Tracking Photoinduced Charge Separation in DNA: from Start to Finish

Tracking Photoinduced Charge Separation in DNA: from Start to Finish. Frederick D. Lewis* , Ryan ... Publication Date (Web): August 2, 2018. Copyright...
1 downloads 0 Views 2MB Size
Download PDF
Article Cite This: Acc. Chem. Res. 2018, 51, 1746−1754

pubs.acs.org/accounts

Tracking Photoinduced Charge Separation in DNA: from Start to Finish Frederick D. Lewis,* Ryan M. Young, and Michael R. Wasielewski*

Downloaded via UNIV OF KENTUCKY on August 21, 2018 at 09:13:57 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Department of Chemistry and Institute for Sustainability and Energy at Northwestern, Northwestern University, Evanston, Illinois 60208-3113, United States CONSPECTUS: The initial studies of the dynamics of photoinduced charge separation conducted in our laboratories 20 years ago found strongly distance-dependent rate constants over short distances but failed to detect intermediates in the transport of positive charge (holes). These observations were consistent with the single-step superexchange or tunneling mechanism that had been observed for numerous donor− bridge−acceptor systems at that time. Subsequent studies found weak distance dependence for hole transport over longer distances in DNA, characteristic of incoherent hopping of either localized or delocalized holes. The introduction of synthetic DNA capped hairpin constructs possessing hole donor and acceptor chromophores (or purine bases) at opposite ends of a base-pair domain made it possible to determine the time required for transit of charge from one chromophore to the other and, in some cases, to distinguish between the transit time and the much faster initial charge injection time. These studies eliminated conventional tunneling as a viable mechanism for charge transport in DNA except at very short donor−acceptor separations; however, they did not establish the presence or nature of intermediates in the charge separation process. Recent studies in our laboratories have succeeded in identifying key intermediates as well as untangling the dynamics and efficiency of the charge separation process from start to finish. The dynamics of the initial charge injection process is dependent upon both its free energy and the stacking of the hole donor chromophore and adjacent purine base. The transport of positive charge (holes) over multiple base pairs in duplex DNA occurs most efficiently via repeating adenine bases, known as A-tracts. The transit time across an A-tract is strongly dependent upon the free energy for hole injection, whereas the efficiency of charge separation depends on the competition between charge delocalization and charge recombination in the contact radical ion pair. The guanine cation radical has been detected both by femtosecond transient absorption and by stimulated Raman spectroscopies when the guanine is located near the chromophore employed for hole injection into an A-tract. Replacement of guanine by its derivative 8-phenylethynylguanine (EG), permits tracking of hole transport across longer poly(purine) sequences as a consequence of the stronger transient absorption and stimulated Raman scattering for EG+• vs G+•. We have recently obtained evidence based on femtosecond transient absorption spectroscopy for the formation of delocalized A-polarons in Atracts possessing four or more A-T base pairs. Similar methods have been used to track hole transport across less-common DNA structures including diblock and triblock poly(purines), locked nucleic acids, three-way junctions, and G-quadruplexes. Similar methods are have been applied to the study of photoinduced electron transport in DNA.

■

INTRODUCTION The mechanism of photoinduced charge separation in DNA continues to be an area of active investigation, 25 years after Barton and Turro reported rapid charge transport (CT) between metallo-intercalators separated by ca. 8 DNA base pairs.1 Initial studies of the mechanisms of CT in DNA employed indirect methods such as fluorescence quenching or oxidative strand cleavage at guanine bases to infer the distance and base-sequence dependence of the CT process.2 Studies by the groups of Barton, Schuster, Giese, Saito, and others, which combined DNA base-sequence design and the principles of physical organic chemistry served to elucidate many of the basic features of DNA CT. These studies have been the subject of several collected volumes and comprehensive reviews.3−5 Seeking a direct method to determine the dynamics and efficiency of the initial events in photoinduced charge separation in DNA, the Lewis and Wasielewski groups applied © 2018 American Chemical Society

femtosecond transient absorption (fsTA) spectroscopy to the study of chromophore-linked hairpins, whose synthesis was introduced by our late colleague Robert Letsinger.6 The crystal structure of a Letsinger-type hairpin having a stilbenediether linker (Sd, Chart 1a,b) is shown in Figure 1.7,8 Notable features of this structure are the excellent π-overlap between the Sd chromophore and the adjacent base pair and the B-form structure of the base pair domain. Our initial studies of photoinduced electron transfer employed DNA hairpins possessing a stilbenedicarboxamide hairpin linker (Sa, Chart 1a) and a poly(A-T) base pair domain having a single G-C base pair separated from Sa by 0−4 A-T base pairs (Chart 1c, Sa-AnG). Rate constants for photoreduction of 1*Sa to its anion radical Sa−• (kcs) and return of Received: March 1, 2018 Published: August 2, 2018 1746

DOI: 10.1021/acs.accounts.8b00090 Acc. Chem. Res. 2018, 51, 1746−1754

Article

Accounts of Chemical Research

electron donor (hole acceptor); however, no spectroscopic evidence was obtained for formation of G+•. The absence of transient intermediates other than 1*Sa and Sa−• and the linear dependence of both log(kcs) and log(kcr) on the number of AT base pairs separating Sa and G-C led us to propose that photoinduced charge separation and charge recombination occurred via a single-step tunneling mechanism in which charge does not reside on the (A-T)n bridge (Scheme 1,

