Influencing the stability of diaminomethane-containing azacrown ether ligands in the presence of transition-metal ions

Article abstract of DOI:10.1002/ejic.201201564

The evaluation of a new azacrown ether ligand bis{[3-(pyridin-4-yl)-1H- pyrazol-1-yl]methyl}diaza-18-crown-6 (b3pd), which contains pendant p-pyridylpyrazole arms connected by diaminomethane linkers, identified a tendency to undergo retro-Mannich fragmentation in the presence of transition-metal ions. However, treatment of b3pd with potassium perchlorate or potassium iodide prior to complexation with transition-metal ions imparted a resistance to fragmentation, such that crystallisation of coordination polymers from concentrated solutions after several weeks was possible. Four solid-state structures containing Kb3pd were isolated: a perchlorate salt (1), a divalent manganese 1D coordination polymer [Mn(Kb3pd)(DMF)₄]·2ClO ₄·I₃ (2, DMF = N,N-dimethylformamide), a cuprous 2D coordination network [Cu₂(Kb3pd)₂(I₃)(I) ₃]·3H₂O (3) and a cuprous 1D coordination polymer [Cu(Kb3pd)(I)₂] (4). Additionally, the retro-Mannich process was investigated by in situ FTIR spectroscopy, mass spectrometry, crystallography and by the isolation of a cobalt complex ligated by a partially fragmented ligand, mono{[3-(pyridin-4-yl)-1H-pyrazol-1-yl]methyl}diaza-18-crown-6 (m3pd). The composition of the cobalt complex was found to be [CoCl₃(Hm3pd)] ·H₂O (5). Diaza-18-crown-6 ligands functionalised with diaminomethane linkers are unstable in the presence of selected divalent transition-metal ions. Pretreatment of the ligands with potassium salts prevents degradation by blocking the central coordination region. Iodide redox chemistry also affects the coordination-polymer motif. Copyright

Full text of DOI:10.1002/ejic.201201564

FULL PAPER
DOI:10.1002/ejic.201201564
Influencing the Stability of Diaminomethane-Containing
Azacrown Ether Ligands in the Presence of Transition-
Metal Ions
William J. Gee[a] and Stuart R. Batten*[a]
Keywords: Crown compounds / Polyanions / Transition metals / Structure elucidation / Coordination polymers
The evaluation of a new azacrown ether ligand bis{[3-(pyr-
idin-4-yl)-1H-pyrazol-1-yl]methyl}diaza-18-crown-6 (b3pd),
which contains pendant p-pyridylpyrazole arms connected
by diaminomethane linkers, identified a tendency to undergo
retro-Mannich fragmentation in the presence of transition-
metal ions. However, treatment of b3pd with potassium per-
chlorate or potassium iodide prior to complexation with tran-
sition-metal ions imparted a resistance to fragmentation,
such that crystallisation of coordination polymers from con-
centrated solutions after several weeks was possible. Four
chlorate salt (1), a divalent manganese 1D coordination poly-
mer [Mn(Kb3pd)(DMF)4]·2ClO4·I3 (2, DMF = N,N-dimethyl-
formamide), a cuprous 2D coordination network [Cu2(Kb3pd)
2(I3)(I)3]·3H2O (3) and a cuprous 1D coordination polymer
[Cu(Kb3pd)(I)2] (4). Additionally, the retro-Mannich process
was investigated by in situ FTIR spectroscopy, mass spec-
trometry, crystallography and by the isolation of a cobalt
complex ligated by a partially fragmented ligand, mono{[3-
(pyridin-4-yl)-1H-pyrazol-1-yl]methyl}diaza-18-crown-6
(m3pd). The composition of the cobalt complex was found to
solid-state structures containing Kb3pd were isolated: a per- be [CoCl3(Hm3pd)]·H2O (5).
geometry (e.g., 4,4Ј-bipyridine), which promotes their appli-
Introduction
cation to crystal engineering. Those classed as flexible tend
to exhibit unpredictable ligand geometries and intermetal
bridging lengths [e.g., 1,3-di(pyridyl)propane], which limits
their usefulness in crystal engineering. Functionalised aza-
crown ether ligands represent an intermediate between the
rigid and flexible classes, as they exhibit a range of coordi-
native geometries and lengths; however, these can be con-
trolled by the addition of guest molecules[8] or by varying
the pH.[9] Indeed, post-synthetic changes to the identity of
a guest species or pH could induce structural changes in a
preformed coordination polymer and could potentially lead
to variation in porosity, control over adsorption/desorption
behaviour and the ability to manipulate metal-based elec-
tronic properties such as luminescence and magnetism.
Polyiodide networks exhibit templated growth in the
presence of linear organic molecules,[10] but the plethora of
different metal iodide coordination species known, in par-
ticular for copper,[11] is problematic for crystal engineering
owing to the diverse coordination environments and vari-
able degree of polyiodide formation. However, judicious se-
lection of the oxidation state of the metal centre can be
used to promote or inhibit polyiodide formation by altering
the availability of iodine present in the mother liquor prior
to crystallisation, which imparts some measure of control
over the resulting crystalline product.
Functionalised diaza-18-crown-6 ether moieties have
been applied to a variety of metal-interaction-based appli-
cations and research targets, including colorimetric sensors
for calcium in blood serum,[1] synthetic ion channels,[2]
phase-transfer catalysts[3] and complexing agents for radio-
isotopes in the treatment of cancer,[4] as well as other biolo-
gically relevant interactions.[5] A convenient means of gen-
erating azacrown ether ligands with pendant functionalities
was discovered in the Mannich reaction, which allowed
modification of metal binding properties and variation of
ligand geometries.[6] Variation of the nucleophile to a hy-
droxy, amine, thiol or phosphite (Kabachnik–Fields reac-
tion) species gives related products, each with varied reac-
tivities and stabilities.[7]
A new application of functionalised azacrown ether li-
gands concerns their potential ability to alter their bridging
length when incorporated into coordination polymers. Typi-
cally, the organic linkers that comprise coordination poly-
mers can be classed as either rigid or flexible depending on
their level of conformational freedom. Ligands classed as
rigid have an invariable predetermined bridging length and
[a] School of Chemistry, Monash University, Clayton,
Melbourne, Vic 3800, Australia
Fax: +61-3-9905-4597
E-mail: stuart.batten@monash.edu
Herein, we evaluate a new functionalised diaza-18-
crown-6 ether ligand, bis{[3-(pyridin-4-yl)-1H-pyrazol-1-
yl]methyl}diaza-18-crown-6 (b3pd), for the synthesis of co-
ordination polymers with selected transition-metal ions
Homepage: www.monash.edu/science/about/schools/chemistry/
staff/sbatten/
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201201564.
