J. Phys. Chem. A 2010, 114, 7725–7732
7725
Electronic Structure and Normal Vibrations in (+)-Catechin and (-)-Epicatechin Encapsulated β-Cyclodextrin Jayshree K. Khedkar, Vivekanand V. Gobre, Rahul V. Pinjari, and Shridhar P. Gejji* Department of Chemistry, UniVersity of Pune, Ganeshkhind, Pune 411007, India ReceiVed: March 14, 2010; ReVised Manuscript ReceiVed: June 11, 2010
Host-guest interactions between β-cyclodextrin (β-CD) and flavan-3-Ol enantiomers (guest) namely, (+)catechin (CA) or (-)-epicatechin (EC), have been analyzed within the framework of density functional theory. Both CA and EC consist of two phenol rings, I and II, and a pyran ring, III, which facilitate a variety of binding patterns with the host, β-CD. The minimum energy β-CD-CA complex reveals that ring II of CA interacts with primary hydroxyls of the upper rim and the phenol ring I engenders hydrogen-bonded interactions with secondary hydroxyl from the lower rim of CD. On the other hand, the O-H · · · O interactions between ring I and primary hydroxyls of β-CD along with those between one of hydroxyl of ring II and secondary hydroxyl of the host render large stability to the β-CD-EC complex. Structures of both β-CD-CA and β-CD-EC complexes thus obtained are in consonant with those inferred from the experimental NMR data and exhibit distinct features in infrared spectra. The frequency shifts of characteristic vibrations in infrared spectra of these complexes compared to the unbound individual host or guest in its free state have been analyzed with the use of natural bond orbital analyses and combining difference electron density maps with bond critical points in molecular electron density topography. Introduction Cyclodextrins (CDs) are chiral, cyclic oligosaccharides comprised of D-glucopyaranoside units linked together by R-1,4glycosidic linkage. CDs are widely used receptors in host-guest inclusion chemistry and can be synthesized from starch by simple enzymatic conversion.1 Three important crystalline, homogeneous, and nonhygroscopic cyclodextrins R-, β-, and γ-CD, consisting of 6, 7, and 8 glucopyaranose units, respectively, are known.2-5 CDs are represented as cylindrical tunnels with the upper (narrow) rim consisting of primary hydroxyl groups and the lower (wide) rim attached to secondary hydroxyl groups. Both these attributes, namely, the hydrophilic hydroxyl groups around both rims and the relatively hydrophobic cavity that results from the CH groups and glycosidic oxygens, render these hosts with the unique ability of selective and efficient binding toward several organic and organometallic guests to produce inclusion complexes.1,6 The encapsulation of guest within the host cavity enhances stability and further improves bioavailability of molecules of pharmaceutical interest.7 CDs, therefore, have widely been used as host in molecular recognition for enantiomer separation in racemic mixtures, and their inclusion complexes in solution8 as well as in solid state9,10 have further been explored in numerous applications in pharmaceutical science,11-13 catalysis,14,15 affinity chromatography,16,17 drug research, and asymmetric reactions.18 Inclusion complexes of CD serve as ideal models mimicking enzyme-substrate interactions.19,20 To understand the host-guest interactions following inclusion of the guest within the CD cavity molecular mechanics,21,22 semiempirical23,24 or single-point density functional calculations25 have been utilized. Molecular dynamics simulations have also been carried out on encapsulation of substituted phenols and imidazoles within CD hosts.26 The inclusion of guests within * To whom correspondence should be addressed. E-mail: spgejji@ chem.unipune.ac.in.
the relatively hydrophobic CD host cavity27 was attributed to noncovalent interactions, hydrogen bonding, and release of ring strain in the cavity.28 Complementarity between hosts and guests in terms of their size and shape has been crucial for molecular recognition theories. Recently the present authors have proposed a method based on molecular electrostatic potential (MESP) to estimate effective cavity dimensions of CD conformers exhibiting different hydrogen bonding patterns. It has been shown that shape of host-cavity can be gauged from the charge distribution in terms of MESP topography.29-33 Thus, β-CD, by far the most widely been used host, exhibits either cone-like or barrel-shaped cavity in different conformations. Interestingly, flavan-3-Ol enantiomers, namely, (+)-Catechin (CA) and (-)-epicatechin (EC), consisting of two phenyl rings I and II with four phenolic hydroxyl groups and pyran ring III, can bind to β-CD in a variety of ways. These phenolic compounds, which are present in natural food and drinks such as green tea,34 can be extracted from plants35 and are important in biological functions such as antioxidation,36 deodorization and have antibiotic37 and cancer-inhibiting properties.38,39 Both CA and EC are, however, bitter and easily oxidizable and thus can not be used directly as medicine, cosmetic, or food additives. An inclusion of these phenolic compounds in β-CD host has therefore been envisaged to mask the nasty aspect40 of its odor and taste. To this end, the structure of β-CD complexed with CA and EC has been investigated by different research groups in the literature. Formation of the inclusion complexes of CA and β-CD was evident from the 1H NMR measurements by Smith et al.41 The formation constants for the complex were estimated with the use of circular dichroism. Thus, structures of inclusion complexes of β-CD with CA and EC in solution have been inferred from the 1H NMR experiments.40 The β-CD-EC complex thus revealed that ring II of CA was included deeply in the β-CD cavity from secondary hydroxyl group, whereas
10.1021/jp102304j 2010 American Chemical Society Published on Web 07/02/2010
7726
J. Phys. Chem. A, Vol. 114, No. 29, 2010
Khedkar et al.