Chart 1. Structures of (a) Stilbenediether and -diamide Linkers (Sd and Sa), (b) a Sd-Linked Hairpin Whose Crystal Structure Appears in Figure 1, (c) Sa-Linked Hairpins Having a T6A6 Stem or a Single G-C Base Pair in the First or Fifth Base Pair Position, and (d) Sa/Sd Capped Hairpins Having Three and Six A-T Base Pairs

Scheme 1. Tunneling and Hole Hopping Mechanisms for Charge Separation (kcs) and Charge Recombination (kcr) in Sa-AnG Hairpins Showing the Occurrence of Tunneling When n ≤ 4 and Hopping When n ≥ 4a

a The increase in Sa−•-An+• energies with charge separation reflects increasing electrostatic attraction.

dashed arrow).10,11 Direct evidence for the formation of a charge-separated state with an intervening base-pair domain awaited the development of capped hairpins having Sa hole donor and Sd hole acceptor chromophores whose anion and cation radicals could be distinguished by fsTA spectroscopy.12−14 The maximum in the transient absorption spectrum of Sd+• is blue-shifted from that of Sa−• (535 vs 575 nm) making possible observation of both radical ions (Figure 3). Kinetic analysis of the ratio of the 535/575 nm absorption maxima provides the charge separation rate constants (hole arrival times) for the capped hairpins (Figure 2). The values of kcs are similar to those obtained previously for Sa with G as the hole trap.

Figure 1. X-ray structure of the hairpin Sd-BrU (Chart 1b). (left) Side view with Sd in green at top. Adapted from ref 7. (right) Top-down view with Sd in gray stacked with guanine (blue) and the A-tract in yellow (pdb 1PUY). Adapted with permission from ref 8. Copyright 2003 American Chemical Society.

Sa−• to the neutral Sa ground state (kcr) decreased as the number of A-T base pairs separating Sa and the G-C base pair increased (Figure 2).9 We assumed that guanine served as the

Figure 2. Rate constants (s−1) for charge separation in hairpins (Chart 1b) having n A-T base pairs separating Sa and G (open black squares), Sa and Sd (filled black squares), naphthalenediimide (NDI) and Sd (filled red circles), and diphenylacetylenedicaboxamide (DPA) and G (filled blue triangles).

Figure 3. Femtosecond transient absorption spectra of the Sa-A5Sd capped hairpin with n = 5 at indicated delay times following femtosecond laser excitation. Reproduced with permission from ref 13. Copyright 2006 VCH-Wiley. 1747

DOI: 10.1021/acs.accounts.8b00090 Acc. Chem. Res. 2018, 51, 1746−1754

Article

Accounts of Chemical Research

Figure 4. (a) Energetics for hole injection from selected linker chromophores into adjacent purine bases and dynamics for charge injection (τci) and charge recombination (τcr). For Sa-A6, τdf is the decay time for delayed fluorescence. (b) Singlet energies and reduction potentials of the chromophores obtained from fluorescence spectra and electrochemical data, respectively. (c) Oxidation potentials of purine bases in aqueous solution.

■

PHOTOCHEMICAL HOLE INJECTION In our initial fsTA investigation of the hairpin Sa-(A-T)6, the absence of a well-resolved band for Sa−• and the observation of unquenched 1*Sa fluorescence were attributed to endergonic hole injection into adenine.9 Subsequent reinvestigation of this hairpin using both time-resolved fluorescence and fsTA spectroscopy with global analysis established that hole injection into the A-tract does in fact occur but that it is reversible with an equilibrium constant favoring 1*Sa (Figure 4a).19 Reversible hole injection accounts for both the observation of delayed fluorescence and the formation of Sa−• upon hole arrival at G in the Sa-AnG hairpins.19 When GC is the adjacent base pair, rapid and irreversible exergonic charge separation occurs followed by fast charge recombination (Figure 4a). The results obtained for several other singlet electron acceptors in hairpins having (A-T)6 stems are shown in Figure 4a. Rate constants for charge injection (kci = τci−1) and charge recombination (kcr = τcr−1) depend primarily upon the free energy changes for these processes, which can be estimated using eqs 1 and 2, which are based on a theory due to Weller.20

In the case of Sa/Sd capped hairpins possessing a small number of intervening A-T base pairs (n < 3), the conversion of 1*Sa to Sa−• and Sd+• both occurred within the ca. 200 fs time resolution of our fsTA instrumentation; however for larger Sa−Sd separations, the formation of Sa−• occurred prior to the formation of Sd+• (Figure 3). This led to the proposal that photoinduced charge separation occurred via tunneling at short distances but by charge injection followed by multistep hopping mechanism (hole arrival) at longer distances (Scheme 1).15,16 A distance-dependent change in mechanism from tunneling to hopping had previously been proposed on the basis of strand cleavage studies by Giese.17 Rate constants of 1.2 × 109 s−1 and 4.3 × 109 s−1 for hole hopping in extended Atracts and G-tracts, respectively, were obtained from kinetic analyses of our hole arrival data.16,18 Strand cleavage and transient absorption measurements provided a basic understanding of the mechanism and dynamics of hole transport in DNA. However, the studies outlined above did not resolve some of the fundamental questions concerning the photoinduced charge separation process. These include the nature of the initial intermediates and the kinetics of their formation and decay. For example, does the electronically excited electron-acceptor chromophore undergo single step tunneling to yield a base-pair separated radical ion pair? Is the cation radical formed upon hole injection into a purine base localized on a single base or delocalized over several bases? Does the presence of a shallow hole trap between the donor and acceptor chromophores inhibit the hole transport process? Finally, and of greatest importance, can the intermediates in the hole transport process be directly detected and characterized? The objective of this Account is to describe the answers to these questions that we have obtained in our recent collaborative studies.

