Correlation between Structure and Shape of the Polarized Infrared

Mar 3, 2011 - Polarized infrared spectra of liquid crystalline 4-chloro-2′-hydroxy-4′-alkyloxyazobenzenes (CHAB) with C5, C6, and C7 alkyl chains ...
0 downloads 0 Views 4MB Size
ARTICLE pubs.acs.org/JPCB

Correlation between Structure and Shape of the Polarized Infrared Absorption Spectra of 4-Chloro-20-hydroxy-40-alkyloxyazobenzenes Paulina Majewska, Maria Rospenk, Boguszawa Czarnik-Matusewicz,* and Lucjan Sobczyk Faculty of Chemistry, University of Wroczaw, F. Joliot-Curie 14, 50-383 Wroczaw, Poland ABSTRACT: Polarized infrared spectra of liquid crystalline 4-chloro-20 -hydroxy-40 -alkyloxyazobenzenes (CHAB) with C5, C6, and C7 alkyl chains measured at 25 C were compared, with particular attention being paid to the influence of chain length on formation of ordered boundary layers. The effect of chain elongation is discussed on the basis of calculations of the optimized geometry of dimeric species, which reflects the role of association in ordered phases. The impact of the varying intermolecular interactions on the spectra of C5, C6, and C7 derivatives measured as a function of the polarization angle is analyzed by principal component analysis (PCA). The polarization-angle-dependent results are contrasted with the transition moment directions obtained from the density functional theory (DFT) calculations. The continuum spread down to 500 cm-1 and a pseudoband at ca. 600 cm-1 ascribed to the Fermi resonance between ν(OH) and γ(OH) modes is an additional phenomenon whose features also depend on alkyl chain length.

1. INTRODUCTION An important subject concerning liquid crystalline (LC) molecules is their alignment at a surface boundary. Many surface treatments that promote the alignment of LC molecules require the application of specific mechanical or chemical processes that modify the substrate.1 An attractive alternative is the photoalignment process, where the reorientation is induced by lightcontrolled changes in the configuration at the molecular level.2 Under the experimental conditions described here, the LC molecules due to their anisotropic properties undergo the selfalignment process at two boundary surfaces, that is, at the KRS-5 windows of the infrared cell. The anchoring strength is controlled by anisotropic intermolecular interactions between the substrate and LC molecules at the boundary regions where the LC phase makes contact with a surface.3,4 The subtle balance in the interactions between LC molecules in a bulk phase and in boundary layers determines the efficiency of the alignment. Studies that show how changes in the properties of the surface and/or in physicochemical properties of LC molecules affect the process are very attractive due to possible application of the process in construction of LC display devices. An analysis of 4-chloro-20 -hydroxy-40 -alkyloxyazobenzenes (CHAB) with C5, C6, and C7 alkyl chains that show liquid crystalline properties5-8 allows the correlation between chain length and vertical and horizontal anchoring strength to be determined. Variations in the balance between them lead to remarkable changes in the orientation direction of the LC molecules that is efficiently probed by a linearly polarized infrared light. The compounds are also the focus of the present paper because the OH group in position 20 to the 4-chloro-40 -alkyloxyazobenzene stabilizes the mesophases and broadens their existence range.7 On the one hand, the OH group forms r 2011 American Chemical Society

intramolecular hydrogen bonds and a chelate ring that has caused considerable interest from the point of view of resonance-assisted hydrogen bonds (RAHB)9-11 and, at the same time, of lowbarrier hydrogen bonds (LBHB).12 Such hydrogen bonds have several properties, the most interesting of which is the formation of continua in the infrared absorption spectrum.9 The intensity of such continua is much less as compared with usual hydrogen bonds without the resonance effect.13 On the other hand, the ordering of molecules in the crystalline or mesophases leads to a considerable increase in the intensity of the continuum spread down to 400 cm-1. The creation of a band at ca. 620 cm-1 with a half-width of ca. 200 cm-1 was simultaneously found in the case of 40 -hexyloxyazobenzene.5 This pseudoband, according to the Fermi resonance approach,14 takes place only in ordered phases due to the coupling of ν(OH) with γ(OH) vibrations, which leads to creation of a deep broad minimum and gives a picture of an independent broad band at ca. 600 cm-1. This problem seems to be most important in our case. Thus we would like to throw some light on the intensity and shape of the low-frequency wing of the broad absorption in the range of ca. 600 cm-1. It seems interesting that the low-frequency wing ascribed to the chelate ring of 40 -hydroxy-40 -alkyloxyazobenzenes is affected by the length of the alkyloxy chain. Thus, in the case of the pentyloxy derivative, the pseudoband at ca. 600 cm-1 is not observed at all. It appears for a hexyloxy derivative and is particularly clear for a heptyloxy derivative. It seems evident that an increase in the intensity of the continuum is due to the mutual Received: August 18, 2010 Revised: January 31, 2011 Published: March 03, 2011 2728

