C-H Nickelation of Aryl Phosphinites: Mechanistic Aspects

Article abstract of DOI:10.1021/acs.organomet.8b00899

The present report describes the results of a combined experimental and computational study on the mechanism of aryl phosphinite cyclonickelation. The reaction of ArOP(i-Pr)₂ with [(i-PrCN)NiBr₂]_n proceeds more readily in

Full text of DOI:10.1021/acs.organomet.8b00899

Article
pubs.acs.org/Organometallics
Cite This: Organometallics XXXX, XXX, XXX−XXX
C−H Nickelation of Aryl Phosphinites: Mechanistic Aspects
Loïc P. Mangin and Davit Zargarian*
́
́
́
́
́
Departement de chimie, Universite de Montreal, Montreal, Quebec, Canada H3C 3J7
S
* Supporting Information
ABSTRACT: The present report describes the results of a
combined experimental and computational study on the
mechanism of aryl phosphinite cyclonickelation. The reaction
of ArOP(i-Pr)2 with [(i-PrCN)NiBr2]n proceeds more readily
in acetonitrile relative to toluene; this is because the greater
nucleophilicity of acetonitrile toward Ni stabilizes a more
reactive acetonitrile adduct bearing one phosphinite ligand (as
opposed to two). A sufficiently strong external base such as
Et3N serves to quench the HBr generated at the nickelation
step, thus allowing isolation of the cyclonickelated species.
However, nickelation tests conducted in the absence of external base revealed that D/H scrambling occurs at the ortho
positions of C6D5OP(i-Pr)2, implying that the cyclonickelation proceeds independently of the base. Thus, the main role of the
external base is to prevent protonation of the Ni−aryl moiety formed via C−H nickelation. Tests have also shown that
nickelation rates are affected by the quantity of the base used: the presence of more than 1 equiv of Et3N generates the less
reactive biphosphinite complex trans-{PhOP(i-Pr)2}2NiBr2, thus inhibiting the desired C−H nickelation. Experimental studies
have shown that nickelation is faster with aryl phosphinites bearing electron-donating substituents (Hammett slope ρ ≈ −4)
and the proton transfer is rate limiting (KIE ≈ 11). The activation parameters were found to be ΔH⧧ = 17.7 kcal mol−1 and ΔS⧧
= −27.1 cal mol−1 K−1. DFT analyses have provided support for these findings and suggest that aryl phosphinite C−H
nickelation proceeds via an ion-pair-assisted deprotonation.
reaction mechanisms. Ideally, therefore, systematic efforts are
needed to isolate model cyclonickelated complexes and
delineate the factors that favor various steps of the C−H
functionalization process.
INTRODUCTION
■
The past decade has witnessed an increasing number of reports
that disclose Ni(II)-catalyzed protocols for the direct
functionalization of unactivated C−H moieties. Such non-
redox-type C−H metalation processes, followed by function-
alization, are facilitated by the use of at least one directing
group (DG) that serves the purpose of binding to the Ni(II)
center and facilitating its encounter with the C−H moiety.
This type of metalation by chelation also helps determine the
regiochemistry of the functionalization, which is likely
controlled by the relative stabilities of the nickelacycles
generated in the C−H nickelation step. In this context, the
most commonly used DGs are mono- and bidentate auxiliaries
based on imine-type donors, such as 2-pyrimidyl anilines
(monodentate)1 and benzamido quinolones and pyridines
(bidentate),2 which were popularized by the groups of
Ackermann and Chatani, respectively. Another effective DG
is pyridine oxide.3
In contrast to the significant progress registered in Ni(II)-
catalyzed direct functionalization of unactivated C−H bonds,
there is relatively little concrete mechanistic knowledge of how
these processes work.4 The main reason for this paucity of
mechanistic knowledge is that most efforts to date have
focused on developing one-pot protocols that combine the
chelation, metalation, and functionalization steps. While such a
one-pot approach is attractive from a practical point of view, it
bypasses the detection and isolation of reaction intermediates
and foregoes the chance to improve our understanding of
The above considerations and our group’s longstanding
interest in organonickel chemistry5 inspired us to investigate
the preparation of thermally stable cyclonickelated species via
C−H nickelation and study the direct functionalization of the
resulting Ni−hydrocarbyl moiety. A search of the literature
revealed that most cyclonickelated complexes accessible via
C−H nickelation are stabilized by pincer-type ligands.6 Our
experience with the facile C−H nickelation of POCOP-type
pincer ligands featuring aryl and alkyl phosphinite donor
moieties7 encouraged us to test the reactivities of substrates
bearing a single phosphinite DG. Initial tests showed that alkyl
phosphinites were insufficiently reactive, whereas aryl
phosphinites underwent relatively facile C−H nickelation.
Thus, we succeeded in accessing the isolable and thermally
stable ortho-metalated complexes {(κP,κC-ArOPR2)Ni(μ-Br)}2
and (κP,κC-ArOPR2)Ni(L)Br (Chart 1; R = i-Pr) and then
studied their structures, stabilities, and reactivities with
electrophiles.8
In addition to the aforementioned promising reactivities of
aryl phosphinites, we should also note that these substrates
present many advantages in the context of the proposed
Received: December 12, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.organomet.8b00899
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
cyclonickelation in acetonitrile might be this solvent’s greater
nucleophilicity toward Ni(II).11 Thus, the more favorable
MeCN→Ni coordination could conceivably generate a solvent
adduct such as (L)(MeCN)NiBr2 (1a, L = PhOP(i-Pr)2) that
would be more reactive toward cyclonickelation relative to the
biphosphinite complex trans-L2NiBr2 (1b) present in tol-
uene.12
Chart 1. Cyclonickelated Species Derived from Aryl
Phosphinites
The assertion that a different intermediate might form in
acetonitrile was consistent with the observation of differently
colored initial reaction mixtures as a function of solvent. For
instance, stirring the insoluble Ni precursor [(i-PrCN)NiBr2]n
in toluene solutions of PhOP(i-Pr)2 at room temperature
generates the toluene-soluble biphosphinite adduct 1b. The
deep red color of this complex dominates the heterogeneous
mixture that forms in the presence of less than 2 equiv of L,
whereas a homogeneous deep red solution is observed with 2
equiv or more of L. In other words, the L:Ni ratio affects the
homogeneity of these toluene mixtures, but not their color.
The 31P NMR spectrum of the latter mixtures displayed,
independently of the L:Ni ratio, a fairly broad peak spanning
the chemical shift interval 132−138 ppm, which could be
attributed to dynamic exchange. Recording the 31P spectrum at
−30 °C gave significant peak sharpening (Figure S15), and
displayed a major peak at 135 ppm, along with a minor peak at
130 ppm which can be attributed to a different conformer of
1b (i.e., C2 symmetric vs centrosymmetric).
Repeating the above experiment in acetonitrile revealed two
different color regimes depending on the L:Ni ratio. Thus,
addition of 1 equiv of phosphinite to the acetonitrile
suspension of [(i-PrCN)NiBr2]n gave a homogeneous green
solution displaying no 31P signal at all, whereas adding another
1 equiv of L led to a deep red solution displaying a very broad
31P signal spanning the chemical shift range of 120−150 ppm.
This latter signal was confirmed to be due to 1b by recording
the 31P NMR spectrum of an authentic sample of this
compound in acetonitrile. Recording the 31P spectrum of 1b at
−30 °C revealed sharp peaks at 135 ppm (major) and 130
ppm (minor), as well as a peak at 150 ppm corresponding to
the free ligand. This indicates that MeCN is nucleophilic
enough to allow displacement of the phosphinite and that
lowering the temperature slows the exchange process
sufficiently to allow us to detect the species involved in the
exchange (Figure S16). Moreover, when the green 1:1 L:Ni
mixture was subjected to the same analysis, the only detectable
peaks were at 135 ppm (major) and 130 ppm (minor, Figure
S17); this implies that 1b also exchanges with the
studies: they are derived from phenols, which are inexpensive
and ubiquitous in organic synthesis; the PR2 directing group is
easily grafted onto phenols by reaction with ClPR2; 31P NMR
spectroscopy provides a convenient and powerful method for
monitoring reactions and detecting intermediates. These
advantages compelled us to undertake detailed mechanistic
studies on C−H cyclonickelation of aryl phosphinites. In a
recent report,9 we presented an optimized synthetic protocol
for the preparation of cyclonickelated complexes shown in
Chart 1 and provided a detailed analysis of the electronic and
steric effect of phenol substituents on cyclonickelation rates,
regiochemistry, and solid-state structures of the cyclonickelated
complexes.
The present contribution reports on our continued
investigations of the reactivities of Ni(II) precursors with
aryl phosphinites. The focus of this report is identification of
the in situ generated Ni species that promote the C−H
nickelation step. Our investigation of the reaction conditions
affecting the rate of this step has revealed a crucial role for the
solvent, as well as the type and quantity of amine bases used.
Deuterium labeling experiments have revealed that the C−H/
C−D nickelation step occurs in the absence of base, but
isolation of the cyclonickelated product arising from this
reaction requires a relatively strong base. Kinetic studies and a
Hammett analysis have provided quantitative data about the
rate and nature of the C−H nickelation step. Finally, DFT
studies have supported the mechanistic insights obtained from
experimental studies.
RESULTS AND DISCUSSION
■
Effect of Solvent on Cyclonickelation. We have
reported previously that ortho-C−H nickelation of aryl
phosphinites is faster in acetonitrile relative to toluene, THF,
and EtOAc.10 Initially, we attributed this observation to the
greater ability of a polar solvent such as acetonitrile for
stabilizing the nickelation transition state, which is believed to
involve significant charge separation over the reacting moieties
Cδ−- - -Hδ+ and Niδ+- - -Brδ−.10 However, on reflection it
occurred to us that another reason for the more facile
Scheme 1. Different Ni Species Postulated To Exist in Toluene and Acetonitrile Prior to the Nickelation of PhOP(i-Pr)2
B
DOI: 10.1021/acs.organomet.8b00899
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
biphosphinite species, even though the latter is present to a
significantly smaller extent in MeCN than in nonpolar solvents.
