- Email: [email protected]
Research of hybrid single cavity linac
NIM B Beam Interactions with Materials & Atoms
Nuclear Instruments and Methods in Physics Research B 261 (2007) 17–20 www.elsevier.com/locate/nimb
Research of hybrid single cavity linac Taku Ito *, Noriyosu Hayashizaki, Naoko Matsunaga, Takuya Ishibashi, Jun Tamura, Liang Lu, Toshiyuki Hattori Research Laboratory for Nuclear Reactors, Tokyo Institute of Technology, N-25 2-12-1 O-okayama Meguro-ku, Tokyo 152-8550, Japan Available online 14 April 2007
Abstract We present a study of the compact proton linac for boron neutron capture therapy (BNCT) by using the nuclear reaction 10B(n, a)7Li. In order to downsize the linac system, we propose a hybrid type linac combined with radio frequency quadrupole (RFQ) electrodes and drift tube electrodes in a single cavity. We designed an interdigital-H (IH) mode linac with high power efficiency to accelerate proton beams from an injection energy of 40 keV to an acceleration energy of 3 MeV and an operation frequency of 80 MHz. The resonance frequencies and electromagnetic fields were optimized by simulation. As a result, we were able to design cavity lengths of less than 2 m in the simulation model. For this purpose, high frequency properties were carried out in the electromagnetic field simulation. The designed concept of the hybrid linac and the results of electromagnetic field simulation will be in this report. Ó 2007 Elsevier B.V. All rights reserved. PACS: 29.17.+w; 29.27. a Keywords: RFQ; IH drift tube linac; BNCT
1. Introduction BNCT is mainly radiotherapy used for brain tumors with cancer cells in which a boron compound has been previously taken in. The tumor area is then irradiated with a thermal or epithermal neutron beam and is treated using cytocidal 10B(n, a)7Li reactions at subcellular locations. Recently, linacs as a neutron source for BNCT have been studied in several laboratories [1–3]. One of the desired factors in this accelerator is downsizing so it can be introduced in general therapeutic facilities. In the neutron source for BNCT, we propose a new concept, a hybrid single cavity (HSC) linac with a novel electrode structure. The HSC linac is combined with the electrodes of a radio frequency quadrupole (RFQ) and an interdigital-H type drift tube (IH) in a single cavity. Conventionally, equipment using RF operation ((e.g.) for high frequency power sources) differs from one accelerator *
Corresponding author. Tel.: +81 3 5734 3055; fax: +81 3 5734 3850. E-mail address: [email protected] (T. Ito).
0168-583X/$ - see front matter Ó 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.nimb.2007.04.101
to another. Therefore, linac system has to need the broad space by just that much. HSC linac is able to downsize its cavity length more than those linacs. Furthermore, space for the linac system is more saving it. In this study, we designed the compact proton linear accelerator linac for BNCT by using the nuclear reaction 7Li(p, n)7Be. In order to realize the possibility of the HSC linac we investigated the possibility of using electric-field excitation for proton acceleration in a HSC cavity. Characterization of electromagnetic field distribution in the elaborated cavity was investigated by a three-dimensional simulator. 2. Basic design The schematic of the HSC linac is shown in Fig. 1. The HSC linac structure was composed of a RFQ section and a IH-drift tube section. The electrode structure of RFQ section adopted the interdigital-H type because this hybrid linac contains the electrodes of IH drift tube within. Therefore, the resonance of electromagnetic-field must be the quasi-TE1 1 0 mode of ion acceleration. On the other hand,
18
T. Ito et al. / Nucl. Instr. and Meth. in Phys. Res. B 261 (2007) 17–20
Fig. 1. Schematic of hybrid single cavity linac.
3. Accuracy evaluation of calculated values
120
Resonant frequency (MHz)
applied alternative phase focusing (APF) was applied to the IH-drift tube section in order to shorten the cell length [4]. The two ridges, each with attached electrodes, are fixed oppositely at the top and bottom of the cavity tank. We have an assumption that the cavity length is within 2 m in order to achieve practical use in mid-level therapeutic facilities.
