Development of a fast scintillator based beam phase measurement systemfor compact superconducting cyclotronsTanushyam Bhattacharjee, Malay Kanti Dey, Partha Dhara, Suvodeep Roy, Jayanta Debnath et al. Citation: Rev. Sci. Instrum. 84, 053303 (2013); doi: 10.1063/1.4807076 View online: http://dx.doi.org/10.1063/1.4807076 View Table of Contents: http://rsi.aip.org/resource/1/RSINAK/v84/i5 Published by the American Institute of Physics. Additional information on Rev. Sci. Instrum.Journal Homepage: http://rsi.aip.org Journal Information: http://rsi.aip.org/about/about_the_journal Top downloads: http://rsi.aip.org/features/most_downloaded Information for Authors: http://rsi.aip.org/authors

REVIEW OF SCIENTIFIC INSTRUMENTS 84, 053303 (2013)

Development of a fast scintillator based beam phase measurement systemfor compact superconducting cyclotrons

Tanushyam Bhattacharjee, Malay Kanti Dey, Partha Dhara, Suvodeep Roy, JayantaDebnath, Rajendra Balakrishna Bhole, Atanu Dutta, Jedidiah Pradhan, Sarbajit Pal,Gautam Pal, Amitava Roy, and Alok ChakrabartiVariable Energy Cyclotron Centre, 1/AF Bidhannagar, Kolkata 700064, India

(Received 23 February 2013; accepted 4 May 2013; published online 21 May 2013)

In an isochronous cyclotron, measurements of central phase of the ion beam with respect to rf andthe phase width provide a way to tune the cyclotron for maximum energy gain per turn and efficientextraction. We report here the development of a phase measurement system and the measurementscarried out at the Variable Energy Cyclotron Centre’s (VECC’s) K = 500 superconducting cyclotron.The technique comprises detecting prompt γ -rays resulting from the interaction of cyclotron ion beamwith an aluminium target mounted on a radial probe in coincidence with cyclotron rf. An assemblycomprising a fast scintillator and a liquid light-guide inserted inside the cyclotron was used to detectthe γ -rays and to transfer the light signal outside the cyclotron where a matching photo-multipliertube was used for light to electrical signal conversion. The typical beam intensity for this measurementwas a few times 1011 pps. © 2013 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4807076]

I. INTRODUCTION

In heavy ion cyclotrons, beam phase measurement sys-tem is used as an effective tool for precise isochronous mag-netic field tuning. The non-intercepting capacitive phase mea-surement, which measures currents associated with the imagecharges produced by the accelerated beam, is a method widelyused for phase measurement in cyclotrons.1, 2 However, thismethod needs intense beam bunches to produce the requiredphase information.3, 4 Moreover, the phase measuring methodin a compact cyclotron, that uses capacitive pick-up, gives anaverage over several turns.5

At MSU, phase measurement was carried out using ascintillator detector positioned on the viewport hole just out-side the beam acceleration zone6 and the γ emission result-ing out of beam hitting the main probe was detected. Thisarrangement, obviously, was not designed for obtaining thecomplete radial phase information. To overcome this prob-lem, Si-PIN diode was used by the same group so that the de-tector can be inserted inside the cyclotron. However, radiallyvarying signal-to-noise ratio due to rf pick-up inside the cy-clotron imposes a serious limitation in accuracy of the phaseinformation.7

Plastic scintillator was also placed inside separated sectorcyclotron at GANIL to measure the phase information of verylow intensity radioactive ion beams.8, 9 This method can beused only with low intensity beam since the beam hits directlythe scintillator detector.

At the extraction zone of the compact superconductingcyclotron the magnetic field falls sharply and the beam phasechanges abruptly. One needs to measure the phase accuratelyat short intervals in this region to ensure efficient beam ex-traction. This has been the main motivation for developinga scintillator based phase measurement system, by detectingprompt γ -rays, reported in this article. In the phase measure-ment methods by detecting γ -rays, one would need to over-

come the challenge of extracting the phase signal in presenceof high γ -background resulting from beam losses.10 In ourdesign the signal to background ratio has been remarkablyimproved. The details of the phase measurement setup andthe measurements carried out using the same are reported inSecs. II and III.

