Molecular structure of mercury(II) thiocyanate complexes based on DFT calculations and experimental UV-electron spectroscopy and Raman studies

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 115 (2013) 574–582

Contents lists available at SciVerse ScienceDirect

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa

Molecular structure of mercury(II) thiocyanate complexes based on DFT calculations and experimental UV-electron spectroscopy and Raman studies  naite˙ a, Erika Elijošiute˙ a,⇑, Olegas Eicher-Lorka b,⇑, Egidijus Griškonis c, Ieva Matulaitiene˙ b, Dalia Janku a Gintaras Denafas a

Department of Environmental Engineering, Kaunas University of Technology, Radvile˙nu˛ pl. 19, LT-50254 Kaunas, Lithuania Center for Physical Sciences and Technology, Savanoriu˛ pr. 231, LT-02300 Vilnius, Lithuania c Department of General Chemistry, Kaunas University of Technology, Radvile˙nu˛ pl. 19, LT-50254 Kaunas, Lithuania b

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

 Raman and UV spectral studies of

[Hg(SCN)n]2n complexes were performed.  Vibrational frequencies and electronic spectra were calculated using DFT theory.  The effect of solvation and H2O/D2O exchange on vibrational spectra was studied.  Hg(II) was always four-coordinated with thiocyanate and (or) water ligands.

a r t i c l e

i n f o

Article history: Received 8 March 2013 Received in revised form 14 June 2013 Accepted 19 June 2013 Available online 29 June 2013 Keywords: Mercury thiocyanate Raman UV DFT Isotopic exchange

a b s t r a c t In this work, we report a combined experimental and theoretical study on molecular structure, vibrational and electronic spectra of [Hg(SCN)n]2n complexes (where n = 2, 3, 4) in the aqueous solution. Molecular modeling of the mercury(II) complexes were done by the density functional theory (DFT) method using B3LYP functional with Stuttgart relativistic ECP 78MWB basis set for Hg and 6-311++G(d,p) basis set for all other atoms. The effect of different solvation models with explicit (ligand) and/or implicit water environment upon its geometry, vibrational frequencies and UV spectrum have been studied. The influence of H2O/D2O exchange on the experimental and calculated vibrational frequencies of studied complexes has been established. The double-peak character of the mHgAS vibrational mode of the all analyzed mercury complexes and mC„N mode of [Hg(SCN)3H2O] complex, respectively, were proposed here for the first time. The formation of four-coordinated Hg(II) complexes with thiocyanate and (or) water ligands was verified. Ó 2013 Elsevier B.V. All rights reserved.

Introduction Mercury is one of the most important elements in the environment [1]. Mercury(II) compounds offer attractive properties in ⇑ Corresponding authors. Tel.: +370 37 300180; fax: +370 37 300152 (E. Elijošiute˙), tel.: +370 52729372 (O. Eicher-Lorka). E-mail addresses: [email protected] (E. Elijošiute˙), [email protected] (O. Eicher-Lorka). 1386-1425/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.saa.2013.06.072

terms of their potential applications in paper industry, paints, cosmetics, preservatives, thermometers, manometers, fluorescent lamps, and mercury batteries [2]. Mercury is capable to form complexes with many inorganic ligands and reagents [3]. A search in the Cambridge Structural Database shows 123 mercury(II) thiocyanate complexes and the most of these compounds are monomeric structures [2].

E. Elijošiute˙ et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 115 (2013) 574–582

Thiocyanate-containing metal complexes are considered to be the most investigated systems because of their diverse structures (linkage isomerism), applications in magnetic materials and luminescence properties [4]. In agreement with the HSAB (hard soft acid base) theory the SCN ion coordinates to hard acids (i.e. Mn2+, Co2+, Ni2+, Mg2+, Fe3+, Na+) through nitrogen atom, and the uncoordinated sulfur atom is involved in hydrogen bonds and sometimes involved in SAS interactions. If the transition metal center is soft acid (i.e. Cd2+, Cu+, Hg2+), then SCN ligand binds to central ion through sulfur atom. Different stretching modes of the thiocyanato ligand can generate various types of supramolecular structures with particular properties [5–8]. The geometry and coordination mode of NCS ion in 3d metal complexes is strongly influenced by the electronic and steric effects around the central ion [5]. However, the final bonding mode depends also on the metal charge, other coordinated ligands and steric effects [9]. Thiocyanato bridges play an important role in the magnetic exchange pathways between paramagnetic centers too [10]. Simultaneous presence of two different metal centers can potentially give rise to interesting physico-chemical properties and lead to attractive novel topologies and intriguing frameworks [11,12]. According to Schmidt et al. [13] powerful and sensitive analytical tools, which provide a wealth of information on the physical/ structural and chemical composition of a sample without a priori knowledge, are vibrational spectroscopic techniques [13]. Dealing with the intrinsic view of molecule in the solution, it is well known that the spectroscopic and structure properties of molecules in solution are influenced by the solvation [14]. Since the first quantitative measurements were made, it has been clear that many types of solvent–solute interactions are possible in the solution. The onset of intramolecular interactions often leads to gross changes in band positions, intensity and shape, which can be explained in terms of changes in force constants, normal coordinates and electron redistribution due to the interaction [15,16]. However, because of extended vibrational coupling and the influence of ionization of ending groups, the interpretation of the vibrational spectrum is not straightforward, and in general cannot be accomplished without the high level of computational modeling [17]. In recent years, the density functional theory (DFT) has been favored as a tool for quantum chemistry because of its efficiency and accuracy with respect to the evaluation of a number of molecular properties and relative fastness in studying relatively large molecules [18–20]. During such quantum chemical calculations inclusion of both bulk effect and interaction with explicit water molecules improves the agreement with experimental results and is essential for the quantitative understanding of the solvent effects [21–24]. Kato et al. [25] found that the inclusion of explicit water molecules is a key factor for obtaining reliable computational results. Mercury(II) complexes are of interest since they are widely found in the aqueous solutions employed for the determination of chloride ions in clinical and industrial laboratories [26] or obtained during mercury determination procedures using titrimetric analysis [27]. The quantum chemical studies of the mercury–thiocyanato complexes have been started in the last two decades. Fukushima et al. [7] reported the formation energies, geometry optimizations of the simple tetrahedral [Hg(SCN)4]2 complex using only theoretical calculations. Šašic´ et al. [14,15] analyzed the influence of various solvents on HgAS, C„N and SAC vibrations of Hg(SCN)2, [Hg(SCN)3], [Hg(SCN)4]2 complexes using Raman analysis. Chillemi et al. [28] has tried to explain the structure and coordination of the pure hydrated Hg(II) complex in aqueous solution. Hofer et al. [29] also reported the results of the study of hydrated mercury(I)-dimer based on molecular dynamics. Additionally, Rosdahl et al. [30] performed the theoretical and vibrational spectroscopic study of the solvated mercury(I) dimer. The behavior of mercury