Eur. J. Inorg. Chem. 2013, 3240–3248
3240
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
(Mn, Cu). Similar ligands have been demonstrated to have the crown and coordinated by two additional oxygen atoms
important implications in the field of crystal engineering, derived from the perchlorate ion. The bond lengths between
both by us[8,9] and by others.[12] Structurally, the ligand con- the azacrown ether and potassium atom approach uniform-
tains two aminal bridges, a functionality rarely observed in ity for the oxygen donors [K–O(1): 2.7075(15) Å, K–O(2):
solid-state coordination species with the exception of 2.8936(15) Å,
K–O(3):
2.7503(15) Å,
K–O(4):
cyclised linkers (e.g., triazacyclohexane or hexamethylene- 2.7615(15) Å], and slightly longer distances are observed for
tetramine)[13] and imidazolidine rings, which are often con- the nitrogen donors [K–N(4): 3.1294(17) Å, K–N(5):
tained within Schiff bases.[14] Two previously reported, crys- 3.0586(17) Å]. The perchlorate anion coordinates through
tallographically characterised examples containing aminal O(5) and O(6) with bond lengths of 2.833(2) and
bridges analogous to b3pd also demonstrated solution in- 3.233(3) Å, respectively. No additional noncovalent interac-
stability and required an excess of formaldehyde to prevent tions were located within the structure.
ligand degradation.[15]
Four solid-state structures containing potassium-coordi-
nated b3pd are described here: the potassium perchlorate
complex of the ligand (1), a divalent manganese 1D coordi-
nation polymer [Mn(Kb3pd)(DMF)4]·2ClO4·I3 (2, DMF =
N,N-dimethylformamide), a cuprous 2D coordination net-
work [Cu2(Kb3pd)2(I3)(I)3]·3H2O (3) and a cuprous 1D co-
ordination polymer [Cu(Kb3pd)(I)2] (4). Unlike previous di-
aza-18-crown-6 ether moieties[8,9] and aminal-bridge-con-
taining ligands,[15] b3pd could be stabilised with guest po-
Figure 2. Solid-state structure of 1. Thermal ellipsoids are shown
at 50% probability.
tassium ions, which inhibited ligand hydrolysis and thus ne-
gated the requirement for excess formaldehyde.[15] The frag-
Mixtures of ligand b3pd with either KClO4 or KI were
next treated with hydrated transition-metal perchlorates to
facilitate the formation of coordination polymers. These ef-
forts were successful in three instances, with one structure
obtained from Mn(ClO4)2·6H2O and one each from two
separate copper perchlorate sources Cu(ClO4)2·6H2O and
[Cu(MeCN)4](ClO4). A summary of the synthesis is shown
in Scheme 1.
mentation of b3pd by zinc chloride was then monitored by
in situ FTIR spectroscopy and mass spectrometry (MS) to
give general insights into the mechanism of this process.
Furthermore, a solid-state structure of a partially frag-
mented ligand, mono{[3-(pyridin-4-yl)-1H-pyrazol-1-yl]-
methyl}diaza-18-crown-6 (m3pd), ligated to cobalt chloride,
[CoCl3(Hm3pd)]·H2O (5), was isolated and provided crys-
tallographic evidence for the proposed fragmentation.
Results and Discussion
The functionalised diaza-18-crown-6 ether ligand bis{[3-
(pyridin-4-yl)-1H-pyrazol-1-yl]methyl}diaza-18-crown-6
(b3pd) was synthesised by aminomethylation of diaza-18-
crown-6 with two equivalents of both p-formaldehyde and
4-(1H-pyrazol-3-yl)pyridine in a manner analogous to the
Mannich reaction (Figure 1).
Scheme 1. Synthetic overview of the synthesis of coordination poly-
mers 2–4. Conditions: (a) reagents dissolved in DMF layered with
Et2O; (b) b3pd and KI dissolved in CHCl3 and MeOH layered with
a buffer layer (1:1 CHCl3/MeOH) and a solution of metal perchlo-
rate in MeCN or MeOH.
Single crystals of coordination polymers 2–4, the molecu-
lar formulae of which are shown in Scheme 1, were ob-
tained by one of two methods. The first required dissolution
of b3pd, potassium iodide and a metal perchlorate salt in
DMF, into which was slowly diffused diethyl ether. The sec-
ond method required the dissolution of b3pd and KI in a
mixture of chloroform and the minimum volume of meth-
anol required to solubilise the ionic species. This solution
was then layered with a buffer layer of chloroform and
methanol (1:1) and a layer of neat methanol or acetonitrile
containing the metal perchlorate. Crystals were found to
deposit in the buffer layer, in which the two solutions met
over several days.
Figure 1. Reagents and conditions: (a) 2 equiv. (CH2O)n, MeOH,
room temp., 12 h; (b) 2 equiv. 4-(1H-pyrazol-3-yl)pyridine, toluene,
ΔT, 4 h.
1
A H NMR spectrum of a sample in deuterated chloro-
form and the elemental analysis confirmed the formation
of b3pd (see Supporting Information). Solid-state structural
information was obtained by the reaction of a chloroform
solution of b3pd with a saturated methanol solution of
KClO4; the product crystallised after slow evaporation of
the solvent. The resultant crystal structure is shown in Fig-
Structures of the Coordination Polymers
Treatment of (Kb3pd)I with Mn(ClO4)2·6H2O resulted in
ure 2. The potassium ion is eight coordinate, situated within the isolation of a 1D coordination polymer with composi-
Eur. J. Inorg. Chem. 2013, 3240–3248
3241
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Figure 3. Solid-state structure of 2. Hydrogen atoms have been omitted for clarity.
