Synthetic, structural, electrochemical and solvent extraction studies of neutral trinuclear Co(ii), Ni(ii), Cu(ii) and Zn(ii) metallocycles and tetrahedral tetranuclear Fe(iii) species incorporating 1,4-aryl-linked bis-β-diketonato ligands

Article abstract of DOI:10.1039/b716653b

Uncharged complexes, formulated as trimeric metallocycles of type [M ₃(L¹)₃(Py)₆] (where M = cobalt(ii), nickel(ii) and zinc(ii) and L¹ is the doubly deprotonated form of a 1,4-phenylene linked bis-β-

Full text of DOI:10.1039/b716653b

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Synthetic, structural, electrochemical and solvent extraction studies of neutral
trinuclear Co(II), Ni(II), Cu(II) and Zn(II) metallocycles and tetrahedral
tetranuclear Fe(III) species incorporating 1,4-aryl-linked bis-b-diketonato
ligands†‡§
Jack K. Clegg,a David J. Bray,a Kerstin Gloe,b Karsten Gloe,*b Katrina A. Jolliffe,c Geoffrey A. Lawrance,d
Leonard F. Lindoy,*a George V. Meehane and Marco Wenzelb
Received 30th October 2007, Accepted 17th December 2007
First published as an Advance Article on the web 16th January 2008
DOI: 10.1039/b716653b
Uncharged complexes, formulated as trimeric metallocycles of type [M3(L1)3(Py)6] (where M =
cobalt(II), nickel(II) and zinc(II) and L1 is the doubly deprotonated form of a 1,4-phenylene linked
bis-b-diketone ligand of type 1,4-bis(RC(O)CH2C(O))C6H4 (R = t-Bu)) have been synthesised, adding
to related, previously reported complexes of these metals with L1 (R = Ph) and copper(II) with L1 (R =
Me, Et, Pr, t-Bu, Ph). New lipophilic ligand derivatives with R = hexyl, octyl or nonyl were also
prepared for use in solvent extraction experiments. The X-ray structures of H2L1 (R = t-Bu) and of its
trinuclear (triangular) nickel(II) complex [Ni3(L1)3(Py)6]·3.5Py (R = t-Bu) are also presented.
Electrochemical studies of H2L1, [Co3(L1)3(Py)6], [Ni3(L1)3(Py)6], [Cu3(L1)3], [Zn3(L1)3(Py)6] and
[Fe4(L1)6] (all with R = t-Bu) show that oxidative processes for the complexes are predominantly
irreversible, but several examples of quasireversible behaviour also occur and support the assignment of
an anodic process, seen between +1.0 and +1.6 V, as involving metal-centred oxidations. The reduction
behaviour for the respective metal complexes is not simple, being irreversible in most cases. Solvent
extraction studies (water/chloroform) involving the systematic variation of the metal, bis-b-diketone
and heterocyclic base concentrations have been performed for cobalt(II) and zinc(II) using a radiotracer
technique in order to probe the stoichiometries of the respective extracted species. Significant extraction
synergism was observed when 4-ethylpyridine was also present with the bis-b-diketone ligand in the
chloroform phase. Competitive extraction studies demonstrated a clear uptake preference for copper(II)
over cobalt(II), nickel(II), zinc(II) and cadmium(II).
development, several groups have extended the chemistry of such
systems by incorporating more than one b-diketone motif,5–12 or
Introduction
a related malonate motif,13 into extended ligand structures before
investigating their metal ion chemistry. For example, we14–19 and
others20,21 have recently reported the use of a series of bis-b-
diketone ligands of types H2L1 and H2L2 for the construction
of a variety of new metallo-supramolecular arrays. The largely
rigid sp2 hybridised backbone of the doubly deprotonated forms
of these ligands results in particular defined geometries, which
when combined with metal ions such as cobalt(II), nickel(II),
copper(II), zinc(II) iron(III), and gallium(III) produce a variety
of discrete and extended neutral metallo-species. For exam-
ple, with copper(II) these include neutral dinuclear species of
The metal coordination chemistry of b-diketone ligands has
now been investigated for over 100 years1 and it has been
well demonstrated that these versatile ligands give rise to a
diverse range of interesting metal coordination behaviour.2 More
recently, there has also been increasing interest in the metallo-
supramolecular structural aspects of such systems.3,4 In a parallel
aCentre for Heavy Metals Research, School of Chemistry, University of
Sydney, NSW, 2006, Australia
bDepartment of Chemistry and Food Chemistry, Technical University
Dresden, 01062, Dresden, Germany
type M2L2 (solvent)n,15,17 as well as neutral trinuclear (triangular)
cSchool of Chemistry, University of Sydney, NSW, 2006, Australia
dDiscipline of Chemistry, School of Environmental and Life Sciences, The
University of Newcastle, Callaghan, NSW, 2308, Australia
2
species of type M3L1 (solvent)n,14,16 with trivalent ions dinuclear
3
triple helical complexes of type M2L2 9,15,22 and tetrahedral-shaped
3
eSchool of Pharmacy and Molecular Sciences, James Cook University,
Townsville, Qld, 4811, Australia
species of type M4L16 are obtained.14 In a recent extension of these
studies we reported19 aspects of structural, electrochemical and
liquid–liquid extraction behaviour involving cobalt(II), nickel(II),
copper(II), zinc(II) and iron(III) complexes incorporating ligands
of type H2L2. In particular, we observed marked synergism in the
extraction of cobalt(II) and zinc(II) when 4-ethylpyridine (EtPy)
was added as a co-extractant to the bis-b-diketone in the organic
phase.
† The HTML version of this article has been enhanced with colour images.
‡ CCDC reference numbers 665749 and 665750. For crystallographic data
in CIF or other electronic format see DOI: 10.1039/b716653b
§ Electronic supplementary information (ESI) available: Cyclic voltam-
mograms for H2L1 (R = t-Bu) together with those for [Co3(L1)3(Py)6]
(R = t-Bu), [Ni3(L1)3(Py)6] (R = t-Bu) and [Zn3(L1)3(Py)6] (R = t-Bu);
representative plots for the extraction experiments; tables of selected bond
lengths and angles. See DOI: 10.1039/b716653b
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1331–1340 | 1331
©

View Article Online
CH2–, 4H), 1.34 (br m, –CH2–, 12H), 0.90 ppm (br m, –CH3, 6H).
(ESI-MS): m/z 387 (M + H)+, 385 (M − H)−. FTIR (KBr) (major
peaks only): 3000 (br), 1601 (vbr), 1559, 1290, 858, 775 cm−1.
Found: C, 73.48; H, 9.55%. Calc. for C24H34O4·(EtOH)0.5: C, 73.30;
H, 9.11%.
1,1ꢀ-(1,4-Phenylene)-bis-undecane-1,3-dione, H2L1 (R = octyl).
In our prior studies it was demonstrated that neutral species
1
Yield 8.48 g (75%), pale yellow platelets. H-NMR d (200 MHz
of stoichiometry Cu3L1 (solvent)n, containing triangular planar
3
CDCl3): 16.30 (br s, enol, 2H), 7.98 (s, aromatic, 4H), 6.23 (s,
–CH– enol, 2H), 2.44 (br m, –CH2–, 4H), 1.70 (br m, –CH2–, 4H),
1.34 (br m, –CH2–, 20H), 0.89 ppm (br m, –CH3, 6H). (ESI-MS):
m/z 443 (M + H)+, 466 (M + Na)+, 441 (M − H)−. FTIR (KBr)
(major peaks only): 3002 (br), 1598 (vbr), 1507, 1292, 1118, 812,
770 cm−1. Found: C, 76.12; H, 9.60%. Calc. for C28H42O4: C, 75.98;
H, 9.56%.
Cu3L1 units, can be used for the formation of extended multi-
3
nuclear architectures by reaction with difunctional heterocyclic
nitrogen ligands that include pyrazine, dabco and 4,4ꢀ-bipyridine.16
In contrast to the situation for copper, the metal ion chemistry
of H2L1 with other divalent metals has received less attention,
although two triangular cobalt(II) and nickel(II) derivatives and a
polymeric zinc(II) complex derived from H2L1 (with R = Ph) have
been reported.20 It is also noted that related triangular complexes
incorporating difunctional, non-b-diketone ligand derivatives of a
range of metals are also known.23
In an extension of our prior studies of copper(II) and iron(III)
complexes of H2L1 derivatives,14 we now report an investigation of
new species incorporating cobalt(II), nickel(II) and zinc(II) derived
from this ligand type; these products also incorporate pyridine
(Py) as a co-ligand. Aspects of the synthesis, electrochemistry
and liquid–liquid extraction properties of these systems are now
presented.
