- Email: [email protected]
Photophysical properties of N-acetyl-menthyl anthranilate
Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109–116 www.elsevier.com / locate / jphotobiol
Photophysical properties of N-acetyl-menthyl anthranilate a, b Andrew Beeby *, Allison E. Jones a b
Department of Chemistry, University of Durham, South Road, Durham DH1 3 LE, UK Centre for Forensic Science, University of Central Lancashire, Preston PR1 2 HE, UK
Abstract The results of a comprehensive investigation of the photophysical properties of the sunscreen analogue, N-acetyl menthyl anthranilate (NAMA), in various solvent systems are reported. Luminescence studies reveal that this compound is fluorescent (Ff 50.1660.01) in toluene and has a solvent dependent emission maximum in the range 363–370 nm. Phosphorescence has also been detected in low temperature glasses with an emission maximum at 420 nm in EPA, and a lifetime of 1.3 s; the triplet energy was found to be 31163 kJ mol 21 . Kinetic UV–visible absorption measurements revealed a transient species with absorption maxima at 450 nm and solvent dependent lifetimes of 120–240 ms which are attributed to the triplet state. The triplet state is efficiently quenched by oxygen, leading to the formation of singlet oxygen in all of the solvent systems studied. The singlet oxygen quantum yields (FD ), determined by time-resolved near-infrared luminescence measurements, were in the range 0.19–0.21. 2001 Elsevier Science B.V. All rights reserved. Keywords: Sunscreen; Singlet oxygen; Sensitiser; UV radiation
1. Introduction The past decade has seen an increase in the use of sunscreen formulations resulting from an increased awareness of the damaging effects of solar UV radiation on human skin [1]. The recent association of UVA exposure with skin-ageing and other dermatological conditions has furthered this increase [2]. Shaath [3] describes the anthranilates as an ‘interesting’ class of UV filters and describes in detail the two derivatives approved for commercial use, menthyl anthranilate in the USA and N-acetyl-homomenthyl anthranilate in the European Union (EU), see Fig. 1. He comments on the ‘ortho-effect’ that is evident in these compounds, namely the red-shift of the absorption maxima and low molar absorption coefficients when compared to their para-disubstituted counterparts. This shift is attributed to the ease of electron delocalisation in the ortho-compounds as opposed to the para-compounds. However, steric crowding due to the ortho-disubstitution in these compounds causes deviations from coplanarity of the molecules resulting in a reduction of the intensity of the absorption and lower molar absorption coefficients. Agrapidis-Paloymis et al. [4] have commented that the *Corresponding author. Tel.: 144-191-3744-623; fax: 144-191-3844737. E-mail address: [email protected] (A. Beeby).
anthranilates, like the salicylates, are stable and safe compounds to use, owing to this ortho-disubstituted relationship, and do not exhibit significant solvent effects in cosmetic formulations. The work presented here is a continuation of our study of the anthranilate esters used as sunscreens. We have previously reported work on menthyl anthranilate [5], a compound that is currently used in formulations sold in the USA. We have found it to be highly fluorescent, both in solution (Ff |0.6) and when part of commercially available sunscreen formulations. It was also found that a significant proportion of the excited state molecules undergo intersystem crossing to the triplet state, which was long lived in solution (20–200 ms) and had a long phosphorescence lifetime (tp 52.5 s). The energy of the triplet state was estimated from the phosphorescence spectrum to be |287
Fig. 1. The molecular structure of N-acetyl-homomenthyl anthranilate.
1011-1344 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S1011-1344( 01 )00197-X
110
A. Beeby, A.E. Jones / Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109 – 116
kJ mol 21 , high enough to efficiently sensitise the formation of singlet oxygen (FD 50.10). Despite the apparent advantages of these compounds the N-acetyl-homomenthyl anthranilate derivative was removed from the list of approved sunscreens in the EU in 1989 [6], although the reasons for its withdrawal have not been published. Furthermore there is surprisingly little information on this compound in the scientific and medical literature given the interest in using N-acetyl-homomenthyl anthranilate in commercial sunscreens. It might be expected that its photophysical and photochemical properties would have been investigated at the time of its admission onto the list of approved sunscreens but there has been little published on its photophysics and photochemistry or that of the parent acid and related esters. Several UV–visible spectroscopy studies have been carried out on N-acetyl anthranilic acids. Dannenberg [7] determined the absorption spectra of benzoic acid and a range of its derivatives, including the acetamido derivative, to compare and quantify the shifts in absorption bands caused by the different ring substitutions. Ungnade [8] has studied the ultraviolet absorption spectra of a range of compounds and interpreted the intensities and lmax values in terms of the electronic and steric effects of the substituents. The spectrum of N-acetyl anthranilic acid has been reported to have bands at 221, 252 and 305 nm, and the molar absorption coefficient at 305 nm has been determined to be 4650 dm 3 mol 21 cm 21 . The red shift and increased intensity of the 252 nm band, and the blue shift and reduction in the intensity of the 305 nm band relative to the spectrum of anthranilic acid have been attributed to the increase in conjugation of the system following acylation of the amine group and reduced conjugation between the nitrogen atom and the aromatic ring arising from the loss of planarity of the molecule, respectively. Grammaticakis [9] recorded the UV spectra of a range of N-substituted o-nitro and o-carboxy anilines including Nacetyl anthranilic acid. The peak shifts and intensity changes of the absorption bands following substitution of the alkyl in the ortho position by –NO 2 , –COOH and –COCH 3 were studied. It was found that substitution had little or no effect on the absorption bands. Further work by Grammaticakis [10] studied the near-ultraviolet and visible absorption of some isomeric arylamines and their nitrogen substituted derivatives. In compounds of the form XC 6 H 5 NHCOCH 3 , acetylation of the amine group was found to cause a red shift in the middle band (|250 nm) and a blue shift in the band at longer wavelength (|300 nm), in agreement with the work by Ungnade. The photolysis of N-acetyl anthranilic acid has been carried out by Staudemayer and Roberts [11], in order to study the neighbouring group participation in the cyclization rearrangements known to occur in acetanilides under UV light. They irradiated solutions of N-acetyl anthranilic acid at 254 nm for 100 h and monitored the changes using
UV spectroscopy. The resulting solution was found to be a mixture containing the starting material, anthranilic acid, 3-aceto-2-aminobenzoic acid, and 5-aceto-2-aminobenzoic acid, all products of photo-Fries-like cleavage and rearrangements. Since photo-Fries and similar reactions have been postulated to occur from singlet excited states [12] further experiments were conducted to probe the reactions of the triplet state of the acid species. In benzene solution they observed sensitisation of the triplet state of the acid by energy transfer from the triplet state of the solvent. The photo-product formed in under these conditions was identified as benzoxazinone. From these studies they concluded that N-acetyl anthranilic acid forms photo-Frieslike cleavage and rearrangement products via the excited singlet state, and that it cyclizes to benzoxazinone via the triplet state. Several workers have reported that N-acetyl anthranilic acid is luminescent, although there is surprisingly little quantitative data in the literature. Staudemayer and Roberts determined the luminescence spectra of N-acetyl anthranilic acid as part of the work described above in order to determine a suitable triplet state sensitiser. They found that in acetonitrile fluorescence emission occurred with a maximum at 367 nm and phosphorescence emission with a maximum at 428 nm. In this paper we present the results of a solution state study of N-acetyl-menthyl anthranilate (Fig. 2) in a range of solvent systems. The menthyl ester was chosen for the study due to its ease of synthesis. Based upon our previous work we have demonstrated that the properties of the esters are essentially independent of the alkyl group: the methyl and menthyl esters of anthranilic acid are identical [13]. The work described here includes measurement of fluorescence and phosphorescence emission spectra and lifetimes, quantum yields of fluorescence, triplet–triplet absorption spectra and the determination of oxygen- and self-quenching rate constants. Significantly we have shown that this N-acetyl anthranilate ester has an energetic, long lived triplet state which can sensitise the formation of singlet oxygen, a species known to be harmful to biological systems [14]. The quantum yields of singlet oxygen formation have been determined to be in the range 0.19– 0.21 in the solvent systems studied.
