The nickel, copper and zinc complexes of a potentially heptadentate nitrogen donor ligands

Article abstract of DOI:10.1016/j.poly.2017.03.050

The synthesis of the potentially heptadentate ligands, tris-(2-aminobenzylidene)amino-ethylamine (L³) and tris-(o-aminobenzyl)aminoethylamine (L⁴) are reported. Complexes of L³ with nickel, copper and zinc have been synthesised and characterised. However, the nickel and zinc compounds are observed to re-arrange during the procedures used to produce samples for X-ray diffraction analysis. In both cases aniline groups are found to migrate to give rise to unique but related coordinated polyamine species. A rational route which allows for the reduction of the imine function in L³ is presented giving rise to a heptadentate ligand (L⁴) which contains primary, secondary and tertiary amines. Having removed the reactive imine function, the synthesized nickel and copper complexes follow the expected synthetic and structural pattern with the nickel complex being observed to be octahedral and the copper(II) complex five coordinate. The zinc complex is in contrast different. As observed for L³, it is possible to generate simple ZnL^4? complexes but these are prone to oxidation and intramolecular cyclisation where an aniline nitrogen couples with the secondary amine to form a coordinated indazine.

Full text of DOI:10.1016/j.poly.2017.03.050

Accepted Manuscript
The nickel, copper and zinc complexes of a potentially heptadentate nitrogen
donor ligands
Abdullahi Mustapha, Graham Steel, John Reglinski, Alan R. Kennedy
PII:
S0277-5387(17)30248-6
DOI:
http://dx.doi.org/10.1016/j.poly.2017.03.050
POLY 12565
Reference:
Polyhedron
To appear in:
Received Date:
Revised Date:
Accepted Date:
2 March 2017
23 March 2017
25 March 2017
Please cite this article as: A. Mustapha, G. Steel, J. Reglinski, A.R. Kennedy, The nickel, copper and zinc complexes
of a potentially heptadentate nitrogen donor ligands, Polyhedron (2017), doi: http://dx.doi.org/10.1016/j.poly.
2017.03.050
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

The nickel, copper and zinc complexes of a potentially heptadentate nitrogen donor ligands.
Abdullahi Mustapha#, Graham Steelƒ, John Reglinski*ƒ and Alan R. Kennedyƒ
ƒ Department of Pure & Applied Chemistry
University of Strathclyde
Glasgow G1 1XL
# Department of Chemistry,
Federal University Dutse,
Jigawa State, Nigeria.
abdullahi.mustapha@fud.edu.ng
Tel: 0141-548-2349
Fax: 0141-552-4822
e-mail j.reglinski@strath.ac.uk
* Corresponding author
Key Words: Multidentate nitrogen donor ligands, Schiff base; reduced Schiff base;
polyamine; intramolecular migration; nickel; copper; zinc
Abbreviations
Tren: tris-(2-aminoethyl)amine
L1:
L2:
L3:
L4:
tris-(2-hydroxybenzylidene)aminoethylamine
tris-(2-hydroxybenzyl)aminoethylamine
tris-(2-aminobenzylidene)aminoethylamine
tris-((2-aminobenzyl)aminoethylamine
Abstract
The synthesis of the potentially heptadentate ligands, tris-(2-aminobenzylidene)amino-
ethylamine (L3) and tris-(o-aminobenzyl)aminoethylamine (L4) are reported. Complexes of
L3 with nickel copper and zinc have been synthesised and characterised. However, the nickel
and zinc compounds are observed to re-arrange during the procedures used to produce
samples for X-ray diffraction analysis. In both cases aniline groups are found to migrate to
give rise to unique but related coordinated polyamine species. A rational route which allows
for the reduction of the imine function in L3 is presented giving rise to a heptadentate ligand
(L4) which contains primary, secondary and tertiary amines. Having removed the reactive
imine function, the synthesized nickel and copper complexes follow the expected synthetic
and structural pattern with the nickel complex being observed to be octahedral and the
copper(II) complex five coordinate. The zinc complex is in contrast different. As observed
for L3, it is possible to generate simple ZnL3 complexes but these are prone to oxidation and
intramolecular cyclisation where an aniline nitrogen couples with the secondary amine to
form a coordinated indazine.

Introduction
Scheme 1
Multidentate phenolate ligands derived from tris-(2-aminoethyl)-amine (tren) and
salicylaldehyde (L1, L2) have been known for over fifty years (scheme 1) [1, 2]. The early
chemistry of these ligand systems was undoubtedly hindered by the inability to
crystallographically characterise the products as many of the complexes generated are
multimetallic in nature [2-14]. Indeed one of the fascinating aspects of this chemistry is the
myriad of motifs which can be produced by challenging a potentially heptadentate ligand with
transition metals which have a preference for low coordination numbers. With divalent
metals the primary complexes produced in combination with L1 and L2 are anionic and are
thus capable of acting as ligand themselves and this gives rise to a related series of
multimetallic and heterometallic complexes [2-14]. Many of these have interesting magnetic
properties [7, 9, 14]. Replacing the phenol residue with an aniline moiety is a simple way of
increasing the versatility of this family of ligands [15]. Multidentate nitrogen donor ligands
have been identified as central to the design and function of ruthenium complexes for water
oxidation catalysis and as models for nitrogen rich donor sites in metalloenzymes [16, 17]. In
the belief that the metal complexes of the amine analogues of L1 and L2 (figure 1) might also
have a rich and diverse chemistry we have sought to identify simple routes into their
synthesis. Our previous studies using L1 and L2 have shown that the divalent metals cations
such as nickel, copper and zinc generate a wide range of structural motifs and as such we have
adopted these metals again to explore the behaviour of these two heptadentate nitrogen donor
ligands (L3, L4 figure 1).
Figure 1. A schematic representation of tris-(2-aminobenzylidene)aminoethylamine (L3) and
tris-((2-aminobenzyl)aminoethylamine (L4).

