This protocol is crucial in the technology of semiconductors for defining the density and chemistry of dislocations, and consequently establishing the nature of structural defects in the as-grown structures. And the method allows three dimensional localization of low concentration impurities in solid-state materials, making it possible to relate their position to certain structural defects. Before attempting the technique, familiarize yourself with the instrument, perform multiple beam stability tests, and determine what circumstances prolong the stability period.
Try working with lower-than-usual current. Demonstrating the procedure will be Iwona Jozwik, a leading SEM specialist from my laboratory. Begin by preparing a eutectic mixture of potassium hydroxide, sodium hydroxide, and magnesium oxide.
Dissolve and mix the composing alkali hydroxides and metal oxide in distilled water and heat the mixture in a flask on a hot plate to 200 degrees Celsius for one hour with a magnetic stirrer. Cool the mixture to about 100 degrees Celsius by reducing the temperature of the hot plate until the remaining liquid completely evaporates. Then transfer the solid etchant into a dried bottle while avoiding exposure to moisture.
For defect-selective etching, place a gallium nitride sample on a 450 degree Celsius hot plate along with a thermocouple to precisely read the real temperature. Then place a piece of solid etchant on top of the gallium nitride and keep it there for three minutes. Remove the sample from the hot plate and place it in a beaker with hot hydrogen chloride for three to five minutes to eliminate any remaining solid etchant.
Transfer the sample to a beaker with deionized water, and expose it to an ultrasonic bath for 5 to 10 minutes, then dry it with nitrogen blowing. Use a diamond pen cutter to mark the sample with an L-shaped scratch and mount it on a metal stub with a conductive adhesive, making sure to use gloves to avoid grease contamination from hands. Add a piece of the conductive tape to connect the sample surface with the metal stub to prevent charge buildup on the specimen surface.
Acquire at least three high resolution SEM micrographs of a top view of the sample with each image displaying an area of at least 25 by 25 micrometers. Avoid taking images from the surface regions with macroscopic surface defects. Calibrate the SIMS equipment using negative polarity, cesium primary ions with 7 to 13 kilo electron volt impact energy, and align the secondary and primary beams.
Keep the beam as small as possible. Prepare five to seven settings for beams with various ion current density. For simplicity, keep the size of the beam intact and change the beam current.
Measure the beam current and size of the beam. Use a 50 by 50 micrometer raster size and a 35 by 35 micrometer analysis area. Choose 256 by 256 pixels for spatial resolution.
If not specified otherwise, use a standard integration time for each signal, typically one to two seconds. Choose a setting with a moderate beam current and obtain a series of images using a 30 silicon(2)anion secondary ion for a blank silicon wafer. For each image, integrate the signal for 5 to 10 minutes.
If the system does not allow longer integration times, select 60 seconds. After obtaining 200 images, group five images into one to further analysis and run the measurements. Perform pixel-to-pixel comparisons of all images with the first image.
If greater than 5%of pixels show a greater than 5%difference from the first image, this indicates that the beam became unstable. Note the time span of beam stability. Perform measurements within a stable time span of the beam.
Using the same beam settings, perform at least five measurements for each beam setting. Obtain a depth profile using a 16 oxygen anion secondary ion, reach a 200 nanometer depth, and measure the intensity of 69 gallium anion secondary ion by integrating the signal for 10 to 15 seconds. Do not perform this in regions where SEM images have been obtained.
Plot the intensity ratio of 16 oxygen anion and 69 gallium anion signals as a function of the inversed primary current density, and estimate the vacuum background contribution. Then choose an intense beam and obtain an image that will be used for flat field correction. Use a 30 silicon(2)anion secondary ion for a blank silicon wafer.
Integrate the signal for 5 to 10 minutes. Perform depth profile measurements in the same regions where SEM images have been obtained. Using a 16 oxygen anion secondary ion, integrate the signal for three to five seconds for each data point.
This protocol can be used to obtain realistic 3D distributions of impurities or dopants in solid state materials. As the reduction procedure is performed, 90%of counts are randomly eliminated from each layer. Very clear pillar-shaped structures are observed in the final 3D image.
A typical result for a single plane is shown here. If the core is smaller than the size of a primary beam, the secondary image will inherit the size and shape of the primary beam. In suboptimal experiments, a random distribution of oxygen counts can be seen.
In certain situations, the beam becomes unstable during the experiment. Specifically, the quality is high for a region close to the surface, but gradually deteriorates during the experiment. A stable beam is required to perform this experiment.
The beam is typically most stable after it has been switched on, so running the experiment for roughly two to three hours after starting the beam is the best option. Sometimes it is better to work faster, even if the depth resolution gets worse. This technique makes it possible to detect and precisely localize low concentration impurities.
It opens up possibilities to study the chemistry of various structural defects.
