Effect of ligand derivatization at different positions on photochemical properties of hybrid Re(I) photocatalysts

Article abstract of DOI:10.1016/j.molcata.2015.10.034

Hybrid CO₂-reduction photocatalysts based on Re(bpy)(CO)₃Cl, where bpy is 2,2′-bipyridine, were synthesized in order to investigate the effect of different ligand derivatization strategies on photochemical properties of the hybrid Re(I) systems. Derivatization of the bpy ligand was carried out at the 4,4′- and 5,5′-positions with electron-withdrawing amide groups. The derivatized ligands were grafted on mesoporous silica via a dipodal silane coupling agent and were further coordinated with Re(I). The synthesized hybrid photocatalysts were studied using spectroscopic techniques, including UV-visible, in situ infrared and electron paramagnetic resonance (EPR) spectroscopy, and tested in photochemical CO₂ reduction. An interesting light-induced color change was observed for the hybrid photocatalyst involving derivatization of the bpy ligand at the 5,5′-positions. Combined UV-visible and EPR studies indicated that in the reduced form of this hybrid photocatalyst electron density was more localized on the bpy ligand than on the Re(I) center. Further studies with in situ infrared spectroscopy demonstrated possible formation of a carbonate-bridged binuclear Re(I) species on this hybrid photocatalyst. Our results are particularly relevant to developing new molecular and hybrid photocatalytic systems which involve extensive derivatization of coordinating ligands with functional groups for improved solar energy conversion.

Full text of DOI:10.1016/j.molcata.2015.10.034

Journal of Molecular Catalysis A: Chemical 411 (2016) 272–278
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Effect of ligand derivatization at different positions on photochemical
properties of hybrid Re(I) photocatalysts
Thomas G. Fenton, Michael E. Louis, Gonghu Li∗
Department of Chemistry and Materials Science Program, University of New Hampshire, Durham, NH 03824, United States
a r t i c l e i n f o
a b s t r a c t
Hybrid CO2-reduction photocatalysts based on Re(bpy)(CO)3Cl, where bpy is 2,2ꢀ-bipyridine, were syn-
thesized in order to investigate the effect of different ligand derivatization strategies on photochemical
properties of the hybrid Re(I) systems. Derivatization of the bpy ligand was carried out at the 4,4ꢀ- and
5,5ꢀ-positions with electron-withdrawing amide groups. The derivatized ligands were grafted on meso-
porous silica via a dipodal silane coupling agent and were further coordinated with Re(I). The synthesized
hybrid photocatalysts were studied using spectroscopic techniques, including UV–visible, in situ infrared
and electron paramagnetic resonance (EPR) spectroscopy, and tested in photochemical CO2 reduction.
An interesting light-induced color change was observed for the hybrid photocatalyst involving derivati-
zation of the bpy ligand at the 5,5ꢀ-positions. Combined UV–visible and EPR studies indicated that in the
reduced form of this hybrid photocatalyst electron density was more localized on the bpy ligand than on
the Re(I) center. Further studies with in situ infrared spectroscopy demonstrated possible formation
of a carbonate-bridged binuclear Re(I) species on this hybrid photocatalyst. Our results are particu-
larly relevant to developing new molecular and hybrid photocatalytic systems which involve extensive
derivatization of coordinating ligands with functional groups for improved solar energy conversion.
© 2015 Elsevier B.V. All rights reserved.
Article history:
Received 29 June 2015
Received in revised form 21 October 2015
Accepted 30 October 2015
Available online 4 November 2015
Keywords:
Rhenium
CO2 reduction
Photocatalysis
Ligand derivatization
Spectroscopy
1. Introduction
In CO2-reduction photocatalysis, hybrid photocatalysts generally
show enhanced reactivity and improved stability under pho-
Photocatalysis represents the most promising approach to
converting CO2 into useful materials, chemicals and fuels in a
sustainable fashion [1–3]. Molecular catalysts based on transi-
tion metal complexes have been extensively studied in solar fuel
production via CO2 reduction [4–6]. Many studies focused on
diimine-tricarbonyl rhenium complexes, including Re(bpy)(CO)3Cl
where bpy is 2,2ꢀ-bipyridine [7–11]. These Re(I)-based complexes
have demonstrated excellent reactivity and selectivity to CO forma-
tion in photocatalytic CO2 reduction. In this study, we investigate
how derivatizing the bpy ligand at different positions affects
photochemical properties of hybrid photocatalysts prepared by
covalently grafting Re(bpy)(CO)3Cl onto mesoporous silica.
A variety of hybrid photocatalysts have been reported for use
in solar fuel production via CO2 reduction [12–15]. The hybrid
photocatalysts are often prepared by coupling molecular cata-
lysts with solid-state surfaces via different strategies including
covalent attachment [16–18], incorporation in frameworks/porous
environment [19–21] and polymerization on surfaces [22,23].
tochemical conditions. More importantly, combining molecular
catalysts with solid-state surfaces allows fabrication of artificial
photosynthetic systems which could potentially use water as a
renewable proton and electron donor for CO2 reduction [24].
