- Email: [email protected]
Influence of conformation on the electronic structure of thiacalixarenes according to DFT calculations and X-ray emission spectroscopy
Journal of Molecular Structure 1006 (2011) 502–507
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Structure journal homepage: www.elsevier.com/locate/molstruc
Influence of conformation on the electronic structure of thiacalixarenes according to DFT calculations and X-ray emission spectroscopy Gennady A. Kostin a,b,⇑, Natalya A. Kryuchkova a,b, Lev N. Mazalov a,b, Vladislav G. Torgov a, Andrej B. Drapaylo c a b c
Nikolaev Institute of Inorganic Chemistry, Siberian Branch, Russian Academy of Sciences, Av. Lavrentyeva, 3, 630090 Novosibirsk, Russia Novosibirsk State University, Pirogova Str., 2, 630090 Novosibirsk, Russia Institute of Organic Chemistry, National Academy of Sciences, Murmanskaya Str., 5, 02660 Kiev, Ukraine
a r t i c l e
i n f o
Article history: Received 5 August 2011 Received in revised form 29 September 2011 Accepted 29 September 2011 Available online 8 October 2011 Keywords: Thiacalixarene DFT calculation X-ray emission spectroscopy Electrophilic substitution
a b s t r a c t The DFT calculations of different thiacalix[4]arenes show that the conformation and geometry of arene core mainly determine the difference in electronic density distribution (frontier occupied MO and ESP charges) between ‘‘cone’’ and ‘‘1,3-alternate’’ conformation. The key factor is orientation of benzene rings relative to average plane through bridging sulfur atoms. Calculated structure of frontier orbitals was verified by experimental sulfur Kb spectra, calculated spectra reproducing main features of experimental ones. Difference in HOMO and ESP distribution on carbon atoms in benzene rings for ‘‘cone’’ and ‘‘1,3alternate’’ conformers correlates with earlier reported data on regioselectivity of electrophilic substitution in these compounds. Ó 2011 Elsevier B.V. All rights reserved.
1. Introduction Thiacalixarenes (TCAs) is a relatively new class of macrocyclic compounds, which were actively studied in recent years. The replacement of the bridging CH2 groups by electron-withdrawing sulfur atoms leads to a substantial difference between TCAs and classic calixarenes [1–3]. Several features of TCA such as increased acidity of OH groups or the possibility of directly replacing them by amino and amide groups [1,4] can be qualitatively explained by the inductive effects of the more electronegative sulfur atoms. The unusual regioselectivity of electrophilic substitution in TCA (Scheme 1) and its dependence on the conformation reported in Lhotak’s recent papers [5–8] require a quantitative description of the highest occupied molecular orbitals and the charge distribution on the carbon atoms in the benzene rings. From the viewpoint of coordination chemistry, the major distinction of TCA is the presence of sulfur atoms as additional donor centers. The coordination chemistry of classic calixarenes without additional donor groups is limited to phenoxo and p-complexes. The sulfur atoms of TCA can participate in complexation either together with OH groups or as single donor atoms, especially for chalcophile metals. The possibility of their participation in com⇑ Corresponding author at: Nikolaev Institute of Inorganic Chemistry, Siberian Branch, Russian Academy of Sciences, av. Lavrentyeva, 3, 630090 Novosibirsk, Russia. E-mail address: [email protected] (G.A. Kostin). 0022-2860/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.molstruc.2011.09.058
plexation is determined by the geometrical characteristics of thiacalixarenes, the electron density on the sulfur atoms, and the contribution of the valence electrons of sulfur to the highest occupied molecular orbitals. If TCA is modified by adding donor groups to the upper or lower rim, the electron interference between the groups should also be taken into consideration. The participation of the valence S 3p AO in the highest occupied molecular orbitals can be experimentally proven by X-ray emission spectroscopy (3p valence-to-1s-core XES, Kb-line). The XES spectra of TCAs obtained earlier [9] revealed a systematic difference between TCAs in the ‘‘cone’’ and ‘‘1,3-alternate’’ conformations, which was evidence of different sulfur contributions to HOMO. The experimental studies should be complemented with a profound theoretical description of the electronic structure of TCA. This study is focused on the interrelationships between the electronic structures of TCAs, their reactivities in electrophilic substitution reactions, and X-ray spectra. Previous quantum chemical calculations of TCAs were mainly focused on the energy aspects of complexation with different metals [10–13]. A qualitative description of HOMO for TCA in the ‘‘cone’’ and ‘‘1,3-alternate’’ conformations was given in [8,10].
2. Experimental Thiacalix[4]arenes 1–6 (Scheme 2) were synthesised as described in [14–18]. Model compounds 7, 8 were not synthesised,
G.A. Kostin et al. / Journal of Molecular Structure 1006 (2011) 502–507
503
structure of 8 is ‘‘pinched cone’’ with C2v symmetry common to TCA in the ‘‘cone’’ conformation tetraalkylated in the lower rim [14,25]. 3.1. Molecular orbitals and XES spectra
Scheme 1.
but DFT calculations were performed to compare the results with those for 1 and 5.The S Kb X-ray fluorescence emission spectra of 1–6 were recorded on a ‘‘Stearate’’ X-ray ultrasoft spectrometer. The X-ray tube operating conditions were V = 8 kV, I = 100 mA. A 0Þ quartz crystal was used as a crystal analyzer. The comð1 3 4 pounds were deposited on the secondary anode of the X-ray tube. The spectra were measured at the liquid nitrogen temperature. The spectrometer resolution in the studied region was 5 103. The electronic structures of synthesised thiacalix[4]arenes in the ‘‘cone’’ (1–4) and ‘‘1,3-alternate’’ (5, 6) conformations and of model compounds (7, 8) were calculated using the Jaguar 6.5 program, Shrodinger Inc. [19]. The geometry of thiacalix[4]arene molecules was optimized by the DFT technique with the B3LYP hybrid exchange-correlation functional [20,21]. The M6-31G(TM)+ basis set [22] was used for all atoms. 3. Results and discussion Table 1 shows the selected interatomic distances, interatomic angles, and interplanar angles for the optimized structures of the molecules. The calculated average bond lengths (S–C, C–O) and C–S–C bond angles are close for all structures and are in reasonable agreement with the experimental data for the crystal phases (the experimental parameters from the original papers or calculated from the corresponding CIF files in CCDC are given in parentheses). For all ‘‘1,3-alternate’’ structures 5–7, the angles between the average plane through four sulfur bridging atoms [S4] and the planes of the benzene rings are approximately 90°, the structure of the arene core being of approximately D2d symmetry. In ‘‘cone’’ conformers 1–4 with OH groups in the lower rim, the calculated interplanar angles vary from 26° to 70°. The calculated structure of 1 has approximately C4v symmetry, the interplanar angle (51.3–52.8°) being the average between the two experimental values corresponding to the ‘‘pinched cone’’ structure determined by single crystal XRD [14]. Due to the presence of substituents in the upper rim of 2–4, the symmetry is lowered from C4v to the ‘‘pinched cone’’ conformation. As thiacalixarenes tend to include small molecules in the inner cavity, the experimental interplanar angles for 2 depend on the nature of the included molecule [23,24]. The calculated values shown in Table 1 are closest to those reported for the structure of a pyridine inclusion compound [23]. The optimized
Scheme 2.
