Distance dependence of bidirectional concerted proton-electron transfer in phenol-ru(2,2'-bipyridine)32+ dyads

Article abstract of DOI:10.1002/chem.201304256

Proton-coupled electron transfer (PCET) was investigated in three covalent donor-bridge-acceptor molecules with different bridge lengths. Upon photoexcitation of their Ru(bpy)₃²⁺ (bpy=2,2,-bipyridine) photosensitizer in acetonitrile, intramolecular long-range electron transfer from a phenolic unit to Ru(bpy)₃²⁺ occurs in concert with release of the phenolic proton to pyrrolidine base. The kinetics of this bidirectional concerted proton-electron transfer (CPET) reaction were studied as a function of phenol-Ru(bpy)₃²⁺ distance by increasing the number of bridging p-xylene units. A distance decay constant (β) of 0.67±0.23 A^-1 was determined. The distance dependence of the rates for CPET is thus not significantly steeper than that for ordinary (i.e., not proton coupled) electron transfer across the same bridges, despite the concerted motion of oppositely charged particles into different directions. Long-range bidirectional CPET is an important reaction in many proteins and plays a key role in photosynthesis; our results are relevant in the context of photoinduced separation of protons and electrons as a means of light-to-chemical energy conversion. This is the first determination of β for a bidirectional CPET reaction. Time for a concert! The dependence of the rates for bidirectional concerted proton-electron transfer (CPET) on the electron donor/electron acceptor distance was determined for the first time (see scheme). The results are relevant in the context of photodriven separation of protons and electrons across natural or artificial membranes as a means of light-to-chemical energy conversion.

Full text of DOI:10.1002/chem.201304256

DOI: 10.1002/chem.201304256
Full Paper
&
Proton-Coupled Electron Transfer
Distance Dependence of Bidirectional Concerted Proton–Electron
2+
Transfer in Phenol-Ru(2,2’-bipyridine)3 Dyads
Jing Chen, Martin Kuss-Petermann, and Oliver S. Wenger*[a]
Abstract: Proton-coupled electron transfer (PCET) was inves-
tigated in three covalent donor–bridge–acceptor molecules
with different bridge lengths. Upon photoexcitation of their
mined. The distance dependence of the rates for CPET is
thus not significantly steeper than that for ordinary
(i.e., not proton coupled) electron transfer across the same
bridges, despite the concerted motion of oppositely charged
particles into different directions. Long-range bidirectional
CPET is an important reaction in many proteins and plays
a key role in photosynthesis; our results are relevant in the
context of photoinduced separation of protons and elec-
trons as a means of light-to-chemical energy conversion.
This is the first determination of b for a bidirectional CPET
reaction.
2+
Ru(bpy)3 (bpy=2,2’-bipyridine) photosensitizer in acetoni-
trile, intramolecular long-range electron transfer from a phe-
2+
nolic unit to Ru(bpy)3 occurs in concert with release of the
phenolic proton to pyrrolidine base. The kinetics of this bi-
directional concerted proton–electron transfer (CPET) reac-
2+
tion were studied as a function of phenol–Ru(bpy)3 dis-
tance by increasing the number of bridging p-xylene units.
A distance decay constant (b) of 0.67ꢀ0.23 ꢀꢁ1 was deter-
Introduction
In many cases of “simple” electron transfer the distance de-
pendence of reaction rates is dominated by the electronic cou-
pling term (HAB),[1,5] but in other cases changes in reorganiza-
tion energy (l) or reaction free energy (DG0) with increasing
donor–acceptor distance are more important.[12,13] It is not
The distance dependence of “simple” electron transfer has
been thoroughly explored in proteins,[1] DNA,[2,3] self-assembled
monolayers,[4] and donor–bridge–acceptor molecules.[5,6] An
important open question is how the distance between an elec-
tron donor and an electron acceptor affects the rates of
proton-coupled electron transfer (PCET) reactions, particularly
when electron and proton transfer occur in concerted fashion
and when the electron and the proton are transferred into dif-
ferent directions. Long-range PCET reactions of this type are
biologically relevant, for instance in photosystem II, in ribonu-
cleotide reductase, and in DNA.[7–9] We are unaware of prior in-
vestigations of the distance dependence of bidirectional con-
certed proton-electron transfer (CPET) reactions.
a priori clear which one of the three factors (HAB,PCET, lPCET
,
DG0PCET) will dominate the distance dependence of a bidirec-
tional CPET reaction. On the one hand, increasing the electron
donor/electron acceptor distance is expected to have a signifi-
[11]
cant influence on HAB,PCET
.
On the other hand, it is known that
bidirectional CPET reactions can be associated with significant-
ly larger reorganization energies (lPCET) than “simple” electron-
transfer reactions,[14] and significant changes in lPCET with dis-
tance could strongly influence CPET rates.
