Breaching the Fortress: Photochemistry of DNA-Caged Ag₁₀⁶⁺

Abstract

A DNA strand can encapsulate a silver molecule to create a nanoscale, aqueous stable chromophore. A protected cluster that strongly fluoresces can also be weakly photolabile, and we describe the laser-driven photochemistry of the green fluorophore C 4 AC 4 TC 3 GT 4 /Ag 10 6+ . The embedded cluster is selectively photoexcited at 490 nm and then bleached, and we describe how the efficiency, products, and route of this photochemical reaction are controlled by the DNA cage. With irradiation at 496.5 nm, the cluster absorption progressively drops to give a photodestruction quantum yield of 1.5 (±0.2) × 10 –4 , ∼10 3 × less efficient than fluorescence. A new λ abs = 335 nm chromophore develops because the precursor with 4 Ag 0 is converted into a group of clusters with 2 Ag 0 – Ag 6 4+ , Ag 7 5+ , Ag 8 6+ , and Ag 9 7+ . The 4–7 Ag + in this series are chemically distinct from the 2 Ag 0 because they are selectively etched by iodide. This halide precipitates silver to favor only the smallest Ag 6 4+ cluster, but the larger clusters re-develop when the precipitated Ag + ions are replenished. DNA-bound Ag 10 6+ decomposes because it is electronically excited and then reacts with oxygen. This two-step process may be state-specific because O 2 quenches the red luminescence from Ag 10 6+ . However, the rate constant of 2.3 (±0.2) × 10 6 M –1 s –1 is relatively small, which suggests that the surrounding DNA matrix hinders O 2 diffusion. On the basis of analogous photoproducts with methylene blue, we propose that a reactive oxygen species is produced and then oxidizes Ag 10 6+ to leave behind a loose Ag + -DNA skeleton. These findings underscore the ability of DNA scaffolds to not only tune the spectra but also guide the reactions of their molecular silver adducts.

Fluorescence Screening of DNA-AgNCs with Pulsed White Light Excitation

“…The purpose of designing new DNA-AgNCs based on the machine learning models is to find bright emitters that can be excited and emit in the Vis and NIR region for sensing and imaging purposes. ,− Additionally, steady state approaches do not allow for discriminating between ns- and μs-lived emission, leading to potential misclassification of the training data for dual emissive DNA-AgNCs. ,,,,, …”

“…Depending on the sequence, one or multiple types of DNA-AgNCs can be formed with photophysical properties that can vary greatly (e.g., emit green/red fluorescence or have a ns/μs decay time; Figure a). Since dual emissive DNA-AgNCs with short- and long-lived bands have been reported with increasing frequency, we speculated that some of the multiband emitters disregarded in previous color-predicting machine learning models could actually have been single DNA-AgNCs with short- and long-lived bands. ,,,,, Accordingly, the 20 sequences studied here were selected from a list of DNA-AgNCs previously identified by Copp et al to yield multiband emission . We performed our synthesis in Eppendorf tubes at a modest scale of 100 μL and stored the samples in the fridge between measurements.…”