The molecular orbital diagram for thiacalix[4]arene 5 in the ‘‘1,3-alternate’’ conformation is shown along with the typical MO schemes in Fig. 1. Four highest occupied orbitals (from HOMO to HOMO-3) are mainly formed from the 3p orbitals of the sulfur atoms and involve the p-bonding orbitals of the benzene rings, the total contribution of all sulfur AOs being 40–70%. The orbitals from HOMO-4 to HOMO-7 are mainly distributed on the carbon atoms of the benzene rings (65–75%) with some contribution from the oxygen and sulfur atomic orbitals. For orbitals from HOMO to HOMO-4, the electron density distribution on the carbon atoms is determined by the «2 + 2» p-bonding orbitals of the benzene ring (Scheme 3). The electron density determined by HOMO-3, 7, and 8 is mainly located on the carbon atoms bonded to the OR group (C1) and in the para-position (C4), the carbon 2p AO forming the «3 + 3» p-bonding orbitals of the benzene ring (Scheme 3). For the doubly degenerate orbitals (HOMO-1, 2 and HOMO-5, 6), one pair of opposite benzene rings have the same electron density distribution as in the ‘‘2 + 2’’ orbital and the other as in the ‘‘3 + 3.’’ The 2p orbitals of the oxygen atoms participate in the formation of molecular orbitals starting from HOMO-9 to lower-lying ones. The total contribution of the atomic orbitals of all four oxygen atoms to the highest occupied molecules orbitals does not exceed 13%. For ‘‘1,3-alternate’’ 7 with OH instead of methoxy groups, the structure of the molecular orbitals changes but slightly. The orbitals that are close in energy are inverted (HOMO-4 of b1 symmetry and HOMO-5, -6 of e symmetry), but the contributions of atomic orbitals to different HOMOs coincide well with the results obtained for 5. The highest occupied molecular orbital of 1 (Fig. 2), having b2 symmetry, corresponds to the linear combination of the ‘‘3 + 3’’ p-bonding orbitals of the benzene rings (70%) and the 2p atomic orbitals of the oxygen atoms (27%). Although the oxygen 2p orbitals in HOMO are perpendicular to the benzene plane, they have different phase signs, thus forbidding the p-p bonding. The contribution of the sulfur AO to HOMO is negligible. The doubly degenerate HOMO-1 and HOMO-2 have similar structures, each orbital being distributed on only two of the four benzene rings. The HOMO-3, HOMO-5, and HOMO-6 are the linear combinations of mainly 3p sulfur atomic orbitals, the total contribution of all sulfur AOs being 70–80%. The occupied molecular orbitals that are lower in energy are formed by the «2 + 2» p-bonding orbitals of the benzene rings with a contribution from the 3p (3s) sulfur orbitals, the total contribution of the sulfur AOs varying from 21 to 42%. The positive overlap of the S3p orbitals with the p-bonding orbitals of the benzene ring in HOMO-7 and HOMO-8 leads to additional S–C pbonding. It is interesting to note that for thiacalixarene 8 in the ‘‘pinched cone’’ conformation, molecular orbitals from HOMO to HOMO-3 are similar in shape to the orbitals of TCA in the ‘‘1,3alternate’’ conformation (Fig. 1) and mainly formed from the sulfur atomic orbitals (50–60%). The doubly degenerate HOMO-4 and -5 in TCA 8 are primarily localized on the benzene rings orientated outwardly relative to the cavity of TCA, the electron density being maximum on the C1 and C4 carbon atoms. As in the case of 5, in TCA 8 the oxygen atomic orbitals make a considerable contribution (20–55%) only to HOMO-9 and lower-energy orbitals (HOMO-10 to -12). The optimized geometry of phosphorylated TCA 4 differs substantially from C4v symmetry, and this is also responsible for the less symmetric character of HOMO (Fig. 3). The structure of the occupied orbitals in 4, however, is similar that in 1 (‘‘cone’’). The highest occupied molecular orbital is a linear combination of
504
G.A. Kostin et al. / Journal of Molecular Structure 1006 (2011) 502–507
Table 1 Geometrical parameters of the optimized structures for the compounds. The experimental single crystal XRD data [14,18,23] are shown in parentheses. L
d(S–C), Å
d(O–C), Å
1 2 3 4
1.80 (1.783) 1.81 (1.780) 1.81 1.86
1.36 (1.363) 1.36 (1.366) 1.38 1.38
5 6 7 8
1.80 (1.772) 1.80 1.85 1.80
1.37 (1.379) 1.37 1.40 1.37
d(P–O), Å
1.624 1.730 1.597
\C–S–C, °
[Cn]–[S4] interplanar angles
104.0 (102.7) 104.4 (101.29) 104.0 104.2
52.8, 55.5, 38.7, 26.7,
51.3, 49.4, 50.9, 59.5,
52.8, 54.8, 41.9, 31.3,
51.3 (62.2, 43.1) 47.8 69.5 70.1
103.2 (102.8) 104.0 102.2 101.2
86.7, 88.6, 88.9, 31.2,
87.1, 88.0, 88.6, 84.9,
86.7, 87.4, 88.9, 31.1,
87.1 (85.1, 86.3) 83.9 88.6 84.3
Fig. 1. MO diagrams shown energy levels of HOMO for thiacalixarenes 5, 7, 8 and selected orbitals for 5.
Fig. 2. MO diagram shown energy levels of HOMO and selected orbitals for thiacalixarene 1.
Scheme 3. Designation of carbon atoms in benzene ring and signs of molecular orbital corresponding to two orbital types.
the «3 + 3» type p-bonding orbital of the benzene ring, the sulfur atomic orbitals (17%), and the atomic orbitals of the oxygen atoms in the lower rim (20%). The HOMOs-1, -4, and -5 are mainly formed by the sulfur AO (50–60%). The atomic orbitals of the oxygen atoms of the phosphoryl groups are involved in the formation of HOMO starting from HOMO-7. The molecular orbitals with energies from 7.44 to 7.78 eV (HOMO-12–HOMO-19) correspond to the lone electron pairs of the oxygen atoms double-bonded to phosphorus. The lower-lying orbitals are mainly formed from the lone electron pairs of the oxygen atoms of the ethoxy groups bonded to phosphorus. The experimental Kb-spectra of sulfur indirectly confirm the structure of the highest occupied orbitals in the studied thiacalixarenes. All the studied compounds have two bands at binding energies of 2466 (A) and 2469 (B) eV in the experimental spectra, maximum B being more intense for all TCAs in the ‘‘cone’’ conformation (Fig. 4). The S(Kb) valence-to-core spectra correspond to the electron transition from the occupied MO including the sulfur 3pAO to the S 1s core level. The intensities of the Kb transitions are proportional to the squares of the coefficients c2i of the sulfur 3p AO in different occupied molecular orbitals. An analysis of the
Fig. 3. MO diagram shown energy levels of HOMO and selected orbitals for phosphorylated thiacalixarene 4.
valence-to-core transitions thus enables us to estimate the contribution of the sulfur 3p AO to the occupied molecular orbitals. The calculated spectral transitions were convolved by 1.5 eV full width at half maximum to model the experimental broadening (Fig. 5, green line). A less line broadening (0.5 eV) was chosen to
505
G.A. Kostin et al. / Journal of Molecular Structure 1006 (2011) 502–507
7.41 eV (the range is twice wider) for the ‘‘1,3-alternate’’ conformer. Both factors result in a broadening of maximum B and lowering of intensity. 3.2. Reactivity of carbon atoms in the benzene rings
Fig. 4. Experimental sulfur Ka spectra for investigated compounds.
