Photoinduced electron transfer to pyrimidines and 5,6-dihydropyrimidine derivatives: Reduction potentials determined by fluorescence quenching kinetics

Article abstract of DOI:10.1021/jp970164a

The dynamics of flourescence quenching of excited state electron donor sensitizers by various pyrimidine and 5,6-dihydropyrimidine substrates was examined. For all of the substrates studied the rate constant of fluorescence quenching (k_q) increases as the excited state oxidation potential (E_ox*) becomes more negative. The dependence of k_q on E_ox* in each case is well described by the Rehm-Weller relationship. Fits of the data to this relationship allow for the estimation of the reduction potentials of the substrates (E_red). The pyrimidines 1,3-dimethylthymine, 1,3-dimethyluracil, and 1,3,6-trimethyluracil give E_red values (in CH₃CN) ranging from -2.06 (vs SCE) to -2.14 V. Their dihydro derivatives, 1,3-dimethyl-5,6-dihydrothymine, 1,3-dimethyl-5,6-dihydrouracil, and 1,3,6-trimethyl-5,6-dihydrouracil gave E_red values ranging from -1.90 to -2.07 V. The higher E_red values for the dihydropyrimidines compared with their unsaturated derivatives is attributed to aromatic stabilization in the pyrimidines, which is not present in the dihydro derivatives. In addition, the E_red for both the trans-syn and cis-syn diastereomers of the dimethylthymine cyclobutane dimer was examined using the same method. The trans-syn dimer gives an E_red of -1.73 V and the cis-syn dimer gives an E_red of -2.20 V. This remarkable difference is attributed to a stereoelectronic effect. The cis-syn dimer anion radical suffers from an unfavorable charge-dipole interaction between the added electron and the O⁴ carbonyl group in the remaining pyrimidine ring. In contrast, the trans-syn dimer anion radical shows mainly a stabilizing inductive electron-withdrawing effect of the remaining O⁴ carbonyl group. Solvent effects on E_red were also examined. It is shown that the protic solvent, CH₃OH, significantly stabilizes the anion radicals, raising E_red by ca. 400 mV over the value in CH₃CN.

Full text of DOI:10.1021/jp970164a

4332
J. Phys. Chem. A 1997, 101, 4332-4337
Photoinduced Electron Transfer to Pyrimidines and 5,6-Dihydropyrimidine Derivatives:
Reduction Potentials Determined by Fluorescence Quenching Kinetics
Michael P. Scannell, Gautam Prakash, and Daniel E. Falvey*
Department of Chemistry and Biochemistry, UniVersity of Maryland, College Park, Maryland 20742
ReceiVed: January 9, 1997; In Final Form: March 25, 1997X
The dynamics of flourescence quenching of excited state electron donor sensitizers by various pyrimidine
and 5,6-dihydropyrimidine substrates was examined. For all of the substrates studied the rate constant of
fluorescence quenching (kq) increases as the excited state oxidation potential (Eox* ) becomes more negative.
The dependence of kq on Eox* in each case is well described by the Rehm-Weller relationship. Fits of the
data to this relationship allow for the estimation of the reduction potentials of the substrates (Ered). The
pyrimidines 1,3-dimethylthymine, 1,3-dimethyluracil, and 1,3,6-trimethyluracil give Ered values (in CH3CN)
ranging from -2.06 (vs SCE) to -2.14 V. Their dihydro derivatives, 1,3-dimethyl-5,6-dihydrothymine, 1,3-
dimethyl-5,6-dihydrouracil, and 1,3,6-trimethyl-5,6-dihydrouracil gave Ered values ranging from -1.90 to -2.07
V. The higher Ered values for the dihydropyrimidines compared with their unsaturated derivatives is attributed
to aromatic stabilization in the pyrimidines, which is not present in the dihydro derivatives. In addition, the
Ered for both the trans-syn and cis-syn diastereomers of the dimethylthymine cyclobutane dimer was examined
using the same method. The trans-syn dimer gives an Ered of -1.73 V and the cis-syn dimer gives an Ered of
-2.20 V. This remarkable difference is attributed to a stereoelectronic effect. The cis-syn dimer anion radical
suffers from an unfavorable charge-dipole interaction between the added electron and the O4 carbonyl group
in the remaining pyrimidine ring. In contrast, the trans-syn dimer anion radical shows mainly a stabilizing
inductive electron-withdrawing effect of the remaining O4 carbonyl group. Solvent effects on Ered were also
examined. It is shown that the protic solvent, CH3OH, significantly stabilizes the anion radicals, raising Ered
by ca. 400 mV over the value in CH3CN.