Grafting molecular catalysts on surfaces has allowed us to
follow CO2 binding/activation and related photochemical events
on transition metal centers using in situ spectroscopic tech-
niques including diffuse reflectance infrared Fourier transform
spectroscopy (DRIFTS). In our previous work, hybrid Re(I) photo-
catalysts were prepared by covalently attaching Re(bpy)(CO)3Cl
on different surfaces [17,25,26]. The covalent attachment was
achieved by derivatizing the bpy ligand at the 4,4ꢀ-positions with
amide groups and further coupling with surfaces. Derivatizing the
bpy ligand with either CONH group or NHCO group at the 4,4ꢀ-
positions resulted in significant changes in photochemical and
catalytic properties of the hybrid Re(I) photocatalysts in CO2 reduc-
tion [25].
Building upon previous work, we synthesized new hybrid
Re(I) photocatalysts by derivatizing the bpy ligand at differ-
ent positions and functionalizing a mesoporous silica surface
with a dipodal silane coupling agent. A bpy ligand derivatized
at the 5,5ꢀ-positions, 2,2ꢀ-bipyridine-5,5ꢀ-dicarboxylic acid, has
∗
Corresponding author.
E-mail address: gonghu.li@unh.edu (G. Li).
http://dx.doi.org/10.1016/j.molcata.2015.10.034
1381-1169/© 2015 Elsevier B.V. All rights reserved.

T.G. Fenton et al. / Journal of Molecular Catalysis A: Chemical 411 (2016) 272–278
273
been extensively employed as a building block for metal-organic
2.3. Synthesis of homogenous Re(I) complexes (4 and 5)
frameworks [27,28] and supramolecular complexes [29]. In this
study, we investigate how derivatizing the bpy ligand at the
5,5ꢀ-positions affects photochemical properties of hybrid Re(I)
photocatalysts using spectroscopic techniques including diffuse
reflectance UV–visible, in situ DRIFTS and electron paramagnetic
resonance (EPR) spectroscopies.
One of our research goals is to achieve cooperative CO2 reduc-
tion via binuclear pathways on hybrid photocatalysts, in which
two metal catalytic centers work together to reduce CO2 to CO
and/or formic acid. Prior studies by several research groups have
clearly demonstrated the involvement of binuclear Re(I) species
in photochemical and electrochemical CO2 reduction [30–34]. We
hypothesize that the use of a dipodal silane coupling agent and
a bpy ligand derivatized at the 5,5ꢀ-positions will enable effective
assembly of molecular Re(I) catalysts on surfaces with close prox-
imity and optimal geometry, respectively, desired for the binuclear
chemistry to occur. In this study, we report spectroscopic evidence
for the possible formation of a binuclear Re(I) complex on a hybrid
photocatalyst under light irradiation.
Homogenous Re(I) complexes were synthesized according to
previous work by Wang et al. [36]. The synthesis of 4 started with
1 mmol of 2,2ꢀ-bipyridine-4,4ꢀ-dicarboxylic acid dispersed in 40 ml
of toluene, to which 1 mmol of pentacarbonylchlororhenium(I) was
added. The resulting mixture was refluxed for 6 h under nitrogen,
and was then stored at 0 ◦C for 1 h. The mixture was purified by
filtration and was concentrated under reduced pressure to give 4.
The synthesis of 5 was achieved via the same method by mixing
1 mmol 2,2ꢀ-bipyridine-5,5ꢀ-dicarboxylic acid in 40 ml of m ethanol
with 1 mmol of pentacarbonylchlororhenium(I).
2.4. Materials characterization
Optical spectra of synthesized samples were obtained using a
Cary 50 Bio spectrophotometer outfitted with both a transmission
cell (for homogeneous solutions) and a Barrelino diffuse reflectance
probe (for powder samples). DRIFTS spectra were collected on a
Nicolet 6700 FTIR spectrometer equipped with a Harrick Pray-
ing Mantis diffuse reflectance accessory. Elemental analysis of the
synthesized hybrid photocatalysts was performed via acid diges-
tion utilizing a Varian Vista AX inductively coupled plasma atomic
emission spectrometer. EPR spectra were collected on an X-band
(9.5 GHz) Bruker ELEXSYS E-500 cw-EPR/ENDOR spectrometer at
room temperature.
2. Experimental
2.1. Materials
Hydrochloric acid, 2,2ꢀ-bipyridine-4,4ꢀ-dicarboxylic acid (98%),
thionyl chloride (99.5%), dichloromethane (DCM), toluene, diethyl
ether, dimethylformamide (DMF, >99.5%), triethanolamine (TEOA,
99+%), pentacarbonylchlororhenium(I) (98%), tris(2,2ꢀ-bipyridine)
dichlororuthenium(II) hexahydrate (99.95% trace metals basis,
denoted “Ru(bpy)32+”) were obtained from Sigma–Aldrich. 2,2ꢀ-
Bipyridine-5,5ꢀ-dicarboxylic acid (98%) was obtained from TCI
2.5. Photocatalytic testing
In photocatalytic testing, a certain amount of 4-D-SiO2 or 5-D-
SiO2 was dispersed in an air-tight pyrex test tube containing 4 ml
of DMF and TEOA (volume ratio 3:1). The loading of Re(I) catalysts
was controlled to be 0.8 ␮mol in the reaction solution by adjust-
ing the mass of the hybrid photocatalysts. In some experiments,
3 mg of Ru(bpy)32+ was added as a photosensitizer to enhance light
harvesting in the visible region. The reaction solution was purged
with CO2 (Airgas, 99.999%) for 30 min prior to photocatalysis. A Xe
arc lamp equipped with a water filter and an AM 1.5 optical filter
was used as the light source. For all testing, the light intensity was
fixed at 30 mW/cm2. While the reaction solution was constantly
stirred under light irradiation, the head space was sampled using
a gas tight syringe for product analysis using an Agilent 7820 GC
equipped with a TCD detector and a 60/80 Carboxen 1000 column.
