Mechanism of the formation of a Mn-based CO2 reduction catalyst revealed by pulse radiolysis with time-resolved infrared detection

Article abstract of DOI:10.1021/ja501051s

Using a new technique, which combines pulse radiolysis with nanosecond time-resolved infrared (TRIR) spectroscopy in the condensed phase, we have conducted a detailed kinetic and mechanistic investigation of the formation of a Mn-based CO₂ reduction electrocatalyst, [Mn(^tBu ₂-bpy)(CO)₃]₂ (^tBu₂-bpy = 4,4′-^tBu₂-2,2′-bipyridine), in acetonitrile. The use of TRIR allowed, for the first time, direct observation of all the intermediates involved in this process. Addition of excess [ⁿBu ₄N][HCO₂] to an acetonitrile solution of fac-MnBr( ^tBu₂-bpy)(CO)₃ results in its quantitative conversion to the Mn-formate complex, fac-Mn(OCHO)(^tBu ₂-bpy)(CO)₃, which is a precatalyst for the electrocatalytic reduction of CO₂. Formation of the catalyst is initiated by one-electron reduction of the Mn-formate precatalyst, which produces the bpy ligand-based radical. This radical undergoes extremely rapid (τ = 77 ns) formate dissociation accompanied by a free valence shift to yield the five-coordinate Mn-based radical, Mn^?( ^tBu₂-bpy)(CO)₃. TRIR data also provide evidence that the Mn-centered radical does not bind acetonitrile prior to its dimerization. This reaction occurs with a characteristically high radical-radical recombination rate (2k_dim = (1.3 ± 0.1) × 10⁹ M^-1 s^-1), generating the catalytically active Mn-Mn bound dimer.

Full text of DOI:10.1021/ja501051s

Communication
pubs.acs.org/JACS
Mechanism of the Formation of a Mn-Based CO2 Reduction Catalyst
Revealed by Pulse Radiolysis with Time-Resolved Infrared Detection
David C. Grills,* Jaime A. Farrington,† Bobby H. Layne, Sergei V. Lymar, Barbara A. Mello,
Jack M. Preses, and James F. Wishart
Chemistry Department, Brookhaven National Laboratory, P.O. Box 5000, Upton, New York 11973-5000, United States
S
* Supporting Information
intermediate, mer-[Mn(Me2-bpy)(CO)3(C(O)OH)]+ that was
identified by pulsed EPR.1c A very recent paper3 also describes
ABSTRACT: Using a new technique, which combines
pulse radiolysis with nanosecond time-resolved infrared
(TRIR) spectroscopy in the condensed phase, we have
conducted a detailed kinetic and mechanistic investigation
of the formation of a Mn-based CO2 reduction electro-
catalyst, [Mn(tBu2-bpy)(CO)3]2 (tBu2-bpy = 4,4′-tBu2-
2,2′-bipyridine), in acetonitrile. The use of TRIR allowed,
for the first time, direct observation of all the intermediates
involved in this process. Addition of excess [nBu4N]-
[HCO2] to an acetonitrile solution of fac-MnBr(tBu2-
bpy)(CO)3 results in its quantitative conversion to the
Mn−formate complex, fac-Mn(OCHO)(tBu2-bpy)(CO)3,
which is a precatalyst for the electrocatalytic reduction of
CO2. Formation of the catalyst is initiated by one-electron
reduction of the Mn−formate precatalyst, which produces
the bpy ligand-based radical. This radical undergoes
extremely rapid (τ = 77 ns) formate dissociation
accompanied by a free valence shift to yield the five-
coordinate Mn-based radical, Mn•(tBu2-bpy)(CO)3. TRIR
data also provide evidence that the Mn-centered radical
does not bind acetonitrile prior to its dimerization. This
reaction occurs with a characteristically high radical−
radical recombination rate (2kdim = (1.3 0.1) × 109 M−1
s−1), generating the catalytically active Mn−Mn bound
dimer.
the use of 1 (R = H) as a precatalyst for the reduction of CO2 to
formic acid in a photocatalytic system that makes use of
[Ru(Me2-bpy)3]2+ as a visible light-absorbing photosensitizer
and 1-benzyl-1,4-dihydronicotinamide (BNAH) as a sacrificial
electron donor. In that case, formation of dimer 4 was initially
observed, followed by its decay upon further irradiation into an as
yet unidentified “Mn species” that is presumed to be the active
catalyst.
