Oxygen reduction reactions of monometallic rhodium hydride complexes

Article abstract of DOI:10.1021/ic300279z

Selective reduction of oxygen is mediated by a series of monometallic rhodium(III) hydride complexes. Oxidative addition of HCl to trans-Rh ^ICl(L)(PEt₃)₂ (1a, L = CO; 1b, L = 2,6-dimethylphenylisocyanide (CNXy); 1c, L = 1-adamantylisocyanide (CNAd)) produces the corresponding Rh^III hydride complex cis-trans-Rh ^IIICl₂H(L)(PEt₃)₂ (2a-c). The measured equilibrium constants for the HCl-addition reactions show a pronounced dependence on the identity of the L ligand. The hydride complexes effect the reduction of O₂ to water in the presence of HCl, generating trans-Rh^IIICl₃(L)(PEt₃)₂ (3a-c) as the metal-containing product. In the case of 2a, smooth conversion to 3a proceeds without spectroscopic evidence for an intermediate species. For 2b/c, an aqua intermediate, cis-trans-[Rh^III(OH₂)Cl ₂(L)(PEt₃)₂]Cl (5b/c), forms along the pathway to producing 3b/c as the final products. The aqua complexes were independently prepared by treating peroxo complexes trans-Rh^IIICl(L) (η²-O₂)(PEt₃)₂ (4b/c) with HCl to rapidly produce a mixture of 5b/c and 3b/c. The reactivity of the peroxo species demonstrates that they are plausible intermediates in the O ₂-reduction chemistry of hydride complexes 2a-c. These results together show that monometallic rhodium hydride complexes are capable of promoting selective reduction of oxygen to water and that this reaction may be controlled with systematic alteration of the ancillary ligand set.

Full text of DOI:10.1021/ic300279z

Article
pubs.acs.org/IC
Oxygen Reduction Reactions of Monometallic Rhodium Hydride
Complexes
Thomas S. Teets and Daniel G. Nocera*
Department of Chemistry, 6-335, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, Massachusetts
02139-4307, United States
S
* Supporting Information
ABSTRACT: Selective reduction of oxygen is mediated by a
series of monometallic rhodium(III) hydride complexes.
Oxidative addition of HCl to trans-RhICl(L)(PEt3)2 (1a, L =
CO; 1b, L = 2,6-dimethylphenylisocyanide (CNXy); 1c, L = 1-
adamantylisocyanide (CNAd)) produces the corresponding
RhIII hydride complex cis-trans-RhIIICl2H(L)(PEt3)2 (2a−c).
The measured equilibrium constants for the HCl-addition
reactions show a pronounced dependence on the identity of
the “L” ligand. The hydride complexes effect the reduction of O2 to water in the presence of HCl, generating trans-
RhIIICl3(L)(PEt3)2 (3a−c) as the metal-containing product. In the case of 2a, smooth conversion to 3a proceeds without
spectroscopic evidence for an intermediate species. For 2b/c, an aqua intermediate, cis-trans-[RhIII(OH2)Cl2(L)(PEt3)2]Cl (5b/
c), forms along the pathway to producing 3b/c as the final products. The aqua complexes were independently prepared by
treating peroxo complexes trans-RhIIICl(L)(η2-O2)(PEt3)2 (4b/c) with HCl to rapidly produce a mixture of 5b/c and 3b/c. The
reactivity of the peroxo species demonstrates that they are plausible intermediates in the O2-reduction chemistry of hydride
complexes 2a−c. These results together show that monometallic rhodium hydride complexes are capable of promoting selective
reduction of oxygen to water and that this reaction may be controlled with systematic alteration of the ancillary ligand set.
metal hydrides and especially those involving palladium have
proliferated in the context of the aerobic oxidation
catalysis.42−48 Mechanistic studies of O2 reactivity with group
9 metal hydrides were less numerous32,35,41 leading up to our
observation that dirhodium hydride complexes, accessed by
reversible HCl addition to two-electron mixed-valent centers,
promote the reduction of oxygen to water.49 The analogous
diiridium hydride complex inserts O2 to form an isolable
hydroperoxo complex.50 This finding, coupled with kinetic51
and computational analyses52 of both the dirhodium and the
diiridium systems, led us to conclude that a hydroperoxo
complex is a key intermediate in the 4e−/4H+ oxygen reduction
reaction (ORR) of O2 to water. Though select examples involve
radical chain mechanisms,32,53 most of these recent mechanistic
studies reveal two common pathways for insertion of O2 into a
metal−hydride bond: (i) the “HX reductive elimination”
(HXRE) mechanism,41,45,48 where reductive elimination
precedes reaction of the reduced metal center with O2, and
(ii) “H-atom abstraction” (HAA),42 where O2 reacts directly
with the metal hydride. It has been observed experimentally
that it is possible in some systems for these two mechanisms to
be competitive and occur simultaneously.46,51,52
INTRODUCTION
■
The fundamental reactivity of O2 at transition metal centers is
at the nexus of bioenergy1,2 and chemical energy3,4 conversion.
In nature, oxidase5 and oxygenase6,7 enzymes utilize O2 as the
oxidant in crucial respiratory and biosynthetic catalyses by
managing both the proton and electron inventory. With an eye
toward designing biomimetic catalysts and gaining a more
thorough understanding of energy-conversion mechanisms,
numerous model complexes designed to retain key features of
the enzyme active sites have appeared.8−10 By appending either
Brønsted acids11,12 or auxiliary metal centers9,13−17 in the
secondary coordination sphere of a metallomacrocycle, biofunc-
tional O2 reduction catalysts have been realized. In addition,
because of oxygen’s abundance and environmental compati-
bility, in the realm of synthetic chemistry there is considerable
appeal in developing catalytic systems that utilize O2 as the sole
oxidizing species.18,19 Complexes of iron,20 copper,21 rho-
dium,22 and select other metals23−25 can catalyze aerobic
oxidation reactions, though palladium catalysis is the most
widespread.18,26−28
Insertion of O2 into metal−hydride bonds often plays a key
role in its activation, especially if the reactions occur in acidic
media and/or are attendant to C−H activation. As such, in-
depth studies of the reactions of O2 with metal hydride
complexes are germane to numerous topics in catalysis. There
are many examples of O2 insertion into metal hydrides to
furnish hydroperoxo complexes,29−41 and in recent years
mechanistic studies of the aerobic oxidation of group 10
The proclivity of the HXRE and HAA mechanisms to
compete in the ORR offers the opportunity to assess the
electronic factors that are determinants of the HXRE and HAA
Received: February 7, 2012
Published: June 18, 2012
© 2012 American Chemical Society
7192
dx.doi.org/10.1021/ic300279z | Inorg. Chem. 2012, 51, 7192−7201

Inorganic Chemistry
Article
Preparation of trans-RhICl(CNXy)(PEt3)2 (1b). In a 20 mL
scintillation vial, [RhI(COD)Cl]2 (100 mg, 0.203 mmol, 1.00 equiv)
was suspended in 2 mL of THF. A solution of PEt3 (120 μL, 0.811
mmol, 4.00 equiv) dissolved in 2 mL of THF was added, producing a
light orange solution. Immediate addition of 2,6-dimethylphenyliso-
cyanide (CNXy) (53 mg, 0.40 mmol, 2.0 equiv) in 2 mL of THF
caused the color to fade to bright yellow. The slightly cloudy mixture
was stirred for 1 h at room temperature and then filtered to remove a
small amount of gray solid. The resulting yellow solution was
concentrated in vacuo to give a sticky yellow solid, which was
redissolved in 8 mL of hexane. The solvent was removed in vacuo, and
the resulting microcrystalline yellow solid was dried in vacuo overnight
to remove residual 1,5-cyclooctadiene. Yield: 200 mg (97.6%). 1H
NMR (500 MHz, C6D6) δ/ppm: 6.76−6.82 (m, 3H), 2.33 (s, 6H),
1.80 (m, 12H) 1.11 (quintet, 18H). 31P{1H} NMR (121.5 MHz,
pathways and thus provide further insight into the preferred
mechanism(s) of the aerobic oxidation reactions. We sought to
tune the reactivity of the metal center by systematic alteration
of the electronic properties of the ligand set of the metal center.
In this regard, the two-electron mixed-valent dirhodium and
diiridium complexes that served as the starting points for our
previous studies49,51 offer few synthetic inroads to tune the
electronic properties of these scaffolds. Such a line of inquiry
thus demands the construction of new transition metal scaffolds
to effect ORR activity. We also sought to better understand the
role of the second metal in the O2 activation chemistry of
bimetallic centers and whether metal cooperativity at two-
electron mixed-valent centers54 offered any benefits for the
ORR. Accordingly, we set out to interrogate O2 activation and
reduction at monometallic rhodium hydride complexes. Our
findings are disclosed herein.
