J. Phys. Chem. B 2006, 110, 13985-13990
13985
Radiative Relaxation in Synthetic Pheomelanin Jennifer Riesz,*,† Tadeusz Sarna,‡ and Paul Meredith† Soft Condensed Matter Physics Group and Centre for Biophotonics and Laser Science, School of Physical Sciences, UniVersity of Queensland, St. Lucia, Brisbane, Queensland 4072, Australia, and Department of Biophysics, Faculty of Biotechnology, Jagiellonian UniVersity, Krakow, Poland ReceiVed: August 29, 2005; In Final Form: May 18, 2006
We report a detailed photoluminescence study of cysteinyldopa-melanin (CDM), the synthetic analogue of pheomelanin. Emission spectra are shown to be a far more sensitive probe of CDM’s spectroscopic behavior than are absorption spectra. Although CDM and dopa-melanin (DM, the synthetic analogue of eumelanin) have very similar absorption spectra, we find that they have very different excitation and emission characteristics; CDM has two distinct photoluminescence peaks that do not shift with excitation wavelength. Additionally, our data suggest that the radiative quantum yield of CDM is excitation energy dependent, an unusual property among biomolecules that is indicative of a chemically disordered system. Finally, we find that the radiative quantum yield for CDM is ∼0.2%, twice that of DM, although still extremely low. This means that 99.8% of the energy absorbed by CDM is dissipated via nonradiative pathways, consistent with its role as a pigmentary photoprotectant.
Introduction Melanin is an important biological macromolecule found throughout nature. Two types, black/brown eumelanin and red/ yellow pheomelanin, are pigments found in human skin and hair, where they are thought to act as photoprotectants.1 Eumelanin, pheomelanin, and their synthetic precursors have been studied extensively due to their connection to melanoma skin cancer.2-5 5-S-cysteinyldopa, a precursor of pheomelanin, is currently the best marker of malignant melanoma.6,7 Melanoma carries a very unfavorable prognosis, and early detection is essential for effective treatment. There remains debate as to the exact relationship between the color of skin, its pheomelanin and eumelanin content, and susceptibility to skin cancer. Although melanin appears to be photoprotective,8,9 it has been suggested that it might also be photosensitizing, especially pheomelanin. It has been shown that pheomelanin becomes mutagenic after irradiation by UV light10,11 and it has been suggested that this is why people with red hair are more susceptible to skin cancer.4,12,13 Other studies suggest that the ratio of pheomelanin to eumelanin is the important parameter; those with a higher ratio having a higher risk of melanoma.14,15 Still other studies suggest that pheomelanin may be photoprotective or photosensitizing depending upon the concentration of the pigment and the wavelength of irradiation.16 This uncertainty is partially due to the fact that melanins, and particularly pheomelanins, are hard to study. They are difficult to extract, insoluble in most common solvents, and strongly absorbing with a very low radiative quantum yield.17 These factors combine to make chemical and spectroscopic analysis challenging. Even very fundamental properties, such as an accurate picture of the molecular structure (at the primary and secondary levels), remain in question.18 It has recently been proposed that disorder plays a significant role in the secondary structure and optical properties of * Corresponding author. Tel.: +61 7 3365 3406; Fax: +61 7 3365 1242; E-mail:
[email protected]. † University of Queensland. ‡ Jagiellonian University.
