Understanding the Early Stages of Nickel Sulfide Nanocluster Growth: Isolation of Ni3, Ni4, Ni5, and Ni8 Intermediates

Article abstract of DOI:10.1002/smll.202003133

Addition of sub-stoichiometric quantities of PEt₃ and diphenyl disulfide to a solution of [Ni(1,5-cod)₂] generates a mixture of [Ni₃(SPh)₄(PEt₃)₃] (1), unreacted [Ni(1,5-cod)₂], and [(1,5-cod)Ni(PEt₃)₂], according to ¹H and ³¹P{¹H} NMR spectroscopic monitoring of the in situ reaction mixture. On standing, complex 1 converts into [Ni₄(S)(Ph)(SPh)₃(PEt₃)₃] (2), via formal addition of a “Ni(0)” equivalent, coupled with a C-S oxidative addition step, which simultaneously generates the Ni-bound phenyl ligand and the μ₃-sulfide ligand. Upon gentle heating, complex 2 converts into a mixture of [Ni₅(S)₂(SPh)₂(PEt₃)₅] (3) and [Ni₈(S)₅(PEt₃)₇] (4), via further addition of “Ni(0)” equivalents, in combination with a series of C–S oxidative addition and C-C?reductive elimination steps, which serve to convert thiophenolate ligands into sulfide ligands and biphenyl. The presence of 1–4 in the reaction mixture is confirmed by their independent syntheses and subsequent spectroscopic characterization. Overall, this work provides an unprecedented level of detail of the early stages of Ni nanocluster growth and highlights the fundamental reaction steps (i.e., metal atom addition, C-S oxidative addition, and C-C reductive elimination) that are required to grow an individual cluster.

Full text of DOI:10.1002/smll.202003133

Full PaPer
www.small-journal.com
Understanding the Early Stages of Nickel Sulfide
Nanocluster Growth: Isolation of Ni3, Ni4, Ni5,
and Ni8 Intermediates
Alexander J. Touchton, Guang Wu, and Trevor W. Hayton*
In contrast, when this synthetic strategy
is applied to other transition metals, such
as Co and Ni, only the products of salt
metathesis, i.e., small M(II) thiolate clus-
ters, are isolated[6,12–14]—the metal reduc-
tion step is not observed. This result is not
entirely surprising, given that it is much
harder to reduce Co2+ and Ni2+ than it
is to reduce the group 11 M+ cations.[15]
Access to these materials is of interest
because Co- and Ni-thiolate nanoclusters
are anticipated to be paramagnetic, unlike
group 11 nanoclusters, which are typically
diamagnetic. Thus, they could be useful
for a variety of magnetism applications,
including ferrofluids and quantum com-
puting.[16–18] Additionally, their properties
can be easily tuned by changing the thi-
olate substituent, in contrast to previously
reported sulfide-capped nanoclusters.[19,20]
In an effort to solve the metal reduction
problem, several groups have employed
the use of M(0)-containing starting mate-
rials, instead of the traditional M(II)
Addition of sub-stoichiometric quantities of PEt3 and diphenyl disulfide
to a solution of [Ni(1,5-cod)2] generates a mixture of [Ni3(SPh)4(PEt3)3] (1),
unreacted [Ni(1,5-cod)2], and [(1,5-cod)Ni(PEt3)2], according to 1H and
31P{1H} NMR spectroscopic monitoring of the in situ reaction mixture. On
standing, complex 1 converts into [Ni4(S)(Ph)(SPh)3(PEt3)3] (2), via formal

addition of a “Ni(0)” equivalent, coupled with a C S oxidative addition
step, which simultaneously generates the Ni-bound phenyl ligand and the
μ3-sulfide ligand. Upon gentle heating, complex 2 converts into a mixture
of [Ni5(S)2(SPh)2(PEt3)5] (3) and [Ni8(S)5(PEt3)7] (4), via further addition of
“Ni(0)” equivalents, in combination with a series of C–S oxidative addition

and C C reductive elimination steps, which serve to convert thiophenolate
ligands into sulfide ligands and biphenyl. The presence of 1–4 in the
reaction mixture is confirmed by their independent syntheses and
subsequent spectroscopic characterization. Overall, this work provides an
unprecedented level of detail of the early stages of Ni nanocluster growth
and highlights the fundamental reaction steps (i.e., metal atom addition,


C S oxidative addition, and C C reductive elimination) that are required
to grow an individual cluster.
salts.[21–23] For example, we recently used [NiII(η3-CPh3)2] as a
source of Ni(0) to generate the low-valent phosphinidene cluster,
[Ni3(PPh)(PPh2)2(PPh3)3].[24] Similarly, Roy and co-workers
recently employed the commercially available Ni(0) synthon,
[Ni(1,5-cod)2], to generate a series of low-valent Ni phosphi-
nidene clusters.[25] Steigerwald and co-workers have also
employed [Ni(1,5-cod)2] as a source of Ni(0) to access low-valent,
chalcogenide-capped Ni nanoclusters.[26–28]
In this work, we sought to evaluate whether [Ni(1,5-cod)2]
could be used to access low-valent [Nix(SR)y]-type (x > y) clus-
ters, a class of materials which remains unknown despite sev-
eral attempts at their isolation.[6,13–14] To access these materials
we envisioned that addition of sub-stoichiometric quantities of
diphenyl disulfide (PhSSPh), which can act as a thiolate source
1. Introduction
The last 10 years have seen significant progress in the synthesis
of thiolate-stabilized group 11 nanoclusters (NCs).[1–6] These
advancements have been driven, in part, by the deployment of
the single phase Brust–Schiffrin synthetic protocol, which has
allowed for the rapid diversification of coinage metal NCs,[4] such
as [Au102(p-MBA)44] (p-MBA = para-mercaptobenzoic acid),[7]
[Ag25(S-2,4-Me2C6H3)18]−,[8] and [Cu61(S)6(StBu)26Cl6H14]+.[9]
In a typical synthesis, small [M(SR)]n oligomers are generated
in the early stages of the reaction by combining a metal halide
precursor with excess HSR (R = alkyl, aryl). These oligomers are
then reduced by NaBH4 to generate low-valent [Mx(SR)y]-type
(x > y) clusters, which feature at least some M(0) character.[10,11]