dx.doi.org/10.1021/jp107826x | J. Phys. Chem. B 2011, 115, 2728–2736

The Journal of Physical Chemistry B

ARTICLE

)

)

following steps, done in a strictly controlled manner. At the beginning, the sample was heated to a temperature of 120 C. Next, by capillary action (CA) the isotropic liquid was introduced into the cell of thickness about 20 μm. The encapsulated LC sample was cooled down to 25 C and warmed up to the isotropic phase. This cycle was repeated a few times until the infrared spectra measured at 25 C did not change between cycles. Finally, the cell was mounted in front of a polarizer. The experimental setup enabled the parallel orientation between E and the directions of filling flow to be preserved for each sample. The scheme of the system is presented in Figure 1. To simplify the general procedure, it was assumed that for the mode i the long axis of the molecule, L, is parallel to the transition dipole moment, M. During the measurements the values of the elevation (vertical) angle j and the azimuthal (horizontal) angle θ were fixed and only the angle of polarization was changed from the parallel E to the perpendicular E^ orientation with respect to the plane of incidence. The quantum chemical calculations at B3LYP and MP2 levels, using the 6-31G(d,p) basis set tending toward geometry optimizations of isolated monomers and dimers and corresponding vibrational spectroscopic data, were performed with the GAUSSIAN program.16 For each derivative the calculations for dimers were initiated from parallel orientation. The optimized structures discussed in a later section are fully geometry-optimized and correspond to the global minimum. Despite high flexibility of the aliphatic chains, we failed in searching for other local minima. For other examined geometries of the chains and relative orientations of the aromatic cores, negative frequencies were obtained. For vibrational spectroscopic data of aromatic molecules, DFT results contrary to the MP2 ones are in very good agreement with experimental frequencies that were discussed in detail elsewhere.17-19 Therefore, for more detailed insight into the direction of the transition dipole moments with respect to the long axis of molecules, additional calculations were performed only for B3LYP results by use of the PQS program.20 Data processing to remove the fluctuating baseline and the calculation of integrated absorbances were accomplished with the GRAMS program.21

)

Figure 1. Definitions of geometry of the most important constituents of the system: CA, capillary action direction; L, long axis of the LC molecule; Mi, transition moment for transition i that is parallel to the axis L; θ and j, two angles that determine the horizontal and vertical anchoring, respectively, of the LC molecule to the KRS-5 surface; E and E^, two limited orientations (parallel and perpendicular) of the electric field vectors with respect to the plane of incidence that, by definition, is perpendicular to the surface of the sample. The vector quantities are denoted by boldface type.

interaction of chelate rings, which are characterized by a high polarizability.15 A most favorable conformation of neighboring rings for which the intensity is maximal probably exists. A maximal spreading of the continuum toward low frequency should be accompanied by its intensity. A careful insight to the structure of dimeric species is provided by a calculation of dimeric structures, which should reflect the most profitable conformation of associates.

2. EXPERIMENTAL AND CALCULATION METHODOLOGY The preparation of samples and a description of results for 5and 6-CHAB derivatives have been presented in previous papers.5,6 Analogous to those studies, the same preparation procedure was applied in measurements of polarized IR spectra for the heptyloxy derivative 7-CHAB in a crystal state that is the subject of the present study. However, two facts are worth mentioning once again. First, the windows of the KRS-5 cell were not patterned by any procedure. Second, even small changes in physicochemical properties of the system might be reflected in the orientation. As was confirmed by us, the equilibrated arrangement of LC molecules depends on the filling conditions. Therefore, to get equilibrium between the boundary layers and the bulk phase, the KRS-5 cell was filled in the

