Synthesis and hydride transfer reactions of cobalt and nickel hydride complexes to BX3 compounds

Article abstract of DOI:10.1021/ic200857x

Hydrides of numerous transition metal complexes can be generated by the heterolytic cleavage of H₂ gas such that they offer alternatives to using main group hydrides in the regeneration of ammonia borane, a compound that has been intensely studied for hydrogen storage applications. Previously, we reported that HRh(dmpe)₂ (dmpe = 1,2-bis(dimethylphosphinoethane)) was capable of reducing a variety of BX₃ compounds having a hydride affinity (HA) greater than or equal to the HA of BEt₃. This study examines the reactivity of less expensive cobalt and nickel hydride complexes, HCo(dmpe)₂ and [HNi(dmpe)₂]⁺, to form B-H bonds. The hydride donor abilities (ΔG_H-°) of HCo(dmpe)₂ and [HNi(dmpe)₂]⁺ were positioned on a previously established scale in acetonitrile that is cross-referenced with calculated HAs of BX₃ compounds. The collective data guided our selection of BX₃ compounds to investigate and aided our analysis of factors that determine favorability of hydride transfer. HCo(dmpe)₂ was observed to transfer H^- to BX₃ compounds with X = H, OC₆F₅, and SPh. The reaction with B(SPh)₃ is accompanied by the formation of dmpe-(BH₃)₂ and dmpe-(BH₂(SPh))₂ products that follow from a reduction of multiple B-SPh bonds and a loss of dmpe ligands from cobalt. Reactions between HCo(dmpe)₂ and B(SPh)₃ in the presence of triethylamine result in the formation of Et₃N-BH₂SPh and Et ₃N-BH₃ with no loss of a dmpe ligand. Reactions of the cationic complex [HNi(dmpe)₂]⁺ with B(SPh)₃ under analogous conditions give Et₃N-BH₂SPh as the final product along with the nickel-thiolate complex [Ni(dmpe)₂(SPh)] ⁺. The synthesis and characterization of HCo(dedpe)₂ (dedpe = Et₂PCH₂CH₂PPh₂) from H ₂ and a base is also discussed, including the formation of an uncommon trans dihydride species, trans-[(H)₂Co(dedpe) ₂][BF₄].

Full text of DOI:10.1021/ic200857x

Article
pubs.acs.org/IC
Synthesis and Hydride Transfer Reactions of Cobalt and Nickel
Hydride Complexes to BX3 Compounds
Michael T. Mock,* Robert G. Potter, Molly J. O’Hagan, Donald M. Camaioni, William G. Dougherty,†
W. Scott Kassel,† and Daniel L. DuBois
Pacific Northwest National Laboratory, P.O. Box 999, Richland, Washington 99352, United States
†Department of Chemistry, Villanova University, Villanova, Pennsylvania 19085, United States
S
* Supporting Information
ABSTRACT: Hydrides of numerous transition metal complexes can be
generated by the heterolytic cleavage of H2 gas such that they offer alternatives
to using main group hydrides in the regeneration of ammonia borane, a
compound that has been intensely studied for hydrogen storage applications.
Previously, we reported that HRh(dmpe)2 (dmpe = 1,2-bis(dimethyl-
phosphinoethane)) was capable of reducing a variety of BX3 compounds
having a hydride affinity (HA) greater than or equal to the HA of BEt3. This
study examines the reactivity of less expensive cobalt and nickel hydride complexes, HCo(dmpe)2 and [HNi(dmpe)2]+, to form
+
−
B−H bonds. The hydride donor abilities (ΔGH °) of HCo(dmpe)2 and [HNi(dmpe)2] were positioned on a previously
established scale in acetonitrile that is cross-referenced with calculated HAs of BX3 compounds. The collective data guided our
selection of BX3 compounds to investigate and aided our analysis of factors that determine favorability of hydride transfer.
HCo(dmpe)2 was observed to transfer H− to BX3 compounds with X = H, OC6F5, and SPh. The reaction with B(SPh)3 is
accompanied by the formation of dmpe-(BH3)2 and dmpe-(BH2(SPh))2 products that follow from a reduction of multiple
B−SPh bonds and a loss of dmpe ligands from cobalt. Reactions between HCo(dmpe)2 and B(SPh)3 in the presence of
triethylamine result in the formation of Et3N−BH2SPh and Et3N−BH3 with no loss of a dmpe ligand. Reactions of the cationic
complex [HNi(dmpe)2]+ with B(SPh)3 under analogous conditions give Et3N−BH2SPh as the final product along with the
nickel−thiolate complex [Ni(dmpe)2(SPh)]+. The synthesis and characterization of HCo(dedpe)2 (dedpe = Et2PCH2CH2PPh2)
from H2 and a base is also discussed, including the formation of an uncommon trans dihydride species, trans-
[(H)2Co(dedpe)2][BF4].
that HRh(dmpe)2 can effectively transfer a hydride ligand
to BX3 compounds, and in some cases multiple B−H bonds
can be formed. These initial results proved favorable,
however, in a practical system for AB regeneration, the use of
transition metals much less expensive than rhodium will be
required.
INTRODUCTION
■
Fundamental studies of dihydrogen binding to transition metal
complexes have been examined in considerable detail.1
Moreover, the use of H2 to prepare transition-metal hydride
complexes that are powerful hydride donors is of interest
because it could provide an alternative to using traditional
borohydride reagents that are generally expensive to produce
and regenerate.2 For example, methods to regenerate the
chemical hydrogen storage material ammonia borane (AB,
NH3BH3), a promising candidate for hydrogen storage3 due to
its gravimetric density of 19.6 wt % H2, have used LiAlH4,
MgH2, SnH, and N2H4 as a means to form B−H bonds.4
Recently, as an alternative approach toward AB regeneration,
we examined hydride transfer reactions between HRh(dmpe)2
(dmpe = 1,2-bis(dimethylphosphinoethane)) and a series of
BX3 complexes (X = OR, SPh, F, H; R = Ph, p-C6H4OMe,
To address this requirement, we have taken advantage of
previous studies regarding the structural and electronic factors
that control the hydride donor ability of transition-metal
hydride complexes with chelating diphosphine ligands, e.g.,
HM(diphosphine)2 and [HM′(diphosphine)2]+ (M = Co or
Rh and M′ = Ni, Pd, or Pt6). For strong hydride donors,
the incorporation of electron-donating substituents on the
diphosphine ligands6,7 and a small natural bite angle8 of the
diphosphine are critical. Furthermore, the identity of the transi-
tion metal plays an important role, where second row transition
metals have been shown to give better hydride donors than first
and third row metals with identical diphosphine ligands.9
Varying these structural and electronic parameters, hydride
donor abilities can span a range of free energies associated with
t
C6F5, Bu, Si(Me)3) using a combination of experiment and
theory.5 Our goals were to provide a reliable and quantitative
method for determining when hydride transfer from transition-
metal hydrides to BX3 compounds will be favorable. HRh-
(dmpe)2 was chosen because it is one of the most powerful
transition-metal hydride donors, and therefore hydride
transfer reactions with a wide variety of BX3 hydride acceptor
compounds could be examined. Indeed, this study established
5,10
−
eq 1a (ΔGH °) from 26 to 76 kcal/mol,
with smaller values
Received: April 23, 2011
Published: October 31, 2011
© 2011 American Chemical Society
11914
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928

Inorganic Chemistry
Article
−
of ΔGH ° representing better hydride donors.
(1a)
Herein, we report the preparation, characterization, and
hydride transfer reactions of first-row transition-metal hydride
systems of cobalt and nickel, with the ultimate goal of
developing catalytic processes based on these less expensive
metals. In the current transition-metal hydride complexes of
interest, i.e., HCo(dmpe)2 and [HNi(dmpe)2]+, dmpe has been
selected as the supporting diphosphine ligand due to the
important structural and electronic considerations noted above.
These transition-metal hydrides are evaluated for their ability to
perform hydride transfer reactions with BX3 compounds. BX3
compounds are selected on the basis of their hydride acceptor
ability and are proposed targets generated from the digestion
of spent ammonia borane material.11 Although a decrease in
hydride donor ability according to eq 1a is expected in these
systems compared to the HRh(dmpe)2 system, our findings
indicate that products containing multiple B−H bonds can be
formed in these reactions. In addition to hydride transfer
studies of [HM(dmpe)2]n+ (M = Co, Ni; n = 0, 1), we report
the synthesis and characterization of HCo(dedpe)2 (dedpe =
Et2PCH2CH2PPh2) from H2 and a base, highlighting
interesting structural findings with the cobalt−dedpe com-
plexes, and discuss initial hydride transfer chemistry.
consequence of the asymmetry of the complex and coupling to
the 59Co nucleus. Additionally, the hydride ligands appear as a
broad resonance at δ −14.5 in the H NMR spectrum. The
addition of Verkade’s base (P(CH3NCH2CH2)3N, VSB; pKa =
32.9 in acetonitrile14) to a THF solution of cis-[(H)2Co(dmpe)2]-
[BF4] results in the precipitation of [HP(CH3NCH2CH2)3N]-
[BF4] and concomitant formation of HCo(dmpe)2. The 31P NMR
spectrum of this fluxional five-coordinate cobalt hydride contains a
broad singlet centered at δ 46.6. The hydride resonance in the 1H
NMR spectrum is a quintet at δ −16.6.
1
RESULTS AND DISCUSSION
■
Synthesis and Characterization of Co(dmpe)2 and
Co(dedpe)2 Complexes. Transition-metal hydrides have
been prepared by a variety of synthetic routes. For example,
HCo(dppe)2 was prepared by reaction of an excess of a
hydridic borohydride (e.g., NaBH4) with [Co(dppe)2-
(CH3CN)][BF4]2 in ethanol.12 Initially, we followed a similar
procedure for the preparation of HCo(dmpe)2. Although
HCo(dmpe)2 was produced, we found that this preparative
route resulted in a variety of products resulting from
undesirable side reactions such as the formation of dmpe−
(BH3)2. HCo(dmpe)2 was first reported by Schunn,13 prepared
from the reduction of (dmpe)2CoBr2 with sodium naphthale-
nide, followed by the addition of a stoichiometric amount of
water or hydrogen gas. In the present work, we prepared
HCo(dmpe)2 from hydrogen using two separate synthetic
routes. The first “one-pot” method, shown in reaction 1, is a
modification of Schunn’s procedure. [Co(dmpe)2(CH3CN)2]-
[BF4]2 is reduced with two equivalents of potassium graphite
(KC8) followed by exposure to hydrogen gas, affording
HCo(dmpe)2 in 60% yield. Using the second preparative
method, reaction 2, two intermediate complexes are isolated on
route to HCo(dmpe)2. First, reduction of paramagnetic
[Co(dmpe)2(CH3CN)2][BF4]2 with one equivalent of KC8 in
acetonitrile produces the diamagnetic Co(I) species, [Co-
(dmpe)2(CH3CN)][BF4], in moderate yield. 1H and 31P NMR
signals for all diamagnetic complexes reported in this study
exhibit broadening due to the S = 7/2 spin of the 59Co nucleus.
In THF-d8, [Co(dmpe)2(CH3CN)][BF4] exhibits a singlet
at δ 50.7 in the 31P NMR spectrum consistent with four
magnetically equivalent phosphorus atoms. The addition of H2
gas to a THF or acetonitrile solution of [Co(dmpe)2-
(CH3CN)][BF4] produces the six-coordinate dihydride, cis-
[(H)2Co(dmpe)2][BF4]. The 31P NMR spectrum recorded in
THF-d8 contains two broad resonances at δ 61.2 and 51.6, a
An analogous three-step procedure, as shown in reaction 2,
was used to prepare the cobalt complexes containing the
asymmetric dedpe ligand. Starting from the paramagnetic
cobalt(II) complex, [Co(dedpe)2(CH3CN)][BF4]2, the addi-
tion of one equivalent of KC8 in acetonitrile affords the
diamagnetic complex [Co(dedpe)2(CH3CN)][BF4]. This
complex is sensitive to solvent conditions. In acetonitrile, the
solution is orange, corresponding to the five-coordinate square-
pyramidal complex [Co(dedpe)2(CH3CN)][BF4] with acetoni-
trile bound in the axial position. Removal of acetonitrile under
vacuum conditions results in a darkening of the solid. The
addition of THF results in a drastic color change to dark blue
due to a loss of the acetonitrile ligand, affording the 16-electron
complex, [Co(dedpe)2][BF4], confirmed by X-ray crystallog-
raphy (discussed below). The 31P NMR spectrum of [Co-
(dedpe)2(CH3CN)][BF4] in CD3CN contains two resonances,
a broad triplet at δ 68.8 (2JPP = 63 Hz) and a very broad singlet
(or more likely an unresolved triplet that is severely broadened
by the cobalt quadrupole) at δ 62.8, due to the two chemically
inequivalent phosphorus environments of the dedpe ligand.
The exposure of a solution of [Co(dedpe)2(CH3CN)][BF4] to
H2 gas results in an immediate color change from orange
to pale yellow, affording cis and trans isomers of the cobalt(III)
1
dihydride complex [(H)2Co(dedpe)2][BF4] according to H
and 31P NMR spectroscopic data. 31P NMR spectra recorded at
25 °C exhibit remarkably broad, unresolved resonances between
δ 67 and 90. Cooling the sample to −40 °C results in a
significant improvement in resolution of the 31P NMR
resonances (Figure S5, Supporting Information). At this
temperature, nine broad signals are observed. The inherent
complexity of the 31P NMR spectrum is attributed to the
asymmetry of the dedpe ligand and the presence of multiple
1
dihydride isomers in solution. Examination of the H NMR
11915
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928

