The working principle of large-area, open-gate field effect transistors (ogFETs) is attractive for the high-sensitivity detection of chemicals and interfacing with single cells. We describe an ogFET composed of a self-assembled, two-dimensional (2D) random network of 1D chains of 10 nm Au particles spanning over 25 μm. The device has a gating gain of 103-fold at room temperature (RT) compared to <50% for reported nanoparticle arrays at RT. The current, I ∼ (V – V T)ζ, is functionally identical to the Coulomb blockade (CB) effect observed at cryogenic temperatures, and the conductance gap, V T, at room temperature cannot be attributed to local charging for large particles (>5 nm). Surprisingly, unlike the effect observed in CB, the V T remains invariant over a large gating potential 0–25 V, leading to a universal behavior where all the I–V curves collapse into a single master curve. We explain the universality as a classical critical behavior by quantitatively mapping the percolation path in real-space images. The paths evolve as self-similar percolation channels in a fractal dimension of 1.88. The device principle enables a 103-fold gating gain in all-metallic nanoparticle arrays at RT and will potentially lead to ogFET sensors and electrochemical devices with liquid-gate junctions. The critical behavior with bias may serve as a model system to study the electronic transport in these exotic systems.

Electrochemical characteristics of immobilized double-stranded DNA (dsDNA) on a Au electrode were studied as a function of coverage using a home-built optoelectrochemical method. The method allows probing of local redox processes on a 6 μm spot by measuring both differential reflectivity (SEED-R) and interferometry (SEED-I). The former is sensitive to redox ions that tend to adsorb to the electrode, while SEED-I is sensitive to nonadsorbing ions. The redox reaction maxima, R max and Δmax from SEED-R and SEED-I, respectively, are linearly proportional to amperometric peak current, I max. The DNA binding is measured by a redox active dye, methylene blue, that intercalates in dsDNA, leading to an R max. Concomitantly, the absence of Δmax for [Fe­(CN)6]4–/3– by SEED-I ensures that there is no leakage current from voids/defects in the alkanethiol passivation layer at the same spot of measurement. The binding was regulated electrochemically to obtain the binding fraction, f, ranging about three orders of magnitude. A remarkably sharp transition, f = f T = 1.25 × 10–3, was observed. Below f T, dsDNA molecules behaved as individual single-molecule nanoelectrodes. Above the crossover transition, R max, per dsDNA molecule dropped rapidly as f –1/2 toward a planar-like monolayer. The SEED-R peak at f ∼ 3.3 × 10–4 (∼270 dsDNA molecules) was (statistically) robust, corresponding to a responsivity of ∼0.45 zeptomoles of dsDNA/spot. Differential pulse voltammetry in the single-molecule regime estimated that the current per dsDNA molecule was ∼4.1 fA. Compared with published amperometric results, the reported semilogarithmic dependence on target concentration is in the f > f T regime.

Circulating microRNA are promising diagnostic and prognostic biomarkers of disease in quantitative blood tests. A label-free, PCR-free, electrochemical microarray technology on a monolith electrode is described, with 10 attomolar (aM) sensitivity and responsiveness to binding of <1 zeptomole of target to immobilized ssDNA probes with zero background. Specificity is 100% in a mixture with five nonspecific miRNA each with a 103-fold higher concentration. Direct measurement on plasma-derived miRNA without cDNA conversion and PCR demonstrated multiplexing and near-ideal quantitative correlation with an equivalent pure sample. The dynamic range is a target concentration ranging from 10–2 to 103 femtomolar (fM). This PCR-free novel technology can be applied as a test for cancer diagnosis/prognosis to detect 103 copies of a miRNA sequence in RNA extracted from 100 μL of plasma.

A network of one-dimensional (1D) necklaces of 10 nm Au nanoparticles was fabricated by a directed self-assembly to synthesize 1D necklaces followed by self-limiting monolayer deposition to form a two-dimensional (2D) network array. Scanning electron microscope (SEM) image analysis revealed a percolation threshold lower than random 2D arrays signifying the local 1D structure. The topology of (shortest) percolation paths (tortuosity) and the fraction of clusters isolated from the percolating array were quantified to relate the network morphology to the observed non-Ohmic (Coulomb blockade effect) behavior. Leveraging charge contrast in SEM, the morphing of the percolation path as a function of the kinetic energy of the conduction electron was visualized and quantified to understand the dynamic nature of the percolation behavior. The morphology can be systematically tailored by tuning the two self-assembly processes to obtain the same coverage of the array with significantly diverse non-Ohmic behavior. It was concluded that tortuosity and void fraction unify the Coulomb blockade behavior for a range of fabrication conditions leading to varying network morphologies with a threshold blockade bias ranging from 0.5 to 5.5 V at room temperature. This self-assembly avenue will allow the development of highly sensitive, allmetal electrochemical field effect transistors for applications in biology.

No abstract