Chelating properties of EDTA-type ligands containing six-membered backbone ring toward copper ion: Structure, EPR and TD-DFT evaluation

Article abstract of DOI:10.1016/j.poly.2016.12.025

The P-APC ligands (EDTA-like aminopolycarboxylate ligands comprising 1,3-propanediamine backbone) H₄pdta, H₄pd3ap, H₄pddadp and H₄pdtp (H₄pdta?=?1,3-propanediamine-N,N,N′,N′-tetraacetatic acid; H₄pd3ap?=?1,3-propanediamine-N,N,N′-triacetic-N′-3-propionic acid; H₄pddadp?=?1,3-propanediamine-N,N′-diacetatic-N,N′-di-3-propionic acid; H₄pdtp?=?1,3-propanediaminetetra-3-propionic acid) were investigated. The chelating ligands coordinate to copper(II) via five or six donor atoms affording distorted trigonal-bipyramid and octahedral structures that were verified by X-ray analysis for Ba[Cu(pd3ap)]·6H₂O (1) and trans(O₆)-Ba[Cu(pddadp)]·8H₂O (2) complexes respectively. The impact of counter-ions on the P-APC complexes is shown in detail together with the analysis of another strain parameters. EPR spectral results confirm the penta-coordination of 1 and hexa-coordination of 2 in aqueous solution, even if several Cu(II) species with different protonation degree exist as a function of pH, and indicate that a hexa-coordinated structure is favored when the two axial COO^?donors close five-membered chelate rings. We also present here the results of molecular mechanics (LFMM) calculations based on our previously-developed force field along with results of DFT (Density Functional Theory). On the basis of extensive DFT and TD-DFT calculations the B1LYP/6-311++G(d,p) level has been seen as an accurate theory for calculating and predicting the UV–Vis spectra in case of copper–P-APC compounds.

Full text of DOI:10.1016/j.poly.2016.12.025

Accepted Manuscript
Chelating properties of EDTA-type ligands containing six-membered backbone
ring toward copper ion: Structure, EPR and TD-DFT evaluation
Marina Ćendić, Robert J. Deeth, Auke Meetsma, Eugenio Garribba, Daniele
Sanna, Zoran D. Matović
PII:
S0277-5387(16)30676-3
DOI:
http://dx.doi.org/10.1016/j.poly.2016.12.025
POLY 12383
Reference:
Polyhedron
To appear in:
Received Date:
Revised Date:
Accepted Date:
25 October 2016
12 December 2016
18 December 2016
Please cite this article as: M. Ćendić, R.J. Deeth, A. Meetsma, E. Garribba, D. Sanna, Z.D. Matović, Chelating
properties of EDTA-type ligands containing six-membered backbone ring toward copper ion: Structure, EPR and
TD-DFT evaluation, Polyhedron (2016), doi: http://dx.doi.org/10.1016/j.poly.2016.12.025
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.

Chelating properties of EDTA-type ligands containing six-
membered backbone ring toward copper ion: Structure, EPR and
TD-DFT evaluation
Marina Ćendić,*a Robert J. Deeth,b Auke Meetsma,c Eugenio Garribba,d Daniele Sannae and Zoran D.
Matović*a
aUniversity of Kragujevac, Faculty of Science, Department of Chemistry, Radoja Domanovića 12, 34000
Kragujevac, Serbia
bDepartment of Chemistry, Inorganic Computational Chemistry Group, University of Warwick, Coventry
CV4 7AL, UK
cStratingh Institute for Chemistry and Chemical Engineering, University of Groningen, Nijenborgh 4, NL-
9747 AG Groningen, Netherlands
dDipartimento di Chimica e Farmacia, Università di Sassari, Via Vienna 2, I-07100 Sassari, Italy
eIstituto CNR di Chimica Biomolecolare, Trav. La Crucca 3, I-07040 Sassari, Italy
*Corresponding authors: E-mail: zmatovic@kg.ac.rs
E-mail: marina.cendic@kg.ac.rs
Keywords: Copper(II)-Aminopolycarboxylates; Syntheses; LFMM; EPR; TD-DFT.

Abstract
The P-APC ligands (EDTA-like aminopolycarboxylate ligands comprising 1,3-propanediamine
backbone) H4pdta, H4pd3ap, H4pddadp and H4pdtp (H4pdta = 1,3-propanediamine-N,N,N',N’-tetraacetatic
acid; H4pd3ap
=
1,3-propanediamine-N,N,N'-triacetic-N'-3-propionic acid; H4pddadp
=
1,3-
propanediamine-N,N'-diacetatic-N,N'-di-3-propionic acid; H4pdtp = 1,3-propanediaminetetra-3-propionic
acid) were investigated. The chelating ligands coordinate to copper(II) via five or six donor atoms
affording distorted trigonal-bipyramid and octahedral structures that were verified by X-ray analysis for
Ba[Cu(pd3ap)]⋅6H2O (1) and trans(O6)-Ba[Cu(pddadp)]⋅8H2O (2) complexes respectively. The impact of
counter-ions on the P-APC complexes is shown in detail together with the analysis of another strain
parameters. EPR spectral results confirm the penta-coordination of 1 and hexa-coordination of 2 in
aqueous solution, even if several Cu(II) species with different protonation degree exist as a function of
pH, and indicate that a hexa-coordinated structure is favored when the two axial COO– donors close five-
membered chelate rings. We also present here the results of molecular mechanics (LFMM) calculations
based on our previously-developed force field along with results of DFT (Density Functional Theory). On
the basis of extensive DFT and TD-DFT calculations the B1LYP/6-311++G(d,p) level has been seen as
an accurate theory for calculating and predicting the UV-Vis spectra in case of copper-P-APC
compounds.

Introduction
EDTA-like aminopolycarboxylate ligands have a remarkable ability to chelate transition metal ions. The
most typical applications lie in: metal detoxification [1], treatment of HIV infection [2], diagnostic X-ray
and magnetic resonance imaging [3] etc. Our interest in APC ligands concerns their potential medicinal
use in the treatment of neural disorders (Wilson’s and Alzheimer's disease) [4]. We are investigating
influence of EDTA-like ligands on homeostasis of copper ion; so far we have obtained promising results
from in vivo tests (the copper level has been monitored in the experimental rats treated with APCs) which
indicate that EDTA-like ligands containing mixed carboxylate arms most dramatically inactivate copper
ions (presumably copper(II) ion however not excluding copper(I)). Accordingly, it is of major importance
to explore behavior of the copper(II) state with structurally different EDTA-like chelate molecules.
Recently, we reported copper(II) complexes of E-APCs (EDTA-like aminopolycarboxylate ligands
comprising ethylenediamine backbone) using structure [5] and theory [6] as a basis for research. In this
paper, we combine both studies on a set of Cu(II)–P-APC (comprised 1,3-propanediamine backbone)
complexes. A part of the present research (EPR based) includes both P-APC and E-APC classes of
complexes with various combinations of acetate and propionate arms (Figure 1). Thus, H4eddadp
(ethylenediamine-N,N'-diacetic-N,N'-dipropionic acid), H41,3-pddadp (1,3-propylenediamine-N,N'-
diacetic-N,N'-dipropionic acid) are of potential interest. Although their copper(II) complexes have
somewhat weaker stability constants than Cu(II)-edta, the work of Chaberek et al. [7] suggests that they
are more selective toward the Cu(II) ion compared to other metals. However, it was clear that the acetate
groups form more stable chelate structures than do β-propionate groups (H41,3-pdta = 18.92 [8];
H4eddadp = 16.3 [8]; H41,3-pddadp = 14.96 [8]) which gives a chance to H41,3-pdta (1,3-
propylenediaminetetracetic acid) chelate holding 1,3-propanediamine backbone to be effective in copper
immobilization. In support of this, Deeth et al. [6] demonstrated that LFMM also gives a smooth