1,1ꢀ-(1,4-Phenylene)-bis-dodecane-1,3-dione, H2L1·0.125CHCl3
(R = nonyl). Yield 8.16 g (68%), pale yellow platelets. 1H-NMR
d (200 MHz CDCl3): 16.02 (br s, enol, 2H), 7.96 (s, aromatic, 4H),
6.20 (s, –CH– enol, 2H), 2.46 (br m, –CH2–, 4H), 1.71 (br m, –
CH2–, 4H), 1.33 (br m, –CH2–, 24H), 0.88 ppm (br m, –CH3, 6H).
(ESI-MS): m/z 471 (M + H)+, 469 (M − H)−. FTIR (KBr) (major
peaks only): 3000 (br), 1601 (vbr), 1525, 1397, 1069, 798, 681 cm−1.
Found: C, 74.77; H, 8.90%. Calc. for C30H46O4·(CHCl3)0.125: C,
74.51; H, 9.57%.
Synthesis of [Co3(L1)3(Py)6] (R = t-Bu), [Ni3(L1)3(Py)6] (R =
t-Bu), [Zn3(L1)3(Py)6] (R = t-Bu)
Experimental
H2L1 (R = t-Bu) (0.4 mmol) in dry tetrahydrofuran (40 ml)
was added to Na2CO3 (1.0 g, ∼0.01 mol) suspended in dry
tetrahydrofuran (10 ml). This mixture was stirred for 1 h before
the appropriate metal(II) chloride (0.4 mmol) in tetrahydrofuran
(40 ml) and pyridine (0.8 mmol) were added dropwise. The mixture
was stirred at 40 ◦C for 16 h. The resulting solution was filtered and
the volume of the filtrate was then reduced to a small volume in a
rotary evaporator to yield the solid complex as a microcrystalline
product which was isolated and washed with ether.
NMR spectra were recorded on Bruker Avance DPX200, DPX300
or DPX400 spectrometers. Low resolution electrospray ionisation
mass spectra (ESI-MS) were obtained on a Finnigan LCQ-8
spectrometer. FTIR (KBr) spectra were collected using a Bio-Rad
FTS-40 spectrometer. UV-Vis spectra were recorded on a Cary
1E spectrophotometer in the solid state (all reported peaks are
broad). Tetrahydrofuran was pre-dried over sodium wire before
use. Ligands of type H2L1 (R = Pr, t-Bu and Ph) were prepared
using the previously reported method14 [Cu3(L1)3] and [Fe4(L1)6]
were also synthesised as reported previously.14,17
[Co3(L1)2(Py)6]·H2O (R = t-Bu). Yield 174 mg (79%), orange
microcrystalline powder. Found: C, 65.37; H, 6.24; N, 4.92%. Calc.
for C90H102N6Co3O12·H2O.: C, 65.31; H, 6.34; N: 5.08%. UV-Vis
(solid state): 350, 375, 415, 460, 560 nm.
General preparation of H2L1 (R = hexyl, octyl and nonyl)
Sodium amide (5.0 g, 0.13 mol) was added to a mixture of dimethyl
terephthalate (5.00 g, 0.025 mol) and 2-octanone (for H2L1; R =
hexyl) (6.4 g, 0.05 mol) or 2-decanone (for H2L1; R = octyl) (7.8 g,
0.05 mol) or 2-undecanone (for H2L1; R = nonyl) (8.5 g, 0.05 mol)
in dry tetrahydrofuran (100 ml) at 0 ◦C. The reaction was stirred for
2 h at this temperature and then for 2 h at 26 ◦C over which time
the mixture turned yellow. The reaction was quenched with ice
water (100 ml) and then acidified to pH 4 using 1 M hydrochloric
acid; two phases formed. The organic (tetrahydrofuran) phase was
separated from the aqueous phase. The aqueous phase was washed
twice with chloroform (100 ml), the organic phases were combined
and the solvent was removed under reduced pressure to yield the
crude product which was recrystallised from ethanol. Yields and
characterisation details for each product are given below.
[Ni3(L1)3(Py)6]·Py·4THF (R = t-Bu). Yield 149 mg (56%),
green microcrystalline powder. Found: C, 66.53; H, 6.79; N, 4.53%.
Calc. for C90H102 N6 Ni3O12·Py·4THF: C, 66.61; H, 7.00; N, 4.90%.
UV-Vis (solid state): 360, 410, 560 nm. Crystals used for the X-ray
study were obtained by dissolving this product in pyridine and
letting the solution stand (see later).
[Zn3(L1)3(Py)6] (R = t-Bu). Yield 165 mg (75%), colourless
microcrystalline powder. Found: C, 65.40; H, 6.42; N, 5.16%. Calc.
1
for C90H102N6O12Zn3 : C, 65.43; H, 6.23; N, 5.09%. H-NMR d
(200 MHz, DMSO-d6, 300 K): 8.61 (m, Py, 4H), 7.96 (s, aromatic,
4H), 7.88 (m, Py, 2H), 7.39 (m, Py, 4H), 6.24 (s, –CH– enol, 2H),
1.94 (s, CH3, 18H). UV-Vis (solid state): 359, 415 nm.
1,1ꢀ-(1,4-Phenylene)-bis-nonane-1,3-dione, H2L1·0.5EtOH (R =
hexyl). Yield 6.85 g (70%), pale yellow platelets. 1H-NMR d
(200 MHz CDCl3): 16.23 (br s, enol, 2H), 7.95 (s, aromatic, 4H),
6.24 (s, –CH– enol, 2H), 2.45 (br m, –CH2–, 4H), 1.70 (br m, –
Electrochemical studies
The electrochemistry of H2L1 (R = t-Bu), along with its trin-
uclear neutral complexes, [M3(L1)3(Py)6] (M = Co, Ni, Zn),
1332 | Dalton Trans., 2008, 1331–1340
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
[Cu3(L1)3] and [Fe4(L1)6], was investigated in acetonitrile or an
acetonitrile/tetrahydrofuran mixture. Cyclic voltammetry studies
were performed using a conventional three-electrode configuration
with iR compensation and a BAS Model 100B electrochemical
system, operated by a computer using the BAS100W software. The
working electrodes were glassy carbon (GCE) or platinum discs,
polished with alumina. The reference electrode was Ag/AgCl and
was separated from the working and Pt wire auxiliary electrodes
in a glass sleeve fitted with a Vycor frit. Solutions were made up in
HPLC-grade acetonitrile or acetonitrile–tetrahydrofuran (usually
1 : 1) and were 0.05–0.1 M in (Et4N)(ClO4); they were purged with
argon gas. Complex concentrations in the millimolar range were
used throughout. Unless otherwise stated, results discussed below
refer to behaviour at the GCE.
4-Ethylpyridine dependence studies were also carried out.
These involved varying the 4-ethylpyridine concentration from
2 × 10−4 M to 2 × 10−3 M while maintaining the metal ion
concentration at 1 × 10−4 M and the ligand concentration at 1 ×
10−3 M. Log DM was then plotted against log [4-ethylpyridine]org
and the slope determined.
A series of experiments were also performed in which the metal
ion concentration was varied from 1 × 10−4 M to 5 × 10−3
M
at pH 8.7 (TAPS/NaOH), both in the absence and presence
of 4-ethylpyridine (2 × 10−3 M), while maintaining the ligand
concentration at 1 × 10−3 M. From the observed maximum metal
loading of the organic phase, the stoichiometry of the extracted
complex can be deduced.26
Multi-metal competitive extraction experiments were also per-
formed using an equimolar mixture of the following metal ions
in the aqueous phase: cobalt(II), nickel(II), copper(II), zinc(II)
and cadmium(II) at pH 7.8 in the absence of 4-ethylpyridine in
the organic phase or at pH 7.4 (HEPES/NaOH buffer) in its
presence. The aqueous phase contained each of the above metal
perchlorate salts at a concentration of 1 × 10−4 M and the organic
phase (chloroform) contained a 1 × 10−3 M concentration of
H2L1 and in the second set of experiments also 4-ethylpyridine at
2 × 10−3 M.
Solvent extraction studies
Liquid–liquid extraction experiments were performed at 24
1
◦C in microcentrifuge tubes (2 ml) with a phase ratio V(org)
:
V(aq) of 1 : 1 (500 ll each). The aqueous phase contained the
metal perchlorate (normally 1 × 10−4 M, except where variable
concentration experiments were employed), and a selected buffer.
The zwitterionic buffer systems, TAPS/NaOH, HEPES/NaOH
and MES/NaOH, were used (depending on the chosen pH)
to maintain the pH in this phase; as a precaution, the pH
of this phase was checked before and after each experiment.
