Fluorescence Quantum Yields and Lifetimes of Aqueous Natural Dye Extracted from Tradescantia pallida purpurea at Different Hydrogen Potentials

: In this work, we monitored the ﬂuorescence quantum efﬁciency ( η ) and the ﬂuorescence lifetime ( τ ) of natural dye extracts from the leaves of Tradescantia pallida purpurea. The natural dye was extracted from leaves in aqueous solutions as a function of the potential of hydrogen (pH). The η was determined from conical diffraction (CD) pattern measurements due to thermally-driven self-phase modulation. The ﬂuorescence spectra and time-resolved ﬂuorescence measurements corroborate the CD results, and the average η ≈ 0.28 and τ ≈ 3.1 ns values were obtained in the pH range 3.96–8.02. In addition, the extracted natural dye was tested as a possible colorimetric and/or ﬂuorometric pH indicator in milk.


Introduction
The growing concern about the environment and the health of the population has expanded the search for more sustainable, natural, and healthy products [1][2][3][4]. Natural dyes have gained the attention of both consumers and researchers due to their potential antioxidant, anticarcinogenic, antimicrobial, fungicidal, biochemical, and pharmacological effects, and due to their biodegradable and renewable sources [5][6][7][8]. Moreover, the use of synthetic dyes has some disadvantages, such as the possibility of carcinogenic agents, water pollution during the dyeing process, non-renewable sources, and degradation time [9][10][11][12].
Anthocyanin is an important source of natural colorant and this dye has been extracted from different plants, flowers, fruits, and tubers [5,13,14]. Anthocyanins are derived from the flavonol compounds with the basic structure of a flavylium ion (C 15 H 11 O + ), or anthocyanidins that do not have glycated groups [14]. Anthocyanidins consist of a substance with an aromatic ring (A) that is attached to a heterocyclic ring (C). Ring (C) has oxygen, which is responsible for making a carbon-carbon bond to a third aromatic ring (B) [14,15]. When anthocyanidins are found in their glycosylated form, they are called anthocyanins [14]. Anthocyanins strongly absorb the visible region of the electromagnetic spectrum, generating a variety of colors in plants such as red, blue, and purple, according to the different patterns of hydroxyl and methoxyl groups that are attached to the aromatic ring (B), and the sugars and acylated sugars [14][15][16]. Furthermore, anthocyanins have a wide field of applications in health, such as suppression of neuroinflammation, neuronal degradation, and brain aging [17]. These natural dyes also hold promise in preventing diseases such as cancer [18], cardiovascular illness [19], obesity [20], and diabetes [21], and improving visual health [22] due to the possibility of anthocyanins having antioxidant, antiangiogenesis, and antimicrobial effects [23]. In addition, anthocyanin has the potential

Sample Preparation
Anthocyanin was extracted from the leaves of Tradescantia pallida purpurea that were collected from the city of Uberlândia (state of Minas Gerais, Brazil). The process that was carried out to extract the natural dye consisted of washing the leaves with tap and distilled water [7,31] and drying them on absorbent paper. After this, the leaves were cut into small strips and their mass (17 g) was determined by a Shimadzu analytical balance (Model AUW220D, Shimadzu Brazil, SP, Brazil). Next, the leaves were manually macerated at room temperature using a mortar and pestle for approximately twenty minutes with 20 mL of aqueous solution at a specific pH. The homogeneous mixture was then sieved and filtered with paper filter to obtain the dye solution. The aqueous dye was centrifuged for 1.5 h at 7200 rpm (BioPet Model 8011154, Biosigma, SP, Brazil). Then, the samples were stored and refrigerated (~5 • C). A similar procedure was repeated for natural dye extraction in aqueous solutions at different pH values (2.87, 3.66, 5.97, 10.05, 10.30, and 10.80). Distilled water with different concentrations of acetic acid (Sigma-Aldrich) was used for the preparation of acidic solutions, and ammonium hydroxide (NH 4 OH) (Synth) was used for the alkaline solutions. The natural dye was tested as a possible pH indicator in whole ultra-high-temperature (UHT) milk (Itambé Alimentos SA, Patos de Minas, Minas Gerais, Brazil) and powdered milk (Nestlé Brasil Ltd.a, Ituiutaba, Minas Gerais, Brazil). The powdered milk samples were prepared using 5 mL of distilled water (at different pH values) and 625 mg of powdered milk. The natural dye (at 0.85 g/mL) was inserted in milk at different pH values, and the samples were homogenized using a magnetic stirrer for 1 h. pH values of 3.47 to 10.43 were used for the natural dye that was inserted in milk that was prepared from powdered milk. A similar procedure was performed using 5 mL of whole UHT milk and natural dye (pH range used between 3.47 and 10.43). Aqueous solutions of natural dye were inserted into the milk samples with different pH values, at a concentration of 40% of natural dye, for fluorescence and lifetime measurements.

