Chemical Society Reviews

Metal Carbonyl Clusters: Synthesis and Catalysis

Journal: Chemical Society Reviews

Manuscript ID CS-REV-02-2021-000161.R2

Article Type: Review Article

Date Submitted by the

04-Jun-2021
Author:

Complete List of Authors: Cesari, Cristiana; University of Bologna, Chimica Industriale "Toso
Montanari"
Shon, JongHwa; UC Davis, Chemistry
Zacchini, Stefano; Universita di Bologna, Dipartimento di Chimica
Industriale "Toso Montanari"
Berben, Louise; UC Davis, Chemistry
Page 1 of 76 Chemical Society Reviews

Metal Carbonyl Clusters of Groups 8 – 10: Synthesis and Catalysis

Cristiana Cesari,‡2 Jong-Hwa Shon,‡1 Stefano Zacchini,*2 Louise A. Berben*1

1 Department of Chemistry, The University of California, Davis CA, 95616, United States
2 Dipartimento di Chimica Industriale "Toso Montanari", Università di Bologna, Viale
Risorgimento 4, 40136 Bologna, Italy

‡ These authors contributed equally.

Corresponding Authors
[email protected], [email protected]

Abstract. In this review article, we discuss advances in the chemistry of metal carbonyl clusters
(MCCs) spanning the last three decades, with an emphasis on the more recent reports and those
involving Groups 8 – 10 elements. Synthetic methods have advanced and been refined, leading to
higher-nuclearity clusters and a wider array of structures and nuclearities. Our understanding of the
electronic structure in MCCs has advanced to a point where molecular chemistry tools and other
advanced tools can probe their properties at a level of detail that surpasses that possible with other
nanomaterials and solid-state materials. MCCs therefore advance our understanding of structure–
property–reactivity correlations in other higher-nuclearity materials. With respect to catalysis, this
article focuses only on homogeneous applications, but it includes both thermally and
electrochemically driven catalysis. Applications in thermally driven catalysis have found success
where the reaction conditions stabilise the compounds toward loss of CO. In more recent years,
MCCs, which exhibit delocalised bonding and possess many electron-withdrawing CO ligands,
have emerged as very stable and effective for reductive electrocatalysis reactions since reduction
often strengthens M–C(O) bonds and since room-temperature reaction conditions are sufficient for
driving the electrocatalysis.

1. Introduction
A review article describing clusters and their analogies with surfaces was published in 1979 by
Muetterties and Rhodin and coworkers; the second page states, “With such riches, a comprehensive
review of chemical bonding in metal clusters compared to metal surfaces is neither practical nor
appropriate here. ... The largest subgroup comprises the metal carbonyl clusters and the size range

1
 Chemical Society Reviews Page 2 of 76

within this subgroup extends through the full, established range for clusters of 3 to 30.”1 Now 42
years later, the size range of homoleptic metal carbonyl clusters (MCCs) has expanded, ranging
from 3 to 56; since then, hundreds of papers reporting the activity targeting the many aspects of
MCC chemistry have been published. Numerous other reviews have appeared since that
time,234567891011121314 and a comprehensive review is certainly not practical or appropriate in the
foregoing pages of this article. Hence, we will focus our insights on synthetic methods for obtaining
Groups 8 – 10 MCCs and on their use as homogeneous catalysts and as homogeneous
electrocatalysts. We will feature work published in the most recent decades, and we will provide a
view of future interesting directions in the field. We have decided to exclude MCCs of Groups 5 – 7
because they span a smaller range of clusters, and examples of their synthesis and catalysis are
primarily older work.
The work of Hieber defined the beginning of modern metal-carbonyl chemistry and at the
same time, of modern organometallic hydride chemistry with the reports of reactions surrounding
Fe(CO)5 and the synthesis of H2Fe(CO)4 in 1931.15 The chemistry of MCCs developed rapidly in
the 1970s and 1980s, whereas there was a marked decline in publications on this topic in the 1990s.
A renewed interest started in the noughties,16 due to the intense research in the more general field of
molecular nanoclusters and ultra-small metal nanoparticles.171819202122 Within this framework, the
recent research in MCC chemistry has focused on the synthesis and structural characterization of
metal carbonyl nanoclusters (high-nuclearity MCCs), the investigation of their physical properties
in relationship with their molecular structure and composition, and their applications in catalysis
and electrocatalysis. Because high temperatures may initiate CO ligand loss, electrocatalysis
enables catalysis at room temperature, where MCC decomposition is less likely. Electrochemical
reduction reactions in MCCs are particularly promising because the many CO ligands are very
efficient at stabilising any electrons added to the cluster core to initiate reductive electrocatalytic
reaction cycles. Some of these aspects will be covered in this review. General references, covering
the full scope of MCC chemistry, may be found in the cited literature.
In the first part of this review, we discuss in detail the synthesis of MMCs from Groups 8 –
10; after general principles, we illustrate representative examples from recent work. In the
Electronic Supporting Information (ESI), we also tabulate many of the known MCCs with citations
to their syntheses. To further narrow the scope of the review, we mainly focus our attention on the
syntheses of homometallic and homoleptic MCCs. Only a few examples of significant
heterometallic and heteroleptic MCCs and of MCCs containing main-group elements will be
discussed. The readers can find pertinent references throughout the text. In the second part of this
review, we discuss both homogeneous catalysis and homogeneous electrocatalysis, along with the

2
Page 3 of 76 Chemical Society Reviews

relative advantages of using MCCs to promote these processes. To clarify, homogeneous refers to
the fact that the (electro)catalysts are in solution-phase along with reactants and other additives;
unless specified, "catalysis" refers to reactions requiring thermal activation while "electrocatalysis"
requires an applied potential. Heterogeneous catalysis, where the electrocatalyst is not in the same
phase as the reactants, falls outside the scope of this review; however, we recognise that MCCs
have been supported by various matrices to modify their catalytic activities, and some of that work
has been summarised in review articles or books.2324252627 For instance, Pt12 and Pt13 clusters have
been selectively synthesised within dendrimer templates and show dramatically different catalytic
activities in the oxygen reduction reaction, which occurs at the cathode of fuel cells.28 MCCs have
also been employed for the preparation of Pt clusters of controlled sizes inside metal–organic
frameworks and employed for catalysis and electrocatalysis.29

2. Metal Carbonyl Clusters: Synthesis

2.1 General Principles
Methodologies for the syntheses of MMCs may be grouped into four broad categories,5 which can
be applied to both homo- and hetero-metallic species (Scheme 1):
1) Direct (reductive) carbonylation;
2) Thermal methods;
3) Redox methods;
4) Other chemically induced methods.

THERMAL METHODS
(a) Pyrolysis (solid)
DIRECT CARBONYLATION (b) Thermolysis (solvent)
CO
METAL
METAL METAL
CARBONYL CARBONYL
METAL SALT CO REDOX METHODS CLUSTER
OR COMPLEX Reducing agent (a) Reduction
(b) Oxidation
(c) Redox condensation
REDUCTIVE CARBONYLATION
(d) Disproportionation
CHEMICAL
MODIFICATION
OTHER CHEMICALLY
INDUCED METHODS

Scheme 1. General methods for the synthesis of MCCs.5 MCCs can be obtained from non-carbonyl
precursors by direct or reductive carbonylation. Low nuclearity metal carbonyls can be transformed
into larger MCCs through thermal and redox methods. Moreover, MCCs can be used as starting

3
 Chemical Society Reviews Page 4 of 76

material in thermal, redox or other chemically induced methods for the preparation of further
MCCs.

It must be remarked that the distinction among these four categories is mainly based on the
experimental conditions adopted; from a mechanistic point of view (when a mechanism may be
devised), they are often overlapping. For instance, heating a metal carbonyl in the presence of a
base is usually classified as a thermal method, even if the mechanism may also involve redox
reactions. Syntheses of category (1) start from a non-carbonyl precursor, whereas those of
categories (2)–(4) employ carbonyl species.
Compared to C–C bonds, M–M interactions are weaker and non-directional. As a
consequence, MCCs adopt a rich variety of structures even in the case of species with similar
compositions. The products obtained are highly dependent on the experimental conditions, which,
therefore, must be very carefully controlled. A tailored synthesis of a particular structure is still a
dream. Nonetheless, it is nowadays possible to outline some general guidelines for the preparation
of MCCs with a given composition and for inducing their growth up to the nanometric level.2,3,5,6,8-
11,30 Moreover, isolated MCCs can be modified and functionalised with ancillary ligands to induce
new properties, to be anchored onto supports, or to be transformed into nanostructured
catalysts.2,3,5,6,8-11,30
Infrared (IR) monitoring of the reactions is a very useful tool. Reactions are often
accompanied by the formation of side products, such as small carbonyl complexes (e.g., Ni(CO)4,
[Co(CO)4]–, Ni(CO)3(PPh3), [Rh(CO)4]–), bulk metals, metal salts, and other MCCs. Separation and
purification of the major product is mainly achieved using solvent extraction, owing to the
solubility differences of the product and side products in water and organic solvents. In the case of
anionic MCCs, this may be enhanced by choosing suitable tetra-substituted-ammonium or
-phosphonium cations. For less charged and neutral MCCs, chromatography is sometimes
employed. The resulting yields vary case by case, but many MCCs can be obtained on the
hundreds-of-milligram or even gram scale.
Direct (reductive) carbonylation. Direct and reductive carbonylation are the major methods
for the preparation of metal carbonyls starting from non-carbonyl precursors.2,3,5 Direct
carbonylation of a finely divided metal under CO works only in a few cases, such as with Fe, Ni,
and Co. Reductive carbonylation is a more general procedure, which requires the reduction of a
metal salt under CO atmosphere. The reducing agent (H2, Na, CO itself, etc.), temperature, and CO
pressure employed depend on the metal. In some cases, harsh conditions are required, and in other
cases, the reactions may be performed at room temperature (RT) with 1 atm of CO. For instance,

4
Page 5 of 76 Chemical Society Reviews

ambient reductive carbonylation may be used for the synthesis of Rh4(CO)12,31 Ir4(CO)12,32
[Rh(CO)4]–, 33 [Ir(CO)4]–,33 and the [Pt3n(CO)6n]2– (n = 1–10).34 Even if Ru3(CO)12 may be prepared
with CO at atmospheric pressure,35 the high-pressure synthesis is often more convenient.36 The use
of CO as a reducing agent is summarised by the half-reaction (Equation 1); a suitable amount of
base is used to tune the reducing ability of CO.

CO + 2OH– CO2 + H2O + 2e (1)

Reductive carbonylation may also be used in the solid state.37 Moreover, reductive carbonylation of
Pd salts in the presence of PR3 has been extensively used by Dahl and Mednikov for the synthesis
of Pd-CO-PR3 clusters (see the ESI: Scheme S1).13
Thermal methods. CO elimination from a MCCs is an endothermic process with a
favourable entropic term; therefore, it is usually promoted by thermal methods, in the solid phase
(pyrolysis) and in solution (thermolysis). Other than decomposition, four possible events may
follow CO elimination from a MCC, with event (4) as the most common:2
1) No reaction. In very few cases, the same cluster cage can be stabilised by different numbers
of CO ligands, as in the case of [Rh12Sn(CO)27–x]4– (n = 0, 1, 2) 38 and [Ni9CoC2(CO)16–x]3–
(x = 0, 1).39
2) Permutation. Elimination of two CO ligands from the trigonal bipyramidal (TBP)
[PtRh4(CO)14]2– generates the TBP [PtRh4(CO)12]2– with the migration of Pt from an
equatorial to an apical position.40
3) Intramolecular condensation. In some cases, CO elimination induces a structural
rearrangement of the cluster with retention of the nuclearity. This is exemplified well by the
trigonal prismatic (TP) MCCs, [M6C(CO)15]2– and [Co6N(CO)15]– (M = Co, Rh), which are
transformed into octahedral (Oh) clusters, [M6C(CO)13]2– and [Co6N(CO)13]–, after
elimination of two carbonyl ligands.41 The number of M–M bonds increases from 9 (TP) to
12 (Oh), and because of this, the process may be referred to as an intramolecular
condensation where M–CO bonds are replaced by M–M bonds.
4) Intermolecular condensation. This is the most common process observed after heating a
MCC. In this case, CO elimination produces an unsaturated species, which condenses with a
second cluster fragment resulting in a higher-nuclearity MCC. Several examples will be
given in the following sections.

5
 Chemical Society Reviews Page 6 of 76

Heating a MCC may also promote CO dismutation to C and CO2, resulting in carbide
MCCs.2,3,5,6,8-11,30 Thermal reactions of neutral clusters are often unselective, leading to mixtures of
products. As demonstrated by Chini, the use of an anionic cluster as starting material or the addition
of a base to a neutral carbonyl in thermal decomposition often provides better selectivity.42
Redox methods. Redox reactions used for the synthesis of MCCs include reduction,
oxidation, redox condensation, and disproportionation.2,3,5,6,8-11,30 Multivalent MCCs can undergo
reversible redox reaction without any major structural rearrangements;6 examples include [Ni22–
n– (n = 5–8),43 [Pt33(CO)38]n– (n = 0–9), and [Pt40(CO)40]n– (n = 4–11).44 In general,
xPd20+x(CO)48]

redox reactions on non-multivalent (electron-precise) MCCs or the use of excess redox reagents on
multivalent MCCs lead to irreversible processes yielding new MCCs or complete decomposition.
Reduction of a MCC can be performed by (1) direct addition of electrons, for instance, using
alkali metals; or (2) nucleophilic attack of the C-atom of a coordinated carbonyl by OH– ions. In the
latter case, the first product is a metallacarboxylic acid [MCOOH]–, which is readily transformed
into a hydridocarbonylate [MH]– upon CO2 elimination or a metallacarboxylate [MCOO]2– upon
deprotonation.45 Following further deprotonation or CO2 elimination, [MH]– and [MCOO]2– are
both transformed into the dianion [M]2–, which can be stable or undergo redox condensation to
larger MCCs (Scheme 2). This approach was developed by Hieber,15 who demonstrated that
different Fe carbonyl cluster anions could be obtained from the treatment of Fe(CO)5 with base.
This method is very versatile; it can be used for the preparation of low-nuclearity carbonylates and
hydridocarbonylates as well as larger MCCs.

+ OH– - CO2
M-CO [MCOOH]– [MH]–

+ OH– -H2O + OH– -H2O

- CO2
[MCOO]2– [M]2–

+ M-CO

[Mn(CO)x]z–

Scheme 2. Condensation induced by reduction. The process is initiated by OH– nucleophlic attack
on a CO ligand and eventually leads to the hydride [MH]– or the dianion [M]2–. This, in turn, can
undergo to a condensation reaction resulting in the higher nuclearity metal carbonyl [Mn(CO)x]z–.

6
Page 7 of 76 Chemical Society Reviews

When reduction of the MCC is achieved by the direct addition of electrons, the M–M and
M–CO bonds often cleave, with the resulting species undergoing fragmentation or condensation to
larger species. Alternatively, reduction can simply result in the replacement of a CO ligand with
two negative charges with or without structural rearrangement. Indeed, [M]2– may be stable or
undergo redox condensation (Scheme 2).
Oxidation of MCCs with innocent (non-coordinating) reagents, such as tropilium
tetrafluoroborate or ferrocenium ions, may result in the formation of new M–M bonds with or
without CO loss, as in the dimerisation of [Ir6(CO)15]2– to [Ir12(CO)26]2–,46 [Pt19(CO)22]4– to
[Pt38(CO)44]2–,47 or [Co6C(CO)15]2– to [Co11C2(CO)23]n– (n = 1–3).48 Using coordinating oxidants
such as H+ and [MLx]n+, the reaction may proceed through oxidation or formation of a Lewis-type
acid–base adduct. In order to increase the nuclearity of MCCs of a few metal atoms, the latter
process has been widely employed through the addition of [ML]+ fragments (M = Cu, Ag, Au; L =
neutral ligand).3,9,30 In some cases (vide infra), the reaction of MCCs with metal salts or complexes
may result in larger MCCs via redox condensation. The use of H+ ions (from acids or hydrolysis of
[M(H2O)x]n+ ions) may result in the oxidation of the MCC or the formation of isostructural MCCs
containing hydride ligands. For instance, [Rh7(CO)16]3– is oxidised by MCl2·xH2O salts (M = Ni,
Zn, Cd) to [HRh14(CO)25]3–, [Rh15(CO)27]3–, [Rh15(CO)25(MeCN)2]3–, and [Rh17(CO)37]3–.49
Although oxidation usually results in larger clusters, in a few cases oxidative degradation is
observed. For example, treatment of [H2Co20Pd16C4(CO)48]4– with strong acids affords
[HCo15Pd9C3(CO)38]2– and, eventually with excess acid, [Co13Pd3C3(CO)29]–.505152 In several cases,
polyhydride clusters may be interconverted by simple acid–base reactions. For instance, protonation
of [Co6(CO)15]2– affords [HCo6(CO)15]–,53 while stepwise protonation of [Co15Pd9C3(CO)38]3–
results in [H2Co15Pd9C3(CO)38]2–, [HCo15Pd9C3(CO)38]–, and H3Co15Pd9C3(CO)38; these reactions
may be reversed by the addition of base.
Redox condensation was introduced and rationalised by Paolo Chini, and it represents one of
the most powerful and versatile procedures for MCC preparation.54 Redox condensation involves a
comproportionation reaction between an MCC anion and a more oxidised species (not necessarily a
cluster nor even a carbonyl compound), which can be either cationic, neutral, or anionic (Scheme 3;
also see ESI: Figure S1). Redox condensation can be used for the preparation of homo- and hetero-
metallic MCCs, even when the second metal does not form a carbonyl compound, such Pd, Cu, Ag,
Au, Hg, or main-group metals.

7
 Chemical Society Reviews Page 8 of 76

[Rh6(CO)15]2– + [Rh(CO)4]– [Rh7(CO)16]3– + 3CO

[Ni5(CO)12]2– + Ni(CO)4 [Ni6(CO)12]2– + 4CO
[Rh5(CO)15]– + [Rh4(CO)11]2– [Rh9(CO)19]3– + 7CO
[Pt6(CO)12]2– + [Pt12(CO)24]2– 2[Pt9(CO)18]2–
[Co3(CO)10]– + Fe(CO)5 [Co3Fe(CO)12]– + 3CO
5[Rh12(CO)30]2– + [Pt12(CO)24]2– 12[PtRh5(CO)15]–

+ Pt(Et2S)2Cl2
[Co6C(CO)15]2- [Co8Pt4C2(CO)24]2-

+ Pd2+
[HRu3(CO)11]- [Ru6Pd6(CO)24]2-
+ [AuCl4]-
[Ni6(CO)12]2- [Ni32Au6(CO)44]6-
+ GaCl3
[Ni6(CO)12]2- [N12Ga(CO)22]3-

+ SnCl2
[Rh7(CO)16]3- [Rh12Sn(CO)27]4-

+ BiCl3
[Rh7(CO)16]3- [Rh12Bi(CO)27]3-

+ SbCl3
[Rh7(CO)16]3- [Rh21Sb2(CO)38]5-
Scheme 3. Some representative examples of redox condensation. Balanced equations have been
written only when the complete stoichiometry of all the reagents and products are known.

Disproportionation is the reverse of redox condensation, and it may be promoted in neutral

Mx(CO)y carbonyls by using polar solvents or bases. Disproportionation can result in (a)
[ML6]2+[Ma(CO)b]2– ionic couples, (b) isocarbonyls adducts, or (c) xenophilic clusters (Scheme
4).55 An example of case (a) is represented by [Fe(DMF)6][Fe4(CO)13], which is formed after
heating Fe(CO)5 in DMF (N,N-dimethylformamide). Xenophilic clusters (case (c)) contain a direct
M–M bond between a metal centre in a formal negative oxidation state and a second metal in a
positive oxidation state, e.g., Fe4(CO)8(py)456 and [Mn][Mn7(THF)6(CO)12]2 (THF =
tetrahydrofuran; see ESI: Figure S2).57

8
Page 9 of 76 Chemical Society Reviews

Mx(CO)y + L
(L = tetrahydrofuran, ketones, amines)

(a) (c)
[ML62+][M(CO)n-]2 (b) L
mononuclear ions M M(CO)n
L
or L
[ML6 ][Ma(CO)b2-]
2+
xenophilic M-M bond
metal cluster ion L
L
(CO)nM C O M O C M(CO)n
L L
isocarbonyl linkage
Scheme 4. Disproportionation of neutral metal carbonyls can result in (a) ionic couples, (b)
isocarbonyl adducts, or (c) xenophilic clusters (M, black: metal in formal oxidation state = 0; M,
red: formal oxidation state > 0; M, blue: formal oxidation state < 0).55

Other chemically induced methods. MCCs can be modified by additional chemical

reactions. The addition of soft nucleophiles (e.g., CO, PR3, RCCR, R2C=CR2, pyridine, NO2–,
halides) may result in the formation of addition or substitution products, cluster breakdown, or even
condensation; the particular reaction route depends on the MCC, the nature of the nucleophile, the
stoichiometry of the reaction, and experimental conditions.2,3,5,6,8-11,30 Moving from cationic to
neutral and anionic MCCs, CO-substitution with stronger -donors becomes less favoured. The
addition of soft nucleophiles typically causes the removal of MLx fragments (e.g., Ni(CO)4,
Ni(CO)2(PR3)2, [Rh(CO)2I2]–), forming lower-nuclearity MCCs, but in a few cases, nucleophilic
attack promotes condensation, resulting in higher-nuclearity clusters (Scheme 5). For instance, CO-
induced cluster condensation is observed after exposing [Ni10Rh2C(CO)20]2– to CO under ambient
conditions, resulting in the bis-acetylide [Ni6Rh8(C2)2(CO)24]4– (Figure 1).58

[Ni12Ge(CO)22]2– + 6CO [Ni10Ge(CO)20]2– + 2Ni(CO)4

[Rh15(CO)27]3– + 2I– [Rh14(CO)25]4– + [Rh(CO)2I2]–

[Ni9C(CO)17]2– + 2PPh3 [Ni7C(CO)12]2– + Ni(CO)2(PPh3)2 + Ni(CO)3(PPh3)

2[Ni10C2(CO)16]2– + 8PPh3 [Ni16(C2)2(CO)23]4– + 4Ni(CO)2(PPh3)2 + CO

Scheme 5. Some representative reactions of MCCs with soft nucleophiles.

9
 Chemical Society Reviews Page 10 of 76

CO

[Ni10Rh2C(CO)20]2- [Ni6Rh8(C2)2(CO)24]4-
Figure 1. CO-induced condensation of [Ni10Rh2C(CO)20]2– to [Ni6Rh8(C2)2(CO)24]4– (green, Ni;
blue, Rh; grey, C). CO ligands have been omitted for clarity.58

As a final remark, main-group elements are most often introduced into MCCs by redox
condensation using suitable EXn compounds (E = main-group element; X = halide). Chalcogens can
also be introduced using elemental O2, S8, Se, or Te, or using chalcogen-containing EOx2–
molecules under reducing conditions.2,3,5,6,8-12,30,59,60 Suitable reagents for the synthesis of nitride-
containing MCCs from metal carbonyls are NO+, NO2–, N3, and NCO–. Carbide MCCs represent the
largest class of MCCs containing a main-group element. They can be prepared by three main
methods: 2,3,5,6,8-11,30,61
1) Thermal disproportionation of CO to C and CO2. This method is very effective in the case of
Re, Fe, Ru, and Os carbonyls.
2) CO splitting by the reduction of a bridging CO ligand. In this case, a CO ligand that C-
bonded to a metal is activated by O-coordination to a Lewis acid, such as RCO+ (generated
from RCOCl). The resulting M–C–O–C(O)R is then reduced to a metal carbide following
elimination of RCOO–. This method has been applied mainly to Fe and Co clusters.
3) Reaction of MCC anions with carbon halides (e.g., CCl4, C2Cl4, C2Cl6, C3Cl6). This method
is largely employed for the synthesis of Co-, Rh-, and Ni-carbide MCCs.

In the following sections, sample syntheses of MCCs from the Groups 8–10 metals will be
illustrated. For each metal, the synthesis of MCCs that can be further employed as precursors for the
preparation of new species will be summarised with key details. These include neutral clusters,
homometallic MCCs, species containing main-group elements, and heterometallic MCCs. Some
representative (not exhaustive) examples on how these species might be employed for the
preparation of new MCCs will then be briefly outlined.

