- Email: [email protected]
Recent developments and upgrades in ion source technology and ion beam systems at HVE
Nuclear Instruments and Methods in Physics Research B 371 (2016) 137–141
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb
Recent developments and upgrades in ion source technology and ion beam systems at HVE Nicolae C. Podaru ⇑, Dirk J.W. Mous High Voltage Engineering Europa B.V., P.O. Box 99, Amersfoort 3800AB, Netherlands
a r t i c l e
i n f o
Article history: Received 12 July 2015 Received in revised form 11 October 2015 Accepted 11 October 2015 Available online 22 October 2015 Keywords: Ion source technology Ion accelerator Ion beam analysis Ion (nuclear) microprobe Brightness
a b s t r a c t In this paper we discuss various ion sources used in particle accelerator systems dedicated to ion beam analysis techniques. Key performance and characteristics of some ion sources are discussed: emittance, brightness, gas consumption, sample consumption efficiency, lifetime, etc. For negative ion sources, we focus on the performance of volume H ion sources (e.g. HVE model 358), the duoplasmatron negative ion source and the magnetically filtered multicusp volume sources (e.g. HVE model SO-120). The duoplasmatron ion source has been recently upgraded with a Ta filament to deliver up to 150 lA H ion beams and in conjunction with the Na charge exchange canal up to 20 lA of He. The available brightness from the duoplasmatron increased from 2 to 6 A m2 rad2 eV1. The ion source has been incorporated in a stand-alone light ion injector, well suited to deliver 20–30 keV negative ion beams of H, He, C, NHx and O to accelerate for most ion beam analysis techniques. Ó 2015 Elsevier B.V. All rights reserved.
1. Introduction Ion beams with energies of 1–4 MeV are commonly used in accelerator facilities to perform routine Ion Beam Analysis (IBA) of various samples, ranging from biological cells to cultural heritage and semiconductor devices [1–4]. Rutherford Backscattering Spectrometry (RBS), Particle Induced X-ray Emission (PIXE), Particle Induced Gamma-Ray Emission (PIGE), Nuclear Reaction Analysis (NRA), and Elastic Recoil Detection (ERD) are well-established techniques that provide quantitative information about structure and elemental composition of materials. The size of the ion beam on target can be reduced to 1 lm or below using high demagnification lenses, enabling techniques such as l-PIXE which provide elemental maps of the samples with (sub)micron resolution [1]. Most common ion beam species required in the techniques mentioned above are 1H, 4He, 12C, 15N, 18O and, in the case of ERD, heavier ions. Typical ion beam currents used in IBA are ranging from tens of pA up to hundreds of nA, for which the ion source is not a limiting factor. However, for microprobe applications an ion source that is able to deliver high ion beam brightness (typically >10 A m2 rad2 eV1) is essential. In this work we investigate the typical ion sources used in accelerators for IBA and we highlight their performance and key attributes. We also highlight the recent developments in ion source technology at High Voltage ⇑ Corresponding author. Tel.: +31 33 461 97 41; fax: +31 33 461 52 91. E-mail address: [email protected] (N.C. Podaru). http://dx.doi.org/10.1016/j.nimb.2015.10.021 0168-583X/Ó 2015 Elsevier B.V. All rights reserved.
Engineering Europa (HVE) with focus on upgraded duoplasmatron (358) ion source, the Na charge exchange canal and a new standalone single source ion injector for production of negative light ions such as H, He, C, NHx, and O.
2. Ion sources and ion beams – attributes and characteristics Key attributes that characterize ion source performance include: ion beam emittance, ion beam brightness, plasma hash, energy spread, and gas/solid target consumption efficiency. The ion beam emittance e is defined as the area S of the twodimensional phase space divided by p; e = S/p and also: ex ¼ x0 x00 and ey ¼ y0 y00 , x0 and y0 are the beam radii in horizontal (x) and vertical (y) planes at a waist, while x00 and y00 are the beam half angle divergences in the x and the y planes. The energy-normalized emittance is obtained by multiplying the emittance e with the square root of ion beam energy. Degradation of normalized beam emittance from the ion source to the target is a measure of the quality of the beam transport system. The ion beam brightness (B) is defined as the ion beam (particle) current, i, that can be transported through two apertures of areas, Ao (object area) and Ac (collimator area), separated by a drift L, at a 2
given ion beam energy (E): B ¼ A0 iL . To allow comparison of the Ac E performance of the microprobes at various facilities, Szymansky and Jamieson defined the normal brightness [5] at a half-angle divergence of 0.07 mrad and at an object slit opening yielding
138
N.C. Podaru, D.J.W. Mous / Nuclear Instruments and Methods in Physics Research B 371 (2016) 137–141
the highest ion beam brightness. Clearly, high beam brightness is advantageous for microprobe applications. The ion beam energy spread can negatively influence the optical properties of high quality lenses, i.e. nuclear microprobe lenses. For tandem accelerators, the ion beam energy spread DE is dependent on mainly three factors: the initial beam energy spread from the ion source, the energy spread introduced by the interaction with the stripper gas, and the terminal voltage ripple. The beam energy spread introduced by the ion source is dependent on many plasma parameters, which ultimately transfer kinetic energy to the ions. Ion sources with ion beam energy spread below 10 eV are considered good for IBA experiments. This value should be compared to the terminal voltage ripple for all-solid-state particle accelerators, typically in the range of 15–100 VRMS.
