IINNEERRTTIIAALL CCOONNFFIINNEEMMEENNTT LLaawwrreennccee LLiivveerrmmoorree NNaattiioonnaall LLaabboorraattoorryy ICF Quarterly Report April–June 1996, Volume 6, Number 3 Ion-Beam Propagation in a Low-Density Reactor Chamber Efficient Production and Applications of 2- to 10-keV X Rays Laser Tissue Interaction Modeling with the LATIS Computer Program The Energetics of Gas-Filled Hohlraums Measurements of Laser-Speckle-Induced Perturbations Fusion Reaction-Rate Measurements— Nova and NIF UCRL-LR-105821-96-3 The Cover: The ICF Quarterly Report Image of a time-resolved neutron sig- is published nal and related timing pulses recorded with an LLNL four times each fiscal year by the Inertial Confinement optical streak camera. Clearly visible detector pixels in Fusion Program at the Lawrence Livermore National this magnified image and a low-level signal accent the Laboratory. The journal summarizes selected current statistical variation in the background region to create an research achievements of the LLNLICFProgram. The interesting mosaic pattern. The encoded neutron signal underlying theme for LLNL’s ICF Program research is shows an abrupt edge corresponding to the instant in defined within DOE’s Defense Programs missions and which neutrons first reach the detector (time goes from goals. In support of these missions and goals, the ICF right to left). Neutron signals like this are used to deter- Program advances research and technology develop- mine reaction rates with ~30 ps resolution for ICF targets. ment in major interrelated areas that include fusion Evenly spaced pulses 500 ps apart (across top of image) target theory and design, target fabrication, target exper- are used to precisely time the neutron signal relative to iments, and laser and optical science and technology. the laser power irradiating a target. For details, see the article “Fusion Reaction-Rate Measurements—Nova and While in pursuit of its goal of demonstrating thermonu- NIF,” beginning on page 115. clear fusion ignition and energy gain in the laboratory, the ICFProgram provides research and development opportunities in fundamental high-energy-density physics and supports the necessary research base for the possible long-term application of inertial fusion energy for civilian power production. ICF technologies continue to have spin-off applications for additional government and industrial use. In addition to these topics, the ICF Quarterly Reportcovers non-ICF funded, but related, laser research and development and associated applica- tions. We also provide a short summary of the quarterly UCRL-LR-105821-96-3 activities within Nova laser operations, Beamlet laser Distribution Category UC-712 April–June 1996 operations, and National Ignition Facility laser design. LLNL’s ICF Program falls within DOE’s national ICF program that includes the Nova and Beamlet (LLNL), OMEGA (University of Rochester Laboratory for Laser Energetics), Nike (Naval Research Laboratory), and Trident (Los Alamos National Laboratory) laser facilities. The Particle Beam Fusion Accelerator and Saturn pulsed Printed in the United States of America power facilities are at Sandia National Laboratory. Available from National Technical Information Service General Atomics, Inc., develops and provides many of U.S. Department of Commerce the targets for the above experimental facilities. Many of 5285 Port Royal Road the Quarterly Reportarticles are co-authored with our col- Springfield, Virginia 22161 Price codes: printed copy A03, microfiche A01. leagues from these other ICF institutions. This document was prepared as an account of work sponsored by an Questions and comments relating to the technical agency of the United States Government. Neither the United States content of the journal should be addressed to the ICF Government nor the University of California nor any of their employ- Program Office, Lawrence Livermore National ees makes any warranty, express or implied, or assumes any legal lia- bility or responsibility for the accuracy, completeness, or usefulness Laboratory, P.O. Box 808, Livermore, CA 94551. of any information, apparatus, product, or process disclosed, or rep- resents that its use would not infringe privately owned rights. Reference herein to any specific commercial products, process, or ser- vice by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or the University of California. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government Work performed under the auspices of the U.S. Department of or the University of California and shall not be used for advertising Energy by Lawrence Livermore National Laboratory under Contract or product endorsement purposes. W–7405–Eng–48. 1 IINNEERRTTIIAALL CCOONNFFIINNEEMMEENNTT ICF Quarterly Report April–June 1996, Volume 6, Number 3 In this issue: Foreword iii Ion-Beam Propagation in a Low-Density Reactor Chamber for Heavy-Ion Fusion 89 We have assessed constraints of spot size and vacuum for ballistic transport of ion beams in inertial fusion energy reactors. We have simulated transport of a partially neutralized beam at low pressures and examined interactions of multiple beams. Further simulations show that partial beam neutralization allows higher chamber pressures and higher charge-to-mass ions, both of which allow for lower reactor cost. Efficient Production and Applications of 2- to 10-KeV X Rays by Laser-Heated Underdense Radiators 96 The proposed National Ignition Facility (NIF) offers the prospect of producing up to several hundred kilojoules of multi-keV x rays. This may allow us to perform experiments and field diagnostics we could never consider with current laser facilities. We discuss applications of high-energy, multi-keV sources with the NIF. Laser–Tissue Interaction Modeling with the LATIS Computer Program 103 A new computer program, based on many years of experience in laser-matter interaction modeling, is being used to design new laser-medical instruments and procedures. The Energetics of Gas-Filled Hohlraums 110 We have measured the effect of gas-fill on the drive temperature in Nova scale-1 hohlraums. Increasing electron density results in a reduced drive temperature. A significant part of the reduced drive can be attributed to stimulated scattering of the incident laser light. This scattering may be reduced by beam smoothing. Scientific Editor Fusion Reaction-Rate Measurements—Nova and NIF 115 Alan Burnham At Nova, we measure ICF targets’ burn history with a resolution of <30 ps. Our neutron-based technique uses a fast-rise-time plastic scintillator and a high-speed optical streak camera. Publication Editors Measured burn durations range from ~50 ps to ~1 ns. For the NIF, we are investigating new Jason Carpenter measurement techniques based on gamma rays released in the fusion process. Al Miguel Ann Parker Laser-Speckle–Induced Perturbations in Laser-Driven Foils 123 Dabbie Schleich We have calibrated the amplitude and Rayleigh–Taylor growth of modulations imprinted by laser speckle in CH foils with single-mode surface perturbations, converting the imprint to an Design Staff 2 equivalent surface finish. The addition of bandwidth and dispersion to the drive laser reduced Daniel Moore the imprinted modulations, with the highest bandwidth showing the largest reduction in imprint, in agreement with LASNEX simulations. Art Staff Sandy Lynn Linda Wiseman Program Updates U-1 Publications P-1 i F OREWORD This issue of the ICF Quarterlypresents a diverse set of articles ranging from interpreta- tion of current Nova experiments to weapons effect testing on the proposed National Ignition Facility (NIF), to medical uses of lasers, and to inertial fusion energy (IFE) drivers. The lead article, “Ion-Beam Propagation in a Low-Density Reactor Chamber for Heavy-Ion Inertial Fusion” (p. 89), explores the ability of heavy-ion beams to be ade- quately transported and focused in an IFE reactor. Such beams are a leading candidate for reactors because of their high efficiency, repetition rate, and lifetime. The vapor pressure of the molten chamber wall has the potential to prevent adequate beam focus- ing because of beam stripping, but the authors show that partial beam neutralization by several candidate techniques can overcome that problem. Neutralization also provides the possibility of using lower-cost, lighter-ion beams such as Cs+. The next article, “Efficient Production and Applications of 2- to 10-KeV X Rays by Laser-Heated Underdense