Spectroscopic Techniques
Absorption spectra measurements were performed using a 1 cm quartz cuvette in a portable spectrometer that was equipped with a halogen lamp (Ocean Optics USB2000+, Dunedin, FL, USA) at room temperature. Fluorescence spectra (excitation wavelength at 532 nm) were obtained using a portable spectrometer (DPSSL Driver model MGL532, Midvale, UT, USA). The pH values were determined using an HI 2221 pH meter from Hanna Instruments. Fourier transform infrared (FTIR) spectroscopy was performed using a Perkin-Elmer Frontier spectrometer (resolution of 2 cm −1 , PerkinElmer Brazil, SP, Brazil). Time-resolved fluorescence lifetimes (TRFL) were performed by using a light source PLS 450 LED (λ = 460 ± 10 nm, 40 µW average power at 40 MHz and 800 ps). TRFL results for liquid samples were analyzed within a FluoTime 100 time-resolved fluorescence spectrometer from PicoQuant (Berlin, Germany) using a 1 cm thick quartz cuvette [41]. A Ludox solution was used as a scattering sample for prompt measurements [41,42].

Thermo-optical Techniques
The CD technique takes advantage of the ring pattern that arises due to thermallyinduced self-phase-modulation (TSPM) effects [7,37,43,44], behind the cuvette of liquid samples when a relatively high-powered laser beam is focused on a sample. These rings usually emerge when the photothermal phase shift (∆ϕ TH ) that is induced is high enough (∆ϕ TH >> 2π). TSPM effects can be understood as the ability of the excitation beam to induce spatial variations in the refractive index, which leads to a phase shift that depends on the transverse distance from the beam axis [45][46][47]. For a thermally-induced phase change, the number of rings (N) can be determined in the from [45,48]: where P e is the power of the excitation laser beam, K is the thermal conductivity, α (cm −1 ) is the optical absorption coefficient at the excitation wavelength (λ), L eff = (1 − exp(−αL))/α is the sample effective length, and L is the sample thickness. The fraction of energy that is converted into heat (ϕ) is related to the fluorescence quantum yield η parameter by the expression [37,43]: where <λ em > is the average emission wavelength. Here, the CD technique was applied using a linearly polarized argon-ion laser at 514.5 nm that passes through a rotating linear polarizer that is controlled by a step motor to modulate the light intensity. The laser beam, after passing through a focusing lens (f = 20 cm), is folded vertically by a mirror and impinges on a quartz cuvette (2 mm thick) that is placed horizontally [37]. Typical ring patterns were observed when the sample was positioned at the focus of the pump beam due to TSPM effects [37,[43][44][45]].

Single Arm Double Interferometer
The refractive index temperature coefficient (dn/dT) was determined by using a singlearm double interferometer [37,38]. The Mach-Zehnder interferometric technique (MZI) at a wavelength value of λ = 532 nm was applied for liquid samples, and dn/dT was determined using the expression [37] dn where L is the cuvette thickness (L = 1 mm) and ∆T is the temperature spacing between two consecutive fringes.  Figure 1 presents the absorbance spectra and photos of the natural dye solutions that were extracted from the leaves of Tradescantia pallida purpurea in different pH values (3.93-8.02). The absorption bands were observed at (508 ± 2), (545 ± 1), and (586 ± 1) nm for all-natural dyes that were extracted in different pH values (Peaks 1, 2, and 3, respectively, as shown in Figure 1). The typical absorbance peaks that were obtained were similar to the results that were reported for the anthocyanin natural dye that was extracted from Tradescantia pallida in different solvents as aqueous solutions, sodium citrate buffer (in the pH range of 4-7), ethanol, and acetone [7,31,49,50]. The relationship between absorbance peaks 2 and 3 for the extracted natural dyes as a function of pH are presented in Table 1. The typical absorption bands that were observed in the 500-590 nm region and mainly at 586 nm (peak 3, Figure 1) correspond to B-ring substituted anthocyanins with a quinonoidal base structure [49][50][51], associated with color stability proprieties due to complex acylation patterns [51]. The pH increases of the aqueous solutions that were used in the natural extraction process promoted a decrease in absorbance peak 2 (Figure 1), as shown in Table 1.