Inorganic Chemistry
Article
Figure 1. The structures of trans-dihydride (structures A and B) and cis-dihydride (structures C−E) isomers formed from the addition of H2 to
[Co(dedpe)2(CH3CN)]+.
Figure 2. Molecular structures of (A) [Co(dmpe)2(CH3CN)2]2+, (B) [Co(dmpe)2(CH3CN)]+, (C) cis-[(H)2Co(dmpe)2]+, and (D) HCo(dmpe)2
with atom numbering. Thermal ellipsoids drawn at 30% probability. Hydrogen atoms omitted except for hydride ligands attached to Co in structures
C and D.
spectrum at 25 °C shows three broad, unresolved peaks in
the hydride region around δ −14. Upon cooling the sample to
−40 °C, four broad signals are present. The three broad features
at δ −14 are assigned as cis-dihydride products, while the new
quintet resonance at δ −10.9 is assigned as the trans-dihydride
isomer (Figure S6, Supporting Information). In a recent report
of cis and trans isomers of Ir(H)2(dppm)2, a similar downfield
hydride resonance was observed for trans-[Ir(H)2(dppm)2] at δ
−7.6 compared to cis-[Ir(H)2(dppm)2]15 at δ −11.4. In the
present case, further examination of the hydride region in the
1H{31P} NMR spectrum displays a singlet for the trans isomer
and four broad, overlapping resonances at δ −13.6, −13.7, −14.0,
and −14.6 for the cis-dihydride isomers. The possible structures
of trans- and cis-dihydride isomers for [(H)2Co(dedpe)2]+ are
shown in Figure 1.
The addition of Verkade’s base to a THF solution of
[(H)2Co(dedpe)2][BF4] affords a red solution of HCo-
(dedpe)2 and the precipitation of the protonated base,
[HP(CH3NCH2CH2)3N][BF4]. The five-coordinate cobalt−
hydride exhibits broad signals at δ 79.1 and 74.2 in the 31P
1
NMR spectrum and a quintet at δ −15.9 in the H NMR
spectrum corresponding to the hydride resonance.
Structural Studies. X-ray crytallographic studies were
carried out for each synthetic step in the preparation of
HCo(dmpe)2 and HCo(dedpe)2. Molecular structures of
[Co(dmpe)2(CH3CN)2]2+, [Co(dmpe)2(CH3CN)]+, [cis-
(H)2Co(dmpe)2]+, and HCo(dmpe)2 are shown in Figure 2,
structures A−D, respectively. Metric parameters for each
structure are listed in Table 1. The molecular structure of
[Co(dmpe)2(CH3CN)2]2+ (Figure 2, structure A) crystallized
11916
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928

Inorganic Chemistry
Article
bite angle is measured at 85.54(3)° and 85.86(3)° for P1−
Co(1)−P2 and P3−Co(2)−P4, respectively.
Table 1. Selected Bond Distances (Å) and Angles (deg) for
HCo(dmpe)2, cis-[(H)2Co(dmpe)2][BF4], [Co(dmpe)2-
(CH3CN)][BF4], [Co(dmpe)2(CH3CN)2][BF4]2
The molecular structure of [Co(dmpe)2(CH3CN)]+
(Figure 2, structure B) is square-pyramidal with the dmpe
ligands occupying the equatorial plane and acetonitrile bound
in the axial site. The Co−P bond distances are shorter than
the analogous Co(II) complex with an average bond length of
2.1736(5) Å. The Co−N bond distance is 1.9841(16) Å.
The molecular structure of the cationic dihydride cis-
[(H)2Co(dmpe)2]+ (Figure 2, structure C) is best described
as a distorted octahedron. The chelating dmpe ligands are
arranged in a cis configuration. The average Co−P bond
distance is 2.184(6) Å. The hydride ligands, located from the
difference map, occupy the two remaining coordination sites
and exhibit Co−H bond distances of 1.45(3) and 1.40(3) Å.
The Co−P bond distances trans to the hydride ligands are
slightly longer, 2.198(3) and 2.1934(8) Å, than those trans to
phosphine ligands, 2.180(6) and 2.166(6) Å. The bond angles
of P(1)−Co(1)−P(2) and P(3)−Co(1)−P(4) for the chelating
dmpe ligands are 91.72(12)° and 87.79(19)°; however,
deviation from ideal octahedral geometry of 180° is more
obvious in the P(1)−Co(1)−H(1D) and P(3)−Co(1)−H(1C)
bond angles of 168.3(14)° and 167.5(11)°, respectively, and
the P(2)−Co(1)−P(4) bond angle of 157.0(2)°.
a
HCo(dmpe)2
Co(2)−P(3)
Co(2)−P(4)
Co(2)−P(5)
Co(2)−P(6)
2.1306(12)
2.0847(12)
2.1325(12)
2.1164(11)
P(3)−Co(2)−P(4)
P(3)−Co(2)−P(5)
P(3)−Co(2)−P(6)
P(4)−Co(2)−P(5)
P(4)−Co(2)−P(6)
P(5)−Co(2)−P(6)
H(2M1)−Co(2)−P(3)
H(2M1)−Co(2)−P(4)
H(2M1)−Co(2)−P(5)
H(2M1)−Co(2)−P(6)
89.24(5)
114.23(5)
110.14(4)
119.76(5)
132.63(5)
91.81(4)
161.5(15)
72.8(15)
79.6(15)
80.3(15)
Co(2)−H(2M1) 1.62(3)
cis-[(H)2Co(dmpe)2][BF4]
Co(1)−P(1)
Co(1)−P(2)
Co(1)−P(3)
Co(1)−P(4)
Co(1)−H(1C)
Co(1)−H(1D)
2.198(3)
2.180(6)
2.1934(8)
2.166(6)
1.45(3)
P(1)−Co(1)−P(2)
P(1)−Co(1)−P(3)
P(1)−Co(1)−P(4)
P(2)−Co(1)−P(3)
P(2)−Co(1)−P(4)
P(3)−Co(1)−P(4)
P(1)−Co(1)−H(1C)
P(2)−Co(1)−H(1C)
P(3)−Co(1)−H(1C)
P(4)−Co(1)−H(1C)
P(1)−Co(1)−H(1D)
P(2)−Co(1)−H(1D)
P(3)−Co(1)−H(1D)
P(4)−Co(1)−H(1D)
H(1C)−Co(1)−H(1D)
91.72(12)
103.57(7)
103.25(17)
105.84(15)
157.0(2)
87.79(19)
86.7(11)
80.6(11)
167.5(11)
82.8(11)
168.3(14)
79.9(14)
86.7(13)
82.5(14)
83.9(16)
1.40(3)
The molecular structure of the neutral five-coordinate cobalt
hydride HCo(dmpe)2 (Figure 2, structure D) exhibits distorted
trigonal bipyramidal coordination geometry. This complex
crystallized with three independent molecules in the asym-
metric unit. In each molecule, the equatorial plane is composed
of three phosphorus atoms, while a fourth phosphorus atom
and a hydride ligand occupy the axial sites. The average Co−P
bond distances for each independent molecule are similar at
2.1002(12), 2.1161(12), and 2.1099(12) Å. The hydride
ligands were located from the difference map and were refined
with all three cobalt hydride distances made equivalent to one
another at 1.62(3) Å. The average bond angle between the axial
hydride ligand and the axial phosphorus atom is 160.7°.
Molecular structures from single crystal X-ray diffraction
studies performed on [Co(dedpe)2(CH3CN)]2+, [Co-
(dedpe)2]+, trans-[(H)2Co(dedpe)2]+, and HCo(dedpe)2 are
depicted in Figure 3, structures A−D, respectively. Table 2
contains selected metric parameters for these complexes.
[Co(dedpe)2(CH3CN)]2+ (Figure 3, structure A) contains
two independent molecules in the unit cell, each molecule
having a five-coordinate square-pyramidal geometry. The dedpe
ligands are arranged about the cobalt center so that the
phosphorus atoms with phenyl substituents from one ligand are
trans to phenyl groups of the opposing ligand. This arrange-
ment is likely favored to decrease steric interactions between
opposing ligands. The average Co−P bond length is 2.2649(5) Å.
The Co−P bonds with ethyl substituents on phosphorus are
slightly longer than those with phenyl substituents. The average
dedpe bite angle (i.e., P(1)−Co(1)−P(2)) is 84.6°.
[Co(dmpe)2(CH3CN)][BF4]
Co(1)−P(1)
Co(1)−P(2)
Co(1)−P(3)
Co(1)−P(4)
Co(1)−N(1)
2.1703(5)
2.1699(5)
2.1732(5)
2.1811(5)
1.9841(16)
P(1)−Co(1)−P(2)
P(1)−Co(1)−P(3)
P(1)−Co(1)−P(4)
P(2)−Co(1)−P(3)
P(2)−Co(1)−P(4)
P(3)−Co(1)−P(4)
N(1)−Co(1)−P(1)
N(1)−Co(1)−P(2)
N(1)−Co(1)−P(3)
N(1)−Co(1)−P(4)
84.78(2)
166.13(2)
93.60(2)
91.90(2)
161.68(2)
85.32(2)
97.75(5)
103.36(5)
96.12(5)
94.94(5)
[Co(dmpe)2(CH3CN)2][BF4]2
Co(1)−P(1)
Co(1)−P(2)
Co(1)−N(1)
Co(2)−P(3)
Co(2)−P(4)
Co(2)−N(2)
2.2419(9)
2.2434(9)
2.254(3)
2.2412(9)
2.2465(9)
2.271(3)
P(1)−Co(1)−P(2)
P(1)−Co(1)−P(2A)
P(1)−Co(1)−P(1A)
P(1)−Co(1)−N(1)
P(3)−Co(2)−P(4)
P(3)−Co(2)−P(3A)
P(3)−Co(2)−P(4A)
P(3)−Co(2)−N(2)
85.54(3)
94.46(3)
180.0
87.14(8)
85.86(3)
180.00(4)
94.14(3)
89.16(8)
a
Metric parameters representing one of the three independent
molecules in the unit cell; see the Supporting Information for
additional bond distances and angles.
Crystals of [Co(dedpe)2]+ (Figure 3, structure B) were grown
by ethyl ether diffusion into an acetonitrile solution of the metal
complex. Our attempts to obtain X-ray quality crystals of
[Co(dedpe)2(CH3CN)][BF4] by this combination of solvents
was unsuccessful. Instead, after two weeks, blue crystals of the
16-electron complex [Co(dedpe)2][BF4] precipitated out of
the yellow-orange solution. The complex exhibits a nearly ideal
square-planar geometry with a dedpe ligand arrangement equi-
valent to the previously described structure. The asymmetric
unit contains four independent molecules with an average
with two independent molecules per asymmetric unit. Each
molecule exhibits octahedral coordination geometry with nearly
identical metric parameters. The complex is composed of two
dmpe ligands occupying the equatorial plane and two
acetonitrile ligands in the axial coordination sites. The average
Co−P bond distance is 2.2433(9) Å, while the Co−N bond
distances are similar at 2.254(3) and 2.271(3) Å. The dmpe
11917
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928