upon S S bond cleavage, to [Ni(1,5-cod)2], in the presence of
a phosphine capping ligand, would generate the desired low-
valent, thiolate-protected Ni nanoclusters.
A. J. Touchton, Dr. G. Wu, Prof. T. W. Hayton
Department of Chemistry and Biochemistry
University of California
Santa Barbara, CA 93106, USA
E-mail: hayton@chem.ucsb.edu
Herein, we report the conversion of [Ni(1,5-cod)2] into the
low-valent Ni thiolate cluster, [Ni3(SPh)4(PEt3)3] (1), which
subsequently converts into a mixture of the low-valent Ni
sulfide-containing nanoclusters, [Ni5(S)2(SPh)2(PEt3)5] (3) and
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/smll.202003133.

[Ni8(S)5(PEt3)7] (4), via a series of Ni(0) addition, C S oxida-

tive addition, and C C reductive elimination steps. In addi-
tion, the isolation of the organometallic phenyl-containing Ni4
DOI: 10.1002/smll.202003133
2003133 (1 of 8)
© 2020 Wiley-VCH GmbH
Small 2020, 2003133

www.advancedsciencenews.com
www.small-journal.com
Figure 1. 31P{1H} NMR spectroscopic monitoring of the reaction of [Ni(1,5-cod)2] (8 equiv) with PEt3 (7 equiv) and PhSSPh (2.5 equiv) in toluene-d8
with Ph3PO added as internal standard.
intermediate, [Ni4(S)(Ph)(SPh)3(PEt3)3] (2), from the reaction
mixture conclusively demonstrates that the sulfide ligands in
assignable to 2 and 5 grow in intensity (Figure 1b and Table S2,
Supporting Information). On standing overnight at room tem-
perature, the resonance assignable to 1 disappears, the reso-
nance assignable to [(1,5-cod)Ni(PEt3)2] decreases in intensity,
while resonances assignable to 2 and 5 continue to grow in inten-
sity (Figure 1c). This spectrum also features minor resonances
at 4.7 and 13.4 ppm, which appear in a 1:4 ratio and are assign-
able to the sulfide thiolate cluster, [Ni5(S)2(SPh)2(PEt3)5] (3).
After heating the sample to 40 °C for 1 h, the resonances
assignable to 2 decrease in relative intensity, while those
assignable to 3 increase in intensity (Figure 1d). Further
heating of the sample at 40 °C for a total of 43 h results in a
disappearance of the resonances assignable to 2, while those
assignable to 3 increase in intensity (Figure 1g). Also pre-
sent in this spectrum are broad resonances at −4.8, 12.8,
and 14.6 ppm. These resonances appear in a 4:2:1 ratio and
are assignable to the Ni8 sulfide nanocluster, [Ni8(S)5(PEt3)7]
(4). Finally, further heating of the solution to 60 °C for
28 h results in the decrease of resonances assignable to
3 and 5, while those assignable to 4 continue to increase in
intensity (Figure 1h). Overall, these data are consistent with
a reaction sequence in which complex 1 is generated initially,
but then converts into 2, which then converts into 3 and then 4
upon heating.

3 and 4 form via C S oxidative addition. The isolation of these
intermediates provides an unprecedented level of detail of Ni
sulfide nanocluster growth, which may enable the targeted syn-
thesis of larger Ni sulfide clusters.
2. Results and Discussion
2.1. 31P NMR Spectroscopic Monitoring of the Reaction
of [Ni(1,5-cod)2] with PEt3 and PhSSPh
Addition of 2.5 equiv of diphenyl disulfide to a suspension
of 8 equiv of [Ni(1,5-cod)2] and 7 equiv of PEt3 in toluene-d8,
results in an immediate color change to dark brown.[29] The
31P{1H} NMR spectrum of this mixture features a sharp reso-
nance at 17.6 ppm assignable to [(1,5-cod)Ni(PEt3)2] (Figure 1a,
and Table S2, Supporting Information),[30] as well as a broad
resonance at 27 ppm, which we have assigned to the trime-
tallic thiolate cluster, [Ni3(SPh)4(PEt3)3] (1). Also present in
this spectrum are minor resonances at 16.8 (t, JPP = 33.8 Hz)
and 10.1 ppm (br d, JPP = 28.4 Hz), which we have assigned to
the sulfide-containing Ni4 cluster, [Ni4(S)(Ph)(SPh)3(PEt3)3] (2).
These resonances are present in a 1:2 ratio, respectively. This
spectrum also features a minor peak at 9.0 ppm, which we have
tentatively assigned to the square planar Ni(II) phenyl complex,
[Ni(Ph)(SPh)(PEt3)2] (5) (see below for more discussion). Upon
standing for 2 h at room temperature, the broad resonance
assignable to 1 decreases in intensity, while the resonances
The in situ 1H NMR spectra and electrospray ionization
mass spectrometry (ESI-MS) data of the reaction mixture are
broadly consistent with the speciation data revealed by our
31P{1H} NMR spectra (Figure 1). These data are discussed in
further detail in the Supporting Information. Additionally, the
1
in situ H NMR spectra reveal the formation of biphenyl as a
Small 2020, 2003133
2003133 (2 of 8)
© 2020 Wiley-VCH GmbH

