The use of hydrogenated Schiff base ligands in the synthesis of multi-metallic compounds II

Article abstract of DOI:10.1016/j.poly.2011.03.004

Tris(2-hydroxybenzylaminoethyl)amine (H₃L) complexes of nickel, copper and zinc are investigated as potential metallo-ligands ([(H _xL)M]; x = 0, 1: M = Ni, Cu, Zn). The homometallic complexes formed are dimetallic ([{(HL)Ni}Ni(OAc)₂] and [{(L)Zn}ZnCl]), tetrametallic ([{(L)Cu}Cu]₂²⁺) and hexametallic ([{(L)Ni}Ni ₂(μ-OH)₂(OEt)(OH₂)]₂). Hetero-dimetallic complexes can be formed with [(HL)Ni] and copper chloride ([{(HL)Ni}CuCl₂]) or zinc bromide ([{(HL)Ni}ZnBr₂]). The metallo-ligand acts as a chelating agent using phenolate pairs. The remaining phenolate either does not coordinate or can be used to increase the number of metals included in the scaffold from two to four or six. Not all combinations are possible and [(HL)Cu]⁺ produces a charge separated species with zinc chloride rather than a complex. An exchange reaction is observed to take place when [(HL)Zn]⁺ is treated with the halides of nickel or copper producing [(HL)M]⁺ (M = Ni, Cu, respectively).

Full text of DOI:10.1016/j.poly.2011.03.004

Polyhedron 30 (2011) 1530–1537
Contents lists available at ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
The use of hydrogenated Schiff base ligands in the synthesis of multi-metallic
compounds II q
⇑
Abdullahi Mustapha, Christoph Busche, John Reglinski , Alan R. Kennedy
Department of Pure and Applied Chemistry, University of Strathclyde, Glasgow G1 1XL, United Kingdom
a r t i c l e i n f o
a b s t r a c t
Article history:
Tris(2-hydroxybenzylaminoethyl)amine (H3L) complexes of nickel, copper and zinc are investigated as
Received 25 January 2011
Accepted 3 March 2011
Available online 21 March 2011
potential metallo-ligands ([(HxL)M]; x = 0, 1: M = Ni, Cu, Zn). The homometallic complexes formed are
2+
dimetallic ([{(HL)Ni}Ni(OAc)2] and [{(L)Zn}ZnCl]), tetrametallic ([{(L)Cu}Cu]2
([{(L)Ni}Ni2(
) and hexametallic
l
-OH)2(OEt)(OH2)]2). Hetero-dimetallic complexes can be formed with [(HL)Ni] and copper
chloride ([{(HL)Ni}CuCl2]) or zinc bromide ([{(HL)Ni}ZnBr2]). The metallo-ligand acts as a chelating agent
using phenolate pairs. The remaining phenolate either does not coordinate or can be used to increase the
number of metals included in the scaffold from two to four or six. Not all combinations are possible and
[(HL)Cu]+ produces a charge separated species with zinc chloride rather than a complex. An exchange
reaction is observed to take place when [(HL)Zn]+ is treated with the halides of nickel or copper producing
[(HL)M]+ (M = Ni, Cu, respectively).
Keywords:
Multidentate
Schiff base
Nickel
Copper
Zinc controlled organisation
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
complexes are five coordinate employing an N4O coordination
sphere. As a consequence of the protonated phenolate (Fig. 1c)
Multidentate ligands such as tris(2-hydroxybenzylaminoeth-
yl)amine (H3L) and related systems have been of some interest
over the past 10 years as they offered a direct route to the synthe-
sis of mixed trimetallic d–f metal complexes (Fig. 1) for magneto-
chemical studies [1–5]. The synthesis of species such as
[{(L)Ni}2Ln(HOMe)]+ (Ln = lanthanoid) were typically achieved by
the direct reaction of the ligand (H3L) with the nitrate salts of nick-
el and the desired lanthanoid. The complexes self-assemble in
solution with the transition metal choosing to bind preferentially
to the four nitrogen donors and the lanthanoid exclusively to the
phenolates. These complexes do not have C3 symmetry as only
two of the phenolates are used to bridge the metals. The absence
of information on the simple complexes of H3L with nickel, which
must form during the synthesis of these species, led us to the iso-
lation of the intermediate complex [(HxL)Ni]+ (x = 1, 2) including
the related copper and zinc complexes [(H2L)M]+ (M = Cu, Zn) [6].
Although the three transition metal complexes have identical
empirical formula they are not isostructural (Fig. 1). The nickel sits
in a six coordinate N4O2 environment whereas the copper and zinc
the nickel complex remains cationic.
The subtle differences in the structures of the nickel and copper
complexes (Fig. 1) go some way to explain why the synthesis of the
analogous mixed copper–lanthanoid complexes of H3L cannot be
isolated [6]. Instead of forming a trimetallic LnCu2 complex with
H3L we obtain the trimetallic copper complex [(HL)2Cu3]4+
(Fig. 2). Extending our studies to include alkali metals we recently
reported that nickel again had a preference for tetrametallic
(Ni2K2) complexes such as [{(L)Ni}K
and zinc generated larger multimetallic complexes such as
[({(L)Cu}K)(OH2)4- 4-OH2] (Fig. 2), which contain a K4O core and
l-(EtOH)]2 whereas copper
l
are essentially 1:1 K:Cu and K:Zn, respectively [7].
We are beginning to regard complexes such as [(HxL)M]+ as po-
tential ligands [7]. Complexes such as [{(L)Ni}2Ln(HOMe)]+ demon-
strate that, with the correct conditions, it is possible to overcome
the positive charge on the primary complex and organise the phe-
nolate donors such that they create a binding site for metals [4–7].