ring I remains outside the cavity near the secondary hydroxyl group. These experiments further revealed that the enantiomer EC was fitted loosely in the cavity when complexed with β-CD. Kriz et al42 have reported complexation of CA with β-CD employing 2D NMR and micro-calorimetric measurements. The experimental data when combined with the force field calculations incorporating docking of ligand and host molecules suggested that ring II encapsulates within the host cavity whereas ring I facilitates interactions with secondary hydroxyls in the β-CD-CA complex. Recent calculations of Yan et al25 on the β-CD-CA complex using semiempirical PM3 geometries support these conclusions. From the calculated binding energies using PM3 geometries it was not clear whether ring II of EC interacts from primary hydroxyl or secondary hydroxyl end of host and thus yield two probable structures for β-CD-EC complex. These calculations further led to the conclusion that the C-H · · · O interactions, along with the O-H · · · O and van der Waals contacts, between β-CD and enantiomer contribute significantly to the stability of the inclusion complex. Recently, Julian et al43 have analyzed the inclusion properties and selectivity of β-CD and modified β-CD hosts toward CA based on 2D NMR methods and further carried out molecular docking studies employing semiempirical PM3 optimizations. It was inferred that the catechol moiety orients toward primary rim of β-CD complex as noticed earlier. The present work focuses on molecular level understanding of binding patterns and vibrational characteristics of β-CD complexed with CA or EC. Computational Method The atomic numbering scheme of β-CD monomer and enantiomers CA or EC is shown in Figure 1. β-CD conformers possessing either clockwise or anticlockwise hydrogen bonding patterns in upper and lower rims, classified as “A”, “B”, and “C”33 are displayed in Figure 2. Accordingly conformers “A” possess O6H · · · O6′ interactions whereas those of “B” are comprised of O6H · · · O5′ interactions. Further the intraglucose O6H · · · O5 interactions are present only in “C” conformers. Here, a single prime refers to atoms from an adjacent glucose unit. Thus, geometries of inclusion complexes of CA and EC with β-CD possessing interactions between either 3′-OH/4′-OH of ring II or 5-OH/7-OH of ring I in CA/EC with primary hydroxyls of β-CD, were considered. Selected complexes from each of aforementioned A-C33 type of β-CD were optimized using the density functional theory incorporating Becke’s three-parameter exchange with Lee, Yang, and Parr’s (B3LYP) correlation functional.44,45 The optimizations on the whole system, without any constraint, were carried out using an internally stored 6-31G(d) basis set employing the Gaussian 09 program.46 Stationary point geometries of these complexes were confirmed to be local minima since all the frequencies of normal vibrations turn out to be real and no imaginary frequency was noticed. Binding energies of guest in gas phase were calculated by subtracting the sum of energies of individual fragments viz., β-CD host and enantiomer (CA or EC) from the total electronic energy of complex. The normal modes were assigned by visualizing displacements of atoms around their equilibrium (mean) positions using the locally written program UNIVIS-2000.47 The natural bond orbital (NBO) analyses were carried out to understand the direction of frequency shifts in the CA or EC complex of β-CD relative to isolated β-CD host or free CA/EC enantiomer.48 The quantum theory of atoms in molecules (QTAIM) was employed to investigate the molecular electron density (MED) topography and the bond critical point (bcp) characterized as
Figure 1. Atomic numbering scheme in (a) CA, EC and (b) glucose unit cut from CD.
