Polyacidic multiloading metal extractants

Article abstract of DOI:10.1039/b810610j

Novel polynucleating, di- and tri-acidic ligands have been designed to increase the molar and mass transport efficiencies for the recovery of base metals by solvent extraction. The Royal Society of Chemistry.

Full text of DOI:10.1039/b810610j

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Polyacidic multiloading metal extractantsw
Ross J. Gordon,a John Campbell,b David K. Henderson,a Dorothy C. R. Henry,a
Ronald M. Swart,b Peter A. Tasker,*a Fraser J. White,a Jenny L. Wooda and
Lesley J. Yellowleesa
Received (in Cambridge, UK) 23rd June 2008, Accepted 9th July 2008
First published as an Advance Article on the web 22nd August 2008
DOI: 10.1039/b810610j
Novel polynucleating, di- and tri-acidic ligands have been
designed to increase the molar and mass transport efficiencies
for the recovery of base metals by solvent extraction.
Solvent extraction provides highly efficient concentration and
separation operations in extractive hydrometallurgy.1 Major new
plants to recover zinc have opened recently2 and ca. 20% of the
global production of copper is achieved using phenolic oxime
extractants3 (e.g. L1H, Fig. 1) in an acid leach–solvent
extraction–electrowinning flowsheet (Scheme 1). The latter gives
an excellent materials balance for recovery from oxidic ores
because the leach, extractant and electrolyte solutions are recycled,
and the process can be adapted to work efficiently with wide
variation of compositions of pregnant leach solutions.1,4
Scheme 1 Materials balance for the recovery of copper from oxidic
ores.
Whilst the flowsheet outlined in Scheme 1 is very efficient in
terms of the overall materials balance, there is scope to
improve the throughput by increasing the transport efficiency
in the solvent extraction step. The mass of copper transferred
to the water-immiscible solvent in the extraction stage is
limited by the 1 : 2 stoichiometry required to generate a
neutral, organic soluble, complex from the monoanionic phe-
nolic oxime reagents. We recognized that the molar ratio of
copper to extractant could be improved if polynucleating
ligands capable of existing as di- or tri-anions after deprotona-
tion were designed. Some systems developed to achieve this are
shown in Fig. 1. A key design feature is that the anionic donor
atoms, Xꢀ or Wꢀ in Fig. 1, are able to bridge two Cu(II) atoms
in planar complexes.
Fig. 1 Functionalization of the conventional type of extractant L1H
to generate di- and tri-acidic reagents L2H2–L6H3 which can form
di- and tri-nuclear Cu(II) complexes (B and C).
a School of Chemistry, The University of Edinburgh, Edinburgh, UK
EH9 3JJ. E-mail: Peter.Tasker@ed.ac.uk; Fax: +44-131-6506453;
Tel: +44-131-6504706
Studies of the pH-dependence of Cu(II) extraction into
chloroform solutions using the ligands (Fig. 2) show that the
metal to ligand stoichiometry can be increased from 1 : 2 to
2 : 2 or 3 : 2, corresponding to 100, 200 and 300% Cu-loading
for ligands L1H, L2H2 to L4H2 or L5H3, respectively. The
loading curves (Fig. 2) for L3H2 and L5H3 show plateaux
which indicate that stepwise loading is possible, e.g. in
b Cytec Industries Inc, PO Box 42, Blackley, Manchester,
UK M9 8ZS
w Electronic supplementary information (ESI) available: Full experi-
mental details. CCDC reference numbers 692604 and 692605. For ESI
and crystallographic data in CIF or other electronic format, see DOI:
10.1039/b810610j
ꢁc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 4801–4803 | 4801

