Nickel(ii) in chelate N2O2 environment. DFT approach and in-depth molecular orbital and configurational analysis

DaltonTransactions

PAPER

Cite this: Dalton Trans., 2013, 42, 13357

Received 26th April 2013,Accepted 24th June 2013

DOI: 10.1039/c3dt51099a

www.rsc.org/dalton

Nickel(II) in chelate N2O2 environment. DFT approachand in-depth molecular orbital and configurationalanalysis†

Srećko R. Trifunović,a Vesna D. Miletić,‡a Verica V. Jevtić,a Auke Meetsmab andZoran D. Matović*a

The O–N–N–O-type tetradentate ligands H2S,S-eddp (H2S,S-eddp stands for S,S-ethylenediamine-N,N’-di-

2-propionic acid) and H2edap (H2edap stands for ethylenediamine-N-acetic-N’-3-propionic acid) and the

corresponding novel octahedral nickel(II) complexes have been prepared and characterized. N2O2 ligands

coordinate to the nickel(II) ion via four donor atoms (two deprotonated carboxylate atoms and two

amine nitrogens) affording octahedral geometry in the case of all investigated Ni(II) complexes. A six

coordinate, octahedral geometry has been verified crystallographically for the s-cis-[Ni(S,S-eddp)(H2O)2]

complex. Structural data correlating similarly chelated Ni(II) complexes have been used to carry out an

extensive configuration analysis. Molecular mechanics and Density Functional Theory (DFT) have been

used to model the most stable geometric isomer, yielding, at the same time, significant structural and

spectroscopic (TDDFT) data. The results from density functional studies have been compared to X-ray

data. Natural Bond Orbital (NBO) and Natural Energetic Decomposition Analysis (NEDA) have been done

for the [Ni(edda-type)(H2O)2−n] and nH2O fragments. Molecular orbital analysis (MPA) is given as well.

The infra-red and electronic absorption spectra of the complexes are discussed in comparison to the

related complexes of known geometries.

Introduction

The excess of copper and nickel or lead and other heavymetals can be removed from the human body by complexingthem with EDTA (ethylenediamine tetraacetic acid) and similarchelating agents. The effects of chelating agents and drugsused clinically as antidotes to metal toxicity were reviewed inseveral articles.1 Metal complexes with open coordination siteshave found wide use in molecular recognition. They serve asbinding sites in the development of chemosensors; they arealso used to study metalloenzyme function in bioinorganicchemistry, or to direct supramolecular self-assembly. Lewis-acidic metal complexes can target a large variety of Lewis basicfunctional groups, making them very suitable for the design ofsynthetic receptors. Coordination to metal ions typically

occurs with large enthalpies compared to those for hydrogenbond formation, salt-bridges, or dipole–dipole interactions.This gives the opportunity to study molecular recognition andself-assembly in solvents competing for binding, such aswater, using coordinatively unsaturated metal complexes asbinding sites.2

A number of tetradentate ligands similar to edda (edda =ethylenediamine-N,N′-diacetate ion), such as: analogues,homologues, C and N-substituted derivatives occasionally withchiral carbon(s) atoms in their structure, have been prepared.Many of these ligands can coordinate via either an α- or β-ciscoordination mode to a metal ion (making different geometri-cal isomers) as proposed by Weyh and Pierce.3 The S,S-eddp(S,S-eddp = S,S-ethylenediamine-N,N′-di-2-propionate ion)ligand has two chiral carbon atoms within carboxylate arms(adopting S configuration). The major number of papersdescribing structural and optical properties of [M(edda-type)L]n

(L = two monodentate or one bidentate ligand) complexes havebeen published with M being Co(III), Cr(III), Cu(II) and Ni(II).4,5

However, metal complexes of prevalently optical AA(Et)AA(AA = amino acid, Et = ethylene) ligands were limited to thetheir IR, NMR and/or CD (circular dichroism) spectralproperties.6–8 Just a few MAA(ET)AA complexes were structurallycharacterized where M = oxovanadium(IV) (i.e. VO2+ or vanadyl)

†Electronic supplementary information (ESI) available: Tables S1–S3. CCDC932753. For ESI and crystallographic data in CIF or other electronic format seeDOI: 10.1039/c3dt51099a‡Deceased 13th June 2013.

aDepartment of Chemistry, Faculty of Science, University of Kragujevac, Kragujevac,

SRB-34000, Serbia. E-mail: [email protected] Institute for Chemistry and Chemical Engineering, University of

Groningen, Nijenborgh 4, NL-9747 AG Groningen, The Netherlands

This journal is © The Royal Society of Chemistry 2013 Dalton Trans., 2013, 42, 13357–13368 | 13357

Publ

ishe

d on

25

July

201

3. D

ownl

oade

d by

Uni

vers

ity o

f B

elgr

ade

on 0

9/05

/201

4 11

:55:

49.

View Article OnlineView Journal | View Issue

and Cr(III), with AA limited to alanine, proline and methionine.9,10

Only one nickel(II) complex of an AA(ET)AA-like ligand such asC2-symmetrical ligand H2baboc [H2baboc stands for (all-R)-1,2-bis(2-aza-3-carboxybicyclo[3.3.0]octan-2-yl)ethane] has been syn-thesized and its structure elucidated by X-ray analysis.11

Weyh’s proposal for α- or β-cis isomer distribution has beenbetter described by introducing a new nomenclature adoptedfor this system.4,12 Therefore, the linear edda-type ligands canform three possible geometrical isomers of the generalformula [M(edda-type)L2]

n (L = monodentate ligand). Theseisomers are denoted as s-cis, uns-cis and trans (Fig. 1).However, when M = Pt(IV) all three isomers have been preparedand characterized in the case where edda-type = edda4 and S,S-eddp.13,14 It was shown that s-cis geometry was favored inmetal complexes with edda-type ligands forming five-mem-bered chelate rings while uns-cis geometry was favored in com-plexes comprehending six-membered chelate rings.4

The observed greater ring strain in the octahedral planethan out of the plane is accepted as a general reason for suchdistribution of geometrical isomers.4,15–17 The s-cis geometrywas also proposed for the [M(edda)(H2O)2] complex (M = Ni(II),Cu(II) and Co(III)).14–16 While s-cis geometry of the pure[Cu(edda)(H2O)2] complex is not yet crystallographically con-firmed, the X-ray structure of the s-cis-[Ni(edda)(H2O)2]complex was reported by several authors.18,19 Furthermore,nickel(II) ion surrounded by bis N-hydroxymethyl-edda ligand(abbreviated as OHMeedda) also adopts s-cis geometry19 whilefor the Ni-1,3-pdda complex (1,3-pdda stands for 1,3-propane-diamine-N,N′-diacetate anion) the uns-cis geometry was foundto be dominant one20 by X-ray analysis. In the case of Ni-eddp(eddp stands for ethylenediamine-N,N′-di-3-propionate ion) theproposed uns-cis geometry accounts for the larger six-mem-bered carboxylate arms being able to encircle nickel(II) in equa-torial plane.20

For the edda-type ligands having mixed (five- and six-mem-bered) carboxylate arms, the geometrical isomers differ in thenumber (0, 1, or 2) of six-membered rings lying in the G (equa-torial) plane. The tetradentate ligand dacoda (1,5-diazacyclo-octane-N,N′-diacetate ion), due to the steric make-up of itsbackbone chelate, allows the coordination of only oneadditional ligand and the square-pyramidal [M(dacoda)(H2O)]complex with the ligand being coordinated in the equatorialplane (M = Co(II), Ni(II), Zn(II) or Cu(II)) has been prepared.21–23

In the present paper we examined how the electronic inter-actions influence energetic and geometric features using DFT

methods and some third party EES such as a natural bondorbital NBO24 or natural energetic decomposition analysisNEDA25–27 and molecular orbital MPA (Mulliken populationanalysis) analysis. The results obtained have been compared tothe experimental data to enhance the bonding nature of thewhole set of nickel(II)-edda-type complexes.

