Tracking Hole Transport in DNA Hairpins Using a Phenylethynylguanine Nucleobase

Article abstract of DOI:10.1021/jacs.7b06998

The hole transport dynamics of DNA hairpins possessing a stilbene electron acceptor and donor along with a modified guanine (G) nucleobase, specifically 8-(4′-phenylethynyl)deoxyguanosine, or EG, have been investigated. The nearly indistinguishable oxidation potentials of EG and G and unique spectroscopic characteristics of EG^+? make it well-suited for directly observing transient hole occupation during charge transport between a stilbene electron donor and acceptor. In contrast to the cation radical G^+?, EG^+? possesses a strong absorption near 460 nm and has a distinct Raman-active ethynyl stretch. Both spectroscopic characteristics are easily distinguished from those of the stilbene donor/acceptor radical ion chromophores. Employing EG, we observe its role as a shallow hole trap, or as an intermediate hole transport site when a deeper trap state is present. Using a combination of ultrafast absorption and stimulated Raman spectroscopies, the hole-transport dynamics are observed to be similar in systems having EG vs G bases, with small perturbations to the charge transport rates and yields. These results show EG can be deployed at specified locations throughout the sequence to report on hole occupancy, thereby enabling detailed monitoring of the hole transport dynamics with base-site specificity.

Full text of DOI:10.1021/jacs.7b06998

Subscriber access provided by Queen Mary, University of London
Article
Tracking Hole Transport in DNA Hairpins
Using a Phenylethynylguanine Nucleobase
Kristen E Brown, Arunoday P.N. Singh, Yi-Lin Wu, Ashutosh Kumar Mishra,
Jiawang Zhou, Frederick D. Lewis, Ryan M. Young, and Michael R. Wasielewski
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b06998 • Publication Date (Web): 02 Aug 2017
Downloaded from http://pubs.acs.org on August 2, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
Tracking Hole Transport in DNA Hairpins Using a
Phenylethynylguanine Nucleobase
9
Kristen E. Brown, Arunoday P. N. Singh, YiꢀLin Wu, Ashutosh Kumar Mishra, Jiawang Zhou, Frederick
D. Lewis*, Ryan M. Young*, Michael R. Wasielewski*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Department of Chemistry, ArgonneꢀNorthwestern Solar Energy Research (ANSER) Center, and Institute
for Sustainability and Energy at Northwestern, Northwestern University, Evanston, Illinois 60208ꢀ31131
ABSTRACT: The hole transport dynamics of DNA hairpins possessing a stilbene electron
acceptor and donor along with a modified guanine (G) nucleobase, specifically 8ꢀ(4′ꢀ
phenylethynyl)deoxyguanosine, or EG, have been investigated. The nearly indistinguishable
oxidation potentials of EG and G and unique spectroscopic characteristics of EG+• make it wellꢀ
suited for directly observing transient hole occupation during charge transport between a stilbene
electron donor and acceptor. In contrast to the cation radical G+•, EG+• possesses a strong
absorption near 460 nm and has a distinct Ramanꢀactive ethynyl stretch. Both spectroscopic
characteristics are easily distinguished from those of the stilbene donor/acceptor radical ion
chromophores. Employing EG, we observe its role as a shallow hole trap, or as an intermediate
hole transport site when a deeper trap state is present. Using a combination of ultrafast
absorption and stimulated Raman spectroscopies, the holeꢀtransport dynamics are observed to be
similar in systems having EG vs. G bases, with small perturbations to the charge transport rates
and yields. These results show EG can be deployed at specified locations throughout the
sequence to report on hole occupancy, thereby enabling detailed monitoring of the hole transport
dynamics with baseꢀsite specificity.
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 32
1
2
3
4
■ INTRODUCTION
5
6
7
8
The observation of wireꢀlike charge transport in duplex DNA1ꢀ3 has prompted extensive,
careful design of DNA sequences, specifically tailored to achieve highꢀefficiency hole transport.
Substitution of diblock polypurine sequences in lieu of naturally occurring purineꢀpyrimidine
mixed sequences has led to significantly higher efficiencies for photoinduced charge transport4ꢀ5
as has modification to natural base pairs.6ꢀ9 Of particular interest is the role of guanine, which has
the lowest oxidation potential of the natural nucleobases.10ꢀ11 As such, guanine acts as a shallow
hole trap that can influence the dynamics of competitive charge recombination and charge
transport in DNA. Chemical reactions of the guanine cation radical can also lead to DNA strand
cleavage,12ꢀ13 and studies of photoinduced oxidative strand cleavage in duplexes containing
multiple GꢀC base pairs by Barton, Schuster, and Giese in the 1990’s led to the current models
for hole hopping in DNA.14ꢀ16
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
While previous studies have demonstrated efficient hole transport in Gꢀtracts across series of
AmGn diblock purine sequences,4 mechanistic understanding of these systems relies on a detailed
kinetic analysis of the donor and acceptor signals. In particular, direct detection of oxidized
intermediates, specifically the guanine cation (G+•), remains challenging. Electronic absorption
spectroscopy has, until recently, been limited to studies of the more stable, neutral radical (GꢀH)•
–
generated by flash photolysis following oneꢀelectron oxidation by the sulfate radical anion (SO4
•).17ꢀ18 Kobayashi and Tagawa have reported the transient absorption spectra of G+• formed upon
pulse radiolysisꢀtransient absorption of several duplexes containing one or more G:C base pairs
and its deprotonation to form the more stable neutral radical (GꢀH)•. Both reactive intermediates
have broad, weak transient spectra in the 350ꢀ700 nm region with no characteristic absorption
bands.19
2
ACS Paragon Plus Environment