"
#
À
Á
2
DG0PCET þ lPCET
4lPCETRT
2p
ꢀh
1
The rate constant for a PCET reaction can be expressed in
a similar way as the rate constant for “simple” electron transfer,
that is, in analogy to what is commonly known as the Marcus
equation.[10] In Equation (1), HAB,PCET is the electronic coupling
between the PCET reactant and product states, lPCET is the reor-
ganization energy associated with the overall PCET process,
2
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
kPCET
¼
HAB;PCET
exp ꢁ
ð1Þ
4plPCETRT
In the work presented herein we aimed to explore these basic
2+
aspects by using the phenol–Ru(bpy)3 (bpy=2,2’-bipyridine)
dyads shown in Scheme 1. They are comprised of a phenolic
and DG0
is the reaction free energy.[11] The interplay be-
PCET
tween these three factors governs the PCET rate constant
(kPCET).
[a] Dr. J. Chen,+ M. Kuss-Petermann,+ Prof. Dr. O. S. Wenger
Department of Chemistry, University of Basel
St. Johanns-Ring 19, 4056 Basel (Switzerland)
Fax: (+41)61-267-09-76
E-mail: oliver.wenger@unibas.ch
[+] These authors contributed equally to this work.
Scheme 1. Molecular structures of the PhOH-xyn-Ru2+ molecules (n=1–3) in-
vestigated in this work and illustration of their light-induced bidirectional
CPET reaction involving pyrrolidine. ET=electron transfer, PT=proton
transfer.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201304256.
Chem. Eur. J. 2014, 20, 4098 – 4104
4098
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
unit acting as a combined electron–proton donor, p-xylene
bridges of variable length, and a ruthenium photosensitizer as
an electron acceptor. The proton acceptor is a base (pyrroli-
dine) which is added separately to the solution. Phenols have
received much attention for mechanistic PCET studies,[15–17] and
there have been several investigations of donor–bridge–ac-
ceptor molecules incorporating phenols.[18–24] Long-range PCET
reactions have received increasing attention recently,[24–28] but
to our knowledge the dependence of the rates for bidirection-
al CPET on the electron donor/electron acceptor distance has
never been investigated before. Given the biological relevance
of this reaction type a systematic investigation of the model
compounds from Scheme 1 seemed to be a worthy research
endeavor.
rates because it is too endergonic. Pyrrolidine does not deprot-
onate our charge-neutral phenols to a significant extent (see
pKa values below).
The situation changes completely when base is present be-
C
cause the neutral phenoxyl radical (PhO ) instead of phenoxyl
radical cation (PhOH+) can be formed upon photoexcitation of
2+
the Ru(bpy)3 photosensitizers. In other words, phenol oxida-
tion becomes thermodynamically possible when combined
with proton release (see below for driving-force estimates). In
acetonitrile solution, addition of 4-aminopyridine quenches the
3MLCT excited state of the PhOH-xy1-Ru2+ dyad but has no
effect on the longer congeners. The stronger base pyrrolidine
leads to a photoreaction in all three dyads. We have therefore
chosen pyrrolidine as a base for our PCET investigations. Im-
portantly, up to concentrations of 1M in CH3CN pyrrolidine
3
does not have any noticeable influence on the MLCT decay of
2+
the Ru(bpy)3 reference complex (Figure S2 in the Supporting
Results and Discussion
Information).
Basic photochemistry and electrochemistry of the dyads
In pure acetonitrile photoexcitation of the three dyads from
Photochemistry in presence of pyrrolidine
3
2+
Scheme 1 leads to MLCT emission of the Ru(bpy)3 moieties,
but there is no photochemistry. In absence of any base the
3MLCT excited state lifetimes of the dyads are within experi-
mental accuracy the same as for isolated Ru(bpy)32+. Although
a significantly better oxidant in the 3MLCT state than in the
Transient absorption spectra of the three dyads in aerated
CH3CN obtained in presence of excess pyrrolidine are shown in
Figure 1a (solid black lines). These spectra were measured
2+
electronic ground state, Ru(bpy)3 is obviously not strongly
oxidizing enough to convert the phenols to phenoxyl radical
cations (PhOH+). This finding can be understood on the basis
of the electrochemical data in Table 1 which were extracted
Table 1. Electrochemical potentials (in [V] vs. Fc+/Fc in CH3CN with 0.1M
Bu4NPF6) of the individual redox-active moieties of the dyads and some
reference molecules.
ꢁ
RuIII/II
bpy/bpyꢁ
PhOH+/PhOH
PhO /PhO
C
Molecule
PhOH-xy1-Ru2+
PhOH-xy2-Ru2+
PhOH-xy3-Ru2+
Ru(bpy)3
2,4,6-tBu2PhOH
0.90
0.90
0.90
0.90
ꢁ1.73
ꢁ1.72
ꢁ1.72
ꢁ1.73
0.88
0.90
0.86
ꢁ0.54
ꢁ0.58
ꢁ0.58
2+
1.18[a]
ꢁ0.70[a]
Figure 1. a) Solid black lines: Transient absorption spectra measured after
excitation of the three dyads in aerated CH3CN (2ꢂ105 M) in presence of pyr-
rolidine at 532 nm. Detection occurred by time averaging over a period of
200 ns starting immediately after excitation with pulses of ca. 10 ns duration.