America.
N,N-bis[3-(trimethoxysilyl)propyl]ethylenediamine
(EDTMS, 95%) was obtained from Gelest. All reagents were used
without further purification.
2.2. Synthesis of hybrid Re(I) photocatalysts
Mesoporous silica (SiO2) was synthesized according to pre-
vious work by Zhao et al. [35]. Two diimine-tricarbonyl Re(I)
complexes (4 and 5, Fig. 1) and a dipodal silane coupling agent
(EDTMS, denoted “D”, Fig. 1) were utilized to synthesize hybrid
Re(I) photocatalysts by grafting the Re(I) complexes on mesoporous
SiO2 surfaces. The Re(I) complexes feature two derivatized bpy
ligands, 2,2ꢀ-bipyridine-4,4ꢀ-dicarboxylic acid and 2,2ꢀ-bipyridine-
3. Results and discussion
5,5ꢀ-dicarboxylic acid. In
a typical synthesis of hybrid Re(I)
3.1. Characterization of hybrid Re(I) photocatalysts
photocatalysts, 100 mg of mesoporous SiO2 was dried at 100 ◦C for
2 h and dispersed in 50 ml of dry toluene, to which 75 ␮l EDTMS
were added under constant stirring and refluxed for 48 h under
nitrogen. The surface functionalized SiO2 was washed with toluene,
diethyl ether, and DCM before collected by centrifugation. In a
separate solution, 25 mg 2,2ꢀ-bipyridine-4,4ꢀ-dicarboxylic acid or
2,2ꢀ-bipyridine-5,5ꢀ-dicarboxylic acid were dissolved in 10 ml of
thionyl chloride. This solution was then refluxed for 12 h and dried
under vacuum. The resulting solid was dispersed in 10 ml DCM
and was added drop wise under nitrogen to the functionalized
SiO2 which was dispersed in 30 ml of DCM. The combined solu-
tion was then refluxed for 12 h and washed with toluene, diethyl
ether, and DCM. The resulting product was dispersed in 40 ml of dry
toluene, to which 50 mg of pentacarbonylchlororhenium(I) were
added. This final solution was refluxed for 12 h under nitrogen and
again washed with toluene, diethyl ether, and DCM. The result-
ing hybrid photocatalyst, denoted as “4-D-SiO2” or “5-D-SiO2”, was
then collected as a yellow powder.
In this study, a dipodal silane coupling agent (D, Fig. 1) was uti-
lized to graft the diimine-tricarbonyl Re(I) catalyst on SiO2 surfaces
following a procedure similar to that in our previous studies, where
amide bonds were formed on surfaces to serve as covalent linkages
[17,25,26]. Each dipodal molecule has two NH groups available
for further functionalization with bpy ligands. Elemental analysis
of the synthesized hybrid photocatalysts shows that the amounts
of rhenium are 2.8 ␮mol and 0.7 ␮mol per 10 mg of 5-D-SiO2 and
4-D-SiO2, respectively. It is unclear why exactly the loadings are
different for these two samples. One contributing factor is the
different orientation of the two derivatized bpy ligands upon sur-
face immobilization. In 5-D-SiO2, the derivatized bpy groups are
likely perpendicular to the SiO2 surface upon the formation of
amide bonds in which one 2,2ꢀ-bipyridine-5,5ꢀ-dicarboxylic acid
only reacted with one surface NH group. In contrast, each 2,2ꢀ-
bipyridine-4,4ꢀ-dicarboxylic acid moiety could form amide bonds
with two adjacent surface NH groups in 4-D-SiO2, and the

274
T.G. Fenton et al. / Journal of Molecular Catalysis A: Chemical 411 (2016) 272–278
Fig. 1. Molecular structures of two diimine-tricarbonyl Re(I) complexes, 4 and 5, and the dipodal silane coupling agent.
Fig. 3. DRIFTS spectra of (a) 4-D -SiO2 and (b) 5-D-SiO2 mixed with KBr.
Fig. 2. Diffuse reflectance UV–visible spectra of (a) 4-D-SiO2 and (b) 5-D-SiO2 in
powder form. Barium sulfate was used as the background.
The synthesized hybrid photocatalysts were tested in CO2
reduction in the presence of triethanolamine (TEOA) as a sacrificial
electron donor under simulated solar irradiation. Under the exper-
imental conditions employed this study, CO was observed as the
major product in CO2 reduction using the hybrid photocatalysts;
no significant formation of H2 or formic acid was observed. In the
derivatized bpy groups in this sample are likely in parallel to the
SiO2 surface [25].
In previous studies, we have utilized spectroscopic techniques
to thoroughly characterize molecular Re(I) catalysts grafted on
silica surfaces to ensure structural integrity of the Re(I) centers
[17,25,26]. In this work, the presence of diimine-tricarbonyl Re(I)
units in the hybrid photocatalysts was monitored using a diffuse
reflectance UV–visible probe and DRIFTS. Both 5-D-SiO2 and 4-D-
SiO2 exhibit a strong absorption band around 400 nm (Fig. 2), which
is associated with the metal-to-ligand charge transfer (MLCT) tran-
sition of the diimine-tricarbonyl Re(I) units. In comparison with the
MLCT band of Re(bpy)(CO)3Cl at 370 nm,this represents a 30 nm
shift in the MLCT band as a result of derivatizing the bpy ligand
with electron-withdrawing amide groups [25]. The results shown
in Fig. 2 indicate that derivatization at the 4,4ꢀ- and 5,5ꢀ-positions
with COOH groups had the same effect on the MLCT band of the
Re(I) compound.