Page 3 of 32
Journal of the American Chemical Society
1
2
Our recent work20 using a diphenylacetylenedicarboxamide hole donor revealed a weak G+•
absorption in the visible spectrum, where it is typically obscured by the much stronger and
broader absorption features of the chromophores typically employed in these studies.21
Moreover, it was demonstrated that the already weak intensity further decreases with the length
of the purineꢀtract in which it is embedded, which makes directly tracking the specific location of
the photoinjected hole quite challenging.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To overcome these issues, we recently synthesized and studied an electronꢀrich guanine
derivative, 8ꢀ(4′ꢀaminophenylethynyl)guanine, GEAn, which provided evidence of hole
delocalization in a Gꢀquadruplex architecture using femtosecond transient absorption (fsTA)
spectroscopy and femtosecond stimulated Raman spectroscopy (FSRS).22 The addition of the
aminophenylethynyl group to the 8ꢀposition of G results in a strong absorption band for the
cation radical species, ca. 510 nm, with only a minor shift in the oxidation potential of the G unit,
making this species an excellent starting point for designing a spectroscopic tag to study hole
transport dynamics within DNA.
Here, we aim to optically identify the role of intermediate hole acceptors in hairpinꢀforming
DNA charge conduits through development of a synthetically modified guanine nucleobase
analogue, specifically 8ꢀ(4′ꢀphenylethynyl)deoxyguanosine, EG (Chart 1a). We observe a strong
EG+• absorption band at 460 nm, far removed from the congested bands in the 500ꢀ600 nm
region of the spectrum of the stilbeneꢀbased donor and acceptor we have previously employed,20
which thereby allows us to track the hole population directly. The similarity of the oxidation
potential to G means that EG can be placed at a specific location within a Gꢀtract and serve as a
siteꢀspecific indicator of hole arrival, with minimal expected perturbation of the dynamics. In
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 32
1
2
3
4
5
6
addition, the phenylethynyl group is strongly Raman active, which provides an alternative, baseꢀ
specific approach to track the hole location using FSRS.
7
8
9
To provide direct spectroscopic evidence of EG as a hole acceptor, we incorporated EG into
a series of DNA hairpins (Chart 1b), capped with a nonꢀreactive thymine bridge (EG-1) or
photooxidant stilbenedicarboximide (Sa) linker (EG-2). We also probe the role of EG in the
presence of a deep hole trap (EG-3, -4), namely stilbenediether (Sd). The stilbene linkers have
been demonstrated to produce stable BꢀDNA hairpins and display distinct spectral tags.23ꢀ24 We
selected A2G as the initial bridge based on the high quantum yield of charge separation (ca. 25%)
in related systems.4 The four hairpins were designed to: 1) identify the spectral signature of EG+•
when employed as a (shallow) terminal trap state, 2) probe EG+• as an intermediary acceptor
state, and 3) distinguish EG+• from G+• within a hole transport conduit. Reference hairpins G-2–4
with G in place of EG were also studied to compare the dynamics and spectra. We find that,
indeed, EG can serve as a siteꢀspecific probe of the hole location in these systems, and maintains
similar transport and recombination behavior compared to sequences with unsubstituted G
despite small differences in hole transport rates.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
ACS Paragon Plus Environment

Page 5 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Chart 1. Structures of a) Sa, Sd, EG and b) the hairpins in this study. Hole transfer pathway highlighted in red.
The control compound EG-H2 is the protonated version of EG.
■ MATERIALS AND METHODS
Synthesis and characterization of 8ꢀ(4′ꢀphenylethynyl)deoxyguanosine (EG-H2), hairpins
EG-1–4, G-1–4, and relevant reference compounds are provided in the Supporting Information
(SI) along with experimental details for the timeꢀresolved spectroscopies. A description of the
fitting methodology and details of the applied kinetic models are also presented in the SI.
■ RESULTS AND ANALYSIS
Electrochemistry. Differential pulse voltammetry provides oxidation potentials of Eox = 1.12 V
and 1.11 V vs. SCE (0.1 M nBu4NPF6 in DMF) for EG-H2 and deoxyguanosine, respectively
(Figure S1), which are effectively the same within the instrumental resolution. The oxidation
potential of the next most easily oxidized nucleobase, adenine, is higher by about 0.4 eV,25 so the
overall energetic landscape of the EGꢀcontaining sequences remains similar to those with G. Any
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 32
1
2
3
4
5
6
small shift in potential resulting from incorporation in the DNA hairpins is expected to only
slightly perturb the hole transport dynamics, as discussed in detail below.
7
8
9
Steady-State Spectroscopy. Figure 1 shows the absorption spectra of the hairpins normalized at
260 nm and their circular dichroism (CD) spectra. The UV spectrum of EG-1 reveals that the
addition of the phenylethynyl substituent at the 8ꢀposition redꢀshifts the guanine groundꢀstate
absorption spectrum from 253 nm with a shoulder at 272 nm to 332 nm with a shoulder at 350
nm. The structured absorption spectrum is reminiscent of the La and Lb bands observed in
guanine.26ꢀ27 The EG absorption coincides with the Sa and Sd steadyꢀstate absorption, as is
apparent in the spectra of EG-2–4. Each of these Saꢀcontaining hairpins shows an increase in the
330 nm absorption band relative to the 260 nm nucleobase absorption band. This increase is
demonstrated in Figure S2 (SI) by comparison of the spectra of compounds EG-ref and Sa-ref
which contain one EG or Sa chromophore, respectively, and 6 A:T base pairs.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a)
EG-1
EG-2
EG-3
1.0
0.8
0.6
0.4
0.2
0.0
EG-4
G-2
G-3
G-4
250
300
350
400
450
Wavelength (nm)
6
ACS Paragon Plus Environment

Page 7 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
b)
EG-1
EG-2
EG-3
EG-4
7
8
9
4
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0
-2
-4
250
300
350
400
450
Wavelength (nm)
Figure 1. a) UV/Vis absorption spectra normalized to the 260 nm band
and b) Circular dichroism spectra of EG hairpins (3 ꢀM) in phosphate
buffer (10 mM phosphate, 0.1 M NaCl, pH 7.2).
The CD spectra of hairpins EG-2–4 indicates they adopt BꢀDNA structures similar to
previously studied Saꢀlinked hairpins,28 as evidenced by the sign and intensity pattern in the
shortꢀwavelength portion of the spectrum (200–300 nm). High values for the first derivatives of
o
the thermal dissociation profiles for sequences EG-1–4 (CDꢀmelting >50 C, Figure S3)
indicates the phenylethynyl substituent at the 8ꢀposition of guanine enhances duplex stability,
presumably by extending into the major grove without disrupting the base pairing. The bands in
the longꢀwavelength region (300–400 nm) are attributed to induced CD of the two or three
achiral chromophores which have absorption maxima in the same wavelength region of the CD
minima.28 In contrast, the EG-1 spectrum has a strong positive band at 260 nm and a negative
band around 215 nm which indicate some distinct features of an AꢀDNAꢀlike structure.29
Poly(dG).poly(dC) sequences similar to that of EG-1 are known to form intermediate or mixed
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 32
1
2
B/A type structures.30,31 This structural difference should not compromise the utility of EG-1 as
a model sequence designed to probe the behavior of EG in the absence of stilbene chromophores.
The decreased intensity of the induced CD band of EG-1 is a consequence of the absence of the
stilbene chromophores.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Transient Absorption Spectroscopy. Femtosecond transient absorption spectroscopy is used to
initiate and monitor hole transport dynamics. Briefly, transient absorption spectra were acquired
after 350 nm excitation followed by a broadband visibleꢀtoꢀNIR probe. At the excitation
wavelength of 350 nm, we expect an approximate 2:1 Sa:EG coꢀexcitation ratio. The normalized
absorption of EG-2 and EG-3, the latter of which contains both stilbene chromophores, reveals
that Sd also contributes to 18% of the total absorption at 350 nm, but is not expected to impact
the hole transport properties of the system.32 In these systems, the fraction of absorbed photons
1*
leading to the population of Sa from which hole injection occurs is approximately 0.55 (See
Figures 1 and S2). All samples were purged with N2 and stirred to minimize the effects of
degradation and local heating.
We first investigated the excitedꢀstate behavior of the EG-H2 chromophore in methanol,
shown in Figure 2a. Excitation of EG at 350 nm results in a sharp excitedꢀstate absorption band
at 500 nm which extends well into the nearꢀinfrared region with two small maxima at 1000 and
1250 nm. The signal decays biꢀexponentially with τ1 = 6 ps and τ2 = 18 ps, with minimal spectral
evolution. The relaxation dynamics are largely similar when EG is placed into a duplex hairpin,
as in EG-1 (Figure 2b): the maximum absorption shifts to 505 nm, and the biꢀexponential decay
is somewhat slower (9 and 27 ps). The location of the 505 nm absorption is reminiscent of the
GEAn+• absorption; however, the similarity of the spectra and dynamics of EG-1 with those of
1*
the isolated EG-H2 strongly suggests that the 505 nm band belongs to EG and that there is no
8
ACS Paragon Plus Environment

