C SPECTRA OF THE 1 , 1 , 3-TRIMETHYL-3-( 4-METHYLPHENYL ) BUTYL HYDROPEROXIDE IN VARIOUS SOLVENTS : MOLECULAR MODELING

GIAO-calculated NMR 13C chemical shifts as obtained at various computational levels are reported for the 1,1,3-trimethyl-3-(4-methylphenyl)butyl hydroperoxide. The data are compared with experimental solution data in chloroform-d, acetonitrile-d3, and DMSO-d6, focusing on the agreement with spectral patterns and spectral trends. Calculation of magnetic shielding tensors and chemical shifts for 13C nuclei of the 1,1,3-trimethyl-3-(4methylphenyl)butyl hydroperoxide molecule in the approximation of an isolated particle and considering the solvent influence in the framework of the continuum polarization model (PCM) was carried out. Comparative analysis of experimental and computer NMR spectroscopy results revealed that the GIAO method with MP2/6-31G(d,p) level of theory and the PCM approach can be used to estimate the NMR 13C chemical shifts of the 1,1,3-trimethyl-3-(4methylphenyl)butyl hydroperoxide.


Introduction
Arylalkyl hydroperoxides are useful starting reagents in the synthesis of surface-active peroxide initiators for the preparation of polymeric colloidal systems with improved stability [5].Thermolysis of arylalkyl hydroperoxides was studied in acetonitrile [11].NMR 1 Н spectroscopy has been already used successfully for the experimental evidence of the complex formation between a 1,1,3trimethyl-3-(4-methylphenyl)butyl hydroperoxide and tetraalkylammonium bromides in acetonitrile [2; 9; 12] and chloroform solution [9].The aim of this work is a comprehensive study of the 1,1,3trimethyl-3-(4-methylphenyl)butyl hydroperoxide (ROOH) by experimental NMR 13 C spectroscopy and molecular modeling methods.
The magnetic shielding tensors (c, ppm) for 13 С nuclei of the hydroperoxide and the reference molecule were calculated with the МР2/6-31G(d,p) and B3LYP/6-31G(d,p) equilibrium geometries by standard GIAO (Gauge-Independent Atomic Orbital) approach [13].The calculated magnetic isotropic shielding tensors,  i , were transformed to chemical shifts relative to TMS molecule,  i , by where both,  ref and  i , were taken from calculations at the same computational level.Table 1 illustrates c values for TMS molecule used for the hydroperoxide 13 C nuclei chemical shifts calculations.
Inspecting the overall agreement between experimental and theoretical spectra RMS errors () were used to consider the quality of the 13 С nuclei chemical shifts calculations.Correlation coefficients (R) were calculated to estimate the agreement between spectral patterns and trends.
Ten signals for the hydroperoxide carbon atoms are observed in the ROOH 13 C NMR spectrum.Signal of the carbon atom bonded with a hydroperoxide group shifts slightly to the str onger fields with the solvent polar ity increasing, while the remaining signals are shifted to weak fields.A linear dependences between the 13 C chemical shifts values of the hydroperoxide are observed in the studied solvents (Fig. 1).This is consistent with the authors [     Note.Experimental values are those for tert-butyl hydroperoxide from [13].
classes.Equations corresponded to the obtained relationships (Fig. 1) are listed below.

Molecular modeling of the 1,1,3t r i m e t h y l -3 -( 4 -m e t h y l p h e n y l ) b u t y l hydroperoxide NMR 13 С spectra by МР2 and B3LYP methods
The hydroperoxide molecule geometry optimization in the framework of МР2/6-31G(d,p) and B3LYP/6-31G(d,p) methods was carried out as the first step of the hydroperoxide NMR 13 С spectra modeling.Initial hydroperoxide configuration chosen for calculations was the one obtained by semiempirical AM1 method and used recently for the hydroperoxide O-O bond homolysis [11] as well as complexation with Et 4 NBr [2; 10] modeling.The main parameters of the hydroperoxide fragment molecular geometry obtained in the isolated particle approximation within the framework of MP2/6-31G(d,p) (Fig. 2) and B3LYP/6-31G(d,p) levels of theory are presented in Table 2. Peroxide bond is a reaction centre in this type of chemical initiators.Thus, the main attention was focused on the geometry of -CO-OH fragment.The calculation results were compared with known experimental values for the tert-butyl hydroperoxide [6], and appropriate agreement between calculated and experimental parameters can be seen in the case of МР2/6-31G(d,p) method.
Calculation of 13 C chemical shifts of the hydroperoxide was carried out by GIAO method in the approximation of an isolated particle as well as in studied solvents within the PCM model, which takes into account the nonspecific solvation.Equilibrium hydroperoxide geometries obtained in the framework of MP2/6-31G(d,p) and B3LYP/ 6-31G(d,p) levels of theory for the isolated particle approximation were used in all cases.
The chemical shift values (, ppm) for 13 C nuclei in the hydroperoxide molecule were evaluated on the base of calculated magnetic shielding constants (, ppm).TMS was used as standard, for which the molecular geometry optimization and  calculation were performed using the same level of theory and basis set.Values of the 13 C chemical shifts were found as the difference of the magnetic shielding tensors of the corresponding TMS and hydroperoxide nuclei (Tables 3 and 4).
The correct spectr al pattern for the hydroperoxide NMR 13 C spectrum was obtained for all methods and basis sets used within the isolated molecule approximation (see Table 3) as well as solvation accounting (see Table 4).Exceptions are aromatic C8 and C9 carbons, which signals are interchanged for all calculations.
The best reproduced experimental chemical shift value for the carbon atom of the CO-OH group is observed in the case of MP2/6-31G(d,p) approximation in all used solvents whereas B3LYP with the same basis set gives slightly worse values.
Basis set extension to 6-311++G(d,p) leads to a deterioration of the calculation results.Calculated value for the carbon of CO-OH group (83.61 ppm) within the isolated molecule approximation is closest to experimental one in acetonitrile (83.74 ppm).When passing to the calculations in the PCM mode solvation accounting leads to more correct results for the MP2 and B3LYP methods.The lowest  values for all solvents are obtained with 6-31G(d,p) basis set.Linear relationships between the experimental NMR 13 C chemical shifts and the calculated values  calc for the hydroperoxide 13 C nuclei (see Fig. 3) have been obtained for both methods and all basis sets.The correlation coefficients (R) corresponding to obtained dependences are shown in Table 4. Joint account of  and R values indicates possibility of MP2 method with 6-31G(d,p) basis set using for the calculation of the hydroperoxide 13 C nuclei chemical shifts.

Conclusions
A comprehensive study of the 1,1,3-trimethyl-3-(4-methyl-phenyl)butyl hydroperoxide by experimental NMR 13 C spectroscopy and molecular modeling methods was performed.A comparative assessment of the 13 C nuclei chemical shifts calculated by GIAO in various approximations.For NMR 13 C spectra of the hydroperoxide in different solvents MP2 and B3LYP methods approximations with 6-31G(d,p), 6-311G(d,p), and 6-311++G(d,p) basis sets allow to obtain the correct spectral pattern.A linear correlations between the calculated and experimental values of the 13 C chemical shifts for the studied hydroperoxide molecule were obtained for all solvents studied.In all cases, the MP method combined with 6-31G(d,p) basis set allows to get a better agreement between the calculated and experimental data as compared to the B3LYP results.
1], who showed linear correlation between the chemical shifts values in chloroform-d and dimethylsulphoxide-d 6 for a large number of organic compounds of different 20