Two approaches can be used for describing the reactivity. The first approach is based on the calculation of relative energies for the activation complex formed after the attack at different positions, but the results of calculation strongly depend on the nature of the attacking particle, the geometry of the intermediate state, and the details of calculations. Earlier, Lhotak reported on similar calculations for the electrophilic substitution in ‘‘1,3-alternate’’ thiacalixarene, but the results obtained for different attacking particles and different calculation techniques agreed with the experimental selectivity only in several instances [8]. The second approach is based on the calculation of molecular descriptors, which are formally independent of the nature of the attacking reagent and the geometry of the activated complex but define only the electron density distribution in the TCA molecule. According to perturbation theory, the first two terms of the series DEtotal ¼ Eel þ Eorb correspond to the electrostatic interactions of molecules and interactions that are due to the overlap of their molecular orbitals [26,27]. With electrostatic regioselectivity control (charge-controlled reactions), the electrophilic agent with high charge density and low polarizability (for example, H+) is expected to attack the ring sites with a maximum negative charge. The direction of electrophilic substitution will thus be determined by the largest negative values of the electrostatic potential (ESP) and the R qðr0 Þ 0 P Zi corresponding atomic charges [26–28]: UðrÞ ¼ M dr . i¼1 jrRi j jrr 0 j
LUMO 0, the overlap between the If the difference EHOMO donor Eacc donor and acceptor molecular orbitals is dominant in energy (the second term of DEtotal). The regioselectivity of similar frontiercontrolled reactions can be determined in terms of the Fukui functions [29–32]. The Fukui function for electrophilic substitution 2 P i ðrÞj is defined by fN ðrÞ ¼ j/N ðrÞj2 þ Ni¼1 ð@j/@N Þ v ðrÞ .
The second term of this equation is responsible for orbital relaxation, the change in the orbital shape accompanying the removal of electrons from the system. When this term is small compared to the first term, it can be neglected and the function is determined by only the frontier highest occupied molecular orbital j/N ðrÞj2 . Table 2 DFT calculated ESP charges and atomic contributions to HOMO for 1, 5, 7, and 8 (the maximum negative charges and the largest contributions of carbon atoms are bold). Atom
Atomic charges C1 0.29 C2 0.00 C3 0.08 C4 0.24 C5 0.08 C6 0.03 O 0.46 S 0.12
Fig. 5. Experimental (bold line) and calculated Ka – spectra for thiacalixarenes 1, 5, 7, 8. Calculated spectra are shown with 1.5 eV (green line) and 0.5 eV (red dotted line) broadening. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
determine the fine structure of the calculated spectra (Fig. 5, red line). A comparison of the experimental and calculated spectra for 1 and 5 shows that maximum B is related to the transitions from the MO involving the lone electron pairs of the sulfur 3p AO, while maximum A is due to the sulfur 3p orbitals involved in the formation of the S–C r-bonds. The decrease in short-wave maximum B for TCA in the ‘‘1,3-alternate’’ conformation is determined by two factors. Firstly, for ‘‘1,3-alternate’’ TCA, the contribution of the 3p sulfur atomic orbitals in different HOMOs is more regular. Secondly, maximum B is determined by the transitions from the molecular orbitals with energies from –6.83 to 7.55 eV (the range is 0.7 eV) for the ‘‘cone’’ conformer and from 5.98 to
1
Contributions to HOMO-0 C1 0.04 C2 0.03 C3 0.01 C4 0.05 C5 0.01 C6 0.03 O 0.07 S 0.01 a b
5
7
8 a
b
0.09 0.37 0.30 0.12 0.30 0.37 0.30 0.27
0.09 0.26 0.23 0.17 0.23 0.26 0.56 0.32
0.44 0.06 0.18 0.06 0.19 0.01 0.21 0.22
0.13 0.18 0.13 0.23 0.15 0.20 0.28
0.01 0.01 0.03 0.00 0.03 0.01 0.00 0.15
0.01 0.02 0.03 0.00 0.03 0.02 0.00 0.15
0.01 0.02 0.03 0.00 0.04 0.01 0.01 0.15
0.01 0.01 0.03 0.00 0.03 0.01 0.00
For the benzene rings lying at an angle of 84–85° to the [S4] plane. For the benzene rings lying at an angle of 31° to the [S4] plane.
506
G.A. Kostin et al. / Journal of Molecular Structure 1006 (2011) 502–507
We analyzed the ESP charges and electron density distributions related to the highest occupied molecular orbital for TCA 1, 5, 7, and 8. The atomic contributions to HOMO (Table 2) were calculated as Pn 2 i¼1 c i , where n is the number of atomic basis functions, and ci is the coefficient for the ith basic function in the HOMO. The results were averaged over four rings for 1, 5, and 7 and over pairs of opposite rings for 8. The differences do not exceed 0.02 for charges and are less than 0.005 for the atomic contributions to HOMO. An analysis of the data of Table 2 indicates that for TCA 1 in the ‘‘cone’’ conformation, both descriptors are favorable for para-substitution; the maximum negative charge and maximum density of HOMO are localized on the carbon atoms in the para-position relative to the OH groups, which agrees with the strong para-orientation of the hydroxo groups. The charge distribution and atomic contributions to HOMO in thiacalixarenes 5 and 7 in the ‘‘1,3-alternate’’ conformation differ considerably from those for 1. In both alternates, the maximum negative charge and maximum density of HOMO are located on the C3 and C5 carbon atoms in the meta-position to the oxygen-containing groups. This effect is more pronounced for 5; in 7, the charge distribution between the C3–C5 atoms is more regular. The most interesting result was obtained for model TCA 8 in the ‘‘pinched cone’’ conformation with C2v symmetry. Two opposite benzene rings in it are almost parallel to each other and perpendicular to the [S4] plane; the other two rings are pointer outside the cavity of TCA, the interplanar angle being 118° (Table 1). The distribution of HOMO here is almost the same for all the four benzene rings and coincides with the distribution found for both TCAs in the ‘‘1,3-alternate’’ conformation. The ESP charge distribution in the benzene rings perpendicular to the [S4] plane in 8 also qualitatively coincides with that in the alternates, the maximum negative charge lying on the C3 and C5 atoms. In the other two rings, however, the maximum negative charge lies on the C4 atom in the para-position relative to the OR group, as it does in TCA 1 in the ‘‘cone’’ conformation. The orientation of the benzene rings with respect to the [S4] plane thus plays the key role in the distribution of ESP charges on the C3–C5 atoms, though alkylation of OH groups also results in a certain increase in the total electron density on the C3 and C5 atoms (cf. 1 and 82 or 7 and 5). Clearly, the calculated descriptors do not take into consideration such factors as steric accessibility of