Radiators” (p. 96), explores the ability of the NIF to produce sufficient high-energy x rays for diagnostic backlighting, target preheating, or uniform irradiation of large test objects for Nuclear Weapons Effects Testing. The latter applica- tion illustrates the increasing emphasis of the NIF for nonignition experiments. By appropriate high-power illumination of low-density, high-Zmaterials, a considerable fraction of the laser energy can be converted into KeV x rays. For example, focusing six of the 10-TW 4-beamlet clusters on the NIF into a Xe-filled hohlraum would convert 30% of the laser energy into x rays with energies greater than 4 KeV and 48% into ener- gies greater than 1 KeV. For capsule implosion experiments, the increasing energies and distances involved in the NIF compared to Nova require the development of new diagnostics methods. The article “Fusion Reaction-Rate Measurements—Nova and NIF” (p. 115) first reviews the use of time-resolved neutron measurements on Nova to monitor fusion burn histo- ries and then explores the limitations of that technique, principally Doppler broadening, for the proposed NIF. It also explores the use of gamma rays to follow burn history and reports the first measurement of ICF fusion gamma rays on Nova, thereby providing a proof-of-principle for using gamma rays for monitoring fusion burn histories on the NIF. The articles “The Energetics of Gas-Filled Hohlraums” (p. 110) and “Measurements of Laser-Speckle-Induced Perturbations in Laser-Driven Foils” (p. 123) report measurements on Nova of two important aspects of implosion experiments. The first characterizes the amount of energy lost from a hohlraum by stimulated Brillioun and Raman scattering as a function of gas fill and laser-beam uniformity. Some amount of hohlraum gas fill is anticipated for indirectly driven implosions on the NIF in order to minimize x-ray drive asymmetry due to ablation of the hohlraum wall. Although energy losses by scattering up to 25% were observed for unsmoothed beams, scattering loss was reduced to about 5% for a single smoothed beam and a gas fill producing a plasma with 0.04 of its critical density. The second of these articles shows that the growth of density nonuniformities implanted on smooth capsule surfaces by laser speckle can be correlated with the effects of physical surface roughness. This work helps define the laser beam smooth- ness required for directly driven implosions. The article “Laser–Tissue Interaction Modeling with the LATIS Computer Program” (p. 103), explores the use of modeling to enhance the effectiveness—maximize desired effects and minimize collateral damage—of lasers for medical purposes. The program LATIS, a two-dimensional, time-dependent simulation tool recently developed using the experience of LASNEX, is used to predict heat deposition and the resultant thermal, material, and hydrodynamic responses from lasers interacting with tissue. When fully developed and validated, LATIS can be used to optimize medical procedures such as photothermal therapy, tissue welding, hard-tissue ablation, and thrombolysis. Alan Burnham Scientific Editor UCRL-LR-105821-96-3 iii I -B P L -D R ON EAM ROPAGATION IN A OW ENSITY EACTOR C H -I I F HAMBER FOR EAVY ON NERTIAL USION D. A. Callahan A. B. Langdon Introduction most conventional ion accelerators, space-charge plays only a small role; in a HIF accelerator, space-charge is Heavy-ion fusion (HIF) is an attractive candidate for very important. inertial fusion energy (IFE) production. In HIF, beams The gain of a heavy-ion target generally increases as of heavy ions (mass ~100 to 200 amu) are accelerated the size of the beam focal spot decreases. In order to and focused on an indirectly driven inertial confine- get a small spot (radius ~2 to 5 mm), the beam emit- ment fusion (ICF) target. At the target, the ion energy tance must be kept small. The beam emittance (a term is converted in a hohlraum into x rays, which implode used in accelerator physics) is proportional to the the capsule. Accelerator drivers have the long lifetime phase space area occupied by the beam and is related (~30 years), high repetition rate (~5 Hz), and high to the beam temperature. Keeping the emittance small