Sputtering crater

In another review, Magee and Honig [24] discuss three important aspects of depth profiling by SIMS depth resolution, dynamic range and sensitivity. First, the depth resolution is a measure of the profile quality. They point out that the depth resolution is limited by atomic mixing effects and the flatness of the sputtered crater within the analyzed area. Second, the dynamic range of depth profiles is limited by crater edge... 

Figure 21 Sputter crater bottoms in (100) Si after 6 keV 00+ sputtering to different depths (left) 2.1 pm (center) 2.8 pm (right) 4.3 pm. (From Ref. 77.)...

In the SIMS technique, an oxygen or cesium ion beam incident on the sample, sputters atoms from the surface. Either negatively or positively charged ions are mass analyzed and their density displayed as a function of sputter time. By using calibration standards, the density is calibrated as concentration/cm, and by measuring the sputter crater depth/ the time axis is converted to a distance axis, giving a dopant concentration vs. depth plot. 

Ball cratering is related to angle lapping,in which the sample is polished at a slight bevel. Similar measurements can also be made on the wall of a sputter crater, but without an easy conversion from a lateral distance to depth. Although these methods are destructive, they have the advantage over ion bombardment that the measurement along the crater wall can be repeated. 

Fig. 4.17 SEM micrographs of poly crystal nickel with (a) an unsputtered polished surface and with sputtered crater bottoms using (b) O2 to a crater depth of 1.5-2.2 pm, (c) O2 with 1 X 10 Torr oxygen backfill to a depth of 1.5-1.8 pm and (d) Cs to a depth of 1.9-2.5 pm. Single crystal Ni is sputtered with O2 to depths of (e) 2.9 pm and (f) 4 pm...

Any neutrals present in the primary ion beam must also be removed because this can result in nonsymmetrical sputter crater formation. Recall The trajectories of neutrals are not inlinenced by electrostatic or magnetic fields, and as a result, they may impinge on areas not specified by the primary ion beam optics. Such neutrals arise from primary ion interaction with apertures present within the primary ion colnmn (typically the last set of apertures) and/or collisions with residual gas molecules. A vacuum of at least 10 Pa should be used to ensure a dynamic range of 10 or better (Wittmaack 1982). 

Note Atomic Force microscopy (AFM), which is conunonly used to derive surface topography (AFM is discussed further in Appendix A. 11.4), is not typically used in evaluating sputter rates as sputtered craters typically cover areas that are at least an order of magnitude larger than the spatial range accessible to AFM (AFM areas are typically limited to less than 10 x 10 pm). 

Interferometric methods operate by splitting a monochromatic coherent beam (all waves in phase) of hght into two parts. One part is directed at the surface from which it then reflects and the other part directed at some reference (optically smooth mirror) from which it reflects. When the two parts are recombined, their waves will either interfere in a constractive manner (waves in phase) or destructive manner (waves out of phase). As constructive interference enhances the amplitude of the combined beam, whereas destructive interference suppresses the amplitude, an interference pattern will be prodnced that is dependent on the crater depth. This interference pattern can also be nsed to derive the surface topography of the sputtered crater in both of the spatial dimensions, with the result being an AFM-like three-dimensional topographic image. 

Since the major limitation on depth resolution arises from the sputtering process itself and from the sputter-induced microtopography, many attempts have been made to minimize such effects. The main aim is to suppress both the crystal orientation effects and the increase in roughness by rotating the sample during sputtering, as suggested by Zalar. If that is done in the correct way, then the sputter crater can be produced with a flat bottom, in the center of which is the analyzed area (79, 80]. 

In Figure 7 is a schematic of the principle underlying rotation and of the shape of the sputtering crater Fig, 8. shows sputter profiles of an Au/AI/Si02/Si multilayer sample without (a) and with (b). sample rotation. The interface, shown by the broken lines, is sharper with sample rotation. 

---