A majority of the experiments were carried out at pH 8.7
using the TAPS/NaOH system. The chloroform organic phase
contained a known concentration of a H2L1 ligand (normally 1 ×
10−3 M, except where variable concentration experiments were
employed) and in some experiments also 4-ethylpyridine (normally
2 × 10−3 M except where variable concentration experiments
were conducted). All solvent extraction experiments involved the
mechanical shaking of the two-phase system until equilibrium
had been established. The overall error for individual extractions
is estimated to be approximately 5%.
In the case of the single metal ion experiments a shaking time
of 3 h was used for the cobalt(II) experiments and 30 min for
the zinc(II) experiments. At the end of these times, the phases
were separated, centrifuged and then duplicate 100 ll samples
of the aqueous and organic phases were removed for analysis.
The concentration of the cobalt or zinc was measured in both
phases radiometrically using 60Co and 65Zn radioisotopes (ROTOP
Pharmaka) by means of a NaI (TI) scintillation counter (Cobra
II/Canberra-Packard).24,25
The competitive (multi-metal) extraction experiments employed
a shaking time of 24 h. For these experiments the depletion
of the metal ion concentrations in the aqueous phase was
measured using an ICP-MS (ELAN 9000/Perkin Elmer) spectro-
meter.
Three sets of variable concentration experiments were also
undertaken. Ligand dependence studies were performed in which
the concentration of H2L1 (R = t-Bu, nonyl)) was varied from
5 × 10−4 to 1.5 × 10−3 M while the zinc(II) concentration was
maintained at 1 × 10−4 M. Similar experiments were carried out
in the presence of 4-ethylpyridine (2 × 10−3 M) for both cobalt(II)
and zinc(II). In each case log DM was plotted against log [H2L1]org
and the slope determined.
X-Ray structural determinations
Data for H2L1 (R = t-Bu) were collected with x scans to
approximately 56◦ 2h using a Bruker SMART 1000 diffractometer
employing graphite-monochromated Mo-Ka radiation generated
˚
from a sealed tube (0.71073 A) at 150(2) K. Data integration
and reduction were undertaken with SAINT and XPREP.27 Data
for [Ni3(L1)3(Py)6]·3.5Py (R = t-Bu) were collected on a Bruker-
Nonius APEX2-X8-FR591 diffractometer employing graphite-
monochromated Mo-Ka radiation generated from a rotating
◦
˚
anode (0.71073 A) with x and w scans to approximately 44 2h at
100(2) K.28 Subsequent computations were carried out using the
WinGX-32 graphical user interface.29 Structures were solved by
direct methods using SIR97.30 Multi-scan empirical absorption
corrections were applied to the data set using the program
SADABS.31 Data were refined and extended with SHELXL-97.32
All non-hydrogen atoms were refined anisotropically. Carbon-
bound hydrogen atoms were included in idealised positions and
refined using a riding model. Oxygen bound hydrogen atoms
were first located in the difference Fourier map before refine-
ment.
Crystal and structure refinement data for H2L1 (R = t-Bu) and
[Ni3(L1)3(Py)6]·3.5Py (R = t-Bu) are summarised below and an
ORTEP33 depiction of the structure of H2L1 (R = t-Bu) is provided
in Fig. 1, schematic representations of [Ni3(L1)3(Py)6]·3.5Py (R =
t-Bu) are provided in Fig. 2 and Fig. 3. Tables of selected bond
lengths and angles are given in the ESI.§
H2L1 (R = t-Bu). Formula C20H26O4, M 330.41, orthorhom-
bic, space group Pbca (#61), a 10.232(3), b 10.158(3), c
3
17.738(5) A, V 1843.8(9) A , Dc 1.190 g cm−3, Z 4, crystal
˚
˚
size 0.600 by 0.370 by 0.032 mm, colour colourless, habit plate,
−1
˚
temperature 150(2) K, k(MoKa) 0.71073 A, l(MoKa) 0.082 mm ,
T(SADABS)min,max 0.864765, 1.000000, 2hmax 56.58, hkl range −13
to 13, −13 to 13, −22 to 22, N 16191, Nind 2229 (Rmerge 0.0612),
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1331–1340 | 1333
©

View Article Online
N
obs 1380 (I > 2r(I)), Nvar 115, residuals¶ R1 (F) 0.0498, wR2 (F2)
keto and enol forms in solution7,11,12—an observation also made
for simple b-diketones some considerable time ago.38 We have
previously reported the X-ray structure of H2L2 (R = t-Bu) in
which it was shown that this ligand exists exclusively in its (bis)
enol form both in solution and in the crystalline state.15 Similar
results have also been reported for the H2L1 (R = Ph) ligand21 and
a ligand related to H2L2 incorporating a resorcinol spacer.10
In view of the above it was of interest to confirm what
configuration the present ligands of type H2L adopt in the
solid state with respect to possible enol/keto tautomerism as
well as to determine the relative orientations of the b-diketone
fragments with respect to each other. Single crystals suitable for
X-ray diffraction were grown by slow evaporation of a diethyl
ether solution of H2L1 (R = t-Bu). The X-ray structure (Fig. 1)
reveals that the b-diketone fragments are essentially in the same
plane as the central phenylene core, reflecting the presence of
the sp2 conjugated backbone and perhaps aided by weak C(H)-
phenylene to oxygen interactions present in this arrangement. The
two b-diketone fragments are orientated with their coordination
vectors aligned approximately 180◦ with respect to each other.
The b-diketone groups exist in their enol form, with the enol
hydrogen atoms located on the oxygens (O(1)) furthest from
the central phenylene ring; this contrasts with the solid state
structure of the H2L2 (R = t-Bu) isomer in which the hydrogens
were located on the oxygens closest to the phenylene ring15 and
H2L1 (R = Ph) in which each hydrogen was partially located
on each b-diketone oxygen atom (with 50% occupancy).17 In the
present structure the expected ‘enolic’ intramolecular hydrogen
bonds are present between each pair of adjacent oxygen atoms
0.1333, A = 0.0638, B = 0.4206, GoF (all) 1.013, Dqmin,max −0.193,
−3
˚
0.234 e A .
[Ni3(L1)3(Py)6]·3.5Py (R = t-Bu). Formula C107.50H119.50N9.50
-
¯
Ni3O12, M 1912.76, triclinic, space group P1 (#2), a 8.9493(10),◦b
˚
23.585(3), c 26.145(3) A, a 104.688(8), b 93.275(8), c 98.843(7) ,
3
−3
˚
V 5247.8(11) A , Dc 1.210 g cm , Z 2, crystal size 0.300 by 0.110
by 0.020 mm, colour yellow, habit needle, temperature 100(2) K,
k(MoKa) 0.71073 A, l(MoKa) 0.594 mm , T(SADABS)min,max
−1
˚
0.456, 0.988, 2hmax 44.00, hkl range −9 to 9, −24 to 24, −27 to 27,
N 127447, Nind 12860 (Rmerge 0.2848), Nobs 6417 (I > 2r(I)), Nvar
1048, residuals¶ R1 (F) 0.0913, wR2 (F2) 0.2723, A = 0.1690, B =
−
−3
˚
0.00, GoF (all) 0.947, Dqmin,max −1.526, 0.930 e A .
Specific details. Despite numerous recrystallisation attempts
from various solvents, the crystals obtained and employed in this
study (despite appearing free from physical imperfections) were
extremely small and diffracted poorly. The crystals were extremely
sensitive to solvent loss and despite rapid handling during the
mounting process (ca. 1 min at 200 K) prior to quenching in the
cryostream less than ideal data were obtained. Even with a high
power machine and long exposure times the highest resolution
data recorded were at ∼ 44◦ 2h. This has resulted in higher than
ideal values of Rint, lower than ideal bond precision and difficulties
in obtaining a satisfactory absorption correction. The C(39) and
C(58) containing tertiary-butyl groups are rotationally disordered
and have been modelled over two positions. The N(10)-containing
pyridine solvent molecule is modelled as 50% occupancy and
is consistent with the rapid solvent loss and poor diffraction
properties observed. A variety of constraints and restraints were
required in the chelate rings to facilitate realistic modelling.
˚
(O(1) · · · O(2) distance 2.479 (2) A; O(1)–H(13)–O(2) angle 154
(2)◦). Analysis of the crystal packing shows the presence of a
number of close CH(phenylene)–O interactions within the lattice
˚
˚
(H(9)–O(2), 2.6 A; CH(t-Bu)–O(H(6), 2.6 A; CH(Hc)–O(3A)–
Results and discussion
˚
O(2), 2.6 A), giving rise to an infinite sheet-like motif.
