Ion Implantation and Surface Modification
Ion implantation is a major application of plasma processing in a variety of applications in which the surfaces of materials are to be treated. The implantation process requires a source of ions and a means to accelerate them toward the surface. Two general methods are in use today: ion beam implantation, in which a beam of ions is directed toward a substrate, and plasma implantation, in which the ions produced in a plasma discharge surrounding or near the object to be implanted are extracted from the plasma and accelerated into the object.
Ion implantation is designed to modify the surface properties of materials without changing their bulk properties. The implantation process may offer improvements in their properties or may actually be used to degrade the surface, depending on the application. It is becoming economically attractive in Japan, Europe, and the United States and can be done on metals and alloys, as well as on semiconductors, ceramics, insulators, and polymers. Some of the surface properties that can be modified by this process are hardness, fatigue, toughness, adhesion, wear, friction, corrosion oxidation, dielectric properties, magnetic properties, superconductivity, resistivity, and catalysis.
In general, both the substrate material and the implanted species, if it can be ionized, can cover a wide range of substances. As a result, a vast array of ions can be implanted into an equally wide range of substrates. Furthermore, both implantation processes tend to operate at low pressure so that (1) a vacuum chamber is required for the workpiece and (2) collisions between charged particles and other species present in the vacuum chamber, for example, neutral particles, free radicals, electrons, and so on, tend to be minimized. The result is that the implantation process is physical, rather than chemical, in nature. However, significant numbers of chemical reactions may often occur within the substrate, since the particle density in the substrate is very high and the ions travel a comparatively short distance, thus making their concentration high near the surface.
As an ion enters the surface of a material, it collides with atoms and interacts with electrons. Each nuclear or electronic interaction reduces the energy of the ion until it finally comes to rest within the target. Typically, interactions follow a statistical process, and the implanted profile is often approximated by a Gaussian distribution as follows:
N(x) = N p exp[−(x − Rp)2 / 2ΔRp]
where the average distance an ion travels before it stops is called the projected range Rp. The peak concentration Np occurs at a range of Rp. Because of the statistical nature of the process, some ions will obviously penetrate beyond the projected range Rp and some will not travel as far as Rp. The spread of the distribution of the implanted ions is characterized by the standard deviation ΔRp and is called the straggle. The area under the Gaussian distribution curve is the implanted dose Q,
If the implant is contained entirely within the target, and the distribution is Gaussian,
The implanted dose can often be controlled to within a few percent. In addition, doses in the range of 1010 to 1018 cm2 are needed for many applications and are almost impossible to achieve by a thermal diffusion process in many applications. Range and straggle are roughly proportional to ion energy.
In conventional ion implantation devices, an ion beam is extracted from a plasma source, accelerated to the desired energy, and then transported to the target. Typical beam currents are very small (in the microampere range) and the beam “footprint” area is less than 1 cm2. To process large-scale targets, and to avoid shadowing if the target is nonplanar, a combination of beam rastering and target manipulation during the process is required.
In the plasma source ion implantation (PSII) process, the object is immersed in a plasma in which the Debye length is much smaller than the dimensions of the object. The strongest electric field is then in the cathode sheath, which accelerates positive ions to the negatively biased object, which serves as a cathode. If the pressure is kept low enough to prevent an arc discharge, positive ions can be accelerated to energies of 100 kV or more and can be implanted into the cathode surface.
To control the process, the implantation voltage is pulsed. The process begins with the application of a high negative potential to the object relative to the vacuum chamber wall. As the potential of the surface becomes more and more negative, the sheath, or the region surrounding the surface from which the electrons have been expelled, expands into the plasma, reflecting the electrons ahead of it. Being more massive, the ions do not have time to move as the boundary sweeps through them. When the ions find themselves on the other side of the boundary, they are in a region of a strong inward electric field, which accelerates them to the cathode.
Currently, the factors and their interactions that influence this process are poorly understood from the standpoint of the behavior of the material that is implanted. To further advance this field, an understanding of how the bombarding particles interact with the base material and how to apply this knowledge to manufacturing techniques is required for industry to further exploit this technology. Three aspects of this technique, surface hardening and wear resistance, corrosion and oxidation resistance, and semiconductor applications, are ideally suited for the use of statistical methods for experimental design and, in particular, of response-surface methods for exploring interaction effects. These are crucial tools for further developments in this area.
To be applied successfully in industry, many of the applications of ion implantation must be demonstrated to be cost-effective. Here two classes of implantation applications—metals and dielectrics—are discussed. The inherent differences in metallic, as compared with dielectric, implantation need to be considered. In both cases, the substrate requires an applied bias voltage for ion acceleration, but dielectrics will undergo charging and therefore inherently degrade the ion acceleration sheath, thus modifying the implantation. Accordingly, the acceleration voltage will be applied either to a backing conductor or to a conducting metallic, graphitic, or diamond-like carbon film deposited on the substate, which will also act to inhibit its charging.
Recent work using this process has shown remarkable improvement in properties of nitrogen-implanted alloy die steel and of nitrogen-implanted aluminum tools for machining high-temperature alloys. The effects of implanting nitrogen into a surface previously enriched, for instance with carbon and/or boron by vapor deposition, are now being examined.
