Experimental Demonstration of Double Emittance Exchange toward Arbitrary Longitudinal Beam Phase-Space Manipulations

Many of the most significant advances in accelerator science have been due to improvements in our ability to manipulate beam phase space. Despite steady progress in beam phase-space manipulation over the last several decades, future accelerator applications continue to outpace the ability to manipulate the phase space. This situation is especially pronounced for longitudinal beam phase-space manipulation, and is now getting increased attention. Herein, we report the first experimental demonstration of the double emittance exchange concept, which allows for the control of the longitudinal phase space using relatively simple transverse manipulation techniques. The double emittance exchange beamline enables extensive longitudinal manipulation, including tunable bunch compression, time-energy correlation control, and nonlinearity correction, in a remarkably flexible manner. The demonstration of this new method opens the door for arbitrary longitudinal beam manipulations capable of responding to the ever increasing demands of future accelerator applications.

Figure 1

Experimental beamline configuration. The beamline is composed of a photoinjector, matching quadrupoles (Quad set 1), a DEEX beamline, and a LPS diagnostic section. The photoinjector includes an rf photocathode gun and four accelerating cavities that accelerate the electron beam energy up to 44.5 MeV. The DEEX beamline consists of two EEX beamlines with a middle section for transverse manipulations between them. The LPS diagnostic section follows the second EEX beamline, and it is equipped with four quadrupole magnets (Quad set 3) for transverse focusing, a horizontal slit with a $100\text{−}\mathrm{\mu }\mathrm{m}$ opening, a TDC, and a spectrometer. Green dots indicate reference positions 1, 2, 3, and 4 on the experimental beamline.

Figure 2

Tunable bunch compression by DEEX beamline. The last quadrupole magnet in the middle section was scanned during the experiment. The red dots (with associated error bars) represent the measured final bunch lengths (${\sigma }_{z,4}$) for each value of the quadrupole gradient ($g_5$). The black curve indicates the final bunch length, which is estimated using Eq. (1). Because the quadrupole scan results are used to estimate ${ϵ}_{x,3}$ in Eq. (1), the black curve includes a finite error range, which is displayed as the blue shaded area. The red dashed line indicates the bunch length (${\sigma }_{z,1}$) at the entrance of the DEEX beamline (position 1).

Figure 3

Longitudinal chirp control by DEEX beamline. (a)–(e) show the measured longitudinal phase spaces with various combinations of the last two quadrupole magnets in the middle section. (f) shows ${\sigma }_{z,4}$ and the final longitudinal chirps corresponding to each case.

Figure 4

Third-order correction using an octupole magnet. (a)–(e) show measured longitudinal phase spaces with different octupole magnet strengths. The strengths of the octupole magnet were $(182,-182,-364,-546)\mathrm{T}/{\mathrm{m}}^3$. Each value corresponds to (a)–(e), respectively. (f) shows the third-order coefficient ($a_3$) of the polynomial fitting and the final bunch length (${\sigma }_{z,4}$) for each of the displayed cases.