MnBr(R2‐bpy)(CO)3(1) + e−
→ [MnBr(R2‐bpy•−)(CO)3]−(2)
(1)
2 → Mn•(R2‐bpy)(CO)3(3) + Br−
(2)
kdim
3 + 3 ⎯⎯→⎯ (CO)3(R2‐bpy)Mn‐Mn(R2‐bpy)(CO)3(4)
(3)
4 + 2e− → 2[Mn(R2‐bpy)(CO)3]−(5)
(4)
Clearly, there is still much to learn about the mechanism of
CO2 reduction catalysis with Mn complexes, and gaining this
knowledge is of critical importance in terms of developing a new
generation of low-cost catalysts that are suitable for practical
applications. The techniques applied so far have revealed
important information, but they have not allowed some of the
intermediate species to be observed, e.g., 2, 3, and any CO2
adducts other than the hydroxycarbonyl species measured at low
temperature by EPR.1c We have begun to address these issues by
making use of pulse radiolysis combined with nanosecond-TRIR
detection, focusing initially on the one-electron reduction of a
new Mn−formate complex, fac-Mn(OCHO)(tBu2-bpy)(CO)3
(6), and its subsequent conversion to the dinuclear catalyst 4
(eqs 9, 10, and 3). Cyclic voltammetry measurements have
shown that 6 is indeed an effective precatalyst for electrocatalytic
CO2 reduction, behaving identically to 1 (see Figure S3).
Pulse radiolysis utilizes a short, high-energy electron pulse for
the rapid production of radical ions.4 The subsequent reactions
of these species are typically monitored by transient absorption
(TA) spectroscopy in the UV/vis/near-IR regions. However, the
often-broad absorption bands can make the identification of
reactive intermediates challenging. To overcome this limitation,
researchers have turned to TRIR detection, which takes
ecent work1 has shown that the manganese complexes, fac-
R
MnBr(R2-bpy)(CO)3 (1; R2-bpy = 4,4′-R2-2,2′-bipyridine;
t
R = H, Me, or Bu), which are analogues of a well-established
family of rhenium catalysts,2 can act as precatalysts for the
efficient and selective electrocatalytic reduction of CO2 to CO in
acetonitrile (CH3CN) in the presence of Brønsted acids. The
Mn-based catalysts are of considerable interest because they
contain an earth-abundant first-row transition metal and facilitate
CO2 reduction at ∼0.3 V lower overpotential than their Re
counterparts, while exhibiting a similar catalytic activity
(particularly for R = tBu).1b The proposed mechanism1a,b begins
with formation of the active catalytic species 4 and 5 from the
precatalyst 1 as shown in eqs 1−4 (see Chart S1 for chemical
structures).
Once formed, species 4 and 5 engage in the catalytic cycle by
two pathways: (i) a two-electron pathway in which 5 is the
catalytically active species that reacts with CO2,1 and (ii) a one-
electron pathway in which the Mn−Mn bound dimer 4 is reactive
toward CO2 and H+, forming a Mn(II) hydroxycarbonyl
Received: January 30, 2014
Published: March 28, 2014
© 2014 American Chemical Society
5563
dx.doi.org/10.1021/ja501051s | J. Am. Chem. Soc. 2014, 136, 5563−5566

Journal of the American Chemical Society
Communication
advantage of the structural specificity of mid-IR spectroscopy.