Ligand synthesis and X-ray structure of H2L1 (R = t-Bu)
Claisen condensations were employed to synthesise the ligands
of type H2L1 in good yield using an adaptation of literature
procedures.5,14,34,35 The 1H NMR spectra of deuterochloroform
solutions of the respective ligands of type H2L1 with R = Pr,
t-Bu, Ph, hexyl, octyl, nonyl in each case indicated that these
species exist almost entirely in their (bis) enol tautomeric forms,
shown by the presence of a broad resonance at ∼16.2 ppm (enol
protons), integrating to two protons, in the respective spectra.
The existence of a sharp singlet in each spectrum at ∼6.2 ppm
(due to Hc), integrating to two protons, is also in harmony
with this assignment. In the solid state, the infrared spectra of
the solid ligands are in keeping with the presence of hydrogen
bonded hydroxy and carbonyl groups (absorptions at ∼3000 and
∼1600 cm−1, respectively);36 the occurrence of O–H · · · O hydrogen
bonding in the enol forms of the b-diketone fragments has been
proposed to confer added stability on such a structure with respect
to its corresponding (bis) keto form.37 Nevertheless, it is noted that
a number of bis- and tris-b-diketone ligands have been reported
recently in which the b-diketone groups exist as a mixture of both
Fig. 1 ORTEP33 representation of the X-ray crystal structure of H2L1
(R = t-Bu) shown with 50% probability ellipsoids. The molecule occupies
a crystallographic 2-fold site (symmetry code; 1 − x, −y, 1 − z).
Complex formation and X-ray structure of [Ni3(L1)3(Py)6]·3.5Py
New mixed ligand complexes of cobalt(II), nickel(II) and zinc(II)
incorporating both doubly deprotonated H2L1 and pyridine have
been synthesised for use in electrochemical studies and to provide a
background for mixed-ligand solvent extraction studies described
later. The products were obtained in good yield and purity
upon reaction of a 1 : 1 ratio of the H2L1 ligand with the
appropriate metal chloride in tetrahydrofuran in the presence
of excess pyridine and solid sodium carbonate. This procedure
differs from those employed previously by Soldatov et al.20 as
part of a crystal engineering study for obtaining related mixed
ꢀ
ꢀ
ꢀ
2
¶ R1 = ꢁFo| − |Fcꢁ/ |Fo| for Fo > 2r(Fo); wR2 = ( w(Fo
−
ꢀ
Fc2)2/ (wFc2)2)1/2 all reflections; w = 1/[r2(Fo2) + (AP)2 + BP] where
2
P = (Fo + 2Fc2)/3.
1334 | Dalton Trans., 2008, 1331–1340
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
ligand metal complexes of deprotonated H2L1 (R = Ph) but is
analogous to that used by us for the synthesis of corresponding
triangular (trinuclear) copper(II) complexes incorporating depro-
tonated H2L1 ligands14,17 and dinuclear cobalt(II), nickel(II) and
zinc(II) complexes containing deprotonated H2L2 ligands.19 The
following new derivatives were isolated in the present study:
[Co3(L1)3(Py)6]·H2O (R = t-Bu), [Ni3(L1)3(Py)6]·Py·4THF (R =
t-Bu) and [Zn3(L1)3(Py)6] (R = t-Bu); all contain M : L : Py ratios
of 1 : 1 : 2 and were formulated as trinuclear by analogy with
the corresponding triangular copper(II) species of L1 (R = Me,
Pr)14 and the analogous cobalt(II) and nickel(II) complexes of this
ligand type (with R = Ph),20 all of which have been characterised
by single crystal X-ray structures. In this context it is noted that
[Zn3(L1)3(Py)6] (R = t-Bu) yields a symmetric 1H NMR spectrum
confirming ligand equivalence in DMSO-d6 solution.
The X-ray structure of the above nickel(II) complex recrys-
tallised from pyridine, [Ni3(L1)3(Py)6]·3.5Py (R = t-Bu), confirms
its triangular nature and shows that each nickel centre adopts
a tetragonally distorted octahedral arrangement (mean Ni–O
˚
˚
lengths, 2.020 A; mean Ni–N lengths, 2.108 A) with the pyridine
ligands occupying axial sites (Fig. 2); the structure is generally
similar to the corresponding structure reported by Soldatov (with
R = Ph).20,21 It is noted that recrystallisation of the present complex
resulted in very small needle-like crystals that rapidly decay out
of pyridine. Despite rapid handling at dry-ice temperature, the
diffraction qualities of the crystals were not ideal. Even though
long exposure times were used and a comparatively high powered
laboratory X-ray source (5 kW from a rotating anode) was
Fig. 3 A schematic representation of part of one of the ‘prismatic’
columns formed by offset face-to-face p–p stacking between the coordi-
nated pyridyl ligands in adjacent molecules and showing the encapsulated
pyridine solvent molecules.
˚
employed, only reflections to ∼1.0 A resolution could be recorded.
one another along the a-axis. The above extended packing motif
results in the formation of a number of voids within the structure.
There are pyridine solvent molecules present between each of
the columns and also encapsulated within the channels formed
by the stacked prisms (Fig. 3). Each of these pyridine solvent
molecules is involved in both edge-to-face and offset face-to-face
p–p interactions with other pyridine molecules (both coordinated
and uncoordinated) and with the phenylene rings of the L1 ligands.
Electrochemical studies
The electrochemistry of H2L1 (R = t-Bu), both protonated
and deprotonated, along with the present trinuclear neutral
complexes, [M3(L1)3(Py)6] (M = Co, Ni, Zn), has been investigated
in acetonitrile or acetonitrile/tetrahydrofuran. Additionally, a
comparative investigation of the previously reported14,17 complexes
[Cu3(L1)3] and [Fe4(L1)6] (both with R = t-Bu) was also under-
taken. The results for these ligand and complexes species are
summarised in Table 1. Since the paramagnetic iron(III), cobalt(II),
nickel(II) and copper(II) centres in the respective complexes are
almost certainly sufficiently separated not to be significantly
magnetically coupled,14,17,19 Mn+M(n+1)+ or Mn+M(n−1)+ intermediate
systems where the metals ‘share’ the higher or lower oxidation
state effectively equally are not expected to occur. Instead,
oxidation or reduction at one metal centre is anticipated to occur
initially, thereby affecting the redox behaviour of the other metal
centres. As a consequence the latter metal centres are expected
to be oxidised or reduced at a somewhat different observed
Fig. 2 A schematic representation of the X-ray structure of [Ni3(L1)3-
(Py)6]·3.5Py (R = t-Bu), hydrogen atoms, solvent molecules and regions of
disorder are not shown for clarity.
The structure shows the presence of disordered pyridine solvent
molecules in the lattice. As can be seen in Fig. 2, the triangular
structure is somewhat distorted from an ideal equilateral triangle,
˚
˚
with Ni–Ni distances ranging from 10.56 A to 10.77 A. The N–
Ni–N axes through each of the metal centres are also significantly
distorted from the ideal perpendicular to the average plane of the
triangle. This is in part due to the presence of strong offset face-
to-face p–p stacking between the coordinated pyridyl ligands in
adjacent molecules. This results in an infinite one-dimensional
‘prismatic’ arrangement (Fig. 3) with triangles stacking above
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1331–1340 | 1335
©

View Article Online
for [Co3(L1)3(Py)6] (R = t-Bu), [Ni3(L1)3(Py)6] (R = t-Bu) and
[Zn3(L1)3(Py)6] (R = t-Bu) are given as Fig. S1 in the ESI.§
1
2
Table 1 Oxidation and reduction processes (Ep or E ,
V versus
Ag/AgCl)a observed for complexes of L1
[Zn3(L1)3(Py)6] (R = t-Bu). Because zinc(II) is not electroac-
tive, any observed redox processes must originate in the ligand
framework, albeit influenced by the coordinated metal ion.
The zinc(II) complex displays an irreversible oxidation process
+1.43 V. This amounts to a cathodic shift of ∼0.6 V as a result
of complexation, although of course, the character of the ligand
has changed from neutral to dianionic. At a GCE, the major
reduction wave lies at −1.61 V, with a broad ill-defined following
wave at −2.0 V, and pre-waves at −1.16 and −1.31 V, as well as a
backwave at −1.45 V that appears from wave shape and size/scan
rate profile to be a desorption process. The presence and position of
pre-waves between −1.2 and −1.5 V vary both with scan rate and
with successive scans, leaving a single consistent reduction process
(at Epc −1.75 V for 10 V s−1 scan rate), although backwaves on
the anodic sweep occur at −1.33 and −1.13 V. Adsorption and
desorption phenomena clearly complicate the voltammograms,
but are less obtrusive than similar affects that occur for the free
ligand. The reduction process is shifted ∼0.3 V cathodically on
complexation (compared to the free ligand), so that, overall, both
oxidation and reduction processes display cathodic shifts as a
consequence of deprotonation and complexation.