View Article Online
Fig. 2 Cu-loading by CHCl3 solutions of L1H ( ), L2H2 ( ), L3H2
), L4H2 (&) and L5H3
) as a function of pH. Loadings of 100,
(
(
Fig. 4 Dinuclear copper complex of a structural analogue of L2H2.
200 and 300% correspond to Cu : 2L, 2Cu : 2L and 3Cu : 2L molar
ratios.
Fig. 3) and at 2 : 2 stoichiometry there is almost no signal.
Antiferromagnetic coupling between the two copper centres in
the dinuclear complexes could account for these changes.
There is some ambiguity as to the structure of the dinuclear
complexes, and in particular which of the oxygen atoms in
addition to forming species with a stoichiometry consistent
with the desired trinuclear complex, [Cu3(L5)2], at pH 4 5 the
ligand L5H3 also forms species corresponding to mono- or
dinuclear complexes, [Cu(L5H2)2] and [Cu2(L5H)2], at pH
values of ca. 2.2 or 3.8, respectively.
The Cu-uptake by chloroform solutions of L2–4H2 can be
followed by monitoring X-band EPR spectra. The intensity of
the signal increases (A in Fig. 3) until the Cu-loading corre-
sponds to a metal to ligand stoichiometry of 1 : 2. Further
loading is accompanied by a decrease in signal intensity (B in
L
2–4H2 or L5–6H3 form bridges between the two Cu atoms. An
analogue of L2H2, with t-Bu and Me groups replacing the
heptyl and nonyl substituents, gives a complex (Fig. 4)z which
is similar to several reported in the CSD5 in which the
phenolate oxygen atom of the salicylaldimine component
bridges the two copper centres. In contrast, in the dinuclear
complex (Fig. 5)z formed by L6H3, it is another type of
phenolate oxygen atom which forms the bridges and much
2ꢀ
more irregular NO3 donor sets are presented to the copper
atoms by the ligands.
Whilst the new di- and tri-acidic extractants increase the
molar transport efficiency two or three fold, it is the mass
transport which is of greater practical significance. Based on
the molecular weights shown in Table 1, it can be seen that
Fig. 3 EPR spectra of CHCl3 solutions of L2H2 as Cu-loading
increases from 0–100% (curves A) and from 100–200% (curves B).
Fig. 5 Crystal structure of dinuclear copper complex of L6H3.
ꢁc
This journal is The Royal Society of Chemistry 2008
4802 | Chem. Commun., 2008, 4801–4803