Results and Discussion
Photochem 2023, 1, FOR PEER REVIEW 4 where L is the cuvette thickness (L = 1 mm) and ΔT is the temperature spacing between two consecutive fringes. Figure 1 presents the absorbance spectra and photos of the natural dye solutions that were extracted from the leaves of Tradescantia pallida purpurea in different pH values (3.93-8.02). The absorption bands were observed at (508 ± 2), (545 ± 1), and (586 ± 1) nm for allnatural dyes that were extracted in different pH values (Peaks 1, 2, and 3, respectively, as shown in Figure 1). The typical absorbance peaks that were obtained were similar to the results that were reported for the anthocyanin natural dye that was extracted from Tradescantia pallida in different solvents as aqueous solutions, sodium citrate buffer (in the pH range of 4-7), ethanol, and acetone [7,31,49,50]. The relationship between absorbance peaks 2 and 3 for the extracted natural dyes as a function of pH are presented in Table 1. The typical absorption bands that were observed in the 500-590 nm region and mainly at 586 nm (peak 3, Figure 1) correspond to B-ring substituted anthocyanins with a quinonoidal base structure [49][50][51], associated with color stability proprieties due to complex acylation patterns [51]. The pH increases of the aqueous solutions that were used in the natural extraction process promoted a decrease in absorbance peak 2 (Figure 1), as shown in Table 1. In Zebrina anthocyanin solution (at pH 1.37), it has been reported that the pigment exists only as the flavylium cation, with a maximum wavelength of 537 nm, and when the acidity of the natural dye decreases, the cation is converted to the quinoidal base [49]. The B-ring substituted anthocyanins completely converted to the quinoidal base are reported at pH ~5.5 [49,50]. Another important aspect of anthocyanin to be addressed is the change in its color due to pH, which is due to the ionic nature of the dye structure. In general, at more acidic pH values (pH 1), the structural base of anthocyanins is the flavylium cation (red color). With a change in pH values, the dye structure becomes a carbinol pseudobase, with quinonoidal and chalcone structures [6,15]. Anthocyanin color (inserted in Figure 1) is reported to be dependent on the structure, pH values, temperature, and other factors [6,15,52]. Furthermore, studies show that anthocyanins can undergo degradation at higher alkaline pH values, depending on the substituent groups that are attached to the structure [6].  In Zebrina anthocyanin solution (at pH 1.37), it has been reported that the pigment exists only as the flavylium cation, with a maximum wavelength of 537 nm, and when the acidity of the natural dye decreases, the cation is converted to the quinoidal base [49]. The B-ring substituted anthocyanins completely converted to the quinoidal base are reported at pH~5.5 [49,50]. Another important aspect of anthocyanin to be addressed is the change in its color due to pH, which is due to the ionic nature of the dye structure. In general, at more acidic pH values (pH 1), the structural base of anthocyanins is the flavylium cation (red color). With a change in pH values, the dye structure becomes a carbinol pseudobase, with quinonoidal and chalcone structures [6,15]. Anthocyanin color (inserted in Figure 1) is reported to be dependent on the structure, pH values, temperature, and other factors [6,15,52]. Furthermore, studies show that anthocyanins can undergo degradation at higher alkaline pH values, depending on the substituent groups that are attached to the structure [6]. Figure 2 presents the fluorescence spectra at an excitation wavelength of λ = 532 nm, for the aqueous dyes that were extracted from the leaves of Tradescantia pallida purpurea at different pH values (~3.9-8.0). The values of <λ em > and Stoke s shift wavelength (∆λ) for each natural dye sample (a-f) are presented in Table 1, and the average result for all the natural dye solutions that were extracted in different pH values were <λ em > = (652 ± 3) nm and ∆λ = (25 ± 1) nm, respectively. The peak emission wavelengths that were obtained were positioned at (611 ± 2) nm and (648 ± 1) nm for all natural dyes that were extracted in different pH values. The fluorescence spectra of natural dyes that were extracted from Tradescantia pallida purpurea in distilled water and <λ em > = 648 nm were reported [7]. The decrease in the fluorescence intensity for more alkaline extractions of the natural dye can be justified due to possible degradation effects [6,52].  