Selectivity of attack on a Si-C(sp3) sigma bond coordinated to NiII

Article abstract of DOI:10.1016/j.ica.2010.12.061

(PNP)Ni⁺ (as its (BAr^F4- salt) adds PhCN to Ni, but HX cleaves the Si-CH₂ bond to form Ni[η^2-( ^tBu₂PCH₂SiMe₂)N(H)(SiMe ₂X)][η²-CH₂

Full text of DOI:10.1016/j.ica.2010.12.061

Inorganica Chimica Acta 374 (2011) 79–87
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Selectivity of attack on a Si–C(sp3) sigma bond coordinated to NiII q
⇑
Matthew F. Laird, Nikolay P. Tsvetkov, Maren Pink, Tao He, René W. Buell, Kenneth G. Caulton
Department of Chemistry, Indiana University, Bloomington, IN, United States
a r t i c l e i n f o
a b s t r a c t
(PNP)Ni+ (as its (BArF salt) adds PhCN to Ni, but HX cleaves the Si–CH2 bond to form Ni[
g
2ꢀ(tBu2PCH2-
ꢀ
Article history:
Available online 30 December 2010
4
SiMe2)N(H)(SiMe2X)][
rotic reagent H2O gives (
g
2–CH2tBu2P]+, for X = OMe, piperidyl, N(H)CH2Ph, N(H)Ph, morpholinyl. The dip-
g
2–tBu2PCH2SiMe2OSiMe2NH2)( 2–tBu2CH2P)Ni+. RCCH (R = Ph, SiMe3, tBu)
g
Dedicated to Professor W. Kaim
reacts, through three detected intermediates, to form (tBu2PCH2SiMe2)N(H)(SiMe2CH2tBu2PCCR)Ni+, a
product where one P has been oxidized and Ni reduced, each by two electrons. This shows the dominant
influence on reactivity of Si–C bond activation by its unconventional donation to nickel in the structure of
(PNP)Ni+.
Keywords:
Nickel
Pincer
Si–C reactivity
Alkyne reactivity
Ó 2011 Elsevier B.V. All rights reserved.
1. Introduction
alcohols, or primary or secondary amines and even unconventional
nucleophiles such as RCCH, reveals that (PNP)Ni+ also shows Bron-
We have shown that the species (PNP)Ni+, 1, where PNP is the
anion (tBu2PCH2SiMe2)2Nꢀ1, synthesized by chloride removal from
(PNP)NiCl using anhydrous NaB(C6H3(CF3)2)4 (NaBArF4) in non-
coordinating solvent, is not simply a T-shaped three coordinate
species [1]. The Lewis acidity of such a structure is apparently very
high, since the ground state structure of the species (Scheme 1) has
one P located transoid to the amide nitrogen, and the electron den-
sity of one Si–C sigma bond donates to nickel, stretching that bond
by ꢁ0.2 Å. Studies with nucleophiles as weak as triflate show that
its addition product to (PNP)Ni+ is not that from binding to nickel,
but rather cleavage of the Si–C bond, to form a tBu2PCH2ꢀ ligand
which binds g2 to the metal. The LUMO of structure 1 [1] is along
the coordinated Si–C bond, indicating that Lewis base attack occurs
perpendicular to that bond, not ‘‘back side,’’ anti to the stretched
Si–C bond. When fluoride is the attacking nucleophile, the analo-
gous reaction happens, and it was possible to show that this Si/C
cleavage product is more thermodynamically stable than the con-
ventional isomeric (PNP)Ni–nucleophile structure. However, in its
reaction with H2, (PNP)Ni+ shows addition of nucleophilic Hꢀ to
nickel, not to the Si–C bond [2].
sted basicity at its amide nitrogen. Efforts to systematically under-
stand the multifunctional reactivity of (PNP)Ni+ are discussed. In
general, this report seeks to understand which of the product
structures in Scheme 1 is produced for a variety of reagents in their
reactions with (PNP)Ni+. Will reaction occur at nickel, 3, with dis-
placement of this unique Si–C donation to the metal, or will the
Si–C bond be directly attacked, 2?
We will explore here OH nucleophiles and also primary amines.
2. Results
2.1. General
The reaction of (PNP)Ni+, as its BArF salt, is conventional with
PhCN: a 1:1 adduct forms completely at nickel, and this adduct is
unchanged by vacuum drying at 25 °C. One additional equivalent
of PhCN causes no change in the spectra of (PNP)Ni(NCPh)+, indi-
cating that this four coordinate planar (C2v) species is not signifi-
cantly Lewis acidic, nor is there any attack at the Si–C bond.
There is no reaction between (PNP)Ni+ and 1 atm of N2 or N2O,
showing the limits of reactivity of this cation; ethylene, 1 atm, is
likewise unreactive, as is styrene, in equimolar amount. There is
no reaction (e.g., no proton transfer) between (PNP)Ni+ and
2,4,6-tBu3C6H2(OH), nor with CO2 or ethyl vinyl ether, all carried
out in CD2Cl2.
Both chloride and carbon monoxide bind at nickel however. In
short, the unconventional (PNP)Ni+ structure makes this cation a
multifunctional reagent, electrophilic at both Ni and Si.
We report here a more general exploration of the reactivity of
(PNP)Ni+ with nucleophiles of a variety of types, which reveal that
attack selectivity is determined by the arriving nucleophile. Using
2.2. Water
q
In honor of Wolfgang Kaim, who showed us much about spin delocalization.
⇑
One reaction was discovered, apparently due to adventi-
tious water in the unreactive reagents N2 and N2O. This product
Corresponding author.
E-mail address: caulton@indiana.edu (K.G. Caulton).
