Computational and Theoretical Chemistry 1095 (2016) 36–43

Contents lists available at ScienceDirect

Computational and Theoretical Chemistry

journal homepage: www.elsevier .com/locate /comptc

A theoretical study of intramolecular H–bonding and metal–ligandinteractions in some complexes with bicyclic guanidine ligands

http://dx.doi.org/10.1016/j.comptc.2016.09.0142210-271X/� 2016 Elsevier B.V. All rights reserved.

⇑ Corresponding authors.E-mail addresses: m.chahkandi@hsu.ac.ir, chahkandimohammad@gmail.com

(M. Chahkandi), mirzaeesh@um.ac.ir (M. Mirzaei).

Mohammad Chahkandi a,⇑, Behnaz Madani Khoshbakht a, Masoud Mirzaei b,⇑aDepartment of Chemistry, Hakim Sabzevari University, Sabzevar 96179-76487, Islamic Republic of IranbDepartment of Chemistry, Ferdowsi University of Mashhad, Mashhad 917751436, Islamic Republic of Iran

a r t i c l e i n f o a b s t r a c t

Article history:Received 2 June 2016Received in revised form 17 August 2016Accepted 8 September 2016Available online 9 September 2016

Keywords:Intramolecular hydrogen bondhppHNBOCharge transferTD–DFT

The reported complexes formulated as [MnCl2(hppH)2] (1), [FeCl2(hppH)2] (2), and [NiCl2(hppH)2] (3) anda new theoretically designed example, [CuCl2(hppH)2] (4), have been used for calculations at the B3LYP/LANL2DZ/6-311G (d, p) level of density functional theory (DFT). The intramolecular hydrogen bonds (HB)NAH� � �Cl could followed through their physicochemical properties such as, vibrational frequency, elec-tronic transmission in UV–Visible spectroscopy, metal–ligand donor–acceptor interactions in NBO anal-ysis, total energy and frontier molecular orbital energy. These properties influenced by the relationship ofstructure and metal–ligand electron density exchange will be discussed. The computations revealed thatthe stronger NAH� � �Cl HB exists in complexes with longer MACl and MANimine bonds and shorter H� � �Clbond, and vice versa that confirms the shortest and longest HBs in 4 and 3, respectively. These resultsagree with the second–order perturbation energies obtained by NBO analysis within charge transfer fromthe proton–acceptor chlorine to the p⁄ orbital of the N atom. The calculated electronic absorption spectraby TD–DFT calculations show the larger Cl� to M2+ donation in 2 and 3 in comparison with 1 and 4 thatconfirms the stronger HB in 1 and 4 compared to 2 and 3. In order to find the basis set effect on the struc-ture, vibrational frequencies, and electronic transitions, we use another basis set def2-TZVP includespolarization function for Mn element in complex 1. The obtained test results showed unremarkable dif-ferences between two basis sets LANL2DZ and def2-TZVP (cf. Figs. S2, S3 and Tables S1 and S2 inSupplementary Materials).

� 2016 Elsevier B.V. All rights reserved.

1. Introduction

The hydrogen bond (HB) is a unique phenomenon in structuralchemistry and biology that plays crucial roles in molecular associ-ation, function, and dynamics of the great numbers of chemical andbiological compounds. The studies of systems including theH–bond have attracted growing interest during the last decades[1–5]. However, the stabilization effect of such interactions in acomplex containing fragments of ligand(s) and metal has receivedsubstantially less attention. Many functional groups containingamine and imine such as oxime and guanidine ligands have theability to form intra and intermolecular HBs playing importantroles in the molecular design [6–9]. They could form NAH� � �Xand OAH� � �X HBs, where X is an atom or group that contains elec-trons available for interaction with the acidic proton (see Fig. 1).

Guanidine derivatives are important ligands in coordinationchemistry [10], especially because of their large basicity[pKA = 13.6 for HN@C(NH2)2] [11,12] and their role in biologicalsystems [10,13,14]. Herein, we theoretically investigate structuresof complexed the bicyclic guanidine (general formula: R2NC{NR0}{NHR0}) 1,3,4,6,7,8-hexahydro-2H-pyrimido[1,2-a]pyrimidine(hppH), bearing the annular framework with the Eanti configurationof the amidine moiety which makes the NH functionality availablefor additional HB interaction with the X fragment to which it iscoordinated. In these complexes the neutral guanidine ligandbonds through the imine nitrogen atom to a tetrahedral metal cen-tre with additional interaction between the secondary amino NHand the metal bound halide [15]. This member of the bicyclicguanidine family is an effective enantioselective catalyst for theasymmetric synthesis of R–amino nitriles and R–amino acids[16,17]. All of the complexes described here contain an intramolec-ular HB involving amine functional groups (see Fig. 1). Differenttransition metal complexes containing guanidine ligands like hppHhave been synthesized [15,18]. It has been shown that this

Fig. 1. B3LYP/LANL2DZ/6-311G (d, p) optimized structures of 1–4.