Figure 2 presents the fluorescence spectra at an excitation wavelength of λ = 532 nm, for the aqueous dyes that were extracted from the leaves of Tradescantia pallida purpurea at different pH values (~3.9-8.0). The values of <λem> and Stoke´s shift wavelength (Δλ) for each natural dye sample (a-f) are presented in Table 1, and the average result for all the natural dye solutions that were extracted in different pH values were <λem> = (652 ± 3) nm and Δλ = (25 ± 1) nm, respectively. The peak emission wavelengths that were obtained were positioned at (611 ± 2) nm and (648 ± 1) nm for all natural dyes that were extracted in different pH values. The fluorescence spectra of natural dyes that were extracted from Tradescantia pallida purpurea in distilled water and <λem> = 648 nm were reported [7]. The decrease in the fluorescence intensity for more alkaline extractions of the natural dye can be justified due to possible degradation effects [6,52]. Fourier-transform infrared (FTIR) spectra for the powder of natural dyes that were extracted in different pH values are presented in Figure 3. The main vibrational modes were observed in the 700-3500 cm −1 range at 3187, 1555, 1380, 1085, and 820 cm −1 for all the natural dyes in different pH values. For powdered anthocyanin that was extracted in distilled water [7], the main bands that were assigned were reported at 3342 cm −1 , attributed to the stretching vibration of hydroxyl (-OH); 1558 cm −1 , due to the stretching vibration in the aromatic rings of carbonyl groups (C=C) or (C=O) for anthocyanin dye; and 1081 cm −1 , attributed to the C-O-C vibration for anthocyanin dye [7,53,54]. A band at 1380 cm −1 belongs to the fingerprint region (1542-965 cm −1 ), where different IR bands exist due to C-O, C-C, C-H, and C-N bonds [55], and additional bands at 820 cm −1 caused by Fourier-transform infrared (FTIR) spectra for the powder of natural dyes that were extracted in different pH values are presented in Figure 3. The main vibrational modes were observed in the 700-3500 cm −1 range at 3187, 1555, 1380, 1085, and 820 cm −1 for all the natural dyes in different pH values. For powdered anthocyanin that was extracted in distilled water [7], the main bands that were assigned were reported at 3342 cm −1 , attributed to the stretching vibration of hydroxyl (-OH); 1558 cm −1 , due to the stretching vibration in the aromatic rings of carbonyl groups (C=C) or (C=O) for anthocyanin dye; and 1081 cm −1 , attributed to the C-O-C vibration for anthocyanin dye [7,53,54]. A band at 1380 cm −1 belongs to the fingerprint region (1542-965 cm −1 ), where different IR bands exist due to C-O, C-C, C-H, and C-N bonds [55], and additional bands at 820 cm −1 caused by ring vibrations C=C-C [55]. The anthocyanin structure is supported by the presence of these functional groups, such as the benzene ring, double bond, carbonyl group, C-H bond, and OH group [54]. The broad OH peaks are also characteristic of anthocyanin [55]. The discrete band at 3023 cm −1 and 2926 cm −1 , due to a methyl group (C-H 3 ), can be attributed to aliphatic CH vibrations [55], and with decreasing pH values, peaks at 3187 cm −1 increased, and other peaks at~3023 cm −1 and 2926 cm −1 appeared. Similar results were reported for anthocyanin in methanol solution with different pH values, where with the decrease of pH (from 3 to 1), the O-H peak gradually shifted from 3310 cm −1 to Photochem 2023, 3 6 3270 cm −1 due to the hydroxyl groups of methanol changing into the hydroxyl groups of anthocyanin [56]. The C-O group was detected at 1022 cm −1 , and the absorption band in the region of 2950-2840 cm −1 can be attributed to the C-H absorption stretching with the sugar group. The peak at 656 cm −1 , due to the presence of aromatic C-H bonds, is more apparent when the pH is more acidic [57]. Furthermore, the anthocyanin cyanidin compound has been reported to absorb in the range 3100-3400 cm −1 due to O-H, 2900-2840 cm −1 due to aliphatic C-H, 675-870 cm −1 due to aromatic C-H, and 1660 cm −1 due to the presence of aromatic C=C [57,58].

