Solvent Effects on the Electronic Spectra of Some Heterocyclic Azo Dyes

Abstract

The influence of solvent polarity on the absorption spectra of some synthesized azo dye with heterocyclic moieties and ${\beta}$-naphthol (1-3) have been investigated using a UV-Visible spectrophotometer. The spectral characteristics of the azo dyes (1-3) in different solvents at room temperature were analyzed. The solvatochromic empirical variables like ${\pi}^*$, ${\alpha}$, and ${\beta}$ have been used to discuss the solvatochromic behaviour of the dyes and to evaluate their contributions to the solute-solvent interactions. A multi-parameter regression model for quantitative assessment of the solute/solvent interaction and the absorption has been used to explain the solvent effect on azo dyes (1-3).

Introduction

The solvent effect on spectra, resulting from electronic transitions, is primarily dependent on the chromophore and the nature of the (σ→σ*, n→σ*, π→π*, n→π* and charge-transfer absorption) transition. The electronic transitions of particular interest in this respect are π→π* and n→π* as well as charge transfer absorptions. Solvatochromism is caused by differential solvation of the ground and first excited state of the chromophore. The positive solvatochromism is due to the increase in stability of the excited state with increasing solvent polarity. Similarly the negative solvatochromism is due to better stabilization of the molecule in the ground-state relative to that in the excited state, with increasing solvent polarity. In this context, “first excited state means the so-called Franck-Condon excited state with the solvation pattern present in the ground state.

The solvent-dependent spectral shifts can arise from either nonspecific (dielectric enrichment) or specific (e.g. hydrogen- bonding) solute-solvent interactions. The solvent effect can be determined by solvent polarity scale or solvatochromic parameters.1 Solvent polarity is a commonly used term related to the capacity of a media for solvating neutral, charged, polar and apolar species. Attempts to express it quantitatively have mainly involved solvent properties such as relative permittivity, dipole moment, or refractive index, but these parameters cannot effectively account for the multitude and specific interactions of solute-solvent on the molecular-microscopic level2. Spectroscopic solvent polarity parameters have been derived from solvent-sensitive standard compounds absorbing radiation in spectral ranges corresponding to UV/Visible, IR, ESR, and NMR spectra.3-7

Among the single-parameter approaches, the ET(30) scale as one of the more comprehensive solvent scales. ET(30) values are based on the negatively solvatochromic pyridinium N-phenolate betaine dye as the probe molecule. The parameters are simply defined, in analogy to Kosower's Zvalues,8 as the molar electronic transition energies (ET) measured in kilocalories per mole (kcal/mol) at room temperature (25 ℃) and normal pressure (1 bar),9 and can be determined by Eq. (1).

where λmax is the wavelength of the maximum of the longest wavelength of π→π* absorption band of betaine dye.10

In spite of the observation that single empirical parameters may serve as good approximations of solvent polarity in the sense defined, there are many examples of solvent-sensitive processes known, which cannot be interrelated to one empirical solvent parameter. However, multi-parameter solvent polarity scale for quantitative assessment of the solute/ solvent interaction and the absorption shifts can be used. The effect of solvent polarity on the absorption spectra are interpreted by means of linear solvation energy relationship (LSER) using a Kamlet-Taft equation.11

where π* is a measure of the solvent dipolarity/polarizability,12 β is the scale of the solvent hydrogen bond acceptor (HBA) basicities,13 α is the scale of the solvent hydrogen bond donor (HBD) acidities14 and is the regression value of the solute property in the reference solvent cyclohexane. The regression coefficients s, b and a (Eq. 2) measure the relative susceptibilities of the solvent-dependent solute property (absorption frequencies) to the corresponding solvent parameters.

