Photochemistry and photophysics of a Pd(II) metalloporphyrin: Re(I) tricarbonyl bipyridine molecular dyad and its activity toward the photoreduction of CO2 to CO

Article abstract of DOI:10.1021/ic200243y

The photochemistry and photophysics of the cationic molecular dyad, 5-{4-[rhenium(I)tricarbonylpicoline-4-methyl-2,2′-bipyridine-4′- carboxyamidyl]phenyl}-10,15,20-triphenylporphyrinatopalladium(II) ([Re(CO) ₃(Pic)Bpy-PdTPP][PF₆]) ha

Full text of DOI:10.1021/ic200243y

Article
pubs.acs.org/IC
Photochemistry and Photophysics of a Pd(II) Metalloporphyrin: Re(I)
Tricarbonyl Bipyridine Molecular Dyad and its Activity Toward the
Photoreduction of CO2 to CO
Jacob Schneider,† Khuong Q. Vuong,‡ James A. Calladine,‡ Xue-Zhong Sun,‡ Adrian C. Whitwood,†
Michael W. George,*,‡ and Robin N. Perutz*,†
†Department of Chemistry, University of York, York, YO10 5DD, United Kingdom
‡School of Chemistry, University of Nottingham, University Park, Nottingham, NG7 2RD, United Kingdom
S
* Supporting Information
ABSTRACT: The photochemistry and photophysics of the
cationic molecular dyad, 5-{4-[rhenium(I)tricarbonylpicoline-
4-methyl-2,2′-bipyridine-4′-carboxyamidyl]phenyl}-10,15,20-
triphenylporphyrinatopalladium(II) ([Re(CO)3(Pic)Bpy-
PdTPP][PF6]) have been investigated. The single crystal X-
ray structure for the thiocyanate analogue, [Re(CO)3(NCS)-
Bpy-PdTPP], exhibits torsion angles of 69.1(9)°, 178.1(7)°,
and 156.8(9)° between porphyrin plane, porphyrin-linked
C6H4 group, amide moiety, and Bpy, respectively. Steady-state
photoexcitation (λex = 520 nm) of [Re(CO)3(Pic)Bpy-
PdTPP][PF6] in dimethylformamide (DMF) results in
substitution of Pic by bromide at the Re(I)Bpy core. When
[Re(CO)3(Pic)Bpy-PdTPP][PF6] is employed as a photocatalyst for the reduction of CO2 to CO in DMF/NEt3 solution with
λex > 420 nm, 2 turnovers (TNs) CO are formed after 4 h. If instead, a two-component mixture of PdTPP sensitizer and
mononuclear [Re(CO)3(Pic)Bpy][PF6] catalyst is used, 3 TNs CO are formed. In each experiment however, CO only forms
after a slight induction period and during the concurrent photoreduction of the sensitizer to a Pd(II) chlorin species.
Palladium(II) meso-tetraphenylchlorin, the hydrogenated porphyrin analogue of PdTPP, has been synthesized independently and
can be substituted for PdTPP in the two-component system with [Re(CO)3(Pic)Bpy][PF6], forming 9 TNs CO. An
intramolecular electron transfer process for the dyad is supported by cyclic voltammetry and steady-state emission studies, from
which the free energy change was calculated to be ΔGox* = −0.08 eV. Electron transfer from Pd(II) porphyrin to Re(I)
tricarbonyl bipyridine in [Re(CO)3(Pic)Bpy-PdTPP][PF6] was monitored using time-resolved infrared (TRIR) spectroscopy in
the ν(CO) region on several time scales with excitation at 532 nm. Spectra were recorded in CH2Cl2 with and without NEt3.
Picosecond TRIR spectroscopy shows rapid growth of bands assigned to the π−π* excited state (2029 cm−1) and to the charge-
separated state (2008, 1908 cm−1); these bands decay and the parent recovers with lifetimes of 20−50 ps. Spectra recorded on
longer time scales (ns, μs, and seconds) show the growth and decay of further species with ν(CO) bands indicative of electron
transfer to Re(Bpy).
reaction.9 Another strategy for CO2 reduction is to employ
electrocatalytic systems that obtain their electrons from solar
INTRODUCTION
■
The conversion of light-to-chemical energy is an attractive
cells and are capable of multielectron multiproton processes
approach to a sustainable future, and solar fuels provide a
adept at yielding fuels beyond CO such as MeOH and
means to exploit that energy through concepts of fundamental
methane.10−12
chemistry.1 In photosynthesis, plants absorb light energy from
In the early 1980s, Lehn and Ziessel reported the
the sun to convert CO2 and water into organic molecules and
photochemical reduction of CO2 to CO using a simple
oxygen, storing energy in the chemical bonds of the newly
formed molecules.2 The artificial photocatalytic conversion of
molecular photocatalyst, Re(CO)3(Bpy)X (Bpy = 2, 2′-
bipyridine, X = Cl, Br) with blue or near UV light (λ > 400
CO2 into liquid fuels with homo- and heterogeneous systems
has been a field of interest for many years,3−5 and though it
nm) in DMF/TEOA (DMF = dimethylformamide; TEOA =
triethanolamine) solution, with TEOA acting as a sacrificial
electron donor.13,14 They found the homogeneous rhenium(I)
remains a viable option, great advances are needed to develop a
sustainable infrastructure for solar fuels. One possible strategy
to utilize CO2 as a raw material for hydrocarbon products6,7 is
to reduce CO2 to CO, a chemical feedstock for such industrial
processes as hydroformylation8 and the Fischer−Tropsch
Received: February 21, 2011
Published: November 1, 2011
© 2011 American Chemical Society
11877
dx.doi.org/10.1021/ic200243y|Inorg. Chem. 2011, 50, 11877−11889

Inorganic Chemistry
Article
catalyst to be highly selective for CO production compared to
quantum yield of 0.0064 for CO formation.38 In a separate
study that involved a Zn(II) pyridylporphyrin directly
coordinated to a Re(I)Bpy fragment via a pyridine,39 Alessio,
Scandola, and co-workers found the photoinduced charge
transfer process from the S1 state of the Zn(II) porphyrin to
Re(I)Bpy to be only slightly exergonic; the complexes were not
tried for photochemical CO2 reduction. Recently, a Zn(II)
pyridylporphyrin was linked to a cobaltbis(dimethylglyoximate)
moiety and used for the photogeneration of hydrogen in the
presence of triethylamine.40 These reports draw obvious
parallels to the current study, but the structural, photochemical,
and photophysical properties of the Pd(II) porphyrin−
Re(I)Bpy dyad described here are distinct.
In this study the tetraphenylporphyrin sensitizer is
coordinated to Pd(II). We investigate the photochemical and
photophysical properties of a previously reported single
component Pd(II)−Re(I) dinuclear complex, [Re(CO)3(Pic)-
Bpy-PdTPP][PF6] (Pic = 3-picoline; Bpy = 2, 2′-bipyridine;
TPP = meso-tetraphenylporphyrin).37 The complex showed
striking activity toward photosubstitution at the Re(I)Bpy
center, but the photophysics were not studied in detail. In
addition, Pd(II) metalloporphyrins are less familiar in the
literature than other metalloporphyrin complexes, especially
with regard to supramolecular chemistry. We have examined
the time-resolved features of [Re(CO)3(Pic)Bpy-PdTPP]-
[PF6] such as photoinduced charge separation and radical
formation that occur from picoseconds up to seconds, as well as
its steady-state reactions including remote-site photosubstitu-
tion of 3-picoline by bromide and CO2 reduction to CO. We
find that saturation of a CC bond of the porphyrin to
produce a chlorin sensitizer occurs during photocatalysis. We
also report systems for the photocatalytic reduction of CO2 in
DMF/NEt3 solution that involve a two-component mixture of
Pd(II) porphyrin or chlorin sensitizer and [Re(CO)3(Pic)-
Bpy][PF6] as a mononuclear catalyst. They yielded slightly
higher but still modest amounts of CO.
15
2+
other photocatalytic systems such as Ru(Bpy)3
,
mixtures of
13,16
Ru(Bpy)3 and Co2+ (Co2+ = CoCl2 or Co(Bpy)32+),
or
2+
mixtures of Re(CO)3(Bpy)X and Co2+,13 for which the CO2
photoreduction products may include formate or H2. The
detailed mechanism of CO2 reduction using the Re(I)
photocatalysts has remained elusive since the seminal work
by Lehn and Ziessel and continues to be an area of
extraordinary interest.3,17−19 Their system was recently
optimized by Ishitani and co-workers in a mechanistic study
that led to the use of two different molecular entities in
combination, Re(CO)3(Bpy)(MeCN)+ serving as a catalyst and
Re(CO)3((4,4′-MeO)2Bpy)(P(OEt)3)+ as the photosensitizer,
with a quantum yield of 0.59 for CO formation.18 One of the
drawbacks to using Re(I) tricarbonyl diimine complexes as
photocatalysts is the high energy light required (λ ∼ 350−450
nm) to induce their metal-to-ligand charge transfer (MLCT)
excited-states.20 Over the past few years Ishitani, Bian, and co-
workers attempted to remedy this by sensitizing a molecular
Re(I) catalyst with one or more visible-light absorbing
Ru(Bpy)3 pendants.21−23 A number of different di- and
2+
multinuclear Ru(II)−Re(I) complexes have been synthesized
that function as single component photocatalysts in the
reduction of CO2 to CO using lower energy visible light (λ >
500 nm).24,25 The most effective of these are the molecular
2+
dyads that have an insulated connection between Ru(Bpy)3
and Re(I),25,26 with P(OEt)3 as the peripheral ligand for
Re(CO)3(Bpy), giving a quantum yield for CO formation of
0.21.27
While frequently studied mononuclear Re(I) photocatalysts
are of the general formula [Re(CO)3(Bpy′)(X/L)]0/+ (X =
halide, NCS, CN; L = phosphine, phosphite; Bpy′ = 2, 2′-
bipyridine and derivatives),18,28−30 or even [Re-
(CO)2(Bpy′)(L)2]+,31,32 there is one study that uses mono-
nuclear [Re(CO)3(Bpy)(4-Xpy)]+ (Xpy = pyridine derivatives)
complexes for the photocatalytic reduction of CO2.33 The
authors showed that the 4-Xpy family of complexes are not as
efficient photocatalysts as their halide and phosphite analogues,
having lower quantum yields (ϕ = 0.03−0.13). In addition,
Ishitani and colleagues suggested that when pyridine is the
peripheral ligand in single component Ru(II)−Re(I) dinuclear
complexes, the decomposition of the active photocatalyst is
accelerated, thus lowering its catalytic activity.27 Apart from
these reports, the reactivity of mono- and dinuclear complexes
with a pyridine bound to Re(I) toward CO2 reduction remains
relatively unexplored. To investigate this further we have built
upon our earlier work34−36 and synthesized a molecular dyad
made up of two components: (1) a low-energy-light absorbing
Pd(II) metalloporphyrin core that is covalently attached to (2)
a photocatalytic Re(I) tricarbonyl bipyridine (Re(I)Bpy) center
with 3-picoline as the ancillary ligand.37 The two components
are linked through an amide bond that aims to optimize
electron-transfer efficiency. It has been determined in
analogous Zn(II) and Mg(II) dyads that there is no
communication between metalloporphyrin and Re(I)Bpy in
the ground state.36 However, upon selective light excitation into
the metalloporphyrin S1 state, photoinduced electron transfer
occurs,36 as well as subsequent substitution of 3-picoline by
other nucleophiles or solvent molecules at the Re(I)Bpy
center.35−37 In addition, a Zn(II) porphyrin-Re(I)Bpy dyad was
recently reported by Inoue and co-workers to undergo electron
transfer from the porphyrin S2 state to the Re(I)Bpy center, a
process that also led to the photoreduction of CO2, with a
EXPERIMENTAL SECTION
■
Characterization and Instrumentation. All NMR spectra were
recorded on a Bruker 500 MHz spectrometer. 1H NMR chemical
shifts (in ppm) are referenced using the residual solvent resonances of
CD2Cl2, CDCl3, or THF-d8; 31P{1H} NMR chemical shifts (in ppm)
are relative to an external H3PO4 standard. NMR spectroscopic
assignments were made with the aid of 2D 1H−1H COSY analysis. We
relied on high resolution mass spectrometry rather than elemental
analysis because of incomplete combustion when attempting micro-
analysis. Solution infrared spectra were measured using a Unicam
Research Series FTIR spectrometer in either a demountable Omni cell
system (Specac) or a Harrick solution cell with 2 mm CaF2 windows.
Electrospray Ionization (ESI) mass spectra (positive) were recorded
on a Bruker microOTOF mass spectrometer coupled to an Agilent
1200 Series LC system. Mass spectra are quoted for 106Pd187Re. The
UV−visible absorption spectra were measured using an Agilent 8453
spectrometer. The emission spectra were measured using a Hitachi F-
4500 Fluorimeter, and all samples (1.0 × 10−5 M) were either
degassed by three freeze−pump−thaw cycles or deaerated by purging
the samples with Ar.
Cyclic voltammograms of [Re(CO)3(Pic)Bpy-PdTPP][PF6],
PdTPP, and PdTPC in CH2Cl2 were measured using a VoltaLab
PST050 potentiostat (Radiometer Analytical) with Pt wire working, Pt
wire auxiliary, and Ag wire pseudoreference electrodes. The cyclic
voltammogram of [Re(CO)3(Pic)Bpy][PF6] in CH2Cl2 was measured
using a BAS 100B Electrochemical Analyzer with glassy carbon
working, Pt wire auxiliary, and SCE reference electrodes. The sample
solutions contained 0.1 M [Bu4N][PF6] (electrochemical grade,
11878
dx.doi.org/10.1021/ic200243y|Inorg. Chem. 2011, 50, 11877−11889