2.2 Iron Carbonyl Clusters

10
Page 11 of 76 Chemical Society Reviews

Fe(CO)5 can be obtained in very large scale by the direct carbonylation of Fe metal. It is the best
commercially available starting material for the preparation of other iron carbonyls (Scheme 6). The
photolysis of Fe(CO)5 in acetic acid affords the almost insoluble Fe2(CO)9,62 which was probably
the first structurally characterised carbonyl compound containing a direct M–M bond, even if
supported by bridging carbonyls. Fe3(CO)12 has been obtained by three different methods:63 (a)
thermal treatment of Fe2(CO)9, (b) oxidation of [HFe(CO)4]–, and (3) reaction of [HFe3(CO)11]–
with strong acids.
[HFe(CO)4]-

KOH
[Fe(CO)4]2- NH3/H2O Fe2(CO)9 [Fe2(CO)8]2-

Na hv in CH3COOH
Fe(CO)5
100 °C in DMF or py
80 °C in NEt3/H2O

H2SO4 H2SO4
[HFe4(CO)13]- [Fe4(CO)13]2- [HFe3(CO)11]- Fe3(CO)12

KOH KOH

H2SO4 [Fe3(CO)11]2-
[H2Fe4(CO)12]2- [HFe4(CO)12]3-
CF3SO3CH3

[HFe5(CO)14]3-

Scheme 6. Synthesis of carbonylferrates and hydridocarbonylferrates from Fe(CO)5 (DMF = N,N-

dimethylformamide, py = pyridine). Anionic Fe carbonyls are obtained by a combination of redox
and thermal reactions as well as the use of acids and bases. See Table S1 in ESI for a full list with
references.
Carbonylferrates and hydridocarbonylferrates. Based on the seminal work of Hieber,15
several carbonylferrates and hydridocarbonylferrates have been prepared and structurally
characterised:45 [Fe(CO)4]2–, [Fe2(CO)8]2–, [Fe3(CO)11]2–, [Fe4(CO)13]2–, [HFe(CO)4]–, [HFe2(CO)8]–
, [HFe3(CO)11]–, [HFe4(CO)13]–, [HFe4(CO)12]3–, [HFe5(CO)14]3–, and [H2Fe4(CO)12]2– (see ESI for a
full list with references: Table S1). The paramagnetic species [Fe2(CO)8]•–, [Fe3(CO)12]•–, and
[Fe3(CO)11]•– have been confirmed and characterised by electron paramagnetic resonance studies on
(13C, 57Fe) isotopically enriched samples.6465 In contrast, variable-temperature (VT) multinuclear
nuclear magnetic resonance (NMR) studies indicate that the purported dihydrides H2Fe3(CO)11 and
H2Fe4(CO)13 should be reformulated to HFe3(CO)10(CO-H) and HFe4(CO)12(CO-H).66
Carbonylferrates and hydridocarbonylferrates are usually obtained from neutral Fe carbonyls
by reduction, treatment with base, or disproportionation. Several procedures have been reported,

11
 Chemical Society Reviews Page 12 of 76

with some species having multiple synthetic routes. Scheme 6 shows only some of the most
significant examples. Reduction of Fe(CO)5 with Na/benzophenone in THF affords the Collman's
reagent Na2[Fe(CO)4]. Depending on the specific reaction conditions, treatment of Fe(CO)5 with
OH– conveniently yields [HFe(CO)4]– or [HFe3(CO)11]–; the former is produced with NH3 in water
at RT, while the latter with NEt3 in water at 80 °C. In contrast, [Fe4(CO)13]2– is prepared in high
yields from the disproprtionation of Fe(CO)5 in either DMF or pyridine (py) at 100 °C (see ESI:
Scheme S2).
Isonuclear carbonylferrates and hydridocarbonylferrates are easily interconverted by simple
acid–base reactions. For instance, [Fe3(CO)11]2– is quantitatively protonated to [HFe3(CO)11]– by
stoichiometric amounts of strong acids, and the reaction is reversed by using 1.2-M KOH in
methanol. Interestingly, the reaction of [HFe3(CO)11]– or [Fe4(CO)13]2– with 6-M KOH results in
[HFe4(CO)12]3–.67 The unique hydride of [HFe4(CO)12]3– is 3-coordinated to a triangular face of the
cluster. This coordination is retained after the addition of two [AuPPh3]+ fragments, yielding
[HFe4(CO)12(AuPPh3)2]–;68 however, further addition of [AuPPh3]+ produces the neutral
HFe4(CO)12(AuPPh3)3, and concomitantly, the hydride migrates from the triangular face to the
tetrahedral cavity (Scheme 7).

Au(PPh3)Cl Au(PPh3)Cl

[HFe4(CO)12]3- [HFe4(CO)12(AuPPh3)2]- HFe4(CO)12(AuPPh3)3

Scheme 7. Hydride migration in [HFe4(CO)12]3– induced by the addition of [Au(PPh3)]+ fragments

(blue, Fe; yellow, Au; grey, C; red, O; white, H). The surface 3-H hydride ligand of
[HFe4(CO)12]3– and [HFe4(CO)12(AuPPh3)2]– is moved to the tetrahedral cavity of
HFe4(CO)12(AuPPh3)3 upon the addition of a third [Au(PPh3)]+ fragment.68 PPh3 ligands have been
omitted for clarity.

Iron Carbide Carbonyl Clusters. By heating Fe(CO)5 in quinoline at 160–180 °C, both CO
and metal disproportionation are observed, resulting in the carbide cluster [Fe6C(CO)16]2– (Equation

12
Page 13 of 76 Chemical Society Reviews

2).69 The same cluster can be alternatively obtained by heating Fe(CO)5 and [Fe(CO)4]2– in diglyme
at 160 °C.

7Fe(CO)5 + 6quinoline  [Fe(quinoline)6]2+ + [Fe6C(CO)16]2– + 17CO + CO2 (2)

[Fe6C(CO)16]2– is a very versatile reagent for the preparation of other Fe carbide carbonyl
clusters (Scheme 8). Its reaction with NaOH in dimethylsulfoxide (DMSO) affords the highly
reduced [Fe6C(CO)15]4–.70 Moreover, [Fe6C(CO)16]2– is oxidised by Fe3+ ions in H2O to form
Fe5C(CO)15, which, in turn, is transformed into [Fe5C(CO)14]2– after reaction with [Fe(CO)4]2–.71 In
contrast, the reaction of [Fe6C(CO)16]2– with Fe3+ in MeOH results in the 4-
(methoxycarbonyl)methylidyne cluster [Fe4(CO)12C(CO2Me)]–;72 which can be further transformed
into Fe4C(CO)13 by reaction with CF3SO3H or into [HFe4C(CO)12]– by reaction with BH3·THF.7374
The related [Fe4C(CO)12]2– dianion can be obtained using a strong base to deprotonate
[HFe4C(CO)12]– or using Na to reduce the [Fe4(CO)12(COC(O)CH3)]– adduct, which is prepared
from the reaction of [Fe4(CO)13]2– and CH3COCl.61

Fe Fe 2-
Fe Fe [Fe(CO)4]2-
Fe Fe Fe Fe 2-
C C Fe
Fe Fe Fe Fe
Fe5C(CO)15 [Fe5C(CO)14]2- Fe
[Fe4(CO)13]2-

Fe3+ in H2O
CH3COCl

O -
Fe 2- Fe 4-
160 °C NaOH C
Fe(CO)5 Fe Fe Fe Fe O CH3
quinoline C DMSO C
Fe Fe Fe Fe C
Fe Fe
Fe Fe Fe
[Fe6C(CO)16]2- [Fe6C(CO)15]4-
Fe

Fe 3+
in CH3OH [Fe4(CO)12(COC(O)CH3)]-

Na
O OCH3 -
C
C C
-
Fe Fe Fe Fe
C C 2-
Fe Fe
CF3SO3H BH3·THF NaOH Fe Fe
Fe Fe Fe Fe
Fe Fe Fe Fe
H
Fe4C(CO)13 [Fe4(CO)12C(CO2CH3)]-
[HFe4C(CO)12]- [Fe4C(CO)12]2-

Scheme 8. Synthesis of Fe carbide carbonyl clusters. All the species have been isolated and fully
characterized.61,69-74 [Fe6C(CO)16]2– is prepared directly from Fe(CO)5 by thermal treatment. Other
Fe carbide carbonyl clusters can be obtained from [Fe6C(CO)16]2– using redox reactions.

13
 Chemical Society Reviews Page 14 of 76

Alternatively, [Fe4C(CO)12]2– can be prepared starting from [Fe4(CO)13]2– in a two steps reaction by
CO scission. CO ligands have been omitted for clarity.

[Fe6C(CO)16]2–, [Fe6C(CO)15]4–, [Fe4C(CO)12]2–, and [Fe5C(CO)14]2– react with metal

carbonyls and metal complexes affording several heterometallic carbide clusters, via redox
condensation, metal replacement, or formation of Lewis-type acid–base adducts.7576 A recent
advance in this area is the reaction of [Fe5C(CO)14]2– with [Mo(CO)3(chpt)] (chpt =
cycloheptatriene) to afford the heterometallic carbide [Fe5MoC(CO)17]2–, which displays selective
alkyne reduction.77 The first Fe carbide–sulfide clusters, Fe6C(CO)16(S) and [Fe6C(CO)14(S)]2–,
were also obtained recently (Scheme 9).78 The addition of SO2 to [Fe6C(CO)16]2– results in
[Fe6C(CO)15(SO2)]2–, which is then O-methylated by MeOTf (methyltriflate = methyl
trifluoromethanesulfonate) to [Fe6C(CO)15(SO2Me)]–. Demethoxylation with BF3 results in
Fe6C(CO)15(SO), which spontaneously converts in solution into Fe6C(CO)16(S) at RT. Reduction
with Na/naphthalene eventually affords [Fe6C(CO)14(S)]2–.

O O -
2- O
Fe 2- Fe S O Fe S OMe Fe S

Fe Fe SO2 Fe Fe MeOTf Fe Fe BF3 Fe Fe

C C C C
Fe Fe Fe Fe Fe Fe Fe Fe
[Bu4N]I
Fe Fe Fe Fe
[Fe6C(CO)16] 2-
[Fe6C(CO)15(SO2)] 2- [Fe6C(CO)15(SO2Me)]- Fe6C(CO)15(SO)

e-
RT
2-
Fe S -
Fe Fe
Fe Fe e- Fe Fe e- Fe Fe
C C S C S
Fe Fe Fe Fe Fe Fe

Fe Fe Fe
2- -
[Fe6C(CO)14(S)] [Fe6C(CO)16(S)] Fe6C(CO)16(S)

Scheme 9. Synthesis of Fe6C(CO)16(S) and [Fe6C(CO)14(S)]2–.[Fe6C(CO)16]2– is transformed into

[Fe6C(CO)15(SO2)]2– by CO substitution with SO2. Further reactions with MeOTf, BF3 and
reduction with Na/naphthalene result in the target compounds.78 CO ligands have been omitted for
clarity.

Iron carbonyl clusters containing other main-group elements. Nitride atoms can be
introduced into Fe carbonyl clusters by the reduction of NO+ ions (see ESI: Scheme S3).3 The
thermal reaction of [NO][BF4] and [Fe2(CO)8]2– in diglyme is highly dependent on the temperature;
14
Page 15 of 76 Chemical Society Reviews

it affords [Fe4N(CO)12]– at 130 °C, whereas it produces [Fe5N(CO)14]– at 145 °C. [Fe4N(CO)12]– can
be alternatively obtained from the redox condensation of Fe3(CO)12 and [Fe(CO)3(NO)]–. Both
[Fe4N(CO)12]– and [Fe5N(CO)14]– are protonated by strong acids, forming the related neutral mono-
hydrides HFe4N(CO)12 and HFe5N(CO)14. The redox condensation of [Fe4N(CO)12]– with
[Fe2(CO)8]2– results in the fully interstitial nitride cluster [Fe6N(CO)15]3–, from which an Fe atom
can be removed by oxidation, affording [Fe5N(CO)14]–.
Oxygen is by far one of the most difficult main-group elements to incorporate into low-
valent metal carbonyl clusters because of its oxidising power, hard base nature, and higher affinity
for high-valent metals (see ESI: Table S2). One of the most significant exceptions is represented by
the 3-oxo carbonyl cluster [Fe3(3-O)(CO)9]2–, which is obtained from the reaction of
[Fe3(CO)11]2– with dry air.79 Indeed, this is a rare case in which letting (dry) air into a Schlenk tube
containing an anionic metal carbonyl cluster results in the selective synthesis of a new species. It is
noteworthy that protonation of [Fe3(3-O)(CO)9]2– occurs on the O atom affording the 3-hydroxo
cluster [Fe3(3-OH)(CO)9]–, whereas [AuPPh3]+ adds to the Fe3 metal cage affording [Fe3(3-O)(3-
AuPPh3)(CO)9]– (Scheme 10).8081 This is due to the fact that the hard H+ acid prefers to add to the
hard oxo base, whereas the soft [AuPPh3]+ acid prefers to interact with the softer base site of the
metal cage of the cluster. The coordination of the oxo ligand can be further expanded by reacting
[Fe3(3-O)(CO)9]2– with metal fragments, such as [Mn(CO)3(MeCN)3]+ and Ru3(CO)10(MeCN)2,
which results in [Fe3Mn(4-O)(CO)12]–82 and [Fe2Ru3(4-O)(CO)14]2–.83

15
 Chemical Society Reviews Page 16 of 76

H -
O
-
Fe Fe O
[Fe3(3-OH)(CO)9] - Fe Mn
Fe Fe
Fe
[Mn(CO)3(MeCN)3]+
H+ [Fe3Mn(4-O)(CO)12]-

O 2-
O2
[Fe3(CO)11]2- Fe Fe
Fe
[Fe3(3-O)(CO)9]2- Ru3(CO)10(MeCN)2

Au(PPh3)Cl Ru 2-
Fe Ru
O - Fe Ru
O
Fe Fe
Fe [Fe2Ru3(4-O)(CO)14]2-

Au
PPh3

[Fe3(3-O)(3-AuPPh3)(CO)9]-

Scheme 10. Reactivity of [Fe3(3-O)(CO)9]2–. This compound is obtained from the reaction of
[Fe3(CO)11]2– with dry air,79 and is protonated on the oxo ligand resulting in [Fe3(3-OH)(CO)9]–.80
The metal cage can be expanded by condensation with metal fragments such as Au(PPh3)Cl,
[Mn(CO)3(MeCN)3]+ and Ru3(CO)10(MeCN)2.81-83 CO ligands have been omitted for clarity.

Other chalcogen atoms (S, Se, Te) can be introduced into Fe carbonyl clusters by various
routes:8485
(1) Reacting Fe carbonyl anions with reagents containing the chalcogen in a positive oxidation
state, such as EO32–, EO2, and SCl2;
(2) Reacting neutral Fe carbonyls—usually Fe(CO)5 or Fe3(CO)12—with En2– polychalcogenide
anions;
(3) Reaction of Fe(CO)5 with the elemental chalcogen in the presence of a base.

Iron carbonyl clusters containing other main-group elements (e.g., As, Bi, Sb, Pb, Tl, Sn)
can be generally obtained from the reaction of iron carbonyl anions with main-group halides or
oxides (see ESI: Scheme S4).86
Heterometallic Fe-based carbonyl clusters. Heterometallic Fe–M carbonyl clusters can be
prepared by the reactions of Fe carbonyl anions with metal salts and complexes.8788 For instance,
reactions of the Collman’s reagent Na2[Fe(CO)4]·2THF with M(NHC)Cl, where M are Cu, Ag, or
16
Page 17 of 76 Chemical Society Reviews

Au and NHC is IMes (C3N2H2(C6H2Me3)2) or IPr (C3N2H2(C6H3iPr2)2), afford the anionic

[Fe(CO)4{M(NHC)}]– and the neutral Fe(CO)4{M(NHC)}2 and Fe(CO)4{M(NHC)}{M'(NHC)}
complexes (Scheme 11).899091 These, in turn, may be transformed into larger clusters by means of
thermal or chemical reactions. Such Fe–M complexes may be viewed as transition-metal Lewis
acid–base pairs92 with potential applications in bimetallic catalysis.93 For instance,
Fe(CO)4{Cu(IPr)}2 was found to be useful in the bifunctional activation of N2O.94

CO CO
+ M(NHC)Cl + M(NHC)Cl OC CO
Na2[Fe(CO)4] (NHC)M Fe CO Fe
(NHC)M M(NHC)
OC CO CO
M, M' = Cu, Ag, Au
+ M'(NHC)Cl
NHC = IMes, IPr
Me
Me i CO
Pr
N Me i
Pr OC CO
IMes = N Fe
IPr =
N Me N i
(NHC)M M'(NHC)
Pr CO
Me i
Pr
Me
Scheme 11. Reactions of Na2[Fe(CO)4] with M(NHC)Cl (M = Cu, Ag, Au; NHC = IMes, IPr; IMes
= C3N2H2(C6H2Me3)2; IPr = C3N2H2(C6H3iPr2)2). One or two [M(NHC)]+ fragments as well as two
different [M(NHC)]+ and [M'(NHC)]+ fragments can be added, resulting in [Fe(CO)4{M(NHC)}]–,
Fe(CO)4{M(NHC)}2 and Fe(CO)4{M(NHC)}{M'(NHC)}, respectively.89-91

Molecular gold nanoclusters protected by Fe–CO fragments have been prepared by redox
condensation of [Fe3(CO)11]2– with [AuCl4]–. The resulting [Au22{Fe(CO)4}12]6–,
[Au21{Fe(CO)4}10]5–, [Au28{Fe(CO)3}4{Fe(CO)4}10]8–, and [Au34{Fe(CO)3}6{Fe(CO)4}8]10– clusters
are reminiscent of thiolate-protected Au nanoclusters, comprising Au cores protected by
organometallic staple motifs (Figure 2).95

17
 Chemical Society Reviews Page 18 of 76

Figure 2. Molecular structures of [Au21{Fe(CO)4}10]5– (left) and [Au28{Fe(CO)3}4{Fe(CO)4}10]8–

(right). Spheres represent atoms as follows: yellow, Au; blue, Fe; red, O; grey, C. Both these
molecular nanoclusters can be obtained from the reaction of [Fe3(CO)11]2– with [AuCl4]– under
different experimental conditions.95

2.3 Ruthenium and Osmium Carbonyl Clusters

Ru3(CO)12 and Os3(CO)12 can be obtained by reductive carbonylation of RuCl3 and OsO4,
respectively, at high temperature and pressure.36,96 An atmospheric-pressure synthesis of Ru3(CO)12
is also available.35 The monometallic M(CO)5 carbonyls are obtained at very high CO pressures
(180–200 atm), but Ru(CO)5 can also be prepared by photolysis of Ru3(CO)12 under CO at
atmospheric pressure.97 Interestingly, treatment of OsO4 with a 3:1 combined pressure (180 atm) of
CO/H2 at 160 °C in THF results in the dihydride H2Os(CO)4, which can be deprotonated to
[HOs(CO)4]– by Na.
Homometallic Ru and Os carbonyl clusters. Pyrolysis and thermolysis of M3(CO)12 have
been thoroughly investigated by Lewis and Johnson,96,97, 98,99 whose work greatly contributed to the
development of the chemistry of high-nuclearity MCCs (see ESI: Tables S3 and S4). The outcome
of these thermal reactions largely depends on temperature, time, atmosphere, and solvent (Scheme
12).98 Their main drawback is that they often lead to product mixtures, which must be separated by
chromatography. In contrast, the reaction of Ru3(CO)12 with H2 (1 atm) in octane at 120 °C
selectively affords H4Ru4(CO)12. Under analogous conditions, Os3(CO)12 is transformed into the
unsaturated H2Os3(CO)10 cluster, whereas higher pressure (60 atm) affords H4Os4(CO)12.99
H4M4(CO)12 and H2Os3(CO)10 are very versatile species for the synthesis of other Ru and Os
clusters.100 The reaction of Os3(CO)12 with stoichiometric amounts of Me3NO in MeCN results in
Os3(CO)12-x(MeCN)x (x = 1, 2).101 The lability of MeCN may be exploited for further
functionalisation102 and applications in biology and medicine.103 Interestingly, pyrolysis of
18
Page 19 of 76 Chemical Society Reviews

Os3(CO)10(MeCN)2 affords different clusters than reactions using Os3(CO)12 (see ESI; Scheme S5).
Moreover, reaction of Os3(CO)10(MeCN)2 with the triosmiumfuryne complex Os3(CO)9(3-2-
C4H2O)(-H)2, which contains a bridging furanyl ligand, produces the bis-triosmium complex
Os3(CO)9(-H)2(3-2-2,3--2-4,5-C4HO)Os3(CO)10(-H)2; CH activation occurs at the
uncoordinated C=C double bond of the furanyl (Scheme 13).104 Further Ru and Os carbonyl clusters
can be obtained by reduction of M3(CO)12, as well as by redox reactions on various preformed
clusters.
Os3(CO)12

200°C, vacuum 270-300°C,

220°C vacuum, H2O
vacuum

Os4(CO)13 HOs3(CO)10(OH) [Os17(CO)36]2-

Os5(CO)16 H2Os4(CO)13 [Os20(CO)40]2-
Os6(CO)18 H2Os5(CO)16
Os8(CO)23 H2Os7C(CO)19
Os8C(CO)21 H2Os8(CO)23

Scheme 12. Synthesis of Os carbonyl clusters by pyrolysis of Os3(CO)12. The obtained products
depend on temperature and/or the addition of some water to solid Os3(CO)12 during the pyrolysis.
When obtained in mixture, different products can be separated by chromatography.98

Os3(CO)10(MeCN)2

O H
Os H
H H Os Os Os
Os Os
H Os Os H
O H Os
Os3(CO)10(MeCN)2
H + O +
O
H Os 80 °C
Os Os H Os
H Os
H Os
Os Os Os Os Os
1 H H H H
Os
2 3 4

Scheme 13. The reaction of Os3(CO)10(MeCN)2 with the triosmiumfuryne complex Os3(CO)9(3-
2-C4H2O)(-H)2. Os3(CO)9(3-2-C4H2O)(-H)2 (1); Os3(CO)9(-H)2(3-2-2,3--2-4,5-
C4HO)Os3(CO)10(-H)2 (2); Os3(CO)9(-H)2(3-2-2,3--2-4,5-C4O)Os3(CO)9(-H)2 (3);
Os3(CO)9(-H)2(3-2-3-2-C-C-C)Os3(CO)9(-H)2 (4).104 The products can be separated by
chromatography. CO ligands have been omitted for clarity.
19
 Chemical Society Reviews Page 20 of 76

Ru and Os carbide carbonyl clusters. Ru and Os carbide carbonyl clusters can be obtained
through CO disproportionation at high temperature.105106107 In particular, Ru6C(CO)17 is
conveniently prepared from the thermal treatment of Ru3(CO)12 at 165 °C under 30 atm of C2H4.
Ru6C(CO)17 is quantitatively transformed into Ru5C(CO)15 after exposure to CO (80 atm).
Ru6C(CO)17 and Ru5C(CO)15 are transformed into [Ru6C(CO)16]2– and [Ru5C(CO)14]2–,
respectively, by Na2CO3 in MeOH. The reaction of Ru3(CO)12 with CaC2 affords the dicarbide
[Ru10(C)2(CO)24]2–,108 whereas the monocarbide [Ru10C(CO)24]2– is obtained from the redox
condensation of [Ru6C(CO)16]2– and Ru3(CO)12.109 In general, Ru carbide clusters may be further
transformed by CO substitution and by the addition of neutral or cationic fragments.110111
Ru and Os carbonyl clusters containing other main-group elements. The preparation of Ru
nitride carbonyl clusters relies on the addition of N3– or NO+ to preformed Ru carbonyls. The
stepwise addition of Ru(CO)2 fragments to [Ru4N(CO)12]– affords the larger [Ru5N(CO)14]–,
[Ru6N(CO)16]–, and [Ru10N(CO)24]– clusters. [Ru6N(CO)16]– is readily prepared from Ru3(CO)12
and N3– at 80 °C. Reactions of [H3M4(CO)12]– with NOBF4 afford H3Ru4N(CO)11 and
[Os4N(CO)12]–.3
Ru and Os carbonyl clusters containing S, Se, and Te may be prepared by various methods
employing the elemental chalcogen, organic chalcogen compounds, or EO32– as sources (Scheme
14).112 The reaction of Ru3(CO)12 with Na2Te2 in water at 110 °C in the presence of [PPh4]Cl
affords the Te-rich [Ru6(Te2)7(CO)12]2– cluster (see ESI: Scheme S6).113 Moreover, reactions of
Ru3(CO)12 with K2TeO3 in the presence of Cu(I) salts under a variety of experimental conditions
result in Te–Ru–Cu carbonyl clusters, which have been recently reviewed.59,60,114

20
Page 21 of 76 Chemical Society Reviews

Os3(3-S)(CO)10

PhSH CO
h, benzene

Os3(CO)12 HOs3(CO)10(-SPh) , benzene

+ Os3(CO)10(MeCN)2

, toluene

Os6(4-S)(CO)17 Os6(3-S)(CO)19

Scheme 14. The reaction of Os3(CO)12 with PhSH (blue, Os; yellow, S; red, O; grey, C; white,
H).112 Hydrogens on organic groups have been omitted for clarity. Light or thermal treatment are
indicated by hν and Δ, respectively.

Similar procedures may be applied to other main-group elements.115 For instance, Ru–Bi
and Os–Bi carbonyl clusters may be obtained by reacting M3(CO)12, as well as other Ru and Os
carbonyls, with NaBiO3. Using THF as solvent, the reaction of [HRu3(CO)11]– with SbPh2Cl affords
Ru3(CO)10(-H)(-SbPh2), whereas using CH2Cl2 results in Ru6(CO)20(-H)2(-SbPh2)2 (see ESI:
Scheme S7).116
Six different Ru–Ge carbonyl clusters have been obtained by heating Ru3(CO)12 and
tBuGeH in heptane (Scheme 15): Ru3(CO)9(3-GetBu)2, Ru2(CO)6(-GetBuH)3, Ru4(CO)10(4-
3

Ge2tBu2)(-GetBuH)2, Ru4(CO)8(4-Ge2tBu2)(-GetBuH)2(3-GetBu)(H), Ru5(CO)12(3-

GetBu)2(4-GetBu)(H), and Ru6(CO)12(3-GetBu)4(H)2.117 Additionally, using different
experimental conditions, Os3(CO)10(3-C6H4), Os3(CO)10Ph(-2-O=CPh), HOs6(CO)20(-2-
C6H4)(4-Bi), Os2(CO)8(-BiPh), and HOs5(CO)18(-2-C6H4)(4-Bi) have been obtained from
Os3(CO)10(MeCN)2 and BiPh3.102

21
 Chemical Society Reviews Page 22 of 76

Ru3(CO)12 + tBuGeH3

Heptane 98 °C

Ru Ru
H
t
H
Bu Ge Ge H
Ru Ge tBu Ge tBu Ge
t
Bu t
Bu
Ru Ru
Ru3(CO)9(3-GetBu)2 Ru2(CO)6(-GetBuH)3

H H H H
Ge Ge
t
Bu Ge Ru t
Bu
t
Bu Ge Ru t
Bu
t t t
Bu Ge Ru Bu Ge Ge Bu
H
Ru Ge t Ru Ru
Bu
Ge t
Ru Ru Bu
t t
Ru4(CO)10(4-Ge2 Bu2)(-Ge BuH)2 Ru4(CO)8(4-Ge2 Bu2)(-GetBuH)2(3-GetBu)(H)
t

t t
Bu Bu t
Bu
Ge Ru Ge Ru Ge
t
Bu Ru Ru
Ru Ge H
H Ge tBu
Ru Ru Ru Ru
H
Ge Ge t
Ru t Ru Bu
Bu
Ru5(CO)12(3-GetBu)2(4-GetBu)(H) Ru6(CO)12(3-GetBu)4(H)2

Scheme 15. The products of the thermal reaction of Ru3(CO)12 and tBuGeH3. The different products
have been separated by thin layer chromatography (TLC).117 CO ligands have been omitted for
clarity.