Page 9 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
charge transfer or strong exciplex interaction between the EG and the neighboring C:G base
pairs.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.16
0.12
0.08
0.04
a)
b)
500 nm
EG-H2 / MeOH
Delay (ps)
-1
1
2.5
5.0
10
25
50
100
250
500
1000
2500
5000
7500
x5
0.00
0.16
505 nm
EG-1
0.12
0.08
0.04
0.00
x5
400
600
800
1000
1200
1400
Wavelength (nm)
Figure 2. Transient absorption for a) EG-H2 in methanol, and b) EG-1 in aqueous buffer acquired
after 350 nm, 500 ꢀJ excitation. The excitedꢀstate maxima of 1*EG in both occur near 500 nm, and
decay biꢀexponentially. Delays are given in ps.
The fsTA spectra for EG-2–3 and G-2–3 following 350 nm excitation are shown in Figures 3
and S4, respectively. The spectra are dominated initially by the absorption of the 1*Sa state at 577
nm.33 Coꢀexcitation of EG is evidenced by the narrow 510 nm band of 1*EG that is likely shifted
1*
by added absorption at redder wavelengths by Sa. Over the next tens of picoseconds, the 577
nm band sharpens, characteristic of Sa–•,33 and new bands at 460 and 1000 nm appear. The 1000
nm band is also assigned to absorption by Sa–• as it appears concomitantly with the 577 nm Sa–•
band in G-2 where there are no additional chromophores.
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 32
1
2
We assign the new 460 nm feature that appears at later times to absorption by EG+•, since it
appears with the same rate as the bands belonging to Sa–• and is also absent in the reference
sequences G-2–4 (Figure S4). The shift of the 460 nm EG+• band relative to the guanineꢀ
quadruplex GEAn+• band plausibly results from the decreased conjugation length owing to
replacement of the aniline moiety with a phenyl substituent.22
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.15
a)
EG-2
x4
Delay (ps)
0.10
-1.1
1.0
2.5
5.0
10
0.05
25
51
0.00
105
249
515
1040
2490
5090
7690
b)
0.15
0.10
0.05
0.00
EG-3
400
600
800
1000
1200
1400
Wavelength (nm)
Figure 3. Femtosecond transient absorption spectra of a) EG-2 and b) EG-3 following 350 nm
excitation. Delays are given in ps. Early time traces (blue) represent the EG excited state (λmax = 510 nm)
and 1*Sa excited state (λmax = 578 nm). At late times, absorption maxima of EG+• (λmax = 460 nm) and Sa–•
(λmax = 577 nm and 1000 nm) are apparent.
To model the hole transport dynamics in EG-2, we employ global kinetic analysis with the
simplified model depicted in Scheme 1a. Excitation of the Saꢀcontaining hairpins results in
10
ACS Paragon Plus Environment

Page 11 of 32
Journal of the American Chemical Society
1
2
relaxation (kR) followed by charge separation (kCS) to form Sa–•/EG+•. As explicit separation of
the 1*EG and 1*Sa (and 1*Sd) populations leads to a large number of parameters in the model, we
simplify the kinetic model to assume a single excitedꢀstate population.34 Based on the data for
EG-1, the initial dynamics during the first ~100 ps will thus contain some evolution from the
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1*
secondary EG population. This observation is corroborated by FSRS measurements discussed
below. However, since hole transport primarily occurs after this early period, the simplified
kinetic model should serve as a good approximation to those dynamics. Details of the kinetic
modelling and results can be found in the SI section 3, and kinetic parameters are summarized in
Table 1.
1*
1*
For EG-2, global analysis of the fsTA data captures the initial decay of the EG and Sa
populations with one process, τR ~ 7 ps. This relaxation comprises the initial decay of 1*EG
observed in EG-1, along with some of the reversible hole injection (1*Sa ⇌ Sa–•) observed when
stilbenedicarboxamide is used as a hole donor and A:T base pair as hole acceptor.35 The speciesꢀ
associated spectrum (SAS, Figure S8) for this state, a combination of 1*EG and 1*Sa ⇌ Sa–•,
supports this assignment. A similar spectrum and rate are obtained for G-2, without the 505 nm
contribution from 1*EG. This composite state then decays to form the Sa–•/EG+• chargeꢀseparated
state in τCS = 51 ps, as indicated by the appearance of the 460 nm band. The Sa–•/EG+• state then
decays in τEG+ = 3.4 ns, while in Gꢀ2 the Sa–•/G+• state decays in τG+ = 1.4 ns. Nanosecond
transient absorption (nsTA) spectroscopy reveals the same decay kinetics on the nsꢀꢀs timescale
(Figures S17 and S20), and shows the evolution to a low intensity, much longerꢀlived 8 ꢀs
species with an absorption maximum at 440 nm. This species is most likely the 3*Sa state, as the
triplet state of unsubstituted transꢀstilbene has an absorption maximum at ca. 390 nm in viscous
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 32
1
2
solution at room temperature.36 The presence of the similar band in G-2 also indicates that the
3
4
5
6
band does not originate from EG.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Kinetic scheme for hole transport in a) EG-2 (SaꢀA2ꢀEGꢀT3) and b) EG-3 (SaꢀA2ꢀEGꢀA2ꢀ
Sd) and EG-4 (SaꢀA2ꢀEGꢀG2ꢀSd). Coꢀexcitation of EG and Sa is followed by relaxation/fluorescence
from 1*EG and 1*Sa, and reversible hole injection from 1*Sa to form Sa–•ꢀꢀA+, all of which are captured
in part by kR. Hole transport to EG occurs with rate kCS; the higher energy level in this state represents
EG-3 with the bridge B2 = A2, while the lower energy state represents EG-4 with the bridge B2 = G2. In
EG-3–4 the rate (kSd) and yield of trapping by Sd+• are determined by the sequence of intervening base
pairs B: either A2 or G2. States in gray are not captured by the data modelling. Similar schemes hold
for G-2 (SaꢀA2ꢀEGꢀT3) and for G-3 (SaꢀA2ꢀGꢀA2ꢀSd) and G-4 (SaꢀA2ꢀGꢀG2ꢀSd).
12
ACS Paragon Plus Environment