Experimental
All experiments were carried out using standard apparatus and commercially available
chemicals. O-aminobenzaldehyde was prepared as previously reported [18]. NMR analysis
was carried out on a Bruker AMX 400 operating at 400 MHz for 1H for proton and 100 MHz
13
for C. Solid reflectance spectra (400 – 900 nm) were recorded on a Photonics CCD array
UV-Vis spectrophotometer. Mass spectra were recorded in house on a Thermo Finnigan
LCQDuo by electrospray ion trap (4.5 kW, 200 oC). Infra red spectra were recorded using an
A2 technologies ATR FT-IR spectrometer. Elemental microanalysis was carried out in house
on a Perkin Elmer 2400 CHN Analyser.
X-Ray measurements were conducted using a Nonius Kappa CCD diffractometer at
123 K using graphite monochromated Mo-Kα radiation. Crystals were coated in mineral oil
and mounted on glass fibers. The heavy atom positions were determined by Patterson
methods and the remaining atoms located in difference electron density maps. Full matrix
least-squares refinement was based on F2, with all non-hydrogen atoms anisotropic. While the
hydrogen atoms were mostly observed in the difference maps, they were placed in calculated
positions riding on the parent atoms except those involved in hydrogen bonding which were
refined isotropically. The structure solution and refinement used the programs SHELXL
SHELXL-97 [19] and the graphical interface WinGX [20]. A summary of the crystallographic
parameters is given in Table 1.
Synthesis of tris-((2-aminobenzylidene)-aminoethylamine (L3): o-amino-benzaldehyde
(0.66 g, 5.5 mmol) was refluxed with tris(2-aminoethyl)amine (0.3 g, 2.0 mmol) in methanol
(50 ml) for 90 minutes. The pale yellow solution produced was reduced in volume to 10 ml.
Diethyl-ether (20 ml) was added to obtain a white solid which was suction filtered and dried.
The solid was recrystallised by vapour diffusion of methanol into a chloroform solution.
Yield 67%. m.p. 136oC. Anal. Found: C, 70.26; H, 8.18; N, 20.92 % Calcd. for
1
C27H33N7.1/2CH3OH: C, 70.03; H, 7.48; N, 20.78 %. H NMR (400 MHz, CDCl3; δ): 8.2 (s,
1H, -HC=N), 7.1 (t, 1H, arom); 6.8 (d, 1H, arom); 6.6 (m, 2H, arom); 6.3 (br s, 2H, -NH2); 3.6
(t, 2H, -CH2-); 3.5 (s, MeOH); 2.8 (t, 2H, -CH2-). 13C-{1H} NMR (100 MHz CDCl3; δ): 165,
148, 134, 131, 118, 116, 115, 60, 56. ν/cm-1: 3231 (NH), 1639 (C=N), 1588 (C=C), 753
(arom). MS (ESI, m/e): 456 (100%),

Synthesis of tris-(2-aminobenzyl)aminoethylamine (L4): Tris-(2-aminoethyl)amine (2 g,
13.7 mmol) and 2-nitrobenzaldehyde (6.2 g, 41 mmol) were refluxed in methanol (50 ml) for
1 hour. The solution was allowed to cool and the orange solid, (L4A), collected by filtration
(Yield 89%).
L
4A was re-dissolved in methanol (50 ml) and treated with NaBH4 (1.5 g, 43 mmol) over a
1 hr period under the blanket of nitrogen. The yellow solution was stirred overnight. The
solvent was removed under reduced pressure to give an orange oil. An aqueous solution of
ammonium acetate (4g in 50 ml) was added to the oil and the mixture extracted with
chloroform (3 x 100 ml). The chloroform solution was dried over anhydrous Na2SO4, filtered
and the solvent reduced in volume to give a yellow solid. The yellow solid (L4B), was
collected by filtration (Yield 68%).
L4B produced above was suspended water (175 ml) and conc. hydrochloric acid (0.5 ml)
in a 500ml round bottom flask containing FeSO4.7H2O (105 g, 0.38 mol). The mixture was
heated to 90oC followed by the addition of conc. ammonia solution (30 ml). The black slurry
was stirred while being allowed to cool (~ 1 hr). The mixture was extracted with chloroform
(3 x 100 ml), dried over anhydrous Na2SO4, filtered and the solvent removed to give a viscous
yellow oil (L4). Yield 54%.
Alternatively the L4B produced above was dissolved in propan-2-ol and palladium on carbon
(10%) (200 mg) was added. The suspension was heated with stirring to 50oC and hydrazine
hydrate (15 ml) added dropwise. The reaction mixture was heated to reflux for 2 hours under
nitrogen before being filtered and the solvent removed in vacuo to yield L4 as a viscous
yellow oil. Yield 75%.
L4A: Anal. Found: C, 59.45; H, 5.48; N, 18.34 % Calcd. for C27H27N7O6: C, 59.44; H, 4.99;
1
N, 17.94 %. H NMR (400 MHz, CDCl3; δ): 8.6 (s, 1H, -HC=N), 8.0 (dd, 1H, arom), 7.9 (dd,
1H, arom); 7.6 (dt, 1H, arom); 7.5 (dt, 1H, arom); 3.8 (t, 2H, -CH2); 3.0 (t, 2H, –CH2-). 13C-
{1H} NMR (100 MHz CDCl3; δ): 157, 133, 131, 130, 129, 124, 69, 55, 31. ν/cm-1 1630,
1575, 1525, 1340, 740. MS (ESI, m/e): 546 ([M+H]+, 100%),
1
L4B: H NMR (400 MHz, CDCl3; δ): 7.9 (d, 1H, arom), 7.6 (d, 1H, arom); 7.5 (t, 1H, arom);
7.4 (t, 1H, arom); 4.0 (s, 2H, -CH2-); 2.7 (t, 2H, -CH2-), 2.6 (t, 2H, –CH2-). 13C-{1H} NMR