2+
presence of Ru(bpy)3 as a photosensitizer, turnover numbers for
CO production of 12.7 and 8.8 were obtained for 4-D-SiO2 and 5-D-
SiO2, respectively, after photocatalysis for 4 h (Fig. 4). In the absence
of Ru(bpy)32+, the amount of CO produced was negligible for both
4-D-SiO2 and 5-D-SiO2. These numbers are relatively lower than
those obtained using similar hybrid Re(I) photocatalysts reported
in our previous studies [25], likely due to the low light intensity
employed in this study. We are also mindful that molecular diffu-
sion of Ru(bpy)32+ in the mesopores could be significant in 4-D-SiO2
and 5-D-SiO2 with different catalyst loadings. However, the same
comparison in activity was observed using similar hybrid photocat-
alysts prepared by grafting 4 and 5 on a non-porous silica, in which
molecular diffusion does not significantly affect photocatalysis.
Fig. 3 shows the DRIFTS spectra of 4-D-SiO2 and 5-D-SiO2 mixed
with KBr. Two intense Re(I)-carbonyl bands at 2035 cm−1 and
1930 cm−1 are seen in both spectra, together with other absorp-
tions (not labeled). These absorption features are similar to those
in the DRIFTS spectra of other hybrid Re(I) photocatalysts reported
in our previous studies [17,25,26]. Therefore, diimine-tricarbonyl
Re(I) units are present in both 4-D-SiO2 and 5-D-SiO2, mostly in the
form of covalently attached molecular catalysts. Since both hybrid
photocatalysts were thoroughly washed with solvents during syn-
thesis, the presence of physically adsorbed Re(I) compound is very
unlikely.
3.2. Spectroscopic investigation of the color change of 5-D-SiO2
upon light irradiation
An interesting observation on the synthesized hybrid photo-
catalysts is the color change of 5-D-SiO2 during photocatalytic
2+
CO2 reduction in the absence of the Ru(bpy)3 photosensitizer.
Upon light irradiation, the color of a reaction solution containing
5-D-SiO2 quickly changed from light yellow to green. At room tem-
perature, the green color gradually decayed in a few hours after
the light was turned off, and re-appeared when light irradiation
resumed. Control experiments indicated that the photocatalyst,

T.G. Fenton et al. / Journal of Molecular Catalysis A: Chemical 411 (2016) 272–278
275
Fig. 5. Optical spectra of a solution containing 5-D-SiO2, TEOA and CO2 before (a,
black line) and after (b, blue line) light irradiation for 5 min. Additional spectra were
collected every 15 min after the light was turned off in order to obtain difference
spectra (inset) which demonstrate the decay of the blue color in dark. (For interpre-
tation of the references to colour in this figure legend, the reader is referred to the
web version of this article.).
Fig. 4. CO production in photocatalytic CO2 reduction using 4-D-SiO2 (a, c) and 5-
D-SiO2 (b, d) under simulated solar irradiation. In testing (a) and (b), Ru(bpy)32+ was
added as an additional photosensitizer. Light intensity was fixed at 30 mW/cm2 for
all photocatalytic testing.
TEOA (or other amine-based electron donor), CO2 and light irradi-
ation were all necessary for the color change to occur, and that the
color change was associated with the Re(I) complex on 5-D-SiO2.
In addition, the color change was immediately reversed as soon as
the reaction solution was exposed to air or purged with Ar, even in
the presence of continuous light irradiation. Such color change was
not observed for 4-D-SiO2.
Close examination of spectroscopic results indicated that the
green color of 5-D-SiO2 upon light irradiation is associated with two
absorption bands around 450 nm and 620 nm. To study this phe-
nomenon, color change of the hybrid photocatalyst was monitored
using a UV–visible spectrophotometer in the transmission mode.
Specifically, a small amount of the reaction solution containing
well-dispersed 5-D-SiO2, TEOA and CO2 in a sealed quartz cuvette
wasexposedto simulatedsolarirradiationfor 5 min. Opticalspectra
of the solution was collected before and after the light irradiation
(Fig. 5). Collection of spectra continued after the light irradiation
was removed. Difference spectra were obtained by subtracting the
initial spectrum (before light irradiation) from subsequent ones
(Fig. 5, inset). From the spectra shown in Fig. 5, it is clearly seen that
the two absorption bands around 450 nm and 620 nm appeared
upon light irradiation and that the two bands gradually decayed
when the irradiated sample was placed in dark (Fig. 5, inset). In con-
trast, formation of neither band was observed for 4-D-SiO2 under
the same experimental conditions (Fig. 6).
Fig. 6. Optical spectra of a solution containing 4-D-SiO2, TEOA and CO2 before (a,
black line) and after (b, red line) light irradiation for 5 min. Similar to Fig. 5, difference
spectra (inset) were obtained by subtracting the spectrum before light irradiation.
(For interpretation of the references to colour in this figure legend, the reader is
referred to the web version of this article.).
formation of one-electron reduced species upon light irradiation
[37].
