Zirconium complexes of amino(polyphenolic) ligands and their hydrolytic stability

Article abstract of DOI:10.1016/j.inoche.2007.06.004

Three amino(polyphenolic) ligands, N,N′-bis(5-tert-butyl-2-hydroxybenzyl)-1,2-diaminoethane 1, tris(5-tert-butyl-2-hydroxy-3-methylbenzyl)amine 3, its 3-chloro analogue 4 and their Zr^IV complexes have been synthesised. The crystal structure of the Zr^IV complex of tris(5-tert-butyl-3-chloro-2-hydroxybenzyl)amine, shows this to be [Zr(4-2H)₂] in which both ligands exists in a zwitterionic form with one alkylammonium and three phenolate groups. The complexes are stable in a two phase, chloroform/water, system at high pH, but the zirconium is stripped at pH < 2.5. The pH value needed to strip 50% of the Zr^IV from the complex [Zr(1-4H)] of the tetraphenolic ligand is ～2.0 whilst the complexes [Zr(3-2H)₂] and [Zr(4-2H)₂] of the triphenolic ligands are slightly more stable with pH_1/2 values of ～1.4. We were unable to use the ligands to extract zirconium from low pH aqueous zirconium oxychloride solutions into an organic phase under a variety of conditions.

Full text of DOI:10.1016/j.inoche.2007.06.004

Inorganic Chemistry Communications 10 (2007) 1154–1158
www.elsevier.com/locate/inoche
Zirconium complexes of amino(polyphenolic) ligands and
their hydrolytic stability
a
b
a,
a
Jy D. Chartres , Ahmed Dahir , Peter A. Tasker *, Fraser J. White
a
School of Chemistry, The University of Edinburgh, Joseph Black Building, Kings Buildings, West Mains Road, Edinburgh EH9 3JJ, UK
b
Nuclear Materials Authority (NMA), P.O. Box 530, Almaadi, Cairo, Egypt
Received 17 May 2007; accepted 8 June 2007
Available online 15 June 2007
Abstract
Three amino(polyphenolic) ligands, N,N0-bis(5-tert-butyl-2-hydroxybenzyl)-1,2-diaminoethane 1, tris(5-tert-butyl-2-hydroxy-3-methyl-
benzyl)amine 3, its 3-chloro analogue 4 and their ZrIV complexes have been synthesised. The crystal structure of the ZrIV complex of tris
(5-tert-butyl-3-chloro-2-hydroxybenzyl)amine, shows this to be [Zr(4-2H)2] in which both ligands exists in a zwitterionic form with one
alkylammonium and three phenolate groups. The complexes are stable in a two phase, chloroform/water, system at high pH, but the zirco-
nium is stripped at pH < 2.5. The pH value needed to strip 50% of the ZrIV from the complex [Zr(1-4H)] of the tetraphenolic ligand is ꢀ2.0
whilst the complexes [Zr(3-2H)2] and [Zr(4-2H)2] of the triphenolic ligands are slightly more stable with pH1/2 values of ꢀ1.4. We were unable
to use the ligands to extract zirconium from low pH aqueous zirconium oxychloride solutions into an organic phase under a variety of
conditions.
ꢀ 2007 Elsevier B.V. All rights reserved.
Keywords: Zirconium complexes; Phenolic ligands; Phenolate complexes
The high affinity of ZrIV and HfIV for anionic oxygen
has led us to consider polyphenolic ligands as solvent
extractants to recover zirconium from aqueous solution.
The only reports in the literature recording the use of phe-
nols in zirconium extraction involve phenolic oximes simi-
lar to those used commercially for copper recovery [1,2].
The potentially sexadentate tetrakis(2-hydroxybenzyl)-
1,2-diaminoethane ligands have been reported by Hefele
et al. [3] to complex TiIV, VIV, MnIV and SnIV. The crystal
structure of a VIV complex has also been reported [4].
Other metal complexes of this ligand type have included
ZnII [5], FeIII [6], and GaIII [7]. Tetraphenols of this type
have been shown to extract gallium from strongly basic
solutions containing aluminates [8], giving high selectivity
for recovery of GaIII over AlIII. In this paper we report
the formation of the ZrIV complex of N,N,N0,N0-tetra-
kis(4-tert-butyl-2-hydroxybenzyl)-1,2-diaminoethane 1.
OH
HO
N
N
tBu
tBu
tBu
tBu
OH
HO
1
The triphenol 2 has been shown to give a 2:1 complex
with ZrIV in which the ligand exists in a zwitterionic form
[9], with phenolate groups defining a pseudo octahedral
ZrIV complex and the two bridgehead nitrogens being
*
Corresponding author. Tel.: +44 131 6504706; fax: +44 131 6506453.
E-mail address: p.a.tasker@ed.ac.uk (P.A. Tasker).
1387-7003/$ - see front matter ꢀ 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.inoche.2007.06.004