RF ion sources (only positive ion extraction) coupled to charge exchange canals (vapors of Li, Na, Rb) can provide negative H, He, C, O ion beams for IBA. Charge exchange efficiency is typically a few percent, and the beam current is from several hundreds of nA’s up to 10 lA. The beam energy spread (100 eV) is large when compared to other direct negative extraction ion sources. The charge exchange process in the alkali vapor introduces additional energy spread. When discussing the alkali vapors used in the charge exchange canals, the work of Slachter et al. [8] indicates that the highest charge exchange efficiency for He is achieved for Na when the beam energy is 20 keV, energy well suited for beam transport and injection into TandetronTM accelerators.
3. Ion sources for different types of particle accelerators
A series of ion sources can produce direct negative extraction of ion beams of e.g. H, C, NHx, O. We will focus on the volume H ion sources, the magnetically filtered multicusp volume source [9–11] (e.g. HVE model SO-120) and the duoplasmatron negative ion source [12] (e.g. HVE model 358). The HVE model SO-120 multicusp ion source has been developed to deliver 1–3 mA intensity H ion beams with direct negative extraction. Details of this ion source has been discussed in several previous papers [13,14], indicating high beam brightness (30–40 A m2 rad2 eV1, value after ion beam extraction and mass analysis), the low beam energy spread (<5 eV), low hash (<1%) and the long filament lifetime (more than 500 h). High ion beam brightness and low beam energy spread delivered by the ion source have been key requirements for the IBA microprobe community, e.g. at the Jozef Stefan (JSI), Slovenia and the Technical University (TU) Munich, accelerator facilities. A sevenfold normal brightness increase (currently 14 A m2 rad2 eV1) at the microprobe lens position has been reported at JSI when compared to the previous duoplasmatron beam brightness. The brightness figure has been recorded when injecting only 20% of the beam current available from the SO-120 multicusp ion source [15]. Results from TU Munich indicate that the SO-120 ion source coupled to the 14 MV vdG tandem particle accelerator increased the brightness available at the SNAKE microprobe from 0.1 A m2 rad2 eV1 (from an electron cyclotron resonance ion source + CEC) to 0.8 A m2 rad2 eV1 [16]. The SO-120 multicusp ion source is housed in a small footprint (1.8 1.4 m2) dual source multicusp injector and the constructive details have been previously given [13]. For negative He ion beam production, the combination between the direct positive extraction HVE model SO-130 multicusp ion source (13 mA of He) and the Na CEC yields negative He ion beams currents in excess of 70 lA [14]. To conclude, the multicusp ion sources can yield high brightness & current hydrogen beams for tandem accelerators, brightness values that start to approach or even exceed beam brightness values typically reserved for single-ended particle accelerators. The duoplasmatron ion source can operate with direct negative extraction for ion beams of H (in off-axis beam extraction mode) or is used for direct positive extraction for He ion beams. Traditionally, the filament is a PtIr mesh coated with an oxide layer (e.g. BaSrCaCO3), for increased electron emission coefficient while operating at low filament temperature (1000 K). Since the duoplasmatron can provide direct extraction of C, NHx, and O, but these gases lower the lifetime of the PtIr gauze, HVE upgraded its duoplasmatron with a Ø1.3 mm Ta filament. The power supply package driving the duoplasmatron was upgraded to support operating the Ta filament. The filament support base has been modified to allow cooling of the filament feedthrough, while the filament shape and location have been optimized to facilitate stable plasma conditions. An XYZ manual stage was added to allow manual optimization of extraction geometry over a large current density range.
Single-ended particle accelerators benefit from the use of direct positive extraction ion sources located in the high voltage terminal. Positive ion source are more prolific in producing high intensity ion beams. Only the ion source energy spread and the accelerator terminal voltage ripple determine the beam energy spread. Typically, single-ended particle accelerators are preferred when highest brightness is required. State-of-the-art work in IBA with single-ended particle accelerators includes high-resolution microprobe work, with typical probe size down to 100 nm and below. The 3.5 MV SingletronTM accelerator systems installed at e.g. the National University of Singapore (NUS) [6] and at the Centre Etudes Nucléaires de Bordeaux, Gradignan [7] are specially designed for high brightness and high stability for microprobe applications. The most common ion source employed in single ended particle accelerators is the RF ion source. A 100 MHz RF field is capacitively coupled to magnetically confined plasma. Such ion sources can produce ion beams of H, He, N, O, Ne, Ar, Kr, Xe. The ion beam energy spread is typically about 100 eV while the normalized emittance is 1.5 p mm mrad MeV1/2. To date, the highest ion beam brightness (74 A m2 rad2 eV1) measured for microprobe applications, has been recorded at the SingletronTM at NUS, which is fed by the RF ion source HVE model 173. The gas consumption of the RF ion source for 3He can be as low as 0.25 sccm. Maintenance cycle of an RF ion source is at approx. 1000 h. Tandem accelerators require negative ion beams for the injection. In contrast to single-ended particle accelerators the ion sources are easily accessible and the injection configuration allows multiple ion sources to be applied with ease. Tandem accelerators have the advantage that the ion source maintenance can be done much faster than in single-ended particle accelerators. Negative ion sources for tandem injectors include gas sources: von Ardenne type, multicusp type or RF type and cesium sputter type ion sources. Cesium sputter type ion sources use energetic Cs+ beam of particles that strike a solid surface with target material. The sputtered atoms may pick up electrons due to a lowered work function introduced by the Cs layer at the surface of the target, thereby forming a negative beam. In such manner, negative ion beams (100’s of nA to hundreds of lA, depending on species) from almost the entire periodic table can be created. He beams cannot be created in Cs sputter type ion sources. Negative H ion beam currents created from TiH target pieces can be prolific (e.g. 100 lA HVE ion source model 860A, SO-110), intensities more than sufficient for IBA. Related to ion beam emittance the typical value for the emittance is 2–12 p mm mrad MeV1/2. However, microprobe work is usually performed with gas type ion sources since the brightness is at least a factor of three higher. The lifetime of the sputter target is a drawback for this type of ion source. Additionally, the ion beam current output may vary with time.