Page 13 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Table 1. Time constants and ΦSd derived from global fits of the fs/nsTA datasets to the first order kinetic
model presented in Scheme 1. τx = (kx)ꢀ1. Uncertainties are reported as the standard error of the fit.
Explicit decays pathways to the ground state are neglected, such that each time constant represents an
effective time.
10
11
12
13
14
15
τCR2
,
τR,
τCS,
τ
EG+ (τG+)
τSd,
τCR1
,
ΦSd
ΦSd
ΦSd
ΦSd
psa
psa
nsa
nsa
nsb
(5 ns) (7 ns)a (10 ns)b (40 ns)c
ꢀsb,e
16
17
18
7.0 ± 0.3
51 ± 1
3.41 ± 0.01
ꢀꢀ
ꢀꢀ
ꢀꢀ
ꢀꢀ
ꢀꢀ
ꢀꢀ
ꢀꢀ
EG-2
1.7 ± 0.3
3.6 ± 0.3
5.0 ± 0.3
2.9 ± 0.3
7.6 ± 0.3
34 ± 1
37 ± 1
40 ± 1
39 ± 1
2.74 ± 0.04
1.92 ± 0.04
1.40 ± 0.01
1.29 ± 0.01
14 ± 1
4.4 ± 0.2
ꢀꢀ
5.8 ± 0.6
23 ± 1
ꢀꢀ
7.1 ± 0.6
3.0 ± 0.1
< 1%
EG-3
EG-4
0.16
ꢀꢀ
0.05
0.26d
0.16
ꢀꢀ
0.03
0.15
ꢀꢀ
0.06
0.05
ꢀꢀ
0.03
19
ꢀꢀ
ꢀꢀ
G-2
20
16 ± 2
~1000
G-3
21
57 ± 1
0.86 ± 0.02
2.2 ± 0.1 7.2 ± 0.2
0.7 ± 0.1
0.21
0.20
0.07
G-4
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a) Determined from fsTA data.
b) Determined from nsTA data.
c) Value corrected for contamination by 3*Sd by subtracting the spectrum at 40 ꢀs.
d) Reference value taken from Ref. 4 and corrected for the fraction of Sa extinction coefficient at 350 nm (
e) May include other decay pathways, see SI.
ηSa = 0.75).
The EG-2 kinetics inform the hole transfer dynamics in sequences EG-3–4, with Sd as the
deeper hole trap. In both EG-3 and EG-4, the decreased initial intensity of the 505 nm 1*EG
band relative to that in EG-2 is attributed to competitive absorption by Sd (Figures 1 and S2).
The spectra for both EG-3 and G-3 (Figure S4) are similar to those for EG-2 and G-2,
respectively, with the initial appearance of the 1*Sa and 1*EG bands followed by the formation of
the Sa–•/EG+• radical pair. Despite the expected slower recombination from Sa–•/EG+• for EG-3,
the fsTA data (Figure 3b) do not clearly exhibit the 533 nm absorption associated with Sd+•
within the first 7 ns.
To model the charge separation and recombination processes, we extend the same kinetic
model used above, allowing the possibility of hole trapping from the Sa–•/EG+• state to form Sa–
•/Sd+• with rate kSd (Scheme 1b). The additional process changes the rate of decay of EG+• to
kEG+ = kCR0 + kSd
(1a)
(1b)
kG+ = kCR0 + kSd
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 32
1
2
where kCR0 is the value of kEG+ for EG-2 and represents the charge recombination from EG•+. A
similar situation exists for G-2–4 with kG+ for charge recombination from G•+. For EG-3, charge
separation occurs with a ~35 ps time constant to populate the Sa–•/EG+• state, with absorption
bands around 460, 577, and 1000 nm. This state decays with minor spectral changes; however, a
small signal persists beyond the ~8 ns window of the fsTA experiment. The measured decay time
of Sa–•/EG+• is 2.7 ns in EG-3, which is faster compared to that in EG-2 (3.4 ns), suggesting a
possible hole trapping pathway. The nsTA data (Figures S16ꢀS23) show that the remaining
population at long times largely decays over the next 5.8 ns (charge recombination) to a longꢀ
lived signal around 530 nm (Figure 4a), near the expected absorption of Sd+•.37
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The spectra at long times (>>10 ꢀs) do not show absorption from Sa–• as would be expected
for the fully chargeꢀseparated state. Instead, the 440 nm feature is reminiscent of the 3*Sa
spectrum. The rapid disappearance suggests that it decays by tripletꢀtriplet energy transfer to a
lower lying triplet state. In this case, the isolated 530 nm absorption feature could be the Tn ← T1
transition of 3*Sd.36 However, it is also possible that some permanent oxidation of Sd is
occurring due to incomplete removal of oxygen, resulting in a longꢀlived Sd+• population that is
not accompanied by the Sa–•, resulting in the 530 nm absorption band.
In G-3, there is more direct spectral evidence for hole trapping in low yield. After fast
relaxation, charge separation occurs in 39 ps, followed by a 1.3 ns decay. Figure 4b shows that at
long delays (3ꢀ7 ns), inversion of the 533 and 577 nm signal is observed. The low amplitude of
the signal results from fast (1.3 ns) recombination.
14
ACS Paragon Plus Environment