3. RESULTS AND DISCUSSION 3.1. Geometry of the System Being Analyzed. Whenever capillary forces initiate the filling of a thin cell with LC molecules of elongated shape, they tend to orient their long axes along the direction of the hydrodynamic flow. Due to the surface-stabilized action at 25 C, ordered boundary layers are created in close proximity to the KRS-5 windows. Because their surface was not pretreated in any way, the total thickness of the ordered boundary layers does not reach the distance between the windows, and between them there is a bulk phase composed of less ordered LC molecules. Thus, the overall orientation depends on the relative thickness of the two differently ordered layers and the strength of the anchoring of the LC molecules by the KRS-5 surface in the boundary layers. As has already been shown on an example of LC molecules that belong to CHAB5-7 and other chemical classes,22-24 polarized infrared spectroscopy is very suitable to analyze the ordering properties. As can be seen in Figure 2, five orientations of LC molecules in the boundary layers are possible, depending on the balance between the intermolecular LC forces and surface anchoring (surface-LC) forces: 2729

dx.doi.org/10.1021/jp107826x |J. Phys. Chem. B 2011, 115, 2728–2736

The Journal of Physical Chemistry B

ARTICLE

Figure 2. Five possible orderings of LC molecules between the two boundary layers along with information about the relative orientation of the vectors defining the system. The ordered LC molecules at the surface of the KRS-5 windows are presented as ellipses of different shapes depending on the orientation. Shadowed areas in the middle between the KRS-5 windows symbolize the bulk, unordered phase.

)

values cos2(Mi, E) were calculated for the system as a function of varying orientation of the E vector. The runs show that with an increase in the elevation angle j the amplitude of absorbance decreases, which leads to smaller values of dichroic ratio R, defined as R = A /A^. For j = 90 the dichroism for the band characterized by Mi parallel to L drops to 0, which corresponds to the case from Figure 2e. Changes of the azimuthal angle θ do not change the R values but shift the position of the absorbance maximum to the θ value. An analysis of the experimental absorbance changes as a function of the polarization angle according to the investigations presented above will allow differences in the ordering phenomenon caused by elongation of the alkyl chain to be identified. On the basis of the PQS results for the C5, C6, and C7, monomers Mi of the mode dominated by the δ(OH) vibration, absorbing at 1624 cm-1, and L form angles of 6, 2, and 4, respectively. According to the calculations, an almost perfect perpendicular orientation characterizes the mode assigned to the γ(OH) vibration at about 840 cm-1. Therefore, from the polarized spectra of the three alkyloxy derivatives shown in Figure 4, the absorbance variations at the two frequencies were analyzed by means of the polar plots shown in Figure 5. The amplitude of absorbance changes for the δ(OH) vibration is largest for the C5 derivative because the calculated angle is not the minimum but the alkyl chain is the shortest one. The run for the C6 derivative has about the same shape, whereas for the C7 derivative the spread of absorbance changes is almost 2 times smaller. It seems that for the C5 and C6 derivatives the LC molecules are more planar anchored to the KRS-5 surface than the C7 compound for that increase in the elevation angle. The fact that it is not so obviously shown by the polar plot done for the γ(OH) vibration is discussed in a later section of this paper. The information derived from Figure 5a corresponds very well with the optimized geometries obtained in the DFT calculations done for the dimers. The calculated structures presented in Figure 6 demonstrate that as the alkyloxy chain elongates, the deviation from parallel orientations of the core rings increases. The heptyloxy derivative shows a particularly remarkable effect. We tried to look for some )

)

