Unexpected fragmentations of triphosphaferrocene-formation of supramolecular assemblies containing the (1,2,4-P3C2Mes2)- ligand

Article abstract of DOI:10.1039/c5dt00303b

While reacting the sterically demanding triphosphaferrocene [CpFe(η⁵-P₃C₂Mes₂)] (1) with Cu(i) halides, the sandwich complex undergoes an unprecedented fragmentation into decamethylferrocene, FeX₂ (X

Full text of DOI:10.1039/c5dt00303b

Dalton
Transactions
View Article Online
View Journal
PAPER
Unexpected fragmentations of triphosphaferrocene –
formation of supramolecular assemblies
Cite this: DOI: 10.1039/c5dt00303b
containing the (1,2,4-P3C2Mes2)− ligand†‡
Claudia Heindl,a Andrea Kuntz,a Eugenia V. Peresypkina,b,c Alexander V. Virovets,b,c
Manfred Zabel,a David Lüdeker,d Gunther Brunklausd and Manfred Scheer*a
While reacting the sterically demanding triphosphaferrocene [Cp*Fe(η5-P3C2Mes2)] (1) with Cu(I) halides,
the sandwich complex undergoes an unprecedented fragmentation into decamethylferrocene, FeX2 (X =
Cl, Br, I) and [P3C2Mes2]− units. Subsequently, these phospholyl ligands act as versatile, negatively charged
building blocks for the formation of supramolecular aggregates representing the monomeric, dimeric and
polymeric (1D and 2D) coordination compounds [(P3C2Mes2)2{Cu7(CH3CN)7(μ4-X)(μ3-X)2(μ-X)}{Cu2(μ2-
X)2X}{Cu(CH3CN)(μ2-X)}]2·6CH3CN (2·6CH3CN:
X = Cl, 3·6CH3CN: X = Br), [(P3C2Mes2)2{Cu-
(CH3CN)}6(μ-Br)2(μ3-Br)2{Cu(CH3CN)2Br}2]·CH3CN (4a·CH3CN), [(P3C2Mes2)4{Cu5(CH3CN)5(μ2-Br)}{Cu-
(CH3CN)2CuBr2}2{Cu(CH3CN)2}]n+[CuBr2]n−·2CH3CN (5·2CH3CN), [(P3C2Mes2){Cu(CH3CN)(μ-I)}4{Cu-
(CH3CN)3}]·0.5C7H8·2.5CH3CN (6·0.5C7H8·2.5CH3CN), [(P3C2Mes2)Cu7(CH3CN)4(μ4-I)2(μ3-I)2(μ-I)2]x·2C7H8
(7·2C7H8), [(P3C2Mes2){Cu(CH3CN)3}2{Cu(μ-I)}6]·0.5CH2Cl2·3CH3CN (8·0.5CH2Cl2·3CH3CN) and [Cp*Fe-
(CH3CN)3]n+[(P3C2Mes2)2{Cu(CH3CN)2}{Cu(μ-I)}6]n−·0.6CH2Cl2 (9·0.6CH2Cl2) with rather non-typical
structural motifs within the large varieties of copper halide chemistry. Besides the X-ray structural analyses
the obtained assemblies were also characterized in solution in which they undergo fragmentation and
re-aggregation processes.
Received 21st January 2015,
Accepted 24th February 2015
DOI: 10.1039/c5dt00303b
www.rsc.org/dalton
gives rise to the class of phosphaferrocenes.3 These 18 VE
sandwich complexes turned out to be very stable and, in con-
Introduction
Ferrocene [Fe(η5-C5H5)2] is one of the fundamental molecules trast to ferrocene, the lone pairs of the P atoms enable them
in organometallic chemistry and until now the flood of publi- to be excellent building blocks in coordination and supra-
cations devoted to its chemical properties has not stopped. molecular chemistry.4 Therefore new perspectives for the
Apart from its use as a reference redox system (Fc/Fc+),1 synthesis of phosphorus-based oligomers and polymers
chiral ferrocene-based ligands represent important classes of open up.
auxiliaries in asymmetric homogeneous catalysis.2 Based on
For the aggregation of these organometallic moieties we
the isolobal principle the substitution of one up to six have broadly used Cu(I) halides.5 To combine the central role
methine moieties by phosphorus atoms is possible, which of copper in catalysis6 with the potential of phosphaferrocenes
in this area,3b it is of special interest to accumulate copper
halides using phosphaferrocenes. For such a purpose the
1,2,4-triphosphaferrocene [CpRFe(η5-P3C2 Bu2)] (CpR = Cp, Cp*,
t
aInstitut für Anorganische Chemie, Universität Regensburg, Universitätsstr. 31,
t
93053 Regensburg, Germany. E-mail: Manfred.Scheer@ur.de
bA. V. Nikolaev Institute of Inorganic Chemistry, SB RAS, Ak. Lavrentiev prosp. 3,
Cp′′′; Cp* = C5Me5; Cp′′′ = C5H2 Bu3) with three possible
coordination sites is very well suited. Recently, we have shown
Novosibirsk 630090, Russia
that its coordination behaviour strongly depends on the steric
demand of the adjacent CpR ring and the nature of the halide.7
A selection of the obtained structural motifs is shown in
Fig. 1. Using less bulky CpR ligands (CpR = Cp, Cp*) the coordi-
nation of two or even three phosphorus atoms leads to the for-
mation of dimeric (A) and polymeric compounds (B and C).7a,b
However, the use of Cp′′′ leads to a much higher steric
cNovosibirsk State University, Pirogova 2, Novosibirsk 630090, Russia
dInstitut für Physikalische Chemie, Westfälische Wilhems-Universität Münster,
Corrensstr. 28, 48149 Münster, Germany
†In memory of Professor Kenneth Wade.