ΔGci° ≈ NA{e[E1/2(D+ /D) − E1/2(A/A−)] − δ} − ΔΕο , ο (1)

ΔGcr° ≈ −NA{e[E1/2(D+ /D) + E1/2(A /A− )]

(2)

where NA and e are Avogadro’s number and the elementary charge, E1/2(D+/D) and E1/2(A/A−) are the half-wave electrode potentials of the purine base and oxidant chromophore, ΔEo,o is the singlet energy of the chromophore (Figure 4b),21 and δ is a solvent-dependent term that corrects for the difference in energy between the contact radical ion 1748

DOI: 10.1021/acs.accounts.8b00090 Acc. Chem. Res. 2018, 51, 1746−1754

Article

Accounts of Chemical Research

Figure 5. Dynamics and efficiency of charge transport for four chromophore linkers. Hole injection kinetics shown in Figure 4 are omitted for clarity. Quantum yields for charge separation are calculated from the charge separation and charge recombination rate constants of the C−•A6+• charge-separated state (C = linker chromophore).

pair and free radical ions.22 The values E1/2(D+/D) are those measured by Jalilov et al.23 in aqueous solution (Figure 4c) rather than the values for A and G measured earlier by Seidel et al. in acetonitrile.24 The large value of E1/2(D+/D) for A vs G measured in acetonitrile is inconsistent with the observation that a single guanine base does not block hole transport in an A-tract base domain (see below).25 In the case of chromophores whose singlet excited states are not energetically capable of hole injection to an adjacent adenine (e.g., phenanthrene-2,7-dicarboxamide and perylene3,9-dicarboxamide hairpin linkers and an intercalated acridine), singlet quenching by G or by the more readily oxidized purine donor deazaguanine (Figure 4c) is only observed with a small number of intervening A-T base pairs (≤3) and is strongly distance dependent, suggestive of a tunneling mechanism (Scheme 1).21,26 Deviations from a simple energy-gap dependence for kci or kcr for different chromophores may reflect either different hairpin structures or different extents of charge separation in the radical ion pair formed upon charge injection. For example, the slow rate constant for PDI-A charge injection may reflect a large reorganization energy for the large hydrophobic surface of PDI.

The fastest A-tract hole transport rate constants that we have observed are for the NDI-An-Sd capped hairpins (Figure 2).14 The distance dependence of the rate constants for the NDI hairpins is much weaker than that for the Sa hairpins, the ratio of rate constants attaining a value of >300 for hairpins with 6 intervening A-T base pairs (Figure 5). Plausibly this difference is a consequence of the much larger driving force for hole injection with the NDI chromophore (ca. 1.0 eV). Investigation of A-tract hole transport dynamics using chromophores that are weaker hole donors than NDI but stronger than Sa might clarify the effect of driving force. We note that Fujitsuka and Majima have observed that the driving force for electron injection into a poly(T) electron transport sequence similarly determines the rate constant for T-to-T hopping.29

■

DIRECT OBSERVATION OF THE GUANINE AND PHENYLETHYNYLGUANINE CATION RADICALS Guanine has the lowest oxidation potential of the common nucleobases (Figure 4c) and, as mentioned in the Introduction, has been employed as a hole trap in studies of photoinduced electron transfer and as an intermediate in multistep long-distance hole transport processes. The formation of the cation radical G+• has been studied by several spectroscopic methods including transient absorption,30,31 vibrational,32 and electron paramagnetic resonance (EPR) spectroscopies.33 However, the contact and base-pair-separated radical ion pairs formed upon photooxidation of G in duplex DNA had not been detected by transient spectroscopy prior to our recent studies.25,34 This failure had been attributed variously to the short-lifetime and broad transient absorption spectra of G+• and its spectral overlap with the much more intense spectra of the excited singlet and anion radical of the chromophores employed for hole injection. Faced with the problem of spectral overlap with the reported absorption spectrum of G+•, we reinvestigated the excited-state behavior of the DPA-linked hairpins (Chart 2), which we first investigated nearly two decades ago.35 The singlet excited state and anion radical of DPA have relatively narrow absorption bands (Figure 6) and 1*DPA undergoes efficient hole injection with both neighboring A-T and G-C base pairs.34 The 525 nm band formed upon 330 nm laser excitation of DPA-AnG is assigned to 1*DPA. It decays and is replaced within several picoseconds by a sharp band at 500 nm and a broad band at 1130 nm, both of which are assigned to DPA−•. A weaker but

■

DYNAMICS AND EFFICIENCY OF HOLE TRANSPORT AND TRAPPING Kinetic analysis of the time-dependent transient absorption spectra of the Sa-An-Sd capped hairpins provided the first values for both the hole arrival times and quantum yields for hole transport and trapping by Sd (Figure 5). Results obtained for several other chromophores as hole donors with either Sd or G as hole acceptors are summarized in Figure 5. Values of kcs for the DPA-An-G hairpins shown in Figure 2 are similar to those for Sa-An-G and Sa-An-Sd, decreasing sharply with increasing values of n for n < 4 before attaining a more constant value when n > 4. The efficiency of charge separation for DPA-A6-G is 0.28, the largest we have observed for an Atract of this length. This suggests that the energetics of both the hole injection and hole transport steps may influence the efficiency of charge separation (Figure 5). No hole transport is observed for either PDI-A6-Sd or PDI-A6-G, presumably as a consequence of rapid charge recombination in the contact radical ion pair.19 In the case of Sa-Gn-Sd capped hairpins no hole transport is observed for n > 3,27 presumably as a consequence of poor stacking of G bases in G-tracks having three or more guanines as well as fast charge recombination.28 1749