View Article Online
Crystal data for [Cu2(L6H)2] (Fig. 5):
Data were collected on a 3 circle Bruker Smart Apex CCD diffrac-
tometer with graphite-monochromated Mo-Ka radiation (l
Table 1 Comparison of the mass transport efficiencies L2–4H2, and
L5H3 with the commercial extractant, 5-nonylsalicylaldoxime (L1H)
=
0.71073 A) and equipped with an Oxford Cryosystems low tempera-
ture device operating at 150 K. The crystal was indexed using the
Cell_now indexing program9 and found to be triclinic with a =
6.8829(7), b = 9.6493(10), c = 12.3771(12) A and a = 88.498(6),
b = 83.850(6), g = 83.339(6)1. The crystal was also twinned and the
twin law obtained after global cell refinement was (ꢀ0.99434 ꢀ0.00064
ꢀ0.02085), (0.00755 ꢀ0.99936 ꢀ0.01644), (ꢀ0.43677 ꢀ0.05385
0.99478). From initial indexing a data collection strategy was refined
which aimed to collect fully complete data to a resolution of 531 in 2y
in as short a time as possible. In total 3802 reflections were collected
Cu Transport
g per kg
reagent
Relative
to L1H
Stoichiometry
(Cu : L)
Reagent
Mr
L1H
263
388
433
325
320
1 : 2
2 : 2
2 : 2
2 : 2
3 : 2
121
164
147
195
298
1.00
1.36
1.21
1.62
2.47
L2H2
L3H2
L4H2
L5H3
ꢀ
and from these the space group was determined to be P1. Absorption
correction was performed using a multi-scan method by applying the
TWINABS10 program to the data. The initial solution was determined
by direct methods with the SHELXS8 program. All heavy atoms were
refined anisotropically and hydrogen atoms were placed geometrically
and allowed to ride on their host atom. Full matrix least squares
refinement was carried out against F2 producing a final conventional
R-factor of 0.0865 based on 3091 reflections.
between 1.36 to 2.47 fold improvements relative to the com-
mercial extractant 5-nonylsalicylaldoxime (L1H) result from
using the reagents L2H2 to L5H3.
Exploiting the improved transport efficiency of these new
reagents will only be possible if they have high solubility in the
hydrocarbon solvents used industrially and are stable to
hydrolysis and show selectivity for Cu-loading in the low pH
ranges used in commercial circuits. Nevertheless, the work
described in this communication has shown that substantial
increases in mass transport efficiency are possible by designing
multi-loading extractants which incorporate into the ligand
super structure several acidic groups which can function in a
metal–metal bridging mode.
1 P. A. Tasker, P. G. Plieger and L. C. West, in Comprehensive
Coordination Chemistry II, ed. J. A. McCleverty and T. J. Meyer,
Elsevier, Oxford, UK, 2004, vol. 9, p. 759.
2 G. Borg, K. Karner, M. Buxton, R. Armstrong and S. W. Van Der
Merwe, Econ. Geol., 2003, 98, 749.
3 P. J. Mackey, CIM Mag., 2007, 2, 35–42.
4 J. Szymanowski, Hydroxyoximes and Copper Hydrometallurgy,
CRC Press, London, 1993.
5 (a) E. W. Ainscough, A. M. Brodie, A. J. Dobbs, J. D. Ranford
and J. M. Waters, Inorg. Chim. Acta, 1998, 267, 27; (b) S. C. Chan,
L. L. Koh, P. H. Leung, J. D. Ranford and K. Y. Sim, Inorg. Chim.
Acta, 1995, 236, 101; (c) P. M. Haba, O. Diouf, A. Sy, M. L. Gaye,
A. S. Sall, A. H. Barry and T. Jouini, Z. Kristallogr. - New Cryst.
Struct., 2005, 220, 479; (d) M. F. Iskander, L. El-Sayed, N. M. H.
Salem, W. Haase, H. J. Linder and S. Foro, Polyhedron, 2004, 23,
23; (e) V. A. Kogan, V. V. Lukov, S. I. Levchenkov, M. Y. Antipin
and O. V. Shishkin, Mendeleev Commun., 1998, 145; (f) W. G. Lu,
C. H. Peng, H. W. Liu and X. L. Feng, Chin. J. Inorg. Chem., 2003,
19, 1222; (g) A. Roth, A. Buchholz, M. Gartner, A. Malassa, H.
Gorls, G. Vaughan and W. Plass, Z. Anorg. Allg. Chem., 2007, 633,
2009; (h) N. R. Sangeetha, K. Baradi, R. Gupta, C. K. Pal, V.
Manivannan and S. Pal, Polyhedron, 1999, 18, 1425; (i) V. F.
Shulgin, O. V. Konnik, K. V. Rabotyagov, V. M. Novotortsev, O.
G. Ellert, V. M. Shcherbakov, I. L. Eremenko, S. E. Nefedov and
Y. T. Struchkov, Zh. Neorg. Khim., 1994, 39, 1486; (j) P. E.
Werner, A. Valent, V. Adelskold and O. Svajlenova, Acta Chem.
Scand., Ser. A, 1983, 37, 51.
6 Smart software, Bruker-Nonius, Bruker-AXS, Madison,
Wisconsin, USA, 2002.
7 G. M. Sheldrick, SADABS, Program for area detector adsorption
correction, Bruker-AXS, Madison, Wisconsin, USA, 2004.
8 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr.,
2008, 64, 112.
9 G. M. Sheldrick, Cell_now indexing program, Bruker-AXS,
Madison, Wisconsin, USA, 2004.
10 G. M. Sheldrick, TWINABS software, University of Gottingen,
Germany, 2003.
The authors thank Cytec Industries Inc. and EPSRC for
funding.
Notes and references
z Crystal data for [Cu2L2]ꢂ2CHCl3 (Fig. 4):
Data were collected on a 3 circle Bruker Smart Apex CCD diffrac-
=
tometer with graphite-monochromated Mo-Ka radiation (l
0.71073 A) equipped with an Oxford Cryosystems low temperature
device operating at 150 K. The crystal was indexed using the Bruker
Smart software6 and found to be triclinic with a = 5.9794(3), b =
9.9352(5), c = 15.3792(9) A, and a = 98.853(4), b = 94.933(4), g =
103.343(4)1. From initial indexing a data collection strategy was
refined which aimed to collect fully complete data to a resolution of
531 in 2y in as short a time as possible. In total 8149 reflections were
ꢀ
collected and from these the space group was determined to be P1.
Absorption correction was performed using a multi-scan method by
applying the SADABS7 program to the data. The data were merged
according to the crystal system in SHELX8 which gave 3052 unique
reflections with a merging R-factor of 0.0442. The initial solution was
determined by direct methods with the SHELXS8 program. All heavy
atoms were refined anisotropically and hydrogen atoms were placed
geometrically and allowed to ride on their host atom. Full matrix least
squares refinement was carried out against F2 producing a final
conventional R-factor of 0.0957 based on 2750 reflections.
ꢁc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 4801–4803 | 4803