The above observations allow us to propose the sequence of
events shown in Scheme 1 for rationalizing how PhOP(i-Pr)2
and [(i-PrCN)NiBr2]n interact prior to the cyclonickelation
stage. We believe that dynamic equilibria are taking place in
both solvents at room temperature. In toluene, these equilibria
always favor the deep red biphosphinite species 1b, whereas in
acetonitrile it is the green monophosphinite adduct 1a that is
dominant in the presence of 1 equiv of phosphinite. The
stability of the monophosphinite species 1a is presumably due
to the nucleophilic character of acetonitrile, whereas related
monophosphinite intermediates do not appear to be stable in
toluene, THF, or EtOAc.13 Furthermore, given the broadness
of the 1H and 31P signals for the acetonitrile solutions of 1b, it
is reasonable to infer that the competitive binding of the
solvent could lead to exchange between bi- and mono-
phosphinite species and that the latter would exist as a
paramagnetic tetrahedral form.
Having established that room-temperature reaction mixtures
contain different Ni−phosphinite species 1a and 1b depending
on the solvent and L:Ni ratio, we proceeded to measure how
easily each of these species undergoes cyclonickelation. Thus,
heating a 1:1:1 L:Ni:Et3N mixture in acetonitrile at 80 °C gave
complete cyclonickelation in 16 h, whereas only 50%
conversion was observed in toluene under analogous
conditions; even heating the toluene mixture to 100 °C gave
only 75% conversion.14 A similar result was obtained when we
measured the cyclonickelation rates in the presence of a 2:1
L:Ni ratio. Indeed, heating 1:1 solutions of 1b:Et3N at 80 °C
for 16 h showed complete cyclonickelation in acetonitrile, but
only 40% in toluene; the latter mixture required heating to ca.
100 °C to achieve ca. 90% conversion.15 It appears, therefore,
that cyclonickelation is faster in acetonitrile likely because of
the reactive monophosphinite intermediate 1a formed in this
solvent (Scheme 1). Conversely, metalation is slower in
toluene because the predominant species in this solvent is the
biphosphinite complex 1b; the latter is less reactive since it
requires the dissociation of one phosphinite to promote the
C−H nickelation. This scenario is also consistent with the
observation that the presence of excess L slows the
cyclonickelation reaction.8
Role and Influence of Base on C−H Nickelation. Our
previous studies had shown that the presence of Et3N is
essential for successful cyclonickelation of aryl phosphinites;
this was attributed to the quenching of the in situ generated
HBr.8,9 However, a more careful analysis of the reaction
mixtures prior to cyclonickelation (i.e., prior to heating) gave
indications that Et3N might hinder the desired reaction. Thus,
although the 1:1 L:Ni mixture in MeCN gives a green solution
that is NMR-silent, addition of 1 equiv of Et3N to this solution
turned it brownish and gave a broad 31P signal (ca. 120−150
ppm) characteristic of biphosphinite complex 1b in acetoni-
trile. We also detected a sharp and minor peak (<2%) on top
of the initial broad peak at ca. 135 ppm, but we were unable to
identify the species responsible for this resonance. Addition of
3 equiv more of Et3N led to a small amount of free phosphinite
(151 ppm), suggesting that competitive binding of the base
would hinder the reaction.
change, and its 31P NMR spectrum displayed only the very
minor sharp peak at 135 ppm, but not the broad signal at ca.
120−150 ppm attributed to 1b.
The above observations imply that the NMR-silent feature
of the 1:1 L:Ni acetonitrile mixture as well as its green color
can be reasonably ascribed to the formation of a paramagnetic
tetrahedral monophosphinite adduct such as 1a. Alternatively,
the absence of NMR signals could also indicate a dynamic
exchange process, for instance involving various mono- and
biphosphinite adducts. Regardless of which interpretation is
more probable, addition of a sufficiently nucleophilic amine
such as Et3N can lead to a competition for binding to Ni, thus
reducing the effective concentration of Ni available for
interaction with the phosphinite. In other words, Et3N leads
to an effective L:Ni ratio greater than 1:1, thereby leading to
the formation of 1b, which is known to hinder the
cyclonickelation reaction. We conclude, therefore, that
although Et3N is necessary for driving the reaction to
completion by capturing the in situ generated HBr, it also
retards the cyclonickelation by promoting the formation of 1b.
The seemingly counteracting roles of Et3N prompted us to
conduct kinetic studies aimed at measuring the effect of
[amine] on cyclonickelation of PhOP(i-Pr)2. The conversion
vs time plots revealed that the C−H nickelation of PhOP(i-
Pr)2 is faster with 1 equiv of Et3N than with 4 equiv (Figure
S38). Plotting ln(1 − conv.) vs time over the first 3 h of the
reaction produced fairly straight lines (Figure 1), revealing that
Figure 1. Plot of ln(1 − conv.) as a function of time and rate
constants under the assumption of a first-order cyclonickelation in
MeCN at 70 °C with various amines and [amine].
the C−H nickelation follows approximately first order rates.
The regression gave rate constants of 0.19 and 0.15 h−1,
respectively, implying a smaller than zeroth order for the base.
This finding is consistent with the aforementioned proposal
that competitive Et3N→Ni coordination can diminish the
concentration of the species responsible for the C−H
nickelation. Moreover, when 1 equiv of base was used, the
reaction was faster with i-Pr2NEt relative to Et3N (0.22 vs 0.19
h−1), which is also consistent with the diminished inhibitory
effect of a more hindered amine on the cyclonickelation.
To further investigate the effect of external base on
cyclonickelation, we studied the reaction of [(i-PrCN)NiBr2]n
with PhOP(i-Pr)2 in the presence of other amines. This
reaction was found to be very sensitive to the specific choice of
external base, strong bases of weak nucleophilic character being
ideal candidates. Thus, little or no reactivity was observed with
amines that are weakly basic (e.g., aniline and N,N-diethylani-
line) or strongly coordinating (e.g., pyridine, imidazole,
To confirm that the observed changes result from the
coordination of the base to Ni, we added 1 equiv of the bulkier
and less nucleophilic amine i-Pr2NEt (Hunig’s base) to a green
̈
1:1 L:Ni mixture in acetonitrile. This mixture showed no color
C
DOI: 10.1021/acs.organomet.8b00899
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
pyrazole, and DBU). In the case of the latter amines,
immediate color changes were observed upon addition of the
amine, confirming the competitive coordination of the amine,
thus hindering the metalation. These findings can be summed
up as follows: external bases can hinder cyclonickelation of aryl
phosphinites if they are strongly coordinating and/or present
in superstoichiometric amounts; nevertheless, they play an
essential role in the fate of cyclonickelation, one that cannot be
fulfilled by excess phosphinite, which is insufficiently basic to
quench HBr.
Reversibility of the Nickelation. It occurred to us that
the requirement for a strong base could be understood if the
nickelated complex was easily protonated. In such a scenario,
protonation of weak bases (including phosphinites) by the in
situ generated HBr would give rise to a strong conjugate acid
capable of protonating the newly formed Ni−C bond, thus
reversing the nickelation step. Conversely, reaction of strong
external bases with HBr would form weak conjugate acids that
would not be effective for protonating the Ni−C moiety; this
would drive the cyclonickelation forward. To determine the
validity of this hypothesis, we tested the stability of the
independently prepared and authenticated complexes [{κP,κC-
C6H4OP(i-Pr)2}Ni(μ-Br)]2 (2-μBr) and 2-NCMe in the
presence of a relatively strong proton source, as described
below in Scheme 2.
but it should be kinetically accessible even in the absence of a
base. To prove this, we sought evidence for H/D scrambling
between a phosphinite and its deuterated analogue under the
conditions of C−H nickelation; if the cyclonickelated species
arising from these phosphinites could survive in the medium
long enough, the ensuing reprotonation by HBr/DBr should
leave a trace of the cyclonickelation event.
The above hypothesis was tested by refluxing a 1:1 mixture
of trans-(C6D5OPR2)2NiBr2 (R = i-Pr, 1b-d5) and trans-(4-Cl-
C6H4OPR2)2NiBr2 (1b-Cl) in d0-acetonitrile over 24 h in the
absence of added base (Scheme 3). The nickel complexes
present in the final reaction mixture were isolated by
1
crystallization and analyzed by H NMR in CDCl3, which
showed H incorporation into the ortho position of the
deuterated ligand but none into meta/para positions.
Surprisingly, however, there was no significant decrease in
1
the H signal for the ortho H in 1b-Cl, indicating that this
compound was not the source of the protons incorporated into
1b-d5. This raised the question of whether the H/D scrambling
occurs with the solvent. In such a scenario, scrambling of
phosphinite ortho positions with CH3CN would occur for both
complexes but would be detectable only in the deuterated
ligand.
To test this second hypothesis, we refluxed the deuterated
complex 1b-d5 alone in CH3CN for 24 h and analyzed the final
product as above, which showed that the ligand had
incorporated 17% H into only the ortho positions (Scheme
Scheme 2. Protonation of the Aryl Moiety by AcOH
2
4). However, H{1H} NMR of the sample in CHCl3 revealed
Scheme 4. Proposed Mechanism for the Role of Solvent in
the Incorporation of H into Deuterio Ligand
Addition of 1 equiv of acetic acid to the room-temperature
acetonitrile solution of 2-NCMe caused a gradual color change
to a deeper red, and the original 31P signal at 196 ppm was
partially replaced over a few hours by a broad signal at ca. 135
ppm; this is the signal for the nonmetalated complexes trans-
L2NiX2 (X = Br, OAc). A similar reactivity was observed when
2 equiv of acetic acid was added to the toluene solution of 2-
μBr at room temperature, but in this case the color change
took place in the time of mixing, and the 31P NMR spectrum
recorded within 10 min confirmed complete disappearance of
the starting material. These observations indicate that the C−
H nickelation of aryl phosphinites is reversible even in the
presence of a relatively weak organic acid. Moreover, tests also
showed that the phosphinite could not be protonated in the
presence of acetic acid or an even stronger acid such as 3-Cl-
C6H4COOH, demonstrating that the conjugate acid [i-
Pr2P(H)OPh]+ must be a sufficiently strong acid to reverse
the cyclonickelation step.
signals corresponding to ortho D in the ligand above natural
abundance. Although the signal was too small to be accurately
quantified, comparison with the natural abundance of CDCl3
in CHCl3 allowed us to evaluate that the extent of D
incorporation is well below 1% (Figure S29). These results can
The above results indicate that the cyclonickelated complex
is thermodynamically unstable toward reprotonation by HBr,
Scheme 3. Observed Ortho H/D Scrambling between Complexes 1b-d5 and 1b-Cl in MeCN
D
DOI: 10.1021/acs.organomet.8b00899
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
be rationalized by invoking H/D exchange between the in situ
generated DBr and the reaction solvent to generate HBr (eq
1). The HBr will then protonate the newly formed Ni−C bond
(eq 2), thus accounting for H incorporation into 1b-d5
(Scheme 4). We should note as well that 13C{1H} NMR
analysis of the final material (Figures S32 and S33) showed
that the i-Pr moieties of the ligand have not incorporated any
D, as we detected no signal for CDMe2 or CHMe (CDH2); this
confirms that the H/D exchange takes place with the solvent.
This sequence of reactions takes place because the H/D
exchange between the in situ generated DBr and CH3CN (eq
1) proceeds with a rate that is similar to the protonation of the
newly formed C−Ni bond by the HBr generated in the
exchange (eq 2). In contrast, the analogous H/D exchange
between 1b and CD3CN is not observed, probably because the
H/D exchange in CD3CN (eq 3) is slower than the
protonation of the C−Ni bond (eq 2).
C−H nickelation is consistent with the previously stated
proposal that the base plays no role in the rate-determining
step.
CH3CN + DBr F CH3CND+ + Br− F
CH2DCNH+ + Br− F CH2DCN + HBr
Figure 2. Plot of ln(1 − conv.) as a function of time for the
cyclonickelation of PhOP(i-Pr)2 with 1 equiv of Ni precursor and 1
equiv of i-Pr2NEt in MeCN at 80 °C.
(1)
(2)
HBr + 2‐L → 1b fast
CD3CN + HBr F CD2HCN + DBr slow
In an effort to shed some light on the possible reaction
pathways, we conducted a Hammett analysis on the cyclo-
nickelation of substituted aryl phosphinites. Monitoring the
cyclonickelation of various meta-substituted aryl phosphinites
3-X-C6H5OP(i-Pr)2 (X = COOMe, Cl, H, Me, OMe) at 80 °C
revealed a strong dependence on the electron-withdrawing
(EW) or electron-donating (ED) character of the substituent
(Figure 3). Indeed, the Hammett plot displays a strongly
(3)
Repeating the above experiments in toluene gave similar
results. Heating 1b-d5 in the absence of a base for 24 h at 100
°C in d0-toluene showed 23% H incorporation into the ortho
position of the ligand under the same conditions (and no D
incorporation into the i-Pr moieties); moreover, 2H{1H} NMR
did not show ortho D incorporation into the ligand, suggesting
this to be below the detection limit (<0.1%). In this case, the
H/D exchange between in situ generated DBr and toluene is
conceivably going through a protonated arene such as a
Wheland intermediate (eq 4).
C6H5CH3 + DBr F [C6H5(D)CH3]+ + Br−
F C6H4DCH3 + HBr
(4)
As was discussed for the analogous reactivity in CD3CN, no
H/D exchange is observed in d8-toluene owing to slow
exchange with this solvent. Although the nickelation is known
to be slower in toluene than in acetonitrile, more H
incorporation has been observed in toluene (23% vs 17%);
this indicates that the exchange of DBr with toluene is faster
than that with acetonitrile and that the rate-limiting step in
acetonitrile is this exchange, and not the nickelation step.
Altogether, the results of the D/H exchange experiments
described above confirm that the C−H nickelation of aryl
phosphinite can take place in the absence of an external base.
Significantly, these observations also imply that capture of the
acid formed in the immediate aftermath of the C−H
nickelation step by the base takes place outside the
coordination sphere.
Kinetics of Aryl Phosphinite Cyclonickelation. The
observations described in the above sections established that
the most favorable conditions for cyclonickelation of aryl
phosphinites are the following: running the reactions in
acetonitrile with the phosphinite ligand, our Ni precursor, and
i-Pr2NEt in a 1:1:1 ratio. Monitoring the reaction under these
conditions at 80 °C and plotting ln(1 − conv.) against time
revealed first-order kinetics for the cyclonickelation up to 90%
conversion, although a somewhat faster rate is evident at the
beginning of the reaction (Figure 2). The observed first-order
Figure 3. Nickelation rates of various 3-X-C6H5OP(i-Pr)2 species in
MeCN at 80 °C.
negative slope (ρ ≈ −4, Figure 4), indicating a much faster C−
H nickelation rate for substrates bearing ED substituents.The
fastest rate, obtained for the cyclonickelation of 3-MeO-
C6H4OP(i-Pr)2, is ca. 20 times greater than the slowest rate,
obtained for the cyclonickelation of 3-COOMe-C6H4OP(i-
Pr)2. Consistent with these results, cyclonickelation of the
highly electron rich substrate 3-Me2N-C6H4OP(i-Pr)2 proved
too fast to be measured conveniently at 80 °C (>70%
conversion within 15 min); indeed, this reaction proceeded to
completion even at room temperature, over 16 h. Finally, the
cyclometalation of sterically hindered 3,5-(MeO)2-C6H5)OP(i-
Pr)2 was found to be faster than the analogous reaction with
the unsubstituted PhOP(i-Pr)2 (Figure S41), demonstrating
that favorable electronic factors can overcome unfavorable
steric effects.
E
DOI: 10.1021/acs.organomet.8b00899
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
so, we set out to isolate Et3N→Ni adducts in the absence of
the phosphinite, as follows.
Adding 2 equiv of Et3N to a green suspension of [(i-
PrCN)NiBr2]n in acetonitrile caused an instantaneous color
change to deep blue and rendered the mixture homogeneous;
addition of Et2O and cooling to −35 °C for a few hours
yielded large royal blue crystals. Exposing these crystals to
ambient air for even a few seconds led to formation of a green
crust on the crystal, presumably hydrates. Crystallographic
analysis of the material proved challenging due to significant
decomposition during crystal mounting and data collection;
nevertheless, the collected data allowed us to unambiguously
identify the main blue species as the homonuclear salt
[Ni(NCMe)6][(Et3N)NiBr3]2 (Scheme 5).
Figure 4. Hammett plot for the nickelation of 3-X-C6H5OP(i-Pr)2
species in MeCN at 80 °C.
Scheme 5. Isolation of Et3N Adducts of Ni
The favorable impact of ED substituents on cyclonickelation
of aryl phosphinites indicates that a positive charge develops
on the Ni center at the transition state, which is stabilized by
the coordination of the C−H moiety. This argues in favor of an
electrophilic C−H metalation. To determine the importance of
the proton transfer on the overall C−H nickelation rate, we
examined the relative reaction rates for the cyclonickelation of
C6H5OP(i-Pr)2 and its deuterio analogue C6D5OP(i-Pr)2. The
reaction with the latter was much slower, and we observed kH/
kD ≈ 11 (Figure S42). The large value for the kinetic isotope
effect suggests that the proton transfer is the rate limiting step
in the mechanism.
To determine if new, acetonitrile-free species could form in
the presence of Et3N, we repeated the above experiment in the
less nucleophilic solvent THF. Adding 4 equiv of Et3N to [(i-
PrCN)NiBr2]n in THF at room temperature also gave a blue
solution that yielded blue crystals upon addition of Et2O and
cooling to −35 °C. Analysis of these crystals revealed the
formation of another homonuclear salt featuring the same
anion, [(Et3N)NiBr3]−, but a different cationic species, namely
the μ-Br bridged dinuclear adduct of THF (Scheme 5 and
Figure 6). It appears, therefore, that formation of the
Next, we studied metalation rates of PhOP(i-Pr)2 between
50 and 80 °C. The linear regression of the Eyring plot (Figure
5) allowed us to extract the idealized kinetic parameters
Figure 5. Eyring plot displaying rates at 50, 60, 70, and 80 °C for the
metalation of PhOP(i-Pr)2 in MeCN.
Figure 6. Molecular diagram of [(THF)6Ni2(μ-Br)3]-[(Et3N)NiBr3].
Thermal ellipsoids are shown at the 50% probability level. Hydrogens
atoms have been omitted for clarity.
according to the transition state theory: ΔH⧧ = 17.7 kcal mol−1
and ΔS⧧ = −27.1 cal mol−1 K−1. The large and negative ΔS⧧
suggests a rate-limiting step with fewer degrees of freedom: i.e.,
a highly ordered transition state. This finding implies that the
pathway from the resting state of the active species to the
transition state for C−H nickelation must involve association
as opposed to dissociation.
Attempts To Isolate Reaction Intermediates in
Acetonitrile. Identification of Et3N→Ni Intermediates. The
aforementioned findings on the formation of the less reactive
biphosphinite complex 1b being favored in the presence of
excess Et3N prompted us to probe how binding of Et3N to Ni
can impact the latter’s reactivity with aryl phosphinites. To do
tetrahedral anion [(Et3N)NiBr3]− is strongly favored and
that this species serves to divert some of the Ni (in a reversible
manner) from the main path of the cyclonickelation reaction.