J.D. Chartres et al. / Inorganic Chemistry Communications 10 (2007) 1154–1158
1155
protonated to give an overall neutral complex [Zr(2-2H)2].
The reported air-stability of this complex was of interest to
us in relation to the hydrolytic stability of this type of com-
plex and whether ligands of this type, in particular 3 and 4,
would be useful extractants for zirconium.
tion [10], and be fully stripped at a low pH which is
conveniently accessible with hydrochloride or similar acids.
Ligand 1 was synthesised by a modified literature proce-
dure [8] outlined in Scheme 1. 1,2-Diaminoethane was
reacted with 4-tert-butlylsalicylaldehyde to give the double
Schiffs’ base which was subsequently reduced with sodium
borohydride to give bis-phenol 5 in 93% yield. Reaction of
5 with 2-bromomethyl-4-tert-butylphenyacetate followed
by removal of the acetate groups gave the desired product
1 in 35% yield.
R
R'
HO
Ligands 3 and 4 were synthesised via a modified litera-
ture procedure [11] (Scheme 2) in which either 4-tert-
butyl-2-methylphenol or 4-tert-butyl-2-chlorophenol is
reacted with hexamethylenetetramine in the presence of a
catalytic amount of p-toluenesulfonic acid. After chroma-
tography, ligands 3 and 4 were isolated in 76% and 66%
yield, respectively.
R'
N
OH
R
OH
R
Compound 4 was recrystalised from acetonitrile to yield
crystals suitable for X-ray structure determination (Fig. 1).
The free ligand, in contrast to its ZrIV complex (vide infra),
does not exist in its zwitterionic form. It displays an anti-
conformation of the phenolic groups in relation to the
R'
2 R = R' = Me
3 R = Me, R' = tBu
4 R = Cl, R' = tBu
˚
bridgehead amine. An intramolecular H-bond (1.86 A)
exists between one of the phenolic protons and the nitrogen
atom. A second phenolic proton forms an intermolecular
H-bond (2.07 A), linking molecules in a 1D chain through
Chloro-substitution of the triphenol (as in 4) was under-
taken to investigate the effect of altering the pKa of the
ligand. The 4-tert-butyl substituted variants of the triphe-
nol, 3 and 4, and the tetraphenol 1 were prepared as candi-
date extractants to enhance solubility in non-polar and/or
hydrocarbon solvents.
˚
the crystal structure.
The complexes were synthesised by reacting chloroform
solutions of the ligand and zirconium(IV) isopropoxide in
the appropriate ligand to metal ratio. Water was added
to hydrolyse and extract any unreacted zirconium isoprop-
oxide allowing the chloroform solution of the complex to
be separated, dried over magnesium sulphate, and evapo-
rated. The proposed formula for each of the complexes in
supported by high resolution FAB mass spectrometry.
The complex [Zr(1-4H)] displays a lack of rotational
flexibility in the 1H NMR spectrum at 25 ꢁC in chloroform.
Severe broadening of all signals is observed, especially for
the benzylic protons which become non-equivalent. Two
In order to operate efficiently as ‘pH-swing’ extractants
for ZrIV, loading and stripping, e.g.,
Lorg þ Zr4þ ꢀ ½ZrðL-4HÞꢁorg þ 4Hþ ðL ¼ 1Þ
2Lorg þ Zr4þ ꢀ ½ZrðL-2HÞ2ꢁorg þ 4Hþ ðL ¼ 3 or 4Þ
must occur over a pH-range to allow the ligand to be fully
loaded at a pH lower than that at which insoluble zirco-
nium(IV) oxy/hydroxides precipitate from aqueous solu-
tBu
NaBH4
NH
OH
HN
HO
H2N
NH2
MeOH
tBu
tBu
OH
O
5
tBu
1. Na2CO3 / cat. NaI
acetone
1
5
2. KOH
3. AcOH
OAc
Br
Scheme 1.