‡Electronic supplementary information (ESI) available: Experimental part, crys-
tallographic data and additional figures. CCDC 1043724–1043731. For ESI and
crystallographic data in CIF or other electronic format see DOI: 10.1039/
c5dt00303b
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Fig. 1 Coordination compounds of triphosphaferrocenes [CpRFe(η5-P3C2tBu2)] (CpR = Cp, Cp*, Cp’’’) and Cu(I) halides.
demand of the triphosphaferrocene, hence different products
could be obtained. Besides a complex similar to A being
Results and discussion
Coordination behaviour of 1 towards Cu(I) halides –
formed, two unexpected fragmentation reactions of the
t
t
unexpected fragmentation
P3C2 Bu2 ring took place. By treating [Cp′′′Fe(η5-P3C2 Bu2)] with
an excess of CuCl a polymeric chain containing triple decker The self-assembly process of 1 and CuX (X = Cl, Br, I) leads to
moieties [(FeCp′′′)2(µ,η4:η4-P4)] is formed (D).7c In D a C2P different coordination compounds depending on the con-
moiety was formally replaced by one phosphorus atom, hence ditions applied. Surprisingly, building block 1 always under-
in the resulting diphosphacyclobutadiene ring in E a formal went fragmentation. Out of the fragmentation products, only
elimination of a P atom occurred.7a
the 1,2,4-triphospholyl ligand remains in the isolated pro-
Since the substitution pattern of the CpR ligand seems to ducts, hence the [Cp*Fe]+ unit must have been split off. These
play a determining role, the question arises, whether a change fragments together with remaining halide anions might
in the steric demand of the triphospholyl ligand itself would initially form the dimeric complex [Cp*Fe(μ-X)]2 (X = Cl, Br, I),
cause consequences. Hence, instead of tert-butyl groups, we which displays a known intermediate and can even be isolated,
decided to use [Cp*Fe{η5-(1,2,4-P3C2Mes2)}] (Mes = 2,4,6-tri- when CpR ligands with a higher steric demand are used.9
methylphenyl) (1)8 with two sterically even more demanding Since this compound is coordinatively unsaturated, it, in all
mesityl groups next to the phosphorus atoms and explore its probability, dissociates into [Cp*2Fe] and FeBr2. To support
coordination behaviour towards Cu(I) halides.
this assumption, the reaction mixtures were analysed by 1H
Herein we report the unexpected fragmentation of 1 by the NMR spectroscopy and mass spectrometry. In fact, a singlet at
1
reaction with CuX (X = Cl, Br, I) yielding a large diversity of 1.64 ppm in the H NMR spectra can be assigned to [Cp*2Fe]
unprecedented coordination compounds and show the useful- and the EI mass spectra display peaks corresponding to deca-
−
ness of the formed P3C2Mes2 five-membered ring as a build- methylferrocene as well. In addition, the negative ion ESI mass
ing block for extended structures in:
spectra of the mother liquors exhibit peaks for [FeCl3]− and
[(μ,η1:η2:η2-P3C2Mes2)(μ,η1:η3:η3-P3C2Mes2){Cu7(CH3CN)7- [FeBr3]−, respectively.
(μ4-X)(μ3-X)2(μ-X)}{Cu2(μ2-X)2X}{Cu(CH3CN)(μ2-X)}]2·6CH3CN
(2·6CH3CN: X = Cl, 3·6CH3CN: X = Br),
Among the stable phosphaferrocenes the rather uncommon
fragmentations are related to the cleavage of C–P bonds in the
[(μ,η1:η2:η3-P3C2Mes2)2{Cu(CH3CN)}6(μ-Br)2(μ3-Br)2{Cu(CH3CN)2- phospholyl ligands (D and E). Hence, the unexpected instabi-
Br}2]·CH3CN (4a·CH3CN),
lity of 1 is unique and can most likely be attributed to the
[(μ,η1:η1:η2-P3C2Mes2)3(μ,η1:η2:η3-P3C2Mes2){Cu5(CH3CN)5- influence of the bulky mesityl ligand. This is confirmed by the
(μ2-Br)}{Cu(CH3CN)2CuBr2}2{Cu(CH3CN)2}]n+[CuBr2]n−·2CH3CN existence of a stable [P3C2Mes*2] (Mes* = 2,4,6-tri-tert-butyl-
(5·2CH3CN),
phenyl) radical, reported by Ionkin et al., by using the similar,
[(μ,η1:η2:η2-P3C2Mes2){Cu(CH3CN)(μ-I)}4{Cu(CH3CN)3}]·0.5C7H8· sterically even more demanding Mes* moiety.10
2.5CH3CN (6·0.5C7H8·2.5CH3CN),
[(μ,η1:η2:η2-P3C2Mes2)Cu7(CH3CN)4(μ4-I)2(μ3-I)2(μ-I)2]x·2C7H8
Solid-state characterization of the coordination compounds
(7·2C7H8),
All isolated coordination compounds contain the remaining
(μ,η1:η3:η3-P3C2Mes2){Cu(CH3CN)3}2{Cu(μ-I)}6]·0.5CH2Cl2· negatively charged [P3C2Mes2]− unit as a building block. So
3CH3CN (8·0.5CH2Cl2·3CH3CN) and
far, the cyclic 1,2,4-triphospholyl ligand was mainly used for
the synthesis of (half-)sandwich complexes.11 Its coordination
behaviour towards Lewis acids, especially coinage metal salts,
[Cp*Fe(CH3CN)3]n+[(μ,η1:η3:η3-P3C2Mes2)2{Cu(CH3CN)2}-
{Cu(μ-I)}6]n−·0.6CH2Cl2 (9·0.6CH2Cl2).
