Addressing the properties of “Metallo-DNA” with a Ag(i)-mediated supramolecular duplex

The silver-nucleoside complex [Ag(i)-(N3-cytidine)2]+, 1, self-assembles to form a supramolecular metal-mediated base-pair array highly analogous to those seen in metallo-DNA.

Emission spectroscopy (in the range of 300 -800 nm) in a gel or liquid state were measured using a Shimadzu RF-6000 Spectro Fluorophotometer using a quartz cuvette (50 μl) with a path length of 1.5 mm. Scan intervals were acquired every 0.5 nm at a scan speed of 600 nm/min, within an emission and excitation bandwidth of 5.0 nm. The system was controlled using LabSolutions RF software.
Circular dichroism spectra were measured in a high precision quartz SUPRASIL (Hellma Analytics®) cell with a light path of 0.1 mm. Data was acquired using a JAS.CO J-810 spectropolarimeter under nitrogen (5-10 ml/min) connected to a PTC-423S temperature controller maintaining 25°C using a Julabo water bath.
For high-resolution electrospray ionisation mass spectrometry (ESI-MS) studies, xerogel samples were mixed in ultrapure (type 1) water at concentrations in the order of 100 ng ml -1 . Measurements were made on a Waters Micromass LCT Premier TOF Mass Spectrometer and data was processed using the Mmass software (version 5.5.0).

Experimental 1; Synthesis and Crystallographic Data.
Synthesis of [Ag(N3-cytidine) 2 ][NO 3 ], (1) Cytidine (200.0 mg, 0.88 mmol, 2.0 equiv.) was dissolved in water (4.5 ml). To this solution, another aqueous solution (0.5 ml) of AgNO 3 (74.7 mg, 0.44 mmol, 1.0 equiv.) was added slowly. The mixture was heated under reflux (100 °C) for 2 h in the dark. The solution was filtered while still warm; Long needle-like crystals were grown over a period of 14-days at room temperature using a thermal vapor diffusion chamber with acetonitrile as anti-solvent. The yield of crystals was 67%. 1      The crystallographic study of 1 was hampered by a number of factors. Available crystals were too small and weakly diffracting for investigation using standard in-house facilities, even with enhanced microsource X-ray intensities. The weakness of the diffraction was subsequently found to be exacerbated by extensive structural disorder and pseudo-symmetry. After repeated attempts an adequate set of single-crystal diffraction data was obtained at beamline I19 of Diamond Light Source, 1 providing a reliable unit cell geometry and a clear indication of the space group (one of an enantiomeric pair), which had been uncertain from earlier attempts, possibly because of some merohedral twinning.
Standard multiscan corrections were applied for absorption and for incident beam intensity variations. 2 The positions of Ag atoms were readily found by a variety of routine direct and heavy-atom methods. These, however, are arranged in a regular and almost linear chain along the crystallographic c axis, with 5 Ag atoms in the asymmetric unit. This leads to a dominance of the diffraction pattern by strong reflections with l = 5n. Use of all available routine methods for structure solution failed due to the pseudo-symmetry of the helical structure of the polymeric cation chain, which was subsequently found to display approximate non-crystallographic fivefold screw axis symmetry; the threefold crystallographic screw axis relate cation chains to each other. The structure was eventually solved with successive manual selection of possible ligand atom positions from the multiple images in electron density maps insufficiently phased by the silver atoms, relying on the recognition of fragments of cytidine rings.
To aid refinement, geometrical similarity restraints were applied to the bond lengths and angles of the 10 ligands in the asymmetric unit, together with similarity and rigid-bond restraints on their anisotropic displacement parameters. 3 H atoms were included in geometrically calculated positions and refined with a riding model; those on OH groups were allocated to give a reasonably self-consistent set of positions appropriate for hydrogen bonds where these could be identified. Reliable positions could not be found for the nitrate counteranions or solvent molecules, which occupy large apparent voids between cation polymer chains; these were treated by the SQUEEZE procedure of PLATON. 4 Nitrate anions, but not solvent molecules, are included in the chemical formula and calculated physical properties.
Resonant ('anomalous') scattering effects of the silver atoms are significant and provide confirmation of the absolute configuration of the overall structure, consistent with that known for the ligand sugar component, 5 giving a refined value of 0.08(4) for the absolute structure parameter.