(a) The LC molecules are planar anchored to the KRS surfaces, and the L vector and E component are parallel to each other. This relates to a homogeneous alignment. (b) The LC molecules remain planar anchored to the KRS surfaces, but their long axes are no longer oriented in parallel with respect to the direction of capillary force action. The azimuthal angle θ is different from 0 but less than 90. (c) After the molecules are loaded between the two KRS surfaces, they undergo further in-plane surface-induced ordering relaxation. The molecules are still planar anchored to the surfaces but the L vector and E component are perpendicular to each other. For the azimuthal angle θ equal to 90, L is parallel with the E^component. In the three cases (a, b, and c) due to the dominant surface anchoring forces, there is planar alignment. (d) Due to the surface-induced ordering relaxation, not only the θ angle but also the elevation (vertical) angles j are changed. The LC molecules are no longer planar anchored. The two θ and j angles determine their orientation with respect to the window surface. Molecules are aligned neither homogenously nor homeotropically due to the increased role of the intermolecular LC forces. (e) The homeotropic alignment when the j angle is 90 is very unlikely. Such an alignment is realized when the intermolecular LC forces are larger than the surface-LC ones. For any polarization there is always a normal angle between E and L directions. Thus all bands characterized by M parallel to L will not show any dependence on the polarization angle. In polarization spectroscopy, absorbance A at the frequency assigned to mode i is proportional to the average over the entire ensemble of molecules, the cosine squared of the angles between transition moment of the mode i and the electric vector E;, that is, A ∼ Æcos2 (Mi, E)æ.25 Figure 3 shows how the changes in the orientation of the samples might influence the absorbance. For the purposes of this investigation a set of 1000 molecules whose Mi vectors follow a normal distribution with respect to the elevation j and the azimuthal θ angles was created. Next, the

2730

dx.doi.org/10.1021/jp107826x |J. Phys. Chem. B 2011, 115, 2728–2736

The Journal of Physical Chemistry B

ARTICLE

Figure 3. Results of a simulation that shows the influence of variations in (a) horizontal and (b) vertical orientation of the Mi vector on the absorbance of the transition that is proportional to cos2 (E  Mi) (panel d). (c) Spread of the orientation in a sample frame. The simulation was based on a set of 1000 Mi vectors normally distributed around θ = 15 and j = 30 with the same standard deviation equal to 10.

)

Figure 4. Infrared spectra of (a) 5-CHAB, (b) 6-CHAB, and (c) 7-CHAB at 25 C as a function of the polarization angle rotation. Arrows indicate direction of absorbance changes at 1624 and 840 cm-1 with incremental rotation of E from E to E^ with a 5 step.

quantitative relationship between geometrical parameters of the calculated dimers and spectroscopic behavior of studied alkyloxy derivatives. Selected distances and angles that show an increased opening of the dimeric structures are compared in Table 1. The

angle between planes defined by the aromatic rings is the most characteristic. In the case of the C7 derivative, this angle is much larger as compared with C5 and C6 derivatives. The second feature concerns the distances between the terminal carbon 2731

dx.doi.org/10.1021/jp107826x |J. Phys. Chem. B 2011, 115, 2728–2736

The Journal of Physical Chemistry B

ARTICLE

Figure 5. Polar plots of the relative changes in absorbance in the regions dominated by (a) δ(OH) mode at 1624 cm-1 and (b) γ(OH) mode at 838, 834, and 849 cm-1 for 5-CHAB (red circles), 6-CHAB (green circles), and 7-CHAB (blue circles) derivatives.

Table 1. Selected Distances between Monomers, Measured for Aromatic Terminal Carbons, and Angles between Core-toCore Planes in Calculated Dimers of C5, C6, and C7 at B3LYP Level distances: (Cl-C T C-O), Å

angles, deg

C5

5.47

6.47

13.6

C6 C7

5.39 5.32

6.23 8.19

22.8 72.9

Table 2. Dichroic Ratios (R) Calculated for the δ(OH) Vibration at 1624 cm-1 phase

Figure 6. Calculated structures of dimeric species of (a) 5-CHAB, (b) 6-CHAB, and (c) 7-CHAB.

atoms bond with the chlorine atom of one monomer and with the alkyloxy oxygen atom of the second monomer. Data for 7-CHAB differ substantially from both remaining derivatives due to the twisted orientation of the two monomers. The data in Table 1 and structures in Figure 6 concern only B3LYP results, as those from MP2 calculations are almost the same. Such a tendency to asymmetrization of a system is in accordance with the polarized infrared spectra. As Table 2 shows the R value for the heptyloxy derivative is much smaller as compared with those for the pentyloxy and hexyloxy derivatives. It can suppose some twist of the long axes of molecules, and at the same vectors of the transition moments have a widespread distribution in mesophases. As a consequence this leads to a decrease of dichroism