However, until we recently developed a nanosecond TRIR
detection capability for condensed-phase pulse radiolysis,5 this
had been restricted to a handful of gas-phase studies on the
microsecond time scale6a,b and an FTIR-based approach with
∼30 s time resolution.7 Laser flash photolysis combined with
TRIR detection is a well-established alternative approach for the
investigation of transient intermediates.8 However, it is not
applicable to the chemistry under investigation here due to the
extreme photosensitivity of 1 that results in CO ligand loss and/
or light-induced isomerization.9
Pulse radiolysis of CH3CN results in the formation of solvated
electrons, es−, solvent radicals,•CH2CN, and protons (eq 5).10
Since es− is a strong reducing agent, rapid one-electron reduction
of a dissolved solute ensues.10a,c In the presence of formate anion,
it has been shown that the oxidizing side product, •CH2CN can
be scavenged via hydrogen atom transfer, resulting in the
production of a second reducing equivalent, CO2•−, as shown in
eq 6.11 In principle, this should result in a net doubling of the
yield of the one-electron reduced (OER) solute (eq 7). Further
details of these processes are provided in the SI.
Figure 1. DFT-calculated IR spectra (9 cm−1 Gaussian fwhm) of species
3−4 (R = Bu) and 6−8, calculated with a conductor polarizable
continuum model for CH3CN.
t
on the Mn center and increased π back-bonding into the π*(CO)
antibonding orbitals, resulting in a weakening of the CO bonds.
The five-coordinate radical 3 has three ν(CO) bands that are
shifted further to the red relative to those of 7, since 3 is a Mn(0)
species with the unpaired electron residing mainly on the metal
center. The calculated IR spectrum of the skewed cis isomer14 of
4 is in good agreement with experiment.1b The solvent-bound,
ligand-localized radical 8 will have ν(CO) IR bands that are very
close to those of 6, since it is a neutral Mn(I) species. The
locations of the unpaired electrons in 7, 3 and 8 can be seen in
their calculated singly occupied molecular orbitals (SOMO’s) in
Figure S1.
•
CH3CN ⇝ es−, CH2CN, H+
(5)
−
•
•−
•CH2CN + HCO2 → CH3CN + CO2
(6)
(7)
•−
2 solute + es−, CO2 → 2 solute•− + CO2
•
Since it is essential to eliminate CH2CN from the reaction
mixture, all of our pulse radiolysis measurements were performed
in the presence of 50 mM formate. IR and NMR spectroscopy
revealed that under these conditions, the bromide ligand of 1 is
quantitatively displaced by formate in the time of mixing, to
generate the formate complex 6 (eq 8, see SI for IR and NMR
spectra). Previous IR spectroelectrochemical (SEC) measure-
ments on 1 (and its chloride analogue) have shown that one-
electron reduction results in dimer 4 as the only observed
product.1b,12 It was assumed that the OER species, 2, is unstable,
rapidly losing Br− (or Cl−), followed by the resulting radicals
dimerizing within the time scale of the electrochemical
experiment. In this work, we are investigating the analogous
conversion of 6 to 4 (eqs 9, 10, and 3). Since 7 and 3 have never
been experimentally observed, we used density functional theory
(DFT)13 to calculate the ν(CO) IR spectra of all the Mn species
in eqs 9, 10, and 3 (6, 7, 3 and 4; R = tBu), together with another
possible intermediate that may result from the reaction of 3 with
CH3CN, i.e., the ligand-localized radical, fac-Mn(tBu2-bpy•−)-
(CO)3(CH3CN) (8). These DFT-calculated spectra will provide
a useful comparison with the pulse radiolysis-TRIR spectra and
are shown in Figure 1, with the band positions listed in Table S1.