Thus far the greatest success has been had with divalent metals,
such as nickel, which prefers an octahedral geometry and which
leave a residual negative charge on the metallo-ligand. However,
our recent work [7] on the alkali metal adducts of [(HL)M]+ sug-
gested that this concept could be extended to the corresponding
metal complexes of copper and zinc as long as the charge on the
metal centre being introduced is not too high (P3). To investigate
this hypothesis we have extended our studies on the H3L adducts
of nickel copper and zinc and their subsequent combination with
the divalent transition metals nickel copper and zinc.
Abbreviations: HBTC, tris(hydroxybenzyl)triazacyclononane; HAME, tris
(hydroxybenzyl)aminomethylethane; HAC, tris(hydroxybenzyl)aminocyclohexane;
Salpr, bis(salicylidene)-1,3-propanediamine; TrenSal, tris(2-hydroxybenzylidene)
aminoethylamine; H3L, tris(2-hydroxybenzyl)aminoethylamine.
q
For part I see Ref. [6].
⇑
Corresponding author. Tel.: +44 0141 548 2349; fax: +44 0141 552 0876.
E-mail address: j.reglinski@strath.ac.uk (J. Reglinski).
0277-5387/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2011.03.004

A. Mustapha et al. / Polyhedron 30 (2011) 1530–1537
1531
+
+
N
N
N
HN
NH
HN
HN
NH
NH
Ni
NH
M
NH
NH
OH
OH
HO
OH
HO
O
O
OH
HO
M = Cu, Zn
b
c
a
Fig. 1. Tris(2-hydroxybenzylaminoethyl)amine (H3L) (a) and a schematic representations of its cationic nickel (b), copper and zinc (c) complexes; [(H2L)M]+ (M = Ni, Cu, Zn).
Fig. 2. (Left) The X-ray crystal structure of [(HL)2Cu3] 4NO3. (Right) The X-ray crystal structure of [({(L)Cu}K)(OH2)4-
l4-OH2]. The expansion in the motif from trimetallic to
octametallic is supported by the bridging nature of the pendant phenolates (cf nickel Fig. 1). These structures suggest that [(H2L)Cu]+ is capable of chelating low valent metals.
The methanol was removed and the product recrystallised from
acetonitrile/diethyl-ether by vapour diffusion. Yield 67%. Anal. Calc.
for C31H40N4Ni2O7: C, 53.33; H, 5.78; N, 8.03. Found: C, 53.19; H,
2. Experimental
All experiments were carried out using standard apparatus and
commercially available chemicals. Tris(2-hydroxybenzylideneami-
noethyl)amine (TrenSal) and tris(2-hydroxy-benzylaminoethyl)-
amine (H3L) were prepared as previously reported [6,8]. Solid
reflectance spectra (400–900 nm) were recorded on a Photonics
CCD array UV–Vis spectrophotometer. Mass spectra were recorded
using a Thermo Finnigan LCQDuo by electrospray ion trap. Infra red
spectra were recorded as potassium bromide discs using a Nicolet
Avatar 360 FTIR spectrometer.
6.10; N, 8.22%. FTIR [(m
/cmꢀ1 (KBr)]: 3430, 2995, 1565, 1480,
1275, 770. Mass spec. (ESI) m/z 521 (LNi+). kmax (solid reflectance)
no bands.
2.2. Preparation of [{(L)Ni}Ni2(l-OH)2(l-OEt)(OH2)]2 NO3 Cl
H3L (0.1 g, 0.22 mmol) was dissolved in ethanol (20 ml), fol-
lowed by addition of Ni(NO3)2ꢁ6H2O (0.06 g, 0.22 mmol). The mix-
ture was stirred at 50 °C after which three drops of triethylamine
was added. NiCl2ꢁ6H2O (0.07 g, 0.30 mmol) was added. The green-
ish solution was refluxed for 1 h, allowed to cool and filtered. Crys-
tals were grown directly from the mother liquors by vapour
X-ray measurements were conducted using Oxford Diffraction
and Nonius Kappa CCD diffractometers using graphite monochro-
mated Mo K
a radiation, apart from [{(L)Zn}ZnCl] and [{(HL)Ni}-
CuCl2] which were collected at Daresbury Synchrotron Radiation
Source on a Bruker Smart APEX2 CCD diffractometer. Crystals were
coated in mineral oil and mounted on glass fibres. Full matrix
least-squares refinement to convergence was based on F2. The
structure solution and refinement used the programs SHELXS and
SHELXL-97 [9] and the graphical interface WinGX [10]. A summary
of the crystallographic and refinement parameters is given in Table
1. Full details are available in cif format as Supplementary
Information.
diffusion with diethyl-ether. Yield 32%. Anal. Calc. for C58H84
-
N9Cl1O17Ni6ꢁ5H2O: C, 42.04; H, 6.81; N, 7.61. Found: C, 41.76; H,
7.05; N, 7.62%. FTIR [(m
/cmꢀ1 (KBr)]: 3450, 2925, 1635, 1600,
1480, 1383, 1285, 760. Mass spec. (ESI) m/z 521 (100% LNi+),
1063 (10% L2Ni2Na), 1099 (1% L2Ni3). kmax (solid reflectance) 560,
630, 950.
2.3. Preparation of [{(L)Cu}Cu]2 [NO3]2
2.1. Preparation of [{(HL)Ni}Ni(OAc)2]
H3L (0.92 g, 2.0 mmol) was dissolved in methanol (15 ml), fol-
lowed by the addition of copper nitrate (0.75 g, 4.0 mmol) in meth-
anol (10 ml). The mixture turned dark green. After stirring for
10 min a few drops of triethylamine were added upon which the
solution became cloudy. The solution was filtered to give a
green powder. X-ray quality crystals were grown from the slow
H3L (1.10 g, 2.36 mmol) was dissolved in methanol (50 ml), fol-
lowed by the addition of nickel (II) acetate tetrahydrate (1.17 g,
4.72 mmol). The mixture was brought to reflux for 60 min, the
resultant bluish solution was allowed to cool and then filtered.

Table 1
Selected crystallographic and refinement parameters.