Introduction
CHART 1: Some Base Modifications Resulting in
Saturation of the 5,6-Double Bond in Thymine
Addition of an electron to DNA is a process of general
interest. One of the mechanisms by which ionizing radiation
damages DNA involves attachment of solvated electrons to the
DNA bases.1-3 It is generally understood that the attached
electrons localize on the pyrimidines. The two pyrimidine bases
in DNA, cytosine and thymine, have similar electron affinities,4
and there has been some discussion5-8 about which of these
bases serves as the ultimate electron “sink”. It is likely that
this is determined by the local environment around the base in
question. The initial electron attachment, or reduction of DNA,
can lead to a variety of genotoxic lesions on the DNA molecule
including strand scission and base modification.1,2
In addition to the question of radiation damage, there is also
fundamental interest in the mechanism and rate by which single
electrons migrate through DNA strands. Recent experiments
using excited state metal complexes as electron donors and/or
acceptors have led Barton and co-workers9-11 to conclude that
electron (and hole) migration through DNA is surprisingly fast.
The charge carriers presumably migrate through the stacked
bases, although the precise mechanism is still under investiga-
tion.12,13
Our interest in single-electron reduction of DNA comes from
studies of photochemically-driven DNA repair enzymes known
as DNA photolyases.14-16 UV irradiation of DNA results in the
formation of cyclobutane dimers between adjacent pyrimidine
bases. The DNA photolyases bind to damaged bases and then,
upon absorption of a photon, repair them to their normal forms.
A number of studies17-21 indicate that the initial photochemical
step is transfer of a single electron from the protein to the
damaged bases.
While electron transfer in nucleic acids has been extensively
studied in the context of the canonical bases, less is known about
how relatively small changes in the structure of the bases affect
electron transfer. Base modification occurs both enzymatically
and through various types of enviromental insults. Pyrimidine
cyclobutane dimers, spore products, and photohydrates (Chart
1) represent base modifications that result from UV irradiation
of DNA.22,23 Exposure of DNA to ionizing radiation generates
(among other products) 5,6-dihydroxypyrimidines and 5,6-
dihydropyrimidines.1,2
Since naturally occuring DNA is likely to have some fraction
of modified bases, it would be useful to understand how these
modifications affect electron transfer processes. If modified
bases are easily reduced, they might serve as traps for electrons.
On the other hand if modifications render the base more difficult
to reduce, then such sites may constitute a barrier for electron
X Abstract published in AdVance ACS Abstracts, June 1, 1997.
S1089-5639(97)00164-3 CCC: $14.00 © 1997 American Chemical Society

Photoinduced Electron Transfer to Pyrimidines
J. Phys. Chem. A, Vol. 101, No. 24, 1997 4333
TABLE 1: Rehm-Weller Parameters for Some Pyrimidine
Derivatives Determined in CH3CN
(10 mL at a time) into rapidly stirring pentane (200 mL
quantities). Under these conditions, the product dissolves in
the pentane. A colorless solution is decanted from an insoluble
colored oil. Crystals of the product form from the colorless
fraction upon storage at 4 °C overnight. 1,3-Dimethyl-5,6-
dihydrouracil crystallizes as short white needles (5.3 g, 7.5%):
mp ) 47-48 °C. 1H NMR (CDCl3): δ 3.32 (m, 2H), 3.13 (s,
3H), 3.00 (s, 3H), 2.69 (m, 2H). 13C NMR (CDCl3): δ 169.35,
154.07, 42.93, 35.83, 31.43, 27.52. Low resolution MS: m/z
(relative intensity) 142 (M+, 100), 112 (23), 84 (12), 57 (31).
1,3-Dimethyl-5,6-dihydrothymine (DMTH2). 1,3-Dimeth-
ylurea (88.11 g, 1 mol) was refluxed with methacrylic acid (43
g, 0.5 mol) and hydroquinone (0.4 g, 3.6 × 10-3 mol) at 200
°C for 2 h. The product was separated as described for 1,3-
dimethyl-5,6-dihydrouracil. 1,3-Dimethyl-5,6-dihydrothymine
cystallizes as long white needles (15 g, 10%), mp 37-38 °C.
1H NMR (CDCl3): δ 3.28 (dd, J ) 6.2, 6.1 Hz, 1H), 3.13 (s,
3H), 3.09 (m, 1H), 3.03 (s, 3H), 2.71 (m, 1H), 1.21 (d, J ) 6.9
Hz, 3H). 13C NMR (CDCl3): δ 172.43, 154.09, 49.83, 35.80,
35.59, 27.79, 13.28. Low resolution MS: m/z (relative intensity)
156 (M+, 100), 112 (35), 72 (27), 58 (35), 55 (46).
1,3,6-Trimethyl-5,6-dihydrouracil (TMUH2). 1,3-Di-
methylurea (88.11 g, 1 mol) was refluxed with crotonic acid
(43 g, 0.5 mol) at 200 °C for 2 h. The product was separated
as described for 1,3-dimethyl-5,6-dihydrouracil, with the excep-
tion that the CHCl3 layers were not washed with K2CO3. 1,3,6-
Trimethyl-5,6-dihydrouracil crystallizes as white cubes (34 g,
44%): mp 37-38 °C;. 1H NMR (CDCl3): δ 3.49 (m, 1H),
3.17 (s, 3H), 3.01 (s, 3H), 2.84 (dd, J ) 16.4, 6.4 Hz, 1H),
2.53 (dd, J ) 16.4, 2.7 Hz, 1H), 1.20 (d, J ) 6.6 Hz, 3H). 13
C
NMR (CDCl3): δ 168.83, 153.10, 49.33, 38.09, 34.09, 27.47,
17.64. Low resolution MS: m/z (relative intensity) 156 (M+,
62), 141 (84), 84 (100), 55 (40).