www.advancedsciencenews.com
www.small-journal.com
Scheme 1. Synthesis of 1.
major by-product of the reaction (see below for more discus-
sion of this observation).
2.2. Independent Synthesis and Characterization
of Complexes 1, 2, 3, and 4
To confirm the formulation of complex 1, we endeavored to
independently prepare and isolate this species. In this regard,
complex 1 could be successfully prepared by addition of 2 equiv
of PhSSPh to a suspension of [Ni(1,5-cod)2] (3 equiv) and PEt3
(3 equiv) in cold diethyl ether (−25 °C). Work-up of the reaction
mixture furnished [Ni3(SPh)4(PEt3)3] (1), which could be iso-
lated in 62% yield as a deep brown crystalline solid (Scheme 1).
Its room temperature 31P{1H} NMR spectrum in toluene-d8
features a broad singlet at 32 ppm, consistent with the results
from the in situ spectroscopic monitoring study (Figure S5,
Supporting Information, and Figure 1). On cooling the NMR
sample to −79 °C, however, this resonance decoalesces into
two peaks at 2 ppm and −7 ppm, which are assignable to
two magnetically inequivalent PEt3 environments in a 2:1
ratio (Figure S8, Supporting Information), consistent with
the solid-state molecular structure (see below). Its room tem-
Figure 2. Oak ridge thermal ellipsoid plot (ORTEP) diagram of 1 with
thermal ellipsoids plotted at 50%. All hydrogen atoms omitted for
clarity. Carbon atoms shown as wireframe. Selected bond lengths
(Å): Ni1–Ni2 = 2.6481(6), Ni1–Ni3 = 2.6699(6), Ni2–Ni3 = 2.5100(6).
in Et2O at room temperature. Work-up of the reaction mixture
after 18 h, resulted in the isolation of 2 in 15% yield as a dark
brown crystalline solid (Scheme 2). The room temperature
31P{1H} NMR spectrum of 2 in benzene-d6 features a triplet
at 17.1 ppm (JPP = 33.8 Hz) and a broad doublet at 10.3 ppm
(JPP = 30.4 Hz) in a 1:2 ratio (Figure S12, Supporting Informa-
tion). Importantly, these resonances confirm the presence of 2
1
in the in situ monitoring experiment (Figure 1). The H NMR
1
perature H NMR spectrum in benzene-d6 also supports its
spectrum of 2 in benzene-d6 at room temperature features a
doublet at 8.08 ppm, which is assignable to ortho-CH environ-
ment of the Ni-bound phenyl ligand. Also present in this spec-
trum are resonances at 7.96 (4H) and 7.58 ppm (2H), which are
assignable to the ortho-CH environments of two magnetically
inequivalent thiophenolate ligands (Figure S11, Supporting
Information). Finally, the ESI-MS spectrum of 2 in THF, col-
lected in positive mode, features an intense peak at 1065.20 m/z,
assignable to [M − Ph + PEt3]+, as well as smaller peaks
at 947.10 m/z and 988.14 m/z, which are ascribable to [M − Ph]+
(calcd 947.01 m/z) and [M − 2Ph + PEt3]+ (calcd 988.06 m/z),
respectively (Figures S27–S32, Supporting Information). The
failure to observe [M]+ in this spectrum is consistent with the
expected higher reactivity of an organometallic Ni species.
presence in the in situ spectroscopic monitoring experiment
(Figures S2–S4, Supporting Information). The ESI-MS of 1
in tetrahydrofuran (THF), acquired in positive ion mode, fur-
ther supports the proposed formulation. Specifically, peaks
at 966.21 m/z and 857.19 m/z are ascribable to [M]+ (calcd
966.12 m/z) and [M − SPh]+ (calcd 857.11 m/z), respectively
(Figures S23–S26, Supporting Information).
Complex 1 crystallizes in the monoclinic space group P21/n
and its solid-state molecular structure is shown in Figure 2.
It features a triangular arrangement of nickel atoms with two
unique thiophenolate environments and two unique PEt3 envi-
1
ronments, consistent with its low temperature H and 31P{1H}

NMR spectra. The Ni Ni bond lengths are 2.5100(6), 2.6699(6),
and 2.6481(6) Å, which are consistent with those expected for

Ni Ni single bonds (formal shortness ratio (FSR) = 1.01, 1.07,
and 1.06, respectively).[31,32] For further comparison, the Ni Ni

distances in the Ni(II) thiolate cluster, [Ni(SCH2CH2Ph)2]6, are
much longer (2.84–2.94 Å).[13] Complex 1 features an average

oxidation state of +1.33, but given the Ni Ni bonding and
the observed diamagnetism, the Ni valence electrons in 1 must
be delocalized over all three Ni centers.
We also attempted to independently generate complex 2.
This species could be specifically targeted by reaction of 4 equiv
of [Ni(1,5-cod)2] with 3 equiv of PEt3 and 2 equiv of PhSSPh
Scheme 2. Synthesis of 2.
2003133 (3 of 8)
© 2020 Wiley-VCH GmbH
Small 2020, 2003133

