Solid-State Luminescence with a Large Stokes Shift in Starch Functionalized with Low-Content ESIPT Dye

Abstract

Herein, we present the preparation of solid-state photoactive starches with a large Stokes shift, along with the resulting materials. In this investigation, 2-(2′-hydroxyphenyl)benzazole derivatives responsive to intramolecular proton transfer in the excited state (ESIPT) were covalently bonded to the polymeric structure of starch through a reaction involving an isothiocyanate group and the hydroxyl groups of starch. These compounds exhibit absorption at approximately 350 nm, which is related to fully spin- and symmetry-allowed π → π* electronic transitions, and solid-state fluorescence at approximately 500 nm, which features a significant separation between the absorption and emission maxima (~9000 cm⁻¹). Due to the minimal use of fluorophores in functionalized starch preparation, this modification does not affect the original properties of the starch. Finally, photoactive starch-based films with significantly high transparency were successfully produced.

1. Introduction

Plastics have become ingrained in daily routines, as they find applications across various industries [1]. Originating from synthetic polymers derived from nonrenewable petroleum sources, these materials present environmental challenges due to their resistance to degradation and prolonged persistence in the environment [2]. As concerns about their environmental impact intensify, there is growing interest in alternative materials that can replace traditional polymers. In response to these concerns, biopolymers are emerging as promising substitutes, primarily sourced from renewable raw materials and often exhibiting biodegradable properties [3,4,5]. Starch, an abundantly available natural resource, serves as a cost-effective renewable raw material for biopolymer production [6,7,8,9,10]. The conversion of starch into a thermoplastic polymer, known as thermoplastic starch (TPS), is facilitated through the incorporation of plasticizers [11,12]. Plasticizers, such as glycerol, sorbitol, glycols, and urea, enable the processing of TPS by reducing intermolecular hydrogen bonds within starch grains. The type and proportion of plasticizer used significantly impact the physical properties of the processed TPS, including the glass transition temperature and mechanical characteristics. Due to its hydrophilic nature, TPS is highly susceptible to moisture absorption, which affects its mechanical properties. Efforts to alleviate this sensitivity involve blending TPS with other materials to enhance its final properties and reduce vulnerability to environmental factors such as humidity, temperature, and radiation exposure [13,14,15,16].

Starch can undergo functionalization by introducing chemical groups through hydroxyl groups, significantly broadening the range of applications for the resulting material [17,18,19]. This expansion is not limited to the acquisition of thermoplastic materials. In the literature, light-emitting groups are attached to polysaccharide skeletons, demonstrating diverse applications [20,21,22,23]. However, there are currently few reports on the functionalization of starch with fluorescent compounds using excited-state intramolecular proton transfer (ESIPT), usually related to advancing methods for detecting and visualizing latent fingermarks, with improvements in sensitivity, selectivity, and ease of visualization on diverse substrates [24,25].

In 1955, Weller discovered the excited-state intramolecular proton transfer (ESIPT) process in salicylic acid, which exhibited a remarkable Stokes shift compared to that of analogs [26]. ESIPT-active fluorophores, known for their photophysical stability and large Stokes shift (6000 to 12,000 cm⁻¹), have become a focal point in research [27,28,29,30]. These compounds, which have dual fluorescence emission and sensitivity to the environment, have diverse applications [31,32,33,34,35,36,37,38,39,40,41,42]. ESIPT in benzazole derivatives relies on factors such as intramolecular hydrogen bonds, electronic effects of donor (–OH) and acceptor (–C=N–) groups, and molecular geometry. The delicate balance between structural and electronic aspects determines the occurrence of ESIPT, which is essential for its successful execution [31]. Benzazole heterocycles, aromatic compounds with a benzene-azole ring fusion, play a crucial role in heterocyclic compounds, forming benzoxazoles, benzimidazoles, benzothiazoles, or benzoselenazoles [43]. These compounds, which contain diverse heteroatoms, are fundamental building blocks in medicinal chemistry for synthesizing biologically active molecules [44,45,46,47] and photoactive polymeric materials [48,49,50,51,52,53]. Regarding ESIPT and benzazoles, hydroxyphenylbenzazoles, which can undergo proton transfer in the excited state, exist predominantly as enol (E) species in the ground state (Figure 1) [28]. Upon excitation, a redistribution of charges leads to intramolecular proton transfer, giving rise to a new keto tautomer and fluorescence emission with a large Stokes shift. In addition, these fluorophores can also exhibit dual fluorescence emission, with distinct bands corresponding to additional enol conformers (blue line) and keto tautomers (red line) influenced by the surrounding medium [54].