Inorganic Chemistry
Article
Aldrich) and were continuously stirred and purged with a steady flow
of N2 or Ar in between measurements. Ferrocene (sublimed, Aldrich)
was used as an internal standard during each experiment, and the
electrochemical potentials are reported versus the internal ferroce-
nium/ferrocene (Fc+/0) redox couple.
0.68890 Å, selected by a Si(111) double-crystal monochromator at the
Zr edge. The beam was focused by vertical and horizontal focusing
mirrors. Rigaku CrystalClear software was used to record images.44
Frame integration and unit-cell refinement were carried out with
Bruker APEX2.45 Absorption corrections were applied by SADABS
(v2.03, Sheldrick). The structures were solved by direct methods using
SHELXS-97 and refined by full-matrix least-squares using SHELXL-
97.46,47 Hydrogen atoms bound to carbon and nitrogen were placed at
calculated positions and included in the refinement using a riding
model. The benzene of crystallization was disordered and modeled
over two sites with a refined relative occupancy of 80:20. The carbons
of the benzene were modeled isotropically. Because of their close
proximity, the pairs C(61a)/C(61b) and C(62a)/C(62)b were
constrained to have identical ADP. The benzene was constrained to
be hexagonal with a C−C distance of 1.39 Å. Crystallographic
parameters are listed in Table 1.
Steady-state photolysis experiments were performed using either a
(1) ILC 302UV xenon high intensity light source set to visible (ILC
Technology, Inc.) or a (2) Hitachi XGA ED-X3270A multimedia
mobile data projector (150 W UHB mercury lamp) or a (3) BenQ
MP77 data projector (Philips 280 W UHB mercury lamp). The light
was directed through a water filter (10 cm) and a cutoff or interference
filter before reaching the sample solution. Experiments with lamp (2)
are summarized in the Supporting Information, Table S1; lamp (3)
was only used for the quantum yield measurements outlined below.
The quantum yield (ϕ) for the photosubstitution of 3-picoline by
bromide in [Re(CO)3(Pic)Bpy-PdTPP][PF6] was calculated using a
method developed by Lees,41 and accounts for the inner filter
absorbance of the [Re(CO)3BrBpy-PdTPP] product that is formed
(see the Steady-State Photochemistry section of the manuscript for
more details). The incident light intensity (BenQ MP777; λ = 520 nm
narrow bandwidth filter) reaching the sample during photolysis was
7.5 × 10−9 einsteins s−1 (0.51 mW/cm2), determined using an
ILT1400 radiometer (International Light Technologies).
Table 1. Unit Cell, Data Collection, and Refinement
Parameters for [Re(CO)3(NCS)Bpy-PdTPP]·C6H6
formula
Fw
C60H37N8O4PdReSC6H6
1336.74
T (K)
120(2)
The specifications of pico-, nano-, and microsecond time-resolved
infrared spectroscopy experiments have been reported.42,43 A Nicolet
680 FTIR spectrometer capable of rapid-scanning was synchronized to
the laser for the longer time scale measurements. For all TRIR and
FTIR experiments the second harmonic output (532 nm) of a
Nd:YAG laser was used and the concentrations of [Re(CO)3(Pic)-
Bpy-PdTPP][PF6] and NEt3 (when present) were approximately 0.5
mM and 1 M in CH2Cl2, respectively.
Photochemical Reduction of CO2. All photochemical CO2
reduction experiments were carried out in a custom-designed
photolysis kit fitted with a gas infrared cell (100 mm long, 25 mm
diameter CaF2 windows) connected to a UV−visible cuvette (10 mm
path length) (See the Supporting Information, Figure S1). A 3 mL
solution of either 0.05 or 0.27 mM complex in DMF/NEt3 (5:1 v/v)
was transferred to the UV−visible cuvette and sealed with a red rubber
Suba-seal and wire. The sample solution was deaerated by purging
with CO2/CH4 (1% CH4, BOC gases) for 5 min. When two
components were involved, the sensitizer (PdTPP, PdTPC, or
[Ru(Bpy)3][PF6]2) and the catalyst ([Re(CO)3(Pic)Bpy][PF6])
were present in equimolar concentrations. The solution was then
irradiated, and the reactions were monitored by gas chromatography
(GC), UV−vis, and gas infrared spectroscopies.
For GC analysis, a Unicam ProGC+ (ThermoONIX) fitted with a
thermal conductivity detector was used. The gases, air, CO, CH4, and
CO2, were separated with a Restek ShinCarbon ST 100/120
micropacked column (2 m, 1/16″ OD, 1.0 mm ID) fitted with
“pigtails” of Restek intermediate-polarity deactivated guard column on
either end (fused silica, 0.53 mm ID, 0.69 ± 0.05 mm OD). Ultra high
purity He (N6.0, BOC gases) was used as the carrier gas and was
passed through a GC triple carrier gas filter (Focus Technical) to
remove trace impurities (oxygen, moisture, and hydrocarbons) prior to
reaching the detector. Gas samples were injected manually (100 μL)
using a Hamilton gastight locking syringe (500 μL). The amount of
CO produced was quantified using a calibration plot. Known volumes
of CO were mixed with a standard volume (large excess) of CO2/1%
CH4. Samples were removed and analyzed by GC. The area of the CO
peak was plotted versus the area of the CH4 peak in the GC
chromatogram. Moles of CO were calculated using the volume of an
ideal gas (i.e., 22.4 L mol−1 at 293 K), and turnover numbers (TNs) of
CO were calculated based on the moles of complex catalyst present in
the reaction solution and the volume of cell headspace.
λ (Å)
0.68890
crystal system
space group
a (Å)
triclinic
P1
̅
9.401(2)
12.936(3)
23.410(5)
91.932(2)
94.160(2)
105.874(3)
2726.9(10)
2
b (Å)
c (Å)
α (deg)
β (deg)
γ (deg)
V (Å3)
Z
ρcald (Mg/m3)
μ (mm−1
1.628
)
2.425
crystal color/shape
crystal size (mm3)
θ range (deg)
red/block
0.04 × 0.04 × 0.02
0.85−24.16
9300
no. of data
no. of parameters
701
GOF
1.047
R1, wR2 (F2, I > 2σ(I))
R1, wR2 (F2, all data)
CCDC no.
0.0470, 0.1051
0.0734, 0.1151
843937
Materials. NaNCS, AgOTf, [Bu4N][Br], [H4N][PF6], aluminum
oxide (activated acidic, Brockmann I), pyrrolidine (Aldrich), toluene,
tetrahydrofuran (THF, Fisher), Pd(OAc)2 (Alfa Aesar), meso-
tetraphenylporphyrin (Frontier Scientific), and 3-picoline (BDH)
were purchased commercially and used without further purification;
NEt3 (Aldrich) was dried over KOH or CaH2 at room temperature
and then distilled under argon; [Ru(Bpy)3]Cl2·6H2O (Aldrich) was
converted to the PF6 salt ([Ru(Bpy)3][PF6]2),48 CH2Cl2 and DMF
−
(Fisher) were dried on a solvent purification system and deaerated
with argon prior to use. DMF was further dried over molecular sieves
followed by distillation under argon. Re(CO)5Br,49 PdTPP,37 Bpy-
PdTPP,37 [Re(CO)3(Pic)(Bpy)][PF6],29,37 meso-tetraphenylchlorin
(TPC),50,51 and 1-benzyl-1,4-dihydronicotinamide (BNAH)52,53 were
prepared according to literature methods. All other solvents and
reagents were purchased commercially and used as received unless
otherwise noted.
Synthesis of [Re(CO)3BrBpy-PdTPP]. This complex was synthe-
sized as reported previously.37 Since the compound was not fully
characterized spectroscopically, we include a full set of data here.
1H NMR (500 MHz, THF-d8): δ 2.65 (3 H, s, Bpy CH3); 7.55 (1
H, d, J 5.5 Hz, Bpy H5′); 7.77 (9 H, m, 9 m-, p-phenyl); 8.14−8.20 (7
Crystallographic Structure Determination. The data for the
crystallographic structure determination of [Re(CO)3(NCS)Bpy-
PdTPP] were collected on beamline I19 of the Diamond Light
Source, using a Crystal Logic kappa geometry four-circle diffrac-
tometer equipped with a Rigaku Saturn 724+ CCD detector and a
Cryostream cooler (at 120 K). The synchrotron X-ray wavelength was
11879
dx.doi.org/10.1021/ic200243y|Inorg. Chem. 2011, 50, 11877−11889

