On the Thermal Transformation of All-trans-1,6-Diphenyl-1,3,5-hexatriene (All-trans-DPH) into Its s-cis Conformer (s-cis-DPH) in Solution

Abstract

The behavior of the energy of the peaks of the first UV/Vis absorption band and the presence or absence of isosbestic points in this band with changing temperature for all-trans-DPH and all-trans-β-carotene, dissolved in 1-chlorobutane or hydrocarbon solvents, allows us to show conclusively whether these compounds transform their all-trans-molecular structures into a structure of their conformers. From these analyses, it is concluded that in these solvents, all-trans-DPH is not thermally transformed to its conformer s-cis-DPH in a temperature range from 133 K to 350 K. On the other hand, all-trans-β-carotene, as a model-compound, does show changes in its molecular structure in these solvents with changing temperature. We also show that a portion of all-trans-DPH dissolved in cis-Decalin, at room temperature, slowly transforms into its conformer s-cis-DPH.

1. Introduction

Polyene compounds contain chromophores which are involved in important processes such as vision, coloration of fruits, and generation of vitamins and in protection mechanisms against sunlight. In these biochemical processes, the photochemistry of these compounds plays a very important role and, therefore, it has been imperative to know their photophysical functioning [1].

Recently, we provided experimental results [2] that allowed us to conclude that all-trans-DPH, dissolved in a hydrocarbon solvent within a temperature range of 77–365 K, was not thermally transformed into its conformer s-cis-DPH. This experimental evidence contradicted what was proposed over the last 32 years by Saltiel et al. [3,4]. It seems clear that this disagreement can be solved only by providing more experimental pieces of evidence.

Thus, in this work, new experimental evidence is provided, obtained in a wide range of temperatures in order to evidence the following: (a) the temperature-dependent shift in the peaks of the first UV/Vis absorption band of the analyzed polyene compound and (b) the existence of isosbestic points in the first UV/Vis absorption band of these compounds. We believe that this information will allow us to clarify the possibility of thermal changes in the transformation of the molecular structure of these compounds. To this end, we studied the UV/Vis spectral behavior of two polyenes, one that has been shown to thermally change its structure, such as β-carotene [5,6,7,8], and another, the DPH molecule, for which doubts remain on its possible thermal transformation [2,9]; see Scheme 1.

The UV/Vis absorption spectra of different symmetrical di-substituted derivatives of hexatriene, X-(CH=CH)₃-X, with X = H [10], X = Me [11,12], tert-butyl [13], and phenyl [14,15], show a first UV/Vis absorption band, which is evidenced in solution at rt, with three more or less resolved peaks. The first peak corresponds to its 0-0 component, the second one marks the maximum of the band, and the third one is less intense and tends to disappear above room temperature.

In principle, it is to be supposed that two behaviors exist for the energy of the peaks of the first absorption band of the compound with increasing temperature:

If the energy of the peaks of the first absorption band of all-trans-DPH varies linearly with temperature, it is acceptable to conclude that the compound keeps the structure of its polyene chain unchanged within the considered temperature range and, consequently, there exists neither isomerization nor change between conformers when changing the temperature.

When we change the temperature and all-trans-DPH generates stereoisomers or conformers, we can find two different situations: (a) The first absorption band of the new generated structure overlaps with the corresponding band of all-trans-DPH and plotting the energy of the absorption peaks against temperature would consequently show bilinear behavior. It is also important to note that this situation should always imply the presence of an isosbestic point in the spectral behavior of the compound. (b) The first absorption band of the generated structure does not present intensity in the spectral zone corresponding to the first absorption band of all-trans-DPH and, consequently, their peaks would not reflect such a transformation when varying the temperature.

In his brilliant contribution to the isomerization of polyene compounds, Zechmeister [16,17] reports that α,ω-diphenylpolyenes and carotenoids can be isomerized by different methods such as the following: (a) thermal cis-trans isomerization in solution, (b) cis-trans isomerization by melting crystals, (c) cis-trans isomerization by iodine catalysis at room temperature, (d) cis-trans isomerization by acid catalysis, and (e) photochemical cis-trans isomerization. However, when Zechmeister [16,17] analyzed the thermal isomerization of compounds with long polyene chains and with structural peculiarities such as carotenoid compounds, he left aside thermal isomerization in shorter chain compounds as they are undetectable. Thus, it is noteworthy that Lunde and Zechmeister [9] in their work on the isomerization of all-trans-DPH indicate, when exposing the different procedures that they used to isomerize this compound, that in this compound the thermal route is not viable, writing “Refluxing (45 min. in n-propyl alcohol, b.p. 97 °C, 7 mg. per 100 mL; in the darkness) only traces of cis forms were observed”.