In this context, our investigation focused on creating fluorescent starches by introducing excited-state intramolecular proton transfer (ESIPT) dyes through starch functionalization. The objective of this study was to incorporate the inherent photophysical properties of the ESIPT process, such as solid-state emission, fluorescence with a large Stokes shift, and photostability, into starch matrices. The ultimate aim is to produce novel biocompatible and biodegradable fluorescent materials. The development of these new materials is crucial for merging the advantages of fluorescence with environmentally friendly attributes. These materials have diverse applications, ranging from environmental monitoring to medical imaging, aligning with the increasing demand for sustainable and eco-friendly solutions. For this purpose, benzazoles functionalized with isothiocyanate will be used to covalently attach these fluorophores to starch, forming a thiocarbamate bond. The resulting functionalized starches will be employed in the fabrication of solid-state photoactive thermoplastic films.

2. Experimental Section

2.1. Materials and Methods

Natural corn starch (Amisol 3408, Ingredion Brasil Ing. Ind. Ltda, Mogi Guaçu, SP, Brazil) was used as received. Commercially available reagents and solvents (acetone, dichloromethane, dimethylsulfoxide, methanol, glycerin, and distilled water) were used as received without additional purification. Thin-layer chromatography (TLC) was performed using the supported silica gel GF254 (0.25 mm thickness). For visualization, TLC plates were placed under UV light at 254 nm. X-ray diffraction analyses were performed on a diffractometer (Shimadzu XRD 6000, Shimadzu, Kyoto, Japan) with k-α radiation and a wavelength of 1.5406 Å. The reading range was from 4 to 45° with a step of 2° min⁻¹. Thermogravimetric analysis (TGA) of the photoactive starches and respective thermoplastic films was conducted using a Shimadzu TGA-50 (Shimadzu) thermogravimetric analyzer under a nitrogen atmosphere with a flow rate of 50 mL·min⁻¹ and a heating rate of 10 °C·min⁻¹. Fourier transform infrared (FT-IR) spectra were recorded on a Shimadzu IR Prestige-21 (Shimadzu) spectrometer in the range of 400–4000 cm⁻¹ with a spectral resolution of 4 cm⁻¹ using KBr pellets. UV–Vis absorption spectra were obtained using a Shimadzu UV2450PC (Shimadzu) spectrophotometer. All the experiments were conducted at room temperature (25 °C) for both the powder and film samples. Solid-state measurements utilized an ISR2200 integrating sphere. Barium sulfate (BaSO₄) (Wako Pure Chemical Industries, Ltd., Tokyo, Japan) was used for the experiments at room temperature. Fluorescence emission spectra were obtained using a Shimadzu RF5301PC (Shimadzu) spectrofluorometer. All the experiments were conducted at room temperature (25 °C) for both the powder and film samples. The wavelength of the maximum UV–Vis absorption band was used as the excitation wavelength for fluorescence emission measurements. For solid-state samples, support was used where the angle of incidence of radiation was 45° relative to the detector. The solid-state relative fluorescence quantum yields were determined by employing sodium salicylate as the standard (QY_SS = 0.55) [55,56]. This involved a comparison of the integrated area of the emission spectrum of the benzazole derivatives (A_BZ) with that of sodium salicylate (A_SS). The percentage reflectance of the benzazole (R_BZ) and the reference sodium salicylate (R_SS) was measured relative to a plate coated with barium sulfate at the wavelength of maximum excitation. The fluorescence quantum yield of the benzazoles (QY_BZ) was calculated using Equation (1), where Φ_FL ss represents the fluorescence quantum yield of the reference material. In this experiment, the samples were treated as powders.