efficiency (~30%) needed for commercial energy pro- throughout the accelerator in the presence of nonlinear duction. In addition, final focusing is accomplished via fields (resulting from imperfect accelerator elements, magnetic fields, which are not damaged by the blast. misalignments, fringe fields, etc.) while preformingthe The accelerator requirements are set by the target. In necessary beam manipulations (beam bending, beam order to get efficient coupling of ion energy into x rays, combining, etc.) is a challenge. Recent Quarterlyarti- the amount of radiator material in the target (Fig. 1) cles by W. M. Sharp1and A. Friedman2discuss issues that needs to be heated must be kept reasonably small. and progress towards understanding the accelerator Because the ions must be stopped in the target, the ion needed for a HIF power plant. range (equal to the stopping distance times the material Once the beams leave the accelerator, they must be density) is related to the amount of material necessary. focused and transported through the reactor chamber In order to keep the target mass reasonably small, an to the target. Transporting the heavy-ion beams from ion range (cid:178)0.1 g/cm2is used. For ions of mass 100 to the accelerator to the target is extremely important for 200 amu, this corresponds to an ion energy of 3 to 10 GeV. a successful HIF reactor. Chamber transport (along To get the required beam power on target (5 to 10 MJ with the target) sets the requirements on the accelera- in 10ns = 0.5–1.0 · 1015W), the beam current must be tor driver.3 As a result, improvements in chamber greater than 50 kA. This current is broken into several transport and final focus can significantly reduce the beams to keep the space-charge forces manageable. In cost of electricity. Relaxing the requirements on the accelerator reduces the cost of the driver, which directly impacts the cost of electricity; reducing the beam spot size at the target allows a larger target gain, Symmetry shields which also reduces the cost of electricity. Improvements in beam transport and final focusing can be exploited Ion Ion beam beam in optimizing the end-to-end HIF system. The main-line approach to chamber transport is low-density, ballistic or nearly ballistic transport. The HYLIFE-II reactor4uses a low-density chamber with a Beam absorber and radiator pressure of a few millitorr. Even at the low density of FIGURE1. Schematic IFE heavy-ion indirect-drive target with end the HYLIFE-II chamber (5 · 1013cm–3(cid:197)0.003 torr), radiators. (40-00-1096-2401pb01) partial beam-charge neutralization is needed to overcome UCRL-LR-105821-96-3 89 ION-BEAMPROPAGATIONINALOW-DENSITYREACTORCHAMBER the effects of beam stripping. In this case, beam strip- angle of the beam. Including space charge reduces the ping refers to collisional ionization of the beam ions to chromatic aberrations by about a factor of 3/4, so a higher charge state by the background gas. Beam stripping and neutralization have been recognized as d (cid:112) important issues in chamber transport for many years.4, 5 (cid:114) (cid:32)» (cid:32)(cid:54)(cid:70)q (cid:46) (2) (cid:99) (cid:112) Low-density transport is the most conservative option, but puts strict requirements on the beam quality out of the accelerator. Generally Fis set by the reactor geometry and has a In this article, we first discuss different effects that value of ~5 m. The convergence angle q is generally impact the beam spot size at the target. We then show limited to ~15 mrad to avoid geometric aberrations in that ballistic transport in a near vacuum is possible, but the final focusing magnets. It is possible to use larger puts undesirable constraints on the reactor chamber design. angles and correct the geometric aberrations using an We then discuss simulations of partially neutralized beam octupole correction.8 A larger convergence angle results transport in a low-density chamber. Finally, we discuss cal- in a larger aperture for the final focusing magnets, how- culations of multiple beam effects in the chamber. Partial ever,which increases the cost of those magnets. For the neutralization allows chamber operation at higher purposes of these calculations, we will keep q (cid:197)15 mrad. pressures, and the use of ions with higher charge-to- Limiting the spot radius from chromatic aberrations to mass ratio which are easier to accelerate. Both increases 1 to 1.5 mm sets a limit on the allowable momentum in flexibility allow for lower reactor cost. spread in the final focusing system. Using F= 5 m and q = 15 mrad, we find d p/p(cid:178)2.2 – 3.3 · 10–3in the final Total Beam Spot Size at the Target focusing system. In the conventional HIF driver scenario, the beam is drift compressed by a factor of 10 or more In this section we describe the different effects that between the end of the accelerator and the final focusing lead to the final beam spot size at the target. The beam system. This is accomplished by giving the beam a veloc- spot size is increased by the beam’s space charge and ity “tilt” so that the beam tail is moving faster than the emittance, chromatic aberrations in the final focusing sys- beam head. After the tilt is applied, the beam is allowed tem, and errors in aiming the beams at the target. These to drift and it compresses as the tail catches up with the sources are roughly independent and add in quadrature,6 head. This drift compression increases the longitudinal momentum spread, and, as a result, d p/p(cid:178)2.2 – 3.3 · 10–4 at the end of the accelerator if the beam is to be drift-com- (cid:114)(cid:50) » (cid:114)(cid:50)+ (cid:114)(cid:50)+ (cid:114)(cid:50) (cid:44) (1) pressed by a factor of 10 prior to focusing. (cid:116)(cid:97)(cid:114)(cid:103)(cid:101)(cid:116) (cid:115) (cid:99) (cid:97) If r = 0.4 mm, and r = 1.5 mm, then 2.57 mm are a c left for space charge and emittance (r )in a total spot s where r is the final spot radius at the target, and (r )of 3 mm. An optimization needs be done to target target r , r ,and r are the spot radius due to space charge and weigh the relative costs of each of the spot size contri- s c a emittance, chromatic aberrations, and aiming, respectively. butions. We need to assess the cost of increasing d p/p The simulation results presented in the sections below at the expense of beam emittance, for example. do not include chromatic aberrations or aiming errors; The beam-spot radius from space charge and emit- Eq. (1) is used to estimate the total spot radius. The tance can be estimated using the envelope equation spot radius necessary for a high gain ((cid:179)40) ranges from 2 to 6 mm depending on the details of the target design. (cid:75) e (cid:50) In the HYLIFE-II reactor concept, the targets are (cid:97)¢¢ (cid:32)=(cid:32) (cid:32)+(cid:32) (cid:44) (3) (cid:97) (cid:97)(cid:51) injected into the chamber at a rate of six per second using a gas gun. Petzoldt7estimates that the spot radius due to errors in aiming the beams at the target where each ¢ indicates a derivative with respect to z, K is 0.4 mm. This estimate takes into account transla- is the perveance (which is a measure of the beam’s tional positioning errors and rotations of the target. space charge), e is the unnormalized emittance, and a Chromatic aberrations occur in the final focusing is the beam edge radius. Multiplying by a¢ and inte- system because particles with different longitudinal grating gives momenta are focused at different distances. This causes a radial spread in the particles at target, which is pro- ( ) ( ) (cid:230) (cid:246) (cid:230) (cid:246) portional to d p/p, the longitudinal momentum spread (cid:97)¢ (cid:50)(cid:32)(cid:177)(cid:32) (cid:97)¢ (cid:50)(cid:32)=(cid:32)(cid:50)(cid:75)(cid:108)(cid:110)(cid:231) (cid:97)(cid:102)(cid:247) (cid:32)(cid:177)(cid:32)e (cid:50)(cid:231) (cid:49) (cid:32)(cid:177)(cid:32) (cid:49) (cid:247) (4) divided by the longitudinal momentum. For a focusing (cid:102) (cid:48) Ł (cid:97)(cid:48)ł Ł (cid:97)(cid:102)(cid:50) (cid:97)(cid:48)(cid:50)ł system of four thin lenses, single-particle calculations (neglecting space charge) show that r = 8Fdq p/p, where F where 0 and f denote the initial and final values. At c is the focal distance, and q is the half convergence the beam waist a = r and (cid:97) ¢ = 0. At the entrance, f s (cid:102) 90 UCRL-LR-105821-96-3 ION-BEAMPROPAGATIONINALOW-DENSITYREACTORCHAMBER (cid:97)¢ = q and a (cid:197)Fq . Using these