Inorganic Chemistry
Article
Figure 3. Molecular structures of (A) [Co(dedpe)2(CH3CN)]2+, (B) [Co(dedpe)2]+, (C) trans-[(H)2Co(dedpe)2]+, and (D) HCo(dedpe)2 with
atom numbering. Thermal ellipsoids drawn at 30% probability. Hydrogen atoms omitted except for hydride ligands attached to Co in structures C
and D.
Co−P bond length of 2.1864(7) Å, shorter than the average
Co−P bond length in [Co(dedpe)2(CH3CN)]2+.
Co−H bond length is 1.42(2) Å, similar to that of HCo(dppe)2
and HCo(dppp)2.12,17
The angles between the hydride ligand and the equatorial
phosphorus atoms, namely, P2−Co(1)−H1C and P4−Co(1)−
H1C, are acute at 75.3(9)° and 76.6(9)°, while the bond angle
in P3−Co(1)−H1C is larger at 90.6(9)°. The P1−Co(1)−H1C
bond angle deviates significantly from linearity at 156.3(9)°.
Hydride Donor Abilities of Transition-Metal Hydrides
and Hydride Affinities of BX3 Compounds. The hydride
donor abilities of HCo(dmpe)2 and HCo(dedpe)2 were
estimated by comparing the difference in hydride donor ability
of two systems that have been determined experimentally.
Estimates of hydride donor abilities using this approach are
a useful tool to check experimentally determined values.5
For example, the experimentally determined hydride donor
abilities of HCo(dppe)2 (49 kcal/mol) and [HNi(dppe)2]+
(63 kcal/mol; dppe = 1,2-bis(diphenylphosphinoethane))
differ by 14 kcal/mol. A similar difference would be expected
between HCo(dmpe)2 and [HNi(dmpe)2]+. The hydride
donor ability of [HNi(dmpe)2]+ was determined to be 51
kcal/mol. On this basis, we estimate a value of 37 kcal/mol for
The structure obtained from colorless crystals of trans-
[(H)2Co(dedpe)2]+ (Figure 3, structure C) has nearly ideal
octahedral geometry with two dedpe ligands occupying the
equatorial plane and the hydride ligands, located from the dif-
ference map, positioned in the axial coordination sites in the
less commonly observed trans stereochemistry.15,16 The average
Co−P bond length is 2.1861(8) Å, similar to other octahedral
cobalt(III) cations with chelating diphosphine ligands.12 The
Co−H bond distances were found to be slightly different at
1.40(3) and 1.53(3) Å, and the H(1A)−Co(1)−H(1B) bond
angle is nearly linear at 177.9(16)°.
The molecular structure of HCo(dedpe)2 has distorted
trigonal-bipyramidal geometry. The equatorial plane consists of
phosphorus atoms P3 and P4 from one of the chelating dedpe
ligands and P2 of the second chelating dedpe ligand, while the
hydride ligand and phosphorus atom P1 of the second dedpe
ligand occupy the axial positions. The equatorial Co−P bond
lengths are 2.1489(4), 2.1397(5), and 2.1120(4) Å, respec-
tively, while the axial Co−P distance is 2.1357(4) Å. The
−
ΔGH ° for HCo(dmpe)2. Likewise, using the experimentally
11918
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928

Inorganic Chemistry
Article
Table 2. Selected Bond Distances (Å) and Angles (deg) for HCo(dedpe)2, trans-[(H)2Co(dedpe)2][BF4], [Co(dedpe)2][BF4],
and [Co(dedpe)2(CH3CN)][BF4]2
HCo(dedpe)2
P(1)−Co(1)−P(2)
[Co(dedpe)2][BF4]
2.2108(7)
Co(1)−P(1)
Co(1)−P(2)
Co(1)−P(3)
Co(1)−P(4)
Co(1)−H(1C)
2.1357(4)
2.1120(4)
2.1489(4)
2.1397(5)
1.42(2)
88.387(17)
112.374(17)
107.613(17)
117.100(18)
140.121(18)
90.808(17)
156.3(9)
Co(2)−P(4)
Co(3)−P(5)
Co(3)−P(6)
Co(4)−P(7)
Co(4)−P(8)
P(3)−Co(2)−P(4)
P(3)−Co(2)−P(3A)
P(3)−Co(2)−P(4A)
P(5)−Co(3)−P(6)
P(5)−Co(3)−P(5A)
P(5)−Co(3)−P(6A)
P(7)−Co(4)−P(8)
P(7)−Co(4)−P(7A)
P(7)−Co(4)−P(8A)
85.27(2)
179.998(9)
94.73(2)
94.57(3)
180.0
P(1)−Co(1)−P(3)
P(1)−Co(1)−P(4)
P(2)−Co(1)−P(3)
P(2)−Co(1)−P(4)
P(3)−Co(1)−P(4)
H(1C)−Co(1)−P(1)
H(1C)−Co(1)−P(2)
H(1C)−Co(1)−P(3)
H(1C)−Co(1)−P(4)
2.1682(7)
2.1999(7)
2.1709(7)
2.2045(7)
85.43(3)
84.63(3)
179.999(1)
95.37(3)
75.3(9)
90.6(9)
76.6(9)
[Co(dedpe)2(CH3CN)][BF4]2
trans-[(H)2Co(dedpe)2][BF4]
Co(1)−P(1)
Co(1)−P(2)
Co(1)−P(3)
Co(1)−P(4)
Co(1)−N(1)
Co(2)−P(5)
Co(2)−P(6)
Co(2)−P(7)
Co(2)−P(8)
Co(2)−N(2)
2.2543(5)
2.2898(5)
2.2538(5)
2.2657(5)
2.0128(16)
2.2439(5)
2.2732(5)
2.2614(5)
2.2772(5)
2.0253(16)
P(1)−Co(1)−P(2)
P(1)−Co(1)−P(3)
P(1)−Co(1)−P(4)
P(2)−Co(1)−P(3)
P(2)−Co(1)−P(4)
P(3)−Co(1)−P(4)
N(1)−Co(1)−P(1)
N(1)−Co(1)−P(2)
N(1)−Co(1)−P(3)
N(1)−Co(1)−P(4)
P(5)−Co(2)−P(6)
P(5)−Co(2)−P(7)
P(5)−Co(2)−P(8)
P(6)−Co(2)−P(7)
P(6)−Co(2)−P(8)
P(7)−Co(2)−P(8)
N(2)−Co(2)−P(5)
N(2)−Co(2)−P(6)
N(2)−Co(2)−P(7)
N(2)−Co(2)−P(8)
84.646(18)
177.47(2)
94.663(19)
95.756(19)
169.42(2)
84.483(18)
91.43(5)
Co(1)−P(1)
Co(1)−P(2)
Co(1)−P(3)
Co(1)−P(4)
Co(1)−H(1B)
Co(1)−H(1C)
2.1719(8)
2.1964(9)
2.1707(8)
2.1893(8)
1.40(3)
P(1)−Co(1)−P(2)
P(1)−Co(1)−P(3)
P(1)−Co(1)−P(4)
P(2)−Co(1)−P(3)
P(2)−Co(1)−P(4)
P(3)−Co(1)−P(4)
H(1B)−Co(1)−P(1)
H(1B)−Co(1)−P(2)
H(1B)−Co(1)−P(3)
H(1B)−Co(1)−P(4)
H(1C)−Co(1)−P(1)
H(1C)−Co(1)−P(2)
H(1C)−Co(1)−P(3)
H(1C)−Co(1)−P(4)
H(1B)−Co(1)−H(1C)
85.85(3)
177.64(4)
94.34(3)
94.05(3)
176.87(4)
85.89(3)
79.6(13)
92.1(13)
98.2(13)
91.0(13)
99.5(10)
89.7(10)
82.8(10)
87.1(10)
177.9(16)
1.53(3)
94.38(5)
91.03(5)
96.18(5)
84.383(18)
176.06(2)
95.297(19)
94.822(19)
172.95(2)
85.017(18)
92.15(5)
[Co(dedpe)2][BF4]
91.07(5)
Co(1)−P(1)
Co(1)−P(2)
Co(2)−P(3)
2.1663(6)
2.2036(7)
2.1676(6)
P(1)−Co(1)−P(2)
P(1)−Co(1)−P(1A)
P(1)−Co(1)−P(2A)
85.18(2)
179.999(10)
94.82(2)
91.72(5)
95.97(5)
measured value of 60 kcal/mol for [HNi(dedpe)2]7+, we predict
a hydride donor ability of 46 kcal/mol for HCo(dedpe)2. In
addition, the estimated hydride donor ability of 37 kcal/mol for
HCo(dmpe)2 is in good agreement with the computationally
show that HRh(dmpe)2 and [HBEt3]− have the same hydride
donor abilities. This allows us to relate the hydride donor
ability and hydride affinity scales shown in Figure 4. In so
−
determined hydricity (ΔGH °) of 36.3 kcal/mol for HCo-
(dmpe)2 from Liu and co-workers.18
In the previous study, we predicted hydride affinity (HA)
values for selected BX3 complexes using electronic structure
calculations.5 Table 3 lists the calculated HA values for BX3
Table 3. Calculated Hydride Affinity Values of BX3
a
Compounds
BX3 compound
hydride affinity −ΔH° (kcal/mol)
BEt3
58.7
73.0
Figure 4. Scales (kcal/mol) of experimentally determined hydride
donor abilities of selected transition-metal hydride complexes in
acetonitrile (blue, bottom) and the calculated gas-phase hydride
affinity values for selected BX3 compounds (red, top).
B(OPh)3
BH3
b
73.7
BH2(SPh)
BH(SPh)2
B(SPh)3
B(OC6F5)3
78.3
80.8
86.6
doing, we assume (1) that the difference between ΔG and ΔH
is approximately constant, dominantly due to the loss of
translational entropy of H−, and (2) that the differences in
93.3
a
−
From reaction enthalpies for hydride transfer from BH4 to BX3
calculated at the B3LYP/aug-cc-pVTZ//B3LYP/DZVP2 level of
−
b
theory. CCSD(T)/CBS value.38
solvation free energies (ΔGsolv) of BHX3 and BX3 are
constant. However, we realize that this second assumption
may break down because the solvation of ions is more strongly
dependent on size than is the solvation of neutrals. Therefore,
it needs to be considered when estimating relative acceptor
compounds used in this study. We include the HA values for
BH(SPh)2 and BH2(SPh) that are intermediates in the
reduction of B(SPh)3. Our previous experimental studies
11919
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928

Inorganic Chemistry
Article
abilities. Although Figure 4 is semiquantitative, it was a useful
guide in selecting BX3 compounds to react with the Co and Ni
hydrides of this study.
Hydride Transfer Reactions from Cobalt to BX3
Complexes: (X = OC6H5, OC6F5). Experiments also tested
the reactivity of HCo(dmpe)2 with B(OPh)3 (HA = 73.0).
Because the hydride transfer from HCo(dmpe)2 to THF-BH3
(HA = 73.7) at room temperature produced [BH4]− by 11B
NMR spectroscopy, reactions in which a 3-fold excess of
HCo(dmpe)2 was mixed with B(OPh)3 were expected to be
favorable. We found that HCo(dmpe)2 did not produce hydride
transfer products even after refluxing the solution for 12 h. It is
notable that, given the nearly identical calculated hydride affinity
values for BH3 and B(OPh)3, hydride transfer was not observed.
The difference in reactivity suggests that in cases where the HAs
of BX3 compounds are similar, other factors, such as sterics and
solvation free energies of BX3/BHX3−, need to be explicitly
Hydride Transfer from HCo(dmpe)2 to THF-BH3.
Hydride transfer to THF-BH3 was expected to be thermody-
namically close to zero on the basis of our observations in the
previous section requiring excess NaBH4 to prepare HCo-
(dmpe)2, as well as its position in Figure 4. In Figure 4,
transition-metal hydrides to the left of a boron complex on the
top scale are expected to transfer a hydride from the metal to
boron. The addition of 0.5 equiv of THF-BH3 to a THF
solution of HCo(dmpe)2 results in an immediate color change
from yellow to orange. The 11B{1H} NMR spectrum recorded
immediately after mixing, Figure 5, trace A, indicates the
−
considered. In this instance, BH4 and BH(OPh)3− likely have
very different solvation free energies such that the acceptor
abilities BH3 and B(OPh)3 are not as similar as Figure 4 indicates.
In order to increase the hydride acceptor ability of the subject
borate ester, electron-withdrawing substituents were incorpo-
rated on the aromatic ring. Computational results (Table 3 and
Figure 4) indicated that electron-withdrawing groups dramat-
ically increase the hydride acceptor ability of the BX3
compounds.5,38 For example, a 3-fold excess of HCo(dmpe)2
reacts with B(OC6F5)3, one of the best hydride acceptors, HA =
93 kcal/mol, according to reaction 3. The 1H-coupled 11B NMR
spectrum, Figure 6, trace A, recorded immediately after mixing
Figure 5. (A) 11B{1H} NMR spectrum recorded at 25 °C from the
hydride transfer reaction of HCo(dmpe)2 to THF-BH3. (B) The
1
corresponding H-coupled 11B NMR spectrum.
formation of three new boron-containing products. Indicating
that hydride transfer has occurred, a singlet at δ −37.6, a
1
quintet in the H-coupled 11B NMR spectrum, Figure 5, trace
B, is assigned as the tetrahydroborate anion of [Co(dmpe)2]-
[BH4]. The corresponding 31P NMR spectrum contains a
singlet at δ 51.8 consistent with [Co(dmpe)2]+. By comparison,
the 11B NMR spectrum of [Rh(dmpe)2][BH4], prepared in the
hydride transfer reaction between HRh(dmpe)2 and THF-BH3,
appears at δ −37.4.5 Of the remaining boron-containing products
in Figure 5, a doublet centered at δ −39.0 in the 11B{1H} NMR
Figure 6. (A) 1H-coupled 11B NMR spectrum recorded at 25 °C from
the hydride transfer reaction of HCo(dmpe)2 to B(OC6F5)3. (B) The
corresponding 11B{1H} NMR spectrum.
1
spectrum, a doublet of quartets in the H-coupled 11B NMR
spectrum, has been identified as dmpe-(BH3)2, the product of a
side reaction where the dmpe ligand has been removed from
the cobalt center. The identity of the singlet at δ −38.3, a quintet
indicates the formation of [HB(OC6F5)3]−, a broad doublet at δ
4.9, and [B(OC6F5)4]−, a sharp singlet at δ 1.0.
1
in the H-coupled 11B NMR spectrum, remains tenuous. The
minor variation in chemical shift (δ −37.6 versus −38.3) suggests
a slightly different chemical environment or degree of ion pairing
for the tetrahydroborate anion than in the discrete complex
[Co(dmpe)2][BH4]. This could be due to the formation of a
bridging BH4− anion with the cobalt(I) center supported by one
or two diphosphines. This type of interaction has been observed
in trans-[Fe(H)(dmpe)2BH4], located at δ −38.1 in the 11B
NMR spectrum,19 and transition metal-tetrahydroborate com-
plexes such as [(C6H5)3P)2]Cu(BH4) have been reported.20 Last,
the 31P NMR spectrum after 14 h contained only HCo(dmpe)2
and dmpe-(BH3)2, while the corresponding 11B NMR indicated
dmpe-(BH3)2 as the only detectible product. This result suggests
that the kinetically favored reaction is the hydride transfer to
BH3, while the thermodynamically favored reaction results in the
transfer of dmpe from the cobalt center to BH3.
(3)
The 31P NMR spectrum contains a sharp singlet at δ 52.9 for
[Co(dmpe)2]+ and a broad resonance at δ 46.4 for unreacted
HCo(dmpe)2. Heating the reaction at 50 °C for 6 h produced
only a minor increase in the amount of [HB(OC6F5)3]− by 11
B
NMR. There were no resonances in the 11B NMR spectra,
suggesting the formation of products containing multiple B−H
bonds.
Hydride Transfer from HCo(dmpe)2 to B(SPh)3. The
hydride transfer reaction between HCo(dmpe)2 and B(SPh)3
11920
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928