[{(HL)Ni}Ni(OAc)2]
[{(L)Ni}Ni2(
(OH2)]2 [Cl1.33 (NO3)0.67]ꢁxH2O
C59.6H86Cl1.34N8.66Ni6O16.77
l
-OH)2(OEt)-
[{(L)Cu}Cu]2 2NO3ꢁ3MeOH
[{(L)Zn}ZnCl]ꢁ
[{(HL)Ni}CuCl2] MeNO2
[{(HL)Ni} ZnBr2]ꢁMeOH
[(H2L)Cu]+ Et3NH+ [ZnCl4]2ꢀ
MeOH
A
Empirical formula
Formula weight
Crystal system
Space group
a (Å)
C31H40N4Ni2O7
698.09
P21/n
C57H78Cu4N10O15
1397.45
P21
C28H37ClN4O4Zn2
659.81
P21/c
C28H37Cl2CuN5NiO5
716.78
P21/c
C28H37Br2N4NiO4Zn
777.52
C33H51Cl4CuN5O3Zn
836.50
1591.89
ꢀ
ꢀ
ꢀ
P1
P1
P1
monoclinic
13.2085(3)
17.7615(3)
13.7552(3)
triclinic
11.1482(5)
11.5773(3)
15.9213(7)
103.717(3)
91.233(3)
114.348(3)
1
monoclinic
12.6994(5)
17.2400(7)
14.8768(5)
monoclinic
18.744(3)
8.8981(12)
17.531(2)
monoclinic
8.9235(9)
22.086(2)
15.3655(15)
triclinic
11.5412(4)
15.2088(5)
17.8402(6)
94.130(3)
98.336(3)
94.964(3)
4
triclinic
10.9046(4)
12.9133(4)
14.9948(4)
81.563(2)
69.780(2)
71.069(2)
2
b (Å)
c (Å)
a
(°)
b (°)
98.173(2)
108.853(2)
92.1470(10)
90.841(2)
c
(°)
Z
4
2
4
4
V (Å3)
3194.23(11)
123
1802.14(12)
123
3082.3(2)
123
2921.9(7)
120
3027.0(5)
120
3075.37(18)
123
1872.76(10)
120
T (K)
k (Å)
0.71073
1.231
0.71073
1.653
0.71073
1.434
0.67100
1.773
0.67100
1.547
0.71073
2.194
0.71073
1.533
lcalc (mmꢀ1
)
No. of reflections measured
No. of unique reflections
Rint
34 059
7698
0.0324
6023
33 023
7051
0.0431
5348
29 534
13 472
0.080
27 063
8701
0.0604
7329
22 905
5351
0.0782
3922
27 496
12 065
0.0683
5564
41 215
8581
0.0626
6320
No observed
9998
No. of parameters
400
431
0.0575
0.1483
1.032
782
354
385
685
0.0417
0.0774
0.732
451
0.0425
0.0848
1.015
Ra (I > 2
r
(I))
0.0299
0.0791
0.978
0.0538
0.0912
1.039
0.0467
0.1367
0.990
0.0484
0.1288
1.027
b
Rw (all reflections)
Goodness-of-fit (GOF)
Instrument
Nonius Kappa
Oxford Gemini S
Nonius Kappa
Daresbury SRS
Daresbury SRS
Oxford Gemini S
Nonius Kappa
A
Indicates that the program SQUEEZE was used to remove small amounts of disordered solvent from the model crystal structure. This has not been included in the given formula but is calculated to be roughly equivalent to five
water molecules.

A. Mustapha et al. / Polyhedron 30 (2011) 1530–1537
1533
evaporation of the mother liquor. Yield 67%. Anal. Calc. for
54H66N10O12Cu4 2MeOH: C, 49.26; H, 5.46; N, 10.26. Found: C,
49.21; H, 5.32; N, 10.54%. FTIR [(
/cmꢀ1 (KBr)]: 2925, 1634, 1562,
2.7.1. Reactions of [(H2L)Cu]+ with nickel halides
C
This reaction was carried as described above on a 0.5 mmol
scale. The products obtained from the reaction was consistent with
[(H2L)Cu]+ [7]. The parent ion (m/z 521) of [(H2L)Ni]+ is a dominant
feature of the mass spectrum of the reaction solution indicative of
some metal exchange.
m
1408, 1383, 1276, 1020, 871, 753. Mass spec. (ESI) m/z 526
(100%, LCu+). kmax (solid reflectance) 670 nm.
2.4. Preparation of [{(L)Zn}ZnCl]
2.7.2. Reactions of [(H2L)Zn]+ with nickel or copper halides
These reactions were carried as described above on a 0.5 mmol
scale. The products obtained were consistent with [(H2L)Ni]+ and
[H2L)Cu]+, respectively [7].
H3L (0.1 g, 0.22 mmol) was dissolved in methanol (15 ml), to
which was added Zn(NO3)2ꢁ6H2O (0.06 g, 0.22 mmol). The mixture
was stirred at 50 °C for 5 min, followed by addition of two drops of
triethylamine. A methanolic (5 ml) solution of ZnCl2 (0.03 g,
0.22 mmol) was added to the mixture drop-wise to obtain a
brownish solution. This was allowed to cool and filtered. Crystals
were grown by the slow evaporation of the mother liquor. Yield
47%. Anal. Calc. for C27H33N4O3Zn2ClꢁMeOHꢁ3H2O: C, 47.11; H,
2.8. Preparation of [(HL)Fe]NO3
H3L (0.13 g, 0.28 mmol) was dissolved in methanol (30 ml) fol-
lowed by addition of Fe(NO3)3ꢁ9H2O (0.11 g, 0.28 mmol). To the
brown mixture three drops of KOH solution (in methanol) was
added, the mixture was refluxed for 1 h, allowed to cool and
filtered. Crystals were grown directly from the mother liquors by
vapour diffusion with diethyl-ether. Although the crystals were
subjected to crystallographic analysis the data obtained failed to
refine to a satisfactory solution (see Supplementary material).