Inorganic Chemistry
Article
Figure 1. Chemical structures of complexes studied.
H, m, 6 o-phenyl, Bpy H5); 8.20−8.25 (4 H, m, bridging C6H4); 8.67
(1 H, s, Bpy H3′); 8.80 (4 H, s, β-pyrrolic Ha and Hb); 8.82 (2 H, d, J
5.0 Hz, β-pyrrolic Hc); 8.87 (2 H, d, J 5.0 Hz, β-pyrrolic Hd); 8.96 (1
H, d, J 5.5 Hz, Bpy H6′); 9.04 (1 H, s, Bpy H3); 9.29 (1 H, d, J 5.5 Hz,
Bpy H6); 10.36 (1 H, s, CONH). See the Supporting Information,
Figure S2 for labeling of atoms.
Single crystals suitable for X-ray diffraction were grown by layering a
benzene solution of the complex with diethyl ether.
1H NMR (500 MHz, CD2Cl2): δ 2.59 (3 H, s, Bpy CH3); 7.52 (1
H, d, J 5.5 Hz, Bpy H5′); 7.75−7.80 (9 H, m, 9 m-, p-phenyl); 8.17−
8.18 (7 H, m, o-phenyl and Bpy H5); 8.24−8.30 (4 H, m, bridging
C6H4); 8.38 (1 H, s, Bpy H3′); 8.73 (1 H, s, Bpy H3); 8.84 (4 H, s, β-
pyrrolic Ha and Hb); 8.86 (2 H, d, J 4.9 Hz, β-pyrrolic Hc); 8.89 (1 H,
d, J 5.7 Hz, Bpy H6′); 8.92 (2 H, d, J 4.9 Hz, β-pyrrolic Hd); 8.94 (1 H,
s, CONH); 9.25 (1 H, d, J 5.5 Hz, Bpy H6).
IR(THF): ν(CO)/cm−1 2020 (s), 1921 (s), 1897 (s).
ESI-MS (m/z): calcd for [C59H37BrN7O4106Pd187Re] 1279.07;
observed mass [M − Br + DMSO]+ 1278.2 and [M − Br +
MeCN]+ 1241.2.
IR(THF): ν(CN)/cm−1 2090 (br s); ν(CO)/cm−1 2023 (s), 1922
and 1912 (s, overlapping peaks).
[Re(CO)3(Pic)Bpy-PdTPP][PF6]. [Re(CO)3(Pic)Bpy-PdTPP]+
was synthesized as the triflate salt,37 and was converted to the PF6− salt
by anion metathesis with [H4N][PF6] in MeOH, and was isolated by
precipitation with H2O. The product was purified by column
chromatography using silica gel 60 with CH2Cl2/THF (<10% THF)
as the eluting solvent mixture.
ESI-MS (m/z): calcd for C60H38N8O432S106Pd187Re 1259.1329;
observed mass [M + H]+ 1259.1346; difference 1.7 mDa.
Palladium(II) meso-Tetraphenylchlorin (PdTPC). In a flask
charged with TPC (>97% pure, 70 mg, 0.11 mmol) and Pd(OAc)2 (54
mg, 0.24 mmol, 2.2 equiv) was added toluene (30 mL) and the
suspension was refluxed (130 °C) under Ar for 18 h. The solvent was
removed under vacuum, the dark purple residue redissolved in a
minimum amount of CH2Cl2, and the solution passed through a plug
of acidic alumina (to remove any unreacted free base TPC), rinsing
with CH2Cl2 until the eluent was nearly colorless. The solution was
concentrated, and addition of MeOH caused a dark purple precipitate
to form, which was collected on a sintered glass funnel (note that
filtrate still dark blue-green in color). The purple precipitate was very
fine and difficult to handle; 50 mg recovered (63% yield), isolated in
1H NMR (500 MHz, CD2Cl2): δ 2.28 (3 H, s, Bpy CH3); 2.70 (3
H, s, 3-picoline CH3); 7.28 (1 H, dd, J 7.7 and 5.9 Hz, 3-picoline H5);
7.60 (1 H, d, J 5.7 Hz, Bpy H5′); 7.68 (1 H, d, J 7.8 Hz, 3-picoline H4);
7.75−7.81 (9 H, m, 9 m-, p-phenyl); 7.88 (1 H, d, J 5.6 Hz, 3-picoline
H6); 8.08 (1 H, s, 3-picoline H2); 8.18−8.19 (6 H, m, o-phenyl);
8.21−8.25 (4 H, m, bridging C6H4); 8.35 (1 H, d, J 5.6 Hz, Bpy H5);
8.50 (1 H, s, Bpy H3′); 8.84 (4 H, s, β-pyrrolic Ha and Hb); 8.86 (2 H,
d, J 4.9 Hz, β-pyrrolic Hc); 8.88 (1 H, s, Bpy H3); 8.91 (2 H, d, J 4.9
Hz, β-pyrrolic Hd); 8.99 (1 H, d, J 5.7 Hz, Bpy H6′); 9.28 (1 H, s,
CONH); 9.33 (1 H, d, J 5.6 Hz, Bpy H6). 31P{1H} NMR (202 MHz,
1
92% purity, the remaining impurity was PdTPP as determined by H
NMR (Supporting Information,Figure S3).
−
1H NMR (500 MHz, CDCl3): δ 4.21 (4 H, s, pyrrolic C2H4); 7.60−
7.69 (12 H, m, 8 m-, 4 p-phenyl); 7.80 (4 H, m, o-phenyl); 8.00 (6 H,
m, 4 o-phenyl, 2 β-pyrrolic); 8.33 (4 H, m, β-pyrrolic).
PdTPP (for Comparison). 1H NMR (500 MHz, CDCl3): δ 7.72−
7.79 (12 H, m, 8 m-, 4 p-phenyl); 8.17 (8 H, o-phenyl); 8.81 (8 H, s,
pyrrolic).
CD2Cl2): δ −144.24 (septet, JP−F 712.8 Hz, PF6 ).
IR(THF): ν(CO)/cm−1 2032 (s), 1928 (br s).
ESI-MS (m/z): calcd for [C65H44N8O4106Pd187Re]+ 1293.2085;
observed mass [M]+ 1293.2050, difference 3.5 mDa.
[Re(CO)3(NCS)Bpy-PdTPP]. A solution of [Re(CO)3BrBpy-
PdTPP] (0.084 mmol, 108 mg) was refluxed in the presence of
excess NaNCS (9.37 mmol, 0.76 g) in THF (∼25 mL) for 21 h under
Ar. The cooled solution was diluted with saturated aqueous NaNCS
(100 mL) and extracted with CH2Cl2 until no color remained in the
aqueous layer. The combined CH2Cl2 portions were washed with H2O
(2 × 50 mL) and dried over MgSO4. The CH2Cl2 solution was
filtered, concentrated, and slow addition of hexane caused a precipitate
to form that was collected by filtration. The crude product was purified
by column chromatography using silica gel 60 with CH2Cl2 as the
eluting solvent; 56 mg orange-red product recovered (53% yield).
RESULTS
■
Syntheses. The structures of the complexes studied here
are shown in Figure 1. The cationic [Re(CO)3(Pic)Bpy-
PdTPP][PF6] was synthesized from its [Re(CO)3BrBpy-
PdTPP] precursor,37 followed by anion metathesis to the
−
PF6 salt. In the earlier report [Re(CO)3BrBpy-PdTPP] had
not been characterized, but we have now done so using NMR,
11880
dx.doi.org/10.1021/ic200243y|Inorg. Chem. 2011, 50, 11877−11889

Inorganic Chemistry
Article
Table 2. Selected Bond Lengths and Angles for [Re(CO)3(NCS)Bpy-PdTPP]·C6H6
Bond Lengths (Å)
Re(1)−N(6)
Re(1)−N(8)
Re(1)−C(59)
C(58)−O(2)
C(60)−O(4)
C(57)−S(1)
Pd(1)−N(2)
Pd(1)−N(4)
2.131(9)
2.162(6)
1.912(10)
1.158(9)
1.168(10)
1.639(10)
2.019(5)
2.011(5)
Re(1)−N(7)
Re(1)−C(58)
Re(1)−C(60)
C(59)−O(3)
N(6)−C(57)
Pd(1)−N(1)
Pd(1)−N(3)
2.163(5)
1.912(8)
1.904(10)
1.166(10)
1.111(10)
2.026(5)
2.020(5)
Bond Angles (deg)
N(7)−Re(1)−N(8)
N(6)−Re(1)−C(60)
N(6)−Re(1)−N(8)
N(6)−C(57)−S(1)
N(1)−Pd(1)−N(3)
N(1)−Pd(1)−N(2)
75.8(2)
176.4(3)
85.0(3)
176.2(8)
178.6(2)
89.4(2)
C(58)−Re(1)−C(59)
C(58)−Re(1)−N(7)
N(6)−Re(1)−C(59)
C(57)−N(6)−Re(1)
N(2)−Pd(1)−N(4)
N(2)−Pd(1)−N(3)
87.5(3)
172.9(3)
92.7(3)
175.9(7)
179.2(2)
90.6(2)
Torsion Angles (deg)
C(40)−C(39)−C(6)−C(5)
C(22)−C(23)−C(24)−N(5)
C(32)−C(31)−C(33)−N(8)
119.4(7)
178.1(7)
179.2(7)
C(2)−C(1)−C(21)−C(22)
O(1)−C(27)−C(28)−C(32)
69.1(9)
156.8(9)
UV−vis, and IR spectroscopies, and mass spectrometry. It is
worth noting that in the mass spectrum only solvent complexes
of [Re(CO)3BrBpy-PdTPP] could be observed, namely,
[Re(CO)3(DMSO)Bpy-PdTPP]+ and [Re(CO)3(MeCN)-
Bpy-PdTPP]+.
The free base TPC was prepared50 and isolated51 by
modified literature methods. Palladium(II) was then inserted
by refluxing TPC with excess Pd(OAc)2 to give PdTPC.
Because of trace amounts of meso-tetraphenylporphyrin (TPP)
impurity in the isolated TPC (<3%), we were only able to
1
obtain purity up to 92% (based on H NMR) of PdTPC, the
only impurity being PdTPP (See the Supporting Information,
Figure S3). Insertion of Pd(II) into TPC required longer
periods of refluxing relative to the corresponding insertion into
TPP. To the best of our knowledge the only studies of PdTPC
described either (i) methods used to determine complex
purity,54 (ii) association properties in aqueous solution,55 or
(iii) triplet emission quenching under various conditions.56−59
The latter reports by Sapunov and co-workers dealt with effects
on triplet emission quenching that included concentration
dependence in aqueous56 and organic media,57 quenching by
heme proteins and cyanocobalamin,58 and the triplet state
emission quenching of PdTPP by PdTPC in toluene.59
Crystallography. After several unsuccessful attempts at
growing single crystals of [Re(CO)3(Pic)Bpy-PdTPP][PF6]
and [Re(CO)3BrBpy-PdTPP], we instead synthesized [Re-
(CO)3(NCS)Bpy-PdTPP] by reaction of [Re(CO)3BrBpy-
PdTPP] with NaNCS. Its X-ray structure was determined from
data collected at the Diamond synchrotron because of the small
size of the crystals (Table 2, Figure 2). The Re(I) center has a
geometry representative of the analogous mononuclear [Re-
Figure 2. ORTEP drawing of [Re(CO)3(NCS)Bpy-PdTPP]·C6H6;
hydrogen atoms and C6H6 solvent molecule omitted for clarity.
Anisotropic displacement parameters shown at 50% probability.
The chromophoric porphyrin constituent of [Re-
(CO)3(NCS)Bpy-PdTPP] adopts a square planar geometry
at the Pd(II) nucleus having N−Pd−N bond angles of nearly
90 and 180°, and is comparable to the structure of PdTPP,
reported in 1964 by Fleischer et al.61 The overall porphyrin
moiety is nonplanar in both [Re(CO)3(NCS)Bpy-PdTPP]
and mononuclear PdTPP. The nonplanar porphyrin in
[Re(CO)3(NCS)Bpy-PdTPP] forms a saddle with pairs of
opposite β-carbons. For example, C(3) and C(4) paired with
C(13) and C(14) make up the high point of the saddle, while
C(8) and C(9) paired with C(18) and C(19) make up the low
point of the saddle. In contrast, the saddle structure in
mononuclear PdTPP has the meso-carbons of the porphyrin
backbone as the highest and lowest points in the saddle.61
Also of interest in the structure of [Re(CO)3(NCS)Bpy-
PdTPP] is the planarity of the phenylamide link between
porphyrin and Re(I)Bpy. Should the two components of the
dyad arrange in a planar conformation, this will optimize
electron transfer efficiency from chromophore to catalyst. As it
turns out, the phenylamide linker is nearly planar in itself with
the C(22)−C(23)−C(24)−N(5) torsion angle being
178.1(7)°, but is out of the plane of the porphyrin (C(2)−
C(1)−C(21)−C(22)) by 69.1(9)° and is out of the plane of
̌
(CO)3(NCS)(Bpy)] complex reported by Vlcek, and co-
workers.60 The three CO ligands are in a facial arrangement
around the Re(I) center with near 90° angles between ligands,
while the N(7)−Re(1)−N(8) Bpy bite angle is comparatively
smaller at 75.8(2)°. As reported for the mononuclear
[Re(CO)3(NCS)(Bpy)],60 the Re−N(CS) bond in [Re-
(CO)3(NCS)Bpy-PdTPP] is angled slightly, only about 4°,
toward the chelating Bpy ligand.
11881
dx.doi.org/10.1021/ic200243y|Inorg. Chem. 2011, 50, 11877−11889

