Introduction & History

Elettra is the third generation storage ring (2 and 2.4 GeV ) that has been in operation since October 1993. It has been optimised to provide the scientific community with photons in the energy range from a few to several tens of KeV with spectral brightness of up to 10¹⁹ photons/s/mm²/mrad²/0.1%bw and is continuously upgraded in order to be competitive with the most recent sources.

There were three phases and two main upgrades in the operations of Elettra. In the first phase from 1993 to 2007 Elettra operated in the ramping mode. The injector was providing 1 GeV electrons to the storage ring that were further accelerated by ramping the storage ring to the final operational energy 2 or 2.4 GeV for users. Refill occurred once per day. Ramping operations, intensity in mA ( y-axis)  Since 2007 Elettra constructed a full energy injector consisting of a small LINAC (LINear ACcelerator) and a full energy booster that accelerates and injects electrons at the final energy of 2 and 2.4 GeV (the 1 GeV old injector has been partially refurbished and used in the new FERMI@Elettra FEL facility, the new 1.2 – 1.8 GeV  Free Electron Laser in construction here).   Full energy injection operations, intensity in mA (y-axis )      Finally since 2010 Elettra operates in the top up mode for both 2 and 2.4 GeV operational energy. Top up operations, intensity in mA (y-axis)

Additionally to Elettra, a storage ring Free Electron Laser(SR-FEL) is operating keeping Elettra at the forefront of fourth generation photon sources. The SR-FEL using SRHG method operates at 2 GeV, producing coherent fs-light pulses with variable polarization in the vacuum ultraviolet since some years. Already this beamline is used for experiments. Brightness of the synchrotron radiation for various ID's and bending magnets as a function of photon energy Synchrotron radiation is produced when electrons travelling at relativistic speeds are deflected in magnetic fields. The storage ring is made up of four types of magnets: bending magnets that deflect the circulating electron beam into a closed circular path, quadrupoles that focus the beam, sextupoles that compensate chromatic and non-linear effects and steerer magnets that perform small adjustments to the circular trajectory. The arrangement of magnets forms a lattice of magnetic confinement elements. The lattice used for ELETTRA is an expanded Chaseman Green type also known as a double bend achromat. The ring is made up of twelve identical groups of magnets forming a ring roughly 260 m in circumference. A characteristic of third generation synchrotrons is the space that is made available in the lattice, the so called long straight sections, to install insertion devices of lengths up to 4.5 m. These are the principal sources of high brightness photons and are composed of arrays of magnetic poles that force the circulating electrons along serpentine trajectories. Insertion devices are of many types (electromagnets or permanent magnets) and depending on the magnetic configuration can be made to produce linear to circular polarised light. The wavelength of the light is tuneable by changing the magnetic field acting on the electron beam. For the electromagnets this is performed by changing the current flowing in the coils, whilst for the permanent magnets the field is changed by varying the distance between the top and bottom magnet arrays. One of these sections is used for injecting electrons into the storage ring.

Elettra configuration before 2007

The electrons were generated in a LINAC and then injected in the ring. Electrons start off from a ceramic disc that is heated to very high temperature. An electric field of up to 100 kV draws out the electrons that are then accelerated through various radio-frequency structures that make up the LINAC. The LINAC was composed of the Gun (that houses the ceramic disc), a low energy bunching section and several high-energy sections. The overall length of the LINAC was 66 m and contained seven high-energy sections each 6m long. Between accelerating sections, quadrupole magnets keep the beam focused. The LINAC operated at 3 GHz and generated a pulse of electron bunches that are accelerated to a final energy which can be as high as 1.2 GeV. In order not to lose electrons as they are accelerated, the entire LINAC is under vacuum, as is the transfer line and storage ring. In this way collisions with gas molecules, that would otherwise cause the loss of electrons, are avoided.

 

 

 

The electrons are generated in a small linac. They start off from a ceramic disc that is heated to very high temperature. An electric field of up to 80 kV draws out the electrons that are then accelerated through two radio-frequency structures that make up the linac. The linac is composed of the Gun (that houses the ceramic disc), a low energy bunching section and two high-energy sections. The overall length of the linac is 12 m and contains two high-energy sections each 5m long. Between accelerating sections quadrupole magnets keep the beam focused. The linac operates at 3 GHz and generates a pulse of electron bunches that are accelerated to the final energy of 100 MeV. In order not to lose electrons as they are accelerated the entire linac is under vacuum, as is the transfer lines the booster and storage ring. In this way collisions with gas molecules, that would otherwise cause the loss of electrons, are avoided.









General layout of the accelerator complex showing the
position of the linac, transfer line and storage ring

The cathode (part of the gun)	The linac tunnel with the accelerating sections and magnets, in the front hanging from the ceiling the SLED cavities, nowadays used for the FERMI FEL

The electrons exiting the LINAC were then transported to the inner side of the storage ring by a transfer line (a series of deflection and focussing magnets). Both the LINAC and transfer line were below ground so as not to interfere with beamlines in the experimental hall. The storage ring was filled by a multi-turn injection process whereby pulses of electrons were gradually fed into the ring ten times a second until the desired current is achieved. Refilling the ring to high currents took 45 minutes.