Results and Discussion
coupled with a quadrupole time-of-flight high-resolution mass spectrometer, and the results were described in our Reference [7]. The anthocyanin structures consisted mainly of cyanidin, three glucose molecules, arabinose, and three ferulic molecules [7,59]. Furthermore, the literature reports that the main anthocyanin structures of cyanidin-3,7,3′-triglucosides with three ferulic acid molecules and an additional terminal glucose molecule were obtained from dyes that were extracted from Tradescantia pallida leaves [51,60]. Figure 4 shows the results that were obtained using the conical diffraction technique; the number of rings (N) is presented as a function of beam power (Pe). The linear behavior of N versus Pe is presented in the inset of Figure 4, and the slope that was obtained was (18.0 ± 0.5) W −1 . The value of ϕ was determined for the natural dye using Eq. 1, the K parameters for aqueous solutions [37], and dn/dT determined using the MZI technique [37,38]. The average value that was obtained for dn/dT = −(0.92 ± 0.03) × 10 −4 K −1 (at 22 0 C) was in good agreement with the value that was reported in the literature for the pure aqueous solutions [7,37,61]. Using <λem> (Table 1), λ = 514.5 nm and Eq. 2, the radiative quantum yield η values were determined for the all-natural dye samples.  Natural dyes that were extracted from Tradescantia pallida purpurea in aqueous solutions were already analyzed using a high-performance liquid chromatography system coupled with a quadrupole time-of-flight high-resolution mass spectrometer, and the results were described in our Reference [7]. The anthocyanin structures consisted mainly of cyanidin, three glucose molecules, arabinose, and three ferulic molecules [7,59]. Furthermore, the literature reports that the main anthocyanin structures of cyanidin-3,7,3 -triglucosides with three ferulic acid molecules and an additional terminal glucose molecule were obtained from dyes that were extracted from Tradescantia pallida leaves [51,60]. Figure 4 shows the results that were obtained using the conical diffraction technique; the number of rings (N) is presented as a function of beam power (P e ). The linear behavior of N versus P e is presented in the inset of Figure 4, and the slope that was obtained was (18.0 ± 0.5) W −1 . The value of ϕ was determined for the natural dye using Equation (1), the K parameters for aqueous solutions [37], and dn/dT determined using the MZI technique [37,38]. The average value that was obtained for dn/dT = −(0.92 ± 0.03) × 10 −4 K −1 (at 22 • C) was in good agreement with the value that was reported in the literature for the pure aqueous solutions [7,37,61]. Using <λ em > (Table 1), λ = 514.5 nm and Equation (2), the radiative quantum yield η values were determined for the all-natural dye samples. Figure 5 presents the results of ϕ and η for natural dyes that were obtained from CD as a function of pH. The ϕ and η results were dependent on the pH of the aqueous solutions that were used for natural dye extractions. The decrease of η (Table 2) for more alkaline solutions can be attributed to the possible degradation [6,52], aggregation, and/or co-pigmentation effects [62,63]. Self-aggregation has been reported for low anthocyanin concentrations [62], and the weakening of their emission efficiency as typically observed in AIE (aggregation-induced emission) active molecules could be due to the activation of non-radiative decay by free molecular rotation [64]. Similar results were obtained for the intensity of fluorescence spectra as a function of pH (Figure 2), which decreased at higher pH values. The η average result for samples at different pH values in the range of (~4.0-8.0) was η = (0.28 ± 0.02). For comparison, for annatto that was extracted from the fruit of the Bixa Orellana trees at a pH between 5.4 and 11.5, the average value of η = (0.40 ± 0.08) was reported [41], and for the Carthamus tinctorius L. petal, η ≈ 0.004 was obtained for the main colored species that was extracted [65].  Figure 5 presents the results of ϕ and η for natural dyes that were obtained from CD as a function of pH. The ϕ and η results were dependent on the pH of the aqueous solutions that were used for natural dye extractions. The decrease of η (Table 2) for more alkaline solutions can be attributed to the possible degradation [6,52], aggregation, and/or co-pigmentation effects [62,63]. Self-aggregation has been reported for low anthocyanin concentrations [62], and the weakening of their emission efficiency as typically observed in AIE (aggregation-induced emission) active molecules could be due to the activation of non-radiative decay by free molecular rotation [64]. Similar results were obtained for the intensity of fluorescence spectra as a function of pH (Figure 2), which decreased at higher pH values. The η average result for samples at different pH values in the range of (~4.0-8.0) was η = (0.28 ± 0.02). For comparison, for annatto that was extracted from the fruit of the Bixa Orellana trees at a pH between 5.4 and 11.5, the average value of η = (0.40 ± 0.08) was reported [41], and for the Carthamus tinctorius L. petal, η ≈ 0.004 was obtained for the main colored species that was extracted [65].    