- Email: [email protected]
Commissioning of a powerful electron gun for electron–ion crossed-beams experiments
Nuclear Instruments and Methods in Physics Research B 408 (2017) 317–322
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb
Commissioning of a powerful electron gun for electron–ion crossed-beams experiments Benjamin Ebinger a,⇑, Alexander Borovik Jr. a, Tobias Molkentin a, Alfred Müller b, Stefan Schippers a a b
I. Physikalisches Institut, Justus-Liebig-Universität Gießen, 35392 Giessen, Germany Institut für Atom- und Molekülphysik, Justus-Liebig-Universität Gießen, 35392 Giessen, Germany
a r t i c l e
i n f o
Article history: Received 29 November 2016 Received in revised form 19 March 2017 Accepted 26 March 2017 Available online 7 April 2017 Keywords: Electron gun Electron target Electron–ion collisions Electron-impact ionization Crossed-beams
a b s t r a c t The sensitivity of an electron–ion crossed-beams experiment is mainly determined by the densities of both beams in the interaction region. Aiming at the extension of the available range of accessible electron energies and densities, a new high-power electron gun has been developed and fabricated. It delivers a ribbon-shaped beam with currents of up to 1 A at variable energies reaching 3500 eV at the maximum. The design goals of the gun are being met by using a configuration of electrodes, which allows for a variety of operation modes. Here, we report on the results of the ongoing tests of the new electron gun. In particular, electron-impact ionization cross sections of He+ and Xe5+ ions were measured employing the animated crossed-beams technique. The results are in excellent agreement with literature data. Ó 2017 Elsevier B.V. All rights reserved.
1. Introduction In electron–ion collision studies, it is highly desirable to have an electron beam available with high intensity over a wide energy range as these two parameters define the experimental accessibility of the cross sections to be investigated. Moreover, a welldefined energy in the interacton region is desirable for the investigation of, e.g., resonant ionization processes with high resolution. An electron gun producing an electron beam between 10 and 1000 eV with currents of up to 450 mA at 1000 eV has been built and successfully operated at our institute for about three decades [1]. However, as techniques for the production of highly charged ions have improved and the focus of interest has shifted to more highly charged ions (see, e.g., [2,3]), access to higher collision energies became necessary. Thus, a project for the development of a new electron gun with wider accessible electron-energy range was started [4]. The envisaged capabilities of the new electron gun included the energy range between 10 and 3500 eV as well as the ability to deliver high electron currents at low energies. With these parameters a continuation of projects for the determination of electron-impact ionization cross sections of, e.g., xenon and tungsten ions and the extension to higher charge states than the formerly accessible Xe24+ [2] and W19+ [3] will be possible. The determination of these data is highly desirable for plasma⇑ Corresponding author. E-mail address: [email protected] (B. Ebinger). http://dx.doi.org/10.1016/j.nimb.2017.03.136 0168-583X/Ó 2017 Elsevier B.V. All rights reserved.
modeling purposes and the validation of theoretical calculations that have already been carried out for charge states not experimentally accessible yet (e.g., W25+ [5]). After thorough simulations [4] and the fabrication of the first prototype, further developments and improvements were carried out addressing, e.g., thermal loads, high-voltage operation, the cooling of the electron collector as well as prevention of the ion beam from being influenced by the fields of focusing electrodes outside the electron gun [6]. Meanwhile, the gun is integrated into the electron–ion crossed-beams setup in Giessen (see [3] and references therein). The commissioning of the gun is currently underway. This report aims at giving an overview of the experimentally observed characteristics and capabilities of the present version of the new electron gun and its current commissioning status. 2. Characteristics of the electron gun The present electron gun comprises ten different electrodes which are electrically insulated from each other and can be controlled and monitored independently (Fig. 1). Electrons are emitted by a rectangular cuboid-like tungsten dispenser cathode with a curved emitting surface (60 5 mm) mounted into the Pierce-type focusing electrode. Thus, the produced electron beam has the shape of a ribbon. The first electrodes (focusing electrode, controlling electrode 1 and 2) focus the electron beam into the interaction region where it has a height of approximately 1–2 mm. Subsequently, the beam is defocused (by the controlling electrodes 3 and 4 as well as
318
B. Ebinger et al. / Nuclear Instruments and Methods in Physics Research B 408 (2017) 317–322
Fig. 1. Geometrical sketch of the electron gun with its various electrodes as well as the equipotential lines and electron trajectories for operation in the high-energy mode with w = 0.25 as defined in Table 1.