DOI: 10.1021/acs.accounts.8b00090 Acc. Chem. Res. 2018, 51, 1746−1754

Article

Accounts of Chemical Research

excitation with much slower time resolution.36 The short lifetime of our G+• 575 nm band can be attributed to fast charge recombination with DPA−•. Assignment of this band to the contact- and chargeseparated radical ion pairs DPA−•-G+• and DPA−•-A1G+• was corroborated by femtosecond stimulated Raman spectroscopy (FSRS). The FSRS spectra of all of the DPA hairpins display bands that match those previously assigned to DPA−•.37 However, only the spectra of DPA-G and DPA-A1G (Figure 6, top) display additional bands consistent with the formation of G+•.38 The appearance of fsTA and FSRS bands characteristic of G+• was also observed upon laser excitation of the minihairpins DPA-G2 and DPA-G3 (Chart 2c).39 The absence of G+• transient absorption or Raman bands for DPA-AnG hairpins with n > 1 is somewhat surprising in view of efficient hole transport to G for n = 2 and 3 (Φcs = 0.85 and 0.61, respectively).34 Similarly, bands associated with the formation of G+• are not observed in the fsTA and FSRS spectra of the hairpin Sa-A2GT3.39 The significantly longer lifetimes of the charge-separated radical ion pairs for these hairpins (e.g., 0.13 ns for DPA-A1G vs 100 ns for DPA-A3G, Chart 2) may allow for charge delocalization of G+• with the neighboring A bases or deprotonation of G+• to form the guanine radical.31 Our failure to characterize G+• by means of fsTA or FSRS spectroscopies upon photoinduced charge separation in hairpins having 2 or more A-T base pairs separating the DPA from G led us to investigate the use of phenylethylguanine (EG, Chart 3) in place of guanine as the hole trap.25 EG absorbs at longer wavelengths than G and undergoes characteristic shifts in both its transient absorption band and in its CC Raman stretch upon oxidation. The spectra of EG-1 and its picosecond decay are similar to those of the isolated chromophore (EGH2), indicating that no charge transfer occurs with the neighboring G-C base pairs. In the case of EG-2, excitation of Sa results in the appearance of bands assigned to Sa−• and EG+• in both the fsTA and FSRS spectra. The observation of the transient spectra of Sa−• and EG+• upon laser excitation of EG-2 suggested the potential use of capped hairpins such as EG-3 or EG-4 to directly track the dynamics and efficiency of hole transport across a duplex sequence containing an intermediate shallow hole trap by means of transient spectroscopy. More efficient charge transport from EG+• to Sd (ΦEG‑Sd = 0.26) is observed for EG-4 (Chart 3) than for the analogous hairpin having G in place of EG (ΦG‑Sd = 0.16).

Chart 2. Structures of (a) the DPA-Diol Linker, (b) DPAlinked Hairpins, and (c) DPA Minihairpins

Figure 6. fsTA and nsTA spectra for several hairpins in Chart 2 at the indicated delay times following a 330 nm laser pump pulse in aqueous buffer. The nsTA spectra shown (>8 ns) were scaled to fsTA by comparing the 7 ns spectra when necessary. Adapted with permission from ref 34. Copyright 2016 American Chemical Society.

distinct band is observed at 575 nm for DPA-G and DPA-A1G, but for none of the other DPA-AnG hairpins. This band is sharper than those reported previously, using pulse radiolysis

Chart 3. Structures of (a) the Guanine Derivative EG as a Base Surrogate and (b) the Hairpins EG-1−4

1750

DOI: 10.1021/acs.accounts.8b00090 Acc. Chem. Res. 2018, 51, 1746−1754

Article

Accounts of Chemical Research

■

band assigned to DPA−•, indicating that it is not directly associated with charge recombination of DPA−•-AnG+• and thus cannot be assigned to G+•. The absence of Raman bands for the (An)+• or AnG+•Am (n > 1) species is plausibly a consequence of their delocalized structures.34 The 565 nm absorption maximum is also present in the fsTA spectrum of DPA-A6.34 However, it is absent in the stable minihairpins DPA-A1−3.39 It should be noted that the intensity and lifetime of the A-polaron reaches a maximum when n = 4− 5, the same as the number of A-T base pairs required for the characteristic features of A-tract DNA in the ground state, such as its narrow minor groove and specific hydration motifs.43,44 Less obvious than the specific hydration is the linear helical axis of the A-tract, a consequence of efficient A-A base stacking, which is clearly seen in the top-down view of the Xray structure of hairpin Sd-BrU (Figure 1)! The presence of a G-C base pair in the middle of a longer A-tract alters the Atract local structure but preserves its linearity.44 This suggests that the ground state structure of the DNA A-tract may be a crucial factor in its ability to serve as a conduit for charge transport over multiple base pairs in DNA.