3.1. Gas discharge type ion sources
N.C. Podaru, D.J.W. Mous / Nuclear Instruments and Methods in Physics Research B 371 (2016) 137–141
Fig. 1. 3D CAD model of the Stand-Alone Light Ion Injector (SALII) including the upgraded duoplasmatron ion source, the utilities cabinet (source deck: power supplies, gas box, control hardware & extraction power supply, Einzellens power supply, control hardware, power distribution unit, etc.), Na CEC, 30° analyzing magnet (ME/q2 = 0.6 a.m.u. MeV).
In addition, any mechanical misalignments between the ion source outlet aperture and extraction electrode can be corrected. The filament lifetime is in excess of 500 h when running on inert gasses such as H and He. For H ion beam extraction the emittance of the ion source is <1.5 p mm mrad MeV1/2. In direct negative extraction we measured ion beam currents in 14 excess of >1 lA 12C; >10 lA 12CH NH; >5 lA 14NH 4 ; >1 lA 2; and >30 lA 16O. The typical gas consumption of the
139
duoplasmatron ion source is about 0.5 sccm with a 0.64 mm plasma outlet aperture. In addition to the different ion beam species, the total H beam current output has been increased from 50 lA up to 150 lA. When combining this duoplasmatron ion source with a Na CEC, the negative He beam current available for injection into the tandem accelerators is as high as 20 lA. The upgraded duoplasmatron has been integrated in a newly developed Stand-Alone Light Ion Injector (SALII), 3D CAD model depicted in Fig. 1. An Einzellens, a 30° analyzing magnet and an electrostatic Y-steerer allow ion optical matching with the beam line and/or the tandem accelerator that follows. This injector can be integrated with an existing HVE Multi-Purpose Injectors (MPI), at the 0° port of the MPI Magnet. The integration of a SALII with the MPI does not compromise the serviceability of MPI. One of the SALII injectors has been integrated in the accelerator facility from the Nuclear Physics Institute, Academy of Sciences, Rez, Czech Republic. The end-user indicates a beam brightness increase of 2–3 times when compared to ion beams obtained from the duoplasmatron located in the MPI injector [17]. The duoplasmatron performance improvements were quantified on a 3 MV TandetronTM accelerator system. This system is equipped with a Na CEC and with a terminal voltage stabilizing system based on a high dispersion 90° – 1500 mm radius analyzing magnet. We present the results on: hydrogen ion beam brightness at 3 MeV, ion beam current (H&He) particle transmission as a function of terminal voltage, and the ion beam energy stability using the 7Li(p,n)7Be nuclear reaction. 3.2. Performance measurements on accelerator system equipped with upgraded duoplasmatron ion source For this experiment, the duoplasmatron ion source delivered 75 lA H and 10 lA He at the entrance of the TandetronTM accelerator. The ion beam particle transmission tests were done in the range of 150 kV and 3000 kV (5–100% of terminal voltage range). The stripper gas thickness has been optimized to obtain maximum particle transmission through the accelerator, for the given energy
Fig. 2. Hydrogen and Helium ion beam particle transmission through the Tandetron accelerator as a function of accelerator terminal voltage between 150 kV and 3 MV. The stripper gas thickness has been optimized for maximum particle transmission.
140
N.C. Podaru, D.J.W. Mous / Nuclear Instruments and Methods in Physics Research B 371 (2016) 137–141
Fig. 3. Ion beam brightness measured for 3 MeV hydrogen at full object slits opening (X&Y) 0.1 and 0.15 mm as a function of half angle beam divergence.
turns increased the beam plasma density. Fig. 3 also shows reduced beam brightness for a half angle divergence below 50 lrad. This effect is often observed in microprobes and is likely related to hollow beam formation direct after extraction. Further investigations could clarify the reason for the reduced beam brightness with reducing half angle divergence and reduced aperture size. Microprobe experiments require long-term (several hours) ion beam energy stability for high demagnification microprobe lenses, enabling a constant probe resolution at target position. The functional description of the terminal voltage stabilization loop used in HVE accelerator systems has been previously reported in detail for a high beam power (in excess of 3 kW) accelerator system and one accelerator system used for microprobe applications [14,18]. In this work we discuss the results on the terminal voltage stability of a 3 MV T-shape TandetronTM obtained when using a 1500 mm radius, 90° analyzing magnet with ME/q2 of 48 a.m.u. MeV and powered by high stability power supply (10 ppm/h or better). The terminal voltage stability was investigated using the 7Li(p, n)7Be nuclear reaction, threshold energy 1880 keV. The results are summarized in Fig. 4. 11 runs (each lasting 10 min) have been recorded while the accelerator was controlled in slit stabilization mode. A linear fit on the experimentally determined energy variation yields that the ion beam energy variation is 13 eV/h, implying that the accelerator terminal voltage variation in slit stabilization mode is 6.5 ppm/h. The beam spot size increase is estimated to 3.4 nm/h due to such a beam energy drift, when the system has the following characteristics: microprobe lens with working distance of 150 mm, demagnification of 50 and microprobe acceptance of 70 lrad. 4. Summary
Fig. 4. Hydrogen ion beam energy stability function of time.