Azo dyes are normally known to show a positive solvatochromism.15-22 The first examples of negative solvatochromism in neutral azo dyes containing both strongly electron-donating and -withdrawing moieties were reported by Kim et al..23 Recently Mohammadi and coworkers24 prepared five azo disperse dyes and studied the electronic absorption spectra of dyes in fifteen solvents with different polarities at room temperature. The solvent dependent maximum absorption band shifts were investigated using dielectric constant (ε), refractive index (n) and Kamlet-Taft polarity parameters (hydrogen bond donating ability (α), hydrogen bond accepting ability (β) and dipolarity/polarizability polarity scale (π*). Acceptable agreement was found between the maximum absorption band of dyes and solvent polarity parameters especially with ε and π*.

The aim of the present work is to investigate the influence of solvent polarity on the UV-Visible absorption spectra of some synthesized azo dye using heterocyclic moieties and β-naphthol (1-3) to evaluate the intermolecular interactions occurring in solutions. The spectral characteristics of the azo dyes (1-3) in different solvents at room temperature were analyzed. The solvatochromic empirical variables like π*, α, and β have been used to discuss the solvatochromic behaviour of the dyes and to evaluate their contributions to the solute-solvent interactions.

 

Experimental Section

Azo dyes using heterocyclic moieties (1-3) were prepared by the method reported earlier.25 Spectroscopic grade solvents (Merk) were used throughout the experiment. Stock solutions of the dyes (1 ×10˗3 M) were prepared in methanol and its concentration in the experimental condition was maintained at 5 × 10˗5 M. Absorption spectra were recorded on Shimadzu-160 UV-Visible spectrophotometer at 300 K with sensitivity ± 0.1 nm.

 

Results and Discussion

Electronic Absorption Spectra. The electronic (UV-Visible) spectra of heterocyclic azo dyes (1-3) were investigated in various solvents at a concentration of ~10˗5 M. The different types solvents selected are (i) polar protic solvents such as methanol, ethanol, n-propanol, n-butanol and secbutanol (ii) polar aprotic solvents such as dichloromethane, acetone, acetonitrile, ethyl acetate, DMF and DMSO and (iii) nonpolar solvents such as hexane, benzene, toluene, dioxane, chloroform and diethylether. The values of λmax of the azo dyes (1-3) in different solvents are summarized in Table 1. The characteristics visible spectral band (λmax) appears in the range 485-528 nm depending on the nature of the solvent. The electronic spectra of the dye 1 in different class of solvents (polar protic, polar aprotic and nonpolar) are given in Figure 1.

Table 1.Absorption maxima (λmax) of azo dyes (1-3) in organic solvents at 27 ℃

The data of Table 1 reveals that the absorption maxima of the dyes (1-3) are affected by solvent type and has a maximum shift of Δλ = 35 nm for the solvents used in this work. Thus this change in spectral position can be used as a probe to analyze various types of interactions between solute and solvent. The electronic spectra of these dyes in visible region is due to π→π* transition involving the whole π- electronic system. In acid media, an n→π* band disappears due to protonation of the lone pair. The protonation may increase the excitation energy to an extent that the band may shift far out into the UV region and not be observed. In this case the possibilities of n→π* transition is ruled out as the λmax does not change in acidic condition.

The broadness in the electronic spectra suggests the existence of a considerable charge transfer (CT) character.26-28 The charge delocalization seems to originate mainly from the naphthyl to heterocyclic moiety, which acts as an electron acceptor. Amrallah and coworkers26 have reported similar behaviour for the UV-Visible spectra of some aryl azo barbituric acids and aryl azo pyrimidine in pure and mixed organic solvents of varying polarities. This band is red shifted when passing from the non-polar solvent (n-hexane; λmax = 485 nm for dye 1, 494 nm for dye 2 and 496 nm for dye 3) to polar solvent (DMSO; λmax = 498 nm for dye 1, 511 nm for dye 2 and 517 nm for dye 3). The solvatochromism is caused by differential solvation of the ground and Franck-Condon excited state, due to the absorption of electromagnetic radiation in the visible region. As the polarity of solvent increases, the excited state is more stabilized than the ground state due to solvation.