0020-1693/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2010.12.061

80
M.F. Laird et al. / Inorganica Chimica Acta 374 (2011) 79–87
Scheme 1.
Scheme 2.
(Scheme 2) has an AX 31P{1H} NMR pattern, and shows only Cs
symmetry in its 1H NMR spectrum. A crystal structure determina-
tion of 5 shows this to be the product of nucleophilic attack of
water oxygen on one silicon of the PNP ligand. Rearrangement, to-
its characteristic reactivity, and the reaction does not stop at the
simple adduct, (PNP)Ni(H2O)+. At shorter reaction times, an inter-
mediate is seen, then disappears; this species has the 31P NMR sig-
nature of the single Si/C cleavage product, 4, with trans phosphorus
groups, by comparison to other products reported later here.
This prompted us to consider more broadly the question of
where various weakly acidic nucleophiles will attack the unusual
structure of (PNP)Ni+, given that the Si–C bond is ‘‘activated’’ by
being bound to the metal. We have undertaken this study with
progressively more complicated analogs of water, being amines
(primary and secondary) and alcohols, to evaluate the effect of dif-
ferent numbers of mobile protons.
t
gether with two proton transfer steps, leaves one Bu2PCH2SiMe2–
O–SiMe2–NH2 ligand, together with the tBu2PCH2 ligand (Fig. 1). It
is thus a cation with atom composition being that of (PNP)Ni+ plus
one H2O. In the solid state, this exists as an ion pair via hydrogen
bonding of one of the NH protons to one F of CF3 of the BAr4 anion.
F
The identity of this product clearly establishes that silicon shows
2.3. MeOH [3–5]
Methanol was chosen to prevent formation of two Si/O bonds.
(PNP)Ni+ reacts with equimolar methanol in dichloromethane
within time of mixing at ꢀ78 °C to give complete conversion to a
single product with an AX 31P{1H} NMR spectrum whose JPP value,
0
182 Hz, indicates the phosphorus atoms to be trans in the coordi-
nation sphere. Structure 6, Scheme 3, is consistent with the 1H
NMR spectrum for this species, including an NH proton observed
(very broad) at 6.1 ppm. This shows that isomerization of the
tBu2PCH2 ligand to a location with the two P mutually trans has al-
ready occurred, but there is no proton migration to this CH2 on
nickel.
2.4. Primary amines
Benzyl amine reveals behavior similar to that of methanol in
spite of it being a primary amine, hence having two mobile pro-
tons. The reaction of equimolar reagents in dichloromethane gives
Fig. 1. ORTEP view (50% probabilities) of the non-hydrogen atoms of
(tBu2PCH2SiMe2OSiMe2NH2)(tBu2CH2P)Ni+ from its B(C6H3(CF3)2)4 salt, showing
selected atom labeling. Unlabelled atoms are carbons, and the two H on nitrogen
are indicated. Selected structural parameters (Å, °): Ni1–C14, 1.961(5); Ni1–N1,
2.001(4); Ni1–P2, 2.1802(13); Ni1–P1, 2.2574(12); Si1–O1, 1.650(3); Si2–O1,
1.625(3); C14–Ni1–N1, 159.00(17); N1–Ni1–P2, 109.61(11); C14–Ni1–P1,
99.61(14); N1–Ni1–P1, 101.08(10); P2–Ni1–P1, 148.82(6); Si2–O1–Si1,
137.42(19); Si2–N1–Ni1, 121.3(2).
0
an AX pattern (JPP = 178 Hz); all of these chemical shifts and J val-
ues are very similar to those observed with methanol, consistent
with the ring-opened product.
t
Reaction of (PNP)Ni+ with BuNH2 (chosen to simplify the 1H
NMR spectrum of the amine substituent) in CD2Cl2 proceeds to a

M.F. Laird et al. / Inorganica Chimica Acta 374 (2011) 79–87
81
three other singlet signals. This AX NMR pattern immediately con-
firms attack on the Si–C bond, to cleave the ligand backbone. If the
reaction is carried out at ꢀ78 °C, only the AX species is produced;
this solution converts to one of the three singlet species slowly
(2 days) at 25 °C, indicating the reaction path to involve an AX
intermediate which transforms to a species 8 with one phosphorus
NMR signal.
In the proton chemical shift region near 3 ppm of this final
product we observe a non-first order NMR pattern due to two
non-equivalent protons on the CH2 group adjacent to piperidine
N; these are well separated in chemical shift from the other
CH2CH2CH2 protons. This non-equivalence alone indicates that
the nitrogen is four coordinate (no rapid inversion at N), hence
bound to nickel. These signals undergo no change under conditions
4
Scheme 3.
of phosphorus decoupling, so any JPH is unresolved. Extraction
into pentane shows NMR spectra indicative of Cs symmetry: one
tBu chemical shift (a doublet), 2 SiMe2 chemical shifts, one CH2
chemical shift and no signals for BAr4F . The ESI-mass spectrum
showed the presence in 8 of one (solvent-derived) chloride, and
only one phosphorus in this compound, which is thus not a salt,
but a neutral molecule.
steady state in less than 10 min to produce one major product, hav-
ing an AX 31P{1H} NMR pattern. The J value in the AX pattern,
182 Hz, shows that these two phosphorus are mutually trans.
The AX product is unchanged in solution over a period of 5 days.