tion [Mn(Kb3pd)(DMF)4]·2ClO4·I3 (2, Figure 3). The di-
valent Mn atom is located on an inversion centre, is six
coordinate, and is ligated equatorially by four molecules of
DMF and apically by the pyridyl groups of two Kb3pd li-
gands. Analysis of the bond lengths identified a greater af-
finity for the oxygen atoms of the DMF molecules [Mn(1)–
O(3): 2.188(5) Å, Mn(1)–O(4): 2.151(5) Å] relative to the
pyridyl donors [Mn(1)–N(1): 2.293(6) Å]. The cationic
charge was offset by perchlorate and triiodide anions, the
Scheme 2. Means of triiodide formation within solutions of redox-
active transition-metal ions[17] prior to coordination-polymer for-
mation.
latter situated across an inversion centre with an I(1)–I(2) potassium ion range from 2.756(6) Å for K(1)–O(2) to
bond length of 2.9081(10) Å. Two features merit further dis- 2.997(7) Å for K(1)–N(2), and intermediate distances of
cussion: the redox formation of triiodide and the unique 2.834(6) and 2.902(6) Å are observed for K(1)–O(1) and
ligand geometry coupled with pyrazole coordination to the K(1)–N(2), respectively. The coordination of the pyrazole
potassium ion.
group to the potassium ion is highly atypical regarding the
The formation of triiodide occurs in solutions of iodine obtuse angle of coordination (110.3°) between the plane of
and iodide.[16] The formation of iodine occurs by oxidation the pyrazole and the coordination bond between K(1) and
of two iodide ions with the concomitant reduction of two N(2). The aromatic nature of the pyrazole ring would yield
Mn ions.[17] The iodine then undergoes addition to an ad- an orbital containing a lone pair of electrons in the plane
ditional iodide ion to yield the triiodide anion.[18] The over- of the ring and an orthogonal delocalised electron-rich p
all equilibrium is shown in Scheme 2. Clearly, an oxidant is orbital, neither of which are well oriented to interact and
required to drive the iodide/triiodide redox couple, which stabilise the electrophilic potassium ion. Consequently, al-
may form by air oxidation of Mn2+ to Mn3+ or of iodide though the distance between the two groups is indicative of
directly;[19] however, neither the identity nor means of gen- an interaction, such an interaction would likely be weak-
eration of the oxidant could be unambiguously determined. ened by an unfavourable coordination geometry; hence, it
The ligand geometry observed for Kb3pd is interesting is possible that the ligand S conformation arises because of
for several reasons. The eight coordinate binding of the po- Mn coordination, and packing effects cannot be dis-
tassium ion instils an S conformation in the ligand (Fig- counted. Interestingly, all prior studies of potassium coordi-
ure 4, left). Related ligands are known to adopt this confor- nation in diazacrown ether ligands with pyridylmethyl arms
mation,[8,9] however this is the first example of alkali-metal resulted in inhibition of the S configuration.[8–9] In these
stabilisation of this ligand geometry. Bond lengths to the instances, the K ions solely templated a linear conforma-
Figure 4. Potassium coordination within the b3pd host in 2. Hydrogen atoms omitted for clarity. Thermal ellipsoids are shown at 50%
probability.
Eur. J. Inorg. Chem. 2013, 3240–3248
3242
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
tion, which maximises the bridging ligand length and is the ide coordination are filled by lattice water molecules. The
opposite of the behaviour exhibited by Kb3pd here. azacrown functionality exhibits a boatlike conformation
The use of Cu(ClO4)2·6H2O resulted in the iodide-medi- with close contacts to oxygen donors ranging from 2.727(9)
ated reduction of cupric perchlorate to cuprous perchlorate to 2.800(6) Å. The nitrogen donor distances are slightly
shown in Scheme 1. Such redox chemistry between iodine longer at 2.984(8) and 2.996(8) Å. The K–I bond lengths
and copper has a rich history and is covered by numerous are 3.574(2) and 3.713(2) Å for I(1) and I(2), respectively.
review articles.[20] The resultant [Cu2(Kb3pd)2(I3)(I)3]·3H2O The extended 2D network propagates through the afore-
(3) network contains both Cu+ and triiodide ions, as well as mentioned K–I(1) interactions between the Cu2I2 nodes
K-coordinated b3pd (Figure 5). The two Cu+ atoms exhibit and the K ions of adjacent chains (Figure 6).
tetrahedral coordination geometries, each ligated by two
iodide anions and the pyridyl groups of two Kb3pd ligands.
Near-equivalent Cu–I and Cu–Py bond lengths were ob-
served between Cu(1) and Cu(2), measured at 2.6330(11)
[Cu(1)–I(1)], 2.6244(11) [Cu(2)–I(1)], 2.057(7) [Cu(1)–N(5)]
and 2.059(7) Å [Cu(2)–N(1)].
A Cu–Cu distance of
2.588(2) Å was observed, which is comparable to the dis-
tance inferred from that of metallic copper (2.56 Å).[21] The
potassium ion is coordinated to eight donor atoms, six from
b3pd and one [I(1)] from the iodide anion of the Cu2I2 di-
mer. The remaining coordination site is occupied by I(2),
which may exist as either an iodide ion or as part of a tri-
iodide ion (1:1 ratio). When a triiodide ion is present, it is
positionally distorted over two sites to give I(3)–I(6) occu-
pancy factors of one quarter. The voids resulting from iod-
Figure 6. 2D network structure of 3. A single chain motif is high-
lighted in bold to highlight the linkage between Cu2I2 nodes and
K+ ions. Hydrogen atoms and lattice solvent molecules omitted for
clarity.
When Cu(ClO4)2·6H2O was substituted with [Cu-
(MeCN)4](ClO4), no triiodide ions were found in the struc-
ture, owing to inhibition of the initial iodine-forming step
of Scheme 2. This resulted in a 1D coordination polymer
with a composition of [Cu(Kb3pd)(I)2] (4). The solid-state
structure of 4 is shown in Figure 7.
The asymmetric unit of 4 was found to contain two dis-
crete chain units, each virtually indistinguishable in terms
of bond lengths and angles. Despite the similarity between
the two units, no evidence for missed symmetry was re-
ported by PLATON.[22] The structure contains 1D chains
of CuI2 moieties bridged by Kb3pd ligands. The Cu atoms
are tetrahedral and coordinated to two iodide ions and two
Kb3pd pyridyl groups. The chains are crosslinked into pairs
by interactions between the iodide ions on one chain and
potassium ions on the adjacent, symmetry related chain and
vice versa to give a zigzag ladder motif with CuI2K “rungs”
(Figure 7). There are two crystallographically unique (but
chemically similar) zigzag ladders, which are parallel to
Figure 5. Solid-state structure of 3. Lattice water molecules omitted
for clarity. The triiodide ion displays 1:1 substitutional disorder
with the iodide ion [I(2)] as well as site disorder over two positions,
which gives I(3)–I(6) each an occupancy factor of one quarter.