Crystal data: [(HL)Fe], orthorhombic Pc21a, a = 21.7392(18) Å,
b = 7.2251(4) Å, c = 38.0527(31) Å, V = 5979.61(8) Å3. Anal. Calc.
for C27H34N5O6Fe: C, 56.01; H, 5.90; N, 11.61. Found: C, 56.01; H,
6.07; N, 7.85. Found: C, 47.25; H, 5.15; N, 8.12%. FTIR [(m
/cmꢀ1
(KBr)]: 3426, 2934, 1485, 1275, 763. Mass spec. (ESI) m/z 527
(100%, LZn+), 593 (20%, LZn2+).
2.5. Preparation of [{(HL)Ni}CuCl2]
H3L (0.1 g, 0.22 mmol) was dissolved in methanol (15 ml), to
which was added Ni(NO3)2ꢁ6H2O (0.06 g, 0.22 mmol). The mixture
was stirred at 50 °C for 5 min, followed by the addition of two
5.90; N, 12.06%. FTIR [(m
/cmꢀ1 (KBr)]: 1475, 1265, 1388,, 615. Mass
drops triethylamine.
A methanolic solution (5 ml) of CuBr2
spec. (ESI) m/z 518 (100%, LFe+); 1090 (29%, KL2Fe3+). kmax (solid
(0.05 g, 0.22 mmol) was added dropwise to the mixture. The solu-
tion changed colour immediately from purple to brown. The solu-
tion was allowed to cool and filtered to produce a brown powder
which was recrystallised from nitromethane. Yield 68%. Anal. Calc.
for C27H34N4Cl2Cu1O3Ni1ꢁMeNO2: C, 46.92; H, 5.20; N, 9.77. Found:
reflectance) 475 nm.
3. Results and discussion
C, 46.44; H, 5.05; N, 9.17%. FTIR [(m
/cmꢀ1 (KBr)]: 2925, 1610, 1480,
The multidentate ligand systems H3L can be prepared with ease
by hydrogenating the Schiff base species generated from the tris
(2-aminoethyl)amine and salicylaldehyde [8]. Treatment of H3L
in situ with three equivalents of nickel acetate produced an off-
white material. It is notable that the parent [(H2L)Ni]+ has a prom-
inent band in the near infra-red (950 nm) which is absent in this
product. Recrystallisation revealed the complex to be dimetallic;
namely [{(HL)Ni}Ni(OAc)2] (Fig. 3). The internalised nickel is coor-
dinated by an N4O2 motif in a similar manner to [(H2L)Ni]+ [6].
However, the additional pendant phenoxide is not deprotonated
and consequently the metallo-ligand does not carry a negative
charge. Nickel acetate chelates to the phenolates (O1, O2) which
1390, 760. Mass spec. (ESI) m/z 521 (100%, LNi+). The parent ion
(m/z 526) of [(H2L)Cu]+ is a prominent feature of the mass spec-
trum of the reaction solution indicative of some metal exchange.
kmax (solid reflectance) 480, 820.
2.6. Preparation of [{(HL)Ni}ZnBr2]
H3L (0.1 g, 0.22 mmol) was dissolved in methanol (15 ml), to
which was added Ni(NO3)2ꢁ6H2O (0.06 g, 0.22 mmol). The mixture
was stirred at 50 °C for 5 min, followed by the addition of two
drops of triethylamine, after which methanolic (5 ml) solution of
ZnBr2 (0.05 g, 0.22 mmol) was added drop-wise. The light purple
mixture was stirred for further 10 min, allowed to cool and filtered.
Crystals were obtained by slow evaporation of the saturated solu-
tion. Yield 72%. Anal. Calc. for C27H33N4O3Ni1Zn1Br2ꢁMeOH: C,
43.20; H, 4.80; N, 7.04. Found: C, 42.90; H, 5.01; N, 7.04%. FTIR
[(m
/cmꢀ1 (KBr)]: 3406, 2980, 1475, 1255, 758. Mass spec. (ESI) m/
z 521 (100%, LNi+). kmax (solid reflectance) 555, 950.
2.7. Preparation of [(H2L)Cu] Et3NH+ [ZnCl4]
H3L (0.23 g, 0.5 mmol) was dissolved in methanol (15 ml), fol-
lowed by addition of Cu(NO3)2ꢁ2.5H2O forming a green solution
of [(H2L)Cu]+. The solution was stirred at 50 °C for 5 min. Four
drops of triethylamine was added and the mixture stirred for a fur-
ther 2 min.
A methanolic (5 ml) solution of ZnCl2 (0.07 g,
0.05 mmol) was added drop-wise and the pale green mixture
allowed to cool. The solution was filtered and allowed to stand.
X-ray quality crystals were grown by slow evaporation of the
saturated solution. Yield 82%. Anal. Calc. for C33H51Cl4Cu1N5O3Zn1:
C, 42.78; H, 6.64; N, 7.56. Found: C, 42.77; H, 5.19; N, 7.69%. FTIR
Fig. 3. The X-ray crystal structure of [{(HL)Ni}Ni(OAc)2]. The metrical parameters of
the [(H2L)Ni] moiety are given in Table 2. The thermal ellipsoids are drawn at 50%
probability.
[(
m
/cmꢀ1 (KBr)]: 2980, 1603, 1480, 1265, 765. Mass spec. (ESI)
m/z 526 (100%, LCu+).