115 110 105 100 95 90 85 80 75 70 60
In design of linac using simulator, we previously investigated its precision in advance. The low level RF measurements in the IH linac model resonator have previously been studied at the Institute for Nuclear Study, University of Tokyo (INS) [5]. This model was constructed to demonstrate the operational capabilities of the tuning method while permitting the length, width and height of the ridges to be variable in order to allow for systematic measurements of the RF characteristics. To examine the reliability of the calculated value, we created the IH models using high frequency structure simulator (HFSS), a three-dimen450
80
100
120
140
Inner diameter of IH section cavity (cm) Fig. 3. The examples of electric-field distribution and the coordinate on beam line. (d): field was concentrated RFQ side, j: field was concentrated IH side, (m): higher order (no TE111 mode).
sional simulation software tool for electric design, and computed the resonant frequencies, electric fields [6]. We simulated the electromagnetic fields of those cold models and modified the distance between opposite ridges as shown Fig. 2. The relative error of the low-order mode frequency between the actual measurements and the calculated values was about 1.74%. When the actual facility is designed, this result will be considered in terms of HSC linac size.
350 f1 cal. f1 act. f2 cal. f2 act. f3 cal. f3 act. f4 cal. f4 act.
300
250
200
150 100 50
100
150
200
250
The distance between opposite ridges (cm) Fig. 2. The resonant frequencies of calculated value and actual measurement in the distance between opposite ridges. d: calculated value, (j): actual measurement value.
0.9
Electric-field strength (V/m)
Resonant Frequency (MHz)
400
0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 260
280
300
320
340
360
380
ETR length (mm) Fig. 4. Resonance frequency in the cavity for different ERT. d: RFQ section, (j): IH section.
T. Ito et al. / Nucl. Instr. and Meth. in Phys. Res. B 261 (2007) 17–20
4. Adjustment for constant capacity distribution along the cavity The electric-field strength of the HSC cavity must be constant through out the beam line. However, the field of the inner HSC cavity is inclined towards the beam injection side because the RFQ section, more crowded with the electrodes than IH section, is condensation of capacitances. To adjust this unequal field distribution, we improved the HSC cavity geometry. First, in order to modify the cavity diameter at the IH section, the field distribution was spread along the beam output side. Secondly, as the End Ridge tuner ERT length was adjusted, the cavity was made to adopt a constant capacity distribution along the cavity. The electric-field strength was investigated by means of field simulations. Adjustments of the resizing cavity diameter at the IH section and the ERT length were performed. The more the diameter of IH drift tube section was increased, the more the frequency of both sides were approached, and electric field was exited from input side to output side when inside diameter was 900–1100 mm as shown in Fig. 3. We measured the electric field distribution E from 280 to 360 cm at 20 cm intervals of the ERT at the exit of the linac and set the diameter of IH drift tube side at cavity 900 mm. The optimum electric-field distribution was obtained with an ERT length of 320 mm as shown in Figs. 4 and 5. It suggested that the HSC linear accelerator of complicated structure could excite the electric field that was necessary for acceleration from these results. To evaluate a linac performance of the HSC linac we calculated shunt impedance in this structure. The effective shunt impedances Zeff of the simulated models were calculated separately and divided into RFQ and IH sections. RFQ section’s shunt impedance Zr is given as Zeff = ZT2 = V2/PL where T is the transit time factor, V is the voltage of interelectrode gap, P is the average power loss, and L is the cavity length. The IH section’s shunt impedance Zi is given as Zi = Cb 2D3f 3.5 where b is the synchronous particle velocity divided by light velocity, D is the diameter of the cavity, f is the resonant frequency and C is the coefficient of effective shunt impedance for an IH structure
19
LINAC, as C = Zeffb2D 3f 3.5. Therefore, theZr and Zi were confirmed to be 754 kX/m and 360 MX/m. Fig. 6 shows a comparison of the effective shunt impedance of the linacs structures as suggested by Hattori et al. [7]. This indicates that the HSC linac can demonstrate high shunt impedance as just as well as other linacs do. Based on those results, the basic parameters of the HSC linac were chosen
Fig. 6. Effective shunt impedance of the IH structure and other acceleration structures.