(3,-1) CP, of different hydrogen bonds.49,50 An excellent review of MED topography is given in refs 49 and 50. A program developed in our laboratory was used to calculate the density at the bcp (Fbcp) in MED topography.51 Difference electron density (∆F) was calculated by subtracting the sum of electron densities of individual β-CD and the guest CA or EC from that of the host-guest complex. Positive and negative values of (∆F) refer to enhanced and depleted electron density, respectively, at the given grid point, which was visualized using the program UNIVIS.47 Calculated ∆F was mapped on bcp in the MED with a color mapping function that renders the appropriate color to bcp depending on enhanced (blue) or depleted (red) electron density. Results and Discussion Complexes of CA and EC, each from aforementioned A-C conformation shown in Figure 2, were subjected to optimizations using the B3LYP/6-31G(d) theory. The complexes with increasing stabilization energies are designated by ICA1, ICA2. . . for β-CD-CA and by IEC1, IEC2. . . in case of β-CD-EC. The hydrogen bond distances in ICA1 complexes are reported in Table 1. Thus, the optimized structure of ICA1 comprises of hydrogen bonded interactions between hydroxyls of ring II of
Host-Guest Interactions with β-Cyclodextrin
J. Phys. Chem. A, Vol. 114, No. 29, 2010 7727
Figure 2. Optimized structures of A, B, and C β-CD conformers.
TABLE 1: Selected Hydrogen Bond Distances (in Å) and Electronic Density at bcp (Gbcp) in MED Topography in ICA1 and IEC1 Complexes O-H · · · O distances
Fbcp
O6H · · · O3′ O6H · · · O4′ O6 · · · H-O4′ O6 · · · H-O3′ O3 · · · H-O5
ICA1 1.878 1.788 1.701 1.678 1.887
0.031 0.036 0.046 0.047 0.030
O6H · · · O7 O6H · · · O5 O6 · · · H-O7 O6 · · · H-O5 O3H · · · O3′ O2 · · · H-O3′
IEC1 1.794 1.820 1.668 1.646 1.827 1.714
0.037 0.034 0.049 0.058 0.036 0.044
CA and primary hydroxyl groups of β-CD, whereas ring I interacts with secondary hydroxyls from the host, leaving 6H protons outside the host cavity. Geometries of the energetically less favorable complexes are given in Figure 1S of Supporting Information. In ICA1 complex, the C-C bond joining ring II and ring III orients at an angle of θ ∼ 21° with cavity axis of β-CD. The optimized geometry of ICA1 is depicted in Figure 3a. Interestingly, H5 and H6 protons of β-CD exhibit upshifted signals in the experimental NMR spectra owing to interactions of β-CD and ring II of CA while the H3 proton within the host cavity interacts with ring I of CA, concomitantly leading to upfield signals in the spectra. Broadening of CA protons, except for 6H, led to the conclusion that ring II, ring III and the portion of ring I of CA are encapsulated within the β-CD host.40 B3LYP/6-31G(d) optimizations yield ICA1 complex as the minimum, and the predicted structure is consistent with that inferred from the NMR experiments. On the other hand, enantiomer EC upon complexation with β-CD exhibits different binding patterns than those of CA. In IEC1 the guest orients at an angle of θ ∼ 38° with the cavity axis of β-CD. This is depicted in Figure 3b, which shows that ring I is bound to primary hydroxyls whereas ring II interacts with secondary hydroxyls of β-CD. The calculations predict that ring I of EC facilitats hydrogen bonding interactions with primary hydroxyls of β-CD (the distance O-H · · · O being 1.668 Å), rendering large stability to it. The corresponding bond distances from the host-guest interactions are reported in Table 1. It should be remarked here that the measured 1H NMR spectra, suggests two possibilities for β-CD-EC complex where either ring I or ring II interacts with primary hydroxyl of β-CD.
Figure 3. The minimum energy inclusion complexes (a) ICA1 and (b) IEC1. The tilt angle θ (in °) represents orientation of CA/EC with the host cavity-axis.
It was not conclusive, however, whether the primary hydroxyls of β-CD interact with protons from ring I or ring II of EC. The work of Ishitzu et al40 suggested plausibility of interactions between ring II and primary hydroxyl of β-CD. On the contrary, molecular modeling investigations incorporating semiempirical
7728
J. Phys. Chem. A, Vol. 114, No. 29, 2010
Khedkar et al.