II. EXPERIMENTAL DETAILS

A. Experimental arrangement

The phase probe assembly comprises an aluminiumtarget-plate where the ion beam would be allowed to hit toproduce prompt γ -rays, a glass window with “O” ring sealto isolate the scintillator assembly from the cyclotron beamchamber vacuum, the scintillator placed behind the glass win-dow, and the liquid light-guide to carry the scintillation signaloutside the cyclotron where a matching PMT is coupled tothe light-guide. The front-end of the probe that is inside thecyclotron is shown in Figure 1. The aluminium target plate ismounted on a radial steel tube. Except the aluminium platethe entire assembly is isolated from the beam chamber vac-uum. The beam induced prompt γ -rays produced due to fu-sion reaction of neon (beam) with aluminium (target) weredetected by the cylindrical scintillator detector, inserted in-side the 20 mm bore of the phase probe tube at a distance of50 mm from the target plate tip. The scintillator detector wascovered by an aluminium casing and one end of the liquidlight-guide was inserted into the casing ensuring proper cou-pling with the outer circular edge of the scintillator detectorto minimize light leakage. A liquid light-guide of 5 m lengthwas used to transfer the light signal from the scintillator to thePMT. The PMT was placed outside the cyclotron to minimisethe interference from the magnetic field and the rf interfer-ence. A remotely operated linear drive system was used tomove the probe along the radius and the radial information

0034-6748/2013/84(5)/053303/5/$30.00 © 2013 AIP Publishing LLC84, 053303-1

053303-2 Bhattacharjee et al. Rev. Sci. Instrum. 84, 053303 (2013)

FIG. 1. Sectional view of phase probe head.

was obtained from a position encoder attached with the probeassembly.

The insertion diameter in the mechanical assembly of theprobe was 20 mm and because of this constraint we have useda cylindrical shape plastic scintillator detector of diameter16 mm. The outer surface of the scintillator was wrapped witha few layers of Teflon and aluminium foil as reflective layeron the scintillator surface to maximize light transfer towardslight-guide. The aluminium foil, outside the scintillator sur-face, blocks the external light and also acts as a good heatabsorber. UV optical glue was used to couple the scintillatorand PMT with the light-guide as better count rate was ob-served in the off-line test with the optical glue as compared tobare coupling.

To get good time resolution, a BC418 plastic scintilla-tor was chosen for its fast rise (decay) time of 0.5 (1.4) ns.The TAC full width at half maximum (FWHM) obtained forγ -γ coincidence using a 60Co source placed in between twoBC418 scintillators, positioned face to face in close geom-etry, was about 450 ps. For this measurement, Ortec 467TAC and Ortec 935 CFDs were used. This time resolution(FWHM) is good enough for measurement of central phaseand phase width of the beam with respect to rf since beamphase width in our cyclotron is typically about 30◦, whichat 19 MHz rf frequency represents a time width of littlemore than 4 ns. This scintillator has other desirable propertiestoo. For example, the wavelength of maximum emission is391 nm, which perfectly matches with the spectral responseof the light-guide. Also, BC418 does not exhibit any self-radiation property and the scintillator is quite insensitive toother particles such as electrons, protons, alpha particles, neu-tron, etc. The liquid light guide of model Newport-77554 waschosen for its high transmittance in the UV and visible range.The transmittance of the light guide was measured in the lab-oratory and was found to be 90%. A ten-stage, UV-sensitive,Photonis-XP2978 PMT was used in the final experimentalsetup. The PMT was coupled with the light guide by Tefloncoupler to minimize light leakage and also to prevent sparkscaused due to high voltage bias of PMT. The PMT was thenshielded by a mu-metal to minimise the magnetic field that inturn ensures efficient light conversion.

The K500 superconducting cyclotron at Variable EnergyCyclotron Centre (VECC) has two beam diagnostics probes.One of them moves along the spiral central line of a hill, calledmain-probe (MP) and the other moves straight along a radialline across another hill which was used to place borescopeearlier, as shown in Figure 2. The Al target with the scintilla-tor assembly was installed on this second probe, called here-

FIG. 2. The median plane sectional view of the VECC K500 superconduct-ing cyclotron, showing three dees, the main probe (MP), and the phase probe(PP) on which the target plate was mounted.

after as the phase-probe (PP). The phase-probe can be insertedup to a radial distance of 400 mm starting from the extractiontrajectory at 670 mm radius.