1,3-Dimethyluracil (DMU). Uracil (10 g, 0.9 mol) was
added to H2O (50 mL) containing KOH (9 g, 0.16 mol). Under
rapid stirring, the solution turned clear and was placed in an
ice bath. To it, dimethyl sulfate (20 mL, 0.2 mol) was added
dropwise with stirring. After the addition was complete, the
solution was removed from the ice bath and heated until it
boiled. Upon cooling, the solution was extracted with three
aliquots of CHCl3 (50 mL). The CHCl3 layers were combined,
dried over MgSO4, filtered, and rotary evaporated to remove
the solvent. The solid was recrystallized twice from C2H5OH
yielding 1,3-dimethyluracil (9.4 g, 75%): mp 121-123 °C (lit.25
mp 120-121 °C). 1H NMR (CDCl3): δ 7.09 (d, J ) 7.8 Hz,
1H), 5.63 (d, J ) 7.8 Hz, 1H), 3.31 (s, 3H), 3.23 (s, 3H). 13C
NMR (CDCl3): δ 163.19, 151.76, 142.69, 101.11, 36.80, 27.52.
Low resolution MS: m/z (relative intensity) 140 (M+, 100), 83
(43), 54 (48).
migration along the DNA strand. The present study focuses
on how removal, or saturation, of the 5,6-double bond affects
the reduction potentials of various thymine derivatives. As
indicated in Chart 1, this particular structural motif is found in
a variety of DNA photolysis and radiolysis products.
We have examined the pyrimidine derivatives listed in Table
1 and determined their one-electron reduction potentials using
a kinetic method. Analysis of fluorescence quenching kinetics
shows that saturation of the 5,6-double bond makes a given
base easier to reduce. Additionally, we have determined the
reduction potential for the trans-syn-cyclobutane dimer of 1,3-
dimethylthymine and show that is significantly easier to reduce
than than the previously examined cis-syn cyclobutane dimer.
1,3-Dimethylthymine (DMT). Thymine (10 g, 0.8 mol) was
added to H2O (50 mL) containing KOH (9 g, 0.16 mol). The
procedure followed was similar to that for 1,3-dimethyluracil.
1,3-Dimethylthymine (10.4 g, 85%), is recovered as colorless
crystals mp 152-153 °C (lit.26 mp 153 °C); 1H NMR (CDCl3):
δ 7.26 (s, 1H), 3.37 (s, 3H), 3.36 (s, 3H), 1.94 (s, 3H). 13C
NMR (CDCl3): δ 164.12, 152.00, 138.91, 109.64, 36.64, 27.96,
12.97. Low resolution MS: m/z (relative intensity) 154 (M+,
100), 97 (26), 70 (59).
Experimental Section
1,3-Dimethyl-5,6-dihydrouracil (DMUH2). The general
procedure for the synthesis of the dimethyldihydropyrimidines
is a modification of the procedure for the synthesis of hydro-
uracils by Zee-Cheng et al.24 1,3-Dimethylurea (88.11 g, 1 mol)
was refluxed with acrylic acid (36 g, 0.5 mol) and hydroquinone
(0.4 g, 3.6 × 10-3 mol) at 200 °C for 2 h. After 2 h, the solution
was poured into H2O (200 mL) and then extracted three times
with aliqouts of CHCl3 (200 mL). The CHCl3 layers were
combined, washed with two aliquots of 0.5 M K2CO3 (200 mL),
dried over MgSO4, filtered, and rotary evaporated to yield a
pale-yellow viscous liquid. This liquid was added dropwise
1,3,6-Trimethyluracil (TMU). 6-Methyluracil (5 g, 0.4 mol)
was added to H2O (100 mL) containing NaOH (12 g, 0.3 mol).
The solution turned clear and was placed in an ice bath.
Dimethyl sulfate (50 mL, 0.5 mol) was added dropwise with
stirring. After the addition, the solution was heated until it
boiled, cooled, and extracted with three aliquots of CHCl3 (100
mL). The CHCl3 layers were combined, dried over MgSO4,

4334 J. Phys. Chem. A, Vol. 101, No. 24, 1997
Scannell et al.
and filtered, and the the solvent was removed by rotary
evaporation. The solid was recrystallized twice from C2H5OH
held fixed at a number (20-50) of preselected values. The best
fit was selected from the case where the sum of the squares
was minimized and through visual inspection of the 20-50 fits.
This procedure was repeated three times for each reported fit,
“fixing” each of the three parameters in turn and allowing the
remaining two to relax. For all of the cases reported here, this
procedure converged on the same best fit values (to within the
stated uncertainties) regardless of which parameter was “fixed”
and which were allowed to relax. We estimate the uncertainty
in Ered to be (0.08 V and in λ to be (5 kcal/mol.