$${\mathrm{QY}}_{\mathrm{SS}}={\mathrm{QY}}_{\mathrm{BZ}}\frac{{\mathrm{A}}_{\mathrm{BZ}}}{{\mathrm{A}}_{\mathrm{SS}}}\frac{100-{\mathrm{R}}_{\mathrm{SS}}}{100-{\mathrm{R}}_{\mathrm{BZ}}}$$

(1)

2.2. Synthesis

The studied 2-(2′-hydroxyphenyl)benzazole derivatives 4a–4b were prepared as described in the literature [57,58]. In general, the respective amino precursors 2-(5′-amino-2′-hydroxyphenyl)benzazoles 3a–3b were prepared from the condensation of 5-aminosalicylic acid (1) with o-substituted anilines 2a or 2b in polyphosphoric acid, resulting in the formation of the corresponding aminobenzazoles 3a–3b [59]. For purification, chromatography in a column was performed on silica gel (230–400 mesh) using dichloromethane as the eluent. To obtain isothiocyanate derivatives 4a–4b (Scheme 1), a solution of 3a (or 3b) in dry acetone was slowly added dropwise to a solution of thiophosgene (Cl₂CS) in dry acetone (1:5) at 0 °C. The reaction mixture was stirred for two hours at room temperature, leading to the formation of the desired products. The resulting precipitate was filtered, washed with cold acetone, and subsequently dried at 60 °C. Purification was accomplished through chromatography in a column on silica gel (230–400 mesh) using dichloromethane as the eluent. The spectroscopic characterization aligns with the literature and conforms to the anticipated chemical structures.

2-(5′-Isothiocyanate-2′-hydroxyphenyl)benzoxazole (4a). Yield: 90%. M.p.: 169–171 °C. ¹H NMR (400 MHz, CDCl₃) δ (ppm): 11.62 (s, 1H), 7.08 (dd, 1H, J = 8.8 Hz), 7.88 (d, 1H, J = 2,4 Hz), 7.74 (m, 1H), 7.62 (m, 1H), 7,42 (m, 2H), 7.42 (m, 2H), 7.28 (dd, 1H, J = 9.2, 2.8 Hz). FTIR (wavenumber, cm⁻¹): 3062, 2100, 1629, 1488, 1231.

2-(5′-Isothiocyanate-2′-hydroxyphenyl)benzothiazole (4b). Yield: 92%. M.p.: 152–157 °C. ¹H NMR (400 MHz, CDCl₃) δ (ppm): 12.74 (s, 1H), 7.99 (m, 1H), 7.92 (m, 1H), 7.53 (m, 2H), 7.45 (m, 1H), 7.24 (dd, 1H, J = 8.8, 2.4 Hz), 7.06 (d, 1H, J = 8.8 Hz). FTIR (wavenumber, cm⁻¹): 3022, 2101, 1585, 1483, 1263.

2.3. Preparation of Fluorescent Starch

The general procedure used to prepare the fluorescent starch is presented in Scheme 2. An amount of 2 g of pure starch was dissolved in 50 mL of DMSO at 65 °C for 30 min. Subsequently, different amounts of fluorophore 4a or 4b were introduced into the reaction system, and the mixture was stirred at 65 °C for an additional 4 h. During this phase, we conducted reactions using different amounts of fluorophores to starch (w/w), as outlined in

Table 1. Reaction conditions for the preparation of modified starches.

Fluorophore	Starch (g)	Fluorophore Amount (w%) 1	Nomenclature
4a	2.0	0.1	Starch@4a01
		0.5	Starch@4a05
		1.0	Starch@4a1
		5.0	Starch@4a5
4b		0.1	Starch@4b01
		0.5	Starch@4b05
		1.0	Starch@4b1
		5.0	Starch@4b5

¹ Derived from the initial amount of starch employed in the calculation.