Figure 7. Solid-state structure of 4. Hydrogen atoms omitted for clarity.
Eur. J. Inorg. Chem. 2013, 3240–3248
3243
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
each other and alternate to yield layers (Figure 8). These contains a guest water molecule, which gives the structure
layers are separated by large voids that are filled with sol- an overall composition of [CoCl3(Hm3pd)]·H2O (5). The
vent molecules. The void volume was determined to be metal centre exhibits tetrahedral coordination, in which
1976 Å3, some 19% of the cell volume. The residual elec- three sites are occupied by chloride anions and the fourth
tron density located in these voids was unable to be as- by N(1) of Hm3pd. Protonation occurs at the secondary
signed unambiguously and so was removed by using the amine formed by the fragmentation, and there are hydrogen
SQUEEZE command.[23] Microanalyses of air-dried crys- bonds to both the guest water molecule within the crown
talline samples suggest the presence of one molecule of [N(5)–O(5) 2.699(5) Å, O(5)–O(1) 2.751(4) Å, O(5)–O(4)
methanol and 17 water molecules per unit of 4. Owing to 2.772(5) Å] and to a chloride ion of a second complex
the similarity between the two chain units in the unit cell, [N(5)–Cl(1) 3.277(4) Å]. Reciprocal hydrogen bonding be-
only one will be discussed in detail. Longer Cu–I bond tween complexes gives an extended pseudodimeric motif
lengths of 2.6919(17) [Cu(1)–I(1)] and 2.774(2) Å [Cu(1)– (Figure 10).
I(2)] were observed relative to those of 3 and are offset by
shorter Cu–N bond lengths of 2.004(9) [Cu(1)–N(8)] and
2.040(9) Å [Cu(1)–N(1)]. The K ion remains coordinated to
the crown and to iodide ions and exhibits similar bond
lengths to those in 3. The oxygen donor distances range
from 2.693(10) to 2.826(8) Å, and the nitrogen donor dis-
tances are 2.926(8) and 3.025(9) Å. The K–I bond lengths
of 4 [3.474(3) and 3.508(3) Å for I(1) and I(2), respectively]
are much shorter than those of 3. The I(1)–K(1)–I(2) angle
changed substantially, shifting from 83.36(5)° for 3 to
115.11(6)° for 4.
Figure 9. Solid-state structure of 5. Thermal ellipsoids are shown
at 50% probability.
Figure 10. Pseudodimeric motif exhibited by complexes of 5. Hy-
drogen bonds are shown as red dashed lines. Thermal ellipsoids are
shown at 50% probability.
Figure 8. The layers of zigzag ladder motifs of 4 viewed down the
a axis and displaying the large interlayer voids. Hydrogen atoms
omitted for clarity.
Many attempts to synthesise coordination polymers from
Fragmentation Study
b3pd in the absence of a potassium source were undertaken,
however all failed to yield crystalline products in which the
A study of the fragmentation of b3pd in the absence of
ligand remained intact. Indeed, unexpected gellation behav- guest potassium ions was undertaken by using in situ FTIR
iour was recently observed with CuCl2; a fragmentation spectroscopy to visualise the region from 4000 to 650 cm–1.
product of b3pd was crystallographically identified and evi- Changes in band frequency and intensity following the ad-
denced the instability of the ligand in the presence of transi- dition of zinc chloride to a 3:1 solution of DMF/H2O were
tion-metal salts.[24] A related experiment in which b3pd was monitored in real time by using the ConcIRT software.[25]
treated with cobalt chloride yielded further crystallographic The software interprets the net change in all peak intensities
evidence of the fragmentation process. The obtained cobalt over a region of interest, in this instance 1800 to 700 cm–1,
crystals were extremely small and necessitated the use of and deconstructs that data into individual component spec-
a synchrotron radiation source. The structure contains the tra, which allow changes in the reaction profile to be visual-
partially fragmented ligand mono{[3-(pyridin-4-yl)-1H-pyr- ised.[26] This provides a rapid qualitative “fingerprint” of
azol-1-yl]methyl}diaza-18-crown-6 (m3pd) coordinated to reaction behaviour intended for general insights into the
cobalt chloride (Figure 9). The ligand is protonated and mechanism.[27] Zinc chloride was chosen for this study ow-
Eur. J. Inorg. Chem. 2013, 3240–3248
3244
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
ing to its high Lewis acidity and lack of redox behaviour. Table 1. Observed cationic fragments after treatment of b3pd with
ZnCl2.
The percentage fragmentation of b3pd mediated by ZnCl2
versus time is plotted in Figure 11. Owing to the simulta-
neous co-evolution of diaza-18-crown-6 and 4-(1H-pyrazol-
3-yl)pyridine during fragmentation, the software was unable
to distinguish between these two species and instead yields
a retro-Mannich (RM) product spectrum. The fragmenta-
tion of b3pd in the presence of ZnCl2 occurs rapidly in the
absence of guest potassium ions and nears completion after
ca. 40 min. To confirm the identity of the fragmentation
products in the composite spectrum of the RM products, a
comparison was made to the solid-state IR spectra of diaza-
18-crown-6 and 4-(1H-pyrazol-3-yl)pyridine. Although sub-
tle variations between the composite solution-state spec-
trum and the two component solid-state spectra are to be
expected owing to variations in solvation and crystalline lat-
tice interactions, peaks present in the fragmented finger-
print region were observed at wavelengths consistent with
the reference spectra of diaza-18-crown-6 and 4-(1H-pyr-
azol-3-yl)pyridine (Figure 12).
