Detecting chirality-induced spin selectivity in chromophore-linked DNA hairpins using photogenerated radical pairs

Abstract

Chirality-induced spin selectivity (CISS) results in spin polarization of electrons transmitted through chiral molecules and materials. Since CISS results in spin polarization even at room temperature, it affords the possibility of using it to develop quantum technologies that can operate under ambient conditions. We have shown previously that photo-driven hole transfer within DNA hairpins provides a facile route to generate spin-correlated radical pairs (SCRPs). To study the effect of CISS on the spin dynamics of SCRPs in DNA hairpins, we prepared a series of electron donor-chiral bridge-acceptor molecules where the chiral bridge is a B-form DNA helix consisting of 4 to 6 base pairs. Naphthalene-1,8:4,5-bis(dicarboximide) (NDI) serves as the hairpin linker chromophore and electron acceptor. Photoexcitation of NDI results in rapid hole transfer through the π-stacked purine bases of the DNA and trapping of the hole on a terminal stilbene diether (Sd) to generate the NDI^•-- Sd^•+ SCRP. Time-resolved electron paramagnetic resonance spectra of the SCRPs at X- (9.6 GHz), Q- (34 GHz), and W- (94 GHz) bands show that the CISS effect imparts significant triplet character to the SCRP. We do not observe a significant dependence of CISS on DNA length, likely resulting from hole delocalization over the guanine bases in the G-tract. Interestingly, we find that the CISS contribution significantly increases with magnetic field strength. These findings should be considered in any future modeling of CISS.

Keywords: CISS; chirality; electron transfer; radical ion pair; spin dynamics.

Fig. 1.

(A) Structures of the NDI chromophore/electron acceptor and the Sd terminal electron donor. (B) Hairpin structures.

Fig. 2.

(A) UV-Vis absorption of D-1G and L-1G in buffer (100 mM sodium chloride and 10 mM sodium phosphate) at room temperature. (B) CD spectra of D-1G and L-1G in buffer (100 mM sodium chloride and 10 mM sodium phosphate) at room temperature.

Fig. 3.

fsTA and nsTA spectra of D-1G (A and B), respectively, and L-1G (C and D), respectively, following 355 nm excitation in buffer (100 mM sodium chloride and 10 mM sodium phosphate) at room temperature at the indicated pump–probe delay times.

Fig. 4.

(A) X-band and (B) Q-band TREPR spectra of D-1G, D-2G, and D-3G in 50% glycerol/50% buffer (100 mM NaCl and 10 mM Na₃PO₄) 85 K, 100 ns after a 355 nm, 7 ns laser pulse. The smooth curves overlaying the experimental spectra are simulations using the parameters given in SI Appendix, Table S4. The field was frequency-corrected.

Fig. 5.

(A) Q-band and (B) X-band TREPR spectra of D-1G in 50% glycerol/50% buffer (100 mM NaCl and 10 mM Na₃PO₄) 85 K, 100 ns after a 355 nm, 7 ns laser pulse (black circles). The smooth curves overlaying the experimental spectra are simulations using the parameters given in SI Appendix, Table S4 with CISS contributions of 0% (blue curve), 100% (red curve), or the optimal fit (black curve).

Fig. 6.

W-band EDFS of D-1G in 50% glycerol/50% buffer (100 mM NaCl and 10 mM Na₃PO₄) 85 K, 150 ns after a 355 nm, 7 ns laser pulse (black circles). The smooth curves overlaying the experimental spectra are simulations using the parameters given in SI Appendix, Table S4 with CISS contributions of 0% (blue curve), 100% (red curve), or the optimal fit (black curve).