correlation between stability constant and the number of six-membered chelate rings with logβ decrease
as the number of six-membered rings increase.
Figure 1. The general structure of APCs and geometrical isomerism of six-coordinate [M(edta-type)]n−
complexes. The trans(On) nomenclature refers to the size of the carboxylate chelate rings for the oxygen
donors; G stands for equatorial plane and R stands for axial line.
Two or three geometrical isomers are possible for APCs with different lengths of carboxylate arms:
trans(O5), trans(O5O6) and trans(O6) (O5 and O6 stands for five- and six-membered carboxylate arms)
(Figure 1). For the H4pddadp, all three isomers of [Cr(pddadp)]- complex [9] were isolated. The
octahedral P-APC complex [Cu(pdta)]2- [10] has been prepared with pdta4- acting as hexadentate. There
are no preparative and structural data for H4pd3ap and H4pdtp ligands and their corresponding copper(II)
complexes. We also attempt to investigate the nature of solution and solid state influence, monitoring the
level of axial or equatorial distortion. With this idea in mind, we prepared an additional [Cu(pddadp)]2–
complex anion bearing Ba2+ as a counter-ion. The Ba2+ complex is placed along with Na+ and H+
(physiologically present in the body fluids) complexes and a comparative structural study has been made.

The computational study has been undertaken using different DFT methods and basis sets with an aim to
establish the best approach with respect to the energy and geometry of the complexes. The DFT
calculations are compared with a recently developed LFMM force field [6] implemented in the molecular
operating environment (MOE) [11]. To be able to calculate, predict and monitor spectral parameters of
complexes the details of the comprehensive TD-DFT analyses are given as well. Finally, we would like to
clarify P-APCs/copper interactions by general correlation of their solution, electronic and structural
properties.
Experimental Section
Materials and measurements
Reagent-grade, commercially available chemicals were used without further purification. The 1,3-
propanediamine, 3-chloropropionic and chloroacetic acids were purchased from Fluka and used as
supplied. The Ba[Cu(edta)]·6H2O [12], K2[Cu(edtp)]·4H2O [13] and Mg[Cu(pdta)]·8H2O [10] were
prepared according to procedures given elsewhere.
Synthesis of ligands and complexes
Preparation of (1,3-propanediamine-N,N,N′-triacetic-N′-3-propionic) acid, H4pd3ap (Condensation
mixture)
Solution A: Monochloroacetic acid (11.50 g, 120 mmol) was dissolved in 42 ml of demineralized water
and 4.70 g (84 mmol) of calcium oxide was added to the solution. The pH rose from 1.0 to 11.2.
Subsequently, 3.11 g (42 mmol) of 1,3-propanediamine was added within 10 min. The temperature was
kept at about 50 °C without additional heating. After 1 h the temperature was increased to 70 °C and the
resulting mixture was left for an additional 5.5 h. During the entire reaction the pH was kept constant at
7.5-8.0 by the addition of calcium oxide. At the end of the reaction, a total of 1.90 g (34 mmol) of calcium

oxide was added. The reaction mixture was filtered warm over a glass filter to remove the excess of
calcium hydroxide. The reaction mixture (volume 50 ml) contains H3pd3a (50%), H4pdta (18%) and
H2pdda (23%) ligands. This was checked by means of chromatography of copper(II) complexes within
the above mixture. To the reaction mixture a solution of 9.44 g (236 mmol) NaOH in 15 ml of water was
added. The deposited Ca(OH)2 was separated by filtration under vacuum; Solution B: A solution obtained
by dissolving 4.53 g (42 mmol) of 3-chloropropionic acid in 10 ml of water was cooled in an ice bath and
carefully neutralized with a cold solution of 1.68 g NaOH (42 mmol) in 5 ml of water, making sure that
the solution temperature did not exceed 15 ºC; The Solution B was then added to Solution A (pH ~ 7.5).
The reaction mixture was gradually warmed up with stirring to 55 ºC during 5h (where pH was
maintained at ~ 8 by NaOH (1.68 g, 42 mmol in 5 ml of water). The volume of solution was held on 50
ml.
Preparation of barium(1,3-propanediamine-N,N,N′-triacetato-N′-3-propionato) cuprate(II)
hexahydrate, Ba[Cu(pd3ap)]·6H2O (1)
A 6.84 g (42 mmol) amount of CuCl2·2H2O was dissolved in 15 ml of water and was added to equivalent
amounts of the mixture of P-ACP acids described above with mixing and heating at 65 ºC for one hour. In
order to keep the pH at 7, a solution of NaHCO3 was added in small portions. The volume was maintained
by the addition of hot water. The blue colored synthesized mixture was then filtered. The filtrate was then
evaporated to a small volume and desalted by the technique of gel-filtration (Sephadex G-10 was used in
this purpose). The resulting blue solution was poured into a 5 × 60 cm column containing a Dowex 1-X8
(200−400 mesh) anion-exchange resin in the Cl- form. The column was then washed with water and
eluted with a 0.1 M solution of BaCl2. Three bands were obtained. The eluates were evaporated to a small
volume and desalted by passage through a G-10 Sephadex column, with distilled water as the eluent. The
first and third eluate corresponds to [Cu(pddadp)]2- and [Cu(pdta)]2- complexes respectively. The second
eluate was concentrated to a volume of 3 mL and stored in a desiccator over ethanol for several days. The
blue crystals were collected and air-dried. Yield: 5.1 g (19.5 %) of Ba[Cu(pd3ap)]⋅6H2O. Elemental