Fig. 2. The molecular structure of N-acetyl-menthyl anthranilate.
A. Beeby, A.E. Jones / Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109 – 116
2. Materials and methods
2.1. Materials 2.1.1. Preparation of N-acetyl-menthyl anthranilate A mixture of menthyl anthranilate (5.75 g, 20 mM, Aldrich) and acetic anhydride (4.08 g, 40 mM, Aldrich) was stirred and heated to 608C for 2 h. The mixture was cooled to room temperature, aqueous sodium hydrogencarbonate was added and the product extracted using dichloromethane. The extract was dried and evaporated to yield the crude product which was chromatographed on silica using dichloromethane as the eluent to yield the pure product as a white solid, m.p. 67–698C. Yield 3.90 g, 56%. Particular care was taken to completely remove traces of the highly fluorescent menthyl anthranilate from the product that could be readily detected by thin-layer chromatography (TLC). Nuclear magnetic resonance (NMR), CDCl 3, d (ppm) 0.8–2.3 (m, 18H), 2.23 (s, 3H), 4.82 (m, 1H), 7.01–8.78 (m, 4H), 11.19 (m, 1H). Elemental analysis: found C, 71.76%, H 8.55%, N 4.41%, C 19 H 27 NO 3 requires C, 71.92%, H 8.52%, N, 4.42%. Analar-grade ethanol was obtained from BDH and used without further purification, as was HPLC-grade acetonitrile obtained from Aldrich and cyclohexane obtained from Fischer. 2.2. Instrumentation UV–visible absorption spectra were recorded on an ATI Unicam UV-2 spectrophotometer. Luminescence measurements were recorded using a Perkin-Elmer LS50B luminescence spectrometer, and fluorescence quantum yields were determined using an ISA Fluorolog-3-11 spectrophotometer. Emission spectra were corrected for the spectral response of the spectrometers using correction curves provided by the manufacturers. Low-temperature spectra were recorded using samples held in an Oxford Instruments DN1704 cryostat, and the sample temperature monitored using an ITC-6 controller. Kinetic absorption and singlet oxygen measurements were recorded using the spectrometers described elsewhere; the samples were excited at 266 nm [5]. Fluorescence lifetime measurements were recorded using the method of time-correlated single photon counting. The samples were excited using the 337 nm output of a nanosecond nitrogen flashlamp (IBH) running at 20–40 kHz. The fluorescence was collected at 908 to the excitation source and the emission wavelength was selected using a monochromator (IBH) and detected using a photomultiplier (Hamamastsu R928). The detected fluorescence signal was processed by a constant fraction discriminator (Ortec 584) and was used as the ‘stop’ pulse for the time to amplitude converter (Ortec 457). The start
111
pulse was derived from the flashlamp. The output of the TAC was analysed by a Ortec Trump-2K pulse height analyser and the acquired histogram analysed by iterative reconvolution of the observed instrument response function with an exponential decay. In all cases excellent fits were obtained for a single exponential decay as judged by reduced chi-squared, random residuals and an autocorrelation function.
2.3. Methods Fluorescence quantum yield, Ff , values were determined relative to 2-aminopyridine in 0.05 M H 2 SO 4 (Ff 5 0.6060.05) and quinine bisulfate in 0.5 M H 2 SO 4 (Ff 5 0.55) [15,16], using the method described previously [5]. Low-temperature phosphorescence spectra of this compound were recorded in organic glasses at 77 K. The glasses used were EPA (diethyl ether–iso-pentane–ethanol, 5:5:2) [17]. The phosphorescence lifetime was recorded in an EPA glass following 310 nm excitation, with the emission monitored at 420 nm. Transient lifetimes and absorption spectra were obtained for degassed solutions held in 1 cm pathlength quartz degassing cells. Degassing was achieved using the freeze– pump–thaw technique. Difficulties were experienced in studying the triplet state due to interference in the probe from delayed fluorescence. To overcome this measurements were also made using more dilute solutions held in long pathlength quartz degassing cells (l54 cm) in an attempt to reduce the effects of triplet–triplet annihilation. Under these conditions the pump and probe beams were made co-linear. Singlet oxygen quantum yield, FD , values were determined relative to meso-tetraphenylporphyrin, TPP, in cyclohexane (FD 50.58), acridine in acetonitrile (FD 50.82) and perinaphthenone in ethanol (FD 50.97) [18], using the method recently described by Nonell and Braslavsky [19]. In the singlet oxygen quenching studies, singlet oxygen was generated by irradiation of solutions containing NAMA and the photosensitizer TPP at 532 nm. Use of this wavelength ensured that only the TPP was excited and quenched by oxygen. The lifetime of the singlet oxygen produced was monitored as a function of the NAMA concentration, allowing the combined physical and chemical quenching rate constant to be determined. The rate constants for the self-quenching of the triplet state of this compound were determined using the Stern– Volmer relationship. The experimental set-up and data analysis have been described previously [5]. Oxygen quenching rate constants were determined using a similar treatment to that used above. The samples were first degassed and then exposed to a controlled air pressure using an MKS Baratron pressure gauge. Corrections were then made for the contribution to the total pressure by the solvent [20].
A. Beeby, A.E. Jones / Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109 – 116
112
a higher degree than the shorter wavelength bands. The data are tabulated in Table 1.
3.2. Fluorescence spectra
Fig. 3. (a) The UV–visible absorption spectrum of NAMA in ethanol at a concentration of 1.2310 24 mol dm 23 in a 1 cm pathlength quartz cuvette. (b) The corrected fluorescence emission spectrum of NAMA in ethanol at a concentration of 3310 25 mol dm 23 , lex 5310 nm. (c) The delayed fluorescence spectrum of NAMA in ethanol following 266 nm laser excitation of a degassed solution. The absorbance of the solution was 0.5 in a 1 cm pathlength cuvette and the laser fluence was 1 mJ cm 22 .
The fluorescence properties of this compound were studied and quantified in a range of solvents. The emission spectrum in ethanol is shown in Fig. 3b and shows an emission maximum at 370 nm. The position of the fluorescence emission maximum is slightly solvent dependent showing a red shift of 7 nm on changing from non-polar to polar solvents. This red-shift with increasing solvent polarity is indicative of an increased polarity of the ester in its excited state relative to that in ground state. The quantum yields and lifetimes are given in Table 2. It can be seen that the fluorescence is short lived in the range 3–4 ns and that the quantum yields lie in the range 0.13–0.16 in all solvent systems studied.
3.3. Phosphorescence spectra 3. Results
3.1. Absorption spectra The UV absorption spectrum of NAMA in ethanol is shown in Fig. 3a. It exhibits three peaks at 224, 252 and 310 nm and has a molar absorption coefficient of 5400 dm 3 mol 21 cm 21 at 310 nm. The absorption maxima and absorption coefficients only vary slightly with changing solvent. However, the peak at 310 nm is shown to vary to
The low-temperature phosphorescence spectrum of this compound in EPA shows a structureless emission centred at 420 nm, and the spectrum allows the energy of the triplet state to be estimated to be 31163 kJ mol 21 . Because the phosphorescence spectrum is a single, broad, featureless band the wavelength of the (0,0) transition was taken to be that at which the intensity of the phosphorescence at the blue edge of the emission was 10% of the maximum. The phosphorescence from the low temperature
Table 1 UV spectroscopic data for NAMA Solvent
lmax 1 (nm)
lmax 2 (nm)
lmax 3 (nm)
e at lmax 3 (dm 3 mol 21 cm 21 )
Toluene Cyclohexane Ethanol Acetonitrile Methanol Dichloromethane Chloroform EPA MCH:i-P
– 224 224 223 222 – – 226 226
– 253 252 252 252 253 253 254 254
313 315 310 310 307 310 312 315 317
70006100 68006100 54006100 57006100 54006100 63006100 59006100 – –
Table 2 Photophysical data for NAMA Photophysical parameter
Acetonitrile Toluene Ethanol Cyclohexane
l *max (nm)
t †f (ns)
F ‡f
t §T (ms)
F uuD
k ¶D (10 4 dm 3 mol 21 s 21 )
k ** O2 (10 8 dm 3 mol 21 s 21 )
k †† self (10 6 dm 3 mol 21 s 21 )
369 368 370 363
3.460.1 3.960.1 3.660.1 4.060.1
0.1460.01 0.1660.01 0.1360.01 0.1360.01
240625 – 180615 120610
0.2160.02 – 0.2060.02 0.1960.02
– – – 2.460.3
4.660.2 – 5.060.1 4.060.1
5.260.8 – 2.060.1 32.067.0
*Fluorescence emission maximum, following 330 nm excitation. † Fluorescence lifetime. ‡ Fluorescence quantum yield. § Triplet state lifetime. uu Singlet oxygen generation quantum yield. ¶ Singlet oxygen quenching rate constant. ** Oxygen quenching rate constant. †† Self quenching rate constant.