Based on both the above crystallographic evidence and
the mass spectrum, the variable behaviour of b3pd in the
presence of transition-metal and potassium ions can be ex-
plained. Coordination of a transition-metal ion to either the
azacrown or pyrazole group activates the diaminomethane
group to cleavage (Figure 13) to yield a terminal iminium
group, as evidenced by the mass spectrum. This species
quickly hydrolyses.[15] This can also be implied from the IR
experiment, in which the bands expected for a secondary
amine were not observed, yet the prominent signal expected
for a primary iminium group was also absent. These species
Figure 11. The fragmentation of b3pd by ZnCl2 in a solution of 3:1
are prone to hydrolysis to give the secondary amine starting
materials.
DMF/H2O. Changes in the reaction profile were visualised by using
the ConcIRT software. Green trace: b3pd; blue trace: fragmentation
products.
Figure 13. Two postulated fragmentation pathways of b3pd in the
presence of transition-metal ions.
The fragmentation process requires the coordination of
a highly Lewis acidic, divalent metal ion to a non-pyridyl
nitrogen atom. Hence, the coordination of a weaker Lewis
Figure 12. Comparison between the composite spectrum of the
retro-Mannich products with reference crystalline samples of diaza-
acid (K+) protects b3pd from fragmentation by blocking co-
ordination at these sites. Ionic repulsion is the likely cause
of this effect coupled with selective guest preference for po-
tassium ions within the crown.
18-crown-6 and 4-(1H-pyrazol-3-yl)pyridine. Blue trace: fragmenta-
tion products; red trace: diaza-18-crown-6; black trace: 4-(1H-pyr-
azol-3-yl)pyridine.
A mass spectrum obtained from the fragmentation study
after 120 min determined a range of cationic fragmentation
products. Five prominent species were identified, each of
which is shown in Table 1.
Conclusions
As a result of this study with b3pd, we have isolated four
structures, namely, the potassium perchlorate salt of the li-
Eur. J. Inorg. Chem. 2013, 3240–3248
3245
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Bis{[3-(pyridin-4-yl)-1H-pyrazol-1-yl]methyl}diaza-18-crown-6
(b3pd): To a nitrogen-flushed round-bottomed flask were added di-
aza-18-crown-6 (500 mg, 1.9 mmol) and paraformaldehyde
(142 mg, 3.8 mmol). Dry methanol (20 mL) was added, and the
mixture was stirred for 12 h, wherein the paraformaldehyde slowly
dissolved. The solvent was removed under vacuum, and 4-(1H-pyr-
azol-3-yl)pyridine (553 mg, 3.8 mmol) was added. Toluene (30 mL)
was added, and the solution was heated to reflux for 5 h. The sol-
vent was removed, and the remaining solid was dried for several
hours under high vacuum to yield the title compound (1.090 g,
gand, two divalent 1D coordination polymers,
[Mn(Kb3pd)(DMF)4]·2ClO4·I3 and [Cu(Kb3pd)(I)2], and a
cuprous 2D coordination polymer [Cu2(Kb3pd)2(I3)(I)3]·
3H2O. The choice of redox-active metal ion, coupled with
a redox-active iodide counteranion, yielded an interesting
means to vary the product. The formation of the triiodide
ion was found to predominantly occur in the presence of a
reducible metal ion, and the unique geometry of the triiod-
ide ion relative to the iodide ion influenced the product.
Copper species 3 and 4 demonstrated this effect; the reduc- 99%). C30H40N8O4 (576.70): calcd. C 62.48, H 6.99, N 19.43; found
1
ible Cu2+ ion in 3 promotes the formation of the triiodide
ion, which was incorporated into a 2D network structure,
C 62.45, H 6.95, N 19.44. H NMR (500 MHz, CDCl3): δ = 8.61
(d, J = 2.5 Hz, 4 H, 1-H), 7.68 (d, J = 2.5 Hz, 4 H, 2-H), 7.29 (d,
whereas the already reduced Cu+ ion in 4 results solely in
J = 1.5 Hz, 2 H, 4-H), 6.65 (d, J = 1.5 Hz, 2 H, 3-H), 5.18 (s, 4 H,
5-H), 3.70 (t, J = 5.0 Hz, 8 H, 7-H), 3.67 (s, 8 H, 8-H), 2.95 (t, J
iodide incorporation to yield a 1D coordination polymer.
Of further interest in this study was the fragmentable na-
ture of b3pd. This was demonstrated crystallographically
with the isolation of a single-armed derivative of b3pd, ab-
breviated to m3pd, which was incorporated into a cobalt
structure with the composition [CoCl3(Hm3pd)]·H2O. A
= 5.0 Hz, 8 H, 6-H) ppm. 13C NMR (125 MHz, CDCl3): δ =
150.19, 148.61, 141.07, 131.97, 131.67, 120.01, 103.37, 70.65, 69.99,
52.18 ppm. IR (ATR): ν
= 2875 (m), 2833 (m), 1604 (s), 1558
˜
max
(w), 1487 (m), 1458 (m), 1421 (m), 1355 (s), 1336 (m), 1305 (s),
1247 (m), 1206 (s), 1167 (m), 1135 (m), 1113 (s), 1061 (s), 996 (s),
961 (s), 879 (w), 830 (s), 803 (w), 778 (s), 744 (s), 699 (m), 685 (s),
fragmentation study was undertaken by using mass spec- 629 (m) cm–1. Single crystals of [Kb3pd](ClO4) (1) were obtained
by dissolving b3pd and K(ClO4) in chloroform and a small quantity
of methanol and allowing the solution to slowly evaporate.
trometry and in situ FTIR spectroscopy, which determined
that coordination to transition-metal ions effected fragmen-
tation analogous to the retro-Mannich reaction. Interest-
ingly, the inclusion of a potassium guest ion into b3pd pre-
vented this fragmentation, likely because the preferential af-
finity of the crown moiety for K relative to transition-metal
[Mn(Kb3pd)(DMF)4]·2ClO4·I3 (2): To a solution of b3pd (45 mg,
0.078 mmol) in chloroform (ca. 1 mL) was added KI (25 mg,
0.151 mmol) dissolved in the minimum quantity of methanol. A
buffer layer of chloroform and methanol (1:1, 2 mL) was placed
ions shields the inner coordination sites by ionic repulsion. above the b3pd solution. A final layer of methanol (1 mL) contain-
ing Mn(ClO4)2·6H2O (25 mg, 0.069 mmol) was placed over the
buffer layer. Brown crystals were found to deposit in both
This strategy allows the use of diaminomethane-fused aza-
crown ether ligands in the synthesis of coordination poly-
mers in the presence of transition-metal ions.
the
buffer
and
chloroform
layers
(37 mg,
35%).