Three isostructural rhodium hydride complexes of the type
cis-trans-RhIIICl2H(L)(PEt3)2 (where L = CO, 2,6-dimethyl-
phenylisocyanide (CNXy), or 1-adamantylisocyanide (CNAd))
are found to promote the ORR cleanly in the presence of HCl
with the attendant production of the corresponding trans-
RhIIICl3(L)(PEt3)2. Alteration of L has a dramatic effect on
both the HCl-addition equilibrium constant and the O2-binding
thermodynamics of the parent RhI complexes. We show that
when L is an isocyanide, an aqua−rhodium(III) intermediate
forms prior to generation of the final product. By interrogating
the reactivity of the O2 adducts trans-RhIIICl(CNR)(η2-
O2)(PEt3)2 (R = 2,6-dimethylphenyl, 1-adamantyl) we also
demonstrate that these η2-peroxo complexes and the
spectroscopically unobserved hydroperoxo complexes are
plausible intermediates in the ORR of the hydride complexes.
1
C6D6) δ/ppm: 23.6 (d, JRh−P = 124 Hz). UV−vis (THF): λ/nm (ε/
M−1 cm−1) 265 (24 000), 302 (sh) (8100), 367 (sh) (2600). IR
(Nujol): νCN
̃
= 2054 cm−1. Anal. Calcd for C21H39ClNP2Rh: C,
49.86; H, 7.77; N, 2.77. Found: C, 49.59; H, 7.69; N, 2.64.
Preparation of trans-RhI(CNAd)Cl(PEt3)2 (1c). THF solutions (2
mL) of [RhI(COD)Cl]2 (100 mg, 0.203 mmol, 1.00 equiv) and PEt3
(120 μL, 0.811 mmol, 4.00 equiv) were combined to afford a pale
orange solution. 1-Adamantylisocyanide (CNAd) (66 mg, 0.41 mmol,
2.0 equiv) in 2 mL of THF was introduced, causing the color to fade
to yellow. After stirring for 1 h at room temperature, the solution was
concentrated in vacuo to leave a yellow residue, which was redissolved
in 4 mL of pentane. The yellow solution was concentrated and dried in
vacuo overnight, leaving the product as a bright yellow solid. Yield:
208 mg (95.8%). 1H NMR (500 MHz, C6D6) δ/ppm: 1.88 (m, 12H),
1.80 (br, d, 6H), 1.73 (br, m, 3H), 1.33 (br, m, 6H), 1.20 (quintet,
1
18H). 31P{1H} NMR (121.5 MHz, C6D6) δ/ppm: 22.8 (d, JRh−P
=
127 Hz). UV−vis (THF): λ/nm (ε/M−1 cm−1) 282 (sh) (5700), 308
(9700), 369 (3900). IR (Nujol): νCN
̃
= 2068 cm−1. Anal. Calcd for
C23H45ClNP2Rh: C, 51.55; H, 8.46; N, 2.61. Found: C, 51.42; H, 8.15;
N, 2.53.
Preparation and NMR Characterization of cis-trans-RhIII(CO)-
Cl2H(PEt3)2 (2a). A J. Young NMR tube was charged with 1a (10 mg,
0.025 mmol) dissolved in 0.7 mL of THF-d8. The solution was freeze−
pump−thaw degassed three times at ∼10−6 Torr on a high-vacuum
manifold. Anhydrous HCl, generated by dropping concentrated
H2SO4 onto anhydrous NaCl (36 mg, 0.62 mmol, 25 equiv), was
vacuum transferred to the solution of 1a. While frozen, the NMR tube
was evacuated to ∼10−6 Torr and then thawed to reveal a pale yellow
solution. Under these conditions (where HCl transfer is not
quantitative), 2a and 1a were observed to be present in a ca. 2:1
ratio. Removal of the volatiles resulted in complete reversion to 1a, as
judged by 31P{1H} NMR. 1H NMR (500 MHz, THF-d8) δ/ppm: 2.10
EXPERIMENTAL SECTION
■
General Considerations. All reactions involving air-sensitive
materials were executed in a N2-filled glovebox, on a Schlenk line,
or on a high-vacuum manifold using solvents previously dried by
passage through an alumina column under Ar. HCl (4 M in dioxane, 1
M in Et2O), anhydrous NaCl, concentrated H2SO4, anhydrous 1,4-
dioxane, and 2,6-dimethylphenylisocyanide (CNXy) were obtained
from Sigma-Aldrich, [RhI(CO)2Cl]2, [RhICl(COD)]2 (COD = 1,5-
cyclooctadiene), and PEt3 were purchased from Strem, and O2 was
purchased from Airgas. All commercially available starting materials
were used as received. The complex trans-RhICl(CO)(PEt3)2 (1a) was
prepared by a modified procedure,55 starting with [RhI(CO)2Cl]2 and
using toluene as the solvent. The ligand 1-adamantylisocyanide
(CNAd) was prepared as described previously.56 Elemental analyses
were performed by Midwest Microlab LLC.
Physical Methods. NMR spectra were recorded at the MIT
Department of Chemistry Instrumentation Facility on Varian
Mercury-300 or Inova-500 NMR spectrometers. 31P{1H} NMR
spectra were referenced to an external standard of 85% D3PO4, and
1H spectra were referenced to the residual proteo solvent resonances.
UV−vis spectra were recorded at 295 K in THF solutions in quartz
cuvettes on a Varian Cary 5000 UV−vis−NIR spectrophotometer.
Extinction coefficients were determined over a concentration range
from ∼10−6 to 10−4 M, for which all compounds obeyed Beer’s Law.
For hydride complexes 2b/c, UV−vis spectra were recorded in the
presence of 52 mM HCl, whereas for 4b/4c, spectra were recorded
with 1 atm of O2 present to prevent reversion to 1b/c. Spectral data is
summarized below, and full electronic spectra are presented in the
Supporting Information, Figures S4−S13. IR spectra were recorded on
a PerkinElmer Spectrum 400 FT-IR/FT-FIR spectrometer outfitted
with a Pike Technologies GladiATR attenuated total reflectance
accessory with a monolithic diamond crystal stage and pressure clamp.
Samples were suspended in Nujol for all IR measurements.
(m, 12H), 1.18 (quintet, 18H), −13.16 (dt, 1JRh−H = 16.6 Hz, 2JP−H
=
10.4 Hz, 1H). 31P{1H} NMR (121.5 MHz, THF-d8) δ/ppm: 26.4 (d,
1JRh−P = 81 Hz).
Preparation of cis-trans-RhIIICl2(CNXy)H(PEt3)2 (2b). A 25 mL
Schlenk tube with a Teflon plug seal was charged with 1b (100 mg,
0.198 mmol) dissolved in 3 mL of THF. The solution was freeze−
pump−thaw degassed three times at ∼10−6 Torr. Anhydrous HCl,
generated by dropping concentrated H2SO4 onto anhydrous NaCl (58
mg, 0.99 mmol, 5.0 equiv), was vacuum transferred onto the still-
frozen solution. The vessel was pumped down to ∼10−6 Torr and then
allowed to thaw slowly with stirring. Upon thawing, the now colorless
solution was stirred for 10 min. The volatiles were removed in vacuo,
and the resulting residue was taken back into the glovebox, dissolved
in 4 mL of THF, and transferred to a scintillation vial. Solvent was
removed in vacuo to yield a pale residue, which was suspended in 0.25
mL of toluene. Addition of 4 mL of hexane separated a white solid,
1
which was decanted and dried in vacuo. Yield: 100 mg (93.4%). H
NMR (500 MHz, C6D6) δ/ppm: 6.74 (t, 3JH−H = 7.6 Hz, 1H), 6.64 (d,
3JH−H = 7.6 Hz, 2H), 2.29 (s, 6H), 2.08 (m, 6H), 1.90 (m, 6H), 1.05,
1
2
(quintet, 18H), −14.48 (dt, JRh−H = 18.1 Hz, JP−H = 11.4 Hz, 1H).
31P{1H} NMR (121.5 MHz, C6D6) δ/ppm: 24.7 (d, JRh−P = 86 Hz).