Inorganic Chemistry
Article
was anticipated to have a considerable driving force due to the
high hydride affinity value of 87 kcal/mol. Additionally, due to a
weaker heterolytic B−S bond strength, the formation of
multiple B−H bonds was expected in these reactions. For
example, in the reaction between a 3-fold excess of HRh-
(dmpe)2 with B(SPh)3 in THF, a mixture of products was
obtained over the course of the reaction including [B(SPh)4]−
and [HxB(SPh)4−x]− (x = 1−4). When solid B(SPh)3 was added
to a THF solution containing a 3-fold excess of HCo(dmpe)2 at
room temperature according to reaction 4, the solution color
immediately changed from a golden yellow to a red-brown.
δ 0.22 and 9.31, respectively. An additional observation from
the 31P NMR spectrum was the absence of the signal for
[Co(dmpe)2]+ (singlet at ∼ δ 51) despite the formation of a
small amount of the [H2B(SPh)2]− anion. One consideration
could be that upon the loss of the dmpe ligands the cobalt
cation is stabilized by the formation of a cobalt-THF adduct,
[Co(THF)x]+ (x = 4 or 5). THF adducts such as [Co(thf)6]2+
have been structurally characterized.21 Efforts to identify this
cationic cobalt species product were not pursued.
Hydride transfer between excess HCo(dedpe)2 and B(SPh)3
was surprisingly sluggish. Immediately after mixing, only the
starting materials were observed in the 31P and 11B NMR
spectra. In fact, no hydride transfer was observed until the reac-
tion was heated at 50 °C for 4 days. After this time, the products
noted by 11B NMR included [B(SPh)4]− and a small amount of
[HB(SPh)3]−. Decomposition of the cobalt complex was
evident due to the presence of a free dedpe ligand in the 31P
NMR spectrum. We attribute the limited ability of HCo-
(dedpe)2 to transfer the hydride ligand to the steric bulk of the
dedpe ligands.
Hydride Transfer from HCo(dmpe)2 to B(SPh)3 in the
Presence of NEt3. The results in the previous section
demonstrated that hydride transfer from HCo(dmpe)2 to
B(SPh)3 led to the formation of products containing multiple
B−H bonds. However, during the course of the reaction, an
undesirable side reaction occurs where the dmpe ligands of
[Co(dmpe)2]+ are removed from the metal complex and act as
a Lewis base to form borane adducts. In an effort to avoid or
significantly reduce the decomposition of the cobalt complex,
hydride transfer reactions between HCo(dmpe)2 and B(SPh)3
were performed in the presence of triethylamine. The addition
of an exogenous donor should provide an alternative Lewis
base other than dmpe ligands and drive the reactions toward
producing a neutral amine borane product, Et3N-BH3. In a
typical reaction, depicted in reaction 5, solid B(SPh)3 was
added to a solution of HCo(dmpe)2 in 2 mL of THF and
100 μL of NEt3, resulting in an immediate color change from
yellow to orange.
(4)
The 1H-coupled 11B NMR spectrum Figure 7, trace A,
collected immediately after mixing indicated the formation of
Figure 7. (A) 1H-coupled 11B NMR spectrum from the hydride
transfer reaction from HCo(dmpe)2 to B(SPh)3 in THF recorded
immediately after mixing. (B) The corresponding 11B{1H} NMR
spectrum.
three products containing B−H bonds. A triplet at δ −14.7 was
assigned as [H2B(SPh)2]−, the result of two hydride transfers
to boron. Additionally, multiplets at δ −25.1 and −38.7 are
assigned as the phosphine adducts, dmpe-(H2B(SPh))2 and
dmpe-(BH3)2, based on the presence of boron−phosphorus
(5)
The H-coupled 11B NMR spectrum recorded immediately
1
1
1
coupling of JBP = 67 Hz and JBP = 56 Hz, respectively, in the
11B{1H} NMR spectrum, Figure 7, trace B. No B(SPh)3 or
[B(SPh)4]− was observed in the 11B NMR spectrum
immediately after mixing. Formation of these B−H containing
products indicates that multiple hydride transfers have
occurred; however, as observed in the reaction with THF-
BH3, decomposition of the cobalt complex resulting from the
loss of dmpe ligands is evident. The 31P NMR spectrum shows
complete consumption of HCo(dmpe)2 immediately after
mixing. The spectrum contains a singlet at δ 44.7 assigned as
Co(dmpe)2(SPh) (structurally characterized product isolated
from the reaction, Figure S3 in the Supporting Information).
Co(dmpe)2(SPh) was prepared independently by the reaction
of [Co(dmpe)2(CH3CN)][BF4] with NaSPh in THF to
confirm the 31P NMR chemical shift. Also present in the 31P
NMR spectrum of the hydride transfer reaction are multiplets
corresponding to dmpe-(H2B(SPh))2 and dmpe-(BH3)2 at
after mixing, Figure 8, trace A, a singlet at δ 4.2 which indicated
the formation of [B(SPh)4]− as the major boron-containing
product. No B(SPh)3 was observed in the reaction mixture
immediately after mixing. In addition, two minor hydride
transfer products are observed, a doublet at δ −4.5 and a triplet
at δ −6.8, assigned as [HB(SPh)3]− and Et3N−H2B(SPh),
respectively. The 31P NMR spectrum indicates that HCo-
(dmpe)2 and Co(dmpe)2(SPh) are present in roughly equal
amounts. Interestingly, there is no evidence for the formation of
dmpe-(BH3)2 that had been observed without triethylamine
present in the reaction. Heating the reaction for 6 h at 55 °C
results in additional formation of B−H bonds and near
complete consumption of HCo(dmpe)2 with the concomitant
1
formation of Co(dmpe)2(SPh) by 31P NMR. The H-coupled
11B NMR spectrum shown in Figure 8, trace B, contains two
borohydride amine adducts: Et3N-BH2(SPh) present as a
triplet at δ −6.8 and modest conversion to Et3N-BH3, a quartet
11921
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928

Inorganic Chemistry
Article
of a small amount of [B(SPh)4]− at δ 4.15. The 1H-coupled 11
B
NMR spectrum recorded after 18 h at 25 °C indicates that a
small amount of unreacted B(SPh)3 remained and two B−H
containing products were formed. The major product is a
triplet corresponding to Et3N-BH2(SPh) at δ −7.1, overlapping
with [B(Ph)4]− at δ −6.7. The minor product is a doublet at δ
4.9, tentatively assigned as Et3N-BH(SPh)2 (JBH = 130 Hz), a
product not previously observed in hydride transfer reactions
with cobalt or rhodium where the driving force for hydride
transfer is greater than for nickel. After heating the reaction
mixture at 50 °C for 24 h, as shown in reaction 6, dark red
crystals precipitated from the orange solution. This material
was identified as the nickel−thiolate complex [Ni(dmpe)2-
(SPh)][B(Ph)4] (see Figure S4 Supporting Information for
X-ray crystal structure).
Figure 8. 1H-coupled 11B NMR spectra from the hydride transfer
reaction from HCo(dmpe)2 to B(SPh)3 in the presence of NEt3: (A)
Recorded immediately after mixing, (B) recorded after heating for 6 h
at 55 °C, showing triethylamineborane, Et3N-BH3, at δ −12.8.
centered at δ −12.8, in which case all B−S bonds have been
cleaved. Despite initial formation of the four-coordinate
thioborate ester, [B(SPh)4]−, nearly all B−S bonds were
cleaved, and multiple B−H bonds were formed with no loss of
dmpe ligands from the cobalt complex.
(6)
The 11B NMR of the orange solution indicated that the
remaining B(SPh)3 was consumed, and Et3N-BH2(SPh) was
observed as the only boron containing product, Figure 9, traces
The formation of Co(dmpe)2(SPh) is prevalent in the
hydride transfer reactions noted above. The formation of this
product is the result of multiple hydride transfer and ligand
redistribution reactions occurring between boron and cobalt,
ultimately leading to the thermodynamically favored products,
Co(dmpe)2(SPh) and Et3N-BH3. The ligand redistribution
reactions are expected to contribute to the driving force for
hydride transfer and the formation of B−H bonds. Although a
mechanistic study has not been performed, [Co(dmpe)2]+
clearly participates in assisting with SPh− redistribution, thus
facilitating additional hydride transfer or substituent exchange
reactions. A related reaction involving alkali-metal cation
participation in a ligand redistribution process was proposed
by Noth and Knizek.22 Notably, the formation of Co(dmpe)2-
̈
(SPh) poses a challenge in the development of a catalytic
process, as coordination of SPh− to cobalt hinders the ability of
the metal center to activate H2.
Hydride Transfer to B(SPh)3 Using [HNi(dmpe)2][BPh4]
in the Presence of Triethylamine. Thus far, the transition-
metal hydride complexes of interest contain chelating diphos-
phine ligands, e.g., HM(diphosphine)2 complexes where M is Co
or Rh. Next, we turned our attention toward transition-metal
complexes of the type [HM′(diphosphine)2]+ where M′ is Ni.
The group 10 transition-metal hydride complexes exhibit an
overall positive charge, making them less powerful hydride
donors compared to the neutral group 9 analogues. As mentioned
above, thermochemical data collected for [HNi(dmpe)2]+ reveal a
Figure 9. (A) 1H-coupled 11B NMR spectrum from the hydride
transfer reaction of [HNi(dmpe)2][B(Ph)4] to B(SPh)3 in the
presence of triethylamine after heating at 50 °C for 24 h. (B) The
corresponding 11B{1H} NMR spectrum.
A and B. Unlike the reactions with cobalt and rhodium, the
completely reduced product, Et3N-BH3, was not observed after
heating despite the presence of excess [HNi(dmpe)2]+. The
positioned hydride donor ability of [HNi(dmpe)2]+ (ΔGH
=
−
−
hydride donor ability of ΔGH = 50.7 kcal/mol, and the low cost
50.7 kcal/mol), relative to the hydride affinity values for
HB(SPh)2 and H2B(SPh) (81 and 78 kcal/mol, respectively) in
Figure 4, suggests that hydride transfer should be unfavorable.
While [HNi(dmpe)2]+ appears not to be a strong enough
hydride donor to transfer a hydride to BH2(SPh), we propose
the formation of the Ni−S bond in [Ni(dmpe)2(SPh)]+ and the
B−N dative bond in Et3N-BH2SPh may offset the unfavorable
thermodynamics for hydride transfer to HB(SPh)2, which is a
slightly better hydride acceptor than BH2(SPh). Additionally,
the overall positive charge of the complex may contribute to
slower reaction rates compared to reactions with cobalt and
rhodium hydrides.
of nickel compared to Pd or Pt or Rh is an attractive feature in
examining the nickel system. Our preliminary reactivity studies
focused on reactions between [HNi(dmpe)2][BPh4] and
B(SPh)3, with the intent of demonstrating that multiple B−H
bonds can be formed using transition-metal hydrides that are
considerably less hydridic than the Co and Rh analogues.
The reaction of 3 equiv of [HNi(dmpe)2][BPh4] with 1
equiv of B(SPh)3 performed in THF in the presence of
triethylamine does not immediately form products containing
B−H bonds. 11B NMR data recorded immediately after mixing
indicates the starting materials B(SPh)3 and [HNi(dmpe)2]-
[BPh4] at δ 62.3 and −6.7, respectively, as well as the formation
11922
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928

