Commissioning of Bunch Compressor to Compress Space Charge-Dominated Electron Beams for THz Applications

Abstract

The high peak current of the electron beam was found to be the key parameter for the THz SASE FEL at the Photo Injector Test facility at DESY in Zeuthen (PITZ). A multipurpose bunch compressor was implemented at PITZ to expand the parameter space of proof-of-principle studies on the tunable high-power accelerator-based THz source for pump-probe experiments at the European XFEL. The magnetic chicane, consisting of four rectangular dipole magnets, is designed with a bending angle of 19 degrees, due to limited space in the PITZ original beamline, to compress electron bunches with a beam momentum of 15–20 MeV/c and a charge up to 2 nC. The space charge effect and coherent synchrotron radiation are expected to drastically affect the bunch compressor performance for these parameters, thereby challenging the beam transport throughout the bunch compressor. A staged commissioning strategy was developed in order to achieve optimum bunch compressor operation. The first commissioning procedure establishes electron beam transport throughout the reference path and provides minimum beam momentum dispersion after the bunch compressor. This procedure yielded correlations between dipole magnet currents. As a result, the first bunch compression experiments were performed.

1. Introduction

An accelerator-based THz source prototype for pump-probe experiments at the European XFEL was recently constructed at the Photo Injector Test Facility at DESY in Zeuthen (PITZ), which was established for the commissioning and testing of electron sources with various diagnostic systems from measurement of beam charge, beam momentum to transverse and longitudinal phase space characterization [1]. At PITZ, a space charge-dominated electron beam from the radio-frequency (RF) gun is accelerated to a beam momentum of 15–20 MeV/c by the Cut Disk Structure (CDS) booster cavity. In 2022, new elements such as an undulator, THz test stations, and a bunch compressor were installed at the Photo Injector Test Facility at DESY in Zeuthen (PITZ) beamline as parts of the THz source prototype.

To achieve a THz Self-Amplified Spontaneous Emission (SASE) Free Electron Laser (FEL) with pulse energy in the mJ-range, electron beams of 1.5–2.5 nC compressed by the bunch compressor to a peak bunch current near 200 A are considered. Figure 1 shows a brief PITZ layout for the THz SASE FEL experiments, including the RF gun, the booster, the bunch compressor, the coherent transition radiation (CTR) measurement station, and the undulator. Furthermore, the bunch compressor has been foreseen and assigned as a part of the prototype for various applications such as the SASE FEL, seeded FEL, superradiant radiation, etc. [2].

The bunch compressor using a magnetic chicane consists of four rectangular dipole magnets (with the pole gap of 8 cm and the pole size of 15.5 × 30 cm) obsoleted from the HERA beamline [3]. The magnets are reused, as the pole shoes are added with four shims optimized by the program CST Studio Suite 2018 [4] to flatten the magnetic field profile of the magnet. Thus, the electron beam experiences a magnetic field that is <0.1% different from the maximum field throughout the chicane dipole magnet within the constant field region (5 × 7 × 20 cm in the horizontal, vertical, and longitudinal axis, respectively). Moreover, these chicane dipole magnets have identical strength and length and a vertical bending plane with a deflection angle of 19 degrees. Given the angle, the beam remains within this “good field” region during transport through the magnet. In addition, the angle is designed due to the limited installation space along the original beamline. As the effective length of each chicane dipole magnet of 0.327 m is estimated from the magnetic field simulation via the program CST, $R_{56}$ of this chicane is 0.215 m. However, there is no tuning knob for varying the chicane height (or $R_{56}$), as $R_{56}$ is fixed to one value. Note that in general, the vertical chicane structure has few benefits. First, the emittance growth from collective effects in the vertical axis could be beneficial to an undulator with a large vertical pipe size (not in our case). Secondly, there is no interference from the earth magnetic field in bending (vertical).

