Reactivity of ruthenium phosphido species generated through the deprotonation of a tripodal phosphine ligand and implications for hydrophosphination

Article abstract of DOI:10.1021/ja5006724

The fragmentation of the 1,1,2-tris(diphenylphosphino)ethane ligand in [RuCp((Ph₂P)₂CHCH₂PPh₂)][PF ₆] (1) was explored through treatment with base under aprotic conditions. The neutral phosphido complex RuCp(PPh₂CH - CHPPh ₂)(PPh₂) (2) with a (Z)-1,2-bis(diphenylphosphino)ethene (dppen) ligand was generated through a base-facilitated dehydrophosphination reaction. Installation of a bis(p-tolyl)phosphido ligand was attempted by combining bis(p-tolyl)phosphine with RuCp(dppen)Cl in the presence of KOtBu, but surprisingly, the unsymmetrical diphenylphosphido compound RuCp(Ph ₂PCHCHP(p-tol)₂)(PPh₂) (5) was generated instead. The ligand rearrangement reaction was driven by the greater electron density on the bis(p-tolyl)phosphido moiety. Density functional theory calculations showed that fragmentation to the 1,2-disubstituted ligand was thermodynamically favored over the 1,1-disubstituted ligand and that intramolecular phosphido exchange was kinetically accessible at room temperature. The greater basicity of the bis(p-tolyl)phosphido ligand was experimentally verified by the measured pKaTHF of 28 for the acid/base pair [RuCp(Ph₂P(o-C₆H₄)PPh₂)(P(p-tolyl) ₂H)]⁺/RuCp(Ph₂P(o-C₆H ₄)PPh₂)(P(p-tolyl)₂) versus 25 for the acid/base pair [RuCp(Ph₂P(o-C₆H₄)PPh ₂)(PPh₂H)]⁺/RuCp(Ph₂P(o-C ₆H₄)PPh₂)(PPh₂) (7). For comparison, the approximate pKa THF values for free P(p-tolyl)₂H/[K(crypt)]P(p- tolyl)₂ and free PPh₂H/[K(crypt)]PPh₂ are 43 and 38, respectively. This is the first quantitative measurement of the large effect that coordination to a metal center, in this case ruthenium(II), has on the acidity of secondary phosphines. This is useful information for designing and understanding hydrophosphination catalysts. Complexes 2 and 7 are catalysts for the addition of PPh₂H to acrylonitrile, but they deactivate fairly rapidly. The pKa THF measurements are consistent with a catalytic cycle involving a Michael addition step. Complex 2 in solution underwent a slow, unprecedented rearrangement of P-C, C-C, and C-H bonds to give crystalline Ru(C₅(CH₃)₄(CH₂C₆H ₅))(Ph₂PCH₂CH₂PPh(o-C ₆H₄)PPh) (9) in high yields, demonstrating the unpredictable reactivity of phosphido ligands.

Full text of DOI:10.1021/ja5006724

Article
pubs.acs.org/JACS
Reactivity of Ruthenium Phosphido Species Generated through the
Deprotonation of a Tripodal Phosphine Ligand and Implications for
Hydrophosphination
Peter E. Sues, Alan J. Lough, and Robert H. Morris*
Department of Chemistry, University of Toronto, Toronto, Ontario M5S 3H6, Canada
S
* Supporting Information
ABSTRACT: The fragmentation of the 1,1,2-tris-
(diphenylphosphino)ethane ligand in [RuCp*-
((Ph2P)2CHCH2PPh2)][PF6] (1) was explored through
treatment with base under aprotic conditions. The neutral
phosphido complex RuCp*(PPh2CHCHPPh2)(PPh2) (2)
with a (Z)-1,2-bis(diphenylphosphino)ethene (dppen) ligand
was generated through a base-facilitated dehydrophosphina-
tion reaction. Installation of a bis(p-tolyl)phosphido ligand was
attempted by combining bis(p-tolyl)phosphine with RuCp*-
(dppen)Cl in the presence of KOtBu, but surprisingly, the unsymmetrical diphenylphosphido compound RuCp*(Ph2PCHCHP-
(p-tol)2)(PPh2) (5) was generated instead. The ligand rearrangement reaction was driven by the greater electron density on the
bis(p-tolyl)phosphido moiety. Density functional theory calculations showed that fragmentation to the 1,2-disubstituted ligand
was thermodynamically favored over the 1,1-disubstituted ligand and that intramolecular phosphido exchange was kinetically
accessible at room temperature. The greater basicity of the bis(p-tolyl)phosphido ligand was experimentally verified by the
measured pKTa HF of 28 for the acid/base pair [RuCp*(Ph2P(o-C6H4)PPh2)(P(p-tolyl)2H)]+/RuCp*(Ph2P(o-C6H4)PPh2)(P(p-
tolyl)2) versus 25 for the acid/base pair [RuCp*(Ph2P(o-C6H4)PPh2)(PPh2H)]+/RuCp*(Ph2P(o-C6H4)PPh2)(PPh2) (7). For
comparison, the approximate pKTa HF values for free P(p-tolyl)2H/[K(crypt)]P(p-tolyl)2 and free PPh2H/[K(crypt)]PPh2 are 43
and 38, respectively. This is the first quantitative measurement of the large effect that coordination to a metal center, in this case
ruthenium(II), has on the acidity of secondary phosphines. This is useful information for designing and understanding
hydrophosphination catalysts. Complexes 2 and 7 are catalysts for the addition of PPh2H to acrylonitrile, but they deactivate
fairly rapidly. The pKTa HF measurements are consistent with a catalytic cycle involving a Michael addition step. Complex 2 in
solution underwent a slow, unprecedented rearrangement of P−C, C−C, and C−H bonds to give crystalline
Ru(C5(CH3)4(CH2C6H5))(Ph2PCH2CH2PPh(o-C6H4)PPh) (9) in high yields, demonstrating the unpredictable reactivity of
phosphido ligands.
reactivity similar to that of terminal phosphidos in that they
are very basic and nucleophilic; they have even been reported
to deprotonate phenyl groups.17,28,36
INTRODUCTION
■
The coordination chemistry of negatively charged deprotonated
secondary phosphines, or phosphido ligands, is incredibly
diverse, and the reactivity of these ligands is contingent on their
specific coordination mode.1 Bridging phosphidos, which form
relatively stable metal clusters or bimetallic species, tend to be
less reactive because both lone pairs on phosphorus are
occupied in metal−phosphorus bonds,1−12 whereas trigonal-
pyramidal phosphidos or terminal phosphidos, where the
phosphorus atom is bound to only one metal center, are much
more nucleophilic because of their unquenched lone pair.1,13−21
As such, terminal phosphidos are known to react with various
electrophiles, including HX, MeI, and Et3SiH, among several
other reagents.15,16,22−26 The third possible coordination mode
for phosphidos occurs when a metal ion would otherwise be
coordinatively unsaturated without additional bonding from the
phosphorus-based ligand.1,17,27−35 In this case, the lone pair on
phosphorus forms a π bond with the metal center and the
phosphido adopts a trigonal-planar geometry. Despite the
metal−phosphorus double bond, planar phosphidos have
Phosphido ligands are especially important in the late-
transition-metal-catalyzed hydrophosphination of unsaturated
substrates, as conventional organometallic catalytic mechanisms
involve the insertion of an alkene or alkyne into the M−P bond
of a phosphido intermediate.22−26,37−59 A typical organo-
metallic mechanism for alkene hydrophosphination involves
oxidative addition of a primary or secondary phosphine to an
electron-rich metal center followed by coordination of the
olefin substrate and subsequent insertion to form a P−C bond
(Scheme 1).22−26,37,38,41,42,44,47,48,50−63 Reductive elimination
of the hydride and alkyl moieties completes the catalytic cycle,
generating the hydrophosphinated product and recovering the
electron-rich metal center (Scheme 1).22−26,37,38,41,42,44,47−63
Received: January 27, 2014
Published: February 25, 2014
© 2014 American Chemical Society
4746
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760

Journal of the American Chemical Society
Article
Scheme 1. (Right) Organometallic and (Left) Michael Addition Mechanisms for the Hydrophosphination of
Alkenes22−26,37,38,41,42,44,47,48,50−63
Another possible mechanism for hydrophosphination that
also includes metal phosphido intermediates is a Michael
addition route, where attack of the phosphido ligand on an
alkene is followed by protonation of the resulting carbanion
(Scheme 1).22−26,37,38,41,42,44,47−54,56−63 Subsequent coordina-
tion and deprotonation of a phosphine then closes the catalytic
cycle. Exploring the behavior and reactivity of phosphido
ligands is therefore crucial for better understanding either an
organometallic or Michael addition mechanism of hydro-
phosphination as well as for developing more efficient and
selective catalytic systems.
In this paper, we explore the fragmentation of the 1,1,2-
tris(diphenylphosphino)ethane (tppe) ligand in [RuCp*-
(tppe)][PF6] (1) upon treatment of the cationic ruthenium
complex with KOtBu under aprotic conditions. The base
facilitates dehydrophosphination of the tppe ligand, producing
a deprotonated ruthenium phosphido species. In addition, we
examine the unexpected exchange of a phosphido ligand with a
phosphine moiety of an “inert” bidentate phosphine ligand and
compare the measured pKa values of free and coordinated
secondary phosphines in order to understand the high
nucleophilicity of phosphido groups here and in catalytic
hydrophosphination reactions. Lastly we discuss a remarkable
rearrangement of the deprotonated ruthenium complex, which
illustrates the diverse and sometimes unpredictable reactivity of
phosphido ligands.
Scheme 2. Synthesis of the Ruthenium Half-Sandwich
Phosphido Species 2
respectively. Upon deprotonation, a doublet representing the
dppen phosphorus nuclei appeared downfield at 79.2 ppm,
while the triplet representing the phosphido donor was upfield
at 23.2 ppm. If deprotonation of the trisubstituted carbon had
occurred without carbon−phosphorus bond cleavage, then the
signal for the equivalent phosphines would remain upfield
because of the adjacent localized negative charge, as was seen
by Ruiz and co-workers.65,66 Examination of the two-dimen-
sional (2D) gradient heteronuclear single-quantum coherence
(gHSQC) spectrum revealed the presence of alkene carbons at
147 ppm in the 13C NMR spectrum correlated to vinylic
1
protons at around 7.3 ppm in the H NMR spectrum (no
methylene protons were detected). This identified the presence
of the dppen unsaturated ligand backbone, while the chemical
shift at 23.2 ppm in the 31P{1H} NMR spectrum was diagnostic
of a phosphido ligand (this value is consistent with the chemical
shifts seen by the Gladysz group for similar ruthenium half-
sandwich phosphido complexes18).
RESULTS AND DISCUSSION
Base-induced bond cleavage to produce a phosphido ligand is
a process similar to other phosphorus−carbon bond cleavage
reactions known in the literature.67−71 Moreover, the KOtBu-
promoted dehydrophosphination of the tppe ligand is the
reverse of several related hydrophosphination reactions.41,44
X-ray diffraction (XRD) studies revealed that complex 2
adopts a distorted-piano-stool geometry in which the Cp*
ligand occupies one-half of the coordination sphere while the
dppen ligand and a trigonal-pyramidal diphenylphosphido
ligand occupy the other three coordination sites (Figure 1).
In the parent compound 1 there is a considerable amount of
ring strain in the tripodal phosphine ligand, as reflected by the
P−Ru−P angles, which are all compressed below the optimal
octahedral angle of 90° [P(1)−Ru(1)−P(2), P(1)−Ru(1)−
P(3), and P(2)−Ru(1)−P(3) bond angles of 79.0(2), 80.8(2),
and 69.4(2)°, respectively].64 The base-induced elimination of
the diphenylphosphido and formation of dppen to give 2,
however, relieves the geometric constraints imposed by the
tripodal ligand architecture. This allows the phosphorus donors
to move apart and expand the P−Ru−P bond angles: the
bidentate phosphine ligand has a P(1)−Ru(1)−P(2) bite angle
■
Deprotonation of Parent Complex 1 To Yield
Complex 2. The previously reported ruthenium complex 1
was found to be air- and moisture-stable as well as completely
unreactive toward small molecules, even at elevated temper-
atures.64 However, reaction with a strong base such as KOtBu
or KH under aprotic conditions (in THF, toluene, or benzene)
caused a dramatic color change from light yellow to deep red.
Upon further inspection, the reaction with base resulted in
quantitative conversion of the parent compound into a new
deprotonated species, 2, (Scheme 2) which could be isolated as
a brick-red powder in excellent yield (94%).
The NMR spectra of 2 signaled the formation of cis-1,2-
bis(diphenylphosphino)ethene (dppen, also known as dppv)
and diphenylphosphido ligands, suggesting that deprotonation
occurred at the disubstituted carbon instead of the
trisubstituted carbon and caused an elimination reaction to
occur within the ruthenium complex. The 31P{1H} NMR
spectrum of the parent complex 1 displayed a triplet at 60.8
ppm and a doublet at 39.8 ppm for the single inequivalent
phosphorus donor and the two equivalent phosphorus donors,
4747
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760