(100 MHz CDCl3; δ): 149, 136, 133, 131, 128, 124, 54, 50, 47. ν/cm-1 1520, 1340, 725. MS
(ESI, m/e): 552 ([M+H]+, 100%),
1
L4: H NMR (400 MHz, CDCl3; δ): 7.3 (br s, 2H, arom), 7.0 (d, 1H, arom); 6.6 (d, 1H,
arom); 4.5 (br s, 2H, -NH2); 3.7 (s, 2H, -CH2-); 2.6 (br s, 2H, -CH2-), 2.5 (br s, 2H, –CH2-) 1.5
(br s, 2H, -NH). 13C-{1H} NMR (100 MHz CDCl3; δ): 147, 129, 128, 124, 117, 115, 54, 53,
47. MS (ESI, m/e): 462 ([M+H]+, 100%),
Schiff Base Chemistry
Synthesis of nickel, copper and zinc complexes of L3: L3 (0.05 g, 0.11 mmol) dissolved in
methanol (15 ml) was treated with 2 mol equivalents of the relevant metal halide. The
mixture was stirred at 50oC for 15 minutes, allowed to cool and filtered.
[NiL3] 2Cl: Anal. Found: C, 52.79; H, 6.08; N, 15.41% Calcd. for C27H33N7NiCl2.2H2O: C,
52.21; H, 6.00; N, 15.78 %. FTIR [(ν/cm-1 (KBr)]: 2925 (-NH2), 1610 (-C=N), 1500 (C=C),
760 (arom). MS (ESI, m/e): [M-H]+ 512 (25%), [M+R+H]+ 615 (25%) [M+2R+H]+ 718
(100%), [M+3R+H]+ 823 (10%), [M-R+H]+ 409 (10%) [M-2R+H]+ 306 (25%), R = anilino
group. These are observed to migrate during the mass spectroscopic analysis
[NiL3] 2I: Anal. Found: C, 44.83; H, 5.11; N, 14.65% Calcd. for C27H33N7I2Ni: C, 42.22; H,
4.33; N, 12.76 %. FTIR [(ν/cm-1 (KBr)]: 2925 (-NH2), 1610 (-C=N), 1500 (C=C), 760
(arom). [M-H]+ 512 (10%), [M+3Na-2H]+ 581 (35%) [M+R-2H]+ 684 (15%), [M+2R-2H]+
787 (100%). R = anilino group. These are observed to migrate during the mass spectroscopic
analysis. The poor agreement of the microanalysis is believed to stem from the partial
decomposition of the product.
Crystals suitable for X-ray diffraction were obtained from the very slow evaporation of a
methanol solution of [(L3)Ni][I2]. The crystals were taken from an oily residue in the base of
the flask which mitigated against further analysis.
[CuL3] 2Cl: Anal. Found: C, 53.15; H, 5.98; N, 15.49% Calcd. for C27H33N7CuCl2.H2O: C,
53.33; H, 5.80; N, 16.12 %. FTIR [(ν/cm-1 (KBr)]: 2925 (-NH2), 1615 (-C=N), 1505 (C=C),

760 (arom). MS (ESI, m/e): [M-H]+ 517 (25%), [M+2R-H]+ 723 (45%) [M-2R+2H]+ 310
(100%), λmax = 595 nm (solid)
[ZnL3][ZnCl4]: Anal. Found: C, 42.51; H, 4.98; N, 12.60
%
Calcd. for
C27H33N7Zn2Cl4.2H2O: C, 42.43; H,4.87; N,12.83 %. FTIR [(ν/cm-1 (KBr)]: 2800-3200 (-
NH2), 1610 (-C=N), 1495 (C=C), 760 (arom). MS (ESI, m/e): [M-H]+ 518 (30%), [M-2H-
2R]+ 310 (100%). 1H NMR (400 MHz, CDCl3; δ): 7.02 (s, 1H, -CH=N), 7.00 (s, 1H, -CH=N),
6.92 (d, 1H, arom), 6.9 (s, 1H, -CH=N), 6.88 (d, 1H, arom); 6.6-6.4 (m, 10H, arom); 3.7 (br s,
2H, -NH2); 3.3 (br s, 4H, -NH2); 2.7-2.5 (m, 12H, -CH2-).
Zinc crystals were grown by slow evaporation of the mother liquors over 24 hours leading to
a ligand rearrangement.
[ZnL3][ZnCl4]: Anal. Found: C, 42.51; H, 4.98; N, 12.60
%
Calcd. for
C27H33N7Zn2Cl4.2H2O: C, 42.43; H,4.87; N,12.83 %. FTIR [(ν/cm-1 (KBr)]: 3440 (-NH2),
1650 (-C=N), 1495 (C=C), 760 (arom). MS (ESI, m/e): 518 (100%), 259 (55%).
Secondary Amine Chemistry
Synthesis of (tris-(2-aminobenzyl)-aminoethylamine) nickel) nitrate: L4 (0.07 g, 0.16
mmol) was dissolve in methanol (20 ml). The mixture was stirred at 50oC followed by
addition of NiNO3.6H2O (0.045 g, 0.16 mmol). The brown solution was stirred for a further
10 minutes, allowed to cool and filtered. Crystals (deep lilac) were grown by slow
evaporation of the saturated methanol solution. Yield 73%. Anal. Found: C, 50.54; H, 6.05;
N, 19.57 % Calcd. for C27H39N9NiO6: C, 50.33; H, 6.10; N, 19.56 %. ν/cm-1 2960, 1600,
1500, 1380, 760. MS (ESI, m/e): 518 (100%), λmax (solid) 550 (shoulder) 940.
Synthesis of (tris-(2-aminobenzyl)-aminoethylamine)-nickel) chloride: L4 (0.053 g, 0.11
mmol) and nickel chloride hexahydrate (0.023 g, 0.1 mmol) were dissolved separately in the
minimum amount of ethanol. The solutions were filtered and the solution of L4 added to the
nickel solution. The mixture was stirred for 1 hr during which time the solution turned green
and a yellow/green precipitate formed. This was collected by filtration and washed with
diethyl-ether. Yield 50%. Anal. Found: C, 47.66; H, 5.96; N, 14.92%. Calcd. for