1
UV−vis (THF): λ/nm (ε/M−1 cm−1) 251 (27 000). IR (Nujol): νCN
̃
7193
dx.doi.org/10.1021/ic300279z | Inorg. Chem. 2012, 51, 7192−7201

Inorganic Chemistry
Article
= 2147 cm−1. Anal. Calcd for C21H40Cl2NP2Rh: C, 46.51; H, 7.43; N,
2.58. Found: C, 46.29; H, 7.27; N, 2.41.
glovebox coldwell. They were removed, and upon thawing the PhICl2
solution was added dropwise to the 1c solution, yielding a bright
yellow-orange solution which was warmed to room temperature and
stirred for 30 min. The solution was diluted with 4 mL of hexane and
concentrated in vacuo to afford a yellow-orange residue. The residue
was taken up in 0.25 mL of CH2Cl2, to which was added 4 mL of
hexane to produce a cloudy mixture. Upon concentrating in vacuo to
<2 mL, a yellow-orange solid precipitated. The supernatant was
decanted, and the product was dried in vacuo. Yield: 53 mg (93%). 1H
NMR (500 MHz, C6D6) δ/ppm: 2.25 (m, 12H), 1.72 (br, d, 6H), 1.61
(br, m, 3H), 1.13−1.25 (m, 24H). 31P{1H} NMR (121.5 MHz, C6D6)
Preparation of cis-trans-RhIII(CNAd)Cl2H(PEt3)2 (2c). A 25 mL
Schlenk flask was charged with 1c (100 mg, 0.186 mmol) dissolved in
6 mL of Et2O. After cooling to −78 °C in dry ice/acetone, a solution
of HCl in Et2O (1.03 M, 0.90 mL, 5.0 equiv) was added via syringe.
The reaction mixture was stirred for 10 min before removing the cold
bath. A white precipitate that formed initially redissolved upon
warming to room temperature, yielding a colorless solution that was
stirred for 30 min. The solution was concentrated in vacuo to give a
white solid. In the glovebox, the solid was suspended in 8 mL of Et2O
and transferred to a scintillation vial. Solvent was removed in vacuo,
and the solid was redissolved in 0.5 mL of toluene. With stirring, 6 mL
of hexane was added, freeing a colorless solid, which was decanted and
1
δ/ppm: 14.4 (d, JRh−P = 78 Hz). UV−vis (THF): λ/nm (ε/M−1
cm−1) 222 (25 000), 262 (15 000), 272 (sh) (14 000), 341 (2100),
403 (560). IR (Nujol): νCN
̃
= 2194 cm−1. Anal. Calcd for
C23H45Cl3NP2Rh: C, 45.52; H, 7.47; N, 2.31. Found: C, 45.26; H,
1
dried in vacuo. Yield: 97 mg (91%). H NMR (500 MHz, C6D6) δ/
7.29; N, 2.37.
ppm: 2.14 (m, 6H), 1.98 (m, 6H), 1.65 (br, d, 6H), 1.61 (br, m, 3H),
1.17−1.27 (br, m, 6H), 1.15, (quintet, 18H), −15.09 (dt, 1JRh−H = 17.9
Hz, 2JP−H = 11.6 Hz, 1H). 31P{1H} NMR (121.5 MHz, C6D6) δ/ppm:
Preparation of trans-RhIIICl(CNXy)(η2-O2)(PEt3)2 (4b). A 10 mL
Schlenk flask was charged with 1b (50 mg, 0.099 mmol) dissolved in 4
mL of Et2O. With vigorous stirring, the headspace was purged with O2
for 1 min, and after removing the O2 stream the solution was allowed
to stir for an additional 5 min, leaving a dull yellow-brown solution.
The solvent was removed in vacuo to give a brown solid, which was
redissolved in Et2O and transferred to a scintillation vial in the
glovebox. Solvent was removed again in vacuo, and the resulting solid
was washed with 2 mL of hexane and dried in vacuo briefly (<30 min).
NMR spectra show ca. 7% of 1b, though the microanalytical data and
IR spectrum suggest high purity for the isolated solid. Yield: 45 mg
(85%). 1H NMR (500 MHz, C6D6) δ/ppm: 6.73−6.79 (m, 1H),
6.66−6.69 (m, 2H), 2.35 (s, 6H), 1.90 (m, 6H), 1.72 (m, 6H), 1.09
(quintet, 18H). 31P{1H} NMR (121.5 MHz, C6D6) δ/ppm: 26.3 (d,
1JRh−P = 87 Hz). UV−vis (THF): λ/nm (ε/M−1 cm−1) 243 (40 000).
1
24.1 (d, JRh−P = 88 Hz). UV−vis (THF): λ/nm (ε/M−1 cm−1) 233
(sh) (14 000), 277 (7200). IR (Nujol): νCN
̃
= 2170 cm−1. Anal.
Calcd for C23H46Cl2NP2Rh: C, 48.26; H, 8.10; N, 2.45. Found: C,
48.33; H, 7.81; N, 2.32.
Preparation of trans-RhIII(CO)Cl3(PEt3)2 (3a). In a 20 mL
scintillation vial, 1a (50 mg, 0.12 mmol) was dissolved in 1 mL of
CH2Cl2. Separately, PhICl2 (28.5 mg, 0.104 mmol, 1.05 equiv) was
also dissolved in 1 mL of CH2Cl2. Both solutions were frozen in the
coldwell of the glovebox. They were removed, and upon thawing the
PhICl2 was added dropwise to the stirred solution of 1a, giving a bright
yellow solution which was allowed to warm to room temperature and
stirred for 30 min. At this time, 4 mL of hexane was added, and the
solution was concentrated in vacuo to produce a sticky yellow solid.
Washing the product with 2 mL of hexane at −20 °C gave a yellow
solid, which was dried in vacuo. The spectral data reported here are a
good match for those reported previously.57,58 Yield: 55 mg (93%). 1H
NMR (500 MHz, C6D6) δ/ppm: 2.02 (M, 12H), 0.98 (quintet, 18H).
IR (Nujol): ν
̃
= 2122 cm−1, ν
̃
= 876 cm−1. Anal. Calcd for
CN
O−O
C21H39ClNO2P2Rh: C, 46.90; H, 7.31; N, 2.60. Found: C, 46.81; H,
7.24; N, 2.55.
Preparation of trans-RhIII(CNAd)Cl(η2-O2)(PEt3)2 (4c). A
solution of 1c (50 mg, 0.093 mmol) in 4 mL of Et2O was prepared
in a 10 mL Schlenk flask. The headspace was flushed with O2 for 1 min
with vigorous stirring, and after removing the O2 flow stirring was
continued for an additional 5 min. The resulting yellow-brown
solution was concentrated in vacuo to give an olive green solid. In the
glovebox, the solid was dissolved in Et2O and transferred to a
scintillation vial. Volatiles were removed in vacuo, and the product was
washed with 2 mL of hexane before drying briefly in vacuo. Yield: 49
1
31P{1H} NMR (121.5 MHz, C6D6) δ/ppm: 19.3 (d, JRh−P = 72 Hz).
UV−vis (THF): λ/nm (ε/M−1 cm−1) 295 (20 000), 371 (2000). IR
(Nujol): νCO
̃
= 2059 cm−1. Anal. Calcd for C13H30Cl3OP2Rh: C,
32.97; H, 6.38. Found: C, 32.48; H, 6.16.
Preparation of trans-RhIIICl3(CNXy)(PEt3)2 (3b). A sample of 1b
(50 mg, 0.099 mmol) was dissolved in 1 mL of toluene. In a separate
vial, PhICl2 (28.5 mg, 0.104 mmol, 1.05 equiv) was also dissolved in 1
mL of toluene. Both solutions were frozen in the coldwell of the
glovebox. They were removed, and upon thawing the PhICl2 was
added dropwise to the stirred solution of 1b with a slight darkening in
color observed. The solution was allowed to warm to room
temperature and stirred for a total of 40 min. At this time, the solvent
was removed in vacuo to leave a yellow-orange residue, which was
redissolved in 0.5 mL of CH2Cl2. With stirring, 3 mL of hexane was
added, and the mixture was concentrated to ca. one-half its original
volume, liberating a yellow-orange solid. The supernatant was
decanted, and the product was dried in vacuo. The solid was
redissolved in 2 mL of toluene, and after sitting for 4 days at room
temperature complete conversion from a mixture of trans-
RhIIICl3(CNXy)(PEt3)2 and mer-cis-RhIIICl3(CNXy)(PEt3)2 to the
desired trans product was achieved. Toluene was removed in vacuo to
reveal a yellow solid, which was dissolved in a mixture of 0.5 mL of
CH2Cl2 and 4 mL of hexane. After concentrating in vacuo to <2 mL,
the supernatant was separated from the yellow-orange product, which
1
mg (92%). H NMR (500 MHz, C6D6) δ/ppm: 1.93 (m, 6H), 1.82
(m, 6H), 1.76 (br, d, 6H), 1.64 (br, m, 3H), 1.15−1.28 (m, 24H).
31P{1H} NMR (121.5 MHz, C6D6) δ/ppm: 25.3 (d, JRh−P = 89 Hz).