TABLE 2: Selected B3LYP Vibrational Frequencies of CA, EC, β-CD, ICA1, and IEC1a CA
EC
ν(-O5H) ν(-O7H) ν(-O3′H) ν(-O4′H)
3623 3609 3574 3627
(35) (43) (90) (72)
3610 3608 3570 3624
(48) (47) (94) (56)
ν(-O3H) ν(CH2) (III)
3594 2935 2909 2903 1626
(13) (28) (65) (49) (216)
3568 2976 2920 2910 1616 1598 1497 1610 1600 1518 1280 1198 1034
(21) (11) (38) (9) (257) (191) (131) (35) (15) (100) (139) (210) (292)
ν(C-C) (I) ν(C-C) (II)
1614 (6) 1283 (187)
ν(C-C) + δ(OH)
1212 (191)
ν(C-O-C) + ν(C-C) ν(OHsec) (NHB)
1085 (157)
β-CD
3576 (134)
β-CDexpb
3504
ICA1 3618 3314 3086 3236 3053 3233 2861 2831
(32) (563) (989) (1296) (1695) (281) (41) (23)
3002 3081 3150 3532
1629 1590 1500 1601 1575 1505 1271 1264 1042 3577
(172) (236) (57) (13) (25) (121) (69) (135) (24) (102)
3505 3481 3471 3442 3439 3381 3376 3347 3341
(141) (188) (288) (431) (278) (254) (421) (294) (1365)
1619 (190) 1602 (55) 1514 (58) 1603 (84) 1595 (43) 1508 (184) 1279 (158) 1274 (102) 1020 (187) 3577 (42) 3576 (38) 3482 (218) 3453 (667) 3450 (242) 3440 (511) 3362 (629) 3447 (360) 3408 (869) 3389 (668) 3378 (321) 3293 (833) 2965 (58) 2963 (28) 2959 (21) 1483 (44) 904 (164) 1393 (98) 1399 (45) 1135 (220)
ν(OHsec) (HB)
3467 (1026)
ν(OHpri) (HB)
3325 (2412)
ν(C-H3+C-H5)
2922 (43)
2926 (89) 2923 (37)
δ(OHpri)
1512 (69) 660 (694) 1394 (25) 1392 (86) 1134 (523)
1491 (20) 674 (135) 1393 (46) 1396 (80) 1134 (217) 1129 (116) 1061 (278)
δ(CH1+CH2+CH3+CH4+CH5) δ(OHsec) ν(C-O-C) interglucose ν(C-C+C-O-C) ν(C-O-C) intraglucose ν(C-C) + δ(C-H) CH2 rock + ν(C-O)
1110 (174) 1087 (174) 1030 (858) 975 (175) 928 (2) 860 (37)
3336
1390
1080 1030
IEC1
1032 (293) 1029 (467)
(1131) (1007) (1020) (62)
3578 (34) 2928 (28)
1061 (172) 1032 (355) 1029 (384)
950 860
ν ) stretching, δ ) bending. (I), (II), and (III) refer to phenolic and pyran rings in guest. (HB) ) hydrogen bonded, (NHB) ) non-hydrogen bonded. b Reference 53. a
PM3 optimized geometries by Yan et al.25 predicted preference for structures where ring I interacts with primary hydroxyls of the host in β-CD-EC complex. Thus, it has been predicted that IEC1 possessing ring I bound to primary hydroxyls of β-CD turns out to be 60 kJ mol-1 lower in energy than IEC3 where ring II interacts with primary hydroxyl groups of the host. From the NMR experiments it was inferred that CA binds more strongly to β-CD than EC. As opposed to this molecular modeling, studies by Yan et al.25 concluded that EC binds more strongly to β-CD host. Calculated binding energies of EC in IEC1 in gas phase, using the B3LYP/6-31G(d) framework of theory, has been predicted to be 79.76 kJ mol-1, which is nearly 3.64 kJ mol-1 higher than that in the β-CD-CA complex (zeropoint corrected energies were incorporated for binding energy calculations). It may be remarked here that IEC1 possesses 11 hydrogen bond interactions from primary and secondary hydroxyl groups and from cavity protons of host compared to 9 noticed in the ICA1 complex. These hydrogen bond interactions
emerge with signature as bcp in MED topography, and the corresponding electron densities predict the primary hydroxyl oxygens of β-CD facilitate relatively strong hydrogen bond interactions with guest protons in ICA1, which is evident from O6 · · · H bond distances reported in Table 1. B3LYP/6-31G(d) vibrational frequencies, scaled by 0.9614,52 of CA, EC, β-CD, and ICA1 as well as IEC1 complexes are reported in Table 2. FTIR measurements53 of infrared spectra of solid samples of β-CD by Egyed showed a broad band near ∼3400 cm-1, a composite of four vibrations that upon deconvolution yielded two intense peaks near 3504 and 3336 cm-1. These peaks were assigned to primary hydroxyl and secondary hydroxyl stretching vibrations of β-CD, respectively. A direct comparison of experimentally measured infrared spectra with those derived from the present calculations is far from straightforward since the calculations refer to a β-CD molecule in its free state. B3LYP/6-31G(d)-derived vibrational spectra of β-CD exhibit different bands for OH stretchings, which by and large
Host-Guest Interactions with β-Cyclodextrin
J. Phys. Chem. A, Vol. 114, No. 29, 2010 7729
Figure 4. IR spectra of ICA1, β-CD, and CA in (a) 3700-2800 cm-1 and (b) 1700-500 cm-1 regions.
Figure 5. IR spectra of EC, β-CD, and IEC1 in (a) 3700-2800 cm-1 and (b) 1700-500 cm-1 regions.
can be classified as nonhydrogen bonding interactions (NHB) (3576 cm-1); and those from two types of hydrogen-bonded protons (HB), one from secondary O-H assigned to (3467 cm-1) and other as primary O-H stretching (3325 cm-1). The higher wavenumbers band in the infrared spectra here should result from secondary hydroxyl stretching since they possess relatively strong hydrogen-bonded interactions. This inference is consistent with the molecular electrostatic potential topography investigations by Pinjari et al.32 that revealed shallow minima near oxygens of secondary hydroxyls than those near primary hydroxyl oxygens in β-CD. Hence, the 3504 cm-1 band in experimental FTIR spectra should be assigned to secondary hydroxyl stretching. The 3336 cm-1 vibration in the observed spectra results from interactions between primary OH groups. Furthermore the ν(C-O-C) intraglucose vibration appearing at 1030 cm-1 has also been observed in the measured FTIR spectra. Likewise, the band near 950 cm-1 in the observed spectra compares well with the vibration at 975 cm-1, which was assigned to ν(C-C) + δ(C-H) mixed vibrational mode. To understand the effect of encapsulation of CA within the β-CD cavity on the normal vibration frequencies in isolated β-CD or guest, B3LYP-calculated infrared spectra of ICA1 have been compared with those of individual β-CD and CA shown in Figure 4. Thus, secondary hydroxyl (not participating in hydrogen bonding) stretching was assigned to 3576 cm-1 for β-CD, which has been noticed to be unchanged on complexation. The intense secondary hydroxyl 3467 cm-1 stretching of free β-CD yield a doublet at (3505, 3481 cm-1) and a near triplet, namely, (3471, 3442, 3439 cm-1) in ICA1, while primary
hydroxyls (3325 cm-1) engender a pattern consisting of four bands, all of which exhibit frequency upshift relative to that in free host and are found in the 3381-3341 cm-1 region. The interglucose (1134 cm-1) and intraglucose (1030 cm-1) C-O-C stretching frequencies of free β-CD are insensitive to encapsulation. A comparison with free CA further reveals that 1283 and 1212 cm-1 bands disappear in ICA1. Further, the interactions of ring II of CA that penetrates deeper within the host cavity manifest in appearance of new bands at 3236, 3053, and 3086 cm-1 in the infrared spectra of ICA1 complex, which is also accompanied by disappearance of bands at 3627 and 3574 cm-1. It may as well be remarked here that optimized structure of ICA1 suggests interaction between only one of ring I protons (O7H) and the host, whereas the O5H proton of CA does not facilitate any such interaction with secondary hydroxyls of β-CD. This has been manifested in a marginal shift of 5 cm-1 of the 3623 cm-1 band assigned to O5H stretching. The remaining 3609 cm-1 band from the O7H hydroxyls, however, downshifts to 3314 cm-1 in ICA1. Moreover, OH stretching of ring III in CA interacting with host cavity protons (H3 and H5) has been downshifted to 3233 cm-1 in the spectra. Encapsulation of EC within the β-CD does not influence secondary hydroxyl stretching significantly; accordingly, the 3577 cm-1 vibration in IEC1 is unchanged as compared to β-CD. The hydrogen-bonded secondary hydroxyl stretching in IEC1 split into three distinct vibrations at 3482, 3440, and 3362 cm-1 along with a near doublet (3450, 3453 cm-1). The degenerate 3325 cm-1 of
7730
J. Phys. Chem. A, Vol. 114, No. 29, 2010
Khedkar et al.
TABLE 3: Electron Density in Antibonding Orbital (σ* in au), Bond Distances (r in Å), and Frequency of Vibration (ν in cm-1) in CA, EC, ICA1, and IEC1 Complexesa CA -O7H -O5H -O3′H -O4′H -O3H a
EC
ICA1
ICE1
σ*
r
ν
σ*
r
ν
σ*
r
ν
σ*
r
ν
0.00808 0.00777 0.01186 0.00931 0.01099
0.970 0.969 0.973 0.969 0.970
3609 3623 3573 3627 3593
0.00892 0.00862 0.00934 0.01845 0.01098
0.970 0.970 0.973 0.969 0.973
3608 3610 3569 3623 3568
0.04415 0.00782 0.07060 0.07194 0.0111
0.984 0.969 0.998 1.001 0.972
3314 3618 3233 3236 3569
0.07088 0.08029 0.06780 0.01897 0.01257
0.999 1.003 0.996 0.975 0.972
3081 3002 3149 3532 3578
O7H, O5H from ring I; O3′H, O4′H from ring II; and O3H from ring III. See text for details.