and charge state. The convolution between the stripping efficiency and the particle transmission through the accelerator yield the experimental curves depicted in Fig. 2. The ion beams were stripped in Argon gas. The optimum stripper gas thickness to obtain 2 MeV H or He beams was determined: 1H1+ 0.94 lg/cm2 (80%), 4He1+ 0.24 lg/cm2 (35.5%), and 4He2+ 0.56 lg/cm2 (23%). The ion beam particle transmission value is given in between brackets. It can be seen that for creating He ion beams with high energy <2.8 MeV, ion beam stripping for charge state 1+ is more prolific, whereas for He beam energy >2.8 MeV, charge state 2+ is more prolific. At 150 kV (only 5% of the nominal terminal voltage) the He beam particle transmission is 25%. The 3 MeV H ion beam brightness of the modified duoplasmatron has been measured with a setup using two micrometric precision XY slits separated by 4.25 m. The ion beam current was recorded in an electrostatically suppressed Faraday cup with 10 pA reading accuracy. The results on hydrogen brightness for the upgraded duoplasmatron ion source are summarized in Fig. 3. At 70 lrad half angle divergence, the ion beam brightness is 6.2 A m2 rad2 eV1 for 100 100 lm2 object size and 9 A m2 rad2 eV1 for 150 150 lm2 object size. This result indicates a threefold improvement in normal beam brightness when compared to the previous version of the duoplasmatron ion source (brightness of 2 A m2 rad2 eV1). The improvement in beam brightness value is attributed to the increased arc current that in
We have presented and discussed the most common ion sources used to create ion beam for IBA applications. Some key characteristics of the ion sources have been highlighted, with a strong emphasis on ion beam current output and ion beam brightness. The ion sources have been split in two groups, for positive ion production and negative ion production (with and without charge exchange). The RF ion sources are presently the most common for single-ended particle accelerators. They offer high ion beam brightness, and have a reasonable long lifetime. When discussing negative ion sources, we have focused on multicusp type ion sources and duoplasmatron type ion sources. Multicusp ion sources are very well suited to provide high brightness ion beams for microprobe work, brightness values bridging the gap between the tandem accelerators and single-ended accelerators. We also described the performance of the upgraded HVE duoplasmatron ion source with Ta filament when coupled to a tandem accelerator well suited for microprobe IBA applications. References [1] P. Vavpeticˇ, K. Vogel-Mikuš, L. Jeromel, N. Ogrinc Potocˇnik, P. Pongrac, D. Drobne, Zˇ. Pipan Tkalec, S. Novak, M. Kos, Š. Koren, M. Regvar, P. Pelicon, Nucl. Instr. Meth. B 348 (2015) 147. [2] L. Beck, Nucl. Instr. Meth. B 332 (2014) 439. [3] A. Zucchiatti, A. Climent Font, P.C. Gutierrez Neira, A. Perea, P. Fernandez Esquivel, S. Rovira Llorens, J.L. Ruvalcaba Sil, A. Verde, Nucl. Instr. Meth. B 332 (2014) 160. [4] A. Mackova, P. Malinsky, H. Pupikova, P. Nekvindova, J. Cajzl, Z. Sofer, R.A. Wilhelm, A. Kolitsch, J. Oswald, Nucl. Instr. Meth. B 332 (2014) 74. [5] R. Szymanski, D.N. Jamieson, Nucl. Instr. Meth. B 130 (1997) 80. [6] F. Watt, J.A. van Kan, I. Rajita, A.A. Bettiol, T.F. Choo, M.B.H. Breese, T. Osipowicz, Nucl. Instr. Meth. B 210 (2003) 14–20. [7] Ph. Barberet, L. Daudin, N. Gordillo, S. Sorieul, M. Simon, H. Seznec, I. Idarrage, S. Incerti, A. Balana, Ph. Moretto, Nucl. Instr. Meth. B 269 (2011) 2163–2167. [8] A.S. Schlachter, A.I.P. Conf. Proc. 111 (1984) 300–332. [9] M. Bacal, G.W. Hamilton, Phys. Rev. Lett. 42 (1979) 1538–1540.
N.C. Podaru, D.J.W. Mous / Nuclear Instruments and Methods in Physics Research B 371 (2016) 137–141 [10] M. Bacal, G.W. Hamilton, A.M. Bruneteau, H.J. Doucet, J. Taillet, Rev. Sci. Instr. 50 (1979) 719–721. [11] R.L. York, R.R. Stevens, K.N. Leung, K.W. Ehlers, Rev. Sci. Instr. 55 (1984) 681. [12] M. Von Ardenne, Technik 11/65, Tabellen der Elektronenphysik, Ionenphysik und Übermikroscopie, VEB Deutscher Verlag der Wissenschaften Berlin, 1956. [13] G.W.W. Quax, A. Gottdang, D.J.W. Mous, Rev. Sci. Instr. 81 (2010) 02A701. [14] N.C. Podaru, A. Gottdang, FRENA Group, D.J.W. Mous, Nucl. Instr. Meth. B 273 (2012) 231–233.
141
[15] P. Pelicon, N.C. Podaru, P. Vavpeticˇ, L. Jeromel, N. Ogrinc Potocnik, S. Ondracˇka, A. Gottdang, D.J.W. Mous, Nucl. Instr. Meth. B 332 (2014) 229–233. [16] M. Moser, P. Reichart, W. Carli, C. Greubel, K. Peeper, P. Hartung, G. Dollinger, Nucl. Instr. Meth. B 273 (2012) 226–230. [17] Václav Vosecˇek (personal communication, January 10, 2015). [18] D.J.W. Mous, R.G. Haitsma, T. Butz, R.-H. Flagmeyer, D. Lehmann, J. Vogt, Nucl. Instr. Meth. B 130 (1997) 31.