This finding allows us to propose the scenario depicted in
Scheme 6 for the speciation of Ni in the initial stages of the
cyclonickelation.
Isolation of a Proposed Monophosphinite Intermediate.
In an effort to isolate the cyclonickelation-active mono-
phosphinite species depicted in Scheme 6, we stirred an
equimolar acetonitrile solution of [(i-PrCN)NiBr2]n (ca. 0.1
M) and PhOP(i-Pr)2 and cooled the resulting green mixture to
F
DOI: 10.1021/acs.organomet.8b00899
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
The unanticipated formation of an anionic trihalide species can
be ascribed to the high halide:Ni ratio of 4:1 in this mixture vs
the 2:1 ratio present under the usual conditions of cyclo-
nickelation.
Scheme 6. Postulated Mechanism of Action of Et3N in
Slowing Cyclonickelation in MeCN
In conclusion, our attempts to isolate or detect the putative
cyclonickelation intermediate were unsuccessful. Nevertheless,
the existence of a reactive, monophopsphinite species in
acetonitrile mixtures of the Ni precursor and phosphinites is
supported by (a) isolation of the monophopsphinite
zwitterionic complex illustrated in Scheme 7 and (b) the fact
that this tetrahedral compound has the same color (green) as
the species generated in an equimolar solution of [(i-
PrCN)NiBr2]n and PhOP(i-Pr)2. Further support for the
intermediacy of such species in the cyclonickelation process
was provided by the DFT calculations presented in the
following section.
DFT Computational Analysis. Our DFT calculations
started with model compounds based on the simpler
phosphinite PhOPMe2. This simplification allowed us to
explore a larger range of mechanistic scenarios at a lower
computational cost and eventually eliminate the least realistic
pathways; subsequent DFT studies were then undertaken with
PhOP(i-Pr)2 phosphinites in order to describe our system
more accurately. All energies discussed in the following
sections for local minima and for transition states are
calculated Gibbs free energies.
−35 °C. This gave deep red crystals of the biphosphinite
complex 1b and green-yellow crystals that were identified as
the coordination polymer of the formula [NiBr2(NCMe)2]n.
The failure to isolate the target monophosphinite species is
attributed to the propensity of 1b and [NiBr2(NCMe)2]n to
equilibrate readily and also to their limited solubility in
acetonitrile at lower temperatures.
In a second effort to isolate the putative monophosphinite
adduct, we prepared the parent charge-tagged phosphinite salt
[3-Me3N-C6H4OP(i-Pr)2]X, L′+X− (X = Cl, I) and repeated
the above experiment. To avoid potential complications arising
from the presence of different halides, we treated L′+X− with 2
equiv of Ag(OTs) (Ts = p-Me-C6H4SO2) in order to replace
Cl− and I− anions by the less coordinating tosylate anion. This
led to immediate precipitation of a white solid (presumably
silver salts), which was filtered off. Subsequent addition of 1
equiv of the Ni precursor to the filtrate and stirring at room
temperature led to a green solution that displayed broad 31P
NMR signals at 161 ppm (major) and 135 ppm (minor).
Allowing this green solution to stand at room temperature gave
only colorless crystals that turned out to be the dimeric silver
salt [(L′+Ag)2(μ-Br)3]Br (Scheme 7; see Figure S36 for a
molecular diagram). It is worth emphasizing that the
anticipated Ni compound was not isolated even after cooling
the green filtrate at −35 °C for several days. Evidently, the
greater affinity of the phosphinite for Ag(I) appears to
circumvent the isolation of Ni−phosphinite species.
Ground State. M06 was used to start the optimization of
the tetracoordinated, 16-electron model compound
(PhOPMe2)NiBr2(NCMe), which was allowed to adopt the
1
cis or trans square-planar singlet geometry (1cis-1 and trans-
1) or a tetrahedral triplet geometry (1tet-1). For each of these,
we considered as starting points different rotamers (rotation
around the Ni−P bond) as well as different orientations for the
Ph moiety, pointing away from the Ni center or pointing
toward it. In each of the three geometries considered, several
local minima were found within a range of 1−2 kcal/mol, with
those that featured the Ph group pointing toward the Ni atom
being more stable. Taking into account the more stable
1
rotamer found for each geometry (Figure 7), trans-1 was
3
Finally, we treated the in situ generated L′+X− directly with
1 equiv of [(i-PrCN)NiBr2]n in MeCN at room temperature
and analyzed the green solution by ESI-MS. The result was
most consistent with the formation of (MeCN)n(L′+)NiBr2
species (n = 1, 2).16 Cooling this solution to −35 °C overnight
gave deep green crystals that turned out to be the zwitterionic
monophosphinite adduct {3-Me3N-C6H4OP(i-Pr)2)}NiX3 (X
= Cl, Br, Scheme 7; see Figure S37 for a molecular diagram).
found to be the most stable, followed closely by tet-1 (1.0
kcal/mol higher), and 1cis-1, which was significantly less stable
(4.7 kcal/mol higher).
For the sake of completeness, we also probed the viability of
the pentacoordinated 18-electron species (PhOPMe2)Br2Ni-
(NCMe)2, optimizing it as a square-pyramidal singlet or a
trigonal-bipyramidal (tbp) triplet. The first approach led to
dissociation of the apical acetonitrile, implying that this option
Scheme 7. Species Obtained from Reactions of a Charge-Tagged Ligand with Ni Precursor
G
DOI: 10.1021/acs.organomet.8b00899
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 7. Energies (calculated at the M06 level of theory and
expressed in kcal/mol) for optimized structures of the potential
ground states for the model compound (PhOPMe2)NiBr2(NCMe).
Figure 9. Energies and kinetic barriers (calculated at the M06 level of
theory and expressed in kcal/mol) for the species resulting from C−H
oxidative addition.
The very high energies found for the mechanisms involving
Ni(IV) species thus rule out scenario a.
Next, we considered scenario b. Proceeding with M06 in the
search for a transition state for the concerted σ bond
metathesis always led to a transition state for the ion-pair-
assisted deprotonation, implying that scenario b is also
unlikely. In contrast, the search for an ion-pair-assisted
deprotonation (scenario c) led to proton transfer transition
states for each of the precursor geometries 1trans, 1cis, and 3tet
(Figure 10). The energy barriers found with M06 were 26.4
Figure 8. Energies (calculated at the M06 level of theory and
expressed in kcal/mol) for optimized structures as the potential
ground states for the pentacoordinated model compound
(PhOPMe2)NiBr2(NCMe)2.
is unlikely (Figure 8). The second approach led to two distinct
options featuring equatorial Br ligands and the phosphinite
occupying either an equatorial (3tbp-1-eq) or an axial position
(3tbp-1-ax). Both of these species proved to be less stable (by
1
ca. 7 kcal/mol) with respect to trans-1, implying that the
latter four-coordinated species is the most realistic option.
Optimization of the geometries discussed above with B3LYP
led to different conclusions. The ground state was found to be
3tet-1, whereas trans-1 and cis-1 were found to be more
energetic by 8.3 and 11.6 kcal/mol, respectively. It should be
noted here that the much greater stability observed for the
triplet tetrahedral isomer likely results from B3LYP’s
propensity to artificially stabilize high-spin over low-spin
states.17 Finally, the tbp geometries were disfavored by more
than 11 kcal/mol.
1
1
Mechanistic Pathways. In accordance with our exper-
imental data, we tested three plausible mechanistic scenarios:
(a) C−H oxidative addition followed by H−Br reductive
elimination, (b) concerted σ bond metathesis, and (c) ion-
pair-assisted electrophilic C−H deprotonation/metalation. To
examine the premises of the first scenario, we evaluated the
kinetic accessibility of the Ni(IV) intermediates originating
from a C−H bond oxidative addition. The three octahedral
species shown in Figure 9 were considered as probable
intermediates differing in the relative position of the MeCN
relative to the aryl moiety.
Figure 10. Energies (calculated at the M06 level of theory and
expressed in kcal/mol) of the optimized transition states for the
1
1
3
deprotonation on the cis, trans and tet surfaces as well as the ion
pair on the triplet surface.
kcal/mol for 1TS-trans-dep and 32.2 kcal/mol for 1TS-cis-dep.
These transition states were linked by the intrinsic reaction
coordinate (IRC) path to the nickelated species and HBr
All attempts to minimize 1Oxadd-1-mciseq with M06
converged to its precursor Ni(II) square planar species
1trans-1, suggesting that this species is not viable. Optimization
of the species 1Oxadd-1-mcisax and 1Oxadd-1-mtrans did
converge to local minima, albeit at very high energies of 43.9
and 48.2 kcal/mol, respectively. The transition states for
accessing these intermediates were found quite close in
geometry and displayed barriers of 44.6 and 48.5 kcal/mol,
respectively. Using B3LYP also gave very high energies for
barriers and intermediates (>52 kcal/mol above ground state).
1
1
(respectively trans-2 + HBr and cis-2 + HBr) and to the
nonmetalated complexes (1trans-1 and cis-1, respectively),
1
albeit without finding ion pairs as local minima.
Exploration of the ion-pair-assisted deprotonation scenario
on the triplet surface gave the following results. A transition
state was found for this pathway at a barrier of 45.8 kcal/mol
(3TS-tet-1-dep). The IRC path revealed an ion pair as a local
minimum at 36.1 kcal/mol, which can be considered an
intermediate (3tet-1-Ionpair) for the heterolytic dissociation
H
DOI: 10.1021/acs.organomet.8b00899
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
a
Scheme 8. Proposed Energy Surface for the Nickelation of PhOPMe2
a
Energies were calculated at the M06 level of theory and are expressed in kcal/mol.
3tet-2 + HBr, 31.3 kcal/mol. However, in this case the process
is more endergonic, such that nickelation would not proceed
even in the presence of Et3N. We infer here that the
step (leading to the ion pair). Moreover, no transition state
was found for the heterolytic dissociation of the bromide
3
3
ligand, i.e. tet-1 → tet-1-Ionpair, suggesting a TS very close
to the ion pair. It is worth emphasizing here that the ion pair
on the triplet surface is 5−10 kcal/mol higher in energy than
the two deprotonation barriers on the singlet surfaces.
Exploring scenario b with B3LYP (Scheme S2) gave transition
3
thermodynamics is biased by the overstabilized tet-1.
Reaction Surface with P Substituents i-Pr. Having
completed the analysis of the C−H nickelation pathway with
the PMe2 model, we set out to explore the analogous reaction
surface for the PhOP(i-Pr)2 species. The above results for the
PMe2 model served to guide our study on the P(i-Pr)2 species.
Specifically, we focused on the four-coordinated species only
and considered the ion-pair-assisted deprotonation pathway as
the most viable reaction mechanism. To prevent falling into
local minima defined by various rotamers arising from rotation
of the i-Pr moieties, we studied the different conformations of
these substituents. Among all possible intermediates studied,
the more stable conformations were always those that
minimize the Me−Me steric interactions, as observed in the
crystal structures for {ArOP(i-Pr)2}Ni complexes.
1
states on the triplet surface as high as 34.1 kcal/mol for TS-
1
trans-1-dep and 39.5 kcal/mol for TS-cis-1-dep. The IRC
1
path of the latter provided cis-1-Ionpair as a local minimum
36.5 kcal/mol above the ground state; as was the case with
1
M06, no ion pair was found on the trans pathway. The
deprotonation on the triplet surface going through 3tet-1-
Ionpair (32.2 kcal/mol) was, once again, found prohibitive,
3
with TS-tet-1-dep being as high as 47.2 kcal/mol. Overall,
both M06 and B3LYP favor the ion-pair-assisted deprotona-
1
tion process on the trans pathway.
Thermodynamic Considerations. Scheme 8 summarizes
the M06 analysis of the cyclonickelation reaction in acetonitrile
leading to 2 + HBr. This was found to be an endergonic
process with the following ΔG values (in kcal/mol with respect
to 1trans-1): 1trans-2 + HBr, 15.8; 1cis-2 + HBr, 21.4; 3tet-2 +
HBr, 33.2. The endergonic character of this process echoes the
need for a sufficiently strong base to quench the in situ
generated HBr and driving the C−H nickelation to
completion. Given the Gibbs free energy of 18.1 kcal/mol
for the experimental ΔG for protonation of Et3N by HBr in
acetonitrile,18 we can conclude that this is an effective base for
rendering the overall C−H nickelation process exergonic by
2.3 kcal/mol. This accounts for the feasibility of the nickelation
in the presence of Et3N. In contrast, the value of −9.7 kcal/mol
we obtained by DFT calculation of the ΔG value for
phosphinite protonation by HBr would be insufficient (by
ca. 6 kcal/mol) for rendering the overall nickelation exergonic.
The ΔG values of the nickelation process computed using
B3LYP revealed the same trend as above (in kcal/mol with
As was the case with the simplified PMe2 model, analysis of
the ground state with M06 showed that the cis isomer (1cis-
iPr-1) was higher in energy than the trans isomer (1trans-iPr-
1, Figure 11). However, the triplet tetrahedral species (3tet-
iPr-1) is significantly more stable than the square-planar forms
Figure 11. Optimized structures for the {PhOP(i-Pr)2}NiBr2(NCMe)
isomers and their energies calculated at the M06 level of theory.
Energies are expressed in kcal/mol.
3
1
1
respect to tet-2): trans-2 + HBr, 21.7; cis-2 + HBr, 25.2;
I
DOI: 10.1021/acs.organomet.8b00899
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
a
Scheme 9. Proposed Energy Surface for the Nickelation of PhOP(i-Pr)2 on 1trans and Local Minima on 1cis and 3tet Pathways
a
Energies were calculated at the M06 level of theory and are expressed in kcal/mol.
for the P(i-Pr)2 system, presumably because of the greater
importance of steric demands of the i-Pr substituents. This
result is also more consistent with the fact that the 1:1 L:Ni
mixture is green and NMR-silent, features usually observed for
tetrahedral Ni(II) compounds such as the emerald green
zwitterionic complex {Me3N-PhOP(i-Pr)2}NiX3 (vide supra,
Scheme 7). Ion-pair intermediates were found for each of these
1
geometries, with trans-iPr-1-Ionpair being the most stable,
1
24.7 kcal/mol above the ground state, followed by cis-iPr-1-
3
Ionpair (29.1 kcal/mol) and tet-iPr-1-Ionpair (34.1 kcal/
mol).
Analysis of the nickelation thermodynamics gave the
following ΔG values for the formation of iPr-2 + HBr (in
kcal/mol with respect to 3tet-iPr-2): 1trans-iPr-2 + HBr, 15.2;
1cis-iPr-2 + HBr, 18.1; 3tet-iPr-2 + HBr, 31.1. Scheme 9 gives
a brief analysis of the metalation process for the i-Pr species.
1
1
Thus, conversion of the trans isomer to trans-iPr-1-Ionpair
−
via Br dissociation revealed a transition state at 25.7 kcal/
mol, and the deprotonation from the latter ion pair toward the
2trans-iPr-2 + HBr went through a slightly higher transition
state (26.5 kcal/mol, Figure 12). The ion pairs for the 1cis and
Figure 12. Optimized structures for the P(i-Pr)2 intermediates on the
1trans surface and their energies calculated at the M06 level of theory.
Energies are expressed in kcal/mol.
1
3tet geometries were also found to be higher than TS-trans-
1
iPr-1-dep, suggesting that the surface via trans is the most
probable. Although the strongly negative slope of the Hammett
plot is consistent with this pathway, this finding on its own
does not allow us to determine if the rate-limiting step is the
dissociation or the deprotonation. To settle this question, we
resorted to calculating the KIE values associated with different
steps to see which would give a KIE value closer to the
experimentally determined value.
observed high KIE is indeed the deprotonation and not the
dissociation.
The C−H nickelation process is still endergonic by 15.2
kcal/mol, and the protonation of the ligand PhOP(i-Pr)2 by
HBr in MeCN was computed to be exergonic by only 9.7 kcal/
mol. Therefore, the quenching of HBr by Et3N (−18.1 kcal/
mol) is necessary in order to make the overall reaction
exergonic by 2.9 kcal/mol. The MECP between the singlet
Frequency calculations on the 1trans surface with the
aromatic protons replaced by deuterons allowed us to evaluate
the magnitude of the kinetic isotope effect for both
dissociation and deprotonation barriers. According to the
transition state theory, the kH/kD value at 25 °C was found to
be 1.2 (1.1 at 80 °C) for the heterolytic Ni−Br dissociation
(i.e., the formation of the ion pair) but 8.4 (6.0 at 80 °C) for
the deprotonation. The finding that this last step gives a much
closer value to the experimentally determined kH/kD value of
11 suggests that the rate-limiting step responsible for the
1
pathways involving trans/1cis-iPr-1 and the triplet surface
3
involving tet-iPr-1 was not sought, since the experimental
observations suggest that the system intercrossing would not
be rate limiting in the reaction.
As was the case with the M06 calculations, the metalation
reaction analyzed with B3LYP was found to be endergonic,
with nickelated adducts iPr-2 + HBr reaching 18.4, 21.0, and
27.0 kcal/mol above ground state for 1trans, 1cis, and 3tet. The
computational analysis with B3LYP also showed that the
J
DOI: 10.1021/acs.organomet.8b00899
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
3
under nitrogen, and stored over 4 Å molecular sieves. Triethylamine
was dried over CaH2. Synthesis of the nickel precursor [(i-
PrCN)NiBr2]n used throughout this study has been described
previously.19 All other reagents, including deuterated phenol, were
purchased from Sigma-Aldrich and used without further purification.
All kinetic measurements or studies were conducted by NMR on a
Bruker AV400 instrument equipped with a cryoprobe cooled to liquid
N2 temperature. All other NMR spectra were recorded on regular
Bruker AV400 or AV500 spectrometers. The C−H nickelation
progress was followed by monitoring the 31P NMR signal for {κP,κC-
ArOP(i-Pr)2}Ni(NCMe)Br (resonance around 196 ppm in CH3CN),
using TBAPF6 as the internal standard (septet around −150 ppm).
The substituted aryl phosphinites were synthesized directly from
commercially available phenols and ClP(i-Pr)2 using a reported
protocol.10
ground state is tet-iPr-1, its stability being greater than those
of 1trans-iPr-1 (by 9.7 kcal/mol) and 1cis-iPr-1 (by 10.6 kcal/
mol). The ion pairs for each geometry were shown to follow
the same trend as the starting material, with trans-iPr-1-
Ionpair, cis-iPr-1-Ionpair, and tet-iPr-1-Ionpair found at
33.0, 34.0, and 27.30 kcal/mol above the ground state,
respectively. On the 1trans pathway, the barrier for the
dissociation of the bromide to yield the ion pair reached 31.4
kcal/mol and the proton transfer was also found as the rate-
limiting step with a barrier of 33.4 kcal/mol. The quenching of
HBr by the ligand PhOP(i-Pr)2 turned out to be exergonic by
only 8.5 kcal/mol. Indeed, these calculations showed that even
1
1
3
1
Et3N would barely drive the reaction on the trans pathway,
which reflects the overestimated stability of the high spin
ground state with B3LYP
[3-Me3N-C6H4OH]I. 3-N,N-Dimethylaminophenol (686 mg, 5.0
mmol) was dissolved in 20 mL of CH3CN to give a deep red solution.