reaction centers, possible solvation effects in condensed phases, etc. Nevertheless, the regularities obtained are in complete agreement with both new experimental data on regioselectivity obtained by Lhotak and numerous data found earlier for the electrophilic substitution in 1. Thiacalixarene 1 was the starting compound for different modifications, and electrophilic substitution occurred at the paraposition in all cases [2,3], as could be predicted from both HOMO and charge distributions (Table 2). In 1,3-alternates, however, both factors are favorable for electrophilic substitution at the meta-position. Lhotak discovered that nitration or formylation (both Duff and Gross conditions) of similar TCAs resulted in only meta-substituted products [5,8]. For tetraalkylated thiacalixarene 8 in the ‘‘pinched cone’’ conformation, the situation is more complex. For frontier-controlled reactions favored by HOMO distribution, meta-substitution should be expected according to Table 2, while the ESP charge distribution favors meta-substitution in the arene rings perpendicular to the [S4] plane and para-substitution in the other two rings. Depending on the conditions of formylation for similar TCA with OPr groups in the lower rim, Lhotak described the para-substituted products predominant for the Duff reaction (51% yield vs 15% for meta-substituted product) [6] and only meta-substituted products for the Gross reaction [7]. Studies of the transformations of the CHO group in the meta-substituted products of the Gross reaction showed that the –CHO and –CN groups in the meta-position of the benzene ring shifted the
equilibrium between the two ‘‘pinched cone’’ forms in such way that the electron-withdrawing meta substituents were located in the vertical benzene rings. In contrast, the reduction of the formyl group to CH2OH or CH3 led to a conformer in which the electron donor meta substituent was in the rings rotated out of the cavity of TCA. Both facts correlate well with the ESP charge distribution on the C3 and C5 atoms in thiacalixarene 8. In the rings rotated out of the cavity, the negative ESP charges on these atoms are lower, thus favoring the electron donor groups. 4. Conclusions The DFT calculations of thiacalixarenes in the ‘‘cone’’, ‘‘pinched cone,’’ and ‘‘1,3-alternate’’ confirmations revealed that the orientation of the benzene rings relative to the mean [S4] plane through four bridging sulfur atoms played a critical role in electron density distribution (the highest occupied molecular orbital and ESP charges). In the benzene rings that form angles of approximately 90° (84–89°) with the [S4] plane, the carbon atoms in the meta positions to the phenol group are the most favorable sites for an electrophilic attack. This result could help to understand the unusual regioselectivity found by Lhotak for tetraalkylated thiacalix[4]arenes in the ‘‘pinched cone’’ or ‘‘1,3-alternate’’ conformation. In the benzene rings lying at an angle of 30–50° to the [S4] plane of tetrahydroxo thiacalix[4]arenes in the ‘‘cone’’ or ‘‘pinched cone’’ conformation, the maximum negative ESP charge and the maximum density of HOMO are located on the carbon atom in the para position to the phenol oxygen. The calculated difference in the structure of the highest occupied molecular orbitals between the ‘‘1,3-alternate’’ and ‘‘cone’’ conformations was experimentally confirmed by the sulfur Kb spectra of thiacalix[4]arenes. The experimentally observed decrease in the relative intensity of the short-wave band (2469 eV) for the alternate conformers was explained by the more regular distribution of the sulfur 3p AO in the HOMO and the wider range of energies of the molecular orbitals that formed this maximum. References [1] N. Iki, S. Miyano, J. Incl. Phenom. 41 (2001) 99–105. [2] E.A. Shokova, V.V. Shokova, Russ. J. Org. Chem. 39 (2003) 1–28. [3] N. Morohashi, F. Narumi, N. Iki, T. Hattori, S. Miyano, Chem. Rev. 106 (2006) 5291–5316. [4] P. Rao, M.W. Hosseini, A.D. Cian, J. Fischer, J. Chem. Soc., Chem. Commun. (1999) 2169–2170. [5] O. Kundrat, J. Kroupa, S. Bohm, J. Budka, V. Eigner, P. Lhotak, J. Org. Chem. (2010) 8372–8375. [6] O. Kundrat, H. Dvorakova, I. Cisarova, M. Pojarova, P. Lhotak, Org. Lett. 11 (2009) 4188–4191. [7] O. Kundrat, H. Dvorakova, V. Eigner, P. Lhotak, J. Org. Chem. 75 (2010) 407– 411. [8] O. Kundrat, I. Cisarova, S. Bohm, M. Pojarova, P. Lhotak, J. Org. Chem. 74 (2009) 4592–4596. [9] V.G. Torgov, L.N. Masalov, G.A. Kostin, T.V. Us, T.M. Korda, N.A. Kryuchkova, E.A. Korotaev, A.D. Fedorenko, A.B. Drapailo, J. Struct. Chem. 52 (4) (2011) 738– 745. [10] M.-L. Fu, N.L. Rangel, R.D. Adams, J.M. Seminario, J. Clust. Sci. 21 (2010) 867– 878. [11] A. Suwattanamala, A.L. Magalhaes, J.A.N.F. Gomes, Theor. Chem. Acc. 117 (2007) 431–440. [12] A. Suwattanamala, A.L. Magalhaes, J.A.N.F. Gomes, J. Phys. Chem. A 109 (2005) 10742–10752. [13] L.I. Shamova, G.A. Shamov, I.S. Antipin, A.I. Konovalov, J. Phys. Chem. A 113 (2009) 5691–5699. [14] H. Akdas, L. Bringel, E. Graf, M.W. Hosseini, G. Mislin, J. Pancanel, A.D. Cian, J. Fischer, Tetrahedron Lett. 39 (1998) 2311–2314. [15] O. Kasyan, D. Swierczynski, A. Drapailo, K. Suwinska, J. Lipkowski, V. Kalchenko, Tetrahedron Lett. 44 (2003) 7167–7170. [16] H. Kumagai, M. Hasegawa, S. Miyanari, Y. Sugawa, Y. Sato, T. Hori, S. Ueda, H. Kamiyama, S. Miyano, Tetrahedron Lett. 38 (1997) 3971–3972. [17] J. Lang, H. Dvorakova, I. Bartosova, I. Stibor, R. Hrabal, Tetrahedron Lett. 40 (1999) 373–376. [18] P. Lhoták, M. Himl, I. Stibor, J. Sykora, H. Dvorakova, J. Lang, H. Petrı´ckova, Tetrahedron 59 (2003) 7581–7585.
G.A. Kostin et al. / Journal of Molecular Structure 1006 (2011) 502–507 [19] [20] [21] [22] [23]
Jaguar 7.5, Shrodinger Inc. – Portland OR. – (2008). A. Becke, J. Chem. Phys. 98 (1993) 5648–5653. C. Lee, W. Yang, R.G. Parr, Phys. Rev. B. 37 (1988) 785–789. A.V. Mitin, J. Baker, P. Pulay, J. Chem. Phys. 118 (2003) 7775–7783. A. Bilyk, A.K. Hall, J.M. Harrowfield, M.W. Hosseini, B.W. Skelton, A.H. White, Inorg. Chem. 40 (2001) 672. [24] N. Iki, C. Kabuto, T. Fukushima, H. Kumagai, H. Takeya, S. Miyanari, T. Miyashi, S. Miyano, Tetrahedron 56 (2000) 1437. [25] M. Himla, M. Pojarova, I. Stibora, J. Sykorab, P. Lhoták, Tetrahedron Lett. 46 (2005) 461–464.
[26] [27] [28] [29]
507
G. Klopman, J. Am. Chem. Soc. 90 (1968) 223–233. L.J. Bartolotti, P.W. Ayers, J. Phys. Chem. A 109 (2005) 1146–1151. T. Brinck, J. Phys. Chem. A 101 (1997) 3408–3415. F. Gilardoni, J. Weber, H. Chermette, T.R. Ward, J. Phys. Chem. A 102 (1998) 3607–3613. [30] C. Cardenas, N. Rabi, P.W. Ayers, C. Morell, P. Jaramillo, P. Fuentealba, J. Phys. Chem. A 113 (2009) 8660–8667. [31] R.K. Roy, J. Phys. Chem. A 107 (2003) 397–404. [32] T. Mineva, V. Parvanov, I. Petrov, N. Neshev, N.J. Russo, Phys. Chem. A 105 (2001) 1959–1967.
