Electron transfer from photoexcited naphthalene-1,4:5,8-bis(dicarboximide) radical anion to Mn(bpy)(CO)3X and Re(bpy)(CO)3X CO2 reduction catalysts linked via a saturated methylene bridge

Article abstract of DOI:10.1016/j.jphotochem.2018.11.047

Supramolecular systems that connect a naphthalene-1,4:5,8-bis(dicarboximide) (NDI) radical anion donor to Mn(bpy)(CO)₃Br or Re(bpy)(CO)₃Cl CO₂ reduction catalysts via a methylene bridge have been synthesized and studied by femtosecond transient visible, near-infrared and mid-infrared spectroscopy. The use of the methylene bridge to link NDI to the complexes does not affect the reduction potentials of the metal complexes. Selective photoexcitation of NDI^?? to ^2*NDI^?? results in ultrafast reduction of the bipyridine (bpy) ligands on both the Mn and Re complexes to form Mn(I)(bpy^??)(CO)₃X and Re(I)(bpy^??)(CO)₃X in near unity quantum yield, respectively. The initial formation of Mn(I)(bpy^??)(CO)₃X is unexpected based on previous electrochemical data that indicates the Mn(I) center is reduced at a more positive potential than the bpy ligand. Moreover, the rate of forward electron transfer in the Mn complex was found to be faster than in the Re complex, while the rate of the back electron transfer in the Re complex was faster than in the Mn complex.

Full text of DOI:10.1016/j.jphotochem.2018.11.047

Journal of Photochemistry & Photobiology A: Chemistry 372 (2019) 21–28
Contents lists available at ScienceDirect
Journal of Photochemistry & Photobiology A: Chemistry
journal homepage: www.elsevier.com/locate/jphotochem
Electron transfer from photoexcited naphthalene-1,4:5,8-bis(dicarboximide)
radical anion to Mn(bpy)(CO) X and Re(bpy)(CO) X CO reduction catalysts
3
3
2
linked via a saturated methylene bridge
⁎
Jose F. Martinez, Nathan T. La Porte, Michael R. Wasielewski
Department of Chemistry and Institute for Sustainability and Energy at Northwestern, Northwestern University, Evanston, IL 60208-3113, United States
A R T I C L E I N F O
A B S T R A C T
Keywords:
Supramolecular systems that connect a naphthalene-1,4:5,8-bis(dicarboximide) (NDI) radical anion donor to Mn
(bpy)(CO) Br or Re(bpy)(CO) Cl CO reduction catalysts via a methylene bridge have been synthesized and
Femtosecond spectroscopy
Electrochemistry
Electron transfer
Radical anions
3
3
2
studied by femtosecond transient visible, near-infrared and mid-infrared spectroscopy. The use of the methylene
bridge to link NDI to the complexes does not affect the reduction potentials of the metal complexes. Selective
•−
2*
•−
photoexcitation of NDI to NDI results in ultrafast reduction of the bipyridine (bpy) ligands on both the Mn
CO
2
reduction
•
−
•−
and Re complexes to form Mn(I)(bpy )(CO)
3
X and Re(I)(bpy )(CO)
3
X in near unity quantum yield, respec-
Solar energy
•
−
tively. The initial formation of Mn(I)(bpy )(CO) X is unexpected based on previous electrochemical data that
3
indicates the Mn(I) center is reduced at a more positive potential than the bpy ligand. Moreover, the rate of
forward electron transfer in the Mn complex was found to be faster than in the Re complex, while the rate of the
back electron transfer in the Re complex was faster than in the Mn complex.
1
. Introduction
of supramolecular systems for CO reduction, the acceptor is the cata-
2
lyst and the electron donor is a photosensitizer that can be selectively
In the past several decades, many research groups have focused a
excited to ultimately transfer an electron to the catalyst. For example,
II
2+
great deal of eff ;ort on harnessing solar photons to power chemical
reactions for artificial photosynthesis that produce fuels [1–7]. One way
to do so is by transforming solar energy into chemical bonds, as plants
do during photosynthesis through the transformation of carbon dioxide
into glucose [8,9]. In green plant photosynthesis, this is accomplished
by the absorption of a photon by chlorophyll that initiates a series of
electron transfer events that reduce nicotinamide adenine dinucleotide
phosphate (NADP), so that it can effectively bind a proton to give
NADPH, which serves as the reducing agent in the biochemical reac-
Ishitani and co-workers covalently linked
a
Ru (bpy)
3
photo-
sensitizer to Re(bpy)(CO)
3
X via a series of saturated alkyl (C
x
H )
y
bridges [10–12]. They demonstrated that the nature of the linking
bridge can play a major role in the efficiency and turnover number
(TON) for CO reduction of these supramolecular systems. In contrast,
2
they saw a dramatic decrease in TON by using a phenanthroline-imi-
im
II
im
2+
dazolyl ( phen) motif to link Ru (bpy)
2
(
phen)
and Re(dmb)
and Re(dmb)(CO) Cl
II
2+
(CO)
3
Cl or an ethylene bridge to link Ru (bpy)
3
3
[10]. Perutz and Whitwood utilized an amide bridge to link zinc por-
phyrins to Re analogues, and saw a decrease in TON when compared to
a saturated alkyl linkage [13]. Ishitani showed further that the decrease
in photocatalytic performance is due to the use of bridges that extend
the bpy conjugation in either the photosensitizer/donor or the catalyst.
Linkers that extend the bpy π-conjugation shift its reduction potential
to more positive potentials. As the bpy reduction potential of the cat-
alyst shifts more positive, the one-electron reduced complex that is
formed upon electron transfer from the photosensitizer has a decreased
tions that convert CO into glucose [9]. In artificial photosynthesis, this
2
can be done by utilizing solar energy to transfer electrons to catalysts
that can bind and convert carbon dioxide into liquid fuel. Natural
photosynthesis ultimately provides the design principles and demon-
strates how to eff ;ectively balance light absorption, charge separation,
and catalysis, giving chemists a blueprint for developing effective ar-
tificial supramolecular systems.
