ALBUQUERQUE, NM — Using a low-power laser similar to those at supermarket checkout counters and a half-dollar-sized piece of specially coated optical glass, a simple, reliable method to determine stresses in very thin films of materials used in advanced electronic, optical and magnetic devices has been developed by scientists at Sandia National Laboratories.
Thin films and their support structures are critical components of CD-ROMs, semiconductor chips, computer hard drives, and telecommunication devices.
Sandia is applying for a patent on the inexpensive technology and licensing it on a non-exclusive basis to industry.
Developed by physicist Eric Chason with colleague Jerry Floro, the method, though highly sensitive, is impervious to vibrations caused by thin-film processing equipment or the rumble of passing trucks.
From a single laser beam, the coated glass creates an array of parallel beams aligned like the teeth on a comb. By monitoring small deflections of the beams after they bounce off a target, its warp and hence its stress can be measured.
The new capability makes it possible for researchers to measure stress as it builds in a noisy environment rather than interrupt the process to remove the product for off-line analysis.
Stresses develop during production due to, for example, different rates of contraction between films and support materials during cooling.
Thin-film stress may cause cracking, buckling or even delamination, limiting applications of the film. Yet some stress may be necessary to optimize performance — for example, in lasers and detectors with precisely-controlled optical properties.
“Stress can be harmful or useful, depending on the application — but in either case, we need to measure it before we can control it,” says Chason.
To calibrate the amount of stress in thin films, Chason and Floro monitor the curvature of the supporting material to determine its warp while varying the temperature or thickness of the film.
By beaming a one-half milliwatt laser into an angled, partially reflective mirror coated on both sides, called an etalon (here, not used as an interferometer), the researchers create a series of internal reflections which emerge as parallel beams of light. The beams are directed into the process chamber and onto the material surface over a region of about one centimeter. The spacing between reflections of the beams are monitored electronically using a simple video camera. Distortions in the material as small as one-hundredth of a micron — a micron is about the size of a bacterium — can be measured by the change in distance between the reflected beams. The changing distance becomes a continuous measure of warpage and thus, stress.
Instrument jiggle never appears in the readings because the rigid connection between laser and etalon causes all the beams to move together.
Earlier methods scan a laser light sequentially at different points of the material and then measure the distance between reflections. Because the measurements are taken at different times, these methods are more vulnerable to vibrations.
According to Harvard physicist Frans Spaepen, “What Eric has done is to measure all the laser points simultaneously. By doing so, he removed a significant source of error.”
In a further improvement, the Sandia researchers place a second etalon rotated at right angles to the first. This piggyback arrangement creates a square array of beams from the single source, making possible simultaneous measurement over a square area rather than just along a line, and allowing analysis of shapes as complex as those on a potato chip.
A uniform substrate and a uniformly distributed surface film are two necessary components for the method. The method was presented at the Spring meeting of the Materials Research Society meeting in San Francisco in April.