substitutions and molten LiF-BeF chamber wall) range from 1.3 to (cid:48) 0 2 assuming a » r , Eq. (4) becomes 4.0 · 10–16cm2. Stripping only about 1% of the beam 0 s ions requires a chamber density (cid:197)1011cm–3. This (cid:230) (cid:97) (cid:246) e (cid:50) density is two orders of magnitude below the cham- q (cid:50)(cid:32)=(cid:32)(cid:50)(cid:75)(cid:108)(cid:110)(cid:231) (cid:48)(cid:247) (cid:32)+(cid:32) (cid:46) (5) ber density in the HYLIFE-II reactor ((cid:197)5 · 1013cm–3). Ł (cid:114) ł (cid:114)(cid:50) (cid:115) (cid:115) Since a HIF reactor has a repetition rate of about 5 Hz, it is difficult to achieve densities lower than used in In the absence of space charge (i.e., a perfectly neutral- HYLIFE-II after each shot. Thus, purely ballistic ized beam), K= 0 and the spot radius due to emittance r transport puts an undesirable restriction on the e is given by chamber density. e (cid:114)(cid:101)(cid:32)=(cid:32)q (cid:46) (6) Low-Density, Nearly Ballistic Transport If r = 1 mm and q = 15 mrad, then transverse beam e emittance is restricted to e (cid:178)15 mm-mrad. One option for improving the beam focus at the Using Eq. (5), we can estimate the maximum beam density of HYLIFE-II is to partially charge neutralize perveance allowed for a given r . Using q = 15 mrad, the beam. At the density of the HYLIFE-II chamber, the s a = 7.5 cm, r = 2.5 mm, and e = 15 mm-mrad, Eq. (5) stripping mean-free-path is 0.5 to 1.5 m. In the 0 s gives a maximum perveance of K= 2.8 · 10–5. The per- HYLIFE-II design, the chamber radius from “first veance is related to the beam current by wall” to the target is 3 m (from target to the center of the last focusing magnets is about 5 m, as used in the (cid:50)(cid:90) (cid:73) “Total Beam Spot Size at the Target” section). Therefore, (cid:75)(cid:32)=(cid:32) (cid:32) (cid:98) (cid:44) (7) (gb )(cid:51)(cid:65) (cid:73)(cid:48) tphoer tb. e Saimm wulialtl isotnrisp w 2i ttho t6h teim BIeCs rdzu croindge 1c2h–1a5msbheorw tr tahnast- most of the stripped electrons tend to stay with the beam. where Zis the ion charge state, b is the beam velocity However, in the higher charge state, the ions respond over the speed of light, g is the Lorentz factor, I is more strongly to the electric fields and the spot size b the beam current, Ais the ion mass in amu, and increases. Simulations with a stripping mean-free-path I = m c3/e= 31 MA. For K= 2.8 · 10–5, this leads of 1.2 m in a 3-m chamber show an increase in the beam 0 amu to a maximum current per beam of 2.3 kA for a spot radius from 2.6 mm (vacuum transport) to 8 mm. mass-200 ion and 1.6 kA for a mass-135 ion, both at This is an unacceptably large spot, and partial beam bg = 0.3. Neutralization is needed if the current per neutralization must be used to offset this increase. beam is larger than these values. While beam stripping makes chamber propaga- tion more difficult, ionization of the background gas Ballistic Transport in a Near by the beam ions can partially neutralize the beam and aid transport. Cross sections for collisional ion- Vacuum ization of BeF by the beam have a larger uncertainty 2 than stripping cross sections because calculating Using the equations introduced in the previous sec- molecular cross sections is more difficult than tion, we can show that a beam of 10-GeV, singly charged, calculating atomic cross sections. Estimates of the heavy (~200-amu) ions can be ballistically transported mean-free-path for ionizing the background gas with a reasonable spot size provided the chamber den- range from 0.7 to 25 m. For BeF , the ratio of strip- 2 sity is low enough to avoid beam stripping. The target ping to gas ionization cross sections is not favorable requires a main pulse with 4 MJ of energy in 10 ns (an (i.e., more stripping than gas ionization). Other additional 1 MJ is carried by a low-power prepulse). chamber gases, such as Li, may have a more favor- The total current necessary in the main pulse is able cross section ratio.16 4 MJ/(10GeV · 10 ns) = 40 kA. Without neutralization, Simulations show that including a stripping the maximum current per beam found in the previous mean-free-path of 1.2 m and a gas ionization mean- section was 2.3 kA, so 17 beams are needed