The experiment was carried out with Ne4+ beam accel-erated in 2nd harmonic mode at 19 MHz rf frequency. Theexpected beam energy was about 8 MeV/A at the extractionradius. The energy of Ne beam was sufficient to overcome thecoulomb barrier and induce nuclear reaction in the Al targetat 400 mm. The length of the aluminium target was more thanthe beam size which ensures that the target stops one or atmost a few turns. The rf signal was picked up from the dee,which is adjacent to the PP (B-DEE in Figure 2). The circu-lating beam first crosses the reference dee and then hits thealuminium target.

B. Data acquisition and processing

The NIM electronics and CAMAC-based data acqui-sition system were used to extract the phase information.The signal from the PMT-anode was amplified by CAENN978 variable-gain fast-amplifier module having four chan-nels, where two of its channels were cascaded to increase thesignal level of the weak PMT-output. The PMT-output wastransmitted using a coaxial cable and finally fed to the inputof a Constant Fraction Discriminator (CFD) module (Ortec935) to generate the timing pulse that used as the START sig-nal for the ORTEC 567 Time-to-Amplitude Converter (TAC),as shown in Figure 3. The STOP signal for the TAC wastaken from rf and fed through the Leading Edge Discrimi-nator (LED) module (Phillips-Scientific). The full scale timeinterval of TAC was set as 100 ns as one full cycle of 19 MHzRF signal is equivalent to 52 ns. The LED timing pulse wasdelayed by a variable delay module (CAEN N108A) to keepthe spectrum within the selected range of TAC. The analogoutput of the TAC (10 V full scale), representing the time

053303-3 Bhattacharjee et al. Rev. Sci. Instrum. 84, 053303 (2013)

FIG. 3. Schematic diagram of the phase measurement setup.

difference between the PMT output and the rf signal, wasdigitized by the Ortec AD 413 CAMAC peak sensing ADCsuitable for high resolution spectroscopic applications. Themaximum conversion gain of 8 K of AD 413 was used for thedata acquisition.

Figure 4 shows the measured raw TAC spectrum at threedifferent radii of the cyclotron. The peak of the spectrum rep-resents the central phase of the beam at a particular radiusand phase width is obtained from the FWHM value of thespectrum. The peak and FWHM was measured by fitting thecurve with standard Gaussian distribution. A typical exampleis shown in Figure 5.

III. RESULTS AND DISCUSSION

In the VECC’s K = 500 superconducting cyclotron, thereare 14 trim coils (TC-0 to TC-13) for the tuning of the beam.The second and the 14th trim coils are also used as harmonic

FIG. 4. Raw TAC spectrum at three different radial positions.

FIG. 5. A typical Gaussian fit of the TAC spectrum at radial position of520 mm.

coils. Figure 6 shows the un-calibrated central phase of thebeam along the radius of the cyclotron for three sets of trimcoil currents. The curve with symbol triangle (�) continuestill extraction radius (∼670 mm). The beam current profilefor this trim-coil setting on the main probe, moving on a ra-dial range corresponding to phase probe movement range, isshown in Figure 7. The other two curves in Figure 6 wereobtained with different current settings of the trim coil 11(TC-11). The curve with symbol circle (◦) was obtained byincreasing TC-11 current and the curve with symbol box (�)is obtained decreasing TC-11 current.

The phase measurements with respect to rf voltage asshown in Figure 6 were done at radial positions along thestraight line path of the PP movement. Since the angulardistance of PP from the reference dee increases with radius(Figure 2) due to spiral structure of hills and dees, the phase

FIG. 6. Un-calibrated beam phase for three sets of trim coil currents.

053303-4 Bhattacharjee et al. Rev. Sci. Instrum. 84, 053303 (2013)

FIG. 7. Beam current versus the probe position for a tuned setting of trimcoils.

FIG. 8. Un-calibrated beam phase after “spiral-correction” of the data.

