Electron Energy Spectrometer for MIR-THz FEL Light Source at Chiang Mai University

Abstract

The linear accelerator system of the PBP-CMU Electron Linac Laboratory has been designed with the aim of generating free-electron lasers (FELs) in the mid-infrared (MIR) and terahertz (THz) regions. The quality of the radiation is strongly dependent on the properties of the electron beam. Among the important beam parameters, the electron beam energy and energy spread are particularly important. To accurately measure the electron beam energy, the first dipole magnet in the bunch compressor system and the downstream screen station are employed as an energy spectrometer. The A Space Charge Tracking Algorithm (ASTRA) software is used for the design and optimization of this system. Simulation results demonstrate that the developed spectrometer is capable of accurately measuring the energy within the 5–25 MeV range. The screen station system is designed and constructed to have the ability to capture a beam size with a resolution of 0.1 mm per pixel. This resolution is achieved with a screen-to-camera distance of 1.2 m, which proves sufficient for precise energy measurement. The systematic error in energy measurement is found to be less than 10%, with a minimum energy spread of 0.4% achievable when the horizontal beam size remains below 3 mm.

1. Introduction

The PBP-CMU Electron Linac Laboratory (PCELL) at Chiang Mai University in Thailand is currently in the process of developing a mid-infrared free-electron laser (MIR-FEL) oscillator and super-radiant terahertz free-electron laser (THz-FEL) [1]. These advanced accelerator and laser systems are being constructed and commissioned at the facility. In the electron injector system of PCELL, electrons are generated using a thermionic cathode installed in a standing wave radio-frequency (RF) electron gun. In the RF gun, the emitted electrons are accelerated using a 2856 MHz RF wave, reaching a kinetic energy of approximately 2 MeV. Subsequently, the electron beam undergoes further acceleration through a 2856 MHz travelling-wave linear accelerator (linac). This additional acceleration elevates the electron beam’s energy within the range of 10–25 MeV. After exiting the linac, the beam passes through a magnetic bunch compressor system specifically designed for the MIR-FEL and the THz-FEL beamlines. The layout of the injector system, including the first two dipole magnets and other associated components, is shown in Figure 1. The required electron beam properties for the generation of high-brightness radiation for both MIR and THz FEL are a proper electron beam energy with low energy spread, short pulse, high bunch charge, and low transverse emittance [2].

To characterize the aforementioned electron beam properties, diagnostic devices are employed. An alpha magnet equipped with energy slits and a downstream current transformer is used to measure the energy and energy spread of the electron beam produced from the RF gun [3,4]. A dipole magnet and a Faraday cup at the end of the straight section are employed together as an energy spectrometer for measuring the electron after the linac acceleration. Furthermore, the first dipole magnet in the bunch compressor, as described in [5], along with the view screen station positioned downstream of this magnet, serve as an energy spectrometer for the MIR and THz FEL beamlines. The electron beam energy spectrometer employing the dipole magnet and beam current monitor or screen is widely used in several facilities. A dipole magnet and a beam current monitor in the form of a Faraday cup are used at KU-FEL to measured the electron beam with an energy spread of below 5% when applying a slit with proper width [6]. At the Helmholtz-Zentrum Berlin for Materials and Energy Research (HZB), a dipole magnet and a screen station are employed to measure the energy and energy spectrum of the electron beam produced from a superconducting RF photoelectron gun (SRF gun) of the Berlin Energy Recovery Linac (BERLinPro). This spectrometer is expected to provide measurements with an energy resolution of about 0.1% [7]. A dipole magnet and a screen station are also utilized at the Photo Injector Test facility at DESY, Zeuthen site (PITZ) to measure the slice momentum spread of the electron beam at a high resolution, down to 1 keV/c [8].

The aforementioned examples clearly demonstrate the effectiveness of applying a dipole magnet and a beam current monitor or a screen with a well-designed setup as an electron beam energy spectrometer, allowing for precise measurements with small energy spread and high energy resolution. In our facility, we employ the first dipole magnet in the magnetic bunch compressor and its downstream screen station for the sake of convenience and cost efficiency. In this paper, we thus focus on the physical design and development of an energy spectrometer, utilizing a dipole magnet and a screen station for measuring electron beam energy and energy spread.