The information provided by Lunde and Zechmeister [9] allows us to (a) discard in a solution of all-trans-DPH the thermal generation of isomeric structures, that is to say molecular structures generated by a torsion around one of the double bonds of its hexatriene chain. According to these authors, there are only six isomeric structures feasible in this compound (ttt-DPH, ctt-DPH, tct-DPH, ctc-DPH, cct-DPH, and ccc-DPH), but only two of these (ctt-DPH and tct-DPH) are generated directly from the structure of all-trans-DPH. These two structures are given in Scheme 2. (b) In principle, all-trans-DPH could also generate two conformers (tstt-DPH and tstst-DPH), but only one of these (s-cis-DPH) is generated directly for the torsion of the first simple bond of the hexatriene chain for all-trans-DPH; see Scheme 3.

In this work, we will compare UV/Vis spectroscopic evidence obtained in a solution of a polyene that is thermally isomerized [16,17], such as all-trans-β-carotene (trans-βCa), with that obtained in identical conditions of a polyene such as all-trans-1,6-diphenyl-1,3,5-hexatrien, which we will call all-trans-DPH. We will also analyze the UV/Vis spectroscopic behavior of all-trans-DPH dissolved in n-dodecane in a temperature range between 283 and 363 K. We will do this in order to rule out the existing controversy about its possible thermal transformation of all-trans-DPH to generate s-cis-DPH.

We will also analyze the possibility of generating s-cis-DPH in a solution of all-trans-DPH dissolved in cis-Decalin, a solvent that has a viscosity of 3.381 cP at 293 K [18].

2. Experimental Sections

The all-trans-β-carotene was of Fluka purum grade (purity > 97.0%). The all-trans-1,6-diphenyl-1,3,5-hexatriene (DPH) was from Sigma-Aldrich (St. Louis, MO, USA) (98% pure) and it was purified by means of crystallization from ethyl acetate before being used. 1-chlorobutane (1CB) was of Chromasolv grade (purity 99.8%). DEMCHEM is a mixture of 1:1 (in volume) decalin and methylcyclohexane. Methylcyclohexane and anhydrous decalin were purchased from Sigma-Aldrich (St. Louis, MO, USA) and possessed a purity higher than 99%. n-Dodecane, n-Hexadecane, and cis-Decalin were anhydrous and had over 99% purity; they were purchased from Sigma-Aldrich (St. Louis, MO, USA).

Between 133 and 293 K, the temperature of the solution was controlled with an Oxford DN1704 cryostat (OXFORD INSTRUMENTS, UV, Oxon, OX13 5QX, England) that was purged with dried nitrogen (99.99% pure) and equipped with an ITC4 controller interfaced to the UV/vis spectrometer. Between 273 and 373 K, the solution temperature was controlled using a Fison Haake D8 GH thermostat (Allscwil Dieselstrabe 4, D-76227 Karsruhe, Germany). All UV/Vis absorption spectra were recorded on a Cary-5 spectrophotometer (Varian Australia Pty Ltd., Mulgrave VIC 3170, Belrose, NSW, Australia) at variable temperatures, using Quartz Suprasil cells of 1 cm path length that were fixed to the cryostat or to the Cary-5 cell holder according to the temperature range analyzed.

3. Results and Discussions

The UV/Vis absorption spectra of trans-β-carotene and all-trans-DPH in the solvents studied, together with the behavior shown by the absorption spectra of all-trans-DPH in n-dodecane obtained between 293 and 363 K, will allow us to establish whether in these solutions these molecules deform their polyene chain thermally, generating their conformers.

3.1. On the Thermal Transformation of All-trans-βCa Dissolved in 1CB and DEMCHEM

From the UV-Vis absorption spectra of all-trans-βCa dissolved in 1CB and measured at temperatures of 293 and 353 K (also see Figure 5 of Ref. [19]), the following can be deduced: (a) at 353 K, the presence of a cis peak is evident, with its maximum around 340 nm, clearly assignable to a cis-βCa isomer structure, which confirms that this compound at 353 K in 1CB is thermally transformed generating a cis-βCa structure, and (b) the corresponding spectrum obtained at 293 K does not show this cis peak, which indicates that at this temperature this thermal transformation of trans-βCa does not occur.

To ensure the possible presence of a new molecular structure in the solution of trans-βCa in 1CB, in Figure 1 we show the absorption spectra obtained in the 283–343 K temperature range. From these spectra, the following can be deduced: (a) a cis-isomer structure of trans-βCa is detected in this solution only above 303 K and (b) below this temperature, this compound does not show signs of thermal transformation.