Even a less nucleophilic primary amine also attacks away from
the metal, to cleave the Si–C bond. Reaction with aniline is com-
plete in less than 5 min to form a product with spectra somewhat
analogous to the others reported above. The proton NMR shows a
In order to simplify the 1H NMR spectral interpretation for the
amine ring hydrogens, the reaction of (PNP)Ni+ with morpholine
was studied, since then only CH2CH2 groups are involved. Reaction
(1:1, at 25 °C) in CD2Cl2 shows primarily production of an AX
t
molecule with no symmetry, and Bu doublets indicate no strong
31P{1H} NMR pattern. The JPP value in the AX pattern, 180 Hz, indi-
0
coupling between phosphorus nuclei (hence phosphorus are not
trans). The absence of any symmetry indicates that the PNP nitro-
gen has been protonated (i.e., chiral N), as well as its backbone rup-
tured. The curious feature of the room temperature 31P{1H} NMR
spectrum is the fact that one of the peaks (ꢁ9 ppm) is very broad,
and the other peak, while sharper, is broad enough to not resolve
any PP0 coupling of less than 200 Hz. Variable temperature 31P
NMR spectra show reversible sharpening of this 9 ppm feature,
cates the two phosphorus nuclei to be mutually trans. After 5 days
in CD2Cl2, extraction of the dried residue into pentane shows this
fraction to have a phosphorus NMR singlet; the proton NMR spec-
trum of this pentane soluble fraction, recorded in benzene, shows
four chemical shifts in the 2.8–3.8 ppm region for the morpholine
hydrogens, thus indicating that this ring is not rapidly inverting,
consistent with the morpholine nitrogen being coordinated to Ni.
Two of these protons show geminal coupling, while the other
two show one additional coupling which we attribute to one re-
solved mutual vicinal coupling; this is characteristic of the dihedral
angle dependence of these hydrogens (Karplus relationship). The
proton NMR integrations are consistent with the presence of two
(non-equivalent) SiMe2 groups, but only one tBu2P group (doublet)
in this species, consistent with structure 8, Scheme 5. The formula,
and the presence of chloride, was established by ESI-MS of the pen-
tane soluble product; chloride was suggested to originate from
0
but never sharpening to the point where JPP is resolved, even at
ꢀ60 °C. Thus, the product detected has inequivalent phosphorus
nuclei, consistent with the other nucleophile reactions. The dy-
namic effects on phosphorus NMR are suggested to originate in
alternate conformations of the phenyl substituent, or also perhaps
hydrogen bonding involving either of the two NH protons.
2.5. Secondary amines
CH2Cl2. The pentane soluble fraction showed no BAr4 1H NMR sig-
F
The secondary amine piperidine was added to (PNP)Ni+ in
dichloromethane at 25 °C. An evident color change and 31P NMR
were used to establish complete conversion (7, Scheme 4) of 1 to
nals, consistent with this neutral complex identity. If the reaction
is repeated in fluorobenzene solvent, the product, which now per-
sists at 25 °C, has an AX 31P{1H} NMR spectrum with the large
(180 Hz) coupling constant characteristic of structure 7.
0
an AX spectrum (JPP = 183 Hz), together with lesser amounts of
Scheme 4.

82
M.F. Laird et al. / Inorganica Chimica Acta 374 (2011) 79–87
Scheme 5.
The recurrent abstraction of chloride from solvent represented
an undesirable feature, but is evidence of the intrinsic electrophi-
licity of the primary product 7, Scheme 5 and the need for product
stabilization if the amine proton is to migrate to the NiCH2PtBu2
carbon, forming PMetBu2. With this in mind, we attempted to
quench this product by adding MeI or benzyl bromide after the
morpholine attack at silicon was completed. Benzyl bromide (equi-
molar) was the more effective of these two, and led to formation of
neutral 8 and [tBu2PMe(CH2Ph)][BAr4F ], Scheme 5, which were sep-
arated by their distinct solubilities in pentane; the phosphonium
salt was identified by comparison to an independently synthesized
sample (from PMetBu2 and PhCH2Br in benzene). We interpret
these results as indicating that the ammonium center is acidic en-
ough to protonate the Ni/C bond if there is a nucleophile (better
than BAr4F ) to coordinate to nickel.
The silyl morpholine complex was identified by its distinctive
1H and 31P NMR spectra and (for the bromide) by ESI-MS. We pre-
fer coordination of nickel to the morpholine N based on the better
ring size involving nickel than by morpholine oxygen.
Fig. 2. ORTEP view (50% probabilities) of the non-hydrogen atoms of (tBu2PCH2Si-
Me2)N(H)(SiMe2CH2tBu2PCCtBu)Ni+ from its B(C6H3(CF3)2)4 salt, showing selected
atom labeling. Unlabelled atoms are carbons, and one H on nitrogen is shown.
Selected structural parameters(Å, °): Ni1–C24, 1.887(4); Ni1–C23, 1.896(4); Ni1–
N1, 2.087(3); Ni1–P1, 2.2225(10); C23–C24, 1.295(5); P2–C23, 1.729(3);C24–Ni1–
N1, 144.36(14); C23–Ni1–N1, 104.89(14); C24–Ni1–P1, 121.36(11); C23–Ni1–P1,
159.34(11); N1–Ni1–P1, 94.28(9); P2–C23–C24, 157.73(3).
There is no reaction between (PNP)Ni+ and equimolar NEt3.
2.6. Terminal alkynes
Terminal acetylenes, HC„CR might be expected to protonate
the amide nitrogen of (PNP)Ni+, to form a terminal acetylide com-
plex of the neutral ligand PN(H)P; RCCH are not nucleophiles capa-
ble of attack at silicon. In fact, in time of mixing, (PNP)Ni+ reacts
with RCCH where R = SiMe3, Me3C and phenyl. In each case the
product has Cs symmetry, but inequivalent phosphorus nuclei.