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb
Recent developments and upgrades in ion source technology and ion beam systems at HVE Nicolae C. Podaru ⇑, Dirk J.W. Mous High Voltage Engineering Europa B.V., P.O. Box 99, Amersfoort 3800AB, Netherlands
a r t i c l e
i n f o
Article history: Received 12 July 2015 Received in revised form 11 October 2015 Accepted 11 October 2015 Available online 22 October 2015 Keywords: Ion source technology Ion accelerator Ion beam analysis Ion (nuclear) microprobe Brightness
a b s t r a c t In this paper we discuss various ion sources used in particle accelerator systems dedicated to ion beam analysis techniques. Key performance and characteristics of some ion sources are discussed: emittance, brightness, gas consumption, sample consumption efficiency, lifetime, etc. For negative ion sources, we focus on the performance of volume H ion sources (e.g. HVE model 358), the duoplasmatron negative ion source and the magnetically filtered multicusp volume sources (e.g. HVE model SO-120). The duoplasmatron ion source has been recently upgraded with a Ta filament to deliver up to 150 lA H ion beams and in conjunction with the Na charge exchange canal up to 20 lA of He. The available brightness from the duoplasmatron increased from 2 to 6 A m2 rad2 eV1. The ion source has been incorporated in a stand-alone light ion injector, well suited to deliver 20–30 keV negative ion beams of H, He, C, NHx and O to accelerate for most ion beam analysis techniques. Ó 2015 Elsevier B.V. All rights reserved.
1. Introduction Ion beams with energies of 1–4 MeV are commonly used in accelerator facilities to perform routine Ion Beam Analysis (IBA) of various samples, ranging from biological cells to cultural heritage and semiconductor devices [1–4]. Rutherford Backscattering Spectrometry (RBS), Particle Induced X-ray Emission (PIXE), Particle Induced Gamma-Ray Emission (PIGE), Nuclear Reaction Analysis (NRA), and Elastic Recoil Detection (ERD) are well-established techniques that provide quantitative information about structure and elemental composition of materials. The size of the ion beam on target can be reduced to 1 lm or below using high demagnification lenses, enabling techniques such as l-PIXE which provide elemental maps of the samples with (sub)micron resolution [1]. Most common ion beam species required in the techniques mentioned above are 1H, 4He, 12C, 15N, 18O and, in the case of ERD, heavier ions. Typical ion beam currents used in IBA are ranging from tens of pA up to hundreds of nA, for which the ion source is not a limiting factor. However, for microprobe applications an ion source that is able to deliver high ion beam brightness (typically >10 A m2 rad2 eV1) is essential. In this work we investigate the typical ion sources used in accelerators for IBA and we highlight their performance and key attributes. We also highlight the recent developments in ion source technology at High Voltage ⇑ Corresponding author. Tel.: +31 33 461 97 41; fax: +31 33 461 52 91. E-mail address: [email protected] (N.C. Podaru). http://dx.doi.org/10.1016/j.nimb.2015.10.021 0168-583X/Ó 2015 Elsevier B.V. All rights reserved.
Engineering Europa (HVE) with focus on upgraded duoplasmatron (358) ion source, the Na charge exchange canal and a new standalone single source ion injector for production of negative light ions such as H, He, C, NHx, and O.
2. Ion sources and ion beams – attributes and characteristics Key attributes that characterize ion source performance include: ion beam emittance, ion beam brightness, plasma hash, energy spread, and gas/solid target consumption efficiency. The ion beam emittance e is defined as the area S of the twodimensional phase space divided by p; e = S/p and also: ex ¼ x0 x00 and ey ¼ y0 y00 , x0 and y0 are the beam radii in horizontal (x) and vertical (y) planes at a waist, while x00 and y00 are the beam half angle divergences in the x and the y planes. The energy-normalized emittance is obtained by multiplying the emittance e with the square root of ion beam energy. Degradation of normalized beam emittance from the ion source to the target is a measure of the quality of the beam transport system. The ion beam brightness (B) is defined as the ion beam (particle) current, i, that can be transported through two apertures of areas, Ao (object area) and Ac (collimator area), separated by a drift L, at a 2
given ion beam energy (E): B ¼ A0 iL . To allow comparison of the Ac E performance of the microprobes at various facilities, Szymansky and Jamieson defined the normal brightness [5] at a half-angle divergence of 0.07 mrad and at an object slit opening yielding
138
N.C. Podaru, D.J.W. Mous / Nuclear Instruments and Methods in Physics Research B 371 (2016) 137–141
the highest ion beam brightness. Clearly, high beam brightness is advantageous for microprobe applications. The ion beam energy spread can negatively influence the optical properties of high quality lenses, i.e. nuclear microprobe lenses. For tandem accelerators, the ion beam energy spread DE is dependent on mainly three factors: the initial beam energy spread from the ion source, the energy spread introduced by the interaction with the stripper gas, and the terminal voltage ripple. The beam energy spread introduced by the ion source is dependent on many plasma parameters, which ultimately transfer kinetic energy to the ions. Ion sources with ion beam energy spread below 10 eV are considered good for IBA experiments. This value should be compared to the terminal voltage ripple for all-solid-state particle accelerators, typically in the range of 15–100 VRMS.