The pyrrolidine concentrations were: 25 mM (PhOH-xy1-Ru2+), 96 mM
(PhOH-xy2-Ru2+), and 3M (PhOH-xy3-Ru2+). Dashed grey traces: Difference
spectra obtained by subtraction of the solid black traces in Figure 1b from
the dotted grey traces in Figure 1b. b) Solid black traces: UV/Vis absorption
of the dyads in pure CH3CN. Dotted grey traces: UV/Vis spectra of the depro-
tonated dyads in CH3CN.
These data were extracted from the cyclic voltammograms shown in Fig-
ure S1 in the Supporting Information. [a] From Ref. [15].
from cyclic voltammetry measurements (Figure S1 in the Sup-
2+
porting Information). Reduction of Ru(bpy)3 is ligand based
and occurs at approximately ꢁ1.7 V vs. Fc+/Fc in CH3CN in all
cases considered here. Oxidation of phenol (PhOH) to phenox-
yl radical cation (PhOH+) occurs at approximately 0.9 V vs. Fc+
/Fc in the three dyads. Given a 3MLCT energy (DEMLCT) of
2.1 eV,[29] one estimates a free energy of DG0ET1 =ꢁeꢂ[E(bpy/
bpyꢁ)ꢁE(PhOH+/PhOH)]ꢁDEMLCT ꢂ +0.5 eV for photoinduced
electron transfer.[30] Such driving-force estimates are usually ac-
curate to ꢀ0.1 eV. In the present case it seems safe to con-
clude that “simple” electron transfer from phenol to photoex-
2+
after selective excitation of the Ru(bpy)3 moieties at 532 nm.
The dyad concentrations were always 2ꢂ10ꢁ5 M, but with in-
creasing bridge length increasing pyrrolidine concentrations
were necessary in order to induce efficient photochemistry. For
PhOH-xy1-Ru2+ a pyrrolidine concentration of 25 mM was suffi-
cient whereas for PhOH-xy2-Ru2+ and PhOH-xy3-Ru2+ concen-
trations of 96 mM and 3M, respectively, were employed.
2+
cited Ru(bpy)3 in pure CH3CN does not occur at measurable
Chem. Eur. J. 2014, 20, 4098 – 4104
www.chemeurj.org
4099
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Under these conditions, conversion to the photoproducts was
essentially complete after the 10 ns duration of the laser exci-
tation pulses in all three cases. Measurement of the transient
absorption data in Figure 1a occurred by time integration over
a period of 200 ns. The three resulting spectra are relatively
similar to each other and feature a prominent band near
365 nm in addition to weaker bands at longer wavelengths.
The shape of the transient absorption spectra can be under-
stood readily on the basis of the UV/Vis data shown in Fig-
ure 1b. The solid black traces in Figure 1b are the absorption
spectra of the PhOH-xyn-Ru2+ dyads in pure CH3CN, whereas
the dotted grey traces are the absorption spectra of their de-
protonated forms (PhOꢁ-xyn-Ru2+). Deprotonation occurred by
addition of excess tetrabutylammonium hydroxide to the
CH3CN solutions. When subtracting the solid black traces from
the dotted grey traces in Figure 1b, one obtains the difference
spectra shown as dashed grey traces in Figure 1a. The similari-
ty between these derived difference spectra and the experi-
mental transient absorption spectra (solid black lines in Fig-
ure 1a) is evident, and we conclude that the photoproducts
which are detected on the nanosecond timescale are com-
(pyr/PhOH+-xyn-Ru+ state at 2.6 eV) is endergonic by 0.5 eV.
The absence of any photochemical reaction in pure CH3CN
without base strongly suggests that the pyr/PhOH+-xyn-Ru+
photoproduct is never formed, not even as a short-lived inter-
mediate when base is present. However, on the basis of Equa-
tion (2) photoinduced CPET producing protonated pyrrolidine
+
C
(pyrH ), phenoxyl radical (PhO ), and reduced ruthenium com-
plex (Ru+) is estimated to be exergonic by approximately
0.8 eV.
DG0 ¼ 0:059 eV ꢃ ½pKaðPhOHþÞꢁpKaðpyrHþÞꢄ
ð2Þ
PT1
Specifically, Equation (2) is used to estimate the free energy as-
sociated with protonation of pyrrolidine by phenoxyl radical
cation.[32] The pKa of PhOH+ in CH3CN is ꢁ3 and the pKa of the
conjugate acid of pyrrolidine (pyrH+) in CH3CN is 19.56
(Table 2),[15,33] resulting in DG0PT1 ꢂꢁ1.3 eV. Consequently, the
Table 2. Acidity constants in CH3CN.
2+
prised of the deprotonated phenols and Ru(bpy)3 in the elec-
Acid
pKa
tronic ground state. It is as if the three dyads from Scheme 1
merely acted as photoacids.
pyrH+
PhOH
PhOH+
19.56[a]
28[b]
ꢁ3[b]
2+
Excitation at 532 nm occurs selectively into the Ru(bpy)3
moieties. The phenols have insignificant extinction at this
wavelength, and it is not obvious why the phenolic units
[a] From Ref. [33]. [b] Values for 2,4,6-tBu3PhOH from Ref. [15].