The coupling constant between the two phosphorus is small
(10 Hz) indicating a structure without two trans phosphorus. A sin-
gle crystal structure determination of the Me3C analog (Fig. 2 and
9) shows that there has been P/C bond formation to the terminal
acetylide carbon,[6] and the alkyne is g2 bonded to the nickel.
The presence of the former acetylene proton on N of the pincer li-
gand is established by the fact that this nitrogen is now pyramidal
with two Si and one Ni substituents, as well as by the longer Ni/N
distance than it is in related amide PNP/Ni complexes. The product
is thus a phosphonium species, so this is the site of the positive
charge in this monocation, and thus nickel has been reduced to oxi-
dation state zero in this reaction. The product contains three-coor-
dinate planar Ni(0). The source of the two reduction equivalents is
identified as the phosphorus, which is now pentavalent. The best
way to understand the formation of this P/C bond is to consider
functionality is formed in the product with the other two acety-
lenes. It is of interest that the cations with R = tBu and SiMe3 both
show (by 1H NMR) mirror symmetry in the molecular plane: gem-
t
inal Bu are pairwise equivalent as are geminal silyl methyls; in
contrast, for R = Ph, there are four tBu chemical shifts and four SiMe
chemical shifts. We attribute this to inversion of the amine proton
being rapid for the former two cases and slower for R = Ph, perhaps
enhanced by the ring current effect of phenyl increasing chemical
shift differences of geminal groups.
Since the observed terminal acetylene products involve multi-
ple bond making and breaking, can an intermediate (Scheme 6)
be detected? In fact, the slower case, where R = tBu, shows, at
25 °C at intermediate reaction times, two species with singlet
31P{1H} NMR signatures, but these are converted to the AX pattern
species after heating; they are thus candidate intermediates. If
equimolar (PNP)Ni+ and tBuCCH are combined at ꢀ80 °C in CD2Cl2
and monitored by 1H and 31P NMR with 10° increments of temper-
ature, one observes, from ꢀ40 to ꢀ20°, one broad 31P NMR peak
whose chemical shift is temperature dependent, indicating a rapid
equilibrium showing only one population-weighted average chem-
that an intermediate nickel acetylide is electrophilic at C , espe-
a
cially as P migrates to this carbon, and two electrons can ‘‘flow’’
to nickel. Given the similar spectroscopic properties, the same
ical shift. We suggest that this is an
g
2-adduct of the intact alkyne

M.F. Laird et al. / Inorganica Chimica Acta 374 (2011) 79–87
83
Scheme 6.
on (PNP)Ni+, where the adduct may have either one or both phos-
phorus donors coordinated, A. The proton NMR spectrum shows
two tBuP and one tBuCCH chemical shifts and three SiMe chemical
shifts. Beginning at ꢀ10 °C, the above species declines in popula-
tion smoothly with formation of a new species B with an AX
A recent report [8] describes reactions of the phenylene-back-
bone-ligated (pincer)Pd(OTf), where triflate leaving group should
provide a synthon for (pincer)Pd+, with various weak acids HX
(X = OR, PR2, CCR, SR. . .). That work, where the absence of silicon
eliminates any pincer backbone cleavage reactivity, shows either
addition of the H–X bond across the N/Pd bond, or simply coordi-
nation of the lone pair of X in HX to form (pincer)Pd(XH)+. The out-
come is thermodynamically controlled.
31P{1H} NMR pattern with a small JPP (5 Hz), suggesting a structure
0
similar to the final product (hence the P/C bond has formed). The
fate of the acetylenyl hydrogen is established, in B, as being a hy-
dride on Ni, which couples differently to the two inequivalent P
(28 and 4 Hz). Already at 0 °C, this species begins to transform into
the final product: hydrogen migration thus is somewhat slow from
Ni to amide N. At ꢀ20 °C, a second 1H NMR peak seen at ꢀ2.2 ppm
is a doublet (140 Hz) of doublets (15 Hz), and this we attribute (C)
to the acetylenic proton already on P (the larger J value indicates[7]
direct P–H bonding), which is then hydrogen bonded to the metal,
to give a second resolved JPH value. This is now zerovalent nickel in
C, so it is a good candidate for hydrogen bonding. A proton on
phosphorus is an attractive mechanism for transport of this H from
nickel to N, and indeed this species (as well as B) decreases in
intensity as the temperature rises further, with growth of the crys-
tallographic product. Note that in species B the redox change at Ni
has not yet occurred, and that the redox change involves phospho-
rus reducing nickel by 2 e as H migrates to P, forming the pentava-
lent phosphorus of the phosphonium functionality.
4. Experimental
4.1. General considerations
All manipulations were performed using standard Schlenk tech-
niques or in an argon-filled glovebox unless otherwise noted. Pen-
tane and THF were purified using an Innovative Technologies
solvent purification system Pure Solv 400-6-MD. Ethyl vinyl ether
(CH2@CH–OC2H5) was dried under molecular sieves, benzene D-6
was dried under Ph2CO/Na, PhCN was dried under CaH2 and each
chemical was vacuum transferred and stored in the glovebox un-
der argon. H2O was degassed by freeze–pump–thaw technique be-
fore use. Carbon dioxide, nitrogen, N2O and ethylene were
purchased through commercial vendors and used without further
purification. All reactions with gases were accomplished with a
gas line equipped with manometer (0–760 Torr) for accurate dos-
age. {[(tBu2PCH2SiMe2)2N]Ni}B(C6H3(CF3)2)4 was prepared follow-
ing the published synthesis [1]. NMR chemical shifts are reported
in ppm relative to protio impurities in the deutero solvents. Cou-
pling constants are given in Hz. 31P NMR spectra are referenced
to external standards of H3PO4. NMR spectra were recorded with
a Varian Unity INOVA instrument (400 MHz 1H; 162 MHz 31P).