A proven strategy to replicate these processes is to use supramole-
cular donor-acceptor assemblies [10]. The simplest supramolecular
ability to bind CO
(CO) X unit is diminished. Additionally, if the bpy reduction potential
of the photosensitizer shifts more positive, electron transfer from the
2
; thus the reductive catalytic activity of the Re(bpy)
systems for CO
2
reduction consist of an electron donor (D) that is ap-
3
pended to an acceptor (A) through a linking bridge unit (B). In the case
⁎
Corresponding author.
E-mail address: m-wasielewski@northwestern.edu (M.R. Wasielewski).
https://doi.org/10.1016/j.jphotochem.2018.11.047
Received 15 October 2018; Received in revised form 26 November 2018; Accepted 30 November 2018
Available online 30 November 2018
1010-6030/ © 2018 Published by Elsevier B.V.

J.F. Martinez et al.
Journal of Photochemistry & Photobiology A: Chemistry 372 (2019) 21–28
A
B (Re) = 0.5 ± 0.1 ps, while the time constant for back electron
•
−
transfer
was
faster
for
Re(4-NDI −CH
2
-bpy)(CO) Cl,
3
B
C (Re) = 0.7 ± 0.1 ps vs.
B
C (Mn) = 1.5 ± 0.1 ps. Additionally, while
the NDI reduction potentials of the two complexes are very similar to
that of NDI itself, the double reduction of the Mn complex occurred at
∼
200 mV more positive than the double reduction of the Re complex,
as expected [27,28].
. Materials and methods
2
2.1. Materials
Dichloromethane (DCM), acetonitrile (MeCN), and N,N-di-
methylformamide (DMF) used for synthesis and spectroscopic experi-
ments were dried using a commercial system (Pure Process Technology,
Nashua, NH). For spectroscopy, DMF and MeCN were transferred under
argon into a N -filled glovebox (MBraun Unilab) for use and storage.
2
•
−
Fig. 1. Previously studied donor acceptor complexes incorporating NDI to Re
bpy)(CO) Cl or Mn(bpy)(CO) Br via a phenyl bridge.
Carbon dioxide (Research Grade) was obtained from Airgas and used
without further purification. Commercially available reagents were
purchased from Sigma-Aldrich or Oakwood Chemicals and used as re-
ceived. Compounds were reduced in the glovebox using tetrakis(di-
methylamino)ethylene (TDAE) from Tokyo Chemical Industries.
Detailed synthetic procedures and compound characterization are
shown in the Supplementary Material.
(
3
3
photosensitizer MLCT state to the catalyst will no longer be sponta-
neous and will decrease the yield of the active catalyst. For these rea-
sons, the supramolecular systems with the highest TON for photo-
catalytic CO reduction currently utilize saturated alkyl bridges to link
2
the donor/photosensitizer to the acceptor/catalyst.
Our previous work examined supramolecular systems that had ei-
2.2. Steady-state spectroscopy
ther
a naphthalene-1,4:5,8-bis(dicarboximide) (NDI) or perylene-
3
,4:9,10-bis(dicarboximide) (PDI) radical anion donor linked to Re
bpy)(CO) Cl via a phenyl bridge to form Re(bpy-Ph-NDI or PDI)
CO) Cl [14–17]. These supramolecular structures were also extended
UV–vis spectra were acquired on a Shimadzu UV-1800 spectro-
photometer at room temperature. The samples were normalized to the
greatest peak. FTIR spectra were measured on a Shimadzu IRAffinity
(
(
3
3
−
1
to form Mn analogues Mn(bpy-Ph-NDI)(CO)
3
Br, see Fig. 1 [18]. The
spectrophotometer in absorbance mode at 2 cm
resolution. Samples
NDI subunit within the supramolecular system can be selectively re-
were prepared in DMF under an argon atmosphere, contained in a li-
quid demountable cell (Harrick Scientific) with 2.0 mm thick CaF
windows and 500 μm Teflon spacers.
duced, either chemically or electrochemically to form Re(bpy-Ph-
2
•
−
•−
NDI )(CO)
3
Cl or Mn(bpy-Ph-NDI )(CO)
3
Br. The NDI radical anion
•
−
(
NDI ) has an excited state oxidation potential of −2.1 V vs. SCE upon
excitation with visible or near-infrared light [19]. This oxidation po-
2.3. Electrochemistry
II
2+
tential is more negative that of Ru (bpy)
3
or other metalorganic
+
chromophores such as Ir(ppy)
3
and Ir(ppy)
2
(bpy) [20,21]. Upon se-
Electrochemical measurements were performed using a CH
•
−
2*
•−
lective excitation of the NDI to form NDI , we observed rapid
Instruments Model 660 A electrochemical workstation. A single-com-
partment cell was used for all cyclic voltammetry experiments with a
1.0 mm diameter glassy carbon disk working electrode, a platinum
counter electrode, a silver wire pseudoreference electrode, and 0.1 M
electron transfer to Re(bpy)(CO)
3
Cl with a quantum yield of near unity
•
−
•−
to form Re(bpy -Ph-NDI°)(CO)
3
Cl or Mn(bpy -Ph-NDI°)(CO)
3
Br, the
first step in the photocatalytic reduction of CO [18,22]. However, in
2
•
−
these complexes, because the NDI chromophore was linked to the bpy
ligand with a phenyl substituent, the reduction potential of bpy was
shifted positive due to the extension of the conjugation onto the brid-
ging phenyl. For this reason, we have now investigated complexes in
tetra-n-butylammonium hexafluorophosphate (TBAPF ) as the sup-
6
porting electrolyte in DMF or MeCN. The ferrocene/ferrocenium redox
couple (0.45 V vs SCE in DMF or 0.40 V vs SCE in MeCN) [29] was used
as an internal standard. TBAPF was recrystallized twice from ethanol
6
•
−
which NDI is linked to bpy via a methylene bridge to maintain the
reducing power of the one-electron reduced catalyst.
prior to use. Electrochemical cells were shielded from light and all so-
lutions were continuously purged with argon before and during the
cyclic voltammetry measurements.