In order to probe possible reasons for the observed color change,
homogenous samples of derivatized Re(I) compounds, 4 and 5 (see
Fig. 1 for molecular structures), were synthesized and tested in
a similar manner. Specifically, the homogeneous samples in DMF
were each mixed with TEOA and saturated with CO2 prior to light
irradiation. Optical spectra of the solutions were collected before
and after irradiation with simulated sunlight (Fig. 7). The color of
the solution containing 5 immediately turned green upon light
irradiation. Two broad peaks around 450 nm (as a shoulder) and
620 nm appeared in the optical spectrum of the solution of 5 col-
lected after exposure to light, while the intensity of the MLCT band
around 400 nm decreased significantly (Fig. 7b). In contract, no
noticeable green color was observed in the solution of 4 upon light
irradiation, nor did the 450 nm and 620 nm peaks show up in its
optical spectrum collected after light irradiation (Fig. 7a). In addi-
tion, the MLCT band of 4 disappeared completely, likely due to the
The comparison shown in Figs. 5–7 clearly indicates that
derivatization of bpy at the 4,4ꢀ-positions with carboxylate moi-
eties has different effects on the photochemical properties of
diimine-tricarbonyl Re(I) complexes than that at the 5,5ꢀ-positions.
Upon photoexcitation, a reduced Re(I) species is formed follow-
ing the reductive quenching of the MLCT excited state of the
Re(I) complex by TEOA, dissociation of Cl− and subsequent bind-
ing of CO2 on the Re(I) center. Previous studies have confirmed
the formation of a CO2-bound Re(I) adduct, Re(L-bpy)(CO)3(OOC-
OCH2CH2NR2) where L-bpy represents a derivatized bpy ligand and
R is CH2CH2OH, in the presence of TEOA and CO2 [26,38]. The green
color observed for reduced 5 and 5-D-SiO2 is likely due to the elec-
tron density in the CO2-bound Re(I) adduct being more localized
on the 5,5ꢀ-derivatized bpy ligand, whereas in reduced 4 and 4-D-
SiO2 the electron density is more localized on the Re(I) center. This
variation in charge density distribution must originate from the

276
T.G. Fenton et al. / Journal of Molecular Catalysis A: Chemical 411 (2016) 272–278
Fig. 7. Optical spectra of (a) 4 and (b) 5 in the presence of TEOA and CO2 after light irradiation. Corresponding spectra before light irradiation are also included (black lines).
Catalyst concentration was 0.45 mM for both samples.
TEOA contributed to the stabilization of the one-electron reduced
Re(I) species.
In addition, the spectrum of 4-D-SiO2 (Fig. 8a) shows some res-
onance features originated from hyperfine interactions between
unpaired electrons with Re(I) nuclei [39], which is similar to the
observations in our previous EPR studies of a hybrid Re(I) photo-
catalyst [25]. Less significant hyperfine coupling features are seen
in the spectrum of 5-D-SiO2. This comparison supports our conclu-
sion that the electron density in reduced 5-D-SiO2 is more localized
on the bpy ligand while in reduced 4-D-SiO2 the electron density
is more localized on the Re(I) center.
In a study by Argazzi et al., Ru(II) polypyridyl complexes were
derivatized with COOH groups for use in dye-sensitized solar cells
[40]. The researchers found that moving the carboxylate moieties
from the 4,4ꢀ-positions to the 5,5ꢀ-positions lowered the ␲* energy
of the bpy ligand [40]. Subsequently, the Ru(II) photosensitizers
based on 2,2ꢀ-bipyridine-5,5ꢀ-dicarboxylic acid were less efficient
at converting photons into electrons because of more significant
nonradiative decay of the excited states of the Ru(II) sensitizers
[40]. In this present study, derivatizing the bpy ligand at different
positions resulted in difference in charge density distribution in the
Fig. 8. EPR spectra of (a) 4-D-SiO2 and (b) 5-D-SiO2 under light irradiation. The
powder samples used in EPR studies were mixed with TEOA and purged with CO2
prior to light irradiation. Spectra were collected at room temperature.
reduced hybrid photocatalysts. This might partially account for the
different photocatalytic activity of 4-D-SiO2 and 5-D-SiO2 shown
in Fig. 4.
different derivatization (4,4ꢀ- vs 5,5ꢀ-positions) on the bpy ligand
with electron-withdrawing COOH or CONH groups. It appears
that derivatizing the bpy ligand at the 5,5ꢀ-positions favored the
localization of electron density on the bpy ring more than the Re(I)
metal center.
3.3. In situ DRIFTS studies of hybrid Re(I) photocatalysts
Light-induced redox processes in the hybrid photocatalysts
were further probed with EPR spectroscopy at room temperature.
Specifically, 4-D-SiO2 or 5-D-SiO2 in powder form was mixed with
TEOA and purged with CO2 prior to collecting EPR spectra. Fig. 8
shows EPR spectra of 4-D-SiO2 and 5-D-SiO2 under continuous light
irradiation. A broad resonance centered at g = 2 is seen in both spec-
tra (Fig. 8). The intensity of this signal continued to increase upon
prolonged light irradiation. For both samples, this EPR signal was
not observed without light irradiation. These observations indicate
that paramagnetic, one-electron reduced Re(I) species formed in
both 4-D-SiO2 and 5-D-SiO2 photochemically in the presence of
TEOA and CO2. As mentioned earlier in the text, the presence of
both an electron donor (TEOA in this study) and CO2 is required for
the light-induced color change observed for 5 and 5-D-SiO2. The
detection of this paramagnetic species with room-temperature EPR
suggests that binding of CO2 to the Re(I) center in the presence of
The surface chemistry and photo-induced events of 4-D-SiO2
and 5-D-SiO2 were further investigated with in situ DRIFTS spec-
troscopy in the presence of TEOA and CO2. Six overlapping
Re(I)-carbonyl bands are clearly seen in the DRIFTS spectra of
both samples collected prior to light irradiation (Fig. 9), indicat-
ing the presence of two different surface Re(I) species. Based on
previous studies [26], the three carbonyl bands at 2021, 1919 and
1897 cm−1 are associated with the CO2-bound Re(I) adduct, Re(L-
bpy)(CO)3(OOC-OCH2CH2NR2). Another set of carbonyl bands at
1995, 1876 and 1851 cm−1 have been assigned to a Re(I) hydroxyl
species, Re(L-bpy)(CO)3(OH).
It should be noted that an infrared band around 2005 cm−1
is present in the spectrum of 5-D-SiO2 (Fig. 9b) but not in the
spectrum of 4-D-SiO2 (Fig. 9a). Although we lack sufficient data
for definitive assignment of this peak, it is likely associated with
the CO2-bound Re(I) adduct, Re(L-bpy)(CO)3(OOC-OCH2CH2NR2),

T.G. Fenton et al. / Journal of Molecular Catalysis A: Chemical 411 (2016) 272–278
277
Fig. 9. DRIFTS spectra of (a) 4-D-SiO2 and (b) 5-D-SiO2 in the presence of TEOA and
CO2. The spectra were collected prior to light irradiation at room temperature.
Fig. 11. Difference DRIFTS spectra of (a) 4-D-SiO2 and (b) 5-D-SiO2 in the presence of
TEOA and CO2. The difference spectra were obtained by subtracting spectra collected
at t = 5 min from corresponding spectra collected at t = 15, 30 and 60 min.
1876 cm−1 and 1689 cm−1 quickly appeared in the spectra of
both samples upon light irradiation (Fig. 10). The peaks at 2005
and 1876 cm−1 are likely associated with the reduced CO2-bound
Re(I) adduct, as discussed earlier. We attribute the absorption at
1689 cm−1 to the formation of a tertiary carbamate [42] as a product
in the reaction between TEOA and CO2, both of which are present
in large excess in the DRIFTS cell used in this study.
We notice that the negative peak around 2025 cm−1 in Fig. 10b
became less negative after t = 5 min, suggesting the formation of a
new Re(I)-carbonyl species on 5-D-SiO2 under continued light irra-
diation. Three additional absorptions at 1994, 1843 and 1720 cm−1
are present in the difference spectra of 5-D-SiO2 only and these
three peaks started to appear as light irradiation continued after
t = 5 min (Fig. 10b). The DRIFTS spectra were further analyzed using
another set of difference spectra which were obtained by sub-
tracting spectra collected after light irradiation for t = 5 min from
corresponding ones collected after light irradiation for t = 15, 30,
and 60 min. These new difference spectra shown in Fig. 11 are used
to describe changes on the hybrid photocatalysts after continued
light irradiation beyond t = 5 min.
The difference spectra shown in Fig. 11a are very similar to
those shown in Fig. 10a, indicating that no new photochemical
events occurred on 4-D-SiO2 after t = 5 min. In Fig. 11a, the negative
peaks observed around 2025, 1920 and 1900 cm−1 (not labeled)
are related to photochemical conversion of Re(L-bpy)(CO)3(OOC-
OCH2CH2NR2). However, the difference spectra shown in Fig. 11b
are different from those in Fig. 10b. Continued light irradiation
(after t = 5 min) resulted in (i) disappearance of the reduced CO2-
bound Re(I) adduct associated with the absorptions at 2005 and
1894 cm−1; and (ii) further formation of the Re(I) hydroxyl species
characterized by the bands at 1994 and 1843 cm−1 in 5-D-SiO2
(Fig. 11b). Significant formation of the Re(I) hydroxyl species
was not observed in 4-D-SiO2 under continued light irradiation
(Fig. 11a). Photochemical production of Re(L-bpy)(CO)3(OH) from
Re(L-bpy)(CO)3(OOC-OCH2CH2NR2) was previously observed on a
similar hybrid system and was suggested to be a possible deac-
tivation pathway for Re(I)-based photocatalysts [26]. Therefore,
formation of the Re(I) hydroxyl species in 5-D-SiO2 might account
for its lower photocatalytic activity than 4-D-SiO2 in CO2 reduction
(Fig. 4).
Fig. 10. Difference DRIFTS spectra of (a) 4-D-SiO2 and (b) 5-D-SiO2 in the presence
of TEOA and CO2. The difference spectra were obtained by subtracting spectra col-
lected prior to light irradiation (t = 0) from corresponding spectra collected after
light irradiation for different times (5, 15, 30 and 60 min).
in the reduced form since peaks at 2005, 1894 and 1878 cm−1
were observed in the infrared spectrum of a reduced Re(I) cata-
lyst [41]. Formation of this reduced Re(I) species in 5-D-SiO2 likely
occurred when the sample was exposed to ambient light conditions
while being loaded into the DRIFTS cell. The comparison shown
in Fig. 9 clearly indicates that certain features of 5-D-SiO2, par-
ticularly ligand derivatization at the 5,5ꢀ-positions, facilitated the
formation of this reduced Re(I) species under ambient light condi-
tions. Further formation of this Re(I) species upon light irradiation
was observed on both 5-D-SiO2 and 4-D-SiO2, as will be discussed
in the following paragraphs.In the DRIFTS cell, the hybrid photo-
catalysts were irradiated with simulated solar light under a CO2
atmosphere and spectra were collected after light irradiation for
various times. Difference spectra were obtained by subtracting
spectra collected prior to light irradiation (i.e., the spectra in Fig. 9)
from corresponding ones collected after light irradiation. Fig. 10
shows the difference spectra of 4-D-SiO2 and 5-D-SiO2, which pro-
vide important information regarding photo-induced events on the
two samples. In the difference spectra shown in Fig. 10, negative
peaks are present around 2025, 1920 and 1900 cm−1 (not labeled)
indicating the photochemical conversion of Re(L-bpy)(CO)3(OOC-
OCH2CH2NR2) on both 4-D-SiO2 and 5-D-SiO2. This CO2-bound
Re(I) adduct was suggested to be the real catalyst in many CO2-
reduction systems using diimine-tricarbonyl Re(I) photocatalysts
[38]. The peak at 2005 cm−1 and two broad absorptions around
The appearance of two additional infrared bands around 2025
and 1720 cm−1 was also observed in the spectra shown in Fig. 11b.
The band at 2025 cm−1 should be more intense than it appears

278
T.G. Fenton et al. / Journal of Molecular Catalysis A: Chemical 411 (2016) 272–278
in Fig. 11b due to concurrent photochemical conversion of Re(L-
bpy)(CO)3(OOC-OCH2CH2NR2) which resulted in a negative peak
in the same position (see Fig. 10b). We attribute this Re(I)-carbonyl
band at 2025 cm−1 and the other absorption at 1720 cm−1 (␯C O in
the carbonate bridge) to possible formation of a carbonate-bridged
binuclear Re(I) complex, (CO)3(L-bpy)Re-OCO2-Re(L-bpy)(CO)3, in
5-D-SiO2. The absorption at 1720 cm−1 would be associated with
the C O stretching mode of the carbonate bridge. Two other Re(I)-
carbonyl bands at 1893 and 1852 cm−1 were expected for this
binuclear complex [43], but were not clearly seen in Fig. 11b due
to the overlapping and subtraction of multiple carbonyl bands in
this region. According to the theoretical studies by Agarwal et al.,
this binuclear, carbonate-bridged Re(I) complex is a byproduct in
CO2 reduction following a binuclear pathway which produces one
CO molecule from two Re(I) centers and two CO2 molecules [33].
In 5-D-SiO2, the close proximity of Re(I) centers on the dipodal
silane coupling agent likely promoted this binuclear pathway. Fur-
ther spectroscopic studies using isotope labeling are underway to
fully explore the binuclear chemistry of hybrid CO2-reduction pho-
tocatalysts including 5-D-SiO2.
[2] C. van der Giesen, R. Kleijn, G.J. Kramer, Environ. Sci. Technol. 48 (2014)
7111–7121.
[3] L. Liu, Y. Li, Aerosol Air Qual. Res. 14 (2014) 453–469.
[4] A.J. Morris, G.J. Meyer, E. Fujita, Acc. Chem. Res. 42 (2009) 1983–1994.
[5] H. Takeda, O. Ishitani, Coord. Chem. Rev. 254 (2010) 346–354.
[6] B. Kumar, M. Llorente, J. Froehlich, T. Dang, A. Sathrum, C.P. Kubiak, Annu.
Rev. Phys. Chem. 63 (2012) 541–569.
[7] J. Hawecker, J.-M. Lehn, R. Ziessel, J. Chem. Soc. Chem. Commun. (1983)
536–538.
[8] T.R. O’Toole, L.D. Margerum, T.D. Westmoreland, W.J. Vining, R.W. Murray, T.J.
Meyer, J. Chem. Soc. Chem. Commun. (1985) 1416–1417.
[9] H. Takeda, K. Koike, H. Inoue, O. Ishitani, J. Am. Chem. Soc. 130 (2008)
2023–2031.
[10] E. Portenkirchner, K. Oppelt, C. Ulbricht, D.A.M. Egbe, H. Neugebauer, G. Knör,
N.S. Sariciftci, J. Organomet. Chem. 716 (2012) 19–25.
[11] E. Portenkirchner, S. Schlager, D. Apaydin, et al., Electrocatalysis 6 (2015)
185–197.
[12] P.D. Tran, L.H. Wong, J. Barber, J.S.C. Loo, Energy Environ. Sci. 5 (2012)
5902–5918.
[13] K.D. Dubois, G. Li, in: S. Suib (Ed.), New and Future Developments in Catalysis:
Solar Photocatalysis, Elsevier B.V., 2013, pp. 219–241.
[14] F. Wen, C. Li, Acc. Chem. Res. 46 (2013) 2355–2364.
[15] T. Zhang, W. Lin, Chem. Soc. Rev. 43 (2014) 5982–5993.
[16] H. Takeda, M. Ohashi, T. Tani, O. Ishitani, S. Inagaki, Inorg. Chem. 49 (2010)
4554–4559.
[17] K.D. Dubois, H. He, C. Liu, A.S. Vorushilov, G. Li, J. Mol. Catal. A: Chem. 363–364
(2012) 208–213.
[18] T. Jin, C. Liu, G. Li, Chem. Commun. 50 (2014) 6221–6224.
[19] C. Wang, Z. Xie, K.E. deKrafft, W. Lin, J. Am. Chem. Soc. 133 (2011)
13445–13454.