M. Chahkandi et al. / Computational and Theoretical Chemistry 1095 (2016) 36–43 37

extremely versatile ligand will bond to a number of main–groupand transition metals and supporting a variety of geometries alongwith NAH� � �X HBs [9] i.e. from trigonal planar copper(I) [19a] andsilver(I) [19b], tetrahedral iron and cobalt [9,19c] to square–planarpalladium complexes [19d].

It has been proposed that the metal could reinforce the hydro-gen bonds while the metal–ligand bond strength showed areversed relation with the size of the metal ion [20]. Theoreticalcalculations as a good complement to experimental results havebeen used for better understanding of the structure and propertiesof synthesized complexes and for designing desired ones [21,22].The density functional theory (DFT) method is a reliable computa-tional approach to determine the most stable of optimized struc-tures of the transition metal complexes, their stabilizationenergy, their frequencies of inter-fragment vibrations, and thepotential and free energy surfaces [23]. It provides impressiveresults in studies of large organic and metal compounds and clus-ters at a relatively low computation value comparing to otherpost–Hartree–Fock methods.

In the present work, using DFT calculations we study geometryand electronic structure of intramolecular H–bonded complexeswith the general formula Cl2(hppH)2 (M = Mn, Fe, Ni, and Cu).The first three are published complexes [15] ([MnCl2(hppH)2] (1),[FeCl2(hppH)2] (2), and [NiCl2(hppH)2] (3)) while the last is a the-oretically proposed compound [CuCl2(hppH)2] (4) which will becompared with the others. To explore the relation between theH–bonding and donor–acceptor interactions in the complexes aquantitative analysis of the metal–ligand interactions was per-formed by a Natural Bond Orbital (NBO) [24] analysis. Moreover,the quantitatively and qualitatively characteristics of the metal–li-gand interactions and the influenced intramolecular HBs of these

complexes with metal centre size using the vibrational and TD–DFT calculations have been performed.

2. Computational details

As starting point to geometry optimization of 1–4 CIFs [15], thenon–hydrogen atoms were fixed at the coordinates taken from thecrystal structures and the positions of the hydrogen atoms wereoptimized. Full geometry optimization, vibrational frequencies,UV–Vis, and NBO calculations were performed through the DFTmethod at the hybrid functional B3LYP level [25–27] with theLANL2DZ (def2-TZVP for Mn in complex 1, see SupplementaryMaterials) basis set for the metal atoms and the 6-311G (d, p) basisset for all other atoms. Employing effective core potential (ECP)functions [28–31] such as LANL2DZ for transition metals, whileusing all electron basis sets for all other non–transition metalatoms has become quite popular in computations on systems con-taining transition metal ions because of the time saved [32]. Thecalculated vibrational frequencies indicated that the structureswere stable (no imaginary frequencies). The computed geometricalparameters have been compared with available experimental data(see Table 1). UV–Vis spectra were calculated for the optimizedgeometry with a time–dependent DFT (TD–DFT) method. Singlet,doublet, and triplet excited states were considered; however asexpected, all excitations to triplet states were prohibited. Netcharges of all compounds are zero. The dissociation energies werecorrected for BSSE using the counterpoise method [33] and forzero–point vibrational energy obtained at the B3LYP/LANL2DZ/6-311G (d, p) level. All calculations were carried out with theGaussian 09 program package [34].

Table 1selected experimental and calculated bond lengths (Å) and angles (�) for B3LYP/LANL2DZ/6-311G (d, p) optimized structure of 1–4.

1 2 3 4

Exp. Calc. Exp. Calc. Exp. Calc. Calc.

MAN 2.107 2.079 2.039 1.978 1.983, 1.977 1.929 2.005MAX 2.368 2.379 2.319 2.337 2.291, 2.275 2.296 2.408C@Nimine 1.354 1.322 1.322 1.319 1.313, 1.316 1.310 1.321CANH 1.348 1.354 1.354 1.358 1.354, 1.359 1.367 1.351CANR2 1.351 1.322 1.352 1.372 1.355, 1.349 1.369 1.370Namine� � �Cl 3.276 3.189, 3.196 3.301 3.172 3.242, 3.213 3.204 3.146H� � �Cl 2.458 2.177, 2.184 2.524 2.178 2.448, 2.465 2.298 2.150NamineAHACl 172.67 170.50, 170.80 169.08 164.41 168.80 147.58 164.35NAMAN0 108.69 108.51 107.52 97.48 103.81 175.71 145.43ClAMACl0 110.67 128.74 111.63 147.67 116.08 175.25 140.33NAMACl 107.97 102.46 107.94 95.34 108.03 88.75 91.10NAMACl0 110.77 107.94 110.87 105.91 111.03 91.43 100.50N0AMACl 110.77 108.06 110.87 105.91 111.12 91.43 100.50N0AMACl0 107.97 100.16 107.94 95.34 106.09 88.75 91.10