Inorganic Chemistry
Article
the Re(I)Bpy with the O(1)−C(27)−C(28)−C(32) torsion
angle at 156.8(9)°.
energy required to sensitize the photocatalytic reduction of
CO2 to CO with a molecular Re(I) tricarbonyl diimine catalyst.
The dyad was tested for photocatalytic CO2 reduction using
literature conditions (see Introduction) that entail steady-state
photolysis of the complex in DMF/NEt3 (5:1 v/v) solution
saturated with CO2. We employ NEt3 as a sacrificial electron
donor to maintain consistency with other steady-state and time-
resolved studies. The photoreduction reaction was monitored
by gas chromatography, UV−vis spectroscopy, and in some
instances solution IR spectroscopy. The photolysis cell used in
each reaction was connected to a infrared gas cell (Supporting
Information, Figure S1), and the gas-phase IR spectra were
measured before and after photolysis to further monitor the
formation of CO. This was especially useful during photo-
reactions where only small amounts of CO could be observed,
as gas-phase IR spectroscopy was a more sensitive technique
than gas chromatography under the conditions in this study.
In the case of [Re(CO)3(Pic)Bpy-PdTPP][PF6], only 2
turnovers of CO could be observed with λ > 420 nm
(porphyrin S2 excitation), and only trace amounts using λ >
495 nm (porphyrin S1) excitation. In all experiments with a
steady-state method, UV−vis spectroscopy detects the
formation of a new porphyrin photoreduction product or
products, one of which has an absorption maximum of 603 nm
and can be identified as a chlorin species (Supporting
Information, Figure S5). This observation is consistent with a
report by Whitten and co-workers that described the formation
of chlorin and iso-bacteriochlorin reduction products of PdTPP
when photolyzed in the presence of dimethylaniline (DMA).63
In that study, one C−C bond of the pyrrole ring (or rings for
iso-bacteriochlorin) was thought to be hydrogenated in one
position and aminated at the other carbon through one of the
methyl groups of the DMA.
With narrow bandwidth excitation (λex = 520 nm) of
[Re(CO)3(Pic)Bpy-PdTPP][PF6], only minimal amounts of
chlorin are formed within 4 h (absorbance ∼0.1, Supporting
Information, Figure S5), but more grows in with longer
photolysis times (absorbance >1). No CO is observed after 4 h
of photolysis with λex = 520 nm and only trace amounts are
observed after 8 h, with no increase up to 21 h photolysis.
With a 420 nm cutoff filter, 2 TNs CO can be obtained after
4 h, but the onset of CO formation occurs only at 2 h, and the
reaction is complicated by the competing photoreduction
products of the porphyrin core. To probe the Re(I)Bpy part of
the dyad, the solution IR spectra were measured after
photolysis showing v(CO) at 2017, 1908, and about 1891
cm−1, suggesting the formation of a formate complex,
[Re(CO)3(OCHO)Bpy-PdTPP]. The proposal of a formate
complex is based on a study by Fujita and co-workers, where
the reaction of Re(CO)3(dmb), a reduced form of [Re-
(CO)3(OTf)(dmb)], with CO2 yielded [Re(CO)3(OCHO)-
(dmb)] (dmb = 4,4′-dimethyl-2,2′-bipyridine) and was reported
to have v(CO) at 2017, 1911, and 1888 cm−1 in DMF
solution.19 The IR spectrum demonstrates that while the Pd(II)
porphyrin core is being reduced, the Re(I)Bpy site undergoes
photosubstitution of 3-picoline by other species present in
solution (Table 3).
Steady-State Photochemistry. Selective excitation of the
[Re(CO)3(Pic)Bpy-PdTPP][PF6] porphyrin core (λ > 495
nm) induces remote-site substitution of 3-picoline at the
Re(I)Bpy center.37 This extraordinary photoreaction can be
observed using steady-state IR spectroscopy under a variety of
conditions in the presence of the sacrificial electron donor,
NEt3. For example, in THF, DMF, or butyronitrile solution
with an added nucleophilic bromide source, 3-picoline is
substituted by bromide at the Re(I)Bpy center forming neutral
[Re(CO)3BrBpy-PdTPP]. In the absence of bromide, 3-
picoline is instead substituted by a solvent molecule such as
DMF, NEt3, THF, and so forth. Alternatively, in the presence
of 1% excess 3-picoline in THF, the rapid formation of a radical
species, believed to be [Re+(CO)3(Pic)Bpy−-PdTPP], can be
detected in the solution IR spectrum with v(CO) at 2011,
1907, and 1893 cm−1.37
Loss of the labile ligand bound to Re(I)Bpy, 3-picoline in the
current study, is one of the steps that is known to take place
during the catalytic cycle of the photoreduction of CO2 by
mononuclear Re(I) tricarbonyl diimine complexes.3,17,18 It is
therefore important to know the efficiency with which this
reaction proceeds, particularly in DMF solution, as it is the
solvent of choice for photochemical CO2 reduction systems
that employ such catalysts.25 A solution of [Re(CO)3(Pic)Bpy-
PdTPP][PF6] in DMF/NEt3 (5:1 v/v) with 0.12 M [Bu4N]-
[Br] was irradiated with a narrow bandwidth filter (λ ex = 520
nm), forming [Re(CO)3BrBpy-PdTPP] quantitatively in less
than 6 min.62 During the photoreaction, the UV−vis
absorbance changes by approximately 5% from time zero,
until the reaction is finished. In contrast to the small changes in
the UV−vis absorbance, the changes in the infrared spectrum
are substantial and there is almost no peak overlap between
[Re(CO)3(Pic)Bpy-PdTPP][PF6] and [Re(CO)3BrBpy-
PdTPP]. Because of this, the quantum yield was obtained
from changes in the infrared spectra over time (see Figure 3).62
Figure 3. Difference IR spectra for the steady-state photosubstitution
reaction of [Re(CO)3(Pic)Bpy-PdTPP][PF6] measured at intervals
from 30 s up to 330 s, showing the replacement of 3-picoline by
bromide in DMF/NEt3 (5:1 v/v) with 0.12 M [Bu4N][Br] and λex
=
520 nm.
Two-Component Photoreduction System. In light of
the observation of chlorin formation using [Re(CO)3(Pic)Bpy-
PdTPP][PF6], we hypothesized that the chlorin may be a
better photosensitizer than the porphyrin. To test this idea we
synthesized PdTPC, the hydrogenated porphyrin derivative of
PdTPP, and employed mixtures of two components, sensitizer
The kinetic models of the infrared absorbances at 2031 and
2019 cm−1 are shown in the Supporting Information, Figure S4.
Photoreduction of CO2. [Re(CO)3(Pic)Bpy-PdTPP][PF6]
as a Molecular Photocatalyst. The molecular dyad [Re-
(CO)3(Pic)Bpy-PdTPP][PF6] was designed to lower the
11882
dx.doi.org/10.1021/ic200243y|Inorg. Chem. 2011, 50, 11877−11889