TRIR spectra at three selected time delays after pulse radiolysis
of a solution of 6 in CH3CN are shown in Figure 2. All ν(CO) IR
Figure 2. IR spectrum of a 1.5 mM solution of 6 in CH3CN containing
50 mM [nBu4N][HCO2] (bottom) and TRIR spectra recorded 40 ns,15
300 ns, and 1.5 ms after pulse radiolysis of this argon-purged solution
(top). The black, red, and blue spectra correspond to completion of the
reactions in eqs 9, 10, and 3, respectively.
band positions of the observed transient species are listed in
Table S1. At 40 ns (black spectrum, shown in more detail in
Figure S4), bleaching of the ν(CO) bands of 6 is observed,
together with two new transient bands at 2004 and 1892 cm−1.15
The lower frequency band is broad and likely an overlap of two
bands. These transient bands decay exponentially (τ ≈ 60 ns, see
Figure S8) to reveal a new pair of ν(CO) bands at 1955 and 1853
cm−1 (red spectrum). Again, the lower frequency band is broad
and likely results from the overlap of two bands. Finally, these
two ν(CO) bands disappear through second-order kinetics on
the millisecond time scale (Figure 3) to yield four new ν(CO)
bands at 1973, 1927, 1880, and 1849 cm−1 (blue spectrum),
which correspond to the dimer 4.1b
−
MnBr(tBu2‐bpy)(CO)3(1) + HCO2
−Br−
⎯⎯⎯⎯→ Mn(OCHO)(tBu2‐bpy)(CO)3(6)
(8)
ke
6 + e− → [Mn(OCHO)(tBu2‐bpy•−)(CO)3]−(7)
(9)
kdis
−
7 ⎯→⎯ Mn•(tBu2‐bpy)(CO)3(3) + HCO2
(10)
In Figure 1, we see that the initial OER product, 7, should
exhibit an ∼25 cm−1 red-shift of its ν(CO) bands relative to 6.
Like 6, 7 is a Mn(I) species, but it is negatively charged with the
extra electron residing largely in the π* orbitals of the bpy ligand.
Due to orbital overlap, this results in additional electron density
Although the 1927 cm−1 band of 4 in the TRIR spectrum at 1.5
ms is obscured by the broad bleach bands in this region, it is
5564
dx.doi.org/10.1021/ja501051s | J. Am. Chem. Soc. 2014, 136, 5563−5566

Journal of the American Chemical Society
Communication
apparatus has a much faster response time than the TRIR
system (∼1.5 vs ∼40 ns). TA spectra measured after pulse
radiolysis are shown in Figure 4. Immediately after the electron
Figure 3. TRIR kinetic traces recorded after pulse radiolysis of an argon-
purged 1.5 mM solution of 6 in CH3CN containing 50 mM
[nBu4N][HCO2] at 1956 cm−1 (red) and 1972 cm−1 (blue).
Figure 4. TA spectra recorded after pulse radiolysis of a 3.0 mM solution
of 6 in argon-purged CH3CN containing 50 mM [nBu4N][HCO2].
evident from the way it diminishes the bleach intensity at this
frequency compared to the other time delays. Note that the
gradually increasing bleach intensity between the successive
TRIR spectra is explained by the ∼40 ns response time of the
TRIR apparatus (black to red spectra, see also Figure S8) and the
reduction of a second equivalent of 6, presumably by CO2•− (red
to blue spectra, eqs 6 and 7 and Figure S7).
pulse, TA bands are observed at 520 and 790 nm. These are
assigned to 7, based on a comparison with the known spectrum18
of fac-[ReCl(bpy•−)(CO)3]− and our TD-DFT calculated
spectrum of 7 (Figure S9). With time, these bands are replaced
by an intense absorption at 780 nm and a weaker band at ∼500
nm. These are assigned to 3, again based on a comparison with
the TD-DFT calculated spectrum of 3 (Figure S9). Exponential
fits of kinetic traces at different wavelengths yielded a rate
constant for formate dissociation from 7 of kdis = 1.3 × 107 s−1
(Figure S11).