Figure 1.UV-Visible spectra of azo dye 1 in (a) polar protic, (b) polar aprotic (c) aprotic solvents at 27 ℃.

The visible spectra of the dyes (1 & 3) exhibit longer shift in DMF and DMSO compared to other solvents (Table 1). DMF and DMSO are characterized by their high basicity and hydrogen-bond accepting ability. By these solvent excited states of the dyes (1 & 3) is stabilized more compared to the ground state leading to the bathochromic shift. Rageh et al. have also observed similar behaviour in DMF and DMSO while investigating the solvent effect of heterocyclic azo dyes and azo coumarin dyes.29 They proposed the formation of solvated complex with DMF and DMSO molecules through intermolecular H-bonding which is in accordance with the earlier observation by Ibrahim et al..30 Formation of solvated complex led to the stability of excited state.

Table 2.Observed ET of azo dye (1-3) in organic solvent at 27 ℃

Correlation of Solvent Parameter. The absorption data of dyes in various solvents were also analyzed with respect to various polarity scales. For the purpose the absorption spectra were converted to corresponding transition energy (Eq. 1) and the data are given in Table 2.

To investigate the effect of solvent on the structural artifact of the dyes, the transition energy of the dyes (1-3) were correlated with each other. The transition energy of the dyes is found to have linear relationship with a slope of 0.666 for dyes 1 and 2, 1.260 for the dyes 2 and 3, and 0.901 for the dyes 1 and 3 (Figure 2).

These slopes refer to the change in solvation pattern of the dyes. For a specific dye the change in transition energy in different solvents may be ascribed to the change in the solvation behaviour, leading to the formation of a solvationcomplex. The slope of the above mentioned plots may be considered as the ratio of the binding efficiency of one dye to the other. Accordingly, the solute-solvent interaction of 2 found to be more significant than 1. However, the interactions of 1 and 3 with the solvents have almost similar characteristics. The structural change in the dyes apparently substantiates the observations. Fusion of benzo-nucleus with the thiazole moiety decreases the overall space of interaction with respect to the substituted phenyl group, while the presence of electron withdrawing halo-group brings similar rigidity due to delocalization of lone pair of electrons of the –Cl group. Further, if the solute-solvent interaction is due to electronic factors, the benzo fusion on the thiazole nucleus in dye 1 decreases its electron density due to anellation effect31 and in dye 3 due to the electron withdrawing –Cl group. As the solute-solvent interaction is mostly contributed by the polarity of the medium, which is evident from the plot of ET(30) verses ET(Dye), the electron donating polar solvents can bind strongly with dye 1 than the other two dyes, which happened in the present case. Earlier Mishra et al. have reported the quaternization kinetics of 2-amino benzothiazole, 4-phenyl-2-amino thiazole and 4-(pchlorophenyl)-2-amino thiazole with phenacyl bromide and found similar trend in reactivity in nitrobenzene medium.31

Figure 2.ET of (a) dye 1 vs dye 2 (b) ET of dye 2 vs dye 3 (c) ET of dye 1 vs dye 3.

When the transition energy of the dyes are compared with that of the standard pyridinium N-phenolate betaine dye by plotting against ET(30) scattered plots are obtained (Figure 3). However, from the linearity of some points in the scattered plots, the solvents can be classified into three distinct categories i.e. aprotic, polar aprotic, and polar protic solvents. The sensitivities of polarity towards the change in transition energy in each class of solvents are determined from the slope of the plots (Table 3).

Figure 3.Plot of ET (Dyes) with ET(30) for (a) dye 1, (b) dye 2 and (c) dye 3 in organic solvents at 27 ℃.