analysis is consistent with the composition of Ba[Cu(pd3ap)]·6H2O: C12H28N2O14CuBa, Mw = 625.23.
Anal. Calc.: C, 23.05; H, 4.51; N, 4.48. Found: C, 23.26; H, 4.54; N, 4.50. Melting point: 217 ºC.
Preparation of the trans(O6) isomer of barium(1,3-propanediamine-N,N′-diacetato-N,N′-di-3-
propionato) cuprate(II) octahydrate, Ba[Cu(pddadp)]·8H2O (2)
This complex was prepared by a previously described procedure [14], using BaCl2 as an eluent for anion
chromatography. Blue crystals of Ba[Cu(pddadp)]⋅8H2O were obtained. Elemental analysis is consistent
with the composition of Ba[Cu(pddadp)]·8H2O: C13H34N2O16CuBa, Mw = 675.29. Anal. Calc.: C, 23.12;
H, 5.07; N, 4.15. Found: C, 23.08; H, 4.98; N, 4.20. Melting point: 234 ºC.
Preparation of (1,3-propanediamine-N,N,N′N′-tetra-3-propionic) acid, H4pdtp (Condensation
mixture)
The solution obtained by dissolving 3-chloropropionic acid (23.87 g, 200 mmol) in 50 ml of water was
cooled in an ice bath and carefully neutralized with cold sodium hydroxide, NaOH (8.00 g, 200 mmol in
20 ml of water) ensuring that the temperature did not exceed 5 °C. With this solution the required amount
of 1.3-propanediamine (3.00 g, 40 mmol, 3.38 ml) was then added. The entire reaction mixture was
heated to a temperature of 50 °C, with stirring, followed by gradual addition of sodium hydroxide, NaOH
(8.00 g, 200 mmol in 20 ml of water) controlling the pH in the range 7-8 over two days.
Preparation of barium(1,3-propanediamine-N,N,N′N′-tetra-3-propionato)cuprate(II)trihydrate,
Ba[Cu(pdtp)]·3H2O (3)
To the condensation mixture containing H4pdtp acid (40 mmol, 50 ml) was added CuCl2·2H2O (6.80 g, 40
mmol). A solution of dark blue color was obtained. The pH was adjusted with sodium hydroxide to 7-8.
The solution was heated for 1 h at 65-70 °C and strained from NaCl. The filtrate was then evaporated to a
small volume and desalted by the technique of gel-filtration (Sephadex G-10). The resulting blue solution

was poured into a 5 × 60 cm column containing a Dowex 1-X8 (200−400 mesh) anion-exchange resin in
the Cl- form. The column was then washed with water and eluted with a 0.05 M solution of BaCl2. Two
bands were obtained. The second dark blue eluate was evaporated to a small volume and desalted by
passage through a G-10 Sephadex column, with distilled water as the eluent. The eluate was concentrated
to a small volume and stored in a desiccator over ethanol for several days. The light blue powder was
collected. Yield: 7.38 g (30%) of Ba[Cu(pdtp)]⋅3H2O. Elemental analysis is consistent with the
composition of Ba[Cu(pdtp)]·3H2O: C15H28N2O11CuBa, Mw = 613.26. Anal. Calc. for the complex salt:
C, 31.64; H, 5.09; N, 6.15. Found: C, 31.54; H, 5.20; N, 6.04. Melting point: 206 ºC.
Physical measurements
Block shaped, light blue (1) and green-blue (2) coloured, crystals were obtained by recrystallization from
water. A crystal with the dimensions of 0.45 x 0.19 x 0.12 mm (1) and 0.31 x 0.21 x 0.17 mm (2) was
mounted on top of a glass fiber and aligned on a Bruker [15] SMART APEX CCD diffractometer
(Platform with full three-circle goniometer). The crystals (1) and (2) were cooled to 100(1) K. Intensity
measurements were performed using graphite monochromated Mo-K radiation from a sealed ceramic
diffraction tube (SIEMENS). The final unit cell was obtained from the xyz centroids of 6710 (1) and 8046
(2) reflections after integration. The structure was solved by Patterson methods and extension of the
models was accomplished by direct methods applied to difference structure factors using the program
2
DIRDIF [16]. Final refinement on Fo was carried out by full-matrix least-squares techniques.
Crystallographic and experimental details of the structures are summarized in S1 and S2 in the Supporting
Information. CCDC entries 1054491 and 1054492 contain the CIF files for this manuscript. All data can
be obtained free of charge from the Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_requests/cif.
Anisotropic EPR spectra were recorded from 2000 to 4000 G at 120 K with an X-band (9.4 GHz) Bruker
EMX spectrometer equipped with an HP 53150A microwave frequency counter. The concentration of
Cu(II) complexes used to record EPR spectra in aqueous solution was 1×10-3 M and the solutions were

obtained dissolving polycrystalline samples in water. EPR spectra on the solid polycrystalline compounds
were recorded at 77 K. The modulation frequency was 100 kHz and modulation amplitude was 4 Gauss.
IR spectra in the 400-4000 cm-1 region were recorded on a Perkin Elmer FTIR spectrophotometer
SpectrumOne, using the KBr pellets technique. Electronic absorption spectra were recorded on a Perkin
Elmer Lambda 35 spectrophotometer. For these measurements 1x10-2 M aqueous solutions (pH~7) of the
complexes under investigation were used.
C, H, N analyses were performed at the Microanalytical laboratory, Faculty of Chemistry, University of
Belgrade, Serbia. Melting points were measured by a Stuart melting device with accuracy 1 °C.
Computational details
Part of DFT calculations used the Amsterdam Density Functional code version ADF2012 [17-19]. Details
on ADF calculation procedure are given in S3 of Supporting Information. The Gaussian09 program [20]
has been used for running DFT part as well. For these calculations, we used the unrestricted BP86 and/or
B3LYP and/or B1LYP hybrid functional and Ahlrich's def2-TZVP [21] or 6-311++G(d)(p) [22] basis
sets. The TD-DFT method required special treatment: all structures were optimized by using the 6–
311++G(d)(p) basis set for all atoms; the solvation was also incorporated by performing the TD-DFT
calculations in the presence of a self-consistent reaction field (SCRF) by using the polarizable continuum
models (PCM); the 3 XC functional investigated can be divided into two categories: (i) pure GGA
functional BP86 and (ii) hybrid GGA functional including B1LYP and B3LYP [20]. For simulation of
TD-DFT spectra (with ∆1/2,I=3000 cm-1) we used AOmix 6.87 software [23]. LFMM calculations [24]
employed the extended version of MOE version 2013.0801. The LFMM and MMFF94 force field
parameters were as described previously. To assist with the correct description of six-membered Cu-
propanediamine chelate rings, a Cu-N torsional restraint was added as described elsewhere [25].
Results and Discussion

P-APCs with an N2O4 chromophore containing five- and/or six-membered carboxylate arms (Figure 2):
H4pdta, H4pd3ap, H4pddadp and H4pdtp have been used for complexation of copper(II) ion. The P-APC
synthesis has followed several methods: i) condensation starting from neutralized α- or β-
monohalogencarboxylic acid and the corresponding diamine; ii) condensation of acrylic acid and diamine,
and iii) condensation of dihalogen derivatives of diamine with diverse amino-acids. Here the hexadentate
P-APCs have been prepared either as a condensation mixture or pure acid starting from neutralized α or β
chloro-aliphatic acid and 1,3-propanediamine.
Figure 2. Synthesis of the [Cu(II)(P-APCs)]n- complexes.

Depending on the structure of the P-APC ligands, the pentadentate form can be expected, as is the case of
the trigonal-bipyramidal [Cu(pd3ap)]2- complex, or hexadentate with symmetrical H4pdta, H4pddadp and
H4pdtp ligands (octahedral coordination). Structural variations of the P-APC's framework involve
increasing the size of the backbone and/or carboxylate rings to generate less strained systems.
Geometrical isomerism is possible for complexes of the P-APCs where the carboxylate arms are replaced
so as to form chelate rings of different size (Figure 1). In the case of H4pd3ap, a mixture of pentadentate
and hexadentate [Cu(pd3ap)]2- complexes was formed. Therefore, there is a delicate balance between
geometries as shown in Figure 2. This depends on many factors such as: the structure of the P-APC
ligand, pH (see later on) and the nature of counter-ion. We have isolated, by means of chromatography, a
pentadentate form of [Cu(pd3ap)]2- complex with one six-membered chelate ring, detached from the
copper (TBP geometry). By the same method, octahedral [Cu(pdtp)]2- has been prepared as well. All the
substances in this study were fully characterized by all available methods (NMR, IR, EPR, UV-vis etc.).
The structures of two prepared complexes (1) and (2) were verified by X-ray analysis.
Structural parameters of (1) and (2)
A structural diagram of the [Cu(pd3ap)]2- anion, with its adopted atom-numbering scheme is shown in
Figure 3. The asymmetric unit consists of eight moieties: a Ba2+ cation, [Cu(pd3ap)]2- anion and six
solvent water molecules with no atom sitting at a special position.

Figure 3. Above: X-ray crystal structure of [Cu(pd3ap)]2- anion and crystal packing along the a axis
showing hydrogen bonds; below: X-ray crystal structure of trans(O6)-[Cu(pddadp)]2- anion and crystal
packing along the a axis. Hydrogen atoms are omitted because of clarity.
Detailed analysis of the structures is presented in the Supporting Information. The triclinic unit cell
contains two asymmetric units, meaning two cations, two anions and twelve water molecules (Figure 3).
The pd3ap4- anion acts as a pentadentate chelate with one detached axial arm forming a TBP complex.
The Cu−L distances range from 1.9329(17) to 2.1269(18) Å (see Table S1 in the Supporting Information)
and are comparable with those in related Cu(APC)-type complexes [26]. An axial Cu-O bond value (Cu2-
O25 = 2.1269(18) Å) is slightly shorter than found in a similar square-pyramidal [Cu(pd3a)]- complex
anion (Cu-O = 2.18; 2.23Å) [27]. The six-membered backbone 1,3-propanediamine ring has a half-chair
conformation while in- and out of plane (G) acetate rings have twisted conformations. The trigonal-based
five-membered chelate ring also adopts a twisted form. The fact that one β-alaninato ring is detached
results in a distorted TBP geometry and therefore the accompanying asymmetry in the axial bond
distances could be due to the steric demands of the in-plane coordinated acetate ring or as a consequence
of the high stability of copper(II)-N2O3 chromophore already seen in the [Cu(pd3a)]- complex. Recently,
we observed that the more six-membered rings, the more pronounced the trigonality [27]. The trigonality
τ (angular structural parameter is defined and proposed as an index of trigonality, as a general descriptor

of five-co-ordinate centric molecules [28]) for the [Cu(pd3ap)]2- complex of 73 % indicates that this
structure has a distorted TBP geometry. A structural diagram of the trans(O6)-[Cu(pddadp)]2- complex is
also shown in Figure 2. Each asymmetric unit contains one formula unit, consisting of ten moieties: the
Ba2+ cation, [Cu(pd3ap)]2- anion and eight solvate water molecules. The triclinic unit cell contains twenty
units, two cations, two anions and sixteen water moieties. The anion pddadp4- (Figure 4a) acts as a
hexadentate ligand with six coordinated bonds forming a distorted octahedral complex. The Cu−L
distances range from 1.971(2) to 2.459(2) Å (see Table S2 in the Supporting Information) are comparable
to related copper(II)-APC systems [26]. There are two longer asymmetric axial bonds (Cu - O1 =
2.334(2)) and Cu - O5 = 2.459(2)). The six-membered backbone 1,3-propanediamine ring has a boat
conformation, the in-plane (G) acetate rings have quasi planar conformations, and the axial chelate ring
(R) adopts a slightly twisted boat form.
Copper-P-APC Geometry Dependence with Respect to the Structure of the Ligand, Geometry of
the Complex and the Nature of Counter-ions
To analyze Cu(II)–P-APC strain in more details the following four terms should be considered: (a) the cis
and trans angles around the central ion; (b) the ring angle sums of the various kinds of rings; (c) the
Cu−O−C or Cu−N−C bond angles; (d) the bond angles that a coordinated N atom makes with its
connectors. The structural data correlating the stereochemistry of the Cu(II) complexes are given in Table
1. These data provide the basis for the detailed analysis of the impact of counter-ions on the structure and
stereochemistry of the [Cu(pddadp)]2- complex. In the case of the quasi-symmetric [Cu(pddadp)]2-
complex ion we compared the influence of Na+, Ba2+ and H+ counter ions. Different ions lead to changes
in strain parameters of the P-APC systems (Table 1). The Σ∆(Oh) parameter (the sum of the absolute
values of the deviations from 90o of the L-M-L’ bite angles) for all hexadentate complexes is similar. In
case of Σ∆(Ring) strain parameter (the deviation from the ideal of the corresponding chelate rings bond
angle sum) the situation is different. The ∆Σ (Ring R) values vary from +7 in case of Ba2+ complex to +25
for the H+ complex (blue column). The Cu−O−C fragment of the carboxylate rings is expected to attain a

bond angle between 109.5 and 120° depending on the degree of covalency of the M−O bond. The
Cu−O−C bond angles of the elongated copper(II) chelates should deviate minimally when there are no 3-
propionato rings (this was demonstrated for the Cu-E-APCs [1]). The Cu-O-C angles for the axial β-
alaninato rings of the elongated Cu(II)–pddadp (with Ba2+, Na+ and H+ cations) deviate minimally from
the ideal bond angle of 109.5°. In the complex with the Ba2+ cation this deviation is negative (-5°) and for
the Na+ and H+ complexes it is positive and varies from 0 to +5° (blue column). The total deviation
around N atoms (Table 1) sums to roughly 13 and 17° for Ba2+ and Na+ complexes, respectively, and the
greatest deviation (20°) was found for the H+ complex (blue column). The extent of distortion is restricted
by chelation of the multidentate ligand and depends on its structure as well as the geometry of the
complex. Owing to the presence of the Jahn–Teller effect (ligand field stabilization), the APCs
preferentially coordinates in an equatorial plane and such a coordination can produce the tetragonally
elongated octahedral complexes. The geometry depends on the in-plane ligand field strength, the structure
of the ligand and the size of its chelate rings. The tetragonality parameter T (taken as the ratio of the
average equatorial Cu−O bond lengths to the average axial Cu−O bond lengths with values typically
around 0.8 0.02) [29] decreases in the order: Ba2+ > Na+ ≥ H+ complexes (Table 1). The same values
have been obtained for tetragonal octahedral complexes [30,31]. We observed that the variation in
tetragonality and axial Cu-O bond asymmetry (Table 1) depends on the nature of the counter-ion and its
covalent interaction with the coordinated carboxylate groups. The Ba2+ complex prepared in this work
shows moderate distortion (T = 0.823). The Jahn-Teller effect is most pronounced for the more distorted
Na+ (T = 0.799) and H+ complex (T = 0.798). In the case of quasi symmetrical complexes such as
[Cu(pddadp)]2-, the asymmetry of axial O-Cu-O bonds arises not only from intramolecular bonding but
primarily from covalent/electrostatic interactions of counter-ion and complex anion. The asymmetry in
bonding of the Ba2+ cation with equatorial oxygen pairs (Figure 4) results in the asymmetric strengthening
of axial Cu-O bonds (2.459 and 2.334 Å).

Table 1. Strain analysis of copper complexes with 1,3-propanediamine ligands with five- and six-membered
carboxylate rings.
∆Σ(ring)[b]
∆(Cu-O-C)[c]
Complex
Σ∆(Oh)[a]
T
R
G
R
G
Σ∆(N)[d]
T[e]
Ref.
This
trans(O6)-Ba[Cu(pddadp)]
51
+27
+7
-9
-5
+4
13
0.823
work
[14]
[32]
trans(O6)-Na2[Cu(pddadp)]
trans(O6)- H2[Cu(pddadp)]
43
44
+27
+23
+14
+25
-11
-11
0
+5
+4
17
20
0.799
0.798
+5
trans(O5O6)-Ba[Cu(pd3ap)][f]
trans(O5)-Ba[Cu(pd3ap)][f]
51
66
+26
+23
+5(0)
0
-12
-6(+1)
+2
+3
16
0.839
0.870
This
work
This
+31(-18)
+17(+3) 17(16)
work
[Cu(pd3ap)][g]
91
+33
-20
-10
+2
+5
20
0.924
This
work
[a] Σ∆(Oh) is the sum of the absolute values of the deviations from 90o of the L-M-L’ bite angles. All values rounded
off to the nearest degree. [b] ∆Σ(ring) is the deviation from the ideal of the corresponding chelate rings’ bond angle
sum. [c] ∆(Cu-O-C) (ring) is the mean value of the deviation of the corresponding rings’ Cu-O-C bond angle from the
109.5o. [d] Σ∆(N) is the sum of the absolute values of the deviations from 109.5o of the six bond angles made by
nitrogen atoms. A mean value for the two nitrogens is reported. [e] Tetragonality, expressed as ratio of mid-length in
plane and medium-length in the axial position. [f] DFT structures: LDA/DZP. [g] TBP complex (X-ray crystal

structure).
The interaction of Ba2+ with the first carboxylate oxygens (2.822 Å) lowers the strength of the matching
nitrogen and oxygen in the equatorial field and amplifies the bonding effect of the corresponding axial
oxygen (shortened length of 2.334 Å) (Figure 4a). Changing the direction of bonding toward the second
oxygen pair results in the opposite effect (longer length of 2.459 Å). The quasi symmetrical structures of
Na+ and H+ complexes derive from balanced electrostatic and covalent influences of Na+ and H+ ions
(Figure 4a). The observed symmetry of the crystal packing unit in the Na+ complex is a result of the
absence of any hydrogen bonds and other short contacts. The sodium ion acts symmetrically on both
equatorial rings which leads to the equal strengthening on the axial field resulting in equal axial Cu-O
bonds of 2.445 Å.

Figure 4. a) Ba2+, Na+ and H+ counter ions in the case of [Cu(pddadp)]2- complex; b) X-ray defined
structure of Ba[Cu(pd3ap)] dimmer.
The hydrogen bonds of the H+ complex are, on the whole, symmetrical (Figure 4a) in the sense that there
is an identical influence of hydrogen bonds and atoms in the equatorial and axial positions and,
accordingly, the regular elongated octahedral geometry was reported. Finally, the impact of counter-ions
to the complex geometry can be of primary or secondary importance. When the influence of ions on the
equatorial or axial plane is too strong, like in the Ba2+ complex, the axial bond of the six-membered ring
already lengthened by the presence of a five-member ring in equatorial plane is further weakened by the
strong influence of the Ba2+ ion in the G plane. One may conclude that the nature of the counter-ion may
have an impact on the complex configuration; in this case it is reflected through asymmetric bond lengths.
The strain parameters of [Cu(pd3ap)]2- (Table 1) derived from DFT optimized trans(O5) and trans(O5O6)
isomers were compared with strain values of the experimental TBP [Cu(pd3ap)]2- complex. The
trans(O5O6) geometrical isomer does not show large deviations (Table 1) which should relate to the
relative stability of this isomer in solution. However, the trans(O5) isomer shows a difference, primarily
in the β-alaninato ring (∆Σ(Ring G) = +31 and ∆(Cu-O-C) (Ring R) = +17 (yellow columns in Table 1).
This causes the large in plane deviation of the trans(O5) six-membered ring (DFT gives an advantage to
trans(O5O6) isomer with energy difference of 1.21 kcal/mol). Therefore, the trans(O5O6) isomer has been
seen as a molecule potentially found in solution, and therefore potentially being able to be converted. We
are of the opinion that divalent barium counter ion play not only charging and solid state cell packing role

but also may contribute to the substantial change of the complex geometry. One may see that barium
cations (Figure 4b) pull down along axial O-Cu-O axes G acetate ring and, therefore, contribute to
detaching of β-propionate oxygen by synchronized effects of Ba2+⋅⋅⋅-OOC attraction and steric
overcrowding.
Spectroscopy
EPR of APCs
In agreement with previously reported data [33], edta forms with Cu(II) ion two complexes, the first at
acid pH values (pH < 4) with (2 N, 3 COO–) coordination (gz 2.330, Az 154.6 × 10–4 cm–1, I in Figure 5
(1)) and the second one from pH 4 with (2 N, 4 COO–) coordination (gz 2.292, Az 160.5 × 10–4 cm–1, II in
Figure 5 (1)). All the EPR parameters are reported in Table 2.
Table 2. EPR parameters of Cu(II) complexes formed with APCs.[a]
Ligand
edta
gz
Az
gz/Az
151
143
138
136
126
128
127
~ 130
Coordination mode
2 N, 3 COO–
2 N, 4 COO–
2 N, 4 COO–
2 N, 3 COO–
2 N, 3 COO–
2 N, 4 COO
2.330
2.292
2.283
2.287
2.272
2.265
2.307
~ 2.26
154.6
160.5
165.1
168.5
180.8
176.4
181.3
~ 175
edta
pdta
edtp
pd3ap
pddadp
pdtp
2 N, 3 COO–
2 N, 4 COO
pdtp
−1
[a] Values of Az reported in 10−4 cm
.
H4edtp forms (6,5,6)-chelate rings in the equatorial plane. Square pyramidal coordination (2 N, 3 COO–)
has been established in the solid state. The same coordination is retained in solution. EPR parameters, gz
2.287, Az 168.5 × 10–4 cm–1 (I in Figure 5 (2)), are very similar to those of the species with (2 N, 4 COO–)
coordination formed by pdta for which (5,6,5)-chelate rings are closed in the equatorial plane. In contrast
to the edta complex, on the basis of the EPR parameters (gz 2.283, Az 165.1 × 10–4 cm–1, I in Figure 5 (3)),

the H4pdta forms only one species in the pH range 3-10 with (2 N, 4 COO–) coordination, which is in
agreement with the X-ray crystal structure of [Cu(pdta)]2- [10]. Comparing the spectroscopic data
measured for H4edta and H4pdta it is possible to observe a slight decrease of gz and a slight increase of Az
in agreement with a more relaxed structure due to the larger size of the chelate ring formed by the two
nitrogen donors (six- vs. five-membered N,N chelate ring) [34]. Increasing the size of the central ring
from (5,5,5)-chelation for H4edta to (5,6,5)-chelation for H4pdta (the two axial position are occupied by
COO- groups) a slight decrease of gz and an increase of Az is expected [34], in agreement with what is
observed experimentally. A good parameter which measures the distortion degree of the equatorial plane
of a Cu(II) species is the ratio gz/Az [35] and it is 106 for a complex with a high degree of planarity such
as [Cu(GlyGlyGlyGlyH–3)]2– with (NH2, N–, N–, N) coordination [34] and 143 and 138 for the hexa-
coordinated species of H4edta and H4pdta.

Figure 5. 1) X-band anisotropic EPR spectra (120 K) recorded on a solution obtained dissolving
Ba[Cu(edta)] in water (CuII concentration 1 mM): (a) pH 3.30; (b) pH 4.15 (spontaneous pH) and (c) pH
10.45; with I and II the first parallel resonances of the species with (2 N, 3 COO–) and (2 N, 4 COO–)
coordination are indicated. 2) X-band anisotropic EPR spectra (120 K) recorded on a solution obtained
dissolving K2[Cu(edtp)] in water (CuII concentration 1 mM): (a) pH 3.60; (b) pH 6.10 (spontaneous pH);
(c) pH 8.00. With I the first parallel resonances of the species with (2 N, 3 COO–) coordination are
indicated. 3) X-band anisotropic EPR spectra (120 K) recorded on a solution obtained dissolving
Ba[Cu(pdta)] in water (CuII concentration 1 mM): (a) pH 3.35; (b) pH 5.30 (spontaneous pH) and (c) pH
9.75. With I the first parallel resonances of the species with (2 N, 4 COO–) coordination are indicated.
EPR spectra of the polycrystalline complex [Cu(pd3ap)]2–, recorded at 77 K, are shown in Figure 6 (1).
The spectrum is isotropic with a g value of 2.131. The broad and unresolved signal is due to the dipolar
interaction which enlarges the spectral lines and obscures the hyperfine structure. The line width is 98.1
G. In solution the species with (2 N, 3 COO–) coordination predominates from pH 4 to pH 11 with gz

2.272, Az 180.8 × 10–4 cm–1 (gz/Az 126). If these parameters are compared with those of the hexa-
coordinated complex of H4pdta with (2 N, 4 COO–) coordination (H4pdta forms the same (5,6,5) rings in
the equatorial plane), a decrease of gz and an increase of Az can be observed; this change confirms the
absence of a COO– group in one of the two axial positions. The spectrum (gz > gx ~ gy) indicates a
tetragonal symmetry with the unpaired electron in the dx2-y2 orbital and a distorted square pyramidal
geometry [36]. Therefore, the penta-coordinated structure of [Cu(pd3ap)]2– in the solid state is observed
also in water, even if in this solvent the geometry can be described as intermediate between the square
pyramid and the trigonal bipyramid rather than close to the TBP limit (see Figure 2); probably, solid state
effects not operating in solution stabilize the structure of 1.
Figure 6. 1) X-band EPR spectra of the polycrystalline complex Ba[Cu(pd3ap)] recorded at 77 K. 2) X-
band anisotropic EPR spectra (120 K) recorded on a solution obtained dissolving Ba[Cu(pd3ap)] in water
(CuII concentration 1 mM): (a) pH 5.90 (spontaneous pH) and (b) pH 10.45; with I the first parallel
resonances of the species with (2 N, 3 COO-) coordination are indicated. 3) X-band EPR spectra of the
polycrystalline complex Ba[Cu(pddadp)] recorded at 77 K. 4) X-band anisotropic EPR spectra (120 K)
recorded on a solution obtained dissolving Ba[Cu(pddadp)] in water (CuII concentration 1 mM): (a) pH
5.80 (spontaneous pH) and (b) pH 9.10-10.70; with I the first parallel resonances of the species with (2 N,
4 COO–) coordination are indicated.

EPR spectra of the polycrystalline complex [Cu(pddadp)]2– were recorded at 77 K (Figure 6 (3)). Three
broad and one sharp signal are detected in the region 2900-3400 G. The resonances at g = 2.214, 2.052
and 2.022 can be attributed to a Cu(II) in an octahedral geometry with a rhombic symmetry, for which a
spectrum with three g values is expected [36,37]; the order gz >> gy > gx > ge indicates that the ground
state is based mainly on the dx2−y2 orbital [36-38]. The resonance at g = 2.078 is characterized by a very
small linewidth (2.4 Gauss). The most plausible assignment is to a double quantum (DQ) transition that
involves the rapid consecutive absorption of two single quanta. The basis for this attribution are: i) the
phase of the resonance is compatible with that of the single quantum transitions; ii) the g factor (2.078) is
close to the average of the three g values of the single quantum transitions (2.096); iii) the absorption
cannot be simulated using only single quantum transitions and iv) the line width of the band (2.4 Gauss)
is significantly less than those measured for the single quantum transitions (> 50 Gauss) [39]. These
transitions have been detected for Fe(II) [40], Mn(II) [41], Ni(II) [42], Cu(II) [43,44] and suggest a strong
interaction in the solid state between two neighboring Cu(II) ions to give a spin state S = 1. When
[Cu(pddadp)]2– is dissolved in water, the hexa-coordinated species with (2 N, 4 COO–) coordination (gz
2.265, Az 176.4 × 10–4 cm–1, I in Figure 6 (4)) is observed at pH > 5. EPR spectrum of the polycrystalline
complex [Cu(pdtp)]2− was recorded at 77 K (Figure 7 (1)). The spectrum is isotropic with a g value of
2.146. The broad absorption is due to the dipolar interaction in this case too. For H4pdtp the species with
(2 N, 3 COO–) coordination is observed around neutral pH (I in Figure 7 (2)), whereas that with (2 N, 4
COO–) coordination (analogous to the solid state compound) is revealed at pH > 9 (II in Figure 7 (2)) and
coexists with the first one (indeed, the spectral resonances become large and little unresolved). This
indicates that when all the chelate rings are six-membered the complexes are less stable. In conclusion, in
aqueous solution H4edta and its derivatives forms Cu(II) species with different coordination modes: hexa-
coordination is favored when the two axial donors close five-membered chelate rings (H4edta, H4pdta).
When the size of the axial chelate rings increases the hexa-coordinated coordination is less stable and

higher pH values are necessary to observe (2 N, 4 COO–) coordination. Comparing the X-ray crystal
structures of 1 and 2 with the EPR spectra recorded in water, we found that the penta- and hexa-
coordination occur also in solution.
Figure 7. 1) X-band EPR spectra of the polycrystalline complex Ba[Cu(1,3-pdtp)] recorded at 77 K; 2)
X-band anisotropic EPR spectra (120 K) recorded on a solution obtained dissolving Ba[Cu(1,3-pdtp)] in
water (CuII concentration 1 mM): (a) pH 3.50; (b) pH 4.05; (c) pH 7.05 (spontaneous pH) and (d) pH
9.50. With I, II and III the first parallel resonances of the species with only one amine nitrogen bound to
Cu(II) (I and II) and with (2 N, 2 COO–) coordination (III) are indicated.
IR data
The complexes have been further analyzed by means of IR and UV−vis spectra (see Table and Figures S4
in the Supporting Information). The IR data (carboxylate region) are in agreement with the structures and
molecular symmetries. As was found in the metal aminocarboxylic acid complexes [45,46], in the
asymmetric carboxyl stretching region, the frequency assigned to five-membered rings [45] lies at higher
energy than the corresponding frequency of six-membered chelate rings [46]. These complexes with C1
and C2 symmetry evidence two strong and separated bands in the asymmetric stretching carboxylate

region (≈ 1640 cm-1 = higher energy of five-membered ring and ≈ 1570 cm-1 = lower energy of six-
membered ring). Bands at 1630 cm-1 and 1564 cm-1 correspond to trans(O6)-[Cu(pddadp)]2-. Accordingly,
the asymmetric stretching vibration of the five-membered glycinate rings, in case of [Cu(pdta)]2- [10],
was at 1601 cm-1. The same vibrations of [Cu(pd3ap)]2- complex are in accordance with a presence of
three acetate rings and one propionate ring: 1622 cm-1 corresponds to the glycinate rings and 1559 cm-1
corresponds to the β-alaninato ring. As expected in [Cu(pdtp)]2- complex there is only one strong band at
1575 cm-1 originating fromfour carboxylate six-membered rings.
Computational experiments
Here we compared four DFT methods, LDA, BP86, B1LYP and B3LYP using the diverse basis sets in
order to see structural and energetic differences. Table 3 gives an overview of the energy differences for
all isomers of [Cu(pddadp)]2– via three methods (LDA, BP86 and B1LYP) and the relative
conformational energies derived from LFMM calculations using the force field developed previously.
Table S5 (Supporting Information) compares the theoretical and experimental structural parameters along
with RMSD values between computed and X-ray determined structures for the P-APC series.
Table 3. Comparison of the relative energy data for [Cu(pddadp)]2- isomers
calculated by DFT and LFMM.
B1LYP/6-
Isomer
LDA/DZP
BP86/DZP
BP86/TZP
BP86/TZ2P
LFMM
311++g(d,p)
trans(O6)[a]
trans(O5O6)
trans(O5)
0
0
0
0
0
0
3.6
2.4
6.1
2.5
5.6
1.9
5.3
1.74
4.93
0.2
1.1
8.7
[a] The isomer with the lowest-energy minimum has been indicated with
kcal·mol−1
0
.
In general, the comparison between the theory and experiment is reasonably good. The DFT energies of
[Cu(1,3-pddadp)]2- isomers are in agreement with predictions based on the geometry of isolated trans(O6)
isomers (Table 3). The DFT results indicate that the crystallographic structure corresponds to the DFT

lowest-energy structure regardless of the type of method or basis set. It can be observed from Table S5
that BP86/TZ2P and B1LYP/6-311++g(d,p) methods give the lower energy differences. The prediction
that the trans(O6) isomer has the lowest energy is consistent with the spectral results and our expectations
from working with such systems. However, BP86 performs less well for structures (Table S5). In
contrast, the LDA functional shows Cu−L distances very close to those observed in the solid state. One
may see that RMSD value (Table S5 Supporting Information) for equatorial atoms is four time lower than
heavy atoms of whole molecule. This is again indication that carboxylate rings out of plane more distort
from DFT calculated due to environmental impact (counter-ions, solid state, solvent) and that simple DFT
job cannot solve this issue. The LFMM optimized, hexadentate [Cu(pdta)]2– and [Cu(pdtp)]2– complexes
have shown usual axial and equatorial Cu-O and equatorial Cu-N distances. Thus a reasonable agreement
between the X-ray and LFMM structure were obtained (Figure 8).
Figure 8. Overlay of LFMM (yellow) and X-ray (blue) calculated structures.
The higher the number of six-membered chelate rings, the greater is the strain in the complex. This comes
out of the structural strain analysis and can also be correlated with LFMM computed strain energies. For
example, if the copper is removed from the complex, the strain energy for the tetraanionic ligand in its
bound conformation can be computed. The ligand is then relaxed to its lowest energy conformation via a

low mode molecular dynamics search. The greater the decrease in energy, the greater was the strain
induced in the ligand upon binding to the metal centre. The energy changes for H4edta, H4pddadp and
H4pdtp are 107, 136 and 180 kcal mol-1 respectively. As expected, the H4pdtp ligand is the most strained
and the associated complex is predicted to be less stable (LFMM bond average lengths are: Cu-O axial of
2.477 Å, Cu-O equatorial of 1.964 Å, and Cu-N of 2.133 Å).
TD-DFT and UV_vis data Typical electronic absorption spectra of the octahedral [Cu(P-APC)]2- complex
is shown in Figure 9. One can see that whole spectral range can be divided into two parts: UV and
Vis/Near_IR regions. Regarding to the shape of spectra the major (energy and absorptivity) region
belongs to UV range. Appearance of one sharp band is common for all the P-APC complexes. Energy
position of these bands is not substantially different too, found within the range of 2-5 nm. This region
was not, evidently, within scope of our interest. The most important one is a visible region reflecting
usually geometry configuration and ligand field strength (LFS) in relation of copper(II) complexes. The
Near_IR region is occasionally characterized by one weak band of low absorbance which is rarely
detected being covered by the higher intensity bands (see Figure 9). All of the complexes are blue and
experimentally exhibit one asymmetric absorption band irrespective of the underlying approximate C1 or
C2 symmetry.
Figure 9. Left: Uv_Vis Spectrum of [Cu(pddadp)]2- obtained from water solution under pH~7; right:
principal MO transitions within Vis_Near_IR region in case of [Cu(pdta)]2- ion.

If we look at Table 4, the highest visible absorption band was obtained for the trans(O6)-[Cu(pddadp)]2–
complex (658 nm) and the lowest for the [Cu(pdtp)]2– complex (752 nm). It was observed that in the
absence of glycinato rings in the axial position the band of complexes with a higher number of glycine
chains in the equatorial plane is shifted toward higher energy. Accordingly, it is expected that
[Cu(pdtp)]2– without any glycine ring shows the lowest energy band. The other two complexes [Cu(1,3-
pd3ap)]2- and [Cu(pdtp)]2– show single asymmetrical absorption bands associated with shoulders,
especially in the case of the [Cu(pd3ap)]2– complex (see spectrum in Supporting Information S6). It is
documented for copper(II) complexes that a single visible band with a high-energy shoulder is indicative
of a trigonal bipyramidal stereochemistry while an absorption envelope with a low-energy shoulder is
characteristic of a square pyramidal geometry [47]. The complex [Cu(pd3ap)]2– shows a visible band with
an obvious shoulder on the higher energy side which is according to the above findings consistent with
trigonal bipyramidal (TBP) geometry and X-ray structure reported in this work.
As an example of TD-DFT transitions we are giving here the full simulated UV-vis spectra of [Cu(pdta)]2-
complex ion (Figure 9). The molecular orbital analysis of TD-DFT data shows that the main UV
transitions are due to ligand to metal charge transfers (LMCT). Within the visible region TD-DFT gave
three transitions: one at higher energy and higher oscillator strength followed by the two additional
transitions of much lower intensity. The most intense transition generally determines position of
convoluted band as well. This excitation line is a consequence of d-d transition mix: 27% HOMO-
20ꢀLUMO and 23% HOMO-21ꢀLUMO. The near IR region is occasionally characterized by one
excitation line of low oscillator strength which is covered by the 63% of HOMOꢀLUMO transition (see
Figure 9). The ground state in all the cases is dx2-y2 and the main percentage contribution of orbitals
involved in singlet transition is attributed to dz2, dxy, dxz and dyz orbitals (see Supporting Information S7).
Several strategies have been proposed to extract, using quantum-mechanic, excited state information from
the ground state density. The most promising of which is time-dependent TD-DFT designed by Runge
and Gross [48]. TD-DFT calculates the frequency-dependent linear response of the charge density with

respect to a time-dependent perturbation in the electric field [49]. The results include calculation of
excitation energies ωI and oscillator strengths fI for transition between the ground and excited states I of a
system [50]. The difficulties associated with the calculation of excited states were overcome describing
these excitations in terms of ground state properties such as molecular orbital (MO) contributions.
Table 4. TD-DFT and experimental Absorption data in water for hexadentate Cu(II)–P-APCs complexes.
Absorption[1]
TD-DFT[2]
Ref.
COMPLEX
B1LYP
B3LYP
BP86
Esf [3]
Esf
Esf
λ (nm) /
λ / ε
λ / ε
λ / ε
ε (mol-1dm3cm-
1
)
[Cu(pdta)]2-
704 / 84
658 / 68
752 / 106
640 / 80
613 /68
1.100
1.073
613 / 78
592 / 86
654 / 186
1.148
1.111
535 / 3325
590 / 2130
1.318
1.115
[10], This
work
[Cu(pddadp)]2-
This work
[Cu(pdtp)]2-
690 / 112
917 / 740
1.089
0.914
1.150
0.922
578 / 3140
905 / 2880
1.301
0.926
This work
This work
[Cu(pd3ap)]2- [4]
680 sh
909 / 985
838 / 69
1020 / 416 (sh)
1010 / 557 (sh)
1020 / 1490
(sh)
[Cu(pda3p)]2- [5]
715 / 90
655 / 90
1.090
This work
predicted
[1] Absorption spectra were recorded under pH~7.
[2] Simulated UV-Vis absorption band (visible or near IR region) from Gaussian TD-DFT calculations. The half-bandwidths, ∆1/2,I, were taken to be equal to 3000 cm-1.
[3] Energy scaling factor.
[4] TBP complex.
[5] Simulated and predicted absorption visible peak position of not synthetized trans(O6)-[Cu(pda3p)]2- complex anion.
Holland and Green compared the performance of 41 different XC functionals in their ability to predict the
UV/vis spectra of copper and zinc bis(thiosemicarbazonato) complexes [51]. Hybrid DFT methods were
found to outperform all pure DFT functionals. Of the functionals tested, B1LYP gave the most accurate

results regarding to the all parameters monitored for the copper complexes [51]. Starting from their results
we performed time-dependent density functional theory calculations using optimized geometries of our
complexes at the three different methods, namely BP86, B1LYP and B3LYP; we used the same basis set
6-311++G(d,p), as Holland and Green, in order to further check the viability of TD-DFT results in our
case. In each case, oscillator strengths and excitation energies for transition between the ground state and
the first 10 excited stated states were calculated. We have used the AOMix program [23] to produce
Gaussian simulated UV/vis spectra based on the TD-DFT calculations. Each calculated transition between
the ground and excited states I, can be modeled by a single Gaussian function in accordance with eq. (1).
ꢁ
ꢃꢊ
ꢆ
ꢂ − ꢂꢇ
ꢇ
ꢄ
ꢁ
ꢃ
ꢀ ꢂ = 2.174X10 ꢅ
ꢋꢌꢍ ꢎ−2.773
ꢏ … … … … … . . ꢁ1ꢃ
ꢈꢉ,ꢇ
ꢈꢊꢉ,ꢇ
ꢐ
ꢊ
ꢊ
ꢓꢔ
ꢁ ꢃ
ꢁ
ꢃ ꢁ
ꢃ
ꢁ ꢃ
ꢅ ꢆ ꢑ = 4.319ꢒ10 ꢕ ꢀ ꢂ ꢖ ꢂ … … … … . … . … … … … … … … . . 2
ꢇ
Where ε stands for molar absorptivity (molar absorption coefficient), ωI stands for TD-DFT calculated
excitation energies, (in cm-1), fI is transition oscillator strength (dimensionless), and ∆1/2,I stands for half-
band width, (in cm-1). The total integrated intensity under the absorption profile is equal to the sum of the
oscillator strengths and is given by eq. (2). We have changed half- band width only during spectra
convolution. Gaussian convoluted TD-DFT data are compared with experimental values as well (see
Table 4 and Figure 10). For low resolution spectroscopy, Gaussian functions have been shown to be
suitable for accurate simulation of experimental band shapes [52]. Experimental half-band widths are
generally unknown, and are difficult to estimate due to their high dependence on solvation and/or vibronic
coupling within the system. After we used different values for this parameter during convolution process
to minimize the difference between the experimental and simulated UV/vis spectra, the AOMix default
value of 3000 cm-1 has been taken as a common value in all the cases. Calculated energies are often
scaled, particularly in vibrational frequency analysis. Therefore, an energy scaling factor, ESF, was also
applied after each simulation. It is evident from Table 4 and Figure 10(a-b), that our results are in
agreement with conclusions of Holland and Green [51]. The parameters of particular importance are band