2. Methodology

The electron energy spectrometer is a device used to measure and analyze the energy distribution of an electron beam. In our design, we employ a combination of a dipole magnet and a screen station. The dipole magnet plays a crucial role in bending the path of the electron beam. As the beam passes through the dipole magnetic field, it experiences deflection due to Lorentz’s force. The deflection angle depends on the energy and momentum of the electrons. By controlling the magnetic field strength and the length of the magnet, the degree of deflection can be calibrated to correspond to specific energy values. After passing through the dipole magnetic field, the deflected electron beam reaches the screen station. The screen station consists of a phosphor screen and a CCD camera that is capable of detecting and recording the positions of the incident electrons. The illuminated image on the screen forms a pattern that represents the distribution of the electron beam. By analyzing the positions of the electrons on the screen, we can determine the energy distribution of the beam. The deflection of the electrons in the dipole magnetic field allows for a separation of the electrons based on their energies. This information is then used to construct an energy spectrum, providing valuable insights on the electron beam mean energy and energy spread.

At our facility, the first dipole magnet in the magnetic bunch compression system and the downstream screen station are employed as the energy spectrometer for the MIR and THZ FEL beamlines. We conducted dynamic simulation of the beam using ASTRA software [9] for investigating the properties of the electron beam when it passes through the dipole magnet and reaches the screen station. The three-dimensional (3D) magnetic field distribution of the dipole magnet obtained from simulation using CST EM Studio software [10] is used in our beam dynamic simulation. The design of this dipole magnet enables it to effectively deflect a 30 MeV electron beam at a 45° angle [5]. The dipole magnetic field is linearly increased with the applied electric current when the current is less than 20 A. At the applied current of 20 A, the maximum vertical magnetic field ($B_y$) is 149.74 mT with an effective length of 349.78 mm. In this study, the electron energy in the range of 5 to 25 MeV was chosen based on the energy of the electron beam that can generate THz transition radiation and MIR/THz FEL [5,11,12,13,14].

2.1. Computation of Electron Kinetic Energy

Since our energy spectrometer consists of the dipole magnet and the screen station, the key parameters to achieve the electron energy are the magnetic field and the effective length of the dipole magnet, the angle shift (${\theta }_i$) and the shifted distance on the screen (D). The setup layout of the designed energy spectrometer is illustrated in Figure 2. The kinetic energy ($E_k$) of an electron traveling through this setup can be calculated from (1) [15].

$$E_k\left[\mathrm{MeV}\right]={\displaystyle \frac{299.8l_{\mathrm{eff}}{B}_0}{\beta \left(0.7854+arctan\left[{\displaystyle \frac{D}{0.945}}\right]\right)}}-0.511,$$

(1)

where $\beta$ is the relative velocity of electrons, $l_{\mathrm{eff}}=(\int B_y\left(z\right)dz)/B_0$ is the effective length of the dipole magnet and $B_0$ is the dipole peak magnetic field. In the simulation, we first investigate the trajectory of a single electron, varying in energy in the range of 5–25 MeV. Electrons are injected into the center of the dipole magnet parallel to the beam axis. All simulation results show electrons bent at an angle of 44.36${}^{\circ }$ due to the effect of the fringe field of the dipole magnet. Therefore, the effective length of the dipole magnet is reduced by 1.61% from the theoretical value and becomes 344.3 mm. This effective length was included in the kinetic energy calculation.

2.2. Optical Resolution Investigation

The screen station used in this energy spectrometer system consists of a phosphor screen with a horizontal length of 33 mm and a CCD camera connected to the computer, as depicted in Figure 3. When the electron beam hits the phosphor screen, the screen is illuminated, reflecting the image of the beam. The transverse beam profile on the screen affects the resolution and accuracy of electron energy measurement. The phosphor screen is placed with an angle of 45${}^{\circ }$ with respect to the electron beam trajectory. Consequently, the screen can capture a beam image with a size of only 23.33 mm. The beam image is captured by the BASLER acA640-120gm CCD camera with a number of pixels in the horizontal and vertical direction of 659 and 494 pixels, respectively. The camera is equipped with an M7528-MP lens, with a maximum magnification of 4.6:1 at the minimum focal point. The distance between the screen and the CCD camera is 1.2 m. Therefore, the resolution of this camera system is 0.1 mm/pixel. We included this resolution value in the computation of the kinetic energy and energy spread values.