solid

smectic A

nematic

isotropic

C5

3.75

3.77

3.15

1.98

C6

3.60

3.57

3.08

1.94

C7

1.47

1.48

1.40

1.35

due to an increase in the value and distribution of the elevation angle. This parameter for the three derivatives obtained from absorbance at 1624 cm-1, measured in solid, smectic A, nematic, and isotropic phases, is compared in Table 2. The comparison performed for the same thickness of the KRS-5 cell close to 20 μm reveals the following regularities. The R values for C5 and C6 derivatives are quite similar in all of the phases investigated, with C6 being a little smaller as compared with C5, whereas for the C7 compound there is a drastic decrease of R, mainly for the ordered phases. This confirms the supposition that elongation of the alkyl chain length causes a drop in the ordering degree due to the increased elevation angle that is most spectacularly reflected for C7. The second point is the fact that the dichroism for crystalline and smectic phases is almost identical and is slightly smaller for the nematic phase. This shows that transition from the smectic A to the crystalline phase is just a freezing of the ordering specific for the smectic phase. This justifies the use of the phrase “surface smectic phase” to describe the boundary layers at 25 C. A peculiar observation is that for the isotropic phase at temperatures a few degrees above the N/iso transitions a nonzero ordering is observed when R is greater than 1, which confirms 2732

dx.doi.org/10.1021/jp107826x |J. Phys. Chem. B 2011, 115, 2728–2736

The Journal of Physical Chemistry B

ARTICLE

Figure 7. Scores plot against PC1 and PC3 obtained from PCA done for 5-CHAB (red circles), 6-CHAB (green circles), and 7-CHAB (blue circles) data combined into one set.

that above the bulk nematic-isotropic transition temperature a residual nematic order persists near the interface.26 By analogy to the “surface smectic phase”, the boundary layers formed on the KRS-5 windows above TN/iso should be called the “surface nematic phase”. According to our observations the R values dropped to 1 at temperatures 10 above the N/iso transition, which means a total decay of the “surface nematic phase”. Our studies of the C5 derivative for various thicknesses of the cell have shown that for thinner cells the system is more ordered,5 which correlates very well with R calculated from absorbance at 1624 cm-1 for 25 C. Cells whose thickness decreases from 20 to 16 to 11 μm have increasing R values of 3.75, 3.95, and 4.05, respectively. The increase of R is a consequence of a reduced width of the bulk phase for thinner cells. The width of the ordered boundary layers is mainly controlled by the surface-LC forces and not by the distance between the KRS-5 windows. The forces are dominated by the surface tension of LC molecules and the surface27 and the interfacial tension between the two components, a physical property that is independent of film thickness.28 Moreover, the terminal molecular groups of LC molecules exhibit an extremely large effect on the surface tension.29 All of these phenomena confirm the specific behavior of LC molecules close to the surface of the cell windows. 3.2. PCA Results. An analysis of polarized infrared spectra and calculations shows how substantial changes proceed in the behavior when the elongation of the alkyl chain going from C5 to C7 occurs. Let us look again at the polarized spectra of the three alkyloxy derivatives shown in Figure 4. One can see the somewhat different behavior of the heptyloxy derivative 7-CHAB as compared with the other two. The difference consists first of the weak dependence of the absorbance on the polarization angle and, second, of the increase in intensity for the pseudoband at ca. 600 cm-1, which is interpreted as a result of the Fermi resonance between the ν(OH) and γ(OH) modes. A weak dependence of absorbance on the polarization angle may be due to two effects,

namely, to a weaker tendency to orient in the boundary layers or to a much larger distribution of the orientation of long axes of molecules as was discussed above. An important result of the continuous absorption registered for the ordered phase and the CCl4 solution was presented in a previous paper6 with 6-CHAB as an example. Similar behavior has been observed for C5 and C7 derivatives. Without a doubt the intensity of the continuum in the CCl4 solution is markedly smaller than in the condensed phases. The calculated structures of dimers for all three derivatives shown in Figure 6 may provide help in solving the problem. They reflect the tendency to form associates with a particular mutual orientation of molecules. To get a more detailed specification of the influence of elongation of the alkyl chain on the ordering properties of CHAB in the condensed phase, principal component analysis (PCA) was employed. It is one of the most common chemometrics methods for the classification of objects being analyzed against variables used for their discrimination.30 As was already shown, this method can be very efficient in an analysis of the spectral properties of liquid crystalline systems.31 First, PCA calculations were performed for all of the data combined into one file with the aim of distinguishing differences between the three systems. Figure 7 shows the distribution of samples against the PC1 and PC3 components as the two components are more strongly correlated with the intra- and intersystem properties. The intrasystem spectral variations are hidden behind the first component that captures the largest amount of spectral changes, that is, 51.73%. The changes are a consequence of a varying polarization angle. An analysis of the absorbance as a function of the angle for an individual system will be facilitated when the PCA is done independently for each system; whereas for the combined system the third component allows the C7 sample that scores have positive values to be separated from the C6 and mainly the C5 sample characterized by negative scores versus PC3. 2733

dx.doi.org/10.1021/jp107826x |J. Phys. Chem. B 2011, 115, 2728–2736

The Journal of Physical Chemistry B

ARTICLE

Figure 8. Loadings against PC3 obtained from PCA done for data combined into one set. Regions attributed to the OH vibrations that are significant for 5-CHAB and 7-CHAB are shaded in light and dark gray, respectively.

Figure 9. Scores plots against PC1 obtained from PCA done for the separate sets: 5-CHAB (red squares), 6-CHAB (green circles), and 7-CHAB (blue diamonds). Dashed lines present cos2 (Mi, E) functions fitted to the scores values.

According to the scores plot, large positive loadings versus PC1 are observed for the transitions that are perfectly aligned with E , whereas large negative loadings indicate the transitions that are parallel to E^. The positive loadings are attributed to in-plane bending vibrations and stretching vibrations arising from the core part of the molecules, whereas the negative loadings characterize bands assigned to the out-of-plane bending vibration of OH, CH2(aliphatic), and CH (aromatic) groups. For each homologue there is a very strongly marked border between the positive and negative values at about 980 cm-1. The loading profiles have been compared with the angles between the L and M vectors, — (M, L), reconstructed from the Hessian matrix by use of PQS software. Only the transitions where the theoretical intensity was larger than one unit were chosen for comparison. The insets in Figure 8, as polar plots, are the angle histogram for the — (M, L)

)

Analysis of the loadings plot for the third component from Figure 8 allows the most characteristic ranges of absorbance responsible for the classification revealed by the scores plot to be selected. Important differences in the hydrogen-bonding properties of the three systems are disclosed by the third component. Among the many spectral features classified by this latent variable are the broad absorbance changes attributed to the ν(OH) stretching vibration characterized by negative loadings and those assigned to the γ(OH) bending vibration involved in Fermi resonance with the ν(OH) stretching vibration displayed by positive loadings expanding from 1000 to 500 cm-1 with the characteristic Evans hole at ∼890 cm-1. Due to this phenomenon, the absorbances read at about 840 cm-1 and presented in Figure 5b do not follow the same pattern of changes as that shown in Figure 5a. The broad absorbance from the ν(OH) vibration with a maximum around 2800 cm-1 is strongest for the C5 sample, whereas the C7 sample has a more pronounced broad absorbance below 1000 cm-1. The changes prove that elongation of the alkyl chain on one side leads to a less homogeneous ordering and on the other side facilitates the formation of stronger hydrogen bonds. Special attention should be paid to the results concerning the orientation of the molecules obtained from PCA done independently for each system. Here, the information on the orientational ordering that reveals the scores plotted in Figure 9 comes from all of the changes in absorbances developed in the range from 3550 to 480 cm-1. In this approach the first component captures the changes correlated with direction of the E vector almost perfectly. Similarly, as was discussed for Figure 7, the C5 system is the most ordered. Elongation of the alkyl chain by one methylene group causes only a slight decrease in the ordering. Elongation by two methylene units is manifested by a drastic loss of ordering. Let us now look at the loading plots in Figure 10, which present the distribution of the contribution of absorbance at each wavenumber for the obtained classification of samples.

