An EBIS-based RHIC Preinjector - Overview

J. AlessiOverview

EBIS Project Technical Review1/27/05

An EBIS-based RHIC Preinjector - Overview

Jim Alessi

Preinjector GroupCollider-Accelerator Department

MotivationRequirements

Intensity, species, q/m, energy, operating modes, …General comments on design choicesProject status

J. AlessiOverview


EBIS - Overview

Presently, one or two ~35-year old Tandem Van de Graaff accelerators are used for RHIC pre-injection, but the recent advances in the state of the art in EBIS performance by more than an order of magnitude now make it possible to meet RHIC requirements with a modern linac-based preinjector.

BNL now has DOE CD0 approval for new pre-injector for RHIC based on the Laboratory’s development of an advanced Electron Beam Ion Source (EBIS).

The new preinjector would consist of an EBIS high charge state ion source, a Radio Frequency Quadrupole (RFQ) accelerator, and a short linac.

J. AlessiOverview


Two Tandems presently serve as RHIC preinjectors

J. AlessiOverview


J. AlessiOverview


EBIS

Tandem to Booster transport line is 860 m long.44 quadrupoles71 steering dipoles6 large dipoles25 multiwires20 Faraday cupsDozens each of pumps, vacuum valves, vacuum gauges, etc.

The transport from the EBIS to the Booster will be only 30 m long!

J. AlessiOverview


Heavy Ion Preinjector for RHIC

The Tandem Van de Graaff is the present RHIC preinjector. Until our recent EBIS development, it was the only option which could meet RHIC requirements, and while presently it is quite reliable, it has disadvantages -

• 860 m transport line to Booster• Stripping foils at terminal and high energy

lead to intensity & energy spread variations • Injection over > 40 turns is required• Limitations on ion species (must start with negative ions)• High maintenance, manpower• Obsolete equipment requires upgrading

J. AlessiOverview


MODE PREINJECTOR OPERATION

Dedicated RHIC running; same species in both rings

5 Hz, 4 pulses every ~3 seconds; single Tandem

Dedicated RHIC running; two species 5 Hz, 4 pulses every ~3 seconds; 2 minute switching time between species; two Tandems

Dedicated NSRL running ~0.3 Hz; single Tandem

RHIC single species & NSRL same species, energy, charge state

5 Hz, 4+1 pulses every ~3 seconds; single Tandem

RHIC single species; NSRL different species of same magnetic rigidity

4 pulses at 5 Hz, ~1.5 second switching, then 1 NSRL pulse; two Tandems

RHIC single species; NSRL different species of different rigidity

~2 minute switching; two Tandems

To improve upon the present Tandem performance, it is desirable for the preinjector to be able to switch both species and transport line rigidity in ~1 second, so that there are no restrictions on compatibility between RHIC and NSRL operations.

REQUIREMENTS

Requirement for RHIC (one example) : 1.7 emA of Au 32+, 10 µs; 5 Hz plus….NSRL (NASA Space Radiation Laboratory) – interleaved second beam at 5 Hz: He 2+, C 6+, O 8+, Si 14+, Ti 18+, Fe 21+, Cu 22+, all at ~2-3 emA, ~ 10 µs

--- short pulses, fast beam changes, any species

J. AlessiOverview


The present control system supports “pulse-to-pulse modulation”The setpoint of any device can be changed pulse-to-pulse, depending on the “user”.

So, within 1 second:the source (EBIS) has to change species,the RFQ and linac have to change gradient (amplitude)transport line elements have to switch to new values

J. AlessiOverview


Intensity required at injection into the Booster, for RHIC:

Au32+ : 2.7 x 109 ions per pulse. (8.6 x 1010 charges/pulse). Sufficient to achieve 1.2 x 109 ions per bunch in RHIC.

D : 2.5 x 1011 ions/pulse. (will be produced from a plasma source, for convenience).

Cu11+ : 1.4 x 1010 ions/pulse. (1.5 x 1011 charges/pulse)

Species Q Ions/pulse Charges/pulse(Booster output)

C 5+ 1.2 x 1010 6 x 1010

O 8+ 4 x 109 3.2 x 1010

Si 13+ 3 x 109 5.2 x 1010

Ti 18+ 8 x 108 1.4 x 1010

Fe 20+ 1 x 109 2 x 1010

Beams for NSRL

With a present efficiency from Booster injection to extraction of ≥ 60% (EBIS should be better), one needs ≤ 1011 charges/pulse in all cases for NSRL. With a trap capacity in the EBIS of 1012 charges, these intensities should be readily achievable.