1
(5.6 g, 92%): mp 111-112 °C (lit.27 mp 113 °C); H NMR
(CDCl3): δ 5.57 (s, 1H), 3.35 (s, 3H), 3.28 (s, 3H), 2.19 (s,
3H). 13C NMR (CDCl3): δ 162.29, 152.52, 151.32, 101.12,
31.57, 27.80, 20.07. Low resolution MS: m/z (relative intensity)
154 (M+, 100), 97 (21), 82 (62), 55 (66).
Cyclobutyldimethylthymine Dimers (csDMTD, tsDMTD).
1,3-Dimethylthymine (8.5 L, 1 mM aqueous solution) was
frozen in batches into sheets (ca. 4 mm thick) and irradiated
for 2 h using a 450 W medium-pressure Hg-vapor lamp fitted
with a Vycor filter. During irradiation the solution was kept
frozen on a bed of dry ice. After 2 h, the solution was thawed,
and the solvent was removed by rotary evaporation. The
residue, a yellow oil, was dissolved in CH3OH and spotted onto
a preparatory TLC plate (2000 µm). The plate was developed
first in 60:40 EtOAc:hexane and then in 85:15 EtOAc:hexane.
The silica gel containing the cis-syn dimer, the lower band (Rf
) 0.06), and the trans-syn dimer, the middle band (Rf ) 0.15),
were isolated and individually washed with CH3OH (ca. 15 mL).
The silica gel was removed by vacuum filteration, and the
solvent was subsequently removed by rotary evaporation. The
resulting solids were individually recrystallized twice from CH3-
OH.
cis-syn-Cyclobutyldimethylthymine Dimer (csDMTD) (0.108
g, 4%): mp 249-253 °C (lit.28 mp 251 °C,); 1H NMR
(CDCl3): δ 3.69 (s, 2H), 3.11 (s, 6H), 2.97 (s, 6H), 1.47 (s,
6H). 13C NMR (CDCl3): δ 169.39, 152.37, 60.51, 47.49, 35.69,
28.07, 19.24. Low resolution MS: m/z (relative intensity) 154
(100), 97 (38), 70 (74), 69 (47) (the M+ peak is not visible
because the compound splits into its constituent halves under
the MS conditions).
trans-syn-Cyclobutyldimethylthymine Dimer (tsDMTD)
(0.22 g, 9%): mp 258-262 °C (lit.29 mp 255 °C). 1H NMR
(CDCl3): δ 3.32 (s, 2H), 3.14 (s, 6H), 3.06 (s, 6H), 1.56 (s,
6H). 13C NMR (CDCl3): δ 169.54, 151.75, 64.35, 36.14, 27.76,
24.92, 12.02. Low resolution MS: m/z (relative intensity) 154
(100), 140 (24), 97 (15), 70 (79), 69 (78) (the M+ peak is not
visible because the compound splits into its constituent halves
under the MS conditions).
Fluorescence Quenching Experiments. A stock solution
of the fluorescent sensitizer is prepared by sonicating the desired
compound (1-3 mg) in spectroscopic grade CH3CN (100 mL)
for 30 min. This process results in a sensitizer concentration
of about 10-2 M. This solution (2.5 mL) is placed in a quartz
cuvette, sealed with a septum, lined with Teflon tape to prevent
contamination, and then purged with Ar for 15 min. A stock
solution of the quencher (200 mM) is prepared by dissolving
the appropriate amount in spectroscopic grade CH3CN and
sonicating the resulting solution for 30 min. Aliquots (12.5 µL)
of the quencher solution are injected into the sealed cuvette
containing the sensitizer. This results in the increase of the
quencher concentration in the cuvette by 1 mM steps. The
fluorescence scan of the sensitizer is recorded at each step (0-6
mM). The excitation wavelength of the sensitizer (330 nm) is
chosen in order to ensure that none of the light is absorbed by
the quencher. The fluorescence quenching rate constants kq were
determined from a Stern-Volmer analysis.30
Results and Discussion
Three types of structural variations were the focus of the
present study. First, it was of interest to determine to what
degree methyl substitution influences the Ered of the pyrimidine
derivatives. To this end we examined 1,3-dimethyluracil
(DMU) and 1,3,6-trimethyluracil (TMU) and compared their
behavior with that of 1,3-dimethylthymine (DMT), which has
been previously examined. The syntheses of these substrates
follow well-known procedures which are described in the
Experimental Section. Second, it was of interest to determine
the effects of saturating the 5,6-double bond in the pyrimidines.
To this end 1,3-dimethyl-5,6-dihydrouracil (DMTH2), 1,3-
dimethyl-5,6-dihydrothymine (DMTH2), and 1,3,6-trimethyl-5,6-
dihydrouracil (TMUH2) were synthesized and examined. Fi-
nally, the trans-syn-cyclobutane dimer of dimethylthymine
(tsDMTD) was examined in order to explore the effect of
stereochemistry on the ability of the bases to accept electrons.
The trans-syn dimer has the same bonds and connectivity as
the previously studied cis-syn dimer (csDMTD). However the
two diastereomers differ in the relative spatial arrangement of
the thymine rings. In the cis-syn both thymine rings are on the
same face of the cyclobutane ring and in the trans-syn they are
on the opposite face. Structures of all of the substrates are
shown in Table 1.