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Structure journal homepage: www.elsevier.com/locate/molstruc
Influence of conformation on the electronic structure of thiacalixarenes according to DFT calculations and X-ray emission spectroscopy Gennady A. Kostin a,b,⇑, Natalya A. Kryuchkova a,b, Lev N. Mazalov a,b, Vladislav G. Torgov a, Andrej B. Drapaylo c a b c
Nikolaev Institute of Inorganic Chemistry, Siberian Branch, Russian Academy of Sciences, Av. Lavrentyeva, 3, 630090 Novosibirsk, Russia Novosibirsk State University, Pirogova Str., 2, 630090 Novosibirsk, Russia Institute of Organic Chemistry, National Academy of Sciences, Murmanskaya Str., 5, 02660 Kiev, Ukraine
a r t i c l e
i n f o
Article history: Received 5 August 2011 Received in revised form 29 September 2011 Accepted 29 September 2011 Available online 8 October 2011 Keywords: Thiacalixarene DFT calculation X-ray emission spectroscopy Electrophilic substitution
a b s t r a c t The DFT calculations of different thiacalix[4]arenes show that the conformation and geometry of arene core mainly determine the difference in electronic density distribution (frontier occupied MO and ESP charges) between ‘‘cone’’ and ‘‘1,3-alternate’’ conformation. The key factor is orientation of benzene rings relative to average plane through bridging sulfur atoms. Calculated structure of frontier orbitals was verified by experimental sulfur Kb spectra, calculated spectra reproducing main features of experimental ones. Difference in HOMO and ESP distribution on carbon atoms in benzene rings for ‘‘cone’’ and ‘‘1,3alternate’’ conformers correlates with earlier reported data on regioselectivity of electrophilic substitution in these compounds. Ó 2011 Elsevier B.V. All rights reserved.
1. Introduction Thiacalixarenes (TCAs) is a relatively new class of macrocyclic compounds, which were actively studied in recent years. The replacement of the bridging CH2 groups by electron-withdrawing sulfur atoms leads to a substantial difference between TCAs and classic calixarenes [1–3]. Several features of TCA such as increased acidity of OH groups or the possibility of directly replacing them by amino and amide groups [1,4] can be qualitatively explained by the inductive effects of the more electronegative sulfur atoms. The unusual regioselectivity of electrophilic substitution in TCA (Scheme 1) and its dependence on the conformation reported in Lhotak’s recent papers [5–8] require a quantitative description of the highest occupied molecular orbitals and the charge distribution on the carbon atoms in the benzene rings. From the viewpoint of coordination chemistry, the major distinction of TCA is the presence of sulfur atoms as additional donor centers. The coordination chemistry of classic calixarenes without additional donor groups is limited to phenoxo and p-complexes. The sulfur atoms of TCA can participate in complexation either together with OH groups or as single donor atoms, especially for chalcophile metals. The possibility of their participation in com⇑ Corresponding author at: Nikolaev Institute of Inorganic Chemistry, Siberian Branch, Russian Academy of Sciences, av. Lavrentyeva, 3, 630090 Novosibirsk, Russia. E-mail address: [email protected] (G.A. Kostin). 0022-2860/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.molstruc.2011.09.058
plexation is determined by the geometrical characteristics of thiacalixarenes, the electron density on the sulfur atoms, and the contribution of the valence electrons of sulfur to the highest occupied molecular orbitals. If TCA is modified by adding donor groups to the upper or lower rim, the electron interference between the groups should also be taken into consideration. The participation of the valence S 3p AO in the highest occupied molecular orbitals can be experimentally proven by X-ray emission spectroscopy (3p valence-to-1s-core XES, Kb-line). The XES spectra of TCAs obtained earlier [9] revealed a systematic difference between TCAs in the ‘‘cone’’ and ‘‘1,3-alternate’’ conformations, which was evidence of different sulfur contributions to HOMO. The experimental studies should be complemented with a profound theoretical description of the electronic structure of TCA. This study is focused on the interrelationships between the electronic structures of TCAs, their reactivities in electrophilic substitution reactions, and X-ray spectra. Previous quantum chemical calculations of TCAs were mainly focused on the energy aspects of complexation with different metals [10–13]. A qualitative description of HOMO for TCA in the ‘‘cone’’ and ‘‘1,3-alternate’’ conformations was given in [8,10].
2. Experimental Thiacalix[4]arenes 1–6 (Scheme 2) were synthesised as described in [14–18]. Model compounds 7, 8 were not synthesised,
G.A. Kostin et al. / Journal of Molecular Structure 1006 (2011) 502–507
503
structure of 8 is ‘‘pinched cone’’ with C2v symmetry common to TCA in the ‘‘cone’’ conformation tetraalkylated in the lower rim [14,25]. 3.1. Molecular orbitals and XES spectra
Scheme 1.
but DFT calculations were performed to compare the results with those for 1 and 5.The S Kb X-ray fluorescence emission spectra of 1–6 were recorded on a ‘‘Stearate’’ X-ray ultrasoft spectrometer. The X-ray tube operating conditions were V = 8 kV, I = 100 mA. A 0Þ quartz crystal was used as a crystal analyzer. The comð1 3 4 pounds were deposited on the secondary anode of the X-ray tube. The spectra were measured at the liquid nitrogen temperature. The spectrometer resolution in the studied region was 5 103. The electronic structures of synthesised thiacalix[4]arenes in the ‘‘cone’’ (1–4) and ‘‘1,3-alternate’’ (5, 6) conformations and of model compounds (7, 8) were calculated using the Jaguar 6.5 program, Shrodinger Inc. [19]. The geometry of thiacalix[4]arene molecules was optimized by the DFT technique with the B3LYP hybrid exchange-correlation functional [20,21]. The M6-31G(TM)+ basis set [22] was used for all atoms. 3. Results and discussion Table 1 shows the selected interatomic distances, interatomic angles, and interplanar angles for the optimized structures of the molecules. The calculated average bond lengths (S–C, C–O) and C–S–C bond angles are close for all structures and are in reasonable agreement with the experimental data for the crystal phases (the experimental parameters from the original papers or calculated from the corresponding CIF files in CCDC are given in parentheses). For all ‘‘1,3-alternate’’ structures 5–7, the angles between the average plane through four sulfur bridging atoms [S4] and the planes of the benzene rings are approximately 90°, the structure of the arene core being of approximately D2d symmetry. In ‘‘cone’’ conformers 1–4 with OH groups in the lower rim, the calculated interplanar angles vary from 26° to 70°. The calculated structure of 1 has approximately C4v symmetry, the interplanar angle (51.3–52.8°) being the average between the two experimental values corresponding to the ‘‘pinched cone’’ structure determined by single crystal XRD [14]. Due to the presence of substituents in the upper rim of 2–4, the symmetry is lowered from C4v to the ‘‘pinched cone’’ conformation. As thiacalixarenes tend to include small molecules in the inner cavity, the experimental interplanar angles for 2 depend on the nature of the included molecule [23,24]. The calculated values shown in Table 1 are closest to those reported for the structure of a pyridine inclusion compound [23]. The optimized
Scheme 2.