1534
A. Mustapha et al. / Polyhedron 30 (2011) 1530–1537
are held close to one another by the encapsulated nickel. Hetero-
atomic nickel species such as [{(Salpr)Ni}M(acac)2] (M = Mn, Cu)
which are loosely related to this complex have been reported pre-
viously by O’Conner et al. Central to this report is the manner in
which the two phenolates of their tetradentate Schiff base ligand
(Salpr) act as a didentate ligand to the diketonate of manganese
and copper [11]. This motif would seem to be maintained here
(Fig. 3) despite the presence of the third potential donor (vide in-
fra). In the absence of any additional stabilising interaction from
the pendant phenolate the dimetallic assembly is somewhat fragile
and its mass spectrum is dominated by [(H2L)Ni]+ (m/z 521). In-
deed its synthesis is best achieved in the presence of excess nickel
acetate (Eq. (1) and (2)).
group. These factors were expected to facilitate an increase in the
metal content of the complex. With Schiff base ligands (e.g. Tren-
Sal, Salpr) the product would be expected to be trimetallic [8,12].
However, with the more flexible H3L the ligand accumulates three
metal centres and dimerises to produce the hexametallic dicationic
complex [[{(L)Ni}Ni2(l
-OH)2(OEt)(OH2)]2]2+ (Fig. 4). Using its phe-
nolate donors (O1, O2), the [(L)Ni] ligand is again didentate to a
second nickel (Ni2). One of the nickel amine (N4) interactions ob-
served in [{(HL)Ni}Ni(OAc)2] (Fig. 3) is severed and its position
occupied by a bridging hydroxide (O5). The amine phenolate re-
leased by Ni1 chelates to the third nickel (Ni3). The presence of a
second bridging hydroxide (O6) completes the basic framework
around which the Ni3 motif dimerises. An ethoxide and water
complete the coordination sphere of the internal nickels (Ni2,
Ni3). The nesting, open faced, cube motif at the centre of this struc-
ture is relatively common within nickel chemistry [13]. However,
the presence of the additional nickel (Ni1) contained within
[(L)Ni] as capping units lifts the complex into a more selective
group of hexametallic nickel complexes [14–16]. Furthermore,
the bond lengths and angles for the species retrieved are typical
for this type of motif (Fig. 4, Table 2).
2NiðOAcÞ þ H3L ! ½ðH2LÞNiꢂþ þ NiðOAcÞ2
ð1Þ
ð2Þ
2
½ðH2LÞNiꢂþ þ NiðOAcÞ2 ¢ ½fðHLÞNigNiðOAcÞ2ꢂ
Employing nickel chloride instead of nickel acetate exchanges a
didentate ligand for unidentate ligand which is also a good leaving
Fig. 4. The X-ray crystal structure of [[{(L}Ni2(l
-OH)2(OEt)(OH2)]2]2+. The disor-
dered chloride and nitrate counter ions are not shown. The metrical parameters of
the [(H2L)Ni] moiety are given in Table 2. Selected bond lengths and distances (Å)
for the Ni4O8 core: Ni2–O3, 2.022(3); Ni1–O4, 2.103(3); Ni2–O2, 2.096(3); Ni3–O4,
2.051(3); Ni3–O5, 2.064(3); bond angles (°): O4–Ni2–O5, 83.4(1); O5–Ni3–O4,
82.9(1). The thermal ellipsoids are drawn at 50% probability.
Fig. 5. The X-ray crystal structure of [{(L)Cu}Cu]22+. The metrical parameters of the
[(H2L)Cu] moiety are given in Table 2. Selected other bond lengths and distances
(Å): Cu1ꢁ ꢁ ꢁCu2, 3.086; Cu2ꢁ ꢁ ꢁCu3, 3.056. Bond angles (°): Cu1–Cu2–Cu3, 146.27. The
thermal ellipsoids are drawn at 50% probability.
Table 2
Selected geometric parameters for [{(HL)Ni}Ni(OAc)2], [{(L)Ni}Ni2(l
-O)2(OEt)(OH2)]2 [{(L)Cu}Cu]2 2NO3, [{(L)Zn}ZnCl], [{(HL)Ni}CuCl2], [{(L)Ni}ZnBr2] and [{(HL)Ni]+.
M1–N2
M1–O1
M1–O2
M2–O1
M2–O2
M1ꢁ ꢁ ꢁM2
N2–M1–N1
N2–M1–O1
N2–M1–O2
M1–O1–M2
M1–O2–M2
O1–M2–O2
N2–M1ꢁ ꢁ ꢁM2
[{(HL)Ni}CuCl2]
[{(HL)Ni}ZnBr2]
2.088(4)
2.108(4)
2.099(3)
2.019(3)
2.089(3)
2.101(3)
1.969(3)
1.940(3)
1.980(4)
1.974(3)
3.1118(8)
3.0482(9)
82.68(15)
81.0(2)
106.98(13)
177.23(13)
105.70(15)
174.23(16)
103.59(13)
99.76(13)
96.99(14)
97.82(15)
80.59(12)
84.64(14)
145.45(10)
145.61(12)
[{(HL)Ni}Ni(OAc)2]
2.114(2)
2.141(4)
2.0795(12)
2.0454(12)
2.057(3)
2.0421(12)
2.0114(11)
2.090(4)
3.1211(3)
81.38(6)
106.11(5)
174.37(5)
101.50(15)
177.27(16)
98.44(5)
81.22(5)
146.30(4)
100.59(5)
82.95(13)
83.39(13)
[[{(L)Ni}Ni2(
l
-OH)2(OEt)(OH2)]2]2+
2.7463(8)
3.0675(8)
(M2–M3)
3.0861(8)
3.0556(8)
(M2–M3)
3.0485(4)
85.63(16)
78.95(14)
132.83(11)
2.032(4)
2.096(3)
2+
[{(L)Cu}Cu]2
2.338(4)
1.947(3)
1.922(3)
1.952(3)
2.225(3)
82.60(16)
91.83(14)
116.61(15)
104.63(15)
95.90(14)
75.02(13)
86.90(6)
[{(L)Zn]}ZnCl]
[(HL)Ni]+ [6]
2.2302(17)
2.103
2.1770(14)
2.0469(15)
1.9820(15)
1.9522(15)
80.28(6)
89.93
111.16(6)
169.23(6)
94.16(6)
99.31(6)
151.55(4)
2.003
2.189
175.13
97.09

A. Mustapha et al. / Polyhedron 30 (2011) 1530–1537
1535
The ability to generate multimetallic complexes of copper
would seem to be more favourable as both Orvig and co-workers
[5] and ourselves [6] have reported trimetallic species using mem-
bers of this family of ligand (Fig. 2). Surprisingly, in this study the
treatment of H3L with copper nitrate generates the tetrametallic
metallic nature through the bridging phenolates (O3, O4). The five
coordinate nature of each of the copper centres help generate the
gentle twist (<Cu–Cu–Cu–Cu, 122.6°) that passes through the core
of the complex. A small number of tetrametallic complexes of cop-
per have been reported. Of interest here is the complex, ([{(Sal-
pr)Cu]2Cu(OAc)(OMe)]2) reported by Fukuhara et al. [17]. This is
constructed from the tetradentate copper Schiff base complex
[(Salpr)Cu], complexed to a second copper, again through the phe-
nolate donors and an acetate bridge (pace an internal phenolate O2,
Fig. 5). The motif dimerises about a bridging methanol (pace the
extended phenolate O3, Fig. 5). The copper atoms in [{(Sal-
pr)Cu]2Cu(OAc)(OMe)]2 have similar spacing (3.124 Å, 3.028 Å)
and a similar undulation (<Cu–Cu–Cu, 147°) within the imaginary
Cu4 chain to the complex reported here (Fig. 5). A major difference
between copper and nickel is the propensity for nickel to utilise a
greater number of donor atom positions than copper. For copper
the remaining donors are not only sufficient in number but due
to the flexibility of H3L, they are placed forward of the second me-
tal thus promoting dimerisation without the need for additional
donors (e.g. hydroxide).