1
UV−vis (THF): λ/nm (ε/M−1 cm−1) 241 (23 000). IR (Nujol): νCN
̃
= 2135 cm−1, νO−O
̃
= 877 cm−1. Anal. Calcd for C23H45ClNO2P2Rh: C,
48.64; H, 7.99; N, 2.47. Found: C, 48.75; H, 7.75; N, 2.44.
O2 Reduction Reactions of 2a−2c. All O2 reduction reactions
were executed and monitored in a screw-cap NMR tube with a PTFE
septum seal. In all cases, the concentration of the hydride complex was
25 mM at the start of the reaction. Hydride complex 2a (L = CO) was
generated in situ by dissolving a sample of 1a (7.0 mg, 0.017 mmol) in
0.35 mL of 1,4-dioxane and adding 0.35 mL of a 4.13 M solution of
HCl in dioxane. For 2b (L = CNXy) and 2c (L = CNAd), an
appropriate amount of the hydride was dissolved in 1,4-dioxane and
the HCl/dioxane solution was added to produce a total volume of 0.7
mL with the desired concentration of HCl. Alternatively, the hydride
complexes 2b and 2c could be generated in situ from 1b and 1c at no
detriment to the observed reaction. After addition of the hydride
complex and HCl, the headspace of the NMR tube was purged for ∼1
min with O2 at atmospheric pressure. The contents of the tube were
shaken vigorously to ensure complete mixing and periodically mixed
throughout the course of the reactions, which were monitored by
31P{1H} NMR spectroscopy.
1
was dried in vacuo. Yield: 46 mg (81%). H NMR (500 MHz, C6D6)
3
3
δ/ppm: 6.73 (t, JH−H = 7.6 Hz, 1H), 6.63 (d, JH−H = 7.6 Hz, 2H),
2.37 (s, 6H), 2.18 (m, 12H), 1.09 (quintet, 18H). 31P{1H} NMR
(121.5 MHz, C6D6) δ/ppm: 15.4 (d, 1JRh−P = 77 Hz). UV−vis (THF):
λ/nm (ε/M−1 cm−1) 254 (37 000), 345 (2800), 398 (sh) (920). IR
(Nujol): νCN
̃
= 2190 cm−1. Anal. Calcd for C21H39Cl3NP2Rh: C,
43.73; H, 6.82; N, 2.43. Found: C, 43.44; H, 6.50; N, 2.31.
Preparation of trans-RhIII(CNAd)Cl3(PEt3)2 (3c). A 20 mL
scintillation vial was charged with 1c (50 mg, 0.093 mmol), and into
a separate vial was weighed PhICl2 (27 mg, 0.098 mmol, 1.05 equiv).
Both solids were dissolved in 1 mL of CH2Cl2 and frozen in the
Addition of HCl to 4b to Generate 3b/5b. Complex 1b (9.0
mg, 0.018 mmol, 1.0 equiv) was dissolved in 0.7 mL of THF-d8 in a
screw-cap, septum-sealed NMR tube. The headspace of the NMR tube
was purged with O2 (1 atm) and manually shaken to mix, generating a
7194
dx.doi.org/10.1021/ic300279z | Inorg. Chem. 2012, 51, 7192−7201

Inorganic Chemistry
Article
Table 1. Crystallographic Summary for Complexes 2b, 3b, 4b, and 5c
2b
3b
4b
5c•CH2Cl2
formula
fw, g/mol
temp./K
color
C21H40Cl2NP2Rh
542.29
C21H39Cl3NP2Rh
576.73
C21H39ClNO2P2Rh
537.83
C24H47Cl5NOP2Rh
707.73
100(2)
200(2)
100(2)
100(2)
colorless
monoclinic
Cc
yellow
brown
yellow
cryst syst
space group
a/Å
orthorhombic
Pna21
monoclinic
P21/n
monoclinic
P21/n
11.4479(12)
14.6692(15)
15.4302(15)
93.787(2)
2585.6(5)
4
20.1861(8)
10.3567(4)
12.9174(5)
90
15.214(2)
11.4156(18)
15.778(3)
113.796(2)
2507.3(7)
4
14.7676(19)
11.0440(14)
19.617(2)
90.952(2)
3199.0(7)
4
b/Å
c/Å
β/deg
V/Å3
2700.53(18)
4
Z
no. of reflns
no. of unique reflns
Rint
25 388
61 558
57 744
68 331
6906
8130
7671
9363
0.0272
0.0353
0.0324
0.0722
a
R1 (all data)
0.0229
0.0203
0.0246
0.0736
b
wR2 (all data)
0.0465
0.0457
0.0508
0.1624
R1 [(I > 2σ)]
0.0205
0.0183
0.0195
0.0609
wR2 [(I > 2σ)]
0.0452
0.0442
0.0472
0.1564
c
GOF
1.042
1.075
1.042
1.220
Flack-x param
−0.030(14)
0.638(13)
c
a
b
R1 = Σ||Fo − |Fc||/Σ|Fo|. wR2 = (Σ(w(Fo2 − Fc2)2)/Σ(w(Fo )2))1/2. GOF = (Σw(Fo2 − Fc2)2/(n − p))1/2, where n is the number of data and p is the
number of parameters refined.
2
dull yellow solution of 4b. A solution of HCl in dioxane (4.2 M, 13 μL,
0.054 mmol, 3.0 equiv) was added via syringe, resulting in an
immediate color change to bright yellow. A 31P{1H} NMR spectrum
recorded immediately after addition of HCl showed a mixture of 5b
(80%), 3b (17%), and ∼3% of unidentified side products. The solution
was transferred to a scintillation vial and concentrated in vacuo, and
the resulting residue was washed with hexane and dried in vacuo.
Spectral data for 5b: 1H NMR (500 MHz, CD3CN) δ/ppm: ∼7.3 (m,
1H, overlapped with 3b), 7.22 (m, 2H, overlapped with 3b), 5.92 (br,
s, 2H), 2.58 (s, 6H), 2.17 (m, 12H), 1.18 (quintet, 18H). 31P{1H}
methods in SHELXS and refined by standard difference Fourier
techniques in the SHELXTL program suite (6.10 v., Sheldrick G. M.,
Siemens Industrial Automation, 2000). Hydrogen atoms bonded to
carbon were placed in calculated positions using the standard riding
model and refined isotropically; all non-hydrogen atoms were refined
anisotropically. In the structure of 2b, the rhodium-bound hydrogen
atom was tentatively located in the difference map and refined
isotropically. The O−H hydrogen atoms in the structure of 5c were
also located in the difference map; they were restrained to a distance of
0.84 Å from the oxygen atom and refined isotropically. The structure
of 3b was refined as a racemic twin. In the structure of 5c, the
adamantyl group, one of the ethyl groups, and a solvent dichloro-
methane molecule were all modeled as two-part positional disorders.
The corresponding 1−2 and 1−3 distances of all disordered parts were
restrained to be identical, and rigid bond restraints were used on all
disordered atoms. Crystallographic details for the structures of 2b, 3b,
4b, and 5c are summarized in Table 1.
1
NMR (121.5 MHz, CD3CN) δ/ppm: 19.1 (d, JRh−P = 74 Hz).
Addition of HCl to 4c to Generate 3c/5c. A solution of 1c (9.0
mg, 0.017 mmol, 1.0 equiv) in 0.7 mL of THF-d8 was prepared in a
screw-cap NMR tube with septum seal. The headspace was purged
with 1 atm of O2 for 1 min, and upon shaking the contents changed to
a dull olive color, which was shown to be complex 4c. Addition of an
HCl/dioxane solution (4.2 M, 12 μL, 0.050 mmol, 2.9 equiv)
produced a bright yellow solution. Alternatively, the reaction can be
executed starting with isolated 4c with no additional O2, and analysis
of the headspace gases by gas chromatography shows no gaseous
products after HCl addition. At this stage, 31P{1H} NMR indicates a
mixture of 5b (80%) and 3b (20%). The solution was concentrated in
vacuo, the product triturated with Et2O/hexane, and the resulting
yellow solid dried in vacuo. Spectral data for 5c: 1H NMR (500 MHz,
C6D6) δ/ppm 6.58 (br, s, 2H), ∼2.22−2.41 (m, 12H, overlapped with
3c), 2.21 (br, s, 6H), 1.74 (br, s, 3H), 1.14−1.39 (m, 24H, overlap
RESULTS
■
Hydride Complexes. Hydride complexes cis-trans-
RhIIICl2H(L)(PEt3)2 (2a−c) were prepared by HCl addition
to RhI complexes 1a−c, as depicted in Scheme 1. Treatment of
1a with a large excess of HCl results in growth of new NMR
Scheme 1
with 3c). 31P{1H} NMR (121.5 MHz, C6D6) δ/ppm: 17.4 (d, 1JRh−P
=
76 Hz).
X-ray Crystallographic Details. Single crystals of 2b were
obtained by cooling a saturated toluene/hexane solution to −20 °C,
3b and 5c crystallized from saturated CH2Cl2/hexane solutions at −20
°C, and crystals of 4b formed by allowing O2 to slowly diffuse into a
hexane solution of 1b at room temperature. Crystals of 3b, 4b, and 5c
were mounted on a Bruker three-circle goniometer platform equipped
with an APEX detector, whereas crystals of 2b were mounted on a
Bruker four-circle goniometer platform with an APEX 2 detector. A
graphite monochromator was employed for wavelength selection of
the Mo Kα radiation (λ = 0.71073 Å). Data were processed and
refined using the program SAINT supplied by Siemens Industrial
Automation. Structures were solved by Patterson methods or direct
7195
dx.doi.org/10.1021/ic300279z | Inorg. Chem. 2012, 51, 7192−7201

Inorganic Chemistry
Article
features attributable to 2a. HCl addition to 1a to form 2a is
reversible,59 such that removal of HCl from 2a results in
complete reversion to 1a. Complexes 2b and 2c featuring
isocyanide ligands have not been previously described, though
an isolated example of a close relative prepared by hydride
transfer from an alcoholic reaction medium is known.60
Hydride complexes 2b and 2c were found to form
quantitatively upon treatment of 1b and 1c with HCl.