2+
should become more acidic upon excitation of the Ru(bpy)3
moieties which are more than 12 ꢀ away. However, the appar-
ent photoacid behavior can be explained by a reaction se-
quence of concerted proton–electron transfer (CPET) followed
by “simple” electron transfer (back-ET) in the reverse direction
(Scheme 2).[31] In order to understand this reaction sequence it
+
+
C
pyrH /PhO -xyn-Ru state must be approximately 1.3 eV below
the pyr/PhOH+-xyn-Ru+ state, that is, at 2.6 eVꢁ1.3 eV=1.3 eV.
3
This is 0.8 eV below the initially populated MLCT state. Thus,
the CPET reaction in which phenol oxidation by photoexcited
2+
Ru(bpy)3
occurs in concert with release of the phenolic
proton to pyrrolidine (arrow marked with “CPET” in Scheme 2)
has DG0CPET =ꢁ0.8 eV. It seems reasonable to conclude that
this is indeed the first reaction step occurring after photoexci-
tation.
+
+
C
The pyrH /PhO -xyn-Ru photoproduct at 1.3 eV is com-
+
posed of a strongly reducing ruthenium complex (Ru(bpy)3
)
C
connected to a relatively good electron acceptor (PhO ). Using
Equation (3), one can estimate the reaction free energy associ-
+
ated with intramolecular electron transfer from Ru(bpy)3 to
C
PhO .
ꢁ
ꢁ
DG0 ¼ ꢁe ꢃ ½EðPhO =PhO ÞꢁEðbpy=bpy Þꢄ
C
ð3Þ
Scheme 2. Energy-level diagram established on the basis of the electro-
chemical potentials in Table 1 and the acidity constants in Table 2 (see the
Supporting Information for details).
ET2
With the electrochemical potentials in Table 1, Equation (3)
yields DG0ET2 ꢂꢁ1.1 eV. In other words, after initial CPET ther-
+
C
is necessary to work out the thermodynamics of the different
photochemical reaction pathways for the PhOH-xyn-Ru2+ dyads
in presence of pyrrolidine (pyr).
mal electron transfer from Ru(bpy)3 to PhO is exergonic by
1.1 eV. This second reaction step (downward arrow in
Scheme 2 labeled “back-ET”) leads to the photoproducts which
are detected by transient absorption spectroscopy in Fig-
The starting point in the energy-level diagram of Scheme 2
is the 3MLCT state 2.1 eV above the ground state.[29] As dis-
cussed above, “simple” photoinduced electron transfer to pro-
duce phenoxyl radical cations and reduced ruthenium complex
2+
ure 1a, that is, to Ru(bpy)3 in the ground state and pheno-
late. Using the thermodynamic cycle described above, one ar-
rives at the conclusion that the pyrH+/PhOꢁ-xyn-Ru2+ state is
Chem. Eur. J. 2014, 20, 4098 – 4104
www.chemeurj.org
4100
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
approximately 0.2 eV above the ground state. Alternatively,
using the pKa values of phenol (28)[15] and the conjugate acid
of pyrrolidine (19.56)[33] in CH3CN one arrives at the conclusion
that the respective state is 0.5 eV above the ground state
(DG0PT2). The deviation between the two energy estimates for
the same state (0.2 eV vs. 0.5 eV) simply reflects the approxi-
mate character of these thermodynamic considerations and
the uncertainty in the electrochemical potentials and acidity
constants.
The conclusion from this section is that the apparent photo-
acid behavior of the three dyads in presence of pyrrolidine is
due to a sequence of two consecutive reaction steps: Photoin-
duced CPET followed by electron transfer in the reverse direc-
Figure 2. Temporal evolution of the transient absorption signal of PhOH-xy1-
Ru2+ (2ꢂ10ꢁ5 M) in aerated CH3CN at 450 nm after excitation at 532 nm
with pulses of ca. 10 ns duration. Pyrrolidine concentrations are given in the
inset.
+
tion (two thick grey arrows in Scheme 2).[31] The pyrH /PhO -
C
xyn-Ru+ intermediate (at 1.3 eV) produced after initial CPET es-
capes detection because it decays more rapidly than it is
formed.[34] In other words, CPET is the rate-determining step
leading to the observable pyrH+/PhOꢁ-xyn-Ru2+ photoprod-
ucts. This view is consistent with previously reported rates ex-
ceeding 109 sꢁ1 for “simple” electron transfer over three adja-
cent p-phenylene units as well as with the relative sluggishness
these bleach recovery times essentially reflect the kinetics of
the CPET reaction.
The observed rate constants (kobs, extracted from single ex-
ponential fits to the data in Figure 2 and Figure S3 in the Sup-
porting Information) are plotted as a function of pyrrolidine
concentration in Figure 3. Squares, circles, and triangles are
used to visually set apart data for dyads with 1, 2, and 3 p-
xylene spacers; open symbols represent data for ordinary
dyads, filled symbols represent data obtained for their deuter-
ated analogs in presence of deuterated pyrrolidine. In this con-
text “deuterated” means replacement of the phenolic proton
3+
of CPET reactions in related phenol-Ru(bpy)3 and phenol–
rhenium(I) systems.[14,23,24,35–38] Some of the H/D kinetic isotope
effects reported below provide additional support for the con-
clusion that CPET is rate-determining.
