Photon Diagnostics for the VUV-FEL

Transcription

1 Photon Diagnostics for the VUV-FEL R. Treusch for the VUV-FEL team In order to fully underst the performance of the VUV-FEL at DESY, a large variety of photon diagnostics tools has been developed over the last years. An ideal detector for the VUV-FEL covers the full dynamic range of about 7 orders of magnitude from spontaneous emission to SASE in saturation, is suitable for single-pulse measurements (response time < 100ns), exhibits low degradation under radiant exposure in the VUV, must be ultra-high vacuum compatible suitable for being assembled under clean room conditions, such as the rest of the linear accelerator (in the following for brevity called linac ) the attached photon beamlines. For the characterisation of the VUV-FEL properties, we adopted to a large extent concepts that worked already successfully during the FEL operation around 100nm at the TESLA Test Facility phase 1 in the years [1,2]. In addition, new approaches were required for the VUV-FEL user facility since the user experiments dem online diagnostics that exhibit a minimum interference with the FEL beam. Minimum interference in this context means that one should, respectively can not use detectors that either block the FEL beam, like e.g. photodiodes, or detectors that alter crucial beam properties such as coherence or intensity to an undesirable extent. In addition to the data acquisition at the experiment, one has to provide a shot-to-shot information on the beam properties in order to underst the subtleties of the physics that happen during the interaction of matter with the FEL beam. Whereas these online diagnostics for the user facility have been described in detail in the 2003 annual report [3], we will focus here on the photon diagnostics which are used to optimize the VUV-FEL performance which guarantee that the beam is accurately transported across the beamline towards an experiment. Most of these diagnostics are already implemented in the machine control system DOOCS [4] are routinely used by the VUV-FEL (linac) operators to tune FEL beam properties such as intensity, pointing direction spectrum. The VUV-FEL single-shot spectrum can be monitored with a grating spectrometer that was built in collaboration with INFM Padova [5]. The spectrometer setup at the end of the linac tunnel is sketched in Figure 1. Top view (end of linac tunnel) FEL beam spectrometer deflecting mirror Top view Entrance Slit α β Meridional plane (spectral plane) Intensified CCD (inside lead) Spectrometer (grating + premirror) F Side view F Sagittal plane (spatial plane) piezo actuated entrance slit SVLS grating Spherical mirror Figure 1: Sketch photo of spectrometer setup for single shot FEL spectra. In the technical drawing at bottom left, left right are flipped. For more details see insets ref. [5]. 159

2 It consists of a spherical VLS grating with a spherical premirror. A gated intensified 1280x1024 pixel CCD camera views the spectrum on a Ce:YAG fluorescent screen which is positioned in the exit (focal) plane of the spectrometer. With exposure times down to 3ns the camera is well suited for the acquisition of single shot spectra, even at the future minimum FEL pulse spacing of 111ns. The resolution of the spectrometer, that reaches values of λ/ λ up to 1500, can be controlled with a piezo-controlled entrance slit. From the characteristic spikes in the spectrum [6], one obtains the number of modes in the pulse can also derive the pulse length. The determination of the wavelength helps to accurately tune it to the experimental needs, which will soon be possible at the VUV-FEL by varying the electron beam energy while keeping high SASE output intensity. Recording the spectrum at the same time can be used for cross-checking the electron energy. The intensity of the individual pulses is measured with a multitude of detectors, such as a gas monitor detector [7], an MCP based detector [8], thermopiles, PtSi Photodiodes, etc.[1]. While the latter two detector types can only be applied within a limited intensity range (see [1] comments in figure 3 below), the first two are routinely applied from spontaneous undulator emission towards SASE (self amplified spontaneous emission) in saturation, covering a dynamical range of about 7 orders of magnitude. The MCP detector uses a gold mesh with 65% transmission or alternatively a thin (50 µm diameter) gold wire to reflect part of the VUV-FEL beam onto a carefully calibrated MCP. Keeping the MCP in the low gain regime, one can determine the SASE gain with respect to spontaneous emission knowing the applied MCP voltage. The GMD detector, on the other h, is a state-of-the-art ionization chamber that determines absolute photon numbers by monitoring both, the ion electron current that result from the interaction of the FEL beam with either nitrogen or a rare gas at low pressures of typically 10-6 mbar. Four GMDs with differential pumping stages at both ends are integrated into the FEL beamline on the way into the experimental hall. They have been absolutely calibrated to an accuracy of about 15% by the PTB, the German institute of stards. The principle of the GMD is depicted in figure 2. Faraday cup Ions Drift tube 10-9 hpa 10-5 hpa hν e - Differential pumping Interaction volume Figure 2: Principle of the gas monitor detector (GMD), a calibrated ionization chamber to determine the VUV-FEL intensity in absolute photon numbers [7]. The MCP /or the GMD are applied by the machine operators to optimize SASE, determine the shot-to-shot intensity statistics, proceed on the way towards saturation. They also serve for a normalization of the signals at the experiments. For most of the experiments, the GMD is the preferred intensity monitor, since it only absorbs a negligible number of photons. Furthermore, the GMD does not alter any beam properties, in contrast to the MCP detector that employs an Au mesh which produces often unwanted diffraction, resulting in an inhomogeneous intensity distribution at the sample position. Figure 4 shows the diagnostics panel of the GMD in the control system. This panel is available to the machine operators as well as all experimenters shows the pulse to 160

