Accelerated Plasmas and Intense Ion Beams for Commercial Processing

H. A. Davis [(505) 665-8373], J. T. Scheuer, K. F. Schoenberg (P-24)

Introduction

Two technologies using intense-pulsed streams of particles are being developed in Group P-24 primarily as flash-heat sources for the advanced and environmentally conscious commercial processing of materials. The first technology, the coaxial plasma accelerator (Fig. 1), is a simple, compact, and mechanically robust device that uses the Lorentz force (j x B) to impulsively accelerate plasma to very high velocities (1 to 5 cm/ms). The plasma is produced in a coaxial accelerator by discharging energy from a capacitor bank with a few kJ of stored energy through a gas puff injected into the muzzle region of the gun. Radial current flow and the azimuthal self-magnetic field give rise to an axial acceleration of the plasma. Typical ion energies are in the range of 10 to 100 eV with 0.1 to 1.0 J/cm2 delivered to the target in a few microseconds. Any gas species can be used to generate the plasma, but typically a nonreactive species such as argon is used. A thin skin of energy from the plasma is deposited on the surface of the target and is transported into the bulk of the material through thermal conduction.

The second technology uses an intense-ion beam with higher particle energy and energy density at the target than accelerated plasmas. As shown in Fig. 2, the beam is produced by accelerating the ions in a diode. The anode of the diode consists of a dielectric ring inserted inside an aluminum plate. The cathodes consist of the tips of two thin concentric conical sections. Just before the accelerating voltage is applied, a fast-risetime (200-ms) magnetic field of a few kG is applied transverse to the anode-cathode gap by two magnetic field coils: one located inside the inner cone and one located outside the outer cone. At the peak magnetic field, a positive accelerating voltage from a Marx generator is applied to the anode. A combination of tangential electric fields at the anode surface and weak electron loss to the anode cause the dielectric anode insert to flashover, providing a source of ions for the beam. The applied magnetic-field strength is adjusted to prevent electrons from crossing the anode-cathode gap, but the more massive ions, which are only very weakly deflected by the magnetic field, are ballistically focused to a spot located 35 cm from the anode. Peak particle energy is 400 keV with up to 30 J/cm2 delivered to the target over a 10-cm-diam spot. The 1-ms beam is composed of a combination of protons and carbon and oxygen ions. Beam energy is deposited in targets over typically a 1- mm ion range. This range is much deeper than that of the accelerated plasmas. Thermal conduction into the target is relatively unimportant for many target materials.

Accelerated Plasma Applications

The heat pulse delivered to a target by accelerated plasmas can be used to modify the chemistry, crystal morphology, topography, or density of the thin surface layer of a target material while not affecting the underlying bulk properties of the material. One key application is large-area patterned etching of thin-surface deposited layers as shown in Fig. 1. This technology provides clear advantages over conventional "wet" chemical etching techniques by being more environmentally benign and for potentially reducing manufacturing costs per component. Furthermore, accelerated plasmas offer significant cost advantages over competing advanced etching technologies by virtue of their high-energy deposition and broad-area coverage on the target, which provide very high etch rates with concomitant high-manufacturing throughput. However, full commercialization of accelerator technology requires the means to control the distribution of directed energy on a target located downstream of the accelerator.

Magnetic nozzles hold great promise for the enhanced control necessary for manufacturing applications. Conceptually, the magnetic nozzle is analogous to a gas-dynamic nozzle, where the nozzle shape can modify flow-velocity and flow-streamline distributions, both within the accelerator and downstream of the accelerator. Recent Los Alamos research using numerical simulation with experiments on the CTX (compact torus experiment) coaxial plasma accelerator demonstrated that magnetic nozzles provide a method to improve accelerator reproducibility, efficiency, and energy/power distribution on a downstream target. For example, without the benefit of a magnetic nozzle, plasma accelerators typically concentrate energy along the axis of symmetry of the device. This "magnetic pinching" results in an undesirable nonuniform energy distribution on the target. Magnetic nozzles can reduce magnetic pinching and spread the plasma plume evenly over a wide area, resulting in a more uniform energy distribution at the target. A new accelerator with industrially relevant size and parameters, currently under development at Los Alamos, will be used to investigate materials applications and to study the fundamental physics of magnetic nozzles.

Intense Ion Beam Applications

There are two primary materials treatment methods that use intense-ion beams as flash-heat sources. In the first, the ion beam is focused on a target to rapidly heat and vaporize target material. The energetic ions deposit their energy in a typical depth of 1 mm; the target material is heated to tens of eV. The vaporized and ionized material expands in a plume normal to the surface of the target and is condensed onto a substrate forming a high-quality film. The numerous advantages of this type of deposition process over standard vapor deposition include the following:

very high rates and low cost are possible;

very energetic, strongly ionized plume material allows for the formation of materials unattainable otherwise (e.g., the formation of diamond-like carbon requires a minimum kinetic energy of about 10 eV);

target-to-substrate stoichiometry is preserved (congruent deposition) and allows for the formation of complex films (e.g., high-temperature superconducting materials); and

crucibles and filaments are eliminated, and thus film purity is improved.

The process is similar to the well-established pulsed laser deposition (PLD), but it has a number of advantages over the laser technique. The total energy of the beam is much higher, there is no reflected energy (i.e., much of the laser energy can be reflected from the target), and ion-beam production is cheaper and more efficient than with the eximer lasers required for PLD. Pulsed ion beam deposition is estimated to be a few hundred to a few thousand times cheaper than PLD and will have much higher throughput. Experiments at Los Alamos have demonstrated, for example, the deposition of diamond-like carbon (e.g., for field-emissive displays, candidates for the next generation of high-definition, flat-panel displays) at 1-mm/s instantaneous rates with state-of-the-art electron-emission properties and the congruent deposition of YBa2Cu3O7-x (1-2-3) superconducting material.

The second process uses ion beams for surface treatment of solids. Here an intense ion beam rapidly melts the near surface of the target. Thermal diffusion is minimized by employing pulses of less than 1.0-ms duration. Rapid resolidification of the target material produces amorphous layers, dissolves precipitates, and forms nonequilibrium microstructures. Experiments at Los Alamos and elsewhere have demonstrated a threefold increase in the hardness of O1 tool steel, a more than sixfold increase in the pitting time of 2024-T3 aluminum alloy, and a decrease in surface roughness of Ti-6Al-4V from 5 to 0.1 mm. Increases in the surface hardness of polymers by a factor of 20 has been demonstrated at Oak Ridge National Laboratory through cross-linking; this accomplishment will, for example, open the way for processing polymer stamping dies.

The next step in demonstrating commercial viability is to move from our current single-shot capability to repetitive high-average-power beam technology. When achieved this will open the door to high-payoff applications such as low- cost photovoltaics and advanced composite materials. High-average power technology requires two new elements: (1) a diode capable of repetitive firing and (2) a modulator system to drive the diode. Of the two new elements, the diode is the most complex. The surface-flashover diode currently used, an inherently single-shot device, will be replaced with an active-anode plasma source whereby the anode is the surface of a plasma generated on each pulse. The anode cannot erode, the ion species can be selected readily (i.e., any gas can be used), and the beam should be much more uniform and reproducible because the unreliable flashover process is eliminated.

Return to Applied Plasma Technologies homepage
Return to P-24 homepage