Figure 5 presents the results of ϕ and η for natural dyes that were obtained from CD as a function of pH. The ϕ and η results were dependent on the pH of the aqueous solutions that were used for natural dye extractions. The decrease of η ( Table 2) for more alkaline solutions can be attributed to the possible degradation [6,52], aggregation, and/or co-pigmentation effects [62,63]. Self-aggregation has been reported for low anthocyanin concentrations [62], and the weakening of their emission efficiency as typically observed in AIE (aggregation-induced emission) active molecules could be due to the activation of non-radiative decay by free molecular rotation [64]. Similar results were obtained for the intensity of fluorescence spectra as a function of pH (Figure 2), which decreased at higher pH values. The η average result for samples at different pH values in the range of (~4.0-8.0) was η = (0.28 ± 0.02). For comparison, for annatto that was extracted from the fruit of the Bixa Orellana trees at a pH between 5.4 and 11.5, the average value of η = (0.40 ± 0.08) was reported [41], and for the Carthamus tinctorius L. petal, η ≈ 0.004 was obtained for the main colored species that was extracted [65].   The time-resolved fluorescence (TRFL) spectroscopy measurements were also performed for natural dyes that were extracted in different pH values. TRFL measurements of the excited states were obtained and the experimental results were adjusted using the single exponential A + B × Exp (−t/τ), where A and B are constants, and τ is the fluorescence lifetime. The TRFL setup was tested by using the sodium fluorescein (Synth, molecular weight 376.28) in an alkaline aqueous solution (pH 11.24), and the transient emission was adjusted with a single exponential (τ = (4.13 ± 0.01) ns and χ 2 = (1.03 ± 0.01)) [41,42]. Figure 6 shows the TRFL for aqueous natural dye that was extracted at different pH values (Figure 6 (a) 3.98, (b) 5.34, (c) 6.75, and (d) 7.99) from ornamental leaves of Tradescantia pallida purpurea. The fluorescence decays were well fitted with a single exponential, and the χ 2 value was less than 1.15 for all the natural dye aqueous solutions that were extracted as a function of pH ( Table 2). ular weight 376.28) in an alkaline aqueous solution (pH 11.24), and the transient emission was adjusted with a single exponential (τ = (4.13 ± 0.01) ns and χ 2 = (1.03 ± 0. 01)) [41,42]. Figure 6 shows the TRFL for aqueous natural dye that was extracted at different pH values (Figure 6 (a) 3.98, (b) 5.34, (c) 6.75, and (d) 7.99) from ornamental leaves of Tradescantia pallida purpurea. The fluorescence decays were well fitted with a single exponential, and the χ 2 value was less than 1.15 for all the natural dye aqueous solutions that were extracted as a function of pH (Table 2).  The lifetime results that were obtained for aqueous natural dyes are dependent on the pH (Figure 7), with τ increasing for more acidic solutions. A similar dependence on pH was observed for the fluorescence area that was obtained from fluorescence spectra ( Figure 7) and η ( Figure 5). The average lifetime that was obtained for aqueous anthocyanin dyes that were extracted in different pH values (between ~ 4 and 8) was (3.1 ± 0.5) ns. The decrease in the fluorescence response generally indicates increased quenching of the emitting center [66], solution degradation, and/or aggregation [6,52,62]. For comparison, the natural dyes from annatto that were extracted from the seeds of the tropical shrub Bixa The lifetime results that were obtained for aqueous natural dyes are dependent on the pH (Figure 7), with τ increasing for more acidic solutions. A similar dependence on pH was observed for the fluorescence area that was obtained from fluorescence spectra ( Figure 7) and η ( Figure 5). The average lifetime that was obtained for aqueous anthocyanin dyes that were extracted in different pH values (between~4 and 8) was (3.1 ± 0.5) ns. The decrease in the fluorescence response generally indicates increased quenching of the emitting center [66], solution degradation, and/or aggregation [6,52,62]. For comparison, the natural dyes from annatto that were extracted from the seeds of the tropical shrub Bixa Orellana L. were reported present τ = (1.9 ± 0.3) ns when extracted in aqueous solutions with different pH values (~5. 6-11.5), and other typical probes (such as fluorescein, rhodamine 6G, and bacteriochlorophyll), and some fluorescent proteins, present lifetimes in the range of 1-4 ns [41,42,67]. In addition, the fluorescent NIR dye cypate has been reported to present a longer fluorescence lifetime when the pH of the dye changed from a neutral to an acidic value [67].
