Quantum Tunnelling Combined With Dielectrophoresis… Single-Molecule Detection Events Amplified Up to 100,000-Fold

Abstract

Researchers at Zhejiang University in China and Imperial College London in the UK built standalone tunnelling electrodes with an average gap of 1.6 nm at the tip of a nanopipette, and used them to identify single nucleotides and proteins electrically. By combining the probes with dielectrophoretic (DEP) trapping, which uses an alternating electric field to concentrate molecules, the team got past the diffusion limit, raising event detection rates by up to five orders of magnitude (100,000-fold) and reaching sub-femtomolar (fM) sensitivity. The paper frames nucleic acid sequencing as a future prospect rather than something achieved here, and notes that at the lowest concentrations the same molecules are likely being recaptured near the probe tip and detected repeatedly. [Quantum Biology Society] Quantum tunnelling, in which an electron passes through a potential barrier it could not classically cross by virtue of its wave nature, has long served as a high-precision measurement tool, because the current varies exponentially with the width of the gap and is therefore sensitive to changes in distance at the atomic scale. Forming a gap narrower than 5 nm between nanoelectrodes and reading the characteristic tunnelling current of a single molecule passing through it has raised hopes for a next generation of nucleic acid sequencing and, potentially, even protein sequencing. Existing approaches based on the scanning tunnelling microscope (STM), however, require a conductive substrate and precise piezo controllers, which makes the system unwieldy. They also have to wait for molecules to diffuse into a gap only nanometres wide, a process that is entropically unfavourable, so the efficiency of real sample analysis has been extremely low. A collaboration led by Longhua Tang at Zhejiang University in China, together with Aleksandar P. Ivanov and Joshua B. Edel at Imperial College London, published a standalone nanoprobe platform in Nature Communications in February 2021 that addresses both limitations at once. ■ Self-Terminating Electrodeposition: Forming Nanometre Gaps With Precision The researchers laser-pulled a theta-shaped dual-barrel quartz capillary into a nanopipette whose tip terminates in two closely spaced nanopores, 25 ± 12 nm in diameter, separated by a quartz septum 15 ± 5 nm wide. Butane was then passed through the pipette and pyrolytically deposited to form two coplanar carbon nanoelectrodes, onto which gold was electrochemically plated. The key is a self-terminating mechanism that applies tunnelling current feedback during plating. The preset current is the sum of a Faradaic deposition component and a tunnelling component. As the gap between the electrodes narrows into the tunnelling regime, the tunnelling component comes to dominate, the Faradaic current falls towards zero, and deposition stops of its own accord. Of the 650 probes fabricated, roughly 85% ran through self-termination successfully. The freshly made gaps were then immersed in ultrapure deionised water for 12 to 48 hours, after which conductance had dropped by an average of 55%, consistent with a widening of the tunnelling gaps. The authors tentatively attribute this change to surface diffusion of gold atoms minimising the total interfacial free energy at a fixed volume. The resulting probes held consistent I-V characteristics over several days. Fitted with the Simmons model, 418 standalone probes had gap widths ranging from sub-nanometre to more than 3 nm, with an average of 1.6 ± 0.6 nm. ■ Verifying Tunnelling by Measuring Solvent Barrier Heights To check that the current flowing across the fabricated gaps really came from tunnelling, the researchers measured the tunnelling potential barrier of the surrounding medium. The values came out at 0.37 ± 0.21 eV for deionised water, 0.78 ± 0.14 eV for dimethyl sulfoxide and 0.97 ± 0.21 eV for hexane, all in agreement with the literature. In air the figure was 1.04 ± 0.83 eV, and the large spread was attributed to possible condensation of water vapour in the gap. In control experiments with bridged, short-circuited junctions, the I-V characteristics showed no dependence on the medium, confirming that the junctions were working as tunnelling junctions rather than through physical contact between the electrodes. ■ Identifying Nucleotides and Proteins by Their Characteristic Conductance Measuring four deoxymononucleotides with a probe of roughly 1.1 nm gap at a bias of 50 mV, the team found a clear ordering in the conductance change (ΔG): dGMP (240 ± 36 nS) > dAMP (180 ± 33 nS) > dCMP (161 ± 5 nS) > dTMP (120 ± 10 nS). The authors offer a partial, qualitative interpretation in terms of the highest occupied molecular orbital (HOMO): dGMP shows the highest conductance because its HOMO level sits closer to the Fermi level of the electrodes. They are careful to add that the actual electron transport mechanism across mononucleotides remains an open question. In a mixed solution of dTMP and dGMP, the two characteristic conductance peaks separated clearly. With a probe of roughly 1.8 nm gap, three proteins chosen for their differing molecular weights and charges were likewise told apart by clearly different conductance values: streptavidin (1.56 ± 0.19 nS), bovine serum albumin (BSA, 1.11 ± 0.21 nS) and immunoglobulin G (IgG, 0.52 ± 0.08 nS). Although the gap was narrower than the proteins themselves, characteristic tunnelling signals were still obtained. ■ Combining Dielectrophoresis: Molecular Trapping and a Five-Order Gain in Detection Rate The biggest obstacle to single-molecule sensing in a nanogap, the mass transport limit, was tackled with dielectrophoresis (DEP). Applying an AC field between the two electrodes (100 kHz, 10 Vpp, 10 seconds) generates exceptionally steep field gradients that pull target molecules in solution towards the probe tip and concentrate them there. Measurements could not be run in parallel, because applying the AC field brought a significant rise in the low-frequency noise associated with conductance fluctuations (flicker noise), along with a milder increase in higher-frequency noise attributed to capacitance. The team therefore ran DEP concentration and DC tunnelling detection sequentially. At a poly-A20 DNA concentration of 1 fM, no significant tunnelling events were seen without DEP, whereas clear spike signals appeared after trapping, and event detection rates rose by up to five orders of magnitude. That brought detection down to 0.1 fM for poly-A20 and 0.15 fM for streptavidin. ■ Significance and Limits: Single-Molecule Identification and the Recapture Mechanism The significance of the work lies in realising an ultrasensitive nanoscale tunnelling sensor that operates in solution as a standalone capillary probe, breaking away from the conventional STM architecture in which a conductive substrate is essential. The authors also state that, to their knowledge, this is the first example of dynamic control of molecular transport used in any tunnelling system. The paper treats nucleic acid sequencing as a future prospect rather than something completed here; what is demonstrated is the identification of single mononucleotides and the detection of oligomers and proteins. The authors also note that at the lowest concentrations probed, around 0.1 fM, the event rate is very high, above 200 events per second, and loses any significant dependence on concentration. They explain this as likely arising because trapped molecules localise and accumulate around the tip, so the same molecules have a much higher probability of being recaptured and interacting with the gap, and are probably detected multiple times. Read that way, in this regime the platform looks less like a quantitative counting instrument and more like an ultrasensitive qualitative test for the presence of trace molecules. #QuantumTunnelling #Dielectrophoresis #SingleMoleculeDetection #Nanoelectrode #TunnellingCurrent #Nucleotide #ProteinConductance #SimmonsModel #Nanopipette #BiomolecularSensing #QuantumBiology #NatureCommunications Source (Nature Communications): https://www.nature.com/articles/s41467-021-21101-x Follow-up study (Sci. Adv. 2022): https://doi.org/10.1126/sciadv.abm8149