Production and separation of positron emitters for hadron therapy at FRS-Cave M

Abstract

The FRagment Separator FRS at GSI is a versatile spectrometer and separator for experiments with relativistic in-flight separated short-lived exotic beams. One branch of the FRS is connected to the target hall where the bio-medical cave (Cave M) is located. Recently a joint activity between the experimental groups of the FRS and the biophysics at the GSI and Department of physics at LMU was started to perform biomedical experiments relevant for hadron therapy with positron emitting carbon and oxygen beams. This paper presents the new ion-optical mode and commissioning results of the FRS-Cave M branch where positron emitting ¹⁵O-ions were provided to the medical cave for the first time. An overall conversion efficiency of 2.9±0.2×10^{−4 15}O fragments per primary ¹⁶O ion accelerated in the synchrotron SIS18 was reached.

1. Introduction

The European project on Biomedical Applications of Radioactive Beams, BARB,¹ was launched at GSI in 2021. It aims at pre-clinical validation of in-vivo beam visualization and ion-beam therapy with positron-emitting isotopes of carbon and oxygen [1–3].

The fragment separator FRS [4] at GSI, a versatile separator and spectrometer, is ideal for the production and in-flight separation of positron emitters. Although both the FRS and the biomedical Cave M at GSI are long existing, the possibility of using fragment beams from FRS in the Cave M has never been explored before. The BARB project triggered this development and first commissioning results are presented here.

2. New ion-optical mode from FRS target to Cave M

As part of the planned experiments with radioactive ion beams at Cave M, an ion-optical mode using the existing GSI beamlines as depicted in Fig. 1 was calculated. The upper and lower half of Fig. 1 shows the magnetic elements of the FRS in the horizontal x- and vertical y-plane, respectively. The quadrupole magnets are shown in red (x-focusing) and blue (y-focusing), and dipole magnets in cyan. The two quadrupole magnets shown in gray next to F3 and F5 are at zero field in this optics due to present operation limitations. The hexapole magnets, located in front of and behind the large dipole magnets, are shown in gray and are at zero field. The size of these ion-optical elements in Fig. 1 corresponds to the apertures (vertical scale) and lengths (horizontal scale). The FRS quadrupole magnets (starting from the production target at F0 up to the dipole magnet between F7 and F8) have a special star-shaped aperture to maximize the transmission, the dimension shown in the figure show the aperture in planes x and y, whereas the inscribed radius of the star-shaped chamber is limited to 85 mm. A particular challenge of the beam line starting from F8 is the small apertures, because this part of the beam line was designed and built for transportation of primary beam with much smaller emittance. Therefore the dispersion should be kept small between F8-Cave M to ensure high transmission of the secondary beams, which are characterized by a large momentum spread of the order of a few precent. A second constraint of the ion-optics was to use the same polarities of the quadrupole magnets as they are when beam is sent directly from SIS18 to Cave M.

Fig. 1. Ion-optical calculation for FRS-Cave M.

The upper and lower half shows the beam envelopes and the magnetic elements of the FRS in the x- and y-plane, respectively. The beam envelopes (solid lines) are shown for a elliptical phase space with ε_x = 30 π mm mrad, ε_y = 20 π mm mrad, x₀ = 1.5 mm, y₀ = 2.2 mm and δp₀ = 0%. The dispersion line (dotted, black) represents a momentum deviation of 1%. For this case, the vertical scale represents the positive and negative coordinate in the horizontal x-direction. The vertical dashed line in Cave M indicates the position 161.38 m where the waist in x and y is fitted. The names of the different focal planes are indicated on top and follows the convention used at the FRS. The missing planes F4 and F6 belong to other branches of the FRS. The size of magnets correspond to their apertures (vertical scale) and lengths (horizontal scale). The elements between F8 and Cave M belong to the high-energy beam transport line of GSI, which was designed for primary beams; therefore they have smaller apertures than the FRS magnetic elements.

