The last two years saw major efforts to reduce the vertical emittance. Benefits are manifold: higher brilliance, lower radiation dose during injections and, possibly, higher photon flux towards high energies by further reducing the vertical aperture of in-vacuum undulators.

In an ideal machine like the ESRF storage ring, the vertical emittance should be below the pico-metre (pm) level. However, in reality, if no correction is carried out it may reach a few hundreds of pm. There are two main sources of vertical emittance: vertical dispersion and betatron coupling. The dispersion represents the dependency of the electron off-axis orbit upon its energy. Turn by turn, electrons lose energy as synchrotron radiation and gain some energy back thanks to the RF cavities. Non-zero vertical dispersion results then in electrons subjected to vertical displacements, as in the horizontal plane. When integrating such displacements over all electrons inside bunches, the result is a nonzero vertical beam size and emittance. Tilts of quadrupole magnets and vertical misalignment of sextupoles are instead responsible for betatron coupling: The natural horizontal oscillations of the electrons within a bunch, induced by the emission of synchrotron radiation, are partially transferred to the vertical plane. As for dispersion, when integrating the vertical oscillations of all electrons, a thick vertical beam size and nonzero divergence are generated. An important difference between the two sources is that the contribution to the vertical emittance from vertical dispersion is constant along the storage ring, whereas the fraction of emittance generated by coupling varies along the machine. Values ranging from 100 pm to 450 pm have been measured at twelve emittance monitors when no correction is applied, see Figure 156.

Fig. 156: Vertical emittance measured at twelve monitors along the ring with all correctors turned off (January 16, 2010).

The minimisation of the vertical emittance is carried out in two steps. The vertical dispersion and betatron coupling need to be precisely measured and then they have to be efficiently corrected by means of the skew quadrupole correctors installed around the storage ring. Two main factors in the two year period 2009/2010 made an unprecedented coupling measurement and correction possible. Installation of the most advanced beam position monitoring system, LIBERA, greatly increased the resolution in the measurement of the beam orbit response matrix, from which coupling is inferred. A new correction algorithm developed at the ESRF resulted in a faster and more effective simultaneous reduction of betatron coupling and vertical dispersion. Already in June 2010 a record-low vertical emittance of 5 pm was achieved during machine dedicated time.

However, at that time, it was difficult to preserve such a low value during beam delivery, because of the continuous changes in the apertures of insertions devices (IDs) carried out by users: despite the high magnetic quality of the great majority of IDs, a few of the earliest ones are known to be a gap-dependent source of coupling. To cope with this issue, two countermeasures were undertaken. The coupling induced by two of these IDs was measured and corrected locally by dedicated correctors placed at the ends of their straight sections. A feed-forward loop powers these magnets and it depends on the ID gap aperture, according to a look-up table created during the measurement. On top of this, a unique coupling feedback was also implemented to periodically trim the currents of the 32 skew quadrupole correctors distributed around the storage ring so to retrieve the lowest emittance possible. At the end of 2010, the ESRF storage ring running in the 7/8+1 mode was routinely operating with a vertical emittance at the level of 7 pm, down from the typical 20-30 pm of 2009.

After the successful 2010 campaign, the late Pascal Elleaume wondered whether an ultra-low vertical emittance of 2 pm could be achieved by increasing the number of skew quadrupole correctors. Besides the dedicated IDs correctors, at that time 32 skew quadrupoles were used. Even though 52 trim coils installed in the storage ring were still available, studies demonstrated that 2 pm could be achieved by adding just 32 new skew correctors, the benefit of going beyond this number being minimal. During the winter shutdown of 2010 those coils were powered up and connected to the control system. After a few months of fine-tuning the correction with the new magnets, a vertical emittance between 3 and 4 pm could be routinely achieved, see Figure 157. While the larger number of correctors further reduced betatron coupling efficiently, the vertical dispersion remained almost unchanged and is now believed to be the main source of vertical emittance and the missing stage on the way towards the target of 2 pm. Studies are on-going to better understand and to overcome this issue.

Fig. 157: Vertical emittance measured at twelve monitors along the ring after coupling correction (June 1, 2011)

During the initial period of low vertical emittance, the beam lifetime after refill deteriorated from 45 hours to 30 (in the 7/8 +1 mode), hence casting some concerns on the continuous use of this special mode. Further corrections of the sextupolar resonances, however, allowed the traditional lifetime of 45 hours at 200 mA to be recovered. Since spring 2011, therefore, the 7/8+1 mode has been delivered routinely at the lowest achievable emittance. Vertical emittance and beam lifetime during a typical week of delivery are reported in Figure 158.

Fig. 158: Beam lifetime (top) and average vertical emittance (bottom) measured during a week of beam delivery (7/8+1 filling mode) in June 2011. Data acquired during refills are not displayed.




J. Chavanne et al. Vertical emittance reduction and preservation at the ESRF electron storage ring, in Proceedings of the IPAC2011 Conference, San Sebastián, Spain (2011).