Angle dependence of SIMS measurements
In a representative study on the aforementioned hard material films, the quantification error as a function of the impact angle α and the take-off angle β was determined in a range of approximately 20° - 60°, where 45° is the standard angle for flat specimens.
It was found that the raw counting rates for Cr, Ti, Al, W, and N vary by a factor of 5-30 with the angle, the count rates of H, C on the other hand considerably less. The dependence on the take-off angle β is similar in size. For quantification using so called relative sensitivity factors (RSF), the ratio of the counting rates for the elements to the matrix element carbon (C) or nitrogen (N) respectively is determined. This value is proportional to the chemical composition. Thus its variation with the angle α and β indicates the possible quantification error. The graphic below shows exemplarily the error for the H/C ratio of a DLC film and the Cr/N ratio of a CrN film. Especially with H, an angle error of 10° can lead to quantification errors of up to 20 percent which means e.g. that 25at% H instead of 20at% H is detected. The relative error is usually lower for nitride film systems (10%).
The studies show that extremely careful adjustment of the measuring position is of special importance with curved samples. With cylindrical specimens, the ion beam has to be adjusted precisely to the cylinder’s highest horizontal point. Based on the results, for known, well-defined tilt angles, for example a blade cutting edge, the angle dependence can be corrected using the known angle dependence.