Analysis of fluorescence quenching rate constants was used
to determine the reduction potentials of the substrates. Similar
methods have been previously employed by us31,32 and others.33-35
This technique holds several advantages over the more typical
electrochemical methods. First, the measurements can be made
in homogeneous solutions in the absence of added salts. Second,
the potentials of very unstable organic radical ions can be
measured with reasonable accuracy. Of course, fluorescence
quenching is less direct than any equilibrium-based measure-
ment. The accuracy of the values thus derived are dependent
on the accuracy of the model which relates the kinetic
information to the desired thermodynamic quantities. However,
as the goal here is to identify how small structural alterations
affect Ered, the absolute values of Ered are of less importance
than how they change and in what direction.
Electron transfer from the excited state sensitizers (S*) to the
pyrimidine derivatives (Q) follows the kinetic scheme given in
eq 1.36 The quenching of S* involves a diffusive encounter of
k3
9
S + Q
9
hν8 S* + Q {kkdiff } (SQ)* {kkct } (S+•Q-•)*
8 S+• + Q-•
(1)
-diff
-ct
S* with Q. (kdiff) to form the so-called precursor complex,
followed by a charge transfer step (kct) to form the successor
complex.
Fluorescence quenching rate constants, kq, for a series of
sensitizers with each of the substrates were determined by
Stern-Volmer analysis.30 The sensitizers employed are the
same as the series that was used in a previously published
study.31 The list of sensitizers and their photophysical properties,
Data Analysis. The procedures for fitting the kq data to the
Rehm-Weller relationship is described in some detail else-
where.31 Basically a simplex minimization algorithm was used
to minimize the sum of the squares of the differences between
the experimental data and a theoretical curve that was generated
from the parameters Ered, λ, and kmaxKdiff. In order to determine
the uniqueness of the fits, only two of the parameters were
adjusted in the algorithm while the remaining parameter was

Photoinduced Electron Transfer to Pyrimidines
J. Phys. Chem. A, Vol. 101, No. 24, 1997 4335
along with the kq values for each substrate, is available as
Supporting Information. In the series the excited state oxidation
potential, Eox* varies from -2.1 to -3.3 V (vs SCE). For each
of the substrates, kq increases as Eox* for the sensitizer becomes
increasingly negative. As shown below in Figure 1, log(kq)
approaches an asymptotic value that is at or near the diffusion
limit, kdiff, in the solvent (CH3CN; taken as 1.91 × 1010 M-1
s-1).37
The strong correlation of kq with Eox* supports an electron-
transfer mechanism for quenching. Energy transfer is ruled out
for two reasons. First the lowest energy absorption bands of
the substrates are all well below 300 nm. Thus, their singlet
energies can be safely assumed to be >100 kcal/mol. This value
is more than 15 kcal/mol higher than that of the highest energy
sensitizer employed in this study. Secondly, the a plot of log-
(kq) vs the sensitizers singlet energy reveals no discernable
correlation (data not shown).
In principle, some sort of bimolecular photochemical reaction
could also give rise to quenching. To test for this, several
combinations of sensitizers and substrates were subjected to
prolonged irradiation and analyzed by HPLC. No decomposi-
tion of the substrates could be detected. The results from this
control experiment are also consistent with the electron transfer
mechanism.
The plots of log(kq) vs Eox* for each of the pyrimidine
derivatives were fit to the Rehm-Weller relationship (eqs
2-4).38,39 The adjustable parameters were Ered, the reduction
kdiff
kq )
(2)
q
Figure 1. Rehm-Weller plots for trans-syn-pyrimidine-cyclobutane
dimer of dimethylthymine (tsDMTD, upper panel) and 1,3-dimethyl-
uracil (DMU, lower panel). The kq values were measured in CH3CN
solvent.
kdiff
∆Gct
1 +
exp
(
)
kdiffkmax
RT
2
2
1/2
∆Gct
∆Gct
λ
4
q
CHART 2: Aromaticity vs Increased Conjugation
∆Gct
)
+
+
(3)
(4)
(
[
)
( )
]
RT
2
q2
rꢀ
∆G ) 23.03 Eox - Ered
-
- E
(
)
ct
oo
potential of the pyrmidine derivative, λ, the reorganization
energy, and kmaxKdiff. The latter is the product of kmax, the pre-
exponential term in the Eyring expression for the rate constant
of the charge transfer step (kct), and Kdiff, which is the
equilibrium constant for formation of the reactive precursor
complex. The curves calculated from the Rehm-Weller
relationship show excellent agreement with the experimental
data. Figure 1 shows representative fits for tsDMTD and DMU.
The best fit parameters of these and the other substrates are
presented in Table 1. Some data from the earlier study are also
included for comparison purposes.
We also attempted to fit the data to the Marcus equation.
However, in many cases, this model failed to give reasonable
fits with physically meaningful parameters. It has been widely
known that Marcus theory, while highly successful in predicting
electron transfer rates in rigid systems40,41 and ion pairs,42 often
fails to predict bimolecular excited state quenching behavior.
This is because the inverted region predicted by Marcus theory
is often absent43,44 or obscured.45
To the extent that reasonable fits could be obtained, the same
trends in Ered were observed with the Marcus relationship. The
latter consistently predicts higher (i.e. less negative) values for
Ered and much higher values for λ than the fits to the Rehm-
Weller relationship. Soumillion33 and Schuster34,35 have both
studied ion-neutral excited state quenching processes. In their
systems agreement with Marcus theory could be obtained when
kmaxKdiff was assumed to be larger (1012-1014 M-1 s-1) than
originally assumed by Rehm and Weller. However with the
compounds in our study, reasonable fits to the Marcus theory
required kmaxKdiff to be much smaller than this (109-1011 M-1
s-1).
One objective of this study was to determine how saturation
of the 5,6-double bond in pyrimidines affects Ered. Examination
of similar compounds in the literature reveals two limiting cases.
First, reducing the extent of conjugation in a π-system usually
makes one-electron reduction more difficult. For example,
cyclohexanone has a much more negative Ered (-2.33 V) than
its R-â unsaturated analog, cyclohex-2-en-1-one (-1.62 V)46
(see Chart 2). This analogy would predict that the 5,6-
dihydropyrimidines would have more negative Ered. On the
other hand, uracil and thymine are, at least formally, aromatic
systems. Generally aromatic molecules are more difficult to
reduce than their homologs where one of the double bonds is
reduced. For example, benzene has a lower gas phase electron
affinity (-1.15 eV) than 1,3-cyclohexadiene (-0.85 eV).47 This
analogy would predict that the the 5,6-dihydropyrimidines would
have more positive Ered.
The results from the current study show that the latter analogy
is more accurate. That is, the effect of aromatic stabilization,
which makes one-electron reduction more difficult, predominates
over the effect of increasing conjugation, which should have
made one-electron reduction easier. The 5,6-dihydropyrmin-

4336 J. Phys. Chem. A, Vol. 101, No. 24, 1997
Scannell et al.
TABLE 2: Solvent Effects on the Reduction of TMUH2
THF
CH3CN
CH3OH
Ered (V)a
-1.97
21
80
-1.91
22
40
-1.51
36
160
λ (kcal/mol)b
k
maxKdiff (× 10-9 M-1 s-1
)
a (0.08 V. b (10 kcal/mol.
Solvent effects were also examined. We were particularly
interested in determining whether the hydrogen bonding would
strongly alter Ered. Duplex DNA is a complex environment that
is difficult to model accurately with any particular solvent. On
one hand, the bases in duplex DNA are buried in a hydrophobic
core and are thus not in a truly aqueous environment. On the
other hand, the complementary base on the opposite strand
provides specific and directional hydrogen bonds. Thus the
duplex DNA environment also differs from an aprotic organic
solvent. Both of these effects should be considered when
comparing data from free solution to behavior in native DNA
strands.
Previous studies on thymine and cytosine have revealed
significant discrepencies in the Ered values determined in aqueous
media compared with those determined in aprotic solvents. For
example DMT gives an Ered of -2.14 V in CH3CN,31,32 but a
value of -1.1 V was reported in H2O.4 In an earlier paper we
attributed this large difference to specific H-bonding interactions
between the solvent and the anion radical which were operative
in protic solvents such as H2O but not in aprotic solvents such
as CH3CN.31 In addition to making Ered less negative such a
solvent effect would also be expected to increase the λ value.
Fluorescence quenching experiments with Rehm-Weller
analysis were carried out on TMUH2 in three solvents: CH3-
OH, tetrahydrofuran (THF), and CH3CN. The best fit param-
eters from these experiments are presented in Table 2. The
aprotic solvents, THF and CH3CN, give almost indistinguishable
values for Ered and λ. On the other hand, CH3OH shows a large
(ca. 400 mV) stabilization of the anion radical as seen in its
less negative Ered. Moreover, as predicted, λ is increased by
14 kcal/mol from its value in the aprotic media. Thus, the data
here are consistent with the earlier proposal of H-bonding to
the anion radicals. We would therefore predict that the Ered
values should be increased (i.e. made more favorable) by at
least several hundred millivolts in aqueous media.
Figure 2. Stereoelectronic effects in cyclobutane dimer anion radicals
csDMTD and tsDMTD.
idines are slightly more easily reduced having Ered values that
range from -1.90 to -2.07 V. The corresponding pyrimidines
have Ered values that range from -2.06 to -2.14 V. The
differences are not dramatic, but they are consistent throughout
the series when each pyrimidine is compared to its hydrogenated
analog. We would therefore predict that, in the absence of any
additional structural perturbations, modifications of pyrimidines
that involve loss of the 5,6-double bond are likely to make the
base more easily reduced.
The two cyclobutane dimers, csDMTD and tsDMTD, il-
lustrate an interesting stereoelectronic effect on Ered. These two
diastereomers are similar to the 5,6-dihydropyrimidines in that
they have saturated 5,6-double bonds. However, unlike the 5,6-
dihydro derivatives, the cyclobutyl dimers have inductively
electron-withdrawing carbonyl groups attached directly to the
C-5 position. One might predict that this additional carbonyl
group would make the system easier to reduce, and thus Ered
would be expected to become less negative. This does in fact
appear to be the case for tsDMTD. The Ered for this compound
(-1.73 V) is considerably less negative than its 5,6-dihydro
analog, DMTH2 (-2.07 V).
In contrast, the cis-syn diastereomer csDMTD (-2.14 V) is
considerably more difficult to reduce than either DMTH2 or
tsDMTD. We attribute this to unfavorable charge-dipole
interactions that are present in csDMTD but not in the other
two species. This is illustrated in Figure 2.
Semiempricial calculations done on uracil-cyclobutane dimer
anion radical by Ro¨sch48 show that the charge is localized on
only one of the pyrimidine residues. Furthermore the negative
charge is localized mostly on O4 (i.e. the carbonyl group R to
the 5,6-bond). In csDMTD this charge is held next to the
unreduced carbonyl group. The unreduced carbonyl has the
negative part of its dipole directly across from the anion site.
This results in an unfavorable charge-dipole interaction which
destabilizes the anion radical and thus makes Ered more negative.
In the tsDMTD the unreduced carbonyl group is held further
away from the anion radical center, and the favorable inductive
effects prevail.
Conclusion
Three general trends in the reduction potentials are identified
here. First, it is shown that 5,6-dihydropyrimidines undergo
one-electron reduction more easily than their unsaturated
analogs. This is attributed to aromatic stabilization which is
present in the parent pyrimidines but not in the 5,6-dihydro
derivatives. Second, the cis-syn-cyclobutane dimer of dimeth-
ylthymine is more difficult to reduce than its trans-syn diaste-
reomer. In the cis-syn case, the corresponding anion radical
apparently suffers from a destabilizing charge-dipole interaction
that balances, and in fact exceeds, the stabilization afforded by
the inductively withdrawing carbonyl groups. Finally it is
shown that protic solvents significantly stabilize the anion radical
of 5,6-dihydro-1,3,5-trimethyluracil, presumably through a
specific hydrogen-bonding interaction.
Methyl substitution has a much less pronounced effect on
the ability of the modified bases to accept an electron.
Comparison of the three 5,6-unsaturated pyrimidines, DMU,
DMT, and TMU, reveal no change in Ered beyond experimental
uncertainty. With the dihydro systems DMTH2 shows a more
negative Ered than either DMUH2 or TMUH2. Methyl groups
are weak, inductive electron-donating groups. It could be argued
that the 5-methyl group in DMTH2 is located more closely to
the site of higher charge density (i.e. the O4 in the anion radical)
than it is in TMUH2 (where it is in the 6-position) or DMUH2
(which lacks the methyl group). A similar effect is not observed
in the 5,6-unsaturated systems because the negative charge in
the anion radical state is presumably much delocalized over the
entire aromatic ring.
Acknowledgment. This work is supported by the National
Institues of Health (GM45856-01A1). We thank Arun Prakash
for synthetic advice and Sundeep P. Mattanama for help with
recrystallization of the 5,6-dihydro compounds.
Supporting Information Available: List of sensitizers
employed along with their photophysical parameters and their

Photoinduced Electron Transfer to Pyrimidines
J. Phys. Chem. A, Vol. 101, No. 24, 1997 4337
(23) Go¨rner, H. J. Photochem. Photobiol. B 1994, 26, 117-139.
(24) Zee-Cheng, K.-Y.; Robins, R. K.; Cheng, C. C. In Synthetic
Procedures in Nucleic Acid Chemistry; Zorbach, W. W., Timpson, R. S.,
Ed.; Interscience: New York, 1968; Vol. 1.
(25) Burkhalter, J. H.; Scarborough, H. C. J. Am. Pharm. Assoc. 1955,
44, 545-550.
(26) Johnson, T. B.; Clapp, S. H. J. Biol. Chem. 1908, 5, 49-70.
(27) Ainley, A. D.; Curd, F. H. S.; Hepworth, W.; Murray, A. G.; Vasey,
C. H. J. Chem. Soc. 1953, 59-70.
(28) Weinblum, D.; Ottensmeyer, F. P.; Wright, G. F. Biochim. Biophys.
Acta 1968, 155, 24-31.
(29) Kloepfer, R.; Morrison, H. J. Am. Chem. Soc. 1972, 94, 255-264.
(30) Turro, N. J. Modern Molecular Photochemistry; Benjamin/
Cummings: Menlo Park, CA, 1978.
kq values for each of the quenchers (1 page). Ordering
information is given on any current masthead page.
References and Notes
(1) Sonntag, C. v. The Chemical Basis of Radiation Biology; Taylor
& Francis: Philadelphia, 1987.
(2) Becker, D.; Sevilla, M. D. AdV. Radiat. Biol. 1993, 17, 121-180.
(3) O’Neill, P.; Fielden, E. M. AdV. Radiat. Biol. 1993, 17, 53-120.
(4) Steenken, S.; Todo, J. P.; Novais, H. M.; Candeias, L. P. J. Am.
Chem. Soc. 1992, 114, 4701-4709.
(5) Wang, W.; Sevilla, M. D. Radiat. Res. 1994, 138, 9-17.
