Excited-State Isomerization of Leuco Indigo - The ... - ACS Publications

The photoreaction of indigo and two other derivatives in its reduced (leuco) form was investigated by absorption and fluorescence (steady-state and ...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/JPCA

Excited-State Isomerization of Leuco Indigo R. Rondaõ ,† J. Seixas de Melo,*,† M. J. Melo,‡,§ and A. J. Parola§ †

Department of Chemistry, University of Coimbra, P3004-535 Coimbra, Portugal Departamento de Conservaçaõ e Restauro, UNL, Portugal § Requimte, CQFB, Departamento de Química, FCT-UNL, 2825 Monte de Caparica, Portugal ‡

ABSTRACT: The photoreaction of indigo and two other derivatives in its reduced (leuco) form was investigated by absorption and fluorescence (steady-state and time-resolved) techniques. The fluorescence quantum yield (ϕF) dependence with the UV irradiation time was found to increase up to a value of ϕF ≈ 0.2−0.3 (after 16 min) for indigo and ϕF = 0.2 (at ∼150 min) for its derivative 4,4′-dibutoxy-7,7′-dimethoxy5,5′-dinitroindigo (DBMNI). With a model compound, where rotation around the central C−C bond is blocked, the ϕF value was found constant with the UV irradiation time. Timeresolved fluorescence revealed that initially the decays are fitted with a biexponential law (with 0.12 and 2.17 ns), ending with an almost monoexponential decay (∼2.17 ns). Quantum yields for the isomerization photoreaction (ϕR) were also obtained for indigo and DBMNI with values of 0.9 and 0.007, respectively. The results are rationalized in terms of a photoisomerization (conversion) reaction occurring in the first excited singlet state of trans to cis forms of leuco indigo.



molecule in a trans planar configuration,6 preventing photochemical trans−cis isomerization,4,11 with the leuco species, as with butadiene, the central bond now possesses considerable single-bond character. Furthermore, in compounds related to indigo where the NH groups have been replaced by other heteroatoms such as oxygen, sulfur, or selenium, intense fluorescence and clear photochemical cis−trans isomerization have been reported.12−14 Indeed, with thioindigo, cis−trans photoisomerization (through the triplet state) has been reported to occur,15 and alkyl substitution at the nitrogen atom also allows photoisomerization to take place with indigo itself.16,17 This fascinating molecule has gained a renewed interest because (1) it is important for many applications to have a library of highly stable dyes, (2) the mechanism of photostability3,5,18−20 is not yet fully clarified, (3) in solution, the photodegradation quantum yields (ϕR) of its keto form were found to be in the order of 10−4 (that is a low value, thus revealing indigo as a highly stable dye), with isatin the major degradation product,21 and (4) with a few recent exceptions,8,10,22,23 the reduced leuco (Chart 1) form of indigo (and its derivatives) has been a forgotten species, in terms of both spectroscopic and photochemical properties. Because this is the water-soluble form, still used to dye “blue jeans”, it merits a deeper understanding of its properties, and our aim is to contribute to this through a study of the dependence of the fluorescence emission properties with the time of exposure to

INTRODUCTION Indigo and some of its derivatives (Tyrian Purple, for example) are in the category of ancient dyes used to give color to the world: in textiles, paintings, illumination, etc.1 Indigo has a mythical status; it was known to the Hebrews as “tekhelet” (although in this case it is believed that the sacred color consists of a mixture of indigo with Tyrian Purple), and is still used to dye the famous “blue jeans”. Besides its unique deep “blue color”, it demonstrates a remarkable stability, which has been linked to the combination of inefficient intersystem crossing and low triplet energy, coupled with fast internal conversion of the singlet state.2 The deactivation of indigo excited state has been associated with a process resulting from fast intramolecular proton transfer between the two adjacent carbonyl and N−H groups:3−5 although there are differing views regarding this mechanism:3,6−9 leading to excited-state deactivation mainly through internal conversion with negligible contributions from other deactivation processes (fluorescence, intersystem crossing to the triplet, or photochemistry).10 Moreover and in contrast with indigo’s keto (neutral) form (Chart 1), where it is generally accepted that intra- and intermolecular hydrogen bridges are formed keeping the Chart 1. Structures of Indigo’s Neutral (Keto) and Reduced (Leuco, Cis and Trans) Forms