Re(bpy)(CO)
reduce CO
3
Cl and its derivatives have been shown repeatedly to
2
within donor-acceptor systems, while Mn(bpy)(CO) Br has
3
received little attention despite having similar electro- and photo-
catalytic properties as the Re analogues [10,23–27]. Only recently has
2.4. Femtosecond transient visible absorption spectroscopy
photoinduced electron transfer to a Mn(bpy)(CO)
3
Br complex been
Femtosecond transient absorption experiments were performed
employing a regeneratively amplified Ti:sapphire laser system oper-
ating at 828 nm and a 1 kHz repetition rate as previously described
[30,31]. The output of the amplifier was frequency-doubled to 414 nm
using a BBO crystal and the 414 nm pulses were used to pump a la-
boratory-built collinear optical parametric (OPA) amplifier for visible-
light excitation [32] or a commercial non-collinear optical parametric
amplifier (TOPAS-White, Light-Conversion, LLC.) for NIR excitation.
Approximately 1–3 mW of the fundamental was focused into a con-
•
−
investigated [18]. For these reasons, we sought to link NDI to both Re
bpy)(CO) Cl and Mn(bpy)(CO) Br via a CH bridge. In this report, we
show that the incorporation of the NDI to a CH bridge does not affect
the bpy reduction potential. Excitation of NDI to form NDI leads
(
3
3
2
•
−
2
•
−
2*
•−
•
−
to the rapid formation of Re(bpy −CH
the photocatalytic reduction of CO , and the quantum yield of forward
electron transfer is near unity. We also extended this study to the
analogous Mn complex to photochemically reduce Mn(bpy)(CO) Br to
X is
2
-NDI°)(CO)
3
Cl, the first step in
2
3
3
•
−
•−
Mn(bpy )(CO)
3
Br. The initial formation of Mn(I)(bpy )(CO)
tinuously rastered CaF disk to generate the UV–vis white light probe
2
unexpected based on previous electrochemical data that indicate the
Mn(I) center is reduced at a more positive potential than is the bpy
ligand. [28] The time constant of electron transfer was found to be
spanning 340–800 nm, or into a proprietary medium (Ultrafast Systems,
LLC) to generate the NIR white-light probe spanning 850–1620 nm. The
total instrument response function was 300 fs. Experiments were per-
formed at a randomized pump polarization to suppress contributions
•
−
faster in Mn(4-NDI −CH
2
-bpy)(CO)
3
Br,
A
B (Mn) = 0.3 ± 0.1 ps vs.
22

J.F. Martinez et al.
Journal of Photochemistry & Photobiology A: Chemistry 372 (2019) 21–28
from orientational dynamics. Spectral and kinetic data were collected
with a CMOS or InGaAs array detector for visible and NIR detection,
respectively, and a 8 ns pump-probe delay track (customized Helios,
Ultrafast Systems, LLC). Transient spectra were averaged for at least 3 s.
Gaps shown in the spectra are due to either scattering of the pump or
idler beam, or regions not covered by the detectors. Samples prepared
in DMF had an absorbance of 0.2–0.7 at the excitation wavelength and
were irradiated in 2 mm quartz cuvettes with 0.4–0.8 μJ/pulse focused
to ∼0.2 mm diameter spot. Samples were stirred to avoid effects of
local heating or sample degradation. The samples were prepared in the
glovebox. The fsTA data were corrected for group delay dispersion, or
chirp, and scattered light prior to the kinetic analysis using Surface
Xplorer (Ultrafast Systems, LLC). The kinetic analysis was performed
using home-written programs in MATLAB [33] and was based on a
global fit to selected single-wavelength kinetics. The time resolution is
given as w = 300 fs (full width at half-maximum); the assumption of a
uniform instrument response across the frequency domain and a fixed
time zero are implicit to global analysis. The data were globally fit ei-
ther to specified kinetic models or a sum of exponentials. The data set
was then deconvoluted with the resultant populations or amplitudes to
reconstruct the species-associated or decay-associated spectra. The
procedures and kinetic models used to fit the data have been previously
described in detail [14].
dicarboxyanhydride [14] led to the final ligand. The ligand was then
refluxed in dry dichloromethane under argon with rhenium(I)penta-
carbonylchloride or bromopentacarbonylmanganese(I) to form com-
plexes 1 and 2, respectively.
3.2. Electrochemistry
The cyclic voltammograms of complexes 1 and 2 are shown in Fig. 2
with their reduction potentials summarized in Table 1. Complex 1
displays four reduction waves with NDI having reversible first and
second reduction potentials at −0.48 and −0.98 V, respectively. The
third and fourth reductions are well separated from those of NDI as well
as from each other. The reversible reduction at −1.45 V is assigned to
bpy, while the irreversible reduction at −1.94 V is assigned to Re(I) in
the complex. Complex 2 also displays four reduction waves with the
first two reversible waves at −0.48 and −0.99 V, once again resulting
from NDI. The third and fourth reductions at −1.38 and −1.73 V are
well separated from one another but irreversible, and in contrast to the
Re complex are assigned to Mn(I) and bpy reduction, respectively [35].
These assignments are in good agreement with previous work that in-
tegrates NDI with Re(bpy)(CO) X and other electrochemical studies of
3
Re and Mn(N^N)(CO) X complexes [16,18,35].
3
3.3. Electron transfer energetics
2.5. Femtosecond time-resolved Mid-IR spectroscopy
Based on the redox potentials discussed above, the Gibbs free energy
Femtosecond transient mid-IR absorption (fsIR) spectroscopy was
for the excited state electron transfer reactions of complexes 1 and 2
performed using a commercial Ti:sapphire oscillator/amplifier (Solstice
can be estimated using the following equation:
3
.5 W, Spectra-Physics) to pump two optical parametric amplifiers
TOPAS-C, Light Conversion), one which provided a 100 fs, 605 nm
excitation pulse and the other provided 100 fs pulses at 2150–1800
•
2
•
G
ET = E (NDI /NDI) E (A• /A)
E
D1 ( *NDI ).