Kinetics of the photoreaction in presence of pyrrolidine
In order for the photoinduced CPET reaction to occur, pyrroli-
dine must be in immediate proximity to the phenolic units be-
cause the proton cannot tunnel very far.[39] It is therefore rea-
sonable to expect CPET to occur predominantly from the hy-
drogen-bonded encounter adducts on the right-hand side of
the chemical equilibrium in Equation (4), in analogy to prior
studies of PCET with hydrogen-bonded phenols.[22,31,39–46]
pyrþPhOH-xyn-Ru2þ Ð pyr ꢅ ꢅ ꢅ PhOH-xyn-Ru2þ
ð4Þ
Any analysis of the CPET kinetics must take this equilibrium
into account. The CPET kinetics of the three dyads can be mea-
sured by monitoring the transient absorption signal at 450 nm
as a function of time after excitation at 532 nm with laser
pulses of 10 ns width. Exemplary bleach recoveries for the
PhOH-xy1-Ru2+ dyad are shown in Figure 2, all others are re-
ported in Figure S3 in the Supporting Information. Immediately
after the laser pulses one detects a bleach at 450 nm due to
3
Figure 3. Experimentally observable MLCT-depopulation rate constants
(kobs) as a function of pyrrolidine concentration for the three dyads and their
deuterated analogues. The actual experimental data are shown in Figure 2
and Figure S3 in the Supporting Information. The solid lines are fits to the
experimental data with Equation (5) as described in the text. Results from
these fits are reported in Table 3.
1
the disappearance of the MLCT absorption as a consequence
2+
of Ru(bpy)3 excitation.[47] In absence of pyrrolidine the bleach
by a deuteron and replacement of the easily exchangeable Nꢁ
H proton of pyrrolidine by a deuteron. The most important ob-
servation is that upon bridge lengthening increasing pyrroli-
dine concentrations are necessary to induce similar reaction ki-
netics. This is a manifestation of the fact that the CPET reaction
rate decreases with increasing bridge length.
recovery time is about 200 ns in aerated CH3CN for all three
3
dyads, corresponding to relaxation of the MLCT state to the
ground state like in free Ru(bpy)32+. In presence of pyrrolidine
the bleach recovery becomes significantly more rapid
(Figure 2, Figure S3 in the Supporting Information) due to the
sequence of photoreactions (CPET, back-ET) discussed in the
previous paragraph. Since the CPET step is rate determining,
The experimentally determined bleach recovery rate con-
3
stants (kobs) are principally a function of the intrinsic MLCT ex-
Chem. Eur. J. 2014, 20, 4098 – 4104
www.chemeurj.org
4101
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
cited-state decay rate constant (k0) and the rate constant for
CPET (kCPET). Equation (5) further contains a rate constant for
“simple” electron transfer (kET) and a kQ[base] term describing
nificant KIE is not an argument against CPET.[50] Trends in the
magnitudes of KIEs associated with PCETs are generally difficult
to rationalize because the KIE depends on many different pa-
rameters.[11]
3
direct MLCT excited-state quenching by the base.[22,40,41]
The distance dependence of the CPET rate constants is
shown in Figure 4 in semilogarithmic representation; RDA corre-
sponds to center-to-center distances between the phenol and
KA½baseꢄ
ð5Þ
kobs ¼ k0 þ kET þ kQ½baseꢄ þ k
CPET 1 þ KA½baseꢄ
2+
Ru(bpy)3 moieties as obtained by molecular modeling. Linear
3
As discussed above, in absence of pyrrolidine the MLCT life-
time of the three dyads in aerated CH3CN is within experimen-
2+
tal accuracy the same as that of isolated Ru(bpy)3 hence we
conclude that kET is negligible compared to k0. Addition of up
2+
to 1M pyrrolidine to a 2ꢂ10ꢁ5 M solution of Ru(bpy)3 in aer-
3
ated CH3CN has no influence on the MLCT emission and its
lifetime (Figure S2 in the Supporting Information). From this
we conclude that kQ ꢆ105 Mꢁ1
s
ꢁ1. Fits with Equation (5) to the
experimental data in Figure 3 are therefore possible with only
two adjustable parameters: The CPET rate constant (kCPET) and
the association constant (KA) describing the chemical equilibri-
um in Equation (4). We attempted to determine KA independ-
ently on the basis of steady-state luminescence data as de-
scribed previously (Figure S4 and S5 in the Supporting Infor-
mation) but have not been able to extract meaningful KA
values from this procedure (see the Supporting Information).[48]
Two-parameter fits to single sets of data from Figure 3
turned out to be not meaningful,[49] and therefore we decided
to fit all six data sets globally with the restriction to one
common KA value for all six cases. In other words, we assume
that the equilibrium in Equation (4) is affected neither by the
number of p-xylene spacers nor by deuteration. Our fits em-
ployed k0 =5.2ꢂ106 sꢁ1, kET =0 sꢁ1, and kQ =105 Mꢁ1 sꢁ1 as non-
adjustable parameters and produced the kCPET values reported
in the third column of Table 3 and KA =1.4ꢀ0.2Mꢁ1. The fits
are shown as solid lines in Figure 3. The quality of these fits
supports the CPET hypothesis based on the chemical equilibri-
um in Equation (4) and speaks against any stepwise proton
transfer, electron transfer mechanisms.