Infrared spectra were recorded on a Nicolet 510P FT-IR spectrom-
eter. Mass spectra were acquired on a MAT-95-XP spectrometer
(Thermo Electron Corp., Bremen, Germany). ‘‘PNP’’ is N(Si-
3. Results and discussion
This and earlier work confirms (Scheme 7) some of the reactiv-
ity expected from a Si–heteroatom bond, but in a different version.
In our earlier observations of this reaction, fluoride or even triflate
t
nucleophiles gave neutral products with an Bu2PCH2 ligand g2
-
coordinated to Ni. In the present cases, an active proton leads to
protonation of the amide nitrogen. The case of terminal alkynes
represent different behavior since the acetylide ligand on nickel
is apparently sufficiently electrophilic that it succumbs to attack
by phosphorus.
Me2CH2PBu2 )2. ArF is 3,5-bis-trifluoromethylphenyl.
t

84
M.F. Laird et al. / Inorganica Chimica Acta 374 (2011) 79–87
4.1.2. Reaction of (PNP)Ni+ (BArF4ꢀ) with H2O
4.1.1. Synthesis of [(PNP)Ni(PhCN)](BAr4F)
30.0 mg (0.022 mmol) of [(PNP)Ni](BAr4F ) was dissolved in
0.5 mL of CD2Cl2 in a J-Young NMR tube yielding a dark reddish
brown solution. To this solution, 0.0004 mL of H2O (0.4 mg,
0.022 mmol) was added at 25 °C via syringe. In time of mixing,
the solution became orange in color. Full conversion of the starting
material to the mixture of two products was observed. One could
see complete conversion of one of these two into another in 16 h
in CD2Cl2. Trans (thermodynamic) isomer: 1H NMR (25 °C, CD2Cl2):
30.0 mg (0.022 mmol) of [(PNP)Ni](BAr4F ) was dissolved in
0.5 mL of CD2Cl2 in a J-Young NMR tube yielding a dark reddish
brown solution. To this solution, 0.0022 mL of benzonitrile
(2.3 mg, 0.022 mmol) was added at 25 °C via syringe. In time of
mixing, the solution became brown in color. Full conversion of
the starting material to the single product, [(PNP)Ni(PhCN)](BArF4ꢀ
)
was observed. 1H NMR (25 °C, CD2Cl2): 0.23 (s, SiMe, 12H), 0.84 (t,
t
J = 5.8, 4H, CH2), 1.45 (t, J = 6.3, 36H, Bu), 7.50–7.90 (m, 17H, Ar).
t
31P{1H} NMR (25 °C, CD2Cl2): 58.8 (s). This solution was concen-
trated in vacuum and pumped for 30 min at 25 °C to give
28.4 mg (88%) of the product. The residue was dissolved in CD2Cl2
and NMR spectra did not show any changes. One more equivalent
of benzonitrile (0.0022 mL) was added and no changes were seen
in the 1H and 31P NMR.
1.29 and 1.40 (both d, 18H each, Bu, J = 13.6 and 15.2), 7.56 and
7.71 (both s, Ar, 4 and 8H), other signals (SiMe and CH2 protons
and NH2) were not identified due to the broad nature of these
peaks. 31P{1H} NMR (25 °C, CD2Cl2): 2.0 and 38.6 (both d,
J = 164). The intermediate, containing only one Si–O bond, was
identified by 31P{1H} NMR (25 °C, CD2Cl2): ꢀ14.9 and 51.2 (both
Scheme 7.

M.F. Laird et al. / Inorganica Chimica Acta 374 (2011) 79–87
85
d, J = 174) as a structure similar to the ‘‘MeOH product’’ (see be-
low). Crystals suitable for X-ray diffraction analysis were grown
in 2 days at 25 °C after a solution of the final product in CD2Cl2
was layered with pentane.
4.1.5. Reaction of (PNP)Ni+ (BArF4ꢀ) with PhCH2NH2
30.0 mg (0.022 mmol) of [(PNP)Ni](BAr4F ) was dissolved in
0.5 mL of CD2Cl2 in a J-Young NMR tube yielding a dark reddish
brown solution. To this solution, 0.0024 mL of benzylamine
(2.35 mg, 0.022 mmol) was added at 25 °C via syringe. Full conver-
sion of the starting material to the mixture of products was ob-
served. 31P{1H} NMR (25 °C, CD2Cl2): ꢀ32.4 and 47.0 (both d,
J = 178). Other products are seen as singlets at 11.4, 58.9 and
64.5 ppm in 31P NMR. Ratio (AX):singlets = 2:1. There is no change
in population of these products after 12 h at 25 °C.
4.1.6. Reaction of (PNP)Ni+ (BArF4)ꢀ with PhNH2
15.5 mg (0.0113 mmol) of [(PNP)Ni](BAr4F ) was dissolved in
0.5 mL of CD2Cl2 at room temperature in a J-Young NMR tube,
yielding a dark reddish brown solution. To this solution, 1.01 lL
of aniline (1.03 mg, 0.0113 mmol) was added at 25 °C via syringe.
Five minutes after mixing, a reddish-purple product was observed,
with compete consumption of starting material, with the following
spectroscopic features.