By reacting carbonyls with boranes, boron may be incorporated into Ru and Os carbonyl
clusters. Heating Ru3(CO)12 in toluene at 75 °C with THF·BH3 affords the octahedral
HRu6B(CO)17, whereas the redox condensation of Ru3(CO)10(MeCN)2 and [Ru3(CO)9(B2H5)]–
yields the trigonal prismatic cluster [H2Ru6B(CO)18]–.118 The reaction of H2Os3(CO)10 with B2H6
produces the carbonylborylidyne cluster H3Os3(CO)9(BCO), which is transformed into a mixture of
HOs4(CO)12(BH2) and HOs5B(CO)12 after heating in toluene at 110 °C.119 The cluster
Os3(CO)10(MeCN)2 may be used for the cage opening of carboranes; its reaction with closo-o-
C2B10H10 yields two interconvertible isomers: Os3(CO)9(3-4,5,9-C2B10H8)(-H)2 and Os3(CO)9(3-
3,4,8-C2B10H8)(-H)2 (Figure 3).120

22
Page 23 of 76 Chemical Society Reviews

Figure 3. Molecular structures of Os3(CO)9(3-4,5,9-C2B10H8)(-H)2 and Os3(CO)9(3-3,4,8-

C2B10H8)(-H)2 (blue, Os; yellow, B; red, O; grey, C; white, H). These two interconvertible isomers
have been obtained from the reaction of Os3(CO)10(MeCN)2 with closo-o-C2B10H10.120

Heterometallic Ru- and Os-based carbonyl clusters. Heterometallic Ru–M and Os–M
carbonyl clusters can be obtained by redox condensation of preformed Ru and Os carbonyls with
metal salts, complexes, and carbonyls. For instance, the reaction of [Ir(CO)4]– with 1 equiv.
(equivalent(s)) of Ru3(CO)12 yields the tetrahedral cluster [IrRu3(CO)13]–, whereas employing 2
equiv. of Ru3(CO)12 results in [IrRu6(CO)23]– (Scheme 16).121 Moreover, the redox condensation of
[Os10C(CO)24]2– with an excess of [Pd(MeCN)4][BF4]2 results in the large cluster,
[Os18Pd3C2(CO)42]2– (see ESI: Figure S3).122

Ru3(CO)12 Ru3(CO)12
[Ir(CO)4]-
1 equiv 2 equiv

[IrRu3(CO)13]-
[IrRu6(CO)23]-
Scheme 16. The reactions of Ru3(CO)12 and [Ir(CO)4]– (yellow, Ir; blue, Ru; red, O; grey, C).
Depending on the stoichiometry of the reaction, [IrRu3(CO)13]– or [IrRu6(CO)23]– are formed.121

2.4 Cobalt Carbonyl Clusters

The commercially available Co2(CO)8 is the best starting material for the preparation of cobalt
carbonyl clusters (see ESI: Table S5). In the solid state, it adopts the bridged Co2(-CO)2(CO)6
structure with C2v symmetry, whereas in solution there is spectroscopic evidence of two additional

23
 Chemical Society Reviews Page 24 of 76

unbridged isomers with D3d and D2d symmetry. Co-crystallization with C60 traps the unbridged D3d
isomer in the Co2(CO)8·C60 adduct.123
Homometallic cobalt carbonyl clusters. The tetrahedral Co4(CO)12 cluster is obtained after
heating Co2(CO)8 in low-polarity solvents such THF, dioxane, or iPrOH. The reaction of Co4(CO)12
with halides or pseudo-halides results in the [Co4(CO)11X]– mono-anions (X = Br, I, SCN).
Octahedral Co6(CO)16 can be prepared by oxidation of [Co6(CO)15]2– under a CO atmosphere.
Co2(CO)8 is quantitatively reduced by Na/naphthalene or Na/Hg to Na[Co(CO)4], which
contains the very versatile cobaltate carbonyl anion. Alternatively, [Co(CO)4]– can be obtained as
the [Co(MeOH)6]2+ salt by disproportionation of Co2(CO)8 in methanol at 50 °C. In both cases, a
variety of quaternary ammonium or phosphonium salts of [Co(CO)4]– are readily prepared by
metathesis of the Na+ or [Co(MeOH)6]2+ cations.
[Co(CO)4]– may be considered a pseudo-halide and forms the HCo(CO)4 hydride, which
possesses acidic character. Well-defined isocyano analogues of HCo(CO)4 have been recently
investigated.124 The reaction of [Co(CO)4]– with M+ (M = Cu, Ag, Au) or Hg2+ in a 2:1
stoichiometric ratio affords the linear [M{Co(CO)4}2]– and Hg{Co(CO)4}2 complexes reminiscent
of [MX2]– and HgX2 (X = Cl, Br, I, CN).125 Transmetalation reactions involving 1:2 molar ratio of
Ln metal (Ln = Yb, Eu) and Hg{Co(CO)4}2 in Et2O afford the isocarbonyl polymeric arrays,
[{(Et2O)3Ln[Co4(CO)11]}, which contain the reduced [Co4(CO)11]2– anion (see ESI: Figure S4).126
Co2(CO)8 undergoes different disproportionation/condensation equilibria depending on the
polarity of the solvent, the presence of base, and the application of vacuum. In very polar solvents
under nitrogen, only disproportionation to Co2+ and [Co(CO)4]– is observed (see ESI: Scheme S8).
In contrast, heating Co2(CO)8 in dry ethanol at 60 °C under vacuum affords the octahedral
[Co6(CO)15]2– cluster as the result of the partial condensation between Co2+ and [Co(CO)4]–. The
stoichiometric reaction between Co2(CO)8 and pyridine in hexane results in the triangular
[Co3(CO)10]– anion.127
[Co6(CO)15]2– is protonated by strong acids to [HCo6(CO)15]–, which contains a fully
interstitial hydride at the centre of the octahedral cage of the cluster. Reduction of [Co6(CO)15]2–
with alkali metals affords the [Co6(CO)14]4– tetra-anion (Scheme 17).

24
Page 25 of 76 Chemical Society Reviews

H+ Na

[HCo6(CO)15]- [Co6(CO)15]2- [Co6(CO)14]4-

Scheme 17. The reactions of [Co6(CO)15]2– with acids and Na (blue, Co; red, O; grey, C; white, H).
Its protonation results in the fully interstitial hydride [HCo6(CO)15]–, whereas the addition of two
electrons promotes the elimination of one CO ligand and affords the [Co6(CO)14]4– tetra-anion. See
Table S5 in ESI for details.

Cobalt carbide carbonyl clusters. Several homometallic Co carbide carbonyl clusters are
known, and most of them can be prepared starting from the trigonal prismatic [Co6C(CO)15]2–
cluster128 via thermal or redox reactions (Scheme 18).48 In addition, Co3(3-CCl)(CO)9 is obtained
by refluxing Co2(CO)8 in CCl4.

[Co11(C2)(CO)22]3–

[Co6C(CO)13]2– Ox
 [Co13C2(CO)24]4– [Co13C2(CO)24]3–
 

Red Redox condensation

[Co7C(CO)15]3– [Co6C(CO)15]2– [Co8C(CO)18]2–

Red Ox
Ox  [Co10(C2)(CO)21]2–
[Co8C(CO)17]4–

Red
[Co6C(CO)14]– [Co9(C2)(CO)19]2–
[Co6C(CO)12]3–
Ox

Red Ox
[Co11C2(CO)23]3– [Co11C2(CO)23]2– [Co11C2(CO)23]–
Scheme 18. Synthesis of Co carbide carbonyl clusters (Ox = oxidation; Red = reduction;  =
thermal reaction). Several homometallic Co carbide carbonyl clusters are prepared by thermal or
redox reactions starting from [Co6C(CO)15]2–.48 See text for details.

25
 Chemical Society Reviews Page 26 of 76

[Co6C(CO)15]2– can be prepared from the reaction of [Co(CO)4]– with CCl4 at a 6:1 molar
ratio or from [Co(CO)4]– and Co3(3-CCl)(CO)9 at a 3:1 molar ratio. By performing the latter
reaction at a 2:1 molar ratio, the paramagnetic [Co6C(CO)14]– cluster is obtained. [Co6C(CO)14]– can
be alternatively obtained by mild oxidation of [Co6C(CO)15]2– or from the reaction of [Co6(CO)15]2–
with MeCOCl via CO scission induced by the addition of MeCO+ to a coordinated CO ligand. In
the presence of even weak bases, [Co6C(CO)14]– is converted to [Co6C(CO)15]2–.
Gently heating [Co6C(CO)15]2– in THF results in an intramolecular condensation to yield
[Co6C(CO)13]2– via the loss of two CO ligands and the rearrangement of the Co6 cage from trigonal
prismatic to octahedral geometry. Under more severe thermal conditions, larger clusters such as
[Co13C2(CO)24]4– are obtained. The stepwise reduction of [Co6C(CO)15]2– with Na/naphthalene
affords, in sequence, the highly reduced clusters [Co7C(CO)15]3– and [Co8C(CO)17]4–, and
eventually, [Co6C(CO)12]3–.48
Several heterometallic Co–M carbide carbonyl clusters have been reported. Although they
will not be reviewed in detail, some general strategies for their synthesis, along with pertinent
examples, are listed below:
1) Reaction of Co3(3-CCl)(CO)9 with metal carbonyl complexes or clusters. Co3(3-
CCl)(CO)9 reacts with [Rh(CO)4]– yielding [Co2Rh4C(CO)13]2–,129 whereas it reacts with
Mo(CO)3(MeCN)3 producing [Mo3Co3C(CO)18]–.130 Interestingly, [Co3Ni9C(CO)20]3– results
from the addition of Co3(3-CCl)(CO)9 to [Ni6(CO)12]2–,131 whereas a different product,
[Co6Ni2C2(CO)16]2–, is obtained by doing the opposite—adding [Ni6(CO)12]2– to Co3(3-
CCl)(CO)9.132 C-C coupling and redox condensation are observed in the reaction of Co3(3-
CCl)(CO)9 with [Ni9C(CO)17]2–, which affords the [Co3Ni7(C2)(CO)15]3– mono-acetylide.133
The Ni–Co hexacarbide carbonyl cluster [H6–nNi22Co6C6(CO)36]n– (n = 3–6) has been
obtained by the redox condensation of [Ni10(C2)(CO)16]2– and Co3(3-CCl)(CO)9 (Scheme
19). Thermolysis of [H2Ni22Co6C6(CO)36]4– in THF affords the larger octa-carbide
[Ni36Co8C8(CO)48]6–.39,134

26
Page 27 of 76 Chemical Society Reviews

Co3(3-CCl)(CO)9 [Ni10(C2)(CO)16]2-

THF, reflux [H2Ni22Co6C6(CO)36]4-

[Ni36Co8C8(CO)48]6-

Scheme 19. Synthesis of [H2Ni22Co6C6(CO)36]4– and [Ni36Co8C8(CO)48]6– (green, Ni; blue, Co;
yellow, Cl; red, O; grey, C). Redox condensation of [Ni10(C2)(CO)16]2– and Co3(3-CCl)(CO)9
affords [H2Ni22Co6C6(CO)36]4–, which is thermally converted into [Ni36Co8C8(CO)48]6–.39,134

2) Redox condensation involving a Co carbide carbonyl cluster and a metal complex, salt or
cluster. Some recent examples are represented by the reactions of [Co6C(CO)15]2– with
Pt(Et2S)2Cl2 and with Pd(Et2S)2Cl2, which result in the formation of [Co8Pt4C2(CO)24]2– 135
and [H6–nCo20Pd16C4(CO)48]n– (n = 3–6), respectively (Scheme 20).50-52

27
 Chemical Society Reviews Page 28 of 76

Pd(Et2S)2Cl2 HBF4
[Co6C(CO)15]2-
CH2Cl2

[H6-nCo20Pd16C4(CO)48]n- [H3-nCo15Pd9C3(CO)38]n-
(n = 3-6) (n = 2, 3)
HBF4
CH2Cl2

HBF4 excess
CH2Cl2

[Co13Pd3C3(CO)29]- [H3-nCo15Pd9C3(CO)38]n-
(n = 0, 1)
Scheme 20. Synthesis and reactivity of [H6–nCo20Pd16C4(CO)48]n– (n = 3–6) (orange, Pd; blue, Co;
red, O, grey, C). This compound is obtained from the redox condensation of [Co6C(CO)15]2– and
Pd(Et2S)2Cl2. [H6–nCo20Pd16C4(CO)48]n– (n = 3–6) is transformed into [H3–nCo15Pd96C3(CO)38]n– (n
= 0–3) and, eventually, [Co13Pd3C3(CO)29]– upon reaction with increasing amounts of strong
acids.50-52

3) Addition of cationic metal fragments to an anionic Co carbide carbonyl cluster anion with
the formation of a Lewis-type acid–base adduct. This method is represented well by the
reaction of Au(PPh3)Cl with [Co6C(CO)15]2–, which affords several Co carbide carbonyl
clusters decorated by AuPPh3 fragments, depending on the experimental conditions (see
ESI: Scheme S9).136

Cobalt carbonyl clusters containing other main-group elements. The [Co6N(CO)15]–

nitride, which is isostructural to [Co6C(CO)15]2–, can be obtained from [Co6(CO)15]2– and
[NO][BF4]. Removal of two CO ligands from the trigonal prismatic [Co6N(CO)15]– by means of
Me3NO affords the octahedral nitride [Co6N(CO)13]–. Redox condensation between [Co6N(CO)15]–
and [Co(CO)4]– results in [Co7N(CO)15]2–, whereas the treatment of [Co6N(CO)15]– with OH– ions
28
Page 29 of 76 Chemical Society Reviews

leads to the formation of the di-nitride [Co10N2(CO)19]4–. Thermal treatment of [Co6N(CO)15]– in

diglyme at 140–150 °C results in the tri-nitride [Co14N3(CO)26]3–, which can be transformed into
[Co13N2(CO)24]3– by heating at 100 °C in a water solution buffered at pH 11.137
The incorporation of other main-group elements into Co carbonyl clusters usually involves
starting with suitable anions, such as [Co(CO)4]–, [Co6(CO)15]2–, and even the neutral Co2(CO)8, in
the presence of a main-group compound, which is often a halide, EXn. This method is exemplified
well by the synthesis Co phosphide carbonyl clusters (Figure 4), which strongly depends on the
precursors and experimental conditions. The reaction of [Co(CO)4]– with PCl3 affords
[Co6P(CO)16]–, whereas using PCl5 results in [Co9P(CO)21]2–. The triangular Co3P(CO)9 cluster can
be prepared from [Co6(CO)15]2– and PBr3, whereas [HCo10P2(CO)23]– has been isolated mixing
[Co(CO)4]–, [Co6(CO)15]2–, and PBr3 at a 4:1:2 molar ratio.138 Cluster expansion may be achieved
by redox condensation, as exemplified by the synthesis of [Co10P(CO)22]3– from [Co9P(CO)21]2– and
[Co(CO)4]–.139 Alternatively, P-atoms may be directly introduced as in the reaction between
Co2(CO)8 and W(CO)4(PH3)2, which results in the formation of Co8P2(CO)19 or Co10P2(CO)24 in a
mixture with [{Co3(CO)8{4-PW(CO)5}}{(4-P)Co3(CO)9}], depending on the stoichiometric
ratio.140

(a) (b) (c)

Figure 4. Molecular structures of (a) [Co10P(CO)22]3–, (b) [HCo10P2(CO)23]– (the hydride ligand has
not been located by X-ray crystallography and, thus, is not represented in the figure), and (c)
Co10P2(CO)24 (blue, Co; orange, P; red, O; grey, C).138-140

2.5 Rhodium and Iridium Carbonyl Clusters

Rh4(CO)12 and Ir4(CO)12 are the major carbonyl species isolated following the reductive
carbonylation of Rh and Ir salts at atmospheric pressure.31-33 The dinuclear species M2(CO)8 (M =
Rh, Ir) have been observed only by spectroscopic methods at very high CO pressures or in matrix
experiments at low temperatures.141 Rh4(CO)12 spontaneously loses CO under N2 atmosphere

29
 Chemical Society Reviews Page 30 of 76

affording Rh6(CO)16. Meanwhile, the analogous Ir6(CO)16 is conveniently obtained from the
oxidation of [Ir6(CO)15]2–.142
Homometallic Rh and Ir carbonyl clusters. Neutral Rh and Ir carbonyls are suitable
precursors for the synthesis of larger clusters (see ESI: Tables S6 and S7). Moreover, CO ligands
can be replaced by phosphines or other soft nucleophiles, resulting in heteroleptic clusters.
Rh4(CO)12 and Rh6(CO)16 disproportionate in the presence of bases, such as amines. For instance,
the ionic couple [Rh(CO)2(py)]+[Rh5(CO)13(py)2]– is formed in pyridine.143 The most powerful
technique for obtaining anionic Rh and Ir carbonyl clusters involves the reduction of M4(CO)12 (M
= Rh, Ir).144 Several different species may be obtained, depending on the nature (alkali metals or
alkali hydroxides) and amount of reducing agent, the temperature, and the reaction atmosphere (N2,
CO, H2, vacuum).
The reduction of Rh4(CO)12 with increasing amounts of NaOH or KOH under CO at
atmospheric pressure affords [Rh12(CO)30]2–, [Rh6(CO)15]2–, [Rh7(CO)16]3–, [Rh4(CO)11]2–, and
[Rh(CO)4]–. Again with increasing base, the reduction of the analogous Ir4(CO)12 produces
[HIr4(CO)11]–, [Ir4(CO)11]2–, [HIr4(CO)10]3–, [Ir4(CO)10]4–, and [Ir(CO)4]– (see ESI: Scheme S10).33
Careful dosage of NaOH during thermolysis of Rh4(CO)12 under different atmospheres enables the
more or less selective preparation of several high-nuclearity Rh carbonyl clusters (see ESI: Scheme
S11), including [H5-nRh13(CO)25]n– (n = 2, 3), [Rh22(CO)37]4–, [H8-nRh22(CO)35]n– (n = 4, 5),
[Rh19(CO)31]5–, and [Rh33(CO)47]5– (Figure 5).145146 Similarly, the base-induced condensation of
Ir4(CO)12 may lead to products such as [Ir8(CO)22]2–, [Ir9(CO)20]3–, [HIr9(CO)19]4–, and
[Ir10(CO)21]2–, depending on the solvent and reaction conditions.147 Moreover, [Ir6(CO)15]2– may be
obtained from the reductive carbonylation of K2IrCl6 in 2-methoxy-ethanol/water or the redox
condensation of Ir4(CO)12 and [Ir(CO)4]–; this is a valuable precursor for the preparation of larger Ir
clusters. For instance, oxidation of [Ir6(CO)15]2– with [Cp2Fe]+ (Cp = cyclopentadienyl) affords
[Ir14(CO)27]– (Figure 5), which is the largest homometallic Ir carbonyl cluster reported to date.148

30
Page 31 of 76 Chemical Society Reviews

(a) (b)
Figure 5. Molecular structures of (a) [Rh33(CO)47]5– and (b) [Ir14(CO)27]–, the largest homometallic
homoleptic Rh and Ir carbonyl clusters (yellow, Ir; blue, Rh; red, O; grey, C).145-147

The robustness of the Rh–Rh and Ir–Ir bonds also allows the isolation of unusual species,
such as the -1-1-peroxo Ir4(CO)5(PR3)3(O2)2 149 and the Rh6(CO)12(3-GaCp*)4 and Rh6(CO)16-
x(3-InCp*)x (x = 1, 2) derivatives containing ECp* ligands (E = Ga, In; Cp* =
pentamethylcyclopentadienyl) (Figure 6).150

(a) (b)
Figure 6. Molecular structures of (a) Rh6(CO)12(3-GaCp*)4 and (b) Rh6(CO)15(3-InCp*) (blue,
Rh; yellow, Ga (a) or In (b); red, O; grey, C; white, H).150 These compounds have been obtained
upon reactions of Rh6(CO)16 with GaCp* and Rh6(CO)15(MeCN) with InCp*, respectively (Cp* =
pentamethylcyclopentadienyl).

Rhodium carbonyl clusters containing main-group elements. Rh forms several carbonyl

clusters containing interstitial heteroatoms (E = C, N, P, S, Ge, As, Sn, Sb, Bi), whereas the
analogous interstitial heteronuclear Ir carbonyl clusters are not known. This is due to the fact that
Ir–Ir bonds are stronger than Rh–Rh bonds. Consequently, Ir clusters prefer to maximise Ir–Ir
31
 Chemical Society Reviews Page 32 of 76

contacts, rather than allowing interstitial heteroatoms to replace them with Ir–E bonds. Indeed, as
also observed for Pt, a few heteronuclear Ir carbonyl clusters have been reported, but they contain
an iridium core with the heteroatoms on the surface.151152 An exciting advancement in this field is
the recent discovery of the 15-vertex deltahedral cluster [(3-Bi3)2(IrCO)6(4-Bi)3]3–, which is
obtained from Ir(CO)2(acac) (acac = acetylacetonate) and K5Bi4 (Scheme 21).153

Ir(CO)2(acac) + K5Bi4

[(3-Bi3)2(IrCO)6(4-Bi)3]3-

Scheme 21. Synthesis of the 15-vertex deltahedral cluster [(3-Bi3)2(IrCO)6(4-Bi)3]3– (yellow, Ir;
orange, Bi; red, O; grey, C).153

A general strategy for the synthesis of Rh carbonyl clusters containing heavy main-group
elements such as Ge, Sn, Sb, and Bi is represented by the reaction of [Rh7(CO)16]3– with EXn
compounds.11,38,154155 Conversely, lighter elements, P and As, have been introduced into
[Rh10As(CO)22]3–,156 [Rh10P(CO)22]3–, and [Rh9P(CO)21]2–157 by reductive carbonylation of Rh(I)
complexes at high pressure and temperature in the presence of EPh3 (E = P, As). [Rh10S(CO)22]2–
was obtained from Rh4(CO)12 and SCN–.158
[Rh6C(CO)15]2– may be obtained from the reductive carbonylation of RhCl3 at atmospheric
pressure and RT in MeOH in the presence of CHCl3, or from the reaction of [Rh(CO)4]– and
CCl4.159 Refluxing the trigonal prismatic [Rh6C(CO)15]2– in iPrOH generates the octahedral
[Rh6C(CO)13]2– by the removal of two CO ligands.41 The self-assembly of [Rh6C(CO)15]2– with Ag+
ions affords oligomeric and polymeric species, such as [Ag{Rh6C(CO)15}2]3– and
[{AgOC4H8{Rh6C(CO)15}AgOC4H8}pyz] (pyz = pyrazine).160 Oxidation of [Rh6C(CO)15]2– with
Fe3+ ions results in Rh12C2(CO)25, whereas heating [Rh6C(CO)15]2– at 70 °C in the presence of
H2SO4 produces [Rh12C2(CO)24]2–. This latter cluster may be further transformed into
[Rh12C2(CO)23]4– by treatment with alkali hydroxides or into [Rh12C2(CO)20{Au(PPh3)}4] and
[Rh12C2(CO)18{Au(PPh3)}4] after reaction with Au(PPh3)Cl. In turn, several multivalent clusters
based on the Rh10(C)2Au4-6 framework may be obtained via the reduction of
[Rh12C2(CO)20{Au(PPh3)}4].161
32
Page 33 of 76 Chemical Society Reviews

The chemistry of Rh nitride carbonyl clusters is quite rich. [Rh6N(CO)15]3– can be prepared
from the reaction of [Rh6(CO)15]2– with NOBF4, or by treating [Rh7CO)16]3– with base under
NO/CO atmosphere.3 [Rh6N(CO)15]3– undergoes nucleophilic attack when reacted with OH–,
affording the hydride [HRh6N(CO)14]3–.162 Thermal treatment of [Rh6N(CO)15]3– under different
experimental conditions (e.g., temperature, solvent, time, pH) results in larger nitride and
polynitride clusters, such as [HRh12N2(CO)23]3–, [HRh12N2(CO)24]–, [Rh12N2(CO)24]2–,
[Rh14N2(CO)25]2–, [Rh23N4(CO)38]3–, and [H6-nRh28N4(CO)41]n– (n = 4, 5).163 Redox condensation of
[Rh6N(CO)15]3– with [PtRh4(CO)14]2– and with [M(CO)4]– (M = Co, Ir) results in the heterometallic
nitrides [PtRh10N(CO)21]3– and [MRh6N(CO)15]2–, respectively.164
Anionic Rh and Ir carbonyl clusters may be used in combination with metal salts and
complexes for the preparation of heterometallic clusters via redox condensation.

2.6 Nickel Carbonyl Clusters

Ni(CO)4 was the first metal carbonyl to be discovered and is nowadays produced at industrial
quantities for the obtainment of ultra-pure Ni metal. Its main drawback is the fact that Ni(CO)4 is
highly toxic and carcinogenic; therefore, it must be handled with care.
Homometallic nickel carbonyl clusters. The reduction of Ni(CO)4 with base was
investigated first by Hieber,165 and then, rationalised in the seminal work of Chini and Longoni (see
ESI: Table S8).166167 The most suitable and versatile precursor cluster for Ni carbonyl cluster
chemistry is [Ni6(CO)12]2–, which can be prepared by the reduction of Ni(CO)4 using a variety of
experimental conditions. The most efficient synthesis for a gram-scale yield is the reaction of
Ni(CO)4 with alkali hydroxides (NaOH, KOH) in DMF or DMSO followed by the addition of water
(Scheme 22). In this reaction, the main product of the reduction is initially [Ni5(CO)12]2–, and water
is required to oxidise it to [Ni6(CO)12]2–. It must be remarked that if aqueous NH4Cl is present,
[HNi12(CO)21]3– is produced rather than the hydrolysis to [Ni5(CO)12]2–; if H3PO4 is present, the
dihydride [H2Ni12(CO)21]2– is formed. These hydrides [HNi12(CO)21]3– and [H2Ni12(CO)21]2– can
easily interconvert to one another as simple acid–base reactions. In contrast, their full deprotonation
to [Ni12(CO)21]4– occurs only under very drastic basic conditions. [Ni5(CO)12]2– can be prepared in
pure form by treating [Ni6(CO)12]2– with CO at atmospheric pressure. [Ni5(CO)12]2– is readily
oxidised back to [Ni6(CO)12]2– simply using water, and thus, it must be handled under very
anhydrous conditions.
[Ni9(CO)18]2– can be obtained by redox condensation of [Ni6(CO)12]2– with Ni(CO)4, or
oxidation of [Ni6(CO)12]2– with Ni2+, Fe3+ or H+ ions. In the latter case, the reaction must be
carefully controlled in order to avoid the formation of [HNi12(CO)21]3– and [H2Ni12(CO)21]2–.