2+
cation [{(L)Cu}Cu]2 (Fig. 5). Each of the copper centres sits in
a
five coordinate environment. Furthermore the terminal
copper atoms (Cu1, Cu4) are in an unfamiliar N3O2 motif. Similar
to [[{(L)Ni}Ni2(
-OH)2(OEt)(OH2)]2]2+ and in contrast with
l
[(H2L)Cu]+, one of the amine donors (N4) is displaced from the
coordination sphere of the encapsulated metal in favour of a sec-
ond phenolate (O2). The central copper atoms (Cu2, Cu3) sit in
an NO4 motif forming two chelates; one to the cleft of the L moiety
(O1, O2) and a second with the pendant phenolate (O3) and its
attendant secondary amine (N4). The complex achieves its tetra-
The zinc chemistry of TrenSal and H3L demonstrates similarities
to both nickel and copper. With TrenSal, zinc adopts the motif dis-
played by nickel namely an octahedral trimetallic complex [8],
whereas with H3L zinc follows copper [7] forming a five coordinate
N4O complex (Fig. 1). On treating H3L with excess zinc chloride we
obtained a second dimetallic complex (Fig. 6). The motif is a hybrid
of that displayed by nickel (Figs. 3 and 4) and copper (Fig. 5). Sim-
ilar to nickel the internalised zinc (Zn1) sits within an N4O2 motif.
The bond distances and angles differ only by what might be ex-
pected when introducing a cation of slightly larger radius (Table
2). Similar to [{(HL)Ni}Ni(OAc)2], the external zinc chloride (Zn2)
is again held between the chelating phenolates (O1, O2). However,
in contrast to [{(HL)Ni}Ni(OAc)2] and similar to [{(L)Cu}Cu]22+, the
remaining phenolate coordinates. However, again it is projected
beyond the second zinc filling one of the chloride coordination
2+
sites. Dimerisation to a tetrametallic complex in [{(L)Cu}Cu]2 is,
we believe, driven by the reorganisation of the pendant phenolate
which is used as a chelating site. In the zinc complex, the amine
phenolate (N4) does not migrate from Zn1 to Zn2 and dimerisation
does not occur. A number of simple dimetallic zinc complexes are
known [18–19]. These comprise a zinc dihalide chelated to the two
Fig. 6. The X-ray crystal structure of [{(L)Zn]}ZnCl]. The metrical parameters of the
[(H2L)Zn] moiety are given in Table 2. The thermal ellipsoids are drawn at 50%
probability.
Fig. 7. The X-ray crystal structure of [{(HL)Ni]}CuCl2] and [{(HL)Ni]}ZnBr2]. The metrical parameters can be found in Table 2. The thermal ellipsoids are drawn at 50%
probability.

1536
A. Mustapha et al. / Polyhedron 30 (2011) 1530–1537
phenolates of a zinc Schiff base complex such as [(Salpr)Zn].
Displacement of a halide from these cannot occur as these tetra-
dentate ligands lack the essential additional donor. The motif
described here would seem to be somewhat rare [20] as reactions
of Schiff base ligands in which additional donors (e.g. salicylidene,
acetate) are present typically promote the formation of trimetallic
complexes [8,21–24]. However, it is likely that this process is pre-
vented here by the extended phenolate (O3) which encroaches on
the space required by a second [(H2L)Zn] unit.
again [6] the pKa of the phenols in these species is playing major
role in the complexes which form.
The relationship between the motifs displayed by [{(HL)-
Ni]}ZnBr2] and [{(HL)Zn]}ZnCl] was tested. An attempt to remove
a halide and deprotonate the pendant phenoxide (O3 Fig. 7) was at-
tempted using sodium hydroxide. Rather than rotate the phenox-
ide into the coordination sphere of the zinc, the complex
dissociated. An attempt to construct a monohalide by limiting
the stoichiometric amount of halide using [(HL)Ni]Br and zinc tet-
rafluoroborate was also unsuccessful, the major product obtained
from solution being [{(HL)Ni]}ZnBr2].