EVIDENCE FOR FORMATION OF AN A-POLARON As noted previously, the base sequence employed most frequently in studies of photoinduced hole injection and transport in DNA is poly(A).3 Conwell was an early proponent of A-polaron formation in DNA and proposed that a hole injected into an extended A-tract leads to the formation of a polaron delocalized over 3−5 A-T base pairs.40,41 We first encountered experimental evidence that implicated A-polaron formation in a study of the fsTA and time-resolved EPR spectra of DNA hairpin conjugates possessing a perylenediimide (PDI) chromophore base surrogate separated from a GC base pair by a variable number of A-T base pairs.42 In the course of investigating the FSRS and fsTA spectra of the DPA-AnG hairpins, we noted that hairpins for which n ≥ 2 (Chart 2) do not display Raman or transient absorption bands characteristic of G+•, which were observed for n < 2 (Figure 7).34 However, hairpins with n > 2 show 565 nm fsTA bands,

■

SUMMARY AND OUTLOOK The key steps in the photoinduced charge separation process in DNA that we have identified are summarized in Scheme 2. Irradiation of Sa or another suitable hole donor (D) results in formation of a locally excited singlet state that can oxidize a neighboring purine base (A or G, Figure 4) with a rate constant that is dependent upon the thermodynamics of the hole injection process (eq 1). When DPA is the excited chromophore D and G is located in the first or second base pair direct spectroscopic evidence has been obtained for the formation of the guanine cation radical G+•.34 Spectroscopic evidence for the formation of the cation radical of the guanine derivative EG+• has been obtained when Sa is the excited chromophore and EG is located in the third base pair.25 When 3 or more A-T base pairs separate a DPA chromophore and G, spectroscopic and kinetic evidence have been obtained for the formation of a delocalized A-tract intermediate in hole transport to G.34 The hole transport time (τtG) is dependent upon both the length of the A-tract and the free energy of hole injection, and the hole transport efficiency is dependent upon the competition between the rate constants for hole transport and the contact radical ion pair charge recombination (τcr1).14 G or EG base can serve as a shallow hole trap in an A-tract and reduce the efficiency of hole transport from Sa to Sd by introducing a second point of competition between hole transport (τtSd) and charge recombination (τcr2). When G is followed by a G-tract the efficiency of hole transport becomes

Figure 7. Normalized femtosecond stimulated Raman spectra, obtained using a 575 nm Raman pump following 330 nm actinic excitation of the hairpins at a time point roughly 6τinj. The peaks are labeled accordingly and the major G+• features are denoted with vertical dashed lines. Reproduced with permission from ref 38. Copyright 2015 American Chemical Society.

which increase in intensity with n, attaining a maximum intensity in their normalized transient absorption spectra when n = 4 or 5. The assignment of these bands to the A-polaron is consistent with the kinetic behavior of the DPA-AnG hairpins. The 565 nm band decays well before the stronger 500 nm

Scheme 2. Summary of Key Steps Identified in Photoinduced Charge Separation in DNA

1751

DOI: 10.1021/acs.accounts.8b00090 Acc. Chem. Res. 2018, 51, 1746−1754

Accounts of Chemical Research

■

relatively insensitive to the number of G-bases, as we have previously found in a study of hole transport in diblock purines.27 Whereas key steps in Scheme 2 have been determined for several chromophores (Figures 4 and 5), the key intermediates G+• and An+• have been detected only when DPA is the hole donor. It would be highly desirable to characterize both intermediates and determine the dynamics of their formation and decay as a function of the free energy of their formation. Unfortunately, spectral overlap between the chromophores and intermediates complicates such an investigation. Several theories have been advanced to explain the distance dependence of the charge separation dynamics in Sa-An-Sd hairpins as well as a more recent theory to explain the much faster dynamics for NDI-An-Sd hairpins.14 As mentioned previously, the change in slope for plots of log(kcs) vs n for Sa-An-Sd and DPA-An-Sd (Figure 2) can be accounted for by a change in mechanism from tunneling to hopping (Scheme 1). However, values of kci are faster than kcs making it unlikely that tunneling occurs in a single step, unless preceded by hole injection.16 More recent theories include “quantum filling” in which tunneling at short distances is replaced by charge transport via hole delocalization over multiple bases at longer distances45 and “flickering resonance” in which tunneling is replaced by transient degeneracy of the bridging states, which supports coherent or “ballistic” charge transfer at short distances.46 Renaud et al. recently proposed that NDI is able to inject holes into occupied orbitals of the neighboring adenine base below the highest occupied molecular orbital.14 These hot holes are able to produce delocalized adenine cation radicals, which extend further into the stacked bases than thermalized holes formed by less potent donors such as 1*Sa or 1 *DPA, thus forming a conduction channel for rapid, longdistance hole transport. This Account brings to a close the most recent chapter of a collaborative investigation of photoinduced charge transfer in DNA begun over 20 years ago. In addition to the studies of hole transport described herein, we have extended the methods developed to charge transfer in less-common DNA structures such as diblock and triblock purines,27,47 locked nucleic acids (LNAs),48 3-way-junctions,49 G-quadruplexes,50 and i-motifs.51 The synergy of molecular design, structure, dynamics, and theory has led to our current state of understanding of DNA charge separation. We and others have applied similar methods to the study of electron transport in DNA.29,52 Future work will focus on a number of new research directions that have become apparent from the work carried out thus far. These involve the need to develop a deeper understanding of the structural and electronic changes within the DNA duplex that accompany charge generation and transport as well as how to take advantage of charge and spin motion within the DNA duplex to develop new device structures. For example, we expect that substituted guanine derivatives such as EG may be particularly effective in tracking and controlling hole transport in such systems. In addition, our observation of photogenerated entangled multispin states within DNA makes it possible to take advantage of spin dynamics to manipulate quantum information at a molecular level and to create DNA structures that may form the basis of quantum circuitry.