Journal of the American Chemical Society
Article
Scheme 3. Acid/Base Reactivity of Complex 2
center. In the 31P{1H} NMR spectrum, the triplet representing
the phosphido group shifted downfield from 23.2 ppm in 2 to
37.9 ppm in 3 and displayed a strong P−H coupling (with a
coupling constant of JP−H = 357 Hz). In the 1H NMR
spectrum, a P−H resonance could also be detected at 6 ppm as
a doublet of triplets due to a strong P−H coupling (JP−H = 357
Hz) as well as weaker coupling to the other phosphorus nuclei
(JP−H = 6 Hz). In addition, no methylene protons could be
detected, and much like 2, analysis of the 2D gHSQC spectrum
revealed the presence of a dppen ligand (olefinic carbons at 146
ppm in the 13C NMR spectrum coupled to vinylic protons at
Figure 1. ORTEP3 representation and atom numbering for 2 (thermal
ellipsoids at 50% probability; most of the hydrogens omitted for
clarity).
1
around 7.5 ppm in the H NMR spectrum).
Along with NMR experiments, complex 3 was also
characterized utilizing XRD. The ruthenium center exhibits a
distorted-piano-stool geometry very similar to that of 2, with
Cp* opposite cis dppen and diphenylphosphine ligands (Figure
2). Upon protonation of the phosphido ligand, the P−Ru−P
of 82.48(9)°, while the P(1)−Ru(1)−P(3) and P(2)−Ru(1)−
P(3) bond angles are 91.12(6) and 96.01(6)°, respectively.
Furthermore, in the parent complex 1, all of the Ru−P
distances are greater than 2.3 Å, whereas in 2 the Ru(1)−P(1)
and Ru(1)−P(2) bond lengths, representing the dppen donors,
contract to 2.269(2) and 2.254(2) Å, respectively.64 Con-
versely, the Ru(1)−P(3) bond distance lengthens to 2.461(2)
Å, which is characteristic of terminal phosphidos (see Table 1
for other notable bond lengths and angles).14,15,17 The long
ruthenium−phosphorus distance is attributed to the “transition-
metal gauche effect”, in which the lone pair on phosphorus is
repelled by a filled d orbital on the metal center, and the
conformation that the phosphido ligand adopts minimizes this
electronic repulsion as well as steric interactions.13,14
Protonation and Regeneration of Complex 2 Using
Acid and Base. In order to probe the basicity and
nucleophilicity of the phosphido ligand, the deprotonated
species was exposed to acid under moisture-free conditions.
When a solution of 2 in toluene was treated with an excess of
HCl in ether, the cationic diphenylphosphine complex 3 could
be isolated as a yellow powder in good yields (77%).
Furthermore, complex 2 could be recovered quantitatively
from 3 through deprotonation with excess base, indicating that
the protonation and deprotonation of the phosphido ligand are
completely reversible (Scheme 3).
Figure 2. ORTEP3 representation and atom numbering for 3 (thermal
ellipsoids at 50% probability; solvent and most of the hydrogens
omitted for clarity).
bond angles remain unchanged for the most part (Table 1)
while the Ru−P bonds for the dppen ligand lengthen by about
0.04 Å and the Ru(1)−P(3) bond contracts by 0.14 Å [from
1
The 31P and H NMR spectra of 3 in THF-d8 unequivocally
demonstrated that protonation occurred at the phosphido
Table 1. Selected Bond Lengths (Å) and Angles (deg) for 1, 2, 3, 5, 7, and 9
a
1
2
3
5
7
9
Bond Lengths (Å)
Ru(1)−P(1)
Ru(1)−P(2)
Ru(1)−P(3)
C(1)−C(2)
2.309(5)
2.333(6)
2.307(6)
1.56(4)
2.265(2)
2.253(2)
2.461(4)
1.316(9)
2.301(2)
2.302(2)
2.324(2)
1.326(9)
2.276(3)
2.258(3)
2.468(3)
1.31(2)
2.254(2)
2.269(2)
2.441(2)
1.40(2)
2.269(2)
2.254(2)
2.366(2)
1.519(9)
Bond Angles (deg)
P(1)−Ru(1)−P(2)
P(1)−Ru(1)−P(3)
P(2)−Ru(1)−P(3)
79.0(2)
80.8(2)
69.4(2)
82.48(9)
91.12(6)
96.01(6)
82.48(6)
89.19(6)
97.03(6)
82.50(9)
91.46(9)
95.87(9)
80.94(6)
97.30(6)
87.22(6)
83.83(6)
99.53(6)
83.77(6)
a
Data from ref 64.
4748
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760

Journal of the American Chemical Society
Article
2.461(4) Å for the phosphido ligand in 2 to 2.324(2) Å for the
phosphine ligand in 3]. Another feature of note is the presence
of the hydrogen dichloride anion, where the additional
equivalent of HCl is not removed when the crystals are
exposed to reduced pressure. The proton sits between the two
chlorine atoms with Cl(1)−HCl(2) and Cl(2)−HCl(1) bond
lengths of 1.55(9) and 1.58(9) Å, respectively, which are
comparable to the distances for other hydrogen dichloride
anions reported in the literature.72−74
Alternative Synthesis of 2 and Synthesis of a Bis(p-
tolyl)phosphino Analogue, Complex 5. Scheme 4 shows a
Scheme 4. Synthesis of Complexes 2 and 5 via Intermediate
Complex 4
Figure 3. ORTEP3 representation and atom numbering for 5 (thermal
ellipsoids at 50% probability; solvent and most of the hydrogens
omitted for clarity).
groups. Moreover, resonances for three inequivalent phospho-
rus nuclei were evident in the 31P{1H} NMR spectrum: two
doublets of doublets at 79.3 and 78.0 ppm as well as a triplet at
23.0 ppm.
second route to 2 involving a synthesis of phosphido complexes
that is more general and versatile than that in Scheme 2.
Starting with the commercially available cis-dppen ligand and a
suitable ruthenium(II) precursor, RuCp*(cod)Cl, the neutral
chloride complex 4 was generated in good yield (84%; Scheme
4). The NMR spectra of 4 in CD2Cl2 unambiguously showed
that coordination of the dppen ligand had occurred (a singlet at
81 ppm in the 31P{1H} NMR spectrum, a doublet at 7.5 ppm in
NMR studies of the formation of 5 were also conducted in
order to determine whether hydrophosphination of the dppen
ligand occurred upon the addition of bis(p-tolyl)phosphine to 4
after treatment with AgOTf (i.e., via intermediate A in Scheme
5). A 31P NMR spectrum of the reaction mixture showed a
strong P−H coupling for the inequivalent phosphorus donor in
the crude reaction mixture, indicating that intermediate A
containing dppen and P(p-tol)2H ligands (Scheme 5) was
present in solution and that swapping of the phosphorus
functionalities took place only after the addition of base. We
believe that the driving force for this exchange can be attributed
to electronic factors relating to the gauche effect. A bis(p-
tolyl)phosphido is more electron-rich and thus would
experience greater repulsion from a filled d orbital on the
ruthenium center. By exchange with a less electron-rich
diphenylphosphido, the electronic repulsion could be mini-
mized, affording a more stable species. This difference in
electron density is reflected by the pKa of bis(p-tolyl)phosphine
in THF, which was determined by NMR experiments to be
approximately 43 on the basis of the K value for the reaction
shown in eq 1:
1
the H NMR spectrum representing vinylic protons, and a
doublet of doublets at 148 ppm in the 13C NMR spectrum for
the olefinic carbons). Removal of the chloride in 4 with AgOTf
in THF followed by coordination of diphenylphosphine and
then deprotonation with KOtBu afforded 2 in moderate yield
(53%; Scheme 4). It should be noted that hydrophosphination
of the dppen ligand did not occur upon addition of
diphenylphosphine to 4; instead, 3 was detected by 31P NMR
spectroscopy.
The synthesis of a bis(p-tolyl)phosphido analogue of 2 was
also targeted by this route. The chloride of 4 was abstracted
with AgOTf, and then bis(p-tolyl)phosphine was added to the
reaction mixture. Upon reaction with KOtBu in THF, the
yellow solution turned dark red, and complex 5 was isolated as
a red powder in moderate yield (72%; Scheme 4).
XRD studies (Figure 3) revealed that 5 contains a 1-bis(p-
tolyl)phosphino-2-diphenylphosphinoethene ligand as well as a
diphenylphosphido group as a result of an unexpected ligand
rearrangement: the bis(p-tolyl)phosphino moiety was incorpo-
rated into the bidentate ligand at the expense of a
diphenylphosphido functionality. This reaction is reminiscent
of a transformation seen by the Stradiotto group in which a
phosphido exchanged with the amine functionality of a
bidentate ligand with an indenyl backbone.75 The bond lengths
and angles of complex 5 were nearly identical to those seen for
2 (Table 1).
HP(p‐tolyl)2 + [K(crypt)]NPh2
K
⇄ [K(crypt)][P(p‐tolyl)2] + HNPh2
(1)
where “crypt” denotes 2.2.2-cryptand. This is 5 orders of
magnitude more basic than diphenylphosphine in THF (pKa ≈
38, as reported previously).76 Furthermore, the reaction may
also be driven by statistical factors in that the unsymmetrical
complex has a higher entropy.
This unanticipated reactivity could have implications for
hydrophosphination reactions, where a common goal is to
more efficiently produce chiral, multidentate phosphorus
ligands from alkene substrates with phosphorus substituents.
In most studies, diphenylphosphine is used as a standard
substrate, but there are many cases in which more electron-rich
phosphines would be preferable, and ligands with different
The spectroscopic data for 5 were completely consistent with
the results from the X-ray diffraction study. Two distinct methyl
1
peaks were observed at 2.4 and 2.3 ppm in the H NMR
spectrum for the inequivalent bis(p-tolyl)phosphino methyl
4749
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760