Inorganic Chemistry
Article
dried over NaK, and vacuum transferred before use. Triethylamine was
purchased from Alfa-Aesar and distilled over CaH2 before use. Vinyldi-
phenylphosphine, diethylphosphine, 1,2-bis(dimethylphosphino)-
ethane, and Verkade’s base (P(CH3NCH2CH2)3N; 2,8,9-trimethyl-
2,5,8,9-tetraaza-1-phosphabicyclo[3.3.3]undecane) were purchased
from Strem Chemicals, Inc. 2,2′-Azobis(2-methylpropionitrile)
(AIBN), borane-THF, ammonium tetraphenylborate, thiophenol,
pentafluorophenol, and boron trichloride were purchased from
Sigma-Aldrich. High purity grade (99.9%) H2 gas was passed through
a column of Drierite before use. [Co(CH3CN)6][BF4]223 was prepared
as previously described. [HNi(dmpe)2][B(Ph)4] was prepared as
previously described24 substituting [NH4][B(Ph)4] for [NH4][PF6].
Preparation of dedpe Phosphine Ligand. A reaction flask was
charged with 1.2 mL (0.010 mmol) of diethylphosphine, 2.1 mL
(0.010 mmol) of vinyldiphenylphosphine, and 0.160 g of 2,2′-azobis(2-
methylpropionitrile). The reaction was heated at 70 °C for 2 h.
A second portion (0.100 g) of AIBN was added after 2 h, and the
reaction was heated at 70 °C for an additional 5 h. Volatile materials
were removed under reduced pressure, affording a clear yellow oil.
Spectroscopic data matched the previously reported values for dedpe
prepared using a Rayonet photoreactor.7 31P{1H} NMR (toluene-d8):
δ −12.0 (d, JPP = 25 Hz, PPh2), −18.2 (d, JPP = 25 Hz, PEt2).
Synthesis of BX3 Compounds. All aryl borates were synthesized
from BCl3 and the corresponding alcohol or thiol using a low
temperature version of the Michaelis method25 developed by Lappert
and co-workers.26 B(SPh)3 was prepared as previously described.5
HCo(dmpe)2, Method A. A solution of Verkade’s Base, (P-
(CH3NCH2CH2)3N; 0.057 g, 0.26 mmol) in 2 mL of THF was added
to a stirring solution of cis-[(H)2Co(dmpe)2][BF4] (0.121 g, 0.27
mmol) in 10 mL of THF, resulting in a dark yellow solution and the
precipitation of a white solid, [HP(CH3NCH2CH2)3N][BF4]. The
solution was stirred for 2 h, filtered through Celite, and the solvent
removed under vacuum conditions. The crude solid was extracted with
pentane and passed through a plug of APTS-coated silica gel, eluting
with pentane. The volatiles were removed under vacuum conditions,
affording a bright yellow solid. Yield: 0.050 g, 52%. 1H NMR
(THF-d8): δ 1.47 (s, 8H, PCH2CH2), 1.26 (s, 24H, P(CH3)2), −16.6
(quintet, 1H, Co−H). 31P{1H} NMR (THF-d8): δ 46.6 (br). IR
(KBr): νCo−H 1856 cm−1. Anal. Calcd for C12H33CoP4: %C, 40.01; %
H, 9.23. Found: %C, 40.28; %H, 9.37.
SUMMARY AND CONCLUSIONS
■
This study examines the formation of B−H bonds using
nonprecious transition-metal hydrides supported by diphos-
phine ligands. These five-coordinate transition-metal hydride
complexes, HCo(dmpe)2 and HCo(dedpe)2 and [HNi-
(dmpe)2][B(Ph)4], can be prepared from H2 and a base.
On route to the five-coordinate monohydrides, reaction of
the cobalt(I) cations, [Co(dmpe)2(CH3CN)]+ and [Co-
(dedpe)2(CH3CN)]+, with H2 affords the six-coordinate
dihydride complexes. Structural characterization by X-ray
crystallography indicated the formation of cis-[(H)2Co-
(dmpe)2]+ and a less commonly observed trans-dihydride
isomer for trans-[(H)2Co(dedpe)2]+. The hydride donor
abilities of HCo(dmpe)2 and HCo(dedpe)2 were estimated
on the basis of comparisons to hydride donor abilities
determined experimentally and from computational studies.18
The hydride donor abilities for the cobalt and nickel systems
were cross-referenced to calculated hydride affinities of BX3
complexes in order to predict when hydride transfer may be
favorable. Though the cobalt and nickel hydride complexes
exhibit hydride donor abilities significantly less than the
previously studied HRh(dmpe)2 system,5 we observed hydride
transfer from HCo(dmpe)2 to BH3, B(SPh)3, and B(OC6F5)3
and from [HNi(dmpe)2]+ to B(SPh)3. Multiple B−H bonds
were formed in the reactions with B(SPh)3, and in the case of
Co, the reaction was accompanied by a loss of the dmpe ligands
to form dmpe-(BH3)2. Decomposition was avoided by the
addition of triethylamine, which led to the formation of Et3N-
BH2(SPh) and Et3N-BH3 in which nearly all B−S bonds were
cleaved. Complete reduction to Et3N-BH3 did not occur in
reactions of the cationic complex [HNi(dmpe)2]+ with
B(SPh)3. Rather, Et3N-BH2(SPh) was the final product.
Although hydride transfer from [HNi(dmpe)2]+ to HB(SPh)2
is predicted to be unfavorable on the basis of the HA of
HB(SPH)2, the formation of a Ni−S bond in [Ni-
(dmpe)2(SPh)]+ and subsequent formation of Et3N−H2B-
(SPh) may facilitate hydride transfer. These fundamental
hydride transfer studies further our thermodynamic under-
standing of the factors that govern hydride transfer reactions to
BX3, especially in reactions where there is less driving force for
hydride transfer.
3
3
HCo(dmpe)2, Method B. [Co(dmpe)2(CH3CN)2][BF4]2 (0.461
g, 0.74 mmol) was dissolved in 40 mL of acetonitrile. KC8 (0.210 g,
1.54 mmol) was added slowly to the stirring solution, resulting in a
dark red mixture. After 20 min of stirring, H2 gas was vigorously
bubbled through the solution for 15 min, resulting in a color change to
yellow-brown. The mixture was filtered through Celite and the solvent
removed under vacuum conditions. The yellow-brown residue was
extracted with pentane (∼5−10 mL), filtered through a Celite plug,
and concentrated to a volume of 1−2 mL. The dark yellow solution
was passed through a plug of APTS-coated silica gel and the yellow
product eluted with pentane. The volatiles were removed under
vacuum conditions, affording a bright yellow solid. Yield: 0.160 g, 60%.
cis-[(H)2Co(dmpe)2][BF4]. KC8 (0.110 g, 0.81 mmol) was added
to a stirring solution of [Co(dmpe)2(CH3CN)2][BF4]2 (0.455 g, 0.74
mmol) in 40 mL of acetonitrile. Hydrogen gas was vigorously bubbled
through the red-brown solution, resulting in a color change to pale
yellow. The solution was filtered through Celite and the volatiles
removed under vacuum conditions. The remaining solids were
dissolved in a minimal amount of THF (∼5 mL) and filtered through
Celite. Slow diffusion of ether into the THF solution affords an
off-white microcrystalline solid of cis-[(H)2Co(dmpe)2][BF4]. Yield:
EXPERIMENTAL SECTION
■
Physical Measurements and General Procedures. NMR
spectra were recorded in thin-walled quartz NMR tubes (25 °C
unless otherwise noted) on a Varian Inova 500 MHz spectrometer
equipped with a Sun workstation or a 500 MHz Varian NMR S
system. 1H chemical shifts are referenced to residual protons in
deuterated solvent. 31P chemical shifts are proton-decoupled unless
otherwise noted and referenced to H3PO4 as an external reference. 11
B
chemical shifts are reported relative to BF3−OEt2 as an external
reference. 11B NMR spectra were integrated to determine the ratio of
total boron-containing products in solution. Infrared spectra were
recorded on a Nicolet Magna 860 FT-IR spectrometer at ambient
temperature and under a purge stream of nitrogen gas. Solid-state
FT-IR samples were prepared as KBr pellets. Elemental analysis
was performed by Columbia Analytical Services, Tucson, Arizona or
Atlantic Microlabs, Norcross, Georgia.
Synthesis and Materials. All synthetic procedures were
performed under an atmosphere of N2 using standard Schlenk or
glovebox techniques. Unless described otherwise, all reagents were
purchased from commercial sources and were used as received.
Solvents were dried by passage through activated alumina in an
Innovative Technology, Inc., PureSolv solvent purification system.
Deuterated THF was purchased from Cambridge Isotope Laboratories,
1
0.156 g, 47%. H NMR (THF-d8): δ 1.96 (s, 8H, PCH2CH2), 1.53
(s, 24H, P(CH3)2), −14.5 (broad singlet, 2H, CoH). 31P{1H} NMR
(25 °C) (THF-d8): δ 61.2 (br), 51.6 (br). IR (KBr): νCo−H 1877 cm−1.
ESI-MS (m/z+, exptl (calcd)): 360.0814 (360.0821). Anal. Calcd for
C12H34CoBF4P4: %C, 32.17; %H, 7.65. Found: %C, 32.43; %H, 7.92.
[Co(dmpe)2(CH3CN)][BF4]. Potassium graphite, KC8 (0.050 g,
0.37 mmol), was slowly added to a stirring solution of [Co(dmpe)2-
(CH3CN)2][BF4]2 (0.20 g, 0.35 mmol) in 15 mL of acetonitrile,
resulting in a dark red-purple solution. The solution was filtered
11923
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928