www.advancedsciencenews.com
www.small-journal.com
Figure 3. ORTEP diagram of 2⋅0.5C6H14 with thermal ellipsoids plotted at
50% probability. Hydrogen atoms and hexane solvate molecule omitted
for clarity. Carbon atoms shown as wireframe. Only one set of ethyl groups
are displayed for the disordered ethyl groups of P1 and P2. Selected bond
lengths (Å): Ni1–Ni4 =2.5314(8), Ni2–Ni4 =2.5572(9), Ni1–Ni2 =2.5642(8),
Ni3–Ni4 = 2.6492(8), Ni2–Ni3 = 2.6853(9), Ni1–Ni3 = 2.7022(7).
Figure 4. ORTEP diagram of 3 with thermal ellipsoids displayed at
50% probability. Hydrogen atoms are omitted for clarity. Selected dis-
tances (Å): Ni5–Ni1 = 2.645(7), Ni5–Ni2 = 2.655(8), Ni5–Ni4 = 2.663(7),
Crystals of 2 suitable for single-crystal X-ray diffraction
were grown from a hexanes solution of 2 stored at −25 °C for
two months. Compound 2 crystallizes in the triclinic space
group P-1 as the hexane solvate, 2⋅0.5C6H14 (Figure 3). Impor-
tantly, the solid-state structure of 2⋅0.5C6H14 reveals the pres-
ence of the Ni-bound phenyl moiety, as well as a μ3-sulfide
Ni5–Ni3
= 2.623(8), Ni1–Ni2 = 2.786(8), Ni3–Ni4 = 2.789(8),
Ni3–Ni1 = 2.994(7), Ni4–Ni2 = 2.983(8).
(Scheme 3). A 31P{1H} NMR spectral analysis of this mixture
revealed the presence of 3 and 4 in an approximate 1:2 ratio
(Figure S22, Supporting Information). Because of their similar
solubilities, we could not separate 3 and 4 by selective crystal-
lization, and so we characterized them as a mixture.

ligand. Evidently, these ligands are derived from a C S oxida-
tive addition across a low-valent Ni center. Complex 2 features
a distorted tetrahedral arrangement of its four Ni atoms. The

Ni Ni distances between Ni1, Ni4, and Ni2 are relatively short
Crystals of 3 suitable for X-ray crystallography were grown
from a concentrated pentane solution of a mixture of 3, 4, and
5 left to slowly evaporate at −25 °C for two months. Complex
3 crystalizes in the monoclinic space group P21/c (Figure 4).
It features a square pyramidal arrangement of its five Ni atoms
with overall Cs symmetry. Each Ni atom is ligated by one PEt3
ligand. In addition, 3 possesses two μ3-sulfide ligands and two
μ2-thiophenolate ligands. One phenyl group exhibits axial stere-
ochemistry, while the other exhibits equatorial stereochemistry.
(Ni1–Ni4 = 2.5314(8), Ni2–Ni4 = 2.5572(9), Ni1–Ni2 = 2.5642(8) Å;
FSR = 1.02, 1.03, and 1.03, respectively), whereas those from Ni1,
Ni2, and Ni4 to Ni3 are slightly longer (Ni3–Ni4 = 2.6492(8),
Ni2–Ni3 = 2.6853(9), Ni1–Ni3 = 2.7022(7) Å; FSR = 1.06, 1.08,
and 1.09, respectively). Finally, the Ni1–C1 distance (1.887(4) Å)
is comparable to those found for other low-valent Ni-phenyl
complexes.[33,34] Overall, complex 2 features an average oxidation
state of +1.5. Its diamagnetism (Figures S11 and S12, Supporting


Information), along with its short Ni Ni distances, suggest
The average Ni Ni distance from the apical Ni (Ni5) to the
that the Ni valence electrons in 2 are antiferromagnetically
coupled and delocalized over all four Ni centers, as was observed
for 1.
We next attempted to independently generate complex 3.
Thus, a mixture of 1.5 equiv of PhSSPh, 5 equiv of [Ni(1,5-cod)2],
and 4.5 equiv of PEt3 was heated to 90 °C in toluene for 18 h.
Work-up of the reaction mixture and crystallization from
MeCN/Et2O resulted in the deposition of a mixture of [Ni5(S)
2(SPh)2(PEt3)5] (3) and [Ni8(S)5(PEt3)7] (4), as a brown powder
square base is 2.65 Å (range = 2.623(8) to 2.663(7) Å), which is

consistent with a long Ni Ni bonding interaction (FSR = 1.07).

The Ni Ni distances within the square base alternate

between long and short interactions (e.g., Ni1 Ni2 = 2.786(8),
Ni3 Ni4 = 2.789(8), Ni3–Ni1 = 2.994(7), Ni4–Ni2 = 2.983(8) Å),

suggesting that a minimal bonding interaction exists between
Ni3 and Ni1 and Ni4 and Ni2, respectively. Nonetheless, the
diamagnetism of 3, as evidenced by its 31P{1H} NMR spectrum
(Figure S22, Supporting Information), suggests that the valence
d electrons of the five Ni centers are strongly coupled. Finally,
complex 3 possesses an average oxidation state of +1.20.
Crystals of 4 suitable for X-ray crystallography were grown
from the slow diffusion of acetonitrile into a diethyl ether solu-
tion of a mixture 3 and 4 at −25 °C for 7 d. Complex 4 crystal-
lizes in the triclinic space group P-1 (Figure 5). It is isostructural
with [Ni8(S)5(PPh3)7], which was reported by Fenske et al.[19]
Interestingly, [Ni8(S)5(PPh3)7] was synthesized by reaction
of [NiCl2(PPh3)2] and S(SiMe3)2, followed by reduction with
Scheme 3. Synthesis of 3 and 4.
Small 2020, 2003133
2003133 (4 of 8)
© 2020 Wiley-VCH GmbH