the defocusing electrode) as it moves further towards the watercooled collector. No magnetic field is involved. The properties of the tungsten dispenser cathode set strict requirements for the vacuum conditions. Stable gun operation can only be assured at residual-gas pressures not exceeding the lower 10 9 mbar range. Considering substantial gas emission from the collector surfaces as a result of the bombardment by energetic electrons, the task to maintain the required pressure is challenging, despite the availability of high-speed pumping (around 2400 l/s for N2). For this reason, the electron beam can be decelerated immediately before entering the collector thus reducing the deposited power and the corresponding gas emission. This feature is very useful at high electron energies, where deposited powers may easily exceed 1 kW. Compared to the prototype version of this electron gun described in [4], some minor changes to the design have been realized. Most notably, a formerly existing interaction box has been replaced by two separate, electrically insulated electrodes named interaction region 1 and 2. This change allows for an independent monitoring of the loss of electron current before and after the interaction with the ion beam and, thus, a more accurate determination of the electron current that is available for the ionization process. However, when the electron gun is in operation, both electrodes are set to the same potential resulting in a uniform potential in the whole interaction region and, thus, no change in the calculated electron trajectories. Moreover, separate insulated plates at the entrance and the exit of the electron–ion interaction region, as seen from the perspective of the ion beam, were installed in order to shield the primary ion beam from the electrode potentials outside of the electron gun (see Fig. 2). The given configuration of electrodes allows for a variety of different operation modes, which can be classified into four groups (Table 1): High-energy modes, high-current modes, modes without a potential trap for positive ions and modes with the electron beam decelerated at the collector to reduce the deposited power and, correspondingly, the heat load on and gas emission from the collector surfaces. The high-energy modes are suitable for higher electron energies, where intense currents can be reached straightforwardly. To obtain sufficient currents at low energies, the high-current modes can be used. Here, the potential on the first controlling electrode (also termed extraction electrode) is increased resulting in a higher electron current drawn from the cathode. Before entering
the interaction region, electrons are decelerated thus achieving the desired electron energy. In both types of modes, the intense electron beam induces a potential trap in the interaction region, which may influence the energy resolution and trap positive ions of the residual gas leading to distortions of the measured absolute cross sections. This can be avoided by applying negative potentials on the controlling electrodes 2 and 3 with respect to the interaction region (modes without a potential trap). Finally, the power deposited on the collector surfaces can be reduced by applying a lower potential to the collector such that the strict vacuum conditions described above can be satisfied also at high electron energies. For a better understanding of the rather theoretical paragraph above and to supply specific values, a complete set of potentials for the different operation modes at an electron energy of 500 eV is given here. In order to achieve the desired electron energy of 500 eV, the cathode together with the focusing and the defocusing electrode is set to 500 V, while the interaction region 1 and 2 are set to the ground potential for all modes. When using the highenergy mode with w = 0.25, the controlling electrodes 1, 2, 3 and 4 are set to 125 V and the collector is on the ground potential. However, for the corresponding high-current mode with, e.g., x = 1.5, the potential on the controlling electrodes 1 and 4 is raised to 750 V thus achieving a raise of the extracted electron current from 65 mA to 160 mA. To eliminate the potential trap in the interaction region in this particular case, a value of y = 0.55 with the resulting potential on the controlling electrodes 2 and 3 of 275 V is needed. Finally, to reduce the power deposited on the collector, a value of, e.g., z = 0.5 results in an applied potential of 250 V and the decrease of power from 80 W to 40 W. For the investigations described here, the potentials are set symmetrically with respect to the interaction region, which means that the potentials on the controlling electrodes 1 and 4 as well as on the controlling electrodes 2 and 3 are the same. 3. Experimental technique The cross-section measurements are performed with the new electron gun integrated into the Giessen electron–ion crossedbeams setup. This apparatus has already been described in detail elsewhere (see, e.g., [3] and references therein) and, thus, will only be briefly described here. In the setup, ions are produced by
B. Ebinger et al. / Nuclear Instruments and Methods in Physics Research B 408 (2017) 317–322
319
Fig. 2. A picture of the electron gun with the cathode unmounted.
Table 1 Overview of the potential settings for different operation modes of the new electron gun. The electrodes are termed as in Fig. 1. Column 2 provides the ratios of potentials on the cathode/focusing electrode, controlling electrode 1, controlling electrode 2, interaction region, controlling electrode 3, controlling electrode 4, defocusing electrode and collector, respectively. Shown are the potentials on the corresponding electrodes relative to the potential difference between the cathode/focusing electrode and interaction region, i.e. the voltage which determines the laboratory energy of the electrons where they are colliding with the ions. The variables can have the following values: w between 0.05 and 0.25; x up to 7 at low energies, less at higher energies; y between 0.55 and 0; z down to about 0.8. Operation mode High-energy High-current Without potential trap Reduced power on collector
Potentials on electrodes 1:w:w:0:w:w: 1:0 1:x:w:0:w:x: 1:0 1:x:y:0:y:x: 1:0 1:x:w:0:w:x: 1:z
Potential trap Yes Yes No Yes
a 10-GHz electron-cyclotron-resonance ion source. Extracted ions are mass-over-charge analyzed while passing a dipole magnet. This assures charge-state and isotope-pure beams in the experiment. After entering the UHV section of the setup, the ion beam is once again charge-state purified by a 90° spherical deflector. Two pairs of four-jaw slits collimate the ion beam to sizes typically below 1 mm2 before it enters the interaction region. Following the exposure of the ion beam to the ribbon-shaped electron beam, the ionization products are separated from the primary ion beam in the field of a second dipole magnet. The product ions are registered by a single-particle detector and the primary ions are collected in a Faraday cup. Absolute cross sections for electron-impact ionization of the corresponding ions are recorded by employing the animated crossed-beams technique [7]. This method determines the spatial overlap of the beams by mechanically moving the electron through the ion beam. One well-known difficulty upon utilization of this method originates from a potential trap in the interaction region caused by the space-charge of the highly intense electron beam [8]. This results