4. Conclusions
[20] M.L. Macnaughtan, H.S. Soo, H. Frei, J. Phys. Chem. C 118 (2014) 7874–7885.
[21] Y. Ueda, H. Takeda, T. Yui, K. Koike, Y. Goto, S. Inagaki, O. Ishitani,
ChemSusChem 8 (2015) 439–442.
[22] S. Sato, T. Arai, T. Morikawa, K. Uemura, T.M. Suzuki, H. Tanaka, T. Kajino, J.
Am. Chem. Soc. 133 (2011) 15240–15243.
[23] K. Maeda, K. Sekizawa, O. Ishitani, Chem. Commun. 49 (2013) 10127–10129.
[24] S. Sato, T. Arai, T. Morikawa, Inorg. Chem. 54 (2015) 5105–5113.
[25] C. Liu, K.D. Dubois, M.E. Louis, A.S. Vorushilov, G. Li, ACS Catal. 3 (2013)
655–662.
[26] H. He, C. Liu, M.E. Louis, G. Li, J. Mol. Catal. A: Chem. 395 (2014) 145–150.
[27] C. Wang, Z. Xie, K.E. deKrafft, W. Lin, J. Am. Chem. Soc. 133 (2011)
13445–13454.
[28] T.L. Easun, J. Jia, J.A. Calladine, D.L. Blackmore, C.S. Stapleton, K.Q. Vuong, N.R.
Champness, M.W. George, Inorg. Chem. 53 (2014) 2606–2612.
[29] K.F. Cheng, C.M. Drain, K. Grohmann, Inorg. Chem. 42 (2003) 2075–2083.
[30] Y. Hayashi, S. Kita, B.S. Brunschwig, E. Fujita, J. Am. Chem. Soc. 125 (2003)
11976–11987.
[31] Z.-Y. Bian, K. Sumi, M. Furue, S. Sato, K. Koike, O. Ishitani, Inorg. Chem. 47
(2008) 10801–10803.
[32] Z.-Y. Bian, K. Sumi, M. Furue, S. Sato, K. Koike, O. Ishitani, Dalton Trans. (2009)
983–993.
In summary, two hybrid CO2-reduction photocatalysts based
on Re(bpy)(CO)3Cl have been successfully synthesized and char-
acterized with UV–visible, in situ DRIFTS and EPR spectroscopies.
Synthesis of the hybrid photocatalysts involved derivatization of
the bpy ligand with amide groups and grafting on mesoporous sil-
ica via a dipodal silane coupling agent. In this study, derivatization
of the bpy ligand at the 5,5ꢀ-positions resulted in an interesting
color change of the hybrid photocatalyst upon light-induced reduc-
tion, presumably due to localization of electron density on the bpy
ligand. In comparison, electron density was found to be localized
more on the Re(I) center in the reduced hybrid photocatalyst fea-
turing ligand derivatization at the 4,4ꢀ-positions. Such differences
in electron density distribution, as well as the difference in prox-
imity and geometry of the Re(I) catalysts on surfaces, are likely
correlated with photochemical events relevant to CO2-reduction
catalysis observed in the hybrid photocatalysts. In addition, our
results using in situ DRIFTS spectroscopy demonstrated possible
formation of a carbonate-bridged binuclear Re(I) species on the
hybrid photocatalyst prepared by derivatizing the bpy ligand at
the 5,5ꢀ-positions under light irradiation and in the presence of an
electron donor and CO2.
[33] J. Agarwal, E. Fujita, H.F. Schaefer, J.T. Muckerman, J. Am. Chem. Soc. 134
(2012) 5180–5186.
[34] C.W. Machan, S.A. Chabolla, J. Yin, M.K. Gilson, F.A. Tezcan, C.P. Kubiak, J. Am.
Chem. Soc. 136 (2014) 14598–14607.
[35] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D.
Stucky, Science 279 (1998) 548–552.
[36] C. Wang, Z. Xie, K.E. deKrafft, W. Lin, J. Am. Chem. Soc. 133 (2011)
13445–13454.
[37] Y. Hayashi, S. Kita, B.S. Brunschwig, E. Fujita, J. Am. Chem. Soc. 125 (2003)
11976–11987.
Acknowledgements
[38] T. Morimoto, T. Nakajima, S. Sawa, R. Nakanishi, D. Imori, O. Ishitani, J. Am.
Chem. Soc. 135 (2013) 16825–16828.
[39] T. Scheiring, A. Klein, W. Kaim, J. Chem. Soc., Perkin Trans. 2 (1997)
2569–2571.
[40] R. Argazzi, C.A. Bignozzi, T.A. Heimer, F.N. Castellano, G.J. Meyer, Inorg. Chem.
33 (1994) 5741–5749.
This material is based upon work supported by the U.S. National
Science Foundation under a CAREER Award (#CHE-1352437). We
are grateful to Chao Liu and Dr. Roy Planalp for helpful discussions
and assistance in various aspects of experiments.
[41] J.M. Smieja, C.P. Kubiak, Inorg. Chem. 49 (2010) 9283–9289.
[42] S. Pinchas, D. Ben-Ishai, J. Am. Chem. Soc. 79 (1957) 4099–4104.
[43] D.H. Gibson, X. Yin, H. He, M.S. Mashuta, Organomet 22 (2003) 337–346.
References
[1] A.M. Appel, J.E. Bercaw, A.B. Bocarsly, et al., Chem. Rev. 113 (2013) 6621–6658.