Dalton Trans.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Paper
has been rather neglected, although interesting coordination rally coordinated copper ions. The unique P atom of the peri-
modes to gold centres have been observed by the reaction with pheral [P3C2Mes2] fragment is additionally coordinated at
[(PPh3)AuCl] and [(PEt3)AuCl].12 However concerning Cu, to each side of the dimer by a terminal Cu2Cl3 fragment, in
the best of our knowledge, only two reactions are known: which the threefold-coordinated copper ions are in a triangu-
t
Nixon et al. treated K[P3C2 Bu2] with Cu2I2 in the presence lar planar environment. In total, four [P3C2Mes2]− rings and
t
of PMe3, yielding the binuclear complexes [{P3C2 Bu2}- 16Cl− balance the positive charge of 20Cu+. The central struc-
t
{Cu(PMe3)2}2I] and [{P3C2 Bu2}2{Cu(PMe3)2}2].12b Zenneck et al. tural motif can be described as an unprecedented cage of 15
t
succeeded in the synthesis of the versatile [{η5-(P3C2 Bu2)}Cu- inorganic atoms (4P, 7Cu, 4Cl), consisting of one Cu ion and
(PPh3)] complex.13
three Cu2 dimers (2.546(1)–2.617(1) Å), connected by one μ2-Cl,
In general, Cu(I) halides display versatile building blocks in two μ3-Cl and one μ4-Cl (2.303(1)–2.840(1) Å). The unique phos-
coordination chemistry, and therefore a large variety of phorus atom of the [P3C2Mes2]− ligand is η1-coordinated to
different coordination compounds can be obtained. Since the copper (2.173(1)–2.153(1) Å), while the adjacent P atoms each
structural motifs differ when different halides are used, they show a η2- and η3-coordination with longer Cu–P distances to
will be described separately.
two Cu2 dimers and two Cu2 dimers and one unique Cu ion,
respectively. The Cu–P distances involving Cu2 dimers
(2.221(1)–2.406(1) Å) are systematically shorter than those
of the unique Cu ion (2.419(1)–2.501(1) Å) (see also ESI‡
and Table 1). The P3C2 rings of the phospholyl ligands are
planar (deviation 0.01°) with the mesityl ligands rotated by
82.6° and 72.6° at the peripheral phospholyl rings and by
89.1° and 79.9° in the central ones, respectively. Thus, the
rotation is much more distinctive than in 1 (46.07(1)° and
40.85(1)°, respectively),8 most likely due to the absence of
a Cp* ligand.
CuCl-containing compound (2)
The reaction of 1 with CuCl selectively leads to the form-
ation of the dimeric complex [(μ,η1:η2:η2-P3C2Mes2)(μ,η1:η3:η3-
P3C2Mes2){Cu7(CH3CN)7(μ4-Cl)(μ3-Cl)2(μ-Cl)}{Cu2(μ2-Cl)2Cl}-
{Cu(CH3CN)(μ2-Cl)}]2 (2), which can be isolated as solvate
2·6CH3CN in quantitative yields (Fig. 2c).
Compound 2·6CH3CN displays a dimer of two central
[P3C2Mes2]2Cu7Cl4 cages (Fig. 2a and b), which are bridged by
a {Cu(CH3CN)}2Cl2 four-membered ring formed by tetrahed-
CuBr-containing compounds (3–5)
The diffusion of a CH3CN solution of CuBr into a CH2Cl2 solu-
tion of 1 leads to the crystallization of three different coordi-
nation compounds depending on the applied conditions:
[(μ,η1:η2:η2-P3C2Mes2)(μ,η1:η3:η3-P3C2Mes2){Cu7(CH3CN)7(μ4-Br)-
(μ3-Br)2(μ-Br)}{Cu2(μ2-Br)2Br}{Cu(CH3CN)(μ2-Br)}]2·6CH3CN
(3·6CH3CN),
[(μ,η1:η2:η3-P3C2Mes2)2{Cu(CH3CN)}6(μ2-Br)2-
(μ3-Br)2{Cu(CH3CN)2Br}2]·CH3CN (4a·CH3CN) and [(μ,η1:η1:η2-
P3C2Mes2)3(μ,η1:η2:η3-P3C2Mes2){Cu5 (CH3CN)5(μ2-Br)}{Cu(CH3CN)2-
CuBr2}2{Cu(CH3CN)2}]n+[CuBr2]n−·2CH3CN (5·2CH3CN). If
a
molar ratio of 1 : CuBr = 1 : 2 is used, compound 5·2CH3CN is
formed selectively, which is in agreement with the stoichio-
metric ratio of (P3C2Mes2) : Cu = 1 : 2 in 5·2CH3CN. An excess
of CuBr (1 : 5 or 1 : 10) therefore enables the assembly of 3
and 4. However, since the ratio (P3C2Mes2) : Cu is the same in
3 and 4 (1 : 5), a selective synthesis is rather difficult. Nonethe-
less a preferred crystallization of 4·CH3CN is achieved by apply-
ing more diluted and pure CH3CN solutions, while the use of
more concentrated and CH2Cl2–CH3CN solvent mixtures
Fig. 2 (a) ‘Cage’ motif in 2·6CH3CN (X = Cl) and 3·6CH3CN (X = Br), (b)
section of the molecular structure of 2·6CH3CN illustrating the ‘cage’
motif, (c) molecular structure of the dimer of 2·6CH3CN. H atoms and
solvent molecules are omitted for clarity.