Received: December 13, 2011 Revised: February 17, 2012 Published: February 21, 2012 © 2012 American Chemical Society

2826

dx.doi.org/10.1021/jp211996f | J. Phys. Chem. A 2012, 116, 2826−2832

The Journal of Physical Chemistry A

Article

The quantum yield of reaction for leuco indigo (at λexc = 416 nm) and its derivative DBMNI (λexc = 405 nm) was calculated from eq 2:21

light. This clearly demonstrates a photoinduced chemical transformation. To our best knowledge, it is the first time that a photoisomerization process is reported for the leuco form of indigo.



EXPERIMENTAL SECTION

ΔAV ε510d ϕR t

(2)

where I0 is the light absorbed by the solution at the irradiation wavelength (calculated from eq 1), V is the volume of irradiated solution (in mL), ΔA is the change in absorbance (at the monitoring wavelength) over the irradiation time period, Δt, and ε is the molar absorption coefficient of the compound at the monitoring wavelength. Considering that in the current experiments the changes observed were registered in terms of fluorescence emission, instead of absorbance, the factor (ΔA/ε) in eq 2 has been replaced by the change in normalized concentration C(t) along the irradiation time, according to eq 3:

Indigo and 1,10-phenanthroline were purchased from Aldrich, potassium ferrioxalate was from Alfa Aesar, and the indigo derivatives (4,4′-dibutoxy-7,7′-dimethoxy-5,5′-dinitroindigo (DBMNI) and methylated preciba (MpreCiba)) were synthesized as elsewhere reported.8,24 The solvents used were all of spectroscopic or equivalent grade. The leuco species was prepared by adding 2−3 drops of concentrated sodium dithionite/NaOH solution (0.15 g of Na2S2O4 in 10 mL of NaOH (1 M)) to the dye in dioxane, submitted to constant and gentle bubbling with Ar. The solution was left bubbling for further 20 min and then sealed in a proper device described elsewhere.25 Absorption and fluorescence spectra were recorded on Shimadzu UV-2100 and Horiba-Jobin-Yvon Spex Fluorolog 32.2 spectrophotometers, respectively. Fluorescence spectra were corrected for the wavelength response of the system. The fluorescence quantum yield of the compounds was determined using quinquethiophene (ϕF = 0.36 in dioxane26) as standard. Photoreaction figures and yields for the leuco forms were performed with a Horiba-Jobin-Yvon Spex Fluorolog 3-2.2 spectrophotometer, with irradiation at 370 nm for indigo, 408 nm for Mpreciba, and 405 nm for 4,4′-dibutoxy-7,7′-dimethoxy5,5′-dinitroindigo (3 mm bandwidth). To obtain the I0 from the irradiation source (450 W Xe lamp), the actinometer potassium ferrioxalate was used [ϕR = 1.14 (405 nm) and ϕR = 1.12 (416 nm) in aqueous 0.05 M H2SO4], using the “micro-version” procedure,27 and by collecting the change in the optical density at 510 nm, after several periods of time of irradiation (λexc = 405 nm, 4 mm bandwidth for DBMNI and λexc = 416 nm, 2 mm bandwidth for indigo) in the fluorimeter. At each irradiation wavelength (405 and 416 nm), a solution of 3 mL (0.006 M) of the actinometer was irradiated, keeping constant stirring, for 10 min. At the end of this, the cell was immediately taken out of the fluorimeter, and 0.5 mL of buffered phenanthroline was added. The absorbance at 510 nm was measured immediately and after 10, 20, and 30 min. Meanwhile, the solution was kept in the dark. The same was done with a reference solution that has been kept in the dark. The experiment was repeated, irradiating for 20 and 30 min. Assuming that the entire incident light is absorbed by the solution, the intensity of light is given by eq 1: I0 =

ΔAV ε1000I0Δt

ϕR =

C(t ) =

fluor. int.(t ) × final concentration final fluor. int.