Figure 4. Distance dependence of kCPET. The solid lines are linear regression
fits yielding the b-values reported in the text.
regression fits to the data in Figure 4 yields distance decay
constants (b) of 0.67ꢀ0.23 ꢀꢁ1 and 0.64ꢀ0.06 ꢀꢁ1 for the
PhOH-xyn-Ru2+ and PhOD-xyn-Ru2+ series, respectively. These
b values are much in line with those extracted for “simple”
(i.e., not proton-coupled) electron tunneling across oligo-p-
xylene spacers (0.77 ꢀꢁ1, 0.52 ꢀꢁ1 [51,52]
and oligo-p-phenylene
)
bridges.[35,36,53,54] Extrapolation of these exponential fits to van
2+
der Waals contact distance between phenol and Ru(bpy)3 re-
actants (RDA =4 ꢀ) yields
k
CPET =4.5ꢂ1011 sꢁ1 for ordinary
phenol and kCPET =1.8ꢂ1011 sꢁ1 for deuterated phenol. These
rates compare favorably to those of “simple” electron transfer
between reactants in van der Waals contact.[1,10]
A value of 1.4ꢀ0.2Mꢁ1 for KA is in line with previously re-
ported equilibrium constants of hydrogen-bonded phenols in
benzonitrile.[40,41] The H/D kinetic isotope effect (KIE) of 3.91ꢀ
Conclusion
0.95 (ratio between kCPET of PhOH-xy2-Ru2+ and PhOD-xy2-Ru2+
)
The distance dependence of the rates for bidirectional CPET in
the systems investigated here does not deviate drastically from
that of ordinary (i.e., not proton-coupled) electron transfer. The
dominant contribution to this distance dependence likely
comes from HAB,PCET [Eq. (1)], which can be described as a prod-
uct of electronic coupling matrix elements for the electron
transfer (HAB,ET) and proton transfer components (HAB,PT) of the
overall CPET reaction.[11,39] Increasing the electron transfer dis-
tance decreases HAB,ET whereas HAB,PT remains constant. From
this perspective the observed similarity between the distance
dependences of bidirectional CPET (b=0.67ꢀ0.23 ꢀꢁ1) and or-
found for the dyad with two p-xylene bridging units is signifi-
cant and supports the hypothesis of a rate-determining CPET
step. The two other dyads exhibit smaller KIEs (1.26ꢀ0.40, xy1
system) and (0.95ꢀ0.44, xy3 system), but the absence of a sig-
Table 3. CPET rate constants (kCPET) as a function of phenol–ruthenium
(center-to-center) distance (RDA).
Molecule
RDA [ꢀ]
kCPET [sꢁ1
]
PhOH-xy1-Ru2+
PhOD-xy1-Ru2+
PhOH-xy2-Ru2+
PhOD-xy2-Ru2+
PhOH-xy3-Ru2+
PhOD-xy3-Ru2+
12.2
12.2
16.5
16.5
20.8
20.8
(1.07ꢀ0.14)ꢂ109
(0.85ꢀ0.12)ꢂ109
(3.35ꢀ0.41)ꢂ108
(0.86ꢀ0.10)ꢂ108
(3.41ꢀ1.22)ꢂ106
(3.58ꢀ1.14)ꢂ106
dinary electron transfer (b=0.52–0.77 ꢀꢁ1 [51,52,55]
across the
)
same type of bridges is readily understandable. Changes in re-
organization energy (lPCET, [Eq. (1)]) with increasing electron
donor/electron acceptor distance appear to be relatively
minor. It is possible that reorganization energy effects are re-
Chem. Eur. J. 2014, 20, 4098 – 4104
www.chemeurj.org
4102
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[5] V. Balzani, Electron transfer in chemistry, Wiley-VCH, Weinheim, 2001.
[6] P. P. Edwards, H. B. Gray, M. T. J. Lodge, R. J. P. Williams, Angew. Chem.
2008, 120, 6860; Angew. Chem. Int. Ed. 2008, 47, 6758.
[7] B. A. Barry, G. T. Babcock, Proc. Natl. Acad. Sci. USA 1987, 84, 7099.
[8] E. C. Minnihan, D. G. Nocera, J. Stubbe, Acc. Chem. Res. 2013, 46, 2524.
[9] A. Kumar, M. D. Sevilla, Chem. Rev. 2010, 110, 7002.
[10] R. A. Marcus, N. Sutin, Biochim. Biophys. Acta Rev. Bioenerg. 1985, 811,
265.
[11] S. Hammes-Schiffer, A. A. Stuchebrukhov, Chem. Rev. 2010, 110, 6939.
[12] B. S. Brunschwig, S. Ehrenson, N. Sutin, J. Am. Chem. Soc. 1984, 106,
6858.
[13] S. S. Isied, A. Vassilian, J. F. Wishart, C. Creutz, H. A. Schwarz, N. Sutin, J.
Am. Chem. Soc. 1988, 110, 635.
[14] M. Sjçdin, S. Styring, B. ꢀkermark, L. C. Sun, L. Hammarstrçm, J. Am.