Metal Ion
Process
Oxidations
Nil
Zn(II)
Cu(II)
Ni(II)
Co(II)
Fe(III)
Reductions
Nil
Zn(II)
Cu(II)
Ni(II)
∼+2.1 (irrev.)
+1.43 (irrev.)
+1.37 (irrev.), +1.55 (sh, irrev.)
+1.29 (irrev.)
+1.31 (irrev.)
+1.36 (irrev.)
−1.33 (irrev.), −1.50 (irrev.)
−1.61 (irrev)b
−1.19 (irrev.), −2.06 (sh, irrev.)
−1.84 (irrev.), −2.12 (irrev.)
−1.64 (quasirev.; DE 230 mV)c
Co(II)
Fe(III)
−0.65 (irrev.), −1.48 (irrev.), −1.77 (irrev.)
a Glassy carbon electrode; 0.1 M tetraethylammonium perchlorate in
acetonitrile or mixed acetonitrile/tetrahydrofuran solvent. b Pre-waves at
−1.16, −1.31 V; some reversible character at high scan rates. c Additional
backwave at −0.25 V, indicative of competing dissociation.
potential, provided the integrity of the complex is maintained to
a reasonable extent. However, if fragmentation and dissociation
occur rapidly following the first single electron reduction, release
of the remaining metal ions may lead to a significant alteration
of their redox properties, to the point where these other centres
either immediately undergo redox processes or become redox
inert in the electrochemical window under examination. Complex
electrochemical behaviour is hence likely to result from such
sequential processes. As for our previous studies involving the
related dinuclear complexes of H2L2 (R = t-Bu),19 the present
studies were in part undertaken to probe the validity of the above
postulates.
[Co3(L1)3(Py)6] (R
=
t-Bu). The simple [Co(2,4-penta-
nedionato)2(OH2)2] complex displays an irreversible reduction at
−1.79 V, assigned to CoII/I, while an irreversible one-electron
oxidation at +1.07 V is possibly a CoIII/II couple. A further irre-
versible multi-electron oxidation appears at +1.99 V, accompanied
by a broad pre-wave near +1.6 V; this multi-electron process is
almost certainly ligand centred. The related monomer, [Co(2,4-
pentanedionato)3], has been reported to undergo irreversible
reduction at −1.1 V and irreversible oxidation at +1.8 V.39
The present trinuclear cobalt(II) complex exhibits a poorly
irreversible oxidation at +1.31 V. This is similar in position to the
CoIII/II oxidation observed for [Co(2,4-pentanedionato)2(OH2)2],
and can be assigned to the same metal-centred process. The
cobalt(II) complex also exhibits a quasi-reversible wave at −1.64 V
(DE = 230 mV at 100 mV s−1), with no further reductions
until below −2 V. Competing dissociative processes are clearly
visible; an additional backwave, assignable to free solvated cobalt
ion, occurs at −0.25 V. A tied reduction wave for this wave on
subsequent sweeps is consistent with it being a metal-centred
process. The appearance of these waves supports electron transfer
into a metal centre, formally forming cobalt(I), which usually
displays limited stability. Normally, diketonic cobalt complexes
display CoIII/II couples near 0 V,40 so the current process at a much
more negative potential is almost certainly associated formally
with a CoII/I couple.
Free ligand. H2L1 (R = t-Bu) exhibits no oxidation processes
before +2 V (versus Ag/AgCl), and shows only an irreversible
process, presumably ligand oxidation, near the system limit of
∼+2.2 V in acetonitrile. The reduction behaviour in acetonitrile
is more complex, displaying reductions that are shifted anodically
compared with those for H2L2 (R = t-Bu),19 with waves at −1.33
and −1.50 V. Even at 100 mV s−1 there is a backwave at −1.455 V,
with another at −1.33 V observed at higher speed. These reduction
processes are affected (at both glassy carbon and platinum
working electrodes) by strong adsorption spikes coincident with
the positions of the cathodic and anodic waves near −1.3 V. This
suggests that the anodic wave, whose peak current varies linearly
with respect to increasing scan rate, corresponds predominantly
to a desorption phenomenon and not to an electron transfer
process. The observed behaviour does not change significantly
with working electrode, dilution or addition of other aprotic
solvents (such as tetrahydrofuran). In particular, deprotonation
of the ligand through addition of sodium methoxide causes a
clear cathodic shift in the oxidation behaviour, with a poorly
resolved shoulder appearing at ∼+1.8 V. The cathodic processes
are not shifted, remaining effectively the same (Epc at −1.52
V) as occurs for the protonated ligand. The observed shifts
presumably reflect both the influence of deprotonation and the
effect of weak coordination to the introduced sodium cation.
Cyclic voltammograms for H2L1 (R = t-Bu) together with those
[Ni3(L1)3(Py)6] (R = t-Bu). The related nickel(II) complex,
[Ni2(L2)2(Py)4] (R = t-Bu), displays a quasi-reversible oxidation
at +1.04 V (DE = 155 mV at 100 mV s−1 scan rate).19 The
behaviour and position of this process characterises it as a metal-
centred NiIII/II process. For the present complex, the oxidation
wave is shifted to higher potential (Epa = +1.29 V) and is
irreversible, possibly reflecting the structural differences between
the complexes of the two ligand isomers and implying lower
stability of the reduced trimetallic cluster. It exhibits an irreversible
1336 | Dalton Trans., 2008, 1331–1340
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
reduction near −1.84 V, somewhat comparable to that reported
for [Ni(2,4-pentanedionato)2(OH2)2] of −1.70 V.41 Shifts should
arise as a result of electronic effects of the tert-butyl and phenyl
substituents in the present cases, compared with two methyl groups
in 2,4-pentanedionato itself. Notably, a dimeric 1,3,5-triketonate
complex has been reported to show two nearly reversible processes
at −1.6 and −1.90 V,41 assigned to two successive metal-centred
processes with a comproportionation constant of ∼105. A second
process is observed for [Ni3(L1)3(Py)6] (R = t-Bu) at −2.12 V,
although this process is irreversible. Assignment as sequential
metal-centred processes cannot be made with certainty, given the
redox behaviour of the ligand itself in this region. Reduction
processes are not substantially shifted from those found for the
zinc(II) complexes, where the metal is electro-inactive, so there is
a good case for suggesting that these processes at very negative
potentials formally involve ligand-centred molecular orbitals.
4-ethylpyridine.19 In this context, it was of interest to explore the
comparative behaviour of similar ligands of type H2L1 given the
difference in the directions of the coordination vectors in these
two ligand types.
Studies at constant metal and ligand concentration. The ex-
traction of cobalt(II) and zinc(II) using the H2L1 ligands with
R = Pr, Ph, t-Bu, hexyl, octyl and nonyl was performed using
the previously described radiotracer technique.25 Parallel sets of
extractions were carried out both in the absence and presence of 4-
ethylpyridine and the results are summarised in Fig. 4. At pH 8.7 in
the absence of 4-ethylpyridine, cobalt(II) extraction is negligible (1
to 2%) for each member of the above ligand series, while the range
towards zinc(II) is ∼2 to 35%. As expected, the more lipophilic
ligand derivatives give rise to enhanced extraction.
[Cu3(L1)3] (R = t-Bu). The oxidative behaviour is not simple,
exhibiting an irreversible oxidation at +1.37 V and a shoulder
at +1.55 V. Although it is tempting to assign these to successive
CuIII/II oxidations of separate copper centres in the structure, it
is notable that the simple [Cu(2,4-pentanedionato)2] complex also
displays a suite of oxidations, although only one wave (at +1.48 V)
lies near the potential of the [Cu3(L1)3] (R = t-Bu) system and may
be the copper-centred process in that case. The trinuclear complex
displays an irreversible wave at −1.19 V, similar in position to the
wave seen for the simple 2,4-pentanedionato monomer complex.
This suggests that these are metal-centred processes, and are
assigned to CuII/I couples. Other irreversible redox processes are
seen below −1.6 V, possibly associated with sequential metal
reduction concomitant with complex fragmentation.