3 pulse intensity distribution derived from the electron current, as well as the averaged signal from the ions that can be used to monitor the long term SASE fluctuations. Figure 3: Diagnostics panel in the control system showing the yield of the gas monitor detector. In the upper window, the averaged ion current is depicted, whereas the lower window shows the bunch by bunch electron signals (both in µj). The panel has been adapted from a similar control panel for the MCP detector, which was continuously developed refined since the first FEL operation in phase 1 of the TESLA Test Facility. In addition to spectrum intensity, it is crucial to monitor FEL beam position profile. This is required to ensure pointing stability optimum focusing at the experiment. Presently, a Ce:YAG fluorescent crystal, incorporated in one of the photon diagnostic detector units (figure 4) is used to determine the beam pointing with help of a laser engraved cross. The cross centre position, as well as all other detector positions, has been carefully calibrated with help of the DESY surveyors. Centring the beam on the cross ensures that it can propagate across all mirrors in the beamlines towards the experiment. Behind each mirror, another fluorescent screen is positioned monitored by a CCD camera which is remote controlled via FireWire (IEEE1394). Fibre coupled red diode lasers which possess a high pointing stability are set up collinear with the surveyed FEL axis run across all mirrors to serve as alignment lasers for the setup of the experimental chambers at the end. These alignment lasers also define reference points on the Ce:YAG fluorescent screens that the FEL beam is supposed to hit as well. Figure 5 shows an overview photo of the photon diagnostics setup described above. 161

4 Ce:YAG crystal PtSi photodiode (1nJ < E pulse < 1µJ) 100 mm Thermopile (E pulse > few µj) Double slits in thin foil (diffraction) rear front Figure 4: Water cooled detector unit for the VUV-FEL. This compact setup allows to measure pulse energy, beam position profile -- with the double slits -- the degree of transverse coherence. Note that the Ce:YAG crystal in this unit has a scattering screen with a cross mounted behind it. The crystal with the laser engraved cross referred in the text, sits in the second detector unit about 5m behind this one. It is used to image the double slit diffraction pattern. For more details see figure insets text. Detector Unit F2 (apertures, detectors, mirror) Intensity + beam profile + diffraction (coherence) +deflection into spectrometer Detector Unit F1 (apertures + detectors) Intensity + beam profile + double slits (coherence) FEL electrons Beamline for synchrotron radiation from dipole magnet pulse arrival time (evaluated with streak camera in exp. hall) Grazing incidence grating spectrometer with intensified CCD Single shot spectrum MCP detector (Au wire/mesh + MCP) Intensity + beam profile Figure 5: Overview of the VUV-FEL photon diagnostics setup at the end of the linear accelerator tunnel, just behind the last dipole magnet that deflects the electrons into the dump (distances ranging from 247m to 260m behind the electron gun). 162

