Nanospectroscopy and NanoESCA monochromator: performance after upgrade

Aim of the monochromator upgrade project

In early 2014, an internal project was presented to the Elettra - Sincrotrone Trieste (E-ST) management to upgrade the monochromator of the Nanospectroscopy and nanoESCA beamlines. The main aim of the project was to improve the stability of the instrument, which appeared to be particularly sensitive to temperature variations in the experimental hall and the radiative load by the photon beam. Transients used to be in fact observed after opening the beamline to the undulator light or when changing the undulator gap.

The upgrade project was approved in summer 2014 and a call of tender was opened in fall 2014. The call was won by CINEL Scientific Instruments s.r.l., which signed the contract with E-ST in March 2015. The monochromator was opened for inspection in late June 2015 and was dismounted for maintenance shortly afterwards. The new parts were installed at the end of October and the instrument baked shortly after optical alignment was carried out. The monochromator was commissioned during December 2015 - January 2016.

Description of the improvements made

The monochromator was upgraded in various parts:

  • New in vacuo Renishaw Series Tonic encoders were installed in place of the old Heidenhein optical encoders. The new sensors allow a direct measurement of the tilt angle of the monochromator mirror and gratings. The encoders reach an angular resolution of just a few tens nrad, which in turn permit to correct temperature-induced drifts.
  • The monochromator mirror received a new holder. The new mounting system allows side-cooling. Heat exchange with the absorbers has been greatly improved with respect to the previous holder, where the mirror was contacted to a Cu absorber placed on its bottom face. This solution allows minimizing mirror deformation upon illumination under intense undulator light and minimizes transients.
  • New holders were installed for both VLS 200 and VLS 400 gratings. Both holders host piezo actuators allowing in situ fine alignment of the grating roll. In this way, the optics can be aligned so that the photon beam does not move when changing photon energy or grating.
  • All grating holders received new absorbers, larger Cu braids and improved clamps connecting the optics to the water cooling circuit.  
  • The cooling system was re-designed and new anti-vibration water pipes were installed in place of the old ones. The grating carriage was enforced by installing an anti-vibration frame.
  • The grating carriage rails were found worn and damaged. It was therefore necessary to have them replaced by new ones.
  • The coating of the monochromator mirror was found to be damaged (ablated) by prolonged exposure to the intense photon beam of our Sasaki Apple type II undulators. It was therefore necessary to have the mirror cleaned, fine-polished and re-coated with a 50 nm thick Au film. 
  • New motors were installed in place of the old five-phase Berger-Lahr stepper motors; a new dedicated in-house built controller (YAMS) and software was developed to substitute the old one.

A brief report on the tests carried out during the commissioning phase is provided below.


Short term stability

The short term energy stability of the monochromator was tested by changing the radiation load impinging on its optical components, mirror and grating. The test was carried out using the most severe operation conditions, 2 GeV ring energy and 310 mA ring current. In the experiment, we monitored the evolution of the W 4f core level emission as a function of time, checking for possible variations in the W 4f peak energy. Shortly after starting the data acquisition, the beamline shutter was closed for about one hour and then re-opened. Raw experimental data are shown in the Figure on the left hand side, which represents an area plot of the W 4f spectrum as a function of time (y axis). As can be seen from the constant energy of the W 4f doublet,

there are no transients after the shutter has been re-opened. Note also that the W 4f peak energy has not changed from the value it had before closing the shutter. The data acquisition rate was 30 s per spectrum. This value therefore represents an upper limit for the duration of (possible) transients. 

Long term stability

We tested the long term stability of the monochromator by monitoring the energy of the W 4f core level emission as a function of time. The W 4f spectra were collected at photon energy of 400 eV, during a time interval of about 14 hours after the beginning of the measurement. To compensate for temperature induced changes of the monochromator energy, the mirror and grating were re-positioned every 12 s. The series of W 4f core level spectra was fitted to quantitatively determine variations in the peak position. The statistical analysis of the results reveals that the peak energy exhibits a standard deviation of just 13 mV. The deviations seldom observed. are likely due to variations of the vertical position of the photon in the experimental

chamber. The stability of the monochromator thus appears definitely improved with respect to the situation prior to the upgrade. Such good performance is achieved by the fine control over the angles of both grating and mirror, through their accurate repositioning. Without such "feedback", larger variations (about 0.5 eV) were observed in other experiments, which we could correlate to changes in the temperature of the cooling water or the experimental hall. The final version of monochromator control software will offer an option to control and stabilize the energy of the photon beam.