(1)
(
•
−
•−
where E(NDI /NDI) is the reduction potential of NDI and E(A /A) is
−
1
cm . The overall instrument response was 300 fs. The spectra were
the reduction potential of the complex of interest, and E is the energy
D1
2
*
•−
acquired with a liquid N
2
-cooled dual channel (2 × 64) MCT array
of the NDI excited state, assuming electron transfer occurs from the
•
−
detector that is coupled to a Horiba HR320 monochromator as part of a
Helios-IR spectrometer (Ultrafast Systems, LLC). Samples with a max-
imum optical density of 1.5 at the excitation wavelength were prepared
in DMF contained in a liquid demountable cell (Harrick Scientific) with
D state. The D ⟵ D transition of NDI absorbs at 785 nm, making
1
D1
1
0
E
= 1.58 eV. There is no electrostatic work term for this reaction
because the reaction is a charge shift, not a charge separation. In ad-
dition, there is no solvation correction term because the fsTA and
electrochemical experiments are performed in the same high polarity
solvent. The Gibbs free energy for the thermal back electron transfer
reactions can be estimated using the following equation:
2
.0 mm thick CaF windows and a 500 μm Teflon spacer. During data
2
acquisition, the cell was mounted and rastered on a motorized stage to
prevent sample degradation.
ET = E (A /A) E (NDI• /NDI)
•
(2)
G
3
. Results
From the equations above, the Gibbs free energies for each electron
3.1. Synthesis
transfer step are shown in Table 2 below.
Complexes 1 Re(4-NDI-CH
2
-bpy)(CO)
3
Cl and 2 Mn(4-NDI-CH
2
-bpy)
3.4. Steady-state spectroscopy
(
CO) Br were synthesized as illustrated in Scheme 1 and detailed in the
3
Supplementary Material. 4-Methyl-4′-aminomethyl-2,2′-bpy was pre-
3.4.1. Steady-State FTIR
pared as previously reported [34]. Condensation of this amine with N-
The FTIR spectra of complexes 1 and 2 were collected in solution
(Fig. 3) and the ν(CO) are tabulated in Table 3. Complexes 1 and 2 show
(
2,5-di-t-butyl)naphthalene-(1,4)-dicarboximide-(5,8)-
Scheme 1. Synthesis of complexes 1 and 2.
23

J.F. Martinez et al.
Journal of Photochemistry & Photobiology A: Chemistry 372 (2019) 21–28
Fig. 2. Cyclic voltammograms of 1.0 mM solutions of (A) 1 and (B) 2 in DMF, recorded at 100 mV s−1 at room temperature with 0.1 M TBAPF
as supporting
A″
6
electrolyte under argon.
Table 1
Table 3
Redox potentials of complexes 1 and 2 vs SCE in DMF.
CO stretching frequencies of complexes 1 and 2 in DMF.
0
/−
−/2−
Compound
NDI
NDI
3rd
4th
Compound
A′(1)
A′(2)
Re (1)
−0.48
−0.48
−0.98
−0.99
−1.45
−1.38
−1.94
−1.73
Re (1)
2019
2024
1914
1932
1891
1914
Mn (2)
Mn (2)
•
−
Table 2
absorptions disappear while the NDI absorptions appear at 471, 605,
Free energy changes for electron transfer reactions in complex 1•− and 2•−
.
•−
700, and 785 nm. [19] In the chemically reduced complexes 1 and
•
−
2
, the Re(bpy)(CO)
3
X MLCT and Mn(bpy)(CO) X LMCT bands un-
3
Compound
process
ΔG (eV)
•
−
derly the strong absorptions of NDI
.
•
•
•
•
•
•
−
−
−
−
−
−
2*
2*
2*
•−
•−
1
2
2
1
2
2
NDI → Re(bpy)(CO)
3
3
Cl
−0.61
−0.33
−0.68
−0.97
−1.25
−0.90
NDI → Mn(bpy)(CO)
Br
Br
•−
NDI → Mn(bpy)(CO)
3
•−
3.5. Electron transfer dynamics
Re(bpy )(CO)
3
3
3
Cl → NDI°
•−
Mn(bpy )(CO)
Cl → NDI°
•−
Mn (bpy)(CO)
Cl → NDI°
3.5.1. Transient absorption spectroscopy
•
−
The intramolecular electron transfer behavior of complexes 1 and
−
•
2
was investigated using pump-probe femtosecond transient absorp-
three characteristic ν(CO) stretches for facially coordinated tricarbonyl
complexes. It is known that the energies of and separation between the
two lower-energy bands depend on the axial ligand X [36]. Typically,
two well-developed bands are observed in complexes with halide li-
gands [36]. The spectra of complexes 1 and 2 are in agreement with this
observation because the two low energy bands are separated by
tion (fsTA) spectroscopy. Figs. 5 and 6 display the fsTA spectra at se-
•−
lected time delays following selective photoexcitation of NDI with a
pump pulse centered at ex = 605 nm for each complex. Excitation of
•−
•−
complexes
1
and
2
at 605 nm results in instrument-limited
•−
bleaching of the ground-state absorption of NDI at 473 nm, and the
appearance of a broad absorption spanning 410–450 nm, corresponding
−
1
1
8–23 cm
.
2*
•–
2*
•−
to the induced absorptions of NDI . The NDI absorptions decay
in < 1 ps, and are accompanied by the appearance of absorptions at
362 and 382 nm, which result from the absorption of NDI°, indicating
that electron transfer to the metal complex has occurred. For complexes
3
.4.2. Steady state UV–vis absorption
The normalized steady-state electronic absorption spectra of 1 and 2
•
−
•−
are shown in Fig. 4. In complexes 1 and 2, the NDI absorptions dom-
inate the spectrum of the unreduced compounds. In complex 1, the Re
1
and 2 , the data are best fit to an A
B
C
GS kinetic model.
The species-associated spectra corresponding to species A, B, and C and
the transient kinetics at selected wavelengths are shown in Figs. 5 and
6. The population dynamics are shown in Figs. S3–S4 and the corre-
sponding rate constants are given in Table 4.
(
bpy)(CO) X MLCT absorption band tails out from underneath the NDI
3
absorptions. This is similar for complex 2 where the Mn(bpy)(CO)
3
X
LMCT absorption band is broad and tails out to ∼500 nm. [24] When
•
−
•−
TDAE is added to complex 1 and 2, to form 1 and 2 , the NDI
Fig. 3. FTIR spectra of complexes (A) 1 and (B) 2 in DMF.