1156
J.D. Chartres et al. / Inorganic Chemistry Communications 10 (2007) 1154–1158
tBu
Hexamethylenetetraamine
3 or 4
cat.TsOH
R
OH
R= Me or Cl
Scheme 2.
Fig. 2. Crystal structure of [Zr(4-2H)2] (50% probability ellipsoids)
showing H-bonding interactions. The ZrIV atom lies on an inversion
˚
centre. Selected bond lengths (A): Zr1–O11 2.0446(11), Zr1–O12
2.0424(11), Zr1–O13 2.0501(11), N1–H1 0.93(0), O11–H1 2.17(9), O12–
H1 2.17(3), O13–H1 2.06(9). The benzene solvent molecules and the
remaining hydrogen atoms have been omitted for clarity.
Fig. 1. Crystal structure of 4 (50% probability ellipsoids) showing
H-bonding interactions. The remaining hydrogen atoms have been
omitted for clarity.
Despite the similarity between the crystal structures of
[Zr(2-2H)2] and [Zr(4-2H)2], the 1H NMR spectrum of
[Zr(4-2H)2] is much more complicated (and severely broad-
ened) compared to that of [Zr(2-2H)2]. This suggests that in
solution the ligands of [Zr(4-2H)2] can exist in a number of
conformations. While heating the solution began to shar-
pen the signals slightly, even at 110 ꢁC in DMSO-d6 the dif-
ferent conformations were not inter-converting rapidly
enough on the NMR-time scale to see only one species.
Likewise, the 13C NMR spectrum could not be assigned.
In order to assess the viability of using ligands of the
types 1, 3 and 4 as pH-swing extractants for ZrIV we first
undertook stripping of the neutral complexes,
signals are also observed for the aromatic protons that are
ortho to the phenolic oxygen. The spectrum sharpens sig-
nificantly at 80 ꢁC in DMSO-d6, but the signals for the ben-
zylic protons fail to coalesce. Apart from the signals
relating to the tert-butyl groups, the 13C NMR spectrum
could not be assigned as it consisted of severely broadened
signals due to restricted rotation of the bonds.
1
The H NMR spectrum of [Zr(3-2H)2] is similar to that
reported [9] for the zirconium complex of 2. The ammo-
nium proton appears as a broad signal at d 11.8. The ben-
zylic CH2 protons appear as an AB spin system which is
attributed [9] to the C3-symmetry observed in the solid
state being retained in solution, with slow interconversion
of structures.
½ZrðL-4HÞꢁorg þ 2HCl þ H2O ꢀ Lorg þ ZrOCl2 ðL ¼ 1Þ
½ZrðL-2HÞ2ꢁorg þ 2HCl þ H2O ꢀ 2Lorg þ ZrOCl2 ðL ¼ 3 or 4Þ
The crystal structure of [Zr(4-2H)2], crystallised from
benzene, is shown in Fig. 2. The complex is isostructural
with the reported [9] [Zr(2-2H)2] complex, displaying the
same pseudo C3-symmetric coordination mode. Two trian-
ionic ligands are bound to the ZrIV metal centre in an octa-
hedral arrangement of the phenolic oxygen donors. Charge
neutrality is obtained by protonation of the two bridgehead
amines, such that the ligands exist in a zwitterionic form.
As for [Zr(2-2H)2], the same trifurcated N–Hꢂ ꢂ ꢂO3 hydro-
gen bonding between the ammonium proton and the phe-
nolic oxygens is observed. The average Zr–O bond
to establish the pH1/2 values (the pH for 50% loading) for
the ligands. We judged that a viable system should have a
pH1/2 significantly less than 2.8 because we had found that
aqueous zirconium(IV) oxychloride solutions become clou-
dy at a pH of above 2.8.
The pH1/2 values for stripping with aqueous HCl were
determined by stirring aliquots of a 0.005 M solution of
the complex in chloroform with aqueous solutions at vari-
ous pH values which had a constant (0.8 M) chloride con-
centration for 16 hours.
The stripping curve for the tetraphenolate complex
[Zr(1-4H)] is less steep than that of the bistriphenolates
and has a pH1/2 of ca. 2.0 (see Fig. 3). The shallower
˚
˚
length, 2.0457(11) A, is the same as that, 2.059(9) A, in
the complex of the dimethyl-substituted ligand 2 [9].