38 M. Chahkandi et al. / Computational and Theoretical Chemistry 1095 (2016) 36–43

3. Results and discussion

3.1. Molecular geometries of 1–4: Theoretical point of view

The CIFs of 1, 2, and 3 with CCDC 276690, 231326, and 276689show that they crystallize in monoclinic and orthorhombic sys-tems with P21/n, Fdd2, and Fdd2 space groups, respectively.Selected geometrical parameters experimentally (for 1–3)obtained from the published structures [15] and theoretically opti-mized (for 1–4) are collected in Table 1. The asymmetric units arethe neutral monomeric complexes. The metal centre (M(II)) isfour–coordinated through bonding to the imine nitrogen atomsof two monodentate hppH and two chlorides in distorted tetrahe-dral (for 1, 2, and 4) and square planar (for 3) geometries (seeFig. 1). It should be noted that for the optimized structures no sym-metry constraints were applied. Since the structure in the solidstate involves a huge network of inter–molecular interactions,the reported structural parameters for the synthesized complexessometimes differ from the calculated ones which are based on iso-lated units in the gas phase. The largest differences of experimentaland calculated bond angles of ClAMACl and NAMAN are near 60and 72�, respectively, found in 3 as a result of geometry changefrom tetrahedral to distorted square planar. It seems that hppHand chloride ligands move far away thereby changing theClANiACl and NANiAN bond angles from 116.08 and 103.81�(Exp.) to 175.25 and 175.71� (Calc.) and C8AN4ANiAN1 torsionangle from 85.62 (Exp.) to �161.88 (Calc.) and reinforcing a doubleset of NAH� � �Cl hydrogen bonds from 2.465 (Exp.) to 2.298 Å in theoptimized structure of 3 (see Fig. 1 and Table 1). The lesser distor-tion of this torsion angle (1, 26.13 to 121.42; 2, 146.64 to 143.59)could be followed in the isostructural complexes 1 and 2. However,the largest difference of Exp. and Calc. NAH� � �Cl in 1 and 2 are 0.28and 0.35 Å, respectively (see Table 1). Based on geometrical param-eters (H� � �A and DAA bond lengths and the <DHA angle) and theelectron density concept on the D and A groups, HBs can be classi-

Table 2Calculated vibrational frequencies with B3LYP/LANL2DZ/6-311G (d, p) for optimized struc

Assignments 1 2

MACl sym-str. 256.7 vw 242.2 vwMANimine sym-str. 245.9 vw 276.6 vwCH2 bend. 1476.6 m 1479.3 mC@NR2 str. 1581.7–1582.9 m 1580.9–1581.2 mC@Nimine 1632.2 vs 1634.4 vsNAH str. (from NAH� � �Cl HB) 3370.6–3371.2 vs 3387–3388.8 vs

The abbreviations are, vs: very strong, s: strong, m: medium, w: weak, vw: very weak.

fied into three categories, namely strong, moderate and weak [1]and on this basis the HBs in complexes 1–4 belong to the moderateclass (Table 1). The optimized structures of MCl2(hppH)2 showstrong metal–ligand bonds as indicated by very short MANimine

distances (1, 2.08; 2, 1.98; 3, 1.93; and 4, 2.00 Å, respectively),being below the sum of covalent radii of the bonded atoms (1,2.20; 2, 2.03; 3, 1.95; and 4, 2.03 Å, respectively). The covalent radiiof these metals show the trend of Ni < Cu < Fe < Mn. The abovementioned moderate HBs could be confirmed by significantly elon-gation of NAH bond lengths in the optimized structures (in paren-thesis) as compared to the experimental ones [1, 0.823 (1.021); 2,0.786 (1.017); 3, 0.803 (1.021); 4 (1.021)]. According to the crystalfield stabilization energy (CFSE) and Jahn-Teller effect, theMAX bond length (X = Cl, N) in experimental and calculateddistorted tetrahedral complexes 1–4 could be ordered asCuAX > MnAX > FeAX > NiAX. The optimized structures confirmthis trend (see Fig. 1). Regarding the crystallographic and opti-mized structures of these complexes, elongation of MACl corre-lates with a shortening of H� � �Cl and the longer MANimine bondis accompanied by a shorter HB, and vice versa that confirms theshortest and longest HBs in 4 and 3, respectively (see Table 1and Fig. 1).

The vibrational spectra show different frequencies assigned tothe important functional groups of 1–4. The important featuresof their IR spectra (omitted) are the symmetric stretchingof ʋ(C@Nimine) (Exp. 1602–1626, Calc. 1629–1637 cm�1) and ʋ(NamineAH) (Exp. 3351–3366, Calc. 3356–3457 cm�1) [35–37].Other essential characteristic frequencies of 1–4 are compared inTable 2 from which it can be concluded that the experimentallyrecorded vibrational frequencies are in good agreement with calcu-lated ones at B3LYP/LANL2DZ/6-311G (d, p) method. The test cal-culations about complex 1 showed this consistency with theoptimized structure at B3LYP/def2-TZVP/6-311G (d, p) experimen-tal one (see Supplementary Materials). The low energy shift of theʋ(C@Nimine) stretching frequency (corresponding shift in free

tures of 1–4 and free guanidine ligand.