Article

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. ORCID

Frederick D. Lewis: 0000-0002-3669-2796 Ryan M. Young: 0000-0002-5108-0261 Michael R. Wasielewski: 0000-0003-2920-5440 Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Funding

Financial support was provided by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, Chemical Sciences, Geosciences, and Biosciences Division under the Awards DE-FG02-96ER14604 (F.D.L.) and DEFG02-99ER14999 (M.R.W.) and the Office of Naval Research MURI Grant No. N00014-11-1-0729 (F.D.L. and M.R.W.). Notes

The authors declare no competing financial interest. Biographies Frederick D. Lewis received a B. A. from Amherst College and his Ph.D. in physical organic chemistry from the University of Rochester in 1968. After a postdoctoral year studying photochemistry at Columbia University, he joined the faculty at Northwestern University in 1969. Following a long and rewarding career combining research, teaching, and mentoring, he was appointed Professor Emeritus in 2016. Ryan M. Young obtained his B.S. in Chemistry at the University of California, Los Angeles, in 2006. He then studied anion dynamics using time-resolved photoelectron imaging at the University of California, Berkeley, with Daniel Neumark, receiving his Ph.D. in 2011. From there, he joined Michael Wasielewki’s group at Northwestern University as a Camille and Henry Dreyfus Environmental Chemistry Postdoctoral Fellow and is currently a Research Associate Professor at Northwestern University, as well as the Director of Laboratory Research at the Institute for Sustainability and Energy at Northwestern (ISEN). Michael R. Wasielewski received his B.S. (1971), M.S. (1972), and Ph.D. (1975) in Chemistry from the University of Chicago. Following a postdoctoral fellowship at Columbia University, he began his career at Argonne National Laboratory, ultimately becoming Group Leader of the Molecular Photonics Group. In 1994, he joined the faculty of Northwestern University, where he is currently the Clare Hamilton Hall Professor of Chemistry and Executive Director of the Institute for Sustainability and Energy at Northwestern.

■

ACKNOWLEDGMENTS We thank all current and former members of the Lewis and Wasielewski groups and the collaborating groups who have participated in the research featured in this Account.

■

REFERENCES

(1) Murphy, C. J.; Arkin, M. R.; Jenkins, Y.; Ghatlia, N. D.; Bossmann, S. H.; Turro, N. J.; Barton, J. K. Long-Range Photoinduced Electron Transfer through a DNA Helix. Science 1993, 262, 1025−1029. 1752

DOI: 10.1021/acs.accounts.8b00090 Acc. Chem. Res. 2018, 51, 1746−1754

Article

Accounts of Chemical Research

(22) Montalti, M.; Credi, A.; Prodi, L.; Candolfi, M. T. Handbook of Photochemistry, 3rd ed.; Taylor & Francis: Boca Raton, FL, 2006. (23) Jalilov, A. S.; Patwardhan, S.; Singh, A.; Simeon, T.; Sarjeant, A. A.; Schatz, G. C.; Lewis, F. D. Structure and Electronic Spectra of Purine-Methyl Viologen Complexes. J. Phys. Chem. B 2014, 118, 125−133. (24) Seidel, C. A. M.; Schulz, A.; Sauer, M. H. M. NucleobaseSpecific Quenching of Fluorescent Dyes. 1. Nucleobase One-Electron Redox Potentials and Their Correlation with Static and Dynamic Quenching Efficiencies. J. Phys. Chem. 1996, 100, 5541−5553. (25) Brown, K. E.; Singh, A. P. N.; Wu, Y.-L.; Mishra, A. K.; Zhou, J.; Lewis, F. D.; Young, R. M.; Wasielewski, M. R. Tracking Hole Transport in DNA Hairpins Using a Phenylethynylguanine Nucleobase. J. Am. Chem. Soc. 2017, 139, 12084−12092. (26) Lewis, F. D.; Liu, J.; Weigel, W.; Rettig, W.; Kurnikov, I. V.; Beratan, D. N. Donor-Bridge-Acceptor Energetics Determine the Distance Dependence of Electron Transfer in DNA. Proc. Natl. Acad. Sci. U. S. A. 2002, 99, 12536−12541. (27) Vura-Weis, J.; Wasielewski, M. R.; Thazhathveetil, A. K.; Lewis, F. D. Efficient Charge Transport in DNA Diblock Oligomers. J. Am. Chem. Soc. 2009, 131, 9722−9727. (28) Ng, H. L.; Dickerson, R. E. Mediation of the A/B-DNA Helix Transition by G-Tracts in the Crystal Structure of Duplex CATGGGCCCATG. Nucleic Acids Res. 2002, 30, 4061−4067. (29) Fujitsuka, M.; Majima, T. Charge Transfer Dynamics in DNA Revealed by Time-Resolved Spectroscopy. Chem. Sci. 2017, 8, 1752− 1762. (30) Kobayashi, K.; Yamagami, R.; Tagawa, S. Effect of Base Sequence and Deprotonation of Guanine Cation Radical in DNA. J. Phys. Chem. B 2008, 112, 10752−10757. (31) Choi, J.; Yang, C.; Fujitsuka, M.; Tojo, S.; Ihee, H.; Majima, T. Proton Transfer of Guanine Radical Cations Studied by TimeResolved Resonance Raman Spectroscopy Combined with Pulse Radiolysis. J. Phys. Chem. Lett. 2015, 6, 5045−5050. (32) Bucher, D. B.; Pilles, B. M.; Carell, T.; Zinth, W. Charge Separation and Charge Delocalization Identified in Long-Living States of Photoexcited DNA. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 4369− 4374. (33) Adhikary, A.; Kumar, A.; Becker, D.; Sevilla, M. D. Guanine Cation Radical: Investigation of Deprotonation States by Esr and Dft. J. Phys. Chem. B 2006, 110, 24171−24180. (34) Harris, M. A.; Mishra, A. K.; Young, R. M.; Brown, K. E.; Wasielewski, M. R.; Lewis, F. D. Direct Observation of the Hole Carriers in DNA Photoinduced Charge Transport. J. Am. Chem. Soc. 2016, 138, 5491−5494. (35) Lewis, F. D.; Liu, X.; Miller, S. E.; Wasielewski, M. R. Electronic Interactions between π-Stacked DNA Base Pairs and Diphenylacetylene-4,4’-Dicarboxamide in Hairpin DNA. J. Am. Chem. Soc. 1999, 121, 9746−9747. (36) Kobayashi, K.; Tagawa, S. Direct Observation of Guanine Radical Cation Deprotonation in Duplex DNA Using Pulse Radiolysis. J. Am. Chem. Soc. 2003, 125, 10213−10218. (37) Hiura, H.; Takahashi, H. Time-Resolved Resonance Raman Spectroscopy of Diphenylacetylene: Structures and Dynamics of the Lowest Excited Triplet State, Radical Cation, and Radical Anion. J. Phys. Chem. 1992, 96, 8909−8915. (38) Wu, Y.-L.; Brown, K. E.; Gardner, D. M.; Dyar, S. M.; Wasielewski, M. R. Photoinduced Hole Injection into a SelfAssembled π-Extended G-Quadruplex. J. Am. Chem. Soc. 2015, 137, 3981−3990. (39) Mishra, A. K.; Harris, M. A.; Young, R. M.; Wasielewski, M. R.; Lewis, F. D. Dynamics of Charge Injection and Charge Recombination in DNA Mini-Hairpins. J. Phys. Chem. B 2017, 121, 7042−7047. (40) Conwell, E. M.; Rakhmanova, S. V. Polarons in DNA. Proc. Natl. Acad. Sci. U. S. A. 2000, 97, 4556−4560. (41) Conwell, E. M.; Bloch, S. M.; McLaughlin, P. M.; Basko, D. M. Duplex Polarons in DNA. J. Am. Chem. Soc. 2007, 129, 9175−9181. (42) Zeidan, T. A.; Carmieli, R.; Kelley, R. F.; Wilson, T. M.; Lewis, F. D.; Wasielewski, M. R. Charge-Transfer and Spin Dynamics in