The old transfer line	The old transfer line

Elettra configuration since 2007

A new building accommodating a small LINAC injector and a booster has been built in the central empty space of the storage ring building. The full energy injector project started in 2005 and finished by providing beam in March 2008 on time and within budget. This new injection chain consists of a 100 MeV linear accelerator and a 2.5 GeV booster that sends the beam into the storage ring at a rate of up to 3 Hz.

The electrons are generated in a small linac. They start off from a ceramic disc that is heated to very high temperature. An electric field of up to 100 kV draws out the electrons that are then accelerated through two radio-frequency structures that make up the linac. The linac is composed of the Gun (that houses the ceramic disc), a low energy bunching section and two high-energy sections. The overall length of the linac is 12 m and contains two high-energy sections each 5m long. Between accelerating sections quadrupole magnets keep the beam focused. The linac operates at 3 GHz and generates a pulse of electron bunches that are accelerated to the final energy of 100 MeV. In order not to lose electrons as they are accelerated the entire linac is under vacuum, as is the transfer lines the booster and storage ring. In this way collisions with gas molecules, that would otherwise cause the loss of electrons, are avoided.

The small linac

The electrons exiting the linac are then transported to the booster by a transfer line (a series of deflection and focussing magnets). The booster is a simple synchrotron of 118 m of circumference that can accelerate a maximum of 6 mA current from 100 MeV up to 2.5 GeV with a repetition rate of 3 Hz. It operates always at full cycle (100 MeV to 2.5 GeV) and the electrons are extracted at the needed energy by adjusting the extraction kicker time. 

A view to a part of the booster

The booster is using the on-axis injection scheme therefore can not accumulate electrons (since it is not needed). Once the electrons arrive at the requested energy are extracted to a long transfer line that arrives to the storage ring.

Booster to Storage ring transfer line

The storage ring is filled by a multi-turn injection process whereby pulses of electrons from the booster are gradually fed into the ring three times a second until the desired current is achieved using special magnets called septa and kickers.

The injection elements fo Elettra

Usually Elettra is filled with 310 mA when at 2 GeV and 150 mA when at 2.4 GeV. More than 500 mA have been stored at 1.5 GeV and more than 700 mA at 1 GeV. The maximum intensity is limited by the radiofrequency power and the thermal load in the vacuum chamber due to synchrotron radiation.

Storage ring showing the superconducting wiggler in front and the superconducting third harmonic cavity behind.	The variable polarization undulators for the SR-FEL serving also the nano- spectroscopy beam line and further in front a short undulator for the TweenMic beam line

The electrons circulating in the ring do so in a metal vacuum chamber. The vacuum that is maintained in the ring must be of very high quality, since unlike the LINAC and transfer line where an electron passes through once, in the ring the electrons, travelling close to the speed of light, traverse a given point more than a million times in one second. To maintain a long beam lifetime we must therefore reduce the chance of electrons colliding with gas molecules. The situation is further complicated by the copious emission of synchrotron radiation - around 90 kW of power just from the bending magnets. The unused radiation must be absorbed in special places otherwise chamber deformation and photo-electron release of surface gasses will occur.



The brightness of the photon beam is derived from the small transverse size and divergence of the electron beam. A parameter that encompasses these dimensions is the emittance defined as the area occupied by the beam in phase space. To obtain a small emittance the beam is strongly focused by the ring quadrupoles. Furthermore the bending magnets have a gradient to provide additional focussing. The use of strong focusing magnets leads to increased chromatic aberrations simply because the beam contains electrons with a distribution of energies (up to a few percent of the total energy). If uncorrected these chromatic effects will limit the current that can be stored to a few mA. Sextupole magnets, placed at points in the ring where electrons with different energies travel different paths (dispersion regions), are used to compensate these unwanted effects. The story does not end here though because the chromatic sextupoles themselves introduce non-linear motion of the electrons that may lead to particle loss. This effect is in turn compensated by additional sextupoles place at points in the ring where the electrons travel the same path even if they have different energies.

 SRPM image and the dimensions of the beam

The energy lost by the electrons when emitting synchrotron radiation is compensated by radio-frequency cavities. Four single cell cavities are used and have been positioned between dipole magnets in the dispersion region thereby allowing maximum use of the long straight sections for insertion devices. The cavities, operating at 500 MHz, produce a longitudinally bunched beam, simply because only those electrons arriving at the right time will be accelerated and the rest lost. The maximum number of electron bunches, separated by 2 ns, that can fit in the ring circumference is 432. There is a great deal of flexibility in filling the ring: from one bunch to any combination. The usual mode of operation is multi-bunch where roughly 95% of the ring circumference is filled with electron bunches. The summed voltages of the cavities determine the (longitudinal) energy acceptance and electrons having energies outside the acceptance are lost. The energy acceptance is an important parameter and partially governs the lifetime of the stored beam. This arises because collisions between electrons within the highly dense bunches (a consequence of low emittance) transfer energy from the transverse plane to the longitudinal (the Touschek effect). The lifetime is therefore essentially determined by the quality of the vacuum and the quality of the emittance.

