[ Home ]

Industrial Applications of
Low Temperature Plasma Physics

        Plasmas are gases that are so hot that electrons have come off the atoms, forming a negatively charged fluid which is mixed with the positively charged ions.  The entire plasma is electrically neutral, but its behavior is complex because the particles' motions are controlled by electric and magnetic fields.  Though the main applications of plasma physics are to space exploration and to controlled fusion (reproducing on earth the energy source of stars), the aspect that affects our everyday lives is the use of plasmas in manufacturing.  For example, metals, such as propeller blades, are hardened with a plasma surface treatment; plastics, such as automobile bumpers and potato chip bags, are plasma treated for better paint adhesion; camera lenses have optical coatings applied with a plasma; and synthetic diamonds can be made with a plasma arc.  The most important practical use of plasmas is, however, in the manufacture of semiconductor circuits, particularly computer chips.

        In making a computer chip, which may contain as many as 100M transistors, the feature sizes have to be smaller than 0.13mm.  Purely chemical etching is much too crude; plasma etching is required.  In processing a silicon wafer, perhaps half of the steps require plasma processing, and the machines required for this account for over half the cost of the factory.  These plasma "reactors" were developed originally by trial and error without the help of plasma physicists.  In recent years there has been a continuous improvement in the accuracy and efficiency of these reactors, and it is becoming more clear how they work. 

        In UCLA's Low Temperature Plasma Technology Laboratory (LTPTL), experimental and theoretical work is done to understand the physical mechanisms in plasma generators so that one can design better ones.  For instance, the helicon source, which we have been studying for a decade, can produce denser plasmas with less radiofrequency (RF) power.  The reasons for its superiority, however, were mysterious, and the understanding of this source led to many fascinating research results.  One ongoing project is to make an array of helicon sources which can produce a wide plasma capable of etching a large substrate uniformly, as shown by the diagram on the left.   Click on the thumbnail to expand the image, and again on the icon at the lower right of the image for full screen.  Use the BACK button on your browser to resume.

        Another project has to do with Inductively Coupled Plasmas (ICPs), which are commonly used today.  These machines do not need a magnetic field, as do helicon sources, and it is a puzzling how the RF energy gets into the center of the plasma to produce a uniform density, when the RF field should be shielded by the plasma.  By studying these sources in a laboratory device such as shown on the right, we have discovered that nonlinear effects can cause a number of energetic electrons to reach the interior and produce plasma there.  An interesting problem is how an ICP turns into a helicon source as a small magnetic field is applied.  Nothing much happens until the field is large enough to support a helicon wave.  A computer code, HELIC, has been developed to predict this behavior and enable the design of future helicon sources.  

        In making transistors smaller and smaller, one also has to scale the thickness of the insulating oxide layer on the "gate" down to less than 50 Angstroms.  Etching these oxide layers is problematic, because plasma processing can cause large voltages to build up across these gates and ruin their electrical properties.  A mechanism called Electron Shading Damage was postulated some time ago to explain why this happens.  To test this hypothesis, we have constructed an experiment and a computational code to model the effect on a macroscopic scale.  As shown by this diagram, we plan to measure the charging currents in a structure of mm size rather than the sub-micron size of real semiconductor features.  To do this required making a very tenuous plasma source so that the plasma sheath thickness scales along with the feature size. 

        The physics of partially ionized plasmas has been considered a "dirty science" because adding neutral atoms to ions and electrons can complicate the problem so that the solutions are messy.  Our aim has been to show that low-temperature plasma physics need not be low-brow.  There are challenging problems with elegant solutions.  As an added bonus, this line of research has an impact on devices and materials which we use everyday.

        For further information, please visit the website for the LTPTL.