The molecular orbital diagram for thiacalix[4]arene 5 in the ‘‘1,3-alternate’’ conformation is shown along with the typical MO schemes in Fig. 1. Four highest occupied orbitals (from HOMO to HOMO-3) are mainly formed from the 3p orbitals of the sulfur atoms and involve the p-bonding orbitals of the benzene rings, the total contribution of all sulfur AOs being 40–70%. The orbitals from HOMO-4 to HOMO-7 are mainly distributed on the carbon atoms of the benzene rings (65–75%) with some contribution from the oxygen and sulfur atomic orbitals. For orbitals from HOMO to HOMO-4, the electron density distribution on the carbon atoms is determined by the «2 + 2» p-bonding orbitals of the benzene ring (Scheme 3). The electron density determined by HOMO-3, 7, and 8 is mainly located on the carbon atoms bonded to the OR group (C1) and in the para-position (C4), the carbon 2p AO forming the «3 + 3» p-bonding orbitals of the benzene ring (Scheme 3). For the doubly degenerate orbitals (HOMO-1, 2 and HOMO-5, 6), one pair of opposite benzene rings have the same electron density distribution as in the ‘‘2 + 2’’ orbital and the other as in the ‘‘3 + 3.’’ The 2p orbitals of the oxygen atoms participate in the formation of molecular orbitals starting from HOMO-9 to lower-lying ones. The total contribution of the atomic orbitals of all four oxygen atoms to the highest occupied molecules orbitals does not exceed 13%. For ‘‘1,3-alternate’’ 7 with OH instead of methoxy groups, the structure of the molecular orbitals changes but slightly. The orbitals that are close in energy are inverted (HOMO-4 of b1 symmetry and HOMO-5, -6 of e symmetry), but the contributions of atomic orbitals to different HOMOs coincide well with the results obtained for 5. The highest occupied molecular orbital of 1 (Fig. 2), having b2 symmetry, corresponds to the linear combination of the ‘‘3 + 3’’ p-bonding orbitals of the benzene rings (70%) and the 2p atomic orbitals of the oxygen atoms (27%). Although the oxygen 2p orbitals in HOMO are perpendicular to the benzene plane, they have different phase signs, thus forbidding the p-p bonding. The contribution of the sulfur AO to HOMO is negligible. The doubly degenerate HOMO-1 and HOMO-2 have similar structures, each orbital being distributed on only two of the four benzene rings. The HOMO-3, HOMO-5, and HOMO-6 are the linear combinations of mainly 3p sulfur atomic orbitals, the total contribution of all sulfur AOs being 70–80%. The occupied molecular orbitals that are lower in energy are formed by the «2 + 2» p-bonding orbitals of the benzene rings with a contribution from the 3p (3s) sulfur orbitals, the total contribution of the sulfur AOs varying from 21 to 42%. The positive overlap of the S3p orbitals with the p-bonding orbitals of the benzene ring in HOMO-7 and HOMO-8 leads to additional S–C pbonding. It is interesting to note that for thiacalixarene 8 in the ‘‘pinched cone’’ conformation, molecular orbitals from HOMO to HOMO-3 are similar in shape to the orbitals of TCA in the ‘‘1,3alternate’’ conformation (Fig. 1) and mainly formed from the sulfur atomic orbitals (50–60%). The doubly degenerate HOMO-4 and -5 in TCA 8 are primarily localized on the benzene rings orientated outwardly relative to the cavity of TCA, the electron density being maximum on the C1 and C4 carbon atoms. As in the case of 5, in TCA 8 the oxygen atomic orbitals make a considerable contribution (20–55%) only to HOMO-9 and lower-energy orbitals (HOMO-10 to -12). The optimized geometry of phosphorylated TCA 4 differs substantially from C4v symmetry, and this is also responsible for the less symmetric character of HOMO (Fig. 3). The structure of the occupied orbitals in 4, however, is similar that in 1 (‘‘cone’’). The highest occupied molecular orbital is a linear combination of
504
G.A. Kostin et al. / Journal of Molecular Structure 1006 (2011) 502–507
Table 1 Geometrical parameters of the optimized structures for the compounds. The experimental single crystal XRD data [14,18,23] are shown in parentheses. L
d(S–C), Å
d(O–C), Å
1 2 3 4
1.80 (1.783) 1.81 (1.780) 1.81 1.86
1.36 (1.363) 1.36 (1.366) 1.38 1.38
5 6 7 8
1.80 (1.772) 1.80 1.85 1.80
1.37 (1.379) 1.37 1.40 1.37
d(P–O), Å
1.624 1.730 1.597
\C–S–C, °
[Cn]–[S4] interplanar angles
104.0 (102.7) 104.4 (101.29) 104.0 104.2
52.8, 55.5, 38.7, 26.7,
51.3, 49.4, 50.9, 59.5,
52.8, 54.8, 41.9, 31.3,
51.3 (62.2, 43.1) 47.8 69.5 70.1
103.2 (102.8) 104.0 102.2 101.2
86.7, 88.6, 88.9, 31.2,
87.1, 88.0, 88.6, 84.9,
86.7, 87.4, 88.9, 31.1,
87.1 (85.1, 86.3) 83.9 88.6 84.3
Fig. 1. MO diagrams shown energy levels of HOMO for thiacalixarenes 5, 7, 8 and selected orbitals for 5.
Fig. 2. MO diagram shown energy levels of HOMO and selected orbitals for thiacalixarene 1.
Scheme 3. Designation of carbon atoms in benzene ring and signs of molecular orbital corresponding to two orbital types.
the «3 + 3» type p-bonding orbital of the benzene ring, the sulfur atomic orbitals (17%), and the atomic orbitals of the oxygen atoms in the lower rim (20%). The HOMOs-1, -4, and -5 are mainly formed by the sulfur AO (50–60%). The atomic orbitals of the oxygen atoms of the phosphoryl groups are involved in the formation of HOMO starting from HOMO-7. The molecular orbitals with energies from 7.44 to 7.78 eV (HOMO-12–HOMO-19) correspond to the lone electron pairs of the oxygen atoms double-bonded to phosphorus. The lower-lying orbitals are mainly formed from the lone electron pairs of the oxygen atoms of the ethoxy groups bonded to phosphorus. The experimental Kb-spectra of sulfur indirectly confirm the structure of the highest occupied orbitals in the studied thiacalixarenes. All the studied compounds have two bands at binding energies of 2466 (A) and 2469 (B) eV in the experimental spectra, maximum B being more intense for all TCAs in the ‘‘cone’’ conformation (Fig. 4). The S(Kb) valence-to-core spectra correspond to the electron transition from the occupied MO including the sulfur 3pAO to the S 1s core level. The intensities of the Kb transitions are proportional to the squares of the coefficients c2i of the sulfur 3p AO in different occupied molecular orbitals. An analysis of the
Fig. 3. MO diagram shown energy levels of HOMO and selected orbitals for phosphorylated thiacalixarene 4.
valence-to-core transitions thus enables us to estimate the contribution of the sulfur 3p AO to the occupied molecular orbitals. The calculated spectral transitions were convolved by 1.5 eV full width at half maximum to model the experimental broadening (Fig. 5, green line). A less line broadening (0.5 eV) was chosen to
505
G.A. Kostin et al. / Journal of Molecular Structure 1006 (2011) 502–507
7.41 eV (the range is twice wider) for the ‘‘1,3-alternate’’ conformer. Both factors result in a broadening of maximum B and lowering of intensity. 3.2. Reactivity of carbon atoms in the benzene rings
Fig. 4. Experimental sulfur Ka spectra for investigated compounds.