A. Beeby, A.E. Jones / Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109 – 116
113
glasses was found to be long lived, with a lifetime of 1.360.1 s.
3.4. Kinetic absorption measurements Kinetic absorption measurements in dilute degassed solutions revealed a short-lived transient absorption centred around 480 nm with a solvent dependant lifetime in the range 120–240 ms. The transient absorption spectrum in degassed ethanol is shown in Fig. 4a and the corresponding kinetic absorption trace and first-order fit is shown in Fig. 4b. Aeration of the solutions resulted in the loss of the transient signal. Due to interference from the prompt and delayed fluorescence and limitations in our probe wavelengths, it was not possible to record a bleaching of the ground state, nor was an attempt made to determine eTT or FT . However based on the values of Ff and FD measured here we estimate that 0.2,FT ,0.6. A summary of the data is shown in Table 2. A long-lived emission was observed during the course of these kinetic absorption experiments. The intensity of the emission was found to increase exponentially with laser power and the spectral profile was recorded and found to be identical to the fluorescence spectrum. Aera-
Fig. 5. Kinetic singlet oxygen phosphorescence decay trace following 266 nm laser excitation and the first-order fit obtained.
tion of the solution resulted in the loss of the emission, indicative of P-type delayed fluorescence resulting from a triplet–triplet annihilation process. The quenching of the transient species by ground state molecules was studied in a range of solvent systems. The rate constants for self-quenching were found to lie in the range 0.2–3.2310 7 dm 3 mol 21 s 21 as shown in Table 2. The total loss of the transient signal on aeration of the solution demonstrates that the species formed is efficiently quenched by oxygen, indicating that the species is a triplet state. The oxygen quenching rate constants were found to be in the range 4.0–5.0310 8 dm 3 mol 21 s 21 , as shown in Table 2.
3.5. Singlet oxygen measurements The efficient oxygen-quenching of the transient species formed results in the formation of singlet oxygen, detected by its characteristic luminescence at 1270 nm. A typical singlet oxygen phosphorescence decay trace and the firstorder fit are shown in Fig. 5. The lifetimes of singlet oxygen were determined to be 22 ms in cyclohexane, 68 ms in acetonitrile and 17 ms in ethanol, all of which are consistent with literature values for these solvents [21]. The quantum yields of singlet oxygen formation are shown in Table 2. The quenching of the characteristic singlet oxygen phosphorescence by NAMA was studied in cyclohexane and the rate constant determined to be 2.4360.03310 4 dm 3 mol 21 s 21 .
4. Discussion Fig. 4. (a) Transient absorption spectrum obtained following 266 nm excitation of NAMA in degassed ethanol at 295 K. The absorbance of the solution was 0.5 at the laser wavelength in a 4 cm pathlength cell and the laser energy was 1 mJ per pulse. (b) The kinetic absorption profile at a 450 nm probe wavelength.
NAMA has been synthesised and characterised. Based upon our earlier findings that the starting amine is both highly fluorescent and a sensitiser of singlet oxygen we
114
A. Beeby, A.E. Jones / Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109 – 116
ensured, by exhaustive chromatography, that the materials used in these studies contained no detectable menthyl anthranilate (,0.1% estimated by TLC). Anthranilate esters have been described as being among the few chemical sunscreens that absorb in the UV-A region [3], whilst providing minimal protection in the UV-B region [22]. However, the absorption spectrum of NAMA shows that this compound offers protection in both the UV-B and UV-A ranges. The absorption maximum at 310 nm is situated in the UV-B region but extends from 270 to 340 nm, and therefore provides only limited coverage of the UV-A range. Another disadvantage of this compound is the low molar absorption coefficient, arising from the ortho-disubstitution pattern of the compound. As with menthyl anthranilate the absorption coefficient is higher than in other ortho-disubstituted sunscreens such as the salicylates, but considerably lower than para-disubstituted counterparts such as the para-aminobenzoic acid derivatives. This work, however, highlights other more disturbing photophysical properties of this material. The absorption spectra of this compound in ethanol show three bands at 224, 252 and 310 nm. When these bands are compared to those in the analogous spectra of menthyl anthranilate several differences can be noted due to the acetylation of the amine group. Firstly, there is a red shift in the shortest wavelength band from 220 to 224 nm. Secondly, there is a slight red shift in the 249 nm band to 252 nm and an increase in relative intensity when compared to the shorter wavelength bands. Finally, the longest wavelength band is blue shifted from 340 to 310 nm and shows a reduction in relative intensity. These findings are consistent with those of Ungnade [8] and Grammaticakis [10] who have explained them in terms of an increase in conjugation and the loss of molecular planarity following acetylation. The latter results in the reduction in the tendency of the lone pair of electrons on the nitrogen to delocalise into the aromatic ring. Luminescence studies have determined that this compound is both fluorescent and phosphorescent. The fluorescence was determined to be short lived (tf |4 ns) with a solvent dependent emission maximum in the range 362– 370 nm. This value is in agreement with the findings of Staudemayer and Roberts [11]. Comparison of the fluorescence spectra and quantum yields with those obtained for menthyl anthranilate shows a blue shift of the emission maxima and a reduction in the quantum yields. The fluorescence quantum yields observed in these solution state studies and the small Stokes shift observed for the fluorescence spectrum means that a high proportion of the emitted radiation falls in the UV-A range, i.e., 320–400 nm. In addition to the radiative decay the excited singlet state undergoes intersystem crossing to yield the triplet state. This is very long lived and emissive in low temperature glasses and we have established its triplet energy,
ET |310 kJ mol 21 . From this we can postulate that it may act as a sensitiser for thymine dimerisation. The lowest triplet energy of thymine has been estimated from phosphorescence measurements to be 315 kJ mol 21 [23]. It has also been established that triplet–triplet energy transfer will occur quite efficiently from donors whose triplet energies are either above or 8–12 kJ mol 21 below that of the acceptor [24]. This implies that there is at least the possibility that this compound could sensitise the formation of thymine triplets resulting in the generation of thymine dimers, which are known to be potential precursors to skin cancer [25]. The triplet state has also been found to be a sensitiser of singlet oxygen, which is itself a harmful species in biological systems. We have assigned the transient species formed following excitation of NAMA at 266 nm to the triplet state by virtue of its quenching by oxygen. The absorption spectra of the transient species in all solvent systems were similar, with an absorption maximum at 450 nm, suggesting that the same species is formed in each case. The decrease in triplet state lifetime observed on changing from ethanol through acetonitrile to cyclohexane indicates that the species formed is more stable in polar environments. In addition to the radiative decay and oxygen quenching we have shown that the triplet state can also be deactivated by P-type delayed fluorescence. This was observed from degassed solutions of this compound at room temperature following 266 nm pulsed excitation and assigned on the basis of the quadratic behaviour of intensity with laser power. Finally, the triplet state is also quenched by the ground state, although this is only significant in degassed solutions of higher concentrations. By comparison of the results reported here with work involving sunscreen formulations containing menthyl anthranilate [26–29], it is likely that the use of NAMA within a sunscreen formulation where it is the major absorbing species is likely to result in a fluorescent product. Most of the fluorescent emission will be in the UV-A range and, as discussed previously [5], a significant amount will be transmitted to the skin. This indicates that incorporation of this ingredient into a formulation could result in a product that may, in fact, increase the intensity of radiation at the skin surface in this range. More significantly, however, the results of laser flash photolysis and near infrared luminescence measurements clearly demonstrate the formation of a long-lived triplet state which does sensitise singlet oxygen with a significant quantum yield in dilute solution and, in vivo could sensitise thymine dimerisation. In many respects the compound behaves in a similar fashion to the parent, menthyl anthranilate, the major difference being that acetylation of the amine group gives rise to a blue shift in the absorption and emission spectra, a reduction in the fluorescence quantum yield and an increase in the quantum yield of singlet oxygen formation.
