DNA origami based superconducting nanowires

The continuing quest for reducing the size of electronic components has led to the emergence of a new field of research, aiming at the study and application of molecular building blocks for the fabrication of electronic components.^1–4 In this ambitious endeavor, DNA molecules are expected to play an important role. The unique ability of DNA to self-assemble⁵ into arbitrary structures^6–12 suggests its use as a network on which various molecular electronic components can be placed.^13–15 For this purpose, numerous metallization processes have been attempted aiming to transform DNA molecules^2,16 and DNA origami^13–15 into electrically conducting wires. Various metals have been employed, including silver,¹⁷ palladium,¹⁸ and gold.¹⁹ The reported resistance values have spanned over several orders of magnitude, typically in the range of 10² Ω–10⁶ Ω.¹⁶ To increase conductivity, superconducting materials were used to fabricate superconducting nanowires, using molecules such as carbon nanotubes^20–22 and DNA^22,23 as templates. There are, however, no reports in the literature describing superconducting nanowires that use DNA origami as a template. The flexibility of DNA origami can be used as a powerful technique for fabrication of complex shaped nano-objects^9–12 that can be converted into functional superconducting devices, hence motivating the study of the physical properties of DNA origami based superconducting nanowires.

Here, we investigate the use of DNA self-assembled nanowires as templates for the deposition of superconducting NbN. We demonstrate coating self-assembled DNA nanowires with superconducting NbN and report on magneto-transport properties of the resulting nanowires. Due to the nanometric lateral size of these nanowires, their resistance does not drop to zero as temperature drops below the superconducting transition temperature, T_c. This well-known phenomenon has been observed in studies of various superconducting nanowires and has been ascribed to thermally activated phase slips (TAPSs) near T_c and to quantum phase slips (QPSs) far below T_c (for a review, see Ref. 24). The NbN nanowires also show exceptionally large negative magnetoresistance (NMR) that can be utilized to reduce a significant part of the low temperature resistance resulting from the QPS.

DNA origami nanowires were prepared as described previously in detail.²⁵ A typical TEM image of the resulting DNA origami nanowires is shown in Fig. 1(a), revealing ∼220 nm long and ∼15 nm wide wires. The DNA origami nanowires were drop-casted on a SiN/SiO₂ chip with a ∼50 nm wide channel, as schematically shown in Fig. 1(b). The channel was prepared using e-beam lithography followed by CHF₃–H₂ Reactive Ion Etching (Plasma-Therm, RIE) to remove ∼25 nm layer of SiN. In order to prevent shortcuts in the electrical measurements after NbN deposition, a ∼200 nm deep and ∼170 nm wide undercut was formed in SiO₂ by wet etching, using hydrofluoric acid (HF). The chip with the DNA nanowires was dried for 12 h in vacuum and then coated with ∼10 nm layer of NbN, using the AJA magnetron sputtering system. The chamber pressure was kept at 1.7 mTorr to avoid clustering of NbN during the sputtering process.

Figure 1(c) shows an HR-SEM image of the channel (black) with the DNA wire laid on it. The NbN coated wire and substrate are shown in bright gray color. The location of the DNA wire (which is coated with NbN and, therefore, cannot be seen in the image) is shown schematically by the orange dashed rectangle. The width of the wire after NbN deposition is ∼25 nm at its narrowest point, and as can be seen, it changes along its length. In the range of ∼30 nm, the change in the width is only few nm. In the following, the effective length of the DNA wire is taken as 30 nm.

To measure the transport properties of a single DNA nanowire, we patterned a four-probe setup on the NbN coated film, using negative tone photolithography (MLA Heidelberg Inst.), followed by Cl₂–BCl₃ etching to remove NbN around the pads and the channel. In addition, we disconnected excess DNA wires along the channel using the helium ion beam (Orion Nanolab, Zeiss), which is nondestructive to the superconducting layer.²⁶ 

Figure 3(a) shows the field dependence of the NbN nanowire resistance for temperatures between 2.2 K and 3.7 K, i.e., in the temperature range of the resistance tail. With the increase in the magnetic field, the resistance initially decreases, exhibiting a negative magnetoresistance (NMR), reaching a minimum value at a field H_min, and then continuously increases. The NMR effect is more pronounced as the temperature decreases.

Figure 3(b) shows the size of the effect, measured by the ratio $r=(RH−R(0))/R(0)$⁠, as a function of temperature, where R(0) and R(H) are the resistance at zero field and at H_min, respectively. The figure also shows the values of H_min vs temperature. Apparently, the increase in $r$ is accompanied by an increase in H_min. Although NMR has been previously reported for various superconducting nanowires, see, e.g., Refs. 37–42, we note that the effect observed here is exceptionally large, reaching a value of ∼90% at low temperatures.

In conclusion, this work demonstrates the ability to fabricate superconducting nanowires using DNA origami wires as templates. A similar “bottom-up” approach can be further adopted in fabrication of 2D or 3D superconducting nanostructures,⁴⁴ where the conventional “top-bottom” techniques, such as e-beam lithography, show their limitations. The DNA origami based NbN nanowires exhibit properties characteristic of nanowires fabricated using conventional techniques, namely, phase slips and negative magnetoresistance. Nevertheless, the NMR effect exhibited by the NbN coated DNA wires is exceptionally large, a phenomenon that should be further investigated. The results of this work can be used in various applications, including interconnects in nano-electronics and novel devices based on exploitation of the flexibility DNA origami in fabrication of 3D architectures, e.g., 3D SQUIDs⁴⁵ for the measurement of the magnetic field vector.

The idea for this experiment was first conceived during discussions with Omri Sharon. We thank Naor Vardi for growing the NbN layer and Maria Tckachev for growing the SiN layer. Y.Y. acknowledges financial support from the Israeli Ministry of Science and Technology. L.S. acknowledges the support of Bathsheba de Rothschild Fund and the Monique and Mordecai Katz Foundation. P.T. is grateful for support by the DFG excellence cluster e-conversion.