The ion-optical notation of the transfer matrix elements, R_ij, and sigma matrix elements, σ_ij, follows that of the programs Transport [5] and MIRKO [6] and characteristic elements are listed in Table 1. The mode is overall achromatic, meaning that R₁₆ = 0.0 m and R₂₆ = 0.0 rad for F0-Cave M and the ion-optical condition of a waist in both planes x and y is fulfilled at the position 161.38 m behind the production target, such that a small beam spot is achieved at the target position in Cave M. For this optics we use a production target on the second target ladder at F0, which is located 156 cm in front of the first quadrupole. The degrader needed for isotope separation is placed at the dispersive focal plane F2, where the dispersion is −4.89 m. The momentum resolving power is 870 for an object size of x₀ = 1.5 mm and is reduced compared to the standard ion-optical mode of the FRS.

Table 1.

Calculated ion-optical matrix elements of the optics mode depicted in Fig. 1. For this mode, the momentum resolving power at F2 is 870 for an object size of x₀=1.5 mm. The total length of the system F0-Cave M is 161.38 m.

F0–F1	F0–F2	F0–F8	F0-Cave M	F8-Cave M
R16 = 2.08 m	R16 = −4.89 m	R16 = 0.0 m	R16 = 0.0 m	R16= 0.0 m
R26 =0.1 rad	R26 = 0.0 rad	R26 = 0.0 rad	R26 = 0.0 rad	R26= 0.0 rad
R11 = −1.82	R11 = 1.87
R12 = 0.0 m/rad	R12 = 0.0 m/rad		${\sigma }_{12}^2$ = 0.0
${\sigma }_{34}^2$ = 0.0	R34 = 0.0 m/rad		${\sigma }_{34}^2$ = 0.0

The maximum value and the standard deviation of incident angles and momentum deviation of the ion distributions that are accepted by the optics was calculated in fifth order with MIRKO. In the calculation, the ions were distributed uniformly in the momentum coordinate, δp₀, and also in a four-dimensional $(x_0,x_0^{\prime },y_0,y_0^{\prime })$ ellipse with x₀ = y₀ = 1.5 mm. The acceptance is determined to be $x_{0,max}^{\prime }$ = 31 mrad, ${\sigma }_{x_0^{\prime }}$ = 13.5 mrad, $y_{0,max}^{\prime }$ = 14 mrad, ${\sigma }_{y_0^{\prime }}$ = 7.4 mrad, δp_0,max = 2% and ${\sigma }_{\delta }{{}_{{}_p}}_{{}_{{}_{{}_0}}}$ = 0.87% with δp₀ = +0.2%.

The calculated ion-optical mode was investigated experimentally using a primary beam. Basic properties were measured along the beam line with the profile monitors installed in the sections F1, F2, F3, F5, and two at F8. The central Bρ was scaled by +/−0.5% and the position of the beam determined. In this way, the measured beam positions were compared to the positions expected from the calculated matrix elements. The experimentally determined values for the matrix elements are: R₁₆ = 2.25 m (F0–F1), R₁₆ = -5.5 m (F0–F2), R₁₆ = 0.5 m (F0–F3), R₁₆ = 3.8 m (F0–F5) and R₁₆ = -0.7 m (F0–F8, first profile monitor), R₁₆ = 0.0 m (F0–F8, second profile monitor) and R₂₆ = 2.3 mrad (F0–F8). The section F8-Cave M was confirmed to be achromatic by changing the rigidity of the beam with the insertion of a piece of matter at F8.

3. Measurements with a beam of positron emitting ¹⁵O in Cave M

A 400 MeV/u ¹⁶O beam impinging on a 8045 ± 1 mg/cm² beryllium target located at the entrance of the FRS was used to produce ¹⁵O ions. They were separated by B_ρ − ΔE-B_ρ method [4]. At the midfocal plane an achromatic degrader was placed. The total thickness of the material on the beam axis at the midfocal plane corresponds to the energy loss in 2642 mg/cm² aluminum. The mean energy of the ¹⁵O beam in Cave M was estimated to be 286 MeV/u using a range measurement method as described in Ref. [2].