24

J.F. Martinez et al.
Journal of Photochemistry & Photobiology A: Chemistry 372 (2019) 21–28
Fig. 4. Electronic absorption spectra of complexes (A) 1 and (B) 2 in DMF with and without the TDAE reductant added.
3
.5.2. Time-resolved mid-IR spectroscopy
of complexes 1 and 2 are NDI localized and are similar to that of the
To certify that the metal complexes in 1 and 2 were indeed reduced,
free NDI chromophore. [19] The electrochemical data shown in Table 1
−1
the CO-stretching region of their IR spectra (1850–2100 cm ) was
monitored using femtosecond time-resolved mid-IR transient absorp-
tion (TRIR) experiments. TRIR spectra and kinetic traces at selected
•−
show that the attachment of NDI to bpy via a CH bridge has a negli-
2
gible effect on the reduction potentials of NDI. The third reduction in
complex 1 is reversible, while the third reduction in complex 2 is not. In
complex 1, the third reduction is attributed to bpy reduction to form Re
•
−
energies are shown below in Figs. 7 and 8. Complexes 1 and 2
•
−
exhibited similar spectral features but with different kinetics. For ex-
(bpy )(CO) Cl, while in complex 2, the third reduction is attributed to
3
-
1
−1
ample, induced absorptions appear at 1869 cm and 2001 cm
in
the reduction of the metal center to form Mn°(bpy)(CO) [28]. In
3
•
−
complex 1 following excitation at 605 nm. These induced absorptions
complex 2, the third reduction leads to the simultaneous loss of the
halide ligand to form a five-coordinate Mn° complex. On the electro-
chemical time scale (100 mV/s), rapid dimerization occurs upon the
first metal-based reduction in complex 2. This is evident in the reverse
sweep by the oxidative cleavage of the dimer complex Mn°-Mn° at
−0.18 V. In contrast, the first metal-based reduction in complex 1 is
bpy localized and the loss of the halide ligand does not occur. Upon the
fourth reduction in complex 1, the rhenium center is formally reduced,
-
decay with the same kinetics that the ground state bleaches at 1911 cm
1
-1
•−
and 2019 cm decay. As is the case for 1 , the kinetics of the induced
•
−
-1
−1
absorption decay for complex 2 at 1935 cm and 2024 cm
fol-
lowing 605 nm excitation are essentially the same as the bleach re-
-
1
−1
covery kinetics at 1895 cm and 2003 cm . The magnitudes of these
frequency shifts have been previously reported by us and by other
groups in photo- and electrochemical experiments in which Re(bpy)
•
−
(
(
CO)
3
X or Mn(bpy)(CO)
3
X are reduced by one electron to form Re
and the halide ligand simultaneously dissociates to form Re°(bpy
)
•
−
•−
bpy )(CO)
3
X or Mn(bpy )(CO)
3
X, respectively.
(CO) . Upon the fourth reduction in complex 2, the bpy is reduced, and
3
•
−
The fsIR spectra were fit to the same A→B→C→GS model as the
the dimer complex is cleaved to form Mn°(bpy )(CO) . Complexes 1
3
fsTA data, with a modification to account for the A species (associated
and 2 are freely diffusing in solution according to the Randles-Sevcik
equation (Figs. S1–S2). These assignments and observations are in
complete agreement with the literature. [25,27,28,35,37,38]
•
–
with excitation of the NDI fragment) exhibiting no change in the mid-
IR relative to the ground state and with the τA→B fixed at the value
obtained from the fsTA fit. This fitting gave τB→C = 0.4 ± 0.3 ps and
•
–
τ
C→GS = 18 ± 10 ps
for
1 ,
and
τB→C = 1 ± 0.5 ps
and
•
–
τ
C→GS = 50 ± 40 ps for 2 . The large error bars for τC→GS result from
4.2. Femtosecond transient absorption spectroscopy
the low signal-to-noise of the TRIR data, but in general the values are
similar to those obtained from the fitting the fsTA data.
•
−
The intramolecular electron transfer behavior of complex 1 and
−
•
2
was probed using transient absorption spectroscopy in the visible
region. The fsTA spectrum, the species associated spectra for each decay
component, and the multiple wavelength kinetics and fits are shown in
4
. Discussion
•
−
Figs. 5 and 6. For complex 1 , following excitation at ex = 605 nm,
4
.1. Electrochemistry
the photoinduced charge shift dynamics were fit to an A→B→C→GS
•
−
model, where A represents the promotion of NDI to its excited state
2
*
•−
In complexes 1 and 2, there are four well-separated reduction pro-
Re(4- NDI -bpy)(CO)
3
Cl, B represents the photoinduced charge-
•−
cesses in their cyclic voltammograms. The first two reduction processes
shifted species Re(4-NDI°-bpy )(CO) Cl, C potentially represents a
3
Fig. 5. (A) fsTA spectra of complex 1•− (B) species-associated spectra (C), transient kinetics at selected wavelengths. λex = 605 nm, solvent = DMF.
25

J.F. Martinez et al.
Journal of Photochemistry & Photobiology A: Chemistry 372 (2019) 21–28
Fig. 6. (A) fsTA spectra of complex 2•− (B) species-associated spectra (C), transient kinetics at selected wavelengths. λex = 605 nm, solvent = DMF.