Table 1 Selected bond lengths [Å] of 2–9. Ranges of bond lengths are given whenever more than one bond is present in the asymmetric unit
X
Cu⋯Cu
Cu–Punique
Cudimer–P
Cuunique–P
Cu–X
2
3
4a
5
6
7
8
9
Cl
Br
Br
Br
I
I
I
2.564(1)–2.617(1)
2.531(1)–2.599(1)
2.533(1)
2.527(2)–2.663(2)
2.556(1)–2.561(1)
2.688(2)–2.876(1)
2.471(3)–2.553(3)
2.454(2)–2.535(2)
2.153(1)–2.173(1)
2.169(2)–2.189(2)
2.219(2)
2.239(2)–2.282(2)
2.217(1)
2.191(3)
2.202(4)–2.218(4)
2.256(2)
2.221(1)–2.406(1)
2.230(2)–2.386(2)
2.202(2)–2.521(2)
2.239(2)–2.538(2)
2.292(1)–2.313(1)
2.235(2)–2.428(3)
2.304(4)–2.820(4)
2.251(2)–2.778(2)
2.419(1)–2.501(1)
2.420(2)–2.514(2)
2.202(2)–2.521(2)
2.266(3)–2.373(3)
—
—
—
—
2.303(1)–2.840(1)
2.291(1)–2.660(1)
2.431(1)–2.592(1)
2.295(2)–2.350(2)
2.633(1)–2.694(1)
2.657(2)–2.864(2)
2.550(2)–2.627(2)
2.537(1)–2.617(1)
I
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Fig. 3 (a) ‘Cage’ motif in 4a·CH3CN, (b) molecular structure of
4a·CH3CN, (c) section of the polymeric structure of 4a·CH3CN. H atoms
and solvent molecules are omitted for clarity.
Fig. 4 (a) ‘Cage’ motif in 5·2CH3CN, (b) section of the polymeric struc-
ture of 5·2CH3CN illustrating the mesh-like structure (view along the
crystallographic a-axis). The anions are displayed in the space-filling
model, the network in ‘wires and sticks’. H atoms, mesityl ligands and
solvent molecules are omitted for clarity. (c) Repeating unit in
5·2CH3CN. H atoms, mesityl ligands and solvent molecules are omitted
for clarity. (d) Schematic representation of the cube-like arrangement of
a single layer in 5·2CH3CN.
favours the crystallization of 3·6CH3CN. Compound 3·6CH3CN
is isostructural to the Cl derivative 2·6CH3CN (Fig. 2). The
bond distances in 3·6CH3CN remain in a similar range despite
the presence of the larger halogen anion (see Table 1).
Surprisingly, the X-ray structural analysis of 4·CH3CN
reveals a solid solution of different compounds attended with
partial occupancies and complex disorder of the Cu, Br and
CH3CN positions. The monomeric compound [{μ,η1:η2:η3-
(P3C2Mes2)}2{Cu(CH3CN)}6(μ2-Br)2(μ3-Br)2{Cu(CH3CN)2Br}2]
(4a·CH3CN) displays the structural fragment with major occu-
pation (70%) (Fig. 3b). The central cage motif in 4a·CH3CN is
structurally related to that of the dimer 3·6CH3CN with one
demonstrates the versatility of the phospholyl ligand and CuBr
in coordination chemistry. The central core contains four
[P3C2Mes2] units (Fig. 4a), whose adjacent P atoms coordinate
two Cu2 dimers (2.527(2)–2.663(2) Å) and one single Cu in
either an η1 (3P; 2.239(2)–2.282(2) Å), η2 (4P; 2.282(2)–2.538(2) Å)
or an η3 (1P; 2.331(2)–2.469(2) Å) coordination mode. In
addition, one Br bridges two Cu atoms with a distance of 2.295(2)–
2.350(2) Å, resulting in a distorted tetrahedral environment
for these Cu. One of the four unique P atoms is linked to a
terminal {CuBr2} unit with a trigonal planar Cu, which blocks
further growth into this direction, while the other three coordi-
nate a {Cu(CH3CN)2} unit, whereby polymerization is enabled.
The repeating unit (two ‘cages’ plus the linking and terminal
moieties) therefore consists of eight [P3C2Mes2]−, six Br− and
15Cu+. Its remaining charge is balanced by a linear dibromo-
cuprate counter ion [CuBr2]− (Fig. 4c). The 2D network
of 5·2CH3CN can be described as a sheet-like structure
with porous layers. A single layer consists of ‘cubes’ out of
four repeating units stringing together (Fig. 4d). Since in one
direction polymerization is hindered due to the terminal
{CuBr2} unit, four edges per cube are absent. The view along
the crystallographic a-axis illustrates the resulting mesh-like
structure (Fig. 4b). However, these meshes display no
empty voids, but are occupied by the mesityl ligands and the
counter anion.
missing Cu(CH3CN) unit to give
a
formally neutral
{(P3C2Mes2)2Cu6Br4} cage (Fig. 3a). This difference influences
the connectivity and bond lengths, hence the Br ions connect
the Cu ions in a μ2-(2Br) or a μ3-fashion (2Br) (2.431(1)–2.592
(1) Å), respectively. The coordination mode of the adjacent P
atoms to Cu is also different, one is η2-, and the other is η3-co-
ordinated within a more extensive range of distances (2.202
(2)–2.521(2) Å). In 4a·CH3CN only one Cu2 dimer is present
with a distance of 2.533(1) Å compared to three dimers in 3.
The unique phosphorus atoms each coordinate
a {Cu-
(CH3CN)2Br} unit instead of the Cu2Br3 unit in 3·6CH3CN.
Other similar structural fragments of the solid solution
4·CH3CN with minor occupation factors are possible (for more
information see ESI‡), among them even the polymeric com-
pound [(P3C2Mes2)2Cu11Br9(MeCN)7]x (4b·CH3CN) (occupation
factor 10%) containing a linear coordinated Cu (Cu–Br: 2.226(4) Å)
(Fig. 3c) co-exists.
While with CuCl the used molar ratio does not influence
product formation, with CuBr 5·2CH3CN can be obtained
selectively and reproducibly, when less than five equivalents of
CuBr are used. Its structural analysis reveals a 2D network,
which is rather astonishing in consideration of the bulky
mesityl substituents. A totally different cage motif again
CuI-containing compounds (6–9)
By combining 1 and CuI, products with two different very rare
structural motifs of CuI units can be realized, the ‘crown’
Dalton Trans.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Paper
Fig. 6 Section of the polymeric structure of 7·2C7H8. H atoms, disorder
and solvent molecules are omitted for clarity.