www.advancedsciencenews.com
www.small-journal.com
observed for much larger clusters, such as [Ni21(Se)14(PEt2Ph)12][36]
and [Ni23(Se)12(PEt3)13],[27,37] suggesting that 4 represents a tran-
sition point between small molecular clusters, such as com-
plexes 1, 2, and 3, and much larger NCs.
A
31P{1H} NMR spectrum of a mixture of 3 and 4 features a
doublet at 4.8 ppm (4P, JPP = 25.6 Hz) and a triplet at 13.4 ppm
(1P, JPP = 25.6 Hz), which correspond to the two unique 31P
environments expected for complex 3 (Figure S22, Supporting
Information). Broad singlets are also present in the spectrum
at −4.6, 13.1, and 14.9 ppm, which are assignable to complex 4.
These singlets appear in a 4:2:1 ratio. Importantly, these spec-
tral data are identical to those observed in the in situ 31P{1H}
NMR spectrum (Figure 1), confirming their presence in that
experiment. The 1H NMR spectrum of a mixture of 3 and
4 features a series of broad multiplets between 1 and 2 ppm,
which are assignable to the PEt3 ethyl environments of 3 and 4
(Figure S21, Supporting Information). This spectrum also fea-
tures a multiplet at 7.68 ppm, which is assignable to the lone
ortho-CH phenyl environment of complex 3. This chemical
shift is nearly identical to that observed for 3 in the in situ
1H NMR spectrum (Figures S2 and S3, Supporting Information),
further confirming its presence in that reaction.
Figure 5. ORTEP diagram of 4 with thermal ellipsoids plotted at 50%
probability. All hydrogen atoms omitted for clarity. Carbon atoms shown
as wireframe. Only one set of ethyl groups are displayed for disordered
ethyl groups of P3. Selected bond lengths (Å): Ni1–Ni3 = 2.4988(8),
Ni1–Ni5 = 2.5223(7), Ni1–Ni6 = 2.4847(6), Ni1–Ni4 = 2.4756(6), Ni1–
Ni7 = 2.3982(6), Ni1–Ni8 = 2.4008(8), Ni1–Ni2 = 2.3871(7).
2.3. Mechanistic Considerations
To rationalize our in situ 31P{1H} and 1H NMR spectro-
scopic results, we suggest that the first step involves reaction of
[Ni(1,5-cod)2] with PEt3 to form [(1,5-cod)Ni(PEt3)2] (Scheme 4).
This proposed step fits with the reported reactivity of [Ni(1,5-
cod)2], which is known to readily react with monodentate phos-
phines to form complexes of the type [(1,5-cod)Ni(PR3)2].[30,38–41]
We hypothesize that [(1,5-cod)Ni(PEt3)2], along with unreacted
[Ni(1,5-cod)2], then react with PhSSPh to form complex 1,
Zn metal, and not by S-atom transfer from PhSSPh. The Ni8
core of 4 consists of two edge-sharing trigonal bipyramids. Its
five sulfide ligands each adopt μ3 binding modes. In addition,
PEt3 ligands are attached to all but one Ni center. This unique
Ni atom (Ni1) appears at the center of the cluster is bonded to

the other seven Ni centers with an average Ni Ni distance of
2.453 Å (range = 2.3871(7)–2.5223(7) Å). The FSRs for these dis-
1
tances (0.96–1.01) are consistent with those expected for Ni Ni
which, according to our 31P{1H} and H NMR spectroscopic

single bonds.[31,35] In contrast, the Ni Ni distances between the
results, is the major Ni-containing cluster at short reaction
times. Its formation can be explained by invoking reaction
of PhSSPh with [(1,5-cod)Ni(PEt3)2] to form [Ni(SPh)2(PEt3)2]
via oxidative addition. 2 equiv of [Ni(SPh)2(PEt3)2] could then
couple with either [Ni(1,5-cod)2] or [(1,5-cod)Ni(PEt3)2] to form 1.
It is important to note that all of the PEt3 and PhSSPh is con-
sumed in the initial stages of the reaction, but not all of the
Ni(0) is consumed, which is consistent with the fact that

peripheral Ni atoms are longer (range = 2.6517(8)–2.8216(8) Å;
FSR = 1.07–1.13), suggesting weaker metal–metal interactions

among these atoms. Similar Ni Ni distances are observed for
[Ni8(S)5(PPh3)7].[19] Ni1 is also bound to one sulfide ligand (S2)

with an Ni S bond length of 2.1296(9) Å. The unique coordi-
nation environment (and high coordination number) observed
for Ni1 is reminiscent of the Ni coordination environments
Scheme 4. Proposed reaction sequence that converts [Ni(1,5-cod)2] into complexes 3 and 4.
2003133 (5 of 8)
© 2020 Wiley-VCH GmbH
Small 2020, 2003133

www.advancedsciencenews.com
www.small-journal.com
sub-stoichiometric amounts of PEt3 and PhSSPh were added
to the reaction mixture. Significantly, this pool of unreacted
Ni(0) (i.e., [Ni(1,5-cod)2] and [(1,5-cod)Ni(PEt3)2]) likely plays an
important role in cluster growth (see below).
On standing at room temperature, we observe a decrease
of 1, [Ni(1,5-cod)2], and [(1,5-cod)Ni(PEt3)2], and an increase
in 2, suggesting that complex 1 converts into 2, which occurs
via formal addition of 1 equiv of “Ni(0)” to 1, concomitant
Scheme 5. Thermolysis of 2.