Inorganic Chemistry
Article
Table 3. UV-vis and Emission Data for Selected Complexes
d
complex
λabs/nm (ε/103 dm3 mol−1 cm−1
)
λem/nm
a
[Re(CO)3BrBpy-PdTPP]
417 (249), 484 (3.8), 524 (23.6), 554 (2.3)
416 (275), 484 (3.7), 524 (31.2), 554 (2.7)
565, 610, 705, 775
565, 610, 705, 775
563, 612, 706, 775
614, 659, 794
a
[Re(CO)3(Pic)Bpy-PdTPP][PF6]
b
PdTPP
416 (314), 463 (2.7), 485 (3.3), 523 (31.3), 554 (2.8)
c
PdTPC
410 (192), 496 (6.6), 529 (4.9), 563 (10.0), 603 (55.0)
a
b
c
d
Measured in THF, λex = 524 nm. Measured in CH2Cl2, λex = 524 nm. Measured in CH2Cl2, λex = 603 nm. Emission maximum marked in italics.
and [Re(CO)3(Pic)Bpy][PF6], for CO2 reduction. When
comparing the TNs of all the systems, the system with
PdTPC as the sensitizer was generally more effective in the
photoreduction of CO2 to CO (Figure 4). However, only when
maximum of PdTPC is red-shifted significantly to 603 nm and
has a higher extinction coefficient, ε = 55 000 M−1 cm−1, as well
as another notable Q-band at 563 nm with ε = 10 000
M−1 cm−1 (Table 3). Not only are the Q-peaks shifted
substantially, but PdTPC possesses lower symmetry than
PdTPP, C2v versus D4h, respectively.
Emission Spectroscopy. The steady-state emission spec-
tra of Bpy-PdTPP, [Re(CO)3BrBpy-PdTPP], and [Re-
(CO)3(Pic)Bpy-PdTPP][PF6] in THF solution (λex = 524
nm) are shown in Supporting Information, Figure S7. The
complexes exhibit weak singlet emission at 565 and 610 nm and
strong triplet emission at 705 and 775 nm, assigned to 1Q0,0 (or
E00) and 1Q1,0 and 3Q0,0 and 3Q1,0 transitions, respectively, with
E00 determined to be 2.19 eV for all three complexes. It is
noteworthy that at equimolar concentrations the emission
intensities of Bpy-PdTPP and [Re(CO)3BrBpy-PdTPP]
remain unchanged while that of the cationic [Re(CO)3(Pic)-
Bpy-PdTPP][PF6] is quenched by 37%. This is in contrast to
our earlier report in butyronitrile solution where we observed
no change in the emission intensity of [Re(CO)3(Pic)Bpy-
3
PdTPP][OTf] relative to Bpy-PdTPP, nor the Q1,0
transition.37
Figure 4. Turnovers of CO obtained from steady-state photolysis in
DMF/NEt3 (5:1 v/v) solution saturated with CO2; complexes 0.05
mM except [Re(CO)3(Pic)Bpy-PdTPP][PF6] (0.27 mM); λ > 420
nm except [Ru(Bpy)3][PF6]2 (λ > 495 nm); PdTPP, PdTPC, and
[Ru(Bpy)3][PF6]2 are 1:1 two-component mixtures with [Re-
(CO)3(Pic)Bpy][PF6]. PdRePic = [Re(CO)3(Pic)Bpy-PdTPP]-
[PF6], RuBpy = [Ru(Bpy)3][PF6]2.
The steady-state emission spectra of PdTPP and PdTPC in
CH2Cl2 solution are shown in Supporting Information, Figure
S9. The emission profile of PdTPP (λ ex = 524 nm, CH2Cl2) is
very similar to that of the Pd(II) porphyrins in THF solution,
having weak singlet emission at 563 and 612 nm and strong
triplet emission at 706 and 775 nm (E00 = 2.20 eV). The
emission quenching of PdTPP by NEt3 in CH2Cl2 solution was
found to follow Stern−Volmer behavior (M0/M = 1 + KSV[Q],
exciting into the S2 state did the system produce moderate
amounts of CO.64 To further complicate matters, S2 excitation
where KSV = kqt0) and has a Stern−Volmer constant of KSV
=
of PdTPC results in quantitative chlorin photoreduction to a
24 M−1 (Supporting Information, Figure S10). The emission of
PdTPC is less trivial than that of PdTPP. First, the emission
intensity of PdTPC relative to PdTPP is significantly lower.
Excitation of PdTPC at 603 nm in CH2Cl2 gives an emission
maximum at 794 nm and two weaker bands at 614 and 659 nm.
max
bacteriochlorin species (Q-band λabs
= 588 nm), as
evidenced by steady-state UV−visible spectroscopy50,51 (Sup-
porting Information, Figure S6), while S1 excitation of PdTPC
gives only minimal amounts of bacteriochlorin (ΔAbs ∼0.4).
For comparison, the two-component system with tris-
(bipyridine) ruthenium(II) ([Ru(Bpy)3][PF6]2) and [Re-
(CO)3(Pic)Bpy][PF6] was also studied and proved to be
more effective than either of the two-component systems with
PdTPP or PdTPC sensitizers (Figure 4).
3
1
These have been tentatively assigned to the Q0,0, Q0,0, and
1Q1,0 transitions, respectively. Supporting Information, Figure
S11 shows the emission spectra of PdTPC excited at different
wavelengths, 524, 560, and 603 nm, all of which give the same
emission profile. With the assumption that the band at 614 nm
Ground State Spectroscopy. UV−vis Absorption Spec-
tra. The UV−vis absorption spectrum of [Re(CO)3(Pic)Bpy-
PdTPP][PF6] in THF solution matches that reported
previously,37 and it completely masks the absorption spectrum
of mononuclear [Re(CO)3(Pic)Bpy][PF6], even at low
concentrations (Supporting Information, Figure S7). There is
no discernible difference in the electronic spectra of Bpy-
PdTPP, [Re(CO)3BrBpy-PdTPP], and [Re(CO)3(Pic)Bpy-
PdTPP][PF6]. The mononuclear sensitizer PdTPC has a
distinctly different UV−vis absorption spectrum from that of
PdTPP (Supporting Information, Figure S8). While PdTPP
portrays a similar electronic spectrum to [Re(CO)3(Pic)Bpy-
PdTPP][PF6] with a Q-band absorption maximum at 523 nm
in CH2Cl2 solution (ε = 31 300 M−1 cm−1), the Q-band
1
is the Q0,0 transition, PdTPC has an E00 of 2.02 eV, lower by
0.18 eV than that of PdTPP. One could argue that the band at
614 nm is merely due to the 8% PdTPP impurity. If this were
the case we would expect to see triplet emission from PdTPP as
well, which was not observed.
Electrochemistry. The oxidation and reduction potentials
of [Re(CO)3(Pic)Bpy-PdTPP][PF6], PdTPP, and PdTPC
were determined by cyclic voltammetry in CH2Cl2 solution
with 0.1 M [Bu4N][PF6], and are reported versus the
ferrocenium/ferrocene couple (Fc+/0), which was used as an
internal standard in all experiments (Table 4). The cyclic
voltammograms for [Re(CO)3(Pic)Bpy-PdTPP][PF6] and
PdTPC are shown in Supporting Information, Figure S12. In
11883
dx.doi.org/10.1021/ic200243y|Inorg. Chem. 2011, 50, 11877−11889

Inorganic Chemistry
Article
Thus ΔGox* was calculated to be slightly favorable at −0.08 eV.
This differs to a small degree from our earlier report, ΔGox* =
−0.03 eV, but that calculation was performed using literature
values for PdTPP.37 This favorable electron transfer process is
also supported by the steady-state emission spectrum
(Supporting Information, Figure S7), where we observed the
intensity of the emission of [Re(CO)3(Pic)Bpy-PdTPP][PF6]
to be quenched by 37% relative to Bpy-PdTPP (no Re(I)) and
[Re(CO)3BrBpy-PdTPP] (neutral overall charge). It is also
noteworthy that intramolecular transfer from the Pd(II)
porphyrin T1 excited-state to Re(I)Bpy is a thermodynamically
uphill process, having ΔGox* = +0.35 eV (E00 for T1 is 1.76
eV).
To estimate ΔGox* for intermolecular transfer from PdTPP
or PdTPC to [Re(CO)3(Pic)Bpy][PF6] we simply use the Ered
potential of mononuclear [Re(CO)3(Pic)Bpy][PF6] (−1.57 V)
in eq 1. By doing so, we obtain ΔGox* values of +0.02 and
−0.17 eV for PdTPP and PdTPC, respectively. Thus
intermolecular electron transfer from PdTPC to [Re-
(CO)3(Pic)Bpy][PF6] appears to be a thermodynamically
favorable process, but ΔGox* for intermolecular electron
transfer from PdTPP to [Re(CO)3(Pic)Bpy][PF6] was
estimated to be nearly zero. It also follows that the
thermodynamic driving force for intermolecular electron
transfer from the sensitizer triplet state to catalyst is an uphill
process, having ΔGox* = +0.46 eV (E00 = 1.76 eV) for PdTPP:
[Re(CO)3(Pic)Bpy][PF6], and ΔGox* = +0.35 eV (E00 = 1.56
eV) for PdTPC:[Re(CO)3(Pic)Bpy][PF6].
Table 4. Redox Potentials and Free Energy Change (ΔGox*)
for Intra- and Intermolecular Electron Transfer from Pd(II)
Porphyrin/Chlorin S1 Excited States to Re(I)Bpy Moieties
ΔGox*
a
b
complex
E1/2ox/V (ΔE/mV) E1/2red/V (ΔE/mV)
/eV
[Re(CO)3(Pic)Bpy- +0.63 (66)
PdTPP][PF6]
−1.48 (90)
−0.08
PdTPP
PdTPC
+0.65 (95)
+0.34 (55)
−1.81 (136)
−1.79 (100)
−1.57 (99)
+0.02
−0.17
[Re(CO)3(Pic)Bpy]
[PF6]
a
Measured in CH2Cl2 with 0.1 M [Bu4N][PF6] and Fc+/0 as internal
b
ox
standard; 100 mV/s scan rate. Calculated from ΔGox* = E1/2
E1/2 − E00.
−
red
the experimental window, all complexes show one reversible
oxidation and one reversible reduction. [Re(CO)3(Pic)Bpy-
PdTPP][PF6] and PdTPP exhibit similar oxidation potentials
at +0.63 and +0.65 V, respectively, that are smaller than the
converted literature values for PdTPP (+0.74 V vs Fc+/0 but
measured under different conditions).65 PdTPC, having one of
its pyrrole rings fully hydrogenated, is predictably easier to
oxidize (+0.34 V) than either of the former complexes.
Conversely, the reduction potentials of PdTPP and PdTPC
are similar at −1.81 and −1.79 V, respectively, while that of
[Re(CO)3(Pic)Bpy-PdTPP][PF6] at −1.48 V resembles the
zinc analogue, [Re(CO)3(Pic)Bpy-ZnTPP][OTf] (−1.44 V vs
Fc+/0), yet differs from [Re(CO)3(Pic)Bpy][PF6] by 0.09 V
(Table 4). In an earlier study, we compared the reduction
potentials of [Re(CO)3(Pic)Bpy-ZnTPP][OTf] to that of a
rhenium complex functionalized with an 4-tBu-phenyl amido
unit to mimic the effect of the linker (the diimine ligand was 4-
methyl-2,2′-bipyridine-4′-carboxyamid-(4-tBu-phenyl)yl).35
Their potentials were −1.44 and −1.45 V vs Fc+/0 in THF
showing that the amidophenyl group makes the Re(Bpy)
moiety easier to reduce.
Considering that the oxidations in [Re(CO)3(Pic)Bpy-
PdTPP][PF6], PdTPP, and PdTPC are based on the
macrocycle it is expected that the PdTPC, being saturated at
one pyrrole ring, would be easier to oxidize. The reduction in
[Re(CO)3(Pic)Bpy-PdTPP][PF6] is localized on the bipyr-
idine ligand bound to Re(I) and thus has a reduction potential
similar to its analogous Zn dyad and to that of the mononuclear
[Re(CO)3(Pic)Bpy][PF6], also having a first reduction on the
bipyridine ligand.66 These results again emphasize the lack of
interaction between metalloporphyrin (Pd or Zn) sensitizer and
Re(I)Bpy in the ground state.
Picosecond Time-Resolved Infrared Spectroscopy (ps-
TRIR). We have investigated the photophysics of [Re-
(CO)3(Pic)Bpy-PdTPP][PF6] using ps-TRIR spectroscopy
since this approach probes the initial photoprocesses, and is a
powerful technique for interrogating intramolecular energy or
electron transfer in multinuclear complexes.68−70
Irradiation (λex = 532 nm) of a solution of [Re(CO)3(Pic)-
Bpy-PdTPP][PF6] in CH2Cl2 leads to the production of three
transient v(CO) bands at lower energy relative to those of the
parent complex at 2029, 2008, and 1908 cm−1 (Figure 5). The
Estimation of Driving Force (ΔGox*). From cyclic
voltammetry and steady-state emission experiments it is
possible to estimate the free energy change (ΔGox*) for
intramolecular photoinduced electron transfer from Pd(II)
porphyrin S1 excited-state to Re(I)Bpy in the [Re(CO)3(Pic)-
Bpy-PdTPP][PF6] molecular dyad using the simplified eq 1.
(1)
Figure 5. Picosecond TRIR spectra of [Re(CO)3(Pic)Bpy-PdTPP]-
[PF6] in CH2Cl2 (λex = 532 nm); bleach is due to the depletion of the
ground state complex, while π−π* and charge-separated (CS) are
bands from excited states of the complex. Notice the positive bands
overlapping the bleach (see 100 and 240 ps spectra), leading to its
apparent reformation.
where Eox and Ered are taken to be +0.63 and −1.48 V,
respectively, and E00 is 2.19 eV. While E00 is commonly
estimated from the highest energy emission maximum in a
frozen glass sample,67 a structured emission spectrum of
[Re(CO)3(Pic)Bpy-PdTPP][PF6] can be obtained at room
temperature. The value for E00 was therefore estimated from
the singlet emission band (1Q0,0) of [Re(CO)3(Pic)Bpy-
PdTPP][PF6] at 565 nm (Supporting Information, Figure S7).
transient band at 2029 cm−1 rises on a subpicosecond time
scale while the bands at 2008 and 1908 cm−1 grow in with a
dx.doi.org/10.1021/ic200243y|Inorg. Chem. 2011, 50, 11877−11889
11884