(2) Lewis, F. D. Electron Transfer and Charge Transport Processes in DNA. In Electron Transfer in Chemistry; Balzani, V., Ed.; WileyVCH: Weinheim, Germany, 2001; Vol. 3, pp 105−175. (3) Schuster, G. B., Ed. Long-Range Charge Transfer in DNA, I and II; Springer: Berlin, 2004; Vols. 236 and 237. (4) Wagenknecht, H. A., Ed. Charge Transfer in DNA; Wiley-VCH: Weinheim, 2005. (5) Genereux, J. C.; Barton, J. K. Mechanisms for DNA Charge Transport. Chem. Rev. 2010, 110, 1642−1662. (6) Salunkhe, M.; Wu, T.; Letsinger, R. L. Control of Folding and Binding of Oligonucleotides by Use of a Nonnucleotide Linker. J. Am. Chem. Soc. 1992, 114, 8768−8772. (7) Lewis, F. D.; Liu, X.; Wu, Y.; Miller, S. E.; Wasielewski, M. R.; Letsinger, R. L.; Sanishvili, R.; Joachimiak, A.; Tereshko, V.; Egli, M. Structure and Photoinduced Electron Transfer in Exceptionally Stable Synthetic DNA Hairpins with Stilbenediether Linkers. J. Am. Chem. Soc. 1999, 121, 9905−9906. (8) Egli, M.; Tereshko, V.; Mushudov, G. N.; Sanishvili, R.; Liu, X.; Lewis, F. D. Face-to-Face and Edge-to-Face π-π Interactions in a Synthetic DNA Hairpin with a Stilbenediether Linker. J. Am. Chem. Soc. 2003, 125, 10842−10849. (9) Lewis, F. D.; Wu, T.; Zhang, Y.; Letsinger, R. L.; Greenfield, S. R.; Wasielewski, M. R. Distance-Dependent Electron Transfer in DNA Hairpins. Science 1997, 277, 673−676. (10) Lewis, F. D.; Wu, T.; Liu, X.; Letsinger, R. L.; Greenfield, S. R.; Miller, S. E.; Wasielewski, M. R. Dynamics of Photoinduced Charge Separation and Charge Recombination in Synthetic DNA Hairpins with Stilbenedicarboxamide Linkers. J. Am. Chem. Soc. 2000, 122, 2889−2902. (11) Lewis, F. D.; Wu, Y. Dynamics of Superexchange Photoinduced Electron Transfer in Duplex DNA. J. Photochem. Photobiol., C 2001, 2, 1−16. (12) Lewis, F. D.; Wu, Y.; Zhang, L.; Zuo, X.; Hayes, R. T.; Wasielewski, M. R. DNA-Mediated Exciton Coupling and Electron Transfer between Donor and Acceptor Stilbenes Separated by a Variable Number of Base Pairs. J. Am. Chem. Soc. 2004, 126, 8206− 8215. (13) Lewis, F. D.; Zhu, H.; Daublain, P.; Cohen, B.; Wasielewski, M. R. Hole Mobility in DNA A-Tracts. Angew. Chem., Int. Ed. 2006, 45, 7982−7985. (14) Renaud, N.; Harris, M. A.; Singh, A. P. N.; Berlin, Y. A.; Ratner, M. A.; Wasielewski, M. R.; Lewis, F. D.; Grozema, F. C. Deep-Hole Transfer Leads to Ultrafast Charge Migration in DNA Hairpins. Nat. Chem. 2016, 8, 1015−1021. (15) Lewis, F. D.; Zhu, H.; Daublain, P.; Fiebig, T.; Raytchev, M.; Wang, Q.; Shafirovich, V. Crossover from Superexchange to Hopping as the Mechanism for Photoinduced Charge Transfer in DNA Hairpin Conjugates. J. Am. Chem. Soc. 2006, 128, 791−800. (16) Blaustein, G. S.; Lewis, F. D.; Burin, A. L. Kinetics of Charge Separation in Poly(A)-Poly(T) DNA Hairpins. J. Phys. Chem. B 2010, 114, 6732−6739. (17) Giese, B.; Amaudrut, J.; Köhler, A.-K.; Spormann, M.; Wessely, S. Direct Observation of Hole Transfer through DNA by Hopping between Adenine Bases and by Tunnelling. Nature 2001, 412, 318− 320. (18) Conron, S. M. M.; Thazhathveetil, A. K.; Wasielewski, M. R.; Burin, A. L.; Lewis, F. D. Direct Measurement of the Dynamics of Hole Hopping in Extended DNA G-Tracts. An Unbiased Random Walk. J. Am. Chem. Soc. 2010, 132, 14388−14390. (19) Singh, A. P. N.; Harris, M. A.; Young, R. M.; Miller, S. A.; Wasielewski, M. R.; Lewis, F. D. Raising the Barrier for Photoinduced DNA Charge Injection with a Cyclohexyl Artificial Base Pair. Faraday Discuss. 2015, 185, 105−120. (20) Weller, A. Photoinduced Electron Transfer in Solution: Exciplex and Radical Ion Pair Formation Free Enthalpies and Their Solvent Dependence. Z. Phys. Chem. 1982, 133, 93−98. (21) Lewis, F. D.; Zhang, L.; Kelley, R. F.; McCamant, D.; Wasielewski, M. R. A Perylenedicarboxamide Linker for DNA Hairpins. Tetrahedron 2007, 63, 3457−3464. 1753