Rheology
Rheological measurements were performed with a HR-2 Discovery Hybrid Rheometer (TA Instruments) with a standard steel parallel-plate geometry of 20 mm diameter with a gap of 1 mm. The strain and the frequency were set to 1% and 1 Hz, respectively. Atomic force microscopy (AFM). A Multimode 8 atomic force microscope with Nanoscope V controller and a "E" scanner (Bruker, Germany) was used for acquiring AFM height images. Nanoscope software version 9.1 was used to control the microscope. The system was operated in a peak force tapping mode in air (ScanAsyst) at ultra-low forces to minimise damage to the Xe1 fibres. An isolation table/acoustic enclosure (Veeco Inc., Metrology Group) was used to decrease vibrational noise. Silicon tips on V-shaped silicon nitride cantilevers (ScanAsyst-Air, Bruker) were used for imaging. The nominal tip radius was approximately 2 nm, resonant frequency 70 kHz, and spring constant k ~0.4 N/m. The AFM height images were analysed with NanoScope Analysis 1.5 software (Bruker). The xerogel sample Xe1 was prepared by drop-casting 1 µl of 1 MeOH gel onto a clean silicon wafer. The sample was kept in a 9 ml closed glass vial for 3 days to allow the self-healing of the gel. Gel transferred to the silicon surface by micropipette led to loss of the fibrous structure in keeping with the thixotropic nature of the gel. If the 1 MeOH complex was dried quickly on a silicon surface immediately after it was transferred, it appears in the form of a film of dried material in the AFM images. AFM Height Analysis -The mean diameter of the fibres was estimated using AFM images of the Xe1. First, the distribution of heights in the image was determined using Gwyddion 1 with a bin width of 0.188 nm; it is important the bin width is much smaller than the typical feature size. This distribution is shown in Figure 12(a). Next, a pseudo-Voigt function was fitted to the measured distribution by a standard least-squares method and the residual was plotted in Figure 12(b). Clear peaks (labelled 1-6) are seen in the residual which are a result of unresolved components in Figure 12(a) originating from the stacking of fibres in the AFM image. The mean fibre diameter can then be obtained by linear regression of the heights at the peaks of Figure 12(b) against an integer index. This is shown in Figure 12(c) and the fibre diameter was estimated to be 6.3 +/-0.45 nm.

Conductivity Measurements, IV and EFM.
Electrical measurements (I-V). Current/voltage curves were recorded on the probe station (Cascade Microtech with a B1500A parameter analyser, Agilent). All of the electrical measurements were carried out under dry nitrogen without light illumination. Platinum MBEs (Smart Microsystems Pt MB-4000, Windsor Scientific Ltd., Slough, UK) were used to fabricate electronic devices for electrical characterisation of Xe1. The MBEs were made on Si/SiO 2 substrates. Four independent platinum electrodes were patterned on the top of the SiO 2 layer. The height of the electrodes is 200 nm and their width is 10 µm with 10 µm spaces between them. The surfaces of the MBEs were electrically insulated except for a 2 × 2 mm 2 area for depositing the gel. The platinum MBEs were washed with ethanol and dried with nitrogen gas. The clean platinum electrodes were analysed on a probe station and reference current/voltage curves were recorded, which showed the background currents to be less than 100 fA at 2 V. In the case of the xerogel a drop (0.5 µl) of 1 MeOH was cast onto the platinum MBEs and the sample was kept in a closed vial to allow the self-healing of the gel and then left to dry. The gel droplet dried to produce a film of the Xe1 fibres across the Pt-electrodes. The electrodes were connected to the probe station and current/voltage curves were collected. For single crystal measurements, a crystal of 1 was electrically insulated on a glass slide and the crystal were connected either directly to probe needles or the ends of the crystal were coated with Ga-In eutectic and probe needles were attached to these pads. In both cases the probe needles were positioned at opposite ends of the long crystal axis. Both methods for contacting the crystal gave essentially similar results (see Figure S13a and S13b). Figure S13. I-V curves of (a) the Xe1 (red) and an oriented single crystal of 1 directly contacted to probe needles (blue), and (b) data for an oriented single crystal of 1 contacted to probe station tips through Ga-In eutectic (green) as shown in (e). Optical images of (c) microband electrodes used for I-V measurements of xerogel Xe1, (d) a single crystal of 1 directly contacted to probe needles (blue) and (e) a single crystal of 1 contacted to probe needles through Ga-In eutectic (green).