A. Beeby, A.E. Jones / Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109 – 116
5. Conclusion A thorough study of the photophysical properties of N-acetyl-menthyl anthranilate has been conducted using conventional room temperature and low temperature luminescence spectroscopy, laser flash photolysis, and near infrared luminescence measurements. This compound was synthesised as an analogue of N-acetyl-homomenthyl anthranilate, a compound that, until relatively recently, was approved for use as a sunscreen in the EU. The luminescence studies indicate that, following excitation, radiative decay from the S 1 state resulting in fluorescence occurs to a significant extent (Ff |0.1560.02), but that intersystem crossing to the triplet state is the dominant deactivation pathway (e.g., FD 50.2060.02 in ethanol). The triplet state has been shown to have a long lifetime in degassed solution (120–240 ms) and a long lived phosphorescence lifetime of |1.3 s when held in low temperature organic glasses. From the phosphorescence emission spectrum the triplet state energy, ET , has been estimated to be |310 kJ mol 21 . This relatively high value indicates that this compound has the potential to sensitise the formation of thymine triplets in skin resulting in the generation of thymine dimers. Kinetic absorption measurements on degassed solutions have revealed a triplet–triplet absorption band at |420 nm that is independent of solvent and that is readily quenched by oxygen on aeration. Also, under high intensity pulsed illumination delayed fluorescence has been observed. The intensity of this emission increased exponentially with the laser power and was quenched by oxygen indicating P-type delayed fluorescence resulting from triplet–triplet annihilation. Finally, in aerated solutions this N-acetyl anthranilate ester has been shown to produce singlet oxygen, with a quantum yield of 0.2060.02 in all solvents studied.
6. Abbreviations NAMA UV EPA
N-Acetyl menthyl anthranilate Ultraviolet Ether–pentane–alcohol (5:5:2)
Acknowledgements The authors acknowledge the Royal Society for funding equipment used in this study and the University of Durham for a studentship (A.E.J.).
References [1] F.P. Gasparro, M. Mitchnick, J.F. Nash, A review of sunscreen safety and efficacy, Photochem. Photobiol. 68 (1998) 243–256.
115
[2] L.H. Kligman, F.J. Akin, A.M. Kligman, The contributions of UVA and UVB to connective tissue damage in hairless mice, J. Invest. Dermatol. 84 (1985) 272–276. [3] N.A. Shaath, Evolution of modern sunscreen chemicals, in: N.J. Lowe, N.A. Shaath, M.A. Pathak (Eds.), Sunscreens – Development, Evaluation, and Regulatory Aspects, Marcel Dekker, New York, 1997, pp. 3–33. [4] L.E. Agrapidis-Paloymis, R.A. Nash, N.A. Shaath, The effect of solvents on the ultraviolet absorbance of sunscreens, J. Soc. Cosmet. Chem. 38 (1987) 209–221. [5] A. Beeby, A.E. Jones, The photophysical properties of menthyl anthranilate: a UV-A sunscreen, Photochem. Photobiol. 72 (2000) 10–15. [6] M. Brown, Boots plc, personal communication, 2001. [7] H. Dannenberg, Classification of ultraviolet absorption. I. Substituted acetophenines, benzoic acids and cinnamic acid, Z. Naturforsch. 4b (1949) 327–344. [8] H.E. Ungnade, Ultraviolet absorption spectra of acetanilides, J. Am. Chem. Soc. 76 (1954) 5133–5135. [9] P. Grammaticakis, The ultraviolet absorption of o-substituted anilines III. N-Substituted o-nitro- and o-carboxyanilines, Bull. Soc. Chim. France (1950) 158–166. [10] P. Grammaticakis, Near ultraviolet and visible absorption of some isomeric arylamines and their nitrogen substituted derivatives I, Bull. Soc. Chim. France (1951) 220–226. [11] R. Staudemayer, T.D. Roberts, Neighbouring group participation in photolysis of o-substituted aryls, Tetrahedron Lett. 13 (1974) 1141– 1144. [12] V.I. Stenberg, in: O.L. Chapman (Ed.), Organic Photochemistry, Vol. I, Marcel Dekker, New York, 1967. [13] A.E. Jones, A Spectroscopic Study of Sunscreen, Ph.D. Thesis, University of Durham, 2000. [14] R.V. Bensasson, E.J. Land, T.G. Truscott, Excited States and Free Radicals in Biology and Medicine, Oxford University Press, New York, 1993. [15] D.F. Eaton, Luminescence spectroscopy, in: J.C. Scaiano (Ed.), Handbook of Organic Photochemistry, Vol. 1, CRC Press, Boca Raton, FL, 1989, pp. 231–239. [16] C.A. Parker, in: Photoluminescence of Solutions, Elsevier, Amsterdam, 1968, pp. 266–267. [17] S.L. Murov, I. Carmichael, G.L. Hug, in: Handbook of Photochemistry, 2nd Edition, Marcel Dekker, New York, 1995, pp. 283– 297. [18] F. Wilkinson, W.P. Helman, A.B. Ross, Quantum yields for the photosensitized formation of the lowest excited singlet state of molecular oxygen in solution, J. Phys. Chem. Ref. Data 22 (1) (1993) 113–262. [19] S. Nonell, S.L. Braslavsky, Time resolved singlet oxygen detection, in: L. Packer, H. Sies (Eds.), Methods in Enzymology, Singlet Oxygen, UV-A and Ozone, Vol. 319, Academic Press, San Diego, CA, 2000, pp. 37–49. [20] R.C. Weist (Ed.), Handbook of Chemistry and Physics, 61st Edition, CRC Press, Boca Raton, FL, 1980. [21] A.A. Gorman, M.A.J. Rodgers, Singlet oxygen, in: J.C. Scaiano (Ed.), Handbook of Organic Photochemistry, Vol. II, CRC Press, Boca Raton, FL, 1989, pp. 229–247. [22] A.A. Fisher, Sunscreen dermatitis: part IV – the salicylates, the anthranilates, and physical agents, Cutis 50 (1992) 397–398. [23] A.A. Lamola, Applications of electronic energy transfer in solution, Photochem. Photobiol. 8 (1968) 601–616. [24] D.F. Evans, Perturbation of singlet–triplet transitions of aromatic molecules by oxygen under pressure, J. Chem. Soc. (1957) 1351– 1537. [25] H. Gonzenbach, T.J. Hill, T.G. Truscott, The triplet energy levels of UVA and UVB sunscreens, J. Photochem. Photobiol. B: Biol. 16 (1992) 377–379.
116
A. Beeby, A.E. Jones / Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109 – 116
[26] L.E. Rhodes, B.L. Diffey, Quantitative assessment of sunscreen application technique by in vivo fluorescence spectroscopy, J. Soc. Cosmet. Chem. 47 (1996) 109–115. [27] L.E. Rhodes, B.L. Diffey, Fluorescence spectroscopy: a rapid, noninvasive method for measurement of skin surface thickness of topical agents, Br. J. Dermatol. 136 (1997) 12–17. [28] R.M. Azurdia, J.A. Pagliaro, B.L. Diffey, L.E. Rhodes, Sunscreen
application by photosensitive patients is inadequate for protection, Br. J. Dermatol. 140 (1999) 255–258. [29] R.P. Stokes, B.L. Diffey, The feasibility of using fluorescence spectroscopy as a rapid, non-invasive method for evaluating sunscreen performance, J. Photochem. Photobiol. B: Biol. 50 (1999) 137–143.