33
 Chemical Society Reviews Page 34 of 76

+ NaOH +H2O
Ni(CO)4 [Ni5(CO)12]2- [Ni6(CO)12]2-
+ CO
+H+
+NH4Cl
+H3PO4 +H2O
[Ni9(CO)18]2-
+H2O
+H+

+Base +Base
[H2Ni12(CO)21]2- +
[HNi12(CO)21]3- [Ni12(CO)21]4-
+H +H+

Scheme 22. Synthesis of homoleptic homometallic Ni carbonyl clusters (green, Ni; red, O; grey, C).
The reaction of Ni(CO)4 with NaOH in DMF or DMSO affords [Ni5(CO)12]2–, whose hydrolysis
results in [Ni6(CO)12]2–, [HNi12(CO)21]3– or [H2Ni12(CO)21]2– depending on the pH.166,167

Heterometallic Ni-based carbonyl clusters. [Ni6(CO)12]2– can undergo redox condensation

with metal salts, metal complexes, MCCs, and main-group compounds in order to obtain a large
variety of heterometallic MCCs and MCCs containing main-group elements. The reaction fails only
with very electropositive elements, such as alkali and earth alkali metals, Al, Zn, early transition
metals, and lanthanides. In these cases, a reaction is not observed for elements that are weak acids.
In contrast, oxidation of [Ni6(CO)12]2– to [Ni9(CO)18]2– occurs with stronger acids such as Zn2+, due
to the hydrolysis of H2O and the oxidation of the cluster via the H+/H2 redox couple. When the
electropositive metal is a soft Lewis acid, such as Cd(II) and In(III), the formation of Lewis acid–
base adducts comprising the cluster anion and the metal salt is observed rather than the redox
condensation. As a result, the reaction of [Ni6(CO)12]2– with InBr3 yields [Ni6(CO)11(3-InBr3)(4-
In2Br5)]3–, [Ni6(CO)10(4-In2Br5)2]4–, and [{Ni6(CO)11}2(6-In){6-In2Br4(OH)}]4–, which contain
[Ni6(CO)11]4– and [Ni6(CO)10]6– cores decorated by InBr3, [In2Br5]+, In3+, and [In2Br4(OH)]+
fragments.168 Moreover, the reaction of [Ni6(CO)12]2– with CdCl2 results in the formation of the
Lewis acid–base adduct {Cd2Cl3[Ni6(CO)12]2}3–.169
In contrast, less electropositive transition metals can be incorporated into heterometallic Ni–
M clusters via redox condensation. Limiting our scope to homoleptic carbonyls, these include Cr,
Mo, W, Fe, Ru, Os, Co, Rh, Ir, Pd, Pt, Cu, Ag, and Au (Figure 7; see also ESI: Table S9).170171172173
34
Page 35 of 76 Chemical Society Reviews

The highest nuclearities, which are obtained for Ni–Pd and Ni–Pt clusters, can exceed 40 metal
atoms.

(a) (b)

(c)

Figure 7. The molecular structures of (a) [Ni22-xPd20+x(CO)48]6– (x = 0.62) (green, Ni; purple, Pd;
yellow, Ni/Pd  16:84; blue, Ni/Pd  85:15) (b) [Ni32Au6(CO)44]6– (green, Ni; yellow, Au), (c)
[Ni32Pt24(CO)56]n– (green, Ni; purple, Pt; blue, Ni/Pt  66:34). In all structures, grey and red
represent C and O, respectively. These are representative high nuclearity heterometallic Ni-M
carbonyl clusters obtained from redox condensation of [Ni6(CO)12]2– with metal salts and
complexes.170-173

Nickel carbonyl clusters containing main-group elements. A similar trend has been
observed with post-transition metals; redox condensation yielding heteronuclear Ni–E clusters has
been observed with Ga, Ge, Sn, Sb, and Bi, whereas Lewis-type acid–base adducts are found for In,
which behaves like Cd. In addition, non-metallic main-group elements, such as C, P, As, Se, and
Te, can be included in the metal cage of Ni carbonyl clusters.
Several Ni carbide carbonyl clusters may be obtained from reactions of [Ni6(CO)12]2– with
halocarbons (CCl4, C2Cl4, C2Cl6, C3Cl6, C4Cl6); they contain isolated C atoms or tightly bonded C2
35
 Chemical Society Reviews Page 36 of 76

units (see ESI: Table S10). Ni carbide carbonyl anions can be further reacted with metal salts or
complexes, affording heterometallic Ni–M carbide carbonyl clusters.174175176177178 Alternatively,
such compounds can be prepared starting from a preformed M carbide carbonyl and a Ni (carbonyl
or non-carbonyl) compound.
The reaction of [Ni6(CO)12]2– with GaCl3 in CH2Cl2 under nitrogen atmosphere results in a
mixture of [Ni12+xGa(CO)22+x]3– (x = 0–3) clusters, which is transformed into [Ni12Ga(CO)22]3– after
exposure to CO atmosphere (see ESI: Figure S5).179 Several Ni–Sb carbonyl clusters can be
obtained by the redox condensation of [Ni6(CO)12]2– with SbCl3 using various experimental
conditions (see ESI: Figure S5).180 Moisture must be avoided in order to prevent hydrolysis of the
ECl3 reactant. Indeed, anhydrous PCl3 and POCl3 are required for obtaining the Ni–P carbonyl
clusters [Ni11P(CO)18]3–, [Ni14P2(CO)22]2–, [Ni22-xP2(CO)29-x]4– (x = 0.84), [Ni22P6(CO)30]2–, [Ni23-
4– (x = 0.84), [H6–nNi31P4(CO)39]n– (n = 4, 5), and [Ni39P3(CO)44]6– (Figure 8).181
xP2(CO)30-x]

(a) (b)

(c) (d)

Figure 8. The molecular structures of (a) [Ni14P2(CO)22]2–, (b) [Ni22P6(CO)30]2–, (c) [H6–
n– (n = 4 and 5), and (d) [Ni39P3(CO)44]6– (green, Ni; purple, P; grey, C; red, O).181
nNi31P4(CO)39]

These have been obtained from the reactions of [Ni6(CO)12]2– with PCl3 or POCl3 under different
experimental conditions (e.g., stoichiometry, solvent).
36
Page 37 of 76 Chemical Society Reviews

2.7 Platinum Carbonyl Clusters

Platinum carbonyl clusters have been recently reviewed.8,10 Chini clusters of the general formula
[Pt3n(CO)6n]2– (n = 1–10) represent a milestone in inorganic and cluster chemistry.34 As
demonstrated by Chini and Longoni, they can be obtained by reductive carbonylation of Na2PtCl6 in
MeOH at RT and atmospheric pressure of CO. Modulating the nature and amount of base, it is
possible to control the nuclearity of the [Pt3n(CO)6n]2– (n = 1–10) clusters (Figure 9).182183184 Chini
clusters have been used for the preparation of several homoleptic, heteroleptic, and heterometallic
carbonyl clusters, as well as for the preparation of metal nanoparticles, nanostructured materials,
and heterogeneous catalysts.8,10

Na2PtCl6
+ NaOH or CH3COONa CO (1 atm) / CH3OH / RT

Increasing amount or strength of the base

OX
OX
RED
RED OX
OX
RED
OX
RED
RED OX

RED
Figure 9. Synthesis of Chini clusters, [Pt3n(CO)6n]2– (n = 2–8), using reductive carbonylation of
Na2PtCl6 and their redox interconversion (purple, Pt; red, O; grey, C).8,10,182,183,184 OX and RED
represent oxidation and reduction, respectively.

The reactions of Chini clusters may be grouped into two main categories, depending on
whether the resulting clusters do (A) or do not (B) retain the trigonal prismatic structure of the
parent species (Table 1 and Figure 10).

37
 Chemical Society Reviews Page 38 of 76

Table 1. Categorising the reactions of Chini clusters.

(A) The product retains the trigonal (B) The product does not retain the trigonal
prismatic structure of the Chini Cluster. prismatic structure of the Chini Cluster.
(A-1) Redox reaction (B-1) Redox fragmentation
(A-2) CO/Phosphine substitution (B-2) Thermal reaction
(A-3) Lewis acid–base adduct formation (B-3) Formation of surface-decorated clusters
(B-4) Formation of heterometallic clusters

Figure 10. Illustrating the general reaction categories of Chini clusters.

Redox reactions of Chini clusters (A-1). Chini clusters may be easily and reversibly
interconverted by means of redox reactions (A-1), with full retention of their trigonal prismatic
structures. Reduction of [Pt3n(CO)6n]2– under CO atmosphere results in the gradual decrease of
cluster nuclearity, whereas oxidation leads to higher nuclearities (Figure 9). The oxidation process
eventually results in insoluble, infinite, and conductive molecular Pt carbonyl wires composed of
continuous stacks of Pt3(-CO)3(CO)3 units.182-184
Reactions of Chini clusters with soft nucleophiles: Non-redox substitution (A-2) and
redox fragmentation (B-1). The reactions of Chini clusters with soft nucleophiles may result in
heteroleptic Chini clusters via non-redox substitution (A-2) or in zero-valent neutral Pt–(CO)–L
species via redox fragmentation (B-1), where L is a phosphine-based ligand (Figure 11). The
outcome of the reaction depends on the nature of the nucleophile, the nuclearity of the Chini cluster,
and the stoichiometry of the reaction. In the case of phosphines, non-redox substitution is favoured
by the stoichiometric addition of the phosphine to lower-nuclearity Chini clusters, whereas redox
fragmentation is observed for larger clusters and in the presence of excess of ligand.185186187
38
Page 39 of 76 Chemical Society Reviews

Redox Fragmentation Non-Redox Substitution

Pt4(CO)4(P^P)2 [Pt12(CO)22(PPh3)2]2-
L = P^P
L = PPh3

L = dppm +L +L L = PTA

L = dppBz
CO L = dppe
Pt6(CO)6(dppm)3 [Pt12(CO)24]2-
[Pt12(CO)20(PTA)4]2-

[Pt9(CO)18]2-

[Pt12(CO)20(dppe)2]2-
Pt(dppBz)2

Figure 11. Reactions of Chini clusters with phosphines (L) include redox fragmentation (B-1) and
non-redox substitution (A-2). P^P = H2C=C(PPh2)2; dppm = CH2(PPh2)2; dppBz = 1,2-C6H4(PPh2)2;
PTA = C6H12N3P; dppe = Ph2PCH2CH2PPh2. In the examples reported in this Figure, 1-4 CO
ligands of [Pt12(CO)24]2– can be substituted with phosphines leading to heteroleptic Chini-type
clusters via non-redox substitution.185-187 Alternatively, redox fragmentation leads to the more
reduced homoleptic [Pt9(CO)18]2– cluster and neutral complexes containing Pt(0).

Formation of Lewis acid–base adducts based on Pt Chini clusters (A-3). The external
triangular faces of Chini clusters have been predicted to behave as Lewis bases. Nonetheless, this
feature has only been exploited in two cases, both of which produce Lewis acid–base adducts (A-3).
The reaction of [Pt9(CO)18]2– with CdCl2 affords 1-D {[Pt9(CO)18(3-CdCl2)2]2–}∞ superwires,188
and the reactions of [Pt3n(CO)6n]2– (n = 2, 3) with Ag(IPr)Cl (IPr = C3N2H2(C6H3iPr2)2) result in the
neutral adducts [Pt3n(CO)6n(AgIPr)2] (see ESI: Figure S6).189 Otherwise, the reactions of Chini
clusters with other Lewis acids proceed via oxidation to [Pt3(n+1)(CO)6(n+1)]2– species (A-1) or
formation of surface-decorated clusters (B-3).
Surface-decorated Pt carbonyl clusters (B-3). This reaction path may be considered as an
oxidation accompanied by the loss of CO ligands, resulting in the formation of a Ptn(CO)m core with
MXx fragments decorating its surface. For instance, [Pt3n(CO)6n]2– (n = 2–5) clusters react with
SnCl2 resulting in [Pt8(CO)10(SnCl2)4]2–, [Pt5(CO)5{Cl2Sn(OR)SnCl2}3]3– (R = H, Me, Et, iPr),
39
 Chemical Society Reviews Page 40 of 76

[Pt6(CO)6(SnCl2)2(SnCl3)4]4–, [Pt9(CO)8(SnCl2)3(SnCl3)2(Cl2SnOCOSnCl2)]4–, and

[Pt10(CO)14{Cl2Sn(OH)SnCl2}2]2– (Figure 12).190191192 Moreover, heating [Pt3n(CO)6n]2– (n = 2–10)
in the presence of CdBr2·nH2O affords the larger [Pt13(CO)12Cd10Br14(DMF)6]2–,
[Pt19(CO)17Cd10Br14(DMF)6]2–, and [H2Pt26(CO)20(CdBr)12]8– clusters (Figure 12).193

(a) (b) (c)

(d) (e)
Figure 12. Some examples of surface-decorated Pt carbonyl clusters produced from Chini clusters
via reaction pathway B-3: (a) [Pt6(CO)6(SnCl2)2(SnCl3)4]4–, (b) [Pt8(CO)10(SnCl2)4]2–, (c)
[Pt10(CO)14{Cl2Sn(OH)SnCl2}2]2–, (d) [Pt13(CO)12Cd10Br14(DMF)6]2–, (e) [H2Pt26(CO)20(CdBr)12]8–
(purple, Pt; yellow, Cd; orange, (a-c) Sn and (d,e) Br; blue, N; red, O; grey, C; white, H).190-193

Thermal reactions of Chini clusters (B-2). Thermal treatment of Chini clusters under
controlled conditions causes CO loss and condensation via the formation of additional Pt-Pt bonds
(B-2), eventually leading to Pt–CO globular molecular nanoclusters, often referred to as Pt browns
due to their colour in solution.47194195 The full list of homoleptic homometallic Pt carbonyl clusters
(including Pt brown) that are structurally characterised to date is available in the ESI (Table S11);
some representative examples are given in Figure 13.

40
Page 41 of 76 Chemical Society Reviews

(a) (b) (c)

(d) (e) (f)

Figure 13. The molecular structures of some platinum browns: (a) [Pt14(CO)18]4– (bcc), (b)
[Pt40(CO)40]6– (bcc), (c) [Pt19(CO)22]4– (pp), (d) [Pt26(CO)32]– (hcp), (e) [Pt38(CO)44]2– (ccp) and (f)
[Pt44(CO)45]n– (ccp/hcp) (purple, Pt; red, O; grey, C). The structure of the metal core of the cluster is
given in parentheses: bcc, body-centred cubic; pp, pentagonal prismatic; hcp, hexagonal close
packed; ccp, cubic close packed. All these clusters have been obtained from thermal decomposition
(B-2) of Chini clusters (see ref. 194, 195 and Table S11 in ESI).

Heterometallic clusters (B-4). Heteronuclear Pt–M carbonyl clusters can be prepared by the
redox condensation of Chini clusters with metal salts, complexes, or carbonyls (B-4). Alternatively,
as described in the previous sections, such Pt–M clusters can be obtained from the redox
condensation of carbonyl anions of the second metal and Pt salts or complexes. For instance,
[Co8Pt4C2(CO)24]2– can be synthesised from [Co6C(CO)15]2– and [Pt6(CO)12]2– or more
conveniently, from [Co6C(CO)15]2– and Pt(Et2S)2Cl2.135

3. Metal Carbonyl Clusters: Catalysis

3.1 General Principles
This Section will discuss catalysis performed by MCCs. We will give an overview of homogeneous
catalysts and their proposed mechanisms in Section 3.2 and follow this with a discussion of
homogeneous electrocatalysis and electrochemical mechanistic studies in Section 3.3. Most
examples of MCCs in catalysis come from Groups 8 – 10. This is most likely because those clusters
are the most stable under a variety of reaction conditions. A common mechanistic feature observed
41
 Chemical Society Reviews Page 42 of 76

for both the homogeneous catalysis and electrocatalysis promoted by MCCs is the ability of
substrates, CO ligands, and hydride ligands or protons to migrate around the metal–metal bonds of
the cluster core. In most cases, careful probes for the formation of nanoparticles, using Hg
poisoning196 or measurements of solution dispersivity,197,198 are required to ensure that nanoparticle
side-products are not responsible for observed reactivity.
The geometric and electronic structures of MCCs (Figure 14) fall between those of single-
site metal coordination complexes (molecular catalysts) and those of extended solids
(heterogeneous catalysts). When discussing heterogeneous materials, we use a band diagram as a
model for the energy levels that make up the valence band and conduction band. When we talk
about molecular catalysts, we refer to a molecular orbital diagram. However, as molecules become
larger as in the case of clusters, their molecular orbitals become closer in energy and approach the
band-diagram model. The size and electronic structure of both metal clusters and nanomaterials are
found between those of single-site metal coordination complexes and of heterogeneous materials.
When MCCs are large enough, it has been demonstrated that their properties—such as diffusion
coefficients in solutions and even their size—more closely resemble those of a nanoparticle,
quantum dot, or fullerene than that of a molecule. Many of the more convincing links between the
structure and reactivity of MCCs with those of nanomaterials have been elucidated from
comparisons of their electrochemical properties, e.g., electron-transfer rate constants. Thus, one
motivation for studying MCCs in thermally driven catalysis and electrocatalysis is that we can
characterise their properties using the tools of molecular chemistry, which include single-crystal X-
ray diffraction (SC-XRD), NMR, and IR spectroscopy; these provide atom-level insights into
reactivity, which can also inform our understanding of the reactivity and properties of
nanomaterials.

42
Page 43 of 76 Chemical Society Reviews

Figure 14. Relative energy spacings in the electronic structures of a single-site metal coordination
complex (molecular catalyst), MCC, nanomaterial, and extended solid (heterogeneous material or
electrocatalysts). MCC = metal carbonyl cluster. Reproduced with permission from Ref. 238.

Unique to the homogeneous catalysis of MCCs, the MCC in many cases—but not all—
serves as a catalyst precursor, and therefore, debates regarding the active catalyst remain in many
examples. Fragmentation of the cluster to form a mononuclear complex is one possibility when
ligands are added as co-catalyst and when they include protons, phosphine, silane, and MeCN.
While this article focuses on carbonyl clusters, clusters where one or two of the carbonyl ligands are
replaced by another ligand, such as phosphine or amine, are also discussed. Fragmentation of
clusters was well-characterised in some mechanistic proposals for homogeneous catalysis with
MCCs,199 and it is consistent with the fact that most reactions catalysed by MCCs are thermally
initiated. In the following discussion of homogeneous catalysis by MCCs, we include clusters of
Groups 8, 9, and 10 that are accompanied by various additives used to initiate the catalytic reaction.
Many examples of stoichiometric reaction chemistry have been explored on the MCC framework,
and a review of that work falls outside the scope of this manuscript.77,78,200 MCCs have also served
as precursors to heterogeneous catalysts, which also falls outside the scope of this review article.
Electrocatalytic reactions using MCCs has seen increased interest since about 2011, and
efforts have focused on reduction chemistry—primarily hydride-transfer chemistry—using Fe and
Co clusters containing interstitial atoms. These clusters represent some of the most stable known
species in the transition series. Consequently, their product profiles and the mechanistic details of
their catalysis have been thoroughly characterised, primarily employing the electrochemical
technique of cyclic voltammetry (CV). The two-electron reduction of two protons to yield H2 is the
simplest reaction performed by Fe and Co MCCs. This simple reaction has enabled a mechanistic
study for understanding hydride formation and hydride-transfer reactivity as mediated by MCCs.
The resulting insights have made it possible to tailor hydride transfer to other substrates, such as
CO2 for formate production.

3.2 Homogeneous Catalysis by MCCs

The earliest work on catalysis by MCCs focused on syngas as the substrate. Syngas comprises CO,
CO2, H2, and H2O in various ratios, as produced by petroleum refining processes. Due to the
abundance of CO in syngas, it serves well as a source of CO during catalytic reactions; the CO can
contribute to MCC stability, acting as a substrate for CO insertion reactions and hydroformylation,
or it can regenerate the MCC in a catalytic cycle where one of its CO ligands was consumed. CO-

43
 Chemical Society Reviews Page 44 of 76

containing headspaces have also been shown to enhance MCC stability in electrocatalytic reactions
promoted by MCCs (vide infra). Pettit et al. made a comparison of various metal (Fe, Rh, Ru, Os,
Ir, and Pt) carbonyl complexes and clusters as catalysts for the hydroformylation reaction and the
water-gas shift reaction (Scheme 23). In one of the simpler proposed mechanisms for syngas
reactivity, CO is replaced by a hydride to initiate hydride-transfer chemistry. In the foregoing
paragraphs, we give more specific details of homogeneous catalysis by MCCs, organizing the
reactions by group, i.e., Group 8, 9, and 10.

CO2 CO2
-
[HMx(CO)y-1] OH- OH-
[HMx(CO)y-1]-

RCH=CH
2
CO

Water gas shift

MCC
Hydro
RCH2CH2[Mx
Mx(CO)y formylation (CO)y-1]-

[HMx(CO)y]- CO
O
H2 O -
OH- CO RCH2CH2C[Mx(CO)y-1]
RCH2CH2CHO H2 O
H2+OH-

Scheme 23. Mechanisms of CO-driven MCC-mediated catalysis.

Catalysis by MCCs of Group 8 (Fe, Ru, Os). A variety of MCC structure types were
explored for catalytic applications in the early development of MCC chemistry, and much of the
work focused on Group 8 compounds due to their increased stability relative to MCCs with other
transition elements (Chart 1, Table 2).201,202,203,204,205 Casey and coworkers have shown catalytic
performances of the iron cluster Fe3(CO)12 on the isomerization reaction of the alkene in the 1970s;
more specifically, Fe3(CO)12 catalysed the isomerization of the ethyl pentene substrates. Deuterium-
labeling studies revealed that the olefin isomerization reaction occurs via intramolecular hydrogen
shift.206 Geoffroy and coworkers utilised the radical state of the iron cluster, [Fe3(CO)11]−•, to
perform the reduction of nitrobenzene to aniline. The Fe radical anion was produced from the
disproportionation reaction of the halide anion with Fe3(CO)12. While the reaction is stoichiometric,
PhNO2 convert to the azo-azoxybenzene product.65
In more recent examples using MCCs as catalysts, additives are often used, and their role in
the chemistry varies. Nagashima and coworkers207 reported hydrosilylation of tertiary amides to
amine functional groups with Fe3(CO)12 as the catalyst and with 1,1,3,3-tetramethyl disiloxane
(TMDS) or polymethyl-hydro siloxane (PMHS) as the reducing agent. Nearly concurrently, Beller
44
Page 45 of 76 Chemical Society Reviews

and coworkers demonstrated the catalytic activity of iron- and ruthenium-based MCCs toward
hydrosilylation chemistry.208,209,210 In a report by Enthaler and coworkers, a combination of
Fe3(CO)12 and stoichiometric amounts of silyl reagent results in the activation of a sulfoxide,
yielding sulfide product.211

M M -
H CO
M M
M M
M3(CO)12 [HM3(CO)11]-
M = Fe, Ru, Os M = Fe, Ru, Os
Ru Ru
H H H H
Ru Ru Ru Ru
Ru H Ru H
H2Ru4CO13 H4Ru4CO12

Chart 1. Homogeneous MCC Catalysts from Group 8. Terminal CO ligands have been omitted for
clarity.

Table 2. Summary of catalysis by Fe carbonyl clusters.

