES - high resolution electron spectrometer PHOENEXS - photoemission and near edge X-ray spectroscopy High resolution spinpolarisation photoelectron spectroscopy Stored Nano Particels So-Li-AS - solid-liquid-analysis system VUV/XUV ellipsometry
Ind

Contact
rader bessy.de varykhalov bessy.de c.m.schneider fz-juelich.de ruehl phys-chemie.uni-wuerzburg.de mayerth surface.tu-darmstadt.de esser isas-berlin.de eugen.weschke physik.fu-berlin.de schoenhe mail.uni-mainz.de emad.aziz bessy.de anna.zimina bessy.de cissy hmi.de szargan rz.uni-leipzig.de ruediger.mitdank physik.hu-berlin.de schaefers bessy.de
Ind

BioSR: suitable for biological samples Ind: suitable for industrial users

MUSTANG - multi-user stage for angular resolved photoemission gavrila bessy.de

Scattering experiments in the VUV/XUV-range Photoemission microscope for ps time resolved spectroscopyInd LIQUIDROME for NEXAFS on liquids and degassing samples Soft X-ray emission spectrometer CISSY - CIS- diagnostic using Synchrotron radiation ROSA - rotateable spectrometer apparatus Fluorescence spectroscopy Polarimetry
Ind

Page

61

Introduction

Scientific Highlights

News & Events

������������������������
����������� ������ ������������������� ��������� ��������� ���������� ����������

���������� ���������

Supervisory Board:
Prof. Dr. J. Treusch (Chairman) Prof. Dr. E. Umbach (Vice-Chairman) Frau Dr. B. Vierkorn-Rudolph Prof. Dr. E. O. Göbel Prof. Dr. R. Maschuw Prof. Dr. H. Braun Frau Dr.Ulrike Gutheil Prof. Dr. J. Schneider MinR. N. Barz Prof. Dr. M. Steiner Senatsdirig. W. Eckey Forschungszentrum Jülich Universität Würzburg Bundesministerium für Bildung und Forschung ������������ PTB Braunschweig ��������� Forschungszentrum Karlsruhe Max-Planck-Gesellschaft München Technische Universität Berlin DESY Hamburg Bundesministerium für Wirtschaft und Technologie Hahn-Meitner-Institut Berlin Senatsverwaltung für Wissenschaft, Forschung und Kultur

����������� ��������

��������������

Scientific Advisory Committee:
Prof. Dr. Robert Schlögl (Chairman)) Prof. Dr. Hans Ade Prof. Dr. Vlasta Bonacic-Koutecký Prof. Dr. Reinhard Brinkmann Prof. Dr. Thomas Elsässer Prof. Dr. Roger Falcone Prof. Dr.- Ing. Norbert Langhoff Prof. Dr. Nils Martensson Prof. Dr. Eckhard Rühl Priv.- Prof. Dr. Ilme Schlichting Prof. Dr. Hans-Peter Steinrück Prof. Dr. Metin Tolan Prof. Dr. Richard Walker Permanent Guests Dr. Josef Feldhaus Prof. Dr. Helmuth Möhwald Prof. Dr. Jürgen Richter Dr. Rainer Schuchhardt Fritz-Haber-Institut der MPG Berlin North Carolina State University Humboldt Universität zu Berlin DESY Hamburg Max-Born-Institut Berlin ALS Berkeley (Institutional Member) Institut for Scientific Instruments Berlin MAXLAB, Sweden Freie Universität Berlin Max-Planck-Institut für medizinische Forschung Heidelberg Universität Erlangen-Nürnberg Universität Dortmund Rutherford Appleton Laboratory (Diamond), Chilton HASYLAB/DESY Hamburg Max Planck Institut für Kolloid- und Grenzflächenforschung, Potsdam Bundesministerium für Bildung und Forschung Senatsverwaltung für Wissenschaft, Forschung und Kultur Berlin

Beam Time Committee
Prof. Dr. Wilfried Wurth (Chairman) Prof. Dr. Lorenz Singheiser (Vice-Chairman) Prof. Dr. Stefan Blügel Prof. Dr. Ralph Claessen Prof. Dr. Thomas Elsässer Prof. Dr. Karsten Horn Dr. Michael Meyer Prof. Dr. Anke Rita Pyzalla Prof. Dr. Liu Hao Tjeng Prof. Dr. Hartmut Zabel Sub-Committee Protein Crystallography Prof. Dr. Ralf Ficner Prof. Dr. Peter Lindley Dr. Matthias Wilmanns Universität Hamburg Forschungszentrum Jülich Forschungszentrum Jülich Universität Würzburg Max-Born-Institut Berlin Fritz-Haber-Institut der MPG Berlin Centre Universitaire Paris-Sud Max-Planck-Institut für Eisenforschung Düsseldorf Universität zu Köln Ruhr-Universität Bochum Universität Göttingen Universität Lissabon EMBL Hamburg

Financial Committee
Ass.jur. S. Lettow (Chairman) R. Kellermann (Vice-Chairman) Dr. W. Buck M. Schleier Dr. R. Schuchardt H. Görres Forschungszentrum Karlsruhe Forschungszentrum Jülich PTB Berlin Max-Planck-Gesellschaft München Senatsverwaltung für Wissenschaft, Forschung und Kultur Berlin Bundesministerium für Bildung und Forschung

User Committee
Ralph Püttner (Chairperson) Birgit Kanngießer (Vice-Chairperson) Peter Baumgärtel Wolfgang Braun Ralf Feyerherm Freie Universität Berlin Technische Universität Berlin Universität Potsdam Paul-Drude-Institut Hahn-Meitner-Institut

BESSY - Highlights 2007

Femtoslicing Special

Facility Repor t

User Pages

U49/2 -PGM -1

M-2

U4 9/2

ru
C

sN

dt

Ex E x it I it II

M PG

-PG

P-

IM

Structural biology

1A 6- B -1 -1 IP 16 D IPD

TG M4 TG M7

C-1 itho KM V-L EU -1A -01 1B -2 DIP 01-01 7T-WLS-1 DIP DIP

IR
ID-02-1 ID-02-2

DIP-02-1A DIP-02-1B

µ-focus beamline BAMline

DIP-02-2
U12 5-2
ID DIP -03 DIP 03-1A -03 -1B

ID DIP -15 DΠ15-1A P-1 5- 1 B DIP -15 -2

GRAD-line

MAD-line

MAX-line

U49

DIP

-03 -2

5m

-NIM

U12
10

-2

DIP-14-2

5/2-S GM mP NIM DI-K

MC

PW -M 7T

P U112

ID-14 DIP-14-1A DIP-14-1B

2 PG U11
2a M-

7T-WLS-2

2 11 UE

D IPIP 04 -0 -1 4- A 1B D IP -0 42

D

-1 4 -2 -0 04 IDDI

GM-1

R ED

M-2b

PG

M

ag

S

D ED

I

DIP -13 -2 -1B -13 DIP -13-1A DIP -13 ID

UE 2 56-2 PM 05 ID- -1A -05 -2 DIP 5-1B -05 -0 6T-WLS DIP DIP
b ssi ly ta ca

1 U4 -P G

U4

1X

M
3m

m ea

M

11 UE

lin

-N

e

IM

2

-1

soft-WLS

D

12 -2 D D IPIP 12 ID 12- -1B -1 1A 2
DIP-06-2
UE 56 /1 UE 56/1 -SGM -ZP UE M 56/1 -PG M-a

IP -

ID-06 DIP-06-1A DIP-06-1B

UE 56 /2PG 1 M-

UE5 6/2-P

GM

U

41
-11 -2
DIP DIP -07-1B -07 ID- 1A 07 U18 U49 0 -1

-2

PM -3

DIP

Lithography Lithography

UE56/1-PGM-b

DIP

-07

DIP DIP -11-1B -11 ID- -1A 11

fs-G UE enera 56-1 to U13 r 9

-2

Det.-NIM

STXM

UE46-PGM-2

UE-46

UE46-PGM-1

DIP

-09

2

C KM
M 2-PG UE5 M -SG 52 UE

-2

-1B -09 DIP 09-1A DIP -09 ID

UE

52

D

IP

-0

2 8-

PTB -ID WL PTB -PG M P

PTB -SX PTB 700 PTB -KMC -DW L

Full fi

DIP-10-2
eld X M

DIP-10-1B DIP-10-1A ID-10

1B 8- A -0 -1 IP 8 D IP-0 -08 D ID

E4 U

9

SM A

O

-PG M-a

R

TP

ic pt

G

ea -b

µ-

M

fo

Q-N

-b

m

e lin

HE

s cu

IM

SG

-S

M

TB-ID WL2

G

in operation under construction planned

M
SPE

Scientific Director Prof. Dr. Dr. h.c. Wolfgang Eberhardt Secretary: Ines Maupetit phone +49 (0)30 / 6392 4633 fax +49 (0)30 / 6392 2989 wolfgang.eberhardt bessy.de, ines.maupetit bessy.de Technical Director Prof. Dr. Eberhard Jaeschke Secretary: Dr. Nikoline Hansen phone +49 (0)30 / 6392 4651 fax +49 (0)30 / 6392 4632 eberhard.jaeschke bessy.de, nikoline.hansen bessy.de Administration Thomas Frederking Secretary: Katrin Rosenblatt phone +49 (0)30 / 6392 2901 fax +49 (0)30 / 6392 2920 thomas.frederking bessy.de, katrin.rosenblatt bessy.de

Beamtime Coordination Dr. Walter Braun, Dr. Gerd Reichardt Secretary: Stine Mallwitz phone +49 (0)30 / 6392 2904 fax +49 (0)30 / 6392 4673 beamtime bessy.de User Office Ines Drochner, Cornelia Stürze phone +49 (0)30 / 6392 4734 fax +49 (0)30 / 6392 4746 useroffice bessy.de Public Relations Gabriele André, Dr. Heike Henneken, Dr. Markus Sauerborn phone +49 (0)30 / 6392 4921 fax +49 (0)30 / 6392 4972 pr bessy.de

EM

Credits For providing photographs and drawings we would like to thank: Hendrik Spohler / laif (Front cover) Kai Godehusen, Berlin Alastair Rae, Northern Ireland Books Illustrated, Illustration by Susan Scott, UK Bill Adler, Washington, USA Bureau International des Poids at Mesures, Paris, France Cartoonstock.com Euler Hermes, Hamburg Infineon Technologies, München Juan Salmotal Franco, Barcelona, Spain Keith Petrie, Auckland, NZ Michael Peters, Kürten Mihaela Gorgoi, Berlin National Academy of Sciences, USA Physikalisch-Technische Bundesanstalt, Braunschweig Silko Barth, Berlin Page

63

Good

b y e

H e i

k e

Published by: Berliner Elektronenspeicherring-Gesellschaft für Synchrotronstrahlung m.b.H. – BESSY Albert-Einstein-Straße 15 12489 Berlin, Germany phone +49 (0)30 / 6392 2999 fax +49 (0)30 / 6392 2990 www.bessy.de info@bessy.de Board of Directors: Prof. Dr. Dr. h.c. Wolfgang Eberhardt, Prof. Dr. Eberhard Jaeschke, Thomas Frederking Editors: Dr. Heike Henneken, Dr. Markus Sauerborn Layout: Annette Weber, Stitz & Betz GmbH, Berlin ISSN Number: 1611-6127

Cover Photo by Henrik Spohler / laif

Berliner Elektronenspeicherring-Gesellschaft für Synchrotronstrahlung m.b.H.
Member of the Leibniz Association

Berliner Elektronenspeicherring-Gesellschaft für Synchrotronstrahlung m.b.H.

Highlights 2006

Highlights 2006

Berliner Elektronenspeicherring-Gesellschaft für Synchrotronstrahlung m.b.H.
Member of the Leibniz Association
        
Top of page

Note to user

Dear user,

In response to current developments in the web technology used by the Goobi viewer, the software no longer supports your browser.

Please use one of the following browsers to display this page correctly.

Thank you.