J. AlessiOverview


Species He to U

Intensity in desired charge state ≥ 1 x 1011 charges/pulse

Charge-to-mass ratio, Q/m ≥ 1/6

Repetition rate 5 Hz

Pulse width 10 – 40 µs

Switching time between species 1 second

Output energy 2 MeV/amu

Emittance (full beam, normalized) ≤ 1.4 π mm mrad

Momentum spread, dp/p ≤ ± 0.05%

Summary of Performance Specifications

Injection energy: 2 MeV/amu. Present tandem injection is at 0.92 MeV/amu for Au. At this energy, there is a significant beam loss due to electron capture during Booster injection. By raising the injection energy to 2 MeV/amu, the capture cross section is reduced by a factor of 20-40. In addition, the higher energy reduces the space charge tune shift at injection, which might be important at these higher peak currents. At even higher injection energies one would approach the voltage limit of the inflector, and losses due to ionization would begin to become important.

J. AlessiOverview


12 MVLINAC

The EBIS preinjector offers the following benefits: •Improvements in reliability, setup time, and stability should lead to increased integrated luminosity in RHIC

• Reduced operating costs, and avoidance of ~ 9 M$ in reliability-driven investments in the tandems

• Elimination of two stripping stages and an 860 m long transport line, leading to improved performance

• Simplification of Booster injection (few turn vs. present 40 turn)

• Increased flexibility to handle the multiple simultaneous needs of RHIC, NSRL, and AGS

• Capability to provide ions not presently available, such as noble gas ions (for NSRL), uranium (RHIC), or, with additional enhancements, polarized 3He (eRHIC) (the Tandems must start with negative ions)

•Simpler technology, robust, more modern (Tandem replacement parts becoming difficult to get)

J. AlessiOverview


4 m 4.3 m

RFQ: 17 - 300 keV/u; 100 MHz

Linac: 0.3 - 2.0 MeV/u; 100 MHz

Proposed Linac–Based RHIC Preinjector

J. AlessiOverview


Linac-based Preinjector - Source “requirements”

1. Intensity for 1 x 109 Au ions/bunch in RHIC : ~ 3.4 x 109 Au32+ ions/pulse from the source

2. No stripping before Booster injection : q/m > 0.16 (Au32+, Si14+, Fe21+)

3. 1-4 turn injection into Booster : pulse width 10-40 s

(Note - 1 & 3 result in a Au32+ current of 1.6 - 0.4 mA)

4. Rep rate : ~ 5 Hz

5. Emittance : 0.2 mm mrad, normalized, rms (for low loss at Booster injection)

J. AlessiOverview


Radial trapping of ions by the space charge of the electron beam.Axial trapping by applied electrode electrostatic potentials.

Ion output per pulse is proportional to the trap length and electron current.Ion charge state increases with increasing confinement time.Charge per pulse (or electrical current) ~ independent of species or charge state!

EBIS

J. AlessiOverview


Ions are extracted from an EBIS in pulses of constant charge; one has control over the pulse width

Charge state is selected by choosing theconfinement time of ions in the trap

J. AlessiOverview


EBIS Test Stand

J. AlessiOverview


J. AlessiOverview


RHIC EBIS

Design changes relative to the Test EBIS, needed to meet RHIC requirements:

1. Longer SC solenoid and ion trap region2. Collector design for higher average power

Improvements to increase operational reliability and safety margin:

3. Collector design for higher peak power4. Electron gun capable of 20 A operation5. Increases to the solenoid bore and maximum field

J. AlessiOverview


Advantages of an EBIS (vs. LIS, ECR) An EBIS can produce any type ions – from gas, metals, etc., and

is easy to switch species (pulse-to-pulse!)

One has control over the charge state produced (easy to get intermediate charge states, such as Au32+ or U45+)

One has control over pulse width, extracting a fixed charge – can better match to synchrotron requirements

EBIS produces a narrow charge state distribution ( ≥ 20% in the desired charge state), so there is less of a space charge problem in the extraction and transport of the total current

The scaling laws are understood

The source is reliable, and has excellent pulse-to-pulse stability, long life

Our R&D now shows that an EBIS meeting the RHIC requirements can be built. (mA’s of Au32+ !)

J. AlessiOverview


ECRFeatures, advantages

~ the only choice for high current, high Q, dc applications

Reliable; lots of operating ECRs, lots of experience

TechnologiesSC magnets; At high freq's, need SC sol and SC hexapole 28 GHz VENUS - 4 T injection field, 2 T hexapole at plasma chamberRF power source - 28 GHz gyrotron, 10-15 kW; plus sometimes multiple frequencies

Questions, issues?