MCC Additive, substrate Reaction Product (yield a) Reference

Fe3(CO)12 3-ethyl-1-pentene Isomerization 3-ethyl-2-pentene Casey
of 3-ethyl-1- (97) 1973 206
pentene
[HFe3(CO)11]− CO2, H2, alcohol Formate ester Methyl formate Evans
formation (5.8) 1978 201
[Fe3(CO)11]−• Reduction of Stoichiometric Geoffroy
PhNO2 reaction 1995 65
Fe3(CO)12 TMDS, h, Thermal- or Amine products Nagashima
carboxamide photo-assisted (21–96, thermal; 2009 207
substrates amine 73–95, photo)
reduction
Fe3(CO)12 Silane, methyl phenyl Reduction of Methyl phenyl Enthaler,
sulfoxide methyl phenyl sulfide (99) 2011 211
sulfoxide
Fe3(CO)12 Phosphine, ketone Asymmetric Alcohol products Gao
substrates ketone (66–99) 2014 212,213
hydrogenation
Fe3(CO)12 PhSiH3, N,N- Amide to Amine products Beller
dimethylbenzamide amine (99) 2019 208

[HFe3(CO)11]− (EtO)2MeSiH, primary Dehydration of Nitrile products Beller

amines amides to (52–99) 2009 209
nitriles
[HFe3(CO)11]– P2N2 ligandb, base, Hydrogenation Amine products Beller
iPrOH, N- of imine (35–95) 2010 210
(diphenylphosphinyl)-
imines derivatives
[TeFe3(CO)9]2– Copper catalyst, aryl Homocoupling Biphenyl Shieh

45
 Chemical Society Reviews Page 46 of 76

[SeFe3(CO)9]2– boronic acid reaction products 2015 214

(54–99)
a %yield, b N,N′-bis[o-(diphenylphosphino)benzylidene]cyclohexane-1,2-diamine

The transformation of amide functional groups into nitriles has been explored using the
hydrogenated iron carbonyl cluster, [Et3NH][HFe3(CO)11]. Using various silyl additives such as
Ph2SiH2, (EtO)3SiH, and (EtO)2MeSiH, the amide group is transformed into a nitrile group within
3 h with above 90% yield. The same transformation was achieved in 97% yield using Fe2(CO)9
although the reaction time was longer (20 h);213 83% yield was obtained using Fe3(CO)12 with a 2-h
reaction time. In general, the amide substrates employed in this work were primary carboxylic
amides, which include alkyl, aromatic, heteroaromatic, and aliphatic amides. Reactions with
activated phenyl carboxylic amides achieved completion in 4 h, whereas unactivated substrates took
30 h.
Asymmetric transfer hydrogenation (ATH) converts a ketone or amine into an alcohol or
imine, respectively. Various iron-containing catalysts, including mononuclear catalysts and the iron
carbonyl clusters [Et3NH][HFe3(CO)11] and Fe3(CO)12, show high yields and selectivity in ATH;
however, there are significant differences in reaction outcomes depending on the cluster employed,
as described below. The hydrogenation of imines was successfully achieved and described by Beller
and coworkers using systems with chiral ligands also acting as co-catalysts.210 The scope of the
imine substrates included a ketamine, aromatic imine, and hetero-aromatic imine, while the co-
substrates for ATH included iPrOH and chiral P2N2 ligands, which are tetradentate ligating at two P
and two N centres. Since these observations were made under conditions employing chiral ligand
additives, it was initially inferred that the chiral ligand likely promoted the formation of a chiral
mono-iron coordination complex, which serves as the active catalyst when used with Fe3(CO)12.
However, a recent report by Gao and coworkers has shown that varying the iron carbonyl clusters
for ATH does change the reaction outcomes. As an example, Fe3(CO)12 fully converts
acetophenone to (S)-1-phenylethanol with 97% ee (enantiomeric excess), while
[Et3NH][HFe3(CO)11] does not activate the substrate. One possible explanation for these different
catalytic performances might arise from the varying coordination of the chiral ligands directly onto
different Fe3 cluster cores or perhaps to a fragment of the initial cluster.213 In another report by Gao
and coworkers, the catalytic reaction was monitored by IR spectroscopy, verifying that
[Et3NH][HFe3(CO)11] stayed intact during the reaction. This may serve as evidence for
[Et3NH][HFe3(CO)11] playing the role of active catalyst, or it may be that a very small fraction of
[Et3NH][HFe3(CO)11] is converted to the active catalyst although the event is not detected by IR.212
Taken together, these studies do not yet define a clear role for Fe3(CO)12 or [Et3NH][HFe3(CO)11] in
the catalytic ATH conversion of amines to imines. Shieh and coworkers have recently reported
46
Page 47 of 76 Chemical Society Reviews

homocoupling reactions where a copper catalyst is coordinated with [Se/TeFe3(CO)9]2−.214

Comparing to previous examples, the iron cluster supports the catalysis as a ligand substitute. For
both the Se- and Te-containing species, the reaction yield is above 80% for the homocoupling
reaction.
The study of catalytic reactions using ruthenium carbonyl clusters started with interest in
reactions utilising syngas. Examples reported by Ford and coworkers215 and Shore and coworkers 216
provided evidence for hydrogen evolution from water and CO gas using various Ru carbonyl
clusters (Table 3). Darensbourg and coworkers screened a series of Ru carbonyl clusters and
demonstrated C–C bond formation between CO2 and methanol to afford methyl formate: in the
presence of CO and H2, reactions promoted by [HRu3(CO)11]− had a turnover number (TON) of
around 4, while those promoted by [HRCO2Ru3(CO)10]− had a TON of 5.7. In this work, it was
suggested that the trinuclear Ru cluster is a precursor for the active catalyst, which is formed via a
fragmentation step to a tetranuclear species. Evidence suggested that a mononuclear Ru species is
not the active catalyst. The formation of methyl formate from H2 and CO2 is regarded as a water-gas
shift reaction that can be used to produce CO. For experiments performed in the absence of H2 and
CO2, use of CO led to TONs of up to 106. Thus, it is plausible that the formation of methyl formate
arises from the insertion of carbon monoxide into methanol.204

47
 Chemical Society Reviews Page 48 of 76

Table 3. Summary of catalysis by Ru carbonyl clusters.

MCC Additive, Reaction Product a Reference
substrate
Ru3(CO)12 Amine, CO, H2O, Hydroformylation Aldehydes, Pettit
H4Ru4(CO)12 propylene Water-gas shift alcohol, H2 1977 202
Ru3(CO)12 KOH, H2O, CO, H2 evolution H2 Ford
H2Ru4(CO)13 ethoxyethanol (Catalyst: 1.5– 1978 215
H4Ru4(CO)12 4.4 b)
Ru6C(CO)17
Ru3(CO)12 CO2, H2, MeOH Alkyl formate Methyl formate Darensbourg
production (106 b) 1983 204
[HRu3(CO)11]− H2O, CO H2 evolution H2 Shore
(100) 1985 216
Ru3(CO)12 (EtO)3SiH Hydrosilylation of 1-octene (70 c) Hilal
olefin, acetophenone acetophenone 1993 217
(220 c)
Ru3(CO)12 CO, imines Carbonylative [4+1] Various lactams Murai
cycloaddition product 1999 218
(51–96)
Ru3(CO)12 Acenaphthylene, Hydrosilylation of Silyl ether Nagashima
silane ketone product 2000 219
(76–99)
Alcohol
products
(72–98)
Ru3(CO)12 Silane (Et3SiH), Hydrosilylation of Alkyl silyl acetal Fuchikami
ester esters to alkyl silyl product 2001 220
acetals (36–94)
Ru3(CO)12 CO/C2H4 CO insertion Stoichiometric Imhof
reaction 2005 221
Ru3(CO)12 Acenaphthylene, Amide to amine Amines Nagashima
silane, amides (34–98) 2009 207
a % yield, b activity (Moles of H2 per mole of complex per day), c turnover number

The second area of interest for Ru carbonyl clusters involves reactions with silane
additives. Some of the most successful implementations of this approach include the transformation
of esters to alkyl silyl acetals, which are hydrolysed to aldehyde. The proposed mechanism of this
reaction involves oxidative addition of silane to a metal in Ru3(CO)12, and mechanistic work
suggests that an intermediate silane-ligated Ru cluster is important for Si–H bond activation, which
leads to subsequent reaction with olefin.222 Also using Ru3(CO)12, hydrosilylation of acetophenone
has faster turnover than olefin formation for 1-octene using (EtO)3SiH (TON: 220 and 60,
respectively).217 By replacing the silane with (EtO)3Si(CH2)3NH2, the Ru catalyst becomes more
active for the hydrogenation of olefin substrates.223 Murai and coworkers demonstrated the
carbonylative [4+1] cycloaddition of alkyl and aryl α,β-unsaturated imines with CO using
Ru3(CO)12. Although they were not able to definitively identify the active species, a mechanism has
been proposed (Scheme 24).216

48
Page 49 of 76 Chemical Society Reviews

Ru(CO)4
NR NR
O +CO

NR NR
(OC)3Ru Ru(CO)4
O

Scheme 24. Proposed mechanism for carbonylative [4+1] cycloaddition using Ru3(CO)12.

Fuchikami and coworkers have reported reduction of an ester to aldehyde via a proposed
silyl acetal intermediate that is hydrolyzed to afford the aldehyde.218 The role of the silane addition
is different in this example because it reacts with the substrate directly, rather than activating
Ru3(CO)12 by direct reaction with a Ru centre.220 Nagashima has found that coordination between
Ru3(CO)12 and acenaphthylene enables catalytic hydrosilylation of ketone and reduction of an
amide to an amine group in a presence of silane additive.219
Compared with Fe and Ru carbonyl clusters, there are far fewer reports of catalysis using
Os carbonyl clusters (Table 4).205,224 Muetterties and coworkers investigated the reactivity of
Os3(CO)12 and Ir4(CO)12 towards hydrocarbons, which led to the discovery that Os3(CO)12 can
effect hydrogen–deuterium exchange. Years later, Adams proposed multiple molecular structures
of intermediates to better understand the catalytic reaction of Os3(CO)12 with isocyanides and
related functional groups, using a series of stoichiometric reactions where activation of the nitrile
CN bond was observed.225

Table 4. Summary of catalysis by Os carbonyl clusters.

MCC Additive, Reaction Product a Reference
substrate
CO reduction to Alkane (not Muetterties
Os3(CO)12 CO, H2 mentioned)
methane 1976 205
Deuterated Muetterties
Os3(CO)12 H–D exchange benzene (not
1976 224
mentioned)
C4 aldehyde
Os3(CO)12 Amine, CO, Hydroformylation (13 b) Pettit
H4Os4(CO)12 propylene Water-gas shift H2 1977 202
(270 b)
CO, H2O, Reduction of Pettit
Os3(CO)12 trimethylamine, Aniline (100)
nitrobenzene 1978 203
nitrobenzene
Reactivity with Stoichiometric
reaction Adams
Os3(CO)12 isocyanides and
1982 225
nitrile group
a % yield, b activity (Moles of H2 per mole of complex per day)

49
 Chemical Society Reviews Page 50 of 76

Catalysis by MCCs of Group 9 (Co, Rh, Ir). Pittman and coworkers studied phosphine-
substituted Co carbonyl clusters, Co3(CO)9CPh and Co4(CO)10(PPh)2.226 These clusters promote the
hydroformylation reaction of pentene, which was the first example of homogeneous catalysis with
Co clusters (Chart 2, Table 5). Both catalysts completely converted 1- or 2-pentene to three
different types of formylated products within 24 h. A discussion of possible fragmentation
pathways to afford the active catalyst was presented, but no definitive active catalyst could be
identified in this work.

Ph Ph
CO
CO

C P
M M
M M M M M M
M P
Co2(CO)8 Ph
Co3(CO)9(CPh)
Co4(CO)10(PPh)2
M = Rh M = Ir
OC M
M
CO

O M
C M M
M M
M M M M
C

M
O

M
CO
OC

M
OC

Rh4(CO)12
Rh6(CO)16 Ir4(CO)12

Chart 2. Homogeneous MCC Catalysts from Group 9. Terminal CO ligands have been omitted for
clarity.

Table 5. Summary of catalysis by Co carbonyl clusters.

MCC Additive, substrate Reaction Product (yield a) Reference
Co3(CO)9CPh CO, H2, 1-pentene, Hydroformylation Aldehydes Pittman
Co4(CO)10(PPh)2 2- pentene (99.7–100) 1977 226
Co2(CO)8 Me3SiH, Hydrosilylation of Disilyl amines Murai
benzonitrile benzonitrile (11–91) 1989 227
Co2(CO)8 HSiMe2Ph, Hydrosilylation of Silane product Ojima
Co2Rh2(CO)12 isoprene isoprene (40–100) 1991 228
Co3Rh(CO)12
Co3(CO)9CCl Amine–phosphine Asymmetric ketone Alcohol product Gao
ligand, KOH, hydrogenation (75) 2014 212
propiophenone
a % yield

Homoleptic Co carbonyl clusters have also be used as homogeneous catalysts. Using a

silane additive, Murai and coworkers performed a hydrosilylation reaction of benzonitrile.227 Ojima

50
Page 51 of 76 Chemical Society Reviews

and coworkers explored cobalt–rhodium carbonyl clusters for hydrosilylation of isoprene,

cyclohexanone, and cyclohexanone derivatives.228 Although Rh4(CO)12 and Co–Rh carbonyl
clusters show higher conversion rates, Co2(CO)8 can also perform the catalytic hydrosilylation of
isoprene.229 In recent work by Gao and coworkers, an additional example of asymmetric
hydrogenation of ketone was presented; that was described above in the discussion of Fe carbonyl
cluster catalysts.212
In the catalysis of rhodium carbonyl clusters, Rh4(CO)12 and Rh6(CO)16 are popular (Chart
2, Table 6). Pettit and coworkers used Rh6(CO)16 clusters and CO gas to perform hydroformylation
202 and nitrobenzene reduction.203 Yamazaki and coworkers demonstrated synthetic pathways with
Rh4(CO)12 and Rh6(CO)16, where they focused on C–H activation of an arene followed by the
addition of substrates, including benzene, N-methylpyrrole, thiophene, and furan. They also
introduced hydrocarbonylation of acetylene and olefins in the presence of CO to afford furanone
derivatives.230 Using silane additives, Ojima and coworkers activated Rh4(CO)12, to convert
unsaturated hydrocarbon substrates—such as 1-hexyne, isoprene, and cyclohexenone—into highly
regioselective and stereoselective silylated product.228

Table 6. Summary of catalysis by Rh carbonyl clusters.

MCC Additive, Reaction Product a Reference

substrate
Rh6(CO)16 Amine, CO, H2O, Hydroformylation C4 aldehyde Pettit
propylene (300 b) 1977 202
H2
(1700 b)
Rh6(CO)16 CO, H2O, Reduction of Aniline (100) Pettit
nitrobenzene nitrobenzene 1978 203
Rh4(CO)12 CO, ethanol Carbonylation of Furanone (59–67) Yamazaki
Rh6(CO)16 diphenylacetylene dipheylactylene 1983 230
Rh4(CO)12 HSiMe2Ph, Hydrosilylation of Silane product Ojima
Co2Rh2(CO)12 isoprene isoprene (40–100) 1991 228
Co3Rh(CO)12
Rh4(CO)12 HSiMe2R, 1- Silylformylation Silane product Ojima
Co2Rh2(CO)12 hexyne of 1-hexyne (63–95) 1991 229
Rh6(CO)16 CO, H2O, amine Reduction of nitro Amines Kaneda
groups (74–95) 1994 231
a % yield, b turnover number

In our earlier discussion of Os-carbonyl-mediated catalysis, we mentioned the application

of Ir4(CO)12 in the hydrogen–deuterium exchange reactions of hydrocarbons (Table 7).202,205,224 In
addition, applications of supported Ir carbonyl clusters have been reported by Gates and coworkers.
For example, MgO-supported [HIr4(CO)11]− enables propane hydrogenolysis to yield methane and
ethane when the reaction is carried out at 200 °C and 1 atm of H2. Using Ir6(CO)16 encapsulated by
51
 Chemical Society Reviews Page 52 of 76

NaY zeolite, catalytic CO hydrogenation yields a C2–C4 carbon product. Gates and coworkers have
also investigated catalytic performances of various other MCCs supported by MgO or zeolite.232,233
The encapsulated Ir cluster was prepared by the adsorption of [Ir(CO)2(acac)] into the cage of NaY
zeolite and thermal treatment. IR and extended X-ray absorption fine structure (EXAFS) techniques
were adopted to verify the structure of the final catalyst because Ir6(CO)16 has a distinct signature at
1730 cm−1. Using a feedstock of CO and H2, zeolite-encaged Ir cluster yields various hydrogenated
products, which were detected by gas chromatography (GC). The continuous reaction time of up to
8 days reveals high stability for Ir6(CO)16 within the zeolite cage. In the reactions starting with
MgO-supported Ir6(CO)16, IR spectroscopy detected the structure of Ir4(CO)12, and it is believed
that this cluster forms on the MgO surface. In each of these reports, it is believed that the supporting
substrate enhances the stability of the MCCs.

Table 7. Summary of catalysis by Ir carbonyl clusters.

MCC Additive, Catalysis Product a Reference
substrate
Ir4(CO)12 CO, H2 CO reduction to methane Alkane (not Muetterties
mentioned) 1976 205

Ir4(CO)12 H–D exchange Deuterated benzene Muetterties

(not mentioned) 1976 224

Ir4(CO)12 Amine, CO, H2O Hydroformylation C4 aldehyde Pettit

Water-gas shift (250 b) 1977 202
H2
(300 b)
Ir4(CO)12 KOH, H2O, CO Water-gas shift H2 Ford
(H2 evolution) (5.3 c) 1978 215

Ir6(CO)16 Zeolite Hydrogenation of CO Propane (not Gates

encapsulated Decarbonylation mentioned) 1993 232
HIr4(CO)11− Supported on Propane hydrogenolysis H2 (not mentioned) Gates
MgO surface 1993 233
a % yield, b turnover number, c activity

Catalysis by MCCs of Group 10 (Pd and Pt). Likely due to their limited stability, the
carbonyl clusters of Ni are not represented in the literature for catalysis. As for Pt carbonyl clusters
(Chart 3, Table 8), the first catalytic reaction was reported by Pettit and coworkers in 1977, and that
report described hydroformylation and water-gas shift reactions.202 The next known catalytic
reaction was reported in the late 1990s. It is challenging to stabilise the MCCs of Group 10 under
ambient conditions, especially under catalytic conditions; for this reason, all the examples involve
catalysts supported on mesoporous silica or zeolite. Various evidence for the identity of the active
catalyst in each of these cases is obtained from in-situ IR spectroscopy and EXAFS. In an
alternative mechanism for catalysis in Group 10, Bhaduri and coworkers have employed redox

52
Page 53 of 76 Chemical Society Reviews

pathways where a reduced form of Pt9(CO)18 is used as a reductant to reduce nicotinamide adenine
dinucleotide (NAD+). In this reaction, the platinum cluster takes the role of electron transporter
from the sacrificial electron-donor molecules, which are hydroxide ions (OH−) recovered from
ferric cyanide. In an alternative redox reaction, methylene blue, Safranine O, and methyl viologen
were used as sacrificial electron acceptors, which were then reduced by the Pt cluster.235

2-
O Pt
O C CO
C Pt
CO

OC

Pt Pt OPt Pt Pt
C
CO

C C
O n O
Pt Pt
{[Pt3(CO)6]2-}n C
n = 1-10 O

Chart 3. Homogeneous MCC Catalysts from Group 10. Terminal CO ligands have been omitted
for clarity.

Table 8. Summary of catalysis by Pt carbonyl clusters.

MCC Additive, substrate Reaction Product a Reference

[Pt3(CO)6]5− Amine, CO, H2O, Hydroformylation C4 aldehyde Pettit
propylene Water-gas shift (0.5 b) 1977 202
H2
(700 b)
[Pt3(CO)6]52− Zeolite- Hydrogenation of Alkyl product Ichikawa
encapsulation, ethene, 1,3-butadiene (<0.1 c) 1998 234
ethene, 1,3-
butadiene
[Pt9(CO)18]2− Sacrificial Electron-transfer Not mentioned Bhaduri
electron-donor/- mediator 2000 235
acceptor
[Pt12(CO)24]2− Mesoporous silica, Hydrogenation of Not verified Bhaduri
MCM-41 support, methyl pyruvate or (10‫–ؘ‬40) 2005 236
methyl pyruvate, acetophenone
acetophenone
Pd13(CO)x Zeolite Y, CO, H2 CO hydrogenation Hydrocarbon Sachtler
(1.8–2.3) 1992 237
a % yield, b turnover number, c turnover frequency (s−1)

3.3 Electrocatalysis by MCCs

In an electrocatalytic reaction, an external applied potential is used to initiate the reaction of an
electron-transfer event with the redox-active catalyst, which may be homogeneous or
heterogeneous, depending on whether it can be in solution with the substrate. Homogeneous
electrocatalysts are metal-based coordination complexes or clusters, which will exist in solution
with the other reagents; heterogeneous electrocatalysts may be extended solids or other insoluble
53
 Chemical Society Reviews Page 54 of 76

materials that are commonly deposited on the electrode. We will focus only on homogeneous
electrocatalysis, where the most basic requirements for the electrocatalyst include stability in a polar
solvent system and a redox potential that is accessible without oxidising or reducing the solvent.
These criteria alone limit the number of MCCs that might be suitable electrocatalysts, but those
MCCs that do meet these criteria have many advantages over single-metal-ion homogeneous
electrocatalysts, as described below. Considered from another angle, there are advantages for using
electrocatalysis instead of thermal activation as a means to control the reactivity of MCCs. One big
advantage of electrochemical catalysis is that energy can be added to the system at RT and with
very mild conditions; as a result, the stability of MCCs can be maintained throughout the catalytic
reaction and for many turnovers. In most instances, it is also possible to include a partial
atmosphere of CO (if it is needed), imparting even greater stability to the MCC; the mole-fraction
of CO in the headspace can be tuned at will.
The advantages of using suitable MCCs as homogeneous electrocatalysts, rather than using
single-site metal–ligand catalysts or heterogeneous electrocatalysts, have come to light in recent
years. Suitable MCCs and their reduced forms are readily protonated, allowing the formation of
hydride intermediates at low applied potential; in one instance, catalysis was promoted at a rate of
109 s-1 with an applied potential as low as –0.86 V vs. SCE (versus the saturated calomel
electrode).238 Multiple sites for protonation on the clusters lead to very fast rates of proton transfer
(PT), facilitating diffusion-limited PT chemistry, which can be applied to hydride-transfer reactions.
It is thought that the fast rates arise from a statistical effect of having so many possible protonation
sites and hydride migration sites. Another advantage of MCCs as electrocatalysts involve the easy
substitution reactions at the cluster surface, where CO ligands can be replaced by phosphines
containing an array of functional groups. Those substitutions can be used to tune the secondary
coordination sphere of a catalyst or to tune the reactivity properties by changing the electronic
properties of the cluster core. Substitutions of metal ions within the MCC core or of an interstitial
atom within the cluster core can also change the electronic and reactivity properties. A more
detailed discussion of synthetic modifications to MCCs and their effect on catalysis is included
below.
Electrocatalytic reactions where MCCs have been employed are all reduction reactions;
low-valent clusters can stabilise additional electron density with their delocalised electronic
structures and multiple -acceptor CO ligands. Electrons added to the molecular-orbital manifold of
MCCs often strengthen the bonding between the metal and carbon atoms of the M–CO moiety. In a
typical electrocatalytic reduction reaction—including those mediated by MCCs, catalysis is initiated
by electron transfer (ET) to the MCC as potential is applied; this electrochemical elementary step is
54
Page 55 of 76 Chemical Society Reviews

abbreviated as “E” in the standard nomenclature of the field. After the initiation step, if the
intermediate has enough energy to react with the substrate, it will react chemically, and that
chemical reaction is abbreviated as “C” (Scheme 25). Alternatively, the initial E event can be
followed by a second E event before any chemical steps proceed. Thus, a series of elementary steps
in an electrocatalytic reaction will be described as a series of E or C steps. The most common
reaction mechanisms are ECEC and EECC. The latter is considered equivalent to the ECCE
mechanism; while they differ in their starting points in the catalytic cycle, they have identical
electrochemical responses when CV is used to monitor their reactions.239

H2 [Mx(CO)y]n H2 [Mx(CO)y]n
+ e- + e-
+ + e-
+H

[HMx(CO)y]n-1 ECEC [Mx(CO)y]n-1 [H2Mx(CO)y]n+1 ECCE [Mx(CO)y]n-1

+ e- + H+ + H+ + H+
[HMx(CO)y]n [HMx(CO)y]n

Scheme 25. Examples of electrocatalytic hydrogen-evolution pathways by MCCs.

In any electrochemical reaction, the energy difference between the electron-donor and -
acceptor is an important parameter, and this difference is expressed by their redox potentials with
respect to a reference electrode. The IUPAC (International Union of Pure and Applied Chemistry)
recommends use of the ferrocene/ferrocenium couple as a reference in organic solvent;240 other
commonly employed references include the SCE and the natural hydrogen electrode (NHE).
Provided that the experimental conditions are clearly described, conversion between the different
referencing schemes is generally trivial, allowing catalysts to be compared.241 In this review, we use
SCE referencing throughout. Metrics used to assess the utility of electrocatalytic reactions include
the Faradaic efficiency (FE) and the turnover frequency (TOF). The FE is the percentage of the
charge that is converted into product in a preparative-scale experiment. The TOF is equivalent to
kobs, the observed rate constant, and it is reported in units of s-1; as long as factors such as the
applied potential or the overpotential () are accounted for, it can be used to compare the rates of
reactions.
Two electrochemical reactions have been studied using MCCs as the electrocatalyst:
(a) the two-electron reduction of two protons to H2, and
55
 Chemical Society Reviews Page 56 of 76

(b) the two-electron reduction of protons and CO2 into formate.

These transformations share the same two initial steps in the catalyst cycle, where ET is followed
by PT to generate a reduced hydride intermediate (Scheme 26). Reaction of the hydride with
protons generates H2, and reaction of the hydride with CO2 generates formate in the third step. The
fourth step is another ET, so the overall mechanism in either case is ECCE. Even from this simple
overview, it is clear that controlling the reaction chemistry of the intermediate hydride can lead to
many applications for hydride-transfer chemistry using MCCs and potentially many substrates.

Scheme 26. Schematic showing the two possible reaction pathways for a metal hydride – which can
(a) react with protons to form H2 or (b) react with CO2 to form formate.

Electrocatalysis by MCCs of Group 8 (Fe). It was first reported in 2011 that [Fe4N(CO)12]−
can cause the evolution of H2 from protons or the evolution of formate from CO2, depending on the
reaction conditions 244 (Chart 4, Table 9). The first attempt for the electrocatalytic hydrogen-
evolution reaction (HER) and CO2-reduction reaction (CO2RR) was carried out under 1 atm of N2
or CO2, respectively, using organic acids as the proton source. In the proposed mechanism,
[Fe4N(CO)12]− is first reduced at −1.23 V (vs. SCE), and then PT affords [HFe4N(CO)12]−, which
can go on to react with either H+ or CO2 following an ECCE mechanism. There are very few
molecular electrocatalysts that produce formate selectively,242 and [Fe4N(CO)12]− performs this
reaction at modest rates and a low applied potential. In subsequent work reported in 2013,
replacement of organic acids with water buffered at pH 7 afforded formate with 97% FE at –1.23 V
of applied potential.246 The solvent-dependence reflects a lowering of the activation barrier for
hydride transfer due to solvation effects,243 which will not be discussed in detail here. In addition to
the ability of [Fe4N(CO)12]− to perform catalytic small-molecule reduction chemistry effectively,
these reports have demonstrated that it is stable in water and that it is stable under an applied
potential over reaction times of up to 24 h, although longer reactions were not investigated.

56
Page 57 of 76 Chemical Society Reviews

Fe - 2-
Fe Fe
X C
Fe Fe Fe
Fe Fe
Fe
-
X = N, [Fe4N(CO)12] [Fe5C(CO)15]2-
X = C, [Fe4C(CO)12]2-
2-
Fe
Fe Fe
C
Fe Fe
Fe
[Fe6C(CO)18]2-

Chart 4. Drawings of homometallic Fe carbonyl clusters used as electrocatalysts. CO ligands have

been omitted for clarity.

Table 9. Homometallic clusters of Fe and Co and their reactivity for HER and CO2RR.