. The aim was to evaluate the properties of the resulting materials concerning fluorophore concentration. After this period, the solution was allowed to cool to room temperature (25 °C). Following the completion of the reaction, 150 mL of methanol was added to this solution. The resulting precipitate was then filtered and washed with methanol (3 × 20 mL) and acetone (3 × 20 mL) to remove the unbound fluorophore. The resultant precipitate was air-dried for 24 h at room temperature (25 °C) and for an additional 24 h at 65 °C under vacuum. Finally, the dried precipitate was comminuted in a mortar for characterization purposes.

2.4. Fluorescent Films Based on Starch

The casting technique was employed to prepare fluorescent starch films with glycerol as the plasticizer in different proportions (

Table 2. Reaction conditions for the preparation of fluorescent films based on modified starch.

Component	Amount (w%) 1
	Film Starch	Film Starch@4a	Film Starch@4b
Starch	70	-	-
Glycerol	30	30	30
Starch@4a5	-	70	-
Starch@4b5	-	-	70

¹ Derived from the initial amount of starch employed in the calculation.

). Natural corn starch (used as a blank sample) or fluorescent starch (Starch@4a5 or Starch@4b5) was dissolved in deionized water at 75 °C under mechanical stirring for 1 h. Glycerol was then introduced into the starch solution, and the mixture was stirred for an additional 15 min. Notably, at this stage, starch films were also produced through simple mixing, wherein fluorophores, without covalent bonding to starch, were added during the dissolution of starch. This step served as a control for evaluating the resulting fluorescent materials. The resulting solutions or mixtures were poured into polypropylene Petri dishes and dried in an oven at 65 °C for 24 h.

3. Results and Discussion

3.1. Characterization

The characterization of photoactive-doped materials heavily relies on the concentration of fluorophores, and various techniques can be employed to clarify binding, each operating at different sensitivities. However, the literature notes the intricacy of this characterization when a low content of fluorophores is utilized [60,61], which is precisely the situation with the modified starches prepared in this study. The morphology of the modified starches was first investigated using X-ray diffraction (XRD) (Figure 2). Substituting hydroxyl groups in starch can cause alterations in its crystalline structure. For example, when acetylation occurs on starch, a new X-ray diffraction pattern is generated, providing a means to evaluate the effectiveness of these reactions [62,63,64]. The diffraction pattern of natural starch displays peaks at approximately 15°, 16°, 17°, and 23° (2ϴ), as expected for corn starch [64,65]. X-ray diffraction analyses thus indicate a departure from the usual crystalline structure observed in natural starch when examining photoactive starches. The literature highlights that the processing of starch can induce significant structural alterations. A widely employed processing step is gelatinization, which involves the heating of starch in excess water within the temperature range of 50 to 70 °C. Throughout this procedure, a cooperative transition occurs wherein water infiltrates the amorphous regions of the granules, resulting in their swelling and destabilization of the crystalline regions. Consequently, this induces rapid granule swelling, crystallite melting, and complete loss of ordered structure [66,67,68]. Because starch modification occurs at 65 °C in the presence of undried DMSO, the loss of crystallinity could be related to the gelatinization process. Furthermore, due to the low water content, the efficiency of the process was not absolute, allowing for the discernment of some degree of ordered structure. In addition, despite the limited presence of fluorophores in starch reactions, which leads to minimal substitution of starch hydroxyl groups, the influence of fluorophores on the loss of crystallinity observed in modified starches compared to pure starch cannot be ruled out. With respect to the substitution of starch hydroxyl groups, each repeating unit of α-D-glucopyranose in starch has three hydroxyl groups available for reactions. The literature suggests a maximum possible degree of substitution of 3.0 [69]. For instance, the calculated degree of substitution for the highest concentration (5%) of 2-(5′-isothiocyanato-2′-hydroxyphenyl)benzoxazole (4b), assuming 100% reaction with starch, would be 0.06. Consequently, only approximately 2% of the hydroxyl groups in the α-D-glucopyranose units would undergo substitution.