eumelanin.19-24 Since there are several different eumelanin monomers that may cross-link through various different sites, it is possible that a sample of eumelanin may consist of a large number of distinct oligomers.25 Density functional theory calculations suggest that each of these oligomers may be expected to have a slightly different HOMO-LUMO energy gap.21,24,26-30 A disordered collection of oligomers may therefore exhibit broadband absorbance, since its spectrum will be the sum of many absorbance peaks at slightly different energies. This is currently the favored structural model for eumelanins, but pheomelanins have been studied far less extensively, and their structure should not be assumed to be identical. The pheomelanin precursors, benzothiazines, have fewer likely crosslinking sites than the eumelanin precursors.31 They are also less likely to form flat sheets when polymerized (as has been proposed for eumelanins), for steric reasons. Some structural studies have been conducted for pheomelanins, including radiotracer biosynthetic methods and NMR spectroscopy by Deibel and Chedekel et al.,32-34 biosynthetic studies by Napolitano et al.,35,36 and, recently, studies of whole melanosomes by Liu et al.37 Despite these careful studies, however, a clear and accurate picture of the primary and secondary structure of pheomelanin remains elusive, even for synthetic analogues where it might be thought that the structure would be better known. Determining the nature of the structural differences between eumelanins and pheomelanins is important for understanding their spectroscopic differences.38 A clear characterization of the spectroscopic properties of melanins is essential for understanding their biological functionality as a photoprotectant. Pheomelanins have been studied far less extensively than eumelanins in this respect. The absorption spectrum of pheomelanin is well known to be broadband and exponential (as shown in Figure 1), although the mechanism that produces this unusual shape is not clear. There have been several studies of the ultrafast absorption dynamics and energy transfer of pheomelanins, revealing that they have very short decay lifetimes.39-42 It has also been proposed that the ratio of absorbances at 650 nm and 500 nm
10.1021/jp054869l CCC: $33.50 © 2006 American Chemical Society Published on Web 06/27/2006
13986 J. Phys. Chem. B, Vol. 110, No. 28, 2006
Figure 1. Absorbance of CDM solution (0.001 wt %). The absorbance spectrum stabilized completely after approximately 10 days, and very little further change was observed.
can be used as an indicator of the relative amounts of eumelanin and pheomelanin in solution.43 Otherwise, the bulk of the spectroscopic literature has focused on eumelanins, rather than their more photoactive counterpart, the pheomelanins. As stated earlier, the strong broadband absorbance and very low quantum yields of melanins makes spectroscopic analysis challenging. These factors mean that any photoluminescence spectra will be strongly affected by reabsorption and inner filter effects that will distort their line shapes significantly. Here, we present accurate photoluminescence spectra for pheomelanin that have been corrected for these effects, using a method recently devised, detailed in ref 44. The use of this method has also allowed us to accurately measure the radiative quantum yield of pheomelanin, also presented in this paper. We have used cysteinyldopa-melanin (CDM) and dopa-melanin (DM) as synthetic analogues for pheomelanin and eumelanin, respectively. These are known to be good models for human melanin.31,45 Experimental Section Sample Preparation. CDM was derived from the tyrosinasecatalyzed oxidation of a mixture of L-dopa and L-cysteine in the ratio 1:1.5, as described fully in ref 46. The powder was diluted to a range of concentrations (0.001% to 0.005 wt % macromolecule) in high purity 18.2MΩ MilliQ deionized water. To aid solubility, the solutions were adjusted using NaOH to ∼ pH10 and gently heated with sonication. Pale brown, apparently continuous dispersions were produced. Between measurements solutions were stored in darkness at 4 °C. Dopamelanin derived by the nonenzymatic oxidation of tyrosine was purchased from Sigma Aldrich (Sydney, Australia) and used without further purification. Although DM could be prepared in an identical fashion to the CDM used in this study (in the absence of cysteine), DM from Sigma Aldrich has been used extensively in previous spectroscopic studies and therefore provides an excellent benchmark for these measurements. Solutions of DM were prepared in an idential fashion at pH∼ 11.5 (necessary for complete solubility). Fluorescein was used as a standard for the determination of the radiative quantum yield of the melanins. It was purchased from Sigma Aldrich and used without further purification to prepare standard solutions at five different concentrations (1 × 10-5% to 4 × 10-5% by weight) in 0.1 M NaOH solution (18.2 MΩ MilliQ deionized water). Absorption Spectrometry. Absorption spectra were recorded using a Perkin-Elmer Lambda 40 spectrophotometer. A scan speed of 240 nm/min and slit width of 3 nm band-pass were used. Spectra were collected using a quartz 1 cm square cuvette. Solvent scans (obtained under identical conditions) were used for background correction. To ensure that absorption measure-
Riesz et al.