The three homometallic species of nickel copper and zinc dis-
play widely different behaviour. Each with precedence in Schiff
base chemistry [8]. The zinc species is probably the most easily
understood as homometallic species where ZnCl2 is complexed to
the two phenolates are known. The species reported here is a sub-
tle modification of this motif where the pendant phenolate (O3,
Fig. 6) rotates into the coordination sphere of the metal displacing
a halide. This species can be viewed as a precursor of the family of
trimetallic (L2M3) complexes best exemplified by nickel. These tri-
metallic species readily form the more rigid hexa- and heptaden-
tate Schiff base ligands [8] and with tetradentate ligands
crucially with acetate bridges [11]. Thus when we revert to the
use of halide precursors with nickel the motif can expand but in
a more dramatic manner. Although it continues to dimerise it does
so based on an Ni3 unit rather than around a central nickel. The
Having successfully placed a copper and zinc on the external
face of [(HL)Ni] it was of interest to see if a similar result could
be achieved with the ‘‘respective’’ copper and zinc complexes.
2+
The motifs displayed by copper namely [{(L)Cu}Cu]2 (Fig. 6)
[{(HL)Cu}2Cu]4+ [5,6] would suggest that it should be possible to
collect the phenolates in a manner which forms a chelating centre.
However consistent with our studies on the lanthanoids we failed
to isolate any dimetallic complexes derived from the nickel halides
or nickel acetate. The reactions with zinc chloride confirmed this
lack of reactivity as here we were able to isolate a charge separated
salt namely [(H2L)Cu]+ Et3NH+ [ZnCl4]2ꢀ (Fig. 8) which contained
both metals.
The predominant products formed in the reactions with
[(H2L)Cu] were [(H2L)Cu] itself. The crystallographic data base
however, contains three complexes based on a modified [(Sal-
pr)Cu] framework where different hydrated metal cations
(Cu(OH2)3, Co(OH2)3 and Mn(OH2)3) are chelated to the phenolate
donors [30]. These complexes display behaviour very similar to
those which we report here (e.g. [(HL)Ni}Ni(OAc)2], [(HL)Ni}CuCl2]
and [(HL)Ni}ZnBr2]) and are related to those we have attempted to
prepare here. This suggests that our problems with the preparation
of the copper complexes may rest with our synthetic design. In
contrast, the products obtained from the treatment of [(H2L)Zn]
with nickel and copper are [(H2L)Ni] and [(H2L)Cu], respectively.
This suggests that a metal replacement reaction had taken place
for the zinc precursor. In retrospect this is a possible problem with
the design of mixed metal complexes of this type when the second
metal centre being introduced onto the external sites of the com-
plex has a higher formation constant for the N4 donor set in the
interior of the ligand. In certain instances where the exchange rates
are favourable and the ligand is flexible, metal exchange can be ex-
pected to be a competing reaction. A related process is believed to
be central to the synthesis of the mixed nickel lanthanoid
complexes ([{(L)Ni}2Ln(HOMe)]+, Fig. 1) and responsible for the
2+
copper species [{(L)Cu}Cu]2 taken in conjunction with a similar
copper complex, [{(HL)Cu}2Cu]4+ [5,6], is also instructive in this
analysis. These two motifs show that the two [(H2L)Cu]
ligand can complex at single metal centre ([{(HL)Cu}2Cu]4+) or
individually act as a ligand to a copper centre which then dimerises
([{(L)Cu}Cu]22+). This latter pair of complexes confirm the
importance of the synthetic design to the synthesis of these
compounds.
It is interesting that none of the H3L complexes isolated here
make use of all three phenolate donors as a symmetric facially cap-
ping donor set. This contrasts with the behaviour of the other mul-
timetallic tri-phenolate ligands, HBTC [5,25,26], HAME [27] and
HAC [28], which do have this arrangement. However these are
hexadentate and to encapsulate a metal cation in an octahedral
environment all three phenolates are required. In an attempt to
show that it is possible to bring the three phenolate donors into
the coordination sphere of transition metals we prepared and iso-
lated the ferric complex [(L)Fe]. The crystals which were obtained
for this species refined poorly during crystallographic analysis.
However, although it is impossible to extract any metrical param-
eters, the gross structural features could be identified which con-
firm that the iron coordinates to the three phenolates at the
expense of a bond to one of the secondary amines. The motif is
similar to that of [(L)Al] [29].
removal of the zinc during the synthesis of [{(L)Ce}2-
lO2] from
[(H2L)Zn] [6].
The ability to synthesize homometallic complexes is relatively
straightforward as there can be no problem with site selection by
the metals being used. A real test of the durability of species such
as [(HxL)M] as ligand system would be the ability to build hetero-
nuclear species especially with other transition metals. Thus, treat-
ing H3L sequentially with nickel nitrate and the halides of either
copper and zinc respectively we were again able to retrieve com-
plexes which could be fully characterised (Fig. 7) and which sup-
port our view that we are interrogating the chemistry of a
metallo-ligand.
From the perspective of the external atom (Cu, Zn) both com-
plexes (Fig. 5) are four coordinate. The {(HL)Ni] moiety acts as a
didentate ligand operating via the adjacent phenolates (O1, O2).
The presence of a metal chelated into the phenolates would seem
to have a marginal effect on the encapsulated nickel. It is curious
that in neither complex does the pendant phenolate (O3) rotate
into the coordination sphere of the external metal. This is espe-
cially so considering the structure of [{(HL)Zn]}ZnCl] (Fig. 6). Once
Fig. 8. The X-ray crystal structure of [(H2L)Cu]+ Et3NH+ [ZnCl4]2ꢀ. The thermal
ellipsoids are drawn at 50% probability.

A. Mustapha et al. / Polyhedron 30 (2011) 1530–1537
1537
[2] V. Chandrasekhar, R. Azhakar, G.T.S. Andavan, V. Krishnan, S. Zacchini, J.F.
Bickley, A. Steiner, R.J. Butcher, P. Kogerler, Inorg. Chem. 42 (2003) 5989.
[3] S.R. Bayly, Z.Q. Xu, B.O. Patrick, S.J. Rettig, M. Pink, R.C. Thompson, C. Orvig,
Inorg. Chem. 42 (2003) 1576.