Concerning the spectra obtained for trans-βCa in 1CB between 133 and 343 K (see Figure 1 and Figure 2), we evaluate the position of the peak corresponding to the 0-0 component of the first UV-Vis absorption band of the all-trans-βCa in 1ClB, and we find that these values follow a linear correlation versus the temperature value in a temperature range between 133 and 300 K and another distinct linear behavior above 300 K; see Figure 3a. The same bilinear behavior is found if the evaluated data are those of the peak corresponding to the maximum of this band; see Figure 3b.

If we analyze the spectra of trans-βCa in 1CB, obtained in the temperature range of 283 and 343 K (Figure 1), we observe that they present an isosbestic point around 422 nm. If we perform the same analysis for the data corresponding to the range between 133 and 283 K (Figure 2), we do not observe an isosbestic point.

From the analysis of the absorption spectra of trans-βCa, dissolved in DEMCHEM at the temperature range of 283–363 K (Figure 4), we find that trans-βCa starts the transformation of its molecular structure above 323 K and that this structure also presents a cis peak with its maximum around 340 nm. The trans-βCa in this solvent does change its molecular structure below 323 K.

Concerning the spectra obtained for trans-βCa in DEMCHEM between 133 and 363 K (see Figure 4 and Figure 5), if we evaluate the position of the peak corresponding to the 0-0 band, we find that these values follow a linear correlation versus temperature, in the range between 133 and 310 K, and another distinct linear behavior above 323 K (see Figure 6a). The same bilinear behavior is found if the evaluated data are those of the peak corresponding to the maximum of this band (Figure 6b).

If we analyze the trans-βCa UV/Vis spectra in DEMCHEM, obtained between 283 and 363 K (Figure 4), we observe an isosbestic point around 422 nm. If we perform the same analysis for the data corresponding to the range of 133–283 K (Figure 5), we do not observe an isosbestic point.

3.2. On the Thermal Transformation of All-trans-DPH Dissolved in 1CB and DEMCHEM

In Figure 7, we show the UV/Vis absorption spectra of all-trans-DPH, dissolved in 1CB and measured in a temperature range of 283–343 K. In these spectra, we do not find evidence that leads us to conclude the following: (a) a DPH cis-peak is being generated and (b) an isosbestic point is being generated. In Figure 8, we show the first absorption band of all-trans-DPH, measured in 1CB in a temperature range of 133–293 K.

The spectral data presented in Figure 7 and Figure 8 allow us to evaluate the behavior of the peaks corresponding to the 0-0 component and the maximum of the first absorption band of the compound measured in a temperature range of 133–343 K. The results presented in Figure 9a,b show linear behavior throughout the wide temperature range studied for both spectral peaks.

To this evidence on the behavior of all-trans-DPH in DEMCHEM, we add that presented in Figure 1, Figure 2, Figure 5 and Figure 6 of our previous publication; see reference [1]. We must conclude that for the two solvents, 1CB and DEMCHEM, in which βCa undergoes a thermal transformation of its molecular structure, all-trans-DPH does not provide any evidence to support its thermal transformation.

After an elaborate theoretical treatment of the corresponding spectroscopic data of all-trans-DPH, dissolved in hydrocarbon solvents and measured between 283 and 374 K, Saltiel et al. [3] decomposed the first UV/Vis absorption band of all-trans-DPH dissolved in n-dodecane into two pure spectra, one assigned to all-trans-DPH and the other to its conformer, s-cis-DPH.

The first absorption bands for both conformers, obtained in their spectral decomposition by these authors, coincide in the same spectral zone. From the figures presented (see Ref. [3]), it is evident that the one corresponding to s-cis-DPH is slightly shifted to the blue with respect to the one corresponding to all-trans-DPH. It is also relevant to indicate that the band assigned to the s-cis-DPH shows three peaks, as does all-trans-DPH, although slightly less resolved.

Since the transformation of all-trans-DPH into s-cis-DPH involves a rotation around a single bond of the polyene chain of the compound (see Scheme 3), one assumes that the process, as performed by Saltiel et al. [3], will be very probable, and even more so the more we raise the temperature of the solution. Consequently, these authors, in their work, [3] come to write the following: “The molar fraction of the s-cis-conformer equals, or even exceeds, the molar fraction of the all-trans-conformer as the highest temperatures employed in our study are approached” [3].

In Figure 10, we show the first absorption band of a solution of all-trans-DPH in n-dodecane measured between 283 and 363 K.