Figure 2. Comparison of the absorbance spectrum for CDM (at 10 days) and DM. The curves show signficantly different shapes, although both are broadband and almost entirely featureless. Spectra have been scaled for shape comparison. Both have an excellent exponential fit, parameters of which are shown in Table 1.
ments were not skewed by scattering effects, integrated scattering measurements were made using a Perkin-Elmer Lambda 40 spectrophotometer with an integrating sphere attachment (Labsphere RSA-PE-20 reflectance spectroscopy accessory). Scattering was found to be negligible at all wavelengths.47 Photoluminescence Emission and Excitation Spectrometry. Photoluminescence emission and excitation spectra were recorded using a Jobin Yvon FluoroMax 3 fluorimeter with a slit width of band-pass 3 nm and an integration of 0.5 s. Background scans were performed under identical instrumental conditions using the relevant solvents. Spectra were collected using a quartz 1 cm square cuvette. Spectra were corrected to account for differences in pump beam power at different excitation wavelengths using a reference beam. All spectra were corrected for reabsorption and inner-filter effects using the method outlined in ref 44. Quantum yields were calculated using the method outlined in ref 48 with standard values from ref 49. Since the fluorescein standard has a much higher quantum yield than the melanin samples, it was necessary to use a neutral density filter (OD = 2) to reduce the fluorescence intensity to a level measurable by the fluorimeter at the same settings. The absorbance spectrum of the filter was then measured, and the fluorescein photoluminescence spectra were accurately corrected for its presence. Results and Discussion The absorbance spectrum of CDM was monitored over a period of 24 days to check stability. It was discovered that, unlike DM, CDM took a period of time (approximately 10 days) to stabilize completely (Figure 1). The absorbance is characteristically broadband and increases monotonically toward higher energies. Initially, small peaks were observed at approximately 240 nm and 330 nm; these decreased gradually over a period of 10 days, after which time very little change in the spectrum was observed. These peaks are attributed to small amounts of precursors (potentially benzothiazine units) which eventually polymerized. EPR measurements were conducted (not shown here) to determine the exact nature of the sample change over the first 10 day period. The results suggest that this is a complex process that consists of both melanin autoxidation and high pHinduced polymerization. Further studies are being conducted along these lines and shall be reported in a later publication. Figure 2 compares the absorption spectra of CDM and DM. Both increase monotonically toward higher energies, although CDM does so more steeply. Since the cysteinyldopa monomer
Radiative Relaxation in Synthetic Pheomelanin
J. Phys. Chem. B, Vol. 110, No. 28, 2006 13987
TABLE 1: Fit Parameters for Melanin Absorbance Spectra (shown in Figure 2), as Defined in Eq 1a
CDM DM
k0 (dimensionless)
k1 (dimensionless)
k2 (nm-1)
0.0013 -0.0073
4.19 1.98
0.0092 0.0061
a Both solutions were at a concentration of 0.001 wt %, with a 1 cm absorption path length. CDM values are taken at 10 days, when the solution had stabilized.
has fewer cross-linking sites than either DHI or DHICA (dihydroxyindole and dihydroxyindole-carboxylic acid, the DM monomers,1,50), it could be expected that CDM would be less disordered than DM at the secondary structural level. Possibly, then, there are fewer oligomers with absorbances in the visible range, giving a steeper absorbance curve. Both spectra are extremely close to exponential in shape in wavelength space. Fit parameters are shown in Table 1, when fitted to eq 1.
A ) k0 + k1e-k2λ
(1)
Figure 3 shows the change in the photoluminescence spectrum of CDM as the solution stabilized, excited at 350, 380, and 410 nm. As for the absorbance spectrum, the changes observed are believed to be due to a combination of high pH-induced polymerization of remnant low molecular weight precursors and pheomelanin autoxidation (which is known to increase photoluminescence51). We do not believe this process to be dominated by oxidation, however, since this would also result in increased absorbance at long wavelengths (which is not observed). Figure 4 compares the photoluminescence spectrum of stabilized CDM and DM excited at 380 nm. The accurate quantification of these spectra has been made possible by the use of the reabsorption correction method outlined in ref 44. DM has one broad photoluminescence peak, occurring at approximately 485 nm (when excited at 380 nm; the position of this peak shifts when the sample is excited at different wavelengths17). In contrast, CDM exhibits two clearly resolved broad peaks, which do not shift as the excitation wavelength is varied. This is an extremely important difference between CDM and DM. The CDM spectra can all be fitted very closely to two Gaussians, as given by eq 2. -[(E-E1)/D1]2
I ) N 1e
+ N2e-[(E-E2)/D2 + A ]2
(2)
Fits were performed in energy space (E) since we expect line broadening to be characteristically Gaussian in energy (not in wavelength). All fits to a simple two Gaussian model were excellent; a representative example is show in Figure 5. The parameters of these fits are given in Table 2. The peaks occur consistently at the average values of 2.27 ( 0.08 eV (548 nm) and 2.72 ( 0.08 eV (458 nm) in all spectra. These two peaks could possibly be due to two types of chemically distinct populations with transitions suitably inhomogeneously broadened. The broadness of the peaks may be associated with photoluminescence from a disordered collection of macromolecules, although the prescence of two distinct peaks suggests that the level of disorder may be lower than for DM. First principles quantum chemical calculations are in progress to probe this hypothesis. It is possible that the synthetic method for the CDM used here could produce a sample that is a mixture of CDM and DM. This would produce two peaks in the CDM spectrum (one due to pure CDM, the other due to DM). We believe that this is not the case, however, since neither of the peaks coincide
Figure 3. Photoluminescence of CDM solution (0.001 wt %). Only key time points are shown for clarity. The spectrum stabilized after approximately 10 days, and little further change was observed. The small peak evident at approximately 400 nm in some spectra is a remnant of the Raman scattering peak that was not completely removed by background subtraction due to lamp intensity fluctuations.