For dye 1 the trend of sensitivity of the polarity is polar aprotic > polar protic > aprotic, while for dyes 2 and 3 the trend is polar aprotic > aprotic > polar protic. The insignificant contribution of the protic solvents towards the transition energy may be ascribed to the strong hydrogen bonding of the hydrogen of the protic solvents with three nitrogen centers, one oxygen and one sulfur centers; thereby the contribution of polar characteristics of the solvent is not sensed properly by the dyes. Thus the solute-solvent interaction is mostly due to hydrogen bonding. In polar aprotic solvents, the solute-solvent interaction is mostly due to polar characteristics and thus the sensitivity is maximum in all the dyes.

Table 3.Slope and R2 of the plots between ET (dyes) vs. ET(30) for aprotic, polar protic and polar aprotic solvents

Quantitative Solvent Spectral Relationship. A multi-parameter regression model for quantitative assessment of the solute/solvent interaction and the absorption shifts has been used to explain the solvent effect on azo dyes (1-3). Using Kamlet-Taft equation (Eq. 2), the effect of solvent polarity on the absorption spectra has been investigated with three variables and it was found that the R2 values for all the three dyes are around 0.59. Thus the regression model necessitates more number of variables to explain the solvatochromism. Fourteen different solvent parameters mentioned in Table 4 have been obtained from literature.33 Before their use in regression model their auto-correlation has been tested through correlation matrix.

For compatibility of the parameters (to bring all the data in the same scale), instead of ET(30) its normalized values has been used in the multiple regression analysis.8 The other parameters are hydrogen bond donocity (α), hydrogen bond acceptor ability (β), dipole moment (μ), dielectric constant (D), log P, refractive index (n), Taft’s parameter (π*), solubility (δ), acidity (SA), basicity (SB) and the dipolarity and polarizability (SPP).

Table 4.α: hydrogen bond donocity, β: hydrogen bond acceptor ability, μ: dipole moment, D: dielectric constant, log P, n: refractive index, π*: Taft’s parameter, δ: solubility, SA: acidity, SB: basicity and SPP: dipolarity & polarizability.

All the solvent parameters were subjected to regression analysis to find out the contribution of each parameter towards the ET values of the dye by using a generalized regression model (Eq. 3) where the coefficient of the parameters a………l are the susceptibility of the concerned parameter towards ET value.

The regression coefficient (R2) and statistical F values are the indicators of the validity of the equation. Increase in R2 and F value indicates the fitness of the regression model. The statistical ‘t’ test for the appropriateness of the parameter to be used in the regression model has also been considered for selection of the parameter for the regression model. For optimization of the regression model a successive exclusion of variable (SEV) technique was adopted wherein variables with ‘t’ values less than one were deleted in each successive generation of regression model.

The multiple regression analysis of the ET values of dye 1 using all the 16 solvent parameters results in a general regression model (Eq. 3) with high R2 value (0.97) and a F value of 2.326 delineating 99.1% confidence level. However, the ‘t’ test indicates the inappropriateness of some of the parameters due to low value (< 1.0). The parameter β with the lowest value of 0.07 was excluded in the first step from the regression model and thus a new regression model was obtained with almost similar R2 value as well as with a significant increase in F value (4.99). The exclusion of variables are continued till the minimum value of ‘t’ reaches > 1 i.e. 1.5 for . The optimization regression models of dye (1-3) obtained by above techniques and ET of dye (1-3) were determined and plotted against the experimentally observed ET values (Figure 4). The linearity of the plots validates the regression models.

Figure 4.Plot of calculated ET vs. observed ET of (a) dye 1 (b) dye 2 (c) dye 3.

DMF and DMSO incompatible with the regression model may be attributed to complexicity in solute-solvent interaction which may arise due to a variety of bonding sites of different extent in these solvents and probes.

 

Conclusions

Absorption maxima of dyes are dependent on solvent polarity. Solvation of dye molecules probably occurs via dipole-dipole interactions in non-hydrogen-bond donating solvents, whereas in hydrogen-bond donating solvents the phenomenon is more hydrogen bonding in nature. The unified scale for estimating the solvent effect on the absorption of azo dye is more adopted and more applicable than the π* scale model.