Table 1). The tetrahedral coordination of Cu is accomplished
by CH3CN ligands. The presence of five Cu+, four I− and one
[P3C2Mes2]− ligand affords charge balance.
By formally replacing three CH3CN ligands at the terminal
Cu ion by two CuI fragments, 6 serves as a repeating unit for
the 1D polymer 7·2C7H8 (Fig. 5c and 6). The geometry of the
‘crown’ fragment retains similar Cu–P distances (2.235(2)–
Fig. 5 (a) ‘Crown’ structural motif in 6·0.5C7H8·2.5CH3CN and 7·2C7H8,
(b) molecular structure of 6·0.5C7H8·2.5CH3CN, (c) repeating unit of the
polymeric structure of 7. H atoms and the minor position of the dis-
ordered mesityl ligand are omitted for clarity.
(Fig. 5a) and the ‘hexagram’ motif (Fig. 7a). Both are found 2.428(3) Å) and Cu–I (2.657(2)–2.864(2) Å), but the Cu⋯Cu sep-
either in a monomeric coordination compound or in a 1D arations became significantly more uniform, 2.688(2) and
polymer: [(μ,η1:η2:η2-P3C2Mes2){Cu(CH3CN)(μ2-I)}4{Cu(CH3CN)3}]· 3.664(2) Å (Fig. 7S,‡ and Table 1). To bridge these units, an
0.5C7H8·2.5CH3CN (6·0.5C7H8·2.5CH3CN), [(μ,η1:η2:η2-P3C2Mes2)- additional five-membered Cu3I2 ring is formed, and another
Cu7(CH3CN)4(μ4-I)2(μ3-I)2(μ2-I)2]x·2C7H8 (7·2C7H8), [(μ,η1:η3:η3- Cu2 dimer (Cu⋯Cu 2.876(1) Å) coordinates an iodide of the
P3C2Mes2){Cu(CH3CN)3}2{Cu(μ2-I)}6]·0.5CH2Cl2·3CH3CN ‘crown’. In 7, the iodine atoms do not bridge only two Cu ions
(8·0.5CH2Cl2·3CH3CN) and [Cp*Fe(CH3CN)3]n+[(μ,η1:η3:η3- as in 6·0.5C7H8·2.5CH3CN, but show a μ3 and μ4-coordination
P3C2Mes2)2{Cu(CH3CN)2}{Cu(μ2-I)}6]n−·0.6CH2Cl2 (9·0.6CH2Cl2). mode. A neutral coordination polymer results, since the
A controlled and selective synthesis of one specific compound repeating unit contains seven Cu+, six I− and one [P3C2Mes2]−
is a big challenge due to similar molar ratios ((P3C2Mes2) : Cu ligand. The rotation of the mesityl ligands with respect to the
= 1 : 5, 1 : 7, 1 : 8 and 1 : 7 in 6, 7, 8 and 9, respectively) in the phospholyl plane is comparable to 2·6CH3CN with angles of
products. However, it has never occurred that products 89.5° in 7·2C7H8 and 84.1° and 83.8° in 6·0.5C7H8·2.5CH3CN.
with different structural motifs crystallize out of one
The Cu4I4 ‘crown’ motif in 6 and 7 is uncommon, since this
sample. Hence, either the crystallization of one single com- number of CuI units usually prefers a cubane- or step-like
pound or of mixtures of 6·0.5C7H8·2.5CH3CN and 7·2C7H8 or arrangement.14 Its formation was only observed once by Sugi-
8·0.5CH2Cl2·3CH3CN and 9·0.6CH2Cl2 can be obtained, moto et al., and in this case the copper ions are coordinated
respectively. The crystallization of solely
obtained when a pure CH3CN solution is used. In conclusion,
7
(polymer) is by sulfur atoms.15
Among the large variety of CuX (X = Cl, Br, I) coordination
pure 6·0.5C7H8·2.5CH3CN (monomer) is obtained as the only compounds in the literature, only one example is known for
crystalline product applying CH2Cl2/CH3CN solvent mixtures the ‘hexagram’ structural motif: [{(triphos)-CoP3}2{CuBr}6]
or more diluted conditions, which has been proved reproduci- (triphos = MeC(CH2PPh2)3).16 Hence, 8 and 9 represent the
ble. In addition, an excess of 1 is conducive, since in first examples with an iodine-containing hexagonal core.
6·0.5C7H8·2.5CH3CN the lowest Cu content is present. Unfortu-
The central almost planar Cu6 ring (deviation 0.12 Å) is co-
nately, the formation of 9 has only been observed on very rare ordinated by two phospholyl ligands, which are rotated to each
occasions together with 8·0.5CH2Cl2·3CH3CN. Several attempts other by 66.6° (Fig. 7b). Each of the adjacent P atoms shows a
to reproduce this compound only resulted in the selective iso- η3-coordination mode towards Cu with bond lengths of 2.304(4)–
lation of 8·0.5CH2Cl2·3CH3CN, independent of the solvents 2.820(4) Å. As in the case of the Cu2 dimers in 2–7, Cu⋯Cu
and crystallization procedures.
interactions may be considered in 8·0.5CH2Cl2·3CH3CN as
The ‘crown’ motif in 6·0.5C7H8·2.5CH3CN consists of an well, since the Cu⋯Cu distances are in the range of 2.471(3)–
eight-membered Cu4I4 ring which is significantly distorted by 2.553(3) Å.17 Six I ions bridge these edges in a μ2-fashion
the presence of two opposed Cu2 dimers (2.556(1)–2.561(1) Å) (2.550(2)–2.627(2) Å) so that a distorted ‘hexagram’ structural
which are separated from one another by 4.030(1)–4.032(1) Å motif is formed. This core structure is completed by two term-
(Fig. 5b). The copper atoms of each dimer are coordinated by inal {Cu(CH3CN)3} units, which coordinate the remaining
the adjacent P atoms of the phospholyl ligands (2.292(1)–2.313(1) P atoms (Fig. 7c). With eight Cu+, six I− and two [P3C2Mes2]−
Å). In contrast, the third P atom again shows a η1-coordi- ligands 8·0.5CH2Cl2·3CH3CN also displays a neutral molecular
nation to a single Cu ion revealing a shorter bond length of complex.
