Using the scanning tunneling microsocpe, we report the creation of big patches, on a Si(111) $7\times 7$ substrate, exhibiting the 2×1 reconstruction of this crystal. This can be done by a sudden increase of the sample bias, which produces big modifications of an initially flat area. The inspection of these strongly modified areas reveals a row disposal of maxima that coincides with the reported topographic and spectroscopic description of the $\mathrm{Si}\mathrm{}\left(111\right)2\times 1.$ We propose a mechanism in which a strong removal of sample material creates a contact between the tip and the sample. The breaking of this contact, if crystalline, exposes a bulk termination that reconstructs with the $2\times 1$ reconstruction as it happens when cleaving silicon single crystals in the (111) orientation.

• Received 23 December 1998

DOI:https://doi.org/10.1103/PhysRevB.59.9768