At PITZ, the space charge-dominated electron beam is transported throughout the beamline via the use of periodic and dense focusing elements. However, in addition to the space charge-dominated beam dynamics, this type of bunch compressor design becomes a challenge for beam dynamics and beam transport due to the impact of high bunch charge, high bending angle, and low beam energy on coherent synchrotron radiation (CSR) effects, resulting in energy spread and projected emittance growth [5,6]. Furthermore, such collective effects—both space charge and CSR effects—introduce dispersion leakage due to the imbalance in the longitudinal dynamics of particles [7,8]. According to our previous study via the particle tracking simulation [2], the compression performance (compression ratio of bunch length) is dropped, and projected vertical emittance is increased drastically when increasing bunch charge. While the beam is fully compressed and the beam current is increased, space charge and CSR effects limit the bunch compressor performance and distort the current profile. To satisfy our goal for THz SASE FELs in the simulation, for instance, a 2-nC beam current profile with booster phase of −39 degrees can be compressed (in under-compression mode) into an average bunch current of ∼200 A with normalized vertical emittance under 20 μm.

A commissioning method was developed to achieve the chicane dipole magnet currents to guide the electron beam throughout the reference path. Moreover, the two commissioning objectives are to ensure proper electron beam transport through the vacuum chamber of the chicane and to acquire the vertical bending angle closest to 19 degrees by each chicane dipole magnet, thereby achieving minimum energy (or momentum) dispersion. Figure 2 shows a concept of the first commissioning method based on the dispersion measurements at the screen downstream of the chicane. The method uses two cerium-doped ytterbium aluminium garnet (Ce:YAG) screens called HIGH2.SCR2 and HIGH2.SCR3, located upstream and downstream of the chicane, respectively. The chicane dipole magnets are named CHICANE.D1, CHICANE.D2, CHICANE.D3, and CHICANE.D4, respectively.

2. Commissioning Goal

The bunch compressor will be used to maximize the FEL pulse energy with the booster phase as a beam-energy-chirp tuning knob. However, the bunch compressor was installed downstream of a longitudinal bunch profile measurement station, and the bunch profile after compression cannot be directly measured. In order to predict the bunch compression outcome via the start-to-end simulation results, both beam momentum chirps dp/dt upstream and coherent transition radiation (CTR) downstream of the chicane are measured to benchmark the simulation results in [2].

Firstly, the generation of the beam momentum chirps upstream to the chicane via tuning booster phases was verified to benchmark ASTRA [9] tracking simulations with space charge effects for bunch charges of 10 pC, 30 pC, and 2 nC with the booster phase between 0 and −25 degrees [2,10]. The momentum chirps are extracted from the longitudinal phase space (LPS) measurement at PITZ with the setup consisting of a transverse deflecting system (TDS) and a dipole spectrometer (HEDA2), respectively [11]. Note that in the case of 2 nC, the 50-μm-slit at a distance of 5.3 m upstream of the TDS is used to reduce horizontal emittance influence in order to enhance the LPS resolution. Thus, the beamlet of 2 nC beam momentum chirps is benchmarked ASTRA simulations.

Secondly, the CTR is measured at the measurement station between the chicane and the LCLS-I undulator. The CTR measurement is expected to indicate the fully compressed-bunch booster phases for the beam with the bunch charge up to a few hundred pC, which are simulated [2] by the combination of the simulation programs ASTRA, IMAPCT-t [12], and OCELOT [13,14]. In these simulations, program ASTRA with space charge effects tracks the beam from the RF gun to the chicane entrance, and then program IMPACT-T [15,16] with both space charge and CSR effects tracks the beam throughout the chicane. Ultimately, program OCELOT with only CSR effects also tracks the beam throughout the chicane to benchmark with program IMPACT-T.

The CTR signals from compressed bunches can be measured after the chicane dipole magnet currents are obtained by the commissioning method. This benchmarking also represents that the commissioning method results in bunch compressor properties such as $R_{56}$, as expected from the simulations.

3. Background

In theory, an equal beam bending angle of all dipole magnets throughout the chicane results in zero beam momentum dispersion $R_{36}$. However, the challenge arises regarding how to achieve zero dispersion in practice, due to imperfection in the chicane installation, the inexactly identical magnetic field of each chicane dipole magnet, magnetic fields from other beamline components, etc. Additionally, at PITZ, the chicane was installed with a mechanical constraint for positioning the four dipole magnets, resulting in offsets in the magnet positions. In terms of the mechanical constraint, the second and the third chicane dipole magnet are installed above the original beamline components—an energy measurement station including a horizontal dipole magnet, a vertical beam dump, etc.—with clearance of 3–4 mm. Such clearance is very small due to the attempt to minimize the bending angle for the CSR effects. A vertical offset of any chicane dipole magnets results in the fringe field difference depending on the vertical position at the dipole magnet entrance due to the finite width of the magnets. As a result, the offset could introduce energy dispersion. For example, according to ASTRA tracking simulation results, the dispersion is increased by 0.07 mm/MeV/c, once both the first and fourth chicane dipole magnet vertically are moved by −1 mm from the designed positions.