Photophysical properties of N-acetyl-menthyl anthranilate a, b Andrew Beeby *, Allison E. Jones a b
Department of Chemistry, University of Durham, South Road, Durham DH1 3 LE, UK Centre for Forensic Science, University of Central Lancashire, Preston PR1 2 HE, UK
Abstract The results of a comprehensive investigation of the photophysical properties of the sunscreen analogue, N-acetyl menthyl anthranilate (NAMA), in various solvent systems are reported. Luminescence studies reveal that this compound is fluorescent (Ff 50.1660.01) in toluene and has a solvent dependent emission maximum in the range 363–370 nm. Phosphorescence has also been detected in low temperature glasses with an emission maximum at 420 nm in EPA, and a lifetime of 1.3 s; the triplet energy was found to be 31163 kJ mol 21 . Kinetic UV–visible absorption measurements revealed a transient species with absorption maxima at 450 nm and solvent dependent lifetimes of 120–240 ms which are attributed to the triplet state. The triplet state is efficiently quenched by oxygen, leading to the formation of singlet oxygen in all of the solvent systems studied. The singlet oxygen quantum yields (FD ), determined by time-resolved near-infrared luminescence measurements, were in the range 0.19–0.21. 2001 Elsevier Science B.V. All rights reserved. Keywords: Sunscreen; Singlet oxygen; Sensitiser; UV radiation
1. Introduction The past decade has seen an increase in the use of sunscreen formulations resulting from an increased awareness of the damaging effects of solar UV radiation on human skin [1]. The recent association of UVA exposure with skin-ageing and other dermatological conditions has furthered this increase [2]. Shaath [3] describes the anthranilates as an ‘interesting’ class of UV filters and describes in detail the two derivatives approved for commercial use, menthyl anthranilate in the USA and N-acetyl-homomenthyl anthranilate in the European Union (EU), see Fig. 1. He comments on the ‘ortho-effect’ that is evident in these compounds, namely the red-shift of the absorption maxima and low molar absorption coefficients when compared to their para-disubstituted counterparts. This shift is attributed to the ease of electron delocalisation in the ortho-compounds as opposed to the para-compounds. However, steric crowding due to the ortho-disubstitution in these compounds causes deviations from coplanarity of the molecules resulting in a reduction of the intensity of the absorption and lower molar absorption coefficients. Agrapidis-Paloymis et al. [4] have commented that the *Corresponding author. Tel.: 144-191-3744-623; fax: 144-191-3844737. E-mail address: [email protected] (A. Beeby).
anthranilates, like the salicylates, are stable and safe compounds to use, owing to this ortho-disubstituted relationship, and do not exhibit significant solvent effects in cosmetic formulations. The work presented here is a continuation of our study of the anthranilate esters used as sunscreens. We have previously reported work on menthyl anthranilate [5], a compound that is currently used in formulations sold in the USA. We have found it to be highly fluorescent, both in solution (Ff |0.6) and when part of commercially available sunscreen formulations. It was also found that a significant proportion of the excited state molecules undergo intersystem crossing to the triplet state, which was long lived in solution (20–200 ms) and had a long phosphorescence lifetime (tp 52.5 s). The energy of the triplet state was estimated from the phosphorescence spectrum to be |287
Fig. 1. The molecular structure of N-acetyl-homomenthyl anthranilate.
1011-1344 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S1011-1344( 01 )00197-X
110
A. Beeby, A.E. Jones / Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109 – 116
kJ mol 21 , high enough to efficiently sensitise the formation of singlet oxygen (FD 50.10). Despite the apparent advantages of these compounds the N-acetyl-homomenthyl anthranilate derivative was removed from the list of approved sunscreens in the EU in 1989 [6], although the reasons for its withdrawal have not been published. Furthermore there is surprisingly little information on this compound in the scientific and medical literature given the interest in using N-acetyl-homomenthyl anthranilate in commercial sunscreens. It might be expected that its photophysical and photochemical properties would have been investigated at the time of its admission onto the list of approved sunscreens but there has been little published on its photophysics and photochemistry or that of the parent acid and related esters. Several UV–visible spectroscopy studies have been carried out on N-acetyl anthranilic acids. Dannenberg [7] determined the absorption spectra of benzoic acid and a range of its derivatives, including the acetamido derivative, to compare and quantify the shifts in absorption bands caused by the different ring substitutions. Ungnade [8] has studied the ultraviolet absorption spectra of a range of compounds and interpreted the intensities and lmax values in terms of the electronic and steric effects of the substituents. The spectrum of N-acetyl anthranilic acid has been reported to have bands at 221, 252 and 305 nm, and the molar absorption coefficient at 305 nm has been determined to be 4650 dm 3 mol 21 cm 21 . The red shift and increased intensity of the 252 nm band, and the blue shift and reduction in the intensity of the 305 nm band relative to the spectrum of anthranilic acid have been attributed to the increase in conjugation of the system following acylation of the amine group and reduced conjugation between the nitrogen atom and the aromatic ring arising from the loss of planarity of the molecule, respectively. Grammaticakis [9] recorded the UV spectra of a range of N-substituted o-nitro and o-carboxy anilines including Nacetyl anthranilic acid. The peak shifts and intensity changes of the absorption bands following substitution of the alkyl in the ortho position by –NO 2 , –COOH and –COCH 3 were studied. It was found that substitution had little or no effect on the absorption bands. Further work by Grammaticakis [10] studied the near-ultraviolet and visible absorption of some isomeric arylamines and their nitrogen substituted derivatives. In compounds of the form XC 6 H 5 NHCOCH 3 , acetylation of the amine group was found to cause a red shift in the middle band (|250 nm) and a blue shift in the band at longer wavelength (|300 nm), in agreement with the work by Ungnade. The photolysis of N-acetyl anthranilic acid has been carried out by Staudemayer and Roberts [11], in order to study the neighbouring group participation in the cyclization rearrangements known to occur in acetanilides under UV light. They irradiated solutions of N-acetyl anthranilic acid at 254 nm for 100 h and monitored the changes using
UV spectroscopy. The resulting solution was found to be a mixture containing the starting material, anthranilic acid, 3-aceto-2-aminobenzoic acid, and 5-aceto-2-aminobenzoic acid, all products of photo-Fries-like cleavage and rearrangements. Since photo-Fries and similar reactions have been postulated to occur from singlet excited states [12] further experiments were conducted to probe the reactions of the triplet state of the acid species. In benzene solution they observed sensitisation of the triplet state of the acid by energy transfer from the triplet state of the solvent. The photo-product formed in under these conditions was identified as benzoxazinone. From these studies they concluded that N-acetyl anthranilic acid forms photo-Frieslike cleavage and rearrangement products via the excited singlet state, and that it cyclizes to benzoxazinone via the triplet state. Several workers have reported that N-acetyl anthranilic acid is luminescent, although there is surprisingly little quantitative data in the literature. Staudemayer and Roberts determined the luminescence spectra of N-acetyl anthranilic acid as part of the work described above in order to determine a suitable triplet state sensitiser. They found that in acetonitrile fluorescence emission occurred with a maximum at 367 nm and phosphorescence emission with a maximum at 428 nm. In this paper we present the results of a solution state study of N-acetyl-menthyl anthranilate (Fig. 2) in a range of solvent systems. The menthyl ester was chosen for the study due to its ease of synthesis. Based upon our previous work we have demonstrated that the properties of the esters are essentially independent of the alkyl group: the methyl and menthyl esters of anthranilic acid are identical [13]. The work described here includes measurement of fluorescence and phosphorescence emission spectra and lifetimes, quantum yields of fluorescence, triplet–triplet absorption spectra and the determination of oxygen- and self-quenching rate constants. Significantly we have shown that this N-acetyl anthranilate ester has an energetic, long lived triplet state which can sensitise the formation of singlet oxygen, a species known to be harmful to biological systems [14]. The quantum yields of singlet oxygen formation have been determined to be in the range 0.19– 0.21 in the solvent systems studied.
Fig. 2. The molecular structure of N-acetyl-menthyl anthranilate.