Table 1 Basic parameters of the linac RFQ section Particle Input energy (MeV) Output energy (MeV) Cavity length (m) Cavity diameter (mm) Resonant frequency (MHz) RF power (kW)
Fig. 5. Electric field strength at (a) RFQ and (b) IH drift tube sections.
IH section +
H 0.05 0.25 0.68 400
0.25 3 1.18 900 77.2 31.3
20
T. Ito et al. / Nucl. Instr. and Meth. in Phys. Res. B 261 (2007) 17–20
as seen in Table 1, those parameters suggest that HSC linac has the potential of downsize enough for practical use. 5. Conclusion In these simulated data, the electric-field distribution in the HSC linac, which accelerated H+ up to 40 keV from 3 MeV was well controlled by modifications made to the IH section diameter and the ERT length. In addition, high shunt impedance of the RFQ and IH sections at 754 kX/m and 360 MX/m respectively were performed. Those results suggest that the HSC linac with complex structure is readily achievable for a BNCT neutron source. We plan to conduct further research via a beam dynamics simulation for additional evaluation of the HSC linac performance.
References [1] D.A. Swenson, W.J. Starling, F.W. Guy, K.R. Crandall, in: Proc. Particle Acceleratitor Confeerence, 2003, p. 2889. [2] R. Terlizzi, N.N. Colonna, E. Biscegline, P. Colangelo, S. Marrone, A. Raino, G. Tagliente, V. Variale, Nucl. Instr. and Meth. B 213 (2004) 210. [3] Y.D. Harker, J.F. Harmon, G.W. Irwin, Nucl. Instr. and Meth. B 99 (1995) 843. [4] K. Yamamoto, T. Hattori, M. Okamura, N. Hayashizaki, et al., Nucl. Instr. and Meth. B 244 (2006) 467. [5] S. Yamada, T. Hattori, T. Fujino, T. Fukushima, T. Murakami, E. Tojyo, K. Yoshida, INS-NUMA-57, 1985. [6] http:www.ansoft.com. [7] T. Hattori, K. Sasa, M. Okamura, T. Ito, H. Tomizawa, T. Katayose, N. Hayashizaki, T. Yoshida, K. Isokawa, M. Aoki, N. Fujita, M. Okada, Fusion Eng. Des. 32–33 (1996) 359.
Nuclear Instruments and Methods in Physics Research B 261 (2007) 17–20 www.elsevier.com/locate/nimb
Research of hybrid single cavity linac Taku Ito *, Noriyosu Hayashizaki, Naoko Matsunaga, Takuya Ishibashi, Jun Tamura, Liang Lu, Toshiyuki Hattori Research Laboratory for Nuclear Reactors, Tokyo Institute of Technology, N-25 2-12-1 O-okayama Meguro-ku, Tokyo 152-8550, Japan Available online 14 April 2007
Abstract We present a study of the compact proton linac for boron neutron capture therapy (BNCT) by using the nuclear reaction 10B(n, a)7Li. In order to downsize the linac system, we propose a hybrid type linac combined with radio frequency quadrupole (RFQ) electrodes and drift tube electrodes in a single cavity. We designed an interdigital-H (IH) mode linac with high power efficiency to accelerate proton beams from an injection energy of 40 keV to an acceleration energy of 3 MeV and an operation frequency of 80 MHz. The resonance frequencies and electromagnetic fields were optimized by simulation. As a result, we were able to design cavity lengths of less than 2 m in the simulation model. For this purpose, high frequency properties were carried out in the electromagnetic field simulation. The designed concept of the hybrid linac and the results of electromagnetic field simulation will be in this report. Ó 2007 Elsevier B.V. All rights reserved. PACS: 29.17.+w; 29.27. a Keywords: RFQ; IH drift tube linac; BNCT
1. Introduction BNCT is mainly radiotherapy used for brain tumors with cancer cells in which a boron compound has been previously taken in. The tumor area is then irradiated with a thermal or epithermal neutron beam and is treated using cytocidal 10B(n, a)7Li reactions at subcellular locations. Recently, linacs as a neutron source for BNCT have been studied in several laboratories [1–3]. One of the desired factors in this accelerator is downsizing so it can be introduced in general therapeutic facilities. In the neutron source for BNCT, we propose a new concept, a hybrid single cavity (HSC) linac with a novel electrode structure. The HSC linac is combined with the electrodes of a radio frequency quadrupole (RFQ) and an interdigital-H type drift tube (IH) in a single cavity. Conventionally, equipment using RF operation ((e.g.) for high frequency power sources) differs from one accelerator *
Corresponding author. Tel.: +81 3 5734 3055; fax: +81 3 5734 3850. E-mail address: [email protected] (T. Ito).