MeI (3.11 mL, 50 mmol, 2 equiv) was added slowly and the mixture
stirred at room temperature for 20 h, after which a solid formed. A 30
mL portion of Et2O was added in order to complete precipitation, and
the product was filtered, washed with 2 × 15 mL of Et2O, and dried in
vacuo to yield 1.25 g (4.5 mmol, 90%) of a pale purple powder which
CONCLUSION
■
This report has established that C−H nickelation of aryl
phosphinites is facilitated when this reaction is conducted in
acetonitrile and in the presence of an equimolar quantity of
NEt3 or its less nucleophilic homologue i-Pr2NEt. The
beneficial role of acetonitrile is believed to originate from the
greater nucleophilic character of this solvent (relative to
toluene, THF, or EtOAc), which leads to the stabilization of a
reactive mono(phosphinite) intermediate. In contrast, the
nucleophilic character of a base such as NEt3 hinders the C−H
nickelation by generating various unproductive amine species
such as [(Et3N)NiBr3]− and the biphosphinite species
(ArOPR2)2NiBr2, which is less reactive toward C−H nickel-
ation. DFT calculations have provided support for these
conclusions by showing that tetrahedral (and hence NMR-
silent) monophosphinite Ni(II) complexes are viable species.
Kinetic studies demonstrated that the C−H nickelation
reaction in our system is favored by the presence of electron-
donating substituents on the phosphinite aryl ring, thus
underlining the importance of the latter’s coordination to the
electrophilic Ni(II) center. This conclusion is also consistent
with the importance of directing groups (the phosphinite
moiety in this case) in promoting the C−H nickelation
process. The kinetic studies also revealed a very large KIE
(∼11 at 80 °C), suggesting that C−H bond breaking is rate
limiting. A DFT analysis has suggested that the C−H
nickelation proceeds through an ion-pair mechanism involving
a Ni(II) intermediate.
Another interesting finding touched on the thermodynamics
of the C−H nickelation step: H/D exchange studies in
deuterio and protio solvents confirmed that C−H nickelation
does take place under base-free conditions via an uphill
equilibrium, but the newly formed cyclonickelated species is
unstable in the absence of a sufficiently strong external base
that can quench the in situ generated HBr. The finding that the
metalated intermediate is generated in the absence of Et3N and
can survive long enough (before reprotonation) to allow H/D
scrambling with the solvent implies that such species are
kinetically accessible and should be susceptible to functional-
ization of the Ni−C moiety in the presence of suitable
reagents. This lead will be exploited in our future studies to
develop new ortho-functionalization strategies for phenols.
1
was found to be pure by H NMR in DMSO-d6.
trans-{PhOP(i-Pr)2}2NiBr2 (1b). PhOP(i-Pr)2 (210 mg, 1.0
mmol) was added to a suspension of [(i-PrCN)NiBr2]n (287 mg,
1.0 mmol, 1.0 equiv) in 20 mL of Et2O. The mixture rapidly turned
dark red and was stirred at room temperature for 1 h. The excess Ni
precursor was removed by cannula filtration, and the residues were
extracted twice with 10 mL of Et2O. Evaporation of the ethereal phase
gave 243 mg (0.38 mmol, 76%) of a deep red solid, which was shown
to be pure by 31P and 1H NMR in CDCl3. The spectroscopic data in
CDCl3 matched literature data.9 Analytically pure samples were
obtained by dissolving the crude compound in a minimum amount of
MeCN at room temperature and then cooling to −35 °C for 20 h.
The deep red crystals were washed twice with cold acetonitrile (−35
°C) prior to drying under vacuum.
trans-{C6D5OP(i-Pr)2}2NiBr2 (1b-d5). This compound was
synthesized using the same procedure as described above for 1b,
except for the use of deuterated phenol. The crude powder was
directly purified as described above, yielding deep red crystals (330
1
mg, 0.51 mmol, 51%) . The 31P and H NMR spectra in CDCl3
matched those of the related nondeuterated complex, with an absence
of the aromatic protons.
trans-{4-Cl-C6H4OP(i-Pr)2}2NiBr2 (1b-Cl). This compound was
synthesized using the same procedure described above for 1b, except
for the use of 4-Cl-phenol. The crude product (dark red powder) was
directly purified as described above for 1b to give deep red crystals
(yield 349 mg, 0.59 mmol, 49%). 1H NMR (500 MHz, 20 °C,
3
CDCl3): δ 1.39 (d, 6H, CH(CH3)(CH3), JHH = 6.5 Hz), 1.52 (d,
3
6H, CH(CH3)(CH3), JHH = 6.6 Hz), 2.75 (sept, 2H, CH(CH3)2,
3
3JHH = 7.3 Hz), 7.28 (dm, 2H, Cortho-H or Cmeta-H, JHH = 8.9,
correlated to 13C at δ 129), 7.49 (dm, 2H, Cortho-H or Cmeta-H, 3JHH
=
8.9 Hz, correlated to 13C δ 121). 13C{1H} NMR (125.7 MHz, 20 °C,
CDCl3): δ 18.19 (s, 2C, CH(CH3)(CH3)), 19.52 (s, 2C, CH(CH3)-
(CH3)), 29.76 (s, 2C, CH(CH3)(CH3)), 121.41 (s, 2C, Cortho-H or
C
meta-H), 128.01 (s, 1C, C-Cl), 129.10 (s, 2C, Cortho-H or Cmeta-H),
153.64 (s, 1C, C-OP). 31P{1H} NMR (202.4 MHz, 20 °C, CDCl3): δ
138 (br s).
H/D Scrambling Attempt between 1b-d5 and 1b-Cl. A
Schlenk tube was charged with 100 mg of 1b-Cl, 100 mg of 1b-d5,
and ca. 2 mL of MeCN. The deep red mixture was heated for 24 h at
80 °C and then cooled to −35 °C for 20 h, leading to the formation of
deep red crystals. These were filtered off, washed with 2 × 1.5 mL of
cold MeCN (−35 °C), and then dried under vacuum. The crystals
were analyzed by 1H NMR analysis in CDCl3 and the spectrum
compared to the spectra of approximately equimolar mixtures of
crystalline 1b + 4-Cl-1b and to equimolar mixtures of crystalline 1b-
d5 + 4-Cl-1b. This comparison confirmed the appearance of only the
ortho aromatic protons of 1b and showed no decrease of the
integration for the aromatic protons of 4-Cl-1b.
EXPERIMENTAL SECTION
■
All manipulations were carried out under a nitrogen atmosphere using
standard Schlenk techniques and an inert-atmosphere box. Solvents
were dried by passage over a column of activated alumina, collected
K
DOI: 10.1021/acs.organomet.8b00899
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Protocol for H/D Scrambling with Acetonitrile. A Schlenk
tube was charged with 200 mg of the compound (1b-d5 or 1b) and ca.
2 mL of solvent (CH3CN or CD3CN, respectively). The mixture was
heated at 80 °C for 24 h and then cooled to −35 °C for 20 h to give
crystals, which were isolated, washed twice with 1 mL of cold
acetonitrile, dried under vacuum, and analyzed by 1H NMR in CDCl3.
Protocol for H/D Scrambling with Toluene. A Schlenk tube
was charged with 200 mg of the compound (1b-d5 or 1b) and ca. 2
mL of the solvent (C6H5CH3 or C6D5CD3, respectively). The mixture
was heated at 100 °C for 24 h and the solvent evaporated. The
residues were redissolved in 2 mL of acetonitrile and cooled to −35
°C for 20 h to give crystals, which were isolated, washed twice with 1
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.organo-
met.8b00899.
■
S
NMR spectra, additional plots and figures, and DFT
details (PDF)
Accession Codes
CCDC 1865666−1865668 contain the supplementary crys-
tallographic data for this paper. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by
emailing data_request@ccdc.cam.ac.uk, or by contacting The
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
1
mL of cold acetonitrile, dried under vacuum, and analyzed by H
NMR in CDCl3.
Protocol for Kinetic Measurements. To a solution of ArOP(i-
Pr)2 (200 μmol), [(i-PrCN)NiBr2]n (58 mg, 200 μmol, 1.00 equiv),
and (i-Pr)2NEt (34.8 μL, 200 μmol, 1.00 equiv) in 2.500 mL of
MeCN was added a known amount of the standard TBAPF6 (typically
50−100 mg). A 500 μL aliquot of the solution was placed in a J.
Young NMR tube. This sample was used for the kinetic measure-
ments by one of the following protocols: (i) the NMR tube was
placed in the spectrometer probe and heated to the desired
temperature before recording the NMR spectra over the required
time interval; (ii) the NMR tube was heated in an oil bath at the
desired temperature and for the required time interval before
immersing it into a cold-water bath to cool the sample, recording
the NMR spectra at room temperature, and repeating the heat−cool−
record cycle.
AUTHOR INFORMATION
Corresponding Author
*E-mail for D.Z.: zargarian.davit@umontreal.ca.
ORCID
Davit Zargarian: 0000-0002-0207-7007
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
■
The authors gratefully acknowledge financial support provided
by the NSERC of Canada (Discovery grants to D. Z.), Centre
in Green Chemistry and Catalysis (CGCC/CCVC, summer
For reactions that went to completion within a reasonable time, the
conversion (a number between 0 and 1) at each time t (conv.(t)) was
determined by the ratio of integrals in 31P NMR for the metalated
́
research stipends and travel awards), and Universite de
−
species (AP(t)) and the central peak corresponding to PF6 (AS(t)),
́
Montreal (graduate scholarships to L. P. M.). We also thank
over the same ratio at the end of the reaction (tinf), as follows:
our colleagues from the service laboratories: Dr P. Aguiar (for
his availability and help on setting up the 31P NMR kinetic
measurements), Sylvie Bilodeau (for VT NMR experiments),
AP(t) AP(tinf )
conv.(t) =
AS(t) AS(tinf )
1
2
́
Dr. Cedric Malveau (for H{ H} experiments), and Dr M.
Simard (for help with the resolution of the solid-state structure
for complex {Me3N-PhOP(i-Pr)2}NiBr3). Finally, we are
indebted to Compute Canada-Calcul Canada for access to
Westgrid Computational Facilities, and to Mr J.-P. Cloutier for
discussions on our DFT studies.