Chem. Soc. 2000, 122, 3932.
[15] J. J. Warren, T. A. Tronic, J. M. Mayer, Chem. Rev. 2010, 110, 6961.
[16] D. R. Weinberg, C. J. Gagliardi, J. F. Hull, C. F. Murphy, C. A. Kent, B. C.
Westlake, A. Paul, D. H. Ess, D. G. McCafferty, T. J. Meyer, Chem. Rev.
2012, 112, 4016.
sponsible for the deviation from the exponential distance de-
pendence of CPET rates observed here.[12]
Since this is the first determination of b for a bidirectional
CPET reaction it is delicate to make a general statement, but it
seems that the dependence of rates for bidirectional CPET on
electron donor/electron acceptor distance is not steeper than
that of ordinary electron transfer rates, despite the concerted
motion of oppositely charged particles into different directions.
This finding is of key importance when attempting to perform
light-driven separation of electrons and protons in order to
build up charge gradients across natural or artificial mem-
branes, similar to what occurs in photosystem II. We find that
a long electron transfer distance (up to ca. 20 ꢀ) is no obstacle
to bidirectional CPET.
[17] C. Costentin, M. Robert, J.-M. Savꢃant, Acc. Chem. Res. 2010, 43, 1019.
[18] T. Irebo, M.-T. Zhang, T. F. Markle, A. M. Scott, L. Hammarstrçm, J. Am.
Chem. Soc. 2012, 134, 16247.
[19] M. Sjçdin, T. Irebo, J. E. Utas, J. Lind, G. Merenyi, B. ꢀkermark, L. Ham-
marstrçm, J. Am. Chem. Soc. 2006, 128, 13076.
Experimental Section
1H NMR spectra were measured on a 400 MHz Bruker Avance in-
1
strument. The H spectra were referenced relative to tetramethylsi-
lane by using the solvent signals as internal standards. High-resolu-
tion mass spectra were recorded on a Finnigan MAT8200 mass
spectrometer. Elemental analysis was performed using a Vario EL III
CHNS analyzer from Elementar. For cyclic voltammetry a Versa-
stat3-200 potentiostat from Princeton Applied Research was used,
and a glassy carbon disk served as a working electrode. A silver
wire was used as a counter-electrode, and a second silver wire was
employed as a quasi-reference electrode. Proper referencing of the
potentials occurred through addition of small amounts of ferro-
cene and by recording of the ferrocenium/ferrocene redox wave.
Voltage sweeps were performed with rates of 100 mVsꢁ1. Optical
absorption spectra were recorded on a Shimadzu UV-1800 instru-
ment. Steady-state luminescence was measured on a Fluorolog322
instrument from Horiba Jobin–Yvon. Transient absorption was per-
formed with an LP920-KS spectrometer from Edinburgh Instru-
ments, equipped with an R928 photomultiplier and an iCCD
camera from Andor. The frequency-doubled output from a Quantel
Brilliant b laser was used as an excitation source. The duration of
the laser excitation pulses was approximately 10 ns. Transient ab-
sorption spectra were generally recorded by time-averaging the
signal over a detection period of 200 ns. Quartz cuvettes from
Starna were employed for all optical spectroscopic experiments.
For H/D kinetic isotope effect studies, the deuterated dyads were
dissolved in pure CH3CN and deuterated pyrrolidine was added.
Synthetic protocols and product characterization data are given in
the Supporting Information.
[20] L. C. Sun, M. Burkitt, M. Tamm, M. K. Raymond, M. Abrahamsson, D. Le-
Gourriꢃrec, Y. Frapart, A. Magnuson, P. H. Kenꢃz, P. Brandt, A. Tran, L.
Hammarstrçm, S. Styring, B. ꢀkermark, J. Am. Chem. Soc. 1999, 121,
6834.
[21] T. Lachaud, A. Quaranta, Y. Pellegrin, P. Dorlet, M. F. Charlot, S. Un, W.
Leibl, A. Aukauloo, Angew. Chem. 2005, 117, 1560; Angew. Chem. Int. Ed.
2005, 44, 1536.
[22] A. A. Pizano, J. L. Yang, D. G. Nocera, Chem. Sci. 2012, 3, 2457.
[23] T. Irebo, S. Y. Reece, M. Sjçdin, D. G. Nocera, L. Hammarstrçm, J. Am.
Chem. Soc. 2007, 129, 15462.
[24] M. Kuss-Petermann, H. Wolf, D. Stalke, O. S. Wenger, J. Am. Chem. Soc.
2012, 134, 12844.
[25] O. S. Wenger, Acc. Chem. Res. 2013, 46, 1517.
[26] V. W. Manner, A. G. DiPasquale, J. M. Mayer, J. Am. Chem. Soc. 2008, 130,
7210.
[27] V. W. Manner, J. M. Mayer, J. Am. Chem. Soc. 2009, 131, 9874.
[28] J. J. Warren, A. R. Menzeleev, J. S. Kretchmer, T. F. Miller, H. B. Gray, J. M.
Mayer, J. Phys. Chem. Lett. 2013, 4, 519.
[29] D. M. Roundhill, Photochemistry and Photophysics of Metal Complexes,
Plenum Press, New York, 1994.