Table 4
Observed time constants and rates of electron transfer for complexes 1 and
different effects, the distance between the NDI and Re components and
the Gibbs free energy for back electron transfer. For example, in com-
•
−
•
−
2
.
plexes 1 and 2 that are separated by a CH bridge, the NDI and metal
2
complex are separated by one carbon whereas the complexes that are
separated by a phenyl bridge are separated by four carbons. Ad-
ditionally, the Gibbs free energies for back electron transfer with the
A
B
B
(ps)
C
C
(ps)
GS
k
A
B
kB
C
kC GS
FET
(s−1)
(s−1)
(s−1)
(
ps)
1•−
0.5
±
0.7
±
8.1
±
0.2 2.0 × 1012 1.4 × 1012 1.2 × 1011 0.99
CH bridges (−0.97 V and −1.25 V) are more negative than those of
2
0
.1 ps
0.1 ps
ps
2•−
3.3 × 1012 6.6 × 1011 1.2 × 1010 0.99
the conjugated phenyl substituents (−0.78 V and −1.05 V). With re-
spect to electrochemical reduction, the bpy is easier to reduce in the
complexes with a phenyl bridge. This is expected as the extension of the
conjugation from the bpy into the phenyl substituent decreases the bpy
0.3
±
1.5
±
84.7
±
0
.1 ps
0.1 ps
2.0 ps
0
/−
structural change or vibrational cooling, GS represents the original
π* orbital energy, making the bpy
reduction more facile and shifting
•
−
•−
ground state Re(4-NDI -bpy)(CO)
3
Cl. For complex 2 , following ex-
its reduction potential more positive. In the case of the CH bridges, the
2
citation at ex = 605 nm, the photoinduced charge shift dynamics were
bpy reduction potentials in complexes 1 and 2 are more negative due to
fit to an A→B→C→GS model, where A represents the promotion of
the electron donating ability of the methyl and methylene groups at the
4 positions of bpy that increase the bpy π* orbital energy. [26,27,38]
•
−
2*
•−
NDI to its excited state Mn(4- NDI -bpy)(CO) Br, B represents the
•−
3
photoinduced charge-shifted species Mn(4-NDI°-bpy )(CO)
3
Br, C po-
tentially represents a structural change or vibrational cooling, GS re-
•
−
presents the original ground state Mn(4-NDI -bpy)(CO)
3
Br.
4.3. Time-resolved Mid-IR spectroscopy
The quantum yield of forward electron transfer was estimated by
2
*
•−
comparing the intrinsic NDI excited state lifetime ( = 141 ps) with
its lifetime in each complex [19]. In comparison to the previously
studied complexes that utilized a conjugated phenyl bridge instead of a
Time-resolved mid-IR (TRIR) spectroscopy has been used to observe
facile differentiation among the various oxidation and ligand-field
states of Re(bpy)(CO)
3
X and more recently has been employed for Mn
saturated CH
2
bridge, complexes 1 and 2 also demonstrate a quantum
(bpy)(CO) X complexes. Using TRIR spectroscopy following excitation
3
•
−
yield of forward electron transfer close to unity, indicating that the use
of NDI at 605 nm, we can determine the nature of the charge-shifted
•− •−
of a saturated CH
2
bridge is as effective as the phenyl bridge in facil-
product states. Excitation of complexes 1 and 2 leads to a 25–45 cm
shift of v(CO) to lower frequencies. These induced absorptions are as-
2
*
•−
itating a charge shift from NDI to the corresponding metal complex.
One key difference between the two types of bridges is that the back
•
−
signed to the formation of Re(4-NDI°-bpy )(CO) Cl and Mn(4-NDI°-
3
•
−
•−
electron transfer from the CH
2
bridged complexes is faster than the
bpy )(CO) Br, respectively. [39] For complex 1 , this is expected as
3
conjugated phenyl bridge. For example, in the previously studied Re
the first reduction of the metal complex is bpy localized. In contrast, for
•
−
•−
complex, the time constant for back electron transfer rate for Re(bpy
Ph-NDI°)(CO) Cl is 9.5 ± 0.7 ps. This is approximately 14 times slower
in the CH -linked complex. This is likely a manifestation of two
-
complex 2 , while the electrochemical data shows that first reduction
of the metal complex is expected to be the Mn(I) center, the TRIR data
shows that the electron is transferred initially to bpy. The induced
absorption changes that would correspond to a 5-coordinate Mn species
3
2
Fig. 7. (A) TRIR spectra of complex 1•− (B) transient kinetics and kinetic fits at selected ν(CO) of 1•−. λex = 605 nm, solvent = DMF.
26

J.F. Martinez et al.
Journal of Photochemistry & Photobiology A: Chemistry 372 (2019) 21–28
Fig. 8. (A) TRIR spectra of complex 2•− (B) transient kinetics and kinetic fits at selected ν(CO) of 2•−. λex = 605 nm, solvent = DMF.
are not observed. [40] The charge shift from bpy•− to the Mn center is
work were performed at the IMSERC at Northwestern University, which
has received support from the Soft and Hybrid Nanotechnology
Experimental (SHyNE) Resource (NSF NNCI-1542205); the State of
Illinois and International Institute for Nanotechnology (IIN).
•
−
not observed because the back-electron transfer step from bpy to NDI°
(
k
B
C
) is faster. More extensive studies are currently being pursued in
our laboratory to elucidate the nature of the singly reduced 6-co-
ordinate Mn species and the required energetics/kinetics for the for-
mation of a 5-coordinate Mn species.
References
[
[
1] M.R. Wasielewski, Photoinduced electron transfer in supramolecular systems for
artificial photosynthesis, Chem. Rev. 92 (1992) 435–461.
5
. Conclusions
2] S. Berardi, S. Drouet, L. Francàs, C. Gimbert-Suriñach, M. Guttentag, C. Richmond,
T. Stoll, A. Llobet, Molecular artificial photosynthesis, Chem. Soc. Rev. 43 (2014)
This work has several important implications for designing supra-
7
501–7519.
molecular donor-acceptor systems with Re(bpy)(CO) Cl and Mn(bpy)
3
•−
[3] J. Su, L. Vayssieres, A place in the sun for artificial photosynthesis? ACS Energy
Lett. 1 (2016) 121–135.
(
CO)
CH
transfer. In photocatalytic systems based on longer saturated linker, this
3
Br CO
2
reduction catalysts. First, incorporating NDI to bpy via a
2
bridge does not affect the quantum yield of forward electron
[4] M. Natali, F. Scandola, Supramolecular artificial photosynthesis, in: G. Bergamini,
S. Silvi (Eds.), Applied Photochemistry: When Light Meets Molecules, Springer
International Publishing, Cham, 2016, pp. 1–66.