2.217(1) Å, as it was observed in the previous discussed com-
Formally removing one {Cu(CH3CN)3} unit and CH3CN
pounds (for a comparison of selected bond lengths see ligand from different sides of the molecular complex 8, the
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Fig. 8 Section of the anionic polymeric structure of 9. H atoms and
minor disordered components are omitted for clarity.
Spectroscopic characterization of the products
All obtained compounds 2–9 are completely insoluble in non-
polar solvents like n-hexane or toluene as well as in more
polar solvents like CH2Cl2 and Et2O. But they all show solubi-
lity in CH3CN up to a certain degree, which enables their
characterization in solution by NMR spectroscopy as well as
by mass spectrometry. The solubility decreases when using
heavier halides, but the lowest solubility is found for 5 due to
1
its 2D polymeric structure (characterization in solution by H
and mass spectrometry only). The 31P{1H} NMR spectrum of
3 in CD3CN shows four very broad signals at δ = 136, 160, 204
and 217 ppm with an integral ratio of 4 : 6 : 3 : 2. The broaden-
ing of the signals implies that the coordination of the
P atoms to Cu (nuclear spin I = 3/2) is still present in solution.
In contrast to 1 (δ = 73.2 ppm)8 the signals show a strong
downfield shift, though this comparison is hampered, since
no coordination to Fe is present in the product. In contrast,
the signals of 3 are shifted to higher field when compared to
the free phospholyl ligand in K[P3C2Mes2] (δ = 261.7 and
266.4 ppm).8 This is in agreement with the observations
made for other coordination compounds starting from
different Pn ligand complexes and Cu(I) halides.19 However,
to explain the integral ratios of the observed signals, one
must assume the presence of two different compounds in
solution. In addition, the 31P{1H} NMR spectrum of freshly
dissolved crystals of pure 4·CH3CN surprisingly looks exactly
the same with minimal differences in the integral ratio (δ =
135, 160, 205 and 217 ppm with an integral ratio of
4 : 5 : 2.5 : 2). Since elemental analyses confirm the purity of
the respective products, these results imply a dynamic dis-
and re-assembly behaviour in solution. In addition, for a
plausible assignment variable temperature 31P{1H} NMR
spectra of crystals of 3·6CH3CN were recorded (spectra see
ESI‡). Upon cooling to 273 K the outer signals at δ = 135 and
217 ppm become sharpened and increase in intensity. In con-
trast, the intensity of the inner signals at δ = 160 and
204 ppm decreases and the signals vanish completely upon
further cooling to 253 K. The opposite effect, namely the dis-
appearance of the outer signals, is monitored when the
sample is heated to 343 K. Therefore the inner signals can be
attributed to a rather small, monomeric unit, since its for-
mation requires bond dissociations, which are forced by
higher temperatures. In contrast, cooling favours the existence
of larger aggregates, hence the outer signals are assigned to
dimeric or oligomeric coordination compounds. The values
Fig. 7 (a) ‘Hexagram’ structural motif in 8·0.5CH2Cl2·3CH3CN and
9·0.6CH2Cl2,
(b)
section
of
the
molecular
structure
of
8·0.5CH2Cl2·3CH3CN (top view) illustrating the ‘hexagram’ motif, (c)
molecular structure of 8·0.5CH2Cl2·3CH3CN, (d) repeating unit of the
polymeric structure of 9·0.6CH2Cl2.
H atoms, mesityl ligands and
solvent molecules are omitted for clarity.
resulting complex forms the negatively charged repeating unit
of the 1D polymer 9·0.6CH2Cl2 without significant changes in
the ‘hexagram’ motif (Fig. 7d and 8). The Cu–Cu and Cu–I
bond lengths vary in ranges of 2.454(2)–2.535(2) and 2.537(1)–
2.617(1) Å, respectively (see also ESI‡). The Cu–P bonds are
slightly shorter within the range of 2.251(2)–2.778(2) Å (for a
comparison of selected bond lengths see Table 1). Surpris-
ingly, [Cp*Fe(CH3CN)3]+ turned out to act as a counter ion,
which is known as an extremely water-sensitive complex with
very labile acetonitrile ligands.18 Hence, another reaction
pathway of the [Cp*Fe]+ cation aforesaid becomes apparent in
this case, namely the accomplishment of the free coordination
sites by acetonitrile ligands.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Paper
for the integral ratios are in accordance with this, each group
being 2 : 1 for the adjacent and the unique P atoms, respect-
Conclusions
ively. These dynamic processes are further confirmed by mass In summary, we reported an unexpected fragmentation of
spectrometry. In the obtained positive ion ESI spectra many otherwise very stable phosphaferrocenes by reaction with Cu(I)
different fragments from [{P3C2Mes2}Cu5Br3{CH3CN}]+ up to halides. The observed splitting of [Cp*Fe(η5-P3C2Mes2)] (1) into
[{P3C2Mes2}4Cu17Br12]+ can be identified.
a [Cp*Fe]+ and a [P3C2Mes]− unit can most likely be attributed
The behaviour in solution does not depend on the used to the bulky mesityl ligands. While the [Cp*Fe]+ fragment
halide, since the same observations were made for compounds reacts independently, the self-assembly process of the remain-
2 and 6–9. Interestingly, the 31P{1H} NMR spectrum of dis- ing phospholyl ligands and CuX (X = Cl, Br, I) leads to the for-
solved crystals of 2 also looks similar to compound 3, but after mation of a large variety of coordination compounds, among
many attempts no second compound could be isolated, most them monomeric, dimeric as well as 1D and 2D polymeric
probably due to the preferred crystallization of 2.