Two approaches can be used for describing the reactivity. The first approach is based on the calculation of relative energies for the activation complex formed after the attack at different positions, but the results of calculation strongly depend on the nature of the attacking particle, the geometry of the intermediate state, and the details of calculations. Earlier, Lhotak reported on similar calculations for the electrophilic substitution in ‘‘1,3-alternate’’ thiacalixarene, but the results obtained for different attacking particles and different calculation techniques agreed with the experimental selectivity only in several instances [8]. The second approach is based on the calculation of molecular descriptors, which are formally independent of the nature of the attacking reagent and the geometry of the activated complex but define only the electron density distribution in the TCA molecule. According to perturbation theory, the first two terms of the series DEtotal ¼ Eel þ Eorb correspond to the electrostatic interactions of molecules and interactions that are due to the overlap of their molecular orbitals [26,27]. With electrostatic regioselectivity control (charge-controlled reactions), the electrophilic agent with high charge density and low polarizability (for example, H+) is expected to attack the ring sites with a maximum negative charge. The direction of electrophilic substitution will thus be determined by the largest negative values of the electrostatic potential (ESP) and the R qðr0 Þ 0 P Zi corresponding atomic charges [26–28]: UðrÞ ¼ M dr . i¼1 jrRi j jrr 0 j
LUMO 0, the overlap between the If the difference EHOMO donor Eacc donor and acceptor molecular orbitals is dominant in energy (the second term of DEtotal). The regioselectivity of similar frontiercontrolled reactions can be determined in terms of the Fukui functions [29–32]. The Fukui function for electrophilic substitution 2 P i ðrÞj is defined by fN ðrÞ ¼ j/N ðrÞj2 þ Ni¼1 ð@j/@N Þ v ðrÞ .
The second term of this equation is responsible for orbital relaxation, the change in the orbital shape accompanying the removal of electrons from the system. When this term is small compared to the first term, it can be neglected and the function is determined by only the frontier highest occupied molecular orbital j/N ðrÞj2 . Table 2 DFT calculated ESP charges and atomic contributions to HOMO for 1, 5, 7, and 8 (the maximum negative charges and the largest contributions of carbon atoms are bold). Atom
Atomic charges C1 0.29 C2 0.00 C3 0.08 C4 0.24 C5 0.08 C6 0.03 O 0.46 S 0.12
Fig. 5. Experimental (bold line) and calculated Ka – spectra for thiacalixarenes 1, 5, 7, 8. Calculated spectra are shown with 1.5 eV (green line) and 0.5 eV (red dotted line) broadening. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
determine the fine structure of the calculated spectra (Fig. 5, red line). A comparison of the experimental and calculated spectra for 1 and 5 shows that maximum B is related to the transitions from the MO involving the lone electron pairs of the sulfur 3p AO, while maximum A is due to the sulfur 3p orbitals involved in the formation of the S–C r-bonds. The decrease in short-wave maximum B for TCA in the ‘‘1,3-alternate’’ conformation is determined by two factors. Firstly, for ‘‘1,3-alternate’’ TCA, the contribution of the 3p sulfur atomic orbitals in different HOMOs is more regular. Secondly, maximum B is determined by the transitions from the molecular orbitals with energies from –6.83 to 7.55 eV (the range is 0.7 eV) for the ‘‘cone’’ conformer and from 5.98 to
1
Contributions to HOMO-0 C1 0.04 C2 0.03 C3 0.01 C4 0.05 C5 0.01 C6 0.03 O 0.07 S 0.01 a b
5
7
8 a
b
0.09 0.37 0.30 0.12 0.30 0.37 0.30 0.27
0.09 0.26 0.23 0.17 0.23 0.26 0.56 0.32
0.44 0.06 0.18 0.06 0.19 0.01 0.21 0.22
0.13 0.18 0.13 0.23 0.15 0.20 0.28
0.01 0.01 0.03 0.00 0.03 0.01 0.00 0.15
0.01 0.02 0.03 0.00 0.03 0.02 0.00 0.15
0.01 0.02 0.03 0.00 0.04 0.01 0.01 0.15
0.01 0.01 0.03 0.00 0.03 0.01 0.00
For the benzene rings lying at an angle of 84–85° to the [S4] plane. For the benzene rings lying at an angle of 31° to the [S4] plane.
506
G.A. Kostin et al. / Journal of Molecular Structure 1006 (2011) 502–507
We analyzed the ESP charges and electron density distributions related to the highest occupied molecular orbital for TCA 1, 5, 7, and 8. The atomic contributions to HOMO (Table 2) were calculated as Pn 2 i¼1 c i , where n is the number of atomic basis functions, and ci is the coefficient for the ith basic function in the HOMO. The results were averaged over four rings for 1, 5, and 7 and over pairs of opposite rings for 8. The differences do not exceed 0.02 for charges and are less than 0.005 for the atomic contributions to HOMO. An analysis of the data of Table 2 indicates that for TCA 1 in the ‘‘cone’’ conformation, both descriptors are favorable for para-substitution; the maximum negative charge and maximum density of HOMO are localized on the carbon atoms in the para-position relative to the OH groups, which agrees with the strong para-orientation of the hydroxo groups. The charge distribution and atomic contributions to HOMO in thiacalixarenes 5 and 7 in the ‘‘1,3-alternate’’ conformation differ considerably from those for 1. In both alternates, the maximum negative charge and maximum density of HOMO are located on the C3 and C5 carbon atoms in the meta-position to the oxygen-containing groups. This effect is more pronounced for 5; in 7, the charge distribution between the C3–C5 atoms is more regular. The most interesting result was obtained for model TCA 8 in the ‘‘pinched cone’’ conformation with C2v symmetry. Two opposite benzene rings in it are almost parallel to each other and perpendicular to the [S4] plane; the other two rings are pointer outside the cavity of TCA, the interplanar angle being 118° (Table 1). The distribution of HOMO here is almost the same for all the four benzene rings and coincides with the distribution found for both TCAs in the ‘‘1,3-alternate’’ conformation. The ESP charge distribution in the benzene rings perpendicular to the [S4] plane in 8 also qualitatively coincides with that in the alternates, the maximum negative charge lying on the C3 and C5 atoms. In the other two rings, however, the maximum negative charge lies on the C4 atom in the para-position relative to the OR group, as it does in TCA 1 in the ‘‘cone’’ conformation. The orientation of the benzene rings with respect to the [S4] plane thus plays the key role in the distribution of ESP charges on the C3–C5 atoms, though alkylation of OH groups also results in a certain increase in the total electron density on the C3 and C5 atoms (cf. 1 and 82 or 7 and 5). Clearly, the calculated descriptors do not take into consideration such factors as steric accessibility of