RF ion sources (only positive ion extraction) coupled to charge exchange canals (vapors of Li, Na, Rb) can provide negative H, He, C, O ion beams for IBA. Charge exchange efficiency is typically a few percent, and the beam current is from several hundreds of nA’s up to 10 lA. The beam energy spread (100 eV) is large when compared to other direct negative extraction ion sources. The charge exchange process in the alkali vapor introduces additional energy spread. When discussing the alkali vapors used in the charge exchange canals, the work of Slachter et al. [8] indicates that the highest charge exchange efficiency for He is achieved for Na when the beam energy is 20 keV, energy well suited for beam transport and injection into TandetronTM accelerators.
3. Ion sources for different types of particle accelerators
A series of ion sources can produce direct negative extraction of ion beams of e.g. H, C, NHx, O. We will focus on the volume H ion sources, the magnetically filtered multicusp volume source [9–11] (e.g. HVE model SO-120) and the duoplasmatron negative ion source [12] (e.g. HVE model 358). The HVE model SO-120 multicusp ion source has been developed to deliver 1–3 mA intensity H ion beams with direct negative extraction. Details of this ion source has been discussed in several previous papers [13,14], indicating high beam brightness (30–40 A m2 rad2 eV1, value after ion beam extraction and mass analysis), the low beam energy spread (<5 eV), low hash (<1%) and the long filament lifetime (more than 500 h). High ion beam brightness and low beam energy spread delivered by the ion source have been key requirements for the IBA microprobe community, e.g. at the Jozef Stefan (JSI), Slovenia and the Technical University (TU) Munich, accelerator facilities. A sevenfold normal brightness increase (currently 14 A m2 rad2 eV1) at the microprobe lens position has been reported at JSI when compared to the previous duoplasmatron beam brightness. The brightness figure has been recorded when injecting only 20% of the beam current available from the SO-120 multicusp ion source [15]. Results from TU Munich indicate that the SO-120 ion source coupled to the 14 MV vdG tandem particle accelerator increased the brightness available at the SNAKE microprobe from 0.1 A m2 rad2 eV1 (from an electron cyclotron resonance ion source + CEC) to 0.8 A m2 rad2 eV1 [16]. The SO-120 multicusp ion source is housed in a small footprint (1.8 1.4 m2) dual source multicusp injector and the constructive details have been previously given [13]. For negative He ion beam production, the combination between the direct positive extraction HVE model SO-130 multicusp ion source (13 mA of He) and the Na CEC yields negative He ion beams currents in excess of 70 lA [14]. To conclude, the multicusp ion sources can yield high brightness & current hydrogen beams for tandem accelerators, brightness values that start to approach or even exceed beam brightness values typically reserved for single-ended particle accelerators. The duoplasmatron ion source can operate with direct negative extraction for ion beams of H (in off-axis beam extraction mode) or is used for direct positive extraction for He ion beams. Traditionally, the filament is a PtIr mesh coated with an oxide layer (e.g. BaSrCaCO3), for increased electron emission coefficient while operating at low filament temperature (1000 K). Since the duoplasmatron can provide direct extraction of C, NHx, and O, but these gases lower the lifetime of the PtIr gauze, HVE upgraded its duoplasmatron with a Ø1.3 mm Ta filament. The power supply package driving the duoplasmatron was upgraded to support operating the Ta filament. The filament support base has been modified to allow cooling of the filament feedthrough, while the filament shape and location have been optimized to facilitate stable plasma conditions. An XYZ manual stage was added to allow manual optimization of extraction geometry over a large current density range.
Single-ended particle accelerators benefit from the use of direct positive extraction ion sources located in the high voltage terminal. Positive ion source are more prolific in producing high intensity ion beams. Only the ion source energy spread and the accelerator terminal voltage ripple determine the beam energy spread. Typically, single-ended particle accelerators are preferred when highest brightness is required. State-of-the-art work in IBA with single-ended particle accelerators includes high-resolution microprobe work, with typical probe size down to 100 nm and below. The 3.5 MV SingletronTM accelerator systems installed at e.g. the National University of Singapore (NUS) [6] and at the Centre Etudes Nucléaires de Bordeaux, Gradignan [7] are specially designed for high brightness and high stability for microprobe applications. The most common ion source employed in single ended particle accelerators is the RF ion source. A 100 MHz RF field is capacitively coupled to magnetically confined plasma. Such ion sources can produce ion beams of H, He, N, O, Ne, Ar, Kr, Xe. The ion beam energy spread is typically about 100 eV while the normalized emittance is 1.5 p mm mrad MeV1/2. To date, the highest ion beam brightness (74 A m2 rad2 eV1) measured for microprobe applications, has been recorded at the SingletronTM at NUS, which is fed by the RF ion source HVE model 173. The gas consumption of the RF ion source for 3He can be as low as 0.25 sccm. Maintenance cycle of an RF ion source is at approx. 1000 h. Tandem accelerators require negative ion beams for the injection. In contrast to single-ended particle accelerators the ion sources are easily accessible and the injection configuration allows multiple ion sources to be applied with ease. Tandem accelerators have the advantage that the ion source maintenance can be done much faster than in single-ended particle accelerators. Negative ion sources for tandem injectors include gas sources: von Ardenne type, multicusp type or RF type and cesium sputter type ion sources. Cesium sputter type ion sources use energetic Cs+ beam of particles that strike a solid surface with target material. The sputtered atoms may pick up electrons due to a lowered work function introduced by the Cs layer at the surface of the target, thereby forming a negative beam. In such manner, negative ion beams (100’s of nA to hundreds of lA, depending on species) from almost the entire periodic table can be created. He beams cannot be created in Cs sputter type ion sources. Negative H ion beam currents created from TiH target pieces can be prolific (e.g. 100 lA HVE ion source model 860A, SO-110), intensities more than sufficient for IBA. Related to ion beam emittance the typical value for the emittance is 2–12 p mm mrad MeV1/2. However, microprobe work is usually performed with gas type ion sources since the brightness is at least a factor of three higher. The lifetime of the sputter target is a drawback for this type of ion source. Additionally, the ion beam current output may vary with time.
3.1. Gas discharge type ion sources
N.C. Podaru, D.J.W. Mous / Nuclear Instruments and Methods in Physics Research B 371 (2016) 137–141
Fig. 1. 3D CAD model of the Stand-Alone Light Ion Injector (SALII) including the upgraded duoplasmatron ion source, the utilities cabinet (source deck: power supplies, gas box, control hardware & extraction power supply, Einzellens power supply, control hardware, power distribution unit, etc.), Na CEC, 30° analyzing magnet (ME/q2 = 0.6 a.m.u. MeV).
In addition, any mechanical misalignments between the ion source outlet aperture and extraction electrode can be corrected. The filament lifetime is in excess of 500 h when running on inert gasses such as H and He. For H ion beam extraction the emittance of the ion source is <1.5 p mm mrad MeV1/2. In direct negative extraction we measured ion beam currents in 14 excess of >1 lA 12C; >10 lA 12CH NH; >5 lA 14NH 4 ; >1 lA 2; and >30 lA 16O. The typical gas consumption of the
139
duoplasmatron ion source is about 0.5 sccm with a 0.64 mm plasma outlet aperture. In addition to the different ion beam species, the total H beam current output has been increased from 50 lA up to 150 lA. When combining this duoplasmatron ion source with a Na CEC, the negative He beam current available for injection into the tandem accelerators is as high as 20 lA. The upgraded duoplasmatron has been integrated in a newly developed Stand-Alone Light Ion Injector (SALII), 3D CAD model depicted in Fig. 1. An Einzellens, a 30° analyzing magnet and an electrostatic Y-steerer allow ion optical matching with the beam line and/or the tandem accelerator that follows. This injector can be integrated with an existing HVE Multi-Purpose Injectors (MPI), at the 0° port of the MPI Magnet. The integration of a SALII with the MPI does not compromise the serviceability of MPI. One of the SALII injectors has been integrated in the accelerator facility from the Nuclear Physics Institute, Academy of Sciences, Rez, Czech Republic. The end-user indicates a beam brightness increase of 2–3 times when compared to ion beams obtained from the duoplasmatron located in the MPI injector [17]. The duoplasmatron performance improvements were quantified on a 3 MV TandetronTM accelerator system. This system is equipped with a Na CEC and with a terminal voltage stabilizing system based on a high dispersion 90° – 1500 mm radius analyzing magnet. We present the results on: hydrogen ion beam brightness at 3 MeV, ion beam current (H&He) particle transmission as a function of terminal voltage, and the ion beam energy stability using the 7Li(p,n)7Be nuclear reaction. 3.2. Performance measurements on accelerator system equipped with upgraded duoplasmatron ion source For this experiment, the duoplasmatron ion source delivered 75 lA H and 10 lA He at the entrance of the TandetronTM accelerator. The ion beam particle transmission tests were done in the range of 150 kV and 3000 kV (5–100% of terminal voltage range). The stripper gas thickness has been optimized to obtain maximum particle transmission through the accelerator, for the given energy
Fig. 2. Hydrogen and Helium ion beam particle transmission through the Tandetron accelerator as a function of accelerator terminal voltage between 150 kV and 3 MV. The stripper gas thickness has been optimized for maximum particle transmission.
140
N.C. Podaru, D.J.W. Mous / Nuclear Instruments and Methods in Physics Research B 371 (2016) 137–141
Fig. 3. Ion beam brightness measured for 3 MeV hydrogen at full object slits opening (X&Y) 0.1 and 0.15 mm as a function of half angle beam divergence.
turns increased the beam plasma density. Fig. 3 also shows reduced beam brightness for a half angle divergence below 50 lrad. This effect is often observed in microprobes and is likely related to hollow beam formation direct after extraction. Further investigations could clarify the reason for the reduced beam brightness with reducing half angle divergence and reduced aperture size. Microprobe experiments require long-term (several hours) ion beam energy stability for high demagnification microprobe lenses, enabling a constant probe resolution at target position. The functional description of the terminal voltage stabilization loop used in HVE accelerator systems has been previously reported in detail for a high beam power (in excess of 3 kW) accelerator system and one accelerator system used for microprobe applications [14,18]. In this work we discuss the results on the terminal voltage stability of a 3 MV T-shape TandetronTM obtained when using a 1500 mm radius, 90° analyzing magnet with ME/q2 of 48 a.m.u. MeV and powered by high stability power supply (10 ppm/h or better). The terminal voltage stability was investigated using the 7Li(p, n)7Be nuclear reaction, threshold energy 1880 keV. The results are summarized in Fig. 4. 11 runs (each lasting 10 min) have been recorded while the accelerator was controlled in slit stabilization mode. A linear fit on the experimentally determined energy variation yields that the ion beam energy variation is 13 eV/h, implying that the accelerator terminal voltage variation in slit stabilization mode is 6.5 ppm/h. The beam spot size increase is estimated to 3.4 nm/h due to such a beam energy drift, when the system has the following characteristics: microprobe lens with working distance of 150 mm, demagnification of 50 and microprobe acceptance of 70 lrad. 4. Summary
Fig. 4. Hydrogen ion beam energy stability function of time.
and charge state. The convolution between the stripping efficiency and the particle transmission through the accelerator yield the experimental curves depicted in Fig. 2. The ion beams were stripped in Argon gas. The optimum stripper gas thickness to obtain 2 MeV H or He beams was determined: 1H1+ 0.94 lg/cm2 (80%), 4He1+ 0.24 lg/cm2 (35.5%), and 4He2+ 0.56 lg/cm2 (23%). The ion beam particle transmission value is given in between brackets. It can be seen that for creating He ion beams with high energy <2.8 MeV, ion beam stripping for charge state 1+ is more prolific, whereas for He beam energy >2.8 MeV, charge state 2+ is more prolific. At 150 kV (only 5% of the nominal terminal voltage) the He beam particle transmission is 25%. The 3 MeV H ion beam brightness of the modified duoplasmatron has been measured with a setup using two micrometric precision XY slits separated by 4.25 m. The ion beam current was recorded in an electrostatically suppressed Faraday cup with 10 pA reading accuracy. The results on hydrogen brightness for the upgraded duoplasmatron ion source are summarized in Fig. 3. At 70 lrad half angle divergence, the ion beam brightness is 6.2 A m2 rad2 eV1 for 100 100 lm2 object size and 9 A m2 rad2 eV1 for 150 150 lm2 object size. This result indicates a threefold improvement in normal beam brightness when compared to the previous version of the duoplasmatron ion source (brightness of 2 A m2 rad2 eV1). The improvement in beam brightness value is attributed to the increased arc current that in
We have presented and discussed the most common ion sources used to create ion beam for IBA applications. Some key characteristics of the ion sources have been highlighted, with a strong emphasis on ion beam current output and ion beam brightness. The ion sources have been split in two groups, for positive ion production and negative ion production (with and without charge exchange). The RF ion sources are presently the most common for single-ended particle accelerators. They offer high ion beam brightness, and have a reasonable long lifetime. When discussing negative ion sources, we have focused on multicusp type ion sources and duoplasmatron type ion sources. Multicusp ion sources are very well suited to provide high brightness ion beams for microprobe work, brightness values bridging the gap between the tandem accelerators and single-ended accelerators. We also described the performance of the upgraded HVE duoplasmatron ion source with Ta filament when coupled to a tandem accelerator well suited for microprobe IBA applications. References [1] P. Vavpeticˇ, K. Vogel-Mikuš, L. Jeromel, N. Ogrinc Potocˇnik, P. Pongrac, D. Drobne, Zˇ. Pipan Tkalec, S. Novak, M. Kos, Š. Koren, M. Regvar, P. Pelicon, Nucl. Instr. Meth. B 348 (2015) 147. [2] L. Beck, Nucl. Instr. Meth. B 332 (2014) 439. [3] A. Zucchiatti, A. Climent Font, P.C. Gutierrez Neira, A. Perea, P. Fernandez Esquivel, S. Rovira Llorens, J.L. Ruvalcaba Sil, A. Verde, Nucl. Instr. Meth. B 332 (2014) 160. [4] A. Mackova, P. Malinsky, H. Pupikova, P. Nekvindova, J. Cajzl, Z. Sofer, R.A. Wilhelm, A. Kolitsch, J. Oswald, Nucl. Instr. Meth. B 332 (2014) 74. [5] R. Szymanski, D.N. Jamieson, Nucl. Instr. Meth. B 130 (1997) 80. [6] F. Watt, J.A. van Kan, I. Rajita, A.A. Bettiol, T.F. Choo, M.B.H. Breese, T. Osipowicz, Nucl. Instr. Meth. B 210 (2003) 14–20. [7] Ph. Barberet, L. Daudin, N. Gordillo, S. Sorieul, M. Simon, H. Seznec, I. Idarrage, S. Incerti, A. Balana, Ph. Moretto, Nucl. Instr. Meth. B 269 (2011) 2163–2167. [8] A.S. Schlachter, A.I.P. Conf. Proc. 111 (1984) 300–332. [9] M. Bacal, G.W. Hamilton, Phys. Rev. Lett. 42 (1979) 1538–1540.
N.C. Podaru, D.J.W. Mous / Nuclear Instruments and Methods in Physics Research B 371 (2016) 137–141 [10] M. Bacal, G.W. Hamilton, A.M. Bruneteau, H.J. Doucet, J. Taillet, Rev. Sci. Instr. 50 (1979) 719–721. [11] R.L. York, R.R. Stevens, K.N. Leung, K.W. Ehlers, Rev. Sci. Instr. 55 (1984) 681. [12] M. Von Ardenne, Technik 11/65, Tabellen der Elektronenphysik, Ionenphysik und Übermikroscopie, VEB Deutscher Verlag der Wissenschaften Berlin, 1956. [13] G.W.W. Quax, A. Gottdang, D.J.W. Mous, Rev. Sci. Instr. 81 (2010) 02A701. [14] N.C. Podaru, A. Gottdang, FRENA Group, D.J.W. Mous, Nucl. Instr. Meth. B 273 (2012) 231–233.
141
[15] P. Pelicon, N.C. Podaru, P. Vavpeticˇ, L. Jeromel, N. Ogrinc Potocnik, S. Ondracˇka, A. Gottdang, D.J.W. Mous, Nucl. Instr. Meth. B 332 (2014) 229–233. [16] M. Moser, P. Reichart, W. Carli, C. Greubel, K. Peeper, P. Hartung, G. Dollinger, Nucl. Instr. Meth. B 273 (2012) 226–230. [17] Václav Vosecˇek (personal communication, January 10, 2015). [18] D.J.W. Mous, R.G. Haitsma, T. Butz, R.-H. Flagmeyer, D. Lehmann, J. Vogt, Nucl. Instr. Meth. B 130 (1997) 31.