aggregates. The obtained structural motifs represent novel
Furthermore, it should be noted that in the 31P{1H} NMR ‘cages’ composed of P, Cu and X atoms (X = Cl, Br), a very
spectra of the CuI-containing assemblies 6–9 the signals seldom appearing Cu4I4 ‘crown’ and a Cu6I6 ‘hexagram’ yet
exhibit the same chemical shift like 2–4 despite their unknown for CuI systems. All obtained products show
different structural motifs in the solid state: ‘crown’ (6, 7) and dynamic dissociation and association behaviour in solution,
‘hexagram’ (8, 9) motifs compared to the ‘cage’ motif (2–4). which turns the so far rather neglected 1,2,4-triphospholyl
However, the quality of the received spectra is initially quite ligand into an interesting and attractive ligand in coordination
poor caused by the low solubility so that the extremely broad chemistry. It remains to be investigated further whether the
signals start to disappear into the noise through coordination salt K(P3C2Mes2) can also be used as a building block in metal-
to Cu. Similar association processes in solution can again losupramolecular chemistry.
be proposed by means of the ESI mass spectra, which
show peaks at higher mass numbers than the molecular
weight in the solid state. For example, in the spectrum of the
Acknowledgements
monomeric compound 6 the fragment [{P3C2Mes2}4Cu11I6]+
is observed.
This work was comprehensively supported by the Deutsche
Due to its limited solubility, 31P and 65Cu MAS NMR Forschungsgemeinschaft. C.H. is grateful for a PhD fellowship
spectra were acquired for compound 6. Despite the rather of the Fonds der Chemischen Industrie.
high spinning frequency of 30 kHz, the 31P{1H} MAS NMR
spectrum of 6 exhibits two broad featureless signals at 31.8
and 82.6 ppm with an integrated area ratio of about 2 : 1 in
good agreement with the molecular structure of 6. In com-
Notes and references
parison with the reported data of copper-coordinated phos-
phorus cage compounds,20 all peaks are noticeably shifted
to lower ppm values, reminiscent of Cu3P3I2 in which the
polymeric phosphorus tube is coordinated by more flexible
copper iodide ligands.21 The presence of spinning sidebands
in the 31P MAS NMR spectrum indicates substantial 31P
chemical shift anisotropy typically observed in cases of
locally distorted symmetry, e.g. imposed by the various
ligands coordinated to phosphorus atoms. Any lineshape
splitting due to 31P–63,65Cu heteronuclear scalar interactions
(J-couplings) could not be resolved at the applied magnetic
fields (4.7 and 11.7 T), most likely owing to 31P–63,65Cu
residual dipolar couplings.22 The 65Cu MAS NMR spectrum
of 6 is dominated by a Gaussian-shaped peak at 5.5 ppm,
1 (a) G. Gritzner and J. Kuta, Pure Appl. Chem., 1984, 56, 461;
(b) N. G. Connelly and W. E. Geiger, Chem. Rev., 1996, 96,
877.
2 A. Togni, in Metallocenes: Synthesis - Reactivity - Appli-
cations, ed. A. Togni and R. Halterman, Wiley-VCH, 1998,
vol. 11, pp. 685–718.
3 (a) F. Mathey, J. Organomet. Chem., 2002, 646, 15;
(b) L. Weber, Angew. Chem., Int. Ed., 2002, 41, 563.
4 (a) C. Mueller, R. Bartsch, A. Fischer, P. G. Jones and
R. Schmutzler, J. Organomet. Chem., 1996, 512, 141;
(b) M. M. Al-Ktaifani, P. B. Hitchcock and J. F. Nixon,
J. Organomet. Chem., 2008, 693, 611; (c) C. Mueller,
R. Bartsch, A. Fischer and P. G. Jones, Polyhedron, 1993, 12,
1383; (d) M. H. A. Benvenutti, P. B. Hitchcock, J. F. Nixon
and M. D. Vargas, Chem. Commun., 1996, 441.
where the estimated quadrupolar coupling constant of CQ
≈
0.9 MHz is comparable to the coupling constant determined
for purified solid copper(I) iodide, reflecting reasonable
local symmetry such as a tetrahedral coordination. There-
fore, this peak is tentatively attributed to the copper atom
coordinated to the single phosphorus atom rather than
the P2-‘dumbbell’. In contrast, the copper atoms comprising
the highly distorted ‘crown’ motif are attributed to the
weak featureless shoulders recognized in the region around
−23 ppm.
5 (a) A. Schindler, M. Zabel, J. F. Nixon and M. Scheer,
Z. Naturforsch., B: J. Chem. Sci., 2009, 64, 1429;
(b) C. Schwarzmaier, A. Schindler, C. Heindl,
S. Scheuermayer, E. V. Peresypkina, A. V. Virovets,
M. Neumeier, R. Gschwind and M. Scheer, Angew. Chem.,
Int. Ed., 2013, 52, 10896; (c) A. Schindler, C. Heindl,
G. Balazs, C. Groeger, A. V. Virovets, E. V. Peresypkina and
M. Scheer, Chem. – Eur. J., 2012, 18, 829; (d) F. Dielmann,
A. Schindler, S. Scheuermayer, J. Bai, R. Merkle, M. Zabel,
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.

View Article Online
Paper
Dalton Transactions
A. V. Virovets, E. V. Peresypkina, G. Brunklaus, H. Eckert
and M. Scheer, Chem. Eur. J., 2012, 18, 1168;
J. R. Hanks, P. B. Hitchcock and J. F. Nixon, J. Organomet.
Chem., 2008, 693, 2287.