4.1.3. Reaction of (PNP)Ni+ (BArF4ꢀ) with MeOH
30.0 mg (0.022 mmol) of [(PNP)Ni](BAr4F ) was dissolved in
0.5 mL of CD2Cl2 in a J-Young NMR tube yielding a dark reddish
brown solution. This solution was frozen with liquid nitrogen
and 0.0009 mL of methanol (0.7 mg, 0.022 mmol) was vacuum
transferred. The solution was allowed to warm and the tube was
shaken and one could see a color change to orange. Full conversion
of the starting material to one product was observed. 1H NMR
(25 °C, CD2Cl2): 1.34 and 1.40 (both d, J = 13.7 and 15.5, 18H,
tBu), 3.43 (s, 3H, OCH3), 6.3 (br. s, 1H, N–H), 7.58 and 7.75 (both
s, 4 and 8H, ArF); CH2 and SiMe signals were not assigned with cer-
tainty due to strong interfering signals. 31P{1H} NMR (25 °C,
CD2Cl2): ꢀ26.5 and 49.8 (both d, J = 182).
1H NMR (25 °C, CD2Cl2): 0.05 (s, 3H, SiMe), 0.16 (s, 6H, SiMe2),
0.33 (s, 3H, SiMe), 0.78 (d, 2H, CH2), 1.43 and 1.36 (both d,
J = 16.0 for each, 18H, tBu), 1.48 and 1.56 (both d, J = 12.0 for each,
18H, tBu), 3.23 (d, J = 12.0, 1H, PhNH), 7.36 (m, ꢁ4H, Ph), 7.58 and
7.71 (both s, 4 and 8H, ArF). The other CH2 signal was not identified
due to masking by stronger signals. 31P{1H} NMR (25 °C, CD2Cl2):
90% of the signals are those at 12.36 (s, vb), and 69.68 (s).
31P{1H} NMR (ꢀ60 °C, CD2Cl2): 69.5 (s) and 9.52 (s, br). HRMS
(ESI+, CH2Cl2): Observed at 599.3041 (C28H59N2Si2P2Ni, [M]+), calc.
599.3046.
4ꢀ
t
4.1.4. Reaction of (PNP)Ni+ (BArF ) with BuNH2
4.1.7. Reaction of (PNP)Ni+ (BArF4ꢀ) with piperidine in CD2Cl2
30.0 mg (0.022 mmol) of [(PNP)Ni](BAr4F ) was dissolved in
0.5 mL of CD2Cl2 in a J-Young NMR tube yielding a dark reddish
brown solution. To this solution, 0.0023 mL of t-butylamine
(1.6 mg, 0.022 mmol) was vacuum transferred. After the solution
was allowed to warm to room temperature and shaken, the color
became orange. Full conversion of the starting material to one
NiCl(CH2PtBu2)(N(SiMe2CH2 Bu2)(SiMe2-
t
4.1.7.1.
Synthesis
of
N(CH2)5)]. 30.0 mg (0.022 mmol) of [(PNP)Ni](BAr4) was dissolved
F
in 0.5 mL of CD2Cl2 in a J-Young NMR tube forming a dark reddish
brown solution. To this solution, one equivalent of piperidine
(0.0021 mL) was added at 25 °C via syringe. Initial 31P{1H} NMR
spectra showed an AX pattern with JPP = 180, as well as three other
singlets. 1H NMR after only 20 min showed complete conversion of
(PNP)Ni+. After 30 min the CD2Cl2 was stripped off and replaced
with pentane and filtered. The pentane-soluble portion was sub-
jected to further characterization and a four coordinate, chloride-
containing product with only one phosphorus atom was deduced,
by 1H NMR integrations and especially by mass spectrometry. 1H
NMR (25 °C, C6D6): 0.13, 0.17 (both s, 6H each, SiMe), 1.47 (d,
product was observed. 1H NMR (25 °C, CD2Cl2): 1.18 (s, 9H, Bu),
t
t
1.36 and 1.43 (both d, 18H each, Bu, J = 15.0 and 15.4), 7.58 and
7.75 (both s, Ar, 4 and 8H). 31P{1H} NMR (25 °C, CD2Cl2): ꢀ31.4
and 46.2 (both d, J = 182).
t
J = 13.8, 18H, Bu), 3.26 (d, J = 13.3, 2H, two C–H of two N–CH2),
3.40 (t, J = 13.3, 2H, two C–H of two N–CH2); other CH2 protons
on ligand and piperidine are obscured by stronger ligand
resonances. 31P{1H} NMR (25 °C, C6D6): 59.8. MS (CI, from pen-
tane): Observed at 466.1650 (C18H42N2Cl1Ni1P1Si2, [M]+), calc.
466.1666; obs. 431.1963 (C18H42N2Ni1P1Si2, [MꢀCl]+), calc.
431.1978; obs. 383.0919 (C13H32N1Cl1Ni1P1Si2, [MꢀC5H10N+H]+),
calc. 383.0931.

86
M.F. Laird et al. / Inorganica Chimica Acta 374 (2011) 79–87
was partially dissolved in dry fluorobenzene (2 mL). The suspen-
sion was placed in a freezer and cooled to ꢀ40 °C. To the suspen-
sion, morpholine (7.6 mg, 0.088 mmol) was slowly added via a
pipette. The suspension was left in the freezer for half an hour to
produce a dark red solution. Benzyl bromide (15 mg, 0.088 mmol)
was then added to the solution, and the solution was kept in free-
zer for another half hour. The solution was then stripped to dry-
ness. Pentane (5 mL) was added to extract the non-polar fraction
which was then filtered to yield a purple solution. The volatiles
were removed in vacuo to yield a purple solid for NMR analysis.