Electrostatic Force Microscopy (EFM).
Electrostatic force microscopy (EFM) operated in the dynamic non-contact mode provides a useful contactless tool for qualitative testing of the conductivity of a single nanowire prior to any quantitative measurement of its conductivity. 1,2 The method is based on the storage of energy in the tip/nanowire/substrate capacitor and the nanowire conductance influences the measurement via the RC time constant for polarisation of the nanowire. For EFM studies, the nanowires are aligned on Si/SiO 2 chips with an oxide thickness ~200 nm and the phase shift between the driving force and the tip motion is recorded as a function of applied dc bias (V) as the tip crosses above the fibre at a constant lift height. As shown by Staii and co-workers, 3 a negative phase shift with a V 2 applied bias dependence is an indication of a conductive nanowire. Objects which are merely polarisable and do not have charges which can move away from the vicinity of the tip produce a positive phase shift; this conclusion is independent of the details of the quantitative model of the effect.
Both EFM phase images and the corresponding topographic images were acquired using the Multimode 8 atomic force microscope (Bruker, Germany). Data were analysed using NanoScope Analysis 1.5 software. EFM was performed in air using AFM tips on n-doped Si cantilevers (MESP-LM-V2, Bruker), both of the probe sides were coated with Cobalt/Chromium layer. The nominal tip radius was approximately 25 nm, resonant frequency 75 kHz, and spring constant k ~3 N/m. The EFM phase images has been captured at applied DC bias between +10 and -10 volts with a scan lift height of 30 nm. The sample of 1 MeOH gel was prepared as above and diluted 20 times with further methanol. The sample for analysis was prepared by drop-casting 1 µL of the diluted cytidine-Ag(I) solution onto a Si<100>/200 nm SiO 2 slide. The Si slides were sonicated in methanol for 15 min, dried with nitrogen gas and then exposed to oxygen plasma (Femto -diener electronic, Plasma-Surface-Technology) for 20 min at 80 W. The slides were washed again with methanol and dried with nitrogen gas. The slide with the drop-cast gel was surrounded by 10 µl of methanol in a closed glass vial (with a volume ~ 9 ml) for 30 min. Then, it was left to dry in air for 10 min.
A control experiment has been carried out to confirm that the fibres in the AFM and EFM images do not arise from any solvent impurity. A Si<100>/200 nm SiO 2 slide was sonicated gently in a sonication bath in methanol for 15 min, dried with nitrogen gas and then exposed to oxygen plasma for 20 min at 80 W. The slides were washed again with methanol and dried with nitrogen gas. A drop of methanol (2 µl) was drop-casted onto the Si/SiO 2 slide and left to dry in air. The sample was scanned by AFM at different locations. The AFM images on this sample have showed no fibres on the surface of the Si/SiO 2 . Figure S14. AFM image of a control sample that has been prepared by drop-casting 2 µl of methanol onto a Si<100>/200 nm SiO 2 slide highlighting the absence of fibres from solvent.

Reduction Behaviour
Photoreduction was achieved by suspending vials of 1 MeOH (2 ml) within a Southern New England Ultraviolet Company Photochemical reactor, using eight RMR-3000 (50/60 Hz, 120 Volts, 300 nm) UV bulbs. Exposure times ranged from 1 -90 minutes. Figure S16. An image showing a series of 1 MeOH samples exposed to UV-light (300 nm) for up to 90 minutes.

Band Structure Calculations
All spin-polarized density functional theory (DFT) calculations were performed with the Vienna ab-initio Simulation Package (VASP) 1 and the projector augmented wave (PAW) method. 2,3 The exchange-correlation potential was described by the Perdew-Burke-Ernzerhof (PBE) functional 4 extended to incorporate a Van der Waals correction. 5 We use a k-point gird of 1 × 1 × 5, so that there are 5 k-points in the transport direction.