Fig. 4 Percentage of metal ion extracted by H2L1 (R = Pr, t-Bu, Ph, hexyl,
octyl and nonyl), and on the addition of 4-ethylpyridine (*). [M(ClO4)2] =
1 × 10−4 M, pH = 8.7 (TAPS/NaOH buffer); [H2L1] = 1 × 10−3 M,
[4-ethylpyridine] = 2 × 10−3 M in CHCl3 (*); shaking time 30 min for
zinc(II), 3 h for cobalt(II) or 1.5 h (when 4-ethylpyridine was added); T =
24 1 ◦C. No useful data were obtained for zinc(II) with H2L1 (R = Pr and
Ph) in the presence of 4-ethylpyridine due to precipitation at the respective
aqueous/organic interfaces, possibly due to the formation of polymeric
species similar to that reported by Soldatov et al.20 on complexation of this
ion with H2L (R = Ph) and pyridine.
[Fe4(L1)6] (R = t-Bu). The iron complex displays the most
definitive example of a metal-centred redox process in its reduction
behaviour, with a process well removed from the region of any
free ligand reduction steps being present (irreversible reduction,
Epc = −0.65 V; only at high scan rate (>2 V s−1) is a backwave
observable for this complex). The position and nature of this
process strongly supports a metal-centred process; the couples
occur at very similar potentials to the metal-centred couple in
the simple [Fe(2,4-pentanedionato)3] complex, strengthening this
assignment. Additional irreversible waves occur well below −1 V
(at −1.48, −1.77 V). Whether these are associated with other
metal centres or the ligand itself is difficult to define; however,
what is clear is that the cluster does not retain its integrity
when undergoing further reduction processes. This is consistent
behaviour across all the polymetallic compounds examined in the
present study.
The above experiments were repeated with 4-ethylpyridine
(2 × 10−3 M) added to the organic phase but under otherwise
identical experimental conditions; the results are also summarised
in Fig. 4. On addition of two equivalents (relative to H2L1) of 4-
ethylpyridine, the extraction of cobalt(II) was markedly enhanced
(from ∼2 to 26%) in the case of H2L1 (R = nonyl) and also
enhanced from 35 to 61% for zinc(II) with this ligand. Further,
the time for equilibrium to be reached in the case of cobalt(II)
was also approximately halved (from 3 to 1.5 h). Parallel ‘control’
experiments in the absence H2L1 but at the same concentration
of 4-ethylpyridine were performed; under these conditions metal
extraction was negligible at ≤1%. Overall, it is noted that
synergism is more pronounced for the nonyl derivative, especially
in the case of cobalt(II).
Both in the presence and absence of 4-ethylpyridine, these sys-
tems show no significant extraction of metal below approximately
pH 6 and thus stripping may be achieved by pH adjustment of the
aqueous phase to pH <6.
Clearly the enhancement of extraction in the presence of 4-
ethylpyridine is substantial for both zinc(II) and cobalt(II) for
H2L1 (R = nonyl). It is of interest to compare the results for
Liquid–liquid extraction studies
Simple b-diketones have long been used as extractants in nu-
merous liquid–liquid extraction studies (including, for example,
for mineral processing applications).42 In contrast, there appears,
however, to be no report of the use of bis-b-diketone ligands
of the present type for similar studies even though individual
examples of the latter have also been known for a considerable
time.5,34 Nevertheless we have recently demonstrated that 1,3-
phenylene bridged ligands of type H2L2 give rise to significant
synergistic metal extraction behaviour in the presence of added
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1331–1340 | 1337
©

View Article Online
H2L2 (R = nonyl) under identical experimental conditions: this
ligand gave an increase of 12 to 70% for cobalt(II) while the increase
for zinc was 53 to 85%.43
undertaken as part of the present study. This method is based
on a defined complex being generated in the organic phase
under ‘saturation’ conditions—corresponding to a constant ligand
concentration in the presence of excess metal ion.26 H2L1 (R =
nonyl) at a concentration of 1 × 10−3 M was employed in the
chloroform phase. In the case of zinc the [metal]/[ligand] ratio
of the extracted species was probed by incrementally increasing
H2L1 (R = nonyl) dependence studies. The stoichiometries of
the extracted species were analysed by undertaking the extraction
of cobalt(II) and zinc(II) (at a fixed concentration of 1 × 10−4 M)
under variable H2L1 (R = nonyl) concentration in the presence of
a fixed concentration (2 × 10−3 M) of 4-ethylpyridine as well as in
its absence. For the extractions involving zinc(II) and H2L1 (R =
t-Bu, hexyl, octyl, nonyl) in all cases the concentration of the H2L1
was varied from 5 × 10−4 to 1.5 × 10−3 M (namely, it was always
in excess relative to the metal ion concentration in an attempt to
approximate constant ionic strength conditions).25 The negligible
extraction observed for cobalt(II) in the absence of 4-ethylpyridine
(Fig. 4) across all H2L1 ligands prevented investigation of this
system over the above concentration range.
the zinc concentration in the aqueous phase from 1 × 10−4
M
to 5 × 10−3 M (at pH 8.7) while the degree of zinc extracted
into the chloroform phase was monitored. The inflection in the
corresponding plot obtained from this experiment provides an
indication of the maximum zinc to ligand ratio for complex
formation. The conditions employed for these studies are thus
somewhat different to those used for the extraction experiments
discussed above.
The results are summarised in Table 3. Under the conditions
employed, an equimolar ratio (metal : ligand = 1 : 1) is indicated
for zinc(II) in the absence of added 4-ethylpyridine. This 1 : 1
ratio is comparable to that present in the complexes formulated
For each of the above experiments, log DM (DM
=
[Mn+](org)/[Mn+](aq)) was plotted against log [H2L1]. Provided a
‘simple’ equilibrium is involved, the slope of this plot gives the
as [M3L1 Py6] (M = Ni, Co or Zn; R = t-Bu) (and confirmed for
3
stoichiometry (n) of the extracted species directly since log DM
=
nickel by X-ray diffraction) sythesised in the present study and also
those reported17 for the complexes of H2L1 with R = Ph which,
as mentioned previously, were shown to be M3L3 triangles when
cobalt(II) and nickel(II) were employed (while zinc(II) yielded a
polymeric species in the solid state).
n log[L]org + K (where K is a constant).25,44 The results for both
cobalt(II) and zinc(II) with H2L1 at pH 8.7 in both the absence and
presence of 4-ethylpyridine show that linear relationships occur
between log DM and the log of the ligand concentration in each
case. For the present ligand series, slopes that range from 1.8 to 2.1
were obtained (Table 2), suggesting that predominately 1 : 2 (M :
L) species (unlike H2L2, for which case a mixture of species was
postulated to form) are extracted under the conditions employed.
Further variable metal concentration loading experiments were
undertaken in which the solution stoichiometries of the cobalt(II)
and zinc(II) systems with H2L1 (R = nonyl) were probed under
similar conditions, but this time in the presence of a fixed
concentration of 4-ethylpyridine of 2 × 10−3 M. The presence
of 4-ethylpyridine causes a major change in the species being
extracted, leading to ratios that approximate the formation of ML2
(M = Co, Zn) derivatives. From these experiments, two types of
extracted complexes can be proposed to form under the saturation
conditions discussed above. In the absence of 4-ethylpyridine only
a species with a M : L ratio of 1 : 1 is predicted (most likely of type
Zn3(L1)3) while in the presence of 4-ethylpyridine the dominant
complex appears to be a 1 : 2 (M : L) species where M = Zn
or Co (most likely of type [M(L1)2(4-ethylpyridine)]). These latter
results appear not inconsistent with those obtained from the ligand
4-Ethylpyridine dependence studies. These involved variation
of the 4-ethylpyridine concentration over the range 2 × 10−4
M
to 2 × 10−3 M while the other concentrations were held constant.
In the case of H2L1 (R = nonyl) with both zinc(II) and cobalt(II),
a metal : 4-ethylpyridine ratio of approximately 0.8 was obtained
(Table 2). This suggests that, under the conditions employed, the
predominant species extracted approximates one 4-ethylpyridine
coordinated per each metal centre.
Variable metal concentration studies. A further procedure for
probing the stoichiometry of the extracted species has been
Table 2 Gradients (see text) of the plots obtained for the variable H2L1 and 4-ethylpyridine (EtPy) concentration extraction experiments (water–
chloroform)
Slopes
Run
Ligand
(Zn(II),H2L1)
(Zn(II),H2L1) + EtPy
(Co(II),H2L1) + EtPy
H2L1 dependence
R = t-Bu
1.8
2.0
2.1
2.0
N/A
N/A
N/A
1.9
N/A
N/A
N/A
1.9
R = Hexyl
R = Octyl
R = Nonyl
R = Nonyl
4-Ethylpyridine dependence
N/A
0.8
0.8
Table 3 Maximum metal : ligand ratios for the organic phases obtained from loading experiments involving a fixed concentration of H2L1 (R = nonyl)
and varying metal ion concentration (see text)
Metal : ligand ratio
Ligand H2L1
(Zn(II),H2L1)
(Zn(II),H2L1) + EtPy)
(Co(II),H2L1) + EtPy
0.6
R = Nonyl
1.0
0.6
1338 | Dalton Trans., 2008, 1331–1340
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
dependence studies where under conditions of ligand excess 1 : 2
(M : L) species were proposed to occur in both the presence and
absence of added 4-ethylpyridine.
for 24 h before the metal ion concentrations were determined by
ICP-MS. Selectivity for copper(II) over the other four divalent
transition metal ions present was obtained. Within experimental
error, quantitative (100%) extraction of copper ion occurred in
each case, with zero extraction occurring for the remaining four
metals.