3 4 Free guanidine

239.8 vw 204.7 vw –331.8 vw 276.6 vw –1484.2 w 1469.3 m 1502.5 vw1573.9–1576.0 m 1585.8–1586.0 m 1544.4 m1636.9 vs 1629.2 vs 1694.6 vs (Calc.), 1641 s (Exp.)3454.9–3456.6 vs 3356.3–3356.9 vs 3610.0 w (Calc.), 3419 s (Exp.)

M. Chahkandi et al. / Computational and Theoretical Chemistry 1095 (2016) 36–43 39

hppH = 1641(Exp), 1694.6 (Calc) cm�1) is consistent with bondingto a metal centre via the imine rather than the amine nitrogen [35].Moreover, the strong shift of ʋ(NamineAH) signal in synthesized[35–37] and optimized structures of 1–4 comparison with freehppH (Exp. 3419 [38], Calc. 3610 cm�1) is indicative of NAH� � �ClHB formation. Based on the computed IR spectra the weakest andstrongest NAH bond result in shortest and longest H� � �Cl for 4and 3, respectively which confirms the above mentioned subject(cf. Tables 1 and 2). Generally, HBs can be estimated by the bondlengths e.g. NAH� � �Cl and respective angle (<NHCl), howeverbecause of complexity of the issue, just according to the structuralparameters, its strength could not truly be compared.

3.2. Metal-ligand and HBs interactions

3.2.1. NBO analysisA scientific method such as NBO analysis [24] could be per-

formed for understanding the covalent and non–covalent interac-tions such as HBs [20,39–42]. During H–bonding, based on itsdefinition, electrons may transfer between the donor (D) andacceptor (A) groups (DAH� � �A). This charge transfer (CT) and gen-erally the attractive metal–ligand and HBs interactions can becharacterized by the second–order perturbation energies, E(2), onthe basis of the natural atomic charges and the donor–acceptorinteractions [24].

Eð2Þ ¼ �nrrjFjh r�ier� � er

¼ �nrF2ij

De

where Fij is the Fock matrix element between the NBO i(r) and j(r⁄), er and er⁄ are the energies of r and r⁄ NBOs, and nr is thepopulation (a lone pair in the bond). On the other hand, a decreasingenergy gap between donor (r) to acceptor (r⁄) orbitals enhancesthe CT from r to r⁄ increasing E(2) and making the interactionstronger. However, the summation of all influencing parametersshould be considered.

As mentioned before, although the strength of NAH� � �Cl couldbe followed through geometrical parameters of itself and its sur-roundings (Table 1), HB formation energy should be estimatedfrom the metal–ligand donor–acceptor interactions in NBO analy-sis given in Table 3: these are the longer NAH bond, the shorterNamine� � �Cl and H� � �Cl, and longer MACl and MANimine bonds inthe stronger H–bonded [CuCl2(hppH)2] (4) especially with respectto the Ni complex (3). The considerable lengthening of the NAHbonds is caused by CT from the proton–acceptor chlorine to thep⁄ orbital of the NAH bond. This charge transfer can be character-ized by the second–order perturbation energies obtained from NBO

Table 3Dissociation energy, computed natural charges, and charge transfer interactions in the stu

1 2

D0 81115.1 81267qM 0.793 0.560qCl �0.597 �0.57qNimine �0.678, �0.688 �0.58qNamine �0.634 �0.64qH 0.434 0.434qL 0.205 0.289E(2) LP Cl(p)? p⁄(NAH) 89.58 90.96E(2) p(C@Nimine)? LP M(p) 27.61 51.80E(2) p(CH2ANimine)? LP M(p) 32.22 49.79E(2) LP Nimine(s)? LP M(p) 28.76 46.80E(2) LP Cl(s,p)? LP M(p,d) 270.75 301.8E(2) LP M(d)? p⁄(C@Nimine) 14.10 27.87E(2) LP M(d)? LP⁄(Nimine)(p) 30.42 17.15E(2) LP M(d)? p⁄(CH2ANimine) N.F. 10.54

a From NBO analysis. Second order perturbation energies (E(2) donor? acceptor) in k

[24], giving values of 113.39, 90.96, 89.58, and 51.30 kJ/mol for HBsformation in 4, 2, 1, and 3 complexes, respectively. The calculatedformation energies, E(2), and geometrical parameters of HB of thecrystallographic and optimized structures have a good consistency(see Fig. 1 and Tables 1 and 3). The very stable title complexesshow large dissociation energies (around 81,000 kJ/mol, seeTable 3) of [MCl2(hppH)2]?M2+ + 2(Cl� + hppH). These ratherhigh values can be attributed to the very low stability of the result-ing ions in vacuum. The dissociation energies increase fromM = Mn to Cu due to increasing the strength of MANimine, MACl,and respective HB bonds (cf. Tables 1 and 3).

The formally M2+ metal ions, Cl�, and Nimine in these complexeshave natural charges of around +0.8 to 0.45, �0.53 to �0.61, and�0.58 to �0.69 e, respectively. The smaller positive charge of Niand Fe and smaller negative charge of Cl�, and Nimine in 2 and 3complexes indicate stronger donation from the chloride and hppHligands in these two complexes as compared with the others understudy. These results confirm the shorter bond lengths within Feand Ni in 2 and 3 and vice versa in 1 and 4. E(2) calculated energies(cf. Table 3) point to the significant CT from the ligand to the met-als indicating also that the opposite process, the metal–to–ligandback–donation, must be much less. Comparison to the free hppH,this ligand in complexes 1 and 4 coordinates by a more negativelycharged Nimine (around�0.7 e) while in complexes 2 and 3 by a lessnegatively charged Nimine (around �0.6 e). Obviously, in the lattertwo complexes the coordination of L must be facilitated by stron-ger NimineAM donor–acceptor interactions which are reflected bythe total ligand charges (qL) in Table 3 where the hppH ligandshave larger positive charges (around +0.3 e) in 2 and 3 comparedwith the ligands in 1 and 4 (around +0.2 e). However, the electro-static attraction between the M2+, Nimine, and Cl� playing probablya secondary role beside the DAA interaction.

The following DAA interactions of a Lewis model frameworkcan operate between M2+ and the hppH and chloride ligands inthe title complexes:

� r–donation from the lone pairs (LP) of Nimine and Cl� to unfilledp and or d orbitals of M2+.

� p–donation from the p orbitals of CANimine double and singlebonds to unfilled p orbitals of M2+.

� p–back–donation from occupied d orbitals of M2+ to anti–bonding p⁄ orbitals of CANimine bonds and unfilled p⁄ of Nimine.

Regarding the second–order perturbation energies obtainedfrom NBO in Table 3 the ligand–to–metal r–donation was foundto be the major component accompanied by minor p contributions.The quantitative data of the CT interactions given in Table 3 justify

dies complexes 1-4a.

3 4 Free hppH

.9 81534.1 81649.3 –0.454 0.771 –

8 �0.530 �0.612 –8 �0.584 �0.679 �0.6211 �0.640 �0.631 �0.651

0.432 0.437 0.3910.305 0.228 0.0051.30 113.39 –N.F. 68.03 –N.F. 41.67 –55.62 17.35 –

4 720.40 134.39 –24.69 23.10 –12.26 37.74 –10.04 8.37 –

J/mol. N.F.: Not Found.

40 M. Chahkandi et al. / Computational and Theoretical Chemistry 1095 (2016) 36–43

that the main donor–acceptor interaction is the donation from theLP of the Cl�. The donation is somewhat larger in complexes 2 and3 and results in the somewhat smaller negative charge of chlorideligands in these complexes in comparison with complexes 1 and 4.Obviously, in the latter two complexes the larger donation fromthe LPs of chlorine to the p⁄ orbital of the NAH bond form strongerHBs. The other charge transfer interactions are weaker by an orderof magnitude. The data prove the minor importance of the metal–to–ligand back–donation in the title [MCl2(hppH)2] complexes.However, the p⁄–back–donation from metal to unfilled p⁄ of Nimine

is larger for 4 and 1 than those ones in 2 and 3 that increases pos-itive charge of the metal centre, weaken the respective MANimine

bond length, and in conclusion it could reinforce the related H� � �Clhydrogen bonds in 1 and 4 complexes. The other metal–ligand

Wavelength, nm 300 290 280 270 260 250 240 230 220 210 200

Osc

illat

or S

tren

gth

(a.u

.)

0.07 0.065

0.06 0.055

0.05 0.045

0.04 0.035

0.03 0.025

0.02 0.015

0.01 0.005

0

211.9

254.2

272.3

Wavelength, nm 320 310 300 290 280 270 260 250 240 230 220 210 200

Osc

illat

or S

tren

gth

(a.u

.)

0.14 0.13 0.12 0.11

0.1 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01

0

246.8

226.6

223.8

2

Osc

illat

or S

tren

gth

(a.u

.)O

scill

ator

Str

engt

h (a

.u.)

1

Fig. 2. Calculated electro

Table 4Calculated electronic data for 1–4.

Complex Wave length (nm) DEad (eV)

1 211.87 5.84252.56 4.90254.18 4.87272.31 4.55

2 223.77 5.53226.62 5.46246.80 5.02

3 200.10 6.19222.60 5.56239.76 5.16

4 206.16 6.00311.80 3.97372.51 3.32

donor–acceptor interactions are not very different for 1–4complexes.