Figure 4. Photoluminescence of CDM (stabilized at 10 days) and DM with excitation at 380 nm.
with the separately measured DM peak when excited at the same energy, and they do not shift with excitation energy in the same
13988 J. Phys. Chem. B, Vol. 110, No. 28, 2006
Riesz et al.
Figure 5. Photoluminescence of CDM (6 days, 380 nm excitation) with double Gaussian fit (as per eq 2). The excellent fit of the double Gaussian is representative of all spectra.
way. Interestingly, Gallas and Eisner52 observed a double peaked DM spectrum (not CDM, as observed to have a double peak here), although their spectra showed an isosbestic point at 470 nm which we do not observe in either our CDM or DM spectra. This double peaked spectrum for DM has not been observed in other studies and is possibly due to the different pH at which the measurements were conducted (this will affect the degree of ionization of the sample, which influences the molecular species that are present). Alternatively, it may reflect a slightly different sample preparation, which can affect the degree of carboxylation of the sample. That these small differences can be detected reflects the extreme sensitivity of emission spectra as a melanin probe. The relative heights of the two peaks were determined from the Gaussian fitting (N1 and N2, listed in Table 2). These values have been plotted in Figure 6. Over time we see that the higher energy peak (at 458 nm) increases in intensity and then appers to stabilize (except when excited at 350 nm), whereas the lower energy peak (at 548 nm) increases initially, then decreases. This suggests that the concentrations of the two sub-populations of oligomers responsible for these peaks are changing over time as the CDM polymerizes. Possibly, the lower energy peak is due to a small molecular weight species that gradually polymerizes into larger units. Figure 7 shows photoluminescence excitation data for CDM. These spectra were measured after the CDM had stablized (at 10 days) and have been fully corrected for reabsorption and inner filter effects. The detection wavelengths (465 nm and 550 nm) were selected because they are close to the positions of the two peaks observed in the photoluminescence spectra. We
Figure 6. Relative heights of the two Gaussian CDM emission peaks over time (parameters N1 and N2 as defined in eq 2, and listed in Table 2).
Figure 7. Photoluminescence excitation of CDM solution (0.001 wt %) after 10 days.
can see that the higher energy peak (at 458 nm) will show maximum emission when excited at approximately 380 nm. The lower energy peak (at 548 nm) will show maximum emission when excited at approximately 395 nm. In simple systems, the photoluminescence excitation spectrum typically has the same shape as the absorption spectrum (because exciting at a wavelength where there is greater absorption will produce
TABLE 2: Gaussian Fit Parameters to CDM Photoluminescence Spectra (parameters as defined in eq 2) λexcitation (nm)
E1 (eV)
E2 (eV)
N1 (cps)
N2 (cps)
D1 (eV-1)
D2 (eV-1)
A (cps)
350 (0 days) 380 (0 days) 410 (0 days) 350 (6 days) 380 (6 days) 410 (6 days) 350 (10 days) 380 (10 days) 410 (10 days)
2.17 2.24 2.26 2.21 2.31 2.27 2.22 2.34 2.33
2.59 2.69 2.72 2.73 2.76 2.73 2.76 2.75 2.75
24800 69200 78100 103000 188000 187000 15200 146000 143000
80600 74100 43900 183000 236000 143000 227000 226000 126000
0.227 0.286 0.301 0.248 0.329 0.270 0.181 0.280 0.284
0.483 0.287 0.235 0.369 0.222 0.228 0.540 0.257 0.245
7110 2100 -1350 15500 -1690 -383 1150 4790 -258
Radiative Relaxation in Synthetic Pheomelanin
J. Phys. Chem. B, Vol. 110, No. 28, 2006 13989
Figure 9. Quantum yield of CDM (at 10 days) and DM (data for DM from ref 17, used with author’s permission).