reaction centers, possible solvation effects in condensed phases, etc. Nevertheless, the regularities obtained are in complete agreement with both new experimental data on regioselectivity obtained by Lhotak and numerous data found earlier for the electrophilic substitution in 1. Thiacalixarene 1 was the starting compound for different modifications, and electrophilic substitution occurred at the paraposition in all cases [2,3], as could be predicted from both HOMO and charge distributions (Table 2). In 1,3-alternates, however, both factors are favorable for electrophilic substitution at the meta-position. Lhotak discovered that nitration or formylation (both Duff and Gross conditions) of similar TCAs resulted in only meta-substituted products [5,8]. For tetraalkylated thiacalixarene 8 in the ‘‘pinched cone’’ conformation, the situation is more complex. For frontier-controlled reactions favored by HOMO distribution, meta-substitution should be expected according to Table 2, while the ESP charge distribution favors meta-substitution in the arene rings perpendicular to the [S4] plane and para-substitution in the other two rings. Depending on the conditions of formylation for similar TCA with OPr groups in the lower rim, Lhotak described the para-substituted products predominant for the Duff reaction (51% yield vs 15% for meta-substituted product) [6] and only meta-substituted products for the Gross reaction [7]. Studies of the transformations of the CHO group in the meta-substituted products of the Gross reaction showed that the –CHO and –CN groups in the meta-position of the benzene ring shifted the
equilibrium between the two ‘‘pinched cone’’ forms in such way that the electron-withdrawing meta substituents were located in the vertical benzene rings. In contrast, the reduction of the formyl group to CH2OH or CH3 led to a conformer in which the electron donor meta substituent was in the rings rotated out of the cavity of TCA. Both facts correlate well with the ESP charge distribution on the C3 and C5 atoms in thiacalixarene 8. In the rings rotated out of the cavity, the negative ESP charges on these atoms are lower, thus favoring the electron donor groups. 4. Conclusions The DFT calculations of thiacalixarenes in the ‘‘cone’’, ‘‘pinched cone,’’ and ‘‘1,3-alternate’’ confirmations revealed that the orientation of the benzene rings relative to the mean [S4] plane through four bridging sulfur atoms played a critical role in electron density distribution (the highest occupied molecular orbital and ESP charges). In the benzene rings that form angles of approximately 90° (84–89°) with the [S4] plane, the carbon atoms in the meta positions to the phenol group are the most favorable sites for an electrophilic attack. This result could help to understand the unusual regioselectivity found by Lhotak for tetraalkylated thiacalix[4]arenes in the ‘‘pinched cone’’ or ‘‘1,3-alternate’’ conformation. In the benzene rings lying at an angle of 30–50° to the [S4] plane of tetrahydroxo thiacalix[4]arenes in the ‘‘cone’’ or ‘‘pinched cone’’ conformation, the maximum negative ESP charge and the maximum density of HOMO are located on the carbon atom in the para position to the phenol oxygen. The calculated difference in the structure of the highest occupied molecular orbitals between the ‘‘1,3-alternate’’ and ‘‘cone’’ conformations was experimentally confirmed by the sulfur Kb spectra of thiacalix[4]arenes. The experimentally observed decrease in the relative intensity of the short-wave band (2469 eV) for the alternate conformers was explained by the more regular distribution of the sulfur 3p AO in the HOMO and the wider range of energies of the molecular orbitals that formed this maximum. References [1] N. Iki, S. Miyano, J. Incl. Phenom. 41 (2001) 99–105. [2] E.A. Shokova, V.V. Shokova, Russ. J. Org. Chem. 39 (2003) 1–28. [3] N. Morohashi, F. Narumi, N. Iki, T. Hattori, S. Miyano, Chem. Rev. 106 (2006) 5291–5316. [4] P. Rao, M.W. Hosseini, A.D. Cian, J. Fischer, J. Chem. Soc., Chem. Commun. (1999) 2169–2170. [5] O. Kundrat, J. Kroupa, S. Bohm, J. Budka, V. Eigner, P. Lhotak, J. Org. Chem. (2010) 8372–8375. [6] O. Kundrat, H. Dvorakova, I. Cisarova, M. Pojarova, P. Lhotak, Org. Lett. 11 (2009) 4188–4191. [7] O. Kundrat, H. Dvorakova, V. Eigner, P. Lhotak, J. Org. Chem. 75 (2010) 407– 411. [8] O. Kundrat, I. Cisarova, S. Bohm, M. Pojarova, P. Lhotak, J. Org. Chem. 74 (2009) 4592–4596. [9] V.G. Torgov, L.N. Masalov, G.A. Kostin, T.V. Us, T.M. Korda, N.A. Kryuchkova, E.A. Korotaev, A.D. Fedorenko, A.B. Drapailo, J. Struct. Chem. 52 (4) (2011) 738– 745. [10] M.-L. Fu, N.L. Rangel, R.D. Adams, J.M. Seminario, J. Clust. Sci. 21 (2010) 867– 878. [11] A. Suwattanamala, A.L. Magalhaes, J.A.N.F. Gomes, Theor. Chem. Acc. 117 (2007) 431–440. [12] A. Suwattanamala, A.L. Magalhaes, J.A.N.F. Gomes, J. Phys. Chem. A 109 (2005) 10742–10752. [13] L.I. Shamova, G.A. Shamov, I.S. Antipin, A.I. Konovalov, J. Phys. Chem. A 113 (2009) 5691–5699. [14] H. Akdas, L. Bringel, E. Graf, M.W. Hosseini, G. Mislin, J. Pancanel, A.D. Cian, J. Fischer, Tetrahedron Lett. 39 (1998) 2311–2314. [15] O. Kasyan, D. Swierczynski, A. Drapailo, K. Suwinska, J. Lipkowski, V. Kalchenko, Tetrahedron Lett. 44 (2003) 7167–7170. [16] H. Kumagai, M. Hasegawa, S. Miyanari, Y. Sugawa, Y. Sato, T. Hori, S. Ueda, H. Kamiyama, S. Miyano, Tetrahedron Lett. 38 (1997) 3971–3972. [17] J. Lang, H. Dvorakova, I. Bartosova, I. Stibor, R. Hrabal, Tetrahedron Lett. 40 (1999) 373–376. [18] P. Lhoták, M. Himl, I. Stibor, J. Sykora, H. Dvorakova, J. Lang, H. Petrı´ckova, Tetrahedron 59 (2003) 7581–7585.
G.A. Kostin et al. / Journal of Molecular Structure 1006 (2011) 502–507 [19] [20] [21] [22] [23]
Jaguar 7.5, Shrodinger Inc. – Portland OR. – (2008). A. Becke, J. Chem. Phys. 98 (1993) 5648–5653. C. Lee, W. Yang, R.G. Parr, Phys. Rev. B. 37 (1988) 785–789. A.V. Mitin, J. Baker, P. Pulay, J. Chem. Phys. 118 (2003) 7775–7783. A. Bilyk, A.K. Hall, J.M. Harrowfield, M.W. Hosseini, B.W. Skelton, A.H. White, Inorg. Chem. 40 (2001) 672. [24] N. Iki, C. Kabuto, T. Fukushima, H. Kumagai, H. Takeya, S. Miyanari, T. Miyashi, S. Miyano, Tetrahedron 56 (2000) 1437. [25] M. Himla, M. Pojarova, I. Stibora, J. Sykorab, P. Lhoták, Tetrahedron Lett. 46 (2005) 461–464.
[26] [27] [28] [29]
507
G. Klopman, J. Am. Chem. Soc. 90 (1968) 223–233. L.J. Bartolotti, P.W. Ayers, J. Phys. Chem. A 109 (2005) 1146–1151. T. Brinck, J. Phys. Chem. A 101 (1997) 3408–3415. F. Gilardoni, J. Weber, H. Chermette, T.R. Ward, J. Phys. Chem. A 102 (1998) 3607–3613. [30] C. Cardenas, N. Rabi, P.W. Ayers, C. Morell, P. Jaramillo, P. Fuentealba, J. Phys. Chem. A 113 (2009) 8660–8667. [31] R.K. Roy, J. Phys. Chem. A 107 (2003) 397–404. [32] T. Mineva, V. Parvanov, I. Petrov, N. Neshev, N.J. Russo, Phys. Chem. A 105 (2001) 1959–1967.