According to the thermogravimetric analyses, presented in Figure 2 (bottom), the modified starch exhibited behavior reminiscent of that of natural starch, revealing two distinct stages below 500 °C. Note that an additional sample was prepared for comparison, the so-called StarchP. This sample was processed in the same manner as the modified starches but without the addition of a fluorophore. The initial stage, marked by a slight mass loss just above 100 °C, is linked to water loss, while the second stage, occurring at approximately 300 °C with substantial mass loss, corresponds to polymer decomposition. Both the starch and modified starch samples undergo a mass loss of approximately 15% up to nearly 100 °C, attributed to the absorption of water due to the hygroscopic nature of starch [64,70]. On the other hand, both fluorophores, 4a and 4b, demonstrate a single stage in their TGAs, indicating thermal degradation at approximately 300 °C. Finally, the infrared spectra presented in the Supplementary Material (Figures S4–S6) reveal strong absorption at approximately 2100 cm⁻¹ for both studied fluorophores related to the N=C=S bond vibrations. In addition, the localized absorptions at 1491 cm⁻¹ and 1572 cm⁻¹ (4b), as well as at 1507 cm⁻¹ and 1586 cm⁻¹ (4a), correspond to aromatic C=C bond vibrations. A specific absorption at 1445 cm⁻¹ (C=C aromatic stretching) was observed, along with absorptions at 1653 cm⁻¹ (carbonyl C=O) and 3500 cm⁻¹ (medium intensity, stretching of the primary amine N-H bond). In natural starch, localized absorptions at 1159, 1082, and 1014 cm⁻¹ are related to C-O stretching, while other bands at 992, 929, 861, 765, and 575 cm⁻¹ are due to stretching vibrations of the anhydroglucose unit [62,63]. Similar results were observed for processed starch (StarchP). The broad band at 3425 cm⁻¹ is attributed to hydroxyl groups, and at 1641 cm⁻¹, strongly bound water is observed [62,63]. Notably, no infrared absorption related to the fluorophores was detected in the modified starches. The anticipated band at approximately 1550 cm⁻¹ (NH(C=S) group) is absent, which should result from the formation of a thiocarbamate group during the reaction between –OH and isothiocyanate. Additionally, the characteristic band of N=C=S bond vibrations in unreacted fluorophores was not observed, possibly due to the low fluorophore content used in the starch reactions. Detecting very small quantities or dilutions in the polymer chain is challenging with FTIR [71].

3.2. Photophysics

The obtained modified starches were characterized in the solid state using UV–Vis absorption and fluorescence emission spectroscopies. The relevant data are summarized in

Table 3. Photophysical data in the solid state for modified starches using fluorophores 4a and 4b, where λ_abs and λ_em are the absorption and emission maxima, respectively, Δλ_ST is the Stokes shift, and QY is the fluorescence quantum yield.

Sample	λabs	λem	ΔλST (nm/cm−1)	QY (%)
Starch@4a01	333	514	181/10575	1.1
Starch@4a05	344	512	168/9539	3.4
Starch@4a1	350	511	161/9002	4.5
Starch@4a5	366	512	146/7791	7.6
Starch@4b01	334	501	167/9980	1.8
Starch@4b05	345	503	158/9105	3.3
Starch@4b1	366	503	137/7442	4.2
Starch@4b5	362	540	178/9106	1.1

. All the experiments were performed under the same conditions, using 50 mg of each sample, which was treated as powder. Each modified starch sample exhibited a distinct absorption in the visible region, which intensified at higher energies, below 400 nm (Figure 3a,b). Furthermore, there is an increase in absorption at approximately 350 nm as the concentration of fluorophores increases during starch preparation, which could be related to the absorption of the fluorophore. It is also noteworthy that with an elevated fluorophore concentration, the absorption maxima shifts toward the corresponding maxima observed for pure fluorophores (Figure 3c). Finally, it was observed that the pure starch in the solid state (blank sample) presented absorption below 300 nm, as already reported in the literature [72,73].