C42H68Cl2I3KMnN12O16 (1542.14): calcd. C 32.70, H 4.44, Cl + I
29.28, N 10.9; found C 32.77, H 4.48, Cl + I 30.03, N 11.02. IR
(ATR): ν
= 3399 (m), 2877 (m), 1610 (s), 1454 (m), 1422 (m),
˜
max
Experimental Section
1350 (m), 1303 (m), 1277 (w), 1206 (m), 1081 (s), 1050 (s), 1006
(m), 937 (m), 833 (m), 762 (m), 742 (m), 694 (m), 622 (m) cm–1.
General Considerations: Diaza-18-crown-6[28] and 4-(1H-pyrazol-3-
yl)pyridine[29] were both synthesised according to literature pro-
cedures. A stock solution of cuprous perchlorate was created by
heating cupric perchlorate hydrate to reflux with an excess of cop-
per powder in acetonitrile for two hours. The resultant clear solu-
tion was filtered and stored under nitrogen. Metal salts were pur-
chased from Alfa Aesar. All other chemicals were purchased from
Sigma–Aldrich.
[Cu2(Kb3pd)2(I3)(I)3]·3H2O (3): Synthesised analogously to 2. Red
crystals were isolated from the buffer and chloroform layers (80 mg,
55%). C60H80Cu2I6K2N16O11 (2168.11): calcd. C 33.24, H 3.61, I
35.12, N 10.34; found C 33.18, H 3.90, I 34.84, N 10.44. IR (ATR):
ν
˜
= 2876 (m), 1610 (s), 1554 (w), 1493 (w), 1454 (m), 1425 (m),
max
1348 (m), 1302 (m), 1267 (m), 1205 (s), 1177 (m), 1100 (s), 1067
(s), 1010 (m), 955 (w), 936 (s), 836 (m), 748 (s), 699 (m), 691 (m),
665 (w), 622 (m) cm–1.
Crystals used for elemental analysis and infrared spectroscopy were
removed from the buffer layer and manually separated from colour-
less crystals of KI. The crystals were washed with methanol prior
to analyses. MS measurements were obtained by using a Micro-
mass Platform II QMS spectrometer. Solid-state IR spectra were
recorded with a Bruker Equinox 55 Infrared Spectrometer fitted
with a Specac Diamond attenuated total reflectance (ATR) source.
Solution real-time infrared (RTIR) scanning measurements were
recorded by using a Mettler Toledo ReactIR 10 spectrometer fitted
with a DiComp probe and connected to a mercury cadmium tellu-
ride (MCT) detector by a K6 Conduit. Scanning was performed in
the region 4000–650 cm–1 at 8 cm–1 resolution. Elucidation of the
reaction components in the region 1800–700 cm–1 was performed
by using the ConcIRT software.[25] Elemental analyses were per-
formed by Campbell Microanalytical Laboratory, Department of
Chemistry, University of Otago, Dunedin, New Zealand.
[Cu(Kb3pd)(I)2]·MeOH·17H2O (4): Synthesised analogously to 2
by exchanging the metal-containing layer with [Cu(MeCN)4](ClO4)
(57 mg, 0.174 mmol) in acetonitrile (1 mL). Yellow crystals were
isolated from the buffer and chloroform layers (56 mg, 33%).
C61H118Cu2I6K2N16O29 (2505.54): calcd. C 29.23, H 4.75, I 30.38,
N 8.94; found C 29.01, H 4.88, I 29.84, N 8.72. IR (ATR): ν
=
˜
max
2876 (m), 1610 (s), 1554 (w), 1493 (w), 1454 (m), 1425 (m), 1348
(m), 1302 (m), 1267 (m), 1205 (s), 1177 (m), 1100 (s), 1067 (s), 1010
(m), 955 (w), 936 (s), 836 (m), 748 (s), 699 (m), 691 (m), 665 (w),
622 (m) cm–1.
[CoCl3(Hm3pd)]·H2O (5): A solution of b3pd (45 mg, 0.078 mmol)
in chloroform (ca. 1 mL) was placed beneath a buffer layer of chlo-
roform and methanol (1:1, 2 mL). A final layer of methanol (1 mL)
containing CoCl2·6H2O (25 mg, 0.105 mmol) was placed over the
buffer layer. Blue crystals were found to deposit in the buffer layer
after several days (Ͻ1%). The limited quantity of pure, crystalline
X-Seed[30] and POV-Ray[31] were used for the preparation of the
crystallographic figures.
Eur. J. Inorg. Chem. 2013, 3240–3248
3246
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Table 2. Crystal data and structural refinements for 1–5.
1
2
3
4
5
Chemical formula
Formula mass
Crystal system
Space group
a [Å]
b [Å]
c [Å]
α [°]
β [°]
C30H40ClKN8O8
715.25
monoclinic
P21/n
10.5928(7)
18.5681(15)
17.1466(15)
90
93.572(2)
90
C42H68Cl2I3KMnN12O16 C60H80Cu2I6K2N16O11 C60H80Cu2I4K2N16O8
C21H36Cl3CoN5O5
603.83
1542.72
triclinic
P1
2168.08
monoclinic
C2/c
1866.28
monoclinic
P21/c
triclinic
¯
¯
P1
9.0310(18)
11.448(2)
15.437(3)
69.84(3)
86.54(3)
80.49(3)
1477.6(5)
1
17.898(4)
23.075(5)
20.652(4)
90
103.00(3)
90
21.022(4)
23.280(5)
21.566(4)
90
102.10(3)
90
9.4360(19)
12.301(3)
12.928(3)
87.30(3)
73.84(3)
73.06(3)
1377.8(5)
2
γ [°]
Unit cell volume [Å3]
Z
3366.0(5)
4
8311(3)
4
10320(4)
4
Temperature [K]
Reflections measured
123(3)
30391
123(2)
20247
123(2)
74222
123(2)
70930
123(2)
19628
Independent reflections (Rint
Observed reflections [IϾ2σ(I)] 5169
)
6147 (0.0416)
5360 (0.0333)
5069
10291 (0.1439)
6697
22967 (0.0943)
10867
5188 (0.0707)
4213
Final R1 (observed, all)
Final wR2 (observed, all)
0.0374, 0.0485
0.1028, 0.1238
0.0934, 0.0958
0.2693, 0.2709
0.0874, 0.1283
0.2475, 0.2831
0.1034, 0.1663
0.2863, 0.3081
0.0579, 0.0708
0.1552, 0.1637
material precluded all analyses except the crystallographic determi-
nation.