with a C S oxidative addition, which converts a thiophenolate
internal standard (Figure S10, Supporting Information). Also
present in the reaction mixture is a small amount of complex 3,
which formed in 9% yield. To explain the presence of both
2 and 3 in this sample we suggest that the rates of Ni(0) addi-
tion to their precursor complexes are similar. Importantly, the
observation of both species is evidence for cluster growth by
addition of Ni(0) equivalents. The formation of 5 in the reac-
tion mixture is harder to rationalize, as it is the only mono-
metallic species that appears to form during the reaction. Its
presence in the reaction mixture was confirmed by X-ray crys-
tallography (Figure S1, Supporting Information). It is possible
that 5 forms by reaction of PhSSPh with [(1,5-cod)Ni(PEt3)2]
to form [Ni(SPh)2(PEt3)2], followed by donation of an S-atom
to another cluster. Alternatively, its formation can be rational-
ized by fragmentation of 2 into 5 and the unobserved species,
[Ni3(S)(SPh)2(PEt3)], which can serve as a source of atoms and
ligands for the growth of other clusters. While the latter route
may appear somewhat nonintuitive, it is consistent with the
observed reactivity of 2 (Scheme 5). Specifically, heating an
independently prepared sample of 2 in benzene-d6 at 55 °C
results in formation of complex 5, as well as small amounts of
3, according to the 31P{1H} NMR spectrum (Figure S16, Sup-
porting Information). Also formed is a small amount of an uni-
dentified PEt3-containing material. Interestingly, in the in situ
monitoring experiment (Figure 1e and Table S2, Supporting
Information) the amount of complex 5 begins to decrease
upon prolonged heating, which may be due to its unproductive
thermal decomposition. Alternatively, 5 could be serving as the
source of “NiS” that is apparently required to form complex 4.
Indeed, the involvement of monometallic complexes in nano-
cluster growth has come under increased scrutiny in recent
years.[43] For example, we demonstrated that the conversion of
a Cu25 nanocluster into a Cu29 nanocluster was mediated by
[(Ph2phen)(PPh3)CuCl], which acted as a “CuCl” carrier.[44]
Several research groups have explored the mechanisms of
nanocluster growth over the past decade.[4,45] These studies
have typically employ mass spectrometry to monitor cluster
speciation.[46–52] For example, Xie and co-workers have studied
the growth of Au25 via mass spectrometry. Incredibly, they iden-
tify nearly 30 intermediates along the Au25 reaction pathway.[11]
Likewise, we identified numerous intermediates along the Ni8
reaction pathway using 31P NMR spectroscopy to monitor cluster
ligand into a phenyl ligand and a μ3-sulfide ligand. The source
of “Ni(0)” required to form 2 can be either [Ni(1,5-cod)2] or
[(1,5-cod)Ni(PEt3)2], which are abundant at early time scales,
according to the 31P{1H} and H NMR spectra of the reaction
mixture. The amount of complex 3 also slowly increases during
the early stages of the reaction. Its formation can be rational-
ized by invoking addition of an equivalent of “Ni(0)” to 2, con-
1

comitant with a C S oxidative addition step followed by reduc-
tive elimination of biphenyl. This hypothesis is consistent with
the observation of biphenyl in the reaction mixture as well as
the continued presence of Ni(0) equivalents at this stage of the
reaction. We also note that PhSSPh has previously been shown
to act as an S-atom donor in the synthesis of Ru sulfide clusters,
likely via a similar sequence of oxidative addition and reductive
elimination steps.[42]
Upon heating to 40 °C, complex 2 begins to decrease in con-
centration, while complex 3 continues to increase. Complex 4
also appears in the reaction mixture at this temperature. Its for-
mation can be explained by invoking elimination of biphenyl
from 3, followed by formal addition of 2 equiv of “Ni(0),” 2 equiv
of PEt3, and 1 equiv of “NiS.” Alternatively, complex 4 could
be formed directly from 2 via a similar sequence. With pro-
longed heating, we observe the complete disappearance of 2
and further increases of 3 and 4, which are the only observable
PEt3-containing clusters at this stage of the reaction. Impor-
tantly, this observation is consistent with complex 2 being the
precursor to complexes 3 and 4. While 4 appears to be the ulti-
mate product of the reaction, we suggest that it is just the most
stable cluster at these temperatures and time scales. Other,
larger clusters can probably be accessed by moving to higher
temperatures and longer time scales. Finally, it is important to
note that complete conversion of 3 into 4 is likely not observed
because of the consumption of all the “Ni(0)” equivalents at
later stages of the reaction. It is also interesting to note the
apparent absence of Ni6- and Ni7-containing clusters, which
are probably being generated during the conversion of 3 into
4. We hypothesize that they are too quickly converted into 4 to
be generated in sufficient concentrations to be observed. For
comparison, Fenske et al. isolated a comparable series of sele-
nide-supported Ni clusters, including [Ni4Se3(P(CH2CH2Ph)3)5],
[Ni7Se5(PiPr3)6], and [Ni8Se6(PEt2Ph)6]. These clusters were iso-
lated by varying solvent, stoichiometry, and identity of the sup-
porting ligand and were hypothesized to be intermediates to a
larger Ni21 cluster.[36] While the mechanism of cluster forma-
tion is obviously different than in our case, this example is con-
sistent with the stepwise cluster growth that we observe for 3
and 4.
growth. While not as widely applicable as mass spectrometry, 31
P
NMR spectroscopic monitoring is a complementary technique
that could find broad utility in nanocluster synthesis, especially
given the prevalence of phosphine coligands in nanoclusters.
Consistent with the proposed mechanism of formation of
2 and 3, reaction of 1 with 1 equiv of [Ni(1,5-cod)2] results in
formation of 2 in 21% yield after 17 h at room temperature,
according to the integration of its 31P resonances against an
3. Conclusions
In summary, we have shown that reaction of [Ni(1,5-cod)2] with
sub-stoichiometric amounts of the thiolate source, PhSSPh,
Small 2020, 2003133
2003133 (6 of 8)
© 2020 Wiley-VCH GmbH