In Figure 4a,b, the emission curves of the modified starches are depicted, clearly indicating that all the samples exhibited photoactivity within the 450–650 nm range. Unlike what was noted in absorption, the emission spectra show notable distinctions when the fluorophores are bound to starch, suggesting that the matrix influences the photophysical properties of these compounds. The emission peaks for modified starches prepared with fluorophore 4a are situated at approximately 510 nm (Figure 4a), matching the region observed for the pure fluorophore (Figure 4c). Specifically, the resemblance between the spectra of the pure fluorophore and modified starch samples suggested that the photophysical behavior of 4a was similar, regardless of whether it was covalently linked to the polymer chain of starch. In addition, in the modified starches prepared with fluorophore 4a, there was an increase in the emission intensity of the photoactive starch as the quantity of fluorophore used for the preparation of the materials increased, in accordance with expectations. However, modified starches prepared using 4b as a fluorophore exhibited different photophysical behaviors. Initially, after binding with starch, the emission maxima of fluorophore 4b shifted toward blue (500 nm) compared to that of the pure fluorophore (530 nm). This behavior has already been observed when benzazoles reactive to the ESPT mechanism are covalently attached or doped into organic or inorganic matrices, highlighting the sensitivity of these fluorophores to the environment [74,75]. Additionally, as observed for starch doped with fluorophore 4a, samples prepared with 4b showed that the emission intensity was dependent on the dye concentration in the starch (from Starch@4b01 to Starch@4b1). Additionally, the modified starch prepared with more fluorophore (Starch@4b5) exhibited an emission maximum located at 540 nm, which was less intense than that at lower concentrations. In this case, unlike what was noted in absorption, the emission spectra show notable distinctions when the fluorophores are bound to starch, suggesting an additional nonradiative energy loss in the excited state or even interaction with additional ESIPT dyes in the matrix. The pure starch did not exhibit fluorescence emission, as expected. Furthermore, regarding the Stokes shift, the calculated values are consistent with what is expected for the keto* emission arising from the ESIPT process, indicating the absence of dual emission or the favoring of conformers unresponsive to ESIPT. The low values for the fluorescence quantum yield were expected due to the low dye content and agree with similar structures presenting ESIPT, as reported in the literature [76]. Finally, we would like to emphasize that the correlation between the quantum yields in the solid state, emission curve profiles, and the location of their respective maxima suggests that the emitting species may be different. Although it is not possible to observe any red-shifted absorption band, corresponding to ionized species already present in the ground state, the literature reports that compounds that are reactive to ESIPT can ionize in the excited state, generating species with emissions distinct from those observed in their neutral analogs [76,77]. Although the Stokes shift values indicate the presence of keto emission, the possibility of ionization in the excited state should not be dismissed, since the remaining OH groups in the starch could serve as proton acceptors [77]. This property, stemming from these compounds’ increased photoacidity, may contribute in some way to the distinct emissions observed.

As a proof of concept, films were prepared with modified starches, as shown in Figure 5. As noted in this study, the fluorophores exhibiting fluorescent properties were covalently bonded to the polymeric structure of starch through reactions with its hydroxyl groups. The quantity of compound used in modifying starch does not compromise its original properties, which allowed us to prepare these photoactive films. In this investigation, a control film (Figure 5a) prepared with nondoped starch was prepared for comparison (Table 2, Film Starch). The good transparency of the films prepared with pure starch and modified starches can be observed in Figure 5a–c. The films prepared with photoactive starches exhibited emission in the cyan-green to green regions, as expected from the emission results of the modified starches in the solid state.

4. Conclusions

In this study, we successfully synthesized novel starch-based materials with solid-state fluorescence under mild reaction conditions. Thermal analyses revealed that the thermal behavior of natural starch remained largely unchanged after reactions with fluorophores. Although the X-ray diffraction patterns of the modified starches differed from those of natural starch, a consistent pattern was not observed. The presence of fluorophores in the starch polymeric matrix was confirmed through photophysical characterizations. The similarity between the spectra of pure fluorophores and modified starches, which feature a significant Stokes shift, suggests that the photophysical behavior of fluorophores is consistent, whether covalently linked to the polymeric chain of starch or not. Moreover, the modified starches developed in this study facilitated the creation of photoactive thermoplastic films.