Furthermore, these complexes are isolable and can be obtained
in pure form as colorless solids. Characteristic NMR features,
similar to those described for 2a, are observed for 2b and 2c as
well and are summarized in the Experimental Section.
with the known pKas of 2,6-lutidinium (14.13)62 and HCl
(8.9)63 in acetonitrile,51,64 the KHCl values were determined and
are tabulated in Scheme 1. Though the measured value of KHCl
will somewhat be dependent on solvent choice, the disparate
solvent media for addition reactions of 1a and 1b/c cannot
account for the dramatic changes in KHCl that occur upon
ligand substitution.
ORR Activity. Hydride complexes 2a−c react with O2, in
the presence of additional HCl, as summarized in Scheme 2;
Scheme 2
The stereochemistry of 2b was confirmed by single-crystal X-
ray diffraction. The structure of the complex is shown in Figure
1. The geometry of the RhIII center is octahedral with a trans
though the O2 stoichiometry is ill defined (vide infra), the only
two products readily identified by spectroscopic methods, 3a−c
and water, are quantifiable. In all cases, the final outcome of the
reaction between the hydride complex and O2 and HCl is the
respective trans-RhIIICl3(L)(PEt3)2, 3a−c, which is produced
cleanly. Products 3a−c were prepared independently by PhICl2
oxidation of RhI complexes 1a−c, allowing the identities of the
final products formed in the reactions of Scheme 2 to be
ascertained unequivocally. Complexes 3a−c possess similar
31P{1H} spectral features, and together with the X-ray crystal
structure results of Figure 2, a trans arrangement of the PEt3
ligands and the corresponding meridional arrangement of Cl−
ligands in these products are established.
Figure 1. Thermal ellipsoid plot of 2b, shown at the 50% probability
level. All carbon-bound hydrogen atoms are omitted for clarity. Data
were collected at 100(2) K.
arrangement of the two PEt3 ligands and a cis arrangement of
the two Cl− ligands. Of note is the disparity of the two Rh−Cl
bond distances. The strong trans influence of the hydride ligand
causes the Rh(1)−Cl(2) distance [2.4947(5) Å] to be
substantially longer than the Rh(1)−Cl(1) distance
[2.4041(5) Å]. The stereochemistry of 2b is consistent with
the known stereochemical preference for HCl addition to
square-planar MI centers,61 and this precedent, coupled with
the similarities in the NMR features, led us to conclude that 2a
and 2c, which were not crystallographically characterized, are
isostructural with 2b.
Addition of HCl to 1a−c establishes the equilibrium shown
in Scheme 1. The equilibrium constant for the reaction (KHCl
=
[2]/[1][HCl]) was determined directly by addition of a known
concentration of HCl to a 1,4-dioxane solution of 1a.
Integration of the 31P{1H} NMR spectrum furnished the
equilibrium concentrations of 1a and 2a. The equilibrium
constants for the 1b/c to 2b/c conversions were too large to be
easily determined by direct addition of HCl to solutions of the
starting complexes. Thus, a thermodynamic cycle connecting
HCl and the weaker acid, 2,6-lutinidium hydrochloride, was
constructed. These studies were conducted in acetonitrile,
owing to the considerable solubility of protonated amine bases
and ready availability of the relevant thermodynamic data in
that solvent. Complex 1a is completely unreactive to 2,6-
lutidinium hydrochloride, so it was not possible to confirm its
KHCl value by this method. By using 2,6-lutinidium hydro-
chloride as the acid source in acetonitrile, an equilibrium
between the RhI complex (1b/c) and RhIII hydride (2b/c) was
established. By coupling this measured equilibrium constant
Figure 2. Thermal ellipsoid plot of 3b, shown at the 50% probability
level. All hydrogen atoms are omitted for clarity. Data were collected at
200(2) K.
The reaction progression of hydride 2a significantly deviates
from that of 2b and 2c. Figure 3 shows the evolution of the
31P{1H} NMR spectra upon treatment of 2a with HCl (2.1 M)
and O2 (ca. 0.2 atm) in 1,4-dioxane at room temperature. Over
a long time course, the resonance attributed to 2a (27.0 ppm,
1JRh−P = 80 Hz) gives rise exclusively to the resonance for 3a
1
(19.9 ppm, JRh−P = 72 Hz). In parallel to the dirhodium
system,49,51 no intermediates are spectroscopically observed
along the ORR conversion.
7196
dx.doi.org/10.1021/ic300279z | Inorg. Chem. 2012, 51, 7192−7201

Inorganic Chemistry
Article
observed in comparable amounts. As detailed below, an
alternate reaction strategy furnishes this intermediate in much
higher yields, ultimately allowing us to identify it as a RhIII aqua
complex.
Peroxy Intermediates. Addition of 1 atm of O2 to a
solution of 1a produces only a small amount of a new product,
4a, with mostly 1a remaining. Product 4a, which forms in <5%
yield under these conditions, shows a slightly downfield-shifted
1
31P{1H} resonance (30.0 ppm) with a smaller JRh−P coupling
constant (82 Hz) relative to that of 1a. Over a period of 1 week,
the compound decomposes and OPEt3 and an intractable
mixture of metal-containing products is observed. In contrast,
treatment of 1b/c with O2 at 1 atm leads to clean, quantitative
formation of trans-RhIIICl(L)(η2-O2)(PEt3)2 (4b, L = CNXy;
4c, L = CNAd) as illustrated in Scheme 3. Though these
Figure 3. Temporal evolution of the 31P{1H} NMR spectra when 2a,
in the presence of 2.1 M HCl, is treated with ca. 0.2 atm of O2 at room
temperature. Spectra were collected at the time intervals depicted on
the right of the plot. Resonances attributed to 2a and 3a are labeled.
Scheme 3
As shown in Figure 4, conversion of 2c in the presence of
HCl (77 mM) and O2 (1 atm) in 1,4-dioxane to 3c is also slow.
complexes suffer from limited solution stability, again
decomposing to OPEt3 and a mixture of rhodium-containing
products, the peroxo complexes can be isolated in pure form in
the solid state and do persist long enough in solution to
characterize their spectral properties and reactivity. Isolated
solids of 4b/4c give rise to nearly identical O−O bond-
stretching frequencies in their infrared spectra (νO−O = 876
̃
cm−1 for 4b and 877 cm−1 for 4c); these values are suggestive
of an O−O single bond.65 Figures S14 and S15, Supporting
Information, collect overlaid partial IR spectra for 1b/c−4b/c,
which show that the region of 800−1000 cm−1 is essentially
transparent for all other complexes, allowing the O−O
stretching frequency for 4b/c to be easily identified. The
assignments of these stretching frequencies are further validated
by their similarity to some well-characterized palladium66,67 and
rhodium68,69 peroxo complexes and their good correlation with
the observed O−O bond length,70 which is given below.
The X-ray crystal structure of 4b, shown in Figure 5,
confirms that O2 binds in a η2-peroxo motif. The geometry
about the RhIII center approximates trigonal bipyramidal if the
midpoint of the O−O vector is taken to occupy a coordination
site. As established from the 31P{1H} NMR spectrum the two
Figure 4. Time-resolved 31P{1H} NMR spectra when 2c, in the
presence of 77 mM HCl, is treated with 1 atm of O2 at room
temperature. Spectra were collected at the time intervals depicted on
the right of the plot. Resonances attributed to 2c and 3c are labeled.