(3)

where “fluor. int. (t)” stands for the intensity of the fluorescence emission at time t, “final concentration” stands for the concentration of the leuco form (when the total isomerization has been achieved), and “final fluor. int.” stands for the intensity of fluorescence emission upon total isomerization, that is, at the end of the experiment. As a consequence of this, eq 2 now transforms into eq 4: ϕR =

VsolΔC 1000I0Δt

(4)

where the ratio ΔC/Δt is obtained from the slope of the plot of C(t) versus time. Fluorescence decays were measured using a home-built TCSPC apparatus elsewhere described29 and were analyzed using the modulating functions method implemented by Striker.30 The experimental excitation pulse (fwhm = 21 ps) was measured using a LUDOX scattering solution in water. After deconvolution of the experimental signal, the time resolution of the apparatus is ca. 2 ps.29



RESULTS AND DISCUSSION Indigo. The observation that, upon UV irradiation of the leuco form of indigo, an increase was observed in both the emission wavelength maxima and the fluorescence quantum yield (Figure 1) led us to go into a deeper and detailed study of this and related molecule(s). Considering that in the reduced (leuco) form of indigo the C−C bond connecting the two indole moieties allows considerable rotation (particularly in the lowest excited state), this behavior was interpreted to be due to a (trans−cis) isomerization induced by light (photoisomerization). The experimental support will be given below. In addition, and supporting this, is the fact that we note a similar small red shift in the absorption spectrum of butadiene on going from the trans to the cis isomer.31 As can be seen from Figure 1, the fluorescence quantum yield values are low for t = 0 (ϕF ≈ 10−2), increasing to ϕF ≈ 0.2−0.3 when a steady-state equilibrium has been accomplished, and these latter values are in agreement with those previously published10,23 for the leuco form of indigo (and related molecules) where full conversion was achieved by leaving the sample under daylight exposure for approximately 20 min.

(1)

where ΔA is the absorbance of the irradiated solution at 510 nm, corrected by the absorbance of the reference solution, d is the optical path of the absorption cell measured in centimeters, ε510 = 1.11 × 104 M−1 cm−1, ϕR is the quantum yield of production of Fe2+ at the wavelength used in the photolysis, t is the irradiation time (minutes), and V is the volume (dm3) of the solution used in the determination of the absorbance.28 2827

dx.doi.org/10.1021/jp211996f | J. Phys. Chem. A 2012, 116, 2826−2832

The Journal of Physical Chemistry A

Article

Figure 1. Fluorescence emission spectra of indigo’s leuco species in dioxane obtained with different irradiation times at T = 293 K. The Rayleigh scattering peak at the excitation wavelength (λexc = 370 nm) is presented showing that the variation in the emission intensity is not due to light fluctuations but solely due to the property of the molecule under investigation; also shown as an inset is the plot of intensity at 495 nm versus the irradiation time. The leuco species was generated from a keto indigo solution as described in the Experimental Section and elsewhere.10

It is also worth noting that if this process consists of a pure trans−cis photoisomerization, differences, however likely to be small, are expected both in the absorption and in the fluorescence wavelength maxima. Considering that differences in the conjugation (electron delocalization) between cis and trans leuco forms are not very significant and rotation around the central C−C bond allows the thermodynamically more stable isomer to be formed, the absorption spectra taken at the beginning (before irradiation) and at the end of the experiment do not show appreciable differences (see Figure 2). However, the excitation spectra, obtained under the same conditions, revealed completely different spectral bands both in shape and in maxima (see Figure 2). Moreover, and as can be seen by observation of the absorption spectrum of the keto form (Figure 2), a valley (therefore with negligible absorption) is found precisely in the region where the intense absorption band of the leuco form occurs. This clearly indicates that the changes observed in Figure 1 cannot be attributed to incomplete keto− leuco interconversion, which would partially result from a gradual increment of the more fluorescent leuco species. As can also be seen from Figure 2B, there is an increase in intensity of the excitation spectra with time; however, and in addition to this increment, it is also possible to observe a small change in the shape of the spectra. This mirrors both the gradual conversion of the keto into the leuco form and the possibility of an initial leuco trans to be gradually converted into the (more stable) leuco cis conformer. Model Compound: MpreCiba. Nevertheless, it still may be considered that the above experiments do not constitute a full proof for the prevalence of photoisomerization, and further support for this hypothesis comes with additional evidence. This (attempt of proof) was therefore performed with two analogous systems: one where, for structural reasons, this behavior is not seen and another where the increase of the intensity of the leuco spectra with time is also found to occur. With the first case, we have studied the leuco form of an indigo derivative8 (with the acronym of MpreCiba) whose structure (see inset in Figure 3) precludes the possibility of

