TY - JOUR
T1 - Molecules on silicon
T2 - Self-consistent first-principles theory and calibration to experiments
AU - Rakshit, T.
AU - Liang, G. C.
AU - Ghosh, A. W.
AU - Hersam, M. C.
AU - Datta, S.
PY - 2005/9/15
Y1 - 2005/9/15
N2 - In this paper based on a fully self-consistent first-principles transport calculation we show that in a silicon-molecule-STM (scanning tunneling microscope) structure, negative differential resistance (NDR) occurs for positive substrate voltage, mediated by the molecular levels on p-type Si(100) substrates. The positions of the NDR peaks are determined by (i) the equilibrium location of the relevant level with respect to the Si(100) Fermi energy (Ef) and (ii) how fast the level slips past the band edge. Based on (ii), we predict that by varying the STM tip-to-molecule spacing, the NDR peak location can be shifted in the current-voltage (I-V) characteristics. Recent experiments indeed show the NDR peak movement as a function of tip distance, thus strongly supporting this molecule-mediated mechanism of NDR on p-type substrates. Extrapolation from the NDR peak shift on the voltage axis can be used to predict the equilibrium location of the molecular level with respect to the Si(100) substrate Fermi energy. Based on (i), we conclude that to observe NDR in the negative bias direction on n-type substrates, much higher voltages are required. However, polarity-reversed NDR on n-type substrates may be observed under conditions beyond the model described here, and the relevant scenarios are discussed.
AB - In this paper based on a fully self-consistent first-principles transport calculation we show that in a silicon-molecule-STM (scanning tunneling microscope) structure, negative differential resistance (NDR) occurs for positive substrate voltage, mediated by the molecular levels on p-type Si(100) substrates. The positions of the NDR peaks are determined by (i) the equilibrium location of the relevant level with respect to the Si(100) Fermi energy (Ef) and (ii) how fast the level slips past the band edge. Based on (ii), we predict that by varying the STM tip-to-molecule spacing, the NDR peak location can be shifted in the current-voltage (I-V) characteristics. Recent experiments indeed show the NDR peak movement as a function of tip distance, thus strongly supporting this molecule-mediated mechanism of NDR on p-type substrates. Extrapolation from the NDR peak shift on the voltage axis can be used to predict the equilibrium location of the molecular level with respect to the Si(100) substrate Fermi energy. Based on (i), we conclude that to observe NDR in the negative bias direction on n-type substrates, much higher voltages are required. However, polarity-reversed NDR on n-type substrates may be observed under conditions beyond the model described here, and the relevant scenarios are discussed.
UR - http://www.scopus.com/inward/record.url?scp=29744451341&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=29744451341&partnerID=8YFLogxK
U2 - 10.1103/PhysRevB.72.125305
DO - 10.1103/PhysRevB.72.125305
M3 - Article
AN - SCOPUS:29744451341
SN - 1098-0121
VL - 72
JO - Physical Review B - Condensed Matter and Materials Physics
JF - Physical Review B - Condensed Matter and Materials Physics
IS - 12
M1 - 125305
ER -