2734

dx.doi.org/10.1021/jp107826x |J. Phys. Chem. B 2011, 115, 2728–2736

The Journal of Physical Chemistry B

ARTICLE

Figure 10. Loading against PC1 obtained from PCA done for three separate sets: 5-CHAB (red), 6-CHAB (green), and 7-CHAB (blue). The angle histograms reflect the distribution of the angles of the Mi directions according to their value. The transitions from 1800 to 1000 cm-1 are marked in orange on the right polar plots; the transitions from 1000 to 450 cm-1 are marked in violet. The polar plots in gray on the left identify the C-H aliphatic and aromatic transitions above 3000 cm-1.

continuous absorption in the low-frequency range. The polar plots of relative changes in absorbance in the regions of the δ(OH) and γ(OH) modes for the three derivatives C5, C6, and C7 show similar relationships for C5 and C6 and a markedly different one for C7. The same conclusion can be drawn from the scores plots against PC1 and PC3 obtained by principal component analysis (PCA). The scores values for the three homologues show that the C5 derivative is the most ordered. The addition of one methylene group causes only a small decrease in ordering. Elongation of the alkyl chain by two methylene units is manifested by a remarkable loss of ordering. The question is which factors cause such drastic differences in the behavior of a homologous series of three compounds that differ little in the length of the alkyloxy chain. It seems that such a spectacular difference arises from the proportion between this length and the dimension of the rigid two-ring core. We are dealing with a situation where in the C6 compound the alkyloxy chain length is close to the dimension of the two-ring core, while in the case of the C5 compound this length is somewhat shorter, and in the case of C7 is somewhat longer, than the length of the core.

4. CONCLUSIONS Elongation of the alkyl chain in liquid crystalline 4-chloro-20 hydroxy-40 -alkyloxyazobenzenes C5, C6, and C7 leads to conspicuous changes in the spectroscopic behavior and self-organization in the condensed phases. In infrared spectra the continuous absorption is spread down to low frequencies of ca. 500 cm-1 with a characteristic pseudoband at ca. 600 cm-1, which is ascribed to the Fermi resonance between the ν(OH) and γ(OH) modes. The intensity of this pseudoband is highest for the heptyloxy derivative, which is due to the highest intensity of the

’ AUTHOR INFORMATION

)

)

values. There is a very good convergence between the angles pointed by the loadings values and those obtained from the DFT calculations. As is shown in Figure 10, there is a set of three peaks that have large positive values below 980 cm-1 for each derivative, despite the fact that this range is dominated by a strong, broad band attributed to the γ(OH) transition involved in the Fermi resonance with the ν(OH) transition, mainly for the C6 and C7 systems. According to the DFT calculations, these three peaks arise from the δ(C-C) vibrations that incorporate the aliphatic chain, the first, and the second aromatic ring. The — (M, L) values for the three transitions are in the range from 4 to 12, whereas for the γ(OH) transition it is 90. The large negative loading values attributed to this transition reveal that its involvement in the Fermi resonance did not change — (M, L) of the γ(OH) transition. An analysis of the distribution of — (M, L) for the CH stretching modes shows that most of them have an orientation perpendicular to E . The DFT results reveal that only Mi for the νsym(CH3) vibration characterizes an orientation parallel to E . The results reveal very good agreement between the calculated and experimentally determined IR transition moment directions, which is not a trivial problem.32

Corresponding Author

*E-mail [email protected]; fax þ48 71 3282348; tel þ48 71 3757293.

’ ACKNOWLEDGMENT Calculations were carried out at the Wroczaw Centre for Networking and Supercomputing (http://www.wcss.wroc.pl). 2735