The time scale for formate dissociation from 7 (τ = 77 ns) is
significantly shorter than that of bromide dissociation from the
Re−Br analogue, fac-[ReBr(tBu2-bpy•−)(CO)3]− (10) by at
least 7 orders of magnitude (see Figure S5). The short lifetime of
7 parallels previous observations by IR-SEC on 1, in which 2 and
3 could not be observed upon one-electron reduction of 1 to
generate dimer 4, due to the instability and short lifetime of
2.1b,12 The labilization of the halide ligand in reduced [MX(α-
diimine•−)(CO)3]− complexes (M = Mn or Re) results from an
overlap of the partially filled α-diimine π* orbital with the
σ*(M−X) antibonding orbital.19 Presumably, a similar effect is
occurring in the reduced Mn−formate complex 7, with charge
leakage into the Mn center being more efficient than in the OER
Re−Br complex 10 due to better matching of the Mn 3d orbitals
with the α-diimine π* orbital.20
The magnitude of the rate constant for dimerization of 3 to 4
(kdim, eq 3) provides further insight into the nature of the Mn
coordination environment in 3. To determine kdim, we took
advantage of the well-separated ν(CO) IR bands of 6, 3, and 4
and measured the TRIR decay kinetics of 3 as a function of its
initial concentration; the latter was modulated through changing
the per pulse radiation dose (see Figures S12 and S13 and
discussion). From these experiments, the dimerization rate
constant was determined to be 2kdim = (1.3 0.1) × 109 M−1 s−1.
This value is some orders of magnitude greater than typical
dimerization rate constants observed for solvent-coordinated
metal complexes with ligand-localized radicals; for instance, 2k =
40 M−1 s−1 has been reported for dimerization of Re(Me2-
bpy•−)(CO)3(THF) in THF.17b Although the kdim value that we
have determined for the Mn-based radical is somewhat below the
diffusion-controlled limit of ∼1010 M−1 s−1, this deviation can be
explained by the below unity steric factor. Indeed, only a fraction
of radical−radical encounters can result in dimerization due to
the screening of the Mn-centered free valence by the surrounding
ligands. We thus take the high kdim value as evidence that the Mn-
The 2004 and 1892 cm−1 bands observed immediately after
pulse radiolysis were confirmed to be due to a Mn carbonyl
species by a control experiment in the absence of 6, in which no
significant transient signals were observed. We can confidently
assign these bands to the OER species, 7, by comparison with the
DFT-calculated spectrum of 7 (Figure 1 and Table S1) and a
pulse radiolysis experiment on the related fac-ReBr(tBu2-
bpy)(CO)3 complex 9 (Figure S5). The next transient observed
(ν(CO) at 1955 and 1853 cm−1) is assigned as a neutral radical
species resulting from formate dissociation from 7. The positions
of the ν(CO) bands of this radical provide useful information
about the Mn coordination environment. They are an average of
44 cm−1 lower in frequency than those of 7 and are a close match
to the DFT-calculated IR bands of the metal-based radical 3. In
contrast, the calculated bands of the solvent-bound ligand-
localized radical 8 lie at much higher frequencies (Figure 1 and
Table S1), which indicates that the species generated upon
formate dissociation from 7 exists as the five-coordinate metal-
based radical 3, with no binding of a solvent molecule occurring
at the vacant coordination site. This conclusion agrees with
previous IR-SEC observations on the Re analogue, Re•(tBu2-
bpy)(CO)3,16 that was shown to exist as a five-coordinate radical
in CH3CN. Interestingly, previous data on related [Re(α-
diimine)(CO)3L]n+ and [Re(α-diimine)(CO)3]2 complexes
have indicated that in coordinating solvents, such as CH3CN,
the neutral radical generated after dissociation of L from the OER
species or by homolysis of the Re−Re bond exists as a six-
coordinate solvent complex with the unpaired electron being
shifted from the metal to the α-diimine ligand.17 Thus, the two
tBu substituents on the bpy ligand in 3 must have sufficient
electron donating ability to destabilize such a complex and favor
the five-coordinate metal-based radical, as was previously
suggested for the Re analogue.16 Although we should expect
the radical−radical reaction between CO2•− and 3 to form a CO2
adduct, i.e., fac-[Mn(tBu2-bpy)(CO)3(CO2)]− (11) in the pulse
radiolysis experiments, it can be shown that this reaction does not
significantly interfere with the dimerization of 3 via reaction 3
(see SI).