Similar competitive extraction experiments were carried out
in the presence of 4-ethylpyridine (2 × 10−3 M) at pH 7.4 but
otherwise the conditions were identical to those used in the
above studies. As above, the results show effectively quantitative
extraction of copper(II) across all ligand derivatives. In contrast,
extraction of cobalt(II), nickel(II), zinc(II) and cadmium(II) was
found to be zero (within experimental error) in all cases apart
from nickel(II) with H2L1 (R = nonyl). In this case 63% extraction
of nickel(II) was observed. Thus, the additional presence of 4-
ethylpyridine in this experiment results in a marked synergistic
enhancement of the extraction of nickel(II) (from ∼0 to 63%) under
the conditions employed.
Summary of complex species extracted. Overall, on the basis
of the solvent extraction experiments three complex species may
be proposed to form in the respective organic phases (Scheme 1),
with the particular species generated depending on the conditions
employed. In the absence of 4-ethylpyridine both 1 : 1 and 1 : 2
(M : L) metal to ligand ratios were indicated, with the 1 : 1 species
dominating under higher metal (zinc) concentrations. For the case
where the ligand is in clear excess with respect to the metal ion
(cobalt or zinc), the formation of the corresponding 1 : 2 (M : L)
complex is promoted. In the presence of 4-ethylpyridine a 1 : 2 : 1
(M : L : EtPy) complex is proposed to be the dominant extracted
species in both the cobalt and zinc cases.
Concluding remarks
In the present study we present the synthesis of three new
lipophilic ligands of the 1,4-phenylene-linked, bis-b-diketone
category (H2L1) as well as new solid complexes of cobalt(II),
nickel(II) and zinc(II) of H2L1 with R = t-Bu. Electrochemical and
liquid–liquid extraction studies for individual systems have been
undertaken with, in general, the results paralleling those reported
previously for the related complexes of the 1,3-phenylene-linked
ligands of type H2L2.19
It is clear that the present compounds are associated with a
rich electrochemistry; while many of the observed processes for
the complexes appear to be ligand centred, the occurrence of
peaks removed from the expected positions for normal ligand
processes is in accord with dominantly metal-centred behaviour
also occurring. While the metal centres in the tri- and tetrameric
systems appear sufficiently separate to prevent significant metal
coupling, the distances are still of such a magnitude for a given
metal centre to feel the effect of neighbouring ‘spectator’ metal
ions. As a consequence processes at different potentials might be
expected but these were not able to be definitively assigned because
of interfering adsorption/desorption processes and complications
due to the presence of an electroactive ligand.
The solvent extraction studies collectively show that the sto-
ichiometries of the extracted species clearly depend on both
the relative concentrations of the metal and ligand species as
well as on their respective natures. As expected, enhanced metal
extraction efficiencies in general occur for the more lipophilic
ligand derivatives. An interesting feature of the results obtained
is the substantial extraction synergism obtained on addition
of 4-ethylpyridine to the respective lipophilic ligand/metal ion
systems. As also found for related H2L1 systems, competitive metal
ion studies indicate a clear preference for copper(II) extraction
when equal concentrations of cobalt(II), nickel(II), zinc(II) and
cadmium(II) were also present in the corresponding aqueous
phase.
Scheme 1 Proposed major complex species formed in the respective
organic phases: (a) on metal ion saturation of the organic phase in the
presence of a fixed concentration of H2L1; (b) with excess ligand present in
the organic phase; and (c) in the presence of EtPy when either an excess of
ligand is present or metal ion saturation has occurred in the organic phase.
Competitive extractions. In order to probe the potential metal
ion selectivities of these ligands, competitive extraction experi-
ments were carried out at pH 7.8 (HEPES/NaOH) in the absence
of 4-ethylpyridine for the ligands H2L1 (R = t-Bu, hexyl, octyl and
nonyl). Equal concentrations of five different divalent transition
metal perchlorate salts (cobalt(II), nickel(II), copper(II), zinc(II),
cadmium(II), each 1 × 10−4 M) were added to the aqueous phase
and H2L1 (R = t-Bu, hexyl, octyl or nonyl) at 1 × 10−3 M were
added to the organic phase. The phases were combined and shaken
Acknowledgements
We thank the Australian Research Council and the Deutsche
Forschungsgemeinschaft for support. We thank Dr Paul Jensen
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1331–1340 | 1339
©

View Article Online
for assistance with and advice regarding the crystallographic
refinement.
15 J. K. Clegg, L. F. Lindoy, J. C. McMurtrie and D. Schilter, Dalton
Trans., 2005, 857.
16 J. K. Clegg, L. F. Lindoy, J. C. McMurtrie and D. Schilter, Dalton
Trans., 2006, 3114.
17 J. K. Clegg, K. Gloe, M. J. Hayter, O. Kataeva, L. F. Lindoy, B.
Moubaraki, J. C. McMurtrie, K. S. Murray and D. Schilter, Dalton
Trans., 2006, 3977.
References
1 A. Werner, Ber. Dtsch. Chem. Ges., 1901, 34, 2594.
2 A. G. Swallow and M. R. Truter, Proc. R. Soc. London, Ser. A, 1960,
254, 205; D. Gibson, Coord. Chem. Rev., 1969, 4, 225; F. P. Dwyer and
D. P. Mellor, Chelating Agents and Metal Chelates, Academic Press,
London, 1964; U. Casellato, P. Vigato and M. Vivaldi, Coord. Chem.
Rev., 1977, 23, 31; U. Casellato, Chem. Soc. Rev., 1979, 199.
18 J. K. Clegg, Aust. J. Chem., 2006, 59, 660.
19 J. K. Clegg, D. J. Bray, K. Gloe, K. Gloe, M. J. Hayter, K. A. Jolliffe,
G. A. Lawrance, G. V. Meehan, J. C. McMurtrie, L. F. Lindoy and M.
Wenzel, Dalton Trans., 2007, 1710.
20 D. V. Soldatov, A. S. Zanina, G. D. Enright, C. I. Ratcliffe and J. A.
Ripmeester, Cryst. Growth Des., 2003, 3, 1005.
21 D. V. Soldatov, I. E. Sokolov and K. Sowin’ska, J. Struct. Chem., 2005,
46, S158.
22 V. A. Grillo, E. J. Seddon, C. M. Grant, G. Aromi, J. C. Bollinger,
K. Folting and G. Christou, Chem. Commun., 1997, 1561; M. M.
Matsushita, T. Yasuda, R. Kawano, T. Kawai and T. Iyoda, Chem.
Lett., 2000, 812.
3 See, for example: D. V. Soldatov, J. Chem. Crystallogr., 2006, 36, 747;
D. J. Bray, J. K. Clegg, L. F. Lindoy and D. Schilter, Adv. Inorg. Chem.,
2006, 59, 1; T.-J. Won, J. K. Clegg, L. F. Lindoy and J. C. McMurtrie,
Cryst. Growth Des., 2007, 7, 972 and ref. therein.
4 G. Aromi, P. Gamez and J. Reedijk, Coord. Chem. Rev., 2007,
DOI: 10.1016/j.ccr.2007.07.008.
5 D. E. Fenton, C. M. Regan, U. Casellato, P. Alessandro, Vigato and M.
Vidali, Inorg. Chim. Acta, 1982, 58, 83.
23 See, for example: O. Hallale, S. A. Bourne and K. R. Koch, CrystEng-
Comm, 2005, 7, 161 and ref. therein.
6 See for example: A. W. Maverick and F. E. Klavetter, Inorg. Chem.,
1984, 23, 4130; A. W. Maverick, S. C. Buckingham, Q. Yao, J. R.
Bradbury and G. G. Stanley, J. Am. Chem. Soc., 1986, 108, 7430; J. R.
Bradbury, J. L. Hampton, D. P. Martone and A. W. Maverick, Inorg.