[30] A. Weller, Z. Phys. Chem. (Muenchen Ger.) 1982, 133, 93.
[31] M. Kuss-Petermann, O. S. Wenger, J. Phys. Chem. A 2013, 117, 5726.
[32] J. M. Mayer, Annu. Rev. Phys. Chem. 2004, 55, 363.
[33] I. Kaljurand, A. Kꢄtt, L. Soovali, T. Rodima, V. Maemets, I. Leito, I. A.
Koppel, J. Org. Chem. 2005, 70, 1019.
[34] Note that this is not a matter of temporal resolution of the used equip-
ment. It is easily possible to decelerate the overall reaction to rates well
below 108 sꢁ1 by using lower pyrrolidine concentrations. The intermedi-
ate remains undetectable under such conditions.
[35] O. S. Wenger, Chem. Soc. Rev. 2011, 40, 3538.
[36] E. A. Weiss, M. J. Ahrens, L. E. Sinks, A. V. Gusev, M. A. Ratner, M. R. Wa-
sielewski, J. Am. Chem. Soc. 2004, 126, 5577.
Acknowledgements
[37] T. Irebo, O. Johansson, L. Hammarstrçm, J. Am. Chem. Soc. 2008, 130,
9194.
[38] A. Magnuson, H. Berglund, P. Korall, L. Hammarstrçm, B. ꢀkermark, S.
Styring, L. C. Sun, J. Am. Chem. Soc. 1997, 119, 10720.
[39] M.-T. Zhang, T. Irebo, O. Johansson, L. Hammarstrçm, J. Am. Chem. Soc.
2011, 133, 13224.
This work was supported by the Swiss National Science Foun-
dation (grant number 200021 146231/1) and by the Deutsche
Forschungsgemeinschaft (IRTG-1422).
[40] L. Biczꢅk, N. Gupta, H. Linschitz, J. Am. Chem. Soc. 1997, 119, 12601.
[41] L. Biczok, H. Linschitz, J. Phys. Chem. 1995, 99, 1843.
[42] I. J. Rhile, T. F. Markle, H. Nagao, A. G. DiPasquale, O. P. Lam, M. A. Lock-
wood, K. Rotter, J. M. Mayer, J. Am. Chem. Soc. 2006, 128, 6075.
[43] I. J. Rhile, J. M. Mayer, J. Am. Chem. Soc. 2004, 126, 12718.
[44] J. N. Schrauben, M. Cattaneo, T. C. Day, A. L. Tenderholt, J. M. Mayer, J.
Am. Chem. Soc. 2012, 134, 16635.
Keywords: electron transfer · hydrogen bonds · molecular
wires · photochemistry · time-resolved spectroscopy
[1] H. B. Gray, J. R. Winkler, Proc. Natl. Acad. Sci. USA 2005, 102, 3534.
[2] B. Giese, Annu. Rev. Biochem. 2002, 71, 51.
[3] J. C. Genereux, J. K. Barton, Chem. Rev. 2010, 110, 1642.
[4] H. J. Yoon, N. D. Shapiro, K. M. Park, M. M. Thuo, S. Soh, G. M. White-
sides, Angew. Chem. Int. Ed. 2012, 51, 4658.
[45] C. Costentin, M. Robert, J. M. Savꢃant, C. Tard, Angew. Chem. 2010, 122,
3891; Angew. Chem. Int. Ed. 2010, 49, 3803.
Chem. Eur. J. 2014, 20, 4098 – 4104
www.chemeurj.org
4103
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[46] W. Herzog, C. Bronner, S. Lçffler, B. C. He, D. Kratzert, D. Stalke, A.
Hauser, O. S. Wenger, ChemPhysChem 2013, 14, 1168.
[54] B. Albinsson, M. P. Eng, K. Pettersson, M. U. Winters, Phys. Chem. Chem.
Phys. 2007, 9, 5847.
[47] P. Mꢄller, K. Brettel, Photochem. Photobiol. Sci. 2012, 11, 632.
[48] J. J. Concepcion, M. K. Brennaman, J. R. Deyton, N. V. Lebedeva, M. D. E.
Forbes, J. M. Papanikolas, T. J. Meyer, J. Am. Chem. Soc. 2007, 129, 6968.
[49] Error bars associated with such fits are unreasonably large.
[50] T. F. Markle, I. J. Rhile, J. M. Mayer, J. Am. Chem. Soc. 2011, 133, 17341.
[51] D. Hanss, O. S. Wenger, Inorg. Chem. 2008, 47, 9081.
[55] O. S. Wenger, Acc. Chem. Res. 2011, 44, 25.
Received: October 31, 2013
Published online on February 26, 2014
Please note: Minor changes have been made to this manuscript since its
[52] D. Hanss, O. S. Wenger, Inorg. Chem. 2009, 48, 671.
publication in Chemistry—A European Journal Early View. The Editor.
[53] M. T. Indelli, C. Chiorboli, L. Flamigni, L. De Cola, F. Scandola, Inorg.
Chem. 2007, 46, 5630.
Chem. Eur. J. 2014, 20, 4098 – 4104
www.chemeurj.org
4104
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim