Photon-driven reduction of Zn2+ to Zn metal

Article abstract of DOI:10.1021/ic302629q

Easily oxidized metals are of interest as a means of storing solar energy in the form of fuels. While their efficient metal/air batteries make them attractive solar fuel candidates, the photoreduction of the corresponding metal ions remains difficult. Accordingly, this work describes the photon driven reduction of Zn²⁺ by an iridium(III) photosensitizer (PS) and catalyst. [Ir(ppy)₂(dtbbpy)](PF₆) (ppy = 2-phenylpyridine, dtbbpy = 4,4′-di-tert-butyl-2,2′-bipyridine) was found to be the most robust photocatalyst, and the use of ZnCl₂ as the Zn ²⁺ starting material and acetonitrile as the solvent afforded the highest yield of Zn metal product. Under these conditions, a maximum of 430 catalyst turnover numbers were achieved. Cyclic voltammetry of ZnCl₂ in different solvents and of different zinc salts in acetonitrile (MeCN) demonstrated the roles of MeCN and Cl^- in the photoreduction mechanism. Kinetics measurements revealed a first order dependence of the initial rate on both [Ir(ppy)₂(dtbbpy)](PF₆) and ZnCl ₂. A first order decay of the reaction rate was also observed.

Full text of DOI:10.1021/ic302629q

Article
pubs.acs.org/IC
2+
Photon-Driven Reduction of Zn to Zn Metal
Anthony C. Brooks, Katherine Basore, and Stefan Bernhard*
Department of Chemistry, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213-3890, United States
ABSTRACT: Easily oxidized metals are of interest as a means of
storing solar energy in the form of fuels. While their efficient metal/
air batteries make them attractive solar fuel candidates, the
photoreduction of the corresponding metal ions remains difficult.
Accordingly, this work describes the photon driven reduction of
2
+
Zn by an iridium(III) photosensitizer (PS) and catalyst.
Ir(ppy) (dtbbpy)](PF ) (ppy = 2-phenylpyridine, dtbbpy = 4,4′-
[
2
6
di-tert-butyl-2,2′-bipyridine) was found to be the most robust
2
+
photocatalyst, and the use of ZnCl as the Zn starting material and
2
acetonitrile as the solvent afforded the highest yield of Zn metal
product. Under these conditions, a maximum of 430 catalyst turnover numbers were achieved. Cyclic voltammetry of ZnCl in
2
different solvents and of different zinc salts in acetonitrile (MeCN) demonstrated the roles of MeCN and Cl− in the
photoreduction mechanism. Kinetics measurements revealed a first order dependence of the initial rate on both
[
Ir(ppy) (dtbbpy)](PF ) and ZnCl . A first order decay of the reaction rate was also observed.
2
6
2
INTRODUCTION
battery employing the relevant electrode reactions shown in
Scheme 1.
■
Human energy consumption has been rising for decades. The
resulting economic conditions and the environmental con-
sequences of this growth will provide an opportunity for
a
Scheme 1. Redox Reactions Relevant to the Zn/Air Battery
1,2
renewable energy applications to thrive. Many direct and
3
indirect solar power conversion schemes are currently the
subject of intense research, since they promise abundant energy
that could potentially supplement or replace coal or petroleum
4
on a massive scale. Consequently, storing photon energy as
chemical potential in solar fuels is of particular interest for
replacing hydrocarbons for mobile applications. This approach
alleviates the complications created by the seasonal and diurnal
fluctuations commonly observed in solar energy conversion
systems. Unlike photovoltaic energy, solar fuels are independ-
ent of the electrical grid, allowing them to be stored and
a
A strongly basic electrolyte is commonly used in these mechanically
rechargeable fuel cells.
Neglecting the mass of the O oxidant, zinc has a theoretical
2
transported easily as a mobile fuel. Specifically, H is often
8
2
specific energy density of 1.35 kWh/kg. Without the need for
compression, only the battery is used for storage, which lowers
the practical specific energy density just slightly to 1.084 kWh/
kg. The lightweight and remarkable 98% coulombic discharge
efficiencies of the Zn/air battery exemplify the great potential
of zinc as a transportable, abundant solar fuel.
5
,6
identified as a promising candidate, but storage requirements
partially offset the high gravimetric energy densities desired for
transportation applications. To illustrate, gaseous H has a 39.4
9
2
kWh/kg theoretical specific energy density, but a safe, 700 bar
10
tank system for mobile H use has a theoretical specific energy
2
2
+
density of 1.89 kWh/kg, which further drops because of losses
To make use of zinc as a renewable solar fuel, Zn oxidized
7
0
from compression and fuel cell mass.
in the battery must be reduced back to Zn by light. Also, a
Metals are an often-overlooked alternative to energy-dense
liquid and gaseous solar fuels. Despite this oversight, zinc and
aluminum are well suited for this application because they can
utilize established Zn/air and Al/air batteries which are
efficient, lightweight, and consume only these reactive metals
and air. The very negative reduction potential of Al renders it
an extremely efficient candidate for energy storage, but its
reactivity will greatly complicate the photocatalytic production
of this metal. Consequently, zinc is a more appealing prospect
for a fuel cycle involving a mechanically rechargeable metal/air
complementary reaction that oxidizes hydroxide ions to oxygen
has to be driven by photons concomitantly. A robust and
efficient photosensitizer (PS) responsible for the initial
absorption of a photon is critical for such a scheme.
Photoinduced electron transfer reactions will subsequently
drive OH oxidation and Zn reduction processes and thereby
reverse the redox chemistry responsible for driving the Zn/air
−
2+
Received: November 30, 2012
Published: May 6, 2013
©
2013 American Chemical Society
5794
dx.doi.org/10.1021/ic302629q | Inorg. Chem. 2013, 52, 5794−5800

Inorganic Chemistry
Article
cell. Iridium complexes stand out as PSs because of their
remarkable photophysical properties such as long lifetimes and
Cyclic voltammetry was performed utilizing a CH instruments
00C electrochemical analyzer potentiostat. The setup employed a
6
11
glassy carbon working, platinum counter, and silver wire pseudorefer-
ence electrode, against which a ferrocene internal standard was
photochemical robustness. Furthermore, the highest occupied
molecular orbital (HOMO) and the lowest unoccupied
molecular orbital (LUMO) of cyclometalated iridium com-
plexes can be tuned independently by modification of separate
ligands, which allows the thermodynamics of electron transfer
28
referenced at 0.380 V vs SCE.
Photoreduction samples were prepared in 40 mL EPA vials with
open top caps containing a fluoropolymer resin/silicone septa
(ICHEM SB36-0040). A typical photoreaction contained 9 mL of
acetonitrile, 1 mL of TEA, 1.1 μmol of [Ir(ppy) (dtbbpy)]PF , and
1
2,13
reactions to be fine-tuned.
Cyclometalated Ir(III) com-
2
6
14
plexes are used in organic light-emitting diodes (OLEDs) and
as photocatalysts in many redox reactions including organic
radical polymerizations, cyclizations, aldehyde fluoromethy-
lations, the water gas shift reaction, and more prominently,
the reduction of H O and CO for solar fuels.
420 mg of ZnCl .
2
The samples were illuminated under argon in a custom designed,
multiwell photoreactor: A 5 m long 24 W 460 nm LED strip with 300
diodes (Solid Apollo SA-LS-BL-3528-300-24 V) was mounted inside
an 8 in. (203 mm) galvanized steel tube in a uniform, spiral pattern.
The 60 mm high illumination zone was centered on the samples,
which were magnetically stirred. Samples were held in place by a
circular sample holder that was placed inside the steel tube. Sheaths
were built by cutting transparent 1−1/4″ (31.8 mm) polycarbonate
tube with a 1−1/8″ inner diameter into twelve 42 mm long pieces.
Twelve of these short pieces were then mounted around a clear 3-1/2″
acrylic tube at a height of 40 mm. The septum covered EPA vials were
used as illumination vessels. The sample holder was centered on a
magnetic stirrer, and the distance between all the vials and the LED
ribbon was uniformly kept at 28 mm. Before illumination, sample vials
were degassed by cycling 6 times between vacuum and Ar atmosphere.
Figure 1 depicts the photoreactor.
15
16
1
7
18
19,20
In H2
2
2
evolving, water reduction reactions, turnover numbers (TONs,
moles product/moles PS) of the Ir(III) PS have reached
1
0,000, which was achieved by the judicious design of the
21
ligand sphere of these highly luminescent metal complexes.
The electrochemical and photochemical robustness of these
phosphorescent materials is extraordinary and greatly facilitates
the exploration of new photocatalytic systems relevant to the
renewable energy field.
This work describes the visible-photon driven reduction of
Zn2 to Zn (E = −0.76 V vs NHE) by an iridium(III) PS
+
0
red
and catalyst system. To our knowledge, this is the first
published case of the photocatalyzed reduction of a low redox
potential metal ion by a transition metal complex. In being so,
this initial work demonstrates the feasibility of Zn2+ reduction
by a molecular photocatalyst for eventual use as a solar fuel.
Contrasting, [Ru(bpy) ] has been used to photoreduce Ag+
2
+
3
2
2
to Ag metal, but silver has a high work function, its cation is
0
easy to reduce, and Ag lacks the energy density necessary of a
solar fuel. A dissimilar approach is the reduction of Zn2 to Zn
metal by solar thermal ZnO decomposition. At high temper-
atures achieved by concentrating sunlight, equilibrium shifts
+
23
and ZnO splits into its elemental constituents O and Zn.
2
Though ZnO splitting can occur at 1100 K, temperatures of
300 K or higher are generally required, and low metal yields
2
0
caused by the reoxidation of Zn are commonly observed
during cooling process.
In this paper, [Ir(ppy) (dtbbpy)]PF (ppy = 2-phenyl-
24
2
6
Figure 1. Custom designed photoreactor used in Zn2+ reducing
photoreactions.
pyridine, dtbbpy = 4,4′-di-tert-butyl-2,2′-bipyridine) is em-
2
+
ployed as a photocatalyst for the reduction of Zn .
Triethylamine (TEA) serves as a sacrificial electron donor for
the aforementioned half reaction, an approach that is similar to
work done by the hydrogen community while exploring water
After illumination, the samples were purged with argon for 10 min
to vacate any potential H2 from the headspace. Quantitative
conversion of reduced Zn0 to H2 was achieved by injecting
photoreaction mixtures with 2 mL of concentrated hydrochloric
2
5,26
splitting systems.
While the use of ZnO as the Zn(II)
source is preferential, ZnO is easily converted to ZnCl , so
2
more important is the maintenance of charge balance. Like H
2
2
+
29
evolution systems, Zn photoreduction ultimately requires a
complementary oxidation half reaction, such as water oxidation,
to maintain charge balance necessary for complete solar fuel
photosynthesis. The structure−activity relationships of diverse
Ir(III) PSs are explored, and rate studies revealed insights into
the mechanistic details of these photocatalytic reduction
reactions.
acid, similar to the technique described by Vogel.
The headspace of the acidified photoreduction vials was injected
into a residual gas analyzer to permit calculation of total H evolved by
2
1
3
the procedure used by Curtin et al. Catalytic TONs were calculated
using the 1:1 stoichiometric ratio of H to zinc product. In this case,
2
2+
0
because Zn is reduced to Zn and the Ir catalyst serves the dual
function of PS, electron TONs are simply twice that of catalytic
TONs. All TONs reported in this work are catalytic TONs.
X-ray diffraction was measured using a PANalytical X′Pert PRO
diffraction system. To prepare the completed photoreduction samples
for X-ray diffraction, the solution phase was extracted through the
septum of the capped photoreduction vial to preserve the argon
atmosphere in the headspace. MeCN was then added and extracted
similarly and the samples were purged with argon until dry. The zinc
product was then briefly exposed to air during transfer and diffraction
measurement.
EXPERIMENTAL SECTION
■
All iridium complexes, including [Ir(ppy) (dtbbpy)]PF , were
synthesized according to established procedures.
2
6
1
3,27
Acetonitrile was
purchased from Fischer, triethylamine was purchased from Acros, and
ZnCl as well as the other zinc salts were obtained from Sigma. All
2
chemicals were used as received.
5
795
dx.doi.org/10.1021/ic302629q | Inorg. Chem. 2013, 52, 5794−5800

Inorganic Chemistry
Article
Bernhard group, which demonstrated a dependence of e−
TONs on the lifetime of the Ir PS. In the future, a better
understanding of the photocatalytic mechanism should yield
clearer insights into structure property relationships observed in
25
RESULTS AND DISCUSSION
■
Mixtures of ZnCl in acetonitrile were observed to yield
2
reduced Zn metal after illumination under visible light in the
presence of an iridium(III) PS and triethylamine. Light, TEA,
and the Ir PS all proved necessary; without any given
2+
Zn photoreduction reactions. Among the Ir(III) complexes
tested, [Ir(ppy) (dtbbpy)](PF ) yielded the highest TONs and
was used in subsequent studies.
Although the photoreduction product possessed similar
reactivity and appearance to Zn ; X-ray diffraction (XRD)
was utilized to confirm its identity as such. The measured XRD
pattern is shown in Figure 3. Both the angle and relative
2
+
2
6
component, no reduction of Zn was observed. TEA is
30
known to reductively quench an excited-state iridium PS. The
Ir(III) PS photocatalytically transfers the reducing equivalents
0
2+
from TEA to Zn , yielding the reduced Zn metal product and
oxidizing TEA.
The oxidation mechanism of TEA appears to be the same as
in similar H evolving systems, in which the two electron
2
25
oxidation of TEA yields diethylamine and acetaldehyde. This
mechanism is confirmed by the failure of anhydrous ZnCl to
2
yield any reduced Zn metal; H O is required to produce
2
3
1
acetaldehyde, which is otherwise supplied by H O bound to
2
the Zn salt crystal lattice. However, prior to the involvement of
25
H O, the oxidation of TEA releases two protons. Despite the
2
influx of protons, there is no pH change over the course of a
photoreduction. The protons released are effectively buffered
−
by TEA, which outnumber Cl by a 7:2 excess in a typical
0
photoreduction mixture. This ensures that the Zn product will
−
not be reoxidized and permits the involvement of free Cl in
the Zn2 reduction mechanism.
+
Several iridium(III) PSs were evaluated for their effectiveness
2
+
0
in the reduction of Zn to Zn . TONs of each Ir PS tested are
shown in Figure 2. All Ir PSs demonstrated some degree of
Figure 3. X-ray diffraction pattern of zinc product (red) with accepted
0
literature patterns for Zn (blue) and ZnO (green) shown underneath.
0
The correspondence of the experimental peaks to Zn confirms zinc
metal as the major photoproduct. Θ−2Θ scan performed at 30 kV, 10
mA, from 20° to 60°.
intensities of the pattern obtained from the photoreduction
32
product match reported literature values for Zn metal. Excess
zinc chloride was removed via MeCN rinses and was not
detected. Minor peaks can be attributed to small quantities of
0
zinc oxide formed by oxidation of the Zn product during XRD
sample processing, or by residual water and O in the reaction
2
vial. Importantly, the Zn metal analysis via H evolution is not
2
0
affected by ZnO, thus giving a true value for the yield of Zn
Figure 2. Catalytic TONs afforded by various iridium(III) PS
with or without ZnO present.
The verification of Zn as the photoreaction product allowed
for the systematic optimization of the photocatalytic system. In
addition to iridium PS tuning, studies were conducted varying
0
complexes after 48 h illumination. A constant concentration of 110
−
μM and PF6 counterion to each Ir PS ensured direct comparison.
Additionally, each Ir PS photoreduction mixture contained 9 mL of
MeCN, 1 mL of TEA, and 420 mg of ZnCl . 95% percent confidence
2
the solvent, H O concentration, and zinc salt, with the best
2
intervals are shown. The abbreviations for the ligands used are as
follows: ppy: 2-phenylpyridine, F-mppy: 2-(4-fluorophenyl)-5-meth-
ylpyridine, dF-mppy: 2-(2,4-difluorophenyl)-5-methylpyridine, Ph-
mppy: 2-([1,1′-biphenyl]-4-yl)-5-methylpyridine, MeO-mppy: 2-(4-
methoxyphenyl)-5-methylpyridine, bpy: 2,2′-bipyridine, dtbbpy: 4,4′-
conditions from each experiment being used in the subsequent.
The use of different zinc salts as a Zn source was found to
have a significant effect on the TONs of the Ir(III) PS. Figure 4
2+
shows the results of this study. While ZnCl yielded the most
2
0
reduced Zn , the most remarkable trend is the superior
di-tert-butyl-2,2′-bipyridine, phen: 1,10-phenanthroline, Me -phen:
4
3
,4,7,8-tetramethyl-1,10-phenanthroline, dppe: (Z)-1,2-bis-
performance of ZnX halide salts. Despite higher solubility,
2
0
(
diphenylphosphino)ethane.
other salts afforded much less Zn product. Zn(BF ) was the
4
2
only salt that completely dissolved in the 1 TEA: 9 MeCN
0
photoreaction solution, yet produced only 5 μmol of Zn in
photocatalytic capacity, with the exception of [Ir-
ppy) (dppe)](PF ). Presumably, [Ir(ppy) (dppe)](PF ) is
photoreductions. This indicates that to some extent, the halide
counterion remains coordinated to the Zn(II) reactant.
Furthermore, it suggests halides may play a role in the reaction
(
2
6
2
6
ineffective because it luminesces from its π to π* excited
2+
state; all other Ir PSs tested emit via triplet state metal to ligand
mechanism, aiding in Zn reduction perhaps by stabilizing a
3
27b
+
charge transfer ( MLCT) transitions.
Though a dominant
Zn intermediate.
3
MLCT proved necessary, there was little correlation between
photocatalytic ability and any other photophysical properties of
the Ir PS, including excited state lifetime, extinction coefficient
at 465 nm, and emission wavelength. This stands in contrast to
The redox behavior of fully dissolved ZnI , ZnCl , and ZnBr
2
2
2
was examined through cyclic voltammetry in MeCN to
0
determine if reduction potential correlated with yield of Zn .
These results are shown in Figure 5. In theory, a less negative
similar H O reducing systems previously studied by the
reduction potential of the ZnX salt would indicate a greater
2
2
5
796
dx.doi.org/10.1021/ic302629q | Inorg. Chem. 2013, 52, 5794−5800

Inorganic Chemistry
Article
Similar to the zinc salt study, the photocatalytic ability of the
[
Ir(ppy) (dtbbpy)](PF ) catalyst was tested in different
2
6
solvents to further optimize its performance. TONs of the Ir
PS were maximized in acetonitrile as the solvent (9 mL
MeCN:1 mL TEA); 289 were achieved after 3 days. The
observation that zinc chloride had neither the highest nor
lowest solubility in MeCN prompted the addition of MeCN to
other solvent bodies. Even as the minority component in other
solvents (1 mL MeCN:8 mL solvent:1 mL TEA), the addition
of MeCN increased TONs substantially, as Figure 6
demonstrates.
Figure 4. TONs of [Ir(ppy) (dtbbpy)](PF ) in the reduction of
2
6
0
various Zn(II) salts to Zn . Each photoreduction mixture contained 9
mL of MeCN, 1 mL of TEA and 110 μM [Ir(ppy) (dtbbpy)](PF ) in
2
6
addition to 2.01 mmol of the given Zn(II) salt.
Figure 6. [Ir(ppy) (dtbbpy)](PF ) TONs in solvent bodies of varied
2
6
compositions. Photoreduction mixtures “w/MeCN” contained 1 mL
of TEA, 1 mL of MeCN, and 8 mL of solvent. Other photoreactions
lacked MeCN and contained 1 mL of TEA, 9 mL of solvent. The
samples were 110 μM in Ir PS and contained 420 mg of ZnCl2.
Figure 5. Cyclic voltammograms of ZnCl (red), ZnI (green), and
ZnBr2 (blue). MeCN and 0.1 M tetrabutylammonium hexafluor-
2
2
ophosphate (TBAPF ) served as solvent and electrolyte, at a scan rate
6
of 0.1 V/s.
These results indicate that MeCN may yield higher TONs
because of a participation in the reaction mechanism, and that it
does not play an innocent role in the greater solvent body. In
small quantities, MeCN could still form part of the inner
solvation shell or coordinate directly to a zinc species. It is
unlikely that MeCN interacts directly with the Ir PS to produce
thermodynamic driving force and thus more TONs in
reduction to Zn by an Ir PS. However, no correlation can
0
be found between reduction potential of ZnX from Figure 5
2
0
and yield of Zn in Figure 4.
such results; Ir(III) complexes have shown greater capacity to
This lack of correlation could be explained by different
25
2
+
photoreduce H O in other solvents.
2
mechanisms of Zn reduction by CV and by an Ir(III) PS.
+
Cyclic voltammetry was performed to test whether the effect
of acetonitrile on TONs observed in Figure 6 is a result of a
Figure 5 shows no sign of Zn , a highly unstable species. The
reduction of Zn2+ to Zn0 by CV is a simultaneous 2 electron
shift in the reduction potential of ZnCl . Electrochemistry
2
transfer; Zn(II) is reduced directly to zinc metal. This differs
from the accepted method of reduction by a heteroleptic Ir(III)
experiments were performed at low ZnCl concentrations, well
2
within its solubility limit in tetrahydrofuran (THF) and MeCN.
Figure 7 shows two dramatically different cyclic voltammo-
PS, which was shown to reduce H O in a similar system by
2
1
3,33
consecutive transfers of a single electron.
The possible
involvement of any Ir(I) species can be excluded because it is
not observed electrochemically within the potential window of
this study. The bpy and then ppy ligands are reduced first, and
catalysis by iridium complexes has been shown to involve a
∧
∧
0
singly reduced [Ir(C N) (N N)] species that is characterized
2
−
25
by e population of the π* orbital of the diimine ligand. The
+
lifetime of Zn would be short, but relatively stable Zn(I)
compounds are known to exist as diamagnetic Zn22+ species,
which bind an array of bridging ligands to form dinuclear
3
4
+
complexes. A longer lived Zn species stabilized by a halide
ligand could explain how a 2 e redox reaction could be driven
−
photochemically by an Ir(III) PS. Halides are also known for
their bridging capabilities, and ZnCl is thought to stabilize
2
2
+
Zn2 , something that would also increase TONs in an
environment in which two proximal Zn ions would be likely
Figure 7. Cyclic voltammograms of ZnCl in pure THF and in THF
2
+
with 10% acetonitrile v/v. The scan rate was 0.1 V/s and 0.1 M
35
to disproportionate.
TBAPF served as the electrolyte.
6
5
797
dx.doi.org/10.1021/ic302629q | Inorg. Chem. 2013, 52, 5794−5800

Inorganic Chemistry
Article
grams of ZnCl in THF with and without 10% v/v MeCN. The
scan in pure THF exhibits only slight reduction of Zn and
Zn(II) interaction. Proton NMR studies revealed that the
2
2+
addition of ZnCl to a TEA solution in acetonitrile failed to
2
bears little resemblance to the scan of ZnCl in MeCN depicted
shift TEA peaks, indicating a lack of coordination of the amine
2
2+
in Figure 5. Conversely, upon the addition of 10% MeCN,
there is both a positive potential shift and increase in the
to Zn .
To further elucidate the mechanism of Zn2+ photoreduction
2+
reduction of Zn , as well as a strong resemblance to the ZnCl
by an Ir PS, several kinetics experiments were performed. First,
the photoreduction rate was measured and analyzed, with the
2
CV scan in Figure 5. Addition of H O reveals a dominating
2
0
water reduction peak at less negative voltages before the onset
of zinc reduction, thus obscuring the Zn(II) reduction peak and
results displayed in Figure 9. After 100 h, the Zn production
had mostly ceased, and the experimental data demonstrates a
strong fit with first order rate decay.
making it difficult to determine the effect of H O on the
2
reduction potential of Zn(II).
While the decreased reaction rate over time could be
These results reinforce the hypothesis that acetonitrile
attributed to the consumption of ZnCl starting material or
2
2
+
interacts closely with Zn in its reduction mechanism. The
decay of the Ir(III) PS, the latter appears more likely. Before
increased reduction potential of ZnCl upon addition of MeCN
the illumination of the photoreductions, [Ir(ppy) (dtbbpy)]-
2
2
to THF explains the increased TONs under these conditions,
and also presumably the similar effect observed in other
solvents that were not suitable for CV measurements.
Furthermore, the success of small parts MeCN in yielding
high TONs allows for the possibility of utilizing a more benign
solvent to improve environmental impact while preserving high
(PF ) is completely dissolved but ZnCl is saturated. If the first
6
2
order kinetics observed were a result of a decrease in dissolved
2+
Zn , a longer initial constant rate would be expected in Figure
9 because of ZnCl saturation. In the case of a heterogeneous
2
reaction, more complicated kinetics would be expected.
Conversely, first order deterioration of cyclometalated Ir(III)
0
Zn yields. It is noteworthy that TONs increase as the potential
PS have been observed in similar H systems by the dissociation
2
2
+
36
difference between the onset of Zn reduction and the
reduction potential of [Ir(ppy) (dtbbpy)](PF ) (E = −1.18
of the bipyridine ligand.
Ten days of illumination yielded an average of 473 μmol Zn0
2
6
red
2
7a
V vs SHE ) grows. This energy difference is ultimately
responsible for driving the reaction.
product (same conditions, not shown in Figure 9). Using 473
0
μmol as the maximum yield of Zn , the linearized kinetic first
2
Along the same lines, water was added to the photoreduction
solvent body to investigate its effect on the catalytic system.
Small quantities of water, down to 50 μL/10 mL photo-
reduction vial, inhibited the ability of [Ir(ppy) (dtbbpy)](PF )
order decay model showed a good fit. The high R value
confirms the first order catalytic process observed in this
system. 473 μmol Zn0 is equivalent to 430 TONs of
[Ir(ppy) (dtbbpy)](PF ), which is the highest TON value
2
6
2
6
2+
to reduce Zn . The concurrent production of H was also
reported in this work.
2
measured, both of which are shown in Figure 8. The gradual
Kinetics experiments examining the dependence of the initial
rate on Ir PS concentration were also performed. To achieve
this, Zn yields were repeatedly measured at 1, 2, 4, and 8 h for
0
each concentration of Ir(III) PS. The slopes of the lines
generated represent the initial reaction rate, which in turn were
plotted against concentration of the Ir PS, shown in Figure 10.
The resulting linear graph indicates a first order dependence of
initial rate with respect to catalyst concentration. This
dependence is consistent both with consecutive one electron
0
+
transfers to produce Zn via Zn , or single electron transfers to
2+
two separate Zn+ species and subsequent disproportionation
between two Zn ions.
Similar initial rate experiments were conducted to determine
the dependence of the initial rate on the mass of ZnCl added
2
0
Figure 8. Dependence of Zn and H yields on H O volume in
2
2
to photoreductions. The plot of initial rate vs ZnCl mass
2
reaction mixture. Photoreduction mixtures contained 110 μmol
Ir(ppy) (dtbbpy)](PF ), 420 mg of ZnCl , 1 mL of TEA, H O,
and enough MeCN to bring to volume (10 mL total).
concentration can be seen in Figure 11. Mass concentration is
used because of saturation of the photoreduction mixture. The
linear graph again indicates a first order dependence, though
[
2
6
2
2
these results are difficult to interpret given that the ZnCl is not
2
−
2+
rise in H production indicates that the steep drop of Zn0
2
production is not the result of competitive inhibition. Instead,
completely dissolved and that Cl and Zn both participate in
the reaction mechanism. Furthermore, the positive x intercept
raises questions as to the nature of the Zn2+ species reduced.
These results are difficult to interpret and do not
unambiguously identify the reaction mechanism, but the
saturated photoreduction mixtures and the apparent existence
H O must play an active role in inhibiting the reduction of
2
2
+
Zn .
2
+
Given the affinity of H O for Zn , the addition of water
2
would readily displace a relatively weak acetonitrile ligand
bound to a zinc ion. This explanation is consistent with the
mechanism postulated previously that the coordination of
of a minimum required ZnCl concentration engender the
2
2+
possibility that the photoreduction of Zn is heterogeneous.
2+
MeCN to zinc is required for the photoreduction of Zn . It
also supports the CV evidence which indicated that the
reduction of water occurs at less negative potentials than the
CONCLUSION
■
2+
The photon driven reduction of Zn by an iridium(III) PS and
catalyst is described. TONs of the Ir PS were systematically
optimized through Ir ligand tuning, variation of solvent and
2+
reduction of Zn . In fact, dominance of acetonitrile ligands in
the coordination environment of the ionic zinc species is
further supported by excluding the possibility of any TEA-
zinc salt, as well as by the exclusion of H O. In conditions
2
5
798
dx.doi.org/10.1021/ic302629q | Inorg. Chem. 2013, 52, 5794−5800

Inorganic Chemistry
Article
0
Figure 9. Measured Zn production with respect to time (red), the optimal kinetic trace for first order rate decay (blue), and rate of reaction (green).
Reaction conditions were 1 mL of TEA, 9 mL of MeCN, 420 mg of ZnCl , and 110 μM Ir(ppy) (dtbbpy)](PF ) per vial (A). Linearized
2
2
6
experimental data and kinetic model of first order decay, adapted from data shown in Figure 9A (B).
Figure 10. Initial steady-state zinc production at various concentrations of [Ir(ppy) (dtbbpy)](PF ). Each photoreduction mixture contained 1 mL
2
6
of TEA, 9 mL of MeCN, 420 mg of ZnCl , and the Ir PS (A). First order kinetic dependence of the initial reaction rate on concentration of the
2
[
Ir(ppy) (dtbbpy)](PF ) PS. Each data point is an average of 4 samples (B).
2 6
reduction mechanism, allowing for further optimization of and
investigation into such a catalytic system.
AUTHOR INFORMATION
■
Corresponding Author
*
E-mail: bern@cmu.edu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Figure 11. Kinetic dependence of initial reaction rate on the mass
The authors acknowledges support from the National Science
Foundation through CHE-1055547 and the Princeton MRSEC
Grant DMR-0819860. Dr. Paul Salvador and Yiling Zhang of
Carnegie Mellon Materials Science and Engineering performed
the XRD measurement, which is greatly appreciated. The
authors thank Kristen Hadley for assistance in hydrogen
evolution experiments.
concentration of ZnCl in heterogeneous photoreductions. Reaction
2
conditions were 1 mL of TEA, 9 mL of MeCN, specified mass of
ZnCl , and 110 μM [Ir(ppy) (dtbbpy)](PF ) per vial.
2
2
6
employing [Ir(ppy) (dtbbpy)](PF ) as the PS, ZnCl , and pure
2
6
2
acetonitrile solvent, a maximum of 430 turnovers were
achieved. Cyclic voltammetry was used to help elucidate the
reaction mechanism, which appears to involve acetonitrile
REFERENCES
■
(1) Brabec, C.; Gowrisanker, S.; Halls, J. M. M.; Laird, D.; Jia, S. J.;
Williams, S. P. Adv. Mater. 2010, 22, 3839−3856.
2
+
+
coordination to Zn and the stabilization of a Zn intermediate
−
(2) Hagfeldt, A.; Boschloo, G.; Sun, L. C.; Kloo, L.; Pettersson, H.
Chem. Rev. 2010, 110, 6595−6663.
by a halide X from the starting material ZnX .
2
Kinetics measurements revealed a first order dependence of
the initial rate on the Ir(III) PS, suggesting that each Ir PS will
twice absorb a photon and transfer an electron per Zn2+ ion
(
3) Kraemer, D.; Poudel, B.; Feng, H. P.; Cavlor, J. C.; Yu, B.; Yan,
X.; Ma, Y.; Wang, X. W.; Wang, D. Z.; Muto, A.; McEnaney, K.;
Chiesa, M.; Ren, Z. F.; Chen, G. Nat. Mater. 2011, 10, 532−538.
(4) McDaniel, N. D.; Bernhard, S. Dalton Trans. 2010, 39, 10021−
reduced. Additionally, a first order dependence on ZnCl and a
2
first order decay of reaction rate was found. These results offer
1
0030.
2
+
a proof of concept for the reduction of Zn by a molecular
catalyst as well as a basis of understanding of the photo-
(5) Chen, X. B.; Shen, S. H.; Guo, L. J.; Mao, S. S. Chem. Rev. 2010,
110, 6503−6570.
5
799
dx.doi.org/10.1021/ic302629q | Inorg. Chem. 2013, 52, 5794−5800

Inorganic Chemistry
Article
(
6) Andreiadis, E. S.; Chavarot-Kerlidou, M.; Fontecave, M.; Artero,
V. Photochem. Photobiol. 2011, 87, 946−964.
(
(
(
7) Fichtner, M. J. Alloys Compd. 2011, 509, S529−S532.
8) Beck, F.; Ruetschi, P. Electrochim. Acta 2000, 45, 2467−2482.
9) Lee, J.-S.; Tai Kim, S.; Cao, R.; Choi, N.-S.; Liu, M.; Lee, K. T.;
̈
Cho, J. Adv. Energy Mater. 2011, 1, 34−50.
10) Wen, Y.-H.; Cheng, S.-Q.; Yang, Y.-S. J. Power Sources 2009,
88, 301−307.
11) Flamigni, L.; Barbieri, A.; Sabatini, C.; Ventura, B.; Barigelletti,
F. Top. Curr. Chem. 2007, 281, 143−203.
12) Yuan, Y.-J.; Zhang, J.-Y.; Yu, Z.-T.; Feng, J.-Y.; Luo, W.-J.; Ye, J.-
H.; Zou, Z.-G. Inorg. Chem. 2012, 51, 4123−4133.
13) Curtin, P. N.; Tinker, L. L.; Burgess, C. M.; Cline, E. D.;
Bernhard, S. Inorg. Chem. 2009, 48, 10498−10506.
14) Baranoff, E.; Yum, J.-H.; Graetzel, M.; Nazeeruddin, M. K. J.
Organomet. Chem. 2009, 694, 2661−2670.
15) Lalevee, J.; Tehfe, M. A.; Dumur, F.; Gigmes, D.; Blanchard, N.;
Morlet-Savary, F.; Fouassier, J. P. ACS Macro Lett. 2012, 1, 286−290.
16) Tucker, J. W.; Nguyen, J. D.; Narayanam, J. M. R.; Krabbe, S.
W.; Stephenson, C. R. J. Chem. Commun. 2010, 46, 4985−4987.
17) Nagib, D. A.; Scott, M. E.; MacMillan, D. W. C. J. Am. Chem.
(
1
(
(
(
(
(
(
(
Soc. 2009, 131, 10875−10877.
(
(
(
18) Ziessel, R. J. Am. Chem. Soc. 1993, 115, 118−127.
19) Metz, S.; Bernhard, S. Chem. Commun. 2010, 46, 7551−7553.
20) Kang, P.; Cheng, C.; Chen, Z.; Schauer, C. K.; Meyer, T. J.;
Brookhart, M. J. Am. Chem. Soc. 2012, 134, 5500−5503.
21) DiSalle, B. F.; Bernhard, S. J. Am. Chem. Soc. 2011, 133, 11819−
1821.
(
1
(
22) Whitten, D. G. Acc. Chem. Res. 1980, 13, 83−90.
(
23) Abanades, S.; Charvin, P.; Flamant, G. Chem. Eng. Sci. 2007, 62,
6
323−6333.
24) Keunecke, M.; Meier, A.; Palumbo, R. Chem. Eng. Sci. 2004, 59,
695−2704.
25) Cline, E.; Adamson, S.; Bernhard, S. Inorg. Chem. 2008, 47,
0378−10388.
26) Han, Z.; McNamara, W. R.; Eum, M.-S.; Holland, P. L.;
Eisenberg, R. Angew. Chem., Int. Ed. 2012, 51, 1667−16670.
27) (a) Slinker, J. D.; Gorodetsky, A. A.; Lowry, M. S.; Wang, J. J.;
Parker, S.; Rohl, R.; Bernhard, S.; Malliaras, G. G. J. Am. Chem. Soc.
004, 126, 2763−2767. (b) Lowry, M. S.; Hudson, W. R.; Pascal, R.
A.; Bernhard, S. J. Am. Chem. Soc. 2004, 126, 14129−14135.
c) Goldsmith, J. I.; Hudson, W. R.; Lowry, M. S.; Anderson, T. H.;
Bernhard, S. J. Am. Chem. Soc. 2005, 127, 7502−7510.
28) Pavlishchuk, V. V.; Addison, A. W. Inorg. Chim. Acta 2000, 298,
7−102.
29) Vogel, A. L. A Textbook of Quantitative Inorganic Analysis;
Logmans, Green and Co Ltd: London, U.K., 1951; Vol. 2.
30) Zhang, P.; Jacques, P.-A.; Chavarot-Kerlidou, M.; Wang, M.;
Sun, L.; Fontecave, M.; Artero, V. Inorg. Chem. 2012, 51, 2115−2120.
31) Chow, Y. L.; Danen, W. C.; Nelsen, S. F.; Rosenblatt, D. H.
Chem. Rev. 1978, 78, 243−274.
32) (a) Swanson, H. E.; Tatge, E. Natl. Bur. Stand. Circ. (U. S.)
953, 539, 95. (b) Garcia-Martinez, O.; Rojas, R. M.; Vila, E.; de
Vidales, J. L. M. Solid State Ionics 1993, 63, 442−449.
33) Gartner, F.; Denurra, S.; Losse, S.; Neubauer, A.; Boddien, A.;
(
2
(
1
(
(
2
(
(
9
(
(
(
(
1
(
̈
Gopinathan, A.; Spannenberg, A.; Junge, H.; Lochbrunner, S.; Blug,
M.; Hoch, S.; Busse, J.; Gladiali, S.; Beller, M. Chem.Eur. J. 2012, 18,
3
(
220−3225.
34) (a) Resa, I.; Carmona, E.; Gutierrez-Puebla, E.; Monge, A.
Science 2004, 305, 1136−1138. (b) Li, T.; Schulz, S.; Roesky, P. W.
Chem. Soc. Rev. 2012, 41, 3759−3771.
(
35) Kerridge, B. D. H.; Tariq, S. A. J. Chem. Soc. A 1969, 1122−
1
(
125.
36) Tinker, L. L.; McDaniel, N. D.; Curtin, P. N.; Smith, C. K.;
Ireland, M. J.; Bernhard, S. Chem.Eur. J. 2007, 13, 8726−8732.
5
800
dx.doi.org/10.1021/ic302629q | Inorg. Chem. 2013, 52, 5794−5800