For the pentane soluble fraction: 1H NMR (C6D6, 25 °C): 3.75 (t,
two CH in two equivalent CH2, J = 11.6, 2H), 3.33 (d, two CH in
two equivalent CH2, J = 12.0, 2H), 3.05 (t, two CH in two equivalent
CH2, J = 11.6, 2H), 2.85 (d, two CH in two equivalent CH2, J = 13.6,
4.1.8. Reaction of (PNP)Ni+ (BArF4ꢀ) with morpholine in CD2Cl2
4.1.8.1. Synthesis of [NiCl(N(SiMe2CH2 Bu2)(SiMe2N(CH2CH2)2-
t
t
2H), 1.45 (d, 18H, Bu, J = 13.8), 0.11 (s, 6H, SiMe2), 0.06 (s, 6H,
SiMe2). 31P{1H} NMR (C6D6, 25 °C): 63.9 (s). Pentane insoluble frac-
tion: [Me(C6H5CH2)PtBu2][BAr4F ]: the pentane insoluble residue
above was redissolved in CD2Cl2 for NMR study. 1H NMR (CD2Cl2,
25 °C): 7.76 (d, J = 7.9, Ar, 2H), 7.56 (d, J = 6.0, Ar, 3H), 4.30 (d,
2H, JPH = 13.8, CH2), 2.18 (d, 3H, JPH = 12.0, CH3), 1.61 (d, 18H,
O))]. Reaction was carried out as with piperidine, above, but
substituting morpholine. 1H NMR (25 °C, C6D6): 0.05, 0.09 (both
t
s, 6H each, SiMe), 1.44 (d, J = 13.8, 18H, Bu), 2.89 (d, J = 13.7, 2H,
two CH in two equivalent CH2), 3.04 (t, J = 11.3, 2H, two CH in
two equivalent CH2), 3.34 (d, J = 11.1, 2H, two CH in two equivalent
CH2), 3.64 (t, J = 13.0, 2H, two CH in two equivalent CH2); CH2 pro-
tons on PNP ligand are obscured by stronger resonances. 31P{1H}
NMR (25 °C, C6D6): 60.4.
t
two Bu, JPH = 15.3). 31P{1H} NMR (CD2Cl2, 25 °C): 46.6 (s).
4.1.10.2. Independent synthesis of phosphonium salt. One equivalent
of benzyl bromide and PMetBu2 were mixed in benzene, forming a
colorless precipitate. After 12 h at 25 °C all volatiles were removed
in vacuum and the residue was dissolved in CD2Cl2. 1H and 31P
NMR showed spectral data identical to those of the pentane insol-
uble product (see above).
4.1.9. Reaction of (PNP)Ni+ (BArF4ꢀ) with morpholine in PhF
4.1.9.1. Synthesis of [Ni(CH2PtBu2)(NH(SiMe2CH2 tBu2)(SiMe2N(CH2-
CH2)2O))]+[BArF4]ꢀ. [(PNP)Ni][BAr4F ] (120 mg, 0.088 mmol) was par-
tially dissolved in dry fluorobenzene (2 mL). The suspension was
placed in a freezer and cooled to ꢀ40 °C. To the suspension, mor-
pholine (7.6 mg, 0.088 mmol) was slowly added via a pipette.
The suspension was left in the freezer for another half an hour to
achieve completion: all solids dissolved to produce a dark red solu-
tion. 31P{1H} NMR (PhF, 25 °C): 47.8 (d, J = 180), ꢀ34.3 (d, J = 180).
4ꢀ
t
4.1.11. Reaction of (PNP)Ni+ (BArF ) with BuC„CH
30.0 mg (0.022 mmol) of [(PNP)Ni][BAr4F ] was dissolved in
0.5 mL of THF in a J-Young NMR tube yielding a dark reddish
brown solution. To this solution, 0.0027 mL of tert-butyl acetylene
(0.022 mmol) was added at 25 °C. The solution became increas-
ingly orange in color, after which it was heated in an oil bath at
60 °C for 2 days. After the heating period, full conversion of the
starting material to the single product, [(PNHP+C„C(t-Bu))Ni]
(BAr4Fꢀ) was observed. The solution was dried and dissolved in
dichloromethane and layered with pentane, which produced crys-
tals suitable for X-ray diffraction analysis. 1H NMR (25 °C, d8THF):
0.38 (s, 6H, SiMe2), 0.53 (s.br, 6H, SiMe2), 1.01 (d, J = 9.6, 2H, CH2),
t
t
1.33 (s, 9H, Bu), 1.35 (d, J = 12.8, 18H, Bu), 1.47 (d, J = 15.4, 18H,
tBu), 7.57 and 7.78 (both s, 4 and 8H, Ar); N–H proton was not lo-
cated. 31P{1H} NMR (25 °C, d8THF): 62.2 (d, J = 10), 35.7 (d, J = 10).
