As electronic devices shrink in size, they become more susceptible to scattering from defects such as dangling bonds and traps. While this may seem deleterious, their high sensitivity also allows precise single-molecule characterization and the possibility of engineering of these surface defects. Our group has pioneered the idea of using noise signatures from such traps to characterize them precisely, identifying their unique fingerprints or `bar-codes', as well as signatures of interactions between them.

Trap formation and RTS from single molecular defects

We are studying electronic characterization of single-molecule defects using the concept of surface-modulated field effect transistors (`SurFETs'). Attaching a molecule creates localized `trap' states in the band-gap of the silicon. Resonating with those traps using a back-gate creates a stochastic filling and emptying of these traps. Since the change in trap population blocks the channel through Coulomb interactions, the current shows a concomitant flicker known as random telegraph signals (RTS). While RTS has been studied since the 60s, the novelty is their cleanness (since in the nanoscale we can imagine just one or two defects that may even be engineered), and their impact on the current (since even one defect can occlude a nanoscale channel such as a nanotube). While the noise looks messy, the power spectral density shows clear peaks with corner frequencies given by the trap lifetimes, while the peak values of these signals are reached upon resonance with a gate. We can thus imagine real-time detection of bonded molecular states using a scalable CMOS platform.

GNRBANDS




Extracting Molecular `bar-codes' from trap-induced noise

IVGNR

Since the RTS is very sensitive to the precise trap location in the silicon band-gap, and since the traps are very narrow spectrally (being in the band-gap), we can compute a molecule-specific `bar-code' or fingerprint that shows the placement of the molecular levels within the band-gap. A library of such fingerprints will allow us to identify molecular species with ultra-fast electronic detection techniques.






Blocking and unblocking of channel currents with interacting traps



Geometry

In experiments at UVa, we found that two defects in the oxide near a nanotube FET produced large (80% base current at room temperature) RTS signals. Furthermore, there is a unique window carved out in the ambipolar nanotube I-VG curve, where the current goes through an RTS sequence leading into a blockade, and then subsequently the current fully unblocks by going out of the window through a separate RTS sequence. Our theoretical models, coupling many-body rate equations for the dots with mean-field quantum transport (NEGF) in the channel show that the blockade happens when one trap blocks the channel, while the second trap blocks the first trap and thus unblocks the channel. The physics is aided by the ease with which the two trap levels slide past each other under bias from a distant gate. The slide is enhanced significantly by the 1-D electrostatics around the cylindrical nanotube.






Extracting trap interaction dynamics from electronic noise spectroscopy

Geometry

By extracting the spectral function from the multi-state noise and multiplying by the frequency to eliminate the 1/f tail, we get a nice spectral snapshot of the trap and channel states -- the frequency axis spells out their lifetimes while the voltage axis spells out their energetic locations. Since we are seeing a snapshot of all states near the Fermi energy of the channel (assumed to stay close to equilibrium), we can see when states go above or below the Fermi energy and thus map out dynamics of interactings between the traps.




Multiphonon processes to fill the traps



Since the traps are in the band-gap, they are filled and emptied by non-equilibrium phonon emission and absorption. This makes their effective density-of-states frequency dependent and allows them to be filled from the other bands at deep level.`



PUBLICATIONS

  • "Modeling electrostatic and quantum detection of molecules", S. Vasudevan, K. Walczak, N. Kapur, M. Neurock and A. W. Ghosh, IEEE-Sensors Vol. 8 , 857 (2008).

  • "Reversal of current blockade through multiple trap correlations", J. Chan, B. Burke, K. Evans, K. A. Williams, S. Vasudevan, M. Liu, J. Campbell and A. W. Ghosh, Phys. Rev. B Vol. 80, 033402 (2009)..