When the reaction was too slow to reach completion within a
reasonable time, conv(t) was determined by the amount of TBAPF6
added, where AS,tot is the total area of the peak for the standard (thus
AS,tot = (64/20)AS) and nS,tot the exact amount of TBAPF6 added, in
μmol, as follows:
nP(t)
nP(t)
200 μmol
AP(t)
200 μmol AS,tot(t)
REFERENCES
1
■
conv(t) =
=
=
nS,tot
ns0ubstrate
(1) (a) Song, W.; Ackermann, L. Nickel-catalyzed alkyne annulation
by anilines: versatile indole synthesis by C−H/N−H functionaliza-
tion. Chem. Commun. 2013, 49 (59), 6638−6640. (b) Ruan, Z.;
Lackner, S.; Ackermann, L. A General Strategy for the Nickel-
Catalyzed C−H Alkylation of Anilines. Angew. Chem., Int. Ed. 2016,
55 (9), 3153−3157.
AP(t)
64 nS,tot
AS(t)20
1
=
200 μmol
The first-order rate constants were determined by plotting ln(1 −
conv.) vs the time, giving a slope corresponding to −k, where k is the
rate constant.
(2) (a) Shiota, H.; Ano, Y.; Aihara, Y.; Fukumoto, Y.; Chatani, N.
Nickel-Catalyzed Chelation-Assisted Transformations Involving
Ortho C−H Bond Activation: Regioselective Oxidative Cycloaddition
of Aromatic Amides to Alkynes. J. Am. Chem. Soc. 2011, 133 (38),
14952−14955. (b) Aihara, Y.; Chatani, N. Nickel-Catalyzed Direct
Alkylation of C−H Bonds in Benzamides and Acrylamides with
Functionalized Alkyl Halides via Bidentate-Chelation Assistance. J.
Am. Chem. Soc. 2013, 135 (14), 5308−5311. (c) Aihara, Y.; Chatani,
N. Nickel-Catalyzed Direct Arylation of C(sp3)−H Bonds in
Aliphatic Amides via Bidentate-Chelation Assistance. J. Am. Chem.
Soc. 2014, 136 (3), 898−901. (d) Aihara, Y.; Tobisu, M.; Fukumoto,
Y.; Chatani, N. Ni(II)-Catalyzed Oxidative Coupling between
C(sp2)−H in Benzamides and C(sp3)−H in Toluene Derivatives.
J. Am. Chem. Soc. 2014, 136 (44), 15509−15512. (e) Song, W.;
Lackner, S.; Ackermann, L. Nickel-Catalyzed C-H Alkylations: Direct
Secondary Alkylations and Trifluoroethylations of Arenes. Angew.
Chem., Int. Ed. 2014, 53 (9), 2477−2480. (f) Wu, X.; Zhao, Y.; Ge, H.
Computational Details. The DFT calculations were carried out
in implicit MeCN solvent using the SMD model. The basis sets
employed for all geometry optimizations and frequency calculations
were 6-31g** for light atoms and def2TZVP for Ni and Br, whereas
the electronic energy was computed with def2TZVP for all atoms.
The optimizations started with the M06 functional,20 which is known
to properly handle ligand−metal bonds and give accurate results on
bond dissociation energies.21 For comparison purposes, the local
minima and saddle points were further reoptimized with B3LYP,
which is widely used for various DFT computational analyses.22 High-
resolution representations of 3D models were generated by
CYLview.23 Additional computational details as well as the
coordinates/energies for all intermediates and transition states are
given in the Supporting Information.
L
DOI: 10.1021/acs.organomet.8b00899
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Nickel-Catalyzed Site-Selective Alkylation of Unactivated C(sp3)−H
Bonds. J. Am. Chem. Soc. 2014, 136 (5), 1789−1792. (g) Landge, V.
G.; Shewale, C. H.; Jaiswal, G.; Sahoo, M. K.; Midya, S. P.; Balaraman,
E. Nickel-catalyzed direct alkynylation of C(sp2)−H bonds of amides:
an “inverse Sonogashira strategy” to ortho-alkynylbenzoic acids. Catal.
Sci. Technol. 2016, 6 (6), 1946−1951. (h) Yi, J.; Yang, L.; Xia, C.; Li,
F. Nickel-Catalyzed Alkynylation of a C(sp2)−H Bond Directed by
an 8-Aminoquinoline Moiety. J. Org. Chem. 2015, 80 (12), 6213−
6221. (i) Liu, Y.-J.; Liu, Y.-H.; Yan, S.-Y.; Shi, B.-F. A sustainable and
simple catalytic system for direct alkynylation of C(sp2)−H bonds
with low nickel loadings. Chem. Commun. 2015, 51 (29), 6388−6391.
(j) Yan, S.-Y.; Liu, Y.-J.; Liu, B.; Liu, Y.-H.; Zhang, Z.-Z.; Shi, B.-F.
Nickel-catalyzed direct thiolation of unactivated C(sp3)−H bonds
with disulfides. Chem. Commun. 2015, 51 (34), 7341−7344. (k) Liu,
Y.-H.; Liu, Y.-J.; Yan, S.-Y.; Shi, B.-F. Ni(ii)-catalyzed dehydrative
alkynylation of unactivated (hetero)aryl C−H bonds using oxygen: a
user-friendly approach. Chem. Commun. 2015, 51 (58), 11650−
11653.
Chauvin, R.; Zargarian, D. Nickel(II) complexes of the new pincer-
type unsymmetrical ligands PIMCOP, PIMIOCOP, and NHCCOP:
versatile binding motifs. Chem. Commun. 2012, 48 (84), 10446−
́
10448. (j) Hao, J.; Mougang-Soume, B.; Vabre, B.; Zargarian, D. On
the Stability of a POCsp3OP-Type Pincer Ligand in Nickel(II)
Complexes. Angew. Chem. 2014, 126 (12), 3282−3286. (k) Vabre, B.;
Petiot, P.; Declercq, R.; Zargarian, D. Fluoro and Trifluoromethyl
Derivatives of POCOP-Type Pincer Complexes of Nickel: Prepara-
tion and Reactivities in SN2 Fluorination and Direct Benzylation of
Unactivated Arenes. Organometallics 2014, 33 (19), 5173−5184.
̀
(7) (a) Lefevre, X.; Durieux, G.; Lesturgez, S.; Zargarian, D.
Addition of amines and phenols to acrylonitrile derivatives catalyzed
by the POCOP-type pincer complex [{κP,κC,κP-2,6-(i-Pr2PO)-
2C6H3}Ni(NCMe)][OSO2CF3]. J. Mol. Catal. A: Chem. 2011,
́
335 (1), 1−7. (b) Hao, J.; Vabre, B.; Mougang-Soume, B.; Zargarian,
D. Small Molecule Activation by POCOP-Nickel Complexes. Chem. -
Eur. J. 2014, 20 (39), 12544−12552. (c) Salah, A.; Corpet, M.; ul-
Hassan Khan, N.; Zargarian, D.; Spasyuk, D. M. Synthesis of
unsymmetrical 5,6-POCOP′-type pincer complexes of nickel(ii):
impact of nickelacycle size on structures and spectroscopic properties.
New J. Chem. 2015, 39 (8), 6649−6658.
(3) Zheng, X.-X.; Du, C.; Zhao, X.-M.; Zhu, X.; Suo, J.-F.; Hao, X.-
Q.; Niu, J.-L.; Song, M.-P. Ni(II)-Catalyzed C(sp2)−H Alkynylation/
Annulation with Terminal Alkynes under an Oxygen Atmosphere: A
One-Pot Approach to 3-Methyleneisoindolin-1-one. J. Org. Chem.
2016, 81 (10), 4002−4011.
(8) Vabre, B.; Deschamps, F.; Zargarian, D. Ortho Derivatization of
Phenols through C−H Nickelation: Synthesis, Characterization, and
Reactivities of Ortho-Nickelated Phosphinite Complexes. Organo-
metallics 2014, 33 (22), 6623−6632.
(9) Mangin, L. P.; Zargarian, D. C−H nickelation of phenol-derived
phosphinites: regioselectivity and structures of cyclonickelated
complexes. Dalton Transactions 2017, 46 (46), 16159−16170.
(10) Vabre, B.; Lambert, M. L.; Petit, A.; Ess, D. H.; Zargarian, D.
Nickelation of PCP- and POCOP-Type Pincer Ligands: Kinetics and
Mechanism. Organometallics 2012, 31 (17), 6041−6053.
(11) (a) Baho, N.; Zargarian, D. Diphenyl(dipyrazolyl)methane
Complexes of Ni: Synthesis, Structural Characterization, and
Chromotropism of NiBr2 Derivatives. Inorg. Chem. 2007, 46 (18),
7621−7632. (b) Lapointe, S.; Vabre, B.; Zargarian, D. POCOP-Type
Pincer Complexes of Nickel: Synthesis, Characterization, and Ligand
Exchange Reactivities of New Cationic Acetonitrile Adducts.
(4) See the following report for a rare mechanistic study on this
subject: Beattie, D. D.; Grunwald, A. C.; Perse, T.; Schafer, L. L.;
Love, J. A. Understanding Ni(II)-Mediated C(sp3)−H Activation:
Tertiary Ureas as Model Substrates. J. Am. Chem. Soc. 2018, 140 (39),
12602−12610.
(5) (a) Fontaine, F.-G.; Kadkhodazadeh, T.; Zargarian, D. Nickel
indenyl complexes as precatalysts for dehydropolymerization of
phenylsilane. Chem. Commun. 1998, No. 12, 1253−1254.
́
́
(b) Groux, L. F.; Belanger-Gariepy, F.; Zargarian, D.; Vollmerhaus,
R. Preparation and Characterization of Nickel Complexes with η-
Indenyl Ligands Bearing a Pendant Aminoalkyl Chain. Organo-
́
metallics 2000, 19 (8), 1507−1513. (c) Groux, L. F.; Belanger-
́
Gariepy, F.; Zargarian, D. Phosphino-indenyl complexes of nickel(II).
