Inelastic Electron Tunneling Spectroscopic Analysis of Bias‐Induced Structural Changes in a Solid‐State Protein Junction

Abstract

A central issue in protein electronics is how far the structural stability of the protein is preserved under the very high electrical field that it will experience once a bias voltage is applied. This question is studied on the redox protein Azurin in the solid‐state Au/protein/Au junction by monitoring protein vibrations during current transport under applied bias, up to ≈1 GV m−1, by electrical detection of inelastic electron transport effects. Characteristic vibrational modes, such as CH stretching, amide (NH) bending, and AuS (of the bonds that connect the protein to an Au electrode), are not found to change noticeably up to 1.0 V. At >1.0 V, the NH bending and CH stretching inelastic features have disappeared, while the AuS features persist till ≈2 V, i.e., the proteins remain Au bound. Three possible causes for the disappearance of the NH and CH inelastic features at high bias, namely, i) resonance transport, ii) metallic filament formation, and iii) bond rupture leading to structural changes in the protein are proposed and tested. The results support the last option and indicate that spectrally resolved inelastic features can serve to monitor in operando structural stability of biological macromolecules while they serve as electronic current conduit.

What Can We Learn from Protein-Based Electron Transport Junctions?

“…Underlying NDR is quantum tunneling across a barrier as this phenomenon is associated with getting the junction in and out of energetic resonance: in order to observe NDR, the applied voltage aligns empty (full) semiconductor (in our case protein) energy levels with full (empty) electrode levels, making them equi-energetic (ON resonance) or misaligning them (OFF resonance). The fact that NDR has been observed in the conductance–voltage characteristics of junctions of small proteins (∼2 nm) 35 further supports the operation of ETp by tunneling across these junctions.…”

Biotin Binding Hardly Affects Electron Transport Efficiency across Streptavidin Solid-State Junctions

“…The fabrication procedure of the suspended gold-nanowire (AuNW)-based, non-destructive, top-electrode contact has been described in detail elsewhere. , Here, we slightly modified the procedure to increase the fraction of junctions where only a single AuNW is trapped within an array of microfabricated electrode devices (based on a chip developed by Noy et al) from ∼10 to ∼35% of all possible junctions on the chip. First, monolayers of the n -STV and t -STV proteins (or their respective biotin complexes) were prepared on the chip’s microelectrodes (bottom-electrode) as described above.…”

“…Based on the bias-dependent current responses, some junctions are classified as non-contact (current <10 pA @0.5 V), or shorted (current >1 mA @0.5 V), and the remaining are considered as active ones (10 pA < I < 50 μA @0.5 V). However, most junctions that pass >50 nA currents at 0.5 V were found to be shorted after a few scans and are considered to be partially shorted . The junctions with lower currents (30 pA < I < 50 nA @ 0.5 V) were noticeably less noisy than junctions with higher currents.…”

“…Protein layer defects and possible variations in the dielectrophoretic force distribution may result in some penetration of the AuNW (during the trapping) through the soft-protein monolayer, which can change the effective electric contact area and reduce the separation between top and bottom electrodes. One important cause is that the applied AC field between the reference and working electrodes will vary somewhat from one electrode to another due to differences in the distance between them in the macroscopic (∼5 × 5 mm 2 ) chip (Figure S2). The nanowire trapping force (dielectrophoretic) is directly proportional to the square of the applied electric field .…”

“…Both electrostatic and covalent binding of the protein to the substrate electrode were used for monolayer preparation. First, protein monolayer formation was optimized on the Au/Si substrate; the optimized procedure was then used to deposit protein monolayers on microfabricated Au electrodes. , The Au/Si substrate was first cleaned by bath-sonication, sequentially in acetone, isopropyl alcohol, and double-distilled water for 10 min each, followed by drying in a dry N 2 stream. Next, the N 2 -dried substrate was cleaned by UV-generated O 3 for 10 min.…”