Figure 6. The difference electron density mapped on bond critical point in (a-b): ICA1, (c-d): IEC1 in the range +0.00001 au (blue) to -0.00001 au (red). (Side and top view depicted.)
hydrogen-bonded primary OH stretching of β-CD on complexation with EC engenders four distinct vibrations: 3447, 3408, 3389, and 3378 cm-1. As noticed earlier for ICA1, the C-O-C stretching (interglucose and intraglucose) are unaltered in the IEC1 complex. A comparison of infrared spectra of IEC1 (cf. Figure 5) and isolated EC leads to the following inferences. It should be remarked here that interaction of EC with β-CD is qualitatively different from that of CA: with ring I protons interacting with primary hydroxyls while those of ring II are bound to secondary hydroxyl functionalities of β-CD. A frequency down shift (red shift) of ∼600 cm-1 for O5H and O7H stretching of ring I can also be noticed. These vibrations invoke strong coupling of
different internal coordinates. The hydroxyl (O4′H and O3′H) stretchings from ring II exhibit frequency downshifts in IEC1; the O3′H facilitats relatively strong interactions with secondary hydroxyl groups of β-CD, which engenders relatively large frequency down-shift in the spectra. Furthermore, the band arising from hydroxyls of ring III has been less influenced in the complex. Thus, unlike ICA1, relatively weak interactions between ring III and β-CD cavity protons can be inferred. A comparison of infrared spectra of ICA1 and IEC1 reveals that 904 cm-1 has been predicted only for IEC1. The ICA1 complex, on the other hand, shows a strongly coupled deformation vibration at 674 cm-1. A corresponding 660 cm-1 vibration can
Host-Guest Interactions with β-Cyclodextrin be noticed in isolated β-CD. Binding patterns of CA and EC to β-CD are qualitatively different and engender different patterns in its infrared spectra. To understand bond strength variation on encapsulation of the guest CA/EC we utilize the NBO approach. Accordingly, the electron density of the antibonding orbital (σ*) of -OH bonds of CA/EC and those in ICA1 and IEC1 is enhanced or depleted as shown in Table 3. The corresponding bond distances as well as the vibrational frequency have also been summarized. The direction of frequency shifts of OH stretching of CA/EC can be rationalized from electron density in its antibonding orbital. Thus, the -O3′H and -O4′H of ring II of CA/EC interacting with primary hydroxyls reveal increased electron density (by 0.06 au) in its antibonding (σ*) orbital. Similarly, the electron density in σ* orbital of O5H (ring I) is nearly unchanged and manifests accordingly (marginal downshift of 5 cm-1 as pointed out earlier) in its vibrational frequency. Use of difference electron density (∆F) to analyze host-guest interactions at the molecular level is outlined in the following. As pointed out in the preceding section, ∆F was calculated by subtracting the sum of electron densities of the host and guest from the corresponding electron density of ICA1 or IEC1. A mapping54 of ∆F on bond critical point in MED topography provides understanding of frequency shift in the calculated infrared spectra. Difference electron density (∆F) was mapped in the range of +0.00001 to -0.00001 au on the bond critical points in MED topography. The bcps are rendered with blue, where ∆F values turns out to be less than +0.00001 au, and the bcps with red are located for ∆F less than -0.00001 au. It may as well be remarked here that the (0.00001 au density values here refer to cut off values for the electron density data set utilized in devising the color mapping function. Smaller cutoff values bring about clear distinction of different bond critical points