0168-583X/$ - see front matter Ó 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.nimb.2007.04.101
to another. Therefore, linac system has to need the broad space by just that much. HSC linac is able to downsize its cavity length more than those linacs. Furthermore, space for the linac system is more saving it. In this study, we designed the compact proton linear accelerator linac for BNCT by using the nuclear reaction 7Li(p, n)7Be. In order to realize the possibility of the HSC linac we investigated the possibility of using electric-field excitation for proton acceleration in a HSC cavity. Characterization of electromagnetic field distribution in the elaborated cavity was investigated by a three-dimensional simulator. 2. Basic design The schematic of the HSC linac is shown in Fig. 1. The HSC linac structure was composed of a RFQ section and a IH-drift tube section. The electrode structure of RFQ section adopted the interdigital-H type because this hybrid linac contains the electrodes of IH drift tube within. Therefore, the resonance of electromagnetic-field must be the quasi-TE1 1 0 mode of ion acceleration. On the other hand,
18
T. Ito et al. / Nucl. Instr. and Meth. in Phys. Res. B 261 (2007) 17–20
Fig. 1. Schematic of hybrid single cavity linac.
3. Accuracy evaluation of calculated values
120
Resonant frequency (MHz)
applied alternative phase focusing (APF) was applied to the IH-drift tube section in order to shorten the cell length [4]. The two ridges, each with attached electrodes, are fixed oppositely at the top and bottom of the cavity tank. We have an assumption that the cavity length is within 2 m in order to achieve practical use in mid-level therapeutic facilities.
115 110 105 100 95 90 85 80 75 70 60
In design of linac using simulator, we previously investigated its precision in advance. The low level RF measurements in the IH linac model resonator have previously been studied at the Institute for Nuclear Study, University of Tokyo (INS) [5]. This model was constructed to demonstrate the operational capabilities of the tuning method while permitting the length, width and height of the ridges to be variable in order to allow for systematic measurements of the RF characteristics. To examine the reliability of the calculated value, we created the IH models using high frequency structure simulator (HFSS), a three-dimen450
80
100
120
140
Inner diameter of IH section cavity (cm) Fig. 3. The examples of electric-field distribution and the coordinate on beam line. (d): field was concentrated RFQ side, j: field was concentrated IH side, (m): higher order (no TE111 mode).
sional simulation software tool for electric design, and computed the resonant frequencies, electric fields [6]. We simulated the electromagnetic fields of those cold models and modified the distance between opposite ridges as shown Fig. 2. The relative error of the low-order mode frequency between the actual measurements and the calculated values was about 1.74%. When the actual facility is designed, this result will be considered in terms of HSC linac size.
350 f1 cal. f1 act. f2 cal. f2 act. f3 cal. f3 act. f4 cal. f4 act.
300
250
200
150 100 50
100
150
200
250
The distance between opposite ridges (cm) Fig. 2. The resonant frequencies of calculated value and actual measurement in the distance between opposite ridges. d: calculated value, (j): actual measurement value.
0.9
Electric-field strength (V/m)
Resonant Frequency (MHz)
400
0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 260
280
300
320
340
360
380
ETR length (mm) Fig. 4. Resonance frequency in the cavity for different ERT. d: RFQ section, (j): IH section.