C27H39N9NiCl2.5H2O C, 47.60; H, 7.25; N, 14.39 %. ν/cm-1 2960, 1600, 1500, 1310 (NO3)
MS (ESI, m/e): 259 (M2+ 100%), 518 (M+ 100%), 554 ([M+Cl]+ 65%), 593 ([M+Cl+K]+
64%), 632 ([M+Cl+2K]+ 40%), λmax (solid) 620 nm.
Synthesis of (tris-(2-aminobenzyl)-aminoethylamine)-copper) nitrate: L4 (0.050 g, 0.11
mmol) and copper nitrate trihydrate (0.025 g, 0.1 mmol) were dissolved separately in the
minimum amount of acetonitrile. The solutions were filtered and the solution of L4 added to
the copper solution. The mixture was stirred for 1 hr during which time the solution turned
dark green and a green precipitate formed. This was collected by filtration and washed with
diethyl-ether. Yield 75%. Anal. Found: C, 45.26; H, 5.29; N, 17.80%. Calcd. for
C27H39N9CuO6.4H2O C, 44.97; H, 6.57; N, 17.48 %. ν/cm-1 2960, 1600, 1500, 1380, 1310
(NO3), 750. MS (ESI, m/e): 523 (M+ 25%), 519 ([M-4H]+ 100%), 5154 ([M-8H]+ 64%),
([M+Cl+2K]+ 40%), λmax (solid) 620 nm.
Synthesis of (tris-(2-aminobenzyl)-aminoethylamine)-copper) chloride: L4 (0.053 g, 0.11
mmol) and copper chloride dihydrate (0.023 g, 0.1 mmol) were dissolved separately in the
minimum amount of ethanol. The copper solution was added to the solution of L4. The
mixture was stirred for 30 mins during which time the solution turned deep green and a
precipitate formed. This solid was collected by filtration and washed with diethyl-ether.
Yield 40%.
Anal. Found: C, 50.13; H, 6.47; N, 13.87
%
Calcd. for
C27H39N7CuCl3.EtOH.H2O: C, 50.00; H, 6.94; N, 14.07 %. ν/cm-1 2925, 1620, 1495, 1445,
755. MS (ESI, m/e): 555, 557 (100%) [M+OMe]+, 571, 573 (100%) [M+OEt]+, λmax (solid)
665.
Crystals (green) of [CuL4H] 3Cl were grown by diffusion at the interface of a propanol
solution of copper chloride with that of L4 also in propanol.
Synthesis of tris-(2-aminobenzyl)-aminoethylamine)-zinc salts: L4 (0.053 g, 0.11 mmol)
and zinc salt (chloride/ nitrate, 0.1 mmol) were dissolved separately in the minimum amount
of ethanol. The solutions zinc solutions were added to the solution of L4. The mixture was
stirred for 30 mins during which time the creamy solution turned yellow and a precipitate
formed. This solid was collected by filtration and washed with diethyl-ether.

[ZnL4] 2NO3: Yield 25%. Anal. Found: C, 46.11; H, 5.15; N, 17.98%. Calcd. for
C27H39N9O6Zn.3H2O C, 46.00; H, 6.43; N, 17.88 %. ν/cm-1 2960, 1620, 1480, 1315 (NO3),
1035, 755. MS (ESI, m/e): 524 (M-H+ 100%).
[ZnL4] ZnCl4: Yield 25%. Anal. Found: C, 44.30; H, 4.74; N, 13.35%. Calcd. for
C27H39N7Zn2Cl4, 44.17; H, 5.35; N, 13.35 %. ν/cm-1 2960, 1620, 1480, 1035, 750 524 MS
(ESI, m/e): 524 (M-H+ 30%), 520 (M-5H+ 30%), 516 (M-9H+ 30%).

“NiL2” H3O+ 3I
“ZnL2” [ZnCl4]2-
NiL4 2NO3
CuHL4 Cl3
0.5C3H7OH
L3
Empirical Formula
FW
Crystal system
Space Group
a / Å
C27H35N7O
473.62
trigonal
R-3
12.9828(4)
12.9828(4)
C21H35I3N6NiO2
842.96
C27H36Cl4N7O1.50Zn2 C27H39N9NiO' C31.5H52Cl3CuN7O1.5
C27H35N9O6Zn
647.01
Monoclinic
P21/c
755.17
Monoclinic
P21/n
644.38
Monoclinic
P21/c
722.69
Monoclinic
P21/n
Triclinic
P-1
9.8973(3)
10.8803(5)
14.5836(6)
9.4360(6)
16.0009(8)
20.7248(11)
10.2682(3)
15.9820(5)
18.1017(5)
9.5888(8)
27.554(4)
14.073(2)
18.053(3)
17.308(3)
20.711(3)
b / Å
c / Å
26.1704(10)
90.00
69.044(4)
79.799(3)
90.00
90
90
90
94.408(7)
α / deg
β / deg
98.967(2)
90.552(2)
93.715(10)
90.00
89.366(3)
2
1441.05(10)
3.915
13150
6889
90.00
4
3090.9(3)
1.935
11637
6060
3419
399
0.1410
0.0634
1.038
90
4
90
4
90
8
120.00
6
3820.1(2)
0.079
9712
1854
1567
122
γ / deg
Z
V / Å3
2970.47(15)
0.710
31226
6468
3710.4(8)
3.096
12064
5453
2674
420
0.266
0.089
1.006
6452.2(18)
0.814
36271
10091
4224
µcalc / mm-1
No. Rflns Measd
No. Unique reflns
No Observed
No. parameters
6047
329
0.072
0.028
1.063
4227
439
0.081
0.039
0.934
780
Ra (I > 2
σ(I) )
0.078
0.208
0.954
0.1338
0.0513
1.083
Rwb (all reflns)
GOF
Crystallographic data. The crystallographic data for CuHL4 Cl3.0.5C3H7OH is poor, with the structure not refining to an acceptable solution. The cell
dimensions etc obtained are however, included here for completeness.

Results and Discussion.
Scheme 2
L3 was prepared from the tris(aminoethyl)amine and o-amino-benzaldehyde (scheme 2). It
was isolated as an off-white solid which could be crystallised from chloroform-methanol
(figure 2). Consistent with Garcia et al. [15], structural analysis of L3 reveals that the three
anilino groups are folded inwards from the apical nitrogen such that the three aryl rings are
placed in close proximity to one another. The distance between the rings (2.902 Å) estimated
by distance between H5 and C1 suggests that an interaction of the ring hydrogens with the π-
system has an influence in stabilising the structure. The anilino hydrogens are also directed
towards the Schiff base nitrogens (N2-H2N1; 2.027 Å) thus forming a network of three
internal hydrogen bonds.
Figure 2. The X-ray crystal structure of tris-((2-aminobenzylidene)-aminoethylamine. The thermal
ellipsoids are drawn at 50% probability. This structure has been previously reported by Garcia et al.
[15]. The structure was re-determined for this study because of the re-arrangements discussed below
The synthesise of the reduced multidentate nitrogen ligands such as L1 (Scheme 1) typically
commences by condensing the relevant aldehyde with tren. The resulting Schiff base is then
reduced with sodium borohydride. Consistent with the report of McKee et al, we could

synthesis the required Schiff base using tren and o-amino-benzaldehyde (scheme 2) [21].
However, the subsequent total reduction of the imine and nitro groups using either
borohydride or aluminium hydride was unsuccessful. We thus altered the synthetic strategy.
The Schiff base derived from o-nitro-benzaldehyde with tren was first produced (Scheme 3).
This was subsequently reduced using sodium borohydride to form the corresponding
secondary amine. Subsequently, the nitro group was reduced using acidic ferrous sulfate or
hydrazine over the palladium to produce the desired polyamine ligand (L4) which was
obtained as a viscous oil [16].* The reaction is presented in the experimental as three
independent steps and during the development of the synthesis of the polyamine we took care
to isolate the various intermediates. In practice, however, it is possible to carry out the first
two steps of the synthesis (coupling and reduction) in sequence. Similarly the reduction of
the nitro group to give the desired polyamine can be achieved using the crude liquors obtained
by simply taking the products of imine reduction to dryness.
Scheme 3
The Schiff base chemistry of L3
The halides of nickel, copper and zinc were treated with L3 under conditions typically
employed (methanol ~65 oC for 2 hrs.) for the formation of the related salicylaldehyde Schiff
base species [6, 7]. The mass spectra of the isolated complexes are dominated by ions
associated with the parent species (ML3) and their fragmentation/recombination products. We
could find no evidence to support the formation of multimetallic species at this juncture (c.f.
M3(L1)2, M = Ni, Zn; Cu4(L1)2) [6]. Vibrational spectroscopy of the three complexes
(coincident amide, imine and aromatic bands) indicates that the structures of these products
are of high symmetry and related. It was possible to record the 1H NMR spectrum of the
ZnL3 which shows that there are three imine environments (δ7.00, δ7.02, δ6.90). However,
the chemical shift difference is not sufficient to suggest that any of these remain
uncoordinated. The region assigned to tren is broad but compact indicative that the imines are
all coordinated. Thus we propose that ZnL3 is a simple six coordinate species (c.f. Zn3(L1)2)
[6].

Figure 3. The X-ray crystal structure of the isolated adduct of nickel iodide and L3. The structure is
presented to clearly show the fused aniline moieties which result from Schiff base cleavage/migration.
Selected bond lengths (Å): Ni-N1, 2.186(3); Ni-N2, 2.087(3); Ni-N3 2.018(3); Ni-N4, 2.121(3);
Ni-N5, 2.093(3); Ni-N6, 2.123(3). Bond angles (o) N1-Ni-N2, 81.95(11); N1-Ni-N3,
170.28(11); N1-Ni-N4, 106.19(11); N1-Ni-N5, 87.43(11); N1-Ni-N6, 88.05(11); N2-Ni- N3,
88.58(11); N2-Ni-N4, 171.74(11); N2-Ni-N5, 98.92(11); N2-Ni-N6, 100.65(10), N3-Ni-N4,
83.24(11); N3-Ni-N5, 91.97(11); N3-Ni-N6, 95.83(11); N4-Ni-N5, 80.34(11); N4-Ni-N6,
81.34(10); N5-Ni-N6, 159.08(11). The thermal ellipsoids are drawn at 50% probability.
Previous studies clearly show that it is difficult to predict the structures of the metal
complexes of heptadentate ligands such as those discussed here. Thus, in an effort to gain
greater insight into the structure of these ML3 species we extending the scope of the study to a
range of salts. From the reactions with nickel iodide we were able to obtain suitable crystals
for analysis by diffraction methods (figure 3). The structure obtained is, however, not
consistent with that expected for a simple ML3 complex. It would seem that the protracted
crystallisation process has allowed the ligand to both hydrolyse and re-arrange. Not only is
one of the aniline units lost completely a second is transferred to an adjacent aniline group,
probably via an intramolecular reaction, giving rise to the formation of a coordinated aniline-
methylidene-aniline moiety. The resulting perturbation of the number (seven to six) and
positioning of the aniline donors facilitates the movement of the nickel into the coordinating
environment of all four donors on tren (c.f. NiL2). [11, 13]. Migrations and rearrangements of
this type go some way to explaining the behaviour of the parent species in the mass
spectrometer [22, 23] where ions with masses consistent with the gain and loss of aniline units
are freely seen. Although not as prominent a similar degradation process has been reported
for the nickel salts of L1 where the phenolate arms are lost [6]. Consistent with the
observation reported here, the reaction of the salicylaldehyde species take time to come to
completion.

In the reactions involving zinc chloride a second product crystallised directly from the cooling
mother liquors (figure 4). On analysis by XRD methods we observed that the ligand (L3) has
re-arranged within the lifetime of the crystallisation process. Consistent with the nickel
example above (figure 3) an intermolecular migration has occurred whereby the aniline group
migrates to replace the nitrogen of tren in the Schiff base linkage. However, in this instance a
second coupling takes place and in-so-doing forms a tetracyclic ring system fused to a
nitrogen donor on tren (N4, figure 4). As the reaction proceeds the zinc migrates deep into
the tren moiety maintaining a coordination number of 5.
Figure 4. Left: The X-ray crystal structure of the isolated adduct of zinc chloride and L3. The
structure is presented to clearly show the fused tetracyclic component of the ligand which results from
Schiff base cleavage/migration. Right: A schematic representation of the fused aniline moieties. The
solvent of crystallisation, protons and [ZnCl4]2- have been omitted for clarity. The thermal ellipsoids
are drawn at 50% probability.
In both instances (nickel and zinc) we have evidence that the primary complex (ML3) forms
with ease. Structural analysis of the ligand itself (re-determined, figure 2) clearly shows that
the components are positioned in line with the expected condensation reaction [15]. Thus,
migration must be occurring as a result of the placement of the Lewis acid at the various
donors. Studies on the related Schiff base compounds formed between tren and
pyridinecarboxaldehyde provide some insight into these rearrangements. Here the three imine
moieties fuse to form complex macrocycles and polycyclic systems [24-26]. However, in
these cases only the three nitrogen centres of the tren moiety are involved. In our system the
nitrogen donor has been raised from the aryl ring (pyridine vs aniline) introducing additional
nucleophiles into the reaction profile. Thus instead of crosslinking between the imine groups,
the rearrangement includes the aniline nitrogen. For nickel we see the formation of one new
linkage at N2 (figure 3). However for zinc a second crosslink forms and we arrive at the

fused tetracyclic species (figure 4). Our previous attempts to generate heptadentate N4S3
ligand system based on tren and 2-(tert-butylthio)benzaldehyde were curtailed due to the
limited stability of the imine group within the metal complexes studied [27]. The
salicylidene-tren systems are more robust but even these, when metallated (e.g. nickel) have
demonstrated a tendency to degrade all-be-it over a prolonged time period [6]. In retrospect is
thus unsurprising that the nitrogen donor systems reported here display behaviour
intermediate between that observed for the phenolates and thio-ethers.
Secondary amine chemistry (L4).
The reaction of tris-((2-aminobenzyl)aminoethylamine (L4) with nickel nitrate gives rise to a
purple material which could be crystallographically characterised. X-ray analysis revealed
this to be the desired six-coordinated complex (figure 5) where the nickel sits deep in the
hepta-dentate ligand coordinating to the tertiary amine (N2 figure 5) and the three secondary
amines (N1, N3, N4). The coordination sphere is then completed via the use of two of the
anilino groups. Thus the observed structure of this tripodal anilino- ligand follows that of the
related tripodal phenolate (table 1) in that the geometry at nickel is best described as distorted
octahedral [11, 13]. We were unable to grow crystals of the corresponding chloride species.
However, spectroscopic and elemental analysis support the view that the nickel cation
([Ni(L4)]2+) in this species is, as expected, the same as that formed in the reaction with nickel
nitrate.
The reaction of copper chloride and nitrate with L4 produced products which vibrational and
mass spectroscopies again confirm that they contain the same complex cation ([Cu(L4)]2+).
These products steadfastly refused to crystallise using traditional recrystallisation procedures
and we eventually grew a poor quality crystal by the slow diffusion of a solution of copper(II)
chloride in propanol into a solution of L4 in propanol (figure 5). The resulting structure did
not refine to an acceptable point and as such the data can only be used to confirm the gross
structure of the complex i.e. that the adduct is again a simple 1:1 complex. Once again this
motif follows the established chemistry of the phenolate analogue where the copper sits in a
five coordinate environment using the secondary and tertiary amines derived from tren as the
primary donors. In this instance only one of the aniline groups are incorporated into the metal
coordination sphere. Progressing to zinc we obtained yellow complexes which analysed for
[ZnL4]2+. Although we were unable to confirm a structure using X-ray methods, vibrational

and mass spectroscopies indicate that the [ZnL4]2+ motif is iso structural with that for
[CuL4]2+. This behaviour is in line with the corresponding chemistry of the secondary amine
tren-based, N4O3-donor ligands (L2).
Figure 5. The molecular structures of left: [Ni(L4)] 2NO3 and right: [Cu(HL4)] 3Cl. Three chlorides
are found in the lattice of [Cu(HL4)] and consequently it is believed that one of the aniline
moieties must be protonated. The nitrates and chlorides are not shown for clarity. The key metrical
parameter for [Ni(L4)] 2NO3 can be found in table 1. The thermal ellipsoids are given at 50%
probability. The crystallographic data for CuHL4 Cl3.0.5C3H7OH is poor, with the structure not
refining to an acceptable solution. The picture shown serves only to demonstrate the connectivity of
the atoms.
trans N2-M-D M-N2
M-N (sec amines)
M-D (D = NH2, O)
2.145, 2.205
[Ni(L4)]2+
167.77
175.11
2.103 2.135, 2.171, 2.154
2.102 2.138, 2.062, 2.128
D = NH2 (aniline)
[Ni(L2)]2+
2.189, 2.005
D = OH (phenolate)
Table 1 The metrical parameters for [Ni(L4)]2+ and [Ni(L2)]2+. The distances associated with the tren
moieties would seem to be conserved between the two ligand systems (N4O3; N7) [11, 13]. The
crystallographic data for CuHL4 Cl3.0.5C3H7OH is poor, with the structure not refining to an
acceptable solution. It is not possible to quote accurate metrics for this species.
The complexes of L4 were all prepared by the addition of a metal salt to ligand without
heating in such a manner that sufficient product precipitated quickly from the mother liquors.
This crude approach was adopted as there was evidence that the complexes did not survive
excessive heating in solution even at modest temperatures (e.g. methanol for 1 hr). The
products formed under these reactions gave inconsistent results, but mass spectroscopic
evidence shows clear signs that the molecular weight of the ions was lower than desired.
During our attempts to isolate crystals for analysis by X-ray diffraction methods, compounds
were re-dissolve in a minimum amount of methanol and the samples left to evaporate. This

methodology was successful in the isolation of the crystals of [NiL4]2+ (figure 5). However,
although we were able to isolated crystals suitable for X-ray diffraction from the solution
containing [ZnL4] 2[NO3], the complex isolated for analysis was one in which the ligand (L4)
had rearranged (figure 6). Here one of the aniline moieties and secondary amines come
together in a ring closing reaction which gives rise to a coordinated indazole (figure 6). The
transformation requires the loss of two mole equivalents of hydrogen and reduces the
denticity of the ligand from hepta- to hexa- dentate.
Figure 6 The XRD structure of the material obtained from the slow evaporation of [ZnL4] 2[NO3]
from methanol. The structure is present in a manner which shows the formation of the indazole ring
The nitrates are not shown for clarity. The thermal ellipsoids are given at 50% probability.
Ring closures of this type have been reported before. Ovalle et al. reported on the attempted
synthesis of L4 using a synthesis similar to the one reported here [28]. In their study they
attempted to purify the polydentate amine using column chromatography (silica). However
instead of isolating L4 they obtained the corresponding tris-indazole in good yield. The
mechanism of ring closure in this instance has not been rigorously investigated. Previous
reports on tetradentate N4 secondary amine and Schiff base complexes has shown a tendency
for the anilino group to deprotonate in the presence of metals to form a coordinated amide
[29]. We were hoping to use chemistry of this nature to ultimately activate the aniline and
coax the polydentate ligand to bridge and form multimetallic complexes. Instead the
deprotonation of the aniline in close proximity to the coordinated secondary amine would
seem to initiate oxidative ring closure within the crowded coordination sphere. The ease with
which this ring closure reaction takes place goes some way to explain why the synthetic
protocol (direct precipitation; no heating) employed here for the isolation of the primary
complexes was found to be the most suitable. It also goes some way to explain the
discrepancy in the mass spectra of the nickel complexes compared with the corresponding

copper complexes. The nickel complex, by virtue of its better stability gives rise to a parent
peak at m/e 518. In contrast the copper complex gives rise to a plethora of signals between
m/e 512 (tris-indzole) and 520. The latter is consistent with the sequential loss of H2 during
ligand rearrangement.
Concluding remarks
Our studies on heptadentate ligands (N4O3, N4S3, N4N3) is complete with this partially
successful report on the synthesis on potentially heptadentate ligands. The general absence of
tren based polyamine systems in the literature has always presented itself as a intriguing
omission especially as the tetradentate systems have been know for many years and that
studies on high dentictity salicylaldehyde species were also reported over 60 years ago [1,
30]. The study here finally shows that it is possible to generate these species. While their
primary coordination chemistry is in line with previous studies regrettably their stability is
poor. The placement of an amine adjacent to a Schiff base linkage was, in retrospect,
expected to give rise to intra-molecular chemistry (c.f the tren Schiff base species of
pyridinecarboxaldehyde [24-26]). However reduction of the Schiff base linkage normally
give rise to more robust species and what was not anticipated was that in contrast to the
salicylaldehyde motif that the aniline based species were not completely passivated by the
formation of secondary amines.
Footnotes
* There is evidence that this material is ultimately a yellow solid.
Supporting information
Details of the X-ray crystal structure determinations may be obtained from the Director,
CCDC, 12 Union Road, Cambridge, CB2 1EZ, UK (fax +44-1223-336033; e-mail
deposit@ccdc.cam.ac.uk or www:http://ccdc.cam.ac.uk) on request quoting the depository
numbers CCDC 1535536-1535541
Acknowledgments

JR acknowledges the EPSRC national service at Southampton for X-ray crystallography. AM
would like to express his gratitude to Kano University of Science and Technology, Wudil,
Nigeria for financial assistance.
Reference
1
2
J.A Broomhead, D.J. Robinson Aust. J. Chem. 21, 1365 – 1367 (1968)
S. Liu, L. Gelmini, S.J. Rettig, R.C. Thompson, C. Orvig J. Amer. Chem. Soc., 114,
6081–6087 (1992)
3
4
P. Wei, D.A. Atwood J. Organomet. Chem. 563, 87–93 (1998)
C. Gedye, C. Harding, V. McKee, J. Nelson, J. Patterson J. Chem. Soc., Chem.
Commun., 392-394 (1992)
5
6
7
S.K. Chandra, P. Chakraborty, A. Chakravorty, J. Chem. Soc., Dalton Trans., 863-869,
(1993)
A. Mustapha, K. Busch, J. Reglinski, A.R.Kennedy and T.J.Prior. Polyhedron 27, 868-
878 (2008).
A. Mustapha, J. Reglinski, A.R. Kennedy, D.R. Armstrong, J. Saßmannshausen, M.
Murrie Inorg. Chem. 49, 5350-5352 (2010).
8
9
V.K. Bhardwaj, A. Singh. Inorg. Chem., 53, 10731–10742 (2014)
S.R. Bayly, Z. Xu, B.O. Patrick, S.J. Rettig. M. Pink, R.C. Thompson, C. Orvig Inorg
Chem. 42, 1576–1583 (2003)
10 J. Reglinski, M.K. Taylor, A.R. Kennedy, Inorg. Chem. Commun 9, 736-739 (2006)
11 A. Mustapha, J. Reglinski, A.R. Kennedy Inorg Chim. Acta 362, 1267-1274 (2009)
12 A. Mustapha, J. Reglinski, A.R. Kennedy Inorg. Chem. Commun 13, 464-467 (2010)
13 A. Mustapha, C. Busche, J. Reglinski, A.R. Kennedy. Polyhedron. 30, 1530-1537,
(2011).
14 J-L. Liu, J-Y. Wu, Y-C. Chen, V. Mereacre, A.K. Powell, L. Ungur, L.F. Chibotaru, X-
M. Chen, M-L. Tong. Angew Chem. Int. Ed. 53, 12966–12970, (2014)#
15 M.E. García, S. Bernès, N.P. Rodríguez, P.E. Martínez Acta Crystallogr. E. 66, o3054-
o3055, (2010).
16 J. J. Concepcion, J.W. Jurss, M.K. Brennaman, P.G. Hoertz, A.O.T. Patrocinio, N.Y.M.
Iha, J.L. Templeton, T.J. Meyer Acc. Chem. Res. 42, 1954-1965 (2009)
17 A. Messerschmidt, R. Huber Eur. J. Biochem 187, 341-352 (1990)
18 L.I. Smith, J.W. Opie, Org. Synth. Coll. 3, 56-58 (1955)

19 G.M. Sheldrick, Acta Cryst. A64 (2008) 112
20 L.J. Farrugia, J. Appl. Crystallogr. 32 (1999) 837
21 V. McKee, G.G. Morgan, J. Nelson Acta Cryst E. E62, o3747-o3749, (2006).
22 D. Mishra, S. Naskar, R.J. Butcher, S.K. Chattopadhyay, Inorg. Chim. Acta. 358, 3115–
3121 (2005)
23 A. Knödler, K. Hübler, T. Sixt, W. Kaim, Inorg. Chem. Commun. 3, 182-184, (2000)
24 D.A. Durham, F.A. Hart, D. Shaw, J. Inorg. Nucl. Chem. 29, 509-516, (1967)
25 L.T. Taylor, D.H. Busch, Inorg. Chem. 8, 1366-1371,
26 J.K. Clegg, J. Harrowfield, Y. Kim, Y.H. Lee, J-M. Lehn, W.T. Lim, P. Thuery, Dalton
Trans. 41, 4335-4357, (2012).
27 G. Steel, A. Mustapha, J. Reglinski, A. R. Kennedy. Polyhedron 67 (2014) 360-367.
28 S. Ovalle, P. Elizondo Martínez, N. Pérez Rodríguez, S. Bernès, M. Flores-Alamo Acta
Cryst. E 68, o1879-o1880 (2012).
29 G. Brewer, J. Jasinski, W. Mahany, L. May, S. Prytkov. Inorg.Chim.Acta. 232, 183-187,
(1995),
30 F. Lions, K.V. Martin. J. Amer. Chem. Soc. 80, 3858 – 3865, (1958)

The synthesis of the potentially heptadentate ligands, tris-(2-aminobenzylidene)amino-
ethylamine and tris-(o-aminobenzyl)aminoethylamine are reported. Complexes of nickel
copper and zinc have been synthesised and characterised. However, some compounds are
observed to re-arrange leading to curious coordinated polyamine species.