However, growth of an intermediate species in the 31P{1H}
NMR spectrum is clearly observed. Whereas a 31P{1H} NMR
1
doublet of 2c (24.2 ppm, JRh−P = 87 Hz) is observed at early
times and of 3c (14.5 ppm, 1JRh−P = 78 Hz) is observed at late
times, a unique species, with a chemical shift of 17.4 ppm and
1JRh−P = 75 Hz, is apparent at intermediate times. Under
identical conditions a very similar outcome is observed for 2b;
as shown in Figure S1, Supporting Information, an intermediate
species appears during conversion of 2b to 3b. Figures S2 and
S3, Supporting Information, demonstrate qualitatively the effect
of augmenting the acid concentration to 250 mM on the overall
reaction progress for 2b and 2c, respectively. The observations
are generally similar, though in the case of 2b increasing [HCl]
results in a much smaller amount of the intermediate species
along the path to forming 3b; after the first 2 h of conversion
only 2b and 3b are present. In addition, the time required to
completely consume hydride 2b is substantially longer at higher
acid concentrations. For 2c, increasing the concentration of
HCl also lengthens the reaction time but the intermediate is
Figure 5. Thermal ellipsoid plot of 4b, shown at the 50% probability
level. All hydrogen atoms are omitted for clarity. Data were collected at
100(2) K.
7197
dx.doi.org/10.1021/ic300279z | Inorg. Chem. 2012, 51, 7192−7201

Inorganic Chemistry
Article
PEt3 ligands remain trans to one another, occupying the apical
positions of the trigonal bipyramid. The O(1)−O(2)
internuclear distance of 1.4413(12) Å is consistent with the
formulation of an O22− ligand with an O−O single bond.40,51,70
Although 4c was not structurally characterized, its nearly
identical spectral features to those of 4b point to an analogous
structure for 4c.
With synthetic routes to 4b and 4c in hand, we interrogated
the reactivity of the compounds with HCl; identical outcomes
are obtained whether or not additional O2 is present. This
reactivity is summarized in Scheme 4. Whereas treatment of 4b
additional singlet that integrates to two protons at 5.92 (5b,
CD3CN) and 6.58 ppm (5c, C6D6). With HCl present, these
resonances are broadened considerably and coalesce with the
resonance for free HCl. The positions of these new resonances,
their slightly broadened line shapes, and their rapid exchange
with HCl on the NMR time scale suggest O−H protons. We
also verified that 5c cleanly converts to 3c in C6D6 in the
absence of HCl or O2. Over a period of ca. 24 h, the NMR
features of 5c disappear, with concomitant growth of the peaks
for 3c as well as a broad singlet at 0.60 ppm, which we attribute
to free H2O. The broadening of this water peak and the slight
71
shift from the 0.40 ppm chemical shift of H2O in C6D6
indicate a weak interaction of the liberated water with 3c.
All spectroscopic results were confirmed with the solution of
the single-crystal X-ray structure of 5c. Figure 6 shows two
Scheme 4
and 4c with a single equivalent of HCl produces an intractable
mixture of RhIII products, use of excess HCl cleanly generates a
mixture of two products. After 20 min, the major product (ca.
80%) is the same intermediate observed above in the O2
reduction experiments with the balance accounted for by 3b/c.
1
31P{1H} and H NMR spectra indicate that these are the only
two diamagnetic products containing phosphorus and/or
hydrogen nuclei. Moreover, the IR spectra of solid product
(vide infra), isolated without any purification, shows only
features attributed to these two diamagnetic products,
discounting the presence of substantial paramagnetic impur-
ities. The intermediate gradually converts to 3b/c over time, in
line with the observations in the previous section for the O2
reduction experiments, where 3b/c are formed as the exclusive
products when the aqua intermediate disappears. Gas
chromatographic analysis of the headspace gas, immediately
following HCl addition to anaerobic solutions of 4b/c, shows
no evidence for gaseous products formed during this reaction,
indicating that O2 is not a reaction product. As such, the fate of
the second oxygen atom in the peroxo remains unclear, but
mechanisms which could result in production of 1/2 O2 can be
ruled out.
Figure 6. Thermal ellipsoid plots of 5c, shown at the 50% probability
level. All carbon-bound hydrogen atoms and solvent molecules are
omitted for clarity. Data were collected at 100(2) K. (a) Cation is
shown with the outer-sphere Cl− omitted. (b) Counterion is included
as well as a second molecule of 5c generated by a crystallographic
inversion center. Dashed lines indicate located hydrogen bonds, and
atoms labeled with an asterisk (*) are symmetry equivalents of those
with conventional labels.
Having established a route to prepare the intermediate in
appreciable quantities, interrogation of the compound with a
variety of experimental methods leads to its unequivocal
identification as the aqua complex, cis-trans-[RhIII(OH2)Cl2(L)-
(PEt3)2]Cl (5b, L = CNXy; 5c, L = CNAd). The IR spectrum
of an isolated solid mixture of 5b/c and 3b/c shows only
stretches attributed to the PEt3 and isocyanide ligands. The
region between 800 and 1000 cm−1 is notably barren,
suggesting the absence of an O−O bond in 5b/c and
discounting the formulation of these intermediates as hydro-
views of the structure of 5c. In Figure 6a, the cation cis-trans-
[RhIII(CNAd)(OH2)Cl2(PEt3)2]+ is depicted, which confirms
the stereochemistry about the RhIII center, where a trans
arrangement of the two PEt3 ligands and a cis arrangement of
the two Cl− ligands persists. In Figure 6b, the structure is
extended to show the outer-sphere Cl− counterion as well as a
neighboring molecule of 5c that is related by a crystallographic
inversion center. This view clearly shows a series of hydrogen-
bonding interactions between the aqua protons and the outer-
sphere Cl− anions that stabilize the structure and give rise to a
dimeric motif in the solid state. The hydrogen-bonding donor−
1
peroxo complexes. Furthermore, in the absence of HCl the H
NMR spectra of 5b/c, in addition to the expected resonances
arising from PEt3 and the respective isocyanide, show an
7198
dx.doi.org/10.1021/ic300279z | Inorg. Chem. 2012, 51, 7192−7201

Inorganic Chemistry
Article
acceptor distance between O(1) and Cl(1s) is 2.948(3) Å,
whereas the distance is 3.056(3) Å between O(1) and the
symmetry-generated equivalent of Cl(1s). This reveals a slight
asymmetry in the two crystallographically independent hydro-
gen-bonding interactions.
high concentrations of HCl are required to ensure 2a as the
majority species, so meaningful comparisons of qualitative
reaction rates cannot be easily made for this complex. However,
hydrides 2b/c do react under identical conditions to those
investigated for the dirhodium hydrides, and reaction times of
∼4−5 days are required, as compared to a few hours for
dirhodium complexes. These observations suggest a possible
benefit of bimetallic cooperativity for ORR promoted by late
transition metal systems, but the comparison is muddled by
ligand electronic effects, so additional studies of detailed
reaction kinetics will be required to fully understand bimetallic
effects in this chemistry.
In summary, treatment of 4b/c with HCl results in rapid
cleavage of the O−O bond and instantaneous generation of
aqua intermediate 5b/c, which liberates H2O and forms 3b/c
irrespective of the presence of O2 and HCl. It is not definitive at
this stage whether conversion of peroxo complex 4b/c to
RhIIICl3 complex 3b/c necessarily proceeds through the aqua
intermediate 5b/c or if direct conversion is possible (dashed
arrow in Scheme 4). It seems likely that such a direct pathway
occurs, since there is a ca. 20% population of 3b/c immediately
after HCl addition, and conversion of 5b/c to 3b/c requires
∼24 h, suggesting there is a route to 3b/c that does not require
intermediacy of aqua complexes 5b/c. Ultimately, detailed
reaction kinetics will be turned to to distinguish these
possibilities.
As η2-peroxo species are keystones to ORR, especially for
reactions proceeding by a HXRE mechanism, peroxo
complexes 4b/c were independently prepared by pressurizing
1b/c with 1 atm of O2, and their reactivity was investigated
with HCl under both aerobic and anaerobic conditions.
Analogous reactions have been interrogated on a series of
closely related rhodium peroxo complexes,68 though in these
previous examples the hydroperoxo was stable enough to be
observed and crystallized below room temperature, and further
reaction with HCl liberated hydrogen peroxide and not water.