This work is the continuation of our research with M(II)complexes (M(II) = Cu(II) and Ni(II)) of aminopolycarboxy-lates.16,17 Here we report the preparation and characterizationof the neutral [Ni(S,S-eddp)(H2O)2] and [Ni(edap)(H2O)2] com-plexes with expected s-cis and uns-cis geometry, respectively.The s-cis and uns-cis geometry of the isolated complexes wereassigned through their infrared (IR) and electronic absorption(UV) spectra. The spectra of the complexes are discussed inrelation to the other [Ni(edda-type)(H2O)2] complexes ofknown configuration. The proposed geometry of the complexeswas confirmed by X-ray crystal structure analysis in the case ofs-cis-[Ni(S,S-eddp)(H2O)2] complex and by DFT theory for theunsymmetrical uns-cis-Ni(edap)(H2O)2] complex. DFT theoryalong with its successors NBO, NEDA and MPA has beenemployed to further analyze energy relations between edda-type, nickel(II) and water fragments, since these kinds of com-plexes have not yet been characterized in such a way.

Results and discussionPreparation of aminocarboxylate edda-type ligands andcorresponding Ni(II) complexes and their characterization

We obtained O–N–N–O ligands using neutralized monochloro-acetic and 3-chloropropionic acid in the case of H2edap and1,2-dibromoethane and alanine in the case of H2S,S-eddp. Onthe other hand, for H2edap acid we were able to provide onlycondensation mixtures, being directly used for complexationwith Ni(II). All the complexes have been prepared by complex-ing the chloride salt of Ni(II) with neutralized acids underthe conditions given in the experimental section. Two newnickel(II) complexes have been characterized spectroscopicallyand by means of DFT (NBO/NEDA/MPA) theory.

This paper deals with Ni(II) complexes with ligands beingedda, edap, eddp, S,S-eddp OHMeedda and 1,3-pdda (Fig. 2)and two bulky ligands similar to edda, baboc and dacoda.

We were able to synthesize green-blue colored crystals of[Ni(S,S-eddp)(H2O)2] (1) and blue-green colored [Ni(edap)-(H2O)2] (2) complexes. Fig. 3 shows the isolated isomers of pre-pared Ni(II) complexes. Apart from the isomerism given inFig. 1, the H2edap acid upon complexation with the metal ionmay actually yield two different uns-cis isomers (Fig. 3), i.e.uns-cis (A) with a five-membered acetate ring in the axial posi-tion and uns-cis (B) with switched acetate and β-propionaterings. In the case of (1), s-cis geometry has been found as themost stable one and for Ni-edap complex (2) the favoredisomer has been found to be uns-cis A with a 3-propionate armin the equatorial position. The presence of the 6-memberedβ-alaninato ring relieves the in-plane strain, thus allowing the

Fig. 1 Geometrical isomerism of six-coordinate [Ni(edda-type)(H2O)2]complexes.

Paper Dalton Transactions

13358 | Dalton Trans., 2013, 42, 13357–13368 This journal is © The Royal Society of Chemistry 2013

Publ

ishe

d on

25

July

201

3. D

ownl

oade

d by

Uni

vers

ity o

f B

elgr

ade

on 0

9/05

/201

4 11

:55:

49.

View Article Online

complex to attain bite angles closer to the octahedral idealwith apparently less strain in the Ni–L bonds.

The diaminodicarboxylate ligand, S,S-eddp related to edda,forming five-membered rings with the known absolute con-figuration (Fig. 3) upon coordination to a M(II) ion gives acomplex keeping the S configuration of the both carbon atomswith a ΔΔΔ (net Δ) absolute configuration.28 However, theorientation of the methyl groups of the S,S-eddp ligand, aftercoordination to metal ion, in the complexes can be differentand the two diastereoisomers can theoretically be formed forevery geometrical isomer (Fig. 3) giving rise to SSSS configur-ation or SSRR as found in experimental structures (S,S attribu-ted to carbon and R,R to nitrogen stereochemistry). Apparentlyadditional conformational diastereoisomers arise from adifference of the chirality around one of the asymmetricN atoms in a coordinated edda-type ligand. Though such a dia-stereoisomer has not been found so far, it may be seen for[Ni(baboc)(H2O)]

11 that the total RRRR configuration on bothcarbon and nitrogen atoms can be reached by introducing abulky substituent in AA(Et)AA (Fig. 4).

It is evident from Fig. 4 that in case of nickel(II)-edda-typecomplexes, the s-cis configuration keeps its form whether wechange carboxylate shape (even large enough to preventadditional coordination of the second water molecule aswas the case for [Ni(baboc)(H2O)] complex) or introduce

substituents on the CH2 of carboxylate or nitrogens of diamine(with +I or −I effects: [Ni(S,S-eddp)(H2O)2] or [Ni(OHMeedda)-(H2O)2]

19). However, extending the diamine (1,3-pdda) or car-boxylate (asymmetric “edap”) length for the –CH2– groupusually makes a change in the starting configuration i.e. thesegive rise to favored uns-cis isomers. Here it is worthwhile tomention that uns-cis-[Ni(1,3-pdda)(H2O2)2] gives only one S,Sconformational diastereoisomer. Introducing bulky diamine(1,5-diazacyclooctane diamine), as seen in the square-pyrami-dal [Ni(dacoda)(H2O)] complex, results in a transformationfrom s-cis to quasi-trans geometry.

Let us explain this difference in spatial geometry distri-bution between Ni(II) complexes of edda derivatives. Thechanges from s-cis geometry happen only when one disturbsthe core of the ethylenediamine ring by introducing, forinstance a bulky ring (dacoda ligand) or more than one –CH2–

group (1,3-pdda ligand). These disruptions often lead to adifferent square-pyramidal (or quasi-trans Fig. 1) geometry orthe formation of uns-cis isomer. On the other hand, no matterwhere we bring a substituent in with N(diamine) or C(carboxy-late) atoms, or even introduce a bulky baboc ligand coupledwith both N and C atoms, s-cis geometry always dominates (seeFig. 4).

Description of the crystal structure of s-cis-[Ni(S,S-eddp)-(H2O)2] (1)

The adopted atom-numbering scheme of the atoms and theview of the puckering of the title compound are shown in theORTEP drawings of Fig. 5; the packing of the molecules isshown in the unit cell as well. The asymmetric unit containsone formula unit molecule with no atoms set at a special posi-tion. The monoclinic unit cell contains two formula units ofthe title compound. The chiral centers of N1 and N2 both

Fig. 2 Diaminedicarboxylic acids considered in this work (see Fig. 4).

Fig. 3 Geometrical isomers of Ni(II) complexes prepared with edda-typeligands.

Fig. 4 Distribution of different geometrical isomers deriving from [Ni(edda)-(H2O)2] complex.

Dalton Transactions Paper

This journal is © The Royal Society of Chemistry 2013 Dalton Trans., 2013, 42, 13357–13368 | 13359

Publ

ishe

d on

25

July

201

3. D

ownl

oade

d by

Uni

vers

ity o

f B

elgr

ade

on 0

9/05

/201

4 11

:55:

49.

View Article Online

showed the R-configuration and the chiral centers of C2 andC7 both showed the S-configuration.

The search for the distances yielded intermolecular andintramolecular contacts shorter than the sum of the van derWaals radii for the atoms: the moieties are linked by hydrogenbonds (Table S1, ESI†) forming an infinite three-dimensionalnetwork along the base vectors.