Journal of the American Chemical Society
Article
Scheme 5. The Two Possible Intermediates on the Way to Complex 5
phosphorus donors would also be desirable.25,26,41,44,61 In these
cases, unexpected products could be produced if the substituted
olefin and secondary phosphine starting materials are not
chosen carefully. It may be important to manage how electron-
rich the phosphido ligand is in relation to the other phosphorus
donors on the substrate in order to prevent unwanted swapping
of phosphorus moieties: the more electron-deficient phosphido
ligand should be targeted with the more electron-rich
phosphorus functionalities on the olefin.
and 10.7 ppm, respectively. In the case of 8, the 31P{1H} NMR
spectrum indicated that the dppbz ligand remained intact while
the bis(p-tolyl)phosphido endured as a monodentate ligand.
Much like 2, the XRD structure of the o-phenylene analogue
7 (Figure 4) shows a distorted-piano-stool geometry with the
Synthesis of 1,2-Bis(diphenylphosphino)benzene An-
alogues, Complexes 6−8. The pKa values for diphenylphos-
phine and bis(p-tolyl)phosphine in ruthenium(II) complexes
were also of interest in order to better understand the facile P−
C bond rearrangements seen in the formation of 5 (and 2 as
described below) as well as for hydrophosphination reactions in
general. Only qualitative studies of this type have been reported
to date.18 To this end, analogues of 2 employing the 1,2-
bis(diphenylphosphino)benzene (dppbz) ligand with an
aromatic backbone were targeted to prevent the exchange of
phosphido ligands with the bidentate donor. By means of an
approach similar to that in Scheme 4, dppbz and RuCp*(cod)
Cl were combined to generate the neutral chloride species 6 in
good yield (77%; Scheme 6). The 31P{1H} NMR spectrum was
Figure 4. ORTEP3 representation and atom numbering for 7 (thermal
ellipsoids at 50% probability; most of the hydrogens omitted for
clarity).
Scheme 6. Synthesis of o-Phenylene Analogues 7 and 8 via
Intermediate Complex 6
dppbz and diphenylphosphido ligands opposite the Cp* ligand
(Figure 4). In addition, the P−Ru−P bond angles deviate from
the optimal octahedral angle of 90° [80.94(6), 97.30(6), and
87.22(6)° for P(1)−Ru(1)−P(2), P(1)−Ru(1)−P(3), and
P(2)−Ru(1)−P(3), respectively]. Compared with 2, however,
the Ru−P distances for the bidentate ligand are slightly longer
for 7 [Ru(1)−P(1) and Ru(1)−P(2) bond lengths of 2.254(2)
and 2.269(2) Å, respectively] while the Ru(1)−P(3) bond
length for the phosphido is slightly shorter [2.441(2) Å] (Table
1). The elongation of the Ru(1)−P(3) bond in comparison
with the other Ru−P bonds is attributed again to the gauche
effect.
quite diagnostic, with a singlet at 72.1 ppm. With 6 readily
available, abstraction of the chloride with AgOTf and
coordination of diphenylphosphine followed by deprotonation
with KOtBu allowed the facile synthesis of 7 in 87% yield,
whereas the use of bis(p-tolyl)phosphine gave complex 8 in
76% yield (Scheme 6).
The NMR spectra of complexes 7 and 8 in THF-d8 were very
similar to those of 2, except that the olefinic carbons and vinylic
protons were absent; instead, they were replaced by additional
proton signals associated with the aromatic ring. The 31P{1H}
NMR spectra were especially diagnostic, with doublets at 76.2
and 76.0 ppm, respectively, representing the dppbz phosphorus
atoms and broad singlets for the diarylphosphido ligand at 13.1
pKa Studies. The pKa values for the two dppbz metal
complexes 7 and 8 were experimentally determined by 31P{1H}
NMR spectroscopy through the use of a phosphazenium acid of
known pKa for comparison with the pKa values for the
corresponding free secondary phosphines (Table 2). It was
found that the bis(p-tolyl)phosphido complex 8 is 3 orders of
magnitude more basic than the diphenylphosphido complex 7.
4750
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760

Journal of the American Chemical Society
Article
Table 2. Approximate pKa Values in THF
a
b
acid/base
rel. pKa
0.5
abs. pKa
ref
c
P(p-tol)2H/[K(crypt)][P(p-tol)2]
2
43
41
38
28
26
25
4
this work
ref 76
NPh2H/[K(crypt)][NPh2]
0
4
4
3
3
3
c
PPh2H/[K(crypt)][PPh2]
−3 0.5
ref 76
d
[Ru(Cp*)(dppbz)(P(p-tol)2H)]BF4/Ru(Cp*)(dppbz)(P(p-tol)2) (8)
[P(NMe2)3NP(NMe2)2NHtBu]BF4/P(NMe2)3NP(NMe2)2NtBu
2
0.5
this work
ref 76
0
d
[Ru(Cp*)(dppbz)(PPh2H)]BF4/Ru(Cp*)(dppbz)(PPh2) (7)
−1 0.5
this work
a
The errors in the relative pKa values are significantly smaller than the errors in the absolute pKa values because the equilibrium constants were
determined under the same conditions with the same acid, and therefore, the errors arise only from the NMR instrument and weighing of the
b
reagents. The errors in the absolute pKa values are significantly larger than the errors in the relative pKa values because the errors are accumulated
c
from the entire pKTa HF ladder. See ref 76 for the calculation of the errors in the measured values. The relative pKa values are referenced to PPh2H/
d
[K(crypt)][PPh2]. The relative pKa values are referenced to [P(NMe2)3NP(NMe2)2NHtBu]BF4/P(NMe2)3NP(NMe2)2NtBu.
This result supports the hypothesis that the formation of 5 is
driven by the difference between the electron densities of the
two phosphido ligands and that incorporation of the more
electron-rich phosphorus donor into the bidentate ligand is
favored because it minimizes electronic repulsion due to the
gauche effect.
Scheme 7. Hydrophosphination of Acrylonitrile Catalyzed
by Phosphido Complexes 2 and 7
A comparison of the pKa values in Table 2 reveals that the
acidity of the secondary phosphines increases by 13−15 orders
of magnitude when they are coordinated to ruthenium in the
cationic half-sandwich compounds. These are significant
changes considering that the resulting metal complexes are
extremely electron-rich with Cp* and three strong σ donors as
ligands. For less electron-rich systems, the size of this shift
should be even larger. There are some other literature reports
of pKa values of metal−element−hydrogen groups where the
hydrogen is two bonds from the metal center.77,78 The most
common are for aqua complexes in water, where the pKa can
drop by up to 15 units when water coordinates to cations with
high charge-to-size ratios.79−81 Similarly, amines, which are very
weak acids in water, can have pKa values near 0 when bound to
trivalent cations,82 whereas ammonia can have a very wide
range of pKa values when bound to metals in non-aqueous
solutions, depending on the structure of the complex.83
temperature with a catalyst loading of 1% vs a TOF of 10 h−1 at
50 °C with a catalyst loading of 10% for a platinum-based
system59] but deactivated more rapidly as well. In addition, a
byproduct similar to one seen at −20.1 ppm in the 31P{1H}
NMR spectrum for the platinum system was seen with our
catalysts as well. We have identified this species as the
Markovnikov branched product, which was generated as a
minor product (the ratio of anti-Markovnikov to Markovnikov
product was approximately 4.5:1). The control experiment with
no ruthenium complex and only KOtBu was also carried out,
and it was found that base itself under exactly the same reaction
conditions was not an effective catalyst (less than 10%
conversion after 24 h).
While we have not done mechanistic studies to probe the
catalytic cycle, a Michael addition mechanism (Scheme 8)
Moreover, coordination of an aminoolefin to cationic rhodium-
Scheme 8. Proposed Catalytic Cycle for
Hydrophosphination Catalyzed by 7
84
(I) resulted in a change of 10 units in pKaDMSO
,
while the
pKDa MSO of a secondary amine in a cationic iridium(I) complex
changed from 22 to 15 upon substitution of only one ligand
(PMe3 to CO).85 Astruc and co-workers reported that the
pKDa MSO values for the 18-electron complexes [Fe(η5-C5R5)(η6-
C6Me6)]+ (R = H, Me) and [Fe(η5-C5R5)(η6-C6H5CHPh2)]+
were found to be 12−14 units lower than the pKa of the
benzylic C−H of the free ligand,86 while even the pKaMeCN of
1,3-diketones was found to decrease by 12 units upon
coordination to copper(II).87 Additionally, hydrogen sulfide
coordinated in a neutral ruthenium(II) complex was found to
be significantly more acidic than free H2S, whereas thiols, on
the other hand, were found to be less affected by metal
coordination.88,89
Catalytic Hydrophosphination. The ruthenium phosphi-
do complexes 2 and 7 were tested for hydrophosphination and
were found to catalyze the addition of diphenylphosphine to
acrylonitrile at room temperature (Scheme 7). The former
catalyst was twice as active as the latter, but the reactions did
not proceed to completion because the active species
deactivated rapidly after only 1 h. In comparison with other
acrylonitrile hydrophosphination catalysts reported in the
literature, complexes 2 and 7 operated more rapidly [turnover
frequencies (TOFs) of 30 and 60 h−1, respectively, at room
seems very reasonable on the basis of the pKa values that we
have determined (Table 2). Scheme 8 shows the very basic
phosphido complex 7 undergoing a Michael addition reaction
with acrylonitrile to generate a nitrile-stabilized carbanion. This
carbanion must be more basic than deprotonated acetonitrile
(pKaDMSO = 31,90 pKaTHF = 3576) and thus is likely to
deprotonate free PPh2H (pKTa HF = 38; Table 2). The very
4751
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760