Inorganic Chemistry
Article
lifetime of about 3 ps (Figure 6). The three transient bands are
found to decay fully (τdecay ∼30 ps) with concomitant
Figure 7. Nanosecond TRIR spectra of [Re(CO)3(Pic)Bpy-PdTPP]-
[PF6] in CH2Cl2 with 1 M NEt3 (λex = 532 nm); bleach is due to the
depletion of the ground state complex, and CS indicates a charge-
separated excited-state complex. Notice the positive bands overlapping
the bleach (see 1, 10, and 120 ns spectra), which correspond to those
observed in the latter stages of the ps-TRIR experiments.
Figure 6. Kinetic traces for selected bands in the ps-TRIR experiment
shown in Figure 5 for [Re(CO)3(Pic)Bpy-PdTPP][PF6] in CH2Cl2
solution (λex = 532 nm). The triangles, solid squares, and circles show
experimental points. The continuous lines show the best-fit lines.
formation of two new bands at 2035 and 1935 cm−1, which
overlap with the parent bleaches, leading to its apparent
reformation (see the 100 and 240 ps traces in Figure 5).71
When sacrificial electron donor NEt3 (1 M) is added to the
solution, the same transient bands are observed but these
display different kinetic lifetimes. The species at 2029 cm−1 still
rises on a subpicosecond time scale but the other two bands at
2008 and 1908 cm−1 grow two times more slowly (ca. 6 ps)
than in the absence of NEt3. The decay of the band at 2029
cm−1 occurs faster than in the absence of NEt3 (ca. 20 vs 30 ps,
respectively). Conversely, the bands at 2008 and 1905 cm−1
decay more slowly than in the absence of NEt3 (ca. 45 vs 30 ps).
Again, two new v(CO) bands are observed to form which
overlap the parent bleaches, upon decay of the transients.
Nanosecond Time-Resolved Infrared Spectroscopy
(ns-TRIR). Experiments were performed on the nanosecond
time scale in an attempt to observe the catalytically active
species which leads to photosubstitution of Pic by bromide.
Excitation of [Re(CO)3(Pic)Bpy-PdTPP][PF6] in CH2Cl2
with a 532 nm laser, leads to the formation of two new
positive v(CO) bands overlapping the parent, which corre-
spond to those observed in the latter stages of the ps-TRIR
experiments. These bands decay with a lifetime of ∼1 μs,
leading to the apparent rebleaching of the parent, and the
formation of two additional bands at 2008 and 1903 cm−1. The
bands at 2008 and 1903 cm−1 have a lifetime of 10 μs, which
matches the recovery of the ground-state complex at 2032 and
1931 cm−1. Repeating these measurements in the presence of 1
M NEt3 leads to identical spectra, but with significantly
different kinetic behavior, Figures 7 and 8. The transient bands
Figure 8. Kinetic traces for selected bands in the ns TRIR experiment
shown in Figure 7 for [Re(CO)3(Pic)Bpy-PdTPP][PF6] in CH2Cl2
solution with 1 M NEt3 (λex = 532 nm). The squares, solid squares,
and circles show experimental points. The continuous lines show the
best-fit lines.
at 2008 and 1903 cm−1 now grow in and decay more slowly
than in the absence of NEt3 (τgrowth ∼ 2 μs; τdecay ∼ 45 μs), and
these rates closely match those of the parent rates of recovery.
Even with the experiments described above, there is still a
need to link the ns-TRIR and the steady-state remote-site
photosubstitution reactions, and we have used a combination of
microsecond TRIR and rapid-scan FTIR spectroscopies to
probe this further. We found that a single laser shot (λ ex = 532
nm) on a solution of [Re(CO)3(Pic)Bpy-PdTPP][PF6] in
CH2Cl2 in the presence of NEt3 (1 M) and [Bu4N][Br] (0.1
M) leads to identical results to those obtained in the fast ns-
TRIR experiments. No evidence for growth of the substituted
bromide complex, [Re(CO)3BrBpy-PdTPP], was obtained
during this experiment (Supporting Information, Figure S13).
11885
dx.doi.org/10.1021/ic200243y|Inorg. Chem. 2011, 50, 11877−11889

Inorganic Chemistry
Article
However, when the same solution was subjected to pulsed laser
shots for 3 s at 10 Hz, a long-lived product is observed (v(CO)
2010 and 1900 cm−1) which may indicate formation of a one-
electron reduced (OER) species (Supporting Information,
Figure S13). Monitoring this reaction with rapid-scan FTIR
allows us to measure the rate of substitution of Pic by bromide.
The decay of this species is the same as the growth of the
bromide product, [Re(CO)3BrBpy-PdTPP] (kobs = 0.12
(±0.01) s−1) (Supporting Information, Figure S13).
Re(I)Bpy catalyst upon S2 excitation, but showed minimal
efficiency for CO2 reduction with a quantum yield of 0.0064 for
CO production.38 Charge recombination was thought to
compete with electron transfer quenching from the sacrificial
donor and limit catalytic activity. However, in that report the
authors did not take into account that the Zn(II) porphyrin
precoordinates sacrificial NEt3, and transfers an electron by an
intramolecular process.36,40,77,78 While other Zn(II) porphyr-
in−Re(I)Bpy dyads have been synthesized by Casanova et al.,39
the complexes were not tested for photochemical CO2
reduction. Interestingly, in a separate study Zhang et al. report
the photochemical production of H2 using a dyad with Zn(II)
porphyrin and a cobaloxime catalyst in place of Re(I)Bpy.40
With that dyad the authors obtained TNs H2 on the order of
∼25, and there is no mention of photoreduction at the Zn(II)
porphyrin core.
The dyad [Re(CO)3(Pic)Bpy-PdTPP][PF6] reported here
for the photoreduction of CO2 to CO results in only 2 TNs
CO, detectable after an induction period, and requires
excitation into the Soret (S2) band. The complex undergoes
competing reduction reactions at the porphyrin core that form
chlorin and bacteriochlorin side products. When sensitizer and
catalyst are used as two separate components in place of the
dyad, higher TNs CO, up to 9 with PdTPC chromophore, are
obtained. This observation is an indication that back-electron
transfer from the charge-separated state is a limiting factor for
[Re(CO)3(Pic)Bpy-PdTPP][PF6] to function as a dyad,
though a two-component system still requires collision between
compounds.79
TRIR Product Assignment. Picosecond Time Scale.
The TRIR measurements in the region (ca. 1800−2100 cm−1)
are sensitive to the oxidation state of Re(I) and the LUMO on
the Bpy, as well as mildly sensitive to the peripheral ligand
bound to Re(I), 3-picoline for the current case. The TRIR
spectrum is much less sensitive to the state of the Pd(II)
porphyrin component of the dyad. In the ps-TRIR experiments,
both in the presence and absence of NEt3, a reduced species of
[Re(CO)3(Pic)Bpy-PdTPP][PF6] can be detected, with
product bands shifted to lower wavenumber relative to the
ground state complex. This shift is indicative of a reduction at
the Re(I)Bpy moiety,29,80−82 likely to be localized on the lowest
unoccupied molecular orbital (LUMO) of the coordinated Bpy,
which leads to weakening of the CO bond because of increased
back-bonding with Re(I). The excited-state bands appear at the
same wavenumber in both picosecond experiments, but have
differing kinetic lifetimes. Because it is not possible to
distinguish between a charge-separated (CS) state,
[Re+(CO)3(Pic)Bpy•−-PdTPP•+], and a radical species,
[Re+(CO)3(Pic)Bpy−-PdTPP], with TRIR spectroscopy, we
have distinguished between these product assignments based
on the kinetics.
In the ps-TRIR experiments the growth and decay of product
bands at 2008 and 1908 cm−1 is so rapid that we suggest the
reduced species to be a CS state complex, [Re+(CO)3(Pic)-
Bpy•−-PdTPP•+]. In other words, upon laser excitation an
electron−hole pair is formed in 3−6 ps, the hole being at the
PdTPP•+ core and the electron at the Re(I)Bpy•− center. The
charge recombination then occurs within 30−45 ps. The small
changes in lifetimes are probably caused, at least in part, by
changes in solvent dynamics. While the band at 2029 cm−1
develops on a subpicosecond time scale, its decay time (20−30
ps) is similar to the other, CS excited-state bands, and has been
tentatively assigned to a π−π* transition by comparison to the
DISCUSSION
■
[Re(CO)3(Pic)Bpy-PdTPP][PF6]: CO2 Reduction in Per-
spective. We first compare [Re(CO)3(Pic)Bpy-PdTPP]-
[PF6] to other dyads, but we note that quantitative
comparisons are not possible without using the same
experimental conditions (i.e., solvent system, sacrificial reagent,
light source, reaction vessel, etc.).
Dyads that contain Ru(Bpy)32+, or derivatives thereof, as the
photosensitizer component are by far the most studied
complexes toward the photochemical reduction of CO2 to
CO. Kimura et al. first reported three such dyads with the
sensitizer linked to a Ni(cyclam)2+ (cyclam = 1,4,8,11-
tetraazacyclotetradecane) catalyst.72,73 While Ni(cyclam)2+ is a
highly selective electrocatalyst for CO2 reduction to CO,74 the
molecular dyads proved to be nonspecific and inefficient
photocatalysts, producing mixtures of CO and H2, with TNs
CO amounting to less than one.75 Kasuga and co-workers later
reported dyads with a Co(III) or Ni(II) Bpy catalytic site,76 but
again CO2 reduction was not selective. The systems produced
mixtures of H2, CO, and formic acid, and TNs CO were lower
than 5.75 Further, greater yields of photoreduction products
were obtained, for the most part, when the sensitizer and
catalyst were used as separate entities.
More recent studies of dyads with Ru(Bpy)32+-like
chromophores have extended to heteronuclear complexes
with a Re(I)Bpy catalytic component. Ishitani and co-workers
were the first to study such systems21 and have since developed
the most effective of these Ru(Bpy)3−Re(I)Bpy dyads, a
photocatalyst capable of producing greater than 230 TNs
CO.25,27 Through their studies the authors have initiated a set
of principles for constructing effectual two-component photo-
catalysts, putting emphasis on (1) slow rates of electron transfer
from the chromophore to the catalystas CO2 reduction itself
is the rate-limiting process; (2) the ancillary ligand bound to
the Re(I)Bpy catalyst (i.e., P(OEt)3, pyridine, halide, etc.); (3)
directional electron transfer into the bridging diimine ligand of
the chromophore; and (4) the use of an aliphatic, non-
conjugated connection between chromophore and catalyst. In
contrast to the earlier studies of dyads with Ni(II) or Co(III)
catalytic sites, the dyads reported by Ishitani tend to
outperform systems with the sensitizer and catalyst as separate
components. More recently, Bian et al. have demonstrated
Ru(Bpy)3−Re(I)Bpy dyads that hold up Ishitani’s principles,
particularly that an aliphatic link between Ru(Bpy)3 and
Re(I)Bpy proves a better photocatalyst than the conjugated
analogue.26
To the best of our knowledge there is only one other report
of a dyad that uses a metalloporphyrin sensitizer covalently
linked to a Re(I)Bpy catalytic site for the photoconversion of
CO2 to CO. The preliminary study of a Zn(II) porphyrin−
Re(I)Bpy dyad, having a similar molecular structure to that of
[Re(CO)3(Pic)Bpy-PdTPP][PF6], appeared to undergo ultra-
fast electron transfer from the Zn(II) porphyrin to the
11886
dx.doi.org/10.1021/ic200243y|Inorg. Chem. 2011, 50, 11877−11889