Particle identification was performed event-by-event by B_ρ − ΔE-ToF method [4]. Two TPCs for position tracking and a scintillator for ToF and energy loss measurements were installed at F2 and a second scintillator for position, ToF and energy loss measurements was installed in Cave M. The reconstructed particle identification is shown in Fig. 2 panel (a). The plot shows that the beam reaching Cave M is completely dominated by ¹⁵O, and only a small fraction of about 0.1% of A/q = 2 is seen in addition to the fragment of interest. This contaminant could be further reduced by the use of slits, but this would adversely affect the delivered intensity of ¹⁵O. Considering that planned bio-medical experiments in Cave M require highest possible intensity of ¹⁵O and the observed contaminant level is acceptable, no further efforts were made. Simulations show that the measured contaminant is ¹⁴N produced in the production target at F0.

Fig. 2.

(a) Particle identification scatter plot of ions arriving at Cave M measured event-by-event. Only a small fraction (of about 0.1%) of A/q = 2 is seen in addition to the fragment of interest, ¹⁵O. The main goal was to obtain a maximum intensity of ¹⁵O and therefore no additional separation with slits were applied to reduce the level of contaminants further. Simulations show that the most prominent contaminant is ¹⁴N produced in the production target located at F0. (b) The standard deviation of the momentum spread of the ¹⁵O is measured to be 0.6%. (c) The measured beam spot size on the detector located at the position indicated with a dashed line in Fig. 1 in Cave M has a standard deviation of 7.9 mm.

The measured distribution of the relative momentum deviation of the ¹⁵O ions reaching Cave M is shown in panel (b). The distribution is asymmetrical and has a mean value of +0.4% and the momentum spread has a standard deviation ${\sigma }_{\Delta p/p0}$ The non zero mean value and the asymmetry is caused by the acceptance between F2 and Cave M. The momentum distribution of ¹⁵O ions at F2 is centered and symmetric. The measured position distribution of the ¹⁵O beam at the plane where the scintillator is placed in Cave M is shown in panel (c).

The beam intensity in the SIS18 was measured with a beam transformer and recorded in the data acquisition system. A conversion factor of 2.9±0.2×10^{−4 15}O in Cave M per primary ¹⁶O ion accelerated in the SIS18 was achieved.

4. Summary and conclusions

Recently a joint effort several experimental groups was launched to perform biomedical experiments relevant for a possible future development of hadron therapy with positron emitting carbon and oxygen beams [1]. As one of the first steps of the project, a new ion-optical mode from the FRS production target to the biomedical cave was developed and tested with the goal to produce, separate and transport secondary beams efficiently to Cave M. In a first experimental run, a pure beam of positron emitting ¹⁵O was produced and separated with the FRS and sent to the biomedical cave at GSI. An overall conversion efficiency of about 2.9 ± 0.2 x 10^{−4 15}O fragments per primary ¹⁶O ion accelerated in the synchrotron was reached. This conversion efficiency could be further increased if one overcome present limitations in terms of small apertures and choice of polarities of the magnets in the high-energy beam transport line, and also the possibility use the two FRS quadrupole magnets marked in gray in Fig. 1.

The maximum intensity of an oxygen beam that has been reached in the synchrotron SIS18 is approximately 1×10¹¹ ions per spill and similar values have also been reached for carbon and nitrogen beams [7], consequently a rate of approximately 3 × 10^{7 15}O ions per spill in Cave M should be achievable. A typical cycle time is on the order of 2 s, thus a rate relevant for clinical use is in reach, cf. beam parameter set values for carbon beam intensity at the clinical treatment center HIT Heidelberg is in the range of 2 × 10⁶−8 × 10⁷ ions per second [8].