[4] C.A. Barta, S.R. Bayly, P.W. Read, B.O. Patrick, R.C. Thompson, C. Orvig, Inorg.
Chem. 47 (2008) 2280.
[5] C.A. Barta, S.R. Bayly, P.W. Read, B.O. Patrick, R.C. Thompson, C. Orvig, Inorg.
Chem. 47 (2008) 2294.
[6] A. Mustapha, J. Reglinski, A.R. Kennedy, Inorg. Chim. Acta 362 (2009) 1267.
[7] A. Mustapha, J. Reglinski, A.R. Kennedy, Inorg. Chem. Commun. 13 (2010) 464.
[8] A. Mustapha, K. Busch, J. Reglinski, A.R. Kennedy, T.J. Prior, Polyhedron 27
(2008) 868.
4. Concluding remarks
The systematic construction of complexes which utilise
[(HxL)M] as a ligand is a partial success. Metals such as nickel
and zinc have been successfully encapsulated. We are still unable
to achieve this aim with copper most likely due to the design of
the ligand and the preferred geometry of the metal centre. Placing
metals on the outer face of these ligands has also been a partial
success. Copper sits didentate and zinc sits both didentate and tri-
dentate to these metallo-ligands. However, there remains some
problems with controlling the ultimate positions of the metals
especially when using {(H2L)Zn] as a ligand.
[9] G.M. Sheldrick, Acta Crystallogr. A64 (2008) 112.
[10] L.J. Farrugia, J. Appl. Crystallogr. 32 (1999) 837.
[11] C.J. O’Conner, D.P. Freyberg, E. Sinn, Inorg. Chem. 18 (1979) 1077.
[12] J. Reglinski, S. Morris, D.E. Stevenson, Polyhedron 21 (2002) 2167.
[13] J. Reglinski, M.K. Taylor, A.R. Kennedy, Inorg. Chem. Commun. 9 (2006) 736.
[14] M.M. Levitsky, O.I. Schegolikhina, A.A. Zhdanov, V.A. Igonin, Y.E. Ovchinnikov,
V.E. Shklover, Y.T. Struchkov, J. Organomet. Chem. 401 (1991) 199.
[15] M. Murrie, H. Stoeckli-Evans, H.U. Gudel, Angew. Chem., Int. Ed. 40 (2001)
1957.
[16] S.T. Ochsenbein, M. Murrie, E. Rusanov, H. Stoeckli-Evans, C. Sekine, H.U.
Gudel, Inorg. Chem. 41 (2002) 5133.
[17] C. Fukuhara, K. Tsuneyoshi, K. Katsura, N. Matsumoto, S. Kida, M. Mori, Bull.
Chem. Soc. Jpn. 62 (1989) 3939.
[18] D. Ulku, L. Tatar, O. Atakol, M. Aksu, Acta Crystallogr., Sect. C 57 (2001) 273.
[19] J. Reglinski, S. Morris, D.E. Stevenson, Polyhedron 21 (2002) 2175.
[20] L. San Felices, E.C. Escudero-Adan, J. Benet-Buchholz, A.W. Kleij, Inorg. Chem.
48 (2009) 846.
[21] O. Atakol, L. Tatar, M.A. Akay, D. Ulku, Anal. Sci. 15 (1999) 199.
[22] C. Arici, M. Aksu, Anal. Sci. 18 (2002) 727.
Acknowledgments
AM would like to express his gratitude to Kano University of
Science and Technology, Wudil, Nigeria for financial assistance.
JR acknowledges the EPSRC national crystallography service at
the University of Southampton for data collection on three
compounds.
Appendix A. Supplementary data
CCDC 788756, 788757, 788758, 788759, 788760, 788761 and
788762 contain the supplementary crystallographic data for this
paper;. These data can be obtained free of charge via http://
www.ccdc.cam.ac.uk/conts/retrieving.html, or from the Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ,
UK; fax: (+44) 1223-336-033; or e-mail: deposit@ccdc.cam.ac.uk.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.poly.2011.03.004.
[23] L. Tatar, O. Atakol, D. Ulku, Acta Crystallogr., Sect. E 58 (2002) m83.
[24] X. Yang, R.A. Jones, Q. Wu, M.M. Oye, W.-K. Lo, W.-K. Wong, A.L. Holmes,
Polyhedron 25 (2006) 271.
[25] A. Sokolowski, J. Muller, T. Weyhermuller, R. Schnepf, P. Hildebrandt, K.
Hildenbrand, E. Bothe, K. Wieghardt, J. Am. Chem. Soc. 119 (1997) 8889.
[26] M.D. Snodin, L. Ould-Moussa, U. Wallmann, S. Lecomte, V. Bachler, E. Bill, H.
Hummel, T. Weyhermuller, P. Hildebrandt, K. Wieghardt, Chem. Eur. J. 5
(1999) 2554.
[27] Z. Xu, P.W. Read, D.E. Hibbs, M.B. Hursthouse, K.M. Abdul Malik, B.O. Patrick,
S.J. Rettig, M. Seid, D.A. Summers, M. Pink, R.C. Thompson, C. Orvig, Inorg.
Chem. 39 (2000) 508.
[28] J.E. Bollinger, J.T. Mague, C.J. O’Connor, W.A. Banks, D.M. Roundhill, Dalton
Trans. (1995) 1677.
References
[29] S. Lui, S.J. Rettig, C. Orvig, Inorg. Chem. 31 (1992) 5400.
[30] M. Nayak, R. Koner, H.-H. Lin, U. Florke, H.-H. Wei, S. Mohanta, Inorg. Chem. 45
(2006) 10764.
[1] H. Ohta, K. Harada, K. Irie, S. Kashino, T. Kambe, G. Sakane, T. Shibahara, S.
Takamizawa, W. Mori, M. Nonoyama, M. Hirotsu, M. Kojima, Chem. Lett.
(2001) 842.