To more accurately measure the rate of formate dissociation
from 7, we have performed a complementary pulse radiolysis
experiment with vis/near-IR TA detection, since the TA
5565
dx.doi.org/10.1021/ja501051s | J. Am. Chem. Soc. 2014, 136, 5563−5566

Journal of the American Chemical Society
Communication
centered radical does not bind CH3CN prior to its dimerization
and is best formulated as fac-Mn•(tBu2-bpy)(CO)3.
REFERENCES
■
(1) (a) Bourrez, M.; Molton, F.; Chardon-Noblat, S.; Deronzier, A.
Angew. Chem., Int. Ed. 2011, 50, 9903. (b) Smieja, J. M.; Sampson, M. D.;
Grice, K. A.; Benson, E. E.; Froehlich, J. D.; Kubiak, C. P. Inorg. Chem.
2013, 52, 2484. (c) Bourrez, M.; Orio, M.; Molton, F.; Vezin, H.; Duboc,
C.; Deronzier, A.; Chardon-Noblat, S. Angew. Chem., Int. Ed. 2013, 53,
240.
(2) Hawecker, J.; Lehn, J. M.; Ziessel, R. Helv. Chim. Acta 1986, 69,
1990.
(3) Takeda, H.; Koizumi, H.; Okamoto, K.; Ishitani, O. Chem.
In conclusion, we have used the newly developed technique of
condensed-phase pulse radiolysis combined with TRIR spec-
troscopy, together with conventional TA detection, to examine
the mechanistic details of the formation of a Mn-based CO2
reduction electrocatalyst, [Mn(tBu2-bpy)(CO)3]2, upon one-
electron reduction of a new precatalyst, fac-Mn(OCHO)(tBu2-
bpy)(CO)3 in CH3CN. This precatalyst is generated by
dissolution of the previously investigated1b precatalyst, fac-
MnBr(tBu2-bpy)(CO)3 in CH3CN in the presence of excess
formate, which is used for scavenging the radiolytically generated
solvent radical. The use of TRIR detection allowed, for the first
time, the direct observation and identification of all the
intermediates in the catalyst formation process. The reduced
precatalyst is a bpy ligand-based radical, which undergoes rapid
(τ = 77 ns) formate dissociation to yield a five-coordinate Mn-
based radical, fac-Mn•(tBu2-bpy)(CO)3. The Mn-based radicals
dimerize with a characteristically high radical−radical recombi-
nation rate (2kdim = (1.3 0.1) × 109 M−1 s−1), providing further
evidence that they exist as five-coordinate species in CH3CN.
Future work will focus on the next stage of the catalytic cycle that
involves the reduction of the Mn−Mn dimer and its subsequent
reactions with CO2, in the presence and absence of Brønsted
acids. Since these processes could involve the formation of
formate-like adducts, it may become necessary to replace formate
by a potentially less interfering solvent radical scavenger for these
studies (details in SI).
Commun. 2014, 50, 1491.
(4) (a) Tabata, Y. Pulse Radiolysis; CRC Press: Boca Raton, FL, 1990.
(b) Recent Trends in Radiation Chemistry; Wishart, J. F., Rao, B. S. M.,
Eds.; World Scientific Publishing Co.: Singapore, 2010.
(5) Grills, D. C.; Cook, A. R.; Fujita, E.; George, M. W.; Preses, J. M.;
Wishart, J. F. Appl. Spectrosc. 2010, 64, 563.
(6) (a) Schwarz, H. A. J. Chem. Phys. 1977, 67, 5525. (b) Meunier, H.;
Pagsberg, P.; Sillesen, A. Chem. Phys. Lett. 1996, 261, 277.
(7) Le Caer, S.; Vigneron, G.; Renault, J. P.; Pommeret, S. Chem. Phys.
̈
Lett. 2006, 426, 71.