Accuracy and linearity of the energy readout

Resonant XPS is a powerful technique to detect changes in the valence state of atomic species at surfaces and interfaces. In these experiments, the spectrum of a well-defined core level is acquired as a function of the varying photon energy. The actual photon energy must precisely match the set value. Our new in-vacuo encoders permit a fine control on the tilt angle of both mirror and grating. Compared to the previous ex-situ linear encoders, there is no coupling to the rotation axis of the motor, resulting in a one-to-one relationship between the increase of the photon energy and that of the kinetic energy of the photoelectrons. This can be clearly appreciated in the figure on the left hand side, which shows

a proof-of-principle experiment where photon energy and electron energy are vaired in equal amounts. THe figure shows an area plot of the W 4f core level emission as a function of the increasing photon energy. Note that the W specimen was oxidized at high temperature prior to the experiment. Four peaks can be therefore identified in spectra, corresponding to two spin-split W 4f doublets ascribed to the bulk (low binding energy) and oxide (high binding energy) components.

Precision in setting the desired energy

The monochromator control unit allows the user to reach different levels of precision in positioning the optics. Selectable via the software interface, precision level 2 and 3 produce a very accurate movement, the actual energy of the monochromator  (calculated from the readout of the optical encoders) differing only few meV from the set (desired) value. In the energy range from 90 to 140 eV, see the graph on the left hand side, we measured differences less than 5 meV from the desired value. Statistical analysis of the data shows that the variance of the difference between set and measured photon energy is 2.7 and 2.3 meV for precision level 2 (standard) and 3 (high precision), respectively. 

 

Energy reproducibility after displacing the grating carriage

A critical aspect in the everyday use of the monochromator is the lack of reproducibilty in setting the photon energy after changing the gratings. Typically this happens when going from one to another grating and back to the first one. In this regard, we noted a notable improvement after replacing the rails of the grating carriage. Obviously, to obtain the best reproducibility, one needs to position the grating so that the photon beam impinges on the same optical region. The measurment position must be carefully set always moving the grating carriage in the same verse, for instance towards the accumulation ring, taking care to release strain from the linear transfer rod. The area plot in the figure shows W 4f core level

emission spectra acquired before and after displacing and re-positioning the VLS 400 grating, as indicated by the blue arrows, which has been repeated for four consecutive times. The position of the W 4f7/2 peak, deterimend by fitting the spectra, indicates that the energy reproducibility is better than 70 meV at photon energy of 500 eV.

Energy resolution

The energy resolution of all gratings was evaluated using the NanoESCA and SPELEEM microscopes. To test the SG10 and VLS200 gratings, an Ag surface was used. Then, we acquired k-PEEM spectra of the valence band and fitted the Fermi edge. The plot on left hand side demonstrates the performance of NanoESCA at photon energy of 200 eV with the VLS400 grating. The observed broadening reflcts the different contributions of the beamline (estimated to be less than 33 meV), ESCA analyzer (about 48 meV) and thermal broadening of Fermi (equal to 4kBT, i.e. 46 meV at 135 K). The calculated curves and experimental data for the resolving power of the three gratings are shown in the graph on the right hand side. The best resolving power  is achieved by the VLS 400 grating at energies in the range 100-200 eV. At lower energies, the spherical grating SG10 provides higher flux, but lower energy resolution.

In situ alignment of the gratings: remote-controlled fine adjustment of the roll

Whereas the roll of the monochromator mirror was pre-aligned in factory, both VLS 200 and VLS 400 grating holders permit in situ fine alignment of the roll. Precisely correcting the roll is essential to prevent the photon beam to move horizontally at the microscope focus, which obviously causes undesired variations in the illumination intensity. The piezo actuators are powered by voltage amplifiers that are in turn driven by a voltage provided by our custom built control unit. The plots on the left and right hand side quantify the displacement of the photon beam (imaged by the SPELEEM microscope) as a function of the varying photon energy, for different voltages applied to the piezo actuators. The red curves correspond to the best alignment configuration. As can be seen, the position of the photon beam remains stable within about 1 micron when changing the photon energy throughout the available range.

Vacuum

The vacuum system was modified as to permit independent operation of the NEG pump, which can be now separated from the optics chamber by a gate vale. Thus, the NEG can be activated (or re-generated) without exposing the optics to bad vacuum. The monochromator was baked out at a temperature of 115°C for more than a week. Then, the Varian StarCell triode 500 l/s ion pump was started and operated for 1 day at a temperature of 150°C. Once the ion pump was outgassed, the bake-out was finished and the gate valve opened to the NEG pump. The pressure readily stabilized a value of 8.5·10-10 mbar, presumably due degassing of the Cu braids, braises and piezo actuators, the main residual gas in the vacuum vessel being hydrogen. Two months after bake-out the base pressure of the monochromator chamber has reached a value of 3.5·10-10 mbar, testifying that all installed components are indeed suitable for UHV operations.

Ultima modifica il Mercoledì, 17 Febbraio 2016 15:48