for the main free-path of 3.0 m reduced the spot radius from 8 to pulse. This provides a reasonable, conservative scenario 5.4 mm. Neutralization by gas ionization occurs for transporting the beam to the target. “for free” since we do not have to add anything to Beam stripping is an issue, however. Estimates of the reactor for it to occur; however, gas ionization the cross section9–11for stripping the beam ions by neutralizes the beam slowly, so radial velocities the background gas in HYLIFE-II (BeF from the develop before neutralization occurs. The solid 2 UCRL-LR-105821-96-3 91 ION-BEAMPROPAGATIONINALOW-DENSITYREACTORCHAMBER curve in Fig. 2 shows the neutralization fraction as a electrons are accelerated longitudinally by the zelectric function of distance from the chamber wall for a simu- field of the beam and oscillate back and forth across lation without beam stripping. The beam is more than the beam in both the radial and longitudinal direction. 70% neutralized, but it takes about 2.5 m of propaga- The details of the acceleration and deceleration of the tion distance to reach this level of neutralization. To electrons in the longitudinal direction will depend on get a smaller spot, additional neutralization is needed. the shape of the beam. These simulations used a beam with a current profile that was parabolic in z. Because the BICrz code is axisymmetric, particles cannot gain or lose angular momentum. The only angu- 1.0 lar momentum a particle has is the momentum it is created with. For the plasma annulus, we used an ini- Plasma annulus tial isotropic temperature of 100 eV. The particles can 0.8 heat in rand z, but not in q . As a result, a temperature anisotropy develops, which causes an excess of elec- n trons near the axis. This causes the radial fields to o cti become nonlinear. Figure 3 shows the radial velocity a 0.6 r n f o ati z ali 8 r 0.4 ut Gas ionization e N 7 0.2 –40 6 1 · m 5 0 Vbea 0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 / n 4 Distance from chamber entrance (m) o cti FIGURE2. Neutralization fraction as a function of distance from the efle 3 chamber entrance for neutralization using collisional ionization of d the background gas (black curve) and neutralization using a preformed city plasma annulus in the chamber (gray curve). (50-00-0696-1362pb01) o 2 el V 1 Neutralization Using a Preformed 0 Plasma Annulus 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 r (cm) Neutralizing an ion beam is more difficult than neu- FIGURE3. The radial velocity deflection of particles near the axial tralizing an electron beam. When an electron beam center of the beam as a function of radius shows the electric field is passes through a plasma, the plasma electrons are linear when no electrons are present. (50-00-0696-1363pb01) moved out of the beam path, and the beam is neutral- ized by the immobile ions. For an ion beam, however, electrons must be pulled in from outside the beam path in order to reduce the net charge. deflection of particles near the center of the beam as a One method for neutralizing the beam quickly is to function of radius when no electrons are present. As create a preformed plasma in the chamber before the expected, the field is linear. Figure 4 shows the same beam enters. Simulations used a small (0.3-m = 40% plot when the beam has passed through a plasma annu- of the beam length) annulus of plasma just inside the lus (but with no beam stripping). The field is greatly chamber entrance. A 4-kA beam of Pb+ions entered reduced from the previous case, but the nonlinearity is the chamber through the annulus. The annulus had a also apparent. Because of the abundance of electrons total electron charge of four times the beam charge. near the axis, the field is negative at small radii. We can As the beam entered the chamber, electrons were compensate for linear fields by increasing the focusing pulled from the inner surface of the annulus by the angle at the final optic. We cannot compensate for the large radial electric field of the beam (E = 16 MV/m at nonlinear fields in this way; as a result the nonlinear r the beam edge for a 4-kA beam of radius 5 cm). The fields can make it more difficult to focus the beam. 92 UCRL-LR-105821-96-3