derived from the MED topography54 outlined in the preceding section. It is thus transparent that depletion of electron density can be inferred near O-H bonds of CA or EC in the red region (cf. Figure 6), signifying frequency downshift for corresponding stretching. A frequency upshift (blue) of C-H3/ C-H5 (within the host cavity) can also be rationalized similarly. Conclusions Systematic investigations on the electronic structure of the inclusion complexes of (+)-Catechin (CA) and (-)-epicatechin (EC) in β-CD have been carried out with the use of B3LYP density functional theory. It has been shown that the minimum energy complex ICA1, possess interactions between primary hydroxyl of CD and hydroxyls of ring II, whereas the secondary hydroxyl facilitates hydrogen bonding with ring I and leaves H6 proton of the guest outside the host cavity. On the other hand, encapsulation of EC within β-CD engenders O-H · · · O interactions between ring I and primary -OH groups of β-CD and ring II being bound to secondary OH groups of the host. Different patterns of binding of CA and EC manifest in vibrational spectra of ICA1 and IEC1. The 904 cm-1 band was predicted in IEC1 is not present in ICA1, while the 674 cm-1 band only in ICA1 corresponds to 660 cm-1 vibration in β-CD. The frequency shifts of characteristic vibrations in the complex relative to free or isolated guest or the host thus have been rationalized with use of natural bond orbital analyses and combining difference electron density with MED topography. Acknowledgment. The authors thank the Center for Network Computing, University of Pune, for providing computational facilities. J.K.K. thanks UGC for the award of meritorious
J. Phys. Chem. A, Vol. 114, No. 29, 2010 7731 student fellowship. R.V.P. is grateful to the Council of Scientific and Industrial Research, New Delhi, India for a Senior Research fellowship. S.P.G. acknowledges support from the University Grants Commission (UGC), New Delhi, India (Research Project F34-370/2008(SR)). Supporting Information Available: B3LYP/6-31G(d) optimized geometries of ICA and IEC complexes. Vibrational spectra of CA, EC, β-CD, ICA1, and IEC1. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Szeijtli, J. Chem. ReV. 1998, 98, 1743. (2) Dupuya, N.; Barbryb, D.; Briac, M.; Marquisd, S.; Vrielyncka, L.; Kistera, J. Spectrochim. Acta, Part A 2005, 61, 1051. (3) Bender, M. L.; Komiyama, M.; Springer-Verlag: Berlin, 1978. (4) Saenger, W. Angew. Chem., Int. Ed. Engl. 1980, 19, 334. (5) Szejtil, J. Cyclodextrin Technology; Kluwer: Dor-drecht, 1988. (6) Radecka, H.; Radecki, J. Anal. Lett. 2001, 12. (7) Loftsson, T.; Jarvinen, T. AdV. Drug DeliVery ReV. 1999, 36, 59. (8) Al-Shihry, S. S. Spectrochim. Acta Part A 2005, 61, 2439. (9) Cunha-Silva, L.; Teixeira-Dias, J. J. C. New J. Chem. 2004, 28, 200. (10) Vrielynck, L.; Lapouge, C.; Marquis, S.; Kister, J.; Dupuy, N. Spectrochim. Acta Part A 2004, 60, 2553. (11) Szejtli, J. J. Med. Chem. ReV. 1994, 14, 353. (12) Szejtli, J. J. Incl. Phenom. 1992, 14, 25. (13) Duchene, D.; Wouessidjew, D. J. Coord. Chem. 1992, 27, 223. (14) D’Souza, V. T. Supramol. Chem. 2003, 15, 221. (15) Wenz, G. Angew. Chem., Int. Ed. Engl. 1994, 33, 803. (16) Li, S.; Purdy, W. C. Chem. ReV. 1992, 92, 1457. (17) Schneiderman, E.; Stalcup, A. M. J. Chromatogr. B 2000, 745, 83. (18) Takahashi, K.; Hattori, K. J. Incl. Phenom. 1994, 17, 1. (19) Breslow, R.; Dong, S. D. Chem. ReV. 1998, 98, 1829. (20) Cao, F.; Ren, Y.; Hua, W.-Y.; Ma, K.-F.; Guo, Y.