The Na-atom-tip model

Lang s result from the numerical theory is in excellent agreement with the general theory in the previous section. If both tip state and sample state are. 9-wave states, the tunneling conductance should have a spherical distribution, and the apparent radius should equal the nominal distance between the two nuclei. The radius obtained from the figure fits well with this expectation. 

Similar to the. y-wave model, the Na-atom-tip model predicts a poor resolution. The agreement of the Na-atom-tip model with the y-wave-tip model does not mean that the s-wave-tip model describes the actual experimental condition in STM. According to the analysis of Tersoff and Lang (1990), real tips are neither Na or Ca, but rather transition metals, probably contaminated with atoms from the surface (for example. Si and C are common sample materials). For a Si-atom tip, the p state dominates the Fermi-level LDOS of the tip. For a Mo-atom tip, while the p contribution is reduced, this is more than compensated by the large contribution from states of d like symmetry. The STM images from a Si, C, or Mo tip, as predicted by Tersoff and 

Lang (1990), are dramatically different from the images predicted by the, 9-wave-tip model. 

---

The original 5-wave-tip model described the tip as a macroscopic spherical potential well, for example, with r 9 A. It describes the protruded end of a free-electron-metal tip. Another incarnation of the 5-wave-tip model is the Na-atom-tip model. It assumes that the tip is an alkali metal atom, for example, a Na atom, weakly adsorbed on a metal surface (Lang, 1986 see Section 6.3). Similar to the original 5-wave model, the Na-atom-tip model predicts a very low intrinsic lateral resolution. 

---