Journal of the American Chemical Society
Article
Figure 5. Free energy profile for the two possible deprotonation and fragmentation pathways of the 1,1,2-tris(diphenylphosphino)ethane complex
[M06/6-31++G(d,p)/SDD-Ru/IEF-PCM + SMD(THF)]. All energies are relative to C + tBuO−.
nucleophilic diphenylphosphide formed in this way could then
displace the phosphinonitrile and coordinate to ruthenium, thus
regenerating the catalyst and releasing the hydrophosphinated
product.
give the observed product (D → E) and (ii) deprotonation of
the carbon β to the lone phosphorus donor (C → F) followed
by elimination of a phosphido (F → G). Even though we took
measures to minimize the errors in the energies of the starting
cation and the resulting neutral species by using a solvation
model, it should be kept in mind that comparing the energy of
cation C with those of the neutral species D to G will have a
greater uncertainty than comparing the energies of the neutral
species.
Calculations showed that both deprotonation events were
equally viable, giving ground-state structures D and F with
nearly identical free energies of −8.0 and −8.1 kcal/mol,
respectively. The two paths then diverged in their relative free
energies, with the formation of E clearly shown as both the
kinetic product [ΔGs°o⧧lv(TSD,E) = 0.7 kcal/mol] and the
thermodynamic product (G°solv(E) = −41.6 kcal/mol). The
other pathway was also kinetically feasible [ΔGs°o⧧lv(TSF,G) = 4.0
kcal/mol], but the product, G, was significantly higher in free
energy than E [G°solv(G) = −23.9 kcal/mol]. Although both
activation free energies were quite modest in the forward
direction, for the back reactions they were significantly larger
and differed greatly: 34.3 and 19.8 kcal/mol for E → D and G
→ F, respectively. It is evident that the interconversion between
E and D has a large kinetic barrier, whereas the interconversion
between F and G is more reasonable at ambient temperature.
This likely gives the system a mechanism to equilibrate toward
the thermodynamic sink and the observed product, E. Selected
ground-state and transition-state geometries are shown in
Figure 6. The elongated ruthenium−phosphido bonds in the
structures of E and G should be noted.
The basicity of coordinated secondary phosphines impacts
the way hydrophosphination reactions should be approached.
On the basis of these NMR studies, toluene and THF should
be ideal solvents, and indeed these solvents are the most
popular for hydrophosphination reactions, while more acidic
solvents such as MeOH and DCM should be
avoided.22,26,37,38,44,51,55−62 With respect to substrate choice
for alkene hydrophosphination, activated olefins such as
acrylonitrile and α,β-unsaturated enones are typically
used.22,26,37,38,44,51,55−57,59−62 For acrylonitrile, there are no
acidic protons, but α,β-unsaturated enones may have acidic,
enolizable protons; these should be avoided to prevent
unwanted side reactions and catalyst poisoning. In fact, mainly
aryl-substituted α,β-unsaturated enones are tested for late-
transition-metal-catalyzed alkene hydrophosphination reac-
tions, and this may be due to factors other than the electron-
withdrawing nature of these substituents.25,51,55−62 Lastly, when
developing ligands for hydrophosphination reactions, the
acidity of the ligand backbone should be considered carefully
in order to preclude catalyst degradation.
DFT Calculations on the Formation of 2. Detailed
computational studies utilizing density functional theory
(DFT) with the model complex [Ru(η5-C5H5)-
((PH2)2CHCH2PH2)]+ (C) (in which all of the phenyl and
methyl groups were removed to reduce computational costs)
with THF solvation were conducted in order to probe the
energetics of the fragmentation of tppe upon deprotonation of
complex 1 (Figure 5). The theoretical studies were based on
the initial presence of C as well as the strong base tBuO−, and
therefore, all of the energies are relative to these species. From
this starting point, two distinct reaction pathways were
considered: (i) deprotonation at the carbon α to the lone
phosphorus donor (C → D) and elimination of a phosphido to
Although the initial DFT calculations were useful for a
qualitative examination of the preferential formation of E over
G, they were inadequate when considering the dynamic
behavior of the phosphorus moieties exemplified by the
formation of 5. The barrier between E and D (34.3 kcal/
mol) was unreasonably high and did not agree with the
experimental results, which showed that the phosphorus
4752
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760

Journal of the American Chemical Society
Article
Figure 6. Optimized structures [M06/6-31++G(d,p)/SDD-Ru/IEF-PCM + SMD(THF)] as well as selected bond lengths (Å) and angles (deg) of
D, TSD,E (363i cm−1), E, F, TSF,G (345i cm−1), and G.
functionalities could exchange at room temperature. It is
important to note, however, that the calculations were
performed on a simplified model system where the steric
demands of the system were not modeled. Because the steric
parameters were neglected, fragmentation of the ligand released
an exaggerated amount of energy because it accounted only for
the release of the ring strain imposed by the tripodal ligand
architecture. Upon fragmentation of the ligand, however, the
phosphorus donors move apart, which would bring the phenyl
substituents closer to the methyl groups on the Cp* ligand.
This would increase the steric strain of the system and help
counterbalance the energy gained upon ring opening of the
four-membered metallocycle.
Computational studies with the full structures of D and E
(Dfull and Efull, respectively) were consequently conducted in
order to compare the relative energies of the two species. A
smaller basis set was used to reduce the computational costs,
and Dfull was arbitrarily selected as the reference free energy
(0.0 kcal/mol). The free energy difference between Efull (−17.2
kcal/mol) and Dfull was found to be 17.2 kcal/mol, which was
significantly smaller than that between E and D (33.6 kcal/mol;
Figure 5). This drastic decrease in the energy gap between the
two species indicates that a more reasonable barrier for
phosphido exchange exists and that sterics play a major role in
destabilizing the product Efull in order to allow this
phenomenon to take place.
This observed exchange of phosphorus moieties and outer-
sphere attack of a phosphido ligand on a noncoordinated C−C
double bond provides theoretical and experimental evidence for
proposed steps in catalytic hydrophosphination mechanisms.
These include outer-sphere attack of a phosphido ligand on a
pendant alkene coordinated to the metal through a heteroatom
substituent or, alternatively, a Michael addition to a free alkene.
Although Michael addition mechanisms are popular in the
hydrophosphination literature, in general there is very little
experimental evidence to support these proposed catalytic
cycles.22−26,37,38,41,42,44,47−54,56−63 The most convincing sup-
port was provided by Glueck and co-workers, who invoked an
outer-sphere Michael addition mechanism in platinum-
catalyzed hydrophosphination reactions.23,24 Our findings
suggest that outer-sphere attack of a coordinated phosphido
group on an activated alkene is favorable and that coordination
of an alkene heteroatom substituent may be more important
than initially anticipated.46,48
Decomposition/Rearrangement of Complex 2. Early
attempts to crystallize complex 2 were conducted under N2 at
25 °C, and after approximately 9 days, large orange crystals
suitable for single-crystal XRD studies formed in excellent yield
4753
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760

Journal of the American Chemical Society
Article
(94%). However, the product obtained was not compound 2
but a constitutional isomer, 9, that displayed several interesting
features (Scheme 9).
doublets of doublets and a broad singlet for the three
inequivalent phosphorus nuclei. The two phosphine donors
exhibited peaks at 99.8 and 86.5 ppm, while the phosphido
1
peak was shifted upfield to around 66.4 ppm. The H NMR
spectrum was also very revealing. Four inequivalent methyl
peaks were evident at 1.54, 1.46, 1.41, and 1.35 ppm, indicating
that functionalization of the Cp* ligand had occurred and that
all symmetry in the molecule was lost. Moreover, a methylene
peak at around 3.1 ppm for benzylic protons also confirmed
that activation of the Cp* had taken place, whereas two sets of
overlapping methylene peaks at around 2.2 ppm showed that
the unsaturated backbone of the dppen ligand had been
reduced.
Scheme 9. Decomposition/Rearrangement of Complex 2 To
Give Complex 9 in 94% Yield
It is interesting to note that all atoms are conserved upon
conversion of 2 to 9 and that this process proceeds even in the
dark, demonstrating that it is not a photochemical reaction
(Scheme 9). In order to generate the rearranged species, several
key transformations need to occur. First, activation of a Cp*
methyl group and migration of a phenyl functionality from a
phosphorus center is necessary to produce the 1-benzyl-2,3,4,5-
tetramethylcyclopentadienyl ligand. It is known in the literature
that tetramethylfulvene is susceptible to nucleophilic attack, and
Maitlis and co-workers have shown that even a chloride ligand
can migrate to tetramethylfulvene to produce a functionalized
Cp* on ruthenium.91−96 Furthermore, reports indicate that the
electronic structure of “tucked-in” complexes can be viewed as
either tetramethylfulvene and Mn+ or a tetramethylfulvenide
ligand and M(n+2)+, but regardless, it is known that this ligand
can be converted to a functionalized Cp* ligand, whether
through reductive elimination or intramolecular nucleophilic
attack.97−110 To generate the tridentate (2-((2-
(diphenylphosphino)ethyl)(phenyl)phosphino)phenyl)-
(phenyl)phosphido ligand, three key steps need to take place:
P−C bond cleavage and elimination of a phenyl group, as
mentioned previously; activation of an aromatic o-C−H bond
and coupling to a phosphorus center; and hydrogenation of the
unsaturated dppen backbone. The Rosenberg and Peters
groups have both shown that phosphidos are able to
deprotonate and orthometalate a phenyl substituent of a cis-
phosphine ligand on coordinatively unsaturated ruthenium(II)
complexes.17,28,36 This, followed by deprotonation of the
resulting phosphine and reductive elimination of the alkyl
and phosphido functionalities, could form the 1,2-diphosphi-
nobenzene linkage. In order to reduce the unsaturated
backbone from the dppen ligand, on the other hand, the
hydrogen atoms would need to come from the activated Cp*
methyl group and the aromatic o-C−H bond. Attack by a
hydride on the alkene and subsequent protonation of the
corresponding alkyl species by a phosphine could yield the
ethylidene fragment (see Table 3 for possible elements of the
rearrangement mechanism).
The ruthenium metal center exhibits a distorted-piano-stool
geometry with a functionalized Cp* ligand opposing a fac
tridentate diphosphinephosphido ligand (Figure 7). The Ru−P
Figure 7. ORTEP3 representation and atom numbering for 9 (thermal
ellipsoids at 50% probability; most of the hydrogens omitted for
clarity).
bond lengths for the phosphine donors, Ru(1)−P(1) and
Ru(1)−P(2), are similar to those seen for complex 2 [2.269(2)
and 2.254(2) Å, respectively], while the phosphido donor
displays an elongated Ru(1)−P(3) bond length of 2.366(2) Å
(attributed to the gauche effect). It is also evident from the
crystal structure that the new tridentate ligand constrains the
geometry about the metal center and contracts two of the P−
Ru−P bond angles, P(1)−Ru(1)−P(2) and P(2)−Ru(1)−
P(3), below the optimal octahedral angle of 90° [83.83(6) and
83.77(6)°, respectively]. The P(1)−Ru(1)−P(3) angle, on the
other hand, is obtuse (99.53°) and is likely pulled open by the
two five-membered rings linking the three phosphorus donors
(see Table 1 for other notable bond lengths and angles).
The rearrangement product 9 was characterized not only by
XRD but also by NMR spectroscopy. The 31P{1H} NMR
spectrum showed a very diagnostic pattern of two sets of
Attempts were made to probe the formation of complex 9
1
utilizing both 31P{1H} and H NMR spectroscopy, with an
emphasis on identifying hydride peaks in the 1H NMR
spectrum. A small amount of complex 2 was dissolved in
toluene-d8, placed in a J-Young tube, and layered with pentane
(the addition of pentane was important for isolating the
crystalline product, which decomposed over time when
dissolved in solution). Then, over the course of 13 days, the
pentane was allowed to slowly diffuse into the toluene solution
and NMR spectra were taken periodically (see Figure S1 in the
Supporting Information). After 2 days, three peaks appeared in
the 31P{1H} NMR spectrum: two doublets of doublets at 99.8
4754
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760