Figure 2. (A) Leuco’s indigo absorption and excitation spectra in dioxane obtained before and after irradiation under the conditions of Figure 1. Also shown is the absorption spectrum of indigo’s keto (neutral) form, also in dioxane. (B) Excitation spectra collected at different times for the leuco form of indigo in dioxane at T = 293 K; the total time of irradiation is t = 15 min.

Figure 3. Fluorescence emission spectra of the leuco form of MpreCiba8 (see structure in the inset) in dioxane obtained with different irradiation times at T = 293 K; the excitation wavelength was set to 408 nm.

photoisomerization. From Figure 3, it can be observed that there is effectively no change, upon irradiation, of the low fluorescence quantum yield of the leuco form of this compound. Note that the geometry of this compound (inset in Figure 3) is analogous to the trans form of leuco indigo. Moreover, the emission maximum (500 nm) of this reduced species is different from its keto form where it is found at 680 nm.8 The fluorescence quantum yield and lifetime of leuco2828

dx.doi.org/10.1021/jp211996f | J. Phys. Chem. A 2012, 116, 2826−2832

The Journal of Physical Chemistry A

Article

MpreCiba are, respectively, ϕF = 0.0026 and τF = 40 ps (major component of the decay with λem = 500 nm, although a second component with a value of ∼1.1−1.6 ns has a non-negligible value and an origin that is not yet clear to us). Analogous Indigo Derivative: 4,4′-Dibutoxy-7,7′-dimethoxy-5,5′-dinitroindigo (DBMNI). For the case of the analogous system/compound, we have investigated the leuco form of an hexa-substituted indigo derivative, 4,4′-dibutoxy-7,7′dimethoxy-5,5′-dinitroindigo (DBMNI); see structure in the inset in Figure 4. It can be seen that a behavior similar to that

Figure 5. (A) Leuco’s absorption and excitation spectra obtained before and after irradiation (under the conditions of Figure 4). In the labeling, tinit stands for the initial time (i.e., t ≅ 0) and tfinal for the final time, which is t ≅ 170 min (see Figure 4). Also shown is the absorption spectrum of indigo derivative’s keto (neutral) form. (B) Fluorescence excitation spectra for 4,4′-dibutoxy-7,7′-dimethoxy-5,5′dinitroindigo in dioxane, collected at λem = 500 nm and at different times (periods of irradiation), with a total time of irradiation (tirradiation) of 30 min.

Figure 4. Fluorescence emission spectra of the leuco species 4,4′dibutoxy-7,7′-dimethoxy-5,5′-dinitroindigo (DBMNI) in dioxane obtained with different irradiation times (tirradiation) at T = 293 K and λexc = 405 nm. Shown as insets are the intensity dependence with time obtained at λem = 490 nm and the magnified region of the Rayleigh peak (402−408 nm).