dx.doi.org/10.1021/jp107826x |J. Phys. Chem. B 2011, 115, 2728–2736

The Journal of Physical Chemistry B

’ REFERENCES (1) Jer^ome, B. Rep. Prog. Phys. 1991, 54, 391. (2) Ichimura, K. Chem. Rev. 2000, 100, 1847. (3) Cognard, J. Mol. Cryst. Liq. Cryst. Suppl. Ser. 1982, 1, 1. (4) Dadmun, M. D.; Muthukumar, M. J. Chem. Phys. 1994, 101, 10038. (5) Rospenk, M.; Majewska, P.; Czarnik-Matusewicz, B.; Sobczyk, L. Chem. Phys. 2006, 326, 458. (6) Majewska, P.; Rospenk, M.; Czarnik-Matusewicz, B.; Sobczyk, L. Chem. Phys. Lett. 2009, 473, 75. (7) Majewska, P.; Rospenk, M.; Sobczyk, L. Pol. J. Chem. 2009, 83, 709. (8) Paja- k, J.; Galewski, Z.; Rospenk, M.; Sobczyk, L. Liq. Cryst. 2001, 28, 1003. (9) Sobczyk, L.; Grabowski, S. J.; Krygowski, T. M. Chem. Rev. 2005, 105, 3513. (10) Gilli, G.; Bellucci, F. F.; Ferretti, V.; Bertolasi, V. J. Am. Chem. Soc. 1989, 111, 1023. (11) Gilli, P.; Bertolasi, V.; Pretto, L.; Ferretti, V.; Gilli, G. J. Am. Chem. Soc. 2004, 126, 3845. (12) Gerlt, J. A.; Kreevoy, M. M.; Cleland, W. W. Chem. Biol. 1997, 4, 25. (13) Filarowski, A.; Koll, A.; Sobczyk, L. Curr. Org. Chem. 2009, 13, 172. (14) Claydon, M. F.; Sheppard, N. Chem. Commun. 1969, 1431. (15) Zundel, G. Adv. Chem. Phys. 2000, 111, 1. (16) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A.; Vreven, T.; Kudin, K. N.; Burant, J. C.;. Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Cossi, M.; Scalmani, G.; Rega, N.; Peterson, G. A.; Nakkatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratman, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvator, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, D. K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Ciolowski, B. B.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Gonzales, C.; Pople, J. A. Gaussian 03, revision D.01; Gaussian Inc., Wallingford, CT, 2004. (17) Michalska, D.; Bienko, D. C.; Abkowicz-Bienko, A. J.; Latajka, Z. J. Phys. Chem. 1996, 100, 17786. (18) Michalska, D.; Zierkiewicz, W.; Bienko, D. C.; Wojciechowski, W.; Zeegers-Huyskens, T. J. Phys. Chem. A 2001, 105, 8734. (19) Lampert, H.; Mikenda, W.; Karpfen, A. J. Phys. Chem. A 1997, 101, 2254. (20) PQS version 3.3; Parallel Quantum Solutions, Fayetteville, AR 72703. (21) Grams AI/8.0 spectroscopy software, Thermo Fisher Scientific Inc., 2009. (22) Majewska, P.; Rospenk, M.; Czarnik-Matusewicz, B.; Sobczyk, L. Chem. Phys. 2007, 334, 117. (23) Majewska, P.; Rospenk, M.; Czarnik-Matusewicz, B.; Kochel, A.; Sobczyk, L.; Da-browski, R. Chem. Phys. 2007, 340, 227. (24) Majewska, P.; Rospenk, M.; Czarnik-Matusewicz, B.; Kochel, A.; Sobczyk, L.; Da-browski, R. Chem. Phys. 2008, 354, 186. (25) Thulstrup, E. W.; Thulstrup, P. W. Acta Chim. Slov. 2005, 52, 371. (26) Rai, P. K.; Denn, M. M.; Maldarelli, C. Langmuir 2003, 19, 7370. (27) Georgie, A. K.; Mohandas, K. P. Phys. Chem. Liq. 1995, 30, 233. (28) Mach, P.; Huang, C. C.; Stoebe, T.; Wedell, E. D.; Nguyen, T.; de Jeu, W. H.; Guittard, F.; Naciri, J.; Shashidhar, R.; Clark, N.; Jiang, I. M.; Kao, F. J.; Liu, H.; Nohira, H. Langmuir 1998, 14, 4330.

ARTICLE

(29) Mach, P.; Huang, C. C.; Nguyen, H. T. Phys. Rev. Lett. 1998, 80, 732. (30) Malinowski, E. R. Factor Analysis in Chemistry, 3rd ed.; John Wiley and Sons Inc.: New York, 2002. (31) Czarnik-Matusewicz, B.; Paja-k, J.; Rospenk, M. Spectrochim. Acta, Part A 2005, 62, 157. (32) Radziszewski, J. G.; Downing, J. W.; Gudipati, M. S.; Balaji, V.; Thulstrup, E. W.; Michl, J. J. Am. Chem. Soc. 1996, 118, 10275.

2736

dx.doi.org/10.1021/jp107826x |J. Phys. Chem. B 2011, 115, 2728–2736