J.D. Chartres et al. / Inorganic Chemistry Communications 10 (2007) 1154–1158
1157
Fig. 3. The pH dependence of zirconium stripping of 0.005 M chloroform solutions of [Zr(1-4H)] ¤, [Zr(3-2H)2] j and [Zr(4-2H)2] m by aqueous HCl.
S-curve may be a consequence of regeneration of the free
ligand requiring four independent reprotonation equilibria
for 1 compared with only two to regenerate 3 or 4. A pH1/2
of ca. 2.0 for 1 suggests that recovery of ZrIV could not be
accomplished efficiently from aqueous solutions which are
sufficiently acidic (pH < 2.8, vide supra) to prevent hydroly-
sis/precipitation of ZrIV compounds.
Stripping of the triphenolate complex of ligand 3, [Zr(3-
2H)2], into 0.8 M aqueous chloride gave a pH1/2 of ꢃ1.4,
close to a value which could provide a practical recovery
process, provided that loading/stripping S-curves are steep
and that loading behaviour mirrors that shown by the
stripping experiments. On this basis we undertook the syn-
thesis of ligand 4 in the expectation that the incorporation
of the chloro-substituents would increase the acidity of the
phenolic groups (the pKa of 4-tert-butyl-2-chlorophenol is
8.58 [12] versus 10.59 for 4-tert-butyl-2-methylphenol
[13]) and thus increase its strength as a pH-swing extract-
ant. In practice, the strength of the tri-chloro-substituted
ligand 4 (pH1/2 for ZrIV stripping of ꢃ1.4) is very similar
to that of the tri-methyl-substituted analogue 3.
these ligands to extract zirconium could be attributed to
the lack of an available mechanism for the zirconium spe-
cies present in aqueous solution in the pH ranges studied to
react with the ligand at the interface to give the ZrIV com-
plex. The fact that no zirconium is extracted also indicates
that ZrO2+ is also not extracted by these ligands.
In conclusion, we have been able to synthesise ZrIV com-
plexes of both amino tetraphenolate and amino triphenolate
ligands. These complexes are stable in a two phase aqueous
system at high pH while the zirconium can be stripped from
the complex at low pH. The present ligands were unsuitable
for the recovery of zirconium by solvent extraction from
aqueous zirconium(IV) oxychloride solutions.
Acknowledgements
We thank the EPSRC for a postdoctoral fellowship (JC)
and the Nuclear Materials Authority, Egypt, for a study
fellowship (AD).
Appendix A. Supplementary material
When extraction experiments in which 0.01 M chloro-
form solutions of these ligands were contacted with
0.005 M zirconium oxychloride solutions at pH values
from 0 to 2.5 for 16 hours, no extraction of zirconium
was detected. To determine whether high chloride concen-
tration could promote zirconium extraction the experiment
was repeated for ligand 3 with zirconium oxychloride solu-
tions containing 0.8 M chloride ion. Again no zirconium
uptake was detected in the organic phase.
As the kinetics of complex formation can be an impor-
tant factor in extraction processes, an experiment was per-
formed in which a 0.01 M chlorobenzene solution of ligand
3 was intimately stirred with a 0.005 M aqueous solution of
zirconium oxychloride at 60 ꢁC. Even after 5 days, no zir-
conium was detected in the organic phase. The failure of
Cryatallographic data has been deposited at the Cam-
bridge Crystallographic Data Centre with the Deposition
Numbers 646827 and 646828. Copies of this information
may be obtained free of charge from The Director, CCDC,
12 Union Road, Cambridge, CB2 1EZ, UK (fax: +44 1223
336033; e-mail: deposit@ccdc.cam.ac.uk or http://
www.ccdc.cam.ac.uk). Experimental data associated with
this article can be found, in the online version, at
doi:10.1016/j.inoche.2007.06.004.
References
[1] B.R. Reddy, J.R. Kumar, Hydrometallurgy 74 (2004) 173.
[2] S. Banerjee, S. Basu, J. Radioanal. Nucl. Chem. 262 (2004) 733.

1158
J.D. Chartres et al. / Inorganic Chemistry Communications 10 (2007) 1154–1158
[3] H. Hefele, E. Ludwig, W. Bansse, E. Uhlemann, T. Luegger, E. Hahn,
H. Mehner, Z. Anorg. Allg. Chem. 621 (1995) 671.
[4] A. Neves, A. Ceccato, I. Vencato, Y. Mascarenhas, C. Erasmus-Buhr,
J. Chem. Soc., Chem. Commun. (1992) 652.
[5] C. Higham, D. Dowling, J. Shaw, L. Janet, A. Cetin, C. Ziegler, J.
Farrell, Tetrahedron Lett. 47 (2006) 4419.
[9] M. Davidson, C. Doherty, A. Johnson, M. Mahon, J. Chem. Soc.,
Chem. Commun. (2003) 1832.
[10] A. Veyland, L. Dupont, J.-C. Pierrard, J. Rimbault, M. Aplincourt,
Eur. J. Inorg. Chem. 11 (1998) 1765, and refs therein.
[11] A. Chandrasekaran, R. Day, R. Holmes, J. Am. Chem. Soc. 122
(2000) 1066.
[6] I. Vencato, A. Neves, B. Vincent, C. Erasmus-Buhr, W. Haase, Acta
Crystallogr. C 50 (1994) 3786.
[12] H.N. Simpson, C.K. Hancock, E.A. Meyers, J. Org. Chem. 30 (1965)
2678.
[7] S. Madsen, M. Welch, R. Motekaitis, A. Martell, Nucl. Med. Biol. 19
(1992) 431.
[8] P. Quan, S. Leeming, Patent US No. 5,308,595, 1994.
[13] P. Demerseman, J.P. Lechartier, R. Reynaud, A. Cheutin, R.
Royer, P. Rumpf, M.-L. Desvoye, Bull. Soc. Chim. France (1963)
2559.