A. Beeby, A.E. Jones / Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109 – 116
2. Materials and methods
2.1. Materials 2.1.1. Preparation of N-acetyl-menthyl anthranilate A mixture of menthyl anthranilate (5.75 g, 20 mM, Aldrich) and acetic anhydride (4.08 g, 40 mM, Aldrich) was stirred and heated to 608C for 2 h. The mixture was cooled to room temperature, aqueous sodium hydrogencarbonate was added and the product extracted using dichloromethane. The extract was dried and evaporated to yield the crude product which was chromatographed on silica using dichloromethane as the eluent to yield the pure product as a white solid, m.p. 67–698C. Yield 3.90 g, 56%. Particular care was taken to completely remove traces of the highly fluorescent menthyl anthranilate from the product that could be readily detected by thin-layer chromatography (TLC). Nuclear magnetic resonance (NMR), CDCl 3, d (ppm) 0.8–2.3 (m, 18H), 2.23 (s, 3H), 4.82 (m, 1H), 7.01–8.78 (m, 4H), 11.19 (m, 1H). Elemental analysis: found C, 71.76%, H 8.55%, N 4.41%, C 19 H 27 NO 3 requires C, 71.92%, H 8.52%, N, 4.42%. Analar-grade ethanol was obtained from BDH and used without further purification, as was HPLC-grade acetonitrile obtained from Aldrich and cyclohexane obtained from Fischer. 2.2. Instrumentation UV–visible absorption spectra were recorded on an ATI Unicam UV-2 spectrophotometer. Luminescence measurements were recorded using a Perkin-Elmer LS50B luminescence spectrometer, and fluorescence quantum yields were determined using an ISA Fluorolog-3-11 spectrophotometer. Emission spectra were corrected for the spectral response of the spectrometers using correction curves provided by the manufacturers. Low-temperature spectra were recorded using samples held in an Oxford Instruments DN1704 cryostat, and the sample temperature monitored using an ITC-6 controller. Kinetic absorption and singlet oxygen measurements were recorded using the spectrometers described elsewhere; the samples were excited at 266 nm [5]. Fluorescence lifetime measurements were recorded using the method of time-correlated single photon counting. The samples were excited using the 337 nm output of a nanosecond nitrogen flashlamp (IBH) running at 20–40 kHz. The fluorescence was collected at 908 to the excitation source and the emission wavelength was selected using a monochromator (IBH) and detected using a photomultiplier (Hamamastsu R928). The detected fluorescence signal was processed by a constant fraction discriminator (Ortec 584) and was used as the ‘stop’ pulse for the time to amplitude converter (Ortec 457). The start
111
pulse was derived from the flashlamp. The output of the TAC was analysed by a Ortec Trump-2K pulse height analyser and the acquired histogram analysed by iterative reconvolution of the observed instrument response function with an exponential decay. In all cases excellent fits were obtained for a single exponential decay as judged by reduced chi-squared, random residuals and an autocorrelation function.
2.3. Methods Fluorescence quantum yield, Ff , values were determined relative to 2-aminopyridine in 0.05 M H 2 SO 4 (Ff 5 0.6060.05) and quinine bisulfate in 0.5 M H 2 SO 4 (Ff 5 0.55) [15,16], using the method described previously [5]. Low-temperature phosphorescence spectra of this compound were recorded in organic glasses at 77 K. The glasses used were EPA (diethyl ether–iso-pentane–ethanol, 5:5:2) [17]. The phosphorescence lifetime was recorded in an EPA glass following 310 nm excitation, with the emission monitored at 420 nm. Transient lifetimes and absorption spectra were obtained for degassed solutions held in 1 cm pathlength quartz degassing cells. Degassing was achieved using the freeze– pump–thaw technique. Difficulties were experienced in studying the triplet state due to interference in the probe from delayed fluorescence. To overcome this measurements were also made using more dilute solutions held in long pathlength quartz degassing cells (l54 cm) in an attempt to reduce the effects of triplet–triplet annihilation. Under these conditions the pump and probe beams were made co-linear. Singlet oxygen quantum yield, FD , values were determined relative to meso-tetraphenylporphyrin, TPP, in cyclohexane (FD 50.58), acridine in acetonitrile (FD 50.82) and perinaphthenone in ethanol (FD 50.97) [18], using the method recently described by Nonell and Braslavsky [19]. In the singlet oxygen quenching studies, singlet oxygen was generated by irradiation of solutions containing NAMA and the photosensitizer TPP at 532 nm. Use of this wavelength ensured that only the TPP was excited and quenched by oxygen. The lifetime of the singlet oxygen produced was monitored as a function of the NAMA concentration, allowing the combined physical and chemical quenching rate constant to be determined. The rate constants for the self-quenching of the triplet state of this compound were determined using the Stern– Volmer relationship. The experimental set-up and data analysis have been described previously [5]. Oxygen quenching rate constants were determined using a similar treatment to that used above. The samples were first degassed and then exposed to a controlled air pressure using an MKS Baratron pressure gauge. Corrections were then made for the contribution to the total pressure by the solvent [20].
A. Beeby, A.E. Jones / Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109 – 116
112
a higher degree than the shorter wavelength bands. The data are tabulated in Table 1.
3.2. Fluorescence spectra
Fig. 3. (a) The UV–visible absorption spectrum of NAMA in ethanol at a concentration of 1.2310 24 mol dm 23 in a 1 cm pathlength quartz cuvette. (b) The corrected fluorescence emission spectrum of NAMA in ethanol at a concentration of 3310 25 mol dm 23 , lex 5310 nm. (c) The delayed fluorescence spectrum of NAMA in ethanol following 266 nm laser excitation of a degassed solution. The absorbance of the solution was 0.5 in a 1 cm pathlength cuvette and the laser fluence was 1 mJ cm 22 .
The fluorescence properties of this compound were studied and quantified in a range of solvents. The emission spectrum in ethanol is shown in Fig. 3b and shows an emission maximum at 370 nm. The position of the fluorescence emission maximum is slightly solvent dependent showing a red shift of 7 nm on changing from non-polar to polar solvents. This red-shift with increasing solvent polarity is indicative of an increased polarity of the ester in its excited state relative to that in ground state. The quantum yields and lifetimes are given in Table 2. It can be seen that the fluorescence is short lived in the range 3–4 ns and that the quantum yields lie in the range 0.13–0.16 in all solvent systems studied.
3.3. Phosphorescence spectra 3. Results
3.1. Absorption spectra The UV absorption spectrum of NAMA in ethanol is shown in Fig. 3a. It exhibits three peaks at 224, 252 and 310 nm and has a molar absorption coefficient of 5400 dm 3 mol 21 cm 21 at 310 nm. The absorption maxima and absorption coefficients only vary slightly with changing solvent. However, the peak at 310 nm is shown to vary to
The low-temperature phosphorescence spectrum of this compound in EPA shows a structureless emission centred at 420 nm, and the spectrum allows the energy of the triplet state to be estimated to be 31163 kJ mol 21 . Because the phosphorescence spectrum is a single, broad, featureless band the wavelength of the (0,0) transition was taken to be that at which the intensity of the phosphorescence at the blue edge of the emission was 10% of the maximum. The phosphorescence from the low temperature
Table 1 UV spectroscopic data for NAMA Solvent
lmax 1 (nm)
lmax 2 (nm)
lmax 3 (nm)
e at lmax 3 (dm 3 mol 21 cm 21 )
Toluene Cyclohexane Ethanol Acetonitrile Methanol Dichloromethane Chloroform EPA MCH:i-P
– 224 224 223 222 – – 226 226
– 253 252 252 252 253 253 254 254
313 315 310 310 307 310 312 315 317
70006100 68006100 54006100 57006100 54006100 63006100 59006100 – –
Table 2 Photophysical data for NAMA Photophysical parameter
Acetonitrile Toluene Ethanol Cyclohexane
l *max (nm)
t †f (ns)
F ‡f
t §T (ms)
F uuD
k ¶D (10 4 dm 3 mol 21 s 21 )
k ** O2 (10 8 dm 3 mol 21 s 21 )
k †† self (10 6 dm 3 mol 21 s 21 )
369 368 370 363
3.460.1 3.960.1 3.660.1 4.060.1
0.1460.01 0.1660.01 0.1360.01 0.1360.01
240625 – 180615 120610
0.2160.02 – 0.2060.02 0.1960.02
– – – 2.460.3
4.660.2 – 5.060.1 4.060.1
5.260.8 – 2.060.1 32.067.0
*Fluorescence emission maximum, following 330 nm excitation. † Fluorescence lifetime. ‡ Fluorescence quantum yield. § Triplet state lifetime. uu Singlet oxygen generation quantum yield. ¶ Singlet oxygen quenching rate constant. ** Oxygen quenching rate constant. †† Self quenching rate constant.
