The Development of Pepper-Pot Emittance Monitor in Gunma University

The pepper-pot emittance monitor is a device that can measure the phase space distribution of the beam rapidly. We have introduced one at the ion source test stand of Gunma University. We confirmed that the phase space distribution can be measured accurately comparing to the existing wire-slit type emittance monitor.


Introduction
Emittance measurement of the ion source is crucial because the injected beam into a linear accelerator or a cyclotron must match their acceptance.Slit-and-wire emittance monitors (SWEM), which utilize a beam cutting slit and detecting wires, are commonly used.However, such SWEM require several minutes to measure the emittance because the slit has to be moved physically during the measurement.Pepper-pot emittance monitors (PPEM) consist of a pepper-pot (PP) mask, an MCP and a fluorescent plate positioned downstream of the ion source, and a camera that observes the fluorescence [1,2].Particles passing through the apertures of the PP mask spread due to the angular divergence.By measuring them with the MCP and fluorescent plate, we can determine the phase space distribution of the beam.As measurement time is shorter than SWEM, PPEMs have been developed by several institutions such as RIKEN [3], RCNP [4,5], QST [6], and others.We referenced their work in our study.
Gunma University Heavy Ion Medical Center (GHMC) utilizes a KeiGM [7], a permanent magnet ECR ion source with an RF frequency of 10 GHz, to extract approximately 200 μA of C 4+ at an extraction voltage of 30 kV.It stably produces the C 4+ ion beam from CH4 gas for injection of heavy ion therapy operations for a year without maintenance.Gunma University has another ion source, the KeiGM2, which is equivalent to the KeiGM.This acts as a backup machine for GHMC and is also employed for research purposes.In addition to CH4, it can handle various gases such as He, N2, O2, Ne, Ar, Kr, and Xe.In the ion source test stand, the ion species are separated by 90-degree bending magnets and slits, and the ions are focused by Einzel lenses and triple electrostatic quadrupole lenses (figure 1).
The SWEM in the test stand is the same type used at the GHMC.It was primarily designed to measure the distribution of beams accelerated to 4 MeV/u after the linear accelerator before injection to the synchrotron.As such, the measurement range was somewhat limited, especially for beams larger than 100 π mm mrad.Expanding the observation to a larger phase space area was also an objective in the development of the PPEM.

Principle of PPEM
The working principle of PPEM [1,2] is illustrated in figure 2. When an ion beam arrives from the left side, only the beamlets that pass through the apertures in the PP mask can reach the MCP monitor on the right side.These beamlets make the electrons amplified by the MCP and cause the fluorescent plate to glow.The position of the aperture in the mask through which the particle passed gives positional information of the beamlet.The beamlet which reaches the MCP monitor has angular spread, so the beamlet spreads by the amount of its original divergence at the MCP.So the size of fluorescence on the MCP monitor gives the information of the spread of the beamlet.Plotting the brightness distribution of each location of the apertures in the PP mask, the phase space distribution can be reproduced.If the beam distribution from one aperture on the MCP monitor overlaps with the neighboring distribution, it will be unclear which aperture the particle passed through.So the distance of the PP mask to MCP determines the limit of measurable beam angle spread.

Production of PPEM
A thin (50 μm) etched nickel plate was used for the PP mask because it is a material that can withstand the thermal load of the beam.In order to observe angularly broadened beams, the pinhole spacing was set to a relatively wide 3 mm to allow measurements to ±25 mrad.We considered making the distance between the PP mask and the MCP adjustable [3,4] to change the measurable size of the phase space.However, considering cost issues, we decided to fix the distance at 58 mm.We used a Hamamatsu F2226-24P-Y003 with a diameter of 77 mm for our MCP monitor.The base of the PP mask, MCP, and mirror can be retracted from the beamline by an air cylinder when the PPEM is not in use.The negative high voltage for the MCP the positive high voltage for the fluorescent plate are each supplied through a high voltage feedthrough in the upper part of the chamber.Voltages up to 3 kV can be applied, though lower voltages are typically used.Since the base if electrically floated from the ground and connected to BNC connector with feedthrough, current of the beam hitting the PP mask can also be measured.The schematic view of PPEM is shown in figure 3.

Optical analysis of the fluorescent light of PPEM
We utilized an industrial USB 3.0 camera (DMK33UX252, SONY 1/1.8"CMOS, 2048 × 1536 of resolution, 120 fps) from The Imaging Source, which was capable of high-speed imaging to measure phenomena with small time widths, and the lens of F1.8 and 25 mm focal length.The images shown in figure 4 taken by the camera are sent to a PC for analysis using Python.When analyzing, a constant value was subtracted from the raw image of the beamlets.Then, we summed up the brightness of the pixels in both the X and Y directions.From this modified distribution, we calculated the beam distribution originating from a row of apertures.
When analyzing images, it's essential to determine the relationship between the fluorescence from the PP mask and the beam's position and spread.We observed the beam distribution in a state where the electrostatic Q-lens was turned off and the slit located just downstream of the analyzing magnet was narrowed.In this case, we can assume that the beam comes straight from the slit.We assumed that the center of the beamline cross-section corresponds to the fluorescence position from the beam at the center aperture of the PP mask.Furthermore, because the slit, pepper-pot's aperture and the glowing spots on MCP can be judged to be in a straight line, we determined which the pixel position of the glowing image corresponds to the position and angle.

Results and Discussion
The image obtained from the camera changes in response to the voltage applied to the MCP.The higher the voltage applied to the MCP, the more electrons are multiplied and the signal is enhanced, but the higher the voltage, the more optical noise is generated, especially in the center of the image.Also, the higher the voltage applied to the fluorescent plate, the more the fluorescence intensity increases.If the voltage is too high, the fluorescent agent will become saturated.After some adjustments, we decided to apply a voltage of -0.85 kV to the MCP and 0.85 kV to the fluorescent plate to maximize the fluorescence intensity within the range that does not increase noise.
The He 2+ beam with a current of 270 μA and an extraction voltage of 30 kV was measured with both SWEM (shown in figure 5) and PPEM (shown in figure 6).The beam was focused by the einzel lens and 3 electrostatic quadrupole electrodes to the proper size.The beam current was measured by the Faraday cup and then the Faraday cup was removed from the beamline to input the beam to the emittance monitors.
The phase space distributions taken by the two emittance monitors are compared.The horizontal and vertical emittance slits of the SWEM and the PP mask are 829 mm and 919 mm apart, respectively.So the phase space distribution of PPEM is considered to drift by the length from that of SWEM. Figure 7 shows the result of drifting the phase space distribution of figure 5 from the location of SWEM to PPEM.When comparing the two distributions, one can see that, although the line thicknesses differ, the distributions appear similar.
The horizontal and vertical emittance measured by PPEM was 425 π mm mrad and 362 π mm mrad, respectively.The horizontal and vertical emittance measured by SWEM was 185 π mm mrad and 153 π mm mrad, respectively.the SWEM claims to measure 95% emittance, assuming that it reads out the current linearly.While PPEM observes fluorescence, and there is no linearity between the current and fluorescence, which could lead to an overestimation of the low-density beam region.To improve the accuracy in the weak areas around the beam, it is necessary to further investigate the MCP output for the weak beam.

Conclusion
A PPEM was assembled at the ion source test stand of Gunma University.Comparing the measurement results with those of the PPEM and the SWEM, we managed to obtain reliable phase space plots.In the future, we aim to use the PPEM to measure the emittance of beams of various charge and ion species by adjusting parameters like the RF power, frequency, bias voltage, etc.

Figure 3 .
Figure 3.The schematic view of PPEM.Figure 4. Typical picture of the fluorescence from MCP of PPEM.When analyzing, a constant value was subtracted from each point in the image as a background.

Figure 4 .
Figure 3.The schematic view of PPEM.Figure 4. Typical picture of the fluorescence from MCP of PPEM.When analyzing, a constant value was subtracted from each point in the image as a background.

Figure 5 .
Figure 5.The phase space plot of He 2+ beam taken with SWEM.

Figure 6 .
Figure 6.The phase space plot of He 2+ beam taken with PPEM.

Figure 7 .
Figure 7. Drift space progression (X: 829 mm, Y: 919 mm) of the result of SWEM in the figure 6 to the place of PPEM.