Inorganic Chemistry
Article
results obtained previously in the ps-TRIR experiments for the
analogous [Re(CO)3(Pic)Bpy-ZnTPP][PF6] complex.36
The apparent inconsistencies in the kinetics of the product
bands assigned to π−π* and charge-separated excited-states are
of particular concern. One plausible explanation is that
[Re(CO)3(Pic)Bpy-PdTPP][PF6] adopts multiple conforma-
tions in solution.36 As a result, the rates of forward and back
electron transfer, as determined by ps-TRIR, are dependent on
the torsional angles between the planes of the different
molecular constituents: PdTPP porphyrin, Re(I)Bpy, and the
phenylamide bridge that links them together. The solid state X-
ray structure of [Re(CO)3(NCS)Bpy-PdTPP] gives insight
into one possible conformation (Figure 2, Table 2). The data
show that the phenylamide link is virtually coplanar within itself
(178.1(7)°), but is slightly out of plane with the Re(I)Bpy, by
nearly 23°, and is substantially distorted from the core plane of
the Pd(II) porphyrin having a torsional angle of 69.1(9)°.
Conformers that resemble the solid state [Re(CO)3(NCS)-
Bpy-PdTPP] will most likely have a higher population than
conformers that are nearly coplanar,36 but there will still be a
distribution of conformers. The ps-TRIR data are therefore a
summation of the contributions of different conformers and
excited-state conformers present in solution, and may adopt a
different distribution in CH2Cl2 from that in CH2Cl2 with 1 M
NEt3. Alternatively, another kinetic model postulates an
ultrafast equilibrium between π−π* and CS states.36 This
model gives significance to the only slightly favorable
thermodynamic driving force for intramolecular electron
transfer from Pd(II) porphyrin S1 excited-state to Re(I)Bpy,
estimated to be ΔGox* = −0.08 eV.
Nanosecond Time Scale. While the TRIR spectra in the
nanosecond regime are similar in structure to those on the
picosecond time scale, the kinetics are clearly different. There is
no evidence for a π−π* excited-state in the ns-TRIR, but
product bands at 2008 and 1903 cm−1 grow and decay in about
1 and 10 μs in CH2Cl2, respectively, and about 2 and 45 μs in
CH2Cl2 with 1 M NEt3, respectively. The product bands
observed in the ns-TRIR are tentatively assigned to a charge-
separated excited-state complex, different from the CS excited-
state complex observed in the ps-TRIR experiments. In
addition to changes in solvent dynamics in the presence of 1
M NEt3, two other processes enter the frame on such long time
scales: (a) substitution of coordinated picoline by NEt3, (b)
formation of porphyrin reduction products (especially chlorin).
It would be difficult to distinguish an excited state complex with
coordinated NEt3 from one with coordinated picoline using
TRIR spectroscopy. Porphyrin reduction products should be
observable by transient absorption. We tried such experiments
on pico and nanosecond time scales (but not microsecond) and
observed no evidence for chlorins.
separation and subsequent recombination, with full recovery of
ground state complex in a few picoseconds. Nanosecond TRIR
experiments give evidence to a different CS excited-state species
that grows and decays in microseconds, but without full
recovery of the ground state complex. In microsecond TRIR
experiments that follow a single laser shot of [Re(CO)3(Pic)-
Bpy-PdTPP][PF6] in the presence of excess bromide, the same
CS excited-state complex as that observed in the ns-TRIR can
be detected (Supporting Information, Figure S13). When the
same system is pulsed for 3 s (at 10 Hz), a different, long-lived
species is then detected and has been tentatively assigned to a
one-electron reduced (OER) species, [Re+(CO)3(Pic)Bpy−-
PdTPP]. The OER species and the photosubstituted product,
[Re(CO)3BrBpy-PdTPP], decay and grow at the same rate on
the order of seconds, but the kinetics are hardly trivial
(Supporting Information, Figure S13). It is not possible to
make unambiguous assignments for these processes. They are
complicated by photosubstitution of Pic by NEt3, electronic
states of the electron-rich asymmetric porphyrin core, different
geometric conformers, and porphyrin reduction products that
may be in solution.
CONCLUSIONS
■
We have studied in detail the photophysical and photochemical
properties of a molecular dyad made up of a Pd(II)
metalloporphyrin sensitizer covalently linked to a cationic
Re(I) tricarbonyl bipyridine complex. The pendant arm linking
the porphyrin to the Bpy is distinctly twisted in the structure of
the NCS derivative. Steady-state photolysis of [Re(CO)3(Pic)-
Bpy-PdTPP][PF6] with λex = 520 nm causes substitution of
the labile 3-picoline ligand. When employed as a single
component for the photoreduction of CO2, [Re(CO)3(Pic)-
Bpy-PdTPP][PF6] produces 2 TNs CO when using λex > 420
nm light and only trace amounts of CO when using λex = 520
nm light, and the reaction is complicated by photoreduction to
chlorin. The two-component system involving PdTPP or
PdTPC and [Re(CO)3(Pic)Bpy][PF6] is slightly better for CO
formation, but still lacks the efficacy of the analogous two-
component system with [Ru(Bpy)3][PF6]2 as the sensitizer. As
an alternative approach, one might envisage increasing the
separation between sensitizer and catalyst within a dyad, and we
are currently investigating such methodology. The two systems
in this study that were found to be most effective for the
photoreduction of CO2,64 were the mixture of PdTPC:
[Re(CO)3(Pic)Bpy][PF6] and the mixture of [Ru(Bpy)3]-
[PF6]2:[Re(CO)3(Pic)Bpy][PF6], giving 9 TNs of CO after 4 h
photolysis with λex > 420 nm for the former and 11 TNs CO
with λex > 495 nm for the latter. TRIR spectroscopy of
[Re(CO)3(Pic)Bpy-PdTPP][PF6] with excitation at λ = 532
nm on a picosecond time scale shows formation of a CS state.
The ensuing rapid back electron transfer is a significant
limitation of this dyad. Investigation by TRIR spectroscopy on
longer time scales reveals further species that are kinetically
distinct from the CS state, but have similar IR spectra in the
ν(CO) region.
At this stage the assignment of the bands that overlap with
the parent bleach in both the ps- and the ns-TRIR experiments
is unclear. These bands form at 2035 and 1935 cm−1 in the later
stages of the ps-TRIR, as the transient bands are recovering.
They are present in the very early stages of the ns-TRIR
experiments, and decay at the same rate as the bleach of the
parent bands.
Differences in Time Scales: from Picoseconds up to
Steady-State Conditions. There appears to be some
disparity when comparing the kinetics of the product bands
of [Re(CO)3(Pic)Bpy-PdTPP][PF6] in the time-resolved IR
experiments, starting from ps-TRIR and working up to steady-
state conditions. The ps-TRIR is dominated by ultrafast charge-
ASSOCIATED CONTENT
■
S
* Supporting Information
1
Additional H NMR, UV−vis, and emission spectra, cif file of
[Re(CO)3(NCS)Bpy-PdTPP]·C6H6, cyclic voltammograms,
kinetics of photosubstitution, Stern−Volmer plot, and a
summary of the photoreduction of CO2 in the presence of
11887
dx.doi.org/10.1021/ic200243y|Inorg. Chem. 2011, 50, 11877−11889