A. Beeby, A.E. Jones / Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109 – 116
113
glasses was found to be long lived, with a lifetime of 1.360.1 s.
3.4. Kinetic absorption measurements Kinetic absorption measurements in dilute degassed solutions revealed a short-lived transient absorption centred around 480 nm with a solvent dependant lifetime in the range 120–240 ms. The transient absorption spectrum in degassed ethanol is shown in Fig. 4a and the corresponding kinetic absorption trace and first-order fit is shown in Fig. 4b. Aeration of the solutions resulted in the loss of the transient signal. Due to interference from the prompt and delayed fluorescence and limitations in our probe wavelengths, it was not possible to record a bleaching of the ground state, nor was an attempt made to determine eTT or FT . However based on the values of Ff and FD measured here we estimate that 0.2,FT ,0.6. A summary of the data is shown in Table 2. A long-lived emission was observed during the course of these kinetic absorption experiments. The intensity of the emission was found to increase exponentially with laser power and the spectral profile was recorded and found to be identical to the fluorescence spectrum. Aera-
Fig. 5. Kinetic singlet oxygen phosphorescence decay trace following 266 nm laser excitation and the first-order fit obtained.
tion of the solution resulted in the loss of the emission, indicative of P-type delayed fluorescence resulting from a triplet–triplet annihilation process. The quenching of the transient species by ground state molecules was studied in a range of solvent systems. The rate constants for self-quenching were found to lie in the range 0.2–3.2310 7 dm 3 mol 21 s 21 as shown in Table 2. The total loss of the transient signal on aeration of the solution demonstrates that the species formed is efficiently quenched by oxygen, indicating that the species is a triplet state. The oxygen quenching rate constants were found to be in the range 4.0–5.0310 8 dm 3 mol 21 s 21 , as shown in Table 2.
3.5. Singlet oxygen measurements The efficient oxygen-quenching of the transient species formed results in the formation of singlet oxygen, detected by its characteristic luminescence at 1270 nm. A typical singlet oxygen phosphorescence decay trace and the firstorder fit are shown in Fig. 5. The lifetimes of singlet oxygen were determined to be 22 ms in cyclohexane, 68 ms in acetonitrile and 17 ms in ethanol, all of which are consistent with literature values for these solvents [21]. The quantum yields of singlet oxygen formation are shown in Table 2. The quenching of the characteristic singlet oxygen phosphorescence by NAMA was studied in cyclohexane and the rate constant determined to be 2.4360.03310 4 dm 3 mol 21 s 21 .
4. Discussion Fig. 4. (a) Transient absorption spectrum obtained following 266 nm excitation of NAMA in degassed ethanol at 295 K. The absorbance of the solution was 0.5 at the laser wavelength in a 4 cm pathlength cell and the laser energy was 1 mJ per pulse. (b) The kinetic absorption profile at a 450 nm probe wavelength.
NAMA has been synthesised and characterised. Based upon our earlier findings that the starting amine is both highly fluorescent and a sensitiser of singlet oxygen we
114
A. Beeby, A.E. Jones / Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109 – 116
ensured, by exhaustive chromatography, that the materials used in these studies contained no detectable menthyl anthranilate (,0.1% estimated by TLC). Anthranilate esters have been described as being among the few chemical sunscreens that absorb in the UV-A region [3], whilst providing minimal protection in the UV-B region [22]. However, the absorption spectrum of NAMA shows that this compound offers protection in both the UV-B and UV-A ranges. The absorption maximum at 310 nm is situated in the UV-B region but extends from 270 to 340 nm, and therefore provides only limited coverage of the UV-A range. Another disadvantage of this compound is the low molar absorption coefficient, arising from the ortho-disubstitution pattern of the compound. As with menthyl anthranilate the absorption coefficient is higher than in other ortho-disubstituted sunscreens such as the salicylates, but considerably lower than para-disubstituted counterparts such as the para-aminobenzoic acid derivatives. This work, however, highlights other more disturbing photophysical properties of this material. The absorption spectra of this compound in ethanol show three bands at 224, 252 and 310 nm. When these bands are compared to those in the analogous spectra of menthyl anthranilate several differences can be noted due to the acetylation of the amine group. Firstly, there is a red shift in the shortest wavelength band from 220 to 224 nm. Secondly, there is a slight red shift in the 249 nm band to 252 nm and an increase in relative intensity when compared to the shorter wavelength bands. Finally, the longest wavelength band is blue shifted from 340 to 310 nm and shows a reduction in relative intensity. These findings are consistent with those of Ungnade [8] and Grammaticakis [10] who have explained them in terms of an increase in conjugation and the loss of molecular planarity following acetylation. The latter results in the reduction in the tendency of the lone pair of electrons on the nitrogen to delocalise into the aromatic ring. Luminescence studies have determined that this compound is both fluorescent and phosphorescent. The fluorescence was determined to be short lived (tf |4 ns) with a solvent dependent emission maximum in the range 362– 370 nm. This value is in agreement with the findings of Staudemayer and Roberts [11]. Comparison of the fluorescence spectra and quantum yields with those obtained for menthyl anthranilate shows a blue shift of the emission maxima and a reduction in the quantum yields. The fluorescence quantum yields observed in these solution state studies and the small Stokes shift observed for the fluorescence spectrum means that a high proportion of the emitted radiation falls in the UV-A range, i.e., 320–400 nm. In addition to the radiative decay the excited singlet state undergoes intersystem crossing to yield the triplet state. This is very long lived and emissive in low temperature glasses and we have established its triplet energy,
ET |310 kJ mol 21 . From this we can postulate that it may act as a sensitiser for thymine dimerisation. The lowest triplet energy of thymine has been estimated from phosphorescence measurements to be 315 kJ mol 21 [23]. It has also been established that triplet–triplet energy transfer will occur quite efficiently from donors whose triplet energies are either above or 8–12 kJ mol 21 below that of the acceptor [24]. This implies that there is at least the possibility that this compound could sensitise the formation of thymine triplets resulting in the generation of thymine dimers, which are known to be potential precursors to skin cancer [25]. The triplet state has also been found to be a sensitiser of singlet oxygen, which is itself a harmful species in biological systems. We have assigned the transient species formed following excitation of NAMA at 266 nm to the triplet state by virtue of its quenching by oxygen. The absorption spectra of the transient species in all solvent systems were similar, with an absorption maximum at 450 nm, suggesting that the same species is formed in each case. The decrease in triplet state lifetime observed on changing from ethanol through acetonitrile to cyclohexane indicates that the species formed is more stable in polar environments. In addition to the radiative decay and oxygen quenching we have shown that the triplet state can also be deactivated by P-type delayed fluorescence. This was observed from degassed solutions of this compound at room temperature following 266 nm pulsed excitation and assigned on the basis of the quadratic behaviour of intensity with laser power. Finally, the triplet state is also quenched by the ground state, although this is only significant in degassed solutions of higher concentrations. By comparison of the results reported here with work involving sunscreen formulations containing menthyl anthranilate [26–29], it is likely that the use of NAMA within a sunscreen formulation where it is the major absorbing species is likely to result in a fluorescent product. Most of the fluorescent emission will be in the UV-A range and, as discussed previously [5], a significant amount will be transmitted to the skin. This indicates that incorporation of this ingredient into a formulation could result in a product that may, in fact, increase the intensity of radiation at the skin surface in this range. More significantly, however, the results of laser flash photolysis and near infrared luminescence measurements clearly demonstrate the formation of a long-lived triplet state which does sensitise singlet oxygen with a significant quantum yield in dilute solution and, in vivo could sensitise thymine dimerisation. In many respects the compound behaves in a similar fashion to the parent, menthyl anthranilate, the major difference being that acetylation of the amine group gives rise to a blue shift in the absorption and emission spectra, a reduction in the fluorescence quantum yield and an increase in the quantum yield of singlet oxygen formation.