In the crystal structure the Ni(II) ion is coordinated by sixdonor atoms originating from two deprotonated oxygen atomsof carboxylate groups, two oxygen atoms from water and twonitrogen atoms from diamine ring. The complex has a dis-torted octahedral geometry. The positions of the carboxylategroups define the s-cis geometry of this complex, as twoacetate rings occupy an axial position around nickel, while twowaters are in the equatorial plane. There are two longer Ni–Nbonds: Ni–N1 (2.133 Å) and Ni–N2 (2.118 Å). The other dis-tances and bonds are given in Table 1. The cis angles are inrange 81.229 Å to 97.385 Å and trans angles in the range168.702 Å to 171.955 Å. The [Ni(S,S-eddp)(H2O)2] unit containsa similarly twisted “envelope” conformation of five-membered

rings. The equatorial ethylene-diamine ring is found in afavored twisted conformation. The puckering parameters q2and φ2, which relate to the deviations of the atoms from themean plane of the ring, are q2 = 0.457(3) Å, φ2 = 269.2(2)°. Thetwo axial five-membered acetate rings have a slightly twistedenvelope conformation (almost planar). Their puckering para-meters are q2 = 0.235(2) Å, φ2 = 117.6(5)° (NiO1C1C2N1) andq2 = 0.165(2) Å, φ2 = 127.5(7)° (NiO3C6C7N2). These should becompared to the ideal values of q2 > 0 and φ2 = 0 for the envel-ope conformation versus q2 > 0 and φ2 = 90° for the twistedconformation (see Table S2, ESI† showing the basic confor-mations of N-membered rings). Slightly shorter Ni–O bonds(both Ni–O(ring) and Ni–Ow) for s-cis-[Ni(baboc)(H2O)] withrespect to Ni–O bonds of other complexes may be alsoobserved. This means that a decrease in number of watermolecules effectively strengthens the electrostatic interactionbetween the nickel(II) ion and oxygen donors.

Computational part

Having a series of edda-like complexes with 5- and 6-mem-bered carboxylate chelates, we can now compare their experi-mental spectroscopic and geometric properties with theresults of DFT calculations. We optimized the geometries ofeach of the possible geometric isomers (Fig. 1). Quantumchemical calculations of anions often lead to occupied orbitalshaving physically unreasonable positive energies but, aspointed out by Deeth and Fey,29 ionization and/or ligand dis-sociation is prevented by the finite size of the basis set. In ourcase the neutral species optimized by DFT in gaseous statedistort their geometry caused by intramolecular hydrogenbonds between water hydrogens and carbonyl oxygens fromthe ligand. Therefore, we include a PCM polarizable

Fig. 5 X-ray crystal structure of s-cis-[Ni(S,S-eddp)(H2O)2]: ORTEP diagram ofthe [Ni(S,S-eddp)(H2O)2] and crystal packing view along c axis.

Table 1 Comparison of experimental and DFT data for edda-type-Ni(II) complexes

Parameters

s-cis-[Ni(baboc)(H2O)]11 s-cis-[Ni(S,S-eddp)(H2O)2] s-cis-[Ni(edda)(H2O)2]

18,19

DFT X-ray DFT X-ray DFT X-ray

Ni–N 2.094 2.057 2.125 2.133 2.131 2.1002.144 2.114 2.125 2.118 2.122 2.101

Ni–O 2.004 1.996 2.038 2.054 2.044 2.0502.010 1.985 2.037 2.075 2.044 2.064

Ni–Ow 2.076 1.996 2.153 2.085 2.146 2.0612.150 2.070 2.147 2.077

trans angles 165.813 165.689 178.026 171.955 178.315 174.850156.330 157.220 171.368 168.702 171.722 173.154

171.544 171.015 171.977 169.344cis anglesO–Ni–Ow 88.202 88.745 91.538 95.330 91.276 93.524

88.534 90.211 89.930 91.973 90.775 91.98489.823 87.610 89.510 87.99491.493 95.562 90.743 92.924

N–Ni–Ow 115.593 114.897 92.723 97.385 96.854 93.11492.982 90.165 91.441 92.234

O–Ni–N 83.978 83.257 82.129 81.343 82.506 81.854109.846 108.706 96.398 95.965 96.076 96.11483.334 84.638 82.111 81.229 82.351 82.03494.287 92.336 96.432 90.863 96.726 93.174

N–Ni–N 88.087 87.882 85.120 84.602 84.823 84.654Ow–Ni–Ow 90.329 89.166 87.867 90.927

Paper Dalton Transactions

13360 | Dalton Trans., 2013, 42, 13357–13368 This journal is © The Royal Society of Chemistry 2013

Publ

ishe

d on

25

July

201

3. D

ownl

oade

d by

Uni

vers

ity o

f B

elgr

ade

on 0

9/05

/201

4 11

:55:

49.

View Article Online

continuum model (water as dielectric) as a convenient way ofmitigating the structural effects of charge influence on metal–water bonding and excessively long M–L bonds.

Table 1 compares theoretical and experimental bond dis-tances and angles for the series. Relative energies are obtainedfrom the calculated total energies without corrections for basisset superposition error since these are not very large with DFT,especially with a large TZVP30 basis set. In general, the com-parison between theory and experiment is reasonably good.B3LYP/TZVP(PCM) gives systematically longer Ni–Ow distancesthan those observed in solid state structures by around 0.05 Å.However, the energetic consequences of changing the Ni–Ow

bonds even by a few tenths of an Ångström are minimal.The crystallographically observed geometries correspond to

the lowest energy of the DFT structure although the complexesgive structures which are quite different in energy (Table 2).Since we were not able to prepare suitable crystals for an X-raystructure determination of the [Ni(edap)(H2O)2] complex,despite many crystallization techniques used, the proposeduns-cis (A) structure is a prediction based on the DFT energies(Table 2, Fig. 6). However, as described later on, this predictionis consistent with the spectral results and our expectations arebased on work with such systems.

Stabilization energies

To determine the stabilization energies (SE) of the complexesformed with different ligands we undertook calculations on a

substitution reaction:

½NiðH2OÞ6�2þ þ edda-type2� ¼ ½Niðedda-typeÞðH2OÞn�þ ð6� nÞH2O

using the DFT method, where

SE ¼ Ef½Niðedda-typeÞðH2OÞn�g þ Efð6� nÞH2Og� Ef½NiðH2OÞ6�2þg � Efedda-type2�g

Our data describe a diverse stability trend of chelatesformed, with an increase in the size of the rings being pro-duced upon complexation, and with the N*, C*CO, NC*N-sub-stituted ligands containing the ethylenediamine and 1,3-propanediamine backbones. The stabilities of metal complexesof polyamino dicarboxylates increase in the order Ni-OHMeedda > Ni-edda ∼ Ni-baboc > Ni-eddp > Ni-dacoda > Ni-edap > Ni-1,3-pdda > Ni-S,S-eddp (Table 3).

Courtney, Chaberek and Martell31 have reported stabilityconstant data for edda-type Ni(II) complexes. The measuredlog β values (water–KNO3 media; potentiometry) for Ni(II)-edda, and Ni(II)-eddp are 13.5 and 9.3, respectively, indicatinga trend towards less stable complexes as the number of six-membered chelate rings increases. The corresponding freeenergies for Ni(II)-edda, and Ni(II)-eddp are −18.42 and−12.69 kcal mol−1 respectively. As seen in Table 3, the SE datakeep to the above trend. However, it is noteworthy to empha-size that the somewhat bulky baboc and dacoda ligandsprevent the coordination of a sixth donor atom (water mole-cule), spreading the charge distribution of the whole complexto the atoms in such a way as to keep the complex more elec-trostatically unique (see MPA discussion bellow).

Natural bond orbital analysis

NBO output from the Gaussian calculations was analyzedusing the NBO 5.0 package.24 Energy values based onthe donor–acceptor (D–A) mechanism (a second orderperturbation theory analysis of the Fock matrix in the NBObasis) have been obtained (see Table S3, ESI†). The NBOanalysis does find prevalently three-directional 3-center2-electron (β spin) metal–ligand hypovalent bonds (Table 4,Fig. 7).

It is noteworthy to mention that these hypovalent bonds areof weak character, being built up only from β electrons. β elec-trons are far more sensitive, reflecting many characteristics ofthe molecule such as spectroscopic features or the formationof atom–atom covalent bonds. For example in the case of Ni–edda complex, the α HOMO–LUMO gap of 6.524 eV differsfrom β HOMO–LUMO, being 5.771 eV, according to molecularorbital analysis. β electrons tend to form more bonds than αbecause there are often vacant valence orbitals that serve asgood electron acceptors for bonding interactions in the β-spinsystem, whereas the corresponding orbitals in the α-spinsystem are filled. This is certainly what we are seeing at the Nicenter, which is nominally 3d8 (filled in the α system, but twovacancies in the β).