Can. J. Chem. 2005, 83 (6−7), 634−639. (d) Chen, Y.; Sui-Seng, C.;
Zargarian, D. Tetraphenylborate as a Novel Bridging Ligand in a
Zwitterionic Nickel(I) Dimer. Angew. Chem., Int. Ed. 2005, 44 (47),
7721−7725. (e) Baho, N.; Zargarian, D. Syntheses, Structures,
Spectroscopy, and Chromotropism of New Complexes Arising from
the Reaction of Nickel(II) Nitrate with Diphenyl(dipyrazolyl)-
methane. Inorg. Chem. 2007, 46 (1), 299−308.
̀
Organometallics 2015, 34 (14), 3520−3531. (c) Lefevre, X.;
Spasyuk, D. M.; Zargarian, D. J. Organomet. Chem. 2011, 696, 864−
870. (d) Wang, R.; Groux, L. F.; Zargarian, D. Nickel-Triflate
Complexes as Precursors to Reactive Cations: Preparation and
Reactivities of (1-R-indenyl)Ni(PPh3)(OSO2CF3). Organometallics
2002, 21 (25), 5531−5539. (e) Spasyuk, D. M.; Gorelsky, S. I.; van
der Est, A.; Zargarian, D. Characterization of Divalent and Trivalent
Species Generated in the Chemical and Electrochemical Oxidation of
a Dimeric Pincer Complex of Nickel. Inorg. Chem. 2011, 50, 2661−
2674. (f) Groux, L. F.; Zargarian, D. Aminoalkyl-Substituted
Indenyl−Nickel Compounds: Tuning Reactivities as a Function of
the Pendant, Hemilabile Moiety. Organometallics 2003, 22 (15),
3124−3133.
(6) For examples of Ni complexes based on various symmetrical or
unsymmetrical pincer ligands see the following review and primary
reports. (a) Haenel, M. W.; Jakubik, D.; Kruger, C.; Betz, P. 1,8-
̈
Bis(diphenylphosphino)anthracene and Metal Complexes1). Chem.
́
Ber. 1991, 124 (2), 333−336. (b) Gomez-Benítez, V.; Baldovino-
́
́
Pantaleon, O.; Herrera-Alvarez, C.; Toscano, R. A.; Morales-Morales,
D. High yield thiolation of iodobenzene catalyzed by the phosphinite
nickel PCP pincer complex: [NiCl{C6H3−2,6-(OPPh2)2}]. Tetrahe-
dron Lett. 2006, 47 (29), 5059−5062. (c) Gutsulyak, D. V.; Piers, W.
E.; Borau-Garcia, J.; Parvez, M. Activation of Water, Ammonia, and
Other Small Molecules by PCcarbeneP Nickel Pincer Complexes. J.
Am. Chem. Soc. 2013, 135 (32), 11776−11779. (d) Jonasson, K. J.;
Wendt, O. F. Synthesis and Characterization of a Family of POCOP
Pincer Complexes with Nickel: Reactivity Towards CO2 and
Phenylacetylene. Chem. - Eur. J. 2014, 20 (37), 11894−11902.
(e) Jonasson, K. J.; Wendt, O. F. Synthesis and characterisation of
new PCsp3P-supported nickel complexes. J. Organomet. Chem. 2014,
759, 15−18. (f) Mousa, A. H.; Bendix, J.; Wendt, O. F. Synthesis,
Characterization, and Reactivity of PCN Pincer Nickel Complexes.
Organometallics 2018, 37 (15), 2581−2593. (g) Zargarian, D.;
Castonguay, A.; Spasyuk, D. M., ECE-Type Pincer Complexes of
Nickel. In Organometallic Pincer Chemistry; van Koten, G., Milstein,
D., Eds.; Springer Berlin Heidelberg: Berlin, Heidelberg, 2013; pp
131−173. (h) Castonguay, A.; Beauchamp, A. L.; Zargarian, D. New
Derivatives of PCP-Type Pincer Complexes of Nickel. Inorg. Chem.
2009, 48 (7), 3177−3184. (i) Vabre, B.; Canac, Y.; Duhayon, C.;
(12) This complex has been shown to adopt a square -planar
structure in the solid state. See ref 9 for its preparation, isolation, and
solid-state characterization.
(13) It should be noted that EtOAc and THF mixtures showed the
same colors as the toluene mixtures; to simplify the discussion, we
have focused on toluene mixtures.
(14) The rates of these reactions were monitored by observing the
emergence of 31P signals characteristic of the cyclonickelated
complexes: a singlet at 196 ppm in acetonitrile for (κP,κC-i-
Pr2POC6H4)Ni(NCMe)Br and AB doublets at 184 and 151 ppm in
toluene for (κP,κC-i-Pr2POC6H4)Ni(i-Pr2POPh)Br. Complete con-
version was noted by the complete disappearance of the broad signals
due to the nonmetalated phosphinite adducts, whereas partial
conversions were estimated by integration of the signals for the
metalated and nonmetalated species.
(15) Since toluene samples of both the starting material trans-
L2NiBr2 and the cyclometalated product (κP,κC-i-Pr2POC6H4)Ni(i-
Pr2POPh)Br display sharp 31P signals, the extent of cyclometalation in
toluene mixtures could be determined in a straightforward manner on
M
DOI: 10.1021/acs.organomet.8b00899
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
the basis of NMR. In the case of the reactions conducted in
acetonitrile, however, determination of conversion is complicated by
the fact that the 31P NMR spectrum of the reaction mixture displays
no signals due to the ligand substitution reaction between the
cyclometalated complex and the solvent. In this case, conversions
were determined by evaporating the final mixture, dissolving the solid
residues in toluene, and recording the 31P NMR of this sample.
(16) It should be noted that the experimental m/z values obtained
from this analysis correspond to “M+ − 1” for the parent ion with one
Ni-bound MeCN, M+ = [C17H30Br2N2NiOP]+ = 526.98, and to “M+
+ 2” for the parent ion with two Ni-bound MeCN, M+
=
[C19H33Br2N3NiOP]+ = 568.01. Of course, we are mindful of the
fact that the highly dilute solutions required for ESI-MS experiments
would be expected to affect the equilibria governing the speciation of
Ni, which might cast doubt on the relevance of this result for the
overall reaction mechanism under realistic reaction conditions. This
result is nevertheless significant, because it supports the viability of a
monophosphinite adduct in MeCN.
(17) (a) Paulsen, H.; Duelund, L.; Winkler, H.; Toftlund, H.;
Trautwein, A. X. Free Energy of Spin-Crossover Complexes
Calculated with Density Functional Methods. Inorg. Chem. 2001, 40
(9), 2201−2203. (b) Reiher, M.; Salomon, O.; Artur Hess, B.
Reparameterization of hybrid functionals based on energy differences
of states of different multiplicity. Theor. Chem. Acc. 2001, 107 (1),
48−55. (c) Reiher, M. Theoretical Study of the Fe(phen)2(NCS)2
Spin-Crossover Complex with Reparametrized Density Functionals.
Inorg. Chem. 2002, 41 (25), 6928−6935. (d) Conradie, J.; Ghosh, A.
DFT Calculations on the Spin-Crossover Complex Fe(salen)(NO):
A Quest for the Best Functional. J. Phys. Chem. B 2007, 111 (44),
12621−12624.
(18) See calculations on page S23 of the Supporting Information,
which were based on these references: (a) Raamat, E.; Kaupmees, K.;
Ovsjannikov, G.; Trummal, A.; Kutt, A.; Saame, J.; Koppel, I.;
̈
Kaljurand, I.; Lipping, L.; Rodima, T.; Pihl, V.; Koppel, I. A.; Leito, I.
Acidities of strong neutral Brønsted acids in different media. J. Phys.
̈
Org. Chem. 2013, 26 (2), 162−170. (b) Kaljurand, I.; Kutt, A.;
̈
̈
Soovali, L.; Rodima, T.; Maemets, V.; Leito, I.; Koppel, I. A.
Extension of the Self-Consistent Spectrophotometric Basicity Scale in
Acetonitrile to a Full Span of 28 pKa Units: Unification of Different
Basicity Scales. J. Org. Chem. 2005, 70 (3), 1019−1028.
(19) (a) Vabre, B.; Spasyuk, D. M.; Zargarian, D. Impact of
Backbone Substituents on POCOP-Ni Pincer complexes: A
Structural, Spectroscopic and Electrochemical Study. Organometallics
2012, 31 (24), 8561−8570. (b) Vabre, B.; Lindeperg, F.; Zargarian,
D. Direct, one-pot synthesis of POCOP-type pincer complexes from
metallic nickel. Green Chem. 2013, 15 (11), 3188−3194.
(20) Zhao, Y.; Truhlar, D. G. The M06 suite of density functionals
for main group thermochemistry, thermochemical kinetics, non-
covalent interactions, excited states, and transition elements: two new
functionals and systematic testing of four M06-class functionals and
12 other functionals. Theor. Chem. Acc. 2008, 120 (1), 215−241.
(21) Raju, R. K.; Bengali, A. A.; Brothers, E. N. A unified set of
experimental organometallic data used to evaluate modern theoretical
methods. Dalton Transactions 2016, 45 (35), 13766−13778.
(22) (a) Lee, C.; Yang, W.; Parr, R. G. Development of the Colle-
Salvetti correlation-energy formula into a functional of the electron
density. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37 (2), 785−
789. (b) Becke, A. D. Density-functional thermochemistry. III. The
role of exact exchange. J. Chem. Phys. 1993, 98 (7), 5648−5652.
́
(23) Legault, C. Y. CYLview, 1.0b; Universite de Sherbrooke, 2009;
http://www.cylview.org.
N
DOI: 10.1021/acs.organomet.8b00899
Organometallics XXXX, XXX, XXX−XXX