Table 1. Energy dispersion for different cases with bending angle offsets (errors) $\{\Delta D_1,\Delta D_2,\Delta D_3,\Delta D_4\}$ (unit: deg.) of chicane dipole magnet CHICANE.D1, CHICANE.D2, CHICANE.D3, and CHICANE.D4, respectively.

$\mathsf{\Delta }{\mathit{D}}_1$ (deg.)	$\mathsf{\Delta }{\mathit{D}}_2$ (deg.)	$\mathsf{\Delta }{\mathit{D}}_3$ (deg.)	$\mathsf{\Delta }{\mathit{D}}_4$ (deg.)	Energy Dispersion (mm)
0	0	0	0	0
0.5	0	0	0	40.0
0	0.5	0	0	−49.3
0.5	0	0.5	0	23.1
−0.5	0	−0.5	0	−22.9
0	0.5	0	0.5	−23.1

represents the energy dispersion for different bending angles offset to 19 deg, where $\Delta D_1,\Delta D_2,\Delta D_3,$ and $\Delta D_4$ denote bending angle offsets (errors) of the chicane dipole magnets CHICANE.D1, CHICANE.D2, CHICANE.D3, and CHICANE.D4, respectively. Therefore, in this study, the magnitude of the energy dispersion is ultimately minimized via direct measurements.

According to Table 1, the magnitude of the energy dispersion in the cases of $\Delta D_1=\Delta D_3$ and $\Delta D_2=\Delta D_4$ is smaller than the case of the single bending angle offset. This suggests the possibility of performing the commissioning by tuning the chicane dipole magnet currents with some constraints.

4. Dipole Magnet Current Constraints

To simplify our first commissioning with an assumption, the currents ($D_1,D_3$) of the chicane dipole magnets CHICANE.D1 and CHICANE.D3 are set to equal magnitude, and the same goes to the currents ($D_2,D_4$) of the chicane dipole magnets CHICANE.D2 and CHICANE.D4, which are written as

$$\begin{array}{c}\hfill D_1=-D_3,\mathrm{and}{D}_4=-D_2.\end{array}$$

(1)

The assumption is that the electron beam experiences a fringe field of CHICANE.D1 and CHICANE.D3 in the similar beam path as the beam straight enters them, and the same goes for CHICANE.D2 and CHICANE.D4 as the beam enters them with the same opening angle of 19 degrees. Figure 3 shows the evolution of absolute value of offset momentum in the y-axis of a reference particle throughout each chicane dipole magnet, which is simulated by program IMPACT-T. The simulation is implemented by the 1D magnetic field distribution (Figure 4) along the beamline longitudinal axis [2], simulated by CST Studio Suite [4], as the particle enters and leaves at a beamline longitudinal distance of $\mp 0.645$ m offset from the center of each chicane dipole magnet, respectively. Note that the constant field region (longitudinal length) of each chicane dipole magnet of approximately $0.2$ m is implemented in the program IMPACT-T. In order to achieve the bending angle of 19 deg at all chicane dipole magnets, current $D_1=-D_3$ is set to 1.155 A, while $D_4=-D_2$ is set to 1.220 A for the beam momentum of 17 MeV/c.

Furthermore, beam transport throughout the chicane can be manipulated such that the beam follows the beam reference path by a simulation of beam tracking in program ASTRA, which is implemented by the 2D magnetic field distribution. Note that the simulation is performed without space charge effects. For instance, the chicane dipole magnet currents with the constraints in Equation (1) can be found to achieve both minimum momentum dispersion and y-position $\Delta y=0$ of 17-MeV/c beam at the screen HIGH2.SCR3 (Figure 2) located downstream of the chicane in the simulation results shown in Figure 5. Figure 5 displays both the contour line of the y-position (or offset position $\Delta y$) and density plot of momentum dispersion as a function of its chicane dipole magnet currents with the constraints in Equation (1), which determine at least one set of the chicane dipole magnet currents to achieve both the zero offset position and zero dispersion. In other words, the constraints in Equation (1) can be applied to the commissioning.