-L. Chin. J. Org. Chem. 2002, 22, 827. (21) Cervello´, E.; Jaime, C. J. Mol. Struct. (Theochem) 1998, 428, 195. (22) Madrid, J.; Paozuelo, J.; Mendicuti, F.; Mattice, W. L. J. Colloid Interface Sci. 1997, 193, 112. (23) Nascimento, C. S., Jr.; Dos Santos, H. F.; De Almedia, W. B. Chem. Phys. Lett. 2004, 397, 422. (24) Nascimento, C. S., Jr.; Cleber, P. A.; Dos Santos, H. F.; De Almedia, W. B. J. Phys. Chem. A 2005, 109, 3209. (25) Yan, C.; Xiu, Z.; Li, X.; Hao, C. J. Mol. Graph. Mol. Model. 2007, 26, 420. (26) El-Barghouthi, M. I.; Jaime, C.; Al-Sakhen, N. A.; Issa, A. A.; Adboth, A. A.; Al Omari, M. M.; Badwan, A. A.; Zughul, M. B. J. Mol. Struct. (THEOCHEM) 2008, 853, 45. (27) Inoue, Y.; Liu, Y.; Tong, L.-H.; Shen, B.-J.; Jin, D.-S. J. Am. Chem. Soc. 1993, 115, 10637. (28) Huang, M. J.; Watts, J. D.; Bodor, N. S. Int. J. Quantum Chem. 1997, 64, 711. (29) Kaanumalle, L. S.; Gibb, C. L. D.; Gibb, B. C.; Ramamurthy, V. J. Am. Chem. Soc. 2005, 127, 3674. (30) Smit, B.; Maesen, T. L. M. Nature 2008, 451, 671. (31) Pinjari, R. V.; Khedkar, J. K.; Gejji, S. P. J. Incl. Phenom. Macrocycl. Chem. 2010, 66, 371. (32) Pinjari, R. V.; Joshi, K. A.; Gejji, S. P. J. Phys. Chem. A 2007, 111, 13583. (33) Pinjari, R. V.; Joshi, K. A.; Gejji, S. P J. Phys. Chem. A 2006, 110, 13073. (34) Yanagida, A.; Shoji, A.; Shibhbusawa, Y.; Shindo, H.; Tagashira, M.; Ikeda, M.; Ito, Y. J. Chromatogr. A. 2006, 1112, 195. (35) Harborne, J. B.; Mabry, T. J.; Mabry H. The Flavonoids ed 1975. (36) Katiyar, S.; Elmets, C. A.; Katiyar, S. K. J. Nutrit. Biochem. 2007, 18, 287. (37) Gradisar, H.; Pristovsek, P.; Plaper, A.; Jerala, R. J. Med. Chem. 2007, 2, 264. (38) Fujiki, H. Pharmacia 1998, 34, 223. (39) Yamada, M. Chem. Chem. Ind. 1998, 51, 582. (40) Ishizu, T.; Kintsu, K.; Yamamoto, H. J. Phys. Chem. B 1999, 103, 8992. (41) Smith, V. K.; Ndou, T. T.; Warner, I. M. J. Phys. Chem. 1994, 98, 8627. (42) Kriz, Z.; Koca, J.; Imberty, A.; Charlot, A.; Auzely-Velty, R. Org. Biomol. Chem 2003, 1, 2590. (43) Jullian, C.; Montecinos, J. M.; Torres, G. Z.; Aguilera, B.; Rodriguez, J.; Aran, V.; Olea-Azar, C. Bioorg. Med. Chem. 2008, 16, 5078.
7732
J. Phys. Chem. A, Vol. 114, No. 29, 2010
(44) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (45) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785. (46) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R. G.; Calmani, V.; Barone, B.; Mennucci, G. A.; Petersson, H.; Nakatsuji, M.; Caricato, Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A.; Peralta, Jr., J. E. F.; Ogliaro, M.; Bearpark, J. J.; Heyd, E.; Brothers, K. N.; Kudin, V. N.; Staroverov, R.; Kobayashi, J.; Normand, K.; Raghavachari, A.; Rendell, J. C.; Burant, S. S.; Iyengar, J.; Tomasi, M.; Cossi, N.; Rega, J. M.; Millam, M.; Klene, J. E.; Knox, J. B.; Cross, V.; Bakken, C.; Adamo, J.; Jaramillo, R.; Gomperts, R. E.; Stratmann, O.; Yazyev, A. J.; Austin, R.; Cammi, C.; Pomelli, J. W.; Ochterski, R. L.; Martin, K.; Morokuma, V. G.; Zakrzewski, G. A.; Voth, P.; Salvador, J. J.; Dannenberg, S.; Dapprich, A. D.; Daniels, O.; Farkas, J. B.; Foresman, J. V.; Ortiz, J.; Cioslowski Fox, D. J. Gaussian 09; Gaussian, Inc.: Wallingford CT, 2009.
Khedkar et al. (47) Limaye, A. C.; Gadre, S. R. Curr. Sci. 2001, 80, 1298. (48) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. Re. 1988, 88, 899. (49) Bader, R. F. W. Atoms in Molecules: A Quantum Theory; Oxford University Press: Oxford, UK, 1990. (50) Matta, D. F.; Boyd, R. J. The Quantum theory of Atoms in Molecules. In An Introduction to the Quantum Theory of Atoms in Molecules; Matta, D. F., Boyd, R. J., Eds.; Wiley-VCH: Weinheim, 2007; pp 1-34. (51) Balanarayan, P.; Gadre, S. R. J. Chem. Phys. 2003, 119, 5037. (52) Anthony, P.; Scott, Leo R. J. Phys. Chem. 1996, 100, 16502. (53) Egyed, O. Vibra. Spectro. 1990, 1, 225. (54) Pinjari, R.; Joshi, K.; Gejji, S. P. Spectrochim. Acta, Part A 2008, 71, 1056.
JP102304J