Journal of the American Chemical Society
Article
Table 3. Possible Steps of the Rearrangement Mechanism of the Reaction in Scheme 9 and Similar Reactions Found in the
a
Literature
a
It should be noted that this table simply displays a series of reactions that could be involved in the formation of 9, and the order in which they are
presented is not meant to be taken as a definitive sequence of events that must be followed.
and 86.5 ppm as well as a broad singlet at around 66.4 ppm.
These peaks corresponded to the product 9. Over the next 5
days, the concentration of 9 in solution increased until crystals
began to form in the NMR tube on day 8. The signals for 9
then slowly disappeared over 6 days, while more of the
rearrangement product crystallized out of solution. Throughout
the entire study, no hydrides could be detected in the 1H NMR
spectra.
or functionalizing any of their substituents would provide no
reliable mechanistic information on the formation of 9.
Therefore, mechanistic studies on the rearrangement reaction
have reached an impasse.
Currently it is premature to propose a mechanism for the
formation of the rearrangement product because of a lack of
evidence for any intermediates and because of the many steps
that need to take place in going from complex 2 to complex 9.
What can be proposed with a degree of certainty is that the first
step in the rearrangement process is rate-limiting and is
followed by a series of fast low-barrier transformations. This
may also explain why the yield is so high and the reaction is so
selective: after the initial slow step, the transient intermediates
form the product so quickly that they are not intercepted by
any other reactants.
Although the rearrangement reaction that produces 9 may
not be broadly applicable to other systems, it does illustrate the
high reactivity of the phosphido ligands and the unpredictability
of their transformations. The example of 9 is likely an extreme
case of how a phosphido ligand can facilitate unexpected ligand
rearrangements, but its formation demonstrates how carefully
ancillary ligands must be chosen for metal centers bearing
After failing to glean any significant mechanistic insight from
1
the 31P{1H} and H NMR study of the formation of 9, the
synthesis of 5 was pursued. It was thought that incorporation of
a bis(p-tolyl)phosphido ligand would provide a way to
determine which aromatic group is transferred to the Cp*
moiety and which substituent on phosphorus is activated at the
ortho position. Unfortunately, the migratory nature of the
phosphorus donors in the dppen phosphido species revealed
that complex 5 could not be used for its intended purpose. If a
phenyl or p-tolyl group were transferred to the Cp* ligand
during the rearrangement process, there would be no way to
confidently conclude whether it originated from the bidentate
ligand or the phosphido moiety. Essentially, because the
different phosphorus functionalities can exchange, deuterating
4755
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760

Journal of the American Chemical Society
Article
Sciex QStar mass spectrometer with an ESI source. The electron
impact mass spectrometry (EI-MS) data were collected on a Waters
GC ToF mass spectrometer with an EI/CI source, and the direct
analysis in real time mass spectrometry (DART-MS) data were
collected on a JEOL AccuTOF-DART mass spectrometer with a
DART ion source (no solvent is required). NMR spectra were
recorded at ambient temperature and pressure using a Varian Gemini
400 MHz spectrometer (400 MHz for 1H, 100 MHz for 13C, 376 MHz
for 19F, and 161 MHz for 31P) or an Agilent DD2 600 MHz
phosphido moieties. It is possible that the best ligands for
catalytic hydrophosphination are those that not only provide
the correct stereoelectronic environment to facilitate the
reaction but also are robust enough to withstand the phosphido
intermediates. It is also important to note that many ligands
that are traditionally thought to be stable, such as Cp* and
triphenylphosphine, may not be robust enough under these
conditions.
1
spectrometer (600 MHz for H, 151 MHz for 13C, 564 MHz for 19F,
and 243 MHz for 31P), unless stated otherwise. The 1H and 13C NMR
chemical shifts were measured relative to partially deuterated solvent
peaks but are reported relative to tetramethylsilane (TMS). All of the
31P chemical shifts were measured relative to 85% phosphoric acid as
an external reference. The elemental analyses were performed at the
Department of Chemistry, University of Toronto, on a PerkinElmer
2400 CHN elemental analyzer. Single-crystal XRD data were collected
using a Nonius Kappa-CCD or Bruker Kappa APEX DUO
diffractometer with Mo Kα radiation (λ = 0.710 Å) or Cu Kα
radiation (λ = 2.29 Å). The CCD data were integrated and scaled
using the Denzo-SMN package. The structures were solved and
refined using SHELXTL version 6.1. Refinement was by full-matrix
least-squares on F2 using all data. For compounds where extra solvent
was evident in the elemental analysis, there was evidence for the
additional solvent either in the XRD structure (solvent was seen or
was removed using the SQUEEZE function) or in the NMR spectra of
the purified compounds.
CONCLUSIONS
■
We have described the base-promoted dehydrophosphination
of the tppe ligand in 1 to generate a ruthenium dppen
phosphido complex 2 and explored the reactivity of the
deprotonated species in the context of alkene hydrophosphi-
nation reactions and mechanisms. It was found that the pKa of
the phosphido ligands was important in dictating the stability of
the metal complexes, to the extent that the more electron-rich
bis(p-tolyl)phosphido functionality exchanged with a diphenyl-
phosphino moiety of the dppen ligand to generate the more
stable species 5. This has implications for hydrophosphination
reactions, where the electron richness of the phosphido ligand
has to be managed carefully in order to prevent unwanted
swapping of phosphorus groups during the synthesis of chiral,
multidentate phosphorus ligands. Furthermore, the measured
pKa values of the phosphido complexes were used to rationalize
a possible catalytic mechanism for complexes 2 and 7, which
were capable of effecting the addition of diphenylphosphine to
acrylonitrile. Moreover, the pKa studies illustrated that the
acidity of the solvent, substrates, and ligand backbones have to
be carefully considered in these reactions. DFT studies of the
tppe fragmentation and the exchange of phosphorus moieties
were also enlightening in that they showed that an outer-sphere
attack of a phosphido ligand on a pendant alkene coordinated
to the metal through a heteroatom is energetically accessible at
room temperature. This gives support for Michael addition
mechanisms for the hydrophosphination of alkenes, which are
popular in the hydrophosphination literature. Lastly, the
unexpected reactivity of phosphido ligands was highlighted by
an unprecedented rearrangement reaction in which P−C, C−C,
and C−H bonds were made and broken to produce 9, a
constitutional isomer of 2. The rearrangement reaction stresses
the importance of designing robust ligands for hydro-
phosphination reactions and shows that ligands that are
traditionally considered “stable” may undergo unpredictable
decomposition pathways under relatively mild conditions.
Computational Details. DFT calculations were performed using
the Gaussian 09 package122 and the M06 hybrid functional.123−125
Ruthenium was treated with the SDD relativistic effective core
potential and associated basis set,126,127 while all other atoms were
treated with the 6-31++G(d,p) or 6-31+G(d) basis set.128−130
A
pruned (99,590) integration grid was used throughout (Grid=-
UltraFine). The substituents on phosphorus and Cp* were replaced
with hydrogen atoms to reduce the computational cost, except where
stated otherwise. Optimizations were performed using the integral
equation formalism polarizable continuum model (IEF-PCM)131,132
with radii and nonelectrostatic terms from the SMD solvation model
for THF.133 Ground states were connected to their transition states by
performing intrinsic reaction coordinate (IRC) calculations.134 Odd-
electron species were not considered, and stationary points were
characterized by normal-mode analysis. Full vibrational and
thermochemical analyses (1 atm, 298 K) were performed on optimized
structures to obtain solvent-corrected free energies (G°) and
enthalpies (H°). Optimized ground states were found to have zero
imaginary frequencies, while transition states were found to have one
imaginary frequency. Three-dimensional visualizations of calculated
structures and molecular orbitals were generated by ChemCraft.
Determination of pKa in THF. Equilibrium constants were
determined using gated−decoupled 31P NMR spectroscopy (in
nondeuterated THF). The experimentally determined T1 values
were all less than 10 s on a Varian Gemini 400 MHz spectrometer,
and thus, the recycling time (D1 + AT) was set to 2.363 s and a 30°
pulse was employed (the Ernst equation gives a maximum angle of
37.9° for this D1 + AT, indicating that the signals will be completely
relaxed and integrations will be quantitative in this field).135 To
determine the pKa of bis(p-tolyl)phosphine, 1 equiv of diphenylamine
was completely deprotonated with KH in the presence of
4,7,13,16,21,24-hexaoxa-1,10-diazabicyclo[8.8.8]hexacosane (2.2.2-
cryptand), and a solution of bis(p-tolyl)phosphine was added. Three
independent reactions with varying concentrations of bis(p-tolyl)-
phosphine were utilized to calculate the reported pKa value (see the
Supporting Information for sample calculations). For the metal
complexes, the experimentally determined T1 values were all less than
16 s on a Agilent DD2 600 MHz spectrometer, and thus, the recycling
time (D1 + AT) was set to 2.891 s and a 30° pulse was employed in
order to obtain quantitative integrations.135 To determine the pKa of 7,
the ruthenium complex was dissolved in THF, and a solution of 1-tert-
butyl-2,2,4,4,4-pentakis(dimethylamino)-2λ5,4λ5-catenadi-
(phosphazenium) tetrafluoroborate salt was added. Three independent
EXPERIMENTAL SECTION
■
General Considerations. Procedures and manipulations were
performed under an argon or nitrogen atmosphere using standard
Schlenk line and glovebox techniques, unless stated otherwise.
Solvents were degassed and dried using standard procedures prior to
all manipulations and reactions, unless stated otherwise. The synthesis
of complex 1 was described previously,64 as was the synthesis of
RuCp*(cod)Cl.121 The 1-tert-butyl-2,2,4,4,4-pentakis-
(dimethylamino)-2λ5,4λ5-catenadi(phosphazenium) tetrafluoroborate
salt was synthesized from commercially available 1-tert-butyl-
2,2,4,4,4-pentakis(dimethylamino)-2λ5,4λ5-catenadi(phosphazene)
and tetrafluoroboric acid diethyl ether complex (purchased from
Sigma-Aldrich and used without further purification). Deuterated
solvents were purchased from Cambridge Isotope Laboratories or
Sigma-Aldrich, degassed, and dried over activated molecular sieves
prior to use. All other reagents were purchased from commercial
sources and utilized without further purification. The electrospray
ionization mass spectrometry (ESI-MS) data were collected on an AB/
4756
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760

Journal of the American Chemical Society
Article
reactions with varying concentrations of ruthenium complex and
phosphazenium salt were utilized to calculate the reported pKa value.
The pKa of 8 was determined in a similar fashion using three
independent reactions with varying concentrations of ruthenium
complex and phosphazenium salt as well. The values are listed in Table
2. To reduce sources of error, the amount of solvent added was
determined by weight.
Ar−CH), 129.16 (s, Ar−CH), 127.70 (t, Ar−CH, J = 4.5 Hz), 127.51
(t, Ar−CH, J = 4.8 Hz), 89.64 (t, Cp*-C, J = 2.6 Hz), and 9.51 (s,
Cp*−CH3) ppm. Anal. Calcd for [C36H37ClP2Ru]·0.50CH2Cl2: C,
61.69; H, 5.39. Found: C, 61.88; H, 5.32. MS (ESI+, DCM; m/z+):
668.1 [C36H38ClP2Ru]+.
Alternative Synthesis of Complex 2. A small amount of 4
(0.031, 0.046 mmol) was dissolved in THF (2 mL), and AgOTf (0.011
g, 0.046 mmol) in THF (1 mL) was added. The solution turned
orange with a dark-gray precipitate. The mixture was stirred for 30 min
in the dark and then filtered. A solution of diphenylphosphine (0.009
g, 0.046 mmol) in THF (2 mL) was added to the filtrate, and the
solution turned from orange to yellow in color. The reaction mixture
was stirred for an additional 30 min, and KOtBu (0.005 g, 0.046
mmol) in THF (1 mL) was added. The solution immediately turned
dark red in color. The solvent was removed under reduced pressure,
and the red residue was redissolved in toluene (∼2 mL). The solution
was cooled to −33 °C in a freezer overnight and then filtered through
Celite. Hexanes (∼4 mL) was added, and the solution was once again
cooled to −33 °C in a freezer overnight. The red precipitate was
collected and washed with hexanes (∼2 mL) to give a brick-red
powder. Yield: 53% (0.020 g). All of the spectroscopic data were
identical to those of 2.
Synthesis of RuCp*(Ph2PCHCHP(p-tol)2)(PPh2) (5). This was
similar to the alternative synthesis of 2 (see p S2 in the Supporting
Information).
Synthesis of RuCp*(Ph2PC6H4PPh2)Cl (6). This was similar to
the synthesis of 4 (see pp S2−S3 in the Supporting Information).
Synthesis of RuCp*(Ph2PC6H4PPh2)(PPh2) (7). This was similar
to the alternative synthesis of 2 (see p S3 in the Supporting
Information).
Synthesis of RuCp*(Ph2PCHCHPPh2)(PPh2) (2). A small
amount of 1 (0.020 g, 0.021 mmol) was dissolved in THF (2 mL).
An excess of KOtBu (0.010 g, 0.089 mmol) was added. The resulting
deep-red solution was evaporated to dryness and dissolved in toluene
(2 mL). The solution was left in the freezer at −33 °C overnight and
then filtered. The solvent was removed under reduced pressure to give
a brick-red powder. Yield: 94% (0.016 g). Crystals suitable for XRD
studies were grown by slow diffusion of pentane into a toluene
1
solution of the complex at −33 °C. H NMR (400 MHz, THF-d8) δ:
7.66−7.57 (m, 4H, Ar−CH), 7.43−7.35 (m, 5H, Ar−CH and CH
CH), 7.24−7.17 (m, 3H, Ar−CH and CHCH), 7.13−7.06 (m, 6H,
Ar−CH), 6.89 (t, 4H, J = 8.4 Hz), 6.67 (t, 2H, Ar−CH, J = 7.2 Hz),
6.52 (t, 4H, Ar−CH, J = 7.5 Hz), 6.44−6.36 (m, 4H, Ar−CH), and
1.12 (s, 15H, Cp*−CH3) ppm. 31P{1H} NMR (161 MHz, THF-d8) δ:
79.22 (d, J = 8.1 Hz) and 23.21 (t, J = 8.1 Hz) ppm. 13C{1H} NMR
(100 MHz, THF-d8) δ: 148.18 (d, C−P, J = 44.6 Hz), 146.92 (dd,
CH−P, J = 38.4, 32.7 Hz), 139.31 (d, C−P, J = 38.4 Hz), 136.81 (d,
C−P, J = 44.6 Hz), 136.55 (d, Ar−CH, J = 19.5 Hz), 133.97 (q, Ar−
CH, J = 5.1 Hz), 133.21 (dd, Ar−CH, J = 5.2 Hz), 129.03 (s, Ar−
CH), 128.48 (s, Ar−CH), 127.60 (dd, Ar−CH, J = 4.6 Hz), 126.82
(dd, Ar−CH, J = 4.4 Hz), 125.19 (d, Ar−CH, J = 5.4 Hz), 123.47 (s,
Ar−CH), 93.05 (t, Cp*-C, J = 2.0 Hz), and 9.60 (d, Cp*−CH3, J = 5.3
Hz) ppm. Anal. Calcd for [C48H47P3Ru]·0.50C7H8: C, 71.60; H, 5.95.
Found: C, 71.22; H, 6.34. MS (DART, no solvent; m/z+): 819.0
[C48H48P3Ru]+.
Synthesis of RuCp*(Ph2PC6H4PPh2)(P(p-tol)2) (8). This was
similar to the alternative synthesis of 2 (see pp S3−S4 in the
Supporting Information).
Synthesis of [RuCp*(Ph2PCHCHPPh2)(HPPh2)][HCl2] (3). A
small amount of 2 (0.030 g, 0.037 mmol) was dissolved in toluene (2
mL), and an excess of 1 M HCl/ether (0.08 mL, 0.080 mmol) was
added. The solution turned from red to pale yellow, and a yellow
precipitate formed. The resulting yellow powder was isolated by
filtration and washed with small amounts of ether (1 mL) and pentane
(2 mL). Yield: 76.5% (0.025 g). Crystals suitable for XRD studies were
grown by slow diffusion of pentane into a DCM solution of the
Synthesis of Ru(C5(CH3)4(CH2C6H5))(Ph2PCH2CH2PPh(o-C6H4)-
PPh) (9). A small amount of 2 (0.010 g, 0.012 mmol) was dissolved in
toluene (∼1 mL), and a large amount of pentane (∼3 mL) was slowly
diffused into the solution over 9−10 days. The resulting orange
crystals were filtered and dried under reduced pressure. Yield: 93%
(0.09 g). Crystals suitable for XRD studies were grown by diffusion of
1
pentane into a toluene solution of 1 after 9−10 days. H NMR (600
1
complex. H NMR (400 MHz, CD2Cl2) δ: 7.73−7.65 (m, 6H, Ar−
MHz, THF-d8) δ: 7.69−7.63 (m, 2H, Ar−CH), 7.58−7.52 (m, 2H,
Ar−CH), 7.51−7.46 (m, 1H, Ar−CH), 7.42 (d, 1H, Ar−CH, J = 7.4
Hz), 7.28 (t, 1H, Ar−CH, J = 7.4 Hz), 7.26−7.23 (m, 2H, Ar−CH),
7.12−7.05 (m, 4H, Ar−CH), 7.03−6.94 (m, 9H, Ar−CH), 6.89−6.85
(m, 2H, Ar−CH), 6.82 (t, 2H, Ar−CH, J = 6.9 Hz), 6.61 (t, 1H, Ar−
CH, J = 7.3 Hz), 6.46 (t, 2H, Ar−CH, J = 7.6 Hz), 3.14−3.04 (m, 2H,
CH2), 2.31−2.16 (m, 4H, CH2), 1.54 (s, 3H, Cp*−CH3), 1.46 (s, 3H,
Cp*−CH3), 1.41 (s, 3H, Cp*−CH3), and 1.35 (s, 3H, Cp*−CH3)
ppm. 31P{1H} NMR (243 MHz, THF-d8) δ: 99.79 (dd, J = 23.3, 6.5
Hz), 86.51 (dd, J = 23.3, 7.1 Hz), and 66.74−66.38 (br s) ppm.
13C{1H} NMR (151 MHz, THF-d8) δ: 133.06 (d, Ar−CH, J = 21.1
Hz), 132.15 (d, Ar−CH, J = 10.4 Hz), 132.00 (d, Ar−CH, J = 8.5 Hz),
131.88−131.70 (m, Ar−CH), 129.27−129.18 (m, Ar−CH), 128.91
(d, Ar−CH, J = 1.9 Hz), 128.42 (d, Ar−CH, J = 2.2 Hz), 128.00 (d,
Ar−CH, J = 1.0 Hz), 127.96−127.85 (m, Ar−CH), 127.47 (d, Ar−
CH, J = 9.1 Hz), 127.02 (d, Ar−CH, J = 8.6 Hz), 126.25 (d, Ar−CH, J
= 6.4 Hz), 125.04 (s, Ar−CH), 123.92−123.81 (m, Ar−CH), 94.28−
93.92 (m, Cp*-C), 93.75−93.62 (m, Cp*-C), 31.02−31.0 (m, P−
CH2), 30.04−30.02 (m, CH2), 9.92 (s, Cp*−CH3), 9.84 (s, Cp*−
CH3), 9.79 (s, Cp*−CH3), and 9.69 (s, Cp*−CH3) ppm. Anal. Calcd
for [C48H47P3Ru]: C, 70.49; H, 5.79. Found: C, 70.97; H, 6.17. MS
(DART, no solvent; m/z+): 819.0 [C48H48P3Ru]+.
CH), 7.65−7.57 (m, 5H, Ar−CH and P−CH), 7.45 (d, 1H, P−CH, J
= 24.1 Hz), 7.42 (t, 2H, Ar−CH, J = 7.0 Hz), 7.31 (td, 4H, Ar−CH, J
= 7.6 Hz), 7.14 (t, 2H, Ar−CH, J = 7.0 Hz), 6.98−6.90 (m, 6H, Ar−
CH), 6.46 (dd, 4H, Ar−CH, J = 11.4, 7.2 Hz), 6.05 (dt, 1H, PH, J =
356.7, 5.6 Hz), and 1.44−1.41 (m, 15H, Cp*−CH3) ppm. 31P{1H}
NMR (161 MHz, CD2Cl2) δ: 74.61 (d, J = 35.8 Hz) and 37.92 (t, J =
35.8 Hz) ppm. 13C{1H} NMR (100 MHz, CD2Cl2) δ: 146.22 (dd, P−
CH, J = 35.5, 32.6 Hz), 134.35 (dd, C−P, J = 42.5, 2.9 Hz), 133.87
(dd, C−P, J = 48.8, 3.5 Hz), 132.58 (t, Ar−CH, J = 5.4 Hz), 132.48 (d,
Ar−CH, J = 10.3 Hz), 132.32 (d, Ar−CH, J = 9.0 Hz), 131.91 (d, C−
P, J = 42.4 Hz), 131.56 (s, Ar−CH), 130.64 (s, Ar−CH), 129.57 (t,
Ar−CH, J = 4.9 Hz), 129.56 (s, Ar−CH), 128.23 (t, Ar−CH, J = 4.9
Hz), 128.11 (d, Ar−CH, J = 9.5 Hz), and 9.72 (s, Cp*−CH3) ppm.
Anal. Calcd for [C48H47P3Ru][Cl2H]·0.25CH2Cl2: C, 63.54; H, 5.47.
Found: C, 63.54; H, 5.91. MS (ESI+, DCM; m/z+): 819.2
[C48H47P3Ru]+.
Synthesis of RuCp*(Ph2PCHCHPPh2)Cl (4). Both cis-1,2-bis-
(diphenylphosphino)ethene (0.104 g, 0.26 mmol) and RuCp*(cod)Cl
(0.100 g, 0.26 mmol) were dissolved in dichloromethane (DCM) (10
mL), and the solution was heated at 36 °C overnight. The solvent was
removed under reduced pressure, and the resulting orange solid was
1
washed with hexanes. Yield: 83% (0.147 g). H NMR (400 MHz,
CD2Cl2) δ: 7.73−7.62 (m, 4H, Ar−CH), 7.50 (d, 2H, P−CH, J = 58.5
Hz), 7.44−7.38 (m, 6H, Ar−CH), 7.38−7.26 (m, 6H, Ar−CH), 7.12
(t, 4H, Ar−CH, J = 8.4 Hz), and 1.44 (s, 15H, Cp*−CH3) ppm.
31P{1H} NMR (161 MHz, CD2Cl2) δ: 81.14 (s) ppm. 13C{1H} NMR
(100 MHz, CD2Cl2) δ: 147.52 (dd, P−CH, J = 34.6, 33.5 Hz), 137.86
(dd, C−P, J = 43.5, 4.5 Hz), 133.93 (t, Ar−CH, J = 5.2 Hz), 133.07
(dd, C−P, J = 45.4, 2.2 Hz), 132.53 (t, Ar−CH, J = 5.3 Hz), 129.28 (s,
Hydrophosphination of Acrylonitrile. A small amount of 2
(0.005 g, 0.006 mmol) or 7 (0.005 g, 0.006 mmol) was dissolved in 1
mL of THF. A solution of diphenylphosphine (0.105 g, 0.565 mmol)
in 1 mL of THF and a solution of acrylonitrile (0.031 g, 0.565 mmol)
in 1 mL of THF were simultaneously added. The reaction mixture was
stirred for 1 h, and conversion was determined by 31P{1H} NMR
analysis.
4757
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760