 

The beam train seen on an oscilloscope	Radio-frequency cavity installed in the storage ring

 

Users are sensitive to variations in beam parameters. Unwanted motion of the electron beam translates as an effective emittance growth or worse still as jumps in intensity and loss of brightness. The disturbances have different time scales ranging from months to milliseconds and require different techniques for their suppression or control. The slowest instabilities affect the orbit of the beam and are mainly due to changes in temperature (of buildings, electronic components, ring equipment, etc…). To control the orbit Beam Position Monitors - BPM's are used to provide information to orbit correction programs. The resolution of the BPM's has to be good (sub- micron) to enable effective control of the beam that has typical dimensions of tens of microns. The faster instabilities require feedback systems.

 

 The BPM mounted on a quadrupole

 

 Elettra is having a full set of feedback systems that together with the third harmonic syper conducting passive cavity and the fine tuning of the cavity temperatures eliminate all multibunch instabilities. A fast global orbit feedback is also used to keep the orbit stable with submicron accuracy.

The energy of the circulating electrons can be varied up to 2.4 GeV. A typical annual operating schedule allocates about 25% of beam time at this energy and the remaining time at 2.0 GeV. The storage ring operates on a twenty-four hour basis for up to 6500 hours a year (about 75% of the year). These hours are distributed into so called Runs, i.e., blocks of time that usually last seven to ten weeks. A Run is further split into periods lasting about a week for the production of light for the Users interspersed by one or two days of machine dedicated studies. Machine studies, performed by accelerator physicists and engineers, are all geared towards bettering the quality of the light and the commissioning of new systems. The Runs are separated by the Shutdown periods that usually last from one to four weeks. During Shutdowns maintenance of systems and the installation of new equipment is performed and is an essential activity in the life of the facility.

All systems are controlled from the Elettra control room, whereas the equipments are in special service areas.

 

Elettra control room	Service gallery

Finally the experiments are performed in the experimental hall.

Elettra operation configuration since 2010
In May 2010 Elettra,successfully joined the synchrotron facilities that fully operate in top-up. Elettra operates for users since 1994 and during the past few years a massive upgrade program including the construction and set in operation of a full energy injector took place. The full energy injector and other machine and beam line upgrades as well as the demand for intensity and thermal stability naturally led to top-up operations whereby the storage ring beam current is kept constant during user operations.
 
Although Elettra was not originally designed for this type of operation (and operated for many years even without a full energy injector), almost a year after establishing reliable operations of the full energy injector , successfully operates in top-up in both 2 and 2.4 GeV user energies (since May 2010). Elettra is  thus another example of how a third-generation synchrotron, that previously operated in decay mode, can successfully advance to full top-up operation and at multiple energies too. In top-up mode, the storage ring intensity is kept constant by frequent beam injections, in contrast with the decay mode where the stored beam is allowed to decay to some level before refilling occurs.    
                             
The top-up operation renders the produced photon intensity practically stable whereas the integrated intensity is 60% higher for a time period equal to the beam lifetime. Thus while keeping the optical components of the beam lines in thermal equilibrium also the integrated number of photons is higher, an additional beam time gain for the experiments. At the same time, also the intensity dependent electronics remain stable allowing submicron accuracy in the electron beam position and hence a higher stability of the photon beam.

The upgrade to top-up started in 2009 and included the addition of various diagnostic and radiation safety instruments, modification of the control and interlock software, fine tuning of the kicker and septa timing and a revised operation strategy with the laborious collaboration of the radiation protection team resulting to a high level application a “top-up controller” controlling all aspects of the procedure. The careful radiation measurements at each beam line under various conditions of the injected beam and the high injection efficiencies achieved at both energies (radiation levels in all beam linesremain below 1 µSv/hour for efficiencies higher than 90% ) resulted in no additional shielding to the beam lines.

The storage ring beam current at 2 GeV is set by the users to 310 mA and top-up occurs every 6 min by injecting 1 mA in 4 seconds, keeping thus the current level constant to 3‰. At 2.4 GeV the stored beam current is set to 140 mA, top-up occurs every 20 minutes injecting 1 mA in 4 seconds keeping the current level constant to 7‰.

The users have chosen the fixed current top-up (1 mA) instead of the fixed time interval. The injection system is perfectly tuned and for the majority of the beam lines does not produce interference with data-acquisition processes. A gating signal is also provided but up to now only few, very sensitive, beam lines see some interference and therefore are gated.

No transition period was needed or required and once the top-up started all went exceptionally smooth due to the very good preparation and the high level of expertise of the personnel involved. It is worth mentioning that although at the beginning the operation in top-up was programmed at 20% of users beam time, already from the start it became clear that the users strongly preferred this mode and thus Elettra operated in top-up at 100% of the user dedicated beam time right from the beginning i.e. May 2010.