Determination of Geometries of some complexes of Ni(II), Cu(II)

505 O N N S - N O M OH O OH N NH S NH2 M OH- -H2O M = Ni2+, Cu2+, VO2+,. . . 1 2 M(thsasal) anion Journal of Chemistry, Vol. 44 (4), P. 505 - 509, 2006 Determination of Geometries of some complexes of Ni(II), Cu(II) Received 5 July 2005 Vu Van Van, Nguyen Thanh Cuong, Le Kim Long, Lam Ngoc Thiem, Vu Dang Do Faculty of Chemistry, College of Natural Sciences, Vietnam National University of Hanoi Summary Calculations with 8 complexes of Cu(II) and Ni(II) with thiosemicarbazone bis(salicylaldehyde) were done within HyperChem 7.02 sotfware. The structure of complexes was determined. The UV-VIS spectra, single point of complexes were calculated and compared with experimental data. The calculated results have been confirmed by experimental structure of complexes. I - Introduction Template reactions play a very important role in the formation of many natural macro- heterocyclic compounds that have biological activities such as porphyrine, metalloenzyme etc. and become one type of the most important reactions in bioinorganic Chemistry and other relating fields. In the early 1970’s, some Russian chemists found a novel type of template reactions between thiosemicarbazone salicylaldehyde (1) and salicylaldehyde (2) and the metal ions such as Ni2+, Cu2+ (figure 1) [1]. It is necessary to note that the condensation of 1 and 2 can not occur without metal ions. Figure 1: Template reaction The products of these reactions were studied by methods of element analysis, conductivity measurement, and magnetic moment measure- ment ... but the data were relatively modest. It is experimentally studied the structure of the products by modern spectroscopic methods such as IR, UV-Vis, MS and NMR to examine the proposed mechanism of the reaction [3]. In this paper, some more data obtained about structure of such products by quantum chemistry calculations are given. It mainly deals with electronic spectra and geometry of the complexes formed in the template reaction (figure 2). II - Calculation Methods The calculations have been done within the 506 software HyperChem 7.02 using semi-empirical methods. Calculations, which involve molecules containning transition metals, are only done well with ZINDO/1 and ZINDO/S [4]. All molecules (3-10) are built, geometrically optimized by ZINDO/1 with convergence limit =10-5. All of algorithms have been tried. Only Conjugate Direction algorithm is able to give the good geometry but it requires quite long calculating time. The other algorithms get results fast but with not fine geometry. With the optimized geometry, the Single Point and IR Spectrum computation have been carried out by ZINDO/S. Configuration Interac- tion (CI)-Singlely Excited method is chosen to calculate the electronic spectrum with both two choices: Orbital criterion (the numbers of occupy orbital and unoccupy orbital) and energy criterion (energy excited). Since orbital-crite- rion takes longer time and is not stable [4], only Energy Criterion is used. The maximum chosen energy has been varied to find the suitable one. Complexes with R = Na, K, NH4 are electrolytically dissociated [1, 2], so they can be considered as M (thsasal) anion. However, when M are Ni (II), Cu (II) situation is quite different. III - Results and Discussion 1. Geometry of complexes The data of geometrically optimized complexes of Ni(II) and Cu(II) (figure 3) are listed in table 1. The Ni, Cu complexes usually have square planar or tetrahedron structure. The six torsion angles (table 1) in each complex have approximate zero value and almost unchange from this complex to other complex. This confirms the planar structures of complexes. The lengths of M-N and M-O bonds (M = Ni, Cu) are adequate in each complex (1-9 and 1-21 bonds, 1-10 and 1-13 bonds) and change inconsiderable in different complexes. The bond angles of N-M-O are also adequate in complexes but the O-M-O bond angle is greater than the N-M-N bond angle in each complex. This is due to the strain of the ring MNNCN. Both two-bond angles are almost the same in different complexes. These things show the good optimized result and the suitable optimization method. Perhaps, the structures of complexes have conjugate- electrons, so optimization Conjugated Direction algorithm gave the best result. M R Complex M R Complex Na 3 K 9 K 4 Cu CH3 10 NH4 5 H 6 CH3 7 Ni CH3CO 8 O N N S N O M R Figure 3: Optimized Geometry of 6 Figure 2: Complexes formed in template reaction of 1 and 2 507 Table 1: Bond length and angle of geometrically optimized complexes Atoms Ni(thsasal)- 6 7 8 Cu(thsasal)- 10 Bond length () 1-9 1.827 1.819 1.826 1.826 1.830 1.821 1-21 1.841 1.835 1.843 1.837 1.844 1.837 1-10 1.923 1.922 1.927 1.920 1.921 1.920 1-13 1.910 1.920 1.921 1.932 1.912 1.919 Bond angle, O 9-1-10 96.13 95.60 95.13 94.73 95.53 95.43 13-1-21 95.30 94.54 93.61 94.95 94.89 94.07 9-1-21 86.77 87.43 89.38 87.96 87.53 88.15 10-1-13 81.80 82.43 81.88 82.34 82.05 82.35 Torsion angle ×103, O 8-10-1-9 1.68 -1.08 1.98 -6.65 63.96 -2.04 6-9-1-10 0.96 1.52 -7.19 2.18 7.39 12.39 11-10-1-13 0.19 1.16 0.49 -18.22 40.89 0.02 12-13-1-10 -0.91 -1.76 5.02 26.64 -29.72 0.02 14-13-1-21 -4.68 0.60 3.49 23.72 56.86 9.98 16-21-1-13 12.52 7.56 1.29 -24.04 0.53 9.76 Charge densities of atoms H27 and H28 are listed in table 2 in order to find some relativity with the chemical shifs of these atoms. Only the 1H-NMR spectra of 4 - 8 were recorded. It is impossible to recorded the spectra of 9 and 10 because of magnetic field caused by the unpaired electron of Cu(II). Table 2: Charge (calculated) and Chemical shifts (experimental) on H27 and H28 H27 H28 Complexes Charge (e) , ppm Charge, e , ppm 4, 5 0.044 8.04 0.075 9.11 7 0.053 8.29 0.057 8.51 6 0.054 7.80 0.056 9.30 8 0.061 9.23 0.049 8.24 CH3CO are electrophile group while lonepair, CH3 and H are nucleophile groups with relative strength as following: lonepair > CH3 > H. The increase of charge on H27 from 4 to 8 is agree with the electron supplies of the groups bind to S atom. However, the increase of charge on H28 from 4 to 8 is strange. The order of chemical shifts is suitable with the charge on the atoms H27 and H28 respectively, except 6. Calculated chemical shift 1H-NMR spectra gave the results which suite with experimental spectra. So ZINDO/1 is a good method for optimization and calculations. The planar structure of complexes and the ZINDO optimized method are proved by the comparison between experimental UV-Vis spectra with calculated UV-Vis spectra and density charge of H atoms in the following part. 2. Electronic Spectra of complexes The electronic spectra of complexes were calculated with Criterion energy = 6 eV, 8 eV, 508 Ni NN O O gggyxgyzxz gggyxgz gggyxgxy EAbdeddIII BAbdadII AAbdbdI 1 1 1 1 1 1 1 1 11 2 1 1 1 12 ~)()(, ~)()( ~)()( 22 222 22       10 eV. However, energy of 6 eV is not enough for electrons in Ni(II) complexes transfer from ground state to excited state. Calculation with 10 eV gives complicated results. Therefore, we used Criterion Energy 8 eV spectral results to discuss. Figure 4: UV-Vis spectra of 6, experimental and calculated Table 3: Experimental UV-Vis spectra of Ni(II) complexes -  (nm) Complexes 1 2 3 4 5 6 7 3 235 301 340 370 476 4 237 301 340 368 475 5 238 302 341 367 472 6 238 302 340 368 471 7 241 301 325 397 485 8 241 315  382 Shoulders or bands overlapped by bands 4 527 Weak bands at ~530 nm Inside Ligand Transfer CT d-  gEgA 1 1 1  gBgA 2 1 1 1  gEgA 1 1 1  For each central ion (Ni(II), Cu(II)), the UV-Vis spectra of all the complexes are almost the same. Calculated spectra (Ni(II) and Cu(II) complexes) are similar to experimental spectra (figure 3). According to Ligand Field Theory and Group Theory, square planar complexes of Ni(II) exhibit three d-d bands in range 400 - 600 nm. These bands are geometrically forbidden, so that their intensities are very weak. Both calculated and experimental spectra of 3 - 8 bands have three bands in range 400 - 600 nm with weak intensities (almost coincide with the background). 509 Thus, 3 - 8 can be assigned to have square planar conformation, and their spectra can be assigned as in table 2. Table 4: Experimental UV-Vis spectra of Cu(II) complexes -  (nm) Complexes 1 2 3 4 5 9 227 252 312 348 430 10 239 287 315 344 448 Inside Ligand Transfer CT gg AB 1 2 1 2  The Ligand Field Theory and Group Theory also affirm that there is only one d-d band at about 400 - 600 nm in spectra of square planar Cu(II) complexes. This band lies far from other bands and are symmetric. In fact, there is a symmetrical band at about 430 - 450 nm in both experimental and calculated spectra of each of Cu(II) complexes. Thus, 9 and 10 are square planar. The atoms N10 and N13 are sp2 hybrided while the atom O9 and O21 are sp3 hybrided. The sp2 hybridation make the electronegative higher and the atom’s radius smaller than those in sp3 hybridation. This shows that these N atoms are almost similar to these O atoms. Thus, the four donor atoms are quite identical; and again is the square planar structure proved. IV - Conclusion From the calculated results and experimental results, conclusion can be made that the geometries of transition metal M(Ni,Cu) complexes with thiosemicarbazone are planar. The ZINDO/1 and ZINDO/s methods are suitable for transition metal complexes calculations. References 1. . .    , . . .   , 16, 4 (1971). 2. H. B. Gray, C. J. Ballhausen. J. Chem. Soc., 85, 260 - 265 (1963). 3. Vu Van Van. Vu Dang Do, Chu Dinh Kinh. Advanced in Natural Science, In press. 4. Hyperchem Release 7. Tools for molecular modeling. Copyright @2002 Hypercube Inc, USA. 5. Dang Ung Van, Vuong Minh Chau. Journal of Analytical Sciences (Vietnamese). Vol. 7, No. 1, 53 - 58 (2002).