Figure 8. Quantum yield of CDM over time.
TABLE 3: Quantum Yield Values for CDM (at 10 days) excitation wavelength (nm)
quantum yield (%)
350 380 410
0.143 ( 0.008 0.19 ( 0.01 0.19 ( 0.01
greater photoluminescence, and in simple systems the quantum yield is independent of excitation wavelength), but this is clearly not the case here. This suggests that the processes of photon absorption and photon emission are not closely linked in CDM (they are likely separated by a great deal of coupling to phonon modes of the system, or other nonradiative relaxation processes). This result highlights the need for accurate radiative relaxation quantum yield data for cystienyldopa-melanin, to better understand the relationship between photon absorption and emission. The use of the reabsorption correction method outlined in ref 44 allowed accurate measurement of the radiative quantum yield of CDM, shown in Figure 8. The yield was monitored for a period of 24 days. Note that the yield is extremely small, approximately 0.2%. This means that 99.8% of the energy absorbed by CDM is dissipated via nonradiative pathways. These nonradiative pathways may include vibrational coupling (consistent with pheomelanin’s biological role as a photoprotectant) or excited state chemistry (aerobic photoreactions of melanin are known to be responsible for the generation of superoxide anion and hydrogen peroxide, both of which are cytotoxic and could potentially lead to melanoma skin cancer). Calorimetry measurements for pheomelanins would allow a distinction between these. We see that over time the yield initially increased and appears to have stabilized after the first 10 days, with some minor fluctuations. The stabilized quantum yield values at each excitation wavelength are shown in Table 3. It is clear that, unlike most molecules, the quantum yield of CDM is dependent upon the excitation wavelength.53 This lends weight to the structural model of CDM as a disordered collection of oligomers, since in that model we would be selectively exciting sub-populations of oligomers which may have different quantum yields. It is interesting to note that the yields when excited at 380 nm and 410 nm are very similar at all time points, but the yield at 350 nm is always significantly lower. This is because the consistently higher absorption at 350 nm is not reflected in an increase in photoluminescence when excited at that wavelength (the integrated intensity is very similar to that when excited at 380 nm). This, combined with the fact that the peak height results (Figure 6) showed differing behavior for
excitation at 350 nm, suggests that the relaxation process at this wavelength is different to that at 380 nm and 410 nm. Interestingly, there is a trough centered around 350 nm in the photoluminescence excitation spectra (Figure 7) for both detection wavelengths. Further spectroscopic studies should take this into account by observing behavior when excited at 350 nm, in addition to other wavelengths.Most significantly, the radiative quantum yield of CDM, although very small, is approximately twice that of DM. This perhaps reflects the fact that eumelanin is known to be a better photoprotectant than pheomelanin. Since relaxation via phonon modes occurs on a significantly shorter time scale than photoluminescence (picoseconds, rather than nanoseconds53), a larger number of CDM oligomers will remain excited until relaxing radiatively, potentially explaining pheomelanin’s higher photoreactivity. We also see that the yield for CDM follows a different trend with excitation wavelength to that for DM, indicating that although these two molecules are very similar in many ways, they do have different energy dissipation behavior. Conclusions In this paper we present extensive results of cysteinyldopamelanin (CDM, the synthetic analogue of pheomelanin) photoluminescence emission and excitation. We believe this to be the most complete study to date of these properties. We find that the spectra are fundamentally different to those for dopamelanin (DM, the synthetic analogue of eumelanin), having two distinct broad peaks that do not shift with excitation wavelength. We have also accurately measured radiative quantum yields for CDM. We find that although the yield is twice that of DM, it is extremely low (0.2%), meaning that 99.8% of the absorbed energy is dissipated via nonradiative pathways. This is consistent with melanin’s role as a primary photoprotectant, and potentially leads to an understanding of the greater photoreactivity of pheomelanin over eumelanin. Acknowledgment. This work has been supported in part by the Australian Research Council, the UQ Centre for Biophotonics and Laser Science, and the University of Queensland (RIF scheme). References and Notes (1) Prota, G. Melanins and Melanogenesis; Academic Press: New York, 1992. (2) Wolbarsht, M.; Walsh, A.; George, G. Appl. Opt. 1981, 20, 21842186. (3) Forest, S.; Nofsinger, J.; Simon, J. J. Phys. Chem. B 1999, 103, 11428-11432. (4) Hill, H.; Zeise, L.; Chedekel, M.; Fitzpatrick, T. Is melanin photoprotective or photosensitizing? In Melanin: Its role in human photoprotection; Vladenmar Press, Overland Park, KS., 1995.