www.advancedsciencenews.com
www.small-journal.com
[1] Q. Yao, T. Chen, X. Yuan, J. Xie, Acc. Chem. Res. 2018, 51, 1338.
[2] R. W. Huang, J. Yin, C. Dong, A. Ghosh, M. J. Alhilaly, X. Dong,
M. N. Hedhili, E. Abou-Hamad, B. Alamer, S. Nematulloev, Y. Han,
O. F. Mohammed, O. M. Bakr, J. Am. Chem. Soc. 2020, 142, 8696.
[3] M. Azubel, R. D. Kornberg, Nano Lett. 2016, 16, 3348.
[4] R. Jin, C. Zeng, M. Zhou, Y. Chen, Chem. Rev. 2016, 116, 10346.
[5] I. Chakraborty, T. Pradeep, Chem. Rev. 2017, 117, 8208.
[6] A. W. Cook, T. W. Hayton, Acc. Chem. Res. 2018, 51, 2456.
[7] P. D. Jadzinsky, G. Calero, C. J. Ackerson, D. A. Bushnell,
R. D. Kornberg, Science 2007, 318, 430.
[8] C. P. Joshi, M. S. Bootharaju, M. J. Alhilaly, O. M. Bakr, J. Am. Chem.
Soc. 2015, 137, 11578.
[9] A. Ghosh, R.-W. Huang, B. Alamer, E. Abou-Hamad, M. N. Hedhili,
O. F. Mohammed, O. M. Bakr, ACS Mater. Lett. 2019, 1, 297.
[10] T. Chen, V. Fung, Q. Yao, Z. Luo, D. E. Jiang, J. Xie, J. Am. Chem.
Soc. 2018, 140, 11370.
[11] Z. Luo, V. Nachammai, B. Zhang, N. Yan, D. T. Leong, D. E. Jiang,
J. Xie, J. Am. Chem. Soc. 2014, 136, 10577.
[12] A. W. Cook, G. Wu, T. W. Hayton, Inorg. Chem. 2018, 57, 8189.
[13] H. N. Kagalwala, E. Gottlieb, G. Li, T. Li, R. Jin, S. Bernhard, Inorg.
Chem. 2013, 52, 9094.
[14] M. Zhu, S. Zhou, C. Yao, L. Liao, Z. Wu, Nanoscale 2014, 6, 14195.
[15] S. G. Bratsch, J. Phys. Chem. Ref. Data 1989, 18, 1.
[16] A. Gaita-Ariño, F. Luis, S. Hill, E. Coronado, Nat. Chem. 2019, 11, 301.
[17] J. Tejada, E. M. Chudnovsky, E. Del Barco, J. M. Hernandez,
T. P. Spiller, Nanotechnology 2001, 12, 181.
in the presence of PEt3, results in initial formation of the low-
valent Ni-thiolate cluster, [Ni3(SPh)4(PEt3)3] (1), which then
converts into [Ni4(S)(Ph)(SPh)3(PEt3)3] (2), via formal addition

of “Ni(0),” along with a C S oxidative addition step. Complex 2
then converts into a mixture of [Ni5(S)2(SPh)2(PEt3)5] (3) and
[Ni8(S)5(PEt3)7] (4), via further addition of “Ni(0)” equivalents, in


combination with a series of C S oxidative addition and C C
reductive elimination steps. The isolation of phenyl-containing,
organometallic Ni cluster 2 is particularly significant because it
clarifies the mechanism by which the sulfide ligand is generated
as well as mechanism by which biphenyl is formed in the reac-
tion mixture. Overall, this work provides an unprecedented level
of detail of the early stages of Ni nanocluster growth and identi-
fies the fundamental reaction steps (i.e., metal(0) atom addition,


C S oxidative addition, and C C reductive elimination) that are
required to grow an individual cluster. In addition, this work
uncovers a previously unrecognized route to Ni sulfide NCs,
namely, use of PhSSPh as a S-atom source. Finally, and perhaps
most importantly, the facile formation of sulfide-containing
NCs from oxidative addition of a bound thiophenolate group
demonstrates that the isolation of low-valent, thiolate-stabilized
NCs of Ni will be incredibly challenging. Given the proclivity of