Journal of the American Chemical Society
Article
(25) Glueck, D. S. Top. Organomet. Chem. 2010, 31, 65.
(26) Kovacik, I.; Scriban, C.; Glueck, D. S. Organometallics 2006, 25,
536.
(27) Derrah, E. J.; Pantazis, D. A.; McDonald, R.; Rosenberg, L.
Angew. Chem., Int. Ed. 2010, 49, 3367.
(28) Derrah, E. J.; Pantazis, D. A.; McDonald, R.; Rosenberg, L.
Organometallics 2007, 26, 1473.
(29) Derrah, E. J.; McDonald, R.; Rosenberg, L. Chem. Commun.
ASSOCIATED CONTENT
* Supporting Information
■
S
Extended experimental section; complete ref 122; 31P NMR
spectroscopic study of the formation of 5; sample calculations
for the determination of pKa values; crystallographic data tables
for 2, 3, 5, 7, and 9 (including CIF files); and tables giving
Cartesian coordinates and free energies for optimized
structures. This material is available free of charge via the
Internet at http://pubs.acs.org.
2010, 46, 4592.
(30) Derrah, E. J.; Giesbrecht, K. E.; McDonald, R.; Rosenberg, L.
Organometallics 2008, 27, 5025.
(31) Vaughan, G. A.; Hillhouse, G. L.; Rheingold, A. L. Organo-
metallics 1989, 8, 1760.
(32) Iluc, V. M.; Hillhouse, G. L. J. Am. Chem. Soc. 2010, 132, 15148.
(33) Baker, R. T.; Calabrese, J. C.; Harlow, R. L.; Williams, I. D.
Organometallics 1993, 12, 830.
AUTHOR INFORMATION
Corresponding Author
rmorris@chem.utoronto.ca
■
Notes
(34) Nakazawa, H.; Yamaguchi, Y.; Mizuta, T.; Ichimura, S.; Miyoshi,
K. Organometallics 1995, 14, 4635.
The authors declare no competing financial interest.
(35) Menye-Biyogo, R.; Delpech, F.; Castel, A.; Pimienta, V.;
Gornitzka, H.; Riviere, P. Organometallics 2007, 26, 5091.
(36) Takaoka, A.; Mendiratta, A.; Peters, J. C. Organometallics 2009,
28, 3744.
ACKNOWLEDGMENTS
We thank NSERC Canada for a Discovery Grant to R.H.M and
Demyan Prokopchuk for fruitful discussions.
■
(37) Wicht, D. K.; Kourkine, I. V.; Lew, B. M.; Nthenge, J. M.;
Glueck, D. S. J. Am. Chem. Soc. 1997, 119, 5039.
(38) Du, D.; Lin, Z.-Q.; Lu, J.-Z.; Li, C.; Duan, W.-L. Asian J. Org.
Chem. 2013, 2, 392.
(39) Ananikov, V. P.; Beletskaya, I. P. Chem.Asian J. 2011, 6, 1423.
(40) Ananikov, V. P.; Makarov, A. V.; Beletskaya, I. P. Chem.Eur. J.
2011, 17, 12623.
REFERENCES
■
(1) Rosenberg, L. Coord. Chem. Rev. 2012, 256, 606.
(2) Miyake, Y.; Endo, S.; Yuki, M.; Tanabe, Y.; Nishibayashi, Y.
Organometallics 2008, 27, 6039.
(3) Cauzzi, D.; Graiff, C.; Massera, C.; Predieri, G.; Tiripicchio, A.
Eur. J. Inorg. Chem. 2001, 721.
(4) Alvarez, C. M.; Garcia, M. E.; Ruiz, M. A.; Connelly, N. G.
Organometallics 2004, 23, 4750.
(41) Zhang, Y.; Tang, L.; Pullarkat, S. A.; Liu, F.; Li, Y.; Leung, P.-H.
J. Organomet. Chem. 2009, 694, 3500.
(5) Wong, W.-Y.; Ting, F.-L.; Lin, Z. Organometallics 2003, 22, 5100.
(6) Baldovino-Pantaleon, O.; Rios-Moreno, G.; Toscano, R. A.;
Morales-Morales, D. J. Organomet. Chem. 2005, 690, 2880.
(7) Dick, D. G.; Hou, Z.; Stephan, D. W. Organometallics 1992, 11,
2378.
(42) Sugiura, J.; Kakizawa, T.; Hashimoto, H.; Tobita, H.; Ogino, H.
Organometallics 2005, 24, 1099.
(43) Yeo, W.-C.; Tang, L.; Yan, B.; Tee, S.-Y.; Koh, L. L.; Tan, G.-K.;
Leung, P.-H. Organometallics 2005, 24, 5581.
(44) Yeo, W.-C.; Tee, S.-Y.; Tan, H.-B.; Tan, G.-K.; Koh, L. L.;
Leung, P.-H. Inorg. Chem. 2004, 43, 8102.
(45) Han, L.-B.; Zhao, C.-Q.; Tanaka, M. J. Org. Chem. 2001, 66,
5929.
(46) Zhao, D.; Mao, L.; Yang, D.; Wang, R. J. Org. Chem. 2010, 75,
6756.
(47) Bravo-Altamirano, K.; Abrunhosa-Thomas, I.; Montchamp, J.-L.
(8) Cabeza, J. A.; Mulla, F.; Riera, V. J. Organomet. Chem. 1994, 470,
173.
(9) Suss-Fink, G.; Godefroy, I.; Beguin, A.; Rheinwald, G.; Neels, A.;
̈
Stoeckli-Evans, H. J. Chem. Soc., Dalton Trans. 1998, 2211.
(10) Jungbluth, H.; Stoeckli-Evans, H.; Suss-Fink, G. J. Organomet.
̈
Chem. 1990, 391, 109.
(11) Bottcher, H.-C.; Graf, M.; Merzweiler, K.; Bruhn, C. Polyhedron
1998, 17, 3433.
(12) Eichele, K.; Wasylishen, R. E.; Corrigan, J. F.; Taylor, N. J.;
Carty, A. J.; Feindel, K. W.; Bernard, G. M. J. Am. Chem. Soc. 2002,
124, 1541.
(13) Bader, A.; Nullmeyers, T.; Pabel, M.; Salem, G.; Willis, A. C.;
Wild, S. B. Inorg. Chem. 1995, 34, 384.
(14) Buhro, W. E.; Zwick, B. D.; Georgiou, S.; Hutchinson, J. P.;
Gladysz, J. A. J. Am. Chem. Soc. 1988, 110, 2427.
(15) Fryzuk, M. D.; Bhangu, K. J. Am. Chem. Soc. 1988, 110, 961.
(16) Gloaguen, Y.; Jacobs, W.; de Bruin, B.; Lutz, M.; van der Vlugt,
J. I. Inorg. Chem. 2013, 52, 1682.
(17) Hoyle, M.-A. M.; Pantazis, D. A.; Burton, H. M.; McDonald, R.;
Rosenberg, L. Organometallics 2011, 30, 6458.
(18) Planas, J. G.; Hampel, F.; Gladysz, J. A. Chem.Eur. J. 2005, 11,
1402.
(19) Boni, G.; Blacque, O.; Sauvageot, P.; Poujaud, N.; Moise, C.;
Kubicki, M. M. Polyhedron 2002, 21, 371.
(20) Fryzuk, M. D.; Joshi, K.; Chadha, R. K.; Rettig, S. J. J. Am. Chem.
Soc. 1991, 113, 8724.
J. Org. Chem. 2008, 73, 2292.
(48) Sadow, A. D.; Togni, A. J. Am. Chem. Soc. 2005, 127, 17012.
(49) Chen, K.; Pullarkat, S. A.; Ma, M.; Li, Y.; Leung, P.-H. Dalton
Trans. 2012, 41, 5391.
(50) Sabater, S.; Mata, J. A.; Peris, E. Organometallics 2013, 32, 1112.
(51) Huang, Y.; Pullarkat, S. A.; Li, Y.; Leung, P.-H. Inorg. Chem.
2012, 51, 2533.
(52) Yang, M.-J.; Liu, Y.-J.; Gong, J.-F.; Song, M.-P. Organometallics
2011, 30, 3793.
(53) Hoye, P. A. T.; Pringle, P. G.; Smith, M. B.; Worboys, K. J.
Chem. Soc., Dalton Trans. 1993, 269.
(54) Blank, N. F.; Moncarz, J. R.; Brunker, T. J.; Scriban, C.;
Anderson, B. J.; Amir, O.; Glueck, D. S.; Zakharov, L. N.; Golen, J. A.;
Incarvito, C. D.; Rheingold, A. L. J. Am. Chem. Soc. 2007, 129, 6847.
(55) Rosenberg, L. ACS Catal. 2013, 3, 2845.
(56) Pringle, P. G.; Smith, M. B. J. Chem. Soc., Chem. Commun. 1990,
1701.
(57) Costa, E.; Pringle, P. G.; Smith, B.; Worboys, M. J. Chem. Soc.,
Dalton Trans. 1997, 4277.
(58) Costa, E.; Pringle, P. G.; Worboys, K. Chem. Commun. 1998, 49.
(59) Wicht, D. K.; Kourkine, I. V.; Kovacik, I.; Glueck, D. S.;
Concolino, T. E.; Yap, G. P. A.; Incarvito, C. D.; Rheingold, A. L.
Organometallics 1999, 18, 5381.
(60) Du, D.; Duan, W.-L. Chem. Commun. 2011, 47, 11101.
(61) Huang, Y.; Chew, R. J.; Li, Y.; Pullarkat, S. A.; Leung, P.-H. Org.
Lett. 2011, 13, 5862.
(21) Planas, J. G.; Gladysz, J. A. Inorg. Chem. 2002, 41, 6947.
(22) Glueck, D. S. Synlett 2007, 2627.
(23) Scriban, C.; Glueck, D. S.; Zakharov, L. N.; Kassel, W. S.;
DiPasquale, A. G.; Golen, J. A.; Rheingold, A. L. Organometallics 2006,
25, 5757.
(24) Scriban, C.; Kovacik, I.; Glueck, D. S. Organometallics 2005, 24,
4871.
4758
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760

Journal of the American Chemical Society
Article
(62) Huang, Y.; Pullarkat, S. A.; Li, Y.; Leung, P.-H. Chem. Commun.
2010, 46, 6950.
(63) Chen, Y.-R.; Duan, W.-L. Org. Lett. 2011, 13, 5824.
(64) Sues, P. E.; Lough, A. J.; Morris, R. H. Organometallics 2012, 31,
6589.
(97) Pinkas, J.; Cisarova, I.; Gyepes, R.; Horacek, M.; Kubista, J.;
Cejka, J.; Gomez-Ruiz, S.; Hey-Hawkins, E.; Mach, K. Organometallics
2008, 27, 5532.
(98) Pinkas, J.; Cisarova, I.; Horacek, M.; Kubista, J.; Mach, K.
Organometallics 2011, 30, 1034.
(99) Varga, V.; Sindelar, P.; Cisarova, I.; Horacek, M.; Kubista, J.;
Mach, K. Inorg. Chem. Commun. 2005, 8, 222.
(100) Edelmann, F. T.; Giessmann, S.; Fischer, A. Chem. Commun.
2000, 2153.
(65) Ruiz, J.; Mosquera, M. E. G.; Riera, V.; Vivanco, M.; Bois, C.
Organometallics 1997, 16, 3388.
(66) Riera, V.; Ruiz, J. J. Organomet. Chem. 1986, 310, C36.
(67) Delavaux, B.; Chaudret, B.; Devillers, J.; Dahan, F.; Commenges,
G.; Poilblanc, R. J. Am. Chem. Soc. 1986, 108, 3703.
(68) Toth, I.; Hanson, B. E.; Davis, M. E. Organometallics 1990, 9,
675.
(101) Evans, W. J.; Walensky, J. R.; Ziller, J. W. Chem.Eur. J. 2009,
15, 12204.
(102) Evans, W. J.; Champagne, T. M.; Ziller, J. W. J. Am. Chem. Soc.
2006, 128, 14270.
(103) Legzdins, P.; Sayers, S. F. Organometallics 1996, 15, 3907.
(104) Legzdins, P.; Sayers, S. F. Chem.Eur. J. 1997, 3, 1579.
(105) Cloke, F. G. N.; Day, J. P.; Green, J. C.; Morley, C. P.; Swain,
A. C. J. Chem. Soc., Dalton Trans. 1991, 789.
(106) Cloke, F. G. N.; Hitchcock, P. B.; Nixon, J. F.; Wilson, D. J.
Organometallics 2000, 19, 219.
(107) Bulls, A. R.; Schaefer, W. P.; Serfas, M.; Bercaw, J. E.
Organometallics 1987, 6, 1219.
(108) Riley, P. N.; Parker, J. R.; Fanwick, P. E.; Rothwell, I. P.
Organometallics 1999, 18, 3579.
(109) Schmiege, B. M.; Fieser, M. E.; Ziller, J. W.; Evans, W. J.
Organometallics 2012, 31, 5591.
(110) Takase, M. K.; Siladke, N. A.; Ziller, J. W.; Evans, W. J.
Organometallics 2011, 30, 458.
(111) Li, C.; Luo, L.; Nolan, S. P.; Marshall, W.; Fagan, P. J.
Organometallics 1996, 15, 3456.
(112) Kolle, U.; Kang, B.-S.; Thewalt, U. J. Organomet. Chem. 1990,
(69) Amor, I.; García, M. E.; Ruiz, M. A.; Sae
Jeffery, J. C. Organometallics 2006, 25, 4857.
́
z, D.; Hamidov, H.;
(70) Bandini, A. L.; Banditelli, G.; Minghetti, G. J. Organomet. Chem.
2000, 595, 224.
(71) Ponomarenko, V. I.; Pilyugina, T. S.; Khripun, V. D.; Grachova,
E. V.; Tunik, S. P.; Haukka, M.; Pakkanen, T. A. J. Organomet. Chem.
2006, 691, 111.
(72) Boettcher, H.-C.; Mayer, P.; Schmidbaur, H. Z. Naturforsch., B: J.
Chem. Sci. 2012, 67, 543.
(73) Gellhaar, J.; Knapp, C. Acta Crystallogr., Sect. E 2011, 67, o2546.
(74) Atwood, J. L.; Bott, S. G.; Means, C. M.; Coleman, A. W.;
Zhang, H.; May, M. T. Inorg. Chem. 1990, 29, 467.
(75) Rankin, M. A.; Schatte, G.; McDonald, R.; Stradiotto, M.
Organometallics 2008, 27, 6286.
(76) Abdur-Rashid, K.; Fong, T. P.; Greaves, B.; Gusev, D. G.;
Hinman, J. G.; Landau, S. E.; Lough, A. J.; Morris, R. H. J. Am. Chem.
Soc. 2000, 122, 9155.
(77) Mazzeo, M.; Lamberti, M.; Massa, A.; Scettri, A.; Pellecchia, C.;
386, 267.
Peters, J. C. Organometallics 2008, 27, 5741.
(78) Bauer, R. C.; Gloaguen, Y.; Lutz, M.; Reek, J. N. H.; de Bruin,
B.; van der Vlugt, J. I. Dalton Trans. 2011, 40, 8822.
(113) Burn, M. J.; Fickes, M. G.; Hollander, F. J.; Bergman, R. G.
Organometallics 1995, 14, 137.
(114) Geldbach, T. J.; Drago, D.; Pregosin, P. S. Chem. Commun.
́
(79) Fernandez, R.; Melchart, M.; Habtemariam, A.; Parsons, S.;
2000, 1629.
(115) Geldbach, T. J.; Pregosin, P. S.; Bassetti, M. Organometallics
2001, 20, 2990.
(116) Tan, X.; Li, B.; Xu, S.; Song, H.; Wang, B. Organometallics
Sadler, P. J. Chem.Eur. J. 2004, 10, 5173.
(80) Lehaire, M.-L.; Grundler, P. V.; Steinhauser, S.; Marti, N.; Helm,
L.; Hegetschweiler, K.; Schibli, R.; Merbach, A. E. Inorg. Chem. 2006,
45, 4199.
2011, 30, 2308.
(117) Bennett, D. W.; Siddiquee, T. A.; Haworth, D. T.; Kabir, S. E.;
Hyder, M. I.; Ahmed, S. J. J. Chem. Crystallogr. 2004, 34, 361.
(118) Bruce, M. I.; Humphrey, P. A.; Okucu, S.; Schmutzler, R.;
Skelton, B. W.; White, A. H. Inorg. Chim. Acta 2004, 357, 1805.
(119) Bott, S. G.; Shen, H.; Richmond, M. G. J. Organomet. Chem.
2004, 689, 3426.
(120) Crochet, P.; Demerseman, B.; Rocaboy, C.; Schleyer, D.
Organometallics 1996, 15, 3048.
(121) Boren, B. C.; Narayan, S.; Rasmussen, L. K.; Zhang, L.; Zhao,
H.; Lin, Z.; Jia, G.; Fokin, V. V. J. Am. Chem. Soc. 2008, 130, 8923.
(122) Frisch, M. J.; et al. Gaussian 09, revision B.01; Gaussian, Inc.:
Wallingford, CT, 2010.
(123) Kulkarni, A. D.; Truhlar, D. G. J. Chem. Theory Comput. 2011,
7, 2325.
(124) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215.
(125) Zhao, Y.; Truhlar, D. G. J. Chem. Phys. 2006, 125, No. 194101.
(126) Leininger, T.; Nicklass, A.; Stoll, H.; Dolg, M.; Schwerdtfeger,
P. J. Chem. Phys. 1996, 105, 1052.
(81) Gossens, C.; Dorcier, A.; Dyson, P. J.; Rothlisberger, U.
Organometallics 2007, 26, 3969.
(82) Patel, A.; Ludi, A.; Burgi, H.-B.; Raselli, A.; Bigler, P. Inorg.
̈
Chem. 1992, 31, 3405.
(83) Jayaprakash, K. N.; Conner, D.; Gunnoe, T. B. Organometallics
2001, 20, 5254.
(84) Maire, P.; Breher, F.; Schonberg, H.; Grutzmacher, H.
̈
̈
Organometallics 2005, 24, 3207.
(85) Askevold, B.; Friedrich, A.; Buchner, M. R.; Lewall, B.; Filippou,
A. C.; Herdtweck, E.; Schneider, S. J. Organomet. Chem. 2013, 744, 35.
(86) Trujillo, H. A.; Casado, C. M.; Ruiz, J.; Astruc, D. J. Am. Chem.
Soc. 1999, 121, 5674.
(87) Zhong, Z.; Postnikova, B. J.; Hanes, R. E.; Lynch, V. M.; Anslyn,
E. V. Chem.Eur. J. 2005, 11, 2385.
(88) Ma, E. S. F.; Mudalige, D. C.; Patrick, B. O.; James, B. R. Dalton
Trans. 2013, 42, 7614.
(89) Jessop, P. G.; Morris, R. H. Inorg. Chem. 1993, 32, 2236.
(90) Bordwell, F. G. Acc. Chem. Res. 1988, 21, 456.
(91) Wei, C.; Aigbirhio, F.; Adams, H.; Bailey, N. A.; Hempstead, P.
D.; Maitlis, P. M. J. Chem. Soc., Chem. Commun. 1991, 883.
(92) Fan, L.; Wei, C.; Aigbirhio, F. I.; Turner, M. L.; Gusev, O. V.;
Morozova, L. N.; Knowles, D. R. T.; Maitlis, P. M. Organometallics
1996, 15, 98.
(127) Andrae, D.; Haußermann, U.; Dolg, M.; Stoll, H.; Preuß, H.
̈
Theor. Chim. Acta 1990, 77, 123.
(128) Clark, T.; Chandrasekhar, J.; Spitznagel, G. W.; Schleyer, P. v.
R. J. Comput. Chem. 1983, 4, 294.
(129) Lynch, B. J.; Zhao, Y.; Truhlar, D. G. J. Phys. Chem. A 2003,
107, 1384.
(93) Castro, A.; Turner, M. L.; Maitlis, P. M. J. Organomet. Chem.
(130) Frisch, M. J.; Pople, J. A.; Binkley, J. S. J. Chem. Phys. 1984, 80,
2003, 674, 45.
3265.
(94) Fairchild, R. M.; Holman, K. T. Organometallics 2008, 27, 1823.
(95) Fan, L.; Turner, M. L.; Adams, H.; Bailey, N. A.; Maitlis, P. M.
Organometallics 1995, 14, 676.
(131) Tomasi, J.; Mennucci, B.; Cammi, R. Chem. Rev. 2005, 105,
2999.
(132) Tomasi, J.; Mennucci, B.; Cances
(THEOCHEM) 1999, 464, 211.
̀
, E. J. Mol. Struct.
(96) Koelle, V.; Grub, J. J. Organomet. Chem. 1985, 289, 133.
4759
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760

Journal of the American Chemical Society
Article
(133) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B
2009, 113, 6378.
(134) Fukui, K. Acc. Chem. Res. 1981, 14, 363.
(135) Nagashima, K. J. Magn. Reson. 2008, 190, 183.
4760
dx.doi.org/10.1021/ja5006724 | J. Am. Chem. Soc. 2014, 136, 4746−4760