(6) Close, D. M. Radiat. Res. 1993, 135, 1-15.
(7) Sevilla, M. D.; Besler, B.; Colson, A.-O. J. Phys. Chem. 1995, 99,
1060-1063.
(31) Scannell, M. P.; Fenick, D. J.; Yeh, S.-R.; Falvey, D. E. J. Am.
Chem. Soc. 1997, 119, 1971-1975.
(32) Yeh, S.-R.; Falvey, D. E. J. Am. Chem. Soc. 1992, 114, 7313-
7314.
(33) Legros, B.; Vandereecken, P.; Soumillion, J. P. J. Phys. Chem. 1991,
95, 4752-4761.
(34) Murphy, S.; Schuster, G. B. J. Phys. Chem. 1995, 99, 511-515.
(35) Murphy, S. T.; Zou, C.; Miers, J. B.; Ballew, R. M.; Dlott, D. D.;
Schuster, G. B. J. Phys. Chem. 1993, 97, 13252-13157.
(36) Eberson, L. Electron Transfer Reactions in Organic Chemistry;
Springer-Verlag: Berlin, Germany, 1987.
(37) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of Photochem-
istry, 2nd ed.; Marcel Dekker Inc.: New York, 1993.
(38) Rehm, D.; Weller, A. Isr. J. Chem. 1970, 8, 259-271.
(39) Rehm, D.; Weller, A. Ber. Bunsen-Ges Phys. Chem. 1969, 73, 834-
839.
(8) Cullis, P. M.; Evans, P.; Malone, M. E. J. Chem. Soc. Chem.
Commun. 1996, 985-986.
(9) Hall, D. B.; Holmlin, E.; Barton, J. K. Nature 1996, 382, 731-
735.
(10) Arkin, M. R.; Stump, E. D. A.; Holmlin, R. E.; Barton, J. K.;
Ho¨rmann, A.; Olson, E. J. C.; Barbara, P. F. Science (Washington, DC)
1996, 271, 475-480.
(11) Dandliker, P.; Holmlin, R. E.; Barton, J. K. Science (Washington,
DC) 1997, 275, 1465-1468.
(12) Lincoln, P.; Tuite, E.; Norde´n, B. J. Am. Chem. Soc. 1997, 119,
1454-1455.
(13) Olson, E. J. C.; Lu, D.; Hu¨rmann, A.; Barbara, P. F. J. Phys. Chem.
B 1997, 101, 299-303.
(14) Sancar, A. Biochemistry 1994, 33, 2-9.
(15) Begley, T. P. Acc. Chem. Res. 1994, 27, 394-401.
(16) Park, H.-W.; Kim, S.-T.; Sancar, A.; Diesenhofer, J. Science
(Washington, DC) 1995, 268, 1866-1872.
(40) Closs, G.; Miller, J. R. Science (Washington, DC) 1988, 240, 440-
447.
(41) Wasielewski, M. R.; Niemczyk, M.; Svec, W. A.; Pewitt, E. B. J.
Am. Chem. Soc. 1985, 107, 1080-1082.
(17) Kim, S.-T.; Sancar, A.; Essenmacher, C.; Babcock, G. T. J. Am.
Chem. Soc. 1992, 114, 4442-4443.
(42) Arnold, B. R.; Noukakis, D.; Farid, S.; Goodman, J. L.; Gould, I.
R. J. Am. Chem. Soc. 1995, 117, 4399-4400.
(18) Kim, S.-T.; Volk, M.; Rousseau, G.; Heelis, P. F.; Sancar, A.;
Michel-Beyerle, M.-E. J. Am. Chem. Soc. 1994, 116, 3115-3116.
(19) Pouwels, P. J. W.; Hartman, R. F.; Rose, S. D.; Kaptein, R.
Photochem. Photobiol. 1995, 61, 575-583.
(43) Kikuchi, K.; Niwa, T.; Takahashi, Y.; Ikeda, H.; Miyashi, T. J.
Phys. Chem. 1993, 97, 5070-5073.
(44) Tachiya, M.; Murata, S. J. Phys. Chem. 1992, 96, 8441-8445.
(45) Creutz, C.; Sutin, N. J. Am. Chem. Soc. 1977, 99, 241-243.
(46) Denny, E. J.; Mooney, B. J. Chem. Soc. B 1968, 1410-1415.
(47) Jordan, K. D.; Burrow, P. D. Acc. Chem. Res. 1978, 11, 341-348.
(48) Voityuk, A. A.; Michel-Beyerle, M.-E.; Ro¨sch, N. J. Am. Chem.
Soc. 1996, 118, 9750-9758.
(20) Podmore, I. D.; Heelis, P. F.; Symons, M.; Pezeshk, A. J. Chem.
Soc., Chem. Commun. 1994, 1005-1006.
(21) Fenick, D. J.; Carr, H. S.; Falvey, D. E. J. Org. Chem. 1995, 60,
624-631.
(22) Cadet, J.; Vigny, P. In Bioorganic Photochemistry; Morrison, H.,
Ed.; Wiley: New York, 1990; pp 1-272.