Inorganic Chemistry
Article
2
31P{1H} NMR (CD3CN): δ 68.8 (t, JPP = 63 Hz), 62.8 (br). Anal.
Calcd for C36H48CoBF4P4: %C, 57.62; %H, 6.44. Found: %C, 57.99; %
H, 6.33.
through Celite to remove graphite and dried under vacuum conditions.
The crude solid was extracted with THF and filtered through Celite
and dried under vacuum conditions, affording a dark brown solid.
Yield: 0.100 g, 64%. 1H NMR (THF-d8): δ 2.06 (br, 3H, CoNCCH3),
1.79 (s, 8H, PCH2CH2), 1.45 (s, 24H, P(CH3)2). 31P{1H} NMR
(THF-d8): δ 50.7 (s). IR (KBr): νCN 2243 cm−1. Anal. Calcd for
C14H35CoBF4NP4: %C, 34.52; %H, 7.24; %N, 2.87. Found: %C, 34.56;
%H, 6.76; %N, 1.99.
[Co(dedpe)2(CH3CN)][BF4]2. A solution of dedpe (1.28 g, 4.2
mmol) in 3 mL of acetonitrile was added to a stirring solution of
[Co(CH3CN)6][BF4]2 (1.00 g, 2.1 mmol) in 25 mL of acetonitrile,
resulting in a dark red-brown solution. After stirring for 12 h, the
solution was filtered through Celite and concentrated to ca. 10 mL.
[Co(dedpe)2(CH3CN)][BF4]2 was isolated by slowly diffusing ethyl
ether into the acetonitrile solution, affording a dark orange
[Co(dmpe)2(CH3CN)2][BF4]2. A solution of dmpe (1.60 g, 10.6
mmol) in 4 mL of acetonitrile was added to a stirring solution of
[Co(CH3CN)6][BF4]2 (2.54 g, 5.3 mmol) in 20 mL of acetonitrile,
resulting in a dark red-brown solution. After stirring for 12 h, the
solution was filtered through Celite and the solution concentrated
(ca. 10−12 mL). [Co(dmpe)2(CH3CN)2][BF4]2 was isolated as dark
red crystals by diffusing ethyl ether into the acetonitrile solution. The
red crystals desolvated and turned pale green over the period of several
weeks. The calculated crystalline yield includes two cocrystallized
acetonitrile molecules per [Co(dmpe)2(CH3CN)2][BF4]2 as observed
1
microcrystalline solid. Yield: 1.43 g, 78%. H NMR (acetonitrile-d3):
δ 10.3 (br, 8H, PCH2CH2), 7.2 (br, 20H, P(CH2CH3)2, 5.2 (br, 3H,
CH3CN), −1.8 (br, 20H, P(C6H5)2. No 31P NMR signal was observed.
IR (KBr): νCN 2272 cm−1. Anal. Calcd for C38H51CoB2F8NP4: %C,
51.97; %H, 5.85; %N, 1.59. Found: %C, 51.96; %H, 5.85; %N, 1.59.
Reaction of HCo(dmpe)2 with THF-BH3. A total of 20 μL
(0.02 mmol) of THF-BH3 (1.0 M solution) was added to a solution
of 0.014 g (0.04 mmol) of HCo(dmpe)2 in 1 mL of THF. The tube
was shaken to produce an orange-brown solution. The reaction was
characterized by 11B and 31P NMR spectroscopy. Spectral data
immediately after mixing: 11B{1H}, [1H-coupled] NMR (THF,
1
by X-ray crystallography. Yield: 3.33 g, 90%. H NMR (acetonitrile-
d3): δ −3.6 (br, 24H, P(CH3)2), −6.7 (br, 8H, PCH2CH2). No 31P
NMR signal was observed. Anal. Calcd for C14H35CoB2F8NP4: %C,
29.30; %H, 6.15; %N, 2.44. Found: %C, 29.34; %H, 6.01; %N, 2.47.
HCo(dedpe)2. KC8 (0.016 g, 0.12 mmol) was added to a stirring
solution of [Co(dedpe)2(CH3CN)][BF4]2 (0.100 g, 0.11 mmol) in
10 mL of acetonitrile resulting in a violet mixture. After stirring for
10 min, the solution was filtered through Celite to remove graphite.
Solvent was removed under vacuum conditions, and the brown solids
were dissolved in 10 mL of THF. Verkade’s base, (P(CH3NCH2-
CH2)3N; 0.024 g, 0.11 mmol) was added to the THF solution, and H2
gas was vigorously bubbled through the solution for 2 min, resulting in
a slow color change to deep red and the precipitation of [HP(CH3-
NCH2CH2)3N][BF4]. After stirring for 14 h, the red solution was
filtered through Celite, and the solvent was removed under vacuum
conditions. The red solids were extracted with pentane (∼5−10 mL),
filtered through a Celite plug, concentrated to a volume of 1−2 mL,
and passed through a plug of APTS-coated silica gel, eluted with
pentane. The red solution was concentrated and stored at −35 °C,
1
25 °C): −37.6 (s, [quintet], JBH = 81 Hz, [Co(dmpe)2][BH4],
8%), −38.3 (s, [quintet], 1JBH = 82 Hz, 25%), −39.0 (d, 1JBP = 56 Hz,
[dq], 1JBH = 95 Hz dmpe-(BH3)2, 67%). 31P{1H} NMR (THF): δ 51.8
(s, [Co(dmpe)2]+, 14%), 46.4 (s, HCo(dmpe)2, 82%), 8.9 (m, dmpe-
(BH3)2, 4%). Spectral data after 14 h at 25 °C: 11B{1H}, [1H-coupled]
1
1
NMR (THF, 25 °C): −38.8 (d, JBP = 55 Hz, [dq], JBH = 97 Hz,
dmpe-(BH3)2, 100%). 31P{1H} NMR (THF): δ 46.4 (s, HCo(dmpe)2),
8.9 (m, dmpe-(BH3)2).
Reaction of HCo(dmpe)2 with B(OC6F5)3. In a typical experi-
ment, solid B(OC6F5)3 was added to a stirring solution of HCo-
(dmpe)2 (0.030 mmol) in 2 mL of THF. The reaction was monitored
by 11B and 31P NMR spectroscopy. Spectral data immediately after
mixing with (0.011 mmol) B(OC6F5)3: 11B{1H}, [1H-coupled] NMR
(THF, 50 °C): δ 4.9 (s, [d], JBH = 137 Hz, [HB(OC6F5)3]−, 37%),
1
1.0 (s, [s], [B(OC6F5)4]−, 63%). 31P{1H} NMR (THF): δ 52.9 (s,
[Co(dmpe)2]+, 10%), 46.4 (br, HCo(dmpe)2, 88%), −14.7 (s,
unassigned, 2%). Spectral data after heating at 50 °C for 6 h: 11B{1H},
1
affording red crystals after 18 h. Yield: 0.047 g, 63%. H NMR (THF-
[1H-coupled] NMR (THF, 50 °C): δ 4.9 (s, [d], JBH = 137 Hz,
1
d8): δ 7.43−6.96 (m, 20H, P(C6H5)), 2.17 (br, 2H, PCH2CH2), 2.02
(br, 2H, PCH2CH2), 1.66 (br, 2H, PCH22CH2), 1.56 (m, 6H), 1.05 (m,
10H), 0.73 (m, 6H), −15.9 (quintet, JPH = 24 Hz, 1H, Co−H).
31P{1H} NMR (THF-d8): δ 79.1 (s), 74.2 (s). IR (KBr): νCo−H
1919 cm−1. Anal. Calcd for C36H49CoP4: %C, 65.06; %H, 7.43. Found:
%C, 65.39; %H, 7.40.
[HB(OC6F5)3]−, 31%), 1.0 (s, [s], [B(OC6F5)4]−, 61%), 0.62 (d, [d],
unassigned, 8%). 31P{1H} NMR (THF): δ 52.9 (s, [Co(dmpe)2]+,
30%), 46.4 (br, HCo(dmpe)2, 67%), −14.7 (s, unassigned, 3%).
Reaction of HCo(dmpe)2 with B(SPh)3. In an NMR tube,
0.0030 g (0.008 mmol) of solid B(SPh)3 was added to 0.014 g (0.04
mmol) of HCo(dmpe)2 in 1.5 mL of THF. The tube was shaken to
produce a dark orange solution. The reaction was immediately
monitored by 11B and 31P NMR spectroscopy. Spectral data
immediately after mixing: 11B{1H}, [1H-coupled] NMR (THF,
[(H)2Co(dedpe)2][BF4]. Hydrogen gas was bubbled through an
orange solution of [Co(dedpe)2(CH3CN)][BF4] (0.015 g, 0.019
mmol) in 2 mL of acetonitrile for 1 min, resulting in a pale yellow
solution and full conversion to the dihydride complex. Removal of the
volatiles under vacuum conditions results in a darkening of the
material and conversion back to the dark blue [Co(dedpe)2][BF4].
Room temperature 1H and 31P NMR spectra were exceptionally
broad and unresolved. 1H NMR (−40 °C; CD3CN): δ 8.03−6.83 (m,
P(C6H5)), 1.96 (s, PCH2CH2), 1.53 (s, P(CH3)2), −10.9 (quintet,
trans-[(H)2Co(dedpe)2]), −13.6, −14.0, −14.6 (broad multiplets,
cis-(H)2Co). 1H{31P} NMR (−40 °C; CD3CN): −10.9 (br, trans-
[(H)2Co(dedpe)2]), −13.6, −13.7, −14.1, −14.6 (br, cis-[(H)2Co-
(dedpe)2]). 31P NMR (21 °C; CD3CN): δ 89.50 (br), 82.94 (br),
25 °C): δ −14.7 (s, [t], JBH = 115 Hz, [H2B(SPh)2]−, 6%), −25.1
(d, JBP = 67 Hz, [doublet of triplets], JBH = 103 Hz, dmpe-(H2B-
(SPh))2, 40%), −38.7 (dd, JBP = 56 Hz, [dq], JBH = 97 Hz, dmpe-
(BH3)2, 54%). 31P{1H} NMR (THF): δ 44.7 (s, Co(dmpe)2(SPh),
9.31 (m, dmpe-(BH3)2), 0.22 (m, dmpe-(H2B(SPh))2).
1
1
1
1
1
Reaction of HCo(dmpe)2 with B(SPh)3 in the Presence of
NEt3. A total of 100 μL (0.72 mmol) of NEt3 was added to 0.014 g
(0.038 mmol) of HCo(dmpe)2 in 2 mL of THF, and 0.0035 g (0.010
mmol) of solid B(SPh)3 was added to this mixture. The tube was
shaken, and the reaction mixture began to gradually turn orange. The
reaction was immediately monitored by 11B and 31P NMR spectros-
copy, then again after being heated at 55 °C for 6 h. Spectral data
recorded immediately after mixing: 11B{1H}, [1H-coupled] NMR
2
71.50 (br). 31P NMR (−40 °C; CD3CN): δ 95.04 (t, JPP = 47 Hz),
92.14 (t, 2JPP = 47 Hz), 90.56 (br), 83.71 (br), 82.25 (br), 75.14 (br),
73.23 (br), 71.36 (br), 67.11 (br).
[Co(dedpe)2(CH3CN)][BF4]. Potassium graphite, KC8 (0.013 g,
0.096 mmol), was slowly added to a stirring solution of [Co(dedpe)2-
(CH3CN)][BF4]2 (0.075 g, 0.085 mmol) in 15 mL of acetonitrile,
resulting in a dark red-purple solution. The solution was filtered
through Celite to remove graphite and dried under vacuum conditions.
The crude solid was extracted with THF, filtered through Celite, and
dried under vacuum conditions, affording a dark blue solid. Yield:
(THF, 25 °C): δ 4.2 (s, [s], [B(SPh)4]−, 65%), −4.5 (s, [d], JB−H
=
130 Hz, [HB(SPh)3]−, 19%), −6.8 (s, [t], JBH = 120 Hz, Et3N-
BH2(SPh), 16%). 31P{1H} NMR (THF): δ 46.5 (br, HCo(dmpe)2,
53%), 44.8 (s, Co(dmpe)2(SPh), 47%). Spectral data after heating
6 h at 55 °C: 11B{1H}, [1H-coupled] NMR (THF, 25 °C): δ 21.3 (br,
1
1
[br], B−N oligomer, 13%), −6.8 (s, [t], JBH = 120 Hz, Et3N-
1
1
BH2(SPh), 16%), −12.8 (s, [q], JBH = 98 Hz, Et3N-BH3, 70%).
0.045 g, 70%. H NMR (CD3CN): δ 7.67−7.46 (m, 20H, P(C6H5)),
31P{1H} NMR (THF): δ 46.5 (br, HCo(dmpe)2, 3%), 44.8 (s,
2.36 (br, 4H, PCH2CH2), 1.96 (s, 3H, CoCH3CN)2), 1.46 (br, 4H,
PCH2CH2), 1.15 (m, 8H, PCH2CH3), 0.65 (m, 12H, PCH2CH3).
Co(dmpe)2(SPh), 97%).
11924
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928