575

ions in some environments (i.e. biological, industrial) is not always clearly established at the molecular level. It is helpful to model the complexes and to determine how these ions may act in vivo with each other. It is well known that even slight geometrical distortions or slight modifications in complex structure can modify, enhance or inhibit their behavior and such knowledge is especially important. According to this, in this paper we present the molecular structures and the spectroscopic characterization of mercury(II) thiocyanate complexes in the aqueous media. The studies of the coordination features of [Hg(SCN)n]2n complexes (where n = 2, 3, 4) using different solvation models are based on the comparison of the results of quantum chemical calculations and experimental Raman and UV spectra. Experiment Reagents All chemicals were of analytical-reagent grade, purchased from Sigma Aldrich and used without further purification. Distilled water and heavy water were used throughout the present study as solvents. In order to avoid hydrolysis of standard HgSO4 solutions, both distilled water as well as heavy water was acidified with the least amount of concentrated sulfuric acid. Procedure UV spectroscopic analysis The solutions of [Hg(SCN)n]2n complexes with the metal(II)ligand molar ratios of 1:2, 1:3 and 1:4, all in the level of 104 M, were prepared by mixing the necessary volumes of the standard solutions of 0.02255 M HgSO4 and 0.1 M KSCN. Distilled water for the dilution of mixtures to a known volume was used. UV spectra of the samples were acquired on Lambda 35 UV spectrometer (PerkinElmer) using 2-mm quartz cuvette. Raman spectroscopic analysis The solutions of [Hg(SCN)3], [Hg(SCN)4]2 complexes, both in the level of 102 M, and solution of Hg(SCN)2, in the level of 103 M, were prepared by mixing the necessary volumes of standard solutions of 0.02255 M HgSO4 and 0.5 M KSCN. Distilled water for the dilution of mixtures to a known volume was used. For isotopic exchange analysis heavy water (deuterium oxide D2O) was used as a solvent for the preparation of Hg(SCN)2, [Hg(SCN)3] and [Hg(SCN)4]2 solutions, at the same concentration level. Raman spectra were recorded using Echelle type spectrometer RamanFlex 400 (PerkinElmer, Inc.) equipped with thermoelectrically cooled (50 °C) CCD camera and fiber-optic cable for excitation and collection of the Raman spectra. The 785-nm beam of the diode laser was used as the excitation source. Spectra were collected with an integration time of 10 s. Each spectrum was recorded with accumulation of 30–50 scans, yielding total acquisition time of 300–500 s. The wavenumber axis was calibrated using the polystyrene standard (ASTM E 1840), yielding ±1 cm1 absolute wavenumber accuracy for well defined narrow bands. Intensities were corrected by the NIST intensity standard (SRM 2241). The analysis of Raman spectra was made by subtracting the spectra of acidified distilled water and acidified heavy water from the spectra of each mixture of Hg2+ and SCN ions, respectively. Quantum chemical calculations All the calculations were performed using the Gaussian for Windows package version G03W [31]. The geometry optimization

576

E. Elijošiute˙ et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 115 (2013) 574–582

and frequency calculations were accomplished with DFT method, using B3LYP functional. Almost all calculations except Hg, which was treated with pseudopotentials, were done using 6311++G(d,p) basis set. For the Hg(II) atom Stuttgart relativistic ECP 78MWB basis set was used. To take the solvent effect into account, some calculations were done using Polarizable Continuum Model (PCM), specifically Integral Equation Formalism model, further referenced as IEFPCM. For the calculations involving the simulation of solvent effect, we set SURFACE = WDW, ALPHA = 1.21 and TSNUM = 70 instead of the default settings in order to avoid the oscillatory behavior often encountered during optimization. Several starting geometries failed to converge, and a number of trial structures were required before convergence was achieved. The vibrational frequencies and the intensities of Raman spectra line were calculated using Gaussian 03W package [31]. Only the calculated frequencies of C„N bond were scaled by the 0.9688 scale factor and other were left unchanged [32]. The electronic spectra of Hg(II) complexes were calculated with the B3LYP, PBE1PBE and PBEPBE methods using TD-DFT approach on optimized geometries. For the visualization of vibrational and electronic spectra the Chemcraft graphical program was used [33]. Spectroscopy data were processed and managed using the GRAMS/AI spectroscopy software [34]. Results and discussion Structure modeling of [Hg(SCN)n]2n and SCN In order to elucidate the most reliable structure of free SCN and Hg(SCN)2, [Hg(SCN)3] and [Hg(SCN)4]2 complexes in the aqueous solution, we performed geometry optimizations of each complex using different solvation models (hypothetical cases): with implicit (the solvation effects are calculated by extra terms of the force fields) or explicit (the water molecules are placed around the simulated solute molecule) water molecules in the first coordination shell. All the calculated and experimental structural data of each examined structures are gathered in Table 1. The optimized geometry of SCN ion with CAS and C„N bond distances of 1.67 Å and 1.18 Å correlates well with the experimental values from literature. The differences varying from 0.02 to 0.04 Å may be influenced by the K+ and NHþ 4 cations present in the sample during experimental analysis by other researches. The optimized structures of Hg(SCN)2, [Hg(SCN)3] and [Hg(SCN)4]2 complexes in all hypothetical cases take a bent conformation. The results of optimized structures are given in the Supplementary Fig. S1. Comparing the optimization results of Hg(SCN)2 complex without the solvation effect and using solvation (implicit or/and explicit water molecules), it can be seen that there is a clear differences between the energies of non-solvated complex and using implicit solvation. Also the effect of each implicit or explicit solvation markedly changed the dipole moment (Table 1). According to the calculation results, the most polar molecule is Hg(SCN)2 in nonsolvated environment. Comparing the calculated HgAS bond distances it was observed that both the implicit and explicit solvation separately induces the elongation of bond between Hg and S atoms compared with non-solvated model. But the effect of both solvation models in tandem does not change the HgAS bond length comparing with the explicit solvation effect alone. The CAS and C„N bond lengths almost do not differ in all the cases. The main and clear changes are the energy and dipole moment when the solvation effects are introduced into the system. Based on the simulation data we found that the complex ion [Hg(SCN)3] is the most polar of all analyzed complexes when both the explicit and implicit solvation models were applied. The bond differences between the HgAS atoms of non-solvated [Hg(SCN)3] complex compared with the different solvation models do not vary