progressive conversion of the keto (blue) form, less emissive, into the leuco, more emissive (colorless or pale yellow) form of DBMNI. However, from the absorption (and excitation) spectra, no sign of the keto band (∼610 nm) could be observed, Figure 5A. Thermal Isomerization of DBMNI. The occurrence of thermal isomerization, concomitant with the photochemical isomerization process, was investigated at 20 and 14 °C, by measuring the emission from two samples, one under irradiation at 370 nm and the other kept in the dark, Figure 6. Although at dark there is significant isomerization (right panels in figure 6), it is far less (as seen by the total intensity) when compared to that induced by light, that is, the photoisomerization process. It is also worth remembering that the light exciting the sample (λexc = 370 nm) used to obtain the fluorescence spectra precludes a complete dark experiment even if narrow slits (0.5 mm) are used. When the fluorescence emission intensity at 490 nm is plotted against time, see Figure 7, it becomes clear that thermal isomerization accounts for almost 50% of the observed increase in the fluorescence emission quantum yield. As a consequence, the calculated photoreaction quantum yields are upper limit values of the true photoisomerization quantum yields. From Figure 7, it can also be observed that for an identical time value, the rate of thermal formation of the leuco-cis isomer is higher at 20 °C than at 14 °C, showing the thermal dependence of the isomerization process. Lowering the temperature could in principle decrease the thermal reactivity to a point where the photochemical reactivity would predominate; however, even at 14 °C, the thermal reactivity still accounts for ca. 50% of the total increase in fluorescence emission. Photochemistry (Trans−Cis Conversion) Quantum Yields. Photochemical quantum yields of reaction were also

found for indigo is observed with initial values for ϕF in the order of ∼0.002 and increasing upon photoconversion to 0.2, that is, 2 orders of magnitude higher. Summarizing briefly at this stage, all of the above-reported, dependence of the ϕF values for two indigo compounds together with the absence of change in this value in a compound whose structure precludes rotation around the central C−C bond, strongly support the existence of photoisomerization in the leuco form of indigo starting from a possible trans geometry and leading to a cis isomer. This last hypothesis (initial trans geometry) is reinforced by the photophysical characteristics of MpreCiba whose trans (rigid) structure has ϕF and τF values (0.0026 and 40 ps, see previous section) close to those found for indigo and DBMNI (ϕF ≈ 10−2 and τF ≈ 0.12 ns) at times before (or early stages of) irradiation. From Figure 4, it can be seen that the intensity of the Rayleigh peak is not constant until ∼2 min. It is worth noting that the time zero on our experiments is made immediately after preparation of the solution and avoiding light exposure. However, during the first minutes, it is likely that parts of the sample molecules are still being converted. Indeed, the removal of the two first data points in the inset of Figure 4 does not lead to any deviation of the data and therefore to the overall discussion and conclusions made. The excitation spectra collected at the emission maxima (500 nm) as a function of the time of irradiation are shown in Figure 5B. It is once more shown that the form to which the initial leuco is being converted is gradually becoming more emissive. One may question if the observed pattern is due to a 2829

dx.doi.org/10.1021/jp211996f | J. Phys. Chem. A 2012, 116, 2826−2832

The Journal of Physical Chemistry A

Article

Figure 6. Fluorescence emission spectra for DBMDI in dioxane obtained at two different temperatures (top panels 20 °C and bottom panels 14 °C) with (left) and without (right) irradiation as a function of time. Excitation is at 370 nm.

∼300 min could only be fitted with biexponential decay laws according to eq 5:

obtained for indigo (ϕR = 0.9) and DBMNI (ϕR = 0.007) in dioxane. From the definition of quantum yield (ϕ = amount of reactant consumed or product formed/amount of photons absorbed), it can be considered that for the same amount of light absorbed, the production of the cis-isomer is much more efficient with indigo than it is with DBMNI. This would explain the difference in the irradiation time needed to achieve a total isomerization (∼20 min to indigo and ∼100 min to DBMNI). Time-Resolved Dependence of the Trans−Cis Conversion with DBMNI. Fluorescence decays as a function of the irradiation time were obtained for DBMNI in dioxane at 20 °C. Indeed, if the trans to cis photoisomerization occurs, with coexistence of the two isomers that are both emissive, it is expected that these two will present different decay times, which was found to occur in the present study. In fact, the decays collected at the initial stages of irradiation and up to

I(t ) = a1e−t / τ1 + a2e−t / τ2

(5)