T. Ito et al. / Nucl. Instr. and Meth. in Phys. Res. B 261 (2007) 17–20
4. Adjustment for constant capacity distribution along the cavity The electric-field strength of the HSC cavity must be constant through out the beam line. However, the field of the inner HSC cavity is inclined towards the beam injection side because the RFQ section, more crowded with the electrodes than IH section, is condensation of capacitances. To adjust this unequal field distribution, we improved the HSC cavity geometry. First, in order to modify the cavity diameter at the IH section, the field distribution was spread along the beam output side. Secondly, as the End Ridge tuner ERT length was adjusted, the cavity was made to adopt a constant capacity distribution along the cavity. The electric-field strength was investigated by means of field simulations. Adjustments of the resizing cavity diameter at the IH section and the ERT length were performed. The more the diameter of IH drift tube section was increased, the more the frequency of both sides were approached, and electric field was exited from input side to output side when inside diameter was 900–1100 mm as shown in Fig. 3. We measured the electric field distribution E from 280 to 360 cm at 20 cm intervals of the ERT at the exit of the linac and set the diameter of IH drift tube side at cavity 900 mm. The optimum electric-field distribution was obtained with an ERT length of 320 mm as shown in Figs. 4 and 5. It suggested that the HSC linear accelerator of complicated structure could excite the electric field that was necessary for acceleration from these results. To evaluate a linac performance of the HSC linac we calculated shunt impedance in this structure. The effective shunt impedances Zeff of the simulated models were calculated separately and divided into RFQ and IH sections. RFQ section’s shunt impedance Zr is given as Zeff = ZT2 = V2/PL where T is the transit time factor, V is the voltage of interelectrode gap, P is the average power loss, and L is the cavity length. The IH section’s shunt impedance Zi is given as Zi = Cb 2D3f 3.5 where b is the synchronous particle velocity divided by light velocity, D is the diameter of the cavity, f is the resonant frequency and C is the coefficient of effective shunt impedance for an IH structure
19
LINAC, as C = Zeffb2D 3f 3.5. Therefore, theZr and Zi were confirmed to be 754 kX/m and 360 MX/m. Fig. 6 shows a comparison of the effective shunt impedance of the linacs structures as suggested by Hattori et al. [7]. This indicates that the HSC linac can demonstrate high shunt impedance as just as well as other linacs do. Based on those results, the basic parameters of the HSC linac were chosen
Fig. 6. Effective shunt impedance of the IH structure and other acceleration structures.
Table 1 Basic parameters of the linac RFQ section Particle Input energy (MeV) Output energy (MeV) Cavity length (m) Cavity diameter (mm) Resonant frequency (MHz) RF power (kW)
Fig. 5. Electric field strength at (a) RFQ and (b) IH drift tube sections.
IH section +
H 0.05 0.25 0.68 400
0.25 3 1.18 900 77.2 31.3
20
T. Ito et al. / Nucl. Instr. and Meth. in Phys. Res. B 261 (2007) 17–20
as seen in Table 1, those parameters suggest that HSC linac has the potential of downsize enough for practical use. 5. Conclusion In these simulated data, the electric-field distribution in the HSC linac, which accelerated H+ up to 40 keV from 3 MeV was well controlled by modifications made to the IH section diameter and the ERT length. In addition, high shunt impedance of the RFQ and IH sections at 754 kX/m and 360 MX/m respectively were performed. Those results suggest that the HSC linac with complex structure is readily achievable for a BNCT neutron source. We plan to conduct further research via a beam dynamics simulation for additional evaluation of the HSC linac performance.
References [1] D.A. Swenson, W.J. Starling, F.W. Guy, K.R. Crandall, in: Proc. Particle Acceleratitor Confeerence, 2003, p. 2889. [2] R. Terlizzi, N.N. Colonna, E. Biscegline, P. Colangelo, S. Marrone, A. Raino, G. Tagliente, V. Variale, Nucl. Instr. and Meth. B 213 (2004) 210. [3] Y.D. Harker, J.F. Harmon, G.W. Irwin, Nucl. Instr. and Meth. B 99 (1995) 843. [4] K. Yamamoto, T. Hattori, M. Okamura, N. Hayashizaki, et al., Nucl. Instr. and Meth. B 244 (2006) 467. [5] S. Yamada, T. Hattori, T. Fujino, T. Fukushima, T. Murakami, E. Tojyo, K. Yoshida, INS-NUMA-57, 1985. [6] http:www.ansoft.com. [7] T. Hattori, K. Sasa, M. Okamura, T. Ito, H. Tomizawa, T. Katayose, N. Hayashizaki, T. Yoshida, K. Isokawa, M. Aoki, N. Fujita, M. Okada, Fusion Eng. Des. 32–33 (1996) 359.