13990 J. Phys. Chem. B, Vol. 110, No. 28, 2006 (5) Zeise, L.; Kollias, N.; Sayre, R.; Chedekel, M. Photochem. Photobiol 1991, 9, 135-160. (6) Hartleb, J.; Arndt, R. J. Chromatogr. B 2001, 764, 409-443. (7) Rorsman, H. Pigment Cell Res. 2004, 17, 191-202. (8) De Leeuw, S.; Smit, N.; Van Veldhoven, M.; Pennings, E.; Pavel, S.; Simons, J.; Schothorst, A. J. Photochem. Photobiol. B: Biology 2001, 61, 106-113. (9) Ortonne, J. British J. Dermatol. 2002, 146 (Suppl. 61), 7-10. (10) Harsanyi, Z.; Post, P.; Brinkmann, J.; Chedekel, M.; Deibel, R. Experientia 1980, 36, 291-292. (11) Takeuchi, S.; Zhang, W.; Wakamatsu, K.; Ito, S.; Hearing, V.; Kraemer, K.; Brash, D. Proc. Natl. Acad. Sci. U.S.A. 2004, 101(42), 1507615081. (12) Bliss, J.; Ford, D.; Swerdlow, A.; Armstrong, B.; Cristofolini, M.; Elwood, J.; Green, A.; Holly, E.; Mack, T.; Mackie, R.; Osterlind, A.; Walter, S.; Peto, J.; Easton, D. Int. J. Cancer 1995, 62, 367-376. (13) Elwood, J.; Whitehead, S.; Davison, J.; Stewart, M.; Galt, M. Int. J. Epidemiol. 1990, 19, 801-810. (14) Hennessy, A.; Oh, C.; Diffey, B.; Wakamatsu, K.; Ito, S.; Rees, J. Pigment Cell Res. 2005, 18, 220-223. (15) Vincensi, M.; d’Ischia, M.; Napolitano, A.; Procaccini, E.; Riccio, G.; Monfrecola, G.; Santoianni, P.; Prota, G. Melanoma Res. 1998, 8, 5358. (16) Hill, H.; Hill, G. Pigment Cell Res. 2000, 13 (Suppl. 8), 140-144. (17) Meredith, P.; Riesz, J. Photochem. Photobiol. 2004, 79(2), 211216. (18) Zajac, G.W.; Gallas, J. M.; Cheng, J.; Eisner, M.; Moss, S. C.; Alvarado-Swaisgood, A. E. Biochim. Biophys. Acta 1994, 1199, 217-278. (19) Tran, L.; Powell, B.; Meredith, P. Biophys. J. 2006, 90(3) 743752. (20) Meredith, P.; Powell, B.; Riesz, J.; Nighwander-Rempel, S. P.; Pederson, M. R.; Moore, E. G. Soft Matter 2006, 2, 37-44. (21) Powell, B.; Baruah, T.; Bernstein, N.; Barake, K.; McKenzie, R.; Meredith, P.; Pederson, M. J. Chem. Phys. 2004, 120, 8608. (22) Meredith, P.; Powell, B.; Riesz, J.; Vogel, R.; Blake, D.; Subianto, S.; Will, G.; Kartini, I. Broad Band Photon-harvesting Biomolecules for Photovoltaics. In Artificial Photosynthesis: From Basic Biology to Industrial Application; Collings, A. F., Critchley, C., Eds.; Wiley: Chichester, 2005. (23) Bochenek, K.; Gudowska-Nowak, E. Chem. Phys. Lett. 2003, 373, 532. (24) Powell, B. Chem. Phys. Lett. 2005, 402, 111-115. (25) Galvao, D.; Caldas, M. J. Chem. Phys. 1990, 92(4), 2630-2636. (26) Il’ichev, Y.; Simon, J. J. Phys. Chem. B 2003, 107, 7162-7171. (27) Stark, K.; Gallas, J.; Zajac, G.; Eisner, M.; Golab, J. J. Phys. Chem. B 2003, 107, 3061-3067.