Page 15 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
a)
1.0
0.8
0.6
0.4
0.2
0.0
EG-3
7
8
9
3.0 ns
5.0 ns
7.7 ns
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
450
500
550
600
650
700
Wavelength (nm)
b)
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
G-3
3.0 ns
5.0 ns
7.6 ns
450
500
550
600
650
700
Wavelength (nm)
Figure 4. Normalized fsTA spectra for a) EG-3 and b) G-3 at long times showing the inversion of the
533 and 577 nm bands. Data are normalized to the Sa–• absorption at 577 nm.
The fsTA data for EG-4 and G-4 are both shown in Figure 5. Initially, the spectra appear
similar to those of the hairpins discussed above. However, at long delays, we observe an increase
in the absorption near 533 nm, characteristic of the Sa–•/Sd+• pair. At intermediate times in EG-4
we observe an increased intensity near 460 nm, presumably owing to absorption by transiently
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 32
1
2
oxidized EG before the rise of the Sd+• band (Figure 5a, inset); a feature not observed in G-4
(Figure 5b, inset). The noticeable inversion of the 533 and 577 nm bands occurs after 2 ns,
indicating a significant yield of hole trapping within the experimental window.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a)
0.12
0.08
0.04
0.010
0.005
0.000
EG-4
Delay (ps)
-1.1
1.0
2.5
5.0
10
440
460
480
500
0.00
0.06
25
51
105
249
515
1040
2490
5090
7590
b)
0.010
0.005
0.000
G-4
0.04
0.02
0.00
440
460
480
500
400
600
800
1000
1200
1400
Wavelength (nm)
Figure 5. fsTA spectra of a) EG-4 and b) G-4 following 350 nm excitation. Delays are given in ps. At
long delays hole trapping at Sd+• in moderate yields is observed at 480 and 533 nm. Inset: Increased
amplitude near 460 nm at intermediate (20ꢀ1000 ps) delays shows the characteristic absorption of EG+•
in EG-4 indicating the transit of the hole through EG before trapping at Sd.
The kinetic processes are modelled as previously discussed. The lifetime of the Sa–•/EG+•
intermediate state in EG-4 is 1.9 ns, compared to the longer 3.4 ns lifetime in EG-2, which again
16
ACS Paragon Plus Environment

Page 17 of 32
Journal of the American Chemical Society
1
2
suggests the presence of an additional decay pathway. The characteristic Sa–•/Sd+• spectrum
decays with τCR1 = 23 ns. A second, slower recombination event is observed with τCR2 = 3 ꢀs, as
evidenced by the similar speciesꢀassociated spectrum with characteristic Saꢀ•/Sd+• features. This
process may arise from minor dynamical conformation(s) which disfavor charge recombination.
Following the slower charge recombination, we observe the same lowꢀamplitude signal at 530
nm as in EG-3. The reference sequence G-4 exhibits similar kinetics to those of EG-4, with the
decay of the Sa–•/G+• intermediate occurring with a time constant of 860 ps, compared to 1400 ps
for G-2. Multiple recombination times are also observed by nsTA, with τCR = 7.2 and 650 ns,
although the amplitude of the slower component is low (Figure S22).
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Femtosecond Stimulated Raman Spectroscopy. FSRS data were acquired with a 350 nm
actinic pump, and a ps Raman pump/fs broadband probe pair. The Raman pump wavelength
(λRP) was tuned to 480 nm to be close to resonance with the 1*EG and EG+• transitions. We also
utilized a 600 nm Raman pump to resonantly enhance 1*Sa and Sa–• signals to track the hole
injection process. FSRS spectra of EG-1 (Figure 6a), collected using λRP = 480 nm to resonantly
enhance 1*EG, reveal several excited state bands between 500 cmꢀ1 and 2200 cmꢀ1 which all
decay biꢀexponentially. Time constants are similar to those determined by transient absorption
spectroscopy. The absence of a timeꢀdependent shift in the C≡C stretching mode at 2122 cmꢀ1
corroborates the lack of charge separation in EG-1. Selected Raman mode assignments are given
in the SI. In hairpins EG-2–4, the 480 nm Raman pump wavelength is also resonant with the
transient EG+• absorption. At early times, EG-2 closely matches EG-1, with low intensity peaks
indicative of Sa–• in the 1500ꢀ1600 cmꢀ1 region forming at later times (Figure S25). The C≡C
mode shifts to higher frequency (2172 cmꢀ1) following oxidation of EG, which directly indicates
hole transfer to form EG+•.22
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 32
1
2
3
4
Raman spectra for EG-3 are provided in Figure S25, while those for EG-4 are provided in
5
6
1*
Figure 6b. The C≡C mode of EG initially following excitation appears at 2104 cmꢀ1. At late
7
8
9
times, the spectra closely match those of EG-2 indicating similar hole transfer to EG+•. In
contrast to EG-2, the early time spectra in the 1100ꢀ1500 cmꢀ1 region are broadened with
additional peaks between 1300ꢀ1500 cmꢀ1 and decreased relative intensity of the 1175 cmꢀ1
bands.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a) EG-1, λRP = 480 nm
2122
1186
1236
1125
1581
2%
1001
919
1424
823
608
1*EG
b) EG-4, λRP = 480 nm
2104
1*EG
EG+•
Sa−•
1
%
2172
1562
Sa−•
Delay (ps)
c) EG-4, λRP = 600 nm
1562
-1.5
-1
-0.5
0
0.5
1
1.5
30
Sd+•
1*Sa
Sa−•
100
550
1280
1950
4000
1560
1605
2%
1266
500 750 1000 1250 1500 1750 2000 2250 2500
Raman Shift (cm-1)
Figure 6. FSRS spectra of EG-1 and 4 following 350 nm excitation. a) Spectra of EG-1 acquired with
18
ACS Paragon Plus Environment