DOI: 10.1021/acs.accounts.8b00090 Acc. Chem. Res. 2018, 51, 1746−1754

Article

Accounts of Chemical Research DNA Hairpin Conjugates with Perylenediimide as a Base-Pair Surrogate. J. Am. Chem. Soc. 2008, 130, 13945−13955. (43) Woods, K. K.; Maehigashi, T.; Howerton, S. B.; Sines, C. C.; Tannenbaum, S.; Williams, L. D. High-Resolution Structure of an Extended A-Tract: [d(CGCAAATTTGCG)]2†. J. Am. Chem. Soc. 2004, 126, 15330−15331. (44) Han, G. W.; Kopka, M. L.; Cascio, D.; Grzeskowiak, K.; Dickerson, R. E. Structure of a DNA Analog of the Primer for HIV-1 RT Second Strand Synthesis. J. Mol. Biol. 1997, 269, 811−826. (45) Renaud, N.; Berlin, Y. A.; Lewis, F. D.; Ratner, M. A. Between Superexchange and Hopping: An Intermediate Charge-Transfer Mechanism in Poly(A)-Poly(T) DNA Hairpins. J. Am. Chem. Soc. 2013, 135, 3953−3963. (46) Zhang, Y.; Liu, C.; Balaeff, A.; Skourtis, S. S.; Beratan, D. N. Biological Charge Transfer via Flickering Resonance. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 10049−10054. (47) Thazhathveetil, A. K.; Trifonov, A.; Wasielewski, M. R.; Lewis, F. D. Increasing the Speed Limit for Hole Transport in DNA. J. Am. Chem. Soc. 2011, 133, 11485−11487. (48) Thazhathveetil, A. K.; Vura-Weis, J.; Trifonov, A. A.; Wasielewski, M. R.; Lewis, F. D. Dynamics and Efficiency of Hole Transport in LNA: DNA Hybrid Diblock Oligomers. J. Am. Chem. Soc. 2012, 134, 16434. (49) Young, R. M.; Singh, A. P. N.; Thazhathveetil, A. K.; Cho, V. Y.; Zhang, Y.; Renaud, N.; Grozema, F. C.; Beratan, D. N.; Ratner, M. A.; Schatz, G. C.; Berlin, Y. A.; Lewis, F. D.; Wasielewski, M. R. Charge Transport Across DNA-Based Three-Way Junctions. J. Am. Chem. Soc. 2015, 137, 5113−5122. (50) Thazhathveetil, A. K.; Harris, M. A.; Young, R. M.; Wasielewski, M. R.; Lewis, F. D. Efficient Charge Transport via DNA G-Quadruplexes. J. Am. Chem. Soc. 2017, 139, 1730−1733. (51) Fujii, T.; Thazhathveetil, A. K.; Yildirim, I.; Young, R. M.; Wasielewski, M. R.; Schatz, G. C.; Lewis, F. D. Structure and Dynamics of Electron Injection and Charge Recombination in i-Motif DNA Conjugates. J. Phys. Chem. B 2017, 121, 8058−8068. (52) Gorczak, N.; Fujii, T.; Mishra, A. K.; Houtepen, A. J.; Grozema, F. C.; Lewis, F. D. Mechanism and Dynamics of Electron Injection and Charge Recombination in DNA. Dependence on Neighboring Pyrimidines. J. Phys. Chem. B 2015, 119, 7673−7680.

1754

DOI: 10.1021/acs.accounts.8b00090 Acc. Chem. Res. 2018, 51, 1746−1754