–
(e) M. Scheer, Dalton Trans., 2008, 4372.
12 (a) M. Hofmann, F. W. Heinemann and U. Zenneck, J. Orga-
nomet. Chem., 2002, 643–644, 357; (b) M. M. Al-Ktaifani,
P. B. Hitchcock and J. F. Nixon, J. Organomet. Chem., 2003,
665, 101.
13 F. W. Heinemann, M. Zeller and U. Zenneck, Organometal-
lics, 2004, 23, 1689.
6 (a) O. Daugulis, H.-Q. Do and D. Shabashov, Acc. Chem.
Res., 2009, 42, 1074; (b) C. Zhang, C. Tang and N. Jiao,
Chem. Soc. Rev., 2012, 41, 3464; (c) D. M. D’Souza and
T. J. J. Mueller, Chem. Soc. Rev., 2007, 36, 1095.
7 (a) S. Deng, C. Schwarzmaier, M. Zabel, J. F. Nixon,
M. Bodensteiner, E. V. Peresypkina, G. Balazs and 14 R. Peng, M. Li and D. Li, Coord. Chem. Rev., 2010, 254, 1.
M. Scheer, Eur. J. Inorg. Chem., 2011, 2991; (b) A. Schindler, 15 K. Sugimoto, T. Kuroda-Sowa, M. Munakata and
G. Balazs, M. Zabel, C. Groeger, R. Kalbitzer and M. Scheer,
M. Maekawa, Chem. Commun., 1999, 455.
C. R. Chim., 2010, 13, 1241; (c) S. Deng, C. Schwarzmaier, 16 F. Cecconi, C. A. Ghilardi, S. Midollini and A. Orlandini,
U. Vogel, M. Zabel, J. F. Nixon and M. Scheer, Eur. J. Inorg.
Chem., 2008, 4870.
8 C. Heindl, A. Schindler, M. Bodensteiner, E. V. Peresypkina,
A. V. Virovets and M. Scheer, Phosphorus, Sulfur Silicon Relat.
Elem., 2014, 189, 924–932.
J. Chem. Soc., Chem. Commun., 1982, 229.
17 (a) C.-M. Che and S.-W. Lai, Coord. Chem. Rev., 2005, 249,
1296; (b) M. Jansen, Angew. Chem., Int. Ed. Engl., 1987, 99,
1136.
18 (a) D. Catheline and D. Astruc, Organometallics, 1984, 3,
1094; (b) D. Catheline and D. Astruc, J. Organomet. Chem.,
1983, 248, C9.
9 (a) Y. Li, Y. Li, B. Wang, Y. Luo, D. Yang, P. Tong, J. Zhao,
L. Luo, Y. Zhou, S. Chen, F. Cheng and J. Qu, Nat. Chem.,
2013, 5, 320; (b) M. Wallasch, G. Wolmershäuser and 19 (a) F. Dielmann, A. Schindler, S. Scheuermayer, J. Bai,
H. Sitzmann, Angew. Chem., Int. Ed., 2005, 44, 2597;
(c) H. Sitzmann, T. Dezember, W. Kaim, F. Baumann,
D. Stalke, J. Kärcher, E. Dormann, H. Winter, C. Wachter
and M. Kelemen, Angew. Chem., Int. Ed. Engl., 1996, 35,
2872; (d) M. D. Walter, J. Grunenberg and P. S. White,
Chem. Sci., 2011, 2, 2120.
R. Merkle, M. Zabel, A. V. Virovets, E. V. Peresypkina,
G. Brunklaus, H. Eckert and M. Scheer, Chem. – Eur. J.,
2012, 18, 1168; (b) E.-M. Rummel, M. Eckhardt,
M. Bodensteiner, E. V. Peresypkina, W. Kremer, C. Groeger
and M. Scheer, Eur. J. Inorg. Chem., 2014, 2014, 1625.
20 (a) G. Brunklaus, J. C. C. Chan, H. Eckert, S. Reiser, T. Nilges
and A. Pfitzner, Phys. Chem. Chem. Phys., 2003, 17, 3768;
(b) S. Reiser, G. Brunklaus, J. H. Hong, J. C. C. Chan,
H. Eckert and A. Pfitzner, Chem. – Eur. J., 2002, 8, 4228.
10 A. S. Ionkin, W. J. Marshall, B. M. Fish, A. A. Marchione,
L. A. Howe, F. Davidson and C. N. McEwen, Eur. J. Inorg.
Chem., 2008, 2386.
11 (a) R. Bartsch, P. B. Hitchcock and J. F. Nixon, J. Organomet. 21 (a) A. Pfitzner, M. F. Braeu, J. Zweck, G. Brunklaus and
Chem., 1988, 340, C37; (b) R. Bartsch, P. B. Hitchcock and
J. F. Nixon, J. Organomet. Chem., 1988, 356, C1;
(c) P. B. Hitchcock, J. F. Nixon and R. M. Matos, J. Organo-
H. Eckert, Angew. Chem., Int. Ed., 2004, 43, 4228;
(b) G. Brunklaus, J. C. C. Chan and H. Eckert, Z. Phys.
Chem., 2003, 217, 1627.
met. Chem., 1995, 490, 155; (d) F. G. N. Cloke, J. R. Hanks, 22 (a) S. E. Ashbrook, J. McManus, M. J. Thrippelton and
P. B. Hitchcock and J. F. Nixon, Chem. Commun., 1999,
1731; (e) F. Mathey, J. Organomet. Chem., 2002, 646, 15;
(f) G. K. B. Clentsmith, F. G. N. Cloke, M. D. Francis,
S. Wimperis, Prog. Nucl. Magn. Reson. Spectrosc., 2009, 55,
160; (b) B. Thomas, S. Paasch, S. Steuernagel and
K. Eichele, Solid State Nucl. Magn. Reson., 2001, 20, 108.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2015