Inorganic Chemistry
Article
(27) Sato, S.; Koike, K.; Inoue, H.; Ishitani, O. Photochem. Photobiol.
Sci. 2007, 6, 454−461.
BNAH. This material is available free of charge via the Internet
at http://pubs.acs.org.
(28) Hori, H.; Johnson, F. P. A.; Koike, K.; Ishitani, O.; Ibusuki, T. J.
Photochem. Photobiol., A 1996, 96, 171−174.
AUTHOR INFORMATION
■
(29) Koike, K.; Hori, H.; Ishizuka, M.; Westwell, J. R.; Takeuchi, K.;
Ibusuki, T.; Enjouji, K.; Konno, H.; Sakamoto, K.; Ishitani, O.
Organometallics 1997, 16, 5724−5729.
(30) Kurz, P.; Probst, B.; Spingler, B.; Alberto, R. Eur. J. Inorg. Chem.
2006, 2966−2974.
Corresponding Author
*E-mail: robin.perutz@york.ac.uk (R.N.P.), mike.george@
nottingham.ac.uk (M.W.G.).
(31) Tsubaki, H.; Sekine, A.; Ohashi, Y.; Koike, K.; Takeda, H.;
Ishitani, O. J. Am. Chem. Soc. 2005, 127, 15544−15555.
(32) Ishitani, O.; George, M. W.; Ibusuki, T.; Johnson, F. P. A.;
Koike, K.; Nozaki, K.; Pac, C. J.; Turner, J. J.; Westwell, J. R. Inorg.
Chem. 1994, 33, 4712−4717.
(33) Hori, H.; Ishihara, J.; Koike, K.; Takeuchi, K.; Ibusuki, T.;
Ishitani, O. J. Photochem. Photobiol., A 1999, 120, 119−124.
(34) Aspley, C. J.; Lindsay Smith, J. R.; Perutz, R. N. J. Chem. Soc.,
Dalton Trans. 1999, 2269−2271.
(35) Gabrielsson, A.; Hartl, F.; Lindsay Smith, J. R.; Perutz, R. N.
Chem. Commun. 2002, 950−951.
(36) Gabrielsson, A.; Hartl, F.; Zhang, H.; Lindsay Smith, J. R.;
Towrie, M.; Vlcek, A.; Perutz, R. N. J. Am. Chem. Soc. 2006, 128,
4253−4266.
(37) Gabrielsson, A.; Lindsay Smith, J. R.; Perutz, R. N. Dalton Trans.
2008, 4259−4269.
(38) Kiyosawa, K.; Shiraishi, N.; Shimada, T.; Masui, D.; Tachibana,
H.; Takagi, S.; Ishitani, O.; Tryk, D. A.; Inoue, H. J. Phys. Chem. C
2009, 113, 11667−11673.
(39) Casanova, M.; Zangrando, E.; Iengo, E.; Alessio, E.; Indelli, M.
T.; Scandola, F.; Orlandi, M. Inorg. Chem. 2008, 47, 10407−10418.
(40) Zhang, P.; Wang, M.; Li, C.; Li, X.; Dong, J.; Sun, L. Chem.
Commun. 2010, 46, 8806−8808.
(41) Lees, A. J. Anal. Chem. 1996, 68, 226−229.
(42) Towrie, M.; Grills, D. C.; Dyer, J.; Weinstein, J. A.; Matousek,
P.; Barton, R.; Bailey, P. D.; Subramaniam, N.; Kwok, W. M.; Ma, C.;
Phillips, D.; Parker, A. W.; George, M. W. Appl. Spectrosc. 2003, 57,
367−380.
(43) Brennan, P.; George, M. W.; Jina, O. S.; Long, C.; McKenna, J.;
Pryce, M. T.; Sun, X.-Z.; Vuong, K. Q. Organometallics 2008, 27,
3671−3680.
(44) Rigaku images were converted to Apex II images using the code
called “eclipse” written by Professor Simon Parsons of The University
of Edinburgh.
ACKNOWLEDGMENTS
■
The authors wish to thank the EPSRC for funding (projects
EP/F047770/1 and EP/F047789/1) and our collaborators in
the Consortium for Artificial Photosynthesis (SolarCAP). We
thank the EPSRC National Crystallography Service synchro-
tron component operated by Newcastle University at Diamond
Light Source beamline I19 for X-ray data collection for
[Re(CO)3(NCS)Bpy-PdTPP]. We also acknowledge Dr.
Richard E. Douthwaite for use of the radiometer. MWG
gratefully acknowledges receipt of a Royal Society Wolfson
Merit Award.
REFERENCES
■
(1) Gray, H. B. Nature Chem. 2009, 1, 7.
(2) Barber, J. Chem. Soc. Rev. 2009, 38, 185−196.
(3) Morris, A. J.; Meyer, G. J.; Fujita, E. Acc. Chem. Res. 2009, 42,
1983−1994.
(4) Roy, S. C.; Varghese, O. K.; Paulose, M.; Grimes, C. A. ACS
Nano 2010, 4, 1259−1278.
(5) Yui, T.; Tamaki, Y.; Sekizawa, K.; Ishitani, O. Top. Curr. Chem.
2011, 303, 151−184.
(6) Darensbourg, D. J. Inorg. Chem. 2010, 49, 10765−10780.
(7) Hoffmann, M. R.; Moss, J. A.; Baum, M. M. Dalton Trans. 2011,
40, 5151−5158.
(8) Hebrard, F.; Kalck, P. Chem. Rev. 2009, 109, 4272−4282.
(9) Leckel, D. Energy Fuels 2009, 23, 2342−2358.
(10) Benson, E. E.; Kubiak, C. P.; Sathrum, A. J.; Smieja, J. M. Chem.
Soc. Rev. 2009, 38, 89−99.
(11) Rakowski DuBois, M.; DuBois, D. L. Acc. Chem. Res. 2009, 42,
1974−1982.
́
(12) Saveant, J.-M. Chem. Rev. 2008, 108, 2348−2378.
(13) Hawecker, J.; Lehn, J.-M.; Ziessel, R. J. Chem. Soc., Chem.
Commun. 1983, 536−538.
(45) APEX2 software; Bruker AXS: Madison, WI, 2009.
(46) Sheldrick, G. M. SHELXS-97, a program for crystal structure
(14) Hawecker, J.; Lehn, J. M.; Ziessel, R. Helv. Chim. Acta 1986, 69,
1990−2012.
solution; Gottingen University: Gottingen, Germany, 1997.
̈
̈
(47) Sheldrick, G. M. SHELXL-97, a program for crystal structure
(15) Hawecker, J.; Lehn, J. M.; Ziessel, R. J. Chem. Soc., Chem.
Commun. 1985, 56−58.
determination; Gottingen University: Gottingen, Germany, 1997.
̈
̈
(48) McGee, K. A.; Veltkamp, D. J.; Marquardt, B. J.; Mann, K. R. J.
Am. Chem. Soc. 2007, 129, 15092−15093.
(16) Lehn, J. M.; Ziessel, R. Proc. Natl. Acad. Sci. U.S.A. 1982, 79,
701−704.
(49) Kutal, C.; Weber, M. A.; Ferraudi, G.; Geiger, D. Organo-
metallics 1985, 4, 2161−2166.
(17) Agarwal, J.; Johnson, R. P.; Li, G. J. Phys. Chem. A 2011, 115,
2877−2881.
(50) Harel, Y.; Manassen, J. J. Am. Chem. Soc. 1978, 100, 6228−6234.
(51) Whitlock, H. W.; Hanauer, R.; Oester, M. Y.; Bower, B. K. J.
Am. Chem. Soc. 1969, 91, 7485−7489.
(18) Takeda, H.; Koike, K.; Inoue, H.; Ishitani, O. J. Am. Chem. Soc.
2008, 130, 2023−2031.
(19) Hayashi, Y.; Kita, S.; Brunschwig, B. S.; Fujita, E. J. Am. Chem.
Soc. 2003, 125, 11976−11987.
(52) Mauzerall, D.; Westheimer, F. H. J. Am. Chem. Soc. 1955, 77,
2261−2264.
(20) Kirgan, R. A.; Sullivan, B. P.; Rillema, D. P. Top. Curr. Chem.
2007, 281, 45−100.
(53) Schanze, K. S.; Lee, L. Y. C.; Giannotti, C.; Whitten, D. G. J.
Am. Chem. Soc. 1986, 108, 2646−2655.
(21) Gholamkhass, B.; Mametsuka, H.; Koike, K.; Tanabe, T.; Furue,
M.; Ishitani, O. Inorg. Chem. 2005, 44, 2326−2336.
(22) Bian, Z.-Y.; Sumi, K.; Furue, M.; Sato, S.; Koike, K.; Ishitani, O.
Inorg. Chem. 2008, 47, 10801−10803.
(54) Ion, R. M.; Mandravel, C.; Bercu, C. Rev. Chim. 1998, 49, 121−
127.
(55) Sapunov, V. V.; Egorova, G. D. Russ. J. Phys. Chem. A 1992, 66,
471−477.
(23) Bian, Z.-Y.; Sumi, K.; Furue, M.; Sato, S.; Koike, K.; Ishitani, O.
Dalton Trans. 2009, 983−993.
(24) Koike, K.; Naito, S.; Sato, S.; Tamaki, Y.; Ishitani, O. J.
Photochem. Photobiol., A 2009, 207, 109−114.
(25) Takeda, H.; Ishitani, O. Coord. Chem. Rev. 2010, 254, 346−354.
(26) Bian, Z.-Y.; Chi, S.-M.; Li, L.; Fu, W. Dalton Trans. 2010, 39,
7884−7887.
(56) Sapunov, V. V.; Egorova, G. D. J. Appl. Spectrosc. 1989, 50,
611−615.
(57) Sapunov, V. V.; Egorova, G. D. J. Appl. Spectrosc. 1987, 46,
519−522.
(58) Sapunov, V. V.; Solov’ev, K. N. Vestsi Akad. Navuk BSSR, Ser.
Fiz.-Mat. Navuk 1980, 60−64.
11888
dx.doi.org/10.1021/ic200243y|Inorg. Chem. 2011, 50, 11877−11889

Inorganic Chemistry
Article
(59) Sapunov, V. V.; Solov’ev, K. N.; Egorova, G. D. Russ. J. Phys.
Chem. B 1986, 5, 1044−1048.
(80) Johnson, F. P. A.; George, M. W.; Hartl, F.; Turner, J. J.
Organometallics 1996, 15, 3374−3387.
(81) Stor, G. J.; Hartl, F.; van Outersterp, J. W. M.; Stufkens, D. J.
Organometallics 1995, 14, 1115−1131.
(60) Blanco Rodriguez, A. M.; Gabrielsson, A.; Motevalli, M.;
̌
Matousek, P.; Towrie, M.; Sebera, J.; Zalis, S.; Vlcek, A. J. Phys. Chem.
(82) van Outersterp, J. W. M.; Hartl, F.; Stufkens, D. J.
Organometallics 1995, 14, 3303−3310.
A 2005, 109, 5016−5025.
(61) Fleischer, E. B.; Miller, C. K.; Webb, L. E. J. Am. Chem. Soc.
1964, 86, 2342−2347.
NOTE ADDED AFTER ASAP PUBLICATION
(62) The absolute quantum yield (φ) for product formation was
determined to be φ = 0.12 ± 0.01. Although the efficiency for this
photoreaction was calculated using a procedure that accounts for
substantial inner filter effects41 between the reactant and the product,
there was some evidence for an induction period that makes the
quantum yield difficult to interpret. In each experiment we observed
transformation of [Re(CO)3(Pic)Bpy-PdTPP][PF6] to [Re(-
CO)3BrBpy-PdTPP]; reproducibility was complicated by a variation
of both the induction period and the rate of the photoreaction. The
efficiency reported is for the experiment with a minimal induction
period.
■
This article was published on the Web on November 1, 2011.
Due to a production error, text changes were not incorporated
in this version. The corrected version was reposted on
November 2, 2011.
(63) Mercer-Smith, J. A.; Sutcliffe, C. R.; Schmehl, R. H.; Whitten,
D. G. J. Am. Chem. Soc. 1979, 101, 3995−3997.
(64) When 0.1 M BNAH is added to the reaction mixture, more TNs
CO are formed. However, the reaction becomes quite complicated,
and we do not understand the effect of the BNAH at this point. Details
of preliminary experiments are outlined in the Supporting Information.
(65) Giraudeau, A.; Callot, H. J.; Gross, M. Inorg. Chem. 1979, 18,
201−206.
(66) Sacksteder, L.; Zipp, A. P.; Brown, E. A.; Streich, J.; Demas, J.
N. Inorg. Chem. 1990, 29, 4335−4340.
(67) Turro, N. J. Modern Molecular Photochemistry, 2nd ed.; The
Benjamin/Cummings Publishing Company, Inc.: Menlo Park, 1978; p
628.
(68) Easun, T. L.; Alsindi, W. Z.; Towrie, M.; Ronayne, K. L.; Sun,
X.-Z.; Ward, M. D.; George, M. W. Inorg. Chem. 2008, 47, 5071−5078.
(69) Butler, J. M.; George, M. W.; Schoonover, J. R.; Dattelbaum, D.
M.; Meyer, T. J. Coord. Chem. Rev. 2007, 251, 492−514.
(70) Vergeer, F. W.; Kleverlaan, C. J.; Matousek, P.; Towrie, M.;
Stufkens, D. J.; Hartl, F. Inorg. Chem. 2005, 44, 1319−1331.
(71) Picosecond transient UV/visible data showed that excitation
(λex = 532 nm) in CH2Cl2 leads to production of an intense band at
450 nm within the rise time of the instrument (<1 ps). This decays
partially over the first 10−30 ps and then fully within a few hundred
picoseconds, i.e., on a similar time scale as the rise of the band at 2035
and 1935 cm−1 in the ps-TRIR.
(72) Kimura, E.; Bu, X.; Shionoya, M.; Wada, S.; Maruyama, S. Inorg.
Chem. 1992, 31, 4542−4546.
(73) Kimura, E.; Wada, S.; Shionoya, M.; Okazaki, Y. Inorg. Chem.
1994, 33, 770−778.
(74) Beley, M.; Collin, J. P.; Ruppert, R.; Sauvage, J. P. J. Am. Chem.
Soc. 1986, 108, 7461−7467.
(75) While TNs CO were not reported by the authors, they can be
inferred from their data.
(76) Komatsuzaki, N.; Himeda, Y.; Hirose, T.; Sugihara, H.; Kasuga,
K. Bull. Chem. Soc. Jpn. 1999, 72, 725−731.
(77) Kluwer, A. M.; Kapre, R.; Hartl, F.; Lutz, M.; Spek, A. L.;
Brouwer, A. M.; van Leeuwen, P. W. N. M.; Reek, J. N. H. Proc. Natl.
Acad. Sci. U.S.A. 2009, 106, 10460−10465.
(78) Li, X.; Wang, M.; Zhang, S.; Pan, J.; Na, Y.; Liu, J.; Akermark,
B.; Sun, L. J. Phys. Chem. B 2008, 112, 8198−8202.
(79) If we use the Stern−Volmer constant of KSV = 24 M−1
quenching of PdTPP by NEt3 in CH2Cl2 solution (Supporting
Information, Figure S10) and a literature value for the excited-state
lifetime of PdTPP, τ0 = 385 μs (Harriman, A. J. Chem. Soc., Faraday
1200 Trans. 2 1981, 77, 1281−1291) in the absence of quencher, we
obtain a quenching constant, kq = 6.5 × 104 M−1 s−1. It is therefore
expected that emission quenching of PdTPP or [Re(CO)3(Pic)Bpy-
PdTPP][PF6] by NEt3 will occur under photochemical CO2 reduction
conditions, i.e., 5:1 (v/v) DMF:NEt3 (1.2 M NEt3) with a pseudo first
order rate constant of ca. 104 to 105 s−1.
11889
dx.doi.org/10.1021/ic200243y|Inorg. Chem. 2011, 50, 11877−11889