significantly. Nevertheless, the same tendency as was discussed about the Hg(SCN)2 complex remains: each chemical environment promotes lengthening of HgAS bond. The optimized geometric parameters of [Hg(SCN)4]2 complex ion showed that the difference of the analyzed parameters is high comparing the complex presented in non-solvated environment with the implicit solvation (Table 1). The differences are also much bigger comparing with the results of Hg(SCN)2 and [Hg(SCN)3] complexes discussed above. The effect of explicit and implicit water molecules at the same time renders the complex ion [Hg(SCN)4]2 more polar compared with rest solvation models or non-solvated [Hg(SCN)4]2 complex ion. It was observed that the length of HgAS bond of [Hg(SCN)4]2 complex ion in each case is the only variable parameter while the lengths of CAS and C„N bond remain constant. In contrast to the former discussions it was found out that the weakest HgAS bond becomes when at the same time the explicit and implicit solvation is involved. The calculated length of HgAS bond is in good agreement with experimentally identified bond length given by other authors [35] when the solvent is water. The deviation is ±0.02 Å, while in the case of [Hg(SCN)4]2 and [Hg(SCN)4(H2O)2]2 ions in nonsolvated environment the calculated and experimental values are identical. The bond angles of SAHgAS and SAC„N in each case also do not fluctuate markedly. However the prominent alteration of HgASAC angle was noticed. The disparity is between 0.1° and 6.1°. Generally there are some basic features that are characteristic for the optimized structures. The molecular modeling data clearly indicate the formation of four-coordinated Hg(II) complexes with four thiocyanate and (or) water ligands. The HgASAC fragment is always bent. Moreover, there is always a slight bending involved for the SAC„N bond angle which varies between 1° and 4°. The HgAS bond lengths increased with increasing the number of ligands. Consequently, the weakest bonding mode between metal ion and SCN ligand is for [Hg(SCN)4]2 complex ion. Such tendency also was noticed experimentally by other researches (Table 1) when the solvent was dimethylsulfoxide (DMSO). Vibrational spectra of [Hg(SCN)n]2n complexes in the aqueous solution The vibrational spectroscopy can help to distinguish between the all examined structures and this is explained in the following. The Raman spectra of aqueous solutions with the established molar ratio of Hg2+ and SCN ion were obtained. The vibrational spectral analysis was performed on the basis of literature and calculation data. The frequencies of the HgAS, CAS, C„N vibrational modes have been measured for the free SCN ion, neutral Hg(SCN)2 molecule, anionic [Hg(SCN)3] and [Hg(SCN)4]2 complexes when the solvent was water. The detailed vibrational assignments of fundamental modes along with the calculated Raman intensities are shown in Table 2. For visual comparison, the observed and simulated Raman spectra are presented in Fig. 1. For the interpretation of calculated and experimentally determined Raman spectra the attention mainly was focused on the analysis of the Raman spectra bands arising from the HgASCN and SCN groups. Two characteristic vibration modes have been assigned as valuable: m1 (HgAS stretching mode) and m2(C„N stretching mode), respectively. The results of our performed Raman analysis for SCN anion are in good agreement with the characteristic values for the mCAS, mC„N stretching and dsAC„N bending modes of the thiocyanate anion listed in literature [36] and with the calculated Raman spectra. Three regions with different frequencies are attributed for free SCN ion in the vibrational spectra (465–460 cm1, SAC„N bending vibration modes, 756–730 cm1, CAS stretching, and 2148–2120 cm1, C„N stretching vibration modes). Unfortunately,

577

E. Elijošiute˙ et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 115 (2013) 574–582 Table 1 Geometrical parameters and energy of Hg(SCN)2, [Hg(SCN)3], [Hg(SCN)4]2 complexes in different hypothetical cases. Complex

Bond Calculated bond lengths (average), Å

Experimental bond lengths (solvent), Å

Angle

Calculated bond angle (average), °

SCN

CAS C„N CAS C„N CAS C„N CAS C„N HgAS CAS C„N HgAS CAS C„N HgAS HgAO CAS C„N

1.67 1.18 1.66 1.18 1.66 1.17 1.66 1.17 2.33 1.70 1.16 2.46 1.68 1.17 2.44 2.19 1.69 1.17

1.69a1, 1.63a2 1.15a1, 1.15a2 1.69a1, 1.63a2 1.15a1, 1.15a2 1.69a1, 1.63a2 1.15a1, 1.15a2 1.69a1, 1.63a2 1.15a1, 1.15a2 2.41 (DMSO)b

SAC„N

180.0

308212.7745

1.6562

SAC„N

180

308271.6987

2.3349

SAC„N

178.5

404193.3402

1.4337

SAC„N

178.5

404193.3402

1.4334 2.7168

617033.1889

0

712957.3253

1.0597

2.44 2.18 1.69 1.17

2.41b (DMSO)

712968.8169

0.689

HgAS CAS C„N HgAS CAS C„N HgAS HgAO CAS C„N HgAS HgAO CAS C„N HgAS

2.43 1.69 1.16 2.45 1.69 1.17 2.50 2.21 1.68 1.17 2.50 2.22 1.68 1.17 2.54

2.464b (DMSO)

178.6 94.6 178.1 180.0 104.1 176.9 138.7 103.5 90.8 97.1 175.7 142.6 102.4 93.5 95.9 178.6 120.0 99.9 176.9 120.0 98.7 177.6 114.7 103.4 96.5 176.7 115.9 101.9 95.8 177.7 109.5

616958.5051

HgAS HgAO CAS C„N

SAHgAS HgASAC SAC„N SAHgAS HgASAC SAC„N SAHgAS SAHgAO HgASAC OAHgAO SAC„N SAHgAS SAHgAO HgASAC OAHgAO SAC„N SAHgAS HgASAC SAC„N SAHgAS HgASAC SAC„N SAHgAS SAHgAO HgASAC SAC„N SAHgAS SAHgAO HgASAC SAC„N SAHgAS

925225.7275

0.0039

925268.8072

0.0852

973220.4722

3.5743

973258.1460

7.4747

1233413.0032

0.0049

CAS

1.68

1233553.2590

0.074

1329394.9234

1.364



SCN solvated [(SCN)(H2O)2] [(SCN)(H2O)2] solvated Hg(SCN)2

Hg(SCN)2 solvated

Hg(SCN)2(H2O)2

Hg(SCN)2(H2O)2 solvated

[Hg(SCN)3]

[Hg(SCN)3] solvated

[Hg(SCN)3(H2O)]

[Hg(SCN)3(H2O)] solvated

2

[Hg(SCN)4]

[Hg(SCN)4]2 solvated

C„N 1.17 HgAS 2.52 CAS

[Hg(SCN)4(H2O)2]2

[Hg(SCN)4(H2O)2]2 solvated

2.464b (DMSO)

2.464b (DMSO)

2.464b (DMSO)

b1

b2

2.531 , 2.547 (DMSO), 2.54c (water)

HgASAC 106.3

2.531b1, 2.547b2 (DMSO), 2.54c (water)

2.531b1, 2.547b2 (DMSO), 2.54c (water)

2.531b1, 2.547b2 (DMSO), 2.54c (water)

1.68

C„N 1.17

b

107 (DMSO)

107b (DMSO)

107b (DMSO)

107b (DMSO)

107b (DMSO)

b

107 (DMSO)

108b (DMSO), 102b (water)

b

b

108 (DMSO), 102 (water)

108b (DMSO), 102b (water)

SAC„N 177.1 SAHgAS 108.7 HgASAC 103.5 SAC„N

Dipole moment

107b (DMSO)

SAC„N 177.8 SAHgAS 109.5 HgASAC 103.6

Energy, kcal/ mol

107b (DMSO)

SAC„N 175.8 SAHgAS 109.5 HgASAC 100.2

1.68

C„N 1.17 HgAS 2.56 CAS

2.41b (DMSO)

1.68

C„N 1.17 HgAS 2.54 CAS

2.41b (DMSO)

Experimental bond angle (solvent), °

1329514.7649 12.4823 108b (DMSO), 102b (water)

178.0

a1,a2

Taken from Ref. [38]. b,b1,b2 Taken from Ref. [37]. c Taken from Ref. [35].

the SAC„N deformation vibrational mode was not revealed in the experimental Raman spectra and also was very weak in the calculated spectra. The calculation data from Table 2 show that the mHgAS vibration is sensitive to the number of SCN ligands. As the number of bound SCN ions increases, the effective charge on the Hg metal decreases and the mHgAS vibration number becomes lower. This tendency was confirmed by the Raman spectral analysis. The decrease of mHgAS

vibration number was observed earlier by Šašic´ et al. [14]. Additionally, the behavior of mHgAS vibration in the case of increasing number of SCN ligands remains when the water is changed to other solvents such as DMSO or DMF [14,15,37]. The presence of the third SCN ion reduces the positive charge of Hg2+ and the experimental mHgAS downshifts by about 22–28 cm1, compared with those of the corresponding solution of Hg(SCN)2. Attachment of fourth ligand to the Hg2+ ion causes a further decrease in the

578

E. Elijošiute˙ et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 115 (2013) 574–582

Table 2 Vibrational assignments of Hg(SCN)2, [Hg(SCN)3], [Hg(SCN)4]2 complexes in the different hypothetical cases obtained from DFT calculations and determined experimentally. Complex

Vibrational assignmentsg

Calculated unscaled frequency Intensityf Scaled frequency (scale (average) H2O (D2O) factor 0.9688) H2O (D2O)

SCN

dSAC„N

460 730 2132 465 742 2120 465 756 2148 465 756 2148 290

w m s vw w s vw w s vw w s m

434 687 2253 211

vw vw vs vw

434 716 2197 245, 302 (245, 289)

vw vw vs m

475 708 2209 (2210) 236, 302 (237, 291)

vw w vs m

474 706 2204 (2202) 255

vw m vs m

mCAS mC„N

422 699 2223

vw w vs

mHgAS

239

m

dSAC„N

mCAS mC„N

428 707 2211

w w vs

mHgAS

245, 282 (246, 271)

m

dSAC„N

mCAS mC„N

425 717 2189, 2219 (2188, 2218)

vw w vs

mHgAS

237, 278 (233, 271)

w

dSAC„N

mCAS mC„N

426 715 2188, 2205 (2186, 2206)

vw w vs

[Hg(SCN)4]2

mHgAS

206, 219

m

dSAC„N

[Hg(SCN)4]2 solvated

mCAS mC„N mHgAS

433 709 2192 217, 232

vw vw vs m

[Hg(SCN)4(H2O)2]2

mCAS mC„N mHgAS

434 714 2190 206, 223 (206, 224)

vw w vs m

405, 433 708,722 2199 (2197) 215, 231 (214, 230)

vw w vs m

mCAS mC„N SCN solvated

dSAC„N

mCAS mC„N [(SCN)(H2O)2]

dSAC„N

mCAS mC„N [(SCN)(H2O)2] solvated

dSAC„N

Hg(SCN)2

mCAS mC„N mHgAS

Hg(SCN)2 solvated

mCAS mC„N mHgAS

Hg(SCN)2(H2O)2

mCAS mC„N mHgAS

Hg(SCN)2(H2O)2 solvated

mCAS mC„N mHgAS

[Hg(SCN)3]

mCAS mC„N mHgAS

dSAC„N

dSAC„N

dSAC„N

dSAC„N

dSAC„N

[Hg(SCN)3] solvated



[Hg(SCN)3(H2O)]



[Hg(SCN)3(H2O)] solvated

dSAC„N

dSAC„N

[Hg(SCN)4(H2O)2]2 solvated

mCAS mC„N mHgAS

Experimental frequency in H2O (D2O)

Experimental frequency according to the literature

2066

750 2068

747e 2068e

2054

750 2068

747e 2068e

2081

750 2068

747e 2068e

750 2068 284 (281, 256)

747e 2068e 266a, 264b (DMSO), 271b (DMF), 279b (water)

2081

451 2182

2147 (2146) 284 (281, 256)

2131c (DMSO) 266a, 264b (DMSO), 271b (DMF), 279b (water)

451 2129

2147 (2146) 284 (281, 256)

2131c (DMSO) 266a, 264b (DMSO), 271b (DMF), 279b (water)

451 2141 (2141)

2147 (2146) 284 (281, 256)

2131c (DMSO) 266a, 264b (DMSO), 271b (DMF), 279b (water)

451 2135 (2134)

2153

2142

2121, 2150 (2121, 2150)

2119, 2137 (2119, 2137)

2124

2122

2130 (2130)

2147 (2146) 2131c (DMSO) 256 (255), 262 (261) 242a, 246b (DMSO), 248b (DMF), 252b (water) 453 714c (water) 2131 (2132), 2137 2119c (DMSO). 2117c (2138) (water) 256 (255), 262 (261) 242a, 246b (DMSO), 248b (DMF), 252b (water) 453 714c (water) 2131 (2132), 2137 2119c (DMSO). 2117c (2138) (water) 256 (255), 262 (261) 242a, 246b (DMSO), 248b (DMF), 252b (water) 453 714c (water) 2131 (2132), 2137 2119c (DMSO). 2117c (2138) (water) 256 (255), 262 (261) 242a, 246b (DMSO), 248b (DMF), 252b (water) 453 714c (water) 2131 (2132), 2137 2119c (DMSO). 2117c (2138) (water) 240 (240), 252 (252) 235a1, 234b (DMSO), 234b (DMF), 238b (water) 454 717 (717) 710d, 717e, 714c(water) 2122 (2122) 2117d, 2117c, 2114e (water) 240 (240), 252 (252) 235a1, 234b (DMSO), 234b (DMF), 238b (water) 454 717 (717) 710d, 717e, 714c (water) 2122 (2122) 2117d, 2117c, 2114e (water) 240 (240), 252 (252) 235a1, 234b (DMSO), 234b (DMF), 238b (water) 454 717 (717) 710d, 717e, 714c(water) 2122 (2122) 2117d, 2117c, 2114e (water) 240 (240), 252 (252) 235a1, 234b (DMSO), 234b (DMF), 238b (water) (continued on next page)

579

E. Elijošiute˙ et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 115 (2013) 574–582 Table 2 (continued) Complex

Vibrational assignmentsg

Calculated unscaled frequency Intensityf Scaled frequency (scale (average) H2O (D2O) factor 0.9688) H2O (D2O)

Experimental frequency in H2O (D2O)

dSAC„N

436 720 2202 (2198)

454 717 (717) 2122 (2122)

mCAS mC„N

vw w vs

2130 (2129)

Experimental frequency according to the literature

710d, 717e, 714c(water) 2117d, 2117c, 2114e (water)

a,a1

Taken from Ref. [37]. Taken from Ref. [14]. Taken from Ref. [15]. d Taken from Ref. [35]. e Taken from Ref. [39]. f,g Abbreviations: s, strong; m, medium; w, weak; vw, very weak; vs, very strong; m, stretching; d, deformation. b c

Fig. 1. The calculated and experimental Raman spectrum of each [Hg(SCN)n]2n complex in the aqueous solution in a range of 320–200 cm1 and 2180–2040 cm1.

mHgAS wavenumber of about 5–12 cm1 compared with the [Hg(SCN)3]. The calculated differences of mHgAS vibrations are greater depending on the each hypothetical case. The calculated and experimentally observed mCAS vibration modes were not informative in our study because the signal always was very weak or weak. Furthermore, it was noticed that the signal of mCAS band gets better with the added number of SCN ligand. In the case of [Hg(SCN)4]2 it was possible to identify the mCAS vibration at 717 cm1. There is a lack of such data in literature too. We were able to find only some experimental values of mCAS vibration in water solution (Table 2) and only for [Hg(SCN)3] and [Hg(SCN)4]2 complexes. The greatest changes in Dm are identified for mC„N vibration according to the experimental and calculated data. The tendency is the same: the increased number of SCN ligands shifts the mC„N vibration to the lower wavenumber position. Generally, it can be stated that the more ligands are introduced in the complexation, the weaker HgAS and C„N bond is formed. This was confirmed

by calculated bond length (Table 1). Fig. 1 shows the Raman spectra of Hg2+ ion coordinated series in the mHgAS and mC„N regions. All the peaks of mHgAS and mC„N stretching modes are sufficiently well separated from the peaks of the free SCN. In general the solutions containing [Hg(SCN)3] complex ion gave spectra which were more difficult to analyze. It is known that [Hg(SCN)3] complex ion is prone to partial disproportionation to Hg(SCN)2 and [Hg(SCN)4]2 species [37]. Such feature has been recognized during preparation procedure of the solution containing [Hg(SCN)3] complex ion. The solution became turbid due to the formation of the slightly soluble Hg(SCN)2 complex. Raman spectra of the solution, containing [Hg(SCN)3] complex ion, exhibited three well separated peaks at 2143 cm1, 2131 cm1 and 2070 cm1. The first one is very similar to mC„N vibrational mode of Hg(SCN)2 complex and the third is associated with mC„N vibration of free SCN group. In order to improve the assignment of Raman spectra of [Hg(SCN)3] complex ion, we subtracted the spectra of Hg(SCN)2 and [Hg(SCN)4]2 complexes from the spectrum of [Hg(SCN)3]

580

E. Elijošiute˙ et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 115 (2013) 574–582

complex. Difference spectrum displayed two peaks 2131 cm1 and 2137 cm1 in mC„N region and two 256 cm1, 261 cm1 peaks in mHgAS region, respectively. This double-peak character of mentioned regions was also displayed in the calculation results (Table 2). We presume that the assignment of 2117 cm1 peak to mC„N vibration for both [Hg(SCN)3] and [Hg(SCN)4]2 complexes of other researches [16] is incorrect and the disproportionation feature of [Hg(SCN)3] complex ion was not taken into account. In addition, we observed the peak of mC„N vibration mode of free SCN ion at 2066 cm1 in each analyzed [Hg(SCN)n]2n aqueous solution. On the contrary, this peculiarity is not characteristic for mercury(II) thiocyanate complexes in DMSO solution [37]. Additionally, the calculated and experimental Raman spectra showed previously unnoticed splitting of mHgAS peak of [Hg(SCN)4]2 complex ion (Table 2, Fig. 1). The analysis of experimental Raman spectra of Hg(SCN)2 complex was the most complicated due to its low solubility. Therefore, the weak signals with low signal to noise ratio have been obtained. Also the splitting character of mHgAS compared with [Hg(SCN)3] and [Hg(SCN)4]2 ions in the experimental spectra was not recognized. Based on the comparison of the calculated and experimental Raman spectra of each mercury(II) thiocyanate complex in different hypothetical case (with implicit or explicit water molecules) we tried to determine the most reliable structure of each complex. In the case of Hg(SCN)2 and [Hg(SCN)3] ion it can be seen that the calculated Raman spectra correlates well with the experimental one when the Hg(SCN)2 complex includes two water molecules and the [Hg(SCN)3] complex ion – one water molecule as ligand. The addition of implicit water molecules slightly improves the results. Based on the data for [Hg(SCN)4]2 complex ion the other trend was noticed. The implicit solvation model around the [Hg(SCN)4]2 complex ion in aqueous media represented the most probable mode and the calculated model with the explicitly added water molecules only impaired the results. These data clearly indicate the formation of four-coordinated mercury(II) thiocyanate complexes in the water environment. Vibrational spectra of [Hg(SCN)n]2n complexes in the heavy water solution In order to perform the assignments of fundamental vibrational modes of each [Hg(SCN)n]2n complex properly as well as verify the interaction of ligands with Hg2+ ion, the isotopic substitutions were made. Presumably, if the Hg2+ ion in the all three complexes is four-coordinated, the H2O replacement with D2O will have the minimal influence on fundamental vibrations of [Hg(SCN)4]2 complex and maximal on Hg(SCN)2 complex. The performed

calculations with two H2O/D2O molecules for the [Hg(SCN)4]2 complex ion showed almost identical peaks in the Raman spectra. In excellent agreement with the calculated spectra, the experimental spectra in both these cases were identical. In the case of Hg(SCN)2 the calculations showed that the frequency of mC„N stretching vibration was not sensitive to isotopic exchange. On the contrary, the second mHgAS vibration at 302 cm1 decreased by 13 cm1 (while the first one at 245 cm1 was steady) via H2O/ D2O exchange modeling (Table 2). It should be noted that the vibrations at the 300–250 cm1 region are complex vibrations of HgAS and HgAO stretching. Additionally, it is obvious that the isotopic substitution has substantial influence to the mHgAS and mHgAO stretching modes. This tendency was observed in the experimental Raman spectra of Hg(SCN)2 as well. After the isotopic exchange the mC„N stretching remains unchanged and the mHgAS stretching decreased by 3 cm1. Additionally, the Raman spectra of Hg(SCN)2 in D2O exhibited one further mHgAS peak at 256 cm1 (Fig. 2). The calculated Raman spectra of Hg(SCN)2 in the D2O and H2O also displayed two well separated mHgAS stretching modes. It can be assumed that the H2O environment hides one of HgAS stretching modes due to weak intensity of signal and low solubility of Hg(SCN)2 complex in water. In general, the isotopic changes have played an important role in performing better vibrational assignments. We were able to identify the second mHgAS vibrational mode in the case of Hg(SCN)2 and confirm that the Hg2+ ion in the all complexes is fourcoordinated. UV spectra of [Hg(SCN)n]2n complexes in the aqueous solution The experimental electronic spectra supplemented by simulated spectra of the Hg(SCN)2, [Hg(SCN)3], [Hg(SCN)4]2 complexes are missing. According to this we performed the research in order to assess the performance of the different functionals on the calculation results of UV spectra. We have considered different functionals, including different solvation effects, in order to understand how taking into account these contributions the accuracy of reproducing the experimental spectra can be improved. Starting from the optimized geometry obtained by DFT B3LYP calculations we made next time dependent TD-DFT calculations for the visualization of electronic spectra. The spectra were calculated at B3LYP, PBE1PBE and PBEPBE levels of theory for each hypothetical case. For comparison with the experimental spectra, the obtained intensities were broadened using Lorentzian shape functions with a constant halfwidth, which was adjusted manually to match the corresponding experimental spectrum (25 cm1). Because of difficulties to find the maximum of peaks in the lower intensity region in the experi-

Fig. 2. The experimental Raman mHgAS and mC„N stretching vibrational spectra for the [Hg(SCN)n]2n complexes in the heavy water solution.

E. Elijošiute˙ et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 115 (2013) 574–582

mental electronic spectra, the peak fitting routine, using GRAMS/AI spectroscopy software, was applied. In Fig. 3 we report the experimental spectrum of Hg(SCN)2, [Hg(SCN)3], [Hg(SCN)4]2 complexes together with the TD-DFT calculated ones. The experimental spectrum of Hg(SCN)2 presents an intense and sharp maximum in the region of the spectrum extending from 208 to 240 nm with maximum of 222 nm. Moreover, a much lesser intense absorption band was identified near the latter one centered at about 267 nm. Actually, we observed that the use of PBE1PBE functional corresponds better to the experimentally obtained data in the each hypothetical case. Indeed, the used PBE1PBE functional together with both explicit and implicit solvation models exhibited an absorption maximum and band shape in better agreement with the experimental results. In this case we got the deviation from the experimental maximum of about 4 nm for most intense peak and about 2 nm for the second one, even if the band intensities were less well reproduced. The worst results were obtained using the PBEPBE functional. The experimental absorption spectra of [Hg(SCN)3] complex is similar to experimental spectra of Hg(SCN)2 complex. The only differences lie in the less intensive absorption band region. Here the hipsochromic shift with the maximum of 262.5 nm was observed and the intensity of 222 nm peak is higher of about 0.05 compared with the peak of 222 nm of Hg(SCN)2 complex. The computed spectrum using the PBE1PBE hybrid functional in common with the explicit solvation showed a very good agreement with its experimental counterpart over the entire spectra. Even spectral features with lower intensity were nicely reproduced. The experimental electronic spectra of [Hg(SCN)4]2 complex ion do not differ greatly from the spectra of Hg(SCN)2 and [Hg(SCN)3] complexes. The main peak centers at 222 nm and the second one (identified using fitting) has maximum at 260 nm which shows the small hipsochromic shift compared with the spectra of [Hg(SCN)3] complex. As was previous established for Hg(SCN)2 and [Hg(SCN)3] complexes, the theoretical description using the PBE1PBE functional is the best for [Hg(SCN)4]2 complex too. However, unlike the previous discussed, the implicit solvation model is the most characteristic for [Hg(SCN)4]2 complex. The computed electronic spectrum also showed a good agreement with the experiment one indicating that analyzed systems were treated accurately

581

by means of TD-DFT method. The calculated maximum in the less intensive absorption band region as well as the intense and sharp one differ by 6 nm from experimentally determined values. The obtained TD-DFT calculation results show that the selection of the functional, that sometimes may be not fairly applied, can bring important differences on the final simulated absorption spectra. In general all the computed spectra at PBE1PBE level shows an almost perfect agreement with the experimental electronic spectra and even spectral features with lower intensity are nicely reproduced. Somewhat surprisingly, the arrangement of peaks from the lowest intensity of Hg(SCN)2 complex to highest intensity of [Hg(SCN)4]2 complex is perfectly restored too. Additionally, the experimental and computed data showed that the increasing number of SCN ligands in the analyzed complexes with Hg2+ ion, induces the hipsochromic shift in the less intense absorption band region. Conclusions In this study the vibrational and electronic spectra analysis of named [Hg(SCN)n]2n complexes by comparing the calculated and experimental Raman and UV data were introduced here for the first time. The calculations of geometric parameters, vibrational symmetric and asymmetric frequencies were made by using DFT (B3LYP) method with Stuttgart relativistic ECP 78MWB and 6311++G(d,p) basis set for Hg atoms and H, C, N, O, S atoms, respectively. The simulation of UV spectra was carried out using B3LYP/ B3LYP, B3LYP/PBE1PBE, and B3LYP/PBEPBE levels of theory. The observed and calculated frequencies are in good agreement. B3LYP/PBE1PBE method seemed to be more appropriate in calculating the electronic spectra. The work also delineated the influence of different solvation models on both geometry and characteristics of vibrational and absorption spectra. The calculated and experimental data regarding the H2O/D2O exchange provided a strong support for the clear frequency assignments. Our findings imply that mHgAS peak has a double-peak character in the all titled [Hg(SCN)n]2n complexes. Furthermore, we conclude that the [Hg(SCN)3] and [Hg(SCN)4]2 complexes have their own assignments at the mC„N region. The theoretical and experimental results indicate that the most reliable structure is four-coordinated mercury(II) thiocyanate complexes. Acknowledgement The valuable help from the Department of Organic Technology of Kaunas University of Technology for UV measurements are acknowledged. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.saa.2013.06.072. References

Fig. 3. The calculated and experimental absorption spectra of Hg(SCN)2, [Hg(SCN)3], [Hg(SCN)4]2 complexes in aqueous solution.

[1] G.F. Nordberg, B.A. Fowler, M. Nordberg, L.T. Friberg, Handbook on the Toxicology of Metals, third ed., Academic Press, United States of America, 2007. [2] G. Mahmoudi, A. Morsali, L.G. Zhu, J. Polyhedron 26 (2007) 2885–2893. [3] M.M. Saeed, S.M. Hasany, M. Ahmed, J. Talanta 50 (1999) 625–634. [4] J.G. Małecki, T. Gron, H. Duda, J. Polyhedron 36 (2012) 56–68. [5] C. Buda, A.B. Kazi, A. Dinescu, T.R. Cundari, J. Chem. Inf. Model. 45 (2005) 965– 970. [6] J.G. Małecki, B. Machura, A. S´witlicka, T. Gron, M. Balanda, J. Polyhedron 30 (2011) 746–753. [7] N. Fukushima, G. Iisaka, M. Saito, K. Waizumi, J. Inorg. Chim. Acta 255 (1997) 211–214. [8] B. Machura, M. Wolf, J. Palion, A. S´witlicka, I. Nawrot, K. Michalik, J. Struct. Chem. 22 (2011) 1053–1064. [9] J.S. Gancheff, A. Acosta, D. Armentano, G. De Munno, R. Chiozzone, R. Gonzalez, J. Inorg. Chim. Acta 387 (2012) 314–320.

582

E. Elijošiute˙ et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 115 (2013) 574–582

[10] J.G. Małecki, B. Machura, A. S´witlicka, J. Struct. Chem. 22 (2011) 77–87. [11] A.A. Khandar, A. Klein, A. Bakhtiari, A.R. Mahjoub, R.W.H. Pohl, J. Solid State Chem. 184 (2) (2011) 379–386. [12] B. Machura, A. S´witlicka, I. Nawrot, J. Mrozin´ski, R. Kruszynski, J. Polyhedron 30 (5) (2011) 832–840. [13] D.A. Schmidt, I. Kopf, E. Brundermann, J. Laser Photon. Rev. 6 (3) (2012) 296– 332. [14] S. Šašic´, A. Antic´-Jovanovic´, M. Jeremic´, M. Lalic´, J. Spectrosc. Lett. 28 (8) (1995) 1217–1224. [15] S. Šašic´, A. Antic´-Jovanovic´, M. Jeremic´, J. Chem. Soc. Faraday Trans. 94 (1998) 3255–3260. [16] S. Šašic´, M. Jeremic´, A. Antic´-Jovanovic´, J. Raman Spectrosc. 29 (1998) 321–325. [17] L. Riauba, G. Niaura, O. Eicher-Lorka, E. Butkus, J. Phys. Chem. A 110 (2006) 13394–13404. [18] C. Hopa, R. Kurtaran, A. Azizog˘lu, M. Alkan, N. Burcu Arslan, C. Kazak, Z. Anorg. Allg. Chem. 637 (2011) 1238–1245. [19] C. Jesintha John, T.S. Xavier, M. Amalanathan, I. Hubert Joe, V.K. Rastogi, J. Spectrochim. Acta A 86 (2012) 174–180. [20] M. Shukla, N. Srivastava, S. Saha, T.R. Rao, S. Sunkari, J. Polyhedron 30 (2011) 754–763. [21] A.K. Rai, W. Fei, Z. Lu, Z. Lin, J. Theor. Chem. Acc. 124 (2009) 37–47. [22] C. Cappelli, B. Mennucci, S. Monti, J. Phys. Chem. A 109 (2005) 1933–1943. [23] J. Kim, J. Kapitán, A. Lakhani, P. Bourˇ, T.A. Keiderling, J. Theor. Chem. Acc. 119 (2008) 81–97. [24] J. Kubelka, R. Huang, T.A. Keiderling, J. Phys. Chem. B 109 (2005) 8231–8243. [25] M. Kato, S. Izuka, T. Fujihara, A. Nagasawa, S. Kawai, T. Tanaka, T. Takayanagi, J. Inorg. Chim. Acta 370 (2011) 304–310. [26] J. Cirello-Egamino, I.D. Brindle, J. Anal. 120 (1995) 183–186. [27] X.-G. Li, Q. Dou, M.-R. Huang, J. Montash. Chem. 139 (2008) 1157–1162.

[28] G. Chillemi, G. Mancini, N. Sanna, V. Barone, S. Della Longa, M. Benfatto, N.V. Pavel, P. D’Angelo, J. Am. Chem. Soc. 129 (2007) 5430–5436. [29] T.S. Hofer, B.R. Randolf, B.M. Rode, J. Chem. Phys. 349 (2008) 210–218. [30] J. Rosdahl, I. Persson, L. Kloo, K. Stahl, J. Inorg. Chim. Acta 357 (2004) 2624– 2634. [31] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, 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. Fukada, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Know, H.P. Hratchian, J.B. Cross, 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. Dannenburg, 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, 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, J.A. Pople, Gaussian 03, Revision D.01, Gaussian, Inc., Wallingford, CT, 2004. [32] J.P. Merrick, D. Moran, L. Radom, J. Phys. Chem. A 111 (45) (2007) 11683– 11700. [33] www.chemcraftprog.com. [34] www.thermoscientific.com. [35] T. Yamaguchi, K. Yamamoto, H. Ohtaki, J. Bull. Chem. Soc. Jpn. 58 (1985) 3235– 3243. [36] G. Socrates, Infrared and Raman Characteristic Group Frequencies: Tables and Charts, third ed., Chichester, England, 2001. [37] I. Person, Å. Iverfeldt, S. Ahrland, J. Acta Chem. Scand. A 35 (4) (1981) 295–304. [38] J.A. Dobado, R. Uggla, M.R. Sundberg, J. Molina, J. Chem. Soc. Dalton Trans. 3 (1999) 489–496. [39] K.A. Taylor, T.V. Long, R.A. Plane, J. Chem. Phys. 47 (1967) 138–142.