MCC Additive Potential (V  (mV) Reaction FE (%) a kobs (s-1) Reference

vs. SCE)
[Fe4N(CO)12]− Benzoic acid −1.23 280 HER Quant. 1.0 Berben
2011 244
[Fe4N(CO)12 ]− Benzoic acid, CO2 −1.25 300 formate nr nr Berben
2011
[Fe4N(CO)12 ]− MeCN/H2O (95:5) −1.25 440 HER Quant. nr Berben
2015 245
[Fe4N(CO)12]− MeCN/H2O −1.2 440 formate 95 10 Berben
(95:5), CO2 2015
[Fe4N(CO)12 ]− KHCO3/KCO3 −1.2 formate 95 Berben
pH 6.5 2015
[Fe4C(CO)12 ]2− Acetate buffer −1.25 714 HER 83 368 Berben
[Fe5C(CO)15]2− pH 5 72 nr 2013 246
[Fe6C(CO)18]2− 64 nr
[Co13C2(CO)24]4− Anilinium −0.86 760 HER 78 108 Berben
tetrafluoroborate 2020
a quant. = Quantitative yield, and nr = not reported

In 2013, the electrochemical reduction of protons to H2 was reported using a series of

carbide-centred clusters, [Fe4C(CO)12]2−, [Fe5C(CO)14]2−, and [Fe6C(CO)17]2− (Chart 4).246 Each of
these clusters quickly catalyses H2 evolution from water, and [Fe4C(CO)12]2− was observed to be
particularly stable. Over time, [Fe5C(CO)14]2− and [Fe6C(CO)17]2− converted into [Fe4C(CO)12]2−,
which illustrates a feature of MCC chemistry. One particular structure for a combination of atoms is
often a thermodynamic sink; whereas other analogous clusters are isolable and quite stable, they are
57
 Chemical Society Reviews Page 58 of 76

not a long-term thermodynamic product. This study also enabled a comparison of the nitride-
centred [Fe4X(CO)12]− and carbide-centred [Fe4C(CO)12]2− clusters. Protonation of [Fe4C(CO)12]2−
is concerted with ET and is therefore pH-dependent; meanwhile, [Fe4N(CO)12]− undergoes
sequential ET and PT reactions. The difference in reactivity is attributed to the varied charges on
the clusters. For both [Fe4N(CO)12]− and [Fe4C(CO)12]2−, the protonation site is located along the
bridge of two iron atoms, and the interstitial atom does not get protonated. Another effect of the
increased charge on [Fe4C(CO)12]2- is that PT rates are significantly higher, so H2 evolution is
faster.
Electrocatalysis by MCCs of Group 9 (Co). A cobalt carbonyl cluster has recently been
reported as an electrocatalyst for fast H2 evolution; this cluster has 13 Co atoms,
[Co13C2(CO)24]4−.238 Many of its properties, including its ET and PT rate constants and diffusion
coefficient, are very similar to those of nanoparticles or quantum dots, illustrating the ability of
MCCs to behave like nanoparticles and like heterogeneous electrocatalysts. The larger size of
[Co13C2(CO)24]4−, its multiple Co–Co bonding motifs, and its many electron-withdrawing CO
ligands are just some of the structural features that result in its observed physical properties. When
studying the reduction of protons to hydrogen or the transfer of hydride to CO2 to form formate, the
ability to detect and monitor reaction intermediates is very important: Co carbonyl clusters offer a
unique system for probing protonation chemistry using IR spectra, as was first described by
Zacchini and coworkers.48 When plotting the CO absorption band energy, CO, obtained from IR
data against the x/z values (where x is the number of Co atoms and z is the cluster charge), a linear
relationship is observed for a wide range of clusters—specifically, from 6 to 20 Co atoms. The
charge on the clusters, z, can be varied either by redox events or by protonation of a Co cluster. For
a given cluster charge, protonation events can be monitored accurately using IR spectroscopy,
providing a powerful tool for determining the reaction mechanisms of hydride formation and
hydride transfer to substrates.

Co
Co OC Co
Co CO
OC COCO
OC Co Co Co
Co C C
Co CO
CO Co CO
OC
OC Co
Co Co

Chart 5. Co carbonyl cluster, [Co13C2(CO)24]3− used as an electrocatalyst. Terminal CO ligands

have been omitted for clarity.

58
Page 59 of 76 Chemical Society Reviews

Using [Co13C2(CO)24]3−, the reduction of protons to hydrogen was studied using an applied
potential as the source of electrons (Scheme 26). PT rates were measured at 109 M−1s−1 for the
chemical reaction where electrochemically generated [Co13C2(CO)24]5− reacts with one proton. The
fast PT chemistry supported the HER with an observed rate of 2.3  109 M−1s−1. These PT rates
appear to be limited only by diffusion or the mass transport of protons in solution. The ability of
[Co13C2(CO)24]5− to mediate fast rates of PT was attributed to a statistical effect, which arises from
the multiple Co–Co bonds available for protonation; in heterogeneous catalysis, a similar effect
arises from the multiple PT sites on the electrode surface. Statistical effects that enhance the rates of
PT have also been observed in older work for single-site metal catalysts, i.e., mononuclear
coordination complexes that can act as homogeneous catalysts. In these cases, PT-rate enhancement
is derived from the addition of Lewis bases to the supporting ligands; the multiple Lewis-base sites
deliver protons to the catalytic active site on the metal centre.247 Studies were performed to
elucidate the mechanism for HER using [Co13C2(CO)24]4−. When either H2O or D2O was added to
the CV experiment of [Co13C2(CO)24]3−, an electrochemical kinetic-isotope effect was observed.
The shift in peak potential for the reduction of [Co13C2(CO)24]4− to [Co13C2(CO)24]5− when D2O was
used is indicative of a PT event that occurs in a concerted process with an ET event. Based on the
currently available information, the observed Co–H bond making/breaking process could be either a
protonation of the cluster that is concerted with ET or an intracluster proton migration that is
concerted with ET; the electrochemical data cannot distinguish between these two possibilities.
Both scenarios reflect the ability of large MCCs to interact effectively with protons and to facilitate
proton mobility on the surface of the cluster.

59
 Chemical Society Reviews Page 60 of 76

Scheme 26. Top: Proposed mechanism for proton reduction to H2 in [Co13C2(CO)24]3−; the C and
CO ligands are not shown to simplify the mechanistic scheme. Reproduced with permission from
ref. 238. Bottom: Protonation of [Co13C2(CO)24]3− and migration of CO around the cluster core
(blue, Co; grey, C; red, O).

Atom substitution in the MCC Core: Effects on electrocatalysis. Starting in 2017, Hogarth
and coworkers published several studies on tri-iron carbonyl clusters (Chart 6, Table 10), which
were functionalised by replacing CO ligands or by adding capping ligands to the three-iron
cluster.248,249 Ruthenium clusters were also studied in some cases. For each cluster, two one-electron
reduction events were observed, and H2 evolution was observed at the most negative potential.
Hydride equivalents generated at the first reduction event were not hydridic enough to transfer
hydride to protons, which would have liberated H2. Thiol-capped [Fe3(CO)9(μ-SR)(μ-H)] (where R
= iPr, tBu) was investigated using organic acids, CF3CO2H or HBF4∙Et2O; H2 evolution was
observed at –0.84 (R = iPr) or –1.0 (R = tBu) V vs. SCE, respectively, with the mechanism
assigned as ECEC.249 Although they are less electron-rich, telluride-capped tri-iron clusters also
reduce protons to H2, but with at slower HER rates. Following phosphine substitution, Te-
substituted clusters showed enhanced reactivity. Their proposed mechanism shows that it is possible
to evolve hydrogen with a pathway that starts from the cluster’s first or second reduced state.237

60
Page 61 of 76 Chemical Society Reviews

R
H N S
N
M M Fe Fe
M H Fe H

[Fe3(CO)9(3-SR)(-H)]
[Fe3(CO)9(3-pyNPh)(-H)]
(M = Fe, Ru) (R = iPr, tBu)

Te
Mn
M M M O
Fe Fe
Te Fe
M3(CO)9(3-Te)2 [Fe3MnO(CO)12]-
(M = Fe, Ru)

Chart 6. Heterometallic and capped Fe carbonyl clusters used as electrocatalysts. CO ligands have
been omitted for clarity.

Table 10. Heterometallic and ligand-capped Fe and Ru clusters and their reactivity for HER.

MCC Additive Potential  (mV) Reaction FE (%) kobs Reference

V vs. SCE (s-1)
[Fe3(CO)9(μ3-pyNH)(μ-H)] TsOH −1.21 nr HER nr nr Hogarth
HBF4∙Et2O 2017 248

[Ru3(CO)9(μ3-pyNH)(μ-H)] TsOH −1.61 nr nr nr Hogarth

HBF4∙Et2O 2017

[Fe3(CO)9(μ-SR)(μ-H)] CF3CO2H −0.84 nr HER nr nr Hogarth

R=iPr, tBu HBF4∙Et2O −1.0 2018 249
[Fe(CO)3]3(μ3-Te)2 CF3CO2H nr nr HER nr nr Hogarth
2020 250
[Fe3MnO(CO)12]− MeCN/H2O (95:5) −1.3 540 HER 50 nr Berben
2020 252
[Fe3MnO(CO)12]− CClH2CO2H −1.3 650 nr Berben
2020
Ts = p-toluenesulfonyl; nr = not reported.

Work published in 2020 discussed the effect of Mn substitution in the [Fe4N(CO)12]−

cluster core where the interstitial nitride atom was also replaced by an oxo atom; the subsequent
[Fe3MnO(CO)12]− cluster is therefore isoelectronic to [Fe4N(CO)12]−.251,252 The two clusters,
[Fe3MnO(CO)12]− and [Fe4N(CO)12]- also have similar reduction potentials of –1.3 and –1.2 V vs.
SCE, respectively. Sequential reduction and protonation of [Fe3MnO(CO)12]− produced a hydride
intermediate, [H-Fe3MnO(CO)12]−, which is equivalent to the hydride intermediate observed for
[Fe4N(CO)12]− (Scheme 26); however, it is at this point that the chemistries of [Fe3MnO(CO)12]−
and [Fe4N(CO)12]− begin to differ. Hydride transfer from [H-Fe3MnO(CO)12]− to CO2 was not

61
 Chemical Society Reviews Page 62 of 76

observed due to its reduced ability for hydride donation, which was determined independently using
IR spectroelectrochemical studies.
Ligand substitution on the MCC Core: Effects on electrocatalysis. Ligand substitution on
[Fe4N(CO)12]−, where CO ligands are replaced by phosphine ligands, has been used to study the
mechanism of H2 and formate evolution. Ligand substitution has also been used to modify the
reactivity of [Fe4N(CO)12]−. For these studies, the control compound is [Fe4N(CO)11(PPh3)]−, which
contains a phosphine ligand but does not contain a reactive functional group.253 The phosphine-
substituted compounds (including the control [Fe4N(CO)11(PPh3)]−) have a reduction potential of
around –1.45 V, while that of [Fe4N(CO)12]− is –1.2 V. Phosphine ligands with hydroxyl,
methoxyphenyl, pyridyl, and N,N-dimethylaniline groups have been used to afford
[Fe4N(CO)11(PPh2(CH2)2OH)]−, [Fe4N(CO)11(PPh2(MeOPh))]−, [Fe4N(CO)11(PPh2py)]−, and
[Fe4N(CO)11(PPh2(N,N-Me2NPh))]−, respectively (Chart 6). For each of these examples (Table 11),
MeCN/H2O (95:5) was used for the mechanistic or other electrochemical studies because it
provided adequate solubility; MeCN has a large electrochemical window, and the small amount of
H2O serves as a proton source and as a source of stabilization for the transition state during hydride
transfer. Using multiple approaches, electrochemical mechanistic studies have shown that water
stabilises a charge-separated transition state and enhances the rates of hydride transfer from [H-
Fe4N(CO)12]− to CO2 by five orders of magnitude relative to the rates of hydride transfer observed
in the absence of water. Other research groups working with single-site hydride-transfer catalysts
have also shown using density functional theory (DFT) studies that water may play a role in the
stabilization of the hydride-transfer transition state via a hydrogen-bonding network.242

L = CO
Fe L PPh3
PPh2(CH2)2OH
N
PPh2(2-Me2NPh)
Fe Fe PPh2(2-MeOPh)
Fe
PPh2py

Chart 7. Schematic of the phosphine-substituted derivatives of [Fe4N(CO)12]-. CO ligands have

been omitted for clarity.

62
Page 63 of 76 Chemical Society Reviews

Table 11. Summary of electrocatalysis by phosphine-substituted derivatives of [Fe4N(CO)12]−. All

reactions were performed in MeCN/H2O (95:5) and under 1 atm of CO2 with an applied potential of
–1.47 V vs. SCE.
MCC Reaction Reference
[Fe4N(CO)11(PPh3)]− CO2RR Berben
2016 253
[Fe4N(CO)11(PPh2(CH2)2OH)]− HER Berben
2016
[Fe4N(CO)11(PPh2(N,N-Me2NPh))]− CO2RR Berben
2020 254
[Fe4N(CO)11(PPh2(MeOPh))]− HER Berben
2020
[Fe4N(CO)11(PPh2py)]− HER Berben
2020

When a CO ligand of [Fe4N(CO)12]− is replaced with a phosphine containing a protic

functional group, e.g., [Fe4N(CO)11(PPh2(CH2)2OH)]−, the protic functional group behaves like a
proton relay during eletrocatalysis; the catalysts will only make H2, even under conditions that are
otherwise ideal for the reaction of [Fe4N(CO)12]− with CO2.253 Work using the parent cluster,
[Fe4N(CO)12]−, revealed the possibility of a hydride intermediate in the catalytic cycle, and
evidence for the hydride intermediate was first seen in the cyclic voltammetry data, where the
hydride was shown to be oxidised. Studies on [Fe4N(CO)11(PPh2(CH2)2OH)]−, where H2 evolution
is switched on by the proton relay, provide additional evidence for the hydride intermediate.
In other investigations, the functionalised phosphine ligands added to [Fe4N(CO)12]− were
used to add steric bulk to the secondary coordination sphere. These large functional groups can also
facilitate proton relays, spanning the range of pKa = –0.5 up to 11.2. With bulky functional groups
however, clusters of protonated water cannot approach the active site to produce H2, resulting in the
exclusive production of formate, as is observed in [Fe4N(CO)11(PPh2(N,N-Me2NPh))]−.254 The
hypothesis that steric effects are important was proposed based on measurements of PT rate
constants performed under N2 and CO2 atmospheres, where the source of protons differ; under N2,
water–MeCN clusters dominate as the proton source, while under CO2, carbonic acid clusters
dominate. The carbonic acid clusters are smaller than the water–MeCN clusters, making PT faster
under CO2 than under N2. The selectivity of catalysts such as [Fe4N(CO)11(PPh2(N,N-Me2NPh))]−
for formate formation over H2 evolution is consistent with the small carbonic acid clusters having
easy access to the active site when there are large functional groups present in the secondary
coordination sphere.

63
 Chemical Society Reviews Page 64 of 76

4. Summary and Outlook

Outlook for the Synthesis of MCCs: From a synthetic point of view, the preparation and
molecular-level characterization of larger and larger MCCs remains an ongoing challenge. The total
structural determination of the Au102-thiolate nanocluster in 2007255 by SC-XRD generated a
renewed interest in molecular nanoclusters. It also clearly pointed out that SC-XRD is currently the
only analytical technique available for structural determination with an atomic precision for the
metal core, metal atom surface, and ligand shell of molecular nanoclusters and chemical species in
general. Despite the recent advancements in SC-XRD instrumentation, this poses a serious
analytical limitation to molecular nanochemistry. Consequently, structure–property–reactivity
relationships should be established to provide additional insights for nanoclusters and ultra-small
metal nanoparticles; to achieve this, more and more molecular clusters of increasing size should be
synthesised, structurally characterised, and chemically and physically analysed. Establishing these
relationships may help resolve the persistent open question of how many metal centres are needed
to produce metallic properties.5 The answer depends not only on the dimensions, but also on the
nature of the metal (or metals, for alloys), the type of ligands present, and the property of interest.
Another target of synthetic chemists is to better understand which MCCs will be stable
enough to use in materials and catalysis applications. Work in electrocatalysis has shown that
structures like [Fe4N(CO)12]– are incredibly stable—in water, in air, and over multiple days under
catalytic conditions. But how do we predict this without having to screen many MCCs? The
stability of [Fe4N(CO)12]– was determined empirically, just as experimental work determined that
many clusters in the iron-carbide series, such as [Fe5C(CO)14]– and [Fe6C(CO)17]–, are not stable,
decomposing into [Fe4C(CO)12]– and Fe metal over time. Rules for electron-counting may aide with
predictions, but their validation must be achieved experimentally. Additionally, a role for
theoretical methods in predicting the stability of MCCs is envisioned.
As the applications of MCCs expand further into catalysis, electrocatalysis, and materials
chemistry, advancements in their synthetic chemistry would be beneficial, especially with respect to
the control of ligand substitution on the cluster cores. Replacement of the CO ligands with other
ligands causes many changes in the electronic structure, solubility, and reactivity properties of the
MCC. Almost all ligands available for substitution are less efficient -acceptors than CO, and thus,
the cluster core becomes more electron-rich as more ligand substitutions are made; the remaining
C–O bonds are weakened, and the remaining M–C(O) bonds are strengthened. This latter effect
makes it difficult to substitute multiple CO ligands. Small-cone-angle phosphines and isonitriles
have been substituted on [Fe4N(CO)12]- up to four times, and greater control and tailoring of these
syntheses are needed.256,257 With ligand substitution enabling changes to MCC properties, it can
64
Page 65 of 76 Chemical Society Reviews

also be used to modify the overall charge of a cluster or its solubility, making a cluster more
suitable for catalytic applications. Another challenge facing the chemistry of ligand-substituted
MCCs is how to better predict which cluster–ligand combinations will afford stable ensembles.
Outlook for Tuning the Material Properties of MCCs: Another area needing further
development is the synthesis and characterization of heterometallic and alloy molecular
nanoclusters. Mixing different metals with atomic control may lead to nanomaterials with new
physical or chemical properties. Alloying and metal-doping of molecular nanoclusters is gaining
more and more interest,258,259 and heterometallic MCCs may contribute to our understanding of the
synergetic effects induced by alloying with atomic precision.87 This might also be of some
significance in interpreting the metal migration, chemiadsorption, and catalytic behaviour of alloy
nanoparticles.
The electronic properties of molecular nanoclusters can be experimentally investigated using
spectroscopic and electrochemical techniques as well as magnetic measurements. For additional
insights, they can also be compared to theoretical models. As the size of the cluster increases, an
incipient metallisation of its metal core should be observed.6 Indeed, CV measurements indicate a
decrease of the energy gap (E) between consecutive redox couples in multivalent MCCs as their
size increases. At a certain size, E should be comparable to the thermal energy at RT, leading to
auto-disproportionation and equilibria among isostructural MCCs with different charges. This new
electronic status should also influence the magnetic properties of larger MCCs and nanoclusters in
general. Paramagnetism of even-electron MCCs has been debated for a long time, but recently it has
been clearly demonstrated that they can have unpaired electrons in both the ground and excited
electronic configurations.135 Magnetic measurements of molecular nanoclusters require ultra-pure
samples in order to rule out the presence of paramagnetic impurities, but they can add significant
information to our knowledge of nanomaterials. In addition, by understanding and learning to tune
the electronic and magnetic properties of MCCs, it will be possible to produce suitable candidates
for molecular nanocapacitors, superparamagnetic quantum dots, and nanomagnets.
Outlook for Catalysis with MCCs: From a structural point of view, a regular and simple
structure–size relationship has not been established in the case of MCCs and molecular
nanoclusters. This is likely due to the fact that at these length scales, M–ligand and M–M
interactions contribute similarly to the overall energetic properties. Thus, we can expect that
introducing defects in the metal cages should lead to rearrangements, as shown in the case of some
Pt MCCs. Indeed, there is evidence that the metal cages of some MCCs are fluxional, and generally
speaking, they are soft and deformable. A deeper knowledge of the dynamic behaviour of the metal
cage and ligand shell of molecular clusters is required, and this might have some relevance also to
65
 Chemical Society Reviews Page 66 of 76

heterogeneous catalysis with ultra-dispersed metals. The size, structure, and composition of
molecular nanoclusters strongly affect their catalytic behaviour.
There are still significant opportunities for studying MCCs of different sizes and
compositions in the activation of small molecules, in homogeneous catalysis, as precursors of
heterogeneous nanostructured catalysts, and in homogeneous electrocatalysis. An overview of
homogeneous catalysis and homogeneous electrocatalysis by MCCs has been given in this review,
and other recent reviews have touched on related aspects of catalysis with MCCs.23, 260, 261 General
ideas that can be drawn from these examples include the knowledge that thermal activation of
catalytic cycles often results in MCC decomposition and that many of the most effective thermal
catalytic reactions are those that occur under a MCC-stabilising CO atmosphere. Moving forward, it
is essential that we develop better predictive methods for knowing which MCCs will be stable to
various reaction conditions.
Electrocatalysis has recently provided a method where energy, in the form of the applied
potential, can be added to the system without raising the temperature. This has enabled RT catalytic
reactions, including reactions proceeding via a hydride intermediate. As an example, the reduction
of CO2 into formate can proceed in water for over 24 h when [Fe4N(CO)12]– is electrolyzed at –1.2
V vs. SCE.245 Future work in electrocatalysis with MCCs should take advantage of the tunable
surface of MCCs, where CO ligands can be replaced with phosphine ligands. Inclusion of
functionalised phosphine ligands has the potential to modulate reaction rates by orders of magnitude
and to modulate product selectivity when the correct functional groups are included in the
secondary coordination sphere and installed in appropriate locations. Another frontier in
electrocatalysis using MCCs is the study of larger and larger clusters. Recent work has shown that
larger MCCs, containing 13 cobalt atoms, enhance the rates of proton transfer by many orders of
magnitude compared to the four-iron cluster, [Fe4N(CO)12]–, until they resemble those observed for
heterogeneous electrocatalysts. Proton- and hydride-migration chemistry on these tridecanuclear
cobalt clusters also resembles the behaviour of hydrogen atoms on a surface where hydrogen atoms
are not localised at one place.238 The combination of these nearly diffusion-limited reaction rates
with the ability to use molecular chemistry techniques to characterise structure, reactivity, and
reaction kinetics and mechanism in precise detail is a powerful driving force for the future
development of electrocatalytic reactions promoted by MCCs.

5. Conclusions
The ongoing, new insights into the electronic properties, structural dynamics, and catalytic
mechanisms of MCCs show no sign of abating. Nowadays, we possess a solid and broad knowledge

66
Page 67 of 76 Chemical Society Reviews

of the synthesis of MCCs, which can be exploited for future developments and applications as
outlined in the Summary and Outlook. The relevance of electrocatalysis to fundamental chemistry
and industrial applications is growing; and electrochemical techniques have provided a much-
needed RT approach, enabling MCCs to be stable throughout chemical and catalytic
transformations. Molecular nanoclusters, in general, and MCCs, in particular, can add new
perspectives to electrocatalysis. Being at the cusp of the nanodomain, they can contribute to our
knowledge of nanochemistry and solid-state materials.

6. Acknowledgement

L.A.B.’s work on MCCs has been supported by the Department of Energy, Office of Science, Basic

Energy Sciences with award number DE-SC0016395. S.Z. thanks the University of Bologna for
financial support.

References

1 E. L. Muetterties, T. N. Rhodin, E. BAND, C.F. Brucker, and W. R. Pretzer, Chem. Rev. 1979, 79, 91-137.
2 G. Schmid (Ed.), Clusters and Colloids, Wiley-VCH, New York, 1994.
3 P. Braunstein, L. A. Oro and P. R. Raithby (Eds.), Metal Clusters in Chemistry, Wiley-VCH, New York,

1999.
4 G. Schmid and D. Fenske, Phil. Trans. R. Soc. A, 2010, 368, 1207-1210.
5 S. Zacchini, Eur. J. Inorg. Chem., 2011, 4125-4145.
6 C. Femoni, F. Kaswalder, M. C. Iapalucci, G. Longoni and S. Zacchini, Coord. Chem. Rev., 2006, 250,

1580-1604
7 B. F. G. Johnson and J. S. McIndoe, Coord. Chem. Rev., 2000, 200-202, 901-932.
8 I. Ciabatti, C. Femoni, M. C. Iapalucci, G. Longoni and S. Zacchini, J. Clust. Sci., 2014, 25, 115-146.
9 I. Ciabatti, C. Femoni, M. C. Iapalucci, S. Ruggieri and S. Zacchini, Coord. Chem. Rev., 2018, 355, 27-38.
10 B. Berti, C. Femoni, M. C. Iapalucci, S. Ruggieri and S. Zacchini, Eur. J. Inorg. Chem., 2018, 3285-3296.
11 C. Femoni, M. C. Iapalucci, S. Ruggieri and S. Zacchini, Acc. Chem. Res., 2018, 51, 2748-2755.
12 K. H. Whitmire, Coord. Chem. Rev., 2018, 376, 114-195.
13 E. G. Mednikov and L. F. Dahl, Phil. Trans. R. Soc. A, 2010, 368, 1301-1322.
14 J. A. Cabeza and P. García-Álvarez, Chem. Soc. Rev., 2011, 40, 5389-5405.
15 W. Hieber, F. Leutert, Die Naturwissenschaften. 19, 1931, 360–361.
16 G. Hogarth, S. E. Kabir and E. Nordlander, Dalton Trans., 2010, 39, 6153-5174.
17 R. Jin, C. Zeng, M. Zhou and Y. Chen, Chem. Rev., 2016, 116, 10346-10413.
18 I. Chakraborty and T. Pradeep, Chem. Rev., 2017, 117, 8208-8271.
19 X. Kang, Y. Li, M.Zhu and R. Jin, Chem. Soc. Rev., 2020, 49, 6443-6514.
20 X. Du and R. Jin, Dalton Trans., 2020, 49, 10701-10707.
21 (a) Q. Yao, T. Chen, X. Yuan and J. Xie, Acc. Chem. Res., 2018, 51, 1338-1348; (b) S. Takano, S.

Hasegawa, M. Suyama and T. Tsukuda, Acc. Chem. Res., 2018, 51, 3074-3083.
22 (a) Z. Lei, X. -K. Wan, S. -F. Yuan, Z. -J. Guan and Q.- M. Wang, Acc. Chem. Res., 2018, 51, 2465-2474;

(b) K. Konishi, M. Iwasaki and Y. Shichibu, Acc. Chem. Res., 2018, 51, 3125-3133.
23 P. Buchwalter, J. Rosé and P. Braunstein, Chem. Rev., 2015, 115, 28-126.
24 R. D. Adams and F. A. Cotton (Eds.), Catalysis by Di- and Polynuclear Metal Cluster Complexes, Wiley-

VCH, New York, 1998

25 R. D. Adams and B. Captain, J. Organomet. Chem., 2004, 689, 4521-4529.
26 B. C. Gates, Chem. Rev., 1995, 95, 511-522.

67
 Chemical Society Reviews Page 68 of 76

27 J. M. Thomas, B. F. G. Johnson, R. Raja, G. Sankar and P. A. Midgley, Acc. Chem. Res., 2003, 36, 20-30
28 (a) T. Imoaka, H. Kitazawa, W.-J. Chun, S. Omura, K. Albrecht and K. Yamamoto, J. Am. Chem. Soc.,
2013, 135, 13089-13095; (b) T. Imaoka, H. Kitazawa, W.-J. Chun and K. Yamamoto, Angew. Chem. Int.
Ed., 2015, 54, 9810-9815.
29 K. Kratzl, T. Kratky, S. Günther, O. Tomanec, R. Zbořil, J. Michalička, J. M. Macak, M. Cokoja and R. A.

Fischer, J. Am. Chem. Soc., 2019, 35, 13962-13969.

30 P. R. Raithby, Platinum Metals Rev., 1998, 42, 146-157.
31 S. Martinengo, G. Giordano and P. Chini, Inorg. Synth., 1980, 20, 209-212.
32 L. Malatesta, G. Caglio and M. Angoletta, Inorg. Synth., 1972, 13, 95-99.
33 L. Garlaschelli and P. Chini, Gazz. Chim. Ital., 1982, 112, 285-288.
34 (a) G. Longoni and P. Chini, J. Am. Chem. Soc., 1976, 98, 7225-7231; (b) J. C. Calabrese, L. F. Dahl, P.

Chini, G. Longoni and S. Martinengo, J. Am. Chem. Soc., 1974, 96, 2614-2616.
35 M. Fauré, C. Saccavini and G. Lavigne, Chem. Commun., 2003, 1578-1579.
36 M. I. Bruce, C. M. Jensen and N. L. Jones, Inorg. Synth., 1990, 28, 216-218.
37 (a) M. Ichikawa, Platinum Met. Rev., 2000, 44, 3-14; (b) A. Fukuoka, N. Higashimoto, Y. Sakamoto, M.

Sasaki, N. Sugimoto, S. Inagaki, Y. Fukushima and M. Ichikawa, Catal. Today, 2001, 66, 23-31.
38 (a) C. Femoni, M. C. Iapalucci, G. Longoni, C. Tiozzo, S. Zacchini, B. T. Heaton and J. A. Iggo, Dalton

Trans., 2007, 3914-3923; (b) C. Femoni, M. C. Iapalucci, G. Longoni, C. Tiozzo, S. Zacchini, B. T.

Heaton, J. A. Iggo, P. Zanello, S. Fedi, M. V. Garland and C. Li, Dalton Trans. 2009, 2217-2223.
39 I. Ciabatti, C. Femoni, M. C. Iapalucci, G. Longoni and S. Zacchini, Organometallics, 2012, 31, 4593-

4600.
40 A. Fumagalli, S. Martinengo, P. Chini, D. Galli, B. T. Heaton and R. Della Pergola, Inorg. Chem., 1984,

23, 2947-2954.
41 (a) V. G. Albano, D. Braga and S. Martinengo, J. Chem. Soc., Dalton Trans., 1986, 981-984; (b) V. G.

Albano, D. Braga and S. Martinengo, J. Chem. Soc., Dalton Trans., 1981, 717-720; (c) G. Ciani and S.
Martinengo, J. Organomet. Chem., 1986, 306, C49-C52.
42 P. Chini, J. Organomet. Chem., 1980, 200, 37-61.
43 B. Berti, C. Cesari, C. Femoni, T. Funaioli, M. C. Iapalucci and S. Zacchini, Dalton Trans., 2020, 49,

5513-5522.
44 E. Cattabriga, I. Ciabatti, C. Femoni, T. Funaioli, M. C. Iapalucci and S. Zacchini, Inorg. Chem., 2016, 55,

6068-6079.
45 J. J. Brunet, Chem. Rev., 1990, 90, 1041-1059.
46 R. Della Pergola, F. Demartin, L. Garlaschelli, M. Manassero, S. Martinengo and M. Sansoni, Inorg.

Chem., 1987, 26, 3487-3491.

47 A. Ceriotti, N. Masciocchi, P. Macchi and G. Longoni, Angew. Chem. Int. Ed., 1999, 38, 3724-3727.
48 I. Ciabatti, C. Femoni, M. Hayatifar, M. C. Iapalucci, G. Longoni, C. Pinzino, M. V. Solmi and S.

Zacchini, Inorg. Chem., 2014, 53, 3818-3831.

49 D. Collini, F. Fabrizi De Biani, S. Fedi, C. Femoni, F. Kaswalder, M. C. Iapalucci, G. Longoni, C. Tiozzo,

S. Zacchini and P. Zanello, Inorg. Chem., 2007, 46, 7971-7981.

50 I. Ciabatti, F. Fabrizi de Biani, C. Femoni, M. C. Iapalucci, G. Longoni and S. Zacchini, ChemPlusChem,

2013, 78, 1456-1465.

51 I. Ciabatti, C. Femoni, M. Gaboardi, M. C. Iapalucci, G. Longoni, D. Pontiroli, M. Riccò and S. Zacchini,

Dalton Trans., 2014, 43, 4388-4399.

52 B. Berti, I. Ciabatti, C. Femoni, M. C. Iapalucci and S. Zacchini, ACS Omega, 2018, 3, 13239-13250.
53 (a) D. W. Hart, R. G. Teller, C.-Y. Wei, R. Bau, G. Longoni, S. Campanella, P. Chini and T. F. Koetzle,

Angew. Chem. Int. Ed., 1979, 18, 80-81; (b) D. W. Hart, R. G. Teller, C.-Y. Wei, R. Bau, G. Longoni, S.
Campanella, P. Chini and T. F. Koetzle, J. Am. Chem. Soc., 1981, 103, 1458-1466.
54 P. Chini, Inorg. Chim. Acta, 1968, 2, 31-51.
55 B. R. Whittlesey, Coord. Chem. Rev., 2000, 206-207, 395-418.
56 (a) G. Fachinetti, G. Fochi, T. Funaioli and P. F. Zanazzi, J. Chem. Soc., Chem. Commun., 1987, 89-90;

(b) C. Mealli, D. M. Proserpio, G. Fachinetti, T. Funaioli, G. Fochi and P. F. Zanazzi, Inorg. Chem., 1989,
28, 1122-1127.
57 G. Kong, G. N. Harakas and B. R. Whittlesey, J. Am. Chem. Soc., 1995, 117, 3502-3509.
58 C. Femoni, M. C. Iapalucci, G. Longoni and S. Zacchini, Eur. J. Inorg. Chem., 2009, 2487-2495.

68
Page 69 of 76 Chemical Society Reviews

59 M. Shieh and C.-C. Yu, J. Organomet. Chem., 2017, 849-850, 219-227.

60 M. Shieh, Y.-H. Liu, Y.-H. Li and R. Y. Lin, CrystEngComm, 2019, 21, 7341-7364.
61 G. Longoni, A. Ceriotti, R. Della Pergola, M. Manassero, M. Perego, G. Piro and M. Sansoni, Phil. Trans.

R. Soc. Lond. A, 1982, 308, 47-57.

62 F. A. Cotton and J. M. Troup, J. Chem. Soc., Dalton Trans., 1974, 800-802.
63 (a) C. H. Wei and L. F. Dahl, J. Am. Chem. Soc., 1969, 91, 1351-1361; (b) F. A. Cotton and J. M. Troup, J.

Am. Chem. Soc., 1974, 96, 4155-4159; (c) C. F. Campana, I. A. Guzei, E. G. Mednikov and L. F. Dahl, J.
Clust. Sci., 2014, 25, 205-224.
64 (a) P. J. Krusic, J. R. Morton, K. F. Preston, A. J. Williams and F. L. Lee, Organometallics, 1990, 9, 697-

700; (b) P. J. Krusic, J. San Filippo, B. Hutchinson, R. L. Hance and L. M. Daniels, J. Am. Chem. Soc.,
1981, 103, 2129-2131; (c) P. J. Krusic, J. Am. Chem. Soc., 1981, 103, 2131-2133.
65 F. Ragaini, J.-S. Song, D. L. Ramage, G. L. Geoffroy, G. A. P. Yap and A. L. Rheingold,

Organometallics, 1995, 14, 387-400.

66 (a) H. A. Hodali, D. F. Shriver and C. A. Ammlung, J. Am. Chem. Soc., 1978, 100, 5239-5240; (b) K. H.

Whitmire and D. F. Shriver, J. Am. Chem. Soc., 1981 103, 6754-6755.

67 (a) C. Femoni, M. C. Iapalucci, G. Longoni, S. Zacchini and S. Zarra, Inorg. Chem., 2009, 48, 1599-1605;

(b) C. Femoni, M. C. Iapalucci, G. Longoni and S. Zacchini, Dalton Trans., 2011, 40, 8685-8694.
68 M. Bortoluzzi, I. Ciabatti, C. Femoni, M. Hayatifar, M. C. Iapalucci, G. Longoni and S. Zacchini, Angew.

Chem. Int. Ed., 2014, 53, 7233-7237.

69 (a) M. R. Churchill, J. Wormald, J. Knight and M. J. Mays, J. Am. Chem. Soc., 1971, 93, 3073-3074; (b)

M. R. Churchill and J. Wormald, J. Chem. Soc., Dalton Trans., 1974, 2410-2415.

70 M. Bortoluzzi, I. Ciabatti, C. Cesari, C. Femoni, M. C. Iapalucci and S. Zacchini, Eur. J. Inorg. Chem.,

2017, 3135-3143.
71 A. Gourdon and Y. Jeannin, J. Organomet. Chem., 1985, 290, 199-211.
72 (a) J. S. Bradley, G. B. Ansell and E. W. Hill, J. Am. Chem. Soc., 1979, 101, 7417-7419; (b) J. S. Bradley,

S. Harris, J. M. Newsam, E. W. Hill, S. Leta and M. A. Modrick, Organometallics, 1987, 6, 2060-2069.

73 E. M. Holt, K. H. Whitmire and D. F. Shriver, J. Organomet. Chem., 1981, 213, 125-137.
74 S. Kuppuswamy, J. D. Wofford, C. Joseph, Z.-L. Xie, A. K. Ali, V. M. Lynch, P. A. Lindahl and M. J.

Rose, Inorg. Chem., 2017, 56, 5998-6012.

75 (a) C. Femoni, R. Della Pergola, M. C. Iapalucci, F. Kaswalder, M. Riccò and S. Zacchini, Dalton Trans.,

2009, 1509-1511; (b) R. Della Pergola, A. Sironi, L. Garlaschelli, D. Strumolo, C. Manassero, M.

Manassero, S. Fedi, P. Zanello, F. Kaswalder and S. Zacchini, Inorg. Chim. Acta, 2010, 363, 586-594; (c)
I. Ciabatti, C. Femoni, M. Hayatifar, M. C. Iapalucci, I. Maggiore, S. Stagni and S. Zacchini, J. Clust. Sci.,
2016, 27, 431-456.
76 (a) M. Tachikawa, R. L. Geerts and E. L. Muetterties, J. Organomet. Chem., 1981, 213, 11-24; (b) R.

Reina, L. Rodríguez, O. Rossell, M. Seco, M. Font-Bardia and X. Solans, Organometallics, 2001, 20,
1575-1579.
77 (a) C. Joseph, S. Kuppuswamy, V. M. Lynch and M. J. Rose, Inorg. Chem., 2018, 57, 20-23; (b) J.

McGale, G. E. Cutsail III, C. Joseph, M. J. Rose and S. DeBeer, Inorg. Chem., 2019, 58, 12918-12932.
78 L. Liu, T. B. Rauchfuss and T. J. Woods, Inorg. Chem., 2019, 58, 8271-8274.
79 A. Ceriotti, L. Resconi, F. Demartin, G. Longoni, M. Manassero and M. Sansoni, J. Organomet. Chem.,

1983, 249, C35-C37.

80 C. Femoni, M. C. Iapalucci, G. Longoni, S. Zacchini and E. Zazzaroni, Dalton Trans., 2007, 2644-2651.
81 V. G. Albano, C. Castellari, C. Femoni, M. C. Iapalucci, G. Longoni, M. Monari, M. Rauccio and S.

Zacchini, Inorg. Chim. Acta, 1999, 291, 372-379.

82 C. K. Schauer, S. Harris, M. Sabat, E. J. Voss and D. F. Shriver, Inorg. Chem., 1995, 34, 5017-2028.
83 C. K. Schauer, S. Harris, M. Sabat and D. F. Shriver, J. Am. Chem. Soc., 1989, 111, 7662-7664.
84 M. Shieh, C.-Y. Miu, Y.-Y. Chu, C.-N. Lin, Coord. Chem. Rev., 2012, 256, 637-694.
85 J. P. Shupp and M. J. Rose, Dalton Trans., 2020, 49, 23-26.
86 (a) D. E. Schipper, B. E. Young and K. H. Whitmire, Organometallics, 2016, 35, 471-483; (b) K. H.

Whitmire, M. R. Churchill and J. C. Fettinger, J. Am. Chem. Soc., 1985, 107, 1056-1057; (c) K. H.
Whitmire, C. B. Lagrone, M. R. Churchill, J. C. Fettinger and L. V. Biondi, Inorg. Chem., 1984, 23, 4227-
4232.
87 M. Ferrer, R, Reina, O. Rossell and M. Seco, Coord. Chem. Rev., 1999, 193-195, 619-642.

69
 Chemical Society Reviews Page 70 of 76

88 B. Berti, M. Bortoluzzi, C. Cesari, C. Femoni, M. C. Iapalucci, L. Soleri and S. Zacchini, Inorg. Chem.,
2020, 59, 15936-15952.
89 (a) M. Bortoluzzi, C. Cesari, I. Ciabatti, C. Femoni, M. Hayatifar, M. C. Iapalucci, R. Mazzoni and S.

Zacchini, J. Clust. Sci., 2017, 28, 703-723; (b) B. Berti, M. Bortoluzzi, C. Cesari, C. Femoni, M. C.
Iapalucci, R. Mazzoni, F. Vacca and S. Zacchini, Eur. J. Inorg. Chem., 2019, 3084-3093.
90 B. Berti, M. Bortoluzzi, C. Cesari, C. Femoni, M. C. Iapalucci, R. Mazzoni, F. Vacca and S. Zacchini,

Inorg. Chem., 2019, 58, 2911-2915.

91 (a) B. Berti, M. Bortoluzzi, C. Cesari, C. Femoni, M. C. Iapalucci, R. Mazzoni and S. Zacchini, Eur. J.

Inorg. Chem., 2020, 2191-2202; (b) B. Berti, M. Bortoluzzi, C. Cesari, C. Femoni, M. C. Iapalucci, R.
Mazzoni, F. Vacca and S. Zacchini, Inorg. Chem., 2020, 59, 2228-2240.
92 N. P. Mankad, Chem. Commun., 2018, 54, 1291-1302.
93 N. P. Mankad, Chem. Eur. J., 2016, 22, 5822-5829.
94 Y. Lakliang and N. P. Mankad, Organometallics, 2020, 39, 2043-2046.
95 C. Femoni, M. C. Iapalucci, G. Longoni, C. Tiozzo and S. Zacchini, Agew. Chem. Int. Ed. 2008, 47, 6666-

6669.
96 B. F. G. Johnson and J. Lewis, Inorg. Synth., 1972, 13, 92-95.
97 (a) F. Calderazzo and F. L’Eplattenier, Inorg. Chem., 1967, 6, 1220- 1224; (b) F. L’Eplattener, Inorg.

Chem., 1969, 8, 965-970; (c) B. F. G. Jonhson, J. Lewis and M. V. Twigg, J. Organomet. Chem., 1974,
67, C75-C76.
98 J. Lewis and P. R. Raithby, J. Organoment, Chem., 1995, 500, 227-237.
99 (a) C. R. Eady, B. F. G. Johnson and J. Lewis, J. Chem. Soc., Dalton Trans., 1977, 838-844; (b) S. A. R.

Knox, J. W. Koepke, M. A. Andrews and H. D. Kaesz, J. Am. Chem. Soc., 1975, 97, 3942-3947.
100 M. Yu. Afonin, B. Yu. Savkov, A. V. Virovets, V. S. Korenev, A. V. Golovin and V. A. Maksakov, Eur.

J. Inorg. Chem., 2017, 3105-3114.

101 B. F. G. Johnson, J. Lewis and D. A. Pippard, J. Chem. Soc., Dalton Trans., 1981, 407-412.
102 (a) S. P. Oh, Y.-Z. Li and W. K. Leong, J. Organomet. Chem., 2015, 783, 46-48; (b) R. D. Adams, Y.

Kan and Q. Zhang, J. Organomet. Chem., 2014, 751, 475-481.

103 (a) K. V. Kong, W. K. Leong, S. P. Ng, T. H. Nguyen and L. H. K. Lim, ChemMedChem, 2008, 3, 1269-

1275; (b) K, V. Kong, W. K. Leong and L. H. K. Lim, Chem. Res. Toxicol., 2009, 22, 1116-1122.
104 R. D. Adams, E. J. Kiprotich and M. D. Smith, Chem. Commun., 2018, 54, 3464-3467.
105 P. J. Bailey, B. F. G. Johnson and J. Lewis, Inorg. Chim. Acta, 1994, 227, 197-200.
106 J. P. Canal, A. A. Bengali, M. C. Jennings and R. K. Pomeroy, Inorg. Chem. Commun., 2014, 43, 31-34.
107 S. Takemoto and H. Matsuzaka, Coord. Chem. Rev., 2012, 256, 574-588.
108 M. I. Bruce, N. N. Zaitseva, B. W. Skelton and A. H. White, J. Chem. Soc., Dalton Trans., 2002, 3879-

3885.
109 M. P. Cifuentes, M. G. Humphrey, J. E. McGrady, P. J. Smith, R. Stranger, K. S. Murray and B.

Moubaraki, J. Am. Chem. Soc., 1997, 119, 2647-2655.

110 (a) B. F. G. Johnson, S. Hermans and T. Khimyak, Eur. J. Inorg. Chem., 2003, 1325-1331; (b) R. D.

Adams, M. D. Smith, J. D. Tedder and N. D. Wakdikar, Inorg. Chem., 2019, 58, 8357-8368.
111 (a) R. D. Adams, B. Captain, W. Fu, P. J. Pellechia and M. D. Smith, Angew. Chem. Int. Ed., 2002, 41,

1951-1953; (b) S. Saha, L. Zhu and B. Captain, Inorg. Chem., 2013, 52, 2526-2532.
112 (a) R. D. Adams, I. T. Horváth and H.-S. Kim, Organometallics, 1984, 3, 548-552; (b) R. D. Adams, I. T.

Horváth and P. Mathur, Organometallics, 1984, 3,623-630; (c) R. D. Adams and I. T. Horváth, Inorg.
Chem., 1984, 23, 4718-4722.
113 S.-P. Huang and M. G. Kanatzidis, J. Am. Chem. Soc., 1992, 114, 5477-5478.
114 M. Shieh, M.-H. Hsu, W.-S. Sheu, L.-F. Jang, S.-F. Lin, Y.-Y. Chu, C.-Y. Miu, Y.-W. Lai, H.-L. Liu and

J. L. Her, Chem. Eur. J., 2007, 13, 6605-6616.

115 (a) H. G. Ang, C. M. Hay, B. F. G. Johnson, J. Lewis, P. R. Raithby and A. J. Whitton, J. Organomet.

Chem., 1987, 330, C5-C11; (b) M. D. Randles, A. C. Willis, M. P. Cifuentes and M. G. Humphrey, J.
Organomet. Chem., 2007, 692, 4467-4472.
116 Y.-Z. Li, R. Ganguly, W. K. Leong and Y. Liu, Eur. J. Inorg. Chem., 2015, 3861-3872.
117 S. Saha, D. Isrow and B. Captain, J. Organomet. Chem., 2014, 751, 815-820.

70
Page 71 of 76 Chemical Society Reviews

118 (a) F.-E. Hong, T. J. Coffy, D. A. McCarthy and S. G. Shore, Inorg. Chem., 1989, 28, 3284-3285; (b) C.
E. Housecroft, D. M. Matthews, A. L. Rheingold and X. Song, J. Chem. Soc., Chem. Commun., 1992,
842-843.
119 J.-H. Chung, D. Knoeppel, D. McCarthy, A. Columbie and S. G. Shore, Inorg. Chem., 1993, 32, 3391-

3392.
120 R. D. Adams, J. Kiprotich, D. V. Peryshkov and Y. O. Wong, Chem. Eur. J., 2016, 22, 6501-6504.
121 R. D. Adams, Q. Zhang and X. Yang, J. Am. Chem. Soc., 2011, 133, 15950-15953.
122 K.-F. Yung and W.-T. Wong, Angew. Chem. Int. Ed., 2003, 42, 553-555.
123 T. Y. Garcia, J. C. Fettinger, M. M. Olmstead and A. L. Balch, Chem. Commun., 2009, 7143-7145.
124 (a) A. E. Carpenter, A. L. Rheingold and J. S. Figueroa, Organometallics, 2016, 35, 2309-2318; (b) A. E.

Carpentier, C. Chan, A. L. Rheingold and J. S. Figueroa, Organometallics, 2016, 35, 2319-2326.

125 F. L. Bowles, M. M. Olmstead, C. M. Beavers and A. L. Balch, Chem. Commun., 2013, 49, 5921-5923.
126 C. E. Plečnik, S. Liu, X. Chen, E. A. Meyers and S. G. Shore, J. Am. Chem. Soc., 2004, 126, 204-213.
127 (a) H. N. Adams, G. Fachinetti and J. Strähle, Angew. Chem. Int. Ed., 1980, 19, 404-405; (b) G.

Fachinetti, G. Fochi, T. Funaioli and P. F. Zanazzi, Angew. Chem. Int. Ed., 1987, 26, 680-681.
128 (a) V. G. Albano, P. Chini, S. Martinengo, M. Sansoni and D. Strumolo, J. Chem. Soc., Chem. Commun.,

1974, 299-300; (b) S. Martinengo, D. Strumolo, P. Chini, V. G. Albano and D. Braga, J. Chem. Soc.,
Dalton Trans., 1985, 35-41.
129 V. G. Albano, D. Braga, F. Grepioni, R. Della Pergola, L. Garlaschelli and A. Fumagalli, J. Chem. Soc.,

Dalton Trans. 1989, 879-883.

130 S. Kamiguchi and T. Chihara, J. Clust. Sci., 2000, 11, 483-492.
131 A. Ceriotti, R. Della Pergola, G. Longoni, M. Manassero, N. Masciocchi and M. Sansoni, J. Organomet.

Chem., 1987, 330, 237-252.

132 A. Arrigoni, A. Ceriotti, R. Della Pergola, G. Longoni, M. Manassero, N. Masciocchi and M. Sansoni,

Angew. Chem. Int. Ed., 1984, 23, 322-323.

133 A. Arrigoni, A. Ceriotti, R. Della Pergola, G. Longoni, M. Manassero and M. Sansoni, J. Organomet.

Chem., 1985, 296, 243-253.

134 I. Ciabatti, F. Fabrizi de Biani, C. Femoni, M. C. Iapalucci, G. Longoni and S. Zacchini, Dalton Trans.,

2013, 42, 9662-9670.

135 C. Femoni, M. C. Iapalucci, G. Longoni, J. Wolowska, S. Zacchini, P. Zanello, S. Fedi, M. Riccò, D.

Pontiroli and M. Mazzani, J. Am. Chem. Soc., 2010, 132, 2919-2927.

136 (a) I. Ciabatti, C. Femoni, M. Hayatifar, M. C. Iapalucci and S. Zacchini, Inorg. Chim. Acta, 2015, 428,

203-211; (b) I. Ciabatti, C. Femoni, M. Hayatifar, M. C. Iapalucci, A. Ienco, G. Longoni, G. Manca and S.
Zacchini, Inorg. Chem.2014, 53, 9761-9770; (c) M. Bortoluzzi, I. Ciabatti, C. Femoni, T. Funaioli, M.
Hayatifar, M. C. Iaplaucci, G. Longoni and S. Zacchini, Dalton Trans., 2014, 43, 9633-9646.
137 (a) A. Fumagalli, P. Ulivieri, M. Costa, O. Crispu, R. Della Pergola, F. Fabrizi de Biani, F. Laschi, P.

Zanello, P. Macchi and A. Sironi, Inorg. Chem., 2004, 43, 2125-2131; (b) A. Fumagalli, M. Costa, R.
Della Pergola, P. Zanello, F. Fabrizi de Biani, P. Macchi and A. Sironi, Inorg. Chim. Acta, 2003, 350, 187-
192.
138 C. S. Hong, L. A. Berben and J. R. Long, Dalton Trans., 2003, 2119-2120.
139 G. Ciani, A. Sironi, S. Martinengo, L. Garlaschelli, R. Della Pergola, P. Zanello, F. Laschi and N.

Masciocchi, Inorg. Chem., 2001, 40, 3905-3911.

140 C. Dreher, M. Zabel, M. Bodensteiner and M. Scheer, Organometallics, 2010, 29, 5187-5191.
141 (a) L. A. Hanlan and G. A. Ozin, J. Am. Chem. Soc., 1974, 96, 6324-6329; (b) J. L. Vidal and W. E.

Walker, Inorg. Chem., 1981, 20, 249-254.

142 (a) P. Chini and S. Martinengo, Inorg. Chim. Acta, 1969, 3, 315-318; (b) L. Garlaschelli, S. Martinengo,

P. L. Bellon, F. Demartin, M. Manassero, M. Y. Chiang, C.-Y. Wei and R. Bau, J. Am. Chem. Soc., 1984,
106, 6664-6667.
143 (a) G. Fachinetti, T. Funaioli and P. F. Zanazzi, J. Organomet. Chem., 1993, 460, C34-C36; (b) K. J.