4.1.10. Reaction of (PNP)Ni+ (BAr4Fꢀ) with morpholine in PhF followed
by addition of electrophile
4.1.10.1. Synthesis of [NiBr(N(SiMe2CH2 tBu2)(SiMe2N(CH2CH2)2O))].
Pentane soluble fraction. [(PNP)Ni][BAr4F ] (120 mg, 0.088 mmol)

M.F. Laird et al. / Inorganica Chimica Acta 374 (2011) 79–87
87
4.1.12. Reaction of (PNP)Ni+ (BArF4ꢀ) with Me3SiC„CH
giving a dark reddish brown solution. The tube was placed into a
Dewar containing an acetone/dry ice mixture at ꢀ78 °C with spe-
cial effort to cool the upper glass walls of the tube. To this solution,
one equivalent of tert-butyl acetylene (0.0027 mL) was added via
syringe. The tube was carefully shaken in order to mix the reagents
and NMR spectra were recorded beginning at ꢀ60 °C. Species A:
31P{1H} NMR (CD2Cl2) 63.5 (br) (ꢀ40 °C), 73.4 (br) (ꢀ30 °C), 89.0
(br) (ꢀ20 °C), 110.0 (br) (ꢀ10 °C); 1H NMR: (ꢀ20 °C, CD2Cl2) 0.04,
30.0 mg (0.022 mmol) of [(PNP)Ni][BAr4F ] was dissolved in
0.5 mL of CD2Cl2 in a J-Young NMR tube forming a dark reddish
brown solution. To this solution, 0.0030 mL of Me3SiC„CH
(0.022 mmol) was added at 25 °C. In time of mixing, the solution
became orange in color. Full conversion of the starting material
to the product, [(PNHP+C„CTMS)Ni](BAr4F ) was confirmed by
NMR after ten minutes. 1H NMR (25 °C, CD2Cl2): 0.24 (s, 9H,
TMS), 0.34 (s, 6H, SiMe2), 0.48 (s, 6H, SiMe2), 0.90 (d, J = 9.3, 2H,
t
t
0.07, 0.47 (s, SiMe2), 1.02 (s, BuC), 1.37–1.45 (m, BuP), 7.58, 7.74
(s, Ar). Species B: 31P {1H} NMR (0 °C, CD2Cl2) 46.2, 76.3 (both d,
J = 5); 1H NMR (0 °C, CD2Cl2): ꢀ11.21 (dd, JHP = 4.8, 23, Ni–H),
0.09, 0.12 , 0.21, 0.54 (s, SiMe2), 1.27 (s, tBu–C), 1.33 (d, t Bu–P,
t
t
CH2), 1.30 (d, J = 13.0, 18H, Bu), 1.38 (d, J = 15.5, 18H, Bu), 1.53
(d, J = 12.2, 2H, CH2), 7.57 and 7.74 (both s, 4 and 8H, Ar); N–H pro-
ton was not located. 31P{1H} NMR (25 °C, CD2Cl2): 64.9 (d, J = 10),
32.8 (d, J = 10).
t
t
J = 17), 1.39 (d, BuP, J = 16), 1.48 (d, BuP, J = 13.6). Species C: 1H
NMR (0 °C, CD2Cl2): ꢀ2.2 (dd, J = 14, 140); 31P {1H} NMR (CD2Cl2)
37.1 (d) (ꢀ40 °C), 36.6 (d) (ꢀ30 °C), 35.6 (d) (ꢀ20 °C), 34.1 (d)
(ꢀ10 °C). The second 31P NMR signal for C is considered to be ex-
change-broadened (H+ transfer) beyond detection.
Acknowledgments
We thank the National Science Foundation (NSF CHE-0544829)
for financial support.
Appendix A. Supplementary material
4.1.13. Reaction of (PNP)Ni+ (BArF4ꢀ) with PhC„CH
30.0 mg (0.022 mmol) of [(PNP)Ni][BAr4F ] was dissolved in
0.5 mL of CD2Cl2 in a J-Young NMR tube leaving a dark reddish
brown solution. To this solution, 0.0024 mL of phenyl acetylene
(0.022 mmol) was added at 25 °C. The solution immediately be-
came orange in color. Full conversion of the starting material to
one product, [(PNHP+C„CPh)Ni] (BAr4F ) was confirmed by NMR
after 10 min. 1H NMR (25 °C, CD2Cl2): 0.33, 0.43, 0.44, 0.73 (all s,
3H each, all SiMe2), 0.94 (dd, J = 4.3 and 9.5, 2H, CH2), 1.10 (d,
CCDC 789986 and 789987 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif. Supplementary data associ-
ated with this article can be found, in the online version, at
doi:10.1016/j.ica.2010.12.061.
t
t
References
J = 13.5, 9H, Bu), 1.18 (d, J = 13.1, 9H, Bu), 1.24 (d, J = 14.8, 9H,
t
tBu), 1.39 (d, J = 13.4, 9H, Bu), 7.07 and 7.25 (both m, 2 and 3H,
[1] H. Fan, B.C. Fullmer, M. Pink, K.G. Caulton, Angew. Chem., Int. Ed. 47 (2008)
9112.
[2] T. He, N.P. Tsvetkov, J.G. Andino, X.-F. Gao, B.C. Fullmer, K.G. Caulton, J. Am.
Chem. Soc. 132 (2010) 910.
[3] M.D. Fryzuk, P.A. MacNeil, J. Am. Chem. Soc. 106 (1984) 6993.
[4] M.D. Fryzuk, P.A. MacNeil, S.J. Rettig, J. Organomet. Chem. 332 (1987) 345.
[5] M.D. Fryzuk, P.A. MacNeil, S.J. Rettig, A.S. Secco, J. Trotter, Organometallics 1
(1982) 918.
[6] H. Lang, M. Leise, W. Imhof, Z. Naturforsch., B: Chem. Sci. 46 (1991) 1650.
[7] S.L. Manatt, G.L. Juvinall, D.D. Elleman, J. Am. Chem. Soc. 85 (1963) 2664.
[8] L.C. Gregor, C.-H. Chen, C.M. Fafard, L. Fan, C. Guo, B.M. Foxman, D.G. Gusev, O.V.
Ozerov, Dalton Trans. 39 (2010) 3195.
Ph), 7.58 and 7.74 (both s, 4 and 8H, B–Ar); N–H proton was not
located. 31P{1H} NMR (25 °C, CD2Cl2): 65.7 (d, J = 10), 34.4 (d,
J = 10).
4.2. Low temperature NMR monitoring of the reaction of (PNP)Ni+
(BAr4Fꢀ) with BuC„CH
t
30.0 mg (0.022 mmol) of [(PNP)Ni][BAr4F ] was dissolved in
0.5 mL of CD2Cl2 in a J-Young NMR tube equipped with a septum,