FIG. 9. Beam central phase curve after applying Garren and Smith beamphase detuning technique.

FIG. 10. Phase-width of the beam along radius.

values need to be corrected. The phase curve after “spiral-correction” is shown in Figure 8. The corrected phase curvestill needs an absolute calibration.

For this, we used the Garren and Smith’s beam phase de-tuning technique.11–13 Beam detuning in our case was done byvarying TC-11 current while the rf frequency was kept fixed.The TC-11 current was varied till the beam intensity as mea-sured in the main probe was brought down to half of its value.In such a case the central phase of the beam attained either ofthe limiting values of ±90◦.

All the phase points were again calibrated with respectto the absolute value of ±90◦. The redundancy in calibra-tion (i.e., either by +90◦ or by −90◦) allowed crosschecking.The final phase curves of the beam after spiral-correction andphase calibration are shown in Figure 9. The rising trend ofthe phase at the outer radii is due to falling field at the edge ofthe cyclotron poles.

Figure 10 shows the phase width values along the ra-dius. The phase width remains around 27◦ from 400 mm to600 mm. The increase in the phase width beyond 600 mm isdue to the overlapping turns.

IV. CONCLUSION

A beam phase measurement technique has been devel-oped based on γ detection using a fast plastic scintillator de-tector. The measurements show that this technique can beused as an effective tool for optimum isochronous tuning atthe outer radii, especially at the extreme outer regions closeto the extraction radius, where phase is expected to changerather sharply. This technique can also be used to detect anymagnetic field related error at the outer radii. Further, if nec-essary, the same probe can also be used for the simultaneousmeasurement of the beam current with minor modification.

ACKNOWLEDGMENTS

We thank all our colleagues, related to various sub-system of the cyclotron, who made this experiment successful

053303-5 Bhattacharjee et al. Rev. Sci. Instrum. 84, 053303 (2013)

by providing un-interrupted beam time. We are also thankfulto Dr. S. Bhattacharya, Dr. S. R. Banerjee, and Dr. S. K. Dasfor their valuable suggestions during the laboratory test of var-ious components to design this phase probe.

1J. Gustafsson et al., Nucl. Instrum. Methods Phys. Res. A 335, 417(1993).

2J. H. Timmer et al., Nucl. Instrum. Methods Phys. Res. A 568, 532(2006).

3T. Itahashi et al., in Proceedings of the 13th International Conference onCyclotrons and their Applications, edited by G. Dutto and M.K. Craddock(World Scientific Publishing Co., Singapore, 1993), p. 491.

4J. Zheng et al., in Proceedings of the 19th International Conference onCyclotrons and their Applications, p. 263 (http://accelconf.web.cern.ch/AccelConf/Cyclotrons2010/papers/proceed.pdf).

5P. Miller et al., Annual Report of the Michigan State University CyclotronLaboratory (1974), p. 112.

6B. Milton et al., in Proceedings of the 10th International Conference onCyclotrons and their Applications, edited by F. Marti (IEEE, New York,1984), p. 55.

7J. D. Bailey et al., Proceedings of the 1995 Particle Accelerator Conference1, 369–371 (1995).

8L. Boy et al., Nucl. Instrum. Methods Phys. Res. A 399, 195 (1997).9F. Chautard et al., in Proceedings of the 16th International Conference onCyclotrons and their Applications, edited by F. Marti (American Instituteof Physics, New York, USA, 2001), p. 370.

10M. Fowler et al., Annual Report of the Michigan State University NationalSuperconducting Cyclotron (1985), p. 183.

11A. A. Garren and L. Smith, in Proceedings of the 3rd InternationalConference on Sector-Focused Cyclotrons and Meson Factories, editedby F. T. Howard and N. Vogt-Nilsen, CERN 63-19 (1963), p. 18(http://accelconf.web.cern.ch/AccelConf/c63/INDEX.htm).

12R. E. Berg, H. G. Blosser, and M. M. Gordon, Nucl. Instrum. MethodsPhys. Res. A 58, 327 (1968).

13C. Baumgarten et al., Nucl. Instrum. Methods Phys. Res. A 570, 10 (2007).