2.3. Electron Beam Dynamic Simulation

The electron beam dynamic simulations were conducted for two scenarios: one utilizing a Gaussian beam distribution generated by the Generator program included in the ASTRA software package, and the other based on the electron distribution obtained from the start-to-end beam dynamic simulation from the RF gun to the entrance of the dipole magnet. In both simulation scenarios, the properties of the electron beam were considered by taking into account the influence of accuracy and systematic errors in energy measurement.

In the first scenario, the electron bunches are generated with a Gaussian distribution for both the transverse and longitudinal planes. The average kinetic energy and energy spread were systematically varied within the range of 5 to 25 MeV and 0.25 to 2.0%, respectively. Additionally, the transverse beam size was adjusted from 1 to 6 mm. It is worth noting that the bunch charge remains fixed at 100 pC, while the emittance is set to zero. In the second scenario, the start-to-end beam dynamic simulations were conducted to explore the energy measurement system under more practical conditions for the electron beam. The entire injector system’s magnets were optimized to attain suitable electron beam properties for generating THz transition radiation at its station downstream of the linac section. To achieve a short electron bunch at the transition radiation station, the electrons with energy lower than 2.0 MeV were filtered out utilizing energy slits in the alpha magnet vacuum chamber. The alpha magnet gradient was set at 3.85 T/m for all simulations. The accelerating gradient of the linac was adjusted to obtain an electron beam with a kinetic energy of 20 MeV, while the linac phase was set to the on-crest acceleration condition. These adjustments ensured the desired energy parameters for the electron beam.

The electron beam distributions obtained from both scenarios served as the initial beam distributions for subsequent simulations in the energy spectrometer system. This approach allowed for the construction of an energy spectrum and facilitated an investigation into the systematic error associated with energy measurement. The analysis involved considering various factors that could contribute to inaccuracies, especially the electron beam transverse size and emittance. The investigation performed for this energy spectrometer system provided an understanding of the systematic error associated with energy measurement, allowing for the refinement and improvement of the measurement techniques.

3. Results and Discussion

3.1. Influence of Electron Beam Properties on the Accuracy of Energy and Energy Measurement

The investigation into the influence of electron beam properties on the accuracy of energy measurement was conducted using the initial beam distribution from the first scenario, as described in Section 2.3. The results depicting the systematic error of energy measurement for electron bunches with average energies ranging from 5 to 25 MeV are presented in Figure 4 and Figure 5. Figure 4 demonstrates that the systematic error of average kinetic energy remains below 0.1% for all energy spread values and transverse beam sizes. Meanwhile, Figure 5 shows the systematic error in energy spread measurement. It reveals that electron bunches with larger energy spreads can be measured with lower systematic error compared to those with lower energy spreads, given similar transverse beam sizes. However, for electron bunches with low energy spread, the impact of a larger transverse beam size becomes more pronounced, introducing additional systematic error in the energy spread measurement. The systematic error value converges around 10% as the energy spread increases due to the contribution of the dipole field. In order to maintain a systematic error less than 10%, the electron beam size in the horizontal axis should not exceed 3 mm.

The investigation of the influence of transverse beam emittance on the accuracy of energy measurement was conducted. A Gaussian distribution electron bunch was generated with an rms beam size of 3 mm, an average energy ranging from 5 to 25 MeV, an energy spread of 0.5%, a bunch charge of 100 pC, and emittance values ranging from 0 to 25 mm·mrad. Simulation results for an electron bunch with an average energy of 5 MeV are depicted in Figure 6. The findings reveal that the systematic error in measuring the average kinetic energy remains below 0.1% for all emittance values. However, the systematic error in measuring the energy spread increases as the emittance value increases. When considering a systematic error threshold less than 10%, it is advisable to limit the emittance value to below 12 mm·mrad. For an even more stringent requirement of a systematic error below 1%, the emittance should be maintained below 4 mm·mrad.

The findings in this section emphasize the importance of considering the influence of beam properties, including beam size and emittance, on the accuracy of energy and energy spread measurement. By taking these factors into account, more reliable and precise measurements can be achieved in the analysis of electron beam properties.