CCDC-917061 (for 1), -917062 (for 2), -917063 (for 3), -917064 (for
4) and -917065 (for 5) contain the supplementary crystallographic
data for this paper. These data can be obtained free of charge from
The Cambridge Crystallographic Data Centre via www.ccdc.cam.
ac.uk/data_request/cif.
Crystallographic Studies: X-ray diffraction intensity data for 1 was
collected at 123 K with a Bruker X8 Apex instrument equipped
with a KAPPA CCD detector by using Mo-Kα radiation (λ =
0.71073 Å). The data for 2, 3 and 5 were collected at 123 K at the
MX1 beamline at the Australian Synchrotron using synchrotron
radiation (λ = 0.7107 Å), and the data for 4 was analogously col-
lected at the MX2 beamline (λ = 0.7109 Å). Data collection and
integration for the synchrotron structures were performed with the
Blu-Ice[32] and XDS[33] software programs. The Apex structures
were solved by direct methods in SHELX-97[34] and refined by the
full-matrix method based on F2 with the program SHELXL-97 by
using the X-Seed software package.[35] All hydrogen atoms attached
to carbon atoms were included in the models at idealised positions
and were refined by using a riding model. Details of the data collec-
tions are given in Table 2.
Supporting Information (see footnote on the first page of this arti-
cle): Figure S1 shows the 1H NMR spectrum and signal assignment
for b3pd. A comparison of the copper halide coordination environ-
ments of 3 and 4 is shown in Figure S2 and is supplemented with
a discussion.
Acknowledgments
The authors gratefully acknowledge the Australian Research Coun-
cil and Monash University for financial support. This research was,
in part, undertaken at the macromolecular crystallography MX1
and MX2 beamlines at the Australian Synchrotron, Victoria, Aus-
tralia.
Crystallography Special Details: Manganese chain 2: oxygen atoms
O(5), O(7) and O(8) were constrained to have identical anisotropic
displacement parameters as O(6) by using the EADP command.
Cuprous 2D net 3: carbon atom C(15) was constrained to have
identical anisotropic displacement parameters as C(14) by using the
EADP command. The final difference map showed a large residual
hole (–5.96 eÅ3) 0.41 Å from I(2). Cuprous 1D chain 4: carbon
atoms C(18) and C(19) were constrained to have identical aniso-
tropic displacement parameters as C(17) by using the EADP com-
mand. Large voids with a volume of 1976 Å3 were observed in the
structure. Residual electron density located in these voids was un-
able to be assigned unambiguously and so was removed by using
the SQUEEZE command.[23] Elemental analysis suggests that the
electron density corresponds to a molecule of MeOH and 17 mole-
cules of H2O per unit of 4.
[1] H. Nishida, M. Tazaki, M. Takagi, K. Ueno, Microchim. Acta
1981, 281–287.
[2] G. W. Gokel, O. Murillo, Acc. Chem. Res. 1996, 29, 425–432.
[3] a) M. Cinquini, F. Montanari, P. Tundo, Gazz. Chim. Ital.
1977, 107, 11; b) D. Landini, A. Maia, F. Montanari, P. Tundo,
J. Am. Chem. Soc. 1979, 101, 2526.
[4] a) J. P. L. Cox, K. L. Lankowski, R. Kataky, D. Parker,
N. R. A. Beeley, B. A. Boyce, M. A. W. Eaton, K. Millar, A. T.
Millican, A. Harrison, C. Walker, J. Chem. Soc., Chem. Com-
mun. 1989, 797–780; b) T. A. Kaden, D. Tschudin, M. Studer,
U. J. Brunner, Pure Appl. Chem. 1989, 61, 879–883; c) D. Par-
ker, J. R. Morphy, K. Yankowski, J. Cox, Pure Appl. Chem.
1989, 61, 1637–1641.
[5] a) A. A. Adimado, R. C. Abaidoo, C. Bondzi, Transition Met.
Chem. 1991, 16, 407–409; b) H. O. Hankovszky, K. Hideg, I.
Bodi, L. Frank, Stud. Org. Chem. 1984, 205–215.
[6] a) K.-W. Chi, H.-C. Wei, T. Kottke, R. J. Lagow, J. Org. Chem.
1996, 61, 5684–5685; b) A. V. Bordunov, J. S. Bradshaw, V. N.
Pastushok, R. M. Izatt, Synlett 1996, 10, 933–948; c) N. Su,
J. S. Bradshaw, P. B. Savage, K. E. Krakowiak, R. M. Izatt,
S. L. De Wall, G. W. Gokel, Tetrahedron 1999, 55, 9737–9742.
[7] a) A. V. Bogatskii, N. G. Lukyanenko, V. N. Pastushok, R. G.
Kostyanovsky, Dokl. Akad. Nauk SSSR 1982, 265, 619–621; b)
A. V. Bogatsky, N. G. Lukyanenko, V. N. Pastushok, Synthesis
FTIR Fragmentation Study: Reference solvent measurements were
first obtained by immersing the diamond probe tip in a solution of
DMF and water (3:1, 4 mL). Next, b3pd (100 mg, 0.173 mmol) was
added, and the equilibration was monitored by subtracting the ref-
erence solvent spectrum. Once stable vibrational intensity for b3pd
was observed, zinc chloride (23.6 mg, 0.173 mmol) was added to
the solution, and changes in the reaction profile of b3pd were moni-
tored until completion as evidenced by static vibrational intensities
(120 min).
Eur. J. Inorg. Chem. 2013, 3240–3248
3247
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
1983, 992–994; c) N. Su, J. S. Bradshaw, G. Xue, N. Kent Dal-
ley, X. X. Zhang, P. B. Savage, K. E. Krakowiak, R. M. Izatt,
J. Heterocycl. Chem. 2000, 37, 711–718.
[19] K. Waszkowiak, K. Szymandera, Int. J. Food Sci. Technol.
2008, 43, 895–899.
[20] a) A. R. Hendrickson, R. L. Martin, N. M. Rode, Inorg. Chem.