From these, two decay times with values of 120 ps and 2.17 ns could be obtained and were found roughly constant with the irradiation time; however, the associated pre-exponential factors were found to vary with the same time of irradiation. In Figure 8, the fluorescence decays, collected immediately after the preparation of the solution (Figure 8A) and upon 300 min (Figure 8B) of irradiation, are shown. The decays show the initial presence of the cis and trans leuco forms that gradually converge into a single decay time; this is consistent with the steady-state data (Figure 4) and with the above discussion on the (photo)conversion of the initial trans into the cis leuco form of DBMNI, leading to an almost monoexponential decay (the contribution of the second decay component represents less than 0.5% of the fluorescence decay, i.e., practically 2830

dx.doi.org/10.1021/jp211996f | J. Phys. Chem. A 2012, 116, 2826−2832

The Journal of Physical Chemistry A

Article

negligible), Figure 8B. Indeed, as can be seen from Figure 4, it is likely that the photostationary state still contains both the cis and trans conformers, even though the contribution of the former is strongly dominant. Moreover, this is further shown in the variation of the pre-exponential factors (a1 and a2) that mirror the concentration at time zero of these two forms of leuco-DBMNI, which evolve reciprocally with the time of irradiation, Figure 9.

Figure 7. Changes in fluorescence intensity at 490 nm for 4,4′dibutoxy-7,7′-dimethoxy-5,5′-dinitroindigo (DBMNI) in dioxane at 20 °C (A) and 14 °C (B) as a function of time, with and without irradiation at 370 nm. Figure 9. Dependence of the pre-exponential factors (ai) in eq 1 with the irradiation time for leuco DBMNI. The decay times in the end are roughly constant with values of ∼120 ps and 2.17 ns as indicated in the figure.



SUMMARY AND CONCLUSIONS A detailed investigation of the fluorescence emission dependence of the leuco form of indigo and of its derivative DBMNI with different irradiation times showed that this increases with time, which was found to be related to a photoisomerization process involving the conversion of a trans into a cis conformer. This was further supported by a model compound (where rotation around the central C−C bond is precluded) and by time-resolved data. Photochemical reaction quantum yields were also obtained, revealing a less efficient photoconversion process for the hexa-substituted indigo derivative, DBMNI.



AUTHOR INFORMATION

Corresponding Author

*Fax: 00351 239 827703. E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS



REFERENCES

We are grateful to POCI (project PTDC/QUI-QUI/099388/ 2008), Fundaçaõ para a Ciência e a Tecnologia (FCT), and FEDER for further funding. Dr. Gundula Voss (Bayreuth University) is acknowledged for the MpreCiba and 4,4′dibutoxy-7,7′-dimethoxy-5,5′-dinitroindigo samples. R.R. acknowledges FCT for a Ph.D. grant (SFRH/BD/38882/2007).

Figure 8. (A) Fluorescence decays and pulse instrumental response for 4,4′-dibutoxy-7,7′-dimethoxy-5,5′-dinitroindigo in dioxane, at T = 293 K, at initial time. (B) Fluorescence decay after 300 min of irradiation. Shown as insets are the decay times and pre-exponential factors. Also shown are the weighted residuals, autocorrelation functions (A.C.), and the χ2 values for a better judgment of the quality of the fits. λexc = 372 nm.

(1) Balfour-Paul, J. Indigo; British Museum Press: London, 2000. (2) Seixas de Melo, J. S.; Serpa, C.; Burrows, H. D.; Arnaut, L. G. Angew. Chem., Int. Ed. 2007, 46, 2094−2096. 2831

dx.doi.org/10.1021/jp211996f | J. Phys. Chem. A 2012, 116, 2826−2832

The Journal of Physical Chemistry A

Article

(3) Iwakura, I.; Yabushita, A.; Kobayashi, T. Chem. Lett. 2009, 38, 1020−1021. (4) Kobayashi, T.; Rentzepis, P. M. J. Chem. Phys. 1979, 70, 886− 892. (5) Yamazaki, S.; Sobolewski, A. L.; Domcke, W. Phys. Chem. Chem. Phys. 2011, 13, 1618−1628. (6) Elsaesser, T.; Kaiser, W.; Lüttke, W. J. Phys. Chem. 1986, 90, 2901−2905. (7) Nagasawa, Y.; Taguri, R.; Matsuda, H.; Murakami, M.; Ohama, M.; Okada, T.; Miyasaka, H. Phys. Chem. Chem. Phys. 2004, 6, 5370− 5378. (8) Seixas de Melo, J.; Rondão, R.; Burrows, H. D.; Melo, M. J.; Navaratnam, S.; Edge, R.; Voss, G. J. Phys. Chem. A 2006, 110, 13653− 13661. (9) Seixas de Melo, J. S.; Rondão, R.; Burrows, H. D.; Melo, M. J.; Navaratnam, S.; Edge, R.; Voss, G. ChemPhysChem 2006, 7, 2303− 2311. (10) Seixas de Melo, J.; Moura, A. P.; Melo, M. J. J. Phys. Chem. A 2004, 108, 6975−6981. (11) Miliani, C.; Romani, A.; Favaro, G. Spectrochim. Acta, Part A 1998, 54, 581−588. (12) Kirsch, A. D.; Wyman, G. M. J. Phys. Chem. 1977, 81, 413−420. (13) Rogers, D. A.; Margerum, J. D.; Wyman, G. M. J. Am. Chem. Soc. 1957, 79, 2464−2468. (14) Wyman, G. M.; Zarnegar, B. M. J. Phys. Chem. 1973, 77, 1204− 1207. (15) Wyman, G. M.; Zarnegar, B. M. J. Phys. Chem. 1973, 77, 831− 837. (16) Giuliano, C. R.; Hess, L. D.; Margerum, J. D. J. Am. Chem. Soc. 1968, 90, 587−594. (17) Weinstein, J.; Wyman, G. M. J. Am. Chem. Soc. 1956, 78, 4007− 4010. (18) Iwakura, I.; Yabushita, A.; Kobayashi, T. Bull. Chem. Soc. Jpn. 2011, 84, 164−171. (19) Iwakura, I. Phys. Chem. Chem. Phys. 2011, 13, 5546−5555. (20) Rondão, R. J.; Seixas de Melo, J. S.; Schaberle, F. B.; Voss, G. Phys. Chem. Chem. Phys. 2012, 14, 1778−1783. (21) Sousa, M. M.; Miguel, C.; Rodrigues, I.; Parola, A. J.; Pina, F.; Seixas de Melo, J. S.; Melo, M. J. Photochem. Photobiol. Sci. 2008, 7, 1353−1359. (22) Voss, G. J. Soc. Dyers Colour. 2000, 116, 87−90. (23) Rondao, R.; Seixas de Melo, J. S.; Voss, G. ChemPhysChem 2010, 11, 1903−1908. (24) Voss, G.; Gradzielski, M.; Heinze, J.; Reinke, H.; Unverzagt, C. Helv. Chim. Acta 2003, 86, 1982−2004. (25) Seixas de Melo, J. Chem. Educ. 2005, 10, 29−35. (26) Becker, R. S.; Seixas de Melo, J.; Maçanita, A. L.; Elisei, F. J. Phys. Chem. 1996, 100, 18683−18695. (27) Montalti, M.; Credi, A.; Prodi, L.; Gandolfi, M. T. Handbook of Photochemistry, 3rd ed.; CRC Press: Boca Raton, FL, 2006. (28) Arnaut, L.; Formosinho, S.; Burrows, H. Chemical Kinetics, From Molecular Structure to Chemical Reactivity; Elsevier: Oxford, 2007. (29) Pina, J.; Seixas de Melo, J.; Burrows, H. D.; Maçanita, A. L.; Galbrecht, F.; Bunnagel, T.; Scherf, U. Macromolecules 2009, 42, 1710−1719. (30) Striker, G.; Subramaniam, V.; Seidel, C. A. M.; Volkmer, A. J. Phys. Chem. B 1999, 103, 8612−8617. (31) Squillacote, M. E.; Sheridan, R. S.; Chapman, O. L.; Anet, F. A. L. J. Am. Chem. Soc. 1979, 101, 3657−3659.

2832

dx.doi.org/10.1021/jp211996f | J. Phys. Chem. A 2012, 116, 2826−2832