•
−
ensures that each photon absorbed by NDI results in formation of Re
[
5] J.H. Alstrum-Acevedo, M.K. Brennaman, T.J. Meyer, Chemical approaches to arti-
ficial photosynthesis. 2, Inorg. Chem. 44 (2005) 6802–6827.
•
−
•−
(
bpy )(CO)
3
X or Mn(bpy )(CO), the first intermediate in the catalytic
•
−
cycle. Second, the attachment of the NDI -CH
2
to bpy does not shift
[6] L. Hammarström, Accumulative charge separation for solar fuels production: cou-
pling light-induced single electron transfer to multielectron catalysis, Acc. Chem.
Res. 48 (2015) 840–850.
the reduction of the bpy positive, indicating that the potential needed
•
−
for CO
2
reduction by the singly reduced complex Re(bpy )(CO) X and
3
[
[
[
7] A. Nicola, B. Vincenzo, Solar electricity and solar fuels: status and perspectives in
the context of the energy transition, Chem. A Eur. J. 22 (2016) 32–57.
8] I. McConnell, G. Li, G.W. Brudvig, Energy conversion in natural and artificial
photosynthesis, Chem. Biol. 17 (2010) 434–447.
•
−
Mn(bpy )(CO) X is maintained. This is important as extended con-
3
jugation leads to more positive reduction potentials, which, in turn,
leads to a weaker one-electron reduced species and a decrease in the
9] J. Barber, P.D. Tran, From natural to artificial photosynthesis, J. R. Soc. Interface 10
(2013).
potential required for CO reduction, a fault seen in previous con-
2
[
10] Y. Tamaki, O. Ishitani, Supramolecular photocatalysts for the reduction of CO
Catal. 7 (2017) 3394–3409.
2
, ACS
jugated chromophores. Third, surprisingly, as observed via the afore-
mentioned transient experiments, the initial reduction of the Mn com-
plex occurs at bpy, counter to what is observed in the much slower
electrochemical experiments (femtoseconds vs seconds). Moreover, the
[
11] Y. Tamaki, K. Watanabe, K. Koike, H. Inoue, T. Morimoto, O. Ishitani, Development
of highly efficient supramolecular CO reduction photocatalysts with high turnover
2
frequency and durability, Faraday Discuss. 155 (2012) 115–127.
•
−
[12] K. Koike, S. Naito, S. Sato, Y. Tamaki, O. Ishitani, Architecture of supramolecular
subsequent charge shift from bpy to Mn(I) is not observed in 2 be-
metal complexes for photocatalytic CO
chain connecting photosensitizer to catalyst, J. Photochem. Photobiol. A: Chem.
07 (2009) 109–114.
[13] C.D. Windle, M.W. George, R.N. Perutz, P.A. Summers, X.Z. Sun, A.C. Whitwood,
2
reduction: iii: effects of length of alkyl
•
−
cause the back charge shift repopulating NDI is faster. Fourth, NDI
can easily be modified to attach the complex to an electrode surface,
2
•
−
allowing for quick regeneration of NDI
to inhibit back electron
2
Comparison of rhenium-porphyrin dyads for CO photoreduction: photocatalytic
transfer, thereby rendering the lifetime of the charge-separated state
studies and charge separation dynamics studied by time-resolved IR spectroscopy,
Chem. Sci. 6 (2015) 6847–6864.
2
*
•−
sufficiently long to allow a second electron transfer from NDI to the
[
[
14] N.T. La Porte, J. Martinez, S. Hedstrom, B. Rudshteyn, B.T. Phelan, C.M. Mauck,
R.M. Young, V.S. Batista, M.R. Wasielewski, Photoinduced electron transfer from
Re or Mn complex that can result in CO binding to start the catalytic
2
cycle. These and other modifications are currently being pursued.
rylenediimide radical anions and dianions to Re(bpy)(CO)
frared light, Chem. Sci. (2017).
3
using red and near-in-
15] S. Hedstrom, S. Chaudhuri, N.T. La Porte, B. Rudshteyn, J.F. Martinez,
M.R. Wasielewski, V.S. Batista, Thousandfold enhancement of photoreduction
Competing interests
3
lifetime in Re(bpy)(CO) via spin-dependent electron transfer from a per-
ylenediimide radical anion donor, J. Am. Chem. Soc. 139 (2017) 16466–16469.
16] N.T. La Porte, J.F. Martinez, S. Chaudhuri, S. Hedstrom, V.S. Batista,
M.R. Wasielewski, Photoexcited radical anion super-reductants for solar fuels cat-
alysis, Coord. Chem. Rev. 361 (2018) 98–119.
The authors declare no competing interests.
[
[
Acknowledgements
17] J.F. Martinez, N.T. La Porte, C.M. Mauck, M.R. Wasielewski, Photo-driven electron
transfer from the highly reducing excited state of naphthalene diimide radical anion
We thank Dr. Saman Shafaie for collecting high-resolution mass
spectrometric data. This work was supported by the U.S. Department of
Energy, Office of Science, Office of Basic Energy Sciences under Award
DE-FG02-99ER14999 (M.R.W.). This publication was made possible by
NPRP Grant No. 9-174-2-092 from the Qatar National Research Fund (a
member of Qatar Foundation). The findings achieved herein are solely
the responsibility of the authors. NMR and MS measurements in this
to a CO reduction catalyst within a molecular triad, Faraday Discuss. 198 (2017)
2
2
35–249.
[
18] A. Sinopoli, N.T. La Porte, J.F. Martinez, M.R. Wasielewski, M. Sohail, Manganese
2
carbonyl complexes for CO reduction, Coord. Chem. Rev. 365 (2018) 60–74.
[19] D. Gosztola, M.P. Niemczyk, W. Svec, A.S. Lukas, M.R. Wasielewski, Excited doublet
states of electrochemically generated aromatic imide and diimide radical anions, J.
Phys. Chem. A 104 (2000) 6545–6551.
[
20] Y. Kuramochi, O. Ishitani, Iridium(III) 1-phenylisoquinoline complexes as a
27

J.F. Martinez et al.
