.

.

The BEAR (Bending magnet for Emission, Absorption and Reflectivity) beamline started operation at the beginning of 2003. The beamline aims to provide a number of spectroscopic tools including optical absorption and emission, reflectivity and photoemission to investigate the mechanisms determining the basic properties of the electronic states in solids or molecules. The conception of BEAR was inspired by the whole range of possibilities offered by the use of a photon beam with energy ranging from visible to soft x-rays in the joint study of the electronic and structural properties of a system. Although optical absorption and valence band photoemission are the natural tools to investigate such properties, experience has shown that because of the complexity of the electronic properties of interacting electrons the simple observation of molecular orbital and/or conduction or valence states is in general not sufficient to provide unambiguous insights into the mechanisms governing the finer details of the electronic properties together with the functional properties of systems. The use of higher photon energies give access to core level excitations which allow elemental and chemical selectivity, open up the possibility of disentangling filled from empty states and to derive the local atomic geometry in the modelling of diffraction and interference processes which determine the line-shape of the X-ray absorption edges (NEXAFS and EXAFS – Near edge/X-ray Extended Absorption Fine Structure) and the angular distribution of photoelectrons (PED – Photo Electron Diffraction). Beside this the core level spectroscopies give access to a number of crucial properties ranging from the study of not equivalent chemical sites of the same species, the binding energy shifts of surface core level and atom resolved magnetic moment. Moreover the investigation of anisotropic systems having reduced symmetry, together with the study of magnetic properties, calls for the use of polarised light and the possibility of implementing different scattering geometries. For all the above reasons the apparatus is based on a bending magnet as a source, beamline optics delivering photons from near visible up to about 1600 eV with selectable degree of ellipticity and an end station featuring, in an UHV environment, a high flexibility of scattering geometries and surfaces and ultra thin film and interfaces preparation facilities. Furthermore a spot size with lateral dimensions of the order of a few tens of microns in dimension is provided to ensure grazing incidence of use in surface physics. The bending magnet source features a continuous and smooth spectral distribution of intensity ensuring considerable photon flux and delivering, as a function of the emission angle with respect to the orbit plane, photons with variable ellipticity – ranging from linear polarization up to a circular polarisation degree as high as 80%. The energy range – near visible up to about 1600 eV - covers the range from interband transitions up to the optical absorption of K edges of use for structural investigation including B, C, N, O, F, Na, Mg, the L_2,3 edges of 3d Transition Metals and the M_4,5 edges of the  Rare Earths. This energy range allows also valence band spectroscopy and core level photoemission to be performed, so that binding energy measurements and photoelectron diffraction are possible. Eventually, as far as the flexibility in scattering geometries is concerned, both the photon interaction processes and the emission processes (photon and/or electron) depend on the incidence and collection geometries. The flexibility of the end station permits the implementation of scattering geometries where the direction of the incident photon and the collection geometries of the scattered or emitted particles (photon or electron) can be chosen with a wide flexibility.  The end station features movable electron (hemispherical) analyser and photon detectors (e.g. diodes) allowing angle resolved photoemission (band mapping and dependent selection rules) and angle-resolved optical reflectivity (specular, q-2q scans and diffuse). These characteristics are of use in the case of anisotropic systems including low symmetry bulk materials, surfaces, interfaces and chemisorbed molecule and magnetic systems (e.g. dichroic absorption or reflectivity). In the following a brief presentation of the beamline, of properties of the beam spot at the sample and an account of the features of the end station are given.

.

The photons are emitted at the 8.1 bending magnet of ELETTRA. The optic accepts 3.3 mrad in horizontal and 3.1 mrad in vertical. Fig. 1(a) shows the main characteristics of the light source. At the angle y defined in Fig.1 (a) the resultant light beam is in general elliptically polarized. Different choices of y result in different ellipticities because of the variation of the relative weight of vertical with respect horizontal polarized fields. The two electric fields vertical (b) and horizontal (c) are out of phase ± p/2 according to the sign of the take-off angle y and their analytical expressions are given by [1]: 

.

In Fig.1 (b) and (c) it is shown that y = 0 gives linear polarisation, while  y > 0 gives ellipses with right circular polarisation (RCP) and y < 0 ellipses with left circular polarisation (LCP). The ellipticity increases with y.

.

.

The BEAR optics work in grazing incidence, at an angle of 2.5° . The transport and focussing of the source is achieved by placing the source in the focus of a first parabolic mirror, P1. The outgoing parallel beam is dispersed by the monochromator working in parallel light, based on the plane grating – plane mirror scheme [2], using three gratings and covering the 5 – 1600 eV photon energy range; one grating (1200 l/mm) working in near normal incidence (NIM) covering the 5-40 eV, the other two (1200 l/mm, 1800 l/mm) working in grazing incidence covering the 40 – 1600 eV energy range. Different overall deviation angles (inclusion angle) between plane grating and plane mirror can be chosen. The dispersed radiation is focussed onto the exit slit by a second parabolic mirror, P2. The outgoing radiation is focused onto the sample by an elliptical re-focussing mirror. All the aspherical optical elements work in sagittal focussing to reduce the effects of slope errors in the dispersive plane.  The resolving power is of the order of about 3000 in the whole photon energy range. The typical dimensions of the light spot at the sample are 30x150 mm² . The refocusing mirror is followed by an I₀ section containing a W mesh (cleaned by high temperature annealing by Joule heating)  to monitor the radiation flux, and a couple of Si and Al filters (0.40 mm thickness), which are employed to reduce the higher-order contributions in the energy region below ~100 eV. Two modes of selection of the polarisation ellipse are available corresponding to different selections of the y angle. They include a vertically moving parallel slit (the polarisation selector), with variable aperture, downstream the re-focussing mirror and a vertical moving blade upstream of the first mirror. Fig. 2a shows the simulated bandwidth as a function of the photon energy for various vertical exit slit values (set at 20 mm ´ 900 mm, vertical x horizontal) compared with the experimental photon resolution values at some gas absorption lines. In Fig. 2 b the vertical profile of the zero order beam is reported. The experimental curve is compared with the profiles simulated with the project values and with the best effort values of the slope errors of the parabolic mirrors.

.

.

A summary of the experimental values of energy windows for the grazing channels with 1200 l/mm (G 1200) and 1800 l/mm (G 1800) gratings and the NIM channel is reported in table 1.

 .

table 1: experimental values of energy windows DE (meV)

.

.

In Fig.3a we present an overview of the photon flux for the near normal incidence and grazing incidence (1200 l/mm) gratings as measured with a storage ring current of 200 mA and 50 mm overtical exit slits. In Fig. 3b is reported the measured flux for the G 1800 l/mm grating of the grazing channel at different inclusion angles with a stored current of 100 mA. The vertical exit slit aperture was 30mm.

.

.

The polarization characteristics (major and minor axes of the ellipse and phase) were determined at a number of selected photon energies by using the multilayer polarimeter available at ELETTRA and at a single energy (major and minor axes of the ellipse), by measuring the dichroic signal from a magnetized sample. In fig. 4 the ellipses obtained at a photon energy of 100 eV are shown as a function of the selected vertical portion of the beam (see insets). They provide the vertical and horizontal field of the beam evaluated by the Stokes parameter provided by the polarimeter for the different polarization selector position: fully open; central window and half beam. The measured Stokes parameters corresponding to the ellipses in the order are: (full beam) S1=-0.9, S2 = 0.011, S3 = -0.068;  d =-1.4 r; (horizontal window) S1=-0.97 S2=0.011  S3 = 0.082, d =-1.44 r; (half beam) S1=-0.77  S2 = 0.08  S3 = -0.57, d = 1.43 r. The major and minor axes of the ellipse (ellipticity, e, or degree of circular polarisation, P) were determined by measuring the dichroic signal at the L_2,3 edge of Co in a magnetically saturated sample ( Co(15 Ml)/Cu(001) ). The absorption spectra are reported in fig. 5, in the inset the kind of beam selection used is also reported. The resultant degree of circular polarisation, P, is 70% ± 5% or, equivalently, the ellipticity is e = 0.41.

.

.

.

A photon beam position monitor [3] (BPM) is installed upstream the first optical element of the beamline. This functions using the principle of a four quadrant diode. It allows continuous monitoring of the source position during the beamline operation and data acquisition. It features an absolute overall accuracy in the beam vertical position in the laboratory frame not worse than 0.1 mm and a sensitivity in differential mode of the order of 0.1 mm. A careful correlation and calibration between ELETTRA’s vertical beam position, as obtained from machine data, and BPM readings was carried out. The results demonstrated the reliability of the BPM. This allowed a quantitative correlation between the vertical position of the source and the energy selected by the monochromator; e.g. the test, in close agreement with simulation, showed at 860 eV an almost linear energy shift » 2 eV per mm of source movement. The reliability of the BPM opens in this sense important perspectives allowing automatic correction of delivered energy according to beam vertical position readings.

The end station is based on two UHV chambers coupled in UHV, the first serving as a preparation chamber the second one as a scattering chamber. The preparation chamber is equipped with surface cleaning facilities including ion sputtering and annealing, an evaporation section featuring four ports for evaporation sources and a quartz microbalance, low energy electron diffraction and a cylindrical mirror electron analyser equipped with a coaxial electron gun. The scattering chamber [4] is equipped with an hemispherical electron analyser and photon detectors. Both detectors cover the solid angle above the sample. Rotation of the light ellipse with respect to the sample is made possible by chamber rotation around the beam axis. Chamber rotations performed with the beam selector positioned as in the middle case of Fig. 4 allow to switch from s to p incidence conditions. A load lock for insertion from atmosphere is available. The provided spectroscopies and structural tools include angle resolved and angle integrated photoemission (valence bands and core levels), photoelectron diffraction, X-ray optical absorption (drain current, partial yield and fluorescence yield), specular  and diffuse reflectivity. A schematic top view of the station is shown in Fig.6.

.

.

acknowledgment:

Stimulating and helpful discussion with Ing. R. Baruzzo of the CINEL firm is kindly acknowledged.

.

references:

[1]  see e.g. “Synchrotron radiation, sources and applications”, ed.G.N.Greaves and I.H.Munro, Proc.13^th Scottish Univ. Summer School in Physics, 1985, SUSSP publications;

[2] G.Naletto, M.G.Pelizzo, G.Tondello, A.Giglia and S.Nannarone, “The monochromator for the synchrotron radiation beamline XMOSS at ELETTRA", SPIE Proc.4145,(2001),105;

[3] A.Giglia et al., to be published;

[4] L. Pasquali, A. DeLuisa, S.Nannarone, “The UHV experimental chamber for optical measurements (reflectivity and absorption) and angle resolved photo-emission of the BEAR beamline at ELETTRA”, to appear in the Proceedings of the SRI 13 Conference, San Francisco 2003.

.

.