(8) Butler, J. M.; George, M. W.; Schoonover, J. R.; Dattelbaum, D. M.;
Meyer, T. J. Coord. Chem. Rev. 2007, 251, 492.
(9) In principle, laser flash photolysis could be performed with visible
light (λ > 480 nm, where the Mn complex does not absorb), using a
visible-light absorbing photosensitizer in the presence of a sacrificial
electron donor, resulting in intermolecular electron transfer to the Mn
complex. However, in that case the chemistry may be more complex and
follow a different mechanism, as has been shown in ref 3.
(10) (a) Bell, I. P.; Rodgers, M. A. J.; Burrows, H. D. J. Chem. Soc.,
Faraday Trans. I 1977, 73, 315. (b) Buxton, G. V.; Mulazzani, Q. G.
Radiation-Chemical Techniques. In Electron Transfer in Chemistry;
Balzani, V., Ed.; Wiley-VCH: Weinheim, Germany, 2001; Vol. I-1, pp
503−557. (c) Shkrob, I. A.; Sauer, M. C. J. Phys. Chem. A 2002, 106,
9120.
(11) Burrows, H. D.; Kosower, E. M. J. Phys. Chem. 1974, 78, 112.
(12) Hartl, F.; Rossenaar, B. D.; Stor, G. J.; Stufkens, D. J. Recl. Trav.
Chim. Pays-Bas 1995, 114, 565.
(13) Frisch, M. J. et al. Gaussian 09, revision B.01; Gaussian, Inc.:
Wallingford CT, 2010.
It is clear that the combination of pulse radiolysis and TRIR
spectroscopy is a powerful and useful technique for the rapid
production, identification, and kinetic monitoring of short-lived
intermediates that are involved in redox catalysis. We believe this
method will play an increasing role in unraveling the mechanisms
of such processes, thus contributing to the development of new
generations of catalysts.
(14) The skewed cis isomer was chosen for the calculation since it was
shown in ref 17b to be the most stable isomer of the [Re(bpy)(CO)3]2
dimer in THF solution.
ASSOCIATED CONTENT
■
S
* Supporting Information
(15) Note: weak bands seen at 1955 and 1853 cm−1 in the 40 ns
spectrum are due to a small amount of the second transient species that
has already formed at this time delay (see Figure S8).
(16) Smieja, J. M.; Kubiak, C. P. Inorg. Chem. 2010, 49, 9283.
(17) (a) Hayashi, Y.; Kita, S.; Brunschwig, B. S.; Fujita, E. J. Am. Chem.
Soc. 2003, 125, 11976. (b) Fujita, E.; Muckerman, J. T. Inorg. Chem.
2004, 43, 7636.
Experimental details, characterization data, and complete ref 13.
This material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
dcgrills@bnl.gov
(18) Kalyanasundaram, K. J. Chem. Soc., Faraday Trans. 2 1986, 82,
2401.
(19) (a) Klein, A.; Vogler, C.; Kaim, W. Organometallics 1996, 15, 236.
(b) Stor, G. J.; Hartl, F.; Vanoutersterp, J. W. M.; Stufkens, D. J.
Organometallics 1995, 14, 1115.
Present Address
†Sydor Instruments LLC, 291 Millstead Way, Rochester, NY.
Notes
(20) Johnson, F. P. A.; George, M. W.; Hartl, F.; Turner, J. J.
Organometallics 1996, 15, 3374.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work, and use of the LEAF Facility of the BNL Accelerator
Center for Energy Research, was supported by the US
Department of Energy (DOE), Office of Basic Energy Sciences,
Division of Chemical Sciences, Geosciences & Biosciences under
contract no. DE-AC02-98CH10886 and by BNL Program
Development funding. B.A.M. thanks the DOE and BNL's
Office of Educational Programs for a SULI internship. We are
also grateful to the DOE for FY2010 supplemental capital
funding. We thank Drs. Etsuko Fujita and Jim Muckerman for
helpful discussions.
5566
dx.doi.org/10.1021/ja501051s | J. Am. Chem. Soc. 2014, 136, 5563−5566