A. Beeby, A.E. Jones / Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109 – 116
5. Conclusion A thorough study of the photophysical properties of N-acetyl-menthyl anthranilate has been conducted using conventional room temperature and low temperature luminescence spectroscopy, laser flash photolysis, and near infrared luminescence measurements. This compound was synthesised as an analogue of N-acetyl-homomenthyl anthranilate, a compound that, until relatively recently, was approved for use as a sunscreen in the EU. The luminescence studies indicate that, following excitation, radiative decay from the S 1 state resulting in fluorescence occurs to a significant extent (Ff |0.1560.02), but that intersystem crossing to the triplet state is the dominant deactivation pathway (e.g., FD 50.2060.02 in ethanol). The triplet state has been shown to have a long lifetime in degassed solution (120–240 ms) and a long lived phosphorescence lifetime of |1.3 s when held in low temperature organic glasses. From the phosphorescence emission spectrum the triplet state energy, ET , has been estimated to be |310 kJ mol 21 . This relatively high value indicates that this compound has the potential to sensitise the formation of thymine triplets in skin resulting in the generation of thymine dimers. Kinetic absorption measurements on degassed solutions have revealed a triplet–triplet absorption band at |420 nm that is independent of solvent and that is readily quenched by oxygen on aeration. Also, under high intensity pulsed illumination delayed fluorescence has been observed. The intensity of this emission increased exponentially with the laser power and was quenched by oxygen indicating P-type delayed fluorescence resulting from triplet–triplet annihilation. Finally, in aerated solutions this N-acetyl anthranilate ester has been shown to produce singlet oxygen, with a quantum yield of 0.2060.02 in all solvents studied.
6. Abbreviations NAMA UV EPA
N-Acetyl menthyl anthranilate Ultraviolet Ether–pentane–alcohol (5:5:2)
Acknowledgements The authors acknowledge the Royal Society for funding equipment used in this study and the University of Durham for a studentship (A.E.J.).
References [1] F.P. Gasparro, M. Mitchnick, J.F. Nash, A review of sunscreen safety and efficacy, Photochem. Photobiol. 68 (1998) 243–256.
115
[2] L.H. Kligman, F.J. Akin, A.M. Kligman, The contributions of UVA and UVB to connective tissue damage in hairless mice, J. Invest. Dermatol. 84 (1985) 272–276. [3] N.A. Shaath, Evolution of modern sunscreen chemicals, in: N.J. Lowe, N.A. Shaath, M.A. Pathak (Eds.), Sunscreens – Development, Evaluation, and Regulatory Aspects, Marcel Dekker, New York, 1997, pp. 3–33. [4] L.E. Agrapidis-Paloymis, R.A. Nash, N.A. Shaath, The effect of solvents on the ultraviolet absorbance of sunscreens, J. Soc. Cosmet. Chem. 38 (1987) 209–221. [5] A. Beeby, A.E. Jones, The photophysical properties of menthyl anthranilate: a UV-A sunscreen, Photochem. Photobiol. 72 (2000) 10–15. [6] M. Brown, Boots plc, personal communication, 2001. [7] H. Dannenberg, Classification of ultraviolet absorption. I. Substituted acetophenines, benzoic acids and cinnamic acid, Z. Naturforsch. 4b (1949) 327–344. [8] H.E. Ungnade, Ultraviolet absorption spectra of acetanilides, J. Am. Chem. Soc. 76 (1954) 5133–5135. [9] P. Grammaticakis, The ultraviolet absorption of o-substituted anilines III. N-Substituted o-nitro- and o-carboxyanilines, Bull. Soc. Chim. France (1950) 158–166. [10] P. Grammaticakis, Near ultraviolet and visible absorption of some isomeric arylamines and their nitrogen substituted derivatives I, Bull. Soc. Chim. France (1951) 220–226. [11] R. Staudemayer, T.D. Roberts, Neighbouring group participation in photolysis of o-substituted aryls, Tetrahedron Lett. 13 (1974) 1141– 1144. [12] V.I. Stenberg, in: O.L. Chapman (Ed.), Organic Photochemistry, Vol. I, Marcel Dekker, New York, 1967. [13] A.E. Jones, A Spectroscopic Study of Sunscreen, Ph.D. Thesis, University of Durham, 2000. [14] R.V. Bensasson, E.J. Land, T.G. Truscott, Excited States and Free Radicals in Biology and Medicine, Oxford University Press, New York, 1993. [15] D.F. Eaton, Luminescence spectroscopy, in: J.C. Scaiano (Ed.), Handbook of Organic Photochemistry, Vol. 1, CRC Press, Boca Raton, FL, 1989, pp. 231–239. [16] C.A. Parker, in: Photoluminescence of Solutions, Elsevier, Amsterdam, 1968, pp. 266–267. [17] S.L. Murov, I. Carmichael, G.L. Hug, in: Handbook of Photochemistry, 2nd Edition, Marcel Dekker, New York, 1995, pp. 283– 297. [18] F. Wilkinson, W.P. Helman, A.B. Ross, Quantum yields for the photosensitized formation of the lowest excited singlet state of molecular oxygen in solution, J. Phys. Chem. Ref. Data 22 (1) (1993) 113–262. [19] S. Nonell, S.L. Braslavsky, Time resolved singlet oxygen detection, in: L. Packer, H. Sies (Eds.), Methods in Enzymology, Singlet Oxygen, UV-A and Ozone, Vol. 319, Academic Press, San Diego, CA, 2000, pp. 37–49. [20] R.C. Weist (Ed.), Handbook of Chemistry and Physics, 61st Edition, CRC Press, Boca Raton, FL, 1980. [21] A.A. Gorman, M.A.J. Rodgers, Singlet oxygen, in: J.C. Scaiano (Ed.), Handbook of Organic Photochemistry, Vol. II, CRC Press, Boca Raton, FL, 1989, pp. 229–247. [22] A.A. Fisher, Sunscreen dermatitis: part IV – the salicylates, the anthranilates, and physical agents, Cutis 50 (1992) 397–398. [23] A.A. Lamola, Applications of electronic energy transfer in solution, Photochem. Photobiol. 8 (1968) 601–616. [24] D.F. Evans, Perturbation of singlet–triplet transitions of aromatic molecules by oxygen under pressure, J. Chem. Soc. (1957) 1351– 1537. [25] H. Gonzenbach, T.J. Hill, T.G. Truscott, The triplet energy levels of UVA and UVB sunscreens, J. Photochem. Photobiol. B: Biol. 16 (1992) 377–379.
116
A. Beeby, A.E. Jones / Journal of Photochemistry and Photobiology B: Biology 64 (2001) 109 – 116
[26] L.E. Rhodes, B.L. Diffey, Quantitative assessment of sunscreen application technique by in vivo fluorescence spectroscopy, J. Soc. Cosmet. Chem. 47 (1996) 109–115. [27] L.E. Rhodes, B.L. Diffey, Fluorescence spectroscopy: a rapid, noninvasive method for measurement of skin surface thickness of topical agents, Br. J. Dermatol. 136 (1997) 12–17. [28] R.M. Azurdia, J.A. Pagliaro, B.L. Diffey, L.E. Rhodes, Sunscreen
application by photosensitive patients is inadequate for protection, Br. J. Dermatol. 140 (1999) 255–258. [29] R.P. Stokes, B.L. Diffey, The feasibility of using fluorescence spectroscopy as a rapid, non-invasive method for evaluating sunscreen performance, J. Photochem. Photobiol. B: Biol. 50 (1999) 137–143.