1976, 15, 2115–2119; b) J. G. Wijnhoven, Th. E. M.
Van den Hark, P. T. Beurskens, J. Cryst. Mol. Struct. 1972, 2,
189–196; c) M. Somogyi, J. Biol. Chem. 1937, 117, 771–776; d)
P. A. Shaffer, M. Somogyi, J. Biol. Chem. 1933, 100, 695–713;
e) S. J. Stohs, D. Bagchi, Free Radical Biol. Med. 1995, 18, 321–
336.
[8] M. B. Duriska, S. M. Neville, S. R. Batten, Chem. Commun.
2009, 5579–5581.
[9] W. J. Gee, S. R. Batten, manuscript in preparation.
[10] a) G. J. Reiß, J. S. Engel, CrystEngComm 2002, 4, 155–161; b)
P. H. Svensson, M. Gorlov, L. Kloo, Inorg. Chem. 2008, 47,
11464–11466.
[11] Some examples include: a) J. Y. Lee, H. J. Kim, J. H. Jung, W.
Sim, S. S. Lee, J. Am. Chem. Soc. 2008, 130, 13838–13839; b)
Z. Liu, M. F. Qayyum, C. Wu, M. T. Whited, P. I. Djurovich,
K. O. Hodgson, B. Hedman, E. I. Solomon, M. E. Thompson,
J. Am. Chem. Soc. 2011, 133, 3700–3703; c) T. H. Kim, Y. W.
Shin, J. H. Jung, J. S. Kim, J. Kim, Angew. Chem. 2008, 120,
697; Angew. Chem. Int. Ed. 2008, 47, 685–688; d) J.-K. Cheng,
Y.-G. Yao, J. Zhang, Z.-J. Li, Z.-W. Cai, Z.-Y. Zhang, Z.-N.
Chen, Y.-B. Chen, Y. Kang, Y.-Y. Qin, Y.-H. Wen, J. Am.
Chem. Soc. 2004, 126, 7796–7797; e) H. Hartl, F. Mahdjour-
Hassan-Abadi, Angew. Chem. 1994, 106, 1929; Angew. Chem.
Int. Ed. Engl. 1994, 33, 1841–1842; f) E. Cariati, R. Ugo, F.
Cariati, D. Roberto, N. Masciocchi, S. Galli, A. Sironi, Adv.
Mater. 2001, 13, 1665–1668; g) J.-W. Cheng, J. Zhang, S.-T.
Zheng, G.-Y. Yang, Chem. Eur. J. 2008, 14, 88–97.
[12] E. Graf, M. W. Hosseini, J.-M. Planeix, N. Kyritakas, New J.
Chem. 2005, 29, 343–346.
[13] a) H. Braband, S. Imstepf, M. Felber, B. Spingler, R. Alberto,
Inorg. Chem. 2010, 49, 1283–1285; b) S.-L. Zheng, M.-L. Tong,
X.-M. Chen, Coord. Chem. Rev. 2003, 246, 185–202.
[14] M. Nematirad, W. J. Gee, S. K. Langley, N. F. Chilton, B.
Moubaraki, K. S. Murray, S. R. Batten, Dalton Trans. 2012,
41, 13711–13715.
[15] a) S.-B. Teo, C.-H. Ng, S.-G. Teoh, C. Wei, J. Coord. Chem.
1995, 36, 141–148; b) S.-G. Kang, J. Song, J. H. Jeong, Inorg.
Chim. Acta 2004, 357, 605–610.
[21] K. Takahashi, Y. Mazaki, K. Kobayashi, Chem. Commun.
1996, 2275–2276.
[22] A. L. Speck, PLATON, a Multipurpose Crystallographic Tool,
Utrecht University, The Netherlands, 2001. Available from:
Ͻhttp://www.cryst.chem.uu.nl/platon/Ͼ.
[23] A. L. Spek, J. Appl. Crystallogr. 2003, 36, 7–13.
[24] W. J. Gee, S. R. Batten, Chem. Commun. 2012, 48, 4830–4832.
[25] ConcIRT Opus 1, v. 2.0, Mettler Toledo.
[26] I. Poljansˇek, B. Likozar, M. Krajnc, J. Appl. Polym. Sci. 2007,
106, 878–888.
[27] D. G. Blackmond, Angew. Chem. 2005, 117, 4374; Angew.
Chem. Int. Ed. 2005, 44, 4302–4320.
[28] V. J. Gatto, K. A. Arnold, A. M. Viscariello, S. R. Miller, C. R.
Morgan, G. W. Gokel, J. Org. Chem. 1986, 51, 5373–5384.
[29] H. Adams, S. R. Batten, G. M. Davies, M. B. Duriska, J. C.
Jeffery, P. Jensen, J. Lu, G. R. Motson, S. J. Coles, M. B. Hurst-
house, M. D. Ward, Dalton Trans. 2005, 1910–1923.
[30] L. J. Barbour, J. Supramol. Chem. 2003, 1, 189–191.
[31] POV-Ray, v. 3.6, 2006, Persistence of Vision Raytracer Pty
Ltd., Williamstown.
[32] T. M. McPhillips, S. E. McPhillips, H. J. Chiu, A. E. Cohen,
A. M. Deacon, P. J. Ellis, E. Garman, A. Gonzalez, N. K. Sau-
ter, R. P. Phizackerley, S. M. Soltis, P. Kuhn, J. Synchrotron Ra-
diat. 2002, 9, 401–406.
[33] W. Kabsch, J. Appl. Crystallogr. 1993, 26, 795–800.
[34] G. M. Sheldrick, Acta Crystallogr., Sect. A 2008, 64, 112–122.
[35] L. J. Barbour, X-Seed: A graphical interface for use with
SHELX97 program suite, University of Missouri, Columbia,
1999.
[16] D. A. Palmer, R. W. Ramette, R. E. Mesmer, J. Solution Chem.
1984, 13, 673–683.
[17] F. A. Cotton, G. Wilkinson in Advanced Inorganic Chemistry,
5th ed., John Wiley & Sons, 1988, pp. 1186–1190.
[18] P. H. Svensson, L. Kloo, Chem. Rev. 2003, 103, 1649–1684.
Received: March 11, 2013
Published Online: May 10, 2013
Eur. J. Inorg. Chem. 2013, 3240–3248
3248
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim