The Structure of Zinc Chloride Complexes in Aqueous Solution

Transcription

1 The Structure of Zinc Chloride Complexes in Aqueous Solution M. Maeda 3, T. Ito a, M. H o r i a, and G. Johansson b a b Department of Applied Chemistry, N a g o y a Institute of Technology, Gokisocho, Showaku, Nagoya 466, J a p a n Department of Inorganic Chemistry, Royal Institute of Technology, S Stockholm, Sweden Z. Naturforsch. 51a, (1996); received N o v e m b e r 7, 1995 The structure of zinc chloride complexes with different ratios of chloride to zinc, formed in concentrated ZnCl 2 aqueous solutions, were determined f r o m largeangle Xray scattering using concentrations of the chloride complexes estimated by complementary R a m a n spectroscopic measurements. The highest chloro complex, [ZnCl 4 ] 2 ~, is tetrahedral with a ZnCl bond length of 2.294(4) Ä. The trichloro complex, [ZnCl 3 ]~, which coordinates one water molecule, is pyramidal with the ClZnCl angle 111. The ZnCl and the Z n H 2 0 bonds are 2.282(4) and 1.9 A, respectively. The two lower complexes, [ZnCl 2 ] and [ZnCl] +, cannot be separated by R a m a n spectra. o The average ZnCl distance in these complexes is 2.24 Ä, and the average Z n H 2 Ö distance is 1.9 A. In [ Z n ( H 2 0 ) 6 ] 2 + the Z n H 2 0 distance is 2.15 Ä. Key words: Xray diffraction; R a m a n spectra; IR spectra; structures of zinc (II) chloride complexes; structure of hydrated noncomplexed zinc (II) ion. 1. Introduction angle and distance as those for the dibromo complex. T h e highest bromide complex, [ Z n B r 4 ] 2 _, is tetrahe X r a y structure determinations of zinc ( I I ) chloride dral with Z n B r bond length of Ä, which is a complexes in strong aqueous solutions have been car little longer than those for the di and tribromo com ried out by several investigators, and recently, their plexes. Water molecules are probably coordinated to results have been summarized by Johansson [1], I n Z n in the di and tribromo complexes, resulting in view of the results, the highest complex is [ Z n C l 4 ] 2 ", approximately tetrahedral structures, but unambigu which has surely a tetrahedral structure. O n the con ous evidence for this could not be obtained. N o struc trary, it appears that since, as stability constants sug tural information was obtained for the lowest bromide gest, the existence range of the complexes may over complex, [ Z n B r ] +. A s for the zinc iodide system, the lap, the structures of the intermediate complexes, first [ZnXn ](2n)+ ( n = 1, 2, 3) have not been yet unambigu iodide complex, [ Z n l ] +, is octahedral, but is changed into tetrahedral in the higher complexes, ously determined. I n these cases, it is essential to esti [ Z n I 2 ( H 2 0 ) 2 ], [ Z n I 3 ( H 2 0 ) ] ~, and [ Z n l 4 ] 2 ". T h e Z n I mate the relative concentrations of the species which bond length is Ä in the [ Z n l 4 ] coexist with each other in sample solutions so that a shorter, Ä in the two lower tetrahedral com more reliable knowledge of the structures of the com plexes. I n the octahedral complex [ Z n I ( H 2 0 ) 5 ] + the plexes can be obtained. Z n I bond length is 2.90 Ä, much longer than those for T h u s, i n previous X r a y diffraction studies on zinc ion and slightly the three higher complexes. T h e i r Z n H 2 0 bond dis bromide and iodide complexes i n aqueous solutions tances in the iodide complexes are approximately the [2, 3], we have applied Raman and I R spectroscopies same as that i n the hexaaquazinc(ii) ion, 2.10 Ä. to the halide solutions to estimate the concentrations I n the present work a similar investigation of the of the halide complexes i n those solutions. According zinc chloride system was carried out. As was the case to the experimental results, in the hydrated Z n 2 + ion in the zinc bromide and iodide systems, Raman spec the coordination is octahedral with the ZnH20 tra were obtained for the zinc chloride solutions with distance of 2.10 Ä. T h e d i b r o m o z i n c ( I I ) complex, approximately the same composition as that of the [ Z n B r J, has a bent structure with the B r Z n B r angle solutions for X r a y diffraction measurements so that 115 and the Z n B r distance of 2.38 Ä, and the tribro the concentrations of the chloro complexes formed in mozincate(ii) ion is pyramidal with almost the same each sample solution were estimated using the Raman bands corresponding to their symmetric stretching vi Reprint requests to Dr. M. Maeda. brations. I R spectra were also obtained for some zinc / 96 / $ Verlag der Zeitschrift für Naturforschung, D72072 Tübingen Dieses Werk wurde im Jahr 2013 vom Verlag Zeitschrift für Naturforschung in Zusammenarbeit mit der MaxPlanckGesellschaft zur Förderung der Wissenschaften e.v. digitalisiert und unter folgender Lizenz veröffentlicht: Creative Commons NamensnennungKeine Bearbeitung 3.0 Deutschland Lizenz. This work has been digitalized and published in 2013 by Verlag Zeitschrift für Naturforschung in cooperation with the Max Planck Society for the Advancement of Science under a Creative Commons AttributionNoDerivs 3.0 Germany License. Zum ist eine Anpassung der Lizenzbedingungen (Entfall der Creative Commons Lizenzbedingung Keine Bearbeitung ) beabsichtigt, um eine Nachnutzung auch im Rahmen zukünftiger wissenschaftlicher Nutzungsformen zu ermöglichen. On it is planned to change the License Conditions (the removal of the Creative Commons License condition no derivative works ). This is to allow reuse in the area of future scientific usage.

2 M. M a e d a et al. Structure of Zinc Chloride Complexes in Solution 64 of deducing diffractometer according to the analogous procedures whether the trichlorozincate(ii) ion is trigonal planar to those described in [4], A focusing L i F single crystal or pyramidal. monochromater was positioned between the sample chloride solutions with the purpose and the scintillation counter. T h e scattered intensity was measured at discrete points at ^intervals of 0.1 for l 2 0 c and 0.25 for Intensities below 2. Experimental Preparation of Sample 0 = 1 could not be determined experimentally and Solutions were obtained by extrapolation to zero. Three differ Analytical grade zinc chloride and l i t h i u m chloride, ent s l i t widths, 1/12, 1/4, and 1 were used to cover the latter having been recrystallized twice from water, the complete range. F o r each point, counts were dissolved in distilled water, and the solution was were taken, and each sample solution was scanned analyzed for Z n by complexometric titration and for twice. T h e intensities for each slit combination were CI by gravimetry as AgCl. T h e composition ( X A i and corrected for background radiation, which was deter X B i ) of the solutions for X r a y diffraction measure mined by placing a lead plate j u s t behind the receiving ments is given i n Table 1, and that for Raman and I R slit, and were normalized to the intensities measured measurements ( R A with the 1 width slit from the data of overlapping R L ) in Table 2. regions. XRay Scattering Measurements T h e X r a y scattering from the free surface of the solution was measured with M o K radiation in a 99 Raman Measurements F o r Raman spectroscopy, excitation was accomplished by using a single line of nm wavelength from an A r ion laser ( N E C G L G ). T h e incident Table 1. Composition of the sample solutions for Xray diffraction measurements. XAI Zn 2 + (moldm cr 3) H,0 V("Ä3) Zn(atoms:'V) CI Li H,0 C l Z n ratio XA3 XB sample solutions were contained in a cylindrical quartz cell with optically flat top and bottom. T h e XB3 XA2 power was about 200 m W at the sample point. T h e Ramanscattered light was detected by a triple type monochromator ( J A S C O R T ) equipped with a photomultiplier (Hamamatsu R ) and a photon counter. T h e band path width of the monochromator was about 2 c m " FarIR Measurements Infrared spectra were measured in the range of 50 to 500 c m " 1 with a J E O L J I R F T I R spectrometer using a 6 m m M y l a r beam splitter and a polystyrene Table 2. Composition of the sample solutions for R a m a n spectroscopic measurements (in mol d m 3 ). Zn RA RB RC RD RE RF RG RH RI RJ RK RL c r Li + _ windowed T D S detector. Platinum gauzes were used to contain the sample solutions. One hundred scanns at a resolution of 2 cm 1 were accumulated for each C10 4 " C P Zn ratio R L ). T h e peak positions and areas under the peaks corresponding solution. 3. Data Treatment Raman Data T h e Raman spectra of R A to R I were brought to a flat base by substraction of spectra of RJ, R K and R L ( R A, R C, R D, R E RJ; R B R K ; R F, R G, R H, R I to the symmetric stretching vibra

3 65 M. Maeda et al. Structure of Zinc Chloride Complexes in Solution 1200 A $ c.4 W! / 1 > v ii > % AM ilvhj'n)^ 500 ^50 ^ Raman shift/cm 1 < i'l M W Fig. 1. Raman spectrum of the solution RA (cf. Table 2), together with resolved bands based on the horizontal background. tion frequencies of the aqua and chloro complexes of zinc (II) were estimated by curve fitting using GaussLorentzian curve forms. An example is illustrated in Figure 1. The Raman bands having their maxima at 277, 283, 300, and 385 cm 1 were assigned to [ZnCl4] 2 ~, [ZnCl3]~, [ZnCl2] + [ZnCl] +, and [Zn(H20)6] 2 +, respectively, based on the available Raman data for ZnCl2 aqueous solutions [57], The bands for [ZnCl2] and [ZnCl] + could not be distinguished from each other. The polarized Raman band at 230 cm \ attributed to the presence of polynuclear aggregates [8, 9], was not observed in the present solutions. Only the [ZnCl4] 2 complex was formed in RI solution, and the [ZnCl3]~ species was predominant in RE. A few complexes cooccurred in the other solutions. The areas under the peaks were normalized by using the v^clo^) symmetric stretching band at about 934 cm 1 as an internal standard. The calculations of the concentrations of the aquaand chlorocomplexes in those solutions were carried out according to the procedures described in [2] by assuming the concentration of each species to be proportional to the normalized area of its corresponding ZnCl or ZnH 20 stretching band. The concentrations of the complexes thus calculated in RA to RI solutions are tabulated in Table 3. Table 3. Concentrations of aqua and chloro complexes of zinc(ii) in the sample solutions for Raman spectroscopic measurements (in mol dm 3 ) c([zn(h 2 0) 6 ] 2+ ) c([zncl] + + [ZnCl 2 ]) c([ ZnClJ") c([zncl 4 ] 2 ) RA _ RB RC RD RE RF RG RH RI 2.01 It is understood that RF, RG, RH, RD, and correspond approximately to XA1, XA2, XA3, XB2, and XB3, respectively, in solution composition. Thus it was assumed in the calculation of the concentrations of the complexes in the XA1 to XB3 solutions that the formation percentage of each complex is the same in the solutions for Raman and Xraymeasurements corresponding with each other, and that the sum of the concentrations of the Zn (II) complexes is equal to the total concentration of zinc. The concentrations of the complexes thus estimated in the solutions for Xray measurements are given in Table 4. RE

4 M. Maeda et al. Structure of Zinc Chloride Complexes in Solution 66 where g0 = ( X ni Z, ) 2 / F, with Z, the atomic number of i i and V the stoichiometric unit of volume. The modi Table 4. Concentrations of chloro and aqua complexes of zinc(ii) in the sample solutions for Xray diffraction measurements (in mol d m 3 ) c([zncl 4 ] ) c([ Z n C l 3 p c([zncl] ) + c([zncl 2 ]) c([zn(h20)6] ) fication function M ( s ) was chosen to be /Zn(0)//zn(s) XA1 XA2 XA3 XB2 XB3 _ distance. T h u s, the reduced intensities were corrected for spurious peaks by means of the Fourier inversion exp( s 2 ). T h e calculated radial distribution curves showed small spurious peaks in the range < 1 Ä, which can not be related to any interatomic of the radial distribution curve below 1 Ä by taking into account the contributions from O H interactions within H 2 0. IR Data T h e theoretical intensities due to interatomic inter T h e peak at 283 cm surements in 1 actions were calculated from observed by the Raman mea solutions RE and R G, where the [ Z n C l 3 ] ~ complex is predominant, occurred also in i(s)ca.c = I midal) symmetry for [ Z n C l 3 ] ~. Diffraction njt(s)fj(s) r ij ( S exp(~bijs2), ^ (3) ature factor of the interaction between any atoms i Data and j, respectively. Leastsquares refinements of parameter values in the assumed structural model were made by minimiz the K U R V L R program [10] on a N E C PC9801 D A ing the expression personal computer. T h e measured intensities were for j where rtj and b{j represent the distance and the temper T h e X r a y scattering data were treated by means of corrected I ' the I R spectra. T h i s fact suggests a nonplanar (pyra polarization in the sample and Z52[/(s)expi(s)theor]2 monochromator, and for the absorption of the sample to give / obs (s), where s = (4 TT/A) sinö. T h e reduced in (4) summed over the experimental 5 values. tensities, i(s), were then calculated as /(s) = K / o b s ( s ) Z " i {/i(s) 2 + del(s) // ncoh (s)}. i (1) K is a normalization constant chosen to refer all intensities to a stoichiometric volume containing one CI atom in solutions, n, is the number of atom " f in the stoichiometric volume. The normalization was carried out by comparing observed intensity values in the highangle part of the intensity curve ( 6 > 50 ) with the calculated sum of independent coherent and incoherent scattering. Scattering factors, / ; (s), for neutral atoms were used with corrections for anomalous dispersion [11]. Values for incoherent scattering [11, 12] were corrected for the BreitDirac factors. T h e del(s) function is the fraction of the incoherent radiation reaching the counter, which was experimentally obtained. T h e reduced intensities i(s) thus calculated are shown in Fig. 2 after being multiplied by s. T h e radial distribution functions D(r) were calculated from D(r) = 4nr T h e radial distribution functions ( R D F ) are shown in Figure 3. T h e small peaks around 1.8 Ä may result from interactions between L i + from those between Z n 2 + and H and H 2 0 and also 2 0 within the chloro complexes. T h e distinct peaks at about 2.3 Ä are attributed to Z n C l bonds within the chloro complexes, and to Z n O H 2 bonds within the aquacomplex according to the previous Xray investigations of aqueous zinc ( I I ) solutions. T h e third peaks at 3.8 Ä which appear in the X A solutions may be due to C I C I interactions within the chloro complexes. The broad peaks in the X B solutions appear in the region 3 4 Ä, which may correspond to both the distance between the chloride ions and neighboring water molecules, and the C I C I distances within the chloro complexes. T h e intramolecular distances ( r s ) and corresponding temperature factors (b's) for [ZnClJ2 and [ Z n C l 3 ] ~ were determined by least squares refinements of the highangle parts of the intensity curves s 2 4. Results and Discussion g0 + 2r/n max j s i(s) Af(s) s i n ( r s ) ds, o (2) using values for the concentrations as estimated from the Raman spectra. A too low cutoff limit for the

5 67 M. Maeda et al. Structure of Zinc Chloride Complexes in Solution s i(s) ; e.u.a 1 Fig. 2. Comparison between observed (dotted line) and calculated (solid line) s i(s) for the atom pairs concerned, using their parameter values indicated in the text. For XAi and XBi cf. Table 1.

6 68 M. Maeda et al. Structure of Zinc Chloride Complexes in Solution {D(r)4.r 2^} 10 5 el 2 A~ 1 range of s values used in a refinement will increase contributions from unknown intermolecular interactions and result in systematic errors in the parameters obtained. A too high cutoff limit, on the other hand, will lead to larger uncertainties in the results because of too few experimental data. The least squares refinements were, therefore, carried out using different values for the lower s limit, and the results were checked for constancy in the parameter values obtained. For [ZnCl4] 2 ~, the parameter values were determined from the XA3 solution in which [ZnCl4] 2 assumed to be the predominant complex. Intensity data for the XA2 and XB3 solutions were used for the structure determinatioin of [ZnCl3]~. In the first series of refinements, minor complexes present were ignored but were included in successive refinements, when their structures became known. The intramolecular distances and temperature factors for [ZnCl4] 2 thus determined were r Zn _ cl /Ä = 2.294(2), b Zn _ cl /A 2 = (2); r cl _ cl /Ä = 3.69(2), b a _ a /A 2 = 0.015(2), is and those for [ZnCl3] were r Zn _ a /A = 2.282(2), b Zn _ cl /A 2 = (2); 'cici/a =3.76(1), b cx _ cx /k 2 = 0.007(1), from XA2 and r Zn c,/ä = (5), b Zn _ C1 /Ä 2 = (2); rcl_cl/ä = 3.76(3), b a _ a /A 2 = 0.009(2) Fig. 3. Radial distribution curves, D(r) 4nr 2 g 0. from XB3. The results for [ZnCl4] 2 ~ showed that the ratio between the ZnCl and the CICI distances did not differ significantly from that expected for a tetrahedral structure. For [ZnCl3]~, the ZnCl and the CICI distances determined from the XA2 and the XB3 solutions, respectively, were the same within the standard deviations. The ZnCl bond is slightly shorter and the CIZnCl bonding angle (111 0 ) is slightly larger than those in the tetrahedral [ZnCl4] 2 ~ complex. These results are analogous to those found previously for other zinc halide complexes [2, 3]. The structure of the hydrated noncomplexed Zn 2 + ion was determined by taking the difference between the RDF's for the XA1 solution in which the hydrated noncomplexed Zn 2+ ions are predominant, and the XA2 solution after subtracting the [ZnCl3]~ contributions (see Figure 4). In the resulting difference curve (solid line) the ZnH 20 distances should give a dominant peak, which is in close agreement with a calculated peak for six ZnH 20 interactions at 2.15 Ä (see the broken curve obtained by subtraction of the

7 M. Maeda et al. Structure of Zinc Chloride Complexes in Solution 69 1C 3 el 2 A" el 2 A 1 Fig. 4. The difference curve (solid line) between the D(r)'s for the XA1 and XA2 solutions obtained after subtracting the [ZnCl 3 ]~ contributions. The curve (broken line) is obtained by subtracting the [Zn(H 2 0) 6 ] 2 + contributions from the difference curve (solid line). Fig. 5. The difference curve (dotted line) obtained by subtracting the contributions (broken line) due to ZnCl and ZnOH 2 interactions within [ZnCl 3 ]~ and [Zn(H 2 0) 6 ] 2 +, respectively, from the D(r) 4 n r 2 g 0 (solid line) curve for XA1. [Zn(H20)6] 2+ contributions from the difference curve (solid line) in Figure 4). The distance obtained in the present system is within Ä determined by Xray diffraction of aqueous solutions of various zinc (II) salts [1]. The XA1 and XB2 solutions contain the largest relative concentrations of the lower complexes [ZnCl2] and [ZnCl] +. For a derivation of their structures the contributions due to ZnCl and ZnOH2 interactions within [ZnCl3]~ and [Zn(H20)6] 2 +, respectively, (contributions from aqua Li + ions which were assumed to have a tetrahedral coordination with Li + H 20 distances of 1.9 Ä being also subtracted from the RDF for the XB2 solution) were subtracted from the D(r)'s for XA1 and XB2. In the resulting difference curves (see the dotted curve in Fig. 5 for XA1) peaks occur at 2.25 Ä and 1.9 Ä. The first peak stems obviously from a ZnCl bond, which is slightly shorter than those in the higher complexes. The second peak is interpreted to correspond to ZnH 20 distances within the complexes including [ZnCl3]~. The ZnH 20 bonds within the complexes are shorter than those ( Ä) within the hydrated noncomplexed Zn 2+ ion, [Zn(H20)6] 2 +. This tendency is also observed in crystals. That is, a discrete tetrahedral complex [ZnCl3(H20)]~ has been found in crystals of KZnCl3 (H20)2, in which the ZnOH2 bond distance is 2.02 Ä [13]. The Zn(II) ion in crystal hydrates has a regular octahedral structure, and the Zn OH2 distances are in the range of Ä [14], There is no clear indication of any longer ZnCl bond, which would be expected, as is the case for octahedral [ZnI(H20)5] + complex in aqueous solution [3] and octahedral [ZnCl2(H20)4] complex in crystals of ZnCl2 11/3H20 [15], if an octahedral [Zn Cl (H20)5] + complex rather than a tetrahedral one were present. Since the Raman spectra can not be used to separate the concentrations of [ZnCl] + and [ZnCl2], no definite conclusion about the coordination in [ZnCl] + can be made. The fourcoordination may be retained in the [ZnCl] + complex, or its concentration may be too low to give sufficient contributions to the diffraction data. In Fig. 2 the observed s i (s) values are compared with those calculated for the atom pairs concerned using their parameter values. The agreement is good especially for XA solutions, in which the concentrations of the complexes are much higher than those in XB solutions, except at low s regions where various interactions including intermolecular interactions, which are not taken into account in the present calculations, significantly contribute to the s i(s) values.

8 70 M. Maeda et al. Structure of Zinc Chloride Complexes in Solution [1] G. Johansson, Advances in Inorganic Chemistry, ed. A. G. Sykes, Academic Press, San Diego 1992, vol. 39, pp [2] P. L. Goggin, G. Johansson, M. Maeda, and H. Wakita, Acta Chem. Scand. A38, 625 (1984). [3] H. Wakita, G. Johansson, M. Sandström, P. L. Goggin, and H. Ohtaki, J. Solution Chem. 20, 643 (1991). [4] G. Johansson, Acta Chem. Scand. 20, 553 (1966). [5] M. L. Delwaulle, Bull. Soc. Chim. France 1294 (1955). [6] D. F. C. Morris, E. L. Short, and D. N. Waters, J. Inorg. Nucl. Chem. 25, 975 (1963). [7] J. W. Macklin and R. A. Plane, Inorg. Chem. 4, 821 (1970). [8] D. E. Irish, B. McCaroll, and T. F. Young, J. Chem. Phys. 39, 3436 (1963). [9] G. Gali, S. Magazu, D. Majolino, P. Migliardo, M. C. BellissentFunel, F. Allota, and C. Vasi, Nuovo Cimento D 12, 197 (1990). [10] G. Johansson and M. Sandström, Chem. Scripta 4, 195 (1973). [11] D. T. Cromer, J. Chem. Phys. 50, 4857 (1969). [12] D. T. Cromer and J. B. Mann, J. Chem. Phys. 47, 1892 (1967). [13] P. Süsse and B. Brehler, Beiträge Mineral. Petrograp. 10, 132 (1964). [14] J. Burgess, Ions in Solution Basic Principles of Chemical Interactions, Ellis Horwood, Chichester 1988, Chapt. 3. [15] H. Follner and B. Brehler, Acta Crystallogr. B26, 1679 (1970).