5. Chicane Commissioning

The commissioning method is divided into two steps. The first part is beam transport preparation without the use of the chicane as all chicane dipole magnets are degaussed, and the second part is the momentum dispersion measurement for different dipole magnet current settings. In the second part, the measured results are analyzed to obtain the dipole magnet current settings corresponding to the zero dispersion. Note that the bunch charge is set to 50 pC for simplicity in beam transport during the commissioning phase. The final set of chicane dipole currents as the result of the commissioning will be used as a startup parameter and fine tuned in the THz experiment with higher bunch charges.

5.1. Procedures

In the beam transport preparation, the goal is to achieve the beam that enters the chicane via the reference path. First, the beam straight must propagate through the center of the first chicane dipole magnet. This can be performed by centering the beam to the beam pipe at both screens HIGH2.SCR2 and HIGH2.SCR3 located upstream and downstream of the chicane (Figure 2), respectively. Note that the vertical beam transport and dispersion measurement have no interference from the earth magnetic field; thus, the beam is sent from the booster without the use of y-steerers. Moreover, with the bunch charge of 50 pC, the quadrupole magnets are not used in the commissioning phase. This will ensure the beam propagating straight through the center of the first chicane dipole without any steering force exerted by quadrupole magnets, when practically the beam travels inexactly through the center of all operating quadrupole magnets. In other words, the use of such a low bunch charge simplifies the beam transport and thereby the commissioning.

Then, the momentum dispersion measurement starts by turning on all chicane dipole magnets. To control magnetic hysteresis properly, all dipole magnet currents are set to the magnitude of 3.5 A prior to setting (decreasing the magnitude) to any sets of the testing currents. The set with the constraints in Equation (1) is tuned such that the beam reappears at the center of the beam pipe at screen HIGH2.SCR3 ($\Delta y=0$) again. Then, all chicane dipole magnet currents $D_1$, $D_2$, $D_3$, and $D_4$ are recorded together as a sample set of dipole currents. Note that such a tuning condition with the constraints leads to only one degree of freedom where the dipole magnet current $D_1$ becomes only one tuning knob to minimize dispersion.

For the measurement of the momentum dispersion, beam vertical position $\langle y\rangle$ is recorded at HIGH2.SCR3, where booster phase ${\varphi }_2$ in the range of (0, 10 deg) with a step of 1 deg with respect to the maximum mean momentum gain phase of the CDS booster (MMMG phase) is set to vary the beam energy or momentum. Then, this is repeated without the use of the chicane, as all chicane dipole magnets are degaussed in order to record the $\langle y\rangle$ background of this measurement. Figure 6 shows a plot of the beam vertical position as a function of beam momentum for one sample set of the dipole magnet currents for a bunch charge of 50 pC. Therefore, the momentum dispersion $\mathrm{d}y/\mathrm{d}p$ for the sample set of the dipole magnet currents is obtained by a linear regression fitting with the subtraction of the background. Note that the background is a result of imperfection in the booster beam-based alignment (BBA) and the use of steerer magnets, while it shows that the beamline without the chicane is also dispersive.

5.2. Data Analysis

The sample sets of the dipole magnet currents are set up for momentum dispersion measurement. Thus, the momentum dispersions are obtained from different sample sets of the dipole magnet currents for data analysis. Figure 7 shows a plot of the momentum dispersion as a function of sample sets of the dipole magnet currents for a bunch charge of 50 pC. Note that the sample currents $D_1=-D_3$ and $D_4=-D_2$ are plotted side by side as they correspond to the same momentum dispersion. The relationship between the dispersion and dipole currents suggests the dipole magnet current setting with zero dispersion by the use of linear regression fittings. Thus, the dipole magnet currents $D_1=-D_3=$ 1.17 A and $D_4=-D_2=$ 1.20 A are obtained for 17.05 MeV/c beam to achieve zero dispersion.

This method is repeated for beam momentum of 15, 16, 17, 18, 19, and 20 MeV/c in order to determine the dipole magnet currents to achieve zero dispersion. Figure 8 shows a plot of the dipole magnet currents (with the zero dispersion) as a function of beam momentum with linear regression fitting results for a bunch charge of 50 pC. Therefore, the fitting results or interpolations will be applied for other beam momenta in any bunch compressor application at PITZ.