Inorganic Chemistry
Article
Table 4. Crystallographic Data for HCo(dmpe)2, cis-[(H)2Co(dmpe)2][BF4], [Co(dmpe)2(CH3CN)][BF4], and
[Co(dmpe)2(CH3CN)2][BF4]2
HCo(dmpe)2
cis-[(H)2Co(dmpe)2] [BF4]
[Co(dmpe)2(CH3CN)] [BF4]
[Co(dmpe)2(CH3CN)2] [BF4]2
empirical formula
fw
C30H82.5Co2.5P10
360.19
C12H34BCoF4P4
448.01
C14H35BCoF4NP4
487.05
C20H44B2CoF8N4P4
697.02
color, habit
cryst syst
yellow, plates
monoclinic
C2/c
colorless, blocks
monoclinic
P2(1)/n
9.2170(10)
19.204(2)
12.2850(14)
90
orange, blocks
monoclinic
P2(1)/c
red, blocks
triclinic
space group
a, Å
P1
̅
41.817(2)
22.9904(11)
9.7774(5)
90
12.4445(3)
9.1456(2)
20.0098(4)
90
8.4911(5)
12.9468(8)
15.0180(9)
91.137(4)
92.512(4)
97.786(4)
1633.62(17)
1.54178
b, Å
c, Å
α, deg
β, deg
95.631(2)
90
97.625(5)
90
97.5820(10)
90
γ, deg
V (Å3)
9354.6(8)
(0.71073)
8
2155.2(4)
(0.71073)
4
2257.45(9)
1.54178
λ, Å (Cu (Mo), Kα)
Z
4
2
density (g/cm3)
temperature (K)
2θ range, deg
μ(Cu (Mo), Kα), mm−1
1.279
1.381
1.433
1.417
100(2)
135(2)
100(2)
100(2)
2.59−34.42
(1.242)
1.98−33.26
(1.118)
3.58−68.16
8.932
4.48−68.10
6.551
a
R(F), Rw(F)
0.0508, 0.1110
0.0471, 0.1081
0.0295, 0.0802
0.0562, 0.1421
a
Quantity minimized = R(wF2) = ∑[w(Fo2 − Fc2)2]/∑[(wFo )2]1/2; R = ∑Δ/∑(Fo), Δ = |(Fo − Fc)|, w = 1/[σ2(Fo ) + (aP)2 + bP], P = [2Fc2
+
2
2
Max(Fo,0)]/3.
Reaction of [HNi(dmpe)2][B(Ph)4] with B(SPh)3 in the
Presence of NEt3. B(SPh)3, 0.0040 g (0.012 mmol), was added to
a solution of [HNi(dmpe)2][B(Ph)4], 0.021 g (0.031 mmol), in 0.6
mL of 5% (v/v) triethylamine in THF, slowly producing an orange
solution. Heating the solution resulted in a red solution and the
precipitation of dark red crystals of [Ni(dmpe)2(SPh)][B(Ph)4].
The ratio of products determined by integration of the 11B NMR
spectra was difficult due to overlap of the [B(Ph)4]− anion with
Et3N-BH2(SPh). Spectral data immediately after mixing: 11B{1H},
[1H-coupled] NMR (THF, 25 °C): δ 62.1 (s, [s], B(SPh)3), −6.7 (s,
[s], [B(Ph)4]−). 31P{1H} NMR (THF, 25 °C): δ 32.3 (s,
[Ni(dmpe)2(SPh)]+, 1%), 25.7 (s, [HNi(dmpe)2]+, 99%). Spectral
data after 18 h at 25 °C: 11B{1H}, [1H-coupled] NMR (THF, 25 °C):
[Co(dmpe)2(CH3CN)2][BF4]2·2CH3CN. No symmetry higher than
triclinic was evident from the diffraction data. The data set was
twinned, and two independent domains were located using
CELL_NOW.31 The data set was treated with TWINABS32 absorption
corrections based on redundant multiscan data. A solution in the
centrosymmetric space group option P1 yielded chemically reasonable
̅
and computationally stable results of refinement. Two independent
molecules were located on special positions yielding Z = 2 and Z′ = 1.
The asymmetric unit also contains two molecules of acetonitrile
located in general positions. The ethyl backbone of one of the
chelating dmpe ligands was disordered and subsequently modeled
using a second free variable. All non-hydrogen atoms were refined with
anisotropic displacement parameters. All hydrogen atoms were treated
as idealized contributions. The goodness of fit on F2 was 1.011 with
R1(wR2) = 0.0537 (0.1365) for [Iθ > 2(I)] and with a largest
difference peak and hole of 0.655 and −0.471 e/Å3.
1
δ 62.1 (s, [s], B(SPh)3), 4.6 (s, [d], JBH = 135 Hz, Et3N-BH(SPh)2),
1
−7.1 (s, [t], JBH = 117 Hz, Et3N-BH2(SPh)), −12.4 (m, unassigned,
trace amt). 31P{1H} NMR (THF, 25 °C): δ 32.3 (s, [Ni-
[Co(dmpe)2(CH3CN)][BF4]. The systematic absences in the dif-
fraction data were consistent with the centrosymmetric, monoclinic
space group, P2(1)/c. The asymmetric unit contains one [Co-
(dmpe)2(CH3CN)]+ cation and one [BF4]− anion in general positions
yielding Z = 4 and Z′ = 1. All non-hydrogen atoms were refined with
anisotropic displacement parameters. All hydrogen atoms were treated
as idealized contributions. The goodness of fit on F2 was 1.038 with
R1(wR2) = 0.02957 (0.0802) for [Iθ > 2(I)] and with a largest
difference peak and hole of 0.497 and −0.321 e/Å3.
(dmpe)2(SPh)]+, 43%), 25.7 (s, [HNi(dmpe)2]+, 57%). Spectral
1
data after 24 h 50 °C: 11B NMR (THF, 25 °C): δ −7.1 (s, [t], JBH
=
117 Hz, Et3N-BH2(SPh)). 31P{1H} NMR (THF, 25 °C): δ 32.3 (s,
[Ni(dmpe)2(SPh)]+, 69%), 25.7 (s, [HNi(dmpe)2]+, 31%). Anal.
Calcd for [Ni(dmpe)2(SPh)][B(Ph)4], C42H57BNiP4S: %C, 64.07; %
H, 7.29. Found: %C, 63.77; %H, 7.32.
X-Ray Diffraction Studies. X-ray diffraction data were collected
on a Bruker-AXS Kappa APEX II CCD diffractometer with 0.71073 Å
̈
Mo Kα radiation or 1.54178 Å Cu Kα radiation. Selected crystals were
mounted using Paratone oil onto a glass fiber and cooled to the data
collection temperature of 100 K (135 K for cis-[(H)2Co(dmpe)2]-
[BF4]). For data collection using Mo Kα radiation unit cell parameters
were obtained from 60 data frames, 0.5° Φ, from three different
sections of the Ewald sphere. For data collection using Cu Kα
radiation, unit cell parameters were obtained from 90 data frames, 0.5°
Φ, from three different sections of the Ewald sphere. Cell parameters
were retrieved using APEX II software27 and refined using SAINT+28
on all observed reflections. Each data set was treated with SADABS29
absorption corrections based on redundant multiscan data. The
structure was solved by direct methods and refined by least-squares
method on F2 using the SHELXTL program package.30 Crystallo-
graphic data for each structure are listed in Tables 4 and 5. Details
regarding specific solution refinement for each compound are
provided in the following paragraphs.
cis-[(H)2Co(dmpe)2][BF4]. The systematic absences in the diffrac-
tion data were consistent with the centrosymmetric, monoclinic space
group, P2(1)/n. The molecule was located on a general position
yielding Z = 4 and Z′ = 1. All non-hydrogen atoms were refined with
anisotropic displacement parameters. The [(H)2Co(dmpe)2]+ cation
and the [BF4]− anion are disordered over two positions. The [BF4]−
anion positions were located from the difference map and the
molecules constrained to be equivalent using DFIX and EADP
commands. The [(H)2Co(dmpe)2]+ cation showed the chelating
phosphines in two different positions which were located from the
difference map and allowed to refine freely. The two hydride ligands
were located from the difference map and allowed to refine freely.
All other hydrogen atoms were treated as idealized contributions.
The goodness of fit on F2 was 1.020 with R1(wR2) = 0.0471(0.1081)
for [Iθ > 2(I)] and with a largest difference peak and hole of 0.778 and
−0.465 e/Å3.
11925
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928

Inorganic Chemistry
Article
Table 5. Crystallographic Data for HCo(dedpe)2, trans-[(H)2Co(dedpe)2][BF4]·THF, [Co(dedpe)2][BF4], and
[Co(dedpe)2(CH3CN)][BF4]2·CH3CN
HCo(dedpe)2
trans-[(H)2Co(dedpe)2] [BF4]·THF
[Co(dedpe)2][BF4]
[Co(dedpe)2(CH3CN)] [BF4]2·CH3CN
empirical formula
fw
C36H49CoP4
664.56
C40H58BOCoF4OP4
824.48
C72H96B2Co2F8P8
1500.73
C78H105B4Co2F16N3P8
1797.51
color, habit
cryst syst
space group
a, Å
red, blocks
orthorhombic
Pbca
colorless, blocks
monoclinic
P2(1)/n
10.5119(5)
25.6399(11)
14.7044(12)
90
blue, plates
triclinic
yellow, blocks
monoclinic
P2(1)/c
P1
̅
18.1830(17)
18.9977(17)
19.7711(18)
90
10.4373(13)
17.858(2)
20.266(3)
74.868(6)
82.200(5)
84.762(5)
3606.5(8)
0.71073
22.7320(6)
19.6783(5)
19.1975(5)
90
b, Å
c, Å
α, deg
β, deg
90
91.307(2)
90
91.1260(10)
90
γ, deg
V(Å3)
90
6829.6(11)
0.71073
8
3962.2(3)
0.71073
8585.9(4)
0.71073
λ, Å (Mo, Kα)
Z
4
2
4
density (g/cm3)
temperature (K)
2θ range, deg
μ(Mo, Kα), mm−1
1.293
1.382
1.382
1.391
100(2)
100(2)
100(2)
100(2)
1.86−33.19
0.714
1.60−33.53
0.645
1.97−33.48
0.699
1.37−33.17
0.614
a
R(F), Rw(F)
0.0399, 0.0839
0.0622, 0.1338
0.0728, 0.1873
0.0453, 0.1004
a
Quantity minimized = R(wF2) = ∑[w(Fo2 − Fc2)2]/∑[(wFo )2]1/2; R = ∑Δ/∑(Fo), Δ = |(Fo − Fc)|, w = 1/[σ2(Fo ) + (aP)2 + bP], P = [2Fc2
+
2
2
Max(Fo,0)]/3.
HCo(dmpe)2. The systematic absences in the diffraction data were
consistent with the centrosymmetric, monoclinic space group, C2/c.
The asymmetric unit contained three independent molecules with one
located on a special position and two on general positions yielding
Z = 8 and Z′ = 2.5. All non-hydrogen atoms were refined with
anisotropic displacement parameters. The hydrogen atoms were
treated as idealized contributions except for the cobalt hydride
atoms that were located from the difference map and allowed to refine
with all three cobalt hydride distances made equivalent to one another,
as would be expected chemically. The three molecules all exhibit
disorder in the chelating phosphine ligands over two positions.
Modeling of this disorder was successful for two out of the three
molecules, with the third being minor and unable to be located from
the difference map. The molecule that sits on the special position also
has a hydride disordered over two positions about the symmetry
element. The goodness of fit on F2 was 1.025 with R1(wR2) =
0.0508(0.1110) for [Iθ > 2(I)] and with a largest difference peak and
hole of 1.674 and −2.364 e/Å3. The minor residual density is due to
the unresolved disorder about the phosphine ligands described above.
[Co(dedpe)2(CH3CN)][BF4]2·CH3CN. The systematic absences in the
diffraction data were consistent with the centrosymmetric, monoclinic
space group, P2(1)/c. The asymmetric unit contains two [Co-
(dedpe)2(CH3CN)]+ cations, four [BF4]− anions, and one acetonitrile
molecule yielding Z = 4 and Z′ = 2. Two of the [BF4]− anions were
disordered over two positions and modeled as such using a second
free variable. The acetonitrile molecule was disordered about the
nitrogen and was split to eliminate the large thermal parameter. All
non-hydrogen atoms were refined with anisotropic displacement
parameters. All hydrogen atoms were treated as idealized contribu-
tions. The goodness of fit on F2 was 1.013 with R1(wR2) =
0.0453(0.1004) for [Iθ > 2(I)] and with a largest difference peak and
hole of 1.127 and −0.829 e/Å3.
appropriately. This disorder does not change the overall identity of the
molecules. All non-hydrogen atoms were refined with anisotropic
displacement parameters. Hydrogen atoms were treated as idealized
contributions. The goodness of fit on F2 was 1.024 with R1(wR2) =
0.0728(0.1873) for [Iθ > 2(I)] and with a largest difference peak and
hole of 5.213 and −0.621 e/Å3. The large residual density is due to
disorder of the chelating phosphine ligand about a symmetry element
described above.
trans-[(H)2Co(dedpe)2][BF4]·THF. The systematic absences in the
diffraction data were consistent with the centrosymmetric, monoclinic
space group, P2(1)/n. The asymmetric unit contains one [(H)2Co-
(dedpe)2]+ cation, one [BF4]− anion, and one molecule of THF
located on general positions yielding Z = 4 and Z′ = 1. Both the THF
molecule and the [BF4]− anion were disordered over two positions,
which were located from the difference map, and refined on their own
individual free variable. One ethyl group of the dedpe ligand was
disordered over two positions and refined similarly. All non-hydrogen
atoms were refined with anisotropic displacement parameters.
Hydrogen atoms were treated as idealized contributions except for
the hydrogens attached to Co(1), which were located from the
difference map and allowed to refine freely. The goodness of fit on F2
was 1.004 with R1(wR2) = 0.0622(0.1338) for [Iθ > 2(I)] and with a
largest difference peak and hole of 0.584 and −0.984 e/Å3.
HCo(dedpe)2. The systematic absences in the diffraction data were
consistent with the centrosymmetric, orthorhombic space group, Pbca.
The molecule was located on a general position yielding Z = 8 and
Z′ = 1. All non-hydrogen atoms were refined with anisotropic
displacement parameters. Hydrogen atoms were treated as idealized
contributions except for the hydrogen attached to Co(1), which was
located from the difference map and allowed to refine freely. The
goodness of fit on F2 was 1.010 with R1(wR2) = 0.0399(0.0839) for
[Iθ > 2(I)] and with a largest difference peak and hole of 0.543 and
−0.312 e/Å3.
Computational Methods. Electronic structure calculations using
the NWChem suite of programs33 were performed to determine
hydride affinities of BX3 compounds in Table 3. Minimum energy
structures of the boranes and their corresponding borohydrides were
optimized using Density Functional Theory,34 the B3LYP functional,35
and DZVP2 basis set.36 Minimum energy geometries were confirmed
to be bound states by the absence of imaginary frequencies, and
transition states were confirmed by the presence of exactly one
[Co(dedpe)2][BF4]. No symmetry higher than triclinic was evident
from the diffraction data. Solution in the centrosymmetric space group
P1 yielded chemically reasonable and computationally stable results of
̅
refinement. The asymmetric unit contains four [Co(dedpe)2]+ cations
located on special positions and two [BF4]− anions located on general
positions yielding Z = 2 and Z′ = 2. Two of the four [Co(dedpe)2]+
cations are disordered about their symmetry elements, leading to two
equivalent chelating phosphine positions. The disorder is minimal as
determined by free variable refinement and could not be modeled
11926
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928