Journal of Photochemistry & Photobiology A: Chemistry 372 (2019) 21–28
photosensitizer for photocatalytic CO
2
reduction: a mixed system with a Re(I)
radical, J. Phys. Chem. A 120 (2016) 2841–2853.
catalyst and a supramolecular photocatalyst, Inorg. Chem. 55 (2016) 5702–5709.
21] D.M. Arias-Rotondo, J.K. McCusker, The photophysics of photoredox catalysis: a
roadmap for catalyst design, Chem. Soc. Rev. 45 (2016) 5803–5820.
[32] S.R. Greenfield, M.R. Wasielewski, Near-transform-limited visible and near-IR fs
pulses from OPA using type II β-barium borate, Opt. Lett. 20 (1995) 1394–1396.
[33] I. The Mathworks, Matlab Release 2016a, The Mathworks, Inc., Natick,
Massachusetts, United States, 2016.
[
[
22] J.F. Martinez, N.T. La Porte, M.R. Wasielewski, Electron transfer from photoexcited
naphthalene diimide radical anion to electrocatalytically active Re(bpy)(CO)
a molecular triad, J. Phys. Chem. C 122 (2018) 2608–2617.
3
Cl in
[34] Z. Fang, A. Ito, S. Keinan, Z. Chen, Z. Watson, J. Rochette, Y. Kanai, D. Taylor,
K.S. Schanze, T.J. Meyer, Atom transfer radical polymerization preparation and
photophysical properties of polypyridylruthenium derivatized polystyrenes, Inorg.
Chem. 52 (2013) 8511–8520.
[
23] J. Hawecker, J.-M. Lehn, R. Ziessel, Efficient photochemical reduction of CO2 to CO
2+
by visible light irradiation of systems containing Re(bipy)(CO)
CO combinations as homogeneous catalysts, J. Chem. Soc. Chem. Commun. (1983)
36–538.
24] H. Takeda, H. Koizumi, K. Okamoto, O. Ishitani, Photocatalytic CO
a mn complex as a catalyst, Chem. Commun. (Cambridge, U.K.) 50 (2014)
491–1493.
3
X or Ru(bipy)
3
-
2
[35] M.D. Sampson, A.D. Nguyen, K.A. Grice, C.E. Moore, A.L. Rheingold, C.P. Kubiak,
Manganese catalysts with bulky bipyridine ligands for the electrocatalytic reduction
of carbon dioxide: eliminating dimerization and altering catalysis, J. Am. Chem.
Soc. 136 (2014) 5460–5471.
5
[
[
[
[
[
2
reduction using
1
[36] A. Vlček, Ultrafast excited-state processes in Re(I) carbonyl-diimine complexes:
from excitation to photochemistry, in: A.J. Lees (Ed.), Photophysics of
Organometallics, Springer, Berlin Heidelberg, Berlin, Heidelberg, 2010, pp.
115–158.
25] M. Bourrez, F. Molton, S. Chardon-Noblat, A. Deronzier, [Mn(bipyridyl)(CO)
3
Br]:
an abundant metal carbonyl complex as efficient electrocatalyst for CO
Angew. Chem. Int. Ed. 50 (2011) 9903–9906.
2
reduction,
26] J.M. Smieja, C.P. Kubiak, Re(bipy-tBu)(CO)
3
Cl-improved catalytic activity for re-
[37] S.K. Hayashi, B.S. Brunschwig, E. Fujita, Involvement of a binuclear species with the
duction of carbon dioxide: IR-spectroelectrochemical and mechanistic studies,
Inorg. Chem. 49 (2010) 9283–9289.
2
Re-C(O)O-Re moiety in CO reduction catalyzed by tricarbonyl rhenium(I) com-
plexes with diimine ligands: Strikingly slow formation of the Re-Re and Re-C(O)O-
27] J.M. Smieja, M.D. Sampson, K.A. Grice, E.E. Benson, J.D. Froehlich, C.P. Kubiak,
3
Re species from Re(dmb)(CO) S (dmb = 4,4’-dimethyl-2,2’-bipyridine, S = sol-
Manganese as a substitute for rhenium in CO
acids, Inorg. Chem. 52 (2013) 2484–2491.
2
reduction catalysts: the importance of
vent), J. Am. Chem. Soc. 125 (2003) 11976–11987.
[38] M.L. Clark, P.L. Cheung, M. Lessio, E.A. Carter, C.P. Kubiak, Kinetic and mechan-
istic effects of bipyridine (bpy) substituent, labile ligand, and brønsted acid on
28] C. Riplinger, M.D. Sampson, A.M. Ritzmann, C.P. Kubiak, E.A. Carter, Mechanistic
contrasts between manganese and rhenium bipyridine electrocatalysts for the re-
duction of carbon dioxide, J. Am. Chem. Soc. 136 (2014) 16285–16298.
29] N.G. Connelly, W.E. Geiger, Chemical redox agents for organometallic chemistry,
Chem. Rev. 96 (1996) 877–910.
electrocatalytic CO
2
reduction by Re(bpy) complexes, ACS Catal. (2018)
2021–2029.
[
[
[39] C.W. Machan, M.D. Sampson, S.A. Chabolla, T. Dang, C.P. Kubiak, Developing a
2
mechanistic understanding of molecular electrocatalysts for CO reduction using
30] R.M. Young, S.M. Dyar, J.C. Barnes, M. Juricek, J.F. Stoddart, D.T. Co,
M.R. Wasielewski, Ultrafast conformational dynamics of electron transfer in ex-
box4+⊂perylene, J. Phys. Chem. A 117 (2013) 12438–12448.
infrared spectroelectrochemistry, Organometallics 33 (2014) 4550–4559.
[40] D.C. Grills, J.A. Farrington, B.H. Layne, S.V. Lymar, B.A. Mello, J.M. Preses,
2
J.F. Wishart, Mechanism of the formation of a Mn-based CO reduction catalyst
[
31] N.E. Horwitz, B.T. Phelan, J.N. Nelson, M.D. Krzyaniak, M.R. Wasielewski,
Picosecond control of photogenerated radical pair lifetimes using a stable third
revealed by pulse radiolysis with time-resolved infrared detection, J. Am. Chem.
Soc. 136 (2014) 5563–5566.
28