in the accumulation of slow residual gas ions trapped in the potential well, which can influence the apparent electron–ion collision cross sections as ion-ion collisions occur. With our former electron gun, this issue was resolved by introducing krypton gas (which as a noble gas does not hurt the cathode) up to a high pressure in the 10 7 mbar range to fill the potential trap artificially. Contributions from related ion-ion collisions were eliminated from the measured absolute cross section later on in the data analysis where necessary [8]. The here presented new type of electron gun additionally features the possibility to open the potential trap by an appropriate choice of electrode potentials (Table 1) and thereby directly preventing residual gas ions from accumulating in the electron beam. Two major issues are involved in the successful operation of the new electron gun: First, steady operation of the electron gun itself has to be assured, which includes the stable extraction of the electrons and reliable transmission through the interaction region and onto the collector without significant electron beam losses for all desired operational modes. This is checked by constantly monitoring the electron currents on all electrodes. Second, the experimental parameters of the electron beam must not influence the transmission of the crossing ion beam and its reliable detection. Concerning this point, care has to be taken in properly aligning both beams and in setting the electrode potentials such that field distortions at the position of the ion beam are minimized. In the present experiments this is monitored and assured by observing the product particle beam on a 2D position-sensitive detector. For the first cross-section measurements with the new electron gun, He+ and Xe5+ ions have been chosen as target ions. Both ion species are produced by feeding the corresponding gas into the ECR plasma and extracted with an extraction voltage of 12 kV. Thus, the ion energies are 12 keV for He+ and 60 keV for Xe5+ ions. The ion currents available for the experiments are mainly determined by the settings of the fine collimating four-jaw slits. For He+ ions, 3 nA are obtained with 0.6 0.6 mm2 slits. For Xe5+ the slit size is reduced to 0.2 0.2 mm2 with 150 pA (up to 1600 eV electron energy) and 0.1 0.1 mm2 with 65 pA (above 1600 eV
320
B. Ebinger et al. / Nuclear Instruments and Methods in Physics Research B 408 (2017) 317–322
electron energy), respectively. This reduction of the ion current is necessary because the combination of a large cross section for electron-impact ionization of Xe5+ ions and high electron currents, especially at the higher energies, leads to count rates of several thousand events per second and, thus, dead-time effects on the single-particle detector. For the studied single-ionization process, signal-to-background ratios are 4.0 (for He+) and 101 (for Xe5+) at the peak of the cross section and, due to a more intense electron beam, better at higher energies (19.4 for He+ at 1000 eV and 338 for Xe5+ at 2500 eV). This superior value for Xe5+ compared to He+ is a consequence of the krypton gas introduced into the interaction chamber in the case of He+ leading to an increase of the background signal. For all present measurements, statistical uncertainties at energies well above the ioniza-
tion onset are clearly below 2%. The total error budget of the presently employed setup with the new electron gun is consistent with the one discussed in detail previously [9] except for the error of the electron current which is explained in more detail below. 4. Experimental results Fig. 3 shows the measured electron current emitted from the cathode as a function of electron energy for the different operation modes described in Table 1. The high-energy mode provides decent electron currents at energies above 500 eV, while high currents at lower energies can be reached by employing one of the highcurrent modes. With these, however, operation can only be achieved in a limited energy range. Fig. 3 also shows the
Fig. 3. The electron current emitted from the cathode as a function of electron energy for different operation modes described in Table 1.
Fig. 4. The fraction of electron current transmitted into the interaction region as a function of electron energy for different operation modes described in Table 1.
B. Ebinger et al. / Nuclear Instruments and Methods in Physics Research B 408 (2017) 317–322
321
Fig. 5. Measured cross section for electron-impact ionization of He+ ions compared with experimental literature data [11]. The electron–ion collision energies are given in the center-of-mass frame.
Fig. 6. Measured cross section for electron-impact ionization of Xe5+ ions compared with experimental literature data [12,13]. The electron–ion collision energies are given in the center-of-mass frame.
current-versus-energy behavior for the previous electron gun [1]. As can be seen, the present gun both extends the available electron-energy range and provides higher currents at low electron energies. The ratios of the electron current transmitted into the interaction region relative to the total current emitted from the cathode can be seen in Fig. 4. Apparently, only a negligible portion of electrons is lost if a suitable operation mode is chosen. The accuracy of the measured electron currents is mainly limited by the possibility of secondary-electron emission from the electrodes due to the bombardment of the surfaces with the lost electrons. This can lead to the emission of a large number of electrons thus
resulting in a distortion of the measured currents of electrons lost at a specific electrode. Based on our experimental data, we do not believe this process to influence our determination of the electron current by more than 2% at the maximum, which we, therefore, set as the error of the electron current for all energies. We further verify the appropriate operation of the electron gun by comparative measurements of known absolute cross sections. One of the candidates for this purpose is the electron-impact single-ionization cross section of He+ ions. This cross section is well known from the literature [10,11] and has also been measured many times using the present setup with the previous electron
322
B. Ebinger et al. / Nuclear Instruments and Methods in Physics Research B 408 (2017) 317–322
gun [1] (unpublished data). All these data agree within their respective error bars. In Fig. 5 we, therefore, compare the present results only with the ones of [11]. The agreement between the data sets is well within the estimated uncertainties over the entire energy range. These measurements are carried out employing the high-energy mode with w = 0.05 (Table 1). Krypton gas is introduced into the interaction chamber up to a pressure of 3 10 7 mbar to eliminate effects of electric fields in the interaction region by compensating the electron space charge by the slow krypton ions produced in the electron beam [8]. However, very similar values can be reached without any introduction of krypton gas if the background pressure is sufficiently low (see below). So far, electron-impact ionization of He+ ions is studied only for energies of up to 1000 eV. The extension of these measurements to higher energies is ongoing. Furthermore, the electron-impact single-ionization cross section of Xe5+ ions is measured. The results are displayed in Fig. 6 together with previous experimental data of Griffin et al. [12] and Borovik Jr. [13], the latter of which were obtained at our setup with the former electron gun. The agreement between the data sets is found to be excellent at electron energies up to 500 eV. The small discrepancies around the peak of the cross section can be explained by strong variations of the cross section due to indirect ionization processes [13] and possible deviations of the particular electron energies of just a couple of eV. At higher energies, the present data and those of Borovik Jr. are in a good agreement, while the data of Griffin et al. are up to 25 % higher. This discrepancy may result from the fact that different fractions of ions in long-lived excited states can be admixed to the primary ion beams of the two experiments as discussed previously in [13]. All in all, excellent consistency with the older data carried out at our setup with the former electron gun is reached. It shall be noted that these data are taken using the high-energy mode with w = 0.05 and no addition of any krypton gas. During the operation of the high-power electron gun the background pressure reaches the low 10 9 mbar range at the highest electron energies and stays in the low to mid 10 10 mbar range for all measurements up to 2000 eV. Apparently, space charge compensation is not required for reliable cross-section measurements with the new electron gun in some cases, i.e. if the background pressure is low enough, as the same can be seen for some measurements of the He+ ions. However, comparative measurements with the introduction of krypton gas show a small influence on the measured cross section for the Xe5+ ions, albeit within the experimental error bars. 5. Summary and outlook We reported on the present status of the commissioning of the new high-power electron gun at the electron–ion crossed-beams apparatus in Giessen. Performance tests show that the electron gun is working in accordance with predictions. It delivers the predicted electron currents at a high transmission. Both aims of extending the available electron-energy range and of achieving
high electron currents at low electron energies are reached. Comparative test measurements of electron-impact ionization cross sections of He+ and Xe5+ ions were performed. A very good agreement with the previously published data is found. Thus, the present electron gun is a suitable tool for extending the range of ions for ionization cross-section measurements. Additionally, experiences gained with the present electron gun will be employed in an ongoing project aiming at building a transversal electron target for installation at the heavy-ion storage ring CRYRING at the GSI Helmholtz Center for Heavy Ion Research in Darmstadt [14]. Acknowledgements The authors thank D. Sieben and B. M. Döhring for assistance with some of the measurements. Financial support by the German Federal Ministry of Education and Research (BMBF) within the ”V erbundforschung” funding scheme (contract No. 05P15RGFAA) and by the GSI Helmholtz Center for Heavy Ion Research (Darmstadt, Germany) is gratefully acknowledged. References [1] R. Becker, A. Müller, C. Achenbach, K. Tinschert, E. Salzborn, Nucl. Instrum. Methods Phys. Res., Sect. B 9 (1985) 385. [2] P. Liu, J. Zeng, A. Borovik Jr, S. Schippers, A. Müller, Phys. Rev. A 92 (2015) 012701. [3] A. Borovik Jr, B. Ebinger, D. Schury, S. Schippers, A. Müller, Phys. Rev. A 93 (2016) 012708. [4] W. Shi, J. Jacobi, H. Knopp, S. Schippers, A. Müller, Nucl. Instrum. Methods Phys. Res., Sect. B 205 (2003) 201. [5] A. Kyniene˙, S. Pakalka, Š. Masys, V. Jonauskas, J. Phys. B: At., Mol. Opt. Phys. 49 (2016) 185001. [6] A. Borovik Jr, W. Shi, J. Jacobi, S. Schippers, A. Müller, J. Phys.: Conf. Ser. 488 (2014) 142007. [7] A. Müller, K. Huber, K. Tinschert, R. Becker, E. Salzborn, J. Phys. B: At. Mol. Phys. 18 (1985) 2993. [8] A. Müller, G. Hofmann, K. Tinschert, R. Sauer, E. Salzborn, R. Becker, Nucl. Instrum. Methods Phys. Res., Sect. B 24 (1987) 369. [9] J. Rausch, A. Becker, K. Spruck, J. Hellhund, A. Borovik Jr, K. Huber, S. Schippers, A. Müller, J. Phys. B: At., Mol. Opt. Phys. 44 (2011) 165202. [10] K.T. Dolder, M.F.A. Harrison, P.C. Thonemann, Proc. R. Soc. A 264 (1961) 367. [11] B. Peart, D.S. Walton, K.T. Dolder, J. Phys. B: At. Mol. Phys. 2 (1969) 1347. [12] D.C. Griffin, C. Bottcher, M.S. Pindzola, S.M. Younger, D.C. Gregory, D.H. Crandall, Phys. Rev. A 29 (1984) 1729. [13] A. Borovik Jr., Electron-Impact Ionization of Xenon and Tin Ions, Justus-LiebigUniversity Giessen, 2010, Dissertation. [14] M. Lestinsky, V. Andrianov, B. Aurand, V. Bagnoud, D. Bernhardt, H. Beyer, S. Bishop, K. Blaum, A. Bleile, A. Borovik, F. Bosch, C. Bostock, C. Brandau, A. Bräuning-Demian, I. Bray, T. Davinson, B. Ebinger, A. Echler, P. Egelhof, A. Ehresmann, M. Engström, C. Enss, N. Ferreira, D. Fischer, A. Fleischmann, E. Förster, S. Fritzsche, R. Geithner, S. Geyer, J. Glorius, K. Göbel, O. Gorda, J. Goullon, P. Grabitz, R. Grisenti, A. Gumberidze, S. Hagmann, M. Heil, A. Heinz, F. Herfurth, R. He, P.-M. Hillenbrand, R. Hubele, P. Indelicato, A. Källberg, O. Kester, O. Kiselev, A. Knie, C. Kozhuharov, S. Kraft-Bermuth, T. Kühl, G. Lane, Y. Litvinov, D. Liesen, X. Ma, R. Märtin, R. Moshammer, A. Müller, S. Namba, P. Neumeyer, T. Nilsson, W. Nörtershäuser, G. Paulus, N. Petridis, M. Reed, R. Reifarth, P. Rei, J. Rothhardt, R. Sanchez, M. Sanjari, S. Schippers, H. Schmidt, D. Schneider, P. Scholz, R. Schuch, M. Schulz, V. Shabaev, A. Simonsson, J. Sjöholm, Ö. Skeppstedt, K. Sonnabend, U. Spillmann, K. Stiebing, M. Steck, T. Stöhlker, A. Surzhykov, S. Torilov, E. Träbert, M. Trassinelli, S. Trotsenko, X. Tu, I. Uschmann, P. Walker, G. Weber, D. Winters, P. Woods, H. Zhao, Y. Zhang, Eur. Phys. J. Special Topics 225 (2016) 797.
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb
Commissioning of a powerful electron gun for electron–ion crossed-beams experiments Benjamin Ebinger a,⇑, Alexander Borovik Jr. a, Tobias Molkentin a, Alfred Müller b, Stefan Schippers a a b
I. Physikalisches Institut, Justus-Liebig-Universität Gießen, 35392 Giessen, Germany Institut für Atom- und Molekülphysik, Justus-Liebig-Universität Gießen, 35392 Giessen, Germany
a r t i c l e
i n f o
Article history: Received 29 November 2016 Received in revised form 19 March 2017 Accepted 26 March 2017 Available online 7 April 2017 Keywords: Electron gun Electron target Electron–ion collisions Electron-impact ionization Crossed-beams
a b s t r a c t The sensitivity of an electron–ion crossed-beams experiment is mainly determined by the densities of both beams in the interaction region. Aiming at the extension of the available range of accessible electron energies and densities, a new high-power electron gun has been developed and fabricated. It delivers a ribbon-shaped beam with currents of up to 1 A at variable energies reaching 3500 eV at the maximum. The design goals of the gun are being met by using a configuration of electrodes, which allows for a variety of operation modes. Here, we report on the results of the ongoing tests of the new electron gun. In particular, electron-impact ionization cross sections of He+ and Xe5+ ions were measured employing the animated crossed-beams technique. The results are in excellent agreement with literature data. Ó 2017 Elsevier B.V. All rights reserved.
1. Introduction In electron–ion collision studies, it is highly desirable to have an electron beam available with high intensity over a wide energy range as these two parameters define the experimental accessibility of the cross sections to be investigated. Moreover, a welldefined energy in the interacton region is desirable for the investigation of, e.g., resonant ionization processes with high resolution. An electron gun producing an electron beam between 10 and 1000 eV with currents of up to 450 mA at 1000 eV has been built and successfully operated at our institute for about three decades [1]. However, as techniques for the production of highly charged ions have improved and the focus of interest has shifted to more highly charged ions (see, e.g., [2,3]), access to higher collision energies became necessary. Thus, a project for the development of a new electron gun with wider accessible electron-energy range was started [4]. The envisaged capabilities of the new electron gun included the energy range between 10 and 3500 eV as well as the ability to deliver high electron currents at low energies. With these parameters a continuation of projects for the determination of electron-impact ionization cross sections of, e.g., xenon and tungsten ions and the extension to higher charge states than the formerly accessible Xe24+ [2] and W19+ [3] will be possible. The determination of these data is highly desirable for plasma⇑ Corresponding author. E-mail address: [email protected] (B. Ebinger). http://dx.doi.org/10.1016/j.nimb.2017.03.136 0168-583X/Ó 2017 Elsevier B.V. All rights reserved.
modeling purposes and the validation of theoretical calculations that have already been carried out for charge states not experimentally accessible yet (e.g., W25+ [5]). After thorough simulations [4] and the fabrication of the first prototype, further developments and improvements were carried out addressing, e.g., thermal loads, high-voltage operation, the cooling of the electron collector as well as prevention of the ion beam from being influenced by the fields of focusing electrodes outside the electron gun [6]. Meanwhile, the gun is integrated into the electron–ion crossed-beams setup in Giessen (see [3] and references therein). The commissioning of the gun is currently underway. This report aims at giving an overview of the experimentally observed characteristics and capabilities of the present version of the new electron gun and its current commissioning status. 2. Characteristics of the electron gun The present electron gun comprises ten different electrodes which are electrically insulated from each other and can be controlled and monitored independently (Fig. 1). Electrons are emitted by a rectangular cuboid-like tungsten dispenser cathode with a curved emitting surface (60 5 mm) mounted into the Pierce-type focusing electrode. Thus, the produced electron beam has the shape of a ribbon. The first electrodes (focusing electrode, controlling electrode 1 and 2) focus the electron beam into the interaction region where it has a height of approximately 1–2 mm. Subsequently, the beam is defocused (by the controlling electrodes 3 and 4 as well as
318
B. Ebinger et al. / Nuclear Instruments and Methods in Physics Research B 408 (2017) 317–322
Fig. 1. Geometrical sketch of the electron gun with its various electrodes as well as the equipotential lines and electron trajectories for operation in the high-energy mode with w = 0.25 as defined in Table 1.
the defocusing electrode) as it moves further towards the watercooled collector. No magnetic field is involved. The properties of the tungsten dispenser cathode set strict requirements for the vacuum conditions. Stable gun operation can only be assured at residual-gas pressures not exceeding the lower 10 9 mbar range. Considering substantial gas emission from the collector surfaces as a result of the bombardment by energetic electrons, the task to maintain the required pressure is challenging, despite the availability of high-speed pumping (around 2400 l/s for N2). For this reason, the electron beam can be decelerated immediately before entering the collector thus reducing the deposited power and the corresponding gas emission. This feature is very useful at high electron energies, where deposited powers may easily exceed 1 kW. Compared to the prototype version of this electron gun described in [4], some minor changes to the design have been realized. Most notably, a formerly existing interaction box has been replaced by two separate, electrically insulated electrodes named interaction region 1 and 2. This change allows for an independent monitoring of the loss of electron current before and after the interaction with the ion beam and, thus, a more accurate determination of the electron current that is available for the ionization process. However, when the electron gun is in operation, both electrodes are set to the same potential resulting in a uniform potential in the whole interaction region and, thus, no change in the calculated electron trajectories. Moreover, separate insulated plates at the entrance and the exit of the electron–ion interaction region, as seen from the perspective of the ion beam, were installed in order to shield the primary ion beam from the electrode potentials outside of the electron gun (see Fig. 2). The given configuration of electrodes allows for a variety of different operation modes, which can be classified into four groups (Table 1): High-energy modes, high-current modes, modes without a potential trap for positive ions and modes with the electron beam decelerated at the collector to reduce the deposited power and, correspondingly, the heat load on and gas emission from the collector surfaces. The high-energy modes are suitable for higher electron energies, where intense currents can be reached straightforwardly. To obtain sufficient currents at low energies, the high-current modes can be used. Here, the potential on the first controlling electrode (also termed extraction electrode) is increased resulting in a higher electron current drawn from the cathode. Before entering
the interaction region, electrons are decelerated thus achieving the desired electron energy. In both types of modes, the intense electron beam induces a potential trap in the interaction region, which may influence the energy resolution and trap positive ions of the residual gas leading to distortions of the measured absolute cross sections. This can be avoided by applying negative potentials on the controlling electrodes 2 and 3 with respect to the interaction region (modes without a potential trap). Finally, the power deposited on the collector surfaces can be reduced by applying a lower potential to the collector such that the strict vacuum conditions described above can be satisfied also at high electron energies. For a better understanding of the rather theoretical paragraph above and to supply specific values, a complete set of potentials for the different operation modes at an electron energy of 500 eV is given here. In order to achieve the desired electron energy of 500 eV, the cathode together with the focusing and the defocusing electrode is set to 500 V, while the interaction region 1 and 2 are set to the ground potential for all modes. When using the highenergy mode with w = 0.25, the controlling electrodes 1, 2, 3 and 4 are set to 125 V and the collector is on the ground potential. However, for the corresponding high-current mode with, e.g., x = 1.5, the potential on the controlling electrodes 1 and 4 is raised to 750 V thus achieving a raise of the extracted electron current from 65 mA to 160 mA. To eliminate the potential trap in the interaction region in this particular case, a value of y = 0.55 with the resulting potential on the controlling electrodes 2 and 3 of 275 V is needed. Finally, to reduce the power deposited on the collector, a value of, e.g., z = 0.5 results in an applied potential of 250 V and the decrease of power from 80 W to 40 W. For the investigations described here, the potentials are set symmetrically with respect to the interaction region, which means that the potentials on the controlling electrodes 1 and 4 as well as on the controlling electrodes 2 and 3 are the same. 3. Experimental technique The cross-section measurements are performed with the new electron gun integrated into the Giessen electron–ion crossedbeams setup. This apparatus has already been described in detail elsewhere (see, e.g., [3] and references therein) and, thus, will only be briefly described here. In the setup, ions are produced by
B. Ebinger et al. / Nuclear Instruments and Methods in Physics Research B 408 (2017) 317–322
319
Fig. 2. A picture of the electron gun with the cathode unmounted.
Table 1 Overview of the potential settings for different operation modes of the new electron gun. The electrodes are termed as in Fig. 1. Column 2 provides the ratios of potentials on the cathode/focusing electrode, controlling electrode 1, controlling electrode 2, interaction region, controlling electrode 3, controlling electrode 4, defocusing electrode and collector, respectively. Shown are the potentials on the corresponding electrodes relative to the potential difference between the cathode/focusing electrode and interaction region, i.e. the voltage which determines the laboratory energy of the electrons where they are colliding with the ions. The variables can have the following values: w between 0.05 and 0.25; x up to 7 at low energies, less at higher energies; y between 0.55 and 0; z down to about 0.8. Operation mode High-energy High-current Without potential trap Reduced power on collector
Potentials on electrodes 1:w:w:0:w:w: 1:0 1:x:w:0:w:x: 1:0 1:x:y:0:y:x: 1:0 1:x:w:0:w:x: 1:z
Potential trap Yes Yes No Yes
a 10-GHz electron-cyclotron-resonance ion source. Extracted ions are mass-over-charge analyzed while passing a dipole magnet. This assures charge-state and isotope-pure beams in the experiment. After entering the UHV section of the setup, the ion beam is once again charge-state purified by a 90° spherical deflector. Two pairs of four-jaw slits collimate the ion beam to sizes typically below 1 mm2 before it enters the interaction region. Following the exposure of the ion beam to the ribbon-shaped electron beam, the ionization products are separated from the primary ion beam in the field of a second dipole magnet. The product ions are registered by a single-particle detector and the primary ions are collected in a Faraday cup. Absolute cross sections for electron-impact ionization of the corresponding ions are recorded by employing the animated crossed-beams technique [7]. This method determines the spatial overlap of the beams by mechanically moving the electron through the ion beam. One well-known difficulty upon utilization of this method originates from a potential trap in the interaction region caused by the space-charge of the highly intense electron beam [8]. This results
in the accumulation of slow residual gas ions trapped in the potential well, which can influence the apparent electron–ion collision cross sections as ion-ion collisions occur. With our former electron gun, this issue was resolved by introducing krypton gas (which as a noble gas does not hurt the cathode) up to a high pressure in the 10 7 mbar range to fill the potential trap artificially. Contributions from related ion-ion collisions were eliminated from the measured absolute cross section later on in the data analysis where necessary [8]. The here presented new type of electron gun additionally features the possibility to open the potential trap by an appropriate choice of electrode potentials (Table 1) and thereby directly preventing residual gas ions from accumulating in the electron beam. Two major issues are involved in the successful operation of the new electron gun: First, steady operation of the electron gun itself has to be assured, which includes the stable extraction of the electrons and reliable transmission through the interaction region and onto the collector without significant electron beam losses for all desired operational modes. This is checked by constantly monitoring the electron currents on all electrodes. Second, the experimental parameters of the electron beam must not influence the transmission of the crossing ion beam and its reliable detection. Concerning this point, care has to be taken in properly aligning both beams and in setting the electrode potentials such that field distortions at the position of the ion beam are minimized. In the present experiments this is monitored and assured by observing the product particle beam on a 2D position-sensitive detector. For the first cross-section measurements with the new electron gun, He+ and Xe5+ ions have been chosen as target ions. Both ion species are produced by feeding the corresponding gas into the ECR plasma and extracted with an extraction voltage of 12 kV. Thus, the ion energies are 12 keV for He+ and 60 keV for Xe5+ ions. The ion currents available for the experiments are mainly determined by the settings of the fine collimating four-jaw slits. For He+ ions, 3 nA are obtained with 0.6 0.6 mm2 slits. For Xe5+ the slit size is reduced to 0.2 0.2 mm2 with 150 pA (up to 1600 eV electron energy) and 0.1 0.1 mm2 with 65 pA (above 1600 eV
320
B. Ebinger et al. / Nuclear Instruments and Methods in Physics Research B 408 (2017) 317–322
electron energy), respectively. This reduction of the ion current is necessary because the combination of a large cross section for electron-impact ionization of Xe5+ ions and high electron currents, especially at the higher energies, leads to count rates of several thousand events per second and, thus, dead-time effects on the single-particle detector. For the studied single-ionization process, signal-to-background ratios are 4.0 (for He+) and 101 (for Xe5+) at the peak of the cross section and, due to a more intense electron beam, better at higher energies (19.4 for He+ at 1000 eV and 338 for Xe5+ at 2500 eV). This superior value for Xe5+ compared to He+ is a consequence of the krypton gas introduced into the interaction chamber in the case of He+ leading to an increase of the background signal. For all present measurements, statistical uncertainties at energies well above the ioniza-
tion onset are clearly below 2%. The total error budget of the presently employed setup with the new electron gun is consistent with the one discussed in detail previously [9] except for the error of the electron current which is explained in more detail below. 4. Experimental results Fig. 3 shows the measured electron current emitted from the cathode as a function of electron energy for the different operation modes described in Table 1. The high-energy mode provides decent electron currents at energies above 500 eV, while high currents at lower energies can be reached by employing one of the highcurrent modes. With these, however, operation can only be achieved in a limited energy range. Fig. 3 also shows the
Fig. 3. The electron current emitted from the cathode as a function of electron energy for different operation modes described in Table 1.
Fig. 4. The fraction of electron current transmitted into the interaction region as a function of electron energy for different operation modes described in Table 1.
B. Ebinger et al. / Nuclear Instruments and Methods in Physics Research B 408 (2017) 317–322
321
Fig. 5. Measured cross section for electron-impact ionization of He+ ions compared with experimental literature data [11]. The electron–ion collision energies are given in the center-of-mass frame.
Fig. 6. Measured cross section for electron-impact ionization of Xe5+ ions compared with experimental literature data [12,13]. The electron–ion collision energies are given in the center-of-mass frame.
current-versus-energy behavior for the previous electron gun [1]. As can be seen, the present gun both extends the available electron-energy range and provides higher currents at low electron energies. The ratios of the electron current transmitted into the interaction region relative to the total current emitted from the cathode can be seen in Fig. 4. Apparently, only a negligible portion of electrons is lost if a suitable operation mode is chosen. The accuracy of the measured electron currents is mainly limited by the possibility of secondary-electron emission from the electrodes due to the bombardment of the surfaces with the lost electrons. This can lead to the emission of a large number of electrons thus
resulting in a distortion of the measured currents of electrons lost at a specific electrode. Based on our experimental data, we do not believe this process to influence our determination of the electron current by more than 2% at the maximum, which we, therefore, set as the error of the electron current for all energies. We further verify the appropriate operation of the electron gun by comparative measurements of known absolute cross sections. One of the candidates for this purpose is the electron-impact single-ionization cross section of He+ ions. This cross section is well known from the literature [10,11] and has also been measured many times using the present setup with the previous electron
322
B. Ebinger et al. / Nuclear Instruments and Methods in Physics Research B 408 (2017) 317–322
gun [1] (unpublished data). All these data agree within their respective error bars. In Fig. 5 we, therefore, compare the present results only with the ones of [11]. The agreement between the data sets is well within the estimated uncertainties over the entire energy range. These measurements are carried out employing the high-energy mode with w = 0.05 (Table 1). Krypton gas is introduced into the interaction chamber up to a pressure of 3 10 7 mbar to eliminate effects of electric fields in the interaction region by compensating the electron space charge by the slow krypton ions produced in the electron beam [8]. However, very similar values can be reached without any introduction of krypton gas if the background pressure is sufficiently low (see below). So far, electron-impact ionization of He+ ions is studied only for energies of up to 1000 eV. The extension of these measurements to higher energies is ongoing. Furthermore, the electron-impact single-ionization cross section of Xe5+ ions is measured. The results are displayed in Fig. 6 together with previous experimental data of Griffin et al. [12] and Borovik Jr. [13], the latter of which were obtained at our setup with the former electron gun. The agreement between the data sets is found to be excellent at electron energies up to 500 eV. The small discrepancies around the peak of the cross section can be explained by strong variations of the cross section due to indirect ionization processes [13] and possible deviations of the particular electron energies of just a couple of eV. At higher energies, the present data and those of Borovik Jr. are in a good agreement, while the data of Griffin et al. are up to 25 % higher. This discrepancy may result from the fact that different fractions of ions in long-lived excited states can be admixed to the primary ion beams of the two experiments as discussed previously in [13]. All in all, excellent consistency with the older data carried out at our setup with the former electron gun is reached. It shall be noted that these data are taken using the high-energy mode with w = 0.05 and no addition of any krypton gas. During the operation of the high-power electron gun the background pressure reaches the low 10 9 mbar range at the highest electron energies and stays in the low to mid 10 10 mbar range for all measurements up to 2000 eV. Apparently, space charge compensation is not required for reliable cross-section measurements with the new electron gun in some cases, i.e. if the background pressure is low enough, as the same can be seen for some measurements of the He+ ions. However, comparative measurements with the introduction of krypton gas show a small influence on the measured cross section for the Xe5+ ions, albeit within the experimental error bars. 5. Summary and outlook We reported on the present status of the commissioning of the new high-power electron gun at the electron–ion crossed-beams apparatus in Giessen. Performance tests show that the electron gun is working in accordance with predictions. It delivers the predicted electron currents at a high transmission. Both aims of extending the available electron-energy range and of achieving
high electron currents at low electron energies are reached. Comparative test measurements of electron-impact ionization cross sections of He+ and Xe5+ ions were performed. A very good agreement with the previously published data is found. Thus, the present electron gun is a suitable tool for extending the range of ions for ionization cross-section measurements. Additionally, experiences gained with the present electron gun will be employed in an ongoing project aiming at building a transversal electron target for installation at the heavy-ion storage ring CRYRING at the GSI Helmholtz Center for Heavy Ion Research in Darmstadt [14]. Acknowledgements The authors thank D. Sieben and B. M. Döhring for assistance with some of the measurements. Financial support by the German Federal Ministry of Education and Research (BMBF) within the ”V erbundforschung” funding scheme (contract No. 05P15RGFAA) and by the GSI Helmholtz Center for Heavy Ion Research (Darmstadt, Germany) is gratefully acknowledged. References [1] R. Becker, A. Müller, C. Achenbach, K. Tinschert, E. Salzborn, Nucl. Instrum. Methods Phys. Res., Sect. B 9 (1985) 385. [2] P. Liu, J. Zeng, A. Borovik Jr, S. Schippers, A. Müller, Phys. Rev. A 92 (2015) 012701. [3] A. Borovik Jr, B. Ebinger, D. Schury, S. Schippers, A. Müller, Phys. Rev. A 93 (2016) 012708. [4] W. Shi, J. Jacobi, H. Knopp, S. Schippers, A. Müller, Nucl. Instrum. Methods Phys. Res., Sect. B 205 (2003) 201. [5] A. Kyniene˙, S. Pakalka, Š. Masys, V. Jonauskas, J. Phys. B: At., Mol. Opt. Phys. 49 (2016) 185001. [6] A. Borovik Jr, W. Shi, J. Jacobi, S. Schippers, A. Müller, J. Phys.: Conf. Ser. 488 (2014) 142007. [7] A. Müller, K. Huber, K. Tinschert, R. Becker, E. Salzborn, J. Phys. B: At. Mol. Phys. 18 (1985) 2993. [8] A. Müller, G. Hofmann, K. Tinschert, R. Sauer, E. Salzborn, R. Becker, Nucl. Instrum. Methods Phys. Res., Sect. B 24 (1987) 369. [9] J. Rausch, A. Becker, K. Spruck, J. Hellhund, A. Borovik Jr, K. Huber, S. Schippers, A. Müller, J. Phys. B: At., Mol. Opt. Phys. 44 (2011) 165202. [10] K.T. Dolder, M.F.A. Harrison, P.C. Thonemann, Proc. R. Soc. A 264 (1961) 367. [11] B. Peart, D.S. Walton, K.T. Dolder, J. Phys. B: At. Mol. Phys. 2 (1969) 1347. [12] D.C. Griffin, C. Bottcher, M.S. Pindzola, S.M. Younger, D.C. Gregory, D.H. Crandall, Phys. Rev. A 29 (1984) 1729. [13] A. Borovik Jr., Electron-Impact Ionization of Xenon and Tin Ions, Justus-LiebigUniversity Giessen, 2010, Dissertation. [14] M. Lestinsky, V. Andrianov, B. Aurand, V. Bagnoud, D. Bernhardt, H. Beyer, S. Bishop, K. Blaum, A. Bleile, A. Borovik, F. Bosch, C. Bostock, C. Brandau, A. Bräuning-Demian, I. Bray, T. Davinson, B. Ebinger, A. Echler, P. Egelhof, A. Ehresmann, M. Engström, C. Enss, N. Ferreira, D. Fischer, A. Fleischmann, E. Förster, S. Fritzsche, R. Geithner, S. Geyer, J. Glorius, K. Göbel, O. Gorda, J. Goullon, P. Grabitz, R. Grisenti, A. Gumberidze, S. Hagmann, M. Heil, A. Heinz, F. Herfurth, R. He, P.-M. Hillenbrand, R. Hubele, P. Indelicato, A. Källberg, O. Kester, O. Kiselev, A. Knie, C. Kozhuharov, S. Kraft-Bermuth, T. Kühl, G. Lane, Y. Litvinov, D. Liesen, X. Ma, R. Märtin, R. Moshammer, A. Müller, S. Namba, P. Neumeyer, T. Nilsson, W. Nörtershäuser, G. Paulus, N. Petridis, M. Reed, R. Reifarth, P. Rei, J. Rothhardt, R. Sanchez, M. Sanjari, S. Schippers, H. Schmidt, D. Schneider, P. Scholz, R. Schuch, M. Schulz, V. Shabaev, A. Simonsson, J. Sjöholm, Ö. Skeppstedt, K. Sonnabend, U. Spillmann, K. Stiebing, M. Steck, T. Stöhlker, A. Surzhykov, S. Torilov, E. Träbert, M. Trassinelli, S. Trotsenko, X. Tu, I. Uschmann, P. Walker, G. Weber, D. Winters, P. Woods, H. Zhao, Y. Zhang, Eur. Phys. J. Special Topics 225 (2016) 797.