Orellana L. were reported present τ = (1.9 ± 0.3) ns when extracted in aqueous solutions with different pH values (~5. 6-11.5), and other typical probes (such as fluorescein, rhodamine 6G, and bacteriochlorophyll), and some fluorescent proteins, present lifetimes in the range of 1-4 ns [41,42,67]. In addition, the fluorescent NIR dye cypate has been reported to present a longer fluorescence lifetime when the pH of the dye changed from a neutral to an acidic value [67]. The natural dyes that were extracted from the leaves of Tradescantia pallida purpurea were tested as a bioindicator of pH change. Figure 8 shows the fluorescence spectra for anthocyanin (pH 5.450 ± 0.004) inserted in whole UHT milk that was prepared at different pH values (range ~3. 47-10.43). The change of the pH in milk was performed to simulate adulterated milk due to bacterial contamination (such as Mycoplasma bovis and some Salmonella serotypes) [68][69][70] or other types of contamination, such as urea, starch, water, detergents, caustic soda, and milk from different species [71]. The change in pH values in foods is reported to be due to microbial, chemical, and/or other adulteration actions [30,41,72]. Similar results were obtained for the fluorescence spectra of anthocyanin dye that was inserted into powdered milk at different pH values. The change of the color of the aqueous natural dye that was previously extracted and inserted in milk with different pH values is presented in the photos in Figure 8. For comparison, the results were reported for a natural dye that was extracted from red resin that was removed from the pericarp of annatto seeds [41]. However, in this case, the annatto powder was extracted in milk at different pH values (6.64-10.86) [41]. In comparison with Figure 2, it is possible to observe a change in the fluorescence spectra that is shown in Figure 8. The fluorescence spectra of natural dye that was extracted at different pH values and the anthocyanin that is inserted into milk are different for higher pH values (above ~8). For alkaline pH values, the natural dye fluorescence decreases significantly (Figures 2 and 8), and possible milk fluorescence is predominant. The natural dyes that were extracted from the leaves of Tradescantia pallida purpurea were tested as a bioindicator of pH change. Figure 8 shows the fluorescence spectra for anthocyanin (pH 5.450 ± 0.004) inserted in whole UHT milk that was prepared at different pH values (range~3. 47-10.43). The change of the pH in milk was performed to simulate adulterated milk due to bacterial contamination (such as Mycoplasma bovis and some Salmonella serotypes) [68][69][70] or other types of contamination, such as urea, starch, water, detergents, caustic soda, and milk from different species [71]. The change in pH values in foods is reported to be due to microbial, chemical, and/or other adulteration actions [30,41,72]. Similar results were obtained for the fluorescence spectra of anthocyanin dye that was inserted into powdered milk at different pH values. The change of the color of the aqueous natural dye that was previously extracted and inserted in milk with different pH values is presented in the photos in Figure 8. For comparison, the results were reported for a natural dye that was extracted from red resin that was removed from the pericarp of annatto seeds [41]. However, in this case, the annatto powder was extracted in milk at different pH values (6.64-10.86) [41]. In comparison with Figure 2, it is possible to observe a change in the fluorescence spectra that is shown in Figure 8. The fluorescence spectra of natural dye that was extracted at different pH values and the anthocyanin that is inserted into milk are different for higher pH values (above~8). For alkaline pH values, the natural dye fluorescence decreases significantly (Figures 2 and 8), and possible milk fluorescence is predominant.
The TRFL results were performed in anthocyanin that was inserted in milk with different pH values.
The fluorescence decays were adjusted with one exponential (χ 2 < 1.2 were obtained for all samples), and the τ results that are presented in Figure 9a were determined from natural dye that was inserted in milk with different pH values. Similar results were obtained for the area of fluorescence spectra (AFS) in the function of the pH-changed milk (Figure 9b). The τ and AFS results as a function of the pH range available between 3.47 and 10.43 are presented in Figure 9a,b, and these parameters changed bỹ 12 and 50%, respectively. The change in color (inserted in Figure 8) and fluorescence of the anthocyanin-milk samples as a function of pH raises the possibility of using this natural dye as a colorimetric and/or fluorimetric pH-biosensor probe [29,30,41]. Several applications of natural dyes as possible biosensors for adulterated pH-dependent foods have been reported (Table 3) [30,41,[72][73][74][75]. For applications of natural dyes as possible pH-biosensors, similar behavior of τ and ADF as a pH function were obtained (Figure 9a,b).