Bradd, B. T. Heaton, J. A. Iggo, C. Jacob, J. T. Sampanthar and S. Zacchini, Dalton Trans., 2008, 685-
690.
144 (a) P. Chini and S. Martinengo, Chem. Commun., 1969, 1092-1093; (b) S. Martinengo, A. Fumagalli and

P. Chini, J. Organomet. Chem., 1985, 284, 275-279; (c) L. Malatesta, G. Caglio and M. Angoletta, J.
Chem. Soc., Chem. Commun., 1970, 532-533.

71
 Chemical Society Reviews Page 72 of 76

145 M. R. Churchill and J. P. Hutchinson, Inorg. Chem., 1978, 17, 3528-3535.

146 (a) D, Collini, F. Fabrizi de Biani, D. S. Dolzhnikov, C. Femoni, M. C. Iaplaucci, G. Longoni, C. Tiozzo,
S. Zacchini and P. Zanello, Inorg. Chem., 2011, 50, 2790-2798; (b) D. S. Dolzhnikov, M. C. Iapalucci, G.
Longoni, C. Tiozzo, S. Zacchini and C. Femoni, Inorg. Chem., 2012, 51, 11214-11216.
147 R. Della Pergola, F. Cea, L. Garlaschelli, N. Masciocchi and M. Sansoni, J. Chem. Soc., Dalton Trans.,

1994, 1501-1503.
148 R. Della Pergola, L. Garlaschelli, M. Manassero, N. Masciocchi and P. Zanello, Angew. Chem. Int. Ed.,

1993, 32, 1347-1349.

149 A. P. Palermo, C. Schöttle, S. Zhang, N. A. Grosso-Giordano, A. Okrut, D. A. Dixon, H. Frei, B. C. Gates

and A. Katz, Inorg. Chem., 2019, 58, 14338-14348.

150 (a) E. V. Grachova, P. Jutzi, B. Neumann and H.-G. Stammler, Dalton Trans., 2005, 3614-3616; (b) E. V.

Grachova and G. Linti, Eur. J. Inorg. Chem., 2007, 3561-3564.

151 (a) R. D. Adams, M. Chen, E. Trufan and Q. Zhang, Organometallics, 2011, 30, 661-664; (b) R. D.

Adams and M. Chen, Organometallics, 2012, 31, 445-450.

152 (a) R. D. Adams and M. Chen, J. Organomet. Chem., 2013, 733, 21-27; (b) R. D. Adams and G. Elpitiya,

Inorg. Chem., 2015, 54, 8042-8048.

153 Z. Li, C. Liu, J. Wu, Z. Lin and L. Xu, Dalton Trans., 2019, 48, 12013-12017.
154 (a) C. Femoni, I. Ciabatti, M. C. Iapalucci, S. Ruggieri, S. Zacchini, Progr. Nat. Sci: Mat. Internat., 2016,

26, 461-466; (b) C. Femoni, G. Bussoli, I. Ciabatti, M. Ermini, M. Hayatifar, M. C. Iapalucci, S. Ruggieri
and S. Zacchini, Inorg. Chem., 2017, 56, 6343-6351.
155 (a) A. Boccalini, P. J. Dyson, C. Femoni, M. C. Iapalucci, S. Ruggieri and S. Zacchini, Dalton Trans.,

2018, 47, 15737-15744; (b) C. Femoni, T. Funaioli, M. C. Iapalucci, S. Ruggieri and S. Zacchini, Inorg.
Chem., 2020, 59, 4300-4310.
156 J. L. Vidal, Inorg. Chem., 1981, 20, 243-249.
157 (a) J. L. Vidal, W. E. Walker, R. L. Pruett and R. C. Shoening, Inorg. Chem., 1979, 18, 129-136; (b) J. L.

Vidal, W. E. Walker and R. C. Shoening, Inorg. Chem., 1981, 20, 238-342.

158 L. Garlaschelli, A. Fumagalli, S. Martinengo, B. T. Heaton, D. O. Smith and L. Strona, J. Chem. Soc.,

Dalton Trans., 1982, 2265-2267.

159 V. G. Albano, P. Chini, S. Martinengo, D. J. A. McCaffrey, D. Strumolo and B. T. Heaton, J. Am. Chem.

Soc., 1974, 96, 8106-8107.

160 (a) B. T. Heaton, L. Strona, S. Martinengo, D. Strumolo, V. G. Albano and D. Braga, J. Chem. Soc.,

Dalton Trans., 1983, 2175-2182; (b) R. Della Pergola, A. Sironi, C. Manassero and M. Manassero, Eur.
J. Inorg. Chem., 2009, 4618-4621.
161 L. Cerchi, A. Fumagalli, S. Fedi, P. Zanello, F. Fabrizi de Biani, F. Laschi, L. Garlaschelli, P. Macchi and

A. Sironi, Inorg. Chem., 2012, 51, 9171-9180.

162 G. Ciani, D. M. Proserpio, A. Sironi, S. Martinengo and A. Fumagalli, J. Chem. Soc., Dalton Trans.,

1994, 471-475.
163 A. Fumagalli, S. Martinengo, G. Bernasconi, G. Ciani, D. M. Proserpio and A. Sironi, J. Am. Chem. Soc.,

1997, 119, 1450-1451.

164 (a) S. Martinengo, G. Ciani and A. Sironi, J. Am. Chem. Soc., 1982, 104, 328-330; (b) S. Martinengo, G.

Ciani and A. Sironi, J. Chem. Soc., Chem. Commun., 1984, 1577-1578.

165 W. Hieber, Adv. Organomet. Chem., 1970, 8, 1-28.
166 (a) G. Longoni, P. Chini and A. Cavalieri, Inorg. Chem., 1976, 15, 3025-3029; (b) G. Longoni and P.

Chini, Inorg. Chem., 1976, 15, 3029-3031; (c) A. Ceriotti, P. Chini, R. Della Pergola and G. Longoni,
Inorg. Chem., 1983, 22, 1595-1598.
167 (a) J. K. Beattie, A. F. Masters and J. T. Meyer, Polyhedron, 1995, 14, 829-868; (b) A. F. Masters and J.

T. Meyer, Polyhedron, 1995, 14, 339-365.

168 F. Demartin, M. C. Iapalucci and G. Longoni, Inorg. Chem., 1993, 32, 5536-5543.
169 C. Femoni, M. C. Iapalucci, G. Longoni, F. Ranuzzi, S. Zacchini, S. Fedi and P. Zanello, Eur. J. Inorg.

Chem., 2007, 4064-4070.

170 (a) D. Collini, C. Femoni, M. C. Iapalucci, G. Longoni, P. H. Svensson and P. Zanello, Angew. Chem. Int.

Ed., 2002, 41, 3685-3688; (b) N. T. Tran, M. Kawano, D. R. Powell and L. F. Dahl, J. Chem. Soc., Dalton
Trans., 2000, 4138-4144.

72
Page 73 of 76 Chemical Society Reviews

171 (a) C. Femoni, M. C. Iapalucci, G. Longoni, P. H. Svensson and J. Wolowska, Angew. Chem. Int. Ed.,
2000, 39, 1635-1637; (b) C. Femoni, M. C. Iapalucci, G. Longoni, P. H. Svensson, P. Zanello and F.
Fabrizi de Biani, Chem. Eur. J., 2004, 10, 2318-2326; (c) C. Femoni, M. C. Iapalucci, G. Longoni and P.
H. Svensson, Chem. Commun., 2001, 1776-1777.
172 (a) C. Femoni, M. C. Iapalucci, G. Longoni and P. H. Svensson, Chem. Commun., 2004, 2274-2275; (b) I.

Ciabatti, C. Femoni, M. C. Iapalucci, G. Longoni, S. Zacchini, S. Fedi and F. Fabrizi de Biani, Inorg.
Chem., 2012, 51, 11753-11761.
173 N. T. Tran, M. Kawano, D. R. Powell, R. K. Hayashi, C. F. Campana and L. F. Dahl, J. Am. Chem. Soc.,

1999, 121, 5945-5952.

174 (a) I. Ciabatti, C. Femoni, T. Funaioli, M. C. Iapalucci, S. Merighi and S. Zacchini, J. Organomet. Chem.,

2017, 849-850, 299-305; (b) C. Femoni, M. C. Iapalucci, G. Longoni, S. Zacchini, S. Fedi and F. Fabrizi
de Biani, Dalton Trans., 2012, 41, 4649-4663; (c) C. Femoni, M. C. Iapalucci, G. Longoni and S.
Zacchini, Chem. Commun., 2008, 3157-3159.
175 (a) A. Bernardi, I. Ciabatti, C. Femoni, M. C. Iapalucci, G. Longoni and S. Zacchini, J. Organomet.

Chem., 2016, 812, 229-239; (b) A. Bernardi, I. Ciabatti, C. Femoni, M. C. Iapalucci, G. Longoni and S.
Zacchini, Dalton Trans., 2013, 42, 407-421; (c) I. Ciabatti, C. Femoni, M. C. Iapalucci, A. Ienco, G.
Longoni, G. Manca and S. Zacchini, Inorg. Chem., 2013, 52, 10559-10565.
176 C. Cesari, I. Ciabatti, C. Femoni, M. C. Iapalucci and S. Zacchini, J. Clust. Sci., 2017, 28, 1963-1979.
177 (a) M. Bortoluzzi, I. Ciabatti, C. Femoni, M. Hayatifar, M. C. Iapalucci, G. Longoni and S. Zacchini,

Dalton Trans., 2014, 43, 13471-13475; (b) A. Bernardi, C. Femoni, M. C. Iapalucci, G. Longoni and S.
Zacchini, Dalton Trans., 2009, 4245-4251; (c) A. Bernardi, C. Femoni, M. C. Iapalucci, G. Longoni, F.
Ranuzzi, S. Zacchini, P. Zanello and S. Fedi, Chem. Eur. J., 2008, 14, 1924-1934.
178 (a) A. Bernardi, C. Femoni, M. C. Iapalucci, G. Longoni and S. Zacchini, Inorg. Chim. Acta, 2009, 362,

1239-1246; (b) A. Bernardi, C. Femoni, M. C. Iapalucci, G. Longoni, S. Zacchini, S. Fedi and P. Zanello,
Eur. J. Inorg. Chem., 2010, 4831-4842.
179 C. Femoni, M. C. Iapalucci, G. Longoni and S. Zacchini, Eur. J. Inorg. Chem., 2010, 1056-1062.
180 C. Femoni, M. C. Iapalucci, G. Longoni, S. Zacchini, I. Ciabatti, R. G. Della Valle, M. Mazzani and M.

Riccò, Eur. J. Inorg. Chem., 2014, 4151-4158.

181 (a) C. Capacci, I. Ciabatti, C. Femoni, M. C. Iapalucci, T. Funaioli, S. Zacchiuni and V. Zanotti, Inorg.

Chem., 2018, 57, 1136-1147; (b) C. Capacci, C. Cesari, C. Femoni, M. C. Iapalucci, F. Mancini, S.
Ruggieri and S. Zacchini, Inorg. Chem., 2020, 59, 16016-16026.
182 (a) C. Femoni, F. Kaswalder, M. C. Iapalucci, G. Longoni, M. Mehlstäubl and S. Zacchini, Chem.

Commun., 2005, 5769-5771; (b) C. Femoni, F. Kaswalder, M. C. Iaplaucci, G. Longoni, M. Mehlstäubl, S.

Zacchini and A. Ceriotti, Angew. Chem. Int. Ed., 2006, 45, 2060-2062; (c) C. Femoni, F. Kaswalder, M.
C. Iaplucci, G. Longoni and S. Zacchini, Eur. J. Inorg. Chem., 2007, 1483-1486.
183 C. Femoni, M. C. Iapalucci, G. Longoni, T. Lovato, S. Stagni and S. Zacchini, Inorg. Chem., 2010, 49,

5992-6004.
184 B. Berti, M. Bortoluzzi, A. Ceriotti, C. Cesari, C. Femoni, M. C. Iapalucci and S. Zacchini, Inorg. Chim.

Acta, 2020, 512, 119904.

185 (a) I. Ciabatti, C. Femoni, M. C. Iapalucci, G. Longoni, T. Lovato and S. Zacchini, Inorg. Chem., 2013,

52, 4384-4395; (b) I. Ciabatti, C. Femoni, M. C. Iapalucci, G. Longoni and S. Zacchini, Organometallics,
2013, 32, 5180-5189.
186 (a) C. Cesari, I. Ciabatti, C. Femoni, M. C. Iapalucci, F. Mancini and S. Zacchini, Inorg. Chem., 2017, 56,

1655-1668; (b) L. K. Batchelor, B. Berti, C. Cesari, I. Ciabatti, P. J. Dyson, C. Femoni, M. C. Iapalucci,

M. Mor, S. Ruggieri and S. Zacchini, Dalton Trans., 2018, 47, 4467-4477.
187 B. Berti, C. Cesari, F. Conte, I. Ciabatti, C. Femoni, M. C. Iapalucci, F. Vacca and S. Zacchini, Inorg.

Chem., 2018, 57, 7578-7590.

188 C. Femoni, F. Kaswalder, M. C. Iapalucci, G. Longoni and S. Zacchini, Chem. Commun., 2006, 2135-

2137.
189 M. Bortoluzzi, C. Cesari, I. Ciabatti, C. Femoni, M. C. Iapalucci and S. Zacchini, Inorg. Chem., 2017, 56,

6532-6544.
190 (a) A. Ceriotti, M. Daghetta, S. El Afefey, A. Ienco, G. Longoni, G. Manca, C. Mealli, S. Zacchini and S.

Zarra, Inorg. Chem., 2011, 50, 12553-12561; (b) M. Bortoluzzi, A. Ceriotti, I. Ciabatti, R. Della Pergola,
C. Femoni, M. C. Iapalucci, A. Storione and S. Zacchini, Dalton Trans., 2016, 45, 5001-5013.

73
 Chemical Society Reviews Page 74 of 76

191 (a) M. Bortoluzzi, A. Ceriotti, C. Cesari, I. Ciabatti, R. Della Pergola, C. Femoni, M. C. Iapalucci, A.
Storione and S. Zacchini, Eur. J. Inorg. Chem., 2016, 3939-3949; (b) B. Berti, M. Bortoluzzi, C. Cesari,
C. Femoni, M. C. Iapalucci and S. Zacchini, Inorg. Chim. Acta, 2020, 503, 119432.
192 D. Bonincontro, A. Lolli, A. Storione, A. Gasparotto, B. Berti, S. Zacchini, N. Dimitratos and S.

Albonetti, Appl. Catal. A, 2019, 588, 117279.

193 (a) C. Femoni, M. C. Iapalucci, G. Longoni, S. Zacchini and S. Zarra, J. Am. Chem. Soc., 2011, 133,

2406-2409; (b) I. Ciabatti, C. Femoni, M. C. Iapalucci, G. Longoni, S. Zacchini and S. Zarra, Nanoscale,
2012, 4, 4166-4177.
194 (a) E. Cattabriga, I. Ciabatti, C. Femoni, M. C. Iapalucci, G. Longoni and S. Zacchini, Inorg. Chim. Acta,

2018, 470, 238-249; (b) F. Gao, C. Li, B. T. Heaton, S. Zacchini, S. Zarra, G. Longoni and M. Garland,
Dalton Trans., 2011, 40, 5002-5008.
195 D. M. Washecheck, E. J. Wucherer, L. F. Dahl, A. Ceriotti, G. Longoni, M. Manassero, M. Sansoni and

P. Chini, J. Am. Chem. Soc., 1979, 101, 6110-6112

196 J. A. Widegren and R. G. Finke J. Mol. Catal. A: Chem. 2003, 198, 317–341.

197 L. Liu, T. J. Woods and T. B. Rauchfuss, Eur. J. Inorg. Chem., 2020, 2020, 3460–3465.

198 N. Queyriaux, D. Sun, J. Fize, J. Pécaut, M. J. Field, M. C-. Kerlidou and V. Artero, J. Am. Chem. Soc.,

2020, 142, 274−282.

199 R.M. Laine, J. Mol. Catal., 1982, 14, 137–169.

200 J. McGale, G. E. Cutsail, C. Joseph, M. J. Rose and S. DeBeer, Inorg. Chem., 2019, 58, 12918–12932.

201 G. O. Evans and C. J. Newell, Inorg. Chim. Acta, 1978, L387–L389.

202 H. C. Kang, C. H. Mauldin, T. Cole, W. Slegeir, K. Cann and R. Pettit, J. Am. Chem. Soc., 1977, 99,

8323–8325.
203 K. Cann, T. Cole, W. Slegeir and R. Pettit, J. Am. Chem. Soc., 1978, 100, 3969–3971.

204 D. J. Darensbourg, C. Ovalles and M. Pala, J. Am. Chem. Soc., 1983, 105, 5937–5939.

205 M. G. Thomas, B. F. Beier and E. L. Muetterties, J. Am. Chem. Soc., 1976, 98, 1296–1297.

206 C. P. Casey and C. R. Cyr, J. Am. Chem. Soc., 1973, 95, 2248–2253.

207 Y. Sunada, H. Kawakami, T. Imaoka, Y. Motoyama and H. Nagashima, Angew. Chem. Intl. Ed., 2009, 48,

9511–9514.
208 S. Zhou, K. Junge, D. Addis, S. Das and M. Beller, Angew. Chem., 2009, 121, 9671–9674.

209 S. Zhou, D. Addis, S. Das, K. Junge and M. Beller, Chem. Commun., 2009, 4883.

210 S. Zhou, S. Fleischer, K. Junge, S. Das, D. Addis and M. Beller, Angew. Chem. Intl. Ed., 2010, 49, 8121–

8125.
211 S. Enthaler, ChemCatChem, 2011, 3, 666–670.

212 Y.-Y. Li, S.-L. Yu, W.-Y. Shen and J.-X. Gao, Acc. Chem. Res., 2015, 48, 2587–2598.

213 Y. Li, S. Yu, X. Wu, J. Xiao, W. Shen, Z. Dong and J. Gao, J. Am. Chem. Soc., 2014, 136, 4031–4039.

214 (a) C.-N. Lin, C.-Y. Huang, C.-C. Yu, Y.-M. Chen, W.-M. Ke, G.-J. Wang, G.-A. Lee and M. Shieh,

Dalton Trans., 2015, 44, 16675–16679.; (b) M. Shieh, Y.-H. Liu, C.-C. Wang, S.-H. Jian, C.-N. Lin, Y.-
M. Chen and C.-Y. Huang, New J. Chem., 2019, 43, 11832–11843.
215 P. C. Ford, R. G. Rinker, C. Ungermann, R. M. Laine, V. Landis and S. A. Moya, J. Am. Chem. Soc.,

1978, 100, 4595–4597.

216 J. C. Bricker, C. C. Nagel, A. A. Bhattacharyya and S. G. Shore, J. Am. Chem. Soc., 1985, 107, 377–384.

217 H. S. Hilal, S. Khalaf and W. Jondi, J. Organometallic Chem., 1993, 452, 167–173.

218 T. Morimoto, N. Chatani and S. Murai, J. Am. Chem. Soc., 1999, 121, 1758–1759.

219 H. Nagashima, A. Suzuki, T. Iura, K. Ryu and K. Matsubara, Organometallics, 2000, 19, 3579–3590.

220 M. Igarashi, R. Mizuno and T. Fuchikami, Tet. Lett., 2001, 42, 2149–2151.

221 W. Imhof, A. Göbel, L. Schweda, D. Dönnecke and K. Halbauer, J. Organometallic Chem., 2005, 690,

3886–3897.
222 G. Süss-Fink and J. Reiner, J. Mol. Catal., 1982, 16, 231–242.

223 H. S. Hilal, W. Jondi, S. Khalaf and R. Abu-Halawa, J. Organomet. Chem.,1993, 452, 161–165.

74
Page 75 of 76 Chemical Society Reviews

224 K. G. Caulton, M. G. Thomas, B. A. Sosinsky and E. L. Muetterties, Proc. Natl. Acad. Sci., 1976, 73,
4274–4276.
225 R. D. Adams, Acc. Chem. Res., 1983, 16, 67–72.

226 R. C. Ryan and C. U. Pittman, J. Am. Chem. Soc., 1977, 99, 1986–1988.

227 T. Murai, T. Sakane and S. Kato, J. Org. Chem., 1990, 55, 449–453.

228 I. Ojima, R. J. Donovan and N. Clos, Organometallics, 1991, 10, 2606–2610.

229 I. Ojima, P. Ingallina, R. J. Donovan and N. Clos, Organometallics, 1991, 10, 38–41.

230 H. Yamazaki and P. Hong, J. Mol. Cat., 1983, 21, 133–150.

231 K. Kaneda, H. Kuwahara and T. Imanaka, J. Mol. Catal., 1994, 88, L267–L270.

232 S. Kawi, J. R. Chang and B. C. Gates, J. Am. Chem. Soc., 1993, 115, 4830–4843.

233 S. Kawi, J. R. Chang and B. C. Gates, J. Phys. Chem., 1994, 98, 12978–12988.

234 T. Yamamoto, T. Shido, S. Inagaki, Y. Fukushima and M. Ichikawa, J. Phys. Chem. B, 1998, 102, 3866–

3875.
235 S. Bhaduri, Current Science, 2000, 78, 1318.

236 S. Basu, H. Paul, C. Gopinath, S. Bhaduri and G. Lahiri, J. Catal., 2005, 229, 298–302.

237 Z. Zhang, F. A. P. Cavalcanti and W. M. H. Sachtler, Catal. Lett., 1992, 12 157–169

238 (a) C. R. Carr, A. Taheri and L. A. Berben, J. Am. Chem. Soc., 2020, 142, 12299–12305. (b) S.

Pattanayak, S.; L. A. Berben, ChemElectroChem. 2021, 8, DOI: doi.org/10.1002/celc.202100402

239 C. Costentin, S. Drouet, M. Robert, J.-M. Savéant, J. Am. Chem. Soc. 2012, 134, 11235–11242.

240 G. Gritzner and J. Kuta, Pure & appl. Chem., 1983, 56, 461–466
241 I. Noviandri, K. N. Brown, D. S.Fleming, P. T. Gulyas, P. A. Lay, A. F. Masters and L. Phillips, J. Phys.

Chem. B, 1999, 103, 6713.

242 (a) S. Roy, B. Sharma, J. Pécaut, P. Simon, M. Fontecave, P. D. Tran, E. Derat and V. Artero, J. Am.

Chem. Soc., 2017, 139, 3685–3696; (b) D. C. Cunningham and J. Y. Yang, Chem. Commun., 2020,
56,12965–12968.
243 A. Taheri, C. R. Carr and L. A. Berben, ACS Catal., 2018, 8, 5787–5793

244 (a) M. D. Rail and L. A. Berben, J. Am. Chem. Soc., 2011, 133, 18577–18579; (b) N. D. Loewen, T.

Neelakantan, L. A. Berben, Acc. Chem. Res. 2017,50, 2362-2370; (c) A. Taheri, N. D. Loewen, D. B.
Cluff, J. C. Fettinger, L. A. Berben, Organometallics, 2018, 37, 1087-1091.
245 A. Taheri, E. J. Thompson, J. C. Fettinger and L. A. Berben, ACS Catal., 2015, 5, 7140–7151.

246 (a) A. D. Nguyen, M. D. Rail, M. Shanmugam, J. C. Fettinger and L. A. Berben, Inorg. Chem., 2013, 52,

12847–12854; (b) A. Taheri, L. A. Berben, Inorg. Chem. 2016, 55, 378-385.

247 M. L. Helm, M. Stewart, R. M. Bullock, M. Rakowski DuBois, D. L. DuBois, Science, 2011, 333, 863-

866.
248 S. Ghosh and G. Hogarth, J. Organomet. Chem., 2017, 851, 57–67.

249 S. Ghosh, S. Basak-Modi, M. G. Richmond, E. Nordlander and G. Hogarth, Inorg. Chim. Acta, 2018, 480,

47–53.
250 A. Rahaman, G. C. Lisensky, J. Browder-Long, D. A. Hrovat, M. G. Richmond, E. Nordlander, and G.

Hogarth, Dalton Trans., 2020, 49, 7133–7143

251 M. Tachikawa, J. Stein, E. L. Muetterties, R. G. Teller, M. A. Beno, E. Gebert and J. M. Williams, J. Am.

Chem. Soc., 1980, 102, 6648–6649.

252 C. R. Carr, D. B. Cluff, L. A. Berben, Organometallics 2020, 39, 1658-1663.

253 N. D. Loewen, E. J. Thompson, M. Kagan, C. L. Banales, T. W. Myers, J. C. Fettinger and L. A. Berben,

Chem. Sci., 2016, 7, 2728–2735.

254 (a) N. D. Loewen and L. A. Berben, Inorg. Chem., 2019, 58, 16849–16857; (b) D. B. Cluff, A. Arnold, J.

C. Fettinger and L. A. Berben, Organometallics 2019, 38, 1230-1235

255 P. D. Jadzinsky, G. Calero, D. A. Ackerson, D. A. Bushnell and R. D. Kornberg, Science, 2007, 318, 430-

433.

75
 Chemical Society Reviews Page 76 of 76

256 N. D. Loewen, S. Pattanayak, R. Herber, J. C. Fettinger, L. A. Berben, L. A. J. Phys. Chem. Lett. 2021,
12, 3066-3073.
257 M. J. Drance, C. C. Mokhtarzadeh, M. Melaimi, D. W. Agnew, C. E. Moore, A. L. Rheingold, J. S.

Figueroa, Angew. Chem. Int. Ed., 2018, 57, 13057–13061.

258 S. Hossain, Y. Nihori, L. V. Nair, B. Kumar, W. Kurashige and Y. Negishi, Acc. Chem. Res., 2018, 51,

3114-3124.
259 S. Wang, Q. Li, X. Kang and M. Zhu, Acc. Chem. Res., 2018, 51, 2784-2792.
260 I. G. Powers and C. Uyeda, ACS Catal., 2017, 7, 936–958.
261 M. T. Nielsen, R. Padilla and M. Nielsen, J. Clust. Sci., 2020, 31, 11–61.

76