Chem., 1989, 28, 2392; R. W. Saalfrank, A. Dresel, V. Seitz, S. Trummer,
F. Hampel, M. Teichert, D. Stalke, C. Stadler, J. Daub, V. Schu¨nemann
and A. X. Trautwein, Chem.–Eur. J., 1997, 3, 2058; R. W. Saalfrank,
V. Seitz, D. L. Caulder, K. N. Raymond, M. Teichert and D. Stalke,
Eur. J. Inorg. Chem., 1998, 1313; A. W. Maverick, D. R. Billodeaux,
M. L. Ivie, F. R. Fronczek and E. F. Maverick, J. Inclusion Phenom.
Macrocyclic Chem., 2001, 39, 19; M. Albrecht, S. Schmid, M. deGroot,
P. Weis and R. Fro¨hlich, Chem. Commun., 2003, 2526; G. Arom´ı, C.
Boldron, P. Gaez, O. Roubeau, H. Kooijman, A. L. Spek, H. Stoeckli-
Evans, J. Ribasa and J. Reedijk, Dalton Trans., 2004, 3586; G. Arom´ı,
J. Ribas, P. Gamez, O. Roubeau, H. Kooijman, A. L. Spek, S. Teat,
E. MacLean, H. Stoeckli-Evans and J. Reedijk, Chem.–Eur. J., 2004,
10, 6476; Bai-Shu Zheng, Xiao-Yi Zhang, Huai-Wu Zhu, Shi-Xia Luo,
L. F. Lindoy, J. C. McMurtrie, P. Turner and G. Wei, Dalton Trans.,
2005, 1349; R. W. Saalfrrank, C. Schmidt, H. Maid, F. Hampel, W.
Bauer and A. Scheuer, Angew. Chem., Int. Ed., 2006, 45, 315; X. Zhang,
H. Chen, C. Ma, C. Chen and Q. Liu, Dalton Trans., 2006, 4047; M.
Albrecht, S. Dehn, S. Schmid and M. DeGroot, Synthesis, 2007, 155;
24 K. Gloe and P. Mu¨hl, Isotopenpraxis, 1983, 19, 257.
25 H. Stephan, S. Juran, B. Antonioli, K. Gloe and K. Gloe, in Analytical
Methods in Supramolecular Chemistry, ed. Ch. Schalley, Wiley-VCH,
Weinheim, 2006, pp. 79–103.
26 Y. Marcus, A. S. Kertes, Ion Exchange and Solvent Extraction of Metal
Complexes, Wiley, London, 1969, pp. 480–487.
27 SMART, SAINT and XPREP, Bruker Analytical X-ray Instruments
Inc., Madison, Wisconsin, USA, 1995.
28 APEX v2.1, SAINT v.7 and XPREP v.6.14, Bruker AXS Inc., Madison,
Wisconsin, USA, 2003.
29 L. J. Farrugia, WinGX-32: System of programs for solving, refining
and analysing single crystal X-ray diffraction data for small molecules,
J. Appl. Crystallogr., 1999, 32, 837.
30 A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano, C. Giocavazzo,
A. Guagliardi, A. G. C. Moliterni, G. Polidori and S. Spagna, J. Appl.
Crystallogr., 1999, 32, 115.
31 G. M. Sheldrick, SADABS: Empirical Absorption and Correction
Software, University of Go¨ttingen, Germany, 1999–2003.
32 G. M. Sheldrick, SHELXL-97: Programs for Crystal Structure Analysis,
University of Go¨ttingen, Germany, 1997.
33 L. J. Farrugia, J. Appl. Crystallogr., 1999, 30, 565.
34 D. Martin, M. Shamma and W. Fernelius, J. Am. Chem. Soc., 1958, 80,
4891.
35 D. J. Gosciniak, N. J. Patrick, Ultraviolet radiation absorbing com-
positions of bis-1,3-diketone derivatives of benzene, European Patent
Application, EP 0 376 511 A2, 1990.
36 R. S. Rasmussen, D. D. Tunnicliff and R. R. Brittain, J. Am. Chem.
Soc., 1949, 71, 1068; C. Duval, R. Freymann and J. Lecompte, Bull.
Soc. Chim. Fr., 1952, 106.
37 L. L. Funck and T. R. Ortolano, Inorg. Chem., 1968, 7, 567.
38 J. L. Burdett and M. T. Rogers, J. Am. Chem. Soc., 1964, 86, 2105.
39 J. H. Tocher and J. P. Fackler, Inorg. Chim. Acta, 1985, 102, 211.
40 J. Carr, D. E. Smith and M. Cais, J. Am. Chem. Soc., 1985, 107, 6807.
41 R. L. Lintvedt, G. Ranger and L. S. Kramer, Inorg. Chem., 1986, 25,
2635.
42 P. A. Tasker, P. G. Plieger and L. C. West, in Comprehensive Coordi-
nation Chemistry II, ed. J. A. McCleverty and T. J. Meyer, Elsevier,
Oxford, 2004, vol. 9, pp. 759–808; Solvent Extraction Principles and
Practice, ed. J. Rydberg, M. Cox, C. Musikas and G. R. Choppin,
Marcel Dekker, New York, 2004.
¯
T. Shiga, N. Ito, A. Hidaka, H. Okawa, S. Kitagawa and M. Ohba,
Inorg. Chem., 2007, 46, 3492; R. W. Saalfrank, A. Scheurer, R. Puchta,
F. Hampel, H. Maid and F. W. Heinemann, Angew. Chem., Int. Ed.,
2007, 46, 265; G. Aromi, P. Gamez, J. Krzystek, H. Kooijman, A. L.
Spek, E. J. MacLean, S. J. Teat and H. Nowell, Inorg. Chem., 2007, 46,
2579; C. Pariya, C. R. Sparrow, C.-K. Back, G. Sandi, F. R. Fronczek
and A. W. Maverick, Angew. Chem., 2007, 119, 6421.
7 G. Aromi, P. Gamez, O. Roubeau, P. C. Berzal, H. Kooijman, A. L.
Spek, W. L. Driessen and J. Reedijk, Inorg. Chem., 2002, 41, 3673; G.
Aromi, P. Gamez, P. C. Berzal, W. L. Driessen and J. Reedijk, Synth.
Commun., 2003, 33, 11.
8 G. Arom´ı, H. Stoeckli-Evans, S. J. Teat, J. Cano and J. Ribas, J. Mater.
Chem., 2006, 16, 2635.
9 A. P. Basset, S. W. Magennis, P. B. Glover, D. J. Lewis, N. Spencer, S.
Parsons, R. M. Williams, L. De Cola and Z. Pikramenou, J. Am. Chem.
Soc., 2004, 126, 9413.
10 S. N. Podyachev, I. A. Litvinov, A. R. Mustafina, R. R. Shagidullin,
W. D. Habicher and A. I. Konovalov, Russ. Chem. Bull., 2005, 54, 623.
11 G. Aromi, P. Gamez, C. Boldron, H. Kooijman, A. L. Spek and J.
Reedijk, Eur. J. Inorg. Chem., 2006, 1940.
12 D. J. Bray, K. A. Jolliffe, L. F. Lindoy and J. C. McMurtrie, Tetrahedron,
2007, 63, 1953.
43 J. K. Clegg, D. J. Bray, K. Gloe, K. Gloe, M. J. Hayter, K. A. Jolliffe,
G. A. Lawrance, G. V. Meehan, J. C. McMurtrie, L. F. Lindoy and M.
Wenzel, Dalton Trans., 2007, 1719.
13 R. W. Saalfrank, N. Low, B. Demleitner, D. Stalke and M. Teichert,
Chem.–Eur. J., 1998, 4, 1305 and ref. therein; R. W. Saalfrank, V. Seitz,
F. W. Heinemann, C. Go¨bel and R. Herbst-Irmer, J. Chem. Soc., Dalton
Trans., 2001, 599.
14 J. K. Clegg, L. F. Lindoy, B. Moubaraki, K. S. Murray and J. C.
McMurtrie, Dalton Trans., 2004, 2417.
44 J. Rydberg, G. R. Choppin, C. Musikas and T. Sekine, Solvent
Extraction Equilibria, in Solvent Extraction Principles and Practice, ed.
J. Rydberg, M. Cox, C. Musikas and G. R. Choppin, Marcel Dekker,
New York, 2004, pp. 109–201; J. R. Price, M. Fainerman-Melnikova,
R. R. Fenton, K. Gloe, L. F. Lindoy, T. Rambusch, W. Skelton, P.
Turner, A. H. White and K. Wichmann, Dalton Trans., 2004, 3715.
1340 | Dalton Trans., 2008, 1331–1340
This journal is
The Royal Society of Chemistry 2008
©