3.2.2. Electronic spectrumTo investigate the electronic properties of 1–4, molecular orbital

calculations with natural population analyses (NPA) based on theoptimized ground state geometry have been performed. The opti-mized structures of them have 101, 101, 102, and 103 occupiedmolecular orbitals (MOs) within doublet, singlet, singlet, and dou-blet electron multiplicity, respectively. The complementaryinformation could be found in Supplementarymaterial. The frontiermolecular orbitals (FMOs) can be useful for attributing the calcu-lated UV–Vis bands to the respective electronic transitions and inconclusion comparing the involved metal–ligand interactions. The

Wavelength, nm 400 380 360 340 320 300 280 260 240 220 200 180

0.45

0.4

0.35

0.3

0.25

0.2

0.15

0.1

0.05

0

200.1

222.6

239.8

Wavelength, nm 550 500 450 400 350 300 250 200

0.3 0.28 0.26 0.24 0.22 0.2

0.18 0.16 0.14 0.12 0.1

0.08 0.06 0.04 0.02

0

206.2

311.8

372.5

3

4

nic spectra for 1–4.

Oscillator strength Electronic transition

0.034 99a? 107a0.033 101a? 107a0.047 101a? 106a0.024 98a? 109a

0.130 101a? 109a0.107 100a? 108a0.045 93a? 97a

0.078 94a? 106b0.403 92a? 103a0.263 102a (b)? 104a (b)

0.183 93a? 103a0.014 103a? 107a0.073 99b? 103b

M. Chahkandi et al. / Computational and Theoretical Chemistry 1095 (2016) 36–43 41

main calculated wavelengths of 1–4 have been collected in Table 4(see Tables S1 and S2 in Supplementary materials for complex 1calculated at B3LYP/def2-TZVP/6-311G (d, p)). The experimentalUV–Vis spectra for 1–4 complexes are not found in the literature.As mentioned, vide supra, one of the main donor–acceptorinteraction is the donation from the LP of the Cl� to M2+.Therefore, the main electronic band in the pertinent spectrum;

λ = 211.

CT

99 (HOMO-2)-1

λ = 223

CT

101 (HOMO)-2

λ = 239

CT

102 (HOMO)-3

λ = 206.2

CT

93 (HOMO)-4

Fig. 3. Frontier molecular diagrams for 1 to 4 involving in CT

comparable between complexes 1–4; related to Cl� ?M ligand–to–metal charge transfer (LMCT) appears at 211.9, 223.8, 239.8,and 206.2 nm for 1–4, respectively (cf. Fig. 2). These bands canbe described as n? n from LP of Cl� (s, p) to unfilled p and ord orbitals of M2+ (see Fig. 3). This trend confirms the largerdonation in complexes 2 and 3 resulting in smaller negativecharge on the chloride ligands in these complexes as compared

9

107 (LUMO+6)-1

.8

109 (LUMO+8)-2

.8

104 (LUMO+1)-3

103 (LUMO+3)-4

obtained according to the B3LYP/6-311G (d, p)/LANL2DZ.

42 M. Chahkandi et al. / Computational and Theoretical Chemistry 1095 (2016) 36–43

with complexes 1 and 4. The bands appearing at 272.3, 246.8,222.6, and 311.8 nm for 1–4, respectively, could assigned top–back–donation from M2+ (p and/or d) to hppH (s, p, d) or Cl�

(s, p) metal–to–ligand charge transfer (MLCT). This trend andrelated second–order perturbation energies which are larger forcomplexes 1 and 4 in comparison with complexes 2 and 3 showgood consistencies (Table 3). Moreover, for 4 there is anotherband at 372.5 nm attributed to the Cu? Cu (d? d) crystal fieldtransition (Fig. 2).

4. Conclusion

The formation of intramolecular HBs has very pronouncedeffects on the molecular structure and properties. We have pre-sented here the first systematic DFT theoretical study of bis(bicyc-lic guanidine) complexes of Mn, Fe, Ni, and Cu, [MCl2(hppH)2],focusing on the hydrogen bonding and metal–ligand interactions.The optimized structures have good agreement with the experi-mental ones. However, the largest differences of experimentaland calculated symmetry geometry occurred in 3 due to changingof geometry from tetrahedral to distorted square planar.

According to the crystal field stabilization energy (CFSE), Jahn-Teller effect, and covalent radii of the metal centres, the resultingbond length of MAX (X = Cl, N) in 1–4 could be ordered asCuAX > MnAX > FeAX > NiAX. Therefore, elongation of MACl cou-pled with shortness of H� � �Cl and related HBs and vice versa con-firms that the shortest and longest HBs are in 4 and 3,respectively. Consequently, the differences in the radii of the met-als determine the differences in the lengths of the HBs. The natureof the metal-ligand and NAH� � �Cl HB interactions were evaluatedby NBO analysis. HB could be characterized by the second–orderperturbation energies because of charge transfer from the pro-ton–acceptor chlorine to the p⁄ orbital of the NAH bond resultingin the considerable lengthening of the NAH bonds. HB formationenergy, E(2), shows the strength order of formed HB as4 > 2 > 1 > 3. The calculated formation energies and geometricalparameters of HB of the crystallographic and optimized structureshave a good consistency. However, the computed E(2) revealed thepredominance of the Cl?M r–donation and the secondaryimportance of the other (p–donation from C@Nimine to M, p–backdonation from M to the anti–bonding p⁄ orbitals of the C@Nimine

and Nimine) CT interactions. Based on calculated UV–Vis spectraof 1–4 complexes the main donor–acceptor interaction as thedonation from the LPs of the Cl� to M2+ indicates the larger dona-tion in complexes 2 and 3 bearing Fe and Ni metal centres in com-parison with complexes 1 and 4. Because of more electronegativityand effective nuclear charge of Ni than Fe and Mn, its complex (3)should have intensity and lower energy of ligand to metal chargetransfers (LMCTs) with higher wave lengths. However, the oppositetrend within p–back–donation from M2+ to ligand (MLCT) shouldbe concluded which the larger donation in complexes 1 and 4 incomparison with complexes 2 and 3 found. It should be noted thatthe calculations of optimization, vibrational frequencies, andcharge transfers in UV–vis spectrum of complex 1 at B3LYP/def2-TZVP/6-311G (d, p) method did not shown permanent differenceswith the same calculations at B3LYP/LANL2DZ/6-311G (d, p) (seeSupplementary materials).

Acknowledgements

MCH gratefully acknowledges the financial support by theHakim Sabzevari University, Sabzevar, Iran. Authors thank CinaForoutan-Nejad, CEITEC-Central European Institute of Technology,for performing computations for us. Computational resources wereprovided by the CESNET LM2015042 and the CERIT Scientific Cloud

LM2015085, provided under the programme ‘‘Projects of LargeResearch, Development, and Innovations Infrastructures”.

Appendix A. Supplementary material

Supplementary data contains frontier molecular diagrams for 1to 4 involving in CT obtained according to the B3LYP/6-311G(d, p)/LANL2DZ level of calculations. Also all optimized structural param-eters, vibrational frequencies, and charge transfer wavelengths inUV-Vis spectrum of 1 calculated at B3LYP/def2-TZVP/6-311G(d,p) level of theory have been gathered there. Supplementary dataassociated with this article can be found, in the online version, athttp://dx.doi.org/10.1016/j.comptc.2016.09.014.

References

[1] G.A. Jeffrey, An Introduction to Hydrogen Bonding, Oxford University Press,New York, 1997.

[2] S. Scheiner, Molecular Interactions: From Van Der Waals to Strongly BoundComplexes, Wiley, UK, 1997.

[3] T. Steiner, The hydrogen bond in the solid state, Angew. Chem. Int. Ed. 41(2002) 48.

[4] S.J. Grabowski, Chem. Rev. 111 (2011) 2597.[5] S.J. Grabowski, Hydrogen Bonding – New Insights, Springer, The Netherlands,

2006.[6] H. Raissi, M. Yoosefian, F. Mollania, Comput. Theor. Chem. 996 (2012) 68.[7] S. Köcher, M. Lutz, A.L. Spek, B. Walfort, T. Rüffer, G.P.M. van Klink, G. van

Koten, H. Lang, J. Organomet. Chem. 693 (13) (2008) 2244.[8] G. Tauzher, R. Dreos, A. Felluga, G. Nardin, L. Randaccio, M. Stener, Inorg. Chim.

Acta 355 (20) (2003) 361.[9] S.H. Oakley, D.B. Soria, M.P. Coles, P.B. Hitchcock, Polyhedron 25 (2006) 1247.[10] (a) X. Zhang, C. Wang, C. Qian, F. Han, F. Xu, Q. Shen, Tetrahedron 67 (2011)

8790;(b) U. Wild, P. Roquette, E. Kaifer, J. Mautz, O. Hübner, H. Wadepohl, H.-J.Himmel, Eur. J. Inorg. Chem. 8 (2008) 1248.

[11] S.J. Angyal, W.K. Warburton, J. Chem. Soc. (1951) 2492.[12] A. Gobbi, G. Frenking, J. Am, Chem. Soc. 115 (1993) 2362.[13] A.A. Legin, M.A. Jakupec, N.A. Bokach, M.R. Tyan, V.Yu. Kukushkin, B.K. Keppler,

J. Inorg. Biochem. 133 (2014) 33.[14] M. Said, J. Ahmad, W. Rehman, A. Badshah, H. Khan, M. Khan, F. Rahim, D.M.

Spasyuk, Inorg. Chim. Acta 434 (2015) 7.[15] S.H. Oakley, D.B. Soria, M.P. Coles, P.B. Hitchcock, Polyhedron 25 (2006) 1247.[16] E.J. Corey, Michael J. Grogan, Org. Lett. 1 (1999) 157.[17] D. Enders, John P. Shilvock, Chem. Soc. Rev. 29 (2000) 359.[18] O. Ciobanu, A. Fuchs, M. Reinmuth, A. Lebkücher, E. Kaifer, H. Wadepohl, H.J.

Himmel, Z. Anorg, Allg. Chem. 636 (2010) 543.[19] (a) S.H. Oakley, M.P. Coles, P.B. Hitchcock, Inorg. Chem. 42 (2003) 3154;

(b) S.H. Oakley, D.B. Soria, M.P. Coles, P.B. Hitchcock, Dalton Trans. (2004) 537;(c) F.A. Cotton, C.A. Murillo, D.J. Timmons, Polyhedron 18 (1999) 423;(d) S.H. Oakley, M.P. Coles, P.B. Hitchcock, Inorg. Chem. 43 (2004) 7564.

[20] B. Krámos, A. Kovács, J. Mol. Struct.: THEOCHEM. 867 (2008) 1..[21] J. Chrappov, P. Schwendt, M. Sivk, M. Repisky, V.G. Malkin, J. Marek, Dalton

Trans. (2009) 465.[22] P.E.M. Siegbahn, M.R.A. Blomberg, Chem. Rev. 100 (2000) 421.[23] M.H.N. Assadi, D.A.H. Hanaor, J. Appl. Phys. 113 (2013) 233913.[24] A.E. Reed, L.A. Curtiss, F. Weinhold, Chem. Rev. 88 (1988) 899.[25] C. Lee, R.G. Parr, W. Yang, Phys. Re. 37 (1988) B785.[26] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.[27] A.D. Becke, Phys. Rev. A 38 (1988) 3098.[28] M. Bangesh, W. Plass, J. Mol. Strut. (THEOCHEM) 725 (2005) 163.[29] J.Y. Kravitz, V.L. Pecoraro, H.A. Carlson, J. Chem. Theory Comput. 1 (2005) 1265.[30] M. B}uhl, R. Schurhammer, P. Imhof, J. Am. Chem. Soc. 126 (2004) 3310.[31] M. Rochdi, J.-Y. Saillard, J.-F. Halet, S. Ghosh, H. Rabaâ, Polyhedron 43 (2012)

31.[32] (a) M. Chahkandi, J. Mol. Struct. 1111 (2016) 193;

(b) M. Chahkandi, Polyhedron 109 (2016) 92;(c) M. Mirzaei, H. Eshtiagh-Hosseini, M. Mohammadi Abadeh, M. Chahkandi,A. Frontera, A. Hassanpoor, CrystEngComm 15 (2013) 1404;(d) D. Balcells, F. Maseras, G. Ujaque, J. Am. Chem. Soc. 127 (2005) 3624;(e) H. Eshtiagh-Hosseini, M. Mirzaei, M. Biabani, V. Lippolis, M. Chahkandi, C.Bazzicalupi, CrystEngComm 15 (2013) 6752.

[33] S.F. Boys, F. Bernardi, Mol. Phys. 19 (1970) 553.[34] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,

G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H. Nakatsuji, M. Caricato,X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng, J.L. Sonnenberg, M.Hada, M. Ehara, Toyota, K. Fukuda, R. Hasegawa, J. Ishida, M. Nakajima, T.Honda, Y. Kitao, O.H. Nakai, T. Vreven, J.A.Montgomery, J.E. Peralta, F. Ogliaro,M. Bearpark, J.J. Heyd, E. Brothers, K.N. Kudin, V.N. Staroverov, R. Kobayashi, J.Normand, K. Raghavachari, A. Rendell, J.C. Burant, S.S. Iyengar, J. Tomasi, M.Cossi, N. Rega, J.M. Millam, M. Klene, J.E. Knox, J.B. Cross, V. Bakken, C. Adamo,

M. Chahkandi et al. / Computational and Theoretical Chemistry 1095 (2016) 36–43 43

J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C.Pomelli, J. Ochterski, R.L. Martin, K. Morokuma, V.G. Zakrzewski, G.A. Voth, P.Salvador, J.J. Dannenberg, S. Dapprich, A.D. Daniels, O. Farkas, J.B. Foresman, J.V. Ortiz, J. Cioslowski, D.J. Fox, GAUSSIAN09, revision A.02, 2009. Gaussian Inc.:Wallingford, CT.

[35] P.J. Bailey, K.J. Grant, S. Pace, S. Parsons, L.J. Stewart, J. Chem. Soc., Dalton Trans.(1997) 4263.

[36] P.J. Bailey, K.J. Grant, L.A. Mitchell, A. Parkin, S. Parsons, J. Chem. Soc., DaltonTrans. (2000) 1887.

[37] M.P. Coles, P.B. Hitchcock, Polyhedron 20 (2001) 3027.

[38] W.J. Jones, Trans. Faraday Soc. 55 (1959) 524–531.[39] P.L.A. Popelier, Atoms in Molecules: An Introduction, PrenticeHall, London,

2000.[40] C.F. Matta, R.J. Boyd, The Quantum Theory of Atoms in Molecules: From Solid

State to DNA and Drug Design, Wiley-VCH, Weinheim, 2007.[41] H. Eshtiagh-Hosseini, M. Chahkandi, M.R. Housaindokht, M. Mirzaei,

Polyhedron 60 (2013) 93.[42] S.H. Kazemi, H. Eshtiagh-Hosseini, M. Mirzaei, Comput. Theor. Chem. 1004

(2013) 69.