This process is called ion beam enhanced deposition (IBED), and either ion beam or plasma implantation can be used. Ultimately, this work will be extended to cover the effect of incident nitrogen ion energy (on depth of hardening), concomitant fatigue properties, characterization of the microstructure by transmission electron microscopy or atomic force microscopy, and mechanical properties versus depth by nano-hardness measurements. Such studies are a prerequisite to control of the manufacturing process and require statistical experimental design techniques for their implementation.
It has been conclusively demonstrated that ion implantation can beneficially modify the surface-sensitive mechanical properties of steels. Fatigue life has been extended by as much as a factor of 2, the coefficient of sliding resistance reduced by as much as a factor of 100, and the wear resistance increased by a significant amount. Most studies have concentrated on the use of nitrogen ions, but the use of C, B, Ti, and Mo ions has also shown promising results. When PSII is used in large-scale manufacturing, such as of automobile parts, practical problems arise that may have scientific solutions. For instance, secondary emission from the highly negative object can constitute a dominant fraction of the power that the circuit must supply. This not only increases the cost but also creates a heat problem. Control of secondaries and of unipolar arcs is a familiar problem to plasma physicists.
A very useful application of this process is in the implantation of medical prostheses. For instance, artificial hip joints with complicated shapes have been implanted with nitrogen ions for wear and hardness improvement to increase the lifetime of such devices.
Four major classes of nonmetallic materials are amenable to treatment by PSII: glasses, ceramics, polymers, and semiconductors. Extensive examination of the effectiveness of PSII for substrate modification is envisaged: magnetic domain formation in glasses, erbium doping of polymer waveguides, lattice structure modification of mica, dopant implantation into semiconductors, and surface property modification of plastics and polymers for thin film deposition and for wear improvement of gears and drive units.
The deposition of thin metallic coatings on polymeric surfaces has many important industrial and research applications, for example, in flexible electronic circuits, sensors, electromagnetic shielding, and flexible reflecting surfaces. Historically, however, adhesion of metallic coatings on polymers has suffered from unreliable bonding and delamination. One theory for this frustrating situation is that, in the vacuum environment used for metal-coating deposition, water trapped within the polymer is released, oxidizing the surface of the metal at the metal/polymer interface, thus leading to delamination. Recent research experience indicates that significant adhesion improvement for metallic coatings on polymer surfaces can be realized through the use of IBED. In this process, the metal coating deposition is performed simultaneously with low-energy (< 1 keV) ion beam irradiation. High-energy beams appear to yield results inferior to those of lower-energy beams in this application, presumably because of excessive polymer chain scission and surface nitriding.
Ion implantation of dielectrics, especially polymers, has led to dramatic improvements in their hardness. For example, after implantation, the surfaces of some polymers may become harder than stainless steel, although the process may actually be a result of carbonization of the polymer rather than by the formation of new compounds by implantation itself.
In semiconductor applications, ion implantation is a major component in microfabrication. Typically, beam implantation is used; this often imposes a lower limit on the available ion energies, making it difficult, for example, to produce the shallow-depth implants needed for the next generation of microelectronics. Plasma implantation, however, has been shown to produce shallow implants. In addition, the potential of silicon-
on-insulator (SOI) technology using separation by oxygen, separation by implantation of nitrogen, or separation by implantation of oxygen and nitrogen will permit the manufacture of semiconductor devices on a thin silicon film mechanically supported by a thin insulating substrate. Present technologies limit the thickness of the silicon wafer used for microchip manufacturing to more than 500 µm to avoid unwanted electrical coupling, but only the first few micrometers at the top of the wafer is used for most transistor fabrication. With the use of SOI technology, this limit can be dramatically reduced.
A ROLE FOR NRL
Implantation in industry can take many forms ranging from individual “job shops” to tool and component manufacturers to users with on-site implantation machines. As a result, it is important for NRL to consider how implantation technology can impact all three of these aspects.
NRL can use its expertise in high-energy beams and can build on its existing strengths in this area. It is important to consider how this expertise can be used to expand the applications; these should include IBED, which itself can be used for a much wider range of applications. The energy ranges of the NRL implantation systems should be extended to lower and higher energies. Optimization of the implantation process is extremely important for economic viability, that is, for the shortest possible processing time.
Current research on ion implantation has revealed a number of scientific and technical problems that can serve as examples of areas in which NRL can make a contribution. These areas include design of efficient pulse modulators for high power delivery, control of secondary emission, processes in plasma and sheath formation, computation of ion trajectories near complex boundaries, electric fields in implantation of dielectric materials, implantation of mixtures of ions, numerical simulation of surface kinetics, and implantation and doping of semiconductors.
It is of course not sufficient to understand the physical processes and demonstrate that the implantation can take place; an economic assessment is also required to show feasibility. It is important to concentrate on industrial applications that have a potential for commercialization. NRL should facilitate collaborative activities with ongoing implantation groups that are specifically oriented toward industrial applications.