Page 19 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
1*
λrp = 480 nm. b) Spectra of EG-4 (λRP = 480 nm) showing EG at early times and composite spectra
of EG+• and Sa–• at late times. c) Spectra of EG-4 acquired with λRP = 600 nm showing Sa at early
1*
times and Sa–• at late times. Delay times are provided in ps. See the SI for mode assignments.
We also probe the dynamics of hole injection using a 600 nm Raman pump. Representative
7
8
9
spectra for EG-4 are shown in Figure 6c, showing peaks at 1266, 1562, 1605 cmꢀ1. After 35 ps,
coinciding with the formation of the Sa–•/EG+• state, the 1605 cmꢀ1 mode becomes less
pronounced. The appearance of the 1605 cmꢀ1 mode as a small shoulder on the 1560 cmꢀ1 band,
suggests its attribution to 1*Sa. The DFTꢀcalculated Raman spectra identify the 1605 cmꢀ1 feature
as the amide stretch of 1*Sa, and the 1266 and 1562 cmꢀ1 bands as the vinyl bending and
stretching modes, respectively, of Sa–• (Figure S38). In EG-4, there is a longꢀlived population
with a peak at 1560 cmꢀ1 that is consistent with the vinyl C=C stretches in both Sa–• and Sd+• (see
SI, Figures S39ꢀ40), which keeps with the modelling of the transient absorption data.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
■
DISCUSSION
Ethynylguanine Absorption and Relaxation. The similarity between the EG-1 and the isolated
EG-H2 nucleoside dynamics and the lack of significant spectral evolution (Figure 1) strongly
indicate that electron transfer or other photoprocesses do not occur between 1*EG and the
adjacent C:G base pair in EG-1. The observed dynamics, thus, are the excitedꢀstate decay of
1*EG. It is not unexpected that the addition of the phenylethynyl substituent at the 8ꢀposition
would perturb the relaxation dynamics of unsubstituted guanine.26 Increased radiative decay
times have been observed in nucleobases modified such that torsion is restricted around nuclear
coordinates that gate nonꢀradiative decay via a conical intersection.38 The vertical excited state
may quickly relax to a configuration where internal conversion is slowed, such as through an
intramolecular charge transfer state as was shown in a recent theoretical study of 8ꢀ
vinylguanine.39 Additional work is necessary to identify the nature of biꢀexponential decay in
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 32
1
2
3
4
5
6
7
8
9
EG-H2 and is beyond the scope of the current investigation. Importantly, however, parallel
excitation of EG along with Sa appears to result in a shortꢀlived population of 1*EG that does not
undergo charge transfer, and thus should not greatly alter the hole transport dynamics originating
from 1*Sa.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Hole Injection and Charge Separation. The FSRS spectra, acquired at 600 nm to resonantly
enhance the 600 nm 1*Sa ⇌ Sa–• absorption, indicate a reversible charge injection process. As
seen in Figure 7, in the first few ps following excitation, we observe an apparent combination of
1*
1*Sa and Sa–• Raman features. We have previously observed that hole injection from Sa into a
neighboring Aꢀblock is reversible,35 obstructing resolution of the hole injection step. The FSRS
spectra support this assessment. With a rapid pseudoꢀequilibrium there will be sizeable
population of both species, as both possess an electronic transition at this Raman pump
wavelength, until the population drains into a lower energy state, in this case Sa–•/EG+• in the ~30
ps charge separation. Complete modeling of these processes is prohibitive, as discussed above,
and so they are collectively captured with the relaxation pathway kR. While not fully resolving
these kinetics, the FSRS spectra offer a direct observation of the reversible injection process.
20
ACS Paragon Plus Environment

Page 21 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
EG-4: ꢁt = 1.0 ps
1*Sa (× 0.951)
1605 cm-1
Sa−• (× 0.967)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1565 cm-1
1266 cm-1
1000
1200
1400
1600
1800
2000
Raman Shift (cm-1)
Figure 7. Comparison of the FSRS spectrum of EG-4 acquired at ꢁt = 1.0 ps (λRP = 600 nm) to the
calculated spectra of 1*Sa and Sa–• at the B3LYP/6ꢀ31G* level with 15 cmꢀ1 Lorentzian broadening. The
1*
feature at 1605 cmꢀ1 is the amide stretch of Sa, while the peaks at 1266 and 1565 cmꢀ1 are the vinyl
bending and stretching modes of Sa–•, all of which are present at this delay due to the reversible charge
injection.
The rise in the Sa–• signal also provides insight into the charge transport process in systems
with a small Aꢀblock neighboring the donor. Direct kinetic comparison of the λRP = 480 nm and
λ
RP = 600 nm Raman experiments for EG-2–4 (Figures S28ꢀ30 and S33ꢀ35, respectively) reveal
nearly identical formation times for (isolated) Sa–• and EG+•. This observation is consistent with
a superexchange hole transport mechanism from Sa to Sa–•/EG+•. Under this mechanism, thought
to occur for short Aꢀtracts (n < 3),40ꢀ41 the formation of Sa–• and EG+• should occur concomitantly
and formation of the Sa–•/A+• contact ion pair should not be spectrally or kinetically observable.
In contrast, under a hopping mechanism, Sa–• is fully reduced prior to hole transport from A+• to
EG+•, leading to different appearance times for the two species. We note, however, that the
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 32
1
2
3
4
5
6
reversible hole injection discussed above may complicate the dynamics, and the true mechanism
is likely more subtle.42
7
8
9
Hole Trapping Dynamics. The results of the kinetic modelling of the TA data offer insight into
the hole transport dynamics and hole location in the EG- and Gꢀsequences. As discussed above,
charge separation in EG-2 occurs in 50 ps. This time, and the similar analogous time in G-2 are
also close to the previously observed ~30 ps charge separation times for SaꢀA2 systems.43 The
discrepancy likely arises from the residual parallel decay of 1*EG during this time range (Scheme
1a). As shown in Table 1, the extracted values for the hole trapping time τSd are around two
orders of magnitude larger than the charge separation times, despite the similar distances. This is
likely a consequence of the different charge transport mechanisms at play depending on the
number of intervening base pairs between the two radicals.42
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Following charge separation, the Sa–•/EG+• state decays in 3.4 ns, almost a factor of two
slower than the G-2 reference sequence. The difference in recombination rates is likely a
consequence of the slightly altered ꢁGCR from the shift in oxidation potential of EG vs. G,
although contributions from altered electronic coupling between adjacent nucleobases cannot be
ruled out. While the electrochemically determined oxidation potential of EG was measured to be
the same as that of G, the difference in the potentials is likely exaggerated in the DNA medium.
The dielectric environment is plausibly impacted by the neighboring nucleobases, lowering the
effective dielectric constant relative to the highly polar DMF solvent used for electrochemical
measurements. Additionally, the structure of the duplex in the region of the modified nucleobase
may be perturbed as the phenylethynyl group protrudes into the major groove, possibly altering
the oxidation potential and the electronic coupling. While the altered oxidation potential affects
22
ACS Paragon Plus Environment