Table 2 Comparison of DFT energies for isolated edda-type-Ni(II) complexesa

Ligand S,S-eddp edap

Number of isomers3 4

Relative energies (kcal mol−1)

s-cis 0 21.94uns-cis (A)b 7.71 0uns-cis (B)b 23.21trans 7.65 18.91

a The isomer with the lowest energy minimum has been indicated with0 kcal/mol. b Applied for isomers of Ni-edap complex.

Fig. 6 Left: overlay of X-ray and DFT optimized structures of the s-cis-[Ni(S,S-eddp)(H2O)2] complex. Right: DFT optimized structure of the uns-cis(A)-[Ni(edap)(H2O)2].

Dalton Transactions Paper

This journal is © The Royal Society of Chemistry 2013 Dalton Trans., 2013, 42, 13357–13368 | 13361

Publ

ishe

d on

25

July

201

3. D

ownl

oade

d by

Uni

vers

ity o

f B

elgr

ade

on 0

9/05

/201

4 11

:55:

49.

View Article Online

Rather, the interaction is generally electrostatic and thecomplexes are fragmented into three or four separate units:[Ni(II)(edda-type)] and one or two water molecules. Only Ni(II)with 1,3-pdda gives all four possible fragments, namely: Ni(II)cation, 1,3-pdda dianion, and two molecules of water. Theelectrostatic forces (Donor–Acceptor mechanism) are respon-sible for preserving the framework of Ni-edda-type molecules.The largest D–A energy transfer results from a lone “p” elec-tron pair of carboxylate oxygen atoms as well as from spn

tetrahedral nitrogen atoms, toward the 1-center (lone) virtual“non-Lewis” (excited) LP(S)* orbital of Ni(II).32

Molecular orbital analysis

We used AOMIX33 to calculate percentage contributions ofdifferent molecular fragments (nickel(II), edda-type ligandsand water molecules) to molecular orbitals (MO) using Mulli-ken population analysis (MPA). Therefore, the total, partialand overlap population density-of-states (TDOS, PDOS andOPDOS) have been plotted in order to analyze the nature ofthe chemical bonding of the complexes. The main use of theDOS plots for molecules is to provide a pictorial representationof the MO compositions and their contributions to chemicalbonding through the OPDOS plots which are also referred toin the literature as crystal orbital overlap population (COOP)diagrams.33 In a spin-unrestricted wave function, the α- andβ-spin molecular orbitals are not necessarily orthogonal to oneanother (only within each set, either α-MOs or β-MOs, are allof the molecular orbitals mutually orthogonal to one another).Thus, there are cases of interest such as our case where it isrelevant to evaluate the overlap integrals between α- and β-spinMOs.

We followed the AOMIX recommendation and all the singlepoint (SP) calculations were done with Ahlrichs’ TZVP30 basisset in order to obtain more realistic results. Since there are nooverlapping contributions from α spin molecular orbitals, inFig. 8 selected TDOS, PDOS and OPDOS diagrams were plotted

Table 3 Stabilization energies SE (kcal mol−1)a for reaction: [Ni(H2O)6]2+ + edda-type2− = [Ni(edda-type)(H2O)n] + (6 − n)H2O

b

Complex

Ni-edda Ni-SSeddp Ni-edap Ni-eddp Ni-1,3-pdda Ni-OHMeedda Ni-baboc Ni-dacoda

n = 2 n = 1

SE −35.35 −15.52 −22.34 −29.77 −21.11 −37.35 −35.12 −26.08

a Energies were calculated by their compensation for thermal and ZPE corrections. b The calculation for reactants and products were conductedby the G03 suite under the conditions: (U)B3LYP/TZVP/PCM (water).

Table 4 3-Center, 2-electron A:–B–:C hypovalent bonds (β orbitals) (A–B:C <=> A:B–C) of selected [Ni(edda-type(H2O)n] complexes

Hypovalent bond A:–B–:C %A–B/%B–C OCC.

NBOs 3-Center hybrids

BDa (A–B) LPb (C) h(A) h(B) h(C)

eddaN 20:-Ni 1-:O 2 65.7/34.3 1.9789 2 59 3 4 90N 21:-Ni 1-:O 5 64.2/35.8 1.9789 3 61 5 6 92OHMeeddaN 22:-Ni 1-:O 5 62.3/37.7 1.9683 1 72 1 2 110N 24:-Ni 1-:O 2 62.7/37.3 1.9697 2 70 3 4 108O 37:-Ni 1-:O 35 50.7/49.3 1.9467 3 80 5 6 1181,3-pddaNo 3-center hypovalent bondsBabocO 48:-Ni 1-:O 24 53.1/46.9 1.9553 3 103 5 6 163

a BD means bonding orbital. b LP means lone pair.

Fig. 7 Selected 3-center, 2-electron A:–B–:C hypobonds (β orbitals) (A–B:C <=>A:B–C) of three Ni-edda-type complexes.

Paper Dalton Transactions

13362 | Dalton Trans., 2013, 42, 13357–13368 This journal is © The Royal Society of Chemistry 2013

Publ

ishe

d on

25

July

201

3. D

ownl

oade

d by

Uni

vers

ity o

f B

elgr

ade

on 0

9/05

/201

4 11

:55:

49.

View Article Online

for β spin molecular orbitals in the case of s-cis-[Ni(edda)-(H2O)2].

The first part of Fig. 8 (left) describes the TDOS with thePDOS of particular fragments (nickel(II), water and edda2−) inthe case of the [Ni(edda)(H2O)2] complex. One may see thatedda and nickel ions make the largest PDOS contribution. Theright part shows low overlap integrals (below 0.05) for thewater and edda fragments, leading to low covalent bondingcontribution. On the contrary, nickel(II) and edda fragmentsshow quite positive overlap integrals (about 0.5). Nickel(II) over-laps with water as well, although it shows five times weakercovalent interactions. Fig. 9 shows OPDOS plots for Ni(II)/edda-type and Ni(II)/water fragments of selected [Ni(edda-type)-(H2O)2] complexes. Obviously the principal OPDOS MOs orig-inate from the covalent interaction between nickel(II) and theedda-type ligand. This coincides with NBO results where a3-centric hybrid and in some cases covalent Ni–O(carboxylate)bonds were found. Probably Ni–O(water) NBOs were notfound because of their lower covalency (overlap integralbellow 0.1).

Natural energetic decomposition analysis (NEDA)

We were also interested to find out the energy relationship ofthe (edda-type)Ni⋯OH2 interaction as these usually reflecthow easily the pentacoordinated (or tetracoordinated)complex turns on the hexacoordinated one (Table 5). There-fore, we have done NEDA25–27 calculations using Firefly34 EES.Intramolecular forces and energy interactions inside the mole-cule are the main characteristics of the NEDA results.

In Table 5 the parameters of interest to us are listed asfollows: CT stands for charge transfer; ES stands for electro-static component; POL stands for polarization component; XCstands for exchange-correlation component; DEF stands fordeformation component and ΔE stands for total energy inter-action:

ΔE ¼ CTþ ESþ DEFþ BSSE;

where BSSE stands for basis set superposition error.The charge transfer (CT) and electrostatic component (ES)

terms provide a bigger contribution to the total energy

Fig. 9 Overlap population density-of-states plot of nickel(II)-edda-type andnickel(II)-water within [Ni(edda-type)(H2O)2] complexes.