3.2. Investigation on Energy and Energy Measurement Using Real Beam Simulation

In this section, the energy measurement system was investigated through the beam dynamic simulation of an electron bunch with practical electron beam conditions, following the second scenario described in Section 2.3. The electron distributions obtained from the start-to-end beam dynamic simulation are represented in Figure 7. At the considered experimental station, the electron bunch exhibited a charge of 117.6 pC, a duration of 0.36 ps, and an average kinetic energy of 20.07 MeV, with an energy spread of 0.55% (equivalent to an energy spread of 111 keV). The horizontal and vertical beam sizes are 3.45 mm and 3.60 mm, respectively. Additionally, the horizontal and vertical emittance values are 0.41 mm·mrad and 0.25 mm·mrad, respectively.

The study results in Section 3.1 indicate that to achieve accurate energy measurement below 10%, the horizontal beam size must be less than 3 mm. To fulfill this requirement, two quadrupole magnets placed between the linac exit and the dipole magnet were employed to control the horizontal beam size to remain below 3 mm at the entrance of the dipole magnet, while the vertical beam size did not exceed the vertical screen size. Simulation results determined that the field strengths for the two quadrupole magnets were 1.25 T/m and 1.31 T/m, respectively. Under this condition, the horizontal beam size at the dipole magnet entrance was approximately 2.2 mm. However, due to a large energy spread, the transverse distribution on the screen station is larger than the screen size. To overcome this problem, the dipole magnetic field was adjusted to 153.7 mT for measuring the lost part. The particle distribution of the lost part was then incorporated into the energy spectrum. The complete energy spectrum, encompassing these adjustments, is illustrated in Figure 8. From this simulation, it was determined that with an optimal bunch charge of 117.5 pC, the kinetic energy of 20.06 MeV could be measured with a systematic error of 0.05%. Additionally, the energy spread of 95 keV could be measured with a systematic error of 14%. The energy measurement system exhibited a resolution of 3.5 keV per step.

4. Conclusions

The goal of this study is to design and optimize the electron energy spectrometer through beam dynamic simulations utilizing ASTRA software. In this spectrometer, the electron beam energy and energy spread are measured using a dipole magnet equipped with a screen station. The designed beam monitor system possesses the ability to capture images with a resolution of 0.1 mm per pixel. The 3D magnetic field of the dipole magnet was also included in simulations. By injecting a single electron with an energy range of 5 to 25 MeV into the dipole magnetic field at a bending angle of 45${}^{\circ }$, it was determined that the effective length of the dipole magnet should be reduced by 1.61% to account for the influence of the dipole fringe field. To assess the limitations and accuracy of the energy spectrometer, electron bunches with Gaussian distributions were generated and injected into the dipole magnetic field. The study results reveal that the average kinetic energy can be measured with a systematic error of less than 0.1% for all beam conditions. For a given transverse beam size, a beam with lower energy spread exhibits a larger systematic error compared to a beam with higher energy spread. Furthermore, the impact of transverse beam size on energy measurement is more pronounced with larger beam sizes than with smaller ones. Based on the findings, it was determined that electron bunches with beam sizes smaller than 3 mm and a minimum energy spread of 0.4% result in a systematic error of average energy measurement below 10%. Additionally, the systematic error of energy spread measurement remains below 10% when the emittance is less than 12 mm·mrad.

To verify the performance of the energy measurement system, an example electron bunch was optimized at the experimental station. This involved adjusting the magnetic fields of all magnets along the beam transport line, as well as the linac gradient, phase, and energy filtering in the alpha magnet. The simulation results indicate that, at the experimental station, the electron bunch had an average energy of 20.07 MeV, a charge of 117.6 pC, and an energy spread of 0.55%. The bunch also possessed an rms length of 0.36 ps, with horizontal and vertical emittances of 0.41 and 0.26 mm·mrad, respectively. To maintain a horizontal beam size of less than 3 mm, quadrupole magnets positioned between the linac exit and the dipole magnet were used. Subsequently, the electron beam was directed through the dipole magnetic field towards the screen station for energy measurement. The simulation results demonstrated that the measured average kinetic energy was 20.06 MeV, exhibiting 0.05% deviation from the value at the experimental station. The energy spread was determined to be 95 keV, with a systematic error of 14% in relation to the energy measurement resolution of 3.5 keV per pixel. It can be fairly stated that the results from this work demonstrate the effectiveness of this method in achieving accurate and precise energy and energy spread measurements, while also highlighting the associated limitations and resolutions of the energy measurement system.