Page 23 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
the recombination dynamics, the charge does not become irreversibly trapped at EG and does not
prohibit using EG in the diblock architectures.
7
8
9
The behavior of EG+• as a shallow hole trap can, in principle, be used to enhance the yield of
charge separation in situations where recombination from the poly(purine) sequence outcompetes
the deep hole trapping. By providing a site with an extended lifetime and lower recombination
rate, more of the hole population can make its way to the deep trap state. While in EG-3 we do
not see a strong rise in the characteristic signature of Sd+•, there are indications that the hole is
transferring to the Sd trap site. At longꢀtimes (7.7 ns, Figure 4a), there is a weak growth of the
533 nm Sd•+ signal, though the amplitude suggests a <1% trapping yield. The lifetime of the Sa–
•/EG+• chargeꢀseparated state in EG-3 is, however, significantly shorter than in EG-2, which
indicates that there is a kinetic process competing with charge recombination. We calculate the
rate of hole transfer time from EG+• to Sd+•, τSd (=1/kSd), from eq. 1, which for EG-3 gives τSd =
13.9 ns. This rate would suggest as much as 40% (kSd/kEG+) of the hole population can transfer to
Sd. However, the 5.8 ns decay observed in the nsTA spectra indicates that the kinetics of charge
recombination from the Sd trap are inverted, preventing any population from building up to a
detectable level during the fsTA experiment, leading to the spectra observed in Figure 3b.
The situation is similar in G-3 where the Sa–•/Sd+• radical pair is observed at long times,
as seen in Figure 4a, although the observed yield is quite low. Kinetic analysis gives τSd = 16.4
ns, but a maximum yield of only 8%. This low yield is independently corroborated by
determining the yield directly from the spectra at a specific time. We calculate the yield of hole
trapping on Sd, ΦSd, by obtaining the ratio of the 1*Sa absorption amplitude at 500 fs and the area
of the absorption at 7 ns delay and comparing that ratio to a known standard:17
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
23
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 32
1
2
3
4
5
6










ꢁASd+•
(
t
)
ꢁASd+•
(
t
)


ΦSd
(t
)
=
÷
×
ηSa ×Φref
(2)


ꢁA t ≈ 0
(
)
ꢁA t ≈ 0
(
)
1*Sa
x
1*Sa
ref
7
1*
1*
8
9
The coꢀexcitation of the EG chromophore and the overlapping absorptions of EG and Sa
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
requires a correction for the fraction of absorbed photons that lead to the formation of 1*Sa (
ηSa).
We take G-4 as the reference with Φref = 26% when corrected for the fraction of Sa absorbing at
350 nm.4 Further details are provided in the SI, section 3. This approach provides a timeꢀ
dependent quantum yield of charge trapping, as the ratio is based on the area at a given time
delay; Table 1 reports the yields calculated at 5 and 7 ns using the fsTA measurement, and 10
and 40 ns, based on the nsTA data. Ideally, the yield would be reported when the trapping step is
completed and the spectral intensity is effectively constant. However, as the trapping rate is on
the same order as the charge recombination (τEG+ and τSd ~ 3 ns, and τCR1 ~ 6 ns), the spectrum
continues to decay even as the ratio becomes constant. From this spectral analysis, the yield of
charge separation is ~5% for G-3 toward the end of the fsTA experiment and then decreases in
keeping with the kinetic model. The low yield is also consistent with the SAS of the terminal
state, which shows band inversion indicative of trapping at Sd (Figure 4a).
In contrast to EG-3, the fsTA data for EG-4 show 533 nm absorption within the first few
ns, characteristic of a high yield of Sd+•. The strong amplitude of Sd+• in EG-4 suggests more
efficient hole transport through a GGꢀblock than the AAꢀblock,4 and indicates that the hole is not
trapped in the purine tract when in the presence of a deep hole site. In the model, the trapping
step is combined with injection into the G2 block. This is supported by the concomitant loss of
the 460 nm EG+• absorption with the appearance of the 533 nm Sd+• band and the lack of a strong
+•
apparent G2 absorption.20
Importantly, the characteristic absorption of the EG+• species is
observed prior to hole trapping by Sd, signifying that EG can serve as an optical tag of the hole
24
ACS Paragon Plus Environment

Page 25 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
location during transport. It is also possible, given the similar energetics of EG and G, that the
hole is able to delocalize across GG and move as a polaron, such that there is only one
observable population state. The electronic broadening associated with such polaron formation
may also be suppressing the intensity of the 460 nm band in EG-4 relative to EG-3.20
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The calculated yield of Sd+• for EG-4 using eq. 2 at 5 ns is ΦSd = 16%, compared to ΦCS
=
26% for G-4.4 From eq. 1, we calculate hole trapping times of τSd = 4.4 ns and 2.2 ns for EG-4
and G-4, respectively, and hence the kinetic model predicts similar yields of Sd+• around 40% for
EG-4 and G-4. The discrepancy is likely a result of comparison to a timeꢀdependent quantum
yield, as competitive charge recombination reduces the observable yield in the spectrum by
depleting the Sd+• population at a given observation time. In EG-4, the slower trapping rate is
offset, in part, by the slower charge recombination of the Sa–•/EG+• state, making the hole
transfer to Sd more competitive. However, when the trapping rate is slowed substantially, as in
the A2 block case of EG-3, the kinetics become inverted, and the population of the trap state
does not accumulate to measurable levels.
Despite the different oxidation potentials of EG and G and the slowing of overall charge
transport, the observation of charge trapping at Sd+• in reasonable yield for EG-4 demonstrates
the utility of EG as a hole conduit. Comparison of the τSd times for EG-4 and G-4 shows that
hole transport from EG+• is slower by a factor of two, which manifests in a reduction of the
trapping yield, though by less than a factor of two due to slowed charge recombination. The
resultant yields are still moderate and easily detectable for optimized sequences, allowing for EG
to be deployed in more complicated systems without large disruption of the dynamics.
Additionally, we are able to directly observe the effect of the purine sequence on hole transport.
In EG-3, the hole effectively stops at EG, as the inverted kinetics prevent meaningful population
25
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 32
1
2
3
4
5
6
7
8
9
of the trap state from developing. However, if the sequence is designed properly, as in EG-4, the
hole can easily be transported through the tract and arrive at Sd in reasonable yield, while its
location is easily tracked and timed by the spectral tag on EG. We note that we also examined a
system similar to EG-4 with an A3ꢀblock in place of A2 and observed no significant differences
(data not shown), which suggests the behavior we have observed can be easily extended to larger
diblock systems.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
■
CONCLUSIONS
We have demonstrated the use of a phenylethynlguanine nucleobase to track the location of
positive charge during photoinduced charge transport through DNA hairpins by means of
transient absorption and femtosecond stimulated Raman spectroscopies. Following oxidation of
EG, hole transport to Sd+• through multiple Gs was observed, demonstrating the efficacy of EG
to act as hole conduit in a Gꢀtract. The addition of the phenylethynyl group appears to only
slightly alter the oxidation potential of guanine, resulting in slowed hole transport through the
poly(purine) sequence by a factor of two, and subsequently ~30% lower trapping yields.
However, the changes in the energy landscape are not so drastic as to severely alter the general
hole transport behavior observed in transport through purine block sequences with unsubstituted
guanine. We also examined the formation of Sa–• by FSRS, which directly demonstrates
reversible hole injection between 1*Sa and the neighboring Aꢀblock.
Direct observation of EG as an intermediate hole acceptor demonstrates the utility of EG as a
siteꢀspecific probe of hole migration and charge separation. Its unique electronic and vibrational
spectral tags open the door to study hole migration and charge separation processes in complex
DNA architectures such as 3ꢀ44ꢀ45 and 4ꢀway junctions,46 or Gꢀquadruplexes.47 Furthermore, the
26
ACS Paragon Plus Environment