5.3. CTR Measurement

The goal of the CTR measurement is to find the booster phase corresponding to the maximum bunch compression, which provides the maximum CTR pulse energy. The fully compressed-bunch booster phases are expected to benchmark the simulations for the bunch compressor performance. In the space-charge tracking simulation with the CSR effects, the beam emitted by a Gaussian laser with a root-mean-squared (rms) pulse length of 3.4 ps is fully compressed from the rms bunch length of ∼4 to <0.73 ps or 0.22 mm for a beam momentum of 17 MeV/c and bunch charge below 200 pC. Then, the fully compressed beam is sent to hit an aluminum plate and radiates the CTR, which is detected by a THz pyroelectric detector. Note that the detector can detect signals with a wavelength between 200 and 500 μm while supporting the maximum repetition rate of 1000 pps and the minimum detectable energies in the order of 50 nJ [17].

Throughout the CTR measurement, beam momentum is always set to 17 MeV for every momentum chirp, while the corresponding chicane’s dipole magnet currents from the commissioning result are also fixed. Thus, the beam is first set up for the momentum of 17 MeV/c at the MMMG booster phase. Then, the booster phase offset is set in order to generate momentum chirp, where the beam momentum is reduced. Finally, the booster RF gradient is increased to achieve the momentum of 17 MeV/c again. Note that this momentum-chirp-generation procedure is similar to the simulation via program ASTRA in [2] and the momentum chirp measurement in [10].

After the preparation of the beam momentum and momentum chirp, the beam is transported downstream to the chicane and transversely focused to a CTR measurement station consisting of the aluminum plate target and the pyroelectric detector. Figure 9 displays a plot of CTR pulse energy measured via the pyroelectric detector as a function of the booster phase for bunch charges of 30, 50, 100, and 200 pC. It also shows that the maximum CTR pulse energy from different bunch charges corresponds to different booster phases. Note that the bunch charge higher than 200 pC could be challenging for the CTR measurement of the fully compressed bunch due to the longitudinal space charge effects. In this case, the fully compressed bunch is decompressed as it propagates to the CTR measurement station. Therefore, the CTR measurement of the higher bunch charge should not be conducted for the fully compressed-bunch booster phase benchmarking.

Finally, the booster phase which provides the maximum CTR pulse energy in Figure 10 is benchmarked with the simulation in [2]. Figure 10 also shows simulation results including CSR effects from different programs, which are IMPACT-T and OCELOT. Moreover, the experimental results show a similar trend to the simulation results. This suggests that the bunch compressor properties from the commissioning method are similar to the simulations.

6. Summary

A commissioning procedure based on the dispersion measurements at the screen downstream from the chicane has been developed. In order to achieve optimum bunch compressor operation, the first PITZ bunch compressor commissioning was performed to obtain the set of dipole magnet currents that provides zero momentum dispersion for a bunch charge of 50 pC. The procedure also yielded correlations between dipole magnet currents. By applying these chicane dipole current settings, the first bunch compression experiments were performed, where pyroelectric detector signals from the CTR station downstream from the bunch compressor were used to find a maximum compression phase of the booster cavity.

In other words, we demonstrated zero-dispersion transport of the low-energy electron beam throughout the chicane with the high bending angle, where space charge and CSR effects drastically affect bunch compressor performance. With limited beam diagnostic stations downstream from the bunch compressor, we finally obtained the experimental results to verify the simulation study.

7. Future Plans

Bunch compressor commissioning will be performed to achieve the set of dipole magnet currents that provides zero momentum dispersion for a bunch charge higher than 50 pC to reach THz application requirements. For example, the superradiant radiation experiment at PITZ requires a bunch charge around 250 pC.

The momentum dispersion derivative ($R_{46}$) component is planned to be measured once a new beam position monitor near HIGH2.SCR3 is installed and ready to use. However, during this stage of this commissioning, there was no other position monitor nearby HIGH2.SCR3 to conveniently use for the dispersion derivative measurement. Moreover, the CTR station screen approximately one meter far from HIGH2.SCR3 is used to observe the position and beam focusing prior to the CTR measurement. When introducing the momentum chirp by changing the booster phase (under compression mode), the beam remains focused at the same position. This could indicate a negligible introduction of the momentum dispersion and the dispersion derivative.