It is important to note that Marian et al. [20] in 2011 published a rigorous theoretical paper in which they studied the spectral behavior of all the possible isomers and rotamers of DPH. In their study, they obtained two important conclusions that directly affect what was proposed by Saltiel et al. [3]. A first conclusion reached by these authors is summarized in their work in the following sentence: “Proceeding to the absorption wavelengths, it is noteworthy that the s-cis rotamers tstt and tstst are red-shifted with respect to ttt. In contrast, cis isomerization of a double bons leads to a blue shift in the absorption wavelengths compared to ttt.”

The spectroscopic results shown in Figure 10 rule out the presence of absorption located near the red of the corresponding absorption band of all-trans-DPH that grows in intensity as the solution temperature increases.

The results presented in Figure 11a,b show linear behavior for both spectral peaks of the all-trans-DPH spectrum in n-Dodecane. Clearly, from these results shown in Figure 10 and Figure 11, we must conclude that in these solutions, only all-trans-DPH exists in the wide temperature range analyzed.

3.3. On the Transformation of All-trans-DPH in s-cis-DPH

The second conclusion reached by Marian et al. [20] is that the most energetically stable structure of DPH is all-trans-DPH and that the energy of the conformer directly generable from this, the s-cis-DPH (see Scheme 3), is significantly more unstable, by 5.1 kcal/mol.

Then, one should consider that raising the temperature will decrease the viscosity of the solvent and consequently decrease the potential barrier of the process between these conformers, thus facilitating, in principle, the conversion of all-trans-DPH to s-cis-DPH. But, in this case, we must also keep in mind that at a given temperature the barrier to convert s-cis-DPH into all-trans-DPH will always be about 5.1 kcal/mol lower and consequently this process would be the dominant process among these DPH conformers, making those proposed by Saltiel et al. [3] very unlikely if not impossible in the solvents and at the temperatures used by Saltiel et al. [3].

Figure 12 shows the first UV/Vis absorption band of a solution of all-trans-DPH in cis-Decalin measured at 293 K. The absorption spectrum, ----, corresponds to a freshly prepared solution and the spectrum, ____, corresponds to this solution after being stored for a few days in the dark at room temperature.

The spectra, shown in Figure 12, clearly show that in the initial solution of all-trans-DPH (see the dashed spectrum) stored in the dark at room temperature, a new molecular structure is generated that absorbs more red than all-trans-DPH did (see the spectrum drawn with a continuous line).

It is interesting to see that this spectral behavior does not occur if the solvent used is n-hexadecane (see Figure 13 in which both spectra are identical); so, we must conclude that in n-hexadecane the corresponding conformer structure is not generated.

Using the UV/Vis absorption spectra shown in Figure 12, we can decompose the UV/vis absorption spectrum corresponding to all-trans-DPH kept for several days at room temperature into two spectra: one that would correspond to all-trans-DPH and another clearly red-shifted from it and that in principle we could assign to its conformer s-cis-DPH; see Figure 14.

In Figure 15, we present the emissions obtained by exciting at 423 or 355 nm the solution of all-trans-DPH in cis-Decalin that we kept in the dark for a few days. When this solution is exciting at 423 nm, the emission obtained is located between 425 and 560 nm and is structureless with a maximum at 466 nm, which we initially assign to the s-cis-DPH conformer. While exciting at 355 nm, we obtain a structured emission between 380 and 580 nm mainly due to all-trans-DPH, which is somewhat contaminated by a minor contribution due to s-cis-DPH.

Note also that the emission assigned mainly to all-trans-DPH is structured with a maximum at 429 nm, while the emission assigned to s-cis-DPH is structureless with a maximum at 466 nm, which is significantly red-shifted with respect to the previous one. This is consistent with the coplanar structure of all-trans-DPH versus the rotated out-of-plane structure of the s-cis-DPH conformer, caused by the repulsion of their hydrogen atoms 1 and 4; see Supplementary Material in reference [20].

It should be noted that the spectroscopic behavior of both emissions, see Figure 15, clashes head-on with that proposed by Saltiel et al. [21].

4. Conclusions

We must conclude that all-trans-DPH dissolved in 1CB, in DEMCHEM, or in n-Dodecane is not thermally transformed to s-cis-DPH. Furthermore, none of the experiments supported by Saltiel et al. [3,4,21] reveal the existence of any composite mixture of the conformers mentioned above. On the contrary, the UV-Vis absorption spectra of all-trans-β-Carotene in 1-CB and the DEMCHEM sample isosbestic points revealed the appearance of new species, which is not the case with all-trans-DPH.

This work also demonstrates the clear temperature dependence of the shift in the peaks for the first absorption band of 1,6-diphenylpolyene in predicting structural transformations of these compounds.

Finally, in this work, for the first time it is shown that in a solution of all-trans-DPH in cis-Decalin, the s-cis-DPH conformer is generated.