However, as a possible parameter for biosensor analysis, the ADF versus pH results could be the simplest technique for implementation. The TRFL results were performed in anthocyanin that was inserted in milk with different pH values. The fluorescence decays were adjusted with one exponential (χ 2 < 1.2 were obtained for all samples), and the τ results that are presented in Figure 9(a) were determined from natural dye that was inserted in milk with different pH values. Similar results were obtained for the area of fluorescence spectra (AFS) in the function of the pHchanged milk (Figure 9 (b)). The τ and AFS results as a function of the pH range available between 3.47 and 10.43 are presented in Figure 9 (a,b), and these parameters changed by ~12 and 50%, respectively. The change in color (inserted in Figure 8) and fluorescence of the anthocyanin-milk samples as a function of pH raises the possibility of using this natural dye as a colorimetric and/or fluorimetric pH-biosensor probe [29,30,41]. Several applications of natural dyes as possible biosensors for adulterated pH-dependent foods have been reported (Table 3) [30,41,[72][73][74][75]. For applications of natural dyes as possible pH-biosensors, similar behavior of τ and ADF as a pH function were obtained (Figure 9a,b). However, as a possible parameter for biosensor analysis, the ADF versus pH results could be the simplest technique for implementation.

Natural Dye Extraction Applications
Brassica oleraceae var. capitata (Red Cabbage) [73] Development of chitosan/PVA films doped with anthocyanins to indicate food quality due to pH change. Potential application in the evaluation of milk quality [73]. . Fluorescence (a) lifetime and (b) spectra areas (λ = 532 nm, and P e ≈ 50 mW) for natural dye that was inserted in milk as a function of the pH value (L = 1 cm). The dashed adjustment is shown for eye guidance only. Table 3. Applications of natural dyes as pH indicators in foods.

Natural Dye Extraction Applications
Anthocyanin Brassica oleraceae var. capitata (Red Cabbage) [73] Development of chitosan/PVA films doped with anthocyanins to indicate food quality due to pH change. Potential application in the evaluation of milk quality [73]. Hibiscus rosa-sinensis L. flowers, Clitoria ternatea flowers, Beta vulgaris roots, Opuntia dillenii pricklypears [30] Application of liquid coloring in raw milk to measure microbiological quality control [30].

Curcumin, Quercetin, and Phycocyanin
Vegetable plants, and microalga Spirulina [74] Introduction of natural dyes in nanofibers for use as pH colorimetric indicators as a function of time to monitor food quality [74].

Annatto
Seeds of the shrub Bixa Orellana L. [41] The natural dyes were tested in adulterated milk as colorimetric and/or fluorometric pH-biosensor probes [41].

Litmus
Lichens [75] Development of a pH-sensing film from polysaccharide extracted from tamarind and litmus lichen, as an indicator of deterioration of cream milk [75].
The natural dye that was extracted from the leaves of Tradescantia pallida purpurea was tested in the present work, with a possible colorimetric and fluorometric pH-sensor using different techniques such as AFS, TRFL, and CD. In this case, the aqueous solutions of anthocyanin dye that were previously extracted were inserted in pH-changed milk. The AFS, η, and τ results were obtained with the natural dye that was extracted from the leaves of ornamental plants; these leaves are available in different locations. For comparison, natural dyes that were obtained from annatto seeds were tested as pH sensors in milk [41]. However, in the case of our work [41], the natural dye needed to be extracted in the milk sample at different pH values. The seeds were obtained from the tropical shrub Bixa Orellana L., and these trees are restricted to tropical countries.

Conclusions
The fluorescence quantum yield (η), the area of fluorescence spectra (AFS), and the fluorescence lifetime (τ) parameters were determined for anthocyanin that was extracted from the leaves of Tradescantia pallida purpurea as a function of the potential of hydrogen. The η, AFS, and τ results were dependent on the pH that was used (~3.96-8.02) for aqueous natural dye extraction, with variations of~48, 50, and 12%, respectively. The effect of extractions of natural dyes at different pH values on photophysical and thermooptical properties allows the application of anthocyanin dye as a pH-dependent biosensor. Preliminary results are presented for the insertion of aqueous anthocyanin in milk (pH range 3.47-10.43), as a possible colorimetric and fluorometric pH-probe for food.