These studies are especially valuable inasmuch as the
dirhodium peroxo species are not synthetically accessible,51
and thus, examination of peroxo reactivity in the context of this
original system has not been feasible. The peroxo complexes
described here react rapidly with excess HCl, forming a mixture
of aqua complexes 5b/c and RhIIICl3 complexes 3b/c before
final conversion to 3b/c. Even at early time points, 3b/c and
5b/c are the only two species observed spectroscopically,
indicating facile cleavage of the O−O bond under these
conditions. It is reasonable to assume that a hydroperoxo
intermediate forms initially upon reaction with HCl, given the
reactivity of other late metal peroxo systems.51,68,72,73 However,
in the case of the rhodium complexes described here, such
hydroperoxo complexes are apparently too unstable to observe,
as even a single equivalent of HCl added to 4b/c fails to
produce any spectroscopic features that can be confidently
assigned to a hydroperoxo complex. Several key details of this
O−O bond-cleaving transformation remain unclear, but some
possibilities can be ruled out. Notably, O2 gas is not produced
upon HCl-induced oxygen−oxygen bond scission of 4b/c. This
observation rules out a pathway in which H2O2, liberated by
protonolysis, disproportionates into H2O and 1/2 O2 and also
eliminates any other possibility, such as a bimolecular pathway,
which would result in liberating 1/2 O2. One possibility that
cannot be ruled out is that protonation of a putative
hydroperoxo releases water and leaves an unstable high-valent
oxo species which rapidly decomposes, presumably by reaction
with solvent. The nature of this decomposition is not apparent
from any of the available data, though rhodium-containing
products other than 3b/c and 5b/c are not observed given the
quantitative yields for the overall reactions. Such reactivity with
acids has been characterized for some other synthetic metal−
hydroperoxo complexes, where a high-valent oxo complex is
likewise observed or inferred upon acid-induced O−O bond
heterolysis.74−76 Notwithstanding, the reactivity of peroxo
complexes 4b/c with HCl, which generates water and trichloro
complexes 3b/c as the final outcome, bespeaks to the
plausibility of an HXRE mechanism for ORR by hydride
complexes 2b/c. Forthcoming kinetic studies on these and
related complexes will reveal whether HXRE is the preferred
mechanism for the reactivity of hydrides 2b/c or if other
DISCUSSION
■
Alteration of a single neutral donor ligand can have a profound
effect on O2 activation and reduction by late metal molecular
complexes. The dramatic effect of the ligand environment is
manifested most profoundly in the equilibrium constant for
HCl addition, KHCl, an important parameter for O2 activation,
particularly when a HXRE mechanism is operative.44−46,51 The
strongly π-acidic CO-supporting ligand of 1a renders the RhI
center comparatively electron poor, and a KHCl of 9.6 is
observed. The small value of KHCl precludes isolation of hydride
complex 2a, and in the absence of HCl, reversion to 1a occurs
with facility. Upon changing L to substituted isocyanides, HCl
addition is strongly favored (see Scheme 1) and isolation of
hydrides 2b/c is readily achieved. Similar trends are observed
for addition of O2 to 1a−c to generate peroxo complexes 4a−c.
For 1a, addition of 1 atm of O2 generates only a small amount
of the presumed peroxo complex 4a, whereas quantitative
formation of 4b/c is observed under identical conditions. The
thermodynamic metrics for HCl addition, coupled with the
qualitative observations on O2 addition, readily demonstrate
that the electronic environment of the rhodium center can be
significantly altered by simple ligand substitution.
The O2-reduction chemistry of cis-trans-RhIIICl2H(L)(PEt3)2
(2a−c) is quite general. In all cases, treatment of the hydride
with excess HCl and O2 leads to smooth conversion to trans-
RhIIICl3(L)(PEt3)2 (3a−c) with concomitant formation of
H2O, monometallic analogues to previously reported bimetallic
systems.49 Though an intermediate is not observed for 2a by
traditional spectroscopic techniques, the aqua complex
[RhIII(OH2)Cl2(L)(PEt3)2]Cl (5b/c) appears during the
course of the reaction of 2b/c to 3b/c. Identification of 5b/c
demonstrates that O2 is being reduced to water in these
systems, though it is unclear at this stage if there is a
mechanistic significance of the bound aqua complex, which has
not been detected for the ORR of bimetallic systems. Also
worth noting is that the absence of the intermediate’s
appearance in the aerobic conversion of 2a to 3a may not
have mechanistic significance, given the drastically different
conditions (much higher HCl concentration) required to study
complex 2a.
The reaction times for O2 reduction by the monometallic
systems described here are substantially longer than those
required for bimetallic complexes. For hydride 2a (L = CO),
7199
dx.doi.org/10.1021/ic300279z | Inorg. Chem. 2012, 51, 7192−7201

Inorganic Chemistry
Article
(17) Halime, Z.; Kotani, H.; Li, Y.; Fukuzumi, S.; Karlin, K. D. Proc.
Natl. Acad. Sci. U.S.A. 2011, 108, 13990−13994.
(18) Stahl, S. S. Angew. Chem., Int. Ed. 2004, 43, 3400−3420.
(19) Punniyamurthy, T.; Velusamy, S.; Iqbal, J. Chem. Rev. 2005, 105,
2329−2363.
mechanism(s) prevail. These results set the stage for a
thorough and systematic mechanistic interrogation of O2
activation and reduction by monometallic rhodium hydride
complexes.
To conclude, monometallic RhIII hydride complexes, in the
presence of HCl, quantitatively reduce one-half of an equivalent
of O2 to water. The chemistry of independently prepared
peroxo congeners points to HXRE as the mechanism of ORR.
Mechanistic studies underway to define the precise mechanism
of ORR are bolstered by the ease with which O2 reactivity can
be tuned by ligand design of the monometallic rhodium
platform.
(20) Kunisu, T.; Oguma, T.; Katsuki, T. J. Am. Chem. Soc. 2011, 133,
12937−12939.
(21) Frazier, C. P.; Engelking, J. R.; de Alaniz, J. R. J. Am. Chem. Soc.
2011, 133, 10430−10433.
(22) Liu, L.; Yu, M.; Wayland, B. B.; Fu, X. Chem. Commun. 2010, 46,
6353−6355.
(23) Li, H.; Wei, W.; Xu, Y.; Zhang, C.; Wan, X. Chem. Commun.
2011, 47, 1497−1499.
(24) Fries, P.; Halter, D.; Kleinschek, A.; Hartung, J. J. Am. Chem. Soc.
2011, 133, 3906−3912.
ASSOCIATED CONTENT
■
(25) Jiang, B.; Feng, Y.; Ison, E. A. J. Am. Chem. Soc. 2008, 130,
14462−14464.
S
* Supporting Information
31P{1H} NMR spectra for aerobic conversion of 2b/c and HCl
to 3b/c, electronic spectra for complexes 1a−c, 2b/c, 3a−c,
and 4b/c, partial IR spectra for 1b/c−4b/c, and crystallo-
graphic information file (CIF). This material is available free of
charge via the Internet at http://pubs.acs.org.
(26) Izawa, Y.; Pun, D.; Stahl, S. S. Science 2011, 333, 209−213.
(27) Chuang, G. J.; Wang, W.; Lee, E.; Ritter, T. J. Am. Chem. Soc.
2011, 133, 1760−1762.
(28) Zhang, J.; Khaskin, E.; Anderson, N. P.; Zavalij, P. Y.;
Vedernikov, A. N. Chem. Commun. 2008, 3625−3627.
(29) Roberts, H. L.; Symes, W. R. J. Chem. Soc. A 1968, 1450−1453.
(30) Johnston, L. E.; Page, J. A. Can. J. Chem. 1969, 47, 4241−4246.
(31) Gillard, R. D.; Heaton, B. T.; Vaughan, D. H. J. Chem. Soc. A
1970, 3126−3130.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: nocera@mit.edu.
(32) Endicott, J. F.; Wong, C.-L.; Inoue, T.; Natarajan, P. Inorg.
Chem. 1979, 18, 450−454.
Notes
(33) Atlay, M. T.; Preece, M.; Strukul, G.; James, B. R. J. Chem. Soc.,
Chem. Commun. 1982, 406−407.
The authors declare no competing financial interest.
(34) Atlay, M. T.; Preece, M.; Strukul, G.; James, B. R. Can. J. Chem.
1983, 61, 1332−1338.
ACKNOWLEDGMENTS
■
This work was supported by NSF grant CHE-1112154. Grants
from the NSF also supported the MIT Department of
Chemistry Instrumentation Facility (CHE-9808061 and DBI-
9729592). T.S.T. acknowledges the Fannie and John Hertz
Foundation for a graduate research fellowship.
(35) Bakac, A. J. Am. Chem. Soc. 1997, 119, 10726−10731.
(36) Thyagarajan, S.; Incarvito, C. D.; Rheingold, A. L.; Theopold, K.
H. Chem. Commun. 2001, 2198−2199.
(37) Cui, W.; Wayland, B. B. J. Am. Chem. Soc. 2006, 128, 10350−
10351.
(38) Wick, D. D.; Goldberg, K. I. J. Am. Chem. Soc. 1999, 121,
11900−11901.