Table 5 Contribution of NEDA energies (kcal mol−1) for [Ni(edda-type)(H2O)2−n]–(H2O)n decomposition

edda-type CT ES POL XC DEF(SE)·H2ODEF(SE)Ni-edda-type

μ(ind)a

(edda-type) ΔE

Δ(ΔE)difference

Dimer Trimer

edda n = 1 1 −62.49 −42.72 −60.25 −22.91 121.70 (20.90) 48.99 (9.39) 1.01(9) −17.68 −28.16 −23.68n = 1 2 dimer −106.44 −83.88 −136.75 −42.58 234.47 (47.69) 89.33 (15.56) 1.97(9) −45.84n = 1 1 1 trimer −122.51 −86.58 −117.65 −43.92 118.26(19.48),

121.70(20.92)89.33(18.41) 1.97(9) −41.36

S,S-eddp n = 1 1 −61.17 −42.04 −58.75 −23.79 120.40 (20.19) 48.76 (9.35) 0.96(9) −16.19 −28.56 −23.16n = 1 2 dimer −106.42 −84.04 −136.94 −43.81 234.87 (50.37) 91.60 (19.02) 1.91(3) −44.75n = 1 1 1 trimer −121.93 −86.31 −117.42 −45.16 119.47(19.96),

120.40(20.19)91.60(18.87) 1.91.(3) −39.35

OHMeedda n = 1 1 −72.66 −48.22 −54.42 −24.28 120.35 (19.78) 58.86 (7.67) 1.04(8) −20.37 −25.34 −22.95n = 1 2 dimer −131.67 −94.88 −125.02 −46.70 237.38 (47.69) 115.18 (15.56) 1.73(8) −45.71n = 1 1 1 trimer −145.46 −96.99 −109.04 −47.77 120.42(19.74),

120.35(19.73)115.18(15.48) 1.73(8) −43.32

edap n = 1 1 −70.19 −51.56 −51.45 −23.54 115.52 (20.34) 61.28 (5.74) 1.32(2) −19.94 −30.59 −27.72n = 1 2 dimer −133.61 −104.51 −125.78 −45.46 236.11 (50.60) 122.71 (14.02) 2.41(2) −50.53n = 1 1 1 trimer −148.10 −107.64 −108.17 −46.72 124.75(20.90),

115.52(20.44)122.71(13.93) 2.41(2) −47.66

1,3-pdda n = 1 1 −66.41 −47.67 −48.22 −22.49 110.64(18.63) 56.50(5.78) 1.27(4) −17.65 −28.08 −25.45n = 1 2 dimer −119.48 −96.97 −117.13 −43.16 221.01(45.78) 110.00(14.17) 2.24(8) −45.73n = 1 1 1 trimer −133.64 −99.81 −100.67 −42.92 110.61(18.66),

114.90(18.47)109.93(14.11) 2.24(6) −43.10

dacodab −60.33 −47.39 −64.69 −25.40 124.98(23.04) 50.81(9.52) 0.714 −22.03

aDipole moments in Debye. bOnly one possible [Ni(dacoda)]·OH2 interaction.

Fig. 8 Total and overlap population density-of-states plot of nickel(II), edda-type ligand and water within [Ni(edda)(H2O)2].

Dalton Transactions Paper

This journal is © The Royal Society of Chemistry 2013 Dalton Trans., 2013, 42, 13357–13368 | 13363

Publ

ishe

d on

25

July

201

3. D

ownl

oade

d by

Uni

vers

ity o

f B

elgr

ade

on 0

9/05

/201

4 11

:55:

49.

View Article Online

interaction than the others. The NEDA calculations went intwo directions: (a) firstly we followed interaction between [Ni(edda-type)(H2O)] and the second H2O molecule, and (b) sec-ondly we reviewed NEDA parameters for interaction between[Ni(edda-type)] and both H2O molecules. As seen from Table 5,values of ΔE for the second case are roughly equal for the com-plexes in question. The only exception is [Ni(edap)]⋯2OH2

with ΔE being higher than other energies by about 5 kcalmol−1. The next order has been established in the case of (a):OHMeedda > edap > edda > S,S-eddp following energydecrease: −20.37, −19.94, −17.68 and −16.19 kcal mol−1,respectively. The energy difference (see Table 5) give us true ΔEfor [Ni(edda-type)]⋯O2H forces following a not exactly self-explanatory order: edap(−30.59 kcal mol−1) > S,S-eddp(−28.56 kcal mol−1) > edda(−28.16 kcal mol−1) > 1,3-pdda(28.08 kcal mol−1) > OHMeedda(−25.34 kcal mol−1). However,in the case of trimer H2O⋯[Ni(edap)]⋯OH2, Δ(ΔE) gives morereasonable energies with the next order: edap(−27.72 kcalmol−1) > 1,3-pdda(−25.45 kcal mol−1) > edda(−23.68 kcalmol−1) > S,S-eddp(−23.16 kcal mol−1) > OHMeedda(−22.95 kcal mol−1). NEDA suggests that electrical and chargetransfer interactions both contribute importantly to bondingforces in nickel(II) complexes. Energy interactions in the firstwater molecule release account for CTs, and electrical inter-action (EL = ES + POL + SE) between fragments explain themost stable adduct between [Ni(OHMeedda)(H2O)] and H2Ounits. This is the result of the transfer of electrons from a lonepair of the water oxygen donor into the virtual (LP)Ni* non-Lewis orbital of the [Ni-(OHMeedda)(H2O)] acceptor reflectedby the large charge transfer component (CT = −72.66 kcalmol−1). Polarization effects are modest, contributing−54.42 kcal mol−1 of the total interaction energy. CT and ELalso stabilize the most stable trimer [Ni(edap)]⋯2H2O by−148.10 and −201.88 kcal mol−1, respectively. However, polar-ization effects are enhanced in the trimer, as evidenced by theincreasing dimer or trimer dipole moments, from about 1.32(2) to 2.41(2) D (see Table 5). POL effects make stronger trimerinteractions, by about −57 kcal mol−1. Consequently, abstract-ing the second water molecule from the [Ni(edda-type)] unitcauses an energy penalty for each of the monomers to undergopolarization, reflecting in different interaction energies. Thehighest Δ(ΔE) might be expected for uns-cis-[Ni(edap)(H2O)2]and uns-cis-[Ni(1,3-pdda)-(H2O)2] complexes, meaning that theabstracting of two water molecules from an unsymmetricalframework leads to higher induced polarization than that inthe case for the symmetrical-cis isomer.

The binding energies differ somewhat for the “dimer” and“trimer” calculations. These energies would be identical if thewater molecules were non-interacting, that is if the 2(H2O)units of our complexes were isoenergetic with two free H2Omolecules. These water molecules, which coordinate to ourcomplexes at adjacent octahedral sites, repel each otherslightly, by 4–5 kcal mol−1. Evaluating the energy of 2(H2O) atthe same geometries as possessed by these water molecules inthe complex, we found that the dimer in this geometry isslightly less stable than two separated monomers.

All the NEDA calculations have been done using HFmethod as well. DFT and HF methods differ for just a fewkcal mol−1 but the energy order is almost the same.

Spectral analysis

The complexes have been further analyzed by means of IR andUV-vis spectra.

The IR data (carboxylate region) are in agreement with thestructures and molecular symmetries. In the case of the s-cis-[Ni(S,S-eddp)(H2O)2] complex with C2 symmetry, the IR spec-trum contains only one sharp band centered on 1581 cm−1.Normally, just one band is expected due to the asymmetricvibrations of equivalent five-membered acetate rings. In con-trast, uns-cis-[Ni(edap)(H2O)2] has C1 molecular symmetry andwe obtain a nice correlation in that the spectrum contains oneintense wider band positioned at 1575 cm−1 corresponding toasymmetric vibrations of the β-propionate ring and oneshoulder of moderate intensity located at 1651 cm−1 due toasymmetric stretches of the acetate ring (see Fig. S4 and S5,ESI†). This interpretation is in agreement with the generallyaccepted rule that the frequency assigned to five-memberedrings lies at a higher energy than the corresponding frequencyof six-membered chelate rings.35,36 For protonated carboxylategroups (1700–1750 cm−1) and for coordinated carboxylategroups (1600–1650 cm−1) asymmetric carboxylate stretchingfrequencies have been well established.35,36 Therefore, theabsence of the band in 1700–1750 cm−1 region confirms thatall the carboxylate groups are deprotonated.

The ligand field absorption spectra are now considered forall the complexes. Table 6 lists the relevant electronic absorp-tion data of the investigated complexes including TDDFT cal-culations for the ten lowest energy transitions. Since thecomplexes studied have actual symmetries lower than octa-hedral, it might be expected that the absorption bands wouldshow some evidence of deviation from the ligand field ofpurely Oh symmetry. Indeed in the cases of uns-cis complexes(Ni-eddp, Ni-edap and Ni-1,3pdda), an obvious splitting of thefirst absorption band (low-energy side) was observed due tothe influence of tetragonality on the spectrochemical behaviorof these Ni(II) complexes. Apparently, if the equatorial N2O2

ligand field is lower than the one imposed by two axially co-ordinated oxygen atoms (Ni-edda and Ni-S,S-eddp), the D4h

model fit better with the pattern, with the lower energy peakbeing 3B1g → 3B2g(D4h) and the higher energy peak being3B1g →

3Eg(D4h) both of which derived from the same 3T2g(Oh)state.37 This is further evidence that the geometry of preparedNi-edap differs from Ni-S,S-eddp complex.

Therefore, the best spectral interpretation of Ni(II) com-plexes can be achieved over the (Oh) model: 3A2g → 3T2g(F)(band I), 3A2g →

3T1g(F) (band III) and 3A2g →3T1g(P) (band IV)

(Table 6).All the complexes are light-blue and experimentally exhibit

three absorption bands. Each compound contains shoulder atabout 13 000 cm−1 (band II) on the high-energy side of the firstspin-allowed transition (see Table 6) irrespective of the

Paper Dalton Transactions

13364 | Dalton Trans., 2013, 42, 13357–13368 This journal is © The Royal Society of Chemistry 2013

Publ

ishe

d on

25

July

201

3. D

ownl

oade

d by

Uni

vers

ity o

f B

elgr

ade

on 0

9/05

/201

4 11

:55:

49.

View Article Online

underlying approximate C1 or C2 symmetry. This appearanceoccurs as a result of spin–orbital coupling (see Fig. S6 and S7,ESI†).

The bands I, III and IV (usually reflecting an average ligandfield strength – LFS) of uns-cis Ni-edap, Ni-eddp and Ni-1,3-pdda were compared to the s-cis Ni-edda or Ni-S,S-eddp com-plexes. Generally, the bands belonging to complexes with uns-cis geometry are moved to higher energy, which means thatthey have a stronger ligand field. Comparing Ni-1,3-pdda toNi-edda complexes, the LFS shift occurs as a consequence ofthe presence of one of the carboxylate rings in the equatorialplane that exert greater influence on the d-orbital along thex and y axes.

The TDDFT calculations do not provide as clear a corre-lation with the band positions as in experiment. The tenlowest energy transitions were considered and the transitionwith the largest oscillator strength was taken for comparison.Actually, all the transitions were found, however the secondband (around 13 000 cm−1) appears as a result of spin–orbitalcoupling and often is not positioned well. The position of thefourth band (around 27 000 cm−1) used not to be taken as anaverage of ligand field strength. Accordingly, here we reportedtwo clearly defined DFT bands just for comparison purposes.TDDFT appears systematically to overestimate the transitionenergy. A scaling factor of 0.8 improves the numerical agree-ment between theory and experiment (see Table 6) but thesequence of complexes as a function of band maximum israther different to experiment. For complexes such as [Ni-(edda)(H2O)2] and [Ni(S,S-eddp)(H2O)2], it may be that, in solu-tion, there is an equilibrium between six-coordinated and five-coordinated forms although this would presumably impact thetetragonality.

Conclusions

In summary, we prepared two new nickel(II) complexes withtetradentate ligands of an N2O2 chromophore. In the case ofS,S-eddp the proposed s-cis geometry has been verified byX-ray analysis while the edap ligand encircles nickel(II) ion inthe uns-cis fashion. This prediction has been established bymeans of IR and UV-vis spectroscopy and by DFT modelling ofall possible isomers. The uns-cis(A) geometry corresponds tothe lowest energy structure computed using DFT. Further, wedemonstrated the structural dependence of the favored isomerand substituents introduced in the edda ligand. Disturbingthe diamine core and carboxylate length inevitably lead to achange of s-cis geometry. The structural properties of the [Ni(S,S-eddp)(H2O)2] complex have been explained in detail as well asthe comparison with similar complexes. The stabilization ener-gies (SE) show increased stability in the order Ni-OHMeedda >Ni-edda ∼ Ni-baboc > Ni-eddp > Ni-dacoda > Ni-edap > Ni-1,3-pdda > Ni-S,S-eddp. Computational chemistry involved NBO/NEDA and MPA analysis. The most important finding reflectselectron description of the molecules investigated. NBO findsthat β electrons construct weak bonding orbitals betweennickel(II) and the edda-type ligand (an exception is Ni-1,3-pdda). However, the whole Ni-edda-type-H2O molecule israther held together by electrostatic forces. Molecular orbital(MPA) analysis confirms weak contribution to bonding MOsbetween nickel(II) and the edda-type ligand. An OPDOSdiagram (Fig. 6) shows bonding and antibonding MOs of thesefragments and their energy relationship. NEDA gave us insightinto interactions between [Ni(edda-type)] and coordinatedwater molecules. The total energy interaction of [Ni(edda-type)-(H2O)]⋯H2O fragments accounts for a high charge transfer

Table 6 Experimental and TDDFT UV-Vis spectral data of Ni-edta-type complexes

Complex

Exp. value in TDDFT: scaled by 0.8

Ref.ν (103 cm−1) ε (l mol−1 cm−1) ν (103 cm−1) Assignments (Oh)

[Ni(edda)(H2O)2] 9.90 12.6 10.24 I 3A2g →3T2g(F) 18

13.20 1.9 II →1Eg (D)16.40 4.2 15.36 III →3T1g (F)27.20 7.2 IV →3T1g (P)

[Ni(S,S-eddp)(H2O)2] 9.84 14.1 10.34 I This work13.19 2.6 II16.23 5.8 15.31 III27.03 9.2 IV

[Ni(edap)(H2O)2] 9.94 sh 15.6 10.85 I This work10.57 1713.57 8.3 II16.34 12.1 16.61 III26.95 20.2 IV

[Ni(eddp)(H2O)2] 9.75 sh 8.4 10.14 I 2011.7 15.613.44 4 II16.67 6.6 16.20 III27.10 10.6 IV

[Ni(1,3-pdda)(H2O)2] 9.65 sh 5.8 10.93 I 2011.11 9.213.44 3.3 II16.81 5.0 16.38 III27.32 7.8 IV

Dalton Transactions Paper

This journal is © The Royal Society of Chemistry 2013 Dalton Trans., 2013, 42, 13357–13368 | 13365

Publ

ishe

d on

25

July

201

3. D

ownl

oade

d by

Uni

vers

ity o

f B

elgr

ade

on 0

9/05

/201

4 11

:55:

49.

View Article Online

(CT) between units due to a strong donor–acceptor mechan-ism. However, the whole complex splitting into three units,namely [Ni(edda-type)] and two water molecules, leads to sig-nificant polar deformation of the edda-type unit, changing thedipole moment from 1.32(2) to 2.41(2) D in the case of edapligand. This results in the largest energy contribution of thepolarization (POL) component to the total energy interactionbetween water and the rest of molecule. Spectral analysis (IRand UV-vis) revealed common things observed earlier. Theposition of principal IR and UV-vis bands makes further differ-ence between geometries of complexes prepared in this work.TDDFT appears to systematically overestimate the transitionenergy. A scaling factor of 0.8 improves the numerical agree-ment between theory and experiment (see Table 6) but thesequence of complexes as a function of band maximum israther different to experiment.

ExperimentalMaterials

Ethylenediamine-N,N’-di-S,S-2-propionic acid was preparedusing a previously described procedure.38 Ethylenediamine-N-acetic acid dihydrochloride dihydrate Hedma·2HCl·2H2O wasprepared using a previously described procedure.39 The otherreagents were obtained commercially and used without furtherpurification.

Preparation of s-cis-(ethylenediamine-N,N’-di-S,S-2-propio-nato)(diaqua)nickelate(II), s-cis-[Ni(S,S-eddp)(H2O)2] (1). To asolution of 1.19 g (0.005 mol) of NiCl2·6H2O in 2.0 cm3 ofwater, 1.01 g (0.005 mol) of ethylenediamine-N,N’-di-S,S-2-pro-pionic acid in 3.0 cm3 of water previously neutralized with0.14 g (0.01 mol) of NaOH in 1 cm3 water, was added. Themixture was heated under stirring for 1 hour at 65 °C. Duringthis period, the acidity of the solution was maintained byadding a solution of NaOH (0.16 g, 0.004 mol) at pH = 7. Thesolution temperature was maintained for a further hour. Thismixture was filtrated and the filtrate was concentrated to asmall volume and desalted using Sephadex G-10 column (2.5 ×32 cm). The solution of the complex was evaporated to smallvolume and left in a fridge over night. The complex, as greencrystals, was separated by filtration and air-dried. The obtainedcrystals were suitable for X-ray structure analysis. Yield: 0.525 g(35.4%). Anal. Calcd for s-cis-[Ni(S,S-eddp)(H2O)2] =NiC8H18O6N2 (Mr = 296.93) (%): C, 32.36; H, 6.11; N, 9.44.Found (%): C, 32.57; H, 6.16; N, 9.48.

M.P. = 300 °C (with decomposition).

Ethylenediamine-N-acetic-N′-3-propionic acid, H2edap

The condensation mixture containing edta-type ligands(H2edap and H3eda2p) was prepared by using followingprocedure:

Ethylenediamine-N-acetic acid dihydrochloride dihydrate,Hedma·2HCl·2H2O, (13.62 g, 0.06 mol) was dissolved in H2O(40 mL) containing NaOH (7.20 g, 0.18 mol) (solution I).

3-Chloropropionic acid (6.51 g, 0.06 mol) was dissolved inH2O (7 mL) and cooled in an ice bath. A cooled solution ofNaOH (2.40 g, 0.06 mol in 5 mL of H2O) was added dropwise,the rate of addition being adjusted so that the temperatureremained below 15 °C (solution II). Solutions I and II weremixed, and the reaction mixture was stirred with heating(57 °C) for 9 hours. During this process, the necessary amountof NaOH was added dropwise to keep the pH in the 7–8 range.After that the volume of the resulting mixture was reduced to15 mL, and the deposited NaCl was separated by filtration.The obtained filtrate contains two edta-type ligands (H2edapand H3eda2p). The mixture was used directly for the prepa-ration of nickel(II) complexes that are more easily separated.

Preparation of (ethylenediamine-N-acetato-N′-3-propionato)-(diaqua) nickel(II), [Ni(edap)·(H2O)2] (2). The solution ofNiCl2·6H2O (14.26 g, 0.06 mol in 40 mL of water) was added tothe reaction mixture (0.06 mol) and it was stirred at 65 °C for1 h. After this process, the pH of the solution was 8.6 so thecondensation mixture was adjusted to 7.7 by the addition ofHCl solution and heating (65 °C) with stirring was continuedfor an additional 1 h. The solution was evaporated to 80 mL,and the deposited NaCl was removed by filtration. Theobtained filtrate was then desalted by passing it through G-10Sephadex column while eluting with distilled water. One thirdof this solution was introduced into a 5 × 60 cm column con-taining Dowex 1-X8 (200–400 mesh) anion-exchange resin inthe Cl form. The column was then eluted with H2O. Eluate ofneutral complex was evaporated to 10 mL and desalted bypassage through a G-10 Sephadex column, with distilled wateras the eluent, evaporated again to a small amount and thecomplex crystallized with the addition of 5 mL of ethanol.Light green crystals were collected and air-dried. Yield: 16%(1.0 g). Elemental Anal. Found: C, 27.68; H, 5.76; N, 9.16%.Calcd for C7H18N2O8Ni:C, 26.53; H, 5.72; N, 8.84%. M.P. =264 °C.

Physical measurements

Electronic absorption spectra were recorded on a Perkin-ElmerLambda 35 double-beam UV-Vis spectrophotometer. Aqueous1 × 10–3 mol L−1 solutions of the complexes was used for thesemeasurements. Infrared spectra were recorded on a Perkin/Elmer FTIR 31725-X spectrometer using KBR pellet technique.Elemental microanalyses for C, H, N were performed by stan-dard methods. Melting points were measured via the Stuartmelting device with accuracy ±1 °C.

Crystallographic data for 1 have been deposited with theCambridge Crystallographic Data Center (CCDC referencenumber 932753).† An X-ray analysis was performed on a blue-green crystal of s-cis-[Ni(S,S-eddp)(H2O)2] that was allowed togrow slowly from the water solution. The program suite SAINT-PLUS was used for space group determination (XPREP)40 andthe structure was solved by direct methods using the programSIR2004.41 Crystal data and numerical details on data collec-tion and refinement are given in Table 7 (Table S1, ESI†).

Paper Dalton Transactions

13366 | Dalton Trans., 2013, 42, 13357–13368 This journal is © The Royal Society of Chemistry 2013

Publ

ishe

d on

25

July

201

3. D

ownl

oade

d by

Uni

vers

ity o

f B

elgr

ade

on 0

9/05

/201

4 11

:55:

49.

View Article Online

Computational experiments

All calculations were performed using the Gaussian03 revisionE01 program.42 The calculations for searching local minimawere carried out at the density functional level of theory, usingthe Becke’s hybrid three parameter non-local exchange func-tional combined with the Lee–Yang–Parr gradient correctedcorrelation functional (B3LYP) and employing the all-electrontriple-ξ valence plus polarization basis set (TZVP) developed byAhlrichs and Weigend.30 As shown in our previous work,17 thismethod allows for the accurate description of structural pro-perties of nickel(II) complexes of edta-type ligands. Unrest-ricted formalism and geometry convergence criteria were usedand no symmetry constraints were imposed. All the calcu-lations were done under the Polarizable Continuum Model(PCM) with the solute being water as implemented in G03package. Zero point vibrational energies (ZPE) were evaluatedin frequency calculations at UB3LYP/TZVP level based onanalytical second derivates. These calculations confirmed thatall structures are minima. We found that energies calculatedwith TZVP basis set differ by just a few kcal mol−1 with andwithout correcting the basis-set superposition error (BSSE).Transition energies and oscillator strengths for electronic exci-tation to the first 10 singlet excited states of complexes werecalculated using TDDFT with the accurate UB3LYP/TZVPmethodology.

Orbital interactions have been analyzed by the NBO 5.0Gprogram32 at the UB3LYP/TZVP computational level. Thismethod allows analyses of the interaction between filled and

empty orbitals and associates them with charge-transfer pro-cesses. In addition, natural energy decomposition analysis(NEDA)25–27 has been carried out to obtain insights into mole-cular interactions between [LNi] and coordinated water mole-cules. These calculations have been performed within theFirefly program.34 Further, we used AOMIX software33 to calcu-late percentage contributions of different molecular fragments(nickel(II), edda-type ligands and water molecules) to mole-cular orbitals (MO) from output files generated by G03 suite.

Acknowledgements

The authors are grateful to the Serbian Ministry of Educationand Science for the financial support (project no. III41010 and172016). We also appreciate valuable help for NBO and NEDApart to Professor Eric Glendening. This manuscript is dedi-cated to the memory of Dr Vesna D. Miletić, who sadly passedaway on 13th June 2013.

Notes and references

1 T. Kowalik-Jankowska, H. Kozlowski, E. Farkas andI. Sovago, in Nickel and Its Surprising Impact in Nature, ed.A. Sigel, H. Sigel and R. K. O. Sigel, Wiley, New York, 2007,vol. 2, pp. 31–109, and references therein.

2 M. Kruppa and B. König, Chem. Rev., 2006, 106, 3520–3560.3 J. A. Weyh and R. L. Pierce, Inorg. Chem., 1971, 10, 858–860.4 D. J. Radanović, Coord. Chem. Rev., 1984, 54, 159–261.5 T. Sabo, S. Grguric-Sipka and S. Trifunovic, Synth. React.

Inorg., Met.-Org., Nano-Met. Chem., 2002, 32, 1661–1717.6 H. Nakazawa, H. Ohtsuru and H. Yoneda, Bull. Chem. Soc.

Jpn., 1987, 60, 525–530.7 T. Murakami, I. Hirako and M. Hatano, Bull. Chem. Soc.

Jpn., 1977, 50, 164–168.8 T. Murakami and M. Hatano, Bull. Chem. Soc. Jpn., 1976,

49, 3037–3041.9 V. Glodjović, F. Heinemann and S. Trifunović, J. Chem.

Crystallogr., 2008, 38, 883.10 K. Kawabe, M. Tadokoro, K. Hirotsu, N. Yanagihara and

Y. Kojima, Inorg. Chim. Acta, 2000, 305, 172–183.11 H. Pennemann, S. Wassmann, J. Wilken, H. Gröger,

S. Wallbaum, M. Kossenjans, D. Haase, W. Saak, S. Pohland J. Martens, J. Chem. Soc., Dalton Trans., 2000, 2467–2470.

12 R. J. Bianchini and J. I. Legg, Inorg. Chem., 1986, 25, 3263–3267.

13 T. J. Sabo, V. M. Đinović, G. N. Kaluđerović,T. P. Stanojković, G. A. Bogdanović and Z. D. Juranić, Inorg.Chim. Acta, 2005, 358, 2239–2245.

14 V. M. Djinović, V. V. Glodjović, G. P. Vasić, V. Trajković,O. Klisurić, S. Stanković, T. J. Sabo and S. R. Trifunović,Polyhedron, 2010, 29, 1933–1938.

15 B. E. Douglas and D. J. Radanović, Coord. Chem. Rev., 1993,128, 139–165.

Table 7 Crystal data and details of the structure determination for [Ni(S,S-eddp)-(H2O)2]

Formula C8H18N2NiO6Formula weight (g mol−1) 296.93Crystal system MonoclinicSpace group, no. [22] P21a (Å) 7.5532(5)b (Å) 9.1933(6)c (Å) 8.2722(5)β (°) 100.225(1)°V (Å3) 565.29(6)Θ Range unit cell: min.–max.(°);reflections

2.50–29.75; 3992

Formula Z 2Space Group Z 2Z′ (= Formula_Z/SpaceGroup_Z) 1ρcalc (g cm−3) 1.774F(000), electrons 312µ(Mo Kα) (cm−1) 15.64Color, habit Blue, blockApprox. crystal dimension (mm) 0.47 × 0.43 × 0.39Radiation type; λ (Å) Mo Kα, 0.71073θ range; min. max. (°) 3.12, 28.28Index ranges h: −9 → 9; k: −10 → 10;

l: −20 → 20Min.–max. absorption transmissionfactor

0.4894–0.5406

X-ray exposure time (h) 8.0Total data 7144Unique data 3710Data with criterion: (Fo ≥ 4.0σ(Fo)) 3596Rint = ∑[|Fo

2 − Fo2 (mean)|]/∑[Fo

2] 0.0122Rsig = ∑σ(Fo

2)/∑[Fo2] 0.0195

Dalton Transactions Paper

This journal is © The Royal Society of Chemistry 2013 Dalton Trans., 2013, 42, 13357–13368 | 13367

Publ

ishe

d on

25

July

201

3. D

ownl

oade

d by

Uni

vers

ity o

f B

elgr

ade

on 0

9/05

/201

4 11

:55:

49.

View Article Online

16 Z. D. Matović, V. D. Miletić, M. Ćendić, A. Meetsma,P. J. van Koningsbruggen and R. J. Deeth, Inorg. Chem.,2013, 52, 1238–1247.

17 S. Belološević, M. Čendić, A. Meetsma and Z. D. Matović,Polyhedron, 2013, 50, 473–480.

18 R. Wen, I. Bernal, F. Somoza, W. Li and F. R. Fronczek,Inorg. Chim. Acta, 1998, 282, 96–109.

19 C.-H. Ng, S.-B. Teo, S.-G. Teoh, H.-K. Fun and J.-P. Declercq,Inorg. Chim. Acta, 2002, 340, 81–86.

20 D. J. Radanović, V. Č. Matović, Z. D. Matović, L. P. Battaglia,G. Pelizzi and G. Ponticelli, Inorg. Chim. Acta, 1995, 237,151–157.

21 D. O. Nielson, M. L. Larsen, R. D. Willett and J. I. Legg,J. Am. Chem. Soc., 1971, 93, 5079–5082.

22 D. F. Averill, J. I. Legg and D. L. Smith, Inorg. Chem., 1972,11, 2344–2349.

23 W. E. Broderick, K. Kanamori, R. D. Willett and J. I. Legg,Inorg. Chem., 1991, 30, 3875–3881.

24 E. D. Glendening, C. R. Landis and F. Weinhold, WileyInterdiscip. Rev.: Comput. Mol. Sci., 2012, 2, 1–42.

25 E. D. Glendening and A. Streitwieser, J. Chem. Phys., 1994,100, 2900.

26 E. D. Glendening, J. Am. Chem. Soc., 1996, 118, 2473–2482.

27 E. D. Glendening, J. Phys. Chem. A, 2005, 109, 11936–11940.28 Nomenclature of Inorganic Chemistry: IUPAC Second

Edition - Definitive Rules 1970, Pure Appl. Chem., 1971, 28,1–110.

29 R. J. Deeth and N. Fey, J. Comput. Chem., 2004, 25, 1840–1848.30 F. Weigend and R. Ahlrichs, Phys. Chem. Chem. Phys., 2005,

7, 3297–3305.31 R. C. Courtney, S. Chaberek and A. E. Martell, J. Am. Chem.

Soc., 1953, 75, 4814–4818.32 E. D. Glendening, J. K. Badenhoop, A. E. Reed,

J. E. Carpenter, J. A. Bohmann, C. M. Morales andF. Weinhold, NBO 5.0, Theoretical Chemistry Institute, Uni-versity of Wisconsin, Madison, 2001.

33 S. I. Gorelsky and A. B. P. Lever, J. Organomet. Chem., 2001,635, 187–196.

34 A. A. Granovsky, 2009, http://www.classic.chem.msu.su/gran/firefly/index.html

35 D. H. Busch and J. C. Bailar, J. Am. Chem. Soc., 1953, 75,4574–4575.

36 M. L. Morris and D. H. Busch, J. Am. Chem. Soc., 1956, 78,5178–5181.

37 A. B. P. Lever, Coord. Chem. Rev., 1968, 3, 119–140.38 L. N. Schoenberg, D. W. Cooke and C. F. Liu, Inorg. Chem.,

1968, 7, 2386–2393.39 D. J. Radanović, M. I. Djuran, V. D. Miletić, C. Maricondi

and B. E. Douglas, Inorg. Chem., 1988, 27, 1265–1269.40 Bruker, SMART, SAINTPLUS and XPREP. Area Detector

Control and Integration Software. Smart Apex Software Refer-ence Manuals, Bruker Analytical X-ray Instruments Inc.,Madison, Wisconsin, USA, 2000.

41 M. C. Burla, R. Spagna, G. Polidori, C. Giacovazzo, L. DeCaro, G. L. Cascarano, B. Carrozzini, M. Camalli andR. Caliandro, J. Appl. Crystallogr., 2003, 36, 1103.

42 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,M. A. Rob, J. R. Cheeseman, J. A. Montgomery Jr.,T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam,S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi,G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji,M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai,M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross,V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts,R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi,C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma,G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski,S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas,D. K. Malick, A. D. Rabuck, K. Raghavachari,J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford,J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko,P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith,M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,M. W. Wong, C. Gonzalez and J. A. Pople, Gaussian 03,Gaussian, Inc., Wallingford, CT, 2003.

Paper Dalton Transactions

13368 | Dalton Trans., 2013, 42, 13357–13368 This journal is © The Royal Society of Chemistry 2013

Publ

ishe

d on

25

July

201

3. D

ownl

oade

d by

Uni

vers

ity o

f B

elgr

ade

on 0

9/05

/201

4 11

:55:

49.

View Article Online