Inﬂuence of tip-surface interactions and surface defects on Si„100… surface structures by low-temperature „5K… scanning tunneling microscopy D. Riedel, M. Lastapis, M. G. Martin, and G. Dujardin Laboratoire de Photophysique Mole ´culaire, Ba ˆt. 210, Universite ´ Paris Sud, 91405 Orsay, France Received 10 November 2003; published 8 March 2004 The Si100 surface structures on n-type degenerately doped samples (  0.005  cm) have been investi- gated with a scanning tunneling microscope STM at very low temperature 5K. We have developed a method to monitor quantitatively the proportion of the various observed surface structures  p (2 2), c (4 2) and ﬂickering. This study has been performed as a function of the tunnel current and the presence or not of surface defects in the observed areas. The normal surface areas having a low density of defects 1% have been observed to vary from the p (2 2) to the c (4 2) structures when the tunnel current increases. This indicates that the STM tip-surface interaction strongly inﬂuences the observed structures. Furthermore, surface areas completely free of any defects are dominated by ﬂickering structures. DOI: 10.1103/PhysRevB.69.121301 PACS numbers: 68.35.Bs, 68.37.Ef Over the past 20 years, the atomic structure of the Si100 reconstructed surface has been the subject of intense experi- mental and theoretical work. The ﬁrst low-temperature scan- ning tunneling microscope STM experiment at 120 K by Wolkow 1 nicely conﬁrmed that the ﬂip-ﬂop motion of the Si dimers which occurs at room temperature is frozen at this temperature giving rise to a c (4 2) reconstruction. It was expected that the c (4 2) structure would remain the most stable structure down to very low temperature. However, quite surprisingly, recent STM experiments have shown that new reconstructions such as symmetric dimers 2–3 or static p (2 2) structures 4 can be observed at very low temperature 10 K. There have been several controversial discussions concerning the origin of these new reconstructions. 3–6 Very recently, Sagisaka et al. 6 suggested that the scattered elec- trons issued from the STM tip might be responsible for the observed transition from the c (4 2) to the p (2 2) reconstruction. 7 This structure manipulation has been evi- denced by plotting the evolution of the STM topographies as a function of the surface voltage and tunnel current. Never- theless, we have observed that such surface structure modi- ﬁcations occur when the imaging of the same area is repeated even though the surface voltage and tunnel current are kept constant. Under such conditions, it is difﬁcult to demonstrate the inﬂuence of the STM tip interaction with the surface, only by showing various images recorded at different surface voltages and tunnel currents. In this paper, we propose a statistical approach to clarify this structure manipulation with the STM tip. The Si100n-type, As doped surface after cooling down to 5 K has been observed under a positive and negative surface voltage. For the positive surface voltage, we have counted the proportion of silicon dimers observed in a c (4 2), p (2 2) or ﬂickering surface structure relative to the total number of dimers observed in the scanned area. By plotting this proportion of each of these structures structure probability as a function of the tunnel current, we ﬁnd that the p (2 2) structure shows a strong tendency to be trans- formed into the c (4 2) structure when the STM tip surface interaction increases on normal surface areas defects 1%. Furthermore, we demonstrate that the ﬂickering structure dominates the surface areas that are completely free of de- fects. The inﬂuence of these two effects; the STM-tip surface interaction and surface defects, makes the study of this sur- face reconstruction a very difﬁcult task. The experiments have been performed with an ultrahigh vacuum UHV low-temperature LT scanning tunneling microscope STM. 8 It is composed of a load-lock chamber, a preparation chamber, and a STM chamber. The STM UHV chamber is equipped with a four-liter liquid helium cryostat bath to which the whole STM is connected. This ‘‘beetle’’ type 9 STM is surrounded with a double radiation shield in- side the STM chamber. The external shield is cooled with a liquid nitrogen bath while the internal shield is cooled with the liquid helium bath. The shields are pierced on the sides with removable windows or shutters permitting different functionalities: observation, laser irradiation, or molecular deposition. A front shutter permits samples and tips to be transferred via a cooled manipulator from the preparation chamber. The temperature inside the STM is measured at two points with two silicon diodes DT-470, Lake shore. The temperatures indicated in this paper correspond to the tem- perature of the base plate of the STM situated near the sample holder. It indicates the temperature of the sample, after stabilization of the cooling process, with a precision of 0.25 K in the range 2–100 K. The preparation chamber can receive samples or tips from the load-lock chamber and is equipped with a thermal sensor on its manipulator and elec- trical connection to sample holders or tip holders. The samples used in this experiment are prepared in an ultrahigh vacuum chamber with a base pressure of 7 10 -11 Torr. They are Si100 n-type As doped with a resistivity  0.004– 0.006  cm and a thickness e 100  m. The preparation of the silicon sample starts by resistively heating to 650 °C for a predegassing period of about 12 h. The sample holder is then cooled to 6 K with a liquid He cir- culation through the manipulator while the heating of the sample is kept ﬁxed. We then proceed to rapidly ﬂash the sample as explained in other works, 10 with a slow decline in temperature from 950 to 750 °C. The sample is then cooled back down to 6 K and transferred into the STM chamber at this temperature. This procedure allows us to obtain very RAPID COMMUNICATIONS PHYSICAL REVIEW B 69, 121301R2004 0163-1829/2004/6912/1213014/$22.50 ©2004 The American Physical Society 69 121301-1