Heavy ion fusion (HIF) places great demands on the capabilities of ion sources. There are two approaches to HIF-the RF linac-storage ring approach, and the induction linac approach. The first option is pursued mainly in Europe, where the main centers of source work are GSI in Darmstadt, Germany, and the Institute of Theoretical and Experimental Physics in Moscow. RF linac schemes require low emittance and moderate current levels, because the beam is accumulated in storage rings before being focused on target. The induction linac approach being pursued at the Lawrence Berkeley Laboratory (LBL) requires low emittance and high current, because this is a single pass approach to HIF and one wishes to limit the number of beams in the machine. The RF scheme generally uses long pulse sources together with a buncher or RFQ. The induction linac approach requires sources in the microsecond pulse length range, with good optics being maintained during the pulse. The mainline source being pursued at LBL is a carbon vacuum arc providing 500 mA of ions per beam, which incorporates electrostatic confinement of the plasma to provide a stable ion-extraction surface geometry throughout the pulse duration (plasma switch). Other materials can be used for the arc cathode to provide heavier ions, and indeed some work has been performed with aluminum. At the present stage of the program, the emphasis is on scaled accelerator experiments which require lighter ions. The electrostatic fields used for the plasma confinement have great effect on the output emittance of the source, and particle-in-cell simulations have been performed to find an optimum solution for a given set of plasma parameters that will result in a minimum source emittance. The present design goal for the LBL source which will be used in the 16 beam 2 MV injector is 5 x 10(-7)pi-m-rad normalized. Without plasma switch optimization, the achieved normalized emittance is 2.0-2.5 x 10(-6)pi-m-rad. Experiments have also been performed to show the enhanced yield of multiple charge state ions, which have considerable impact on the cost of any heavy ion driver for fusion. This enhanced yield is accomplished by modifying the plasma potential in the source. Langmuir probe diagnosis of the plasma has been performed as part of the effort to reduce emittance by tailoring the output plasma to the conditions suggested by the simulations. Considerable effort has also been devoted to lifetime enhancement and remote trigger cleaning methods. The normal lifetime of the source is 20-30 K shots, which is fine for an experimental machine but clearly not suitable for a commercial driver.