Broad charge state distribution, so one has to extract & transport a high total current

Performance depends on species, favoring gases and low melting point solids

“Memory” effects, slower beam switching times

J. AlessiOverview


LISFeatures, advantages

Produces high currents, short pulses

TechnologiesHigh power laser – 100 J, CO2, 15-30 nsOpticsTargets – 3 x 1013 W/cm2 on the target

Questions, issues?

Laser reliability, rep rate

Pulse-to-pulse current fluctuations

Target erosion; coating of optics by target material

Species ~ limited to solid targets, high melting point solids are best

J. AlessiOverview


EBIS at different pulse widthsECR

LIS, short pulses

Comment (qualitative) ….

J. AlessiOverview


RFQ:

100 MHz, 4 rod design is conventional. Very similar to GSI, CERN, etc.

Deepak’s talk – 17 keV/u injection energy chosen to reduce space charge problems in LEBT.

LINAC:

IH structure chosen, very similar to CERN Pb linac. (conventional baseline design).

SCL was considered, but abandoned since it is not required to meet design requirements, it has higher cost, and the technology is less familiar in Collider-Accelerator, so would increase operational burden.

J. AlessiOverview


Planned location at the end of the 200 MeV linac

J. AlessiOverview


The EBIS, operating with a 10 A electron beam, will produce in excess of 5 x 1011 charges/pulse, for any desired species.

For heavy ions, ~20% of these ions will be in the desired single charge state, while for light ions this fraction can reach ~ 50%.

In addition, key components (electron gun and collector) will be designed with the capability of operating at up to 20 A electron beam current, so one will have the potential of a factor of 1.6 improvement with further developments.

J. AlessiOverview


EBIS

RFQ

LINAC

BOOSTER

STRIPPER FOIL

AGS

Au32+

Au32+

Au32+

Au32+

Au77+

17 keV/u 3.4 x 109 ions

300 keV/u 3.0 x 109 ions

2 MeV/u 2.7 x 109 ions

70 MeV/u 2.3 x 109 ions

70 MeV/u 1.4 x 109 ions

9 GeV/u 1.2 x 109 ions

STRIPPER FOIL

RHIC

Au77+

Au79+

90%

90%

85%

60%

90%

100%

9 GeV/u 1.2 x 109 ions

(=8.6e10 charges)

(=1e11 charges)

J. AlessiOverview


Intensity required at injection into the Booster, for RHIC:

Au32+ : 2.7 x 109 ions per pulse. (8.6 x 1010 charges/pulse). Sufficient to achieve 1.2 x 109 ions per bunch in RHIC. (met, as shown in previous slide)

D : 2.5 x 1011 ions/pulse. (4 x 1011 charges = 4 x 1011 ions/pulse from EBIS, but will be produced from a plasma source, for convenience).

Cu11+ : 1.4 x 1010 ions/pulse. (1.5 x 1011 charges/pulse) 5e11 charges x 81% x 27% in Cu11+ = 1.1e11 charges/pulse need 70% neutralization, or ~15A electron beam

Species Q Ions/pulse(Booster output)as run

EBIS(see note)

C 5+ 12 x 109 27 x 109

O 8+ 4 x 109 12 x 109

Si 13+ 3 x 109 5.3 x 109

Ti 18+ 0.8 x 109 3.8 x 109

Fe 20+ 1 x 109 3.4 x 109

He 2+ 172 x 109

Beams for NSRL

Note – assumption is 10A electron beam, 80% EBIS to Booster input, 85% Booster input to output.

All NSRL requirements can be met.

J. AlessiOverview


Project Status• DOE gave CD0 approval (“mission need”) in August, 2004.• Preparing all documents required for CD1 – (“alternative selection

and cost range”)

J. AlessiOverview


We are here! “Begin operations”

J. AlessiOverview


J. AlessiOverview


SUMMARY

• The EBIS initiative, a linac-based preinjector, is based on a modern technology, which will be simpler to operate and easier to maintain than the Tandems and will have the potential for future performance improvements.

• It will provide a robust and stable preinjector, which is important for the successful operation of the injectors.

• The RHIC EBIS design has been verified by the present EBIS operating at BNL (next talk).

• No significant improvement in performance is required, other than the straightforward scaling of ion output with an increase in trap length. The RFQ and linac are very similar to devices already operating at other labs.

Documents

An EBIS-based RHIC Preinjector - Overview