Inorganic Chemistry
Article
imaginary frequency. Enthalpies and free energies at 298 K of
optimized structures were calculated from frequency calculations using
the harmonic oscillator-rigid-rotor approximation with entropies
corrected for rotational symmetry number.37 HAs of BX3 were
determined from the calculated enthalpy for the isodesmic metathesis
J. Hydrogen Energy 2008, 33, 608. (c) Davis, B. L.; Dixon, D. A.;
Garner, E. B.; Gordon, J. C.; Matus, M. H.; Scott, B.; Stephens, F. H.
Angew. Chem., Int. Ed. 2009, 48, 6812. (d) Hamilton, C. W.; Baker, R.
T.; Staubitz, A.; Manners, I. Chem. Soc. Rev. 2009, 38, 279. (e) Sutton,
A. D.; Burrell, A. K.; Dixon, D. A.; Garner, E. B.; Gordon, J. C.;
Nakagawa, T.; Ott, K. C.; Robinson, J. P.; Vasiliu, M. Science 2011,
331, 1426.
(5) Mock, M. T.; Potter, R. G.; Camaioni, D. M.; Li, J.; Dougherty,
W. G.; Kassel, W. S.; Twamley, B.; DuBois, D. L. J. Am. Chem. Soc.
2009, 131, 14454.
38
−
reaction 7 with BH4 and HA = 73.7 kcal/mol for BH3.
(7)
(8)
Here, we used single point energies calculated with the aug-cc-pVTZ
basis set39 and the DZVP2 optimized geometries and associated
thermal corrections. We deemed this protocol satisfactory on the basis
of the following results. Grant et al.38a calculated HA = 70.5 kcal/mol
for BF3 using couple cluster theory. We calculated 79.1 and 75.0 at
B3LYP/DZVP2 and B3LYP/6-311+G**40 levels. Calculations using
single point energies at the B3LYP/aug-cc-pVTZ//DZVP2 level give
72.5 kcal/mol.
(6) (a) Berning, D. E.; Noll, B. C.; DuBois, D. L. J. Am. Chem. Soc.
1999, 121, 11432. (b) Curtis, C. J.; Miedaner, A.; Ellis, W. W.; DuBois,
D. L. J. Am. Chem. Soc. 2002, 124, 1918.
(7) Berning, D. E.; Miedaner, A.; Curtis, C. J.; Noll, B. C.; DuBois,
M. R.; DuBois, D. L. Organometallics 2001, 20, 1832.
(8) Raebiger, J. W.; Miedaner, A.; Curtis, C. J.; Miller, S. M.; DuBois,
D. L. J. Am. Chem. Soc. 2004, 126, 5502.
(9) Curtis, C. J.; Miedaner, A.; Raebiger, J. W.; DuBois, D. L.
Organometallics 2004, 23, 511.
ASSOCIATED CONTENT
* Supporting Information
(10) (a) Price, A. J.; Ciancanelli, R.; Noll, B. C.; Curtis, C. J.; DuBois,
D. L.; DuBois, M. R. Organometallics 2002, 21, 4833. (b) Miedaner, A.;
Raebiger, J. W.; Curtis, C. J.; Miller, S. M.; DuBois, D. L.
Organometallics 2004, 23, 2670.
(11) Camaioni, D. M.; Heldebrant, D. J.; Linehan, J. C.; Shaw, W. J.;
Li, J.; DuBois, D. L.; Autrey, T. Prepr. Pap.-Am. Chem. Soc., Div. Fuel
Chem. 2007, 52, 509.
(12) Ciancanelli, R.; Noll, B. C.; DuBois, D. L.; Rakowski DuBois, M.
J. Am. Chem. Soc. 2002, 124, 2984.
(13) Schunn, R. A. Inorg. Chem. 1970, 9, 2567.
(14) Kisanga, P. B.; Verkade, J. G.; Schwesinger, R. J. Org. Chem.
■
S
Experimental details and cyclic voltammetry data for [Co(dmpe)2-
(CH3CN)2][BF4]2 and [Co(dedpe)2(CH3CN)][BF4]2; full X-ray
diffraction refinement data (CIF) for HCo(dmpe)2, cis-[(H)2Co-
(dmpe)2][BF4], [Co(dmpe)2(CH3CN)][BF4], [Co-
(dmpe)2(CH3CN)2][BF4]2, HCo(dedpe)2, trans-[(H)2Co-
(dedpe)2 ][BF4 ]·THF, [Co(dedpe)2 ][BF4 ], [Co-
(dedpe)2(CH3CN)][BF4]2·CH3CN, Co(dmpe)2(SPh)·HSPh,
and [Ni(dmpe)2(SPh)][B(Ph)4]; NMR spectral data for the
reaction of [Co(dedpe)2(CH3CN)]+ and H2; calculated electronic
energies and coordinates for structures in Table 3 and the
complete ref 33. This material is available free of charge via Inter-
net at http://pubs.acs.org.
2000, 65, 5431.
(15) Rossin, A.; Caporali, M.; Gonsalvi, L.; Guerri, A.; Lledos, A.;
Peruzzini, M.; Zanobini, F. Eur. J. Inorg. Chem. 2009, 3055.
(16) (a) Findlater, M.; Bernskoetter, W. H.; Brookhart, M. J. Am.
Chem. Soc. 2010, 132, 4534. (b) Rybtchinski, B.; Ben-David, Y.;
Milstein, D. Organometallics 1997, 16, 3786. (c) Guggenberger, L. J.
Inorg. Chem. 1973, 12, 1317.
AUTHOR INFORMATION
Corresponding Author
*E-mail: michael.mock@pnnl.gov.
■
(17) Holah, D. G.; Hughes, A. N.; Maciaszek, S.; Magnuson, V. R.;
Parker, K. O. Inorg. Chem. 1985, 24, 3956.
(18) Qi, X.-J.; Fu, Y.; Liu, L.; Guo, Q.-X. Organometallics 2007, 26,
4197.
(19) (a) Baker, M. V.; Field, L. D. J. Chem. Soc., Chem. Commun.
1984, 996. (b) Bau, R.; Yuan, H. S. H.; Baker, M. V.; Field, L. D. Inorg.
Chim. Acta 1986, 114, L27. (c) Baker, M. V.; Field, L. D. Appl.
Organomet. Chem. 1990, 4, 543.
(20) Lippard, S. J.; Melmed, K. M. J. Am. Chem. Soc. 1967, 89, 3929.
(21) Schneider, J. J.; Goddard, R.; Kruger, C. Z. Naturforsch., B:
Chem. Sci. 1995, 50, 448.
ACKNOWLEDGMENTS
■
This work was supported by the U.S. Department of Energy’s
(DOE) Office of Energy Efficiency and Renewable Energy,
Center of Excellence for Chemical Hydrogen Storage. M.J.O.
was supported as part of the Center for Molecular Electro-
catalysis, an Energy Frontier Research Center funded by the
U.S. Department of Energy, Office of Basic Energy Sciences,
under FWP 56073. This research was performed in part using
the Molecular Science Computing Facility in the William R.
Wiley Environmental Molecular Sciences Laboratory, a U.S.
Department of Energy (DOE) national scientific user facility
located at the Pacific Northwest National Laboratory. The
Pacific Northwest National Laboratory is operated by Battelle
for DOE. The authors wish to thank Charles Campana for his
assistance with the structural solution for HCo(dmpe)2.
(22) Knizek, J.; Noth, H. J. Organomet. Chem. 2000, 614−615, 168.
̈
(23) Heintz, R. A.; Smith, J. A.; Szalay, P. S.; Weisgerber, A.; Dunbar,
K. R.; Beck, K.; Coucouvanis, D. Inorg. Synth. 2002, 33, 75.
(24) Miedaner, A.; DuBois, D. L.; Curtis, C. J. Organometallics 1993,
12, 299.
(25) Michaelis, A.; Hillringhaus, F. A. Ann. 1901, 315, 41.
(26) (a) Colclough, T.; Gerrard, W.; Lappert, M. F. J. Chem. Soc.
1955, 907. (b) Colclough, T.; Gerrard, W.; Lappert, M. F. J. Chem. Soc.
1956, 3006.
(27) APEX II, v. 2009.3; Bruker AXS: Madison, WI, 2009.
(28) SAINT+, v. 7.56A; Bruker AXS: Madison, WI, 2009.
(29) SADABS, v. 2008/1; Bruker AXS Inc.: Madison, WI, 2008.
(30) Sheldrick, G. M. SHELXTL, v. 2008; Bruker AXS Inc.: Madison,
WI, 2008.
REFERENCES
■
(1) (a) Kubas, G. J. Chem. Rev. 2007, 107, 4152. (b) Heinekey, D.
M.; Oldham, W. J. Chem. Rev. 1993, 93, 913. (c) Jessop, P. G.; Morris,
R. H. Coord. Chem. Rev. 1992, 121, 155.
(31) CELL_NOW, v. 2008/2; Bruker AXS: Madison, WI, 2008.
(32) TWINABS, v. 2008/2; Bruker AXS: Madison, WI, 2008.
(33) Bylaska, E. J.; et al. NWChem, version 5.1.1; Pacific Northwest
National Laboratory: Richland, WA.
(34) (a) Hohenberg, P.; Kohn, W. Phys. Rev. 1964, 136, B864.
(b) Kohn, W.; Sham, L. J. Phys. Rev. 1965, 140, A1133.
(35) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(2) DuBois, D. L.; Blake, D. M.; Miedaner, A.; Curtis, C. J.; DuBois,
M. R.; Franz, J. A.; Linehan, J. C. Organometallics 2006, 25, 4414.
(3) (a) Marder, T. B. Angew. Chem., Int. Ed. 2007, 46, 8116.
(b) Stephens, F. H.; Pons, V.; Baker, T. R. J. Chem. Soc., Dalton. Trans.
2007, 25, 2613.
(4) (a) Ramachandran, P. V.; Gagare, P. D. Inorg. Chem. 2007, 46,
7810. (b) Hausdorf, S.; Baitalow, F.; Wolf, G.; Mertens, F. O. R. L. Int.
11927
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928

Inorganic Chemistry
Article
(36) Godbout, N.; Salahub, D. R.; Andzelm, J.; Wimmer, E. Can. J.
Chem. 1992, 70, 560.
(37) Fernandez-Ramos, A.; Ellingson, B.; Meana-Paneda, R.;
́
̃
Marques, J.; Truhlar, D. Theor. Chem. Acc. 2007, 118, 813.
(38) (a) Grant, D. J.; Dixon, D. A.; Camaioni, D. M.; Potter, R. G.;
Christe, K. O. Inorg. Chem. 2009, 48, 8811. (b) Workman, D. B.;
Squires, R. R. Inorg. Chem. 1988, 27, 1846.
(39) (a) Dunning, T. H. Jr. J. Chem. Phys. 1989, 90, 1007.
(b) Kendall, R. A.; Dunning, T. H. Jr.; Harrison, R. J. J. Chem. Phys.
1992, 96, 6796. (c) Woon, D. E.; Dunning, T. H. Jr. J. Chem. Phys.
1993, 98, 1358.
(40) Schafer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994, 100,
5829.
11928
dx.doi.org/10.1021/ic200857x|Inorg. Chem. 2011, 50, 11914−11928