Page 27 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
use of EG to serve as a hole donor to both G and Sd demonstrates its ability to provide insight
into hole transport through a Gꢀtract. Future work to determine the hole mobility in mixed
G(EG) sequences will seek to provide direct information on the hopping rates through Gꢀtracts.5
Optical detection of G+• is complicated by its weak absorption that overlaps with features from
other typical chromophores including the stilbene radical ions. Additionally, when the hole
encounters a tract of Gs, it has been shown theoretically to delocalize across two or three Gs and
behave as a polaron,40 which broadens and weakens the signal, making it even more challenging
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
+•
to observe.20 Even if Gn is observed, its location within a larger Gꢀtract cannot be easily
determined. However, placement of EG at specific locations within larger runs of Gs would
avoid this problem and assist in directly locating the hole during transit, with minimal alteration
of the dynamics, and offer a promising route to exploring hole transport dynamics with site
specificity.
■
ASSOCIATED CONTENT
Supporting Information
Conjugate mass spectral data, thermal dissociation curves, electrochemical data,
femtosecond/nanosecond transient absorption spectra, femtosecond stimulated Raman spectra,
and data analysis, computational details, and Raman mode assignments. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Authors
fdl@northwestern.edu
27
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 32
1
2
3
4
ryan.young@northwestern.edu
5
6
7
8
9
mꢀwasielewski@northwestern.edu
Notes
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The authors declare no competing financial interests.
■
ACKNOWLEDGMENTS
This research was funded by the U.S. Department of Energy, Office of Science, Office of Basic
Energy Sciences, Chemical Sciences, Geosciences, and Biosciences Division under Awards DEꢀ
FG02ꢀ96ER14604 (F.D.L.) and DEꢀFG02ꢀ99ER14999 (M.R.W).
■
REFERENCES
(1) Lewis, F. D., Isr. J. Chem. 2013, 53, 350ꢀ365.
(2) Kawai, K.; Majima, T., Acc. Chem. Res. 2013, 46, 2616ꢀ2625.
(3) Fujitsuka, M.; Majima, T., Chem. Sci. 2017, 8, 1752ꢀ1762.
(4) VuraꢀWeis, J.; Wasielewski, M. R.; Thazhathveetil, A. K.; Lewis, F. D., J. Am. Chem. Soc.
2009, 131, 9722ꢀ9727.
(5) Conron, S. M. M.; Thazhathveetil, A. K.; Wasielewski, M. R.; Burin, A. L.; Lewis, F. D., J.
Am. Chem. Soc. 2010, 132, 14388ꢀ14390.
(6) Kawai, K.; Kodera, H.; Osakada, Y.; Majima, T., Nat Chem 2009, 1, 156ꢀ159.
(7) Kawai, K.; Hayashi, M.; Majima, T., J. Am. Chem. Soc. 2012, 134, 9406ꢀ9409.
28
ACS Paragon Plus Environment

Page 29 of 32
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
(8) Kawai, K.; Majima, T., Increasing the Hole Transfer Rate Through DNA by Chemical
Modification. In Chemical Science of πꢀElectron Systems, Akasaka, T.; Osuka, A.; Fukuzumi, S.;
Kandori, H.; Aso, Y., Eds. Springer Japan: 2015; pp 751ꢀ760.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(9) Thazhathveetil, A. K.; VuraꢀWeis, J.; Trifonov, A.; Wasielewski, M. R.; Lewis, F. D., J. Am.
Chem. Soc. 2012, 134, 16434ꢀ16440.
(10) Steenken, S.; Jovanovic, S. V., J. Am. Chem. Soc. 1997, 119, 617ꢀ618.
(11) Lewis, F. D.; Kalgutkar, R. S.; Wu, Y.; Liu, X.; Liu, J.; Hayes, R. T.; Miller, S. E.;
Wasielewski, M. R., J. Am. Chem. Soc. 2000, 122, 12346ꢀ12351.
(12) Geacintov, N. E.; Solntsev, K.; Johnson, L. W.; Chen, J.; Kolbanovskiy, A. D.; Liu, T.;
Shafirovich, V. Y., J. Phys. Org. Chem. 1998, 11, 561ꢀ565.
(13) Yun, B. H.; Lee, Y. A.; Kim, S. K.; Kuzmin, V.; Kolbanovskiy, A.; Dedon, P. C.;
Geacintov, N. E.; Shafirovich, V., J. Am. Chem. Soc. 2007, 129, 9321ꢀ9332.
(14) Genereux, J. C.; Barton, J. K., Chem. Rev. 2010, 110, 1642ꢀ1662.
(15) Schuster, G. B., Acc. Chem. Res. 2000, 33, 253ꢀ260.
(16) Giese, B., Acc. Chem. Res. 2000, 33, 631ꢀ636.
(17) Candeias, L. P.; Steenken, S., J. Am. Chem. Soc. 1989, 111, 1094ꢀ1099.
(18) Wu, L.; Liu, K.; Jie, J.; Song, D.; Su, H., J. Am. Chem. Soc. 2015, 137, 259ꢀ266.
(19) Kobayashi, K.; Yamagami, R.; Tagawa, S., J. Phys. Chem. B 2008, 112, 10752ꢀ10757.
(20) Harris, M. A.; Mishra, A. K.; Young, R. M.; Brown, K. E.; Wasielewski, M. R.; Lewis, F.
D., J. Am. Chem. Soc. 2016, 138, 5491ꢀ5494.
(21) Lewis, F. D.; Wu, T. F.; Liu, X. Y.; Letsinger, R. L.; Greenfield, S. R.; Miller, S. E.;
Wasielewski, M. R., J. Am. Chem. Soc. 2000, 122, 2889ꢀ2902.
29
ACS Paragon Plus Environment