REFERENCES
■
(39) Konnick, M. M.; Gandhi, B. A.; Guzei, I. A.; Stahl, S. S. Angew.
Chem., Int. Ed. 2006, 45, 2904−2907.
(1) Barber, J. Chem. Soc. Rev. 2009, 38, 185−196.
(2) Barber, J. Philos. Trans. R. Soc. A 2007, 365, 1007−1023.
(3) Cook, T. R.; Dogutan, D. K.; Reece, S. Y.; Surendranath, Y.;
Teets, T. S.; Nocera, D. G. Chem. Rev. 2010, 110, 6474−6502.
(4) Lewis, N. S.; Nocera, D. G. Proc. Natl. Acad. Sci. U.S.A. 2006, 103,
15729−15735.
(40) Denney, M. C.; Smythe, N. A.; Cetto, K. L.; Kemp, R. A.;
Goldberg, K. I. J. Am. Chem. Soc. 2006, 128, 2508−2509.
(41) Szajna-Fuller, E.; Bakac, A. Inorg. Chem. 2010, 49, 781−785.
(42) Keith, J. M.; Nielson, R. J.; Oxgaard, J.; Goddard, W. A., III J.
Am. Chem. Soc. 2005, 127, 13172−13179.
(43) Keith, J. M.; Muller, R. P.; Kemp, R. A.; Goldberg, K. I.;
Goddard, W. A., III; Oxgaard, J. Inorg. Chem. 2006, 45, 9631−9633.
(44) Popp, B. V.; Stahl, S. S. J. Am. Chem. Soc. 2007, 129, 4410−
4422.
(45) Konnick, M. M.; Stahl, S. S. J. Am. Chem. Soc. 2008, 130, 5753−
5762.
(46) Konnick, M. M.; Decharin, N.; Popp, B. V.; Stahl, S. S. Chem. Sci
2011, 2, 326−330.
(47) Decharin, N.; Stahl, S. S. J. Am. Chem. Soc. 2011, 133, 5732−
5735.
(5) Kaila, V. R. I.; Verkhovsky, M. I.; Wikstrom, M. Chem. Rev. 2010,
̈
110, 7062−7081.
(6) Baik, M.-H.; Newcomb, M.; Friesner, R. A.; Lippard, S. J. Chem.
Rev. 2003, 103, 2385−2419.
(7) Himes, R. A.; Barnese, K.; Karlin, K. D. Angew. Chem., Int. Ed.
2010, 49, 6714−6716.
(8) Collman, J. P.; Boulatov, R.; Sunderland, C. J.; Fu, L. Chem. Rev.
2004, 104, 561−588.
(9) Rosenthal, J.; Nocera, D. G. Acc. Chem. Res. 2007, 40, 543−553.
(10) Rosenthal, J.; Nocera, D. G. Prog. Inorg. Chem. 2007, 55, 483−
544.
(48) Decharin, N.; Popp, B. V.; Stahl, S. S. J. Am. Chem. Soc. 2011,
(11) McGuire, R., Jr.; Dogutan, D. K.; Teets, T. S.; Suntivich, J.;
Shao-Horn, Y.; Nocera, D. G. Chem. Sci. 2010, 1, 411−414.
(12) Dogutan, D. K.; Stoian, S. A.; McGuire, R., Jr.; Schwalbe, M.;
Teets, T. S.; Nocera, D. G. J. Am. Chem. Soc. 2011, 133, 131−140.
(13) Chang, C. J.; Deng, Y.; Shi, C.; Chang, C. K.; Anson, F. C.;
Nocera, D. G. Chem. Commun. 2000, 1355−1356.
(14) Chang, C. J.; Loh, Z.-H.; Shi, C.; Anson, F. C.; Nocera, D. G. J.
Am. Chem. Soc. 2004, 126, 10013−10020.
(15) Collman, J. P.; Wagenknecht, P. S.; Hutchison, J. E. Angew.
Chem., Int. Ed. 1994, 33, 1537−1554.
(16) Collman, J. P.; Fu, L.; Hermann, P. C.; Zhang, X. Science 1997,
275, 949−951.
133, 13268−13271.
(49) Teets, T. S.; Cook, T. R.; McCarthy, B. D.; Nocera, D. G. J. Am.
Chem. Soc. 2011, 133, 8114−8117.
(50) Teets, T. S.; Cook, T. R.; McCarthy, B. D.; Nocera, D. G. Inorg.
Chem. 2011, 50, 5223−5233.
(51) Teets, T. S.; Nocera, D. G. J. Am. Chem. Soc. 2011, 133, 17796−
17806.
(52) Keith, J. M; Teets, T. S.; Nocera, D. G. Submitted for
publication.
(53) Look, J. L.; Wick, D. D.; Mayer, J. M.; Goldberg, K. I. Inorg.
Chem. 2009, 48, 1356−1369.
7200
dx.doi.org/10.1021/ic300279z | Inorg. Chem. 2012, 51, 7192−7201

Inorganic Chemistry
Article
(54) Gray, T. G.; Veige, A. S.; Nocera, D. G. J. Am. Chem. Soc. 2004,
126, 9760−9768.
(55) Chatt, J.; Shaw, B. L. J. Chem. Soc. A 1966, 1437−1442.
(56) Sasaki, T.; Nakanishi, A.; Ohno, M. J. Org. Chem. 1981, 46,
5445−5447.
(57) Al-Jibori, S.; Crocker, C.; Shaw, B. L. J. Chem. Soc., Dalton Trans.
1981, 319−321.
(58) Intille, G. M. Inorg. Chem. 1972, 11, 695−702.
(59) Conkie, A.; Ebsworth, E. A. V.; Mayo, R. A.; Moreton, S. J.
Chem. Soc., Dalton Trans. 1992, 2951−2954.
(60) Masters, C.; Shaw, B. L. J. Chem. Soc. A 1971, 3679−3686.
(61) Vaska, L. J. Am. Chem. Soc. 1966, 88, 5325−5327.
(62) Kaljurand, I.; Kutt, A.; Soovali, L.; Rodima, T.; Maemets, V.;
̈
̈
̈
Leito, I.; Koppel, I. A. J. Org. Chem. 2005, 70, 1019−1028.
(63) Izutsu, K. Electrochemistry in Nonaqueous Solutions; Wiley-VCH:
Weinheim, 2009; Chapter 3.
(64) Wilson, A. D.; Miller, A. J. M.; DuBois, D. L.; Labinger, J. A.;
Bercaw, J. E. Inorg. Chem. 2010, 49, 3918−3926.
(65) Socrates, G. Peroxides and Hydroperoxides: −O−O−Group.
Infrared and Raman Characteristic Group Frequencies: Tables and
Charts, 3rd ed.; Wiley and Sons: Chichester, U.K., 2001; pp 105−106.
(66) Stahl, S. S.; Thorman, J. L.; Nelson, R. C.; Kozee, M. A. J. Am.
Chem. Soc. 2001, 123, 7188−7189.
(67) Konnick, M. M.; Guzei, I. A.; Stahl, S. S. J. Am. Chem. Soc. 2004,
126, 10212−10213.
(68) Ahijado, M.; Braun, T.; Noveski, D.; Kocher, N.; Neumann, B.;
Stalke, D.; Stammler, H.-G. Angew. Chem., Int. Ed. 2005, 44, 6947−
6951.
(69) Carlton, L.; Mokoena, L. V.; Fernandes, M. A. Inorg. Chem.
2008, 47, 8696−8703.
(70) Cramer, C. J.; Tolman, W. B.; Theopold, K. H.; Rheingold, A. L.
Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 3635−3640.
(71) Fulmer, G. R.; Miller, A. J. M.; Sherden, N. H.; Gottlieb, H. E.;
Nudelman, A.; Stoltz, B. M.; Bercaw, J. E.; Goldberg, K. I.
Organometallics 2010, 29, 2176−2179.
(72) Takahashi, Y.; Hashimoto, M.; Hikichi, S.; Akita, M.; Moro-oka,
Y. Angew. Chem., Int. Ed. 1999, 38, 3074−3077.
(73) Konnick, M. M.; Guzei, I. A.; Stahl, S. S. J. Am. Chem. Soc. 2004,
126, 10212−10213.
(74) Wang, W.-D.; Bakac, A.; Espenson, J. H. Inorg. Chem. 1995, 34,
4049−4056.
(75) Bakac, A.; Wang, W.-D. J. Am. Chem. Soc. 1996, 118, 10325−
10326.
(76) Li, F.; Meier, K. K.; Cranswick, M. A.; Chakrabarti, M.; Van
Heuvelen, K. M.; Munck, E.; Que, L., Jr. J. Am. Chem. Soc. 2011, 133,
̈
7256−7259.
7201
dx.doi.org/10.1021/ic300279z | Inorg. Chem. 2012, 51, 7192−7201