5 Another important property of the VUV-FEL is its pulse timing. To fully exploit the extremely short pulse length of the FEL of about 100fs or below, e.g. in a pump-probe experiment, it is absolutely matory to know the shot-to-shot timing of the FEL pulses to the same accuracy. The concepts for efficient time resolved experiments are described in the article by S.Düsterer et al. in the same section of this annual report. The photon diagnostics hardware is complemented by software tools that are to a large extent already implemented in the machine control system DOOCS [4]. One of the main goals was to hide the sometimes complicated geometry technical details behind a graphical user interface (GUI) that should be usable by most of the (experienced) operators, like the rest of the machine control of the VUV-FEL (linear accelerator). The DOOCS servers + GUIs (written by T. Nunez, HASYLAB many colleagues from DESY-MVP) give almost intuitive control via simple buttons supported by graphical sketches (see e.g. figure 3 above, figures 6 7 below). In our case, ONLINE (written by T. Kracht, HASYLAB) is controlling a major part of the diagnostics beamline hardware (stepper motors, encoders, oscilloscopes, etc.). For camera servers some other tools, stards from the linac have been adopted partially modified according to our needs. Figure 6: Main control panel for the VUV-FEL photon diagnostics in the linear accelerator tunnel. The top view onto the diagnostics units is sketched all the individual functions are clickable what opens the underlying menus. Figure 7: Control panel for beamlines, diagnostics optical laser in experimental hall. 163

6 In summary, the photon diagnostics described above are essential for a lot of purposes: Shot-to-shot online diagnostics of properties such as intensity, beam position profile, spectrum timing are matory for most user experiments. Comprehensive photon beam characterization helps in understing the subtleties of SASE serves as a benchmark for advanced start-to-end-simulations from the electron gun in the beginning towards the SASE process at the end. Combining experimental photon beam characterization start-to-end-simulations allows to deduce important parameters of the electron beam is hence a complementary tool to electron beam diagnostics as well. References [1] R. Treusch, T. Lokajczyk, W. Xu, U. Jastrow, U. Hahn, L. Bittner J. Feldhaus, Nucl. Instrum. Meth. A 445, (2000) [2] V. Ayvazyan et al., Eur. Phys. J. D 20, (2002) [3] K. Tiedtke in HASYLAB Annual Report 2003, Section Technical Developments at HASYLAB, subsection VUV-FEL K. Tiedtke, J. Feldhaus, Ch. Gerth, U. Hahn, U. Jastrow, E. Plönjes, B. Steeg R. Treusch, AIP Conf. Proc. 705, (2004) [4] http.//doocs.desy.de [5] L. Poletto, L. Epuli, P. Nicolosi, M. Pelizzo, P. Zambolin, J. Feldhaus, U. Jastrow, U. Hahn, E. Ploenjes K. Tiedtke, SPIE Proc. Vol. 5534, Fourth Generation X-Ray Sources Optics II, (2004) P. Nicolosi, L. Poletto, M.-G. Pelizzo, L. Epuli, P. Zambolin, J. Feldhaus, U. Jastrow, U. Hahn, E. Ploenjes, K. Tiedtke, Proc. VUV2004, to be published in J. Electr. Spectrosc. Relat. Phenom. [6] E. L. Saldin, E. A. Schneidmiller M. Yurkov, The Physics of Free Electron Lasers, Springer, Berlin-Heidelberg (2000) [7] A. A. Sorokin, S. V. Bobashev, J. Feldhaus, Ch. Gerth, A. Gottwald, U.Hahn, U.Kroth, M.Richter, L.A. Schmaenok B. Steeg, K. Tiedtke R. Treusch, AIP Conf. Proc. 705, (2004) M. Richter, A. Gottwald, U. Kroth, A. A. Sorokin, S. V. Bobashev, L. A. Shmaenok, J. Feldhaus, Ch. Gerth, B. Steeg, K. Tiedtke R. Treusch, Appl. Phys. Lett. 83, (2003) [8] A. Bytchkov, A. A: Fateev, J. Feldhaus, U. Hahn, N. I. Lebedev, E. A: Matyushevskiy, E. L. Saldin, E. A. Schneidmiller, A. V. Shabunov, K. P. Sytchev K. Tiedtke, R. Treusch M. V. Yurkov, Nucl. Instrum. Meth. A 528, (2004) [9] 164