low-valent Ni to effect C S oxidative addition, we suggest that

thiolate-stabilized NCs of Ni will only be isolable if C S oxida-
[18] A.-H. Lu, E. L. Salabas, F. Schüth, Angew. Chem., Int. Ed. 2007, 46,
1222.
[19] D. Fenske, J. Hachgenei, J. Ohmer, Angew. Chem., Int. Ed. 1985, 24,
706.
[20] S. Koenig, A. Eichhöfer, N. R. M. Crawford, R. Ahlrichs, D. Fenske,
Z. Anorg. Allg. Chem. 2007, 633, 713.
tive addition can be somehow suppressed.
CCDC 2004575–2004579 contains the supplementary crystal-
lographic 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.
[21] V. W. Day, R. O. Day, J. S. Kristoff, F. J. Hirsekorn, E. L. Muetterties,
J. Am. Chem. Soc. 1975, 97, 2571.
[22] M. G. Thomas, W. R. Pretzer, B. F. Beier, F. J. Hirsekorn,
E. L. Muetterties, J. Am. Chem. Soc. 1977, 99, 743.
[23] E. L. Muetterties, E. Band, A. Kokorin, W. R. Pretzer, M. G. Thomas,
Inorg. Chem. 1980, 19, 1552.
Supporting Information
Supporting Information is available from the Wiley Online Library or
from the author.
[24] A. J. Touchton, G. Wu, T. W. Hayton, Organometallics 2020, 39, 1360.
[25] E. A. Doud, C. J. Butler, D. W. Paley, X. Roy, Chem. - Eur. J. 2019, 25,
10840.
Acknowledgements
[26] J. G. Brennan, T. Siegrist, S. M. Stuczynski, M. L. Steigerwald,
J. Am. Chem. Soc. 1989, 111, 9240.
The authors thank the National Science Foundation (CHE 1764345)
for financial support of this work. This research made use of the
400 MHz NMR Spectrometer of the University of California Santa
Barbara Chemistry Department, an National Institutes of Health
Shared Instrumentation Grant (1S10OD012077-01A1). The Materials
Research Laboratory Shared Experimental Facilities are supported
by the Materials Research Science and Engineering Center Program
of the National Science Foundation under award DMR 1720256; a
member of the National Science Foundation-funded Materials Research
Facilities Network. A.J.T. thanks the Mellichamp Academic Initiative in
Sustainability at UCSB for a summer fellowship.
[27] J. G. Brennan, T. Siegrist, Y. U. Kwon, S. M. Stuczynski,
M. L. Steigerwald, J. Am. Chem. Soc. 1992, 114, 10334.
[28] L. R. Narasimhan, T. T. M. Palstra, S. M. Tanzler, M. L. Steigerwald,
Phys. Rev. B: Condens. Matter Mater. Phys. 1995, 51, 9337.
[29] This stoichiometry was chosen because it matches the relative
abundance of the components in complex 4. However, other
stoichiometries produce similar results.
[30] L. Cronin, C. L. Higgitt, R. Karch, R. N. Perutz, Organometallics
1997, 16, 4920.
[31] L. Pauling, J. Am. Chem. Soc. 1947, 69, 542.
[32] R. J. Eisenhart, P. A. Rudd, N. Planas, D. W. Boyce, R. K. Carlson,
W. B. Tolman, E. Bill, L. Gagliardi, C. C. Lu, Inorg. Chem. 2015, 54,
7579.
[33] A. Zeller, E. Herdtweck, T. Strassner, Eur. J. Inorg. Chem. 2003, 2003,
1802.
Conflict of Interest
The authors declare no conflict of interest.
[34] H.-F. Klein, A. Bickelhaupt, M. Lemke, H. Sun, A. Brand, T. Jung,
C. Röhr, U. Flörke, H.-J. Haupt, Organometallics 1997, 16, 668.
[35] L. Pauling, Proc. Natl. Acad. Sci. U. S. A. 1976, 73, 4290.
[36] D. Fenske, H. Krautscheid, M. Müller, Angew. Chem., Int. Ed. 1992,
31, 321.
Keywords
mechanisms, nanoclusters, nickel, sulfide, synthesis, thiolate
Received: May 19, 2020
Revised: August 5, 2020
Published online:
[37] P. Wix, G. E. Kostakis, V. A. Blatov, D. M. Proserpio, S. P. Perlepes,
A. K. Powell, Eur. J. Inorg. Chem. 2013, 2013, 520.
2003133 (7 of 8)
© 2020 Wiley-VCH GmbH
Small 2020, 2003133

www.advancedsciencenews.com
www.small-journal.com
[38] Q. Feng, R. Tong, J. Am. Chem. Soc. 2017, 139, 6177.
[39] H. Maciejewski, A. Sydor, B. Marciniec, M. Kubicki, P. B. Hitchcock,
Inorg. Chim. Acta 2006, 359, 2989.
[40] E. Despagnet-Ayoub, M. K. Takase, J. A. Labinger, J. E. Bercaw,
J. Am. Chem. Soc. 2015, 137, 10500.
[41] H. Schwager, C. Krüger, R. Neidlein, G. Wilke, Angew. Chem. 1987, 99, 72.
[42] N. Begum, M. I. Hyder, M. R. Hassan, S. E. Kabir, D. W. Bennett,
D. T. Haworth, T. A. Siddiquee, D. Rokhsana, A. Sharmin,
E. Rosenberg, Organometallics 2008, 27, 1550.
[46] A. C. Dharmaratne, T. Krick, A. Dass, J. Am. Chem. Soc. 2009, 131,
13604.
[47] Y. Zhang, Q. Hu, M. C. Paau, S. Xie, P. Gao, W. Chan, M. M. F. Choi,
J. Phys. Chem. C 2013, 117, 18697.
[48] Y. Yu, Z. Luo, Y. Yu, J. Y. Lee, J. Xie, ACS Nano 2012, 6, 7920.
[49] X. Yang, Y. Su, M. C. Paau, M. M. F. Choi, Anal. Chem. 2012, 84,
1765.
[50] M. van der Linden, A. J. van Bunningen, M. U. Delgado-Jaime,
B. Detlefs, P. Glatzel, A. Longo, F. M. F. de Groot, J. Phys. Chem. C
2018, 122, 28351.
[43] X. Kang, M. Zhu, Chem. Mater. 2019, 31, 9939.
[44] T.-A. D. Nguyen, Z. R. Jones, D. F. Leto, G. Wu, S. L. Scott,
T. W. Hayton, Chem. Mater. 2016, 28, 8385.
[45] N. Goswami, Q. Yao, T. Chen, J. Xie, Coord. Chem. Rev. 2016, 329, 1.
[51] H. Qian, Y. Zhu, R. Jin, ACS Nano 2009, 3, 3795.
[52] T. Chen, Q. Yao, R. R. Nasaruddin, J. Xie, Angew. Chem., Int. Ed.
2019, 58, 11967.
Small 2020, 2003133
2003133 (8 of 8)
© 2020 Wiley-VCH GmbH