Riesz et al. (28) Stark, K.; Gallas, J.; Zajac, G.; Eisner, M.; Golab, J. J. Phys. Chem. B 2003, 107, 11558-11562. (29) Bolivar-Marinez, L.; Galvao, D.; Caldas, M. J. Phys. Chem. B 1999, 103, 2993-3000. (30) Stark, K.; Gallas, J.; Zajac, G.; Golab, J.; Gidanian, S.; McIntire, T.; Farmer, P. J. Phys. Chem. B 2005, 109, 1970-1977. (31) Ito, S. Pigment Cell Res. 2003, 16, 230-236. (32) Deibel, R.; Chedekel, M. J. Am. Chem. Soc. 1982, 104, 73067309. (33) Deibel, R.; Chedekel, M. J. Am. Chem. Soc. 1984, 106, 58845888. (34) Chedekel, M.; Subbarao, K.; Bhan, P.; Schultz, T. Biochim. Biophys. Acta 1987, 912, 239-243. (35) Napolitano, A.; Memoli, S.; Crescenzi, O.; Prota, G. J. Org. Chem. 1996, 61, 598-604. (36) Napolitano, A.; Donato, P.; Prota, G. Biochim. Biophys. Acta 2000, 1475, 47-54. (37) Liu, Y.; Hong, L.; Wakamatsu, K.; Ito, S.; Adhyaru, B.; Cheng, C.; Bowers, C.; Simon, J. Photochem. Photobiol. 2005, 81, 135-144. (38) Losi, A.; Bedotti, R.; Brancaleon, L.; Viappiani, C. J. Photochem. Photobiol. B: Biol. 1993, 21, 69-76. (39) Ye, T.; Simon, J. J. Phys. Chem. B 2002, 106, 6133-6135. (40) Ye, T.; Simon, J. Photochem. Photobiol. 2003, 77(1), 1-4. (41) Ye, T.; Simon, J. J. Phys. Chem. B 2003, 107, 11240-11244. (42) Ye, T.; Simon, J. Photochem. Photobiol. 2003, 77(1), 41-45. (43) Ozeki, H.; Ito, S.; Wakamatsu, K.; Thody, A. Pigment Cell Res. 1996, 9, 265-270. (44) Riesz, J.; Gilmore, J.; Meredith, P. Spectrochim. Acta A 2005, 61, 2153-2160. (45) Prota, G.; Nicolaus, R. Gazz. Chim. Ital. 1967, 97, 665-684. (46) Rozanowska, M.; Sarna, T.; Land, E.; Truscott, T. Free Radical Biol. Med. 1999, 26(5-6), 518-525. (47) Riesz, J.; Gilmore, J.; Meredith, P. Biophys. J. 2006, 90(11), 1-8. (48) Horiba, J. Y. http://www.jobinyvon.co.uk/jy/fluorescence/plqy.htm. (49) Magde, D.; Wong, R.; Seybold, P. Photochem. Photobiol. 2002, 75(4), 327-334. (50) Ito, S. Biochim. Biophys. Acta 1986, 883, 155-161. (51) Kayatz, P.; Thumann, G.; Luther, T.; Jordan, J.; Bartz-Schmidt, K.; Esser, P.; Schraermeyer, U. InVest. Opthemol. Visual Sci. 2001, 41(1), 241-246. (52) Gallas, J.; Eisner, M. Photochem. Photobiol. 1987, 45(5), 595600. (53) Lakowicz, J. Principles of Fluorescence Spectroscopy, 2nd ed.; Kluwer Academic/Plenum Publishers: New York, 1999.
